Metabonomics and Gut Microbiota in Nutrition and Disease: Sunil Kochhar François-Pierre Martin Editors

Molecular and Integrative Toxicology
Sunil Kochhar
François-Pierre Martin Editors
Metabonomics and
Gut Microbiota in
Nutrition and
Disease
Molecular and Integrative Toxicology
Series editor
Rodney R. Dietert
More information about this series at http://www.springer.com/series/8792
Sunil Kochhar • François-Pierre Martin
Editors
Metabonomics and Gut

Microbiota in Nutrition
and Disease
Editors
Sunil Kochhar François-Pierre Martin
Analytical Sciences Competence Pillar Molecular Biomarkers
Nestec SA, Nestlé Research Center Nestlé Institute of Health Sciences
Lausanne Lausanne
Switzerland Switzerland
ISSN 2168-4219 ISSN 2168-4235 (electronic)

ISBN 978-1-4471-6538-5 ISBN 978-1-4471-6539-2 (eBook)
DOI 10.1007/978-1-4471-6539-2
Springer London Heidelberg New York Dordrecht
Library of Congress Control Number: 2014953791
© Springer-Verlag London 2015

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection
with reviews or scholarly analysis or material supplied specifically for the purpose of being entered and
executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this
publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s
location, in its current version, and permission for use must always be obtained from Springer.
Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations
are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.
Printed on acid-free paper
Humana Press is a brand of Springer

Springer is part of Springer Science+Business Media (www.springer.com)
Preface
Being the most densely populated microbial ecosystem on earth, gut microbiota
coevolved as a key component of human biology and physiology. Human beings
share an intimate and mutualistic relationship with their gut microbiota, and recent
findings have demonstrated that the role of our gut microbial symbionts into health
and disease has been so far underestimated. Increasingly, scientific evidence identi-
fies gut microbiota as a key biological interface between human genetics and envi-
ronmental conditions encompassing nutrition. Microbiota dysbiosis or variation in
metabolic activity has been associated to metabolic deregulation (e.g., obesity,
inflammatory bowel disease), disease risk factor (e.g., coronary heart disease), or
even in the etiology of various pathologies (e.g., autism), although causal role into
impaired metabolism still needs to be established.
The rise in multifactorial disorders, the lack of understanding of the molecular
processes at play, and the needs for disease prediction in asymptomatic conditions
are some of the many questions that systems biology approaches are well suited to
address. Achieving this goal lies in our ability to model and understand the complex
web of interactions between genetics, metabolism, environmental factors, and gut
microbiota. The adaptation of systems biology to translational and clinical sciences
has been termed network medicine and is changing the way we think about prevent-
ing, predicting, diagnosing, and treating complex human diseases. Through consoli-
dating knowledge across intermediate organizational levels of life such as the
epigenome, transcriptome, proteome, metabolome or microbiome, and its integra-
tion with gene-disease traits, systems biology approaches are becoming highly rel-
evant for assessing the clinical characteristics of human health and disease.
The advent of metabonomics as a powerful systems biology approach opens new
opportunities to deepen and model the complex web of molecular interactions
between nutrition and health encompassing the understanding on how to modulate
gut microbiota. While the study of the gut microbiota required the development of
modern molecular biology techniques (many of the organisms present cannot be
cultured and are only known as a result of their detection via ribosomal 16S DNA),
we now know that these microbial populations are highly complex with many hun-
dreds of different species cohabiting in the gut and forming a complex ecology.
v
vi Preface
Its composition and activity coevolve with the host from birth and is subject to a
complex interplay that depends on the host genome, nutrition, and lifestyle.
Undoubtedly, the interplay between gut microbiome and host and its modulation
by nutrition will benefit from the integration of information on a systems biology-
wide approach. Integration of gene sequence of the microbiome and metabolomics
is currently envisioned to pave the way towards a better molecular understanding of
the complex mammalian superorganism, which is a prerequisite for optimizing
therapeutic strategies to manipulate the gut microbiota to combat disease and
improve health.
The field of endeavor is expanding rapidly, but we believe that now is a good
time to review the achievements in the area of metabonomics and gut microbiota
research. Therefore, this book will aim at providing state-of-the-art information on
the key findings and methodologies and point future directions towards understand-
ing how to beneficially modulate our gut functional ecology for health and nutritional
benefits.
This book provides a comprehensive overview of metabonomics and gut micro-
biota research from molecular analysis to population-based global health consider-
ations. The topics include the discussion of the applications in relation to
metabonomics and gut microbiota in nutritional research and in health and disease
and a review of future therapeutical, nutraceutical, and clinical applications. It also
examines the translatability of systems biology approaches into applied clinical
research and to patient health and nutrition.
The book begins with a general introduction on major concepts and research
directions with emphasis on core indicators of health and functions in infant, adult,
and elderly populations (Chap. 1). The second and third chapters will provide some
background information on the metabonomics technology and its implementation
in clinical research and data modeling.
Chapters 4, 5, 6, 7, 8, 9, and 10 describe extensively some key applications in
nutritional research. These sections provide examples and expectations in the field
of personalized medicine and nutrition, including the study of infant-nutrition and
healthy aging paradigms and studies of microbial and human metabolism of
macronutrients (protein, fat, carbohydrates, and fiber) and specific non-nutrient
food components (polyphenols, flavonoids) and herbal medicine.
Chapters 11, 12, 13, and 14 describe the applications and perspectives of com-
bining metabonomics and gut microbiota approaches in health and disease research.
Chapter 11 covers the introduction of the field of microbio-immuno-metabolism,
based on the role of the environment, genetic background, and individual diversity
in relation to the onset and development of metabolic diseases. Chapter 12 gives
emphasis on the role of microbiota in modulating lipid metabolism and related dis-
ease risks, while Chap. 13 comprehensively describes the intimate relationship
between gut microbiota metabolism and gastrointestinal disorders with emphasis on
inflammatory bowel diseases. Chapter 14 extends the discussion to the field of the
gut brain axis and its role in the etiology of several disorders with emphasis on
autism.
Preface vii
Chapters 15 and 16 review the applications towards future therapeutical,

nutraceutical, and clinical applications, with emphasis on two main examples, bar-
iatric surgery and drug efficacy and toxicity.
Metabonomics and gut microbiota in nutrition and disease serves as a hand-
book for postgraduate students, researchers in life sciences or health sciences, and
scientists in academic and industrial environments working in application areas as
diverse as health, disease, nutrition, microbial research, and human clinical
medicine.
We would like to acknowledge the chapter authors for their efforts in creating
this book. Even though they all have many demands on their time, they generously
contributed towards this effort to publicize Metabonomics and gut microbiota
research efforts towards deciphering human health and disease.
Lausanne, Switzerland Sunil Kochhar

François-Pierre Martin
Contents
1 Introduction to Metabonomics in Systems Biology Research ............. 1

François-Pierre Martin and Sunil Kochhar
2 Metabonomics in Clinical Practice ........................................................ 25
Sebastiano Collino, François-Pierre Martin, and Sofia Moco
3 Adopting Multivariate Nonparametric Tools to Determine
Genotype-Phenotype Interactions in Health and Disease ................... 45
Ivan Montoliu
4 Metabonomics in Translational Research for Personalized
Medicine and Nutrition .......................................................................... 63
Guoxiang Xie and Wei Jia
5 Can We Use Metabolomics to Understand Changes
to Gut Microbiota Populations and Function?
A Nutritional Perspective ....................................................................... 83
Sofia Moco and Alastair B. Ross
6 Automated Annotation of Microbial and Human
Flavonoid-Derived Metabolites .............................................................. 109
Velitchka V. Mihaleva, Fatma Yelda Ünlü, Jacques Vervoort,
and Lars Ridder
7 Metabonomics in Neonatal and Paediatric Research:
Studying and Modulating Gut Functional Ecology
for Optimal Growth and Development ................................................. 125
Vassilios Fanos and Laura Cuzzolin
8 Metabolomics and Milk: The Development
of the Microbiota in Breastfed Infants .................................................. 147
J. Bruce German, Jennifer T. Smilowitz, Carlito B. Lebrilla,
David A. Mills, and Samara L. Freeman
ix
x Contents
9 Metabonomics and Gut Microbial Paradigm

in Healthy Aging ...................................................................................... 169
Elena Biagi, Marco Candela, François-Pierre Martin,
Sebastiano Collino, Claudio Franceschi, and Patrizia Brigidi
10 Roles of Herbal Medicine in Modulating Gut
Microbiota Associated with Health and Diseases ................................. 185
Yulan Wang and Huiru Tang
11 Gut Microbiota and Metabolic Diseases:
From Pathogenesis to Therapeutic Perspective.................................... 199
Rémy Burcelin, Michael Courtney, and Jacques Amar
12 Role of Microbiota in Regulating Host Lipid
Metabolism and Disease Risk ................................................................ 235
Isabel Bondia-Pons, Tuulia Hyötyläinen, and Matej Orešič
13 Role of the Gut Microbiota in Maintaining GI Health:
Highlights on Inflammatory Bowel Disease ......................................... 261
Lisa Gruber and Dirk Haller
14 Deciphering the Gut Microbial Contribution
to the Etiology of Autism Development ................................................. 311
Ivan K.S. Yap and François-Pierre Martin
15 The Modulation of Drug Efficacy and Toxicity
by the Gut Microbiome........................................................................... 323
Ian D. Wilson and Jeremy K. Nicholson
16 Understanding the Benefits of Bariatric Surgery
on Gut Physiology: Implications for Obesity,
Type 2 Diabetes, and Cardiovascular Disease ...................................... 343
Steven K. Malin and John P. Kirwan
Index ................................................................................................................. 371

Contributors
Jacques Amar Institut National de la Santé et de la Recherche Médicale

(INSERM), Toulouse, France
Université Paul Sabatier (UPS), Unité Mixte de Recherche (UMR) 1048, Institut de
Maladies Métaboliques et Cardiovasculaires (I2MC), Toulouse Cedex 4, France
Département de thérapeutique de l’hôpital Rangueil, Toulouse, France
Elena Biagi Department of Pharmacy and Biotechnology, University of Bologna,
Bologna, Italy
Isabel Bondia-Pons Steno Diabetes Center, Gentofte, Denmark
Patrizia Brigidi Department of Pharmacy and Biotechnology, University of
Bologna, Bologna, Italy
Rémy Burcelin Institut National de la Santé et de la Recherche Médicale
(INSERM), Toulouse, France
Université Paul Sabatier (UPS), Unité Mixte de Recherche (UMR) 1048, Institut de
Maladies Métaboliques et Cardiovasculaires (I2MC), Toulouse Cedex 4, France
Marco Candela Department of Pharmacy and Biotechnology, University of
Bologna, Bologna, Italy
Sebastiano Collino Molecular Biomarkers, Nestlé Institute of Health Sciences,
Lausanne, Switzerland
Michael Courtney Vaiomer S.A.S, Labège, France
Laura Cuzzolin Department of Public Health & Community Medicine, Section
Pharmacology, University of Verona-Italy, Verona, Italy
Vassilios Fanos Neonatal Intensive Care Unit, Puericulture Institute and Neonatal
Section, University of Cagliari-Italy, Cagliari, Italy
xi
xii Contributors
Claudio Franceschi Department of Experimental Diagnostic and Specialty

Medicine and CIG – Interdipartimental Center L. Galvani, University of Bologna,
Bologna, Italy
Samara L. Freeman Department of Food Science and Technology, Foods for
Health Institute, University of California Davis, Davis, CA, USA
J. Bruce German Department of Food Science and Technology, Foods for Health
Institute, University of California Davis, Davis, CA, USA
Lisa Gruber Nutrition and Immunology, Freising, Germany
Dirk Haller Nutrition and Immunology, Freising, Germany
ZIEL – Research Center for Nutrition and Food Sciences, Biofunctionality Unit,
Technische Universität München, Freising, Germany
Tuulia Hyötyläinen Steno Diabetes Center, Gentofte, Denmark
Wei Jia University of Hawaii Cancer Center, Honolulu, HI, USA
John P. Kirwan Department of Pathobiology, Lerner Research Institute, Cleveland
Clinic Foundation, Cleveland, OH, USA
Department of Nutrition, School of Medicine, Case Western Reserve University,
Cleveland, OH, USA
Metabolic Translational Research Center, Endocrine and Metabolism Institute,
Cleveland Clinic, Cleveland, OH, USA
Sunil Kochhar Analytical Sciences Competence Pillar, Nestec SA, Nestlé
Research Center, Lausanne 26, Switzerland
Carlito B. Lebrilla Department of Chemistry, Foods for Health Institute,
University of California Davis, Davis, CA, USA
Steven K. Malin Department of Pathobiology, Lerner Research Institute, Cleveland
Clinic Foundation, Cleveland, OH, USA
François-Pierre Martin Molecular Biomarkers, Nestlé Institute of Health
Sciences, Lausanne, Switzerland
Velitchka V. Mihaleva Laboratory of Biochemistry, Wageningen University and
Research Center, Wageningen, The Netherlands
Netherlands Metabolomics Centre, Leiden, The Netherlands
David A. Mills Department of Viticulture & Enology, Department of Food Science
and Technology, Foods for Health Institute, University of California Davis, Davis,
CA, USA
Sofia Moco Natural Bioactives and Screening, Nestlé Institute of Health Sciences,
Lausanne, Switzerland
Contributors xiii
Ivan Montoliu Analytical Sciences, Applied Mathematics, Nestec Ltd., Nestlé

Research Centre, Lausanne 26, Switzerland
Jeremy K. Nicholson Department of Surgery and Cancer, Faculty of Medicine,
Imperial College, London, UK
Matej Orešič Steno Diabetes Center, Gentofte, Denmark
Lars Ridder Laboratory of Biochemistry, Wageningen University and Research
Center, Wageningen, The Netherlands
Netherlands eScience Center, Amsterdam, The Netherlands
Alastair B. Ross Food and Nutrition Science, Department of Biology and Biological
Engineering, Chalmers University of Technology, Gothenburg, Sweden
Jennifer T. Smilowitz Department of Food Science and Technology, Foods for
Health Institute, University of California Davis, Davis, CA, USA
Huiru Tang Key Laboratory of Magnetic Resonance in Biological Systems, State
Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan
Centre for Magnetic Resonance, Wuhan Institute of Physics and Mathematics,
Chinese Academy of Sciences, Wuhan, People’s Republic of China
State Key Laboratory of Genetic Engineering, Biospectroscopy and Metabolomics,
School of Life Sciences, Fudan University, Shanghai, People’s Republic of China
Fatma Yelda Ünlü Laboratory of Biochemistry, Wageningen University and
Jacques Vervoort Laboratory of Biochemistry, Wageningen University and
Netherlands Metabolomics Centre, Leiden, The Netherlands
Yulan Wang Key Laboratory of Magnetic Resonance in Biological Systems, State
Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan
Centre for Magnetic Resonance, Wuhan Institute of Physics and Mathematics,
Chinese Academy of Sciences, Wuhan, People’s Republic of China
Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases,
Hangzhou, People’s Republic of China
Ian D. Wilson Department of Surgery and Cancer, Faculty of Medicine, Imperial
College, London, UK
Guoxiang Xie University of Hawaii Cancer Center, Honolulu, HI, USA
Ivan K.S. Yap Department of Life Sciences, International Medical University,
Kuala Lumpur, Malaysia
Editor Biographies
François-Pierre Martin, Ph.D., Eng., is Senior Scientist and Research Program

Manager in the Department of Molecular Biomarkers, at the Nestlé Institute of
Health Sciences, Lausanne, Switzerland. He received his Ph.D. in Biochemistry
and Metabonomics from Imperial College London, UK, in 2006, and has a back-
ground in biotechnology and microbiology from the engineering school Polytech-
Marseille, France. From 2006 to 2012, François-Pierre has developed and expanded
metabonomics applications to nutrition and gut microbiota research for health and
disease management, at the Nestlé Research Center, Switzerland. Emphasis was
given to investigate transgenomic interactions between the mammalian and the gut
microbial metabolic space using systems biology. Dr. Martin has developed a
strong expertise in investigating gut microbial effects on diverse mammalian
processes and their impact upon the etiology of several conditions (gastrointestinal
health, diabetes, obesity). In July 2012, Dr. Martin joined the Nestlé Institute of
Health Sciences, Switzerland, to further develop systems biology approaches in
the frame of health, disease, and nutrition research. His main research interests
include gastrointestinal health (e.g., inflammatory bowel disease, irritable bowel
syndrome), metabolic health (metabolic syndrome and diabetes), and the study of the
underlying mechanisms of foods and bioactives in pediatric and adult populations.
This research helps identify new targets for nutraceutical and/or therapeutical
applications. Dr. Martin has authored over 60 research papers published in interna-
tionally peer-reviewed journals, 6 book chapters, and 20 patents.
Sunil Kochhar obtained his Ph.D. from the Indian Institute of Technology, New
Delhi (1986), and after postdoctorate at the University of Delaware, DE USA, and
University of Zurich, Zurich Switzerland, he joined Nestlé Research Center,
Lausanne, Switzerland. Following different lead scientist and line management
functions, he became Head of the BioAnalytical Sciences Department in 2002. He
has authored over 100 scientific articles in the internationally peer-reviewed journals,
20 book chapters, and 14 patents. He is invited lecturer in many international
conferences. His research interests are in enzyme and protein chemistry, functional
genomics, glycobiology, and taste receptors. In early 2000, he initiated the concept
xv
xvi Editor Biographies
of “nutritional metabonomics” to assess metabolic health and its early deviations in

healthy humans with a notion to provide appropriate nutritional solutions to main-
tain and/or prolong the healthier status. Metabonomics research leads to the concept
of “metabotype” that included full metabolite signature of an individual including
metabolites of the gut microbiota. Over the past 15 years, he has intensively studied
the influence of dietary intake and mammalian gut microbiota on the overall health
status in healthy individuals. Currently, Dr. Kochhar is the head of Analytical
Sciences department and holds an honorary professorship at Imperial College,
London UK.
Chapter 1
Introduction to Metabonomics in Systems
Biology Research
François-Pierre Martin and Sunil Kochhar
Abstract There is a growing interest to understand the paradigm of healthy aging

since the aging population and longevity increases worldwide. However, chronic
diseases cover a broad range of organ and biological functions, ranging from
metabolic to digestive health, as well as from mental to physical functions. In the
meantime, rising evidence also point towards the critical and long-term involvement
of prenatal and early nutrition on later health and disease risk predisposition. Dietary
preferences and nutrient composition have been shown to influence human and gut
microbial metabolism, which ultimately have specific effects on health and disease
risk. Increasingly, results from molecular biology and microbiology demonstrate
the key role of the gut microbiota metabolic interface to the overall mammalian
host’s nutritional and disease status.
There is therefore raising interest in nutrition and disease research to character-
ize the molecular foundations of the gut microbial mammalian cross talk at both
physiological and biochemical pathway levels. Tackling these challenges can be
achieved through systems biology approaches, such as metabonomics, to underpin
the highly complex metabolic exchanges between diverse biological compartments,
including organs, systemic biofluids, and microbial symbionts. The generalization
of such approaches has opened new research areas to deepen our current under-
standing on many physiological processes as well as food functionalities in general
and targeted populations. Such novel insights are envisioned for aiding strategies
for a tailored therapeutic and nutritional program. By the development of specific
biomarkers for prediction of health and disease, metabonomics is increasingly used
in clinical applications as regard to disease etiology, diagnostic stratification, and
potentially mechanism of action for therapeutical and nutraceutical solutions.
F.-P. Martin (*)

Molecular Biomarkers, Nestlé Institute of Health Sciences,
EPFL Innovation Park, bâtiment H, 1015 Lausanne, Switzerland
e-mail: [email protected]
S. Kochhar
Analytical Sciences Competence Pillar, Nestec SA, Nestlé Research Center,
Route du Jorat 57, 1000 Lausanne 26, Switzerland
© Springer-Verlag London 2015 1

S. Kochhar, F.-P. Martin (eds.), Metabonomics and Gut Microbiota
in Nutrition and Disease, Molecular and Integrative Toxicology,
DOI 10.1007/978-1-4471-6539-2_1
2 F.-P. Martin and S. Kochhar
This chapter focuses on introducing major concepts and research directions with
emphasis on core indicators of health and functions in infant, adult, and elderly
populations. The different themes will be developed further under the related book
sections.
Keywords Disease • Gut microbiota • Health • Mass spectrometry • Metabonomics

• Metagenomics • Nuclear magnetic resonance spectroscopy • Nutrition • Systems
biology
1.1 Introduction
There is a growing interest to understand the paradigm of healthy aging since the
aging population and longevity increases worldwide. Indeed, it has been estimated
that by 2020 chronic disease in developing countries will account for almost three-
quarters of all deaths worldwide with 75 % of death due to stroke and 70 % of death
due to diabetes [1, 2]. Awareness about the role of nutrition and lifestyle for health
and disease risk management has increased, with key emphasis on the prevention of
metabolic disorders, including cardiometabolic diseases and type 2 diabetes [1–3].
In parallel, rising evidence pointed towards the critical and long-term involvement
of early nutrition and lifestyle on later health and disease risk predisposition [4].
Thus, it becomes pertinent to look at metabolism throughout life, disease develop-
ment, and nutritional requirements to understand the onset of certain child and adult
physiological conditions [3, 4].
With the advent of the post-genomic era, nutrition research benefits from all
state-of-the-art analytical strategies that could be used for understanding the com-
plex relationship between nutrition and health [5]. In the meantime, modern nutri-
tion has shifted its focus towards molecular biology, genetics, and metabolic
pathways with the goal of preventing disease and enhancing the health and
well-being of individuals [6]. Since population and individual physiological fea-
tures are encoded at the different levels of biological compartments and organiza-
tion, there is growing interest in modeling gene expression, protein or metabolite
concentrations, and their dynamic pathways in cells, tissues, and organs to generate
biological system models. Recent revolution in omics technologies has generated
various promising concepts aiming to generate a global systems view of physiologi-
cal and pathological processes [6, 7]. The concept of systems biology has then been
developed and related to the integration of information at the different levels of
genomic expression (mRNA, protein, metabolite). Thus, systems biology generates
pathway information and provides the capacity to measure subtle perturbations of
metabolic pathways resulting from various intrinsic and extrinsic effects, including
disease or nutrition. Dietary preferences and nutrient composition have been shown
to influence human and gut microbial metabolism, which ultimately impact health
and disease risk. It is therefore envisioned that the integration of systems biology
1 Introduction to Metabonomics in Systems Biology Research 3
into nutritional and health research will provide novel perspectives [8, 9]. This has
led to the development of nutrigenomics that tackles how diet influences gene tran-
scription, protein expression, and metabolism [6, 7, 10].
Systems biology applications are rising and are envisioned to drive a change in
clinical practice through generalizing large population-based studies, aiming at
enhancing our understanding of the role of genetics, environmental factors, and
their interactions on individual susceptibility to disease and health [11, 12].
Therefore, one of the most powerful strategies for deciphering gene-environment
interactions and their role in individual variability, disease etiology, and nutritional
outcomes lies in the ability to combine data from different omics technologies,
establishing the way for systems biology approaches [13, 14]. In this, metabonom-
ics is able to generate multivariate information on a wide range of molecules and
provide the ability to measure subtle changes in biological processes as a result of
different nutritional effects [15–17]. Metabolic profiling of biological fluids by
nuclear magnetic resonance (NMR) spectroscopy or mass spectrometry (MS)
ensures a simultaneous analysis of a wide range of metabolites that are the end-
points of molecular regulatory processes, diet, and gut microbiota metabolism and
influenced by other environmental factors (Fig. 1.1). Metabonomics therefore
enables monitoring metabolite concentrations and dynamics in cells, tissues, and
multi-compartmental biological systems [18–21]. By opening a direct biochemical
window into the metabolome, metabonomics is uniquely suited to develop new gen-
erations of biomarkers that are capable of providing a better understanding of com-
plex metabolic phenomena as well as assessing intra- and interindividual differences.
This feature makes metabonomics very efficient for the generation of biomarker
patterns for the comprehensive characterization of metabolic health, the prognostics
and the diagnostics of diseases, and the generation of new insights in the under-
standing of the interactions of diet and metabolism and/or medical conditions.
Therefore, the identification of specific metabolic fingerprints vows strong poten-
tials for nutraceutical and therapeutical surveillance.
To date, numerous successful applications of metabonomics are reported in tox-
icity screening, drug metabolism, and functional genomics, a major part of them
involving animal models rather than human subjects due to the greater control of
conditions to reduce extrinsic variability [22–27]. Metabonomics’ introduction in
the field of nutrition research, i.e., nutritional metabonomics, has already delivered
interesting insights for the understanding of metabolic responses of human or ani-
mals in response to dietary interventions and for the definition of metabolic pheno-
types [5, 28–31]. Defining the metabolic phenotype or “metabotype” of human
populations will offer a great opportunity to evaluate the metabolic response and the
degree of this response to specific dietary modulations at the individual level, simi-
larly to the “pharmaco-metabonomic” concept [32].
One particular feature that such an approach enables is rediscovering and revisit-
ing gut functional ecology and its role in the etiology of many metabolic disorders.
Metabonomics has indeed offered an unprecedentedly new way to capture the com-
plex metabolic interactions of the host with its commensal microbial partners pro-
viding a new way to define individual and population phenotypes [18, 33, 34]. This
Socio-economics Lifestyle, environmental factors

Health status Medical, clinical conditions
Genetics Anthropometric data, gender, age
Host genome & metabolism
Gastro-intestinal
functional ecology
Gut microbiome
& metabolism
Metabolic <Personalized Healthcare and Nutrition>

health <--------------------- Diagnostic ------------>
<----- Prognosis (predisposition, risk factor) ---->
<----- Prediction (predicted response) --------->
Fig. 1.1 Conceptualization of systems biology approaches for gastrointestinal health and risk
management. The metabolic status of individuals results from gene, environment, lifestyle, food,
and gut microbiota interactions. The relationship between host and gut microbiota, with the meta-
bolic influence of gut symbionts at different level of biological organization, underpins the depth
of controls over multiple host cell metabolic functions
feature offers a new perspective to study the role of mammalian gut microbial
metabolic interactions in individual susceptibility to health and disease outcomes,
which is of increasing importance in many health research fields ranging from meta-
bolic to gastrointestinal health and even brain health [35–42]. The mammalian gut
microbial relationship is a key determinant by performing multiple digestive,
immune, and metabolic functions [43–45]. The gastrointestinal tract of adult
humans contains around 1.5 k of biomass composed by gut microbial symbiotic and
commensal organisms that are in intimate communication with the host, which is
the result of a long period of coadaptation between the host genotype and the gut
microbiome [46]. The gut microbiota provides to the host specific capacities rang-
ing from dietary energy recovery from nutrients, generating digestible carbohy-
drates and short-chain fatty acids (SCFAs) from otherwise nondigestible fibers,
amino acids, and vitamins to protect against infectious diseases [47–50]. Gut
microbial activities were shown to be extremely specific, as exemplified via its

essential role in the development and maintenance of the mucosal innate and adap-
tive immune system [51]. In parallel, there is increasing awareness that many health
disorders involve a significant perturbation of immune and energy metabolism
which are intrinsically linked to the gut functional ecology [45, 52, 53]. The consor-
tium of symbiotic gut microorganisms can be viewed as a metabolically adaptable,
rapidly renewable, and metabolically flexible ecosystem varying in addition with
the host’s age, diet, and health status. Consequently, the microbiome is a nutritional
target today and might also become the foundation of future drug targeting and
interventions [54].
Undoubtedly, the interplay between gut microbiome and host and its modulation
by nutrition will benefit from the integration of information on a system’s biology-
wide approach. Integration of gene sequence of the microbiome and metabolomics
is currently envisioned to pave the way towards a better molecular understanding of
the complex mammalian superorganism, which is a prerequisite for optimizing
therapeutic strategies to manipulate the gut microbiota to combat disease and
improve health. The field of endeavor is expanding rapidly, but we believe that now
is a good time to review the achievements in the area of metabonomics and gut
microbiota research. Therefore, this book will aim at providing state-of-the-art
information on the key findings and methodologies and point future directions
towards understanding how to beneficially modulate our gut functional ecology for
health and nutritional benefits.
1.2 Metabonomics-Based Systems Approaches:

Lessons Learnt and Highlights
Over the last three decades, the majority of metabonomic-based investigations in

human populations have applied metabonomic analysis of biofluids to screen and
diagnose certain diseases and monitor physiological changes caused by toxic insults
[23–25, 32, 55–58]. Metabonomics is being widely employed in the area of food
research to investigate the mutual link among the fields of metabonomic and nutri-
tion research [5, 31]. Systems biology approaches employing systemic and com-
partmental metabonomics are therefore applied to assess the molecular processes
reflecting metabolic adaptation to disease etiology, development, or management
with drugs. One key feature of such an approach lies in its ability to capture the
individual status prior to intervention which encapsulates both intrinsic and extrin-
sic determinants of the individual. Increasingly, results from molecular biology and
microbiology demonstrate the key role of the gut microbiota metabolic interface to
the overall mammalian host’s health status. There is therefore raising interest in
nutrition and disease research to characterize the molecular foundations of the gut
microbial mammalian cross talk at both physiological and biochemical pathway
levels. Tackling these challenges can be achieved through systems biology
approaches, such as metabonomics, to underpin the highly complex metabolic
exchanges between diverse biological compartments, including organs, systemic

biofluids, and microbial symbionts. To explore how the changes in environmental
conditions and lifestyle influence human physiology, a large-scale metabonomic
study was conducted to investigate metabolic phenotype variation across and within
four human population groups. Holmes et al. used metabonomics in the context of
a large-scale epidemiological study to identify metabolic signatures across and
within selected human populations in relation to geography, diet-related major risk
factors, and cardiovascular diseases [59]. In this study, urinary metabolite excretion
patterns differ between East Asian and Western populations, Japanese individuals
living in Japan or in the USA, and Chinese participants living in the northern and
southern parts of China. For example, urinary excretion of formate, a mammalian
gut microbial co-metabolite, was shown for the first time to be inversely correlated
with blood pressure.
Overall, the increasing rate of metabolic syndrome is a worldwide challenge for
both health and nutritional research [60, 61]. By 2015, the world health organization
(WHO) projections predict 2.3 billion overweight adults and more than 700 million
obesity cases [62]. This overweight and obesity pandemic continues to rise particu-
larly among children [63, 64]. This rapid pandemic, which cannot be explained on
the sole basis of genetic heritability and susceptibility, has raised awareness on the
critical role of changes in dietary habits and lifestyle in the metabolic etiology of the
diseases. It is nowadays well admitted that a main obesity determinant is the energy
imbalance between calorie intakes and expenditures which can be ascribed to a
global dietary shift towards high energy density foods and sedentary behaviors.
Existing therapeutic strategies to obesity lies in changing dietary habits and lifestyle
or pharmacotherapies, yet, with only marginally beneficial effects for morbidly
obese patients. Nowadays, much interest is therefore given for a tailor-made weight
management program and more recently to body composition rather than BMI-
driven approaches. Body fat distribution, and visceral fat in particular, was demon-
strated as a key determinant of increased risk of cardiovascular disease [65–67],
diabetes [68, 69], hypertension [70], nonalcoholic fatty liver disease [71], and a
higher risk of mortality [72]. Over the last decades, genetic and environmental pro-
moters for obesity-related metabolic disorders were investigated, including genes
and transcription factors associated with fat storage and obesity [73, 74], genetic
inheritability [75], or even gut microbiota influence [76]. Nevertheless, the gener-
ated insights face the challenges of explaining why similar obesogenic and diabeto-
genic conditions do not necessarily lead to a universal response to insulin and
adiposity-associated cardiometabolic risks [77, 78]. Individuals with normal weight
(body mass index, BMI <25) are increasingly shown to also express cardiometa-
bolic abnormalities [77]. Increasing evidence pinpoints the key role of region-
specific body composition and metabolism underpinning individual susceptibility
to metabolic disease risks. In this search for expanding our knowledge on the
etiology of metabolic disorders, systems biology models may lead to new working
hypotheses underlying one’s predisposition to develop specific disorders. For
instance, in an elegant series of studies integrating metabolic, endocrine, inflamma-
tory, and physiologic differences between obese and lean subjects, a branched-chain
amino acid (BCAA)-related metabolic signature was highlighted as a robust

metabolic readout of insulin resistance (IR) [79], even present under subclinical
conditions [80]. The catabolism of BCAAs was tightly intertwined with the IR lev-
els, while greater levels of BCAAs were detected in the obese and IR phenotype
[81–83]. Further studies confirmed the predictive values of these markers for the
development of diabetes in prospective human studies [84, 85]. Five branched-chain
and aromatic amino acids were indeed associated with IR, namely, isoleucine, leu-
cine, valine, tyrosine, and phenylalanine, and a combination of three amino acids
(isoleucine, phenylalanine, tyrosine) could predict future diabetes (>5-fold higher
risk for individuals in the top quartile) [85]. However, supplementation of BCAAs
in a diet-induced obesity rat model did not result in any improvement of the meta-
bolic status and IR and only led to reduced food intake and weight gain. Together
these key findings demonstrate a critical role of BCAA metabolism in the early
onset of IR and diabetes type 2 developments, but the understanding of the biologi-
cal specificities in certain organ still remains to be understood in order to better
decipher how the modulation of BCAA and amino acid metabolism could help dis-
ease management. Recent report on obese Japanese subjects further demonstrated a
physiological inference between IR; plasma levels of alanine, glycine, glutamate,
tryptophan, tyrosine, and BCAA; and visceral fat metabolism [86]. Complementary
findings further described a complex relationship between dyslipidemia and IR
development [84], while there is growing interest towards understanding the speci-
ficities of fatty liver and visceral adiposity in the development of metabolic disor-
ders. For instance, a complex remodeling of triglycerides [84] and phospholipids
[87] species were highlighted in relation to the distribution of visceral and subcuta-
neous fat within the body. These observations may result from a multifactorial ori-
gin, including dietary factors, gut functional ecology, intestinal absorption, as well
as platelet-activating factor metabolic pathways [88, 89], which are modulated by
obesity and insulin resistance [90, 91]. Another physiological peculiarity of the
metabolic syndrome encompasses nonalcoholic fatty liver disease, ranging from
steatosis to nonalcoholic steatohepatitis [92, 93]. For instance, using a parallel
animal model/human design, biochemical perturbations linked to liver dysfunction
through increased concentrations in bile acids and eicosanoids were highlighted.
Such findings corroborate metabolic steatosis markers in liver tissues, including
altered levels of bile acids, glutathione, and lipids. All together, these metabolic
findings provided further insight into specific metabolic contribution of liver
dysfunction in an overall metabolic syndrome context, as per specific cholesterol,
lipid, and oxidative stress readouts. More recently, metabonomics was successfully
employed to demonstrate a relationship between gut microbial metabolism of
dietary phosphatidylcholine and cardiovascular pathogenesis in humans [94].
Over the last two decades, gastrointestinal surgery has been revealed as a poten-
tial “holy grail” solution to treat diabetes in several obese populations, and its appli-
cations is envisioned to be increasingly applied worldwide [95]. Effective weight
loss was achieved in morbidly obese patients after undergoing bariatric surgery.
A substantial majority of patients with diabetes, hyperlipidemia, and hypertension
generally experience complete resolution or improvement [95]. However, major
gaps still obscure proper understanding on the mechanisms at play for improving
diabetes. In the meantime, clinicians and patients are also facing the challenge of
understanding why a significant percentage of the subjects do not respond post-
surgery and reacquire insulin-dependent disorders and how to tackle all the subse-
quent adverse events, including micronutrient deficiencies. Metabolic insights are
envisioned to help generate reference data that will help further the understanding
of the underlying biological deregulations, for instance, through better monitoring
of the metabolic and nutritional requirements of patients who underwent various
interventions (e.g., gastric banding, gastric bypass, gastroplasty, biliopancreatic
diversion, or duodenal switch). Recent findings in animal models reported how uri-
nary phenotyping may indicate weight loss-independent metabolic effects of Roux-
en-Y gastric bypass, which sets a first step towards promising novel insights for
human applications [96].
Furthermore, the gastrointestinal tract (GIT) is one of the most essential inter-
faces of mammalian organisms interacting with nutrients, exogenous compounds,
and gut microbiota, and its condition is influenced by the complex interplay between
these environmental factors and host genetic elements. Along the GIT, the gut
microbiota is a key determinant of the gut functional ecology and regulatory
processes involved in the absorption, digestion, metabolism, and excretion of dietary
nutrients as well as barrier integrity, motility, and mucosal immunity [35, 97]. The
evolution of nutrition, sanitary, and medical care conditions has led to rediscovering
host-gut microbial metabolic interactions in health and disease [98]. Gut microbial
activities can be extremely complex, such as in the etiology and development of
several chronic inflammatory disorders, including inflammatory bowel disease
(IBD) or irritable bowel syndrome (IBS) [99, 100]. Since normal aging is associated
with a number of significant changes in GIT function and with the development and
progression of chronic disease, such insights will be key for tailoring personal nutri-
tional intervention and monitoring patient response on an individual basis [101].
Advancing knowledge regarding the mechanisms of IBD has led, for instance, to the
development of different therapeutic solutions based on surgery [102], cannabi-
noids [103], or immunosuppression [104]. Although prognostic and monitoring
tools are currently lacking, the use of metabolic readouts in combination with state-
of-the-art clinical and medical readouts is to be a valuable tool to differentiate and
follow-up disease evolution and respond to disease-modifying interventions.
This is, for instance, exemplified through the discovery of metabolic indicators of
different IBD determinants, including mucosal healing, gut permeability, absorp-
tion, digestion, or inflammatory states. As an example, Winterkamp et al. reported
previously how N-methylhistamine, a key metabolite in mast cell metabolism
involved in the pathogenesis of IBD, could be used as an indicator of disease activ-
ity in patients [105]. In this study, the urinary excretion of N-methylhistamine was
associated with elevated histamine production and metabolism in IBD and could be
used as a reliable diagnostic tool to monitor clinical and endoscopic disease activity
in IBD. Additional animal studies may further substantiate proofs of concept on the
feasibility to identify blood-related metabolic indicators of early onsets of chronic
inflammatory development for patient monitoring [106]. Finally, noninvasive stool
analysis was shown to be very promising for monitoring the remodeling of the gut
functional ecology, concomitant to malabsorption and element of protein losing
enteropathy [107, 108]. Other noninvasive approaches to monitor gut permeability
functions may also be promising prospects for future translation to patient monitor-
ing [109].
It can be forecasted that such integrative systems biology approaches would help
to delineate different behavioral and response phenotypes, with which to personal-
ize the disease and nutritional management programs. For instance, beneficial
effects of exercise in patients with type 1 diabetes were assessed using such a meta-
bonomic approach [110], a concept that can be extended to a larger range of indi-
viduals under different age, nutritional, and health conditions. Nevertheless, the
translation of this unprecedented source of metabolic knowledge which is now at
hand into actionable clinical practices for healthcare requires addressing properly
the dynamics of physiological and biological processes when defining metabolic
phenotypes. This implies the development of analytical strategies based on the mea-
surement of metabolites to assess the effects of nutrition at both the organ-specific
compartment and system levels [25, 26, 111]. Therefore, the understanding of regu-
latory metabolic processes of a complex living organism at the system level involves
the assessment of spatiotemporal interorgan metabolic cross talks through the
analysis of biofluids. Recently, Montoliu et al. applied unsupervised chemometric
methods for integrating 1H NMR metabolic profiles from mouse plasma, liver, pan-
creas, adrenal gland, and kidney cortex matrices in order to infer inter-compartment
functional links [112]. The authors showed that integration of metabolic profiles
using advanced mathematical modeling approaches provided an overview of func-
tional relationships across matrices and enabled the characterization of compartment-
specific metabolite signatures, the spectrotypes. In particular, the methodology
enables the modeling of biochemical signatures common to different biological
matrices, which may highlight specific metabolic processes or cross talks shared by
different organs or specific to a given biological tissue or biofluid. In parallel, devel-
opments on computational modeling have established a global systems view of
human metabolism. A global genome- and bibliome-wide reconstruction of the
human metabolism was built using 3,311 reactions [113]. Metabolic activities at the
tissue level were predicted by a constraint-based computational model, from inte-
gration a genome-scale model with tissue-specific gene and protein data. Results
were obtained for ten human tissues (the brain, heart, kidney, ling, pancreas, pros-
tate, spleen, thymus, skeletal muscle, and liver) [114]. A genome-scale stoichiomet-
ric model of hepatic metabolism was obtained by combining literature-based
knowledge, transcriptomic, proteomic, metabolomic, and phenotypic data [115].
Likewise, modeling of the gut ecosystem could be applied as a combinatory analy-
sis of individual genome-scale metabolic models of gut bacteria, taking into account
their interactions [116]. System-wide computational approaches can be then useful
in nutritional applications to test hypothesis in silico on whole systems, in order to
study potential effects of diet or modulation of metabolic diseases. Such a combina-
torial approach could provide new research avenues to assess the role of gut micro-
biota and nutritional modulation of bacterial metabolic activities on host metabolism.
In the meantime, it is expected that analytical technologic developments will enable

sustainable deployment of such tests into clinical routine setups thanks to improved
automation, miniaturization, time to result, and cost per analysis.
The assessment of the inherent interindividual variability especially as regards
variable environmental conditions and genetic background will have to be properly
considered through longitudinal studies of a large cohort of subjects. The existence
of unique individual metabolic phenotypes has been hypothesized, but the experi-
mental evidence has been only recently collected [117, 118]. This was exemplified
by Bertini et al. through the analysis of individual phenotypes over the timescale of
years, which shows that the metabolic phenotypes are largely invariant. Such reports
also support the idea that the individual metabolic phenotype can also be considered
a metagenomic entity that is strongly affected by both gut microbiome and host
metabolic phenotype, the latter defined by both genetic and environmental contribu-
tions [117, 118]. Longitudinal metabolic profiling (e.g., generating individual meta-
bolic trajectories) and metagenomics are therefore foreseen as phenotypic variables
to stratify genetic backgrounds, offering an alternative to the averaging of genetic
variability within study groups defined using clinical criteria currently available.
One can foresee that such phenotypic-determined groups will aid in exploring the
underlined causes of inherent individual response to clinical therapeutic and nutri-
tion schemes (e.g., responders and nonresponders). This is also a feature of key
importance in the early life period for gastrointestinal and immune system matura-
tions and understanding metabolic and nutritional requirements for optimal growth
and development and their inferred effects with later-life diseases. The applications
of metabonomics have been so far very promising in the field of neonatology,
including intrauterine growth restriction, perinatal transition, asphyxia, brain injury
and hypothermia, metabolic diseases, perinatal programming, as well as inborn
error metabolism screening [3, 4, 119]. Such data will serve as reference to provide
understanding at the cellular and molecular level of the relationships between early
life nutritional status and the later disease risk predisposition and formulate future
nutritional concepts.
1.3 Gut Microbiota Metabolism in Nutrition,

Health, and Disease
Recently, an exhaustive gene catalogue containing virtually all of the prevalent gut
microbial genes in large human cohort and reported to which extent many bacterial
species are shared by different individuals [120]. Such an approach could be used
for global characterization of the genetic potential of ecologically complex environ-
ments [120] but also to help understand how gut microbiota specificities could be
exploited to develop new therapeutic and nutritional strategies. In particular, demo-
graphics have made aging and age-related chronic disease an enormous and grow-
ing biomedical and societal challenge [38]. The immune system undergoes profound
and multifaceted changes with aging. In particular, the homeostatic balance between
the proinflammatory and the antiinflammatory arms of the immune system is
skewed, resulting in a state of persistent low-grade systemic inflammation. In this
the origins and drivers of the “inflammaging” process are still poorly understood
[121], for instance, but intestinal dysbiosis appears a cornerstone in growth, devel-
opment, and aging.
The gastrointestinal tract is a complex ecosystem host to a diverse and highly
evolved microbial community composed of hundreds of different microbial species.
The human body contains circa 10 trillion parenchymal cells, but the gut contains
approximately 100 trillion of microbial organisms, which vary in community
composition through life according to lifestyle and nutrition [48, 122–124]. On
average, each individual harbors a unique combination of about 500–1,000 different
bacterial species [48, 122–124]. In mammals, microbial communities differ in com-
position from the stomach to the colon, where the competition for space and nutri-
ents in the large bowel contributes to the microbial composition of this internal
ecosystem. The main human intestinal bacteria coexist in a dynamic ecological
equilibrium together with various yeasts and other microorganisms [125]. The
members of the gut microbiota consortium are diverse and impart to the host-spe-
cific capacities ranging from dietary energy recovery from nutrient load, generating
digestible carbohydrates, short-chain fatty acids, amino acids, and vitamins, to pro-
tection against infectious diseases [47–50]. Activities of the diverse gut microbiota
can be highly specific, and it has been reported that the establishment of
Bifidobacteria is important for the development of the immune system and manage-
ment of gut functions [126–128]. As the microbiome interacts strongly with the host
to determine the metabolic phenotype [129, 130], and the metabolic phenotype
influences the outcomes of drug interventions [18, 27], understanding these interac-
tions as part of personalized healthcare solutions is clearly an important role [26].
Advances in metabonomic applications are providing novel insights into the molec-
ular foundations of these host-microbial relationships and their influence onto
health and disease risks [131, 132]. In particular, a series of investigations in humans
[133], rats [134], and gnotobiotic mice [131, 135] have provided a set of reference
metabolic profiles of gut intestinal biopsies that can be used to assess not only the
compartment structure and function but also the gut microbial impact at the tissue
level [135]. These studies therefore indicated that the type of gut microbiota may be
a key factor in the determination of the intestinal homeostasis, osmo-protection,
motility, and calorie recovery from the diet. For instance, the ileum is regarded as
the major site for absorption of luminal bile acids and emulsified dietary lipids, and
the report of higher concentrations of glycerophospholipids, glutathione, taurine,
and betaine in this tissue was consistent with its role in lipid and bile acid metabo-
lism. The jejunum of the mice harboring a nonadapted microbiota showed meta-
bolic similarities to the ileal profile, marked by higher levels of glutathione and
lower concentrations of its precursors in the γ-glutamyl cycle when compared to
conventional animals. Such an observation illustrates the essential role of the gut
microbiota to influence the surrounding tissue metabolic pathways, ultimately to
shape a host environment that fosters their implantation and persistence. The
investigations illustrated how microbial-dependent variations along the upper

intestine, an element often underestimated due to low bacterial populations, may
affect utilization efficiency of dietary proteins and amino acids and their subsequent
availability to extra-intestinal tissues. Moreover, some reference data were gener-
ated to investigate changes in gut functionality, such as gut permeability, using
metabolic profiling of biofluids [109, 136]. Since growth and aging are associated
with a number of significant changes in GI function, which may impact the daily
energy intake [101], such insights will be key for tailoring personal nutritional inter-
vention and monitoring patient response on an individual basis.
1.4 System Biology Approaches: Applications with Specific

Reference to Gut Microbial Metabolic Interactions
Recent findings are describing the deep and fundamental role of gut microbiota in
both positive and negative triggers of specific metabolic states of individuals and
populations [37, 98]. In this the generation of novel and more specific metabolic
readouts will help the deciphering of gut microbial influence on human health and
nutritional status [31, 36, 119]. In particular, both system-wide and organ-specific
metabolism may have components driven by gut microbial activities [137, 138],
which suggests that the dynamics of the gut microbiome could help maintain or
restore host metabolic homeostasis in disease and early onsets of metabolic
deregulations.
Today, direct metabolic insights into gut microbiota metabolism remain limited
due to the inaccessibility of the intestinal habitat and by the sheer complexity of the
gut microbiota [139]. The measurement of the gut microbial metabolism is gener-
ally confined to fecal samples, which is generally limited due to the elevated colonic
absorption of bacterial metabolites [36, 140]. However, such measurable outcomes
provide some essential insights – yet limited – into a small range of microbial activi-
ties within the colon. In particular, fecal analysis provides essential information on
the impact of microbial activities on specific biological processes, including the
metabolism of bile acids, SCFAs, or heterocyclic amines [141–143]. The metabolic
composition of fecal extracts provides therefore a window for elucidating the com-
plex metabolic interplay between mammals and their intestinal ecosystems, and
these metabolic profiles can yield information on a range of gut diseases. Saric et al.
employed 1H NMR-based metabolic analysis of fecal water extracts to assess inher-
ent similarities and dissimilarities across different mammals, namely, humans,
mice, and rats [144]. This study provided an important milestone as per character-
ization of many fecal metabolites common to the three species, such as SCFAs and
branched-chain amino acids. The authors also described how not only the presence
but also the proportion of the different biochemical compounds resulted in a species-
specific profile, encapsulating interindividual variations (i.e., the natural genetic and
environmental diversity in human populations). Furthermore, NMR spectroscopy,
using high-resolution magic angle spinning NMR (HR-MAS) [145], offers a unique
prospect by the holistic and simultaneous profile of hundreds of metabolites in
intact tissue biopsies. Therefore, HR-MAS presents the unique feature of ensuring
the integrity and organizational compartmentation of the biological samples.
Advanced data modeling strategies enabled the visualization of metabolic similari-
ties and differences along the gut tract that were induced by different gut microbial
ecologies, namely, mice with a conventional mouse microbiota or inoculated with a
simplified model of human-derived microbiota [131, 132]. These studies indicated
that the type of gut microbiota may be a key factor in the determination of intestinal
homeostasis, osmo-protection, motility, and calorie recovery from the diet. Indeed,
the symbiotic gut microbiome exerts a strong influence on the metabolic phenotype
of the mammalian host and participates in extensive microbial-mammalian co-
metabolism [26, 112, 146, 147]. The integrated metabolism of the bile acid pools in
mammals is a good example of the complex transgenomic biochemical interactions
between host and microbiome symbionts [33], which are crucial for the absorption
of dietary fats and lipid-soluble vitamins in the intestine [148].
Additional applications of top-down systems biology approaches revealed the
depth and width of the long-range effects of gut microbiota modulation in complex
organisms, resulting in modulation of host lipid, carbohydrate, and amino acid
metabolism at a panorganismal scale [44, 137, 149–151]. Wikoff et al. provided
additional evidence that the specific metabolic activities of a single gut bacterial
species can provide the host with new biochemical compounds in sufficient amount
to be detected in the general blood circulation [152]. Martin et al. exemplified how
the gut microbial modulation of the gastrointestinal system [131] and extensive
microbial-mammalian co-metabolism may fine-tune host metabolic processes and
may induce metabolic deregulations [33]. In this case, gut bacteria exert modulation
over the host metabolism via reprocessing of signaling molecules, i.e., bile acids.
As such, bile acids may be an example of transgenomic mechanism of quorum sens-
ing [44] whereby microbial cells communicate with each other and disperse their
metabolic functions, thus behaving like a multicellular organism [33]. Martin et al.
evaluated the effects of the induction of a nonadapted microbiota in a murine model
(human-derived microbiota) on the host metabolism by comparison with animals
colonized with a natural gut microbiota (conventional), the result of a long period of
coevolution [33]. The simplified human microbiota was not adapted to mice, which
modified the physiology of the murine host towards a pre-pathologic state. While
conventionalized mice evolve to normal gut flora from an ecological point of view,
the simplified human microbiota maintained the gut tract and the liver out of a sus-
tainable mouse ecological equilibrium, as denoted by increased lipid accumulation
in the liver and lower concentrations of glutathione, which together are associated
with a higher lipoperoxidation risk. In another study, Claus et al. described the
determinant role of the gut microbiota on the metabolism of bile acids through the
enterohepatic cycle, as noted by higher levels of phospholipids in the liver of germ-
free mice and higher levels of bile acids in gut tissues [153]. The presence of the gut
microbiota was shown as a key regulator of the bile acids metabolism and some
CYP family enzymes.
In addition to the building awareness that specific environmental exposure

(e.g., stress perception) may induce specific and chronic gut functional activities
[109, 154], increasing evidence tends to show that long-term dietary habits may also
have a significant impact in determining chronically the gut functional ecology,
which may have a long-term health consequence [24, 155, 156]. Therefore, the gut
microbiota functionality remains to be further explored and understood, in order to
assess how its nutritional modulation at individual and population level may be used
for future nutritional management solutions. It is therefore crucial to decipher the
foundations of the reciprocal metabolic influences between host and microbiota to
better define the role of gut microbes in determining gastrointestinal functional ecol-
ogy. The gut microbiota does not only determine absorption, digestion, metabolism,
and excretion of dietary nutrients but may also determine the metabolic fate of both
ingested nutrients and host cell molecular machinery, which nowadays become fun-
damental for developing individual disease and nutritional management solutions.
Recent evidence has shown that the influence of the gut microorganisms might
be more important in the progression of human diseases than was previously sus-
pected [157] and is of main concern in the etiology and/or maintenance of gut dys-
functions, such as IBS [158] or IBD [159]. Metabonomics was successfully applied
to characterize the metabolic response to the induction of intestinal ischemia reper-
fusion injury by portal vein outflow occlusion in rats [160]. This combinatorial
approach provides novel insights into metabolic signatures associated with oxida-
tive stress, as noted by increased glycolysis and fatty acid and amino acid accumula-
tion. Metabonomics was proven to be a valuable diagnostic tool to differentiate
active and quiescent ulcerative colitis, as per the analysis of intact gut biopsies and
colonocytes [161]. This approach was also successfully employed to provide
insights into the molecular processes associated with the development of ulcerative
colitis, using blood plasma [162] or urine [163] analysis, the latter revealing a pos-
sible contribution of gut microbiota via methylamine metabolism. Bertini et al. also
demonstrated that combinatorial metabonomic analysis of blood sera and urine
could help further the understanding of Celiac disease [164]. The authors high-
lighted major urinary changes in gut microbial co-metabolites phenylacetylglycine,
indoxylsulfate, and meta-hydroxy-phenylpropionate, which may be associated with
aberrant microbiota previously characterized in the small bowel of subjects suffer-
ing from Celiac disease [165].
More recently, metabonomics was successfully employed to demonstrate a
relationship between gut microbial metabolism of dietary phosphatidylcholine and
cardiovascular pathogenesis in humans [94]. Intestinal microbiota metabolism of
choline and phosphatidylcholine produces trimethylamine (TMA), which is further
metabolized to a proatherogenic species, trimethylamine-N-oxide (TMAO). Wang
et al. described how the circulating levels of choline, trimethylamine-N-oxide, and
betaine were shown to predict the risk for cardiovascular events, which offer novel
perspectives for therapeutical and nutraceutical approaches, as demonstrated by
their follow-up [166, 167]. They further demonstrated how metabolism by intestinal
microbiota of dietary l-carnitine, a trimethylamine abundant in red meat, also pro-
duces TMAO and accelerates atherosclerosis in mice. In humans they reported how
the presence of specific bacterial taxa in human feces was associated with both
plasma TMAO concentration and dietary status. More interestingly, they described
that high plasma TMAO levels combined with plasma l-carnitine levels in subjects
undergoing cardiac evaluation predicted increased risks for both prevalent cardio-
vascular disease and incident major adverse cardiac events (myocardial infarction,
stroke, or death) [166]. In parallel, significant scientific efforts were put in identify-
ing specific disease risk genotypes. But it is only recently that the incorporation of
metabonomics in genome-wide association studies has offered novel opportunities
for exploring disease-related metabolic deregulations and interactions between
environmental exposure, lifestyle, genetic predisposition, and actual metabolic
phenotype at individual and population scale [28]. Such an approach has recently
described an association between urinary trimethylamine and plasma dimethyl-
amine – two gut microbial metabolites – and the pyridine nucleotide-disulfide oxi-
doreductase gene PYROXD2 [168]. The present association is of importance and
suggests how the conversion rate TMAO/TMA can be related with key hepatic
functions that are under genetic determinant and may predispose individuals to spe-
cific disease risks. Such approaches offer novel avenues to screen individuals with
specific predisposition and determine candidate metabolic targets to be used to
develop tailor-made nutritional management program.
1.5 Novel Directions and Perspectives
Metabonomics is increasingly becoming popular in clinical research due to its

unique attractiveness to generate functional and system readouts of individuals,
building the elementary steps for future personalized nutraceutical and therapeutical
care. Yet, the clinical translation of this unprecedented source of metabolic knowl-
edge, which is now at hand of clinical practices, requires properly addressing the
dynamics of physiological and biological adaptations throughout lifetime, includ-
ing the various intrinsic and extrinsic influential factors. In addition, the develop-
ment of systems biology approaches and the new generation of biomarker patterns
will provide the opportunity to associate complex metabolic regulations with key
mammalian and gut microbial biological processes. The link between nutrition,
host-microbe interactions, and metabolism is so tightly interlinked in mammalian
systems that it is still difficult to understand and unravel causality and molecular
mechanisms of action. In this the lessons learnt and to be learnt from epidemiologi-
cal metabonomic studies, including twin cohorts, will help in delineating the
interaction of nutrients, gene expression, and metabolism from environmental pres-
sures which ultimately determine the physiological and the disease risk status of
individuals. Furthermore, metabolic phenotypes will need to capture these dynamics
at the molecular and system levels, requiring longitudinal sampling and proper
assessment of inherent interindividual variability. Therefore, advances in linking
metabolite data to known and validated clinically relevant indices will have to be
seriously considered in addition to consolidate metabonomics and metabolic mark-
ers in clinical translation/applications.
The generalization of systems biology approaches will subsequently lead to the

development of system mechanistic hypotheses that could be targeted with new
nutritional and therapeutic personalized concepts. As displayed above, both system-
wide (i.e., whole organism) and organ-specific changes in metabolic profiles may
have components driven by gut microbial activities which are of relevance in nutri-
tion, health, and disease research as exemplified hereafter in the book. In particular,
high-resolution metabonomics data combined with high-resolution microbial pro-
filing are envisioned as a way forward to demonstrate direct and indirect functional
relationships of gut microbiota metabolic activities on host cell metabolic pathways
in key organs and biofluids. But we are far from the complete elucidation of the
molecular processes linking bacteria, metabolic enzymes, and metabolites with
their corresponding biological functions. There are indeed several diverting issues
to be taken into account, such as the fact that abundant concentrations of either pro-
teins or metabolites may actually be mediated by a low-abundance microbe. Because
all studies so far have relied on fecal samples as a mirror of the whole gut microbi-
ome, these might not accommodate local functionalities, as environments between
distal, transverse, and proximal colon are known to be divergent [169]. Undoubtedly,
the interplay between gut microbiome and host and its impact in nutrition will ben-
efit from the integration of information on a systems biology-wide approach. Gene
sequencing of the microbiome, metaproteomics, and metabolomics are starting
points to understand the complex mammalian superorganism.
Such future scientific developments are foreseen as a promising step forward to
study the spatiotemporal changes in interorgan metabolic cross talks and quantify
the metabolic contribution of the gut microbiota to host cell machinery and physi-
ological states. Moreover, such novel concepts have great potential for advancing
our mechanistic knowledge of how different environmental triggers, including
nutrition, can impact metabolic health of complex mammalian organisms.
References
1. Abegunde DO, Mathers CD, Adam T, Ortegon M, Strong K. The burden and costs of chronic
diseases in low-income and middle-income countries. Lancet. 2007;370(9603):1929–38.
2. Strong K, Mathers C, Leeder S, Beaglehole R. Preventing chronic diseases: how many lives
can we save? Lancet. 2005;366(9496):1578–82.
3. Antonucci R, Atzori L, Barberini L, Fanos V. Metabolomics: the “new clinical chemistry” for
personalized neonatal medicine. Minerva Pediatr. 2010;62(3 Suppl 1):145–8.
4. Atzori L, Antonucci R, Barberini L, Griffin JL, Fanos V. Metabolomics: a new tool for the
neonatologist. J Matern Fetal Neonatal Med. 2009;22 Suppl 3:50–3.
5. Rezzi S, Ramadan Z, Fay LB, Kochhar S. Nutritional metabonomics: applications and per-
spectives. J Proteome Res. 2007;6(2):513–25.
6. Van Ommen B, Stierum R. Nutrigenomics: exploiting systems biology in the nutrition and
health arena. Curr Opin Biotechnol. 2002;13(5):517–21.
7. Trujillo E, Davis C, Milner J. Nutrigenomics, proteomics, metabolomics, and the practice of
dietetics. J Am Diet Assoc. 2006;106(3):403–13.
8. Van Ommen B. Nutrigenomics: exploiting systems biology in the nutrition and health arenas.
Nutrition. 2004;20:4–8.
9. Muller M, Kersten S. Nutrigenomics: goals and strategies. Nat Genet. 2003;4:315–22.
10. Corthesy-Theulaz IE, Dunnen J-T, Ferre P, Geurts J-MW, Mueller M, Belzen N, et al.
Nutrigenomics: the impact of biomics technology on nutrition research. Ann Nutr Metab.
2005;49(6):355–65.
11. Ordovas JM. Integrating environment and disease into “omic” analysis. Rev Esp Cardiol.
2009;62 Suppl 2:17–22.
12. Ordovas JM, Shen J. Gene-environment interactions and susceptibility to metabolic syndrome
and other chronic diseases. J Periodontol. 2008;79(8 Suppl):1508–13.
13. van der Greef J, Hankemeier T, McBurney RN. Metabolomics-based systems biology and
personalized medicine: moving towards n = 1 clinical trials? Pharmacogenomics.
2006;7(7):1087–94.
14. van der GJ. Systems biology, connectivity and the future of medicine. Syst Biol (Stevenage).
2005;152(4):174–8.
15. Nicholson JK, Lindon JC. Systems biology: metabonomics. Nature. 2008;455(7216):
1054–6.
16. Holmes E, Wilson ID, Nicholson JK. Metabolic phenotyping in health and disease. Cell.
2008;134(5):714–7.
17. Nicholson JK, Wilson ID. Opinion: understanding ‘global’ systems biology: metabonomics
and the continuum of metabolism. Nat Rev Drug Discov. 2003;2(8):668–76.
18. Nicholson JK, Holmes E, Lindon JC, Wilson ID. The challenges of modeling mammalian
biocomplexity. Nat Biotechnol. 2004;22(10):1268–74.
19. Keurentjes JJ. Genetical metabolomics: closing in on phenotypes. Curr Opin Plant Biol.
2009;12(2):223–30.
20. Saito K, Matsuda F. Metabolomics for functional genomics, systems biology, and biotechnol-
ogy. Annu Rev Plant Biol. 2008.
21. Weckwerth W. Metabolomics in systems biology. Annu Rev Plant Biol. 2003;54:669–89.
22. Lindon JC, Keun HC, Ebbels TM, Pearce JM, Holmes E, Nicholson JK. The Consortium for
Metabonomic Toxicology (COMET): aims, activities and achievements. Pharmacogenomics.
2005;6(7):691–9.
23. Nicholson JK, Connelly J, Lindon JC, Holmes E. Metabonomics: a platform for studying
drug toxicity and gene function. Nat Rev Drug Discov. 2002;1(2):153–61.
24. Nicholson JK, Wilson ID, Lindon JC. Pharmacometabonomics as an effector for personal-
ized medicine. Pharmacogenomics. 2011;12(1):103–11.
25. Nicholson JK. Omics dreams of personalized healthcare. J Proteome Res.
2006;5(9):2067–9.
26. Nicholson JK. Global systems biology, personalized medicine and molecular epidemiology.
Mol Syst Biol. 2006;2:52.
27. Clayton TA, Lindon JC, Cloarec O, Antti H, Charuel C, Hanton G, et al. Pharmaco-
metabonomic phenotyping and personalized drug treatment. Nature. 2006;440(7087):
1073–7.
28. Montoliu I, Genick U, Ledda M, Collino S, Martin FP, le Coutre J, et al. Current status on
genome-metabolome-wide associations: an opportunity in nutrition research. Genes Nutr.
2013;8(1):19–27.
29. Moco S, Martin FP, Rezzi S. A metabolomics view on gut microbiome modulation by
polyphenol-rich foods. J Proteome Res. 2012;11:4781–90.
30. Kussmann M, Rezzi S, Daniel H. Profiling techniques in nutrition and health research. Curr
Opin Biotechnol. 2008;19(2):83–99.
31. Claus SP, Swann JR. Nutrimetabonomics: applications for nutritional sciences, with specific
reference to gut microbial interactions. Annu Rev Food Sci Technol. 2013;4:381–99.
32. Nicholson JK, Everett JR, Lindon JC. Longitudinal pharmacometabonomics for predicting
patient responses to therapy: drug metabolism, toxicity and efficacy. Expert Opin Drug Metab
Toxicol. 2012;8(2):135–9.
33. Martin FP, Dumas ME, Wang Y, Legido-Quigley C, Yap IK, Tang H, et al. A top-down
systems biology view of microbiome-mammalian metabolic interactions in a mouse model.
34. Martin FP, Verdu EF, Wang Y, Dumas ME, Yap IK, Cloarec O, et al. Transgenomic metabolic
interactions in a mouse disease model: interactions of Trichinella spiralis infection with
dietary Lactobacillus paracasei supplementation. J Proteome Res. 2006;5(9):2185–93.
35. Sommer F, Backhed F. The gut microbiota – masters of host development and physiology.
Nat Rev Microbiol. 2013;11(4):227–38.
36. Martin FP, Collino S, Rezzi S, Kochhar S. Metabolomic applications to decipher gut micro-
bial metabolic influence in health and disease. Front Physiol. 2012;3:113.
37. Nicholson JK, Holmes E, Kinross J, Burcelin R, Gibson G, Jia W, et al. Host-gut microbiota
metabolic interactions. Science. 2012;336(6086):1262–7.
38. Rehman T. Role of the gut microbiota in age-related chronic inflammation. Endocr Metab
Immune Disord Drug Targets. 2012;12(4):361–7.
39. Zheng X, Xie G, Zhao A, Zhao L, Yao C, Chiu NH, et al. The footprints of gut microbial-
mammalian co-metabolism. J Proteome Res. 2011;10(12):5512–22.
40. Tiihonen K, Ouwehand AC, Rautonen N. Human intestinal microbiota and healthy ageing.
Ageing Res Rev. 2010;9(2):107–16.
41. Biagi E, Nylund L, Candela M, Ostan R, Bucci L, Pini E, et al. Through ageing, and beyond:
gut microbiota and inflammatory status in seniors and centenarians. PLoS One.
2010;5(5):e10667.
42. Cani PD, Delzenne NM. Gut microflora as a target for energy and metabolic homeostasis.
Curr Opin Clin Nutr Metab Care. 2007;10(6):729–34.
43. Ley RE, Hamady M, Lozupone C, Turnbaugh PJ, Ramey RR, Bircher JS, et al. Evolution of
mammals and their gut microbes. Science. 2008;320(5883):1647–51.
44. Nicholson JK, Holmes E, Wilson ID. Gut microorganisms, mammalian metabolism and per-
sonalized health care. Nat Rev Microbiol. 2005;3(5):431–8.
45. Dethlefsen L, Fall-Ngai M, Relman DA. An ecological and evolutionary perspective on
human-microbe mutualism and disease. Nature. 2007;449(7164):811–8.
46. Gill SR, Pop M, Deboy RT, Eckburg PB, Turnbaugh PJ, Samuel BS, et al. Metagenomic
analysis of the human distal gut microbiome. Science. 2006;312(5778):1355–9.
47. Gilmore MS, Ferretti JJ. The thin line between gut commensal and pathogen. Science.
2003;299:1999–2002.
48. Tannock GW. New perceptions of the gut microbiota: implications for future research.
Gastroenterol Clin North Am. 2005;34(3):361–82, vii.
49. Xu J, Bjursell MK, Himrod J, Deng S, Carmichael LK, Chaing HC, et al. A genomic view of
the human-Bacteroides thetaiotaomicron symbiosis. Science. 2003;299:2074–6.
50. Cummings JH, Pomare EW, Branch WJ, Naylor CP, Macfarlane GT. Short chain fatty acids
in human large intestine, portal, hepatic and venous blood. Gut. 1987;28(10):1221–7.
51. Guigoz Y, Dore J, Schiffrin EJ. The inflammatory status of old age can be nurtured from the
intestinal environment. Curr Opin Clin Nutr Metab Care. 2008;11(1):13–20.
52. Dunne C. Adaptation of bacteria to the intestinal niche: probiotics and gut disorder. Inflamm
Bowel Dis. 2001;7(2):136–45.
53. Turnbaugh P, Ley R, Mahowald M, Magrini V, Mardis E, Gordon J. An obesity-associated gut
microbiome with increased capacity for energy harvest. Nature. 2006;444:1027–31.
54. Jia W, Li H, Zhao L, Nicholson JK. Gut microbiota: a potential new territory for drug target-
ing. Nat Rev Drug Discov. 2008;7(2):123–9.
55. Lindon JC, Nicholson JK, Holmes E, Antti H, Bollard ME, Keun H, et al. Contemporary
issues in toxicology the role of metabonomics in toxicology and its evaluation by the COMET
project. Toxicol Appl Pharmacol. 2003;187(3):137–46.
56. Clayton TA, Baker D, Lindon JC, Everett JR, Nicholson JK. Pharmacometabonomic identi-
fication of a significant host-microbiome metabolic interaction affecting human drug
metabolism. Proc Natl Acad Sci U S A. 2009;106(34):14728–33.
57. Cunningham K, Claus SP, Lindon JC, Holmes E, Everett JR, Nicholson JK, et al.
Pharmacometabonomic characterization of xenobiotic and endogenous metabolic pheno-
types that account for inter-individual variation in isoniazid-induced toxicological response.
J Proteome Res. 2012;11:4630–42.
58. Coen M, Goldfain-Blanc F, Rolland-Valognes G, Walther B, Robertson DG, Holmes E, et al.
Pharmacometabonomic investigation of dynamic metabolic phenotypes associated with vari-
ability in response to galactosamine hepatotoxicity. J Proteome Res. 2012;11(4):2427–40.
59. Holmes E, Loo RL, Stamler J, Bictash M, Yap IK, Chan Q, et al. Human metabolic phenotype
diversity and its association with diet and blood pressure. Nature. 2008;453(7193):
396–400.
60. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, et al. Diagnosis
and management of the metabolic syndrome: an American Heart Association/National Heart,
Lung, and Blood Institute scientific statement: executive summary. Crit Pathw Cardiol.
2005;4(4):198–203.
61. Wirfalt E, Hedblad B, Gullberg B, Mattisson I, Andren C, Rosander U, et al. Food patterns
and components of the metabolic syndrome in men and women: a cross-sectional study
within the Malmo Diet and Cancer cohort. Am J Epidemiol. 2001;154(12):1150–9.
62. Flegal KM, Carroll MD, Kit BK, Ogden CL. Prevalence of obesity and trends in the distribu-
tion of body mass index among US adults, 1999–2010. JAMA. 2012;307(5):491–7.
63. Lasserre AM, Chiolero A, Paccaud F, Bovet P. Worldwide trends in childhood obesity. Swiss
Med Wkly. 2007;137(9–10):157–8.
64. Chiolero A, Lasserre AM, Paccaud F, Bovet P. Childhood obesity: definition, consequences,
and prevalence. Rev Med Suisse. 2007;3(111):1262–9.
65. Donahue RP, Abbott RD. Central obesity and coronary heart disease in men. Lancet.
1987;2(8569):1215.
66. Lapidus L, Bengtsson C, Larsson B, Pennert K, Rybo E, Sjostrom L. Distribution of adipose
tissue and risk of cardiovascular disease and death: a 12 year follow up of participants in the
population study of women in Gothenburg, Sweden. Br Med J (Clin Res Ed).
1984;289(6454):1257–61.
67. Larsson B, Renstrom P, Svardsudd K, Welin L, Grimby G, Eriksson H, et al. Health and age-
ing characteristics of highly physically active 65-year-old men. Eur Heart J. 1984;5(Suppl
E):31–5.
68. Kalkhoff RK, Hartz AH, Rupley D, Kissebah AH, Kelber S. Relationship of body fat distribu-
tion to blood pressure, carbohydrate tolerance, and plasma lipids in healthy obese women. J
Lab Clin Med. 1983;102(4):621–7.
69. Hartz AJ, Rupley Jr DC, Kalkhoff RD, Rimm AA. Relationship of obesity to diabetes: influ-
ence of obesity level and body fat distribution. Prev Med. 1983;12(2):351–7.
70. Despres JP, Tremblay A, Leblanc C, Bouchard C. Effect of the amount of body fat on the
age-associated increase in serum cholesterol. Prev Med. 1988;17(4):423–31.
71. Park SH, Kim BI, Kim SH, Kim HJ, Park DI, Cho YK, et al. Body fat distribution and insulin
resistance: beyond obesity in nonalcoholic fatty liver disease among overweight men. J Am
Coll Nutr. 2007;26(4):321–6.
72. Folsom AR, Kaye SA, Sellers TA, Hong CP, Cerhan JR, Potter JD, et al. Body fat distribution
and 5-year risk of death in older women. JAMA. 1993;269(4):483–7.
73. Sorensen TI, Boutin P, Taylor MA, Larsen LH, Verdich C, Petersen L, et al. Genetic polymor-
phisms and weight loss in obesity: a randomised trial of hypo-energetic high- versus low-fat
diets. PLoS Clin Trials. 2006;1(2):e12.
74. Klannemark M, Orho M, Langin D, Laurell H, Holm C, Reynisdottir S, et al. The putative
role of the hormone-sensitive lipase gene in the pathogenesis of Type II diabetes mellitus and
abdominal obesity. Diabetologia. 1998;41(12):1516–22.
75. Teran-Garcia M, Bouchard C. Genetics of the metabolic syndrome. Appl Physiol Nutr Metab.
2007;32(1):89–114.
76. Ley R, Turnbaugh P, Klein S, Gordon J. Microbial ecology: human gut microbes associated
with obesity. Nature. 2006;444(7122):1022–3.
77. Wildman RP, Muntner P, Reynolds K, McGinn AP, Rajpathak S, Wylie-Rosett J, et al. The
obese without cardiometabolic risk factor clustering and the normal weight with cardiometa-
bolic risk factor clustering: prevalence and correlates of 2 phenotypes among the US popula-
tion (NHANES 1999–2004). Arch Intern Med. 2008;168(15):1617–24.
78. Smith Jr SC, Haslam D. Abdominal obesity, waist circumference and cardio-metabolic risk:
awareness among primary care physicians, the general population and patients at risk–the
Shape of the Nations survey. Curr Med Res Opin. 2007;23(1):29–47.
79. Newgard CB, An J, Bain JR, Muehlbauer MJ, Stevens RD, Lien LF, et al. A branched-chain
amino acid-related metabolic signature that differentiates obese and lean humans and contrib-
utes to insulin resistance. Cell Metab. 2009;9(4):311–26.
80. Suhre K, Meisinger C, Doring A, Altmaier E, Belcredi P, Gieger C, et al. Metabolic footprint
of diabetes: a multiplatform metabolomics study in an epidemiological setting. PLoS One.
2010;5(11):e13953.
81. Huffman KM, Shah SH, Stevens RD, Bain JR, Muehlbauer M, Slentz CA, et al. Relationships
between circulating metabolic intermediates and insulin action in overweight to obese, inac-
tive men and women. Diabetes Care. 2009;32(9):1678–83.
82. Fiehn O, Garvey WT, Newman JW, Lok KH, Hoppel CL, Adams SH. Plasma metabolomic
profiles reflective of glucose homeostasis in non-diabetic and type 2 diabetic obese African-
American women. PLoS One. 2010;5(12):e15234.
83. Boulange CL, Claus SP, Chou CJ, Collino S, Montoliu I, Kochhar S, et al. Early metabolic
adaptation in C57BL/6 mice resistant to high fat diet induced weight gain involves an activa-
tion of mitochondrial oxidative pathways. J Proteome Res. 2013;12:1956–68.
84. Rhee EP, Cheng S, Larson MG, Walford GA, Lewis GD, McCabe E, et al. Lipid profiling
identifies a triacylglycerol signature of insulin resistance and improves diabetes prediction in
humans. J Clin Invest. 2011;121(4):1402–11.
85. Wang TJ, Larson MG, Vasan RS, Cheng S, Rhee EP, McCabe E, et al. Metabolite profiles and
the risk of developing diabetes. Nat Med. 2011;17(4):448–53.
86. Yamakado M, Tanaka T, Nagao K, Ishizaka Y, Mitushima T, Tani M, et al. Plasma amino acid
profile is associated with visceral fat accumulation in obese Japanese subjects. Clin Obes.
2012;2(1–2):29–40.
87. Floegel A, Stefan N, Yu Z, Muhlenbruch K, Drogan D, Joost HG. et al. Identification of
serum metabolites associated with risk of type 2 diabetes using a targeted metabolomic
approach. Diabetes. 2013;62(2):639–48.
88. Snyder F. Platelet-activating factor and its analogs: metabolic pathways and related intracel-
lular processes. Biochim Biophys Acta. 1995;1254(3):231–49.
89. McIntyre TM, Prescott SM, Stafforini DM. The emerging roles of PAF acetylhydrolase. J
Lipid Res. 2009;50(Suppl):S255–9.
90. Davi G, Guagnano MT, Ciabattoni G, Basili S, Falco A, Marinopiccoli M, et al. Platelet acti-
vation in obese women: role of inflammation and oxidant stress. JAMA.
2002;288(16):2008–14.
91. Basili S, Pacini G, Guagnano MT, Manigrasso MR, Santilli F, Pettinella C, et al. Insulin
resistance as a determinant of platelet activation in obese women. J Am Coll Cardiol.
2006;48(12):2531–8.
92. Fabbrini E, Sullivan S, Klein S. Obesity and nonalcoholic fatty liver disease: biochemical,
metabolic, and clinical implications. Hepatology. 2010;51(2):679–89.
93. Johnson NA, Walton DW, Sachinwalla T, Thompson CH, Smith K, Ruell PA, et al.
Noninvasive assessment of hepatic lipid composition: advancing understanding and manage-
ment of fatty liver disorders. Hepatology. 2008;47(5):1513–23.
94. Wang Z, Klipfell E, Bennett BJ, Koeth R, Levison BS, Dugar B, et al. Gut flora metabolism
of phosphatidylcholine promotes cardiovascular disease. Nature. 2011;472(7341):57–63.
95. Buchwald H, Avidor Y, Braunwald E, Jensen MD, Pories W, Fahrbach K, et al. Bariatric
surgery: a systematic review and meta-analysis. JAMA. 2004;292(14):1724–37.
96. Seyfried F, Li JV, Miras AD, Cluny NL, Lannoo M, Fenske WK, et al. Urinary phenotyping
indicates weight loss-independent metabolic effects of Roux-en-Y gastric bypass in mice. J
Proteome Res. 2013;12(3):1245–53.
97. Tremaroli V, Backhed F. Functional interactions between the gut microbiota and host metabo-
lism. Nature. 2012;489(7415):242–9.
98. Holmes E, Kinross J, Gibson GR, Burcelin R, Jia W, Pettersson S, et al. Therapeutic modula-
tion of microbiota-host metabolic interactions. Sci Transl Med. 2012;4(137):137rv6.
99. Collino S, Martin FP, Rezzi S. Clinical metabolomics paves the way towards future health-
care strategies. Br J Clin Pharmacol. 2013;75(3):619–29.
100. Martin FP, Collino S, Rezzi S. 1H NMR-based metabonomic applications to decipher gut
microbial metabolic influence on mammalian health. Magn Reson Chem. 2011;49 Suppl
1:S47–54.
101. Parker BA, Chapman IM. Food intake and ageing–the role of the gut. Mech Ageing Dev.
2004;125(12):859–66.
102. Harper PH, Truelove SC, Lee EC, Kettlewell MG, Jewell DP. Split ileostomy and ileocolos-
tomy for Crohn’s disease of the colon and ulcerative colitis: a 20 year survey. Gut.
1983;24(2):106–13.
103. Duncan M, Davison JS, Sharkey KA. Review article: endocannabinoids and their receptors
in the enteric nervous system. Aliment Pharmacol Ther. 2005;22(8):667–83.
104. Actis GC, Pazienza P, Rosina F. Mesalamine for inflammatory bowel disease: recent reap-
praisals. Inflamm Allergy Drug Targets. 2008;7(1):1–5.
105. Winterkamp S, Weidenhiller M, Otte P, Stolper J, Schwab D, Hahn EG, et al. Urinary excre-
tion of N-methylhistamine as a marker of disease activity in inflammatory bowel disease. Am
J Gastroenterol. 2002;97(12):3071–7.
106. Baur P, Martin FP, Gruber L, Bosco N, Brahmbhatt V, Collino S, et al. Metabolic phenotyping
of the Crohn’s disease-like IBD etiopathology in the TNF(DeltaARE/WT) mouse model. J
Proteome Res. 2011;10(12):5523–35.
107. Marchesi JR, Holmes E, Khan F, Kochhar S, Scanlan P, Shanahan F, et al. Rapid and nonin-
vasive metabonomic characterization of inflammatory bowel disease. J Proteome Res.
2007;6(2):546–51.
108. Le GG, Noor SO, Ridgway K, Scovell L, Jamieson C, Johnson IT, et al. Metabolomics of
fecal extracts detects altered metabolic activity of gut microbiota in ulcerative colitis and
irritable bowel syndrome. J Proteome Res. 2011;10(9):4208–18.
109. Rezzi S, Martin FP, Alonso C, Guilarte M, Vicario M, Ramos L, et al. Metabotyping of bio-
fluids reveals stress-based differences in gut permeability in healthy individuals. J Proteome
Res. 2009;8(10):4799–809.
110. Brugnara L, Vinaixa M, Murillo S, Samino S, Rodriguez MA, Beltran A, et al. Metabolomics
approach for analyzing the effects of exercise in subjects with type 1 diabetes mellitus. PLoS
One. 2012;7(7):e40600.
111. Rezzi S, Martin FP, Kochhar S. Defining personal nutrition and metabolic health through
metabonomics. Ernst Schering Found Symp Proc. 2007;4:251–64.
112. Montoliu I, Martin FP, Collino S, Rezzi S, Kochhar S. Multivariate modeling strategy for
intercompartmental analysis of tissue and plasma 1H NMR spectrotypes. J Proteome Res.
2009;8(5):2397–406.
113. Duarte NC, Becker SA, Jamshidi N, Thiele I, Mo ML, Vo TD, et al. Global reconstruction of
the human metabolic network based on genomic and bibliomic data. Proc Natl Acad Sci U S
A. 2007;104(6):1777–82.
114. Shlomi T, Cabili MN, Herrgard MJ, Palsson BO, Ruppin E. Network-based prediction of
human tissue-specific metabolism. Nat Biotechnol. 2008;26(9):1003–10.
115. Jerby L, Shlomi T, Ruppin E. Computational reconstruction of tissue-specific metabolic
models: application to human liver metabolism. Mol Syst Biol. 2010;6:401.
116. Karlsson FH, Nookaew I, Petranovic D, Nielsen J. Prospects for systems biology and
modeling of the gut microbiome. Trends Biotechnol. 2011;29(6):251–8.
117. Bernini P, Bertini I, Luchinat C, Nepi S, Saccenti E, Schafer H, et al. Individual human phe-
notypes in metabolic space and time. J Proteome Res. 2009;8(9):4264–71.
118. Assfalg M, Bertini I, Colangiuli D, Luchinat C, Schafer H, Schutz B, et al. Evidence of dif-
ferent metabolic phenotypes in humans. Proc Natl Acad Sci U S A. 2008;105(5):1420–4.
119. Moco S, Collino S, Rezzi S, Martin FP. Metabolomics perspectives in pediatric research.
Pediatr Res. 2013;73(4 Pt 2):570–6.
120. Qin J, Li R, Raes J, Arumugam M, Burgdorf KS, Manichanh C, et al. A human gut microbial
gene catalogue established by metagenomic sequencing. Nature. 2010;464(7285):59–65.
121. Franceschi C, Capri M, Monti D, Giunta S, Olivieri F, Sevini F, et al. Inflammaging and anti-
inflammaging: a systemic perspective on aging and longevity emerged from studies in
humans. Mech Ageing Dev. 2007;128(1):92–105.
122. Xu J, Gordon JI. Inaugural article: honor thy symbionts. Proc Natl Acad Sci U S A.
2003;100(18):10452–9.
123. Xu J, Mahowald MA, Ley RE, Lozupone CA, Hamady M, Martens EC, et al. Evolution of
symbiotic bacteria in the distal human intestine. PLoS Biol. 2007;5(7):e156.
124. Beale B. Probiotics: their tiny worlds are under scrutiny. Scientist. 2002;16(15):20–2.
125. Eckburg PB, Bik EM, Bernstein CN, Purdom E, Dethlefsen L, Sargent M, et al. Diversity of
the human intestinal microbial flora. Science. 2005;308(5728):1635–8.
126. Ouwehand AC. Antiallergic effects of probiotics. J Nutr. 2007;137(3 Suppl 2):794S–7.
127. Salminen SJ, Gueimonde M, Isolauri E. Probiotics that modify disease risk. J Nutr.
2005;135(5):1294–8.
128. Blum S, Schiffrin EJ. Intestinal microflora and homeostasis of the mucosal immune response:
implications for probiotic bacteria? Curr Issues Intest Microbiol. 2003;4(2):53–60.
129. Gavaghan McKee CL, Wilson ID, Nicholson JK. Metabolic phenotyping of nude and normal
(Alpk:ApfCD, C57BL10J) mice. J Proteome Res. 2006;5(2):378–84.
130. Holmes E, Nicholson J. Variation in gut microbiota strongly influences individual rodent
phenotypes. Toxicol Sci. 2005;87(1):1–2.
131. Martin FP, Wang Y, Yap IK, Sprenger N, Lindon JC, Rezzi S, et al. Topographical variation in
murine intestinal metabolic profiles in relation to microbiome speciation and functional eco-
logical activity. J Proteome Res. 2009;8(7):3464–74.
132. Martin FP, Montoliu I, Kochhar S, Rezzi S. Chemometric strategy for modeling metabolic
biological space along the gastrointestinal tract and assessing microbial influences. Anal
Chem. 2010;82(23):9803–11.
133. Wang Y, Holmes E, Comelli EM, Fotopoulos G, Dorta G, Tang H, et al. Topographical varia-
tion in metabolic signatures of human gastrointestinal biopsies revealed by high-resolution
magic-angle spinning (1)H NMR spectroscopy. J Proteome Res. 2007;6(10):3944–51.
134. Wang Y, Tang H, Holmes E, Lindon JC, Turini ME, Sprenger N, et al. Biochemical character-
ization of rat intestine development using high-resolution magic-angle-spinning 1H NMR
spectroscopy and multivariate data analysis. J Proteome Res. 2005;4(4):1324–9.
135. Martin FP, Wang Y, Sprenger N, Holmes E, Lindon JC, Kochhar S, et al. Effects of probiotic
Lactobacillus paracasei treatment on the host gut tissue metabolic profiles probed via magic-
angle-spinning NMR spectroscopy. J Proteome Res. 2007;6(4):1471–81.
136. Alonso C, Guilarte M, Vicario M, Ramos L, Rezzi S, Martinez C, et al. Acute experimental
stress evokes a differential gender-determined increase in human intestinal macromolecular
permeability. Neurogastroenterol Motil. 2012;24(8):740-e349.
137. Martin FP, Sprenger N, Yap IK, Wang Y, Bibiloni R, Rochat F, et al. Panorganismal gut
microbiome-host metabolic crosstalk. J Proteome Res. 2009;8(4):2090–105.
138. Martin FP, Sprenger N, Montoliu I, Rezzi S, Kochhar S, Nicholson JK. Dietary modulation
of gut functional ecology studied by fecal metabonomics. J Proteome Res. 2010;9(10):
5284–95.
139. Tuohy KM, Gougoulias C, Shen Q, Walton G, Fava F, Ramnani P. Studying the human gut
microbiota in the trans-omics era–focus on metagenomics and metabonomics. Curr Pharm
Des. 2009;15(13):1415–27.
140. Cummings JH, Macfarlane GT. The control and consequences of bacterial fermentation in
the human colon. J Appl Bacteriol. 1991;70(6):443–59.
141. Woodmansey EJ, McMurdo ME, Macfarlane GT, Macfarlane S. Comparison of composi-
tions and metabolic activities of fecal microbiotas in young adults and in antibiotic-treated
and non-antibiotic-treated elderly subjects. Appl Environ Microbiol. 2004;70(10):6113–22.
142. Coldham NG, Darby C, Hows M, King LJ, Zhang AQ, Sauer MJ. Comparative metabolism
of genistin by human and rat gut microflora: detection and identification of the end-products
of metabolism. Xenobiotica. 2002;32(1):45–62.
143. Gonthier MP, Verny MA, Besson C, Remesy C, Scalbert A. Chlorogenic acid bioavailability
largely depends on its metabolism by the gut microflora in rats. J Nutr. 2003;133(6):
1853–9.
144. Saric J, Wang Y, Li J, Coen M, Utzinger J, Marchesi JR, et al. Species variation in the fecal
metabolome gives insight into differential gastrointestinal function. J Proteome Res.
2008;7(1):352–60.
145. Beckonert O, Coen M, Keun HC, Wang Y, Ebbels TM, Holmes E, et al. High-resolution
magic-angle-spinning NMR spectroscopy for metabolic profiling of intact tissues. Nat
Protoc. 2010;5(6):1019–32.
146. Wilson ID, Nicholson JK. The role of gut microbiota in drug response. Curr Pharm Des.
2009;15(13):1519–23.
147. Wei H, Dong L, Wang T, Zhang M, Hua W, Zhang C, et al. Structural shifts of gut microbiota
as surrogate endpoints for monitoring host health changes induced by carcinogen exposure.
FEMS Microbiol Ecol. 2010;73(3):577–86.
148. Staggers JE, Frost SC, Wells MA. Studies on fat digestion, absorption, and transport in the
suckling rat. III. Composition of bile and evidence for enterohepatic circulation of bile salts.
J Lipid Res. 1982;23(8):1143–51.
149. Claus SP, Ellero SL, Berger B, Krause L, Bruttin A, Molina J, et al. Colonization-induced
host-gut microbial metabolic interaction. MBio. 2011;2(2):e00271–10.
150. Mestdagh R, Dumas ME, Rezzi S, Kochhar S, Holmes E, Claus SP, et al. Gut microbiota
modulate the metabolism of brown adipose tissue. J Proteome Res. 2011;11(2):620–30.
151. Merrifield CA, Lewis M, Claus SP, Beckonert OP, Dumas ME, Duncker S, et al. A metabolic
system-wide characterisation of the pig: a model for human physiology. Mol Biosyst.
2011;7(9):2577–88.
152. Wikoff WR, Anfora AT, Liu J, Schultz PG, Lesley SA, Peters EC, et al. Metabolomics analy-
sis reveals large effects of gut microflora on mammalian blood metabolites. Proc Natl Acad
Sci U S A. 2009;106(10):3698–703.
153. Claus SP, Tsang TM, Wang Y, Cloarec O, Skordi E, Martin FP, et al. Systemic multicompart-
mental effects of the gut microbiome on mouse metabolic phenotypes. Mol Syst Biol.
2008;4:219.
154. Alonso C, Guilarte M, Vicario M, Ramos L, Ramadan Z, Antolin M, et al. Maladaptive intes-
tinal epithelial responses to life stress may predispose healthy women to gut mucosal inflam-
mation. Gastroenterology. 2008;135(1):163–72.
155. Rezzi S, Ramadan Z, Martin FP, Fay LB, Bladeren PV, Lindon JC, et al. Human metabolic
phenotypes link directly to specific dietary preferences in healthy individuals. J Proteome
Res. 2007;6(11):4469–77.
156. Stella C, Beckwith-Hall B, Cloarec O, Holmes E, Lindon JC, Powell J, et al. Susceptibility of
human metabolic phenotypes to dietary modulation. J Proteome Res. 2006;5(10):2780–8.
157. Guarner F, Malagelada JR. Gut flora in health and disease. Lancet. 2003;361(9356):512–9.
158. Madden JA, Hunter JO. A review of the role of the gut microflora in irritable bowel syndrome
and the effects of probiotics. Br J Nutr. 2002;88 Suppl 1:S67–72.
159. Bickston SJ, Comerford LW, Cominelli F. Future therapies for inflammatory bowel disease.
Curr Gastroenterol Rep. 2003;5(6):518–23.
160. Vincenti M, Behrends M, Dang K, Park YH, Hirose R, Blasi-Ibanez A, et al. Induction of
intestinal ischemia reperfusion injury by portal vein outflow occlusion in rats. J Gastroenterol.
2010;45(11):1103–10.
161. Bjerrum JT, Nielsen OH, Hao F, Tang H, Nicholson JK, Wang Y, et al. Metabonomics in
ulcerative colitis: diagnostics, biomarker identification, and insight into the pathophysiology.
J Proteome Res. 2010;9(2):954–62.
162. Martin FP, Rezzi S, Philippe D, Tornier L, Messlik A, Holzlwimmer G, et al. Metabolic
assessment of gradual development of moderate experimental colitis in IL-10 deficient mice.
J Proteome Res. 2009;8(5):2376–87.
163. Murdoch TB, Fu H, Macfarlane S, Sydora BC, Fedorak RN, Slupsky CM. Urinary metabolic
profiles of inflammatory bowel disease in interleukin-10 gene-deficient mice. Anal Chem.
2008;80(14):5524–31.
164. Bertini I, Calabro A, De Carli V, Luchinat C, Nepi S, Porfirio B, et al. The metabonomic
signature of celiac disease. J Proteome Res. 2009;8(1):170–7.
165. Tjellstrom B, Stenhammar L, Hogberg L, Falth-Magnusson K, Magnusson KE, Midtvedt T,
et al. Gut microflora associated characteristics in children with celiac disease. Am J
Gastroenterol. 2005;100(12):2784–8.
166. Koeth RA, Wang Z, Levison BS, Buffa JA, Org E, Sheehy BT, et al. Intestinal microbiota
metabolism of l-carnitine, a nutrient in red meat, promotes atherosclerosis. Nat Med.
2013;19(5):576–85.
167. Tang WH, Wang Z, Levison BS, Koeth RA, Britt EB, Fu X, et al. Intestinal microbial metabo-
lism of phosphatidylcholine and cardiovascular risk. N Engl J Med. 2013;368(17):1575–84.
168. Nicholson G, Rantalainen M, Maher AD, Li JV, Malmodin D, Ahmadi KR, et al. Human
metabolic profiles are stably controlled by genetic and environmental variation. Mol Syst
Biol. 2011;7:525.
169. Payne AN, Zihler A, Chassard C, Lacroix C. Advances and perspectives in in vitro human gut
fermentation modeling. Trends Biotechnol. 2012;30(1):17–25.
Chapter 2
Metabonomics in Clinical Practice
Sebastiano Collino, François-Pierre Martin, and Sofia Moco
Abstract Metabonomics is recognized as a powerful top-down system biology

approach to understand genetic-environment-health paradigm paving new avenues
to identify clinically relevant biomarkers. It is nowadays commonly used in clinical
applications shedding new light on physiological regulatory processes of complex
mammalian systems as regards disease etiology, diagnostic stratification, and poten-
tially mechanism of action of therapeutic solutions. It therefore offers opportunities
to associate complex metabolic regulations to the etiology of multifactorial diseases
and metabolic dysfunctions, which may subsequently lead to mechanistic hypotheses
and targets for new nutritional concepts. This review aims at describing recent
applications of metabonomics in clinical fields with insight into diseases diagnostics/
monitoring and improvement of homeostasis metabolic regulation.
Keywords Aging • Biomarkers • Cancer • Metabonomics • Lipidomics • Metabolic

syndrome • Mental health
2.1 Introduction
Nutritional research has shifted from the measurements of a few, but key, physiolog-
ical descriptors to a large-scale screening of molecular processes at different levels
of biological organization, from gene to mRNA, to proteins and enzymes, and to
metabolic pathways [1]. Integration of this biological information is a vital factor
for nutritional research, as population and individual physiological features are not
only reflected in protein concentrations and gene expression but also in metabolite
S. Collino (*) • F.-P. Martin

e-mail: [email protected]; [email protected]
S. Moco
Natural Bioactives and Screening, Nestlé Institute of Health Sciences,

DOI 10.1007/978-1-4471-6539-2_2
26 S. Collino et al.
concentrations and their kinetic changes in cells, tissues, and organs. However,
studies at individual and population scale are often a complex task because biologi-
cal processes are under the influence of numerous intrinsic and extrinsic factors
such as environmental agents, drugs, diet, lifestyle, stress, and microbiome modula-
tions [2, 3].
Recent revolutions in omics technologies are promising today substantial capa-
bility in achieving a global systems view of physiological and pathological pro-
cesses [4]. Their application is driving a rapid shift from the clinic style investigations
to much more large population-based studies, aiming at enhancing our understand-
ing of the role of genetics, environmental factors, and their interactions on individ-
ual susceptibility to disease and health [5, 6]. Therefore, one of the most powerful
strategies for deciphering gene-environment interactions and their role in individual
variability, to pathology and nutritional outcome, lies in the ability to combine data
from different omics technologies, establishing the way for system biology
approaches [7, 8].
2.2 State-of-the-Art Metabonomics and Lipidomics

Technologies
Metabonomics is able to generate multivariate information on a wide range of

molecules and provides the ability to measure subtle changes in biological pro-
cesses as a result of different nutritional effects [2, 9, 10]. Based on the study of
metabolic profiles of biofluids from a complex biological system, over the past
decade, metabonomics appeared rapidly as a powerful technology to diagnose and
identify biomarkers for a medical condition. Metabonomics measures and moni-
tors metabolite concentrations and dynamics in cells, tissues, and multi-
compartmental biological systems, revealing not only the end products of enzyme
expression and activity but as well the ultimate information contained in the genetic
code [11–13]. Because specific physiological states, gene expression, and environ-
mental stressors can cause changes in the steady state of a biological system, moni-
toring the resulting metabolic variations provides a unique insight into intra- and
extra-cellular regulatory processes involved in maintaining homeostasis. Therefore,
the identification of specific metabolic fingerprints vows strong potentials for
nutriceutical and therapeutic surveillance. Nowadays, metabonomics applications
have evolved towards deciphering the cellular and molecular processes in response
to different individual dietary modulations, predicting health and disease out-
comes. By the global study of low molecular weight compounds (<1,500 Da) in
biofluids (plasma/serum and urine), complex biological matrixes, and tissues, it
assures the characterization of individual metabolic phenotypes, or metabotypes.
Metabonomics employs mainly two analytical techniques based on 1H nuclear
magnetic resonance (NMR) and mass spectrometry coupled to gas/high perfor-
mance liquid chromatography (GC-MS and LC-MS) with lately the addition
of ultrahigh performance liquid chromatography systems coupled to mass
2 Metabonomics in Clinical Practice 27
Fig. 2.1 Metabonomics as a tool to probe human metabolism and health. By using NMR- and
MS-based metabonomics, biomarkers of disease, nutrition, and general development can be
assessed in a qualitative and quantitative way
spectrometry (UPLC/MS) (Fig. 2.1). Both methods are then comprehensively used
to generate multivariate datasets which are then employed, in coordination with
advance statistical tools, to recover meaningful biological information from the
complex metabolic profiles.
NMR-based metabonomics provides efficient high-throughput analysis of bio-
logical samples, making it a relatively cost-effective approach. NMR spectroscopy
offers the unique prospect to holistically profile hundreds of metabolites with no a
priori selection. In proton NMR spectroscopy (1H NMR), all covalently attached
protons from mobile molecules within a very high dynamic range of concentrations,
i.e., from millimolar to nanomolar range, are simultaneously scanned, thus provid-
ing a biochemical fingerprint of biological sample. 1H NMR-based metabonomics
is generally preferred to other nuclei-like carbon-13 due to highest sensitivity and
the relative short experimental time needed to acquire metabolic profiles. However,
resonances of metabolites may be highly overlapped within the proton resonance
window. In such case, ultrahigh magnetic field and/or two-dimensional (2D) NMR
spectroscopy can be used to resolve overlapped resonances.
Urine and blood plasma are the most commonly used biofluids for metabonomics
studies due to their intrinsic richness in metabolic information and their relatively
easy and noninvasive access. Detailed procedures to collect, store, and prepare bio-
fluids or tissue samples for NMR analysis have been provided as guidelines for
metabonomics [14]. Urine, serum, and plasma usually require minimal pretreatment
such as the addition of sodium azide to control bacterial growth, phosphate buffer to
control pH-induced shift in resonance, and deuterated water to lock the magnetic field,
TSP (3-(trimethylsilyl)-propionate, sodium salt), and DSS (2,2-dimethyl-2-silapentane-
5-sulfonate, sodium salt) for chemical shift calibration. Recent introduction of
cryoprobes has strongly improved NMR sensitivity to generally a fourfold factor

relatively to conventional room temperature probes [15]. Furthermore, NMR spec-
troscopy, using high-resolution magic angle spinning NMR (HR-MAS) [16], offers
a unique prospect to generate metabolic profiles from intact tissues, ensuring thus
the biological integrity of the investigated sample. One of the key features of NMR
remains definitely the robustness as showed, i.e., reproducibility >98 % [17].
Along with NMR, LC-MS is probably the most widely used technique in meta-
bonomics. Compared to NMR, LC-MS allows the detection of a wide range of
metabolites, reaching higher sensitivity levels. Essentially two different
approaches have been developed: the detection of all instrumentally possible
metabolites and detection of specific classes of metabolites, named targeted and
untargeted methods, respectively. The detection of metabolites by LC-MS is
firstly obtained with optimization of the sample preparation, chromatographic
separation, and ionization. The sample preparation in metabonomics is perhaps
the most underestimated step and is crucial as the first selection of the aimed
metabolite classes is according to their chemical properties. Typically samples are
quenched and then extracted, often by liquid extraction with organic solvents or
solid-phase extraction, so that eventual enzymatic activity is stopped and the con-
servation of the metabolite pool is insured [18, 19]. The integrity and recovery of
the metabonome is an important parameter, as well as repeatability, so that com-
parison between samples is made possible in a robust way. Sample preparation
strategies have been described for biofluids [19, 20]), tissues [21], and cells [22,
23]. Similarly to GC-MS, the use of internal and external standards, such as non-
endogenous compounds or isotopically labeled species, is recommended when-
ever possible for better compensation of liabilities during sample preparation [24,
25]. The ionization of analytes prior to MS analysis is a prerequisite for the detec-
tion of metabolites. Electrospray ionization (ESI) and atmospheric pressure
chemical ionization (APCI), which are chemical ionization techniques, are the
most used in LC-MS. The chemistry involved in ionization is complex and
strongly depends on the characteristics of the solvents and additives (volatility,
surface tension, viscosity, conductivity, ionic strength, dielectric constant, elec-
trolyte concentration, pH, and gas-phase ion-molecule reactions), analyte (acid
dissociation constant, hydrophobicity, surface activity, ion salvation energy, pro-
ton affinity), and operational parameters such as flow rate, temperature, and ESI
voltage [26]. Nordström and co-workers [27] have applied a multiple ionization
strategy, using ESI and APCI, in both positive and negative ionization modes,
matrix-assisted laser desorption ionization (MALDI), and desorption ioniza-
tion on silicon (DIOS) for the analysis of human serum. The combination of
ionization techniques maximizes the coverage of measured metabolite classes.
The variety of hardwares in mass spectrometry offers a wide-range technical solu-
tion for metabonomics. Some of these include quadrupole (Q), time of flight
(TOF), orbitrap, ion trap (IT), ion cyclotron resonance (ICR), and combinations
(such as QQQ, QTOF, TOF-TOF, IT-orbitrap). Typically in targeted methods,
QQQ-MS or QTrap-MS is used, allowing the optimization of parameters for each
selected ion pairs (typically molecular ion and abundant fragment) or transitions
called selected reaction monitoring (SRM) or multiple reactions monitoring
(MRM) can be used for quantitative purposes. IT allows the isolation of ions for
fragmentation up to nth time (with n being the number of collision events), while
TOFs, ICRs, and orbitraps provide very high mass accuracy and resolution which
can be useful for metabolite identification purposes. Plasma has been analyzed by
untargeted GC-TOF-MS [28] leading to the detection of 46 endogenous metabo-
lites. Using a 12T Fourier transform ICR-MS, about 570 distinct metabolite fea-
tures, represented by monoisotopic masses above signal-to-noise ratio 3 within
the mass range m/z 90–570, were detected in murine plasma.
The central carbon metabolism, including glycolysis, pentose phosphate path-
way, tricarboxylic acid cycle (TCA cycle), and surrounding metabolic reactions,
contains mostly polar compounds such as sugars, sugar phosphates, and organic and
amino acids and has been covered by GC-MS after derivatization [28], ion pairing
LC-MS [24], hydrophobic interaction LC-MS [29], and CE-MS [30].
In addition, a comprehensive analysis of biological lipids can be performed
using MS-based lipidomics. A number of analytical approaches can be deployed
to access the lipid inventory of a given biological matrix [31]. Recent advances
in LC-MS technologies make this field a promising area of lipid biochemical
research [32]. Although such approaches are reasonably well established for
high-throughput analysis of the major lipid classes (phospholipids, sphingolip-
ids) [33–36], they still have to be fine-tuned for the quantification of low abun-
dant signaling lipids such as sphingosine, sohingosine-1-phosphate, or
lysophosphatidic acid [37–39]. Standardization of metabonomics practices has
been proposed by the Chemical Analysis Working Group from the metabonomics
standards initiative where a general consensus concerning the minimum report-
ing standards for metabonomics experiments has been described [40]. Such a
position document sets the basis for better intra- and interlaboratory reproduc-
ibility. The comparison between LC-MS urinary profiles obtained in three differ-
ent labs has shown a reproducibility of over 0.95 (coefficient of determination)
[41]. The compliance to quality controls, signal intensity checks, and post-anal-
ysis signal drift corrections is some of the solutions to warrant high-quality
LC-MS datasets [42].
Yet, one of the bottlenecks in metabonomics remains the identification of metab-
olites in particular biological matrices. Traditionally, full identification of molecules
is a laborious procedure, including enrichment, isolation, purification, and full char-
acterization by several analytical strategies, to reach unambiguous identification.
MS offers valuable pieces of information in the identification of metabolites: molec-
ular mass and molecular fragmentation pattern, taken from MS and MS/MS frag-
mentation, respectively. In combination with LC, a relative polarity index can be
obtained, according to the used stationary phase. Often, efficient identification relies
with the use of a combination of analytic techniques such as NMR and MS and
access to biochemical database and computational techniques (data preprocessing,
statistical modeling, data mining) [43–49].
2.3 Metabonomics Flow Chart in Epidemiological Studies
Epidemiological studies aim to substantiate the causes, distribution, and control of

health and diseases in populations. They integrate scientific information derived
from complementary disciplines (e.g., microbiology, genetics) across a large-scale
population to assess etiological hypotheses and provide the basis for developing and
evaluating health management solutions [50]. Typically, epidemiological studies
deal with a large number of subjects and observations, relying on statistics to estab-
lish validity levels and predict risk factors. Metabonomics has been recently used as
an analytical tool in large epidemiological studies assessing several public health
challenges, including cardiovascular disease [51], diabetes [52], schizophrenia [53],
and aging [54]. On one hand, having access to a large number of samples is essential
to validate health or disease trends, at a population level. On the other hand, the
demand in organization, compliance to ethic procedures, bookkeeping, and stan-
dardized sampling procedures is imperative to carry out such a study. Therefore,
this type of studies is undertaken under the control of dedicated study centers, where
rigorous standardized procedures, under safety precautions, can be applied [52].
Epidemiological studies can produce thousands of samples and involve different
centers, where different groups work as a consortium. In order to make sure all
sample information is recorded, and traced, sample management becomes an impor-
tant part of the process, dealing with sample identification, sample labeling, storage
and access to analytical data, sample history, etc. Normally unique identifiers should
be given to each sample (e.g., sample barcodes), and all information should be
stored in established databases [55].
Preferably, samples are collected in dedicated centers under defined protocols
and upon ethical approval. According to the type of sample to collect and analysis
to follow, different procedures are applied [56]. Urine is obtained by noninvasive
means which is an advantage, as subjects are more prone to participate in the study,
and it allows the access to otherwise difficult/impossible studies. Typically, urine
can be collected in random samples (any time of the day), timed samples (at a spe-
cific time in the day or at x hours after a specific intervention), and 24-h samples
(pooling of urine during 24 h). The last allows compensating for large variability
owed to short collection times. Obtaining blood plasma is more invasive, but remains
largely applicable in the pediatric, child, and elderly fields, mainly through techno-
logical development requiring low sample volumes. For instance, successful analy-
ses can be conducted on blood serum, plasma, or even blood spot (collected using
Gutri paper or capillaries in combination to finger prick). The analysis of stool
samples is also widely generalized and provides information on digestive and gut
microbial metabolism, but suffers of the individual acceptance in providing samples
from children and adults. In the context of disease diagnostics requiring biopsies
analyses, metabonomics is increasingly envisioned as a new way to investigate the
physiological integrity of tissue and personalize healthcare. Under these conditions,
the collection of tissue samples will be conducted according to best clinical practice
and ensuring sample biochemical integrity prior to analysis. Finally, metabonomics
analysis is also applicable to food itself to identify major nutritional constituents

and/or particular metabolites showing a significant bioactivity, including milk, cere-
als, wine, fruits, vegetables, and meat.
Integrity of biological samples is of utmost importance when the aim of the analysis
is to obtain a reflex of the metabolism taking place. Alteration of integrity of sam-
ples due to long periods at room temperature is sample dependent and should be
avoided, as potentially leading to an unreal representation of physiological status.
Samples should be therefore stored at −80 °C whenever possible to allow a maxi-
mum storage time [57]. In fact, the cold chain from sample collection to sample
storage should be preserved, especially when home collection is involved. Samples
should be collected in sterile containers and sodium azide should be added to pre-
vent bacterial growth whenever possible. In conclusion, when dealing with a large
number of samples, as in epidemiological studies, planning and automation of pro-
cedures is imperative, including study design, sample collection and storage, meta-
bonomics analysis, data analysis, and storage and retrieval of metadata (Fig. 2.2).
Fig. 2.2 Flowchart for epidemiological metabonomics studies. Planning and automation of
procedures is crucial, including study design, as sample collection and storage, data analysis, and
retrieval of metadata
2.4 Clinical Applications of Metabonomics

for Aging Biomarkers
Decades of research on aging have found hundreds of genes [58, 59] and many
biological processes [60–62] that are associated to the aging process, but up to date,
metabonomics applications covering aging as a multifactorial event are sparse [63].
In particular, the frailty syndrome is increasingly being considered as a risk indica-
tor of adverse health outcomes. Elderly may be also prone to be resistant to anabolic
stimuli which is likely a key factor in the loss of skeletal muscle mass with aging.
Vital to understand these biological processes is the development of biological
markers, through system biology approaches, aiding at strategies for tailored thera-
peutic and personalized nutritional program. The overall aim is to prevent or attenu-
ate decline of key physiological functions required to live an active and independent
life. Therefore, it is crucial to develop core indicators (biomarkers) of health and
function in older adults, where nutrition and tailored personalized programs could
exhibit preventive roles and where the aid of omics technologies is increasingly
displaying potential in revealing key molecular mechanisms/targets linked to spe-
cific aging and/or healthy aging processes (Fig. 2.3). In one of the first aging study,
Lawton et al. analyzed the human plasma metabonome in 269 individuals (both men
and women, 20–65 years old) identifying significant changes in relative concentra-
tions of more than 100 metabolites [64]. Here changes in protein, energy, oxidative
stress, and lipid metabolism were observed with increasing age. In addition, certain
xenobiotics (i.e., caffeine) were higher in older subjects, displaying possibly
decreased hepatic cytochrome P540 activity. Further, Nikkila and colleagues per-
formed a metabonomics study on early childhood, following 59 children from birth
to an age of 4 years old and identified that major developmental state differences
between girls and boys are attributed to sphingolipids metabolism [65]. Moreover,
comparison of longitudinal metabolic trajectories between boys and girls revealed
Fig. 2.3 Selected biomarker applications for aging and clinical research
higher levels of sphingomyelins in girls than in boys. More recently Zhonghao et al.
characterized the metabolic profile of a large group of subjects with a wide age
range (32–81 years) and identified metabolites related to chronological age, inde-
pendent of BMI [54]. Here, two population-based studies were used: a German
aging population as a discovery cohort, with 1,038 female and 1,124 male partici-
pants (32–81 years), and a UK study as replication, with 724 female participants.
Targeted metabonomics of fasting serum samples quantified 131 metabolites.
Among these, 71 out of 34 metabolites were significantly associated with age in
women/men reflecting mostly incomplete mitochondrial fatty acid oxidation
(elevated serum levels of acylcarnitines).
Overall, while the identification of biological markers specific of aging is still on
its infancy, their characterization is crucial in providing insights into mechanisms or
strategies that can prevent or reverse the decline in certain of the affected networks
and as such could extend health span, preventing accelerated aging.
2.5 Clinical Application of Metabonomics

for Biomarkers of Metabolic Syndrome
The continuously increasing prevalence of obesity in many countries around the

world is strongly linked to the projected pandemic of type 2 diabetes (T2D) and its
cardiovascular complications [66, 67]. However, there are many individuals under
the same obesogenic and diabetogenic environments who remain metabolically
healthy. Newgard et al. have studied metabolic, endocrine, inflammatory, and physi-
ologic differences between obese and lean subjects and reported a branched-chain
amino acids (BCAAs)-related metabolic signature contributing to insulin resistance
[68]. Suhre et al. recently reported the outcomes from a multiplatform metabonomics
analysis of an epidemiological study on diabetes in which diabetes-related compli-
cations could be detected already under subclinical conditions in a general German
population [52]. In addition to previously reported T2D biomarkers, including sugar
metabolites, ketone bodies, and BCAA, metabolites resulting from perturbations of
metabolic pathways linked to kidney dysfunction (3-indoxyl sulfate), lipid metabo-
lism (glycerophospholipids, free fatty acids), and bile acid metabolism. Additional
metabonomics investigations suggested that the catabolism of BCAAs was tightly
intertwined with the levels of insulin resistance, while greater levels of BCAAs
were detected in the obese and IR phenotype [69, 70]. Several by-products of BCAA
catabolism, such as glutamate, α-ketoglutarate, propionylcarnitine, and 2-methyl-
butyryl and isovalerylcarnitines, were showing a very strong contribution to the
metabolic signature for obesity and insulin resistance (IR) phenotype. The authors
further tested their hypothesis by supplementing BCAAs in a diet-induced obesity
rat model. However, while having reduced food intake and weight gain, no improve-
ment of the IR levels was detected. Very recently, blood plasma profiling was suc-
cessfully employed to provide predictive markers of the development of diabetes in
prospective human studies [71, 72]. In a first report, five branched-chain and
aromatic amino acids were strongly associated with IR, namely, isoleucine, leucine,
valine, tyrosine, and phenylalanine. The authors demonstrated that a combination of
three amino acids (isoleucine, phenylalanine, tyrosine) could predict future diabetes
(>5-fold higher risk for individuals in top quartile) [72]. Together these key findings
demonstrate a critical role of BCAA metabolism in the early onset of IR and T2D
development. In a second report, specific interrelationships between dyslipidemia
and IR development were evaluated [71]. Interestingly, this work reports how lipids
of lower carbon number and double bond content were associated with an increased
risk of T2D, unlike higher carbon number and double bond content lipids [71]. In
particular, a combination of two triacylglycerols further improved diabetes predic-
tion and could aid in clinical risk assessment. Recently, a lipidomics approach was
applied to reveal the molecular phenotype in prediction of type 1 diabetes (T1D)
[73]. The authors found that T1D progressors are characterized by a distinct cord
blood lipidomic profile which includes reduced major choline-containing phospho-
lipids including sphingomyelins and phosphatidylcholines. A molecular signature
was developed comprising seven lipids which predicted high risk for progression to
T1D, with an odds ratio of 5.94 (95 % confidence interval, 1.07–17.50). Reduction
in choline-containing phospholipids in cord blood is therefore specifically associ-
ated with progression to T1D but not with development of β-cell autoimmunity in
general. Several studies also investigated the interactions between lifestyle,
diet, and metabolic disorders associated with IR. In particular, Huffman
et al. explored the impact of exercise training on insulin sensitivity (IS) in combina-
tion with monitoring of circulating concentrations of metabolic intermediates, hor-
mones, and inflammatory mediators. Improvement in IS was associated with
reduced levels of fatty acid oxidation by-products and increased concentrations in
glycine and proline [74]. Moreover, metabonomics was also employed to decipher
indicators of early onsets of prediabetes status. Zhao et al. investigated the blood
plasma composition in normal and impaired glucose-tolerant populations and dem-
onstrated that prediabetes was associated with alterations in fatty acid, tryptophan,
uric acid, bile acid, and gut microbial metabolism. In parallel, a great amount of
knowledge was also consolidated in the field of T1D, with patients also showing a
variety of metabolic abnormalities including hyperglycemia, ketogenesis, and mus-
cle proteolysis [75]. Lanza et al. analyzed plasma from T1D humans during insulin
treatment and acute insulin deprivation [75] and provided additional evidence on the
disease etiology including protein synthesis and breakdown, gluconeogenesis, keto-
genesis, amino acid oxidation, mitochondrial bioenergetics, and oxidative stress.
There is increasing evidence that the specific metabolic disturbances preceding
β-cell autoimmunity in humans are of relevance for preventive medicine and poten-
tial prognosis of children who subsequently progress to T1D [65, 76, 77]. In a series
of studies, the specificity of the pre-autoimmune metabolic changes was tested both
in non-obese prediabetic mouse models and in prospective human cohorts [65, 76,
77]. Of particular interest is the observation that autoimmune diabetes is preceded
by a state of increased metabolic demands from the islets resulting in elevated insu-
lin secretion and suggest alternative metabolic-related pathways as therapeutic tar-
gets to prevent diabetes.
Nonalcoholic fatty liver disease (NAFLD) is increasingly considered as a main

pathological determinant in various metabolic deregulations such as obesity, insulin
resistance, hypertension, dyslipidemia, and cardiovascular disease (CVD) [78, 79].
NAFLD is characterized by fatty acid infiltration in the liver in the absence of alco-
hol abuse [80]. NAFLD ranges from simple steatosis to nonalcoholic steatohepatitis
(NASH), the latest being marked by increased inflammation status [81, 82]. In the
absence of validated biomarkers of NAFLD alternative to liver biopsy, metabonomics
and lipidomics are foreseen promising to deliver both a new set of minimally inva-
sive clinical classifiers, i.e., biomarkers and metabolic mechanistic insights into the
disease etiology and progression. Recently, Vinaixa et al. reported the use of meta-
bonomics for quantitative profiling of liver extracts from LDLr−/− mice [83]. NMR-
based metabonomics was used to investigate the metabolic effects and implications
of the dietary cholesterol in the etiology of progression from hepatic steatosis to
NASH. Dietary cholesterol increased the hepatic concentrations of cholesterol, tri-
glycerides, and oleic acid but also decreased the polyunsaturated fatty acids
(PUFAs)/monounsaturated fatty acids ratio as well as the relative amount of long-
chain PUFAs in the liver. Changes in hepatic concentration of taurine, glutathione,
methionine, and carnitine were also observed. Likewise, Li et al. used a methionine-
and choline-deficient diet to describe metabolic changes associated to different
stages of NAFLD in male C57BL/6 mice [84]. Four potential biomarkers including
serum glucose, lactate, glutamate/glutamine, and taurine were selected and used to
stratify NAFLD severity. In addition, using a parallel NAFLD animal model/human
design, Barr et al. analyzed 42 serum samples collected from nondiabetic, morbidly
obese, biopsy-proven NAFLD patients and 17 animals belonging to the glycine
N-methyltransferase knockout (GNMT-KO) NAFLD mouse model [85]. MS-based
metabonomics revealed similarities in the GNMT-KO and human NAFLD patients
with relevant biochemical perturbations linked to liver dysfunction through reduced
levels of creatinine and increased concentrations in bile acids as well as in eico-
sanoids. Metabonomics was also employed by Kalhan et al. [86] to provide potential
metabolic steatosis markers in biopsy confirmed NASH subjects. While steatosis and
NASH could not be distinguished, NASH metabolic signature was marked by altered
levels of bile acids, glutathione, lipids, and amino acids. More recently, Feldstein
et al. used a targeted isotope dilution MS-targeted technique to quantify 9- and
13-HODEs and 9- and 13-oxoODEs as circulating biomarkers of NASH [87].
2.6 Clinical Application for Neurological

and Psychiatric Disorders
Diagnostic markers of clinical metabonomics can also find applications in sociopsy-

chological and neurodevelopment disorders. Yap et al. displayed, by the use of
NMR spectroscopy, biochemical signature of autistic individuals [88]. Urinary
metabolic phenotypes of autistic individuals were marked by increased levels
of N-methyl-2-pyridone-5-carboxamide, N-methyl nicotinic acid, N-methyl
nicotinamide, taurine, and a lower concentration of glutamate. Abnormalities in gut

microbiota metabolism were also suggested through lower levels of urinary dimeth-
ylamine, hippurate, and phenylacetylglutamine in autistic children.
Early detection, risk assessment, and therapeutic monitoring of Alzheimer’s dis-
ease (AD) were also studied with metabonomics [89]. Shotgun lipidomics indicated
reductions of sphingomyelin and significant increases in two ceramide species
(N16:0 and N21:0) in plasma of AD patients. Lastly, Oresic et al. reported that
serum metabolic profiles of persons with schizophrenia had significantly higher
metabolite levels in six lipid clusters encompassing saturated triglycerides and in
two small-molecule clusters containing BCAAs, phenylalanine, tyrosine, proline,
glutamate, lactate, and pyruvate [90]. A GC-MS-based metabonomics profiling
approach was used to detect potential biomarkers associated with schizophrenia and
risperidone treatment [91]. Here, 22 marker metabolites provided separation of
schizophrenic patients from matched healthy controls, with citrate, palmitic acid,
myoinositol, and allantoin exhibiting the best combined classification performance.
Moreover, 20 markers displayed the complete separation between posttreatment
and pretreatment patients, with myoinositol, uric acid, and tryptophan showing the
maximum combined classification performance.
A general comprehensive metabonomics population-based study in Finland was
applied [90] to determine metabolic differences between persons included in three
main psychotic disorders (schizophrenia, n = 45; other non-affective psychosis
(ONAP), n = 57; affective psychosis, n = 37) and controls matched by age, sex, and
regions. Here, global lipidomics displayed that compared to healthy controls, those
with schizophrenia had significantly higher metabolite levels in six lipid clusters
containing mainly saturated triglycerides. In addition, a combined GC metabonom-
ics approach revealed, in persons with schizophrenia, two small-molecule clusters
containing, among other metabolites, branched-chain amino acids, phenylalanine
and tyrosine, and proline, glutamic, lactic, and pyruvic acids. Among these, serum
glutamic acid was elevated in all psychoses (p = 0.0020) compared to controls, while
proline upregulation (p = 0.000023) was specific to schizophrenia.
2.7 Clinical Applications for Cancer Diagnosis
Metabonomics is nowadays foreseen as a promising high-throughput, automated

approach in addition to functional genomics and proteomics for analyses of molecu-
lar changes in malignant tumors [92–95]. Metabolite profiling approach was, for
instance, successfully employed to characterize molecular changes in ovarian tumor
tissues [95]. Sixty-six invasive ovarian carcinomas and nine borderline tumors of
the ovary were analyzed by GC-MS. A total of 51 metabolites were significantly
different between borderline tumors and carcinomas, which encompassed glycero-
lipid, pyrimidine, purine, amino acid, propanoate, and free fatty acid metabolism
[95]. In addition, the potential of applying metabonomics to explore metabolic path-
ways modulation specific to organ-confined disease or metastatic disease may lead
to the identification of new early disease biomarkers. MS-based metabonomics
analysis of prostate cancer patients based on tissue biopsies, urine, and plasma
samples was able to distinguish benign prostate, clinically localized prostate cancer,
and metastatic disease [92]. Sarcosine, an N-methyl derivative of the amino acid
glycine, was identified as a differential metabolite that was highly increased during
prostate cancer progression to metastasis and can be detected noninvasively in
urine. Pasikant et al. displayed the potential and validity in the staging, grading, and
diagnosis capabilities of urinary metabonomics in bladder cancer tumors [96]. Here,
100 % sensitivity in detecting bladder cancer was observed using urinary metabo-
nomics versus 33 % sensitivity achieved by urinary cytology, the current standard
for tumor detection and monitoring of recurrence or progression of bladder cancer.
Using plasma-free amino acids profiling, Miyag et al. described metabonomics
applications for lung, gastric, colorectal, breast, and prostate cancer disease diagno-
sis [97]. Cancer patients and controls could be discriminated using multivariate
analysis where significant alterations in plasma-free amino acids profiles were
observed in the disease cancer stage. Interestingly, tryptophan was identified as a
key amino acid associated with cancer progression. New breast cancer diagnostic
measures had also been developed by high-resolution magic angle spinning
(HR-MAS) NMR spectroscopy [98]. This technique provides a means to generate
metabolic profiles of intact tissues. HR-MAS MR spectroscopic studies on breast
tissue biopsies revealed elevated levels of taurine- and choline-containing com-
pounds, especially phosphocholine in the cancer samples. Moreover, metabolic pro-
filing allowed a clinical prediction with 69 % sensitivity and 94 % specificity in a
validation cohort. NMR and MS metabolic profiles were also used to develop a
specific prediction model for early detection of recurrent breast cancer [99], dis-
playing capabilities of metabonomics in providing predictive biomarkers.
Interestingly, 55 % of the patients could be correctly predicted to have recurrence
13 months before the recurrence was clinically diagnosed. A MS-based quantitative
metabonomics method to analyze plasma samples from 55 breast cancer patients
and 25 healthy controls was applied [100 ]. A number of 30 patients and 20
age-matched healthy controls were used as a training dataset to establish a diagnos-
tic model and to identify potential biomarkers. Here, 39 differentiating metabolites
were identified, including significantly lower levels of lysophosphatidylcholines
and higher levels of sphingomyelins in the plasma samples obtained from breast
cancer patients compared with healthy controls. Using logical regression, a diag-
nostic equation based on three metabolites (LPC16:0, PCae 42:5, and PCaa 34:2)
successfully differentiated breast cancer patients from healthy controls, with a sen-
sitivity of 98.1 % and a specificity of 96.0 %.
Recent technological advances facilitate automated analyses of biological
samples, and installations of NMR equipment in close proximity to the surgical
theaters are in a growing phase. Metabolic profiling thus has the potential to
become a method for rapid characterization of cancerous biopsies in the operation
theater. Bathen et al. analyzed 328 tissue samples from 228 breast cancer patients
using solid-state NMR [101]. Using double cross validation, high sensitivity, and
specificity of 91 % and 93 %, respectively, was achieved. Analysis of the loading
profiles from both principal component analysis (PCA) and PLS-DA showed
the choline-containing metabolites as main biomarkers for tumor content, with
phosphocholine being especially high in tumor tissue. Other indicative metabolites

include glycine, taurine, and glucose.
Being independent of prior assumptions, metabonomics approaches also allow
hypothesis generation on how nutritional intervention might be beneficial to malig-
nant cancers. NMR-based metabonomics was used to determine the effects of a diet
rich in whole grain (WG) rye products on the profile of metabolites in the plasma of
prostate cancer (PC) patients [102]. Seventeen PC patients received 485 g rye bran
product (RP) or refined white wheat product (WP) in a randomized, controlled,
crossover design during a period of 6 week with a 2-week washout period.
Metabonomics analysis of plasma showed an increase in 3-hydroxybutyric acid,
acetone, betaine, N,N-dimethylglycine, and dimethyl sulfone after RP. Plasma
homocysteine concentration was lower (p = 0.017) and that of leptin tended to be
lower (p = 0.07) after RP intake compared to WP intake.
2.8 Conclusion
It has been estimated that by 2020 chronic disease in developing countries will
account for almost three-quarters of all deaths worldwide with 75 % of death due to
stroke and 70 % of death due to diabetes. In such a context, there is a clear need to
develop new predictive approaches for preventive medicine and prognostic strate-
gies for personalized therapeutic management and monitoring (Fig. 2.3). The devel-
opment of systems biology approaches and the new generation of biomarker patterns
will provide the opportunity to associate complex metabolic regulations with key
biological processes. By opening a direct biochemical window into the metabo-
nome, metabonomics is a unique science perfectly suited for the identification of
biomarkers capable of providing better understanding of the complex metabolic
phenomenon. Metabonomics is then foreseen to deliver in clinical settings a new
generation of endpoints, e.g., biomarkers, to describe healthy and abnormal devel-
opmental metabolic trajectory such as in aging studies. This makes clinical metabo-
nomics a very efficient approach for generation of metabolic patterns for the
comprehensive characterization of metabolic health, the prognosis and diagnostic
of diseases, and the generation of new insights in the understanding of the interac-
tions between diet and metabolism (Table 2.1).
Table 2.1 Overview of metabolomics applications in selected human studies

Metabolomics applications Methods Reference
Aging MS [64, 65]
Diabetes MS [52, 68, 71–77]
Insulin resistance MS [69, 70]
Nonalcoholic fatty liver MS, NMR [83–87]
disease (NAFLD)
Autism NMR [88–91]
Cancer-ovarian tumor MS, NMR [92, 95–102]
References
1. Tiret L. Gene-environment interaction: a central concept in multifactorial diseases. Proc Nutr

Soc. 2002;61(4):457–63.
2. Nicholson JK, Wilson ID. Opinion: understanding ‘global’ systems biology: metabonomics
and the continuum of metabolism. Nat Rev Drug Discov. 2003;2(8):668–76.
3. Nicholson JK, Lindon JC, Holmes E. ‘Metabonomics’: understanding the metabolic
responses of living systems to pathophysiological stimuli via multivariate statistical analysis
of biological NMR spectroscopic data. Xenobiotica. 1999;29(11):1181–9.
4. Trujillo E, Davis C, Milner J. Nutrigenomics, proteomics, metabolomics, and the practice of
dietetics. J Am Diet Assoc. 2006;106(3):403–13.
5. Ordovas JM, Shen J. Gene-environment interactions and susceptibility to metabolic syn-
drome and other chronic diseases. J Periodontol. 2008;79(8 Suppl):1508–13.
6. Ordovas JM. Integrating environment and disease into “omic” analysis. Rev Esp Cardiol.
2009;62 Suppl 2:17–22.
7. van der Greef J, Hankemeier T, McBurney RN. Metabolomics-based systems biology and
personalized medicine: moving towards n = 1 clinical trials? Pharmacogenomics. 2006;7(7):
1087–94.
8. van der Greef J. Systems biology, connectivity and the future of medicine. Syst Biol
(Stevenage). 2005;152(4):174–8.
9. Nicholson JK, Lindon JC. Systems biology: metabonomics. Nature. 2008;455(7216):
1054–6.
10. Holmes E, Wilson ID, Nicholson JK. Metabolic phenotyping in health and disease. Cell.
2008;134(5):714–7.
11. Keurentjes JJ. Genetical metabolomics: closing in on phenotypes. Curr Opin Plant Biol.
2009;12(2):223–30.
12. Saito K, Matsuda F. Metabolomics for functional genomics, systems biology, and
biotechnology. Annu Rev Plant Biol. 2010;61:463–89.
13. Weckwerth W. Metabolomics in systems biology. Annu Rev Plant Biol. 2003;54:669–89.
14. Beckonert O, Keun HC, Ebbels TMD, Bundy J, Holmes E, Lindon JC, Nicholson
JK. Metabolic profiling, metabolomic and metabonomic procedures for NMR spectroscopy
of urine, plasma, serum and tissue extracts. Nat Protocol. 2007;2(11):2692–703.
15. Keun HC, Beckonert O, Griffin JL, Richter C, Moskau D, Lindon JC, Nicholson JK. Cryogenic
probe 13C NMR spectroscopy of urine for metabonomic studies. Anal Chem.
2002;74(17):4588–93.
16. Beckonert O, Coen M, Keun HC, Wang Y, Ebbels TM, Holmes E, Lindon JC, Nicholson
JK. High-resolution magic-angle-spinning NMR spectroscopy for metabolic profiling of
intact tissues. Nat Protoc. 2010;5(6):1019–32.
17. Dumas ME, Maibaum EC, Teague C, Ueshima H, Zhou B, Lindon JC, Nicholson JK, Stamler
J, Elliott P, Chan Q, Holmes E. Assessment of analytical reproducibility of 1H NMR spec-
troscopy based metabonomics for large-scale epidemiological research: the INTERMAP
Study. Anal Chem. 2006;78(7):2199–208.
18. Álvarez-Sánchez B, Priego-Capote F, Castro MDLD. Metabolomics analysis II. Preparation
of biological samples prior to detection. TrAC Trends Anal Chem. 2010;29(2):120–7.
19. Bojko B, Cudjoe E, Pawliszyn J, Wasowicz M. Solid-phase microextraction. How far are we
from clinical practice? TrAC Trends Anal Chem. 2011;30(9):1505–12.
20. Bruce SJ, Tavazzi I, Parisod V, Rezzi S, Kochhar S, Guy PA. Investigation of human blood
plasma sample preparation for performing metabolomics using ultrahigh performance liquid
chromatography/mass spectrometry. Anal Chem. 2009;81(9):3285–96.
21. Römisch-Margl W, Prehn C, Bogumil R, Röhring C, Suhre K, Adamski J. Procedure for tissue
sample preparation and metabolite extraction for high-throughput targeted metabolomics.
Metabolomics. 2011;8(1):133–42.
22. Sellick CA, Hansen R, Maqsood AR BW, Stephens GM, Goodacre R, Dickson AJ, Dunn
WB. Effective quenching processes for physiologically valid metabolite profiling of suspen-
sion cultured mammalian cells effective quenching processes for physiologically valid
metabolite profiling of suspension cultured mammalian cells. Anal Chem. 2009;81:174–83.
23. Villas-Bôas SG, Højer-Pedersen J, Åkesson M, Smedsgaard J, Nielsen J. Global metabolite
analysis of yeast: evaluation of sample preparation methods. Yeast. 2005;22(14):1155–69.
24. Buescher JM, Moco S, Sauer U, Zamboni N. Ultrahigh performance liquid chromatography-
tandem mass spectrometry method for fast and robust quantification of anionic and aromatic
metabolites. Anal Chem. 2010;82(11):4403–12.
25. Wu L, Mashego MR, Van Dam JC, Proell AM, Vinke JL, Ras C, Van Winden WA, Van Gulik
WM, Heijnen JJ. Quantitative analysis of the microbial metabolome by isotope dilution mass
spectrometry using uniformly 13C-labeled cell extracts as internal standards. Anal Biochem.
2005;336(2):164–71.
26. Kostiainen R, Kauppila TJ. Effect of eluent on the ionization process in liquid chromatography
mass spectrometry. J Chromatogr A. 2009;1216(4):685–99.
27. Nordström A, Want E, Northen T, Lehtiö J, Siuzdak G. Multiple ionization mass spectrometry
strategy used to reveal the complexity of metabolomics. Anal Chem. 2008;80(2):421–9.
28. Bruce SJ, Breton I, Decombaz J, Boesch C, Scheurer E, Montoliu I, Rezzi S, Kochhar S, Guy
PA. A plasma global metabolic profiling approach applied to an exercise study monitoring the
effects of glucose, galactose and fructose drinks during post-exercise recovery. J Chromatogr
B. 2010;878(29):3015–23.
29. Tolstikov VV, Fiehn O. Analysis of highly polar compounds of plant origin: combination of
hydrophilic interaction chromatography and electrospray ion trap mass spectrometry. Anal
Biochem. 2002;301(2):298–307.
30. Soga T, Igarashi K, Ito C, Mizobuchi K, Zimmermann HP, Tomita M. Metabolomic profiling
of anionic metabolites by capillary electrophoresis mass spectrometry. Anal Chem.
2009;81(15):6165–74.
31. Griffiths WJ, Ogundare M, Williams CM, Wang Y. On the future of “omics”: lipidomics. J
Inherit Metab Dis. 2011;34(3):583–92.
32. Wenk MR. The emerging field of lipidomics. Nat Rev Drug Discov. 2005;4(7):594–610.
33. Liebisch G, Binder M, Schifferer R, Langmann T, Schulz B, Schmitz G. High throughput
quantification of cholesterol and cholesteryl ester by electrospray ionization tandem mass
spectrometry (ESI-MS/MS). Biochim Biophys Acta. 2006;1761(1):121–8.
34. Liebisch G, Drobnik W, Reil M, Trumbach B, Arnecke R, Olgemoller B, Roscher A, Schmitz
G. Quantitative measurement of different ceramide species from crude cellular extracts by
electrospray ionization tandem mass spectrometry (ESI-MS/MS). J Lipid Res.
1999;40(8):1539–46.
35. Liebisch G, Lieser B, Rathenberg J, Drobnik W, Schmitz G. High-throughput quantification
of phosphatidylcholine and sphingomyelin by electrospray ionization tandem mass spec-
trometry coupled with isotope correction algorithm. Biochim Biophys Acta.
2004;1686(1–2):108–17.
36. Schuhmann K, Herzog R, Schwudke D, Metelmann-Strupat W, Bornstein SR, Shevchenko
A. Bottom-up shotgun lipidomics by higher energy collisional dissociation on LTQ Orbitrap
mass spectrometers. Anal Chem. 2011;83(14):5480–7.
37. Scherer M, Gnewuch C, Schmitz G, Liebisch G. Rapid quantification of bile acids and their
conjugates in serum by liquid chromatography-tandem mass spectrometry. J Chromatogr B
Analyt Technol Biomed Life Sci. 2009;877(30):3920–5.
38. Scherer M, Leuthauser-Jaschinski K, Ecker J, Schmitz G, Liebisch G. A rapid and quantita-
tive LC-MS/MS method to profile sphingolipids. J Lipid Res. 2010;51(7):2001–11.
39. Liebisch G, Schmitz G. Quantification of lysophosphatidylcholine species by high-throughput
electrospray ionization tandem mass spectrometry (ESI-MS/MS). Methods Mol Biol.
2009;580:29–37.
40. Sumner L, Amberg A, Barrett D, Beale M, Beger R, Daykin C, Fan T, Fiehn O, Goodacre R,
Griffin J, Hankemeier T, Hardy N, Harnly J, Higashi R, Kopka J, Lane A, Lindon J, Marriott P,
Nicholls A, Reily M, Thaden J, Viant M. Proposed minimum reporting standards for chemical
analysis. Metabolomics. 2007;3(3):211–21.
41. Benton HP, Want E, Keun HC, Amberg A, Plumb RS, Goldfain-Blanc F, Walther B, Reily
MD, Lindon JC, Holmes E, Nicholson JK, Ebbels TM. Intra- and interlaboratory reproduc-
ibility of ultra performance liquid chromatography-time-of-flight mass spectrometry for uri-
nary metabolic profiling. Anal Chem. 2012;84(5):2424–32.
42. Kamleh MA, Ebbels TM, Spagou K, Masson P, Want EJ. Optimizing the use of quality control
samples for signal drift correction in large-scale urine metabolic profiling studies. Anal
Chem. 2012;84:2670–7.
43. Brown M, Dunn WB, Dobson P, Patel Y, Winder CL, Francis-McIntyre S, Begley P, Carroll
K, Broadhurst D, Tseng A, Swainston N, Spasic I, Goodacre R, Kell DB. Mass spectrometry
tools and metabolite-specific databases for molecular identification in metabolomics. Analyst.
2009;134(7):1322–32.
44. Exarchou V, Godejohann M, van Beek TA, Gerothanassis IP, Vervoort J. LC-UV-solid-phase
extraction-NMR-MS combined with a cryogenic flow probe and its application to the identi-
fication of compounds present in Greek oregano. Anal Chem. 2003;75(22):6288–94.
45. Lommen A, Gerssen A, Oosterink JE, Kools HJ, Ruiz-Aracama A, Peters RJ, Mol HG. Ultra-
fast searching assists in evaluating sub-ppm mass accuracy enhancement in U-HPLC/
Orbitrap MS data. Metabolomics. 2011;7(1):15–24.
46. Moco S, Bino RJ, De Vos RCH, Vervoort J. Metabolomics technologies and metabolite iden-
tification. Trends Anal Chem. 2007;26:855–66.
47. Moco S, Forshed J, De Vos RCH, Bino RJ, Vervoort J. Intra- and inter-metabolite correlation
spectroscopy of tomato metabolomics data obtained by liquid chromatography-mass spec-
trometry and nuclear magnetic resonance. Metabolomics. 2008;4:202–15.
48. Tautenhahn R, Bottcher C, Neumann S. Highly sensitive feature detection for high resolution
LC/MS. BMC Bioinforma. 2008;9:504.
49. Neumann S, Bocker S. Computational mass spectrometry for metabolomics: identification of
metabolites and small molecules. Anal Bioanal Chem. 2010;398(7–8):2779–88.
50. Coughlin SS. Ethical issues in epidemiologic research and public health practice. Emerg
Themes Epidemiol. 2006;3:16.
51. Holmes E, Loo RL, Stamler J, Bictash M, Yap IK, Chan Q, Ebbels T, De Iorio M, Brown IJ,
Veselkov KA, Daviglus ML, Kesteloot H, Ueshima H, Zhao L, Nicholson JK, Elliott
P. Human metabolic phenotype diversity and its association with diet and blood pressure.
Nature. 2008;453(7193):396–400.
52. Suhre K, Meisinger C, Doring A, Altmaier E, Belcredi P, Gieger C, Chang D, Milburn MV,
Gall WE, Weinberger KM, Mewes HW, Hrabe de Angelis M, Wichmann HE, Kronenberg F,
Adamski J, Illig T. Metabolic footprint of diabetes: a multiplatform metabolomics study in an
epidemiological setting. PLoS One. 2010;5(11):e13953.
53. He Y, Yu Z, Giegling I, Xie L, Hartmann AM, Prehn C, Adamski J, Kahn R, Li Y, Illig T,
Wang-Sattler R, Rujescu D. Schizophrenia shows a unique metabolomics signature in
plasma. Transl Psychiatry. 2012;2:e149.
54. Yu Z, Zhai G, Singmann P, He Y, Xu T, Prehn C, Römisch-Margl W, Lattka E, Gieger C,
Soranzo N, Heinrich J, Standl M, Thiering E, Mittelstraß K, Wichmann H-E, Peters A, Suhre
K, Li Y, Adamski J, Spector TD, Illig T, Wang-Sattler R. Human serum metabolic profiles are
age dependent. Aging Cell. 2012;11(6):960–7.
55. Haquin S, Oeuillet E, Pajon A, Harris M, Jones A, Tilbeurgh H, Markley J, Zolnai Z, Poupon
A. Data management in structural genomics: an overview. In: Kobe B, Guss M, Huber T, editors.
Structural proteomics, vol 426. Methods in Molecular Biology™. Humana Press; 2008.
pp 49 – 79. doi:10.1007/978-1-60327-058-8_4.
56. Holland NT, Smith MT, Eskenazi B, Bastaki M. Biological sample collection and processing
for molecular epidemiological studies. Mutat Res. 2003;543(3):217–34.
57. Lauridsen M, Hansen SH, Jaroszewski JW, Cornett C. Human urine as test material in 1H
NMR-based metabonomics: recommendations for sample preparation and storage. Anal
Chem. 2007;79(3):1181–6.
58. Singh R, Kolvraa S, Rattan SI. Genetics of human longevity with emphasis on the relevance
of HSP70 as candidate genes. Front Biosci. 2007;12:4504–13.
59. De Benedictis G, Carotenuto L, Carrieri G, De Luca M, Falcone E, Rose G, Cavalcanti S,
Corsonello F, Feraco E, Baggio G, Bertolini S, Mari D, Mattace R, Yashin AI, Bonafe M,
Franceschi C. Gene/longevity association studies at four autosomal loci (REN, THO, PARP,
SOD2). Eur J Hum Genet. 1998;6(6):534–41.
60. Schmitt K, Grimm A, Kazmierczak A, Strosznajder JB, Gotz J, Eckert A. Insights into mito-
chondrial dysfunction: aging, amyloid-beta and tau – a deleterious trio. Antioxid Redox
Signal. 2011;16:1456.
61. Castro MD, Suarez E, Kraiselburd E, Isidro A, Paz J, Ferder L, Ayala-Torres S. Aging
increases mitochondrial DNA damage and oxidative stress in liver of rhesus monkeys. Exp
Gerontol. 2011;47:29–32.
62. Radak Z, Zhao Z, Goto S, Koltai E. Age-associated neurodegeneration and oxidative damage
to lipids, proteins and DNA. Mol Asp Med. 2011;32(4–6):305–15.
63. Xue H, Xian B, Dong D, Xia K, Zhu S, Zhang Z, Hou L, Zhang Q, Zhang Y, Han JD.
A modular network model of aging. Mol Syst Biol. 2007;3:147.
64. Lawton KA, Berger A, Mitchell M, Milgram KE, Evans AM, Guo L, Hanson RW, Kalhan
SC, Ryals JA, Milburn MV. Analysis of the adult human plasma metabolome.
Pharmacogenomics. 2008;9(4):383–97.
65. Nikkila J, Sysi-Aho M, Ermolov A, Seppanen-Laakso T, Simell O, Kaski S, Oresic
M. Gender-dependent progression of systemic metabolic states in early childhood. Mol Syst
Biol. 2008;4:197.
66. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, Gordon DJ,
Krauss RM, Savage PJ, Smith Jr SC, Spertus JA, Costa F. Diagnosis and management of the
metabolic syndrome. An American Heart Association/National Heart, Lung, and Blood
Institute Scientific Statement. Executive summary. Cardiol Rev. 2005;13(6):322–7.
67. Wirfalt E, Hedblad B, Gullberg B, Mattisson I, Andren C, Rosander U, Janzon L, Berglund
G. Food patterns and components of the metabolic syndrome in men and women: a cross-
sectional study within the Malmo Diet and Cancer cohort. Am J Epidemiol.
2001;154(12):1150–9.
68. Newgard CB, An J, Bain JR, Muehlbauer MJ, Stevens RD, Lien LF, Haqq AM, Shah SH,
Arlotto M, Slentz CA, Rochon J, Gallup D, Ilkayeva O, Wenner BR, Yancy Jr WS, Eisenson
H, Musante G, Surwit RS, Millington DS, Butler MD, Svetkey LP. A branched-chain amino
acid-related metabolic signature that differentiates obese and lean humans and contributes to
insulin resistance. Cell Metab. 2009;9(4):311–26.
69. Huffman KM, Shah SH, Stevens RD, Bain JR, Muehlbauer M, Slentz CA, Tanner CJ,
Kuchibhatla M, Houmard JA, Newgard CB, Kraus WE. Relationships between circulating
metabolic intermediates and insulin action in overweight to obese, inactive men and women.
Diabetes Care. 2009;32(9):1678–83.
70. Fiehn O, Garvey WT, Newman JW, Lok KH, Hoppel CL, Adams SH. Plasma metabolomic
American women. PLoS ONE. 2010;5(12):e15234.
71. Rhee EP, Cheng S, Larson MG, Walford GA, Lewis GD, McCabe E, Yang E, Farrell L, Fox
CS, O’Donnell CJ, Carr SA, Vasan RS, Florez JC, Clish CB, Wang TJ, Gerszten RE. Lipid
profiling identifies a triacylglycerol signature of insulin resistance and improves diabetes
prediction in humans. J Clin Invest. 2011;121(4):1402–11.
72. Wang TJ, Larson MG, Vasan RS, Cheng S, Rhee EP, McCabe E, Lewis GD, Fox CS, Jacques
PF, Fernandez C, O’Donnell CJ, Carr SA, Mootha VK, Florez JC, Souza A, Melander O,
Clish CB, Gerszten RE. Metabolite profiles and the risk of developing diabetes. Nat Med.
2011;17(4):448–53.
73. Oresic M, Gopalacharyulu P, Mykkanen J, Lietzen N, Makinen M, Nygren H, Simell S,
Simell V, Hyoty H, Veijola R, Ilonen J, Sysi-Aho M, Knip M, Hyotylainen T, Simell O. Cord
serum lipidome in prediction of islet autoimmunity and type 1 diabetes. Diabetes.
2013;62:3268–74.
74. Huffman KM, Slentz CA, Bateman LA, Thompson D, Muehlbauer MJ, Bain JR, Stevens RD,
Wenner BR, Kraus VB, Newgard CB, Kraus WE. Exercise-induced changes in metabolic
intermediates, hormones, and inflammatory markers associated with improvements in insulin
sensitivity. Diabetes Care. 2011;34(1):174–6.
75. Lanza IR, Zhang S, Ward LE, Karakelides H, Raftery D, Nair KS. Quantitative metabolomics
by H-NMR and LC-MS/MS confirms altered metabolic pathways in diabetes. PLoS ONE.
2010;5(5):e10538.
76. Sysi-Aho M, Ermolov A, Gopalacharyulu PV, Tripathi A, Seppanen-Laakso T, Maukonen J,
Mattila I, Ruohonen ST, Vahatalo L, Yetukuri L, Harkonen T, Lindfors E, Nikkila J, Ilonen J,
Simell O, Saarela M, Knip M, Kaski S, Savontaus E, Oresic M. Metabolic regulation in pro-
gression to autoimmune diabetes. PLoS Comput Biol. 2011;7(10):e1002257.
77. Oresic M, Simell S, Sysi-Aho M, Nanto-Salonen K, Seppanen-Laakso T, Parikka V,
Katajamaa M, Hekkala A, Mattila I, Keskinen P, Yetukuri L, Reinikainen A, Lahde J, Suortti
T, Hakalax J, Simell T, Hyoty H, Veijola R, Ilonen J, Lahesmaa R, Knip M, Simell
O. Dysregulation of lipid and amino acid metabolism precedes islet autoimmunity in children
who later progress to type 1 diabetes. J Exp Med. 2008;205(13):2975–84.
78. Fabbrini E, Sullivan S, Klein S. Obesity and nonalcoholic fatty liver disease: biochemical,
metabolic, and clinical implications. Hepatology. 2010;51(2):679–89.
79. Johnson NA, Walton DW, Sachinwalla T, Thompson CH, Smith K, Ruell PA, Stannard SR,
George J. Noninvasive assessment of hepatic lipid composition: advancing understanding
and management of fatty liver disorders. Hepatology. 2008;47(5):1513–23.
80. Tiniakos DG, Vos MB, Brunt EM. Nonalcoholic fatty liver disease: pathology and pathogenesis.
Annu Rev Pathol. 2010;5:145–71.
81. James OF, Day CP. Non-alcoholic steatohepatitis (NASH): a disease of emerging identity and
importance. J Hepatol. 1998;29(3):495–501.
82. Day CP, James OF. Steatohepatitis: a tale of two “hits”? Gastroenterology.
1998;114(4):842–5.
83. Rull A, Vinaixa M, Angel RM, Beltran R, Brezmes J, Canellas N, Correig X, Joven
J. Metabolic phenotyping of genetically modified mice: an NMR metabonomic approach.
Biochimie. 2009;91(8):1053–7.
84. Li H, Wang L, Yan X, Liu Q, Yu C, Wei H, Li Y, Zhang X, He F, Jiang Y. A proton nuclear
magnetic resonance metabonomics approach for biomarker discovery in nonalcoholic fatty
liver disease. J Proteome Res. 2011;10(6):2797–806.
85. Barr J, Vazquez-Chantada M, Alonso C, Perez-Cormenzana M, Mayo R, Galan A, Caballeria
J, Martin-Duce A, Tran A, Wagner C, Luka Z, Lu SC, Castro A, Le Marchand-Brustel Y,
Martinez-Chantar ML, Veyrie N, Clement K, Tordjman J, Gual P, Mato JM. Liquid
chromatography-mass spectrometry-based parallel metabolic profiling of human and mouse
model serum reveals putative biomarkers associated with the progression of nonalcoholic
fatty liver disease. J Proteome Res. 2010;9(9):4501–12.
86. Kalhan SC, Guo L, Edmison J, Dasarathy S, McCullough AJ, Hanson RW, Milburn M. Plasma
metabolomic profile in nonalcoholic fatty liver disease. Metabolism. 2011;60(3):404–13.
87. Feldstein AE, Lopez R, Tamimi TA, Yerian L, Chung YM, Berk M, Zhang R, McIntyre TM,
Hazen SL. Mass spectrometric profiling of oxidized lipid products in human nonalcoholic
fatty liver disease and nonalcoholic steatohepatitis. J Lipid Res. 2010;51(10):3046–54.
88. Yap IK, Angley M, Veselkov KA, Holmes E, Lindon JC, Nicholson JK. Urinary metabolic
phenotyping differentiates children with autism from their unaffected siblings and age-
matched controls. J Proteome Res. 2010;9(6):2996–3004.
89. Han X, Rozen S, Boyle SH, Hellegers C, Cheng H, Burke JR, Welsh-Bohmer KA,
Doraiswamy PM, Kaddurah-Daouk R. Metabolomics in early Alzheimer’s disease: identifi-
cation of altered plasma sphingolipidome using shotgun lipidomics. PLoS ONE.
2011;6(7):e21643.
90. Oresic M, Tang J, Seppanen-Laakso T, Mattila I, Saarni SE, Saarni SI, Lonnqvist J, Sysi-Aho
M, Hyotylainen T, Perala J, Suvisaari J. Metabolome in schizophrenia and other psychotic
disorders: a general population-based study. Genome Med. 2011;3(3):19.
91. Xuan J, Pan G, Qiu Y, Yang L, Su M, Liu Y, Chen J, Feng G, Fang Y, Jia W, Xing Q, He
L. Metabolomic profiling to identify potential serum biomarkers for schizophrenia and ris-
peridone action. J Proteome Res. 2011;10(12):5433–43.
92. Sreekumar A, Poisson LM, Rajendiran TM, Khan AP, Cao Q, Yu J, Laxman B, Mehra R,
Lonigro RJ, Li Y, Nyati MK, Ahsan A, Kalyana-Sundaram S, Han B, Cao X, Byun J, Omenn
GS, Ghosh D, Pennathur S, Alexander DC, Berger A, Shuster JR, Wei JT, Varambally S,
Beecher C, Chinnaiyan AM. Metabolomic profiles delineate potential role for sarcosine in
prostate cancer progression. Nature. 2009;457(7231):910–4.
93. Sawyers CL. The cancer biomarker problem. Nature. 2008;452(7187):548–52.
94. Kind T, Tolstikov V, Fiehn O, Weiss RH. A comprehensive urinary metabolomic approach for
identifying kidney cancer. Anal Biochem. 2007;363(2):185–95.
95. Denkert C, Budczies J, Kind T, Weichert W, Tablack P, Sehouli J, Niesporek S, Konsgen D,
Dietel M, Fiehn O. Mass spectrometry-based metabolic profiling reveals different metabolite
patterns in invasive ovarian carcinomas and ovarian borderline tumors. Cancer Res.
2006;66(22):10795–804.
96. Pasikanti KK, Esuvaranathan K, Ho PC, Mahendran R, Kamaraj R, Wu QH, Chiong E, Chan
EC. Noninvasive urinary metabonomic diagnosis of human bladder cancer. J Proteome Res.
2010;9(6):2988–95.
97. Miyagi Y, Higashiyama M, Gochi A, Akaike M, Ishikawa T, Miura T, Saruki N, Bando E,
Kimura H, Imamura F, Moriyama M, Ikeda I, Chiba A, Oshita F, Imaizumi A, Yamamoto H,
Miyano H, Horimoto K, Tochikubo O, Mitsushima T, Yamakado M, Okamoto N. Plasma free
amino acid profiling of five types of cancer patients and its application for early detection.
PLoS ONE. 2011;6(9):e24143.
98. Li M, Song Y, Cho N, Chang JM, Koo HR, Yi A, Kim H, Park S, Moon WK. An HR-MAS
MR metabolomics study on breast tissues obtained with core needle biopsy. PLoS ONE.
2011;6(10):e25563.
99. Gu H, Pan Z, Xi B, Asiago V, Musselman B, Raftery D. Principal component directed partial
least squares analysis for combining nuclear magnetic resonance and mass spectrometry data
in metabolomics: application to the detection of breast cancer. Anal Chim Acta.
2011;686(1–2):57–63.
100. Qiu Y, Zhou B, Su M, Baxter S, Zheng X, Zhao X, Yen Y, Jia W. Mass spectrometry-based
quantitative metabolomics revealed a distinct lipid profile in breast cancer patients. Int J Mol
Sci. 2013;14(4):8047–61.
101. Bathen TF, Geurts B, Sitter B, Fjosne HE, Lundgren S, Buydens LM, Gribbestad IS, Postma
G, Giskeodegard GF. Feasibility of MR metabolomics for immediate analysis of resection
margins during breast cancer surgery. PLoS ONE. 2013;8(4):e61578.
102. Moazzami AA, Zhang JX, Kamal-Eldin A, Aman P, Hallmans G, Johansson JE, Andersson
SO. Nuclear magnetic resonance-based metabolomics enable detection of the effects of a
whole grain rye and rye bran diet on the metabolic profile of plasma in prostate cancer
patients. J Nutr. 2011;141(12):2126–32.
Chapter 3
Adopting Multivariate Nonparametric
Tools to Determine Genotype-Phenotype
Interactions in Health and Disease
Ivan Montoliu
Abstract This chapter describes the role of machine learning approaches such as
random forests in holistic discovery applications and provides a background for its
better understanding. Their suitability for feature selection, data integration, and
network modelling are also evaluated through recent examples in the literature.
These examples cover a variety of fields, ranging from ecology to metabolomics.
Keywords Random forests • Chemometrics • Classification and regression trees

• Data integration • Network modelling • Metabolomics
3.1 M
etabolomics: Introducing the Paradigm Shift
in Data Analysis
The arrival of “-omics” into the scientific scene resulted in a breakthrough for the
data analysis community. The generalization of sequencing methodologies, with the
introduction of gene expression microarray technology, contributed to the genera-
tion of huge datasets covering many aspects related to phenotypic changes present
in biological samples. This avenue facilitated a renovation of the bioinformatics
concept known so far, moving from the general concept of studying information
processes in biological systems to a more complex one: the storing, retrieving, orga-
nization, and analysis of biological data. Elements such as control, system, and
information theory and statistics were widely introduced in the field to tackle new
challenges. In particular, factors such as the dimensionality of datasets and the low
number of available samples hardly challenged the established concepts in data
analysis. For some applications, such as the analysis of differential gene expression
readouts, traditional univariate statistics remained prevalent, being widely used in
I. Montoliu (*)
Analytical Sciences, Applied Mathematics, Nestec Ltd., Nestlé Research Centre,
PO Box 44, Route du Jorat 57, CH-1000 Lausanne 26, Switzerland

DOI 10.1007/978-1-4471-6539-2_3
46 I. Montoliu
large-scale calculations. In this way, they became, by default, the core of data
analysis techniques in the fields of biomarker discovery research. This led to several
issues due to the high number of tests performed and the need to address the statisti-
cal significance of the results. Research in univariate statistics concepts addressed
this point with the introduction of suitable preprocessing methods and concepts
such as false discovery rate and multiplicity testing corrections [1, 2].
Metabolomics [3–5] was a late comer to the “-omics” party. Committed to
address the metabolic changes associated to a specific phenotype due to interven-
tion, environment, or genetic predisposition showed soon the constraints of the
field. The use of advanced analytical chemistry tools, based on MS and NMR tech-
nologies, introduced higher complexity in the data structure and higher covariance
between variables. Moreover, it became a mandatory step the development of
specific tools for preprocessing, in particular for MS holistic approaches [3]. Factors
such as analysis time and cost per sample contributed hardly to keep the issue of the
curse of dimensionality (low n-to-p ratio), also present in differential gene expres-
sion analyses. Furthermore, the indeterminacy on the p-value, strongly dependent
on the preprocessing used, questioned the suitability of the established analysis
methodologies standardly used in differential gene expression.
Other intrinsic biogenic elements, such as reduced fold changes, created in many
cases a more complex framework for biomarker discovery. Being at the lower level
of the biochemical expression chain (genome > proteome > metabolome), the inter-
individual variability was expected to be high and dependent on cofactors such as
the environment. In fact, metabolomics seems to combine environmental and
genetic variability, as partially showed by the genetic origin of metabolic respon-
siveness of human subjects challenged to an intervention [6].
All these elements, data and biology driven, soon faced serious hurdles in
providing relevant results using a standard approach rooted in univariate analysis
plus multiplicity testing correction. In a univariate approach, chances were that the
low number of objects, associated to low fold of changes, would be the source of an
eventual high proportion of type II errors (false negatives). As a counterpart, the use
of multivariate analysis started providing a summarized view of the simultaneous
changes throughout the experiment, disregarding its univariate significance. These
characteristics did not help in biomarker discovery by itself, but contributed
positively to the understanding of coordinated biochemical changes in complex
biological systems.
Chemometrics provide a descriptive or predictive assessment of chemical sys-
tems in experimental life sciences using data analysis techniques [7]. It is focused
on the analysis of analytical data, and it is rooted in an ensemble of tools coming
from different fields such as multivariate statistics, applied mathematics, and com-
puter science. With all those values at hand, chemometrics has become the preferred
partner for analyzing metabolomics data.
From the set of techniques driven by a pure chemometrics approach, principal
component analysis (PCA) [8], partial least squares regression (PLS) [9, 10], and
their derivates, such as orthogonal projection on latent structures (OPLS) [11, 12],
soon became the main workhorses of metabolomics data analysis. These are both
3 Adopting Multivariate Nonparametric Tools to Determine Genotype-Phenotype… 47
soft modelling methods which perform well in low n, high p datasets through the
projection of multivariate data onto a reduced subspace, even with highly covarying
variables. Moreover, in spite of being linear, PLS models handle acceptably well
soft nonlinearities. However, because both methods are designed with predictive
purposes, a strict validation process is needed to assess its generalization ability
(minimize overfit risk) [13].
Most of the metabolomics studies where the PLS methodology was applied
were typical case-control studies. In this case, classifier variants based both on PLS
and OPLS were used, converting both regression models into classifiers through
binary class encoding. From this moment on, partial least squares discriminant
analysis (PLS-DA) [14] and orthogonal projection on latent structures discriminant
analysis (OPLS-DA) [15] became preferred tools for most of the metabolomics
studies handling spectroscopic data [16]. To better cover metabolomics goals, both
regression methods were supported with interesting contributions that improved
somehow the simplicity of interpretation of its parameters for feature extraction
[17]. Both PLS-related techniques were recognized as good performers in the
metabolomics area, and its use was generalized to other types of data (semiquanti-
tative and MS data). This generalization was also supported by the availability of
chemometrics data analysis packages that implemented, in a convenient way, a
variety of algorithms including the most common data pretreatments.
One of the key points of the success of PLS is its utility in providing feature
highlighting through the analysis of its parameters. This strong asset imposes
constraints when selecting alternatives to this model. Even if it highlights metabolic
profiles responding to the design, several aspects, often overlooked, need to be
recall in PLS analysis: PLS has an assumption on the variable distribution and
linearity of the model and needs of a careful validation.
3.2 Expanding the View: Machine Learning
In the last years, an increasing number of new classification algorithms have been
proposed, many of them focused on solving low n-to-p ratio issues. It is hard to
determine which of them have the best performance, and on which conditions. This
difficulty in the assessment hinders a lack of consensus for the best one, being likely
that the best classifier/regressor does not exist for all conditions. Moreover, chances
are the best methodology might be probably problem specific.
In some cases, proposed models are complex, with several parameters to tune,
and each model needs to be tailored at hand to be more on target. In situations as
nowadays, when huge amounts of data need of feature selection, there is a strong
request for models that use few parameters, minimal human input and low compu-
tational cost with simple parallelization.
In search of more suitable algorithms for data analysis and interpretation in
metabolomics, machine learning provides a wealth of tools highly performing for
data representation and generalization (classification and regression). Despite most
48 I. Montoliu
of them offer properties of interest, not all provide relevant information for feature
extraction. Support vector machines (SVM) [18] and multilayer perceptrons
(ANN-MLP) [19] are two examples, where the use of kernels and weight connec-
tion layers removes any traceability of the role of the individual variables in the
model. There have been other alternatives proposed to find this variable relevance,
but they are mostly linked to the application of recursive feature elimination patterns
[20] that make the procedure computer intensive for long datasets.
3.3 D
ata Structure and Inference: Classification
and Regression Trees (CART)
When dealing with metabolomics multivariate datasets (typically low number of

objects, high intragroup variance, unknown individual variable distribution), the use
of decision tree learning appears to be a good option to tackle these issues, still
keeping some of the advantages of bilinear methods such PLS. Thus, this family of
methods can describe the structure of multivariate datasets while providing predic-
tive models of the outputs given. They are built following certain simple rules that
create multiple linear boundaries in the multivariate space. According to their main
goal, they are known as classification (categorical output) and regression (quantita-
tive output) trees [21].
Decision tree learning [22, 23] has several advantages and limitations. Trees are
simple to interpret, and they need simple or no preprocessing, handle well both
numerical and categorical data, are possible to validate (act as predictors of external
objects), are robust, and perform well with large datasets. In addition, CARTs are
able to handle missing attributes. These methods offer also advantages in visualiz-
ing the structure of data: how clusters are built and the distribution of the samples
within each cluster. Different data types: categorical, (un-)ordered, and continuous,
can be handled and can be related nonlinearly to a response.
Their main limitation comes from how the decisions are taken at each branch of
the tree, being usually locally optimal. This point does not warrant the global opti-
mality of the tree. Due to its greedy nature during growth, they are also very depen-
dent on the composition of the training set, which makes them naturally instable.
Moreover, they have the risk of overcomplexity in defining the boundaries, even for
easy discrimination cases. In such cases, using linear bound classifiers can lead to
better performance. At the end, a “pruning” step is needed, where the generalization
error is minimized to decrease the risk of overfit.
Briefly, CART algorithm splits the overall group of subjects (so-called main
node) in several groups or nodes according to the variable that minimizes a prese-
lected criterion of impurity, which in turn accounts for degree of matching with the
target pattern. In principle, each split can be performed in n sub-nodes, but the most
frequent are binary splits due to its easiness of computation when optimizing a tree.
At each node, there is an impurity factor that needs to be calculated for each variable
xi to perform the right selection.
For numerical data, this step is relatively easy, as hyperplane decision boundaries
can be calculated in the form: Is xi ≤ xi split?. The driving criteria throughout all the
tree buildup is to keep the model as sparse (with fewer nodes) as possible. With this
purpose, at each node CART looks out for the query that makes the subsequent
descending node as pure as possible (minimize impurity).
The calculation of node impurity in regression trees is basically driven by the
mean standard error (MSE). Alternatively, in classification of trees, several mea-
sures are available:
• Entropy impurity (also called information impurity), which is defined through
the following expression: I(N) = − ∑ jp(ωj)log2 p(ωj), where p(ωj) accounts for
fraction of patterns at node N that are in category ωj. If all patterns are the same
(i.e., the node is pure), then I(N) = 0.
• Variance impurity, defined in the two-category case as I(N) = p(ω1)p(ω2), which
gives I(N) = 0 when all patterns belong to the same class, either ω1 or ω2.
• Gini impurity, which consists in the extension of variance impurity to more than
two classes: I(N) = ∑ i ≠ jp(ωi)p(ωj). When compositions between ωi and ωj are
unbalanced, it is necessary to weight I(N) with a weight matrix with γij elements,
thus giving I(N) = Σi ≠ jγijp(ωi)p(ωj).
• Misclassification impurity, measuring the minimum probability that a training
pattern must be misclassified.
As stated above, the achievement of the minimum node impurity is the driver of
the tree optimization. To select the best split, the decrease of impurity after the split
is calculated for each variable, and the variable that maximizes this drop is the one
selected.
From all the impurity measurements available, the Gini index and entropy impu-
rity are often preferred due to its computational simplicity. However, in spite of the
diversity of measures, very often the choice if the impurity functions does not affect
excessively the accuracy of the final classifier. This leaves the stopping criterion and
the pruning method as main drivers of the model accuracy.
The stopping criterion controls predictive performance, limiting the excessive
growth of the tree and thus controlling the overfit risk. It may be determined by
using standard validation procedures such as cross-validation and test set. An alter-
native way to limit the tree growth implies setting up a threshold β, which has the
obvious advantage that all samples can be used for training the classifier. This
approach often leads to an unbalanced tree, where the leaf nodes lay at different
levels. However, finding optimal β is not an easy task, as it has little to say with
model performance. An alternative, simpler in concept, is to set a minimum node
size of a fixed number or percentage of objects. This procedure has the advantage
that it adapts the partition size to the density of objects in that region of the multi-
variate space. Complexity of the tree can also be used as an alternative to stop the
growth of the tree, using a balance between the number of nodes present in the tree
and the uncertainty of the tree on the training data. Alternatively, the splitting can be
stopped using hypothesis testing on the increase of node impurity between the tree
layers (usually a χ2 test).
50 I. Montoliu
Using the stopping criterion directly has important drawbacks. Often it leads to
biased trees, because the biggest decreases of impurity are achieved in the nodes
closer to the root node. Consequently, tree growth stops too early, ignoring further
splits. For this reason, growing full trees and posterior pruning has gained accep-
tance. In this approach, all pairs of leaf (terminal) nodes connected to the same
ancestor (one level above) are considered for elimination. Each pair providing a
small increase in node impurity is selected for elimination. At this point, the ances-
tor becomes a leaf. This is basically the inverse of the splitting process described
previously, and commonly delivers unbalanced trees. Its main drawback is that it is
computationally intensive, which in turn limits its application in very big datasets.
3.4 E
nsemble Classifiers: From Single Trees
to Random Forests
If they are grown deep enough, trees are learners with low bias and high variance
that reflect the data structure quite well. These characteristics make them suitable
for performance improvement using general methodologies such as bootstrap
aggregation (bagging) [24]. The main idea of bagging is to reduce the variance of
the prediction through averaging (regression) or aggregated voting (classification)
among several classifiers. This is achieved through the average of many noisy (and
approximately unbiased) models. Models are built taking a bootstrap sample with
replacement of the hold out data. This strategy gives a trade-off solution, as this
averaging provides a decrease in the variance, but with a little increase in bias.
Another positive effect of aggregation and majority voting is the decrease in chances
of overfit.
Targets of boosting [25] are weak learning algorithms (as trees). The approach
generates m different individual classifiers to create an ensemble classifier G(X).
In such approach, weight is given to the individual classifiers Gm(x) according to
their accuracy. This weighting is used to provide a final result, weighting accord-
M
ingly the output of the different classifiers G(X) = sign[Σ m = 1αmGm(x)]. In this way,
misclassified observations are scaled by an exponential factor that increases the
importance of this pattern in the next Gm + 1(x) model. One popular version of boosting,
with good performance, is ADABOOST [26].
The concept of ensemble classifiers is in the deep roots of the random forests
(RF) algorithm [27]. Random forests are an ensemble of classification (regression)
trees that are trained using bootstrapped (with replacement) samples of the training
data. Trees are fully grown and posteriorly pruned to a certain node level (specified)
and the remaining patterns (“out-of-bag,” OOB samples) are predicted after passing
them through the whole ensemble (forest). As in bagged classifiers, majority voting
is used to assign the target class. When used in regression, averaging is used instead.
One of the novel points introduced in the approach is the application of a secondary
randomization scheme, taking a bootstrap sample of the variables at each splitting
node, to infer variable relevance (feature extraction). In this way, RF reduces the
Table 3.1 Algorithm: random forests for regression and classification

In training
For b = 1 to B:
Draw a bootstrap sample Z* of size N from training data
Grow a tree Tb to the bootstrapped data, by recursively repeating the following steps (I–III)
for each terminal node of the tree, until minimum node size nmin is reached
I. Select m variables at random from the p variables
II. Pick the best variable/split point among the m
III. Split the node into two daughter nodes
Output the ensemble of trees {Tb}1B
In prediction
To make a prediction of the pattern x|x ∉ Z (out-of-bag samples)
1 B
Regression: f rf ( x ) = Σ b=
B
1Tb ( x )
B

Classification: Let C  b ( x ) be the class prediction of pattern x at bth tree of the RF. Obtain
{ }
B
 rfB = majority vote C
the prediction for the overall forest through C b ( x)
1
Table modified from Hastie et al. [23], Copyright 2009, with permission from Springer-Verlag
New York
variance through bagging, but simultaneously de-correlating the variables. This

dual randomization scheme has other interesting properties, such as the reduction of
the dependence between trees and the suitability of the OOB error rate of an estima-
tor of the generalization ability (Table 3.1).
Number of parameters to be tuned in RF is relatively small: number of variables
selected in the node variable subsampling (np), total number of trees (nt), and tree
depth (td). Indeed, its sensitivity to them is reduced. Furthermore, the literature on
the topic provides good practice recommendations [28]. Thus, a good value of np
should be around p for RF in classification and p/3 in regression mode. Once set
to this suggested values (or around), RF models usually provide a good predictive
performance. This parameter may become critical when the dimensionality of the
set (p) is small. In this very particular case, it is necessary to consider which the
expected ratio between informative and noninformative variables is. Low values of
the number of selected variables can lead to a big decrease in predictive perfor-
mance of the RF, as some of the important variables can be ignored. Even if this
ratio is very hard to know at the very beginning of the analysis, it is rarely the case
in metabolomics datasets (usually highly p dimensional).
One of the main claims of RF is that they do not overfit [27]. In general, this is a
true assumption, but with some remarks. Number of trees (nt) can be determined
with relative simplicity, just plotting the evolution of the OOB error rate according
to the number of trees included in the forest. If nt increases too much, chances are
that all variance (relevant or not) is contained in the forest. In these conditions, the
model may become too rich and too close to the training data, the right conditions
for an overfitted model [23].
52 I. Montoliu
Similarly to np, proposed tree depths are 1 and 5 depending on the RF mode
( classification and regression, respectively). This is a parameter that makes little sense
to optimize, as the little gains in predictive performance described in the literature do
not justify optimizing one more parameter.
The ensemble nature of the RF limits one of the nice features of trees, which is the
convenient representation of the data structure. The average of bnt trees makes rather
complex to determine the relevant structure common to the different bootstrapped
samples of the main dataset. However, RF provides key parameters to identify the
importance of the variables and to visualize the similarities among samples.
Variable importance measures allow the identification of their role in the struc-
ture of the dataset. The two most used in available RF implementations are based on:
I. Changes in node impurity. At each split, the improvement criterion (impurity
decrease) is stored and cumulated individually for each variable. The procedure
is applied to each of the bnt trees of the forest and averaged. This estimator is
mostly used in classification, using the Gini index as improvement criterion.
II. Changes in the accuracy of the model. In this estimator, the importance of the
variable is linked to its predictive strength. For its calculation, there are used the
samples in the OOB, using the following scheme:
(a) At each b tree, OOB samples are passed down to the tree to get the model
accuracy.
(b) The values of the jth variable are permuted randomly in the OOB samples.
Accuracy is recalculated accordingly.
(c) Decreases of accuracy due to permutation are averaged over all the b trees
of the random forest B. This gives an estimation of the importance of the
variable j. Importance is then converted to percentage.
In general, this approach provides more gradual variable importance estimations
than the changes in node impurity. It is mostly used in regression, estimating the
1 b
{
accuracy as mean standard error MSEOOB = ∑ yi − ŷiOOB (MSE). }
2
b 1
On the other side, proximities can be calculated while building the forest with the
aim of visualizing similitude patterns between objects. To construct such proximity
matrix, for each tree b(x; θb), any pair of observations in the OOB set sharing a leaf
(terminal node) have their proximity increased by one. To visualize these proximi-
ties between patterns, a multidimensionality scaling (MDS) plot is used. In brief,
this approach consists in an eigenvalue decomposition of the distance matrix and is
key in approaches such principal coordinate analysis [29].
The interpretation of these plots in RF is often controversial, as their utility is
often questioned by some authors. Often they show a similar star shape, with each
arm corresponding to an individual subclass. The more pronounced is the separation
between groups, the better the performance of the RF is expected. Some numerical
experiments show how the classifier boundaries are usually in the center of the star-
shaped cloud, while better separated samples lay in the edges. This point makes
sense, as similar patterns have more chances to end up in the same terminal node.
On the opposite, more dissimilar patterns have much less chances of sharing those
kinds of nodes (Fig. 3.1).
Fig. 3.1 Analysis of transcriptomics data from seven mice strains [73]. Graphical evaluation of
Random Forests results based in sample proximities: multidimensional scaling (a) shows differences
between groups of samples (b) Highlighting samples with extreme behavior (outliers). (c) Variable
importance measure (classification) based in the mean Gini index decrease (scaled) after variable
permutation. Expression levels at position expressing a highest mean decrease in the Gini index (d)
Other visualizations of the role of the objects in the model can be achieved using the
information contained in the proximity matrix. Thus, proximity information may also
be used to evaluate the outlier character of the samples in RF. To determine this feature,
the reciprocal of the sum of squared proximities is calculated between that observation
and the remaining ones within each class. Extreme values will point out those objects
with high influence in the overall performance of the model. Often, they correspond to
patterns that have been incorrectly classified or with high standard errors.
In principle, RF can handle continuous and categorical discrete variables.
However, in these cases, there has been detected possible bias in the assignment of
variable importance [30]. There have been identified two major sources for this
behavior, mainly the important changes in the measurement scale and the oscillation
in the number of categories. Two explanations on the mechanism underlying those
54 I. Montoliu
−0.2 0.0 0.2
0.2
0.0
Dim 1
−0.4 −0.2
0.2
Dim 2
0.0
−0.2
0.2
0.0
Dim 3
−0.2
−0.4 −0.2 0.0 0.2 −0.2 0.0 0.2
Fig. 3.2 Analysis of transcriptomics data from seven mice strains [73]. Graphical evaluation of
unsupervised random forests results based in sample proximities: multidimensional scaling shows
the differences between groups of samples on the first three dimensions
deficiencies are the biased variable selection in each individual classification tree
and the side effects of a bootstrapping with replacement (as used in RF).
RF can be also used in unsupervised mode to describe the structure of the dataset
[28] alike in principal component analysis (PCA) [8]. The key point is to create a
two-class case, where one class is the real dataset and the other class is generated
synthetically to be as close as possible as the original dataset. Both sets are com-
bined, and a RF model is built to predict both classes. As an outcome, objects that
are similar will lay in the same terminal nodes. This information is recorded in the
proximity matrix and can be visualized using the MDS plot (Fig. 3.2).
Summarizing, RF can be used in the buildup of supervised models of classifica-
tion or regression, since they are models with good performance, even in highly
complex real data. They are tolerant with missing data and can be easily trans-
formed in an unsupervised model, if used smartly, to describe the structure of the
data. They provide efficient tools for variable selection for homogeneous data sets,
even if the validity of the approach for heterogeneous datasets (continuous and
discrete data) can be argued. Moreover, the implementation of outlier detection
features enables their use for event detection.
3.5 R
ecent Contributions from RF in the Metabolome-
Microbiome Arena
Since the publication of the initial algorithm in 2001 [27], the interest in RF has
undergone an exponential growth. The amount of publications has increased from
less than 10 in 2002 to more than 500 in 2012 (www.scopus.com), covering a wide
spectrum of areas of interest in science. From the very beginning, computer science
and biochemistry, genetics, and molecular biology publications have been maintain-
ing a sustained lead, partially due to the strong interaction they have through multi-
disciplinary fields such as bioinformatics. The progressive growth in interest in RF
has been especially relevant in the medical community, wide-spreading applications
that have gone from support to diagnostics to biomedical signal processing (imag-
ing and multiparametrical monitoring). Other fields such as engineering, agriculture
sciences, environmental science, and chemistry have also devoted a remarkable
interest in RF as a partner tool for their applications in several fields, such as remote
sensing, prediction of environmental parameters, and quantitative structure activity
relationships in chemistry. In parallel to the exponential increase in publications
using the RF approach, it is also remarkable a qualitative aspect: the progressive
increase of heterogeneity of fields of application. This growth that can be considered
normal in a new coming procedure has stabilized at a high number of fields. One
possible interpretation of this fact is the awareness of the global scientific community
on the performance of the algorithm, which is in the process of its consolidation as a
standard tool of machine learning.
3.6 I ntegration Between Information Sources: RF

and Networks
Data fusion using RF has been performed mostly in engineering on several domains,
from medical to environmental applications. The approach is often used in high-
dimensional datasets coming from imaging, multisensorial applications and even
microarrays. A medical application example of this integration can be found in the
integration between transcriptomics and dermatoscopy in cutaneous melanoma
patients, which has led to improved classifications using RF [31], or eventually in
their key role in handling structural and functional imaging for prostate cancer
diagnosis [32]. Environmental sciences have also benefited of the approach, applying
RF regression models onto datasets fusing Lidar, Radar, and multispectral remote
sensing to predict multiyear bird detections (migratory activity) of eight bird species
[33] and structural forest attributes [34].
The integration of random forests as a preliminary feature extraction step in
network analysis has become a relevant procedure in several areas such as pathway
analysis and data integration and in describing the relationship between chemical
structure and interactions. Pathway analysis has benefited from the generalization of
56 I. Montoliu
the use of random forests as feature extraction method in microarray data to rank
important pathways from externally available databases [35]. The feature extraction
properties of RF have been also exploited in feeding networks integrating several
“-omics” datasets (metabolomics, gene expression) and phenotypic traits [36–38].
Moreover, RF have demonstrated to be useful, and even better performing than
other methods, in describing protein-protein interactions [39, 40], the prediction of
long disordered regions within proteins [41], the identification of protein complexes
using topological structure from local subgraphs [42], and even predicting binding sites
by using structural information [43]. RNA-protein interactions using sequence infor-
mation were tested with success with two classifiers based in RF and SVM as a basis
to further create RNA-protein interaction networks [44]. Still in this area, the integra-
tive analysis of co-expression modules (gene ontology, protein-protein interaction data
and literature) using RF enabled finding interesting gene-phenotype associations [45].
In general, the use of structural information is important to understand the interac-
tion mechanisms but can also be relevant to facilitate drug design. A good example
of the application with success of RF into this setup can be found in the prediction of
sulfotyrosine binding sites, where the RF are winning option to SVM, ANN, and
hidden Markov models (HMM) [46]. Their performance was still highly scored in
the integration of chemical, genomic [47], and pharmacological information to deter-
mine drug-target interactions [48], where they can be compared to other methods
such SVM [49], to help in the assessment of the reliability of the results.
Furthermore, a two-stage-based RF analysis was performed to characterize the
functional effects of single amino acid variants (SAV) combining sequence, struc-
ture, and residue-contact network features [50]. RF models showed their utility not
only in highlighting functional regions but also in scoring protein interactions from
one organism model into another to be used in protein interaction networks [51]. The
antioxidant biological activity of proteins associated to star graph topographical indi-
ces was better highlighted with RF, when compared with other methods [52].
Classifier performance is one of the main points of interest in all applications
where those models are the working horse to provide insight on data, or they are
even the main outcome to be achieved. RF has been explored in combination with
other classifiers to provide better overall predictions through using consensual out-
puts. In these setups, it is necessary to find a precision index measure that summa-
rizes individual performances, incorporating concepts such as maximum posterior
probability [53]. Patient treatment strategies can also benefit from this multi-
classifier strategy, for instance, the prediction of the coreceptor usage in HIV-1
patients using translated V3 genotypes as input [54].
3.7 Application-Driven Improvements to RF Scheme
The standard RF method based in the routines proposed initially by Breiman is still
actual, very performing and widely applied into different fields. Notwithstanding,
there have appear modifications to the original algorithm, with the aim of adapting
it better for the purpose of the application [55]. GWAS analyses are a very particular
setup, with a very unbalanced n-to-p ratio. Analyzing them using a RF approach
should have advantages over other strategies, in particular for the detection of inter-
actions between single nucleotide polymorphisms (SNPs). Despite this, careful
evaluation of the RF performance in such huge p, low n analyses has detected limi-
tations in the utilization of RF to detect interactions [56]. Thus, GWAS analysis
with RF is an example of special request that needs to be addressed with “ad hoc”
customizations, for instance, the modification of the sampling scheme introducing a
stratified sampling of the SNPs [57]. The aim of the approach is dual: avoiding
highly computational costs derived from an exhaustive analysis and keeping enough
informative SNPs at the same time. A similar example of a modification of the RF
scheme to limit the number of variables in GWAS analyses is the application of
search algorithms based on simulated annealing and genetic programming, basis of
the random forests fishing (RFF) model. In this approach, the dimensionality is
reduced updating repeatedly a limited set of variables obtained by RF tests to find
groups of variables predictive of the target phenotype.
Not only variable subset selection, but also choosing the number of trees to be
included in the classifier has received attention. In this way, there has been proposed
a dynamic determination of the number of trees during the growing of the forest
[58]. This approach goes in the opposite direction to growing an excess of trees and
select afterwards their optimal number checking the model performance.
Furthermore, changes in the voting mechanism based on weighting [59], feature
selection, clustering, nearest neighbors, and optimization techniques have been pro-
posed [60].
In this process of revisiting some of the key aspects of RF, variable importance
measures were also reevaluated focusing in their performance on extreme cases of
recognized limited performance of RF, such in highly unbalanced datasets [61].
With this aim, a more robust variable importance indicator was proposed, now using
a variant at the variable permutation step based in the area under the curve (AUC)
[30]. The application of clustering techniques (partition around medoids, PAM) to
the RF proximity matrix also enabled the identification of regulatory cliques in tran-
scriptomics data from yeast [62]. When assessing the performance of variable
importance measures in feature ranking, RF showed a limited performance in some
cases. With this purpose, average gain measure and the similarity-weighted estimate
were introduced with success to replace information gain and maximum likelihood
estimates [63].
More from an algorithmically point of view, improvements in predictive perfor-
mance have been achieved by proven variable selection schemes, such backward
variable elimination, and introducing changes in the tree induction procedure that
attempt to complement the trees in the ensemble [64]. The application of backward
variable elimination outperforms both single classification trees and standard RF
and provides similarity measures that successfully cluster samples per molecular
pattern [65]. Moreover, some of the algorithm basics implemented in RF methodol-
ogy (bagging and model aggregation) have been also a source for inspiring new
modelling approaches, such the stabilization of recursive partitioning models [66]
58 I. Montoliu
and the adaptation of RF into semi-supervised learning [67]. In a kind of loop-back,

boosting approaches have also benefited from RF, through their use as weak learn-
ers in ADABOOST to decrease chances in over fit [68].
3.8 New Challenges to RF: Network Inference
The possibilities of random forests have been extensively tested in many environ-
ments in biology, even as part of complex strategies comprising ensembles of data,
models, and methodologies. Aside from those cases in which the predictive step is
key for the application, their properties in feature extraction have become extremely
relevant with a strong impact in the field, feeding networks for further graph-based
modelling. Even if useful, this kind of approach reduces RF to a filtering step to
select the relevant variables from a set of measurements. Then, the challenge now is
to determine if RF could be directly incorporated into a model to benefit from their
statistical properties for network inference.
One possible approach of network inference using RF goes through the decom-
position of the regulatory network of p variables onto p regression problems and its
posterior evaluation with RF in regression [69]. Edge’s estimation is performed
building a RF model of pi ≠ k variables to predict pk and using the variable importance
measure to estimate the importance of the link. These putative link strengths are
further used to build the network model. Using this setup, properties from RF are
inherited into the network: no assumptions on the nature of the variable (both linear
and nonlinear interactions are allowed), and they provide directed graphs, are fast to
compute, and are relatively easy to scale up. Furthermore, this is an approach able
to generalize easily up to many types of data and even capable to integrate data from
different sources (microbiome-metabolome).
Pathway analysis is a powerful approach to add interesting insights to the out-
come of genome-wide association study (GWAS) analyses. Most of the pathway
methods are based in testing the cumulative main effects associated to a phenotype
(disease). However, gene-gene interactions are also expected to have relevance on
the etiology of disease. To tackle this information, a two-stage RF-based algorithm
has been proposed, which is a restricted variant of a previously published one [70].
In this algorithm, RF is performed twice in the set of SNPs to reduce the variable
size and thus increase the power in classification. In a first round, all SNPs corre-
sponding to a user-specified pathway are modelled according to the target stratifica-
tion, and SNPs above a certain threshold in their variable importance are kept to be
used in a second round. During this step, the dataset is reanalyzed using just the
relevant SNPs, and the prediction error rate of the model is used to generate a score
of the pathway. The significance of the score is assessed empirically using a permu-
tation test [71].
Integrating pathway information to identify similarities among them has been
also explored using RF [72]. In this setup, RF is used to build pathway clusters
using tight clustering, which are supposed to agglomerate pathways sharing similar
functions (even if they do not share specific variables). The approach models the
phenotype using sets of variables from different pathways, and their prediction
errors are used for further clustering. In some way, the approach is linking similitude
between RF models with pathway similitude. To sustain these proximities, OOB
errors rate the performance of the pathway to describe a specific phenotype, thus
giving an idea of RF model – pathway potential interest. In this approach, the
handling of the OOB errors is quite interesting, as it uses class votes to define
pathway distances that can be clustered, later on. These class votes are defined
using the proportion of votes of each class, for a specific subject, along all trees in
the forest. An interesting point of the approach is that enables the comparison
between pathways, even if they do not share variables (genes, metabolites). This
is an interesting point for data integration, as it enables the comparisons between
pathways of different organisms (host, microbiome) sharing same phenotype.
3.9 Conclusions and Perspective
Since their introduction, RF has become an alternative for all those cases in which
reliability of other models are compromised. Their nonparametric character, its
capability in handling low n, high p datasets, its predictive performance, and its
low tendency to over fit have been largely in favor of its consolidation as a general
purpose tool.
This consolidation does not imply this is a completely frozen approach. New
application-driven developments, improvements in the internals of the algorithm,
and changes in the mechanisms for feature selection reflect the research area is fully
active and still under development. Moreover, the validity of the approach is
sustained by proposed enhancements to tailor other analysis setups, as network
modelling or data integration. These ongoing proposals make integration between
data from different organisms/compartments and their association with phenotype
highly susceptible to benefit from RF approach.
References
1. Dudoit S, Shaffer JP, Boldrick JC. Multiple hypothesis testing in microarray experiments. Stat
Sci. 2003;18(1):71–103.
2. Shaffer JP. Multiple hypothesis testing. Annu Rev Psychol. 1995;46(1):561–84.
3. Smith CA, Want EJ, O’Maille G, Abagyan R, Siuzdak G. XCMS: processing mass spectrom-
etry data for metabolite profiling using nonlinear peak alignment, matching, and identification.
Anal Chem. 2006;78(3):779–87.
4. Nicholson JK, Lindon JC, Holmes E. ‘Metabonomics’: understanding the metabolic responses
of living systems to pathophysiological stimuli via multivariate statistical analysis of biological
NMR spectroscopic data. Xenobiotica. 1999;29(11):1181–9.
5. Fiehn O. Metabolomics – the link between genotypes and phenotypes. Plant Mol Biol.
2002;48(1–2):155–71.
60 I. Montoliu
6. Montoliu I, Genick U, Ledda M, Collino S, Martin FP, Le Coutre J, et al. Current status on
genome-metabolome-wide associations: an opportunity in nutrition research. Genes Nutr.
2013;8(1):19–27.
7. Massart DL, Vandeginste BGM, Buydens LMC, De Jong S, Lewi PJ, Smeyers-Verbeke
J. Handbook of chemometrics and qualimetrics. Amsterdam: Elsevier Science B.V.; 1997.
8. Jolliffe IT. Principal component analysis. New York: Springer; 2002.
9. Wold S, Sjöström M, Eriksson L. PLS-regression: a basic tool of chemometrics. Chemom
Intell Lab Syst. 2001;58(2):109–30.
10. Geladi P, Kowalski BR. Partial least-squares regression: a tutorial. Anal Chim Acta.

1986;185(C):1–17.
11. Trygg J, Wold S. Orthogonal projections to latent structures (O-PLS). J Chemom.

2002;16(3):119–28.
12. Trygg J, Wold S. O2-PLS, a two-block (X-Y) latent variable regression (LVR) method with an
integral OSC filter. J Chemom. 2003;17(1):53–64.
13. Westerhuis JA, Hoefsloot HCJ, Smit S, Vis DJ, Smilde AK, Velzen EJJ, et al. Assessment of
PLSDA cross validation. Metabolomics. 2008;4(1):81–9.
14. Barker M, Rayens W. Partial least squares for discrimination. J Chemom. 2003;17(3):166–73.
15. Bylesjö M, Rantalainen M, Cloarec O, Nicholson JK, Holmes E, Trygg J. OPLS discriminant
analysis: combining the strengths of PLS-DA and SIMCA classification. J Chemom.
2006;20(8–10):341–51.
16. Westerhuis JA, van Velzen EJJ, Hoefsloot HCJ, Smilde AK. Multivariate paired data analysis:
multilevel PLSDA versus OPLSDA. Metabolomics. 2010;6(1):119–28.
17. Cloarec O, Dumas ME, Craig A, Barton RH, Trygg J, Hudson J, et al. Statistical total correla-
tion spectroscopy: an exploratory approach for latent biomarker identification from metabolic
1H NMR data sets. Anal Chem. 2005;77(5):1282–9.
18. Cortes C, Vapnik V. Support-vector networks. Mach Learn. 1995;20(3):273–97.
19. Hagan MT, Menhaj MB. Training feedforward networks with the Marquardt algorithm. IEEE
Trans Neural Netw. 1994;5(6):989–93.
20. Guyon I, Weston J, Barnhill S, Vapnik V. Gene selection for cancer classification using support
vector machines. Mach Learn. 2002;46(1–3):389–422.
21. Breiman L, Friedman J, Olshen R, Stone C. Classification and regression trees. Boca Raton:
CRC Press LLC; 1984.
22. Duda RO, Hart PE, Stork DG. Pattern classification. 2nd ed. New York: Wiley; 2001.
23. Hastie T, Tibshirani R, Friedman J. The elements of statistical learning. Data mining, inference
and prediction. 2nd ed. New York: Springer; 2009. p. 588.
24. Breiman L. Bagging predictors. Mach Learn. 1996;24(2):123–40.
25. Friedman J, Hastie T, Tibshirani R. Additive logistic regression: a statistical view of boosting.
Ann Stat. 2000;28(2):337–407.
26. Schapire RE, Singer Y. Improved boosting algorithms using confidence-rated predictions.
Mach Learn. 1999;37(3):297–336.
27. Breiman L. Random forests. Mach Learn. 2001;45(1):5–32.
28.
Liaw A, Wiener M. Classification and regression by randomForest. R News.
2002;2(3):18–22.
29. Borg I, Groenen P. Modern multidimensional scaling: theory and applications. New York:
Springer Verlag; 2005.
30. Janitza S, Strobl C, Boulesteix AL. An AUC-based permutation variable importance measure
for random forests. BMC Bioinforma. 2013;14:119.
31. Moutselos K, Maglogiannis I, Chatziioannou A, editors. Heterogeneous data fusion and selec-
tion in high-volume molecular and imaging datasets. IEEE 12th conference on Bioinformatics
and Bioengineering proceedings 2012;407–412.
32. Viswanath S, Bloch BN, Rosen M, Chappelow J, Toth R, Rofsky N, et al. Integrating structural
and functional imaging for computer assisted detection of prostate cancer on multi-protocol
in vivo 3 tesla MRI. SPIE Medical Imaging 2009;7260.
33. Swatantran A, Dubayah R, Goetz S, Hofton M, Betts MG, Sun M, et al. Mapping migratory
bird prevalence using remote sensing data fusion. PLoS ONE. 2012;7(1):e28922.
34. Latifi H, Nothdurft A, Straub C, Koch B. Modelling stratified forest attributes using optical/
LiDAR features in a central European landscape. Int J Digit Earth. 2012;5(2):106–32.
35. Pang H, Lin A, Holford M, Enerson BE, Lu B, Lawton MP, et al. Pathway analysis using ran-
dom forests classification and regression. Bioinformatics. 2006;22(16):2028–36.
36. Acharjee A, Kloosterman B, de Vos RCH, Werij JS, Bachem CWB, Visser RGF, et al. Data
integration and network reconstruction with -omics data using Random Forest regression in
potato. Anal Chim Acta. 2011;705(1–2):56–63.
37. Chen Z, Zhang W. Integrative analysis using module-guided random forests reveals correlated
genetic factors related to mouse weight. PLoS Comput Biol. 2013;9(3):e1002956.
38. Tang X, Xiao J, Li Y, Wen Z, Fang Z, Li M. Systematic analysis revealed better performance
of random forest algorithm coupled with complex network features in predicting microRNA
precursors. Chemom Intell Lab Syst. 2012;118:317–23.
39. Lin N, Wu B, Jansen R, Gerstein M, Zhao H. Information assessment on predicting protein-
protein interactions. BMC Bioinforma. 2004;5:154.
40. Lee J, Lee J. Hidden information revealed by optimal community structure from a protein-
complex bipartite network improves protein function prediction. PLoS ONE. 2013;8(4):e60372.
41. Han P, Zhang X, Norton RS, Feng ZP. Large-scale prediction of long disordered regions in
proteins using random forests. BMC Bioinforma. 2009;10:8.
42. Li ZC, Lai YH, Chen LL, Zhou X, Dai Z, Zou XY. Identification of human protein complexes
from local sub-graphs of protein-protein interaction network based on random forest with
topological structure features. Anal Chim Acta. 2012;718:32–41.
43. Zheng C, Wang M, Takemoto K, Akutsu T, Zhang Z, Song J. An integrative computational
framework based on a two-step random forest algorithm improves prediction of zinc-binding
sites in proteins. PLoS ONE. 2012;7(11):e49716.
44. Muppirala UK, Honavar VG, Dobbs D. Predicting RNA-protein interactions using only
sequence information. BMC Bioinforma. 2011;12:489.
45. Mehan MR, Nunez-Iglesias J, Dai C, Waterman MS, Zhou XJ. An integrative modular
approach to systematically predict gene-phenotype associations. BMC Bioinforma. 2010;11
Suppl 1:S62.
46. Yang ZR. Predicting sulfotyrosine sites using the random forest algorithm with significantly
improved prediction accuracy. BMC Bioinforma. 2009;10:361.
47. Cao DS, Liang YZ, Deng Z, Hu QN, He M, Xu QS, et al. Genome-scale screening of drug-
target associations relevant to Ki using a chemogenomics approach. PLoS ONE.
2013;8(4):e57680.
48. Heider D, Verheyen J, Hoffmann D. Predicting Bevirimat resistance of HIV-1 from genotype.
BMC Bioinforma. 2010;11:37.
49. Yu H, Chen J, Xu X, Li Y, Zhao H, Fang Y, et al. A systematic prediction of multiple drug-target
interactions from chemical, genomic, and pharmacological data. PLoS ONE. 2012;7(5):e37608.
50. Wang M, Zhao XM, Takemoto K, Xu H, Li Y, Akutsu T, et al. FunSAV: predicting the func-
tional effect of single amino acid variants using a two-stage random forest model. PLoS ONE.
2012;7(8):e43847.
51. Pesch R, Zimmer R. Complementing the eukaryotic protein interactome. PLoS ONE.

2013;8(6):e66635.
52. Fernandez-Blanco E, Aguiar-Pulido V, Robert Munteanu C, Dorado J. Random forest classification
based on star graph topological indices for antioxidant proteins. J Theor Biol. 2013;317:331–7.
53. Ko D, Windle B. Enriching for correct prediction of biological processes using a combination
of diverse classifiers. BMC Bioinforma. 2011;12:189.
54. Masso M, Vaisman II. Accurate and efficient gp120 V3 loop structure based models for the
determination of HIV-1 co-receptor usage. BMC Bioinforma. 2010;11:494.
55. Liu S, Chen Y, Wilkins D. Large margin classifiers and random forests for integrated biological
prediction. Int J Bioinforma Res Appl. 2012;8(1–2):38–53.
56. Winham SJ, Colby CL, Freimuth RR, Wang X, de Andrade M, Huebner M, et al. SNP interac-
tion detection with Random Forests in high-dimensional genetic data. BMC Bioinforma.
2012;13:164.
62 I. Montoliu
57. Wu Q, Ye Y, Liu Y, Ng MK. SNP selection and classification of genome-wide SNP data using
stratified sampling random forests. IEEE Trans Nanobiosci. 2012;11(3):216–27.
58. Tripoliti EE, Fotiadis DI, Manis G. Automated diagnosis of diseases based on classification:
dynamic determination of the number of trees in random forests algorithm. IEEE Trans Inf
Technol Biomed. 2012;16(4):615–22.
59. Robnik-Sikonja M. Improving random forests. 2004.
60. Tripoliti EE, Fotiadis DI, Manis G. Modifications of the construction and voting mechanisms
of the random forests algorithm. Data Knowl Eng. 2013;87:41–65.
61. Anaissi A, Kennedy PJ, Goyal M, Catchpoole DR. A balanced iterative random forest for gene
selection from microarray data. BMC Bioinforma. 2013;14:261.
62. Xiao Y, Segal MR. Identification of yeast transcriptional regulation networks using multivari-
ate random forests. PLoS Comput Biol. 2009;5(6):e1000414.
63. Jiang L. Learning random forests for ranking. Front Comput Sci China. 2011;5(1):79–86.
64. Bernard S, Adam S, Heutte L. Dynamic random forests. Pattern Recogn Lett.

2012;33(12):1580–6.
65. Li S, Fedorowicz A, Singh H, Soderholm SC. Application of the random forest method in
studies of Local Lymph Node Assay based skin sensitization data. J Chem Inf Model.
2005;45(4):952–64.
66. Garge NR, Bobashev G, Eggleston B. Random forest methodology for model-based recursive
partitioning: the mobForest package for R. BMC Bioinforma. 2013;14:125.
67. Leistner C, Saffari A, Santner J, Bischof H, editors. Semi-supervised random forests. 2009.
68. Zeng JY, Cao XH, Gan JY. An improvement of AdaBoost for face detection with random
forests. ed. CCIS; 2010;93: 22–9.
69. Huynh-Thu VA, Irrthum A, Wehenkel L, Geurts P. Inferring regulatory networks from expres-
sion data using tree-based methods. PLoS ONE. 2010;5(9):e12776.
70. Chang JS, Yeh RF, Wiencke JK, Wiemels JL, Smirnov I, Pico AR, et al. Pathway analysis of
single-nucleotide polymorphisms potentially associated with glioblastoma multiforme suscep-
tibility using random forests. Cancer Epidemiol Biomarkers Prev. 2008;17(6):1368–73.
71. Chung RH, Chen YE. A two-stage random forest-based pathway analysis method. PLoS ONE.
2012;7(5):e36662.
72. Pang H, Zhao H. Building pathway clusters from Random Forests classification using class
votes. BMC Bioinforma. 2008;9:87.
73. Collino S, Martin F-P, Montoliu I, Barger J, Da Silva L, Prolla T, et al. Transcriptomics and
metabonomics identify essential metabolic signatures in calorie restriction (CR) regulation
across multiple mouse strains. Metabolites. 2013;3(4):881–911. PubMed PMID: d oi:10.3390/
metabo3040881.
Chapter 4
Metabonomics in Translational Research
for Personalized Medicine and Nutrition
Guoxiang Xie and Wei Jia
Abstract Personalized medicine promises prediction, prevention, and treatment of

illness that is targeted to individuals’ needs. New technologies for detailed biologi-
cal profiling of individuals at the molecular level have been crucial in initiating the
move to personalized medicine. Metabonomics is promising to contribute signifi-
cantly to the characterization of various disease phenotypes and to the identification
of personal metabolic features that can predict response to therapies. Based on ana-
lytical platforms such as mass spectrometry or nuclear magnetic resonance spec-
troscopy, the metabonomic approach enables a comprehensive overview of the
metabolites, leading to the characterization of the metabolic profiles of a given
sample. These metabolic profiles can then be used to distinguish between different
disease phenotypes and to predict a drug’s effectiveness and/or toxicity.
Metabonomics has tremendous potential to advance our understanding of human
health and disease and to inform the development of personalized approaches to
disease prevention, diagnosis, and treatment.
Keywords Gut microbiota • Metabonomics • Personalized medicine • Personalized

nutrition • Pharmacometabonomics • Pharmacogenetics • Biomarkers • Mass
spectrometry • Nuclear magnetic resonance
4.1 Metabolomics and Metabonomics
The “-omics” sciences (Fig. 4.1), including genomics, transcriptomics, proteomics,

and metabonomics, have emerged over the last two decades as a systems biology
approach to obtain important insights into the role of host–gut microbial metabolic
interactions in an individual’s susceptibility to disease and treatment outcomes [1].
In particular, metabonomics will have a particular role with respect to other “-omics”
sciences because of its ability to detect, in real time, the adaptive multiparametric
responses of an organism to pathophysiological stimuli or genetic modifications [2].
G. Xie, M.D. • W. Jia, Ph.D. (*)

University of Hawaii Cancer Center, 701 Ilalo Street, Honolulu, HI 96813, USA

DOI 10.1007/978-1-4471-6539-2_4
64 G. Xie and W. Jia
Fig. 4.1 The flow of the “omics” sciences: genomics, transcriptomics proteomics, and metabo-
nomics technologies in individualized medicine
Metabolomics [3], or metabonomics [4], which is the quantitative measurement of

dynamic metabolic changes of living systems using nuclear magnetic resonance
(NMR) spectroscopy and/or mass spectrometry (MS), offers new opportunities to
explore individual needs, foods, and nutrients functionalities and to formulate
tangible biological hypotheses that can be tested at the individual and population
scales. Monitoring the metabolic variations provides a unique insight into intra- and
extracellular regulatory processes involved in our metabolic regulation and homeo-
stasis. Application of metabonomics to diagnostics, drug research, and nutrition
might be integral to improved health and personalized medicine [5]. To date, numer-
ous metabolic profiling studies involving both animal models and human subjects
have been reported in the field of personalized medicine or nutrition [6–9]. For more
details, see Chaps. 1 and 2.
Metabonomics has been applied in clinical studies in two major areas. First area
is the early diagnosis and characterization of disease phenotypes, where metabo-
nomic analysis can detect a panel of metabolites that discriminate between groups
of subjects, enabling the metabolic characterization of a disease, or of a disease
phenotype. This is an exploratory process, since unexpected or even unknown
metabolites may turn out to be important in this discrimination, paving the way to
the formulation of new pathophysiological hypotheses [10, 11]. The second area of
application is the identification of individual metabonomic characteristics able to
predict drug effectiveness and/or toxicity – an approach termed as “pharmacometa-
bonomics.” The pharmacometabonomics is believed as a promising metabonomic
approach for screening human populations, leading to a concrete possibility of a
genuinely individualized approach to treatment [12, 13].
4 Metabonomics in Translational Research for Personalized Medicine and Nutrition 65
4.2 Metabonomics: The Link Between

Genotypes and Phenotypes
Generally, the phenotype is not necessarily predicted by the genotype. The gap
between the genotype and the phenotype is spanned by many biochemical processes,
each with individual dependencies on various influences, including drugs, nutrition,
and environmental factors. In this chain of biomolecules from the genes to the phe-
notype, metabolites are the quantifiable molecules with the closest link to the phe-
notype [14]. Many phenotypic and genotypic states, such as a toxic response to a
drug, are predicted by differences in the concentrations of functionally relevant
metabolites in biological fluids and tissues. Personalized medicine is currently
based on the concept of pharmacogenomics that studies the influence of an indi-
vidual’s genotype and/or single-nucleotide polymorphisms (SNPs) on their response
to a drug or medical treatment. Despite enormous efforts, pharmacogenomics has
had limited success in clinical pharmacology to predict drug response with absolute
certainty using single or multiple SNPs as biomarkers [15–18]. The main reason for
the limitations is that the response is dependent upon the phenotype of an individ-
ual, which is determined by both genotype and the complex interactions between
genes and other environmental factors [19, 20]. These environmental factors include
diet, lifestyle, gut microbiota, nutrition, medications, age, and exposures to toxins
or dietary supplements, as well as the individual physical and pathological condi-
tions (e.g., type 2 diabetes and obesity). Therefore, it is critical to be able to assess
an individual’s metabolic phenotype, which will provide useful information for
determining the correct drug and dose treatment and predicting the individual
response following a therapeutic intervention.
The metabolic phenotype (metabotype) is a result of the overall influences of the
patient’s physiological status, gut microbiome status, and chemical, genetic, and
other environmental factors. Changes in the metabotype reflected in the biofluid or
tissue evaluated occur downstream of alterations in gene and protein expression.
As such, the metabotype, which comprises the genotype and phenotype, represents
the ultimate biological endpoint and can provide useful information about an indi-
vidual’s current physiological status that can be used for predicting the outcome
prior to a therapeutic intervention.
The “-omics” technologies have been used to obtain a more holistic view of how
biological systems work and underpin the base of functional genomics and systems
biology. They are also widely employed to identify biomarkers for use in the
diagnosis and monitoring of human disease. While “-omics” technologies such as
transcriptomics and proteomics are now well established and widely used across the
biological sciences, each has limitations and only provides part of the picture.
Metabonomics is clearly complementary to other “-omics” approaches but may
have a special role in bridging the phenotype–genotype gap, since metabonomics
provides the capability to analyze large arrays of metabolites for extracting bio-
chemical information that reflects true functional endpoints of overt biological
events, whereas other functional genomics technologies such as transcriptomics and
proteomics merely indicate the potential cause for phenotypic response.

Metabonomics bridges this information gap by depicting, in particular, such func-
tional information because metabolite differences in biological fluids and tissues
provide the closest link to the various phenotypic responses. Such changes in the
biochemical phenotype are of direct interest to pharmaceutical, biotech, and health
industries once appropriate technology allows the cost-efficient mining and integra-
tion of this information. Metabonomic approaches can also provide unique insights
into metabolic dynamics. Specifically, analysis of metabolite changes over time,
and stable isotope labeling experiments can be used to infer metabolic fluxes that
cannot be deduced by any of the other “-omics” methods. Therefore, they can
necessarily predict drug effects, toxicological response, or disease states at the phe-
notype level unless functional validation is added.
Over the last few years, genome-wide association studies (GWAS) have been
carried out and reported an extraordinary harvest of new genetic associations with
metabolic traits as phenotypic traits [21]. GWAS technology has been particularly
productive in the area of human autoimmune disorders, where over 100 confirmed
associations have been reported in diseases including type 1 diabetes [22], depres-
sion [23], inflammatory bowel disease [24, 25], and rheumatoid arthritis [26]. The
metabotypes of individuals result from gene, environment, lifestyle, food, and host–
gut microbial interactions have been identified by simultaneous measurements of
SNP and blood concentrations of endogenous metabolites in human population
[27]. Individuals with polymorphisms in genes coding for well-characterized
enzymes of the lipid metabolism have significantly different metabolic capacities
with respect to the synthesis of some polyunsaturated fatty acids, the beta-oxidation
of short- and medium-chain fatty acids, and the breakdown of triglycerides [21].
Thus, the concept of “genetically determined metabotype” as an intermediate phe-
notype provides a measurable quantity in the framework of GWAS with metabo-
nomics and might help to better understand the pathogenesis of common diseases
and gene–environment interactions. The metabotypes, in interactions with environ-
mental factors such as nutrition and lifestyle, may influence the susceptibility of an
individual for certain phenotypes. For example, there are potential links between
long-chain fatty acid metabolism and attention-deficit hyperactivity syndrome [28].
Understanding these connections, in turn, may eventually lead to more targeted
nutrition or therapies and more refined disease risk stratification. These could result
in a critical step towards personalized health care and nutrition based on a combina-
tion of genotyping and metabolic characterization.
Personalized medicine promises prediction, prevention, and treatment of illness
that is targeted to individuals’ needs. A primary goal of personalized medicine is to
provide the best medical treatment for each individual patient by determining which
drug will have the best efficacy with the least toxicity and/or adverse effects [29].
Figure 4.2 describes the different flows from clinical presentation through treatment
for standard clinical practice and a personalized medicine approach. Moving for-
ward, the use of a combination of “-omics” technologies aiming to develop bio-
markers tailored for individual responses will provide a more personalized approach
to patient treatment with a more positive outcome by diagnosing not only the disease
but also the disease phenotype [29].
Fig. 4.2 Differences between standard clinical practice and a personalized medicine approach to
diagnosis and treatment of disease (Adapted from Ref. [30])
4.3 Role of Pharmacometabonomics

in Personalized Medicine
In order to choose the right therapy for the right patient at the right time, more
accurate tests that can diagnose and classify a patient’s disease and their response to
medical treatment are needed. Pharmacometabonomics, focusing on small mole-
cule metabolite profiles and concerning the inherent different metabotypes, is an
emerging approach that combines metabolic profiling and bioinformatics to link the
inherent variation of a metabotype to the prediction of drug efficacy or toxicity in
patients [30]. Nicholson [30, 31] described the potential for pharmacometabonom-
ics in clinical trials and in longitudinal studies of individual patients or groups of
patients prescribed with particular therapies.
The main potential application for pharmacometabonomics is in personalized
health care. The advantages of pharmacometabonomics over the other targeted
“-omics” technologies are its unbiased opportunity for finding nonpreselected,

and hence unexpected, biomarkers and biomarker combinations, as multiple
analytes are quantified simultaneously from biofluids. The metabolic profile
represents the phenotype of the organism and reflects the overall biological
influences, including interactions between multiple genomes (e.g., genomes
from animals or humans and their gut microbiome). Pharmacometabonomics
uses the pre-dose metabolite profiling in the biofluids or fecal extracts to predict
the responses of an individual to a drug/nutritional intervention and to identify
surrogate markers for subsequent drug administration. Furthermore, pharmaco-
metabonomics is capable of providing useful drug pharmacokinetic and drug
metabolite information for an individual, which can provide a mechanistic
understanding of varied responses between individuals to the efficacy, side
effects, and toxicity of a drug. Thus, it needs not be limited by prior biological
understanding or hypotheses and can indeed be a powerful hypothesis-generating
scenario (Fig. 4.3) [3].
Fig. 4.3 An illustration of how pharmacometabonomics could be incorporated into clinical opera-
tions after suitable validation. In addition, metagenomic data on the gut microbiota can be included,
as shown by the dotted arrows (Adapted from Ref. [31])
4.4 Methodologies Used in Metabonomics
Analysis of the metabolome is challenging, not only because of the chemical

diversity and dynamic range of cellular metabolites, but also the redundancy of
cellular metabolic pathways that complicate data interpretation. However, recent
advances in analytical platforms, particularly MS, as well as bioinformatic and com-
putational tools, have made metabonomics much more accessible to researchers.
In view of the chemical and physical diversity of small biological molecules, the
challenge remains in developing protocols to gather the whole “metabolome.” It is
generally accepted that no single analytical technique such as NMR, gas chroma-
tography–MS (GC–MS), and liquid chromatography–MS (LC–MS) could capture
the entire composition of endogenous metabolites, and therefore, multiple comple-
mentary analytical platforms for global metabolic profiling are needed for an
enhanced metabolic measurement and visualization.
In general, analytical techniques such as NMR spectroscopy or MS are the
primary analytical methods employed in metabolite profiling. NMR metabonomics
is characterized by a higher throughput potential and more reliable determination of
metabolite concentrations: the intensities of the peaks in NMR spectra are propor-
tional to the metabolite concentrations, making it a true quantitative technique [11].
The basic workflow for NMR-based as well as MS-based studies is as follows:
quenching/extraction of metabolites → data collection → data processing/analysis
[32–34]. Metabonomics is usually carried out on samples that are available either
noninvasively (e.g., urine, feces, saliva, sweat, or exhaled breath condensate) or
minimally invasively (e.g., serum or plasma from blood). Tissues, intact cells, or
extracts thereof can also be analyzed.
NMR spectroscopy [34, 35] is a nondestructive technique that is highly effective
for structural elucidation. Compared to MS, it is less sensitive, but recent develop-
ment in operating methods and technology has increased its sensitivity and spectral
dispersion considerably. NMR data acquisition on intact tissue samples using a
high-resolution 1H magic angle spinning NMR spectroscopy is also possible. All of
the metabonomic datasets generated from different instruments can be, and should
be, combined if possible so that an integrated metabonomic approach across differ-
ent types of samples is possible.
MS is an effective analytical technique for the identification of metabolic
biomarkers [32]. It is inherently more sensitive than NMR spectroscopy but, to
ensure maximal metabolome coverage and minimize technical problems associated
with ion suppression, it requires a separation technique prior to MS. Both GC and
LC directly coupled to MS have been employed for metabolic profiling, but in most
studies MS has been coupled to HPLC or UPLC for metabonomic studies on bioflu-
ids and tissue extracts [36]. UPLC, which employs sub-2-μm particle columns, pro-
vides superior sensitivity and resolution compared to conventional LC [37, 38].
Moreover, the coupling of this technique to tandem MS (MS/MS) further enhances
specificity, provides an improved signal-to-noise ratio compared to single-stage
MS, and has radically improved the ability to obtain robust and comprehensive
metabolite profiles such as bile acids [39–41]. The combination of NMR and MS
data may improve the identification of unknown metabolites [42].
Both NMR and MS spectra are highly complex, and the biological information
can only be extracted by applying bioinformatic or computational tools, such as
pattern recognition methods. Simple methods such as principal component analysis
(PCA) allow visualization of clustering of similar samples and the determination of
aberrant or outlier samples. The supervised methods, such as partial least squares
(PLS), use a training set of samples (of known classification) to create a mathemati-
cal model that is then used to test an independent dataset, which enables us to pre-
dict to which group a new sample belongs on the strength of the characteristics of
its spectra [43].
Later, new methods for improving the structural identification of metabolites
from NMR data have been established. Statistical total correlation spectroscopy
(STOCSY) takes advantage of the colinearity of the intensity variables for the mul-
tiple peaks of a metabolite in a set of NMR spectra, so that correlations from NMR
peaks belonging to the same molecule can be identified [44]. This is particularly
useful for biomarker identification from complex biofluid analysis. An extension of
STOCSY, statistical heterospectroscopy, allows for the coanalysis of datasets
obtained by both NMR spectroscopy and MS [45]. Moreover, the principles of this
approach have the potential to correlate data from any two spectroscopic techniques
or, by extension, to correlate such data with that from other “-omics” fields such as
proteomics. These two tools, and various extensions of the methods that have been
published recently, may become major approaches for biomarker identification and
biochemical pathway information in “-omics” sciences and systems biology. Other
metabonomic and lipidomics strategies are being discussed in Chap. 2.
4.5 Metabonomic Technologies for Toxicology Studies
Metabonomics evaluation is nontargeted, enabling users to gain a comprehensive

evaluation of the systemic response of the subject (preclinical or clinical) to patho-
physiological stimuli or genetic modification [2]. Metabonomics studies demon-
strate its potential impact on the drug discovery process by enabling the incorporation
of safety endpoints much earlier in the drug discovery process, reducing the likeli-
hood (and cost) of later stage attrition [14]. The advantage of metabonomics to
convey phenotype and to be obtained from peripheral samples makes it very attrac-
tive from a translation standpoint for rapid throughput in vivo toxicity screening
particularly within the pharmaceutical industry [46]. To maximize the information
obtainable from multivariate datasets, a high-throughput technology is desirable so
that the data matrices produced can fully define both the variations associated with
a disorder and the innate variations associated with the biological system, while
minimizing false positives associated with such global multivariate analyses.
A metabolic profile needs not be a comprehensive survey of composition, nor needs
it be completely resolved and assigned, although these are all desirable attributes.
For the profile to be useful across a range of problems, however, it must be amena-
ble to quantitative interpretation, and it should be relatively unbiased in its scope.
A further requirement for the analytical platforms used to generate profiles is that
the analytical variations introduced after collection be less than the typical variation
in the normal population of interest, so as not to reduce significantly the opportunity
to detect treatment/group-related differences. Fulfilling this condition is dependent
on the actual system and questions to be addressed, in addition to the means of qual-
ity control used in the analytical methods.
In a study conducted by Clayton et al., NMR-based metabonomics was applied
to profile pre- and post-dose urine samples from 65 rats given a single toxic dose of
acetaminophen [13]. The metabolic profile of the pre-dose urine samples can pre-
dict both individual susceptibility to acetaminophen-induced toxicity and liver
injury and also can predict the relative excretion levels of acetaminophen metabo-
lites in the forms of glucuronide and sulfate conjugates. Later, the same group pro-
vided the first demonstration of pharmacometabonomics in humans by demonstrating
a clear connection between an individual’s basal urinary metabolic phenotype and
the metabolic fate of a standard dose of the widely used analgesic acetaminophen
[47]. NMR-based metabonomics approaches were employed to profile pre- and
post-dose urinary metabolites and discovered that human subjects with high
pre-dose levels of p-cresol (one of the metabolites related to an individual’s gut
microbiome) had lower concentrations of acetaminophen metabolites. From post-
dose urine samples, it was possible to determine the proportions of the various drug
metabolites excreted by each subject, which was known to show considerable inter-
subject variation. The findings indicate that each individual, colonized by a unique
assortment of trillions of microbes, responds to a drug differently, either beneficially
or adversely. It provides the information of how a particular drug is metabolized and
excreted by each individual. Such information may have a major influence on the
drug safety and efficacy. This study demonstrates that evaluation of a metabolic
phenotype by metabolic profiling could play an important role in drug metabolism
and toxicity, as well as in personalized health care.
One other study in animals reported by Li et al., who used two established
experimental models, the streptozotocin-induced diabetic model and a high-energy,
diet-induced obesity model, both in rats, demonstrated that the different outcomes
of streptozotocin-induced diabetes or dietary intervention could be correlated to
variations in pre-dose urinary metabolites of the rats, mainly those from gut micro-
biota [48]. It appears that these predispose the animals to different pathophysiologi-
cal outcomes upon diet alteration or chemical stimulus. They also surmised that,
based on their animal model observations, for obese/diabetic human subjects, subtle
variations in metabolic phenotype may predetermine their responses to xenobiotic
perturbation, ultimately leading to variability in pathophysiological processes.
Cisplatin has been one of the most widely used anticancer agents, but its nephro-
toxicity remains a dose-limiting complication. The idiopathic nature and the
pre-dose prediction of cisplatin-induced nephrotoxicity were evaluated using a
NMR-based pharmacometabonomic approach [49]. Significant individual differ-
ence was observed that cisplatin produced serious toxic responses in some animals
(toxic group) but had little effect in others (nontoxic group). The individual
metabolic profiles, assessed by urine NMR spectra, showed large differences
between the post-administration profiles of the two groups. Importantly, multivari-
ate analysis of the NMR data showed that the toxic and nontoxic groups can be
differentiated based on the pretreatment metabolite profiles. This study provides a
working model that can predict the idiopathic toxicity outcome based on metabolite
markers found by metabonomics approach. Thus, a pharmacometabonomic
approach using pretreatment metabolite profiles may help expedite personalized
chemotherapy of anticancer drugs.
In both preclinical screening and mechanistic exploration, metabolic profiling
can offer rapid, noninvasive toxicological information that is robust and reproduc-
ible, with little or no added technical resources to existing studies in drug metabolism
and toxicity. Metabonomics is an integral component of the pharmacogenomics
toolbox, especially as its ultimate goal is personalized health. The metabolic profile
contains information about the metabotype in addition to the genotype. The infor-
mation obtained from genotyping and metabotyping will allow a specific course of
treatment to be defined that will have the potential for a successful outcome, thus
making personalized health care a reality.
4.6 Metabonomics and Personalized Nutrition
The metabolome, or the complete metabolite composition of a system such as a cell

or organism, is the end product not only of the genetic blueprint of an organism but
also all influential factors to which the organism is exposed, such as nutrition,
environmental factors, or treatments. Changes in an individual’s metabolome occur
immediately or on a more gradual basis, partially due to the constancy of an indi-
vidual’s genetic makeup and lifestyle/environment. Holmes et al. suggest that com-
mon diets, gut microbes, medicinal practices, genetics, and other lifestyle and
environmental factors give rise to regional metabonomic phenotypes [50]. It has
long been understood that nutrition plays a role in human health. However, many of
the links between an individual’s diet and specific health outcomes are still not com-
pletely understood, for example, why one person easily develops obesity and
another, with the same diet, does not (responders and nonresponders). Metabonomics,
generating profiles of metabolites in biofluids, including urine, plasma, and fecal
water, provides a systems approach to understanding global metabolic regulation of
an organism and its commensal and symbiotic microbiota. In particular, it focuses
on the measurements of metabolite concentrations, fluxes, and secretions in cells
and tissues in which there is a direct connection between gene expression, protein
activity, and metabolic activity [4]. Metabonomic strategies together with advanced
chemometric and bioinformatic tools [44, 51, 52] can help track the interaction
between nutrients and human metabolism, as well as the involvement of the genome
and the gut microbiome, in overall human health, and can be considered critical
measures of function or phenotype [53]. This makes it possible to assess the
Fig. 4.4 Application of nutrigenomics and systems biology together with new bioinformatics
tools to unravel disease mechanisms, define biomarkers, or apply personalized nutrition. The
nutrigenomics approach extracts relevant differences, which become leads for further mechanistic
research, while the nutritional systems biology approach aims at a complete description of the
physiologic response by exploiting the complete data sets, thus targeting a new concept of
biomarker (Adapted from Ref. [55])
metabolic component of nutritional phenotypes and will enable individualized

dietary recommendations. The relation between diet and metabonomic profiles as
well as between those profiles and health and disease needs to be established. In the
past three decades, nutritional research has undergone an important shift in the
focus from physiology and epidemiology to genetics and molecular biology. Many
progresses are made through a systematic inventory of all relevant parameters by
using different “-omics” technologies and application of new bioinformatics tools
together with extensive data warehousing to unravel disease mechanisms, define
biomarkers, or apply personalized medication (Fig. 4.4).
Personalized nutrition is the outcome for individuals who will adapt their diet
and lifestyle according to the knowledge about their current or future healthy status,
and their subsequent nutritional requirements [54]. The knowledge could be built
around the characterization of different metabolic phenotypes in human population.
The pharmacometabonomics, discussed in this chapter, is sensitive to both genetic
and environmental influences, and addresses the metabolic response at the individ-
ual level. This concept could be alternatively applied to nutritional research as a
means of assessing individual response to diets or phytochemicals. In the future,
researchers could use such metabolic profiling to measure, predict, and optimize the
metabolic response of individual response to dietary interventions or modulations
[55]. Likely, in cases of impairment of human homeostasis, the patients would thus
develop a coordinated approach to reestablish a metabolic trajectory for the
individual consistent with their metabolic phenotype.
A depletion–repletion study of choline conducted by Zeisel et al. showed that an
individual’s metabonomics profile at baseline could predict whether or not the
individual would develop liver dysfunction as a result of inadequate choline intake

[56]. A study funded through the European Nutrigenomics Organisation (NuGO)
found that it is possible to characterize individual responses to fasting and to take a
“snapshot” of small metabolic changes after 36 h of fasting [57]. A number of
metabolites and hormones not previously associated with fasting were identified
and high variability between individuals for certain markers, including leptin, the
satiety hormone, which shows how it is possible to use metabonomics to character-
ize the different responses of individuals to nutritional or physiological stress.
A metabonomics study by Wang et al. [58] was performed to investigate the
human metabolic response to nutritional intervention with chamomile, an important
alternative and functional food. The results showed a clear clustering of the subjects
as a function of chamomile tea intake characterized by a decreased urinary creati-
nine level and an increase in glycine and hippurate. Samples are obtained 2 weeks
after daily chamomile intake deviating from the samples collected before chamo-
mile intervention, which was partially ascribed to chamomile-induced changes in
gut microbial metabolism. The results of this study highlight the diversity of physi-
ological variations of human metabolism and emphasize the effect of nutritional
phytochemicals in modulating human metabolism and maintaining homeostasis of
human gut eco-system.
Xie et al. [59] performed a study on 20 volunteers to investigate the human meta-
bolic response to drinking Pu-erh tea over a 6-week period, using a UPLC-
QTOFMS-based metabonomics approach. The final metabolic profile was greatly
altered by Pu-erh tea consumption. The trajectory of the PCA scores plot based on
urine data revealed a clear separation tendency of samples obtained before (days 1
and 7), during (days 16, 21, and 28), and after tea ingestion (washout period; days
30, 36, 42). Interestingly, the metabolic patterns of samples obtained 2 weeks after
tea intake are still distinct from the pre-dose pattern, probably due to the possibility
that Pu-erh tea may change the structure of the resident gut microbiota.
This was followed by a more in-depth study of Pu-erh tea in human subjects [60].
Urine samples were collected at 0, 1, 3, 6, 9, 12, and 24 h within the first 24 h of tea
intake and once a day during a 2-week daily Pu-erh tea ingestion phase and a 2-week
“washout” phase. The dynamic concentration profile of bioavailable plant mole-
cules (due to in vivo absorption and the hepatic and gut bacterial metabolism) and
the human metabolic response profile were identified and correlated with each
other, highlighting the great potential of metabonomic strategy to unravel the com-
plex interactions between multicomponent nutraceuticals and human metabolic sys-
tem in nutritional studies.
A metabonomics study by Rezzi et al. [61] demonstrated that metabonomics can
be used to predict whether an individual will respond to a certain dietary treatment.
Twenty-two healthy male volunteers selected from 75 volunteers based on their
chocolate preferences (chocolate loving or chocolate hating) underwent a 1-week
double-cross-over study in which they consumed either chocolate or a bread placebo
on the two test days and followed a standardized diet throughout the study. NMR
analysis of 24 h blood and urine samples revealed that the chocolate preference of
the individual could be predicted from both biofluid samples even in the absence of
the chocolate stimulus. Such a prediction of the dietary preference of the individual
indicates that the metabolic profile may indeed contain a wealth of information
relating to the diet of an individual, and it may be possible to predict dietary
response, thus proposing a role for metabonomics in personalized nutrition [62].
4.7 Gut Microbiota and Phytochemical Profiling

in Nutrition Science
It is widely accepted that many major human diseases have significant genetic and
environmental factors and that the impairment of human homeostasis in an indi-
vidual or population is a complex result of the conditional probabilities of certain
genes interacting with a variety of environmental triggers.
Diet (nutrients) has a major influence on many diseases and modulates the com-
plex intercommunity of gut microbiota [63]. Understanding the relationships
between the host genome, nutrient (phytochemicals) intervention, and the highly
variable gut microbiota with their genomes is a sophisticated challenge in modern
nutritional science. In the search for new natural bioactive phytochemicals (drugs)
and therapies, most consideration in toxicology and efficacy is given to the genetic
components of the host (animal model or human). Little attention, however, is
directed to the individual microbiome or species variation in the microbiome that
might contribute to the interaction of the potent phytochemicals or drugs within the
human global system. Although the application of some “functional foods” and
“nutraceuticals” has been promoted by food companies, and probiotics are specifi-
cally designed to nourish gut microbiota, the identification and function of many gut
microbial species remain largely unknown. There is an urgent need to understand
the global function of these gut organisms in terms of their impact on human health,
taking into account host–gut microbiota interactions at gene, protein, and metabo-
lite expression levels. Bioanalytical profiling, such as metabonomics, of human
urinary or fecal samples can be utilized to provide holistic and dynamic information
to assist dietary and medical researchers on disease diagnosis, stratification, and
personalized gut microbiota targeted treatment. On the other hand, elucidating the
molecular mechanisms underlying host–gut microbial interactions during both
healthy and pathological conditions should be pursued in order to obtain a deeper
understanding to aid nutritional intervention and drug development.
The goal of nutrition has extended beyond just ameliorating or curing diseases
and now aims to achieve an overall objective in preventing diseases and improving
health. Therefore, the pivotal scientific objective has become understanding the
relationship between diet (both macro- and micronutrients) and health/diseases.
The comprehensive analysis of the metabolome via metabonomics will serve as the
bioinformational base for modern nutritional science. Biomarkers and/or patterns of
expression will undoubtedly have the potential to be used for human health assess-
ment (Fig. 4.5). Together this indicates that the future goal of nutritional research
Fig. 4.5 Conceptualization of nutritional metabonomics/nutrimetabonomics for health and risk

management. Integration of nutritional metabonomics/nutrimetabonomics and systems biology at
the population scale may lead to enhanced use of nutrients to prevent or delay the onset of disease
and to optimize human health at an extensive scale. The metabotypes of individuals result from
gene, environment, lifestyle, food, and host–gut microbiota interactions. Different metabotypes
(represented by green, blue, and red lines/ellipse) are under homeostasis that aims to maintain
metabolic fluctuations within a healthy range (green ellipse). Metabonomics-generated prognostic
biomarkers can be used to assess homeostasis loss and likelihood for future diseases. Nutritional
metabonomics/nutrimetabonomics aims at optimizing nutrition for health maintenance and to
restore homeostasis as illustrated by the blue line/ellipse (Adapted from Refs. [54, 55, 73])
will be to predict the likelihood of future diseases within the context of an individ-
ual’s overall heath and identify causal risk factors, leading to recommendations for
appropriate intervention, such as to change dietary habits or to avoid homeostasis
loss and maintain healthy status.
4.8 The Importance of the Gut Microbiota

on Metabolic Profiles
Symbiotic relationships between microbes and the mammalian and plant hosts
shape our world. The symbiotic microbiome tremendously increases the diversity of
metabolic pathways accessible to mammalian hosts, enabling them to metabolize
many things that they otherwise could not. As a result, gut microbes have been
associated with various essential biological functions in humans through a “net-
work” of microbial–host co-metabolism to process nutrients and drugs and modu-
late the activities of multiple pathways in a variety of organ systems [64–66]. The
studies by Nicholson et al. [64–67] demonstrate that the metabolic variations in GI
compartments such as duodenum, jejunum, and ileum, mammalian tissues such as
kidney and liver, and biofluids such as blood and urine are directly related to the
activities of various microorganisms that coexist in the gut. Perhaps most impor-
tantly, gut microbes enable us to digest cellulose, the single largest nutritional
energy source on the planet, and to survive on diets with low levels of particular
nutrients and high levels of particular toxins. For example, gut microbes metabolize
unabsorbed carbohydrates to short-chain fatty acids (SCFA), CO2, and H2 in the
colon. SCFAs are monocarboxylic acids with a chain length up to six carbon atoms,
i.e., acetic, propionic, butyric, iso-valeric, valeric, iso-caproic, and caproic acids.
SCFAs function both as an energy source and as a signaling molecule, and their
abundance and type (e.g., butyric, propionic, acetic acids) are directly related to the
speciation of the microbiota and their syntrophic interactions. Other signaling path-
ways (e.g., such as through the SCFA receptor GPR43) are similarly involved in
host energy balance, and different microbial communities interact differently with
these molecules [68]. Most mammals can obtain essential amino acids, such as
lysine, from their diet, but there is evidence that they also obtain them from their gut
microbes [69]. Many amino acids and perhaps other nitrogen-containing com-
pounds may be cycling between humans and their microbiota, a process that could
reduce dietary requirements for those nutrients [69]. However, whether the fluxes of
those amino acids or other essential nutrients between microbes and humans are
sufficient in quantity and diversity to meet the nutritional requirements is yet to be
resolved.
Diversity in gut microbial communities and function creates differences in nutri-
ent milieu, digesta retention times, and temperatures that create diverse microbial
niches and inhabitants. With the recent advances in the new molecular profiling
technologies such as metagenomics and metabonomics [69, 70], the direct correla-
tion of global metabolic changes with gut microbiome becomes increasingly impor-
tant to decipher the host–microbe relationships and to gain mechanistic understanding
of nutritional and drug intervention in the “gateway” of host–microbe co-metabolism.
Scientists from different disciplines are working together and beginning to deter-
mine the details of gut microbial diversity and manipulate the complex interactions
between the host metabolism and its symbionts for improved nutrition and disease
treatment [9, 70, 71].
4.9 Conclusion and Future Perspectives
Metabonomics is a novel approach that promises to enable the detection of states of

disease, to stratify patients based on metabolic profiles and to monitor disease pro-
gression. Metabonomic analysis may also be able to orient the choice of therapy,
identify responders, and predict toxicity (pharmacometabonomics), paving the way
to a personalized therapy.
Whereas the human genome is the set of all genes in a human being, the human
metabolome is the set of all metabolites in a human being. Metabonomics bridges
the gap between the genotype and the phenotype and is an important basis of
personalized medicine. Metabonomics has been used to identify biomarkers for

disease and the effects of drugs. Various metabonomic technologies including NMR
and MS have been intensively applied to metabonomics study. Pharmacometabonomic
approach to personalizing drug treatment uses a combination of pre-dose metabolite
profiling and chemometrics to model and predict the responses of individual sub-
jects. Metabonomics also has a role to play in assessing drug toxicity and in guiding
nutrition.
The advantage of the personalized medicine approach includes the fact that the
“-omics”-derived biomarkers will potentially provide better, earlier diagnosis of
disease and disease phenotype, leading to a better clinical outcome for the patient
[29]. Recent efforts that have correlated GWAS data with downstream metabolite
concentrations or ratios [21] will further drive these technologies towards personal-
ized medicine. Such associations provide a unique approach to stratify an individu-
al’s genotype and phenotype. On a population basis, such associations will also
provide more descriptive information about a gene’s function as well as pathway-
specific information about various diseases that will ultimately lead to the develop-
ment of more targeted therapies. An example of how the “-omics” datasets may
move forward the field of personalized medicine has been reported by the group of
Chen et al. [72]. An approach referred to as integrative personal “-omics” profile
evaluated genomic, transcriptomic, proteomic, metabonomic, and antibody profiles
from a single individual over a 14-month period. The study revealed changes in the
“-omics” profiles between healthy and viral states and between nondiabetic and
diabetic states throughout the study period. Furthermore, it was noted that disease
risk could be assessed from the individual and maternal genome sequences. This
study demonstrated that the integration of genomics data with other dynamic
“-omics” datasets can be used to predict various medical risks and the health status
of an individual. Such datasets for many individuals may provide a database that can
be used for enhancing diagnostics, monitoring, and treatment in the future with
metabonomics playing a critical role.
References
1. Nicholson JK. Global systems biology, personalized medicine and molecular epidemiology.
2. Nicholson JK, Lindon JC, Holmes E. Metabonomics’: understanding the metabolic responses
of living systems to pathophysiological stimuli via multivariate statistical analysis of biologi-
cal NMR spectroscopic data. Xenobiotica. 1999;29(11):1181–9 [Review].
3. Fiehn O. Metabolomics – the link between genotypes and phenotypes. Plant Mol Biol.
2002;48(1–2):155–71.
4. Nicholson JK, Connelly J, Lindon JC, Holmes E. Metabonomics: a platform for studying drug
toxicity and gene function. Nat Rev Drug Discov. 2002;1(2):153–61.
5. Hunter P. Reading the metabolic fine print. The application of metabolomics to diagnostics,
drug research and nutrition might be integral to improved health and personalized medicine.
EMBO Rep. 2009;10(1):20–3.
6. German JB, Roberts MA, Watkins SM. Personal metabolomics as a next generation nutritional
assessment. J Nutr. 2003;133(12):4260–6.
7. Scalbert A, Brennan L, Fiehn O, Hankemeier T, Kristal BS, van Ommen B, et al. Mass-
spectrometry-based metabolomics: limitations and recommendations for future progress with
particular focus on nutrition research. Metabolomics. 2009;5(4):435–58.
8. Martin F-PJ, Rezzi S, Pere-Trepat E, Kamlage B, Collino S, Leibold E, et al. Metabolic effects
of dark chocolate consumption on energy, gut microbiota, and stress-related metabolism in
free-living subjects. J Proteome Res. 2009;8(12):5568–79.
9. Li M, Wang B, Zhang M, Rantalainen M, Wang S, Zhou H, et al. Symbiotic gut microbes
modulate human metabolic phenotypes. Proc Natl Acad Sci U S A. 2008;105(6):2117–22
[Research Support, Non-U.S. Gov’t].
10. Sauer U, Heinemann M, Zamboni N. Genetics. Getting closer to the whole picture. Science.
2007;316(5824):550–1 [Comment].
11. Nicholson JK, Lindon JC. Systems biology: metabonomics. Nature. 2008;455(7216):1054–6
[Historical Article News].
12. Meyer JM, Ginsburg GS. The path to personalized medicine. Curr Opin Chem Biol.
2002;6(4):434–8 [Review].
13. Clayton TA, Lindon JC, Cloarec O, Antti H, Charuel C, Hanton G, et al. Pharmaco-
metabonomic phenotyping and personalized drug treatment. Nature. 2006;440(7087):
1073–7.
14. Jain KK. Textbook of personalized medicine. New York: Springer; 2009.
15. Becquemont L, Alfirevic A, Amstutz U, Brauch H, Jacqz-Aigrain E, Laurent-Puig P, et al.
Practical recommendations for pharmacogenomics-based prescription: 2010 ESF-UB
Conference on Pharmacogenetics and Pharmacogenomics. Pharmacogenomics. 2011;12(1):
113–24 [Congresses].
16. Cecchin E, Agostini M, Pucciarelli S, De Paoli A, Canzonieri V, Sigon R, et al. Tumor response
is predicted by patient genetic profile in rectal cancer patients. Pharmacogenomics
J. 2011;11(3):214–26.
17. Tabernero J, Cervantes A, Rivera F, Martinelli E, Rojo F, von Heydebreck A, et al.
Pharmacogenomic and pharmacoproteomic studies of cetuximab in metastatic colorectal can-
cer: biomarker analysis of a phase I dose-escalation study. J Clin Oncol. 2010;28(7):1181–9
[Clinical Trial, Phase I Comment].
18. Benhaim L, Labonte MJ, Lenz HJ. Pharmacogenomics and metastatic colorectal cancer: cur-
rent knowledge and perspectives. Scand J Gastroenterol. 2012;47(3):325–39 [Review].
19. Loscalzo J, Kohane I, Barabasi AL. Human disease classification in the postgenomic era: a
complex systems approach to human pathobiology. Mol Syst Biol. 2007;3:124.
20. Wilson ID. Drugs, bugs, and personalized medicine: pharmacometabonomics enters the ring.
Proc Natl Acad Sci U S A. 2009;106(34):14187–8 [Comment].
21. Gieger C, Geistlinger L, Altmaier E, Hrabe de Angelis M, Kronenberg F, Meitinger T, et al.
Genetics meets metabolomics: a genome-wide association study of metabolite profiles in
human serum. PLoS Genet. 2008;4(11):e1000282 [Research Support, Non-U.S. Gov’t].
22. Barrett JC, Clayton DG, Concannon P, Akolkar B, Cooper JD, Erlich HA, et al. Genome-wide
association study and meta-analysis find that over 40 loci affect risk of type 1 diabetes. Nat
Genet. 2009;41(6):703–7 [Research Support, N.I.H., Extramural Research Support,
Non-U.S. Gov’t].
23. Shyn SI, Shi J, Kraft JB, Potash JB, Knowles JA, Weissman MM, et al. Novel loci for major
depression identified by genome-wide association study of sequenced treatment alternatives to
relieve depression and meta-analysis of three studies. Mol Psychiatry. 2011;16(2):202–15
[Research Support, N.I.H., Extramural Research Support, Non-U.S. Gov’t].
24. Barrett JC, Hansoul S, Nicolae DL, Cho JH, Duerr RH, Rioux JD, et al. Genome-wide
association defines more than 30 distinct susceptibility loci for Crohn’s disease. Nat Genet.
2008;40(8):955–62 [Meta-Analysis Research Support, N.I.H., Extramural Research Support,
Non-U.S. Gov't].
25. Jostins L, Ripke S, Weersma RK, Duerr RH, McGovern DP, Hui KY, et al. Host-microbe
interactions have shaped the genetic architecture of inflammatory bowel disease. Nature.
2012;491(7422):119–24. doi:10.1038/nature11582.
26. Gregersen PK, Olsson LM. Recent advances in the genetics of autoimmune disease. Annu
Rev Immunol. 2009;27:363–91 [Research Support, N.I.H., Extramural Research Support,
Non-U.S. Gov’t Review].
27. Martin FP, Collino S, Rezzi S, Kochhar S. Metabolomic applications to decipher gut microbial
metabolic influence in health and disease. Front Physiol. 2012;3:113.
28. Burgess JR, Stevens L, Zhang W, Peck L. Long-chain polyunsaturated fatty acids in children
with attention-deficit hyperactivity disorder. Am J Clin Nutr. 2000;71(1 Suppl):327S–30
[Review].
29. Sun J, Beger RD, Schnackenberg LK. Metabolomics as a tool for personalizing medicine:
2012 update. Pers Med. 2013;10(2):149–61. 2013/03/01.
30. Nicholson JK, Wilson ID, Lindon JC. Pharmacometabonomics as an effector for personalized
medicine. Pharmacogenomics. 2011;12(1):103–11.
31. Nicholson JK, Everett JR, Lindon JC. Longitudinal pharmacometabonomics for predicting
patient responses to therapy: drug metabolism, toxicity and efficacy. Expert Opin Drug Metab
Toxicol. 2012;8(2):135–9 [Editorial].
32. Want EJ, Masson P, Michopoulos F, Wilson ID, Theodoridis G, Plumb RS, et al. Global meta-
bolic profiling of animal and human tissues via UPLC-MS. Nat Protocs. 2013;8(1):17–32.
doi:10.1038/nprot.2012.135.
33. Dunn WB, Broadhurst D, Begley P, Zelena E, Francis-McIntyre S, Anderson N, et al.
Procedures for large-scale metabolic profiling of serum and plasma using gas chromatography
and liquid chromatography coupled to mass spectrometry. Nat Protocs. 2011;6(7):1060–83.
doi:10.1038/nprot.2011.335.
34. Beckonert O, Keun HC, Ebbels TMD, Bundy J, Holmes E, Lindon JC, et al. Metabolic profil-
ing, metabolomic and metabonomic procedures for NMR spectroscopy of urine, plasma,
serum and tissue extracts. Nat Protocs. 2007;2(11):2692–703. doi:10.1038/nprot.2007.376.
35. Lindon JC, Nicholson JK. Spectroscopic and statistical techniques for information recovery in
metabonomics and metabolomics. Annu Rev Anal Chem (Palo Alto, Calif). 2008;1:45–69
[Review].
36. Lenz EM, Wilson ID. Analytical strategies in metabonomics. J Proteome Res. 2007;6(2):443–
58 [Review].
37. Russo R, Guillarme D, TTN D, Bicchi C, Rudaz S, Veuthey JL. Pharmaceutical applications
on columns packed with sub-2 microm particles. J Chromatogr Sci. 2008;46(3):199–208.
38. Xie G, Plumb R, Su M, Xu Z, Zhao A, Qiu M, et al. Ultra-performance LC/TOF MS analysis
of medicinal Panax herbs for metabolomic research. J Sep Sci. 2008;31(6–7):1015–26.
39. Want EJ, Coen M, Masson P, Keun HC, Pearce JT, Reily MD, et al. Ultra performance liquid
chromatography-mass spectrometry profiling of bile acid metabolites in biofluids: application
to experimental toxicology studies. Anal Chem. 2010;82(12):5282–9.
40. Huang JG, Bathena SPR, Csanaky IL, Alnouti Y. Simultaneous characterization of bile acids
and their sulfate metabolites in mouse liver, plasma, bile, and urine using LC-MS/MS. J Pharm
Biomed Anal. 2011;55(5):1111–9.
41. Xie GX, Zhong W, Li HK, Li Q, Qiu YP, Zheng XJ, et al. Alteration of bile acid metabolism in
the rat induced by chronic ethanol consumption. FASEB J. 2013;27:3583–93.
42. Pan Z, Raftery D. Comparing and combining NMR spectroscopy and mass spectrometry in
metabolomics. Anal Bioanal Chem. 2007;387(2):525–7 [Review].
43. Baraldi E, Carraro S, Giordano G, Reniero F, Perilongo G, Zacchello F. Metabolomics: mov-
ing towards personalized medicine. Ital J Pediatr. 2009;35(1):30 [Editorial].
44. Cloarec O, Dumas ME, Craig A, Barton RH, Trygg J, Hudson J, et al. Statistical total
correlation spectroscopy: an exploratory approach for latent biomarker identification from
metabolic 1H NMR data sets. Anal Chem. 2005;77(5):1282–9 [Research Support,
Non-U.S. Gov’t].
45. Crockford DJ, Holmes E, Lindon JC, Plumb RS, Zirah S, Bruce SJ, et al. Statistical
heterospectroscopy, an approach to the integrated analysis of NMR and UPLC-MS data sets:
application in metabonomic toxicology studies. Anal Chem. 2006;78(2):363–71 [Research
Support, Non-U.S. Gov’t].
46. Harrigan GG, Yates LA. High-throughput screening, metabolomics and drug discovery.
IDrugs. 2006;9(3):188–92.
47. Clayton TA, Baker D, Lindon JC, Everett JR, Nicholson JK. Pharmacometabonomic identifi-
cation of a significant host-microbiome metabolic interaction affecting human drug metabo-
lism. Proc Natl Acad Sci U S A. 2009;106(34):14728–33 [Clinical Trial Research Support,
Non-U.S. Gov’t].
48. Li H, Ni Y, Su M, Qiu Y, Zhou M, Qiu M, et al. Pharmacometabonomic phenotyping reveals
different responses to xenobiotic intervention in rats. J Proteome Res. 2007;6(4):1364–70
[Comparative Study Research Support, Non-U.S. Gov’t].
49. Kwon HN, Kim M, Wen H, Kang S, Yang HJ, Choi MJ, et al. Predicting idiopathic toxicity of
cisplatin by a pharmacometabonomic approach. Kidney Int. 2011;79(5):529–37 [Research
50. Holmes E, Loo RL, Stamler J, Bictash M, Yap IK, Chan Q, et al. Human metabolic phenotype
diversity and its association with diet and blood pressure. Nature. 2008;453(7193):396–400
[Research Support, N.I.H., Extramural Research Support, Non-U.S. Gov’t].
51. Holmes E, Foxall PJ, Nicholson JK, Neild GH, Brown SM, Beddell CR, et al. Automatic data
reduction and pattern recognition methods for analysis of 1H nuclear magnetic resonance
spectra of human urine from normal and pathological states. Anal Biochem.
1994;220(2):284–96.
52. Wang Y, Bollard ME, Keun H, Antti H, Beckonert O, Ebbels TM, et al. Spectral editing and
pattern recognition methods applied to high-resolution magic-angle spinning 1H nuclear mag-
netic resonance spectroscopy of liver tissues. Anal Biochem. 2003;323(1):26–32.
53. Go VL, Nguyen CT, Harris DM, Lee WN. Nutrient-gene interaction: metabolic genotype-
phenotype relationship. J Nutr. 2005;135(12 Suppl):3016S–20 [Research Support, N.I.H.,
Extramural Review].
54. Rezzi S, Ramadan Z, Fay LB, Kochhar S. Nutritional metabonomics: applications and per-
spectives. J Proteome Res. 2007;6(2):513–25.
55. Xie G, Li X, Li H, Jia W. Toward personalized nutrition: comprehensive phytoprofiling and
metabotyping. J Proteome Res. 2013;12:1547–59.
56. Sha W, da Costa KA, Fischer LM, Milburn MV, Lawton KA, Berger A, et al. Metabolomic
profiling can predict which humans will develop liver dysfunction when deprived of dietary
choline. FASEB J. 2010;24(8):2962–75.
57. Rubio-Aliaga I, de Roos B, Duthie SJ, Crosley LK, Mayer C, Horgan G, et al. Metabolomics
of prolonged fasting in humans reveals new catabolic markers. Metabolomics.
2011;7(3):375–87.
58. Wang Y, Tang H, Nicholson JK, Hylands PJ, Sampson J, Holmes E. A metabonomic strategy
for the detection of the metabolic effects of chamomile (Matricaria recutita L.) ingestion. J
Agric Food Chem. 2005;53(2):191–6.
59. Xie G, Ye M, Wang Y, Ni Y, Su M, Huang H, et al. Characterization of pu-erh tea using chemi-
cal and metabolic profiling approaches. J Agric Food Chem. 2009;57(8):3046–54.
60. Xie G, Zhao A, Zhao L, Chen T, Chen H, Qi X, et al. Metabolic fate of tea polyphenols in
humans. J Proteome Res. 2012;11(6):3449–57.
61. Rezzi S, Ramadan Z, Martin FP, Fay LB, van Bladeren P, Lindon JC, et al. Human metabolic
phenotypes link directly to specific dietary preferences in healthy individuals. J Proteome Res.
2007;6(11):4469–77 [Research Support, Non-U.S. Gov’t].
62. Brennan L. Session 2: personalised nutrition. Metabolomic applications in nutritional research.
Proc Nutr Soc. 2008;67(4):404–8 [Research Support, Non-U.S. Gov’t Review].
63. Nicholson JK, Holmes E, Lindon JC, Wilson ID. The challenges of modeling mammalian
biocomplexity. Nat Biotechnol. 2004;22(10):1268–74.
2008;4:219 [Research Support, Non-U.S. Gov’t].
65. Martin FP, Dumas ME, Wang Y, Legido-Quigley C, Yap IK, Tang H, et al. A top-down systems
biology view of microbiome-mammalian metabolic interactions in a mouse model. Mol Syst
Biol. 2007;3:112 [Research Support, Non-U.S. Gov’t].
66. Martin FP, Wang Y, Yap IK, Sprenger N, Lindon JC, Rezzi S, et al. Topographical variation in
murine intestinal metabolic profiles in relation to microbiome speciation and functional eco-
logical activity. J Proteome Res. 2009;8(7):3464–74 [Research Support, N.I.H., Extramural
Research Support, Non-U.S. Gov’t].
67. Martin FP, Wang Y, Sprenger N, Yap IK, Rezzi S, Ramadan Z, et al. Top-down systems biology
integration of conditional prebiotic modulated transgenomic interactions in a humanized
microbiome mouse model. Mol Syst Biol. 2008;4:205 [Research Support, N.I.H., Extramural
Research Support, Non-U.S. Gov’t].
68. Ley RE, Hamady M, Lozupone C, Turnbaugh PJ, Ramey RR, Bircher JS, et al. Evolution of
mammals and their gut microbes. Science. 2008;320(5883):1647–51.
69. Metges CC, Eberhard M, Petzke KJ. Synthesis and absorption of intestinal microbial lysine
in humans and non-ruminant animals and impact on human estimated average requirement
of dietary lysine. Curr Opin Clin Nutr Metab Care. 2006;9(1):37–41 [Research Support,
Non-U.S. Gov’t Review].
70. Pang X, Hua X, Yang Q, Ding D, Che C, Cui L, et al. Inter-species transplantation of gut
microbiota from human to pigs. ISME J. 2007;1(2):156–62 [Evaluation Studies Research
71. Wei H, Dong L, Wang TT, Zhang MH, Hua WY, Zhang CH, et al. Structural shifts of gut
microbiota as surrogate endpoints for monitoring host health changes induced by carcinogen
exposure. FEMS Microbiol Ecol. 2010;73(3):577–86.
72. Chen R, Mias GI, Li-Pook-Than J, Jiang L, Lam HY, Miriami E, et al. Personal omics profiling
reveals dynamic molecular and medical phenotypes. Cell. 2012;148(6):1293–307 [Research
Support, N.I.H., Extramural Research Support, Non-U.S. Gov’t].
73. van der Greef J, Stroobant P, van der Heijden R. The role of analytical sciences in medical
systems biology. Curr Opin Chem Biol. 2004;8(5):559–65.
Chapter 5
Can We Use Metabolomics to Understand
Changes to Gut Microbiota Populations
and Function? A Nutritional Perspective
Sofia Moco and Alastair B. Ross
Abstract Food is an integral part of human life, and the composition of our diet is
an important determinant of our health and well-being. Food is also the main source
of energy and nutrients for the gut microbiota, the 100 trillion cells that coexist
inside us. The impact of macronutrients (protein, fat, carbohydrates, and fiber) and
specific non-nutrient food components (polyphenols) will be reviewed in the con-
text of gut microbial function and interaction with the host. Colonic microbiota
provides diverse enzymatic activities differing from our own, which lead to the
production of metabolites essential for key metabolic functions, including carbohy-
drate and amino acid metabolism. Certain gut metabolites are specific to microbial
activity and confer functionalities beyond energy production, such as signalling cas-
cades across cells, tissues, and organs. Metabolomics has proven to be a useful tool
to measure the effects of food on the gut microbiota and its interaction with host
metabolism.
Keywords Nutrition • Gut • Metabolomics • Digestion • Phase II metabolism •

Food • One carbon metabolism • Polyphenols • Fiber • Microbiota • Diet • Fat •
Protein • Carbohydrate • Choline • Short-chain fatty acids • Phenolic and phenyl
metabolites • Indole metabolites • Hippurate • p-Cresol sulfate • Trimethylamine
oxide • Metabolism • Metabolomics • Colon • Intestine • Butyrate • Gut-liver • Gut-
brain • Pathways • Chocolate • Whole-grain cereals • Carnitine • Branched-chain
fatty acids • Prebiotics
S. Moco (*)
Natural Bioactives and Screening, Nestlé Institute of Health Sciences,
A.B. Ross
Food and Nutrition Science, Department of Biology and Biological Engineering,
Chalmers University of Technology, Gothenburg, Sweden

DOI 10.1007/978-1-4471-6539-2_5
84 S. Moco and A.B. Ross
5.1 Introduction
The sum of all small molecules in a system (i.e. the metabolome) not only reflects
the metabolic response of the subject of interest but also the organisms living in
symbiosis with the subject – in the case of humans, the gut microbiota is an example.
The gut microbiota produces thousands of metabolites through their reproduction,
interaction with other microorganisms, the host and with partially digested food.
Many of these metabolites are specific for microbial metabolism, and cannot be
synthesized by mammalian enzymes. These specific microbial metabolites can be
absorbed from the gut, adding to the diversity of the metabolome, and at the same
time providing a window into the interaction between host, food and gut microbiota.
In this chapter, we examine what dietary components are known to have an impact
on gut microbial metabolism, which biochemical classes of gut metabolites are pro-
duced from different diets, and how metabolomics can be a powerful tool to mea-
sure the effect of food on the gut microbiota, and its interaction with mammalian
metabolism.
5.2 Colonic Digestion
The large intestine is a digestive organ where dietary substrates not absorbed in the
small intestine, are further broken down by anaerobic bacteria (Fig. 5.1). The major
substrates for colonic fermentation include carbohydrates that have escaped diges-
tion in the upper gut (mainly dietary fibers: resistant starch and non-starch polysac-
charide such as celluloses, pectins and gums, and non-digestible oligosaccharides).
The main products of carbohydrate fermentation are short-chain fatty acids (SCFAs),
such as butyrate, propionate, and acetate, which are then absorbed and used as an
energy source. In Western diets, SCFAs contribute less than 10 % to the total energy
obtained from food, although in some cases this value can be up to 30 % [1]. Bacteria
well adapted for fermenting carbohydrate come from the Prevotella and Xylanibacter
genuses [2, 3].
In addition, residual amounts of protein (such as elastin, collagen and albumin),
peptides and amino acids can also reach the colon. Proteolytic bacteria in human
feces are predominantly Bacteroides and Propionibacterium, with lesser numbers of
the genera Streptococcus, Clostridium, Bacillus and Staphylococcus [4]. Low levels
of the amino acid fermentation products ammonia and branched-chain fatty acids
(BCFA) are found in ileal contents indicating that little amino acid fermentation
occurs in the small intestine, underlining the importance of the gut microbiota for
producing these compounds. Protein fermentation leads not only to the production
of BCFAs but also to relatively low amounts of a variety of products, such as
branched-chain amino acids (BCAAs), phenols, and amines which are both absorbed
into the host as well as excreted.
The proportion of carbohydrates to protein in the colon has been estimated as
3–4:1. Regional differences occur in the gut, where the right (proximal) gut has a
5 Can We Use Metabolomics to Understand Changes to Gut Microbiota Populations… 85
Fig. 5.1 Fermentation in the large bowel (size of compounds’ font indicate approximate relative
concentration)
higher saccharolytic activity while the left (distal) is more proteolytic. These fer-
mentation processes provide the nutrients required for gut microbiota biomass
growth while at the same time producing waste products hydrogen, carbon dioxide,
ammonia, and methane which are excreted via flatus. More complex chemical
structures such as polyphenols or alkaloids, mostly from plant foods, are also
degraded during colonic fermentation [1].
Beyond its role in digestion and absorption, the large intestine contributes to
health in a variety of ways: as a physical barrier preventing invasion of pathogenic
bacteria and stimulating immune function and as a site for biosynthesis of vitamins
and metabolism of xenobiotics.
5.3 Gut Modulation by Foods and Diet
5.3.1 How Do Different Foods Alter Gut Microbiota

and Their Metabolism?
There is little controversy in the idea of using food or food ingredients to alter both gut
microbiota populations and gut microbiota metabolism. Directly or indirectly, food is
the main source of energy and nutrients for the gut microbiota and evolutionary pres-
sure to adapt to the gastrointestinal environment and a major determinant of which
microbial genes are upregulated. Technological developments over the past two
decades in the areas of genetic sequencing, to determine the gut microbiome from
fecal samples without the need for culture techniques, have led to a rapid explosion of
our understanding of the importance of the gut microbiota and how it changes with
changing diet. An example of this are pre- and probiotics, where fermentable fiber
sources or live bacterial cultures (often in dairy products), are given with the aim to
positively alter the gut microbiota.
There is still discussion about what constitutes an “ideal” gut microbiota popula-
tion, though favorable changes to gut microbiota are generally described towards
bacterial genus or species that succeed when carbohydrate is the abundant energy
source, while “negative” bacterial species are those that are well adapted to ferment-
ing protein. Arguably, quantifying the population of different bacterial families or
species provides little direct information about actual gut bacterial metabolism:
many can switch between proteolytic and saccharolytic metabolism. It may be that
the end products of microbial metabolism are able to help build the best possible
picture of how gut microbiota are collectively responding to different diets or condi-
tions. Some metabolites of dietary substrates are well known and are summarized in
Table 5.1.
5.3.2 Microbial Metabolism and a Carbohydrate-Rich Diet
Carbohydrates are an important energy source for both humans and our gut micro-
biota. They are found in foods in several different forms, including monosaccha-
rides (e.g., glucose), disaccharides (e.g., sucrose or lactose), starch, and a range of
different types of dietary fiber, carbohydrates not broken down by human digestive
enzymes, but are often fermentable by gut microbiota. While traditional diets are
generally rich in complex carbohydrates (e.g., starch) and high in diverse forms of
fiber, in “Western” pattern diets, simple sugars (e.g., glucose and sucrose) dominate
the carbohydrate fraction of the diet, with low diversity in the small amount of fiber
present [2]. It is likely that the difference between traditional and Western dietary
patterns also leads to an impact on the gut microbiota and intestinal milieu leading
to an increased risk of gastrointestinal disorders including large bowel cancer, gall
stones, and Crohn’s disease. Diet intervention studies high in refined sugar have
found an altered gut metabolism, increased mouth-to-cecum transit time, and
increased production of secondary bile acids [5].
One of the main sources of dietary fibers is cereal-based foods. Cereal grains that
have the bran and germ fractions removed (refined or “white” flour) are also largely
depleted in dietary fiber. Whole grains are cereal grains that still have all the three
grain components in their correct proportions (bran, germ, and endosperm) and are
generally rich in both insoluble and soluble fibers [6]. Soluble dietary fibers are by
definition water soluble and tend to be highly fermentable by the gut microbiota,
producing SCFAs. SCFAs appear to have a wide range of roles, including as an
energy source and for reducing gut inflammation [2]. Diets high in fiber such as
Table 5.1 Common gut microbial metabolites of dietary substrates detected using metabolomics
(urine, feces, blood). oxidation is typical of phase i metabolism and glycine, glutamine and sulfate
conjugations are typical of phase ii metabolism occurring in the liver. glycine or glutamine
conjugation is pathway- and species-specific (differences may occur between mice, rats and humans)
Microbial metabolite Dietary precursors Specific bacterial species References
Phenolic, benzoyl, and phenyl Phenolic Lactobacillus [34, 43,
derivatives compounds; 44, 75–77]
Hippurate flavonoids; protein Clostridium difficile
Cinnamoyl (glycine) (phenylalanine, Clostridium scatologenes
tyrosine)
Proteus vulgaris
Phenol (sulfate) E. coli [77]
Clostridium bifermentans
Clostridium specticum
Bacteroids fragilis
Bifidobacterium longum
p-Hydroxyphenylacetate Clostridium difficile [77]
Bacteroides ovatus
Bifidobacterium bifidum
Bifidobacterium
adolescentis
Bifidobacterium infantis
Bifidobacterium
pseudolongum
p-Hydroxyphenylpropionate Clostridium bifermentans [77]
Clostridium
paraputrificum
Clostridium specticum
Bacteroides
thetaiotaomicron
p-Cresol (sulfate) Clostridium difficile [77]
Clostridium
paraputrificum
Clostridium perfringens
Bacteroids fragilis
Bacteroides
thetaiotaomicron
Bifidobacterium
adolescentis
Bifidobacterium
pseudolongum
Bacteroides
thetaiotaomicron
(continued)
Table 5.1 (continued)
Phenylacetate (glycine) Clostridium bifermentans [77]
Clostridium difficile
Bacteroids fragilis
Bacteroides ovatus
Phenylpropionate (glycine) Clostridium difficile [77]
Peptostreptococcus
asaccharolyticus
Bacteroides ovatus
Phenyllactate (glycine) Clostridium perfringens [77]
Bacteroides ovatus
Indole derivatives Protein E. coli [43, 44,
Tryptamine (tryptophan) Clostridium bifermentans 65, 78]
Proteus vulgaris
Serotonin Paracolobactrum
coliforme
Achromobacter
liquefaciens
Bacteroides spp
Clostridia
Indole (sulfate) E. coli [77]
Bacteroides ovatus
Indole pyruvate E. coli [77]
Peptostreptococcus
asaccharolyticus
Bacteroides ovatus
Bifidobacterium
adolescentis
Bifidobacterium
pseudolongum
Indole lactate E. coli [77]
Peptostreptococcus
asaccharolyticus
Bacteroides ovatus
Bacteroides
thetaiotaomicron
Bifidobacterium
adolescentis
Bifidobacterium
pseudolongum
(continued)

Indole-3-acetate (glutamine) E. coli [77]
Clostridium difficile
Clostridium
paraputrificum
Peptostreptococcus
asaccharolyticus
Bacteroides fragilis
Bacteroides
thetaiotaomicron
Bifidobacterium
pseudolongum
Indole-3-propionate Clostridium [77]
paraputrificum
Peptostreptococcus
asaccharolyticus
Bacteroides fragilis
Bifidobacterium
adolescentis
Bifidobacterium
pseudolongum
Choline metabolites Choline Bacteroides fragilis [15, 79]
Methylamine Carnitine Clostridium perfringens
Dimethylamine Faecalibacterium
Trimethylamine (-N-oxide) prausnitzii
Flavones Lactobacillus mucosae [43, 80]
Equol (sulfate) Flavonoids Enterococcus faecium
Methyl equol (sulfate) Finegoldia magna
Short-chain fatty acids Bifidobacterium [81]
Acetate Glucose and starch Propionibacterium
Propionate Polysaccharides Lactobacilllus
Butyrate Fiber Clostridium
Branched-chain fatty acids Protein (branched- Bacteroides ruminicola [82]
Isobutyrate chain amino acids: Megasphaera elsdenii
Isovalerate leucine, isoleucine,
valine)
those rich in whole grains can alter gut microbiota populations [7–9] and gut
microbiota metabolism [10, 11]. Of the few metabolomics studies that compared
intake of whole grains with refined grains, one rat study found that urinary hippu-
rate was increased on a whole grain diet, along with the aromatic amino acids phe-
nylalanine, tryptophan, and tyrosine [12]. This finding was not replicated in human
urine samples after a whole grain intervention, though other biomarkers of gut
microbiota activity were decreased on a whole grain diet, including 4-hydroxyphen-
ylacetate, a microbial metabolic product of aromatic amino acid metabolism and
trimethylamine, a microbial metabolic product of choline and carnitine [11]. As in
many areas of nutritional science, results on the impact of whole grains on gut
microbiota are variable, with some studies not finding any changes to gut microbial
species measured [13]. This variation in gut microbial response to an admittedly
broad and heterogeneous food group may explain in part some of the variation in
results between intervention studies, an area that will be further explored as more
advanced microbial sequencing techniques become routine [9].
5.3.3 Fat-Rich Diet Interactions with Gut Microbiota
High-fat diets are frequently used in metabolic studies for testing diet-induced
metabolic syndrome (increased risk of developing cardiovascular disease and diabe-
tes), especially in rodent models. A direct relationship has been established between
high-fat feeding and metabolic disorders, where altered gut flora is causal in induc-
ing gut permeability, increasing lipopolysaccharide (LPS) absorption, and inflam-
mation [14, 15]. Given this association between diet and gut microflora, specific
strategies for modifying gut microbiota may be a useful means to reduce the impact
of high-fat feeding on the occurrence of metabolic diseases. However, as these
results mainly stem from rodent models, where high-fat diets represent a far greater
proportion of energy intake than would normally be found among humans, caution
is required until definitive clinical studies are performed.
5.3.4 Choline Metabolism: An Interaction

Node Between Diet, Host, and Gut Microbiota?
Recent studies from a cohort of non-Caucasians based in Cleveland, United States,

have highlighted that gut microbial metabolism of specific dietary components may
result in toxic metabolites that lead to cardiovascular disease. Using LC-MS metabo-
lomics, Wang et al. found that high plasma concentrations of a microbial metabolite
of choline, trimethylamine (TMA), was related to cardiovascular disease risk, con-
cluding that whether the gut microbiota converted choline into TMA or not was a key
modifiable risk factor for development of cardiovascular disease [16]. The active mol-
ecule mediating increased disease risk was identified as trimethylamine oxide
(TMAO), a toxic metabolite of liver metabolism of TMA. Choline, and its related
metabolite, TMA, and betaine (a downstream mammalian metabolite of choline)
were related to cardiovascular disease in this cohort. These metabolomics results
were cross validated, and biomarkers confirmed in mouse models of cardiovascular
disease, though do not fully explain other findings with the same biomarkers. For
example, a comprehensive study on the sources of TMA in humans found that very
little choline was converted into TMA [17]. While this could be explained by differ-
ences in gut microbiota, the intake of fish and shellfish led to extremely high produc-
tion of TMA [17], with some fish species leading to an excretion of over 4,000 μmol
of TMA and TMAO in urine over 8 h. Meat, eggs, and dairy products conversely did
not lead to more TMA and TMAO excretion compared to a control diet [17]. If these
results in urine are reflected in plasma, any increase in TMA due to nonoptimal gut
microbiota metabolism of choline from fatty foods would be “drowned” out by that
due to fish intake. Thus, TMAO being a biomarker for cardiovascular disease risk
would be at odds with a high intake of fish being associated with a decrease in cardio-
vascular disease risk [18–20], which would suggest that TMAO is not a good bio-
marker of cardiovascular disease in populations where fish intake is common.
Similarly, the finding that elevated betaine may be associated with cardiovascular
disease risk goes against other work finding that betaine is substantially associated
with a decreased risk of cardiovascular disease risk factors [21]. The same study how-
ever also found that plasma choline was associated with risk factors for cardiovascu-
lar disease [21]. Betaine is one of the main phytochemicals present in whole grain
wheat [22], and fasting betaine concentrations can be increased on a whole grain-rich
diet [8], and both oral choline and betaine lead to decreased circulating homocysteine
[23, 24], a cholesterol-independent risk factor for cardiovascular disease. In the con-
text of these other findings, it is possible that elevated choline and TMAO are bio-
markers of cardiovascular disease risk in this population, if fish intake is low.
In a follow-up study using stable isotope-labelled phosphatidylcholine, the role
of gut microbiota in the formation of TMAO from choline was clearly established,
along with choline being the main source of circulating betaine [25]. So in this
population, elevated betaine probably comes from a high intake of choline, rather
than a high intake of betaine-containing foods. The complementary analysis of food
records and use of dietary biomarkers of intake (e.g., alkylresorcinols for whole
grains [26] or omega-3 fatty acids for fish intake [27, 28]) along with gut microbiota
measurements and metabolomics may be instructive for determining if elevated
concentrations of these biomarkers are related to disease risk or diet.
A second study by the same group hypothesized that another TMA, L-carnitine,
may also be a risk factor for cardiovascular disease as it can also be metabolized by
gut microbiota to TMAO [29]. Carnitine, which is a major component of red meat
and is conditionally essential for fatty acid transport for mitochondria, was found to
lead to increased TMAO production that depended on gut microbiota. Of interest
for metabolomic methodology is that in the initial screening of the same cohort
where choline was suggested to be a risk factor for CVD [16], carnitine was only
found to be a significant risk factor if correction for multiplicity was not used in the
statistical analysis. While statistical considerations are important, the possibility
that associations of interest may be lost when using stringent tests should not be
overlooked. A recent study comparing a whole grain diet to a refined grain diet
found that urinary excretion of carnitine and acetylcarnitine was reduced when con-
suming whole grains, along with a decrease in the TMAO precursor TMA [11], in a
study where gut microbial populations were also altered due to the whole grain
intervention [8]. It is clear from this work on precursors of TMAO that while there
may be several complexities in assigning biomarkers to disease risk that are also
derived from diet, the one carbon metabolism pathway and phospholipid metabo-
lism are likely key areas of interaction between diet, physiology, gut microbiota,
and cardiovascular disease.
5.3.5 Protein-Rich Diets and Gut Microbiota
While human protein digestion and amino acid absorption are efficient, some
proteins and free amino acids still reach the colon and are associated to increased
production of potentially toxic substances such as volatile sulfur compounds,
ammonia, and p-cresol [30, 31]. Experimental evidence from animal models and
in vitro data shows that dietary proteins can influence cancer expression. Increased
dietary protein consumption can cause increased colonic DNA damage and thinning
of the colonic mucosal barrier [32]. Production of microbial metabolites from amino
acids can be reduced by dietary fiber (via increasing the proportion of carbohydrate
reaching the colon), as carbohydrate appears to be a preferred substrate for most gut
microbiota species [32].
The molar ratios of the BCFAs isovalerate and isobutyrate, compounds resulting
from the bacterial fermentation of valine and leucine, were found to be increased
relative to total fecal SCFAs with high-protein diets. A marked increase in fecal
nitrogenous organic compounds (NOC) amounts was also found when subjects con-
sumed high-protein diets. NOCs are carcinogens in vitro; although the toxicological
significance of increased fecal NOCs is uncertain, NOCs, at concentrations present
in the colonic lumen, contribute to DNA damage in the colon and rectum and pos-
sibly to increased risk of human cancer [33]. Broadly speaking, much evidence
suggests that interaction between protein and amino acids is negative for the host,
though secondary roles of these metabolites on gene signalling and immune func-
tion have not been researched.
5.3.6 Interaction of a Polyphenol-Rich Diet

with the Gut Microbiota
While not a nutrient in the strict sense, polyphenols, or at least polyphenol-rich

foods may also lead to a change in gut microbiota metabolism, notable examples
being coffee and chocolate [34]. This metabolic interaction may lead to many
downstream effects and it has even been suggested that there is a link between
preference for chocolate and gut microbiota, depending on how cocoa polyphenols
are metabolized [35]. While conceptually it makes sense that people who regularly
consume chocolates have gut microbiota more readily adapted to metabolizing
cocoa polyphenols compared to those who avoid chocolate, it remains an intriguing
question as to whether there are wider effects beyond gut microbial metabolism and
into the realm of “gut-host interactions.” Certainly recent studies in both humans
[36] and rodents [37] clearly demonstrate that cocoa polyphenols can alter gut
microbiota populations. In vitro colon model studies find that cocoa polyphenols
increase butyric acid production and formation of 3-hydroxyphenylpropionic acid
from cocoa flavanols [38]. Consumption of dark chocolates also increases
3-hydroxyphenylpropionic acid and hippurate excretion in urine [35]. As will be
addressed below, phenolic compounds may also be metabolized into hippuric acid
[39], and this convergence with amino acid metabolism may lead to some problems
in interpreting metabolomics data relating to polyphenols and amino acid intake.
To be confident of identifying changes in gut microbial metabolism, several related
changes may need to be identified, preferably with concurrent changes to gut micro-
biota populations.
While it is clear that there is a relationship between diet, gut microbiota, and
certain metabolites resulting from gut microbial metabolism, the link between gut
microbial metabolites and systemic effects remains largely unclear. Are they simply
markers, or also mediators? A number of dietary phenolic compounds act as signal-
ling molecules for regulating various metabolic cascades [40], though no data exists
on whether of the common aromatic metabolites identified as being produced by gut
microbiota have any role in influencing gene expression.
There is much that remains to be studied in terms of the diet and gut microbiota –
protein-rich or sulfur-rich diets have received relatively little attention compared to
high-fat diets or different sources of carbohydrates and prebiotics. Beyond specific
pre- and probiotics, several different food groups are also known to have an effect
on the gut microbiota, with consequent possible effects on gut microbial metabo-
lism, though to date whether these effects have long-term effects on the host is less
clear. This is further complicated by the apparent resistance of gut microbial popu-
lations to long-term change, as demonstrated by fecal transplantation studies, where
host populations frequently revert towards pretransplantation levels [41].
5.4 Nutritional Metabolomics: A Methodology Well Suited

for Understanding the Effects of Food on Gut Microbiota
Metabolomics is the study of multiple metabolites (small molecules, generally

<1,500 Da) in response to different stimuli or conditions and generally involves
the measuring of several to hundreds of metabolites or thousands of features in a
metabolic profile [42]. This is followed by multivariate analysis to determine
what metabolite(s) best explain(s) the research question. Metabolomics is
complementary to other omics such as genomics and proteomics. So while genetics

is often seen as a “blueprint,” genetic disposition is often not reflected in phenotype.
As metabolites are the end product of genotype differences, they reflect how a sys-
tem is responding to different stimuli. Simplistically, metabolites can be seen as the
result of genotype + epigenetic modification + posttranslational modification of
proteins + interaction with the environment. Chapters 2 and 3 elaborate on general
metabolomics methodologies and data modelling.
5.4.1 Metabolomics Methods to Study Gut Activity

Effects on Metabolism
Metabolomics is mainly based on two technologies: nuclear magnetic resonance

(NMR) and mass spectrometry (MS). Among a wide variety of applications, it has
been used in characterizing the metabolic fingerprint of mammalian hosts under
conditions designed to alter the microbial communities in the gastrointestinal tract.
While a wealth of studies have found associations between metabolic patterns and
diseases to (deregulated) gut microbiota, the full biochemical characterization of
the gut microbial activity is yet to be defined. To define the metabolome of the gut
microbiome, Wikoff et al. [43] used an untargeted MS-based strategy to compare
plasma extracts of germ-free mice to conventional mice. Indole-containing metabo-
lites, phenylated-organic acids, and phase II metabolites of these (sulfated and gluc-
uronidated species) were found in conventional mice and either absent or poorly
represented in germ-free animals. The absence of phase II metabolites in germ-free
mice suggests a direct impact of the gut microflora on the drug metabolism capacity
of the host, where interplay between gut (microbes) and liver (mammal) takes place.
Other strategies to investigate the function of the metabolite influenced by the
gut microbiota have included urinary and fecal MS-based profiling of metabolites
from Wistar rats exposed to a broad-spectrum β-lactam antibiotic (imipenem/cilas-
tatin sodium) and were compared before and after exposure [44]. An apparent meta-
bolic switchover is observed within 24 h of antibiotic exposure and affecting a wide
range of central metabolic pathways (mainly amino acid metabolism, but also
organic acid metabolism, oligopeptides, carbohydrate metabolism, purine and
pyrimidine metabolism, and the TCA cycle). Benzene- and indole-containing sub-
stances, including tryptophan and hippuric acid, were dramatically reduced by the
antibiotic treatment.
These two studies [43, 44], using different strategies to remove the influence of
the gut microbiota, lead to consistent results on the chemical nature of metabolites
produced by gut microbiota activity. The fact that different biological matrices were
used for metabolomics analysis (plasma [43] and urine and feces [44]) suggests
comparable effects in all systemic biofluids, at least in rodents.
While most metabolomics studies have focused on metabolite analysis of plasma,
urine, and fecal water, there are other potential methods for understanding gut
microbiotal metabolism that to date have not had widespread use. These include
headspace GC-MS analysis of volatile organic compounds from feces [45], which
can measure up to 90 compounds present in the fecal headspace. Breath analysis
may also be a fruitful area for understanding gut microbiota activity; several breath
analyses are already used in nutrition to measure gut microbiota activity including
breath hydrogen to monitor gut fermentation and the urea breath test for Helicobacter
pylori infection [46]. Gasses produced by gut bacterial fermentation not only exit
via flatus but can also be measured in breath – especially those that are active in the
stomach, as is the case with H. pylori. The lactose breath test is another practical
example, where lactose tolerant people can break down lactose before reaching the
intestine, while in lactose-intolerant people, lactose reaches the intestine and is rap-
idly fermented, leading to the production of hydrogen gas. Hydrogen is normally
present in low concentrations in breath, so any spike in breath hydrogen is clearly
linked to lactose intolerance. The same concept is also used for measuring the
fermentation of dietary fibers by gut microbiota in clinical trials [47].
5.5 Metabolites of Gut Activity
The ensemble of bacterial species in the gut can modulate metabolic reactions
essential to the host’s metabolism and health. There are a set of metabolites that
consistently directly or indirectly stand out in association studies on diseases such
as obesity, insulin resistance, type II diabetes, cancer, cardiovascular disease,
chronic systemic inflammation, and autism and related neurological conditions
[15]. These metabolites include SCFAs, bile acids, choline metabolites, phenolic,
benzoyl, and phenyl derivatives and indole derivatives (Figs. 5.1 and 5.2). Given the
range of conditions where these metabolites may be involved, there is now little
doubt about the contribution of the gut microbiota to host metabolism and
maintenance.
5.5.1 Short-Chain Fatty Acids
Possibly the best known examples of gut microbiota metabolites are the SCFAs ace-
tate, propionate, and butyrate, which result from bacterial fermentation of carbohy-
drates. These are water-soluble and readily absorbed respiratory fuels used by the
colonic epithelial cells (colonocytes) produced by anaerobic bacteria. Luminal fatty
acids are the preferred fuels of colonocytes and the order of preference is SCFAs >
ketone bodies > amino acid > glucose [48]. Butyrate is the preferred source of energy
for colonic epithelial cells. Butyrate is transported into colonocytes, enters the mito-
chondria, and undergoes β oxidation to acetyl-CoA, which enters the TCA cycle
resulting in the reduction of NAD+ to NADH. NADH enters the electron transport
chain culminating in ATP production with CO2 as a by-product [49]. Butyrate is
associated with a decreased risk of colon cancer proliferation, modulation of
Fig. 5.2 Proposed indole signalling and metabolism in the intestine (inspired by [43, 44, 65])
inflammation, and an increase in satiety [50–54]. Absorbed acetate and propionate

are delivered to hepatocytes, which consume most of the propionate for gluconeo-
genesis. Although acetate can be used for lipogenesis in colonocytes, hepatocytes
and adipocytes are the principal sites for de novo lipogenesis, at least in rodents.
SCFAs also act as signalling molecules. Propionate, acetate, and to a lesser extent
butyrate and pentanoate are ligands for at least two G protein-coupled receptors
(GPCRs), Gpr41 and Gpr43. Both GPCRs are broadly expressed, including in the
distal small intestine, colon, and adipocytes. SCFAs (C1–C6), which are ligands for
Gpr41, stimulate expression of leptin in mouse-cultured adipocytes [55]. Leptin is a
polypeptide hormone with pleiotropic effects on appetite and energy metabolism.
This possible link between fiber intake, gut microbiota, and satiety has opened up a
new area of possibilities for nutrition research. Clostridia are saccharolytic and amino
acid fermenting species and are able to produce the three main SCFAs in the colon.
Many other bacteria such as Bacteroides, Eubacteria, and Propionibacteria are known
for producing specific SCFA from various substrate sources [4, 49]. Most SCFAs are
produced from the fermentation of carbohydrates, with smaller amounts by microbial
protein degradation in the large intestine. In addition to providing energy for the
colon, SCFAs are important energy sources for muscle, kidneys, heart, and brain.
5.5.2 Phenolic and Phenyl Metabolites
Hippuric acid is a conjugate of glycine with benzoic acid and is a common end
product of metabolism excreted in urine. Dietary sources of protein and polyphenols
(fruits, vegetables, coffee, tea, chocolate) ultimately lead to the degradation into
quinic acid and benzoic acid by colonic microflora which are then oxidized to hip-
purate by hepatic mitochondrial function, in a CoA-dependent fashion [34, 39, 56].
Therefore, diets rich in protein and polyphenols lead to increased excretion of hip-
purate and other phenol-related metabolites such as p-cresol, phenol, p-hydroxybenzoic
acid, and p-hydroxyphenylacetic acid [34, 57]. p-Hydroxyphenylacetic acid has
been identified as an intermediate of p-cresol production from tyrosine and is ele-
vated in a variety of conditions [34, 57]. Hippurate is possibly the most recognized
gut microbial co-metabolite and has been associated with a variety of conditions or
disease status, such as obesity, high blood pressure, Crohn’s disease and ulcerative
colitis, autism (decreased excretion), type I and II diabetes, and anxiety (increased
excretion) [39]. As an example of the relative amount of hippurate precursors,
Table 5.2 summarizes different foods and their potential to form hippurate. The
amount of potential hippurate from coffee and tea is far greater than the other types
of food listed, most notably dark chocolate, also considered a major source of
polyphenols in the diet. Aromatic amino acids are also potential sources of hippu-
rate, though the amount of these actually reaching the large intestine is potentially
lower than for polyphenols. Caution is needed with the interpretation that elevated
hippuric acid is mainly due to polyphenol-rich foods in the diet, as aromatic amino
acid-rich foods such as meat and fish could lead to elevated levels in cases of protein
malabsorption.
p-Cresol-sulfate is an abundant compound in urine that is obtained from protein
fermentation in the human gut, derived from tyrosine and phenylalanine metabo-
lism. Gut bacteria [57, 58] such as the pathogen Clostridium difficile [59] are able
to convert tyrosine into p-cresol. In humans, p-cresol is almost completely sulfonated
into p-cresol sulfate by SULT1A1 (human cytosolic sulfotransferase) which is an
enzyme able to sulfonate various substrates, including xenobiotics [60]. High
amounts of p-cresol in urine are found in adult celiac disease patients [57]. p-Cresol
can exert a variety of effects such as activation of leukocyte free radical production
[61] and blocking the conversion of dopamine into noradrenaline [62]. It is argued
that given the high amount of p-cresol produced in the body, depending on the diet
and eventual modulation of gut bacterial composition, impaired sulfation and events
thereof (drug metabolism) might take place, depending on the individual [60].
In addition to p-cresol, phenol is also produced in the gut, mostly attributed to aero-
bic bacteria (E. coli, Proteus sp, S. faecalis, Staphylococcus albus), while p-cresol
is produced by anaerobic bacteria [58]. As anaerobic bacteria outnumber aerobic
bacteria in the gastrointestinal tract, it is expected that there is greater excretion of
p-cresol than phenol.
Phenylacetylglutamine is derived from β-phenylethylamine formed in the large
bowel by decarboxylation of phenylalanine released by bacterial proteolysis of
98
Table 5.2 Relative amounts of aromatic amino acids and common phenolic precursors for hippurate in different foods. Several of these precursors may be
metabolized through different pathways, so this table should only be considered as a relative indication of the hippuric acid potential of different foods
Catechin Total hippurate
Chlorogenic and mole equiva- Total glycine
Phenylalanine Tyrosine Glycine acids Caffeic acid epicatechin lents equivalents
(g/100 g) (g/100 g) (g/100 g) (mg/100 g) (mg/100 g) (mg/100 g) (mmol/100 g) (mmol/100 g) References
Milk 0.34 0.23 0.17 – – 67 3.56 2.26 [83–85]
chocolate
Dark 0.2 0.074 0.18 – – 200 2.31 2.40 [83–85]
chocolate
Fish 0.86 0.72 0.99 – – – 9.18 13.19 [83]
(smoked
salmon)
Fish 0.82 0.73 0.92 – – – 8.99 12.26 [83]
(smoked
herring)
Beef (rump 0.75 0.62 1.1 – – – 7.96 14.65 [83]
steak)
Chicken 0.74 0.61 1.1 – – – 7.85 14.65 [83]
(breast
meat)
S. Moco and A.B. Ross
5
Refined 0.51 0.22 0.35 – – – 4.30 4.66 [83]

wheat
Whole grain 0.52 0.29 0.48 – – – 4.75 6.39 [83]
wheat
Refined rice 0.45 0.24 0.38 – – – 4.05 5.06 [83]
Brown rice 0.48 0.26 0.41 – 0.3 6 4.36 5.46 [83, 86]
Coffee 0.24 0.16 0.42 8,800 1,740 – 36.83 5.59 [83, 87, 88]
(soluble,
dry)
Tea (black, Trace Trace Trace 100 56.8 3,700 13.34 0.00 [87, 89]
leaves)
Tea (green, Trace Trace Trace 150 52 5,960 21.24 0.00 [87, 89]
leaves)
Trace: Trace amounts found, but close to methodological limit of detection
–: No data found, probably not present in foodstuff
Can We Use Metabolomics to Understand Changes to Gut Microbiota Populations…
99
unabsorbed protein residues [63]. Indoxyl sulfate and phenylacetylglutamine have

been found in higher concentrations in the plasma of diabetic individuals compared
to nondiabetics. Abnormal urinary excretion of phenylacetylglutamine, hippurate,
and hydroxyhippurates has been reported in autistic children [64] (see Chap. 16).
5.5.3 Indole Metabolites
Copious amounts of indole are produced by the human body and ultimately excreted
in urine, in the form of indoxyl sulfate. Indole has been thus associated to gut micro-
bial activity and is produced by tryptophanase (TnA) that reversibly converts
tryptophan into indole, pyruvate, and ammonia [65]. The elimination of tryptophan,
instead of indole, in urine can be associated to altered microbial activity in the gut.
Over 85 bacterial species are known as indole-producing bacteria [65] and in the
gut, indole is a signalling molecule recognized by intestinal epithelial cells and is
used to strengthen the host cell barrier, maintain controlled inflammation, and
increase resistance to pathogen colonization [66]. It is not known if E. coli is able to
degrade indole, but many non-indole-producing bacteria encode various oxygen-
ases that can modify or degrade indole, producing indole derivatives, such as indi-
goid compounds [43, 65]. Independent from gut microbial activity, indole can be
further modified into sulfated, glucuronidated, and fatty acid-conjugated species,
such as indoxyl sulfate, indoxyl glucuronide, and indole-3-propionic acid (IPA) and
indole-2-acetic acid (IAA) [43] (Fig. 5.2). Indole has been compared to the known
autoinducer-2 (AI-2), a furanosyl borate diester, from the family of signalling mol-
ecules used in quorum sensing, although it is still unclear how the roles of two
molecules are connected to each other [65].
Tryptamine is another metabolite in the tryptophan metabolism that is decarbox-
ylated by mammalian L-tryptophan decarboxylyase from tryptophan, as well as
degraded into indole-3-acetaldehyde by gut bacteria. Low tryptamine levels in urine
have been detected in patients with severe depression [67], while high levels of this
molecule in urine and feces have been found in antibiotic-treated subjects [44].
Abnormal tryptophan metabolism is indicated in cognitive disorders, such as
depression, and 5-hydroxy-L-tryptophan has been used clinically for decades to
increase serotonin production in the brain [68].
IAA is a known phytohormone, which belongs to the auxin class of compounds
that regulates cell growth and development. Diverse bacterial strains produce IAA,
especially endophytes, which signal biofilm formation. In the gut, indoles have been
described to lead to biofilm formation [65] and regulation of virulence in vitro and
in vivo [69] and specifically IAA has been identified as a marker of gut activity [44]
and enhancer of cellular defense [70]. Thus, it could be speculated that the indole
class of compounds may act as inter-kingdom signalling molecules regulating mam-
malian, bacterial, and plant signalling.
IPA is a potent antioxidant and neuroprotective molecule. IPA completely
protected primary neurons and neuroblastoma cells against oxidative damage and
death caused by exposure to Alzheimer β-amyloid protein, by inhibition of

superoxide dismutase, or by treatment with hydrogen peroxide [71]. The capacity of
IPA to scavenge hydroxyl radicals exceeded that of melatonin, an indoleamine con-
sidered to be the most potent naturally occurring scavenger of free radicals [71].
In addition, 6-hydroxymelatonin, an oxidation product of melatonin was also
identified as a marker of gut activity, as well as other known neurotransmitters such
as DOPA, dopamine, norepinephrine, and epinephrine which play important roles in
the brain [44]. Several stress mechanisms have been correlated to alteration of bac-
terial composition of the gastrointestinal tract (GI), altering epithelial cell function,
motility, and mucus secretion. Upon stress, norepinephrine is released into the GI
tract, potentially altering gut microbial composition and function [72]. These find-
ings are evidence of the strong association of microbial-mammalian metabolism to
the gut-brain axis.
5.5.4 Choline Metabolites
Eggs, milk, liver, red meat, poultry, shell fish, and fish are natural sources of phos-
phatidylcholine and choline. As described earlier, microbial conversion of dietary
phosphatidylcholine and choline (or betaine) leads to the production of TMA in the
gut which is oxidized in the liver to TMAO by the hepatic flavin monooxygenase
(FMO) family of enzymes, FMO3. A study on mice has shown that dietary supple-
mentation with choline, TMAO, or betaine was found to promote upregulation of
multiple receptors linked to atherosclerosis [16]. Increased levels of TMAO were
also associated with nonalcoholic fatty liver disease [73].
5.6 Gut-Liver Interplay
A portion of the products of colonic fermentation are absorbed by the colonocytes

and via specific pathways lead to the production of ATP, while others undergo bio-
transformation in the liver entering phase I and phase II types of metabolism
(Fig. 5.3). Especially for phenyl, phenol, and indole derivatives, sulfation, gluc-
uronidation, and glycine conjugation occur and have been described [43]. Most
commonly these metabolites are more water soluble and increased polarity of con-
jugates may limit passive partitioning into cells, thus increasing excretion.
Biotransformation capability of the host is dependent on several factors, includ-
ing age, gender, genetic variability, nutrition, disease, exposure to other chemicals
that can inhibit or induce enzymes, and dose levels. For instance, the elderly shows
decreased biotransformation capabilities and gender may also influence the effi-
ciency of biotransformation for specific metabolites or xenobiotics, as this is usually
limited to hormone-related differences in the oxidizing cytochrome P-450 enzymes.
This area is especially deserving of attention as we seek to further our understanding
of what role a dynamic gut microbiota may play in the host aging process.
Fig. 5.3 Overview of the interaction between host organs and gut microbiota
5.7 Future Directions
Gut microbiota has relevance for human health and disease beyond the gastrointes-
tinal tract, appearing to have a systemic impact on human metabolism, through
interaction with multiple organs. The gut microflora has proven causality in the
induction of some metabolic disorders (see Chap. 12), and therapies that target the
gut microbiota are at the forefront of nutrition research. Modulation of the gut
microbiota is potentially attainable by altering dietary habits; however, we are still
far away from understanding either general effects of macronutrients or specific
effects of ingredients on gut microbiotal metabolism. There are cautionary tales
too – while it is tempting to propose a role for gut microbiota in all observed health
benefits related to food, this is not always the case. In one study, cereal fiber changed
gut microbiota composition, but there was no association between these changes
and an observed improvement in insulin sensitivity [74]. It is possible that more
focus is needed on microbial metabolism, rather than population shifts, an area
where metabolomics may be a particularly useful methodology for helping to find
answers.
While metabolomics has been instrumental as an exploratory tool to fuel ideas
and propose novel hypotheses, we believe that strategies to monitor the gut micro-
bial metabolome will be crucial to define gut activity and its impact on metabolism.
To achieve this, targeted metabolomics methods should be implemented to follow
the different classes of gut microbial metabolites in health and disease. The quanti-
tation of metabolites will becoming increasingly important to define the kinetics of
metabolic fluxes, and to determine mechanisms of action and their association with
functionality.
Studying the potential activity of the gut cannot be deduced by solely looking at
fecal samples, as fecal transit can vary considerably (12–120 h) and bacterial gradi-
ents in the colon exist and thus fecal samples may only be a poor approximation of
metabolism along the colon. Gut microbiota metabolites seem to be not only prod-
ucts of digestion, and therefore simpler molecules to be either taken up as energy or
discarded by the host, but also signalling functions are being unraveled that prospect
a more complex interplay between microbiota and host. It is clear that in terms of
our knowledge on the relationship between the diet and nutrition of the host, and our
gut microbiota, we are at the beginning of an area that will have a profound impact
on our current understanding of human nutrition.
References
1. Salminen S, Bouley C, Boutron-Ruault MC, Cummings JH, Franck A, Gibson GR, Isolauri E,
Moreau MC, Roberfroid M, Rowland I. Functional food science and gastrointestinal physiol-
ogy and function. Br J Nutr. 1998;80 Suppl 1:S147–71.
2. De Filippo C, Cavalieri D, Di Paola M, Ramazzotti M, Poullet JB, Massart S, Collini S,
Pieraccini G, Lionetti P. Impact of diet in shaping gut microbiota revealed by a comparative
study in children from Europe and rural Africa. Proc Natl Acad Sci U S A. 2010;107(33):
14691–6.
3. Wu GD, Chen J, Hoffmann C, Bittinger K, Chen Y-Y, Keilbaugh SA, Bewtra M, Knights D,
Walters WA, Knight R, Sinha R, Gilroy E, Gupta K, Baldassano R, Nessel L, Li H, Bushman
FD, Lewis JD. Linking long-term dietary patterns with gut microbial enterotypes. Science.
2011;334(6052):105–8.
4. Macfarlane GT, Cummings JH, Allison C. Protein degradation by human intestinal bacteria. J
Gen Microbiol. 1986;132(6):1647–56.
5. Kruis W, Forstmaier G, Scheurlen C, Stellaard F. Effect of diets low and high in refined sugars
on gut transit, bile acid metabolism, and bacterial fermentation. Gut. 1991;32(4):367–71.
6. van der Kamp JW, Poutanen K, Seal CJ, Richardson DP. The HEALTHGRAIN definition of
‘whole grain’. Food Nutr Res. 2014;58:22100.
7. Costabile A, Klinder A, Fava F, Napolitano A, Fogliano V, Leonard C, Gibson GR, Tuohy
KM. Whole-grain wheat breakfast cereal has a prebiotic effect on the human gut microbiota: a
double-blind, placebo-controlled, crossover study. Br J Nutr. 2008;99(1):110–20.
8. Ross AB, Bruce SJ, Blondel-Lubrano A, Oguey-Araymon S, Beaumont M, Bourgeois A,

Nielsen-Moennoz C, Vigo M, Fay LB, Kochhar S, Bibiloni R, Pittet AC, Emady-Azar S,
Grathwohl D, Rezzi S. A whole-grain cereal-rich diet increases plasma betaine, and tends to
decrease total and LDL-cholesterol compared with a refined-grain diet in healthy subjects. Br
J Nutr. 2011;105(10):1492–502.
9. Martínez I, Lattimer JM, Hubach KL, Case JA, Yang J, Weber CG, Louk JA, Rose DJ,
Kyureghian G, Peterson DA, Haub MD, Walter J. Gut microbiome composition is linked to
whole grain-induced immunological improvements. ISME J. 2012;7(2):269–80.
10. McIntosh GH, Noakes M, Royle PJ, Foster PR. Whole-grain rye and wheat foods and markers
of bowel health in overweight middle-aged men. Am J Clin Nutr. 2003;77(4):967–74.
11. Ross AB, Pere-Trepat E, Montoliu I, Martin FP, Collino S, Moco S, Godin JP, Cleroux M, Guy
PA, Breton I, Bibiloni R, Thorimbert A, Tavazzi I, Tornier L, Bebuis A, Bruce SJ, Beaumont
M, Fay LB, Kochhar S. A whole-grain-rich diet reduces urinary excretion of markers of protein
catabolism and gut microbiota metabolism in healthy men after one week. J Nutr.
2013;143(6):766–73.
12. Fardet A, Canlet C, Gottardi G, Lyan B, Llorach R, Remesy C, Mazur A, Paris A, Scalbert
A. Whole-grain and refined wheat flours show distinct metabolic profiles in rats as assessed by
a 1H NMR-based metabonomic approach. J Nutr. 2007;137(4):923–9.
13. Lappi J, Kolehmainen M, Mykkanen H, Poutanen K. Do large intestinal events explain the
protective effects of whole grain foods against type 2 diabetes? Crit Rev Food Sci Nutr.
2013;53(6):631–40.
14. Cani PD, Bibiloni R, Knauf C, Waget A, Neyrinck AM, Delzenne NM, Burcelin R. Changes in
gut microbiota control metabolic endotoxemia-induced inflammation in high-fat diet-induced
obesity and diabetes in mice. Diabetes. 2008;57(6):1470–81.
15. Nicholson JK, Holmes E, Kinross J, Burcelin R, Gibson G, Jia W, Pettersson S. Host-gut
microbiota metabolic interactions. Science. 2012;336(6086):1262–7.
16. Wang Z, Klipfell E, Bennett BJ, Koeth R, Levison BS, Dugar B, Feldstein AE, Britt EB, Fu X,
Chung YM, Wu Y, Schauer P, Smith JD, Allayee H, Tang WH, DiDonato JA, Lusis AJ, Hazen
SL. Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease. Nature.
2011;472(7341):57–63.
17. Zhang AQ, Mitchell SC, Smith RL. Dietary precursors of trimethylamine in man: a pilot study.
Food Chem Toxicol. 1999;37(5):515–20.
18. Takata Y, Zhang X, Li H, Gao YT, Yang G, Gao J, Cai H, Xiang YB, Zheng W, Shu XO. Fish
intake and risks of total and cause-specific mortality in 2 population-based cohort studies of
134,296 men and women. Am J Epidemiol. 2013;178(1):46–57.
19. Li YH, Zhou CH, Pei HJ, Zhou XL, Li LH, Wu YJ, Hui RT. Fish consumption and incidence
of heart failure: a meta-analysis of prospective cohort studies. Chi Med
J. 2013;126(5):942–8.
20. Raatz SK, Silverstein JT, Jahns L, Picklo MJ. Issues of fish consumption for cardiovascular
disease risk reduction. Nutrients. 2013;5(4):1081–97.
21. Konstantinova SV, Tell GS, Vollset SE, Nygård O, Bleie Ø, Ueland PM. Divergent associations
of plasma choline and betaine with components of metabolic syndrome in middle age and
elderly men and women. J Nutr. 2008;138(5):914–20.
22. Bruce SJ, Guy PA, Rezzi S, Ross AB. Quantitative measurement of betaine and free choline in
plasma, cereals and cereal products by isotope dilution LC-MS/MS. J Agric Food Chem.
2010;58(4):2055–61.
23. Zeisel SH, Caudill MA. Choline. Adv Nutr. 2010;1:46–8.
24. Olthof MR, Van Vliet T, Boelsma E, Verhoef P. Low dose betaine supplementation leads to
immediate and long term lowering of plasma homocysteine in healthy men and women. J Nutr.
2003;133(12):4135–8.
25. Tang WHW, Wang Z, Levison BS, Koeth RA, Britt EB, Fu X, Wu Y, Hazen SL. Intestinal
microbial metabolism of phosphatidylcholine and cardiovascular risk. N Engl J Med.
2013;368(17):1575–84.
26. Ross AB. Present status and perspectives on the use of alkylresorcinols as biomarkers of
wholegrain wheat and rye intake. J Nutr Metab. 2012;12:46297.
27. Markhus MW, Graff IE, Dahl L, Seldal CF, Skotheim S, Braarud HC, Stormark KM, Malde
MK. Establishment of a seafood index to assess the seafood consumption in pregnant women.
Food Nutr Res. 2013;57:19272.
28. Chien KL, Lee MS, Tsai YT, Chen PR, Lin HJ, Hsu HC, Lee YT, Chen MF. A Taiwanese food
frequency questionnaire correlates with plasma docosahexaenoic acid but not with plasma
eicosapentaenoic acid levels: questionnaires and plasma biomarkers. BMC Med Res Methodol.
2013;13:23.
29. Koeth RA, Wang Z, Levison BS, Buffa JA, Org E, Sheehy BT, Britt EB, Fu X, Wu Y, Li L,
Smith JD, Didonato JA, Chen J, Li H, Wu GD, Lewis JD, Warrier M, Brown JM, Krauss RM,
Tang WH, Bushman FD, Lusis AJ, Hazen SL. Intestinal microbiota metabolism of l-carnitine,
a nutrient in red meat, promotes atherosclerosis. Nat Med. 2013;19(5):576–85.
30. Geypens B, Claus D, Evenepoel P, Hiele M, Maes B, Peeters M, Rutgeerts P, Ghoos Y. Influence
of dietary protein supplements on the formation of bacterial metabolites in the colon. Gut.
1997;41(1):70–6.
31. Windey K, de Preter V, Verbeke K. Relevance of protein fermentation to gut health. Mol Nutr
Food Res. 2012;56(1):184–96.
32. Le Leu RK, Brown IL, Hu Y, Morita T, Esterman A, Young GP. Effect of dietary resistant
starch and protein on colonic fermentation and intestinal tumourigenesis in rats. Carcinogenesis.
2007;28(2):240–5.
33. Russell WR, Gratz SW, Duncan SH, Holtrop G, Ince J, Scobbie L, Duncan G, Johnstone AM,
Lobley GE, Wallace RJ, Duthie GG, Flint HJ. High-protein, reduced-carbohydrate weight-loss
diets promote metabolite profiles likely to be detrimental to colonic health. Am J Clin Nutr.
2011;93(5):1062–72.
34. Moco S, Martin FP, Rezzi S. Metabolomics view on gut microbiome modulation by polyphenol-
rich foods. J Proteome Res. 2012;11(10):4781–90.
35. Martin FPJ, Montoliu I, Nagy K, Moco S, Collino S, Guy P, Redeuil K, Scherer M, Rezzi S,
Kochhar S. Specific dietary preferences are linked to differing gut microbial metabolic activity
in response to dark chocolate intake. J Proteome Res. 2012;11(12):6252–63.
36. Tzounis X, Rodriguez-Mateos A, Vulevic J, Gibson GR, Kwik-Uribe C, Spencer JP. Prebiotic
evaluation of cocoa-derived flavanols in healthy humans by using a randomized, controlled,
double-blind, crossover intervention study. Am J Clin Nutr. 2011;93(1):62–72.
37. Massot-Cladera M, Perez-Berezo T, Franch A, Castell M, Perez-Cano FJ. Cocoa modulatory
effect on rat faecal microbiota and colonic crosstalk. Arch Biochem Biophys. 2012;527(2):
105–12.
38. Fogliano V, Corollaro ML, Vitaglione P, Napolitano A, Ferracane R, Travaglia F, Arlorio M,
Costabile A, Klinder A, Gibson G. In vitro bioaccessibility and gut biotransformation of poly-
phenols present in the water-insoluble cocoa fraction. Mol Nutr Food Res. 2011;55 Suppl
1:S44–55.
39. Lees HJ, Swann JR, Wilson ID, Nicholson JK, Holmes E. Hippurate: the natural history of a
mammalian-microbial cometabolite. J Proteome Res. 2013;12(4):1527–46.
40. Mitjavila MT, Moreno JJ. The effects of polyphenols on oxidative stress and the arachidonic
acid cascade. Implications for the prevention/treatment of high prevalence diseases. Biochem
Pharmacol. 2012;84(9):1113–22.
41. Brandt LJ, Aroniadis OC. An overview of fecal microbiota transplantation: techniques, indica-
tions, and outcomes. Gastrointest Endosc. 2013;78(2):240–249.
42. Moco S, Bino RJ, De Vos RCH, Vervoort J. Metabolomics technologies and metabolite identi-
fication. Trends Anal Chem. 2007;26:855–66.
43. Wikoff WR, Anfora AT, Liu J, Schultz PG, Lesley SA, Peters EC, Siuzdak G. Metabolomics
analysis reveals large effects of gut microflora on mammalian blood metabolites. Proc Natl
Acad Sci U S A. 2009;106(10):3698–703.
44. Zheng X, Xie G, Zhao A, Zhao L, Yao C, Chiu NH, Zhou Z, Bao Y, Jia W, Nicholson JK. The
footprints of gut microbial-mammalian co-metabolism. J Proteome Res. 2011;10(12):5512–22.
45. Dixon E, Clubb C, Pittman S, Ammann L, Rasheed Z, Kazmi N, Keshavarzian A, Gillevet P,
Rangwala H, Couch RD. Solid-phase microextraction and the human fecal VOC metabolome.
PLoS One. 2011;6(4):e18471.
46. Parra MD, Martinez JA. Nutritional aspects of breath testing based on stable isotopes. Nutr
Rev. 2006;64(7 Pt 1):338–47.
47. Rosén LAH, Silva LOB, Andersson UK, Holm C, Östman EM, Björck IM. Endosperm and
whole grain rye breads are characterized by low post-prandial insulin response and a beneficial
blood glucose profile. Nutr J. 2009;8:42.
48. Roediger WE, Moore A. Effect of short-chaim fatty acid on sodium absorption in isolated
human colon perfused through the vascular bed. Dig Dis Sci. 1981;26(2):100–6.
49. Donohoe DR, Garge N, Zhang X, Sun W, O’Connell TM, Bunger MK, Bultman SJ. The
microbiome and butyrate regulate energy metabolism and autophagy in the mammalian colon.
Cell Metab. 2011;13(5):517–26.
50. Mansour A, Hosseini S, Larijani B, Pajouhi M, Mohajeri-Tehrani MR. Nutrients related to
GLP1 secretory responses. Nutrition. 2013;29(6):813–20.
51. Viladomiu M, Hontecillas R, Yuan L, Lu P, Bassaganya-Riera J. Nutritional protective mecha-
nisms against gut inflammation. J Nutr Biochem. 2013;24(6):929–39.
52. Kaczmarczyk MM, Miller MJ, Freund GG. The health benefits of dietary fiber: beyond the
usual suspects of type 2 diabetes mellitus, cardiovascular disease and colon cancer. Metab Clin
Exp. 2012;61(8):1058–66.
53. Slavin J. Fiber and prebiotics: mechanisms and health benefits. Nutrients. 2013;5(4):
1417–35.
54. Caricilli AM, Saad MJA. The role of gut microbiota on insulin resistance. Nutrients.
2013;5(3):829–51.
55. Samuel BS, Shaito A, Motoike T, Rey FE, Backhed F, Manchester JK, Hammer RE, Williams
SC, Crowley J, Yanagisawa M, Gordon JI. Effects of the gut microbiota on host adiposity are
modulated by the short-chain fatty-acid binding G protein-coupled receptor, Gpr41. Proc Natl
Acad Sci. 2008;105(43):16767–72.
56. Cummings JH, Hill MJ, Bone ES, Branch WJ, Jenkins DJ. The effect of meat protein and
dietary fiber on colonic function and metabolism. II. Bacterial metabolites in feces and urine.
Am J Clin Nutr. 1979;32(10):2094–101.
57. Lord RS, Bralley JA. Clinical applications of urinary organic acids. Part 2: Dysbiosis markers.
Alternat Med Rev. 2008;13(4):292–306.
58. Bone E, Tamm A, Hill M. The production of urinary phenols by gut bacteria and their possible
role in the causation of large bowel cancer. Am J Clin Nutr. 1976;29(12):1448–54.
59. Dawson LF, Donahue EH, Cartman ST, Barton RH, Bundy J, McNerney R, Minton NP, Wren
BW. The analysis of para-cresol production and tolerance in clostridium difficile 027 and 012
strains. BMC Microbiol. 2011;11:86.
lism. Proc Natl Acad Sci U S A. 2009;106(34):14728–33.
61. Schepers E, Meert N, Glorieux G, Goeman J, Van der Eycken J, Vanholder R. P-cresylsulphate,
the main in vivo metabolite of p-cresol, activates leucocyte free radical production. Nephrol
Dial Transplant. 2007;22(2):592–6.
62. DeWolf WE, Carr SA, Varrichio A, Goodhart PJ, Mentzer MA, Roberts GD, Southan C, Dolle
RE, Kruse LI. Inactivation of dopamine Beta-hydroxylase by p-cresol: isolation and character-
ization of covalently modified active site peptides. Biochemistry. 1988;27(26):9093–101.
63. Seakins JW. The determination of urinary phenylacetylglutamine as phenylacetic acid. Studies
on its origin in normal subjects and children with cystic fibrosis. Clin Chem Acta.
1971;35(1):121–31.
64. Holmes E, Li JV, Athanasiou T, Ashrafian H, Nicholson JK. Understanding the role of gut
microbiome-host metabolic signal disruption in health and disease. Trends Microbiol.
2011;19(7):349–59.
65. Lee JH, Lee J. Indole as an intercellular signal in microbial communities. FEMS Microbiol
Rev. 2010;34(4):426–44.
66. Bansal T, Alaniz RC, Wood TK, Jayaraman A. The bacterial signal indole increases epithelial-
cell tight-junction resistance and attenuates indicators of inflammation. Proc Natl Acad Sci U
S A. 2010;107(1):228–33.
67. Coppen A, Shaw DM, Malleson A, Eccleston E, Gundy G. Tryptamine metabolism in depres-
sion. Br J Psychiatry J Ment Sci. 1965;111(479):993–8.
68. Metzner L, Kottra G, Neubert K, Daniel H, Brandsch M. Serotonin, l-tryptophan, and trypt-
amine are effective inhibitors of the amino acid transport system PAT1. FASEB
J. 2005;19(11):1468–73.
69. Bommarius B, Anyanful A, Izrayelit Y, Bhatt S, Cartwright E, Wang W, Swimm AI, Benian
GM, Schroeder FC, Kalman D. A family of indoles regulate virulence and Shiga toxin produc-
tion in pathogenic E. coli. PLoS One. 2013;8(1):e54456.
70. Bianco C, Imperlini E, Calogero R, Senatore B, Amoresano A, Carpentieri A, Pucci P, Defez
R. Indole-3-acetic acid improves Escherichia coli’s defences to stress. Arch Microbiol.
2006;185(5):373–82.
71. Chyan YJ, Poeggeler B, Omar RA, Chain DG, Frangione B, Ghiso J, Pappolla MA. Potent
neuroprotective properties against the Alzheimer beta-amyloid by an endogenous melatonin-
related indole structure, indole-3-propionic acid. J Biol Chem. 1999;274(31):21937–42.
72. Collins SM, Bercik P. The relationship between intestinal microbiota and the central nervous
system in normal gastrointestinal function and disease. Gastroenterology. 2009;136(6):
2003–14.
73. Dumas ME, Barton RH, Toye A, Cloarec O, Blancher C, Rothwell A, Fearnside J, Tatoud R,
Blanc V, Lindon JC, Mitchell SC, Holmes E, McCarthy MI, Scott J, Gauguier D, Nicholson
JK. Metabolic profiling reveals a contribution of gut microbiota to fatty liver phenotype in
insulin-resistant mice. Proc Natl Acad Sci U S A. 2006;103(33):12511–6.
74. Weickert MO, Arafat AM, Blaut M, Alpert C, Becker N, Leupelt V, Rudovich N, Möhlig M,
Pfeiffer AF. Changes in dominant groups of the gut microbiota do not explain cereal-fiber
induced improvement of whole-body insulin sensitivity. Nutr Metab. 2011;8:90.
75. Ward LA, Johnson KA, Robinson IM, Yokoyama MT. Isolation from swine feces of a bacte-
rium which decarboxylates p-hydroxyphenylacetic acid to 4-methylphenol (p-cresol). Appl
Environ Microbiol. 1987;53(1):189–92.
76. Yokoyama MT, Carlson JR. Production of Skatole and para-Cresol by a Rumen Lactobacillus
sp. Appl Environ Microbiol. 1981;41(1):71–6.
77. Smith EA, Macfarlane GT. Enumeration of human colonic bacteria producing phenolic and
indolic compounds: effects of pH, carbohydrate availability and retention time on dissimila-
tory aromatic amino acid metabolism. J Appl Bacteriol. 1996;81(3):288–302.
78. Keszthelyi D, Troost FJ, Masclee AA. Understanding the role of tryptophan and serotonin
metabolism in gastrointestinal function. Neurogastroenterol Motil. 2009;21(12):1239–49.
79. Allison C, Macfarlane GT. Influence of pH, nutrient availability, and growth rate on amine
production by Bacteroides fragilis and Clostridium perfringens. Appl Environ Microbiol.
1989;55(11):2894–8.
80. Decroos K, Vanhemmens S, Cattoir S, Boon N, Verstraete W. Isolation and characterisation of
an equol-producing mixed microbial culture from a human faecal sample and its activity under
gastrointestinal conditions. Arch Microbiol. 2005;183(1):45–55.
81. Nyangale EP, Mottram DS, Gibson GR. Gut microbial activity, implications for health and
disease: the potential role of metabolite analysis. J Proteome Res. 2012;11(12):5573–85.
82. Allison MJ. Production of branched-chain volatile fatty acids by certain anaerobic bacteria.
Appl Environ Microbiol. 1978;35(5):872–7.
83. Saxholt E, Christensen AT, Møller A, Hartkopp HB, Hess Ygil K, Hels OH. Fødevaredatabanken,
version 7 (Danish Food Database, version 7). Afdelning for Enæring, Fødevareinstituttet,
Danmarks Tekniske Universitet (2008). Accessed 6 May 2013.
84. Cooper KA, Campos-Gimenez E, Jimenez Alvarez D, Rytz A, Nagy K, Williamson
G. Predictive relationship between polyphenol and nonfat cocoa solids content of chocolate. J
Agric Food Chem. 2008;56(1):260–5.
85. Actis-Goretta L, Ottaviani JI, Fraga CG. Inhibition of angiotensin converting enzyme activity
by flavanol-rich foods. J Agric Food Chem. 2006;54(1):229–34.
86. Irakli MN, Samanidou VF, Biliaderis CG, Papadoyannis IN. Simultaneous determination of
phenolic acids and flavonoids in rice using solid-phase extraction and RP-HPLC with photodi-
ode array detection. J Sep Sci. 2012;35(13):1603–11.
87. Mattila P, Hellstrom J, Torronen R. Phenolic acids in berries, fruits, and beverages. J Agric
Food Chem. 2006;54(19):7193–9.
88. Renouf M, Marmet C, Guy P, Fraering AL, Longet K, Moulin J, Enslen M, Barron D, Cavin
C, Dionisi F, Rezzi S, Kochhar S, Steiling H, Williamson G. Nondairy creamer, but not milk,
delays the appearance of coffee phenolic acid equivalents in human plasma. J Nutr.
2010;140(2):259–63.
89. Kyle JA, Morrice PC, McNeill G, Duthie GG. Effects of infusion time and addition of milk on
content and absorption of polyphenols from black tea. J Agric Food Chem. 2007;55(12):
4889–94.
Chapter 6
Automated Annotation of Microbial
and Human Flavonoid-Derived Metabolites
Velitchka V. Mihaleva, Fatma Yelda Ünlü, Jacques Vervoort, and Lars Ridder
Abstract Flavonoids are a class of natural compounds essentially produced by

plants that are part of animal and human diets and have assumed health-promoting
benefits. Upon human consumption, these flavonoids are to a modest extent absorbed
in the small intestines. The major part arrives in the colon where the microflora
utilises and converts the flavonoids to a wide range of products. Many of these prod-
ucts are absorbed in the major intestines and subsequently metabolised by the host.
To understand the impact of the microflora on the metabolism and possible effects
on human health, complete (and quantitative) identification of the microbial as well
as human metabolic conversion products of flavonoids is required. This is a chal-
lenging task, as these bioconversion products are often present in relatively small
amounts, making classical identification strategies based on (accurate) mass infor-
mation or nuclear magnetic resonance, not straightforward. In the absence of refer-
ence compounds, annotation of a component may be achieved by detailed expert
evaluation, e.g. by searching for similar fragmentation patterns in spectral databases
of known compounds. However, such manual analysis is a tedious task, and in
advanced metabolite profiling experiments, with large numbers of unknown metab-
olites, this is a major bottleneck. Therefore, new strategies are needed for quick and
reliable identification of the diverse range of molecules in complex matrices (faeces,
blood, urine). Intelligent software for annotation and identification of unknowns is
crucial to fully exploit complex datasets. We developed a new software tool
(MAGMA) for (sub)structure-based annotation of LC-MSn datasets which,
V.V. Mihaleva • J. Vervoort (*)

Laboratory of Biochemistry, Wageningen University and Research Center,
Dreijenlaan 3, 6703 HA Wageningen, The Netherlands
Netherlands Metabolomics Centre, Einsteinweg 55, 2333 CC Leiden, The Netherlands
F.Y. Ünlü
L. Ridder
Netherlands eScience Center, Science Park 140, 1098 XG Amsterdam, The Netherlands

DOI 10.1007/978-1-4471-6539-2_6
110 V.V. Mihaleva et al.
combined with a newly established database for phenolic molecules (MetIDB),

enables semiautomated identification of flavonoid derivatives.
Keywords Flavonoids • Identification • Automation • LC-MS • NMR • Microflora

• MAGMA • PERCH NMR Software • Microbiota • MetIDB • Profiling • Metabolites
• Glucuronide • Sulphate • Valerolactone • Hippuric acid • Epicatechin • Beta oxida-
tion • Alpha oxidation • Lactone hydrolysis • Gallocatechins • Urolithin
6.1 Introduction
Flavonoids are a class of compounds with diverse biological functions in plants and
animals. They are almost exclusively biosynthesised in plants and are involved in
many biologically relevant functions [1]. Flavonoids are probably most commonly
known for their antioxidant activity and assumed health-promoting benefits such as
anti-proliferative and antitumour behaviour. The consumption of flavonoid-
containing food products (e.g. vegetables, fruit, soybean, olive oil) and drinks (e.g.
tea, wine) has been subject of numerous reflections about the link between dietary
habits and health effects [2]. Although flavonoids are usually considered to have
positive health effects, they have also been assigned potential negative properties.
For example, isoflavones were shown to interfere with estradiol binding to the estro-
gen receptor, and catechins and flavonols were shown to be able to bind to DNA and
RNA [3, 4].
About 6,000 flavonoids have been identified from natural sources [5, 6], but due
to the known diversity in the plant kingdom, many more are bound to exist. In addi-
tion, a large range of flavonoids metabolites are formed by microbial, animal and
human biotransformation. There have been numerous nutritional, mechanistic and
physiological studies of flavonoids, with large differences in biological outcomes.
This diversity in results is probably a consequence of the many (biochemical) modi-
fications flavonoids can be subject to. Flavonoids can be methylated, acylated, sul-
phated, glycosidated or glucuronidated by humans, but flavonoids are also
extensively metabolised by the microbial intestinal flora. These modifications and/
or conjugations have impact on the physicochemical characteristics of the flavo-
noids, such as solubility, receptor-binding abilities and antioxidant or prooxidant
behaviour. It is therefore difficult to predict a priori what the effect of flavonoids in
a diet for a human individual will be, as the kinetics of modifications and conjuga-
tion reactions can be different for each person.
Upon consumption, flavonoids are only to a small extend absorbed in the small
intestines [7]. Most of the flavonoids present in the diet are converted by the micro-
bial flora in the colon. In the microbial breakdown pathway of flavonoids, C-ring
fission of aglycones is usually observed. This results in hydroxylated aromatic mol-
ecules that can be further metabolised to smaller phenylic acids and alcohols,
including valerolactones, valeric acids and phenyl-acetic acids [8]. Figure 6.1 shows
6 Automated Annotation of Microbial and Human Flavonoid-Derived Metabolites 111
B
A C
Quercetin Epicatechin
C-ring fission
3,4-dihydroxyphenyl-acetic acid 5-(3’,4’ -dihydroxyphenyl)-g-valerolactone
dehydroxylation lactone hydrolysis
3-hydroxyphenyl-acetic acid 4-hydroxy-5-(3’,4’ -dihydroxyphenyl)-valeric acid
alpha oxidation beta oxidation
3-(3-hydroxy-phenyl)-propionic acid
3-hydroxybenzoic acid
dehydroxylation
glycination
hippuric acid
Fig. 6.1 Examples of microbial conversion of flavonoids
the microbial metabolic pathways of quercetin and epicatechin. Microbial conversion

of anthocyanins occurs also via C-ring fission, resulting in many different aromatic
compounds with hydroxyl, carboxy, methoxy or acetaldehyde functional groups.
The products formed reflect the substitution pattern of the original anthocyanins [9].
Flavan-3-ols, like epicatechin in Fig. 6.1, are converted into different valerolactone,
valeric acid and aromatic ring structures. The substitution pattern of these products
reflects the substitution pattern of the original flavan-3-ols. Ellagitannins are large
polyphenolic molecules which are not absorbed by the intestines but are converted
by the microflora into a series of lactones, urolithins [9]. Ellagitannins are abun-
dantly present in different kinds of nuts (walnut, chestnut, pecan nut) and berries
(strawberry, raspberry, blueberry) but also in pomegranate, grapes and tea [9]. Due
to intra- and interindividual variation in human microbial composition of the colon,
major differences are expected to occur in the microbial breakdown pathways of
flavonoid molecules [10].
Upon absorption, a large part of the microbiotic conversion products are metabo-
lised by human biotransformation enzymes predominantly present in the gut epithe-
lial cells and in the liver. Xenobiotic metabolism is generally classified into two
phases. Phase 1 biotransformation involves a range of oxidation reactions, catalysed
primarily by cytochrome P450 enzymes as well as by flavinmonooxygenases,
resulting in hydroxylation, epoxidation, heteroatom dealkylation, heteroatom oxi-
dation, alcohol oxidation, oxidative deamination, dehydrogenations, hydrolysis,
etc. These reactions generally lead to less lipophilic molecules, which are more
susceptible to phase 2 biotransformations. The latter consist of conjugation of a
range of (solubilising) moieties mostly to hydroxyl groups, including sulfation,
glucuronidation, acetylation, glycination and methylation. Glutathione conjugation
has an important role in detoxifying reactive electrophilic species. As these detoxi-
fication processes must operate on wide ranges of xenobiotics, they are generally
catalyzed by families of enzymes, whose members have different, but broad and
partly overlapping, substrate specificities.
While it is thought that the main function of human biotransformation is to solu-
bilise foreign lipophilic compounds, in order to detoxify and excrete them from the
body, the same types of reactions occur with common dietary phytochemicals and
their microbiotic conversion products. As most polyphenols and phenolic acids dis-
cussed in previous paragraphs are only moderately lipophilic and comprising mul-
tiple hydroxylic and carboxylic moieties, they are more typical substrates for phase
2 than for phase 1 biotransformation enzymes. Indeed polyphenols and phenolic
acids undergo predominantly phase 2 conjugations with glucuronyl, sulphate, gly-
cine and methyl groups. Nevertheless, several studies suggest, mainly on the basis
of in vitro experiments, that phase 1 reactions are also involved in hydroxylation of
isoflavones and flavonoids [11, 12].
The combination of the variety of metabolic conversions of dietary compounds
by the microbiota in the intestinal tract with the large range of human biotransfor-
mations reactions results in a vast number of possible metabolites. The actual range
and relative quantities of metabolites circulating in blood, and excreted in urine, can
vary significantly in time but also between individuals due to differences in compo-
sition of the microbiota and the activities of the human biotransformation enzymes.
For example, the microbiotic conversion of the isoflavone daidzein to the isoflavan
equol was shown to differ significantly between individuals as determined from
urine excretion [13]. Equol has an inhibitory effect on topoisomerase [14] which has
been associated with preventive effects towards hormone-related cancers [15]. The
difference in equol formation was suggested to be a consequence of differences in
gut microflora induced by habitual fat versus carbohydrate consumption [16].
Differences in human biotransformation may be caused by genetic polymorphisms.

A well-studied example is the effect of glutathione S-transferase polymorphism on
the pharmacokinetics of isothiocyanates. Similarly, genetic polymorphisms, espe-
cially phase 2 biotransformation enzymes, are indicated to influence the fate of
phytochemicals in mammals [12]. While phase 2 conjugation is mainly considered
to be important for rapid excretion of xenobiotics, some glucuronyl conjugates have
been suggested to have biological activities that can potentially contribute to in vivo
effects on health [17].
6.2 Profiling of Metabolites
To understand the impact of diet and the role of the microflora, the full range of
metabolites including human biotransformation products need to be identified and
quantified [18]. Hitherto the systematic study of the intra- and interindividual varia-
tion of flavonoid metabolism has been hampered by the large diversity of products
formed and the low amounts of metabolites present in the complex matrices under
study. Continuous developments in the area of metabolomics enable comprehensive
profiling of metabolites in complex matrices like blood plasma and urine. Mass
spectrometry, coupled to gas (GC-MS) or liquid (LC-MS) chromatography, and
NMR are the major platforms for metabolite profiling and identification. Due to its
sensitivity, mass spectrometry is the principle technology for large-scale metabolite
profiling. However, with MS-based platforms, chemical structure elucidation is
often only possible on the basis of comparison to reference compounds. As a result,
in practice, large fractions of detected features in LC-MS-based metabolomics
remain “unknown.” This is also the case of flavonoid metabolic products for which
reference molecules are not available. Fragmentation can help to go one step further
and to (partially) annotate the detected molecules. Characteristic fragmentations
may be recognised by comparison to literature reports or databases. GC-MS is com-
monly based on relatively reproducible “hard” electron impact ionisation, which
allows successful searching in available databases for similar spectra of known
compounds. Fragmentation data from tandem MS or ion trap-based MSn experi-
ments, based on collision-induced dissociation, are more instrument-specific.
Many flavonoids and their metabolic products occur as different positional and
stereochemical isomers, which can usually not be differentiated with mass spec-
trometry. For example, conjugation with glycosyl groups can occur at different
positions on the flavonoid backbone, yielding different molecules with identical
mass. NMR can help in the identification of the large number of positional and ste-
reochemical isomers. However, direct 1H-NMR analysis of crude extracts can result
in spectral data that are difficult to interpret, due to overlapping NMR signals of
different metabolites as well as dominating NMR signals of highly abundant metab-
olites. For example, hippuric acid [19] which is abundantly present in urine samples
can give rise to strong aromatic proton signals which hinder identification of the
lower abundant intact phenol conjugates in urine. Therefore, full structure
elucidation by NMR often requires purification of the metabolites using, for exam-
ple, MS-based SPE trapping procedure, where LC-MS is coupled to NMR [20].
Structure elucidation of purified components is mostly achieved by a combination
of mass spectrometry and NMR, preferentially one-dimensional 1H-NMR measure-
ments combined with two-dimensional (or multidimensional) spectra. However, for
mass-limited samples, one-dimensional 1H-NMR measurements can easily be
achieved if the amount is 1 μm of compound, but two-dimensional NMR spectra are
with the current state of technology not feasible for 1 μm of compound. For homo-
nuclear 2D NMR, 10 μm of compound (or more) is requested, for hetero-nuclear 2D
NMR, 30 μm of compound (or more) is needed. In addition, extensive purification
is often needed, as background signals can obscure the molecule of interest in the
NMR spectrum. 1H-NMR spectra combined with MS data on molecular mass, MS/
MS fragments and expected molecular formula can be used for structure elucidation
and identification of metabolites isolated from complex mixtures [8, 21].
6.3 Automated Structure Elucidation
MS/MS measurements using soft ionisation techniques, such as electro-spray, pro-

vide information about the mass of parent molecular ions as well as their fragments,
which provide the possibility to interpret the data in the absence of reference spec-
tra. In the case of high-accuracy MS data, the molecular m/z value can be used to
derive the likely elemental composition of the molecule. Computational approaches
allow automatic retrieval of candidate molecules with matching monoisotopic mass,
or elemental formula, from large chemical databases and subsequent assessment of
the chemical structures in comparison of the observed fragmentation patterns. A
number of methods have been developed that annotate m/z values in a fragmenta-
tion spectrum with in silico generated substructures of possible candidate mole-
cules. Methods based on the concept include EPIC [22] (available as MassFragment
software from Waters), Mass-Frontier [23], FiD [24], MetFrag [25] and Mass-
MetaSite [26]. We recently described an extended algorithm that can handle multi-
stage spectral trees, obtained from LC-MSn experiments, where n > 2 [27]. This
algorithm results in a hierarchical tree of substructures, where substructure assign-
ments at each MS level take the assignments of the precursor as well as subsequent
fragmentations into account. The substructure annotation results in a matching
score that can be used to rank different candidate structures for an unknown metabo-
lite, to support its identification. This method has been applied to a complete
LC-MSn metabolite profile of a green tea extract, demonstrating that the majority of
compounds that had been assigned by rigorous expert analysis [8] were successfully
ranked among extensive sets of candidate structures retrieved from the largest pub-
lic chemical database, PubChem [28]. In addition, the computational annotation of
the complex LC-MSn profile resulted in the additional putative annotation of a
diverse set of components, some of which had not been reported in tea before.
Reliable annotation of metabolites is of great help in subsequent identification
efforts, which are commonly based on NMR measurements.
Identification of unknown metabolites based on NMR is facilitated by databases

of (already assigned and characterised) 1H-NMR spectra of known molecules,
which can be queried for the observed spectral features. The Human Metabolome
Database (http://www.hmdb.ca/), the Spectral Database for Organic Compounds
SDBS (http://www.aist.go.jp/RIODB/SDBS/cgi-bin/cre_index.cgi), the NMRShiftDB
(http://nmrshiftdb.ice.mpg.de/) and the Biological Magnetic Resonance Bank
(http://www.bmrb.wisc.edu/) are publicly available databases that contain NMR
spectra of metabolites. A major limitation of existing NMR databases is that their
content is mostly limited to primary metabolites. Recently we developed a new
database, systematically including 1H-NMR spectra of all known glycoside, gluc-
uronide, sulphate and methyl conjugates of the phenolic molecules, supporting
more reliable and complete identification (MetIDB, www.metidb.org). The advan-
tage of using 1H-NMR spectra for phenolic compounds is the availability of signals
in the 9.0–5.5 ppm range of the aromatic and sp2 photons. These signals are usually
well resolved and the chemical shift positions can be easily extracted and used for
querying the MetIDB database. The majority of the 1H-NMR spectra in MetIDB are
predicted spectra from 3D structure of the molecules with correct stereochemistry
and conformation of the molecule and especially of the glycosyl and the glucuronyl
fragments. These predicted spectra provide good starting values for structure verifi-
cation by iterative fit of the predicted and experimental spectrum as we will
demonstrate.
6.4 Examples
Tea contains a diverse set of flavonoids, which upon consumption are extensively
metabolised by microbiota in the colon and via human biotransformation subse-
quently are excreted in the urine. Many of these metabolites are not yet present in
databases. Therefore, in order to perform automatic annotation to LC-MS profiles
of such urine samples, we applied in silico metabolic reaction rules to generate rel-
evant candidate molecule structures, starting from the compounds present in green
tea. The reaction rules included basic hydrolysis reactions as catalysed by protease,
esterase and glycosidase digestive enzymes, a number of well-studied conversions
of flavonoids by microflora resulting in valerolactones and valeric acids, as well as
alpha and beta oxidations, para-dehydroxylation and decarboxylation (see Fig. 6.1
for examples). Large molecules with a molecular mass above 600 Da., which are
less likely to be absorbed directly from the intestinal tract [29], were removed from
the resulting library. To account for human metabolism after uptake, a number of
phase 2 biotransformations [30] were applied, resulting in a total set of almost 5,000
compounds. Figure 6.2 illustrates the in silico generation metabolites for (epi)gal-
locatechin gallate. After formation of 5-(3′,4′,5′-trihydroxyphenyl)-γ-valerolactone
systematically, all possible conjugates are generated, of which a subset of confirmed
metabolites is shown. The in silico generated metabolites were automatically
matched against LC-MSn datasets obtained with urine samples after tea consump-
tion, and fragmentation spectra were annotated by in silico substructures as
Gallocatechins
2x
2x
Trihydroxyphenyl
valerolactone
2x
2x
2x
4x
Trihydroxyphenyl
valerolactone conjugates
Fig. 6.2 Overview of the in silico metabolite generation process, including the example of gallo-
catechin. For each type of conjugate, only one positional isomer is shown. In the urine samples,
often multiple stereo- or positional isomers are detected. These different isomers can usually not
be uniquely assigned on the basis van MS fragmentation
described in the previous section. As an example, Fig. 6.3 shows the automatic
substructure annotation of the spectral data obtained with 3′,4′-dihydroxyphenyl-
valerolactone-O-glucuronide. The fragment annotation confirms that the structure is
a dihydroxyphenyl-valerolactone that is O-glucuronidated. Ninety percent of the
expert assignments that had been made prior to the analysis, for over 60 conjugated
(poly)phenolic components in urine, were reproduced by this automated in silico
113.0246
163.0764
383.0989
175.0249 122.0373
207.0664 163.0763
MS1 MS2 MS3

Fig. 6.3 MSn fragmentation of dihydroxyphenyl-valerolactone-glucuronide
workflow. Furthermore, the in silico analysis helped to make several additional

annotations. It illustrates how MAGMa facilitates annotation of the LC-MSn data at
sufficient chemical detail to provide a general biochemical interpretation. However,
the precise chemical rearrangement of the structure, e.g. which positional isomer,
cannot be determined from the MS fragmentation data. Additional structural infor-
mation from nuclear magnetic resonance spectroscopy (NMR) is then needed to
elucidate the complete structure [20].
1
H-NMR structure confirmation can be done based on predicted 1H-NMR spec-
tra. These spectra provide good starting values for querying the public databases of
known compounds or performing structure verification by iterative fitting of the
experimental spectrum. A crucial step in the 1H-NMR spectra prediction is the gen-
eration of a reliable 3D structure of the molecule. We have developed an algorithm
for 3D structure generation of flavonoids and their glycosylated secondary metabo-
lites. Briefly the procedure is as follows. The glycosyl, glucuronide, sulphate or
methoxyl fragments are detected in the molecule together with all hydroxyl groups.
Then a set of compounds is generated representing all possible combinations for
positioning these fragments in the molecule. In case of glucopyranosyl and glucuro-
nyl fragments, the structures are generated with the correct stereochemistry and
chair conformation with as many as possible hydroxyl groups at equatorial position
using the CORINA software (http://www.molecular-networks.com). Then the 3D
structure is used for 1H-NMR spectra predictions in a specified deuterated solvent
using the PERCH NMR Software (www.perchsolutions.com) following a similar
procedure as described by [31]. The structures are optimised in three-dimensional
space; a statistical set of conformers is generated using Monte-Carlo/molecular
dynamic analysis. Solvent effects are taken into account by a dielectric continuum
Table 6.1 NMR shifts (in ppm) of aromatic protons of 5-(3′,4′-dihydroxyphenyl)-γ-valerolactone

and the 3′- and 4′- O-glucuronides. The induced chemical shift of the glucuronide fragment is
given in parenthesis
ID OH O-Glc H2′ H5′ H6′
1 3′,4′ 6.75 6.76 6.69
2 3′,4′ 3′ 7.07 (0.32) 6.80 (0.04) 6.84 (0.15)
3 3′,4′ 4′ 6.81 (0.06) 7.01 (0.25) 6.72 (0.03)
model. By taking into account the stereochemistry, intramolecular interactions and

solvent effects, chemical shifts and couplings are predicted accurately. The root
mean square error (RMSE) of prediction of the chemical shifts of the sp3 and the
aromatic protons was lower than 0.14 ppm, and those of the geminal couplings
(2JH,H) and the vicinal couplings (3JH,H) were lower than 0.84 and 0.30 Hz, respec-
tively. The predicted chemical shifts can be used for querying databases of NMR
spectra of known compounds, and when experimental spectra with good signal-to-
noise ratio are available, these can be fitted using the predicted spectrum as a start-
ing point. It is not always possible to obtain clean 1H-NMR spectra that allow
iterative fitting. In such cases, it might still be possible to extract the chemical shifts
and compare these to the predicted ones. For example, the aromatic region of com-
pounds matching the exact mass of 3′,4′-dihydroxyphenyl)-γ-valerolactone gluc-
uronides contained additional signals which made the fit of the spectra difficult. The
MS/MS analysis limited possible structures to four: two positions for the glucuronyl
moiety and R and S chirality of the C4 atom (see the scheme in Table 6.1). It is not
possible to resolve the stereochemistry of the C4 atom on the basis of 1H-NMR
spectra only. Therefore, in the analysis, only the R isomer was included which
reduced the problem to resolving the position of the glucuronyl fragment. When one
of the hydroxyl groups is replaced by a glucuronyl moiety, major changes in the
1
H-NMR spectra are observed for the protons adjacent to the glucuronyl moiety
while retaining the splitting pattern. In Table 6.1 the extracted experimental chemi-
cal shifts of the aromatic protons of the aglycone (compound 1) and the two possi-
ble glucuronide conjugates are listed. The three aromatic protons have very distinct
splitting patterns: H2′ and H5′ are doublets with coupling constants of 2.1 and
8.1 Hz, respectively, whereas H6′ is a double doublet coupled to H2′ and H5′. In the
3′-O-glucuronide, the H2′ position was increased by 0.32 ppm when compared to
the aglycone. In the 4′-O-glucuronide, the adjacent H5′ proton was the most affected
a b
8.2 8 7.8 7.6 7.4 7.2 7 6.8 5.15 5.1 5.05 4.05 3.6 3.5
experimental
Urolithin-A-3-O-glucuronide
Urolithin-A-8-O-glucuronide
8.2 8 7.8 7.6 7.4 7.2 7 6.8 5.15 5.1 5.05 4.05 3.6 3.5
chemical shift, ppm
Fig. 6.4 Experimental (black) and fitted (red and green) spectra of urolithin-B-3-O-glucuronide
(upper) and a mixture of urolithin-A-3-O-glucuronide (in red) and 8-O-glucuronide (in green)
(lower). The aromatic region is shown on the left and the signals of the glucuronide moiety on the
right. The strong signal at 8.081 ppm is residual formic acid which was included in the fit in the
mixture of urolithin-A-3-O-glucuronides so that the overlapping signals could be resolved
and the chemical shift was increased by 0.25 ppm. The predicted chemical shifts for
the aromatic protons of the correctly assigned isomer were within 0.15 ppm from
the extracted experimental values. The prediction error was clearly much smaller
than the induced change in the chemical shift which allowed the assignment of the
correct structure. Such specific differences predicted for the 3′-O-glucuronide and
4′-O-glucuronide isomers resulted in more reliable identification of these structures
from their experimental NMR spectra.
Another example of efficient structure identification is the identification/distinc-
tion of the different forms of conjugated urolithins. Urolithins are known to be
microbial breakdown products of ellagic acid, a compound that has been reported to
be present in strawberries, raspberries, pomegranates, walnuts and tea [32]. We were
able to obtain good quality 1H-NMR spectra of glucuronidated conjugates (Fig. 6.4)
of urolithin-A and urolithin-B, purified from urine samples. There were two uro-
lithin-A glucuronides that co-eluted and showed similar fragmentation patterns. The
extracted chemical shifts of the aromatic protons are shown in Table 6.2 together
with the literature data of the aglycones [33, 34]. The structure of urolithin (see
scheme Table 6.2) can be considered as built of a chromene (H1–H4 protons) and a
benzyl (H7–H10 protons) fragments. The identification of urolithin-B-3-O-glucuronide
was straightforward based on the observed changes in the chemical shifts when com-
pared to the aglycone as the largest change in the chemical shifts was observed for
Table 6.2 NMR shifts (in ppm) of the aromatic protons of urolithin-B (ID 1) and urolithin-A (ID
3) and their glucuronide conjugates. The induced chemical shift of the glucuronide fragment is
given in parenthesis
ID OH O-Glc H1 H2 H4 H7 H8 H9 H10
1 3 8.15 6.85 6.75 8.25 7.57 7.88 8.18
2 3 8.204 7.158 7.113 8.31 7.598 7.891 8.252
(0.054) (0.308) (0.363) (0.060) (0.011) (0.072)
3 3,8 7.98 6.86 6.76 7.64 7.36 8.06
4 8 3 8.104 7.12 7.081 7.638 7.351 8.077
(0.124) (0.260) (0.321) (−0.002) (−0.004) (0.017)
5 3 8 8.008 6.849 6.748 7.91 7.608 8.147
(0.028) (−0.011) (−0.012) (0.270) (0.248) (0.087)
the H2 and H4 protons. In urolithin-A there is an additional hydroxyl group in the

benzene fragment. Two isomers are known in the literature [9, 32] in which the
hydroxyl group is at position 8 (urolithin-A) and position 9 (isourolithin-A). These
two isomers have the same splitting pattern of the benzene fragment. We have stud-
ied all four possible positions for the hydroxyl group, namely, hydroxyl at position
3 in the chromen fragment and positions 8 and 9 in the benzyl fragment. The best
agreement between the predicted and experimental chemical shifts was found for the
urolithin-A-3-O-glucuronide and urolithin-A-8-3-O-glucuronide. Although the two
urolithin-A glucuronides were co-eluting, the aromatic signals were well resolved in
the 1H NMR spectra, whereas the signals of the glucuronide moiety were heavily
overlapped (see Fig. 6.4). The samples obtained by SPE extraction always contain
residual amounts of formic acid which gives a singlet at 8.081 ppm. This singlet
overlapped with the two doublets (H1 and H10) of urolithin-A-3-O-glucuronide. The
final fit was performed using the predicted spectra of the two urolithin-A glucuro-
nides and formic acid. In the iterative procedure, the concentrations were optimised
together with the chemical shifts, couplings and line widths.
Intramolecular interactions can induce changes in the chemical shift positions of
the substituent protons and sometimes also second-order effects which result in
modified splitting pattern. We observed these effects in the spectra of the glucuronyl
fragment in all three urolithin conjugates. The strong overlap of the H2′ and H3′
protons of the glucuronyl fragment (difference in chemical shifts of 0.006, 0.009

and 0.015 ppm for the urolithin-A-3-O-glucuronide, urolithin-B-3-O-glucuronide
and urolithin-A-9-O-glucuronide, respectively) results in a complex splitting pat-
tern of the H1′ and the H4′ protons (see Fig. 6.4). Despite this heavy overlap, we
were able to extract the chemical shifts for the H2′ and H3′ protons for all urolithin
glucuronides which would not be possible by manual analysis of 1H-NMR data
only. These examples illustrate that predicted changes of the chemical shift of the
protons adjacent to the substituents when compared to the aglycone can guide the
assignment of the position of the substituent.
6.5 Discussion
There is increasing awareness of the importance of the microbiota in the intestinal

tract for the digestion of food and for the health of the host organism. Furthermore,
evidence accumulates that functional differences exist in the microbiotic composi-
tion between individuals, which may have consequences in terms of effects of diet
on various aspects of human health. To obtain a more detailed understanding,
metabolomics is an essential tool, with the prerequisite that the large range of the
metabolites involved in digestion still need be elucidated at the biochemical detail.
Such large-scale metabolite profiling experiments will be crucial to increase our
knowledge of the bioactivities of specific components of the human diet, of which
flavonoids are currently the most well-known example, as well as their bioactive
metabolites upon conversion by the microbiota and human biotransformation
enzymes. Insight into differences in digestion between individuals is needed at a
biochemical level to identify reliable biomarkers for the actual intake of bioactive
dietary constituents and for the personal consequences of the diet.
Advancement in analytical mass spectrometry instrumentation (UPLC coupled
to Orbitrap, TOF-based or FT mass spectrometry systems) has boosted the output
obtained in metabolomics studies, yielding up to terabytes of information for a sin-
gle scientific study. Data analysis of many mass directed metabolomics studies
relies on an untargeted analysis of the large datasets obtained with subsequent or
concomitant statistical analysis hoping to find biomarkers which are often not
clearly defined nature. It is evident that a big leap forward can be obtained in these
studies if components observed in the mass directed metabolomics studies could be
annotated in an automated fashion, enabling a more holistic approach of the
subject(s) under study. Several software approaches have been tried in the recent
past to facilitate a better annotation of mass spectrometry studies by automatically
retrieving candidate molecules with matching monoisotopic mass (which can be
converted into elemental formula) from chemical databases and subsequent
assessment of the possible structures of the observed fragments. We recently
described a new extended algorithm, MAGMA, which enables to apply such
approach to multistage MSn data [27]. Since the automated approaches rely on
matching candidate molecules, methods are needed to extend chemical databases
with new structures covering potential metabolites not yet observed or described.
One possibility is to use in silico reaction rules to predict relevant molecules. The
value of such an approach was demonstrated in a workflow for the automatic anno-
tation of metabolites of polyphenols present in urine data after consumption of tea
(in press). To support the identification of unknown metabolites with NMR, MetiDB
(www.metidb.org) was recently created, which contains 6,000 phenolic molecules
and their calculated NMR spectra. This database will be extended with phase 2
biotransformation molecules of these phenolic molecules, which is of great impor-
tance for the quick and reliable identification of the components purified from urine.
The tight integration of in silico generation of biochemically relevant molecules and
their annotation and identification in LS/MS- and NMR-based metabolite profiling
data have great potential to accelerate the profiling of novel metabolites and assess
their potential biochemical source.
6.6 Conclusions
Computational approaches are essential for the efficient annotation and identification
of the large range of metabolites of dietary constituents, produced by microbiota in
the human intestinal tract and by human biotransformation upon absorption.
Systematic identification will help to understand the role of the microbiota in diges-
tion and human health.
References
1. Harborne J. Encyclopedia of plant physiology. Berlin/Heidelberg/New York: Springer; 1980.

2. Moco S, Ross A. Can we use metabolomics to understand changes to gut microbiota. In:
Martin F-PJ, Kochhar S, editors. Metabolomics and gut microbiota in nutrition and disease.
Springer; 2014.
3. Rietjens IMCM, Sotoca AM, Vervoort J, Louisse J. Mechanisms underlying the dualistic mode
of action of major soy isoflavones in relation to cell proliferation and cancer risks. Mol Nutr
Food Res. 2013;57(1):100–13.
4. van der Woude H, Alink GM, van Rossum BEJ, Walle K, van Steeg H, Walle T, et al. Formation
of transient covalent protein and DNA adducts by quercetin in cells with and without oxidative
enzyme activity. Chem Res Toxicol. 2005;18(12):1907–16.
5. Mihaleva VV, te Beek TAH, van Zimmeren F, Moco S, Laatikainen R, Niemitz M, et al.
MetIDB: a publicly accessible database of predicted and experimental 1H NMR spectra of
flavonoids. Anal Chem. 2013;85:8700.
6. Arita M, Suwa K. Search extension transforms Wiki into a relational system: a case for flavo-
noid metabolite database. Bio Data Min. 2008;1(1):7.
7. de Boer VCJ, Dihal AA, van der Woude H, Arts ICW, Wolffram S, Alink GM, et al. Tissue
distribution of quercetin in rats and pigs. J Nutr. 2005;135(7):1718–25.
8. van der Hooft JJJ, de Vos RCH, Mihaleva V, Bino RJ, Ridder L, de Roo N, et al. Structural
elucidation and quantification of phenolic conjugates present in human urine after tea intake.
Anal Chem. 2012;84(16):7263–71.
9. Aura A-M. Microbial metabolism of dietary phenolic compounds in the colon. Phytochem
Rev. 2008;7(3):407–29.
10. van Velzen EJJ, Westerhuis JA, van Duynhoven JPM, van Dorsten FA, GruМ€n CH, Jacobs
DM, et al. Phenotyping tea consumers by nutrikinetic analysis of polyphenolic end-metabolites.
J Proteome Res. 2009;8(7):3317–30.
11. Hodek P, Trefil P, Stiborová M. Flavonoids-potent and versatile biologically active compounds
interacting with cytochromes P450. Chem Biol Interact. 2002;139(1):1–21.
12. Lampe JW. Interindividual differences in response to plant-based diets: implications for cancer
risk. Am J Clin Nutr. 2009;89(5):1553S–7.
13. Kelly GE, Nelson C, Waring MA, Joannou GE, Reeder AY. Metabolites of dietary (soya) iso-
flavones in human urine. Clin Chim Acta. 1993;223(1–2):9–22.
14. Chang Y, Nair M, Nitiss J. Metabolites of daidzein and genistein and their biological activities.
J Nat Prod. 1995;58(12):1901.
15. Blaut M, Clavel T. Metabolic diversity of the intestinal microbiota: implications for health and
disease. J Nutr. 2007;137(3):751S–5.
16. Rowland IR, Wiseman H, Sanders TAB, Adlercreutz H, Bowey EA. Interindividual variation
in metabolism of Soy isoflavones and lignans: influence of habitual diet on equol production
by the gut microflora. Nutr Cancer. 2000;36(1):27–32.
17. Zhang Y, Song TT, Cunnick JE, Murphy PA, Hendrich S. Daidzein and genistein glucuronides
in vitro are weakly estrogenic and activate human natural killer cells at nutritionally relevant
concentrations. J Nutr. 1999;129(2):399–405.
18. Collino S, Martin F-P, Moco S. Metabonomics in clinical practice. In: Kochhar S, Martin F-P,
editors. Metabonomics and gut microbiota in nutrition and disease. London: Springer; 2014.
19. Daykin CA, Van Duynhoven JPM, Groenewegen A, Dachtler M, Van Amelsvoort JMM,
Mulder TPJ. Nuclear magnetic resonance spectroscopic based studies of the metabolism of
black tea polyphenols in humans. J Agric Food Chem. 2005;53(5):1428–34.
20. Van der Hooft JJJ, Mihaleva V, Bino RJ, de Vos RCH, Vervoort J. A strategy for fast structural
elucidation of metabolites in small volume plant extracts using automated MS-guided LC-MS-
SPE-NMR. Magn Res Chem. 2011;49(S1):S55–60.
21. van der Hooft JJJ, Akermi M, Unlu FY, Mihaleva V, Roldan VG, Bino RJ, et al. Structural
annotation and elucidation of conjugated phenolic compounds in black, green, and white tea
extracts. J Agric Food Chem. 2012;60(36):8841–50.
22. Hill AW, Mortishire-Smith RJ. Automated assignment of high-resolution collisionally acti-
vated dissociation mass spectra using a systematic bond disconnection approach. Rapid
Commun Mass Spectrom. 2005;19(21):3111–8.
23. Hill DW, Kertesz TM, Fontaine D, Friedman R, Grant DF. Mass spectral metabonomics
beyond elemental formula: chemical database querying by matching experimental with com-
putational fragmentation spectra. Anal Chem. 2008;80(14):5574–82.
24. Heinonen M, Rantanen A, Mielikäinen T, Kokkonen J, Kiuru J, Ketola RA, et al. FiD: a soft-
ware for ab initio structural identification of product ions from tandem mass spectrometric
data. Rapid Commun Mass Spectrom. 2008;22(19):3043–52.
25. Wolf S, Schmidt S, Muller-Hannemann M, Neumann S. In silico fragmentation for computer
assisted identification of metabolite mass spectra. BMC Bioinforma. 2010;11(1):148.
26. Bonn B, Leandersson C, Fontaine F, Zamora I. Enhanced metabolite identification with MSE
and a semi-automated software for structural elucidation. Rapid Commun Mass Spectrom.
2010;24(21):3127–38.
27. Ridder L, van der Hooft JJJ, Verhoeven S, de Vos RCH, van Schaik R, Vervoort
J. Substructure-based annotation of high-resolution multistage MSn spectral trees. Rapid
Commun Mass Spectrom. 2012;26(20):2461–71.
28. Ridder L, van der Hooft JJJ, Verhoeven S, de Vos RCH, Bino RJ, Vervoort J. Automatic chemi-
cal structure annotation of an LC–MSn based metabolic profile from green tea. Anal Chem.
2013;85(12):6033–40.
29. Ridder L, Wang H, de Vlieg J, Wagener M. Revisiting the rule of five on the basis of pharma-
cokinetic data from rat. ChemMedChem. 2011;6(11):1967–70.
30. Ridder L, Wagener M. SyGMa: combining expert knowledge and empirical scoring in the
prediction of metabolites. ChemMedChem. 2008;3(5):821–32.
31. Lehtivarjo J, Hassinen T, Korhonen SP, Perakyla M, Laatikainen R. 4D prediction of protein
H-1 chemical shifts. J Biomol NMR. 2009;45(4):413–26.
32. Gonzalez-Barrio R, Truchado P, Ito H, Espin JC, Tomas-Barberan FA. UV and MS identifica-
tion of urolithins and nasutins, the bioavailable metabolites of ellagitannins and ellagic acid in
different mammals. J Agric Food Chem. 2011;59(4):1152–62.
33. Ito H, Iguchi A, Hatano T. Identification of urinary and intestinal bacterial metabolites of ella-
gitannin geraniin in rats. J Agric Food Chem. 2007;56(2):393–400.
34. Nealmongkol P, Tangdenpaisal K, Sitthimonchai S, Ruchirawat S, Thasana N. Cu(I)-mediated
lactone formation in subcritical water: a benign synthesis of benzopyranones and urolithins
A-C. Tetrahedron. 2013;69:9277–83.
Chapter 7
Metabonomics in Neonatal and Paediatric
Research: Studying and Modulating Gut
Functional Ecology for Optimal Growth
and Development
Vassilios Fanos and Laura Cuzzolin
Abstract Gut microbiota play a fundamental role in human health by promoting

intestinal homeostasis, stimulating development of the immune system, and provid-
ing protection against pathogens. Relatively little is known about the acquisition
and development of this complex microbial community during infancy. However,
emerging ‘omics’ technologies are now being applied to the study of the gut micro-
bial ecology, generating new opportunities to deepen the functions of the gut micro-
biota in human health. All the published literature on paediatric and neonatal
nutrimetabonomics is presented in a synthetic way, including studies on maternal
milk and formula. The role of the disruption of the gut microbiota in various gastro-
intestinal diseases is considered, focusing the metabonomics approach in gut isch-
emia, chronic inflammatory diseases, cystic fibrosis, diabetes, and obesity. As a
general rule, the best biofluid to study nutrition or to identify food-specific biomark-
ers is urine. In food consumption monitoring, the same concepts behind drug testing
and drug monitoring can be used. In the opinion of the authors, in the near future,
improved tools for the analysis of the metabolic profile (simplified like ‘dipsticks’
for urine) and its integration with the other ‘omics’ data will move metabonomics
beside the child, from top research to bedside.
Keywords Child • Development • Growth • Metabonomics • Microbiome •

Newborn • Gut • Ecology
V. Fanos (*)
Neonatal Intensive Care Unit, Puericulture Institute and Neonatal Section,
University of Cagliari-Italy, Via Ospedale 119, 09124 Cagliari, Italy
L. Cuzzolin
Department of Public Health & Community Medicine, Section Pharmacology,
University of Verona-Italy, Verona, Italy

DOI 10.1007/978-1-4471-6539-2_7
126 V. Fanos and L. Cuzzolin
7.1 Introduction
The human host-microbe symbiosis is initiated in early life and its establishment is
an important biological process.
Gut microbiota enhances the host’s metabolic capacity to process nutrients and
drugs and modulate the activities of multiple pathways in different organ systems [1].
So, gut microbiota play a fundamental role in human health by promoting intestinal
homeostasis, stimulating development of the immune system, and providing protec-
tion against pathogens [2, 3]. The disruption of the gut microbiota has been linked
to an increasing number of diseases such as necrotizing enterocolitis, inflammatory
bowel disease, obesity, cancer, diabetes, and allergies [3].
Despite this evidence, relatively little is known about the acquisition and devel-
opment of this complex microbial community during infancy [4]. However, emerg-
ing ‘omics’ technologies are now being applied to the study of the gut microbial
ecology, generating new opportunities to deepen the functions of the gut microbiota
in human health.
7.2 Microbiota Composition and Activity
The human intestinal tract harbours a complex microbial ecosystem of 100 trillion
bacteria [5], consisting of many more types of bacteria than originally thought [6],
that undergoes dynamic changes during development.
Infants are born with an essentially sterile gut, but colonization starts immedi-
ately during and after delivery: the initial inoculation of the intestinal microbiota is
a key step, because this initial phase will probably influence the composition of the
human microbiota throughout life [5, 7]. In general, the first settlers that initially
colonize infants are facultative anaerobic bacteria (mainly E. coli and Streptococci spp.),
followed by Staphylococcus spp., Enterococcus spp., and Lactobacillus spp. that
contribute to provide a favourable condition for anaerobic bacteria [8, 9]. After the
first week of life, Bifidobacterium spp., Bacteroides spp., and Clostridium spp. are
present and Bifidobacterium spp. species become predominant in human milk-fed
infants [10].
This dynamic microbial ecosystem stabilizes during the first 2–3 years [5] and
reaches the highest complexity in the human adult [11].
7.3 Factors Affecting Intestinal Microbiota Composition

and Activity
The connection between both hereditary and environmental factors plays an impor-
tant role in every stage, from conception to the early postnatal period [12] (Fig. 7.1).
The human microbiota is established after birth and starts out as a dynamic
7 Metabonomics in Neonatal and Paediatric Research… 127
Fig 7.1 Factors influencing the development and composition of microbiota in the newborn
and child
ecosystem, dominated by bifidobacteria, that stabilizes during the first 2–3 days [5, 13],
reaching a homeostatic climax composition that remains relatively stable during
most of adult’s life [14]. Influenced by a variety of early-life exposures, the infant
gut microbiota plays a crucial role in life-long health. The composition and activity
of the microbiota is affected by the genetic background of the host and by the diet.
Moreover, the mode of delivery (natural birth, caesarean section) and the gestational
age (preterm, late preterm, and term newborns) together with health status and anti-
biotic treatments could affect human microbiota composition.
7.3.1 Genetic Background
Host genotype is among the factors that influence the composition of gut microbiota
[15]. A single gene difference in the host can affect the population structure of gut
microbiota. The genotype of the host may influence its microbiota composition
either directly (secretions into the gut, control of gut motility, modification of epi-
thelial cell surfaces) or indirectly, through food and lifestyle preferences [16]. Some
authors [17] studied the degree of similarity in the predominant faecal microflora of
identical twin pairs, fraternal twin pairs, and unrelated controls: the highest levels of
similarity were found in genetically identical twins.
7.3.2 Mode of Delivery
The type of delivery strongly affects the composition of the microbiota. In the case
of caesarean section (CS) delivery, instead of faecal bacteria derived from the
mother, other environmental bacteria could affect the microbiota composition.
In particular, the microbiota of CS newborns is similar to the skin communities of

the mothers and thus dominated by Staphylococcus spp., Corynebacterium spp.,
and Propionibacterium spp. [18]. Instead, vaginally delivered infants acquire bacteria
resembling the vaginal microbiota of their mothers dominated by Lactobacillus
spp., Prevotella spp., and Sneathia spp. [18]. Moreover, CS newborns are colonized
later and less frequently by bifidobacteria [19]: this delay in the bifidobacteria colo-
nization has been shown to sustain until the first month of life, while in vaginally
delivered infants, it occurs at 10 days [20]. These differences could be due not only
to the CS itself but also to the prophylactic use of antibiotics frequently applied dur-
ing a CS [21].
As regard other bacterial groups, it has been reported that newborns born by CS
are more frequently colonized by Clostridium spp. and less by Bacteroides spp. [19, 22],
even if these data are more controversial. The observed differences can still remain
months after birth and perhaps even longer [20, 22].
Some authors [23] studied the composition of the intestinal microflora in 46 term
infants, 23 CS delivered and 23 spontaneously delivered using two PCR different
techniques: during the first 3 days of life, intestinal bacteria resulted strongly influ-
enced by mode of delivery, with marked differences among the two groups. PCR-
denaturating gradient gel electrophoresis (DGGE) analysis carried out with
Bifidobacterium-specific primers revealed the presence of this genus in 13 of 23
(56.5 %) faecal samples obtained from vaginally delivered newborns, while this
genus was absent in all samples obtained from subjects delivered by CS. Moreover,
some differences were found as regard E. coli, found in 9 of 23 (39.1 %) spontane-
ously delivered newborns and in only 2 of 23 (8.7 %) CS-delivered infants. PCR-
temperature gradient gel electrophoresis (TGGE) analysis showed greater variations
among the two groups. Finally, in all infants enrolled, Ruminococcus spp. were
absent and Bacteroides spp. were found only in 8.7 % of vaginally delivered
newborns.
Other authors [24] examined the intestinal microflora of term infants and found
a high variability in the profiles of faecal microbiota among the studied subjects,
according to previous reports [25–27]. Compared with vaginally delivered new-
borns, infants born by CS had particularly low bacterial richness and diversity, with
significantly lower abundance of Escherichia spp.-Shigella spp. and an absence of
Bacteroides spp. These results are consistent with a previous work that reported a
delayed colonization by E. coli and the phylum Bacteroidetes undetectable in
infants born by CS [19].
7.3.3 Gestational Age
A consequence of premature birth appears to be a delayed colonization of the gut

with a limited number of bacterial species [28, 29]. These differences in timing and
diversity are mainly due to the aseptic neonatal intensive care environment and the
extensive use of antibiotics shortly after birth [30].
The observed bacterial diversity has been associated with recurrent C. difficile
infections and other disease states [31]. In fact, compared to term infants, the colo-
nization dynamics is different in preterm newborns, often colonized by potentially
pathogenic species such as Klebsiella spp., Enterobacter spp., and Clostridium spp.,
instead of normal commensal microbiota such as Bifidobacterium spp. and
Lactobacillus spp. [32, 33].
Some authors [30] showed a positive correlation between diversity of intestinal
microbiota and digestive tolerance and weight gain. Moreover, the delayed coloni-
zation and reduced diversity observed in preterm newborns render these subjects
more susceptible to bacterial disturbances and therefore at higher risk of necrotizing
enterocolitis and sepsis [28, 34]. Recently, stools of 32 preterm infants exposed to
current NICU practices were analyzed for assessment of the total and viable bacte-
rial communities in the gut. Among the studied population, 7 and 13 preterm infants
developed, respectively, NEC and sepsis. Total bacterial profiles of infants with
NEC and total and viable profiles of infants with sepsis significantly differed from
those of healthy infants, supporting a role for bacterial colonization in the patho-
physiology of these diseases. Importantly, Sphingomonas spp. colonization was sig-
nificantly associated with NEC [34].
7.3.4 Infant Diet
The composition of the intestinal microbiota can be modulated as a result of dietary

exposure (human milk, formulas) as well as of intentional diet supplementations
(prebiotics or probiotics).
Human milk is normally the first dietary exposure in infancy and is considered
the best nutrition for growth and healthy development of the newborn, containing a
wide range of health-promoting constituents [5, 35]. In particular, oligosaccharides
present in human milk have prebiotic effects, fermenting in the colon and stimulat-
ing the growth of specific bacteria [36]. The metabolic profile of infants receiving
human milk is characterized by a relatively high presence of acetate, a lower content
of propionate, and the absence of butyrate [37]. Human milk has also been shown to
a be a source of live bacteria, including staphylococci, streptococci, and bifidobac-
teria [38]. In particular, breast-fed infants contain a high abundance of Bifidobacterium
breve [39]. The origin of bacteria present in human milk remains controversial, even
if it is generally accepted that the newborn acquires the mother’s microbiota during
delivery and then transfers these bacteria to the breast skin [5] or to the mammary
gland through an endogenous route [40].
Bovine milk, the most common base for infant formulas, contains insignifi-
cant levels of prebiotic oligosaccharides and this partly explains the differences
observed between human milk-fed and formula-fed babies. Some authors [24]
detected significant effects of diet on the several bacterial taxa, characterized by
lower bacterial richness and diversity in infants who were breastfed. In detail,
infants receiving formula had bacterial communities with significantly higher
abundance of the families Peptostreptococcaceae and Verrucomicrobiaceae and a

significantly higher prevalence of C. difficile, a pathogen associated with enteric
and atopic diseases [22].
The supplementation of infant formulas with prebiotics has been recently
reviewed [41] and a mixture of short chain galacto-oligosaccharides and long chain
fructo-oligosaccharides has been shown to selectively stimulate the growth of bifi-
dobacteria and lactobacilli, generating metabolic profiles similar to those observed
in human milk-fed infants [37, 42]. An alternative approach regards the oral admin-
istration of viable bacteria (probiotics) [43]: a few probiotics supplemented to
formula-fed infants may be associated with some clinical benefits, such as a reduc-
tion in the risk of non-specific gastrointestinal infections [44] and necrotizing
enterocolitis [45]. An updated meta-analysis of all relevant controlled trials per-
formed to assess the benefits of probiotic supplementation for preterm newborns
underlined a significant decreased risk of necrotizing enterocolitis and mortality,
while no difference in the risk of sepsis was observed [45]. The effects of live bac-
teria and combinations of prebiotics and probiotics are gaining interest, but need
further exploration particularly as regard long-term effects.
7.3.5 Antibiotic Treatments
The early exposure to antibiotics has significant immediate effects and probably also
sustainable effects on the gut microbiota composition. Variables associated to antibiotic
use (dose, length of treatment, route of administration) make difficult to draw strong
conclusions on the exact impact on the microbiota. However, an antibiotic treatment
surely causes disturbances in the early colonization by Bifidobacterium species, predis-
posing to an overgrowth of Enterococcus and Enterobacteriaceae species [46–48].
Recently, some authors underlined that antibiotic administration not only alters
the total counts of Bifidobacterium but can also has an impact at the species level,
with a reduction of Bifidobacterium bifidum and Bifidobacterium adolescentis [47,
48]. Despite the conviction that an early antibiotic treatment does not seem to have
major long-term effects on the faecal intestinal microbiota, a complete recovery of
initial bacterial community composition is rarely achieved [49]. Therefore, it has
been hypothesized that exposure to antibiotic in early life may trigger the subse-
quent development of immune disorders, such as asthma, wheezing, and other aller-
gic manifestations [50, 51].
7.4 Nutrimetabonomics: A Revolutionary Tool

in Neonatology and Paediatrics
The metabonome is the sum of all endogenous and exogenous metabolites. The
intrinsic factors are body composition, tissue turnover, resting metabolic rate, age,
genotype, health status, reproductive status, and diurnal cycle, and the extrinsic
factors are diet (nutrients and nonnutrients), drugs, physical activity, colonic flora,
pollution, lifestyle, and stress [52]. All these factors contribute to the metabolic
genotype-phenotype relationship [53].
Metabotype is the whole set of metabolites which can be detected in body fluids
and which characterizes the metabolic phenotype of an individual.
Gone are the days when one is constrained to study only a single gene or gene
product in a biologically out-of-context situation’.
The term nutrimetabonomics illustrates the mutual link among nutrition and
metabonomics. More practically, it describes how a biological system varies following
a nutritional stimulus. The research in this field studies the effects of specific ingre-
dients and food components elucidating the effects of specific ingredients behind
individual responses. Thus, the emerging faces of nutrition is to achieve the ambitious
goal of optimizing an individual’s health via nutritional intervention [54].
It has been underlined that metabonomics is the logical approach to assess
dysfunction and metabolic imbalances caused by dietary components [55]. This
new approach will surely influence public health practice in the future [56].
Many years ago, Thomas Edison predicted that doctors of the future would no
longer treat the human frame with drugs, but rather would cure and prevent diseases
with nutrition.
Now, deciphering the complex interactions between nutrients and the human
organism constitutes a considerable challenge for the twenty first century [57, 58].
We can add the microbiome to the previous sentence.
However, despite this background, while the number of systems biology publica-
tions has risen rapidly in the last decade, the percentage of these related to nutri-
tional sciences research has remained constant at 3–4 % of the total [59].
Very recently, a review described recent applications of metabonomics in pre-
clinical and clinical fields anticipating novel therapeutic and nutrition advances in
paediatric research [60].
The diet has a key role in the gut microbiota modulation and shaping and in a
metabolic signalling network construction. Metabolic profiling has a wide potential
for the following goals: (a) understanding the complex interactions between compo-
nents of the gut microbiota, (b) elucidating the cause/effect relationships associated
with specific nutritional choices, and (c) evaluating the related shifts in the micro-
biota composition. The symbiosis between mammalians and the microbial system
can play a role in the aetiology and development of several diseases, e.g. insulin
resistance, Crohn’s disease, irritable bowel syndrome (IBS), food allergies, gastritis
and peptic ulcers, obesity, cardiovascular disease, and gastrointestinal tract. In
particular gut, microbiome, and nutrients strongly interact with the host genetic
elements to determine the metabotype. The knowledge of these complex interac-
tions can provide personalized plans of treatment and prevention [61].
This is true also for long-term outcomes. In fact, since ‘physiological’ ageing is
associated with a number of significant changes in gastrointestinal function, the
development and progression of chronic diseases could be prevented, minimized, or
better managed monitoring patient response on an individual basis [62].
In healthy humans, urine represented a sensitive metabolic profile that reflected
acute dietary intake, whereas plasma and saliva did not [63].
As a general rule, the best biofluid to study nutrition or to identify food-specific

biomarkers is urine. Urine is essentially the body’s liquid repository and any nutri-
ent or nonnutrient that is not needed or present in excess in the body will find its way
in urine food-specific biomarkers that are present in the blood or urine for 5–10 h,
with some persisting as long as 48 h. Thus, in food consumption monitoring, the
same concepts behind drug testing and drug monitoring can be used [64].
This assumption has been recently confirmed: although both plasma and urine
can be analyzed for the presence of food-derived biomarkers, with the present state
of development of nutritional metabonomics, urine appears to be the preferred bio-
fluid because (1) a greater diversity of metabolites derived from the food metabo-
nome is observed in the urine, (2) collecting urine samples is relatively easy and
noninvasive, and (3) spot urine samples collected at particular times in relation to
meals and sleep period (‘behavioural phase’ urines) can be informative, and collec-
tion of 24 h urine samples may not be required [65, 66].
For example, it was possible to identify urinary metabolite profiles that discrimi-
nate between high and low intake of dietary protein during a dietary intervention
[67]. The proposed dietary biomarkers have been recently reviewed [68].
In paediatrics, a special problem is represented that age influences metabonom-
ics results. Age-related metabolic changes in children aged 12 years and below
were investigated by Gu et al. [69] using 1H NMR-based metabonomic analysis of
urine. Unsupervised PCA analysis showed a distinct age-dependent clustering,
indicating the effect of age on the urinary metabolite profile. Further statistical
analysis led to the identification of age-related metabolic profiles. Among the
metabolites that were found to correlate with age, creatinine increased with age,
while creatine, glycine, betaine/TMAO, citrate, succinate, and acetone decreased.
This investigation has shown that metabonomic approach has the potential to be
useful in assessing the biological age of young humans as well as in providing
more information about the confounding factors in the clinical application of meta-
bonomics [69].
In neonatology, metabonomics has been extensively studied. The latest papers
are related to intrauterine growth-restricted and small for gestational age neonates,
prematurity, mode of delivery, hypoxic-ischemic encephalopathy, persistent ductus
arteriosus, respiratory syndrome and surfactant therapy, cytomegalovirus infection,
nephrouropathy, inborn errors of metabolism, pharmacometabonomics, and
nutrimetabonomics (including study of maternal milk and formula). Also numerous
papers have been presented in experimental neonatology. In particular, the fluids
most frequently used were as follows: urine (by far the most used fluid), cord blood
plasma, but also milk and stools [70–80].
Finally, we want to stress that intestinal microbiota highly influences the colonic
luminal metabonome, and a comprehensive understanding of intestinal luminal
metabonome is critical for clarifying host-intestinal bacterial interactions. Thus,
low-molecular-weight metabolites produced by intestinal microbiota play a direct
role in health and disease. Matsumoto et al. analyzed in germ-free (GF) mice and
Ex-GF mice the colonic luminal metabonome using capillary electrophoresis mass
spectrometry with time of flight (CE-TOFMS). CE-TOFMS identified 179 metabolites
from the colonic luminal metabonome and 48 metabolites were present in significantly
higher concentrations and/or incidence in the germ-free (GF) mice than in the
Ex-GF mice (p < 0.05), 77 metabolites were present in significantly lower concen-
trations and/or incidence in the GF mice than in the Ex-GF mice (p < 0.05), and 56
metabolites showed no differences in the concentration or incidence between GF
and Ex-GF mice [81].
7.4.1 Maternal Milk, Formula, Metabonome, and Microbiome
The studies published in this field are very few.

In an earlier study, the choline content of human breast milk in the first 3 weeks
after birth was compared with bovine milk and infant formula by use of 1H NMR
spectroscopy [82]. The observed choline species included free choline, phospho-
choline, glycerophosphocholine, phosphatidylcholine, and sphingomyelin. Holmes
et al. identified that total choline content in human colostrum at birth is lower than
in mature milk 7 days post-partum, which correlates well with the acceleration in
growth that the neonate experiences at this time point [82]. Moreover, it was specu-
lated that for preterm infants, the choline content available in human milk is not
enough for their rapid growth, as their metabolic activity is higher compared with
full-term infants [82].
Some preliminary results of the authors of this chapter provide information on
the biochemical variability of preterm HBM and on the potentiality of the metabo-
nomic approach in nutrition and health [83].
The metabolic profile of preterm human breast milk (HBM) was investigated by
using a metabonomics approach. To this aim, NMR spectroscopy and GC/MS, in
combination with multivariate statistical analysis, were used to analyze the water-
soluble and lipid fractions extracted from human milk samples, respectively, com-
pared with preterm formula milk (FM), commonly prescribed. HBM contains
relatively higher contents of this sugar with respect to FM samples. By contrary, the
commercial products were suggested to be richer in maltose. Furthermore, other
bins, belonging to unidentified metabolites, not present in HBM (‘aliens’ metabo-
lites) were found to be important for the sample clustering. A deep examination of
the score plot showed that milk samples of term and late preterm infants were
located at the opposite side with respect those expressed by mothers at the lowest
GA under investigation (i.e. 26 weeks).
Considering the lipidic part, the level of oleic and linoleic acids appeared to be
higher in the artificial formulas than in HBM [83]. Finally, we observed a progres-
sive change of the metabolic profile of milk from the right to the left part of the plot
over the first month of lactation, suggesting a temporal variation in the carbohydrate
composition. In particular, an increase of the lactose level was observed during the
lactation period, in good agreement with the literature data [84].
In another experience, complimentary animal and human studies were con-
ducted on young piglets and premature infants (34–36 weeks). Breast milk-fed vs.
formula-fed groups were analyzed by GC/MS. Metabonomics clearly was able to

identify differences between breast milk-fed and formula-fed groups in the gut
environment of piglets and humans. Among the most important discriminating
metabolites between breast milk-fed and formula-fed groups, the authors found
sugars, amino-sugars, fatty acids (namely, unsaturated fatty acids), and sterols.
Thus, metabonomics and microbiota pinpointed specific sets of metabolites associ-
ated with the dominant bacterial taxa [85].
In conclusion, according with a very recent review on this topic, milk metabo-
nomics represents a very promising field of study and with a potential to impact
primary producers, industry, and consumers. Evidence has been obtained that the
milk metabolites detected by NMR-based metabonomics are of importance in rela-
tion to milk nutritional quality, technological properties, quality control, and bioac-
tivity [86].
7.5 The Complex Partnership of Disease and Nutrition:

The Role of Metabonomics
7.5.1 Metabonomics, Gut Ischemia, and Necrotizing

Enterocolitis
Intestinal ischemia/reperfusion (I/R) injury initiates a systemic inflammatory

response syndrome with a high associated mortality rate. Early diagnosis is essen-
tial for reducing surgical mortality, yet current clinical biomarkers are insufficient is
a novel strategy for studying intestinal I/R, which metabonomics has the potential
for personalized risk stratification in patients exposed to intestinal I/R and may be
used as part of a systems approach for quantitatively analysing the intestinal micro-
biome during gut injury [87].
In a mouse model for intestinal ischemia, sera were analyzed 4 h after mesenteric
artery ligation by gas chromatography–mass spectrometry for 40 small molecules
as their trimethylsilyl and O-methyloxime derivatives. The following molecular sig-
natures were found: three highly significantly upregulated (fold-change) serum
molecules in intestinal ischemia were inorganic phosphate (2.4), probably due to
exit from the gut filter; urea (4.3), likely related to a reduction in glomerular filtra-
tion rate; and threonic acid (2.9). Threonic acid, which is a vit. C metabolite and is
related to oxidative stress, seems to be the most specific metabolite. Five highly
significantly downregulated (fold-change) serum molecules were stearic acid (1.7),
arabinose (2.7), xylose (1.6), glucose (1.4), and ribose (2.2). Lactic acid, differently
from other reports, remained unchanged in intestinal ischemia. Stearic acid seems
to be of nutritional origin and is a precursor of oleic acid in the liver, very reduced
in intestinal ischemia. The four monosaccharides (arabinose, xylose, glucose, and
ribose) are produced by the microbiome and are the prove of its involvement in
intestinal ischemia.
Taken all together, these data reveal alterations of gut microbiota metabolism,
intestinal absorption, and renal function, together with increased oxidative stress
[88].
Matching metabonomics and necrotizing enterocolitis (NEC), no results can be
found. Although diet composition has been implicated as a major factor in the aeti-
ology of various gastrointestinal diseases, such as NEC, conclusive evidence
remains elusive. What is known on this topic is that breast milk, as opposed to
commercial formula, appears to confer a ‘protective effect’ to the ‘immature gut’.
Yet the mechanism by which this occurs continues to remain speculative.
7.5.2 Metabonomics in Chronic Inflammatory Diseases
Gut microbial activities can be extremely important in the aetiology and develop-
ment of several chronic inflammatory disorders, including inflammatory bowel disease
(IBD) or irritable bowel syndrome (IBS). This topic will be covered in Chap. 14.
Here we will underline only few concepts related to this section.
Urinary metabolite profiling was carried out on a mouse model of Crohn’s dis-
ease suggesting alterations of tryptophan metabolism, fucosylation and fatty acid
metabolism in Crohn’s disease fucose, and xanthurenic acid could be useful markers
of gut inflammation [89].
A mouse model of inflammatory bowel disease (IBD) was used to investigate
urinary metabolites using NMR; it was found that there was an increase in trimeth-
ylamine (TMA) and fucose compared to controls. The increase in TMA was parallel
to the progression of IBD [90].
Metabonomic analysis of faecal extracts of patients with both inflammatory dis-
eases showed reduced levels of butyrate, acetate, methylamine, and TMA compared
to control [91].
Urinary metabolites have also been used to distinguish CD and UC in humans.
Hippurate was lowest in CD and differed significantly between CD, UC, and con-
trols. Hippurate has been shown to be modulated according to gut microbes and this
difference is likely to reflect changes in intestinal microbes [92].
By a practical point of view, urine and stools can be used as useful monitoring
tools [93]. Attempt was made to study IBS with a combination of microbial and
metabonomic outcomes using stool analysis. Using a GC–MS methodology, the
authors highlighted higher levels of specific amino acids (alanine and pyrogluta-
mate) and phenolic compounds (hydroxyphenylacetate and hydroxyphenylpropio-
nate) with IBS, alterations possibly associated with specific gut microbial
populations, including the abundance of lactobacilli and Clostridium [58, 94].
Celiac disease (CeD) is a unique autoimmune multifactorial gastrointestinal
disorder in which the genetic factors (DQ2/DQ8) and the environmental trigger
(gluten) are known and necessary but not sufficient for its development. Bertini
et al. [95] highlighted changes in gut microbial cometabolites may be associated to
aberrant microbiota in the small bowel of patients with CeD [96].
Sellitto et al. [97] characterized the longitudinal changes in the microbial

communities that colonize infants from birth to 24 months and the impact of two
patterns of gluten introduction (early vs. late) on the gut microbiota and metabo-
nome, and the switch from gluten tolerance to immune response, including onset of
CeD autoimmunity. CD is probably associated with intestinal and faecal dysbiosis,
which is related to certain bacterial species. As shown by, the gluten-free diet lasting
at least 2 years did not completely restore the microbiota and, consequently, the
metabonome of CD children.
Some molecules (e.g. ethyl-acetate and octyl-acetate, some short chain fatty
acids and free amino acids, and glutamine) together with microbial indices
(e.g. ratio between faecal cell density of lactic acid bacteria-Bifidobacterium vs.
Bacteroides-Enterobacteria) seem to be metabolic signatures of CD patients [98].
These data confirm results published in previous papers [95].
7.5.3 Cystic Fibrosis
In cystic fibrosis (CF), airway inflammation leads to an increased production of

reactive oxygen species, resulting in the degradation of cell membranes and the
generation of volatile organic compounds (VOCs). The study by Wolak et al. [99]
demonstrated that metabonomic analysis of bronchoalveolar fluid can differentiate
between different degrees of inflammation in children affected by cystic fibrosis and
has the potential to identify new biomarkers of inflammation.
A more recent study was carried out with the aim of investigating whether mass
spectrometry-based metabonomic analysis of volatile organic compounds (VOCs)
in exhaled breath was able to discriminate between CF subjects and controls and
between CF subjects with and without Pseudomonas colonization. Samples from 48
children with CF and 57 controls were examined. Analysis revealed that 1099 VOCs
exhibited a prevalence of at least 7 %. A 100 % correct identification of CF subjects
and controls was possible by using 22 VOCs. Therefore, metabonomic analysis of
VOCs in exhaled breath appears to be a reproducible technique and is able to dis-
criminate not only between CF subjects and controls but also between CF subjects
with or without Pseudomonas colonization [100].
7.5.4 Diabetes
This part will be covered by Chap. 12. Different studies suggest that branched-chain
amino acids (BCAAs) related to metabolic signature is a robust metabolic readout
of insulin resistance (IR) [101]: high levels of BCAAs were detected in the obese
and IR phenotype [102–104].
In particular, five branched-chain and aromatic amino acids were indeed associ-
ated with IR, namely, isoleucine, leucine, valine, tyrosine, and phenylalanine, and a
combination of three amino acids (isoleucine, phenylalanine, tyrosine) could

predict future diabetes (>fivefold higher risk for individuals in top quartile) [105].
During the last decade, the rapidly growing research field of metabonomics has
introduced new insights into the pathology of diabetes as well as methods to predict
disease onset and has revealed new biomarkers [106].
In the adult, Wang–Sattler identified three metabolites (glycine, lysophosphati-
dylcholine (LPC) (18:2), and acetylcarnitine) that had significantly altered levels.
The most important results of this study are that metabonomics is enough and it is
not necessary to perform other ‘omics’ and that those who have that three metabo-
lites altered will develop surely diabetes [107].
A recent review summarizes the current findings of metabolic research regarding
diabetes in animal models and human investigations [106].
Finally, very recently it has been stressed that systems biology methodologies
can identify disease biomarkers and uncover potential therapeutic targets from a
combination of ‘omics’ datasets. Relevant examples are diabetes and obesity [107].
7.5.5 Obesity
Metabonomics has been widely studied in adult obese [108].

For example, a recent study on obese Japanese subjects reported a physiological
inference between insulin resistance (IR), plasma levels of alanine, glycine, glutamate,
tryptophan, tyrosine and BCAAs, and visceral fat metabolism [109]. Again, a complex
relationship between dyslipidemia and IR development has been described [110].
Differently from the adults, only few studies have been dedicated to childhood
obesity. In a study by Walsh et al., serum metabolite concentration profiles of obese
children could be distinguished from those of normal-weight children. The identified
metabolite markers are indicative of oxidative stress and of changes in sphingomy-
elin metabolism, in β-oxidation, and in pathways associated with energy expendi-
ture. The altered metabolites might be considered as potential biomarkers in the
generation of new hypotheses on the biological mechanisms behind obesity [111].
Mihalic et al. compared acylcarnitine (AcylCN) species, common amino acid and fat
oxidation (FOX) by-products, and plasma amino acids in normal weight (NW;
n = 39), obese (OB; n = 64), and type 2 diabetic (n = 17) adolescents. The observations
of the authors are consistent with early adaptive metabolic plasticity in youth, which
over time—with continued obesity and ageing—may become dysfunctional, as
observed in adults [112]. An experimental paper has been published on piglets by He
et al. [113]. All the papers on paediatric obesity are presented in Table 7.1 [114].
What is becoming increasingly important is the role of perinatal programming in
the development of paediatric and adult obesity. This could involve either newborns
with ‘not enough’ or newborns with ‘too much’ birth weight. The common sign of
hypoglycemia at birth and common pathways such us some metabolites like myo-
inositol can be responsible for later appearance of metabolic syndrome and obesity
later in life [79, 115–118].
138
Table 7.1 Metabonomics studies that evaluated the metabolic condition in infant obesity
Type of Metabonomic
Author Year patient Sample analysis Metabolites results
He et al. 2012 Newborn Serum NMR-based HDL, VLDL, lipids, unsaturated lipids, glycoprotein, myo-inositol, pyruvate,
piglet metabonomic threonine, tyrosine and creatine > in obese than in lean pigs (p < 0.05). Serum
technology glucose and urea < in obese pigs (p < 0.05)
Wahl et al. 2012 Human Serum Mass spectrometry- 14 metabolites (glutamine, methionine, proline, nine phospholipids, and two
children based metabonomics acylcarnitines, p < 3.8 × 10−4) and 69 metabolite ratios (p < 6.0 × 10−6) to be
approach significantly altered in obese children
Mihalik 2012 Human Plasma Tandem mass Fasting lipolysis and fat oxidation were higher in obese and type 2 diabetes
et al. adolescents spectrometry compared with normal weight. Insulin sensitivity was lower in obese and
type 2 diabetes
From Ref. [114] with permission
V. Fanos and L. Cuzzolin
7.6 Conclusions
The human host-microbe symbiosis is initiated in early life and it is an important

biological process.
Gut microbiota enhances the host’s metabolic capacity to process nutrients and
drugs and modulate the activities of multiple pathways in different organ systems.
The disruption of the gut microbiota has been linked to an increasing number of
diseases such as necrotizing enterocolitis, inflammatory bowel disease, obesity,
cancer, diabetes, and allergies.
Now, deciphering the complex interactions between nutrients and the human
organism constitutes a considerable challenge for the twenty fist century. The role
of microbiome in this setting is of increasing importance. Metabolic profiling has a
wide potential in deciphering for the following goals: (a) understanding the com-
plex interactions between components of the gut microbiota, (b) elucidating the
cause/effect relationships associated with specific nutritional choices, and (c) evalu-
ating the related shifts in the microbiota composition.
As a general rule, the best biofluid to study nutrition or to identify food-specific
biomarkers is urine. In food consumption monitoring, the same concepts behind
drug testing and drug monitoring can be used.
The way to clinical implementation of metabonomics is still hampered by many
of the problems that had to be solved for genomics and proteomics in the past, as
well as new ones that require the creation of new analytic, computational, and inter-
pretative techniques.
Some problems are similar with those observed for pharmacometabonomics:
significant individual variability, issues surrounding methods for metabolite detec-
tion (NMR, MS); extremely complex datasets; possible over-interpretation of data;
necessity for skilled and experienced technicians and well-trained practitioners,
time-consuming processing and analysis of patient samples, resulting in delayed
treatment; and high cost of the processing and analytical platforms [119, 120].
In the opinion of the authors, in the near future, improved tools for the analysis
of the metabolic profile (simplified like ‘dipsticks’ for urine) and its integration with
the other ‘omics’ data will move metabonomics beside the child [121].
The greatest challenge, however, will be the integration of information from
different ‘omics’, for example, in the form of new, superior ‘meta-markers’ [122].
References
1. Claus SP, Ellero SL, Berger B, et al. Colonization-induced host-gut microbial metabolic
interaction. MBio. 2011;2(2):e00271–10. doi:10.1128/mBio.00271-10.
2. Maynard CL, Elson CO, Hatton RD, Weaver CT. Reciprocal interactions of the intestinal
microbiota and immune system. Nature. 2012;489:231–41. doi:10.1038/nature11551.
3. Young VB. The intestinal microbiota in health and disease. Curr Opin Gastroenterol.
2012;28(1):63–9. doi:10.1097/MOG.0b013e32834d61e9.
4. Fouhy F, Ross RP, Fitzgerald G, Stanton C, Cotter PD. Composition of the early intestinal
microbiota: knowledge, gaps and the use of high-throughput sequencing to address these
gaps. Gut Microbes. 2012;3:203–20. doi:10.4161/gmic.20169.
5. Scholtens PA, Oozeer R, Martin R, Amor KB, Knol J. The early settlers: intestinal
microbiology in early life. Annu Rev Food Sci Technol. 2012;3:425–47. doi:10.1146/
annurev-food-022811-101120.
6. Zoetendal EG, Rajilic-Stojanovic M, de Vos WM. High-throughput diversity and functionality
analysis of the gastrointestinal tract microbiota. Gut. 2008;57:1605–15. doi:10.1136/
gut.2007.133603.
7. Faa G, Gerosa C, Fanni D, Nemolato S, Fanos V. Factors influencing the development of a
personal tailored microbiota in the neonate, with particular emphasis on antibiotic therapy.
J Matern Fetal Neonatal Med. 2013;26:35–43.
8. Fanaro S, Chierici R, Guerrini P, Vigi V. Intestinal microflora in early infancy: composition
and development. Acta Paediatr. 2003;91(441):48–55.
9. Orrhage K, Nord CE. Factors controlling the bacterial colonization of the intestine in breast
fed infants. Acta Paediatr. 1999;88(430):47–57.
10. Yoshioka H, Iseki K, Fujita K. Development and differences of intestinal flora in the neonatal
period in breast-fed and bottle-fed infants. Pediatrics. 1983;72:317–21.
11. Rajilic-Stojanovic M, Heilig HG, Molenaar D, et al. Development and application of the
human intestinal tract chip, a phylogenetic microarray: analysis of universally conserved
phylotypes in the abundant microbiota of young and elderly adults. Environ Microbiol.
2009;11:1736–51. doi:10.1111/j.1462-2920.2009.01900.x.
12. Sanz Y. Gut microbiota and probiotics in maternal and infant health. Am J Clin Nutr.
2011;94:2000S–5. doi:10.3945/ajcn.110.001172.
13. Koenig JE, Spor A, Scalfone N, et al. Succession of microbial consortia in the developing
infant gut microbiome. Proc Natl Acad Sci U S A. 2011;108 suppl 1:4578–85. doi:10.1073/
pnas.1000081107.
14. Ottman N, Smidt H, de Vos WM, Belzer C. The function of our microbiota: who is out there
and what do they do? Front Cell Infect Microbiol. 2012;2:104. doi:10.3389/fci.
15. Zhao L, Wang G, Siegel P, et al. Quantitative genetic background of the host influences gut
microbiomes in chickens. Sci Rep. 2013;3:1163. doi:10.1038/srep01163mb.2012.00104.
16. Khachatryan ZA, Ktsoyan ZA, Manukyan GP, Kelly D, Ghazaryan KA, Aminov RI. pre-
dominant role of host genetics in controlling the composition of gut microbiota. PLoS One.
2008;3(8):e3064. doi:10.1371/journal.pone.0003064.
17. Stewart JA, Chadwick VS, Murray A. Investigations into the influence of host genetics on
the predominant eubacteria in the faecal microflora of children. J Med Microbiol.
2005;54:1239–42. doi:10.1099/jmm. 0.46189-0.
18. Dominguez-Bello MG, Costello EK, Contreras M, et al. Delivery mode shapes the acquisi-
tion and structure of the initial microbiota across multiple body habitats in newborns. Proc
Natl Acad Sci U S A. 2010;107(26):11971–5. doi:10.1073/pnas.1002601107.
19. Adlerberth I, Strachan DP, Matricardi PM, et al. Gut microbiota and development of atopic
eczema in European birth cohorts. J Allergy Clin Immunol. 2007;120:343–50. doi:10.1016/j.
jaci.2007.05.018.
20. Gronlund MM, Lehtonen OP, Eerola E, Kero P. Fecal microflora in healthy infants born by
different methods of delivery: permanent changes in intestinal flora after caesarean delivery.
J Pediatr Gastroenterol Nutr. 1999;28:19–25.
21. Chen J, Cai W, Feng Y. Development of intestinal bifidobacteria and lactobacilli in breast-fed
neonates. Clin Nutr. 2007;26:559–66. doi:10.1016/j.clnu.2007.03.003.
22. Penders J, Thijs C, Vink C, et al. Factors influencing the composition of the intestinal micro-
biota in early infancy. Pediatrics. 2006;118:511–21. doi:10.1542/peds.2005-2824.
23. Biasucci G, Rubini M, Riboni S, Morelli L, Bessi E, Retetangos C. Mode of delivery affects
the bacterial community in the newborn gut. Early Hum Dev. 2010;86:S13–5. doi:10.1016/j.
earlhumdev.2010.01.004.
24. Azad MB, Konya T, Maughan H, et al. Gut microbiota of healthy Canadian infants: profiles
by mode of delivery and infant diet at 4 months. CMAJ. 2013. doi:10.1503/cmaj.121189
(Epub ahead of print).
25. Hong PY, Lee BW, Aw M, et al. Comparative analysis of fecal microbiota in infants with and
without eczema. PLoS One. 2010;5:e9964. doi:10.1371/journal.pone.0009964.
26. Roger LC, Costabile A, Holland DT, Hoyles L, McCartney AL. Examination of faecal
Bifidobacterium populations in breast- and formula-fed infants during the first 18 months of
life. Microbiology. 2010;156:3329–41. doi:10.1099/mic.0.043224-0.
27. Abrahamsson TR, Jakobsson HF, Andersson AF, Bjőrkstén B, Engstrand L, Jenmalm
MC. Low diversity of the gut microbiota in infants with atopic eczema. J Allergy Clin
Immunol. 2012;129:434–40. doi:10.1016/j.jaci.2011.10.025.
28. de la Cochetiere M, Piloquet H, Des Robert C, Darmaun D, Galmiche J, Roze J. Early
intestinal bacterial colonization and necrotizing enterocolitis in premature infants:
the putative role of Clostridium. Paediatr Res. 2004;56:366–70. doi:10.1203/01.
PDR.0000134251.45878.D5.
29. Rougè C, Goldenberg O, Ferraris L, et al. Investigation of the intestinal microbiota in pre-
term infants using different methods. Anaerobe. 2010;16(4):362–70. doi:10.1016/ j.
anaerobe.2010.06.002.
30. Jacquot A, Neveu D, Aujoulat F, et al. Dynamics and clinical evolution of bacterial gut micro-
flora in extremely premature patients. J Pediatr. 2011;158:390–6. doi:10.1016/j.jpeds.2010.09.
31. Chang JY, Antonopoulos DA, Kaira A, et al. Decreased diversity of the fecal microbiome in
recurrent Clostridium difficile-associated diarrhea. J Infect Dis. 2008;197(3):435–8.
doi:10.1086/525047.
32. Butel MJ, Suau A, Campeotto F, et al. Conditions of bifidobacterial colonization in preterm
infants: a prospective analysis. J Pediatr Gastroenterol Nutr. 2007;44:577–82. doi:10.1097/
MPG.Ob013e3180406b20.
33. Magne F, Abely M, Boyer F, Morville P, Pochart P, Suau A. Low species diversity and high
interindividual variability in faeces of preterm infants as revealed by sequences of 16S rRNA
genes and PCR-temporal temperature gradient gel electrophoresis profiles. FEMS Microbiol
Ecol. 2006;57:128–38. doi:10.1111/j.1574-6941.2006.00097.x.
34. Stewart CJ, Nelson A, Scribbins D, et al. Bacterial and fungal viability in the preterm gut:
NEC and sepsis. Arch Dis Child Fetal Neonatal Ed. 2013. doi:10.1136/archdischild-
2012-302119 (Epub ahead of print).
35. Walker A. Breast milk as the gold standard for protective nutrients. J Pediatr. 2010;156:3–7.
doi:10.1016/j.jpeds.2009.11.021.
36. Thurl S, Munzert M, Henker J, et al. Variation of human milk oligosaccharides in relation to
milk groups and lactational periods. Br J Nutr. 2011;104:1261–71. doi:10.1017/
S0007114510002072.
37. Knol J, Scholtens P, Kafka C, et al. Colon microflora in infants fed formula with galacto- and
fructo-oligosaccharides: more like breast-fed infants. J Pediatr Gastroenterol Nutr.
2005;40(1):36–42.
38. Collado MC, Delgado S, Maldonado A, Rodriguez JM. Assessment of the bacterial diversity
of breast milk of healthy women by quantitative real-time PCR. Lett Appl Microbiol.
2009;48:523–8. doi:10.1111/j.1472-765X.2009.02567.x.
39. Boesten R, Schuren F, Ben Amor K, Haarman M, Knol J, de Vos WM. Bifidobacterium popu-
lation analysis in the infant gut by direct mapping of genomic hybridization patterns: poten-
tial for monitoring temporal development and effects of dietary regimens. Microb Biotechnol.
2011;4:417–27. doi:10.1111/j.1751-7915.2010.00216.x.
40. Martin R, Langa S, Reviriego C, et al. The commensal microflora of human milk: new per-
spectives for food bacteriotherapy and probiotics. Trends Food Sci Technol. 2004;15:121–7.
41. Rao S, Srinivasjois R, Patole S. Prebiotic supplementation in full-term neonates: a systematic
review of randomized controlled trials. Arch Pediatr Adolesc Med. 2009;163:755–64.
doi:10.1001/archpediatrics.2009.94.
42. Scholtens PA, Alliet P, Raes M, et al. Fecal secretory immunoglobulin A is increased in
healthy infants who receive a formula with short-chain galacto-oligosaccharides and long-
chain fructo-oligosaccharides. J Nutr. 2008;138(6):1141–7.
43. Parracho H, McCartney A, Gibson G. Probiotics and prebiotics in infant nutrition. Proc Nutr
Soc. 2007;66:405–11. doi:10.1017/S0029665107005678.
44. Braegger C, Chmielewska A, Decsi T, et al. Supplementation of infant formula with probiotics
and/or prebiotics: a systematic review and comment by the ESPGHAN committee on nutri-
tion. J Pediatr Gastroenterol Nutr. 2011;52:238–50. doi:10.1097/MPG.Ob013e3181fb9e80.
45. Wang Q, Dong J, Zhu Y. Probiotic supplement reduces risk of necrotizing enterocolitis and
mortality in preterm very low-birth-weight infants: an updated meta-analysis of 20 random-
ized, controlled trials. J Pediatr Surg. 2012;47:241–8. doi:10.1016/j.jpedsurg.2011.09.064.
46. Fallani M, Young D, Scott J, et al. Intestinal microbiota of 6-week-old infants across Europe:
geographic influence beyond delivery mode, breast-feeding, and antibiotics. J Pediatr
Gastroenterol Nutr. 2010;51:77–84. doi:10.1097/MPG.0b013e3181d1b11e.
47. Hussey S, Wall R, Gruffman E, et al. Parenteral antibiotics reduce bifidobacteria colonization
and diversity in neonates. Int J Microbiol. 2011;2011:130574. doi:10.1155/2011/130574.
48. Mangin I, Suau A, Gotteland M, Brunser O, Pochart P. Amoxicillin treatment modifies the
composition of Bifidobacterium species in infant intestinal microbiota. Anaerobe.
2010;16:433–8. doi:10.1016/j.anaerobe.2010.06.005.
49. Dethlefsen L, Relman DA. Incomplete recovery and individualized responses of the human
distal gut microbiota to repeated antibiotic perturbation. Proc Natl Acad Sci U S A. 2011;108
suppl 1:4554–61. doi:10.1073/pnas.1000087107.
50. Kummeling I, Stelma FF, Dagnelie PC, et al. Early life exposure to antibiotics and the sub-
sequent development of eczema, wheeze, and allergic sensitization in the first 2 years of
life: the KOALA Birth Cohort Study. Pediatrics. 2007;119:e225–31. doi:10.1542/
peds.2006-0896.
51. Marra F, Lynd L, Coombes M, et al. Does antibiotic exposure during infancy lead to develop-
ment of asthma? A systematic review and metaanalysis. Chest. 2009;136:e30.
52. Gibney MJ, Walsh M, Brennan L, Roche HM, German B, van Ommen B. Metabolomics in
human nutrition: opportunities and challenges. Am J Clin Nutr. 2005;82(3):497–503.
53. Go VL, Nguyen CT, Harris DM, Lee WN. Nutrient-gene interaction: metabolic genotype-
phenotype relationship. J Nutr. 2005;135(12 Suppl):3016S–20.
54. Mutch DM, Wahli W, Williamson G. Nutrigenomics and nutrigenetics: the emerging faces of
nutrition. FASEB J. 2005;19(12):1602–16.
55. Astle J, Ferguson JT, German JB, Harrigan GG, Kelleher NL, Kodadek T, Parks BA, Roth
MJ, Singletary KW, Wenger CD, Mahady GB. Characterization of proteomic and metabolomic
responses to dietary factors and supplements. J Nutr. 2007;137(12):2787–93.
56. Zeisel SH. Nutrigenomics and metabolomics will change clinical nutrition and public health
practice: insights from studies on dietary requirements for choline. Am J Clin Nutr.
2007;86(3):542–8.
57. Scalbert A, Brennan L, Fiehn O, Hankemeier T, Kristal BS, van Ommen B, Pujos-Guillot E,
Verheij E, Wishart D, Wopereis S. Mass-spectrometry-based metabolomics: limitations and
recommendations for future progress with particular focus on nutrition research.
Metabolomics. 2009;5(4):435–58. Epub 2009 Jun 12.
58. Martin FP, Collino S, Rezzi S, Kochhar S. Metabolomic applications to decipher gut micro-
bial metabolic influence in health and disease. Front Physiol. 2012;3:113. doi:10.3389/
fphys.2012.00113. Epub 2012 Apr 26.
59. Moore JB, Weeks ME. Proteomics and systems biology: current and future applications in the
nutritional sciences. Adv Nutr. 2011;2(4):355–64.
60. Moco S, Collino S, Rezzi S, Martin FP. Metabolomics perspectives in pediatric research.
Pediatr Res. 2013;73:570–6.
61. Vernocchi P, Vannini L, Gottardi D, Del Chierico F, Serrazanetti DI, Ndagijimana M,
Guerzoni ME. Integration of datasets from different analytical techniques to assess the impact
of nutrition on human metabolome. Front Cell Infect Microbiol. 2012;2:156.
62. Parker BA, Chapman IM. Food intake and ageing – the role of the gut. Mech Ageing Dev.
2004;125:859–66.
63. Walsh MC, Brennan L, Malthouse JP, Roche HM, Gibney MJ. Effect of acute dietary stan-
dardization on the urinary, plasma, and salivary metabolomic profiles of healthy humans. Am
J Clin Nutr. 2006;84(3):531–9.
64. Wishart DS. Metabolomics application to food science and nutrition research. Trends Food
Sci Technol. 2008;19:482–93.
65. Primrose S, Draper J, Elsom R, Kirkpatrick V, Mathers JC, Seal C, Beckmann M, Haldar S,
Beattie JH, Lodge JK, Jenab M, Keun H, Scalbert A. Metabolomics and human nutrition.
Br J Nutr. 2011;105(8):1277–83. doi:10.1017/S0007114510004812. Epub 2011 Jan 24.
66. Beckmann M, Lloyd AJ, Haldar S, Favé G, Seal CJ, Brandt K, Mathers JC, Draper J. Dietary
exposure biomarker-lead discovery based on metabolomics analysis of urine samples. Proc
Nutr Soc. 2013;72(3):352–61.
67. Rasmussen LG, Winning H, Savorani F, Toft H, Larsen TM, Dragsted LO, Astrup A, Engelsen
SB. Assessment of the effect of high or low protein diet on the human urine metabolome as
measured by NMR. Nutrients. 2012;4(2):112–31. doi:10.3390/nu4020112. Epub 2012 Feb
20.
68. Hedrick VE, Dietrich AM, Estabrooks PA, Savla J, Serrano E, Davy BM. Dietary biomarkers:
advances, limitations and future directions. Nutr J. 2012;11:109. doi:10.1186/1475-2891-11-10.
69. Gu H, Pan Z, Xi B, Hainline BE, Shanaiah N, Asiago V, Gowda GA, Raftery D. 1H NMR
metabolomics study of age profiling in children. NMR Biomed. 2009;22(8):826–33.
70. Fanos V, Chevalier R, Faa G, Cataldi L. Developmental nephrology: from embryology to
metabolomics, vol 10. In: Quartu S Elena (Cagliari), editors. Hygeia Press; 2011. p. 149–215
(available also as ebook)
71. Fanos V, Barberini L, Antonucci R, Atzori L. Pharma-metabolomics in neonatology: is it a
dream or a fact? Curr Pharm Des. 2012;18:2996–3006.
72. Fanos V, Antonucci R, Barberini L, Atzori L. Urinary metabolomics in the newborn and
infants. Adv Clin Chem. 2012;58:193–223.
73. Fanos V, Antonucci R, Barberini L, Noto A, Atzori L. Clinical application of metabolomics
in neonatology. J Matern Fetal Neonatal Med. 2012;25 Suppl 1:104–9.
74. Fanos V, Mussap M, Noto A, Atzori L. Metabolomics in perinatology: where are we now?
Acta Med Port. 2012;25(S2):117–20.
75. Fanos V, van den Anker J, Noto A, et al. Metabolomics in neonatology: fact or fiction? Semin
Fetal Neonatal Med. 2013;18:3–12.
76. Fanos V, Locci E, Noto A, et al. Urinary metabolomics in newborns infected by cytomegalo-
virus: a preliminary investigation. Early Hum Dev. 2013;89:S58–61.
77. Fanos V, Fanni C, Ottonello G, et al. Metabolomics in adult and pediatric nephrology.
Molecules. 2013;18:4844–5.
78. Fanos V, Iacovidou N, Puddu M, et al. Metabolomics in neonatal life. Early Hum Dev.
2013;89:S7–S10C.
79. Fanos V, Atzori L, Makarenko K, Melis GB, Ferrazzi E. Metabolomics application in
maternal-fetal medicine. Biomed Res Int. 2013;2013:720514. doi:10.1155/2013/720514.
Epub 2013 Jun 9. Review.
80. Fanos V, Antonucci R, Atzori L. Metabolomics in the developing infant. Curr Opin Pediatr.
2013;25:604–11.
81. Matsumoto M, Kibe R, Ooga T, Aiba Y, Kurihara S, Sawaki E, Koga Y, Benno Y. Impact of
intestinal microbiota on intestinal luminal metabolome. Sci Rep. 2012;2:233. doi:10.1038/
srep00233.
82. Holmes HC, Snodgrass I, Iles RA. Changes in the choline content of human breast milk in the
first 3 weeks after birth. Eur J Pediatr. 2000;159:198–204.
83. Marincola FC, Noto A, Caboni P, Reali A, Barberini L, Lussu M, Murgia F, Santoru ML,
Atzori L, Fanos V. A metabolomic study of preterm human and formula milk by high resolu-
tion NMR and GC/MS analysis: preliminary results. J Matern Fetal Neonatal Med. 2012;25
Suppl 5:62–7.
84. Miller JB, Bull S, Miller J, McVeagh P. The oligosaccharide composition of human milk:
temporal and individual variations in monosaccharide components. J Pediatr Gastroenterol
Nutr. 1994;19:371–6.
85. Poroyko V, Morowitz M, Bell T, et al. Diet creates metabolic niches in the “immature gut”
that shape microbial communities. Nutr Hosp. 2011;26:1283–95.
86. Sundekilde UK, Larsen LB, Bertram HC. NMR-based milk metabolomics. Metabolites.
2013;3:204–22. doi:10.3390/metabo3020204.
87. Kinross J, Warren O, Basson S, Holmes E, Silk D, Darzi A, Nicholson JK. Intestinal ischemia/
reperfusion injury: defining the role of the gut microbiome. Biomark Med. 2009;3(2):175–92.
doi:10.2217/bmm.09.11.
88. Fahrner R, Beyoğlu D, Beldi G, Idle JR. Metabolomic markers for intestinal ischemia in a
mouse model. J Surg Res. 2012;178(2):879–87. doi:10.1016/j.jss.2012.08.011. Epub 2012
Aug 25.
89. Lin HM, Edmunds SJ, Helsby NA, Ferguson LR, Rowan DD. Nontargeted urinary metabolite
profiling of a mouse model of Crohn’s disease. J Proteome Res. 2009;8(4):2045–57.
90. Murdoch TB, Fu H, MacFarlane S, Sydora BC, Fedorak RN, Slupsky CM. Urinary metabolic
profiles of inflammatory bowel disease in interleukin-10 genedeficient mice. Anal Chem.
2008;80(14):5524–31.
91. Marchesi JR, Holmes E, Khan F, Kochhar S, Scanlan P, Shanahan F, Wilson ID, Wang
Y. Rapid and noninvasive metabonomic characterization of inflammatory bowel disease.
J Proteome Res. 2007;6:546–51.
92. Williams HR, Cox IJ, Walker DG, et al. Characterization of inflammatory bowel disease with
urinary metabolic profiling. Am J Gastroenterol. 2009;104:1435–44.
93. Winterkamp S, Weidenhiller M, Otte P, Stolper J, Schwab D, Hahn EG, Raithel M. Urinary
excretion of N-methylhistamine as a marker of disease activity in inflammatory bowel disease.
Am J Gastroenterol. 2002;97:3071–7.
94. Ponnusamy K, Choi JN, Kim J, Lee SY, Lee CH. Microbial community and metabolomics
comparison of irritable bowel syndrome faeces. J Med Microbiol. 2011;60(Pt 6):817–27.
95. Bertini I, Calabro A, De Carli V, Luchinat C, Nepi S, Porfirio B, Renzi D, Saccenti E, Tenori
L. The metabonomic signature of celiac disease. J Proteome Res. 2009;8:170–7.
96. Tjellstrom B, Stenhammar L, Hogberg L, Falth-Magnusson K, Magnusson KE, Midtvedt T,
Sundqvist T, Norin E. Gut microflora associated characteristics in children with celiac dis-
ease. Am J Gastroenterol. 2005;100:2784–8.
97. Sellitto M, Bai G, Serena G, Fricke WF, Sturgeon C, et al. Proof of concept of microbiome-
metabolome analysis and delayed gluten exposure on celiac disease autoimmunity in geneti-
cally at-risk infants. PLoS One. 2012;7(3):e33387. doi:10.1371/journal.pone.0033387.
98. Di Cagno R, De Angelis M, De Pasquale I, Ndagijimana M, Vernocchi P, Ricciuti P, Gagliardi
F, Laghi L, Crecchio C, Guerzoni ME, Gobbetti M, Francavilla R. Duodenal and faecal
microbiota of celiac children: molecular, phenotype and metabolome characterization. BMC
Microbiol. 2011;11:219. doi:10.1186/1471-2180-11-219.
99. Wolak JE, Esther Jr CR, O’Connell TM. Metabolomic analysis of bronchoalveolar lavage
fluid from cystic fibrosis patients. Biomarkers. 2009;14:55–60.
100. Robroeks CM, van Berkel JJ, Dallinga JW, Jöbsis Q, Zimmermann LJ, Hendriks HJ, Wouters
MF, van der Grinten CP, van de Kant KD, van Schooten FJ, Dompeling E. Metabolomics of vola-
tile organic compounds in cystic fibrosis patients and controls. Pediatr Res. 2010;68(1):75–80.
101. Newgard CB, An J, Bain JR, Muehlbauer MJ, Stevens RD, Lien LF, Haqq AM, Shah SH,
Arlotto M, Slentz CA, Rochon J, Gallup D, Ilkayeva O, Wenner BR, Yancy Jr WS, Eisenson
H, Musante G, Surwit RS, Millington DS, Butler MD, Svetkey LP. A branched-chain amino
acid-related metabolic signature that differentiates obese and lean humans and contributes to
insulin resistance. Cell Metab. 2009;9:311–26.
102. Huffman KM, Shah SH, Stevens RD, Bain JR, Muehlbauer M, Slentz CA, Tanner CJ,
Kuchibhatla M, Houmard JA, Newgard CB, Kraus WE. Relationships between circulating
metabolic intermediates and insulin action in overweight to obese, inactive men and women.
Diabetes Care. 2009;32:1678–83.
103. Fiehn O, Garvey WT, Newman JW, Lok KH, Hoppel CL, Adams SH. Plasma metabolomics
American women. PLoS One. 2010;5:e15234.
104. Boulange CL, Claus SP, Chou CJ, Collino S, Montoliu I, Kochhar S, Holmes E, Rezzi S,
Nicholson JK, Dumas ME, Martin FP. Early metabolic adaptation in C57BL/6 mice resistant
to high fat diet induced weight gain involves an activation of mitochondrial oxidative path-
ways. J Proteome Res. 2013;12:1956–68.
105. Wang TJ, Larson MG, Vasan RS, Cheng S, Rhee EP, McCabe E, Lewis GD, Fox CS, Jacques
PF, Fernandez C, O’Donnell CJ, Carr SA, Mootha VK, Florez JC, Souza A, Melander O,
Clish CB, Gerszten RE. Metabolite profiles and the risk of developing diabetes. Nat Med.
2011;17:448–53.
106. Friedrich N. Metabolomics in diabetes research. J Endocrinol. 2012;215:29–42.
107. Wang-Sattler R, Yu Z, Herder C, Messias AC, Floegel A, He Y, Heim K, Campillos M,
Holzapfel C, Thorand B, Grallert H, Xu T, Bader E, Huth C, Mittelstrass K, Döring A,
Meisinger C, Gieger C, Prehn C, Roemisch-Margl W, Carstensen M, Xie L, Yamanaka-
Okumura H, Xing G, Ceglarek U, Thiery J, Giani G, Lickert H, Lin X, Li Y, Boeing H, Joost
HG, de Angelis MH, Rathmann W, Suhre K, Prokisch H, Peters A, Meitinger T, Roden M,
Wichmann HE, Pischon T, Adamski J, Illig T. Novel biomarkers for pre-diabetes identified
by metabolomics. Mol Syst Biol. 2012;8:615. doi:10.1038/msb.2012.43.
108. Meng Q, Mäkinen VP, Luk H, Yang X. Systems biology approaches and applications in obesity,
diabetes, and cardiovascular diseases. Curr Cardiovasc Risk Rep. 2013;7(1):73–83. Epub
2012 Oct 18.
109. Yamakado M, Tanaka T, Nagao K, Ishizaka Y, Mitushima T, Tani M, Toda A, Toda E, Okada
M, Miyano H, Yamamoto H. Plasma amino acid profile is associated with visceral fat accu-
mulation in obese Japanese subjects. Clin Obes. 2012;2:29–40.
110. Rhee EP, Cheng S, Larson MG, Walford GA, Lewis GD, McCabe E, Yang E, Farrell L, Fox
CS, O’Donnell CJ, Carr SA, Vasan RS, Florez JC, Clish CB, Wang TJ, Gerszten RE. Lipid
prediction in humans. J Clin Invest. 2011;121:1402–11.
111. Wahl S, Yu Z, Kleber M, Singmann P, Holzapfel C, He Y, Mittelstrass K, Polonikov A, Prehn
C, RömischMargl W, Adamski J, Suhre K, Grallert H, Illig T, Wang-Sattler R, Reinehr
T. Childhood obesity is associated with changes in the serum metabolite profile. Obes Facts.
2012;5:660–70.
112. Mihalik SJ, Michaliszyn SF, de las Heras J, Bacha F, Lee S, Chace DH, DeJesus VR, Vockley
J, Arslanian SA. Metabolomic profiling of fatty acid and amino acid metabolism in youth
with obesity and type 2 diabetes: evidence for enhanced mitochondrial oxidation. Diabetes
Care. 2012;35(3):605–11.
113. He Q, Ren P, Kong X, Wu Y, Wu G, Li P, Hao F, Tang H, Blachier F, Yin Y. Comparison of
serum metabolite compositions between obese and lean growing pigs using an NMR-based
metabonomic approach. J Nutr Biochem. 2012;23(2):133–9.
114. Dessì A, Fanos V. Pediatric obesity: could metabolomics be a useful tool? J Pediatr Neonatal
Individ Med. 2013;2(2):e020205. doi:10.7363/020205.
115. Bassareo PP, Fanos V, Noto A, Solla P, Barberini L, Flore G, Puddu M, Mercuro G. Clinical
metabolomics and hematic ADMA predict the future onset of cardiorenal syndrome in young
grown-up subjects who were born preterm . Clin Biochem. 2013;47(6):423–6
116. Dessì A, Puddu M, Ottonello G, Fanos V. Metabolomics and fetal-neonatal nutrition: between
“not enough” and “too much”. Molecules. 2013;18(10):11724–32. doi:10.3390/
molecules181011724.
117. Fanos V, Iacovidou N, Puddu M, Ottonello G, Noto A, Atzori L. Metabolomics in neonatal
life. Early Hum Dev. 2013;89 Suppl 1:S7–10. doi:10.1016/S0378-3782(13)70003-3. Review.
118. Dessì A, Fanos V. Myoinositol: a new marker of intrauterine growth restriction? J Obstet
Gynaecol. 2013;33(8):776–80. doi:10.3109/01443615.2013.831046.
119. Antonucci R, Pilloni MD, Atzori L, Fanos V. Pharmaceutical research and metabolomics in
the newborn. J Matern Fetal Neonatal Med. 2012;25 Suppl 5:22–6.
120. Syggelou A, Iacovidou N, Atzori L, Xanthos T, Fanos V. Metabolomics in the developing

human being. Pediatr Clin North Am. 2012;59(5):1039–58.
121. Mussap M, Antonucci R, Noto A, Fanos V. The role of metabolomics in neonatal and pediatric
laboratory medicine. Clin Chim Acta. 2013;426:127–38.
122. Leichtea AB, Dufourb JF, Fedlera GM. Potentials and pitfalls of clinical peptidomics and
metabolomics. Swiss Med Wkly. 2013;143:w13801.
Chapter 8
Metabolomics and Milk: The Development
of the Microbiota in Breastfed Infants
J. Bruce German, Jennifer T. Smilowitz, Carlito B. Lebrilla, David A. Mills,

and Samara L. Freeman
Abstract Metabolomics provides a valuable strategy for describing and annotating

the structures, compositions, and functions of mammalian milk. Detailed analyses
of the complex components of milk have revealed an unexpected diversity of
glycans consisting of oligosaccharides, glycoproteins, and glycolipids, all of which
help shape the intestinal environment and in particular the intestinal microbiome of
breastfed babies. Using complete and partial ensembles of glycan mixtures, the
holistic principles of metabolomics analytics were leveraged for microbial screen-
ing studies. The complex glycans of human milk proved to be highly selective in
their ability to support the growth of only a very rare group of enteric bacteria.
These studies led to the conclusion that a signature achievement of breast milk is the
development of a unique milk-oriented intestinal microbiota that results from a
functional overlap of stereospecific glycan biosynthesis in maternal mammary epi-
thelia with equally stereospecific glycosidase enzymes encoded within the genome
of the commensal bacteria. Clinical evidence in support of that hypothesis has now
been generated by the simultaneous administration and quantitation of the entire
repertoire of glycans in the milk going in and the feces coming out of human infants.
These platforms of systems biology combining separation technologies coupled to
highly accurate and sensitive mass spectrometry with exhaustive library develop-
ment and computational tools provide a model for success in understanding biologi-
cal processes. Metabolomics is now extending that understanding of the infant
J.B. German (*) • J.T. Smilowitz • S.L. Freeman

Department of Food Science and Technology, Foods for Health Institute,
University of California Davis, Davis, CA 95616, USA
e-mail: [email protected]; [email protected]; [email protected]
C.B. Lebrilla
Department of Chemistry, Foods for Health Institute, University of California Davis,
Davis, CA 95616, USA
D.A. Mills
Department of Viticulture & Enology, Department of Food Science and Technology,
Foods for Health Institute, University of California Davis, Davis, CA 95616, USA

DOI 10.1007/978-1-4471-6539-2_8
148 J.B. German et al.
microbiota and its phenotype to the role of complex glycans in the microbiota of all
ages. The relentless selective pressure on the process of lactation within the
mammary epithelial cell over millennia of evolution has been to nourish, protect,
and support the survival of the mother infant pair. The principles that have emerged
to nourish infants provide a guiding model for diet and health of all humans. The
tools of metabolomics are proving successful in revealing the mechanisms behind
milk’s “genius.”
Keywords Milk • Glycobiology • Microbiota • Infant • Oligosaccharides •

Glycosidase • Glycan • Glycomics • Evolution • Lactation
8.1 Introduction
The mother and breastfed infant dyad provides a model to understand diet in its
larger context. The scientific opportunities afforded by this model are transforma-
tive. Mechanistic insights to the targets of dietary inputs can be revealed by studying
milk genomics, chemical composition, biological properties, and its diversity across
mammals and temporally across lactation. The clinical comparison of exclusive
breastfeeding against various formulas provides a powerful framework to study
structurally defined diets and monitor the consequences of those structures on health
and disease. In this respect, studies on milk and the lactating mammary gland pro-
vide unique insights into the mechanisms by which diet can act in protection and
prevention. The mammary epithelial cell is a bioreactor for bioengineering complex
structures and activities that act upon virtually all of the infant’s processes: immu-
nity, growth and development, metabolism, physiology, neurological development,
and microbiota colonization and maturation. The components of milk execute on
this biological blueprint using integrative and pleiotropic mechanisms that are dif-
ficult if not impossible to identify using the reductionist strategies of traditional
biological chemistry. The comprehensive nature of the omic sciences and in particu-
lar metabolomics is changing the way milk is studied. Milk is the functional output
of mammary metabolism, and it’s a biofluid representing maternal genetics, health
status, and environments.
A striking example of the principles and technologies of metabolomics applied
to breast milk is in the area of glycomics; glycomics has revealed the diversity and
abundance of glycans notably the free human milk oligosaccharides (HMOs) that
are relatively unique to lactation and the glycosylated proteins, peptides, and lipids.
Interestingly, glycans reach the large intestine and can ultimately be excreted and
measured in the stool in healthy infants. This is a paradox if milk is considered a
source of digestible nutrients for the infant. The resolution of this paradox is found
in the fact that in most breastfed infants, these glycans disappear from stool, coinci-
dent with the appearance of a group of bacteria capable of digesting and utilizing
them as growth substrates. The value of this relationship between maternal milk and
8 Metabolomics and Milk: The Development of the Microbiota in Breastfed Infants 149
intestinal bacteria to shape infant postnatal development is still being revealed,

ranging from protection from pathogenic bacteria, viruses, and toxins, to promoting
neurological and immune systems and enhancing barrier function of intestinal epi-
thelia. The highly selective digestibility of milk glycans act to shape the intestinal
microbiome orchestrating its transition from the sterile uterus through the chaotic
introduction of environmental bacteria at birth to a stable milk-oriented microbiome
(MOM). This convergence of an entire metabolite class, glycans with the selective
metabolism of bacteria and their interaction with the human host, is an opportunity
to define metabolism as a dietary variable and study the structure-function relation-
ships between diet, metabolism, and intestinal bacteria development. These studies
provide broader principles for nourishing complex microbiota throughout life.
At the core, comprehensive and accurate measurement of the structures and compo-
sition of milk’s glycan metabolome is required.
Analytical chemistry has only recently brought the tools needed to measure
glycobiology, the free oligosaccharides and glycans bound to proteins, peptides,
and lipids in milk. Instrumentation is not sufficient; mass spectrometry must be
coupled to separation technologies, enzyme biotechnologies, and bioinformatics
tools to assemble all of the information into computationally accessible libraries.
These technological advancements have led to the discovery that glycans are a cen-
tral component of all mammalian milks, are variable across lactation and among
women, and provide a wide diversity of structures to diverse functions [1–6].
8.2 Metabolomics and Human Milk
The simplifying elegance of the linear encoding of protein structure from DNA,
RNA to protein sequence that is so empowering to biology from evolution to func-
tion is equally enabling to scientific research. Scientists have been wonderfully
successful in annotating DNA-dependent biological processes because of the sim-
plicity of linear sequence. Metabolism does not possess this simplicity. The dizzy-
ing complexity of metabolism must now be studied the old fashioned way, by
measuring it. Scientists are beginning to assemble the technologies to measure
metabolites in the accuracy, sensitivity, and comprehensiveness that reflect actual
biology. A broad goal of food research is to build a linear understanding from the
genetics of agricultural commodities, through their metabolism and thence compo-
sition as foods to the specific actions of those components on the metabolism and
ultimately health of individual consumers (Fig. 8.1). Step one is to define the genetic
and phenotypic basis of food composition through commodity growth and process-
ing. The next step is to understand the principles by which human metabolism is
controlled via these exogenous dietary components. This challenge will be particu-
larly daunting in higher organisms due to the importance of structure to function. In
higher organisms metabolites are distributed according to the cells, tissues, organs,
and whole bodies. This structural dimension will demand that metabolites are mea-
sured as a function of the 3-dimensional structures of their immediate environment,
150
Fig. 8.1 The bidirectional flow of metabolic information through agriculture, food, and health. The genomics of agricultural commodities defines the metabolic
machinery of farmed organisms which, once measured by metabolomics, can be guided by selective breeding and explicit genetic engineering. The metabolome
compositions of harvested organisms that are the result of their metabolism are in turn alterable post-harvest by a wide variety of processing alternatives. The
final compositions of chosen foods measured by food metabolomics define the overall diet compositions of individuals/families. Their diets influence their acute
health which is in turn measureable by metabolomics of various body fluids. Departures from desired health trajectories detected by individual and population
J.B. German et al.
measurements then feed back into all of the agriculture and food input variables to alter diet composition and guide individuals to improved health
Fig. 8.2 The tripartite evolutionary relationship for mothers, infants, and their microbiota. The
importance of the infant microbiota to its survival and genetic success is implied by the substantial
investment of lactation in the control of this ecosystem. Understanding how and why lactation con-
trols the infant microbiota provides scientific insights to microbial ecosystems in human intestine and
beyond to all complex ecosystems in which food is a selective and discriminating input variable
techniques for which are only beginning to emerge. We have taken the approach of
using the interaction between milk and microorganisms as a model for dietary
metabolomics (Fig. 8.2). The principles emerging from this research provide scien-
tists with examples of the interactions between diet and metabolism that may be
instructive for higher animals.
8.3 Milk Glycomics
Glycans are the biopolymer class that has been largely ignored in spite of their
abundance across the phylogenetic tree and through evolution [7]. Despite their
importance in health and disease, they are not sequence encoded but rather the prod-
ucts of enzymatic metabolism. As a result of metabolic synthesis, the number of
potential structures is massive contributing to the structural diversity seen in milk.
The complexity of a glycan is the result of a number of factors including

branching, the number of different sugars, the stereospecific linkages of those sug-
ars all leading to multiple isomers even for a single net mass to which must be added
the dimension of conjugation: they are free or bound to proteins, peptides, or lipids
again in a heterogeneous but stereospecific manner. Biology has apparently
employed the combination of metabolism and structural diversity to leverage glyco-
biology into a variety of functions and most notably recognition. Perhaps not
surprisingly, throughout evolution, the glycans on surfaces of cells are distinct both
to “self” and to “foreign” organisms as the molecular basis for individuality. This
dimension of diversity that is clearly of value to biological organisms is instead to
scientists a nightmare precisely because there is no corresponding genetic template;
every glycan must be explicitly analyzed to be identified. Research into the tech-
nologies and methodologies to routinely and comprehensively measure the glycans
in biological and clinical samples is only now emerging, and as a result, a quantita-
tive, metabolomics approach to glycobiology is becoming possible. Glycomics is
defined as the systematic study of the total complement of sugars present in an
organism in their free or protein and lipid-bound states [8, 9]. Researchers are now
using glycomics to understand diet and health in the context of lactation, milk, and
the role of complex glycans in various aspects of the development of the mamma-
lian neonate.
8.3.1 Milk Oligosaccharides
The field of milk glycobiology is not new, and in fact many of the presumed roles of
glycans throughout biology have been first discovered by examining milk.
Nonetheless, the complexity of glycan structures has been a major hurdle to under-
standing specific structure – function relationships beyond binding assays.
The oligosaccharides of human milk have become of particular interest in large part
because they are an abundant (1–2 % w/v) and yet indigestible by the neonate. The
biological challenges posed by this apparent paradox propelled a few key laborato-
ries to pursue the analytical challenges of identifying and quantifying them. The
revolution that they are bringing to analytical glycomics has been the result of inno-
vations in separation science, enzyme biotechnology, mass spectrometry, automated
library development, and computational toolsets.
8.3.2 Separation Science for Oligosaccharides
Liquid chromatography has been applied to glycan separation, yet neither normal
nor reverse stationary phases provide sufficient separation power to successfully
resolve glycan diversity in structure. Porous graphitized carbon (PGC) has been
used as a uniquely selective stationary phase for bulk purification of glycans for
many years. Only recently has it been possible to formulate PGC as an HPLC
stationary phase for the analysis of native oligosaccharides [10]. This combination
of high-efficiency, high reproducibility, and stationary phase selectivity in an HPLC
system has provided the first generation of separation platforms capable of the
extensive separation of glycan isomers [11]. The future of oligosaccharide and gly-
can analyses will be extending this selectivity and efficiency to stereospecificity,
quantitation, and throughput.
8.3.3 Stereospecificity of Glycan Structures
Glycans are nothing if not a forest of stereospecificity. The total number of possible
structures imaginable is astronomical, yet because these glycan polymers are all the
results of stereospecific enzymatic reactions, the actual number of structures found
in biology is manageably finite and approachable by modern analytics and library
systems. Nonetheless, precise oligosaccharide structures cannot be unequivocally
identified on chiral separation phases and instead must still be determined the old
fashioned way, by cleaving with stereospecific enzymes as an explicit step in the
analysis [12]. The use of stereospecific enzymes will likely remain the most efficient
means of assigning precise structures to oligosaccharides precisely because once a
biological source is accurately described, it is not necessary to perform stereospe-
cific analyses every subsequent analysis. The pragmatic proof of this principle has
been demonstrated for the various milk oligosaccharides that have been analyzed
for their stereospecificity [13].
The establishment of accurate metabolomics today cannot be achieved with
online identification systems due to the complexity of the possible glycan structures
relative to the separation platforms and mass spectrometry accuracy available.
Instead effective methods for the structural identification of HMOs requires the
construction of detailed libraries that map structures into analytical platforms taking
advantage of the combinations of MS, tandem MS, and exoglycosidase digestion
[12]. Neutral [14] and anionic milk oligosaccharides from humans [15] totaling 75
structural isomers have been annotated in this approach (Figs. 8.3, 8.4, and 8.5).
Once begun, the library strategy has been extended to milks from other mammals to
over 200 complete structures. This basic strategy is appropriate for the vast majority
of applications to human milk biology since 50 structures represent 99 % of the total
abundances of oligosaccharides in human milks [14–16].
Metabolomics is of relatively little utility if it only identifies structures and
cannot quantify the absolute amounts of metabolites within biological samples.
As a subset of the metabolome, glycan quantitation remains a major obstacle to
metabolomics of glycobiology. Oligosaccharides lack discriminating chromophores
for spectral detectors. As a result, oligosaccharides are often derivatized with
absorbing labels including anthranilic acid (AA) or 2-aminobenzamide (AB) for
quantitation [17]. To date the varying ionization efficiencies of glycans compromise
the use of mass spectrometry of oligosaccharides. The alternative, using isotopically
Fig. 8.3 Basic oligosaccharide structures in milk. The structural core of oligosaccharides is illus-
trated, the key sugar monomers that make up the oligosaccharide compositions and the possible
stereospecific glycosidic bonds that are possible. Bifidobacteria longum subspecies infantis con-
tains the genetic capability to synthesize ostensibly all of the enzymes necessary to cleave this
array of complex glycans
enriched internal standards, while making MS highly accurate [18], requires the
synthesis of the entire library of potential structures which is not currently available.
Quantitation of metabolites remains the great challenge for the applications of
metabolomics to its most relevant applications in health.
The presence of oligosaccharides has been confirmed in very early mammals and
marsupials [19]. Thus, indigestible carbohydrate biopolymers provided a selective
advantage throughout mammalian lactation. This advantage has apparently contin-
ued up to humans. Human milk contains greater concentration and diversity of sol-
uble oligosaccharides than other mammalian milks [20] ranging from on average
7 g/L mature to 23 g/L in colostrum [21, 22]. These soluble oligosaccharides are
composed of glucose (Glc), galactose (Gal), N-acetylglucosamine (GlcNAc),
fucose (Fuc), and sialic acid (NeuAc) monosaccharides. The basic biochemistry of
oligosaccharide synthesis in the mammary gland is initiated by a lactose core of Gal
and Glc catalyzed by β-galactotransferase in the presence of α-lactalbumin. The
vast majority of HMO structures are based on this lactose core [23]. Lactose is then
decorated by β1–3 linkage to lacto-N-biose (GlcNAc linked to Gal by β1–3 linkage)
or by β1–6 linkage to N-acetyllactosamine (GlcNAc linked to Gal by β1–4 linkage).
HMO Structural Diversity FOS/GOS
lacto-N-biose lactose β1-4
β1-4 β1-4
α1-2
β1-6 β1-4
α1-3 β1-4
α1-4 β1-3 β1-3 β1-4
β1-3 β1-3
β1-4 β1-6 β1-3 β1-3

α2-3
α2-6 n=0-15 β1-3
Lacto-N-tetraose (LNT) Lacto-N-neotetraose (LNnT)
N-acetyllactosamine (type 1 chain) (type 2 chain) Lacto-N-hexaose (LNH)
β
β1-4 β1-4
β
β1-4 β1-4 β1-4
α1-2 α1-3 β1-3 β1-3
β1-4 β1-4
β1-6 β1-4 β1-6 β1-4 β1-3
2’-fucosyllactose 3-fucosyllactose β1-4
(2’FL) (3FL) β1-3 β1-3
β1-4 β1-3 β1-3 β1-3

α2-6 β1-3 β1-3
β1-4 β1-3
α2-3
3’-sialyllactose 6’-sialyllactose iso-Lacto-N-octaose iso-Lacto-N-neooctaose para-Lacto-N-octaose

(3’SL) (6’SL)
β1-4 β1-4 β1-4 β1-4

α1-4 β1-4 β1-3
β1-3 β1-3 β1-3 α1-3
β1-3 β1-3 α1-3 β1-3

α1-2
Lacto-N-fucopentaose I Lacto-N-fucopentaose II Lacto-N-fucopentaose III Lacto-N-fucopentaose V

(LNFP I) (LNFP II) (LNFP III) (LNFP V)
β1-4 β1-4 β1-4 β1-4

α2-6 α2-6 α2-6
β1-3 β1-3 β1-3 β1-3
β1-4
β1-3 β1-3 β1-3
α2-3 α 2-3
LS-Tetrasaccharide a LS-Tetrasaccharide b LS-Tetrasaccharide c Disialyllacto-N-tetraose

(LST a) (LST b) (LST c) (DSLNT)
Glucose (Glc) Galactose (Gal) N-acetylglucosamine (GlcNAc) Fucose (Fuc) N-acetylneuraminic acid (NeuAc)
8 Metabolomics and Milk: The Development of the Microbiota in Breastfed Infants
Fig. 8.4 Structural diversity of complex free glycans in human milk. The cartoon approximations of a subset of oligosaccharide structures in human milk attest
to the diversity and yet specificity of this biomolecule class. Milk oligosaccharides as highly structured biomolecules are the result of multiple sugars joined by
155
multiple stereospecific linkages

Composition Abundance (counts per second)

HMO Structure Mass RT Hex Fuc HexNAc NeuAc Milk Feces Urine
1 3’FL 490.190 1.31 2 1 0 0 NO NO NO

1-0
2 2’FL 490.190 11.69 2 1 0 0 NO NO NO

1-2
3 3’SL 635.227 22.330 2 0 0 1 58916 191249 370266

2-3
2-6
4 6’SL 635.227 15.231 2 0 0 1 30524 50386 NO
5 LOFT 636.248 14.470 2 2 0 0 NO 11395 11197

1-2 1-3
2-6
6 6’SLN 676.254 14.998 1 0 1 1 NO 59563 115044
7 3’SLN 676.254 23.061 1 0 1 1 NO NO 227705

2-3
8 LNT 709.264 15.090 3 0 1 0 21650694 11349391 4096506
9 LNOT 709.264 15.200 3 0 1 0 410714 772639 83004
1-4
10 3’Sle 822.312 14.609 1 1 1 1 NO NO 4368
2-3
Fig. 8.5 Extracted page of the table from the entire library of oligosaccharides in milks. Table
illustrates the data possible to acquire with modern glycomics platforms including accurate mass,
retention time, sugar subunits, and the abundance in milk and in various biofluids from infants
This growing chain structure can be further elongated with lacto-N-biose and
N-acetyllactosamine by β1–3 and β1–6 linkages; Fuc connected with α1–2, α1–3,
or α1–4 linkages and/or NeuAc residues attached by α2–3 or α2–6 linkages at the
terminal positions (Fig. 8.3). The terminal sugars are particularly diagnostic of dif-
ferent mammalian milks, 60–80 % of HMOs are fucosylated, and 10–15 % of
HMOs are sialylated in human milk [24].
8.4 Annotating the Functions of Human Milk Glycans
The structures of milk oligosaccharides have been selected for an unusual biological
value: not to be consumed by infants. This selective pressure on lactation has been
particularly intense since milk is the sole source of nourishment for mammalian
infants. The genetics, synthesis, and structures of oligosaccharides in milk are
unequivocally discoverable. However, the functions of oligosaccharides that were
the basis for their emergence and persistence through evolution are not as easily
discovered. The process of understanding their actions must first identify their
actions, and then each of these actions must then be tested mechanistically as an
actual valuable function in vivo. The most extensive approach to evaluating the
actions and potential functions of oligosaccharides in human milk has been to estab-
lish the detailed support of the growth of specific strains of bacteria notably bifido-
bacteria [25, 26]. While the mechanisms and extent of microbial diversity in
breastfed infants are still being actively documented, the basic observation that
bifidobacterial species dominate the microbiota of breastfed infants around the
world compared with formula-fed infants has been well established [27]. How an
intestinal microbial ecosystem maintains a dominant and consistent bacterial popu-
lation in the face of repeated and diverse inoculations with environmental microor-
ganisms has been largely speculative until recently. Research has revealed the
remarkable interaction between the stereospecific linkages defining the structures of
milk oligosaccharides and the genetic repertoire of stereospecific glycosidases and
solute-binding proteins that provide these bacteria a distinct competitive growth
advantage within the intestine of the breastfed infant.
8.4.1 Screening Bacteria for Growth on Oligosaccharides
In an ongoing search for biological activities of these molecules to justify their

abundance and diversity in milk, a prevailing hypothesis was that they are sub-
strates for bacterial growth. However, no studies had yet documented that fact nor
whether growth was selective among bacteria. Initial growth experiments in fact
failed to demonstrate significant growth of bacteria when human milk oligosac-
charides were the sole source of carbon in an otherwise supportive medium [25].
A series of subsequent experiments revealed that among gut-related bacteria
tested (including Lactobacillus, Clostridium, Eubacterium E. coli, Veillonella,
Enterococcus isolates), only Bifidobacterium and Bacteriodes species grew to high
cell densities [28]. Growth on HMO was found in a select group of B. bifidum and
B. longum subsp. infantis strains. In these same isolated growth conditions, even
isolates of B. longum subsp. longum and B. breve showed poor growth, and other
strains of B. adolescentis and B. animales were ostensibly unable to grow on HMO
[29] (Fig. 8.6).
HMO abundance in pooled breast milk
% HMO abundance in breast milk

25
B. infantis
2.5 20
2 15
OD (600 nm)
1.5 10
5
1
B. breve 0
0.5 B. longum HMO m/z
Post B. infantis growth

0 Time(hours) 100
% HMO Consumed 50
50
100
HMO m/z
Fig. 8.6 Growth curves and human milk oligosaccharide consumption by Bifidobacteria longum
subspecies infantis. The growth curves of isolated bacteria (indicated) on media containing only
isolated milk oligosaccharides as carbon source of B. longum, B. longum subspecies infantis, and
B. breve are shown, illustrating the conspicuously greater growth of B. infantis. The most abundant
oligosaccharides in human milk are histogrammed in the figure on the right before and after con-
sumption by B. infantis illustrating the complete consumption of the majority of oligosaccharides
by this bacterium [25]
The complex mechanisms by which milk oligosaccharides guide bacterial

growth within the ecosystem of the infant intestine have been elaborated in a series
of microbial studies. Among the bifidobacteria that are able to consume HMO, dif-
ferent strategies are present to use HMO as a substrate. In isolated growth studies of
HMO consumption, B. longum subsp. infantis ATCC15697 most efficiently con-
sumed oligosaccharides seven sugars (DP) or below [25]. Oligosaccharides below
ten sugars are the majority of species human milk [10]. Other bifidobacteria includ-
ing B. longum subsp. longum DJO10A and B. breve ATCC15700 that grew slowly
on pooled HMO were found to be consuming mostly a single, nonfucosylated/non-
sialylated species, LNnT. LNnT is present in breast milk yet a small portion of the
overall HMOs. B. breve did grow in culture on all the monomer constituents of
HMO and thus if present within the gastrointestinal tract could grow on liberated
monosaccharides.
The bacteria that were found to be capable of growing on HMO were analyzed
for the presence of metabolic activities towards complex oligosaccharides including
the key sialidase and fucosidase activities required to deconstruct complex glycan
structures. Among the strains examined, fucosidase activity was present in B.

longum subsp. infantis and was only detected upon growth on HMO [25].
Distinct strategies for catalytic activity on complex biopolymer destruction are
known among intestinal bacteria. The majority of intestinal bacteria secrete extra-
cellular glycosidase enzymes that liberate free sugars that are then subsequently
taken up by bacteria and metabolized. Select bifidobacteria use lacto-N-biosidase
activity to break down oligosaccharides [30]. LNB is transported into B. bifidum via
an ABC transporter and an associated LNB-specific solute-binding lipoprotein
whereby it is further processed and fed into the central metabolic pathway [31].
The discovery that B. longum subsp. infantis ATCC15697 was uniquely capable
of growing on human milk oligosaccharides led to an immediate project to sequence
its genome. No prior experience prepared the investigators for the elegance of the
genetic repertoire of this organism’s sequence. B. longum subsp. infantis ATCC15697
has become the blueprint for understanding the genetic basis of glycan-specific
growth and phenotype [32]. This specific strain possesses clusters of genes associ-
ated with its unique phenotype distributed in the genome into four loci. The most
informative, HMO cluster 1 (Fig. 8.7), contains all the necessary glycosidases (siali-
dase, fucosidase, galactosidase, and hexosaminidase) and transporters necessary for
importing and metabolizing HMO. Sequencing more isolates for HMO-related
genomic architecture among B. longum subsp. infantis isolates provides a detailed
genetic map of the mechanisms behind the vigorous growth of this clade on HMO.
The bacterial model of metabolomics illustrates the complexity of structure
within metabolic pathways. Within the large HMO cluster (Fig. 8.7) are genes
encoding an interesting group of extracellular solute-binding proteins (SBP; pfam
01547) demonstrated to bind oligosaccharides. These proteins provide two func-
tions for the bacteria in their ecological niche of the breastfed infant intestine. These
solute-binding proteins would tether the bacteria to glycans on the luminal side of
the infant intestine and provide a net coverage of microbial binding sites thus block-
ing potential pathogens from the infant. Of more direct value to the bacterium, these
solute-binding proteins would internalize free oligosaccharides directly infusing
substrate into its endogenous metabolism. This substrate sequestering mechanism
provides the B. longum subsp. infantis a unique foraging advantage in the overall
microbial community. These solute-binding proteins also appear to be of singular
advantage to the mammalian infant gut. A subset of these genes shows a pronounced
evolutionary divergence from other SBP family 1 proteins in bifidobacteria [32].
The emergence of these genes is consistent with their functions as a mechanistic
basis of symbiosis with humans through their interaction with milk oligosaccha-
rides. The B. longum subsp. infantis genome has been shown to contain 21 family 1
SBP, more than most bifidobacteria.
The results of genomic analyses of bifidobacteria illustrate that HMO-related
clusters are shared among all B. longum subsp. infantis isolates that have been
examined to date, yet they are notably absent in other sequenced bifidobacteria,
such as B. longum subsp. longum DJO10A [33] and B. adolescentis ATCC15703
All 4 glycosyl hydrolases Array of oligosaccharide transporters

0 10 kb 20 kb 30 kb 40 kb
permease permease SBP permease
ESB ?-
galactosidase fucosidase SBP SBPsialidase SBP SBP SBP hexosaminidase ATPase
Extracelluar Solute binding

protein (SBP)
Glycolytic enzymes permease cytoplasm
ATP
ATPase
ADP
Fig. 8.7 Gene cluster 4 from the complete genome of B. infantis illustrating the location of the
genes encoding glycosidases and oligosaccharide transporters [32]. The putative glycosidic
linkages on which the glycosidic enzymes react are shown below left and the model of the cell
membrane-bound solute-binding protein complex is shown below right
(GenBank AP009256), which grow weakly or not all (respectively) on HMO [29].
The elegance of microbial genetics is illustrated by the HMO-related gene set shared
between ATCC15697 and DJO10A. This seven-gene operon is responsible for LNB
metabolism further evidence of evolution selecting for metabolic substrate utiliza-
tion [34]. Given that DJO10A is able to weakly grow on HMO and glycoprofiling
indicated a small consumption of LNnT, it is tempting to speculate that this operon
is linked to consumption of that particular HMO moiety.
While it is very hard to generalize the mechanisms of HMO catabolism across
bifidobacteria because of strain heterogeneity and taxonomic confusion [35] within
the genera, several important trends have emerged. The most common infant-borne
bifidobacteria, B. bifidum, B. longum subsp. infantis, B. longum subsp. longum, and
B. breve, possess different modes for consumption of HMO (Fig. 8.6). B. longum
subsp. infantis likely imports the lower molecular weight oligosaccharides via an
army of dedicated ABC transporters. Once inside the cell, these oligosaccharides
are catabolized by a complement of glycosidases prior to entry of the monosaccha-
rides into central metabolic pathways. In contrast, B. bifidum exports fucosidases
and lacto-N-biosidase to remove LNB from the HMO structure (leaving the free
fucose and sialic acid behind) [48], internalize the free LNB and catabolize it intra-
cellularly. Both B. breve and B. longum subsp. longum are able to consume free
LNnT from an HMO pool, whereas B. breve can also grow on the various monomer
constituents of HMO [48]. These different strategies suggest a possible mechanism
for niche partitioning among the different bifidobacterial species within the devel-
oping infant gastrointestinal tract microbiota. Taken together this data provides a
mechanism of action for glycan structures.
8.4.2 Prebiotics for Infant Gut Bifidobacteria
Breastfed infants that are colonized by protective strains of bifidobacteria benefit

from the microbial activities within their developing intestine [36], which supports
a valuable function in vivo for the emergence and persistence of glycans. Henry
Tissier observed by microscopic analysis and culture techniques that the feces of
breastfed infants were unique in containing a bacterial isolate he termed “Bacillus
bifidus communis” [37]. For the 100 years since that initial identification, method-
ological techniques have wrestled to accurately type much less understand the spe-
cific bacteria within breastfed infants largely due to technical problems [38–40].
The challenges have become understandable in retrospect. Initial, culture-based
studies failed to isolate significant proportions of bifidobacteria from infants, but
these culture techniques failed to appreciate the oxygen sensitivity of infant bifido-
bacteria and were omitted. The major breakthroughs in DNA-based culture-
independent methods should have identified bifidobacteria, yet unfortunately the
16s rDNA primers that are the basis of detection in these methods were not designed
to effectively amplify bifidobacteria. Finally, both 16s rDNA surveys and metage-
nomic techniques that ostensibly sequence all DNA and again should have unequiv-
ocably identified bifidobacteria failed to appreciate the physical integrity of the
double cell wall of bifidobacteria and the need to selectively handle the disruption
of these barriers to DNA release for sequencing. These technical difficulties are now
being resolved, and accurate measures of infant fecal microbiota are now available.
With these techniques in place, studies are demonstrating very high proportions of
specific strains of bifidobacteria in breastfed infants prior to transition to an adult
microbiota [41]. The analyses of the breastfed intestinal track have revealed
Bifidobacterium longum and B. breve with B. bifidum and B. pseudocatenulatum
and B. catenulatum also present [42].
The basic concept that milk itself was influencing the microbial population was
proposed by Gorgy and colleagues [43] on the basis of observations that B. bifidum
(then termed Lactobacillus bifidus) grew on human milk fractions. The concept
however implied that there was a single component responsible, the so-called
Bifidus factor. Various studies since have demonstrated that human milk does indeed
contain indigestible matter that since humans cannot break them down into digest-
ible monomers would invariably reach the intestine [44–46]. The selectivity of
growth promotion by bifidobacterial species growing on human milk oligosaccha-
rides was first demonstrated in vitro by Ward et al. [47, 48]. A series of detailed
studies have extended this initial observation demonstrating that only certain bifido-
bacterial species consume the majority of the stereospecific oligosaccharides of
human milk [1, 25, 42]. Within specific strains, growth on oligosaccharides differed
leading to the conclusion that B. infantis and select B. breve preferentially consume
fucosylated and sialylated HMOs. These results indicate that bifidobacterial strains
that grow well on specific glycan structures possess genetic adaptations for select
growth on human milk in the infant intestine [32, 49].
The interaction between human milk oligosaccharides and bifidobacteria

provides a unique opportunity to map the continuum of metabolites from a food,
through the genetics of their disassembly by a “consumer” through the metabolic
pathways that utilize them for growth. The genes in bifidobacteria that specifically
bind and catabolize HMOs for energy have been identified, expressed, and verified
for enzymatic activity [25, 28, 50–53]. The process of annotating these genes has
demonstrated that different bifidobacterial species grow on HMO by distinct cata-
lytic mechanisms. B. infantis possess a 43-kb gene cluster (termed HMO cluster I),
encode for glycosyl hydrolases, and transport systems using a unique and highly
efficient pathway to internalize and metabolize milk oligosaccharides [54, 55]. In
contrast, B. bifidum is equipped with genes encoding a different set of catalytic
activities toward HMO consumption. This strain exports fucosidases and a lacto-N-
biosidase to hydrolyze lacto-N-biose from HMO structures which is in turn trans-
ported into the bacterium and metabolized [56].
The process of annotating the detailed mechanisms of the metabolism of human
milk oligosaccharides by bifidobacteria has revealed consequences of that metabo-
lism that were unanticipated. This group of metabolites causes a fundamental shift
in the phenotype of the bacterium itself. Milk oligosaccharides trigger a specific
HMO phenotype to B. infantis. In effect the bacterium shifts to a phenotypic state
that is linked to its competitive success in establishing itself within the microbial
ecosystem. The phenotype is also associated with interactions between the bacte-
rium and the infant host. Chichlowski et al. [57] reported that the HMO-specific
phenotype of B. infantis ATCC15697 on HMOs increases binding to intestinal epi-
thelial cells in vitro. These studies suggest that the specific phenotype of bifidobac-
terial populations grown on human milk oligosaccharides as metabolites provides
mechanisms to the organism supporting greater growth, microbiota persistence
interactions with the host epithelium. This model of a metabolically distinct bacte-
rial population induced by its “food” source is supported by in vivo administration
of B. infantis to premature infants fed either formula or breast milk. The human
breast milk-fed infants, when supplemented with B. infantis, had increases in fecal
bifidobacteria and decreases in γ-Proteobacteria compared with the formula-fed
group [58]. The ability of these specific bacteria to deconstruct HMOs that is
encoded in their genome suggests the co-evolution of human lactation and specific
commensal organisms. Thus, mothers are shaping the protective milk-oriented
microbiota (MOM) of their infants through breast milk (Fig. 8.8) [53, 59]. This is
one example of how milk glycans are being annotated.
8.5 A Vision for Metabolomics in the Future
The science of nutrition is faced with a daunting challenge: improving human

health. The enabling principles of reductionist chemistry that were so effective in
identifying essential nutrients are failing to address the more complex problems of
non-communicative but diet-dependent diseases that are epidemic around the world.
8 Metabolomics and Milk: The Development of the Microbiota in Breastfed Infants
Fig. 8.8 Fecal analyses of a single infant at several time points through the first 3 months of life. Above are shown the oligosaccharides from human milk in
the infant feces, normalized to week 1 in which all oligosaccharides from milk appeared unmodified in the feces. Successive weeks later time points illustrate
163
the oligosaccharide abundances in the feces relative to week 1. Below is shown the bacteria profile of the same fecal samples through the first 3 months of life.
(modified from de Leoz [60] unpublished)
Single molecules delivered to everyone in the population will not solve these
problems. Nutrition as a field must lead the world into more integrative, biology-
driven strategies that are not only quantitatively precise and mechanistically com-
plete but mapped to actual foods and deployable as individual solutions. The first
proofs of principle of such strategies are emerging from the, admittedly more nar-
rowly defined, nourishment of breastfed infants. The toolsets of systems biology
including genomics, metabolomics, proteomics, and glycomics have shown their
power to interrogate the complexity of milk and reveal how it accomplishes an
astonishingly successful biological feat, the colonization and development of the
infant microbiota. Evolution clearly identified this to be an important target for
mammalian health. Human mothers are nourishing the bacteria within their infants
almost as enthusiastically as their infants. Yet, the strategy of nourishing the infant
microbiota is a lesson for all of nutrition research. Rather than a single, simple mol-
ecule, the mammary gland produces an entire metabolome that includes: a spectrum
of complex oligosaccharides and glycans that evade digestion by the infant and
continue through to the infant’s lower intestine. The complexity of glycans provides
an intense selectivity that rewards only those bacteria genetically capable digesting
the glycans and accessing their sugars. The combination of glycan complexity as
available substrate and genetic capability as enzymatic specificity is a model for
nutrition’s microbiota research going forward. The knowledge assembled to date
has begun the process of mapping the detailed mechanistic understanding of the
functions of different microbial ecosystems in the infant. Key questions remain:
How does a particular microbiota protect infants from pathogens and what are its
weaknesses? How does a particular microbiota educate immunity in the face of the
bewildering array of both pathogenic insults and completely benign passersby, and
what are the causes of its failures? How does the microbiota prevent the massive
activation of immunity and the anticipated increase in inflammation that would be
expected from dropping a naïve, ostensibly sterile infant in the “real world,” and can
we apply these same principles to adults? How does a particular influence whole
body metabolism and ensure appropriate food intake and suitable direction of fuels
to peripheral tissues, and could these same mechanisms take visceral fat out of
adults and put back “baby fat”? The successes of the first generation of metabolo-
mics research tools applied to understand the interactions between mammary-
produced oligosaccharides, and the infant microbiota are a glimpse of what this new
field of biology can achieve.
Acknowledgements We acknowledge all of the researchers in the UC Davis Foods for Health
Institute and the Milk Bioactives Program for their enthusiasm, imagination, and collective contri-
bution to this subject matter. The work by the Milk Bioactives Program has been supported by the
UC Davis Research Investments in the Sciences and Engineering Program; the UC Discovery
Grant Program; the California Dairy Research Foundation; the Dairy Research Institute; the Bill &
Melinda Gates Foundation; and the National Institutes of Health awards R01HD059127,
R01HD065122, R01HD061923, R21AT006180, and R01AT007079. D.A.M. acknowledges
support as the Peter J. Shields Endowed Chair in Dairy Food Science.
References
1. Smilowitz JT, O’Sullivan A, Barile D, German JB, Lönnerdal B, Slupsky CM. The human milk
metabolome reveals diverse oligosaccharide profiles. J Nutr. 2013;143(11):1709–18.
2. Molinari CE, Casadio YS, Hartmann BT, Arthur PG, Hartmann PE. Longitudinal analysis of
protein glycosylation and β-casein phosphorylation in term and preterm human milk during
the first 2 months of lactation. Br J Nutr. 2012;1(1):1–11.
3. Froehlich J, Dodds E, Barboza M, McJimpsey E, Seipert R, Francis J, et al. Glycoprotein
expression in human milk during lactation. J Agric Food Chem. 2010;58(10):6440–8.
4. Uchiyama S-i, Sekiguchi K, Akaishi M, Anan A, Maeda T, Izumi T. Characterization and
chronological changes of preterm human milk gangliosides. Nutrition. 2011;27(10):
998–1001.
5. Liao Y, Alvarado R, Phinney B, Lönnerdal B. Proteomic characterization of human milk whey
proteins during a twelve-month lactation period. J Proteome Res. 2011;10(4):1746–54.
6. Ruvoën-Clouet N, Le Pendu J, Mas E, Marionneau S, Guillon P, Lombardo D. Bile-salt-
stimulated lipase and mucins from milk of ‘secretor’ mothers inhibit the binding of Norwalk
virus capsids to their carbohydrate ligands. Biochem J. 2006;393:627–34.
7. Varki A. Sialic acids in human health and disease. Trends Mol Med. 2008;14(8):351–60. Epub
2008/07/09.
8. Aoki-Kinoshita KF. An introduction to bioinformatics for glycomics research. PLoS Comput
Biol. 2008;4(5):e1000075.
9. Reinhold V, Zhang H, Hanneman A, Ashline D. Toward a platform for comprehensive glycan
sequencing. Mol Cell Proteomics. 2013;12(4):866–73.
10. Ninonuevo MR, Perkins PD, Francis J, Lamotte LM, LoCascio RG, Freeman SL, et al. Daily
variations in oligosaccharides of human milk determined by microfluidic chips and mass spec-
trometry. J Agric Food Chem. 2008;56(2):618–26. Epub 2007/12/20.
11. De Leoz ML, Wu S, Strum JS, Ninonuevo MR, Gaerlan SC, Mirmiran M, et al. A quantitative
and comprehensive method to analyze human milk oligosaccharide structures in the urine and
feces of infants. Anal Bioanal Chem. 2013;405(12):4089–105. Epub 2013/03/08.
12. Lebrilla CB. Is high throughput glycomics possible? Mass Spectrom (Tokyo). 2013;2(Spec
Iss):S0016. doi:10.5702/massspectrometry.S0016. Epub 2013 Apr 15.
13. Sela DA, Garrido D, Lerno L, Wu S, Tan K, Eom HJ, et al. Bifidobacterium longum subsp.
Infantis ATCC 15697 α-fucosidases are active on fucosylated human milk oligosaccharides.
Appl Environ Microbiol. 2012;78(3):795–803.
14. Wu S, Tao N, German JB, Grimm R, Lebrilla CB. Development of an annotated library of
neutral human milk oligosaccharides. J Proteome Res. 2010;9(8):4138–51.
15. Wu SA, Grimm R, German JB, Lebrilla CB. Annotation and structural analysis of sialylated
human milk oligosaccharides. J Proteome Res. 2011;10(2):856–68.
16. Ninonuevo M, An H, Yin H, Killeen K, Grimm R, Ward R, German B, Lebrilla
C. Electrophoresis. 2005;26:3641–9.
17. Leo F, Asakuma S, Fukuda K, Senda A, Urashima T. Determination of sialyl and neutral oli-
gosaccharide levels in transition and mature milks of Samoan women, using anthranilic
derivatization followed by reverse phase high performance liquid chromatography. Biosci
Biotechnol Biochem. 2010;74(2):298–303.
18. Ninonuevo MR, Ward RE, LoCascio RG, German JB, Freeman SL, Barboza M, et al. Methods
for the quantitation of human milk oligosaccharides in bacterial fermentation by mass spec-
trometry. Anal Biochem. 2007;361(1):15–23.
19. Messer E, Trifonoff W, Stern JG, Collins JH. Bradbury structure of a marsupial-mild trisac-
charide. Carbohydr Res. 1980;83:327–34.
20. Tao N, Wu S, Kim J, An HJ, Hinde K, Power ML, et al. Evolutionary glycomics: characteriza-
tion of milk oligosaccharides in primates. J Proteome Res. 2011;10(4):1548–57. Epub
2011/01/11.
21. Coppa GV, Gabrielli O, Pierani P, Catassi C, Carlucci A, Giorgi PL. Changes in carbohydrate
composition in human milk over 4 months of lactation. Pediatrics. 1993;91(3):637–41.
22. Gabrielli O, Zampini L, Galeazzi T, Padella L, Santoro L, Peila C, et al. Preterm milk oligosac-
charides during the first month of lactation. Pediatrics. 2011;128(6):e1520–31.
23. Kunz C, Rudloff S, Baier W, Klein N, Strobel S. Oligosaccharides in human milk: structural,
functional, and metabolic aspects. Annu Rev Nutr. 2000;20:699–722.
24. Ninonuevo MR, Park Y, Yin H, Zhang J, Ward RE, Clowers BH, et al. A strategy for annotating
the human milk glycome. J Agric Food Chem. 2006;54(20):7471–80. Epub 2006/09/28.
25. LoCascio RG, Ninonuevo MR, Freeman SL, Sela DA, Grimm R, Lebrilla CB, et al.
Glycoprofiling of bifidobacterial consumption of human milk oligosaccharides demonstrates
strain specific, preferential consumption of small chain glycans secreted in early human lacta-
tion. J Agric Food Chem. 2007;55(22):8914–9. Epub 2007/10/06.
26. Sela DA, Mills DA. Nursing our microbiota: molecular linkages between bifidobacteria and
milk oligosaccharides. Trends Microbiol. 2010;18(7):298–307. Epub 2010/04/23.
27. Roger LC, Costabile A, Holland DT, Hoyles L, McCartney AL. Examination of faecal
Bifidobacterium populations in breast-and formula-fed infants during the first 18 months of
life. Microbiology. 2010;156(11):3329–41.
28. Marcobal A, Barboza M, Sonnenburg ED, Pudlo N, Martens EC, Desai P, et al. Bacteroides in
the infant gut consume milk oligosaccharides via mucus-utilization pathways. Cell Host
Microbe. 2011;10(5):507–14. Epub 2011/11/01.
29. Locascio RG, Ninonuevo MR, Kronewitter SR, Freeman SL, German JB, Lebrilla CB, et al. A
versatile and scalable strategy for glycoprofiling bifidobacterial consumption of human milk
oligosaccharides. Microb Biotechnol. 2009;2(3):333–42. Epub 2009/05/01.
30. Wada J, et al. Bifidobacterium bifidum lacto-N-biosidase, a critical enzyme for the degradation
of human milk oligosaccharides with a type 1 structure. Appl Environ Microbiol. 2008;74(13):
3996–4004.
31. Suzuki R, et al. Structural and thermodynamic analyses of solute-binding protein from
Bifidobacterium longum specific for core 1 disaccharide and lacto-N-biose I. J Biol Chem.
2008;283(19):13165–73.
32. Sela D, Chapman J, Adeuya A, Kim J, Chen F, Whitehead T, et al. The genome sequence of
Bifidobacterium longum subsp. infantis reveals adaptations for milk utilization within the
infant microbiome. Proc Natl Acad Sci. 2008;105(48):18964.
33. Lee JH, et al. Comparative genomic analysis of the gut bacterium Bifidobacterium longum
reveals loci susceptible to deletion during pure culture growth. BMC Genomics. 2008;9:247.
34. Mattarelli P, Bonaparte C, Pot B, Biavati B. Proposal to reclassify the three biotypes of
Bifidobacterium longum as three subspecies: Bifidobacterium longum subsp. longum subsp.
nov., Bifidobacterium longum subsp. infantis comb. nov. and Bifidobacterium longum subsp.
suis comb. nov. Int J Syst Evol Microbiol. 2008;58(Pt 4):767–72.
35. Falk P, Roth KA, Boren T, Westblom TU, Gordon JI, Normark S. Proc Natl Acad Sci U S A.
1993;90:2035.
36. Du X-L, Edelstein D, Rossetti L, Fantus IG, Goldberg H, Ziyadeh F, et al. Hyperglycemia-
induced mitochondrial superoxide overproduction activates the hexosamine pathway and
induces plasminogen activator inhibitor-1 expression by increasing Sp1 glycosylation. Proc
Natl Acad Sci. 2000;97(22):12222–6.
37. Tissier H. Recherches sur la flore intestinale des nourrissons: état normal et pathologique:
G. Carré et C. Naud; 1900.
38. Sakata S, Tonooka T, Ishizeki S, Takada M, Sakamoto M, Fukuyama M, et al. Culture‐inde-
pendent analysis of fecal microbiota in infants, with special reference to Bifidobacterium spe-
cies. FEMS Microbiol Lett. 2005;243(2):417–23.
39. Roger LC, McCartney AL. Longitudinal investigation of the faecal microbiota of healthy full-
term infants using fluorescence in situ hybridization and denaturing gradient gel electrophore-
sis. Microbiology. 2010;156(11):3317–28.
40. Harmsen HJ, Wildeboer-Veloo AC, Raangs GC, Wagendorp AA, Klijn N, Bindels JG, et al.
Analysis of intestinal flora development in breast-fed and formula-fed infants by using molecular
identification and detection methods. J Pediatr Gastroenterol Nutr. 2000;30(1):61–7.
41. Yatsunenko T, Rey FE, Manary MJ, Trehan I, Dominguez-Bello MG, Contreras M, et al.
Human gut microbiome viewed across age and geography. Nature. 2012;486:222–7.
42. Turroni F, Bottacini F, Foroni E, Mulder I, Kim JH, Zomer A, et al. Genome analysis of
Bifidobacterium bifidum PRL2010 reveals metabolic pathways for host-derived glycan forag-
ing. Proc Natl Acad Sci U S A. 2010;107(45):19514–9. Epub 2010/10/27.
43. György P, Norris RF, Rose CS. Bifidus factor. I. A variant of Lactobacillus bifidus requiring a
special growth factor. Arch Biochem Biophys. 1954;48(1):193–201.
44. Coppa G, Pierani P, Zampini L, Bruni S, Carloni I, Gabrielli O. Characterization of oligosac-
charides in milk and feces of breast-fed infants by high-performance anion-exchange chroma-
tography. Bioactive components of human milk. Springer; 2001. p. 307–14. ISBN
978-1-4613-5521-2.
45. Chaturvedi P, Warren CD, Buescher CR, Pickering LK, Newburg DS. Survival of human milk
oligosaccharides in the intestine of infants. Bioactive components of human milk. Springer;
2001. p. 315–23. ISBN 978-1-4613-5521-2.
46. Engfer MB, Stahl B, Finke B, Sawatzki G, Daniel H. Human milk oligosaccharides are resis-
tant to enzymatic hydrolysis in the upper gastrointestinal tract. Am J Clin Nutr.
2000;71(6):1589–96.
47. Ward RE, Ninonuevo M, Mills DA, Lebrilla CB, German JB. In vitro fermentation of breast
milk oligosaccharides by Bifidobacterium infantis and Lactobacillus gasseri. Appl Environ
Microbiol. 2006;72(6):4497–9.
48. Ward RE, Ninonuevo M, Mills DA, Lebrilla CB, German JB. In vitro fermentability of human milk
oligosaccharides by several strains of bifidobacteria. Mol Nutr Food Res. 2007;51(11):1398–405.
49. Ruiz-Moyano S, Totten SM, Garrido DA, Smilowitz JT, German JB, Lebrilla CB, et al.
Variation in consumption of human milk oligosaccharides by infant gut-associated strains of
bifidobacterium breve. Appl Environ Microbiol. 2013;79(19):6040–9. Epub 2013/07/31.
50. Garrido D, Nwosu C, Ruiz-Moyano S, Aldredge D, German JB, Lebrilla CB, et al. Endo-beta-
N-acetylglucosaminidases from infant gut-associated bifidobacteria release complex N-glycans
from human milk glycoproteins. Mol Cell Proteomics. 2012;11(9):775–85. Epub 2012/06/30.
51. Garrido D, Ruiz-Moyano S, Jimenez-Espinoza R, Eom HJ, Block DE, Mills DA. Utilization
of galactooligosaccharides by Bifidobacterium longum subsp. infantis isolates. Food
Microbiol. 2013;33(2):262–70. Epub 2012/12/04.
52. Garrido D, Ruiz-Moyano S, Mills DA. Release and utilization of N-acetyl-D-glucosamine
from human milk oligosaccharides by Bifidobacterium longum subsp. infantis. Anaerobe.
2012;18(4):430–5. Epub 2012/05/15.
53. Zivkovic AM, Lewis ZT, German JB, Mills DA. Establishment of a milk-oriented microbiota
(MOM) in early life: how babies meet their MOMs. Funct Food Rev. 2013;5:3–12.
54. Garrido D, Dallas DC, Mills DA. Consumption of human milk glycoconjugates by infant-
associated bifidobacteria: mechanisms and implications. Microbiology. 2013;159(Pt 4):649–
64. Epub 2013/03/06.
55. Garrido D, Kim JH, German JB, Raybould HE, Mills DA. Oligosaccharide binding proteins
from Bifidobacterium longum subsp. infantis reveal a preference for host glycans. PLoS One.
2011;6(3):e17315.
56. Kitaoka M. Bifidobacterial enzymes involved in the metabolism of human milk oligosaccha-
rides. Adv Nutr: Int Rev J. 2012;3(3):422S–9S.
57. Chichlowski M, De Lartigue G, German JB, Raybould HE, Mills DA. Bifidobacteria isolated
from infants and cultured on human milk oligosaccharides affect intestinal epithelial function.
J Pediatr Gastroenterol Nutr. 2012;55(3):321–7. Epub 2012/03/03.
58. Underwood MA, Kalanetra KM, Bokulich NA, Lewis ZT, Mirmiran M, Tancredi DJ, et al. A
comparison of two probiotic strains of bifidobacteria in premature infants. J Pediatr.
2013;163(6):1585–91.e9. doi:10.1016/j.jpeds.2013.07.017. Epub 2013 Aug 29.
59. Bode L, Jantscher-Krenn E. Structure-function relationships of human milk oligosaccharides.
Adv Nutr: Int Rev J. 2012;3(3):383S–91.
60. De Leoz ML, Kalanetra K, Bokulich N, Strum J, Underwood M, German J, Mills D, Lebrilla
C. Determination of human milk glycomics and gut microbial genomics in infant feces shows
correlation between lactation and development of gut microbiota. (2014 unpublished)
Chapter 9
Metabonomics and Gut Microbial Paradigm
in Healthy Aging
Elena Biagi, Marco Candela, François-Pierre Martin, Sebastiano Collino,

Claudio Franceschi, and Patrizia Brigidi
Abstract Metabonomics, aiming to characterize the pattern of low molecular

weight metabolites participating in metabolic pathways, is a valuable tool to
detect changes in metabolic regulation and subsequently link them to the health
outcome, emerging as a powerful technology to capture the complexity of human
aging. The growing interest of the research community in this field is demon-
strated by a large number of recently published researches, here summarized, that
aim to associate complex metabolic regulations with age-related biological pro-
cesses. The involvement of the gut microbiota, our “forgotten organ” with its
impressive metabolic capability, is a natural extension of this interest, even if still
largely unexplored. In this chapter, we aim to summarize the potential of metabo-
nomics in exploring the impact of the combined metabolism of human host and
gut microbiota on aging, as well as on the health outcome of age-related processes
and the probability to attain longevity.
Keywords Aging • Intestinal microbiota • Probiotics • Longevity • Metabonomics

• Gut–brain axis • Nutrition aging • Biomarker • Short chain fatty acids
E. Biagi • M. Candela • P. Brigidi (*)

Department of Pharmacy and Biotechnology, University of Bologna,
Via Belmeloro 6, 40126 Bologna, Italy
F.-P. Martin • S. Collino
C. Franceschi
Department of Experimental Diagnostic and Specialty Medicine and CIG – Interdipartimental
Center L. Galvani, University of Bologna, Bologna, Italy

DOI 10.1007/978-1-4471-6539-2_9
170 E. Biagi et al.
9.1 Metabonomics in Healthy Aging and Brain

Aging Research
The increase in the aging population and incidences of chronic diseases raises new
challenges for global public healthcare in which preventive medicine approaches,
particularly by the means of optimal nutrition, will be crucial. Aging can commonly
be characterized as a progressive, generalized impairment of biological functions
resulting in an increased vulnerability to environmental challenge and a higher risk
of disease and death [1]. Cross talk can occur between multiple physiological sys-
tems, i.e., cognitive, metabolic, and gut systems, which influence the immune
system [2]. Environmental factors such as lifestyle choices as well as genetic factors
all contribute to healthy aging. Among these factors, the environment is the most
easily modifiable. Understanding the physiology of aging is of tremendous impor-
tance to allow populations to grow old disease-free and with a good quality of life.
In this respect, it is important to understand the natural aging process and to
elucidate where lifestyle and/or dietary interventions can have an impact. Despite
the enormous complexity of the aging process, a small number of basic molecular
mechanisms underpin the aging process, including a set of highly conserved mecha-
nisms. One of the key mechanisms is inflammation as a typical feature of the aging
process is the development of a chronic, low-grade inflammatory status named
“inflamm-aging” [3–5], and this condition has emerged as critical in the pathogen-
esis of major age-related chronic diseases such as atherosclerosis, type 2 diabetes,
and neurodegeneration. Inflamm-aging plays a pivotal role in the most important
geriatric conditions, such as sarcopenia, frailty, and disability, thus contributing to
elderly mortality [5]. Interestingly, a variety of tissues (adipose tissue, muscle),
organs (brain, liver), systems (immune system), and ecosystems (gut microbiota) of
the body can contribute to the onset and progression of such a systemic inflamma-
tory state [6], by increasing the production of a number of proinflammatory mediators
or lowering that of the anti-inflammatory ones, thus tilting the equilibrium toward
inflammation [7].
Physiological, genetic, and environmental changes can cause modifications in
existing homeostatic conditions, which are ultimately reflected in the metabolic
composition of the different biological compartments. Metabonomics attempts
therefore to quantitatively profile small molecules endogenously and exogenously
present in a complex biological system, relying on the analysis of fluids (blood,
urine), tissue biopsies, or stools. Metabonomics is therefore a very powerful tool
for capturing the complexity of the aging process. Mass spectrometry (MS) and
nuclear magnetic resonance (1H-NMR) spectroscopy are the most commonly used
analytical tools to obtain profiles. Imaging techniques and flux analysis using sta-
ble isotopes are parallel technologies to obtain metabolite information. Multivariate
statistical and bioinformatics techniques are ultimately used for data mining the
complex metabolic profiles which encapsulate information on genetics, environ-
mental factors, gut microbiota activity, and lifestyle and food habits. This com-
bined strategy sustains the complex process of identifying emerging biomarkers
9 Metabonomics and Gut Microbial Paradigm in Healthy Aging 171
indicative of the individual response to specific physiological factors and/or

nutritional/physical interventions.
The age-related chronic inflammation is believed to be pathogenic, especially
with regard to its contribution to frailty and degenerative disorders. In particular,
the frailty syndrome is increasingly being considered as a key risk indicator of
adverse health outcomes [8]. Frailty is characterized by a decline in the functional
reserve with several alterations in diverse physiological systems, including altered
energy metabolism, skeletal muscle mass and strength (also termed sarcopenia),
and hormonal and inflammatory functions [8]. A disturbance of the balance between
the synthesis and breakdown of muscle proteins can lead to the loss of muscle mass.
In addition, elderly may be also prone to be resistant to anabolic stimuli which is
likely a key factor in the loss of skeletal muscle mass with aging. Such a perturba-
tion of muscle metabolism with aging has been proposed to play a role in the devel-
opment of sarcopenia [9], which ultimately alters walking performance and physical
endurance and results in a perception of exhaustion and fatigue. It has been there-
fore foreseen that well-adapted exercise training and nutritional management pro-
grams should be an effective means of counteracting muscle weakness and physical
frailty in elderly [10]. In particular, combining resistance exercise with essential
amino acid supplementation restores the muscle protein anabolic response in older
men [8]. Future research is needed to determine whether these novel interventions
will be successful in preventing sarcopenia and improving muscle strength and
function in older adults, and in this system, biology approaches will help not only
personalize the management program but may also help understand specific require-
ments for nonresponders. For instance, metabolic profiling was successfully applied
to blood profiling in the frame of a human exercise study focused on the effects of
beverages containing glucose, galactose, or fructose taken after exercise and
throughout a recovery period [11]. Others have shown novel perspectives in moni-
toring the metabolic significance of long-term strength and endurance training [12].
These applications showed how we could further our understanding on individual
predisposition and inferences on nutritional/exercise response by monitoring energy
(glucose, lipid, amino acid) and oxidative stress metabolism [12].
Very recently, a series of metabonomics studies were expanded to report for the
first time the metabolic phenotype of longevity. Plasma metabolic phenotype of
three long-lived murine models was characterized by NMR metabonomics. Here a
panel of metabolic differences were generated for each model, 30 % dietary
restricted, insulin receptor substrate 1 null (Irs1−/−), and Ames dwarf (Prop1df/df),
relative to their controls, and, subsequently, the three long-lived models were
compared to one another. Concentrations of mobile very low-density lipoproteins,
trimethylamine, and choline were significantly decreased in the plasma of all three
models. Such comparative approach suggests that the metabolic networks underly-
ing lifespan extension are not exactly the same for each model of longevity and are
consistent with multifactorial control of the aging process [13].
As centenarians well represent the model of successful and healthy aging [14],
there are many important implications in revealing the underlying molecular mech-
anisms behind such acquired longevity. Centenarians avoid or delayed the major
172 E. Biagi et al.
inflammation-driven age-related diseases, such as cardiovascular diseases, diabetes

mellitus, Alzheimer disease (AD), and cancer [4, 7]. A NMR-/MS-based metabo-
nomics and targeted lipidomics approaches were applied in a representative aging
cohort of centenarians (mean age 100.9 years), elderly (mean age 70 years, includ-
ing offspring of centenarians and their age-matched controls), and young indi-
viduals (mean age 31 years) to reveal the molecular footprints of aging and longevity.
With increasing age, targeted MS profiling of blood serum displayed a marked
decrease in tryptophan concentration, while unique alterations in specific glycero-
phospholipids and sphingolipids were seen in the longevity phenotype. Untargeted
metabonomics profiling of urine revealed that the longevity process is marked by
changes in gut microbial metabolites, as displayed by increase urinary excretions of
phenylacetylglutamine, p-cresol sulfate, and 2-hydroxybenzoate. Moreover, cente-
narian offsprings, who are reported to have delay in age-related diseases, have a
distinct serum metabolic phenotype from siblings of non-long-living parents, with
changes in amino acids (serine, phenylalanine) and lysophosphatidylcholines.
Overall, these results describe different biological processes associated to aging and
longevity, including differential lipid mediator networks and increased gut micro-
biota dysbiosis [15]. Additionally, an investigation on specific lipids associated with
familial longevity in females was explored by Gonzalez-Covarrubias et al. in the
plasma lipidome by measuring 128 lipid species in 1,526 offspring of nonagenari-
ans (59 years ± 6.6) and 675 (59 years ± 7.4) controls from the Leiden Longevity
Study [16]. Here in women 19 lipid species associated with familial longevity with
ether phosphocholine and sphingomyelin species are identified as candidate longev-
ity markers. While this population reflects a different cohort with plausible differ-
ences in lifestyle and dietary habits, common to the previous study, the authors
postulated that lipid signatures in plasma lipidome of female individuals could sug-
gest a better antioxidant capacity and lower lipid peroxidation capabilities with
probable effects on the longevity process. The contrast between offspring of nona-
genarian siblings that express the propensity for longevity and their partners was
also found to be reflected by cellular characteristics in vitro. In dermal fibroblast
strains from offspring, oxidative stress induced less reactive oxygen species, less
senescence, more apoptosis, and slower growth speed when compared to strains
from the partners of the offspring [17]. Aiming to find differences in cellular metab-
olism in vitro between these fibroblast strains, cell culture supernatants collected at
24 h and 5 days were analyzed using 1H nuclear magnetic resonance [18]. Between
24 h and 5 days of incubation, supernatants of all fibroblast strains showed decreased
levels of glucose, pyruvate, alanine–glutamine, valine, leucine, isoleucine, serine,
and lysine and increased levels of glutamine, alanine, lactate, and pyroglutamic
acid. The alanine–glutamine and glucose consumption were higher for fibroblast
strains derived from offspring when compared to strains of their partners. Also, the
production of glutamine, alanine, lactate, and pyroglutamate was found to be higher
for fibroblast strains derived from offspring.
Aging is characterized by a common development of physical and mental disor-
ders that implies metabolic decline with loss of hepatic, renal, coronary, or cardiac
function, with increased risk of suffering cancer, inflammatory, cardiovascular, and
neurodegenerative diseases. Among these factors, AD is the most common dementia

among all the clinically recognized dementias in the aging population. Although the
cause is not known, there are profound biochemical alterations in multiple pathways
in the AD brain including changes in amyloid beta-protein metabolism, tau phos-
phorylation, membrane lipid deregulation, and synaptic neurotransmission [19].
Lipids are key regulators of brain function and have been increasingly implicated in
neurodegenerative disorders including AD. Han et al. analyzed plasma from 26 AD
patients and 26 cognitively normal controls in a nontargeted approach using multidi-
mensional mass spectrometry-based shotgun lipidomics [20]. Here a significant dis-
ruption of plasma sphingolipidome was detected in AD, with long aliphatic chains
sphingomyelin species, 22 and 24 carbon atoms, significantly lower in AD, com-
pared to controls, and ceramide species (N16:0 and N21:0) significantly higher in
AD. Sato et al. [21] applied lipidomics to focus on steroid-related compounds to
identify novel AD plasma biomarkers with significant correlation with mini-mental
state examination scores. A systems-based approach was also employed to determine
the lipidome changes in brain tissues affected by AD. Using liquid chromatography–
mass spectrometry, the profile extracts from the prefrontal cortex, entorhinal cortex,
and cerebellum of late-onset AD were generated by Chan et al. [22]. Here 26 different
lipid subclasses including lysophospholipids, glycerophospholipids, sphingolipids,
glycerolipids, and sterols were studied. While the cerebellum lacked major altera-
tions in lipid composition, an elevation of a signaling pool of diacylglycerols as well
as sphingolipids in the prefrontal cortex of AD patients was found. Furthermore, the
diseased entorhinal cortex showed specific enrichment of lysobisphosphatidic acid,
sphingomyelin, the ganglioside GM3, and cholesterol esters, suggesting common
pathogenic mechanisms associated with endolysosomal storage disorders.
The development of systems biology approaches and the new generation of
biomarker patterns will provide the opportunity to associate complex metabolic
regulations with key aging biological processes. This will subsequently lead to the
development of systems mechanistic hypotheses that could be targeted with new
nutritional and therapeutic personalized concepts aimed at healthy aging.
9.2 The Gut Microbiota in Human Aging
The gastrointestinal tract (GIT) is one of the most essential interfaces of mam-
malian organism interacting with nutrients, exogenous compounds, and gut
microbiota, and its condition is influenced by the complex interplay between these
environmental factors and host genetic elements. Along the GIT, the gut micro-
biota is a key determinant of the gut functional ecology and metabolic homeosta-
sis, through fine interactions with regulatory processes involved in the absorption,
digestion, metabolism, and excretion of dietary nutrients as well as barrier integ-
rity, motility, and mucosal immunity [23, 24]. The GIT ecosystem hosts a diverse
and highly evolved microbial community composed of hundreds of different
microbial species [25, 26], which can be viewed as a metabolically adaptable,
174 E. Biagi et al.
a b
AGEING IMMUNOSENESCENCE &
DIETARY INFLAMM-AGING
• masticatory
CHANGES:
dysfunction
PRO-INFLAMMATORY LOOP
reduction in fruits
• reduction of appetite
and vegetables
• constipation
• reduction of motility PRO-INFLAMMATORY
• depression MICROBIOTA
LOW FIBER DIET
DISBIOSYS
Pathobionts
(Enterobacteriaceae)
Immunomodulatory
groups
AGED-TYPE GUT MICROBIOTA (e.g. Faecalibacterium
prausnitzii)
LOW BIODIVERSITY
REDUCTION OF
DECREASE OF
FIBER-DEGRADING
SCFAs IN THE
SCFAs PRODUCERS GUT
(Clostridium cl. XIVa and IV)
Fig. 9.1 Aging and gut microbiota. (a) Dietary habits usually change during the aging process,
because of several age-related physiological and environmental factors; the resulting low-fiber diet
is among the causes of reduction of gut microbiota diversity and changes in its phylogenetic com-
position and, consequently, reduction of some gut microbiota functions, i.e., production of SCFAs.
(b) Immunosenescence and inflamm-aging nurture the overgrowth of pathobionts in the gut eco-
system to the detriment of immunomodulatory bacterial groups; the resulting gut microbiota dys-
biosis promotes inflammation itself in a sort of self-sustaining loop
rapidly renewable, and metabolically flexible ecosystem varying in addition with

the host’s age, diet, and health status [27]. For instance, bacterial colonization of
the gut by commensal bacteria has been shown to alter intestinal physiology of the
host by modulation of genes and metabolic processes implicated in nutrient
absorption, mucosal defenses, and xenobiotic metabolism.
Age-related physiological changes in the GIT, as well as modification in life-
style, nutritional behavior, and functionality of the host immune system, inevitably
affect the gut microbiota (Fig. 9.1a).
The most informative studies about the age-related modifications of the gut
microbiota structure and composition have been conducted over the last 10 years,
since when 16S rRNA gene-based molecular characterization technologies have
been made available [28–30]. Briefly, studies are in general agreement in reporting
a large interindividual variability, as well as a reduced biodiversity, and compro-
mised stability of the intestinal microbiota in older subjects with respect to younger
individuals [31–35]. An age-related increase in facultative anaerobes, including
streptococci, staphylococci, enterococci, and enterobacteria has also been com-
monly reported in elderly [32, 33, 36–38].
A certain country specificity in how the aging process impacts on the intestinal
microbiota, possibly related to differences in lifestyle and dietary habits, has been
reported [36, 39], particularly in relation to the dominant components of the gut
microbiota, Firmicutes and Bacteroidetes. For what concerns Firmicutes, members

of the Clostridium cluster XIVa were found to decrease in Japanese, Finnish, and
Italian elderly and centenarians, whereas an inverse trend was found in German old
adults [32, 33, 36, 40]. The species Faecalibacterium prausnitzii (Clostridium clus-
ter IV), known for its anti-inflammatory properties and ability to produce short
chain fatty acids (SCFAs), markedly decreased in Italian elderly and centenarians
[32, 36], as well as in frail, hospitalized, and antibiotic- and anti-inflammatory-
treated elderly [41–44]. Conversely, an age-related increase in Bacteroidetes was
found in German, Austrian, Finnish, and Irish elderly [33, 34, 36, 44], but this was
not confirmed in Italian elderly and centenarians [32, 36]. In the case of Irish elderly,
Bacteroidetes were found to be the dominant phylum instead of Firmicutes, which
has always been regarded as the most abundant in healthy adults [34].
Similarly to what is observed in inflammatory bowel diseases and obesity [45–
47], the age-related changes in gut microbiota composition summarized above may
concur to the complex process that both sustains and nurtures inflamm-aging.
Indeed, several bacterial groups known to increase in the elderly (e.g., enterococci
and Enterobacteriaceae) are often classified as “pathobionts,” defined as bacteria
present in the healthy gut microbiota in low concentration, which are able to thrive
in inflamed conditions, promoting the inflammation itself [31].
The age-related changes in the phylogenetic architecture of the gut microbiota
impact on those metabolic and physiological functions for which the human meta-
organism depends on its microbial counterpart, with consequences at the functional
level and possibly affecting the risk of elderly people to develop some types of
diseases. For instance, the production of SCFAs, i.e., butyrate, acetate, and propio-
nate, is an essential feature of a healthy gut ecosystem, because SCFAs, especially
butyrate, have nutritive, anti-inflammatory, antineoplastic properties and exert a
protective role for the intestinal epithelium, increasing its resistance [48–50]. The
lower capacity to produce butyrate in the elderly gut microbiome [32, 51], corre-
lated with decreased amounts of F. prausnitzii, Eubacterium hallii, and bacteria
belonging to the Eubacterium rectale/Roseburia group, together with the increase
in proinflammatory pathobionts (i.e., bacteria present in the healthy gut microbiota
in low concentration, which are able to thrive in inflamed conditions, sustaining
inflammation itself), may nurture the inflamm-aging process in the GIT (Fig. 9.1b)
[31]. Moreover, the declined butyrate-producing capacity may contribute to the
development of degenerative diseases and anorexia [52, 53].
SCFAs are also a fundamental component of the microbiota–host bio-network to
maintain the GI epithelial integrity, by stimulating the release of mucins, the gel-
forming protein component of the colonic mucus layer that contributes to the physi-
cal separation between microbiota and enterocytes [54], and enhancing transepithelial
resistance [55]. The age-related microbiota depletion in SCFA producers may con-
cur in weakening gut epithelium, allowing the passage of whole bacterial cells and
their products, ending up in an abnormal inflammatory response [56].
Moreover, a proinflammatory dysbiosis, together with the decreased butyrate
production in the intestine, has also been linked to an increased risk of colorectal
cancer [57, 58], whose incidence is augmented in aged people [59].
176 E. Biagi et al.
Finally, changes in the gut microbiota composition may be rated as one of the
age-related physiological processes that, all together, are linked to “how healthily”
a human being will age. It has recently been demonstrated by Claesson et al. that the
healthiest elderly, living in community setting and having a high-quality diet, pos-
sess a distinct gut microbiota from less healthy elderly living in residential care
[60]. However, whether these changes are simply a biomarker of the aging process
or could be counted among its driver cannot yet be decided based on the available
data. Certainly, the aging gut microbiota possesses features that can actively affect
the health status of old people, i.e., contributing to the onset of pathological condi-
tions known to affect the elderly with a higher incidence than the young adults, such
as C. difficile-associated diarrhea and colon cancer, as well as their eventual respon-
siveness to therapies [30]. Considering the aging human being as a metaorganism,
composed by both human and bacterial cells that together answer to the changing
metabolic requests of the human body, could be the right approach for promoting
healthy aging and longevity.
9.3 Variation of Microbiota–Host Co-metabolism in Aging
Age-related variations in microbiota–host transgenomic metabolism can impact dif-

ferent ecological services provided by our symbiotic gut microbial community.
Indeed, the microbiota–host co-metabolism has been reported as strategic for sev-
eral host physiological functions – such as regulation of extraction and storage of
food energy, epigenome modulation, and brain function – and its impairment can be
detrimental for health and longevity (Table 9.1).
As an end product of the gut microbiota fermentation processes, gut butyrate
shows a multifactorial role in host nutrition, and its reduction as a consequence
of the age-related microbiota deterioration can provide several consequences for
the host nutritional state. For instance, microbial butyrate is essential to enhance
the efficiency of food energy extraction, representing the primary energy source
for colonic epithelium and providing the 15 % of the total energy request [46,
61]. On the other hand, supporting insulin secretion, microbial butyrate is also
involved in the regulation of the host energy storage. Finally, butyrate has been
reported to be a key regulator of appetite, enhancing the production of leptin and
inducing pyy expression [23]. Beside butyrate, also the age-related decrease of
microbial propionate and acetate can concur in weakening of the host nutritional
state. In particular, these metabolites have a role in the regulation of food energy
storage; acetate is involved in hepatic liponeogenesis, while propionate regulates
gluconeogenesis [61].
SCFAs, in particular butyrate, have also been recently shown to be involved in
the modulation of the host epigenome by means of the inhibition of the histone
deacetylase. Acting as regulators of the epithelial cell transcriptome, the decrease of
microbial SCFA along aging can influence different physiological aspects of the
host biology, such as digestion, immunity, gut–brain function, and hormone
Table 9.1 End products of microbial metabolism in the gut and their activities with respect to the
human host physiology
Metabolite Origin Functions/activities
Butyrate Carbohydrate Energy source; stimulates leptin
fermentation by gut production; regulates neutrophils;
microorganisms inhibits inflammatory cytokines;
increase TJ expression; histone
deacetylase inhibitor; binding to
G-protein-coupled receptors: improves
of insulin secretion; induction of Pyy
expression, supports fat deposition;
antimicrobial activity; modulation of
the sympathetic nervous system and
gut motility
Acetate Supports lipid synthesis; supports
energy metabolisms; binding to
G-protein-coupled receptors;
gut motility
Propionate Supports gluconeogenesis; binding to
G-protein-coupled receptors;
gut motility
Methylamine Metabolism of choline Altered product of choline
Dimethylamine by gut microorganisms metabolisms; has been associated with
Trimethylamine obesity, diabetes and cardiovascular
disease, liver steatosis
Trimethylamine-N-oxide
4-Cresyl sulfate Tyrosine putrefaction Association with cancer; diabetes;
4-Cresil glucuronide by gut microorganisms autism; depression
5-Hydroxytryptamine Bacterial metabolism of Influences brain and behavior
tryptophan
GABA Neurotransmitters of Influences behavior
Noradrenaline bacterial origin
Dopamine
Acetylcholine
LPS Gram – surface Metabolic endotoxemia
component
secretion [62]. Interestingly, recent findings demonstrated that neurotransmitters or

neuroactive metabolites of microbial origin, such as SCFA and 5-hydroxytryptamine,
are effective in the stimulation of the sympathetic nervous system and gut motility
[23]. The reduction of these neuroactive microbial metabolites along aging can
result in changing in brain function and behavior, supporting cognitive deficiencies
and depression which are typical of the frail elderly [63]. Finally, the impairment of
microbiota–host mutualism in the elderly can drive to modifications of the choline
178 E. Biagi et al.
metabolism by gut microorganism, resulting in an increase risk to develop diabetes

and cardiovascular diseases [64]. Analogously, an increase of tyrosine putrefaction
by gut microorganism in an aged-type microbiota can eventually result in the
increased risk of developing cancer and depression.
9.4 Metabonomics Potential in Aging and Gut

Microbiota Research
Increasing scientific evidence has been reported on the fundamental role of gut
microbiota in both positive and negative triggers of specific metabolic states of indi-
viduals and populations [61, 65]. Therefore, it is today critical to understand the
molecular foundations of the impact of the gut microbial activity on human health
and nutritional status [66–68].
The current omics revolution offers an unprecedented opportunity to explore how
our gut symbionts contribute to our physiology and human health. Future systems
biology approaches combining state-of-the-art microbial and metabolic modeling
and discovery approaches, including metagenomics and metabonomics, will help in
deciphering the molecular foundations of these transgenomics interactions. Recently,
an exhaustive gene catalogue has been published containing virtually all of the preva-
lent gut microbial genes in large human cohort, and reporting to which extent many
bacterial species are shared by different individuals [69]. Such an approach could be
used for global characterization of the genetic potential of ecologically complex
environments [69] but also to help understand how gut microbiota specificities could
be exploited to develop new therapeutic and nutritional strategies. Systems biology
approaches, including metabonomics, have emerged over the last two decades as a
novel way forward to provide insights into the role of mammalian gut microbial
metabolic interactions in individual susceptibility to health and disease outcomes.
Since it is more likely that gut microbiome required for proper functioning of the gut
ecosystem in the elderly is different from that of the young subjects, knowledge and
systems models will help in assessing earlier deviations that lead to unhealthy aging
and development of chronic inflammatory conditions. In particular, both system-
wide (i.e., whole organism) and organ-specific metabolism may have components
driven by gut microbial activities [70, 71], which suggests that the dynamics of the
gut microbiome could help to maintain or reestablish host metabolic homeostasis in
disease and early onsets of metabolic deregulations. Gut microbial activities can be
extremely specific, as for the development and maintenance of the mucosal innate
and adaptive immune system [52], but also very complex, such as in the etiology and
development of several chronic inflammatory disorders, including inflammatory
bowel diseases (IBD), irritable bowel syndrome (IBS), and gastrointestinal cancers
[72, 73]. A series of investigations in human [74] and animal models [75–77] have
provided a set of reference metabolic profiles of gut intestinal biopsies that can be
used to assess not only compartment structure and function but also the gut microbial
impact at the tissue level [78]. Such applications will help in identifying main
metabolic processes conserved across species on which gut microbiota modulates to

shape the microenvironment. For instance, the investigations illustrated how micro-
bial-dependent variations along the upper intestine, an element often underestimated
due to low bacterial populations, may affect utilization efficiency of dietary proteins
and amino acids and their subsequent availability to extra-intestinal tissues. Moreover,
some reference data were generated to investigate changes in gut functionality, such
as gut permeability, using metabolic profiling of biofluids [79, 80], since normal age-
ing is associated with a number of significant changes in intestinal function, which
may impact on daily energy intake [81]. Since under-nutrition is common in older
people and has been implicated in the development and progression of chronic dis-
ease commonly affecting the elderly, such insights will be key for tailoring personal
nutritional intervention and monitoring patient response on an individual basis.
Furthermore, indirect and minimally invasive metabolic indicators of specific intes-
tinal structural or functional states (e.g., digestion, mobility, permeability) will be
crucial for patient risk assessment and monitoring. For instance, such applications
seem promising in the field of chronic intestinal inflammatory conditions. Both man-
ifestations of IBD, ulcerative colitis (UC) and Crohn’s disease (CD), are mediated by
common and distinct mechanisms influenced by multiple environmental factors and
specific genetic predispositions, including gut microbiota. Advancing knowledge
regarding the mechanisms of IBD has led to the development of different therapeutic
solutions based on surgery [82], cannabinoids [83], immunosuppression [84], and
alternatively probiotic supplementation [85]. Although prognostic and monitoring
tools are currently lacking, metabolic profiling in combination with state-of-the-art
clinical and medical readouts is foreseen to be a valuable tool to differentiate and
follow-up IBD evolution and response to disease-modifying interventions. This is,
for instance, exemplified through the discovery of metabolic indicators of different
IBD determinants, including mucosal healing, gut permeability, absorption, diges-
tion, or inflammatory states. As an example, Winterkamp et al. reported previously
how N-methylhistamine, a key metabolite in mast cell metabolism involved in the
pathogenesis of IBD, could be used as an indicator of disease activity in patients
[86]. In this study, the urinary excretion of N-methylhistamine was associated with
elevated histamine production and metabolism in CD and UC and could be used as a
reliable diagnostic tool to monitor clinical and endoscopic disease activity in
IBD. Additional proofs of concept on the feasibility to identify some metabolic indi-
cators of early onsets of chronic inflammatory development offer also novel promis-
ing directions for patient monitoring and early patient stratification [87]. Additional
applications of noninvasive profiling of stool from patients provided novel insights
into the remodeling of the gut microbial communities and activities, concomitant to
malabsorption and element of protein-losing enteropathy [88, 89]. Another severe
GIT disorder for which the pandemic is affecting mainly the elderly is colorectal
cancer (CRC) [90]. Effective screening methods have significantly improved the
treatment modalities and effectiveness in the geriatric population. Nowadays, screen-
ing methods are mainly based on fecal assays for detecting blood presence or altered
DNA and endoscopic and computerized radiologic techniques. Therefore, CRC early
diagnosis still suffers from the absence of noninvasive or minimally invasive
180 E. Biagi et al.
diagnostics for early detection. Numerous studies reported that the general popula-
tion of intestinal bacteria is associated with initiation of colon cancer via the produc-
tion of carcinogens, cocarcinogens, or pro-carcinogen substances [91]. Recently,
metabonomics applications identified novel lipid markers of early CRC in American
and Japanese populations using MS analyses of serum samples [92, 93]. Moreover,
the application of NMR metabolic profiling to fecal water extracts has interesting
potential as a diagnostic tool for detecting CRC [94]. NMR metabolic profiles of
fecal water extracts from CRC patients were markedly distinct from controls with
lower concentrations of SCFA (acetate and butyrate) and higher levels of proline and
cysteine, the latter being two major components of most colonic epithelium mucus
glycoproteins. Conventional tumor markers are relatively unsuitable for detecting
carcinoma at an early stage and such novel insights may become critical in the devel-
opment of tailored approaches to reduce disease burden.
9.5 Conclusions
The gut microbiota influences human health status throughout the whole lifespan, from
birth to aging. It is generally accepted that the disruption of the finely balanced interac-
tion between host and microbiota can contribute to the progression of diseases and/or
predispose the human host to develop age-related disorders later in life. Yet, recent
advances in systems biology approaches including metabonomics and microbial ecol-
ogy have shown that the contribution of the intestinal microbiota to the overall health
status of the host has been so far underestimated. It is therefore of crucial importance
for nutrition and health to understand and to metabolically characterize the interactive
molecular processes between the host and its microbiome. By opening a direct bio-
chemical window into the metabonome, metabonomics is a unique science perfectly
suited for the identification of biomarkers capable of providing better understanding of
the complex metabolic phenomenon. This makes metabonomics a very efficient
approach for generation of metabolic patterns for the comprehensive characterization
of metabolic health, including healthy aging, and in the generation of new insights in
the understanding of the interactions between diet and metabolism. By understanding
the mechanisms by which diet and lifestyle influence metabolism, also by means of
influencing the microbiota composition and functionality, it will be possible to develop
personalized strategies to maintain a reasonable health status during aging and possibly
to attain longevity, maintaining a high quality of life as long as possible.
References
1. Kirkwood TB. Understanding the odd science of ageing. Cell. 2005;120:437–47.

2. Weinert BT, Timiras PS. Invited review: theories of ageing. J Appl Physiol. 2003;95:
1706–16.
3. Franceschi C, Bonafè M, Valensin S, et al. Inflamm-ageing. An evolutionary perspective on

immunosenescence. Ann N Y Acad Sci. 2000;908:244–54.
4. Franceschi C. Inflammaging as a major characteristic of old people: can it be prevented or
cured? Nutr Rev. 2007;65:S173–6.
5. Cevenini E, Monti D, Franceschi C. Inflamm-ageing. Curr Opin Clin Nutr Metab Care.
2013;16:14–20.
6. Cevenini E, Caruso C, Candore G, et al. Age-related inflammation: the contribution of differ-
ent organs, tissues and systems. How to face it for therapeutic approaches. Curr Pharm Des.
2010;16:609–18.
7. Franceschi C, Capri M, Monti D, et al. Inflammaging and anti-inflammaging: a systemic per-
spective on ageing and longevity emerged from studies in humans. Mech Ageing Dev.
2007;128:92–105.
8. Evans WJ, Paolisso G, Abbatecola AM, et al. Frailty and muscle metabolism dysregulation in
the elderly. Biogerontology. 2010;11:527–36.
9. Fry CS, Rasmussen BB. Skeletal muscle protein balance and metabolism in the elderly. Curr
Ageing Sci. 2011;4:260–8.
10. Fiatarone MA, O’Neill EF, Ryan ND, et al. Exercise training and nutritional supplementation
for physical frailty in very elderly people. N Engl J Med. 1994;330:1769–75.
11. Bruce SJ, Breton I, Decombaz J, et al. A plasma global metabolic profiling approach applied
to an exercise study monitoring the effects of glucose, galactose and fructose drinks during
post-exercise recovery. J Chromatogr B Analyt Technol Biomed Life Sci. 2010;878:3015–23.
12. Yan BAJ, Wang G, et al. Metabonomic investigation into variation of endogenous metabolites in
professional athletes subject to strength-endurance training. J Appl Physiol. 2009;106:531–8.
13. Wijeyesekera A, Selman C, Barton RH, et al. Metabotyping of long-lived mice using 1H NMR
spectroscopy. J Proteome Res. 2012;11:4425–35.
14. Franceschi C, Motta L, Valensin S, et al. Do men and women follow different trajectories to
reach extreme longevity? Italian Multicenter Study on Centenarians (IMUSCE). Ageing
(Milano). 2000;12:77–84.
15. Collino S, Montoliu I, Martin FP, et al. Metabolic signatures of extreme longevity in northern
italian centenarians reveal a complex remodeling of lipids, amino acids, and gut microbiota
metabolism. PLoS One. 2013;8:e56564.
16. Gonzalez-Covarrubias V, Beekman M, Uh HW, et al. Lipidomics of familial longevity. Aging
Cell. 2013;12:426–34.
17. Dekker P, Maier AB, et al. Stress-induced responses of human skin fibroblasts in vitro reflect
human longevity. Ageing Cell. 2009;8:595–603.
18. Dekker P, Meissner A, Dirks RW, et al. Human in vivo longevity is reflected in vitro by dif-
ferential metabolism as measured by (1)H-NMR profiling of cell culture supernatants. Mol
Biosyst. 2012;8:783–9.
19. Brookmeyer R, Johnson E, Ziegler-Graham K, et al. Forecasting the global burden of
Alzheimer’s disease. Alzheimers Dement. 2007;3:186–91.
20. Han X, Rozen S, Boyle SH, et al. Metabolomics in early Alzheimer’s disease: identification of
altered plasma sphingolipidome using shotgun lipidomics. PLoS One. 2011;6:e21643.
21. Sato Y, Suzuki I, Nakamura T, et al. Identification of a new plasma biomarker of Alzheimer’s
disease using metabonomics technology. J Lipid Res. 2012;53:567–76.
22. Chan RB, Oliveira TG, Cortes EP, et al. Comparative lipidomic analysis of mouse and human
brain with Alzheimer disease. J Biol Chem. 2012;287:2678–88.
23. Tremaroli V, Bäckhed F. Functional interactions between the gut microbiota and host
metabolism. Nature. 2012;489:242–9.
24. Sommer F, Backhed F. The gut microbiota – masters of host development and physiology. Nat
Rev Microbiol. 2013;11:227–38.
25. Tannock GW. New perceptions of the gut microbiota: implications for future research.
Gastroenterol Clin North Am. 2005;34:361–82.
26. Xu J, Mahowald MA, Ley RE, et al. Evolution of symbiotic bacteria in the distal human intes-
tine. PLoS Biol. 2007;5:e156.
182 E. Biagi et al.
27. Jia W, Li H, Zhao L, Nicholson JK. Gut microbiota: a potential new territory for drug targeting.
Nat Rev Drug Discov. 2008;7:123–9.
28. Cheng J, Palva AM, de Vos WM, et al. Contribution of the intestinal microbiota to human
health: from birth to 100 years of age. Curr Top Microbiol Immunol. 2013;358:323–46.
29. Biagi E, Candela M, Fairweather-Tait S, et al. Ageing of the human metaorganism: the micro-
bial counterpart. Age (Dordr). 2012;34:247–67.
30. Biagi E, Candela M, Turroni S, et al. Ageing and gut microbes: perspectives for health main-
tenance and longevity. Pharmacol Res. 2013;69:11–20.
31. Pédron T, Sansonetti P. Commensals, bacterial pathogens and intestinal inflammation: an
intriguing ménage à trois. Cell Host Microbe. 2008;3:344–7.
32. Biagi E, Nylund L, Candela M, et al. Through ageing, and beyond: gut microbiota and inflam-
matory status in seniors and centenarians. PLoS One. 2010;5:e10667.
33. Mäkivuokko H, Tiihonen K, Tynkkynen S, et al. The effect of age and non-steroidal anti-
inflammatory drugs on human intestinal microbiota composition. Brit J Nutr. 2010;103:227–34.
34. Claesson MJ, Cusack S, O’Sullivan O, et al. Composition, variability, and temporal stability of
the intestinal microbiota of the elderly. Proc Natl Acad Sci U S A. 2011;108:4586–91.
35. Rea MC, O’Sullivan O, Shanahan F, et al. Clostridium difficile carriage in elderly subjects and
associated changes in the intestinal microbiota. J Clin Microbiol. 2012;50:867–75.
36. Mueller S, Saunier K, Hanisch C, et al. Differences in fecal microbiota in different European
study populations in relation to age, gender, and country: a cross-sectional study. Appl Environ
Microbiol. 2006;72:1027–33.
37. Mariat D, Firmesse O, Levenez F, et al. The firmicutes/bacteroidetes ratio of the human micro-
biota changes with age. BMC Microbiol. 2009;9:123.
38. Rajilić-Stojanović M, Heilig HGHJ, Molenaar D, et al. Development and application of the
human intestinal tract chip, a phylogenetic microarray: analysis of universally conserved phy-
lotypes in the abundant microbiota of young and elderly adults. Environ Microbiol.
2009;11:1736–51.
39. Biagi E, Candela M, Franceschi C, et al. The ageing gut microbiota: new perspectives. Ageing
Res Rev. 2011;10:428–9.
40. Hayashi H, Sakamoto M, Kitahara M, et al. Molecular analysis of fecal microbiota in elderly
individuals using 16S rDNA library and T-RFLP. Microbiol Immunol. 2003;47:557–70.
41. Bartosch S, Fite A, Macfarlane GT, et al. Characterization of bacterial communities in feces
from healthy elderly volunteers and hospitalized elderly patients by using real-time PCR and
effects of antibiotic treatment on the fecal microbiota. Appl Environ Microbiol.
2004;70:3575–81.
42. Tiihonen K, Tynkkynen S, Ouwehand A, et al. The effect of ageing with and without non-
steroidal anti-inflammatory drugs on gastrointestinal microbiology and immunology. Br J
Nutr. 2008;100:130–7.
43. van Tongeren SP, Slaets JPJ, Harmsen HJM, et al. Fecal microbiota composition and frailty.
Appl Environ Microbiol. 2005;71:6438–42.
44. Zwielehner J, Liszt K, Handschur M, et al. Combined PCR-DGGE fingerprinting and
quantitative-PCR indicates shifts in fecal population sizes and diversity of Bacteroides, bifido-
bacteria and Clostridium cluster IV in institutionalized elderly. Experim Gerontol.
2009;44:440–6.
45. Cani PD, Delzenne NM. The role of the gut microbiota in energy metabolism and metabolic
disease. Curr Pharm Des. 2009;15:1546–58.
46. Neish AS. Microbes in gastrointestinal health and disease. Gastroenterology. 2009;136:
65–80.
47. Maslowski KM, Mackay CR. Diet, gut microbiota and immune responses. Nat Immunol.
2011;12:5–9.
48. Barcenilla A, Pryde SE, Martin JC, et al. Phylogenetic relationships of butyrate-producing
bacteria from the human gut. Appl Environ Microbiol. 2000;66:1654–61.
49. Louis P, Flint HJ. Diversity, metabolism and microbial ecology of butyrate-producing bacteria
from the human large intestine. FEMS Microbiol Lett. 2009;294:1–8.
50. Thibault R, Blachier F, Darcy-Vrillon B, et al. Butyrate utilization by the colonic mucosa in
inflammatory bowel diseases: a transport deficiency. Inflamm Bowel Dis. 2010;16:684–95.
51. Hippe B, Zwielehner J, Liszt K, et al. Quantification of butyryl CoA:acetate CoA-transferase
genes reveals different butyrate production capacity in individuals according to diet and age.
FEMS Microbiol Lett. 2011;316:130–5.
52. Guigoz Y, Doré J, Schiffrin EJ. The inflammatory status of old age can be nurtured from the
intestinal environment. Curr Opin Clin Nutr Metab Care. 2008;11:13–20.
53. Donini LM, Savina C, Cannella C. Nutritional interventions in the anorexia of ageing. J Nutr
Health Ageing. 2010;14:494–6.
54. Petersson J, Schreiber O, Hansson GC, et al. Importance and regulation of the colonic mucus
barrier in a mouse model of colitis. Am J Physiol Gastrointest Liver Physiol.
2011;300:G327–33.
55. Suzuki T, Yoshida S, Hara H. Physiological concentrations of short-chain fatty acids immedi-
ately suppress colonic epithelial permeability. Br J Nutr. 2008;100:297–305.
56. Macia L, Thorburn AN, Binge LC, et al. Microbial influences on epithelial integrity and
immune function as a basis for inflammatory diseases. Immunol Rev. 2012;245:164–76.
57. Candela M, Guidotti M, Fabbri A, et al. Human intestinal microbiota: cross-talk with the host
and its potential role in colorectal cancer. Crit Rev Microbiol. 2011;37:1–14.
58. Ou J, DeLany JP, Zhang M, et al. Association between low colonic short-chain fatty acids and
high bile acids in high colon cancer risk populations. Nutr Cancer. 2012;64:34–40.
59. Boyle P, Ferlay J. Mortality and survival in breast and colorectal cancer. Nat Clin Pract Oncol.
2005;2:424–5.
60. Claesson MJ, Jeffery IB, Conde S, et al. Gut microbiota composition correlates with diet and
health in the elderly. Nature. 2012;488:178–84.
61. Nicholson JK, Holmes E, Kinross J, et al. Host-gut microbiota metabolic interactions. Science.
2012;336:1262–7.
62. Kau AL, Ahern PP, Griffin NW, et al. Human nutrition, the gut microbiome and the immune
system. Nature. 2011;474:327–36.
63. Collins SM, Surette M, Bercik P. The interplay between the intestinal microbiota and the brain.
Nat Rev Microbiol. 2012;10:735–42.
64. Holmes E, Li JV, Marchesi JR, et al. Gut microbiota composition and activity in relation to
host metabolic phenotype and disease risk. Cell Metab. 2012;16:559–64.
65. Holmes E, Kinross J, Gibson GR, et al. Therapeutic modulation of microbiota-host metabolic
interactions. Sci Transl Med. 2012;4:137rv6.
66. Martin FP, Collino S, Rezzi S, et al. Metabonomic applications to decipher gut microbial meta-
bolic influence in health and disease. Front Physiol. 2012;3:113.
67. El AS, Hooiveld G, Tremaroli V, et al. The gut microbiota and mucosal homeostasis: Colonized
at birth or at adulthood, does it matter? Gut Microbes. 2013;4:118–24.
68. Moco S, Collino S, Rezzi S, et al. Metabonomics perspectives in pediatric research. Pediatr
Res. 2013;73:570–6.
69. Qin J, Li R, Raes J, et al. A human gut microbial gene catalogue established by metagenomic
sequencing. Nature. 2010;464:59–65.
70. Martin FP, Sprenger N, Yap IK, et al. Panorganismal gut microbiome-host metabolic crosstalk.
J Proteome Res. 2009;8:2090–105.
71. Martin FP, Sprenger N, Montoliu I, et al. Dietary modulation of gut functional ecology studied
by fecal metabonomics. J Proteome Res. 2010;9:5284–95.
72. Martin FP, Collino S, Rezzi S. 1H NMR-based metabonomic applications to decipher gut
microbial metabolic influence on mammalian health. Magn Reson Chem. 2011;49 Suppl
1:S47–54.
73. Collino S, Martin FP, Rezzi S. Clinical metabolomics paves the way towards future healthcare
strategies. Br J Clin Pharmacol. 2013;75:619–29.
74. Wang Y, Holmes E, Comelli EM, et al. Topographical variation in metabolic signatures of
human gastrointestinal biopsies revealed by high-resolution magic-angle spinning (1)H NMR
spectroscopy. J Proteome Res. 2007;6:3944–51.
184 E. Biagi et al.
75. Wang Y, Tang H, Holmes E, et al. Biochemical characterization of Rat intestine development
using high-resolution magic-angle-spinning 1H NMR spectroscopy and multivariate data anal-
ysis. J Proteome Res. 2005;4:1324–9.
76. Martin FP, Wang Y, Yap IK, et al. Topographical variation in murine intestinal metabolic pro-
files in relation to microbiome speciation and functional ecological activity. J Proteome Res.
2009;8:3464–74.
77. Martin FP, Montoliu I, Kochhar S, et al. Chemometric strategy for modeling metabolic bio-
logical space along the GIT and assessing microbial influences. Anal Chem.
2010;82:9803–11.
78. Martin FP, Wang Y, Sprenger N, et al. Effects of probiotic lactobacillus paracasei treatment on
the host Gut tissue metabolic profiles probed via magic-angle-spinning NMR spectroscopy. J
Proteome Res. 2007;6:1471–81.
79. Alonso C, Guilarte M, Vicario M, et al. Acute experimental stress evokes a differential gender-
determined increase in human intestinal macromolecular permeability. Neurogastroenterol
Motil. 2012;24:740–e349.
80. Rezzi S, Martin FP, Alonso C, et al. Metabotyping of biofluids reveals stress-based differences
in gut permeability in healthy individuals. J Proteome Res. 2009;8:4799–809.
81. Parker BA, Chapman IM. Food intake and ageing–the role of the gut. Mech Ageing Dev.
2004;125:859–66.
82. Harper PH, Truelove SC, Lee EC, et al. Split ileostomy and ileocolostomy for Crohn’s disease
of the colon and ulcerative colitis: a 20 year survey. Gut. 1983;24:106–13.
83. Duncan M, Davison JS, Sharkey KA. Review article: endocannabinoids and their receptors in
the enteric nervous system. Aliment Pharmacol Ther. 2005;22:667–83.
84. Actis GC, Pazienza P, Rosina F. Mesalamine for inflammatory bowel disease: recent reapprais-
als. Inflamm Allergy Drug Targets. 2008;7:1–5.
85. Reid G, Jass J, Sebulsky MT, et al. Potential uses of probiotics in clinical practice. Clin
Microbiol Rev. 2003;16:658–72.
86. Winterkamp S, Weidenhiller M, Otte P, et al. Urinary excretion of N-methylhistamine as a
marker of disease activity in inflammatory bowel disease. Am J Gastroenterol.
2002;97:3071–7.
87. Baur P, Martin FP, Gruber L, et al. Metabolic phenotyping of the Crohn’s disease-like IBD
etiopathology in the TNF(DeltaARE/WT) mouse model. J Proteome Res. 2011;10:5523–35.
88. Marchesi JR, Holmes E, Khan F, et al. Rapid and noninvasive metabonomic characterization
of inflammatory bowel disease. J Proteome Res. 2007;6:546–51.
89. Le GG, Noor SO, Ridgway K, et al. Metabolomics of fecal extracts detects altered metabolic
activity of gut microbiota in ulcerative colitis and irritable bowel syndrome. J Proteome Res.
2011;10:4208–18.
90. Holt PR, Kozuch P, Mewar S. Colon cancer and the elderly: from screening to treatment in man-
agement of GI disease in the elderly. Best Pract Res Clin Gastroenterol. 2009;23:889–907.
91. O’Keefe SJ. Nutrition and colonic health: the critical role of the microbiota. Curr Opin
Gastroenterol. 2008;24:51–8.
92. Ritchie SA, Ahiahonu PW, Jayasinghe D, et al. Reduced levels of hydroxylated, polyunsaturated
ultra long-chain fatty acids in the serum of colorectal cancer patients: implications for early
screening and detection. BMC Med. 2010;8:13.
93. Ritchie SA, Heath D, Yamazaki Y, et al. Reduction of novel circulating long-chain fatty acids
in colorectal cancer patients is independent of tumor burden and correlates with age. BMC
Gastroenterol. 2010;10:140.
94. Monleon D, Morales JM, Barrasa A, et al. Metabolite profiling of fecal water extracts from
human colorectal cancer. NMR Biomed. 2009;22:342–8.
Chapter 10
Roles of Herbal Medicine in Modulating Gut
Microbiota Associated with Health
and Diseases
Yulan Wang and Huiru Tang
Abstract The perturbation of gut microbiota is increasingly recognized to be

associated with human health and diseases. The modulation of the gut microbiotal
community as a means to alleviate disease conditions provides a unique opportunity
for herbal medicine, due to the two-way interaction between gut microbiota and
herbal medicine. Herbal medicine contains a range of polyphenols that require
action from gut microbiota to effectively perform their biological function. The gut
microbiota are subsequently stimulated through this action. In this chapter, we out-
line the associations between gut microbiota and disease (particularly inflammatory
bowel disease (IBD), diabetes, and cancer), and the roles of herbal medicine in
alleviating disease conditions through modulating gut microbiota. In addition, we
discuss the functional uses and challenges of herbal medicine, which include the
quality control and elucidating mechanisms of action. Finally, we describe how a
metabonomics technique can provide a means for the quality control of herbal med-
icines and can be an efficient tool for elucidating the molecular mechanisms of dif-
ferent herbal treatments. Future research on herbal medicine should be focused on
Y. Wang (*)
Key Laboratory of Magnetic Resonance in Biological Systems, State Key Laboratory
of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Centre for Magnetic
Resonance, Wuhan Institute of Physics and Mathematics, Chinese Academy of Sciences,
Wuhan 430071, People’s Republic of China
Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases,
Hangzhou, People’s Republic of China
H. Tang
Key Laboratory of Magnetic Resonance in Biological Systems, State Key Laboratory
of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Centre for Magnetic
Resonance, Wuhan Institute of Physics and Mathematics, Chinese Academy of Sciences,
Wuhan 430071, People’s Republic of China
State Key Laboratory of Genetic Engineering, Biospectroscopy and Metabolomics,
School of Life Sciences, Fudan University, Shanghai 200433, People’s Republic of China

DOI 10.1007/978-1-4471-6539-2_10
186 Y. Wang and H. Tang
directly measuring altered gut microbiota and integrating this with a mechanistic
metabonomics evaluation.
Keywords Gut microbiota • Diseases • Metabonomics • Herbal medicine
10.1 Introduction
Humans have a large number of microbiota resident in the gut. Interactions between
host metabolism and gut microbiota, between gut microbiota themselves, and
between food/medicine ingested and gut microbiota make the human gut a very
complex system. There are approximately about 1,000 microbial species present in
the mammalian gut, which codevelop with the host throughout the host’s lifetime.
Microbes play an important role in host nutrition and health by supplying and digest-
ing nutrients, preventing pathogen colonization, and maintaining balanced mucosal
immunity, which is vital for our health. A shift in the gut microbial composition can
stimulate a specific disease-prone (dysbiosis) or disease-protective activity (probio-
sis). For example, Lactobacillus reuteri strains can produce thiamine to benefit the
host [1], whereas Bifidobacteria may inhibit the colonization of potential pathogens
by competing with the nutrients and the binding site on the mucosa [2]. The gut
microbial community is very complex and dynamic [3], which can be affected by
the host’s genome, birth, age [4], nutrition [5], lifestyle, disease [6], and therapeutic
interventions (e.g., antibiotics [7], herbal medicines [8], and surgery treatments [9]).
In turn, the unbalanced microbial colonies may disturb the physiological homeosta-
sis, leading to various diseases such as colon cancer, inflammatory bowel disease
(IBD) [10], irritable bowel syndrome (IBS) [11], obesity [12–14], diabetes [15],
cardiovascular disease [16], autism [17], and allergic asthma [18]. The close asso-
ciations between the gut microbiotal community and disease status give a unique
opportunity for treatment by using traditional herbal medicine, via restoring the bal-
ance of the gut microbiotal community. Shaping the balance of the gut microbiotal
community by herbal medicine involves a two-way interaction. The active biologi-
cal ingredients of the herbal material are largely polyphenolic compounds, which
often cannot be absorbed directly by humans. Fortunately, however, enzymes
secreted by gut microbiota can metabolize these non-bioavailable phytochemicals,
facilitating their utilization. The enzyme-producing bacteria in return are selectively
stimulated, thereby modifying the balance of the gut microbiota [8, 19] (Fig. 10.1).
The beneficial effects of herbal medicine have gained growing interest in herbal
remedies, leading to a strong growth in consumer demand in plant-based products.
The current global market for plant-based health products is estimated to hit 93.15
billion dollars by 2015 (http://www.nutraingredients-usa.com). Prominent concerns
regarding the quality of phytomedicines are that they contain mixtures of many com-
pounds, which are often derived from plants or animal origins. Traditionally, the qual-
ity of these herbs is assessed by the experiences of herbal practitioners, who sometimes
10 Roles of Herbal Medicine in Modulating Gut Microbiota Associated with Health… 187
IBD
Cancer
Metabolic Diseases
Gut microbiota
Herbal Medicine
Quality Metabolic
control consequence
Metabonomics
Fig. 10.1 A summary of the key roles of metabonomics in herbal medicine and the herbal medi-
cine alleviating disease conditions via modifications of gut microbiota
refer to “active” molecules or arbitrarily chosen “marker” compounds in complex

plant extracts. The lack of scientifically accepted standards for herbal medicine has
long been an obstacle preventing herbal medicine from being recognized and inte-
grated into the standard healthcare system in Western countries. Therefore, one of the
challenges that phytomedicinal practitioners have encountered is the quality control of
herbal medicine. As a consequence of the lack of efficient quality control tools, large
variations of phytomedical products are encountered. For example, an analysis of 14
commercially available feverfew samples showed that each batch generated a unique
and characteristic spectra profile, with two of the batches being markedly different
from the other 12 [20]. The development of new quality control methods based on the
entire biochemical composition of the preparation without reference to “active”
molecules will help improve the quality and will make it more acceptable. Modern
metabonomics techniques appear to be well suited for this purpose. Metabonomics
involves the study of multivariate metabolic responses of complex organisms to
physiological and/or pathological stressors, including the consequent disruption of
systems regulation [21–24]. Metabonomics involves multivariate analysis of data
from Nuclear magnetic resonance (NMR) and mass spectrometry (MS) spectroscopy.
The commonly used multivariate data analysis tool includes a range of pattern recog-
nition techniques and random forest method as detailed in Chaps. 2 and 3. Using 1H
NMR spectroscopy, a wide range of plant metabolites can be detected including
sugars, amino acids, organic acids, and polyphenols. In such cases, all the chemical
components present in a single plant extract can be viewed simultaneously as a
Table 10.1 Summary of alternations of gut microbiota associated with diseases and corresponding
herbal medicine treatment
Diseases Gut microbiota Herbal medicine
IBD Bacteroides fragilis, Quercetin [26, 27]
enterococci, lactobacilli, Curcumin [28]
bifidobacteria [34] Prebiotics [32, 33]
Echinacea purpurea [34]
Metabolic Lactobacillus, Clostridium Jiangtang Xiaozhi [39, 40]
diseases [38]
Enterococcus faecalis [42] Mimosa pudica, Emblica officinalis, Azadirachta
indica [42]
Huangguaxiang [43]
Cancer H. pylori Juzentaihoto [45–48]
Pelargonium sidoides [49, 50]
Calophyllum brasiliense Camb. (Clusiaceae),
Mouriri elliptica Martius (Melastomataceae),
and Hancornia speciosa Gomez (Mangaba)
[51–53]
“metabolic fingerprint.” The holistic nature of metabonomics can overcome the draw-
back of considering “active components” (Table 10.1).
A single-component drug is the major characteristic of many Western medicines,
and the metabolism of the drug and the molecular target are often clearly defined.
However, herbal medicine consists of multiple components in nature, with mixtures
of various herbs; thus, the identification of molecular mechanisms and acting targets
is hindered by the complexity of the herbal medicine. Traditional drug discovery
routes are unsuitable for evaluating the mechanism of action of herbal medicines.
Metabonomics, on the other hand, measures the endpoint metabolic perturbations of
an intervention, including the multiple components of the herb. Therefore, the meta-
bonomics technique can also be adopted to evaluate mechanisms of a given com-
plex herbal medicine. The advantage of utilizing such a technique is its holistic
nature, which can provide important clues as to how herbal medicine really works.
In this chapter, we discuss the roles of gut microbiota in diseases and the ways in
which herbal medicine could alleviate disease conditions via modification of the gut
microbiotal community. In addition, the challenges encountered in herbal medicine
will be discussed and examples of possible solutions are shown.
10.2 Herbal Medicine Affects the Association

Between Disease State and Gut Microbiota
Many diseases are closely associated with an imbalanced gut microbiota, and herbal
medicines are capable of alleviating these disease conditions by offsetting
the imbalance of the microbiotal community. Inflammatory bowel disease (IBD) is
one such disease. IBD primarily includes ulcerative colitis (UC) and Crohn’s disease
(CD) and is a significant public health burden. The pathogenesis of IBD has not yet
been identified, but it has been widely accepted that the combination and interaction
of genetics, environmental influences, and immunologic abnormalities contribute to
the occurrence and perpetuation of this disease [25]. Aminosalicylates, antibiotics,
and a range of immunomodulation drugs are used to manage IBD conditions.
Complementary improvement of IBD conditions using nondigestible food ingredi-
ents has also gained increased recognition. Quercetin is a flavonoid commonly pres-
ent in food and other plant materials. Reports have shown that quercetin has antiviral,
antioxidant, and anti-inflammatory properties [26]. Quercetin-containing microcap-
sules have been given to mice with acetic acid-induced colitis, which have demon-
strated that quercetin treatment is able to decrease neutrophil recruitment, attenuate
histological alterations, and reduce macroscopical damage. In addition, quercetin-
containing microcapsules can also prevent the reduction of the anti-inflammatory
cytokine IL-10 and provide antioxidant properties [27]. Curcumin is another compo-
nent abundant in plant foods that possesses anti-inflammatory and antioxidant activ-
ity and has been demonstrated to be a therapeutic agent for IBD [28]. These
phytochemicals are known to be metabolized by gut microbiota [29]; hence, despite
the anti-inflammatory and antioxidant actions of these phytochemicals, the ameliora-
tion of IBD conditions via modification of gut ecosystems could be one of the under-
lying mechanisms. This is because the bioavailability and bioactivity of these
phytochemicals depend on enzymes of the gut microbiota [30]. The gut ecosystem
could in turn be modified by the phytochemical ingested. Indeed, probiotics supple-
mentations have been suggested to be beneficial in IBD [31], and long-term ingestion
of prebiotics can selectively stimulate or limit the growth and/or activity of bacteria in
the colon, and thus provides a more sustainable effect on colonic bacteria, which
improves host health [32]. Nondigestible dietary fibers can act as prebiotics, and it
has been reported that ingesting a number of different dietary fibers seems to be ben-
eficial in IBD [33]. They also demonstrated that children with CD who achieved
remission by either complete or partial enteral feeding displayed significant modifica-
tions to their fecal microbiota. The profiles of fecal microbiota were stable over time
for healthy children, which suggested that targeting enteral microbiota using phyto-
chemicals or prebiotics can help achieve remission for IBD patients. Many herbal-
based materials have shown antimicrobial effects against certain pathogenic bacteria
in vitro; hence, there is an expectation for significant changes in gut microbiota
in vivo. A study was performed on the effects of the dietary supplement of Echinacea
purpurea on aerobic and anaerobic bacteria common to the human gastrointestinal
tract. They found that human subjects who consumed 1,000 mg of standardized
E. purpurea for 10 days had significantly increased total aerobic bacteria, particularly
Bacteroides fragilis. Supplementation did not significantly alter the number of enteric
bacteria, enterococci, lactobacilli, bifidobacteria, or total anaerobic bacteria [34].
Similar to IBD, type 2 diabetes is believed to be a result of complex
gene-environmental interactions. Recent evidence points to the importance of gut
microbiota as an environmental factor in metabolic diseases, including obesity
[14, 35] and diabetes [36]. Diabetic patients showed clear dysfunction of choline
metabolism, indicating a close association between gut microbiota and diabetes [37].
It was further suggested that the presence of Lactobacillus species correlated
positively with the levels of fasting glucose and glycosylated hemoglobin, whereas
the presence of Clostridium species correlated negatively with the levels of fasting
glucose, glycosylated hemoglobin, insulin, C peptide, and plasma triglycerides
[38]. Metformin is the first line of drug in delaying the onset of diabetic condition,
and the low-risk alternative herbal medicines are often used to improve glucose
tolerance. The Jiangtang Xiaozhi capsule is a traditional Chinese herbal formula-
tion, consisting of radix astragali and rehmannia root, radix pseudostellariae and
Mongolian snakegourd root. There have been animal studies and a small clinical
trial, along with studies of the effects of the individual herbs, which showed that the
formula has a great potential to improve diabetic condition [39]. Recently, a large
clinical trial, including 71 participants treated with Jiangtang Xiaozhi capsules for
16 weeks, showed a significant difference in the levels of fasting insulin between the
treatment group and the placebo group. Patients taking Jiangtang Xiaozhi had a
significant improvement in high-density lipoprotein (HDL) level compared to the
placebo group [40]. Although attention has not yet been paid to scrutinizing changes
of gut microbiota after treating the diabetic patients on this particular case, many of
the diabetes-alleviating drugs possess strong antimicrobial properties [41]. Other
studies have shown that antidiabetic herbal plants have antibacterial activity, which
is not surprising given the fact that most of the phytochemicals become bioavailable
after interacting with gut microbiota. Some Indian herbal medicines, which have
been known for their hypoglycemic activities, were screened for four Gram-negative
and three Gram-positive bacteria. The results showed that the extract of Mimosa
pudica has a strong antibacterial activity against Gram-positive bacteria such as
Enterococcus faecalis and Proteus vulgaris. Authors have also demonstrated that
extracts of Emblica officinalis have a broad spectrum of antimicrobial activity fol-
lowed by Syzygium cumunii and Azadirachta indica [42]. Huangguaxiang
(Matteuccia struthiopteris) was investigated on the hypoglycemic activities of
streptozotocin-induced diabetic rats. The results showed that treatment with
Huangguaxiang for 8 weeks significantly reduced the levels of triglycerides, low-
density lipoprotein, and cholesterol, while levels of bifidobacterium and lactobacil-
lus were also altered following Huangguaxiang treatment [43].
The potential roles of herbal medicine in treating cancer or preventing cancer
metastasis mainly lie in the possible synergic effects of phytochemicals with che-
motherapies and/or the promotion of the immunologic response of the host via
interactions with enteric microbiota [44]. The traditional Japanese medicine juzen-
taihoto, containing ten different herbal plants, has been widely used for the preven-
tion of various types of cancer metastasis [45, 46]. Antifungal effects have also been
shown for juzentaihoto in preventing Candida infection [47, 48]. The most direct
evidence for herbal medicine as a treatment for cancer via targeting microbiota is
their use in treating Helicobacter pylori (H. pylori)-induced gastric cancer.
H. pylori, a Gram-negative bacterium, is commonly found in the human stomach
and can cause various diseases including gastritis, peptic ulcer, and gastric cancer.
In addition to antibiotics treatment against H. pylori, treatments using alternative
herbal medicine are common worldwide. The mode of action could be antibacterial
activity, inhibition of adhesion of H. pylori to gastric mucus, or both in combina-
tion. The extract of Pelargonium sidoides roots, a South African herbal medicine,
has been shown to inhibit H. pylori growth and has a strong adhesion to AGS cells
and to intact gastric tissues from H. pylori-infected humans [49, 50]. A range of
plants native to Brazil, including Calophyllum brasiliense Camb. (Clusiaceae),
Mouriri elliptica Martius (Melastomataceae), and Hancornia speciosa Gomez
(Mangaba), have also displayed anti-H. pylori activity among others [51–53].
Phytomedicine has been commonly used for treating many conditions, apart
from the aforementioned cases, due to the fact that phytomedicines are normally
comprised of many plant extracts with perhaps thousands of metabolites. In most
cases, the molecular mode of action of the active ingredients of these herbal extracts
is unknown [54]. In addition, the origins of the plant, time of harvest, and prepara-
tion methods will affect the efficacy of the phytomedicine. Therefore, authenticat-
ing the medicinal plant and elucidating the mechanism of action using an objective
tool, such as metabonomics, would be a way forward to tackle the many research
challenges in this area.
10.3 Challenges and Solutions in Herbal Medicine
10.3.1 Assessing the Quality of Phytomedicine
Currently, the quality control of phytomedicines is carried out based on the active
ingredients present in a plant. A drawback of this method is that some of the
unknown ingredients may have potential synergic interactions with each other and
may have certain biological functions. As a consequence, the mechanism of action
of these products becomes an impossible task. Thus, the quality control of both raw
and final products in a holistic manner is necessary to ensure the consistency of
these products and to provide a fundamental ground for further understanding the
molecular mechanisms of these products.
Metabonomics that employs 1H NMR spectroscopy facilitates the simultaneous
detection of chemical components present in a plant extract as a “metabolic finger-
print” and can meet the requirements for the quality control of an herbal medicine
in a holistic manner. Multiple-component analysis, based on the combination of
high-resolution NMR spectroscopy with pattern recognition, has been employed to
investigate the effects of origin on the chemical compositions of chamomile
(Matricaria recutita). Clear differences between chamomiles from Northern Africa
(Egypt) and Eastern Europe (Hungary and Slovak Republic) can be seen.
Chamomiles also have distinguished profiles from Hungary and Slovak Republic
based on their metabolomic compositions, despite being close in terms of geo-
graphic location. Furthermore, this method is effective for monitoring the “purity”
of chamomile samples, such as the percentage of stalks mixed with flowers,
suggesting that this is an excellent method for authenticity and quality control [55].
From a processing point of view, NMR-based metabonomics methods have also
been extremely powerful in distinguishing samples extracted with different methods
and samples collected at different seasons and dried with different procedures [56].
Metabonomics studies have also been carried out on the extracts of Artemisia annua
to discriminate samples from different sources and classify them according to their
antiplasmodial activity, without preknowledge of this activity [57]. The use of par-
tial least squares analysis also allows the predictions of actual values of such activi-
ties for independent samples not used in the model construction. Another study was
conducted on the complex pharmaceutical preparations, such as St. John’s wort,
using multivariate analysis of full-resolution 1H NMR spectral data [58]. The results
showed that ten preparations from markets were compositionally diverse, and such
diversity resulted from plant extract preparation rather than post-extraction pro-
cesses. The combination of NMR technique and LC-DAD-MS method has been
used to investigate the differences between three Salvia miltiorrhiza Bunge (SMB)
cultivars. The study demonstrated that the combinational use of these methods was
effective for plant metabonomics phenotype analysis [59]. These examples are only
a reflection on the developments in this area and are by no means exhaustive. In fact,
many studies have also been carried out in terms of phytomedicines and authentic-
ity, and it is conceivable that such applications of metabonomics technology will be
extended much further in the near future.
10.3.2 Elucidating the Mechanisms of Phytomedicine
Since herbal medicine is a mixture of many plant materials with many chemical
components, it is almost impossible to conduct classic pharmacological assays to
clearly demonstrate metabolisms of each chemical or unravel specific drug target in
an herbal-based drug. These difficulties have restricted their use worldwide.
Metabonomics simplifies the complexity by measuring the endpoints of an inter-
vention or a drug effect and hence could provide an alternative strategy for the
assessment of herbal medicine. The utilization of a metabonomic approach to evalu-
ate the metabolic action of an herbal medicine has been demonstrated in the human
ingestion of chamomile tea [60]. A total of 14 participants were given chamomile
tea every morning, and urine samples were collected after one and a half hours after
drinking the tea. The metabonomics analysis of urine samples was able to show that
chamomile ingestion is a mild intervention to the human body in general, and it
causes a reduction in oxidative stress and alters the state of gut microbiota, which
was reflected by alternations in the levels of hippurate. The effects of chamomile
ingestion on human metabolism were not completely recoverable within a succes-
sive week after ingestion. It could suggest that the recovery of gut microbiota is a
long process. Metabonomics was also used to evaluate the mechanism of Xia Yu Xue
decoction, which is a traditional Chinese medicine used for treating liver diseases.
Metabolic trajectory showed the trend of renormalization of the Xia Yu Xue
decoction to CCl4-induced liver dysfunction. In addition, changed metabolites

indicated modulations of energy metabolism, microflora metabolism, amino acid
and fatty acid metabolism, which are found to be associated with Xia Yu Xue decoc-
tion ingestion [61]. The traditional medical treatment targets on the balances of yin
and yang. “Kidney-yang deficiency syndrome” is one of the conditions that relate to
functional disorders associated with the hypothalamic-pituitary-gland axis [62].
Epimedium (Berberidaceae) is known in Chinese medicine as able to strengthen
“yang” and is often used to treat “kidney-yang deficiency.” The metabonomics tech-
nique was applied to investigate the capability of Epimedium koreanum to restore
metabolic disorder in animals with “kidney-yang deficiency.” After 15 days of
orally administering Epimedium koreanum extract, metabolic disorders associated
with “kidney-yang deficiency” returned to normal. These disordered metabolic
pathways included amino acid metabolism, lipid metabolism, and energy metabo-
lism. Importantly, Epimedium koreanum ingestion possessed effects on balancing
gut microbiota [63]. Most herbal medicines contain a range of polyphenolic
compounds, which often act as active ingredients. Gallic acid is one of these poly-
phenolic compounds, which has been shown to inhibit xanthine oxidase [64, 65],
ribonucleotide reductase [66, 67], and histamine release in mast cells [68].
Systematically analyzing the metabolic effects of gallic acid to the metabonomes of
rat plasma, liver, urine, and feces showed that gallic acid promoted oxidative stress
and resulted insignificant metabolic changes involving glycogenolysis, glycolysis,
tricarboxylic acid cycle (TCA), and the metabolism of amino acids, purines, and
pyrimidines, together with gut microbiota functions [69].
10.4 Concluding Remarks
In this chapter, we have shown examples of the close associations between human
diseases and the gut microbiota living within us, and we have exemplified how
herbal medicine could treat disease and maintain health by modifying gut microbi-
ota. In addition, we have emphasized the advantages and challenges of herbal medi-
cine. The literature has shown that there is a potential for future research on the
efficacies of herbal medicine and the requirement for international standards for
herbal medicine to be established. We have shown that metabolomics can facilitate
the understanding of the intrinsic quality of herbal medicine and the evaluation of
the therapeutic effects of the complex herbal formulas. This technique should be
equally effective for evaluating mineral treatment and acupuncture, which have not
been discussed in this chapter. Understanding the mechanisms of action is a neces-
sary step for herbal medicine to be more adoptable worldwide, which should cer-
tainly be warranted some attention. It is with no doubt that further development of
the metabonomics technique, such as high-sensitivity and high-specificity detection
of metabolites, as well as advances in effective molecular identification, would pro-
mote the process of evaluation and acceptance of herbal medicine. Although meta-
bonomics can provide a great understanding of the molecular mechanisms of herbal
medicine, direct measurement of alterations in gut microbiota associated with

herbal treatment using microbiological assays is still lacking. Biological assays
evaluating changes in gut microbiotia should be concurrently joined with a metabo-
nomics evaluation of the mechanisms of specific herbal formulas to provide a com-
prehensive view on the action of herbal medicine.
Acknowledgment We acknowledge financial support from the National Basic Research Program
of China (2012CB934004, 2009CB118804) and the National Natural Science Foundation of China
(20825520, 21221064, 31070854). We would also like to thank S. X. Tang for proofreading.
References
1. Carroll IM, et al. Alterations in composition and diversity of the intestinal microbiota in
patients with diarrhea-predominant irritable bowel syndrome. Neurogastroenterol Motil.
2012;24(6):521-+.
2. Hooper LV, Gordon JI. Commensal host-bacterial relationships in the gut. Science.
2001;292(5519):1115–8.
3. Gordon JI. Honor thy gut symbionts redux. Science. 2012;336(6086):1251–3.
4. Kurokawa K, et al. Comparative metagenomics revealed commonly enriched gene sets in
human gut microbiomes. DNA Res. 2007;14(4):169–81.
lism. Nature. 2012;489(7415):242–9.
6. Le Gall G, et al. Metabolomics of fecal extracts detects altered metabolic activity of gut micro-
biota in ulcerative colitis and irritable bowel syndrome. J Proteome Res.
2011;10(9):4208–18.
7. Rea MC, et al. Effect of broad- and narrow-spectrum antimicrobials on Clostridium difficile
and microbial diversity in a model of the distal colon. Proc Natl Acad Sci. 2011;108(Supplement
1):4639–44.
8. Li HK, et al. Traditional Chinese medicine: balancing the gut ecosystem. Phytother Res.
2009;23(9):1332–5.
9. Zhang HS, et al. Human gut microbiota in obesity and after gastric bypass. Proc Natl Acad Sci
U S A. 2009;106(7):2365–70.
10. Nagalingam NA, Lynch SV. Role of the microbiota in inflammatory bowel diseases. Inflamm
Bowel Dis. 2012;18(5):968–80.
11. Ponnusamy K, et al. Microbial community and metabolomic comparison of irritable bowel
syndrome faeces. J Med Microbiol. 2011;60(6):817–27.
12. Greenblum S, Turnbaugh PJ, Borenstein E. Metagenomic systems biology of the human gut
microbiome reveals topological shifts associated with obesity and inflammatory bowel dis-
ease. Proc Natl Acad Sci U S A. 2012;109(2):594–9.
13. Turnbaugh PJ, Gordon JI. The core gut microbiome, energy balance and obesity. J Physiol
Lond. 2009;587(17):4153–8.
14. Turnbaugh PJ, et al. A core gut microbiome in obese and lean twins. Nature.
2009;457(7228):480–U7.
15. Mathis D, Benoist C. The influence of the microbiota on type-1 diabetes: on the threshold of a
leap forward in our understanding. Immunol Rev. 2012;245:239–49.
16. Wang HY, et al. Bioactivity enhancement of herbal supplements by intestinal microbiota
focusing on ginsenosides. Am J Chin Med. 2011;39(6):1103–15.
17. Topping DL, Clifton PM. Short-chain fatty acids and human colonic function: roles of resistant
starch and nonstarch polysaccharides. Physiol Rev. 2001;81(3):1031–64.
18. Russell SL, et al. Early life antibiotic-driven changes in microbiota enhance susceptibility to
allergic asthma. EMBO Rep. 2012;13(5):440–7.
19. Zhao L, et al. Targeting the human genome-microbiome axis for drug discovery: inspirations
from global systems biology and traditional Chinese medicine. J Proteome Res.
2012;11(7):3509–19.
20. Bailey NJC, et al. Multi-component metabolic classification of commercial feverfew prepara-
tions via high-field H-1-NMR spectroscopy and chemometrics. Planta Med.
2002;68(8):734–8.
21. Nicholson JK, Lindon JC, Holmes E. ‘Metabonomics’: understanding the metabolic responses
of living systems to pathophysiological stimuli via multivariate statistical analysis of biologi-
cal NMR spectroscopic data. Xenobiotica. 1999;29(11):1181–9.
22. Nicholson JK, et al. Metabonomics: a platform for studying drug toxicity and gene function.
Nat Rev Drug Discov. 2002;1(2):153–61.
23. Tang HR, Wang YL. Metabonomics: a revolution in progress. Prog Biochem Biophys.
2006;33(5):401–17.
24. Lindon JC, Holmes E, Nicholson JK. Metabonomics: systems biology in pharmaceutical
research and development. Curr Opin Mol Ther. 2004;6:265–72.
25. Hanauer SB. Inflammatory bowel disease: epidemiology, pathogenesis, and therapeutic oppor-
tunities. Inflamm Bowel Dis. 2006;12:S3–9.
26. Vasanthi HR, ShriShriMal N, Das DK. Phytochemicals from plants to combat cardiovascular
disease. Curr Med Chem. 2012;19(14):2242–51.
27. Guazelli CFS, et al. Quercetin-loaded microcapsules ameliorate experimental colitis in mice
by anti-inflammatory and antioxidant mechanisms. J Nat Prod. 2013;76(2):200–8.
28. Ali T, Shakir F, Morton J. Curcumin and inflammatory bowel disease: biological mechanisms
and clinical implication. Digestion. 2012;85(4):249–55.
29. Atkinson C, Frankenfeld CL, Lampe JW. Gut bacterial metabolism of the soy isoflavone daid-
zein: exploring the relevance to human health. Exp Biol Med. 2005;230(3):155–70.
30. Karlsson FH, et al. A closer look at bacteroides: phylogenetic relationship and genomic impli-
cations of a life in the human gut. Microb Ecol. 2011;61(3):473–85.
31. Ewaschuk JB, Dieleman LA. Probiotics and prebiotics in chronic inflammatory bowel dis-
eases. World J Gastroenterol. 2006;12(37):5941–50.
32. Lim CC, Ferguson LR, Tannock GW. Dietary fibres as “prebiotics”: implications for colorectal
cancer. Mol Nutr Food Res. 2005;49(6):609–19.
33. Lionetti P, et al. Enteral nutrition and microflora in pediatric Crohn’s disease. J Parenter Enteral
Nutr. 2005;29(4):S173–8.
34. Hill LL, et al. Echinacea purpurea supplementation stimulates select groups of human gastro-
intestinal tract microbiota. J Clin Pharm Ther. 2006;31(6):599–604.
35. Noble D, et al. Risk models and scores for type 2 diabetes: systematic review. Br Med
J. 2011;343:1–31.
36. Qin JJ, et al. A metagenome-wide association study of gut microbiota in type 2 diabetes.
Nature. 2012;490(7418):55–60.
37. Zhang X, et al. Human serum metabonomic analysis reveals progression axes for glucose
intolerance and insulin resistance statuses. J Proteome Res. 2009;8(11):5188–95.
38. Karlsson FH, et al. Gut metagenome in European women with normal, impaired and diabetic
glucose control. Nature. 2013;498(7452):99-+.
39. Ge Z, Jin L, Guo Y. Experimental study of jiangtang xiaozhi tablet on decreasing the levels of
blood glucose and serum lipids in transgenic mice with diabetes mellitus. Chin J Integ Tradit
West Med. 2012;32(8):1095–8.
40. Grant SJ, et al. Chinese herbal medicine for impaired glucose tolerance: a randomized placebo
controlled trial. BMC Complement Altern Med. 2013;13:104. p. (14 May 2013)-(14 May
2013).
41. Baharum SN, Nathan ME, Mokhtar MAA. Molecular characterization of lipase producer from
local marine environment. J Pure Appl Microbiol. 2009;3(2):387–92.
42. Jyothi KS, et al. Anti-bacterial response of twelve plants of folk-lore medicine used against
diabetes mellitus. J Phytol Res. 2007;20(2):161–70.
43. Hai-bo W, et al. Role of Huangguaxiang in lipid metabolism and its effect on normal intestinal
bacteria of diabetic rats. Chin J Microecol. 2009;21(10):916–7.
44. Wang C-Z, Calway T, Yuan C-S. Herbal medicines as adjuvants for cancer therapeutics. Am J
Chin Med. 2012;40(4):657–69.
45. Saiki I. A Kampo medicine “Juzen-taiho-to” – prevention of malignant progression and
metastasis of tumor cells and the mechanism of action. Biol Pharm Bull. 2000;23(6):677–88.
46. Sakamoto S, et al. Anticancer effects of a Chinese herbal medicine, Juzen-taiho-to, in combi-
nation with or without 5-fluorouracil derivative on DNA-synthesizing enzymes in
1,2-dimethylhydrazine induced colonic-cancer in rats. Am J Chin Med. 1991;19(3–4):
233–41.
47. Abe S, et al. Protective effect of oral administration of a traditional medicine, Juzen-taiho-to,
and its components on lethal Candida albicans infection in immunosuppressed mice.
Immunopharmacol Immunotoxicol. 1998;20(3):421–31.
48. Akagawa G, Abe S, Tansho S. Protection of C3H/HE J mice from development of Candida
albicans infection by oral administration of Juzen-taiho-to and its component, ginseng radix:
possible roles of macrophages in the host defense mechanisms. Immunopharmacol
Immunotoxicol. 1996;18(1):73–89.
49. Beil W, Kilian P. EPs (R) 7630, an extract from Pelargonium sidoides roots inhibits adherence
of Helicobacter pylori to gastric epithelial cells. Phytomedicine. 2007;14:5–8.
50. Wittschier N, Faller G, Hensel A. An extract of Pelargonium sidoides (EPs 7630) inhibits in
situ adhesion of Helicobacter pylori to human stomach. Phytomedicine. 2007;14(4):285–8.
51. Moleiro FC, et al. Mouriri elliptica: validation of gastroprotective, healing and anti-
Helicobacter pylori effects. J Ethnopharmacol. 2009;123(3):359–68.
52. Moraes TM, et al. Hancornia speciosa: Indications of gastroprotective, healing and anti-
Helicobacter pylori actions. J Ethnopharmacol. 2008;120(2):161–8.
53. Souza MC, et al. In vitro and in vivo anti-Helicobacter pylori activity of Calophyllum brasil-
iense Camb. J Ethnopharmacol. 2009;123(3):452–8.
54. Vitor JMB, Vale FF. Alternative therapies for Helicobacter pylori: probiotics and phytomedi-
cine. FEMS Immunol Med Microbiol. 2011;63(2):153–64.
55. Wang YL, et al. Metabolomic strategy for the classification and quality control of phytomedi-
cine: a case study of chamomile flower (Matricaria recutita L.). Planta Med. 2004;70(3):
250–5.
56. Xiao C, et al. Revealing the metabonomic variation of rosemary extracts using H-1 NMR
spectroscopy and multivariate data analysis. J Agric Food Chem. 2008;56(21):10142–53.
57. Bailey NJC, et al. Prediction of anti-plasmodial activity of Artemisia annua extracts: applica-
tion of H-1 NMR spectroscopy and chernometrics. J Pharm Biomed Anal.
2004;35(1):117–26.
58. Rasmussen B, et al. Multivariate analysis of integrated and full-resolution H-1-NMR spectral
data from complex pharmaceutical preparations: St. John’s wort. Planta Med. 2006;72(6):
556–63.
59. Dai H, et al. Combined NMR and LC-DAD-MS analysis reveals comprehensive metabonomic
variations for three phenotypic cultivars of Salvia Miltiorrhiza Bunge. J Proteome Res.
2010;9(3):1565–78.
60. Wang YL, et al. A metabonomic strategy for the detection of the metabolic effects of chamo-
mile (Matricaria recutita L.) ingestion. J Agric Food Chem. 2005;53(2):191–6.
61. Gou X, et al. Urinary metabonomics characterization of liver fibrosis induced by CCl4 in rats
and intervention effects of Xia Yu Xue Decoction. J Pharm Biomed Anal. 2013;74:62–5.
62. Yang J, et al. The total flavones from Semen cuscutae reverse the reduction of testosterone
level and the expression of androgen receptor gene in kidney-yang deficient mice. J
Ethnopharmacol. 2008;119(1):166–71.
63. Huang D, et al. An integrated plasma and urinary metabonomic study using UHPLC-MS:
intervention effects of Epimedium koreanum on ‘Kidney-Yang Deficiency syndrome’ rats. J
Pharm Biomed Anal. 2013;76:200–6.
64. Lespade L, Bercion S. Theoretical study of the mechanism of inhibition of xanthine oxidase by
flavonoids and gallic acid derivatives. J Phys Chem B. 2010;114(2):921–8.
65. Masuoka N, Nihei K-i, Kubo I. Xanthine oxidase inhibitory activity of alkyl gallates. Mol Nutr
Food Res. 2006;50(8):725–31.
66. De Bona KS, et al. Syzygium cumini extract decrease adenosine deaminase, 5′ nucleotidase
activities and oxidative damage in platelets of diabetic patients. Cell Physiol Biochem.
2010;26(4–5):729–38.
67. Madlener S, et al. Gallic acid inhibits ribonucleotide reductase and cyclooxygenases in human
HL-60 promyelocytic leukemia cells. Cancer Lett. 2007;245(1–2):156–62.
68. Kim SH, et al. Gallic acid inhibits histamine release and pro-inflammatory cytokine produc-
tion in mast cells. Toxicol Sci. 2006;91(1):123–31.
69. Shi X, et al. Gallic acid intake induces alterations to systems metabolism in rats. J Proteome
Res. 2013;12(2):991–1006.
Chapter 11
Gut Microbiota and Metabolic Diseases:
From Pathogenesis to Therapeutic Perspective
Rémy Burcelin, Michael Courtney, and Jacques Amar
Abstract Intestinal microbiota is now considered as a “new organ” which, over

and above their genetic origin, de-orphans the pandemic development of metabolic
diseases. The trillions of bacteria and their corresponding million genes which
inhabit our gut provide a unique source of molecular hypotheses to explain the wide
diversity of metabolic diseases and hence form a basis to reach the important objec-
tive of personalized medicine. The gut and more recently the tissue microbiome
could be the source of: (1) new biomarkers predicting and classifying metabolic
diseases to help the clinician to propose the best therapeutic strategy, and (2) new
pharmacological and nutritional strategies to treat the cause rather than the conse-
quence of diabetes and obesity. The field of immunometabolism should be extended
to microbio-immunometabolism, thus reconciling the role of the environment, the
genetic background, and individual diversity in relation to the onset and develop-
ment of metabolic diseases.
Keywords Diabetes • Bacterial translocation • Immune system • Incretins

• Inflammation • Immunometabolism • Intestine • Metabolic diseases • Metabolic
inflammation • Metabolic infection • Microbiome • Obesity • Tissue microbiota
R. Burcelin, Ph.D. (*)

Institut National de la Santé et de la Recherche Médicale (INSERM), Toulouse, France
Université Paul Sabatier (UPS), Unité Mixte de Recherche (UMR) 1048, Institut de Maladies
Métaboliques et Cardiovasculaires (I2MC), 31432 Toulouse Cedex 4, France
M. Courtney, Ph.D.
Vaiomer S.A.S, 516 Rue Pierre et Marie Curie, 31670 Labège, France
J. Amar, M.D.
Institut National de la Santé et de la Recherche Médicale (INSERM), Toulouse, France
Université Paul Sabatier (UPS), Unité Mixte de Recherche (UMR) 1048, Institut de Maladies
Métaboliques et Cardiovasculaires (I2MC), 31432 Toulouse Cedex 4, France
Département de thérapeutique de l’hôpital Rangueil, Toulouse, France

DOI 10.1007/978-1-4471-6539-2_11
200 R. Burcelin et al.
11.1 Introduction
Evidence now shows that the pandemic progression of diabetes and obesity, as well
as their numerous complications (cardiovascular, hepatic, renal, neuronal), is caused
by lifestyle factors including sedentarity and fat-enriched diet which has replaced
fiber-enriched diet (Fig. 11.1). A dramatic outcome of this epidemic is the increas-
ing number of cardiovascular events leading to mortality [1–4] – the mechanisms at
play need to be delineated in order to define new therapeutic strategies. Over the last
decade lethal cardiovascular events associated with diabetes have progressed by
62 % [5, 6]. This is much higher than the risk linked to cholesterol levels or hyper-
tension. The incidence of type 2 diabetes is 4–5 % in Europe, 8–10 % in the USA,
and higher in South Asia [7]. These numbers have more than doubled over the last
20 years. Therefore, one can suggest that even if genetic analyses provide the basis
for such an epidemic, changes in our genome cannot be solely responsible. One
interpretation is that our genome is no longer adapted to environmental factors.
Numerous environmental hypotheses have been proposed. First, epigenetic non-
coded functions that are independent of genomic factors could have an impact.
Second, perhaps more realistically, the impact of changes in feeding habits and
Metabolic diseases: the pandemic origin
Stress, sedentarity…
Genetic few %
Nutritional
Most % microbiome
Epidemic of Metabolic diseases

complications, cardiovascular,
kidney, skin, liver, eyes….
Fig. 11.1 The pandemic origin of metabolic diseases. The impact of genetics, diet, and other fac-
tors on the incidence and the development of metabolic disease pandemic could have a change in
gut microbiota as an origin. Gut microbiota would also favor the development of complications of
metabolic diseases such as cardiovascular, liver, eye, and skin complications
11 Gut Microbiota and Metabolic Diseases: From Pathogenesis to Therapeutic… 201
social behavior is likely to be an important cause of the growing incidence of

metabolic diseases. This phenomenon could be linked to the microbiota, our “second
genome” harboring almost ten times more prokaryotic cells than eukaryotic cells in
our body [8]. All mammalian organisms are born sterile, without any microbiota
which is inherited at birth. During the first hours, days, and weeks of life, microbiota
from the mother and the environment colonize the body of the newborn in a specific
order [9]. The microbiota colonizes the intestine during the first 3–5 years of life
and is species-, age-, and sex-dependent [10]. The initial infant gut microbiota is a
simple structure usually dominated by bifidobacteria, and through a series of suc-
cessions and replacements, it shifts to a more complex adult pattern [11–15]. The
microbiota also undergoes substantial changes at the extremes of life, in infants and
older people, the ramifications of which are still being explored [16]. Then, through-
out life the bacterial ecology tends to vary mostly according to the environment
[17]. Each individual has at least 160 shared species and a number of well-balanced
host-microbial molecular relationships that define groups of individuals [18, 19].
This second genome is the metagenome. The importance of this metagenome
resides in its gene repertoire, 100 times superior to the eukaryotic nuclear genome
[18, 19], thus providing a huge genetic diversity susceptible to convey a plethora of
functions [18, 19]. Indeed, the tremendous efforts that have been made in bioinfor-
matic analyses have allowed the encoding and the deciphering of all sequences.
Humans host different metagenomes from multiple locations such as the skin, lungs,
vagina, and the mouth in addition to the intestine [8, 20]. The human gut hosts 100
trillion microorganisms, encompassing up to thousands of species at an average
concentration of 1014 per ml and weighing in average 1.5 kg [21]. A major observa-
tion is that the metagenomic diversity is extremely large and represents a signature
of each individual. Its plasticity is significant and depends on numerous environ-
mental and genetic factors that can evolve over time and could explain the rapid
development of metabolic diseases. The adult intestinal microbiota has been shown
to be relatively stable over time [22] and is sufficiently similar between individuals
to allow identification of a core microbiome comprising 66 dominant operational
taxonomic units (OTUs) that correspond to 38 % of sequence reads from 17 indi-
viduals [18, 19]. The core microbiota changes to become distinct in elderly subjects
from that observed for younger adults with a greater proportion of Bacteroides spp.
and typical abundance patterns of Clostridium groups. Interestingly, the onset of
metabolic diseases increases with age and is associated with a change in intestinal
microbiota as observed during aging [16, 23, 24]. Similarly, it was shown that
changes in gut microbiota characterize obesity and diabetes [18, 19]. This suggests
that each member can interact in a perfect mutual symbiosis defining a steady
microbiota [8, 20, 25]. Hence, a new concept has emerged with an important influ-
ence on our understanding of these pathologies. Major advances have been made
over the course of the last decade, thanks to the development of high-throughput
sequencing of the microbiota and to the use of germ-free mice. These have allowed
the demonstration of the causality of the microbiota from the gut [26], from the oral
cavity, and more recently from the tissue [27] on the development of metabolic
diseases. The following decades will be dedicated to the identification of the
molecular crosstalk between the microbiota and the host to understanding

mechanisms controlling diabetes and obesity. The intestinal immune system appears
to be a major player in the field since at birth bacterial colonization of the intestine
represents an outbreak of antigens that can educate the immune system, as well
other major functions such as the vascular and the nervous systems. The latter
through its connection with the brain could participate in the maturation of the cen-
tral nervous system over the course of a lifetime [28]. Consequently, beyond the key
physiological role of the intestinal microbiota in normal development, numerous
pathological issues could be the consequence of an impaired microbiota. In addition
to the first described role of the microbiota on the development of intestinal bowel
diseases, there is now evidence that a change in the intestinal bacterial ecology
could affect metabolic, vascular, liver, heart, oral, and neurodegenerative diseases.
More evidence is required to confirm a role in arthritis and skin immunological
diseases. The recent increase in the use of bariatric surgery for the treatment of
massive obesity and incidental diabetes has generated further evidence reinforcing
the role of intestinal microbiota in the control of metabolic diseases.
New clinical approaches for prevention and therapy are now being planned.
Functional food and pharmaceutical strategies, based on the targeting of the
microbiota to host interactions, can be initiated. In the face of the large diversity of
metabolic phenotypes, i.e., a large spectrum of fasted and postprandial glycemia
and different fat mass distribution, “intelligent and directed” food complements can
be proposed. They should not prevent or treat massively the populations but will
focus on subgroups of patients with similar microbiota-related diseases. The rele-
vance of treating the impact of gut microbiota on the diseases requires the develop-
ment of companion biomarkers. They should first define subgroups of patients with
similar microbiota profiles and should be able to follow up the efficacy of the micro-
biota change in correlation with the metabolic phenotype treated. Hence, they
will drive the therapeutic approach toward appropriate subgroups of patients to
improve the efficacy of the treatment and reduced secondary effects. Pharmaceutical
approaches targeting a mechanism central to the molecular crosstalk between
microbiota and the host, such as metabolic inflammation, have currently been devel-
oped by means of fecal transplant and immunomodulation. Therefore, it is now time
to consider the intestinal microbiota as a new organ controlling metabolism. This
organ is characterized by a high level of plasticity so that it can adapt to a change in
the host behavior in a reciprocal manner for the control of broad host physiological
functions. The understanding of its molecular components will lead to a totally new
way of interpreting physiological, clinical, and therapeutic data. Therefore, the limit
between nutritional and pharmacological strategies is vanishing. Intelligent food
supplements and cause-based pharmacological approaches will be the treatments of
tomorrow that still have to be identified.
Pharmaceutical strategies will benefit from the gain of knowledge generated
based on the intestinal microbiota to host relationship. The corresponding molecu-
lar crosstalk is currently being assessed and will be reviewed below. It includes
mechanisms controlling immunomodulation, bile acid conjugation, the intestinal
barrier, energy harvesting, and entero-endocrine hormone secretion. The bacterial
factors, i.e., the metafactors controlling these eukaryotic functions, are of inflammatory
origin such as the lipopolysaccharides, peptidoglycan fragments, flagellin, or
derived from fermentation products. Hence, pharmacological strategies should
emerge from these concepts and target subgroups of patients with these new drugs
acting on the host to microbiota relationship.
An important matter relates to the identification of subgroups of patients sensi-
tive to these new approaches, most likely characterized by a specific dysbiosis and
impaired gut and tissue microbiota. Hence, microbiota-based biomarkers precisely
identifying these subgroups will help the clinician to treat the patients with the
appropriate pharmaceutical strategy.
11.2 From the Main Features of the Pathophysiology

of Metabolic Diseases to Microbiota
Metabolic diseases are all characterized by alterations in energy balance which

explains, at least in part, the occurrence of obesity. The disease is the consequence
of either an increased energy uptake or a reduced energy expenditure demonstrating
that different mechanisms are responsible for increased body weight. Furthermore,
the accumulation of fat in the body is compartmentalized, i.e., abdominal, subcuta-
neous. or intratissular such as in the liver, the heart, or the pancreatic islets with
different consequences on overall health. The incidence of diabetes is much
increased in patients with normal body weight but with abdominal fat accumulation
which is considered the most deleterious for health [29]. Similarly, the accumula-
tion of triglycerides within the Langerhans islets impairs insulin secretion [30] that
can be restored by leptin treatment [31] activating uncoupling proteins such as
UCP2 [32]. At the onset of obesity, the storage of energy is associated with hyper-
trophy of the adipocytes and their hyperplasia to ensure a sufficient number of cells
[33–35]. The signals favoring adipose depot development could be linked to
increased adipocyte metabolism leading to a local hypoxia [36] and the recruitment
of adipocyte precursors [37, 38]. The maintenance of the adipose depot architecture
is ensured by a concomitant proliferation of endothelial precursors [39, 40],
increasing the adipose tissue capillaries and hence blood and energy supply [40].
Hyperphagia, supplying large amounts of energy, and hyperinsulinemia, although
associated with normal blood glucose profiles, also suggest a neuroendocrine origin
of the disease notably, the gut to brain axis which is recruited in response to an oral
glucose load [41] or food intake [42, 43]. A key enteroendocrine factor, glucagon-
like peptide 1 (GLP-1), triggers the gut–brain axis [42, 44, 45] which is impaired
during high-fat-diet-induced metabolic diseases [46]. The brain is also sensitive to
hormones from the periphery such as insulin and leptin which no longer efficiently
control food intake and energy partitioning. Altogether, impaired gut and peripheral
nutrients and hormone sensing systems, as well as signaling effectors toward the
adipose depot via the brain, are impaired during obesity. The increase of brain to
adipose depot signals then enhances the capacity of the adipose depot to store
Metabolic functions putatively controled by gut microbiota
Adipose tissue
Gene expression
development
SCFA,
LPS, PGN
Flagelin, Pancreatic endocrine
TMAO, metafactors Secretions??
Methane,
Sulfate
Nitrites … -Food
Firmicutes/ Intake
100 101 102 103
Proteobacteria
Bacterioidetes
Verrucomicrobias Muscle energy
expenditure
100 101 102

Cytokines -Steatosis
Metabolic
inflammation -Glucose
production
Fig. 11.2 Metabolic functions putatively controlled by gut microbiota. Metafactors, i.e., bacterial
compounds such as short-chain fatty acids, acetate, propionate, butyrate (SCFA), lipopolysaccha-
rides (LPS), peptidoglycan (PGN), flagellin, and trimethylamine oxide (TMAO), could be involved
in the control of metabolic functions, by means of gene expression and by installing a metabolic
inflammation leading to cytokine production which could then impair all metabolic functions
energy (Fig. 11.1). However, it should be noted that increasing food intake and
activating the storage of energy in fat depots cannot be considered as a pathological
mechanism since it is a normal behavior of the body following a fasting period to
replenish the fat stores. The pathology starts when this process is no longer down-
regulated. Different levels of deregulation occur involving either early control of
food intake and energy distribution or processes that control each step of the obesity
process. The intestinal microbiota has been proposed to interfere with hyperphagia
and to some extent with energy storage – this will be reviewed below. The emerging
role of gut microbiota in the gut to brain axis for the control of neural development,
behavior, and food intake will certainly generate new molecular hypotheses regard-
ing the development of obesity [28, 47–49].
Type 2 diabetes arises due to impaired secretion and action of insulin which
evolve either simultaneously or independently according to numerous influencing
factors (Fig. 11.2). An increased glucose intolerance with glycemic profiles higher
than controls is a common feature at the onset of the disease [50, 51]. This is associ-
ated with hepatic insulin resistance where insulin does not efficiently reduce hepatic
glucose production following a meal [50–52]. Liver insulin resistance is hence
involved in glucose intolerance and type 2 diabetes and explains fasting glycemia
[53–56]. Simultaneously, insulin-stimulated glucose uptake is reduced which is

most likely a consequence of an impaired cellular lipid metabolism [57]. Numerous
hypotheses could explain insulin resistance and among those the role played by a
low-grade inflammation, called metabolic inflammation since it is chronic and not
related to infection or cancer [58]. Briefly, metabolic inflammation is characterized
by an increased infiltration of immune cells within organs involved in the glycemic
control such as the adipose tissue, the liver, and muscles [59, 60]. Cells from the
innate and adaptive immune systems secrete cytokines such as TNFa and IL1-b that
impair insulin signaling, thus affecting both glucose and lipid metabolism (Fig. 11.2)
[59, 60]. Inflammation is increasingly regarded as a key process underlying meta-
bolic diseases [59–64]. In the adipose tissue of individuals with metabolic diseases,
this mechanism includes features characteristic of active local inflammation [63, 65,
66]. The cytokines released impair insulin signaling [67–69], thus leading to
cytokine-mediated insulin resistance [70, 71]. Macrophage infiltration of adipose
tissue has been described in both mice and humans [64–66, 72–74]. It is suggested
that these cells express TNF-alpha and iNOS that are characteristic of M1-type
macrophages which are responsible for almost all adipose tissue TNF-alpha expres-
sion and significant amounts of iNOS and IL-6 expression [73, 75, 76]. In addition,
using functional analyses and microarray technology, it has been demonstrated that
adipocyte progenitors and macrophages are characterized by a closed genome and
phenotypome [77, 78]. Based on measurements of phagocytic activity and gene
profiling analysis of different progenitor cells, we revealed that the origin of inflam-
mation could also be attributed to cells initially present in adipose fat pads such as
preadipocytes [77, 78]. Therefore, both infiltrating and resident cells are most likely
involved in the processes characterizing adipose tissue inflammation. Similarly,
lymphocytes are associated with adipose tissue inflammation [64, 79, 80]. CD8 T
cells are present in obese mouse adipose tissue even before the infiltration of the
tissue by macrophages. The results also showed that the immunological and genetic
depletion of CD8+ T cells lowered macrophage infiltration and adipose tissue
inflammation and improved systemic insulin resistance. Conversely, adaptive trans-
fer of CD8+ T cells to CD8-deficient mice aggravated adipose inflammation.
Co-culture and other in vitro experiments revealed a cycle of interactions between
CD8+ T cells, macrophages, and adipose tissue. Moreover, CD4+ regulatory T lym-
phocytes (Treg) (CD4+FoxP3+), inhibitory cells of the immune system, decrease in
obese adipose tissue [80]. Increases in Treg by antibody treatment (IL-2/anti-IL-2-
induced Treg proliferation) improve HFD-induced insulin resistance [80].
Furthermore, transfer of CD4+ T lymphocytes with anti-inflammatory properties
decreases HFD-induced glucose intolerance and insulin resistance [79]. The rea-
sons for adipose tissue T lymphocyte and macrophage infiltration are unknown, but
it was suggested that the corresponding antigens could be related to intestinal
microbiota [6, 27, 81].
Altogether, little is known about the sequence of events which lead to an
increased number of macrophages and lymphocytes in metabolic tissues. The
origin of this cascade of events could be related to a change in intestinal microbiota
and, as detailed below, to a change in tissue microbiota [27]. This is linked to a
translocation of bacteria and bacterial components from the intestine to tissues

establishing a tissue microbiota leading to metabolic inflammation. Hence, the role
of gut microbiota could be related to its direct role on molecular targets controlling
insulin secretion and action, hepatic glucose production, adipose tissue develop-
ment, and therefore, the incidence of diabetes. These hypotheses will be detailed
below.
Although more than 150 genetic loci are associated with the monogenic or
multifactorial forms of obesity and type 2 diabetes, their impact on the incidence of
the disease is rather low (5–10 % and 2 %, respectively) [82]. In genetically identi-
cal twins the incidence of type 2 diabetes is 20–60 % [83, 84]. Even in type 1 dia-
betes, genetically identical twins develop the disease with an incidence of no more
than 50 %. A major impact of the environment has recently been attributed to the
role of intestinal microbiota where genes involved in the recognition of bacterial
patterns were involved [85, 86]. The innate immunity to microbiota relationship was
causally implied. It was shown that the commensal microbial community alters sex
hormone levels and regulates autoimmune disease fate in individuals with high
genetic risk [87]. On the other hand, environmental factors such as stress, a seden-
tary lifestyle, and nutritional habit could explain the pandemic progression of meta-
bolic impairment. Humans could be considered as “super-organisms” as a result of
their symbiotic association with the gut microbiota [88].
11.3 Lessons from the Gut Microbiota to Metabolic

Diseases Relationship
Complex microbial ecosystems occupy the skin, mucosa, and alimentary tract of all
mammals, including humans [8]. The species that make up these communities vary
between hosts as a result of restricted migration of microorganisms between weak
and strong ecological interactions within hosts. Furthermore, diet, genotype, and
colonization history also influence this ecology [20]. Hence, a mutual relationship
characterizes the host to microbiota crosstalk by which each partner has its own
interest and informs the other of the environmental and metabolic situation. The
microbiota is now considered as a symbiont that shares with its host the influence of
the environment, diet, stress, and the physiological state. Specific communities
inhabit the different epithelia according to the physical and biochemical characteris-
tics of each location. With regard to the gastrointestinal tract, it is colonized by a vast
community of symbionts and commensals that have important effects on immune
function, nutrient processing, and a broad range of other host activities [89].
The precise role of intestinal microbiota on the control of metabolic diseases
emerged in 2004 with the discovery that germ-free mice resist high-fat-diet-induced
obesity [90]. It was shown that germ-free mice colonized with microbiota harvested
from the cecum of a healthy mouse gain 60 % body fat content and became insulin
resistant within 14 days despite reduced food intake. The mechanism was due to an
increased production of monosaccharides. These molecules are generated from the
metabolism of polysaccharides by the cecal microbiota, thus providing an additional

source of carbohydrate to the body. However, a rate-limiting factor was that a large
part of the microbiota cannot be identified since it is highly difficult to culture. This
major problem has been overcome by the use of very-high-throughput sequencing
techniques coupled with new bioinformatics approaches. This strategy allowed the
identification of the taxons within the microbiota from human and animal intestinal
content [91, 92] as well as the overall catalog of bacterial genes [18, 19]. This has
allowed the correlation of metabolic diseases to specific bacteria or groups of
bacterial genes [93]. An increased Firmicutes-to-Bacteroidetes ratio seems to be a
signature of metabolic diseases influencing processes related to energy harvesting,
intestinal permeability, bile acid metabolism, brain functions related to metabolism,
and immunomodulation.
11.3.1 The Gut Microbiota as a Signature

of Metabolic Diseases
An altered gut microbiota has been linked to metabolic diseases including obesity
[26, 94], diabetes [19], and cardiovascular diseases [95]. A core microbiome can be
found at the gene level, despite large variations in community membership, and that
variations from the core are associated with obesity [25, 94, 96–98]. Using a proto-
col for a metagenome-wide association study (MGWAS) based on deep shotgun
sequencing of the gut microbial DNA from 345 Chinese individuals, approximately
60,000 type 2 diabetes-associated markers have been identified and validated, thus
establishing the concept of a metagenomic linkage group [19]. This MGWAS analy-
sis showed that patients with type 2 diabetes were characterized by a moderate
degree of gut microbial dysbiosis, a decrease in the abundance of some universal
butyrate-producing bacteria, and an increase in various opportunistic pathogens, as
well as an enrichment of other microbial functions conferring sulfate reduction and
oxidative stress resistance. Three enterotypes could be identified from the Chinese
samples which were primarily made up of several highly abundant genera, includ-
ing Bacteroides, Prevotella, Bifidobacterium, and Ruminococcus. However, no
significant relationship between enterotype and type 2 diabetes status was found.
However, when using the gene reference (KEGG orthologue genes and eggNOG
group profiles) rather than the phylogenic profiles, a total of 1,345 biomarkers were
identified. Type 2 diabetes-associated biomarkers were mostly involved in mem-
brane transport systems. By contrast, control-enriched markers were frequently
involved in cell motility and metabolism of cofactors and vitamins. When studying
pathway levels, the gut microbiota of type 2 diabetic patients showed enrichment in
membrane transport of sugars, branched-chain amino acid (BCAA) transport, meth-
ane metabolism, xenobiotics degradation and metabolism, and sulfate reduction.
By contrast, there was a decrease in the level of bacterial chemotaxis, flagellar
assembly, butyrate biosynthesis, and metabolism of cofactors and vitamins. Seven
of the markers were also related to oxidative stress resistance suggesting that the gut
environment of type 2 diabetic patients stimulates bacterial defense mechanisms

against oxidative stress [19]. A further specific mathematical model for biomarker
identification has been developed and showed compositional and functional altera-
tions in the metagenomes of a specific cohort of 145 women with type 2 diabetes
[99]. Type 2 diabetic women were characterized by an increased abundance of four
Lactobacillus species and decreases in the abundance of five Clostridium species.
The total Lactobacillus species correlated positively with fasting glucose and
HbA1c (glycosylated hemoglobin), whereas the Clostridium species correlated
negatively with fasting glucose, HbA1c, insulin, C-peptide, and plasma triglycer-
ides. Importantly, impaired glucose tolerance could be identified on the basis of this
mathematical modeling. The impact of medication and hyperglycemia on the
metagenome was not considered as major confounding factors. Comparisons with a
Chinese cohort demonstrated that the biomarkers identifying type 2 diabetes were
different from the European population suggesting that the metagenome analysis to
predict type 2 diabetes should be specific for the age and geographical location of
the populations studied. It should be further noted that the mechanisms at the origin
of this observation are numerous and hence will most likely depend upon each
individual.
11.3.1.1 Impacts of Diet on Microbiota-Related Metabolic Diseases
Socio-demographic and environmental factors have a great influence on the

incidence of metabolic [100] and cardiovascular diseases [101], introducing the
possibility of identifying functional metagenomic factors under the control of envi-
ronmental factors such as stress, food habits, and sedentarity. Diet and nutritional
status are among the most important modifiable determinants of human health, and
gut bacteria feed on the nutrients absorbed during a meal leading to changes in
metabolism, the overall intestinal ecology, and the way bacteria interact with the
host. The first analyses of metagenomic sequencing have been performed on obese
patients followed up during 1 year of a restricted calorie diet [97, 98, 102]. A clear
metagenomic signature was identified in obese patients characterized by a reduction
in the relative abundance of the Bacteroidetes-to-Firmicutes ratio which represents
more than 80 % of the overall bacterial population in feces [97, 98] (Fig. 11.3). This
ratio evolved toward that of lean patients during weight loss showing that the micro-
biota can evolve according to the environmental factors. It is clear that dietary
manipulation, including HF feeding, profoundly alters the profile of the gut micro-
biota [27, 81, 103–107]. An enrichment in gram-negative to gram-positive bacteria
appeared to be associated with the early onset of high-fat-diet-induced diabetes [81].
An elegant validation of the role of diet on human microbiota has been performed
in germ-free mice colonized with human microbiota and fed with a fat-enriched
diet [104]. A single day of fat-enriched diet was sufficient to change the overall
ecological homeostasis within the gut microbiome, the corresponding metabolic
pathways, and hence the microbiome gene expression. The influence of high-fat diet
on gut microbiota has been validated in other species such as the pig [108, 109]
Intestinal functions and microbiota interplay
Fecalibacterium p.
Clostridium Bacterial
translocation
Change in diet Tissue microbiota
dysbiosis
TL
Firmicutes/ TLRs NLRs
Bacterioidetes Butyrate APC inflammasome Target tissues
Intestinal Cytokines
permeability
Change in Metabolic
microbiota Proteobacteria LPS+PGN endotoxemia
Ag
Blood
Lipid
Bile acid Proteobacteria
metabolism
Metabolism?
Intestinal
Akkermansiam. defenses,
defensins,
Genetic SFB mucus …
Fig. 11.3 Intestinal functions and microbiota interplay. Upon a change in diet or a specific genetic
background, the gut microbiota changes. A modification of the Firmicutes-to-Bacteroidetes ratio
leading to a change in polysaccharide fermentation, bile acid metabolism, an accumulation of
Proteobacteria (gram negative; releasing LPS, peptidoglycans, and other antigens), a reduced in
Faecalibacterium- and Clostridium-producing butyrate, a reduced in Akkermansia muciniphila,
and segmented filamentous bacteria (at least in rodents, SFB) would alter intestinal functions.
Intestinal permeability, defenses (mucin and defensin production), innate (APC, antigen-presenting
cells) and adaptive (T lymphocytes, TL) immunity, and lipid metabolism could lead to bacterial
translocation and tissue microbiota dysbiosis, metabolic endotoxemia, and metabolic inflamma-
tion of the targeted tissues. Altogether, the tissue biology would be impaired and be considered as
a risk factor for the development of cardiometabolic diseases
which is highly influenced by pre- and probiotics [110] and sucrose [111]. Defined
lipids such as oleic-acid-derived molecules and a combination of n-3 fatty acids
markedly increased total bacterial density and restored the proportions of clostridial
cluster, Enterobacteriales and Bifidobacterium that were changed during HFD
feeding [112] along with the metabolic status. Conversely saturated fat increased
the Firmicutes-to-Bacteroidetes ratio [113, 114]. In addition to the tremendous
impact of dietary fibers, other molecules such as polyphenols influence gut micro-
biota and further control metabolism [115, 116]. Type 2 diabetes has also its own
metagenomic signature, and type 2 diabetes-associated markers have been associ-
ated with the disease and can be useful to classify the different subgroups of type 2
diabetic patients [19]. Among these Akkermansia muciniphila seems to be a good
candidate with a recently identified role in the control of the intestinal barrier [107].
A reduction of Akkermansia muciniphila, which is involved in mucus degradation,
has also been observed [107]. These bacteria, when used as a probiotic, control
body weight gain [107] and its prevalence increased in response to prebiotics [117]
accompanying metabolic control through a mechanism that could involve gut

peptide secretion [118]. Insulin action also has a metagenomic signature in humans
[119]. Furthermore, hypotheses regarding the functional role of gut microbiota can
be generated by studying the genes characterizing the microbiome. A decrease in
the abundance of some universal butyrate-producing bacteria and an increase in
various opportunistic pathogens, as well as an enrichment of other microbial
functions conferring sulfate reduction and oxidative stress resistance, were also
characterized. A change in intestinal microbiota has also been characterized in non-
alcoholic fatty liver diseases (NAFLD) where the incidence is between 16 % and
30 % of the general population [120] and further rises toward 80 % in patients with
obesity and type 2 diabetes [121]. Some biomarkers have been identified such as a
reduction of Faecalibacterium and Anaerosporobacter and a higher abundance of
Parabacteroides, Allisonella, certain Lactobacillus species, and selected members
of the phylum Firmicutes (Lachnospiraceae; genera, Dorea, Robinsoniella, and
Roseburia) [122]. The causal role of gut microbiota on NAFLD has been shown by
microbiota transfer in the mice. The colonization of germ-free mice with the gut
microbiota from a high-fat-diet-induced NAFLD mouse induced the disease.
Sequencing of the 16S ribosomal RNA revealed differences at the phylum, genera,
and species levels [123]. Some mechanisms have been proposed that implicate
intestinal permeability, low-grade inflammation and immune balance in the devel-
opment of hepatosteatosis [124], the modulation of dietary choline and bile acid
metabolism, and the production of endogenous ethanol [125] that will be discussed
below. Another important feature of the change in intestinal microbiota is that in
addition to the change of phylum ratio obesity is associated with phylum-level
changes in the microbiota and reduced bacterial diversity [97, 98, 126].
11.3.1.2 Impact of the Host Genome
In addition to the role of the environment on the shaping of gut microbiota during
metabolic diseases, the impact of the host genome cannot be totally ruled out. The
concordance of type 2 diabetes within the homozygote population is higher than
between heterozygote twins [83, 84, 127] or in response to overfeeding [128, 129]
suggesting the important role of the genetic background. The results reveal that the
human gut microbiome is shared among family members but that each individual’s
gut microbial community varies in the specific bacterial lineages present, with a
comparable degree of covariation between adult monozygotic and dizygotic twin
pairs [126]. These results demonstrate that a diversity of organismal assemblages
can however yield a core microbiome at a gene level and that deviations from this
core are associated with different physiological states, for example, obese versus
lean. However, the concordance of the metabolic phenotype is not absolute between
twins – this could be linked to differences in gut microbiota since the adult mono-
zygotic twins are no more similar to one another in terms of their gut bacterial com-
munity structure than are adult dizygotic twins [126, 130]. However, the impact of
the host genome on the microbiota seems to depend on the microbiota location
considered. In the saliva twins resemble each other more closely than the whole
population at all time points but become less similar to each other when they age
and no longer cohabit [131]. The sequencing of gut microbiota from the general
population and across countries and ethnic origin showed that shared features of the
functional maturation of the gut microbiome are identified during the first 3 years of
life [24], suggesting an imprinting from the mother which remains but is not abso-
lute. Hence, room is available for the impact of environmental changes throughout
life. Numerous studies now report that the influence of the genetic background is due
to the impact of the immune system that shapes the microbial community [132–135].
Hence, the immunogenetic traits of an individual appear to be major regulators of
gut microbiota. Mutations in the receptors to bacterial determinants such as TLRs
which are pathogen-associated molecular pattern recognition receptors that recog-
nize highly conserved microbial molecules (PAMPs) notably TLR2 [136], TLR4
[137], and TLR5 [132] and NLRs notably NOD2 [138–140] shape the microbiota
although some controversies do exist [141].
11.3.1.3 Impact of Birth
Importantly, the host immune system to microbiota relationship continues to be

educated throughout life since the immune system matures along with the micro-
biota at birth. The gastrointestinal tract of a normal fetus is sterile. During the birth
process and rapidly thereafter, microbes from the mother and surrounding environ-
ment colonize the gastrointestinal tract of the infant until a dense, complex micro-
biota develops [9]. It is now clear that the composition and temporal patterns of the
microbial communities vary widely from baby to baby [142]. The distinct features
of each baby’s microbial community are recognizable for intervals of weeks to
months which show that each baby’s microbiota, until the first year of life, follows
an idiosyncratic law. However, it then converges toward a profile characteristic of
the adult gastrointestinal tract. Therefore, during the first year of life, while the
microbial ecology is not yet set up, the inheritance of gut microbiota could also be
influenced during pregnancy and at delivery according to the influence of the moth-
er’s microbiota. Furthermore, the mode of delivery influences the infant’s microbi-
ota since it was shown that vaginally delivered infants acquire bacterial communities
resembling their mother’s vaginal microbiota, dominated by Lactobacillus,
Prevotella, or Sneathia spp., whereas C-section infants harbored bacterial commu-
nities similar to those found on the skin surface, dominated by Staphylococcus,
Corynebacterium, and Propionibacterium spp. [11, 143]. However, the delivery
mode did not influence the prevalence of obesity in children [144]. Conversely,
antibiotic treatment during the first 6 months of life increased the risk of overweight
in children with normal-weight mothers but reduced the risk of obesity in children
with overweight mothers. Hence, at birth and during the first year of life, the early
colonization of the gut by microbiota can influence the incidence of metabolic dis-
eases [93]. Alterations in the microbiota composition of mothers may be transferred
to infants and lead to an increased risk of weight gain. During pregnancy, gut
microbiota changes dramatically between the first and third trimesters, with a vast
expansion of diversity among mothers, an overall increase in Proteobacteria and
Actinobacteria, and reduced richness [145]. This change in gut microbiota during
pregnancy was functional since colonization of germ-free mice with the microbiota
from the third trimester induced greater adiposity and insulin resistance when com-
pared to mice colonized with the microbiota from the first trimester [145]. This
original observation could suggest that the change in gut microbiota during preg-
nancy programs infant body weight. It has also been suggested that the microbiota
from obese mothers could influence the prevalence of obesity in children. This was
further supported by the fact that the infants’ fecal microbial composition was
related to the weight and weight gain of their mothers during pregnancy [146]. The
fecal Bacteroides and Staphylococcus concentrations were significantly higher in
infants of overweight mothers, whereas the prevalences of Akkermansia muciniph-
ila, Staphylococcus, and Clostridium difficile groups were lower in infants of
normal-weight mothers and of mothers with normal weight gain during pregnancy.
This was confirmed in that the concentration of the gram-negative family
Enterobacteriaceae was significantly higher in obese/overweight children and the
levels of Desulfovibrio and Akkermansia muciniphila-like bacteria were signifi-
cantly lower when compared to lean controls [147]. Chapter 8 “Metabonomics in
neonatal and pediatric research: Studying and modulating gut functional ecology
for optimal growth and development” provides a comprehensive overview of the
state of the art of metabonomics and gut microbiota studies in neonatal and pediatric
research.
11.3.2 Molecular Crosstalk Between Gut Microbiota

and the Host for the Control of Metabolic Diseases
11.3.2.1 The Lipopolysaccharides Hypothesis and Metabolic
Inflammation
Whereas extensive analyses demonstrate the important impact of gut microbiota on

host biology, the mechanisms of the crosstalk between the host and the microbiota
remain to be delineated. As mentioned above, metabolic inflammation is a leading
mechanism responsible for the impairment of glycemia and body weight regulation.
One hypothesis involves bacterial factors from the gut, such as lipopolysaccharides
(LPS). Lipopolysaccharides are components of the wall from gram-negative bacte-
ria [1]. They are potent endotoxins, involved in the acute-phase response to bacterial
infection, inducing a cytokine-mediated systemic inflammatory response that can
cause shock and severe multiple organ failure [2, 3]. These bacterial antigens bind
to their receptors TLR4 and CD14 on numerous cells types notably those of the
immune system, i.e., macrophages and dendritic cells. Adipocytes also express
TLR4 and can bind LPS and could be involved directly in the activation of intracel-
lular inflammatory pathways [148, 149].
Mechanisms of High-Fat-Diet-Induced Blood Bacterial Molecular Patterns
LPS accumulates in blood and contributes to inflammation and insulin intolerance

[6, 81]. A 1-month high-fat feeding in humans increased endotoxemia by 71 %
[150], suggesting that therapeutic agents that reduce intestinal LPS permeability
could control metabolic endotoxemia and hence systemic inflammation in patients
with metabolic syndrome. Similarly, in type 1 diabetic patients, metabolic endotox-
emia was associated with dyslipidemia, insulin resistance, obesity, and chronic
inflammation [151]. The mechanism is associated with an increase of the gram-
negative to gram-positive ratio within the intestinal microbiota [81]. LPS need to be
transported across the intestinal epithelial barrier to reach the blood. A mechanism
involving the synthesis and production of chylomicron is required [152]. The uptake
of LPS is observed within hours of lipid absorption in human [153] and could hence
be considered as a blood nutritional signal informing the body of a change in feed-
ing behavior and in microbiota ecology. The accumulation of LPS in the blood is
also linked to an increased intestinal permeability where tight junctions become
leaky due to a reduction of the expression of specific proteins such as zonula
occludens [81, 106] and the putative role of endocannabinoids [154] that could con-
trol intestinal permeability through a GLP-2-dependent mechanism [155]. This is
reversed upon treatment with prebiotics [156, 157] or probiotics [107]. The increase
of plasma LPS concentration is called metabolic endotoxemia. Lipoproteins bind
LPS to attenuate the biological inflammatory response of this bacterial factor [158,
159] through all classes of lipoprotein (chylomicrons, VLDL, LDL, and HDL) [9].
Chylomicrons and VLDL have been shown to reduce LPS-induced toxicity in mice
[10]. Similarly, the binding of LPS to LDL reduced endothelial cell activity [9, 11].
In one study performed in 10 individuals, HDL has been shown to be the main LPS
carrier holding 60 % of the LPS, with LDL and VLDL carrying, respectively, 25 %
and 12 % [14]. Among three ex vivo studies, two have indicated that LPS, in
humans, was mainly located in VLDL and LDL [15, 16] and one has more precisely
indicated that LDL was the major carrier for LPS [9]. Therefore, it is now evident
that nutritional and lipid metabolism are tightly linked to the intestinal absorption of
gut bacterial factors including LPS for the triggering of metabolic inflammation.
The pharmacological control of intestinal LPS absorption could be of importance
for the prevention or treatment of metabolic diseases. A change in nutrition involv-
ing an increase in fat content is a risk factor for metabolic endotoxemia as demon-
strated in epidemiological studies in humans [151, 160, 161]. In the mouse the
ablation of various pattern recognition receptors (PRRs) such as TLR4, CD14, and
NLRP3 protects mice from diet-/obesity-induced inflammation and insulin resis-
tance [6, 162–164]. In human myotubes, LPS increased JNK phosphorylation and
MCP-1 and IL-6 gene expression [165] and could play a role in the pathogenesis of
insulin resistance. Therefore, antagonists of CD14/TLR4 may improve insulin
action in type 2 diabetic patients. In the mouse the chronic subcutaneous infusion of
LPS at low rates favors liver, adipose tissue, and muscle cytokine production and
macrophage accumulation showing that in vivo metabolic endotoxemia is an impor-
tant risk factor of insulin resistance and hence metabolic diseases. Direct evidence
shows that the LPS to ligand crosstalk controls adipose tissue inflammation [6, 162,
164, 166], insulin sensitivity [148, 167], and obesity [6, 168] through molecular
mechanisms most likely linked to the activation of NFkB and the transcription of
genes coding for cytokines. The latter, such as TNFα, will then interfere with the
insulin receptor leading to insulin resistance [169]. LPS are also involved in the
reduction of adipogenesis [170], as well as hepatic steatosis [171, 172] again
through mechanisms involving TNFα production or SREBP-1 activation [173].
Other bacterial components are involved in the inflammatory process characterizing
metabolic inflammation. The peptidoglycan is a complex structure of the bacterial
wall found mostly in gram-positive bacteria but also found, to a lower extent, in
gram-negative bacteria. Its polysaccharide component is different between the two
types of bacteria and binds to pattern recognition receptors such the nucleotide
oligomerization domain (NOD)1 and NOD2. NOD1 and NOD2 are currently the
only known sensors of bacterial cell wall peptidoglycan (PGN) that elicit inflamma-
tion by increasing cytokine production, defensin expression by Paneth cells [174–
177], and stress kinase responses [178]. In addition to LPS, PGN could be positioned
as a component of metabolic endotoxemia that contributes to inflammation and
metabolic defects [27, 179]. This is important because gut-derived LPS alone
appears to be insufficient for establishing glucose/insulin tolerance in poorly
immune-responding germ-free mice [180]. Furthermore, PGN containing meso-
DAP motifs (generally dominant in gram-negative bacteria) caused profound insu-
lin resistance through actions on NOD1 directly in metabolic cells, including
adipocytes [179]. The NOD2 activation with the minimal bioactive PGN motif,
muramyl dipeptide (MDP), mostly present in gram-positive bacteria, elicited cell
autonomous inflammation and impaired insulin action directly in muscle cells [181]
and caused acute, peripheral insulin resistance in vivo [179]. Mice lacking NOD2
have improved insulin sensitivity during obesity illustrating the protective role of
NOD2 as a sensor of bacterial motifs derived from the gut microbiota on the control
of metabolic diseases [27, 179]. However, defects in NOD2 immunity have been
associated with promoting other chronic proinflammatory pathologies, and human
NOD2 variants have the highest risk associated with Crohn’s disease [182–184].
NOD2 immunity is known to contribute to homeostasis of the gut microbiota dem-
onstrating that the microbiota is in a tight relationship with the immune system that
secondarily controls the metabolism [138].
Metabolic Endotoxemia and Gut Microbiota Induced Hepatic Steatosis
Possible mechanisms leading to hepatosteatosis in obese and type 2 diabetic patients

that involve gut microbiota implicate intestinal permeability, low-grade inflamma-
tion and immune balance [124], the modulation of dietary choline and bile acid
metabolism, and the production of endogenous ethanol [125]. A first hypothesis is
proposed regarding the key role of the intestinal epithelium as a barrier between the
environment, i.e., the luminal side of the intestine and the body. A leaky gut has
been described in patients with NAFLD [185] and linked to a change in intestinal
microbiota [122, 186]. A second hypothesis is linked to the role of choline which is
a major phospholipid component of the cell membrane involved in signal transduc-
tion and the control of lipoprotein metabolism, notably the very-low-lipoprotein
assembly [187–189]. Therefore, a choline-deficient diet promotes liver steatosis.
Such depletion could be induced by the gut microbiota that would convert dietary
choline into toxic methylamines [190, 191]. It was shown that circulating levels of
plasma phosphatidylcholine were low, whereas urinary excretion rates of methyl-
amines dimethylamine, trimethylamine, and trimethylamine-N-oxide were high and
were coprocessed by symbiotic gut microbiota and mammalian enzyme systems.
An inflammatory feature is required to qualify hepatic steatosis (NASH) from a
NAFLD phenotype. It could originate from the gut microbiota since it interacts with
the overall immunity of the host via PAMPs TLRs and NLRs [175]. Similarly, the
role of components from the inflammasome such as NLRP6 and NLRP3 has been
demonstrated [124]. These inflammation sensors could modulate the gut microbiota
and hence the crosstalk of host immunity and gut microbiota leading to the develop-
ment of the pathology. Altogether, the triggering of innate immunity could lead to
the production of cytokines which contribute to the inflammatory phenotype of
NASH through a mechanism similar to that suggested for the development of type
2 diabetes and obesity.
11.3.2.2 The Energy Harvesting, Expenditure, and Short-Chain Fatty

Acid Hypotheses
The change in intestinal microbiota (increased Firmicutes and decreased

Bacteroidetes), observed during obesity, was associated with an increased efficiency
of energy harvest in human and mouse [26, 105]. The fecal content in energy was
lower in obese mice than in controls by about 150 kCal/day, suggesting that the
energy was better absorbed by the gut. This observation was attributed to the exces-
sive hydrolysis of polysaccharides into monosaccharides indispensable to the body
as well as to the production of short-chain fatty acids (SCFA) [26]. However, it
seems that the production of SCFA diminished over time during long-term high-fat
feeding or aging in ob/ob mice [105]. The role of gut microbiota in the production
of SCFA is however strongly supported since it was shown that germ-free mice are
devoid of SCFAs [192] and could control AMP-activated protein kinase activity and
macrophage infiltration in adipose tissue [193]. Human colonic butyrate producers
are phylogenetically diverse, with the two most abundant groups related to E.
rectale/Roseburia species and to Faecalibacterium prausnitzii. The precise SCFA
responsible for the control of metabolism seems to be multiple with a significant
role for acetate. In addition, SCFAs such as propionate can be used for de novo
glucose or lipid synthesis and serve as an energy source for the host. It has also been
demonstrated that butyrate lowers fatty acid content in liver and plasma, reduces
food intake, exerts immunosuppressive actions, and probably improves tissue insu-
lin sensitivity [194, 195]. The mechanism could involve notably the promotion of
glucagon-like peptide 1 (GLP-1) secretion via the binding of SCFA to
G-protein-coupled receptors such as GPR41/43 at the surface of the enteroendocrine

cells [195]. Cascades of events such as the control of glucagon and insulin secretion,
gastric emptying, and satiety would secondary to GLP-1 secretion control glucose
homeostasis. Butyrate is also involved in the improvement of the overall glucose
metabolism as demonstrated by its supplementation to a high-fat diet [196]. Other
experiments show that by high-pressure liquid chromatography analysis, authors
identified significantly higher concentrations of butyrate and propionate in feces
from obese versus normal-weight children [197]. Significantly lower concentrations
of intermediate metabolites were detected in obese children suggesting exhaustive
substrate utilization by obese gut microbiota [197]. Despite this evidence, the role
of SCFA on metabolism still remains unclear. Recent data demonstrate that the
activation of GPR43 by SCFA at the surface of adipocytes reduced insulin signaling
[198]. This could certainly reduce insulin-induced lipid accumulation, and hence
body weight gain, but could also be considered as a factor inducing adipose tissue
insulin resistance which could prevent glucose to be taken up by adipocytes and
hence accumulate in the blood or the liver to induce hyperglycemia and hepatic
steatosis. Thus, data relating to the production of SCFA most likely cover other
unidentified mechanisms that need to be identified to fully understand the benefit of
polysaccharide fermentation.
The role of gut microbiota could also be on energy expenditure since germ-free
mice are leaner despite a dramatically increased food intake [90]. The lean pheno-
type of these mice is associated with increased skeletal muscle and liver levels of
phosphorylated AMP-activated protein kinase (AMPK) and its downstream targets
involved in fatty acid oxidation such as acetylCoA carboxylase and carnitine palmi-
toyltransferase [199]. AMPk is a master switch considered to be a molecule recruited
in case of stress and energy deprivation [200, 201]. Bacterial factors could activate
this enzyme in muscles and the liver to increase energy expenditure. Conversely,
the microbiota inhibits fasting-induced adipose factor (Fiaf) which downregulates
circulating lipoprotein lipase. Hence, free fatty acids are stored in the liver of con-
ventional mice which is not the case in germ-free animals [199].
Altogether, numerous eukaryotic targets are currently being identified using
germ-free mice. An important matter will be to validate them in physiological con-
ditions such as in conventional mice and in humans.
11.3.3 Gut Microbiota During Obesity Surgery
The last 40 years has seen the emergence of the treatment of obesity by surgery.
Several types of surgery involving the stomach and the intestine are used. The
Roux-en-Y gastric bypass (RYGB) surgery which is to date the major bariatric
intervention to treat morbid obesity involves the direct connection of the jejunum to
the stomach so that nutrients are no longer in contact with the duodenum. In addi-
tion, 90 % of the stomach is removed. The sleeve gastrectomy involves the removal
of a major part of the stomach so that food directly reaches the duodenum. Other
types of surgery such as the ileal bypass transposition are being developed. All sur-
gical procedures dramatically affect gut microbiota [202, 203]. A first set of experi-
ments demonstrated that before surgery, Firmicutes were dominant in normal-weight
and obese individuals but significantly decreased in post-gastric bypass [204].
Interestingly, the gut microbiota from patients after surgery was not similar to that
of lean individuals since it was enriched in Gammaproteobacteria which are H(2)-
producing Prevotellaceae [204]. These changes were independent of weight change
and caloric restriction, were detectable throughout the length of the gastrointestinal
tract, and were most evident in the distal gut, downstream of the surgical manipula-
tion site [205]. A precise study identified 14 discriminant bacterial genera (7 were
dominant and 7 were subdominant) and 202 genes changed in the white adipose
tissue that correlated with RYGB as well as with both clinical phenotypes [206].
Faecalibacterium prausnitzii species was lower in patients with metabolic diseases
but associated negatively with inflammatory markers even throughout the follow-up
after surgery and independently with changes in food intake suggesting an imprint-
ing of the microbiota that cannot be easily changed [207].
11.3.3.1 The Bile Acids to Microbiota Hypothesis
Intestinal microbiota also has a role in the metabolism of bile acids, which, with
other sterols [208], are important regulators of metabolic diseases, as also described
in Chaps. 13 and 14. Bile acid can control insulin secretion [209], GLP-1 secretion
[210], energy expenditure [211], as well as atherosclerosis [212]. The role of bile
acid is also suspected during bariatric surgery and could be the molecular link
between the change in intestinal microbiota [213] and the improvement of metabo-
lism [214] notably through the TGR5 receptor [215]. Bile acids are secreted into
the duodenum and work to emulsify liposoluble dietary nutrients to facilitate their
digestion and absorption. Studies have demonstrated that bile acid composition
and secretion in response to fat intake modifies markedly gut microbiota thereby
inducing a dysbiosis [216]. The dietary fat can alter the gut microbiota of mice
indirectly by changing the animals’ pool of bile acids and steroids that are pro-
duced by the liver and secreted into the intestine [216]. Bile acids are synthesized
from cholesterol in the liver and further metabolized by the gut microbiota into
secondary bile acids [217] which can change the metabolism by acting on the
farnesoid X receptor to exert some negative feedback control. When compared to
germ-free mice, the conventionally raised mice are characterized by a dramatic
reduction in muricholic acid, but not cholic acid levels [218]. The microbiota con-
trols fibroblast growth factor 15 in the ileum and cholesterol 7α-hydroxylase
(CYP7A1) in the liver by FXR-dependent mechanisms thereby influencing the
metabolism and secretion of bile acids.
Hence, a new ecology is expected from the host and the microbiota through bile
acid metabolism.
11.3.3.2 The Immunomicrobiota Crosstalk and Metabolic Diseases
Metabolic diseases are now considered as low-grade immunomodulatory diseases

[58–60]. The origin of the antigen is unknown but has been suspected to be from
autoantigens, which notably form the adipose tissue [219–221]. The discovery of
intestinal microbiota opens new avenues regarding the origin of the metabolic
inflammation. This hypothesis stemmed from the observation that intestinal perme-
ability was increased in high-fat-diet-fed mice, leading to accumulation in the blood
of bacterial fragments such as LPS [6, 160]. Whereas, during diabetes and obesity,
LPS and peptidoglycan were shown to be transported from the intestinal lumen to
the blood through the intestinal epithelium, it was initially thought that whole com-
mensal bacteria would be arrested by the mucosal layer and the immune system
within Peyer’s patches or the lamina propria [222, 223], thus preventing transloca-
tion across the intestinal epithelial layer [224, 225].
A deficiency in host immune defenses and increased permeability and damage to
the intestinal mucosal barrier represent mechanisms through which bacteria of
intestinal origin accumulate into the tissues. In the healthy situation macrophages,
dendritic cells, and antibodies within the lumen restrain, along with the mucus layer,
the bacteria within the luminal side of the gut to ensure a tight intestinal permeabil-
ity. In the proximal intestinal segments where the microbiota is sparse and the
mucus layer is thin or absent, commensal bacteria are in close contact with the
epithelial cells, and host immunosuppression synergistically promotes bacterial
translocation from the gastrointestinal tract resulting in accumulation of bacteria
within the mesenteric lymph node [225]. Therefore, lymphocytes from mesenteric
lymph nodes, Peyer’s patches, intraepithelial cells, and the lamina propria are spe-
cifically educated to recognize commensal bacteria, which are therefore considered
as self-antigens. The change of intestinal microbiota that occurs during a fat-
enriched diet [81, 104] leads to the production of new antigens that are no longer
recognized as self by the intestinal immune cells [134]. This change in gut micro-
biota can also be controlled by natural antibiotics such as defensins secreted by
Paneth cells [226, 227].
11.3.3.3 Intestinal Bacterial Translocation and Metabolic Diseases
Bacterial translocation is defined as the passage of viable indigenous bacteria from

the gastrointestinal tract to extra intestinal sites, such as the mesenteric-lymph-node
complex, liver, spleen, and bloodstream [224]. This mechanism is largely observed
during intestinal bacteria overgrowth leading to cirrhosis [228] or sepsis [229].
Although it is intuitively considered as deleterious for the organism, it could be sug-
gested that the bacterial translocation helps the immune system to be prepared
against infections. Another pathological situation of increased bacterial transloca-
tion is AIDS where, due to impaired intestinal immunosuppression, bacteria trans-
locate toward tissues and could lead to inflammation and increased mortality [230].
The immune inhibitory receptor programmed death-1 (PD-1) regulates the function
of CD8+ cells and the translocation of bacteria [230]. Importantly AIDS is associated
with a dramatic increase in the incidence of diabetes and lipodystrophy [231, 232].
The CD4 T helper and Th17 cells appear to be critical for regulating gut mucosal
immune responses to extracellular microbial pathogens and therefore could be
involved in bacterial translocation [233]. Hence, an impaired intestinal immune sys-
tem could lead to bacterial translocation and therefore to metabolic diseases. This
hypothesis has been validated during high-fat diet-induced metabolic diseases [27].
First it was observed that a rapid augmentation of bacterial adherence to the intesti-
nal epithelium layer occurred suggesting an impaired mucosal defense. Then CD11c
positive phagocytes harvested the transepithelial bacteria and translocated with the
live bacteria to adipose tissue. The bacteria were co-localized within the adipose
depot with CD11c positive cells and most likely were intracellular. Importantly, the
bacterial accumulation was considered as a predictive biomarker of type 2 diabetes
[234]. The translocated bacteria were also detected in the blood in humans and con-
versely to the adipose tissue. The Proteobacteria phylum represented more than
80 % of the blood microbiota as assessed by 16S RNA DNA. Interestingly, other
bacterial DNA fragments could predict the onset of cardiovascular events in a large
cohort of type 2 diabetic patients [235]. The processes involved in bacterial translo-
cation at the onset of type 2 diabetes could be related to molecular determinants
involved in bacterial recognition. The NOD1 and LPS-CD14 bacterial receptors
were involved in this translocation mechanism since their deletion dramatically
reduced the amount of bacterial DNA present within the adipose tissue under high-
fat diet [27]. Interestingly, both deletions were associated with improved insulin
sensitivity and reduced glycemia suggesting that the tight control of bacterial trans-
location could be a master regulator of the onset of insulin resistance and diabetes.
Conversely, the deletion of NOD2 induced over-accumulation of bacterial DNA
within the tissues suggesting a protective effect. This was also observed in mice
treated with NOD2 ligands [179]. The importance of the immune system in the
translocation mechanism was also illustrated in mice carrying a deleted MyD88
gene and which were prone to diabetes [236]. These mice also displayed a dramatic
accumulation of bacteria in the mesenteric lymph nodes and adipose tissue.
Therefore, metabolic regulators of bacterial translocation could be linked to hor-
mones controlling the immune system such as estrogens [237] or leptin [238]. This
shift in the paradigm is supported by data which show that cardiovascular disease is
associated with the role of microbiota in the control of lipid metabolism leading to
the development of atherosclerosis [95, 239, 240]. Surprisingly, microbes associ-
ated with periodontitis were at the origin of the bacteria present in the plaques of
atherosclerotic patients [240]. Studies in animals have revealed a mechanistic link
between intestinal microbial metabolism of the choline moiety in dietary phospha-
tidylcholine (lecithin) and coronary artery disease through the production of a pro-
atherosclerotic metabolite, trimethylamine-N-oxide (TMAO) [241, 242]. This has
been confirmed in humans [243]. Importantly, antibiotic administration reduced the
TMAO concentration suggesting that the microbiota was indeed the source of this
proatherogenic molecule. Furthermore, in addition to bacterial factors, whole bac-

teria have been identified within the atherosclerotic plaques. The taxa have been
linked to periodontal diseases [240]. This finding bridges the gap between metabolic
and cardiovascular diseases by means of opportunistic bacteria from the oral cavity.
Its proportion increases in response to a change of diet [244]. In addition, genes
closely involved in the regulation of cardiovascular diseases such as angiotensin
I-converting enzyme (peptidyl-dipeptidase A) 2 (Ace2), which encodes a key regu-
latory enzyme of the renin-angiotensin system, also have an impact on gut micro-
biota leading to inflammation [245]. Furthermore, statins, classical anti-dyslipidemic
agents prescribed to reduce the incidence of cardiovascular events, have been shown
to be associated with microbial-derived agents that could be responsible for the
cholesterol-lowering effect, explaining the efficacy of the treatment [246].
Cholesterol metabolism is indeed regulated by gut microbiota since there is a dra-
matic reduction of muricholic acid, a farnesoid X receptor (FXR) antagonist in the
ileum, in conventional mice when compared with germ-free mice. Therefore, a
given microbiota could be important for the efficacy of a drug treatment as well as
for the susceptibility of developing cardiometabolic diseases, thus reconciling met-
abolic and cardiovascular diseases under the paradigm of the gut to tissue microbi-
ota crosstalk.
11.3.4 Tissue Microbiota and Metabolic Diseases:

The Paradigm Shift of Bacteria Translocation
Bacterial factors such as LPS, peptidoglycans, and bacterial DNA can be absorbed
by the gut and found within the blood. These metafactors could serve as biomark-
ers and also as regulators since they can then activate cells from the immune
system to generate inflammation. Pyrosequencing of the 16S RNA DNA from the
stroma vascular fraction of human adipose tissue from lean, overweight, and
obese patients showed that the diversity of the tissue microbiota was mainly
related to Firmicutes and Proteobacteria (in press). Interestingly, although no
changes were observed among the Firmicutes phylum, a major dysbiosis was
detected within the Proteobacteria phylum and specifically the Ralstonia genus.
Ralstonia was dramatically overrepresented within the family and, furthermore,
its presence was proportional to increased BMI. This suggested that this bacterial
genus might be causal. The bacterial translocation process also leads to the accu-
mulation of bacterial DNA in the blood fraction [234], suggesting that either bac-
teria migrate through the blood to reach the tissues or are released by the tissues
into the blood.
The role of this tissue microbiota is not known; similarly the proportion of live
bacteria compared to dead bacteria or fragments is not perfectly identified and
will most likely depend on diet, age, and other genetic and environmental factors.
This paradigm shift is promising but will require much work to determine its
physiological role.
11.3.4.1 Tissue Microbiota as Biomarkers of Metabolic

and Cardiovascular Diseases
The blood tissue microbiota has been recently described in humans and mice [27,
234, 235]. The presence of bacterial DNA has been revealed by qPCR from human
cohorts from the general population [234] and from diabetic patients [235]. It could
be shown that the increased 16S RNA DNA concentration in blood predicts the
onset of diabetes 6–9 years later [234], whereas the quantification of other bacterial
factors could predict the onset of cardiovascular events in a population of type 2
diabetics [235]. In both cases it is suggested that the bacterial fragments which
accumulate in the blood could be causal of the disease. Since the bacteria are inher-
ited at birth, the accumulation of bacterial DNA initiates at a very early stage, link-
ing the environment with the host genome. In risk situations, such as when feeding
on a fat-enriched diet, the bacterial DNA increases reflecting the new nutritional
situation and could be involved in the triggering of metabolic adaptation. Adipose
tissue is targeted [6, 27, 137] and adipogenesis is induced [170]. Hence, these
bacteria can be considered as nutrient sensors informing the tissues.
11.4 Therapeutic Perspectives
The therapeutic perspectives are huge but so far at their infancy, as presented and
discussed in Chap. 19. One should separate the nutritional approaches aiming at
maintaining health in the general population. This would first involve phenotyping
of the human microbiome [21] and second performing studies to demonstrate the
prevention of the risk of disease development. A second strategy would involve
preventing the disease in patients at risk to develop metabolic diseases. In these
patients subpopulations should be defined and studied in order to perfectly adapt the
nutritional strategy. This should be a precisely directed strategy to treat patients for
the prevention of developing a metabolic disease. This would involve overweight
individuals, smokers, sedentary individuals, and hypertensive patients. With this
aim pre- and probiotics selected for their capacity to influence all the above physi-
ological mechanisms at the cross road of microbiota and the host should be delin-
eated. Other nutritional approaches could be pursued aiming to target intestinal
functions [115] by using antioxidants like polyphenols such as resveratrol. From a
therapeutic point of view, pharmacological strategies can be envisaged that involve
small molecules to target molecular mechanisms such as the immune system, muco-
sal defense, bile acid synthesis, incretin secretion, or the production of short-chain
fatty acids. Again, they should address well-defined subgroups of patients charac-
terized by their specific microbiota or blood biomarkers. Companion blood bacte-
rial biomarkers could be used to monitor the impact of the intervention over a
long-term treatment. This reasoned strategy should help to control metabolic
diseases and associated cardiovascular events. The advent of next-generation
sequencing strategies along with the development of bioinformatics and biostatistical
skills are now available to set up these programs.
11.5 Conclusions
In the quest of a mechanism explaining the pandemic development of metabolic

diseases, the consequent cardiovascular events, and further reconciling the key role
of the adaptive and innate immune system, the role of the intestinal microbiota has
emerged as a very promising candidate. The diversity and huge complexity of the
microbiome precludes a rapid and clear identification of the molecular mechanisms
at the crosstalk between the host and the microbiota. The molecular hypotheses for
the explanation of the metabolic phenotype are numerous and probably related to the
large number of molecular origins of the disease. This suggests that patients should
be screened for their microbiota to host crosstalk. Thus, therapeutic strategies or
preventive programs could be successful. We have entered a new era and one can no
longer disregard the microbiome from the gut or other locations since it is now
clearly involved at the onset and during the development of metabolic disease.
References
1. Haffner SM, Ruilope L, Dahlof B, Abadie E, Kupfer S, Zannad F. Metabolic syndrome, new
onset diabetes, and new end points in cardiovascular trials. J Cardiovasc Pharmacol.
2006;47(3):469–75.
2. Juonala M, Magnussen CG, Berenson GS, Venn A, Burns TL, Sabin MA, et al. Childhood
adiposity, adult adiposity, and cardiovascular risk factors. N Engl J Med.
2011;365(20):1876–85.
3. Franks PW, Hanson RL, Knowler WC, Sievers ML, Bennett PH, Looker HC. Childhood
obesity, other cardiovascular risk factors, and premature death. N Engl J Med.
2010;362(6):485–93.
4. Virmani R, Kolodgie FD, Burke AP, Farb A, Schwartz SM. Lessons from sudden coronary
death: a comprehensive morphological classification scheme for atherosclerotic lesions.
Arterioscler Thromb Vasc Biol. 2000;20(5):1262–75.
5. Danaei G, Lawes CM, Vander Hoorn S, Murray CJ, Ezzati M. Global and regional mortality
from ischaemic heart disease and stroke attributable to higher-than-optimum blood glucose
concentration: comparative risk assessment. Lancet. 2006;368(9548):1651–9.
6. Cani PD, Amar J, Iglesias MA, Poggi M, Knauf C, Bastelica D, et al. Metabolic endotoxemia
initiates obesity and insulin resistance. Diabetes. 2007;56(7):1761–72.
7. Ford ES, Li C, Sattar N. Metabolic syndrome and incident diabetes: current state of the
evidence. Diabetes Care. 2008;31(9):1898–904.
8. Dethlefsen L, Eckburg PB, Bik EM, Relman DA. Assembly of the human intestinal micro-
biota. Trends Ecol Evol. 2006;21(9):517–23.
9. Mackie RI, Sghir A, Gaskins HR. Developmental microbial ecology of the neonatal gastroin-
testinal tract. Am J Clin Nutr. 1999;69(5):1035S–45.
10. Penders J, Thijs C, Vink C, Stelma FF, Snijders B, Kummeling I, et al. Factors influencing the
composition of the intestinal microbiota in early infancy. Pediatrics. 2006;118(2):511–21.
11. Azad MB, Konya T, Maughan H, Guttman DS, Field CJ, Chari RS, et al. Gut microbiota of
healthy Canadian infants: profiles by mode of delivery and infant diet at 4 months. Can Med
Assoc J. 2013.
12. Barrett E, Guinane CM, Ryan CA, Dempsey EM, Murphy BP, O’Toole PW, et al. Microbiota
diversity and stability of the preterm neonatal ileum and colon of two infants.
Microbiologyopen. 2013;2:215.
13. Gronlund MM, Lehtonen OP, Eerola E, Kero P. Fecal microflora in healthy infants born by
different methods of delivery: permanent changes in intestinal flora after cesarean delivery.
J Pediatr Gastroenterol Nutr. 1999;28(1):19–25.
14. Orrhage K, Nord CE. Factors controlling the bacterial colonization of the intestine in breast-
fed infants. Acta Paediatr Suppl. 1999;88(430):47–57.
15. Stratiki Z, Costalos C, Sevastiadou S, Kastanidou O, Skouroliakou M, Giakoumatou A, et al.
The effect of a bifidobacter supplemented bovine milk on intestinal permeability of preterm
infants. Early Hum Dev. 2007;83(9):575–9.
16. Claesson MJ, Cusack S, O’Sullivan O, Greene-Diniz R, de Weerd H, Flannery E, et al.
Composition, variability, and temporal stability of the intestinal microbiota of the elderly.
Proc Natl Acad Sci U S A. 2011;108 Suppl 1:4586–91.
17. Maslowski KM, Mackay CR. Diet, gut microbiota and immune responses. Nat Immunol.
2011;12(1):5–9.
19. Qin J, Li Y, Cai Z, Li S, Zhu J, Zhang F, et al. A metagenome-wide association study of gut
microbiota in type 2 diabetes. Nature. 2012;490(7418):55–60.
20. Dethlefsen L, McFall-Ngai M, Relman DA. An ecological and evolutionary perspective on
human-microbe mutualism and disease. Nature. 2007;449(7164):811–8.
21. Turnbaugh PJ, Ley RE, Hamady M, Fraser-Liggett CM, Knight R, Gordon JI. The human
microbiome project. Nature. 2007;449(7164):804–10.
22. Zoetendal EG, Akkermans AD, De Vos WM. Temperature gradient gel electrophoresis analy-
sis of 16S rRNA from human fecal samples reveals stable and host-specific communities of
active bacteria. Appl Environ Microbiol. 1998;64(10):3854–9.
23. Biagi E, Candela M, Fairweather-Tait S, Franceschi C, Brigidi P. Aging of the human meta-
organism: the microbial counterpart. Age (Dordr). 2012;34(1):247–67.
24. Yatsunenko T, Rey FE, Manary MJ, Trehan I, Dominguez-Bello MG, Contreras M, et al.
Human gut microbiome viewed across age and geography. Nature. 2012;486(7402):222–7.
25. Ley RE, Peterson DA, Gordon JI. Ecological and evolutionary forces shaping microbial
diversity in the human intestine. Cell. 2006;124(4):837–48.
26. Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V, Mardis ER, Gordon JI. An obesity-associated
gut microbiome with increased capacity for energy harvest. Nature. 2006;444(7122):1027–31.
27. Amar J, Chabo C, Waget A, Klopp P, Vachoux C, Bermudez-Humaran LG, et al. Intestinal
mucosal adherence and translocation of commensal bacteria at the early onset of type 2 dia-
betes: molecular mechanisms and probiotic treatment. EMBO Mol Med. 2011;3(9):559–72.
28. Chen X, D’Souza R, Hong ST. The role of gut microbiota in the gut-brain axis: current chal-
lenges and perspectives. Protein Cell. 2013;4:403.
29. Lebovitz HE. The relationship of obesity to the metabolic syndrome. Int J Clin Pract Suppl.
2003;134:18–27.
30. Chen G, Koyama K, Yuan X, Lee Y, Zhou YT, O’Doherty R, et al. Disappearance of body
fat in normal rats induced by adenovirus-mediated leptin gene therapy. Proc Natl Acad Sci
U S A. 1996;93(25):14795–9.
31. Shimabukuro M, Koyama K, Chen G, Wang MY, Trieu F, Lee Y, et al. Direct antidiabetic
effect of leptin through triglyceride depletion of tissues [see comments]. Proc Natl Acad Sci
U S A. 1997;94(9):4637–41.
32. Zhou YT, Shimabukuro M, Koyama K, Lee Y, Wang MY, Trieu F, et al. Induction by leptin of
uncoupling protein-2 and enzymes of fatty acid oxidation. Proc Natl Acad Sci U S A.
1997;94(12):6386–90.
33. Shepherd PR, Gnudi L, Tozzo E, Yang H, Leach F, Kahn BB. Adipose cell hyperplasia and
enhanced glucose disposal in transgenic mice overexpressing GLUT4 selectively in adipose
tissue. J Biol Chem. 1993;268(30):22243–6.
34. Casteilla L, Planat-Benard V, Cousin B, Silvestre JS, Laharrague P, Charriere G, et al.
Plasticity of adipose tissue: a promising therapeutic avenue in the treatment of cardiovascular
and blood diseases? Arch Mal Coeur Vaiss. 2005;98(9):922–6.
35. Caspar-Bauguil S, Cousin B, Galinier A, Segafredo C, Nibbelink M, Andre M, et al. Adipose

tissues as an ancestral immune organ: site-specific change in obesity. FEBS Lett.
2005;579(17):3487–92.
36. Galinier A, Carriere A, Fernandez Y, Carpene C, Andre M, Caspar-Bauguil S, et al. Adipose
tissue proadipogenic redox changes in obesity. J Biol Chem. 2006;281(18):12682–7.
37. Sengenes C, Lolmede K, Zakaroff-Girard A, Busse R, Bouloumie A. Preadipocytes in the
human subcutaneous adipose tissue display distinct features from the adult mesenchymal and
hematopoietic stem cells. J Cell Physiol. 2005;205(1):114–22.
38. Maumus M, Peyrafitte JA, D’Angelo R, Fournier-Wirth C, Bouloumie A, Casteilla L, et al.
Native human adipose stromal cells: localization, morphology and phenotype. Int J Obes
(Lond). 2011;35:1141.
39. Planat-Benard V, Silvestre JS, Cousin B, Andre M, Nibbelink M, Tamarat R, et al. Plasticity
of human adipose lineage cells toward endothelial cells: physiological and therapeutic per-
spectives. Circulation. 2004;109(5):656–63.
40. Villaret A, Galitzky J, Decaunes P, Esteve D, Marques MA, Sengenes C, et al. Adipose
tissue endothelial cells from obese human subjects: differences among depots in angio-
genic, metabolic, and inflammatory gene expression and cellular senescence. Diabetes.
2010;59(11):2755–63.
41. Adachi A, Shimizu N, Oomura Y, Kobashi M. Convergence of hepatoportal glucose-sensitive
afferent signals to glucose-sensitive units within the nucleus of the solitary tract. Neurosci
Lett. 1984;46(2):215–8.
42. Burcelin R. The gut-brain axis: a major glucoregulatory player. Diabetes Metab. 2010;36
Suppl 3:S54–8.
43. Konturek SJ, Konturek JW, Pawlik T, Brzozowski T. Brain-gut axis and its role in the control
of food intake. J Physiol Pharmacol. 2004;55(1 Pt 2):137–54.
44. Holst JJ. The physiology of glucagon-like peptide 1. Physiol Rev. 2007;87(4):1409–39.
45. Cabou C, Burcelin R. GLP-1, the gut-brain, and brain-periphery axes. Rev Diabet Stud.
2012;8:418–31.
46. Knauf C, Cani PD, Kim DH, Iglesias MA, Chabo C, Waget A, et al. Role of central nervous
system glucagon-like peptide-1 receptors in enteric glucose sensing. Diabetes.
2008;57(10):2603–12.
47. Collins SM, Surette M, Bercik P. The interplay between the intestinal microbiota and the
brain. Nat Rev Microbiol. 2012;10(11):735–42.
48. Forsythe P, Kunze WA, Bienenstock J. On communication between gut microbes and the
brain. Curr Opin Gastroenterol. 2012;28(6):557–62.
49. Foster JA, McVey Neufeld KA. Gut-brain axis: how the microbiome influences anxiety and
depression. Trends Neurosci. 2013;36:305.
50. DeFronzo RA, Ferrannini E, Simonson DC. Fasting hyperglycemia in non-insulin-dependent
diabetes mellitus: contributions of excessive hepatic glucose production and impaired tissue
glucose uptake. Metab: Clin Exp. 1989;38(4):387–95.
51. Baudon MA, Ferre P, Penicaud L, Maulard P, Ktorza A, Castano L, et al. Normal insulin
sensitivity during the phase of glucose intolerance but insulin resistance at the onset of diabe-
tes in the spontaneously diabetic BB rat. Diabetologia. 1989;32(12):839–44.
52. Burcelin R, Eddouks M, Maury J, Kande J, Assan R, Girard J. Excessive glucose production,
rather than insulin resistance, accounts for hyperglycaemia in recent-onset streptozotocin-
diabetic rats. Diabetologia. 1995;38(3):283–90.
53. Del Prato S, Bonadonna RC, Bonora E, Gulli G, Solini A, Shank M, et al. Characterization of
cellular defects of insulin action in type 2 (non-insulin-dependent) diabetes mellitus. J Clin
Invest. 1993;91(2):484–94.
54. Golay A, Felber JP, Jequier E, DeFronzo RA, Ferrannini E. Metabolic basis of obesity and
noninsulin-dependent diabetes mellitus. Diabetes Metab Rev. 1988;4(8):727–47.
55. Hundal RS, Krssak M, Dufour S, Laurent D, Lebon V, Chandramouli V, et al. Mechanism
by which metformin reduces glucose production in type 2 diabetes. Diabetes.
2000;49(12):2063–9.
56. Martin BC, Warram JH, Krolewski AS, Bergman RN, Soeldner JS, Kahn CR. Role of glucose
and insulin resistance in development of type 2 diabetes mellitus: results of a 25-year follow-
up study [see comments]. Lancet. 1992;340(8825):925–9.
57. Bergman RN, Ader M. Free fatty acids and pathogenesis of type 2 diabetes mellitus. Trends
Endocrinol Metab. 2000;11(9):351–6.
58. Pickup JC, Crook MA. Is type II diabetes mellitus a disease of the innate immune system?
Diabetologia. 1998;41(10):1241–8.
59. Shoelson SE, Lee J, Goldfine AB. Inflammation and insulin resistance. J Clin Invest.
2006;116(7):1793–801.
60. Hotamisligil GS. Inflammation and metabolic disorders. Nature. 2006;444(7121):860–7.
61. Wellen KE, Hotamisligil GS. Inflammation, stress, and diabetes. J Clin Invest.
2005;115(5):1111–9.
62. Manco M, Fernandez-Real JM, Equitani F, Vendrell J, Valera Mora ME, Nanni G, et al. Effect
of massive weight loss on inflammatory adipocytokines and the innate immune system in
morbidly obese women. J Clin Endocrinol Metab. 2007;92(2):483–90.
63. Dandona P, Aljada A, Bandyopadhyay A. Inflammation: the link between insulin resistance,
obesity and diabetes. Trends Immunol. 2004;25(1):4–7.
64. Nishimura S, Manabe I, Nagasaki M, Eto K, Yamashita H, Ohsugi M, et al. CD8+ effector T
cells contribute to macrophage recruitment and adipose tissue inflammation in obesity. Nat
Med. 2009;15(8):914–20.
65. Lumeng CN, Bodzin JL, Saltiel AR. Obesity induces a phenotypic switch in adipose tissue
macrophage polarization. J Clin Invest. 2007;117(1):175–84.
66. Nguyen MT, Favelyukis S, Nguyen AK, Reichart D, Scott PA, Jenn A, et al. A subpopulation of
macrophages infiltrates hypertrophic adipose tissue and is activated by free fatty acids via Toll-
like receptors 2 and 4 and JNK-dependent pathways. J Biol Chem. 2007;282(48):35279–92.
67. Ueki K, Kondo T, Kahn CR. Suppressor of cytokine signaling 1 (SOCS-1) and SOCS-3 cause
insulin resistance through inhibition of tyrosine phosphorylation of insulin receptor substrate
proteins by discrete mechanisms. Mol Cell Biol. 2004;24(12):5434–46.
68. Plomgaard P, Bouzakri K, Krogh-Madsen R, Mittendorfer B, Zierath JR, Pedersen BK. Tumor
necrosis factor-alpha induces skeletal muscle insulin resistance in healthy human subjects via
inhibition of Akt substrate 160 phosphorylation. Diabetes. 2005;54(10):2939–45.
69. Federici M, Hribal ML, Menghini R, Kanno H, Marchetti V, Porzio O, et al. Timp3 deficiency
in insulin receptor-haploinsufficient mice promotes diabetes and vascular inflammation via
increased TNF-alpha. J Clin Invest. 2005;115(12):3494–505.
70. Tanti J, Gual P, Gremeaux T, Gonzalez T, Barres R, Le Marchand-Brustel Y. Alteration in
insulin action: role of IRS-1 serine phosphorylation in the retroregulation of insulin signal-
ling. Ann Endocrinol. 2004;65:43–8.
71. Bouzakri K, Roques M, Gual P, Espinosa S, Guebre-Egziabher F, Riou JP, et al. Reduced
activation of phosphatidylinositol-3 kinase and increased serine 636 phosphorylation of insu-
lin receptor substrate-1 in primary culture of skeletal muscle cells from patients with type 2
diabetes. Diabetes. 2003;52(6):1319–25.
72. Bouloumie A, Curat CA, Sengenes C, Lolmede K, Miranville A, Busse R. Role of macrophage
tissue infiltration in metabolic diseases. Curr Opin Clin Nutr Metab Care. 2005;8(4):347–54.
73. Kanda H, Tateya S, Tamori Y, Kotani K, Hiasa K, Kitazawa R, et al. MCP-1 contributes to
macrophage infiltration into adipose tissue, insulin resistance, and hepatic steatosis in obe-
sity. J Clin Invest. 2006;116(6):1494–505.
74. Suganami T, Tanimoto-Koyama K, Nishida J, Itoh M, Yuan X, Mizuarai S, et al. Role of the
Toll-like receptor 4/NF-kappaB pathway in saturated fatty acid-induced inflammatory
changes in the interaction between adipocytes and macrophages. Arterioscler Thromb Vasc
Biol. 2007;27(1):84–91.
75. Weisberg SP, Hunter D, Huber R, Lemieux J, Slaymaker S, Vaddi K, et al. CCR2 modulates
inflammatory and metabolic effects of high-fat feeding. J Clin Invest. 2006;116(1):115–24.
76. Lumeng CN, Deyoung SM, Bodzin JL, Saltiel AR. Increased inflammatory properties of adi-
pose tissue macrophages recruited during diet-induced obesity. Diabetes. 2007;56(1):16–23.
77. Charriere G, Cousin B, Arnaud E, Andre M, Bacou F, Penicaud L, et al. Preadipocyte

conversion to macrophage. Evidence of plasticity. J Biol Chem. 2003;278(11):9850–5.
78. Charriere GM, Cousin B, Arnaud E, Saillan-Barreau C, Andre M, Massoudi A, et al.
Macrophage characteristics of stem cells revealed by transcriptome profiling. Exp Cell Res.
2006;312(17):3205–14.
79. Winer S, Chan Y, Paltser G, Truong D, Tsui H, Bahrami J, et al. Normalization of obesity-
associated insulin resistance through immunotherapy. Nat Med. 2009;15(8):921–9.
80. Feuerer M, Herrero L, Cipolletta D, Naaz A, Wong J, Nayer A, et al. Lean, but not obese, fat
is enriched for a unique population of regulatory T cells that affect metabolic parameters.
Nat Med. 2009;15(8):930–9.
81. Cani PD, Bibiloni R, Knauf C, Waget A, Neyrinck AM, Delzenne NM, et al. Changes in gut
microbiota control metabolic endotoxemia-induced inflammation in high-fat diet-induced
obesity and diabetes in mice. Diabetes. 2008;57(6):1470–81.
82. Drong AW, Lindgren CM, McCarthy MI. The genetic and epigenetic basis of type 2 diabetes
and obesity. Clin Pharmacol Ther. 2012;92(6):707–15.
83. Medici F, Hawa M, Ianari A, Pyke DA, Leslie RD. Concordance rate for type II diabetes mel-
litus in monozygotic twins: actuarial analysis. Diabetologia. 1999;42(2):146–50.
84. Poulsen P, Grunnet LG, Pilgaard K, Storgaard H, Alibegovic A, Sonne MP, et al. Increased
risk of type 2 diabetes in elderly twins. Diabetes. 2009;58(6):1350–5.
85. Vaarala O, Atkinson MA, Neu J. The “perfect storm” for type 1 diabetes: the complex inter-
play between intestinal microbiota, gut permeability, and mucosal immunity. Diabetes.
2008;57(10):2555–62.
86. Wen L, Ley RE, Volchkov PY, Stranges PB, Avanesyan L, Stonebraker AC, et al. Innate
immunity and intestinal microbiota in the development of type 1 diabetes. Nature.
2008;455(7216):1109–13.
87. Markle JG, Frank DN, Mortin-Toth S, Robertson CE, Feazel LM, Rolle-Kampczyk U, et al.
Sex differences in the gut microbiome drive hormone-dependent regulation of autoimmunity.
Science. 2013;339:1084.
88. Nicholson JK, Holmes E, Wilson ID. Gut microorganisms, mammalian metabolism and per-
sonalized health care. Nat Rev Microbiol. 2005;3(5):431–8.
89. Hooper LV, Wong MH, Thelin A, Hansson L, Falk PG, Gordon JI. Molecular analysis of
commensal host-microbial relationships in the intestine. Science. 2001;291(5505):881–4.
90. Backhed F, Ding H, Wang T, Hooper LV, Koh GY, Nagy A, et al. The gut microbiota as an
environmental factor that regulates fat storage. Proc Natl Acad Sci U S A.
2004;101(44):15718–23.
91. Hooper LV, Gordon JI. Commensal host-bacterial relationships in the gut. Science.
2001;292(5519):1115–8.
92. Serino M, Chabo C, Burcelin R. Intestinal MicrobiOMICS to define health and disease in
human and mice. Curr Pharm Biotechnol. 2012;13(5):746–58.
93. Backhed F. Programming of host metabolism by the gut microbiota. Ann Nutr Metab.
2011;58 Suppl 2:44–52.
94. Turnbaugh PJ, Gordon JI. The core gut microbiome, energy balance and obesity. J Physiol.
2009;587(Pt 17):4153–8.
95. Karlsson FH, Fak F, Nookaew I, Tremaroli V, Fagerberg B, Petranovic D, et al. Symptomatic
atherosclerosis is associated with an altered gut metagenome. Nat Commun. 2012;3:1245.
96. Ley RE, Turnbaugh PJ, Klein S, Gordon JI. Microbial ecology: human gut microbes associ-
ated with obesity. Nature. 2006;444(7122):1022–3.
97. Ley RE, Turnbaugh PJ, Klein S, Gordon JI. Human gut microbes associated with obesity.
Nature. 2006;444:1022–3.
98. Ley RE, Backhed F, Turnbaugh P, Lozupone CA, Knight RD, Gordon JI. Obesity alters gut
microbial ecology. Proc Natl Acad Sci U S A. 2005;102(31):11070–5.
99. Karlsson FH, Tremaroli V, Nookaew I, Bergstrom G, Behre CJ, Fagerberg B, et al. Gut
metagenome in European women with normal, impaired and diabetic glucose control. Nature.
2013;498:99.
100. Noble D, Mathur R, Dent T, Meads C, Greenhalgh T. Risk models and scores for type 2
diabetes: systematic review. BMJ. 2011;343:d7163.
101. van Dieren S, Beulens JW, Kengne AP, Peelen LM, Rutten GE, Woodward M, et al. Prediction
models for the risk of cardiovascular disease in patients with type 2 diabetes: a systematic
review. Heart. 2012;98(5):360–9.
102. Spor A, Koren O, Ley R. Unravelling the effects of the environment and host genotype on the
gut microbiome. Nat Rev Microbiol. 2011;9(4):279–90.
103. Serino M, Luche E, Gres S, Baylac A, Berge M, Cenac C, et al. Metabolic adaptation to a
high-fat diet is associated with a change in the gut microbiota. Gut. 2012;61(4):543–53.
104. Turnbaugh P, Ridaura V, Faith J, Rey F, Knight R, Gordon J. The effect of diet on the human
gut microbiome: a metagenomic analysis in humanized gnotobiotic mice. Sci Transl Med.
2009;1:6–14.
105. Murphy EF, Cotter PD, Healy S, Marques TM, O’Sullivan O, Fouhy F, et al. Composition and
energy harvesting capacity of the gut microbiota: relationship to diet, obesity and time in
mouse models. Gut. 2010;59(12):1635–42.
106. Moreira AP, Texeira TF, Ferreira AB, Peluzio Mdo C, Alfenas RC. Influence of a high-fat diet
on gut microbiota, intestinal permeability and metabolic endotoxaemia. Br J Nutr.
2012;108(5):801–9.
107. Everard A, Belzer C, Geurts L, Ouwerkerk JP, Druart C, Bindels LB, et al. Cross-talk between
Akkermansia muciniphila and intestinal epithelium controls diet-induced obesity. Proc Natl
Acad Sci U S A. 2013;498:99.
108. Pedersen R, Andersen AD, Molbak L, Stagsted J, Boye M. Changes in the gut microbiota of
cloned and non-cloned control pigs during development of obesity: gut microbiota during
development of obesity in cloned pigs. BMC Microbiol. 2013;13:30.
109. Yan H, Potu R, Lu H, Vezzoni de Almeida V, Stewart T, Ragland D, et al. Dietary fat content
and fiber type modulate hind gut microbial community and metabolic markers in the pig.
PLoS One. 2013;8(4):e59581.
110. Park DY, Ahn YT, Park SH, Huh CS, Yoo SR, Yu R, et al. Supplementation of lactobacillus
curvatus HY7601 and lactobacillus plantarum KY1032 in diet-induced obese mice is associ-
ated with gut microbial changes and reduction in obesity. PLoS One. 2013;8(3):e59470.
111. Parks BW, Nam E, Org E, Kostem E, Norheim F, Hui ST, et al. Genetic control of obesity and
gut microbiota composition in response to high-fat, high-sucrose diet in mice. Cell Metab.
2013;17(1):141–52.
112. Mujico JR, Baccan GC, Gheorghe A, Diaz LE, Marcos A. Changes in gut microbiota due to
supplemented fatty acids in diet-induced obese mice. Br J Nutr. 2013;110:1–10.
113. de Wit N, Derrien M, Bosch-Vermeulen H, Oosterink E, Keshtkar S, Duval C, et al. Saturated
fat stimulates obesity and hepatic steatosis and affects gut microbiota composition by an
enhanced overflow of dietary fat to the distal intestine. Am J Physiol Gastrointest Liver
Physiol. 2012;303(5):G589–99.
114. Liu T, Hougen H, Vollmer AC, Hiebert SM. Gut bacteria profiles of Mus musculus at the phylum
and family levels are influenced by saturation of dietary fatty acids. Anaerobe. 2012;18(3):331–7.
115. Dao TM, Waget A, Klopp P, Serino M, Vachoux C, Pechere L, et al. Resveratrol increases
glucose induced GLP-1 secretion in mice: a mechanism which contributes to the glycemic
control. PLoS One. 2011;6(6):e20700.
116. Neyrinck AM, Van Hee VF, Bindels LB, De Backer F, Cani PD, Delzenne NM. Polyphenol-
rich extract of pomegranate peel alleviates tissue inflammation and hypercholesterolaemia in
high-fat diet-induced obese mice: potential implication of the gut microbiota. Br J Nutr.
2013;109(5):802–9.
117. Tachon S, Zhou J, Keenan M, Martin R, Marco ML. The intestinal microbiota in aged mice
is modulated by dietary resistant starch and correlated with improvements in host responses.
FEMS Microbiol Ecol. 2013;83(2):299–309.
118. Cani PD, Knauf C, Iglesias MA, Drucker DJ, Delzenne NM, Burcelin R. Improvement of
glucose tolerance and hepatic insulin sensitivity by oligofructose requires a functional
glucagon-like peptide 1 receptor. Diabetes. 2006;55(5):1484–90.
119. Serino M, Fernandez-Real JM, Fuentes EG, Queipo-Ortuno M, Moreno-Navarrete JM,

Sanchez A, et al. The gut microbiota profile is associated with insulin action in humans. Acta
Diabetol. 2012.
120. Vernon G, Baranova A, Younossi ZM. Systematic review: the epidemiology and natural his-
tory of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis in adults. Aliment
Pharmacol Ther. 2011;34(3):274–85.
121. Wong VW, Tse CH, Lam TT, Wong GL, Chim AM, Chu WC, et al. Molecular characteriza-
tion of the fecal microbiota in patients with nonalcoholic steatohepatitis - a longitudinal
study. PLoS One. 2013;8(4):e62885.
122. Raman M, Ahmed I, Gillevet PM, Probert CS, Ratcliffe NM, Smith S, et al. Fecal microbi-
ome and volatile organic compound metabolome in obese humans with nonalcoholic fatty
liver disease. Clin Gastroenterol Hepatol: Off Clin Pract J Am Gastroenterol Assoc.
2013;11:868.
123. Le Roy T, Llopis M, Lepage P, Bruneau A, Rabot S, Bevilacqua C, et al. Intestinal microbiota
determines development of non-alcoholic fatty liver disease in mice. Gut. 2012.
124. Henao-Mejia J, Elinav E, Jin C, Hao L, Mehal WZ, Strowig T, et al. Inflammasome-mediated
dysbiosis regulates progression of NAFLD and obesity. Nature. 2012;482(7384):179–85.
125. Aron-Wisnewsky J, Gaborit B, Dutour A, Clement K. Gut microbiota and non-alcoholic fatty
liver disease: new insights. Clin Microbiol Infect. 2013;19(4):338–48.
126. Turnbaugh PJ, Hamady M, Yatsunenko T, Cantarel BL, Duncan A, Ley RE, et al. A core gut
microbiome in obese and lean twins. Nature. 2009;457(7228):480–4.
127. Newman B, Selby J, King M, Slemenda C, Fabsitz R, Friedman G. Concordance for type 2
(non-insulin-dependent) diabetes mellitus in male twins. Diabetologia. 1987;30:763–8.
128. Bouchard C, Tremblay A, Despres J, Nadeau A, Lupien P, Theriault G, et al. The response to
long-term overfeeding in identical twins. N Engl J Med. 1990;322(21):1477–82. 1477-82.
129. Bouchard C, Tremblay A, Despres J, Nadeau A, Lupien P, Moorjani S, et al. Overfeeding in
identical twins: 5-year postoverfeeding results. Metab: Clin Exp. 1996;45:1042–50.
130. Turnbaugh PJ, Quince C, Faith JJ, McHardy AC, Yatsunenko T, Niazi F, et al. Organismal,
genetic, and transcriptional variation in the deeply sequenced gut microbiomes of identical
twins. Proc Natl Acad Sci U S A. 2010;107(16):7503–8.
131. Stahringer SS, Clemente JC, Corley RP, Hewitt J, Knights D, Walters WA, et al. Nurture
trumps nature in a longitudinal survey of salivary bacterial communities in twins from early
adolescence to early adulthood. Genome Res. 2012;22(11):2146–52.
132. Vijay-Kumar M, Aitken JD, Carvalho FA, Cullender TC, Mwangi S, Srinivasan S, et al.
Metabolic syndrome and altered gut microbiota in mice lacking Toll-like receptor 5. Science.
2010;328(5975):228–31.
133. Round JL, O’Connell RM, Mazmanian SK. Coordination of tolerogenic immune responses
by the commensal microbiota. J Autoimmun. 2010;34(3):J220–5.
134. Burcelin R, Garidou L, Pomie C. Immuno-microbiota cross and talk: the new paradigm of
metabolic diseases. Semin Immunol. 2012;24(1):67–74.
135. Natividad JM, Verdu EF. Modulation of intestinal barrier by intestinal microbiota: pathological
and therapeutic implications. Pharmacol Res: Off J Ital Pharmacol Soc. 2013;69(1):42–51.
136. Caricilli AM, Picardi PK, de Abreu LL, Ueno M, Prada PO, Ropelle ER, et al. Gut microbiota
is a key modulator of insulin resistance in TLR 2 knockout mice. PLoS Biol.
2011;9(12):e1001212.
137. Kim KA, Gu W, Lee IA, Joh EH, Kim DH. High fat diet-induced gut microbiota exacerbates
inflammation and obesity in mice via the TLR4 signaling pathway. PLoS One. 2012;7(10):e47713.
138. Rehman A, Sina C, Gavrilova O, Hasler R, Ott S, Baines JF, et al. Nod2 is essential for tem-
poral development of intestinal microbial communities. Gut. 2011;60(10):1354–62.
139. Mondot S, Barreau F, Al Nabhani Z, Dussaillant M, Le Roux K, Dore J, et al. Altered gut
microbiota composition in immune-impaired Nod2(-/-) mice. Gut. 2012;61(4):634–5.
140. Rogler G. The importance of gut microbiota in mediating the effect of NOD2 defects in
inflammatory bowel disease. Gut. 2010;59(2):153–4.
141. Robertson SJ, Zhou JY, Geddes K, Rubino SJ, Cho JH, Girardin SE, et al. Nod1 and Nod2
signaling does not alter the composition of intestinal bacterial communities at homeostasis.
Gut Microbes. 2013;4(3):222–31.
142. Palmer C, Bik EM, Digiulio DB, Relman DA, Brown PO. Development of the human infant
intestinal microbiota. PLoS Biol. 2007;5(7):e177.
143. Dominguez-Bello MG, Costello EK, Contreras M, Magris M, Hidalgo G, Fierer N, et al.
Delivery mode shapes the acquisition and structure of the initial microbiota across multiple
body habitats in newborns. Proc Natl Acad Sci U S A. 2010;107(26):11971–5.
144. Ajslev TA, Andersen CS, Gamborg M, Sorensen TI, Jess T. Childhood overweight after
establishment of the gut microbiota: the role of delivery mode, pre-pregnancy weight and
early administration of antibiotics. Int J Obes (Lond). 2011;35(4):522–9.
145. Koren O, Goodrich JK, Cullender TC, Spor A, Laitinen K, Backhed HK, et al. Host remod-
eling of the gut microbiome and metabolic changes during pregnancy. Cell.
2012;150(3):470–80.
146. Collado MC, Isolauri E, Laitinen K, Salminen S. Effect of mother’s weight on infant’s micro-
biota acquisition, composition, and activity during early infancy: a prospective follow-up
study initiated in early pregnancy. Am J Clin Nutr. 2010;92(5):1023–30.
147. Karlsson CL, Onnerfalt J, Xu J, Molin G, Ahrne S, Thorngren-Jerneck K. The microbiota of
the gut in preschool children with normal and excessive body weight. Obesity (Silver Spring).
2012;20:2257.
148. Song MJ, Kim KH, Yoon JM, Kim JB. Activation of Toll-like receptor 4 is associated with
insulin resistance in adipocytes. Biochem Biophys Res Commun. 2006;346(3):739–45.
149. Zu L, He J, Jiang H, Xu C, Pu S, Xu G. Bacterial endotoxin stimulates adipose lipolysis via
toll-like receptor 4 and extracellular signal-regulated kinase pathway. J Biol Chem.
2009;284(9):5915–26.
150. Pendyala S, Walker J, Holt PR. A high-fat diet is associated with endotoxemia that originates
from the gut. Gastroenterology. 2012;142:1100.
151. Lassenius MI, Pietilainen KH, Kaartinen K, Pussinen PJ, Syrjanen J, Forsblom C, et al.
Bacterial endotoxin activity in human serum is associated with dyslipidemia, insulin resis-
tance, obesity, and chronic inflammation. Diabetes Care. 2011;34(8):1809–15.
152. Ghoshal S, Witta J, Zhong J, de Villiers W, Eckhardt E. Chylomicrons promote intestinal
absorption of lipopolysaccharides. J Lipid Res. 2009;50(1):90–7.
153. Laugerette F, Vors C, Geloen A, Chauvin MA, Soulage C, Lambert-Porcheron S, et al.
Emulsified lipids increase endotoxemia: possible role in early postprandial low-grade inflam-
mation. J Nutr Biochem. 2010;22:53.
154. Geurts L, Lazarevic V, Derrien M, Everard A, Van Roye M, Knauf C, et al. Altered gut micro-
biota and endocannabinoid system tone in obese and diabetic leptin-resistant mice: impact on
apelin regulation in adipose tissue. Front Microbiol. 2011;2:149.
155. Cani PD, Possemiers S, Van de Wiele T, Guiot Y, Everard A, Rottier O, et al. Changes in gut
microbiota control inflammation in obese mice through a mechanism involving GLP-2-
driven improvement of gut permeability. Gut. 2009;58(8):1091–103.
156. Russo F, Linsalata M, Clemente C, Chiloiro M, Orlando A, Marconi E, et al. Inulin-enriched
pasta improves intestinal permeability and modifies the circulating levels of zonulin and
glucagon-like peptide 2 in healthy young volunteers. Nutr Res. 2012;32(12):940–6.
157. Stoidis CN, Misiakos EP, Patapis P, Fotiadis CI, Spyropoulos BG. Potential benefits of pro-
and prebiotics on intestinal mucosal immunity and intestinal barrier in short bowel syndrome.
Nutr Res Rev. 2010:1–9.
158. Kitchens RL, Thompson PA, Viriyakosol S, O’Keefe GE, Munford RS. Plasma CD14
decreases monocyte responses to LPS by transferring cell-bound LPS to plasma lipoproteins.
J Clin Invest. 2001;108(3):485–93.
159. Vreugdenhil AC, Rousseau CH, Hartung T, Greve JW, van ’t Veer C, Buurman
WA. Lipopolysaccharide (LPS)-binding protein mediates LPS detoxification by chylomi-
crons. J Immunol. 2003;170(3):1399–405.
160. Amar J, Burcelin R, Ruidavets JB, Cani PD, Fauvel J, Alessi MC, et al. Energy intake is
associated with endotoxemia in apparently healthy men. Am J Clin Nutr. 2008;87(5):1219–23.
161. Moreno-Navarrete JM, Manco M, Ibanez J, Garcia-Fuentes E, Ortega F, Gorostiaga E, et al.
Metabolic endotoxemia and saturated fat contribute to circulating NGAL concentrations in
subjects with insulin resistance. Int J Obes (Lond). 2010;34(2):240–9.
162. Shi H, Kokoeva MV, Inouye K, Tzameli I, Yin H, Flier JS. TLR4 links innate immunity and
fatty acid-induced insulin resistance. J Clin Invest. 2006;116(11):3015–25.
163. Vandanmagsar B, Youm YH, Ravussin A, Galgani JE, Stadler K, Mynatt RL, et al. The
NLRP3 inflammasome instigates obesity-induced inflammation and insulin resistance. Nat
Med. 2011;17(2):179–88.
164. Poggi M, Bastelica D, Gual P, Iglesias MA, Gremeaux T, Knauf C, et al. C3H/HeJ mice
carrying a toll-like receptor 4 mutation are protected against the development of insulin
resistance in white adipose tissue in response to a high-fat diet. Diabetologia.
2007;50(6):1267–76.
165. Liang H, Hussey SE, Sanchez-Avila A, Tantiwong P, Musi N. Effect of lipopolysaccharide on
inflammation and insulin action in human muscle. PLoS One. 2013;8(5):e63983.
166. Suganami T, Mieda T, Itoh M, Shimoda Y, Kamei Y, Ogawa Y. Attenuation of obesity-induced
adipose tissue inflammation in C3H/HeJ mice carrying a Toll-like receptor 4 mutation.
Biochem Biophys Res Commun. 2007;354(1):45–9.
167. Xu H, Barnes GT, Yang Q, Tan G, Yang D, Chou CJ, et al. Chronic inflammation in fat plays
a crucial role in the development of obesity-related insulin resistance. J Clin Invest.
2003;112(12):1821–30.
168. Gerhardt CC, Romero IA, Cancello R, Camoin L, Strosberg AD. Chemokines control fat
accumulation and leptin secretion by cultured human adipocytes. Mol Cell Endocrinol.
2001;175(1–2):81–92.
169. Hotamisligil GS, Peraldi P, Budavari A, Ellis R, White MF, Spiegelman BM. IRS-1-mediated
inhibition of insulin receptor tyrosine kinase activity in TNF-alpha- and obesity-induced
insulin resistance. Science. 1996;271(5249):665–8.
170. Wang L, Li L, Ran X, Long M, Zhang M, Tao Y, et al. Lipopolysaccharides reduce adipogen-
esis in 3T3-L1 adipocytes through activation of NF-kappaB pathway and downregulation of
AMPK expression. Cardiovasc Toxicol. 2013;13:338.
171. Gabele E, Dostert K, Hofmann C, Wiest R, Scholmerich J, Hellerbrand C, et al. DSS induced
colitis increases portal LPS levels and enhances hepatic inflammation and fibrogenesis in
experimental NASH. J Hepatol. 2011;55(6):1391–9.
172. Tomofuji T, Ekuni D, Yamanaka R, Kusano H, Azuma T, Sanbe T, et al. Chronic administra-
tion of lipopolysaccharide and proteases induces periodontal inflammation and hepatic ste-
atosis in rats. J Periodontol. 2007;78(10):1999–2006.
173. Endo M, Masaki T, Seike M, Yoshimatsu H. TNF-alpha induces hepatic steatosis in mice by
enhancing gene expression of sterol regulatory element binding protein-1c (SREBP-1c). Exp
Biol Med (Maywood). 2007;232(5):614–21.
174. Sansonetti P, Di Santo J. Debugging how bacteria manipulate the immune response.
Immunity. 2007;26:149–61.
175. Kufer TA, Sansonetti PJ. Sensing of bacteria: NOD a lonely job. Curr Opin Microbiol.
2007;10(1):62–9.
176. Sansonetti P. Host-pathogen interactions: the seduction of molecular cross talk. Gut. 2002;50
Suppl 3:III2–8.
177. Stappenbeck TS, Hooper LV, Gordon JI. Developmental regulation of intestinal angiogenesis
by indigenous microbes via Paneth cells. Proc Natl Acad Sci U S A. 2002;99(24):15451–5.
178. Carneiro LA, Magalhaes JG, Tattoli I, Philpott DJ, Travassos LH. Nod-like proteins in
inflammation and disease. J Pathol. 2008;214(2):136–48.
179. Schertzer JD, Tamrakar AK, Magalhaes JG, Pereira S, Bilan PJ, Fullerton MD, et al. NOD1
activators link innate immunity to insulin resistance. Diabetes. 2011;60(9):2206–15.
180. Caesar R, Reigstad CS, Backhed HK, Reinhardt C, Ketonen M, Ostergren Lunden G, et al.
Gut-derived lipopolysaccharide augments adipose macrophage accumulation but is not
essential for impaired glucose or insulin tolerance in mice. Gut. 2012;61:1701.
181. Tamrakar AK, Schertzer JD, Chiu TT, Foley KP, Bilan PJ, Philpott DJ, et al. NOD2 activation
induces muscle cell-autonomous innate immune responses and insulin resistance.
Endocrinology. 2010;151(12):5624–37.
182. Ogura Y, Lala S, Xin W, Smith E, Dowds TA, Chen FF, et al. Expression of NOD2 in Paneth
cells: a possible link to Crohn’s ileitis. Gut. 2003;52(11):1591–7.
183. Ogura Y, Saab L, Chen FF, Benito A, Inohara N, Nunez G. Genetic variation and activity of
mouse Nod2, a susceptibility gene for Crohn’s disease. Genomics. 2003;81(4):369–77.
184. Inohara N, Ogura Y, Nunez G. Nods: a family of cytosolic proteins that regulate the host
response to pathogens. Curr Opin Microbiol. 2002;5(1):76–80.
185. Miele L, Valenza V, La Torre G, Montalto M, Cammarota G, Ricci R, et al. Increased intesti-
nal permeability and tight junction alterations in nonalcoholic fatty liver disease. Hepatology.
2009;49(6):1877–87.
186. Lichtman SN, Keku J, Schwab JH, Sartor RB. Hepatic injury associated with small bowel
bacterial overgrowth in rats is prevented by metronidazole and tetracycline. Gastroenterology.
1991;100(2):513–9.
187. Vance JE, Vance DE. The role of phosphatidylcholine biosynthesis in the secretion of lipo-
proteins from hepatocytes. Can J Biochem Cell Biol. 1985;63(8):870–81.
188. Ortu G, Rivabene R, Cantafora A, Bravo E. Age-related variations in hepatic biosynthesis of
phosphatidylcholine: a study of choline metabolism with perfused rat liver. Proc Soc Exp
Biol Med. 1997;216(1):44–51.
189. Vance DE. Role of phosphatidylcholine biosynthesis in the regulation of lipoprotein homeo-
stasis. Curr Opin Lipidol. 2008;19(3):229–34.
190. Dumas ME, Barton RH, Toye A, Cloarec O, Blancher C, Rothwell A, et al. Metabolic profil-
ing reveals a contribution of gut microbiota to fatty liver phenotype in insulin-resistant mice.
Proc Natl Acad Sci U S A. 2006;103(33):12511–6.
2008;4:219.
192. Martin FP, Wang Y, Sprenger N, Yap IK, Lundstedt T, Lek P, et al. Probiotic modulation of
symbiotic gut microbial-host metabolic interactions in a humanized microbiome mouse
model. Mol Syst Biol. 2008;4:157.
193. Carvalho BM, Guadagnini D, Tsukumo DM, Schenka AA, Latuf-Filho P, Vassallo J, et al.
Modulation of gut microbiota by antibiotics improves insulin signalling in high-fat fed mice.
Diabetologia. 2012;55(10):2823–34.
194. Lin HV, Frassetto A, Kowalik Jr EJ, Nawrocki AR, Lu MM, Kosinski JR, et al. Butyrate and
propionate protect against diet-induced obesity and regulate gut hormones via free fatty acid
receptor 3-independent mechanisms. PLoS One. 2012;7(4):e35240.
195. Al-Lahham SH, Peppelenbosch MP, Roelofsen H, Vonk RJ, Venema K. Biological effects of
propionic acid in humans; metabolism, potential applications and underlying mechanisms.
Biochim Biophys Acta. 2010;1801(11):1175–83.
196. Gao Z, Yin J, Zhang J, Ward RE, Martin RJ, Lefevre M, et al. Butyrate improves insulin
sensitivity and increases energy expenditure in mice. Diabetes. 2009;58(7):1509–17.
197. Payne AN, Chassard C, Zimmermann M, Muller P, Stinca S, Lacroix C. The metabolic activ-
ity of gut microbiota in obese children is increased compared with normal-weight children
and exhibits more exhaustive substrate utilization. Nutr Diabetes. 2011;1:e12.
198. Kimura I, Ozawa K, Inoue D, Imamura T, Kimura K, Maeda T, et al. The gut microbiota sup-
presses insulin-mediated fat accumulation via the short-chain fatty acid receptor GPR43. Nat
Commun. 2013;4:1829.
199. Backhed F, Manchester JK, Semenkovich CF, Gordon JI. Mechanisms underlying the resistance
to diet-induced obesity in germ-free mice. Proc Natl Acad Sci U S A. 2007;104(3):979–84.
200. Hardie DG, Carling D, Carlson M. The AMP-activated/SNF1 protein kinase subfamily: met-
abolic sensors of the eukaryotic cell? Annu Rev Biochem. 1998;67:821–55.
201. Bergeron R, Russell 3rd RR, Young LH, Ren JM, Marcucci M, Lee A, et al. Effect of AMPK
activation on muscle glucose metabolism in conscious rats. Am J Physiol. 1999;276(5 Pt
1):E938–44.
202. Aron-Wisnewsky J, Dore J, Clement K. The importance of the gut microbiota after bariatric
surgery. Nat Rev Gastroenterol Hepatol. 2012;9(10):590–8.
203. Sweeney TE, Morton JM. The human gut microbiome: a review of the effect of obesity and
surgically induced weight loss. JAMA Surg. 2013;148:1–7.
204. Zhang H, DiBaise JK, Zuccolo A, Kudrna D, Braidotti M, Yu Y, et al. Human gut microbiota
in obesity and after gastric bypass. Proc Natl Acad Sci U S A. 2009;106(7):2365–70.
205. Liou AP, Paziuk M, Luevano Jr JM, Machineni S, Turnbaugh PJ, Kaplan LM. Conserved
shifts in the gut microbiota due to gastric bypass reduce host weight and adiposity. Sci Transl
Med. 2013;5(178):178ra41.
206. Kong LC, Tap J, Aron-Wisnewsky J, Pelloux V, Basdevant A, Bouillot JL, et al. Gut micro-
biota after gastric bypass in human obesity: increased richness and associations of bacterial
genera with adipose tissue genes. Am J Clin Nutr. 2013.
207. Furet JP, Kong LC, Tap J, Poitou C, Basdevant A, Bouillot JL, et al. Differential adaptation of
human gut microbiota to bariatric surgery-induced weight loss: links with metabolic and low-
grade inflammation markers. Diabetes. 2010;59(12):3049–57.
208. Cummings JH, Macfarlane GT. Role of intestinal bacteria in nutrient metabolism. JPEN J
Parenter Enter Nutr. 1997;21(6):357–65.
209. Seyer P, Vallois D, Poitry-Yamate C, Schutz F, Metref S, Tarussio D, et al. Hepatic glucose
sensing is required to preserve beta cell glucose competence. J Clin Invest.
2013;123(4):1662–76.
210. Zarrinpar A, Loomba R. Review article: the emerging interplay among the gastrointestinal
tract, bile acids and incretins in the pathogenesis of diabetes and non-alcoholic fatty liver
disease. Aliment Pharmacol Ther. 2012;36(10):909–21.
211. Watanabe M, Morimoto K, Houten SM, Kaneko-Iwasaki N, Sugizaki T, Horai Y, et al. Bile
acid binding resin improves metabolic control through the induction of energy expenditure.
PLoS One. 2012;7(8):e38286.
212. Meissner M, Wolters H, de Boer RA, Havinga R, Boverhof R, Bloks VW, et al. Bile acid
sequestration normalizes plasma cholesterol and reduces atherosclerosis in hypercholesterol-
emic mice. No additional effect of physical activity. Atherosclerosis. 2013;228(1):117–23.
213. Yokota A, Fukiya S, Islam KB, Ooka T, Ogura Y, Hayashi T, et al. Is bile acid a determinant
of the gut microbiota on a high-fat diet? Gut Microbes. 2012;3(5):455–9.
214. Pournaras DJ, Glicksman C, Vincent RP, Kuganolipava S, Alaghband-Zadeh J, Mahon D,
et al. The role of bile after Roux-en-Y gastric bypass in promoting weight loss and improving
glycaemic control. Endocrinology. 2012;153(8):3613–9.
215. Keitel V, Haussinger D. Perspective: TGR5 (Gpbar-1) in liver physiology and disease. Clin
Res Hepatol Gastroenterol. 2012;36(5):412–9.
216. Devkota S, Wang Y, Musch MW, Leone V, Fehlner-Peach H, Nadimpalli A, et al. Dietary-fat-
induced taurocholic acid promotes pathobiont expansion and colitis in Il10-/- mice. Nature.
2012;487(7405):104–8.
217. Zhao Y, Wu J, Li JV, Zhou NY, Tang H, Wang Y. Gut microbiota composition modifies fecal
metabolic profiles in mice. J Proteome Res. 2013.
218. Sayin SI, Wahlstrom A, Felin J, Jantti S, Marschall HU, Bamberg K, et al. Gut microbiota
regulates bile acid metabolism by reducing the levels of tauro-beta-muricholic acid, a natu-
rally occurring FXR antagonist. Cell Metab. 2013;17(2):225–35.
219. Hersoug LG, Linneberg A. The link between the epidemics of obesity and allergic diseases:
does obesity induce decreased immune tolerance? Allergy. 2007;62(10):1205–13.
220. Lolmede K, Duffaut C, Zakaroff-Girard A, Bouloumie A. Immune cells in adipose tissue: key
players in metabolic disorders. Diabetes Metab. 2011;37(4):283–90.
221. Decaunes P, Esteve D, Zakaroff-Girard A, Sengenes C, Galitzky J, Bouloumie A. Adipose-
derived stromal cells: cytokine expression and immune cell contaminants. Methods Mol Biol.
2011;702:151–61.
222. Macpherson AJ, Geuking MB, McCoy KD. Immune responses that adapt the intestinal
mucosa to commensal intestinal bacteria. Immunology. 2005;115(2):153–62.
223. Owen RL, Jones AL. Epithelial cell specialization within human Peyer’s patches: an ultra-
structural study of intestinal lymphoid follicles. Gastroenterology. 1974;66(2):189–203.
224. Berg RD. Bacterial translocation from the gastrointestinal tract. Trends Microbiol.
1995;3(4):149–54.
225. Berg RD, Wommack E, Deitch EA. Immunosuppression and intestinal bacterial overgrowth
synergistically promote bacterial translocation. Arch Surg. 1988;123(11):1359–64.
226. Ouellette AJ, Bevins CL. Paneth cell defensins and innate immunity of the small bowel.
Inflamm Bowel Dis. 2001;7(1):43–50.
227. Schenk M, Mueller C. The mucosal immune system at the gastrointestinal barrier. Best Pract
Res Clin Gastroenterol. 2008;22(3):391–409.
228. Steed H, Macfarlane GT, Blackett KL, Macfarlane S, Miller MH, Bahrami B, et al. Bacterial
translocation in cirrhosis is not caused by an abnormal small bowel gut microbiota. FEMS
Immunol Med Microbiol. 2011;63(3):346–54.
229. Koh IH, Liberatore AM, Menchaca-Diaz JL, Ruiz-Silva M, Vilela-Oliveira L, Watanabe AY,
et al. Bacterial translocation, microcirculation injury and sepsis. Endocr Metab Immune
Disord Drug Targets. 2006;6(2):143–50.
230. Klatt NR, Funderburg NT, Brenchley JM. Microbial translocation, immune activation, and
HIV disease. Trends Microbiol. 2013;21(1):6–13.
231. Jevtovic DJ, Dragovic G, Salemovic D, Ranin J, Djurkovic-Djakovic O. The metabolic syn-
drome, an epidemic among HIV-infected patients on HAART. BioMed Pharmacother.
2009;63(5):337–42.
232. Potthoff A, Brockmeyer NH, Gelbrich G, Neuhaus K, Esser S, Reinsch N, et al.
Lipodystrophy – a sign for metabolic syndrome in patients of the HIV-HEART study. J Dtsch
Dermatol Ges. 2010;8(2):92–8.
233. Dandekar S, George MD, Baumler AJ. Th17 cells, HIV and the gut mucosal barrier. Curr
Opin HIV AIDS. 2010;5(2):173–8.
234. Amar J, Serino M, Lange C, Chabo C, Iacovoni J, Mondot S, et al. Involvement of tissue
bacteria in the onset of diabetes in humans: evidence for a concept. Diabetologia.
2011;54(12):3055–61.
235. Amar J, Lange C, Payros G, Garret C, Chabo C, Lantieri O, et al. Blood microbiota dysbiosis
is associated with the onset of cardiovascular events in a large general population: the
D.E.S.I.R. study. PLoS One. 2013;8(1):e54461.
236. Hosoi T, Yokoyama S, Matsuo S, Akira S, Ozawa K. Myeloid differentiation factor 88
(MyD88)-deficiency increases risk of diabetes in mice. PLoS One. 2010;5:9.
237. Riant E, Waget A, Cogo H, Arnal JF, Burcelin R, Gourdy P. Estrogens protect against high-fat
diet-induced insulin resistance and glucose intolerance in mice. Endocrinology.
2009;150(5):2109–17.
238. Otero M, Lago R, Gomez R, Lago F, Gomez-Reino JJ, Gualillo O. Leptin: a metabolic
hormone that functions like a proinflammatory adipokine. Drug News Perspect.
2006;19(1):21–6.
239. Caesar R, Fak F, Backhed F. Effects of gut microbiota on obesity and atherosclerosis via
modulation of inflammation and lipid metabolism. J Intern Med. 2010;268(4):320–8.
240. Koren O, Spor A, Felin J, Fak F, Stombaugh J, Tremaroli V, et al. Human oral, gut, and plaque
microbiota in patients with atherosclerosis. Proc Natl Acad Sci U S A. 2011;108 Suppl
1:4592–8.
241. Griffin JL, Scott J, Nicholson JK. The influence of pharmacogenetics on fatty liver disease in
the wistar and kyoto rats: a combined transcriptomic and metabonomic study. J Proteome
Res. 2007;6(1):54–61.
242. Wang Z, Klipfell E, Bennett BJ, Koeth R, Levison BS, Dugar B, et al. Gut flora metabolism
of phosphatidylcholine promotes cardiovascular disease. Nature. 2011;472(7341):57–63.
243. Tang WH, Wang Z, Levison BS, Koeth RA, Britt EB, Fu X, et al. Intestinal microbial
metabolism of phosphatidylcholine and cardiovascular risk. N Engl J Med.
2013;368(17):1575–84.
244. Blasco-Baque V, Serino M, Vergnes JN, Riant E, Loubieres P, Arnal JF, et al. High-fat diet
induces periodontitis in mice through lipopolysaccharides (LPS) receptor signaling: protec-
tive action of estrogens. PLoS One. 2012;7(11):e48220.
245. Hashimoto T, Perlot T, Rehman A, Trichereau J, Ishiguro H, Paolino M, et al. ACE2 links
amino acid malnutrition to microbial ecology and intestinal inflammation. Nature.
2012;487(7408):477–81.
246. Kaddurah-Daouk R, Baillie RA, Zhu H, Zeng ZB, Wiest MM, Nguyen UT, et al. Enteric
microbiome metabolites correlate with response to simvastatin treatment. PLoS One.
2011;6(10):e25482.
Chapter 12
Role of Microbiota in Regulating Host Lipid
Metabolism and Disease Risk
Isabel Bondia-Pons, Tuulia Hyötyläinen, and Matej Orešič
Abstract The gut microbiota is an environmental factor which affects host metabolism
and correspondingly contributes to obesity and its cardiometabolic comorbidities.
However, the mechanisms behind the regulation of host lipid metabolism by gut
microbiota are poorly understood. The mechanistic in vivo studies over the past
decade combining gnotobiotic animal models, metagenomics, and metabolomics
have begun to decipher the role of the gut microbiota in the regulation of host physi-
ology. However, since the animal studies cannot be easily extrapolated to humans,
it still remains unclear if an altered microbiota associated with a specific disease in
humans is a causative factor or merely a consequence of the disease state or both.
Cheaper and more comprehensive sequencing tools for the analysis of gut micro-
biota, together with the recent advances in mass spectrometry-based analysis of
molecular lipids, are expected to contribute to our understanding of the mechanisms
linking gut microbiota, host lipid metabolism, and how they together contribute to
metabolic comorbidities of obesity.
Keywords Analytics • Bile acids • Bioinformatics tools • Gnotobiotics • Lipid

characterization • Lipid metabolism • Lipidomics • Mass spectrometry •
Metabolomics • Metagenomics • Metataxonomics • Microbiota • Signaling mole-
cules • Systems biology
12.1 The Link Between Lipids and Gut Microbiota
Advances in the study of microbiomes by culture-independent molecular methods

based on the ribosomal 16S sequence have revolutionized the analysis of gut microbi-
ota. Metagenomic sequencing has confirmed that the human gut microbiota is a very
complex community of about 100 trillion archaeal and bacterial cells corresponding to
over more than 1,000 species [1]. The community is dominated by bacteria belonging
I. Bondia-Pons • T. Hyötyläinen • M. Orešič (*)

Steno Diabetes Center, Niels Steensens Vej 2, 2820 Gentofte, Denmark

DOI 10.1007/978-1-4471-6539-2_12
236 I. Bondia-Pons et al.
to three major groups (phyla), Firmicutes, Bacteroidetes, and Actinobacteria, that

together represent >95 % of the total microbiota [2].
Diet, genetic background, and immune system status are among the factors
affecting the composition of gut microbiota [3, 4], justifying the high interindivid-
ual variation observed in human microbiota. However, it is believed that at least a
common core gut microbiota, and core microbiome, shared among individuals [3,
5] is needed for a correct functioning of the gut. Recent research shows that micro-
bial community has indeed a beneficial role during normal homeostasis, modulating
the host’s immune system as well as influencing host development, physiology, and
metabolism [6], but the underlying mechanisms remain largely unknown. By apply-
ing the “omics” approach, molecular snapshots of biological systems can be gener-
ated, allowing the study of comprehensive molecular profiles in time as dependent
on genetic or environmental variation. Systems biology approach, as discussed in
Chap. 1, is therefore essential in order to deal with the “omics data,” thus shifting
the research emphasis from single molecular components to how they together
contribute as parts of a complex network to a specific phenotype or biological
function [7]. In this context, changes in the concentration of specific groups of
metabolites are sensitive and specific to pathogenically relevant factors including
gut microbiota. As discussed from a nutritional perspective in Chap. 5, metabolo-
mics is expected to help in understanding changes to gut microbiota populations and
function. How do gut microbiota and other environmental factors affect the
Fig. 12.1 Main functions of lipids in the human body. Lipids function as energy source, partici-
pate in essential signaling pathways, and constitute the cellular structural building blocks in cell
and organelle membranes
12 Role of Microbiota in Regulating Host Lipid Metabolism and Disease Risk 237
lipidome and the underlying pathways is indeed one of the research questions to
address in current and future studies in the nutrigenomics field.
Lipid homeostasis is fundamental to the maintenance of health. Lipids are a
diverse group of compounds with many biological functions. The structural diver-
sity of lipids is due to many combinatorial possibilities of how the lipid building
blocks are put together, so that hundreds of thousands of distinct lipid species are
theoretically possible [8, 9]. Lipids function as energy storage sources, participate
in essential signaling pathways, and constitute the cellular structural building blocks
in both cell and organelle membranes [10, 11] (Fig. 12.1). Lipids are thus directly
involved in membrane trafficking, regulating membrane proteins, creating specific
subcompartments in membranes that contribute to cellular function [12, 13], and in
providing dynamic highly specialized molecular scaffolds for the construction of
microscopic and macroscopic chemical assemblies needed for life processes [14].
Structurally, lipids are classified into eight main groups, namely, fatty acyls, glyc-
erolipids, glycerophospholipids, sphingolipids, saccharolipids and polyketides
(derived from condensation of ketoacyl subunits), and sterol lipids and prenol lipids
(derived from condensation of isoprene subunits) [15] (Fig. 12.2).
Category Abbreviation Sub-category
Fatty acyls FA Fatty acids and conjugates

Octadecanoids
Eicosanoids
Docosanoids
Fatty alcohols
Fatty aldehydes
Fatty esters
Glycerolipids GL Monoradylglycerols
Diradylglycerols
Triradylglycerols
Glycerophospholipids GP Phosphatidic acids
Phosphatidylcholines
Phosphatidylserines
Phosphatidylglycerols
Phosphatidylethanolamines
Phosphatidylinositols
Phosphatidylinositides
Cardiolipins
Sphingolipids SP Sphingoid bases
Ceramides
Phosphosphingolipids
Phosphonosphingolipids
Neutral glycosphingolipids
Acidic glycosphingolipids
Sterol lipids ST Sterols
Steroids
Secosteroids
Bile acids and derivatives
Prenol lipids PR Isoprenoids
Quinones and hydroquinones
Polyphenols
Saccharolipids SL Acylaminosugars
Acylaminosugar glycans
Acyltrehaloses
Acyltrehalose glycans
Polyketides PK Macrolide polyketides
Aromatic polyketides
Non-ribosomal
peptide/polyketide hybrids
Fig. 12.2 Structural classification of lipids

12.2 Lipidomics
Lipidomics, a subdiscipline of metabonomics focused on the global study of molec-

ular lipids, including pathways and networks of cellular lipids in biological systems
[16], has been progressing rapidly over the past decade due to the advances in mass
spectrometry (MS) [17, 18], computational methods [19, 20], and systems biology
approaches [21]. These recent advances have highly influenced the evolution of
lipidomics, together with the recognition of the major role that lipids play in health
maintenance and in many metabolic diseases and their comorbidities. Lipidomic
studies play indeed an essential role in defining the biochemical mechanisms of
lipid-related diseases through identifying alterations in cellular lipid metabolism,
trafficking, and homeostasis. Lipidomics is therefore envisaged to valuably contrib-
ute in a systems biology approach to understand the role that microbiota plays in the
complex lipid host metabolism and, consequently, in human health and disease.
Increasing evidence has showed that lipid profiling might powerfully aid in clinical
risk assessment. A simple example can be found in the study of triglycerides (TG).
Owing to many possible combinations of acyl chains esterified to a glycerol backbone,
many different molecular species of TG can be found in the human body. However, the
current standard clinical measurement of TG relies on the measurement of total glyc-
erol after acyl chain hydrolysis [22], obscuring this underlying diversity. Discriminating
plasma lipids at a molecular level might shed insight on the intersection between dys-
lipidemia and metabolic risk in several metabolic diseases. For instance, the application
of LC–MS-based lipidomic profiling has shown that specific TG containing FAs with
lower carbon number and double bond content are associated with insulin resistance
[23]. In agreement with the previous study, Rhee and colleagues found that the same
TG signature was also predictive of type 2 diabetes in subjects who were followed up
for over 12 years in the Framingham Heart Study [24].
The gut microbiota is one of the many environmental factors influencing the host
lipid metabolism, but detailed lipid characterization is needed in future studies if we
want to elucidate the complete mechanisms behind the regulation of the lipid host
metabolism by the gut microbiota. Advances in this field can only be understood
thanks to the recent (and future) advances in both analytical and computational meth-
ods and to its mandatory integration in a systems biology approach with other “omics”
technologies such as metagenomics. The purpose of this section is to familiarize the
reader with the core tools needed in lipidomics, by briefly overviewing the state-of-
the-art and recent advances in the lipidomic field from an analytical perspective.
12.2.1 Analytical Lipidomics
Comprehensive analysis of molecular lipids requires specific analytical proce-

dures [25] (Fig. 12.3). Generally, two types of analytical strategies are applied in
lipidomic analysis: (a) the hypothesis-driven targeted selective analysis and,
Fig. 12.3 Typical workflow used in lipidomics, starting from sampling, followed by sample prep-
aration, analysis, data preprocessing, statistical analysis, and bioinformatics
more comprehensive, (b) the hypothesis-generating nontargeted profiling analysis.

In the targeted analysis, only preselected lipids are analyzed. This approach
allows very sensitive and robust determination of the selected metabolites, but it
gives only limited information about the global lipidome. The nontargeted
approaches aim to cover as many lipids as possible in a single analysis, across
several lipid classes. As a drawback, the nontargeted approaches are typically
only semiquantitative [26, 27].
Mass spectrometry has a central role in lipidomics [28]. Lipids are commonly
analyzed either by direct infusion into the MS instrument (the so-called shotgun
MS) or with MS in combination with a chromatographic separation method liquid
chromatography (LC) or gas chromatography (GC). The LC–MS approaches are
well suited for targeted as well as nontargeted analysis of lipids. While most of the
complex lipids are too nonvolatile for the analysis by GC–MS, this approach is
commonly applied for the analysis of FAs and sterols. Tandem or hybrid mass
spectrometry (MS/MS) is used both for nontargeted analyses and for increasing the
sensitivity and selectivity of quantitative analysis. In MS/MS experiments, the first
analyzer is used to select a precursor ion which is fragmented in a collision cell.
The product ions, i.e., the fragments of the precursor ion (collision-induced disso-
ciation (CID)), are then detected in the second mass analyzer. For further identifi-
cation, this MS/MS process can be iteratively repeated with sequential selection of
resultant ions for fragmentation in MSn experiments. Suitable MS systems for CID
include both quadruple-based tandem in-space instruments (e.g., triple quadrupole
(QqQ) or quadrupole time-of-flight (QTOF)) and ion trap-based tandem in-time
instruments (e.g., quadrupole-ion trap (QIT), linear trap quadrupole (LTQ)-Orbitrap,
or linear trap quadrupole Fourier-transform ion cyclotron resonance (LTQ-FTICR)).

Among these, the QqQ is typically used for targeted quantitative analysis and the
high-resolution systems for global lipid profiling as well as for the identification
of lipids.
Novel hybrid high-resolution instruments, such as combination of ion mobility
and TOFMS (IM–TOFMS), have recently been launched. They are promising tools
to assist in the still great challenge of identification and quantification of all poten-
tial enantiomeric, stereoisomeric, and regioisomeric lipid species that can be found
in biological systems. The main advantages and limitations of the recent analytical
Table 12.1 Main advantages and limitations of the recent analytical approaches used in lipidomics
Analytical approach Main advantages Main limitations References
Shotgun lipidomics Simple and fast approach Ion suppression causes Gross and Han
Use of high-resolution that compounds present [29]; Isaac [30]
mass spectrometers in trace amounts are
(QTOF-MS, FT-MS) often not detected
Novel hybrid high- Its applicability to the
resolution instruments, search of novel,
such as a combination of previously unknown
ion mobility (IM) and lipids is relatively
TOFMS, are suitable for restricted
the shotgun approach
Lipidomics by Versatile (widely used for Matrix effects (although Nygren et al. [31]
ultrahigh both targeted and less than in shotgun
performance liquid nontargeted analyses, methods), as it is not
chromatography using various types of possible to use labelled
coupled to mass mass spectrometers) standards for all
spectrometry Typically high sensitivity compounds
(UHPLC–MS) (especially when used in Memory effects due to,
targeted approaches e.g., sample carryover
(picomoles))
High-throughput analyses
with high separation
efficiency in a short
analysis time
Identification of novel
lipids is possible
Lipidomics by Ion mobility analysis has Expensive Kliman et al. [32]
structurally selective the ability to differentiate instrumentation
ion mobility analytes which are
coupled to mass isobaric in mass but differ
spectrometry in structure
(IM-MS) Promising role for
fundamental lipid
characterization in future
applications
(continued)

Analytical approach Main advantages Main limitations References
Comprehensive Combination of the Ideally, comprehensive Guo and
multidimensional advantages of both multidimensional Lankmayr [33];
approach high-resolution approaches are expected Han et al. [34]
instrumentation (i.e., to inherit the advantages
Orbitrap MS, FT ion of the existing
cyclotron resonance) methodologies and
and a two-dimensional overcome the
chromatographic limitations of any
separation individual, which
High resolving power further develop in the
Separation of isomers, future
conformers, and Complicated
enantiomers instrumentation not well
suited for routine
analysis
Nuclear magnetic Quantitative, Rather modest Lindon and
resonance (NMR) nondestructive technique sensitivity Nicholson [35];
spectroscopy Useful in elucidation of The similarity of the Maher et al. [36]
molecular structures of spectra of lipids with
purified lipids and respect to the limited
structural analysis structural carbon chain
High-resolution information is
magic-angle-spinning challenging
(HR-MAS) NMR NMR interpretation is
spectroscopy potential use complicated by the
for metabolic profiling of considerable number of
intact tissues spin-coupled multiplets
that result in spectral
crowding
approaches used in lipidomics are summarized in Table 12.1. The application of

sophisticated histochemical, cytochemical, and physical imaging techniques [37] is
also expected to contribute by adding valuable information in lipid localization in
different cells and tissues.
As in other “omics” disciplines, data analysis plays a key role in lipidomics.
Particularly in the global profiling approaches, the amount of data is large and it is
challenging to interpret the data without proper bioinformatics. Advanced statistical
analysis tools and strict quality assurance regimens are a must to provide reliable
and meaningful lipidomic results. Before any statistical analysis can even be per-
formed, data preprocessing is required, including signal processing, data normaliza-
tion, and transformation, so that the raw data signals are transformed into the format
which can be used for the statistical data analysis. Some techniques utilized in
lipidomics include artificial neural networks, self-organizing maps, and linear

discriminant analysis, among many others [38–40].
12.3 Gut Microbiota and Host Lipid Metabolism
Research over the past decade has accumulated a large body of evidence linking
alterations in the gut microbiota composition to several diseases [5, 6]. The gut
microbiota is indeed currently accepted as an environmental factor that affects host
metabolism and contributes to major disorders burdening the healthcare systems
today, such as metabolic syndrome, cardiovascular diseases, and type 2 diabetes.
Recent applications in relation to metabonomics and gut microbiota in diabetes are,
for instance, discussed in detail in Chap. 12. From a mechanistic point of view, only
a few recent studies have identified some key signaling pathways of the cross-
species homeostatic regulation between the gut microbiota and its host [5]. In this
section, the impact of the gut microbiota in lipid host metabolism is addressed from
evidence accumulated in both animal and human studies.
12.3.1 In Vivo Studies
In the last decade, the combination of gnotobiotics (the study of animals living in a
microbiologically defined environment, either germ-free or colonized with known
bacteria) and metagenomics (the study of the metagenome, i.e., the collective
genomic content of a microbiota) has elegantly begun to decipher the role of the gut
microbiota in host metabolism and host physiology.
Bäckhed and colleagues [41] were the first to suggest that gut microbiota is an
environmental factor affecting adiposity. The authors used normal and genetically
engineered gnotobiotic mice (germ-free (GF), conventionally raised (CONV-R),
and conventionalized (CONV-D) mice) to test the hypothesis that the microbiota
acts through host signaling pathways to regulate energy storage in the host. GF mice
are raised in the absence of any microbiota, while CONV-D mice are initially germ-
free but then colonized with the microbiota from CONV-R donors. The convention-
alization of adult germ-free (GF) mice with a normal microbiota harvested from the
distal intestine of CONV-R mice resulted in 60 % increase in body fat content
despite reduced food intake [41]. Conventionalization was accompanied by a
significant increase in two liver mRNAs encoding key enzymes in the de novo fatty
acid biosynthetic pathway, acetyl-CoA carboxylase (Acc1), and fatty acid synthase
(Fas). Both enzymes are known targets of the two transcriptor factors, SREBP-1
(sterol response element binding protein 1) and ChREBP (carbohydrate response
element binding protein), which mediate hepatocyte lipogenous responses to insulin
and glucose [42]. Conventionalization clearly increased liver ChREBP mRNA and
to a lesser extent also SREBP-1 mRNA levels. Additionally, the presence of the
microbiota promoted increased monosaccharide uptake from the gut, their increased
delivery to the liver, and a resulting induction of de novo hepatic lipogenesis.
Lipoprotein lipase (LPL) is a key regulator of fatty acid release from
triglyceride-rich lipoproteins in fat and muscle. Increased adipocyte LPL activity
leads to increased cellular uptake of fatty acids and adipocyte TG accumulation.
Fasting-induced adipocyte factor (Fiaf), a member of the angiopoietin-like family
of proteins, is produced by brown and white fat, liver, and intestine [43].
Interestingly, Fiaf was selectively suppressed in the intestine epithelium of con-
ventionalized mice. By using normal and Fiaf knockout mice, Fiaf was estab-
lished as a circulating LPL inhibitor, highlighting that its suppression is essential
for the microbiota-induced deposition of triglycerides in adipocytes [41]. The
relevance of Fiaf expression, which is selectively suppressed in the gut epithelium
by the microbiota, was established when Fiaf-deficient mice fed a Western diet
gained significantly more weight and had significantly greater epididymal fat pads
than their wild-type littermates [44].
The levels of leptin, which is an adipocyte-derived hormone known to reduce
food intake and increase energy expenditure, were increased upon colonization and
were proportional to the observed increase in body fat in [41]. The ob/ob mouse
model, characterized by increased food consumption due to leptin deficiency, was
then firstly used to provide evidence that the obesity-associated gut microbiome has
an increased capacity for energy harvest from the diet [45]. The same group had
shortly before revealed that the two most abundant bacterial divisions in mice were
the phyla Firmicutes and Bacteroidetes and that their proportions were increased
and reduced, respectively, in obese mice relative to their lean counterparts [46] in a
study that can be considered the first approach of DNA sequencing focused on the
16S rRNA gene in the context of obesity. The next step was to apply shotgun pyro-
sequencing technology in the same mouse model [45]. In addition, in order to
confirm the increased ratio of Firmicutes to Bacteroidetes in the obese mice, the
predicted increased capacity for dietary energy harvest by the ob/ob microbiome
was validated using biochemical assays and by transplantation of lean and obese
cecal microbiotas into GF wild-type mouse recipients.
Interestingly, the persistently lean phenotype in GF animals, which are protected
against the obesity that develops after consuming a Western-style, high-fat, sugar-
rich diet, was associated with increased skeletal muscle and liver levels of phos-
phorylated AMP-activated protein kinase (AMPK) and its downstream targets
involved in fatty acid oxidation, acetyl-CoA carboxylase, and carnitine palmitoyl-
transferase (Cpt1) [44]. AMPK is a heterotrimeric enzyme that functions as a “fuel
gauge” monitoring cellular energy status, which is activated in response to meta-
bolic stresses that result in an increased intracellular ratio of AMP to ATP, such as
exercise, hypoxia, and glucose deprivation [47] and adipocyte-derived leptin levels
[48]. All together, the findings reported by Gordon’s group indicated that gut
microbes can affect both sides of the energy balance equation, as a factor that influ-
ences the harvest of energy from components of the diet and as a factor that affects
host genes that regulate how energy is expended and stored.
Wikoff and colleagues [49] were among the first demonstrating the large effect
of the microbiome on mammalian plasma metabolic profile. The untargeted
MS-based profiling of serum from GF and CONV-R mice showed that concentra-
tions of many circulating metabolites are affected by the presence of the microbi-
ome. Several pathways including the metabolic processing of indole-containing
molecules were seen to particularly interact with the microbiome. Multiple organic
acids containing phenyl groups were also greatly increased in the presence of gut
microbes, and a broad, drug-like phase II metabolic response of the host to metabo-
lites generated by the microbiome suggested that the gut microbiota has also a
direct impact on the drug metabolism capacity of the host [49]. In another study,
global analysis of polar metabolites and molecular lipids in serum, white adipose
tissue, and liver of GF and CONV-R mice was applied to delineate how the gut
microbiota affects host energy and lipid metabolism [50]. Analysis of the serum
metabolome showed that energy metabolites were increased in CONV-R mice,
which was consistent with higher-energy metabolism in the presence of gut micro-
biota [51]. Notably, lipidomic analysis highlighted systemic effects of gut micro-
biota on host lipid metabolism, especially in a large number of individual TG
species and in several phosphatidylcholine species. TG levels were lower in serum
but higher in adipose tissue and liver of CONV-R mice, consistently with increased
lipid clearance. Thus, the study demonstrated that gut microbiota affects both
energy-storing and signaling lipids.
The investigation of the microbial communities from both wild-type and
resistin-like molecule (RELM) β knockout mice fed a standard chow diet and a
high-fat diet [52] demonstrated the importance of diet as a determinant of gut
microbiome composition and suggests the need to control for dietary variation
when evaluating the composition of the human gut microbiome. The expression of
the RELMβ gene has been shown to be dependent on the gut microbiome and can
be induced by a high-fat diet [53]. In Hilderbrant et al. study [52], the RELMβ
knockout mice consuming the HFD remained lean, whereas the corresponding
wild-type mice became obese. Higher levels of RELMβ expression were observed
in HFD mice when compared with mice fed a standard chow diet. Further analysis
also revealed that the expression of a collection of genes encoding ABC transport-
ers was increased in wild-type mice fed the HFD when compared with expression
of the same genes in wild-type mice on a standard chow diet. The corresponding
proteins are responsible for the transport of lipids, sugars, and peptides. Altogether
and because the general changes in the composition of the gut microbiota were
similar in both types of mice, the authors concluded that the HFD, and not the
obese state, accounted for the alteration in the gut microbial communities, high-
lighting therefore the dominant effect of diet.
Research in Nicholson’s group added further knowledge to understand the
bidirectional interaction between the host metabolism and its symbionts. Firstly,
the metabolic phenotypes of GF and conventional C3H mice were characterized
by applying an NMR-based metabolic profiling, providing evidence that the
microbiota status affects the systemic metabolism of host modulating the metabolic
fingerprint of topographically remote organs such as the liver and the kidney [54].
In a second study, the adaptive mechanisms of gut colonization by microbiota

using a similar systems biology approach in the same mouse strain were explored
by combining NMR-based profiling to gut microbial monitoring by 16SrRNA
gene pyrosequencing [55]. This kind of approach allows the simultaneous study
of the composition of the microbial ecosystem with the modifications of the host
metabolism induced by the colonization process. In particular, the study focused
on the evolution of liver metabolism, but similar research in other tissue-specific
metabolism would highly benefit the global understanding of the complex lipid
host metabolism.
Claus et al. study [55] findings can be considered as the first evidence of an
in vivo association between a family of bacteria and hepatic lipid metabolism.
Acquisition of the gut microbiota was associated with rapid increase in body weight
over the first days of colonization with parallel changes in multiple pathways in all
compartments analyzed (liver, kidney, colon, urine, and plasma). The colonization
process stimulated glycogenesis in the liver prior to triggering TG increases in
hepatic TG synthesis. These changes were associated with modifications of the
expression of hepatic Cyp8b1 (sterol 12α-hydroxylase), and the subsequent altera-
tion of bile acid metabolites, which are essential regulators of lipid absorption, and
will be further discussed later in this chapter. Statistical regression OPLS-based
models between hepatic metabolic profiles and microbial composition revealed
strong associations of the Coriobacteriaceae family with hepatic TG, glucose, and
glycogen levels. Two bacterial phyla (Actinobacteria and Tenericutes) were signifi-
cantly predicted by the liver metabolic profiles and were both associated with high
hepatic levels of TG and low hepatic levels of glycogen and glucose. A complemen-
tary study performed on hamsters [56] previously reported a strong correlation
between unidentified bacteria of the same Coriobacteriaceae family and non-HDL
plasma cholesterol when the metabolism was challenged using grain sorghum lipid
extract to improve the HDL/non-HDL ratio.
Brown adipose tissue (BAT) is a central organ involved in the regulation of
energy expenditure in mice. BAT is increasingly considered as a new target in the
battle against obesity after recent discovery of BAT in humans [57, 58]. BAT and
its specific uncoupling protein UCP1 have been linked to the development of obe-
sity in C57B1/6J mice ablated for the UCP1 [59]; and interestingly, the regulator
role that BAT plays for TG-rich lipoprotein clearance and for the control of blood
lipid abundance has been recently demonstrated [60], highlighting the potential
role of BAT, with its main depots localized in the supraclavicular and neck regions,
for reducing the risk of metabolic syndrome [61]. In this context, an NMR
spectroscopy-based metabolic profiling approach was applied in order to investi-
gate the influence of GF state and gender on energy metabolism in urine, plasma,
liver, and BAT of C3H mouse [62]. Interestingly, conventional male mice had a
significantly higher total body fat content compared to conventional female mice,
whereas this sexual dimorphism disappeared in GF animals. Among the metabolite
changes, choline, phosphocholine, and glycerophosphocholine were observed in
higher levels in GF males’ BAT metabolic profiles when compared to their conven-
tional counterparts. As it is well known, these metabolites are involved in the
formation and maintenance of cellular membranes through structural lipids, and

according to [62], their observed increased levels may indicate a variation of brown
adipocyte sizes between GF and conventional male animals. It is then likely that
conventional BAT contains larger mature adipocytes and GF display a higher
number of small undifferentiated preadipocytes.
Higher levels of (D)-3-hydroxybutyrate and lower levels of lactate were also
observed in GF mice compared to their conventional counterparts. On one hand, a
strong elevation of (D)-3-hydroxybutyrate was also observed in plasma and liver of
the GF mice. 3-hydroxybutyrate is the major ketone body produced in the mitochon-
dria of adipocytes and hepatocytes, initiated by the condensation of two molecules of
acetyl-CoA derived from the beta-oxidation of lipids. This ketone body plays an
important role in the host energy metabolism, acting on noradrenaline receptors to
inhibit BAT thermogenesis and to regulate appetite in mice [63]. On the other hand,
adipose tissue is a major site of glucose conversion to lactate, and lactate overproduction
could be associated with metabolic abnormalities related to obesity development. In
addition, decreased circulating VLDL levels in GF mice suggested hepatic inhibition
of lipogenesis. Altogether, the findings in [62] indicate that GF BAT overactivated
the catabolism of lipids compared to conventional animals.
Recently, fatty liver showed to accompany an increase in Lactobacillus species
in the hindgut of C57BL/6 mice fed an HFD [64]. The C57BL/6 mouse model of
obesity uses a diet high in total fat, and particularly n-6 FA, to produce outcomes
similar to those observed in obese humans, namely, increased adiposity, production
of proinflammatory cytokines, and fatty infiltration of the liver. Emerging evidence
suggests that hindgut microbiota may contribute to liver pathology [65, 66]. It is
well known that HF feeding increases the secretion of bile acids, which are impor-
tant regulators of hepatic lipid metabolism that are believed to be a determinant of
the gut microbiota in response to HFD [67]. The obesity-related inflammatory fatty
liver caused by the HFD in the previously mentioned model was interestingly
accompanied by a large increase in hindgut L. gasseri and/or L. taiwanensis, both of
which are part of L. acidophilus species group of bacteria. Both bacteria have been
recently suggested to play a role in body weight control [68]. Further studies with
GF mice are therefore needed to determine the mechanistic role of these bacteria in
the development of inflammatory liver fat due to HF feeding.
These last 10 years of gnotobiotics research have highly contributed in the under-
standing of the interactions of microbiota and host lipid metabolism (Fig. 12.4).
Nevertheless, it is important to point out that the results obtained in gnotobiotics-
based studies cannot be automatically extrapolated to humans, and it still remains
unclear if an altered microbiota associated with a disease in humans is causing,
contributing, or merely a consequence of the disease state. Cheaper sequencing and
improved bioinformatics tools for the analysis of the gut microbiota, together with
the recent advances in detailed lipid characterization, are expected to contribute to
the future studies that are still needed to elucidate the complete mechanisms con-
necting microbiota and human metabolism.
↑ Insulin sensitivity
↓ Lipogenesis Liver
↓ Glycogenesis
↓ TG synthesis
Gall bladder
↑ BA concentration in bile Portal vein
Jejunum
↑ Levels of conjugated BA
↑ Reabsorption of BA Ileum
Colon
↓ Levels of SCFA
↑ Lipid excretion
↓ BA excretion
Fig. 12.4 Effect of microbiota in lipid metabolism of germ-free mice (GF) vs. conventionally
raised (CONV-R) mice according to gnotobiotics studies. Arrows mean increased or decreased
levels in GF vs. CONV-R mice
12.3.2 Human Studies
Accumulating evidence indicates that the gut microbiota may be a future target for
treating metabolic diseases [69]. Supplementing the diet with nondigestible foods, or
prebiotics, that stimulate the expansion of specific microbes to improve metabolic
regulation can be a nutritional therapy for overweight and obesity management [70].
However, to determine the effects of these therapies, appropriate human intervention
studies are still required. Human studies are helping to show what effect the gut
microbiota has on host metabolism by improving energy yield from food and modu-
lating dietary or the host-derived metabolites that alter host metabolic pathways. Due
to the heterogeneous etiology of lipid metabolism-related diseases, such as obesity
and diabetes, the role of gut microbiota in the development of these disorders is still
unclear. Many studies so far are underpowered, include participants with diverse
ethnic origin and dietary habits, and have used different methods with specific biases
to profile the microbiota. In addition, studies in humans tend to be associative, so the
mechanistic role of the microbiota in obesity and its comorbidities in humans remains
to be proven. Relevant human studies in the field are briefly discussed in this section.
Results from bariatric surgery studies are not included as they are discussed in detail
in another chapter of this book.
Gut microbiota composition is altered in people who are obese, and it can
respond to changes in body weight. Ley and colleagues [71] revealed that the
Bacteroidetes and Firmicutes divisions dominate the human microbiota, with obese
subjects having lower levels of Bacteroidetes and higher levels of Firmicutes than
lean subjects. Later on, a study with monozygotic and dizygotic twins and their
mothers showed that the composition of the gut microbiota is more similar between
family members than unrelated individuals [3]. Each individual’s gut microbiota
was thought distinct, and a similar degree of covariation existed between adult
monozygotic and dizygotic twin pairs. The previously reported lower levels of
Bacteroidetes in obese than lean subjects were confirmed, but no significant differ-
ences in Firmicutes levels were detected. Remarkably, the microbial population was
in general less diverse in obese individuals.
Some other studies have shown discrepancies in the Firmicutes and Bacteroidetes
proportions with respect to obesity in humans [72, 73]. The different clinical
criteria, such as the level of obesity, degree of weight loss, and duration of caloric
restriction, together with different geographical locations, ages, population sizes,
and microbiota-profiling methodologies, can be responsible of the differences
observed between studies. A low-cost clinical method for monitoring the variations
of bacterial phyla of the gut using real-time PCR assay was able to confirm a reduc-
tion in the Bacteroidetes community in obese subjects and found an increase in
Lactobacillus species in obese subjects and an increase in methanogens (M. smithii)
in anorexic patients [74]. This preliminary data that links Lactobacillus levels with
obesity needs further study with other advanced methodologies.
Consumption of lactic acid bacteria marketed as probiotics is indeed a common
approach to maintain health [75]. Lactobacillus rhamnosus GG is one of the most
widely used probiotic bacteria that is assumed to interact with the host via binding
to human mucus via its extracellular pili [76]. However, further molecular details of
probiotics signaling are not yet understood. In a recent study, high-throughput
screening of the intestinal microbiota was performed using a phylogenetic HITChip
microarray and qPCR methodology and integrated with serum lipidomic profiling
data to study the impact of probiotic intervention on the intestinal ecosystem and to
explore the associations between the intestinal bacteria and serum lipids [77].
Healthy subjects received either L. rhamnosus GG or placebo for a 3-week period
following a randomized, double-bind intervention design. While a specific increase
in the L. rhamnosus-related bacteria was observed during the intervention, no other
changes in the composition or stability of the microbiota were detected. The most
prevailing association between the gut microbiota and lipid profiles was a strong
positive correlation between uncultured phylotypes of Ruminococcus gnavus group
and polyunsaturated serum TG species of dietary origin. Actinomycetaceae corre-
lated negatively with TG of highly unsaturated FA while a set of Proteobacteria
showed negative correlation with ether PCs. Altogether, these results suggest that
several members of the Firmicutes, Actinobacteria, and Proteobacteria may be
involved in the metabolism of dietary and endogenous lipids. This data supports the
concept that the overall lipid content in human serum is a composite of host and
microbial metabolic activity, and the intestinal commensals are implicated in the
metabolism of various lipid species that the human body uses for membranes,
energy storage, and signaling. Considering that a single gene in an intestinal bacte-
rium could alter host FA composition [78], the potential metabolic capacity and the
functional consequences from the million genes in the intestinal microbiome are
overwhelming. Future studies combining metagenomics and lipidomics, involving
controlled diet and dyslipidemic subjects, are indeed needed to provide further
insights on the role of intestinal microbiota on human lipid metabolism.
The human intestinal microbiota is immensely complex and includes thousands
of species that have a collective genome of close to five million genes. High through-
put of this metagenome is increasingly replacing the characterization of individual
microbes [79]. Notably, the MetaHIT Consortium was the first in addressing the
feasibility of comparative metagenomics of the human gut across cohorts and pro-
tocols and in obtaining first insights into commonalities and differences between
gut microbiomes across different populations [80]. They sequenced 22 European
metagenomes from Danish, French, Italian, and Spanish subjects and combined
them to existing Japanese [81] and American [3, 82] datasets. Three enterotypes
were identified on the basis of variations in the relative levels of Bacteroidetes,
Prevotella, and Ruminococcus, which were not nation or content specific.
Interestingly, the Enterotype 1 is enriched in Bacteroides and the co-occurring
Parabacteroides, which derived energy mainly from carbohydrates and proteins by
fermentation [83]. Enterotypes appear complex, but there are functional markers
such as genes or modules that correlate remarkably well with individual features.
This might potentially be used for diagnostic and prognostic tools for numerous
human disorders, including those related to lipid metabolism.
The combination of metagenomic analysis with clinical data is the base of the
recently emerging metagenome-wide association studies (MGWAS), which are
important contributions to reveal the associations of gut microbiota with health and
disease. To date, the metagenomes of relatively few individuals have been sequenced.
Recently, Karlsson and colleagues [84] applied shotgun sequencing to character-
ize the fecal metagenome of 145 European 70-year-old women with normal,
impaired, or diabetic glucose control. Interestingly, the study reported composi-
tional and functional alterations in the metagenomes of women with type 2 diabetes,
such as increases in the abundance of four Lactobacillus species and decreases in
the abundance of five Clostridium species. In the total cohort, Lactobacillus species
correlated positively with fasting glucose and glycosylated hemoglobin (HbA1c),
which is a long-term measure of blood glucose control. By contrast, Clostridium
species correlated negatively with fasting glucose, HbA1c, insulin, C-peptide, and
plasma TG and positively with adiponectin and HDL. These correlations are relevant
for T2D because high TG and low HDL levels are components of the dyslipidemia
typically found in T2D, and reduced levels of adiponectin have been reported in
people at risk of T2D [85]. Interestingly, the authors developed a mathematical
model based on metagenomic clusters (MGC) to test whether the microbiota
composition can identify diabetes status. MGC identified T2D more accurately than
species, indicating that several important gut species still need to be characterized.
L. gasseri had the highest score for the identification of T2D women in the model.
Notably, the model identified Roseburia and Faecalibacterium prausnitzii as highly
discriminant for T2D. These bacteria are known human gut colonizers and butyrate
producers [86] and have been linked to improved insulin sensitivity and diabetes
amelioration in studies of the human fecal microbiota [87, 88]. Gut microbiota
transplantations from lean donors to recipients with metabolic syndrome have
indeed been shown to increase Roseburia and butyrate levels together with improved
insulin sensitivity [87].
The pathways that showed the highest scores for enrichment in T2D metage-
nomes included KEGG orthologues for starch and glucose metabolism, fructose
and mannose metabolism, and ABC transporters for amino acids, ions, and simple
sugars. These findings agree with previous studies showing an increase in microbial
functions for energy metabolism and harvest in the obese microbiome [45]. Other
metabolic pathways containing KEGG orthologues enriched in women with T2D
included glycerolipid metabolism and fatty acid biosynthesis [84]. When applying
the model to a recently described Chinese cohort [89], the authors observed that the
most discriminatory MCG differed between the European and Chinese T2D cohorts.
This observation underscores the need to sample human populations and perform
parallel studies in different continents.
Given that the cost for sequencing is rapidly declining, the major challenge in
metagenome studies will be data analysis rather than data generation [90].
Detailed studies on the metagenome early in life as well as regional metagenomes
will be needed in the future to determine whether or not it can be programmed.
In addition, future studies might benefit from the use of tracers to look at lipid
metabolic functions in order to have a readout to investigate host-bacterial co-
metabolism and signaling.
12.4 Microbiota and Host Lipid Metabolism: Potential Role

of Signaling Molecules
Animal and in vitro studies have shown that the intestinal microbiota can regulate
host lipid metabolism via numerous microbial activities [91]. So far, short-chain
fatty acids (SCFA) have shown to be key microbial products with multiple effects
on host metabolism, while the best characterized mechanism of microbiota and host
lipid metabolism is through the biotransformation of bile acids [92]. Combination
of metagenomics and metataxonomics with comprehensive metabolomics of bioflu-
ids and stool samples has a potential to identify novel metabolites associated with
specific microbes. These could be functionally studied in vitro and in vivo for their
potential role in the regulation of host metabolism.
Complex carbohydrates
PA influences the production of INITIATION OF
hormones by AT and liver ENDOCRINE
Colonic AND/OR
(e.g. leptin)
microbiota NEURONAL SIGNALS
SHORT-CHAIN
FATTY ACIDS
LIVER ↓ FOOD INTAKE
BUTYRIC PROPIONIC
↓ BODY WEIGHT
ACID ACID (PA)
↑ ENERGY EXPEDITURE
ACETIC ADIPOSE
ACID TISSUE (AT)
COLON
PA reduces the production and INHIBITION OF

release of FA from AT and liver INFLAMMATION
Fig. 12.5 Main metabolic effects of the short-chain fatty acid propionic acid (PA) produced by
colonic fermentation
12.4.1 Microbiota and Short-Chain Fatty Acids
Short-chain fatty acids are clearly one of the most important microbial products and
have shown to affect a range of host processes, including energy utilization, host-
microbiome signaling, and control of colonic pH [93]. Complex carbohydrates can
be digested and subsequently fermented in the colon by gut microbes into SCFA,
such as acetate, propionate, and butyrate. Their profile in the gut reflects the meta-
bolic cooperation between different microbial types because no genus of bacteria
can hydrolyze all substrates and none produce all four SCFA upon carbohydrate
fermentation [94]. SCFA have different metabolic features. Among them, butyrate
is used as an energy substrate for colonocytes, whereas acetate is potentially used as
a cholesterol or fatty acid precursor [95]. Propionate is a gluconeogenetic substrate
in the liver, but it may also counteract de novo lipogenesis from acetate or glucose
in the same tissue [96] (Fig. 12.5).
More recently, SCFA have been identified as the physiological ligands of two
receptors for fatty acids, the G protein-coupled receptors GPR41 and GPR43, which
are expressed in several cell types (immune cells, endocrine cells, and adipocytes)
and in a wide variety of host tissues [97]. Both receptors show differences in SCFA
specificity, intracellular signaling, and tissue localization [98]. GPR41 knockout
mice colonized with a specific fermentative microbial community have shown to
resist fat mass gain compared to their wild-type littermates [99]. By using GPR45
knockout animals, it was shown that the activation of GPR43 in adipocytes leads to
inhibition of lipolysis and results in the reduction of plasma FFA levels in vivo [100].
These findings therefore suggest a potential role for GPR43 in regulation of plasma
lipid profiles, but its use as a potential target for the treatment of dyslipidemia
requires further research. By using GPR43-deficient mice fed an HFD, the activa-
tion of GPR43 by the acetate and propionate has also shown to contribute to adipo-
cyte differentiation [101], thereby promoting the expansion of adipose tissue, which
is in agreement with a gnotobiotics studies performed by Bäckhed et al. [41].
Interestingly, studies adding fermentable carbohydrates with prebiotic proper-
ties, such as insulin-type fructans, into the diet did not increase but even lessened fat
mass development in obese mice and humans [102, 103]. Notably, the supplementa-
tion with those prebiotics blunted the overexpression of GPR43 occurring in HFD-
fed animals, a phenomenon that contributes to lower adiposity [104]. It is well
known that specific changes in the gut microbiota composition by using prebiotics
strongly promote SCFA production [105]. Studies related to the potential of
metabonomics and metagenomics to promote health via prebiotics are discussed in
Chap. 9. Altogether the previous studies support the idea that SCFA coming from the
gut act not only as energy substrates but also as important metabolic regulators.
12.4.2 Microbiota and Bile Acid Metabolism
The interaction between bile acids and gut microbiota is complex. However, recent
studies have added novel insights into the regulation of BA metabolism by gut
microbiota. Bile acids (BA) play an important role in lipid metabolism. They func-
tion as detergents by forming micelles that facilitate solubilization, digestion, and
absorption of dietary lipids and lipid-soluble vitamins and represent the major route
of cholesterol excretion, which is critical in whole body sterol metabolism [106].
Remarkably, recent research has highlighted the role of BA as signaling molecules.
The discovery that specific BA differentially activates three nuclear receptors,
namely, farnesoid X receptor (FXR), pregnane X receptor (PXR), and vitamin D
receptor (VDR), and one G protein-coupled receptor (TGR5) identified BA as
hormones that alter multiple metabolic pathways [107]. The activation of these
receptors alters gene expression in multiple tissues, leading to changes not only in
BA metabolism but also in glucose homeostasis, lipid and lipoprotein metabolism,
energy expenditure, inflammation, and liver regeneration processes.
Bile acids are produced in hepatocytes, stored in the gallbladder, and released
into the duodenum upon ingestion of food. After having traveled through the small
intestine, >95 % of all liver-secreted BA are reabsorbed in the ileum to be taken up
by the liver in a process known as enterohepatic circulation [108]. Only a small part
of the BA pool escapes the enterohepatic cycle and travels toward the large intestine
to be excreted in the feces. This excretion is accompanied by microbial deconjuga-
tion of glycine(predominant in humans)- and taurine(predominant in mice)-conju-
gated bile acids [54]. Intestinal microbiota readily deconjugate and 7-α-dehydroxylate
the primary BA (cholic acid (CA) and chenodeoxycholic acid (CDCA) in humans,
and CA and β-muricholic acid (βMCA) in mice [109], converting them into secondary
BA (mainly DCA, UDCA, and LCA in humans) [109] (Fig. 12.6).
O O
R4 R4
R5 R5
Gut microbiota
R1 R3 R3
R1
R2
R2
PRIMARY BILE ACID SECONDARY BILE ACID

Cholic acid: Deoxycholic acid:
R1,3,4,5 = -OH R1,4,5 = -OH
R2 = -H R2,3 = -H
Fig. 12.6 Conversion of primary bile acids to secondary bile acids by the action of gut microbiota.
The basic C24-bile acid structure is illustrated in the figure, showing the conversion of cholic acid
into deoxycholic acid
Recent animal studies have shown that the microbiome affects not only the
composition of the BA pool but also the expression of genes controlled by the
BA-activated nuclear receptor FXR [110]. Bile acid synthesis is indeed under
negative feedback control through activation of FXR in the ileum and liver [111].
At least 14 liver enzymes have been reported to be required in the BA synthesis
from cholesterol. Briefly, the rate-limiting enzyme cholesterol 7α-hydroxylase
(CYP7A1) initiates the classic pathway for BA synthesis, while CYP27A1 initiates
the alternative pathway [112]. Sterol 12α-hydroxylase (CYP8B1) is needed for
CA synthesis, and the hepatic expression of both, CYP7A1 and CYP8B1, is
regulated by FXR [113].
Already 40 years ago, Wostmann et al. showed that the BA concentration in bile
was three times increased and cholesterol absorption was 25 % greater in the
absence of gut microbiota in GF rats when compared to their conventionally raised
counterparts [114]. The cholesterol accumulation was thought to be due to an
increase in intestinal BA reabsorption. Follow-up studies supported Weismann’s
hypothesis by showing that GF animals have elevated levels of conjugated BA
throughout the intestine with no deconjugation and strongly decreased fecal excretion
[115]; and later it was also confirmed in studies with mice treated with antibiotics
[116]. Those previous studies showed that the gut microbiota reduces the bile acid
pool size with its greatest effect on βMCA rather than CA levels. However, the
molecular mechanisms for how the gut microbiota suppresses BA synthesis are
currently unknown. A comprehensive profile of the BA composition of CONV-R
and GF mice through the enterohepatic system and in serum identified a profound
role of the gut microbiota not only on secondary BA but also as regulator of hepatic
BA synthesis [108]. In addition to confirm a dramatic reduction in MCA, but not CA,
levels in CONV-R mice, rederivatization of FXR-deficient mice as GF demonstrated
that the gut microbiota regulated expression of fibroblast growth factor 15 (Fgf15)
in the ileum and CYP7A1 in the liver, by FXR-dependent mechanisms. Remarkably,

tauro-conjugated beta- and alpha-MCA were identified as potent FXR antagonist,
proposing that the higher levels of MCA in GF mice at least partially account for the
lower expression of FXR-dependent genes in the ileum of GF mice [117].
In addition to FXR-based studies, TGR5 is the only G protein-coupled receptor
that has been reported to respond to BAs by the production of cAMP and the subse-
quent activation of PKA signaling pathways [118]. By treating brown adipocytes
and human skeletal myocytes with BA, Watanabe et al. [119] showed an increase in
type 2 iodothyronine deiodinase (D2) activity and oxygen consumption, highlight-
ing that bile acids can induce energy expenditure by promoting intracellular thyroid
hormone activation. Since then, the BA–TGR5–cAMP–D2 signaling pathway has
been considered as a key mechanism for fine-tuning energy homeostasis that can be
targeted to improve metabolic control.
Interestingly, a recent study in which treatment of primary rodent hepatocytes
with conjugated BA led to activation of extracellular regulated kinase (ERK) 1/2
and protein kinase B (AKT), in a sphingosine-1-phosphate receptor-dependent
pathway [120], suggests that other BA receptors remain to be identified. The
generation of new knockout mice lacking potential BA receptors is therefore
needed. Nevertheless, accumulating evidence has proved that the modulation of
FXR and TGR5 activity either directly by BA or pharmacological compounds or
indirectly by intestinal BA sequestration has helped to unravel the function of
these BA receptors in metabolic control. Consequently, both receptors might be
promising targets for the treatment of metabolic disorders associated with the
metabolic syndrome as recently reviewed by Porez et al. [121]. Although some
clinical trials in phases I–III have been already performed by using synthetic
FXR agonists, large-scale clinical trials will be needed to objectively assess their
therapeutic possibilities in the treatment of type 2 diabetes, metabolic syndrome,
or nonalcoholic steatohepatitis.
The understanding of the role of BA plays in human health and disease will
also likely benefit from recent advances in technology that enables genome-wide
association studies. Some inflammatory genes has been, for instance, identified
in the GWAS for primary biliary cirrhosis [122, 123], and some genes involved
in BA metabolism have been associated with other traits such as the association
of CYP7A1 with total and LDL cholesterol [124]. Further studies will probably
open up new insights for discovery of novel genes involved in BA metabolism.
Similar scientific strategies may also be applied to study the role of other bioac-
tive lipids in the regulation of host lipid metabolism. Comprehensive metabolo-
mics of biofluids and stool samples, associated with genomic and metagenomic
strategies, may help identify the microbes as well as genes associated with the
specific lipids. These lipids may then be studied in the context of their role in
health and disease, and mechanistic studies similarly as described above may
contribute to the elucidation of the mechanisms behind their regulation of host
metabolism.
12.5 Future Directions
In order to account for the enormous functional and structural diversity of lipids and
their complex regulation at multiple spatial and temporal scales, a systems biology
approach is needed for the study of lipids. Lipid signaling pathways are complex
and the therapeutic potential of modulation of intracellular and systemic lipid
metabolism is well recognized [10, 125]. Understanding the integrated lipidomic
networks and decoding the coordinately regulated pathways will therefore consti-
tute major goals for the following years in applied lipidomic research.
A better understanding of the lipidome at the physiological level lipids does not
only have to include lipid modelling at the level of biological pathways, but also at
the level of the biophysical systems the lipids are part of, such as cellular mem-
branes and lipoproteins particles [20]. Rapidly accumulating information about the
importance of gut microbiota in many lipid-related disorders together with the
advances in lipid analytical technologies and modelling approaches are likely to
contribute to better understanding of the role gut microbiota plays in the regulation
of cellular and systemic lipid metabolism, and vice versa, and how the dysregula-
tion of these physiological systems may contribute to many devastating diseases.
Future studies should rely on a systems medicine approach, where instead of
focusing on each disease individually, the aim is to account for the complex
gene–environment, socioeconomic interactions and comorbidities that lead to
individual-specific complex phenotypes. An in-depth understanding of the meta-
bolic phenotypes in health and disease is crucial if one is to implement personalized
medicine and nutrition. How different metabolic phenotypes can be implemented as
diagnostic tools in clinic needs therefore to be addressed at a clinical translational
research level, as pointed out in the last chapter of this book.
Acknowledgments I.B-P. is grateful to the Spanish Carlos III Health Institute for her Sara Borrell
postdoctoral fellowship.
References
1. Qin J, Li R, Raes J, et al. A human gut microbial gene catalogue established by metagenomic
sequencing. Nature. 2010;464:59–65.
2. Greiner T, Bäckhed F. Effects of the gut microbiota on obesity and glucose homeostasis.
Trends Endocrinol Metab. 2011;22:117–23.
3. Turnbaugh PJ, Hamady M, Yatsunenko T, et al. A core gut microbiome in obese and lean
twins. Nature. 2009;457:480–4.
4. Benson AK, Kelly SA, Legge R, et al. Individuality in gut microbiota composition is a com-
plex polygenic trait shaped by multiple environmental and host genetic factors. Proc Natl
Acad Sci U S A. 2010;107:18933–8.
lism. Nature. 2012;489:242–9.
6. Sommer F, Backhed F. The gut microbiota – masters of host development and physiology.
7. Ideker T, Galitski T, Hood L. A new approach to decoding life: systems biology. Annu Rev
Genomics Hum Genet. 2001;2:343–72.
8. Yetukuri L, Ekroos K, Vidal-Puig A, Oresic M. Informatics and computational strategies for
the study of lipids. Mol BioSyst. 2008;4:121–7.
9. Buckingham J, editor. Dictionary of natural products on CD-ROM, version 19.1. London:
Chapman and Hall; 2010.
10. Orešič M, Hänninen VA, Vidal-Puig A. Lipidomics: a new window to biomedical frontiers.
Trends Biotechnol. 2008;26:647–52.
11. Fahy E, Cotter D, Sud M, Subramaniam S. Lipid classification, structures and tools. Biochim
Biophys Acta. 1811;2011:637–47.
12. German JB, Gillies LA, Smilowitz JT, Zivkovic AM, Watkins SM. Lipidomics and lipid
profiling in metabolomics. Curr Opin Lipidol. 2007;18:66–71.
13. Shevchenko A, Simons K. Lipidomics: coming to grips with lipid diversity. Nat Rev Mol Cell
Biol. 2010;11:593–8.
14. Klose C, Ejsing CS, García-Sáez AJ, et al. Yeast lipids can phase-separate into micrometer-
scale membrane domains. J Biol Chem. 2010;285:30224–32.
15. Fahy E, Subramaniam S, Brown HA, et al. A comprehensive classification system for lipids.
J Lipid Res. 2005;46:839–62.
16. Wenk MR. Lipidomics: new tools and applications. Cell. 2010;143:888–95.
17. Jung HR, Sylvänne T, Koistinen KM, Tarasov K, Kauhanen D, Ekroos K. High throughput
quantitative molecular lipidomics. Biochim Biophys Acta. 1811;2011:925–34.
18. Murphy RC, Axelsen PH. Mass spectrometric analysis of long-chain lipids. Mass Spectrom
Rev. 2011;30:579–99.
19. Niemelä PS, Castillo S, Sysi-Aho M, Orešič M. Bioinformatics and computational methods
for lipidomics. J Chromatogr B. 2009;877:2855–62.
20. Orešič M. Informatics and computational strategies for the study of lipids. Biochim Biophys
Acta Mol Cell Biol Lipids. 1811;2011:991–9.
21. Gross R, Han X. Lipidomics at the interface of structure and function in systems biology.
Chem Biol. 2011;18:284–91.
22. Stein EA, Myers GL. National cholesterol education program recommendations for triglyc-
eride measurement: executive summary. The National Cholesterol Education Program
Working Group on lipoprotein measurement. Clin Chem. 1995;41:1421–6.
23. Kotronen A, Johansson LE, Johansson LM, Roos C, et al. A common variant in PNPLA3,
which encodes adiponutrin, is associated with liver fat content in humans. Diabetologia.
2009;52:1056–60.
24. Rhee EP, Cheng S, Larson MG, Walford GA, Lewis G, McCabe E, Yang E, Farrell L. Lipid
prediction in humans. J Clin Invest. 2011;121:1402–11.
25. Wang JH, Byun J, Pennathur S. Analytical approaches to metabolomics and applications to
systems biology. Semin Nephrol. 2010;30:500–11.
26. Navas-Iglesias N, Carrasco-Pancorbo A, Cuadros-Rodríguez L, From lipid analysis towards lipido-
mics, a new challenge for the analytical chemistry of the 21st century. TrAC. 2009;28:393–403.
27. Hyötyläinen T, Bondia-Pons I, Orešič M. Lipidomics in nutrition and food research. Mol
Nutr Food Res. 2013;57:1306–18.
28. Harkewicz R, Dennis EA. Applications of mass spectrometry to lipids and membranes. Annu
Rev Biochem. 2011;80:301–25.
29. Gross RW, Han X. Shotgun lipidomics of neutral lipids as an enabling technology for eluci-
dation of lipid-related diseases. Am J Physiol Endocrinol Metab. 2009;297:E297–303.
30. Isaac G. Electrospray ionization tandem mass spectrometry (ESI-MS/MS)-based shotgun
lipidomics. Methods Mol Biol. 2011;708:259–75.
31. Nygren H, Seppänen-Laakso T, Castillo S, Hyötyläinen T, Orešič M. Liquid chromatogra-
phy-mass spectrometry (LC-MS)-based lipidomics for studies of body fluids and tissues.
Methods Mol Biol. 2011;708:247–57.
32. Kliman M, May JC, McLean JA. Lipid analysis and lipidomics by structurally selective ion
mobility-mass spectrometry. Biochim Biophys Acta. 1811;2011:935–45.
33. Guo X, Lankmayr E. Multidimensional approaches in LC and MS for phospholipid bioanalysis.
Bioanalysis. 2010;2:1109–23.
34. Han X, Yang K, Gross RW. Multi-dimensional mass spectrometry-based shotgun lipidomics
and novel strategies for lipidomic analyses. Mass Spectrom Rev. 2012;31:134–78.
35. Lindon JC, Nicholson JK. Spectroscopic and statistical techniques for information recovery in
metabonomics and metabolomics. Annu Rev Anal Chem (Palo Alto Calif). 2008;1:45–69.
36. Maher AD, Lindon JC, Nicholson JK. 1H NMR-based metabonomics for investigating diabe-
tes. Future Med Chem. 2009;1:737–47.
37. Bou Khalil M, Hou W, Zhou H, et al. Lipidomics era: accomplishments and challenges. Mass
Spectrom Rev. 2010;29:877–929.
38. Katajamaa M, Orešič M. Data processing for mass spectrometry-based metabolomics. J
Chromatogr A. 2007;1158:318–28.
39. Sumner LW, Urbanczyk-Wochniak E, Broeckling CD. Metabolomics data analysis, visual-
ization, and integration. Methods Mol Biol. 2007;406:409–36.
40. Wheelock CE, Goto S, Yetukuri L, D’Alexandri FL, Klukas C, Schreiber F, Orešič
M. Bioinformatics strategies for the analysis of lipids. Methods Mol Biol. 2009;580:339–68.
41. Bäckhed F, Ding H, Wang T, et al. The gut microbiota as an environmental factor that regu-
lates fat storage. Proc Natl Acad Sci U S A. 2004;101:15718–23.
42. Dentin R, Pegorier JP, Benhamed F, Foufelle F, Ferre P, et al. Hepatic glucokinase is required
for the synergistic action of ChREBP and SREBP-1c on glycolytic and lipogenic gene
expression. J Biol Chem. 2004;279:20314–26.
43. Hooper LV, Wong MH, Thelin A, Hansson L, Falk PG, Gordon JI. Molecular analysis of
commensal host-microbial relationships in the intestine. Science. 2001;291:881–4.
44. Bäckhed F, Manchester JK, Semenkovich CF, Gordon JI. Mechanisms underlying the resis-
tance to diet-induced obesity in germ-free mice. Proc Natl Acad Sci U S A. 2007;104:979–84.
45. Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V, Mardis ER, Gordon JI. An obesity-associated
gut microbiome with increased capacity for energy harvest. Nature. 2006;444:1027–131.
46. Ley RE, Bäckhed F, Turnbaugh P, Lozupone CA, Knight RD, Gordon JI. Obesity alters gut
microbial ecology. Proc Natl Acad Sci U S A. 2005;102:11070–5.
47. Kahn BB, Alquier T, Carling D, Hardie DG. AMP-activated protein kinase: ancient energy
gauge provides clues to modern understanding of metabolism. Cell Metab. 2005;1:15–25.
48. Minokoshi Y, Kim YB, Peroni OD, Fryer LG, Müller C, Carling D, Kahn BB. Leptin stimulates
fatty-acid oxidation by activating AMP-activated protein kinase. Nature. 2002;415:339–43.
49. Wikoff WR, Anfora AT, Liu J, et al. Metabolomics analysis reveals large effects of gut micro-
flora on mammalian blood metabolites. Proc Natl Acad Sci U S A. 2009;106:3698–703.
50. Velagapudi VR, Hezaveh R, Reigstad CS, et al. The gut microbiota modulates host energy
and lipid metabolism in mice. J Lipid Res. 2010;51:1101–12.
51. Wostmann B, Larkin C, Moriarty A, Bruckner-Kardoss E. Dietary intake, energy metabolism,
and excretory losses of adult male germfree wistar rats. Lab Anim Sci. 1983;33:46–50.
52. Hildebrandt MA, Hoffmann C, Sherrill-Mix SA, Keilbaugh SA, Hamady M, Chen YY,
Knight R, Ahima RS, Bushman F, Wu GD. High-fat diet determines the composition of the
murine gut microbiome independently of obesity. Gastroenterology. 2009;137:1716–24.
53. Shojima N, Ogihara T, Inukai K, et al. Serum concentrations of resistin-like molecules are
elevated in high-fat-fed and obese db/db mice, with increased production in the intestinal
tract and bone marrow. Diabetologia. 2005;48:984–92.
54. Claus SP, Tsang TM, Wang Y, et al. Systemic multicompartmental effects of the gut microbi-
ome on mouse metabolic phenotypes. Mol Syst Biol. 2008;4:291.
55. Claus SP, Ellero SL, Berger B, et al. Colonization-induced host-gut microbial metabolic
interaction. MBio. 2011. doi:10.1128/mBio.00271-10.
56. Martinez AN, Lahiri R, Pittman TL, Scollard D, Truman R, Moraes MO, Williams
DL. Molecular determination of Mycobacterium leprae viability by use of real-time PCR. J
Clin Microbiol. 2009;47:2124–30.
57. Virtanen KA, Lidell ME, Orava J, et al. Functional brown adipose tissue in healthy adults. N
Engl J Med. 2009;360:1518–25.
58. Lidell ME, Enerback S. Brown adipose tissue – a new role in humans? Nat Rev Endocrinol.
2010;6:319–25.
59. Feldmann HM, Golozoubova V, Cannon B, Nedergaard J. UCP1 ablation induces obesity and
abolishes diet-induced thermogenesis in mice exempt from thermal stress by living at ther-
moneutrality. Cell Metab. 2009;9:203–9.
60. Bartelt A, Bruns OT, Reimer R, et al. Brown adipose tissue activity controls triglyceride
clearance. Nat Med. 2011;17:200–5.
61. Nedergaard J, Bengtsson T, Cannon B. New powers of brown fat: fighting the metabolic
syndrome. Cell Metab. 2011;13:238–40.
62. Mestdagh R, Dumas M, Rezzi S, et al. Gut microbiota modulate the metabolism of brown
adipose tissue in mice. J Proteome Res. 2012;11:620–30.
63. Canas X, Fernandez-Lopez J, Remesar X, Alemany M. 3-hydroxybutyrate inhibits
noradrenaline-induced thermogenesis in lean but not in obese zucker rats. Int J Obes Relat
Metab Disord. 1998;22:734–40.
64. Zeng H, Liu J, Jackson MI, Zhao F, Yan L, Combs GF. Fatty liver accompanies an increase in lac-
tobacillus species in the hind gut of C57BL/6 mice fed a high-fat diet. J Nutr. 2013;143:627–31.
65. Compare D, Coccoli P, Rocco A, et al. Gut–liver axis: the impact of gut microbiota on non
alcoholic fatty liver disease. Nutr Metab Cardiovasc Dis. 2012;22:471–6.
66. Quigley EMM, Stanton C, Murphy EF. The gut microbiota and the liver. Pathophysiological
and clinical implications. J Hepatol. 2013;58:1020–7.
67. Yokota A, Fukiya S, Islam KBMS, et al. Is bile acid a determinant of the gut microbiota on a
high-fat diet? Gut Microbes. 2012;3(5):455–9.
68. Million M, Angelakis E, Paul M, Armougom F, Leibovici L, Raoult D. Comparative meta-
analysis of the effect of lactobacillus species on weight gain in humans and animals. Microb
Pathog. 2012;53:100–8.
69. Murphy EF, Cotter PD, Hogan A, et al. Divergent metabolic outcomes arising from targeted
manipulation of the gut microbiota in diet-induced obesity. Gut. 2013;62:220–6.
70. Delzenne NM, Neyrinck AM, Cani PD. Gut microbiota and metabolic disorders: how prebi-
otic can work? Br J Nutr. 2013;109 Suppl 2:81–5.
ated with obesity. Nature. 2006;444:1022–3.
72. Duncan SH, Lobley GE, Holtrop G, et al. Human colonic microbiota associated with diet,
obesity and weight loss. Int J Obes. 2008;32:1720–4.
73. Schwiertz A, Taras D, Schäfer K, et al. Microbiota and SCFA in lean and overweight healthy
subjects. Obesity. 2010;18:190–5.
74. Armougom F, Henry M, Vialettes B, Raccah D, Raoult D. Monitoring bacterial community
of human gut microbiota reveals an increase in Lactobacillus in obese patients and
Methanogens in anorexic patients. PLoS One. 2009. doi:10.1371/journal.pone.0007125.
75. Saxelin M, Tynkkynen S, Mattila-Sandholm T, de Vos WM. Probiotic and other functional
microbes: from markets to mechanisms. Curr Opin Biotechnol. 2005;16:204–11.
76. Kankainen M, Paulin L, Tynkkynen S, et al. Comparative genomic analysis of lactobacillus
rhamnosus GG reveals pili containing a human- mucus binding protein. Proc Natl Acad Sci
U S A. 2009;106:17193–8.
77. Lahti L, Salonen A, Kekkonen RA, Salojärvi J, Jalanka-Tuovinen J, Palva A, Orešič M, de
Vos WM. Associations between the human microbiota, Lactobacillus rhamnosus GG and
serum lipids indicated by integrated analysis of high-throughput profiling data. Peer J. 2013.
doi:10.7717/peerj.32.
78. Rosberg-Cody E, Liavonchanka A, Gobel C, et al. Myosin-cross-reactive antigen (MCRA)
protein from bifidobacterium breve is a FAD-dependent fatty acid hydratase which has a
function in stress protection. BMC Biochem. 2011;12:9.
79. De Voos WM, Nieuwdorp M. A gut prediction. Nature. 2013;498:48–9.
80. Arumugam M, Raes J, Pelletier E, et al. Enterotypes of the human gut microbiome. Nature.
2011;473:174–80.
81. Kurokawa K, Itoh T, Kuwahara T, et al. Comparative metagenomics revealed commonly
enriched gene sets in human gut microbiomes. DNA Res. 2007;14:169–81.
82. Gill SR, Pop M, DeBoy RT, et al. Metagenomic analysis of the human distal gut microbiome.
Science. 2006;312:1355–9.
83. Martens EC, Koropatkin NM, Smith TJ, Gordon JI. Complex glycan catabolism by the human
gut microbiota: the bacteroidetes sus-like paradigm. J Biol Chem. 2009;284:24673–7.
84. Karlsson FH, Tremaroli V, Nookaew I, et al. Gut metagenome in European women with
normal, impaired and diabetic glucose control. Nature. 2013;498:99–103.
85. Marques-Vidal P, Schmid R, Bochud M, et al. Adipocytokines, hepatic and inflammatory
biomarkers and incidence of type 2 diabetes. The CoLaus study. PLoS One. 2012. doi:10.1371/
journal.pone.0051768.
86. Louis P, Young P, Holtrop G, Flint HJ. Diversity of human colonic butyrate-producing bacte-
ria revealed by analysis of the butyryl-CoA:acetate CoA-transferase gene. Environ Microbiol.
2010;12:304–14.
87. Vrieze A, Van Nood E, Holleman F, et al. Transfer of intestinal microbiota from lean donors
increases insulin sensitivity in individuals with metabolic syndrome. Gastroenterology.
2012;143:913–6.
88. Furet J, Kong L, Tap J, et al. Differential adaptation of human gut microbiota to bariatric
surgery–induced weight loss: links with metabolic and low-grade inflammation markers.
Diabetes. 2010;59:3049–57.
89. Qin J, Li Y, Cai Z, et al. A metagenome-wide association study of gut microbiota in type 2
diabetes. Nature. 2012;490:55–60.
90. Bäckhed F. Programming of host metabolism by the gut microbiota. Ann Nutr Metab.
2011;58 Suppl 2:44–52.
91. Fava F, Lovegrove JA, Gitau R, Jackson K, Tuohy K. The gut microbiota and lipid metabolism:
implications for human health and coronary heart disease. Curr Med Chem. 2006;13:3005–21.
92. Ridlon JM, Kang D, Hylemon PB. Bile salt biotransformations by human intestinal bacteria.
J Lipid Res. 2006;47:241–59.
93. Musso G, Gambino R, Cassader M. Interactions between gut microbiota and host metabolism
predisposing to obesity and diabetes. Annu Rev Med. 2011;62:361–80.
94. Louis P, Flint HJ. Diversity, metabolism and microbial ecology of butyrate-producing bacte-
ria from the human large intestine. FEMS Microbiol Lett. 2009;294:1–8.
95. Delzenne NM, Cani PD. Interaction between obesity and the gut microbiota: relevance in
nutrition. Annu Rev Nutr. 2011;31:15–31.
96. Al-Lahham SH, Peppelenbosch MP, Roelofsen H, Vonk RJ, Venema K. Biological effects of
propionic acid in humans; metabolism, potential applications and underlying mechanisms.
Biochim Biophys Acta. 2010;1801:1175–83.
97. Stoddart LA, Smith NJ, Jenkins L, Brown AJ, Milligan G. Conserved polar residues in transmem-
brane domains V, VI, and VII of free fatty acid receptor 2 and free fatty acid receptor 3 are required
for the binding and function of short chain fatty acids. J Biol Chem. 2008;283:32913–24.
98. Brown AJ, Goldsworthy SM, Barnes AA, et al. The orphan G protein-coupled receptors
GPR41 and GPR43 are activated by propionate and other short chain carboxylic acids. J Biol
Chem. 2003;278:11312–9.
98. Samuel BS, Shaito A, Motoike T, et al. Effects of the gut microbiota on host adiposity are
modulated by the short-chain fatty-acid binding G protein-coupled receptor, Gpr41. Proc
Natl Acad Sci U S A. 2008;105:16767–72.
100. Ge H, Li X, Weiszmann J, et al. Activation of G protein-coupled receptor 43 in adipocytes
leads to inhibition of lipolysis and suppression of plasma free fatty acids. Endocrinology.
2008;149:4519–26.
101. Bjursell M, Admyre T, Göransson M, et al. Improved glucose control and reduced body fat
mass in free fatty acid receptor 2-deficient mice fed a high-fat diet. Am J Physiol Endocrinol
Metab. 2011;300:E211–20.
102. Cani PD, Delzenne NM. Interplay between obesity and associated metabolic disorders: new
insights into the gut microbiota. Curr Opin Pharmacol. 2009;9:737–43.
103. Cani PD, Delzenne NM. The role of gut microbiota in energy metabolism and metabolic
disease. Curr Pharm Des. 2009;15:1546–58.
104. Dewulf EM, Cani PD, Neyrinck AM, et al. Inulin-type fructans with prebiotic properties
counteract GPR43 overexpression and PPARγ-related adipogenesis in the white adipose
tissue of high-fat diet-fed mice. J Nutr Biochem. 2011;22:712–22.
105. Roberfroid M, Gibson GR, Hoyles L, et al. Prebiotic effects: metabolic and health benefits.
Br J Nutr. 2010;104 Suppl 2:1–63.
106. Lefebvre P, Cariou B, Lien F, Kuipers F, Staels B. Role of bile acids and bile acid receptors
in metabolic regulation. Physiol Rev. 2009;89:147–91.
107. de Aguiar VT, Tarling E, Edwards P. Pleiotropic roles of bile acids in metabolism. Cell Metab.
2013;17:657–69.
108. Kootte RS, Vrieze A, Holleman F, et al. The therapeutic potential of manipulating gut micro-
biota in obesity and type 2 diabetes mellitus. Diabetes Obes Metab. 2012;14:112–20.
109. Russell DW. The enzymes, regulation, and genetics of bile acid synthesis. Annu Rev Biochem.
2003;72:137–74.
110. Swann JR, Want EJ, Geier FM, Spagou K, Wilson ID, Sidaway JE, Nicholson JK, Holmes
E. Systemic gut microbial modulation of bile acid metabolism in host tissue compartments.
Proc Natl Acad Sci U S A. 2011;108 Suppl 1:4523–30.
111. Russell DW. Fifty years of advances in bile acid synthesis and metabolism. J Lipid Res.
2009;50 Suppl 1:S120–5.
112. Chiang JYL. Bile acids: regulation of synthesis. J Lipid Res. 2009;50:1955–66.
113. Sinal CJ, Tohkin M, Miyata M, Ward JM, Lambert G, Gonzalez FJ. Targeted disruption of the
nuclear receptor FXR/BAR impairs bile acid and lipid homeostasis. Cell. 2000;102:731–44.
114. Wostmann BS. Intestinal bile acids and cholesterol absorption in the germfree rat. J Nutr.
1973;103:982–90.
115. Jones BV, Begley M, Hill C, Gahan CGM, Marchesi JR. Functional and comparative metage-
nomic analysis of bile salt hydrolase activity in the human gut microbiome. Proc Natl Acad
Sci U S A. 2008;105:13580–5.
116. Miyata M, Sakaida Y, Matsuzawa H, Yoshinari K, Yamazoe Y. Fibroblast growth factor 19
treatment ameliorates disruption of hepatic lipid metabolism in farnesoid X receptor (fxr)-
null mice. Biol Pharm Bull. 2011;34:1885–9.
117. Sayin S, Wahlström A, Felin J, et al. Gut microbiota regulates bile acid metabolism by reduc-
ing the levels of tauro-beta-muricholic acid, a naturally occurring FXR antagonist. Cell
Metab. 2013;17(2):225–35.
118. Kawamata Y, Fujii R, Hosoya M, et al. A G protein-coupled receptor responsive to bile acids.
J Biol Chem. 2003;278:9435–40.
119. Watanabe M, Houten SM, Mataki C, et al. Bile acids induce energy expenditure by promot-
ing intracellular thyroid hormone activation. Nature. 2006;439:484–9.
120. Studer E, Zhou X, Zhao R, et al. Conjugated bile acids activate the sphingosine-1-phosphate
receptor 2 in primary rodent hepatocytes. Hepatology. 2012;55:267–76.
121. Porez G, Prawitt J, Gross B, Staels B. Bile acid receptors as targets for the treatment of dys-
lipidemia and cardiovascular disease: thematic review series: new lipid and lipoprotein tar-
gets for the treatment of cardiometabolic diseases. J Lipid Res. 2012;53:1723–37.
122. Liu X, Invernizzi P, Lu Y, et al. Genome-wide meta-analyses identify three loci associated
with primary biliary cirrhosis. Nat Genet. 2010;42:658–60.
123. Mells GF, Floyd JAB, Morley KI, et al. Genome-wide association study identifies 12 new
susceptibility loci for primary biliary cirrhosis. Nat Genet. 2011;43:329–32.
124. Teslovich TM, Musunuru K, Smith AV, et al. Biological, clinical and population relevance of
95 loci for blood lipids. Nature. 2010;466:707–13.
125. Quehenberger O, Dennis EA. The human plasma lipidome. N Engl J Med. 2011;365:1812–23.
Chapter 13
Role of the Gut Microbiota in Maintaining
GI Health: Highlights on Inflammatory
Bowel Disease
Lisa Gruber and Dirk Haller
Abstract The human gut harbors trillions of microbes, with an area of at least
300 m2 intestinal tissue in constant contact with this microbial ecological sys-
tem (Artis, Nat Rev Immunol 8(6):411–420, 2008, Ley et al. Cell 124(4):837–
848, 2006). The intestinal epithelium with a number of different cell types
shapes the frontier between microbes and the host. Intestinal epithelial cells
(IECs) as well as cells of the immune system guard the local interface of
microbes and host and actively tolerate selected commensal microbiota while
mounting an adequate inflammatory response toward pathogens in the context
of infection or disease. Although there is no clear definition of a “normal” intes-
tinal microbiota as such, it is apparent that perturbations of a certain homeo-
static system may lead to a dysregulated interaction between microbes and the
host intestinal mucosal immune system, resulting in aberrant or disproportion-
ate inflammatory conditions.
The following chapter gives a general introduction to the role of the microbiota
in gut health, focuses on aberrations in microbe-host mutualism that are implicated
in the etiopathology of inflammatory bowel diseases, and then briefly addresses the
possibilities of dietary modulation of intestinal microbiota in the context of inflam-
matory bowel diseases.
Keywords Adaptive immunity • Antibiotics • Antigen-presenting cells • Crohn’s

disease • Barrier • Bile acids • Dendritic cells • Diet • Elemental diet • Enteral
nutrition • Fecal microbial transplantation • Fat • Genome-wide association
studies • Gnotobiology • IBD • Inflammatory bowel disease • Innate immunity
L. Gruber
Nutrition and Immunology, 85350 Freising, Germany
D. Haller (*)
Nutrition and Immunology, 85350 Freising, Germany
ZIEL – Research Center for Nutrition and Food Sciences, Biofunctionality Unit,
Technische Universität München, Gregor-Mendel Str. 2, 85350 Freising, Germany

DOI 10.1007/978-1-4471-6539-2_13
262 L. Gruber and D. Haller
• Intestinal health • Iron • Metabonomics • Metagenomics • Metaproteomics •

Metatranscriptomics • Microbiota • Mouse models • Nutrigenetics • Nutrition •
Pathobionts • Permeability • Phylogeny • Prebiotics • Probiotics • Short-chain fatty
acids • Ulcerative colitis
13.1 Host-Microbe Mutualism in Gut Health
Microbial organisms and the host mucosal immune system have established fine-
balanced interactions during coevolution. In the last decades, the use of gnotobiol-
ogy has brought forward the studies of host-microbe interactions to a substantial
extent. These experiments in rodents manipulate the microbiota by introducing
selected bacteria. Comparative studies allow the investigation of factors directing
the establishment and maintenance of bacterial communities in the intestine, as well
as investigating the impact of microbial factors on gastrointestinal functions. It is
now known that the presence and composition of microbiota affect gut morphology;
metabolic, absorptive, nerval, and endocrine functions; as well as mucosal and sys-
temic immune functions.
However, the specific microbial and host factors that regulate the aspects of
mutualism in a complex and changing environment are poorly understood. The
most important and best studied host-microbe metabolic interactions will be pre-
sented in the following chapters.
13.1.1 Colonization: A Matter of Interaction
Besides major time-dependent variation of the intestinal microbiome and its spatial
distribution along the intestinal tract, there are also major differences between
mucosa-associated resident bacteria and fecal bacteria [3], deriving from different
abilities of the microbiota to persist in and attach to a given host environment. While
there exist at least 55 divisions of bacteria, the human gut microbial community is
dominated by members of only four phyla, namely, Bacteroidetes, Firmicutes,
Actinobacteria, and Proteobacteria. It is suggested that a few early colonizers of the
human intestine have established an exclusive population and then diversified into a
high variety of species and strains [2, 4].
From the viewpoint of bacteria, colonization is driven mainly by the availabil-
ity of nutrients and the ability to attach to surfaces for building up residency.
Many species possess large numbers of genes encoding for carbohydrate metabo-
lism enzymes and can switch between different energy sources depending on their
availability, also including protein sources [5]. Some strains are capable of turn-
ing to host-derived mucus when dietary saccharides are rare, as can Akkermansia
muciniphila [6, 7], Bacteroides acidifaciens [7], or Bacteroides thetaiotaomicron
[8, 9]. Although microbial organisms benefit differentially from components of
13 Role of the Gut Microbiota in Maintaining GI Health… 263
host-secreted compounds, the activity of host foragers was shown to depend on

other community members [7]. The mucus layer also serves as a major source of
attachment sites [10–12]. Given the high variability among mucins and glycopro-
tein structures on the host cell surface and the spatial distribution of the different
forms, this suggests a mechanism of host-microbe interaction for the regulation of
microbial composition.
At a second stage, the ability of a bacterial strain to persist in the gut ecosystem
also depends on its power to compete with other strains. Microbes can compete
with bacteria or parasites by the competition for nutrients (“exploitation competi-
tion”); by direct inhibition, for example, via toxic substances (“interference com-
petition”); or by the induction of immune responses that target competitors
(“apparent competition”) [13]. Apparently, a driving force for successful coloniza-
tion is to keep the own vulnerability toward host-derived defense mechanisms low
(also see Sect. 13.1.3) while promoting immune responses toward competitors.
13.1.2 Immune Maturation Processes upon Microbial

Stimulation
Microbial organisms derive clear advantage from ensuring tolerance of the mucosal
immune system toward them and therefore have evolved to influence the host
immune system development profoundly. Studies in germfree animals show dis-
rupted organization of mesenteric lymph nodes and Peyer’s patches and reduced
numbers of dendritic cells (DCs), T cells, and B cells in the lamina propria [14–20].
These findings implicate that stimulation from microbes and/or microbial structures
or metabolites is needed for functional development of the immune system.
Antigen-presenting cells such as DCs constantly sample and process luminal
compounds and can be regarded as major directors of subsequent modulation of
immune response. Reduced numbers of intestinal DCs are observed in germfree
animals, whereas microbial stimulation by mono-colonization with Escherichia
coli was sufficient to recruit DCs to the intestine [20]. Different subsets of DCs are
thought to exert specialized functions during antigen sampling and information
dissemination, as visualized for the murine system in Fig. 13.1a. Murine intestinal
CX3CR1+ antigen-presenting cells extend protrusions into the lumen for uptake of
dietary and microbial antigens [21, 22]. Signals from the intestinal epithelium
shape the properties of DCs [23]. Conditioned antigen-presenting cells then pass
on information to T cells which, depending on co-stimulatory factors, differentiate
into one of the major phenotypes upon this stimulation: T helper cell type (Th) 1
and Th17, both implicated in inflammation; Th2, implicated in allergic reactions;
as well as Treg, regulating inflammatory responses. Other types of Th cells such as
Th9 or Th22 are less well described today. Specialized CD103+ DCs develop a
tolerogenic phenotype after antigen sampling and upon stimulation by IEC-derived
factors such as retinoic acid or transforming growth factor-β (TGF-β). In contrast
a
IEC
CX3CR1+ CD103+
DC DC
CD103+
DC
Th17 Treg
Th1?
MLN
b protection against intracellular

Tbet Th1 pathogens implicated in autoimmunity
PSA from
B. fragilis
(systemic)
protection against extracellular bacteria

ATP Th17
RORyt and fungi implicated in autoimmunity
SFB
antigen CD4+ T cell protection against parasites

GATA3 Th2 implicated in allergies
presenting cell
Clostridium
PSA from B. fragilis
(local)
Th3 tolerance
FoxP3 Treg regulation of immune response
Th9
Th22
ThFH
…
Fig. 13.1 (a) Antigen sampling and information dissemination by murine lamina propria antigen-
presenting cell subsets. CD103−CX3CR1+ cells sample bacteria and induce Th17 and probably also
Th1 cell differentiation, without migration to lymph nodes. In the presence of RA, TGF-β, and TSLP
released by IEC, CD103+ DCs develop a tolerogenic phenotype and induce Treg cell differentiation
after their migration to the mesenteric lymph nodes. (b) Microbes and microbial compounds can
induce T-cell differentiation into one of the major T helper cell types: Th1, Th17, Th2, and Treg.
Differentiation requires the induction of transcription factors – the most important are outlined in the
figure. The bacteria or bacteria-derived factors that have been shown to promote the induction of a
particular Th subtype are indicated in the figure. Each Th subset then has a specific function as indi-
cated. IEC intestinal epithelial cell, Th T helper cell, Treg regulatory T cell, MLN mesenteric lymph
node, ATP adenosine 5′-triphosphate, SFB segmented filamentous bacteria, PSA polysaccharide A
to CX3CR1+ cells, these CD103+ DCs then migrate to the mesenteric lymph nodes
and generate Tregs, inducing tolerance to constantly sampled, harmless antigens
[24, 25]. Far less is known about subsets of human specialized antigen-presenting
cells. While CD103+SIRPα+ DCs correspond to murine CD103+CD11b+ DCs
and account for the majority of small intestinal dendritic cells [26, 27], CD103-
SIRPα+ DCs, likely derived from blood monocytes, are more abundant in the large
intestine and accumulate during inflammation [27]. CD141+CD103+ SIRPα- with
a repertoire of expressed proteins involved in the induction of cytotoxic T-cell
response have been described in the human small intestine and likely correspond
to murine CD103+CD11b- DCs [27].
In general, germfree mice display reduced total numbers of T cells, bias toward
Th2 response [28], and especially low numbers of Th17 cells [29, 30]. Specific
microbes and microbial compounds have been shown to profoundly shape the bal-
ances of T-cell subsets, presumably via conditioning of both IECs and antigen-pre-
senting cells (Fig. 13.1b). Segmented filamentous bacteria (SFB) promote Th17
response [30, 31]. Colonization with SFB confers resistance to Citrobacter roden-
tium infection, indicating that Th17-cell induction by SFB is responsible for this
protective immune response. Bacteria-derived flagellin triggers the differentiation
of Th1 and Th17 cells after stimulation of DCs [32]. Also commensal bacteria-
derived adenosine 5′-triphosphate has been found to activate a unique subset of
lamina propria DCs, resulting in differentiation of Th17 cells [29]. Clostridia, par-
ticularly of clusters IV and XIVa, are capable of inducing Tregs [33]. Polysaccharide
A, derived from the commensal Bacteroides fragilis, also mediates Treg cell
responses and suppresses Th17, resulting in protection from inflammation [28, 34].
However, a Th1-/Th17-driven pro-inflammatory response is induced upon the sys-
temic presence of Bacteroides fragilis [35, 36]. This example highlights the plastic-
ity of immune responses toward microbial antigens and the fine regulation of
tolerogenic and inflammatory mechanisms.
The induction of B-cell maturation and recruitment is also mediated by the micro-
biota. The organized structures in lymphoid organs where differentiation and matura-
tion of B cells occur are disrupted in germfree animals [17, 37], and they exhibit
reduced plasma cell numbers and decreased levels of secretory immunoglobulin A
(sIgA) [38]. In contrast, allergy-associated IgE is increased in germfree rats [39].
As another aspect of microbial influence on the host immune system, microbial
metabolites, especially short-chain fatty acids (SCFA), exert anti-inflammatory
actions on the intestinal immune system. SCFA are produced by microbes upon
fermentation of polysaccharides. Butyrate, propionate, and acetate are the most
abundant and described SCFA, with almost all the produced amount absorbed in the
colon. Butyrate is regarded as the primary energy source for colonic IEC but also
displays anti-inflammatory effects such as decreasing cytokine production [40, 41].
Indeed, treatment with acetate or butyrate resulted in the amelioration of inflamma-
tion in rodents [40, 42] and humans [43].
Interestingly, the immune maturation processes depend on the presence of
coevolved host-specific microbiota, as cross-species colonization cannot induce full
expansion of lymphocytes in mice and also fails to protect against Salmonella infec-
tion compared to self-species colonization [44].
13.1.3 Shaping and Confining the Gut Microbiota
Despite the beneficial mutual effects, the host has to keep the microbiota at bay and
therefore has developed a broad range of innate protection mechanisms. Besides the
selection of colonizing strains, the host also actively shapes the resident community
by strategies related to innate immune response – probably in part as a result of
bacteria modulating the host.
Antimicrobial peptides (AMPs) form a barrier with maximal activity in the intes-
tinal crypts as well as in the mucus layer, preventing contact or penetration of the
host epithelium [45, 46]. AMPs are a diverse group of compounds, including defen-
sins, cathelicidins, lectins, and angiogenins, exerting their effect on bacteria by
destructing bacterial cell wall integrity. AMPs are secreted by Paneth cells upon
stimulation by live bacteria, bacterial components, or metabolites [46–50].
Especially Ang4 and REGIIIγ expression is induced by mono-colonization of
germfree mice [48] or during weaning [51, 52]. Both of them exclusively target
Gram-positive bacteria while sparing Gram negatives, thus influencing microbial
composition and presumably contributing to the shift observed during weaning.
Mice expressing a human α-defensin gene show a loss of SFB and fewer IL17-
producing T cells [53], providing a clear example for the bidirectional effects of
host immune system and microbial community.
Bacteria also induce the production and secretion of sIgA by B cells [17, 38].
SIgA translocates into the lumen, binding bacteria or antigens of other origin, and
therefore prevents attachment on IEC and regulates colonization [54]. IgA can also
bind to bacteria that have gained access to the serosal side and translocate them back
to the lumen. Bound to bacteria, sIgA induces the clearance of the bacteria by DCs,
phagocytes, or neutrophils [55]. DCs then can retain bacteria and selectively induce
the secretion of IgA. These activated DCs are restricted to the mucosal immune
compartment by the mesenteric lymph nodes, ensuring a local immune response
[56]. Taking into consideration that IgA deficiency results in intestinal dysbiosis
[57] and that bacterial species are differentially capable of inducing IgA [58], IgA
secretion is a means of both confining bacteria and shaping the intestinal ecology.
Cells of the innate immune system constantly recognize conserved microbe-
associated molecular patterns, mainly via Toll-like receptors (TLRs) or nucleotide-
binding oligomerization domain (NOD)-like receptors. In a healthy host, commensal
bacteria thereby direct a tolerogenic immune response. On the other hand, the bind-
ing of bacteria-derived flagellin or lipopolysaccharides induces TRL5 and TLR4
signaling in DCs, respectively, alerting the immune system [59]. NODs initiate
innate immune responses upon intracellular sensing of bacterial cell wall compo-
nents, which results in elimination of intruded bacteria. Although the targeted struc-
ture muramyl dipeptide is highly conserved throughout bacterial phyla, NOD2 has
been found to shape the intestinal microbiota in mice [60–62] and IBD patients
(also see Fig. 13.7) [63, 64]. In this context, it has been reported that a colitogenic
microbial ecology shaped by a certain host physiology is even transferring disease
from this host into another: the dysbiosis reported for NOD2−/− mice results in
enhanced disease risk in wild-type mice [65]. In other examples for host genotype
shaping microbial composition, multidrug resistance gene-deficient mice exhibit
altered microbiota independent of their intestinal inflammation (also see Fig. 13.7)
[66]. Furthermore, a deficiency of T-bet, implicated in directing host innate immu-
nity, resulted in colitis that was antibiotic sensitive and communicable to wild-type
mice [67].
It thus becomes evident that the microbiota plays a critical role in inducing host
mechanisms for the defense against pathogens and that at the same time host
responses toward microbiota shape the intestinal ecology. The host genetic back-
ground associated with the function of Paneth cells which secrete antimicrobial
compounds into the lumen is considered especially critical in influencing the intes-
tinal microbial composition [68]. In conclusion, the physiology and genetic back-
ground of the host determine the ecology of the intestinal tract (though with scope
for modulation by environmental factors), and the microbiota is under control in a
healthy individual [69].
13.1.4 Inflammatory Bowel Disease: Loss of Homeostasis
A loss of homeostasis concerning microbiota and host mucosal immune system is

the basis for the pathogenesis of inflammatory conditions of the gastrointestinal
tract, including the complex etiology of inflammatory bowel disease (IBD). IBD,
with its two main idiopathic pathologies, ulcerative colitis (UC) and Crohn’s dis-
ease (CD), is regarded as a multifactorial disease in which a certain host susceptibil-
ity regarding perturbations of barrier or microbe-host interactions combined with
microbial aggressiveness leads to inappropriate host immune response toward the
microbiota. The microbiota is affected in composition and functionality by both
environmental triggers and host genetic background at the same time. Environmental
triggers such as diet, drugs, or infectious agents can also affect the host condition
and thereby impact on disease development. Figure 13.2 gives an overview of all
these aspects in the current paradigm of IBD development. Alternating phases of
active disease (relapse) and freeness of symptoms (remission) are characteristics
of IBD.
Although described by the collective term IBD, UC and CD are two diseases
with different symptoms and distinct pathogenesis. CD potentially extends to the
submucosa and may occur anywhere along the gastrointestinal tract – with a pre-
dominantly ileal phenotype referred to as ICD and a colonic phenotype as CCD. In
UC, inflammation involves mucosa only and is restricted to the large intestine.
Interestingly, the inflammatory profiles of UC and CD are very distinct. UC displays
predominantly Th2 phenotype, associated with IL5 and TGF-β as predominant
cytokines, whereas CD is associated with type Th1 and Th17 immune responses
dominated by IL12, IL23, IFNγ, and TNF.
The prevalence of CD and UC has risen considerably over the last decades. In the
1990s and 2000s, reported incidence rates for CD in Western countries were 4 [70]
intestinal microbiota
host genetic background

CD
healthy IBD
environment
healthy IBD
UC
host barrier
and immune system
Fig. 13.2 IBD is a multifactorial disease, resulting from aberrant host responses to intestinal
microbiota. The host genetic background establishes susceptibility in regard to barrier function and
immune response. This can result in an aberrant response toward intestinal microbiota. At the same
time, the host condition and genetic background shape the intestinal microbiota. Environmental
triggers can affect both barrier and immune functions of the host and composition as well as func-
tionality of intestinal microbiota. IBD, inflammatory bowel disease, CD Crohn’s disease, UC
ulcerative colitis. The histological pictures show distal ileum tissues of TNFΔARE/WT mice dis-
playing CD-like ileitis and the corresponding wild-type mouse, as well as proximal colon tissues
of IL10−/− mice developing UC-like colitis and the corresponding wild-type mouse
to 8 cases [71] per 100,000 person-years with a tendency of higher rates in women
and an incidence peak among 15- to 30-year-olds [70, 72–75]. UC is more prevalent
than CD, with North America and northern Europe showing the highest incidence
rates varying from 9 [71] to 20 cases [76] per 100,000 person-years, showing a simi-
lar age distribution to CD without sex differences.
Despite the lack of data for many areas (especially Africa), there seems to be a
north-south axis of IBD prevalence within continents or even countries [77–79]. In
Asia, incidence rates of IBD have been low in the past but are now rising as well
[80, 81]. The fact that genetic factors are critical in IBD pathogenesis is stressed by
the discrepancies in data sets of Caucasian versus Asian populations (see Sect. 13.2)
[82, 83]. Most importantly, concordance rates for CD in monozygotic twins are
20–50 % in northern Europe, meaning that the relative risk is at least 40-fold higher
for individuals with an affected twin compared to the general population [84–88].
According to these studies, concordance rates are lower for UC. Indeed, the disease
variance that can be explained by genetic variations is higher in CD than UC [89],
implicating that genetic factors might contribute less significantly than in CD or that
environmental and lifestyle triggers overwrite genetics.
Irrespective of the differences in pathophysiology and genetic contribution, there
is remarkable evidence for a role of the intestinal microbiota and loss of host toler-
ance toward it during the pathogenesis of both IBD etiologies [90, 91]. Surgical
bypass of the ileum prevents inflammation in patients with active CD, and inflam-
mation reoccurs rapidly after the reinfusion of the bypassed segments [92]. IBD
patients develop serological responses toward their microbiota [93, 94], and much
higher rates of colonic bacteria are coated with immunoglobulin in patients with
active CD compared to controls or patients in remission [95]. In addition, antibiotic
treatment may result in the amelioration or even abrogation of inflammation in CD
patients [96–98]. Finally, as listed in Table 13.1, treatment with broad-spectrum
antibiotics as well as housing under germfree conditions ameliorates or abrogates
inflammation in genetically engineered rodent models of intestinal inflammation.
These studies provide substantial evidence for the role of microbiota and the devel-
opment of chronic intestinal inflammatory diseases and in addition suggest a crucial
role for the host genetic background in modeling susceptibility.
13.2 IBD Susceptibility Genes: Mucosal Immunology

and Microbial Defense
According to the current paradigm, IBD results from a disturbed host-microbe

interaction and loss of tolerance to nonpathogenic microbiota, leading to a chronic
inflammatory response. The concept of homeostasis in this context not necessarily
means the absence of inflammatory responses, but the ability of the host to mount
an appropriate response toward any changes of the microbial community. The nec-
essary fine-balanced interaction of microbe and host physiology can be perturbed
by primary alterations on both sides. In the following chapter, possible disturbances
of this homeostasis on host side will be highlighted, with a focus on genes that have
been assigned as susceptibility factors for the development of IBD.
More than 160 genetic loci have been revealed implicated in IBD by genome-
wide association studies (GWAS) up to date [89]. In CD 13.6 % of the disease vari-
ance can be explained by the means of variations in these genes and 7.5 % in UC
[89]. It is noteworthy that the ethnicity of the study population has to be kept in
mind when looking at the data sets: along with differences in IBD prevalence, the
identification of relevant susceptibility genes varies in Asians versus Caucasians
[82, 83, 99] and in Jewish versus non-Jewish populations [100–102].
GWAS highlight the impact of microbiota and microbial components in IBD
pathogenesis. A substantial number of genes are implicated in both CD and UC,
many of which involved in barrier function, primary defense mechanisms such as
Table 13.1 Rodent models of inflammatory intestinal diseases and the effect of manipulation of
their microbiota
Manipulation
Animal model Phenotype of microbiota Effect
IL10−/− Colitis GF No disease [132]
GF + E. faecalis Recurrence of
disease [230, 313]
AB† (before onset No disease
of disease) [314–316]
IL2−/− Colitis GF Decreased
severity [317]
HLA-B27 (rat) Colitis GF No disease [177,
318]
Tcrα−/− Colitis GF No disease [319]
(Th2 phenotype)
Colitis GF + L. plantarum, No disease [319]
(Th2 phenotype) S. faecalis, S.
faecium, E. coli
SAMP1/Yit CD-like ileitis GF Decreased
severity [320]
AB* Decreased
severity [321]
TRUC Colitis AB‡ No disease [67]
K8–/– Colitis AB‖ No disease [322]
Mdr1a−/− Colitis AB# No disease [323]
dnTGFβRII x IL10R2−/− Colitis AB* No disease [324]
AB* + B. Recovery of
thetaiotaomicron disease [168]
AB* + E. coli No disease [168]
NOD2−/− + AOM Colitis and AB○ Reduced
inflammation- severity [65]
associated
colorectal cancer
STAT3-IKO Colitis and AB‡ Decreased
inflammation- severity, no tumor
associated formation [325]
colorectal cancer
SCID + CD44RBhigh Lymphoid cell AB Decreased
accumulation severity [326]
and hyperplasia
GF germfree conditions, AB antibiotic treatment; antibiotics used are indicated with the following
symbols, if described
*
Ciprofloxacin and metronidazole
○
Streptomycin, gentamicin, ciprofloxacin, and bacitracin
‖Vancomycin and imipenem
#
Streptomycin, neomycin, bacitracin, and amphotericin
‡
Vancomycin, neomycin, metronidazole, and ampicillin
†
The following combinations: ciprofloxacin, neomycin and metronidazole, metronidazole alone,
vancomycin and imipenem, and roxithromycin
Note that this list may not be complete and not every reference applicable for a certain example
may be indicated
intestinal microbiota
barrier
AMPs
mucus layer
IgA
cell junction proteins
basal membrane
IEC
innate mucosal defense

autophagy
plasma cell DC pathogen sensing
immune cell differentiation

cytokine production/signaling
recruitment chemokine production/signaling
B/T cell regulation

B cell T cell and functionality
see Figure 4
Fig. 13.3 GWAS reveal a critical role of microbe-host interactions on several levels. A compro-
mised barrier function of the host, whether resulting from impaired AMP production (e.g., due to
Paneth cell dysfunction), mucus layer aberrations, inadequate IgA production, loss of cell junction
proteins, or disorganized basal membrane structure, may lead to intestinal inflammation as
observed in IBD. Inadequate innate mucosal defense mechanisms involving autophagy and patho-
gen sensing can lead to inappropriate responses of the host immune system. Immune cell recruit-
ment and lymphocyte regulation and functionality are regulated by a complex network of cytokines,
chemokines, and differentiation processes, and a perturbation of these interactions can result in
IBD. AMPs antimicrobial peptides, IgA immunoglobulin A, DC dendritic cell, IEC intestinal epi-
thelial cell, P Paneth cell
sensing and processing of bacteria, cytokine signaling, and adaptive immunity [90,
103]. The stratification of these processes and tasks in the intestinal cell network is
depicted in Fig. 13.3. A selection of susceptibility genes is introduced in the follow-
ing sections, focusing on the just mentioned functions.
13.2.1 Decreased Barrier Function
The mucosal barrier is built up by different components: commensal microbiota

blocking niches for pathogens, IgA and AMPs secreted to the lumen, mucins above
the epithelial layer, and barrier proteins sealing the IEC layer. Several IBD-
associated loci suggest a critical role for barrier integrity in IBD susceptibility.
GWAS report significant association of mutation of CDH1, resulting in truncated
forms of E-cadherin, with CD [104]. Intestinal biopsies of patients carrying mutated
alleles show defective E-cadherin localization [104]. MUC19, encoding for a
secreted protein forming a chemical barrier together with other mucins and
embedded AMPs, is also implicated in IBD [105–107]. Further, Paneth cell function
is critical in maintaining barrier integrity, as they are the producers of AMPs. Paneth
cell dysfunction is a common consequence of mutations of IBD susceptibility
genes, such as NOD2, ATG16L1, or XBP1 [108] (also see Sect. 13.2.2). Extracellular
matrix proteins ECM1 and LAMB1 are also implicated in IBD, predominantly in
UC [89, 103]. As laminins are the major non-collagenous constituent of basement
membranes, defective variants could allow the penetration of microbes [109]. UC
patients exhibited abrogation of laminin in the epithelial basement membranes
surrounding the crypts in affected tissues [110].
13.2.2 Impaired Pathogen Sensing and Processing/Cellular

Innate Immunity
Bacteria that have penetrated the epithelial barrier have to be sensed and pro-
cessed for the induction of defense mechanisms. Upon intruding into the cell,
bacteria undergo lysis in autophagolysosomes, and their structures are sensed by
NOD-like receptors. NODs contain caspase recruitment domains (CARD) which
mediate downstream signaling pathways such as nuclear factor κB (NFκB) and by
this activate an appropriate immune response, resulting in the clearance of the
infection [111].
GWAS identified various polymorphisms of NOD2/CARD15 as more prevalent
in Caucasian CD patients compared to healthy controls [112–114]. NOD2 malfunc-
tioning leads to inefficient clearance of intruded bacteria, likely in part mediated by
reduced expression of Paneth cell-derived defensins. Paneth cell dysfunction and
reduced AMP production have also been reported for CD patients with NOD2
mutations [115, 116]. Although NOD2-deficient mice do not spontaneously develop
IBD, studies reveal increased susceptibility toward Listeria monocytogenes and
Helicobacter hepaticus infection [62, 117].
Other CARD structures are implicated as susceptibility genes: variants of
CARD11 are significantly associated with UC and variants of CARD9 with both
IBD etiologies [89, 106]. Besides NOD2, ATG16L is regarded as one of the stron-
gest genetic contributors to CD [107]. ATG16L1 is critical for the formation of
autophagosomes and thus for the degradation and processing of microbial proteins
[118]. Patients with either NOD2 or ATG16L1 mutations show ineffective induc-
tion of autophagy and bacterial processing [119, 120]. Indeed, bacterial killing is
abrogated in IEC with disease-associated ATG16L mutation after stimulation with
muramyl dipeptide [121], affirming a functional link of NOD2 and ATG16L1. In
addition, other factors mediating autophagy, such as T-cell protein tyrosine phos-
phatase (PTPN2) [122–124] and interferon-inducible protein 1 (IRGM) [89, 114,
124, 125], are implicated in IBD. IRGM-deficient mice exhibit decreased bacterial
killing in macrophages and are more susceptible to infections [126], whereas
PTPN2−/− mice exhibit compromised T-cell functions [127, 128], systemic inflam-
mation [127], and increased susceptibility to chemically induced colitis [129].
This convergence of several strong genetic risk factors highlights the importance
of pattern recognition and autophagic processes in the clearance of bacteria and thus
maintenance of intestinal health.
13.2.3 Aberrant T Helper Cell Immune Response
Upon sensing and processing of bacteria by innate mechanisms, a complex network

of signals leads to the induction of tolerance or immune response toward the respec-
tive organisms. Of note, it was found that human IECs promote the differentiation
of tolerogenic DCs driving the development of adaptive Foxp3+ Treg cells, as men-
tioned above, and that this mechanism is lost in patients with CD, with concomi-
tantly reduced expression of tolerogenic factors by IECs [23].
Chemokines recruit immune cells and cytokines regulate inflammatory activity
of adaptive immune cells. In general, the regulation of cytokine production seems to
be the most overrepresented functionality in IBD GWAS (Fig. 13.4), especially for
IL12R*
Tyk2*
Tbet
STAT1*
Th1 IFNy*, IL2*, IL12*, TNF
STAT4*
IFNy* IFNyR2
IL12*#
CD phenotpye
IL23R*#
Tyk2*
IL6 IL21*, IL22, IL17A, IL17F
CXCR5* RORyt* Th17
IL23 CCR6*
STAT3*
IL1b
TGFb
IL6ST
IL6R
IL4
antigen CD4+ T cell IL10
presenting cell GATA3 Th2 IL4, IL5, IL10, IL13, IL15
STAT6 UC phenotpye
TGFb
FoxP3
Th3 STAT5 Treg IL10*
Th9
Th22
ThFH
…
Fig. 13.4 GWAS reveal a critical role of the induction of T-cell differentiation and responses in
IBD development. Antigen-presenting cells can induce T-cell differentiation into one of the major
T helper cell types: Th1, Th17, Th2, and Treg. Differentiation depends on co-stimulatory factors of
the antigen-presenting cell surface or by the network of cytokines present that are sensed by recep-
tors on the T-cell surface. Lineage commitment then requires the induction of transcription fac-
tors – the most important of them are outlined in bold in the figure. The different T-cell subsets are
characterized by cytokine profiles. Reported susceptibility genes for the development of CD or UC
are printed in red. Association of a gene with both etiologies is indicated by *. If there are func-
tional studies available for a gene, this is indicated by #. IEC intestinal epithelial cell, Th T helper
cell, Treg regulatory T cell, UC ulcerative colitis, CD Crohn’s disease
IFNγ, IL12, TNF, and IL10 signaling [89]. IFNγ, IL12, and TNF are the predominant
cytokines in Th1 immune response, whereas IL10 is a major anti-inflammatory
cytokine. IL10 signaling is required for the generation of regulatory T-cell responses
via FoxP3 [130]. The impact of IL10 functionality on intestinal health is further
emphasized by the fact that IL10-deficient mice develop colitis when colonized and
have become a widely used model of IBD [131, 132].
Genes annotated to the ontology term “regulation of interleukin-17 produc-
tion” are strongly enriched in GWAS data sets (Fig. 13.4). Among these are
genes driving lineage commitment toward Th17 (RORγt, STAT3, IL23R, IL12B,
IL6ST) and genes related to Th17 signaling (IL21) and chemotaxis (CCR6) [89,
103, 105, 114, 133]. A subunit of IL12, a cytokine driving Th1 differentiation,
also serves as a subunit of IL23 and therefore can be regarded as interface
between Th1- and Th17-directed cell commitment [134], as can TYK2, process-
ing both signaling from IFNγ/IL12 and IL23. Further, Th1-associated genes
identified as susceptibility loci include genes encoding for factors driving Th1
differentiation (STAT1) and mediating Th1 signaling such as IL2, IFNG with
respective receptor forms, and a broad range of receptor and signaling proteins
associated with TNF [89, 103].
Most of these genes are implicated in both etiologies of IBD, although only CD
displays Th17 and Th1 immune response, whereas UC is Th2 dominated. These
genetic observations support the view that the chronic inflammation in IBD is a
consequence of innate immune dysfunction toward microbial stimuli, with environ-
mental risk factors shaping the relationship between microbiota and the immune
system. A substantial number of the susceptibility genes for IBD mentioned here
overlap with other diseases, such as Coeliac disease, atopic dermatitis and psoriasis
[89, 103], all of which result from faulty response of the immune system toward
nonpathogenic microbiota or antigens.
Together with subsequent functional analyses, the results of GWAS have contrib-
uted to a great extent to the understanding of IBD pathogenesis, although they have
not revealed potential for diagnostics as the predictive power is low, even at combi-
nation of all genetic susceptibility loci known today.
13.3 Microbial Diversity, Composition, and Function

in Health and IBD
The comparison between health and diseased state helps to better understand the
aspects of the host-microbe mutualism that sustain health, and the studies on micro-
bial ecology performed in IBD patients offer a broad range of indications for mech-
anisms involved in health maintenance versus pathogenesis.
Despite the obvious contribution of microbiota and microbe-host interactions
to the development of IBD, there is no specific microorganism described which
can be consistently isolated in each IBD case and is absent in healthy hosts.
Hypotheses that a particular bacterial agent causes IBD can therefore not be
supported. Several bacterial species are reported to correlate with the disease,
although no evidence exists for any of them to be the causative factor of human
IBD. Adherent-invasive E. coli (AIEC), for example, are found in abnormal
numbers in ileal mucosa of CD patients, presumably due to enhanced expression
of adherence factors [135, 136]. Besides this example of one defined species,
IBD may in general be caused by overall changes in the composition and func-
tionality of the intestinal microbiota, termed dysbiosis. A change of phenotypic
features of the microbiota has been observed in IBD patients. Also the functional
activity of these organisms has major impact on host-microbe signaling and is
studied using global approaches such as metagenomics, metatranscriptomics,
metaproteomics, and metabonomics. The phenomena observed so far on differ-
ent levels are summarized in the following sections.
13.3.1 Gut Microbial Phenotyping: Alterations in Microbial

Composition in IBD
There are many studies addressing the mucosa-associated microbial diversity in

IBD patients but with major discrepancies concerning the study specimen and the
way of data analysis. First, while some studies do not see differences between CD
and UC samples and therefore regard them as one sample set, others demonstrate
shifts in certain phyla specific for one of the etiologies. Second, the localization of
the biopsy sample taken is critical and differs depending on the etiology studied.
The definition of the control group is critical and may vary from unaffected people
to unaffected relatives to samples from an unaffected site of an IBD patient. In con-
clusion, only the findings that have so far been reproduced in different experiments
can be addressed here.
As a high-throughput method, 16SrDNA-based sequencing allows the rapid and
parallel processing of many samples under the same conditions and is therefore
commonly used in studies focusing on the microbial environment associated with
IBD. Regarding the total numbers of bacterial population, no differences have been
observed for fecal samples. Recently, studies have focused on mucosa-associated
microbiota as the population with proximal contact to the host. This is important as
luminal and mucosal communities have been shown to be distinct [3]. There are
discrepancies in the data for the total numbers of mucosa-associated bacteria [137,
138], but dysbiosis and reduced diversity are a common feature of this compartment
in the context of IBD.
Several studies have demonstrated reduced diversity of the intestinal ecology in
CD [138–141] and UC [140] compared to unaffected control patients. This reduc-
tion in diversity is connected to reduced abundance of the dominant members of the
human gut microbiota in IBD patients. In general, a decrease of Firmicutes [138,
140, 142, 143], and often Bacteroidetes [140, 142, 144], with concomitant increase
of Proteobacteria [141–143, 145], and Actinobacteria [141, 142], was reported for
mucosal biopsy samples or feces of IBD patients.
13.3.1.1 Firmicutes and Bacteroidetes
For the Firmicutes phylum, there is in particular substantial evidence for decreased
abundance of populations belonging to Clostridia IXa and IV [139, 142, 146–148].
The depletion of Faecalibacteria, especially F. prausnitzii, seems to be a common
feature in IBD patients, especially in CD [142, 145, 149–152]. With F. prausnitzii
being a member of the Firmicutes<Clostridia, this could therefore contribute to the
relative loss of this class observed in IBD. Supplementation of live F. prausnitzii or
its supernatant ameliorated chemically induced colitis in animal studies and tended
to correct the connected dysbiosis [151].
Regarding Bacteroidetes abundance, findings are not as consistent as for
Firmicutes. While some comparisons find Bacteroidetes depleted in IBD [140, 142,
144], others do not observe significant alterations [64] or even report increased
abundance [138].
13.3.1.2 Actinobacteria and Proteobacteria
Despite an overall tendency to relative increase in members of the Actinobacteria

phylum, a reduction in populations of Bifidobacteria has been reported for UC
patients [153]. It has been observed that the abundance of Bifidobacteria inversely
correlates with abdominal pain in healthy subjects [154]. Probiotic approaches
therefore consider Bifidobacteria as potential candidate for therapeutic use in IBD
(see Sect. 13.13.2).
The growth of sulfate-reducing bacteria (SRB) seems increased in IBD, espe-
cially UC, along with the rate of sulfidogenesis [155–157]. With about 23 genera, the
largest group of SRB is found among the Deltaproteobacteria, for example, within
the orders Desulfovibrionales and Desulfobacterales [158]. An increase in relative
amounts of SRB could thus contribute to the often observed increases in Proteobacteria
abundance. SRB can metabolize sulfate via dissimilatory reduction and use it as the
terminal electron acceptor in the electron transport chain. Sulfate or elemental sulfur
is also converted into hydrogen sulfide, which can act on IECs by inhibiting butyrate
utilization [159], as well as proliferation [160]. Desulfovibrionales member Bilophila
wadsworthia utilizes taurine for energy generation [161] and is involved in high-fat
diet-associated colitis in IL10−/− mice [162]. There is another important aspect of the
possible involvement of SRB in colitis: dextran sodium sulfate (DSS) is used to
chemically induce colitis in rodent models. Indeed, colonotoxic effects of sulfur
compounds were initially reported with the observation that DSS treatment induces
colitis and colorectal tumors [163, 164]. The microbial reduction of sulfate in the
DSS molecule to the inflammatory, barrier-breaking hydrogen sulfide may be an
initial trigger for the development of colitis [165]. DSS alters the microbial commu-
nity already before the onset of intestinal pathology [166], possibly by the selection
of bacteria that can degrade DSS or metabolize sulfur, such as Proteus mirabilis
[167] or Akkermansia muciniphila (Verrucomicrobiae) – the latter is only detectable
in mice treated with DSS [166]. These findings emphasize the role of SRB in the
induction of inflammation.
Concerning the increase in relative abundance of Proteobacteria in IBD, there is
also particular evidence for the overgrowth of facultative anaerobe Enterobacteriaceae
[138, 141, 143, 144, 147]. However, this family does not seem to be associated with
disease in a causative way, as Enterobacteriaceae (E. coli) isolates did not induce
colitis in antibiotic-pretreated susceptible mice despite robust colonization [168].
This example emphasizes that IBD-associated alterations in microbial composition
do not necessarily reflect a causative relation.
13.3.1.3 Conclusions on Phylogenic Approaches
Many of the studies investigating dysbiosis in IBD revealed ICD as very distinct
from healthy and UC state, whereas CCD and UC were often not distinguishable
from each other or from healthy state. It is noteworthy that the donor patients of
specimen used to study the microbiota often have different backgrounds, as ICD is
more likely to be treated with immunosuppressant therapy and less likely to be
treated with mesalamine or antibiotics than colonic phenotypes [169, 170]. These
confounders might strongly affect the correlations seen between microbial compo-
sition and disease state. Basically, only findings for increased abundance in
Enterobacteriaceae for CD, Faecalibacterium for ICD specifically, and Clostridia in
both etiologies could be upheld after correction for all available covariates [169].
So far, all reported findings are correlations only and it therefore cannot be ruled
out that dysbiosis may simply be a consequence of inflammation, with microbiota
adapting to a changing environment. Related to the abovementioned example, the
growth of Enterobacteriaceae can be enhanced by an inflammatory state per se or by
genetic predisposition of the host in animal models [66, 171]. The overgrowth of an
introduced species of Salmonella enterica (S. typhimurium) was also observed in an
adoptive transfer model of colitis, in which cytotoxic T cells destroy the epithelium
[172]. This highlights the fact that there is a clear impact of host genotypes and
phenotypic conditions on microbial composition. Strikingly, genotypic effects on
bacterial ecology have also been reported for humans since unaffected twins from
UC patients also showed lower bacterial diversity than healthy, unrelated individu-
als and the same tendencies concerning relative increase of Actinobacteria [144].
This again implicates that a certain host genetic background may select pathogenic
microbiota, comparable to the observations associated with NOD2 polymorphisms
mentioned above.
Another drawback of using phylogenic characterization of the microbiota as a
tool to study host-microbe interactions in health versus disease are the substantial
intra- and interindividual variances in microbial composition, often outranging
the differences between the study groups [138]. Defining alterations in relative
abundance of certain bacteria and their possible relation to disease is therefore
difficult. Still, the characterization of the intestinal microbial composition may be
a promising tool to aid in diagnostics of IBD. A noninvasive method applying
16S-based sequencing and mapping to a database of healthy and IBD samples was
successful in identifying pediatric patients with IBD (in contrast to patients with
other gastrointestinal symptoms) and could even distinguish between ICD and UC
both with reasonable sensitivity and specificity [173].
13.3.2 Metagenomics, Transcriptomics, and Proteomics:

Microbial Gene Repertoire and Activity in IBD
Compositional analyses reflect the abundance of different microbial communities,

but cannot give information on their activity or functionality. When aiming to eluci-
date the possible impact of the microbiota on gut health, the logical consequence is
to study the functionality of the present populations. A higher transcriptional activ-
ity suggests higher cell replication rate or increased protein expression. Using this
approach, it was observed that the prevalence of Enterobacteriaceae and in particu-
lar E. coli also translated to a high transcriptional activity [143, 174].
Metagenomic approaches study the whole repertoire of genes present in the
microbiome. In the next steps, metatranscriptomics can give a picture of what genes
are expressed, and finally, metaboproteomics reveal the resultant protein expres-
sion. However, in comparison to compositional analysis, metagenomics as well as
metatranscriptomics and metaproteomics have rarely been applied in the study of
microbial contribution to IBD until now.
The human intestinal microbiome contains an estimated number of 3.3 million
genes, referred to as the metagenome [175]. Considering that there are about 150-
fold more microbial genes present than human genes, it seems plausible to study
the prospective functional activity assigned to the intestinal microbiota when aim-
ing to elucidate host-microbe interactions. Indeed, microbial function might be
more consistently perturbed than composition in IBD [169]. Generally speaking,
metagenomic approaches have shown topological shifts of the microbiome associ-
ated with IBD [176]. The enzymes correlating with the host state were related to
metabolic processes likely to use or produce metabolites that build an interface
between bacterial and host metabolism, therefore probably affecting microbe-host
interaction in the gut environment. Interestingly, similar changes were observed
for IBD and obesity.
Chip-based metatranscriptomics have been applied in studying the transcrip-
tional profiles of bacteria when present in different hosts, thereby substantiating the
concept that microbes are affected by host genotype and the created intestinal envi-
ronment. For example, the gene expression profile of Bacteroides thetaiotaomicron
varied depending on whether it was mono-associated into transgenic, inflamed
HLA-B27, or non-inflamed wild-type rats, with a downregulation of pathways
involved in bacterial growth and metabolism in transgenic rats with colitis [177]. In
contrast, bacterial genes in the ontology of the molecular function “receptor activity,”
mostly encoding nutrient binding proteins, were significantly upregulated in this
colitis model. In another approach, nonpathogenic E. coli exhibited upregulated
expression of stress-response-associated genes upon mono-association into inflamed

IL10−/− compared to healthy wild-type mice [178].
First attempts in applying untargeted shotgun proteomics to the microbiome
could show a common core metaproteome, enriched in proteins related to transla-
tion, energy production, and carbohydrate metabolism in comparison to what could
be predicted from metagenomics [179, 180]. Furthermore, the metaproteome of an
individual proved to be reasonably stable throughout time [180]. Metagenomic and
metaproteomic analyses of stool samples of a cohort of twin pairs, for which data
on bacterial diversity were available, have reported a consistency in the findings of
these three approaches regarding microbial alterations in IBD [141, 181]. The
metagenomic reads of F. prausnitzii were significantly reduced in ICD, and in ten-
dency also the reads of Roseburia, in accord with reduced abundance of these spe-
cies confirmed by phylogenic sequencing. In addition to the metagenomic data, also
proteome analysis indicated the depletion of proteins derived from these species.
Especially bacterial proteins assigned to processes of replication, recombination,
and repair were enriched in CD, whereas those related to energy production and
nutrient transport and metabolism were depleted. In particular, in metagenomic as
well as proteomic approach, SCFA production pathways were underrepresented in
ICD. In total, there were significantly less microbial genes expressed and translated
to proteins in CD compared to healthy state (2 % vs. 8 %). This is consistent with
the finding that the richness of the metagenome is reduced in CD.
13.3.3 Metabonomics: Microbial Metabolic Products in IBD
In contrast to limited data from metatranscriptomics and metaproteomics, metabo-

nomics have been more frequently used to study the role of microbiota in maintain-
ing gut health and also in addressing differences between healthy state and
inflammatory situation. Despite sophisticated approaches [182], no early IBD bio-
markers could be established by the means of metabonomics up to now, but the
applications of metabonomics certainly helped to describe molecular patterns. Fecal
water, intestinal tissue, urine, or plasma is often used as specimen. As not all metab-
olites originating from bacteria will translocate or are absorbed and can be found in
the plasma or urine, the type of specimen influences the outcome of analysis. In the
following chapters, metabolites implicated in IBD are therefore discussed as
grouped according to the respective compartment in which they have been described
as regulated, with emphasis on the potential contribution of the microbiota to the
metabonome profiles. An overview of the findings discussed in the following is
given in Fig. 13.5. Sample specimens obtained by invasive methods (plasma and
intestinal tissue) have been used less frequently in IBD studies, probably explaining
a certain lack of described regulations for these compartments. Regardless of the
sample type, all the differences described can either stem from metabolic profiles
assigned to the host system or from bacteria if their metabolites are absorbed unless
the respective metabolite is uniquely produced by only one of the involved parties.
non-invasive invasive
urine plasma
TMA ?
AA inflammatory mediators
N-methylhistamine glycerophospholipids,
lipoproteins, lipids, AA
putative extent of penetration

of microbial metabolites
fecal water intestinal tissue

SCFA ?
secondary bile acids inflammatory mediators
pancreatic enzymes lipoproteins, lipids
primary bile acids myo-inositol
amino acids lactate
intestinal barrier function
Fig. 13.5 Samples of different origins are used for metabonomic approaches studying IBD-related
changes, describing the regulation of both host-derived (blue) and microbiota-derived (red) metab-
olites. Bacteria-derived metabolites may penetrate the intestinal barrier, depending on their size,
chemical properties, and the intestinal permeability. Their abundance and the extent to which they
are found regulated decreases in the subsequent compartments of the intestinal tissue, plasma, and
urine compared to fecal water. Host metabolites implicated in IBD in the respective sample types
are written in blue, bacteria-derived metabolites in red, and metabolites of unidentified origin in
black. TMA trimethylamine, AA amino acids, SCFA short-chain fatty acids
If conclusions shall be deduced for IBD etiopathology, it is therefore necessary to

validate any correlations by targeted approaches and in regard to the origin of the
metabolite.
13.3.3.1 Fecal Water
The analysis of metabolites in fecal water seems a direct approach to link the com-
position of bacteria to their activity reflected by their metabolites and to draw con-
clusions on how this might affect host intestinal health. Studies on the fecal
microbiota in twin cohorts have revealed that the respective metabonomes of fecal
water cluster in the same way as data for phylogenic composition and metapro-
teomics for the same samples do [182]. All these approaches highlight the fact that
CD samples are distinct from healthy samples but also that CCD samples differ
from ICD. Differences between CCD and healthy samples were not reported by
other -omics approaches in the same samples, indicating that the groups were more
distinct in metabolite profiles than other microbe-associated profiles. This suggests
that the fecal water metabonome might be a good means of studying potential
mechanisms in IBD etiopathology.
Amino Acids
The abundance of amino acids in fecal water is increased in IBD patients [183].
This might be due to reduced absorption of dietary amino acids or due to cata-
bolic actions of the host epithelium, but also an alteration of microbial metabo-
lism could result in changes in amino acids. Besides this ambivalent finding, also
clearly bacteria-derived metabolites are reportedly regulated in fecal water of
IBD patients.
Bile Acids
Metabolites related to bile acid (BA) synthesis and metabolism are often described
as differentially regulated between healthy and IBD samples and even between ICD
and CCD. Especially in ICD but also in irritable bowel syndrome or mouse models
of colitis, luminal contents of total BA are significantly elevated [182, 184, 185].
This increase in the total luminal BA concentration in ICD might reflect a loss of
transport activity as most of the BA should be reabsorbed in the ileum and undergo
enterohepatic circulation. Associations of specific microbial populations with the
respective metabolic profile of the feces were successfully established, with bacteria
of higher abundance in ICD correlating to BA and fatty acid metabolites. As bacte-
ria are differentially prone to antimicrobial effects exerted by BA and are differen-
tially able to metabolize BA species, alterations in BA concentration or composition
presumably result in concomitant shifts in microbial composition. Indeed, feeding
of BA-enriched diets altered microbial composition in rodents profoundly [162,
186]. Targeted quantifications have revealed particular implication of tauroconju-
gated (-SO3H) BA species in the lumen of UC patients [187]. Studies on animal
models of colitis have reported that higher abundance of taurocholic acid [162]
promotes the growth of SRB pathobionts that aggravate disease. Interestingly, cer-
tain luminal BA also decrease barrier function [188–190], probably via epidermal
growth factor receptor and/or modulation of the nuclear farnesoid X receptor in IEC
[191]. One study reported that the anti-inflammatory effect potentially exerted by
secondary BA is abolished upon sulfation of the respective species [187]. As
impaired barrier integrity and inept inflammatory response are the main features of
IBD, BA may trigger disease progression double-tracked – via direct action on IEC
and indirectly via modulation of the microbiota.
SCFA
The amount of SCFA, in particular butyrate, was markedly reduced in the fecal
water of CD patients, and to a lesser extent UC patients, compared to healthy con-
trols [183, 192]. The importance of butyrate for IEC functionality is described in
Sect. 13.1.2. The depletion of these bacteria-derived metabolites may reflect altera-
tions in microbial composition. In general, butyrate-producing bacteria have been
reported as diminished in IBD [142]. Especially Faecalibacterium prausnitzii is
regarded as an important butyrate producer and as outlined above is depleted in
IBD-affected mucosa. Also metagenomic and metaproteomic approaches have
revealed a depletion of butyrate and other SCFA production pathways in IBD [181]
and concomitant reduction of Faecalibacterium prausnitzii abundance. Decreased
SCFA production in IBD, especially CD, can thus be substantiated throughout the
levels of -omics approaches (also see Fig. 13.6). Indeed, enemas of butyrate have
been successfully used in UC treatment [43].
Fig. 13.6 IBD-related phenomena observed are partly interrelated across the different approaches
of assessing total microbial functionality. Dashed lines indicate the approach used, continuous
lines indicate the level of microbial functionality, and dotted lines indicate interrelations of find-
ings on these different levels
Arachidonic Acid Metabolites
Metabonomic analyses on fecal samples also revealed arachidonic acid (AA) and its
metabolites as implicated in IBD [182]. AA is converted by host enzymes into eico-
sanoids, mediators of inflammation. A regulation of eicosanoids was not convincingly
reflected in fecal water, but this might be attributed to the spatial resolution, as these
molecules exert their functions not in the lumen but at the basolateral side of the epithe-
lium. This highlights the restriction of metabonomics using fecal water: it may be
applied to study microbial metabolites but is limited to luminally secreted metabolites
on the host side, or those low amounts of metabolites shed with epithelial cells.
13.3.3.2 Intestinal Tissue
In contrast to abundance in fecal water samples, tissue levels of both AA and eico-
sanoids were reportedly elevated in animal models [193, 194] and IBD patients
[194], emphasizing the importance of metabolite compartmentalization. Untargeted
metabonomic studies on intestinal biopsy tissues were effective in distinguishing
healthy from UC [195] or CD samples [196] and also could discriminate between
CD and UC [196]. Differences described within the metabolite profiles of healthy
versus inflamed intestine lie especially within the group of glycerophospholipids,
with decreases in glycerophosphocholine [195, 197] and lysophosphatidylcholines
[193]. These findings are likely associated with the generation of inflammatory lipid
mediators and altered tissue morphology, such as the infiltration of immune cells
and changes in membrane fluidity. Tissue concentrations of myoinositol are
decreased in UC patients [195, 197]. As myoinositol could be food or host derived
[198], the meaning of this finding is not yet clear. It either reflects impaired absorp-
tion or could be related to alterations in membrane-incorporated phosphatidylinosi-
tol [199]. Lactate concentrations in intestinal tissue seem to be reduced in both UC
and CD patients [197]. Despite the mechanistic inexplicability of this finding, this
might serve as justification for the use of lactate-producing bacterial strains as pro-
biotics in IBD therapy (see Sect. 13.13.2).
Although metabonomic studies of intestinal tissue have not revealed further
microbiota-related alterations, these studies have elucidated molecular mechanisms
during microbiota-driven pathogenesis.
13.3.3.3 Plasma
Up to now, there is no convincing description of IBD-associated alterations of micro-

bial metabolites in human plasma samples. This may also be due to a limited number
of studies analyzing plasma metabolites in IBD patients compared to fecal water or
urinary metabolites. Inflammatory mediators including cytokines, eicosanoids, and
prostaglandins are elevated in the plasma of IBD patients. In addition, many of the
plasma metabolites found implicated in murine IBD-like pathologies reflect changes

in host energy homeostasis, such as lipoproteins, glycerophospholipids, lipids, or
amino acids [200]. These changes probably result from inflammation-associated
symptoms such as malabsorption, wasting, and protein catabolism.
13.3.3.4 Urine
As it is easy to collect and to process, urine seems a useful reservoir to apply metabo-
nomic approaches. Indeed, several studies demonstrate that metabonomics can be
applied on urinary metabolite profiles for the discrimination of healthy state from coli-
tis in mice [201] and IBD in humans [170, 202]. Increased levels of amino acids prob-
ably reflect a catabolic state of the host. N-Methylhistamine, presumably deriving from
mast cells, was found elevated in the urine of IBD patients [203]. The majority of the
regulated metabolites described for urinary samples are host derived. It can be specu-
lated that urinary samples may reflect the absorbed and excreted part of the microbial
metabonome. Trimethylamine, a metabolite uniquely produced by the microbiota, was
significantly elevated in the urine of colitic mice and correlated to disease progression
[201]. But as IBD is also characterized by impaired barrier integrity, increased abun-
dance of microbial metabolites could simply be an artifact of increased translocation
and is not necessarily directly associated with IBD etiopathology.
13.3.4 Integration of Meta-omics Approaches Regarding

Bacterial Functionality in IBD
In summary, -omics approaches allow the deduction of aspects of microbe-host

interactions that can contribute major knowledge to the field of IBD research. The
fact that metagenomics, metatranscriptomics, metaproteomics, and metabonomics
(on fecal water at least) are able to similarly reflect changes in microbial functional-
ity, as described above and summarized in Fig. 13.6, is auspicious. Nevertheless,
these techniques need extensive data mining and are limited by the availability of
the characterization of genes, transcripts, proteins, and metabolites, respectively.
13.3.5 Integrating Microbial and Host -Omics

in Regard to IBD
An appealing approach to gain better insight into microbe-host interactions is the

integration of data sets of both microbial composition or functionality and host
molecular processes. Using microbe compositional data and host transcriptional
profiles, it was described that there are correlations between colitis-associated
intestinal bacteria and potentially disease-relevant transcripts in the colonic mucosa

[144]. In this context, reduced overall correlations and thus a loss of microbe-host
cross talk in UC patients and unaffected twins of the same genotypic background as
compared to unaffected healthy donors have been reported.
Furthermore, using metaproteomics on fecal samples gives the opportunity to
detect both bacterial and human proteins, respectively. Unlike metabonomics, meta-
proteomic analyses can discriminate between bacterial and host sources. Using the
fecal samples from a cohort of twins with healthy, concordant, or disconcordant
status for CD, for which bacterial -omics data are available, and then characterizing
the host proteome allowed drawing conclusions about the relationship between host
status and observed microbial differences. The host proteome in stool of CD
patients, predominantly ICD, was enriched in proteins associated with inflamma-
tory response and host defense (e.g., α-defensin), as well as wound healing [181].
Proteins associated with barrier integrity were depleted in ICD patients (e.g., proto-
cadherin). These findings go along with the notion that IBD patients suffer from a
defective epithelial barrier and a compromised defense toward microbiota at the
epithelial interface. Furthermore, there were high amounts of pancreatic enzymes
found in ICD feces [181], indicating impaired absorption and correlating to the find-
ing that luminal BA metabolites are increased in the same CD patients [182].
13.4 Dietary Manipulation of Intestinal Microbiota

in the Context of IBD
The intestinal microbiota is not only a key driver of inflammation in the context of IBD
[90] but also represents an interface between environment (diet) and host. The manipu-
lation of the bacterial composition of the microbiota via diet might be a promising
means of IBD treatment, especially in pediatrics where aggressive therapies are
eschewed. It is crucial to discriminate between preventive and therapeutic strategies, as
well as applications that intend to prolong phases of remission. Some alimentary strate-
gies and food components have been reportedly successful regarding the latter. However,
the effects of some approaches differ between the etiopathologies of CD, UC, and pou-
chitis, an IBD-like inflammation of an artificial rectum surgically formed out of ileal
tissue during colectomy. Besides the nutritional therapies, research also focuses on the
identification of dietary components that may affect the risk of IBD development.
13.4.1 Prebiotics
Given the potential of diet to modulate intestinal microbial composition and func-
tionality, there are many approaches to implement nutritional intervention in IBD
treatment strategies, in particular in the context of prebiotics. The concept of
prebiotics is the indirect modulation of microbial composition via the selective

enhancement of favorable strains. When indigestible dietary fibers reach the
colon, they are fermented by anaerobic bacteria, resulting in the production of
SCFA. Importantly, SCFA not only serve as energy substrates for bacteria and thus
enhance growth but can also alter intestinal microbial composition via a drop of the
luminal pH and selective fostering of certain bacteria [204]. In humans, the prebi-
otic compounds inulin, lactulose, and fructo- and galacto-oligosaccharides have
been shown to promote the growth of Bifidobacteria, Eubacteria, and Lactobacilli
[205–209], which are considered to exhibit health benefits [210]. Randomized con-
trolled trials using prebiotics in IBD patients are scarce, but small trials indicate
reduced disease activity with concomitant alterations of bacterial composition upon
inulin or fructo-oligosaccharide ingestion in the condition of CD [211] and of pou-
chitis [212]. Positive effects of prebiotics were also shown in animal studies on
chemically induced colitis [213–216] and transgenic rodent models [217, 218], with
increased amounts of Lactobacilli and Bifidobacteria and decreased macroscopic
inflammation. Some of these studies in both humans and mouse models also report
increased levels of butyrate [212–214].
13.4.2 Probiotics
Probiotics aim to actively alter microbial composition by introducing live bacteria. The
prerequisite for a successful probiotic action therefore is the survival and sustaining of
the bacteria in the complex and competitive microbial network. Originally derived
from fermented food, especially dairy products, probiotic strains studied in regard to
IBD include above all Lactobacilli, Bifidobacteria, and nonpathogenic E. coli Nissle
1917, as well as mixtures of these strains and a handful of other species. Human studies
are limited in patient numbers and are focused on a variety of probiotic strains or com-
binations, different outcomes, and diverse kinds of cohorts. Meta-analyses try to esti-
mate the overall beneficial effect of probiotics and indicated that in summary probiotics
might be applicable in patients with colonic phenotype for the maintenance of remis-
sion rather than induction of remission or treatment of active disease [219].
The use of Lactobacilli has been studied in several mouse model experiments as
well as clinical trials. Different strains (GG, paracasei, plantarum, salivarius) were
found to exert a variety of effects on host immune response, such as induction of
mucin expression [220] or suppression of cytokine production [221, 222] and T-cell
proliferation [223]. Although there is a lack of trials in IBD patients that convinc-
ingly show a beneficial effect regarding disease activity or remission maintenance
[224, 225], there is evidence for the amelioration of colitis in different mouse mod-
els in the case of Lactobacilli [221, 222, 226–228]. It became apparent that the
probiotic strains tested for effects on IBD have to be chosen carefully, and also in
regard to the experimental setup, when a study found L. plantarum and LGG to
aggravate DSS-induced colitis, whereas L. paracasei exhibited protective effects
[229]. In addition, E. faecalis, a member of the core gut microbiome historically
applied as probiotic strain, has been shown to compromise the intestinal barrier via
the matrix metalloprotease GelE and to thus contribute to IBD pathogenesis in sus-
ceptible mouse models [230].
Probiotic Bifidobacteria of various strains (infantis, breve, bifidum) resulted in
reduced inflammation in mouse models of colitis [221, 227, 231–234], and
Bifidobacterium-fermented milk was found to ameliorate disease in UC patients
[235, 236]. The mechanisms observed are similar to those reported for Lactobacilli,
such as alterations of cytokine levels [231, 233]. As in the case of Lactobacilli,
randomized placebo-controlled studies successfully using Bifidobacteria in IBD
patients are scarce.
E. coli Nissle, a nonpathogenic E. coli strain, was effective in attaining as well as
maintaining remission in UC patients [237–239]. This strain might exert beneficial
effects by strengthening the intestinal barrier [240], similarly to the barrier-related
effects of both Lactobacilli and Bifidobacteria [222, 241–243].
Also mixtures of probiotics are commonly used. The formula VSL#3 combines
eight different probiotic lactic acid bacteria: four strains of Lactobacilli (acidophi-
lus, bulgaricus, casei, plantarum), three strains of Bifidobacteria (breve, infantis,
longum), and Streptococcus thermophilus, all of which are usually present in the
human intestinal microbiota. This mixture has been shown to induce or maintain
remission in pouchitis [244, 245], mild to moderate UC [246, 247], and pediatric
UC patients [248]. In contrast to these promising results in pouchitis and UC, pro-
biotic therapies failed to show effects in CD patients [249, 250]. The mechanism of
action is poorly understood and the combination of the different strains impedes the
study of molecular mechanisms. Animal models have been used to study the effects
of VSL#3 concerning immune stimulation and regulation, some of which have
reported a DNA-dependent effect of the probiotic mixture [251, 252]. One secreted
component of VSL#3, the L. paracasei-derived prtP-encoded lactocepin, was
shown to reduce inflammation in a murine colitis model [253], due to anti-
inflammatory effects via the degradation of chemokines [254].
Noteworthy are also the effects observed for the use of synbiotics. The idea
behind synbiotic approaches is to promote the survival and growth of the applied
probiotics by simultaneous ingestion of prebiotics. Indeed, clinical trials revealed
beneficial effects of synbiotics in UC patients [255, 256].
In summary, there is a need for more randomized, double-blind clinical trials focus-
ing on the feasibility of probiotics in IBD therapy. In many cases, it is not clear if live
bacteria, bacterial surface compounds, or secreted bacterial compounds are needed for
the effects observed. Evidence for beneficial effects of probiotics, mainly in the case of
E. coli Nissle and VSL#3, in both animal models and clinical trials is promising, though
for colitis and pouchitis only and not for inflammation of the small intestine. The dif-
ferent etiologies of CD and UC might be responsible for this discrepancy. Recent stud-
ies indicate that eukaryotes such as Saccharomyces boulardii [257, 258] or helminths
[259] might be effective in prolonging remission phases in CD, probably due to selec-
tive attenuation of Th1 antigenic responses (implicated in IBD) [260], and shift to Th2
(directed against parasites; also see Fig. 13.4) [261, 262]. However, the research on the
use of these organisms in IBD is still in its infancy.
13.4.3 Enteral Nutrition and Elemental Diet
Enteral nutrition has been applied successfully in the treatment of IBD [263], espe-
cially in children with CD [264–266]. As IBD is characterized by a reduced bacte-
rial diversity, it might appear contradictory that enteral nutrition seems to reduce
bacterial diversity by depriving the microbiota of nutrients [267]. The reduction of
diversity might correlate to a reduced antigenic load of the microbiota in this case.
But a reduction of mucosal antigen exposure could also result from the nature of the
feed, as semisynthetic enteral nutrition formulas are very low in bacterial antigens.
In addition, a liquid diet results in faster intestinal transit, reducing the time of expo-
sure to ingested putative antigens. Alternatively, enteral nutrition might also be able
to modulate the metabolic activity of the microbiota [268].
A reduction of microbial diversity is also observed with the therapeutic use of
elemental diets in IBD [269]. The term elemental diet describes easily digestible
and peptide- or amino acid-based formulas. Some clinical trials support the effica-
ciousness of commercially available formulas in decreasing disease activity indices
and relapse frequency comparably to steroid treatments [270–273].
The use of enteral nutrition or also certain elemental diets in IBD therapy seems
promising, but still these treatment approaches do not always produce equally ben-
eficial results as standard therapies using medication [274]. In addition, the mecha-
nisms behind the amelioration of disease are not clearly understood. In general, as
both nutritional approaches involve only easily digestible formulas, usually contain-
ing low amounts of fat, a general relief of the digestive systems in terms of enzyme
production, chylomicron formation, and possibly mechanical stimulation is induced.
In addition, non-polymeric diets are free of gluten: mouse studies indicated a poten-
tial aggravation of disease by gluten-containing diets [275], and gastrointestinal
symptoms of non-celiac disease patients are ameliorated by gluten-free diet [276].
These possibly affected digestive processes concern the small intestine rather
than the colon, which goes along with the finding that enteral and elemental diets
are more effective in ileal than colonic IBD phenotypes [277]. Together with pre-
sumably altered composition or functionality of microbiota, there are numerous
effects of these approaches on the intestinal tract, hampering the elucidation of
molecular mechanisms.
13.4.4 Indications for Nutrients as Selection Factors

for Pathobionts
In the search of dietary factors contributing to IBD development, various nutritional

compounds and their putative molecular functions have been thoroughly discussed
[278–280]. Attempts to link pre-illness diet to IBD risk did not provide consistent
results [281–284].
Among the dietary factors suspected to trigger IBD are mainly carbohydrates
and fat, in particular refined sugar, saturated fatty acids, and omega-6 fatty acids,
whereas dietary fibers and food groups such as fruit and vegetables may be
protective [280]. The lack of unequivocal evidence from epidemiological studies
for the implication of specific dietary factors might result from the complex etiol-
ogy of IBD and also indicates that diet probably only affects IBD development in
combination with host genetics and other environmental factors. However, there
are accumulating indications for a role of dietary fat in IBD pathogenesis. A sys-
tematic literature review of retrospective epidemiologic data concluded that a high
dietary intake of fat, adjusted for total energy intake and behavioral confounders,
increases risk of IBD [280]. In contrast, body mass index or total energy intake did
not correlate to the development of incident IBD in a prospective cohort study
[285]. This indicates a specific role for dietary lipids in the pathogenesis of
IBD. Indeed, high-fat diets (HFD) have been shown to impair intestinal health in
chemical and genetic animal models of colitis [162, 286, 287] and CD-like ileitis
[288]. There are indications from animal studies that HFD promotes intestinal
inflammatory processes [287, 289, 290] and perturbs barrier function [189, 288,
291], but the mechanisms underlying the facilitation of disease onset or its aggra-
vation remain not fully elucidated. HFD modulates the composition of the intesti-
nal microbiota [287, 292]. The modulation of the bacterial composition by HFD
ingestion can result in increased pathogenicity of the intestinal microbiota, espe-
cially by fostering the growth of pathobionts that induce inappropriate inflamma-
tory response in a susceptible host [162]. There are indications that HFD exerts this
modulation of the microbial ecosystem via shifts in BA composition [162, 186,
189]. Only 1 day after facing a diet high in animal fat, the microbial community
starts to shape accordingly, with increased abundance of bile acid-tolerant species
and reduction of plant carbohydrate-metabolizing species [293]. HFD-associated
modification of the microbial composition, reduction of barrier function, and
aggravation of pathology are reportedly independent of an obese phenotype [189,
288, 292]. In summary, diets rich in fat may aggravate IBD via the modulation of
intestinal luminal factors. This mechanism could also contribute to the beneficial
effects reported for enteral and elemental nutrition, as these forms of alimentation
generally lead to decreased fat consumption compared to a typical Western diet.
Likewise, there is evidence for iron playing a crucial role in IBD pathogenesis from
a mouse model with CD-like phenotype: dietary but not systemic iron aggravated dis-
ease compared to a low-iron diet, probably via the modulation of intestinal microbial
composition, and in combination with increased cellular stress in IECs [294]. These
findings might particularly impact on IBD treatment in the future, considering the fact
that IBD patients are often anemic and treated with iron supplements.
13.4.5 Nutrients: Nutrigenetics in IBD
Besides the effects of diet on intestinal microbial composition and function, diet
may also impact on IBD in a more direct way. Nutrigenetics in the context of IBD
aims at identifying nutrient-gene interactions associated with disease-conditioning
and disease-modulating situation. Thereby, two main aspects have come into
Fig. 13.7 Nutrigenetics in the context of IBD investigate nutrient-gene interactions associated
with disease-conditioning and disease-modulating situation. The regulation of cellular energy
demand, for example, via butyrate import, and the effect of transcription factors activated by nutri-
ents are critical in IBD. MCT1 monocarboxylate transporter 1, ATP adenosine triphosphate, PepT1
peptide transporter 1, MDP muramyl dipeptide, NOD2 nucleotide-binding oligomerization domain
2, NFκB nuclear factor κB, MDR1 multidrug resistance protein 1, AhR aryl hydrocarbon receptor,
Vit A vitamin A, XRE xenobiotic response element, RXR retinoid X receptor, RARE retinoic acid
response element, Vit D vitamin D, VDR vitamin D receptor, VDRE VDR responsive element. FA
fatty acid, PPAR peroxisome proliferator-activated receptor, PRE PPAR responsive element, * IBD
susceptibility gene, ** regulated in IBD
play: (1) the general regulation of cellular energy demand and (2) transcription fac-
tors activated by nutrients as the closest interface of diet and regulation of gene
expression. Major findings are summarized in Fig. 13.7.
13.4.5.1 Cellular Energy Demand
Meeting the cellular energy demand might be of particular importance in IECs

that are challenged by energy-consuming inflammatory processes. Besides
impaired β-oxidation resulting from polymorphisms in the carnitine trans-
porter OCTN2 [ 105 ], there are reports of decreased butyrate oxidation in the
inflamed mucosa of UC patients [295, 296 ] and in animal models of experi-
mental colitis [ 297]. Additionally, the monocarboxylate transporter MCT1,
responsible for butyrate uptake in colonocytes [298], is reportedly downregu-
lated in IBD [299]. Beneficial modulation of the microbial composition could
enhance butyrate supply and attenuate energy deficiency in IEC. Notably,
luminal levels of acetate and butyrate were reported to decrease rapidly with
ingestion of diets high in fat [293].
13.4.5.2 Nutrient-Induced Signaling
Diet- and microbiota-derived xenobiotics are actively removed from IEC via the
multidrug resistance protein (MDR1). Variants of MDR1 are associated with UC
[300, 301]. Xenobiotic substances remaining in the host cell may activate the tran-
scription factor aryl hydrocarbon receptor (Ahr) [302]. Derivatives of diet-derived
vitamins A and D bind to transcription factors retinoid X receptor (RXR) and vita-
min D receptor (VDR), respectively, whereas fatty acids activate peroxisome
proliferator-activated receptors (PPARs). VDR polymorphisms are associated with
IBD [303], and also contribution of PPARs is under debate [302, 304]. PepT1, a
major brush border peptide transporter, is also implicated as IBD susceptibility gene
[305]. It allows uptake of muramyl dipeptide which activates NOD2 and thus
inflammatory response pathways via NFκB [306, 307]. Activated transcription fac-
tors such as NFκB, Ahr, RXR, VDR, or PPAR modulate immune response down-
stream and thus in turn also potentially impact on microbiota, emphasizing the close
interrelations of host, dietary factors, and microbiota.
In summary, it seems likely that dietary components or alimentary strategies
impact on IBD directly or through alterations of the composition or metabolic activ-
ity of the microbiota. Therefore, personalized nutrition strategies might profoundly
improve disease activity or remission maintenance in the future.
13.5 Conclusions and Perspective
While the human intestinal microbiota is needed to build up and maintain an

adequate immune system, a loss of homeostasis regarding the interactions between
microbes and the host intestinal mucosal immune system can result in aberrant or
disproportionate inflammatory responses, with IBD as an example with numerous
cases. Given that the microbial ecology is a major modulator of gastrointestinal
health, the manipulation of its composition and functionality seems a promising
therapeutic strategy for acute infections as well as chronic diseases such as
IBD. Metabonomics as well as other -omics techniques will be useful monitoring
the degree of modulation and may help reveal functional interactions.
While dietary manipulation via nutrients, probiotics, or prebiotics is a rather
mild way of interference, more dramatic changes can be achieved using antibiotics
or fecal microbial transplantation (FMT). Transplanting donor microbiota has been
successfully performed in cases of Clostridium difficile infections [308], but its
application in IBD is still rather new and controversial. Mainly, case reports are
available so far regarding the management of IBD by FMT, with many of them
suggesting a beneficial effect as reviewed/summarized in [309, 310]. A pilot trial
reported the alleviation of symptoms in pediatric UC patients [311]. No serious

adverse effects were reported in this study, whereas fever and an increase in plasma
inflammatory markers were observed in a study involving five adult UC patients,
while only one patient showed symptom improvement [312].
It is believed that FMT can restore antibiotic-related disruption intestinal of the
microbial diversity [308]. However, not much is known about the mechanisms that
render one microbial community of a certain shape harmless while another turns out
deleterious. The research for defining optimal donor microbiota characteristics,
donor-recipient match parameters, and pre-FMT conditioning is crucial for FMT to
be further adapted and developed into an approved therapy and might at the same
time provide more insight into mechanisms of microbe-host cross talk during intes-
tinal health and disease.
References
1. Artis D. Epithelial-cell recognition of commensal bacteria and maintenance of immune

homeostasis in the gut. Nat Rev Immunol. 2008;8(6):411–20. Epub 2008/05/13.
2. Ley RE, Peterson DA, Gordon JI. Ecological and evolutionary forces shaping microbial
diversity in the human intestine. Cell. 2006;124(4):837–48. Epub 2006/02/25.
3. Zoetendal EG, von Wright A, Vilpponen-Salmela T, Ben-Amor K, Akkermans AD, de Vos
WM. Mucosa-associated bacteria in the human gastrointestinal tract are uniformly distrib-
uted along the colon and differ from the community recovered from feces. Appl Environ
Microbiol. 2002;68(7):3401–7. Epub 2002/06/29.
4. Dethlefsen L, McFall-Ngai M, Relman DA. An ecological and evolutionary perspective
on human-microbe mutualism and disease. Nature. 2007;449(7164):811–8. Epub
2007/10/19.
5. Scott KP, Gratz SW, Sheridan PO, Flint HJ, Duncan SH. The influence of diet on the gut
microbiota. Pharmacol Res: Off J Ital Pharmacol Soc. 2013;69(1):52–60. Epub
2012/11/14.
6. Derrien M, Vaughan EE, Plugge CM, de Vos WM. Akkermansia muciniphila gen. nov., sp.
nov., a human intestinal mucin-degrading bacterium. Int J Syst Evol Microbiol.
2004;54(Pt 5):1469–76.
7. Berry D, Stecher B, Schintlmeister A, Reichert J, Brugiroux S, Wild B, et al. Host-compound
foraging by intestinal microbiota revealed by single-cell stable isotope probing. Proc Natl
Acad Sci U S A. 2013;110(12):4720–5. Epub 2013/03/15.
8. Martens EC, Koropatkin NM, Smith TJ, Gordon JI. Complex glycan catabolism by the human
gut microbiota: the Bacteroidetes Sus-like paradigm. J Biol Chem. 2009;284(37):24673–7.
Epub 2009/06/26.
9. Sonnenburg JL, Xu J, Leip DD, Chen CH, Westover BP, Weatherford J, et al. Glycan foraging
in vivo by an intestine-adapted bacterial symbiont. Science. 2005;307(5717):1955–9. Epub
2005/03/26.
10. Granato D, Bergonzelli GE, Pridmore RD, Marvin L, Rouvet M, Corthesy-Theulaz IE. Cell
surface-associated elongation factor Tu mediates the attachment of Lactobacillus johnsonii
NCC533 (La1) to human intestinal cells and mucins. Infect Immun. 2004;72(4):2160–9.
Epub 2004/03/25.
11. Ruiz-Palacios GM, Cervantes LE, Ramos P, Chavez-Munguia B, Newburg DS. Campylobacter
jejuni binds intestinal H(O) antigen (Fuc alpha 1, 2Gal beta 1, 4GlcNAc), and fucosyloligosac-
charides of human milk inhibit its binding and infection. J Biol Chem. 2003;278(16):14112–20.
Epub 2003/02/04.
12. Clavel T, Charrier C, Braune A, Wenning M, Blaut M, Haller D. Isolation of bacteria from the
ileal mucosa of TNFdeltaARE mice and description of Enterorhabdus mucosicola gen. nov.,
sp. nov. Int J Syst Evol Microbiol. 2009;59(Pt 7):1805–12.
13. Begon M, Harper JL, Townsend CR. Ecology. Individuals, populations and communities. 3rd
ed. Wiley-Blackwell; 1996. Oxford.
14. Macpherson AJ, Harris NL. Interactions between commensal intestinal bacteria and the
immune system. Nat Rev Immunol. 2004;4(6):478–85. Epub 2004/06/03.
15. Bouskra D, Brezillon C, Berard M, Werts C, Varona R, Boneca IG, et al. Lymphoid tissue
genesis induced by commensals through NOD1 regulates intestinal homeostasis. Nature.
2008;456(7221):507–10. Epub 2008/11/07.
16. Tlaskalova-Hogenova H, Sterzl J, Stepankova R, Dlabac V, Veticka V, Rossmann P, et al.
Development of immunological capacity under germfree and conventional conditions. Ann N
Y Acad Sci. 1983;409:96–113. Epub 1983/06/30.
17. Moreau MC, Ducluzeau R, Guy-Grand D, Muller MC. Increase in the population of duodenal
immunoglobulin A plasmocytes in axenic mice associated with different living or dead bacte-
rial strains of intestinal origin. Infect Immun. 1978;21(2):532–9. Epub 1978/08/01.
18. Umesaki Y, Setoyama H, Matsumoto S, Okada Y. Expansion of alpha beta T-cell receptor-
bearing intestinal intraepithelial lymphocytes after microbial colonization in germ-free mice
and its independence from thymus. Immunology. 1993;79(1):32–7. Epub 1993/05/01.
19. Duan J, Chung H, Troy E, Kasper DL. Microbial colonization drives expansion of IL-1
receptor 1-expressing and IL-17-producing gamma/delta T cells. Cell Host Microbe.
2010;7(2):140–50. Epub 2010/02/18.
20. Williams AM, Probert CS, Stepankova R, Tlaskalova-Hogenova H, Phillips A, Bland
PW. Effects of microflora on the neonatal development of gut mucosal T cells and myeloid
cells in the mouse. Immunology. 2006;119(4):470–8. Epub 2006/09/26.
21. Niess JH, Brand S, Gu X, Landsman L, Jung S, McCormick BA, et al. CX3CR1-mediated den-
dritic cell access to the intestinal lumen and bacterial clearance. Science. 2005;307(5707):254–8.
Epub 2005/01/18.
22. Rescigno M, Urbano M, Valzasina B, Francolini M, Rotta G, Bonasio R, et al. Dendritic cells
express tight junction proteins and penetrate gut epithelial monolayers to sample bacteria.
Nat Immunol. 2001;2(4):361–7. Epub 2001/03/29.
23. Iliev ID, Mileti E, Matteoli G, Chieppa M, Rescigno M. Intestinal epithelial cells promote
colitis-protective regulatory T-cell differentiation through dendritic cell conditioning.
Mucosal Immunol. 2009;2(4):340–50. Epub 2009/04/24.
24. Siddiqui KR, Powrie F. CD103+ GALT DCs promote Foxp3+ regulatory T cells. Mucosal
Immunol. 2008;1 Suppl 1:S34–8. Epub 2008/12/23.
25. Schulz O, Jaensson E, Persson EK, Liu X, Worbs T, Agace WW, et al. Intestinal CD103+, but
not CX3CR1+, antigen sampling cells migrate in lymph and serve classical dendritic cell
functions. J Exp Med. 2009;206(13):3101–14. Epub 2009/12/17.
26. Schlitzer A, McGovern N, Teo P, Zelante T, Atarashi K, Low D, et al. IRF4 transcription
factor-dependent CD11b+ dendritic cells in human and mouse control mucosal IL-17 cyto-
kine responses. Immunity. 2013;38(5):970–83. Epub 2013/05/28.
27. Watchmaker PB, Lahl K, Lee M, Baumjohann D, Morton J, Kim SJ, et al. Comparative tran-
scriptional and functional profiling defines conserved programs of intestinal DC differentia-
tion in humans and mice. Nat Immunol. 2013;8:22. Epub 2013/12/03.
28. Mazmanian SK, Round JL, Kasper DL. A microbial symbiosis factor prevents intestinal
inflammatory disease. Nature. 2008;453(7195):620–5. Epub 2008/05/30.
29. Atarashi K, Nishimura J, Shima T, Umesaki Y, Yamamoto M, Onoue M, et al. ATP drives lam-
ina propria T(H)17 cell differentiation. Nature. 2008;455(7214):808–12. Epub 2008/08/22.
30. Ivanov II, Atarashi K, Manel N, Brodie EL, Shima T, Karaoz U, et al. Induction of intestinal
Th17 cells by segmented filamentous bacteria. Cell. 2009;139(3):485–98. Epub 2009/10/20.
31. Gaboriau-Routhiau V, Rakotobe S, Lecuyer E, Mulder I, Lan A, Bridonneau C, et al. The key
role of segmented filamentous bacteria in the coordinated maturation of gut helper T cell
responses. Immunity. 2009;31(4):677–89. Epub 2009/10/17.
32. Uematsu S, Fujimoto K, Jang MH, Yang BG, Jung YJ, Nishiyama M, et al. Regulation of
humoral and cellular gut immunity by lamina propria dendritic cells expressing Toll-like
receptor 5. Nat Immunol. 2008;9(7):769–76. Epub 2008/06/03.
33. Atarashi K, Tanoue T, Shima T, Imaoka A, Kuwahara T, Momose Y, et al. Induction of colonic
regulatory T cells by indigenous Clostridium species. Science. 2011;331(6015):337–41.
Epub 2011/01/06.
34. Round JL, Lee SM, Li J, Tran G, Jabri B, Chatila TA, et al. The Toll-like receptor 2 path-
way establishes colonization by a commensal of the human microbiota. Science.
2011;332(6032):974–7. Epub 2011/04/23.
35. Chung DR, Kasper DL, Panzo RJ, Chitnis T, Grusby MJ, Sayegh MH, et al. CD4+ T cells
mediate abscess formation in intra-abdominal sepsis by an IL-17-dependent mechanism. J
Immunol. 2003;170(4):1958–63. Epub 2003/02/08.
36. Gibson 3rd FC, Onderdonk AB, Kasper DL, Tzianabos AO. Cellular mechanism of intraab-
dominal abscess formation by Bacteroides fragilis. J Immunol. 1998;160(10):5000–6. Epub
1998/05/20.
37. Bauer H, Horowitz RE, Levenson SM, Popper H. The response of the lymphatic tissue to the
microbial flora. Studies on germfree mice. Am J Pathol. 1963;42:471–83. Epub 1963/04/01.
38. Hapfelmeier S, Lawson MA, Slack E, Kirundi JK, Stoel M, Heikenwalder M, et al. Reversible
microbial colonization of germ-free mice reveals the dynamics of IgA immune responses.
Science. 2010;328(5986):1705–9. Epub 2010/06/26.
39. Durkin HG, Chice SM, Gaetjens E, Bazin H, Tarcsay L, Dukor P. Origin and fate of IgE-
bearing lymphocytes. II Modulation of IgE isotype expression on Peyer’s patch cells by feed-
ing with certain bacteria and bacterial cell wall components or by thymectomy. J Immunol.
1989;143(6):1777–83. Epub 1989/09/15.
40. Segain JP, Raingeard, de la Bletiere D, Bourreille A, Leray V, Gervois N, Rosales C, et al.
Butyrate inhibits inflammatory responses through NFkappaB inhibition: implications for
Crohn’s disease. Gut. 2000;47(3):397–403. Epub 2000/08/15.
41. Saemann MD, Bohmig GA, Osterreicher CH, Burtscher H, Parolini O, Diakos C, et al. Anti-
inflammatory effects of sodium butyrate on human monocytes: potent inhibition of IL-12 and
up-regulation of IL-10 production. FASEB J: Off Pub Fed Am Soc Experi Biol.
2000;14(15):2380–2. Epub 2000/10/12.
42. Maslowski KM, Vieira AT, Ng A, Kranich J, Sierro F, Yu D, et al. Regulation of inflammatory
responses by gut microbiota and chemoattractant receptor GPR43. Nature. 2009;461(7268):1282–
6. Epub 2009/10/30.
43. Scheppach W, Sommer H, Kirchner T, Paganelli GM, Bartram P, Christl S, et al. Effect of
butyrate enemas on the colonic mucosa in distal ulcerative colitis. Gastroenterology.
1992;103(1):51–6. Epub 1992/07/01.
44. Chung H, Pamp SJ, Hill JA, Surana NK, Edelman SM, Troy EB, et al. Gut immune matura-
tion depends on colonization with a host-specific microbiota. Cell. 2012;149(7):1578–93.
Epub 2012/06/26.
45. Meyer-Hoffert U, Hornef MW, Henriques-Normark B, Axelsson LG, Midtvedt T, Putsep K,
et al. Secreted enteric antimicrobial activity localises to the mucus surface layer. Gut.
2008;57(6):764–71. Epub 2008/02/06.
46. Vaishnava S, Behrendt CL, Ismail AS, Eckmann L, Hooper LV. Paneth cells directly sense gut
commensals and maintain homeostasis at the intestinal host-microbial interface. Proc Natl
Acad Sci U S A. 2008;105(52):20858–63. Epub 2008/12/17.
47. Ayabe T, Satchell DP, Wilson CL, Parks WC, Selsted ME, Ouellette AJ. Secretion of micro-
bicidal alpha-defensins by intestinal Paneth cells in response to bacteria. Nat Immunol.
2000;1(2):113–8. Epub 2001/03/15.
48. Cash HL, Whitham CV, Behrendt CL, Hooper LV. Symbiotic bacteria direct expression of an
intestinal bactericidal lectin. Science. 2006;313(5790):1126–30. Epub 2006/08/26.
49. Hooper LV, Stappenbeck TS, Hong CV, Gordon JI. Angiogenins: a new class of microbicidal
proteins involved in innate immunity. Nat Immunol. 2003;4(3):269–73. Epub 2003/01/28.
50. Kida Y, Shimizu T, Kuwano K. Sodium butyrate up-regulates cathelicidin gene expression via
activator protein-1 and histone acetylation at the promoter region in a human lung epithelial
cell line, EBC-1. Mol Immunol. 2006;43(12):1972–81. Epub 2006/01/21.
51. Burger-van Paassen N, Loonen LM, Witte-Bouma J, Korteland-van Male AM, de Bruijn AC,
van der Sluis M, et al. Mucin Muc2 deficiency and weaning influences the expression of the
innate defense genes Reg3beta, Reg3gamma and angiogenin-4. PLoS ONE. 2012;7(6):e38798.
Epub 2012/06/23.
52. Menard S, Forster V, Lotz M, Gutle D, Duerr CU, Gallo RL, et al. Developmental switch of
intestinal antimicrobial peptide expression. J Exp Med. 2008;205(1):183–93. Epub
2008/01/09.
53. Salzman NH, Hung K, Haribhai D, Chu H, Karlsson-Sjoberg J, Amir E, et al. Enteric defen-
sins are essential regulators of intestinal microbial ecology. Nat Immunol. 2010;11(1):76–83.
Epub 2009/10/27.
54. Peterson DA, McNulty NP, Guruge JL, Gordon JI. IgA response to symbiotic bacteria as a
mediator of gut homeostasis. Cell Host Microbe. 2007;2(5):328–39. Epub 2007/11/17.
55. Pasquier B, Launay P, Kanamaru Y, Moura IC, Pfirsch S, Ruffie C, et al. Identification of
FcalphaRI as an inhibitory receptor that controls inflammation: dual role of FcRgamma
ITAM. Immunity. 2005;22(1):31–42. Epub 2005/01/25.
56. Macpherson AJ, Uhr T. Induction of protective IgA by intestinal dendritic cells carrying com-
mensal bacteria. Science. 2004;303(5664):1662–5. Epub 2004/03/16.
57. Suzuki K, Meek B, Doi Y, Muramatsu M, Chiba T, Honjo T, et al. Aberrant expansion of
segmented filamentous bacteria in IgA-deficient gut. Proc Natl Acad Sci U S A.
2004;101(7):1981–6. Epub 2004/02/10.
58. Yanagibashi T, Hosono A, Oyama A, Tsuda M, Hachimura S, Takahashi Y, et al. Bacteroides
induce higher IgA production than Lactobacillus by increasing activation-induced cytidine
deaminase expression in B cells in murine Peyer’s patches. Biosci Biotechnol Biochem.
2009;73(2):372–7. Epub 2009/02/10.
59. Trinchieri G, Sher A. Cooperation of Toll-like receptor signals in innate immune defence. Nat
Rev Immunol. 2007;7(3):179–90. Epub 2007/02/24.
60. Rehman A, Sina C, Gavrilova O, Hasler R, Ott S, Baines JF, et al. Nod2 is essential for tem-
poral development of intestinal microbial communities. Gut. 2011;60(10):1354–62. Epub
2011/03/23.
61. Girardin SE, Boneca IG, Viala J, Chamaillard M, Labigne A, Thomas G, et al. Nod2 is a
general sensor of peptidoglycan through muramyl dipeptide (MDP) detection. J Biol Chem.
2003;278(11):8869–72. Epub 2003/01/16.
62. Petnicki-Ocwieja T, Hrncir T, Liu YJ, Biswas A, Hudcovic T, Tlaskalova-Hogenova H, et al.
Nod2 is required for the regulation of commensal microbiota in the intestine. Proc Natl Acad
Sci U S A. 2009;106(37):15813–8. Epub 2009/10/07.
63. Li E, Hamm CM, Gulati AS, Sartor RB, Chen H, Wu X, et al. Inflammatory bowel diseases
phenotype, C. difficile and NOD2 genotype are associated with shifts in human ileum associ-
ated microbial composition. PLoS ONE. 2012;7(6):e26284.
64. Frank DN, Robertson CE, Hamm CM, Kpadeh Z, Zhang T, Chen H, et al. Disease phenotype
and genotype are associated with shifts in intestinal-associated microbiota in inflammatory
bowel diseases. Inflamm Bowel Dis. 2011;17(1):179–84. Epub 2010/09/15.
65. Couturier-Maillard A, Secher T, Rehman A, Normand S, De Arcangelis A, Haesler R, et al.
NOD2-mediated dysbiosis predisposes mice to transmissible colitis and colorectal cancer. J
Clin Invest. 2013;3(4):384–7. Epub 2013/01/03.
66. Nones K, Knoch B, Dommels YE, Paturi G, Butts C, McNabb WC, et al. Multidrug resis-
tance gene deficient (mdr1a-/-) mice have an altered caecal microbiota that precedes the onset
of intestinal inflammation. J Appl Microbiol. 2009;107(2):557–66. Epub 2009/03/24.
67. Garrett WS, Lord GM, Punit S, Lugo-Villarino G, Mazmanian SK, Ito S, et al.
Communicable ulcerative colitis induced by T-bet deficiency in the innate immune sys-
tem. Cell. 2007;131(1):33–45. Epub 2007/10/10.
68. Gulati AS, Shanahan MT, Arthur JC, Grossniklaus E, von Furstenberg RJ, Kreuk L, et al.
Mouse background strain profoundly influences Paneth cell function and intestinal microbial
composition. PLoS ONE. 2012;7(2):e32403. Epub 2012/03/03.
69. Rawls JF, Mahowald MA, Ley RE, Gordon JI. Reciprocal gut microbiota transplants from
zebrafish and mice to germ-free recipients reveal host habitat selection. Cell.
2006;127(2):423–33. Epub 2006/10/24.
70. Fonager K, Sorensen HT, Olsen J. Change in incidence of Crohn’s disease and ulcerative
colitis in Denmark. A study based on the National Registry of Patients, 1981–1992. Int J
Epidemiol. 1997;26(5):1003–8.
71. Loftus CG, Loftus Jr EV, Harmsen WS, Zinsmeister AR, Tremaine WJ, Melton 3rd LJ,
et al. Update on the incidence and prevalence of Crohn’s disease and ulcerative colitis in
Olmsted County, Minnesota, 1940–2000. Inflamm Bowel Dis. 2007;13(3):254–61. Epub
2007/01/09.
72. Bernstein CN, Blanchard JF, Rawsthorne P, Wajda A. Epidemiology of Crohn’s disease and
ulcerative colitis in a central Canadian province: a population-based study. Am J Epidemiol.
1999;149(10):916–24. Epub 1999/05/26.
73. Loftus Jr EV, Silverstein MD, Sandborn WJ, Tremaine WJ, Harmsen WS, Zinsmeister
AR. Crohn’s disease in Olmsted County, Minnesota, 1940–1993: incidence, prevalence, and
survival. Gastroenterology. 1998;114(6):1161–8. Epub 1998/06/03.
74. Moum B, Vatn MH, Ekbom A, Aadland E, Fausa O, Lygren I, et al. Incidence of Crohn’s
disease in four counties in southeastern Norway, 1990–93. A prospective population-based
study. The Inflammatory Bowel South-Eastern Norway (IBSEN) Study Group of
Gastroenterologists. Scand J Gastroenterol. 1996;31(4):355–61.
75. Vind I, Riis L, Jess T, Knudsen E, Pedersen N, Elkjaer M, et al. Increasing incidences of
inflammatory bowel disease and decreasing surgery rates in Copenhagen City and County,
2003–2005: a population-based study from the Danish Crohn colitis database. Am J
Gastroenterol. 2006;101(6):1274–82. Epub 2006/06/15.
76. Manninen P, Karvonen AL, Huhtala H, Rasmussen M, Collin P. The epidemiology of
inflammatory bowel diseases in Finland. Scand J Gastroenterol. 2010;45(9):1063–7. Epub
2010/05/07.
77. Jussila A, Virta LJ, Salomaa V, Maki J, Jula A, Farkkila MA. High and increasing prevalence
of inflammatory bowel disease in Finland with a clear North-South difference. J Crohns
Colitis. 2012;42(3):482–9. Epub 2012/11/13.
78. Nerich V, Monnet E, Etienne A, Louafi S, Ramee C, Rican S, et al. Geographical variations
of inflammatory bowel disease in France: a study based on national health insurance data.
Inflamm Bowel Dis. 2006;12(3):218–26. Epub 2006/03/15.
79. Shivananda S, Lennard-Jones J, Logan R, Fear N, Price A, Carpenter L, et al. Incidence of
inflammatory bowel disease across Europe: is there a difference between north and south?
Results of the European Collaborative Study on Inflammatory Bowel Disease (EC-IBD).
Gut. 1996;39(5):690–7. Epub 1996/11/01.
80. Ahuja V, Tandon RK. Inflammatory bowel disease in the Asia-Pacific area: a comparison
with developed countries and regional differences. J Dig Dis. 2010;11(3):134–47. Epub
2010/06/29.
81. Thia KT, Loftus Jr EV, Sandborn WJ, Yang SK. An update on the epidemiology of inflamma-
tory bowel disease in Asia. Am J Gastroenterol. 2008;103(12):3167–82. Epub 2008/12/18.
82. Probert CS, Jayanthi V, Hughes AO, Thompson JR, Wicks AC, Mayberry JF. Prevalence
and family risk of ulcerative colitis and Crohn’s disease: an epidemiological study
among Europeans and south Asians in Leicestershire. Gut. 1993;34(11):1547–51. Epub
1993/11/01.
83. Cheon JH. Genetics of inflammatory bowel diseases: a comparison between Western and
Eastern perspectives. J Gastroenterol Hepatol. 2012;27:919–27. Epub 2012/11/30.
84. Halme L, Paavola-Sakki P, Turunen U, Lappalainen M, Farkkila M, Kontula K. Family and twin
studies in inflammatory bowel disease. World J Gastroenterol: WJG. 2006;12(23):3668–72.
Epub 2006/06/15.
85. Halfvarson J, Bodin L, Tysk C, Lindberg E, Jarnerot G. Inflammatory bowel disease in a

Swedish twin cohort: a long-term follow-up of concordance and clinical characteristics.
Gastroenterology. 2003;124(7):1767–73. Epub 2003/06/14.
86. Orholm M, Binder V, Sorensen TI, Rasmussen LP, Kyvik KO. Concordance of inflamma-
tory bowel disease among Danish twins. Results of a nationwide study. Scandinavian J
87. Thompson NP, Driscoll R, Pounder RE, Wakefield AJ. Genetics versus environment in
inflammatory bowel disease: results of a British twin study. BMJ. 1996;312(7023):95–6.
Epub 1996/01/13.
88. Tysk C, Lindberg E, Jarnerot G, Floderus-Myrhed B. Ulcerative colitis and Crohn’s disease
in an unselected population of monozygotic and dizygotic twins. A study of heritability and
the influence of smoking. Gut. 1988;29(7):990–6. Epub 1988/07/01.
89. Jostins L, Ripke S, Weersma RK, Duerr RH, McGovern DP, Hui KY, et al. Host-microbe
interactions have shaped the genetic architecture of inflammatory bowel disease. Nature.
2012;491(7422):119–24. Epub 2012/11/07.
90. Sartor RB. Mechanisms of disease: pathogenesis of Crohn’s disease and ulcerative colitis.
Nat Clin Pract Gastroenterol Hepatol. 2006;3(7):390–407. Epub 2006/07/05.
91. Clavel T, Haller D. Bacteria- and host-derived mechanisms to control intestinal epithelial cell
homeostasis: implications for chronic inflammation. Inflamm Bowel Dis. 2007;13(9):1153–64.
Epub 2007/05/04.
92. D’Haens GR, Geboes K, Peeters M, Baert F, Penninckx F, Rutgeerts P. Early lesions of
recurrent Crohn’s disease caused by infusion of intestinal contents in excluded ileum.
93. Duchmann R, Kaiser I, Hermann E, Mayet W, Ewe K. Meyer zum Buschenfelde KH. Tolerance
exists towards resident intestinal flora but is broken in active inflammatory bowel disease
(IBD). Clin Exp Immunol. 1995;102(3):448–55.
94. Macpherson A, Khoo UY, Forgacs I, Philpott-Howard J, Bjarnason I. Mucosal antibodies in
inflammatory bowel disease are directed against intestinal bacteria. Gut. 1996;38(3):365–75.
Epub 1996/03/01.
95. van der Waaij LA, Kroese FG, Visser A, Nelis GF, Westerveld BD, Jansen PL, et al.
Immunoglobulin coating of faecal bacteria in inflammatory bowel disease. Eur J Gastroenterol
Hepatol. 2004;16(7):669–74. Epub 2004/06/18.
96. Greenbloom SL, Steinhart AH, Greenberg GR. Combination ciprofloxacin and metronidazole
for active Crohn’s disease. Can J Gastroenterol = J canadien de gastroenterologie.
1998;12(1):53–6.
97. Rutgeerts P, Van Assche G, Vermeire S, D’Haens G, Baert F, Noman M, et al. Ornidazole for
prophylaxis of postoperative Crohn’s disease recurrence: a randomized, double-blind,
placebo-controlled trial. Gastroenterology. 2005;128(4):856–61. Epub 2005/04/13.
98. Arnold GL, Beaves MR, Pryjdun VO, Mook WJ. Preliminary study of ciprofloxacin in active
Crohn’s disease. Inflamm Bowel Dis. 2002;8(1):10–5. Epub 2002/02/12.
99. Yang SK, Jung Y, Hong M, Kim H, Ye BD, Lee I, et al. No association between TNFSF15 and
IL23R with ulcerative colitis in Koreans. J Hum Genet. 2011;56(3):200–4. Epub 2011/01/14.
100. Silverberg MS, Duerr RH, Brant SR, Bromfield G, Datta LW, Jani N, et al. Refined genomic
localization and ethnic differences observed for the IBD5 association with Crohn’s disease.
EJHG. 2007;15(3):328–35. Epub 2007/01/11.
101. Picornell Y, Mei L, Taylor K, Yang H, Targan SR, Rotter JI. TNFSF15 is an ethnic-specific
IBD gene. Inflamm Bowel Dis. 2007;13(11):1333–8. Epub 2007/08/01.
102. Kenny EE, Pe’er I, Karban A, Ozelius L, Mitchell AA, Ng SM, et al. A genome-wide scan of
Ashkenazi Jewish Crohn’s disease suggests novel susceptibility loci. PLoS Genet.
2012;8(3):e1002559. Epub 2012/03/14.
103. Lees CW, Barrett JC, Parkes M, Satsangi J. New IBD genetics: common pathways with other
diseases. Gut. 2011;60(12):1739–53. Epub 2011/02/09.
104. Muise AM, Walters TD, Glowacka WK, Griffiths AM, Ngan BY, Lan H, et al. Polymorphisms
in E-cadherin (CDH1) result in a mis-localised cytoplasmic protein that is associated with
Crohn’s disease. Gut. 2009;58(8):1121–7. Epub 2009/04/29.
105. Barrett JC, Hansoul S, Nicolae DL, Cho JH, Duerr RH, Rioux JD, et al. Genome-wide asso-
ciation defines more than 30 distinct susceptibility loci for Crohn’s disease. Nat Genet.
2008;40(8):955–62. Epub 2008/07/01.
106. Rivas MA, Beaudoin M, Gardet A, Stevens C, Sharma Y, Zhang CK, et al. Deep resequencing
of GWAS loci identifies independent rare variants associated with inflammatory bowel dis-
ease. Nat Genet. 2011;43(11):1066–73. Epub 2011/10/11.
107. Umeno J, Asano K, Matsushita T, Matsumoto T, Kiyohara Y, Iida M, et al. Meta-
analysis of published studies identified eight additional common susceptibility loci for
Crohn’s disease and ulcerative colitis. Inflamm Bowel Dis. 2011;17(12):2407–15. Epub
2011/02/26.
108. Kaser A, Lee AH, Franke A, Glickman JN, Zeissig S, Tilg H, et al. XBP1 links ER stress to
intestinal inflammation and confers genetic risk for human inflammatory bowel disease. Cell.
2008;134(5):743–56. Epub 2008/09/09.
109. Spenle C, Hussenet T, Lacroute J, Lefebvre O, Kedinger M, Orend G, et al. Dysregulation of
laminins in intestinal inflammation. Pathol Biol. 2012;60(1):41–7. Epub 2011/11/22.
110. Schmehl K, Florian S, Jacobasch G, Salomon A, Korber J. Deficiency of epithelial basement
membrane laminin in ulcerative colitis affected human colonic mucosa. Int J Colorectal Dis.
2000;15(1):39–48. Epub 2000/04/15.
111. Bertin J, Wang L, Guo Y, Jacobson MD, Poyet JL, Srinivasula SM, et al. CARD11 and
CARD14 are novel caspase recruitment domain (CARD)/membrane-associated guanylate
kinase (MAGUK) family members that interact with BCL10 and activate NF-kappa B. J Biol
Chem. 2001;276(15):11877–82. Epub 2001/03/30.
112. Hugot JP, Chamaillard M, Zouali H, Lesage S, Cezard JP, Belaiche J, et al. Association
of NOD2 leucine-rich repeat variants with susceptibility to Crohn’s disease. Nature.
2001;411(6837):599–603. Epub 2001/06/01.
113. Lesage S, Zouali H, Cezard JP, Colombel JF, Belaiche J, Almer S, et al. CARD15/NOD2
mutational analysis and genotype-phenotype correlation in 612 patients with inflammatory
bowel disease. Am J Hum Genet. 2002;70(4):845–57. Epub 2002/03/05.
114. Consortium TWTCC. Genome-wide association study of 14,000 cases of seven common
diseases and 3,000 shared controls. Nature. 2007;447(7145):661–78. Epub 2007/06/08.
115. Wehkamp J, Harder J, Weichenthal M, Schwab M, Schaffeler E, Schlee M, et al. NOD2
(CARD15) mutations in Crohn’s disease are associated with diminished mucosal alpha-
defensin expression. Gut. 2004;53(11):1658–64. Epub 2004/10/14.
116. Kobayashi KS, Chamaillard M, Ogura Y, Henegariu O, Inohara N, Nunez G, et al. Nod2-
dependent regulation of innate and adaptive immunity in the intestinal tract. Science.
2005;307(5710):731–4. Epub 2005/02/05.
117. Biswas A, Liu YJ, Hao L, Mizoguchi A, Salzman NH, Bevins CL, et al. Induction and rescue
of Nod2-dependent Th1-driven granulomatous inflammation of the ileum. Proc Natl Acad
Sci U S A. 2010;107(33):14739–44. Epub 2010/08/04.
118. Saitoh T, Fujita N, Jang MH, Uematsu S, Yang BG, Satoh T, et al. Loss of the autophagy pro-
tein Atg16L1 enhances endotoxin-induced IL-1beta production. Nature. 2008;456(7219):264–
8. Epub 2008/10/14.
119. Travassos LH, Carneiro LA, Ramjeet M, Hussey S, Kim YG, Magalhaes JG, et al. Nod1 and
Nod2 direct autophagy by recruiting ATG16L1 to the plasma membrane at the site of bacterial
entry. Nat Immunol. 2010;11(1):55–62. Epub 2009/11/10.
120. Wehkamp J, Salzman NH, Porter E, Nuding S, Weichenthal M, Petras RE, et al. Reduced
Paneth cell alpha-defensins in ileal Crohn’s disease. Proc Natl Acad Sci U S A.
2005;102(50):18129–34. Epub 2005/12/07.
121. Homer CR, Richmond AL, Rebert NA, Achkar JP, McDonald C. ATG16L1 and NOD2 inter-
act in an autophagy-dependent antibacterial pathway implicated in Crohn’s disease pathogen-
esis. Gastroenterology. 2010;139(5):1630–41. 41 e1-2. Epub 2010/07/20.
122. Glas J, Wagner J, Seiderer J, Olszak T, Wetzke M, Beigel F, et al. PTPN2 gene variants are
associated with susceptibility to both Crohn’s disease and ulcerative colitis supporting a
common genetic disease background. PLoS ONE. 2012;7(3):e33682. Epub 2012/03/30.
123. Scharl M, Mwinyi J, Fischbeck A, Leucht K, Eloranta JJ, Arikkat J, et al. Crohn’s disease-
associated polymorphism within the PTPN2 gene affects muramyl-dipeptide-induced cyto-
kine secretion and autophagy. Inflamm Bowel Dis. 2012;18(5):900–12. Epub 2011/10/25.
124. Parkes M, Barrett JC, Prescott NJ, Tremelling M, Anderson CA, Fisher SA, et al. Sequence
variants in the autophagy gene IRGM and multiple other replicating loci contribute to Crohn’s
disease susceptibility. Nat Genet. 2007;39(7):830–2. Epub 2007/06/08.
125. McGovern DP, Gardet A, Torkvist L, Goyette P, Essers J, Taylor KD, et al. Genome-wide asso-
ciation identifies multiple ulcerative colitis susceptibility loci. Nat Genet. 2010;42(4):332–7.
Epub 2010/03/17.
126. MacMicking JD, Taylor GA, McKinney JD. Immune control of tuberculosis by IFN-gamma-
inducible LRG-47. Science. 2003;302(5645):654–9. Epub 2003/10/25.
127. Wiede F, Shields BJ, Chew SH, Kyparissoudis K, van Vliet C, Galic S, et al. T cell protein
tyrosine phosphatase attenuates T cell signaling to maintain tolerance in mice. J Clin Invest.
2011;121(12):4758–74. Epub 2011/11/15.
128. You-Ten KE, Muise ES, Itie A, Michaliszyn E, Wagner J, Jothy S, et al. Impaired bone mar-
row microenvironment and immune function in T cell protein tyrosine phosphatase-deficient
mice. J Exp Med. 1997;186(5):683–93. Epub 1997/08/29.
129. Hassan SW, Doody KM, Hardy S, Uetani N, Cournoyer D, Tremblay ML. Increased suscep-
tibility to dextran sulfate sodium induced colitis in the T cell protein tyrosine phosphatase
heterozygous mouse. PLoS ONE. 2010;5(1):e8868. Epub 2010/01/30.
130. Murai M, Turovskaya O, Kim G, Madan R, Karp CL, Cheroutre H, et al. Interleukin 10 acts
on regulatory T cells to maintain expression of the transcription factor Foxp3 and suppressive
function in mice with colitis. Nat Immunol. 2009;10(11):1178–84. Epub 2009/09/29.
131. Kuhn R, Lohler J, Rennick D, Rajewsky K, Muller W. Interleukin-10-deficient mice develop
chronic enterocolitis. Cell. 1993;75(2):263–74. Epub 1993/10/22.
132. Sellon RK, Tonkonogy S, Schultz M, Dieleman LA, Grenther W, Balish E, et al. Resident
enteric bacteria are necessary for development of spontaneous colitis and immune system acti-
vation in interleukin-10-deficient mice. Infect Immun. 1998;66(11):5224–31. Epub 1998/10/24.
133. Duerr RH, Taylor KD, Brant SR, Rioux JD, Silverberg MS, Daly MJ, et al. A genome-wide
association study identifies IL23R as an inflammatory bowel disease gene. Science.
2006;314(5804):1461–3. Epub 2006/10/28.
134. Oppmann B, Lesley R, Blom B, Timans JC, Xu Y, Hunte B, et al. Novel p19 protein engages
IL-12p40 to form a cytokine, IL-23, with biological activities similar as well as distinct from
IL-12. Immunity. 2000;13(5):715–25. Epub 2000/12/15.
135. Barnich N, Darfeuille-Michaud A. Abnormal CEACAM6 expression in Crohn disease
patients favors gut colonization and inflammation by adherent-invasive E. coli. Virulence.
2010;1(4):281–2. Epub 2010/12/24.
136. Carvalho FA, Barnich N, Sivignon A, Darcha C, Chan CH, Stanners CP, et al. Crohn’s
disease adherent-invasive Escherichia coli colonize and induce strong gut inflammation in
transgenic mice expressing human CEACAM. J Exp Med. 2009;206(10):2179–89. Epub
2009/09/10.
137. Swidsinski A, Ladhoff A, Pernthaler A, Swidsinski S, Loening-Baucke V, Ortner M, et al.
Mucosal flora in inflammatory bowel disease. Gastroenterology. 2002;122(1):44–54. Epub
2002/01/10.
138. Walker AW, Sanderson JD, Churcher C, Parkes GC, Hudspith BN, Rayment N, et al. High-
throughput clone library analysis of the mucosa-associated microbiota reveals dysbiosis and
differences between inflamed and non-inflamed regions of the intestine in inflammatory
bowel disease. BMC Microbiol. 2011;11:7. Epub 2011/01/12.
139. Manichanh C, Rigottier-Gois L, Bonnaud E, Gloux K, Pelletier E, Frangeul L, et al. Reduced
diversity of faecal microbiota in Crohn’s disease revealed by a metagenomic approach. Gut.
2006;55(2):205–11. Epub 2005/09/29.
140. Ott SJ, Musfeldt M, Wenderoth DF, Hampe J, Brant O, Folsch UR, et al. Reduction in diver-
sity of the colonic mucosa associated bacterial microflora in patients with active inflamma-
tory bowel disease. Gut. 2004;53(5):685–93. Epub 2004/04/15.
141. Willing BP, Dicksved J, Halfvarson J, Andersson AF, Lucio M, Zheng Z, et al. A pyrose-
quencing study in twins shows that gastrointestinal microbial profiles vary with inflammatory
bowel disease phenotypes. Gastroenterology. 2010;139(6):1844–54. e1. Epub 2010/09/08.
142. Frank DN, St Amand AL, Feldman RA, Boedeker EC, Harpaz N, Pace NR. Molecular-
phylogenetic characterization of microbial community imbalances in human inflammatory
bowel diseases. Proc Natl Acad Sci U S A. 2007;104(34):13780–5. Epub 2007/08/19.
143. Rehman A, Lepage P, Nolte A, Hellmig S, Schreiber S, Ott SJ. Transcriptional activity of the
dominant gut mucosal microbiota in chronic inflammatory bowel disease patients. J Med
Microbiol. 2010;59(Pt 9):1114–22. Epub 2010/06/05.
144. Lepage P, Hasler R, Spehlmann ME, Rehman A, Zvirbliene A, Begun A, et al. Twin study
indicates loss of interaction between microbiota and mucosa of patients with ulcerative coli-
tis. Gastroenterology. 2011;141(1):227–36. Epub 2011/05/31.
145. Willing B, Halfvarson J, Dicksved J, Rosenquist M, Jarnerot G, Engstrand L, et al. Twin stud-
ies reveal specific imbalances in the mucosa-associated microbiota of patients with ileal
Crohn’s disease. Inflamm Bowel Dis. 2009;15(5):653–60. Epub 2008/11/22.
146. Gophna U, Sommerfeld K, Gophna S, Doolittle WF, Veldhuyzen, van Zanten SJ. Differences
between tissue-associated intestinal microfloras of patients with Crohn’s disease and ulcer-
ative colitis. J Clin Micro. 2006;44(11):4136–41. Epub 2006/09/22.
147. Baumgart M, Dogan B, Rishniw M, Weitzman G, Bosworth B, Yantiss R, et al. Culture inde-
pendent analysis of ileal mucosa reveals a selective increase in invasive Escherichia coli of
novel phylogeny relative to depletion of Clostridiales in Crohn’s disease involving the ileum.
ISME J. 2007;1(5):403–18. Epub 2007/11/29.
148. Mylonaki M, Rayment NB, Rampton DS, Hudspith BN, Brostoff J. Molecular characteriza-
tion of rectal mucosa-associated bacterial flora in inflammatory bowel disease. Inflamm
Bowel Dis. 2005;11(5):481–7. Epub 2005/05/04.
149. Fujimoto T, Imaeda H, Takahashi K, Kasumi E, Bamba S, Fujiyama Y, et al. Decreased abun-
dance of Faecalibacterium prausnitzii in the gut microbiota of Crohn’s disease. J Gastroenterol
Hepatol. 2012. Epub 2012/12/12.
150. Sokol H, Seksik P, Furet JP, Firmesse O, Nion-Larmurier I, Beaugerie L, et al. Low counts of
Faecalibacterium prausnitzii in colitis microbiota. Inflamm Bowel Dis. 2009;15(8):1183–9.
Epub 2009/02/25.
151. Sokol H, Pigneur B, Watterlot L, Lakhdari O, Bermudez-Humaran LG, Gratadoux JJ, et al.
Faecalibacterium prausnitzii is an anti-inflammatory commensal bacterium identified by gut
microbiota analysis of Crohn disease patients. Proc Natl Acad Sci U S A. 2008;105(43):16731–
6. Epub 2008/10/22.
152. Martinez-Medina M, Aldeguer X, Gonzalez-Huix F, Acero D, Garcia-Gil LJ. Abnormal
microbiota composition in the ileocolonic mucosa of Crohn’s disease patients as revealed by
polymerase chain reaction-denaturing gradient gel electrophoresis. Inflamm Bowel Dis.
2006;12(12):1136–45. Epub 2006/11/23.
153. Macfarlane S, Furrie E, Cummings JH, Macfarlane GT. Chemotaxonomic analysis of bacte-
rial populations colonizing the rectal mucosa in patients with ulcerative colitis. Clin Infect
Dis: Off Pub Infect Dis Soc Am. 2004;38(12):1690–9. Epub 2004/07/01.
154. Jalanka-Tuovinen J, Salonen A, Nikkila J, Immonen O, Kekkonen R, Lahti L, et al. Intestinal
microbiota in healthy adults: temporal analysis reveals individual and common core and rela-
tion to intestinal symptoms. PLoS ONE. 2011;6(7):e23035. Epub 2011/08/11.
155. Loubinoux J, Bronowicki JP, Pereira IA, Mougenel JL, Faou AE. Sulfate-reducing bacteria in
human feces and their association with inflammatory bowel diseases. FEMS Microbiol Ecol.
2002;40(2):107–12. Epub 2002/05/01.
156. Gibson GR, Macfarlane GT, Cummings JH. Sulphate reducing bacteria and hydrogen metab-
olism in the human large intestine. Gut. 1993;34(4):437–9. Epub 1993/04/01.
157. Pitcher MC, Beatty ER, Cummings JH. The contribution of sulphate reducing bacteria and
5-aminosalicylic acid to faecal sulphide in patients with ulcerative colitis. Gut.
2000;46(1):64–72. Epub 1999/12/22.
158. Muyzer G, Stams AJ. The ecology and biotechnology of sulphate-reducing bacteria. Nat Rev
Microbiol. 2008;6(6):441–54. Epub 2008/05/08.
159. Roediger WE, Duncan A, Kapaniris O, Millard S. Reducing sulfur compounds of the
colon impair colonocyte nutrition: implications for ulcerative colitis. Gastroenterology.
1993;104(3):802–9. Epub 1993/03/01.
160. Wu YC, Wang XJ, Yu L, Chan FK, Cheng AS, Yu J, et al. Hydrogen sulfide lowers prolifera-
tion and induces protective autophagy in colon epithelial cells. PLoS ONE. 2012;7(5):e37572.
Epub 2012/06/09.
161. Laue H, Denger K, Cook AM. Taurine reduction in anaerobic respiration of Bilophila wad-
sworthia RZATAU. Appl Environ Microbiol. 1997;63(5):2016–21. Epub 1997/05/01.
162. Devkota S, Wang Y, Musch MW, Leone V, Fehlner-Peach H, Nadimpalli A, et al. Dietary-fat-
induced taurocholic acid promotes pathobiont expansion and colitis in Il10-/- mice. Nature.
2012;487(7405):104–8. Epub 2012/06/23.
163. Ishioka T, Kuwabara N, Oohashi Y, Wakabayashi K. Induction of colorectal tumors in rats by
sulfated polysaccharides. Crit Rev Toxicol. 1987;17(3):215–44. Epub 1987/01/01.
164. Ohkusa T. Production of experimental ulcerative colitis in hamsters by dextran sulfate sodium
and changes in intestinal microflora. Nihon Shokakibyo Gakkai zasshi: Japan J Gastroenterol.
1985;82(5):1327–36. Epub 1985/05/01.
165. Fava F, Danese S. Intestinal microbiota in inflammatory bowel disease: friend of foe? World
J Gastroenterol: WJG. 2011;17(5):557–66. Epub 2011/02/26.
166. Nagalingam NA, Kao JY, Young VB. Microbial ecology of the murine gut associated with the
development of dextran sodium sulfate-induced colitis. Inflamm Bowel Dis. 2011;17(4):917–
26. Epub 2011/03/11.
167. Araki Y, Mukaisho K, Sugihara H, Fujiyama Y, Hattori T. Proteus mirabilis sp intestinal micro-
flora grow in a dextran sulfate sodium-rich environment. Int J Mol Med. 2010;25(2):203–8.
Epub 2010/01/01.
168. Bloom SM, Bijanki VN, Nava GM, Sun L, Malvin NP, Donermeyer DL, et al. Commensal
Bacteroides species induce colitis in host-genotype-specific fashion in a mouse model of
inflammatory bowel disease. Cell Host Microbe. 2011;9(5):390–403. Epub 2011/05/18.
169. Morgan XC, Tickle TL, Sokol H, Gevers D, Devaney KL, Ward DV, et al. Dysfunction of the
intestinal microbiome in inflammatory bowel disease and treatment. Genome Biol.
2012;13(9):R79. Epub 2012/09/28.
170. Stephens NS, Siffledeen J, Su X, Murdoch TB, Fedorak RN, Slupsky CM. Urinary NMR
metabolomic profiles discriminate inflammatory bowel disease from healthy. J Crohns
Colitis. 2013;7(2):e42–8. Epub 2012/05/26.
171. Lupp C, Robertson ML, Wickham ME, Sekirov I, Champion OL, Gaynor EC, et al. Host-
mediated inflammation disrupts the intestinal microbiota and promotes the overgrowth of
Enterobacteriaceae. Cell Host Microbe. 2007;2(2):119–29. Epub 2007/11/17.
172. Stecher B, Robbiani R, Walker AW, Westendorf AM, Barthel M, Kremer M, et al. Salmonella
enterica serovar typhimurium exploits inflammation to compete with the intestinal microbi-
ota. PLoS Biol. 2007;5(10):2177–89. Epub 2007/09/01.
173. Papa E, Docktor M, Smillie C, Weber S, Preheim SP, Gevers D, et al. Non-invasive mapping
of the gastrointestinal microbiota identifies children with inflammatory bowel disease. PLoS
ONE. 2012;7(6):e39242. Epub 2012/07/07.
174. Sokol H, Lepage P, Seksik P, Dore J, Marteau P. Temperature gradient gel electrophoresis of
fecal 16S rRNA reveals active Escherichia coli in the microbiota of patients with ulcerative
colitis. J Clin Microbiol. 2006;44(9):3172–7. Epub 2006/09/07.
Epub 2010/03/06.
176. Greenblum S, Turnbaugh PJ, Borenstein E. Metagenomic systems biology of the human gut
microbiome reveals topological shifts associated with obesity and inflammatory bowel dis-
ease. Proc Natl Acad Sci U S A. 2012;109(2):594–9. Epub 2011/12/21.
177. Hansen JJ, Huang Y, Peterson DA, Goeser L, Fan TJ, Chang EB, et al. The colitis-associated
transcriptional profile of commensal Bacteroides thetaiotaomicron enhances adaptive
immune responses to a bacterial antigen. PLoS ONE. 2012;7(8):e42645. Epub 2012/08/11.
178. Patwa LG, Fan TJ, Tchaptchet S, Liu Y, Lussier YA, Sartor RB, et al. Chronic intestinal
inflammation induces stress-response genes in commensal Escherichia coli. Gastroenterology.
2011;141(5):1842–51. e1-10. Epub 2011/07/06.
179. Verberkmoes NC, Russell AL, Shah M, Godzik A, Rosenquist M, Halfvarson J, et al. Shotgun
metaproteomics of the human distal gut microbiota. ISME J. 2009;3(2):179–89. Epub
2008/10/31.
180. Kolmeder CA, de Been M, Nikkila J, Ritamo I, Matto J, Valmu L, et al. Comparative meta-
proteomics and diversity analysis of human intestinal microbiota testifies for its temporal
stability and expression of core functions. PLoS ONE. 2012;7(1):e29913. Epub 2012/01/27.
181. Erickson AR, Cantarel BL, Lamendella R, Darzi Y, Mongodin EF, Pan C, et al. Integrated
metagenomics/metaproteomics reveals human host-microbiota signatures of Crohn’s disease.
PLoS ONE. 2012;7(11):e49138. Epub 2012/12/05.
182. Jansson J, Willing B, Lucio M, Fekete A, Dicksved J, Halfvarson J, et al. Metabolomics reveals
metabolic biomarkers of Crohn’s disease. PLoS ONE. 2009;4(7):e6386. Epub 2009/07/29.
183. Marchesi JR, Holmes E, Khan F, Kochhar S, Scanlan P, Shanahan F, et al. Rapid and non-
invasive metabonomic characterization of inflammatory bowel disease. J Proteome Res.
2007;6(2):546–51. Epub 2007/02/03.
184. Duboc H, Rainteau D, Rajca S, Humbert L, Farabos D, Maubert M, et al. Increase in fecal
primary bile acids and dysbiosis in patients with diarrhea-predominant irritable bowel syn-
drome. Neurogastroenterol Motil: Off J Eur Gastrointestin Motility Soc. 2012;24(6):513–20.
e246-7. Epub 2012/02/24.
185. Wohlgemuth S, Keller S, Kertscher R, Stadion M, Haller D, Kisling S, et al. Intestinal steroid
profiles and microbiota composition in colitic mice. Gut Microbes. 2011;2(3):159–66. Epub
2011/08/27.
186. Islam KB, Fukiya S, Hagio M, Fujii N, Ishizuka S, Ooka T, et al. Bile acid is a host factor
that regulates the composition of the cecal microbiota in rats. Gastroenterology.
2011;141(5):1773–81. Epub 2011/08/16.
187. Duboc H, Rajca S, Rainteau D, Benarous D, Maubert MA, Quervain E, et al. Connecting
dysbiosis, bile-acid dysmetabolism and gut inflammation in inflammatory bowel diseases.
Gut. 2012;40(18):9308–18. Epub 2012/09/21.
188. Stenman LK, Holma R, Korpela R. High-fat-induced intestinal permeability dysfunction
associated with altered fecal bile acids. World J Gastroenterol: WJG. 2012;18(9):923–9.
Epub 2012/03/13.
189. Suzuki T, Hara H. Dietary fat and bile juice, but not obesity, are responsible for the increase
in small intestinal permeability induced through the suppression of tight junction protein
expression in LETO and OLETF rats. Nutr Metab. 2010;7:19. Epub 2010/03/13.
190. Raimondi F, Santoro P, Barone MV, Pappacoda S, Barretta ML, Nanayakkara M, et al. Bile
acids modulate tight junction structure and barrier function of Caco-2 monolayers via EGFR
activation. Am J Physiol Gastrointest Liver Physiol. 2008;294(4):G906–13. Epub 2008/02/02.
191. Gadaleta RM, van Erpecum KJ, Oldenburg B, Willemsen EC, Renooij W, Murzilli S, et al.
Farnesoid X receptor activation inhibits inflammation and preserves the intestinal barrier in
inflammatory bowel disease. Gut. 2011;60(4):463–72. Epub 2011/01/19.
192. Takaishi H, Matsuki T, Nakazawa A, Takada T, Kado S, Asahara T, et al. Imbalance in
intestinal microflora constitution could be involved in the pathogenesis of inflammatory
bowel disease. IJMM. 2008;298(5–6):463–72. Epub 2007/09/28.
193. Baur P, Martin FP, Gruber L, Bosco N, Brahmbhatt V, Collino S, et al. Metabolic phenotyping
of the Crohn’s disease-like IBD etiopathology in the TNF(DeltaARE/WT) mouse model. J
Proteome Res. 2011;10(12):5523–35. Epub 2011/10/28.
194. Fernandez-Banares F, Esteve-Comas M, Mane J, Navarro E, Bertran X, Cabre E, et al. Changes

in mucosal fatty acid profile in inflammatorybowel disease and in experimental colitis: a com-
mon response to bowel inflammation. Clin Nutr. 1997;16(4):177–83. Epub 1997/08/01.
195. Bjerrum JT, Nielsen OH, Hao F, Tang H, Nicholson JK, Wang Y, et al. Metabonomics in
ulcerative colitis: diagnostics, biomarker identification, and insight into the pathophysiology.
J Proteome Res. 2010;9(2):954–62. Epub 2009/10/29.
196. Bezabeh T, Somorjai RL, Smith IC, Nikulin AE, Dolenko B, Bernstein CN. The use of
1H magnetic resonance spectroscopy in inflammatory bowel diseases: distinguishing
ulcerative colitis from Crohn’s disease. Am J Gastroenterol. 2001;96(2):442–8. Epub
2001/03/10.
197. Balasubramanian K, Kumar S, Singh RR, Sharma U, Ahuja V, Makharia GK, et al. Metabolism
of the colonic mucosa in patients with inflammatory bowel diseases: an in vitro proton
magnetic resonance spectroscopy study. Magn Reson Imaging. 2009;27(1):79–86. Epub
2008/07/05.
198. Holub BJ. The nutritional significance, metabolism, and function of myo-inositol and phos-
phatidylinositol in health and disease. Adv Nutr Res. 1982;4:107–41. Epub 1982/01/01.
199. Holub BJ. Metabolism and function of myo-inositol and inositol phospholipids. Annu Rev
Nutr. 1986;6:563–97. Epub 1986/01/01.
200. Martin FP, Rezzi S, Philippe D, Tornier L, Messlik A, Holzlwimmer G, et al. Metabolic
assessment of gradual development of moderate experimental colitis in IL-10 deficient mice.
J Proteome Res. 2009;8(5):2376–87. Epub 2009/03/28.
201. Murdoch TB, Fu H, MacFarlane S, Sydora BC, Fedorak RN, Slupsky CM. Urinary metabolic
profiles of inflammatory bowel disease in interleukin-10 gene-deficient mice. Anal Chem.
2008;80(14):5524–31. Epub 2008/06/19.
202. Williams HR, Cox IJ, Walker DG, North BV, Patel VM, Marshall SE, et al. Characterization
of inflammatory bowel disease with urinary metabolic profiling. Am J Gastroenterol.
2009;104(6):1435–44. Epub 2009/06/06.
203. Winterkamp S, Weidenhiller M, Otte P, Stolper J, Schwab D, Hahn EG, et al. Urinary excre-
tion of N-methylhistamine as a marker of disease activity in inflammatory bowel disease. Am
J Gastroenterol. 2002;97(12):3071–7. Epub 2002/12/21.
204. Macfarlane S, Macfarlane GT, Cummings JH. Review article: prebiotics in the gastrointesti-
nal tract. Aliment Pharmacol Ther. 2006;24(5):701–14. Epub 2006/08/22.
205. Langlands SJ, Hopkins MJ, Coleman N, Cummings JH. Prebiotic carbohydrates modify the
mucosa associated microflora of the human large bowel. Gut. 2004;53(11):1610–6. Epub
2004/10/14.
206. Gibson GR, Beatty ER, Wang X, Cummings JH. Selective stimulation of bifidobacteria in the
human colon by oligofructose and inulin. Gastroenterology. 1995;108(4):975–82. Epub
1995/04/01.
207. Gibson GR, Probert HM, Loo JV, Rastall RA, Roberfroid MB. Dietary modulation of the
human colonic microbiota: updating the concept of prebiotics. Nutr Res Rev. 2004;17(2):259–
75. Epub 2004/12/01.
208. Vulevic J, Drakoularakou A, Yaqoob P, Tzortzis G, Gibson GR. Modulation of the fecal
microflora profile and immune function by a novel trans-galactooligosaccharide mixture
(B-GOS) in healthy elderly volunteers. Am J Clin Nutr. 2008;88(5):1438–46. Epub
2008/11/11.
209. Depeint F, Tzortzis G, Vulevic J, I’Anson K, Gibson GR. Prebiotic evaluation of a novel
galactooligosaccharide mixture produced by the enzymatic activity of Bifidobacterium bifi-
dum NCIMB 41171, in healthy humans: a randomized, double-blind, crossover, placebo-
controlled intervention study. Am J Clin Nutr. 2008;87(3):785–91. Epub 2008/03/11.
210. Hormannsperger G, Clavel T, Haller D. Gut matters: Microbe-host interactions in allergic
diseases. J Allergy Clin Immunol. 2012;129(6):1452–9. Epub 2012/02/11.
211. Lindsay JO, Whelan K, Stagg AJ, Gobin P, Al-Hassi HO, Rayment N, et al. Clinical, micro-
biological, and immunological effects of fructo-oligosaccharide in patients with Crohn’s dis-
ease. Gut. 2006;55(3):348–55. Epub 2005/09/16.
212. Welters CF, Heineman E, Thunnissen FB, van den Bogaard AE, Soeters PB, Baeten
CG. Effect of dietary inulin supplementation on inflammation of pouch mucosa in
patients with an ileal pouch-anal anastomosis. Dis Colon Rectum. 2002;45(5):621–7.
Epub 2002/05/11.
213. Lara-Villoslada F, de Haro O, Camuesco D, Comalada M, Velasco J, Zarzuelo A, et al.
Short-chain fructooligosaccharides, in spite of being fermented in the upper part of the
large intestine, have anti-inflammatory activity in the TNBS model of colitis. Eur J Nutr.
2006;45(7):418–25. Epub 2006/07/28.
214. Cherbut C, Michel C, Lecannu G. The prebiotic characteristics of fructooligosaccharides are
necessary for reduction of TNBS-induced colitis in rats. J Nutr. 2003;133(1):21–7. Epub
2003/01/07.
215. Camuesco D, Peran L, Comalada M, Nieto A, Di Stasi LC, Rodriguez-Cabezas ME, et al.
Preventative effects of lactulose in the trinitrobenzenesulphonic acid model of rat colitis.
Inflamm Bowel Dis. 2005;11(3):265–71. Epub 2005/03/01.
216. Rumi G, Tsubouchi R, Okayama M, Kato S, Mozsik G, Takeuchi K. Protective effect of
lactulose on dextran sulfate sodium-induced colonic inflammation in rats. Dig Dis Sci.
2004;49(9):1466–72. Epub 2004/10/16.
217. Hoentjen F, Welling GW, Harmsen HJ, Zhang X, Snart J, Tannock GW, et al. Reduction of
colitis by prebiotics in HLA-B27 transgenic rats is associated with microflora changes and
immunomodulation. Inflamm Bowel Dis. 2005;11(11):977–85. Epub 2005/10/22.
218. Gopalakrishnan A, Clinthorne JF, Rondini EA, McCaskey SJ, Gurzell EA, Langohr IM, et al.
Supplementation with galacto-oligosaccharides increases the percentage of NK cells and reduces
colitis severity in Smad3-deficient mice. J Nutr. 2012;142(7):1336–42. Epub 2012/04/13.
219. Sang LX, Chang B, Zhang WL, Wu XM, Li XH, Jiang M. Remission induction and mainte-
nance effect of probiotics on ulcerative colitis: a meta-analysis. World J Gastroenterol: WJG.
2010;16(15):1908–15. Epub 2010/04/17.
220. Mack DR, Ahrne S, Hyde L, Wei S, Hollingsworth MA. Extracellular MUC3 mucin secretion
follows adherence of Lactobacillus strains to intestinal epithelial cells in vitro. Gut.
2003;52(6):827–33. Epub 2003/05/13.
221. Matsumoto S, Hara T, Hori T, Mitsuyama K, Nagaoka M, Tomiyasu N, et al. Probiotic
Lactobacillus-induced improvement in murine chronic inflammatory bowel disease is associ-
ated with the down-regulation of pro-inflammatory cytokines in lamina propria mononuclear
cells. Clin Exp Immunol. 2005;140(3):417–26. Epub 2005/06/04.
222. Liu Z, Zhang P, Ma Y, Chen H, Zhou Y, Zhang M, et al. Lactobacillus plantarum prevents the
development of colitis in IL-10-deficient mouse by reducing the intestinal permeability. Mol
Biol Rep. 2011;38(2):1353–61. Epub 2010/06/24.
223. Peluso I, Fina D, Caruso R, Stolfi C, Caprioli F, Fantini MC, et al. Lactobacillus paraca-
sei subsp. paracasei B21060 suppresses human T-cell proliferation. Infect Immun.
2007;75(4):1730–7. Epub 2007/01/24.
224. Oliva S, Di Nardo G, Ferrari F, Mallardo S, Rossi P, Patrizi G, et al. Randomised clinical trial:
the effectiveness of Lactobacillus reuteri ATCC 55730 rectal enema in children with active
distal ulcerative colitis. Aliment Pharmacol Ther. 2012;35(3):327–34. Epub 2011/12/14.
225. Zocco MA, Dal Verme LZ, Cremonini F, Piscaglia AC, Nista EC, Candelli M, et al. Efficacy
of Lactobacillus GG in maintaining remission of ulcerative colitis. Aliment Pharmacol Ther.
2006;23(11):1567–74. Epub 2006/05/16.
226. Dieleman LA, Goerres MS, Arends A, Sprengers D, Torrice C, Hoentjen F, et al. Lactobacillus
GG prevents recurrence of colitis in HLA-B27 transgenic rats after antibiotic treatment. Gut.
2003;52(3):370–6. Epub 2003/02/14.
227. Osman N, Adawi D, Ahrne S, Jeppsson B, Molin G. Modulation of the effect of dextran sul-
fate sodium-induced acute colitis by the administration of different probiotic strains of
Lactobacillus and Bifidobacterium. Dig Dis Sci. 2004;49(2):320–7. Epub 2004/04/24.
228. Schultz M, Veltkamp C, Dieleman LA, Grenther WB, Wyrick PB, Tonkonogy SL, et al.
Lactobacillus plantarum 299V in the treatment and prevention of spontaneous colitis in
interleukin-10-deficient mice. Inflamm Bowel Dis. 2002;8(2):71–80. Epub 2002/02/21.
229. Mileti E, Matteoli G, Iliev ID, Rescigno M. Comparison of the immunomodulatory proper-
ties of three probiotic strains of Lactobacilli using complex culture systems: prediction for
in vivo efficacy. PLoS ONE. 2009;4(9):e7056. Epub 2009/09/17.
230. Steck N, Hoffmann M, Sava IG, Kim SC, Hahne H, Tonkonogy SL, et al. Enterococcus fae-
calis metalloprotease compromises epithelial barrier and contributes to intestinal inflamma-
tion. Gastroenterology. 2011;141(3):959–71. Epub 2011/06/28.
231. McCarthy J, O’Mahony L, O’Callaghan L, Sheil B, Vaughan EE, Fitzsimons N, et al. Double
blind, placebo controlled trial of two probiotic strains in interleukin 10 knockout mice and
mechanistic link with cytokine balance. Gut. 2003;52(7):975–80. Epub 2003/06/13.
232. Setoyama H, Imaoka A, Ishikawa H, Umesaki Y. Prevention of gut inflammation by
Bifidobacterium in dextran sulfate-treated gnotobiotic mice associated with Bacteroides
strains isolated from ulcerative colitis patients. Micro Infect/ Institut Pasteur. 2003;5(2):115–
22. Epub 2003/03/26.
233. Kim N, Kunisawa J, Kweon MN, Eog Ji G, Kiyono H. Oral feeding of Bifidobacterium bifi-
dum (BGN4) prevents CD4(+) CD45RB(high) T cell-mediated inflammatory bowel disease
by inhibition of disordered T cell activation. Clin Immunol. 2007;123(1):30–9. Epub
2007/01/16.
234. Philippe D, Favre L, Foata F, Adolfsson O, Perruisseau-Carrier G, Vidal K, et al.
Bifidobacterium lactis attenuates onset of inflammation in a murine model of colitis. World J
Gastroenterol: WJG. 2011;17(4):459–69. Epub 2011/01/29.
235. Ishikawa H, Akedo I, Umesaki Y, Tanaka R, Imaoka A, Otani T. Randomized controlled trial
of the effect of bifidobacteria-fermented milk on ulcerative colitis. J Am Coll Nutr.
2003;22(1):56–63. Epub 2003/02/06.
236. Kato K, Mizuno S, Umesaki Y, Ishii Y, Sugitani M, Imaoka A, et al. Randomized placebo-
controlled trial assessing the effect of bifidobacteria-fermented milk on active ulcerative coli-
tis. Aliment Pharmacol Ther. 2004;20(10):1133–41. Epub 2004/12/01.
237. Rembacken BJ, Snelling AM, Hawkey PM, Chalmers DM, Axon AT. Non-pathogenic
Escherichia coli versus mesalazine for the treatment of ulcerative colitis: a randomised trial.
Lancet. 1999;354(9179):635–9. Epub 1999/08/31.
238. Kruis W, Fric P, Pokrotnieks J, Lukas M, Fixa B, Kascak M, et al. Maintaining remission of
ulcerative colitis with the probiotic Escherichia coli Nissle 1917 is as effective as with stan-
dard mesalazine. Gut. 2004;53(11):1617–23. Epub 2004/10/14.
239. Henker J, Muller S, Laass MW, Schreiner A, Schulze J. Probiotic Escherichia coli Nissle
1917 (EcN) for successful remission maintenance of ulcerative colitis in children
and adolescents: an open-label pilot study. Z Gastroenterol. 2008;46(9):874–5. Epub
2008/09/24.
240. Ukena SN, Singh A, Dringenberg U, Engelhardt R, Seidler U, Hansen W, et al. Probiotic
Escherichia coli Nissle 1917 inhibits leaky gut by enhancing mucosal integrity. PLoS ONE.
2007;2(12):e1308. Epub 2007/12/13.
241. Gotteland M, Cruchet S, Verbeke S. Effect of Lactobacillus ingestion on the gastrointestinal
mucosal barrier alterations induced by indometacin in humans. Aliment Pharmacol Ther.
2001;15(1):11–7. Epub 2001/01/03.
242. Cani PD, Possemiers S, Van de Wiele T, Guiot Y, Everard A, Rottier O, et al. Changes in gut
microbiota control inflammation in obese mice through a mechanism involving GLP-2-
driven improvement of gut permeability. Gut. 2009;58(8):1091–103. Epub 2009/02/26.
243. Hummel S, Veltman K, Cichon C, Sonnenborn U, Schmidt MA. Differential targeting of the
E-Cadherin/beta-Catenin complex by gram-positive probiotic lactobacilli improves epithelial
barrier function. Appl Environ Microbiol. 2012;78(4):1140–7. Epub 2011/12/20.
244. Mimura T, Rizzello F, Helwig U, Poggioli G, Schreiber S, Talbot IC, et al. Once daily high
dose probiotic therapy (VSL#3) for maintaining remission in recurrent or refractory pouchi-
tis. Gut. 2004;53(1):108–14. Epub 2003/12/20.
245. Gionchetti P, Rizzello F, Venturi A, Brigidi P, Matteuzzi D, Bazzocchi G, et al. Oral bacterio-
therapy as maintenance treatment in patients with chronic pouchitis: a double-blind, placebo-
controlled trial. Gastroenterology. 2000;119(2):305–9. Epub 2000/08/10.
246. Bibiloni R, Fedorak RN, Tannock GW, Madsen KL, Gionchetti P, Campieri M, et al. VSL#3
probiotic-mixture induces remission in patients with active ulcerative colitis. Am J
247. Sood A, Midha V, Makharia GK, Ahuja V, Singal D, Goswami P, et al. The probiotic prepara-
tion, VSL#3 induces remission in patients with mild-to-moderately active ulcerative colitis.
Clin Gastroenterol Hepatol: Off Clin Prac J Am Gastroenterol Assoc. 2009;7(11):1202–9. 9
e1. Epub 2009/07/28.
248. Miele E, Pascarella F, Giannetti E, Quaglietta L, Baldassano RN, Staiano A. Effect of a pro-
biotic preparation (VSL#3) on induction and maintenance of remission in children with
ulcerative colitis. Am J Gastroenterol. 2009;104(2):437–43. Epub 2009/01/29.
249. Malchow HA. Crohn’s disease and Escherichia coli. A new approach in therapy to maintain
remission of colonic Crohn’s disease? J Clin Gastroenterol. 1997;25(4):653–8.
250. Chermesh I, Tamir A, Reshef R, Chowers Y, Suissa A, Katz D, et al. Failure of Synbiotic 2000
to prevent postoperative recurrence of Crohn’s disease. Dig Dis Sci. 2007;52(2):385–9. Epub
2007/01/11.
251. Rachmilewitz D, Katakura K, Karmeli F, Hayashi T, Reinus C, Rudensky B, et al. Toll-like
receptor 9 signaling mediates the anti-inflammatory effects of probiotics in murine experi-
mental colitis. Gastroenterology. 2004;126(2):520–8. Epub 2004/02/06.
252. Jijon H, Backer J, Diaz H, Yeung H, Thiel D, McKaigney C, et al. DNA from probiotic bac-
teria modulates murine and human epithelial and immune function. Gastroenterology.
2004;126(5):1358–73. Epub 2004/05/08.
253. Hoermannsperger G, Clavel T, Hoffmann M, Reiff C, Kelly D, Loh G, et al. Post-translational
inhibition of IP-10 secretion in IEC by probiotic bacteria: impact on chronic inflammation.
PLoS ONE. 2009;4(2):e4365. Epub 2009/02/07.
254. von Schillde MA, Hormannsperger G, Weiher M, Alpert CA, Hahne H, Bauerl C, et al.
Lactocepin secreted by Lactobacillus exerts anti-inflammatory effects by selectively
degrading proinflammatory chemokines. Cell Host Microbe. 2012;11(4):387–96. Epub
2012/04/24.
255. Ishikawa H, Matsumoto S, Ohashi Y, Imaoka A, Setoyama H, Umesaki Y, et al. Beneficial
effects of probiotic bifidobacterium and galacto-oligosaccharide in patients with ulcerative
colitis: a randomized controlled study. Digestion. 2011;84(2):128–33. Epub 2011/04/29.
256. Furrie E, Macfarlane S, Kennedy A, Cummings JH, Walsh SV, O’Neil DA, et al. Synbiotic
therapy (Bifidobacterium longum/Synergy 1) initiates resolution of inflammation in patients
with active ulcerative colitis: a randomised controlled pilot trial. Gut. 2005;54(2):242–9.
Epub 2005/01/14.
257. Guslandi M. Treatment of irritable bowel syndrome with Saccharomyces boulardii. J Clin
Gastroenterol. 2011;45(8):740–1. author reply 1-2. Epub 2011/05/11.
258. Guslandi M, Mezzi G, Sorghi M, Testoni PA. Saccharomyces boulardii in maintenance treat-
ment of Crohn’s disease. Dig Dis Sci. 2000;45(7):1462–4. Epub 2000/08/29.
259. Summers RW, Elliott DE, Urban Jr JF, Thompson RA, Weinstock JV. Trichuris suis therapy
for active ulcerative colitis: a randomized controlled trial. Gastroenterology. 2005;128(4):825–
32. Epub 2005/04/13.
260. Dalmasso G, Cottrez F, Imbert V, Lagadec P, Peyron JF, Rampal P, et al. Saccharomyces
boulardii inhibits inflammatory bowel disease by trapping T cells in mesenteric lymph nodes.
261. Kullberg MC, Pearce EJ, Hieny SE, Sher A, Berzofsky JA. Infection with Schistosoma
mansoni alters Th1/Th2 cytokine responses to a non-parasite antigen. J Immunol.
1992;148(10):3264–70. Epub 1992/05/15.
262. Pearlman E, Kazura JW, Hazlett Jr FE, Boom WH. Modulation of murine cytokine responses
to mycobacterial antigens by helminth-induced T helper 2 cell responses. J Immunol.
1993;151(9):4857–64. Epub 1993/11/01.
263. Watanabe O, Ando T, Ishiguro K, Takahashi H, Ishikawa D, Miyake N, et al. Enteral nutrition
decreases hospitalization rate in patients with Crohn’s disease. J Gastroenterol Hepatol.
2010;25 Suppl 1:S134–7. Epub 2010/07/14.
264. Yamamoto T, Nakahigashi M, Saniabadi AR. Review article: diet and inflammatory bowel
disease–epidemiology and treatment. Aliment Pharmacol Ther. 2009;30(2):99–112. Epub
2009/05/15.
265. Yamamoto T, Shiraki M, Nakahigashi M, Umegae S, Matsumoto K. Enteral nutrition to sup-
press postoperative Crohn’s disease recurrence: a five-year prospective cohort study. Int J
Colorectal Dis. 2012;65(10):866–73. Epub 2012/09/28.
266. Day AS, Whitten KE, Lemberg DA, Clarkson C, Vitug-Sales M, Jackson R, et al. Exclusive
enteral feeding as primary therapy for Crohn’s disease in Australian children and adolescents:
a feasible and effective approach. J Gastroenterol Hepatol. 2006;21(10):1609–14. Epub
2006/08/25.
267. Leach ST, Mitchell HM, Eng WR, Zhang L, Day AS. Sustained modulation of intestinal
bacteria by exclusive enteral nutrition used to treat children with Crohn’s disease. Aliment
Pharmacol Ther. 2008;28(6):724–33. Epub 2009/01/16.
268. Hunter JO. Is diet a factor in the pathogenesis of IBD? Inflamm Bowel Dis. 2008;14 Suppl
2:S35–6. Epub 2008/09/26.
269. Kajiura T, Takeda T, Sakata S, Sakamoto M, Hashimoto M, Suzuki H, et al. Change of intes-
tinal microbiota with elemental diet and its impact on therapeutic effects in a murine model
of chronic colitis. Dig Dis Sci. 2009;54(9):1892–900. Epub 2008/12/06.
270. Pereira SP, Cassell TB, Engelman JL, Sladen GE, Murphy GM, Dowling RH. Plasma arachi-
donic acid-rich phospholipids in Crohn’s disease: response to treatment. Clin Sci (Lond).
1996;91(4):509–12. Epub 1996/10/01.
271. Zoli G, Care M, Parazza M, Spano C, Biagi PL, Bernardi M, et al. A randomized controlled
study comparing elemental diet and steroid treatment in Crohn’s disease. Aliment Pharmacol
Ther. 1997;11(4):735–40. Epub 1997/08/01.
272. Hanai H, Iida T, Takeuchi K, Arai H, Arai O, Abe J, et al. Nutritional therapy versus
6-mercaptopurine as maintenance therapy in patients with Crohn’s disease. Digest Liver Dis:
Off J Ital Soc Gastroenterol Ital Assoc Stud Liver. 2012;44(8):649–54. Epub 2012/05/01.
273. Gorard DA, Hunt JB, Payne-James JJ, Palmer KR, Rees RG, Clark ML, et al. Initial response
and subsequent course of Crohn’s disease treated with elemental diet or prednisolone. Gut.
1993;34(9):1198–202. Epub 1993/09/01.
274. Zachos M, Tondeur M, Griffiths AM. Enteral nutritional therapy for induction of remission in
Crohn's disease. Cochrane Database Syst Rev. 2007(1):CD000542. Epub 2007/01/27.
275. Wagner SJ, Schmidt A, Effenberger MJ, Gruber L, Danier J, Haller D. Semisynthetic diet ame-
liorates Crohn’s disease-like ileitis in TNFDeltaARE/WT mice through antigen-independent
mechanisms of gluten. Inflamm Bowel Dis. 2013;19(6):1285–94. Epub 2013/04/10.
276. Newnham ED. Does gluten cause gastrointestinal symptoms in subjects without coeliac dis-
ease? J Gastroenterol Hepatol. 2011;26 Suppl 3:132–4. Epub 2011/04/02.
277. Esaki M, Matsumoto T, Nakamura S, Yada S, Fujisawa K, Jo Y, et al. Factors affecting
recurrence in patients with Crohn’s disease under nutritional therapy. Dis Colon Rectum.
2006;49(10 Suppl):S68–74. Epub 2006/11/16.
278. Gentschew L, Ferguson LR. Role of nutrition and microbiota in susceptibility to inflamma-
tory bowel diseases. Mol Nutr Food Res. 2012;56(4):524–35. Epub 2012/04/13.
279. Gruber L, Lichti P, Rath E, Haller D. Nutrigenomics and nutrigenetics in inflammatory bowel
diseases. J Clin Gastroenterol. 2012;46(9):735–47. Epub 2012/09/04.
280. Hou JK, Abraham B, El-Serag H. Dietary intake and risk of developing inflammatory bowel
disease: a systematic review of the literature. Am J Gastroenterol. 2011;106(4):563–73. Epub
2011/04/07.
281. Miller B, Fervers F, Rohbeck R, Strohmeyer G. [Sugar consumption in patients with Crohn’s
disease]. Verhandlungen der Deutschen Gesellschaft fur Innere Medizin. 1976;82 Pt 1:922–4.
Epub 1976/01/01. Zuckerkonsum bei Patienten mit Morbus Crohn.
282. Mayberry JF, Rhodes J, Newcombe RG. Increased sugar consumption in Crohn’s disease.
Digestion. 1980;20(5):323–6. Epub 1980/01/01.
283. Sonnenberg A. Geographic and temporal variations of sugar and margarine consumption in
relation to Crohn’s disease. Digestion. 1988;41(3):161–71. Epub 1988/01/01.
284. Shoda R, Matsueda K, Yamato S, Umeda N. Epidemiologic analysis of Crohn disease in

Japan: increased dietary intake of n-6 polyunsaturated fatty acids and animal protein relates
to the increased incidence of Crohn disease in Japan. Am J Clin Nutr. 1996;63(5):741–5.
Epub 1996/05/01.
285. Chan SS, Luben R, Olsen A, Tjonneland A, Kaaks R, Teucher B, et al. Body mass index and
the risk for Crohn’s disease and ulcerative colitis: data from a European prospective cohort
study (The IBD in EPIC Study). Am J Gastroenterol. 2013;108(4):575–82. Epub 2013/01/16.
286. Teixeira LG, Leonel AJ, Aguilar EC, Batista NV, Alves AC, Coimbra CC, et al. The combina-
tion of high-fat diet-induced obesity and chronic ulcerative colitis reciprocally exacerbates
adipose tissue and colon inflammation. Lipids Health Dis. 2011;10:204. Epub 2011/11/15.
287. Kim KA, Gu W, Lee IA, Joh EH, Kim DH. High fat diet-induced gut microbiota exacerbates
inflammation and obesity in mice via the TLR4 signaling pathway. PLoS ONE.
2012;7(10):e47713. Epub 2012/10/24.
288. Gruber L, Kisling S, Lichti P, Martin FP, May S, Klingenspor M, et al. High fat diet acceler-
ates pathogenesis of murine Crohn’s disease-like ileitis independently of obesity. PLoS ONE.
2013;8(8):e71661. Epub 2013/08/27.
289. Ding S, Chi MM, Scull BP, Rigby R, Schwerbrock NM, Magness S, et al. High-fat diet:
bacteria interactions promote intestinal inflammation which precedes and correlates with
obesity and insulin resistance in mouse. PLoS ONE. 2010;5(8):e12191. Epub 2010/09/03.
290. Erdelyi I, Levenkova N, Lin EY, Pinto JT, Lipkin M, Quimby FW, et al. Western-style diets
induce oxidative stress and dysregulate immune responses in the colon in a mouse model of
sporadic colon cancer. J Nutr. 2009;139(11):2072–8. Epub 2009/09/18.
291. Lam YY, Ha CW, Campbell CR, Mitchell AJ, Dinudom A, Oscarsson J, et al. Increased
gut permeability and microbiota change associate with mesenteric fat inflammation and
metabolic dysfunction in diet-induced obese mice. PLoS ONE. 2012;7(3):e34233. Epub
2012/03/30.
292. Hildebrandt MA, Hoffmann C, Sherrill-Mix SA, Keilbaugh SA, Hamady M, Chen YY, et al.
High-fat diet determines the composition of the murine gut microbiome independently of
obesity. Gastroenterology. 2009;137(5):1716–24. e1-2. Epub 2009/08/27.
293. David LA, Maurice CF, Carmody RN, Gootenberg DB, Button JE, Wolfe BE, et al. Diet
rapidly and reproducibly alters the human gut microbiome. Nature. 2014;505(7484):559–63.
Epub 2013/12/18.
294. Werner T, Wagner SJ, Martinez I, Walter J, Chang JS, Clavel T, et al. Depletion of lumi-
nal iron alters the gut microbiota and prevents Crohn’s disease-like ileitis. Gut.
2011;60(3):325–33. Epub 2010/11/16.
295. Den Hond E, Hiele M, Evenepoel P, Peeters M, Ghoos Y, Rutgeerts P. In vivo butyrate
metabolism and colonic permeability in extensive ulcerative colitis. Gastroenterology.
1998;115(3):584–90. Epub 1998/08/28.
296. Roediger WE. The colonic epithelium in ulcerative colitis: an energy-deficiency disease?
Lancet. 1980;2(8197):712–5. Epub 1980/10/04.
297. Ahmad MS, Krishnan S, Ramakrishna BS, Mathan M, Pulimood AB, Murthy SN. Butyrate
and glucose metabolism by colonocytes in experimental colitis in mice. Gut.
2000;46(4):493–9. Epub 2000/03/15.
298. Ritzhaupt A, Wood IS, Ellis A, Hosie KB, Shirazi-Beechey SP. Identification and characterization
of a monocarboxylate transporter (MCT1) in pig and human colon: its potential to transport
L-lactate as well as butyrate. J Physiol. 1998;513(Pt 3):719–32. Epub 1998/11/24.
299. Thibault R, De Coppet P, Daly K, Bourreille A, Cuff M, Bonnet C, et al. Down-regulation of
the monocarboxylate transporter 1 is involved in butyrate deficiency during intestinal inflam-
mation. Gastroenterology. 2007;133(6):1916–27. Epub 2007/12/07.
300. Annese V, Valvano MR, Palmieri O, Latiano A, Bossa F, Andriulli A. Multidrug resistance
1 gene in inflammatory bowel disease: a meta-analysis. World J Gastroenterol: WJG.
2006;12(23):3636–44. Epub 2006/06/15.
301. Onnie CM, Fisher SA, Pattni R, Sanderson J, Forbes A, Lewis CM, et al. Associations of
allelic variants of the multidrug resistance gene (ABCB1 or MDR1) and inflammatory bowel
disease and their effects on disease behavior: a case-control and meta-analysis study. Inflamm
Bowel Dis. 2006;12(4):263–71. Epub 2006/04/25.
302. Annese V, Rogai F, Settesoldi A, Bagnoli S. PPARgamma in inflammatory bowel disease.
PPAR Res. 2012;2012:620839. Epub 2012/09/22.
303. Simmons JD, Mullighan C, Welsh KI, Jewell DP. Vitamin D receptor gene polymorphism:
association with Crohn’s disease susceptibility. Gut. 2000;47(2):211–4. Epub 2000/07/18.
304. Wu GD. Is there a role for PPAR gamma in IBD? Yes, no, maybe. Gastroenterology.
2003;124(5):1538–42. Epub 2003/05/06.
305. Zucchelli M, Torkvist L, Bresso F, Halfvarson J, Hellquist A, Anedda F, et al. PepT1 oligo-
peptide transporter (SLC15A1) gene polymorphism in inflammatory bowel disease. Inflamm
Bowel Dis. 2009;15(10):1562–9. Epub 2009/05/23.
306. Dalmasso G, Nguyen HT, Ingersoll SA, Ayyadurai S, Laroui H, Charania MA, et al. The
PepT1-NOD2 signaling pathway aggravates induced colitis in mice. Gastroenterology.
2011;141(4):1334–45. Epub 2011/07/19.
307. Merlin D, Si-Tahar M, Sitaraman SV, Eastburn K, Williams I, Liu X, et al. Colonic epithelial
hPepT1 expression occurs in inflammatory bowel disease: transport of bacterial peptides
influences expression of MHC class 1 molecules. Gastroenterology. 2001;120(7):1666–79.
Epub 2001/05/29.
308. van Nood E, Vrieze A, Nieuwdorp M, Fuentes S, Zoetendal EG, de Vos WM, et al. Duodenal
infusion of donor feces for recurrent Clostridium difficile. N Engl J Med. 2013;368(5):407–15.
Epub 2013/01/18.
309. Anderson JL, Edney RJ, Whelan K. Systematic review: faecal microbiota transplantation in
the management of inflammatory bowel disease. Aliment Pharmacol Ther. 2012;36(6):503–
16. Epub 2012/07/26.
310. Damman CJ, Miller SI, Surawicz CM, Zisman TL. The microbiome and inflammatory bowel
disease: is there a therapeutic role for fecal microbiota transplantation? Am J Gastroenterol.
2012;107(10):1452–9. Epub 2012/10/05.
311. Kunde S, Pham A, Bonczyk S, Crumb T, Duba M, Conrad Jr H, et al. Safety, tolerability, and
clinical response after fecal transplantation in children and young adults with ulcerative coli-
tis. J Pediatr Gastroenterol Nutr. 2013;56(6):597–601. Epub 2013/04/02.
312. Angelberger S, Reinisch W, Makristathis A, Lichtenberger C, Dejaco C, Papay P, et al.
Temporal bacterial community dynamics vary among ulcerative colitis patients after fecal
microbiota transplantation. Am J Gastroenterol. 2013;108(10):1620–30. Epub 2013/09/26.
313. Balish E, Warner T. Enterococcus faecalis induces inflammatory bowel disease in interleu-
kin-10 knockout mice. Am J Pathol. 2002;160(6):2253–7. Epub 2002/06/12.
314. Madsen KL, Doyle JS, Tavernini MM, Jewell LD, Rennie RP, Fedorak RN. Antibiotic therapy
attenuates colitis in interleukin 10 gene-deficient mice. Gastroenterology. 2000;118(6):1094–
105. Epub 2000/06/02.
315. Hoentjen F, Harmsen HJ, Braat H, Torrice CD, Mann BA, Sartor RB, et al. Antibiotics with a
selective aerobic or anaerobic spectrum have different therapeutic activities in various regions
of the colon in interleukin 10 gene deficient mice. Gut. 2003;52(12):1721–7. Epub 2003/11/25.
316. Tamagawa H, Hiroi T, Mizushima T, Ito T, Matsuda H, Kiyono H. Therapeutic effects of
roxithromycin in interleukin-10-deficient colitis. Inflamm Bowel Dis. 2007;13(5):547–56.
Epub 2007/01/25.
317. Schultz M, Tonkonogy SL, Sellon RK, Veltkamp C, Godfrey VL, Kwon J, et al. IL-2-deficient
mice raised under germfree conditions develop delayed mild focal intestinal inflammation.
Am J Physiol. 1999;276(6 Pt 1):G1461–72. Epub 1999/06/11.
318. Taurog JD, Richardson JA, Croft JT, Simmons WA, Zhou M, Fernandez-Sueiro JL, et al. The
germfree state prevents development of gut and joint inflammatory disease in HLA-B27
transgenic rats. J Exp Med. 1994;180(6):2359–64. Epub 1994/12/01.
319. Dianda L, Hanby AM, Wright NA, Sebesteny A, Hayday AC, Owen MJ. T cell receptor-alpha
beta-deficient mice fail to develop colitis in the absence of a microbial environment. Am J
Pathol. 1997;150(1):91–7. Epub 1997/01/01.
320. Bamias G, Okazawa A, Rivera-Nieves J, Arseneau KO, De La Rue SA, Pizarro TT, et al.
Commensal bacteria exacerbate intestinal inflammation but are not essential for the develop-
ment of murine ileitis. J Immunol. 2007;178(3):1809–18. Epub 2007/01/24.
321. Bamias G, Marini M, Moskaluk CA, Odashima M, Ross WG, Rivera-Nieves J, et al. Down-
regulation of intestinal lymphocyte activation and Th1 cytokine production by antibiotic
therapy in a murine model of Crohn’s disease. J Immunol. 2002;169(9):5308–14. Epub
2002/10/23.
322. Habtezion A, Toivola DM, Butcher EC, Omary MB. Keratin-8-deficient mice
develop chronic spontaneous Th2 colitis amenable to antibiotic treatment. J Cell Sci.
2005;118(Pt 9):1971–80. Epub 2005/04/21.
323. Panwala CM, Jones JC, Viney JL. A novel model of inflammatory bowel disease: mice defi-
cient for the multiple drug resistance gene, mdr1a, spontaneously develop colitis. J Immunol.
1998;161(10):5733–44. Epub 1998/11/20.
324. Kang SS, Bloom SM, Norian LA, Geske MJ, Flavell RA, Stappenbeck TS, et al. An antibiotic-
responsive mouse model of fulminant ulcerative colitis. PLoS Med. 2008;5(3):e41. Epub
2008/03/06.
325. Deng L, Zhou JF, Sellers RS, Li JF, Nguyen AV, Wang Y, et al. A novel mouse model of
inflammatory bowel disease links mammalian target of rapamycin-dependent hyperprolifera-
tion of colonic epithelium to inflammation-associated tumorigenesis. Am J Pathol.
2010;176(2):952–67. Epub 2010/01/01.
326. Morrissey PJ, Charrier K. Induction of wasting disease in SCID mice by the transfer of nor-
mal CD4+/CD45RBhi T cells and the regulation of this autoreactivity by CD4+/CD45RBlo
T cells. Res Immunol. 1994;145(5):357–62. Epub 1994/06/01.
Chapter 14
Deciphering the Gut Microbial Contribution
to the Etiology of Autism Development
Ivan K.S. Yap and François-Pierre Martin
Abstract Autistic spectrum disorder (ASD) is a spectrum of early-onset lifelong

neurodevelopmental disorders that severely impact social and behavioral functioning.
It is a debilitating disorder that affects 1 % of the global children population with
increasing prevalence and presented huge economic burden to the family and the
nation. Current diagnosis for ASD is very subjective mainly because of the multi-
factorial nature of the disorders. The etiology of ASD is highly complex and multi-
faceted involving the gene, environment, and diet and is associated with various
abnormalities that include immunologic, metabolic, and, more recently, the host–
gut microbiome stability (Fig. 14.1). The gut microbiota is a consortium of bacteria
that coexisted and coevolved with the host from the time of birth. As such the gut
microbiome–mammalian “superorganism” represents a level of biological evolu-
tionary development where true symbiosis is characterized by extensive “transge-
nomic” modulation of metabolism and functions between the two entities. The gut
microbiota is involved in various mammalian biological processes including defense
against pathogens, immunity, intestinal microvilli development, and recovery of
metabolic energy through fermentation of otherwise nondigestible dietary fiber. In
addition, the gut microbiota has been shown to communicate with the brain via the
gut–brain axis to modulate brain development, function, and behavior. Recent evi-
dence indicated that the gut microbiota influenced central nervous system develop-
ment and responses to stress. Current understanding on the potential and extend of
gut microbe involvement in brain development and host metabolic signaling is still
in its infancy. Coupled with ever-increasing awareness on the importance of the gut
microbiome in health and disease particularly autism, understanding the fundamen-
tal mechanistic interaction between host brain development and gut microbiota is
crucial for unraveling the mystery behind the etiopathology of autism.
I.K.S. Yap (*)

Department of Life Sciences, International Medical University,
126 Jalan Jalil Perkasa 19, Bukit Jalil, 57000 Kuala Lumpur, Malaysia
F.-P. Martin

DOI 10.1007/978-1-4471-6539-2_14
312 I.K.S. Yap and F.-P. Martin
Healthy Autism
Central
Healthy CNS nervous Emotional,
function system (CNS) behavioural and
neurodevelopmental
dysfunction
Normal gut
Abnormal gut
development and
functions
function
Normal gut Perturbed gut

microflora and Digestive system microflora and
metabolism metabolism
Fig. 14.1 Scheme summarizing the main factors related to gut–brain axis associated with brain
development and autism
Keywords Autism • Disease • Gut–brain axis • Gut microbiota • Health • Mass

spectrometry • Metabonomics • Metagenomics • Nuclear magnetic resonance spec-
troscopy • Nutrition • Systems biology
14.1 Gut–Brain Axis in Health and Disease
The term “gut–brain” or “brain–gut” axis is increasingly employed to define a bidi-

rectional neurohumoral communication system. It comprised of neural pathways
and humoral pathways, which include cytokines, hormones, and neuropeptides as
signaling molecules [1]. The brain–gut–enteric microbiota axis includes the central
nervous system, the neuroendocrine and neuroimmune systems, the sympathetic
and parasympathetic arms of the autonomic nervous system, the enteric nervous
system, and the intestinal microbiota [2]. Through this bidirectional communication
network, signals from the brain can influence the motor, sensory, and secretory
modalities of the GIT, and reciprocally, visceral messages from the GIT can influ-
ence brain function [2]. Recent evidences showed that gut microbiota communi-
cates with the brain via the gut–brain axis to modulate brain development and
behavioral phenotypes. In particular, this system provides the intestinal microbiota
and its metabolites with a potential route towards the brain. Consequently, the gut
microbiota could associate with brain functions as well as neurological diseases via
the gut–brain axis [3], and further insights would require a better characterization of
the composition and the metabolic activities of the gut microbiota. However,
14 Deciphering the Gut Microbial Contribution to the Etiology of Autism Development 313
identification of microbes constituting gut microbiota has been the main technological
challenge currently due to massive amount of intestinal microbes and the difficulties
in culture of gut microbes [3]. In parallel, many challenges remain to better assess,
understand, and modulate gut microbial metabolic activities and their influence at
panorganismal scale [4].
If recent studies have highlighted the depth of microbiota influence on the devel-
opment and function of the host brain, one of the first observations resulted from the
beneficial impact of orally administered antibiotics in reversing encephalopathy in
patients with decompensated liver disease [5]. In addition, accumulating evidence
further describes a relationship between psychiatric and gastrointestinal tract (GIT)
disorders, such as irritable bowel syndrome (IBS) or inflammatory bowel disease
(IBD), that are also associated with disturbances of the intestinal microbiota [6]. As
highlighted in Chap. 14 of this book, there is compelling evidence that the brain
may influence gastrointestinal functions (such as motility, secretion, and mucin pro-
duction) as well as immune functions, and therefore being a direct vehicle for medi-
ating the effects of emotional factors such as stress or depression influence [7–10].
These GIT disorders are also intimately related with gut dysbiosis [11], illustrat-
ing the potential brain influences on the microbial composition and activity along
the GIT and reciprocally for the microbiota to modulate host metabolism [1].
Moreover, the gut microbiota, the intestinal mucosa, and the intestinal immune sys-
tem issue multiple signals from the gut to the brain carried by sensory neurons,
immune mediators, gut hormones, and microbiota-derived signaling molecules
[12]. For instance, the influence of the gut microbiota on the development of the
central nervous systems and stress responses was recently documented [13]. In this,
two specific interactive systems are being highlighted, namely, the hypothalamus–
pituitary–adrenal axis and the vagus nerve, as important means of communicating
signals from gut microbes to the central nervous systems. Furthermore, recent
efforts focused on the members of the neuropeptide Y (NPY) family of biologically
active peptides, NPY, peptide YY (PYY), and pancreatic polypeptide (PP) [12].
PYY and PP are exclusively expressed by endocrine cells of the digestive system,
whereas NPY is found at all levels of the gut–brain and brain–gut axis. Recent studies
have extensively described how PYY is influenced by the intestinal microbiota, with
particular interest in appetite regulation in the context of obesity pandemic. Due to
its multilevel homeostatic mechanism, pharmacological manipulation of NPY–Y
receptor system may have considerable therapeutic efficacy in many common meta-
bolic and GIT disease in addition to psychiatric disorders.
Several lessons learned so far are mainly based and limited to preclinical studies
[7], especially using gnotobiotic and germfree animal models. Such systems models
have enabled rediscovering the multiple and complex facet of the interaction with
the gut microbiota at multi-compartmental levels [4, 14–16]. The influence of the
gut microbiota on the nervous system, brain development, and behavior, in particu-
lar during microbial colonization of the host, has recently been receiving profound
interest [17]. In particular, the metabolic modulation of metabolites influencing
functions of the nervous system, such as tryptophan and kynurenine levels, further illus-
trates the functional microbiota–neurohumoral relationship during gut colonization.
Moreover, novel evidence describes how gut microbiota type and presence can
impact the cerebral biochemical profiles [18], including cerebral glycolytic
metabolism.
14.2 Brain–Gut–Microbe Communication in Health

and Disease
Autistic spectrum disorder (ASD) is a spectrum of neurological disorders character-

ized by a complex lifelong neurodevelopmental and sociological disorder with
poorly defined etiology. ASDs are associated with an array of disabilities such as
social withdrawal, speech impairment, and repetitive behavior [19] (Fig. 14.1).
According to a recent estimate from the Centers for Disease Control and Prevention
(CDC)’s Autism and Developmental Disabilities Monitoring (ADDM) Network,
about 1 in 88 children, from the 14 communities of the network within the United
States, have been identified with an ASD [20]. A recent global estimate by Elsabbagh
and coworkers puts the global prevalence estimate to be about 62 in 10,000 [21] or
1 % of the global children population had an ASD [20]. More alarmingly, there has
been a significant increase in incidence of ASD worldwide, and within the United
States, between 2002 and 2008, there has been a 57 % increase in incidence of ASD
[20]. ASD places a large economic burden on society with the cost of the disorder
on the UK economy estimated to be £2.7 billion [22]. The lifetime cost for someone
with ASD and intellectual disability is estimated at approximately £1.23 million
and for someone with ASD without intellectual disability is approximately £0.80
million [22]. It has been reported that the ASDs affect all racial, ethnic, and socio-
economic group and that boys are five times more likely to have ASDs as compared
to girls [20]. Studies have shown that among identical twins, if one child has an
ASD, then the other will be affected about 36–95 % of the time. In nonidentical
twins, if one child has an ASD, then the other is affected about 0–31 % of the time
[23–25]. Parents who have a child with an ASD have a 2–18 % chance of having a
second child who is also affected [26, 27]. ASDs tend to occur more often in people
who have certain genetic or chromosomal conditions. About 10 % of children with
autism are also identified as having Down syndrome, fragile X syndrome, tuberous
sclerosis, and other genetic and chromosomal disorders [28–31]. A number of
known disorders such as phenylketonuria and Smith–Lemli–Opitz syndrome have
been shown to be associated with the ASD behavioral traits in children. These meta-
bolic disorders are mainly autosomal recessive genetic disorders that present within
the first 3 years of life. Current diagnosis for ASD is through a set of criteria defined
in the Diagnostic and Statistical Manual of Mental Disorders 4th Edition (DSM-
IV-TR) and behavioral observations made by clinician thus making diagnosis very
subjective. The ongoing controversy about the precise definition of ASD stems from
the current lack of understanding of the underlying causes of ASD and its multifac-
torial nature.
14.3 Gut Microbiota and ASD
Gastrointestinal dysfunction has been reported in ASD children [32], and studies
have suggested that the condition may be associated with abnormal gut microbiota.
Given the importance of the microbiome in mammalian metabolism, e.g., bile
acid metabolism, there is a possibility of previously unrecognized etiologic con-
nections between microbiome disorders and childhood developmental problems.
Individuals with ASD are commonly exposed to repeated courses of multiple anti-
biotic therapies, and this may contribute to the complex relationships between
gastrointestinal symbiosis and ASD by altering the composition or stability of
their gut microbiota.
One of the very first suggestions of the potential involvement of bacteria in ASD
was the publication by Bolte et al. where it was hypothesized that ASD may be
linked to low-grade intestinal infection with Clostridium tetani [33]. It is well
known that majority of children with ASD undergo extensive antibiotic therapy.
Oral antibiotics can disrupt the stability and integrity of the “normal” gut microbiota
thus resulting in an environment for opportunistic pathogens to colonize. One of this
opportunistic pathogen is C. tetani. Clostridium belongs to the phylum of Firmicutes.
They are rod-shaped obligate anaerobes that produce endospores [34]. Some of the
most important biological pathogens belong to this genus of bacteria, namely, C.
botulinum, C. difficile, and C. tetani, associated with botulism, antibiotic-associated
diarrhea, and tetanus, respectively. Clostridium species such as C. botulinum and C.
tetani are known to produce neurotoxins, which trigger the very clinical pathologi-
cal manifestation that they are associated with. It has been shown that toxin pro-
duced by C. tetani in the intestine of experimental animals can be transported to the
central nervous system via the vagus nerve resulting in the disruption of neurotrans-
mitters release [35, 36]. It was suggested that such inhibition may lead to the myri-
ads of behavioral deficits observed in children with ASD.
Sandler and colleagues conducted a small cohort study of oral vancomycin in
autistic children which subsequently proved the hypotheses [37]. It was reported in
the study that children receiving vancomycin treatment showed improvement in
gastrointestinal problems such as abdominal pain, constipation, and/or diarrhea.
In addition, behavioral improvements were observed with significant reduction in
aggressive behavior, increased eye contact, and significant improvement in lan-
guage and speech. However, such changes were dependent on vancomycin treat-
ment and all children relapsed after discontinuation of the antibiotic. The work by
Sandler et al. presented one of the first clear scientific evidences on the link between
gut–brain axis and ASD. Several studies had subsequently indicated that children
with ASD have perturbed gut microbiota as compared to typically developing children.
To investigate further the potential involvement of Clostridium species (spp.) in
ASD, Finegold et al. studied the feces of children with ASD and compared with
healthy controls. It was found that children with ASD have higher levels of
Clostridium spp. as well as greater species variation [38]. A subsequent study by the
same group showed that children with ASD have elevated C. bolteae and Clostridium
clusters I and XI [39]. Around the same time, a study conducted by Parracho et al.
comparing feces of children with ASD versus healthy siblings and healthy unrelated
controls showed that children with ASD has higher levels of certain Clostridium
spp. and, more interestingly, the healthy siblings had an intermediate level between
their siblings with ASD and those of the healthy unrelated controls [40]. Further, a
recent pyrosequencing study on fecal microbiota composition between typically
developing controls versus ASD children and their normal functioning siblings
showed that children with ASD had higher level of Bacteroidetes and lower level of
Firmicutes as compared to the controls [41]. In addition, children with ASD also
had lower level of several Bifidobacterium species while Desulfovibrio was higher
[32, 41]. A study by Wang et al. also showed lower level of Bifidobacterium and
Akkermansia muciniphila in ASD children [42].
In fact, the study by Finegold et al. showed that Desulfovibrio was present in half
of the autistic subjects and in some siblings. More interestingly, none of the control
subjects had Desulfovibrio [32]. Children with ASD are known to be sulfur defi-
cient. Aldred et al. showed that individuals with autism have lower levels of plasma
sulfate but considerably elevated levels of urinary sulfate as compared to normal
individuals [43, 44]. These data suggest that autistic individuals may have impaired
detoxification potential involving sulfation as evidenced by their inability to sulfate
acetaminophen [44]. The presence of sulfate-reducing bacteria such as Desulfovibrio
in ASD children could be one of the reasons behind the abnormality observed in
sulfur metabolism. Moreover, the severity of ASD behavior is positively correlated
with increased Desulfovibrio species [32].
14.4 Metabonomics in ASD
Finding the cause of ASD has proved challenging due to the multifactorial nature of
the disorder, which also means that finding biochemical markers for ASD has
remained elusive thus far. However, successful discovery of a set of specific and
accurate biomarkers for ASD would not only help in understanding the pathophysi-
ology of the condition but would, together with behavioral assessment, immensely
help in the diagnosis of ASD thus allowing the possibility of early detection and
thereby allowing early targeted intervention, which could possibly reduce severity
of ASD [19]. Metabonomics or metabolic profiling approach is becoming increas-
ingly important in identifying biomarkers of disease progression and drug interven-
tion and can provide additional information to support or aid the interpretation of
genomic and proteomic data. Since metabolic phenotypes are the results from the
interaction between host genome [45] and the environment including diet and host
microbiome [46, 47], perturbation in such complex interactions will lead to altered
metabolic profiles, which can be studied using metabonomic approaches [48–52].
One of the earlier works on urinary phenotyping of autistic children was car-
ried out by Lis et al. [53], where urine samples from autistic (n = 19) and normal
(n = unknown) children were analyzed using anion exchange chromatography.
The study showed that autistic children have abnormal levels of urinary hippurate,
4-hydroxyhippurate, and N-methyl-2-pyridone-5-carboxamide (2PY) as com-
pared to normal controls. It was postulated in the study that such observation in
the urinary profiles could be due to several factors including involvement of gut
microbiota and potential perturbation in endogenous metabolism. Hippurate is
predominantly formed by hepatic glycine conjugation of dietary and gut microbial-
derived benzoate, which is derived from plant phenolics [54]. Decreased urinary
levels of hippurate could be an indication of reduced benzoic acid synthesis by the
gut microbiota. The work by Lis et al. highlighted the potential of urine as a bio-
chemical window into understanding ASD and a viable biomatrix for biomarker
discovery and potentially diagnosis. The first metabonomic study on ASD was
conducted by Yap et al. utilizing proton nuclear magnetic resonance (1H NMR)
spectroscopy-based metabonomic approach [55]. The study looked at the meta-
bolic profiles of children diagnosed with ASD together with their non-autistic
siblings and age-matched healthy volunteers. The main findings from the study
showed alterations in nicotinic acid metabolism and gut microbe metabolism with
increased urinary levels of 2PY, N-methyl nicotinic acid, and N-methyl nicotin-
amide and decreased levels of urinary gut microbe co-metabolites such as hippu-
rate, phenylacetylglutamine, and 4-cresol sulfate. These observations were in
agreement with the findings from Lis et al. [53] implicating the involvement of
gut microbe in ASD and proved that metabonomic is an effective tool to aid
understanding of the etiology of ASD and biomarker discovery. The study by Yap
et al. also revealed differences in urinary amino acid levels such as glutamate,
alanine, glycine, and taurine [55]. More intriguingly, the study also showed that
the metabolic profiles of non-autistic siblings were quite different from their
autistic siblings as well as the age-matched healthy volunteers. Such pattern was
also observed by Parracho et al. when comparing the levels of Clostridium spp.
between children with ASD versus healthy siblings and healthy unrelated con-
trols, which showed that the healthy siblings had an intermediate level between
their ASD siblings and the healthy unrelated controls. Such observation could
indicate that the presence and levels of certain gut microbe could trigger the onset
of ASD, which could lead to perturbation in the metabolic profiles and warrant
further investigation.
Ming et al. utilized mass spectrometry-based metabonomic approach to further
investigate metabolic perturbations in ASD children versus controls with the aim
of identifying more specific biochemical disturbances linked to the pathogenesis of
ASD [56]. The study showed that individuals with ASD showed differences in
urinary amino acids such as glycine, alanine, and taurine as well as gut microbe
co-metabolites such as propionic acid derivatives and bile acids. Results from this
study validated the findings by Yap et al. [55], which showed perturbation in gut
microbial co-metabolism. Furthermore, the study by Ming et al. also showed lower
levels of urinary antioxidants carnosine and urate indicating potential increase
in oxidative stress. More recently, Emond et al. evaluated the use of gas
chromatography-coupled mass spectrometry metabonomic approach to study the
urinary biochemical profiles of autistic versus healthy children [57]. The study
successfully differentiated autistic from healthy children, and several metabolites

were identified to be significantly contributing the differences. Urinary metabolites
succinate and glycolate were found to be higher in autistic children, whereas
metabolites such as hippurate, 3-hydroxyphenylacetate, 3-hydroxyhippurate, and
several other metabolites were found to be lower in autistic children. Interestingly,
the urinary metabolites that were lower in autistic children were largely gut microbe
co-metabolites.
Results from all four studies indicated a common factor that is the perturbation
of gut microbe co-metabolites in ASD individuals [53, 55–57]. The later three studies
[55–57] also demonstrated the potential of metabonomic as a noninvasive tool to
study and understand the etiopathophysiology of ASD. In addition, the biochemical
changes observed from these studies may provide novel biomarker information
applicable for diagnostic and monitoring therapeutic interactions in the condition.
The key findings on the role of the microbiome in brain development and the etiology
and development of autism were summarized in Table 14.1.
Table 14.1 Role of the microbiome and microbial–host co-metabolites associated with brain
development and autism: overview of key references
Biological
compartment Main findings Refs
Microbiome Certain bacteria produce neurotoxin that can be transported to the [35,
central nervous system via the vagus nerve. Bacteria implicated: 36]
Clostridium species; Clostridium tetani
Vancomycin-treatment leads to short-term improvement in [37]
gastrointestinal problems and behavioral improvements
Higher levels and greater variation of Clostridium species were [38–
found in children with ASD, and normal functioning siblings of 40]
autistic children have intermediate level of Clostridium species
between ASD and healthy unrelated controls. Bacteria implicated:
Clostridium species
Children with ASD were reported to have higher level of [41]
Bacteroidetes and lower level of Firmicutes. Bacteria implicated:
Bacteroidetes; Firmicutes
Metabolome Abnormal levels of urinary hippurate, 4-hydroxyhippurate, and [53]
N-methyl-2-pyridone-5-carboxamide (2PY) found in ASD children
First metabonomics study on ASD. Reported increased urinary [55]
levels of 2PY, N-methyl nicotinic acid, and N-methyl nicotinamide
and decreased levels of urinary gut microbe co-metabolites such as
hippurate, phenylacetylglutamine, and 4-cresol sulfate in ASD
children
Metabolic profiling showed differences in urinary amino acids, i.e., [56]
glycine, alanine, and taurine, as well as gut microbe co-metabolites,
i.e., propionic acid derivatives and bile acids, between normal and
ASD children
Found levels of urinary succinate and glycolate higher in ASD [57]
children and levels of urinary hippurate, 3-hydroxyphenylacetate,
3-hydroxyhippurate lower in ASD children
The growing ASD incidence attracts interest in better defining the role of recent
changes in food intake and exposure in the etiology of the disease. For instance,
essential fatty acids taken in diets mediate brain functions and structures during
development and are involved in many brain-related disorders like autism [58].
Among the various lipid species, cell membrane components, including mainly
phospholipids, are very rich in PUFAs in brain tissue, with AA and DHA represent-
ing up to 20 % of the dry brain weight [58]. Abnormalities in the fatty acid composi-
tions of phospholipids have been implicated in several neurodevelopmental
disorders that manifest with psychiatric symptoms. In particular, alteration in fatty
acids and phospholipids, including not only reduced levels of n-3 PUFAs but also
increased levels of saturated fatty acids in the red blood cell membrane [59] or in
plasma [21], was described in autistic subjects. In particular, blood plasma of autistic
patients showed an increase in most of the saturated fatty acids except for propionic
acid and a decrease in most of polyunsaturated fatty acids, which could relate to
multifactorial processes ranging from oxidative stress to mitochondrial dysfunction
and lead to induced metabolic alterations in Saudi autistic patients [60]. The
concomitant alteration in phospholipase activity associated with decreased levels of
AA, docosatetraenoic acid, and DHA in red blood cell membranes from autistic
subjects further supports a fundamental role of the phospholipid metabolic regula-
tion in autism and the potential role of nutritional intervention for future prevention
strategies [61]. This was recently exemplified by El-Ansary et al. [58] in a study
comparing the relative concentrations of essential fatty acids (linoleic and alpha-
linolenic), their long chain polyunsaturated fatty acids, and phospholipids in plasma
of autistic patients from Saudi Arabia with age-matching controls. They reported a
significant modulation of the metabolism of fatty acids, as assessed via an alteration
of the ratio between essential fatty acids/long chain polyunsaturated fatty acids and
omega-3/omega-6 fatty acids, and a decrease in circulating levels of phospholipids.
The authors provide particular emphasis on phosphatidylethanolamine, phosphati-
dylserine, and phosphatidylcholine lipid species which could be used as potential
biomarkers for future treatment or prevention strategies.
14.5 Therapeutic Perspectives
The potentiality of using metabolic and gut microbial metabolic markers for future
therapeutic perspectives is significant but so far at their infancy, due to the yet limited
definition and understanding of the processes leading to the gut–brain axis dysfunc-
tion in ASD. One should separate the nutritional approaches aiming at prevention
from management of specific conditions or disease stages. Once additional and more
consolidated phenotype characterizations of the human host–microbiome are avail-
able, more studies should be dedicated to investigate metabolic features associated
with the gradual development of the ASD dysfunction. In particular, family at-risk
subpopulations should be defined and studied in order to generate further hypotheses
on environmental and nutritional strategies for prevention and management.
14.6 Conclusions
Comprehensive and long-term phenotyping of populations at risk is envisioned to

provide some key and still missing insights into understanding mechanisms involved
into the pandemic development of ASD. In this, novel methodologies, enabling to
rediscover the intimate relationships with the gut functional ecology and interac-
tions along the gut–brain axis, are foreseen as a fundamental cornerstone of the
molecular mechanisms at play. The molecular hypotheses about etiology of the
metabolic phenotype are still highly debated, but they suggest that patients should
be screened for their microbiota for therapeutic strategies or preventive programs,
which could benefit from novel and minimally invasive systems biology approaches.
References
1. Bercik P. The microbiota-gut-brain axis: learning from intestinal bacteria? Gut.

2011;60:288–9.
2. Grenham S, Clarke G, Cryan JF, et al. Brain-gut-microbe communication in health and disease.
Front Physiol. 2011;2:94.
3. Chen X, D’Souza R, Hong ST. The role of gut microbiota in the gut-brain axis: current chal-
lenges and perspectives. Protein Cell. 2013;4:403–14.
4. Martin FP, Sprenger N, Yap IK, et al. Panorganismal gut microbiome-host metabolic crosstalk.
J Proteome Res. 2009;8:2090–105.
5. Schiano TD. Treatment options for hepatic encephalopathy. Pharmacotherapy. 2010;30:16S–21.
6. Wu JC. Psychological co-morbidity in functional gastrointestinal disorders: epidemiology.
Mech Manag J Neurogastroenterol Motil. 2012;18:13–8.
7. Collins SM, Surette M, Bercik P. The interplay between the intestinal microbiota and the brain.
8. Alonso C, Guilarte M, Vicario M, et al. Acute experimental stress evokes a differential gender-
determined increase in human intestinal macromolecular permeability. Neurogastroenterol
Motil Off J Eur Gastrointest Motil Soc. 2012;24(740–6):e348–9.
9. Rezzi S, Martin FP, Alonso C, et al. Metabotyping of biofluids reveals stress-based differences
in gut permeability in healthy individuals. J Proteome Res. 2009;8:4799–809.
10. Santos J, Perdue MH. Stress and neuroimmune regulation of gut mucosal function. Gut.
2000;47 Suppl 4:iv49–51; discussion iv52.
11. Backhed F. Host responses to the human microbiome. Nutr Rev. 2012;70 Suppl 1:S14–7.
12. Holzer P, Reichmann F, Farzi A. Neuropeptide Y, peptide YY and pancreatic polypeptide in the
gut-brain axis. Neuropeptides. 2012;46:261–74.
13. Forsythe P, Kunze WA, Bienenstock J. On communication between gut microbes and the
brain. Curr Opin Gastroenterol. 2012;28:557–62.
14. Claus SP, Tsang TM, Wang Y, et al. Systemic multicompartmental effects of the gut microbiome
on mouse metabolic phenotypes. Mol Syst Biol. 2008;4:219.
15. Martin FP, Dumas ME, Wang Y, et al. A top-down systems biology view of microbiome-
mammalian metabolic interactions in a mouse model. Mol Syst Biol. 2007;3:112.
16. Claus SP, Ellero SL, Berger B, et al. Colonization-induced host-gut microbial metabolic inter-
action. mBio. 2011;2:e00271–10.
17. El Aidy S, Kunze W, Bienenstock J, et al. The microbiota and the gut-brain axis: insights from
the temporal and spatial mucosal alterations during colonisation of the germfree mouse
intestine. Benefic Microbes. 2012;3:251–9.
18. Matsumoto M, Kibe R, Ooga T, et al. Cerebral low-molecular metabolites influenced by intestinal
microbiota: a pilot study. Front Syst Neurosci. 2013;7:9.
19. Walsh P, Elsabbagh M, Bolton P, et al. In search of biomarkers for autism: scientific, social and
ethical challenges. Nat Rev Neurosci. 2011;12:603–12.
20. Autism and Developmental Disabilities Monitoring Network Surveillance Year Principal
Investigators, Centers for Disease Control and Prevention, et al. Prevalence of autism spectrum
disorders – autism and developmental disabilities monitoring network, 14 sites, United States,
2008. Morb Mortal Wkly Rep Surveill Summ. 2012;61:1–19.
21. Elsabbagh M, Divan G, Koh YJ, et al. Global prevalence of autism and other pervasive devel-
opmental disorders. Autism Res Off J Int Soc Autism Res. 2012;5:160–79.
22. Knapp M, Romeo R, Beecham J. Economic cost of autism in the UK. Autism Int J Res Practice.
2009;13:317–36.
23. Rosenberg RE, Law JK, Yenokyan G, et al. Characteristics and concordance of autism spec-
trum disorders among 277 twin pairs. Arch Pediatr Adolesc Med. 2009;163:907–14.
24. Hallmayer J, Cleveland S, Torres A, et al. Genetic heritability and shared environmental
factors among twin pairs with autism. Arch Gen Psychiatry. 2011;68:1095–102.
25. Ronald A, Happe F, Bolton P, et al. Genetic heterogeneity between the three components of the
autism spectrum: a twin study. J Am Acad Child Adolesc Psychiatry. 2006;45:691–9.
26. Ozonoff S, Young GS, Carter A, et al. Recurrence risk for autism spectrum disorders: a baby
siblings research consortium study. Pediatrics. 2011;128:e488–95.
27. Sumi S, Taniai H, Miyachi T, et al. Sibling risk of pervasive developmental disorder estimated
by means of an epidemiologic survey in Nagoya, Japan. J Hum Genet. 2006;51:518–22.
28. DiGuiseppi C, Hepburn S, Davis JM, et al. Screening for autism spectrum disorders in children
with down syndrome: population prevalence and screening test characteristics. J Dev Behav
Pediatr: JDBP. 2010;31:181–91.
29. Cohen D, Pichard N, Tordjman S, et al. Specific genetic disorders and autism: clinical contri-
bution towards their identification. J Autism Dev Disord. 2005;35:103–16.
30. Hall SS, Lightbody AA, Reiss AL. Compulsive, self-injurious, and autistic behavior in chil-
dren and adolescents with fragile X syndrome. Am J Ment Retard: AJMR. 2008;113:44–53.
31. Zecavati N, Spence SJ. Neurometabolic disorders and dysfunction in autism spectrum disor-
ders. Curr Neurol Neurosci Rep. 2009;9:129–36.
32. Finegold SM. State of the art; microbiology in health and disease. Intestinal bacterial flora in
autism. Anaerobe. 2011;17:367–8.
33. Bolte ER. Autism and clostridium tetani. Med Hypotheses. 1998;51:133–44.
34. Bruggemann H, Gottschalk G. Clostridia : molecular biology in the post-genomic era.
Wymondham: Caister Academic; 2009.
35. Ahnert-Hilger G, Bigalke H. Molecular aspects of tetanus and botulinum neurotoxin poison-
ing. Prog Neurobiol. 1995;46:83–96.
36. Manning KA, Erichsen JT, Evinger C. Retrograde transneuronal transport properties of frag-
ment c of tetanus toxin. Neuroscience. 1990;34:251–63.
37. Sandler RH, Finegold SM, Bolte ER, et al. Short-term benefit from oral vancomycin treatment
of regressive-onset autism. J Child Neurol. 2000;15:429–35.
38. Finegold SM, Molitoris D, Song Y, et al. Gastrointestinal microbiota studies in late-onset
autism. Clin Infect Dis Off Publ Infect Dis Soc Am. 2002;35:S6–16.
39. Song Y, Liu C, Finegold SM. Real-time PCR quantitation of clostridia in feces of autistic chil-
dren. Appl Environ Microbiol. 2004;70:6459–65.
40. Parracho HM, Bingham MO, Gibson GR, et al. Differences between the gut microbiota of
children with autistic spectrum disorders and that of healthy children. J Med Microbiol.
2005;54:987–91.
41. Finegold SM, Dowd SE, Gontcharova V, et al. Pyrosequencing study of fecal microbiota of
autistic and control children. Anaerobe. 2010;16:444–53.
42. Wang L, Christophersen CT, Sorich MJ, et al. Low relative abundances of the mucolytic
bacterium Akkermansia muciniphila and Bifidobacterium spp. in feces of children with autism.
Appl Environ Microbiol. 2011;77:6718–21.
43. Aldred S, Moore KM, Fitzgerald M, et al. Plasma amino acid levels in children with autism
and their families. J Autism Dev Disord. 2003;33:93–7.
44. Alberti A, Pirrone P, Elia M, et al. Sulphation deficit in “low-functioning” autistic children: a
pilot study. Biol Psychiatry. 1999;46:420–4.
45. Sabeti PC, Varilly P, Fry B, et al. Genome-wide detection and characterization of positive
selection in human populations. Nature. 2007;449:913–8.
46. Nicholson JK, Holmes E, Kinross J, et al. Host-gut microbiota metabolic interactions. Science.
2012;336:1262–7.
47. Holmes E, Loo RL, Stamler J, et al. Human metabolic phenotype diversity and its association
with diet and blood pressure. Nature. 2008;453:396–400.
48. Kinross JM, Alkhamesi N, Barton RH, et al. Global metabolic phenotyping in an experimental
laparotomy model of surgical trauma. J Proteome Res. 2011;10:277–87.
49. Saric J, Li JV, Wang Y, et al. Panorganismal metabolic response modeling of an experimental
Echinostoma caproni infection in the mouse. J Proteome Res. 2009;8:3899–911.
50. Skordi E, Yap IK, Claus SP, et al. Analysis of time-related metabolic fluctuations induced by
ethionine in the rat. J Proteome Res. 2007;6:4572–81.
51. Thomas EL, Parkinson JR, Hyde MJ, et al. Aberrant adiposity and ectopic lipid deposition
characterize the adult phenotype of the preterm infant. Pediatr Res. 2011;70:507–12.
52. Yap IK, Brown IJ, Chan Q, et al. Metabolome-wide association study identifies multiple
biomarkers that discriminate north and south Chinese populations at differing risks of cardio-
vascular disease: intermap study. J Proteome Res. 2010;9:6647–54.
53. Lis AW, McLaughlin I, Mpclaughlin RK, et al. Profiles of ultraviolet-absorbing components of
urine from autistic children, as obtained by high-resolution ion-exchange chromatography.
Clin Chem. 1976;22:1528–32.
54. Schwab AJ, Tao L, Yoshimura T, et al. Hepatic uptake and metabolism of benzoate: a multiple
indicator dilution, perfused rat liver study. Am J Physiol Gastrointest Liver Physiol.
2001;280:G1124–36.
55. Yap IK, Angley M, Veselkov KA, et al. Urinary metabolic phenotyping differentiates children
with autism from their unaffected siblings and age-matched controls. J Proteome Res.
2010;9:2996–3004.
56. Ming X, Stein TP, Barnes V, et al. Metabolic perturbance in autism spectrum disorders: a meta-
bonomics study. J Proteome Res. 2012;11:5856–62.
57. Emond P, Mavel S, Aidoud N, et al. GC-MS-based urine metabolic profiling of autism spectrum
disorders. Anal Bioanal Chem. 2013;405:5291–300
58. El-Ansary AK, Bacha AG, Al-Ayahdi LY. Impaired plasma phospholipids and relative amounts
of essential polyunsaturated fatty acids in autistic patients from Saudi Arabia. Lipids Health
Dis. 2011;10:63.
59. Bell JG, Sargent JR, Tocher DR, et al. Red blood cell fatty acid compositions in a patient with
autistic spectrum disorder: a characteristic abnormality in neurodevelopmental disorders?
Prostaglandins Leukot Essent Fatty Acids. 2000;63:21–5.
60. El-Ansary AK, Bacha AG, Al-Ayahdi LY. Plasma fatty acids as diagnostic markers in autistic
patients from Saudi Arabia. Lipids Health Dis. 2011;10:62.
61. Bell JG, MacKinlay EE, Dick JR, et al. Essential fatty acids and phospholipase A2 in autistic
spectrum disorders. Prostaglandins Leukot Essent Fatty Acids. 2004;71:201–4.
Chapter 15
The Modulation of Drug Efficacy
and Toxicity by the Gut Microbiome
Ian D. Wilson and Jeremy K. Nicholson
Abstract The gut microbiota have been shown to have an important influence,
direct and indirect, on the metabolism and toxicity of a wide range of drugs and
xenobiotics. The major drug-metabolizing capability of the gut microbiota is reduc-
tive metabolism, but demethylation, dehydroxylation, deacylation, decarboxylation,
and hydrolysis reactions have also been demonstrated as well as acetylation.
Microbiome-driven drug metabolism can result positively in the activation of pro-
drugs to their pharmacologically active forms or alternatively result in adverse con-
sequences such as toxicity. In addition, the gut microbiota can affect drug metabolism
and toxicity indirectly via, e.g., competition of bacterial-derived metabolites for
xenobiotic metabolism pathways or the modulation of host metabolic systems.
Keywords Deconjugation • Drug metabolism • Drugs • Gut bacteria • Metabolism

• Microbiome • Prodrug activation • Toxicity • Xenobiotics
15.1 Introduction
With our expanding knowledge of the composition and role of the gut microbiota in
human health and disease, interest has once more focused on the role of this impor-
tant “external” organ in modifying the pharmacological effects of drugs, or their
toxicity, and it currently forms an area of active research. In addition there is the
interesting question of the modification of the microbiome using pharmaceuticals.
This can be either intentional, in order to modify some of the activities of the gut
microbiota beneficially, or unintentional, as a result of “collateral damage” caused
by, e.g., exposure of the host to antibiotics, and cytotoxics etc. The latter poses the
further question of how far one should go to mitigate this unintended damage to the
gut ecosystem in the light of our increasing understanding of the importance of the
I.D. Wilson (*) • J.K. Nicholson

Department of Surgery and Cancer, Faculty of Medicine, Imperial College,
Exhibition Road, South Kensington, London SW7 2AZ, UK

DOI 10.1007/978-1-4471-6539-2_15
324 I.D. Wilson and J.K. Nicholson
gut microbiome to the host. It is now clearer than ever that these organisms are not
merely passengers in the hosts digestive system, but important members of the crew.
As is now well accepted, in the adult human, the gut contains up to ca. a kilogram
of bacteria, the bulk of which comprise species of obligate anaerobes from the gen-
era Bacteroides, Clostridium, Lactobacillus, Escherichia, Bifidobacteria, and
others, together with a variety of yeasts and other microorganisms. This results in a
complex and still incompletely characterized ecology comprising ca. 1,000 species
that coexist in a dynamic equilibrium. For the host the microbiome provides benefits
via enhanced energy recovery from food, defense against pathogens, and interac-
tions with the immune and nervous systems through a number of signaling mole-
cules and metabolites. The composition of the gut microbiome is different between
individuals and depends on many factors such as whether birth was conventional or
by Cesarean section, diet, antibiotic treatment, environment, etc.
However, for the drug metabolism and toxicology communities, despite many
early studies showing its importance in some instances of xenobiotic biotransforma-
tion, it is arguable that gut microbiota of animals and man has (at least until recently)
become a “forgotten organ” for modulating drug metabolism, disposition, and tox-
icity (this is despite an early focus in vitro studies employing gut microbiota incuba-
tions where their influence of drug metabolism was investigated in depth (e.g., see
Refs. [1, 2]). This lack of current awareness is potentially a serious oversight as the
microbiota are not only capable of a very wide range of biotransformations but are
also a source of physiological variability between individuals with clear potential to
affect the disposition, and toxicity, of drugs and metabolites via secondary interac-
tions mediated through, e.g., the metabolic exchange and the co-metabolism and
processing of many diverse substrates [3]. Indeed, thanks to the gut microbiota, the
host is continually exposed to bacterially derived metabolites and waste products
that not only require processing and elimination but that may also modulate, to a
greater or lesser extent, the host’s drug-metabolizing systems via, e.g., the induction
of drug-metabolizing enzymes (or their inhibition) or by the production of second-
ary metabolites that compete for particular drug-metabolizing pathways to give, in
an analogous way to drug-drug interactions (DDIs), drug-microbiome interactions
(DMIs). That this should be the case is unsurprising as undoubtedly an important
driver for the development of the, so-called, phase I and II drug-metabolizing sys-
tems was the need to eliminate unwanted microbiota-derived metabolites such as,
e.g., ethanol and benzoic acid. Such factors may be important in both personalized
medicine and so-called idiosyncratic drug toxicity.
15.2 Drug-Metabolizing Capabilities of the Gut Microbiota
The metabolic capacity of the gut microbiota is large and is capable of undertaking a
wide number of biotransformations on drugs and other xenobiotics. These metabolic
capabilities include reductive and hydrolytic reactions as well as decarboxylation,
dehydroxylation, dealkylation, de-halogenation, deamination, etc. (Fig. 15.1).
15 The Modulation of Drug Efficacy and Toxicity by the Gut Microbiome 325
Oral i.v.
DRUG
Gut Microbiota GI Tract Liver Kidney
Oxidation
Reduction Reduction Oxidation
Hydrolysis Hydrolysis Conjugation
Dehydroxylation Conjugation
Demethylation
Dealkylation
Decarboxylation
Excretion Excretion
Acetylation
via bile via urine
Elimination via Urine and Faeces
Fig. 15.1 A schematic highlighting the major metabolic reactions that are undertaken by the gut
microbiota and illustrating the metabolic linkage with the host liver where biliary excretion and
enterohepatic cycling of metabolites can result in host-microbiome co-metabolism
Another important set of reactions includes the modification of xenobiotic metabolites

produced by the host but excreted into the gut (generally via the bile) such as
glucuronides, sulfates, and glutathione conjugates. Examples, and potential conse-
quences, of these types of drug metabolism are given below.
15.3 Microbiome-Based Metabolism of Drugs

and Other Bioactive Compounds
15.3.1 Reductive Metabolism
The environment of the gut is well suited to non-oxidative metabolism, and the
microbiome therefore provides the host with a metabolic organ of considerable
capacity for the reductive biotransformation of drugs, and drug-like compounds,
such as the polar azo dyes prontosil (p-(2,4-diaminophenylazo)benzenesulfon-
amide) [4] and neoprontosil (sodium 2-(p-sulphamyl-phenylazo)-7-acetamido-1-
hydroxynaphthalene-3,6-disulfonate) [5]. These compounds are both prodrugs of
sulfanilamide. Such reductions can therefore have important consequences result-
ing in the conversion of “prodrugs” into their active forms, the inactivation of com-
pounds, or the toxification of otherwise innocuous compounds. Rafii and Cerniglia
[6] demonstrated that a number of anaerobic species of bacteria present in the GI
tract of humans could reductively metabolize a range of azo dyes and nitropolycyclic
aromatic hydrocarbons to aromatic amines. This activity resided in species of the
genera Clostridium and Eubacterium. They studied the azo- and nitroreductases
from three Clostridium strains and one Eubacterium strain, noting that these
enzymes were produced constitutively in each of the bacteria, with the enzymes
exhibiting different electrophoretic mobilities depending upon the bacteria from
which they had been isolated. However, based on comparison of their electropho-
retic mobilities, antibody affinity, and the fact that nitroaromatic compounds could
act as competitive inhibitors against azoreductase activity, they considered that both
azo- and nitroreductase activities resided in a single enzyme.
As well as prontosil and neoprontosil, important reductive bioactivations include
those of, e.g., the 5-aminosalicylic acid prodrugs sulfasalazine (salicyl-azo-sulfapyridine)
[7, 8], olsalazine [9], ipsalazide (5-Carboxymethylcarbamoyl-4-phenylazo)-salicylic
acid), and balsalazide (5-(carboxyethtylcarbamoyl-4-phenylazo)-salicylic acid) [10]
used in ulcerative colitis and inflammatory bowel conditions. Reductive metabolism of
these compounds by the gut microbiota is essential to obtain the desired pharmacological
activity in the gut via the release of the anti-inflammatory aminosalicylic acid moiety.
The consequences of reductive metabolism by the mictobiota can however, in
addition to beneficial drug activation, potentially be quite serious. Thus the toxifi-
cation of drugs via reductive metabolism has been implicated in the case of
nitrazepam-induced teratogenicity in rats. In studies in the rat, Takeno and Sakai
[11] found that incubation of nitrazepam in caecal content-derived bacterial sus-
pensions resulted in extensive nitroreduction of the drug reduction to 7-aminoni-
trazepam (rat liver homogenates could also perform this biotransformation but
only under anaerobic conditions). At an oral dose of 300 mg/kg nitrazepam, to
pregnant rats, the nitro-reduced metabolites 7-aminonitrazepam and 7-acetylami-
nonitrazepam accounted for ca. 30 % of the drug-related material recovered in the
excreta, falling to only 2 % following pretreatment with antibiotics. A direct rela-
tionship with antibiotic treatment and the decline in nitroreduction was confirmed
when the nitroreductase activity of caecal contents was found to be almost com-
pletely eliminated by the antibiotic pretreatment (the activity in the liver homoge-
nates was unaffected). In addition antibiotic pretreatment was observed to reduce
the teratogenic effects of the drug. The combination of these experiments led the
authors to conclude that the intestinal microbiota not only played an important role
in the reductive metabolism of nitrazepam but that the teratogenic effects of the
drug were not unrelated to this.
Similarly, when the in vitro and in vivo metabolism of clonazepam was investi-
gated using germfree and ex-germfree rats, significant amounts of reduction of the
drug clonazepam to 7-aminoclonazepam were seen that appeared to be related to the
gut microbiota [12]. As with nitrazepam hepatic microsomes were also capable of
this biotransformation but only under anaerobic conditions. When [14C]-clonazepam
was dosed orally to germfree rats, the reduced metabolites of the drug accounted for
only 15 % of the radioactivity in the urine, with over 70 % of the dose as a phenolic
metabolite. However, following colonization with an intestinal flora, some 77 % of
the metabolites observed resulted from nitroreduction with the major product
identified as 7-acetamidoclonazepam.
The reductive cleavage of hydrazone linkages in drugs has also been observed.
Thus, when the drug levosimendan was administered to dogs, Antila et al. [13]
found evidence that the hydrazone linkage was cleaved by microbial action in the
lower gastrointestinal tract. It was subsequently observed in humans that the
cleavage product, (R)-6-(4-aminophenyl)-5-methyl-4,5-dihydropyridazin-3(2H)-
one (OR-1855), was then absorbed from the gut and subjected to further biotrans-
formation to a pharmacologically active metabolite [14]. Similarly, following oral
dosing to man, the drug eltrombopag was found to undergo extensive cleavage of its
hydrazine linkage due to metabolism by gut microbiota [15]. This was evidenced by
the production of the cleavage products following anaerobic incubation with rodent
caecal contents or human fecal homogenate that was suppressed by antibiotics in
both in vitro and in vivo experiments. However, on the basis of further in vivo stud-
ies, the authors concluded that concomitant antibiotic treatment in patients was
unlikely to affect the pharmacokinetics of the drug in humans.
Reduction is, however, not limited to nitrogen-containing functional groups.
Strong et al. [16] demonstrated the microbiota-driven reduction of the sulfoxide-
containing drugs sulfinpyrazone and sulindac when these were incubated with
human or rabbit feces. Sulfinpyrazone was reduced to a greater extent under
anaerobic compared to aerobic conditions and it was noted that the reduction of
sulindac was more extensive than that of sulfinpyrazone in human-derived feces.
The presence of the antibiotics metronidazole and lincomycin in the growth
medium markedly inhibited the reduction of sulfinpyrazone in human feces
while tetracycline reduced sulfide production in rabbit feces/caecal contents.
The formation of the sulfides of both sulindac and sulfinpyrazone ex vivo was
found to be reduced for feces obtained from patients treated with metronida-
zole, and this antibiotic also decreased the in vivo reduction of sulfinpyrazone
by the rabbit. More recently the H+/K+ATPase inhibitor omeprazole has also
been shown in vitro to be reduced to its sulfide metabolite by the intestinal
microbiota [17].
A particularly interesting example of reductive drug metabolism by the micro-
biota relates to the clinically important cardiac drug digoxin which is subject to
quite extensive reductive metabolism to less pharmacologically active metabolites
such as dihydrodigoxin and related compounds. Thus Lindenbaum et al. [18] deter-
mined the urinary excretion of the relatively cardioinactive reduced metabolites of
digoxin (dihydrodigoxin and related compounds) in 131 normal subjects during
studies of digoxin bioavailability. They found that in one-third of the subjects,
reduced metabolites of the drug accounted for more than 5 % of the excretion of
digoxin and its metabolites (after either single or multiple doses). This result was
stable such that volunteers continued to excrete the same proportion of the reduced
metabolites on repeated exposure to the drug digoxin. However, the exposure of
some subjects to erythromycin resulted in a failure to excrete further quantities of
the reduced metabolites following subsequent doses. In addition the presence of the
reduced metabolites in urine was lower if the drug was administered via the intrave-
nous route (with oral administration the excretion of the reduced metabolites of
digoxin was seen to vary inversely with bioavailability). On the basis of these find-
ings, the authors hypothesized that this metabolic reduction was the result “of the
activity of a variable component of the intestinal flora.” Further studies [19, 20]
implicated Eubacterium lentum as the sole organism forming these cardioinactive
reduced metabolites of digoxin. However, the authors noted that the presence of this
bacterium in the stools did not automatically mean that the host would excrete the
reduced metabolite and further noted that there was an inverse relationship between
the presence of increasing amounts of arginine in the growth medium and produc-
tion of the reduced metabolites (observations that have only recently been explained
as described below). Subsequently stool samples acquired from a group of 77 nurs-
ing home residents were analyzed to determine if there was a relationship between
colonization with digoxin-reducing strains of E. lentum and infection with
Clostridium difficile and the effects of previous antibiotic treatment, enteral feeding,
and bowel movement habits [21]. This investigation found that colonization with
the digoxin-reducing E. lentum was less prevalent in patients infected with C. diffi-
cile, previously treated with either antibiotics or enteral feedings. In addition nor-
mal bowel habits were more commonly associated with subjects who were not
colonized with C. difficile.
As well as the factors described above, the same group also demonstrated that
age was also important in the development of the ability of the gut microbiota to
inactivate digoxin [22]. Studies in children showed that none of those less than
8 months of age produced reduced digoxin metabolites. A more adult pattern of
digoxin metabolism was seen with age with one-third of the subjects studied pro-
ducing reduced metabolites of the drug after 16 months of age. However, the levels
of reductive digoxin metabolism seen in ca. 10 % of the adult populations were not
observed in patients that were under 9 years of age. Bacterial cultures prepared from
feces obtained from 73 babies aged less than 8 months were, however, shown to
contain the appropriate digoxin-reducing bacteria, leading the authors to conclude
that “maturation of the gut microbiota with respect to digoxin metabolism appears
to be a protracted process. The relative digoxin resistance of infants and children is
not due to bacterial inactivation” [22].
Differences in the production of reduced digoxin metabolites between populations
have also been noted with one investigation showing that the drug was converted to
reduced metabolites to a much greater extent by North American subjects (36 %)
compared with a South Indian population (13.7 %). This difference was maintained
when subjects from India went to live in the USA [23]. Within the
Indian population there were also significant differences with the rural village popu-
lation observed to produce ca. 5 % of the reduced metabolites as opposed to 23 %
by urban dwellers.
Most recently Haiser et al. [24] have reinvestigated the reductive metabolism of
digoxin using modern methods of molecular biology. Using a combination of tran-
scriptional profiling, comparative genomics, and culture-based assays, they found a
cytochrome-encoding operon that was upregulated by digoxin and inhibited by
arginine. This was not present in strains of E. lenta that did not metabolize the drug
and was predictive of digoxin inactivation by the human gut microbiome. In vivo
studies in gnotobiotic mice colonized with the digoxin-reducing strain of E. lenta
revealed that dietary protein reduced the microbial metabolism of digoxin, resulting
in changes in serum pharmacokinetics and urinary excretion (there was no such
effect in mice colonized with the nonreducing strain of these bacteria). What then
emerges from this most recent study is a more comprehensive picture of a microbi-
ome-drug interaction that requires an appropriate strain of the bacteria involved to
be present, with the resulting activity observed possibly also modulated by diet.
The gut reductive microbiota metabolism of metronidazole via the ring nitro
group to its amino metabolite 1-(2-aminoimidazol-1-yl)-3 methoxypropanol-2-ol
and acetamide (a known rat carcinogen) has been demonstrated both in vitro and
in vivo in the rat [25, 26]. The suggested mechanism involved nucleophilic attack
at carbons 2 and 4 of a partially reduced nitroimidazole ring, which was subse-
quently cleaved between positions 1 and 2 and 3 and 4 to yield both acetamide and
N-(2-hydroxyethyl)-oxamic acid [27]. In vitro studies on cultures of rat caecal con-
tents or Clostridium perfringens gave yields of 8–15 % of the metabolite (much
higher than that of the oxamic acid metabolite). In vivo studies employing oral
administration (200 mg/kg) of [14C]-labeled metronidazole to rats resulted in 1.3–
1.8 % of the dose being excreted in urine as acetamide with a further 0.9–2.4 %
eliminated via the feces. However, no such excretion of acetamide was obtained
with germfree rats.
A more recent example of the reductive metabolism of drugs involves the
anticonvulsant zonisamide (1,2-benzisoxazole-3-methanesulfonamide) [28]. This
drug has been shown to be mainly converted to 2-sulphamoylacetylphenol via the
reduction of the benzisoxazole ring. While mammalian liver enzymes can perform
this biotransformation, the role of the intestinal microbiota was also evaluated
in vivo in the rat. In this study it was shown that antibiotic treatment significantly
reduced the urinary and fecal excretion of the reduced metabolite following oral
dosing with the drug. When the animals were recolonized, the presence of the
microbiota restored the excretion of the metabolite. Further studies showed that the
caecal contents obtained from control rats possessed zonisamide reductase activity
that was absent from that of the antibiotic-treated rats. A number of strains of intes-
tinal bacteria were tested for the ability to reduce zonisamide with Clostridium spo-
rogenes showing the highest activity. On the basis of these studies, the authors
concluded that, in vivo, the gut microbiota played a major role in the production of
2-sulphamoylacetylphenol. These studies contrasted with the conclusions of earlier
in vitro studies that suggested that bacterial involvement in this metabolism, while
it did occur, was not particularly important quantitatively compared to that observed
in hepatic microsomes [29].
The in vitro reduction of the prodrug loperamide N-oxide to loperamide (used
for the symptomatic treatment of diarrhea) was investigated using gut microbiota,
gut contents, intestinal cells, and hepatocytes [30]. Efficient reduction was seen in
the caecal contents from rat, dog, and human, and this activity was greatly reduced
in the germfree rat.
Other losses of oxygen have been noted for the H2 receptor antagonists ranitidine
[31] and nitazidine [32], but not cimetidine or famotidine, as a result of the action
of the gut microbiota in vitro [32].
As well as the toxicity associated with the reduction of nitrazepam described
above, the reductive metabolism of the antibiotic chloramphenicol by the gut micro-
biota has been associated with often fatal, idiosyncratic, bone marrow aplasia [33].
The metabolite responsible was suggested as p-aminophenyl-2-amino-1,3-
propanediol [34]. This metabolite is only generated by a small percentage of patients
who take the drug orally and have a high percentage of the coliform bacteria that are
capable of metabolizing chloramphenicol to this metabolite. However, other metab-
olites responsible for this toxicity have been proposed [32, 34] including
p-nitrophenyl-2-dicloroacetamido-1,3-propanediol and 2-dichloroacetamid-3-
hydroxypropio-p-nitrophenone [35].
15.4 Demethylations, Dehydroxylations,

Deacylations, and Decarboxylations
Another well-known set of biotransformations of the gut microbiota is its ability to

undertake metabolic demethylations, decarboxylations, and dehydroxylations. The
demethylation of methamphetamine and 4′-hydroxymethamphetamine by the gut
microbiota was shown by Caldwell and Hawksworth in the guinea pig [36]. The
O- and N-demethylation of a range of compounds were explored by Smith and
Griffiths for a range of compounds, incubated with rat-derived microbiota, who
found that N-dealkylation did not occur for any of the drug molecules studied while
O-dealkylation occurred only for relatively simple aromatic compounds [37]. A
recent example of O-dealkylation is provided by studies in man on the metabolic
fate of N4-(2,2-dimethyl-3-oxo-4-pyrid[1, 4]oxazin-6-yl)-5-fluoro-N2-(3,4,5-
trimethyoxyphenyl)-2,4-pyrimidinediamine (R406) and its oral prodrug N4-(2,2-
dimethyl-4-[(dihydrogenphosphonoxy)methyl]-3-oxo-5-pyrid[1, 4]oxazin-6-yl)-5-
fluoro-N2-(3,4,5-trimethyoxyphenyl)-2,4-pyrimidinediamine disodium hexahydrate
(fostamatinib), a spleen tyrosine kinase inhibitor [38]. Metabolic studies performed
in vitro showed the prodrug to be rapidly converted to the active compound by
human intestinal microsomes. The active species was also the major drug-related
compound detected in human plasma following oral dosing. The bulk (ca. 80 %) of
a [14C]-dose of the drug was eliminated via the feces (with the remainder in urine)
where the drug and a unique 3,5-benzene diol metabolite were detected. This
3,5-benzene diol metabolite appeared to result from the O-demethylation and dehy-
droxylation of one of the systemic metabolites of the drug (designated as R529)
undertaken by anaerobic gut bacteria. This conclusion was based on the observation
that R529 was converted to the 3,5-benzene diol on the in vitro incubation of this
metabolite with human-derived feces. The experiments performed to investigate the
overall metabolism of the drug suggested that production of the 3,5-benzene diol
metabolite in man resulted from an initial hepatic cytochrome P450-mediated
p-O-demethylation of the drug followed by further gut microbiota O-demethylations

and dehydroxylation.
As well as their studies on O- and N-dealkylations described above, Smith and
Griffiths also investigated, in the same study [37], the in vitro gut microbial metabo-
lism of N-acetylated compounds and found that phenacetin, acetanilide, and
paracetamol, amongst others, all underwent significant amounts of N-deacylation.
Such reactions, which unmask the aniline (phenetidine and p-aminophenol in the
case of phenacetin and paracetamol, respectively), are clearly of interest with
respect to potential toxicity.
Another interesting example of dehydroxylation/decarboxylation concerns
L-DOPA (levodopa, L-3,4-dihydroxyphenylalanine) which is administered orally to
treat Parkinson’s disease. That the gut microbiota might be involved in the metabo-
lism of the drug was first suggested by studies by Sandler et al. [39, 40] who noted,
when treating patients suffering from Parkinson’s disease with L-DOPA, that the
urinary excretion of 3-hydroxyphenylacetic acid was increased. However, they also
observed that concentrations this metabolite were significantly reduced in quantity
by administration of neomycin suggesting that some microbial dehydroxylation of
dopamine or L-DOPA occurs in man. Studies in rats [41] showed that 3-hydroxy-
phenylacetic acid was only seen in the urine of conventional, but not germfree,
animals administered L-DOPA or dopamine. Both germfree and conventional ani-
mals were capable of converting m-tyramine to 3-hydroyphenylacetic acid suggest-
ing that the conversion of dopamine to m-tyramine is bacterial in origin. In another
study the microbial degradation of L-DOPA, 14C-DL-DOPA (and potential phenolic
metabolites) was investigated by means of incubations with rat caecal contents [42].
This work suggested that the biotransformation of DOPA was by way of 3,4-dihy-
droxyphenylacetic acid and decarboxylation or dehydroxylation to 4-methylcate-
chol or 3-hydroxyphenylacetic acid, respectively. In some incubations
3-hydroxyphenylacetic acid was further metabolized via decarboxylation to m-
cresol. 3-Hydroxy-phenylpropionic acid was also detected as a metabolite. This is
interesting as, to be effective, DOPA is decarboxylated following its passage across
the blood-brain barrier to give dopamine, and Peppercorn and Goldman [43] sug-
gested that this reaction could also be catalyzed by the gut microbiota which reduces
the amount of dopamine reaching the brain. Results obtained by examining drug
bioavailability after administration of L-DOPA to the dog gave average AUCs after
hepatoportal and IV administration that were essentially the same [44]. However,
when dosed duodenally, the AUC for L-DOPA was reduced, and this was combined
with an increase in that of dopamine. When paromomycin and kanamycin were
administered to suppress the gut microbiota, similar plasma and urinary excretion
profiles were observed between control and antibiotic-treated animals. Further
investigations of this decarboxylation determined the distribution of the decarboxyl-
ase activity of homogenates prepared from segments of the intestine. These showed
that the jejunum had the highest enzyme activity, followed by the ileum and duode-
num, leading the authors to conclude that, in the dog, the reduced bioavailability of
orally administered levodopa occurred as a consequence L-DOPA decarboxylation
in the gut wall rather than gut microbiota.
15.5 Effects of Bacterial Hydrolases

on Drug Action and Toxicity
One of the more obvious and widely known effects that the gut microbiota exert on
drug and xenobiotic metabolism, disposition, and toxicity relates to the effects of
the hydrolytic enzymes responsible for the deconjugation of glucuronides, gluco-
sides, etc. on both dietary pharmacologically active compounds and xenobiotic
metabolites. Investigations have shown that the activities of the β-glucosidase
and β-glucuronidase enzymes (as well as nitrate reductase and nitroreductase)
determined in Escherichia coli, Clostridium sp., Streptococcus sp., Bacteroides sp.,
and Lactobacillus salivarius differ between strains of gut bacteria. Thus, Clostridium
sp. showed the greatest enzymatic activity for β-glucosidase and β-glucuronidase
(and nitroreductase, with E. coli the most active for nitrate reductase). The study
was conducted both in vitro and in vivo and the authors found that, in general, the
in vivo activity of the enzymes in was highest (although there were instances when
this relationship was reversed) [45].
Effects on the bioavailability of natural products such as the flavone glucuronide
baicalin (present in Scutellariae radix), where prior hydrolysis to the aglycone
baicalein is followed by reconjugation following absorption [46], or effects on soy
isoflavones to produce phytoestrogens such as equol are well known [47, 48]. In a
similar way the effect of these gut microbiota-derived hydrolytic enzymes on the
disposition of drugs and their metabolites is readily apparent in the case of conju-
gated metabolites (e.g., glucosides, glucuronides, and sulfates) that, following their
production in the liver, are excreted in the bile. Hydrolysis by microbial enzymes
results in their being resorbed, and this enterohepatic recycling can result in the
modulation of the pharmacokinetic properties of the drug. In fact hydrolysis of
glucuronide metabolites by bacterial enzymes can be responsible for considerable
unwanted drug toxicity, and this has provided the impetus for the synthesis of specific
inhibitors of bacterial glucuronidase to prevent this from occurring. The first use of
this approach was applied to the DNA topoisomerase I inhibiting anticancer
drug irinotecan by Wallace et al. [49]. The dose-limiting side effect of the drug,
which is commonly used in the treatment of colon cancer, is the severe diarrhea that
results from the hydrolysis of an inactive glucuronide metabolite by bacterial
B-glucuronidases thereby reactivating the drug in the gut. By designing specific
bacterial B-glucuronidase inhibitors, that had no effect on the mammalian
B-glucuronidase, with crystal structures establishing that this selectivity was due to
interaction with a bacterial B-glucuronidase-specific loop on the enzyme, the authors
were able to eliminate irinotecan-induced toxicity in mice also orally administered
the inhibitor. More recently the same approach was demonstrated to have similar
effects on the small intestinal injury resulting from nonsteroidal anti-inflammatory
drug (NSAID)-induced toxicity [50, 51]. Many of these NSAIDs contain a carboxylic
acid and this represents a common site for glucuronidation, followed by excretion
via the bile. In the first example of this approach [52], the bacteria-specific
glucuronidase inhibitor (1-((6,8-dimethyl-2-oxo-1,2-dihydroquinolin-3-yl)-3-
(4-ethoxyphenyl)-1-(2-hydroxyethyl))thiourea) used with irinotecan was shown to

inhibit the hydrolysis of diclofenac acyl glucuronide. Control or inhibitor-pretreated
mice were administered an ulcerogenic dose of diclofenac and the effects noted. As
expected, without pretreatment with the inhibitor, numerous large ulcers were
detected in the small intestine while treated animals showed much reduced mucosal
injury and enteropathy. This suggests that it is the liberation of the drug (and ulcero-
genic metabolites) rather than the glucuronides themselves that is responsible for
gut toxicity. Perhaps surprisingly, given the potential to reduce enterohepatic recy-
cling, measurement of the pharmacokinetics of the drug with and without inhibitor
pretreatment showed no differences in systemic drug exposure to diclofenac.
Subsequently the same authors showed that the inhibitor was equally effective at
preventing NSAID-related damage when used when indomethacin, ketoprofen, and
diclofenac were also administered to mice. Interestingly they demonstrated that the
protective effects were maintained to some extent even if the inhibitor was adminis-
tered several hours after the drug diclofenac [53], which undergoes extensive
enterohepatic recirculation, had been dosed. These data are compatible with the
hypothesis that pharmacological inhibition of bacterial β-glucuronidase-mediated
cleavage of NSAID glucuronides in the small intestinal lumen can protect against
NSAID-induced enteropathy caused by locally high concentrations of NSAID
aglycones.
Another example of the potential of microbiome-driven hydrolysis to result in
serious consequences for patients is provided by the antiviral drug sorivudine
(1-β-D-arabinofuranosyl-5-(E)-(2-bromovinyl)uracil) used to combat infections
by varicella-zoster virus and herpes simplex virus type 1 [54]. Recent studies
have indicated that the drug should not be combined with anticancer drugs
such as 5-fluorouracil (5-FU) or prodrugs such as tegafur because a metabolite,
(E)-5-(2-bromovinyl)uracil (BVU), has been shown to cause the accumulation of
5-FU in the systemic circulation by inhibiting its degradation, resulting in
enhanced, sometimes fatal, toxicity. The mechanism behind this toxicity seems to
result from the inactivation of the hepatic dihydropyrimidine dehydrogenase
(DPD) enzyme by BVU. The production of BVU from sorivudine was ascribed to
bacterial phosphorolytic enzymes with high enzymatic activity detected in the caecal
and large intestinal contents of the rat [54]. The measurement of phosphorylase
activity performed on the cell-free extracts obtained from some 40 species of
microbes showed that high activity to convert sorivudine to BVU was present in
the Bacteroides species B. vulgatus, B. thetaiotaomicron, B. fragilis, B. uniformis,
and B. eggerthii. Further studies showed that in rats treated with ampicillin or a
mixture of bacitracin, neomycin and streptomycin, or metronidazole, only low
concentrations of BVU were detectable in serum. In contrast when the amounts of
BVU in serum were determined after the administration of kanamycin (used to
selectively decrease the number of aerobes, rather than anaerobes), they were
found to be higher in untreated animals. From these results the authors concluded
that BVU is produced in vivo from sorivudine by intestinal anaerobic bacteria
especially Bacteroides species.
15.6 Bacterial Acetylation
While many microbiome-driven biotransformations result in reduction or hydrolysis

of drugs and their metabolites, an interesting exception is bacterial acetylation via
bacterial N-acetyl transferases (NATs). Both N- and O-acetylation reactions have
been shown to occur, and these reactions have been highlighted as being important
in the bioactivation of genotoxic aromatic amines [55, 56]. In addition they are
involved in the metabolism of drugs such as 5-aminosalyclic acid. Thus following
oral dosing of 5-aminosalyclic acid, significant amounts of N-acetylated 5-amino-
salycylic acid are excreted via the feces. When the drug was incubated with feces,
using both aerobic and anaerobic conditions, this conjugation reaction was evident
and this activity was demonstrated in a number of species [57]. Indeed, the acetyla-
tion of the aminosalicylate isomers 5-aminosalicylic acid (5-ASA) and 4-aminosali-
cylic acid (4-ASA), together with that of p-aminobenzoic acid, has been investigated
in some detail, examining steady-state kinetics, time-dependent inhibition, and
DNA hybridization in some 40 bacterial species, mostly from the human intestinal
microbiota. This study showed N-acetyltransferase activity in 11 species of
Proteobacteriaceae from seven genera (the Citrobacter amalonaticus, Citrobacter
farmeri, Citrobacter freundii, Klebsiella ozaenae, Klebsiella oxytoca, Klebsiella
rhinoscleromatis, Morganella morganii, Serratia marcescens, Shigella flexneri,
Plesiomonas shigelloides, and Vibrio cholerae). The authors found that 5-amino-
salicylic acid was acetylated much more efficiently than the 4-amino isomer (27–645
times), while p-aminobenzoic acid was only poorly acetylated. Pseudomonas aeru-
ginosa was found to be the best acetylator with regard to both substrate spectrum
and catalytic efficiency [57].
There have been suggestions that the adverse side effects of drugs that form
5-aminosalyclic acid such as olsalazine and sulfasalazine, which can result in
pancreatitis in children, may be a result of the toxicity of the acetylated metabolite
[58].
15.7 Microbial Metabolism of Glutathione Conjugates
While not direct metabolism of the xenobiotics themselves, the role of the micro-
biota in further processing glutathione conjugates formed in the liver from reactive
metabolites, and then excreted in the bile, should be noted. Extensive metabolism
of such conjugates, of a range of agrochemicals, was reviewed by Bakke and
Gustafsson [59]. These studies showed the production of a large number of further
metabolites via the degradation of the glutathione moiety for the conjugates of, e.g.,
2-chloro-N-isopropylacetanilide (propachlor) and naphthalene, phenanthrene, etc.,
involving the bacterial C-S-lyases. Such metabolism has been shown to have the
potential to result in the almost complete removal of the glutathione part of the con-
jugate to produce a free thiol group. This type of reaction has also been shown for
drugs such as paracetamol with the thiol subsequently further metabolized to give
the methylthio metabolite of the drug [60]. The complete removal of the glutathione
conjugate to regenerate the parent compound has also been observed [59].
15.8 Miscellaneous Biotransformations
The microbiota have demonstrated a range of other biotransformation capabilities,

ether in vivo or in vitro, with regard to drug metabolism. For example, in the case of
levamisole, used extensively as an anthelmintic drug, in both animals and humans,
in vitro incubations under anaerobic conditions produced a number of ring-opened
metabolites of the thiazole ring system [61]. The metabolites were mainly produced
by the Bacteroides and Clostridia, and one of them, levametabol I, has been
proposed as active for anti-colon tumor activity [61]. Another example of ring open-
ing has been seen for the antipsychotic drug risperidone where cleavage of the
benzisoxazole ring, in both parent compound and various hydroxylated metabolites,
has been demonstrated both in vitro (caecal contents) and in vivo in the rat and
in vivo in the dog [62].
15.9 Microbiome-Conditional Effects and Consequences
In the preceding text we have summarized the direct effects that the bacteria that
comprise the gut microbiota can exert on the metabolism and toxicity of drugs, their
metabolites, and related xenobiotics. In all likelihood these reported effects repre-
sent the tip of the iceberg as gut microbiota metabolism of drugs is not routinely
assessed. However, the gut microbiota is not limited to direct effects and there are a
number of indirect mechanisms whereby the microbiome can affect the metabolism,
disposition, and toxicity of xenobiotics. Such effects include the modulation of host
metabolic enzymes/transporters, competition for metabolism via particular host
metabolic routes/enzymes, and enhancement of toxicity as a result of other effects
on host biochemistry.
So, effects on the complement of the various xenobiotic metabolizing capabili-
ties of important detoxification organs such as the gut and liver have been noted
with the modulation of the levels of both cytochrome P450s and conjugating enzyme
systems (e.g., [63–66]).
This includes the induction of P450s involved in the bioactivation of mutagens
including the heterocyclic aromatic amine [63] 2-amino-3-methylimidazo[4,5-f]
quinolone, amongst others. To date much of the information that we have on the
effects of the microbiome on drug-metabolizing systems comes from a comparison
of germfree animals with microbiome-competent controls. In a study examining
the effects of the colonization of germfree mice with either individual strains of
bacteria or complete ileal/caecal microbiota from conventionally raised mice,
DNA microarray analysis was used to determine the response of the intestine [64]
revealing a number of effects on xenobiotic metabolizing capabilities. Thus colo-
nization with Bacteroides thetaiotaomicron resulted in, amongst a range of other
changes, decreases in glutathione S-transferase (GST), CYP2D2 (also known as
debrisoquine hydroxylase), and the transporter “multidrug resistance protein 1a”
(Mdr 1a). However, colonization with other species of gut-dwelling microbes, such
as Escherichia coli and Bifidobacterium infantis, or indeed complete microbiota
resulted in different outcomes with either no change (conventional gut microbiota)
or increased expression of these enzymes. The authors concluded that these com-
mensal bacteria were capable of modulating the expression of a range of host genes
involved in a variety of “diverse and fundamental physiological functions” with the
selective effects seen as a result of the type of bacteria used for colonization reveal-
ing how changes in the composition of the gut microbiome could have significant
physiological effects.
In another study [65] the amounts of a number of enzymes involved in the Phase II
conjugation of drugs and other xenobiotics were determined in livers and the small
intestine, caecum, and colon of both germfree rats and those inoculated with micro-
biota from normal rats. Effects were noted on enzymes involved in glutathione
metabolism such as GST glutathione peroxidase (GPX2). In addition differences in
the amounts of epoxide hydrolases (EPHXs) and N-acetyltransferases (NAT) 1 and
sulfotransferases (SULTs) were seen. Some of these enzymes were expressed in
both liver and gut, while others showed regional differences (e.g., the SULTs were
found in liver and large, but not the small intestine) and, in addition, gender effects
were noted for the liver, but not gut, for a number of conjugating enzymes. Thus
hepatic SULT1A1, SULT1C1, and SULT1C2 were seen to be elevated in germfree
animals in both male and females (1.5- to 2.6-fold) while hepatic EPHX2 was
1.6-fold higher in female rats. The colonic germfree rats showed large differences
compared to normal animals with GSTA1/2 4.0- and 5.0-fold higher in males and
females, respectively, GSTA4 between 1.5 and 1.9-fold higher, and GSTM1
elevated by 1.1/1.5-fold. The epoxide hydrolases, EPHX1 and EPHX2, were
3.5/2.4- and 1.4/2.1-fold higher in male and female germfree rats, respectively. In
the case of the sulfotransferases SULT1B1 and SULT1C2, the increases were
0.4/0.6- and 1.3/1.6-fold and for NAT2 amounted to 1.4/1.5-fold for male and
female germfree rats, respectively.
When human gut microbiota was used for recolonization, the effects on the
expression of such enzymes in the colon were smaller than those seen for recoloni-
zation with rat gut microbiota.
Further literature examples also show that germfree rats, and those inoculated
with human gut microbiota, can exhibit differences in glucuronidating (UGT) and
GST enzymes when dosed with (+)-catechin or (-)-epicatechin. In addition, animals
with human gut microbiota had reduced CYP2C11 induction compared to the
germfree animals [66].
Other gut microbial-conditional effects resulting from exposure to soy-derived
phytoestrogens on host endogenous steroid (and thereby potentially xenobiotic)
metabolism via a reduction in the excretion of 4-hydroxyestrogen and increased
2-hydroxyestrogen have been noted in postmenopausal women [67]. The authors

suggested but did not actually demonstrate that this might result from changes in the
expression of the CYPs involved in estrogen hydroxylation. Were this the case it
would also be expected that there would also be knock on effects on the metabolism
of drugs and other xenobiotics as well drug transporters.
Other indirect effects could easily result from competition between xenobiotics
and microbial metabolites for metabolic pathways such as sulfation, often preferen-
tially used for the conjugation of phenolic hydroxyls, but where capacity is limited.
This appears to be the case for metabolism of compounds such as paracetamol [68].
For example, a study in human subjects administered 1 g of paracetamol indicated
that the ratio of glucuronide to sulfate metabolites excreted in the urine, which var-
ies considerably between individuals, was affected by competition for sulfation by
microbially generated p-cresol (produced by bacterial metabolism of tyrosine and
phenylalanine). If large amounts of p-cresol are present that compete with
paracetamol for sulfation, there is the potential to reduce the ability of the host to
detoxify the drug with potentially adverse consequences by directing metabolism
toward the pathways that result in the production of the reactive quinone imine.
Effects on the pharmacokinetics of orally administered paracetamol have been
noted in a recent study in animals treated with antibiotics to eliminate the gut micro-
biota using a cocktail of bacitracin, streptomycin, and neomycin [69]. Paracetamol
was dosed to both control and antibiotic-treated rats, with plasma concentrations of
paracetamol and six metabolites determined via LC-MS/MS. The authors noted that
the AUCs of the drug and its glutathione conjugate were higher in the antibiotic-
treated animals while the metabolic efficiency of sulfate conjugation, as indicated
by the ratio of the AUCs of paracetamol sulfate to paracetamol, was lower in
antibiotic-treated compared to control rats (surprising given the induction of SULTs
seen in the germfree animals described above) [65]. Irrespective of the reasons for
the decreased sulfation seen in these pseudo-germfree animals, this study provides
another example of the gut microbiota, or its absence, affecting drug conjugation.
Effects on toxicity have also been noted between normal and germfree rats, even
when metabolism seems relatively unaffected, for the model hepatotoxin hydrazine.
Thus, when administered to germfree rats at what was effectively a no effect dose
level in controls, hydrazine resulted in profound effects, with some of the test animals
becoming moribund [70]. Clearly, although an extreme example, this study demon-
strates the value of a functional gut microbiome with respect to toxicity and suggests
that different microbiomes might provide more subtle modulations of toxicity.
Such effects have obvious implications for drug toxicity testing, where differences
in outcome may reflect not only strain but microbiome. Clearly, such effects also
have the potential to produce unexpected, and potentially unwelcome, variability in
response to the administration of drugs and exposure to toxins between individual
patients and populations (and indeed some limited evidence suggests that there are
microbially driven differences in drug metabolism between populations e.g., [23]).
However, despite a resurgence in interest in this “forgotten organ” (evidenced by
a number of recent reviews, e.g., [71–76]) there is currently little, if any, real con-
sideration of the potential of this to affect the various phases of drug absorption,
disposition, metabolism, excretion, pharmacology, or toxicity in either discovery or

drug development programs. Similarly there is little evidence that regulatory bodies
are aware of the potential importance of the gut microbiome. This is, potentially, a
serious oversight and in our view a better understanding of these complex interac-
tions could provide novel insights for drug discovery and development, and signifi-
cant benefits for personalized medicine. The microbiome undoubtedly represents a
“druggable target,” and there is no doubt that it is possible to modulate both its
composition and metabolic activity. It clearly deserves more attention from the drug
metabolism community.
References
1. Illing HPA. Techniques for microfloral and associated metabolic studies in relation to the
absorption and enterohepatic circulation of drugs. Xenobiotica. 1981;11:815–30.
2. Boxenbaum HG, Bekersky I, Jack MJ, Kaplan SA. Influence of gut microflora on bioavail-
ability. Drug Metab Rev. 1979;9:259–79.
3. Nicholson JK, Wilson ID. Understanding ‘global’ systems biology: metabonomics and the
continuum of metabolism. Nat Rev Drug Discov. 2003;2:668–76.
4. Gingel R, Bridges JW. Intestinal azo-reduction and glucuronide conjugation of prontosil.
Xenobiotica. 1973;9:599–604.
5. Gingel R, Bridges JW, Williams RT. The role of the gut flora in the metabolism of prontosil and
neoprontosil in the rat. Xenobiotica. 1971;1:143–56.
6. Rafii F, Cerniglia CE. Reduction of azo dyes and nitroaromatic compounds by bacterial
enzymes from the human intestinal tract. Environ Health Perspect. 1995;103 Suppl 5:17–9.
7. Peppercorn MA, Goldman P. The role of intestinal bacteria in the metabolism of salicylazosul-
fapyridine. J Pharmacol Exp Ther. 1972;181:151–62.
8. Schroder H, Gustafsson BE. Azo reduction of salicyl-azosulphapyridine in germ free and con-
ventional rats. Xenobiotica. 1973;3:225–31.
9. Truelove SC. Evolution of olsalazine. Scand J Gastroenterol. 1988;23(s148):3–6.
10. Chan RP, Pope DJ, Gilbert AP, Sacra PJ, Baron JH, Lennard-Jones JE. Studies of two novel
sulfasalazine analogs, ipsalazide and balsalazide. Dig Dis Sci. 1983;28:609–15.
11. Takeno S, Sakai T. Involvement of the intestinal microflora in nitrazepam-induced teratogenic-
ity in rats and its relationship to nitroreduction. Teratology. 1991;44:209–14.
12. Elmer GW, Remmel RP. Role of the intestinal microflora in clonazepam metabolism in the rat.
Xenobiotica. 1984;14:829–40.
13. Antila S, Huuskonen H, Nevalainen T, Kanerva H, Vanninen P, Lehtonen L. Site dependent
bioavailability and metabolism of levosimendan in dogs. Eur J Pharm Sci. 1999;9:85–91.
14. Antila S, Pesonen U, Lehtonen L, Tapanainen P, Nikkanen H, Vaahtera K, Scheinin
H. Pharmacokinetics of levosimendan and its active metabolite OR-1896 in rapid and slow
acetylators. Eur J Pharm Sci. 2004;23:213–22.
15. Deng Y, Rogers M, Sychterz C, Talley K, Qian Y, Bershas D, Ho M, Chen EP, Shi W, Chen EP,
Serabjit-Singh C, Goryckj PD. Investigations of hydrazine cleavage of eltrombopag in humans.
Investigations of hydrazine cleavage of eltrombopag in humans. Drug Metab Dispos.
2011;39:1747–54.
16. Strong HA, Renwick AG, George CF, Liu YF, Hill MJ. The reduction of sulphinpyrazone and
sulindac by intestinal bacteria. Xenobiotica. 1987;17:685–96.
17. Watanabe K, Yamashita S, Furruno K, Kawasaki H, Gomita Y. Metabolism of omeprazole by
gut flora in rats. J Pharm Sci. 1995;84:516–7.
18. Lindenbaum J, Tse-Eng D, Butler Jr VP, Rund DG. Urinary excretion of reduced metabolites
of digoxin. Am J Med. 1981;71(1):67–74.
19. Dobkin JF, Saha JR, Butler Jr VP, Neu HC, Lindenbaum J. Inactivation of digoxin by
Eubacterium lentum, an anaerobe of the human gut flora. Trans Assoc Am Physicians.
1982;95:22–9.
20. Saha JR, Butler Jr VP, Neu HC, Lindenbaum J. Digoxin-inactivating bacteria: identification in
human gut flora. Science. 1983;220:325–7.
21. Bennet RG, Beamer BA, Greenough WB, Lindenbaum J, Bartlett JG. Colonisation with
digoxin reducing strains of Eubacterium lentum and Clostridium difficile infection in nursing
home patients. J Diarrhoeal Dis Res. 1992;10:87–92.
22. Linday L, Dobkin JF, Wang TC, Butler Jr VP, Saha JR, Lindenbaum J. Digoxin inactivation by
the gut flora in infancy and childhood. Pediatrics. 1987;79:544–8.
23. Mathan VI, Wiederman J, Dobkin JF, Lindeenbaum J. Geographic differences in digoxin inac-
tivation, a metabolic activity of the human anaerobic gut flora. Gut. 1989;30:971–7.
24. Haiser HJ, Gootenberg DB, Chatman K, Sirasani G, Balskus EP, Turnbaugh PJ. Predicting and
manipulating cardiac drug inactivation by the human gut bacterium Eggerthella lenta. Science.
2013;341:295–8.
25. Koch RL, Chrystal EJT, Beaulieu Jr BB, Goldman P. Acetamide a metabolite of metronidazole
formed by the intestinal flora. Biochem Pharmacol. 1979;28:3611–5.
26. Koch RL, Beaulieu BB, Goldman P. Role of the intestinal flora in the metabolism of misoni-
dazole. Biochem Pharmacol. 1980;29:3281–4.
27. Koch RL, Goldman P. The anaerobic metabolism of metronidazole forms n-(2-hydroxyethyl)-
oxamic acid. J Pharmacol Exp Ther. 1979;208:406–10.
28. Kitamura S, Sugihara K, Kuwasako M, Tatsumi K. The role of mammalian intestinal bacteria
in the reductive metabolism of zonisamide. J Pharm Pharmacol. 1997;49:253–6.
29. Stiff DD, Robicheau JT. Reductive metabolism of the anticonvulsant agent zonisamide, a
1,2-benzisoxazole derivative. Xenobiotica. 1992;22:1–11.
30. Lavrijsen K, Van Dyck D, Van Houdt J, Hendrickx J, Monbaliu J, Woestenborghs R,
Meuldermans W, Heykants J. Reduction of the prodrug loperamide oxide to its active drug
loperamide in the gut of rats, dogs, and humans. Drug Metab Dispos. 1995;23:354–62.
31. Basit AW, Lacey LF. Colonic metabolism of ranitidine; implications for its delivery and
absorption. Int J Pharm. 2001;227:157–65.
32. Basit AW, Newton JM, Lacey LF. Susceptibility of the H2-receptor antagonists cimetidine,
famotidine and nizatidine, to metabolism by the gastrointestinal microflora. Int J Pharm.
2002;237:23–33.
33. Yunis AA. Chloramphenicol toxicity: 25 years of research. Am J Med. 1989;87(3N):44N–8.
34. Holt R. The bacterial degradation of chloramphenicol. Lancet. 1967;289(7502):1259–1260.
35. Jimenez JJ, Arimura GK, Abou-Khalil WH, Isildar M, Junis AA. Chloramphenicol-induced
bone marrow injury: potential role of bacterial metabolites of chloramphenicol. Blood.
1987;70:1180–5.
36. Caldwell J, Hawksworth GM. The demethylation of methamphetamine by intestinal micro-
flora. J Pharm Pharmacol. 1973;25:422–4.
37. Smith GE, Griffiths LA. Metabolism of N-acylated and O-alkylated drugs by the intestinal
microflora during anaerobic incubation in vitro. Xenobiotica. 1974;4:477–87.
38. Sweeny DJ, Li W, Clough J, Bhamidipati S, Singh R, Park G, Baluom M, Elliott G, Lau
DT-W. Metabolism of fostamatinib, the oral methylene phosphate prodrug of the spleen tyro-
sine kinase inhibitor R406 in humans: contribution of hepatic and gut bacterial processes to the
overall biotransformation. Drug Metab Dispos. 2010;38:1166–76.
39. Calne DB, Karoum F, Ruthven CRJ, Sandler M. The metabolism of orally administered
L-DOPA in Parkinsonism. Br J Pharmacol. 1969;37:57–68.
40. Sandler M, Goodwin BL, Ruthven CRJ, Calne DB. Therapeutic implications in Parkinsonism
of m-tyramine formation from L-Dopa in man. Nature. 1971;229:414–6.
41. Bakke OM. Degradation of DOPA by intestinal microorganisms in vitro. Acta Pharmacol
Toxicol. 1971;30:115–21.
42. Goldin BR, Peppercorn MA, Goldman P. Contribution of host and intestinal microflora in the
metabolism of L-Dopa by the rat. J Pharmacol Exp Ther. 1973;186:160–6.
43. Peppercorn MA, Goldman P. Drug-bacteria interaction. Rev Drug Metab Drug Interact.
1976;11:75–88.
44. Sasahara K, Nitanai T, Habara T, Kojima T, Kawahara Y, Morioka T, Nakajima E. Dosage form
design for improvement of bioavailability of levodopa IV: possible causes of low bioavailabil-
ity of oral levodopa in dogs. J Pharm Sci. 1981;70:730–73.
45. Cole CB, Fuller R, Mallet AK, Rowland IR. The influence of the host on expression of intesti-
nal microbial enzyme activities involved in metabolism of foreign compounds. J Appl
Bacteriol. 1985;59:549–53.
46. Akao T, Kawabata K, Yanagisawa E, Ishiahara K, Mizuhara Y, Wakuia Y, Sakashita Y, Koashi
K. Baicalin, the predominant flavone glucuronide of scutellariae radix, is absorbed from the rat
gastrointestinal tract as the aglycone and restored to its original form. J Pharm Pharmacol.
2000;52:1563–8.
47. Bowey E, Adlercreutz H, Rowland I. Metabolism of isoflavones and lignans by the gut micro-
flora: a study in germ-free and human flora associated rats. Food Chem Toxicol.
2003;41:631–6.
48. Turner NJ, Thomson BM, Shaw IC. Bioactive isoflavones in functional foods: the importance
of gut microflora and bioavailability. Nutr Rev. 2003;61:204–13.
49. Wallace BD, Wang H, Lane KT, Scott JE, Orans J, Koo JS, Venkatesh M, Jobin C, Yeh L-AL,
Mani S, Redinbo MR. Alleviating cancer drug toxicity by inhibiting a bacterial enzyme.
Science. 2010;330:831–5.
50. LoGuidice A, Wallace BD, Bendel L, Redinbo MR, Boelsterli UA. Pharmacologic targeting of
bacteria-glucuronidase alleviates nonsteroidal anti-inflammatory drug-induced enteropathy in
mice. J Pharmacol Exp Ther. 2012;341:447–54.
51. Saitta KS, Carmen Z, Lee KK, Fujimoto K, Redinbo MR, Boelsterli UA. Bacterial
β-glucuronidase inhibition protects mice against enteropathy induced by indomethacin, keto-
profen or diclofenac: mode of action and pharmacokinetics. Xenobiotica. 2014;44:28–35.
52. Nakayama H, Kinouchi T, Kataoka K, Akimoto S, Matsuda Y, Ohnishi Y. Intestinal anaerobic
bacteria hydrolyse sorivudine, producing the high blood concentration of 5-(E)-(2-bromovinyl)
uracil that increases the level and toxicity of 5-fluorouracil. Pharmacogenetics.
1997;7:35–43.
53. Okuda H, Ogura K, Kato A, Takubo H, Watabe T. A possible mechanism of eighteen patient
deaths caused by interactions of sorivudine, a new antiviral drug, with oral 5-fluorouracil pro-
drugs. J Pharmacol Exp Ther. 1998;287:791–9.
54. Grant DM, Josephy PD, Lord HL, Morrison LD. Salmonella typhimurium strains expressing
human arylamine N-acetyltransferases: metabolism and mutagenic activation of aromatic
amines. Cancer Res. 1992;52:3961–4.
55. Hein DW, Doll MA, Rustan TD, Gray K, Feng Y, Ferguson RJ, Grant DM. Metabolic activa-
tion and deactivation of arylamine carcinogens by recombinant human NAT1 and NAT2 acet-
yltransferases. Carcinogenesis (Land). 1993;14:1633–8.
56. van Hogezand RA, Kennis HM, van Schaik A, Koopman JP, van Hees PA, van Tongeren
JH. Bacterial acetylation of 5-aminosalicylic acid in faecal suspensions cultured under aerobic
and anaerobic conditions. Eur J Clin Pharmacol. 1992;43:189–92.
57. Delome’nie C, Fouix S, Longuemaux S, Brahimi N, Bizet C, Picard B, Denamur F, Dupreet
J-M. Identification and functional characterization of arylamine N-acetyltransferases in eubac-
teria: evidence for highly selective acetylation of 5-aminosalicylic acid. J Bacteriol.
2001;183:3417–27.
58. Garau P, Orenstein SR, Neigut DA, Kocoshis SA. Pancreatitis associated with olsalazine and
sulfasalazine in children with ulcerative colitis. J Pediatr Gastroenterol Nutr. 1994;18:481–5.
59. Bakke JE, Gustafsson JA. Role of intestinal flora in metabolism of agrochemicals conjugated
with glutathione. Xenobiotica. 1986;16:1047–56.
60. Mikov M, Caldwell J, Dolphin CT, Smith RL. The role of intestinal microflora in the formation
of the methylthio adduct metabolites of paracetamol. Studies in neomycin-pretreated and
germ-free mice. Biochem Pharmacol. 1988;37:1445–9.
61. Shu YZ, Kingston DGI, Van Tassall RL, Wilkins TD. Metabolism of levamisole, an anti-colon
cancer drug by human intestinal bacteria. Xenobiotica. 1992;21:737–50.
62. Meuldermans W, Hendrickx J, Mannens G, Lavrijsen K, Janssen C, Bracke J, Le Jeune L,
Lauwers W, Heykants J. The metabolism and excretion of risperidone after oral administration
to rats and dogs. Drug Metab Dispos. 1994;22:129–38.
63. Rowland IR, Mallet AK, Cole CB, Fuller R. Mutagen activation by hepatic fractions from
conventional, germ free and monoassociated rats. Arch Toxicol. 1987;11(Suppl):261–3.
64. Hooper LV, Wong MH, Thelin A, Hansson L, Falk PG, Gordon JI. Molecular analysis of com-
mensal host-microbial relationships in the intestine. Science. 2001;291:881–3.
65. Meinl W, Sczesny S, Brigelius-Flohé R, Blaut M, Glatt H. Impact of gut microbiota on intes-
tinal and hepatic levels of phase 2 xenobiotic-metabolizing enzymes in the rat. Drug Metab
Dispos. 2009;37:1179–86.
66. Lhoste EF, Ouriet V, Bruel S, Flinois J-P, Brezillon C, Magdalou J, Cheze C, Nugon-Baudon
L. The human colonic microbiota influences the alterations of xenobiotic-metabolizing
enzymes by catechins in male F344 rats. Food Chem Toxicol. 2003;41:695–703.
67. Xu X, Duncan AM, Wangen KE, Kurzer MS. Soy consumption alters endogenous estrogen
metabolism in postmenopausal women. Cancer Epidemiol Biomark Prev. 2000;9:781–6.
lism. Proc Natl Acad Sci U S A. 2009;106:14728–33.
69. Lee SH, An JH, Lee HJ, Jung BH. Evaluation of pharmacokinetic differences of acetamino-
phen in pseudo germ-free rats. Biopharm Drug Dispos. 2012;33:292–303.
70. Swann J, Wang Y, Abecia L, Costabile A, Tuohy K, Gibson G, Roberts D, Sidaway J, Jones H,
Wilson ID, Nicholson J, Holmes E. Gut microbiome modulates the toxicity of hydrazine: a
metabonomic study. Mol Biosyst. 2009;5:351–5.
71. Kang MJ, Kim HG, Kim JS, Oh DG, Um YJ, Seo CS, Han JW, Cho HJ, Kim GH, Jeong TC,
Jeong HG. The effect of gut microbiota on drug metabolism. Expert Opin Drug Metab Toxicol.
2013;0:1–14.
72. Haiser HJ, Turnbaugh PJ. Is it time for a metagenomic basis of therapeutics? Science.
2012;336:1253–5.
73. Ursell LK, Knight R. Xenobiotics and the human gut microbiome: metatranscriptomics reveal
the active players. Cell Metab. 2013;17:317–8.
74. Li H, Jia W. Cometabolism of microbes and host: implications for drug metabolism and drug-
induced toxicity. Clin Pharmacol Ther. 2013;94:574–81.
75. Saad R, Rizkallah MR, Aziz RK. Gut Pharmacomicrobiomics: the tip of an iceberg of complex
interactions between drugs and gut-associated microbes. Gut Pathog. 2012;4:16.
76. Jeong HG, Kang MJ, Kim HG, Oh DG, Kim JS, Lee SK, Jeong TC. Role of intestinal micro-
flora in xenobiotic‐induced toxicity. Mol Nutr Food Res. 2013;57:84–99.
Chapter 16
Understanding the Benefits of Bariatric
Surgery on Gut Physiology: Implications
for Obesity, Type 2 Diabetes,
and Cardiovascular Disease
Steven K. Malin and John P. Kirwan
Abstract Seminal discoveries in the bariatric surgery field have revealed a remarkable
link between gastrointestinal physiology and obesity, a link that extends to remission
of many metabolic diseases including type 2 diabetes and to risk of cardiovascular
disease. Much of the beneficial health effects of bariatric surgery can be ascribed to
weight loss. However, in the case of type 2 diabetes, resolution is so acute that it
appears to be independent of the weight loss. These observations have created
the intriguing scenario whereby altering gastrointestinal anatomy creates rapid
physiological adaptations that manifest in normalization of glucose homeostasis.
The cellular and molecular mechanisms that produce these favorable health changes
are an area of intense scientific investigation. One leading hypothesis suggests that
rerouting nutrient flow to the gut alters enteroendocrine signals and bile acid
secretion that favors appetite suppression; increased energy expenditure and insulin
action; and decreased inflammation, blood lipids, and hypertension. In addition, the gut
microbiome has emerged as an area of particular interest with a focus on bacteria
and metabolites that interact to influence weight regulation and metabolic health.
S.K. Malin
Department of Pathobiology, Lerner Research Institute, Cleveland Clinic Foundation,
9500 Euclid Ave (NE40), Cleveland, OH 44195, USA
J.P. Kirwan, Ph.D. (*)
Department of Pathobiology, Lerner Research Institute, Cleveland Clinic Foundation,
9500 Euclid Ave (NE40), Cleveland, OH 44195, USA
Department of Nutrition, School of Medicine, Case Western Reserve University,
Cleveland, OH 44195, USA
Metabolic Translational Research Center, Endocrine and Metabolism Institute,
Cleveland Clinic, Cleveland, OH 44195, USA

DOI 10.1007/978-1-4471-6539-2_16
344 S.K. Malin and J.P. Kirwan
The collective evidence presented herein provides strong support for an essential
role for the gastrointestinal tract as a modulator of obesity-related disease. The gut
is now considered a novel endocrine organ and a therapeutic target for prevention
and treatment strategies that will benefit human health.
Keywords Appetite-hormones • Bariatric surgery • Bile acids • Cardiovascular

disease • Diabetes • Glucose • Gut microbiome • Incretins • Inflammation • Insulin
resistance • Microbiome • Nutrition • Obesity • Weight loss
16.1 Overview of Obesity
Obesity is a major public health concern because it is associated with increased

morbidity and mortality. The World Health Organization defines overweight as a
body mass index (BMI) of 25 kg/m2 or more and obesity as a BMI of >30 kg/m2.
Obese patients are further characterized into class I (BMI 30–34.9 kg/m2), class II
(BMI 35–39.9 kg/m2), and class III (BMI >40 kg/m2). These subcategories are
particularly important when considering disease risk, since chronic metabolic disease
risk increases progressively from a BMI >20 kg/m2 [1].
According to the National Health and Nutrition Examination Surveys (NHANES)
in the United States, nearly 78 million (or 35.7 %) adults and 12.5 million (16.9 %)
children/adolescents are obese. Although the prevalence of obesity in adults aged
20–74 years has more than doubled over the last 40 years (13.4 % in 1960–1962 vs.
35.1 % in 2005–2006), it appears to have plateaued in the last several years [2].
Unfortunately, it has become clear that the distribution among obesity status in
adults has shifted, such that a change in prevalence of superobesity (>50 BMI kg/m2
or class IV) has occurred from 0.9 % in 1960–1962 to 6.2 % in 2005–2006 [3].
Obesity is responsible for more than 2.8 million deaths worldwide per year,
owing to an increased prevalence of related comorbidities, including hypertension,
heart disease, stroke, back and lower extremity weight-bearing degenerative
problems, cancer, and type 2 diabetes [4]. Moreover, obesity is an independent risk
factor for death, and some reports indicate that there is a 20–40 % increase in
mortality in those who are overweight and upward of 300 % increased risk among
those who are obese [5]. Lifestyle modification, consisting of a combination of
nutrition, physical activity, and behavioral modification, is the first-line approach to
promote weight loss. Although these weight loss interventions typically show initial
promise, the long-term ability to maintain the desired weight benefit is typically
lost even with use of pharmacotherapy. In fact, most patients who lose weight via
behavioral changes and anti-obesity drugs tend to gain the weight back. The only
known medical treatment for severe obesity that produces durable effects on body
weight is bariatric surgery. In 1991, the National Institutes of Health established
guidelines for surgical therapy for morbid obesity (BMI >40 kg/m2 or BMI >35 kg/m2
in the presence of two or more comorbidities) [6]. In fact, bariatric surgery has
proven efficacy in not only ameliorating type 2 diabetes and cardiovascular disease
16 Understanding the Benefits of Bariatric Surgery on Gut Physiology… 345
Fig. 16.1 Effects of bariatric surgery on physiological systems and disease states in obesity
risk factors but also mitigates non-alcoholic fatty liver disease, sleep apnea, and
gastroesophageal reflux disorder (GERD) as well as reducing mortality (Fig. 16.1).
These effects on multiple clinical conditions make bariatric surgery an important
treatment option for many obese individuals.
This chapter examines the impact of bariatric surgery on gut physiology. There is
increasing evidence that alterations in the gut after bariatric surgery not only con-
tribute to the long-term weight loss maintenance but also to the treatment of type 2
diabetes and cardiovascular disease risk factors in patients with a BMI >35 kg/m2.
Central to the improvement in body weight is the increase in gut hormones known
to suppress appetite. Special attention is also given to gut microbiome-related
mechanisms that appear to be altered by bariatric surgery and may contribute to
weight loss. The chapter also discusses efficacy of the bariatric surgery as a thera-
peutic modality to improve glycemic control and cardiometabolic health in relation
to altered gut physiology and low surgical risk.
16.2 Bariatric Surgery Overview
Bariatric surgery is an effective therapy improving weight loss and metabolic health
and is currently recommended for adults with a BMI of at least 40 kg/m2 or 35 kg/m2
with comorbidities. Although bariatric procedures are commonly referred to as
restrictive and/or malabsorptive (Fig. 16.2) based on the presumed mechanism of
weight loss [7], it is worth noting that not all bariatric surgeries produce the same
effect on body weight, diabetes remission, and cardiometabolic resolution (Table 16.1).
Fig. 16.2 Conventional bariatric surgery operations. (a) Laparoscopic adjustable gastric banding.
(b) Sleeve Gastrectomy. (c) Roux-en-Y gastric bypass
Table 16.1 Metabolic effects of conventional bariatric techniques [19]

Rates of improvement after surgery (%)
Improvement LAGB RYGB BPD
Excess weight loss 46.2 59.5 63.3
Resolution of type 2 diabetes 56.7 80.3 95.1
Remission of dyslipidemia 59 97 99
Resolution of hypertension 43 68 83
Operative mortality 0.1 0.5 1.1
LAGB laparoscopic adjustable gastric banding, RYGB Roux-en-Y gastric bypass, BPD biliopan-
creatic diversion
Restrictive procedures decrease the functional volume of the stomach, thereby

increasing satiety due to limiting the intake of calories. These procedures often
involve some form of material or “band” (i.e., laparoscopic adjustable gastric banding
(LAGB)) and/or surgically resizing of the stomach with a stapler to create a small
pouch (i.e., vertical gastroplasty (VBG) or sleeve gastrectomy (SG)) [8]. In LAGB,
an adjustable plastic and silicone band is placed around the upper stomach to reduce
the size of the channel between the proximal and distal stomach to reduce food
consumption [9]. In VBG, a procedure not performed routinely any longer, part of the
stomach is permanently stapled to create a smaller pouch along the lesser curvature
of the stomach. Although the risk of death and major morbidity is low following
LAGB or VBG, the amount of excess weight loss obtained is inferior compared to
SG or malabsorptive procedures. After LAGB or VGB, patients lose nearly 46 % of
their excess body weight, while over half experience type 2 diabetes remission.
Alternatively, SG is a relatively new surgical approach for obesity management.
This nonreversible procedure involves resection of the greater curvature of the
stomach by stapling it over a sizing tube 11–20 mm in diameter [10]. Although the
effectiveness of SG with respect to weight loss and resolution of comorbidities is
less than that of Roux-en-Y gastric bypass (RYGB), but greater than LAGB, SG is
being used with increasing frequency due to fewer complications and reduced risk
(SG accounted for 7.8 % of primary bariatric operations in 2010) [11]. Current
advancements in restrictive procedures performed endoscopically have also the
potential to expand bariatric intervention for weight loss management. In fact,
endoluminal sleeves and intragastric balloons have demonstrated short-term benefit
for weight loss and comorbidity improvements, although long-term, randomized
trials are lacking [12].
Malabsorptive procedures are designed to reduce the area of intestinal mucosa
available for nutrient absorption and restrict caloric intake similar to LAGB or
SG. However, because the small intestine is shortened, they have added the compo-
nent of malabsorption of fat and nutrients. Afterward, more patients experience
remission of type 2 diabetes (82–99 %) compared with restrictive operations, even
in patients with longer duration of disease, including those treated with insulin
(Table 16.1). In biliopancreatic diversion with duodenal switch (BPD-DS), part of
the stomach is resected, and the duodenum is cut just distal to the pylorus and
Table 16.2 Changes in digestive and gut physiology after bariatric surgery [73]
LAGB SG RYGB
Masticating time Higher Higher Higher
Food intake amount Lower Lower Lower
Food transit time Slowed No change Faster
Food preferences More pureed/ No change Decrease high
less fiber fat/sugary food
Acid production No change Lower Decreased
Ghrelin No change Lower Lower
GLP-1 and PYY No change No change Increased
LAGB laparoscopic adjustable gastric banding, RYGB Roux-en-Y gastric bypass, SG sleeve
gastrectomy
reattached to the ileum, bypassing the duodenum and jejunum (i.e., digestive limb) [8].
The bypassed duodenum and jejunum (i.e., biliopancreatic limb) only pass bile
and pancreatic juices. These loops converge at a common channel at the end of the
small intestine, and the contents then pass normally through the large intestine.
Theoretically, although BPD provides a more physiologic digestive behavior and
diminishes the risk of dumping syndrome, ulcerogenicity, and hypocalcaemia, the
procedure is labor intensive and places the patient at high risk [13]. On the other
hand, RYGB is considered the gold standard for bariatric surgery and is the most
commonly performed operation [7, 13]. The procedure involves creating a gastric
pouch, Roux limb (jejunum and ileum), and biliary limb. The small gastric pouch is
connected to the mid-jejunum, bypassing the majority of the stomach, the entire
duodenum, and part of the proximal jejunum. Similar to the BPD-DS, the biliary
limb is reconnected approximately 150 cm distal relative to the pylorus so that
bilo-pancreatic juices can facilitate digestion. After RYGB, the size of the pouch
greatly limits the amount of food that can be eaten [7, 13].
Collectively, the efficacy of weight loss will differ depending upon the type of
surgery. Further, the speed at which type 2 diabetes remits also varies with restrictive
versus malabsorptive procedures (Table 16.2). For instance, after RYGB and BPD-DS,
diabetes remits within days, even before the patients have lost much weight, while
this does not occur after restrictive procedures [14]. Subsequently, newer restrictive
techniques are currently under investigation to improve personalized approaches
that best fit patients’ weight loss and comorbidity resolution needs. The current
short-term research in assessing endoluminal sleeves and intragastric balloons on
weight regulation and metabolic health improvements appears promising, although
longer-term follow-up data are needed [15].
16.2.1 Effects of Bariatric Surgery on Obesity
Weight loss success following bariatric surgery has been described by 50–75 %
excess weight loss (EWL), 20–30 % initial weight loss, and achieving a BMI
<35 kg/m2 [16]. The largest, prospective interventional-based trial that examined
the effects of bariatric surgery (i.e., LAGB vs. VBG vs. RYGB) was on 4,047 obese
patients with healthy matched treated control and is known as the Swedish Obesity
Study [17]. The results demonstrated that, while the control group gained weight
over time, the surgical groups lost on average 23 %, 17 %, and 18 % body weight
at 2, 10, and 20 years, respectively [18]. Buchwald and colleagues conducted a
meta-analysis on the effects of bariatric surgery-induced weight loss and obesity-
related comorbidities. It was reported that at 2 years post-surgery, the overall excess
weight loss for 10,172 patients was 61.2 % [19]. Patients typically lose less weight
after LAGB than RYGB, and the peak excess weight loss typically occurs at
2–3 years with LAGB compared with 1–1.5 years with RYGB. The success of
LAGB at inducing weight loss, however, at 4 years is comparable to RYGB [20].
Further, superobese (BMI >50 kg/m2) individuals have less excess weight loss than
patients with lower BMIs after RYGB, suggesting that bariatric surgery is success-
ful at inducing weight loss, but the magnitude of this response varies among patients
depending on surgical type and preoperative BMI [21].
It is important to recognize that while many patients experience successful weight
loss, up to 20–25 % of patients experience weight regain [16]. These statistics do
not negate the benefit of surgery, but rather suggest the need for all individuals to
modify behavior to focus on food tolerance, proper energy requirements, eating
triggers, beverage selection, and patient nutritional knowledge in conjunction with
appropriate amounts of exercise.
16.2.2 Effects of Bariatric Surgery on Blood Glucose
Obesity is a major risk factor for type 2 diabetes and contributes to its development
by inducing insulin resistance and inflammation, which in turn reduce beta-cell
function [22–24]. The notion that bariatric surgery “cures” diabetes was recognized
over 20 years ago. Pories et al. [25] demonstrated in 141 patients with type 2
diabetes or impaired glucose tolerance that all but two individuals had normalized
glucose tolerance within 10 days after RYGB. At 7.6 years after surgery, 83 % of the
diabetic patients were off their antidiabetic drugs, and 99 % of those with impaired
glucose tolerance were normoglycemic with a normal fasting glucose and hemoglo-
bin A1c [26]. In the Swedish Obesity Study, at 2 years post-surgery with an average
weight loss of nearly 28 kg, 72 % of patients had complete resolution of type 2
diabetes compared with 21 % of controls [17]. Many of these patients had been able
to stop taking oral hypoglycemic drugs or insulin, which is in contrast to the control
group who had an increased need for these agents. These results are similar to those
of Scopinaro et al. [27, 28] who reported long-term follow-up data on 312 patients
with type 2 diabetes undergoing BPD and indicated that 99 % of patients achieved
normal glucose concentrations by 1 year after surgery. At 10 years after surgery,
98 % of the patients were still in complete remission of diabetes (i.e., normal
blood glucose without antidiabetic medication use). However, not all surgical
approaches induce comparable glycemic benefit. Diabetes resolution was observed
in approximately 98 % of patients who underwent BPD (with or without DS),
84 % who underwent RYGB, 72 % who underwent VBG, and 48 % who underwent

adjustable gastric banding [19]. In addition, it is also worth considering that more
recent guidelines for diabetes remission have been established, and current work
suggests that approximately 50–70 % of individuals undergoing RYGB or SG may
not meet remission criteria at 5 years post-operation [29]. This would suggest that
biological factors, such as weight regain or insulin resistance, may contribute to
diabetes relapse. Indeed, resolution of type 2 diabetes is likely to occur in those
with the shortest duration of diabetes (<5 years) or milder forms of diabetes (diet-
controlled), lower central obesity, and/or the greatest weight loss after surgery [30].
Conversely, patients who do not resolve diabetes post-surgery are usually older or
have a more prolonged surgical disease course [25, 31, 32]. Thus, further work
understanding how to optimize diabetes remission rates is needed following surgery.
Nevertheless, the majority of these observational studies have been supported by
randomized control trials (RCTs) in obese cohorts with type 2 diabetes [33–35].
For instance, the Surgical Treatment and Medications Potentially Eradicate Diabetes
Efficiently (STAMPEDE) trial recently demonstrated the effects of bariatric surgery
on controlling glycemia in obese individuals with type 2 diabetes. In the STAMPEDE
trial, Schauer et al. [36] compared the effects of RYGB and SG versus intensive
medical therapy in 150 obese patients with uncontrolled type 2 diabetes at 1 year
post-operation. People were randomly assigned to surgical or medical therapy
groups, and the primary end point was an HbA1c <6.0 %. The results indicated that
RYGB and SG each produced significant improvements in HbA1c in 42 % and
37 % of patients, respectively, meeting glycemic control criteria. Taken together,
bariatric surgery appears to result in dramatic glycemic control and weight loss
improvements in obese patients with type 2 diabetes.
16.2.3 Effects of Bariatric Surgery on Cardiovascular Disease
In addition to glycemic control benefits, bariatric surgery reduces cardiovascular

disease risk by, in part, improving dyslipidemia and hypertension. Obesity is strongly
associated with atherogenic dyslipidemia, which is often defined as elevated triglyc-
erides, high apolipoprotein B, small low-density lipoprotein (LDL) particles, and
low high-density lipoprotein (HDL) cholesterol. Results of a meta-analysis showed
marked decreases in levels of total cholesterol, LDL, and triglycerides after bariatric
procedures [37]. In fact, approximately 70 % of patients experience an improve-
ment in hyperlipidemia with optimal improvements typically derived after BPD and
RYGB [19]. In the Swedish Obesity Study, significant improvements were observed
in triglyceride and HDL concentrations at 2 and 10 years in the surgical versus
the control group [17]. In recent RCTs, including the Diabetes Surgery Study and
STAMPEDE, bariatric surgery decreased plasma triglycerides and increased
circulating HDL more than medical therapy alone [34–36]. Collectively, these data
demonstrate that bariatric surgery is not only effective at regulating blood glucose
levels and sustaining weight loss but also an effective treatment option for improving
blood lipid profiles in obese people. These findings are likely to contribute to the
overall reduction in cardiovascular disease events seen 20 years post-surgery [18].
Hypertension is also highly associated with obesity, and there is good evidence
that weight loss reduces blood pressure [38]. In general, a decrease of 1 % body
weight leads to a 1 mmHg decrease in systolic blood pressure and a 2 mmHg
decrease in diastolic blood pressure [39, 40]. Similar to effects on dyslipidemia,
hyperglycemia, and weight loss, bariatric surgery across all procedures has good
effects on reducing blood pressure. In particular, approximately 61 and 79 % of the
total population with hypertension had it either resolved or improved for up to
2 years post-surgery [19]. The Swedish Obesity Study examined the effect of obesity
on hypertension by investigating the 8-year incidence of hypertension in obese
patients treated with bariatric surgery (VGB, GB, and RYGB) versus matched obese
controls [41]. Although there was no overall difference in systolic blood pressure
and an increase in diastolic blood pressure at 8 years compared to the control group,
it is worth noting that RYGB did appear to be the more favorable surgical procedure
for decreasing systolic and diastolic blood pressure at 10 years (by 4.7 and 10.4 %,
respectively, P < 0.10) [17]. To understand why systolic blood pressure was unchanged,
an examination of weight loss and age was analyzed [41]. Despite rapid improvements
in body weight and blood pressure by 1 year, the slight increase in systolic and
diastolic blood pressure over subsequent years was linked to the rate of weight
gain and age. In fact, in the surgical group, the effect of blood pressure of 1 year
(time between baseline and last observation in the study) was up to four times greater
than the effect of 1 kg regained. Together, these results suggest that the direction of
weight change is more closely linked to blood pressure than the initial weight loss,
but age is an important factor. Bariatric surgery did not decrease diastolic blood
pressure. In fact, diastolic blood pressure increased post-surgery. Given that pulse
pressure is associated with increased risk for coronary artery disease [42], Sjostrom
et al. examined whether surgery could lower pulse pressure compared to a control
group [41]. The results indicated that the weight reduction post-surgery lowered the
rate of increase in pulse pressure seen in obese patients. Taken together, the result of
bariatric surgery on blood pressure is not a simple relationship, but there does seem
to be some protective effects on risk for future coronary heart disease.
16.3 Role of Bariatric Surgery Linking Gut Physiology

to Obesity-Related Disease
Although weight loss is likely important for gains in insulin sensitivity and beta-cell
function following bariatric surgery [43], restrictive procedures do not induce
comparable diabetes resolution rates as compared to BPD or RYGB. Moreover,
despite malabsorption explaining to some extent reductions in reactive oxygen
species and inflammation derived from excess glucose and lipid digestion in obese
individuals, nutrient malabsorption does not occur after standard RYGB [44],
Fig. 16.3 Hypothetical model of nutrient-stimulated gastrointestinal dysfunction in type 2 diabetes
implying that additional factors likely drive the improvements in glycemic control
post bariatric surgery [45]. Altered gut physiology is currently the leading candidate
as the major mechanism to explain improved weight regulation, type 2 diabetes
remission, and reduced risk of cardiovascular disease following bariatric surgery
(Table 16.2).
Rubino and colleagues have hypothesized that overeating causes excessive
stimulation of the gastrointestinal tract leading to metabolic disturbances that
promote hyperglycemia, whereas restricting food contact with the gastrointestinal
tract improves these conditions [43] (Fig. 16.3). Thus, rerouting of food through the
gut may provide a way to alter the secretion of enteroendocrine factors that regulate
insulin sensitivity and/or beta-cell function [46]. Surgical exclusion of the duode-
num in the RYGB procedure and exclusion of the duodenum and jejunum in BPD
result in altered sites – or at least altered relative distribution – of carbohydrate and
fat absorption. This in turn is associated with an increase in anorectic hormones that
induce satiety (e.g., GLP-1, PYY, OXY) and a decrease in the orexigenic hormone
ghrelin that promotes hunger. These observations have led to the establishment
of overlapping and discrete mechanisms that have been termed the “hindgut,”
“midgut,” and “foregut” hypothesis. The hindgut hypothesis suggests that diabetes
remission occurs because of accelerated delivery of nutrients to the distal intestine,
which then augments an insulinotropic signal (e.g., GLP-1) that improves glucose
homeostasis via enhanced insulin action [47]. Indeed, augmentation of GLP-1
secretion increases the insulin response to nutrient intake and, at least in animal
models, induces beta-cell proliferation [48], which together contribute to lowering

blood glucose to normal levels. In addition to a rapid delivery of nutrients to the
distal intestine and increased GLP-1 secretion, the “midgut” hypothesis suggests
increased intestinal gluconeogenesis activating a neuro-hepato-portal sensor that
regulates food intake and lowers hepatic glucose production [49]. In contrast, the
“foregut hypothesis” suggests that nutrient interactions in the duodenum and proxi-
mal jejunum are diabetogenic and, hence, bypassing the duodenum, alleviate the
intestinal factor that induces insulin resistance and beta-cell dysfunction [50, 51].
However, this later hypothesis has been questioned because a diabetic intestinal
factor has not been identified and SG, which does not bypass the duodenum, also
improves glycemic control. It is important to note that although the “hindgut,”
“midgut,” and “foregut” hypotheses are often explained in terms of hormonal
changes, they are not exclusive of altered nutrient flow that affects neural signaling.
Interestingly, patients following surgery often report a reduction in snack numbers
and/or portion size as well as food preference. In fact, these individuals often have
reduced preference for sweet and fat-tasting foods [52]. In addition, although the
gut hypotheses are often presented as mutually exclusive theories, no data actually
exist excluding portions of the upper or lower intestine. Moreover, the exact molec-
ular mechanism underlying the improvement in metabolism following RYGB is
unknown, and it is likely that a number of gut hormones and neural signals produced
at various sites of the gastrointestinal tract elicit unique mechanisms of action.
Indeed, RYGB was reported to enhance intestinal glucose uptake and utilization,
leading to overall improvements in systemic glucose control [53].
16.3.1 Modulation of Diabetes Remission via Altered

Gut Hormones
Gastrointestinal hormones that augment insulin secretion following meal intake are
known as incretins. Note, this effect is only observed when glucose or nutrients are
ingested, not when they are given by an intravenous route [54]. GLP-1 (glucagon-
like polypeptide 1) and GIP (glucose-dependent insulinotropic peptide) account
for approximately 60 % of nutrient-related insulin secretion. In addition, GLP-1
suppresses glucagon and ghrelin and delays gastric emptying, which delays digestion
and reduces postprandial hyperglycemia [55]. GLP-1 also acts on the brain to induce
satiety, although the mechanism(s) remains largely unknown. Laferrere et al. [56]
and others reported an increase in postprandial GLP-1 within 4 weeks following
RYGB, whereas levels of GLP-1 did not rise with comparable weight loss induced
by diet. These findings are consistent with data in patients with type 2 diabetes
1 year following RYGB in which elevated GLP-1 was significantly associated with
insulin action [57]. Moreover, in the STAMPEDE trial, RYGB was shown to augment
GLP-1 stimulation and acylated ghrelin suppression in association with beta-cell
function to a greater extent than either SG or intensive medical therapy 2 years
post-surgery [58, 59]. In general, RYGB is reported to enhance insulin secretion,
whereas gastric restrictive procedures reduce the need to secrete insulin [60]. On the
other hand, GIP is secreted in the K cells located mainly in the duodenum and
proximal jejunum and released in response to nutrients (mainly lipid). Unlike, GLP-
1, GIP is more involved in lipid metabolism (storage) and is thus thought to play a
more direct role in the pathogenesis of obesity. The effect of bariatric surgery on
GIP is more controversial than the findings on GLP-1 such that the role of GIP is
less clear in the regulation of lower fat mass and/or weight maintenance [52].
Non-insulinotropic gut hormones are altered after RYGB and include polypeptide
tyrosine-tyrosine (PYY), oxyntomodulin (OXY), ghrelin, and cholecystokinin (CCK).
Like GLP-1, PYY and OXY are co-secreted by the L cells of the distal small intes-
tine and are responsible for reducing hunger, decreasing food intake, and delaying
gastric emptying after meals. Moreover, while both OXY and PYY inhibit gastric acid
secretion, only PYY reduces pancreatic and intestinal secretions [61]. While several
studies have consistently documented increases in postprandial PYY and GLP-1
after gastric bypass [62–64], with some also reporting enhanced OXY [65], SG and
BPD have also been shown to elevate PYY and GLP-1 [61]. Fewer studies have
examined the role of CCK following bariatric surgery, but in general, CCK levels
increase following RYGB, SG, and LAGB [61]. Ghrelin is a gastric hormone
produced primarily in the stomach with secondary secretion emanating from the
proximal small intestine. Ghrelin is best known as an appetite-stimulating hormone,
but it also has additional effects on impairing insulin sensitivity and reducing
glucose-stimulated insulin secretion [66]. Ghrelin suppression is usually improved
following RYGB or SG, suggesting that suppression of hunger signals helps sustain
weight loss. In contrast, ghrelin levels typically rise following diet-induced weight
loss [67]. It is important to recognize, however, that the effect of bariatric surgery on
ghrelin is controversial [57, 68].
16.4 Link Between Gut Microbiota, Excess Body Weight,

and Metabolic Disease Risk
The gut microbiome has emerged as an important regulator of obesity, metabolism,

and inflammation (Fig. 16.4). In the human intestine, approximately 400 bacterial
species are present and together resemble a multicellular organ that has evolved to
provide complex nutrient signaling and metabolic functions [69]. The vast majority
of these microorganisms belong to three main groups Firmicutes, Bacteroidetes,
and Actinobacteria (comprising ~95 % of total intestinal bacteria) and reside in the
distal portions of the intestine. The gut microbiome is a dynamic organ that changes
in response to the environment. In animals provided a high-fat diet, the gut microbiota
resulted in increased levels of Firmicutes prior to the development of obesity [70].
This renders the microbiota to be more obesogenic and may result in increased
energy harvest from the diet. In line with this observation, Ley et al. demonstrated
that food restriction, not macronutrient content per se (low-carbohydrate vs. low-fat),
was linked to decreased levels of Firmicutes and elevated levels of Bacteroidetes [71],
Fig. 16.4 Summary of the effects of the gut microbiome on weight regulation and energy metabolism
suggesting that nutrient overload is important for gut bacteria modifications.

For instance, altered gut microbiome by high-fat feeding has been linked to intestinal
inflammation and systemic insulin resistance [72]. Further evidence for the role of
the gut microbiome in metabolic health is apparent using germfree rodents, which
are protected from developing diet-induced obesity. In fact, germfree mice have
lower adiposity and have higher food consumption when compared with control
mice. But when germfree mice are colonized with cecal content from control mice,
weight is rapidly gained and food consumption decreases, suggesting that the gut
microbiome regulates fat mass and energy stores [73]. Further, genetically identical
mice fed high-fat diets differ in their metabolic phenotype due to variation in gut
microbiota composition [69], indicating that the cause of metabolic disease may be
due to a complex interaction of environmental factors as well as gut microbiota
profiles. In either event, germfree mice have reduced expression of SGLT-1 and
CD36, which are important for fatty acid and glucose transport in the intestine [74].
These data together highlight gut bacteria as important modulators of nutrient
absorption and suggest that at least two mechanisms are likely involved in explain-
ing the link between obesity and gut microbiome: first, increased capacity to
process otherwise indigestible polysaccharides, leading to subsequent rises in
nutrients, and second, increased gene expression promoting fat storage in adipose
tissue [73]. Thus, the current working hypothesis is that obese animals are better
able to extract energy from dietary intake and store those calories as fat. Indeed,
there are similar observations from obese human studies [75].
16.4.1 Role of Bariatric Surgery on the Gut Microbiota
Bariatric surgery changes gut microbiota to reflect more of a lean phenotype.

Zhang et al. [76] showed that microbiota functional differences occurred in obese
individuals post-RYGB surgery. In fact, RYGB markedly altered Gammaproteobacteria
(member of Enterobacteriaceae), decreased Firmicutes, and lowered methanogens
(key for energy harvest). Similarly, Furet et al. [77] reported that gut microbiota
adapted to RYGB surgery by increasing the Bacteroides/Prevotella ratio, which was
correlated with reduced body fat. Moreover, the rise in Escherichia coli species
also increased in individuals undergoing RYGB, and this change was significantly
correlated with fat mass and leptin, independent of caloric intake restriction. Changes
in adipose tissue mass are strongly linked to systemic inflammation, and consistent
with weight loss following RYGB, the rise in Faecalibacterium prausnitzii species
in type 2 diabetes was related to lower inflammation (hs-CRP and IL-6). Together,
these findings strengthen the view that obesity and gut microbiota are intimately
involved in the regulation of metabolic health. In fact, Kong et al. [78] demonstrated
that RYGB increased gut microbiota richness, and this change in gut microbiota
was directly correlated with genes encoding white adipose tissue mass, metabolism,
and inflammation. Interestingly, approximately 50 % of these relationships were
independent of caloric intake, suggesting that RYGB uniquely alters gut physiology in
favor of weight reduction maintenance. Although distal portions of the small intestine
contribute to the majority of nutrient absorption and gut microbiome in humans,
the upper portions include gut bacteria that have metabolic function. Indeed, when
microbiota from lean individuals are administered into the duodenum of humans
with metabolic syndrome, insulin resistance declines independent of weight loss [79],
suggesting that exclusion of the duodenum contributes to the regulation of glucose
metabolism.
The cause for this altered gut flora milieu is presently an area of intense research,
and several proposed mechanisms are currently being investigated. First, as demon-
strated in rodent studies, the surgically induced restriction in food intake and/or
change in food preference, including lower sugary foods and dietary fat, may
explain modifications in the gut microbiome composition because of the smaller
stomach size and shorter intestinal length [73]. Next, by diverting nutrients away
from the proximal intestine, gut microbiota are exposed to more rapid food delivery
and adapt accordingly. For instance, intestinal cells are exposed to more oxygen
than usual due to the shorter intestinal lengths, and facultative anaerobes develop.
Lastly, from an anatomical perspective, pH levels rise after RYGB surgery in the
stomach and upper intestine. Although a pH <4 is potentially “deadly” for many
microorganisms, some reports suggest that pH modification affects the overall gut
microbiota composition. Further, modification of acid secretion following RYGB not

only lowers distal small intestine pH and influences the production of deconjugated
primary bile acids but also increases acidification in the distal small intestine thereby
increasing secondary bile acid levels via gut bacteria. Indeed, elevated secondary bile
acid levels are reported to decrease hepatic fatty acid uptake, which may contribute
to improvements in hepatic triglyceride metabolism and hepatic steatosis [80].
Alterations in gut microbiome in the colon have also been linked to lower pH, and
these altered gut bacteria favor the rise in short-chain fatty acid production, which
may help regulate lipid metabolism (see below: Inflammation and Innate Immune
Response Related to Gut Physiology) [81]. Taken together, these findings following
RYGB surgery support the gut microbiome as a key physiologic player involved in
fostering nutrient sensing for both weight regulation and glucose homeostasis [78].
16.4.2 Inflammation and Innate Immune Response

Related to Gut Physiology
Low-grade inflammation is a common comorbidity of type 2 diabetes and cardio-

vascular disease. Locally, the gut microbiome is directly linked to intestinal inflam-
mation via changes in bacterial fragments and/or metabolites known to increase
innate immune system responses. Subsequently, these bacterial components and
metabolites have implicated the gut microbiome as a key factor in the development
of obesity and metabolic disease [82] (Fig. 16.5). The innate immune system has the
capacity to sense various bacterial components via pattern recognition receptors
(PRRs). In general, there are two types of PRRs: Toll-like receptors (TLRs) and
Nod-like receptors (NLRs). Although NLRs have important physiologic roles in gut
health [83], TLRs have been well characterized and are the focus of the bacterial
components discussed in this chapter.
16.4.2.1 Role of Lipopolysaccharides
Lipopolysaccharide (LPS) originating from Gram-negative bacteria in the gut

induces low-grade inflammation and insulin resistance, thereby contributing to
disturbances in energy metabolism that promote metabolic disease. LPS is sensed
by TLR4, which is also recognized by nonmicrobial compounds such as saturated
fatty acids [83]. LPS entry into the general circulation is elevated following high-fat
diets (via a “leaky gut” phenomena or chylomicron transport mechanism) [84], and
treatment with prebiotics significantly mitigates the development of glucose intoler-
ance and inflammation derived from the liver and adipose of mice [84]. The mechanism
by which prebiotics reduce LPS is unclear, but animal work suggests intestinal
permeability improves via glucagon-like peptide-2 and cannabinoid-receptor-1
receptor-mediated pathways [85, 86]. These observations may be of particular
Fig. 16.5 Mechanisms that underpin the effects of RYGB on gut physiology
relevance for humans, as obesity is linked to elevated circulating LPS levels when
compared with healthy controls [87]. Human work also supports the LPS linkage to
hyperglycemia, as treatment with insulin-sensitizing agents in individuals with type
2 diabetes lowered LPS in line with greater rises in insulin action [88]. Further,
elevated LPS in type 1 diabetic and vascular kidney disease was highly associated
with serum triglycerides, diastolic blood pressure, and inflammation markers
(e.g., MCP-1) [89], suggesting that metabolic LPS is linked to cardiovascular disease.
Indeed, LPS levels are elevated in individuals with acute heart failure as compared
to stable heart failure or controls, although clinical trials showing that reduced
LPS leads to lower cardiovascular disease are lacking. Given that bariatric surgery
alters gut microbiome composition, it would seem reasonable to suspect changes

in LPS. The limited evidence suggests that not only does bariatric surgery reduce
LPS but that this reduction is also directly linked to lower adipose mass and
improved HbA1c levels [87]. Taken together, alterations in gut microbiome appear
to contribute to reductions in LPS that reduce metabolic disease risk.
16.4.2.2 Role of Short-Chain Fatty Acids
Short-chain fatty acids (SCFAs) are produced in the colon by gut microbes that
ferment nondigestible polysaccharides (e.g., inulin) [85, 86]. The rise in SCFA
(i.e., acetate, butyrate, and propionate) levels is important for weight regulation as
they are implicated in satiety and decreased food intake, although SCFAs may act
as substrate for lipogenesis. As mentioned previously (see above: Modulation of
Diabetes Remission via Altered Gut Physiology), improved appetite regulation is in
part related to elevated GLP-1 and PYY as well as lower ghrelin, which collectively
contribute to reduced hunger and food intake. The effects of bariatric surgery on gut
hormones and the gut microbiome would by inference suggest that the changes in
SCFA contribute to the overall regulation of appetite. In fact, SCFAs have been
shown to influence gut peptides, appetite, and energy expenditure [84]. For example,
intracolonic and ileum infusion of mixed SCFA increased PYY secretion in rats
and pigs, and supplementation of rat diets with acetate or butyrate lowered body
weight [90]. Although studies in humans are limited, Arora et al. [91] suggested that
propionate may reduce appetite. In addition to the effects on gut peptides, butyrate and
propionate may increase leptin secretion from adipocytes [92]. This later mechanism
may provide an alternative by which the gut directly communicates with adipose
tissue to regulate body weight and feeding behavior. Interestingly, RYGB surgery
enhanced propionate and lowered acetate levels in rodents, and this alteration in
SCFA levels was attributed to the compositional change in gut microbiota. Increased
propionate levels could contribute to improved weight regulation by blocking the
conversion of acetate to triglyceride synthesis in the liver and blood, thereby reducing
ectopic lipid depots. These findings are consistent with germfree animal work
showing that transferring gut microbiota increases insulin and glucose-stimulated
hepatic triglyceride synthesis via sterol response element binding protein (SREBP-1)
and carbohydrate response element binding protein (ChREBP), respectively [90].
Moreover, decreased acetate levels per se may contribute to lower central and
peripheral adiposity following bariatric surgery by reducing substrate availability
for lipogenesis [93]. Concomitantly, bariatric surgery is reported to increase energy
expenditure, and this change in thermogenesis may be attributable to gut microbial-
derived changes in SCFA production. Interestingly, AMPK phosphorylation in the
skeletal muscle and liver is increased in germfree mice on a Western diet [84].
Therefore, the gut microbiota through an unknown mechanism appear capable of
altering peripheral and hepatic fatty acid oxidation and hepatic insulin sensitivity [94],
an observation that is further strengthened by the fact that RYGB in rats increases
hepatic AMPK activity [95]. In addition, the overall change in energy status of
adipose, hepatic, and skeletal muscle following surgery may be partially related
to changes in fasting-induced adipose factor (FIAF). Indeed, colonization of germ-
free mice is characterized by downregulation of FIAF expression, which in turn leads
to increased lipoprotein lipase activity, enhanced lipid storage in adipocytes, and
downregulation of PGC-1 α-related genes in peripheral tissue [90]. These later find-
ings are consistent with evidence suggesting that propionate activates sympathetic
nerve activity and has high binding affinity for GPR41 (G protein-coupled receptor),
which is important for overall energy expenditure [93]. In fact, the effect of SCFA
on GPR41 (as well as GPR43) may be relevant for modulating lipolysis, as infusion
of acetate reduced circulating free fatty acids. Moreover, GPR43 is expressed in
PYY containing L cells of the small intestine. Since L cells are also responsible for
GLP-1 secretion, it is reasonable that SCFA may influence insulin secretion, and this
is supported by rodent work [96]. Overall, these data highlight SCFA as an impor-
tant factor that reduces appetite and/or increases energy metabolism to promote
weight regulation.
16.4.2.3 Role of Trimethylamine-N-Oxide
Gut microbiota release choline from dietary phosphatidylcholine to form trimethyl-

amine (TMA). TMA is transported to the liver via the portal vein and is oxidized to
trimethylamine-N-oxide (TMAO). Elevated levels of plasma TMAO, choline, and
betaine have dose-dependent associations with the presence of cardiovascular disease
independent of conventional risk factors (e.g., blood pressure, triglycerides, etc.)
and medication use [97]. Rats fed high choline diets or TMAO diets had elevated
circulating TMAO levels and greater aortic root atherosclerotic plaque without
alterations in plasma glucose or blood lipids. Although obesity may influence circu-
lating TMAO levels, dietary manipulation appears important for modulating TMAO,
such that high-fat and increased meat consumption lead to greater elevated TMAO
compared with low-fat and vegetarian style meals. Data on TMAO following bariatric
surgery is sparse. There are some urinary data to suggest that trimethylamine is
elevated in rodents following bariatric surgery [98]. However, more work is needed
to determine if the use of gut microbiota therapies (prebiotics and/or bariatric surgery)
influences TMAO since this metabolite appears to be independently associated with
cardiovascular disease risk [99].
16.4.2.4 Role of Hippurate
Intestinal bacteria metabolize low-weight aromatic compounds and polyphenols

from the diet resulting in benzoic acid. In the liver, benzoic acid is conjugated with
glycine to form hippurate, which is excreted in the urine [100]. Hippurate levels are
lower in obese insulin-resistant rodents compared with wild-type controls [101].
In comparison to lean humans, morbidly obese individuals have low levels of
hippurate, which may be clinically relevant since low hippurate is linked with elevated
blood pressure [102]. In animals, bariatric surgery increases urinary hippurate in
parallel with weight loss [98], strengthening the notion that RYGB alters various
aspects of gut microbiota in relation to obesity and metabolic health.
16.5 Bile Acids as Regulators of Energy Metabolism

and Body Weight
Bile acids are produced in the liver, stored in the gall bladder, and secreted into the
duodenum upon meal consumption. Bile acid levels nearly triple following meal
consumption as they are important in not only the facilitation of micelle formation,
which promotes the processing/digestion of dietary fat and fat-soluble vitamins, but
also energy metabolism. Although fasting bile acids do not differ between lean and
obese individuals, obesity appears to blunt the rise in some circulating postprandial
bile acids, while other circulating glycemic conjugated forms preferentially decrease
[103, 104]. The rise in postprandial bile acids is particularly relevant to energy
metabolism and weight regulation as their hormonal effects include stimulating
FGF19, GLP-1, and brown adipose activity. Interestingly, fasting total serum bile acids
and individual levels of taurodeoxycholic, glycocholic, glycochenodeoxycholic, and
glycodeoxycholic acids are elevated after RYGB surgery compared to preoperatively
and when compared with weight-matched nonsurgical controls [103, 105]. The exact
mechanism responsible for elevated bile acids following bariatric surgery is unclear,
but animal work suggests that increased nutrient delivery to the ileum leads to
increased satiety hormone levels and weight loss [106]. This suggests that nutrient
flow to the distal small intestine is a potentially important mechanism linking altered
bile acid levels seen following RYGB with changes in gut hormone secretion.
However, despite immediate elevations in bile acids following RYGB, the rise in
bile acids appears even greater several months post-operation. As such, it is likely
that intestinal adaptation, including genes that synthesize and regulate transporters
important for bile acid uptake, plays a key role in explaining elevated postprandial
bile acid levels [103]. RYGB also alters intestinal gut microbiota, which are key
regulators of bile acid conjugation and secondary bile acid formation [77, 93].
Interestingly, germfree mice also have low bile acid levels and diversity compared
with wild-type controls, highlighting that gut microbiota may contribute to bile
acid diversity and impact not only GLP-1 secretion but also energy expenditure.
Concomitantly, it is important to note that conjugated bile acids entering the duodenum
from the enterohepatic cycle circulate to the ileum where they are deconjugated,
and the portion of these later bile acids might directly affect the composition of the
microbiota [73].
Bile acids have been implicated in the improvement in weight and glucose
metabolism following bariatric surgery. Pournaras et al. [106] demonstrated that
fasting total serum bile acids are elevated within days following RYGB, but not
LAGB, suggesting that bile acids may contribute to weight-independent improvements

in glucose homeostasis. Indeed, fasting total bile acids are inversely correlated with
postprandial glucose and positively correlated with peak GLP-1 levels [107]. In line
with this observation, bile acids are known to act on TGR5 receptors located on
enteroendocrine cells and promote the secretion of GLP-1 release [108], which may
contribute to satiety and beta-cell insulin secretion. In addition, farnesoid X receptor
(FXR) in pancreatic beta-cells may directly respond to the rise in bile acids, thereby
increasing insulin release [109]. In parallel, through an FXR-mediated pathway in
the intestine, bile acids stimulate the secretion of FGF19, a protein that contributes to
improved peripheral glucose disposal and lipid homeostasis [110, 111]. Thus, the
physiologic effects of bile acids likely extend beyond that of gut-pancreas “cross-
talk,” since TGR5 receptors are also located on the skeletal muscle. TGR5 receptors
are also present in brown adipose tissue, and the binding of bile acids to TGR5 in
the skeletal muscle and brown adipose tissue may contribute to enhanced action
of thyroid hormones to foster weight loss by increasing energy expenditure.
Watanabe et al. [112] have reported that bile acids increase energy expenditure in
both skeletal muscle and brown adipose tissue, and Ockenga et al. [113] found that
postprandial levels of circulating bile acids are strongly associated with postpran-
dial energy expenditure in lean individuals. Collectively, these findings suggest that
restoration of elevated bile acids in obese individuals after bariatric surgery may
facilitate weight loss.
16.6 Risks of Bariatric Surgery and Nutrient Deficiencies
A misconception about bariatric surgery is that it is highly related with risk of

postsurgical complications and mortality. However, the prospective Longitudinal
Assessment of Bariatric Surgery study [113] reported that the 30-day death rate of
adults undergoing bariatric surgery (RYGB or LAGB) was 0.3 % [114]. Thus, there is
little evidence for higher risk of mortality than standard operation (e.g., cardiovascular),
and if anything, data suggest that bariatric surgery increases life expectancy [5, 115]
due to reductions in obesity-induced cardiovascular risk factors such as diabetes.
Although in rare cases, symptomatic hypoglycemia has been documented after
RYGB and is associated with postprandial hyperinsulinemia due to elevated GLP-1
levels [116], the incidence of neuroglycopenia and seizures is rare. The conventional
treatment of hypoglycemia in these patients involves carbohydrate restriction to
minimize hypoglycemic-related episodes [117]. Despite little evidence for higher
death rates after bariatric surgery, it is important to consider these risks against the
long-term cardiovascular risk of continued obesity.
LAGB is viewed as the safest of the current bariatric procedures. It does not
involve bowel anastomosis, and the risk of major hemorrhage, gastric perforation,
and pulmonary embolism is less than 1 %. Late complications requiring reoperation
include band slippage or prolapse (5–10 %) and band erosion (1–3 %). The entire
intestinal tract is left intact, so subsequent nutritional deficiencies are rare [118].
RYGB, on the other hand, carries an overall risk of major complication of 10–15 %.
Anastomotic leak (1–5 %), pulmonary embolism (<1 %), and hemorrhage (1–4 %)
can be life-threatening but are rare. Late complications such as ulcer or stricture
formation at the gastrojejunostomy site occur in 5–10 % of cases and are managed
nonoperatively [46].
Nutritional deficiencies occur in 30–70 % of patients. Patients at high risk of
developing severe nutritional deficiencies include those who have lost more than
10 % of their body weight by 1 month and those with: anastomotic stenosis, surgical
revision requirements, and persistent vomiting [119]. Protein calorie malnutrition is
also a concern and can be recognized by signs such as edema, hypoalbuminemia,
anemia, and hair loss. To minimize these effects, it is generally recommended
that patients consume between 60 and 80 g/day of protein and approximately
800 kcal/day. Vitamin deficiencies can lead to peripheral neuropathy (B12), Wernicke
encephalopathy (B1), and metabolic bone disease (Vitamin D). Subsequently, in
addition to multivitamin supplementation, monitoring nutrient and vitamin levels
after bariatric surgery is recommended at least every 6 months [119–121].
16.7 Implications for Gut Physiology Following Bariatric

Surgery in Metabolic Disease
Bariatric surgery has proven valuable in identifying the gut as a critical organ
regulating energy balance and glucose homeostasis (Fig. 16.5). The resulting weight
loss from bariatric surgery is persistent in most patients, and the durability in weight
loss is considered an underlying mechanism responsible for the majority of long-
term glycemic control benefit and cardiovascular risk reduction. However, it is clear
that alterations in gut physiology have distinct effects on not only appetite but also
on insulin resistance, beta-cell function, hypertension, and blood lipids. Specifically,
the prevailing view at this point in time is that alterations in enteroendocrine cells
following bariatric surgery induce elevations in gut hormones (e.g., GLP-1, GIP,
PYY, etc.) that promote satiety and improve insulin action, whereas the changes in
gut microbiome and bile acids are new physiologic mechanisms contributing to the
overall improvement in body fat and cardiovascular benefits following bariatric
surgery. Taken together, the interaction between the gut and metabolic improvements
seen following bariatric surgery warrants consideration of obesity-induced type 2
diabetes and cardiometabolic risk as “intestinal-related diseases.” Thus, there is
an opportunity going forward to stratify patients prior to bariatric surgery from a
metabolic and gut microbial point of view, and this may impact the success of surgery
on obesity-related disease. Further understanding of mechanisms related to gut
metabolism following bariatric surgery, with or without lifestyle modification, will
likely promote new medical strategies that will improve treatment and resolution of
obesity, type 2 diabetes, and cardiovascular disease.
References
1. Obesity and overweight. Fact sheet no 311. Geneva: World Health Organization; 2013.
2. Flegal KM, Carroll MD, Kit BK, Ogden CL. Prevalence of obesity and trends in the distribu-
tion of body mass index among US adults, 1999–2010. JAMA. 2012;307(5):491–7.
3. Ogden C, Carroll M, Curtin L, McDowell M, Tabak C, Flegal K. Prevalence of overweight
and obesity in the United States, 1999–2004. JAMA. 2006;295(13):1549–55.
4. Noria SF, Grantcharov T. Biological effects of bariatric surgery on obesity-related comor-
bidities. Can J Surg. 2013;56(1):47–57.
5. Adams KF, Schatzkin A, Harris TB, Kipnis V, Mouw T, Ballard Barbash R, Hollenbeck A,
Leitzmann MF. Overweight, obesity, and mortality in a large prospective cohort of persons
50 to 71 years old. N Engl J Med. 2006;355(8):763–78.
6. NIH conference. Gastrointestinal surgery for severe obesity. Consensus development confer-
ence panel. Ann Intern Med. 1991;115(12):956–61.
7. Steinbrook R. Surgery for severe obesity. N Engl J Med. 2004;350(11):1075–9.
8. Colquitt JL, Picot J, Loveman E, Clegg AJ. Cochrane Database Syst Rev. 2009;2:CD003641.
doi:10.1002/14651858.CD003641.pub3. Review.
9. Parikh MS, Fielding GA, Ren CJ. U.S. experience with 749 laparoscopic adjustable gastric
bands: intermediate outcomes. Surg Endosc. 2005;19(12):1631–5.
10. Gagner M, Deitel M, Kalberer TL, Erickson AL, Crosby RD. The second international
consensus summit for sleeve gastrectomy. Surg Obes Relat. 2009;5(4):476–85.
11. Hutter MM, Schirmer BD, Jones DB, Ko CY, Cohen ME, Merkow RP, Nguyen NT. First
report from the American college of surgeons bariatric surgery center network: laparoscopic
sleeve gastrectomy has morbidity and effectiveness positioned between the band and the
bypass. Ann Surg. 2011;254(3):410–20.
12. Kumar NT, Christopher CT. Endoscopic solutions for weight loss. Curr Opin Gastroenterol.
2011;27(5):407–11.
13. Rubino F. Bariatric surgery: effects on glucose homeostasis. Curr Opin Clin Nutr Metab
Care. 2006;9(4):497–507.
14. Buchwald H, Estok R, Fahrbach K, Banel D, Jensen M, Pories W, Bantle J, Sledge I. Weight
and type 2 diabetes after bariatric surgery: systematic review and meta-analysis. Am J Med.
2009;122(3):248–56. e5.
15. Stimac DM, Sanja. Endoscopic approaches to obesity. Dig Dis. 2012;30(2):187–95.
16. Blomain ES, Dirhan DA, Valentino MA, Kim GW, Waldman SA. ISRN Obes.
2013;2013:210524. doi:10.1155/2013/210524. eCollection 2013. Review.
17. Sjostrom L, Lindroos AK, Peltonen M, et al. Lifestyle, diabetes, and cardiovascular risk
factors 10 years after bariatric surgery. N Engl J Med. 2004;351(26):2683–93.
18. Sjostrom L, Peltonen M, Jacobson P, et al. Bariatric surgery and long-term cardiovascular
events. JAMA. 2012;307(1):56–65.
19. Buchwald H, Avidor Y, Braunwald E, Jensen MD, Pories W, Fahrbach K, Schoelles
K. Bariatric surgery: a systematic review and meta-analysis. JAMA. 2004;292(14):1724–37.
20. Chapman AE, Kiroff G, Game P, Foster B, O’Brien P, Ham J, Maddern G. Laparoscopic
adjustable gastric banding in the treatment of obesity: a systematic literature review. Surgery.
2004;135(3):326–51.
21. Brethauer SA, Chand B, Schauer PR. Risks and benefits of bariatric surgery: current evidence.
Cleve Clin J Med. 2006;73(11):993–1007.
22. Despres JP, Lemieux I, Prud’homme D. Treatment of obesity: need to focus on high risk
abdominally obese patients. BMJ. 2001;322(7288):716–20.
23. DeFronzo RA, Abdul Ghani MA. Preservation of ß-cell function: the key to diabetes preven-
tion. J Clin Endocrinol Metab. 2011;96(8):2354–66.
24. Samuel VT, Shulman GI. Mechanisms for insulin resistance: common threads and missing
links. Cell. 2012;148(5):852–71.
25. Pories WJ, Caro JF, Flickinger EG, Meelheim HD, Swanson MS. The control of diabetes
mellitus (NIDDM) in the morbidly obese with the greenville gastric bypass. Ann Surg.
1987;206(3):316–23.
26. Pories WJ, Swanson MS, MacDonald KG, Long SB, Morris PG, Brown BM, Barakat HA,
deRamon RA, Israel G, Dolezal JM. Who would have thought it? An operation proves to be
the most effective therapy for adult-onset diabetes mellitus. Ann Surg. 1995;222(3):339–50.
27. Scopinaro N, Papadia F, Marinari G, Camerini G, Adami G. Long-term control of type 2
diabetes mellitus and the other major components of the metabolic syndrome after biliopan-
creatic diversion in patients with BMI < 35 kg/m2. Obes Surg. 2007;17(2):185–92.
28. Scopinaro N, Marinari GM, Camerini GB, Papadia FS, Adami GF. Specific effects of bilio-
pancreatic diversion on the major components of metabolic syndrome: a long-term follow-up
study. Diabetes Care. 2005;28(10):2406–11.
29. Brethauer SA, Aminian A, Romero-Talamás H, Batayyah E, Mackey J, Kennedy L, Kashyap
SR, Kirwan JP, Rogula T, Kroh M, Chand B, Schauer PR. Can diabetes be surgically cured?
Long-term metabolic effects of bariatric surgery in obese patients with type 2 diabetes mellitus.
Ann Surg. 2013;258(4):628–36.
30. Torquati A, Lutfi R, Abumrad N, Richards WO. Is roux-en-Y gastric bypass surgery the most
effective treatment for type 2 diabetes mellitus in morbidly obese patients? J Gastrointest
Surg. 2005;9(8):1112–6.
31. Pories WJ, MacDonald KG, Flickinger EG, Dohm GL, Sinha MK, Barakat HA, May HJ,
Khazanie P, Swanson MS, Morgan E. Is type II diabetes mellitus (NIDDM) a surgical disease?
Ann Surg. 1992;215(6):633–42.
32. Sugerman HJ, Wolfe LG, Sica DA, Clore JN. Diabetes and hypertension in severe obesity and
effects of gastric bypass-induced weight loss. Ann Surg. 2003;237(6):751–6.
33. Dixon JB, O’Brien PE, Playfair J, Chapman L, Schachter LM, Skinner S, Proietto J, Bailey
M, Anderson M. Adjustable gastric banding and conventional therapy for type 2 diabetes: a
randomized controlled trial. JAMA. 2008;299(3):316–23.
34. Mingrone G, Panunzi S, De Gaetano A, Guidone C, Iaconelli A, Leccesi L, Nanni G, Pomp A,
Castagneto M, Ghirlanda G, Rubino F. Bariatric surgery versus conventional medical therapy
for type 2 diabetes. N Engl J Med. 2012;366(17):1577–85.
35. Ikramuddin S, Korner J, Lee WJ, Connett JE, Inabnet WB, Billington CJ, Thomas AJ, Leslie DB,
Chong K, Jeffery RW, Ahmed L, Vella A, Chuang LM, Bessler M, Sarr MG, Swain JM,
Laqua P, Jensen MD, Bantle JP. Roux-en-Y gastric bypass vs. intensive medical management
for the control of type 2 diabetes, hypertension, and hyperlipidemia: the diabetes surgery
study randomized clinical trial. JAMA. 2013;309(21):2240–9.
36. Schauer PR, Kashyap SR, Wolski K, Brethauer SA, Kirwan JP, Pothier CE, Thomas S, Abood
B, Nissen SE, Bhatt DL. Bariatric surgery versus intensive medical therapy in obese patients
with diabetes. N Engl J Med. 2012;366(17):1567–76.
37. Bouldin MJ, Ross LA, Sumrall CD, Loustalot FV, Low AK, Land KK. The effect of obesity
surgery on obesity comorbidity. Am J Med Sci. 2006;331(4):183–93.
38. Neter JE, Stam BE, Kok FJ, Grobbee DE, Geleijnse JM. Influence of weight reduction
on blood pressure: a meta-analysis of randomized controlled trials. Hypertension.
2003;42(5):878–84.
39. Dornfeld LP, Maxwell MH, Waks AU, Schroth P, Tuck ML. Obesity and hypertension: long-
term effects of weight reduction on blood pressure. Int J Obes. 1985;9(6):381–9.
40. Hypertension Prevention Trial Research Group. The hypertension prevention trial: three-year
effects of dietary changes on blood pressure. Hypertension prevention trial research group.
Arch Intern Med. 1990;150(1):153–62.
41. Sjöström CD, Peltonen M, Sjöström L. Blood pressure and pulse pressure during long-term
weight loss in the obese: the Swedish obese subjects (SOS) intervention study. Obes Res.
2001;9(3):188–95.
42. Benetos A, Rudnichi A, Safar M, Guize L. Pulse pressure and cardiovascular mortality in
normotensive and hypertensive subjects. Hypertension. 1998;32(3):560–4.
43. Rubino F, R’bibo SL, del Genio F, Mazumdar M, McGraw T. Metabolic surgery: the role of
the gastrointestinal tract in diabetes mellitus. Nat Rev Endocrinol. 2010;6(2):102–9.
44. Marceau P, Hould FS, Simard S, Lebel S, Bourque RA, Potvin M, Biron S. Biliopancreatic
diversion with duodenal switch. World J Surg. 1998;22(9):947–54.
45. Brolin R, LaMarca L, Kenler H, Cody R. Malabsorptive gastric bypass in patients with
superobesity. J Gastrointest Surg. 2002;6(2):195–203.
46. Kashyap SR, Gatmaitan P, Brethauer SA, Schauer PR. Bariatric surgery for type 2 diabetes:
weighing the impact for obese patients. Cleve Clin J Med. 2010;77(7):468–76.
47. Cummings DE. Endocrine mechanisms mediating remission of diabetes after gastric bypass
surgery. Int J Obes. 2009;33 Suppl 1:S33–40.
48. Pournaras DJ, le Roux CW. Obesity, gut hormones, and bariatric surgery. World J Surg.
2009;33(10):1983–8.
49. Mithieux G. A novel function of intestinal gluconeogenesis: central signaling in glucose and
energy homeostasis. Nutrition. 2009;25(9):881–4.
50. Rubino F. The early effect of the roux-en-Y gastric bypass on hormones involved in body
weight regulation and glucose metabolism. Ann Surg. 2004;240(2):236–42.
51. Salinari S, Debard C, Bertuzzi A, Durand C, Zimmet P, Vidal H, et al. PLoS One.
2013;8(2):e56258. doi:10.1371/journal.pone.0056258. Epub 2013 Feb 20.
52. Ionut V, Bergman RN. Mechanisms responsible for excess weight loss after bariatric surgery.
J Diabetes Sci Technol. 2011;5(5):1263–82.
53. Saeidi N, Meoli L, Nestoridi E, Gupta NK, Kvas S, Kucharczyk J, Bonab AA, Fischman AJ,
Yarmush ML, Stylopoulos N. Reprogramming of intestinal glucose metabolism and glycemic
control in rats after gastric bypass. Science. 2013;341(6144):406–10.
54. Vollmer K, Holst JJ, Baller B, Ellrichmann M, Nauck MA, Schmidt WE, Meier JJ. Predictors
of incretin concentrations in subjects with normal, impaired, and diabetic glucose tolerance.
Diabetes. 2008;57(3):678–87.
55. Holst JJ, Vilsbll T, Deacon CF. The incretin system and its role in type 2 diabetes mellitus.
Mol Cell Endocrinol. 2009;297(1–2):127–36.
56. Laferrère B, Teixeira J, McGintry J, Tran H, Egger JR, Colarusso A, Kovack B, Bawa B,
Koshy N, Lee H, Yapp K, Olivan B. Effect of weight loss by gastric bypass surgery versus
hypocaloric diet on glucose and incretin levels in patients with type 2 diabetes. J Clin
Endocrinol Metab. 2008;93(7):241979–85.
57. Samat A, Malin SK, Huang H, Schauer PR, Kirwan JP, Kashyap SR. Ghrelin suppression is
associated with weight loss and insulin action following gastric bypass surgery at 12 months
in obese adults with type 2 diabetes. Diabetes Obes Metab. 2013;15(10):963–6.
58. Kashayp SR, Bhatt DL, Wolski K, Wantanabe RM, Abdul-Ghani M, Abood B, Pothier CE,
Brethauer SA, Nissen S, Gupta M, Kirwan JP, Schauer PR. Metabolic effects of bariatric
surgery in patients with moderate obesity and type 2 diabetes: analysis of a randomized
control trial comparing surgery vs. intensive medical treatment. Diabetes Care.
2013;36(8):2175–82.
59. Malin SK, Samat A, Wolski K, Abood B, Pothier CE, Bhatt DL, Nissen S, Brethauer SA,
Schauer PR, Kirwan JP, Kashyap SR. Improved acylated ghrelin suppression at 2 years in
obese patients with type 2 diabetes: effects of bariatric surgery vs standard medical therapy.
Int J Obes. 2013;36(8):2175–82.
60. Kashyap SR, Daud S, Kelly KR, Gastaldelli A, Win H, Brethauer S, Kirwan JP, Schauer
PR. Acute effects of gastric bypass versus gastric restrictive surgery on beta-cell function and
insulinotropic hormones in severely obese patients with type 2 diabetes. Int J Obes.
2010;34(3):462–71.
61. Quercia I, Dutia R, Kotler DP, Belsley S, Laferrerre B. Gastrointestinal changes after bariatric
surgery. Diabetes Metab. 2013;S1262–3636
62. Korner J, Bessler M, Inabnet W, Taveras C, Holst JJ. Exaggerated glucagon-like peptide-1
and blunted glucose-dependent insulinotropic peptide secretion are associated with roux-en-Y
gastric bypass but not adjustable gastric banding. Surg Obes Relat Dis. 2007;3(6):597–601.
63. Hanusch Enserer U, Ghatei MA, Cauza E, Bloom SR, Prager R, Roden M. Relation of fasting
plasma peptide YY to glucose metabolism and cardiovascular risk factors after restrictive
bariatric surgery. Wien Klin Wochenschr. 2007;119(9–10):291–6.
64. le Roux CW, Aylwin SJ, Batterham RL, Borg CM, Coyle F, Prasad V, Shurey S, Ghatei MA,
Patel AG, Bloom SR. Gut hormone profiles following bariatric surgery favor an anorectic state,
facilitate weight loss, and improve metabolic parameters. Ann Surg. 2006;243(1):108–14.
65. Laferrère B, Swerdlow N, Bawa B, Arias S, Bose M, Olivan B, Teixeira J, McGinty J, Rother
KI. Rise of oxyntomodulin in response to oral glucose after gastric bypass surgery in patients
with type 2 diabetes. J Clin Endocrinol Metab. 2010;95(8):4072–6.
66. Tong J, Prigeon RL, Davis HW, Bidlingmaier M, Kahn SE, Cummings D, Tschop M,
D’Alessio D. Ghrelin suppresses glucose-stimulated insulin secretion and deteriorates
glucose tolerance in healthy humans. Diabetes. 2010;59(9):2145–51.
67. Cummings DE, Weigle DS, Frayo RS, Breen P, Ma M, Dellinger EP, Purnell J. Plasma
ghrelin levels after diet-induced weight loss or gastric bypass surgery. N Engl J Med.
2002;346(21):1623–30.
68. Falken Y, Hellstrom PM, Holst JJ, Naslund E. Changes in glucose homeostasis after roux-en-
Y gastric bypass surgery for obesity at day three, two months, and one year after surgery: role
of gut peptides. J Clin Endocrinol Metab. 2011;96(7):2227–35.
69. Breen DM, Rasmussen BA, Cote CD, Jackson VM, Lam TK. Nutrient-sensing mechanisms
in the gut as therapeutic targets for diabetes. Diabetes. 2013;62(9):3005–13.
70. Hildebrandt MA, Hoffmann C, Sherrill-Mix SA, Keilbaugh SA, Hamady M, Chen YY,
Knight R, Ahima RS, Bushman F, Wu GD. High-fat diet determines the composition of the
murine gut microbiome independently of obesity. Gastroenterology. 2009;137(5):1716–24.
ated with obesity. Nature. 2006;444(7122):1022–3.
72. Ding S, Chi MM, Scull BP, Rigby R, Schwerbrock NM, Magness S, et al. PLoS One.
2010;5(8):e12191. doi:10.1371/journal.pone.0012191.
73. Aron-Wisnewsky J, Dore J, Clement K. The importance of the gut microbiota after bariatric
surgery. Nat Rev Gastroenterol Hepatol. 2012;9(10):590–8.
74. Swartz TD, Duca FA, de Wouters T, Sakar Y, Covasa M. Up-regulation of intestinal type 1
taste receptor 3 and sodium glucose luminal transporter-1 expression and increased sucrose
intake in mice lacking gut microbiota. Br J Nutr. 2012;107(5):621–30.
75. Schwiertz A, Taras D, Schafer K, Beijer S, Bos NA, Donus C, Hardt PD. Microbiota and
SCFA in lean and overweight healthy subjects. Obesity. 2010;18(1):190–5.
76. Zhang H, DiBaise JK, Zuccolo A, Kudrna D, Braidotti M, Yu Y, Parameswaran P, Crowell
MD, Wing R, Rittmann BE, Krajmalnik-Brown R. Human gut microbiota in obesity and after
gastric bypass. Proc Natl Acad Sci U S A. 2009;106(7):2365–70.
77. Furet JP, Kong LC, Tap J, Poitou C, Basdevant A, Bouillot JL, Mariat D, Corthier G, Dore J,
Henegar C, Rizkalla S, Clement K. Differential adaptation of human gut microbiota to bariat-
ric surgery-induced weight loss: links with metabolic and low-grade inflammation markers.
Diabetes. 2010;59(12):3049–57.
78. Kong LC, Tap J, Aron-Wisnewsky J, Pelloux V, Basdevant A, Bouillot JL, Zucker JD,
Dore J, Clement K. Gut microbiota after gastric bypass in human obesity: increased rich-
ness and associations of bacterial genera with adipose tissue genes. Am J Clin Nutr.
2013;98(1):16–24.
79. Vrieze A, Van Nood E, Holleman F, Salojarvi J, Kootte RS, Bartelsman JF, Dallinga-Thie GM,
Ackermans MT, Serlie MJ, Oozeer R, Derrien M, Druesne A, Van Hylckama Vlieg JE,
Bloks VW, Groen AK, Heilig HG, Zoetendal EG, Stroes ES, de Vos WM, Hoekstra JB,
Nieuwdorp M. Transfer of intestinal microbiota from lean donors increases insulin sensitivity
in individuals with metabolic syndrome. Gastroenterology. 2012;143(4):913–6.e7.
80. Nie B, Park HM, Kazantzis M, Lin M, Henkin A, Ng S, Song S, Chen Y, Tran H, Lai R, Her C,
Maher JJ, Forman BM, Stahl A. Specific bile acids inhibit hepatic fatty acid uptake in mice.
Hepatology. 2012;56(4):1300–10.
81. Walker AW, Duncan SH, McWilliam Leitch EC, Child MW, Flint H. pH and peptide supply
can radically alter bacterial populations and short-chain fatty acid ratios within microbial
communities from the human colon. Appl Environ Microbiol. 2005;71(7):3692–700.
82. Laferrère B. Gut feelings about diabetes. Endocrinol Nutr. 2012;59(4):254–60.
83. Harris K, Kassis A, Major G, Chou CJ. J Obes. 2012;2012:879151. doi:10.1155/2012/879151.
Epub 2012 Jan 24. Erratum in: J Obes. 2012;2012:782920.
84. Tilg H, Kaser A. Gut microbiome, obesity, and metabolic dysfunction. J Clin Invest.
2011;121(6):2126–32.
85. Cani PD, Knauf C, Iglesias MA, Drucker DJ, Delzenne NM, Burcelin R. Improvement of
glucose tolerance and hepatic insulin sensitivity by oligofructose requires a functional
glucagon-like peptide 1 receptor. Diabetes. 2006;55(5):1484–90.
86. Cani PD, Neyrinck AM, Fava F, Knauf C, Burcelin RG, Tuohy KM, Gibson GR, Delzenne
NM. Selective increases of bifidobacteria in gut microflora improve high-fat-diet-induced
diabetes in mice through a mechanism associated with endotoxaemia. Diabetologia.
2007;50(11):2374–83.
87. Troseid M, Nestvold TK, Rudi K, Thoresen H, Nielsen E, Lappegard KT. Plasma lipopoly-
saccharide is closely associated with glycemic control and abdominal obesity: evidence from
bariatric surgery. Diabetes Care. 2013;36(11):3627–32.
88. Creely SJ, McTernan PG, Kusminski CM, Fisher fM, Da Silva NF, Khanolkar M, Evans M,
Harte AL, Kumar S. Lipopolysaccharide activates an innate immune system response in
human adipose tissue in obesity and type 2 diabetes. Am J Physiol Endocrinol Metab.
2007;292(3):E740–7.
89. Lassenius MI, Pletillainen KH, Kaartinen K, Pussinen PJ, Syrjanen J, Forsblom C, Porsti I,
Rissanen A, Kaprio J, Mustonen J, Groop PH, Lehto. Bacterial endotoxin activity in human
serum is associated with dyslipidemia, insulin resistance, obesity, and chronic inflammation.
Diabetes Care. 2011;34(8):1809–15.
90. Shen J, Obin MS, Zhao L. The gut microbiota, obesity and insulin resistance. Mol Aspects
Med. 2013;34(1):39–58.
91. Arora T, Sharma R, Frost G. Propionate. Anti-obesity and satiety enhancing factor? Appetite.
2011;56(2):511–5.
92. Soliman MM, Ahmed MM, Salah-Eldin AE, Abdel-Aal A. Butyrate regulates leptin expres-
sion through different signaling pathways in adipocytes. J Vet Sci. 2011;12(4):319–23.
93. Liou AP, Paziuk M, Luevano Jr JM, Machineni S, Turnbaugh PJ, Kaplan LM. Sci Transl
Med. 2013;5(178):178–41. doi:10.1126/scitranslmed.3005687.
94. Backhed F, Manchester JK, Semenkovich CF, Gordon J. Mechanisms underlying the resistance
to diet-induced obesity in germ-free mice. Proc Natl Acad Sci U S A. 2007;104(3):979–84.
95. Peng Y, Rideout DA, Rakita SS, Gower Jr WR, You M, Murr MM. Does LKB1 mediate
activation of hepatic AMP-protein kinase (AMPK) and sirtuin1 (SIRT1) after roux-en-Y gastric
bypass in obese rats? J Gastrointest Surg. 2010;14(2):221–8.
96. Bjursell M, Admyre T, Goransson M, Marley AE, Smith DM, Oscarsson J, Bohlooly-Y
M. Improved glucose control and reduced body fat mass in free fatty acid receptor 2-deficient
mice fed a high-fat diet. Am J Physiol Endocrinol Metab. 2011;300(1):E211–20.
97. Wang Z, Klipfell E, Bennett BJ, Koeth R, Levison BS, Dugar B, Feldstein AE, Britt EB, Fu
X, Chung YM, Wu Y, Schauer PR, Smith JD, Allayee H, Tang WH, DiDonato JA, Lusis AJ,
Hazen SL. Gut flora metabolism of phosphatidylcholine promotes cardiovascular disease.
Nature. 2011;472(7341):57–63.
98. Seyfried F, Li JV, Miras AD, Cluny NL, Lannoo M, Fenske WK, Sharkey KA, Nicholson JK,
le Roux CW, Holmes E. Urinary phenotyping indicates weight loss-independent metabolic
effects of roux-en-Y gastric bypass in mice. J Proteome Res. 2013;12(3):1245–53.
99. Tang WH, Wang Z, Levison BS, Koeth RA, Britt EB, Fu X, Wu Y, Hazen SL. Intestinal
microbial metabolism of phosphatidylcholine and cardiovascular risk. N Engl J Med.
2013;368(17):1575–84.
100. Nicholson JK, Holmes E, Wilson ID. Gut microorganisms, mammalian metabolism and
personalized health care. Nat Rev Microbiol. 2005;3(5):431–8.
101. Calvani R, Miccheli A, Capuani G, Tomassini Miccheli A, Puccetti C, Delfini M, Iaconelli A,

Nanni G, Mingrone G. Gut microbiome-derived metabolites characterize a peculiar obese
urinary metabotype. Int J Obes. 2010;34(6):1095–8.
102. Holmes E, Loo RL, Stamler J, Bictash M, Yap IK, Chan Q, Ebbels T, De Iorio M, Brown IJ,
Veselkov KA, Daviglus ML, Kesteloot H, Ueshima H, Zhao L, Nicholson JK, Elliott
P. Human metabolic phenotype diversity and its association with diet and blood pressure.
Nature. 2008;453(7193):396–400.
103. Ahmad NN, Pfalzer A, Kaplan LM. Roux-en-Y gastric bypass normalizes the blunted
postprandial bile acid excursion associated with obesity. Int J Obes. 2013;37(12):1553–9.
104. Glicksman C, Pournaras DJ, Wright M, Roberts R, Mahon D, Welbourn R, Sherwood R,
Alaghband-Zadeh J, le Roux CW. Postprandial plasma bile acid responses in normal weight
and obese subjects. Ann Clin Biochem. 2010;47(5):482–4.
105. Patti ME, Houten SM, Bianco AC, Bernier R, Larsen PR, Holst JJ, Badman MK, Maratos-
Flier E, Mun EC, Pihlajamaki J, Auwerx J, Goldfine AB. Serum bile acids are higher in
humans with prior gastric bypass: potential contribution to improved glucose and lipid
metabolism. Obesity. 2009;17(9):1671–7.
106. Pournaras DJ, Glicksman C, Vincent RP, Kuganolipava S, Alaghband-Zadeh J, Mahon D,
Bekker JH, Ghatei MA, Bloom SR, Walters JR, Welbourn R, le Roux CW. The role of bile
after roux-en-Y gastric bypass in promoting weight loss and improving glycaemic control.
Endocrinology. 2012;153(8):3613–9.
107. Katsuma S, Hirasawa A, Tsujimoto G. Bile acids promote glucagon-like peptide-1 secretion
through TGR5 in a murine enteroendocrine cell line STC-1. Biochem Biophys Res Commun.
2005;329(1):386–90.
108. Thomas C, Gioiello A, Noriega L, Strehle A, Oury J, Rizzo G, Macchiarulo A, Yamamoto H,
Mataki C, Pruzanski M, Pellicciari R, Auwerx J, Schoonjans K. TGR5-mediated bile acid
sensing controls glucose homeostasis. Cell Metab. 2009;10(3):167–77.
109. Ryan KK, Tremaroli V, Clemmensen C, Kovatcheva-Datchary P, Myronovych A, Karns R, et
al. Nature. 2014;509(7499):183–8. doi:10.1038/nature13135. Epub 2014.
110. Düfer M, Hörth K, Wagner R, Schittenhelm B, Prowald S, Wagner TF, Oberwinkler J,
Lukowski R, Gonzalez FJ, Krippeit-Drews P, Drews G. Bile acids acutely stimulate insulin
secretion of mouse beta-cells via farnesoid X receptor activation and K(ATP) channel
inhibition. Diabetes. 2012;61(6):1479–89.
111. Lundasen T, Galman C, Angelin B, Rudling M. Circulating intestinal fibroblast growth factor
19 has a pronounced diurnal variation and modulates hepatic bile acid synthesis in man.
J Intern Med. 2006;260(6):530–6.
112. Watanabe M, Houten SM, Mataki C, Christoffolete MA, Kim BW, Sato H, Messaddeq N,
Harney JW, Ezaki O, Kodama T, Schoonjans K, Bianco AC, Auwerx J. Bile acids induce energy
expenditure by promoting intracellular thyroid hormone activation. Nature. 2006;439(7075):484–9.
113. Ockenga J, Valentini L, Schuetz T, Wohlgemuth F, Glaeser S, Omar A, Kasim E, duPlessis D,
Featherstone K, Davis JR, Tietge UJ, Kroencke T, Biebermann H, Köhrle J, Brabant
G. Plasma bile acids are associated with energy expenditure and thyroid function in humans.
J Clin Endocrinol Metab. 2012;97(2):535–42.
114. Flum DR, Belle SH, King WC, Wahed AS, Berk P, Chapman W, Pories W, Courcoulas A,
McCloskey C, Mitchell J, Patterson E, Pomp A, Staten MA, Yanovski SZ, Thirlby R, Wolfe
B. Perioperative safety in the longitudinal assessment of bariatric surgery. N Engl J Med.
2009;361(5):445–54.
115. Sjöström L, Narbro K, Sjöström CD, Karason K, Larsson B, Wedel H, Lystig T, Sullivan M,
Bouchard C, Carlsson B, Bengtsson C, Dahlgren S, Gummesson A, Jacobson P, Karlsson J,
Lindroos AK, Lönroth H, Näslund I, Olbers T, Stenlöf K, Torgerson J, Agren G, Carlsson
LM. Effects of bariatric surgery on mortality in Swedish obese subjects. N Engl J Med.
2007;357(8):741–52.
116. Service GJ, Thompson GB, Service FJ, Andrews JC, Collazo-Clavell ML, Lloyd
RV. Hyperinsulinemic hypoglycemia with nesidioblastosis after gastric-bypass surgery.
N Engl J Med. 2005;353(3):249–54.
117. Goldfine AB, Mun EC, Devine E, Bernier R, Baz Hecht M, Jones DB, Schneider BE, Holst
JJ, Patti ME. Patients with neuroglycopenia after gastric bypass surgery have exaggerated
incretin and insulin secretory responses to a mixed meal. J Clin Endocrinol Metab.
2007;92(12):4678–85.
118. Tucker ON, Szomstein S, Rosenthal RJ. Nutritional consequences of weight-loss surgery.
Med Clin N Am. 2007;91(3):499–514. xii.
119. Davies DJ, Baxter JM. Nutritional deficiencies after bariatric surgery. Obes Surg.
2007;17(9):1150–8.
120. Ritz P, Becouarn G, Douay O, Salle A, Topart P, Rohmer V. Gastric bypass is not associated
with protein malnutrition in morbidly obese patients. Obes Surg. 2009;19(7):840–4.
121. Angstadt JD, Bodziner RA. Peripheral polyneuropathy from thiamine deficiency following
laparoscopic roux-en-Y gastric bypass. Obes Surg. 2005;15(6):890–2.
Index
A Biomarkers, 3, 15, 26, 27, 32–38, 46, 65, 66,

Adaptive immunity, 5, 178, 209, 271, 273 68–70, 73, 75, 76, 78, 90–92, 121, 132,
Aging, vi, 2, 10–12, 30, 32–33, 37, 38, 134, 136, 137, 139, 170, 173, 176, 180,
101, 131, 137, 169–180, 201, 215 202, 203, 207, 208, 210, 219–221, 279,
Alpha oxidation, 111, 115 316–319
Analytics, 2, 9, 10, 26, 29, 30, 46, 69, 71, Branched-chain fatty acids (BCAA),
75, 121, 139, 149, 152, 153, 170, 6–7, 12, 33, 34, 36, 84, 89, 136,
238–242, 255 137, 207
Antibiotics, 94, 100, 127, 128, 130, 175, 186, Butyrate, 84, 89, 92, 95, 96, 129, 135,
189, 190, 211, 218, 219, 253, 267, 269, 175–177, 180, 204, 207, 209, 210, 215,
270, 277, 291, 292, 313, 315, 323, 324, 216, 246, 250, 251, 265, 276, 282, 286,
326–331, 337 290–291, 359
Antigen-presenting cells (APCs),
209, 263–265, 273
Autism, v, vi, 38, 95, 97, 177, 186, 311–320 C
Automation, 10, 31, 36, 37, 109–122, Cancer, 36–38, 55, 86, 92, 95, 112, 126, 139,
152, 246, 328 172, 175–180, 186–188, 190, 205, 270,
332, 333, 344
Carbohydrates, vi, 4, 11, 13, 77, 84–90,
B 92–96, 112, 133, 154, 177, 207, 242,
Bacterial translocation, 209, 218–221 249, 251, 252, 262, 279, 288, 289, 352,
Bariatric surgery, 7, 202, 216, 217, 248, 354, 359, 362
343–363 Cardiovascular disease (CVD), 6, 7, 14, 15,
Barrier, 8, 85, 92, 100, 149, 161, 173, 30, 35, 90–92, 95, 131, 172–173, 177,
202, 209, 213, 214, 218, 266–269, 178, 186, 200, 207, 208, 219–221, 242,
271–272, 276, 280, 281, 284, 285, 343–363
287, 289, 331 Carnitine, 14, 15, 33, 35, 89–92, 137, 138,
Beta oxidation, 66, 111, 115, 246 216, 243, 290
Bile acids, 7, 11–13, 33–35, 70, 86, 95, Chemometrics, 9, 46, 47, 72, 78
202, 207, 209, 210, 214, 217, 221, Child, 2, 6, 30, 32, 34, 36, 100,
237, 245, 246, 250, 252–254, 280, 127, 132, 136–139, 189,
281, 289, 315, 317, 318, 357, 211, 212, 216, 288, 314–318,
361–363 328, 334, 344
Bioinformatics tools, 73, 149, 246 Chocolate, 74–75, 92–93, 97, 98

DOI 10.1007/978-1-4471-6539-2
372 Index
Choline, 7, 14, 34, 35, 37–38, 73–74, 89–92, Drug metabolism, 3, 71, 72, 94, 97, 244, 324,
95, 101, 133, 137, 171, 172, 177–178, 325, 327, 335, 337
189–190, 210, 214, 215, 219, 244, 245, Drugs, 3, 5, 11, 26, 56, 64–68, 70–72, 75–78,
283, 319, 360 94, 97, 126, 131, 132, 188–190, 192,
Classification and regression trees (CARTs), 203, 220, 244, 267, 290, 291, 316,
48–50, 54, 57 323–338, 344, 349
Colon, 11–14, 16, 71, 77, 84–85, 92, 93,
95–97, 100, 101, 103, 110, 112, 115,
126, 128–133, 136, 148, 161, 164, E
174–176, 180, 186, 189, 201, 202, Ecology, v, vi, 3–5, 7–9, 14, 125–139, 173,
206, 208, 210–212, 215, 242, 243, 180, 201, 202, 206, 208, 211–213, 217,
245, 250, 251, 262–263, 265–269, 266, 267, 274, 275, 277, 291, 324
274, 276, 277, 285, 286, 288, 290, Elemental diet, 288
313, 315, 326, 328, 329, 332, 335, Enteral nutrition, 288, 289
336, 355, 357, 359, 360 Epicatechin, 98, 111, 336
p-Cresol sulphate, 71, 92, 97, 172, 337 Evolution, 8, 51, 85, 149, 151, 152, 157, 159,
Crohn’s disease, 86, 97, 131, 135, 179, 189, 160, 162, 164, 179, 238, 245
214, 267, 268, 273
F
D Fat, vi, 4, 6, 7, 11, 13, 14, 33–35, 38, 66, 77,
Data integration, 55–56, 58, 59 84, 89–91, 93, 95–96, 100, 101, 112,
Deconjugation, 252, 253, 332, 357, 361 134–138, 164, 175, 193, 200, 202–206,
Dendritic cells (DCs), 212, 218, 263–266, 208–210, 213–219, 221, 237, 242–246,
271, 273 250–252, 265, 276, 280, 281, 288–291,
Development, v, vi, 2, 3, 5, 7–11, 14–16, 319, 330, 333, 345, 347, 348, 352–357,
27, 30, 32–35, 38, 46, 59, 69, 75, 78, 359–361, 363
86, 90, 100, 113, 125–139, 147–164, Fecal microbial transplantation (FMT),
170–173, 175, 178–180, 187, 192, 193, 291–292
200–204, 206, 209, 210, 212, 215, 219, Fibers, vi, 4, 84, 86, 89, 92, 95, 96, 102, 174,
221, 236, 245–247, 252, 263, 267, 269, 189, 200, 209, 286, 289, 348
273, 274, 276, 285, 288, 289, 311–320, Flavonoids, vi, 87, 89, 109–122, 189
324, 328, 338, 349, 354, 357 Foods, vi, 4–7, 31, 33, 64, 66, 74–76,
Diabetes, 2, 6–9, 30, 33–35, 38, 65, 66, 71, 84–95, 97–99, 102, 110, 121, 127,
78, 90, 95, 97, 100, 126, 136–139, 170, 131–132, 139, 149–151, 162, 164,
172, 177, 178, 186, 189–190, 200–210, 170, 176, 186, 189, 202–204, 206,
213–215, 218, 219, 221, 238, 242, 247, 208, 215–217, 242, 243, 247, 251,
249, 250, 254, 343–363 252, 283, 285, 286, 289, 319, 324,
Diet, 2–9, 11–15, 26, 33–35, 38, 65, 347–349, 352–356, 359
71–77, 84–93, 95, 97, 101–103,
110–113, 121, 127, 129–132, 135,
136, 148–152, 162, 170–174, 176, G
179, 189, 200, 203, 206, 208–210, Gallocatechins, 115, 116
213–221, 236, 243, 244, 246–249, Genome-wide association studies (GWAS),
252, 262, 263, 267, 276, 281, 285–291, 15, 57, 58, 66, 78, 207, 249, 254, 269,
316, 317, 319, 324, 329, 332, 350, 271–274
353–357, 359–361 Glucose (Glc), 34, 35, 38, 86, 89, 95, 134,
Digestion, 4, 8, 11, 14, 30, 77, 84–86, 92, 103, 138, 154, 155, 171, 172, 190, 203–206,
115, 121, 129, 148, 149, 152–154, 161, 208, 214–216, 242, 243, 245, 246,
164, 173, 176, 179, 186, 189, 217, 247, 249–252, 349–357, 359–363
251, 252, 286, 288, 312, 313, 324, 348, Glucuronide, 71, 94, 100, 101, 110, 115–121,
351, 353, 355, 359, 361 177, 325, 332, 333, 337
Diseases, 2, 26, 58, 63, 86, 126, 148, 170, 186, Glycans, 148, 149, 151–162, 164, 203, 204,
200, 238, 266, 312, 323, 344 209, 214, 218, 220
Index 373
Glycobiology, 149, 152, 153 Inflammation, v, 6, 8, 11, 33–35, 66, 86, 90,
Glycomics, 148, 151–156, 164 95–96, 100, 126, 134–136, 164,
Glycosidase, 115, 153, 157, 159, 160 170–175, 177–179, 186, 188–189, 202,
Gnotobiology, 262 203, 205, 206, 210, 212–215, 217, 218,
Gnotobiotics, 11, 242, 246, 247, 252, 220, 246, 251, 252, 254, 261–292, 326,
313, 329 332, 349, 351, 354–361
Growth, 10–12, 27, 31, 48–50, 55, 85, Inflammatory bowel disease (IBD), v, vi, 8,
100, 125–139, 148, 149, 157–162, 14, 66, 126, 135, 139, 175, 178, 179,
172, 174, 186, 189, 191, 212, 217, 186–189, 261–292, 313
218, 253, 263, 276–278, 281, 286, Innate immunity, 206, 215, 267, 272–273
287, 289, 327, 328 Insulin resistance (IR), 7, 33–35, 38, 95, 131,
Gut, 2, 30, 63, 84, 112, 126, 157, 170, 186, 136, 137, 204, 205, 212–214, 216, 219,
200, 235, 262, 312, 323, 346 238, 349, 350, 353, 355–357, 363
Gut bacteria, 9, 13, 74, 86, 95, 97, 100, 208, Intestinal health, 4, 273, 274, 280, 289, 291, 292
210, 213, 330, 332, 355–357 Intestinal microbiota, 14, 126–130, 132, 174,
Gut–brain, 101, 176, 203, 312–315, 319, 320 180, 201, 202, 204–206, 210, 213, 215,
Gut–brain axis, 203, 312–315, 319, 320 217, 218, 222, 248–250, 252, 266, 268,
Gut–liver, 101–102 269, 271, 275, 278, 285–291, 312, 313,
Gut microbiota, 2, 36, 65, 84, 112, 126, 148, 326, 327, 329, 334
170, 186, 200, 235, 262, 312, 323, 354 Intestine, 4, 84, 110, 126, 148, 173, 174, 189,
201, 242, 262, 312, 326, 347
Iron, 289
H
Health, v–vii, 2–6, 8–12, 14, 16, 26, 27,
30, 32, 33, 36–38, 45–59, 64, 66, L
67, 71–76, 78, 85, 95, 102, 103, Lactation, 133, 148, 149, 151, 152, 154,
110, 113, 121, 126, 127, 129–133, 157, 162
148–152, 154, 162, 164, 169–180, Lactone hydrolysis, 111–112
185–194, 203, 206, 208, 218, 221, Lipid characterization, 238, 240, 246
237, 238, 242, 248, 249, 252, 254, Lipid metabolism, vi, 32, 33, 66, 92, 193, 205,
255, 261–292, 312–318, 323, 344, 209, 213, 219, 235–255, 354, 357
346, 348, 349, 355–358, 361 Lipidomics, 26–29, 34–36, 70, 172, 173,
Herbal medicine, vi, 185–194 238–242, 244, 248, 249, 255
Hippurate, 36, 74, 87, 90, 93, 97, 98, 100, 135, Liquid chromatography–mass spectrometry
192, 317, 318, 360–361 (LC–MS), 26, 28, 29, 69, 90, 113–115,
Hippuric acid, 93, 94, 97, 98, 111, 113 117, 238–240, 337
Longevity, 2, 171, 172, 176, 180
I
Identification, 3, 26, 29, 30, 33, 36, 38, M
52, 56, 57, 64, 69, 70, 75, 113–115, MAGMA, 117, 121
119, 122, 132, 136, 153, 161, 180, Mass spectrometry (MS), 3, 26, 28, 29, 35–38,
188, 193, 201–203, 207, 208, 222, 46, 47, 64, 69–70, 78, 90, 94, 95,
239, 240, 250, 269, 285, 313 113–118, 121, 132–136, 138, 149, 152,
Immune system, 5, 10–11, 126, 149, 170, 174, 153, 170, 172, 173, 180, 187, 238–241,
178, 202, 205, 211, 212, 214, 218–222, 244, 312, 317, 337
236, 262, 263, 265–268, 271, 274, 291, Mental health, 30, 172, 173
312, 313, 357 Metabolic diseases, vi, 2, 6, 9, 10, 90,
Immunometabolism, 203, 205–210, 189–190, 199–222, 238, 247, 344,
214–215, 217 354–361, 363
Incretins, 221, 353 Metabolic infection, 200, 203–206,
Indole metabolites, 100–101 208–210, 212
Infants, vi, 87–89, 126–130, 133, 136, 138, Metabolic inflammation, 202, 204–206, 209,
147–164, 201, 211, 212, 287, 328, 336 212–215, 218
374 Index
Metabolic syndrome, 6, 7, 33–35, 90, 137, Nutrition, v, vi 2–6, 8–12, 14–16, 25–27, 31,
213, 242, 245, 250, 254, 356 32, 38, 63–78, 83–103, 110, 129–139,
Metabolism, 2, 27, 66, 84, 112, 132, 148, 171, 162, 164, 170, 171, 173, 174, 176,
186, 202, 236, 262, 313, 324, 353 178–180, 186, 200, 202, 206, 208,
Metabolites, 2, 26, 59, 64, 84, 110, 213, 221, 236, 247, 285, 288, 289,
130, 149, 170, 187, 216, 236, 291, 319, 344, 349, 363
263, 312, 324, 357 Nutrition aging, 171, 172, 174–176
Metabolomics, 5, 9, 16, 38, 45–48, 51, 56,
63–64, 83–103, 113, 121, 147–164,
191, 193, 236, 250, 254 O
Metabonomics, 1–16, 25–38, 63–78, 125–139, Obesity, v, 6, 7, 33, 35, 65, 71, 72, 95, 97, 126,
169–180, 187, 188, 191, 193, 194, 212, 131, 137, 138, 175, 177, 186, 189,
238, 242, 252, 275, 279, 280, 283–285, 200–204, 206, 207, 210–218, 243,
291, 316–319 245–248, 278, 313, 343–363
Metagenomics, 10, 68, 77, 161, 178, 201, Oligosaccharides, 84, 129, 130, 148, 149,
207–210, 235, 238, 242, 249–250, 252, 152–164, 286
254, 275, 278–279, 284 One carbon metabolism, 92
Metaproteomics, 16, 275, 278–280, 282,
284, 285
Metataxonomics, 250 P
Metatranscriptomics, 275, 278, 279, 284 Pathobionts, 174, 175, 281, 288–289
MetIDB, 115, 122 Pathways, 1, 2, 5, 7, 11, 16, 25, 29, 33, 36,
Microbiome, v, vi, 4, 5, 10–13, 16, 26, 55, 58, 55–56, 58–60, 69, 70, 76–78, 87, 92,
59, 65, 68, 71, 72, 75–77, 86, 94, 131, 94, 98, 101, 110–112, 126, 137, 139,
133–134, 139, 149, 175, 178, 180, 201, 159, 160, 162, 173, 193, 207, 208, 212,
207, 208, 210, 211, 221, 222, 235, 236, 236–238, 242, 244, 245, 247, 250,
243, 244, 249–251, 253, 262, 278, 279, 252–255, 272, 278, 279, 291, 312, 324,
315, 316, 318, 319, 323–338, 346, 337, 357, 362
354–357, 359, 363 PERCH NMR Software, 117
Microbiota, 3, 36, 65, 84, 112, 126, 148, 170, Permeability, 8, 9, 12, 90, 179, 207, 209, 210,
186, 200, 235, 262, 312, 323, 354 213, 214, 218, 280, 357
Microflora, 90, 94, 97, 102, 112, 113, 115, Personalized medicine, vi, 63–78, 255, 324, 338
127, 128, 193 Personalized nutrition, vi, 9, 12, 32, 63–78,
Milk, 31, 101, 126, 129, 130, 132–135, 255, 291
147–164, 287 Pharmacogenetics, 65, 72
Mouse models, 34, 35, 91, 134, 135, 243, 246, Pharmacometabonomics, 64, 67–68, 71–73,
281, 286, 287, 289 77, 78, 132, 139
Phase II metabolism, 87, 94, 101, 244
Phenolic and phenyl metabolites, 97–100
N Phylogeny, 207, 277–280
Network modelling, 58, 59 Polyphenols, vi, 85, 92–93, 97, 111–112, 186,
Newborn, 127–130, 137, 201 187, 193, 209, 221, 360
Nuclear magnetic resonance (NMR), 3, 9, 12, Prebiotics, 86, 93, 129, 130, 161–162,
13, 26–29, 35, 37, 38, 46, 64, 69–72, 189, 209–210, 213, 252, 285–287,
74, 78, 94, 113, 114, 117–122, 291, 357, 360
132–135, 138, 139, 170–172, 180, 187, Probiotics, 75, 86, 92, 130, 189, 213, 248, 283,
192, 241, 244, 245, 317 286, 287, 291
Nuclear magnetic resonance (NMR) Prodrug activation, 325, 330
spectroscopy, 3, 12, 27, 28, 35, 64, 69, Profiling, 3, 10, 12, 16, 33, 35, 64, 67, 69,
70, 72, 114–120, 122, 133, 170, 187, 71–73, 75–78, 94, 113–114, 121, 122,
191, 192, 241 135, 139, 171, 172, 179, 205, 238–241,
Nutrigenetics, 289–291 244, 248, 316, 328–329
Index 375
Protein, vi, xv, 2, 3, 9, 12, 16, 25, 32, 34, 56, Toxicity, vii, 3, 5, 64–68, 70–72, 77, 78,
65, 72, 75, 84, 86, 92–94, 96, 97, 100, 90–92, 112, 149, 213, 215, 263,
101, 132, 148, 149, 152, 157, 159, 160, 323–338
171, 175, 179, 203, 213, 215, 237, Trimethylamine oxide (TMAO), 14–15,
242–245, 249, 262, 265, 271, 272, 274, 90–92, 101, 132, 204, 219–220, 360
278, 279, 284, 285, 329, 336, 362, 363
U
R Ulcerative colitis (UC), 14, 97, 135, 179, 189,
Random forests, 50–56, 58, 187 267–269, 272–278, 281–283, 285, 287,
290–292, 326
Urolithin, 111–112, 119–121
S
Short chain fatty acids (SCFAs), 4, 11, 12, 66,
77, 84, 86, 92, 95–96, 136, 174–177, V
180, 204, 215–216, 221, 250–252, 265, Valerolactone, 110, 111, 115–118
279, 280, 282, 286, 357, 359–360
Signaling molecules, 13, 77, 93, 96, 100,
250–254, 312, 313, 324 W
Sulphate, 110, 112, 115, 117 Weight loss, 7, 8, 208, 248, 344, 346–351,
Systems biology, v, vi, 1–16, 38, 63, 65, 70, 353, 354, 356, 361–363
73, 76, 131, 137, 164, 173, 178, 180, Whole grain cereals, 86
236, 238, 245, 255, 320
X
T Xenobiotics, 32, 71, 85, 97, 101, 112, 113,
Tissue microbiota, 203, 205, 206, 209, 174, 207, 290, 291, 324, 325, 332,
220–221 334–337

Metabonomics and Gut Microbiota in Nutrition and Disease: Sunil Kochhar François-Pierre Martin Editors

Uploaded by

Copyright:

Available Formats

Metabonomics and Gut Microbiota in Nutrition and Disease: Sunil Kochhar François-Pierre Martin Editors

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Metabonomics and Gut Microbiota in Nutrition and Disease: Sunil Kochhar François-Pierre Martin Editors

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What topics are covered in the book?

What topics are covered in the book?

Molecular and Integrative Toxicology

Metabonomics and Gut

ISSN 2168-4219 ISSN 2168-4235 (electronic)

Library of Congress Control Number: 2014953791

© Springer-Verlag London 2015

Printed on acid-free paper

Humana Press is a brand of Springer

Chapters 15 and 16 review the applications towards future therapeutical,

Lausanne, Switzerland Sunil Kochhar

1 Introduction to Metabonomics in Systems Biology Research ............. 1

9 Metabonomics and Gut Microbial Paradigm

Index ................................................................................................................. 371

Jacques Amar Institut National de la Santé et de la Recherche Médicale

Claudio Franceschi Department of Experimental Diagnostic and Specialty

Ivan Montoliu Analytical Sciences, Applied Mathematics, Nestec Ltd., Nestlé

François-Pierre Martin, Ph.D., Eng., is Senior Scientist and Research Program

of “nutritional metabonomics” to assess metabolic health and its early deviations in

François-Pierre Martin and Sunil Kochhar

Abstract There is a growing interest to understand the paradigm of healthy aging

F.-P. Martin (*)

© Springer-Verlag London 2015 1

Keywords Disease • Gut microbiota • Health • Mass spectrometry • Metabonomics

Socio-economics Lifestyle, environmental factors

Host genome & metabolism

Metabolic <Personalized Healthcare and Nutrition>

microbial activities were shown to be extremely specific, as exemplified via its

1.2 Metabonomics-Based Systems Approaches:

Over the last three decades, the majority of metabonomic-based investigations in

exchanges between diverse biological compartments, including organs, systemic

amino acid (BCAA)-related metabolic signature was highlighted as a robust

In the meantime, it is expected that analytical technologic developments will enable

1.3 Gut Microbiota Metabolism in Nutrition,

investigations illustrated how microbial-dependent variations along the upper

1.4 System Biology Approaches: Applications with Specific

In addition to the building awareness that specific environmental exposure

1.5 Novel Directions and Perspectives

Metabonomics is increasingly becoming popular in clinical research due to its

The generalization of systems biology approaches will subsequently lead to the

Sebastiano Collino, François-Pierre Martin, and Sofia Moco

Abstract Metabonomics is recognized as a powerful top-down system biology

Keywords Aging • Biomarkers • Cancer • Metabonomics • Lipidomics • Metabolic

S. Collino (*) • F.-P. Martin

© Springer-Verlag London 2015 25

2.2 State-of-the-Art Metabonomics and Lipidomics

Metabonomics is able to generate multivariate information on a wide range of

cryoprobes has strongly improved NMR sensitivity to generally a fourfold factor

2.3 Metabonomics Flow Chart in Epidemiological Studies

Epidemiological studies aim to substantiate the causes, distribution, and control of

analysis is also applicable to food itself to identify major nutritional constituents

2.4 Clinical Applications of Metabonomics

2.5 Clinical Application of Metabonomics

The continuously increasing prevalence of obesity in many countries around the

Nonalcoholic fatty liver disease (NAFLD) is increasingly considered as a main

2.6 Clinical Application for Neurological

Diagnostic markers of clinical metabonomics can also find applications in sociopsy-

3.2 Expanding the View: Machine Learning

3.6 I ntegration Between Information Sources: RF

3.7 Application-Driven Improvements to RF Scheme

3.8 New Challenges to RF: Network Inference

3.9 Conclusions and Perspective