Irit Sagi - Nikolaos A Afratis - Collagen - Methods and Protocols-Humana Press (2019)

Methods in
Molecular Biology 1944
Irit Sagi · Nikolaos A. Afratis Editors
Collagen
Methods and Protocols
METHODS IN MOLECULAR BIOLOGY
Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK
For further volumes:

http://www.springer.com/series/7651
Collagen
Methods and Protocols
Edited by
Irit Sagi and Nikolaos A. Afratis

Department of Biological Regulation, Weizmann Institute of Science, Rehovot, Israel
Editors
Irit Sagi Nikolaos A. Afratis
Department of Biological Regulation Department of Biological Regulation
Weizmann Institute of Science Weizmann Institute of Science
Rehovot, Israel Rehovot, Israel
ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-4939-9094-8 ISBN 978-1-4939-9095-5 (eBook)
https://doi.org/10.1007/978-1-4939-9095-5
Library of Congress Control Number: 2019930144
© Springer Science+Business Media, LLC, part of Springer Nature 2019

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.
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.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to
be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors or omissions that may have been made.
The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Humana Press imprint is published by the registered company Springer Science+Business Media, LLC, part of
Springer Nature.
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
Preface
Collagen is the most abundant protein in mammals, making up about 25% of the whole-
body protein content. Collagen is involved in many physiological and pathological situa-
tions. Specifically, collagen has been shown to be involved either directly or indirectly in the
promotion of cell adhesion and differentiation, cell functions that are related to the control
of embryonic development and regeneration. However, the uncontrolled synthesis and
accumulation of collagen from fibroblasts lead to pathological conditions in fibrotic-related
diseases and tumorigenesis.
Collagen is the main structural protein in extracellular matrix and connective tissue. The
basic fundamental structure consists of three coiled subunits: two α1(I) chains and one α2
(I). There are at least 26 different types of collagen, and all of them contain three-stranded
helical segments of similar structure. The differentiation between each type of collagen is
due to segments that interrupt the triple helix and that fold into other kinds of three-
dimensional structures. Collagen types I, II, and III consist the 80% of the collagen in
the body.
Collagen is widely used in many applications like food, cosmetics, pharmaceuticals,
cosmetic surgery, artificial skin, and glue (the name collagen comes from the
Greek κóλλα (kolla), meaning “glue,” and suffix -γεν, -gen, denoting “producing”). The
importance of collagen in health and disease and the interest of the researchers to further
investigate the in vitro and in vivo mechanisms of collagen were the objective reasons that
motivated us to create this Collagen: Methods and Protocols book. This volume compiles a
collection of state-of-the-art protocols that will serve not only as recipes for bench scientists
but also as accepted methods for the collagen field. Each chapter describes a specific topic
and technique by an expert in the field and provides a detailed protocol, which includes
some most helpful tips in the Notes section.
The content of the book is separated into four different parts. Specifically, Part I focuses
on in vitro models for the characterization of collagen formation. In detail, it describes a
mimetic model for mineralization of type-I collagen fibrils, cross-linked collagen scaffolds,
and methods to clone, mutate, and isolate collagen. Part II highlights a large-scale analysis
of collagen with mass spectrometry in order to elucidate the proteomics, degradomics,
interactomes, and cross-linking of collagen. It also presents the triple-helix hydroxylation
of collagen by solid-state NMR spectroscopy.
Furthermore, Part III describes high-resolution imaging approaches for collagen by the
use of scanning electron microscopy and multiphoton imaging. In addition, a new tool for
degraded and denatured collagen histopathology and collagen-binding integrin analysis is
presented. Finally, Part IV introduces the role of collagen during physiological and patho-
logical conditions and provides an overview of collagenous matrix in tumor microenviron-
ment. The reader will find protocols to analyze the collagen levels from primary fibroblasts,
to produce collagen-based peptides for inhibition of matrix metalloproteinases, and to assess
collagen deposition during aging and liver fibrosis.
v
vi Preface
We hope that this volume on collagen will be a useful tool for the scientists working in
the field and will inspire them to develop new techniques and models that will create the next
generation of collagen protocols. We would also like to thank all the authors and co-authors
that participated on this book for their tremendous contributions. It was a great pleasure to
work with all of them, and we believe that their effort was critical for the high quality and
repeatability of the research protocols.
Rehovot, Israel Irit Sagi

Nikolaos A. Afratis
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
PART I IN VITRO MODELS FOR COLLAGEN CHARACTERIZATION

1 Cloning and Mutagenesis Strategies for Large Collagens. . . . . . . . . . . . . . . . . . . . . 3
Olivier Bornert and Alexander Nyström
2 Multi-well Fibrillation Assay as a Tool for Analyzing Crosslinking
and Stabilization of Collagen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Venkat Raghavan Krishnaswamy
3 Production and Characterization of Chemically Cross-Linked
Collagen Scaffolds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Ignacio Sallent, Héctor Capella-Monsonı́s, and Dimitrios I. Zeugolis
4 A Biomimetic Model for Mineralization of Type-I Collagen Fibrils. . . . . . . . . . . . 39
Shasha Yao, Yifei Xu, Changyu Shao, Fabio Nudelman,
Nico A. J. M. Sommerdijk, and Ruikang Tang
PART II PROTEOMICS ANALYSIS OF COLLAGEN
5 Investigation of Triple-Helix Collagen Hydroxylation

by Solid-State NMR Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Wing Ying Chow
6 Measurement of Collagen Cross-Links from Tissue Samples
by Mass Spectrometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Amar Joshi, Amna Zahoor, and Alberto Buson
7 Mass Spectrometry-Based Proteomics to Define Intracellular
Collagen Interactomes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Ngoc-Duc Doan, Andrew S. DiChiara, Amanda M. Del Rosario,
Richard P. Schiavoni, and Matthew D. Shoulders
8 Exploring Extracellular Matrix Degradomes
by TMT-TAILS N-Terminomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
Elizabeta Madzharova, Fabio Sabino, and Ulrich auf dem Keller
PART III IMAGING APPROACHES FOR COLLAGEN
9 Sample Preparation of Extracellular Matrix of Murine

Colons for Scanning Electron Microscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
Elee Shimshoni and Irit Sagi
10 In Situ Detection of Degraded and Denatured Collagen
via Triple Helical Hybridization: New Tool in Histopathology . . . . . . . . . . . . . . . 135
Yang Li and S. Michael Yu
vii
viii Contents
11 Combination of Traction Assays and Multiphoton Imaging

to Quantify Skin Biomechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Stéphane Bancelin, Barbara Lynch, Christelle Bonod-Bidaud,
Petr Dokládal, Florence Ruggiero, Jean-Marc Allain,
and Marie-Claire Schanne-Klein
12 Analysis of Collagen-Binding Integrin Interactions
with Supramolecular Aggregates of the Extracellular Matrix . . . . . . . . . . . . . . . . . . 157
Uwe Hansen
PART IV COLLAGEN IN HEALTH AND DISEASE
13 Assessing Collagen Deposition During Aging in Mammalian

Tissue and in Caenorhabditis elegans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Alina C. Teuscher, Cyril Statzer, Sophia Pantasis, Mattia R. Bordoli,
and Collin Y. Ewald
14 Procollagen C-Proteinase Enhancer 1 (PCPE-1) in Liver Fibrosis. . . . . . . . . . . . . 189
Efrat Kessler and Eyal Hassoun
15 Tumorigenic Interplay Between Macrophages and Collagenous
Matrix in the Tumor Microenvironment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Chen Varol
16 FACS Analysis of Col1α Protein Levels in Primary Fibroblasts. . . . . . . . . . . . . . . . 221
Noam Cohen and Neta Erez
17 Methods for the Construction of Collagen-Based Triple-Helical
Peptides Designed as Matrix Metalloproteinase Inhibitors . . . . . . . . . . . . . . . . . . . 229
Gregg B. Fields
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
Contributors
JEAN-MARC ALLAIN Laboratoire de Mécanique des Solides (LMS), Ecole Polytechnique,

CNRS, Université Paris-Saclay, Palaiseau, France; Inria, Université Paris-Saclay,
Palaiseau, France
ULRICH AUF DEM KELLER Department of Biotechnology and Biomedicine, Technical
University of Denmark, Kongens Lyngby, Denmark
STÉPHANE BANCELIN Laboratoire d’Optique et Biosciences (LOB), Ecole Polytechnique,
CNRS, INSERM, Université Paris-Saclay, Palaiseau, France; Institut Interdisciplinaire
de Neurosciences, CNRS, Université de Bordeaux, Bordeaux, France
CHRISTELLE BONOD-BIDAUD Institut de Génomique Fonctionnelle de Lyon, ENS-Lyon,
CNRS UMR 5242, Université de Lyon, Lyon, France; Tissue Biology and Therapeutic
Engineering Unit, LBTI, CNRS UMR 5305, Univ. Lyon, Université Claude Bernard
Lyon 1, Lyon, France
MATTIA R. BORDOLI Department of Biology, Institute of Molecular Health Sciences,
Eidgenössische Technische Hochschule (ETH) Zürich, Zürich, Switzerland
OLIVIER BORNERT Department of Dermatology, Faculty of Medicine, Medical Center,
University of Freiburg, Freiburg, Germany
ALBERTO BUSON Pharmaxis Drug Discovery, Frenchs Forest, NSW, Australia
HÉCTOR CAPELLA-MONSONÍS Regenerative, Modular and Developmental Engineering
Laboratory (REMODEL), National University of Ireland Galway (NUI Galway),
Galway, Ireland; Science Foundation Ireland (SFI) Centre for Research in Medical Devices
(CÚRAM), National University of Ireland Galway (NUI Galway), Galway, Ireland
WING YING CHOW Leibniz-Forschungsinstitut für Molekulare Pharmakologie (FMP),
Berlin, Germany
NOAM COHEN Department of Pathology, Sackler School of Medicine, Tel Aviv University,
Tel Aviv, Israel
AMANDA M. DEL ROSARIO Koch Institute, Massachusetts Institute of Technology,
Cambridge, MA, USA
ANDREW S. DICHIARA Department of Chemistry, Massachusetts Institute of Technology,
Cambridge, MA, USA
NGOC-DUC DOAN Department of Chemistry, Massachusetts Institute of Technology,
Cambridge, MA, USA
PETR DOKLÁDAL Centre for Mathematical Morphology, MINES ParisTech, PSL Research
University, Fontainebleau, France
NETA EREZ Department of Pathology, Sackler School of Medicine, Tel Aviv University,
Tel Aviv, Israel
COLLIN Y. EWALD Department of Health Sciences and Technology, Institute of Translational
Medicine, Eidgenössische Technische Hochschule (ETH) Zürich, Zürich, Switzerland
GREGG B. FIELDS Department of Chemistry and Biochemistry, Florida Atlantic University,
Jupiter, FL, USA; Department of Chemistry, Scripps Research, Jupiter, FL, USA
UWE HANSEN Institute of Musculoskeletal Medicine, University Hospital Münster, Münster,
Germany
EYAL HASSOUN Clinical Biochemistry and Pharmacology Institute, Tel Aviv Sourasky
Medical Center, Tel Aviv, Israel
ix
x Contributors
AMAR JOSHI Pharmaxis Drug Discovery, Frenchs Forest, NSW, Australia

EFRAT KESSLER Maurice and Gabriela Goldschleger Eye Research Institute, Sheba Medical
Center, Sackler Faculty of Medicine, Tel Aviv University, Tel-Hashomer, Israel; Clinical
Biochemistry and Pharmacology Institute, Tel Aviv Sourasky Medical Center, Tel Aviv,
Israel
VENKAT RAGHAVAN KRISHNASWAMY Department of Biological Regulation, Weizmann
Institute of Science, Rehovot, Israel
YANG LI Department of Bioengineering, University of Utah, Salt Lake City, UT, USA;
3Helix Inc., Salt Lake City, UT, USA
BARBARA LYNCH Laboratoire de Mécanique des Solides (LMS), Ecole Polytechnique, CNRS,
Université Paris-Saclay, Palaiseau, France; L’Oréal Research and Innovation, Aulnay-
sous-Bois, France
ELIZABETA MADZHAROVA Department of Biotechnology and Biomedicine, Technical
University of Denmark, Kongens Lyngby, Denmark
FABIO NUDELMAN School of Chemistry, University of Edinburgh, Edinburgh, UK
ALEXANDER NYSTRÖM Department of Dermatology, Faculty of Medicine, Medical Center,
University of Freiburg, Freiburg, Germany
SOPHIA PANTASIS Department of Biology, Institute of Molecular Health Sciences,
Eidgenössische Technische Hochschule (ETH) Zürich, Zürich, Switzerland
FLORENCE RUGGIERO Institut de Génomique Fonctionnelle de Lyon, ENS-Lyon, CNRS
UMR 5242, Université de Lyon, Lyon, France
FABIO SABINO Department of Biotechnology and Biomedicine, Technical University of
Denmark, Kongens Lyngby, Denmark
IRIT SAGI Department of Biological Regulation, Weizmann Institute of Science, Rehovot,
Israel
IGNACIO SALLENT Regenerative, Modular and Developmental Engineering Laboratory
(REMODEL), National University of Ireland Galway (NUI Galway), Galway, Ireland;
Science Foundation Ireland (SFI) Centre for Research in Medical Devices (CÚRAM),
National University of Ireland Galway (NUI Galway), Galway, Ireland
MARIE-CLAIRE SCHANNE-KLEIN Laboratoire d’Optique et Biosciences (LOB), Ecole
Polytechnique, CNRS, INSERM, Université Paris-Saclay, Palaiseau, France
RICHARD P. SCHIAVONI Koch Institute, Massachusetts Institute of Technology, Cambridge,
MA, USA
CHANGYU SHAO Department of Chemistry, Center for Biomaterials and Biopathways,
Zhejiang University, Hangzhou, Zhejiang, China
ELEE SHIMSHONI Department of Biological Regulation, Weizmann Institute of Science,
Rehovot, Israel
MATTHEW D. SHOULDERS Department of Chemistry, Massachusetts Institute of Technology,
Cambridge, MA, USA
NICO A. J. M. SOMMERDIJK Laboratory of Materials and Interface Chemistry, Department
of Chemical Engineering and Chemistry, Center for Multiscale Electron Microscopy,
Eindhoven University of Technology, Eindhoven, The Netherlands; Institute for Complex
Molecular Systems, Eindhoven University of Technology, Eindhoven, The Netherlands
CYRIL STATZER Department of Health Sciences and Technology, Institute of Translational
Medicine, Eidgenössische Technische Hochschule (ETH) Zürich, Zürich, Switzerland
RUIKANG TANG Department of Chemistry, Center for Biomaterials and Biopathways,
Zhejiang University, Hangzhou, Zhejiang, China
Contributors xi
ALINA C. TEUSCHER Department of Health Sciences and Technology, Institute of

Translational Medicine, Eidgenössische Technische Hochschule (ETH) Zürich, Zürich,
Switzerland
CHEN VAROL The Research Center for Digestive Tract and Liver Diseases, Sourasky Medical
Center, Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel; Department of
Clinical Microbiology and Immunology, Sackler School of Medicine, Tel Aviv University,
Tel Aviv, Israel
YIFEI XU Laboratory of Materials and Interface Chemistry, Department of Chemical
Engineering and Chemistry, Center for Multiscale Electron Microscopy, Eindhoven
University of Technology, Eindhoven, The Netherlands; Institute for Complex Molecular
Systems, Eindhoven University of Technology, Eindhoven, The Netherlands
SHASHA YAO Department of Chemistry, Center for Biomaterials and Biopathways, Zhejiang
University, Hangzhou, Zhejiang, China
S. MICHAEL YU Department of Bioengineering, University of Utah, Salt Lake City, UT,
USA; Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah,
Salt Lake City, UT, USA
AMNA ZAHOOR Pharmaxis Drug Discovery, Frenchs Forest, NSW, Australia
DIMITRIOS I. ZEUGOLIS Regenerative, Modular and Developmental Engineering
Laboratory (REMODEL), National University of Ireland Galway (NUI Galway),
Galway, Ireland; Science Foundation Ireland (SFI) Centre for Research in Medical Devices
(CÚRAM), National University of Ireland Galway (NUI Galway), Galway, Ireland
Part I
In Vitro Models for Collagen Characterization

Chapter 1
Cloning and Mutagenesis Strategies for Large Collagens

Olivier Bornert and Alexander Nyström
Abstract
The size and relatively high GC content of cDNAs are challenges for efficient targeted engineering of large
collagens. There are both basic biological and therapeutic interests in the ability to modify collagens, as this
would allow for studies precisely describing interactions of collagens with specific interaction partners,
addressing consequences of individual disease-causing mutations, and assessing therapeutic applicability of
precision medicine approaches. Using collagen VII as an example, we will here describe a strategy for rapid
and simple modification of cDNAs encoding large collagens. The method is flexible and can be used for the
creation of point mutations, small or large deletions, and insertion of DNA.
Key words Collagen VII, Gibson assembly, Dystrophic epidermolysis bullosa, Collagenopathy, Exon
skipping
1 Introduction
Size, repetitive sequences, and GC-rich content are factors that

pose challenges for cloning and targeted modification of large
collagens [1]. The ability to recombinantly express collagens with
specific point mutations and small or large deletions is an important
tool that can be used for addressing a multitude of biological and
clinical questions surrounding individual collagens or a group of
them. These questions encompass understanding of the specific
interactome of a collagen, understanding of disease mechanisms
underlying specific mutations causing a collagenopathy, and the
potential therapeutic applicability of collagens shortened by tar-
geted deletion [2–5]. Multiple strategies are available that can be
applied to successfully mutate or delete parts in long, complex
cDNAs. However, a majority of these strategies involve multiple
cloning steps with cutting and pasting out and in of different
vectors. We and others have previously used such approaches to
successfully clone and express recombinant wild-type collagen VII
or collagen VII variants with targeted deletions [2, 3,
Irit Sagi and Nikolaos A. Afratis (eds.), Collagen: Methods and Protocols, Methods in Molecular Biology, vol. 1944,
https://doi.org/10.1007/978-1-4939-9095-5_1, © Springer Science+Business Media, LLC, part of Springer Nature 2019
3
4 Olivier Bornert and Alexander Nyström
5–7]. By incorporating enzymatic assembly of overlapping DNA

fragments developed by Gibson and co-workers [8, 9] in our work-
flow of mutating, expressing, and purifying modified recombinant
collagen VII, we have been able to minimize the steps needed for
this, leading to a significantly streamlined process with higher
fidelity.
We focus our work on collagen VII. Collagen VII is a large
collagen with a length and molecular weight close to 400 nm and
1 MDa, respectively [10]. It has a big N-terminal non-collagenous
(NC1) domain, followed by the largest uninterrupted collagenous
domain in mammalian, and a short C-terminal NC2 domain, which
is essential for antiparallel dimerization [10, 11]. Outside the cell,
two collagen VII molecules align in a tail-to-tail fashion; processing
of the NC2 domain by astacin-like family proteases allows for the
formation of disulfide bonds between two molecules leading to the
creation of stable antiparallel dimers [12–14]. Lateral aggregation
of the dimers produces electron-dense structures that loop through
the lamina densa down to the superficial papillary dermal matrix—
these structures are known as anchoring fibrils. The anchoring
fibrils confer dermal stability through high affinity to laminin-332
and collagen IV in the epidermal basement membrane and physical
entrapment of collagen fibrils in the dermal loops [11]. In addition
to skin stability, collagen VII is essential for correct epidermal and
dermal healing after wounding by aiding organization of the epi-
dermal basement membrane and regulating fibroblast activation
[15]. Genetically evoked loss of abundance or function of collagen
VII causes the skin blistering disease dystrophic epidermolysis bul-
losa (DEB), which in its most severe forms is a severely debilitating
disease [16]. There are active research efforts toward improving the
treatment of DEB. One feature that can be exploited for therapy
which collagen VII also shares with most fibrillar collagens is that
the majority of its exons is in frame and ends with a full codon
[2, 17]. Skipping or deletion of exons carrying mutations could be
used as one means to correct the mRNA, resulting in a shorter but
still functional protein [17]. Indeed, multiple reports of naturally
occurring skipping of mutated exons have shown that this can lead
to a milder DEB phenotype [18–21]. Preclinical studies have also
shown that targeted permanent deletion or antisense
oligonucleotide-mediated exon skipping on the pre-mRNA level
can increase the abundance of collagen VII molecules, which are
slightly shortened but retain sufficient functionality to restore
dermal-epidermal cohesion [2, 22–25]. However, studies on
other diseases suggest that in some cases exon skipping worsen
the phenotype [26] and thus the functionality of proteins resulting
from exon skipping cannot be generalized. Accordingly, clinical
applicability of these approaches has to be assessed on an exon-to-
exon-based basis, to determine the functionality of protein
variants lacking the amino acids encoded by specific exons.
Cloning and Mutagenesis Strategies for Large Collagens 5
We will here present our new optimized workflow for precision

engineering of large collagens. The approach is rapid and flexible,
allowing for direct cDNA creation of point mutations, insertion of
DNA, and deletion of small to large regions. It can also be applied
for other large proteins or proteins encoded by GC-rich genes.
2 Materials
2.1 Polymerase 1. Phusion® high-Fidelity PCR master mix with high-fidelity

Chain Reaction (PCR) buffer.
2. 10 mM DNA primers in ultrapure water.
3. Thermocycler.
4. 1% Agarose gel: Dissolve 1 g of agarose in 100 ml of Tris-
acetate-EDTA (TAE) buffer (40 mM Trizma® base, 40 mM
acetate, 1 mM EDTA pH 8.0) by boiling the solution. Add 2 μl
GelRed® fluorescent nucleic acid dye and cast the gel with the
appropriate gel comb.
5. 6 loading dye.
6. Amersham imager 600.
7. GeneJET Extraction Kit.
8. NanoDrop™ 1000 Spectrophotometer.
2.2 Gibson Reaction 1. NEBuilder® HiFi DNA Assembly Cloning Kit.

and Bacterial 2. Thermocycler.
Transformation
3. Luria broth (LB)-ampicillin agar plates: dissolve 3.7 g of LB
agar MILLER in 100 ml of ultrapure water. After sterilization
of the medium with an autoclave, 100 μg/ml of ampicillin is
added to the medium when the temperature of the mixture is
lower than 50 C. Pour the solution into petri dish plates until
it covers the bottom.
4. NEB® 10-alpha competent E. coli bacteria.
5. Water bath at 42 C.
6. Incubator at 37 C.
7. LB medium: Dissolve 2.5 g of LB medium in 100 ml of
ultrapure water, and sterilize the medium with an autoclave.
8. 100 mg/ml ampicillin.
9. 15 ml round-bottom tube.
10. 500 ml Erlenmeyer flask.
11. PCR products and plasmids purification kit, e.g., GeneJET
plasmid Miniprep Kit.
12. HiSpeed Plasmid Maxi Kit.
13. Incubator at 37 C with shacking plate.
2.3 Recombinant 1. Human Embryonic Kidney (HEK) 293 cells.

Expression 2. DMEM/F12 medium: Dulbecco’s Modified Eagle medium/
of Collagen VII F12 supplemented with 10% Bovine Serum Albumin, 1%
penicillin–streptomycin, 1% antibiotic–antimycotic, and 1%
L-glutamine.
3. Lipofectamine® 2000.
4. Opti-MEM Reduced Serum Medium.
5. 20 mg/ml phleomycin
6. 10 cm cell culture dish
7. Blue sample buffer (BSB) 4: 4 M Urea, 0.5 M Trizma® base,
0.05% bromophenol blue, 10% SDS, 50% glycerol, 0.25 M
DTT, pH 6.8.
8. 7.5% Mini-PROTEAN® TGX precast gels (see Note 1)
9. Mini-PROTEAN® Tetra Vertical Electrophoresis cell.
10. TGS electrophoresis buffer: 25 mM Trizma® base, 192 mM
glycine, 0.1% SDS, pH 8.3.
11. Criterion™ blotter.
12. Transfer buffer: 1 Tris-borate (TB) buffer, 50 mM Trizma®
base, and 50 mM boric acid (see Note 2).
13. Amersham Protran nitrocellulose membrane.
14. Antibodies: Rabbit polyclonal antibody directed against the
epitope of the mouse monoclonal LH7.2 antibody (home-
made and verified to be specific for collagen VII [27], the
epitope is in the NC1 domain of collagen VII) and rabbit
polyclonal ProC7 antibody (homemade, raised against the
end of the NC2 domain of collagen VII; specific for the
pro-form of collagen VII) (see Note 3).
15. Tris-buffered saline-Tween 20 (TBS-T): 50 mM Trizma®
base, 150 mM NaCl, 0.05% Tween 20, pH 7.4.
16. Blocking buffer: TBS-T + 5% nonfat dry milk.
17. Enhanced chemiluminescence detection reagent, e.g., ECL™
Prime.
18. Developing system, e.g., Amersham™ imager 600.
19. 1 Phospho-buffered saline (PBS).
20. 50 μg/ml L-Ascorbic acid.
21. 50 ml Falcon Conical Centrifuge Tube.
22. 1 mM Pefabloc.
23. 500 mM EDTA pH 8.0.
24. Collagen IV.
25. 100 ng laminin-332.
3 Methods
3.1 Strategy Design 1. First, a strategy that will be used to modify the DNA sequence
and Generation of DNA of the target cDNA has to be designed. As an example, we will
Fragments with describe a procedure to remove a specific DNA region (from a
Overlapping Ends few amino acids to deletion of large sequences) from COL7A1
cDNA cloned into the vector pcDNA3.1 (Fig. 1).
2. The strategy is based on the use of PCR dividing the target
cDNA and vector, in this case COL7A1 and pcDNA3.1, into
fragments. In order to allow efficient amplification of the frag-
ments by PCR, dividing the mother sequence into fragments of
2.5 to 3 kb maximum is recommended.
3. As shown in Fig. 1, the strategy described here can be used to
delete a specific region of DNA. Not described here, by using
Fig. 1 Mutagenesis of COL7A1 using Gibson cloning. This figure presents an overview of the strategy used to
modify the COL7A1 cDNA sequence in the pcDNA3.1 vector by Gibson assembly. In step 1, specific pairs of
primers with overlapping sequence are designed to ensure reconstitution of the mother DNA construct, as well
as excision of a specific DNA sequence in COL7A1. These primers are then used to amplify selected regions of
the mother DNA construct (step 2). Finally, through Gibson reaction (step 3), the fragments are assembled
together in order to create a new DNA construct lacking the target DNA sequence
Table 1
Primers used for amplification of COL7A1 and deletion of exon73. Primers used to delete exon73 are
shown in red. Overlap is indicated with lower case
overlapping primer containing point mutations, one or more

nucleotides in a DNA sequence can be modified. Insertion of
DNA is also possible by in vitro synthesis or PCR amplification
of a piece of DNA with ends that overlap with the region where
it should be inserted.
4. To illustrate the approach, we have chosen to present deletion
of the nucleotides corresponding to exon73 from COL7A1
cDNA. Listed in Table 1 are the primers used to achieve this;
the overlap and specific sequences are shown by lower case and
capital letters, respectively. The primers designed to delete
exon73 are shown in red (Table 1).
5. PCR primers that will be used to amplify the DNA fragments
must contain an overlap sequence of 10 nucleotides from the
adjacent fragments together with a specific sequence for the
region to be amplified (15–25 nucleotides) (Table 1). Even
though the design of the primers can be done manually [9],
we recommend to use the online tool, NEBuilder® assembly
tool, available at nebuilder.neb.com.
6. On ice, add 0.05 mM of each primer together with 1 ng of the
vector pcDNA3.1 containing COL7A1 cDNA to a 1.5 ml
Eppendorf tube. Increase volume to 10 μl with ultrapure sterile
water. Then, combine 10 μl of Phusion® High-Fidelity PCR
Master Mix with high-fidelity buffer to the mixture.
7. The following thermocycling conditions are commonly used
by us to amplify the DNA fragments:
Initial denaturation 98 C 2 min 1 cycle

Denaturation 98 C 30 s 35 cycles
Annealinga 60 C 30 s
Extensiona 72 C 1 min
Final extension 72 C 5 min 1 cycle

Hold 4 C
See Note 4
a
8. After amplification, add 4 μl of loading dye to the PCR reac-

tion. Load samples on a 1% agarose gel, and perform migration
at 120 V for approximately 1 h in TAE buffer. PCR products
can be visualized using Amersham™ imager 600 equipped
with a UV lamp. Next, bands corresponding to the desired
products are cut out from the gel using a scalpel. The DNA
fragments are then extracted from the gel and purified using
GeneJET Extraction Kit. The concentration of the extracted
DNA fragments can be determined using a NanoDrop™ 1000
Spectrophotometer (see Note 5).
9. As assembly of more than four fragments using Gibson cloning
is challenging, we recommend a minimal concentration of
20 ng/μl in a volume of at least 20 μl for the DNA fragments.
PCR can be repeated if the quantity is not sufficient.
3.2 Gibson Assembly 1. Mix 50 ng of each PCR product together in a PCR tube. For
and Bacterial optimal reaction, the total volume should not exceed 20 μl
Transformation (see Note 6). Add a 1:1 volume of NEBuilder® HiFi DNA
assembly 2 Master Mix to the mixture. A control reaction
lacking one of the fragments is highly recommended to assess
the efficacy and specificity of the reaction.
2. Incubate reactions in a thermocycler at 50 C for 60 min.
Figure 2 provides a detailed step-by-step illustration of the
Gibson assembly® strategy [9]. Following incubation, samples
can be used directly or stored at 20 C for subsequent
transformation.
3. Thaw on ice NEB 10-alpha (see Note 7) Competent E. coli
bacteria, and incubate 2 μl of the assembled product with 50 μl
the bacteria for 30 min on ice. Heat shock the samples at 42 C
during 45 s, and cool down on ice for 2 min. Resuspend the
transformed bacteria in 200 μl Super Optimal broth with
Catabolite repression (SOC) medium, and plate the mixture
on a LB-agar-ampicillin petri dish. Incubate the plate at 37 C
overnight.
4. Select ten clones for analysis, and resuspend them individually
in 3 ml of LB medium with 100 μg/ml ampicillin in a 15 ml
Fig. 2 Gibson cloning in detail. In order to assemble two pieces of DNA together using Gibson assembly, it is
necessary to design primers containing a sequence that binds to the mother sequence and an overlapping
end, which correspond to the second DNA fragment. After amplification by PCR, the 50 end will be digested
with a 50 exonuclease. Then, complementary DNA sequences will anneal together, and a DNA polymerase will
be used to extend the fragment. After sealing of the gaps with a DNA ligase, the construct can be used to
transform bacteria. Finally, the newly generated product can be analyzed by sequencing
round-bottom tube. Incubate the mixture at 37 C overnight

under vigorous shacking (230 rpm).
5. Purify plasmids from the bacteria by using GeneJET Plasmid
Miniprep Kit, and perform analysis by DNA sequencing (see
Note 8).
6. Positive clone(s) can be resuspended in 100 ml of LB medium
with 100 μg/ml ampicillin in a 500 ml Erlenmeyer flask. Incu-
bate at 37 C overnight under vigorous shacking (230 rpm).
7. Purify DNA plasmid by using HiSpeed Plasmid Maxi Kit, and
store the construct at 20 C.
3.3 Collagen VII 1. HEK293 cells in DMEM/F12 medium are grown until a
Expression in HEK293 confluence of 70% and then transfected with Lipofectamine®
Cells 2000 with purified plasmids in a 10 cm cell culture dish. First,
mix 74 μl of Lipofectamine® 2000 with 500 μl of Opti-MEM

medium (tube A). In a second tube (tube B), add 24 μg of
plasmid DNA to 500 μl of Opti-MEM medium. Then, transfer
mixture in tube B to tube A, and incubate the reaction for
5 min at room temperature (RT) after mixing.
2. Dropwise add the transfection reaction to the cell culture dish
containing HEK293 cells, and incubate for 2 days at 37 C with
5% CO2.
3. To select cells that have been transfected with the construct,
change the medium, and add phleomycin to a final concentra-
tion of 100 μg/ml (see Note 9). Change the medium every
other day, and add phleomycin to the fresh medium at 100 μg/
ml. Keep selecting the cells for 2 weeks.
4. Synthesis of collagen VII can be detected by Western blotting
directly of the medium. Add 5 μl of BSB 4 to 15 μl of
conditioned medium, and heat the mixture for 5 min at 98 C.
5. Place 7.5% mini-PROTEAN® TGX precast gels into a mini-
PROTEAN® Tetra Vertical Electrophoresis tank previously
filled with TGS electrophoresis buffer. Load sample, and per-
form migration at 110 V until dye front exit the gel (see Note 1).
6. Transfer proteins on a nitrocellulose membrane by using a
Criterion™ blotter. Migration is performed at 4 C with
ice-cold transfer buffer (we prefer to transfer large proteins
with TB buffer) at 300 mA during at least 2 h (see Note 2).
7. Incubate membrane in blocking buffer for 1 h at 4 C. Then,
incubate the membrane at 4 C overnight under gentle agita-
tion with rabbit polyclonal LH7.2 or ProColVII antibody
diluted 1000 times in blocking buffer (see Note 10).
8. After extensive wash with TBS-T, incubate the membrane with
HRP-conjugate goat anti-rabbit for 2 h at 4 C (1:10.000 in
blocking buffer). Finally, wash the membrane four times with
50 ml of TBS-T for 5 min at RT.
9. Collagen VII single α1 chains can then be revealed with ECL
and visualized by using Amersham™ 600 imager (Fig. 3).
3.4 First Assessment 1. Seed selected HEK293 cells carrying the construct of interest
of Functionality on poly-L-lysine-coated 10 cm tissue culture plates, and grow
of Engineered Collagen the cells to confluence.
VII Variants 2. Wash the plates three times with 3 ml PBS.
3. Add 10 ml serum-free DMEM and 50 μg/ml L-ascorbic acid
(see Note 11).
4. Next day add an additional 50 μg/ml L-ascorbic acid.
5. After 2 days harvest the medium, and add it to a Falcon 50 ml
Conical Centrifuge Tube, and centrifuge at 1000 g for 5 min
Fig. 3 Analyses of mutated collagen VII. Western blot for collagen VII on
conditioned medium from HEK293 cells expressing wild-type (WT) or collagen
VII with an internal deletion (the nucleotides corresponding to exon73 have been
deleted (Δ73)). The modified collagen VII has a deletion of 201 bp in the central
region of the COL7A1 cDNA (bp 6015 to 6216). As the deletion is small in
comparison to the large size of collagen VII, no significant difference in size can
be seen by Western blot. Usage of polyclonal LH7.2 and ProColVII antibodies
allows ensuring that the collagen VII is produced in full, as these antibodies,
respectively, bind the NC1 and the NC2 domains of collagen VII
to remove cells and debris (see Note 12). Add the medium to a
fresh 50 ml Falcon tube.
6. Cool the medium to 4 C, and add ammonium sulfate to a
saturation of 20%.
7. Precipitate the medium for 1 h at 4 C by rotating the tube
head over head.
8. Centrifuge 6000 g for at 4 C for 20 min.
9. Discard the supernatant and dissolve the pellet in 200 μl TBS.
Optional, if the protein is not to be used for limited trypsin
digestion, add 1 mM Pefabloc and 5 mM EDTA.
10. Limited trypsin digestion can be performed to give quick
information of correct folding of collagen VII (for a detailed
protocol please see [6]). The functionality of the protein can
further be assessed by determining binding to 100 ng immo-
bilized laminin-332 or collagen IV, as described in detail in
ref. 2.
4 Notes
1. Mini-PROTEAN® TGX precast gels are suitable for routine

analyses. For optimized results it is recommended to cast
SDS-polyacrylamide gels with a polyacrylamide percentage
that allows the best separation while keeping the sharpness of
the bands of the protein of interest. For analysis of collagen
VII, we routinely use 4–12% polyacrylamide gradient gels.
2. To transfer large and complex proteins, we prefer to use TB

buffer. Standard transfer buffer consisting of 25 mM Trizma®
base, 192 mM glycine, and 20% methanol can also be used;
however, the transfer efficiency of collagens is increased by
supplementing this buffer with 0.1–0.25% SDS.
3. Multiple commercial antibodies exist that detects the NC1
domain in Western blots, e.g., the mouse monoclonal LH7.2
antibody and a rabbit antihuman collagen VII from Millipore
[28]. We are unfortunately not aware of any good commer-
cially available antibodies that recognize the NC2 domain.
4. As annealing temperature is dependent on each pair of primer,
optimal annealing temperature needs to be determined indi-
vidually. Nevertheless, primers designed with NEBuilder®
Assembly Tool (nebuilder.neb.com) are usually suitable to be
used with an annealing temperature of 60 C. Also, the exten-
sion time can be increased or decreased based on the length of
the fragment to amplify. An extension of 1 min is ideal for a
3 kb fragment.
5. As shown in Fig. 1, additional bands may appear after PCR
amplification (label with #). These bands are not problematic if
they are far enough from the desired product and if they are not
predominantly amplified. Optimization of the PCR is required
if this is not the case.
6. In order to ensure an optimal Gibson reaction in a volume
suitable for subsequent bacteria transformation, the total vol-
ume of the reaction needs to be lower than 40 μl. Then, a
minimal concentration of 20 ng/μl of purified PCR product is
mandatory. As an example, 50 ng of each fragment is necessary
for a five-fragment Gibson reaction. With a concentration of
20 ng/μl, 2.5 μl of each fragment will be used. The volume will
be then 12.5 μl. In this case 12.5 μl of NEBuilder HiFi DNA
Assembly 2 Master Mix will be added to the mixture.
7. We recommend to use NEB 10-alpha competent E. coli bacte-
ria as these bacteria show increased transformation ability of
large DNA construct than others in our hands.
8. It is important to take care when designing the internal primers
in the cDNA used for sequencing. In some cases the primers
used for fragment amplification can also be used for
sequencing.
9. In this protocol phleomycin is used to select transfected cells as
the pcDNA3.1 plasmid used here has a Zeocin selectable
marker.
10. It is often sufficient to incubate the primary antibody for 1 h
at RT.
11. Ascorbic acid is essential for synthesis of hydroxyproline and

hydroxylysine and thus formation of stable collagen triple
helices.
12. It is not recommended to filter the medium as collagen VII
sticks to the filter leading to significantly reduced yield of
purified collagen VII.
Acknowledgments
Part of the work was financed by a research collaboration with

ProQR, Netherlands. AN’s research was supported by grants
from the German Research Foundation, DFG (grants NY90/2-1,
NY90/3-2, SFB850-B11 to AN, BR1475/12-1 to LBT), and
from the Dystrophic Epidermolysis Bullosa Research Association
(DEBRA) (grant Nystrom Bruckner-Tuderman 1) to AN.
References
1. Phillips CL, Lever LW, Pinnell SR et al (1991) 7. Fritsch A, Spassov S, Elfert S et al (2009)
Construction of a full-length murine pro alpha Dominant-negative effects of COL7A1 muta-
2(I) collagen cDNA by the polymerase chain tions can be rescued by controlled overexpres-
reaction. J Invest Dermatol 97:980–984 sion of normal collagen VII. J Biol Chem
2. Bornert O, Kühl T, Bremer J et al (2016) 284:30248–30256. https://doi.org/10.
Analysis of the functional consequences of tar- 1074/jbc.M109.045294
geted exon deletion in COL7A1 reveals pro- 8. Gibson DG (2011) Enzymatic assembly of
spects for dystrophic epidermolysis bullosa overlapping DNA fragments. Methods Enzy-
therapy. Mol Ther J Am Soc Gene Ther mol 498:349–361. https://doi.org/10.1016/
24:1302–1311. https://doi.org/10.1038/ B978-0-12-385120-8.00015-2
mt.2016.92 9. Gibson DG, Young L, Chuang R-Y et al (2009)
3. Chmel N, Bornert O, Hausser I et al (2018) Enzymatic assembly of DNA molecules up to
Large deletions targeting the triple-helical several hundred kilobases. Nat Methods
domain of collagen VII Lead to mild Acral 6:343–345. https://doi.org/10.1038/
dominant dystrophic Epidermolysis Bullosa. J nmeth.1318
Invest Dermatol 138:987–991. https://doi. 10. Ricard-Blum S (2011) The collagen family.
org/10.1016/j.jid.2017.11.014 Cold Spring Harb Perspect Biol 3:a004978.
4. Nyström A, Bornert O, Kühl T et al (2018) https://doi.org/10.1101/cshperspect.
Impaired lymphoid extracellular matrix a004978
impedes antibacterial immunity in epidermoly- 11. Varki R, Sadowski S, Uitto J, Pfendner E
sis bullosa. Proc Natl Acad Sci U S A 115: (2007) Epidermolysis bullosa II. Type VII col-
E705–E714. https://doi.org/10.1073/pnas. lagen mutations and phenotype-genotype cor-
1709111115 relations in the dystrophic subtypes. J Med
5. Chen M, Costa FK, Lindvay CR et al (2002) Genet 44:181–192. https://doi.org/10.
The recombinant expression of full-length type 1136/jmg.2006.045302
VII collagen and characterization of molecular 12. Chen M, Keene DR, Costa FK et al (2001) The
mechanisms underlying dystrophic epidermo- carboxyl terminus of type VII collagen med-
lysis bullosa. J Biol Chem 277:2118–2124. iates antiparallel dimer formation and constitu-
https://doi.org/10.1074/jbc.M108779200 tes a new antigenic epitope for epidermolysis
6. Nyström A, Bruckner-Tuderman L, Kern JS Bullosa acquisita autoantibodies. J Biol Chem
(2013) Cell- and protein-based therapy 276:21649–21655. https://doi.org/10.
approaches for epidermolysis bullosa. Methods 1074/jbc.M100180200
Mol Biol Clifton NJ 961:425–440. https:// 13. Rattenholl A, Pappano WN, Koch M et al
doi.org/10.1007/978-1-62703-227-8_29 (2002) Proteinases of the bone morphogenetic
protein-1 family convert procollagen VII to 38:489–492. https://doi.org/10.1111/j.

mature anchoring fibril collagen. J Biol Chem 1346-8138.2010.01008.x
277:26372–26378. https://doi.org/10. 22. Goto M, Sawamura D, Nishie W et al (2006)
1074/jbc.M203247200 Targeted skipping of a single exon harboring a
14. Moali C, Font B, Ruggiero F et al (2005) premature termination codon mutation: impli-
Substrate-specific modulation of a multisub- cations and potential for gene correction ther-
strate proteinase. C-terminal processing of apy for selective dystrophic epidermolysis
fibrillar procollagens is the only BMP-1-depen- bullosa patients. J Invest Dermatol
dent activity to be enhanced by PCPE-1. J Biol 126:2614–2620. https://doi.org/10.1038/
Chem 280:24188–24194. https://doi.org/ sj.jid.5700435
10.1074/jbc.M501486200 23. Wu W, Lu Z, Li F et al (2017) Efficient in vivo
15. Nyström A, Velati D, Mittapalli VR et al (2013) gene editing using ribonucleoproteins in skin
Collagen VII plays a dual role in wound heal- stem cells of recessive dystrophic epidermolysis
ing. J Clin Invest 123:3498–3509. https://doi. bullosa mouse model. Proc Natl Acad Sci U S A
org/10.1172/JCI68127 114:1660–1665. https://doi.org/10.1073/
16. Bruckner-Tuderman L (2010) Dystrophic epi- pnas.1614775114
dermolysis bullosa: pathogenesis and clinical 24. Turczynski S, Titeux M, Tonasso L et al (2016)
features. Dermatol Clin 28:107–114. https:// Targeted exon skipping restores type VII colla-
doi.org/10.1016/j.det.2009.10.020 gen expression and anchoring fibril formation
17. Nyström A, Bernasconi R, Bornert O (2018) in an in vivo RDEB model. J Invest Dermatol
Therapies for genetic extracellular matrix dis- 136:2387–2395. https://doi.org/10.1016/j.
eases of the skin. Matrix Biol 71–72:330–347. jid.2016.07.029
https://doi.org/10.1016/j.matbio.2017.12. 25. Bremer J, Bornert O, Nyström A et al (2016)
010 Antisense oligonucleotide-mediated exon skip-
18. McGrath JA, Ashton GH, Mellerio JE et al ping as a systemic therapeutic approach for
(1999) Moderation of phenotypic severity in recessive dystrophic Epidermolysis Bullosa.
dystrophic and junctional forms of epidermo- Mol Ther Nucleic Acids 5:e379. https://doi.
lysis bullosa through in-frame skipping of org/10.1038/mtna.2016.87
exons containing non-sense or frameshift 26. Toh ZYC, Thandar Aung-Htut M, Pinniger G
mutations. J Invest Dermatol 113:314–321. et al (2016) Deletion of Dystrophin in-frame
https://doi.org/10.1046/j.1523-1747.1999. exon 5 leads to a severe phenotype: guidance
00709.x for exon skipping strategies. PLoS One 11:
19. Schwieger-Briel A, Weibel L, Chmel N et al e0145620. https://doi.org/10.1371/journal.
(2015) A COL7A1 variant leading to pone.0145620
in-frame skipping of exon 15 attenuates disease 27. Kühl T, Mezger M, Hausser I et al (2015)
severity in recessive dystrophic epidermolysis High local concentrations of intradermal
bullosa. Br J Dermatol 173:1308–1311. MSCs restore skin integrity and facilitate
https://doi.org/10.1111/bjd.13945 wound healing in dystrophic Epidermolysis
20. Cserhalmi-Friedman PB, McGrath JA, Mel- Bullosa. Mol Ther J Am Soc Gene Ther
lerio JE et al (1998) Restoration of open 23:1368–1379. https://doi.org/10.1038/
reading frame resulting from skipping of an mt.2015.58
exon with an internal deletion in the 28. Wullink B, Pas HH, Van der Worp RJ et al
COL7A1 gene. Lab Investig J Tech Methods (2018) Type VII collagen in the human accom-
Pathol 78:1483–1492 modation system: expression in Ciliary body,
21. Koga H, Hamada T, Ishii N et al (2011) Exon Zonules, and lens capsule. Invest Ophthalmol
87 skipping of the COL7A1 gene in dominant Vis Sci 59:1075–1083. https://doi.org/10.
dystrophic epidermolysis bullosa. J Dermatol 1167/iovs.17-23425
Chapter 2
Multi-well Fibrillation Assay as a Tool for Analyzing

Crosslinking and Stabilization of Collagen
Venkat Raghavan Krishnaswamy
Abstract
Collagen is the most widely used substratum in cell culture and biomaterials applications. In this chapter, we
describe a simple procedure to isolate collagen, which can be employed to a wide range of tissue sources,
and subsequently use it to study the collagen crosslinking and stabilization abilities of various compounds.
The protocol is designed for a multi-well format assay and thus can be used for simultaneous assessment of
multiple number of compounds and can be easily adapted to a high-throughput screening setup.
Key words Collagen, Crosslinking, Extracellular matrix, Fibrillation, Quarter staggered structure
1 Introduction
Collagen is the most abundant protein in the human body and plays
a vital role in several physiological and pathological conditions
[1–3]. It is used in several biomedical applications and holds a
central role in treating a myriad of abnormalities such as abnormal
wound healing, cancer, etc. [4]. Because of its wide presence and
diverse functions, isolation and characterization of collagen are
important aspects in biomaterials preparation and understanding
the molecular mechanisms of various diseases. One of the
intriguing properties of the isolated collagen fibrils, in vitro, is its
ability to arrange into an quarter-staggered structure in physiologi-
cal conditions resembling the in vivo fibril formation [5]. This
property can be utilized to comprehend the influence of molecules
on the three-dimensional spatial organization of collagens, which
are crucial in dictating the cellular behavior.
17
18 Venkat Raghavan Krishnaswamy
2 Materials
All solutions should be prepared using analytical grade reagents.

Follow all safety precautions and waste disposal regulations
diligently.
2.1 Collagen This protocol can be employed to isolate collagen from a variety of
Isolation sources including skin and tendon. The most commonly used
source is rat tail tendon which is almost entirely made up of colla-
gen type I. The isolation of tendon from rat tail is explained by
Rajan et al. [6].
1. Wash buffer 1: Mix three volumes of methanol with one vol-
ume of chloroform. Store the solution in an airtight bottle at
4 C.
2. Wash buffer 2: Prepare 0.5 M sodium acetate by dissolving
41.02 g of sodium acetate in 1 L of deionized water.
3. Extraction reagent: Make up 28.73 mL of pure glacial acetic
acid to 1 L with deionized water to prepare 0.5 M acetic acid.
Store at 4 C.
4. Precipitating solution: To prepare 2.8 M sodium chloride,
dissolve 163.6 g of sodium chloride in 800 mL, and make up
the volume to 1 L using extraction reagent, store at 4 C.
5. Dialysate: Make up 1.14 mL of pure glacial acetic acid to 1 L
with deionized water to prepare 0.02 M acetic acid. Store
at 4 C.
2.2 Collagen 1. Sodium phosphate buffer (component a): To prepare 0.2 M

Fibrillation sodium phosphate buffer, dissolve 35.61 g of disodium hydro-
gen phosphate dihydrate (Na2HPO4·2H2O) and 27.6 g of
2.2.1 Fibrillation Buffer
sodium dihydrogen phosphate monohydrate (NaH2-
PO4·H2O) separately in 800 mL of deionized water. Make up
the volumes to 1 L. Store these stock solutions at 4 C. Before
the assay, mix 770 mL of Na2HPO4·2H2O and 230 mL of
NaH2PO4·H2O; adjust the pH to 7.2 if necessary.
2. Sodium chloride (component b): To prepare 2 M sodium
chloride, dissolve 116.88 g of sodium chloride in 800 mL,
and the make up the volume to 1 L using deionized water,
store at 4 C.
3. Sodium hydroxide (component c): To prepare 1.25 N sodium
hydroxide, dissolve 50 g of sodium hydroxide in 800 mL, and
the make up the volume to 1 L using deionized water.
For 1 mL collagen, mix 100 and 75 μL of components a and b
respectively. The volume of component c may vary for each batch of
collagen (see Note 6). The mixing of the fibrillation buffer compo-
nents is always done on the day of the assay.
Collagen Isolation and Fibrillation Assay 19
3 Methods
3.1 Collagen 1. After harvesting the tissue, remove the unwanted tissues like
Isolation soft tissue, clotted blood, etc. that are present. Wash the sam-
ples twice thoroughly with cold deionized water.
2. Wash the tissue samples with wash buffer 1. Repeat two more
times, and ensure all the fat particles are removed.
3. Wash the samples once with wash buffer 2.
4. Wash the samples extensively with extraction reagent, and
incubate the tissue in at least 10 volumes of extraction solution
for 48 to 72 h (see Note 1).
5. Homogenize the samples in a blender with extraction reagent.
Centrifuge the samples at 14,000 g for 30 min at 4 C. Save
the supernatant, and re-extract the collagen once from the
pellet by repeating the homogenization. Pool both the super-
natants (see Note 2).
6. For 750 mL of collagen supernatant, gradually add 250 mL of
precipitating solution with mild stirring overnight at 4 C.
7. Centrifuge the solution at 6000 g for 10 min at 4 C. Dis-
card the supernatant, and collect the pellet.
8. Dissolve the pellet in a suitable volume of cold extraction
reagent.
9. Dialyze the collagen solution extensively against dialysate at
least for 24–48 h (see Note 3).
10. Freeze the samples as a thin layer on a lyophilized flask by
rotating the flask with the sample on an ice-salt slurry. Freeze
the collagen coated flasks at 80 C for a minimum of
12–16 h.
11. Freeze dry the samples thoroughly. Store the samples at
80 C in an airtight container.
12. Dissolve a portion of the lyophilized sample in the dialysate,
and estimate the collagen content using standard procedures.
The purity of the collagen can be assessed by a standard sodium
dodecyl sulfate-polyacrylamide gel electrophoresis (see Note 4)
(Fig. 1).
3.2 Collagen 1. Prepare 0.5 mg mL 1 collagen solution by dissolving the

Fibrillation required amount of lyophilized collagen in the dialysate with
mild stirring at 4 C (see Note 5).
2. Add required volume of fibrillation buffer in different wells of a
multi-well plate (see Note 6).
3. Mix the test compounds to the fibrillation buffer at this stage
by replacing with the same volume of sodium phosphate buffer.
Fig. 1 The electrophoretic mobility of collagen isolated from normal (N),

hypertrophic scar (H), and keloid (K) skin tissue using the present method [7]
Fig. 2 Representative graph showing typical pattern of the collagen fibrillation
4. Switch on the plate reader, and pre-warm the plate chamber to

37 C.
5. Add 200 μL of collagen to all the tubes mix gently with the
pipette to make a uniform solution. Set up controls in parallel
(see Note 7).
6. Quickly spin the contents of the tube to collect any solution on
the walls and to remove air bubbles, if any.
7. Measure the absorbance kinetics at 313 nm for at least
120 min.
8. Calculate the T1/2, which is the time taken to reach half the
value of the plateau. Compare the T1/2 of treated and
untreated samples to understand the effect of the treated com-
pounds (Fig. 2).
Collagen Isolation and Fibrillation Assay 21
4 Notes
1. If isolating from skin, remove the epidermis using a scalpel

blade. Care must be taken not to expose the tissue for a long
period at room temperature. One easy way is to perform this
step after soaking the tissue in extraction reagent which would
loosen up the epidermis. After washing, record the wet weight
of the samples.
2. Place the blending jar at 4 C for at least 6 h before using
it. The pulse blending option should be used rather than
continuous mode to prevent heating of the samples. The acid
soluble collagen in the supernatant and the insoluble portion in
the pellet can be separated at this point. The insoluble pellet
can be further digested using pepsin (1 mg per 100 mg of wet
weight). The fibrillation assay can be performed with both the
collagens; typically pepsin-solubilized atelocollagen has a
slower rate of fibril formation as they lack the terminal
telopeptides.
3. Proper removal of salts is important for downstream applica-
tions of collagen. The volume of the dialysate should be
250 times the sample volume. In cases of large sample volumes,
the dialysate can be changed every 24 h, and the dialysis timing
can be increased. The dialysate pH can be measured to assess
the efficiency of dialysis.
4. Hydroxyproline assay is a reliable method for quantification of
the collagen content. However, dye-based assay kits with sim-
ple colorimetric estimation are now commercially available.
Weighing the lyophilized collagen before dissolving helps in
assessing the concentration of the entire batch.
5. Obtaining a homogeneous solution is important. The lyophi-
lized collagen should be weighed and transferred as quickly as
possible. Using pre-chilled acetic acid is essential.
6. The pH of the final solution is very critical for proper fibril
formation. Therefore, for each batch of collagen, run a trial to
determine the volume of fibrillation buffer component “c”
required to adjust the pH to 7.2. As temperature greatly influ-
ences pH, after mixing all the components, warm the solution
to 37 C and check pH quickly. Once pH is adjusted, leave the
tube at 37 C in a water bath, and note the time taken for the
fibrillation. Alternatively, an aliquot of it can be transferred to a
microplate and read at 313 nm till the optical density reaches
plateau. This will also indicate whether the collagen concentra-
tion is optimum.
7. Use multichannel pipette to dispense collagen simultaneously

to multiple wells. Do not introduce air bubble as it interferes
with the reading. Set controls (without test compounds), posi-
tive controls (50 μM glutaraldehyde or other known crosslink-
ing agents) and negative controls (replace collagen with
dialysate) in parallel for comparison.
References
1. Prockop DJ, Kivirikko KI (1995) Collagens: new target to limit fibrosis. J Biol Chem 283
molecular biology, diseases, and potentials for (38):25879–25886. https://doi.org/10.1074/
therapy. Annu Rev Biochem 64:403–434. jbc.M804272200
https://doi.org/10.1146/annurev.bi.64. 6. Rajan N, Habermehl J, Cote MF, Doillon CJ,
070195.002155 Mantovani D (2006) Preparation of ready-to-
2. Mienaltowski MJ, Birk DE (2014) Structure, use, storable and reconstituted type I collagen
physiology, and biochemistry of collagens. Adv from rat tail tendon for tissue engineering appli-
Exp Med Biol 802:5–29. https://doi.org/10. cations. Nat Protoc 1(6):2753–2758. https://
1007/978-94-007-7893-1_2 doi.org/10.1038/nprot.2006.430
3. Myllyharju J, Kivirikko KI (2001) Collagens and 7. Krishnaswamy VR, Lakra R, Korrapati PS
collagen-related diseases. Ann Med 33(1):7–21 (2014) Keloid collagen–cell interactions: struc-
4. Lee CH, Singla A, Lee Y (2001) Biomedical tural and functional perspective. RSC Adv 4
applications of collagen. Int J Pharm 221 (45):23642–23648. https://doi.org/10.1039/
(1–2):1–22 C4RA01995D
5. Chung HJ, Steplewski A, Chung KY, Uitto J,
Fertala A (2008) Collagen fibril formation. A
Chapter 3
Production and Characterization of Chemically

Cross-Linked Collagen Scaffolds
Ignacio Sallent, Héctor Capella-Monsonı́s, and Dimitrios I. Zeugolis
Abstract
Chemical cross-linking of collagen-based devices is used as a means of increasing the mechanical stability
and control the degradation rate upon implantation. Herein, we describe techniques to produce cross-
linked with glutaraldehyde (GTA; amine terminal cross-linker), 4-arm polyethylene glycol succinimidyl
glutarate (4SP; amine terminal cross-linker), diphenyl phosphoryl azide (DPPA; carboxyl terminal cross-
linker), and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC; carboxyl terminal cross-linker) colla-
gen films. In addition, we provide protocols to characterize the biophysical (swelling), biomechanical
(tensile), and biological (metabolic activity, proliferation and viability using human dermal fibroblasts and
THP-1 macrophages) properties of the cross-linked collagen scaffolds.
Key words Collagen cross-linking, Biophysical characterization, Biomechanical properties, Cytotox-

icity, Macrophage response
1 Introduction
Collagen is the most abundant extracellular matrix (ECM) protein

in mammals. Fibril-forming collagens (type I, II, III, V, XI) repre-
sent 90% of the total collagen content in the body, while their
fibrous structure empower containing tissues with their specific
viscoelastic properties [1]. Collagen type I comprises of two α1
(I) and one α2(I) polypeptide chains that tangle up in a triple-helix
conformation [1, 2]. It is stabilized via two cross-links in the helical
region and one more in each telopeptide region (N- and
C-terminal), through the action of lysyl oxidase [3, 4], sugar-
mediated [5, 6], and transglutaminase [7] cross-linking. These
cross-links are responsible for the quarter-staggered arrangement
of collagen molecules in lateral and head-to-tail fashion, resulting in
the formation of cross-striated fibrils with exceptional mechanical
properties [8, 9].
23
24 Ignacio Sallent et al.
Considering its abundance in native ECM supramolecular

assemblies and its cell recognition signals, collagen type I is favored
for tissue engineering and drug delivery applications [10]. How-
ever, collagen type I extraction from animal tissues involves break-
age of the native cross-links and consequent loss of its mechanical
integrity/stability [11, 12]. Thus, reconstituted forms of collagen
need to be exogenously cross-linked, through chemical, physical, or
biological approaches [12–14] to tune the mechanical properties
and the degradation rate of the produced scaffolds upon implanta-
tion [15–17]. Although physical (e.g., dehydrothermal [18, 19],
ultraviolet irradiation [20, 21]) and biological (e.g., microbial
[22, 23] or tissue type II [24, 25] transglutaminase) cross-linking
methods have been assessed, the resultant cross-linking is weak. To
this end, amine or carboxyl terminal chemical cross-linking meth-
ods are preferred. However, customarily employed chemical cross-
linkers, such glutaraldehyde (GTA, amine terminal) and 1-ethyl-3-
(3-dimethylaminopropyl)carbodiimide (EDC, carboxyl terminal),
are often associated with cytotoxicity [26–29], foreign body
response [30, 31], and macrophage pro-inflammatory response
[32–34]. Although alternative cross-linkers, such as 4-arm polyeth-
ylene glycol succinimidyl glutarate (4SP; amine terminal) [35–38]
and diphenyl phosphoryl azide (DPPA; carboxyl terminal)
[39, 40], have been proposed and have shown promise, they have
yet to be directly compared.
We have previously described assays to characterize collagen
and to quantify thermal and biochemical properties of collagen-
based preparations, biomaterials, and tissue specimens [41–45].
Herein, we provide protocols to fabricate cross-linked with GTA,
4SP, EDC, and DPPA collagen type I films and to characterize the
effect of the cross-linking treatment on the biophysical (swelling),
biomechanical (tensile), and biological (metabolic activity, prolifer-
ation, and viability using human dermal fibroblasts and THP-1
macrophages) properties of the resultant scaffolds.
2 Materials
2.1 Collagen Cross- 1. 1 M sodium hydroxide (NaOH) in water.

Linking and Film 2. 1 M hydrochloric acid (HCl) in water.
Preparation
3. 1 and 10 phosphate-buffered saline (PBS).
4. 6 mg/ml collagen type I stock solution: Dissolve 180 mg of
freeze-dried collagen type I in 30 ml of 0.05 M acetic acid.
5. 0.625% GTA cross-linker solution: Dilute 6.25 μl of 25% GTA
in 244 μl of PBS 1 (see Note 1).
6. 20 mM 4SP cross-linker solution: Dissolve 50 mg of 4SP-SG in
0.25 ml PBS 1. Vortex thoroughly until the powder dissolves.
Chemically Cross-Linked Collagen Scaffolds 25
7. 1% DPPA cross-linker solution: Dilute 2.5 μl of DPPA in

0.25 ml of 2% dimethylformamide (DMF) (see Note 1).
8. 1000 mM EDC cross-linker solution: Dissolve 98 mg of 2-(N-
morpholino)ethanesulfonic acid (MES) in 10 ml of distilled water,
and adjust the pH to 5.5. Dissolve 115 mg of N-hydroxysuccini-
mide (NHS) in 5 ml of MES buffer. Dissolve 48 mg of N
(3-dimethylaminopropyl)-N0 -ethylcarbodiimide (EDC) hydro-
chloride in 0.25 ml of the MES + NHS solution (see Note 1).
9. Silicon or PDMS nonadherent mold (3 4 cm).
10. 15 ml screw-cap tubes.
11. Benchtop vortex.
12. Laminar flow hood.
2.2 Swelling Assay 1. 1 phosphate-buffered saline (PBS).

2. Tweezers.
3. Laboratory digital milligram scale.
2.3 Tensile Test 1. Scalpel and blades.

2. Ruler.
3. Cutting board.
4. 1 phosphate-buffered saline (PBS).
5. Tweezers.
6. Digital caliper.
7. Tensile test machine with 10 N load cell (Zwick Roell).
2.4 Cell Culture 1. Primary human dermal fibroblasts (DF; ATCC).

2. Human-derived leukemic monocyte cells (THP-1; ATCC).
3. Culture media for DF: DMEM—High glucose supplemented
with 10% fetal bovine serum (FBS) and 1% penicillin and
streptomycin.
4. Culture media for THP-1: RPMI-1640 supplemented with 10%
fetal bovine serum (FBS) and 1% penicillin and streptomycin.
5. Phorbol 12-myristate 13-acetate (PMA).
6. Lipopolysaccharides from Escherichia coli O55:B5 (LPS).
7. Cell culture silicone rings to fix the films on the plates:
18.6 mm of diameter ace O-rings.
8. Sterile 24-well plates.
2.5 Cell Proliferation 1. 40 ,6-diamidino-2-phenylindole (DAPI): 1:3000 dilution in

and Morphology Assay PBS 1 from 1 mg/ml stock.
2. Optional: Rhodamine phalloidin, 1:150 dilution in PBS 1
from 0.2 mg/ml stock.
3. 4% paraformaldehyde (PFA).
4. 1 Phosphate-buffered saline (PBS).
5. Triton-X 100 solution in 0.2% PBS.
2.6 Metabolic 1. 1 sterile phosphate-buffered saline (PBS).

Activity Assay 2. alamarBlue® solution in 10% PBS.
(alamarBlue®)
2.7 Cell Viability 1. Dimethyl sulfoxide (DMSO).

Assay 2. Phosphate-buffered saline 1 (PBS).
3. 4 mM calcein AM in DMSO.
4. 2 mM ethidium homodimer-1 in DMSO.
3 Methods
3.1 Collagen Cross- All steps detailed in this paragraph need to be performed on ice to
Linking and Film avoid denaturation of collagen and prevent its premature gelation.
Preparation The temperature of the centrifuge, if possible, should be set to 4 C
before starting the procedure. The following instructions describe
the production of a film of 3 4 cm dimensions, for other mold
dimensions the volume of the reagents need to be scaled up.
1. Mix 0.42 ml of PBS 10 with 0.2 ml of 1 M NaOH in a 15 ml
screw-cap tube using a benchtop vortex.
2. Add 4.2 ml of the collagen stock solution to the mixture, and
mix thoroughly with the vortex.
3. Use the pH meter and 1 M NaOH to neutralize the solution
(pH 7.2–7.4).
4. Add the 0.25 ml of the chosen cross-linker, and mix with the
benchtop vortex (see Note 2).
5. Centrifuge the solution at 2000 rpm for 10 s to remove the
bubbles.
6. Pour 4 ml of the solution on the nonadherent mold (3 4 cm).
7. Incubate the sample (mold and collagen) at 37 C for 1 h.
8. Dry the sample in the flow hood at room temperature (RT) for
at least 24 h.
9. Use tweezers to peel off the collagen film from the nonadher-
ent surface.
10. Store the films in a dark and dry environment until use.
3.2 Swelling Assay 1. Use a blade and a scalpel to cut five pieces (n ¼ 5) from the
collagen film, and record their dry weight using a laboratory
scale (see Note 3).
2. Incubate the samples in PBS 1 at RT overnight.

3. Blot the samples with filter paper to remove the excess surface
water.
4. Weigh the samples in the laboratory scale, and record their wet
weight.
5. Calculate the swelling ratio (%) as follows: [(wet weight dry
weight)/dry weight] 100%.
3.3 Tensile Test 1. Use a scalpel with a sharp blade and a ruler to cut the collagen
film in three stripes of 1 4 cm (see Note 4).
2. Hydrate the samples overnight in PBS 1 at 37 C (see
Note 5).
3. Blot the samples with filter paper to remove the excess surface
water.
4. Using the calipers, take five measurements of the thickness of
the sample in the middle region (see Note 6).
5. Insert the sample into the upper grip of the tensile tester, and
tighten (see Note 7).
6. Bring the crosshead into an appropriate range for the sample
and affix the sample to the lower grip (see Note 8).
7. Set up the software from the mechanical tester to the following
tensile test parameters: Preload, 0.01 N; preload speed, 5 mm/
min; test speed, 10 mm/min.
8. Initiate the test, and export the results to excel.
9. Calculate the following parameters: stress at break (MPa),
strain at break (%) and Young’s modulus (MPa), using the
width and the thickness of the samples. Stress at break ¼ force
at break/original cross-sectional area of the sample, Young’s
modulus ¼ slope of the linear region of the stress (MPa)/strain
(no units) curve (see Note 8) (Table 1).
Table 1
Swelling ratio, thickness, and mechanical data of the produced collagen films
Cross-linking Swelling ratio Thickness Stress at break Strain at break Young’s modulus
agent (%) (μm) (MPa) (%) (MPa)
NC 100 16 178 4 0.03 0.01 8.33 3.00 0.20 0.04
GTA 21 9 # 67 6 # 0.19 0.04 # 4.19 2.82 1.95 0.33 #
4SP 80 9 232 33 0.07 0.03 # 13.28 3.13 0.38 0.04
EDC 125 7 185 39 0.03 0.01 15.78 3.08 0.10 0.03
DPPA 91 23 83 12 # 0.05 0.01 18.22 1.89 # 0.22 0.01
GTA significantly decreased water absorption. GTA and DPPA significantly reduced thickness. GTA induced the highest
stress at break. DPPA induced the highest strain at break. GTA induced the highest modulus. # indicates statistically
significant difference ( p < 0.05)
Fig. 1 Metabolic activity (a), proliferation (b), viability (c), and rhodamine phalloidin (red) and DAPI (blue)
images (d) of DF seeded onto the collagen scaffolds. DF on GTA cross-linked scaffolds presented reduced
3.4 Cell Culture 1. Using a scalpel with a blade cut the collagen films into 1 1 cm
squares.
3.4.1 Scaffold
Preparation 2. In the laminar flow hood, place the films in 24-well plates (see
Note 9).
3. Use the UV from the laminar flow hood for 1 h to partially
sterilize the collagen scaffolds in the plates. Keep the lid of the
plates open during the process.
4. Fix the collagen films to the bottom of the plates with the help
of autoclaved tweezers and silicone rings.
5. Remove the excess of cross-linker by washing the films three
times with sterile PBS 1 for 20 min.
6. Sterilize the films with 0.5 ml of 70% ethanol for 30 min.
7. Wash the films three times in sterile PBS 1.
3.4.2 DF Seeding 1. Incubate the sterilized films at 37 C and 5% CO2 overnight

with specific media for the DF cells.
2. Discard the media, and seed the DF onto the various samples at
20 103 cells/cm2 with 0.5 ml of fresh DMEM supplemented
media. Incubate the fibroblast for 3, 5, and 7 days at 37 C and
5% CO2. Change media every 3–4 days (Fig. 1).
3.4.3 THP-1 Seeding 1. Incubate the sterilized films at 37 C and 5% CO2 with 1 ml
cell-specific media for THP-1 cells for 3 days.
2. Collect the RPMI supplemented media incubated with the
scaffolds (preconditioned media), and filter it through 0.2 μm
sterile filters. Keep it at 37 C in the water bath until
further use.
3. Seed THP-1 into new 24-well plates and onto the scaffolds at
26 103 cells/cm2 with THP-1 complete media supplemen-
ted with 100 ng/ml of PMA to induce adherent mature
macrophage-like state for 6 h.
4. Wash off nonattached cells with sterile PBS 1.
5. Add 0.5 ml of the filtered preconditioned media to the macro-
phages attached to the 24-well plates TCP and 0.5 ml of fresh
RPMI supplemented media to the macrophages attached to
the scaffolds. Create a positive control for macrophage
ä
Fig. 1 (continued) metabolic activity at day 3 and 7 and reduced proliferation at all three time points in
comparison to the NC samples. None of the cross-linking treatments significantly affected DF viability (c). DF
adopted a spindle-shaped morphology in all the groups except for GTA, on which cells spread in all directions
(d). Treatments: non-cross-linked collagen films (NC), collagen films cross-linked with glutaraldehyde (GTA),
4-arm PEG succinimidyl glutarate (4SP), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), and diphe-
nylphosphoryl azide (DPPA). Tissue culture plastic (TCP) was used as control. Significance compared to *NC
group ( p < 0.05)
Fig. 2 Metabolic activity (a), proliferation (b), viability (c), elongated cells proportion (d), and rhodamine
phalloidin (red) and DAPI (blue) images (e) of THP-1 seeded onto the collagen scaffolds. By day 1, LPS
significantly increased THP-1 metabolic activity (a); and 4SP significantly reduced THP-1 proliferation (b) in
inflammatory response by culturing TCP-seeded macrophages

with 100 ng/ml of LPS in RPMI complete media and a nega-
tive control with only complete media. Incubate macrophages
for 24 h and 48 h at 37 C and 5% CO2 (Figs. 2 and 3).
3.5 Cell Proliferation 1. At the appropriate time point remove media from cells.
and Morphology Assay 2. Gently wash scaffolds with PBS 1.
3. Add 0.5 ml of 4% PFA, and allow the sample to fix for 30 min at
RT (see Note 10).
4. Wash the sample with 1 PBS.
5. Incubate the samples with 0.2 ml of 0.2% Triton-X solution for
5 min.
6. Wash the sample with PBS 1.
7. Add rhodamine-phalloidin solution enough to cover the bot-
tom of the well and/or scaffolds (around 150 μl).
8. Incubate 1 h at RT. Protect the samples from the light during
this procedure and in subsequent manipulations.
9. Wash the sample with PBS 1.
10. Add 150 μl of DAPI solution, and incubate for 5 min at RT.
11. Wash the sample three times with PBS 1 (see Note 11).
12. Image the samples in a fluorescence or confocal microscope
using a 10 objective and an excitation/emission wavelength
of 358/461 nm for DAPI and 540/565 nm for rhodamine
phalloidin.
13. Take 4–5 representative images of every sample.
14. Use ImageJ or other software for an automated counting of the
nuclei per image (see Note 12) and for morphology assessment
(see Note 13).
15. Graph the number of cells per area to evaluate cell
proliferation.
Fig. 2 (continued) comparison to the NC samples. By day 2, GTA, EDC, and 4SP significantly reduced THP-1
metabolic activity (a); and all the cross-linking treatments significantly reduced THP-1 proliferation (b) in
comparison to the NC. None of the cross-linking treatments significantly affected macrophage viability (c). All
treatments elicited a significant lower proportion of elongated cells than LPS, whereas THP-1 in GTA, EDC and
DPPA presented a significant higher number of elongated cells than TCP at both time points, likewise NC at
day 2 (d). THP-1 adopted a rounded morphology, independently of the treatment and time in culture. Some
elongated macrophages (white arrows) were observed predominantly in LPS- and GTA-treated samples,
whereas cell clusters (black arrows) were mainly observed in LPS condition (e). Treatments: non-cross-linked
collagen films (NC), collagen films cross-linked with glutaraldehyde (GTA), 4-arm PEG succinimidyl glutarate
(4SP), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), and diphenylphosphoryl azide (DPPA). Tissue
culture plastic (TCP) and lipopolysaccharides in TCP (LPS) were used as controls. Significance compared to
*NC group, compared to #TCP and compared to +LPS group ( p < 0.05)
Fig. 3 Metabolic activity (a), proliferation (b), viability (c), elongated cells proportion (d), and rhodamine
phalloidin (red) and DAPI (blue) images (e) of THP-1 cultured with preconditioned media from the cross-linked
3.6 Cell Metabolic 1. Cell metabolic activity can be assessed by the incubation of the
Activity Assay cells with 10% alamarBlue® for 2 h on the day of the time
(alamarBlue®) point, as per manufacturer’s protocol (see Note 14).
2. Cell metabolic activity is expressed as % of reduced alamar-
Blue® and normalized to the activity of cells in TCP.
3.7 Cell Viability 1. Prepare staining solution by diluting 1:1000 the calcein AM
Assay and the ethidium homodimer-1 stock solutions with PBS 1.
2. Add 150 μl of DMSO to a well with cells alone (no scaffold) to
create a negative control with dead cells. Place the plate in the
incubator for 20 min.
3. Remove culture medium from the wells, and wash cells
with PBS.
4. Add 150 μl of staining solution to each scaffold.
5. Incubate at 37 C, 5% CO2 for 30 min.
6. Wash the sample with PBS 1 (see Note 11).
7. Image the samples in a fluorescence or confocal microscope
using a 10 objective and an excitation/emission wavelength
of 495/515 nm and 528/617 nm for Calcein AM and ethi-
dium homodimer-1, respectively.
8. Take 3–5 images of every sample.
9. Count viable (green) and dead (red) cells with ImageJ or
another software.
10. Calculate cell viability (%) by dividing the number of alive cells
by the total number of cells.
Fig. 3 (continued) collagen films. None of the cross-linking treatments significantly affected THP-1 metabolic
activity (a). Only cells treated with LPS showed a decreased proliferation in comparison with the NC samples
(b). None of the cross-linking treatments significantly affected macrophage viability (c). LPS showed a higher
number of elongated cells than TCP at both times points. All treatments showed values higher than TCP and
lower than LPS. Significant differences were found between LPS and TCP at both time points, whereas only
4SP showed a significant difference with LPS at day 1 (d). THP-1 adopted a rounded morphology, indepen-
dently of the treatment and time in culture. Some elongated macrophages were observed (white arrows)
predominantly in LPS-, NC-, and GTA-treated samples. Some cell clusters were observed (black arrows) in all
the conditions except for TCP (e). Treatments: non-cross-linked collagen films (NC), collagen films cross-
linked with glutaraldehyde (GTA), 4-arm PEG succinimidyl glutarate (4SP), 1-ethyl-3-(3-dimethylaminopropyl)
carbodiimide (EDC), and diphenylphosphoryl azide (DPPA). Tissue culture plastic (TCP) and lipopolysacchar-
ides (LPS) in TCP were used as controls. Significance compared to *NC group, compared to #TCP and
compared to +LPS group ( p < 0.05)
4 Notes
1. For safety reasons, prepare these cross-linking solutions in the

chemical hood. GTA, DPPA, and EDC are toxic if swallowed,
inhaled, or in contact with eyes or skin. Use personal protective
equipment (PPE) when manipulating these samples outside
the fume hood.
2. After the addition of the cross-linker, collagen gelation will take
place within 1–5 min (depending on the cross-linker used). To
avoid collagen gelling in the tube, the next two steps (centrifu-
gation and pouring of the solution on the mold) need to be
performed rapidly. Alternatively, devices like dual syringes or
3-way stopcocks might be used to mix the collagen solution
with the cross-linker, followed by direct pouring of the result-
ing mixture on the PDMS molds (skip the centrifugation step).
3. For an accurate estimation of the swelling of the collagen
samples, the minimum recommended weight per sample
piece is 10 mg.
4. When cutting the collagen film in stripes for tensile testing,
caution must be taken not to damage the middle region of the
sample (test region). Any defects in that region should result in
automatic rejection of the sample. If possible, samples cut in
the dog bone shape are preferred, as a narrower than the edges
middle area promotes stress concentration in the middle of the
sample (this may not be possible in very fragile samples). Ten-
sile test assays are typically carried out in quintuplicate (n ¼ 5),
meaning that in this case at least two collagen films (3 4 cm)
per condition are needed to obtain an accurate measurement of
the scaffolds’ mechanical properties.
5. C-shaped sample holders made of water-resistant materials can
be used to maintain the collagen films in a stretched position
preventing the samples from folding over onto themselves
when incubated in PBS 1, which facilitates sample thickness
measurement and loading in the tensile test machine. Bench
protector sheets made of paper and a laminated waterproof
polyethylene layer can be used to make these sample holders.
Dry collagen stripes can be glued at their ends to two sample
holders in a “sandwich” conformation. After affixing the sam-
ple in the tensile test machine, sample holders need to be cut in
their middle region to negate potential bias during testing.
6. Two glass panes can be used to give a consistent thickness
measurement for the samples: Use the calipers to measure the
thickness of the glass panes alone, and then repeat the measure-
ment with the sample inserted between the panes.
7. It is essential to ensure that the placement of the sample is at

90 angle to the grip so as to negate potential bias upon test.
8. Young’s modulus should be extracted from the linear region of
the stress/strain curve in between the toe and the failure
regions typical of collagen materials.
9. The number of technical replicates (wells per condition) typi-
cally ranges between 3 and 5. The number of plates to consider
depends on the number of time points and experiments
planned.
10. To minimize the toxicity risks associated with PFA, the fixation
step should be performed in the chemical hood.
11. GTA cross-linking can lead to the autofluorescence of the
collagen scaffolds when imaged with 358, 495, 528, and
540 nm wavelengths. Washing the scaffolds thoroughly with
PBS 1 diminishes the effect of this phenomenon.
12. Automated counting with ImageJ in confluent cells can lead to
inaccuracies. If cells are confluent or over-confluent (nuclei
superposition), manual counting is recommended.
13. Elongation degree of THP-1cells can be employed to assess the
immune reaction in vitro. To do so, using ImageJ manually
describe the shape of 25–50 cells per image and measure the
shape features aspect ratio, circumference, and roundness.
Consider as elongated cells those with an aspect ratio higher
than 3.0 or with an aspect ratio superior to 2 and circumfer-
ence and roundness lower than 0.5. Calculate the percentage
of elongated cells in 4–5 images for each condition. The pres-
ence of a high number of clusters (aggregates of 5 or more cell)
can also be assessed as an indication of inflammatory reaction.
14. When assessing the metabolic activity of cells on scaffolds,
these must be moved to new wells with the help of sterile
tweezers to prevent the action of cells attached to
surrounding TCP.
Acknowledgments
This work has been supported by the Science Foundation Ireland,

Career Development Award Programme (grant agreement num-
ber: 15/CDA/3629); Science Foundation Ireland and the
European Regional Development Fund (grant agreement number:
13/RC/2073); and EU H2020, ITN award, Tendon Therapy
Train Project (grant agreement number: 676338).
References
1. Gelse K, Poschl E, Aigner T (2003) Col- 14. Zeugolis DI, Paul GR, Attenburrow G (2009)
lagens–structure, function, and biosynthesis. Cross-linking of extruded collagen fibers -- a
Adv Drug Deliv Rev 55(12):1531–1546 biomimetic three-dimensional scaffold for tis-
2. Brodsky B, Ramshaw JA (1997) The collagen sue engineering applications. J Biomed Mater
triple-helix structure. Matrix Biol 15 Res A 89(4):895–908. https://doi.org/10.
(8–9):545–554 1002/jbm.a.32031
3. Smith-Mungo LI, Kagan HM (1998) Lysyl 15. Weadock K, Olson RM, Silver FH (1983) Eval-
oxidase: properties, regulation and multiple uation of collagen crosslinking techniques.
functions in biology. Matrix Biol 16 Biomater Med Devices Artif Organs 11
(7):387–398 (4):293–318
4. Yamauchi M, Taga Y, Hattori S, Shiiba M, Ter- 16. Jorge-Herrero E, Fernandez P, Turnay J,
ajima M (2018) Analysis of collagen and elastin Olmo N, Calero P, Garcia R, Freile I,
cross-links. Methods Cell Biol 143:115–132. Castillo-Olivares JL (1999) Influence of differ-
https://doi.org/10.1016/bs.mcb.2017.08. ent chemical cross-linking treatments on the
006 properties of bovine pericardium and collagen.
5. Fu MX, Wells-Knecht KJ, Blackledge JA, Biomaterials 20(6):539–545
Lyons TJ, Thorpe SR, Baynes JW (1994) Gly- 17. van Wachem PB, Zeeman R, Dijkstra PJ,
cation, glycoxidation, and cross-linking of col- Feijen J, Hendriks M, Cahalan PT, van Luyn
lagen by glucose. Kinetics, mechanisms, and MJ (1999) Characterization and biocompati-
inhibition of late stages of the Maillard reac- bility of epoxy-crosslinked dermal sheep col-
tion. Diabetes 43(5):676–683 lagens. J Biomed Mater Res 47(2):270–277
6. Reddy GK (2004) Cross-linking in collagen by 18. Haugh MG, Jaasma MJ, O’Brien FJ (2009)
nonenzymatic glycation increases the matrix The effect of dehydrothermal treatment on
stiffness in rabbit achilles tendon. Exp Diabe- the mechanical and structural properties of
sity Res 5(2):143–153. https://doi.org/10. collagen-GAG scaffolds. J Biomed Mater Res
1080/15438600490277860 A 89(2):363–369. https://doi.org/10.1002/
7. Greenberg CS, Birckbichler PJ, Rice RH jbm.a.31955
(1991) Transglutaminases: multifunctional 19. Wess TJ, Orgel JP (2000) Changes in collagen
cross-linking enzymes that stabilize tissues. structure: drying, dehydrothermal treatment
FASEB J 5(15):3071–3077 and relation to long term deterioration. Ther-
8. Bailey AJ, Light ND, Atkins ED (1980) Chem- mochim Acta 365(1-2):119–128. https://doi.
ical cross-linking restrictions on models for the org/10.1016/S0040-6031(00)00619-5
molecular organization of the collagen fibre. 20. Weadock KS, Miller EJ, Bellincampi LD,
Nature 288(5789):408–410 Zawadsky JP, Dunn MG (1995) Physical cross-
9. Orgel JP, Irving TC, Miller A, Wess TJ (2006) linking of collagen fibers: comparison of ultra-
Microfibrillar structure of type I collagen in violet irradiation and dehydrothermal
situ. Proc Natl Acad Sci U S A 103 treatment. J Biomed Mater Res 29
(24):9001–9005. https://doi.org/10.1073/ (11):1373–1379. https://doi.org/10.1002/
pnas.0502718103 jbm.820291108
10. Glowacki J, Mizuno S (2008) Collagen scaf- 21. Torres DS, Freyman TM, Yannas IV, Spector M
folds for tissue engineering. Biopolymers 89 (2000) Tendon cell contraction of collagen-
(5):338–344. https://doi.org/10.1002/bip. GAG matrices in vitro: effect of cross-linking.
20871 Biomaterials 21(15):1607–1619
11. Cliche S, Amiot J, Avezard C, Gariepy C 22. Chen RN, Ho HO, Sheu MT (2005) Charac-
(2003) Extraction and characterization of col- terization of collagen matrices crosslinked
lagen with or without telopeptides from using microbial transglutaminase. Biomaterials
chicken skin. Poult Sci 82(3):503–509. 26(20):4229–4235. https://doi.org/10.
https://doi.org/10.1093/ps/82.3.503 1016/j.biomaterials.2004.11.012
12. Friess W (1998) Collagen-biomaterial for drug 23. Stachel I, Schwarzenbolz U, Henle T, Meyer M
delivery. Eur J Pharm Biopharm 45 (2010) Cross-linking of type I collagen with
(2):113–136 microbial transglutaminase: identification of
cross-linking sites. Biomacromolecules 11
13. Charulatha V, Rajaram A (2003) Influence of (3):698–705. https://doi.org/10.1021/
different crosslinking treatments on the physi- bm901284x
cal properties of collagen membranes. Bioma-
terials 24(5):759–767
24. Zeugolis DI, Panengad PP, Yew ES, Biomater 9(7):7191–7199. https://doi.org/
Sheppard C, Phan TT, Raghunath M (2010) 10.1016/j.actbio.2013.02.021
An in situ and in vitro investigation for the 33. Orenstein SB, Qiao Y, Klueh U, Kreutzer DL,
transglutaminase potential in tissue engineer- Novitsky YW (2010) Activation of human
ing. J Biomed Mater Res A 92(4):1310–1320. mononuclear cells by porcine biologic meshes
https://doi.org/10.1002/jbm.a.32383 in vitro. Hernia 14(4):401–407. https://doi.
25. Orban JM, Wilson LB, Kofroth JA, El-Kurdi org/10.1007/s10029-010-0634-7
MS, Maul TM, Vorp DA (2004) Crosslinking 34. Witherel CE, Graney PL, Freytes DO, Wein-
of collagen gels by transglutaminase. J Biomed garten MS, Spiller KL (2016) Response of
Mater Res A 68(4):756–762. https://doi.org/ human macrophages to wound matrices
10.1002/jbm.a.20110 in vitro. Wound Repair Regen 24
26. Moshnikova AB, Afanasyev VN, Proussakova (3):514–524. https://doi.org/10.1111/wrr.
OV, Chernyshov S, Gogvadze V, Beletsky IP 12423
(2006) Cytotoxic activity of 1-ethyl-3- 35. Delgado LM, Fuller K, Zeugolis DI (2017)
(3-dimethylaminopropyl)-carbodiimide is Collagen cross-linking: biophysical, biochemi-
underlain by DNA interchain cross-linking. cal, and biological response analysis. Tissue
Cell Mol Life Sci 63(2):229–234. https://doi. Eng Part A 23(19–20):1064–1077. https://
org/10.1007/s00018-005-5383-x doi.org/10.1089/ten.TEA.2016.0415
27. Speer DP, Chvapil M, Eskelson CD, Ulreich J 36. Lotz C, Schmid FF, Oechsle E, Monaghan
(1980) Biological effects of residual glutaralde- MG, Walles H, Groeber-Becker F (2017)
hyde in glutaraldehyde-tanned collagen bioma- Cross-linked collagen hydrogel matrix resisting
terials. J Biomed Mater Res 14(6):753–764. contraction to facilitate full-thickness skin
https://doi.org/10.1002/Jbm.820140607 equivalents. ACS Appl Mater Interfaces 9
28. Gough JE, Scotchford CA, Downes S (2002) (24):20417–20425. https://doi.org/10.
Cytotoxicity of glutaraldehyde crosslinked col- 1021/acsami.7b04017
lagen/poly(vinyl alcohol) films is by the mech- 37. Monaghan M, Browne S, Schenke-Layland K,
anism of apoptosis. J Biomed Mater Res 61 Pandit A (2014) A collagen-based scaffold
(1):121–130. https://doi.org/10.1002/jbm. delivering exogenous microRNA-29B to mod-
10145 ulate extracellular matrix remodeling. Mol
29. Hass V, Luque-Martinez IV, Gutierrez MF, Ther 22(4):786–796. https://doi.org/10.
Moreira CG, Gotti VB, Feitosa VP, Koller G, 1038/mt.2013.288
Otuki MF, Loguercio AD, Reis A (2016) Col- 38. Sanami M, Sweeney I, Shtein Z, Meirovich S,
lagen cross-linkers on dentin bonding: stability Sorushanova A, Mullen A, Miraftab M,
of the adhesive interfaces, degree of conversion Shoseyov O, O’Dowd C, Pandit A, Zeugolis
of the adhesive, cytotoxicity and in situ MMP D (2016) The influence of poly(ethylene gly-
inhibition. Dent Mater 32(6):732–741. col) ether tetrasuccinimidyl glutarate on the
https://doi.org/10.1016/j.dental.2016.03. structural, physical, and biological properties
008 of collagen fibers. J Biomed Mater Res B 104
30. Delgado LM, Bayon Y, Pandit A, Zeugolis DI (5):914–922
(2015) To cross-link or not to cross-link? 39. Marinucci L, Lilli C, Guerra M, Belcastro S,
Cross-linking associated foreign body response Becchetti E, Stabellini G, Calvi EM, Locci P
of collagen-based devices. Tissue Eng Part B (2003) Biocompatibility of collagen mem-
21(3):298–313. https://doi.org/10.1089/ branes crosslinked with glutaraldehyde or
ten.TEB.2014.0290 diphenylphosphoryl azide: an in vitro study. J
31. Brown BN, Londono R, Tottey S, Zhang L, Biomed Mater Res A 67A(2):504–509.
Kukla KA, Wolf MT, Daly KA, Reing JE, Bady- https://doi.org/10.1002/jbm.a.10082
lak SF (2012) Macrophage phenotype as a pre- 40. Petite H, Frei V, Huc A, Herbage D (1994)
dictor of constructive remodeling following Use of diphenylphosphorylazide for cross-
the implantation of biologically derived surgi- linking collagen-based biomaterials. J Biomed
cal mesh materials. Acta Biomater 8 Mater Res 28(2):159–165. https://doi.org/
(3):978–987. https://doi.org/10.1016/j. 10.1002/Jbm.820280204
actbio.2011.11.031 41. Capella-Monsonı́s H, Coentro J, Graceffa V,
32. McDade JK, Brennan-Pierce EP, Ariganello Wu Z, Zeugolis D (2018) An experimental
MB, Labow RS, Michael Lee J (2013) Interac- toolbox for characterization of mammalian col-
tions of U937 macrophage-like cells with lagen type I in biological specimens. Nat Pro-
decellularized pericardial matrix materials: toc 13(3):507–529
influence of crosslinking treatment. Acta
42. Coentro J, Capella-Monsonı́s H, Graceffa V, 44. Zeugolis D, Li B, Lareu R, Chan C, Raghunath

Wu Z, Mullen A, Raghunath M, Zeugolis D M (2008) Collagen solubility testing, a quality
(2017) Collagen quantification in tissue speci- assurance step for reproducible electro-spun
mens. Methods Mol Biol 1627:341–350 nano-fibre fabrication. A technical note. J Bio-
43. Lareu R, Zeugolis D, Abu-Rub M, Pandit A, mater Sci Polym Ed 19(10):1307–1317
Raghunath M (2010) Essential modification of 45. Zeugolis D, Raghunath M (2010) The physio-
the Sircol collagen assay for the accurate quan- logical relevance of wet versus dry differential
tification of collagen content in complex pro- scanning calorimetry for biomaterial evaluation:
tein solutions. Acta Biomater 6(8):3146–3151 a technical note. Polym Int 59(10):1403–1407
Chapter 4
A Biomimetic Model for Mineralization of Type-I Collagen

Fibrils
Shasha Yao, Yifei Xu, Changyu Shao, Fabio Nudelman,
Nico A. J. M. Sommerdijk, and Ruikang Tang
Abstract
The bone and dentin mainly consist of type-I collagen fibrils mineralized by hydroxyapatite (HAP)
nanocrystals. In vitro biomimetic models based on self-assembled collagen fibrils have been widely used
in studying the mineralization mechanism of type-I collagen. In this chapter, the protocol we used to build a
biomimetic model for the mechanistic study of type-I collagen mineralization is described. Type-I collagen
extracted from rat tail tendon or horse tendon is self-assembled into fibrils and mineralized by HAP in vitro.
The mineralization process is monitored by cryoTEM in combination with two-dimensional (2D) and
three-dimensional (3D) stochastic optical reconstruction microscopy (STORM), which enables in situ and
high-resolution visualization of the process.
Key words Bone, Type-I collagen, Calcium phosphate, CryoTEM, STORM
1 Introduction
Type-I collagen is the main organic constitution of the hard tissues

of vertebrates, e.g., the dentin and bone [1–3]. The type-I collagen
fibrils are assembled by ~300-nm-long triple helix molecules which
are ~1.5 nm in diameter [4]. Recent X-ray study indicates that in
type-I collagen fibrils, subunit called “microfibrils” is formed by
five 1D staggered and twisted collagen molecules [5]. These micro-
fibrils pack into superstructures with triclinic symmetry (P1,
a 40.0 Å, b 27.0 Å, c 678 Å, α 89.2 , β 94.6 ,
γ 105.6 ) and form larger collagen fibrils which display a
~67 nm periodical structure. Each ~67 nm unit contains a less
dense gap region and an overlap region, which are, respectively,
~37 nm and ~30 nm long [5], while the values vary slightly
depending on the sources of the collagen [6]. The type-I collagen
Shasha Yao and Yifei Xu contributed equally to this work.
39
40 Shasha Yao et al.
fibrils in bone are embedded by ~4-nm-thick-carbonated hydroxy-

apatite (HAP) nanoplatelets with their c-axis oriented along the
fibril axis [7, 8], which significantly improves the mechanical per-
formances of the fibrils [9, 10]. The formation mechanism of the
intrafibrillar HAP crystals has long been an attractive research topic.
Due to the difficulty of in vivo study of bone formation process, the
majority of the studies on collagen mineralization was performed
in vitro based on biomimetic models, using demineralized collagen
from the bone or dentin [11, 12], collagen sponge [8, 13], but
most commonly self-assembled collagen fibrils [14–17]. Type-I
collagen molecules could self-assemble in vitro at pH ¼ 7 to 10 at
temperatures from 20 to 37 with the presence of electrolytes (e.g.,
KCl) [18, 19], forming fibrils with ~67 nm periodical structures
similar to those found in vivo. It has been shown that collagen
could be intrafibrillarly mineralized in vitro by HAP platelets ori-
ented in c-axis [8–15], with electron diffraction pattern indistin-
guishable to mineralized collagen fibrils found in bones. Charged
macromolecules mimicking the acidic non-collagenous proteins
(NCP) [20] in bone or dentin such as polyaspartic acid (pAsp)
[8, 15], polyacrylic acid (pAA) [17], fetuin [15, 21], and poly
(allylamine) hydrochloride (pAH) [16] were found to be essential
for the in vitro mineralization of the type-I collagen fibrils. This is
probably related to their ability to stabilize amorphous calcium
phosphate (ACP) precursor and inhibit the formation of HAP.
Recently we have found that charged small biomolecules, such as
citrate, could further promote the type-I collagen mineralization by
reducing the interfacial energy between collagen and the ACP
precursor [14]. In summary, the in vitro biomimetic models have
been shown to be a powerful tool which provide profound insight
into the collagen mineralization process [22]. However, these
models were often based on different collagen sources and different
methods to obtain and mineralize the collagen fibrils, which brings
uncertainty for interpreting and comparing experimental results.
In this chapter we describe the protocol we used for assembling
and mineralizing type-I collagen fibrils [14, 15]. In a typical proce-
dure, a type-I collagen (from rat tail tendon or horse tendon) stock
solution was mixed with an assembling buffer solution and then
dripped onto transmission electron microscopy (TEM) grids to
assemble the collagen fibrils on the grids. The grids were then
exposed to the mineralization solution to induce the HAP mineral-
ization. After mineralization, the products were visualized using a
conventional cryogenic TEM (cryoTEM) in combination with
2D/3D STORM images in order to examine the intrafibrillar and
extrafibrillar HAP mineralization of type-I collagen fibrils in situ
with nano-level accuracy. We hope our method will contribute for
establishing a standard experimental model, which will benefit all
the researchers in this field.
Biomimetic Model for Mineralization of Type-I Collagen 41
2 Materials
1. Type-I collagen stock solution (3 mg ml1) from rat tail ten-

don was purchased from Gibco-Invitrogen (USA). Type-I col-
lagen extract from horse tendon was kindly provided by
Dr. Giuseppe Falini (Department of Chemistry, University of
Bologna, Italy).
2. Assembling buffer: 0.3754 g glycine and 1.481 g KCl were
dissolved in 100 ml of deionized water and ultrasonicated for
20 min at room temperature to obtain an assembling buffer
solution containing 50 mM glycine and 200 mM KCl. The pH
value was adjusted to 9.2 with 1 M NaOH solution. The
assembling buffer solution was kept at 4 C prior to use.
3. Mineralization solution: 2 ml of 3.34 mM CaCl2 was dripped
into a 10 ml petri dish, and then 0 to 240 μl of 10 mg ml1
pAsp (27 kDa) stock solutions was added into the calcium
solution. 2 ml phosphate solution containing 19 mM
Na2HPO4 and 300 mM NaCl was then mixed with the above
solution to obtain the mineralization solution.
4. Imaging buffer: 1.6 mg ml1 glucose oxidase, 50 mM Tris,
10 mM NaCl, 50 mM cysteamine, 10% (w/v) glucose,
32 μg ml1 catalase, and 1% (v/v) β-mercaptoethanol. The
pH of the imaging buffer was tuned to 8 by 1 M NaOH and
HCl solution.
5. Assembling buffer for cryoTEM study: 10 mM Hepes
(pH ¼ 7.4), 200 mM NaCl, and 30 mM KH2PO4 (see Note 1).
6. Hepes buffer: 10 mM Hepes (pH 7.4), 150 mM NaCl.
7. Mineralization solution C1: 10 mM Hepes (pH ¼ 7.4),
5.4 mM CaCl2.
8. Mineralization solution C2: 2.7 mM K2HPO4, 10 mM Hepes,
20 mg l1pAsp (Mw ¼2000~11,000 Da).
9. APTES solution: Dissolve 0.5 ml (3-aminopropyl)
triethoxysilane (APTES) into 9.5 ml anhydrous ethanol.
10. The blocking buffer and the washing buffer were purchased
from Beyotime (Product Code, P0023B, China).
11. The type-I collagen rabbit anti-mouse antibody (1:50) and
mouse anti-rabbit Cy3B-conjugated secondary antibodies
(1:100) were supplied by Proteintech (USA).
12. Deionized water and all the solutions were filtered through
0.22 μm Millipore films prior to use.
3 Methods
3.1 Conventional 1. In order to assemble the collagen on TEM grids, 8.33 μl of

TEM Experiments type-I collagen stock solution from rat tail tendon was dripped
into 0.5 ml of assembling buffer solution and incubated at
3.1.1 Self-Assembly
37 C for 20 min with stirring (see Note 2).
of Type-I Collagen Fibrils
2. 200-mesh gold TEM grids (see Note 3) coated by carbon films
were glow discharged for 40 s using a Cressington 208 carbon
coater and then transferred into a glass petri dish.
3. 3 μl of incubated type-I collagen solution was gently dripped
into the carbon film face of the TEM grids (see Note 4).
4. Sealed TEM grids were incubated at 37 C for 12 h (see Note
5) and then gently rinsed with deionized water.
5. The type-I collagen fibrils were further cross-linked with 0.05%
glutaraldehyde for 2 h and gently rinsed with deionized water
and air-dried.
6. To increase the contrast of the periodical structure within the
type-I collagen fibrils, 2% (w/v) uranyl acetate staining solu-
tion was dripped onto the TEM grids loaded with type-I colla-
gen fibrils and then washed after 15 s with deionized water and
air-dried.
3.1.2 Collagen 1. 8 μl of 5% (w/v) NaN3 was added to the mineralization solu-

Mineralization tion to retard bacterial growth. The solutions should be mixed
quickly, and the as-prepared mineralization solution should be
colorless and transparent. Slow mixing would lead to a light
blue-colored, cloudy dispersion (Fig. 1).
2. TEM grids loaded with type-I collagen fibrils were floated
upside down on the mineralization solution at 37 C for desig-
nated times.
3. After mineralization, the TEM grids were taken out and
washed in sequence by deionized water, 25% ethanol, 50%
ethanol, and 100% ethanol.
Fig. 1 Photos showing the effect of mixing speed on preparing the mineralization
solution. (a) The colorless and transparent mineralization solution obtained by
fast mixing. (b) The light blue-colored, cloudy dispersion obtained by slow mixing
3.1.3 TEM Imaging The assembled (Fig. 2), stained (Fig. 3), and mineralized (Figs. 4
and Mineralization Degree and 5) type-I collagen fibrils were visualized by TEM operating at
Calculation 100 kV or 200 kV. The mineralization degree increased with
increasing the pAsp concentration from 0 to 240 μg ml1 (Fig. 4)
and increasing the mineralization time from 6 h to 24 h (Fig. 5).
The mineralization degree (m.d.) was calculated by first
splitting the pixel intensity histogram of the TEM images into
three Gaussian distributions (Fig. 6), which correspond to the
Fig. 2 Conventional TEM image of the assembled type-I collagen fibrils
Fig. 3 Conventional TEM image of the assembled type-I collagen fibrils after
uranyl acetate positive staining, displaying the ~67 nm periodical structures
together with fine band structures. Reproduced with permission [14]. Copyright
2018, Wiley-VCH
Fig. 4 Conventional TEM images of the mineralized collagen-I fibrils in the presence of 0 (a), 10 (b), 15 (c),
50 (d), 120 (e), and 240 (f) μg ml1 pAsp. Insets: selected area electron diffraction (SAED) patterns of the
samples, which match with the HAP diffractions and indicate the oriented crystallization. Reproduced with
permission [14]. Copyright 2018, Wiley-VCH
background, collagen fibrils, and mineralized collagen fibrils,

respectively. The thresholds (T) between the three areas are deter-
mined by T ¼ I2σ [23], where I is the mean pixel intensity of
each Gaussian distribution and σ is the standard. The areas
corresponding to all of the collagen fibrils (S1) and the mineralized
Fig. 5 Conventional TEM images of the collagen-I fibrils mineralized at 6 h (a) and 24 h (b). Insets: SAED
patterns indicating HAP mineralization. Reproduced with permission [14]. Copyright 2018, Wiley-VCH
Fig. 6 Mineralization degree calculation. (a) Segmentation thresholds determined by analyzing the pixel
intensity histogram of the TEM image and splitting the histogram into three Gaussian distributions. Determi-
nation of the regions of the mineralized collagen-I fibrils (b), unmineralized collagen-I fibrils (c), and the
background (d). Reproduced with permission [14]. Copyright 2018, Wiley-VCH
part of the collagen fibrils (S2) could then be segmented from the
image, and m.d. could then be given by comparing S1 and S2
S1
m:d: ¼
S2
For each analysis, at least five samples were examined to obtain

the mean value and standard deviation of m.d.
3.2 CryoTEM 1. 3 mg ml1 of type-I collagen stock solution from rat tail was
Experiments mixed with the assembling buffer for cryoTEM study in a 1:29
ratio. 300 μl of as-prepared incubation solution was applied on
3.2.1 Self-Assembly
a parafilm.
of Type-I Collagen
Extracted from Rat Tail 2. Gold Quantifoil grids were floated upside down on the incuba-
Tendon tion solution droplet (Fig. 7). The parafilm was then covered
by a glass petri dish and incubated for 200 min.
3. After incubation, the grids were transferred to the Vitrobot
(FEI, Mark III) in 100% humidity (see Note 6), blotted for 5 s
by filter papers, and then vitrified in liquid ethane to visualize
the collagen fibrils or transferred to the mineralization solution
(see Subheading 3.2.3).
3.2.2 Self-Assembly 1. 1 g of type-I collagen extracted from horse tendon was mixed
of Type-I Collagen overnight with 10 ml of 50 mM aqueous acetic acid, pH 2.5,
Extracted from Horse centrifuged at 3500 g for 10 min, and the supernatant was
Tendon collected and stored at 4 C.
2. CryoTEM grids were laid on a 15 μl drop of collagen solution
for 10 s. The excess of collagen solution was manually blotted,
and the grids were transferred to a 15 μl drop of Hepes buffer,
for 30 min. This procedure was performed inside a glove box,
where temperature and humidity are controlled at 22 C and
100% relative humidity.
3. The grids were then vitrified as described in Subheading 3.2.1
or transferred to the mineralization solution (see Subheading
3.2.3).
Fig. 7 Photo of the gold Quantifoil TEM grids floating on a droplet of buffer
solution applied on a parafilm
Fig. 8 CryoTEM image of assembled type-I collagen fibrils. (a) Rat tail type-I collagen fibrils, which are
60–80 nm thick. (b) Zoom-in image of the fibrils showing no periodical structure. (c) A horse tendon type-I
collagen fibril, which is ~300 nm thick and clearly shows the ~67 nm periodical structure. The 10 nm particles
highlighted by yellow circle are gold fiducial markers used for tomography
3.2.3 Collagen 1. Mineralization solutions C1 and C2 were mixed 1:1 to prepare

Mineralization the final mineralization solution in a glass petri dish.
2. The gold Quantifoil grids were rinsed in two 300 μl water
droplets applied on parafilm for 1 min each immediately after
collagen assembly to remove the residual salts and then trans-
ferred into the petri dish containing the mineralization
solution.
3. The petri dish was then sealed by parafilm and heated at 37 C
for 72 h for mineralization.
4. After mineralization, the grids were transferred to the Vitrobot
(FEI, Mark III) in 100% humidity, blotted for 5 s by filter
papers, and then vitrified in liquid ethane.
3.2.4 CryoTEM Imaging 1. CryoTEM imaging of the assembled (Fig. 8 and see Note 7)
and then mineralized (Fig. 9) collagen was performed under
~3 μm defocus on a FEI Titan cryoTEM equipped with a field
emission gun (FEG) and operating at 300 kV or a FEI Tecnai
20 (Type Sphera) TEM equipped with a LaB6 filament
operating at 200 kV and a Gatan cryo-holder operating at
170 C.
2. Images were recorded in FEI Titan using a 2 k 2 k Gatan
CCD camera equipped with a post-column Gatan energy filter
(GIF) or in FEI Tecnai using a 1 k 1 k Gatan CCD camera.
The electron dose used is 10 e Å2 per image (see Note 8).
Fig. 9 CryoTEM image of mineralized type-I collagen fibrils. (a) Image showing a mineralized rat tail type-I
collagen fibril, displaying stacks of HAP platelets. (b) SAED pattern of the mineralized rat tail collagen fibril,
which shows a narrow but very weak HAP (002) diffraction arc in the collagen fibril axis direction as
highlighted by the yellow arrow. (c) Image showing a mineralized horse tendon type-I collagen fibril. The
fibril is completely impregnated with HAP platelets, and therefore the ~67 nm periodical structures are not
distinguishable. (d) SAED pattern of the mineralized horse tendon type-I collagen fibril, which clearly shows
the HAP (002) diffraction arc in the collagen axis as highlighted by the yellow arrow. The collagen fibril axis
directions in (c) and (d) are highlighted by the blue arrows
3.3 STORM 1. Type-I collagen fibrils (rat tail) were immobilized on the
Experiments amino-silanized LCCD. In the experiments, 200 μl APTES
was dripped onto a LCCD substrate avoiding the light. The
3.3.1 Modification
APTES solution was then removed and dried in an oven at
of the Laser Confocal
100 C prior to use (see Note 9).
Culture Dish (LCCD)
and Self-Assembly 2. 100 μl of 50 μg ml1 type-I collagen (rat tail) solution was
of Type-I Collagen Fibrils dripped on the modified LCCD substrate and incubated at
37 C for 12 h and then gently rinsed with deionized water.
3. The type-I collagen fibrils were further cross-linked with
0.05 wt% glutaraldehyde for 2 h and then gently rinsed with
deionized water (see Note 10).
3.3.2 Staining The immunofluorescent staining was performed according to the

and Mineralization following steps:
of the Type-I Collagen
1. The type-I collagen (rat tail) fibrils were incubated with block-
Fibrils
ing buffer for 1 h and then incubated with the type-I collagen
rabbit anti-mouse antibody at 37 C for another 2 h.
2. The samples were washed by the washing buffer for three times
(10 min each) and incubated with mouse anti-rabbit Cy3B-
conjugated secondary antibodies for 1 h at 37 C.
3. The samples were washed by three extra times in the washing
buffer (10 min each) and mineralized with the mineralization
solution.
The mineralization process was as follows:
1. 2 ml of 3.34 mM CaCl2 solution and 48 μl of 10 mg ml1 pAsp
solution were dropped into a LCCD which was loaded with
immunofluorescence-stained collagen.
2. 2 ml of solution containing 19 mM Na2HPO4 and 300 mM
NaCl was mixed into the solution.
3. The samples were mineralized in the as-prepared mineraliza-
tion solution at 37 C for 24 h and rinsed with deionized water
for three times.
4. The mineralized type-I collagen fibrils were labeled by 2 ml of
10 μM calcein which specifically bind with HAP for 20 min at
room temperature. Residual calcein was removed by washing
using deionized water.
3.3.3 STORM Imaging The type-I collagen (rat tail) fibrils were immersed in the imaging
buffer before STORM imaging.
1. 2D STORM imaging experiments were performed on inverted
optical microscope. For z-stack images of the mineralized type-
I collagen fibrils, the images were recorded, processed, and
analyzed using the Imaris software.
2. 3D STORM imaging experiments were performed on a Nikon
N-STORM microscope equipped with a 100 oil immersion
objective and an Andor camera [24].
3. To identify the zone of interest, a low-magnification fluores-
cence image was acquired prior to STORM imaging.
4. After switching to higher magnification (100 objective), con-
ventional fluorescence images were first acquired. STORM data
acquisition was then started using imaging cycles at one frame
of activation laser illumination (405 nm laser) followed by five
frames of imaging laser illumination (561 nm laser). The inte-
gration time of the camera was set to the one frame mode with
an EM gain of 305,000–10,000 cycles per channel, which was
Fig. 10 The xy projects of the 2D STORM images of the mineralized collagen-I

fibrils, with collagen labeled by Cy3B and showing red fluorescence. The HAP
was labeled with calcein, which emitted green fluorescence. Reproduced with
permission [14]. Copyright 2018, Wiley-VCH
used for the reconstruction of each 3D super-resolution image.

The STORM images confirm the intrafibrillar mineralization of
type-I collagen fibrils.
5. In the 2D images, the type-I collagen fibrils were labeled before
mineralization, which emitted red fluorescence. After minerali-
zation, the HAP was labeled with calcein, which emitted fluo-
rescence with the wavelength of green light. The data showed
that the ACP infiltrated into the type-I collagen fibrils and
transformed into HAP inside the type-I collagen fibrils
(Fig. 10).
6. The z-slice of the 3D STORM images of the mineralized type-I
collagen fibrils indicated the mineralization of HAP inside and
outside the type-I collagen fibrils. Besides, the HAP was homo-
geneously dispersed in the type-I collagen fibrils (Fig. 11).
4 Notes
1. This buffer solution induces a faster assembly of rat tail type-I

collagen, which benefits the cryoTEM study.
Fig. 11 The z-slices of the STORM image of the mineralized collagen-I fibrils,
indicating that the mineralization of HAP inside and on the surface of the
collagen-I fibrils. Reproduced with permission [14]. Copyright 2018, Wiley-VCH
2. Incubation time less than 20 min will lead to an inhomoge-

neous dispersion of the type-I collagen fibrils on the TEM grid
(Fig. 12).
3. The mineralization experiment could not work on copper grids
because of copper leaching in the solution which inhibits the
crystallization of HAP [25].
4. 3 μl is the optimal volume which neither dries very quickly (less
than 4 h) nor leads to overflow of the TEM grids.
5. The mineralization solution was replaced each 6 h to keep the
supersaturation constant.
6. The solution left between the tweezers has to be carefully
removed by filter paper before the grids were transferred into
Vitrobot, in order to prevent the tweezers from being frozen
during the vitrification process.
Fig. 12 Conventional TEM image of the TEM grid inhomogeneously covered by

collagen-I fibrils
Fig. 13 CryoTEM image of collagen fibrils after exposure of 200 e/Å2, showing
the beam damage (white spots, as highlighted by the yellow circle)
7. No ~67 nm periodical structure could be visualized without

staining for the assembled rat tail collagen fibrils, while the
horse tendon collagen fibrils clearly show the periodical struc-
ture. This is due to the fact that these rat tail collagen fibrils
(60–80 nm in diameter) are much thinner than the horse
tendon collagen fibrils (~300 nm in diameter).
8. Collagen fibril is very sensitive to the electron beam, and beam
damage could be clearly observed after the fibril was exposed to
200 e/Å2 (Fig. 13).
Fig. 14 2D STORM image of the nonhomogeneous stained collagen-I fibrils
9. The LCCD must be fresh and clean prior to use. After the
removal of the APTES solution, the modified LCCD was ultra-
sonicated with anhydrous ethanol and deionized water for
20 min, respectively, to remove the physical absorbed chemi-
cals. 50 μg ml1 of type-I collagen solution was obtained by
mixing 5 μl of type-I collagen stock solution with 295 μl of
assembling solution and then incubating the mixed solution at
room temperature for 20 min. After 12 h, the LCCD was
gently rinsed with deionized water for at least three times to
remove unassembled type-I collagen.
10. Incubation times of antibodies less than 2 h or calcein labeling
time less than 20 min would lead to inhomogeneous stained
type-I collagen samples (Fig. 14). The labeling process should
avoid the light.
References
1. Weiner S, Traub W, Wagner HD (1999) 4. Hodge AJ (1963) Recent studies with the elec-
Lamellar bone: structure-function relations. J tron microscope on ordered aggregates of the
Struct Biol 126:241–255 tropocollagen macromolecule. In: Ramachan-
2. Beniash E, Traub W, Veis A, Weiner S (2000) A dran GN (ed) Aspects of protein structure.
transmission electron microscope study using Academic Press, New York, pp 289–300
vitrified ice sections of predentin: structural 5. Orgel JPRO, Irving TC, Miller A, Wess TJ
changes in the dentin collagenous matrix (2006) Microfibrillar structure of type I colla-
prior to mineralization. J Struct Biol gen in situ. Proc Natl Acad Sci U S A
132:212–225 103:9001–9005
3. Goldberg M, Kulkarni AB, Young M, Boskey A 6. Quan BD, Sone ED (2013) Cryo-TEM analy-
(2011) Dentin: structure, composition and sis of collagen fibrillar structure. Methods
mineralization. Front Biosci (Elite Ed) Enzymol 532:189–205
3:711–735
7. Weiner S, Traub W (1986) Organization of 17. Wang Y et al (2012) The predominant role of
hydroxyapatite crystals within collagen fibrils. collagen in the nucleation, growth, structure
FEBS Lett 206:262–266 and orientation of bone apatite. Nat Mater
8. Olszta MJ et al (2007) Bone structure and 11:724
formation: A new perspective. Mater Sci Eng 18. Jiang F, Hörber H, Howard J, Müller DJ
R Rep 58:77–116 (2004) Assembly of collagen into microrib-
9. Gao HJ, Ji BH, Jager IL, Arzt E, Fratzl P bons: effects of pH and electrolytes. J Struct
(2003) Materials become insensitive to flaws Biol 148:268–278
at nanoscale: Lessons from nature. Proc Natl 19. Williams BR, Gelman RA, Poppke DC, Piez
Acad Sci U S A 100:5597–5600 KA (1978) Collagen fibril formation. Optimal
10. Nair AK, Gautieri A, Chang S-W, Buehler MJ in vitro conditions and preliminary kinetic
(2013) Molecular mechanics of mineralized results. J Biol Chem 253:6578–6585
collagen fibrils in bone. Nat Commun 4:1724 20. George A, Veis A (2008) Phosphorylated pro-
11. Chen J, Burger C, Krishnan CV, Chu B, Hsiao teins and control over apatite nucleation, crys-
BS, Glimcher MJ (2005) In vitro mineraliza- tal growth, and inhibition. Chem Rev
tion of collagen in demineralized fish bone. 108:4670–4693
Macromol Chem Phys 206:43–51 21. Price PA, Toroian D, Lim JE (2009) Minerali-
12. Wang J et al (2013) Remineralization of dentin zation by inhibitor exclusion the calcification of
collagen by meta-stabilized amorphous calcium collagen with fetuin. J Biol Chem
phosphate. CrystEngComm 15:6151–6158 284:17092–17101
13. Olszta M, Douglas E, Gower L (2003) Scan- 22. Nudelman F, Lausch AJ, Sommerdijk NAJM,
ning electron microscopic analysis of the min- Sone ED (2013) In vitro models of collagen
eralization of type I collagen via a polymer- biomineralization. J Struct Biol 183:258–269
induced liquid-precursor (PILP) process. Cal- 23. Smeets PJM et al (2017) A classical view on
cif Tissue Int 72:583–591 nonclassical nucleation. Proc Natl Acad Sci U S
14. Shao C et al (2018) Citrate improves collagen A 114:E7882–E7890
mineralization via interface wetting: a physico- 24. Bates M, Jones SA, Zhuang X (2013) Stochas-
chemical understanding of biomineralization tic optical reconstruction microscopy
control. Adv Mater 30:1704876 (STORM): a method for superresolution fluo-
15. Nudelman F et al (2010) The role of collagen rescence imaging. Cold Spring Harb Protoc
in bone apatite formation in the presence of 2013(6):498–520 pdb.top075143
hydroxyapatite nucleation inhibitors. Nat 25. Nudelman F, Bomans PH, George A, Sommer-
Mater 9:1004–1009 dijk NA (2012) The role of the amorphous
16. Niu LN et al (2017) Collagen intrafibrillar phase on the biomimetic mineralization of col-
mineralization as a result of the balance lagen. Faraday Discuss 159:357–370
between osmotic equilibrium and electroneu-
trality. Nat Mater 16:370–378
Part II
Proteomics Analysis of Collagen

Chapter 5
Investigation of Triple-Helix Collagen Hydroxylation

by Solid-State NMR Spectroscopy
Wing Ying Chow
Abstract
Solid-state nuclear magnetic resonance spectroscopy (ssNMR) is an emerging technique in structural
methods of studying collagen proteins, capable of identifying features on an atomic length scale in tissues
and protein samples without extensive extraction or purification. Hydroxylation is a key posttranslational
modification of collagen that gives rise to distinctive signals in the ssNMR spectrum of collagen proteins.
Here we outline the type of information that ssNMR can provide and describe the procedures involved in a
ssNMR structural study, with particular focus on using dynamic nuclear polarization to enhance sensitivity
for detecting hydroxylysine residues by ssNMR.
Key words Solid-state nuclear magnetic resonance spectroscopy, NMR, Proline, Lysine, Hydroxyl-
ation, Hydroxyproline, Hydroxylysine, Isotopic enrichment/labelling, Dynamic nuclear polarization
1 Introduction
Hydroxylation is an important posttranslational modification

(PTM) for all families of collagen proteins, contributing to the
stability of the triple helix [1] and also providing sites for glycosyla-
tion and cross-linking [2]. Mass spectrometry-based studies can
yield highly sensitive and quantitative information on the extent
and location of modification, yet the information of the structural
role of the hydroxylated species is lost in the extraction and purifi-
cation processes required. Solid-state NMR, X-ray crystallography,
solution-state NMR, and more recently cryo-electron microscopy
are the key techniques to generate atomic-level structural informa-
tion on biomolecules. Compared to other techniques, ssNMR has
the advantage that it can be directly applied to highly heteroge-
neous and solid samples, such as tissue and extracellular matrix
(ECM) samples.
57
58 Wing Ying Chow
1.1 NMR Overview Nuclear magnetic resonance spectroscopy [3–5] relies on the fact
that nuclei have a quantum property called spin. Spin arises from
the number of protons and neutrons in the nuclei, taking values of
0, 1/2, 1, 3/2, and so on; certain nuclei (e.g., 1H, 13C) have a value
of spin >0, and these nuclei can be detected by NMR, i.e., they are
NMR-active. For elements with different isotopes, the nuclei of
these isotopes often have different values of spin. For example, of
the isotopes of carbon, the most abundant (12C) has zero spin and
hence cannot be detected by NMR. On the other hand, 1H and 31P,
both highly abundant, have spin and can be detected by NMR.
Almost all of the nuclei that are used for detection in NMR are
stable (i.e., nonradioactive) isotopes.
In most biomolecular NMR, the focus is on experiments that
involve spin-1/2 nuclei, commonly 1H, 13C, 15N, and 31P. These
nuclei have the property of having two quantum states which have
different energies when placed in a magnetic field. There are appli-
cations detecting NMR of higher-spin nuclei such as 2H (spin-1),
14
N (spin-1), 17O (spin-5/2), 23Na (spin-3/2), and 43Ca (spin-7/
2); however, these often give rise to more complicated NMR
spectra and therefore are beyond the scope of this chapter.
Transitions between the two states of a spin-1/2 nuclei in a
magnetic field occur at a frequency that is determined by an intrin-
sic property of the nuclei and by the strength of the magnetic field.
It is therefore necessary to note which nuclei is being detected, as at
9.4 T, the 1H (proton) frequency is 400 MHz, while 13C would
have a frequency of 100 MHz and 15N would have a frequency of
40 MHz. It is common to refer to NMR spectrometers by the 1H
frequency, so a 9.4 T magnet is commonly called a 400 MHz
spectrometer and an 18.8 T magnet, an 800 MHz spectrometer.
A key part of NMR equipment is the probe head, which sits in
the core of the magnet and plays several important roles: firstly, it
places the sample in the region of the magnet with the most
homogeneous magnetic field; secondly, it regulates the sample
temperature; thirdly, for solid-state NMR, the probe head contains
the stator, which can be adjusted to the magic angle and regulates
the spinning of the rotor, necessary for producing resolved spectra
(see below); fourthly and perhaps most importantly, a coil in the
stator surrounds the sample and converts magnetic behavior of the
spins to an electric current, which is amplified and can be detected
by the resonant circuits. The circuits can be tuned to the specific
characteristic frequency for 1H and other nuclei to deliver radio-
frequency (RF) pulses to manipulate the spin transitions; the com-
bination of the number, timing, spacing, length, and power of the
pulses is what makes up different NMR experiments by selecting for
specific spin interactions.
Since the magnetic behavior of the spins occur at specific fre-
quencies, the electric current induced and measured also shows an
oscillating behavior over time, which has an overall decrease in
Solid-state NMR Investigations of Collagen Hydroxylation 59
signal due to relaxation and is called the free induction decay (FID).
Generally, the FID is not analyzed directly but instead subjected to
Fourier transform which provides a spectrum with a frequency axis.
Fourier transform procedures are routine on modern spectrometers
and computers and are therefore not further described in this
chapter.
1.2 Chemical Shift One of the key advantages of NMR is its sensitivity to small varia-
tions in chemical structure. Differences in the number and types of
bonds lead to relatively small (parts per million level) changes in the
transition frequency. The simplest NMR spectra, so-called 1D
spectra, are plots of the intensity of transition frequencies of one
particular type of nuclei, giving insight into the types of different
chemical environments for these nuclei in the sample.
The transition frequency scales with the external applied mag-
netic field. Therefore, the spectrum is commonly presented with a
frequency scale called the chemical shift (δ), which is relative to parts
per million (ppm) of the transition frequency of the nuclei in a
reference sample, rather than the directly measured frequencies in
MHz. The chemical shift scale has the useful property of enabling
direct comparison of spectra obtained on spectrometers with dif-
ferent magnetic field strengths.
While the method of referencing the chemical shift scale can
seem like a technical detail, it is important for accurate reporting
and correct comparisons to literature-reported chemical shift
values. For 1H and 13C NMR, the chemical shift ppm scale is
referenced by setting the signal of TMS (tetramethylsilane) or
DSS (2,2-dimethyl-2-silapentane-5-sulfonic acid) to zero ppm.
TMS is used by publications in synthetic chemistry as per IUPAC
recommendations [6], while DSS is more commonly used in bio-
molecular NMR spectroscopy due to its better water solubility
[7]. For 13C, the DSS-referenced chemical shift values are greater
than the TMS-referenced values by approximately 2.7 ppm, so it is
important to take this conversion factor into account when com-
paring to literature values.
In solution-state NMR, referencing is often carried out using
the solvent signal. However, this is not the recommended approach
in solid-state NMR due to the highly variable amount of solvent,
the wider temperature range, and the potential inhomogeneity in
solvent distribution, all of which can lead to shifting or broadening
of the solvent signal, complicating attempts to reference and/or
reproduce the spectra. Instead, DSS is sometimes added to wet
pellet protein samples as an internal reference or external referen-
cing is used where DSS or TMS is not added to the sample but
measured separately. In this chapter, we use an external referencing
scheme based on TMS, as we would prefer to facilitate comparisons
with synthetic compounds that model collagen glycosylation [8].
60 Wing Ying Chow
Using 1D NMR spectra and chemical shifts alone, it is possible

to quickly identify the components present in the tissue or collagen
sample. In particular, solid samples containing a high proportion of
collagen should show 13C signals at 42.5 ppm (glycine Cα), 70 ppm
(hydroxyproline Cγ), and a broad signal around 170.5 ppm (amide
C0 ) with a shoulder at 169–170 ppm (glycine C0 ). Representative
spectra of collagen-containing samples are shown in Subheading 3.
NMR is highly sensitive to hydroxylation of proline and lysine,
as hydroxylation leads to a large change of the chemical shift of the
carbon signal. Proline Cγ is found at 25 ppm, but hydroxyproline
Cγ is found at 70 ppm. Similarly, lysine Cδ is found at 28 ppm, but
hydroxylysine Cδ is found at 70 ppm. While we can distinguish
between the unmodified and the hydroxylated forms, we see that
both hydroxyproline and hydroxylysine have a 13C signal at
70 ppm, and indeed, in a 1D NMR spectrum, it is not possible to
determine whether the signal arises from a hydroxylated proline or
lysine without other prior knowledge, such as by selectively label-
ling one of these amino acids.
Multidimensional NMR spectra can improve resolution and
enable signals that overlap in 1D to be more clearly identified and
assigned, thus allowing more structural information to be
extracted. Such spectra provide information on correlations
between pairs or a series of nuclei, usually in a distance- or bond-
dependent manner. The nuclei correlated can be the same type
(e.g., 13C–13C), or different (e.g., 13C–15N). In the specific case
of hydroxylated residues in collagen, it is possible to distinguish
hydroxyproline and hydroxylysine since the closest carbons have
different chemical shifts (Fig. 1), meaning that crosspeaks are
resolved from each other in 2D 13C–13C spectra. Another method
to distinguish between the two is to find correlations between the
hydroxylysine sidechain nitrogen and Cγ in 2D 13C–15N spectra, as
the 15N chemical shift of the lysine and hydroxylysine side chain is
very different from that of the backbone amide.
While there is mathematically no limit to the number of dimen-
sions and the length of NMR-active nuclei series thus detected,
with more transfers/dimensions leading generally to better resolu-
tion, NMR phenomena are subject to relaxation, and there is a
sensitivity loss for each additional dimension. While it is useful to
have multiple dimensions for separating overlapping signals, in this
chapter we will only cover experiments up to two dimensions which
are routinely achievable.
1.3 Solid-State NMR The lack of tumbling motion of the sample in solid-state NMR
and Magic Angle complicates the spectra compared to those of solution-state NMR.
Spinning Since the interaction of the nuclei and the magnetic field is aniso-
tropic and depends on the orientation of surrounding atoms and
bonds, the signals in solid-state NMR spectra of static noncrystal-
line samples are greatly broadened. Moreover, through-space
Fig. 1 Typical 13C and 15N chemical shifts of proline, hydroxyproline, lysine, and
hydroxylysine. For proline, hydroxyproline, and lysine, variations are usually less
than 2 ppm. For hydroxylysine, variations of up to 5 ppm are possible due to
structural changes induced by glycosylation, glycation, and cross-linking.
Changes due to hydroxylation are shown in bold text. Chemical shift values
are referenced to TMS to enable easy comparison to the synthetic chemistry
literature on various hydroxylysine modifications
dipolar interactions between nuclei are also highly anisotropic and

further broaden and distort the NMR signals. Magic angle spinning
(MAS) is a key technique for overcoming these anisotropic effects.
By rotating the sample in a rotor that is oriented along the magic
angle (54.7 ) at thousands or tens of thousands of rotations per
second, it is possible to average out the effect of these anisotropic
interactions, significantly reducing the line-broadening effects and
enabling us to recover the isotropic chemical shifts that are rou-
tinely observed in solution-state NMR experiments.
The effectiveness of MAS depends on the strength of the
interaction and the inherent frequency distribution arising from
chemical structures. It is possible to record well-resolved 13C and
15
N spectra with a modest 8 Hz MAS rate, but it is not possible to
recover well-resolved 1H spectra even at a 60 Hz MAS rate, due to
the much stronger homonuclear 1H–1H dipolar interactions.
Therefore, in this chapter, we will focus on obtaining solid-state
NMR spectra detected on 13C.
When the MAS rate is lower than or not significantly higher
than the strength of the interaction, spinning sidebands are
observed as smaller intensity signals located at multiples of the
spinning rate from the centerband in the spectrum. Sidebands can
be easily recognized by changing the MAS rate, as sidebands would
62 Wing Ying Chow
change location while the centerband would not. In proteins, the

aliphatic signals (5–65 ppm) generally do not give rise to significant
spinning sidebands above a MAS rate of 5Hz, while the amide
carbon signals (168–175 ppm) can give spinning sidebands. Some
MAS rates are suggested in the Notes to avoid placing amide
carbon spinning sidebands in the aliphatic carbon region.
It is important to ensure that the rotor containing solid samples
are packed in a balanced manner, with the sample mass evenly
distributed through the rotor, to enable stable MAS. In the case
of imbalance, the sample can fail to spin faster than 1 Hz and can
also lead to unstable spinning and, in the worst case, a rotor crash.
Apart from destroying the sample, rotor crashes often also damage
the NMR equipment. If in doubt, the sample should be repacked
before attempting MAS.
1.4 Sensitivity NMR spectroscopy at room temperature can be rather insensitive.

of NMR Experiments The insensitivity, in the case of biomolecules, is due to two factors:
firstly, the low natural abundance of NMR-active nuclei and, sec-
ondly, the physics of NMR phenomena. Here we outline strategies
to overcome both of these issues.
1.4.1 Isotopic To overcome the challenge of low natural abundance of

Enrichment NMR-active nuclei, the common strategy is to produce isotopically
enriched samples, particularly in 13C and 15N. These types of
samples are frequently referred to as “labelled” in NMR literature.
In biomolecular NMR, the standard protocol is to overexpress the
protein of interest in E. coli, and supplement the culture medium
with 13C-enriched glucose and 15N ammonium salts [9]. While a
recombinant collagen analogue can be produced by bacteria [10],
successful production of native triple-helical collagen protein gen-
erally requires PTMs that are not found in bacterial expression
systems. We have previously shown that it is possible to generate
isotopically enriched collagen-containing ECM for ssNMR both by
using mammalian cell culture and by feeding mice with a suitable
diet [11, 12], based on the stable isotope labelling with amino acid
(SILAC) methodology used in mass spectrometry [13].
It is important to define the amino acids of interest at the outset
of the study, prior to any sample preparation, in order to select the
most cost-effective labelling scheme. Since solid-state NMR reveals
all labelled species within the sample, it is important to ensure that
samples are not contaminated with other labelled metabolites.
For the case of tissue samples, care should be taken so that a
homogeneous sample of a single tissue type is dissected. Ideally, the
tissue type investigated should be predominantly collagen, even
better if it is one particular type of collagen. In the case of cell
culture samples, it is ideal to select a culture that produces ECM
rich in a single type of collagen. Currently, it is challenging to use
ssNMR to study completely heterogeneous protein samples where
collagen is only a minor component, as signals from other proteins

are likely to overlap with those arising from collagen. However,
there is an inherent advantage in studying hydroxylation of prolines
and lysines, which are PTMs exclusively found in collagen proteins.
Even if collagen is a minor component of the tissue, the methods
described in this chapter should be able to detect these types of
residues and identify the presence of collagen.
Prolines are nonessential. When supplemented in the medium/
diet, it does not show a large degree of conversion into other amino
acids apart from generating hydroxyproline. Samples labelled with
13
C-proline show ssNMR signals from both proline and hydroxy-
proline under routine ssNMR conditions.
Lysines are essential for mammals. Therefore, in order to
achieve isotope enrichment of lysine residues, it is necessary to
provide lysine as 13C-only or 13C, 15N labelled amino acids in the
cell culture media for generating labelled ECM or in the diet for
generating labelled tissue. In our experience, lysine labelling of
mammalian ECM/tissues is generally clean, i.e., with essentially
no transfer or scrambling of the isotope label to other non-lysine
residues or sugars. During collagen biosynthesis, a small part of
labelled lysine is converted to labelled hydroxylysine, at a level of
less than 2% of the collagen triple helix. At this level, even with
labelling, routine ssNMR is unable to detect lysine, and further
signal enhancement is required, such as by dynamic nuclear
polarization.
1.4.2 Dynamic Nuclear The detectable signal in NMR phenomena is based on population
Polarization in Solid- differences between the spin quantum states, which is termed the
State NMR polarization. The energy differences between spin quantum states
are much smaller than thermal energy at room temperatures, which
leads to a very small population difference, a very small polariza-
tion, and consequently a very small detectable signal. One of the
techniques to improve polarization in solid-state NMR is dynamic
nuclear polarization (DNP) [14], which has undergone commer-
cialization and rapid development in the recent years. DNP
enhancement relies on the transfer of polarization from electrons
to nuclei, either directly or through a spin diffusion process,
thereby achieving signal enhancement up to a theoretical limit of
660 for 1H (and even larger for other nuclei). Assuming no loss of
sensitivity via depolarization effects, the time saving is equal to the
signal enhancement squared. In other words, when an enhance-
ment factor of 10 is obtained relative to some reference spectrum,
an experiment with the same signal-to-noise ratio as the reference
spectrum can be conducted in 1/100th of the time. Signals that
would take weeks of signal averaging before accumulating sufficient
intensity above the noise level under routine ssNMR conditions can
be observed in hours with sufficient DNP enhancement.
64 Wing Ying Chow
DNP enhancement requires three main components: one or

more unpaired electron(s) that is in close proximity to the nuclear
spins of interest, a microwave source to enable the polarization
transfer, and a temperature of below 200 K to ensure the efficiency
of the transfer. Unpaired electrons are generally provided in the
form of a small molecule containing stable radicals, and biradicals
were found to be effective as they can produce DNP enhancement
via a cross effect mechanism involving two electron spins and one
nuclear spin, with water-soluble nitroxide-based biradicals being
the current conventional choice for biological samples. The current
commercial implementation of equipment for DNP enhancement
uses a gyrotron as the microwave source, with a low temperature
MAS cabinet that can achieve running temperatures of 100 K. In
general, DNP enhancement improves by lowering the temperature;
therefore, we recommend that experiments be conducted at the
lowest stable temperature whenever possible.
There are a few drawbacks of the DNP enhancement proce-
dure. Firstly, due to the low temperatures required, molecular
motion is significantly slowed down. The lack of molecular motion
means that molecules are trapped in a range of conformations, all
with slightly different chemical shifts, thus leading to broadening of
the signals and a reduction in the resolution. Fortunately, the
resolution loss in DNP is not a significant issue for detecting
hydroxylated residues in collagen since the unmodified and the
hydroxylated forms have such different chemical shifts (45 ppm
difference) that there is effectively no chance of overlap. Secondly,
as a stable biradical is doped into the sample, DNP-enhanced
ssNMR cannot be considered a completely nondestructive tech-
nique where the sample can be cleanly and easily recovered, unlike
normal ssNMR.
2 Materials
2.1 NMR Equipment 1. A solid-state NMR spectrometer equipped with a MAS

probe head tuned to the frequencies of the required nuclei
(see Note 1). For the rest of this chapter, quantities mentioned
are specific for a restricted volume 3.2 mm diameter MAS rotor
with a volume of 30 μL.
2. For studies involving hydroxylysines or other residues or com-
ponents that occur at a level less than 2% of the collagen triple
helix, DNP enhancement is strongly recommended. A system
capable of DNP enhancement includes in addition to the
spectrometer:
(a) A microwave source (usually a gyrotron) at a commensu-
rate electron transition frequency (e.g., 263 GHz gyro-
tron at 9.7 T for a 400 MHz NMR spectrometer at 9.4 T).
(b) A waveguide from the gyrotron to the DNP probe head.

(c) A cooling cabinet capable of maintaining MAS gas flows at
temperatures below 150 K (see Note 2).
(d) A DNP probe head that permits microwave irradiation
and also thermally isolates the cooling gas flows of the
sample rotor from the inner bore of the magnet.
The system should provide sufficient stability in MAS and
thermal isolation for a minimum of 12 h.
3. Zirconia or sapphire MAS rotors of a commensurate diameter
to the stator within the probe (see Note 3).
4. Vespel or zirconia MAS rotor caps of a commensurate diameter
(see Note 4).
2.2 Solid Biological While general procedures of preparing samples for DNP NMR
Samples experiments on water-soluble proteins, assemblies, and precipitates
have been described [21], since most collagen samples are poorly
soluble in water, we provide the procedures described below, which
were inspired by prior applications of DNP signal enhancement on
porous materials [22] and can be applied on a wide range of solid
samples. For example, ssNMR experiments have been successfully
performed on tissues dissected from patients or model organisms,
on protein pellets such as those obtained from cell culture-derived
ECM, and on freeze-dried triple-helical peptides obtained from
solid-phase peptide synthesis [11, 12]. 13C-enriched material is
necessary for studies involving hydroxylysine and recommended
for studies involving hydroxyprolines. While this chapter will not
provide a full description of isotope enrichment procedures, which
can vary greatly depending on the scientific question and the type of
samples involved, some general requirements of samples suitable
for solid-state NMR experiments are provided below.
1. NMR experiments can be successfully carried out with slightly
less than 10 mg of isotope-enriched sample, but more is better
for both detection and spinning stability. Up to 30 mg of sample
can usually be packed into a 3.2 mm diameter rotor (see Note 5).
2. Although a high level of isotope enrichment is desirable, a level
of 20–50% can be sufficient (see Note 6).
3. Avoid freeze-drying tissue samples (see Note 7).
4. Avoid mechanical milling of tissue samples (see Note 8).
5. The use of fixatives or preservatives are generally not recom-
mended, as these substances can give rise to irrelevant but
strong signals in the NMR spectra, especially in the case of
natural abundance samples (see Note 9).
6. If a cell lysis step is carried out, it is recommended to rinse the
ECM with phosphate buffered saline (1) or distilled water to
remove any cell debris.
66 Wing Ying Chow
2.3 Stable Radicals 1. Water-soluble binitroxide radicals such as bcTol-M [17] and
for DNP Enhancement AMUPol [18]. Store away from light.
2. Heavy water (D2O).
3. Distilled water (H2O) (for freeze-dried samples only).
3 Methods
3.1 Preparing Carry out this procedure at room temperature. Try to avoid work-
the Radical Stock ing at a bench which is under direct sunlight, as this can lead to
Solution for DNP quenching of the radicals.
Enhancement 1. If the sample is hydrated, use D2O as the solvent. Otherwise, if
the sample is freeze-dried, use a mix of D2O and H2O in a ratio
of 9:1 (v/v) as the solvent. Allow for 4–10 μL of radical
solution for each ssNMR sample to estimate the volume of
radical stock solution required.
2. If using AMUPol, calculate the mass required for a 30 mM
solution. If using bcTol-M, calculate the mass required for a
40 mM solution. Weigh out the mass required using a balance
with sufficient accuracy (see Note 10).
3. Dissolve the solid radical thoroughly with the required amount
of water (pure D2O or mixed D2O/H2O). Check carefully for
any undissolved crystals and sonicate if necessary. Label, wrap
the radical stock solution container with tinfoil to protect from
light, and store in the refrigerator until use.
3.2 Packing For most collagen-containing samples, this procedure can be car-
the ssNMR Rotor ried out at room temperature. In the case of highly temperature-
sensitive biological samples, the procedure can be carried out at low
temperature. The procedures below are specifically for 3.2 mm
diameter rotors.
1. Weigh the empty ssNMR rotor with the cap prior to sample
packing.
2. Weigh the sample. If there is more than 30 mg of sample,
return the excess to storage.
3. If the sample is solid tissue, cut or break up the sample into
pieces that are no larger than 1.5 mm 1.5 mm 4 mm,
keeping the cut pieces in relatively similar sizes where possible.
4. Fill the sample into the rotor with occasional compaction to
ensure an even weight distribution, which is important for
successful MAS. Stop filling when the sample height in the
rotor is about 3 mm below the top edge of the rotor. When
this happens, or when the sample is all filled into the rotor,
weigh the packed rotor to determine the mass of sample in the
rotor.
5. If there is insufficient sample mass to fill the rotor to this point,

check whether the filling length of the sample is centered in the
rotor. If not, remove the sample completely from the rotor, and
use spacing material below the sample to ensure that the sample
is centered lengthwise in the rotor. Weight the rotor to deter-
mine the mass of the spacing material.
If DNP enhancement is not required, skip to step 9 at this
point.
6. If less than 12 mg of sample was filled into the rotor, pipette
4 μL of radical solution into the rotor, directly on top of the
packed sample (see Note 11).
7. If more than 12 mg of sample was filled into the rotor, first take
out half of the packed sample, and then pipette 4 μL of radical
solution into the rotor. Then replace the other half of the
sample in the rotor, and pipette another 4 μL of radical solution
into the rotor (see Note 12).
8. Weigh the rotor to ensure that there is a mass increase due to
the addition of the radical solution.
9. If there is more than 3 mm of space between the top of the
sample and the top edge of the rotor, add a small amount of
Teflon as spacing material.
10. Close the rotor with the cap carefully, taking care not to
damage the vanes on the cap (see Note 13).
11. Check the bottom of the rotor for the optical mark (a black
line along half the circumference) by which the MAS rate is
tracked, ensuring that it is continuous and clear. If not, rein-
force it with a black/dark felt-tip pen.
3.3 Inserting Rotor 1. Set the temperature of the NMR probe head to the desired
into Magnet and Magic value (see Note 14).
Angle Spinning 2. Press “insert” on the MAS control panel.
3.3.1 Insertion 3. Insert the sample by placing the rotor into the sample transfer
on a Probe Head Without line or directly into the opening of the stator within the
the Option of DNP probe head.
Enhancement
3.3.2 Insertion 1. Cool the DNP NMR probe head down to 100 K using the
on a Probe Head MAS and variable temperature gas flow.
with the Option of DNP 2. Press “eject” on the MAS control panel (see Note 15).
Enhancement
3. Place the rotor in the sample holder of the sample transfer line.
4. Press “insert” on the MAS control panel (see Note 16).
68 Wing Ying Chow
3.3.3 Magic Angle 1. Stop the insertion gas flow.

Spinning 2. Spin the sample to 5 Hz MAS (see Note 17).
3. Once the spinning has stabilized, increase MAS rate to the
desired rate (see Note 18).
3.4 Solid-State NMR Here we present three experiments that are particularly useful for
Experiments studying collagen hydroxylation. We assume that the solid-state
NMR spectrometer had already been used previously for other
13
C/15N experiments and therefore does not cover basic spectrom-
eter setup such as setting the magic angle and initial calibration of
the RF pulse powers, and we assume that a functional pulse
sequence has been installed. For each experiment, some represen-
tative spectra are included to aid interpretation and analysis. Please
check with the local facility manager to ensure that maximum
power levels of the NMR spectrometer are not exceeded.
3.4.1 1D Cross This experiment yields a 1D 13C or 15N NMR spectrum and is the
Polarization (CP) basis of many other ssNMR experiments. Below, the 13C/15N
nucleus is denoted with X (see Note 19).
1. Create a new dataset from a previous CP experiment.
2. Optimize the following parameters for maximum signal-to-
noise ratio (see Note 20):
(a) 1
H 90 pulse length
1
(b) H contact pulse power
1
(c) H-X contact pulse length
1
(d) H decoupling sequence pulse power during signal
acquisition
3. Signal average the optimized CP experiment for at least
4–10 min. Usually this would equate to a minimum of
128 transients.
4. For measuring DNP signal enhancement, the same CP experi-
ment (with same number of transients and parameters) should
be run twice: once with the microwave irradiation switched off
and once again with the microwave irradiation switched
on. The relative ratio of intensity of the two 1D CP spectra is
commonly referred to as the enhancement value (ε). An accept-
able value of ε is 10—many experiments are already much more
feasible at this level of enhancement. On collagen samples, we
have thus far obtained values of ε up to 57.
5. For all other DNP NMR experiments, the microwave irradia-
tion should be switched on prior to the start and kept at the
same level throughout the experiment.
If the referenced spectrum contains a 13C signal at 70 ppm, it
strongly indicates that the sample contains hydroxylated carbon
Fig. 2 1D 13C NMR spectra of adult human bone (natural abundance), 297 K,
400 MHz spectrometer, 10 Hz MAS. Residues that feature prominently in
collagen-containing samples are labelled in bold text
atoms. By eliminating the possibility of other alcohol or glycerol-

like substances, and confirming the presence of other distinctive
protein and collagen signals, it is possible to identify collagen
hydroxylation within the sample using this experiment. If present
in unlabelled or generally isotopically labelled samples, the 70 ppm
signal is likely to be dominated by hydroxyproline, unless lysine is
specifically 13C labelled. Figure 2a shows a representative 1D 13C
ssNMR spectrum, with each signal assigned to the specific types of
carbon atoms in specific residues. Figure 2b shows the effect of
DNP enhancement by comparing the microwave on/off 1D 13C
ssNMR spectra.
3.4.2 2D 13C–13C Proton-driven spin diffusion (PDSD) and dipolar-assisted rotary

Correlation resonance (DARR) are two very similar experiments that can yield
informative 2D (usually 13C–13C) spectra and are fairly simple to
set up on isotope-enriched samples.
1. Create a new dataset from a previous PDSD/DARR experi-
ment, ensuring that it is a 2D dataset with two 13C dimensions.
2. Transfer the optimized CP parameters to the new dataset.
3. Optimize the 13C 90 pulse length, generally achieved by
looking for a zero crossing of the signal when a variable length
carbon pulse is applied after the CP step prior to detection.
4. Select the mixing time and pulse power (see Note 21).
5. Check that the indirect dimension sweep width is sufficiently
wide (for 13C, around 250 ppm) to contain the whole 13C
chemical shift range, but not so excessively wide that it leads
to an unnecessarily long experiment.
70 Wing Ying Chow
Fig. 3 2D 13C–13C DARR spectra of (a) lysine 13C-labelled mouse skin, no DNP enhancement, 285 K, 20 ms
mixing time, 600 MHz spectrometer, 13,333 Hz MAS; (b) lysine 13C-labelled mouse skin, 110 K, with DNP
enhancement (AMUPol, 30 mA microwave current, ε ¼ 27), 5 ms mixing time, 400 MHz spectrometer,
8889 Hz MAS; (c) proline and glycine 13C, 15N-labelled fetal sheep osteoblast ECM, 110 K, no DNP
enhancement, 20 ms mixing time, 400 MHz spectrometer, 8889 Hz MAS. Annotations indicate signals for
unmodified lysine (orange/light gray) and proline (light gray/light gray) residues, and bold-type labels indicate
signals for hydroxylysine (dark brown/dark gray) and hydroxyproline (dark green/dark gray). We can see from
(b) and (c) that hydroxylysine and hydroxyproline signals do not overlap with each other; thus, we can use the
2D spectral pattern to distinguish between these two hydroxylated residues
6. Signal average the PDSD/DARR experiment for 5–20 h.

Using this experiment, it is possible to distinguish the signals
arising from hydroxyproline and the hydroxylysine in triple-helical
collagen. The additional dimension provides the resolution
required to separate the signals that normally overlap in a 1D
experiment. Figure 3 shows two DARR experiments on lysine
13
C-labelled samples acquired without and with DNP enhance-
ment, highlighting the hydroxylysine signals that only appear in
the latter. A third DARR spectrum labelled with proline but not
lysine illustrates how the crosspeak positions of hydroxyproline are
distinct from those of hydroxylysine.
3.4.3 15N–13C This experiment is also sometimes called a double CP experiment,

Heteronuclear Correlation as the magnetization is first transferred by CP from 1H to 15N, then
transferred a second time, again by CP, from 15N to 13C. This
experiment requires 15N and 13C labelling and ideally a relatively
good DNP enhancement in order to obtain sufficient sensitivity.
DNP enhancement is optional for observing hydroxyprolines using
this experiment but necessary if hydroxylysines are of interest.
It is possible to transfer the magnetization further, thus overall
from H ! N ! C ! C, with the third step mediated by DARR/
PDSD. This transfer scheme enables correlations between carbon
atoms that are not directly bonded to nitrogen to be detected.
Fig. 4 2D 15N–13C spectra of lysine 13C, 15N-labelled adult bovine vascular smooth muscle cell ECM, 110 K,
with DNP enhancement (AMUPol, 30 mA microwave current, ε ¼ 57), 10 ms DARR mixing time, 400 MHz
spectrometer, 8889 Hz MAS. The signals specific to hydroxylysine (a, b; in red/dark gray) and hydroxylysine
(c, d; in teal/light gray) are indicated on the figure. Arrows on the chemical structure indicate the magnetiza-
tion transfer pathways, omitting the initial 1H to 15N CP step, but starting on 15N and ending on 13C, that give
rise to the signals observed. Due to the low number of hydroxylysine residues in the ECM relative to lysine
residues, the sidechain signals from the hydroxylysine pathway (a) are much weaker than those from the
lysine pathway (c)
1. Create a new dataset from a previous 15N–13C experiment,

ensuring that it is a 2D dataset with a directly detected 13C
dimension and an indirect 15N dimension.
2. Transfer the optimized CP and DARR parameters to the new
dataset.
3. Calculate the NC(A) contact matching condition. A reasonable
set of initial values are:
13
(a) C pulse power: 1.5 MAS rate with a tangential shape
(1.4 to compensate for power loss during the shaped
pulse).
15
(b) N pulse power: 2.5 MAS rate with a square shape
(1).
1
(c) H pulse power: use the same power as the 1H decoupling
power during acquisition.
(d) NC contact pulse length: 5 ms.
72 Wing Ying Chow
4. Optimize (a), (c), and (d) on a 1D version of the NC/double

CP experiment for maximum signal-to-noise ratio (see Note
22).
5. Check that the indirect acquisition sweep width is sufficiently
wide (for 15N, around 200 ppm) and centered correctly to
contain the entire 15N chemical shift range, including the
amide signals up to 120 ppm and the lysine sidechain amine
signal at around 40 ppm.
6. Signal average the NC experiment for 5–20 h.
Using this experiment, it is possible to identify signals arising
from lysine and hydroxylysine due to the distinctive sidechain
amine 15N chemical shift. As the amine sidechain and the hydro-
xylated carbon are not directly bonded, the additional 13C–13C
PDSD/DARR step is required in order to generate a spectrum
with a Cδ-Nζ correlation that is unique to hydroxylysine. Figure 4
shows a 15N–13C NMR spectrum on 13C, 15N-lysine-labelled vas-
cular smooth muscle cell ECM.
4 Notes
1. For tissue samples, rotor diameters of 3.2 mm, 4.0 mm, and
7.0 mm have all been successfully used to achieve stable
spinning. Smaller rotor diameters such as 2.5 mm and
1.9 mm may also be used, but more care is needed to balance
the sample.
2. From our experience, the commercial DNP-ssNMR 3.2 mm
diameter MAS probe head produced by Bruker requires cool-
ing with approximately 250 L of liquid nitrogen every 24 h
when running at a set temperature of 100 K.
3. In the cases where the sample mass is limited, restricted depth
rotors should be used to center the sample in the center of the
rotor to improve RF pulse homogeneity and experiment sensi-
tivity. Inert spacers made of materials such as Teflon can be
used for restricting the sample height to the center of the rotor.
Materials low in proton and carbon (where possible) are pre-
ferred, as otherwise the spacer material could be observed in
the NMR spectra; care should also be taken to avoid materials
that show large volume change with temperature variations.
4. For experiments at the low temperature conditions for DNP
enhancement (<200 K), zirconia caps are recommended,
though Vespel caps have also been successfully used. It is
recommended to test shrinkage of the rotor cap by placing
the packed rotor into liquid nitrogen and observing whether
there is any movement or slippage of the cap before inserting
the rotor into the probe head. Use rotors with regular wall
thickness (1 mm) when possible. Some rotors with thin walls

(0.5 mm) appear to have looser rotor caps and are more prone
to cap slippage.
5. The amount of sample required depends on the proportion of
sites of interest within the sample and the degree of isotope
enrichment. Due to the sensitivity challenges of NMR, it is
advantageous to fill the volume of the rotor with as much
sample as possible while maintaining stable magic angle
spinning.
6. The isotope enrichment method or labelling scheme depends
on the sites of interest. In general, it is better to enrich only the
amino acids that would be of interest, to avoid irrelevant signals
overlapping with the signals of interest and loss of resolution in
the NMR spectra. For example, it is ideal to only 13C-enrich
lysine residues by adding 13C lysine to the cell culture media or
to the diet if the sites of interest are hydroxylysines. With DNP
enhancement, it is possible to detect hydroxylysine sites in
tissues with 45% 13C lysine enrichment.
7. While more protein sample can be filled into the rotor after
freeze-drying, the procedure can lead to structural damage that
manifests as a loss of resolution in the NMR spectra. Whether
to sacrifice the resolution for more sensitivity has to be deter-
mined on a case-by-case basis. For hydrated samples, excess
water can be removed in a more gentle manner by placing the
sample between two layers of Parafilm and gently pressing. The
water expelled can be removed with a pipette.
8. We have found that cryomilling led to significant loss in reso-
lution in mouse bone 13C spectra, likely due to dehydration.
Tissue obtained from small animals may be particularly suscep-
tible. On the other hand, milling may be necessary for miner-
alized tissues that are significantly larger than the rotor in size.
The bone obtained from larger mammals, such as horses and
humans, appears to be only minimally affected by cryomilling.
9. If fixatives or preservatives were used, the tissue sample should
be soaked for 24 h or more in a significant excess of water to
remove any remaining fixative and rinsed prior to use.
10. Ensure that the container containing the stable radical is at
room temperature before opening, to avoid buildup of con-
densation. As a rather small amount of radical mass (under
0.5 mg) is usually required to be weighed out, we often use
the tip of a needle for transferring the solid radical from the
stock container and carry out the weighing process in the lid of
a small Eppendorf tube, which can then be used directly for
mixing the radical solution. The concentrations recommended
for AMUPol and bcTol-M are different due to the higher
solubility of bcTol-M in water.
74 Wing Ying Chow
11. The amount of radical solution has to match the amount of

sample in the rotor. Excessive radical concentration will bleach
the NMR signal, leading to poor or no enhancement. For DNP
enhancement of biomolecules, a concentration of 10 mM is
generally recommended. However, it can be difficult to esti-
mate the volume of a solid sample; hence, the procedure here is
based on sample mass. If the sample is unexpectedly bulky, such
that the rotor volume is full with less than 15 mg of sample, use
the procedure in the next step.
12. The aim here is to achieve a final concentration of approxi-
mately 10 mM of biradical in the rotor. Assuming that the
sample completely fills the volume of the 3.2 mm diameter
rotor (30 μL), adding 8 μL 40 mM bcTol-M stock solution
results in a final concentration of 10.7 mM. If the sample mass
is more than 12 mg but less than 20 mg, reduce the second
addition of the radical solution proportionally.
13. If radical solution was added to the rotor, use a tabletop
spinner to further distribute the radical solution through the
packed sample with several inversions of the rotor. This can
help to maximize the wetting of the sample and optimize the
DNP enhancement. Check the cap carefully for any signs of
damage, and also check to make sure that the rotor cap is flush
against the top of the rotor, without any gaps, as otherwise
MAS would fail.
14. Generally, a temperature between 0 and 20 C is selected. Due
to air friction during MAS, there can be a difference between
the reported and the actual temperatures. For up to 14 Hz
MAS, the difference should not be more than 10 C.
15. Due to the low operating temperature of the DNP probe head,
it is important to avoid moisture ingress. Therefore, when the
probe head is cold, the eject gas flow must always be enabled
prior to opening the sample transfer line.
16. It is recommended to monitor this process via one of two ways:
Either set the bearing gas flow to 200–300 mBar, and look for a
decrease in the bearing gas flow on insertion of the sample, or
display the frequency sweep (Bruker Topspin spectrometer
software “wobb” command) and look for a change in the
absorption minimum frequency on insertion of the sample.
The tendency of the DNP NMR probe to condense moisture
from the air leads to a higher chance of the rotor being stuck in
the transfer line. In case there is a problem with insertion, test
whether the rotor spins; if not, try to eject the rotor; if the
rotor fails to eject, warm up the probe head before attempting
ejection again. It is possible as well to start with a warm probe,
and then carry out one of the two following variations: Spin the
sample to the desired MAS rate, and then cool the probe down
to 100 K; or, spin the sample to a moderate rate (e.g., 5 Hz),

cool the sample to 100 K, and then increase the MAS rate
further. In both of these variations, it is important to increase
the cooling gas slowly, as the cooling gas introduces turbulence
and destabilizes MAS. If the temperature is changed too sud-
denly, there is a risk of a rotor crash.
17. If the rotor fails to spin faster than 900–1000 Hz at this stage,
it indicates a mass imbalance within the rotor. Sometimes this
resolves itself in a few minutes, but at other times, the rotor has
to be ejected and the sample repacked or redistributed. If a
larger gap has appeared between the top of the sample and the
top edge of the rotor, consider adding a small amount of Teflon
tape as a spacer.
18. Some suggested MAS rates: For a 400 MHz spectrometer,
8889 Hz MAS; for a 600 MHz spectrometer, 13,333 Hz
MAS; and for a 700 or an 800 MHz spectrometer,
11,000 Hz MAS. At these rates, the spinning sidebands of
the amide signals do not tend to overlap with the signals arising
from hydroxylation.
19. CP is usually carried out by transferring magnetization from
1
H to 13C or 1H to 15N and detecting on the X nuclei. This
technique has the advantage of increasing the sensitivity of the
nuclei being observed, while at the same time increasing sensi-
tivity per unit time as the relaxation behavior follows that of the
more abundant (usually 1H) spins and is of the order of several
hundred milliseconds, rather than several seconds or more in
the case of 13C, thus enabling more transients to be obtained in
a shorter time. The CP experiment is quite forgiving, and it is
possible to obtain a good 1D spectrum on samples containing
collagen even without isotope labelling, provided that a full
rotor is used and signal averaging is carried out for 4–12 h (for
13
C) and 12–24 h (for 15N). These signal averaging times can
be drastically reduced by DNP enhancement, potentially down
to 10–30 min. The CP experiment relies on the Hartmann-
Hahn matching condition [19] during the contact time, when
pulses are simultaneously applied on both channels. We use a
ramp on the 13C pulse during the contact pulse to broaden the
matching condition [20].
20. For an optimized CP sequence, here are some characteristic
parameters obtained from our spectrometers. We do not report
the powers as these values are not transferable between differ-
ent spectrometers. 1H 90 pulse length: 2.5–4 μs and 1H-X
contact time: 750–1750 μs.
21. Both the PDSD and DARR experiments correlate pairs of 13C
spins that are within a certain distance of each other. As part of
the experiment, we can control a time duration (the “mixing
76 Wing Ying Chow
time) where there is either no pulses at all (PDSD) [21] or a

low power constant amplitude pulse on the 1H channel
(DARR) [22]. In both of these cases, when the mixing time
is short (5–20 ms), only pairs of 13C that are within 2 Å would
show crosspeaks. However, when the mixing time is long
(100þ ms), 13C–13C pairs within 6 Å would show crosspeaks.
Due to the low temperature of the DNP enhancement condi-
tions, crosspeaks tend to appear at shorter mixing times.
DARR is believed to lead to a quicker buildup than PDSD,
though the differences are relatively minor at a 400 MHz
spectrometer.
22. Compare the 1D double CP experiment signal intensity with a
normal 13C CP acquired with the same number of transients.
Due to the extra transfer step, the intensity of the double CP
experiment should be around 5–40% of that obtained from a
CP. If the double CP intensity is more than 20% of that
obtained from a CP, that is a reasonably optimized experiment
for collagen-containing samples, and the parameters should
yield a good 2D spectrum. If the double CP intensity is less
than 5%, then the parameters should be further optimized to
try to increase the transfer efficiency.
Acknowledgments
The author thanks Dr. Jonathan Clark from the Babraham Institute
for providing the lysine 13C-labelled mouse skin sample,
Mr. Rakesh Rajan and Professor Melinda Duer from the University
of Cambridge for providing the proline and glycine 13C, 15N-
labelled fetal sheep osteoblast ECM sample, Mr. Robert Hayward
and Professor Cathy Shanahan at King’s College London for
providing the lysine 13C, 15N-labelled adult bovine vascular smooth
muscle cell ECM sample, and Dr. Kelsey Collier at FMP Berlin for
useful discussions and comments.
References
1. Shoulders MD, Raines RT (2009) Collagen 4. Cavanagh J, Fairbrother WJ, Palmer AG, Skel-
structure and stability. Annu Rev Biochem ton NJ, Rance M (2010) Protein NMR spec-
78:929–958. https://doi.org/10.1146/ troscopy: principles and practice. Elsevier,
annurev.biochem.77.032207.120833 Amsterdam
2. Yamauchi M, Sricholpech M (2012) Lysine 5. Keeler J (2011) Understanding NMR spectros-
post-translational modifications of collagen. copy. John Wiley & Sons, New Jersey
Essays Biochem 52:113–133. https://doi. 6. Harris RK, Becker ED, De Menezes SMC,
org/10.1042/bse0520113 Granger P, Hoffman RE, Zilm KW (2008)
3. Levitt MH (2013) Spin dynamics: basics of International Union of Pure and Applied
nuclear magnetic resonance. John Wiley & Chemistry physical and biophysical chemistry
Sons, New Jersey division (2008) further conventions for NMR
shielding and chemical shifts IUPAC
recommendations. Magn Reson Chem https://doi.org/10.1016/j.pnmrs.2017.06.

46:582–598. https://doi.org/10.1002/mrc. 002
2225 15. Itin B, Sergeyev IV (2018) Strategies for effi-
7. Wishart DS, Bigam CG, Yao J, Abildgaard F, cient sample preparation for dynamic nuclear
Dyson HJ, Oldfield E, Markley JL, Sykes BD polarization solid-state NMR of biological
(1995) 1H, 13C and 15N chemical shift refer- macromolecules. In: Ghose R (eds) Protein
encing in biomolecular NMR. J Biomol NMR NMR. Methods in Molecular Biology, vol
6:135–140 1688. https://doi.org/10.1007/978-1-
8. Allevi P, Paroni R, Ragusa A, Anastasia M 4939-7386-6_7
(2004) Hydroxylysine containing glycoconju- 16. Lesage A, Lelli M, Gajan D, Caporini M A,
gates: an efficient synthesis of natural galacto- Vitzthum V, Miéville P, Alauzun J, Roussey A,
sylhydroxylysine (gal-Hyl) and Thieuleux C, Mehdi A, Bodenhausen G,
glucosylgalactosylhydroxylysine (Glu-gal-Hyl) Copéret C, Emsley L (2010) Surface enhanced
and of their (5S)-epimers. Tetrahedron Asym- NMR spectroscopy by dynamic nuclear polari-
metry 15:3139–3148. https://doi.org/10. zation. J Am Chem Soc 132:15459–61.
1016/j.tetasy.2004.08.006 https://doi.org/10.1021/jacs.5b08423
9. LeMaster DM (1994) Isotope labeling in solu- 17. Geiger M-A, Jagtap AP, Kaushik M, Sun H,
tion protein assignment and structural analysis. Stöppler D, Sigurdsson ST, Corzilius B, Osch-
Prog Nucl Magn Reson Spectrosc kinat H (2018) Efficiency of water-soluble
26:371–419. https://doi.org/10.1016/ Nitroxide Biradicals for dynamic nuclear polar-
0079-6565(94)80010-3 ization in rotating solids at 9.4 T: bcTol-M and
10. An B, Kaplan DL, Brodsky B (2014) Engi- cyolyl-TOTAPOL as new polarizing agents.
neered recombinant bacterial collagen as an Chemistry 24(51):13485–13494. https://doi.
alternative collagen-based biomaterial for tis- org/10.1002/chem.201801251
sue engineering. Front Chem 2:40. https:// 18. Sauvée C, Rosay M, Casano G, Aussenac F,
doi.org/10.3389/fchem.2014.00040 Weber RT, Ouari O, Tordo P (2013) Highly
11. Chow WY, Rajan R, Muller KH, Reid DG, efficient, water-soluble polarizing agents for
Skepper JN, Wong WC, Brooks RA, dynamic nuclear polarization at high frequency.
Green M, Bihan D, Farndale RW, Slatter DA, Angew Chem Int Ed Engl 125:11058–11061.
Shanahan CM, Duer MJ (2014) NMR spec- https://doi.org/10.1002/ange.201304657
troscopy of native and in vitro tissues implicates 19. Hartmann SR, Hahn EL (1962) Nuclear dou-
polyADP ribose in biomineralization. Science ble resonance in the rotating frame. Phys Rev
344:742–746. https://doi.org/10.1126/sci 128:2042–2053. https://doi.org/10.1103/
ence.1248167 physrev.128.2042
12. Wong VWC, Reid DG, Chow WY, Rajan R, 20. Metz G, Wu XL, Smith SO (1994) Ramped-
Green M, Brooks RA, Duer MJ (2015) Prepa- amplitude cross polarization in magic-angle-
ration of highly and generally enriched mam- spinning NMR. J Magn Reson A
malian tissues for solid state NMR. J Biomol 110:219–227. https://doi.org/10.1006/
NMR 63:119–123. https://doi.org/10. jmra.1994.1208
1007/s10858-015-9977-9 21. Szeverenyi NM, Sullivan MJ, Maciel GE
13. Ong S-E, Blagoev B, Kratchmarova I, Kristen- (1982) Observation of spin exchange by
sen DB, Steen H, Pandey A, Mann M (2002) two-dimensional fourier transform 13C cross
Stable isotope labeling by amino acids in cell polarization-magic-angle spinning. J Magn
culture, SILAC, as a simple and accurate Reson 47:462–475. https://doi.org/10.
approach to expression proteomics. Mol Cell 1016/0022-2364(82)90213-x
Proteomics 1:376–386 22. Takegoshi K, Nakamura S, Terao T (2001)
13
14. Lilly Thankamony AS, Wittmann JJ, C–1H dipolar-assisted rotational resonance
Kaushik M, Corzilius B (2017) Dynamic in magic-angle spinning NMR. Chem Phys
nuclear polarization for sensitivity enhance- Lett 344:631–637. https://doi.org/10.
ment in modern solid-state NMR. Prog Nucl 1016/s0009-2614(01)00791-6
Magn Reson Spectrosc 102-103:120–195.
Chapter 6
Measurement of Collagen Cross-Links from Tissue

Samples by Mass Spectrometry
Amar Joshi, Amna Zahoor, and Alberto Buson
Abstract
All tissues contain an extracellular matrix (ECM) which is constantly and dynamically remodeled, either in
physiological or pathological processes, such as fibrosis or cancer. One of the key contributors in the
establishment of a fibrotic state is the abnormal deposition of extracellular matrix and cross-linked proteins,
in particular collagen, leading to tissue stiffening and disruption of organ function. The precise and sensitive
measurement of these cross-links by LC-MS/MS is a very powerful tool for providing a quantitative and
qualitative analysis of fibrosis and is a key requirement in the study of this state, as well as in the development
of drugs for this unmet clinical need.
Key words Collagen, Tissue cross-linking, Immature cross-links, Mature cross-links, Solid phase
extraction (SPE), Mass spectrometry
1 Introduction
One of the most relevant hallmarks of a fibrotic state in a tissue is

the accumulation of heavily cross-linked collagen fibers, primarily
due to the activity of enzymes belonging to the family of the
lysyl oxidases, i.e., LOX and lysyl oxidase like 1 through 4 -
(LOXL1–LOXL4). The expression of these enzymes is often upre-
gulated in course of diseases such as cancer and fibrotic conditions,
resulting in the oxidative deamination of lysine and hydroxylysine
residues in collagen precursor proteins and lysine residues in elastin.
These residues are modified to reactive aldehydes, which spontane-
ously further react in various combinations, giving rise to a variety
of lysine- and hydroxylysine-derived cross-links [1]. The hydroxy-
lysine residues, in particular, are product of the posttranslational
modification of specific lysine residues by members of the lysine
hydroxylase family of enzymes and bear a particular importance in
terms of stability and mechanical properties of the collagen in the
tissue. These modifications are of significant biological importance,
since dysregulations of the system lead to disruption of the
79
80 Amar Joshi et al.
physiological homeostasis, with pathological consequences that

range from the various syndromes linked to insufficient hydroxyl-
ation to the exacerbation of the fibrotic state in case of excessive
hydroxylation. The direct quantification of the cross-links gener-
ated in the fibrotic process is a key complement to the quantifica-
tion and localization of total collagen, providing additional
information and a more thorough evaluation of the degree of
fibrosis of a given tissue. Furthermore, the collagen cross-links
dynamically shift from immature to mature, in parallel with the
maturation and the mechanical properties of the collagen fibers.
The relative content of the immature over the mature cross-links
thus provides additional qualitative information on the trajectory
and reversibility of the fibrotic state [2]. In the type of measure-
ment, we do in our lab, the tissue undergoes homogenization and
reduction steps to obtain a full release of the cross-links, and before
the analysis by UHPLC-MS/MS, the samples are purified by a solid
phase extraction step. The method can be applied to most tissues,
e.g., liver, lung, heart, kidney, skin, and others.
2 Materials
Prepare all solutions using ultrapure water (prepared by purifying

deionized water, to attain a sensitivity of 18.2 MΩ-cm at 25 C) and
analytical grade reagents. All reagents should be prepared and
stored at room temperature (unless otherwise specified). All waste
should be disposed according to proper waste regulations.
2.1 Tissue Sample 1. Freeze dryer.

Preparation 2. Weighing scales.
2.1.1 Tissue Preparation 3. 5 mL polystyrene tubes with cell-strainer caps.
4. 2 mL metal bead lysing matrix tubes.
5. Metal spatulas.
6. Benchtop tissue homogenizer.
7. Phosphate buffered saline (PBS).
8. 2 mL acid-/heat-resistant screw cap tubes.
9. Genevac/evaporator.
Tissue Preparation (Skin) 1. 10 mg/mL Collagenase/Dispase: sourced commercially,

source is V. alginolyticus/B. polymyxa: prepare 10 mg/mL
collagenase solution by dissolving 100 mg of collagenase in
10 mL of PBS. Invert gently to mix.
Measurement of Collagen Cross-Linking by Mass Spectrometry 81
2.1.2 Reduction Step 1. 12 mg/mL sodium borohydride (NaBH4) buffer: the buffer is
prepared in 20 mM sodium hydroxide (NaOH) solution. The
NaOH stock solution can be prepared at 2 M in water and then
diluted to 20 mM with water for NaBH4 buffer preparation (see
Note 1).
2. Neat acetic acid: >99.7% purity.
3. Benchtop centrifuge.
4. Benchtop heat block.
5. 6 M hydrochloric acid (HCL): 12 M HCL sourced commer-
cially is diluted 1:2 to obtain 6 M HCL (see Note 2).
2.1.3 Sample Cleanup 1. Automated SPE machine (e.g., GX-271 ASPEC™ Liquid
Using Solid Phase Handler).
Extraction (SPE) 2. Benchtop quick spin.
3. Benchtop shaker.
4. Capless graduated microtubes (1.7 mL).
5. C18 columns (e.g., GracePure™ C18).
6. Solid phase extraction (SPE) strong cation exchange (SCX)
columns (e.g., GracePure™ SPE).
7. 5 mL round bottom tubes.
8. Solution 1: 10 mM ammonium formate, 0.1% formic acid,
0.1% heptafluorobutyric acid (HFBA) in water. For a 500 mL
solution preparation, weigh 315.3 mg of ammonium formate
and transfer to a 500 mL glass bottle. Add 500 μL formic acid,
500 μL HFBA and make up to 500 mL with water. Mix (see
Note 2).
9. Solution 2: 10 mM ammonium formate, 0.1% formic acid,
0.1% HFBA, 40% methanol in water. For a 500 mL solution
preparation, weigh 315.3 mg ammonium formate and transfer
to a 500 mL Schott glass bottle. Add 500 μL formic acid,
500 μL HFBA, 200 mL methanol and make up to 500 mL
with water. Mix (see Note 2).
10. Solution 3: 25% ammonium hydroxide solution in water (see
Note 2).
11. Solution 4: 0.1% formic acid in water. For a 500 mL solution
preparation, add 500 μL formic acid and make up to 500 mL
with water. Mix (see Note 2).
12. HFBA solution: 5% HFBA in water (see Note 2).
13. SnapStrip 8-strip PCR tubes.
14. High-performance liquid chromatography (HPLC) screw vials
(with fixed inserts) and lids.
2.1.4 Sample Total protein assay kit.

Normalization
2.2 Ultrahigh- 1. Ultrahigh-performance liquid chromatographic (UHPLC)

Performance Liquid (e.g., Thermo Scientific Dionex ultimate 3000).
Chromatography- 2. UHPLC column (e.g., Agilent SB C18 RRHD, 50 2.1 mm,
Triple Quadrupole 1.8 μm reversed phase column).
Mass Spectrometer 3. Triple quadrupole mass spectrometer.
(UHPLC-MS/MS)
4. Mobile phase A: 10 mM ammonium formate, 0.1% formic acid,
Method
0.12% heptafluorobutyric acid (HFBA) in water. Weigh
630.6 mg of ammonium formate and transfer to the 1 L glass
bottle. Add 900 mL of water; add 1 mL of formic acid and
1.2 mL of HFBA. Make up to 1 L with water and mix. Protect
the prepared solution from light by preparing it in amber
colored bottle, or cover the normal glass bottle with aluminum
foil (see Notes 2 and 3).
5. Mobile phase B: 10 mM ammonium formate, 0.1% formic acid,
0.12% HFBA in LCMS grade methanol. Weigh 315.3 mg of
ammonium formate and transfer to the 500 mL glass bottle.
Add 400 mL of LCMS grade methanol; add 0.5 mL of formic
acid and 0.6 mL of HFBA. Make up to 500 mL with LCMS
grade methanol and mix well. Protect the prepared solution
from light by preparing it in amber colored bottle, or cover the
normal glass bottle with aluminum foil (see Notes 2 and 3).
6. Autosampler needle wash solution: 20% methanol in water.
Take 800 mL of water in a 1 L glass bottle and add 200 mL
of LCMS grade methanol. Mix well.
7. Column wash solution A: 0.1% formic acid in water. Take
900 mL of water in a 1 L glass bottle; add 1 mL of formic
acid and make up to 1 L with water. Mix well.
8. Column wash solution B: 0.1% formic acid in LCMS grade
acetonitrile. Take 900 mL of LCMS grade acetonitrile in a 1 L
glass bottle; add 1 mL of formic acid and make up to 1 L with
LCMS grade acetonitrile. Mix well.
9. The working standard were dihydroxylysinonorleucine
(DHLNL), hydroxylysinonorleucine (HLNL), pyridinoline
(PYD) and deoxypyridinoline (DPD) and hydroxyproline.
For DHLNL, dissolve 2.5 mg of the working standard in
1 mL of water to obtain a concentration of 2.5 mg/mL,
which is equivalent to 8.14 nmol/μL. For HLNL, dissolve
0.5 mg of the working standard in 172 μL water to obtain a
concentration of 2.9 mg/mL, which is equivalent to 10 nmol/
μL. For hydroxyproline, dissolve 13.1 mg of working standard
in 1 mL of water to obtain a concentration of 13.1 mg/mL,
which is equivalent to 100 nmol/μL. PYD and DPD are
available as a solution mixture where the concentration

labeled on the vial for PYD is 5.88 μg/mL and for DPD is
3.06 μg/mL.
10. Graph Pad Prism software.
3 Methods
Carry out all procedures at room temperature unless otherwise

specified.
3.1 Extraction For immature and mature collagen cross-links and hydroxyproline.
and Purification
of Cross-Links
3.1.1 Tissue Sample Note that this tissue preparation is generic and applies for most soft
Preparation tissues, except skin (see Subheading “Skin Preparation”).
1. Weigh out relevant amount of tissue and carry out freeze-
drying. Cut the tissue into smaller pieces to increase the surface
area for freeze-drying (see Note 4).
2. Grind the freeze-dried tissue manually using a spatula and
then collect 20 mg in a 2 mL metal bead lysing matrix tube
(see Note 5).
3. Homogenize the dry tissue twice. The sample becomes a dry
powder.
4. Add 1 mL of PBS and homogenize two more times (see
Note 6).
5. Transfer 500 μL of the homogenate (equivalent to 10 mg dry
tissue) into a 2 mL heat/acid resistant screw cap tube (see
Note 7). Tissue preparation is complete for mature collagen
and mature elastin cross-links samples.
6. For immature collagen cross-links extraction and hydroxypro-
line extraction samples, top up with 370 μL of PBS, to gain a
total volume of 870 μL.
7. Reduction step (see Subheading 3.1.2).
Skin Preparation 1. Weigh out each sample and insert into a 2 mL metal bead lysing
matrix tube (see Note 8).
2. For homogenization of skin samples, add 1 mL of collagenase
solution per 10 mg of tissue.
3. Incubate for 24 h at room temperature. During the incuba-
tion, the samples should be homogenized intermittently (see
Note 9).
4. Transfer 500 μL of the homogenate (equivalent to 5 mg dry

tissue) into a 2 mL acid-/heat-resistant screw cap tube (see
Note 7).
5. Top up with 370 μL PBS, to gain a total volume of 870 μL.
6. Reduction step (see Subheading 3.1.2).
3.1.2 Reduction Step For immature collagen cross-links extraction and hydroxyproline.
For mature collagen and elastin cross-links, go straight to step 5.
1. Add 30 μL of 12 mg/mL NaBH4 buffer into the 870 μL tissue
homogenate (equivalent to a 30 dilution of the buffer). After
addition, shake the tubes manually by hand to ensure the
reduction reaction can occur.
2. Incubate at room temperature for 30 min.
3. Stop the reaction by adding 50 μL of neat acetic acid, which
will bring the pH below 2.5 (see Note 10).
4. Quick spin the samples to remove liquid contents from the lid.
5. Dry the samples using the Genevac at 60 C with the aqueous
program settings.
6. Add 1600 μL of 6 M HCL to the pellet. Heat at 105 C to
hydrolyse for 24 h (see Note 11).
7. Dry the HCL using the Genevac at 60 C with the H2O + NH3
program settings (see Note 12).
8. Reconstitute the samples in 200 μL of water, and place on
shaker for 10 min at high speed.
9. Spin dow the samples at 20,000 g for 1 min.
3.1.3 Sample Cleanup 1. Prepare C18 and SCX columns by inserting adaptors onto the
Using SPE (Automated) columns (see Note 13).
2. Take 50 μL of the sample hydrolysate (i.e., from step 8 of the
reduction procedure) into 400 μL of solution 1.
3. Load onto a SPE C18 column.
4. Perform steps 5–9 in a fume hood (see Note 2).
5. Add 600 μL of solution 1 to the C18 column and elute into a
C18-labeled 5 mL round bottom tube.
6. Add 500 μL of solution 2 to the C18 column and elute into the
abovementioned 5 mL round bottom tube.
7. Add the C18 eluate (1550 μL total) directly onto a SCX
column (which has been pre-rinsed with 2 mL of water and
500 μL of solution 4).
8. Wash the loaded SCX column with 1 mL of water.
9. Elute the SCX cartridge with 3 mL of solution 3 and collect
into a SCX-labeled 5 mL round bottom tube.
10. Dry the SCX eluate using the Genevac at 60 C with the
H2O + NH3 program settings.
11. Add 500 μL of 5% HFBA solution.
12. Dry the sample using the Genevac at 60 C with the aqueous
program settings.
13. Reconstitute the dried samples in 50 μL of solution 1 and
transfer to SnapStrip PCR tubes (see Note 14).
14. Spin the samples using a benchtop spin at high speed for 5 min.
15. Transfer 40 μL of the supernatant into HPLC vials; take extra
care not to disturb the pellet.
3.1.4 Protein Assay Perform a protein assay on the samples as per instructions detailed
in the total protein assay kit (see Note 15).
3.2 Ultrahigh- 1. Flow rate: 0.3 mL/min using a multistep gradient method.
Performance Liquid 2. Gradient method: Hold mobile phase A at 96.2% for 1.0 min
Chromatography- and then decrease linearly to 80% within 5 min. Hold mobile
Triple Quadrupole phase A at 80% from 5 min to 7 min. Switch the divert valve
Mass Spectrometer from ion source to the waste at 7 min. Decrease mobile phase A
(UHPLC-MS/MS) linearly to 0% from 7 min to 9 min. Bring back mobile phase A
concentration from 0% to 96.2% linearly at 9.5 min and hold
3.2.1 Chromatographic
for the next 1.5 min, i.e., until 12.0 min. At the end of the
Conditions
gradient program, for the duration of 0.2 min from 11.8 to
12.0 min, decrease the flow rate from 0.3 mL/min to
0.05 mL/min.
3. Total run time: 12.0 min.
4. Column oven temperature: 40 C.
5. Autosampler temperature: Room temperature.
6. Injection volume: 10 μL for the skin, liver, and kidney tissues.
3.2.2 Mass Spectrometry Quantitation of the immature cross-links dihydroxylysinonorleu-

Conditions cine (DHLNL) and hydroxylysinonorleucine (HLNL), the triva-
lent mature cross-links pyridinoline (PYD) and deoxypyridinoline
(DPD), as well as hydroxyproline (HYP) to determine total colla-
gen in the prepared samples by ultra-performance liquid
chromatography-electrospray ionization tandem mass spectrome-
try (UPLC-ESI-MS/MS).
1. Operate the ESI source as follows: Spray voltage, 4000 V;
sheath gas, 35 psi; auxiliary gas, 20 psi; sweep gas, 0 psi; ion
transfer tube temperature, 350 C; vaporizer temperature,
300 C.
2. The MS acquisition parameters for DHLNL: Radio frequency
(RF) lens voltage, 100 V; collision energy, 23 V. Collect data in
single reaction monitoring (SRM) mode using retention time
window from 3.19 min to 4.0 min and a transition of precursor

ion m/z 308.1 to a product ion m/z 128.246. The confirma-
tion peak transitions have precursor m/z 308.1 to a product
ion m/z 82.136 with a collision energy as 33 V and m/z 308.1
to a product ion m/z 146.109 with a collision energy as 25 V.
3. The MS acquisition parameters for HLNL: Radio frequency
(RF) lens voltage, 100 V; collision energy, 20.921 V. Collect
data in single reaction monitoring (SRM) mode using reten-
tion time window from 4.33 min to 5.0 min and a transition of
precursor ion m/z 292.139 to a product ion m/z 128.097.
The confirmation peak transitions have precursor m/z
292.139 to a product ion m/z 130.183 with a collision energy
as 19.303 V and m/z 292.139 to a product ion m/z 229.111
with a collision energy as 20.466 V.
4. The MS acquisition parameters for PYD: Radio frequency (RF)
lens voltage, 180 V; collision energy, 30 V. Collect data in
single reaction monitoring (SRM) mode using retention time
window from 4.7 min to 5.0 min and a transition of precursor
ion m/z 429.2 to a product ion m/z 412.13. The confirmation
peak transitions have precursor m/z 429.2 to a product ion m/
z 266.98 with a collision energy as 36 V and m/z 429.2 to a
product ion m/z 367.06 with a collision energy as 30 V.
5. The MS acquisition parameters for DPD: Radio frequency
(RF) lens voltage, 140 V; collision energy, 30 V. Collect data
in single reaction monitoring (SRM) mode using retention
time window from 5.8 min to 8.0 min and a transition of
precursor ion m/z 413.2 to a product ion m/z 267.1. The
confirmation peak transitions were having m/z 413.2 to a
product ion m/z 351.12 with a collision energy as 28 V and
m/z 413.2 to a product ion m/z 396.16 with a collision energy
as 30 V.
6. The MS acquisition parameters for HYP: Radio frequency
(RF) lens voltage, 81 V; collision energy, 13.135 V. Collect
data in single reaction monitoring (SRM) mode using reten-
tion time window from 0.85 min to 1.5 min and a transition of
precursor ion m/z 132.2 to a product ion m/z 96.125. The
confirmation peak transitions have precursor m/z 132.2 to a
product ion m/z 68.389 with a collision energy as 22.236 V
and m/z 132.2 to a product ion m/z 78.214 with a collision
energy as 29.618 V.
7. Operate the collision induced gas at 1.5 mTorr.
3.2.3 Calibration Carry out all procedures at room temperature unless otherwise
Standards Preparation specified.
1. Pipette out 650 μL of the PYD/DPD mixture from the vial out
of the total volume of 750 μL and transfer it into a plastic test
tube. The concentration of PYD is 3.82 μg/mL equivalent to

8.9 nmol/mL and DPD is 1.99 μg/mL equivalent to
4.8 nmol/mL.
2. Dry the samples using the Genevac EZ-2 plus at 60 C using
the aqueous program.
3. Reconstitute the dried sample with 2 mL of mobile phase A,
add 446.5 μL of hydroxyproline stock solution equivalent to
100 nmol/μL, then add 2.74 μL of DHLNL stock solution
(8.14 nmol/μL), then add 0.89 μL of HLNL stock solution
(10 nmol/μL) and then finally add 2.015 mL of mobile phase
A. The final concentrations of the biomarkers in the prepared
solution are 100 nmol/10 μL hydroxyproline, 50 pmol/10 μL
DHLNL, 20 pmol/10 μL HLNL, 20 pmol/10 μL PYD, and
10 pmol/10 μL DPD. Mix well. Name the solution as “Cross-
links Calibration solution 1.”
4. Make 2.5 dilutions of Crosslinks Calibration solution 1; take
2060 μL of Cross-links Calibration solution 1 and add into
3090 μL of mobile phase A. The final concentrations of the
biomarkers in the prepared solution are 40 nmol/10 μL
hydroxyproline, 20 pmol/10 μL DHLNL, 8 pmol/10 μL
HLNL, 8 pmol/10 μL PYD, and 4 pmol/10 μL DPD. Mix
well. Name the solution as “Cross-links Calibration solution 2.”
5. Make 2.5 dilutions of Cross-links Calibration solution 2; take
biomarkers in the prepared solution are 16 nmol/10 μL
hydroxyproline, 8 pmol/10 μL DHLNL, 3.2 pmol/10 μL
HLNL, 3.2 pmol/10 μL PYD, and 1.6 pmol/10 μL DPD.
Mix well. Name the solution as “Cross-links Calibration solu-
tion 3.”
6. Make 2.5 dilutions of Crosslinks Calibration solution 3; take
biomarkers in the prepared solution are 6.4 nmol/10 μL
hydroxyproline, 3.2 pmol/10 μL DHLNL, 1.28 pmol/10 μL
HLNL, 1.28 pmol/10 μL PYD, and 0.64 pmol/10 μL DPD.
Mix well. Name the solution as “Cross-links Calibration solu-
tion 4.”
hydroxyproline, 1.28 pmol/10 μL DHLNL, 0.512 pmol/
10 μL HLNL, 0.512 pmol/10 μL PYD, and 0.256 pmol/
10 μL DPD. Mix well. Name the solution as “Cross-links
Calibration solution 5.”

hydroxyproline, 0.512 pmol/10 μL DHLNL, 0.2048 pmol/
10 μL HLNL, 0.2048 pmol/10 μL PYD, and 0.1024 pmol/
10 μL DPD. Mix well. Name the solution as “Cross-links
Calibration solution 6.”
hydroxyproline, 0.2048 pmol/10 μL DHLNL,
0.08192 pmol/10 μL HLNL, 0.8192 pmol/10 μL PYD, and
0.04096 pmol/10 μL DPD. Mix well. Name the solution as
“Cross-links Calibration solution 7.”
0.0328 pmol/10 μL HLNL, 0.328 pmol/10 μL PYD, and
0.0164 pmol/10 μL DPD. Mix well. Name the solution as
“Cross-links Calibration solution 8.”
0.01312 pmol/10 μL HLNL, 0.01312 pmol/10 μL PYD,
and 0.00656 pmol/10 μL DPD. Mix well. Name the solution
as “Cross-links Calibration solution 9.”
12. Transfer a required amount of the prepared Calibration stan-
dards to a HPLC vial and load it into the autosampler for
analysis of cross-links samples. Store the remaining amount of
the standards at 20 C and use them further as and when
needed.
3.2.4 Sample Acquisition 1. Start up the LC-MS/MS system (see Note 16). Perform at least
and Data Analysis five injections of water onto the column using the cross-links
method. This is to equilibrate the column and the mass spec-
trometer ion source with the cross-link method conditions.
2. Following equilibration, inject a few Cross-links Calibration stan-
dard for sensitivity and chromatography check (see Note 17).
3. Inject all the Cross-links Calibration standards (see Note 18),
followed by injecting the processed tissue samples.
Immediately following tissue samples, inject all the Cross-links

Calibration standards again to complete one bracket of the run.
Similarly, inject the other tissue samples by bracketing them
between the Cross-links Calibration standards.
4. Wash the column and the ion source by injecting water samples
using column wash method.
5. Column wash method is as follows: Flow rate, 0.5 mL/min
using a multistep gradient method; total run time, 40.0 min.
The gradient is as follows: Hold column wash solution B at 5%
for 1.0 min, and then increase linearly to 95% within 8 min.
Decrease the column wash solution B at 5% within 8.5 min and
hold for up to 10 min. Again linearly increase the column wash
solution B to 95% within 18 min. Decrease the column wash
solution B at 5% within 18.5 min and hold for up to 20 min.
Linearly increase the column wash solution B to 95% within
28 min. Decrease the column wash solution B at 5% within
28.5 min and hold for up to 30 min. Linearly increase the
column wash solution B to 95% within 38 min. Decrease the
column wash solution B at 5% within 38.5 min and hold for up
to 40 min.
6. Calibration parameters are as follows: Use “External method”
for a standard type method (see Note 19). Set the Curve type as
“Quadratic,” Origin as “Ignore,” and Weighting as “Equal.”
Use the averaged out calibration curve from each of the two
adjacent calibrator levels. Select the sample type for Calibration
standards as “Cal STD,” and assign the correct levels to each
bracketing Calibration standard (see Note 20).
7. Units used for analysis of Cross-links Calibration standards:
pmols/10 μL.
8. After sample acquisition onto the LC-MS/MS, evaluate the
generated chromatographic data of each biomarker individually
by observing the Cross-links Calibration standards and tissue
samples in terms of peak shape and peak integrity (manually
integrate the peak if needed). Also evaluate % CV (coefficient of
variance) of peak responses of adjacent calibrators injected
before and after the tissue samples to monitor for any drift in
peak responses onto the detector and % error of calculated
amount of each calibrator on comparing it with its actual
amount and the r2 (coefficient of determination) of the calibra-
tion curve.
9. Check the peak responses of the tissue samples and check where
they lie on the calibration curve. Use the most relevant calibra-
tion levels, which cover the peak response range of the tissue
samples and exclude the rest of the calibration points from the
analysis (see Note 21).
10. Once the chromatographic data are finalized, apply the correct
conversion factor to the tissue samples. This would give the
actual amount of biomarker present in the total volume of
hydrolysate sample. Conversion factor is the ratio of the vol-
ume of the hydrolysate sample to the volume of the injected
sample (e.g., for a 200 μL hydrolysate sample and an injected
10 μL sample onto the column, the conversion factor would be
200/10 ¼ 20).
11. Generate the final LC-MS/MS results summary file into an
Excel file for further calculations.
12. The amount generated are in terms of pmols/volume of sam-
ple hydrolysate. Convert the amount in terms of pmols/mL.
13. Normalize the amount (pmols/hydrolysate sample) of
DHLNL, HLNL, PYD, and DPD against HYP amount
(pmols/hydrolysate sample). Normalize the amount (pmols/
hydrolysate sample) of all biomarkers against dry tissue weight
(mg) and normalize the amount (pmols/mL) of all biomarkers
against protein content (mg/mL).
14. Conduct the statistical data analysis on Graph Pad Prism soft-
ware, version 6 or latest. Use unpaired t-test for the compari-
son of healthy and diseased animal groups. Use one-way
ANOVA for all treated groups against the untreated group to
compare the effect of one variable at the time (dose or different
drugs). Conduct all the Prism analysis on cleaned data by
excluding the outliers, using the Rout test, with the Q value
set at 1%.
4 Notes
1. The NaBH4 buffer is not stable; make it right before the

reduction reaction. Use and discard appropriately.
2. To avoid exposure to dangerous fumes, prepare the solutions
inside a fume hood.
3. Heptafluorobutyric acid (HFBA) is sensitive, and hence if the
solution exposes to light, the concentration of HFBA tends to
deteriorate having an adverse impact on the sensitivity of bio-
markers and their retention times onto the column.
4. We use 5 mL polystyrene tubes with cell-strainer caps to insert
the tissue in for freeze-drying. The tissue samples should be
placed on dry ice and allowed to thaw such that they can be
easily cut with a blade. Dry ice is a hazard and should be
handled carefully in accordance with laboratory guidelines.
Freeze-drying results in approximately 70% loss of weight of
tissue, and at the end of the procedure, minimum 10 mg dry
tissue must be gained; therefore, the initial weighing of the wet
tissue for each sample must be done accordingly (i.e., for liver
samples, 300 mg wet weight is a good starting point). After
each sample has been weighed out, it must be snap frozen in
dry ice (all tubes can be stored at 80 C for 30 min before
being placed on the freeze dryer). It can take up to 2 days for
the samples to freeze-dry completely; at the end of the proce-
dure, perform a qualitative (tissue should appear pale and
lighter in color) and quantitative (check to see approximately
70% of the weight has been deducted).
5. For the cross-links procedure, 10 mg has been determined to
be the ideal amount for the assay (although in some instances
5 mg is used). We usually weigh out 20 mg, add 1 mL PBS, and
take 500 μL of the homogenate across to further steps (which is
equivalent to 10 mg).
6. For certain tissue types, homogenizing two more times is not
sufficient. The tube must be observed thoroughly to ensure an
evenly reconstituted homogenate. The homogenization can be
performed a number of times (when dissolved evenly, the
homogenate usually appears pink). If after 5–6 cycles the pow-
der is not dissolved into the PBS, use a pipette tip to poke at the
tissue that may be stuck to the walls of the tube and then
homogenize again. Occasionally, the metal beads lead to
break of the integrity of the tube, which may in turn lead to
leakage of the homogenate. If this occurs midway, stop the
machine and reweigh the tissue. A good safety catch is to
check the tubes after each homogenization cycle and change
the lids if scratches are observed.
7. It is critical the homogenate is evenly reconstituted before
transferring. Therefore, pipette up and down carefully before
transferring the 500 μL. If any large tissue bits are observed,
repeat the homogenization.
8. If the skin sample is in excess of 14 mg or so, then the sample
should be divided into two, and both subsections should be
taken across in individual metal bead lysing matrix tubes.
9. The homogenization should be performed every 2–3 h (5–6
at every time point) when feasible. Before the end of the 24-h
period, the tissue should be entirely homogenized and no skin
tissue should be visible.
10. Concentrated acetic acid is corrosive to the skin and must
therefore be handled with appropriate care. This step should
be performed in a fume hood.
11. If after the 24-h hydrolysis, the procedure cannot be continued
straight away, the samples containing HCl can be stored a
20 C.
12. If the sample cleanup using SPE (i.e., next step) is to be

performed later, the dried pellet after the HCl has dried can
be stored at 4 C as these samples are stable.
13. On the automated SPE machine, the columns must have adap-
tors inserted onto them. Otherwise, the probe will not be able
to create pressure in the column during the SPE procedure.
14. The reconstituted sample can be transferred to any tube for a
quick spin. However, it is easier to use SnapStrip PCR tubes.
The sample should be pipetted up and down 5–6 to ensure
the pellet is evenly reconstituted in solution 1 before transfer-
ring to the SnapStrip tubes.
15. For cross-links analysis, the data is normalized against different
parameters. Protein measurement is one of the parameters used
in our laboratory. The protein assay can be conducted at the
end of the entire experimental procedure. The sample to be
used is the remaining sample after it was reconstituted in
200 μL of water and 50 μL was taken for SPE. If the dry weight
initially was 10 mg, a 1 in 100 dilution is to be performed for
the assay, and if the initial dry weight was 5 mg, a 1 in 50 dilu-
tion is to be performed for the protein assay. Nevertheless, the
dilution can be adjusted based on the results obtained. Prepare
the samples in this manner, and then 15 μL of each diluted
sample can be pipetted into the appropriate wells (step 4 in the
protein assay kit booklet). The plate readout obtained
(mg/mL) should be multiplied by the dilution factor used
for sample preparation. Protein assay measurements should
be reported as mg/mL.
16. Switch on all modules of UHPLC and switch the mass spec-
trometer from standby mode to scan mode. Allow the ion
source temperature to achieve the desired temperature as per
the cross-links method and then turn on the solvent delivery
pumps.
17. Inject calibrators from the middle range (e.g., Cross-links Cal-
ibration solution 5, 4, and 3). Check for the retention time,
peak shape, and sensitivity of the standards and compare with
the previously injected reference ones. In cases where one of
the parameters does not match with the previous results,
re-prepare the mobile phase using a new bottle of HFBA. It
is critical to procure HFBA in small pack sizes and do not reuse
from the same pack for LC-MS/MS analysis.
18. Inject the calibration standards in the reverse order starting
from “Cross-links Calibration solution 9” to “Cross-links Cal-
ibration solution 1” to avoid any autosampler carry-over effect.
Inject a water sample before and after each calibration cycle to
monitor contamination and carry-over, if any.
19. Deuterated internal standards of each biomarker are very diffi-

cult to synthesis and thus very tough to get. Hence, an external
standard method is used to conduct the cross-links analysis.
20. In the method setup section of the LC-MS/MS software,
create nine calibration levels for each biomarker. Capture the
actual amounts of each calibration standard. Allocate the cor-
rect levels to their corresponding calibrators and analyze.
21. The calibration range for each biomarker is very broad, extend-
ing beyond the perfectly linear portion of the curve. However,
to analyze all the tissue samples in a single run, such broad
range is necessary. Hence, use only the relevant calibration
levels for the calculations, ignoring the other ones.
References
1. Trackman PC (2016) Enzymatic and Zuurmond AM, Bank RA (2005) The type of
non-enzymatic functions of the lysyl oxidase collagen cross-link determines the reversibility of
family in bone. Matrix Biol 52-54:7–18 experimental skin fibrosis. Biochim Biophys Acta
2. van der Slot-Verhoeven AJ, van Dura EA, 1740:60–67
Attema J, Blauw B, De Groot J, Huizinga TW,
Chapter 7
Mass Spectrometry-Based Proteomics to Define

Intracellular Collagen Interactomes
Ngoc-Duc Doan, Andrew S. DiChiara, Amanda M. Del Rosario,
Richard P. Schiavoni, and Matthew D. Shoulders
Abstract
We present the development, optimization, and application of constructs, cell lines, covalent cross-linking
methods, and immunoprecipitation strategies that enable robust and accurate determination of collagen
interactomes via mass spectrometry-based proteomics. Using collagen type-I as an example, protocols for
working with large, repetitive, and GC-rich collagen genes are described, followed by strategies for
engineering cells that stably and inducibly express antibody epitope-tagged collagen-I. Detailed steps to
optimize collagen interactome cross-linking and perform immunoprecipitations are then presented. We
conclude with a discussion of methods to elute collagen interactomes and prepare samples for mass
spectrometry-mediated identification of interactors. Throughout, caveats and potential problems research-
ers may encounter when working with collagen are discussed. We note that the protocols presented herein
may be readily adapted to define interactomes of other collagen types, as well as to determine comparative
interactomes of normal and disease-causing collagen variants using quantitative isotopic labeling (SILAC)-
or isobaric mass tags (iTRAQ or TMT)-based mass spectrometry analysis.
Key words Collagen proteostasis network, Co-immunoprecipitation, Proteomics, Cross-linking,

Mass spectrometry
1 Introduction
Collagen is the molecular scaffold for animal life [1, 2]. In addition
to forming the primary proteinaceous component of bone, skin,
cartilage, basement membranes, and other tissues, collagen also
facilitates diverse phenomena such as cell–cell signaling, wound
healing, and cell migration [3]. Proper execution of the folding,
modification, and quality control processes required for biogenesis
of this complex protein is, therefore, critical [4, 5]. Collagen pro-
teostasis, encompassing all of these factors, can be disrupted by
mutations in collagen molecules themselves or by dysfunction of
the proteostasis network [6–9], which is the integrated system of
intracellular chaperones, quality control factors, and trafficking
95
96 Ngoc-Duc Doan et al.
mechanisms that addresses cellular protein production challenges

[10, 11]. The hierarchical nature of collagenous matrices allows the
resulting defects in individual collagen molecules to propagate to
the level of tissues [12, 13], engendering a broad category of
disease known as the collagenopathies [7, 14]. Current therapies
remain inadequate for alleviating pathologic manifestations of these
diseases, owing in part to an incomplete understanding of collagen
biogenesis.
Developing a molecular-level understanding of intracellular
collagen folding, processing, assembly, and quality control requires
that we first identify the players. Unfortunately, the lack of an
amenable collagen-I expressing cell model system, the absence of
high quality immunoprecipitation (IP)-grade collagen antibodies,
and the challenges associated with performing molecular biology
on a lengthy, GC-rich, and highly repetitive gene had, prior to our
work [15], precluded systematic study of the collagen proteostasis
network. After addressing practical problems associated with
handling collagen genes, we generated panels of fibrosarcoma and
osteosarcoma cell lines that inducibly express collagen variants
tagged with distinct antibody epitopes to overcome these road-
blocks. Covalent cross-linking and selective immunoprecipitation
of collagen-I from these cells, followed by mass spectrometry-based
proteomic analysis (Fig. 1), yielded the first unbiased map of the
wild-type collagen-I proteostasis network (Table 1). Our strategy
robustly identified virtually all previously known collagen-I inter-
actors, as well as ~30 putative new players in collagen-I biogenesis.
The novel interactors we identified encompass a wide range of
proteostasis-relevant endoplasmic reticulum proteins, including
co-chaperones, oxidoreductases, and trafficking mechanisms. We
validated a number of the newly identified players in collagen-I
proteostasis by immunoprecipitations and genetic knockdowns in
osteoblast-like cells that natively produce collagen-I. We also dis-
covered a novel collagen-I posttranslational modification, aspartyl
hydroxylation, via these interactome studies.
Using our work with wild-type collagen-I as a template for the
description of a generalizable strategy, we present a detailed proto-
col to map collagen interactomes, including approaches for manip-
ulating collagen genes, developing stable cell lines expressing
antibody epitope-tagged collagens, optimizing and validating
Fig. 1 Schematic representation of collagen interactomics workflow

Mass Spectrometry-Based Proteomics to Define Intracellular Collagen Interactomes 97
Table 1
Selected wild-type collagen-I interactors reproducibly identified by mass spectrometry [15]
Protein name Gene name Total peptides Unique peptides

Collagen α1(I) chain Col1A1 915 63
Protein disulfide isomerase P4HB 129 20
Prolyl 4-hydroxylase subunit α2 P4HA2 69 15
Prolyl 4-hydroxylase subunit α1 P4HA1 68 15
78 kDa glucose-regulated protein (BiP/GRP78) HSPA5 56 15
Collagen α2(I) chain Col1A2 42 13
Endoplasmin (GRP94) HSP90B1 35 12
Prolyl 3-hydroxylase 1 LEPRE1 25 9
Peptidyl-prolyl cis-trans isomerase FKBP10 FKBP10 19 8
Prelamin-A/C LMNA 14 8
Transitional endoplasmic reticulum ATPase VCP 13 8
Protein disulfide isomerase A3 PDIA3 16 7
Fibronectin FN1 11 7
Cytoskeleton-associated protein 4 CKAP4 10 7
Peptidyl-prolyl cis-trans isomerase B PPIB 25 6
Procollagen-lysine,2-oxoglutarate 5-dioxygenase 2 PLOD2 20 6
Procollagen-lysine,2-oxoglutarate-5-dioxygenase 3 PLOD3 17 6
Annexin A2 ANXA2 15 6
C-type mannose receptor 2 MRC2 12 6
Procollagen-lysine,2-oxoglutarate-5-dioxygenase 1 PLOD1 16 5
Cartilage-associated protein CRTAP 14 5
Serpin H1 (HSP47) SERPINH1 10 5
Galectin-3-binding protein LGALS3BP 7 4
Peroxiredoxin-1 PRDX1 5 3
Peptidyl-prolyl cis-trans isomerase A PPIA 5 3
Endoplasmic reticulum resident protein 44 ERP44 5 3
Calreticulin CALR 5 3
Golgi apparatus protein-1 GLG1 4 3
(continued)
Table 1
(continued)
Protein name Gene name Total peptides Unique peptides

Thioredoxin domain-containing protein 5 TXNDC5 4 3
Hypoxia up-regulated protein 1 HYOU1 3 3
Calnexin CANX 5 2
Polyubiquitin-C UBC 5 2
Reticulocalbin-1 RCN1 5 2
Endoplasmic reticulum resident protein 29 ERP29 4 2
Golgi integral membrane protein 4 GOLIM4 4 2
Thioredoxin domain-containing protein 12 TXNDC12 3 2
Calumenin CALU 3 2
DnaJ homolog subfamily B member 11 DNAJB11 2 2
Reticulocalbin-3 RCN3 2 2
Thioredoxin-related transmembrane protein 1 TMX1 2 2
mass spectrometry-grade immunoprecipitations and covalent

cross-linking reactions, preparing samples for proteomic analysis,
and performing mass spectrometry experiments. The protocol
should prove useful not just for studying wild-type collagen-I but
also for quantitatively evaluating how cells engage misfolding,
disease-causing collagen-I variants. Moreover, the protocol is read-
ily adaptable to studying the interactome of any of the other
27 types of collagen [16].
2 Materials
2.1 Molecular 1. Genes, vectors, and DNA oligomers: Human Col1A1 and
Biology and Cell Col1A2 genes (Origene), pENTR1A vector (Thermo Fisher
Culture Scientific), pTRE-Tight vector (Clontech), pLenti.CMV/TO.
DEST vectors with assorted resistance cassettes (Addgene)
[17], puromycin and hygromycin linear selection markers
(Clontech), DNA primers (Sigma), and RRE, REV, and
VSVG vectors for lentivirus production (Addgene).
2. Enzymes: Q5 Polymerase, BamHI, EcoRV, NotI, T4 ligase,
Antarctic phosphatase (New England BioLabs; NEB), and LR
Clonase (Thermo Fisher Scientific).
3. Gel Purification Kit (Omega BioTech).

4. Bacterial cell lines: Sure2 E. coli (Agilent) and DH5α electro-
competent and Stbl3 competent E. coli (Thermo Fisher
Scientific).
5. Human cell lines: Human bone osteosarcoma Saos-2 cells
(ATCC), human fibrosarcoma HT-1080 Tet-Off cells that
constitutively express the tetracycline (Tet) transactivator
(Clontech), and HEK293FT cells (Thermo Fisher Scientific).
6. Cell culture: Dulbecco’s Modified Eagle Medium (DMEM),
Minimum Essential Medium (MEM), fetal bovine serum
(FBS), and trypsin-EDTA (all from Corning Cellgro),
tetracycline-free FBS (Clontech), and Opti-MEM media
(Thermo Fisher Scientific).
7. Transfection reagents: XFect (Clontech) and Lipofectamine
2000 (Thermo Fisher Scientific).
8. Antibiotics: Penicillin-streptomycin (Thermo Fisher Scien-
tific), puromycin (Corning Cellgro), hygromycin (Enzo Life
Sciences), and blasticidin (Sigma).
9. Lenti-X GoStix (Takara).
10. Chemicals: Doxycycline hyclate, L-glutamine, poly-D-lysine,
and polybrene (Sigma).
11. Antibodies: anti-HSP47 (ADI-SPA-470; Enzo Life Sciences),
anti-HA (sc-7392; Santa Cruz), anti-PDI (sc-20,132; Santa
Cruz), anti-Colα1(I) (LF68; Kerafast), anti-Colα2
(I) (SAB4500363; Sigma), and secondary antibodies 800CW
goat anti-rabbit, 800CW goat anti-mouse, 680LT goat anti-
rabbit, and 680LT goat anti-mouse (LiCor Biosciences).
2.2 Preparation All chemicals and supplies used in these protocol steps must be mass
of Protein Samples, spectrometry-grade. RNase-/DNase-free microcentrifuge tubes
Immunoprecipitation, employed can be from a variety of vendors, but should not be
Protein Elution, autoclaved prior to use.
and Precipitation 1. Radioimmunoprecipitation assay (RIPA) buffer: 50 mM Tris
(VWR), 0.5% sodium deoxycholate (Alfa Aesar), 1% Triton
(INTEGRA Chemical), 0.1% sodium dodecyl sulfate (SDS;
Alfa Aesar), 50 mM NaF (Alfa Aesar), 5 mM β-glycerol phos-
phate (Sigma), 1 mM sodium metavanadate (Sigma), 5 mM
ethylene glycol-bis(β-aminoethyl ether)-N,N,N0 ,N0 -tetraacetic
acid (EGTA; STREM Chemicals), 5 mM ethylenediaminete-
traacetic acid (EDTA; Aqua Solutions), 100 μM sodium pyru-
vate (Sigma), 1 mM phenylmethylsulfonyl fluoride (PMSF;
Thermo Fisher Scientific), and protease inhibitor mixture
(Pierce) at pH 7.4 in dd-H2O.
2. Lysis buffer: 50 mM Tris, 1% Triton, 1 mM PMSF, 150 mM

NaCl, 1.5 mM MgCl2, 1 mM EDTA, and protease inhibitor
mixture (Biotool) or tablets (Pierce) at pH 7.4.
3. Phosphate-buffered saline (Corning).
4. Dithiobis(succinimidyl propionate) (DSP; Sigma).
5. Ultrasound processor (Cole Parmer), tube rotator (VWR),
refrigerated benchtop centrifuge (Thermo Scientific), vortex
mixer (Corning).
6. Anti-HA agarose beads (A2095; Sigma).
7. Elution buffer: 6% SDS in 1 M aqueous Tris buffer at pH 6.8.
8. Solvents: MeOH and CHCl3 (Thermo Fisher Scientific).
9. Vacuum centrifuge dryer (Cole Parmer).
2.3 Preparation All chemicals and supplies used in these protocol steps must be mass
of Tryptic Peptides spectrometry-grade. RNase-/DNase-free microcentrifuge tubes
and Mass employed can be from a variety of vendors, but should not be
Spectrometry autoclaved prior to use.
1. Chemicals: Iodoacetamide (IAA; Sigma), 1,4-dithiothreitol
(DTT; Sigma), urea and formic acid (Fluka).
2. NH4HCO3 buffer: 100 mM NH4HCO3(aq) at pH 8.0.
3. TFA buffer: 0.1% aqueous trifluoroacetic acid (TFA; Sigma).
4. Formic acid buffer: 0.1% aqueous formic acid.
5. TFA/MeCN buffer: 0.1% TFA in 9:1 MeCN:H2O (MeCN
from Sigma).
6. Formic acid/MeCN buffer: 0.1% formic acid in 4:1 MeCN:
H2O.
7. Pierce C18 Peptide Desalting Spin Column (Thermo Fisher
Scientific).
8. Aluminum foil (Reynolds).
9. Water bath at 56 C.
10. EASY-nLC 1000 or similar HPLC with autosampler
connected to a Thermo Fisher Q Exactive Hybrid
Quadrupole-Orbitrap Mass Spectrometer or any similar high-
resolution, accurate-mass mass spectrometer.
11. Proteome Discoverer 2.2 with Mascot search engine.
3 Methods
Stable cell systems are valuable to enable robust and reproducible

determination of collagen-I interactomes via an IP-/MS-based
proteomics approach. In the absence of MS-grade collagen-I
3.1 Creating Cellular antibodies, to permit the necessary IPs, we use cloning strategies to
Platforms introduce short, well-defined HA or FLAG antibody epitopes
for Biochemical between an ER-targeting signal sequence and collagen-I’s
Studies of Collagen N-propeptide. More recently, Cas9-based approaches may make it
possible to introduce such epitopes directly in endogenous gene
loci [18, 19]. With necessary genes encoding antibody epitope-
tagged collagens in hand, we next engineer cell lines that stably and
inducibly express the tagged collagen proteins. As described below,
in HT-1080 cells, we used a transfection-based strategy to deliver
genes [15]. In Saos-2 cells, we used a lentivirus-based strategy
[15, 20]. Work in both cell types is described below. Either may
be used to introduce genes for other collagens or collagen disease-
causing variants into essentially any cell system of interest. An
advantage of HT-1080 cells for collagen-I studies is that, although
they are capable of synthesizing and secreting various collagen
types, they do not express collagen-I, and therefore the collagen-I
interactome can be readily enriched. On the other hand, Saos-
2 cells do endogenously express collagen-I. In such a system, it is
more straightforward to mimic autosomal dominant collagen-I
pathologies by simply introducing the disease-causing collagen-I
variant in a background of endogenous wild-type collagen-I expres-
sion. Note that we have found that adenoviruses may also be used
for robust transient expression of collagen (not described below, as
the focus is on stable cell line development). An adenovirus-based
strategy can be especially valuable for transient collagen variant
expression in primary cells.
3.1.1 Transfection Col1A1 and Col1A2 genes were obtained prior to transfer into an
Approach to Prepare Stable appropriate expression vector (Fig. 2a), as described below (see
HT-1080 Cells Inducibly Note 1).
Expressing Collagen Genes
1. Using appropriate restriction sites (e.g., BamHI and EcoRV),
insert a DNA sequence (IDT) encoding the preprotrypsin signal
sequence (PPT; to target collagen strands to the secretory path-
way), the desired epitope tag (e.g., HA or FLAG), and a NotI
restriction digestion site (i.e., PPT.HA.NotI or PPT.FLAG.
NotI) into a pTRE-Tight vector (see Notes 2 and 3).
2. PCR-amplify (using the primers shown below which remove
the native signal sequence) collagen-I genes over 35–40 PCR
cycles. For amplification of collagen-I genes, we recommend
the use of Q5 Polymerase with the addition of the High GC
Enhancer (protocol from NEB) and a 2 min extension cycle (see
Note 4).
Col1A1: For. 50 -ACATCAGCGGCCGCACAAGAGGAAGG
CCAAGTCGAG-30 .
Rev. 50 - AAAAAAGTCGACTTACAGGAAGCAGACAGGG-30 .
Fig. 2 (a) Generalized schematic for collagen-I expression constructs. Either the
endogenous collagen signal sequence or the preprotrypsin signal sequence can
be employed. The tag can be HA, FLAG, c-MYC, or any of a number of other short
antibody epitope tags. (b) Structure of dithiobis(succinimidyl propionate) (DSP). (c)
Optimizing DSP cross-linker concentration by evaluating the co-immunoprecipitation
of known collagen-I interactors in the HT-1080 cell system. The control sample
represents HT-1080 cells that do not express HA-tagged collagen-I upon induction.
This figure was adapted from DiChiara et al. [15]
Col1A2: For. 50 -AAAAAAGCGGCCGCAACATGCCAATC

TTTACAAGAGGAAAC-30 .
Rev. 50 -AAAAAAGATATCTTATTTGAAACAGACTGGGCC
AATG-30
3. Restriction digest amplicons using the NotI and EcoRV restric-
tion enzymes (protocol from NEB).
4. Gel purify samples using Gel Purification Kits.
5. Ligate into digested pTRE-Tight vectors prepared in Subhead-
ing 3.1.1, step 1, using T4 ligase (protocol from NEB)
followed by transformation into competent E. coli (see Note 5).
6. Transfect antibody epitope-tagged, tet-regulated collagen-I
expression vectors using XFect into HT-1080 Tet-Off cells
that constitutively express the tetracycline transactivator along
with either a puromycin or hygromycin linear selection marker
(protocol from Clontech). Co-transfection with a linear selec-
tion marker at a fixed ratio of 25:1 (5 μg: 0.2 μg) reduces the
probability of a cell being transfected with the linear selection
marker alone, potentially conferring resistance without integra-
tion of the collagen gene.
7. Change the media 18–24 h post-transfection.

8. 48 h post-transfection, treat cells with antibiotic (experimen-
tally optimized to be 0.25 μg/mL puromycin or 150 μg/mL
hygromycin) for 10–12 days (changing the media every 2 days)
for stable selection.
9. Amplify heterostable cell lines, and then split to establish genet-
ically homogenous single colonies to analyze for collagen-I
expression and secretion (Fig. 2a). Repeat above process to
create cell lines that express both Col1A1 and Col1A2 or
other collagen variants, as desired (see Note 6).
3.1.2 Lentiviral Approach 1. Construct pENTR1A vectors encoding the Col1A1 and
to Prepare Stable Saos- Col1A2 genes by ligating Col1A1 or Col1A2 extracted from
2 Cells Inducibly pTRE.Tight vectors (described above) using the NotI and
Expressing Collagen Genes EcoRV restriction enzymes into a pENTR1A vector already
encoding an ER-targeting PPT signal sequence and an HA
antibody epitope tag (or other desired tag; see Note 7). All
pENTR1A vectors must be sequence-confirmed before
proceeding.
2. Introduce the resulting construct into the appropriate pLenti.
CMV/TO.DEST Gateway destination vector via LR clonase-
mediated recombination (protocol from Thermo Fisher Scien-
tific). The resulting pLenti vectors must be sequence-
confirmed before proceeding to lentiviral production.
3. Produce lentivirus using the third-generation system by
co-transfecting HEK293FT cells with the lentiviral plasmids
and packaging vectors encoding RRE, REV, and VSVG using
Lipofectamine 2000. Briefly, (1) incubate the plasmid mixture
(pLenti vector, 15 μg; RRE, 15 μg; REV, 6 μg; and VSVG,
3 μg) with 60 μL of Lipofectamine 2000 in 3 mL of Opti-MEM
media for 45 min at room temperature (RT); (2) add the
resulting mixture dropwise to a 10 cm dish of HEK293FT
cells (~8 106 cells plated on poly-D-lysine-coated plates
~12 h pre-transfection) and incubate at 37 C for 12 h;
(3) remove and replace media with DMEM (6 mL) and incu-
bate the plates for another 36 h, monitoring for successful
production of lentivirus using Lenti-X GoStix (protocol from
Takara); (4) collect viral supernatant 48 h post-transfection;
and (5) use supernatant immediately for transductions of Saos-
2-TREx cells.
4. Transduce Saos-2TREx (see Note 8) cells with a range of titers
of crude viral supernatants of Tet-responsive HA-collagen-α1
(I). Add polybrene to the viral supernatant (final concentration
of 4 μg/mL during transductions) to increase transduction
efficiency.
5. Select heterostable cell lines using 250 μg/mL hygromycin or

other appropriate selection agent and isolate single colonies
(Fig. 2a; see Note 9).
6. Assay for HA-collagen-α1(I) expression in isolated single colo-
nies using immunoblotting.
3.2 Immunoisolation The protocol below uses Saos-2TREx cells inducibly expressing
of the Collagen-I PPT.HA.Colα1(I) (Saos-2Colα1(I) cells) as an example but can be
Interactome adapted for essentially any cell type of interest. Note that a control
for Proteomic Analysis IP should also always be performed on cells not expressing the
antibody epitope-tagged collagen.
1. Plate Saos-2Colα1(I) cells (~6 106) in 10 cm dishes with 10 mL
of DMEM supplemented with 15% heat-inactivated FBS,
100 IU/mL penicillin, 100 μg/mL streptomycin, 150 μg/
mL hygromycin, and 2 mM L-glutamine. Maintain cells as a
monolayer overnight at 37 C in a humidified atmosphere of
5% CO2(g).
2. Induce HA-tagged wild-type collagen-I expression by over-
night incubation with media containing 4 μg/mL doxycycline
(see Note 10) in the presence of 50 μM ascorbate, a required
cofactor for proper collagen hydroxylation and for induction of
synthesis and secretion of the endogenous, untagged collagen-I
(see Note 11).
3. After 48 h, remove media, wash with PBS, and add 1 mL of
trypsin-EDTA for 5 min. Neutralize trypsin by adding 7 mL of
complete DMEM. Pellet cells by spinning at 1000 g for
5 min.
4. Remove media and wash cells twice with PBS at (RT).
5. Resuspend cell pellet in a 15 mL Falcon tube containing 10 mL
of PBS and an experimentally optimized concentration of DSP
(75 μM for Saos-2Colα1(I) cells; Fig. 2b). Rotate tubes for 30 min
at RT to allow the cross-linking to proceed (see Note 12).
6. Quench excess DSP by adding 1 mL of 1 M aqueous Tris,
pH 8, and rotating for 10–15 min at RT.
7. Pellet cells by centrifuging at 1000 g for 5 min at 4 C,
followed by removal of the supernatant.
8. Lyse cells by resuspending them in 1 mL of RIPA buffer at
4 C. Vortex vigorously for 30 s, and then immediately sonicate
(130 Watt, 20 kHz, 70% amplitude) for 6 cycles of one 15 s
pulse each, returning tube to ice for 15 s between pulse cycles.
Lysates should be clear and fluid after sonication (see Note 13).
9. Post-sonicating, place the lysate solution on ice for 30 min to
ensure complete lysis.
10. Transfer cell lysates to 1.7 mL microcentrifuge tubes, and spin

for 15 min (21,100 g) at 4 C.
11. Transfer to new 1.7 mL tubes. Lysates may be stored at 80 C
at this stage, if desired.
12. Add 50 μL of HA bead slurry (25 μL of dried beads) to 1 mL of
lysate solution (see Note 14), and incubate the sample over-
night on a tube roller at 4 C (see Note 15).
13. Pellet the beads at 2000 g for 5 min at 4 C.
14. Wash beads 4 with 1 mL RIPA buffer to remove non-specific
binders, rotating 5 min at 4 C during each wash.
15. Pellet beads at 2000 g for 5 min at 4 C after the last wash,
and remove the supernatant (see Note 16).
3.3 Collagen To enhance the sensitivity of proteomic detection by minimizing

Interactome Elution the presence of any eluted IgG, we elute the proteins from beads
and Precipitation using a nonreducing, denaturing buffer.
1. Elute pelleted beads using 50 μL of elution buffer by heating to
100 C for 15 min (mix by shaking every 5 min during heat-
ing). Do not include a reducing agent during the elution, to
minimize release of any antibody fragments via disulfide reduc-
tion (see Note 17).
2. Spin down at 2000 g for 5 min, carefully collecting the
supernatant and avoiding uptake of any beads.
3. Repeat steps 1 and 2 above to elute a second time. Combine all
supernatants.
4. Add 450 μL MeOH to microcentrifuge tubes containing 90%
of the supernatant (see Note 18). Vortex vigorously for 30 s.
5. Add 150 μL CHCl3. Vortex vigorously for 30 s.
6. Add 450 μL H2O. Vortex vigorously 3 30 s. Spin down at
10,000 g for 2 min, rotate the tube 180 , and spin down
again at 10,000 g for 2 min (see Note 19).
7. Discard all but ~100 μL of the upper layer of supernatant,
being careful not to disturb the thin, white, protein-containing
film at the interface between the H2O/MeOH upper layer and
the CHCl3 lower layer.
8. Add 500 μL MeOH. Vortex vigorously for 30 s. White protein
particles should be observable. Spin down at 16,300 g for
2 min, followed by rotating the tube 180 , and spin down again
at 16,300 g for 2 min.
9. Repeat Subheading 3.3, step 8, 3 to remove all residual SDS,
which would otherwise interfere with MS analysis.
10. Use a speed vacuum to dry the pellet. Samples can then be
maintained at 20 C until use.
3.4 Protein Digestion All procedures should be carried out at RT unless otherwise
and Mass specified.
Spectrometry-Based
1. Add 22.5 μL of 8 M aqueous urea to the precipitated protein
Proteomic Studies pellet from Subheading 3.3, step 10. Vortex vigorously
3.4.1 Trypsin Digestion 3 30 s. Centrifuge at 3000 g for 1 min.
of Protein Pellets 2. Add 2.5 μL of 100 mM DTT dissolved in NH4HCO3 buffer to
and Preparation yield a final concentration of 10 mM DTT.
of Desalted Peptides
3. Incubate at 56 C for 1 h.
for Mass Spectrometry
Analysis 4. Cool samples to RT, and centrifuge at 3000 g for 1 min to
collect any H2O condensed on the internal surface of the tube.
5. Add 2.75 μL of 550 mM iodoacetamide to yield a final con-
centration of 55 mM iodoacetamide. Wrap tubes in foil to
protect from light, and mix end-over-end for 45 min (see
Note 20).
6. Add 169.75 μL of NH4HCO3 buffer. Vortex briefly, and then
add 2.5 μL of sequencing grade modified trypsin (0.4 μg/μL)
to yield a final volume of 200 μL (see Note 21).
7. Mix end-over-end at RT for 12 h or for 3 h at 37 C with
agitation on a thermomixer with hot plate.
8. Acidify the sample by adding 11 μL of 98% formic acid to
achieve a final concentration of 5% formic acid.
9. Remove white tip from the bottom of the peptide desalting
spin column. Place column into a 2 mL microcentrifuge tube.
Centrifuge at 3000 g for 1 min and discard flow through.
10. Wash the column with 300 μL of TFA/MeCN buffer, centri-
fuge at 3000 g for 1 min, and discard flow through (see Note
22). Repeat once.
11. Wash the column with 300 μL TFA buffer, centrifuge at
3000 g for 1 min, and discard flow through. Repeat once.
12. Add sample to the column, centrifuge at 3000 g for 3 min,
and discard flow through. Repeat centrifugation if not all of the
sample passed through the column (see Note 23).
13. Wash the sample on the column by adding 50 μL of formic acid
buffer. Centrifuge at 500 g for 3 min. Repeat twice.
14. Place column into a fresh microcentrifuge tube, and elute the
peptides from the column by adding 50 μL of formic acid/
MeCN buffer. Centrifuge at 3000 g for 3 min. Repeat this
elution step once.
15. Evaporate solvent in a vacuum centrifuge until near dryness.
Store at 20 C until mass spectrometry analysis.
3.4.2 Mass Spectrometry 1. Solubilize digested peptide mixture into a working volume
and Data Analysis (we typically use 30 μL) of HPLC-grade 0.1% aqueous formic
acid by vortexing 3 for 30 s each.
2. Inject samples (typically 1–5 μL volumes are reasonable) into a
C18 reversed-phase resin (3 μM resin; 120 Å pore size) column
(15 cm in length; 50 μM in diameter) on HPLC with auto-
sampler connected to high-resolution, accurate mass/mass
spectrometer. Elute peptides into the mass spectrometer for
analysis over a 140-min gradient of formic acid buffer ranging
to a 95:5 MeCN:H2O in 0.1% formic acid solution (see Notes
24 and 25).
3. Operate the mass spectrometer in data-dependent mode with a
full scan MS spectrum followed by MS/MS for the top ten
precursor ions in each cycle. Depending on instrumentation,
the number of precursor ions selected in each cycle may be
increased.
4. Analyze raw data using Proteome Discoverer 2.2, performing a
database search using Mascot for protein identification
[21]. Depending on instrumentation and study objectives,
search parameters and data analysis used may vary. For qualita-
tive mapping of the collagen-I interactome in these example
cellular systems, Mascot search parameters typically used were
as follows: mass tolerance for precursor ions, 10 ppm; fragment
ion mass tolerance, 15 mmu; missed cleavages of trypsin, 2;
fixed modifications were carbamidomethylation of cysteine;
and variable modifications were methionine oxidation and
hydroxylation of proline, lysine, aspartate, and/or asparagine.
Peptides with unambiguous peptide spectrum match, Mascot
scores 25, and isolation inference 30 were considered iden-
tified, resulting in an average false discovery rate of 0.0077 (see
Notes 26 and 27).
4 Notes
1. Sequencing of numerous collagen genes received from com-

mercial sources and other investigators revealed a number of
missense mutations as well as, in some cases, large or small
deletions of gene segments. We strongly recommend full
sequencing of any collagen vector obtained from any source
and correction of defects using site-directed mutagenesis or
other approaches. We also recommend complete sequencing
of collagen genes after any manipulation, including even simple
cut-and-paste reactions, not just PCR amplifications.
2. We incorporated an HA- or FLAG-epitope tag downstream of
PPT but otherwise at the N-terminus of collagen strands to
enable straightforward, orthogonal immunoblotting and IPs of

collagen-I polypeptides. Other antibody epitopes such as
cMyc, 3 FLAG, cleavable tags, etc. can also be used depend-
ing on the purpose of the study. The proteomic studies pre-
sented here use the HA epitope as an example. We note that,
because collagen folding begins at the C-terminus, placement
of these small tags at the N-terminus is unlikely to disrupt
folding and assembly, which we confirmed for collagen-I via a
variety of biochemical assays [15]. Note also that the
N-propeptide and thus the epitope-tag will be cleaved extracel-
lularly by N-propeptidases, depending on the cell system used.
3. Tet-inducible vectors were chosen to inducibly control the
expression of collagen in engineered stable cell lines. Inducible
expression minimizes potential negative selection pressure
placed on cells forced to continuously express ectopic collagen
that might render continuous propagation of stable cell lines
challenging.
4. Each collagen-I gene is highly repetitive in the triple-helical
domain and has about 65% GC-content overall, with local
regions containing up to 85% GC content in a 50 base-pair
span. These factors render PCR amplification of collagen genes
problematic, resulting in low yields and many small PCR frag-
ments, likely owing to incomplete gene synthesis. We found
that the yield of full-length collagen-I genes was slightly
boosted by increasing the recommended final extension time
by about 1.25-fold.
5. Note that because each collagen-I gene is over 4000 base pairs
in length, more than double the size of the pTRE-Tight expres-
sion vector, an inverted molar ratio is required when ligating
the two linear pieces of DNA (with the insert more abundant
than the vector). Even with inverted ratios, the most common
ligation product observed was the pTRE-Tight vector ends
being ligated together, despite the use of incompatible NotI
and EcoRV restriction enzymes to digest the vector. Sequenc-
ing confirmed that, in a rare event, the sticky end of the NotI
site was destroyed, leaving a blunt end of DNA that ligated to
the EcoRV blunt end. This rare intramolecular ligation was
more favorable and occurred more frequently than insertion
of the collagen-I genes. To avoid this problem, dephosphory-
lation of the vector backbone with Antarctic phosphatase prior
to ligating the two pieces of DNA proved useful in some cases
for successful insertion of collagen genes. We ultimately suc-
ceeded in ligating pTRE-Tight and the collagen-I PCR ampli-
cons, confirmed by sequencing the resulting clones with a
primer that binds to the vector, providing evidence for the
presence of both the 50 - and the 30 -ends of the collagen-I
genes. However, sequencing of the full gene revealed that an
apparent recombination event had removed >1000 base pairs

of the gene. Thereafter, we transformed collagen gene-
containing vectors into recombination-deficient Sure2 E. coli
cells. When grown at 30 C, recombination is effectively shut
down in these cells, allowing for the synthesis and propagation
of intact collagen-I genes. We recommend the use of these
recombination-deficient cells at 30 C for work with collagen
genes whenever possible.
6. Expression of the tetracycline transactivator allows doxycycline
(dox)-inducible expression of the collagen-I genes of interest.
We optimized HT-1080 culture conditions by varying the
concentration of dox for 48 h and then split each concentration
into two conditions: continued culture in the same concentra-
tion of dox or no dox in Tet-free FBS containing media. We
found that continuous culture in 1 ng/mL dox, 1000-fold
lower than the manufacturer’s recommended concentration,
efficiently suppresses both collagen-I genes. Removal of the
dox-containing media and replacement with dox-free media
induces both genes and the collagen proteins can be detected
by immunoblotting in both cell lysates and media. Continuous
culture in higher dox concentrations made it difficult to induce
collagen-I expression via a media change, and so we recom-
mend the use of 1 ng/mL dox during continuous culture.
7. Because we found the ligation of collagen genes into pEN-
TR1A vectors to be extremely slow and inefficient, we used
electroporation instead of heat shock to enhance transforma-
tion efficiency of the ligated vectors.
8. Expression of the tetracycline repressor in Saos-2-TREx cells
allows doxycycline (dox)-inducible expression of the collagen-I
gene under control of the tet operon (in contrast to
dox-repressed expression in Tet-Off cells like the HT-1080s
employed above). To create Saos-2TREx cells, Saos-2 cells
were stably transduced with a lentivirus encoding a constitu-
tively expressed tetracycline repressor protein and a blasticidin
resistance cassette. Stable cells were selected using 2 μg/mL
blasticidin, and single colonies were isolated and assayed for the
colony most responsive to dox treatment (using Tet-responsive
CFP to visualize inducible control of the tetracycline repressor
protein).
9. The production of single colony cell lines from stably trans-
duced Saos-2 cells is challenging. Heterogeneous cells must be
plated at a very low density (~1:10,000) in 10 cm dishes for
2–3 weeks prior to picking and expanding colonies.
10. The dox concentration for collagen induction should be sepa-
rately optimized for each cell system. Note that high concen-
trations ( 10 μg/mL) can be toxic.
11. Depending on the time course used for collagen induction,

ascorbate should be refreshed every 24 h owing to its poor
stability.
12. Transiently interacting chaperones, collagen-modifying
enzymes, and related biogenesis factors are often not observ-
able via traditional IP workflows. Indeed, we find that covalent
cross-linking is essential to co-immunoprecipitate even most of
the well-established collagen-I interactors, such as protein
disulfide isomerase and HSP47. DSP has dual NHS esters (see
DSP structure in Fig. 2b), thus targeting nearby primary
amines from lysine residues in protein complexes and forming
a conjugated network of closely interacting proteins
[22, 23]. DSP also has a disulfide bond that can be reduced
to release cross-linked proteins. Generally, the choice and
length of chemical cross-linkers used will impact the quality
and the relevance of data produced. The optimal choice of
reactive functional groups and the length of cross-linkers vary
based on the objectives of a given study [24, 25]. The optimal
DSP concentration is cell line- and collagen expression-level
dependent and therefore must be experimentally optimized to
maximally immunoprecipitate genuine interactors while mini-
mizing non-specific interactions. For HT-1080 cells, we used
150 μM DSP. For Saos-2 cells, we used 75 μM DSP. To
optimize the cross-linker concentration for immunoprecipita-
tions (Fig. 2c), we recommend using different concentrations
ranging from 0 to 2 mM and then choosing the lowest DSP
concentration that yields excellent signal for HA-collagen and
also allows detection of known, well-established collagen-I
proteostasis network components such as PDI and HSP47.
The optimal concentration is typically also the lowest DSP
concentration at which the cell pellet becomes slightly viscous
in the Falcon tube, providing a rough visual parameter to assess
cross-linking success. If co-immunoprecipitations without
covalent cross-linking are performed, a different protocol
from that described in this protocol is required, employing
much milder, non-denaturing buffers and gentler treatments
of cell lysates—both of which can lead to high background
owing to non-specific protein binding.
13. Likely owing to a small amount of residual DSP activity even
after Tris quenching, the cell pellet must be lysed immediately
with RIPA buffer to avoid the formation of large cell aggre-
gates. Another option to avoid formation of large cell aggre-
gates is to wash the cells repeatedly with PBS post-cross-
linking.
14. Because RIPA buffer contains SDS, which interferes with stan-
dard protein concentration assays, we typically do not measure
the protein concentration and instead assume that the protein
concentration in the same volume of lysate from the same

number of cells is the same.
15. Optional: Lysates can be precleared using 50 μL of packed
sepharose or agarose beads for 1 h at 4 C. The preclear step
can help to minimize non-specific binding of proteins to beads.
16. If the immunoprecipitated samples will not be used immedi-
ately, we have found that dried beads can be stored at 20 C
for at least a few weeks without compromising sample integrity.
However, we find that it is essential to first fully remove any
residual RIPA buffer, which can be achieved using an addi-
tional wash step with a standard lysis buffer (e.g., 50 mM Tris,
1% Triton, 1 mM PMSF, and protease inhibitor mixture at
pH 7.4).
17. Instead of SDS elution, which is our recommended approach,
proteins can alternatively be directly digested on the beads after
the immunoprecipitation to prepare for proteomic analysis.
This approach is less time-consuming, and, in addition,
because it skips elution and protein precipitation steps, the
results can be somewhat less variable. However, a significant
issue is that the HA antibody, which is chemically conjugated
to the bead, will also be digested to some extent and the
resulting abundant IgG tryptic peptides in the final peptide
mixture will substantially reduce sensitivity of the MS-based
proteomics. Therefore, we recommend the described SDS elu-
tion procedure.
18. To verify successful co-immunoprecipitation of collagen-I and
known components of the collagen-I interactome, at this stage
and prior to proceeding with any proteomic studies, we typi-
cally use 10% of the supernatant volume to run a 10%
SDS-PAGE gel, followed by transfer to a nitrocellulose mem-
brane and probing with anti-HA, anti-PDI, and anti-HSP47
(see Fig. 2c).
19. Small white droplet inclusions should form and be clearly
observed by eye. After spinning and re-equilibration, two layers
of solvent should be observed in the tube. Precipitated proteins
will form a thin white layer on the surface between the two
solvent layers. The thin white layer may stick to the wall of the
tube, in which case the spinning should be repeated.
20. Iodoacetamide is unstable and light-sensitive.
21. Sequencing grade modified trypsin has methylated lysine resi-
dues to minimize autoproteolysis during the digestion step.
22. We recommend switching to a fresh microcentrifuge tube after
every wash.
23. Optional: Reload flow through onto the column, and centri-
fuge again to maximize the binding of peptides to the column.
24. Other instruments may be used depending on availability. Note

that the solvent gradient and method setup were optimized for
the instrument listed.
25. We have observed that in a typical experiment, the proportion
of tryptic peptides derived from collagen is >50% of the total
detected peptides. In some cases, to ensure deeper coverage of
the interactome, we remove high levels of collagen-I proteins
after elution from the beads via the following protocol: Treat
eluted samples with 100 μM DTT to reduce the DSP cross-
linker, and release the interactome from the collagen-I bait
protein. Pass each sample over a 100 kDa molecular weight
cutoff filter (Millipore) to remove collagen-I (and other large
proteins). Precipitate the filtrates as in Subheading 3.3,
steps 4–9, and proceed. Another approach to ensure deeper
coverage of the interactome is to (1) run the protein mixture
that was eluted from bead on an SDS-PAGE gel, (2) silver stain
the gel, (3) cut out the collagen band in the gel, and (4) digest
the other bands using trypsin. A disadvantage of both these
approaches is that some high molecular weight protein inter-
actors will also be eliminated.
26. To exclude false positives and non-specific proteins binding to
antibody-conjugated beads, cells expressing HA-tagged CFP
in the cytoplasm (or other appropriate control cells) should be
used as a control. Proteins identified with high abundance in
the control sample should be excluded from the list of collagen
interactors. Proteins with only one unique peptide identified
should also be excluded.
27. Tryptic peptides can be labelled with iTRAQ or TMT tags
[26, 27] before mixing together and injecting into the mass
spectrometer if quantitative comparisons of collagen interac-
tomes are desired (e.g., between wild-type and disease-causing
collagen variants). SILAC proteomics [28] can alternatively be
performed using cells incubated with media containing stable
isotope labeling amino acids. See our recent paper for details
on using the quantitative SILAC technique in our collagen
system [15].
Acknowledgments
We thank Prof. Joseph Genereux (University of California—River-

side) for technical advice. This work was supported by NIH/-
NIAMS Grants R03AR067503 and R01AR071443 (to M.D.S.).
N.-D.D. was supported by Fonds de Recherche du Québec – Santé
(FRQS) and Canadian Institutes of Health Research postdoctoral
fellowships. A.S.D. was supported by a NIH Ruth L. Kirschstein
predoctoral fellowship (F31AR067615). This work was also
supported in part by the NIH/NIEHS under award

P30-ES002109 and by a Cancer Center Support (core) Grant
P30-CA14051 from the NIH/NCI.
References
1. Eyre DR (1980) Collagen: molecular diversity Biochem 84:435–464. https://doi.org/10.

in the body’s protein scaffold. Science 1146/annurev-biochem-060614-033955
207:1315–1322. https://doi.org/10.1126/ 11. Hartl FU, Bracher A, Hayer-Hartl M (2011)
science.7355290 Molecular chaperones in protein folding and
2. Brinckmann J (2005) Collagens at a glance. proteostasis. Nature 475:324–332. https://
Top Curr Chem 247:1–6. https://doi.org/ doi.org/10.1038/nature10317
10.1007/b103817 12. Ottani V, Martini D, Franchi M, Ruggeri A,
3. Ricard-Blum S (2011) The collagen family. Raspanti M (2002) Hierarchical structures in
CSHL Perspect Biol 3:a004978. https://doi. fibrillar collagens. Micron 33:587–596.
org/10.1101/cshperspect.a004978 https://doi.org/10.1016/S0968-4328(02)
4. Shoulders MD, Raines RT (2009) Collagen 00033-1
structure and stability. Annu Rev Biochem 13. Han L, Grodzinsky AJ, Ortiz C (2011) Nano-
78:929–958. https://doi.org/10.1146/ mechanics of the cartilage extracellular matrix.
annurev.biochem.77.032207.120833 Annu Rev Mater Res 41:133–168. https://doi.
5. Ishikawa Y, Bachinger HP (2013) A molecular org/10.1146/annurev-matsci-062910-
ensemble in the rER for procollagen matura- 100431
tion. Biochim Biophys Acta 1833:2479–2491. 14. Jobling R, D’Souza R, Baker N, Lara-Corrales-
https://doi.org/10.1016/j.bbamcr.2013.04. I, Mendoza-Londono R, Dupuis L,
008 Savarirayan R, Ala-Kokko L, Kannu P (2014)
6. Morello R, Bertin TK, Chen YQ, Hicks J, The collagenopathies: review of clinical pheno-
Tonachini L, Monticone M, Castagnola P, types and molecular correlations. Curr Rheu-
Rauch F, Glorieux FH, Vranka J, Bachinger matol Rep 16:394. https://doi.org/10.1007/
HP, Pace JM, Schwarze U, Byers PH, S11926-013-0394-3
Weis M, Fernandes RJ, Eyre DR, Yao ZQ, 15. DiChiara AS, Taylor RJ, Wong MY, Doan
Boyce BF, Lee B (2006) CRTAP is required N-D, Del Rosario AM, Shoulders MD (2016)
for prolyl 3-hydroxylation and mutations cause Mapping and exploring the collagen-I proteos-
recessive osteogenesis imperfecta. Cell tasis network. ACS Chem Biol 11:1408–1421.
127:291–304. https://doi.org/10.1016/j. https://doi.org/10.1021/acschembio.
cell.2006.08.039 5b01083
7. Myllyharju J, Kivirikko KI (2001) Collagens 16. Kuttner V, Mack C, Gretzmeier C, Bruckner-
and collagen-related diseases. Ann Med Tuderman L, Dengjel J (2014) Loss of colla-
33:7–21. https://doi.org/10.3109/ gen VII is associated with reduced transgluta-
07853890109002055 minase 2 abundance and activity. J Invest
8. Barnes AM, Cabral WA, Weis M, Makareeva E, Dermatol 134:2381–2389. https://doi.org/
Mertz EL, Leikin S, Eyre D, Trujillo C, Marini 10.1038/jid.2014.185
JC (2012) Absence of FKBP10 in recessive 17. Campeau E, Ruhl VE, Rodier F, Smith CL,
type XI osteogenesis imperfecta leads to dimin- Rahmberg BL, Fuss JO, Campisi J, Yaswen P,
ished collagen cross-linking and reduced colla- Cooper PK, Kaufmann PD (2009) A versatile
gen deposition in extracellular matrix. Hum viral system for expression and depletion of
Mutat 33:1589–1598. https://doi.org/10. proteins in mammalian cells. PLoS One 4:
1002/Humu.22139 e0006529. https://doi.org/10.1371/journal.
9. Marini JC, Reich A, Smith SM (2014) Osteo- pone.0006529
genesis imperfecta due to mutations in 18. Bressan RB, Dewari PS, Kalantzaki M,
non-collagenous genes: lessons in the biology Gangoso E, Matjusaitis M, Garcia-Diaz C,
of bone formation. Curr Opin Pediatr Blin C, Grant V, Bulstrode H, Gogolok S,
26:500–507. https://doi.org/10.1097/Mop. Skarnes WC, Pollard SM (2017) Efficient
0000000000000117 CRISPR/Cas9-assisted gene targeting enables
10. Labbadia J, Morimoto RI (2015) The biology rapid and precise genetic manipulation of
of proteostasis in aging and disease. Annu Rev mammalian neural stem cells. Development
144:635–648. https://doi.org/10.1242/dev. 4:171. https://doi.org/10.3389/fphar.2013.

140855 00171
19. Savic D, Partridge EC, Newberry KM, Smith 24. Leitner A (2016) Cross-linking and other
SB, Meadows SK, Roberts BS, Mackiewicz M, structural proteomics techniques: how chemis-
Mendenhall EM, Myers RM (2015) CETCh- try is enabling mass spectrometry applications
seq: CRISPR epitope tagging ChIP-seq of in structural biology. Chem Sci 7:4792–4803.
DNA-binding proteins. Genome Res https://doi.org/10.1039/c5sc04196a
25:1581–1589. https://doi.org/10.1101/gr. 25. Holding AN (2015) XL-MS: protein cross-
193540.115 linking coupled with mass spectrometry. Meth-
20. Wong MY, Doan N-D, DiChiara AS, Papa LJ ods 89:54–63. https://doi.org/10.1016/j.
3rd, Cheah JH, Soule CK, Watson N, Hulle- ymeth.2015.06.010
man JD, Shoulders MD (2018) A high- 26. Thompson A, Schafer J, Kuhn K, Kienle S,
throughput assay for collagen secretion sug- Schwarz J, Schmidt G, Neumann T, Hamon
gests an unanticipated role for Hsp90 in colla- C (2003) Tandem mass tags: a novel quantifi-
gen production. Biochemistry 57:2814–2827. cation strategy for comparative analysis of com-
https://doi.org/10.1021/acs.biochem. plex protein mixtures by MS/MS. Anal Chem
8b00378 75:1895–1904. https://doi.org/10.1021/
21. Perkins DN, Pappin DJC, Creasy DM, Cottrell ac0262560
JS (1999) Probability-based protein identifica- 27. Unwin RD, Pierce A, Watson RB, Sternberg
tion by searching sequence databases using DW, Whetton AD (2005) Quantitative proteo-
mass spectrometry data. Electrophoresis mic analysis using isobaric protein tags enables
20:3551–3567. https://doi.org/10.1002/( rapid comparison of changes in transcript and
Sici)1522-2683(19991201)20:18<3551:: protein levels in transformed cells. Mol Cell
Aid-Elps3551>3.0.Co;2-2 Proteomics 4:924–935. https://doi.org/10.
22. Kim KM, Yi EC, Kim Y (2012) Mapping pro- 1074/mcp.M400193-MCP200
tein receptor-ligand interactions via in vivo 28. Ong SE, Blagoev B, Kratchmarova I, Kristen-
chemical crosslinking, affinity purification, and sen DB, Steen H, Pandey A, Mann M (2002)
differential mass spectrometry. Methods Stable isotope labeling by amino acids in cell
56:161–165. https://doi.org/10.1016/j. culture, SILAC, as a simple and accurate
ymeth.2011.10.013 approach to expression proteomics. Mol Cell
23. Corgiat BA, Nordman JC, Kabbani N (2014) Proteomics 1:376–386. https://doi.org/10.
Chemical crosslinkers enhance detection of 1074/mcp.M200025-MCP200
receptor interactomes. Front Pharmacol
Chapter 8
Exploring Extracellular Matrix Degradomes

by TMT-TAILS N-Terminomics
Elizabeta Madzharova, Fabio Sabino, and Ulrich auf dem Keller
Abstract
Global characterization of protein N termini provides valuable information on proteome dynamics and
diversity in health and disease. Driven by the progress in mass spectrometry-based proteomics, novel
approaches for the dedicated investigation of protein N termini and protease substrates have been recently
developed. Terminal amine isotopic labeling of substrates (TAILS) is a quantitative proteomics approach
suitable for high-throughput and system-wide profiling of protein N termini in complex biological matri-
ces. TAILS employs isotopic labeling of primary amines of intact proteins in combination with an amine-
reactive high molecular weight polymer (HPG-ALD) for depletion of internal tryptic peptides and high
enrichment of protein N termini by negative selection. Thereby, TAILS allows simultaneous identification
of the natural N termini, protease-generated neo-N termini, and endogenously modified (e.g., acetylated)
N termini. In this chapter, we provide a protocol for tandem mass tag (TMT)-TAILS analysis and further
discuss specific considerations regarding N-terminome data interpretation using Proteome Discoverer™
software.
Key words N-terminome, Proteome Discoverer™, Proteolysis, TMT, TAILS
1 Introduction
The protein N terminus has a significant influence on a protein’s

biological functions [1]. Its sequence and posttranslational modifi-
cations (PTMs) often affect protein stability, regulate signal trans-
duction, and determine the protein’s final destiny, thereby affecting
cellular localization and function and being crucial for regulation of
cellular homeostasis [2–4]. Thus, comprehensive analysis of protein
N termini can define critical functions of proteins, provide in-depth
knowledge on their multifunctional roles in diverse biological pro-
cesses, and thereby reveal valuable information on the highly
dynamic nature of proteomes [1, 2]. Yet, common bottom-up
mass spectrometry-based techniques often provide limited infor-
mation about N termini, mostly due to the high abundance of
internal tryptic peptides after trypsin digestion of complex protein
115
116 Elizabeta Madzharova et al.
Fig. 1 Standard TAILS workflow. The TAILS experiment starts with the exposure of a complex proteome to a
specific test protease. Next, all primary amines (α-amines on the N-terminal and ε-amines of lysine side
chains) are labeled with isobaric tags (TMT or iTRAQ, indicated by multicolored diamonds), and samples are
TMT-TAILS N-Terminomics 117
samples. This hinders identification of the small fraction of

N-terminal peptides present in the mixture and imposes the use
of enrichment strategies [4, 5].
To specifically identify N termini, various approaches have been
developed to isolate and characterize N-terminal peptides from
highly complex peptide mixtures. These enrichment methods are
divided into positive selection methods, enriching N-terminal pep-
tides by affinity interaction, and negative selection methods, where
the N-terminome is enriched by selectively removing internal pep-
tides. Positive selection methods have the disadvantage of detecting
only endogenously free N termini, since naturally blocked
N-terminal peptides do not react with the affinity tags and therefore
are excluded from the analysis. By contrast, the negative selection
procedures gather information on both endogenously modified
natural N termini and unmodified natural and neo-N termini,
thereby offering a comprehensive analysis and annotation of the
N-terminome [2, 3, 5–7].
Among the N-terminomics strategies, terminal amine isotopic
labeling of substrates (TAILS) has gained its popularity due to the
highly successful negative selection strategy applied to profile the
N-terminome in complex proteome samples (Fig. 1) [8, 9]. To
identify substrates of a protease of interest using TAILS, a native
proteome is exposed to the protease to generate cleavage events
potentially relevant in physiological contexts. Next, the primary
amines present in the proteome (i.e., α-amines of protein N termini
and ε-amines of lysine side chains) are labeled at the protein level
using tags for isotopic labeling such as iTRAQ (isobaric tag for
relative and absolute quantitation) or TMT (tandem mass tags).
Thereby, the terminal information resulting from proteolytic pro-
cessing by the protease of interest is unambiguously preserved,
since the peptide fragments located downstream of the cleavage
site are covalently labeled at the N terminus and can be posteriorly
identified and assigned to a specific amino acid sequence of the
substrate [1, 3]. The labeled proteomes are digested with trypsin,
which generates internal tryptic and C-terminal peptides carrying a
free amino group at their N terminus. These tryptic peptides are
then depleted by covalently binding to an amine-reactive high
Fig. 1 (continued) pooled and trypsinized. The newly generated internal tryptic and C-terminal peptides are
depleted using a high-molecular aldehyde-derivatized polymer (HPG-ALD) that binds to the newly generated
free amino group at the N terminus. The labeled N termini and neo-N termini, as well as naturally blocked N
termini (acetylation, cyclization, and methylation, indicated by black square), do not react with the polymer
and are recovered by ultrafiltration. The enriched N-terminome is analyzed by tandem mass spectrometry,
which discriminates between protease-generated neo–N termini and mature protein N termini with help of
quantitative information from TMT/iTRAQ reporter ions
molecular weight, aldehyde-derivatized, polyglycerol polymer

(HPG-ALD). Peptides naturally N-terminally blocked (e.g., by
acetylation, cyclization, or methylation) and neo-N termini carry-
ing an isobaric tag at the N terminus have unreactive N termini and
thus remain unbound and can be recovered by ultrafiltration. The
N-terminome is collected in the flow-through fraction and ana-
lyzed by LC-MS/MS.
By enriching the N-terminome by negative selection, TAILS
provides both high coverage of N-terminal peptides present in the
proteome and allows reliable isotopic quantification of multiple
samples or conditions simultaneously in a single experiment.
Hence, TAILS is ideally suited to profile protein N termini and
their modifications on a proteome-wide scale [4, 10].
Here, we present a detailed TMT-TAILS workflow for analysis
of protein N termini from up to six different conditions using
isobaric 6plex-TMT tags. We provide a TMT-TAILS protocol for
analyzing the substrate repertoire of glutamyl endoproteinase
(GluC) as test protease with canonical specificity and secretomes
from mouse fibroblasts as sample proteome. We also discuss specific
considerations regarding the analysis of N-terminomics data using
Proteome Discoverer™ software and offer a possibility to obtain
our MS raw files and Proteome Discoverer™ workflow files to
perform a test run before proceeding with the analysis of your
samples.
2 Materials
2.1 Digestion 1. Proteins or proteome of interest.

with Test Protease 2. 1 M HEPES, pH 7.8.
(Optional)
3. 1 M sodium chloride.
4. GluC endoproteinase (test protease).
5. Digestion buffer: 50 mM HEPES, 100 mM NaCl, pH 7.8.
2.2 Identification 1. 8 M guanidine hydrochloride (GuHCl).

of Protease Substrates 2. 1 M HEPES pH 7.8.
by 6plex-TMT-TAILS
3. 350 mM Tris (2-carboxyethyl) phosphine hydrochloride
(TCEP).
4. 250 mM chloroacetamide.
5. TMT 6plex™ reagents.
6. Dimethyl sulfoxide (DMSO).
7. 1 M ammonium bicarbonate (NH4HCO3).
8. Acetone ( 20 C).
9. Methanol (MeOH; 20 C).
10. 100 mM sodium hydroxide (NaOH).

11. 1 M hydrochloric acid (HCl).
12. Hyperbranched polyglycerol-aldehydes (HPG-ALD) polymer
(available without commercial or company restriction from
Flintbox innovation network, the global intellectual exchange
and innovation network (http://www.flintbox.com/public/
project/1948)).
13. 5 M sodium cyanoborohydride (NaBH3CN).
14. Trypsin.
15. Amicon® Ultra-0.5 Centrifugal Filter Units (30 kDa cutoff).
2.3 Sample Desalting 1. C18 reverse-phase extraction disk.

(StageTip) 2. A blunt-ended syringe needle.
3. 200 μL pipette tips.
4. A 20 mL disposable syringe.
5. Activation buffer: MeOH, 80% ACN, 0.1% acetic acid.
6. Elution buffer: 80% ACN, 0.1% acetic acid.
7. Equilibrium buffer: 3% ACN, 1% trifluoroacetic acid (TFA).
8. Washing buffer: 0.1% formic acid (FA).
9. Resuspension buffer: 3% ACN, 0.1% formic acid.
10. Speed vacuum.
2.4 LC-MS/MS 1. Thermo Scientific Q Exactive Mass Spectrometer.

Analysis 2. Thermo Scientific™ EASY-nLC™ 1000 Liquid Chromatogra-
phy system.
3. Thermo Scientific PepMap™ RSLC C18 column (2 μm,
100 Å, 75 μm 50 cm).
4. Buffer A: 0.1% formic acid, 99.9% water.
5. Buffer B: 80% ACN, 0.1% formic acid, 19.9% water.
6. Xcalibur software v2.0.
2.5 Data Analysis 1. Thermo Scientific™ Proteome Discoverer software v2.2 avail-
able at https://www.thermofisher.com/order/catalog/prod
uct/OPTON-30795.
2. SEQUEST database search engine.
3 Methods
3.1 Proteolysis by The following steps are only required if the proteome is to be
Test Protease exposed to a protease of interest in vitro. For other studies, which
(Optional) focus only on the nature of the N-terminome or using samples that
have already undergone proteolysis in cell culture or in vivo sam-

ples, proceed directly with the 6plex-TMT-TAILS protocol.
1. Start with two samples of 160 μg total protein (concentra-
tion 1 μg/μL) under native conditions in digestion buffer
(see Note 1).
2. Add GluC endoproteinase (or the test protease of interest) at
1:100 (enzyme/proteome) ratio to the test sample and an
equivalent volume of digestion buffer to the control sample.
Important: Different proteases have different buffer require-
ments; hence adjust buffer composition accordingly.
3. Incubate for 16 h at 37 C (see Note 2).
4. Stop reaction by adjusting with 8 M GuHCl and 1 M HEPES,
pH 7.8 to a final concentration of 2.5 M GuHCl, 200 mM
HEPES, pH 7.8 to denature GluC. Important: The protein
concentration should not be lower than ~0.4–0.5 μg/μL.
5. Store the sample at 20 C until further processing.
3.2 6plex-TMT-TAILS 1. To assure total protein denaturation, incubate samples at 65 C

for 15 min.
2. Reduce cysteine residues by adding 1 mM TCEP.
3. Incubate samples at 65 C for 45 min.
4. Alkylate cysteine residues by adding 5 mM chloroacetamide
and mix thoroughly.
5. Incubate samples at 65 C for 30 min.
6. Divide samples (control and protease-treated) into three ali-
quots (technical replicates) containing 50 μg total protein each.
7. Resuspend each aliquot containing 200 μg of TMT reagent in
DMSO with a volume equivalent to the total volume of sample
from step 6. Important: see Note 3 for preparation of aliquots
of TMT reagent.
8. Add TMT reagents to control and protease-treated samples to
achieve 1:4 protein/TMT (w/w) ratio and 50% final DMSO
concentration.
9. Mix gently and incubate for 1 h at room temperature. Impor-
tant: The incubation time for each channel should be measured
as precisely as possible to reduce quantification errors derived
from differing labeling times.
10. To quench the labeling reaction, add 100 mM NH4HCO4 and
incubate for 30 min at room temperature.
11. Combine TMT-labeled samples in a 15 mL tube.
12. Add seven sample volumes of ice-cold acetone and one sample
volume of ice-cold MeOH to the labeled proteins.
13. Incubate samples at 80 C for at least 2 h.
14. Centrifuge samples at 4500 g at 4 C for 20 min.

15. Carefully discard the supernatant and wash with 5 mL of
ice-cold MeOH.
16. Centrifuge at 4500 g at 4 C for 10 min, and discard the
supernatant.
17. Air-dry the pellet. Important: Overdrying the sample may
impede protein solubilization.
18. Resuspend sample in 30 μL of 100 mM NaOH.
19. Add 100 mM HEPES, pH 7.8 and adjust the volume to a final
protein concentration of 1 μg/μL.
20. Take an aliquot of 5 μg of protein as quality control of trypsin
digestion.
21. Add trypsin at a ratio of 1:100 trypsin/protein ratio (w/w).
22. Incubate overnight at 37 C.
23. Remove 5 μg of the digested proteome. Separate digested and
non-digested samples by SDS-PAGE and stain with Coomassie
Blue or silver to assess digestion efficiency.
24. Remove ~10% of the digested proteome (preTAILS sample),
and store it at 20 C.
25. Adjust the pH of the remaining solution to pH 6–7 with
1 M HCl.
26. Add a threefold excess (w/w) of HPG-ALD polymer and
50 mM NaBH3CN reagent to deplete peptides with free
N-terminal amines (fully tryptic peptides) (see Note 4).
27. Incubate overnight at 37 C.
28. Pre-wash a 30 kDa Amicon Ultra-0.5 mL centrifugal filter unit
with water according to the manufacturer’s instructions.
Remove the water from the clean Amicon.
29. Transfer the sample to the Amicon previously washed and filter
by centrifugation at 10,000 g for 10 min at room tempera-
ture. The high molecular weight HPG-ALD polymer bound to
the fully tryptic peptides will remain in the filter. Recover
unbound peptides in the flow-through.
30. Repeat step 29 until the entire sample volume is filtered.
31. Wash the polymer with 30 μL of 100 mM NH4HCO3 and
centrifuge again.
32. Collect the flow-through and combine with the flow-through
from step 29 (TAILS sample).
3.3 StageTip 1. Stamp out two pieces of a C18 Empore disk using a blunt-
Desalting ended syringe needle, and press the cored pieces close to the
end of a P200 pipette tip.
2. Activate the disks by subsequently washing with 40 μL MeOH

and 40 μL of 80% ACN and 0.1% acetic acid solution through
the StageTip using a 20 mL syringe.
3. Equilibrate the disks by passing through 2 60 μL of equilib-
rium buffer. Avoid the tip drying completely.
4. Discard the flow-through.
5. Load 15 μL preTAILS and 150 μL TAILS samples, separately.
6. Wash by adding 2 40 μL of washing buffer, and discard the
flow-through.
7. Elute samples with 2 40 μL of elution buffer.
8. Speed vacuum desalted peptide mixture to dryness.
9. Resuspend samples in resuspension buffer.
10. Determine protein concentration using a NanoDrop
spectrophotometer.
3.4 LC-MS/MS Analyze preTAILS and TAILS samples with a tandem mass spec-
Analysis trometer coupled to a C18 chromatography system. Due to the
large variation of instrumental setups, details of the mass spectrom-
etry analysis are specific to the instrument used. As an example,
peptides from our dataset were analyzed on a Thermo Scientific Q
Exactive Mass Spectrometer coupled to a Thermo Scientific EASY-
nLC™ 1000 Liquid Chromatography system and a Thermo Scien-
tific PepMap™ RSLC C18 column (2 μm, 100 Å, 75 μm 50 cm).
For each analysis, 2 μg of peptide mixture was loaded at a constant
flow rate of 250 nL/min in Buffer A and eluted using a 2-h
gradient from 6% to 95% Buffer B. A data-dependent acquisition
(DDA) mode was applied to record the spectra with the following
MS1 parameters, scan range 300–1750 m/z, resolution of 70,000,
AGC target 3e6; MS2 parameters, resolution of 17,500, AGC
target 1e6; and higher energy collision dissociation (HCD), nor-
malized collision energy of 32%.
3.5 N-Terminome A variety of tools are available for proteomic assessment of TAILS
Data Analysis N-terminomics data with specific search settings, which vary
depending on the algorithm and the instrument used for data
acquisition [11–15]. Here we describe in detail a workflow for
the analysis of a TMT-TAILS dataset collected after treatment of
secretomes from murine embryonic fibroblasts with GluC endo-
peptidase (test protease) using Proteome Discoverer™ 2.2
software.
1. Open Proteome Discoverer™ 2.2 application and select New
Study/Analysis on the start page.
2. Choose a name for the study and a root directory. Click OK.
3. Select the TMTe 6plex as a quantification method to be used in
the study.
4. Click Add Files and select the mass spectrometry raw file.
Important: Proteome Discoverer accepts .raw, .mgf, .mzData,
.mzXML, and .mzML file formats as an input. Optional: As a
test run, use our MS raw file (available upon request from the
corresponding author).
5. Specify the quantification method used in the input file on the
Input Files page.
6. On the Samples tab, specify samples as “control” and/or
“sample” type.
7. Open New Analysis, select and drag the input file from the
input file page to the Analysis window.
8. Select Workflows tab to create workflows to use for the pro-
cessing and consensus step. Optional: Use our processing and
consensus workflow files (available upon request from the
corresponding author).
9. Create Processing Step Workflow by dragging and connect-
ing the following nodes into the Workflow tree window
(Fig. 2a):
l Spectrum File.
l Spectrum Selector.
l Sequest HT.
l Percolator.
l Reporter Ions Quantifier.
10. For N termini profiling, set the parameters for Sequest HT
node (see Note 5) (Fig. 2b):
l Upload an appropriate database.
l Choose enzyme semi-specific ArgC with one missed cleav-
age (see Note 6).
l Set minimum peptide length of 6 amino acids and maxi-
mum peptide length of 30 amino acids.
l Set precursor mass tolerance at 10 ppm and fragment mass
tolerance at 0.02 Da.
l Set TMT6plex lysine (+229.163 Da) and carboxymethyl of
cysteine (+58.005 Da) as static modifications.
l Set TMT-labeled N-termini (+229.163 Da), oxidation of
methionine (+15.995 Da), N-terminal acetylation
(+42.010565 Da), N-terminal pyro-Glu ( 17.026549
Da), or asparagine deamidation (+ 0.984 Da) as dynamic
modifications.
11. Create Consensus Step Workflow by dragging and connect-
ing the following nodes into the Workflow tree window
(Fig. 2c):
Fig. 2 Proteome Discoverer settings. (A) Processing Step Workflow nodes. (B) Sequest HT parameters for
TAILS analysis of mouse secretome. (C) Consensus Step Workflow nodes
l MSF Files.
l PSM Grouper.
l Peptide Validator.
l Peptide and Protein Filter.
l Protein Scorer.
l Protein Grouping.
l Peptide in Protein Annotation.
l Reporter Ions Quantifier.
l Protein FDR Validator.
l As for post-processing nodes, Data Distributions, and Dis-
play Settings.
12. In Grouping and Quantification tab, specify the ratios to
report for quantification and how to group the samples with
respect to the specified factor values.
13. Select Sample type as study variable.
14. Choose Sample type: Control denominators to be used for
bulk ratio generator and press Add Ratio.
15. Press Run in the Analysis window to start the search.
Output of analysis results should be inspected on the “Peptide
Group” level for interpretation of N-terminal peptides and can be
exported to appropriate formats for further analysis, e.g., using
spreadsheet applications. To annotate protein N termini for their
position in the mature protein, several tools are available including
ImproViser [16] and TAILS-ANNOTATOR [4].
4 Notes
1. Avoid using buffers containing primary amines (e.g., Tris,

ammonium bicarbonate) since they could interfere with the
labeling and enrichment step.
2. The incubation times and enzyme/proteome ratio are based
on previously established data for the specific test protease. If
not available, monitor digestion of proteins at increasing time
points by SDS-PAGE to determine the optimal conditions.
3. The TMT reagent tubes as supplied by the manufacturer con-
tain 0.8 mg total reagent. Resuspend the content of each tube
in 400 μL ACN, divide into four tubes containing 100 μL each,
and dry using a SpeedVac instrument, thereby obtaining
200 μg TMT reagent per vial. These vials should be stored at
20 C. Since TMT reagents rapidly hydrolyze in aqueous
solutions, they should be resuspended in 100% DMSO imme-
diately before use. Each aliquot containing 200 μg of TMT is
sufficient for labeling 50 μg of protein.
4. The actual amount of the HPG-ALD used in the experiment

depends on concentration and binding capacity specified by the
manufacturer.
5. Parameters for all other nodes work well with default settings
but might be adjusted to fine-tune analysis according to the
Proteome Discoverer™ 2.2 manual.
6. Lysine labeling by isobaric tags precludes trypsin cleavage;
therefore trypsin adopts ArgC specificity. By setting the diges-
tion enzyme as trypsin and by setting lysine labeling as variable
modification (variable TMT6plex lysine (+229.163 Da), it is
possible to evaluate the efficiency of labeling with TMT.
References
1. Hartmann EM, Armengaud J (2014) analysis of MMP-2 and MMP-9 substrate
N-terminomics and proteogenomics, getting degradomes by iTRAQ-TAILS quantitative
off to a good start. Proteomics 14:2637–2646 proteomics. Mol Cell Proteomics 9:894–911
2. Lai ZW, Petrera A, Schilling O (2015) Protein 10. Kleifeld O, Doucet A, auf dem Keller U,
amino-terminal modifications and proteomic Prudova A, Schilling O, Kainthan RK et al
approaches for N-terminal profiling. Curr (2010) Isotopic labeling of terminal amines in
Opin Chem Biol 24:71–79 complex samples identifies protein N-termini
3. Yeom J, Ju S, Choi Y, Paek E, Lee C (2017) and protease cleavage products. Nat Biotech-
Comprehensive analysis of human protein nol 28:281–288
N-termini enables assessment of various pro- 11. Cox J, Mann M (2008) MaxQuant enables
tein forms. Sci Rep 7:6599 high peptide identification rates, individualized
4. Kleifeld O, Doucet A, Prudova A, auf dem p.p.b.-range mass accuracies and proteome-
Keller U, Gioia M, Kizhakkedathu JN, Overall wide protein quantification. Nat Biotechnol
CM (2011) Identifying and quantifying pro- 26:1367–1372
teolytic events and the natural N terminome by 12. Deutsch EW, Mendoza L, Shteynberg D,
terminal amine isotopic labeling of substrates. Farrah T, Lam H, Tasman N et al (2010) A
Nat Protoc 6:1578–1611 guided tour of the trans-proteomic pipeline.
5. Marino G, Eckhard U, Overall CM (2015) Proteomics 10:1150–1159
Protein termini and their modifications 13. Craig R, Beavis RC (2004) TANDEM: match-
revealed by positional proteomics. ACS Chem ing proteins with tandem mass spectra. Bioin-
Biol 10:1754–1764 formatics 20:1466–1467
6. van den Berg BHJ, Tholey A (2012) Mass 14. auf dem Keller U, Prudova A, Gioia M, Butler
spectrometry-based proteomics strategies for GS, Overall CM (2010) A statistics-based plat-
protease cleavage site identification. Proteo- form for quantitative N-terminome analysis
mics 12:516–529 and identification of protease cleavage pro-
7. Rogers LD, Overall CM (2013) Proteolytic ducts. Mol Cell Proteomics 9:912–927
post-translational modification of proteins: 15. auf dem Keller U, Overall CM (2012) CLIP-
proteomic tools and methodology. Mol Cell PER: an add-on to the trans-proteomic pipe-
Proteomics 12:3532–3542 line for the automated analysis of TAILS
8. Schlage P, Egli FE, Nanni P, Wang LW, Kiz- N-terminomics data. Biol Chem
hakkedathu JN, Apte SS, auf dem Keller U 393:1477–1483
(2014) Time-resolved analysis of the matrix 16. Videm P, Gunasekaran D, Schröder B, Mayer B,
metalloproteinase 10 substrate degradome. Biniossek ML, Schilling O (2014) Automated
Mol Cell Proteomics 13:580–593 peptide mapping and protein-topographical
9. Prudova A, auf dem Keller U, Butler GS, Over- annotation of proteomics data. BMC Bioinfor-
all CM (2010) Multiplex N-terminome matics 15:207
Part III
Imaging Approaches for Collagen

Chapter 9
Sample Preparation of Extracellular Matrix of Murine

Colons for Scanning Electron Microscopy
Elee Shimshoni and Irit Sagi
Abstract
Scanning electron microscopy is a useful tool for high-resolution morphological characterization of the
extracellular matrix. In certain tissues and surfaces, imaging of the extracellular matrix requires cell removal
before sample preparation. In this protocol, we will describe a method for preparing extracellular matrices
derived from murine colon for imaging under a scanning electron microscope.
Key words Tissue, Colon, Scanning electron microscope, Extracellular matrix, Decellularization
1 Introduction
Scanning electron microscopy (SEM) is a tool commonly used for

obtaining nanometer spatial resolution images of the topography of
a wide range of biological specimens, including microorganisms,
cells, and tissues. It is a useful tool for studying how morphology
and topography of a tissue changes due to disease. Samples imaged
under SEM are subjected to high vacuum conditions and must be
fixed, completely dry, conductive, and electrically grounded [1, 2].
The structure and organization of collagen and the other extracel-
lular matrix (ECM) proteins in a tissue can be visualized at high
resolution using SEM [3, 4]. In certain tissue surfaces, such as the
gut mucosa, imaging the ECM is difficult since it is not visible
beyond the cells and therefore would require further slicing of the
sample. For this reason, in order to image the colonic ECM using
SEM, we need to first remove the cells from the tissue via decel-
lularization. Figure 1 demonstrates how the colon tissue looks with
and without decellularization. There are various decellularization
methods [5], and we chose the method described herein, which
includes freeze-thaw cycles followed by ammonium hydroxide
treatment, to be the best for structure preservation. In this proto-
col, we will describe a method for preparing extracellular matrices
129
130 Elee Shimshoni and Irit Sagi
Fig. 1 Scanning electron micrographs of a murine colon with or without cells colon samples were either
prepared as described in this protocol (image on the right), or were immediately fixed after harvest and stained
with an additional dye that binds to cell membranes (OsO4). Both images were taken in a 2500 magnifica-
tion. Images demonstrate the mucosal surface with and without cells, displaying the colonic crypts. As evident
by the images, the extracellular matrix scaffold underlying the gut epithelium that can be observed in the
decellularized sample is covered by cells in the sample presented on the left
derived from murine colon for imaging of the under SEM using the
following processes – decellularization, fixation, staining, dehydra-
tion, and mounting.
2 Materials
Use double-distilled water for all solutions.
2.1 Decellularization 1. Tools: scalpel and forceps.

2. Phosphate buffered saline (PBS).
3. Double-distilled water (ddH2O).
4. 25 mM ammonium hydroxide (NH4OH) in water. Store in
4 C.
2.2 Fixation and 1. Fixation solution: 4% paraformaldehyde and 2% glutaraldehyde

Staining in water. Make fresh or store in 20 C until use.
2. ddH2O.
3. Staining solution: 4% sodium silicotungstate (SST).
4. Dehydration solutions: increasing concentrations of EtOH in
water: 30%, 50%, 70%, 96%, and 100% EtOH.
2.3 Sample 1. Aluminum mounts.

Mounting 2. Carbon stickers and tape.
3. Tools: high-precision tweezers, scissors, and curved forceps.
SEM Sample Preparation of ECM 131
3 Methods
3.1 Decellularization 1. Cut 5–10 mm pieces (in length) of fresh or snap-frozen colon.
2. Wash samples twice with PBS, clearing out any residual feces.
3. Wash samples twice with ddH2O.
4. Place samples in an Eppendorf tube with ddH2O.
5. Freeze samples on dry ice (see Note 1) in ddH2O.
6. Thaw at room temperature and replace ddH2O.
7. Repeat steps 4 and 5 five more times (six times in total).
8. Immerse samples in the ammonium hydroxide solution for
20 min in a rotator.
9. Wash six times in ddH2O (see Note 2).
3.2 Fixation Staining 1. Immerse in fixation solution and incubate overnight at 4 C.

and Dehydration 2. Wash twice in ddH2O.
3. Replace water with SST staining solution and incubate for
30 min in motion (see Note 3).
4. Dehydrate samples by immersing them in increasing concen-
trations of EtOH (30%, 50%, and then 70%) and incubate twice
in each solution for 5 min in motion.
5. Keep samples in 70% EtOH at 4 C until the day of critical
point drying (CPD).
6. Continue dehydration in 96% and 100% EtOH as described in
step 4.
7. Subject samples to CPD using the appropriate device.
3.3 Sample Figure 2 illustrates steps 1–5:

Mounting
1. Take aluminum mounts (one for each sample) and stick a
carbon sticker on each one.
2. Gently take each sample out of the CPD gasket using forceps
and place on the sticker on the mount.
3. Cut the colon piece open so that the lumen faces up.
4. Gently press on the edges of the sample to allow them to stick
to the sticker (see Note 4).
5. Carefully stick 2–3 small pieces of carbon tape on the edges,
leaving as much visible sample surface as possible (see Note 5).
6. Coat samples with either gold/palladium alloy or another
metal in a sputtering machine.
7. Always store samples in a desiccator to keep them dry.
132 Elee Shimshoni and Irit Sagi
Fig. 2 Sample mounting after critical point drying Dehydrated colon samples should be mounted as illustrated.
First the colon tube is placed gently on the carbon sticker stuck to the mount and allowed to adhere. Then the
colon is cut open using scissors while trying not to touch the bottom of the sample to not create dents in the
sample. Following, take forceps and gently pull the edges, and stick them to the sticker as well. Finally, stick
small pieces of carbon tape on the edges of the sample to ground the charges that will be interacting with
sample in the microscope
4 Notes
1. For Subheading 3.1, samples can be frozen in the freezer in

80 C. This is useful for continuing the protocol on a differ-
ent day, if desired. Dry ice is more rapid, of course. Avoid using
liquid nitrogen so freezing occurs more gradually.
2. At the final washes of the decellularization process, make sure
the water in the tube is clear, with no remaining debris or feces.
Wash the inside of the colon if necessary using a 200 μl pipette.
3. Steps 3 and 4 of Subheading 3.2 should be done on a shaker, if
transferred to a 24-well plate or on a rotator if samples remain
in Eppendorf tubes.
4. Make sure not to dent the sample too much during mounting,
and leave as much exposed and untouched surface as possible.
5. It is important to secure the edges of the sample with the tape,
both to prevent it from detaching and most importantly to
improve the conductivity of the sample, so the edge on each
side must have a piece of carbon tape on it.
References
1. Hayat MA (2000) Principles and techniques of protocols in microbiology p. Unit 2B.2. John
electron microscopy: biological applications. Wiley & Sons, Hoboken, NJ
Cambridge University Press, Cambridge, UK 3. Brown BN, Barnes CA, Kasick RT et al (2010)
2. Fischer ER, Hansen BT, Nair V et al (2012) Surface characterization of extracellular matrix
Scanning electron microscopy. In: Current scaffolds. Biomaterials 31:428–437
SEM Sample Preparation of ECM 133
4. Lin CP, Douglas WH, Erlandsen SL (1993) 5. Lu H, Hoshiba T, Kawazoe N et al (2012)

Scanning electron microscopy of type I collagen Comparison of decellularization techniques for
at the dentin-enamel junction of human teeth. J preparation of extracellular matrix scaffolds
Histochem Cytochem 41:381–388 derived from three-dimensional cell culture. J
Biomed Mater Res A 100:2507–2516
Chapter 10
In Situ Detection of Degraded and Denatured Collagen via

Triple Helical Hybridization: New Tool in Histopathology
Yang Li and S. Michael Yu
Abstract
Degraded and denatured collagens are useful markers for physiological events (e.g., bone formation and
aging) and pathologic conditions (e.g., cancer, arthritis, and fibrosis). Here we describe histological staining
of such collagens using fluorescent collagen hybridizing peptide that can specifically bind to collagen
strands by folding into triple helix. The method can report the amount of denatured collagen and/or
collagen remodeling activity in tissues via localized fluorescence intensity and can be used in conjunction
with conventional staining agents. The collagen hybridizing peptide probes can be used across species and
collagen types, providing a versatile tool not only for pathology and developmental biology but also
histology-based disease diagnosis, staging, and therapeutic screening.
Key words Matrix metalloproteinase, Collagenase, Collagen fiber, Collagen remodeling, Bone
resorption, Immunostaining
1 Introduction
Collagen is the major structural component of the extracellular

matrix (ECM) in virtually all mammalian tissue and organs, with
an essential role in supporting cell growth and tissue formation.
While collagen synthesis and degradation are delicately coordinated
during tissue homeostasis, collagen undergoes extensive proteolytic
remodeling during development and as a result of a variety of life-
threatening conditions such as cancer, myocardial infarction, and
fibrosis [1–3]. Although the degraded collagens could be an impor-
tant marker of tissue damage and remodeling, it is difficult to detect
such collagens, particularly in histopathology, using conventional
tools, such as collagen stains (e.g., Masson’s trichrome), antibodies,
and second-harmonic generation microscopy. To address this prob-
lem, we developed the collagen hybridizing peptide (CHP) which
can specifically bind to enzymatically degraded collagen molecules
in tissues [4–6]. The peptide was designed based on the triple
helical structure of collagen, which is nearly exclusively found in
135
136 Yang Li and S. Michael Yu
Fig. 1 (a) Schematic of a CHP strand (labeled with X) hybridizing to denatured collagen chains and forming a
collagen triple helix. During disease progression, tissue development, or aging, collagen can be extensively
degraded by collagen-cleaving proteases, causing its triple helix to unfold at the physiological temperature
due to thermal instability. X represents the biotin or fluorescent tag. (b) The thermal activation step used to
generate monomeric CHP strands for collagen hybridization. A dilute CHP solution (15–30 μM) was heated in a
water bath at 80 C for 5 min to dissociate the folded CHP homotrimers. The hot solution was then quenched
in an ice-water bath to room temperature. The solution of monomeric CHP was then ready for immediate use
to hybridize with denatured collagen in tissue samples. The self-trimerizing of CHP takes several hours to
complete in this concentration range, and the effect of the dead time due to quenching (< 3 min) is negligible.
The figure is reproduced from ref. 4 with permission from American Chemical Society
the collagen family (Fig. 1a). Following degradation by collagen-

cleaving proteases (e.g., MMP), the collagen triple helices become
thermally unstable at body temperature and spontaneously unfold,
leaving the denatured collagen fragments covalently cross-linked
within the matrix (Fig. 1a) [5, 7–9]. The CHP peptide has a
collagen-mimetic repeating sequence of glycine(G)–proline
(P)–hydroxyproline(O) with a strong propensity to fold into the
triple helix. Therefore, it can bind specifically to unfolded collagen
chains by forming a new, hybridized triple helical complex, in a
fashion that is similar to a primer binding to a melted DNA strand
Detecting Degraded Collagen via Triple Helical Hybridization 137
during a polymerase chain reaction (Fig. 1a) [4–6]. Due to the lack
of hybridization sites, the CHP has no affinity toward native, fully
folded collagen molecules [5]. Furthermore, the neutral and
hydrophilic amino acid composition of CHP makes it virtually
nonadherent to other biomolecules [10].
In this chapter, we present detailed procedures for detecting
degraded and denatured collagen in histology slides using biotin-
labeled CHP. The key difference between this method and a stan-
dard immunostaining protocol is that the CHP probe requires a
special heating-and-quenching step before application to samples
(see Fig. 1b). Because of the high triple helix propensity, the CHP
peptides tend to self-assemble into homotrimers in solution over
time (e.g., during storage at 4 C) and lose the driving force for
collagen hybridization. Therefore, the trimeric CHP needs to be
thermally dissociated to single strands by heating briefly at 80 C
before it can be used to bind to unfolded collagen [4]. The hot
CHP solution is quenched quickly to room temperature immedi-
ately prior to addition to tissue samples to prevent undesired ther-
mal denaturation of the tissue. Since the CHP trimerization has a
third-order folding rate with a halftime on the order of hours at low
μM concentrations [11, 12], the heating-and-quenching method
(Fig. 1b) described in this protocol ensures that the active mono-
meric form of CHP is delivered to the denatured collagen allowing
efficient hybridization.
2 Materials
2.1 Equipment 1. Unstained animal or human tissue slides (typically 2–20 μm in

and Tools thickness) which are either fixed and embedded in paraffin or
frozen in OCT compound.
2. A humidity chamber for slide staining.
3. Thirteen tissue-slide-staining jars and a vertical slide-
staining rack.
4. Pipettors (1000 μL and 200 μL) and tips.
5. A benchtop timer.
6. 15 mL and 50 mL centrifuge tubes and several 1.5 mL
microtubes.
7. A hydrophobic barrier pen (PAP pen).
8. A temperature-controlled water bath, or heating block, and a
thermometer.
9. An ice bucket or a glass beaker.
10. A floating microtube rack.
11. A benchtop mini-centrifuge for 1.5 mL microtubes.
12. Standard glass coverslips.

13. The fluoroshield mounting medium.
14. A pair of extra-fine point tweezers.
15. A standard fluorescence microscope with common fluores-
cence channels, such as DAPI, GFP, RFP, Cy5, etc.
2.2 Solvents 1. Xylene, 200 mL.

for Prestaining Steps 2. Ethanol, 200 proof, 200 mL.
3. Ethanol, 190 proof, 200 mL.
4. Ethanol, 50%, 200 mL. Add 100 mL of deionized water and
100 mL of 200 proof ethanol to a glass beaker and mix.
5. Phosphate-buffered saline (1 PBS), 2–2.5 L, pH 7.4.
6. Blocking solution: 5 (v/v) % goat serum in 1 PBS, pH 7.4.
Thaw the pre-aliquoted and frozen goat serum. Add 0.5 mL of
10 PBS, 0.5 mL of goat serum, and 9 mL of deionized water
to a 15 mL centrifuge tube. Mix well and store at 4 C (see
Note 1).
7. Biotin-blocking solution kit (optional, see Note 2), containing
a streptavidin reagent (Component A) and a biotin reagent
(Component B), which is usually purchased from major
vendors.
2.3 Staining 1. CHP staining solution. Biotin- or fluorescein-labeled collagen

hybridizing peptide (B-CHP or F-CHP) conjugates can be
purchased from 3Helix, Utah, USA or other vendors. Dissolve
the CHP powder with calculated amount of 1 PBS buffer to
prepare a stock solution of 100 μM concentration. Aliquot a
desired volume of the B-CHP stock solution (see Note 3), and
dilute it to 20 μM (see Note 4) with 1 PBS. Store at 4 C.
2. Diluent solution: 1% bovine serum albumin (BSA) in 1 PBS,
pH 7.4. Weigh 0.1 g of BSA and transfer to a 15 mL centrifuge
tube containing 10 mL of 1 PBS. Mix well and store at 4 C.
3. Streptavidin solution: aliquot a desired volume of the stock
solution of a streptavidin conjugate, and dilute it to 5 μg/mL
concentration with the diluent solution. A variety of commer-
cial avidin or streptavidin conjugates can be used for signal
detection. Here, AlexaFluor647-labeled streptavidin is used as
an example in this protocol. Store the diluted streptavidin
solution at 4 C.
3 Methods
Carry out all procedures at room temperature unless otherwise

specified.
3.1 Slide Preparation 1. Fill two tissue-slide-staining jars with each of the following
and Prestaining Steps solvents: xylene, 200 proof ethanol, 190 proof ethanol, 50%
ethanol, and deionized water. One hundred milliliter of a sol-
vent is added to each jar. Label the jars with their contents.
2. Place the paraffin-embedded slides onto a staining rack, and
immerse the rack in a xylene-containing jar.
3. Incubate the slides for two 5 min cycles in each solvent, follow-
ing a consecutive order of xylene, 200 proof ethanol, 190 proof
ethanol, 50% ethanol, and deionized water. After the final cycle
of washing in water, these deparaffinized slides can be directly
stained by CHP without any antigen-retrieval process. For
OCT-embedded frozen slides (see Note 5).
4. Fill the bottom of the tissue-staining chamber with water.
5. Using the PAP pen, draw a circle around the specimen(s) on
each slide as a hydrophobic barrier to keep staining reagents
localized on the tissue sections. Align the slides on the rack
inside the staining chamber.
6. Add an enough volume of blocking solution to each tissue
sections to fill the space within the barrier circle. To estimate
the volume of solution needed to cover all slides (see Note 4).
Incubate the slides with the blocking solution in the staining
chamber for 20 min to block the non-specific binding sites.
7. Remove the blocking solution by blotting the slides on a paper
towel.
8. Rinse the slides in 100 mL of 1 PBS in a slide-staining jar for
5 min. Rinse three times. (See Note 6, for the optional biotin-
blocking steps.)
9. Blot the slides on a paper towel, and align the slides on the rack
inside the staining chamber. The slides are now ready to be
stained by CHP.
3.2 Collagen 1. Turn on the temperature-controlled water bath, and set the
Hybridizing Peptide temperature to 80 C (see Note 7).
Staining 2. Fill the ice bucket or a glass beaker with ice and water to make
an ice-water bath.
3. Ensure the microtube containing the diluted B-CHP staining
solution is capped tightly. Add the tube onto a floating rack,
and place the rack into the 80 C water bath.
4. Allow the B-CHP staining solution to be heated for 5 min (see
Note 8) to dissociate the trimeric peptides into single strands
(see Fig. 1b).
5. Following heating, immediately vortex and centrifuge the tube
for 10 s to collect the condensation inside the tube.
6. Quickly immerse the B-CHP microtube in the ice-water bath

and wait for 30 s (see Note 9) to quench the solution to room
temperature (see Note 10).
7. Quickly add desired amounts of the heated-and-quenched
B-CHP solutions onto the slides to fill the space inside the
barrier circles (see Note 11).
8. Close the lid of the staining chamber, and carefully move the
chamber into a refrigerator. Incubate the tissue slides in the
chamber at 4 C overnight (see Note 12).
9. After staining, remove the B-CHP solutions by blotting the
slides on a paper towel.
10. Rinse the slides in 100 mL of 1 PBS in a slide-staining jar for
5 min. Rinse three times.
11. Align the slides on the rack inside the staining chamber. Add
solutions of AlexaFluor647-conjugated streptavidin onto the
slides to fill the space inside the barrier circles (see Note 13).
12. Incubate the slides with the streptavidin solutions in the stain-
ing chamber for 1 h.
13. Remove the streptavidin solutions by blotting the slides on a
paper towel. Rinse the slides as in step 10 three times (see Note
14).
14. Add a drop of the fluoroshield mounting medium onto the
center of each tissue specimen. Use a pair of extra-fine point
tweezers to slowly and carefully place a glass coverslip onto
each slide. Ensure that the mounting medium fill the space
between the slide and the coverslip with no air bubbles
trapped. The slides are now ready to be imaged using a stan-
dard fluorescence microscope.
4 Notes
1. We recommend preparing the protein solutions, such as the

blocking, CHP staining, and streptavidin solutions, within 24 h
before usage to ensure freshness.
2. Biotin occurs naturally, functioning as an enzyme cofactor in
the cytosol and mitochondria of a wide variety of cell types.
Kidney, liver, and spleen tissues may contain a high level of
endogenous biotin. Occasionally, when the slides are analyzed
with B-CHP and avidin/streptavidin detection, blocking the
endogenous biotin in the tissue can significantly reduce the
non-specific background signals. This can be done by using a
commercial biotin-blocking kit that is available from most
reagent vendors. A typical biotin-blocking procedure is
described in Note 6.
3. Two CHP conjugate reagents are available for histology stud-

ies: the fluorescein-labeled F-CHP for fluorescence detection
and the biotin-labeled B-CHP for non-green fluorescence and
HRP-based colorimetric detection. B-CHP may sometimes be
advantageous when the tissue auto-fluoresces in green or when
there is a need to enhance signal using enzymatic methods.
This protocol focuses on B-CHP and streptavidin detection.
To perform F-CHP staining, use F-CHP instead of B-CHP
wherever CHP or B-CHP is mentioned in the protocol, and
skip steps 11–13 in Subheading 3.2.
4. Although the optimal concentrations for tissue staining may be
sample dependent, a concentration of 20 μM is recommended
for initial trials, based on our optimization test (see Fig. 2a). In
addition, the total volume of CHP staining solution needed for
each experiment depends on the number of slides and the area
of each tissue section: approximately, it takes 50 μL of solution
to cover an area of 1 cm2 on a slide. A typical tissue slide
requires 50–250 μL of staining solution to cover. The volume
of the CHP stock solution needed in each experiment can be
calculated:
V CHP stock ¼ ðV total volume to cover all slides 20 μM Þ=100 μM
5. For OCT-embedded frozen slides, skip steps 1–3 in Subhead-

ing 3.1. Instead, take out the slides from the freezer, and let
them warm up to room temperature. Fill two staining jars with
1 PBS. Rinse away the OCT compound by incubating the
slides for 5 min in each staining jar.
6. To block endogenous biotin, following step 8 in Subheading
3.1, apply one or two drops of the streptavidin reagent
(Component A, streptavidin) of a biotin-blocking kit to each
tissue sample, and incubate for 20 min at room temperature in
the humidity chamber. Rinse the slides thoroughly with PBS
for 5 min twice in the staining jars. Add one or two drops of the
biotin reagent (Component B, biotin), and incubate 20 min at
room temperature in the humidity chamber. Rinse thoroughly
with PBS twice.
7. To save time, steps 1–3 in Subheading 3.2 can be performed in
parallel of steps 6–8 in Subheading 3.1, while the slides are
being blocked or washed.
8. It is not a problem to heat the CHP solution for over 5 min,
since the CHP peptide is chemically stable at hot temperature.
It is important to ensure that the slides, the centrifuge, and the
ice-water bath are ready before removing the CHP solution
from heating. Turn off the heater after the heating is finished.
Fig. 2 The optimization tests for CHP staining. (a) Fluorescence images of neighboring cryosections of a
porcine ligament stained with 100 μL of F-CHP solutions at varying concentrations. Before staining, all slides
were denatured in 80 C water to purposefully denature the collagen molecules for CHP hybridization. The
staining time was set at 2 h for all samples. (b) Fluorescence images of a set of heated ligament tissue slides
stained with F-CHP (100 μL, 15 μM) for different incubation time periods. Scale bars: 2 mm. The figure is
reproduced from ref. 4 with permission from American Chemical Society
9. The heated CHP solution must be cooled down to room

temperature to avoid thermal damage to the tissue samples.
Depending on the solution volume (e.g., 200–3000 μL), the
quenching time required varies from 15 s to 90 s.
10. When needed, other staining agents, such as a collagen anti-
body, can be diluted into the quenched CHP solution for
Fig. 3 Histological visualization of collagen degradation during endochondral ossification using CHP. (a)
Localization of CHP binding in a sagittal section of an 18 d.p.c. mouse embryo double stained with B-CHP
(detected by AlexaFluor647-streptavidin, orange) and an anti-collagen I antibody (detected by AlexaFluor555-
labeled donkey anti-rabbit IgG H&L, cyan). mx maxilla, md mandibular bone, bp basisphenoid bone, bo
basioccipital bone, vc vertebral column, rb rib, h hipbone, d digital bones. (b) Magnified views of the
basioccipital bone (bo) beside the C1 vertebra (v) (top images) and the manubrium sterni (ms, bottom images)
in the sagittal section of a 17 d.p.c. mouse embryo stained in the same fashion. High level of CHP binding is
found in the hypertrophic zone surrounding the newly deposited collagen I bone matrix (yellow arrow heads),
which is visualized by the anti-collagen I antibody. Scale bars: 3 mm (a), 0.5 mm (b). The figure is reproduced
from ref. 4 with permission from American Chemical Society
co-staining (see Fig. 3, e.g.). Following step 6 in Subheading

3.2, quickly pipette the desired volume of the antibody stock
solution into the quenched CHP staining solution according
to the predetermined dilution factor. Mix well, and this solu-
tion is ready for double staining.
11. The CHP trimerization has a third-order folding rate with a
halftime on the order of hours at low μM concentrations.
Therefore, the heating-and-quenching protocol used here
ensures that the CHP strands are in the active monomer form
until exposed to denatured collagen. However, it is highly
recommended to spend minimal time on steps 5–7 (
3 min) to limit the CHP trimerization during such dead time
and to ensure consistent results.
12. We suggest staining the slides at 4 C, because the CHP
hybridizes with collagen chains more strongly at low tempera-
ture. Based on our optimization test (see Fig. 2b), we
recommend incubating the slides in the CHP solution for at

least 2 h. For optimal results, overnight staining is
recommended.
13. To detect the primary antibody added during the B-CHP
staining step (see Note 10), dilute the labeled secondary anti-
body (e.g., AlexaFluor555 donkey anti-rabbit) into the
AlexaFluor647-streptavidin solution according to the prede-
termined dilution factor (e.g., 1:200), and incubate the slides
with the two probes for 1 h at room temperature.
14. When needed, the slides can be further stained with other
common staining agents (e.g., Hoechst 33342) following the
CHP staining.
Acknowledgments
This work was supported by the National Institutes of Health grant

(ORIP, R43OD021986) awarded to Y.L. We also thank the
National Institutes of Health and the Department of Defense
(NIH, R01AR060484, R21AR065124, and DOD W81XWH-
12-1-0555 awarded to S.M.Y.) for supporting the research of the
CHP technology.
References
1. Bonnans C, Chou J, Werb Z (2014) Remodel- collagen prior to peptide bond hydrolysis.
ling the extracellular matrix in development EMBO J 23:3020–3030
and disease. Nat Rev Mol Cell Biol 8. Danielsen CC (1987) Thermal stability of
15:786–801 human-fibroblast-collagenase-cleavage pro-
2. Page-McCaw A, Ewald AJ, Werb Z (2007) ducts of type-I and type-III collagens. Biochem
Matrix metalloproteinases and the regulation J 247:725–729
of tissue remodelling. Nat Rev Mol Cell Biol 9. Fisher GJ, Kang S, Varani J et al (2002)
8:221–233 Mechanisms of photoaging and chronological
3. Wahyudi H, Reynolds AA, Li Y et al (2016) skin aging. Arch Dermatol 138:1462–1470
Targeting collagen for diagnostic imaging and 10. Li Y, Ho D, Meng H et al (2013) Direct detec-
therapeutic delivery. J Control Release tion of collagenous proteins by fluorescently
240:323–331 labeled collagen mimetic peptides. Bioconjug
4. Hwang J, Huang Y, Burwell TJ et al (2017) In Chem 24:9–16
situ imaging of tissue remodeling with collagen 11. Ackerman MS, Bhate M, Shenoy N et al (1999)
hybridizing peptides. ACS Nano Sequence dependence of the folding of
11:9825–9835 collagen-like peptides. Single amino acids
5. Li Y, Foss CA, Summerfield DD et al (2012) affect the rate of triple-helix nucleation. J Biol
Targeting collagen strands by photo-triggered Chem 274:7668–7673
triple-helix hybridization. Proc Natl Acad Sci U 12. Boudko S, Frank S, Kammerer RA et al (2002)
S A 109:14767–14772 Nucleation and propagation of the collagen
6. Li Y, Yu SM (2013) Targeting and mimicking triple helix in single-chain and trimerized pep-
collagens via triple helical peptide assembly. tides: transition from third to first order kinet-
Curr Opin Chem Biol 17:968–975 ics. J Mol Biol 317:459–470
7. Chung L, Dinakarpandian D, Yoshida N et al
(2004) Collagenase unwinds triple-helical
Chapter 11
Combination of Traction Assays and Multiphoton Imaging

to Quantify Skin Biomechanics
Stéphane Bancelin, Barbara Lynch, Christelle Bonod-Bidaud,
Petr Dokládal, Florence Ruggiero, Jean-Marc Allain,
and Marie-Claire Schanne-Klein
Abstract
An important issue in tissue biomechanics is to decipher the relationship between the mechanical behavior
at macroscopic scale and the organization of the collagen fiber network at microscopic scale. Here, we
present a protocol to combine traction assays with multiphoton microscopy in ex vivo murine skin. This
multiscale approach provides simultaneously the stress/stretch response of a skin biopsy and the collagen
reorganization in the dermis by use of second harmonic generation (SHG) signals and appropriate image
processing.
Key words Biomechanics, Collagen, Skin, Multiphoton microscopy
1 Introduction
The mechanical behavior of skin at macroscopic scale is strongly

related to its microstructure and mainly to the organization of the
collagen fiber network in the dermis. While many microstructural
models have been investigated theoretically, only few experimental
studies have been reported up to now due to technical limitations.
Indeed, it is necessary to monitor the collagen network at a
micrometer scale during mechanical assays in a centimeter-sized
skin biopsy.
To address this issue, we recently combined traction assays with
multiphoton microscopy [1, 2]. This multiscale device measures
simultaneously the stress/stretch response of a skin biopsy and the
collagen reorganization in the dermis by use of second harmonic
generation (SHG) signals. SHG signals correspond to a multipho-
ton mode of contrast that is specific to collagen fibers and efficiently
reveals the micrometer-scale organization of collagen in unstained
thick tissues [3–5]. Accordingly, we obtained new quantitative
145
146 Stéphane Bancelin et al.
information about skin biomechanics that evidenced a non-affine

reorganization of the collagen network and challenged usual micro-
structural models [6, 7]. This new technique also allowed charac-
terizing the impact of microstructural changes in the dermis due to
genetic modifications, aging, or any other processes [2, 8]. Note
that a similar approach has been applied successfully to other tissues
(see, for instance, [9–18]).
In this chapter, we describe the implementation of incremental
uniaxial traction assays with semicontinuous SHG visualization in
ex vivo murine skin. We also explain how to analyze the mechanical
data and process the SHG images to obtain quantitative para-
meters. Specifically, we describe how to use the network of hair
follicles as endogenous tags to quantify the local deformation and
how to extract the fiber orientation to quantify the collagen reor-
ganization in the field of view.
2 Materials
The traction and imaging devices may be either custom-build

devices or commercial ones. They need to have dimensions and
shapes that enable their combination in the same setup. In concrete
terms, the traction device has to be put horizontally in place of the
microscope stage (Fig. 1). The specific features of these devices are
listed below.
2.1 Traction Device 1. Symmetrical uniaxial traction machine using two identical
motors and sensors, one of each side of the skin sample.
2. Motors: Linear actuators. Resolution below 5 μm, accuracy
below 5 μm, strokes above 40 mm, maximal holding force
above 10 N. Speed between 1 and 100 μm/s (or above).
3. Sensors: Capacity at least 8 N. Resolution below 0.01 N. Over-
load should be important (more than 80 N) so as to avoid
breaking the sensors when screwing the sample in the grips.
2.2 Multiphoton 1. Geometry: laser-scanning upright microscope, with two

Setup epi-detection channels, one for SHG and another one for
two-photon excited fluorescence (2PEF).
2. Objective lens: high Numerical Aperture (0.9–1.1 NA) mod-
erate magnification (20 or 25) water-immersion objective
lens (see Note 1).
3. Excitation wavelength: any wavelength in the Ti-Sa range for
SHG signal (see Note 2).
4. Excitation polarization: circular or quasi-circular (see Note 3).
Multiscale Skin Biomechanics 147
Fig. 1 Experimental setup combining multiphoton microscopy and uniaxial traction assay. (a) Schematic
representation and (b) picture. Adapted from Bancelin et al., Sci Rep 2015, doi:https://doi.org/10.1038/
srep17635, published under license CC BY 4.0
5. Excitation power: typically 20–30 mW. Power is adjusted to

obtain a good signal-to-noise ratio in the SHG images and
avoid both saturation and damage to the skin sample.
3 Methods
3.1 Preparation 1. Sacrifice of mice by cervical dislocation at a specific age.

of Murine Skin Biopsy 2. Depilation of the back skin: first, the back of the mouse is
shaved using an electric trimmer, then depilatory cream
(Nair™) is applied for 15 min, and hairs are removed with a
scraper. This last step is repeated two or three times if needed.
3. Tagging of the right foreleg with a dot of black ink to identify
the anterior-posterior axis before cutting the skin with scalpels
or dissecting scissors. Skin is then gently peeled off and placed
in a Petri dish.
4. Removal of the epidermis by incubating the skin 30 min at room
temperature with 3.8% ammonium thiocyanate and scraping the
epidermis under a binocular magnifier (see Note 4).
5. Storage in culture medium (Dulbecco’s Modified Eagle’s
Medium) without phenol red (see Note 5), supplemented
with 50 μg/mL penicillin/streptomycin at 6 C. Medium is
changed every 2 days. Skin biopsies should be used for experi-
ments within 5 days.
3.2 Skin Cutting 1. The skin sample is cut into a dog-bone shape using curved
and Fastening dissection scissors, taking care to hold the sample on the out-
for Traction Assays skirt part (not used in mechanical test) so as to damage the
sample as little as possible (Fig. 2) (see Note 6).
Fig. 2 Preparation of the depilated and de-epidermized skin sample for mechan-
ical test. (a) Schematic representation and dimensions of the dog-bone-shaped
skin sample. (b) Picture before cutting. The orange arrow shows the black spot
indicating the right foreleg. The red dots show the dog-bone shape the skin will
be cut into. Typical size is 30 20 mm2. Adapted from Bancelin et al., Sci Rep
2015, doi:https://doi.org/10.1038/srep17635, published under license CC BY 4.0
2. The dog-shaped skin sample is attached to the traction device

by means of mechanical jaws with rubber joints that pinch the
sample on both sides (see Note 7).
3. The skin dimensions are measured with a digital caliper: thick-
ness (around 1 mm), width at the center (around 8 mm), and
length between the jaws (around 20 mm).
4. The hydration of the skin sample is maintained during cutting
and fastening with drops of culture medium.
5. The traction device with the attached skin sample is installed in
place of the microscope stage, with the papillary dermis facing
the objective lens (Fig. 1). Immersion gel is used to ensure
optical contact with the objective lens and to prevent dehydra-
tion of the skin sample during the experiment. If needed, this
immersion gel is first centrifuged to remove small air bubbles
that could deteriorate image quality. New drops are added
during the experiment if required.
3.3 Traction Assays 1. The motor displacement and the force are measured continu-
Under the Multiphoton ously (every second) during the whole traction assay (Fig. 3).
Microscope 2. SHG imaging is performed continuously while adjusting the
axial position of the objective lens to find the surface of the skin
sample and focus a few μm below.
3. The jaws are moved apart slowly (usually around 0.01 mm s1)
while continuously recording low-resolution SHG images until
Fig. 3 Mechanical data from wild-type skin sample. (a) Experimental timetable showing stepwise stretching of
the skin sample (green dotted line) with continuous force measurement (blue continuous line) and time-lapse
imaging of immobile sample after each displacement step (red arrows). (b) Nominal stress/stretch curve for
the same sample, showing the processed mechanical parameters. Adapted from Bancelin et al., Sci Rep
2015, doi:https://doi.org/10.1038/srep17635, published under license CC BY 4.0
no more vertical displacement is observed (see Note 8). This

position will be referred to as the reference position.
4. The motors are stopped and a first SHG image stack is recorded
using the following typical parameters: 100 kHz pixel rate,
0.5 μm pixel size, 2 μm axial step, and 480 480 50 μm3
stacks, resulting in 5 min acquisition time (see Note 9). 2PEF
signals are acquired simultaneously in a second channel to
visualize remains of hairs or other fluorescent structures. Typi-
cal SHG images are displayed in Fig. 4.
5. The skin sample is stretched by 5% at a stretch rate of 104 s1.
Given the length of the skin sample (20 mm between the jaws),
it corresponds to a typical speed of 0.002 mm s1. SHG
imaging at low resolution is performed continuously during
stretching to follow the region of interest (ROI) imaged in the
previous step. If needed, the axial position of the objective lens
and the lateral position of the traction device are adjusted with
micrometer screws to make sure the same ROI is imaged (see
Note 10).
6. Motors are stopped and a second SHG stack is recorded in the
same ROI as previously and using the same parameters (see
Note 11).
7. Steps 4 and 5 are performed sequentially until the skin sample
breaks, usually around a stretch ratio of 1.5. The total duration
of this incremental traction assay is 3–4 h (15–20 min per step).
3.4 Mechanical Data 1. The global stretch ratio is obtained as the ratio of the length
Processing between the rubber joints of the jaws to the reference length that
is the one measured at the reference position (see Note 12).
Fig. 4 SHG imaging and image processing of ex vivo stretched skin sample. (a) Raw SHG images at 1.0, 1.2,
and 1.4 stretch ratio. The deformation of the hair follicle network reveals the local skin deformation. (b)
Processed SHG images at the same stretch ratios highlighting fiber orientation for each pixel with a color code.
Adapted from Bancelin et al., Sci Rep 2015, doi:https://doi.org/10.1038/srep17635, published under license
CC BY 4.0
2. The nominal stress is obtained as the measured force divided by

the initial skin section (thickness width) (see Note 12).
3. The nominal stress is plotted as a function of the stretch ratio,
and four quantitative parameters are extracted (see Fig. 3b):
(1) the tangent modulus, which is the slope of the linear part;
(2) the length of the heel region; (3) the ultimate tensile
stretch, which is the maximum stress before rupture; and
(4) the failure stretch ratio, which is the stretch corresponding
to the ultimate tensile strength (see Note 13).
3.5 Image 1. SHG images at approximately the same depth are cropped to
Processing for Local obtain a ROI common to all stretching steps showing the same
Deformation Analysis hair follicles (round or elliptical regions with no SHG, inter-
rupting the collagen fiber network, Fig. 4). The size of this
ROI increases in the loading direction at increasing stretch
ratio.
2. The hair follicles are segmented using the following steps
(Fig. 5): application of a median filter to reduce noise (2 pixel
radius), inversion of contrast, application of circular opening
(8 pixel radius) [19] to smooth the image while preserving
follicle edges, thresholding resulting in a binary image, and
shape detection to identify and label every follicle. If needed,
Fig. 5 Hair follicle segmentation in the SHG image for local deformation measurement. (a) Raw and (b)
enhanced SHG image of mice skin papillary dermis; (c) binary SHG image after thresholding; (d) position
detection and numbering of hair follicles; mapping of (e) λxx, (f) λyy, and (g) ω tensorial components calculated
from the deformation of the hair follicle network—see green triangle in (d)—compared to the initial network
before stretching. Scale bar: 50 μm. Reproduced from Bancelin et al., Sci Rep 2015, doi:https://doi.org/10.
1038/srep17635, published under license CC BY 4.0
the follicles are renumbered to obtain the same order in all

images.
3. The network formed by the follicles is used to calculate the
local deformation by performing a Delaunay triangulation and
calculating for each triangle the deformation tensor relative to
the same triangle in the non-stretched state (Fig. 5).
4. This local deformation tensor map is averaged over the cropped
SHG image to obtain three parameters: the local deformations
in the directions parallel and perpendicular to the traction and
the sliding angle that represents the shear in the ROI (Fig. 5).
5. The local deformation in the traction direction is compared to
the global one applied to the tissue: the local and global stretch
ratios have to be equal to validate the mechanical assay
(no slipping within the jaws, homogeneous tissue response)
(see Note 14).
3.6 Image 1. A fixed number of SHG images (for instance, 6) with maximum
Processing average intensity are selected in every SHG stack for all stretch-
for Collagen ing ratios.
Reorganization 2. Morphological filtering is applied to extract the fiber orienta-
tion in every pixel by means of a mean filter (3 pixel diameter)
and of a morphological opening [19] using a linear structuring
element (Strel; see Note 15). A stack of images is obtained by
rotating the Strel where the pixelwise maximum provides an

enhanced image of the collagen fibers. The position of the
maximum provides their local orientation (Fig. 4).
3. A normalized histogram of fiber orientation is calculated for
every stretch ratio. The evolution of these histograms with
increasing stretch ratio can be quantified using different meth-
ods. Our choice is to calculate [20] (1) the maximum of this
histogram (i.e., the main orientation), (2) the orientation
index, defined as the percentage of fibers oriented along the
main orientation:
" R 90 #
90 I ðθÞ cos 2 ðθ θmax Þdθ
OI ¼ 100 2 R 90 1
90 I ðθÞdθ
and (3) the entropy, defined as the usual statistical entropy, which
measures the degree of organization in the skin sample indepen-
dently of any main orientation:
90∘
X
S¼ pðθÞ ln ½pðθÞ
θ¼90∘
I ðθÞ
where pðθÞ ¼ P90 .
θ¼90 I ðθÞ
4 Notes
1. We need a good resolution, in order to visualize small features,

and a large field of view, in order to detect more signal and to
visualize a significant region of the tissue. The microscope
resolution depends on the NA and on the excitation wave-
length, while the field of view depends on the magnification
[21]. Consequently, we have to use an objective lens with large
NA and moderate magnification.
2. The SHG signal is a non-resonant signal whose intensity does
not vary with the excitation wavelength. Nevertheless, it is
necessary to choose a specific excitation wavelength in the
case of simultaneous detection of specific fluorescent signals.
3. The SHG signal is sensitive to the orientation of the excitation
electric field compared to the orientation of the collagen fibers:
The SHG signal is smaller when exciting with a linear polariza-
tion perpendicular to the fiber. Circular polarization must be
used to mitigate this effect and visualize all the fibers in the
imaging plane with similar efficiency. It is usually obtained by
use of a quarter-wave plate. The resulting polarization is often
only quasi-circular, which is sufficient, because of the ellipticity
introduced by the various optical components in the microscope.
4. The epidermis is removed because of the presence of pigmen-

ted cells. These cells exhibit strong absorption and may there-
fore induce thermal photodamage during multiphoton
imaging. Moreover, the epidermis is a little damaged by the
depilation process, so that there is no reason to keep it for
imaging.
5. It is highly preferable to use culture medium without phenol
red because this dye penetrates into the tissue and exhibits
two-photon absorption of the laser excitation during multi-
photon imaging. This absorption process may result in tissue
heating and eventually damaging.
6. The dog-bone shape is the usual shape for tensile tests because
it ensures homogeneous repartition of the strain in the central
region of the skin sample and avoids rupture near the jaws
(Fig. 2).
7. The skin is pressed on a rubber joint to limit damage and
slippage. Accordingly, when removed after the experiment, it
is thinned down on the joint line but not broken.
8. This step aims to unfold the skin sample, which results in
vertical movements. We make sure that the force does not
exceed the sensor noise in this step.
9. The acquisition parameters are a compromise between acquisi-
tion time, field of view, stack depth, and voxel size. We need a
reasonable acquisition time to be able to perform all the steps
of the traction assay within a few hours and ensure that the skin
biopsy is still in good conditions at the end of the assay. We
need a wide field of view to visualize a representative part of the
tissue. We need a small pixel size appropriate to visualize colla-
gen fibers. Finally, the imaging depth is usually limited to
50–100 μm because the imaging quality is deteriorated by
scattering and optical aberrations due to undulations at the
surface of the sample (which get worse at higher stretch ratios).
10. It is possible to recognize the ROI by looking at the character-
istic patterns in hair follicles on the SHG image (Fig. 3). The
position of this ROI moves slightly under the objective lens
during stretching, and manual readjustment is necessary to
follow the same ROI. Axial movements are compensated by
adjusting the axial position of the objective lens. Lateral move-
ments are compensated by manually adjusting the position of
the traction device with micrometer screws (the traction device
is placed on a breadboard fixed on micrometric stages). This
manual adjustment is possible only for slow and small displace-
ments, which is the reason why we deliberately choose a slow
strain rate.
11. It is necessary to stop the motors when recording the SHG
image stack because the skin must be immobile during the
image acquisition (which is quite slow as multiphoton micros-

copy is a pixel-based imaging technique). As a consequence,
the skin relaxes every time the motors are stopped (Fig. 3). This
relaxation may be analyzed to extract relaxation mechanical
data [6].
12. The absolute stretch ratio is not determined accurately because
of the uncertainty in the reference length measurement. Nev-
ertheless, the variations of this parameter are accurately deter-
mined. The same comment applies to the nominal stress
because of the uncertainty in the skin thickness measurement.
13. The heel region is considered to start when the stress stays over
twice the noise for at least 10 s (the noise is measured at very
small strains, in the so-called toe region). The end of the heel
region is defined as the point at which the stress gets close
enough to the linear fit (the difference is smaller than twice the
noise).
14. Other parameters can be processed from the segmented image
of follicles: The deformation of the follicles shape from round
to circular and the variation of the skin porosity (% of skin
surface occupied by follicles) [2].
15. The length of the linear structuring element (Strel) is typically
10 μm. A long Strel enables fine angular steps, but is not able to
fit curved fibers.
Acknowledgments
This work was supported by grants from Ecole Polytechnique

(interdisciplinary project) and from Agence Nationale de la
Recherche (contracts ANR-10-INBS-04 France BioImaging,
ANR-11-EQPX-0029 Morphoscope2, and ANR-13-BS09-0004-
02 Metis). The authors thank Vincent de Greef for his help in the
technical implantation of the mechanical setup.
References
1. Goulam Houssen Y, Gusachenko I, Schanne- 3. Zipfel WR, Williams RM, Christie R, Nikitin
Klein M-C, Allain J-M (2011) Monitoring AY, Hyman BT, Webb WW (2003) Live tissue
micrometer-scale collagen organization in intrinsic emission microscopy using
rat-tail tendon upon mechanical strain using multiphoton-excited native fluorescence and
second harmonic generation microscopy. J Bio- second harmonic generation. Proc Natl Acad
mech 44:2047–2052 Sci U S A 100:7075–7080
2. Bancelin S, Lynch B, Bonod-Bidaud C, 4. Strupler M, Pena A-M, Hernest M, Tharaux
Ducourthial G, Psilodimitrakopoulos S, P-L, Martin J-L, Beaurepaire E, Schanne-
Dokladal P, Allain J-M, Schanne-Klein M-C, Klein M-C (2007) Second harmonic imaging
Ruggiero F (2015) Ex vivo multiscale quanti- and scoring of collagen in fibrotic tissues. Opt
tation of skin biomechanics in wild-type and Express 15(7):4054–4065
genetically-modified mice using multiphoton
microscopy. Sci Rep 5:17635
5. Chen XY, Nadiarynkh O, Plotnikov S, Cam- 13. Mauri A, Ehret AE, Perrini M, Maake C,
pagnola PJ (2012) Second harmonic genera- Ochsenbein-Kolble N, Ehrbar M, Oyen ML,
tion microscopy for quantitative analysis of Mazza E (2015) Deformation mechanisms of
collagen fibrillar structure. Nat Protoc 7 human amnion: quantitative studies based on
(4):654–669 second harmonic generation microscopy. J Bio-
6. Lynch B, Bancelin S, Bonod-Bidaud C, Gues- mech 48(9):1606–1613
quin J-B, Ruggiero F, Schanne-Klein M-C, 14. Alavi SH, Sinha A, Steward E, Milliken JC,
Allain J-M (2017) A novel microstructural Kheradvar A (2015) Load-dependent extracel-
interpretation for the biomechanics of mouse lular matrix organization in atrioventricular
skin derived from multiscale characterization. heart valves: differences and similarities. Am J
Acta Biomater 50:302–311 Physiol Heart Circ Physiol 309(2):
7. Jayyosi C, Affagard J-S, Ducourthial G, H276–H284
Bonod-Bidaud C, Lynch B, Bancelin S, 15. Nesbitt S, Scott W, Macione J, Kotha S (2015)
Ruggiero F, Schanne-Klein M-C, Allain J-M, Collagen fibrils in skin orient in the direction of
Bruyère-Garnier K, Coret M (2017) Affine applied uniaxial load in proportion to stress
kinematics in planar fibrous connective tissues: while exhibiting differential strains around
an experimental investigation. Biomech Model hair follicles. Materials 8(4):1841–1857
Mechanobiol 16(4):1459–1473 16. Caulk AW, Nepiyushchikh ZV, Shaw R, Dixon
8. Lynch B, Bonod-Bidaud C, Ducourthial G, JB, Gleason RL (2015) Quantification of the
Affagard J-S, Bancelin S, passive and active biaxial mechanical behaviour
Psilodimitrakopoulos S, Ruggiero F, Allain and microstructural organization of rat tho-
J-M, Schanne-Klein M-C (2017) How aging racic ducts. J R Soc Interface 12(108). ARTN
impacts skin biomechanics: a multiscale study 20150280
in mice. Sci Rep 7:13750 17. Benoit A, Latour G, Schanne-Klein M-C,
9. Sinclair EB, Andarawis-Puri N, Ros SJ, Laudier Allain J-M (2016) Simultaneous microstruc-
DM, Jepsen KJ, Hausman MR (2012) Relating tural and mechanical characterization of
applied strain to the type and severity of struc- human corneas at increasing pressure. J Mech
tural damage in the rat median nerve using Behav Biomed Mater 60:93–105
second harmonic generation microscopy. Mus- 18. Krasny W, Morin C, Magoariec H, Avril S
cle Nerve 46(6):899–907 (2017) A comprehensive study of layer-specific
10. Wentzell S, Nesbitt RS, Macione J, Kotha S morphological changes in the microstructure
(2013) Measuring strain using digital image of carotid arteries under uniaxial load. Acta
correlation of second harmonic generation Biomater 57:342–351
images. J Biomech 46(12):2032–2038 19. Serra J (1982) Analysis and mathematical mor-
11. Chow MJ, Turcotte R, Lin CP, Zhang YH phology. Academic Press
(2014) Arterial extracellular matrix: a Mechan- 20. Bayan C, Levitt JM, Miller E, Kaplan D, Geor-
obiological study of the contributions and gakoudi I (2009) Fully automated, quantita-
interactions of elastin and collagen. Biophys J tive, noninvasive assessment of collagen fiber
106(12):2684–2692 content and organization in thick collagen
12. Sigal IA, Grimm JL, Jan NJ, Reid K, Minckler gels. J Appl Phys 105(10):102042
DS, Brown DJ (2014) Eye-specific IOP-in- 21. Zipfel WR, Williams RM, Webb WW (2003)
duced displacements and deformations of Nonlinear magic: multiphoton microscopy in
human Lamina Cribrosa. Invest Ophth Vis Sci the biosciences. Nat Biotech 21
55(1):1–15 (11):1369–1377
Chapter 12
Analysis of Collagen-Binding Integrin Interactions

with Supramolecular Aggregates of the Extracellular Matrix
Uwe Hansen
Abstract
Integrin-mediated interactions of cells with the extracellular matrix (ECM) are important for their activ-
ities. The ECM itself is a complex network of macromolecules forming aggregates or suprastructures.
Moreover, the molecular composition is important for the macromolecular organization and, thereby, the
functional properties of the ECM. In addition, collagen molecules lose their integrin-binding capabilities
after incorporation into fibrils. Therefore, we have established detailed protocols for the analysis of integrin-
matrix interactions at the supramolecular level.
Key words Cell adhesion assay, Integrins, Matrix superstructures, Collagens, Chondrocytes, Trans-
mission electron microscopy
1 Introduction
Integrin-mediated interactions of cells with the extracellular matrix

(ECM) are vitally important in the control and modulation of their
activities, e.g., division, differentiation, survival, metabolism, and
programmed cell death [1, 2]. However, the ECM is a complex
network of macromolecules forming aggregates or suprastructures.
In vivo, ECM suprastructures are always heterotypically assembled
as they predominantly consist of several molecular components
combined in various quantities. The molecular composition is
important for their macromolecular organization and, thereby,
their functional properties within the ECM. During collagen fibril
formation, minor ECM components are crucial for the overall
aggregation of collagen monomers [3]. Consequently, native colla-
gen fibrils are known as macromolecular alloys. Their function in
the ECM depends strongly on the composition of both collagen
and non-collagenous components, and notably, the recognition of
157
158 Uwe Hansen
collagens by integrins is strongly influenced by the suprastructural

organization of the ECM [4, 5]. In most studies of integrin-matrix
interactions, monomolecular ligands have been employed; how-
ever, integrins in situ encounter extracellular matrix aggregates in
which integrin-binding epitopes can be newly created as well as
obliterated. We have demonstrated the discrepancy between integ-
rin sensing of molecular collagens and aggregated collagens by
showing that collagen molecules lose their integrin-binding cap-
abilities after incorporation into fibrils. Nevertheless, these
collagen-binding integrins interact with non-collagenous macro-
molecules of the fibril periphery [5], and so it is therefore crucial to
define integrin-matrix interactions at the supramolecular level.
Here we present established detailed protocols for the analysis of
interactions between collagen-binding integrins and ECM compo-
nents at different stages of aggregation and for the interaction of
collagen-binding integrins of chondrocytes with collagens in their
monomeric and fibrillar states of aggregation.
2 Materials
1. Chondrocytes: Primary chondrocytes were obtained from ster-

nal cartilage of 17-day-old chicken embryos.
2. Collagenase B: 1 mg/ml collagenase B.
3. Collagenase buffer: 1 mM cysteine, 100 U/ml penicillin,
100 μg/ml streptomycin (sterile filtered) in DMEM.
4. Integrins: 25 μg/ml soluble recombinant human integrins
α1β1, α2β1, and α10β1, in which transmembrane and intracel-
lular domains were replaced by a Jun-/fos-zipper motif or the
pUC-HMT-β1jun-construct, respectively, were prepared as
described previously [6, 7].
5. Anti-integrin-β1-antibodies (9EG7): 30 μg/ml of the activat-
ing 9EG7 antibody.
6. I-domains: Both human and chicken GST-α1I and GST-α2I
were produced and purified using glutathione affinity chroma-
tography as described previously [8, 9].
7. Discoidin domain receptor (DDR): Recombinant soluble pro-
tein comprising the entire DDR2 ectodomain fused to the
Fc-sequence of human IgG2 was produced in episomally trans-
fected HEK293-EBNA cells and was purified by affinity chro-
matography as previously described [10, 11].
8. Penicillin: 100 U/ml penicillin.
9. Streptomycin: 100 μg/ml streptomycin.
Integrin Interaction with Supramolecular Aggregates 159
10. Collagens:
Collagens II, IX, and XI in their native molar proportions of
ƒII,IX,XI ¼ 8:1:1.
Purified monomeric cartilage collagens II, IX, or XI.
Authentic fibrils fragments from cartilage of 17-day-old
chicken embryos or adult chicken, juvenile or adult bovine
knee joint cartilage or human osteoarthritic hip cartilage.
11. Tris buffer:
50 mM Tris–HCl, 150 mM NaCl, pH 7.4 (for collagen
purification).
50 mM Tris–HCl, 400 mM NaCl, pH 7.4 (for in-vitro
fibrillogenesis).
12. HEPES buffer: 0.1 M HEPES, 0.4 M NaCl, pH 7.4.
13. Staining wash buffer: 150 mM NaCl, 2 mM sodium phos-
phate, pH 7.4.
14. Adhesion blocking buffer: 2 mM Tris–HCl buffer, 150 mM
NaCl, pH 7.4.
15. Collagen reconstitute buffer: 50 mM HEPES, 200 mM NaCl,
pH 7.4.
3 Methods
3.1 Primary 1. Primary chondrocytes were obtained from sternal cartilage of

Chondrocyte 17-day-old chicken embryos [12].
Isolation (See Note 1) 2. After removal of the perichondria, the dissected sterna were
washed three times in Krebs-Ringer buffer.
3. Chondrocytes were released from their extracellular matrix by
overnight treatment with 1 mg/ml collagenase B in collage-
nase buffer at 37 C and 5% CO2.
4. Sedimented cells were resuspended and separated from undi-
gested matrix by filtration through a triple layer of 100 μm pore
size nylon membranes in a Swinnex filter.
5. Residual collagenase was removed by two gentle centrifugation
steps at 600 g for 7 min with substitution of the supernatant
with fresh Krebs-Ringer buffer.
6. After a further centrifugation step, the cell pellet was resus-
pended in DMEM with 100 U/ml penicillin and 100 μg/ml
streptomycin (sterile filtered) and quantified using a Neubauer
counting chamber.
3.2 Collagen Mixtures of cartilage collagens II, IX, and XI in their native molar
Purification proportions ƒII,IX,XI ¼ 8:1:1 were isolated from chicken embryo
chondrocytes cultured in suspension of agarose as described in
160 Uwe Hansen
[13]. The separation of the three collagen types occurred by

ion-exchange chromatography as described in [3].
Authentic fibril fragments were obtained from sternal cartilage
of 17-day-old chicken embryos or adult chicken, juvenile and adult
bovine knee joint cartilage, and human osteoarthritic hip cartilage
[14] as follows:
1. Homogenize small cartilage samples for 20 s on ice in 500 μl of
Tris buffer (three times).
2. Short centrifugation step at 2500 g for 3 min.
3. Collect the supernatant and homogenize the pellet again.
4. Repeat this procedure three times.
5. Combine all supernatants for use in experiments.
3.3 Cell Adhesion 1. Use non-tissue grade, 96-well microtiter plates.

Assay 2. Before coating, mark the bottom of each well using a pen as a
point of reference for microscopic images before and after the
mechanical removal of non-adherent cells.
3. Experiments with monomeric collagens: Coat cell culture dishes
with 100 μl/well of solutions of collagens II, IX, and XI at
molar fractions ƒII,IX,XI ¼ 8:1:1 (total collagen concentration:
150 μg/ml) in HEPES buffer overnight at 4 C (see Notes
2 and 3).
4. Experiments with collagen fibrils: Induce collagen fibril forma-
tion in the wells by diluting 50 μl of 300 μg/ml total concen-
tration of monomeric collagens II, IX, and XI dissolved in
HEPES buffer with 50 μl of distilled water and incubate for
3 h at 37 C (see Notes 2 and 4).
5. Fibronectin coating as positive control: Add 100 μl of HEPES
buffer, containing 20 μg/ml of fibronectin to the wells.
6. Complete all coatings overnight at 4 C.
7. Remove unbound substrate proteins by washing twice with
0.1% BSA in DMEM.
8. Block unspecific binding sites with 0.5% BSA in DMEM for
60 min.
9. Further washing step with 0.1% BSA in DMEM.
10. Optional second blocking step with 10 μg/ml recombinant
DDR2-receptors in adhesion blocking buffer (in control
experiments without DDR2) for 1 h (blocking of potential
binding sides for DDR2-receptors of chondrocytes on
collagens).
11. Washing of the wells twice with 0.1% BSA in DMEM.
12. Add 7.5 104 cells/100 μl freshly isolated cranial chondro-

cytes diluted in DMEM supplemented with penicillin/
streptavidin.
13. Attachment of the cells to the substrates for 45 min (short-
term adhesion assay) (see Note 5).
14. Take photographs of each well with an inverse microscope
(phase-contrast (pH 1), ten-fold magnification, point of refer-
ence should be visible).
15. Remove unbound cells by gentle agitation on a Vortex shaker
at 800 rpm for 15 s.
16. Wash two times with DMEM.
17. Add 100 μl DMEM to the wells and take photographs of the
previously selected areas.
18. For analysis, merge the corresponding microscopic photo-
graphs in silico and count the cells in the overlapping areas.
19. Quantify bound cells by direct counting in at least five ran-
domly selected representative fields.
20. Pool the results from three independent experiments.
21. Determine means and standard deviations with standard
algorithms.
3.4 Immunoelectron 1. Spot 15 μl aliquots of authentic fibril fragments, reconstituted

Microscopy fibrils, or monomeric collagens onto sheets of Parafilm.
2. Absorption of aggregates on 200 square mesh nickel grids
coated with formvar/carbon by floating on the drops for
15 min.
3. Wash grids with staining wash buffer.
4. Block for 60 min with 2% dried skimmed milk in PBS.
5. Treatment of the grids with PBS, containing 2 mM MgCl2 and
375 ng of soluble recombinant human integrins α1β1, α2β1,
and α10β1 and 450 ng of the activating antibody 9EG7 [15]
for 60 min. In controls: replace MgCl2 by 10 mM EDTA.
6. Washing of grids with PBS.
7. Incubate the grids with a suspension of 18 nm colloidal gold
particles coated with goat antibodies against rat immunoglo-
bulins in PBS, containing 0.2% dried skimmed milk for 60 min.
8. Wash grids five times with distilled water (removal of all
remaining phosphate buffer contaminants).
9. Negative staining of the grids with 2% uranyl acetate in distilled
water for 15 min.
10. Dip grids briefly into a drop of distilled water.
11. Dry grids with filter paper.
12. Analysis by transmission electron microscopy (Fig. 1).
162 Uwe Hansen
Fig. 1 Electron micrographs of ultrastructural immuno-localization of the binding of soluble integrins to

isolated authentic cartilage matrices. After blocking of unspecific binding sites, recombinant integrins were
added and detected by anti-integrin-β1-antibodies (9EG7). The binding of integrins to cartilage fibrils with
peripheral components is visualized by immuno-gold transmission electron microscopy. Scale bar ¼ 200 nm
3.5 Binding 1. Coat grids at room temperature for 15 min with purified
Experiments on Grids monomeric cartilage collagens II, IX, or XI in HEPES buffer
for Transmission in order to prevent fibril formation.
Electron Microscopy 2. Wash grids with PBS.
3.5.1 Purified Monomeric 3. Block for 60 min with 2% dried skimmed milk in PBS.
Cartilage Collagens II, IX, 4. Incubate with 375 μg/ml of soluble recombinant α1β1-, α2β1-,
or XI α10β1-integrins in the presence of the integrin-activating anti-
body 9EG7 and 2 mM MgCl2 or 10 mM EDTA (negative
control).
5. Wash with PBS.
6. Expose the grids to immuno-gold particles coated with anti-
IgG antibodies.
7. Stain samples with 2% uranyl acetate for 15 min.
8. Take electron micrographs from five randomly selected fields.
9. Count gold particles/field.
3.5.2 Reconstituted 1. Reconstitute collagen fibrils from monomeric collagens II, IX,
Collagen Fibrils Containing and XI (300 μg/ml total concentration) as adhesion substrates
Collagen Types II, IX, and XI by adjusting the buffer salinity to collagen reconstitute buffer
and warm to 37 C as described above (see Note 6).
2. Apply 10 μl of 150 μg/ml total concentration of reconstituted
collagen fibril suspension onto the surface of 400 square mesh
uncoated nickel grids.
3. Wash the attached collagen fibrils five times with PBS in order
to remove remaining collagen monomers (see Note 7).
4. Block unspecific binding sites with 2% dried skimmed milk
in PBS.
5. Incubate the reconstituted collagen fibrils with integrins and
primary and secondary antibodies as described under 3.4.
6. Coat the grids with a thin layer of carbon generated on mica
plates and float on distilled water. Put the grid under the
carbon layer floating on water surface and take out both the
grid and the carbon layer.
7. Dry the grid carefully on filter paper.
8. Stain the samples with 2% (v/w) uranyl acetate.
9. Analyze samples by transmission electron microscopy.
3.5.3 Binding 1. Coat grids with 10 μl purified monomeric cartilage collagens

Experiments with Solutions II, IX, or XI in HEPES buffer or 10 μl reconstituted collagen
of 25 μg/ml Recombinant, fibril suspension from monomeric collagens II, IX, and XI at
GST-Tagged Chicken α1I- room temperature for 15 min.
and α2I-Domains 2. Wash grids with PBS.
3. Block for 60 min with 2% dried skimmed milk in PBS.
4. Add solutions of 25 μg/ml recombinant, GST-tagged chicken
α1I- or α2I-domains to the grids.
5. Wash grids with PBS in order to remove unbound material.
6. Incubate grids with a murine monoclonal anti-GST antibody
for the detection of bound I-domains.
7. Wash with PBS.
8. Expose the grids to immuno-gold particles coated with anti-
IgG antibodies.
9. Stain samples with 2% uranyl acetate for 15 min.
10. Analysis by transmission electron microscopy.
11. Quantification of gold particles per unit area (150 μm2) by
counting on several randomly selected EM-observation fields.
Determine means and standard deviations by standard
procedures.
Double-labelling experiments for peripheral non-collagenous
macromolecules and for integrins were carried out in a two-step
164 Uwe Hansen
procedure, firstly with specific antibodies to components of the

fibril periphery, e.g., aggrecan, fibronectin, decorin, COMP,
matrilin-1, or collagen VI, and with 12 nm gold particles coated
with specific anti-immunoglobulins, followed by washing, and
secondly with soluble integrins, the antibody 9EG7 (see Subhead-
ing 3.4), and with 18 nm gold particles coated with goat antibodies
against rat immunoglobulins.
All incubation and washing steps with buffers and reagents as
described before (see Subheading 3.4).
4 Notes
1. The described protocol for the analysis of integrin-mediated

interactions was established by using chondrocytes and
cartilage-specific ECM components; however, the underlying
mechanisms are probably true for many other cell types, and
therefore, the protocol is easy to adapt for use with other cells.
2. The state of aggregation should be controlled in preparation
for transmission electron microscopy. Therefore, after coating
overnight, attached collagens should be scratched off from the
cell culture dishes. 10 μl drops of this material should be placed
on formvar-carbon-coated grids, stained with 2% uranyl acetate
for 15 min, and analyzed by transmission electron microscopy.
3. Collagens will remain in their molecular form, and fibril for-
mation is prevented under these buffer and temperature
conditions.
4. Quantitative conversion of the monomeric collagens into fibril
suspensions by changing the buffer and temperature
conditions.
5. We have used a short-term adhesion assay as cells produce
rapidly their own ECM after seeding to the cell culture dishes.
Therefore, we ensure that we analyze the interaction of cells
with the different coatings, rather than with a mixture of
coating and ECM components produced by the cells.
6. We have changed our previously published method for the
reconstitution of collagen fibril from monomeric collagen
types II, IX, and XI as the efficiency of fibril formation is
more efficient using HEPES buffer instead of Tris buffer
(Fig. 2).
7. Although fibril formation is essentially complete under these
conditions, traces of collagen monomers should be removed by
percolating the above buffer solution through the uncoated
grids into filter paper. This procedure should be repeated sev-
eral times with PBS.
Fig. 2 Characterization of the efficiency of the in vitro fibrillogenesis of ternary

mixtures of cartilage collagens: SDS-PAGE of collagen samples after completion
of in vitro fibrillogenesis on 4.5–15% polyacrylamide gels without prior reduction
with β-mercaptoethanol. The collagens were dissolved in (a) HEPES buffer or (b)
Tris buffer, and fibril formation was initiated by adding an equal volume of
distilled water and by heating up to 37 C. After completion of fibril formation,
collagens were centrifuged, and the formed collagen fibrils were collected as a
pellet (P), whereas still remaining monomeric collagens were found in the
supernatant (SN). In comparison to collagen mixtures dissolved in Tris buffer,
collagen mixtures dissolved in HEPES buffer were nearly completely formed into
collagen fibrils (compare SN in A with SN in B). P pellet, SN supernatant
Acknowledgments
This work was supported by Deutsche Forschungsgemeinschaft

grants BR1497/4-1 and EB177/9-1 to PB and JAE and a grant
from BBSRC UK (BB/I011226/1) to BL.
References
1. Lu P, Weaver VM, Werb Z (2012) The extra- 4. Mercier I, Lechaire JP, Desmouliere A, Gaill F,
cellular matrix: a dynamic niche in cancer pro- Aumailley M (1996) Interactions of human
gression. J Cell Biol 196:395–406 skin fibroblasts with monomeric or fibrillar col-
2. Nelson CM, Bissell MJ (2006) Of extracellular lagens induce different organization of the
matrix, scaffolds, and signaling: tissue architec- cytoskeleton. Exp Cell Res 225:245–256
ture regulates development, homeostasis, and 5. Woltersdorf C, Bonk M, Leitinger B,
cancer. Annu Rev Cell Dev Biol 22:287–309 Huhtala M, K€apyl€a J, Heino J et al (2017)
3. Blaschke UK, Eikenberry EF, Hulmes DJS, The binding capacity of α1β1-, α2β1- and
Galla HJ, Bruckner P (2000) Collagen XI α10β1-integrins depends on noncollagenous
nucleates assembly and limit lateral growth of surface macromolecules rather than the col-
cartilage fibrils. J Biol Chem lagens in cartilage fibrils. Matrix Biol
275:10370–10378 63:91–105
166 Uwe Hansen
6. Eble JA, Beermann B, Hinz HJ, Schmidt- 11. Xu H, Raynal N, Stathopoulos S, Myllyharju J,
Hederich A (2001) α2β1 integrin is not recog- Farndale RW, Leitinger B (2011) Collagen
nized by rhodocytin but is the specific, high binding specificity of the discoidin domain
affinity target of rhodocetin, an receptors: binding sites on collagens II and III
RGD-independent disintegrin and potent and molecular determinants for collagen IV
inhibitor of cell adhesion to collagen. J Biol recognition by DDR1. Matrix Biol 30:16–26
Chem 276:12274–12284 12. Böhme K, Winterhalter KH, Bruckner P
7. de Santana Evangelista K, Andrich F, Figueir- (1995) Terminal differentiation of chondro-
edo de Rezende F, Niland S, Cordeiro MN, cytes in culture is a spontaneous process and is
Horlacher T et al (2009) Plumieribetin, a fish arrested by transforming growth factor-beta2
lectin homologous to mannose-binding B-type and basic fibroblast growth factor in synergy.
lectins, inhibits the collagen-binding α1β1 Exp Cell Res 216:191–198
integrin. J Biol Chem 284:34747–34759 13. Bruckner P, Hörler I, Mendler M, Houze Y,
8. Tulla M, Pentik€ainen OT, Viitasalo T, K€apyl€a J, Winterhalter KH, Eich-Bender SG, Spycher
Impola U, Nykvist P et al (2001) Selective MA (1989) Induction and prevention of chon-
binding of collagen subtypes by integrin α1I, drocyte hypertrophy in culture. J Cell Biol
α2I, and α10I domains. J Biol Chem 109:2537–2545
276:48206–48212 14. Mendler M, Eich-Bender SG, Vaughan L, Win-
9. K€apyl€a J, Ivaska J, Riikonen R, Nykvist P, terhalter KH, Bruckner P (1989) Cartilage
Pentikainen O, Johnson M, Heino J (2000) contains mixed fibrils of collagen types II, IX,
Integrin alpha2 I-domain recognizes type I and XI. J Cell Biol 108:191–197
and type IV collagens by different mechanisms. 15. Lenter M, Uhlig H, Hamann A, Jenö P,
J Biol Chem 275:3348–3354 Imhof B, Vestweber D (1993) A monoclonal
10. Leitinger B (2003) Molecular analysis of colla- antibody against an activation epitope on
gen binding by the human discoidin domain mouse integrin chain β1 blocks adhesion of
receptors, DDR1 and DDR2: identification of lymphocytes to the endothelial integrin α6β1.
collagen binding sites in DDR2. J Biol Chem Proc Natl Acad Sci U S A 90:9051–9055
278:16761–16769
Part IV
Collagen in Health and Disease

Chapter 13
Assessing Collagen Deposition During Aging in Mammalian

Tissue and in Caenorhabditis elegans
Alina C. Teuscher, Cyril Statzer, Sophia Pantasis, Mattia R. Bordoli,
and Collin Y. Ewald
Abstract
Proper collagen homeostasis is essential for development and aging of any multicellular organism. During
aging, two extreme scenarios are commonly occurring: a local excess in collagen deposition, for instance
during fibrosis, or a gradual overall reduction of collagen mass. Here, we describe a histological and a
colorimetric method to assess collagen levels in mammalian tissues and in the nematode Caenorhabditis
elegans. The first method is the polychrome Herovici staining to distinguish between young and mature
collagen ratios. The second method is based on hydroxyproline measurements to estimate collagen protein
levels. In addition, we show how to decellularize the multicellular organism C. elegans in order to harvest its
cuticle, one of the two major extracellular matrices, mainly composed of collagen. These methods allow
assessing collagen deposition during aging either in tissues or in whole organisms.
Key words Collagen, Aging, Tissue, Herovici staining, Hydroxyproline, Cuticle, Isolation, Freeze-
cracking, Age-synchronizing, Extracellular matrix, C. elegans
1 Introduction
Cells and tissues are embedded within extracellular matrices

(ECM), which are important for tissue geometry, integrity, and
function [1]. In mammals, collagens constitute about 30% of the
total protein mass in the body and are the major component of the
ECM [2]. In humans, collagen turnover can be extremely slow as
observed in the lens of the eye and in cartilage structures (with a
half-life time of 114 years [3–5]) or extremely fast, within 72 h after
physical exercise, in Achilles tendons [6]. Hence, depending on the
tissue or organ, some collagens are synthesized, secreted, and
integrated in the ECM once early in life and are not replaced over
Alina C. Teuscher and Cyril Statzer contributed equally to this chapter.
169
170 Alina C. Teuscher et al.
the entire lifetime, whereas other collagens have a higher

turnover rate.
During aging, the connective tissue or ECM integrity declines
due to accumulation of damage from collagen fragmentation, oxi-
dation, and glycation [2, 7–12] leading to a continuous reduction
of collagen mass, as best illustrated by wrinkles and sagging skin
[13, 14], but also observed in other connective tissues that support
organ function. This deterioration in ECM integrity has been
implicated in many age-dependent diseases, such as diabetes, can-
cer, chronic liver diseases, and cardiovascular diseases [2, 7–12]. In
parallel, the accumulation of molecular damage [15], chronic
inflammation (inflammaging) [16], or injury during aging might
locally drive abnormal collagen deposition resulting in fibrosis
[17]. These observations suggest that in general the overall colla-
gen mass tends to decline as a function of age, but due to the
increasing incidence of damage or injury, atypical collagen accumu-
lation might occur more locally in certain parts of the body. There-
fore, proper assessment of the extent of collagen deposition might
reveal novel insights in maintenance and remodelling of the ECM
during aging.
A direct role for proper collagen deposition and remodelling
during aging has become evident by studying model organisms,
such as mice and C. elegans. For instance, deficits in the ECM
remodelling enzyme (MMP14) lead to premature aging, short
lifespan, and cell senescence in mice [18]. Interestingly, long-lived
mice preserve connective tissue elasticity and integrity during aging
[19, 20]. Consistent with these findings, in the nematode
C. elegans, collagen synthesis declines during aging. Moreover,
most if not all long-lived C. elegans mutants delay this progressive
decrease in collagen mass [21]. Delaying this progressive reduction
in collagen mass by actively prolonging the biosynthesis of col-
lagens during lifetime is required and sufficient for healthy aging
and longevity in C. elegans [21].
In addition to studying the progressive age-dependent decrease
in ECM integrity in mice, C. elegans provides additional unique
opportunities to explore this phenomenon during aging, since
(1) C. elegans is transparent, thereby allowing ECM components
to be tagged by fluorescent proteins to directly monitor ECM
homeostasis and integrity noninvasively in vivo [21], and
(2) C. elegans is a well-established aging model because of its
short lifespan (about 3 weeks) and powerful established genetics.
However, histological methods are not as commonly used or even
developed in C. elegans compared to mice. Below, we describe a
histological method (Herovici staining) to be used for both
C. elegans and mammalian tissue, a colorimetric method (measure-
ment of hydroxyproline levels) to estimate collagen levels in both
C. elegans and mammalian tissue, and a technique to isolate the
Collagen Deposition During Aging 171
C. elegans cuticle (extracellular matrix) to be used for further

assessment of collagen deposition or other analysis.
The Herovici staining protocol [22] is one of the several histo-
logical methods used to assess the extent of collagen deposition in a
tissue. Compared to other widely used collagen staining techniques
(e.g., picrosirius red staining or Masson’s trichrome staining), the
Herovici protocol allows to distinguish between young (e.g., colla-
gen III) and mature collagen (e.g., collagen I) without cross-
polarized filters, providing an obvious advantage in the visualiza-
tion and analysis of processes, where changes in collagen ratios are
expected or known to occur. The sequential use of several dyes
results in a polychromatic staining, with young collagen appearing
blue, mature collagen colored pink to red, and cytoplasm counter-
stained in yellow (Fig. 1). Moreover, cell nuclei are stained dark
blue to black by hematoxylin. The Herovici method is mostly used
for histological studies in skin; however, the staining protocol
works well in other tissues or organs too (e.g., liver or lung).
Collagen levels can also be determined by using hydroxyproline
measurements. Collagens have characteristic [Gly-X-Y] repeats,
whereby glycine is at every third position, X is frequently proline,
and Y is frequently 4-hydroxyproline. These prolonged stretches of
[Gly-X-Y] repeats in the collagen protein are important to form a
stable triple helix from the three constitutive collagen chains
[2]. Since collagens are enriched in hydroxyproline and this post-
translational modification is not as common as in other proteins,
measuring hydroxyproline abundancy results in an estimate of total
collagen levels [23, 24]. For a more accurate estimation of collagen
amounts by hydroxyproline measurements, one could decellularize
tissue samples of interest [25] in order to decrease the levels of
other hydroxyproline-containing proteins found in the cytoplasm.
Here we show techniques to isolate the C. elegans cuticle, a
procedure to further enrich for collagen content during sample
harvesting for instance for hydroxyproline measurements. The cuti-
cle and basement membranes are the two major extracellular matri-
ces of C. elegans [26]. The C. elegans cuticle is mainly composed of
collagens [27]. Two methods are described to isolate this cuticle.
The first and simpler method was modified from Leushner et al.
[28], and the second method was adapted from Cox et al. [29]. The
advantages and differences of each method are explained in Sub-
heading 3 below.
In summary, collagen deposition and remodelling become
faulty during aging resulting either in a generalized decline in
collagen biosynthesis or, locally, in excessive and amorphous colla-
gen deposition (fibrosis). The methods described here are tailored
for mammalian tissue samples and the organism C. elegans to
enable the quantification of total collagen protein levels (hydroxy-
proline measurement), to distinguish between young and mature
collagen ratios, and to enrich for collagen-containing extracellular
Fig. 1 Herovici staining for mammalian tissue and C. elegans. (a–c) A schematic of the individual steps of
the Herovici staining described in Subheading 3 (see Subheading 3.1). (a) Freeze-cracking of C. elegans
matrices in C. elegans. In addition, we describe in detail how to

synchronize C. elegans in order to start a time course for aging
studies. With these methods in hand, the investigator should be
able to assess collagen deposition during aging and in age-related
pathologies.
2 Materials
General mouse necropsy and mammalian tissue harvesting or

C. elegans maintenance are described elsewhere [30, 31].
C. elegans strains can be acquired from the Caenorhabditis Genetics
Center (CGC, https://cgc.umn.edu, [31]). For the preparation of
all solution, distilled water and analytical grade reagents are recom-
mended. All Herovici staining solutions are stored in glass bottles at
room temperature protected from light. Disposal of waste materials
should be carried out according to specific local regulations. Suit-
able protection (lab coat, safety glasses, and gloves) and the use of a
fume hood are recommended for the handling of acids and other
hazardous substances.
2.1 Herovici Staining 1. 70–100% EtOH.

for Mammalian Tissue 2. Microscope slides with cut edges and frosted end, 1 mm thick,
and C. elegans 76 26 mm.
3. Poly-L-Solution: 400 mg Sigma P1524 poly-L-Lysine; 0.2 g
sodium azide (0.1%) in 200 ml ddH2O. Can be reused and
stored at 4 C.
4. PBS: 8 g NaCl, 0.2 g KCl, 1.42 g Na2HPO4, 0.24 g KH2PO4,
pH adjusted with HCl to pH 7.4.
5. PFA: 4% paraformaldehyde solution in PBS.
6. Herovici’s Solution 1: Add 270 ml of deionized water to a glass
beaker. Weigh 1.25 g celestine blue powder, and transfer it to
the glass beaker. Weigh 15 g of aluminum potassium sulfate
dodecahydrate (CAS: 7784-24-9; 31242, Merck) and add it to
the glass beaker. Heat and boil the solution until the powders
are dissolved (approximately 3–5 min). Let the solution cool
down to room temperature, and transfer to a 500 ml graduated
cylinder, and add distilled water to a final volume of 300 ml.
Fig. 1 (continued) (see Subheading 3.1.1). (b) Mammalian tissue harvesting and deparaffinization (see
Subheading 3.1.2). (c) Shared part of the Herovici staining protocol for both C. elegans and mammalian
tissue (see Subheading 3.1.3). (d) Herovici-stained mouse skin specimen. The young collagen (collagen type
III) is stained in blue, whereas the mature collagen (collagen type I) is stained in red. Cytoplasm is counter-
stained in yellow and cell nuclei dark blue to black. Three layers forming the skin (epidermis, dermis,
hypodermis) can be identified. (e) Herovici-stained C. elegans at day 1 of adulthood
Add 60 ml of glycerol once the solution has cooled down to

room temperature (see Note 1).
7. Herovici’s Solution 2: Weigh 1 g of hematoxylin powder,
transfer it to a glass beaker, and dissolve it in 100 ml of 95%
ethanol in distilled water (alcoholic hematoxylin). Weigh 4.5 g
of ferrous II sulfate (FeSO4) powder and 2.5 g ferric III chlo-
ride (FeCl3) powder, and transfer them to a separate glass
beaker containing 298 ml distilled water. Add 2 ml of 36%
(concentrated) hydrochloric acid (toxic, see Note 2). Stir the
solution on a magnetic stirrer till powders dissolve (see Note 3).
Mix this solution with the alcoholic hematoxylin from step 1
(see Notes 4–6).
8. Herovici’s Solution 3: Add 300 ml of distilled water to a glass
beaker. Weigh 1.25 g of metanil yellow, and transfer it to the
glass beaker. Add 25 drops (use a 1 ml pipette tip) of glacial
acetic acid (see Notes 7 and 8).
9. Herovici’s Solution 4: Add 1800 ml of distilled water to a glass
beaker. Add 9 ml of glacial acetic acid (toxic, see Notes 7 and 9).
10. Herovici’s Solution 5: Weigh >1 g of lithium carbonate
(Li2CO3) powder, and dissolve it in 78 ml distilled water
(aqueous Li2CO3). The solution will be saturated (see Note
10). Add 1 ml of saturated aqueous Li2CO3 to 500 ml distilled
water to a glass beaker (see Notes 11 and 12).
11. Herovici’s Solution 6: Weigh 0.15 g of methyl blue powder,
and dissolve it in 150 ml distilled water in a glass beaker. Weigh
0.2 g of acid fuchsin powder, and dissolve it in 150 ml of
saturated aqueous picric acid in a separate glass beaker. Mix
the two solutions together. Add 30 ml of glycerol and 0.15 ml
of saturated aqueous Li2CO3 (see Notes 11 and 13).
12. Herovici’s Solution 7: Add 10 ml of glacial acetic acid to 1 L of
distilled water (see Notes 7 and 14).
2.2 Total Collagen 1. 5 N NaOH.

Quantification 2. 5% sodium hypochlorite solution (or household bleach).
for Mammalian Tissue
3. Sterile ddH2O.
and C. elegans
4. M9 buffer: 3 g KH2PO4, 7.52 g Na2HPO4, 5 g NaCl, and
2.2.1 Age 0.05 g MgSO4 in 1 L ddH2O, autoclaved [31].
Synchronization
of C. elegans
2.2.2 Total Protein 1. QuickZyme Protein Assay Kit (QuickZyme Biosciences) or

Quantification equivalent kit from other manufactures suitable for protein
quantification of acid hydrolyzed samples.
2. Pierce™ BCA Protein Assay Kit (Thermo Fisher) or equivalent
kit from other manufactures.
2.2.3 Total Collagen 1. QuickZyme Total Collagen Kit (QuickZyme Biosciences) or

Quantification equivalent kit from other manufactures.
2. Aqueous hydrogen chloride solution: 4 M, 6 M, and 12 M.
2.3 Materials 1. Sodium dodecyl sulfate (SDS) buffer: 1% SDS in ddH2O.

for Cuticle 2. Triton X-100 buffer: 0.5% Triton X-100 in ddH2O.
(Extracellular Matrix)
3. 15 ml glass conical centrifuge tubes with screw gaps.
Isolation in C. elegans
2.3.1 Cuticle Isolation
(Freeze-Thaw Protocol)
2.3.2 Cuticle Isolation 1. M9 buffer: 3 g KH2PO4, 7.52 g Na2HPO4, 5 g NaCl,

(ST Buffer Protocol) 0.0493 g MgSO4 in 1 L ddH2O, autoclaved.
2. Sonication buffer: 10 mM Tris–HCl, pH 7.4; 1 mM EDTA; in
ddH2O.
3. 0.1 M PMSF buffer: 1 mM phenylmethanesulfonylfluoride in
isopropanol.
4. ST buffer: 1% SDS; 0.125 M Tris–HCl, pH 6.8; in ddH2O.
3 Methods
3.1 Herovici Staining All procedures are carried out at room temperature, if not indicated
for Mammalian Tissue otherwise. Glass slides with tissue sections fixed in 4% paraformal-
and C. elegans dehyde (PFA in PBS) are used for the Herovici staining protocol.
Reusable Herovici solutions can be poured back in their original
container. At the end of this procedure, tissue sections will be ready
to be imaged under the microscope. In this section, we first
describe the harvesting and preparation for mammalian tissue (see
Subheading 3.1.1) and then for C. elegans (see Subheading 3.1.2)
followed by the shared methodical part (see Subheading 3.1.4).
3.1.1 Harvesting 1. Organs or tissues of interest are harvested as described else-

and Preparing Mammalian where [30], fixed in 4% paraformaldehyde for 24 h at 4 C,
Tissue Sections (Fig. 1a) submersed in increasing concentrations of ethanol, embedded
into paraffin [32], and sectioned with a microtome (3.5 μm
sections).
2. Deparaffinize tissue sections by immerging the glass slides in
xylol for 10 min. Repeat this step once for a total of two.
3. Rehydrate the tissue sections stepwise from 100% to 50%
(100%, 96%, 90%, 80%, 70%, 50%) ethanol by immerging the
glass slides for 1 min in each solution.
4. Continue the protocol; see Subheading 3.1.4.
3.1.2 Preparation of Poly- 1. Clean microscope slides with Kimwipes and 70%–100% EtOH.
L-Slides for Freeze-
2. Pipette 50 μl Poly-L-Solution on the non-frosted part of the
Cracking of C. elegans microscope slide. Place another microscope slide on top, so
that the transparent parts completely overlay and the frosted
parts are hanging over on the opposite sides.
3. Slide the two slides apart: one side per slide should now be
covered with a thin layer of Poly-L-Solution.
4. Turn the slides around, so that the Poly-L-Solution sides are
facing upward, and place the slides in a 55 C incubator for 1 h
to dry. Slides can now be used directly or stored for several
months in a clean slide box at 4 C.
3.1.3 Harvesting The described C. elegans freeze-cracking method was adopted from
and Preparing C. elegans Duerr 2006 [33]. For freeze-cracking use a minimum of 4000
Samples Using a Freeze- C. elegans or better about 8000–12,000 C. elegans (see Note 15).
Cracking Method (Fig. 1b) For a protocol to age-synchronize and culture C. elegans (see Sub-
heading 3.2.2):
1. Place the fixation solution (4% paraformaldehyde (PFA in
PBS)) and several empty 10 cm petri dishes on ice in order to
pre-chill for later use.
2. Wash C. elegans off culturing plates using M9 buffer, and
collect them in a 15 ml conical centrifuge tube.
3. Centrifuge for 1 min at 1000 rpm (standard table centrifuge is
ca. 200 g; see Note 16), and remove supernatant. Wash the
C. elegans until the supernatant is clean and no bacteria are
visible (see Note 17).
4. Rinse C. elegans by filling up the 15 ml conical centrifuge tube
with 14 ml ddH2O, centrifuge for 1 min at 1000 rpm
(200 g), remove supernatant, and add ddH2O to the pelleted
C. elegans to a total volume of 100 μl.
5. Using a glass Pasteur pipette, place a small drop of animals
(ca.10–20 μl) on a Poly-L-Solution covered slide. Wait shortly
to let the C. elegans settle on the slide. Use the pipette to spread
the C. elegans out so that they do not overlap.
6. Carefully place another Poly-L-Solution covered slide on top of
the slide with the C. elegans so that the Poly-L-covered sides
face each other, while the frosted parts are not overlapping.
7. Put on safety goggles and gloves. Hold the two slides slightly
pressed together into liquid nitrogen until the bubbling stops.
Take the slides out. Wait for ca. 20 s. Then start to snip with
your finger carefully against one of the slides until they rip
apart.
8. Check which slide has most of the C. elegans attached to it, and
place either only this slide or both slides separately into indi-
vidual empty 10 cm petri dishes, which are standing on ice.
9. Pipette 100–200 μl 4% PFA pre-chilled onto the C. elegans
covered part of the slides. Fix the C. elegans by incubating the
slides at 4 C for at least 30 min or up to 24 h.
10. After fixation, rinse the slides gently at least three times with
PBS. Proceed immediately with staining, or store the slides in a
Coplin jar or petri dish containing PBS at 4 C for several days.
3.1.4 The Herovici 1. Immerse slides in Solution 1 for 5 min (see Note 1).
Staining for Both 2. Rinse slides under running tap water for 2 min (see Note 18).
Mammalian Tissue
3. Immerse slides in Solution 2 for 5–6 min (see Notes 5, 6,
and C. elegans (Fig. 1c)
and 19).
4. Rinse slides under running tap water for 2 min (see Note 18).
5. Immerse slides in Solution 3 for 2 min (see Note 8).
6. Transfer slides directly to Solution 4 for 2 min (see Note 9).
7. Rinse slides under running tap water for 2 min (see Note 18).
8. Immerse slides in Solution 5 for 2 min (see Note 12).
11. Immerse the slides in 100% ethanol for 1 min. Repeat this
dehydration step once for a total of two.
12. Immerse the slides in xylol for 2 min. Repeat this step once for
a total of two.
13. Let the slides dry for a minimum of 10 min (see Note 20).
14. Mount the slides with a mounting media (e.g., Eukitt®), and
let them dry under a fume hood (see Note 21).
15. After adding the cover slip, the specimen is ready for imaging.
Examples of a Herovici-stained mammalian tissue section and a
C. elegans are shown in Fig. 1d, e, respectively.
3.2 Total Collagen The collagen content in a tissue or whole-organism sample can be
Quantification determined in multiple ways (Fig. 2). It is generally advised to select
for Mammalian Tissue the normalization entity to be as conclusive as possible. Here, total
and C. elegans collagen quantification is combined with either absolute animal
number (C. elegans) or with total protein determination for both
mammalian tissue and C. elegans. Here, we focus mainly on how to
establish aged C. elegans cultures and on the assessment of collagen
levels in C. elegans, since the use of hydroxyproline measurement in
mammalian tissue is well-established.
a b c
harvest
collagen conc [mg / ml]

1.0 0.3
protein conc [mg / ml]

tissue
(in well)
(in well)
formula (1)
store at-80°C
0.1
a) harvest 0.5
b) wash
95°C, 20 h c) count
in acid L4 L4
collagen in total protein [%]

10
split sonicate
1 2 3 4 5 6 7 8 9 10 11 12
35 µl 15 µl 1 2 3 4 5 6 7 8 9 10 11 12
A A
B B
C C
store at -80°C
D D
E E
F F
G
H
G
H
0
Collagen assay Protein assay L4

(acidic env.)
split
1 2 3 4 5 6 7 8 9 10 11 12
A 1 2 3 4 5 6 7 8 9 10 11 12
B
C
D
95°C, 20 h A
B
C
D
E
E
F
G
H in acid F
G
H
Collagen assay Protein assay
Fig. 2 Total collagen quantification using hydroxyproline assay for mammalian tissue and C. elegans. (a, b)
Schematic of the collagen assay that can be performed using mammalian tissue (a) as well as C. elegans
samples (b). Please see Subheading 3 (see Subheading 3.2) for detailed protocol steps. (c) An example of how
to quantify the collagen-to-protein ratio for larval L4 stage spe-9(hc88) C. elegans samples is shown. (d)
Formula (1) shows how to calculate the ratio of collagen content to total protein levels. The dilution term
referring to the digestion is omitted when the total protein content is determined using the hydrolyzed sample.
Formula (2) shows how to normalize collagen content per individual C. elegans
3.2.1 Mammalian Tissue There are several ways to assess hydroxyproline levels in mammalian
(Fig. 2a) tissues. Here, we briefly outline the use of a colorimetric method
using the QuickZyme Kit for both hydroxyproline and total protein
measurements. The advantage of this protocol is that each HCl
hydrolyzed tissue sample can be split to measure (1) hydroxyproline
levels to estimate collagen protein content and (2) to measure total
protein levels from the same sample in order to normalize the
collagen content to total protein levels. Equivalent kits or protocols
from other manufacturers can be used. For more details, please

follow the manufacturer’s instruction.
1. Organs or tissues of interest are harvested as described else-
where [30]. Frozen organs/tissues and formalin-fixed or
paraffin-embedded tissues can also be used.
2. Hydrolyze the tissue in 6 M HCl in a safety centrifuge tube for
16 h at 95 C in a one 1:10 ratio of wet tissue weight to acid
volume (see Note 22). Split the hydrolysate: 35 μl is used for
collagen quantification and 15 μl to assess the total protein
content (see Note 23).
3. To generate the standard curve for the colorimetric quantifica-
tion of collagen and total protein levels, please follow the
manufacturer’s instructions.
3.2.2 Synchronizing C. elegans samples are prepared in one batch and harvested in user-
C. elegans for an defined intervals over the animal’s lifespan. To age-synchronize the
Age-Dependent Time C. elegans population, we describe a bleach preparation protocol
Course Analysis (Fig. 2b) to harvest eggs from gravid C. elegans adults. For each time point
and condition, 2000–5000 animals are required; therefore
ca. 10,000–100,000 eggs should be harvested in the initial lysis step.
1. Use multiple culturing plates containing a large amount of
gravid C. elegans adults to perform a population lysis. Wash
the culturing plates with sterile H2O. Pipette the H2O across
the plate several times to loosen C. elegans that are stuck onto
the bacterial lawn.
2. Collect the liquid containing the C. elegans in a sterile 15 ml
conical centrifuge tube. Centrifuge for 1 min at 1000 rpm
(standard table centrifuge is ca. 200 g; see Note 24). Discard
supernatant. Wash residual bacteria away from C. elegans by
filling up the 15 ml conical centrifuge tube with sterile H2O,
spin down, and discard the supernatant. Repeat this step until
the supernatant is clear. Spin down to pellet gravid C. elegans,
and add H2O to a total volume of 3.5 ml.
3. Put on gloves and eye protection. Add 1 ml of 5% sodium
hypochlorite solution (or household bleach), and add 0.5 ml
of 5 N NaOH. Close 15 ml conical centrifuge tube, and shake
well. In 1-min intervals, hold the 15 ml conical centrifuge tube
under a dissecting scope to observe lysis of gravid C. elegans,
and shake well after observation to optimize lysis. Once almost
all C. elegans are lysed, add M9 buffer to a total volume of 15 ml,
and centrifuge at 3000 rpm (ca. 1800 g) for 1 min to pellet the
eggs (see Note 25). Discard supernatant, and wash the eggs at
least three times by filling up to 15 ml with M9 buffer and
re-pelleting the eggs by centrifugation at 3000 rpm (ca. 1800 g)
for 1 min. Discard the supernatant (see Note 26).
4. After lysis and subsequent washing steps, the samples are incu-
bated while rotating for 12 h in M9 buffer supplemented with
5 μg/ml cholesterol.
5. 500–1000 larval L1 C. elegans larvae are placed on each 10 cm
culturing plate. If a temperature-sensitive sterile strain is used,
the sterility mechanism has to be activated at the needed time
point (see Note 27). If floxuridine (FUDR) is used to avoid
offsprings, the animals are transferred at the larval L4 stage to
50 μg/ml FUDR culturing plates (see Notes 28 and 29).
6. 2000–5000 C. elegans (2–10 plates) are harvested for each
condition and time point. Transfer all animals from the plates
to a 15 ml conical centrifuge tube. The animals are centrifuged
briefly at low speed (200 rpm, ca. 8 g) to pellet aged C. ele-
gans adults. The suspension containing unhatched eggs or
larvae (if FUDR is used) is aspirated and discarded. Repeat
this washing step three times or until only the desired
C. elegans adults are present and supernatant is clear (see Note
30).
7. From the last wash that was filled up with M9 buffer to a total
volume of 14 ml, take five times 20 μl aliquots, and place these
aliquots on a petri dish lid to count the number of animals in
each 20 μl drop using a dissecting scope (see Notes 31 and 32).
To estimate total number of C. elegans, average the number of
C. elegans from each 20 μl drop and multiply by 700 to get an
estimate of the total number of C. elegans present in 14 ml
sample.
8. Centrifuge for 1 min at 1000 rpm (ca. 200 g) to softly pellet
C. elegans. Discard supernatant. Use a glass Pasteur pipette to
transfer the C. elegans pellet to a labelled 1.5 ml Eppendorf
collection tube, fill up with sterile H2O to a total volume of
200 μl, and store at 80 C (see Notes 16, 33, and 34).
9. At the end of the time course, once all samples have been
collected, proceed by thawing all tubes. At least 3–5 freeze-
thaw cycles should be performed by thawing at room tempera-
ture and freezing in liquid nitrogen (or dry ice with EtOH or
by placing back into the 80 C).
10. Then sonicate all samples on ice until all animals are disrupted.
The C. elegans cuticle is the hardest to break. With a standard
sonication device, it will take several repeats to break down all
C. elegans (see Note 35).
3.2.3 Total Protein To determine the total collagen content as a fraction of total
Quantification protein abundance, the total protein content has to be determined
for C. elegans (Fig. 2) separately.
1. Pipette 25 μl of each sonicated sample as well as of the provided

standard solutions into a 96-well plate, and quantify the total
protein content following the manufacturer’s instructions
(Pierce™ BCA Protein Assay Kit, Thermo Fisher, or equivalent
kit from other manufacturers). If needed, dilute the sample
with M9.
2. Use a plate reader to quantify the color change at the specified
wavelength.
3.2.4 Total Collagen Here, we determine the total collagen abundance by digesting the
Quantification sample and directly measuring the concentration of the amino acid
for C. elegans (Fig. 2) hydroxyproline in each sample.
1. The sonicated samples are transferred into heat-stable tubes
provided by the kit manufacturer (QuickZyme Total Collagen
Kit, QuickZyme Biosciences, or equivalent kit from other man-
ufacturers). The provided collagen standard does not require to
be sonicated but is otherwise treated identically as the samples.
2. According to the manufacturer’s instructions, the sample and
standard are mixed with 12 M HCl solution and incubated for
20 h at 95 C.
3. In brief, after the digestion the supernatant is isolated, the acid
content lowered, assay buffer added, and the plate incubated at
room temperature for 20 min followed by the addition of the
detection reagent and subsequent incubation at 60 C for 1 h.
The color change is quantified using a plate reader (see Note 36).
4. The total collagen content per sample in a well can be directly
determined through the measured color change. In a first step,
the collagen reference sample (1.2 mg/ml of rat tail collagen) is
used to generate a standard curve relating the magnitude of the
color change to the collagen content of the well. Over the used
concentration range, this relationship is approximately linear.
With the standard curve available, the collagen content of the
samples can be quantified. It must be noted that the collagen
standard supplied in the QuickZyme Biosciences Kit contains
rat tail collagen. To validate the use of rat tail collagen as a
reference for hydroxyproline levels for C. elegans collagens, the
potential hydroxyproline occurrences on the Y position in the
C. elegans [Gly-X-Y] triple repeats in collagens were estimated
in silico. The relative proline abundance on the Y position is
similar between C. elegans and rat tail collagens (personal com-
munication Jan M. Gebauer). Example results for C. elegans
protein and collagen concentration are shown in Fig. 2b.
5. The total collagen-to-total protein ratio can be determined by
combining the measured collagen and protein concentrations
according to formula (1). In the case of C. elegans, the total
collagen abundance per animal can be calculated by subjecting
the measured collagen concentration to formula (2). All for-

mulas to determine the relative collagen contents are depicted
in Fig. 2c.
3.3 Methods This method was modified from Leushner et al. [28], for the use in
for Cuticle C. elegans. The two advantages of using this protocol compared to
(Extracellular Matrix) the ST protocol is (1) its simplicity and (2) the SDS-cleaned cuticles
Isolation in C. elegans might still contain Schiff base products or other adducts that might
be of importance of the desired analysis [34]. The major problem
3.3.1 Cuticle Isolation with this protocol is that the C. elegans cuticle tends to stick on
(Freeze-Thaw Protocol) plastic; this can be avoided by using glass materials instead. All steps
after the freeze-thawing should be carried out on ice, except steps
involving SDS. We recommend to use about 12,000 C. elegans per
sample to isolate a sufficient number of cuticles. Please see Subhead-
ing 3.2.2 for a protocol to age-synchronize the C. elegans samples.
An example of an aged C. elegans (7-day adult) is shown in Fig. 3a
for comparison with the isolated cuticles Fig. 3b, c.
1. Use 12,000 adult C. elegans (about 8–12 full culturing plates).
Pipette around 3 ml M9 buffer on each plate. Gently tilt plates
to collect the liquid on one side, and carefully pipette the M9
buffer several times across the plate to loosen up the C. elegans
stuck on the bacterial food lawn. Transfer the animals from the
plates into a 15 ml conical centrifuge tube.
2. Wash the C. elegans three times with 15 ml ddH2O by centri-
fugation at around 1000 rpm (ca. 200 g) for 1 min, and
discard the supernatant (see Note 33).
3. After the last washing step, discard the supernatant again, and
use a glass Pasteur pipette to transfer the C. elegans pellet into a
1.5 ml Eppendorf microfuge tube (see Note 37). Centrifuge
1.5 ml tube at around 100 g for 30 s, discard supernatant,
and fill up with ddH2O to 100 μl.
4. Freeze the tubes in liquid nitrogen for 1 min (or until frozen),
followed by a thawing step at room temperature. Repeat this
freeze-thaw cycle at least three times.
5. Sonicate the samples five times in intervals of 20 s each at an
amplitude of 80% with 20 s breaks on ice in between (Sonoplus
mini20 from Bandelin) (see Note 38).
6. Use a glass Pasteur pipette to transfer the ruptured C. elegans
to 15 ml glass tubes (with screw gaps) (see Note 39), and
suspend them in 10 ml 1% SDS solution, followed by overnight
incubation onto a rotor at 37 C. The 1% SDS should wash out
all the internal cells and organs from the disrupted cuticles
(Fig. 3b).
Fig. 3 Isolated C. elegans cuticles. (a) Differential contrast image of a day 7 adult
C. elegans. Anterior to the left and ventral side down. (b) Intact C. elegans cuticle
after 1% SDS before washing step. Black arrows point to the annuli structure of
the cuticle. White arrowheads point to debris. (c) Intact C. elegans cuticle after
washing. Sometimes other extracellular matrices, such as the basement
membranes of the gonad (white arrowheads), are still attached
7. On the following day, wash the cuticles with sterile ddH2O

containing 0.5% Triton-X by centrifuging at 3000 rpm
(ca. 1800 g) for 1 min; discard supernatant (see Note 40).
8. The cleaned cuticles (Fig. 3c) can now be further processed,
analyzed, or stored at 20 C.
3.3.2 Cuticle Isolation This method was adapted from Cox et al. [29]. Please see Subhead-
(ST Buffer Protocol) ing 3.3.1 for the comparison with the other cuticle isolation freeze-
thaw method, and see Subheading 3.2.2 for a protocol to age-
synchronize the C. elegans samples. All steps after the sonication
should be carried out on ice, except steps involving SDS.
1. Collect 12,000 (adult) C. elegans with M9 buffer in a 15 ml
tube as described in the first step of Subheading 3.3.1.
2. Wash the C. elegans pellet three times by centrifugation at

around 1000 rpm (ca. 200 g) for 1 min, and discard the
supernatant (see Note 33).
3. After the last washing step, discard the supernatant, and resus-
pend the C. elegans pellet in 5 ml sonication buffer. Incubate
on ice for 10 min.
4. Add 30 μl 0.1 M PMSF to the solution, and sonicate five times
for 20 s with an amplitude of 80% with 20 s breaks on ice in
between (Sonoplus mini20 from Bandelin).
5. Wash the ruptured C. elegans cuticles three times with the
sonication buffer by centrifugation at 1500 rpm (around
450 g) for 1 min, and discard the supernatant.
6. Discard the supernatant, and transfer the cuticles with a glass
Pasteur pipette to a 1.5 ml microfuge tube (see Note 37).
Centrifuge the 1.5 ml tube at around 100 g for 30 s, discard
supernatant, and fill up with sonication buffer to 100 μl.
7. Add 1 ml ST buffer, and heat the tube for 2.5 min at 95 C.
8. Incubate the tubes on a rotator overnight at room
temperature.
9. Wash cuticles three times with 0.5% Triton-X in sterile ddH2O
by centrifuging at 3000 rpm (ca. 1800 g) for 1 min; discard
supernatant.
10. The cleaned cuticles can now be further processed, analyzed, or
stored at 20 C.
4 Notes
4.1 Herovici Staining 1. Solution 1 can be reused.

for Mammalian Tissue 2. Caution. Hydrochloric acid should be handled with care and
and C. elegans added to the solution under a fume hood.
3. If required filter the solution to eliminate visible impurities.
4. Alcoholic hematoxylin and the ferrous solution are prepared
separately and mixed at the end.
5. Solution 2 has a shelf life of approximately 2–3 months due to
oxidation.
6. Solution 2 can be reused.
7. Caution. Glacial acetic acid should be handled with care and
added to the solution under a fume hood.
9. Do not reuse Solution 4.
10. By adding more than 1 g of Li2CO3 to 78 ml distilled water,

the solution will be saturated. Do not shake to try to dissolve
the precipitate. Let the solution stand.
11. Pipette the aqueous Li2CO3 from the top without disturbing
the precipitate accumulated on the bottom of the container.
14. Do not reuse Solution 7.
15. Per plate you can make 2–3 freeze-cracked C. elegans slides.
16. To prevent adhesion of animals to plastic pipette tips, one can
either add 0.5% Triton-X as a detergent or switch to glass
pipettes.
17. Bacteria prevent the animals to properly stick to the slides.
18. Regular water. This step can be performed directly in the sink.
19. A longer incubation may be needed if the solution has not been
freshly prepared.
20. A longer drying time (30 min–1 h) is perfectly fine.
21. Eukitt® will need approximately 30 min to dry.
4.2 Total Collagen 22. Ideally, a 1:10 ratio of wet tissue weight to acid volume is
Quantification recommended, with a minimal volume of 200 μl 6 M HCl
for Mammalian Tissue per tube. For example, 50 mg of tissue mixed with 500 μl of
and C. elegans 6 M HCl can be used.
23. To be able to relate the total collagen content of the sample as
robustly as possible to the total protein abundance, we recom-
mend to measure the later directly in the hydrolyzed sample.
To be able to quantify the free amino acid level present in the
acidic hydrolyzed sample, a suitable kit has to be chosen. We
recommend the Total Protein Assay, QuickZyme Biosciences.
24. When centrifuging C. elegans, either turn off the breaks or turn
down the deceleration to a minimum.
25. In general, complete lysis can take about 4–8 min. Hence, it is
important to regularly check if all C. elegans are lysed. If lysis is
performed for too long, the eggs will suffer from increasing
damage.
26. Sometimes three washes are not enough to wash out all bleach
solution. Hence, check by smelling the 15 ml conical centri-
fuge tube if bleach odor is still present.
27. We recommend the C. elegans mutant spe-9(hc88). Due to a
temperature-sensitive defect in spermatogenesis, the animals
are not able to lay fertilized eggs when shifted to 25 C during
development. To ensure all animals are sterile, L1 animals can
be cultured at 25 C until day 1 or 2 of adulthood.
28. To avoid food scarcity, check the grow-up plates often, and if
needed, transfer to fresh plates.
29. To check the effectiveness of the intervention, a selected num-
ber of plates can be used for validation by, for example, placing
a fluorescent reporter strain on them.
30. Since the collagen content is changing over the lifespan of
C. elegans, it is crucial to work with age-synchronous popula-
tions. Thus, all eggs and larvae must be removed as best as
possible. This can be achieved by transferring animals often to
new plates and rigorous washing or through filtration.
31. Because adult C. elegans will sink to bottom of the conical
centrifuge tube, shake the 15 ml conical centrifuge tube well
between taking the aliquots for counting. Try to pipette from
the middle of the 15 ml conical centrifuge tube for consistency.
32. Use a pipette with a large orifice to sample C. elegans during
the counting procedure. Since adults grow in size, a too small
pipette tip can lead to an underestimation of older populations.
33. At this step of the protocol is a good time point for a break or
to wait until all time points from before are collected, for
instance, if you collect daily samples over the 2–3 weeks life-
span of C. elegans.
34. It is recommended to fill up all sample C. elegans pellets to
200 μl to simplify the downstream steps.
35. Example sonication configuration using a Bandelin Sonoplus,
UW mini20 device: Amplitude 80%, 30s intervals, pulse 1.0 s
on, 1.0 s off. However, using a more powerful sonicator is
recommended to accelerate C. elegans fragmentation.
36. It is advised to perform a few test reactions prior to processing
the entire batch for both quantification methods to be able to
adjust sample dilution should it be too concentrated.
4.3 Cuticle 37. C. elegans tend to stick on the plastic tip or on the plastic tubes.
(Extracellular Matrix) Hence, make sure to use a glass Pasteur pipette and pipette the
Isolation in C. elegans C. elegans on the bottom of the plastic tube to avoid sticking
on the side of the tube.
38. If the cuticles in the freezing protocol are ruptured too much,
you can consider performing only the freezing and thawing
cycles or only sonication.
39. One of the major problems is losing too many cuticles or
C. elegans. Therefore, use always glass pipettes to transfer
worms and cuticles.
40. Triton-X prevents the cuticles from sticking to the tubes during
the washing steps.
Acknowledgments
We thank Marjolein Wildwater for sharing her improved freeze-

cracking protocol, Eline Jongsma for her assistance in adapting the
freeze-cracking protocol, Salome Brütsch and Hayley Hiebert for
their help to develop the C. elegans Herovici staining protocol,
Anna Bircher for contributing to the early stages of the cuticle
isolation protocols and imaging cuticles, Max Hess for providing
the C. elegans example image, and Jan M. Gebauer for bioinfor-
matic prediction of potential prolines in collagens that might
become hydroxylated in C. elegans. Some C. elegans strains were
provided by the CGC, which is funded by NIH Office of Research
Infrastructure Programs (P40 OD010440). This work was sup-
ported by the Swiss National Science Foundation [PZ00P3
161512] to S.P. and M.R.B. and [PP00P3 163898] to A.C.T., C.
S., and C.Y.E. Alina C. Teuscher and Cyril Statzer contributed
equally to this work.
References
1. Hynes RO (2009) The extracellular matrix: not 10. Sell DR, Monnier VM (2012) Molecular basis
just pretty fibrils. Science 326:1216–1219 of arterial stiffening: role of glycation–a mini-
2. Ricard-Blum S (2011) The collagen family. review. Gerontology 58:227–237
Cold Spring Harb Perspect Biol 3: 11. Snedeker JG, Snedeker JG, Gautieri A et al
a004978–a004978 (2014) The role of collagen crosslinks in ageing
3. Sivan S-S, Wachtel E, Tsitron E et al (2008) and diabetes–the good, the bad, and the ugly.
Collagen turnover in normal and degenerate Muscles Ligaments Tendons J 4:303–308
human intervertebral discs as determined by 12. Myllyharju J (2004) Collagens, modifying
the racemization of aspartic acid. J Biol Chem enzymes and their mutations in humans, flies
283:8796–8801 and worms. Trends Genet 20:33–43
4. Heinemeier KM, Schjerling P, Heinemeier J 13. Fenske NA, Lober CW (1986) Structural and
et al (2016) Radiocarbon dating reveals mini- functional changes of normal aging skin. J Am
mal collagen turnover in both healthy and Acad Dermatol 15:571–585
osteoarthritic human cartilage. Sci Transl Med 14. Shuster S, Black MM, McVitie E (1975) The
8:346ra90–346ra90 influence of age and sex on skin thickness, skin
5. Toyama BH, Hetzer MW (2013) Protein collagen and density. Br J Dermatol
homeostasis: live long, won’t prosper. Nat 93:639–643
Rev Mol Cell Biol 14:55–61 15. López-Otı́n C, Blasco MA, Partridge L et al
6. Kjaer M, Langberg H, Miller BF et al (2005) (2013) The hallmarks of aging. Cell
Metabolic activity and collagen turnover in 153:1194–1217
human tendon in response to physical activity. 16. Franceschi C, Campisi J (2014) Chronic
J Musculoskelet Neuronal Interact 5:41–52 inflammation (inflammaging) and its potential
7. Myllyharju J, Kivirikko KI (2001) Collagens contribution to age-associated diseases. J Ger-
and collagen-related diseases. Ann Med ontol A Biol Sci Med Sci 69(Suppl 1):S4–S9
33:7–21 17. Wynn TA (2007) Common and unique
8. Fisher GJ, Quan T, Purohit T et al (2009) mechanisms regulate fibrosis in various fibro-
Collagen fragmentation promotes oxidative proliferative diseases. J Clin Investig
stress and elevates matrix metalloproteinase-1 117:524–529
in fibroblasts in aged human skin. Am J Pathol 18. Gutiérrez-Fernández A, Soria-Valles C, Osorio
174:101–114 FG et al (2015) Loss of MT1-MMP causes cell
9. Shoulders MD, Raines RT (2009) Collagen senescence and nuclear defects which can be
structure and stability. Annu Rev Biochem reversed by retinoic acid. EMBO J
78:929–958 34:1875–1888
19. Flurkey K, Papaconstantinou J, Miller RA et al 26. Kramer JM (2005) Basement membranes,

(2001) Lifespan extension and delayed WormBook : the online review of C elegans
immune and collagen aging in mutant mice biology. pp 1–15
with defects in growth hormone production. 27. Page AP and Johnstone IL (2007) The cuticle,
Proc Natl Acad Sci U S A 98:6736–6741 WormBook : the online review of C elegans
20. Wilkinson JE, Burmeister L, Brooks SV et al biology. pp 1–15
(2012) Rapamycin slows aging in mice. Aging 28. Leushner JR, Semple NL, Pasternak J (1979)
Cell 11:675–682 Isolation and characterization of the cuticle
21. Ewald CY, Landis JN, Porter Abate J et al from the free-living nematode Panagrellus silu-
(2015) Dauer-independent insulin/IGF-1-sig- siae. Biochim Biophys Acta 580:166–174
nalling implicates collagen remodelling in lon- 29. Cox GN, Kusch M, Edgar RS (1981) Cuticle
gevity. Nature 519:97–101 of Caenorhabditis elegans: its isolation and par-
22. Herovici C (1963) A polychrome stain for dif- tial characterization. J Cell Biol 90:7–17
ferentiating precollagen from collagen. Stain 30. Treuting PM, Snyder JM (2015) Mouse nec-
Technol 38:204–205 ropsy. Curr Protoc Mouse Biol 5:223–233
23. McAnulty RJ (2005) Methods for measuring 31. Stiernagle T (2006) Maintenance of C. elegans,
hydroxyproline and estimating in vivo rates of WormBook : the online review of C elegans
collagen synthesis and degradation. Methods biology. pp 1–11
Mol Med 117:189–207 32. Fischer AH, Jacobson KA, Rose J et al (2008)
24. Qiu B, Wei F, Sun X et al (2014) Measurement Paraffin embedding tissue samples for section-
of hydroxyproline in collagen with three differing. CSH Protoc 2008:pdb.prot4989
ent methods. Mol Med Rep 10:1157–1163 33. Duerr JS (2006) Immunohistochemistry,
25. Naba A, Clauser KR, Hynes RO (2015) WormBook : the online review of C elegans
Enrichment of extracellular matrix proteins biology. pp 1–61
from tissues and digestion into peptides for 34. Davis BO, Anderson GL, Dusenbery DB
mass spectrometry analysis. J Vis Exp 101: (1982) Total luminescence spectroscopy of
e53057 fluorescence changes during aging in Caenor-
habditis elegans. Biochemistry 21:4089–4095
Chapter 14
Procollagen C-Proteinase Enhancer 1 (PCPE-1)

in Liver Fibrosis
Efrat Kessler and Eyal Hassoun
Abstract
Fibrosis is characterized by excessive deposition of collagen and additional extracellular matrix (ECM)
components in response to chronic injuries. Liver fibrosis often results from chronic hepatitis C virus
infection and alcohol abuse that can deteriorate to cirrhosis and liver failure. Current noninvasive diagnostic
methods of liver fibrosis are limited in their ability to detect and differentiate between early and intermedi-
ate stages of fibrosis. New biomarkers of fibrosis that reflect ECM turnover are therefore badly needed.
Procollagen C-proteinase enhancer 1 (PCPE-1), a connective tissue glycoprotein that functions as a
positive regulator of C-terminal procollagen processing and subsequent collagen fibril assembly, is a
promising candidate. Its tissue distribution and expression profile overlap those of collagen, and its
expression in fibrosis is upregulated in parallel to the increase in collagen expression. The potential of
PCPE-1 as a biomarker of liver fibrosis was recently established using a CCl4 mouse model of liver fibrosis
by showing that the increase in collagen and PCPE-1 content in the fibrotic mouse liver was reflected by
elevated plasma levels of PCPE-1. This was achieved using a newly developed highly sensitive, specific,
accurate, and reproducible ELISA for mouse PCPE-1, which is based on commercially available antibodies
and is offered as a new research tool in the field. A similar ELISA test was developed for human PCPE-1,
and preliminary results with plasma from liver fibrosis patients revealed increased plasma concentrations of
PCPE-1 in some patients. The protocols of both ELISA tests are outlined herein in great detail to permit
their application by any laboratory with similar interests.
Key words Collagen, Procollagen C-proteinase enhancer 1, PCPE-1, ELISA, Liver fibrosis,
Biomarkers
1 Introduction
Fibrosis is defined by excessive accumulation of collagen and other

extracellular matrix (ECM) components in response to chronic
injuries, leading to impairment of organ function. It can affect
many organs and tissues, including the liver [1–3]. In the Western
Hemisphere, the main causes of liver fibrosis are chronic hepatitis C
virus infection and chronic alcohol abuse that in the absence of
effective therapy can deteriorate to cirrhosis with its serious
189
190 Efrat Kessler and Eyal Hassoun
consequences, hepatocellular dysfunction, hepatocellular carci-

noma, and liver failure [4].
Diagnosis and quantitation of liver fibrosis require accurate,
reproducible, and easily applied methods and are important to
make therapeutic decisions, determine prognosis, monitor disease
progression, and evaluate new anti-fibrotic drugs [3–5]. Current
approaches to evaluate liver fibrosis include invasive and noninva-
sive methods. Despite the availability of many noninvasive diagnos-
tic methods for liver fibrosis, liver biopsy, an invasive technique, is
still considered the best (gold standard) method for assessment of
liver fibrosis [5, 6]. Because of its invasiveness, liver biopsy suffers
several limitations, in particular risk of injury, sampling errors, high
cost, and inaccurate interpretation by the observer [6, 7]. The
available noninvasive methods comprise imaging techniques such
as FibroScan and MRI that assess liver stiffness and immunoassays
of blood markers [5, 8–11]. The blood markers are classified into
direct and indirect. Direct markers are ECM components (e.g., the
carboxyl propeptide of type I procollagen (PICP), the amino pro-
peptide of type III procollagen (PIIINP), matrix metalloprotei-
nases (MMPs), etc.) that reflect ECM turnover. The indirect
markers comprise molecules released by the malfunctioning liver,
including (but not limited to) α2-macroglobulin, apolipoprotein
A1, haptoglobin, γ-glutamyl transferase (GGT), and bilirubin,
which are components of an algorithm-based test known as FibroT-
est [5, 10, 11].
Noninvasive methods applied clinically rely on indirect mar-
kers. Such markers however are suitable mainly for detection of
advanced disease. Their ability to detect early stages of liver fibrosis
or monitor progression of fibrosis is limited [4, 12]. New direct
markers are therefore badly needed [12].
Procollagen C-proteinase enhancer 1 (PCPE-1) is a 50–55 kDa
extracellular matrix glycoprotein that binds to the procollagen
C-propeptide and stimulates its removal by BMP-1/tolloid-like
proteinases (BTPs; also known as procollagen C-proteinases or
PCPs), a reaction essential for collagen fibril assembly
[13, 14]. The tissue distribution of PCPE-1 overlaps that of colla-
gen [15, 16], and its expression is upregulated in liver [17–19] and
cardiac fibrosis [20–22], hypertrophic/keloid scars [23], and scar-
ring cornea [24]. Furthermore, PCPE-1 does not affect proteolytic
processing of other BTPs’ substrates in the ECM [25, 26], meaning
that its enhancing activity is specific to procollagen substrates.
PCPE-1 has been identified in human sera [27–29] and plasma
[30] and rat plasma [18]. Together these properties of PCPE-1
suggested it could be useful as a new direct marker of liver fibrosis, a
possibility supported by our recent studies on a mouse CCl4 model
of liver fibrosis [19]. Using histochemical, immunofluorescence,
and immunoblotting analyses, we confirmed that after 4 to 6 weeks
of repeated CCl4 administration, the treated mice developed liver
PCPE-1 in Liver Fibrosis 191
Fig. 1 Plasma concentrations of mPCPE-1 reflect the progression of liver fibrosis in CCl4-treated mice. Mice
were treated biweekly with CCl4 for 6 weeks at which time, treatment was terminated, and the mice were
followed for 7 additional weeks. Liver fibrosis was assessed at the indicated times by Sirius red collagen
staining (a), immunofluorescence (b), and immunoblotting (e). Each panel in a and b displays results for a
single randomly selected mouse from the indicated group. Numbers in brackets in b represent time expressed
in weeks. (c) Plasma concentrations of PCPE-1 after 6 weeks of CCl4 treatment are significantly higher than
those of control mice (n ¼ 12). (d) Plasma concentrations of PCPE-1 in CCl4-treated (~) and control (●) mice
as a function of time. Arrow, termination of CCl4 treatment. Results are presented as mean SD. (n at weeks
0 to 6 ¼ 12; n at weeks 7–10 ¼ 9; n at week 13 ¼ 6). Plasma concentrations of mPCPE-1 were determined by
the sandwich ELISA described herein at the indicated time points. (e) Immunoblot evaluating collagen I and
PCPE-1 content in liver extracts from CCl4-treated and control mice at weeks 6, 10, and 13. Samples from a
single control mouse and two CCl4-treated mice (all randomly selected) are shown. The amounts of protein
applied on each lane were 4 and 36 μg for collagen and PCPE-1 detection, respectively (Reproduced from ref.
19, Open Access)
fibrosis (increased collagen and PCPE-1 content; Fig. 1a, b, e,

6 weeks). The increase in liver collagen and PCPE-1 expression
was accompanied by elevated plasma levels of PCPE-1 (Fig. 1c),
determined using a sandwich ELISA we developed for mPCPE-1
[19]. Furthermore, a gradual increase in the plasma levels of PCPE-
1 during a 6-week CCl4 treatment was evident (Fig. 1d). Regres-
sion of fibrosis after termination of CCl4 treatment (i.e., a gradual
decrease in liver collagen and PCPE-1 expression) was reflected by a
gradual decrease in the plasma levels of PCPE-1, reaching basal
levels 6–7 weeks after terminating CCl4 administration (Figs. 1a, b,
d, e and 2). Surprisingly, in another recent study by other investi-
gators [31], in which PCPE-1 was evaluated as a biomarker of liver
fibrosis in patients with chronic hepatitis B, the serum levels of
PCPE-1 (determined using a commercial ELISA kit for hPCPE-1)
Fig. 2 Plasma concentrations of mPCPE-1 decline during liver recovery in parallel to the decline of type I
collagen and PCPE-1 liver content. (a) The relative amounts of collagen type I and PCPE-1 were determined by
densitometry of the respective protein bands in the immunoblot shown in Fig. 1e. E/C, ratio of absorbance
units obtained for CCl4-treated (E) and control (C) mice, respectively. The values are mean SD (n ¼ 2).
Values of PCPE-1 plasma concentrations represent the mean SD obtained at weeks 6, 10, and 13; n ¼ 12,
9, and 6, respectively. (b) The relative amounts of PCPE-1 and collagen type I in the liver were derived from the
immunofluorescence analysis shown in Fig. 1b. Values represent relative staining, expressed as the percent-
age of stained area out of the total area in each panel. Values correspond to data derived for a single
representative mouse per group at the specified time point. Plasma concentrations of PCPE-1 are those
obtained for the respective mice. Each value is the mean SD; n ¼ 4 (Reproduced from ref. 19, Open Access)
were found to be lower than those of healthy (control) individuals.

While the reason for this difference is not clear, the authors con-
cluded that PCPE-1 might be used as a noninvasive marker of liver
fibrosis.
The ELISA we developed for measuring mPCPE-1 concentra-
tions is specific, accurate, sensitive, and reproducible (inter- and
intra-assay CV <10%) [19], thus meeting the criteria mandatory for
immunoassays of blood markers. Furthermore, the assay, which
permits measurements of PCPE-1 in mice plasma, employs com-
mercially available antibodies and is simple and inexpensive, so it is
offered as a new research tool in the field. To permit its application
in other laboratories, the ELISA procedure is described below in
detail, along with a similar assay for determination of human
(h) PCPE-1 [32], that permits measurements of hPCPE-1 in
human plasma/serum and can be applied in studies on liver fibrosis
patients.
2 Materials
2.1 Proteins 1. Mouse (m) PCPE-1: Purified from the culture medium of 3T6
and Antibodies mouse fibroblasts [13]. The concentration of the stock solu-
tion we used [19] was 250 μg/ml in 50 mM Tris–HCl,
150 mM NaCl, 5 mM CaCl2, 0.1% Brij 35, pH 7.5 (see Note 1).
Purified recombinant mPCPE-1 is available commercially
(R&D 2239-PE) and can be used instead.
2. Human (h) recombinant PCPE-1: Purified from the culture
media of 293-EBNA cells expressing the recombinant protein
[25, 33]. The concentration of the stock solution used by us
[32] was 0.5 mg/ml in 50 mM HEPES, 150 mM NaCl, 5 mM
CaCl2, and pH 7.5. Purified recombinant hPCPE-1 is also
available commercially (R&D 2627-PE).
3. Bovine serum albumin (Sigma A7906).
4. Rat monoclonal antibody against mPCPE-1: The antibody
(R&D MAB2239) is supplied as lyophilized powder and is
stored at 20 C. Before the first use, the content (500 μg) is
dissolved in 1.0 ml of sterile PBS (final concentration 500 μg/
ml) and is stored in aliquots at 80 C.
5. Goat polyclonal rabbit antibody to mPCPE-1: The antibody
(R&D AF2239) is supplied as lyophilized powder and is stored
at 20 C. Before the first use, the content (100 μg) is dis-
solved in 0.5 ml sterile PBS (final concentration 200 μg/ml)
and is stored in aliquots at 80 C.
6. Alkaline phosphatase (APA) conjugated rabbit antibody to
goat IgG (Sigma A4187).
7. Mouse monoclonal antibody 7A1/11 against hPCPE-1: This
antibody was developed in our laboratory and purified by
protein G-Sepharose chromatography followed by affinity
chromatography, i.e., adsorption to and elution from hPCPE-
1-Sepharose beads. The hPCPE-1-specific IgG fraction
obtained by affinity chromatography was stored in aliquots at
80 C at concentrations ranging from 0.15 to 0.4 mg/ml
(depending on the preparation) in 0.05 M HEPES, 0.15 M
NaCl, 5 mM CaCl2, and pH 7.5. Monoclonal antibody
7A1/11 is available commercially (Santa Cruz sc-73001;
Sigma C2122) (see Note 2).
8. Rabbit polyclonal antibody (IgG fraction) against hPCPE-1:
IgG fraction, 16 mg/ml in PBS. Stored in aliquots at 20 C.
Produced and purified in our laboratory using standard proto-
cols. Similar antibodies are available commercially (Protein
Tech 14993-1-AP; Sigma P 6243).
9. APA-conjugated goat antibody against rabbit IgG (Sigma
SAB3700950).
2.2 Reagents/ 1. Alkaline phosphatase substrate: para-Nitrophenylphosphate

Buffers (pNPP) 5 mg tablets (Sigma S-0942).
2. Magnesium chloride (MgCl2): 2 M solution in H2O. Diluted
to the desired concentration before use (see Note 3).
3. Phosphate-buffered saline (PBS): 137 mM NaCl, 2.7 mM KCl,

6.5 mM Na2HPO4.2H2O, 1.5 mM KH2PO4, pH 7.3.
4. PBST buffer (PBST): PBS + 0.05% Tween-20.
5. Blocking buffer (BSA/PBS): 5% BSA in PBS. As a routine 100 ml
solution is prepared. Pipet 20 ml of 5 PBS into a 100 ml beaker.
Add 80 ml of water. Add 5 g BSA. Leave for about 30 min with
gentle stirring until the BSA powder is solubilized. Adjust to
100 ml with water. Store at 20 C until use.
6. Carbonate buffer: 0.1 M sodium bicarbonate (NaHCO3)/
sodium carbonate (Na2CO3), pH 9.2. Titrate a 0.1 M
NaHCO3 solution with a 0.1 M solution of Na2CO3 to
pH 9.2.
7. APA assay buffer: 1 mg/ml pNPP in 0.9 M diethanolamine-
HCl, 0.5 mM MgCl2, pH 9.8. Diethanolamine (DEA) buffer is
prepared by adding 9.7 ml of a 99% DEA solution (Sigma
D8885) to 80 ml of water and titrating to pH 9.8 using 6 N
HCl. The volume is then adjusted with water to 100 ml, and
the buffer is stored at 4 C. Before the reaction (see Subhead-
ings 3.1.6 and 3.2, step 6), pNPP is dissolved in 10 ml DEA
buffer, and MgCl2 (5 μl of a 1 M solution) is added to the DEA
buffer solution.
2.3 Supplies 1. Ninety-six wells plates.

and Equipment 2. Multichannel pipettor: For simultaneous pipetting and
removal of solutions to and from eight wells.
3. Stepper: For stepwise pipetting of a constant volume from a
single tip into individual wells.
4. Orbital shaker.
5. Cold room/refrigerator.
6. Incubator set to 30 C.
7. Microplate reader capable of measuring absorbance at 405 nm,
with shaking at constant temperature.
3 Methods
3.1 Sandwich ELISA 1. Adsorption of the first Ab (MAB2239): Dilute the stock solu-
for mPCPE-1 tion 1:100 in carbonate buffer (final concentration 5 μg/ml;
prepare just enough for the number of wells to be used; discard
remaining solution). Pipet 50 μl of the Ab solution into each
well in a 96 well plate using a stepper (see Note 4). Incubate
over night at 4 C.
2. Blocking: On the following day, remove and discard the first
antibody solution and wash the wells twice with PBST. Add to
each well 200 μl of blocking buffer, and incubate for 2 h at

30 C.
3. Addition of mPCPE-1 samples: Remove and discard the block-
ing buffer. Wash the wells four times with PBST. Tap the plate
gently on the bench to fully remove residual amounts of PBST.
Pipet 50 μl/well of an mPCPE-1 sample diluted in blocking
buffer, in duplicates for the standard curve and in triplicates for
samples with unknown mPCPE-1 concentrations (e.g., diluted
plasma). Assign three wells for determination of basal OD
levels (blanks; all of the reagents except PCPE-1). To these
wells, pipet 50 μl of blocking buffer instead of a PCPE-1
sample. Incubate for 2 h at 30 C (see Note 5).
4. Addition of the second Ab (goat polyclonal antibody to
mPCPE-1). Prepare a 0.1 μg/ml solution of the antibody in
blocking buffer. Remove and discard the PCPE-1 samples, and
wash the wells four times with PBST. Pipet 50 μl of the second
Ab solution into each well using a stepper, and incubate for 1 h
at 30 C.
5. Addition of APA-conjugated rabbit anti goat IgG antibody:
Dilute the antibody 1:2000 in blocking buffer. Remove and
discard the second Ab solution, and wash the wells four times
with PBST, drying the wells by tapping gently on the bench.
Pipet into each well 50 μl of the diluted APA-conjugated
antibody. Incubate for 45 min at 30 C.
6. Addition of APA substrate solution: Remove and discard the
APA-conjugated antibody solution, and wash the wells four
times with PBST. Add to each well 100 μl of a freshly prepared
alkaline phosphatase substrate solution using a stepper. Place
the plate in an ELISA plate reader pre-equilibrated to 30 C
with shaking. Follow the rate of increase in absorbance at
405 nm (due to release of the chromophore para-nitro-phenol)
with constant shaking and recording absorbance at 5 min inter-
vals for about 30 min. Altogether, up to and including this step,
the procedure is accomplished in 2 days.
7. Calculation of mPCPE-1 concentrations: Absorbance values
are converted to mPCPE-1 concentrations expressed as
ng/ml based on the standard curve (Fig. 3a) and corrected
for the dilutions as follows: (1) calculate the average OD/min
value obtained for the blanks and for each PCPE-1 sample;
(2) calculate the net OD/min values for each PCPE-1 concen-
tration by subtracting the average blank value; (3) from the
calibration curve (created using the net average values derived
for known mPCPE-1 concentrations), calculate the mOD/min
value corresponding to 1 ng mPCPE (per well; 50 μl), and use
this value to derive the amount of mPCPE-1 in 50 μl of the
unknown PCPE-1 sample; and (4) calculate concentration first
Fig. 3 Standard curves for determination of mouse and human PCPE-1. (a) Sandwich ELISA for mPCPE-1.
Increasing amounts of purified mPCPE-1 were adsorbed to wells pre-coated with a rat monoclonal antibody to
mPCPE-1. Bound mPCPE-1 was detected using a goat antibody to mPCPE-1 and quantified using an
APA-conjugated rabbit anti goat IgG antibody. (b) Sandwich ELISA for hPCPE-1. Increasing amounts of purified
hPCPE-1 were adsorbed to wells pre-coated with mouse mAb 7A11/1 to hPCPE-1. Bound hPCPE-1 was
detected using a rabbit polyclonal Ab to hPCPE-1 and quantified using an APA-conjugated goat anti rabbit IgG
antibody. Each value in a and b represents mean SD; n ¼ 2. mOD, optical density expressed in milli units
(Adapted from refs. 19 (a) and 32 (b), both Open Access)
in ng/ml (multiply by 20), and then correct for the sample

dilution.
For instance, for a mouse plasma sample diluted 20-fold:
B
X¼ 20 D
A
A ¼ mOD/min/1 ng mPCPE-1 (derived from the calibration
curve).
B ¼ mOD/min for the diluted plasma sample.
D ¼ Plasma dilution.
X ¼ PCPE concentration (ng/ml) of the unknown sample.
The average normal concentration of mouse plasma PCPE-1 is
~200 ng/ml. In CCl4-induced liver fibrosis, it can increase up to
~300 ng/ml, about 50% higher than the control (untreated
mice) [19].
3.2 Sandwich ELISA 1. Adsorption of the first antibody (mAb 7A11/1): Dilute the
for hPCPE-1 antibody solution in carbonate buffer to a final concentration
of 5 μg/ml (see Note 2). Pipet 50 μl of the Ab solution into
each well in a 96 well plate. Incubate over night at 4 C.
2. Blocking: On the following day, remove and discard the first
antibody solution. Wash the wells twice with PBST. Add to
each well 200 μl of blocking buffer, and incubate for 2 h at
30 C.
3. Addition of hPCPE-1 samples: Remove and discard the block-

ing buffer. Wash the wells four times with PBST. Add 50 μl/
well of a hPCPE-1 sample diluted in blocking buffer, in dupli-
cates for the standard curve and in triplicates for samples with
unknown hPCPE-1 concentrations (e.g., diluted human
plasma/serum). Assign three wells for determination of basal
OD levels (blanks; all reagents except for hPCPE-1). To these
wells, pipet 50 μl of blocking buffer instead of a hPCPE-1
sample. Incubate for 2 h at 30 C (see Note 6 for hPCPE-1
sample preparation).
4. Addition of the second Ab (rabbit polyclonal IgG to hPCPE-
1). Prepare a 0.1 μg/ml solution of the antibody in blocking
buffer (see Note 7). Remove and discard the PCPE-1 samples.
Wash the wells four times with PBST. Pipet 50 μl of the second
Ab solution into each well, and incubate for 1 h at 30 C.
5. Addition of APA-conjugated goat anti rabbit IgG: Dilute the
antibody 1:2000 in blocking buffer. Remove and discard the
second Ab solution from the wells. Wash the wells four times
with PBST. Pipet into each well 50 μl of the diluted
APA-conjugated antibody. Incubate for 45 min at 30 C.
6. Addition of APA substrate solution: Remove and discard the
APA-conjugated antibody solution. Wash four times with
PBST. Add to each well 100 μl of a freshly prepared alkaline
phosphatase substrate solution. Place the plate in an ELISA
plate reader pre-equilibrated to 30 C. Follow the rate of
increase in absorbance at 405 nm (due to release of the chro-
mophore para-nitro-phenol) with constant shaking and record-
ing absorbance at 5 min intervals for about 30 min. Altogether,
up to and including this step, the procedure is accomplished in
2 days.
7. Calculate PCPE-1 concentrations based on the standard curve
(Fig. 3b). Absorbance values are converted to hPCPE-1 con-
centrations expressed as ng/ml based on the standard curve
and correcting for the dilutions as follows: (1) calculate the
average OD/min value obtained for the blanks and for each
PCPE-1 sample; (2) calculate the net OD/min values for each
PCPE-1 concentration by subtracting the average blank value;
(3) from the calibration curve (created using the net average
values derived for the known hPCPE-1 concentrations), calcu-
late the mOD/min value corresponding to 1 ng hPCPE/well
(50 μl), and use this value to derive the amount of hPCPE-1 in
50 μl of the unknown PCPE-1 sample; and (4) calculate con-
centration first in ng/ml (multiply by 20), and then correct for
the dilution of the unknown sample.
For instance, for a human plasma sample diluted 20-fold when:

B
X¼ 20 D
A
A ¼ mOD/min/1 ng hPCPE-1.
B ¼ mOD/min for the unknown sample.
D ¼ Dilution of the unknown sample.
X ¼ hPCPE-1 concentration (in ng/ml) of the unknown
sample.
The average plasma concentration of hPCPE-1 in healthy
adults is ~300 ng/ml [28, 32]. In liver fibrosis patients, it may
increase to 450–600 ng/ml (50–100% higher than normal) [32].
4 Notes
1. Purified PCPE-1 is required to create a calibration curve in the

ELISA test. PCPE-1 is a sticky protein. Brij 35 is added to
PCPE-1 solutions of concentrations lower than 0.5 mg/ml to
minimize loss due to PCPE-1 adsorption to the test tube. The
stock solution is stored at 80 C in small aliquots to minimize
repeated thawing and freezing. This is true for both mouse and
human PCPE-1.
2. The commercial 7A1/11 mAb antibody is supplied as the total
IgG fraction, while we used an immuno-affinity-purified IgG
fraction, enriched for hPCPE-1-specific antibodies. When
using a commercial preparation, the amount needed for the
first step in the ELISA should be determined empirically.
3. We used a 4.9 M stock solution of MgCl2 6H2O (Sigma
104-20). This product is discontinued. A 2 M solution is
available (Sigma 68475) and can be used instead. The use of
ready to use MgCl2 solution is recommended because of the
hygroscopic nature of MgCl2.
4. In all of the steps except for the addition of PCPE-1 sam-
ples (step 3), we recommend using automatic pipetting to
speed up the procedure. Blocking (step 2) and washing (after
each incubation step) buffers are added to and removed from
the wells using a multichannel pipettor. The first antibody (step
1), the second antibody (step 4), the APA-conjugated third
antibody (step 5) and the APA substrate (step 6) are all added
using a stepper. The PCPE-1 samples are each pipetted manu-
ally, replacing the tip for each PCPE-1 concentration (step 3).
5. Dilutions of mPCPE-1 samples: Each experiment should
include a standard curve with known mPCPE-1 concentrations
between 0 and 1 ng/well to permit calculation of unknown
mPCPE-1 concentrations. For this purpose, dilute the
Table 1
Preparation of mPCPE-1 solutions with recommended concentrations
(2–20 ng/ml; 0.1–1 ng/50 μl) for a calibration curvea
mPCPE-1 0.5 ng/μl Blocking mPCPE-1 ng/50

Tube in blocking buffer (μl) buffer (μl) μl (well)
1 2 498 0.1
2 4 496 0.2
3 6 494 0.3
4 8 492 0.4
5 10 490 0.5
6 12 488 0.6
7 16 484 0.8
8 20 480 1.0
Total volume for each concentration is 500 μl
a
mPCPE-1 stock solution of 250 μg/ml 1:500 in blocking

buffer to obtain a diluted stock solution of 0.5 ng/μl. This
solution should be used on the same day. The rest should be
discarded. A detailed description on how the 0.5 ng/μl
mPCPE-1 solution is diluted for preparation of mPCPE-1
solutions with known concentrations (0.1–1.0 ng/well) for
the calibration curve is presented in Table 1. Plasma samples
are diluted 1:20 and 1:40 in blocking buffer. PCPE-1 concen-
trations of the diluted plasma samples should be determined in
at least duplicates, preferably, in triplicates. Most importantly,
fresh plasma samples should each be divided into small aliquots
and stored at 80 C until use. Repeated thawing and freezing
should be avoided to prevent loss of the PCPE protein from the
plasma, apparently due to degradation. Thus, once thawed, the
plasma sample should be used only once. The remaining
plasma should be discarded.
6. Dilutions of recombinant hPCPE-1 samples: As in the case of
the ELISA for mPCPE-1, each experiment should include a
calibration curve with known hPCPE-1 concentrations
(0.2–1.2 ng/well). The stock solution of 0.5 mg/ml is diluted
1:1000 in blocking buffer to a final concentration of 0.5 ng/μl.
This solution should be used on the same day. The rest should
be discarded. A detailed description on how the 0.5 ng/μl
hPCPE-1 solution is diluted for preparation of samples with
known hPCPE-1 concentrations required for the calibration
curve is presented in Table 2.
7. Dilutions of the rabbit polyclonal antibody against hPCPE-1:
Dilute the stock antibody solution to a concentration of 1 mg/
Table 2
Preparation of hPCPE-1 solutions with recommended concentrations
(4–24 ng/ml; 0.2–1.2 ng/50 μl) for a calibration curvea
hPCPE-1 0.5 ng/μl Blocking hPCPE-1 ng/50

Tube in blocking buffer (μl) buffer (μl) μl (well)
1 4 496 0.2
2 6 494 0.3
3 8 492 0.4
4 12 488 0.6
5 16 484 0.8
6 20 480 1.0
7 24 476 1.2
a
Total volume for each concentration is 500 μl
ml in PBS. This solution can be stored in aliquots at 20 or

80 C. Further dilutions are prepared in blocking buffer on
the day of the assay, and the remaining diluted solution should
be discarded after use.
References
1. Zeisberg M, Kalluri R (2013) Cellular mechan- for improving accuracy in daily clinical practice.
isms of tissue fibrosis. 1. Common and organ- Ann Hepatol 11:426–439
specific mechanisms associate with tissue fibro- 9. Patel K, Shackel NA (2014) Current status of
sis. Am J Physiol Cell Physiol 304:C216–C225 fibrosis markers. Curr Opin Gastroenterol
2. Wynn TA, Ramalingam TR (2012) Mechan- 30:253–259
isms of fibrosis: therapeutic translation for 10. Gressner OA, Weiskirchen R, Gressner AM
fibrotic disease. Nat Med 18:1028–1040 (2007) Biomarkers of liver fibrosis: clinical
3. Lotersztajn S, Julien B, Teixeira-Clerc F et al translation of molecular pathogenesis or based
(2005) Hepatic fibrosis: molecular mechanisms on liver-dependent malfunction tests. Clin
and drug targets. Annu Rev Pharmacol Toxicol Chim Acta 381:107–113
45:605–628 11. Nallagangula KS, Nagaraj SK, Venkataswamy L
4. Manning DS, Afdhal NH (2008) Diagnosis et al (2018) Liver fibrosis: a compilation on the
and quantitation of fibrosis. Gastroenterology biomarkers status and their significance during
134:1670–1681 disease progression. Future Sci OA 4(1):
5. Papastergiou V, Tsochatzis E, Burroughs AK FSO250
(2012) Non-invasive assessment of liver fibro- 12. Karsdal MA, Krarup H, Sand JMB et al (2014)
sis. Ann Gastroenterol 25:218–231 Review article: the efficacy of biomarkers in
6. Bedossa P, Carrat F (2009) Liver biopsy: the chronic fibroproliferative diseases-early diagno-
best, not the gold standard. J Hepatol 50:1–3 sis and prognosis, with liver fibrosis as an exem-
7. Rockey DC, Caldwell SH, Goodman ZD et al plar. Aliment Pharmacol Ther 40:233–249
(2009) American association for the study of 13. Kessler E, Adar R (1989) Type I procollagen
liver diseases. Liver biopsy. Hepatology C-proteinase from mouse fibroblasts: purifica-
49:1017–1044 tion and demonstration of a 55 kDa enhancer
8. Duarte-Rojo A, Altamirano JT, Feld JJ (2012) glycoprotein. Eur J Biochem 186:115–121
Noninvasive markers of fibrosis: key concepts 14. Kessler E, Takahara K, Biniaminov L et al
(1996) Bone morphogenetic protein-1: The
type I procollagen C-proteinase. Science C-proteinase enhancer-1 in corneal scarring.

271:360–362 Invest Ophthalmol Vis Sci 55:6712–6721
15. Kessler E, Mould PA, Hulmes DJS (1990) 25. Moali C, Font B, Ruggiero F et al (2005)
Procollagen type I C-proteinase enhancer is a Substrate-specific modulation of a multisub-
naturally occurring connective tissue glycopro- strate proteinase. C-terminal processing of
tein. Biochem Biophys Res Commun fibrillar procollagens is the only BMP-1-depen-
173:81–86 dent activity enhanced by PCPE-1. J Biol
16. Takahara K, Kessler E, Biniaminov L et al Chem 280:24188–24194
(1994) Type I procollagen C-proteinase 26. von Marschall Z, Fisher LW (2010) Dentin
enhancer protein: Identification, primary struc- sialophosphoprotein (DSPP) is cleaved into its
ture, and chromosomal localization of the cog- two natural dentin matrix products by three
nate human gene (PCOLCE). J Biol Chem isoforms of bone morphogenetic protein-1
269:26280–26285 (BMP-1). Matrix Biol 29:295–303
17. Ogata I, Auster AS, Matsui A et al (1997) Type 27. Pieper R, Gatin CL, Makusky AJ et al (2003)
I procollagen C-proteinase enhancer protein The human serum proteome: display of nearly
(PCPE) is expressed in cirrhotic but not in 3700 chromatographically separated protein
normal rat liver. Hepatology 26:611–617 spots on two-dimensional electrophoresis gels
18. Ippolito DL, AbdulHameed MDM, Tawa GJ and identification of 325 distinct proteins. Pro-
et al (2016) Gene expression patterns asso- teomics 3:1345–1364
ciated with histopathology in toxic liver fibro- 28. Mesilaty-Gross S, Anikster Y, Vilensky B et al
sis. Toxicol Sci 149:67–88 (2009) Different patterns of human serum pro-
19. Hassoun E, Safrin M, Ziv H et al (2016) Pro- collagen C-proteinase enhancer-1 (PCPE-1).
collagen C-proteinase enhancer 1 (PCPE-1) as Clin Chim Acta 403:76–80
a plasma marker of muscle and liver fibrosis in 29. Olswang-Kuz Y, Liberman B, Weiss I et al
mice. PLoS One 11(7):e0159606 (2011) Quantification of human serum procol-
20. Shalitin N, Schlesinger H, Levy MJ et al (2003) lagen C-proteinase enhancer (hsPCPE) glyco-
Expression of procollagen C-proteinase pattern. Clin Chim Acta 412:1762–1766
enhancer in cultured rat heart fibroblasts: Evi- 30. Grgurevic L, Macek B, Durdevic D et al (2007)
dence for co-regulation with type I collagen. J Detection of bone and cartilage-related pro-
Cell Biochem 90:397–407 teins in plasma of patients with bone fracture
21. Kessler-Icekson G, Schlesinger H, Freimann S using liquid chromatography-mass spectrome-
et al (2006) Expression of procollagen try. Int Orthop 31:743–751
C-proteinase enhancer-1 in the remodeling rat 31. Gokce O, Ozenirler S, Yucel AA et al (2018)
heart is stimulated by aldosterone. Int J Bio- Evaluation of serum procollagen C-proteinase
chem Cell Biol 38:358–365 enhancer 1 level as a fibrosis marker in patients
22. Yu L, Ruifrok WPT, Meissner M et al (2013) with chronic hepatitis B. Eur J Gastroenterol
Genetic and pharmacological inhibition of Hepatol 30(8):918–924. https://doi.org/10.
galectin-3 prevents cardiac remodeling by 1097/MEG.0000000000001123
interfering with myocardial fibrogenesis. Circ 32. Hassoun E, Safrin M, Wineman E et al (2017)
Heart Fail 6:107–117 Data comparing the plasma levels of procolla-
23. Wong VW, You F, Januszyk M et al (2014) gen C-proteinase enhancer 1 (PCPE-1) in
Transcriptional profiling of rapamycin treated healthy individuals and liver fibrosis patients.
fibroblasts from hypertrophic and keloid scars. Data Brief 14:777–781
Ann Plast Surg 72:711–719 33. Moschcovich L, Bernocco S, Font B et al
24. Malecaze F, Massoudi D, Fournié P et al (2001) Folding and activity of recombinant
(2014) Upregulation of bone morphogenetic human procollagen C-proteinase enhancer.
protein-1/mammalian tolloid and procollagen EurJ Biochem 268:2991–2996
Chapter 15
Tumorigenic Interplay Between Macrophages

and Collagenous Matrix in the Tumor Microenvironment
Chen Varol
Abstract
The tumor microenvironment is a heterogeneous tissue that in addition to tumor cells, contain tumor-
associated cell types such as immune cells, fibroblasts, and endothelial cells. Considerably important in the
tumor microenvironment is its noncellular component, namely, the extracellular matrix (ECM). In partic-
ular, the collagenous matrix is subjected to significant alterations in its composition and structure that create
a permissive environment for tumor growth, invasion, and dissemination. Among tumor-infiltrating
immune cells, tumor-associated macrophages (TAMs) are numerous in the tumor stroma and are locally
educated to mediate important biological functions that profoundly affect tumor initiation, growth, and
dissemination. While the influence of TAMs and mechanical properties of the collagenous matrix on tumor
invasion and progression have been comprehensively investigated individually, their interaction within the
complex tumor microenvironment was overlooked. This review summarizes accumulating evidence that
indicate the existence of an intricate tumorigenic crosstalk between TAMs and collagenous matrix. A better
mechanistic comprehension of this reciprocal interplay may open a novel arena for cancer therapeutics.
Key words Tumor-associated macrophages, Tumor microenvironment, Collagenous matrix
1 Introduction
Cancer is not merely a disease of tumor cells, but a multifaceted

show, in which stromal cells and tumor microenvironment play
crucial roles. Imbalance in the noncellular component of the
tumor microenvironment, the ECM, is a characteristic marker of
various solid tumors and is actively involved in their progression
and dissemination [1, 2]. Collagen constitutes the most abundant
scaffold in the tumor microenvironment. During cancer invasion,
the tumor stroma undergoes constant architectural changes, char-
acterized by collagen degradation, redeposition, cross-linking, and
stiffening. These structural transitions yield novel biomechanical
and biophysical cues, which can both inhibit and promote tumor
progression, depending on the stage of cancer development [3].
Coevolving with the tumor ECM are immune cells that constitute
203
204 Chen Varol
the predominant cellular component of the tumor stroma. In par-

ticular, TAMs are abundant in the stroma of experimental and
human tumors and are locally programmed to perform key tumor-
igenic functions [4, 5]. Emerging evidence unravel a pivotal role for
TAMs in the shaping of the dynamically evolving collagenous ECM
within the tumor microenvironment [6]. Reciprocally, alterations
within the tumor collagenous matrix regulate the migration and
function of these phagocytic cells [6]. This overview summarizes
the recent progress in our comprehension of the tumorigenic inter-
play between collagenous matrix and macrophages.
2 Collagen Buildup and Remodeling Promote Tumor Progression, Invasion,

and Metastasis
The tumor cellular ecosystem is nourished by its ECM, comprising

a 3D supramolecular network of proteins, glycoproteins, proteo-
glycans, and polysaccharides. Traditionally, ECM has been
regarded primarily as a physical scaffold that plays a supportive
role in maintaining tissue morphology. However, the tumor ECM
differs considerably from its normal tissue counterpart [7], a prod-
uct of aberrantly expressed or modified structural proteins and
deregulated remodeling events choreographed by specific proteo-
lytic and protein cross-linking enzymes [8]. The ensuing abnormal
ECM actively promotes cancer by providing critical biomechanical
and biochemical cues that drive tumor cell growth, survival, inva-
sion, and metastasis [1, 2, 9].
Force modulates cell fate, directs tissue development, and facil-
itates tumor invasion [10]. Aggressive tumors frequently display
desmoplasia, resulting from enhanced deposition, cross-linking,
and geometrical organization (e.g., linearization) especially of
interstitial collagen fibers that are often positioned perpendicular
to the tumor boundary [10–13]. The resultant increased stiffness
correlates with tumor aggression and an increased propensity
toward metastasis [12, 14, 15]. Cancer and immune cells can
exploit these wrapped stiff collagen bundles as invasion “highways”
[16]. Collagenous ECM stiffening can be enzyme-dependent. In
this case, it is predominantly catalyzed by lysyl oxidases (LOX),
which are capable of cross-linking collagens and elastin, thereby
increasing insoluble matrix deposition and tissue stiffness [17].
Matrix stiffening can also be accomplished by nonenzymatic colla-
gen cross-linking, such as glycation and transglutamination or
increased biglycan and fibromodulin proteoglycan deposition [3].
Specifically, fibronectin polymerization is essential for collagen net-
work formation [18]. The dynamic and reciprocal interactions
between collagen and fibronectin may induce tumor progression
and metastasis. In lung metastases, for example, increased
Tumorigenic Interplay Between Macrophages and Collagenous Matrix. . . 205
fibronectin expression is essential for the recruitment and adher-

ence of pro-metastatic myeloid cells, which express the fibronectin
receptor integrin α4β1 [19]. The secreted protein acidic and rich in
cysteine (SPARC) glycoprotein is an additional example for nonen-
zymatic collagen cross-linking and can either suppress or promote
progression of different cancers [20]. In this respect, orthotopic
pancreatic tumors grown in SPARC-deficient mice display reduced
deposition of fibrillar collagens I and III, basement membrane
collagen IV, and the collagen-associated proteoglycan decorin
[21]. SPARC is expressed by various types of malignant cells and
associated stromal cells and modulates tumor development, inva-
sion, and metastasis [20].
Collagenous matrix buildup has a significant impact on many
hallmarks of cancer [2]. To grow beyond a certain size and metas-
tasize, tumors require angiogenesis, a specialized form of branching
morphogenesis whereby new blood vessels develop from a
pre-existing vascular network [22]. A stiffened collagenous ECM
fosters angiogenesis by providing mechanical signals that facilitate
endothelial cell migration, growth, survival, branching morphol-
ogy, and vessel lumen formation [23–27]. Hypoxia is a key driver of
tumor angiogenesis and can lead to overproduction of LOX-like
protein-2 (LOXL2), which promotes collagen type IV scaffolding
in the endothelial basal lamina, resulting in sprouting angiogenesis
[28]. Interactions between endothelial cells and collagen IV in the
vascular basement membrane play key roles in modulating angio-
genesis [29]. Collagen type IV can directly promote angiogenesis
and neovessel survival in a dose-dependent manner [30]. It is the
triple-helical fragments of intact type IV collagen that stimulate
endothelial cell adhesion and migration [31]. Importantly how-
ever, proteolytic exposure of cryptic sites within specific collagen
subtypes can also release fragments associated with anti-angiogenic
activity. Well-known examples are the non-collagenous domain
1 (NC1 domain) of type IV collagen [32, 33] and the endostatin
fragment of type XVIII collagen [34]. Collagen-rich, stiffened
ECM can also advance cancer growth by impeding anti-
tumorigenic T-cell function. Accordingly, collagen is a ligand for
the inhibitory leukocyte-associated Ig-like receptor (LAIR-1),
which is expressed on various immune cells [35] and has been
linked with suppression of cytotoxic T-cell activity [36]. In another
aspect, cellular quiescence must be overcome to establish a neoplas-
tic lesion. Indeed, mechanical-sensing of stiffer ECM initiates intra-
cellular signaling cascades that promote cell proliferation and
override tumor suppressor activity [12, 37–40]. In addition, the
rigid tumoral ECM also enhances tumor cell invasion and metasta-
sis by inducing focal adhesion assemblies and favoring epithelial-
mesenchymal transition (EMT) [11, 15, 41] and by facilitating the
formation of “premetastatic niches” in destination organs [42, 43].
206 Chen Varol
Collagen proteolysis is a prerequisite for the invasion of tumor

cells through physical barriers that resist tumor expansion into the
surrounding tissue. It is also a rate-limiting step for traversing
metastatic tumor cells trying to exit the blood or lymphatic vessel
across a continuous basement membrane [44]. Several classes of
proteases contribute to ECM breakdown and remodeling, most of
which are upregulated in the course of metastasis in different types
of cancers. Proteolytic enzymes including matrix metalloprotei-
nases (MMPs 1, 2, 13, 14) and cathepsins (B, K, and L) show
activity against fibrillar and non-fibrillar collagens [45–47].
MMPs represent the most prominent family of zinc-dependent
endopeptidases associated with tumorigenesis [46]. Liottta and
colleagues were the first to define tumor cell-derived MMP (now
known as MMP2) that degrades non-fibrillar type IV collagen, the
main constitute of basement membranes [48]. Indeed, higher
levels of basement membrane-degrading MMP activity correlate
with a higher number of metastases in animal models [49]. Mount-
ing evidence supports a dominant role of MMP14 in the migration
and invasion of metastatic tumor cells [50]. Cysteine cathepsins are
another major class of proteolytic enzymes in the tumor microen-
vironment [45] that influence collagenous matrix. Indeed, intracel-
lular lysosomal cathepsin-governed collagen degradation serves as a
major pathway of ECM turnover during malignancy [51]. In par-
ticular, cathepsin B extracted from tumor tissues is capable of
degrading type IV collagen [52]. Imaging has further revealed
that it participates in intracellular digestion of type IV collagen in
human breast cancer and glioma cells [53, 54]. Moreover, degra-
dation of type I collagen by cathepsin B [55] or K [56] facilitates
breast cancer bone metastasis. Secreted cathepsin L can also pro-
mote tumor metastasis through degradation of ECM components
such as laminin, type I and IV collagen, fibronectin, and elastin
[57]. Collectively, these studies highlight that the induction of
collagen deposition and remodeling processes directly facilitate
cancer growth and dissemination by various mechanisms.
3 TAMs Support Cancer Growth and Dissemination
Epidemiological meta-analyses indicate a clear correlation between

the density of TAMs and poor prognosis in many, but not all,
human cancers [58, 59]. TAMs represent a major component of
the lymphoreticular infiltrates in solid tumors and prominently arise
from classical Ly6Chi monocytes [60, 61], which arrive in response
to tumor cell-derived growth factors and chemoattractants, espe-
cially colony-stimulating factor 1 (CSF-1) (reviewed in [62]) and
CCL2 [60, 61, 63], respectively. They respond to local cues within
the tumor microenvironment that usually bias away their functional
polarization from the classically activated (M1) macrophage
phenotype to the alternatively activated type termed “M2” [64].

Yet, TAMs are considered a heterogeneous cell population with
multidimensional functional plasticity and often exhibit markers
that are characteristic of both types. They perform various
pro-tumoral functions that effect every hallmark of cancer progres-
sion [65]. For example, TAMs acquire M2-like properties that
qualify them to subvert adaptive immunity [66]. This is typically
manifested by production of the anti-inflammatory cytokines IL-10
and TGFβ [67] and the release of chemokines that preferentially
attract T lymphocytes devoid of cytotoxic functions [68]. More-
over, TAMs orchestrate the so-called angiogenic switch required
for malignant transition by producing neoangiogenic molecules
that increase vascular density [69]. They are inclined to cluster in
“hot spots” in hypoxic avascular areas with a high level of angio-
genesis [70], where their pro-angiogenic activity is further aug-
mented [71, 72]. TAMs also play pivotal roles in the metastatic
process. Specifically, a CSF1-epidermal growth factor (EGF) para-
crine loop between TAMs and tumor cells has been implicated in
the promotion of breast cancer invasion and dissemination
[73, 74]. In addition, monocytes recruited via the CCR2-CCL2
axis to the premetastatic niche create an environment optimized for
rapid development of metastases [19, 75–78].
4 TAMs Are Builders of the Tumorigenic Collagenous Matrix
The role of TAMs or collagenous matrix remodeling in tumor

development and dissemination has each been the subject in a
plethora of studies. Nevertheless, studies on how these pathways
are intertwined to advance tumorigenesis are only now emerging.
Early reports have demonstrated the accumulation of macrophages
and their association with a fibrous cellular capsule formed around
s.c. implanted mammary carcinoma tumors [79]. A follow-up study
has further provided initial evidence for their capacity to function as
collagen-producing cells during tumor encapsulation [80]. Intravi-
tal imaging of mammary tumors has also demonstrated the accu-
mulation of TAMs at the tumor’s collagen-rich border, where they
support the intravasation of cancer cells [16]. These observations
were the first to introduce the idea that TAMs may be involved with
tumorigenic collagen remodeling, an idea that has been affirmed by
a group of recent studies discussed below.
Utilizing an orthotopic colorectal cancer (CRC) mouse model,
we have found that Ly6Chi monocyte-derived TAMs play a key role
in the buildup of the tumorigenic collagenous niche by directly
contributing to the deposition, cross-linking, and linearization of
fibrillar collagens [60]. Complementary transcriptomic and proteo-
mic analyses revealed that Ly6Chi monocytes directly produce type
I and VI collagen in the tumor milieu, and upon their maturation
208 Chen Varol
into TAMs, they further upregulate the expression of type XIV

collagen. The latter belongs to a family of fibril-associated collagens
and interacts predominantly with type I collagen to promote fibril-
logenesis [81, 82]. It appears that TAMs in this CRC model play a
dominant role in the deposition of these collagen subtypes given
their reduced representation in TAM-deficient vs. TAM-sufficient
tumors [60]. Remarkably, CRC TAMs also express various mole-
cules that participate in posttranslational modifications of intracel-
lular collagen and its assembly, such as the glycoprotein procollagen
C-endopeptidase enhancer (PCOLCE), the enzyme prolyl
4-hydroxylase α polypeptide I (P4HA1), the collagen cross-linkers
procollagen-lysine 2-oxoglutarate 5-dioxygenase-1 and
procollagen-lysine 2-oxoglutarate 5-dioxygenase-3 (PLOD-1,
PLOD-3), the proteoglycan biglycan, and the glycoprotein
SPARC [60]. The expression of the latter has also been detected
in TAMs sorted from human ovarian carcinoma [83] as well as from
experimental and human pancreatic ductal adenocarcinoma
(PDAC) [84]. Other studies have highlighted a key involvement
for immune cell-derived SPARC in the assembly and organization
of collagenous matrix in pancreatic [85] and mammary tumors
[86]. Moreover, macrophage-derived SPARC promotes cancer
cell migration and metastasis by modulating integrin-ECM
interactions [87].
While our study reports a pro-fibrotic role for CCR2+Ly6Chi
monocyte-derived TAMs in CRC [60], a recent study uncovered
that in PDAC this function is in fact associated with tissue-resident
embryonic yolk-sac-derived macrophages [84]. Accordingly, in the
settings of an orthotopic KPC model of PDAC, the embryonic-
derived TAMs, rather than the monocyte-derived TAMs, exhibit a
pro-fibrotic profile as manifested by their increased transcription of
collagen types Ia2, IIIa1, Va1, VIa1, Xa1, and XVIIIa1, their
augmented ex vivo production of type I and IV collagen, and the
reduced collagen density in PDAC tumors following their inducible
ablation [84]. Notably, their expression of collagen types IVa4,
Xa1, XVIIa1, and XVIIIa1 was even higher than cancer-associated
fibroblasts (CAFs). A similar pro-fibrotic phenotype was also found
in human embryonic-derived CXCR4+ macrophages isolated from
PDAC samples, suggesting conservation of function across species
[84]. These two studies [60, 84] underscore the importance of
understanding the task division between TAMs of different onto-
genies in different tumor types.
An association between TAM presence and tumorigenic colla-
gen remodeling is also supported by other studies. In a model of
breast cancer, depletion of TAMs alters tumor collagen fibrillar
microstructures that are important for tumor metastasis [88].
Moreover, TAMs extracted from human ovarian carcinoma express
genes encoding for types I, V, and VI collagen [83]. The synthesis
of the latter has also been demonstrated in monocytes associated
with triple-negative breast cancer xenografts [89] and TGFβ1-sti-

mulated human macrophages [90]. The ability of macrophages to
acquire a fibroblast-like function and directly synthesize specific
collagen subtypes is further supported by studies in other patho-
logical settings, such as renal fibrosis [91–93], atherosclerotic pla-
ques [94, 95], fibrosis and diastolic dysfunction in the aging heart
[96], and helminth infection [97], and very recently also in
ERK-activated brain microglia in the context of neurodegenerative
disease [98] and during wound granulation in a process mediated
by miR-21 [99]. Macrophages (hemocytes) also emerge as a major
source of basement membrane collagen IV in the developing Dro-
sophila embryo [100].
CAFs are considered to be the key producers of collagenous
matrix in developing tumors [101]. Accumulating evidence indi-
cates that TAMs can also elicit the pro-fibrotic activity of CAFs. In
support of this, we have observed reduced gene expression of
collagen type I in CAFs sorted from TAM-deficient colorectal
tumors [60]. In another example, the addition of monocytes to
triple-negative breast cancer xenografts facilitated collagen deposi-
tion by stromal CAFs [89]. TAMs can also serve as a reservoir of
growth factors and cytokines that foster the pro-fibrotic activity of
CAFs, such as the platelet-derived growth factor (PDGF) [102],
epidermal growth factor (EGF) [73], and TGFβ [103]. With
respect to the latter, Ly6Chi monocyte-derived TGFβ1 has been
shown to facilitate the transition of hepatic stellate cells into
collagen-producing myofibroblasts in a model of liver fibrosis
[104]. Similarly, M2 macrophage-derived TGFβ1 induces the pro-
liferation of fibroblasts and their differentiation into myofibroblasts
during wound healing scar formation [105]. Macrophages may also
engage the activity of the LOX collagen cross-linker enzyme in
adjacent fibroblasts as has recently been shown in a model of skin
injury [106]. Finally, utilizing a computational two-cell system, the
Medzhitov and Alon groups have recently discovered that macro-
phages and fibroblasts form a stable cell circuit, which is robust to
perturbations owing to cell-contact-dependent mutual exchange of
growth factors (PDGF and CSF-1, respectively) and negative feed-
back regulation [107]. These types of cell circuits may be important
to stabilize fibrogenic interactions between TAMs and CAFs in the
tumor microenvironment. Collectively, these reports outline a piv-
otal role for TAMs in the tumorigenic collagenous matrix buildup
through their direct deposition of specific types of collagen or via
their governed promotion or stabilization of CAF’s pro-fibrotic
activity.
210 Chen Varol
5 TAMs Are Remodelers of the Tumorigenic Collagenous Matrix
Collagen remodeling events, including collagen cross-linking and

collagenolysis, are typical of various solid tumors and direct tumor
development, invasion, and dissemination through physical barriers
[1, 2, 44]. MMPs represent the most studied and prominent family
of matrix proteinases associated with tumorigenesis [46]. Interstitial
collagens are hydrolyzed by the “classic” collagenases MMP1,
MMP8, MMP13, and MT1-MMP (MMP14). MMP2 digests solu-
bilized monomers of collagens I, II, and III, and MMP9 cleaves
solubilized collagen I and III monomers [108]. The basement
membrane is dominated by a scaffolding of cross-linked type IV
collagen molecules, and its breaching during malignancy is
mediated mainly by MMP2 and MMP9 [109, 110] as well as the
membrane-anchored proteases MMP14, MT2-MMP (MMP15),
and MT3-MMP (MMP16) [111]. TAMs are considered to be a
major source of collagen remodeling enzymes in the tumor micro-
environment [6, 112]. With respect to MMP9, Coussens et al. have
initially reported that MMP9 expression by tumor-infiltrating
BM-derived hematopoietic cells promotes metastatic growth
[113]. Further studies have attributed MMP9-governed angiogen-
esis and tumorigenicity to TAMs [114–117]. Yet, other studies
have highlighted tumor-associated neutrophils as the predominant
and immediate source of active MMP9, which is uniquely unen-
cumbered by tissue inhibitor of metalloproteinases-1 (TIMP-1)
[118]. We have shown that Ly6Chi monocytes recruited to ortho-
topically implanted colorectal tumors upregulate the expression of
MMP12, MMP13, and MMP14 [60]. Recently, Madsen et al.
reported that CCR2+Ly6Chi monocytes acquire a collagen cata-
bolic transcriptomic signature during invasive tumor growth and
degrade collagen through mannose receptor-dependent cellular
uptake [119]. These collagen-endocytosing TAMs expressed
genes encoding for MMP8, MMP9, MMP12, and MMP13.
CAFs on the other hand expressed higher levels of MMP2 and its
activator MMP14, suggesting their possible association with initial
collagen fragmentation [119]. Noteworthy, the proteolytic activity
of MMPs can be regulated at multiple levels, including transcrip-
tion, compartmentalization, conversion from zymogen to active
enzyme, and restraining by endogenous inhibitors such as TIMPs.
Therefore, when judging the pathophysiological relevance of MMP
expression in TAMs, their collagenase activity has to be verified.
Another family of matrix proteases with collagen remodeling
activity is the cysteine cathepsins, which function in proteolytic
pathways that increase neoplastic progression. These enzymes can
be secreted, localized at the cell-surface or intracellularly in associa-
tion with lysosomes [45, 47]. In TAMs, cathepsins may drive
lysosomal degradation of internalized collagens [119]. Indeed,
collagen-endocytosing TAMs express a wide panel of cathepsin

transcripts including B, while L and K are higher in CAFs
[119]. Gene and protein expression of cathepsins B, D, and L has
also been noticed in CRC TAMs [60]. Moreover, TAM polariza-
tion by tumor-derived IL-4 induced the activity of cathepsins B and
S that promote pancreatic tumor growth, angiogenesis, and inva-
sion [120]. TAM-derived cathepsin B has also been implicated in
the dissemination of breast cancer metastases to the lung
[121]. Further studies are required to determine the functional
contribution of specific TAM-derived cathepsins to the remodeling
of the collagenous tumorigenic niche. TAMs may also directly
promote collagen cross-linking via their provision of cross-linker
enzymes. In CRC, we have shown enhanced expression of the
ECM covalent cross-linker enzymes transglutaminase-I and
transglutaminase-II and coagulation factor XIII subunit a1, which
were also reduced in whole tumor tissue of TAM-deficient colorec-
tal tumors [60].
6 Abnormal Tumorigenic Collagenous Matrix May Reciprocally Modulate TAM

Recruitment and Function
As summarized above, accumulating evidence indicate that TAMs

contribute to the increased matrix rigidity in various solid tumors
through their direct and indirect modulation of collagen deposition
and its geometrical organization. These kinds of alterations in
collagenous matrix may in turn affect the migration behavior and
function of TAMs. The ability of macrophages to sense the ECM in
their vicinity relies on mechanosensors such as integrins, which
transmit force from the extracellular environment via their interac-
tion with numerous cytoskeletal and signaling proteins [122].
Force sensing is transmitted from the extracellular environment
via the interaction of integrins with numerous cytoskeletal and
signaling proteins that accumulate in a force-dependent manner
into focal adhesions and other types of cell-matrix adhesions [122].
A specific example is the integrin αvβ3, which binds various matrix
proteins including vitronectin, fibronectin, fibrinogen, as well as
denatured or proteolysed collagen and can affect tumor growth by
various ways [123]. Collagen and its digestion products can act as
chemotactic stimuli for monocytes and macrophages. Accordingly,
activation of the collagen tyrosine kinase receptor discoidin domain
receptor 1 (DDR1) plays a critical role in macrophage accumula-
tion at early and late stages of atherogenesis [124], and type I
collagen degradation products might directly serve as chemotactic
stimuli for human peripheral blood monocytes [125]. Moreover,
LOX-governed collagen matrix modifications promote the recruit-
ment, invasion, and retention of CD11b+ myeloid cells to
212 Chen Varol
premetastatic sites [42]. Biophysical parameters of the collagenous

matrix can also instruct the 3D migration mode of macrophages.
Macrophages can apply two main migration modes while moving
through 3D matrices: amoeboid or mesenchymal [126]. In fibrillar
type I collagen, macrophages tend to use the amoeboid migration
mode, while in denser collagen I matrices, they have a preference
for the mesenchymal migration mode [127]. In case of the latter,
they may form 3D collagenolytic podosome-like structures at the
tips of cell protrusions [127]. In addition, a dense collagenous
matrix may also dictate their pro-tumoral polarization. Indeed,
human monocytes grown on collagen-rich matrix exhibit increased
proliferation and a pro-tumorigenic M2 polarization state
[128, 129], and increased matrix rigidity promotes M2 polariza-
tion of macrophages [130]. Altogether, the altered collagenous
matrix deposition and architecture in the tumor microenvironment
may be essential for the regulation of TAM recruitment, retention,
intra-tumor tissue mobility, and tumorigenic activity. Further stud-
ies are required to determine the type of mechanosensors used by
TAMs to sense the dynamically changing collagenous matrix during
tumor development and their governed molecular programming of
TAM migration and behavior.
7 Concluding Remarks
It is becoming conceded that TAMs fundamentally shape the

tumorigenic collagenous matrix. They can do so by direct deposi-
tion of specific subtypes of collagen and instruction of collagen
cross-linking and linearization, as well as by encouraging the
pro-fibrotic activity of co-localizing CAFs. TAMs also bring into
the tumor microenvironment a unique repertoire of collagen remo-
deling enzymes that are capable of orchestrating collagenous matrix
rearrangement and allow the invasion and dissemination of cancer
cells through physical barriers. In turn, TAMs can sense the dyna-
mically evolving abnormal collagenous matrix and respond by
changing their migration mode and function. This kind of interplay
appears to be indispensable in generating a tumor microenviron-
ment permissive to tumor growth, invasion, and metastatic dissem-
ination. While recent studies have advanced our comprehension of
TAM-mediated collagen remodeling, they remain so far largely
descriptive. Formidable challenges remain to delineate the mecha-
nistic relevance of TAM-derived collagenous matrix deposition,
especially with respect to its biophysical, biochemical, and struc-
tural properties. This includes a better definition of the specific
subtypes of collagen produced by resident versus Ly6Chi
monocyte-derived TAMs in comparison with other collagen-
producing cells such as CAFs. In this respect, there is still a great
unmet need for understanding how TAM-elicited changes in the
structure and composition of collagenous matrix and the ensuing

proteolysis products influence the behavior of different target cells
in the tumor microenvironment. Another remaining challenge is to
unravel the exact enzymatic and nonenzymatic pathways used by
TAMs to instruct collagen assembly, cross-linking, and lineariza-
tion. Such a functional molecular view of TAM-derived collagen
remodeling requires a multidisciplinary integrated research
approach that encompasses various methods. For example, the
combination of specific macrophage ablation tools with optical
imaging and electron microscopy covering wide ranges of spatial
and temporal resolutions may allow in-depth characterization of
TAM-guided (direct or indirect) structural alterations within the
native tumor collagenous scaffolds during distinct developmental
phases. Subsequently, image-guided, site-directed proteomics
might be utilized to reveal collagen composition in defined niches
of the tumor occupied by specific TAM subsets. Another challenge
would be to study how abnormal collagenous matrix affect the
behavior of TAMs in terms of mobility and function. This calls for
studies that would map the expression of specific mechanosensors
on TAM subsets with correlation to the spatial availability of specific
collagen ligands in distinct tumor niches and during distinct devel-
opmental phases. Such studies can benefit from transgenic systems
that target deficiency or overexpression of specific mechanosensors
to distinct TAM subsets and their complementary molecular
profiling using transcriptomic and proteomic approaches. Finally,
ex vivo platforms based on decellularized native 3D-ECM tumor
scaffolds supplemented with specific TAM subsets will enable cou-
pling of remodeled collagen macromolecule architecture and com-
position to functional cell-macromolecular interactions.
Collectively, mechanistic insights derived from such analyses may
be further used to rationalize novel agents targeting TAM-collagen
remodeling reactions as diagnostic or therapeutic agents.
References
1. Lu P, Weaver VM, Werb Z (2012) The extra- 4. Biswas SK, Allavena P, Mantovani A (2013)
cellular matrix: a dynamic niche in cancer pro- Tumor-associated macrophages: functional
gression. J Cell Biol 196(4):395–406. diversity, clinical significance, and open ques-
https://doi.org/10.1083/jcb.201102147 tions. Semin Immunopathol 35(5):585–600.
2. Pickup MW, Mouw JK, Weaver VM (2014) https://doi.org/10.1007/s00281-013-
The extracellular matrix modulates the hall- 0367-7
marks of cancer. EMBO Rep 15 5. Noy R, Pollard JW (2014) Tumor-associated
(12):1243–1253. https://doi.org/10. macrophages: from mechanisms to therapy.
15252/embr.201439246 Immunity 41(1):49–61. https://doi.org/10.
3. Fang M, Yuan J, Peng C, Li Y (2014) Colla- 1016/j.immuni.2014.06.010
gen as a double-edged sword in tumor pro- 6. Varol C, Sagi I (2018) Phagocyte-
gression. Tumour Biol 35(4):2871–2882. extracellular matrix crosstalk empowers
https://doi.org/10.1007/s13277-013- tumor development and dissemination.
1511-7 FEBS J 285(4):734–751. https://doi.org/
10.1111/febs.14317
214 Chen Varol
7. Naba A, Clauser KR, Hoersch S, Liu H, Carr 16. Wyckoff JB, Wang Y, Lin EY, Li JF,
SA, Hynes RO (2012) The matrisome: in Goswami S, Stanley ER, Segall JE, Pollard
silico definition and in vivo characterization JW, Condeelis J (2007) Direct visualization
by proteomics of normal and tumor extracel- of macrophage-assisted tumor cell intravasa-
lular matrices. Mol Cell Proteomics 11(4): tion in mammary tumors. Cancer Res 67
M111 014647. https://doi.org/10.1074/ (6):2649–2656. https://doi.org/10.1158/
mcp.M111.014647 0008-5472.CAN-06-1823
8. Mason SD, Joyce JA (2011) Proteolytic net- 17. Xiao Q, Ge G (2012) Lysyl oxidase, extracel-
works in cancer. Trends Cell Biol 21 lular matrix remodeling and cancer metastasis.
(4):228–237. https://doi.org/10.1016/j. Cancer Microenviron 5(3):261–273. https://
tcb.2010.12.002 doi.org/10.1007/s12307-012-0105-z
9. Hoye AM, Erler JT (2016) Structural ECM 18. Velling T, Risteli J, Wennerberg K, Mosher
components in the premetastatic and meta- DF, Johansson S (2002) Polymerization of
static niche. Am J Physiol Cell Physiol 310 type I and III collagens is dependent on fibro-
(11):C955–C967. https://doi.org/10. nectin and enhanced by integrins alpha
1152/ajpcell.00326.2015 11beta 1 and alpha 2beta 1. J Biol Chem
10. Butcher DT, Alliston T, Weaver VM (2009) A 277(40):37377–37381. https://doi.org/10.
tense situation: forcing tumour progression. 1074/jbc.M206286200
Nat Rev Cancer 9(2):108–122. https://doi. 19. Kaplan RN, Riba RD, Zacharoulis S, Bramley
org/10.1038/nrc2544 AH, Vincent L, Costa C, MacDonald DD, Jin
11. Levental KR, Yu H, Kass L, Lakins JN, DK, Shido K, Kerns SA, Zhu Z, Hicklin D,
Egeblad M, Erler JT, Fong SF, Csiszar K, Wu Y, Port JL, Altorki N, Port ER,
Giaccia A, Weninger W, Yamauchi M, Gasser Ruggero D, Shmelkov SV, Jensen KK,
DL, Weaver VM (2009) Matrix crosslinking Rafii S, Lyden D (2005) VEGFR1-positive
forces tumor progression by enhancing integ- haematopoietic bone marrow progenitors ini-
rin signaling. Cell 139(5):891–906. https:// tiate the pre-metastatic niche. Nature 438
doi.org/10.1016/j.cell.2009.10.027 (7069):820–827. https://doi.org/10.1038/
12. Provenzano PP, Inman DR, Eliceiri KW, Knit- nature04186
tel JG, Yan L, Rueden CT, White JG, Keely PJ 20. Chlenski A, Cohn SL (2010) Modulation of
(2008) Collagen density promotes mammary matrix remodeling by SPARC in neoplastic
tumor initiation and progression. BMC Med progression. Semin Cell Dev Biol 21
6:11. https://doi.org/10.1186/1741-7015- (1):55–65. https://doi.org/10.1016/j.
6-11 semcdb.2009.11.018
13. Paszek MJ, Zahir N, Johnson KR, Lakins JN, 21. Arnold SA, Rivera LB, Miller AF, Carbon JG,
Rozenberg GI, Gefen A, Reinhart-King CA, Dineen SP, Xie Y, Castrillon DH, Sage EH,
Margulies SS, Dembo M, Boettiger D, Ham- Puolakkainen P, Bradshaw AD, Brekken RA
mer DA, Weaver VM (2005) Tensional (2010) Lack of host SPARC enhances vascular
homeostasis and the malignant phenotype. function and tumor spread in an orthotopic
Cancer Cell 8(3):241–254. https://doi.org/ murine model of pancreatic carcinoma. Dis
10.1016/j.ccr.2005.08.010 Model Mech 3(1–2):57–72. https://doi.
14. Barcus CE, O’Leary KA, Brockman JL, org/10.1242/dmm.003228
Rugowski DE, Liu Y, Garcia N, Yu M, Keely 22. Carmeliet P, Jain RK (2000) Angiogenesis in
PJ, Eliceiri KW, Schuler LA (2017) Elevated cancer and other diseases. Nature 407
collagen-I augments tumor progressive sig- (6801):249–257. https://doi.org/10.1038/
nals, intravasation and metastasis of 35025220
prolactin-induced estrogen receptor alpha 23. Liu J, Agarwal S (2010) Mechanical signals
positive mammary tumor cells. Breast Cancer activate vascular endothelial growth factor
Res 19(1):9. https://doi.org/10.1186/ receptor-2 to upregulate endothelial cell pro-
s13058-017-0801-1 liferation during inflammation. J Immunol
15. Pickup MW, Laklai H, Acerbi I, Owens P, 185(2):1215–1221. https://doi.org/10.
Gorska AE, Chytil A, Aakre M, Weaver VM, 4049/jimmunol.0903660
Moses HL (2013) Stromally derived lysyl oxi- 24. Montesano R, Orci L, Vassalli P (1983) In
dase promotes metastasis of transforming vitro rapid organization of endothelial cells
growth factor-beta-deficient mouse mam- into capillary-like networks is promoted by
mary carcinomas. Cancer Res 73 collagen matrices. J Cell Biol 97(5 Pt
(17):5336–5346. https://doi.org/10.1158/ 1):1648–1652
0008-5472.CAN-13-0012
25. Ingber D, Folkman J (1988) Inhibition of 34. Kim YM, Jang JW, Lee OH, Yeon J, Choi EY,
angiogenesis through modulation of collagen Kim KW, Lee ST, Kwon YG (2000) Endosta-
metabolism. Lab Investig 59(1):44–51 tin inhibits endothelial and tumor cellular
26. Myers KA, Applegate KT, Danuser G, Fischer invasion by blocking the activation and cata-
RS, Waterman CM (2011) Distinct ECM lytic activity of matrix metalloproteinase. Can-
mechanosensing pathways regulate microtu- cer Res 60(19):5410–5413
bule dynamics to control endothelial cell 35. Meyaard L (2008) The inhibitory collagen
branching morphogenesis. J Cell Biol 192 receptor LAIR-1 (CD305). J Leukoc Biol 83
(2):321–334. https://doi.org/10.1083/jcb. (4):799–803. https://doi.org/10.1189/jlb.
201006009 0907609
27. Newman AC, Nakatsu MN, Chou W, Ger- 36. Meyaard L, Hurenkamp J, Clevers H, Lanier
shon PD, Hughes CC (2011) The require- LL, Phillips JH (1999) Leukocyte-associated
ment for fibroblasts in angiogenesis: Ig-like receptor-1 functions as an inhibitory
fibroblast-derived matrix proteins are essential receptor on cytotoxic T cells. J Immunol 162
for endothelial cell lumen formation. Mol Biol (10):5800–5804
Cell 22(20):3791–3800. https://doi.org/10. 37. Rosenfeldt H, Grinnell F (2000) Fibroblast
1091/mbc.E11-05-0393 quiescence and the disruption of ERK signal-
28. Bignon M, Pichol-Thievend C, Hardouin J, ing in mechanically unloaded collagen matri-
Malbouyres M, Brechot N, Nasciutti L, ces. J Biol Chem 275(5):3088–3092
Barret A, Teillon J, Guillon E, Etienne E, 38. Provenzano PP, Inman DR, Eliceiri KW,
Caron M, Joubert-Caron R, Monnot C, Keely PJ (2009) Matrix density-induced
Ruggiero F, Muller L, Germain S (2011) mechanoregulation of breast cell phenotype,
Lysyl oxidase-like protein-2 regulates sprout- signaling and gene expression through a
ing angiogenesis and type IV collagen assem- FAK-ERK linkage. Oncogene 28
bly in the endothelial basement membrane. (49):4326–4343. https://doi.org/10.1038/
Blood 118(14):3979–3989. https://doi. onc.2009.299
org/10.1182/blood-2010-10-313296 39. Tilghman RW, Cowan CR, Mih JD,
29. Kalluri R (2003) Basement membranes: struc- Koryakina Y, Gioeli D, Slack-Davis JK, Black-
ture, assembly and role in tumour angiogene- man BR, Tschumperlin DJ, Parsons JT
sis. Nat Rev Cancer 3(6):422–433. https:// (2010) Matrix rigidity regulates cancer cell
doi.org/10.1038/nrc1094 growth and cellular phenotype. PLoS One 5
30. Bonanno E, Iurlaro M, Madri JA, Nicosia RF (9):e12905. https://doi.org/10.1371/jour
(2000) Type IV collagen modulates angio- nal.pone.0012905
genesis and neovessel survival in the rat aorta 40. Mouw JK, Yui Y, Damiano L, Bainer RO,
model. In Vitro Cell Dev Biol Anim 36 Lakins JN, Acerbi I, Ou G, Wijekoon AC,
(5):336–340. https://doi.org/10.1290/ Levental KR, Gilbert PM, Hwang ES, Chen
1071-2690(2000)036<0336:TICMAA>2. YY, Weaver VM (2014) Tissue mechanics
0.CO;2 modulate microRNA-dependent PTEN
31. Herbst TJ, McCarthy JB, Tsilibary EC, expression to regulate malignant progression.
Furcht LT (1988) Differential effects of lami- Nat Med 20(4):360–367. https://doi.org/
nin, intact type IV collagen, and specific 10.1038/nm.3497
domains of type IV collagen on endothelial 41. Leight JL, Wozniak MA, Chen S, Lynch ML,
cell adhesion and migration. J Cell Biol 106 Chen CS (2012) Matrix rigidity regulates a
(4):1365–1373 switch between TGF-beta1-induced apopto-
32. Sudhakar A, Nyberg P, Keshamouni VG, sis and epithelial-mesenchymal transition.
Mannam AP, Li J, Sugimoto H, Cosgrove D, Mol Biol Cell 23(5):781–791. https://doi.
Kalluri R (2005) Human alpha1 type IV col- org/10.1091/mbc.E11-06-0537
lagen NC1 domain exhibits distinct antian- 42. Erler JT, Bennewith KL, Cox TR, Lang G,
giogenic activity mediated by alpha1beta1 Bird D, Koong A, Le QT, Giaccia AJ (2009)
integrin. J Clin Invest 115(10):2801–2810. Hypoxia-induced lysyl oxidase is a critical
https://doi.org/10.1172/JCI24813 mediator of bone marrow cell recruitment to
33. Petitclerc E, Boutaud A, Prestayko A, Xu J, form the premetastatic niche. Cancer Cell 15
Sado Y, Ninomiya Y, Sarras MP Jr, Hudson (1):35–44. https://doi.org/10.1016/j.ccr.
BG, Brooks PC (2000) New functions for 2008.11.012
non-collagenous domains of human collagen 43. Bondareva A, Downey CM, Ayres F, Liu W,
type IV. Novel integrin ligands inhibiting Boyd SK, Hallgrimsson B, Jirik FR (2009)
angiogenesis and tumor growth in vivo. J The lysyl oxidase inhibitor, beta-
Biol Chem 275(11):8051–8061 aminopropionitrile, diminishes the metastatic
216 Chen Varol
colonization potential of circulating breast 55. Withana NP, Blum G, Sameni M, Slaney C,
cancer cells. PLoS One 4(5):e5620. https:// Anbalagan A, Olive MB, Bidwell BN,
doi.org/10.1371/journal.pone.0005620 Edgington L, Wang L, Moin K, Sloane BF,
44. Liotta LA (2016) Adhere, degrade, and move: Anderson RL, Bogyo MS, Parker BS (2012)
the three-step model of invasion. Cancer Res Cathepsin B inhibition limits bone metastasis
76(11):3115–3117. https://doi.org/10. in breast cancer. Cancer Res 72
1158/0008-5472.CAN-16-1297 (5):1199–1209. https://doi.org/10.1158/
45. Gocheva V, Joyce JA (2007) Cysteine cathe- 0008-5472.CAN-11-2759
psins and the cutting edge of cancer invasion. 56. Duong LT, Wesolowski GA, Leung P,
Cell Cycle 6(1):60–64. https://doi.org/10. Oballa R, Pickarski M (2014) Efficacy of a
4161/cc.6.1.3669 cathepsin K inhibitor in a preclinical model
46. Kessenbrock K, Plaks V, Werb Z (2010) for prevention and treatment of breast cancer
Matrix metalloproteinases: regulators of the bone metastasis. Mol Cancer Ther 13
tumor microenvironment. Cell 141 (12):2898–2909. https://doi.org/10.1158/
(1):52–67. https://doi.org/10.1016/j.cell. 1535-7163.MCT-14-0253
2010.03.015 57. Sudhan DR, Siemann DW (2015) Cathepsin
47. Mohamed MM, Sloane BF (2006) Cysteine L targeting in cancer treatment. Pharmacol
cathepsins: multifunctional enzymes in can- Ther 155:105–116. https://doi.org/10.
cer. Nat Rev Cancer 6(10):764–775. 1016/j.pharmthera.2015.08.007
https://doi.org/10.1038/nrc1949 58. Bingle L, Brown NJ, Lewis CE (2002) The
48. Liotta LA, Abe S, Robey PG, Martin GR role of tumour-associated macrophages in
(1979) Preferential digestion of basement tumour progression: implications for new
membrane collagen by an enzyme derived anticancer therapies. J Pathol 196
from a metastatic murine tumor. Proc Natl (3):254–265. https://doi.org/10.1002/
Acad Sci U S A 76(5):2268–2272 path.1027
49. Liotta LA, Tryggvason K, Garbisa S, Hart I, 59. Zhang QW, Liu L, Gong CY, Shi HS, Zeng
Foltz CM, Shafie S (1980) Metastatic poten- YH, Wang XZ, Zhao YW, Wei YQ (2012)
tial correlates with enzymatic degradation of Prognostic significance of tumor-associated
basement membrane collagen. Nature 284 macrophages in solid tumor: a meta-analysis
(5751):67–68 of the literature. PLoS One 7(12):e50946.
https://doi.org/10.1371/journal.pone.
50. Sabeh F, Shimizu-Hirota R, Weiss SJ (2009) 0050946
Protease-dependent versus -independent can-
cer cell invasion programs: three-dimensional 60. Afik R, Zigmond E, Vugman M, Klepfish M,
amoeboid movement revisited. J Cell Biol Shimshoni E, Pasmanik-Chor M, Shenoy A,
185(1):11–19. https://doi.org/10.1083/ Bassat E, Halpern Z, Geiger T, Sagi I, Varol C
jcb.200807195 (2016) Tumor macrophages are pivotal con-
structors of tumor collagenous matrix. J Exp
51. Curino AC, Engelholm LH, Yamada SS, Med 213(11):2315–2331. https://doi.org/
Holmbeck K, Lund LR, Molinolo AA, 10.1084/jem.20151193
Behrendt N, Nielsen BS, Bugge TH (2005)
Intracellular collagen degradation mediated 61. Franklin RA, Liao W, Sarkar A, Kim MV,
by uPARAP/Endo180 is a major pathway of Bivona MR, Liu K, Pamer EG, Li MO
extracellular matrix turnover during malig- (2014) The cellular and molecular origin of
nancy. J Cell Biol 169(6):977–985. https:// tumor-associated macrophages. Science 344
doi.org/10.1083/jcb.200411153 (6186):921–925. https://doi.org/10.1126/
science.1252510
52. Buck MR, Karustis DG, Day NA, Honn KV,
Sloane BF (1992) Degradation of 62. Pollard JW (2004) Tumour-educated macro-
extracellular-matrix proteins by human phages promote tumour progression and
cathepsin B from normal and tumour tissues. metastasis. Nat Rev Cancer 4(1):71–78.
Biochem J 282(Pt 1):273–278 https://doi.org/10.1038/nrc1256
53. Sameni M, Moin K, Sloane BF (2000) Imag- 63. Mantovani A, Ming WJ, Balotta C,
ing proteolysis by living human breast cancer Abdeljalil B, Bottazzi B (1986) Origin and
cells. Neoplasia 2(6):496–504 regulation of tumor-associated macrophages:
the role of tumor-derived chemotactic factor.
54. Sameni M, Dosescu J, Sloane BF (2001) Biochim Biophys Acta 865(1):59–67
Imaging proteolysis by living human glioma
cells. Biol Chem 382(5):785–788. https:// 64. Mantovani A, Sica A (2010) Macrophages,
doi.org/10.1515/BC.2001.094 innate immunity and cancer: balance, toler-
ance, and diversity. Curr Opin Immunol 22
(2):231–237. https://doi.org/10.1016/j. 74. Lin EY, Nguyen AV, Russell RG, Pollard JW
coi.2010.01.009 (2001) Colony-stimulating factor 1 promotes
65. Hanahan D, Coussens LM (2012) Accessories progression of mammary tumors to malig-
to the crime: functions of cells recruited to the nancy. J Exp Med 193(6):727–740
tumor microenvironment. Cancer Cell 21 75. Zhao L, Lim SY, Gordon-Weeks AN, Tapme-
(3):309–322. https://doi.org/10.1016/j. ier TT, Im JH, Cao Y, Beech J, Allen D,
ccr.2012.02.022 Smart S, Muschel RJ (2013) Recruitment of
66. Mantovani A, Sozzani S, Locati M, Allavena P, a myeloid cell subset (CD11b/Gr1 mid) via
Sica A (2002) Macrophage polarization: CCL2/CCR2 promotes the development of
tumor-associated macrophages as a paradigm colorectal cancer liver metastasis. Hepatology
for polarized M2 mononuclear phagocytes. 57(2):829–839. https://doi.org/10.1002/
Trends Immunol 23(11):549–555 hep.26094
67. Sica A, Saccani A, Bottazzi B, Polentarutti N, 76. Lu X, Kang Y (2009) Chemokine (C-C motif)
Vecchi A, van Damme J, Mantovani A (2000) ligand 2 engages CCR2+ stromal cells of
Autocrine production of IL-10 mediates monocytic origin to promote breast cancer
defective IL-12 production and NF-kappa B metastasis to lung and bone. J Biol Chem
activation in tumor-associated macrophages. J 284(42):29087–29096. https://doi.org/10.
Immunol 164(2):762–767 1074/jbc.M109.035899
68. Mantovani A, Savino B, Locati M, 77. Qian BZ, Li J, Zhang H, Kitamura T,
Zammataro L, Allavena P, Bonecchi R Zhang J, Campion LR, Kaiser EA, Snyder
(2010) The chemokine system in cancer biol- LA, Pollard JW (2011) CCL2 recruits inflam-
ogy and therapy. Cytokine Growth Factor Rev matory monocytes to facilitate breast-tumour
21(1):27–39. https://doi.org/10.1016/j. metastasis. Nature 475(7355):222–225.
cytogfr.2009.11.007 https://doi.org/10.1038/nature10138
69. Lin EY, Pollard JW (2007) Tumor-associated 78. Kitamura T, Qian BZ, Soong D, Cassetta L,
macrophages press the angiogenic switch in Noy R, Sugano G, Kato Y, Li J, Pollard JW
breast cancer. Cancer Res 67 (2015) CCL2-induced chemokine cascade
(11):5064–5066. https://doi.org/10.1158/ promotes breast cancer metastasis by enhanc-
0008-5472.CAN-07-0912 ing retention of metastasis-associated macro-
70. Leek RD, Lewis CE, Whitehouse R, phages. J Exp Med 212(7):1043–1059.
Greenall M, Clarke J, Harris AL (1996) Asso- https://doi.org/10.1084/jem.20141836
ciation of macrophage infiltration with angio- 79. Vaage J, Pepin KG (1985) Morphological
genesis and prognosis in invasive breast observations during developing concomitant
carcinoma. Cancer Res 56(20):4625–4629 immunity against a C3H/he mammary
71. Lewis JS, Landers RJ, Underwood JC, Harris tumor. Cancer Res 45(2):659–666
AL, Lewis CE (2000) Expression of vascular 80. Vaage J, Harlos JP (1991) Collagen produc-
endothelial growth factor by macrophages is tion by macrophages in tumour encapsulation
up-regulated in poorly vascularized areas of and dormancy. Br J Cancer 63(5):758–762
breast carcinomas. J Pathol 192(2):150–158. 81. Ansorge HL, Meng X, Zhang G, Veit G,
https://doi.org/10.1002/1096-9896( Sun M, Klement JF, Beason DP, Soslowsky
2000)9999:9999<::AID-PATH687>3.0. LJ, Koch M, Birk DE (2009) Type XIV colla-
CO;2-G gen regulates fibrillogenesis: premature colla-
72. Colegio OR, Chu NQ, Szabo AL, Chu T, gen fibril growth and tissue dysfunction in
Rhebergen AM, Jairam V, Cyrus N, Bro- null mice. J Biol Chem 284(13):8427–8438.
kowski CE, Eisenbarth SC, Phillips GM, https://doi.org/10.1074/jbc.M805582200
Cline GW, Phillips AJ, Medzhitov R (2014) 82. Gerecke DR, Meng X, Liu B, Birk DE (2003)
Functional polarization of tumour-associated Complete primary structure and genomic
macrophages by tumour-derived lactic acid. organization of the mouse Col14a1 gene.
Nature 513(7519):559–563. https://doi. Matrix Biol 22(3):209–216
org/10.1038/nature13490 83. Liguori M, Solinas G, Germano G,
73. Wyckoff J, Wang W, Lin EY, Wang Y, Pixley F, Mantovani A, Allavena P (2011) Tumor-
Stanley ER, Graf T, Pollard JW, Segall J, Con- associated macrophages as incessant builders
deelis J (2004) A paracrine loop between and destroyers of the cancer stroma. Cancer 3
tumor cells and macrophages is required for (4):3740–3761. https://doi.org/10.3390/
tumor cell migration in mammary tumors. cancers3043740
Cancer Res 64(19):7022–7029. https://doi. 84. Zhu Y, Herndon JM, Sojka DK, Kim KW,
org/10.1158/0008-5472.CAN-04-1449 Knolhoff BL, Zuo C, Cullinan DR, Luo J,
218 Chen Varol
Bearden AR, Lavine KJ, Yokoyama WM, 92. Meng XM, Wang S, Huang XR, Yang C,
Hawkins WG, Fields RC, Randolph GJ, Xiao J, Zhang Y, To KF, Nikolic-Paterson
DeNardo DG (2017) Tissue-resident macro- DJ, Lan HY (2016) Inflammatory macro-
phages in pancreatic ductal adenocarcinoma phages can transdifferentiate into myofibro-
originate from embryonic hematopoiesis and blasts during renal fibrosis. Cell Death Dis 7
promote tumor progression. Immunity 47 (12):e2495. https://doi.org/10.1038/
(2):323–338 e326. https://doi.org/10. cddis.2016.402
1016/j.immuni.2017.07.014 93. Yang J, Lin SC, Chen G, He L, Hu Z, Chan L,
85. Puolakkainen PA, Brekken RA, Muneer S, Trial J, Entman ML, Wang Y (2013) Adipo-
Sage EH (2004) Enhanced growth of pancre- nectin promotes monocyte-to-fibroblast tran-
atic tumors in SPARC-null mice is associated sition in renal fibrosis. J Am Soc Nephrol 24
with decreased deposition of extracellular (10):1644–1659. https://doi.org/10.1681/
matrix and reduced tumor cell apoptosis. ASN.2013030217
Mol Cancer Res 2(4):215–224 94. Weitkamp B, Cullen P, Plenz G, Robenek H,
86. Sangaletti S, Stoppacciaro A, Guiducci C, Rauterberg J (1999) Human macrophages
Torrisi MR, Colombo MP (2003) Leukocyte, synthesize type VIII collagen in vitro and in
rather than tumor-produced SPARC, deter- the atherosclerotic plaque. FASEB J 13
mines stroma and collagen type IV deposition (11):1445–1457
in mammary carcinoma. J Exp Med 198 95. Medbury HJ, James V, Ngo J, Hitos K,
(10):1475–1485. https://doi.org/10.1084/ Wang Y, Harris DC, Fletcher JP (2013) Dif-
jem.20030202 fering association of macrophage subsets with
87. Sangaletti S, Di Carlo E, Gariboldi S, atherosclerotic plaque stability. Int Angiol 32
Miotti S, Cappetti B, Parenza M, Rumio C, (1):74–84
Brekken RA, Chiodoni C, Colombo MP 96. Cieslik KA, Taffet GE, Carlson S,
(2008) Macrophage-derived SPARC bridges Hermosillo J, Trial J, Entman ML (2011)
tumor cell-extracellular matrix interactions Immune-inflammatory dysregulation modu-
toward metastasis. Cancer Res 68 lates the incidence of progressive fibrosis and
(21):9050–9059. https://doi.org/10.1158/ diastolic stiffness in the aging heart. J Mol
0008-5472.CAN-08-1327 Cell Cardiol 50(1):248–256. https://doi.
88. Burke RM, Madden KS, Perry SW, Zettel ML, org/10.1016/j.yjmcc.2010.10.019
Brown EB 3rd (2013) Tumor-associated 97. Bertrand S, Godoy M, Semal P, Van Gansen P
macrophages and stromal TNF-alpha regulate (1992) Transdifferentiation of macrophages
collagen structure in a breast tumor model as into fibroblasts as a result of Schistosoma
visualized by second harmonic generation. J mansoni infection. Int J Dev Biol 36
Biomed Opt 18(8):86003. https://doi.org/ (1):179–184
10.1117/1.JBO.18.8.086003 98. Mass E, Jacome-Galarza CE, Blank T,
89. Allaoui R, Bergenfelz C, Mohlin S, Lazarov T, Durham BH, Ozkaya N,
Hagerling C, Salari K, Werb Z, Anderson Pastore A, Schwabenland M, Chung YR,
RL, Ethier SP, Jirstrom K, Pahlman S, Rosenblum MK, Prinz M, Abdel-Wahab O,
Bexell D, Tahin B, Johansson ME, Geissmann F (2017) A somatic mutation in
Larsson C, Leandersson K (2016) Cancer- erythro-myeloid progenitors causes neurode-
associated fibroblast-secreted CXCL16 generative disease. Nature 549
attracts monocytes to promote stroma activa- (7672):389–393. https://doi.org/10.1038/
tion in triple-negative breast cancers. Nat nature23672
Commun 7:13050. https://doi.org/10. 99. Sinha M, Sen CK, Singh K, Das A, Ghatak S,
1038/ncomms13050 Rhea B, Blackstone B, Powell HM, Khanna S,
90. Schnoor M, Cullen P, Lorkowski J, Stolle K, Roy S (2018) Direct conversion of injury-site
Robenek H, Troyer D, Rauterberg J, Lor- myeloid cells to fibroblast-like cells of granu-
kowski S (2008) Production of type VI colla- lation tissue. Nat Commun 9(1):936.
gen by human macrophages: a new dimension https://doi.org/10.1038/s41467-018-
in macrophage functional heterogeneity. J 03208-w
Immunol 180(8):5707–5719 100. Matsubayashi Y, Louani A, Dragu A,
91. Nikolic-Paterson DJ, Wang S, Lan HY (2014) Sanchez-Sanchez BJ, Serna-Morales E,
Macrophages promote renal fibrosis through Yolland L, Gyoergy A, Vizcay G, Fleck RA,
direct and indirect mechanisms. Kidney Int Heddleston JM, Chew TL, Siekhaus DE,
Suppl 4(1):34–38. https://doi.org/10. Stramer BM (2017) A moving source of
1038/kisup.2014.7 matrix components is essential for de novo
basement membrane formation. Curr Biol
27(22):3526–3534 e3524. https://doi.org/ 110. Zeng ZS, Cohen AM, Guillem JG (1999)
10.1016/j.cub.2017.10.001 Loss of basement membrane type IV collagen
101. Kalluri R, Zeisberg M (2006) Fibroblasts in is associated with increased expression of
cancer. Nat Rev Cancer 6(5):392–401. metalloproteinases 2 and 9 (MMP-2 and
https://doi.org/10.1038/nrc1877 MMP-9) during human colorectal tumori-
102. Vignaud JM, Marie B, Klein N, Plenat F, genesis. Carcinogenesis 20(5):749–755
Pech M, Borrelly J, Martinet N, Duprez A, 111. Hotary K, Li XY, Allen E, Stevens SL, Weiss
Martinet Y (1994) The role of platelet- SJ (2006) A cancer cell metalloprotease triad
derived growth factor production by tumor- regulates the basement membrane transmi-
associated macrophages in tumor stroma for- gration program. Genes Dev 20
mation in lung cancer. Cancer Res 54 (19):2673–2686. https://doi.org/10.1101/
(20):5455–5463 gad.1451806
103. Pickup M, Novitskiy S, Moses HL (2013) 112. Jiang D, Lim SY (2016) Influence of immune
The roles of TGFbeta in the tumour micro- myeloid cells on the extracellular matrix dur-
environment. Nat Rev Cancer 13 ing cancer metastasis. Cancer Microenviron 9
(11):788–799. https://doi.org/10.1038/ (1):45–61. https://doi.org/10.1007/
nrc3603 s12307-016-0181-6
104. Karlmark KR, Weiskirchen R, Zimmermann 113. Coussens LM, Tinkle CL, Hanahan D, Werb
HW, Gassler N, Ginhoux F, Weber C, Z (2000) MMP-9 supplied by bone marrow-
Merad M, Luedde T, Trautwein C, Tacke F derived cells contributes to skin carcinogene-
(2009) Hepatic recruitment of the inflamma- sis. Cell 103(3):481–490
tory Gr1+ monocyte subset upon liver injury 114. Huang S, Van Arsdall M, Tedjarati S,
promotes hepatic fibrosis. Hepatology 50 McCarty M, Wu W, Langley R, Fidler IJ
(1):261–274. https://doi.org/10.1002/ (2002) Contributions of stromal
hep.22950 metalloproteinase-9 to angiogenesis and
105. Hesketh M, Sahin KB, West ZE, Murray RZ growth of human ovarian carcinoma in mice.
(2017) Macrophage phenotypes regulate scar J Natl Cancer Inst 94(15):1134–1142
formation and chronic wound healing. Int J 115. Hiratsuka S, Nakamura K, Iwai S,
Mol Sci 18(7):E1545. https://doi.org/10. Murakami M, Itoh T, Kijima H, Shipley JM,
3390/ijms18071545 Senior RM, Shibuya M (2002) MMP9 induc-
106. Knipper JA, Willenborg S, Brinckmann J, tion by vascular endothelial growth factor
Bloch W, Maass T, Wagener R, Krieg T, receptor-1 is involved in lung-specific metas-
Sutherland T, Munitz A, Rothenberg ME, tasis. Cancer Cell 2(4):289–300
Niehoff A, Richardson R, 116. Giraudo E, Inoue M, Hanahan D (2004) An
Hammerschmidt M, Allen JE, Eming SA amino-bisphosphonate targets MMP-9-
(2015) Interleukin-4 receptor alpha signaling expressing macrophages and angiogenesis to
in myeloid cells controls collagen fibril assem- impair cervical carcinogenesis. J Clin Invest
bly in skin repair. Immunity 43(4):803–816. 114(5):623–633. https://doi.org/10.1172/
https://doi.org/10.1016/j.immuni.2015. JCI22087
09.005 117. Du R, Lu KV, Petritsch C, Liu P, Ganss R,
107. Zhou X, Franklin RA, Adler M, Jacox JB, Passegue E, Song H, Vandenberg S, Johnson
Bailis W, Shyer JA, Flavell RA, Mayo A, RS, Werb Z, Bergers G (2008) HIF1alpha
Alon U, Medzhitov R (2018) Circuit design induces the recruitment of bone marrow-
features of a stable two-cell system. Cell 172 derived vascular modulatory cells to regulate
(4):744–757 e717. https://doi.org/10. tumor angiogenesis and invasion. Cancer Cell
1016/j.cell.2018.01.015 13(3):206–220. https://doi.org/10.1016/j.
108. Fields GB (2013) Interstitial collagen catabo- ccr.2008.01.034
lism. J Biol Chem 288(13):8785–8793. 118. Deryugina EI, Zajac E, Juncker-Jensen A,
https://doi.org/10.1074/jbc.R113.451211 Kupriyanova TA, Welter L, Quigley JP
109. Collins HM, Morris TM, Watson SA (2001) (2014) Tissue-infiltrating neutrophils consti-
Spectrum of matrix metalloproteinase expres- tute the major in vivo source of angiogenesis-
sion in primary and metastatic colon cancer: inducing MMP-9 in the tumor microenviron-
relationship to the tissue inhibitors of metal- ment. Neoplasia 16(10):771–788. https://
loproteinases and membrane type-1-matrix doi.org/10.1016/j.neo.2014.08.013
metalloproteinase. Br J Cancer 84 119. Madsen DH, Jurgensen HJ, Siersbaek MS,
(12):1664–1670. https://doi.org/10.1054/ Kuczek DE, Grey Cloud L, Liu S,
bjoc.2001.1831 Behrendt N, Grontved L, Weigert R, Bugge
TH (2017) Tumor-associated macrophages
220 Chen Varol
derived from circulating inflammatory mono- 125. Postlethwaite AE, Kang AH (1976)
cytes degrade collagen through cellular Collagen-and collagen peptide-induced che-
uptake. Cell Rep 21(13):3662–3671. motaxis of human blood monocytes. J Exp
https://doi.org/10.1016/j.celrep.2017.12. Med 143(6):1299–1307
011 126. Wiesner C, Le-Cabec V, El Azzouzi K,
120. Gocheva V, Wang HW, Gadea BB, Shree T, Maridonneau-Parini I, Linder S (2014)
Hunter KE, Garfall AL, Berman T, Joyce JA Podosomes in space: macrophage migration
(2010) IL-4 induces cathepsin protease activ- and matrix degradation in 2D and 3D set-
ity in tumor-associated macrophages to pro- tings. Cell Adhes Migr 8(3):179–191
mote cancer growth and invasion. Genes Dev 127. Van Goethem E, Poincloux R, Gauffre F,
24(3):241–255. https://doi.org/10.1101/ Maridonneau-Parini I, Le Cabec V (2010)
gad.1874010 Matrix architecture dictates three-
121. Vasiljeva O, Papazoglou A, Kruger A, dimensional migration modes of human
Brodoefel H, Korovin M, Deussing J, macrophages: differential involvement of pro-
Augustin N, Nielsen BS, Almholt K, teases and podosome-like structures. J Immu-
Bogyo M, Peters C, Reinheckel T (2006) nol 184(2):1049–1061. https://doi.org/10.
Tumor cell-derived and macrophage-derived 4049/jimmunol.0902223
cathepsin B promotes progression and lung 128. Kaplan G (1983) In vitro differentiation of
metastasis of mammary cancer. Cancer Res human monocytes. Monocytes cultured on
66(10):5242–5250. https://doi.org/10. glass are cytotoxic to tumor cells but mono-
1158/0008-5472.CAN-05-4463 cytes cultured on collagen are not. J Exp Med
122. Doyle AD, Yamada KM (2016) Mechanosen- 157(6):2061–2072
sing via cell-matrix adhesions in 3D microen- 129. Wesley RB 2nd, Meng X, Godin D, Galis ZS
vironments. Exp Cell Res 343(1):60–66. (1998) Extracellular matrix modulates mac-
https://doi.org/10.1016/j.yexcr.2015.10. rophage functions characteristic to atheroma:
033 collagen type I enhances acquisition of resi-
123. Jin H, Varner J (2004) Integrins: roles in dent macrophage traits by human peripheral
cancer development and as treatment targets. blood monocytes in vitro. Arterioscler
Br J Cancer 90(3):561–565. https://doi. Thromb Vasc Biol 18(3):432–440
org/10.1038/sj.bjc.6601576 130. Patel NR, Bole M, Chen C, Hardin CC, Kho
124. Franco C, Britto K, Wong E, Hou G, Zhu AT, Mih J, Deng L, Butler J, Tschumperlin D,
SN, Chen M, Cybulsky MI, Bendeck MP Fredberg JJ, Krishnan R, Koziel H (2012)
(2009) Discoidin domain receptor 1 on Cell elasticity determines macrophage func-
bone marrow-derived cells promotes macro- tion. PLoS One 7(9):e41024. https://doi.
phage accumulation during atherogenesis. org/10.1371/journal.pone.0041024
Circ Res 105(11):1141–1148. https://doi.
org/10.1161/CIRCRESAHA.109.207357
Chapter 16
FACS Analysis of Col1α Protein Levels in Primary

Fibroblasts
Noam Cohen and Neta Erez
Abstract
Chronic inflammatory diseases are often associated with organ fibrosis, a progressive condition in which
excessive deposition of extracellular matrix (ECM), mainly composed of collagen I (Col I), is deposited by
activated fibroblasts and severely impairs tissue architecture and function, eventually resulting in organ
failure. Moreover, enhanced collagen deposition by activated fibroblasts and increased stiffness of the
extracellular matrix were demonstrated to be associated with tumor progression and metastasis. In order
to quantitatively analyze fibrotic activation of fibroblasts and collagen deposition, it is essential to assess
collagen content. While various histological methods allow assessment of collagen in tissue sections (e.g.,
Masson trichrome and Sirius red), reliable measurement and quantification of collagen levels in vitro remain
a challenge in the field. In this protocol, we utilize intracellular staining of Col1α and flow cytometry
analysis to analyze collagen content in primary fibroblasts isolated from fresh single cell suspensions of
metastases-bearing lungs.
Key words Cancer-associated fibroblasts, FACS, Collagen, Lung metastases
1 Introduction
Cancer-associated fibroblasts (CAFs) are a vastly heterogeneous

multifunctional population of fibroblastic cells shown to promote
tumor growth by directly stimulating tumor cell proliferation, by
enhancing angiogenesis, and by modifying the extracellular matrix
(ECM), to support tumor cell invasion [1, 2]. CAFs are also key
mediators of tumor-promoting inflammation by secreting cyto-
kines and chemokines that recruit and modulate the function of
immune cells in the tumor microenvironment [3–8].
In addition, CAFs foster tumor progression by modifying
ECM architecture through enhanced deposition of collagen and
by mediating increased cross-linking of collagen fibers, thus stiffen-
ing the stroma, which was found to correlate with tumor progres-
sion [9–11]. In some cancer types, in particular of the breast and
pancreas, activation of fibroblasts leads to substantial deposition of
221
222 Noam Cohen and Neta Erez
fibrotic ECM. Such aberrant fibrotic responses, termed desmopla-

sia, correlate with poor prognosis [11].
Moreover, several recent studies demonstrated in lung and liver
metastasis that formation of hospitable pre-metastatic niche is asso-
ciated with enhanced collagen expression and collagen cross-
linking, leading to increased tissue stiffness that supports metastatic
growth [12–14]. Thus, studying the molecular pathways in fibro-
blasts that are associated with enhanced collagen deposition is
crucial to our understanding of tumor-promoting ECM modifica-
tions. Such changes can be quantitatively assessed in vivo by histo-
logical analysis of collagen in tissues [15, 16]. However, in order to
specifically analyze factors that instigate pro-fibrotic changes in
fibroblasts, establishment of in vitro assays is required. Neverthe-
less, quantitative analysis of collagen in cultured fibroblasts has
been challenging, since analysis of modifications in gene expression
are not always indicative of protein levels. Analyses at the protein
level (e.g., Western blotting) are also challenging and often do not
provide robust and consistent results, since collagen is a very big
and complex protein, containing several subunits and multiple
posttranslational modifications.
Flow cytometry is a well-established method for quantitative
analysis of cell surface markers, by labeling with specific
fluorophore-conjugated antibodies. In recent years, the applica-
tions of FACS-related methods expanded significantly, and they
can now be utilized for analyzing multiple characteristic of single
cells, including protein levels and phosphorylation levels of intra-
cellular proteins [17]. Analyzing such intracellular characteristics
requires additional steps of cell fixation and permeabilization.
In this protocol, we use FACS analysis to quantify collagen-1α
levels in primary lung fibroblasts, isolated from fresh lungs of mice
and cultured in vitro for several days. This method can be used to
analyze the effect of various treatments/triggers on modifications
of collagen levels in fibroblasts. In combination with other metrics
for pro-fibrogenic activity (expression of pro-fibrotic genes, colla-
gen contraction activity, etc.), this method is useful to study ques-
tions regarding fibrosis and fibroblast activation.
2 Materials
2.1 Lung Dissection 1. Surgery tools (washed with detergent and then dipped in 70%
ethanol).
2. Styrofoam surface to rest the mice on during dissection.
3. Pins or needles to hold mice during dissection.
4. Plastic jar with lid for digestion.
5. Magnetic stir bar.
Analysis of Col1α Levels in Primary Fibroblasts 223
6. Incubator at 37 C.
7. Multipoint magnetic stirrer.
8. PBS.
9. 10 cm plate.
2.2 Single Cell 1. Centrifuge (fit for spinning of 50 ml conical tubes).

Suspension 2. Dissociation solution (made freshly before use): 0.02 g dispase
II, 0.02 g collagenase IV, 0.01 g deoxyribonuclease, 20 ml
serum-free medium (SFM).
3. 10% FCS DMEM.
4. 70 μm cell strainer.
5. Red blood cell lysis buffer: 4 g NH4Cl, 0.5 g KHCO3,
500 ml DDW.
2.3 Cell Culture 1. Collagen mixture: 0.05 mg rat tail collagen I in 1 ml 0.02 N
acetic acid. Adjust volumes as required in specific steps (see
Note 1).
2. 10% FCS DMEM.
3. DMEM.
4. Six-well plate.
5. 10 cm plate.
6. Trypsin-EDTA.
7. PBS.
2.4 Antibody 1. PBS.

Staining 2. Trypsin-EDTA.
3. FACS buffer: PBS, 1% FCS, 2 mM EDTA.
4. Staining buffer: PBS, 1% FCS, 0.09% Na+-azide. Adjust buffer
pH to 7.4–7.6.
5. Anti-mouse Col1α-FITC.
6. Cytofix/Cytoperm Plus kit.
7. FACS tubes with filter.
3 Methods
The protocol described is suitable for six mice, but can be easily
adjusted for smaller or lager mouse numbers.
3.1 Lung Dissection Since isolated fibroblasts are cultured, it is best to perform the
dissection in a laminar flow hood, under sterile conditions.
1. Euthanize mice using carbon dioxide (CO2) inhalation.
Fig. 1 Lung dissection, sequential steps from left to right: pin mouse to diaper coated Styrofoam surface,
dissect out skin, cut through the ribs with scissors and use forceps to disconnect the lungs
Fig. 2 Whole lung in PBS post-dissection
2. On a Styrofoam surface, place the mouse on its back, and pin

firmly the upper limbs using 25-gauge needles (Fig. 1).
3. Spray the mouse generously with 70% ethanol.
4. Use scissors to cut the skin. For sterility purposes, it is recom-
mended to use a different set of scissors for skin and for inner
cavities. Use scissors to cut through the ribs to expose the
thoracic cavity.
5. Remove the lungs gently and transfer into 10 cm dish contain-
ing PBS (Fig. 2).
3.2 Single Cell 1. Place the lungs in digestion jar.

Suspension 2. Mince thoroughly with curved scissors.
3. Add 20 ml dissociation solution (see Note 2).

4. Place on multipoint magnetic stirrer in 37 C incubator, and
stir for 30 min.
5. In the meantime, prepare 6 ml of collagen mixture and coat
six-well plate (1 ml per well). Incubate the plate at 37 C
incubator for 30 min (see Note 3).
*This stage can also be done in advance. Collagen-coated plates
can be kept in 4 C with PBS for 3 days.
6. Take out the digestion jar, and stop the reaction by adding
20 ml of cold 10% FCS DMEM.
7. Strain the tissue/media mixture through a 70 μm cell strainer
placed on top of a 50 ml conical tube.
8. Centrifuge the conical tube for 5 min at 300 g at 4 C.
9. Decant the supernatant using a glass pipette attached to a
vacuum, without disturbing the cell pellet.
10. Resuspend cell pellet in 10 ml RBC lysis buffer and incubate for
5 min at RT.
11. Neutralize RBC lysis buffer by adding 10 ml of 10%
FCS DMEM.
12. Centrifuge the conical tube with cells for 5 min at 300 g at
4 C.
13. Decant the supernatant and resuspend in 12 ml 10% FCS
DMEM (2 ml per mouse).
14. Wash the six-well plate with PBS twice, and seed the cells in a
six-well plate (one well per mouse).
15. Incubate the cells in 5% CO2, 37 C incubator.
3.3 Cell Culture 1. On day 2, wash cells with PBS and change to fresh media to
dispose of any cell debris.
2. On day 3 (or when cells are ~80% confluent, usually takes
1–2 days after isolation), coat six 10 cm plates with collagen
mixture (7 ml per plate), and incubate at 37 C incubator for
30 min (see Note 4) (Fig. 3).
3. Detach cells with trypsin-EDTA and split into 10 cm plates
(one well into one plate).
4. On day 4, wash cells with PBS and change to fresh media.
5. When cells are 80–90% confluent (usually takes 1–2 days),
detach cells, count them, and seed 300,000 cells per well in a
six-well plate for desired experiments (see Note 5).
6. On the next day, wash cell twice with PBS, and incubate with
SFM for 6 h (see Note 6).
7. Wash out the SFM and incubate cells with your treatment
media for 24 h.
Fig. 3 Bright field image of cultured lung fibroblasts at ~90% confluency. Magnification 100
3.4 Antibody 1. Following 24 h of incubation with treatment media, wash cells

Staining twice with PBS.
2. Add to each well 500 μl trypsin-EDTA to detach cells.
3. Neutralize trypsin by adding 1 ml 10% FCS DMEM.
4. Place cells into Eppendorf tubes (one tube per well) and cen-
trifuge for 5 min at 300 g.
5. Resuspend cell pellet in 1 ml PBS and centrifuge for 5 min at
300 g.
6. Decant supernatant and resuspend cells in 100 μl of fixation/
permeabilization buffer.
7. Incubate cells on ice for 20 min.
8. Wash cells with 250 μl 1 BD wash buffer and spin for 5 min at
300 g at 4 C. Repeat twice.
9. Resuspend cell pellet in each tube in 110 μl 1 BD wash buffer.
10. Remove 10 μl from each tube to a new Eppendorf tube, to be
used as unstained control. This control will be composed of a
mixture of 10 μl from each sample (total volume:
6 10 ¼ 60 μl).
11. Bring volume in unstained tube to 200 μl with FACS buffer.
12. Add to all test tubes 2 μl of anti-mouse Col1α-FITC antibody.
13. Incubate cells on ice for 30 min in the dark.
14. Wash cells with 250 μl 1 BD wash buffer, and spin for 5 min
at 300 g at 4 C. Repeat twice.
15. Decant supernatant and resuspend cells in 200 μl FACS buffer.
16. Filter the cells in all tubes into matching FACS tubes with filter
top to avoid cell clumps.
17. Keep samples on ice while analyzing.
Fig. 4 FACS analysis of Col1α in NLF analyzed by Kaluza® 1.5 Flow Analysis Software (Beckman Coulter, Inc.)
3.5 FACS Analysis 1. Run control unstained tube to determine gates.

2. Acquire test tubes using the same parameters.
3. Analyze results with a suitable flow cytometry analysis software
such as Kaluza (Fig. 4).
4 Notes
1. The concentration of rat tail collagen I (BD) varies between

batches. Adjust the required volume to the specific concentra-
tion you have in order to get 0.05 mg rat tail collagen I in 1 ml
of 0.02 N acetic acid.
2. This volume of dissociation solution will efficiently dissociate
up to ten lungs.
3. Fibroblast are activated when they are plated directly on the
stiff surface of a plastic culture dish; therefore, for optimal
culturing of primary fibroblasts, a preliminary step of collagen
coating is required, as detailed above.
4. The times in this section may vary between different experi-
ments. Cells should reach 80–90% confluence before
splitting them.
5. Pause point: At this point cells can be frozen in freezing media

(90% FCS, 10% DMSO) for future analysis.
6. Short starvation is used in order to maintain cells in basal level
before incubation with specific treatments.
References
1. Kalluri R, Zeisberg M (2006) Fibroblasts in 9. Goetz JG, Minguet S, Navarro-lérida I et al

cancer. Nat Rev Cancer 6:392–401. https:// (2011) Biomechanical remodeling of the
doi.org/10.1038/nrc1877 microenvironment by stromal caveolin-1 favors
2. Östman A, Augsten M (2009) Cancer- tumor invasion and metastasis. Cell
associated fibroblasts and tumor growth - 146:148–163. https://doi.org/10.1016/j.
bystanders turning into key players. Curr cell.2011.05.040.Biomechanical
Opin Genet Dev 19:67–73. https://doi.org/ 10. Levental KR, Yu H, Kass L et al (2009) Matrix
10.1016/j.gde.2009.01.003 crosslinking forces tumor progression by
3. Erez N, Glanz S, Raz Y et al (2013) Cancer- enhancing integrin signaling. Cell
associated fibroblasts express pro-inflammatory 139:891–906. https://doi.org/10.1016/j.
factors in human breast and ovarian tumors. cell.2009.10.027.Matrix
Biochem Biophys Res Commun 11. Erler JT, Weaver VM (2009) Three-
437:397–402. https://doi.org/10.1016/j. dimensional context regulation of metastasis.
bbrc.2013.06.089 Clin Exp Metastasis 26:35–49. https://doi.
4. Erez N, Truitt M, Olson P, Hanahan D (2010) org/10.1007/s10585-008-9209-8
Cancer-associated fibroblasts are activated in 12. Cox TR, Bird D, Baker A et al (2013)
incipient Neoplasia to orchestrate tumor- LOX-mediated collagen crosslinking is respon-
promoting inflammation in an NF-kappaB- sible for fibrosis-enhanced metastasis. Cancer
dependent manner. Cancer Cell 17:135–147. Res 73:1721–1732. https://doi.org/10.
https://doi.org/10.1016/j.ccr.2009.12.041 1158/0008-5472.CAN-12-2233.LOX-
5. Feig C, Jones JO, Kraman M et al (2013) mediated
Targeting CXCL12 from FAP-expressing car- 13. Erler JT, Bennewith KL, Nicolau M et al
cinoma-associated fibroblasts synergizes with (2006) Lysyl oxidase is essential for hypoxia-
anti—PD-L1 immunotherapy in pancreatic induced metastasis. Nature 440:1222–1226.
cancer. Proc Natl Acad Sci U S A https://doi.org/10.1038/nature04695
110:20212–20217. https://doi.org/10. 14. Nielsen SR, Quaranta V, Linford A et al (2016)
1073/pnas.1320318110 http://www.pnas. Macrophage-secreted granulin supports pan-
org/cgi/doi/10.1073/pnas.1320318110/-/ creatic cancer metastasis by inducing liver
DCSupplemental fibrosis. Nat Cell Biol 18:549–560. https://
6. Martey CA, Pollock SJ, Turner CK et al (2004) doi.org/10.1038/ncb3340
Cigarette smoke induces cyclooxygenase-2 and 15. Bauman TM, Nicholson TM, Abler LL et al
microsomal prostaglandin E 2 synthase in (2014) Characterization of fibrillar collagens
human lung fibroblasts: implications for lung and extracellular matrix of glandular benign
inflammation and cancer. Am J Physiol Lung prostatic hyperplasia nodules. PLoS One
Cell Mol Physiol 14642:981–991. https://doi. 9:1–9. https://doi.org/10.1371/journal.
org/10.1152/ajplung.00239.2003 pone.0109102
7. Sharon Y, Raz Y, Cohen N et al (2015) Tumor- 16. Street JM, Souza ACP, Alvarez-Prats A et al
derived Osteopontin reprograms normal mam- (2014) Automated quantification of renal
mary fibroblasts to promote inflammation and fibrosis with Sirius red and polarization con-
tumor growth in breast cancer. Cancer Res 75 trast microscopy. Physiol Rep 2:1–9. https://
(6):963–974. https://doi.org/10.1158/ doi.org/10.14814/phy2.12088
0008-5472.CAN-14-1990 17. Adan A, Alizada G, Kiraz Y et al (2017) Flow
8. Lotti F, Jarrar AM, Pai RK et al (2013) Che- cytometry: basic principles and applications.
motherapy activates cancer-associated fibro- Crit Rev Biotechnol 37:163–176. https://doi.
blasts to maintain colorectal cancer-initiating org/10.3109/07388551.2015.1128876
cells by IL-17A. J Exp Med 210:2851–2872.
https://doi.org/10.1084/jem.20131195
Chapter 17
Methods for the Construction of Collagen-Based Triple-

Helical Peptides Designed as Matrix Metalloproteinase
Inhibitors
Gregg B. Fields
Abstract
The triple-helical structure of collagen has been accurately reproduced in numerous chemical and recombi-
nant model systems. Triple-helical peptides have found application for dissecting collagen-stabilizing forces,
isolating receptor and protein binding sites in collagen, evaluating collagen-mediated cell signaling activ-
ities, mechanistic examination of collagenolytic proteases, and developing novel biomaterials and drug
delivery vehicles. Due to their inherent stability to general proteolysis, triple-helical peptides present an
opportunity as in vivo inhibitory agents. The present chapter provides methods for the construction of
collagen-based triple-helical peptides designed as matrix metalloproteinase inhibitors.
Key words Collagen, Matrix metalloproteinases, Triple-helical peptides, Triple-helical inhibitor, Pep-
tide-based drugs, Transition state analog
1 Introduction
1.1 Collagen Collagen is the most abundant protein in animals and is the major
structural protein found in the connective tissues such as basement
membranes, tendons, ligaments, cartilage, bone, and skin. The
defining feature of collagen is the supersecondary structure, com-
posed of three parallel extended left-handed polyproline type II
alpha chains of primarily repeating Gly-Xxx-Yyy triplets. Three
left-handed strands intertwine in a right-handed fashioned around
a common axis to form a triple helix. The packing of the triple-
helical coiled-coil structure favors Gly in every third position. In the
collagen Gly-Xxx-Yyy triplet, the residue in the Xxx position is
often L-proline (Pro), and the residue in the Yyy position is often
4(R)-hydroxy-L-proline (Hyp), accounting for 20% of the total
amino acid composition in collagen [1].
Depending on the primary structure of the alpha chains [2],
collagens have been classified into two main categories,
229
230 Gregg B. Fields
homotrimeric and heterotrimeric. Homotrimeric collagens have

three identical alpha chain sequences (designated α1). Examples
of such type of collagens are types II and III. Alternatively, hetero-
trimeric collagens have either three alpha chain of different
sequences, designated as α1, α2, and α3 (i.e., type V), or two
alpha chains of identical sequence (α1) and a third alpha chain of
different sequences (α2), such as types I and VI [3]. Furthermore,
based on their quaternary structure, collagens are grouped into
subfamilies such as fibrillar, collagen associated with banded fibrils
[i.e., fibril-associated collagens with interrupted triple helices
(FACITs)], network-forming (i.e., basement membrane and short
chain), transmembranous, and membrane associated with inter-
rupted triple helices (MACITs) [3]. The collagen triple helix is
essential for the integrity of multiple connective tissues. Most
collagens assist in anchoring cells to the extracellular matrix and
some function in cellular regulation.
1.2 Collagen-Model For several decades triple-helical peptides (THPs) or “mini-col-

Triple-Helical Peptides lagens” consisting of collagen-model sequences and their three-
dimensional folds have been constructed and studied to fully inves-
tigate the structural and biological roles of collagenous proteins
[4–11]. In order to properly study the biological roles of collagen,
THPs need to form triple-helical structures that are thermally stable
to in vitro and in vivo conditions. Several factors influence the
stability of THPs. The chains are held together by hydrogen
bonds that form between the peptide amine of Gly residues and
peptide carbonyl groups in an adjacent polypeptide chain
[7, 12]. These interstrand hydrogen bonds formed between the
GlyN–H and XxxC¼O are critical for triple-helix stability, as
replacement of a central amide bond of (Pro-Pro-Gly)10 with either
an ester or (E)-alkene, or replacement of the central Pro-Pro with a
Pro-trans-Pro isostere, substantially destabilizes the THP
[13–15]. It has long been noted that the thermal stability of the
collagen triple helix is enhanced by Hyp residues. The stability
arising from Hyp is based on stereoelectronic effects preorganizing
the main chain in the proper conformation for triple-helix forma-
tion [16]. Collagen-model sequences have often been “sand-
wiched” between repeats of Gly-Pro-Hyp to obtain THPs of
reasonable stability.
Incorporation of several non-native amino acids in the proper
position within the peptide sequence can further enhance triple-
helix stability. The first Pro in the Pro-Pro-Gly triplet prefers an
endo ring pucker conformation, while the second Pro prefers the
exo ring pucker conformation [17]. This observation suggested
that Pro derivatives preferring the Cγ-endo ring pucker could pre-
organize triple-helix formation when in the Xaa position, whereas
those that prefer the Cγ-exo ring pucker could preorganize triple-
helix formation when in the Yaa position [18]. One such example of
Collagen Peptide Inhibitors 231
a stabilizing non-native amino acid is 4R-fluoroproline (4R-Flp).

Incorporation of 4R-fluoroproline has been shown to induce
hyperstability in the triple helix of (Pro-4RFlp-Gly)10 compared
to (Pro-4RHyp-Gly)10 [19, 20]. The 4R-fluoro group in 4R-
fluoroproline stabilizes the Cγ-exo pucker conformation. Similar
to 4R-Flp, incorporation of 4R-methylproline (4R-mep) in the
Xaa position significantly increases triple-helix stability [18].
1.3 THPs Designed THPs represent small, well-folded peptidic structures. There has
as Inhibitors been a recent surge in the advancement of peptide drugs, as they
tend to have lower toxicities and greater efficacies than existing
drugs [21–24]. The size of THPs makes immunogenicity unlikely;
in fact, with the exclusion of specific glycosylated sequences, colla-
gen peptides are non-immunogenic. The triple-helical nature of the
inhibitor renders it reasonably stable to proteolysis, as observed
in vitro in mouse, rat, and human serum and/or plasma and
in vivo in rats [25–29]. The stability of THPs has allowed for
their administration orally [30].
Application of THPs as biomaterials and drug delivery vehicles
has been described recently [27, 31–33]. The use of THPs as
inhibitors has been explored to a considerably lesser extent. When
identifying integrin binding sites, THPs have been shown to inhibit
integrin binding to native collagens [34–36]. THPs that bind to
leukocyte-associated Ig-like receptor-1 (LAIR-1) induce inhibition
of T-cell activation [37], while THPs that bind to human osteoclast-
associated receptor (OSCAR) inhibit osteoclastogenesis [38].
1.4 Triple-Helical A well-established strategy to develop protease inhibitors is modifi-

Matrix cation of a bioactive peptide such that the hydrolytically susceptible
Metalloproteinase peptide bond is replaced. Metallo(zinc)-proteases use the nucleo-
Inhibitors philic attack of a water molecule as one of the steps of amide bond
hydrolysis [39]. The tetrahedral intermediate that results from
water addition to the amide carbonyl has given rise to
phosphorus-based amide bond replacements (Fig. 1, lower right).
The use of phosphinic pseudo-dipeptides can be a very effective
approach to develop highly selective and potent inhibitors of a
variety of Zn2+ metalloproteases [40, 41]. Phosphinic peptides
[Ψ{PO2H–CH2}; Y ¼ CH2 in Fig. 1] have been shown to behave
as transition state analog inhibitors of matrix metalloproteinases
(MMPs) [42]. Phosphinate triple-helical MMP inhibitors allow
for interaction with both the enzyme active site and secondary
binding sites (exosites) (Fig. 2) [43, 44].
We have developed selective, high-affinity inhibitors for MMPs
based on triple-helical structure and phosphinate analogs of
Gly-Leu, Gly-Val, and Gly-Ile (Tables 1 and 2) [40, 43,
45–49]. Two homotrimeric triple-helical peptide inhibitors
(THPIs), C6-(Gly-Pro-Hyp)4-Gly-Pro-Pro-GlyΨ{PO2H–CH2}
Val-Val-Gly-Glu-Gln-Gly-Glu-Gln-Gly-Pro-Pro-(Gly-Pro-Hyp)4-
232 Gregg B. Fields
Fig. 1 Tetrahedral intermediate (boxed) and statin and phosphorus-based transition state analog inhibitors.
Used with permission from [40]
Fig. 2 Model structure of a triple-helical peptide inhibitor (THPI) of MMPs. The non-hydrolyzable PO2H–CH2
bond, which functions as a transition state analog, is indicated in purple. Used with permission from [40]
Table 1
Inhibition of MMP-2 and MMP-9 by THPIs
Enzyme Inhibitor Temperaturea ( C) Ki(app) (nM)

MMP-2 C6-α1(V)GlyΨ{PO2H–CH2}Val THPI 10 4.14 0.47
37 19.23 0.64
α1(V)GlyΨ{PO2H–CH2}Val [mep14,32,Flp15,33] THPI 25 189.1 26.54
37 2.24 0.11
MMP-9 C6-α1(V)GlyΨ{PO2H–CH2}Val THPI 10 1.76 0.05
37 1.29 0.00
α1(V)GlyΨ{PO2H–CH2}Val [mep14,32,Flp15,33] THPI 25 90.6 6.67
37 0.98 0.092
a
Assay temperature
NH2 [designated C6-α1(V)GlyΨ{PO2H–CH2}Val THPI] and

(Gly-Pro-Hyp)4-Gly-mep-Flp-Gly-Pro-Gln-GlyΨ{PO2H–CH2}
Leu-Ala-Gly-Gln-Arg-Gly-Ile-Arg-Gly-mep-Flp-(Gly-Pro-Hyp)4-
Tyr-NH2 [designated α1(I–III)GlyΨ{PO2H–CH2}Leu THPI],
were shown to inhibit MMPs in the low nanomolar range
[40, 45]. The THPI sequences were based on either the α1(V)
436–450 collagen region, which is hydrolyzed at the Gly-Val bond
by MMP-2 and MMP-9, or the interstitial collagen consensus
sequence α1(I–III)769–783, which is hydrolyzed by MMP-1,
MMP-2, MMP-8, MMP-9, MMP-13, membrane-type 1 MMP
(MT1-MMP), and MT2-MMP [50, 51].
Table 2
Inhibition of MMPs by GlyΨ{PO2H–CH2}Xxx THPIs
Enzyme Inhibitor Inhibitor Tm ( C) Ki(app)a (nM)

MMP-1 GlyΨ{PO2H–CH2}Leu 30 7.83 1.03
MT1-MMP GlyΨ{PO2H–CH2}Leu 30 111.68 28.7
MMP-1 GlyΨ{PO2H–CH2}Ile-His-Lys-Gln 26 169.2 28.4
MMP-3 GlyΨ{PO2H–CH2}Ile-His-Lys-Gln 26 No inhibition at 1 μM
MMP-8 GlyΨ{PO2H–CH2}Ile-His-Lys-Gln 26 124.6 6.9
MT1-MMP GlyΨ{PO2H–CH2}Ile-His-Lys-Gln 26 4704 708.4
MMP-1 GlyΨ{PO2H–CH2}Ile-Tyr-Phe-Gln 25 110.59 29.8
MT1-MMP GlyΨ{PO2H–CH2}Ile-Tyr-Phe-Gln 25 46.15 4.7
a
Assay performed at room temperature (25 C)
The first-generation THPIs were limited for potential in vivo

applications, due to (a) thermal instability or (b) lack of selectivity.
The melting temperature (Tm) of C6-α1(V)GlyΨ{PO2H–CH2}Val
THPI was 25 C, below that desired for a stable triple helix under
physiological conditions [40]. The α1(I-III)GlyΨ{PO2H–CH2}
Leu THPI was selective toward collagenolytic MMPs, but that
group included seven enzymes. The non-native amino acids
(2S,4R)-4-fluoroproline (Flp) and (2S,4R)-4-methylproline (mep)
were utilized to create a more thermally stable α1(V)GlyΨ
{PO2H–CH2}Val THPI (Table 1). The resulting THPI, α1(V)
GlyΨ{PO2H–CH2}Val [mep14,32,Flp15,33], exhibited an increase
of 18 C in thermal stability over the initial THPI [49]. This
THPI was effective in vivo in a mouse model of multiple sclerosis,
reducing clinical severity and weight loss [49].
Two different approaches were applied to obtain selective
THPIs for MT1-MMP, utilizing data obtained from phage display
generation of MT1-MMP selective substrates and combinatorial
chemistry generation of MT1-MMP selective inhibitors
[49]. Two THPIs were developed to be more selective within the
collagenolytic members of the MMP family (Table 2). One of these
THPIs, GlyΨ{PO2H–CH2}Ile-His-Lys-Gln THPI, was more
effective against MMP-8 than MT1-MMP and was utilized
in vivo in a mouse model of sepsis [49]. The THPI targeting
234 Gregg B. Fields
Table 3
Inhibition of MMPs by heterotrimeric THPIs
Enzyme Inhibitor Ki(app)a,b (nM)

MMP-1 THCP8 NI
MMP-3 18,400 11,600
MMP-13 136 8.7
MT1-MMP 382 111.1
MMP-1 THCP9 NI
MMP-3 12,700 4700
MMP-13 106 71.3
MT1-MMP 210 42.4
MMP-1 THCP12 NI
MMP-3 NI
MMP-13 180 62.4
MT1-MMP 320 56.0
Assay performed at room temperature (25 C)
a
NI ¼ No inhibition
b
MMP-8 minimized lung damage, increased production of the anti-

inflammatory cytokine IL-10, and vastly improved mouse
survival [49].
Utilizing the GlyΨ{PO2H–CH2}Ile-Tyr-Phe-Gln THPI
(Table 2), the intracellular role of MMP-9 on citrate synthase
activity post-MI was evaluated in vivo [52]. Citrate synthase activity
decreased in the infarcted, left ventricle tissue of WT mice at day
1 post-MI, but this reduction was not observed in MMP-9 null
mice or THPI-treated mice, suggesting that MMP-9 deletion or
inhibition helps to maintain mitochondrial activity post-MI.
MMP-9 cleaved citrate synthase in vitro, generating a 20 kDa
fragment [52].
In a further attempt to obtain an MT1-MMP selective THPI,
synthesis of heterotrimeric THPIs has been pursued [48]. Hetero-
trimeric THPIs effectively inhibited MT1-MMP and MMP-13
(Ki ¼ 100–300 nM) but had little or no activity against MMP-1
and MMP-3 (Table 3). Since the homotrimeric THPI analog of the
heterotrimer THPI, α1(I–III)GlyΨ{PO2H–CH2}Leu THPI, was
an effective MMP-1 inhibitor (Table 2), the presence of the phos-
phinate in only one strand was responsible for the selectivity for
MT1-MMP versus MMP-1 [48].
2 Materials
All commercially available materials and reagents are used as

received.
Fig. 3 Structures of Fmoc-phosphinate dipeptide building blocks (3), Fmoc-Flp (4), and Fmoc-mep (5).
The phosphinate dipeptides are analogs of Gly-Val (3a), Gly-Leu (3b), and Gly-Ile (3c). Used with permission
from [49]
2.1 Peptide 1. Knight single-stranded peptide (SSP) [Mca-Lys-Pro-Leu-Gly-

Synthesis Leu-Lys(Dnp)-Ala-Arg-NH2] [53–56].
2. Synthesis of Fmoc-Gly-Val phosphinate dipeptide [(R,S)-2-
isopropyl-3-((1-(N-(Fmoc)amino)-methyl)-adamantyloxy-
phosphinyl) propanoic acid] (Fig. 3a) [46].
3. Synthesis of Fmoc-Gly-Leu phosphinate dipeptide [(R,S)-2-
isobutyl-3-((1-(N-(Fmoc)amino)-methyl)-adamantyloxypho-
sphinyl) propanoic acid] (Fig. 3b) [46].
4. Synthesis of Fmoc-Gly-Ile phosphinic dipeptide [(R,S)-2-
(2-butyl)-3-((1-(N-(Fmoc)amino)-methyl)-adamantyloxy-
phosphinyl) propanoic acid] (Fig. 3c) [47].
5. Fmoc deprotection solution. This solution is prepared by mix-
ing 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), piperidine,
and N-methylpyrrolidone (NMP) in a ratio of 1:5:44 (v/v).
1 L of such solution is composed of 20 mL DBU, 100 mL of
piperidine, and 880 mL of NMP.
6. Dde deprotection solution. The 1-(4,4-dimethyl-2,6-dioxocy-
clohexylidene)ethyl (Dde) protecting groups are removed
using 2% hydrazine in dimethylformamide (DMF). This solu-
tion is prepared by mixing 2 mL hydrazine hydrate and
98 mL DMF.
7. Peptide cleavage cocktail. Under the fume hood, prepare the
peptide cleavage solution containing 5% H2O in trifluoroacetic
acid (TFA, peptide synthesis grade). To prepare 5 mL of such a
solution, mix 250 μL of H2O and 4.75 mL of TFA. For every
100 mg of resin, use 1.5 mL cleavage cocktail (see Note 1).
8. Peptide lyophilization. Synthetic peptides and conjugates are
lyophilized.
2.2 HPLC 1. All analytical runs are performed using a Vydac C18 column
Characterization (5 μm, 300 Å, 150 mm 4.6 mm) or similar.
and Purification 2. Preparative runs are carried out using a Vydac C18 column
(15–20 μm, 300 Å, 250 mm 22 mm) or similar.
236 Gregg B. Fields
3. Water is of HPLC or Milli-Q quality.

4. The following solvent system is used: solvent A contains 0.1%
TFA in H2O, while solvent B contains 0.1% TFA in acetoni-
trile. The analytical gradient ranging from 2 to 100% B over
20 min using 1 mL/min flow rate is used. The preparative
HPLC elution gradients are 7–30% of B in 40 min [for GlyΨ
{PO2H–CH2}Ile-His-Lys-Gln THPI], 10–35% of B in 50 min
[for GlyΨ{PO2H–CH2}Ile-Tyr-Phe-Gln THPI], and 20–50%
of B in 30 min [for α1(V)GlyΨ{PO2H–CH2}Val [mep14,32,
Flp15,33] THPI] with a flow rate of 10 mL/min. In both
cases peptide detection is carried out at λ ¼ 220 and 280 nm.
Homogenous fractions are combined and lyophilized.
2.3 Mass The molecular weight of peptides and peptide conjugates are deter-
Spectrometry mined using matrix-assisted laser desorption/ionization time-of-
flight mass spectrometry (MALDI-TOF MS). MALDI-TOF MS
analysis is performed using an Applied Biosystems Voyager
DE-PRO Biospectrometry workstation or Bruker Microflex LF
mass spectrometer using sinapinic acid (SA), α-cyano-4-hydroxy-
cinnamic acid (HCCA), or 2,5-dihydroxybenzoic acid (DHB) as
matrix.
2.4 Circular Triple-helical peptide structure is evaluated by near-UV CD spec-

Dichroism troscopy. Typically, CD spectra of collagen-like peptides are
(CD) Spectroscopy recorded over the range of λ ¼ 190–250 nm and with a sample
concentration of 0.1 mg/mL in 0.5% acetic acid solution (v/v).
Thermal transition curves are obtained by recording the molar
ellipticity ([Θ]) at λ ¼ 225 nm with an increase in temperature of
20 C/h over the range of 5–80 C. The temperature is controlled
by a JASCO PTC-348WI temperature control unit.
2.5 Absorption Peptide concentration is measured by UV-Vis absorption spectro-

Spectrophotometry photometry. Triple-helical peptide concentration is calculated
according to the Beer-Lambert law, using the following formula:
C ¼ A=ðl 3 εÞ
where C ¼ peptide concentration (M), A ¼ measured absorbance,
and l ¼ cell path length (cm). Please note that in case of triple-
helical peptides, the molar absorption coefficient ε has to be multi-
plied by a number of chromophoric groups present in a molecule.
2.6 Enzymatic Assay 1. 0.1 M ZnCl2.

Buffers and Stock 2. Knight SSP can be prepared as a 20 mM stock solution in water
Solutions and stored at 20 C [55].
3. Enzyme assay buffer (EAB): 50 mM Tris–HCl, pH 7.5,

100 mM NaCl, 10 mM CaCl2, 0.05% Brij 35.
4. TS buffer (TSB): 50 mM Tris–HCl, pH 7.5, 150 mM NaCl,
10 mM CaCl2, 0.01% Brij-35, 1 μM ZnCl2.
5. TSB*Zn: 50 mM Tris–HCl, pH 7.5, 100 mM NaCl, 10 mM
CaCl2, 0.05% Brij-35, 0.02% NaN3, 1 μM ZnCl2.
6. 4-Aminophenylmercuric acetate (APMA) solution:
(a) Prepare APMA stock of 100 mM in dimethyl sulfoxide
(DMSO).
(b) Store at 4 C (this stock is stable for several months).
3 Methods
3.1 Peptide For the incorporation of individual amino acids by Fmoc-solid-

Synthesis Protocols phase methodology, O-(6-chlorobenzotriazol-1-yl)-N,N,N0 ,N-
tetramethyluronium hexafluorophosphate (HCTU)/1-
hydroxybenzotriazole (HOBt) activation is recommended
[57]. Reagent amounts including Fmoc-amino acids (Fmoc-AA),
HCTU, HOBt, and N,N-diisopropylethylamine (DIPEA) are cal-
culated in relation to peptide synthesis scale (typically expressed in
mmol). Microwave-assisted peptide synthesis is achieved on a Lib-
erty Microwave Peptide Synthesizer (CEM Corporation, Mat-
thews, NC).
3.1.1 General Amino Acid For Fmoc-amino acid couplings, the following molar excess is
Coupling Cycles recommended: Fmoc-AA ¼ 5 equiv., HCTU ¼ 4.9 equiv.,
HOBt ¼ 4.9 equiv., and N-methylmorpholine (NMM) or
DIPEA ¼ 10 equiv. Recommended coupling time using conven-
tional automatic peptide synthesizer is 1 h, while microwave-
assisted synthesis proceeds for 5 min at 50 C and 25 W.
3.1.2 General Fmoc The Fmoc-protecting group is removed by gentle treatment of

Deprotection Cycle peptidyl-resin using standard Fmoc deprotection solution. This
step can be performed on an automated peptide synthesizer using
the appropriate deprotection program or manually as follows:
1. Place the resin in a peptide synthesis reaction vessel.
2. Add the appropriate amount of Fmoc deprotection solution
(3–5 times the resin volume).
3. Agitate the resin for 5 min. Drain the solution.
4. Repeat step 2.
5. Agitate the resin for another 15 min. Drain the solution.
6. Wash resin (3 times) using 5 mL of DMF.
238 Gregg B. Fields
3.1.3 General Peptide- 1. Place the peptide-resin in a round bottom flask under inert
Resin Cleavage Protocol atmosphere (Ar) in a fume hood.
2. Mix 50 μL of H2O and 50 μL thioanisole with 900 μL of TFA
to create the cleavage cocktail.
3. Cleave the peptide from the resin using 1.5 mL of cleavage
cocktail per 100 mg of dry resin.
4. Cleavage and side-chain deprotection of the peptide is accom-
plished by treating the resin for 3–4 h (see Note 1).
5. Filter the resin and collect the filtrate.
6. Crude peptides are precipitated from the filtrate with cold
MTBE and centrifuged.
7. Dissolve the pellet in H2O and lyophilize.
3.1.4 Solid-Phase General methods Subheadings 3.1.1 and 3.1.2 are followed, with
Synthesis of GlyΨ the following modifications.
{PO2H–CH2}Ile-His-Lys-Gln
1. Synthesis is performed on a Liberty automated microwave
THPI and GlyΨ {PO2H–CH2}
peptide synthesizer using NovaPEG Rink Amide resin (loading
Ile-Tyr-Phe-Gln THPIs
0.5 mmol/g).
2. The coupling of usual Fmoc-amino acids is achieved with
5 equiv. of each amino acid, 4.9 equiv. of HCTU (0.5 M in
DMF), and 8.0 equiv. of NMM (2 M in DMF) using micro-
wave power of 25 W at 50 C with the reaction time of 300 s
(double coupling for all Arg residues first at 0 W, 25 C for
25 min followed by 25 W, 50 C for 300 s).
3. The removal of Fmoc-protecting group is accomplished using
10% piperazine in DMF (microwave power of 35 W at 75 C
and the reaction time of 30 s, followed by a second treatment
under the same conditions for 180 s).
4. The unusual amino acids Fmoc-4R-fluoro-L-proline (Fmoc-
Flp-OH) and Fmoc-4R-methyl-L-proline (Fmoc-mep-OH)
are coupled using 3 equiv. of amino acid, 2.9 equiv. of
HCTU, and 6 equiv. of NMM with microwave power of
25 W at 50 C for 300 s. The coupling of Fmoc-mep-OH is
repeated.
5. The Fmoc-protected phosphinic dipeptides are attached using
3 equiv. dipeptide, 3 equiv. DIPCDI, 6 equiv. HOBt, and
DIPEA to adjust pH to 8 at 40 C for 40 min and room
temperature overnight.
6. Following dipeptide addition the peptide-resin is capped using
0.5 M Ac2O, 0.125 M DIPEA, and 0.015 M HOBt in DMF at
room temperature for 1 h.
7. The deprotection of Fmoc group, immediately after incorpora-
tion of the phosphinic dipeptide into the sequence, is
performed without microwave power but with increased reac-

tion times of 500 and 800 s.
8. The amino acid immediately after the dipeptide is attached
using the same conditions as for the dipeptide; Pro11 is double
coupled.
9. The final cleavage of peptide from the resin and side-chain
deprotection is carried out using 7 mL of triisopropylsilane-
phenol-H2O-TFA (2:5:5:88) under inert atmosphere
(Ar) for 3 h.
3.1.5 Solid-Phase General methods Subheadings 3.1.1 and 3.1.2 are followed, with
Synthesis of α1(V)GlyΨ the following modifications.
{PO2H–CH2}Val [mep14,32,
1. Synthesis is performed on a CEM liberty blue automated
Flp15,33] THPI
microwave peptide synthesizer using NovaPEG Rink Amide
resin (loading 0.48 mmol/g).
2. The coupling of usual Fmoc-amino acids is achieved with
5 equiv. of each amino acid (0.2 M), 5 equiv. of DIPCDI
(0.5 M in DMF), and 5 equiv. of Oxyma (1 M in DMF).
Fmoc-amino acid coupling cycles are carried out for 15 s at
75 C and 110 s at 90 C.
3. Fmoc deprotection is achieved with 10% piperazine (w/v) in
ethanol:N-methyl-2-pyrrolidone (10:90) for 15 s at 75 C and
50 s at 90 C.
4. Fmoc-Flp-OH is incorporated using 2.5 equiv. of amino acid
for 600 s at 25 C and 300 s at 75 C.
5. Fmoc-mep-OH and Fmoc-GlyΨ{PO2H–CH2}Val are
incorporated using 2.5 equiv. of each compound for 1500 s
at 25 C and 600 s at 75 C.
6. Coupling of Fmoc-Pro-OH following GlyΨ{PO2H–CH2}Val
is performed for 1500 s at 25 C and 600 s at 75 C for using
5 equiv. of amino acid.
7. The final cleavage of peptide from the resin and side-chain
deprotection is carried out using 7 mL of triisopropylsilane-
phenol-H2O-TFA (2:5:5:88) under inert atmosphere
(Ar) for 3 h.
3.1.6 Synthesis General methods Subheadings 3.1.1 and 3.1.2 are followed, with
of Single-Stranded Peptide the following modifications.
1 (SSP1)
1. Synthesis of SSP1 (Fig. 4) is performed in 0.1 mmol scale on
NovaPEG Rink Amide resin (0.48 mmol/g).
2. The coupling of Fmoc-(2S)-propargylglycine (Fmoc-Pra-OH)
is carried out at 45 C for 45 min using 25 W microwave power.
3. The coupling of Pro14 is repeated.
240 Gregg B. Fields
(GPO)nGPQGLAGQRGIR(GPO)5G-HN
(GPO)nGPQGLAGQRGIR(GPO)5G-Lys-Lys-NH2
N3
(GPO)5GPQGLAGQRGIR(GPO)5-Pra-NH2
Cu(I)
(GPO)nGPQGLAGQRGIR(GPO)5G-HN
(GPO)nGPQGLAGQRGIR(GPO)5G-Lys-Lys-NH2
N
N
N
(GPO)5GPQGLAGQRGIR(GPO)5-Pra-NH2
Fig. 4 Schematic of the click-based assembly of THCP6 and THCP7. Assembly

was achieved with DSP3 or DSP4 and SSP1, where n ¼ 5 (DSP3) or 6 (DSP4).
2 (SSP2)
1. Synthesis of SSP2 (Figs. 5 and 6) was performed on TentaGel
resin (0.26 mmol/g).
2. The coupling of Fmoc-Pra is as for SSP1. Residues Arg24 and
Arg27 are double coupled, where the first coupling is for
25 min at 25 C using no power followed by 5 min at 25 C
using 25 W power, and the second coupling is for 5 min at
50 C with microwave power of 25 W.
3. Fmoc removal at Ala21 is achieved by initial treatment for 30 s
and final treatment for 3 min at 75 C with microwave power of
35 W.
4. Coupling of the Fmoc-GlyΨ{P(OAd)(OH)–CH2}Ile-OH
dipeptide (3 equiv.) is performed using N,N0 -diisopropyl-
carbodiimide (DIC)/HOBt/DIPEA (3 equiv./6 equiv./3
equiv.) for 80 min at 40 C with microwave power of 25 W
and then left overnight at ambient temperature.
5. The removal of the Fmoc group, immediately after incorpora-
tion of the dipeptide into the sequence and for the rest of the
synthesis, is performed in two stages: Initial and final
(GPO)6GPQGLAGQRGIR(GPO)5G-HN
(GPO)6GPQGLAGQRGIR(GPO)5G-Lys-Lys-NH2 DSP4
N3
(GPO)5GPQ{GΨ(PO2H-CH2)I}AGQRGIR(GPO)5G-Pra-NH2 SSP2
Cu(I)
(GPO)6GPQGLAGQRGIR(GPO)5G-Lys-Lys-NH2
THCP8
N
N
N
(GPO)5GPQ{GΨ(PO2H-CH2)I}AGQRGIR(GPO)5G-Pra-NH2
Fig. 5 Schematic of the click-based assembly of THCP8. Assembly was achieved with DSP4 and SSP2. Used
with permission from [48]
(GPO)6GPQGLAGQRGIR(GPO)5G-Lys-Lys-NH2 DSP5
N3
(GPO)5GPQ{GΨ(PO2H-CH2)I}AGQRGIR(GPO)5G-Pra-NH2 SSP2
Cu(I)
(GPO)6GPQGLAGQRGIR(GPO)5G-Lys-Lys-NH2
THCP9
N
N
N
(GPO)5GPQ{GΨ(PO2H-CH2)I}AGQRGIR(GPO)5G-Pra-NH2
Fig. 6 Schematic of the click-based assembly of THCP9. Assembly was achieved with DSP5 and SSP2. Used
with permission from [48]
242 Gregg B. Fields
treatments at 50 C without microwave power but with

increased reaction time of 500 s and 800 s, respectively.
6. The coupling of the rest of the amino acids after dipeptide
incorporation is performed with increased reaction time of
20 min, with double coupling of Pro14 and Gln18.
7. Cleavage of the peptide from the resin and side-chain depro-
tection is carried out using 7 mL of triisopropylsilane-phenol-
H2O-TFA (2:5:5:88) under inert atmosphere (Ar) for approx-
imately 4 h.
of Double-Stranded the following modifications.
Peptide 3 (DSP3)
1. The synthesis of DSP3 (Fig. 4) is performed on NovaPEG Rink
Amide resin (0.6 mmol/g).
2. Fmoc-Lys(ivDde)-OH is initially coupled to the resin followed
by Fmoc-Lys(N3)-OH. Both couplings are carried out using
3 equiv. of amino acid, 2.7 equiv. of HCTU, 2.8 equiv. of
HOBt, and 6 equiv. of NMM at 50 C for 20 min with
microwave power of 25 W.
3. The resin-bound dipeptide is removed from the synthesizer,
treated twice with 4 mL of 2% hydrazine hydrate in DMF for
1 h each time to remove the ivDde group, washed with DMF,
and transferred back to the synthesizer reaction vessel.
4. Coupling of the remaining residues is extended to 15 min, and
Pro14 was double coupled.
Peptide 4 (DSP4)
1. Synthesis of DSP4 (Figs. 4 and 5) is carried out by adding one
more Gly-Pro-Hyp repeat on resin-bound DSP3 applying the
standard peptide synthesis protocol.
Peptide 5 (DSP5)
1. Synthesis of DSP5 (Fig. 6) is performed on TentaGel resin
(0.26 mmol/g) using the procedure outlined for DSP3 but
first coupling Fmoc-Lys(N3)-OH followed by Fmoc-Lys
(ivDde)-OH.
10 (SSP10)
1. Synthesis of SSP10 (Fig. 7) is performed on TentaGel S Ram
resin (0.26 mmol/g).
2. Two consecutive Fmoc-Lys(N3) residues are coupled to the
resin as described for DSP3.
(GPO)5GPQGLAGQRGIR(GPO)5G-Pra-NH2 SSP11
N3
(GPO)6GPQ{GΨ(PO2H-CH2)I}AGQRGIR(GPO)5G-Lys-Lys-NH2 SSP10
N3
(GPO)5GPQGLAGQRGIR(GPO)5G-Pra-NH2 SSP11
Cu(I)
(GPO)5GPQGLAGQRGIR(GPO)5G-Pra-NH2
N
N
N
(GPO)6GPQ{GΨ(PO2H-CH2)I}AGQRGIR(GPO)5G-Lys-Lys-NH2 THCP12
N
N
N
(GPO)5GPQGLAGQRGIR(GPO)5G-Pra-NH2
Fig. 7 Schematic of the click-based assembly of THCP12. Assembly was achieved with SSP10 and SSP11.
3. Coupling up to Ala21 is as for SSP2. Arg24 and Arg27 are

double coupled, with the coupling time extended to 10 min.
4. Coupling of the Fmoc-GlyΨ{P(OAd)(OH)–CH2}Ile-OH
dipeptide is performed as described for SSP2.
5. The coupling time of the remainder of the amino acids is
increased to 20 min and Pro14 is double coupled.
6. The removal of Fmoc groups after incorporation of the phos-
phinate moiety is carried out as described for SSP2.
11 (SSP11)
1. The synthesis of SSP11 (Fig. 7) is performed in 0.1 mmol scale
on TentaGel S Ram resin (0.26 mmol/g) by following the
procedure described for SSP1, except an extra Gly residue is
included after the Pra.
244 Gregg B. Fields
3.1.13 Click Chemistry A general click reaction procedure is as follows.

1. A solution of CuSO4·5H2O (5 equiv.) in 5% DMF in H2O
(30 μL/μmol of CuSO4·5H2O) is added to a solution of SSP
(1 equiv.) and DSP (1 equiv.) in H2O (300 μL/μmol of total
peptide, SSP + DSP).
2. The reaction mixture is vortexed and sodium ascorbate
(20 equiv.) is added.
3. The resulting solution is mixed for 1–2 min and kept at room
temperature with occasional mixing.
4. After 12 h, the reaction mixture is diluted with H2O to 2 mL,
filtered through a Titan 2, 30 mm filter green 0.45 μm Nylon
membrane, and purified by RP-HPLC (see below).
5. The pure fractions are combined and lyophilized to afford a
white solid.
3.2 Characterization The peptide molecular mass is confirmed by MALDI-TOF mass

of THPIs spectrometry using HCCA or SA MALDI matrices following tryp-
sin digestion.
3.2.1 Mass Spectrometry
1. 0.5 mg of desired peptide is dissolved in 250 μL of 0.1 M
NaHCO3 solution.
2. 85 μL of immobilized, TPCK-treated trypsin resin solution is
measured into a 2 mL centrifuge tube and washed with 250 μL
of 0.1 M NaHCO3 solution three times.
3. The peptide solution in NaHCO3 is then added to the washed
trypsin and incubated at 37–43 C for 4–12 h.
4. The tube is centrifuged to remove the trypsin.
5. The pH of the solution is adjusted to 4 with 0.1 M HCl.
6. The solution is applied to a pre-moist C18 ZipTip, washed with
0.1% TFA-H2O to remove salt, and finally washed with 0.1%
TFA-acetonitrile to collect peptide fragments.
7. The filtrate is used to record MALDI-TOF mass spectra.
3.2.2 Circular Dichroism Triple helicity is monitored by CD spectroscopy in the far UV

(CD) Spectroscopy wavelengths [53, 58–65]. The general protocol is as follows:
1. Dissolve the peptide in an appropriate buffer to a final concen-
tration of 0.1 mg/mL. Leave the peptide in solution for
24–48 h at 4 C to allow for triple-helix formation.
2. Record a series of spectra (n ¼ 5–10) over the range of
λ ¼ 190–250 nm. A typical spectrum for a triple helix shows a
positive molar ellipticity at λ ¼ 225 nm and a negative molar
ellipticity at λ ¼ 205 nm.
3. Determine the melting temperature by monitoring the change
in molar ellipticity at λ ¼ 225 nm with a constant change in
temperature (20 C/h) from 5 to 80 C. For samples exhibit-

ing sigmoidal melting curves, the inflection point in the transi-
tion region (first derivative) is defined as Tm. The first derivative
from the transition curve can be obtained using JASCO or
GraphPad Prism software (GraphPad, USA). Alternatively, Tm
is evaluated from the midpoint of the transition.
3.3 Enzyme Activation of all MMPs takes place in the TS buffer. See also Note 2.
Activation
3.3.1 Activation 1. Dilute proMMP-1 to 100 μg/mL in TS buffer.

of ProMMP-1 2. Activate proMMP-1 by adding APMA to a final concentration
of 1 mM.
3. Incubate at 37 C for 2–3 h.
4. Immediately after incubation dilute MMP-1 to 20 μg/mL and
store at 37 C until ready to use.
5. Activity is initially measured with the Knight substrate. MMP-1
has a weak activity toward fTHP-15.

of 1 mM.
5. MMP-2 can also be obtained in active form.
3.3.3 Activation 1. Chymotrypsin is prepared as a 1 mg/mL stock in 1 mM HCl.

of ProMMP-3 2. PMSF is prepared as a 0.2 M stock in ethanol.
3. Activate proMMP-3 at 20 μg/mL in EAB containing 5 μg/mL
chymotrypsin.
4. Incubate at 37 C for 30 min.
5. Terminate activation with 2 mM PMSF. Pre-warm the PMSF
to 37 C just prior to adding to the sample.
6. Store at 80 C for future use.
7. Measure activity using the Knight substrate at 10 μM final
substrate concentration.

of 1 mM.
246 Gregg B. Fields


of 1 mM.
5. MMP-9 can also be obtained in active form.

of 1 mM.
4. Immediately after incubation dilute MMP-12 to 20 μg/mL
and store at 80 C until ready to use.

of 1 mM.
4. Immediately after incubation dilute MMP-13 to 20 μg/mL
and store at 80 C until ready to use.
3.3.8 Activation 1. Activation buffer: 50 mM Tris, 1 mM CaCl2, 0.5% (v/v) Brij-

of ProMT1-MMP 35, pH 9.0.
Method 1 2. Activate proMT1-MMP at 40 μg/mL with 0.86 μg/mL
recombinant furin in activation buffer.
3. Incubate at 37 C for 1.5 h.
4. Stop furin activation by adding SSM 3 trifluoroacetate to a final
concentration of 1 μM.
Method 2 1. Prepare 4-(2-aminoethyl-benzenesulfonyl fluoride hydrochlo-

ride) (AEBSF) 100 mM stock in deionized water (see Note 3).
2. Activate proMT1-MMP at 100 μg/mL with 5 μg/mL recom-
binant human active trypsin-3 in TS buffer.
4. Add AEBSF to a final concentration of 1 mM and incubate at
room temperature for 15 min to stop the activation.
5. Store the activated enzyme at 80 C for future use.
3.3.9 Activation 1. Prepare AEBSF 100 mM stock in deionized water.

of ProMT2-MMP (MMP-15) 2. Activate proMT2-MMP at 100 μg/mL with 0.1 μg/mL
trypsin-3 in TS buffer.
4. Add AEBSF to a final concentration of 1 mM and incubate at
room temperature for 15 min to stop the reaction.
5. Store the activated enzyme at 80 C for future use.
3.4 Enzymatic Enzyme kinetics can be determined in a microplate reader such as

Assays Using BioTek Synergy 4 running Gen5 1.11, Gen5 2.00, and Gen5 2.01
Knight SSP software as described previously [45, 66]. Manual assays are per-
formed in black low volume 384 well plates.
1. A 12-point serial dilution of Knight SSP substrate is made
starting at 20 mM.
2. 5 μL of each peptide solution is loaded in 12 different wells in a
384 well plate.
3. The plate is read to obtain initial fluorescence in relative fluo-
rescence units (RFUs).
4. 5 μL of 2 enzyme stock solution is added to each well.
5. The fluorescence is monitored continuously for 15 min to
determine initial reaction rates. The kinetic protocol at 25 C
has 30 s shaking followed by each well read every 20 s.
6. Plates are stored at ambient temperature for 24 h before a final
reading. All measurements are done in triplicate.
7. To determine rates of hydrolysis (Corr Vi; μM/s), fluorescence
(RFU) is plotted over time (s), slopes are divided by fluores-
cence change after 24 h, and then multiplied by the substrate
concentration (μM) {Corr Vi ¼ (slope/change in
RFU) [peptide]}.
8. Kinetic parameters are calculated by Lineweaver-Burk, Hanes-
Woolf, and Eadie-Hofstee linear plots (substrate concentration
versus reaction velocity) as well as nonlinear Michaelis-Menten
regression by GraphPad prism.
9. When kinetics parameters from the above four modes correlate
well, the experiment is considered a success. However, when
linear analysis displays significant exponential character,
enzyme concentration is reduced and the experiment repeated
until linearity is obtained.
3.5 Inhibitor Assays 1. Peptide substrate and inhibitor solutions are prepared using
TSB*Zn and left overnight at 4 C to allow the proper folding
of the peptide triple helix.
248 Gregg B. Fields
2. 1–2.5 nM enzyme is incubated with varying concentration of

inhibitor for 2 h at room temperature. A 2 h incubation is
utilized based on the generally observed behavior of slow on
and off rates for tight-binding inhibitors [67] and studies
demonstrating that high-affinity phosphinate inhibitors of
Zn2+ metalloproteinases are slow binding [68].
3. Residual enzyme activity is monitored by adding Knight sub-
strate solution in TSB*Zn to produce a final concentration of
<0.1KM.
4. Initial velocity rates are determined from the first 10 min of
hydrolysis when product release is linear with time.
5. Fluorescence is measured on a BioTek synergy H1 reader or
H4 hybrid reader using λexcitation ¼ 324 nm and
λemission ¼ 393 nm.
6. Apparent Ki values are calculated from the following formulas
[40]:
( 0:5 ),
2
ðappÞ ðappÞ ðappÞ
vi =vo ¼ Et It Ki þ Et It Ki þ 4Et Ki 2Et
ð1Þ
Ki ðappÞ ¼ Ki ðfAt þ KM g=KM Þ ð2Þ

where It is the total inhibitor concentration, Et is the total
enzyme concentration, At is the total substrate concentration,
vo is the activity in the absence of inhibitor, and KM is the
Michaelis constant.
7. The value of Et/Ki(app) does not exceed 100 so that the inhibi-
tor is distributed in both free and bound forms, and Ki(app) can
be calculated by fitting inhibition data to Eq. 1. Because the
substrate concentration is less than KM/10, Ki(app) values are
insignificantly different from true Ki values.
8. In cases where weak inhibition occurs, Ki(app) values are calcu-
lated using vi ¼ vo/(1 + It/Ki(app)).
4 Notes
1. When crude peptide shows unsatisfactory yield and/or purity, a

different cleavage cocktail could be used. Please refer to [69].
2. ProMMP-1, proMMP-3, proMMP-8, and proMMP-12 can be
activated at 20 ng/μL concentration with a 5 ng chymotryp-
sin/5 ng trypsin mixture for 30 min at 37 C. In our experi-
ence, this mode of activation results in greater activity of these
enzymes compared with the manufacturer’s recommended
activation protocols. The reaction is stopped by addition of
2 mM PMSF. Enzyme activity is tested with 5 μM Knight SSP

in TS buffer.
3. AEBSF solutions in water are slightly acidic. A stock solution is
recommended at a concentration of 20 or 100 mM in water.
Stock solutions should be stored at a pH less than 7. If a final
pH > 7 is required, the pH adjustment should be done shortly
before use. Stock solutions are stable for up to 2 months at
20 C or 1 week at 4 C. AEBSF should be added to an assay
buffer just before use. AEBSF is inactivated by 50% at 37 C
after 5 h.
Acknowledgments
I gratefully acknowledge the National Institutes of Health

(CA098799 and NHLBI Contract 268201000036C-0-0-1), the
Multiple Sclerosis National Research Institute, and the Texas
Higher Education Star Award Program for the support of my
laboratory’s research on THPIs.
References
1. Woodhead-Galloway J (1980) Collagen: the collagen conformation, stability, and self-

anatomy of a protein. Edward Arnold Limited, association. Biopolymers 89:345–353
London, pp 10–19 11. Fields GB (2010) Synthesis and biological
2. Gordon MK, Hahn RA (2010) Collagens. Cell applications of collagen-model triple-helical
Tissue Res 339:247–257 peptides. Org Biomol Chem 8:1237–1258
3. Cole WG (1994) Collagen genes: mutations 12. Bella J, Eaton M, Brodsky B, Berman HM
affecting collagen structure and expression. (1994) Crystal and molecular structure of a
Prog Nucleic Acid Res Mol Biol 47:29–80 collagen-like peptide at 1.9 Å resolution. Sci-
4. Shoulders MD, Raines RT (2009) Collagen ence 266:75–81
structure and stability. Annu Rev Biochem 13. Jenkins CL, Vasbinder MM, Miller SJ, Raines
78:929–958 RT (2005) Peptide bond isosteres: Ester or
5. Brodsky B, Shah NK (1995) The triple-helix (E)-alkene in the backbone of the collagen tri-
motif in proteins. FASEB J 9:1537–1546 ple helix. Org Lett 7:2619–2622
6. Fields GB, Prockop DJ (1996) Perspectives on 14. Dai N, Wang XJ, Etzkorn FA (2008) The effect
the synthesis and application of triple-helical, of a trans-locked Gly-pro alkene isostere on
collagen-model peptides. Biopolymers (Pep- collagen triple helix stability. J Am Chem Soc
tide Sci) 40:345–357 130:5396–5397
7. Jenkins CL, Raines RT (2002) Insights on the 15. Dai N, Etzkorn FA (2009) Cis-trans proline
conformational stability of collagen. Nat Prod isomerization effects on collagen triple-helix
Rep 19:49–59 stability are limited. J Am Chem Soc
8. Koide T (2005) Triple helical collagen-like pep- 131:13728–13732
tides: engineering and applications in matrix 16. Kotch FW, Guzei IA, Raines RT (2008) Stabi-
biology. Connect Tissue Res 46:131–141 lization of the collagen triple helix by
9. Koide T (2007) Designed triple-helical pep- O-methylation of hydroxyproline residues. J
tides as tools for collagen biochemistry and Am Chem Soc 130:2952–2953
matrix engineering. Philos Trans R Soc B 17. DeRider ML, Wilkens SJ, Waddell MJ,
362:1281–1291 Bretscher LE, Weinhold F, Raines RT, Markley
10. Brodsky B, Thiagarajan G, Madhan B, Kar K JL (2002) Collagen stability: insights from
(2008) Triple-helical peptides: an approach to NMR spectroscopic and hybrid density func-
tional computational investigations of the
250 Gregg B. Fields
effect of electronegative substituents on prolyl homotrimers, heterotrimers and higher order

ring conformations. J Am Chem Soc 124 structures. Chem Soc Rev 39:3510–3527
(11):2497–2505 32. Chattopadhyay S, Raines RT (2014) Collagen-
18. Shoulders MD, Satyshur KA, Forest KT, based biomaterials for wound healing. Biopo-
Raines RT (2010) Stereoelectronic and steric lymers 101:821–833
effects in side chains preorganize a protein 33. An B, Lin Y-S, Brodsky B (2016) Collagen
main chain. Proc Natl Acad Sci U S A interactions: drug design and delivery. Adv
107:559–564 Drug Deliv Rev 97:69–84
19. Holmgren SK, Taylor KM, Bretscher LE, 34. Miles AJ, APN S, Furcht LT, Fields GB (1994)
Raines RT (1998) Code for collagen’s stability Promotion of cell adhesion by single-stranded
deciphered. Nature 392:666–667 and triple-helical peptide models of basement
20. Holmgren SK, Bretscher LE, Taylor KM, membrane collagen a1(IV)531-543: evidence
Raines RT (1999) A hyperstable collagen for conformationally dependent and conforma-
mimic. Chem Biol 6:63–70 tionally independent type IV collagen cell
21. Dutton G (2007) Development of peptide adhesion sites. J Biol Chem 269:30939–30945
drugs advances briskly. Gen Engin Biotechnol 35. Knight CG, Morton LF, Peachey AR, Tuckwell
News 27(10):18–20 DS, Farndale RW, Barnes MJ (2000) The
22. Thayer AM (2011) Improving peptides. Chem collagen-binding A-domains of integrin a1b1
Eng News May 30:13–20 and a2b1 recognize the same specific amino
23. Albericio F, Kruger HG (2012) Therapeutic acid sequence, GFOGER, in native (triple-
peptides. Future Med Chem 4:1527–1531 helical) collagens. J Biol Chem 275:35–40
24. DePalma A (2015) Peptides: new processes, 36. Kim JK, Xu Y, Xu X, Keene DR,
lower costs. Gen Eng Biotechnol News 35 Gurusiddappa S, Liang X, Wary KK, Hook M
(13):24–26 (2005) A novel binding site in collagen type III
for integrins a1b1 and a2b1. J Biol Chem
25. Fan C-Y, Huang C-C, Chiu W-C, Lai C-C, 280:32512–32520
Liou G-G, Li H-C, Chou M-Y (2008) Produc-
tion of multivalent protein binders using a self- 37. Lebbink RJ, Raynal N, de Ruiter T, Bihan DG,
trimerizing collagen-like peptide scaffold. Farndale RW, Meyaard L (2009) Identification
FASEB J 22:3795–3804 of multiple potent binding sites for human
leukocyte associated Ig-like receptor LAIR on
26. Ndinguri MW, Zheleznyak A, Lauer JL, Ander- collagens II and III. Matrix Biol 28:202–210
son CJ, Fields GB (2012) Application of
collagen-model triple-helical peptide- 38. Haywood J, Qi J, Chen CC, Lu G, Liu Y, Yan J,
amphiphiles for CD44 targeted drug delivery Shi Y, Gao GF (2016) Structural basis of colla-
systems. J Drug Delivery 2012:592602 gen recognition by human osteoclast-
592613 pages associated receptor and design of osteoclasto-
genesis inhibitors. Proc Natl Acad Sci U S A
27. Yasui H, Yamazaki CM, Nose H, Awada C, 113:1038–1043
Takao T, Koide T (2013) Potential of
collagen-like triple helical peptides as drug car- 39. Babine RE, Bender SL (1997) Molecular recog-
riers: their in vivo distribution, metabolism, nition of protein-ligand complexes: applications
and excretion profiles in rodents. Biopolymers to drug design. Chem Rev 97:1359–1472
(Pept Sci) 100:705–713 40. Lauer-Fields JL, Brew K, Whitehead JK, Li S,
28. Yamazaki CM, Nakase I, Endo H, Kishimoto S, Hammer RP, Fields GB (2007) Triple-helical
Mashiyama Y, Masuda R, Futaki S, Koide T transition-state analogs: a new class of selective
(2013) Collagen-like cell-penetrating peptides. matrix metalloproteinase inhibitors. J Am
Angew Chem Int Ed Engl 52:5497–5500 Chem Soc 129:10408–10417
29. Shinde A, Feher KM, Hu C, Slowinska K 41. Georgiadis D, Dive V (2015) Phosphinic pep-
(2015) Peptide internalization enabled by fold- tides as potent inhibitors of zinc-
ing: triple-helical cell-penetrating peptides. J Metalloproteases. Top Curr Chem 360:1–38
Pept Sci 21:77–84 42. Gall AL, Ruff M, Kannan R, Cuniasse P,
30. Koide T, Yamamoto N, Taira KB, Yasui H Yiotakis A, Dive V, Rio MC, Basset P, Moras
(2016) Fecal excretion of orally administered D (2001) Crystal structure of the stromelysin-
collagen-like peptides in rats: contribution of 3 (MMP-11) catalytic domain complexed with
the triple-helical conformation to their stabil- a phosphinic inhibitor mimicking the
ity. Biol Pharm Bull 39:135–137 transition-state. J Mol Biol 307:577–586
31. Fallas JA, O’Leary LE, Hartgerink JD (2010) 43. Lauer-Fields JL, Chalmers MJ, Busby SA,
Synthetic collagen mimics: self-assembly of Minond D, Griffin PR, Fields GB (2009) Iden-
tification of specific Hemopexin-like domain
residues that facilitate matrix metalloproteinase 53. Lauer-Fields JL, Broder T, Sritharan T,
Collagenolytic activity. J Biol Chem Nagase H, Fields GB (2001) Kinetic analysis
284:24017–24024 of matrix metalloproteinase triple-helicase
44. Bertini I, Fragai F, Luchinat C, Melikian M, activity using fluorogenic substrates. Biochem-
Toccafondi M, Lauer JL, Fields GB (2012) istry 40:5795–5803
Structural basis for matrix metalloproteinase 54. Del Valle JR, Goodman M (2003) Asymmetric
1 catalyzed Collagenolysis. J Am Chem Soc hydrogenations for the synthesis of Boc-Pro-
134:2100–2110 tected-4-Alkylprolinols and Prolines. J Org
45. Lauer-Fields JL, Whitehead JK, Li S, Hammer Chem 68:3923–3931
RP, Brew K, Fields GB (2008) Selective modu- 55. Neumann U, Kubota H, Frei K, Ganu V, Lep-
lation of matrix metalloproteinase 9 (MMP-9) pert D (2004) Characterization of Mca-Lys-
functions via exosite inhibition. J Biol Chem pro-Leu-Gly-Leu-Dpa-Ala-Arg-NH2, a fluoro-
283:20087–20095 genic substrate with increased specificity con-
46. Bhowmick M, Sappidi RR, Fields GB, Lepore stants for collagenases and tumor necrosis
SD (2011) Efficient synthesis of Fmoc- factor converting enzyme. Anal Biochem
protected Phosphinic Pseudodipeptides: build- 328:166–173
ing blocks for the synthesis of matrix metallo- 56. Bhowmick M, Fields GB (2013) Stabilization
proteinase inhibitors (MMPIs). Biopolymers of triple-helical peptides for in vivo applica-
(Pept Sci) 96:1–3 tions. Methods Mol Biol 1081:167–194
47. Bhowmick M, Fields GB (2012) Synthesis of 57. Fields GB, Tian Z, Barany G (1993) Principles
Fmoc-Gly-Ile Phosphinic Pseudodipeptide: and practice of solid-phase peptide synthesis.
residue specific conditions for construction of In: Grant GA (ed) synthetic peptides: a User’s
matrix metalloproteinase inhibitor building guide. W.H. Freeman & Co, New York, pp
blocks. Int J Pept Res Ther 18:335–339 77–183
48. Bhowmick M, Stawikowska R, Tokmina- 58. Fields CG, Lovdahl CM, Miles AJ, Matthias-
Roszyk D, Fields GB (2015) Matrix metallo- Hagen VL, Fields GB (1993) Solid-phase syn-
proteinase inhibition by Heterotrimeric triple- thesis and stability of triple-helical peptides
helical peptide transition state analogs. Chem- incorporating native collagen sequences. Bio-
biochem 16:1084–1092 polymers 33:1695–1707
49. Bhowmick M, Tokmina-Roszyk D, Onwuha- 59. Fields CG, Mickelson DJ, Drake SL, McCarthy
Ekpete L, Harmon K, Robichaud T, Fuerst R, JB, Fields GB (1993) Melanoma cell adhesion
Stawikowska R, Steffensen B, Roush WR, and spreading activities of a synthetic
Wong H, Fields GB (2017) Second generation 124-residue triple-helical “mini-collagen”. J
triple-helical peptide transition state analog Biol Chem 268:14153–14160
matrix metalloproteinase inhibitors. J Med 60. Grab B, Miles AJ, Furcht LT, Fields GB (1996)
Chem 60:3814–3827 Promotion of fibroblast adhesion by triple-
50. Lauer-Fields JL, Sritharan T, Stack MS, helical peptide models of type I collagen-derived
Nagase H, Fields GB (2003) Selective hydroly- sequences. J Biol Chem 271:12234–12240
sis of triple-helical substrates by matrix 61. Lauer-Fields JL, Tuzinski KA, Shimokawa K,
metalloproteinase-2 and -9. J Biol Chem Nagase H, Fields GB (2000) Hydrolysis of
278:18140–18145 triple-helical collagen peptide models by matrix
51. Minond D, Lauer-Fields JL, Cudic M, Overall metalloproteinases. J Biol Chem
CM, Pei D, Brew K, Visse R, Nagase H, Fields 275:13282–13290
GB (2006) The roles of substrate thermal sta- 62. Lauer-Fields JL, Nagase H, Fields GB (2000)
bility and P2 and P0 1 subsite identity on matrix Use of Edman degradation sequence analysis
metalloproteinase triple-helical peptidase activ- and matrix-assisted laser desorption/ionization
ity and collagen specificity. J Biol Chem mass spectrometry in designing substrates for
281:38302–38313 matrix metalloproteinases. J Chromatogr A
52. de Castro Brás LE, Cates CA, DeLeon-Pennell 890:117–125
KY, Ma Y, Iyer RP, Halade GV, 63. Malkar NB, Lauer-Fields JL, Borgia JA, Fields
Yabluchanskiy A, Fields GB, Weintraub ST, GB (2002) Modulation of triple-helical stability
Lindsey ML (2014) Citrate synthase is a novel and subsequent melanoma cellular responses by
in vivo matrix Metalloproteinase-9 substrate single-site substitution of fluoroproline deriva-
that regulates mitochondrial function in the tives. Biochemistry 41:6054–6064
post-myocardial infarction left ventricle. Anti- 64. Yu Y-C, Tirrell M, Fields GB (1998) Minimal
oxid Redox Signal 21:1974–1985 lipidation stabilizes protein-like molecular
architecture. J Am Chem Soc 120:9979–9987
252 Gregg B. Fields
65. Yu Y-C, Berndt P, Tirrell M, Fields GB (1996) inhibitors in drug discovery. John Wiley &
Self-assembling amphiphiles for construction Sons, Inc., Hoboken, NJ, pp 178–213
of protein molecular architecture. J Am Chem 68. Yiallouros I, Vassiliou S, Yiotakis A, Zwilling R,
Soc 118:12515–12520 Stöcker W (1998) Phosphinic peptides, the
66. Palmier MO, Van Doren SR (2007) Rapid first potent inhibitors of astacin, behave as
determination of enzyme kinetics from fluores- extremely slow-binding inhibitors. Biochem J
cence: overcoming the inner filter effect. Anal 331:375–379
Biochem 371:43–51 69. Guy CA, Fields GB (1997) Trifluoroacetic acid
67. Copeland RA (2005) Tight binding inhibition. cleavage and Deprotection of resin-bound pep-
In: Copeland RA (ed) Evaluation of enzyme tides following synthesis by Fmoc chemistry.
Methods Enzymol 289:67–83
INDEX
A E
Age-synchronizing .....................174, 176, 179, 182, 183 ELISA ..................................................191, 192, 194–199
Aging ...................................... v, 136, 146, 169–186, 209 Exon skipping.................................................................... 4
Amino propeptide of type III procollagen Extracellular matrix (ECM)....................................... v, 23,
(PIIINP) ............................................................ 190 57, 61, 63, 65, 71, 72, 115–126, 129–132, 135,
157–164, 169–171, 175, 182–184, 186, 189,
B 190, 204–206, 208, 211, 213, 221, 222, 230
Biomarkers................................................. 87–90, 93, 191
F
Biomedical properties ..................................................... 17
Biophysical characterization .................................. 23, 203 Freeze-cracking ............................................172, 176–177
Biopsy ..................................................145, 147, 153, 190
Bone...................................................................39, 40, 69, G
73, 95, 99, 143, 229 γ-glutamyl transferase (GGT)....................................... 190
Bone resorption............................................................. 135 Gibson assembly..................................................... 7, 8, 10
Glutamyl endoproteinase (GluC) .............. 118, 120, 122
C
Glutaraldehyde (GTA) ............................................ 22, 24,
Caenorhabditis elegans (C. elegans)..................... 169–186 27–29, 31, 33–35, 42, 48, 130
Calcium phosphate.......................................................... 40
Cancer-associated fibroblasts (CAFs)..........208–212, 221 H
Carboxyl propeptide of type I procollagen Herovici staining ......................................... 171, 175, 177
(PICP)................................................................ 190
Histopathology .................................................v, 135–144
Cell adhesion ............................................ v, 160–161, 205 Hybridization ....................................................... 135–144
Chondrocytes ...............................................158–160, 164 Hydroxyprolines............................ 14, 21, 60, 61, 63, 69,
Cloning ...........................................................3, 7, 10, 101 70, 82–84, 87, 88, 136, 170, 171, 177, 178, 181
Co-immunoprecipitations .......................... 102, 110, 111
Collagenase............................ 80, 83, 158, 159, 210, 223 I
Collagen fibers..................................................... 145, 150,
152, 153, 204, 221 Immature crosslinks ........................................... 80, 83–85
Collagen hybridizing peptide (CHP) ................. 135–143 Immunostaining............................................................ 137
Collagenopathies.........................................................3, 96 In situ..................................................... 40, 135–144, 158
Collagen proteostasis network ........................95, 96, 110 Integrins .........................v, 157–164, 205, 208, 211, 231
Collagen remodeling .................207, 208, 210, 212, 213 Interactomes....................................................v, 3, 95–112
Collagen VI ................................................................... 164 Isobaric Tag for Relative and Absolute
Collagen type I .........................18, 23, 24, 173, 192, 209 Quantitation (iTRAQ) .................... 112, 116, 117
Colons................................................................... 129–132
L
Cross-linking ........................................................ v, 23–26,
29, 31, 33–35, 221 Liver fibrosis ............................................. v, 189–200, 209
Cryogenic TEM (cryoTEM) ................ 40, 44–48, 50, 52 Lung fibrosis.................................................................. 204
Cuticles ....................................... 171, 175, 182–184, 186 Lung metastases ............................................................ 204
Cytotoxicity ..................................................................... 24 Lysyl oxidase (LOX) ......................................23, 204, 209
like-1 .......................................................................... 79
D like-2 ........................................................................ 205
Decellularization .................................................. 129–132 like-3 .......................................................................... 79
Dystrophic epidermolysis bullosa (DEB) ........................ 4 like-4 .......................................................................... 79
https://doi.org/10.1007/978-1-4939-9095-5, © Springer Science+Business Media, LLC, part of Springer Nature 2019
253
COLLAGEN
254 Index
M Skin .................................................................. v, 4, 18, 20,
21, 34, 70, 80, 83–85, 91, 95, 145–154, 170, 171,
Macrophages .............................. 24, 29, 31, 33, 203–213 173, 209, 224, 229
Mass spectrometry ...............................................v, 57, 61, Solid phase extraction (SPE) ................80, 81, 84, 85, 92
79–93, 95–112, 115, 117, 122, 123, 236, 244 Stochastic optical reconstruction microscopy
Matrix metalloproteinases (MMPs) ........................ v, 136, (STORM) .........................................40, 48–51, 53
190, 206, 210, 231–234, 245
Matrix superstructures ...................................................... 4 T
Mature crosslinks ............................................... 80, 83–85
Multiphoton microscopy ..................................v, 145–154 Tandem mass tag (TMT).............................112, 115–126
Terminal amine isotopic labeling of substrates
P (TAILS)..................................................... 115–126
Tissue ................................................ v, 11, 19–21, 23, 24,
Peptide-based drugs...................................................... 231 29, 31, 33, 39, 57, 60, 61, 63, 65, 66, 71, 73,
Procollagen C-proteinase enhancer-1 79–93, 95, 129, 135–137, 139–142, 145,
(PCPE-1) .................................................. 189–200 151–153, 169–186, 189, 190, 204, 206, 208,
Proteomics.................................v, 95–112, 122, 207, 213 210–212, 222, 225, 229, 230, 234
Transition state analog......................................... 231, 232
R
Transmission electron microscopy (TEM) ................... 40,
Radio frequency (RF) ...........................58, 68, 71, 85, 86 42–47, 51, 52, 161–164
Triple-helical inhibitor ......................................... 229–249
S Tumor-associates macrophages (TAMs)............. 208–213
Tumor microenvironment ....................... v, 203–213, 221
Scaffolds......................................................... v, 23–35, 95,
130, 203–205, 210, 213
U
Scanning electron microscope (SEM) ................ 129–132
Second harmonic generation (SHG) ..........................135, Ultra-high performance liquid chromatography-triple
145, 146, 148–153 quadrupole mass spectrometry
Single reaction monitoring (SRM) ..........................85, 86 (UHPLC-MS/MS) ................... 80, 82–83, 85–90

Irit Sagi - Nikolaos A Afratis - Collagen - Methods and Protocols-Humana Press (2019)

Uploaded by

Copyright:

Available Formats

Irit Sagi - Nikolaos A Afratis - Collagen - Methods and Protocols-Humana Press (2019)

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

Irit Sagi - Nikolaos A Afratis - Collagen - Methods and Protocols-Humana Press (2019)

Uploaded by

Copyright:

Available Formats

Methods in

Molecular Biology 1944

Irit Sagi · Nikolaos A. Afratis Editors

For further volumes:

Methods and Protocols

Irit Sagi and Nikolaos A. Afratis

ISSN 1064-3745 ISSN 1940-6029 (electronic)

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Rehovot, Israel Irit Sagi

PART I IN VITRO MODELS FOR COLLAGEN CHARACTERIZATION

PART II PROTEOMICS ANALYSIS OF COLLAGEN

5 Investigation of Triple-Helix Collagen Hydroxylation

PART III IMAGING APPROACHES FOR COLLAGEN

9 Sample Preparation of Extracellular Matrix of Murine

11 Combination of Traction Assays and Multiphoton Imaging

PART IV COLLAGEN IN HEALTH AND DISEASE

13 Assessing Collagen Deposition During Aging in Mammalian

JEAN-MARC ALLAIN  Laboratoire de Mécanique des Solides (LMS), Ecole Polytechnique,

AMAR JOSHI  Pharmaxis Drug Discovery, Frenchs Forest, NSW, Australia

ALINA C. TEUSCHER  Department of Health Sciences and Technology, Institute of

In Vitro Models for Collagen Characterization

Cloning and Mutagenesis Strategies for Large Collagens

Size, repetitive sequences, and GC-rich content are factors that

5–7]. By incorporating enzymatic assembly of overlapping DNA

We will here present our new optimized workflow for precision

2.1 Polymerase 1. Phusion® high-Fidelity PCR master mix with high-fidelity

2.2 Gibson Reaction 1. NEBuilder® HiFi DNA Assembly Cloning Kit.

2.3 Recombinant 1. Human Embryonic Kidney (HEK) 293 cells.

overlapping primer containing point mutations, one or more

Initial denaturation 98 C 2 min 1 cycle

8. After amplification, add 4 μl of loading dye to the PCR reac-

round-bottom tube. Incubate the mixture at 37 C overnight

mix 74 μl of Lipofectamine® 2000 with 500 μl of Opti-MEM

1. Mini-PROTEAN® TGX precast gels are suitable for routine

2. To transfer large and complex proteins, we prefer to use TB

11. Ascorbic acid is essential for synthesis of hydroxyproline and

Part of the work was financed by a research collaboration with

protein-1 family convert procollagen VII to 38:489–492. https://doi.org/10.1111/j.

Multi-well Fibrillation Assay as a Tool for Analyzing

All solutions should be prepared using analytical grade reagents.

2.2 Collagen 1. Sodium phosphate buffer (component a): To prepare 0.2 M

3.2 Collagen 1. Prepare 0.5 mg mL 1 collagen solution by dissolving the

Fig. 1 The electrophoretic mobility of collagen isolated from normal (N),

Fig. 2 Representative graph showing typical pattern of the collagen fibrillation

4. Switch on the plate reader, and pre-warm the plate chamber to

1. If isolating from skin, remove the epidermis using a scalpel

7. Use multichannel pipette to dispense collagen simultaneously

Production and Characterization of Chemically

Key words Collagen cross-linking, Biophysical characterization, Biomechanical properties, Cytotox-

Collagen is the most abundant extracellular matrix (ECM) protein

Considering its abundance in native ECM supramolecular

2.1 Collagen Cross- 1. 1 M sodium hydroxide (NaOH) in water.

7. 1% DPPA cross-linker solution: Dilute 2.5 μl of DPPA in

2.2 Swelling Assay 1. 1 phosphate-buffered saline (PBS).

2.3 Tensile Test 1. Scalpel and blades.

2.4 Cell Culture 1. Primary human dermal fibroblasts (DF; ATCC).

2.5 Cell Proliferation 1. 40 ,6-diamidino-2-phenylindole (DAPI): 1:3000 dilution in

2.6 Metabolic 1. 1 sterile phosphate-buffered saline (PBS).

2.7 Cell Viability 1. Dimethyl sulfoxide (DMSO).

JEAN-MARC ALLAIN Laboratoire de Mécanique des Solides (LMS), Ecole Polytechnique,

AMAR JOSHI Pharmaxis Drug Discovery, Frenchs Forest, NSW, Australia

ALINA C. TEUSCHER Department of Health Sciences and Technology, Institute of