BMC Microbiology BioMed Central

Research article Open Access

Impact of a synbiotic food on the gut microbial ecology and
metabolic profiles
Beatrice Vitali1, Maurice Ndagijimana2, Federica Cruciani1, Paola Carnevali3,
Marco Candela1, Maria Elisabetta Guerzoni2 and Patrizia Brigidi*1

Address: 1Department of Pharmaceutical Sciences, University of Bologna, Bologna, Italy, 2Department of Food Science, University of Bologna,
Bologna, Italy and 3R&D Food Microbiology & Bioprocess Research, Barilla G&R f.lli SpA, Parma, Italy
Email: Beatrice Vitali - [email protected]; Maurice Ndagijimana - [email protected];
Federica Cruciani - [email protected]; Paola Carnevali - [email protected]; Marco Candela - [email protected];
Maria Elisabetta Guerzoni - [email protected]; Patrizia Brigidi* - [email protected]
* Corresponding author

Published: 7 January 2010 Received: 17 July 2009

Accepted: 7 January 2010
BMC Microbiology 2010, 10:4 doi:10.1186/1471-2180-10-4
This article is available from: http://www.biomedcentral.com/1471-2180/10/4
© 2010 Vitali et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract
Background: The human gut harbors a diverse community of microorganisms which serve
numerous important functions for the host wellbeing. Functional foods are commonly used to
modulate the composition of the gut microbiota contributing to the maintenance of the host health
or prevention of disease. In the present study, we characterized the impact of one month intake of
a synbiotic food, containing fructooligosaccharides and the probiotic strains Lactobacillus helveticus
Bar13 and Bifidobacterium longum Bar33, on the gut microbiota composition and metabolic profiles
of 20 healthy subjects.
Results: The synbiotic food did not modify the overall structure of the gut microbiome, as
indicated by Polymerase Chain Reaction-Denaturing Gradient Gel Electrophoresis (PCR-DGGE).
The ability of the probiotic L. helveticus and B. longum strains to pass through the gastrointestinal
tract was hypothesized on the basis of real-time PCR data. In spite of a stable microbiota, the intake
of the synbiotic food resulted in a shift of the fecal metabolic profiles, highlighted by the Gas
Chromatography Mass Spectrometry Solid Phase Micro-Extraction (GC-MS/SPME) analysis. The
extent of short chain fatty acids (SCFA), ketones, carbon disulfide and methyl acetate was
significantly affected by the synbiotic food consumption. Furthermore, the Canonical discriminant
Analysis of Principal coordinates (CAP) of GC-MS/SPME profiles allowed a separation of the stool
samples recovered before and after the consumption of the functional food.
Conclusion: In this study we investigated the global impact of a dietary intervention on the gut
ecology and metabolism in healthy humans. We demonstrated that the intake of a synbiotic food
leads to a modulation of the gut metabolic activities with a maintenance of the gut biostructure. In
particular, the significant increase of SCFA, ketones, carbon disulfide and methyl acetate following
the feeding period suggests potential health promoting effects of the synbiotic food.

Page 1 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

Background sugar and lipid profile [21]. Another possibility in the

Humans can be considered as "superorganisms" with an microbiota modulation is the use of synbiotics, in which
internal ecosystem of diverse symbiotic microorganisms probiotics and prebiotics are used in combination. This
and parasites that have interactive metabolic processes. combination improves the survival of the probiotic
Their homeostatic balance is dependent upon the interac- strains, because specific substrates are readily available for
tions between the host and its microbial components [1]. their fermentation, and results in advantages to the host
The human intestine is home to some 100 trillion micro- that the live microorganisms and prebiotics offer [11].
organisms of at least 1000 species. The density of bacterial
cells in the colon has been estimated at 1011 to 1012 per The inadequacy of conventional culture techniques to
ml, which makes it one of the most densely populated reflect the microbial diversity of the intestinal ecosystem
microbial habitats known [2,3]. This microbial ecosystem has triggered the development of culture-independent
serves numerous important functions for the human host, 16S rRNA gene-based techniques for the evaluation of the
including protection against pathogens, nutrient process- effects of functional food administration in humans
ing, stimulation of angiogenesis, modulation of intestinal [22,23]. The latest frontier in the characterization of
immune response and regulation of host fat storage [4,5]. uncultured and complex microbial communities is the
The composition of the adult gastrointestinal microbiota high-throughput technology of pyrosequencing, which
has been intensely studied, using both cultivation and, achieves hundreds of thousands of sequences of a specific
more recently, culture-independent, small subunit (SSU) variable region within the small subunit of rRNA gene,
ribosomal DNA (rDNA) sequence-based methods [6-8]. consequently revealing the full diversity of an ecosystem
Members of the anaerobic genera Bacteroides, Eubacterium, [24,25]. However, since this approach is extremely labor
Clostridium, Ruminococcus, and Faecalibacterium have typi- intensive and time consuming, PCR-DGGE and real-time
cally been found to comprise a large majority of the PCR represent population fingerprinting methods, com-
human adult gut microbial community. In healthy adults, monly used to analyze the intestinal microbiota upon die-
the gut microbiota consists of a stable individual core of tary intervention. PCR-DGGE allows the visualization of
colonizing microorganisms surrounded by temporal visi- the predominant genetic diversity without prior knowl-
tors [9,10]. Fluctuations around this core of phylotypes edge of the composition or complexity of the microbial
are due to host genotype, diet, age, sex, organic disease ecosystem present in the sample [23,26]. Real-time PCR
and drugs (especially antibiotics) [11]. It has been shown enables specific intestinal bacterial populations to be
that the microbiota structure strongly influences the gut directly quantified by using DNA isolated from fecal
metabolic phenotype [12,13]. On short time scales, the material [23,27-29].
host-specific effects are relatively constant and changes in
the gut microbiome composition and activities are closely Gene expression profiling and proteomic approaches
influenced by dietary variations. have been applied to elucidate the molecular mechanisms
underlying symbiotic host-bacterial relationships [30-32].
An increasing awareness of the potential of gut microor- However, gene expression and proteomic data might only
ganisms to influence human health has led to widespread indicate the potential for physiological changes because
investigation of the relationship between the gut microbi- many pathway feedback mechanisms are simply not
ota and nutrients, particularly probiotics [14] and prebi- reflected in protein concentration or gene expression. On
otics [15] and their impact on the digestive system. the other hand, metabolite concentrations and their
Members of the genera Bifidobacterium and Lactobacillus, kinetic variations in tissues or biological matrixes repre-
natural components of the colonic microbiota, are the sent real end-points of physiological regulatory processes
most commonly used probiotic bacteria in many func- [1,33]. Metabonomics is defined as "the quantitative
tional foods and dietary supplements [16]. Postulated measurement of the dynamic multiparametric metabolic
health advantages associated to bifidobacteria and lacto- response of living systems to pathophysiological stimuli
bacilli include the inhibition of pathogenic microorgan- or genetic modification" [34]. Metabonomics provides a
isms, improvement of lactose digestion, reduction of systems approach to understand global metabolic regula-
serum cholesterol levels, prevention of cancer and tion of an organism and its commensal and symbiotic
enhancement of the host's immune system [17,18]. Sev- partners [1]. Recently, complementary metabonomic
eral oligosaccharides have been studied as potential preb- approaches have been employed for the biochemical
iotics, including lactulose, galactooligosaccharides and characterization of metabolic changes triggered by gut
fructooligosaccharides (FOS) [19]. Dietary supplements microbiota, dietary variation and stress interactions [35-
of prebiotics increase the content and proportion of bifi- 39]. Solid phase microextraction followed by gaschroma-
dobacteria [20] and exert positive effects on absorption of tography and mass spectrometry represents a novel
nutrients and minerals, synthesis of vitamins, prevention method for studying metabolic profiles of biological sam-
of constipation, colon cancer, and improvement of blood ples. This approach has been used to compare neonates

Page 2 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

and adult feces [40] and to identify volatile markers of Table 1: Similarity index (SI) of DGGE profiles related to T0 and
gastrointestinal disease [41]. T1

Subject SI (%)
In the present study, we characterized the impact of the
intake of a synbiotic snack on the gut microbiota compo- 1 71.8
sition and metabolic profiles of healthy subjects. The syn- 2 60.6
biotic snack contained the substrate FOS, whose prebiotic 3 79.2
effects are widely documented [42], and the probiotic 4 54.1
5 91.3
strains Lactobacillus helveticus Bar13 and Bifidobacterium
6 55.9
longum Bar33, which were selected on the basis of their 7 77.5
adhesion and immune-regolation properties, as assessed 8 47.7
by both in vitro [43] and in vivo studies on animal models 9 65.0
[44]. Co-variations were searched between the gut micro- 10 89.3
biome structure, as reflected by community DNA finger- 11 80.9
prints derived from PCR-DGGE and real-time PCR data, 12 38.2
13 76.1
and host metabolic phenotypes, as detected by GC-MS/
14 64.7
SPME. 15 66.6
16 59.4
Results 17 80.3
Effects of the synbiotic food on composition of the gut 18 64.3
microbiota 19 72.1
PCR-DGGE analysis with universal primers targeting the 20 46.4
V2-V3 region of the 16S rRNA gene was used to monitor
the impact of the synbiotic food intake on the predomi-
nant bacterial population (Figure 1A). Population finger- observed. Both qualitative (presence or absence of a
print profiles were compared and numerically analyzed band) or quantitative (variable intensity of a band) varia-
by FPQuest Software. DGGE band profiles (mean of tions did occur. These inter-individual variations were
bands: 15.3) were stable for each subject over a month of higher than changes elicited by the functional food con-
feeding with the functional food. Only a slight difference sumed.
in band richness was found between the time points of the
study (T0, mean of bands: 15.8; T1, mean of bands: 14.8). Quantitative variations of bifidobacteria and lactobacilli
DGGE bands were subjected to Mann-Whitney U-test in In order to evaluate the effect of the prebiotic component
order to search for significant differences in the intensities on modulation of bifidobacteria and lactobacilli popula-
between T0 and T1. No band showed a significant varia- tions and the capability of the probiotic bacteria to pass
tion, indicating that the consumption of the synbiotic through the gut of the healthy host, quantitative varia-
food did not alter the concentration of any major species tions of Bifidobacterium and Lactobacillus genera were
of intestinal microbiota. Pearson correlation was used to determined by real-time PCR and compared to the varia-
calculate the similarity index (SI) between DGGE band tions of the species B. longum and L. helveticus (Table 2).
profiles related to the time points T0 and T1 for each All volunteers naturally harbored strains belonging to Bifi-
healthy volunteer (Table 1). The high median value of SI dobacterium and Lactobacillus, as demonstrated by the pres-
(67.1%) revealed that the dominant bacterial composi- ence of these genera in all stool samples recovered before
tion remained constant over the treatment. Only 3 sub- the beginning of the feeding trial. B. longum was also
jects presented SIs lower than 50% (subjects 8, 12 and found in all healthy subjects at the time point T0, in
20). No subject showed significant variations between accordance with previous studies reporting B. longum as
DGGE band profiles related to T0 and T1, as evaluated one of the major bifidobacterial species in the intestinal
using the Pearson correlation analysis (P > 0.05). microbiota of human adults [29]. Differently, L. helveticus
was detected only in 8 subjects at the time point T0, show-
Cluster analysis of DGGE population profiling confirmed ing a frequency of 40%. L. helveticus is not a normal inhab-
the stability of the overall structure of the microbiome, itant of the intestinal microbiota, but strains belonging to
revealing no grouping according to the feeding (Figure 1B- this species are used as starter cultures in the manufactur-
C). T0 and T1 banding patterns were closely related for all ing of a variety of fermented dairy products, to modulate
the volunteers, except for the subject 8 (SI: 47.7%). flavor. Thus, presence of L. helveticus in fecal samples can
Among different subjects, considerable variation in the be related to a diet rich in yogurt and cheese [45]. Table 2
composition of the population fingerprints could be highlights different trends of variation of Bifidobacterium,

Page 3 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

B C

Figure
synbiotic
DGGE analysis
1 intakeof the fecal samples recovered from 20 healthy volunteers (s1-s20) before (T0) and after (T1) one month of the
DGGE analysis of the fecal samples recovered from 20 healthy volunteers (s1-s20) before (T0) and after (T1)
one month of the synbiotic intake. A: DGGE profiles related to fecal samples and L. helveticus Bar13 and B. longum Bar33
probiotic strains. B: line graph. C: Cluster analysis (Pearson correlation was used to calculate the similarity in DGGE profiles).

Page 4 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

Table 2: Real-time PCR quantification of bifidobacteria and lactobacilli

Subject Time point 16S rrn operons/μg fecal genomic DNA (mean ± SD)

Bifidobacterium B. longum Lactobacillus L. helveticus

1 T0 9.4 × 106 ± 3.7 × 106 3.2 × 106 ± 1.5 × 106 2.6 × 106 ± 9.6 × 105 0.0 ± 0.0
T1 4.1 × 106 ± 8.3 × 105 1.1 × 106 ± 2.9 × 105 1.9 × 106 ± 9.9 × 105 4.5 × 102 ± 2.9 × 102
2 T0 8.9 × 107 ± 3.1 × 107 4.2 × 107 ± 3.6 × 107 1.1 × 105 ± 5.6 × 104 9.0 × 101 ± 6.2 × 101
T1 1.6 × 107 ± 5.0 × 106 4.7 × 106 ± 2.9 × 105 5.1 × 105 ± 2.4 × 105 2.6 × 103 ± 2.8 × 102
3 T0 4.0 × 108 ± 3.6 × 107 8.6 × 106 ± 2.6 × 106 5.6 × 104 ± 3.5 × 104 0.0 ± 0.0
T1 2.4 × 108 ± 2.5 × 107 2.4 × 107 ± 2.9 × 106 2.6 × 105 ± 1.6 × 105 2.8 × 103 ± 1.8 × 103
4 T0 2.6 × 108 ± 2.8 × 107 2.3 × 107 ± 2.9 × 106 1.6 × 105 ± 1.0 × 103 2.1 × 103 ± 8.7 × 101
T1 5.8 × 108 ± 1.2 × 107 3.7 × 107 ± 3.1 × 106 1.2 × 105 ± 2.7 × 104 1.6 × 103 ± 2.2 × 102
5 T0 3.1 × 106 ± 8.6 × 105 9.8 × 105 ± 2.8 × 105 1.9 × 104 ± 5.8 × 103 0.0 ± 0.0
T1 2.4 × 106 ± 7.3 × 105 9.5 × 105 ± 3.4 × 105 6.1 × 104 ± 3.4 × 104 3.5 × 102 ± 2.3 × 102
6 T0 1.7 × 108 ± 3.8 × 107 6.5 × 106 ± 2.4 × 105 2.7 × 105 ± 1.2 × 105 0.0 ± 0.0
T1 6.2 × 108 ± 4.2 × 107 3.5 × 107 ± 2.0 × 105 1.7 × 105 ± 1.1 × 105 0.0 ± 0.0
7 T0 6.4 × 107 ± 4.8 × 106 3.4 × 107 ± 1.2 × 106 4.0 × 105 ± 1.7 × 105 9.0 × 101 ± 8.2 × 101
T1 7.5 × 107 ± 1.2 × 106 4.6 × 107 ± 5.5 × 106 9.2 × 105 ± 4.9 × 105 1.4 × 104 ± 3.2 × 103
8 T0 1.8 × 106 ± 5.8 × 105 6.0 × 105 ± 3.6 × 105 1.0 × 106 ± 1.0 × 106 0.0 ± 0.0
T1 4.1 × 106 ± 8.5 × 105 1.3 × 106 ± 9.7 × 105 1.7 × 105 ± 1.7 × 105 0.0 ± 0.0
9 T0 4.4 × 106 ± 2.8 × 105 3.0 × 106 ± 2.3 × 106 9.2 × 105 ± 9.0 × 105 3.0 × 103 ± 1.1 × 103
T1 5.6 × 106 ± 1.4 × 105 3.8 × 106 ± 1.3 × 106 2.0 × 106 ± 1.0 × 106 1.8 × 103 ± 1.7 × 103
10 T0 1.0 × 108 ± 1.8 × 107 7.0 × 107 ± 4.5 × 107 7.7 × 105 ± 7.6 × 105 0.0 ± 0.0
T1 3.3 × 108 ± 7.7 × 107 4.3 × 107 ± 2.5 × 107 1.3 × 106 ± 1.2 × 106 3.2 × 103 ± 2.7 × 103
11 T0 4.1 × 106 ± 7.5 × 105 1.2 × 106 ± 2.5 × 105 5.1 × 105 ± 4.1 × 105 6.0 × 102 ± 3.8 × 102
T1 3.4 × 107 ± 6.2 × 105 3.1 × 107 ± 1.0 × 107 7.8 × 105 ± 7.7 × 105 1.7 × 104 ± 3.1 × 103
12 T0 3.4 × 105 ± 7.6 × 104 7.5 × 102 ± 3.0 × 101 1.7 × 107 ± 1.1 × 107 0.0 ± 0.0
T1 1.3 × 106 ± 7.0 × 105 2.0 × 105 ± 9.3 × 104 5.8 × 105 ± 5.6 × 105 3.6 × 103 ± 6.4 × 102
13 T0 3.5 × 107 ± 1.6 × 106 1.2 × 107 ± 2.6 × 105 1.8 × 105 ± 1.0 × 105 0.0 ± 0.0
T1 2.3 × 107 ± 3.8 × 106 4.6 × 106 ± 4.4 × 105 2.5 × 105 ± 1.8 × 105 1.8 × 102 ± 4.3 × 101
14 T0 1.1 × 107 ± 6.9 × 105 2.3 × 106 ± 1.6 × 106 1.1 × 106 ± 1.8 × 105 0.0 ± 0.0
T1 5.4 × 107 ± 1.7 × 107 1.0 × 107 ± 6.5 × 106 7.2 × 105 ± 6.4 × 105 3.0 × 102 ± 3.0 × 101
15 T0 6.1 × 107 ± 7.4 × 106 1.7 × 107 ± 8.3 × 106 3.9 × 105 ± 2.9 × 105 1.8 × 101 ± 1.6 × 101
T1 2.5 × 107 ± 5.3 × 106 1.0 × 107 ± 5.8 × 106 2.5 × 105 ± 2.2 × 105 3.2 × 102 ± 1.4 × 102
16 T0 1.3 × 109 ± 4.5 × 108 4.0 × 107 ± 1.2 × 107 2.0 × 106 ± 1.1 × 106 0.0 ± 0.0
T1 1.3 × 109 ± 2.0 × 108 2.2 × 107 ± 3.8 × 106 1.0 × 106 ± 8.2 × 105 8.3 × 102 ± 1.4 × 101
17 T0 1.6 × 107 ± 1.6 × 106 5.0 × 106 ± 3.2 × 106 1.3 × 107 ± 2.9 × 106 1.3 × 102 ± 1.1 × 102
T1 2.2 × 107 ± 1.9 × 106 4.0 × 106 ± 2.7 × 106 1.5 × 107 ± 2.0 × 105 6.6 × 102 ± 9.5 × 101
18 T0 1.1 × 105 ± 3.1 × 106 1.4 × 103 ± 4.4 × 102 3.1 × 107 ± 2.7 × 107 0.0 ± 0.0
T1 3.7 × 105 ± 8.9 × 104 1.7 × 105 ± 7.3 × 104 3.0 × 106 ± 1.2 × 106 6.5 × 102 ± 1.2 × 102
19 T0 5.2 × 107 ± 1.7 × 107 4.3 × 105 ± 1.8 × 105 2.5 × 106 ± 1.9 × 106 0.0 ± 0.0
T1 2.0 × 107 ± 8.0 × 106 1.5 × 105 ± 9.4 × 104 2.0 × 106 ± 1.5 × 106 0.0 ± 0.0
20 T0 6.6 × 106 ± 5.2 × 106 4.4 × 106 ± 2.2 × 106 1.0 × 107 ± 8.4 × 106 1.8 × 103 ± 2.6 × 102
T1 7.0 × 106 ± 3.3 × 105 5.5 × 106 ± 3.3 × 106 2.7 × 105 ± 2.6 × 105 0.0 ± 0.0

Lactobacillus, B. longum and L. helveticus concentrations In order to assess the global impact of the functional food
among the subjects enrolled in the trial, suggesting a spe- consumption on the bifidobacteria and lactobacilli popu-
cific individual response to the dietary intervention. This lations, a statistical elaboration of the real-time PCR data
variability is particularly evident for L. helveticus. In the was performed. Box plots in Figure 2 show the amounts of
majority of the volunteers, the synbiotic intake was asso- 16S rrn operons of Bifidobacterium (A), B. longum (B),
ciated to an increase or to the appearance of this species. Lactobacillus (C) and L. helveticus (D) detected at the time
In 2 subjects (4 and 9) no variation was found at the time points T0 and T1 of the feeding study. The intake of the
point T1. In 4 subjects (6, 8, 19 and 20) L. helveticus did synbiotic food did not cause significant variations in the
not appear after the feeding period and in the subject 20 median value of Bifidobacterium (T0: 2.6 × 107; T1: 2.2 ×
it disappeared at the time point T1. These results indicate 107), B. longum (T0: 4.7 × 106; T1: 5.1 × 106) and Lactoba-
that the capability of L. helveticus Bar 13 to persist in the cillus (T0: 8.5 × 105; T1: 6.5 × 105). On the contrary, a sig-
gastrointestinal tract is related to the specific characteris- nificant increase (P < 0.05) of L. helveticus DNA was
tics of the host gut environment. observed after the administration of the functional food

Page 5 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

(T0 median value: 0; T1 median value: 6.6 × 102), demon-

strating the ability of L. helveticus Bar13 to pass through
the gut of healthy humans. The significant increase of L.
helveticus was directly linked to the low incidence of this
species in the intestine of the human host. Analogously,
the absence of significant variations in Bifidobacterium,
Lactobacillus and B. longum could be related to the high T0
amounts of these bacterial groups, natural inhabitants of
the gut microbiota of healthy humans. Amounts of L. hel-
veticus were evaluated by real-time PCR in stool samples
recovered from 10 subjects after a wash-out period of 20
days. Concentration of this species returned to a median
value of 0, supporting the hypothesis of a transient per-
sistence of the probiotic strain Bar13 during the feeding
period (data not shown).

Figure 3 shows the relationship between the variation of

B. longum species, expressed as the ratio of T1 and T0 16S
rrn operons, and the basal concentration of B. longum,
expressed as the number of 16S rrn operons measured at
the time point T0. The trend of the curve indicates a strong
influence of the initial concentration of B. longum on the
variation of B. longum population after the feeding period.
An evident increase of B. longum was observed in subjects
11, 12 and 18, who showed T0 amount of this species
minor or equal to 1.0 × 106 16S rrn operons per μg of total
bacterial DNA. Notably, subject 12, presenting the lowest
B. longum concentration at the time point T0 (7.5 × 102),
showed the highest variation of B. longum (T1/T0: 2.6 ×

Real-time
rium (A),to
related
Figure 2B.PCR
the
longum
time
evaluation
(B),
points
Lactobacillus
of
(T016S
andrrnT1)
(C)
operons
of
andtheL.of
feeding
helveticus
Bifidobacte-
study
(D)
Real-time PCR evaluation of 16S rrn operons of Bifi-
dobacterium (A), B. longum (B), Lactobacillus (C) and
L. helveticus (D) related to the time points (T0 and Figure
Relationship
operons)
16S rrn operons)
3 andbetween
B. longumB.amount
longum variations
before the(T1/T0
feeding16S
trialrrn(T0
T1) of the feeding study. Data are expressed as number Relationship between B. longum variations (T1/T0
of operons in 1 μg of total bacterial DNA extracted from the 16S rrn operons) and B. longum amount before the
feces. The box represents the interquartile range (25-75th feeding trial (T0 16S rrn operons). Empty circles indicate
percentile) and the line within the box is the median value. subjects with T0 value minor or equal to 1.0 × 106 16S rrn
The bottom and top bars indicate the 10th and 90th percen- operons per μg of total bacterial DNA. Filled circles indicate
tiles, respectively. Outlier values are indicated (black circles). subjects with T0 value higher than 1.0 × 106 16S rrn operons
* indicates a significant difference (P < 0.05). per μg of total bacterial DNA.

Page 6 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

102) after the synbiotic intake. The same subject presented changes were positively correlated with the synbiotic
the lowest SI (38.2%) between DGGE band profiles administration principally belonged to the families of
related to the time points T0 and T1. These data suggest ketones (methyl-5-hepten-2-one, 2-propanone, 2-
the capability of B. longum Bar33 to pass through the butanone, 2-pentanone, 2,3-butanedione) and SCFA
human gastrointestinal tract, but this property can be (acetic and valeric acid). Differently, the concentration of
detected only in subjects harboring low basal level of B. 1-octanol, thiophene and nonanone decreased signifi-
longum species. cantly after the feeding period. These results are showed in
the Figure 4, which reports the loadings plot obtained
Changes in intestinal metabolic profiles from the CAP analysis. The scores plot (canonical axe)
In this investigation about 130 different metabolites obtained from the same supervised method showed a per-
belonging to the families of alcohols, ketones, aldehydes, fect classification of the samples, on the basis of the syn-
sulfur compounds, nitrogen compounds and SCFA were biotic food intake (Figure 5). The application of the CAP
detected in feces by means of GC-MS/SPME analysis (see analysis on metabolites data set characterized by GC-MS/
Additional file 1). A two-tailed Mann-Whitney test was SPME resulted in classification and predictive abilities of
performed on the metabolic data matrix in order to iden- 100% (see Additional file 2), as evaluated by the leave-
tify the molecules significantly affected by the consump- four-out procedure used, using only a reduced number of
tion of the functional food. A CAP analysis performed on experimental chromatographic peaks as input variables.
the selected molecules evidenced that metabolites whose

CAP
Figure
loadings
4 plot of metabolites whose concentration was significantly affected by the intake of the synbiotic food (P < 0.05)
CAP loadings plot of metabolites whose concentration was significantly affected by the intake of the synbiotic
food (P < 0.05). Positive and negative coefficients indicate the increase or decrease of metabolite amounts following the feed-
ing period.

Page 7 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

T0 T1

Figure
CAP scores
5 plot of the stool samples collected from the twenty volunteers before (T0) and after (T1) the synbiotic food intake
CAP scores plot of the stool samples collected from the twenty volunteers before (T0) and after (T1) the syn-
biotic food intake.

Discussion DGGE profiles of the predominant fecal microbiota gen-

The significant involvement of the gut microbiota in the erated complex but overall relatively stable and unique
human health suggests that modulation of commensal profiles for each individual. Elaboration of DGGE data
microbial composition and metabolism through combi- revealed high SI values between T0 and T1 profiles, and
nations of probiotics and prebiotics could be a dietary no clustering of banding patterns according to the feed-
strategy to prevent diverse diseases, such as obesity, diabe- ing. These results demonstrated that no significant change
tes, non-alcoholic fatty liver disease, inflammatory bowel in the structure of the gut microbiota of healthy subjects
disease, and even cancers [4]. did occur following dietary intervention, confirming pre-
vious findings regarding the subject specificity of the pre-
In the present study, the impact of a synbiotic food sup- dominant fecal communities and their stability over time
plement on the gut microbiota structure of healthy and resistance to perturbations [9,23]. Notably, we cannot
humans was evaluated by using an integrated molecular exclude an effect of the synbiotic intake on minor bacte-
approach based on PCR-DGGE and real-time PCR. rial species, an effect that could be investigated using high-

Page 8 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

throughput sequencing techniques. However, the impact absorption by the protonic dissociation of ammonia and
on the dominant colonic microbiota represents the main other amines [47]. Other metabolites whose changes were
parameter to evaluate the clinical relevance for the use of positively and significantly correlated with the synbiotic
a functional food. intake belonged to the family of methylketones (methyl-
5-hepten-2-one, 2-propanone, 2-butanone, 2-pentanone,
Because DGGE can be considered a semiquantitative tool 2,3-butanedione). In particular, the significant increase of
for monitoring the dynamics of the predominant bacterial 2-pentanone can be regarded as the most interesting effect
species of an ecosystem, additional analysis with real-time associated with the synbiotic food intake. In fact, 2-pen-
PCR was performed to obtain a quantitative estimation of tanone, which is a naturally occurring compound in
the effect of the synbiotic intake on bifidobacteria and fruits, vegetables and fermented foods, has anti-inflam-
lactobacilli populations. In particular, variations in matory and chemopreventive properties. According to
amounts of B. longum and L. helveticus were evaluated in Pettersson et al. [48], it inhibits the prostaglandin produc-
order to assess the capability of the probiotic species tion and COX-2 protein expression in human colon can-
included in the synbiotic food to pass through the gas- cer cells. The increase of 2,3-butanedione is interesting
trointestinal tract of the human host. Only L. helveticus since it may have health benefits by impacting on the
concentration increased significantly after the ingestion of growth of some bacteria, such as L. delbrueckii subsp. bul-
the functional food, demonstrating the gut persistence of garicus ad Streptococcus thermophilus [41]. Furthermore,
the probiotic L. helveticus strain during the feeding period. during glucose catabolism 2,3-butanedione serves as an
Since L. helveticus species is not a natural inhabitant of the electron acceptor and can be reduced to 2,3-butanediol
human intestine and its presence in feces is diet related via acetoin. This pathway was shown to be important in
[45], this result was not surprising and suggests that low the removal of toxic amounts of pyruvate and in mainte-
abundant species could be optimal models for studying nance of pH homeostasis [49]. A diverse range of sulfur
the gut colonization of probiotic bacteria. On the other compounds has been identified in stool samples [41]. The
hand, visualization of the gut colonization of a high usual source of sulfur compounds is the microbial break-
abundant species, such as B. longum, is strictly related to down of sulfur containing amino acids and the increase of
its basal concentration. For this reason, we observed the B. these compounds suggests an abundance or metabolic
longum increase only in subjects with the lowest concen- activity of bacteria able to breakdown cystein and methio-
tration of B. longum species at the time point T0. nine. In our study, a significant increase of carbon
disulfide was observed following the feeding period. Car-
The intake of the synbiotic food resulted in significant bon disulfide may be produced by carbonation of hydro-
changes in some gut metabolic activities, as highlighted gen sulphide as a detoxification mechanism exerted by
by the CAP analysis of the fecal metabolic profiles, which colonic bacteria. According to Garner et al. [41], carbon
pointed out a separation of fecal samples of the subjects disulfide has been found in 100% of the samples from
on the basis of the synbiotic food intake. Surprisingly little healthy donors and absent in many samples of patients
is known about volatile organic compounds formed in with Campylobacter jejuni and Clostridium difficile. Various
the gut. GC-MS/SPME, detecting volatile molecules with esters were detected in all fecal samples. In particular, a
high sensitivity, represents a suitable approach to identify significant increase of methyl acetate, ester of methanol
microbial metabolites in fecal samples, such as SCFAs, and acetic acid, was evident after the synbiotic intake.
ketones, esters and sulfur compounds [46]. Methanol is rarely found as free alcohol in the gut, where
it is generated from the breakdown of macromolecules
Two SCFAs, acetic and valeric acids, were the metabolites including pectins, bran and aspartame. In general, free
showing the highest increase after the synbiotic adminis- alcohols and endogenous fatty acids are metabolized into
tration. Although a general increase was observed also for fatty acid esters in liver, pancreas and intestine [50]. At the
butyric acid, this variation was not statistically significant intestinal site, esterification of alcohols by colonic bacte-
due to the high variability of the measures. SCFAs are very ria can be regarded as a microbial strategy to remove or
common in the gut environment, arising from metabo- trap toxic molecules such as fatty acids and alcohols.
lism of undigested carbohydrates, such as dietary fiber
and prebiotics, by colonic bacteria. The increase of SCFAs To sum up, the investigation of the fecal volatile metabo-
is particularly interesting, as they play a role in regulation lites by GC-MS/SPME allowed to correlate the consump-
of cell proliferation and differentiation of the colonic epi- tion of the synbiotic food with the stimulation of health-
thelial cells. Increases in SCFA production have been asso- promoting metabolic activities of the gut microbiota, such
ciated with decreased pH, which may reduce potential as regulation of the colonic epithelial cell proliferation
pathogenic clostridia, decreased solubility of bile acids, and differentiation, anti-inflammatory and chemopreven-
increased absorption of minerals, and reduced ammonia tive properties and detoxification processes.

Page 9 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

Conclusion using QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Ger-
In the current study molecular fingerprinting techniques many), according to the manufacturer's instructions.
(PCR-DGGE and real-time PCR) were integrated to the
GC-MS/SPME analysis of the metabolic profiles to inves- PCR-DGGE and cluster analysis
tigate the global impact of a dietary intervention on the Amplification of the V2-V3 region of the bacterial 16S
gut ecology and metabolism in healthy humans. In partic- rRNA gene was carried out using the universal eubacterial
ular, the major findings of this study are the following: (i) primers GCclamp-HDA1 and HDA2 [51], supplied by M-
the synbiotic food does not modify the overall structure of Medical (Milan, Italy). The amplification reactions were
the gut microbiome, as detected by DGGE; (ii) the gut sur- performed in a Biometra Thermal Cycler T Gradient
vival of the probiotic strains may be supposed on the basis (Biometra, Göttingen, Germany). AmpliTaq Gold DNA
of the increase of B. longum and L. helveticus after the syn- Polymerase (Applied Biosystem, Foster City, CA) was used
biotic consumption; (iii) the perturbation of the gut as thermostable DNA polymerase. The reaction mixture
metabolism triggered by a synbiotic food intake generates contained 0.5 μM of each primer, 200 μM of each dNTP,
significant changes in the GC-MS/SPME profiles; (iv) 0.5 U of DNA Polymerase, and 4 μl of the bacterial DNA
changes in metabolic phenotypes seem to indicate poten- template in a final volume of 50 μl. The thermocycle pro-
tial implications of the synbiotic food in health mainte- gram consisted of the following time and temperature
nance and prevention of diverse diseases. profile: 95°C for 15 min; 30 cycles of 95°C for 60 s, 56°C
for 30 s, 72°C for 30 s; and 72°C for 8 min. A volume of
In order to better investigate the mechanistic basis of the 15-20 μl of PCR samples was used for DGGE analysis,
probiotics and prebiotics action on gut microbial activi- which was performed by using the D-Code Universal
ties and the outcomes on human health, it will be neces- Mutation System Apparatus (Bio-Rad, Los Angeles, CA),
sary to integrate the GC-MS/SPME and/or NMR profiles of as previously described [52]. Briefly, the sequence-specific
feces with simultaneous analysis of different biofluids, separation of the PCR fragments was obtained in 8% (w/
including urine and plasma. v) polyacrylamide gels, containing a 30% to 50% gradient
of urea and formamide. Electrophoresis was started at a
Methods voltage of 250 V for 5 minutes and continued at constant
Study population voltage of 90 V and temperature of 60°C for 16 h. Follow-
Twenty randomly selected healthy volunteers (11 women ing electrophoresis, the gel was silver stained [53] and
and 9 men) aged between 20 and 50 (mean: 35) partici- scanned using a Molecular Imager Gel Doc XR System
pated in the study. The Ethics Committee of the University (Bio-Rad). DGGE gel images were analyzed using the
of Bologna (Italy) approved the study, and all subjects FPQuest Software Version 4.5 (Bio-Rad). In order to com-
gave informed consent. None of the subjects had a history pensate for gel-to-gel differences and external distortion
of gastrointestinal or metabolic disease or previous sur- to electrophoresis, the DGGE patterns were aligned and
gery (apart from appendectomy). The subjects did not normalized using an external reference ladder, containing
receive antibiotic treatment or any other medical treat- PCR amplicons from pure cultures of intestinal bacterial
ment influencing intestinal microbiota during 3 months species. A cluster analysis of the DGGE patterns was per-
before the start of the study. Subjects maintained their formed using the FPQuest Software. The similarity in the
usual diet during the study period. All the volunteers had profiles was calculated on the basis of the Pearson corre-
normal weight with a body mass index in the range 18.5- lation coefficient with the Ward clustering algorithm.
24.9. The volunteers received one dose of a synbiotic
snack (Barilla, Parma, Italy), twice a day for a period of 1 Development of L. helveticus species-specific primers
month. The synbiotic bar consisted of a biscuit covered by By using 16S and 16S-23S rRNA sequences obtained from
chocolate. The biscuit contained 500 mg of FOS (Acti- the DDBJ and EMBL databases, multiple alignments of
light® 950P, Marckolsheim, France) and the chocolate sequences related to L. helveticus and reference organisms
included a mixture of the probiotic strains B. longum were constructed with the program Clustal W http://
Bar33 and L. helveticus Bar13 (Barilla culture collection). www.ebi.ac.uk/Tools/clustalw2. Potential target sites for
109CFU of each probiotic strain were present in a dose of specific detection of the species L. helveticus were identi-
the synbiotic bar. fied and the following primers were designed: F_Hel (5'-
GTGCCATCCTAAGAGATTAGGA-3') and R_Hel (5'-
Extraction of DNA from fecal samples TATCTCTACTCTCCATCACTTC-3'). A Blast search http://
Stool samples were collected from volunteers before the www.ncbi.nlm.nih.gov/BLAST was carried out to test the
start of the feeding study (T0) and at the end of the inges- virtual specificity of the primers. Validation of specificity
tion period (T1) and immediately frozen at -80°C until was performed by PCR experiments against different spe-
use. Total DNA was extracted from 230 mg of feces by cies of Lactobacillus (L. acidophilus, L. casei, L. plantarum, L.

Page 10 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

bulgaricus, L. reuteri, L. gasseri, L. johnsonii) and other intes- min at 45°C. The SPME fiber was exposed to each sample
tinal genera (Bifidobacterium, Streptococcus, Escherichia). for 40 min and then was inserted into the injection port
The primers were synthesized by M-Medical (Milan, Italy) of the GC for a 5 min sample desorption. GC-MS analyses
and optimal annealing temperature was established by were performed on an Agilent 7890A gaschromatograph
gradient PCR. (Agilent Technologies, Palo Alto, CA) coupled to an Agi-
lent 5975C mass selective detector operating in electron
Real-time quantitative PCR impact mode (ionization voltage 70 eV). A Supelcowax 10
Quantitative PCR was performed in a LightCycler instru- capillary column (60 m length, 0.32 mm ID) was used
ment (Roche, Mannheim, Germany) and SYBR Green I (Supelco). The temperature program was: 50°C for 1 min,
fluorophore was used to correlate the amount of PCR then programmed at 4.5°C/min to 65°C and finally at
product with the fluorescence signal. The following 10°C/min to 230°C which was maintained for 25 min.
genus- and species-specific primers sets, targeted to 16S or Injector, interface and ion source temperatures were 250,
16S-23S rRNA sequences, were used: Bif164/Bif662 (Bifi- 250 and 230°C, respectively. The mass-to-charge ratio
dobacterium [54]); Lac1/Lab0677r (Lactobacillus [55,56]); interval was 30-350 Da at 2.9 scans per second. Injections
BiLON1/BiLON2 (B. longum [29]); F_Hel/R_Hel (L. helve- were performed in splitless mode and helium (1 ml/min)
ticus [this work]). Three sub-samples of each DNA extract was used as carrier gas. The identification of all the mole-
were amplified in a final volume of 20 μl containing 4 cules detected in fecal samples was based on comparison
mM of MgCl2, 0.5 μM of each primer, 2 μl of LightCycler- of their retention times and spectral data with those of
FastStart DNA Master SYBR Green I (Roche), and either 2 pure compounds (Sigma-Aldrich, Milan, Italy) analyzed
μl of template or water (no-template control). The ther- in the same conditions. The identification was further
mal cycling conditions were as follows: an initial denatur- confirmed by comparing mass spectra of all compounds
ation step at 95°C for 10 min followed by 40 cycles of with those contained in available databases (NIST version
denaturation at 95°C for 15 s; primer annealing at 60°C 2005 and Wiley version 1996) and in literature [41].
(Bifidobacterium), 65°C (Lactobacillus and B. longum) and Quantitative data of the identified compounds were
63°C (L. helveticus) for 25 s; extension at 72°C for 25 s obtained by interpolation of the relative areas versus the
(Bifidobacterium), 20 s (Lactobacillus), 45 s (B. longum) and internal standard area, in calibration curves built with
10 s (L. helveticus) and a fluorescence acquisition step at pure reference compounds. The concentration of volatile
90°C (Bifidobacterium and B. longum) or 85°C (Lactobacil- compounds, for which there were no pure references, was
lus and L. helveticus) for 5 s. For each step the temperature obtained by using the same calibration graphs of the com-
transition rate was 20°C/s. Quantification of rrn operons pounds with the most similar chemical structure.
of Bifidobacterium, Lactobacillus and B. longum was done by
using standard curves made from known concentrations Statistical analyses
of genomic DNA from the sequenced strains B. longum For each subject, variations of the DGGE profiles related
NCC2705 [30] and L. acidophilus NCFM [57]. For L. helve- to the time points T0 and T1 were analyzed by Pearson
ticus species the probiotic strain included in the synbiotic correlation. Significant differences in the intensity of each
was used as standard and the number of rrn operons in DGGE band among all fecal samples were searched by
the genome was deduced from the sequenced genome of using Mann-Whitney U-test. Mann-Whitney U-test was
L. helveticus DPC 4571 [58]. Chromosomal DNA of the also used to analyze differences in total rrn operons of tar-
strains used as standards was extracted by using DNeasy get genera and species and to determine metabolites sig-
Tissue Kit (Qiagen) and serially diluted from 105 to 101 nificantly affected by the synbiotic food intake. A P value
molecules/μl. Results obtained by PCR were converted to below 0.05 was considered statistically significant. Metab-
the average estimate of total rrn operons from each group olites with a P value below 0.05 were then used in further
present in 1 μg of total DNA, and standard deviations multivariate analysis. These selected metabolites formed a
(SD) were calculated. matrix containing two kinds of information: the effects of
the synbiotic food intake (within-individual variability)
GC-MS/SPME and the natural differences between individuals (between-
A carboxen-polydimethylsiloxane coated fiber (85 μm) individuals variability). These two kinds of information
and a manual SPME holder (Supelco, Bellefonte, PA) were were separated following the method of Jansen et al. [59].
used in this study after preconditioning according to the A CAP analysis was then performed on the within-individ-
manufacturer's instruction manual. Before each head ual variability matrix [60]. The CAP constrained ordina-
space sampling, the fiber was exposed to the GC inlet for tion procedure can be summarized as follows: the data
5 min for thermal desorption at 250°C in a blank sample. were reduced by performing a principal coordinate analy-
Five ml of fecal slurries (20%) were placed in 10 ml glass sis (PCO) on the parameters using a dissimilarity measure
vials, added with 4-methyl-2-pentanol (4 mg/l) as inter- based on Euclidean distances; an appropriate number of
nal standard. The samples were then equilibrated for 10 PCOs were chosen non-arbitrarily, which maximize the

Page 11 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

number of observations correctly classified [61,60]. The 6. Vaughan EE, Schut F, Heilig HG, Zoetendal EG, de Vos WM, Akker-
mans AD: A molecular view of the intestinal ecosystem. Curr
robustness of the model obtained was established by a 4- Issues Intest Microbiol 2000, 1:1-12.
fold cross validation method, repeatedly leaving out a 7. Gill SR, Pop M, Deboy RT, Eckburg PB, Turnbaugh PJ, Samuel BS, Gor-
fourth of the samples and predicting them back into the don JI, Relman DA, Fraser-Liggett CM, Nelson KE: Metagenomic
analysis of the human distal gut microbiome. Science 2006,
model [62]. Finally a traditional canonical analysis on the 312:1355-1359.
first three PCOs was performed. The hypothesis of no sig- 8. Palmer C, Bik EM, Eisen MB, Eckburg PB, Sana TR, Wolber PK, Rel-
nificant difference in multivariate location among the man DA, Brown PO: Rapid quantitative profiling of complex
microbial populations. Nucleic Acids Res 2006, 34:e5.
groups was tested by using a permutation test based on 9. Zoetendal EG, Akkermans AD, de Vos WM: Temperature gradi-
9999 permutations. ent gel electrophoresis analysis of 16S rRNA from human
fecal samples reveals stable and host-specific communities of
active bacteria. Appl Environ Microbiol 1998, 64:3854-3859.
Statistical analyses were performed using the software Sig- 10. Zoetendal EG, Rajilic-Stojanovic M, de Vos WM: High-throughput
maStat (Systat Sofware Inc., San Jose, CA) and the package diversity and functionality analysis of the gastrointestinal
tract microbiota. Gut 2008, 57:1605-1615.
Canoco for Windows 4.5 (Microcomputer Power, Ithaca, 11. Collins MD, Gibson GR: Probiotics, prebiotics, and synbiotics:
NY). approaches for modulating the microbial ecology of the gut.
Am J Clin Nutr 1999, 69(Suppl):1052-1057.
12. Li M, Wang B, Zhang M, Rantalainen M, Wang S, Zhou H, Zhang Y,
Authors' contributions Shen J, Pang X, Zhang M, Wei H, Chen Y, Lu H, Zuo J, Su M, Qiu Y,
BV performed the study design, analysis and interpreta- Jia W, Xiao C, Smith LM, Yang S, Holmes E, Tang H, Zhao G, Nichol-
son JK, Li L, Zhao L: Symbiotic gut microbes modulate human
tion of the data and the writing of the paper. FC and MC metabolic phenotypes. Proc Natl Acad Sci USA 2008,
performed the DGGE and real time experiments and sta- 105:2117-2122.
tistical analysis of the data. MN carried out GC-MS/SPME 13. Nicholson JK, Holmes E, Wilson ID: Gut microorganisms, mam-
malian metabolism and personalized health care. Nat Rev
experiments. PC, MEG and PB coordinated the study. All Microbiol 2005, 3:431-438.
authors read and approved the manuscript. 14. Fuller R: A review: probiotics in man and animals. J Appl Bacte-
riol 1989, 66:365-378.
15. Gibson GR, Roberfroid MB: Dietary modulation of the human
Additional material colonic microbiota: introducing the concept of prebiotics. J
Nutr 1995, 125:1401-1412.
16. Guarner F, Malagelada JR: Gut flora in health and disease. Lancet
2003, 361:512-519.
Additional file 1 17. Parvez S, Malik KA, Ah Kang S, Kim HY: Probiotics and their fer-
Metabolites detected by GC-MS/SPME analysis. Metabolites were iden- mented food products are beneficial for health. J Appl Microbiol
tified and quantified (mg/kg) in stool samples collected from 20 volun- 2006, 100:1171-1185.
teers before (T0) and after (T1) the synbiotic food intake. 18. Farnworth ER: The evidence to support health claims for pro-
Click here for file biotics. J Nutr 2008, 138(suppl):1250-1254.
19. Cummings JH, Macfarlane GT, Englyst HN: Prebiotic digestion and
[http://www.biomedcentral.com/content/supplementary/1471-
fermentation. Am J Clin Nutr 2001, 73(suppl):415-420.
2180-10-4-S1.DOC] 20. Gibson GR: Dietary modulation of the human gut microflora
using prebiotics. Br J Nutr 1998, 80(suppl):209-212.
Additional file 2 21. Goetze O, Fruehauf H, Pohl D, Giarrè M, Rochat F, Ornstein K,
Menne D, Fried M, Thumshirn M: Effect of a prebiotic mixture on
Confusion matrix. Confusion matrix derived by 4-fold cross-validation of
intestinal comfort and general wellbeing in health. Br J Nutr
CAP model obtained using metabolites identified in stool samples collected 2008, 100:1077-1085.
from 20 volunteers before (T0) and after (T1) the synbiotic food intake. 22. Suau A, Bonnet R, Sutren M, Godon JJ, Gibson GR, Collins MD, Doré
Click here for file J: Direct analysis of genes encoding 16S rRNA from complex
[http://www.biomedcentral.com/content/supplementary/1471- communities reveals many novel molecular species within
2180-10-4-S2.DOC] the human gut. Appl Environ Microbiol 1999, 65:4799-4807.
23. Vanhoutte T, De Preter V, De Brandt E, Verbeke K, Swings J, Huys G:
Molecular monitoring of the fecal microbiota of healthy
human subjects during administration of lactulose and Sac-
charomyces boulardii. Appl Environ Microbiol 2006, 72:5990-5997.
24. Andersson AF, Lindberg M, Jakobsson H, Bäckhed F, Nyrén P, Eng-
References strand L: Comparative analysis of human gut microbiota by
1. Rezzi S, Ramadan Z, Fay LB, Kochhar S: Nutritional metabonom- barcoded pyrosequencing. PLoS One 2008, 3:e2836.
ics: applications and perspectives. J Proteome Res 2007, 25. Armougom F, Raoult D: Use of pyrosequencing and DNA bar-
6:513-525. codes to monitor variations in Firmicutes and Bacteroidetes
2. Eckburg PB, Bik EM, Bernstein CN, Purdom E, Dethlefsen L, Sargent communities in the gut microbiota of obese humans. BMC
M, Gill SR, Nelson KE, Relman DA: Diversity of the human intes- Genomics 2008, 9:576.
tinal microbial flora. Science 2005, 308:1635-1638. 26. Tannock GW, Munro K, Bibiloni R, Simon MA, Hargreaves P, Gopal
3. Ley RE, Peterson DA, Gordon JI: Ecological and evolutionary P, Harmsen H, Welling G: Impact of consumption of oligosac-
forces shaping microbial diversity in the human intestine. charide-containing biscuits on the fecal microbiota of
Cell 2006, 124:837-848. humans. Appl Environ Microbiol 2004, 70:2129-2136.
4. Bäckhed F, Ley RE, Sonnenburg JL, Peterson DA, Gordon JI: Host- 27. Malinen E, Kassinen A, Rinttilä T, Palva A: Comparison of real-
bacterial mutualism in the human intestine. Science 2005, time PCR with SYBR Green I or 5'-nuclease assays and dot-
307:1915-1920. blot hybridization with rDNA-targeted oligonucleotide
5. Palmer C, Bik EM, Digiulio DB, Relman DA, Brown PO: Develop- probes in quantification of selected faecal bacteria. Microbiol-
ment of the human infant intestinal microbiota. PLoS Biol ogy 2003, 149:269-277.
2007, 5:1556-1573.

Page 12 of 13
(page number not for citation purposes)
BMC Microbiology 2010, 10:4 http://www.biomedcentral.com/1471-2180/10/4

28. Bartosch S, Fite A, Macfarlane GT, McMurdo ME: Characterization human feces by PCR methods. FEMS Microbiol Lett 2004,
of bacterial communities in feces from healthy elderly volun- 231:125-130.
teers and hospitalized elderly patients by using real-time 46. Ndagijimana M, Vallicelli M, Cocconcelli PS, Cappa F, Patrignani F, Lan-
PCR and effects of antibiotic treatment on the fecal micro- ciotti R, Guerzoni ME: Two 2[5H]-furanones as possible signal-
biota. Appl Environ Microbiol 2004, 70:3575-3581. ing molecules in Lactobacillus helveticus. Appl Environ Microbiol
29. Matsuki T, Watanabe K, Fujimoto J, Kado Y, Takada T, Matsumoto K, 2006, 72:6053-6061.
Tanaka R: Quantitative PCR with 16S rRNA-gene-targeted 47. Wong JMW, Jenkins DJA: Carbohydrate digestibility and meta-
species-specific primers for analysis of human intestinal bifi- bolic effects. J Nutr 2007, 137(suppl):2539-2546.
dobacteria. Appl Environ Microbiol 2004, 70:167-173. 48. Pettersson J, Karlsson PC, Göransson U, Rafter JJ, Bohlin L: The fla-
30. Schell MA, Karmirantzou M, Snel B, Vilanova D, Berger B, Pessi G, vouring phytochemical 2-pentanone reduces prostaglandin
Zwahlen MC, Desiere F, Bork P, Delley M, Pridmore RD, Arigoni F: production and COX-2 expression in colon cancer cells. Biol
The genome sequence of Bifidobacterium longum reflects its Pharm Bull 2008, 31:534-537.
adaptation to the human gastrointestinal tract. Proc Natl Acad 49. Ott A, Germond JE, Chaintreau A: Vicinal diketone formation in
Sci USA 2002, 99:14422-14427. yogurt: 13C precursors and effect of branched-chain amino
31. Xu J, Bjursell MK, Himrod J, Deng S, Carmichael LK, Chiang HC, acids. J Agric Food Chem 2000, 48:724-731.
Hooper LV, Gordon JI: A genomic view of the human-Bacter- 50. Diczfalusy MA, Björkhem I, Einarsson C, Hillebrant CG, Alexson SE:
oides thetaiotaomicron symbiosis. Science 2003, 299:2074-2076. Characterization of enzymes involved in formation of ethyl
32. Pridmore RD, Berger B, Desiere F, Vilanova D, Barretto C, Pittet AC, esters of long-chain fatty acids in humans. J Lipid Res 2001,
Zwahlen MC, Rouvet M, Altermann E, Barrangou R, Mollet B, Mer- 42:1025-1032.
cenier A, Klaenhammer T, Arigoni F, Schell MA: The genome 51. Walter J, Tannock GW, Tilsala-Timisjarvi A, Rodtong S, Loach DM,
sequence of the probiotic intestinal bacterium Lactobacillus Munro K, Alatossava T: Detection and identification of gastroin-
johnsonii NCC 533. Proc Natl Acad Sci USA 2004, 101:2512-2517. testinal Lactobacillus species by using denaturing gradient gel
33. Holmes E, Wilson ID, Nicholson JK: Metabolic phenotyping in electrophoresis and species-specific PCR primers. Appl Envi-
health and disease. Cell 2008, 134:714-717. ron Microbiol 2000, 66:297-303.
34. Nicholson JK, Lindon JC, Holmes E: Metabonomics: understand- 52. Vitali B, Pugliese C, Biagi E, Candela M, Turroni S, Bellen G, Donders
ing the metabolic responses of living systems to pathophysi- GGG, Brigidi P: Dynamics of vaginal bacterial communities in
ological stimuli via multivariate statistical analysis of women developing bacterial vaginosis, candidiasis, or no
biological NMR spectroscopic data. Xenobiotica 1999, infection, analyzed by PCR-denaturing gradient gel electro-
29:1181-1189. phoresis and real-time PCR. Appl Environ Microbiol 2007,
35. Wang Y, Tang H, Nicholson JK, Hylands PJ, Sampson J, Holmes E: A 73:5731-5741.
metabonomic strategy for the detection of the metabolic 53. Bassam BJ, Caetano-Anollés G, Gresshoff PM: Fast and sensitive
effects of chamomile (Matricaria recutita L.) ingestion. J Agric silver staining of DNA in polyacrylamide gels. Anal Biochem
Food Chem 2005, 53:191-196. 1991, 196:80-83.
36. Marchesi JR, Holmes E, Khan F, Kochhar S, Scanlan P, Shanahan F, 54. Kok RG, de Waal A, Schut F, Welling GW, Weenk G, Hellingwerf KJ:
Wilson ID, Wang Y: Rapid and noninvasive metabonomic char- Specific detection and analysis of a probiotic Bifidobacterium
acterization of inflammatory bowel disease. J Proteome Res strain in infant feces. Appl Environ Microbiol 1996, 62:3668-3672.
2007, 6:546-551. 55. Walter J, Hertel C, Tannock GW, Lis CM, Munro K, Hammes WP:
37. Martin FP, Dumas ME, Wang Y, Legido-Quigley C, Yap IK, Tang H, Detection of Lactobacillus, Pediococcus, Leuconostoc and
Zirah S, Murphy GM, Cloarec O, Lindon JC, Sprenger N, Fay LB, Weissella species in human feces by using group-specific PCR
Kochhar S, van Bladeren P, Holmes E, Nicholson JK: A top-down primers and denaturing gradient gel electrophoresis. Appl
systems biology view of microbiome-mammalian metabolic Environ Microbiol 2001, 67:2578-2585.
interactions in a mouse model. Mol Syst Biol 2007, 3:112. 56. Heilig HGHJ, Zoetendal EG, Vaughan EE, Marteau P, Akkermans ADL,
38. Martin FP, Wang Y, Sprenger N, Holmes E, Lindon JC, Kochhar S, de Vos WM: Molecular diversity of Lactobacillus spp. and other
Nicholson JK: Effects of probiotic Lactobacillus paracasei treat- lactic acid bacteria in the human intestine as determined by
ment on the host gut tissue metabolic profiles probed via specific amplification of 16S ribosomal DNA. Appl Environ
magic-angle-spinning NMR spectroscopy. J Proteome Res 2007, Microbiol 2002, 68:114-123.
6:1471-1481. 57. Altermann E, Russel WM, Azcarate-Peril MA, Barrangou R, Buck BL,
39. Martin FP, Wang Y, Sprenger N, Yap IK, Lundstedt T, Lek P, Rezzi S, McAuliffe O, Souther N, Dobson A, Duong T, Callanan M, Lick S,
Ramadan Z, van Bladeren P, Fay LB, Kochhar S, Lindon JC, Holmes E, Hamrick A, Cano R, Klaenhammer TR: Complete genome
Nicholson JK: Probiotic modulation of symbiotic gut micro- sequence of the probiotic lactic acid bacterium Lactobacillus
bial-host metabolic interactions in a humanized microbiome acidophilus NCFM. Proc Natl Acad Sci USA 2005, 102:3906-3912.
mouse model. Mol Syst Biol 2008, 4:157. 58. Callanan M, Kaleta P, O'Callaghan J, O'Sullivan O, Jordan K, McAuliffe
40. De Lacy Costello B, Ewen R, Ewer AK, Garner CE, Probert CSJ, Rat- O, Sangrador-Vegas A, Slattery L, Fitzgerald GF, Beresford T, Ross
cliffe NM, Smith S: An analysis of volatiles in the headspace of RP: Genome sequence of Lactobacillus helveticus, an organism
the faeces of neonates. J Breath Res 2008, 2:1-8. distinguished by selective gene loss and insertion sequence
41. Garner EC, Smith S, Costello BL, White P, Spencer R, Probert CSJ, element expansion. J Bacteriol 2008, 190:727-735.
Ratcliffe NM: Volatile organic compounds from feces and their 59. Jansen JJ, Hoefsloot HCJ, Greef J Van der, Timmerman ME, Smilde AK:
potential for diagnosis of gastrointestinal disease. Faseb J Multilevel component analysis of time-resolved metabolic
2007, 21:1675-1688. fingerprint data. Anal Chim Acta 2005, 530:173-183.
42. Probert HM, Gibson GR: Investigating the prebiotic and gas- 60. Anderson MJ, Willis TJ: Canonical analysis of principal coordi-
generating effects of selected carbohydrates on the human nates: a useful method of constrained ordination for ecology.
colonic microflora. Lett Appl Microbiol 2002, 35:473-480. Ecology 2003, 84:511-525.
43. Candela M, Perna F, Carnevali P, Vitali B, Ciati R, Gionchetti P, 61. Anderson MJ, Robinson J: Generalized discriminant analysis
Rizzello F, Campieri M, Brigidi P: Interaction of probiotic Lacto- based on distances. Aust Nz J Stat 2003, 45:301-318.
bacillus and Bifidobacterium strains with human intestinal epi- 62. Lindon JC, Nicholson JK, Holmes E, Keun HC, Craig A, Pearce JT,
thelial cells: adhesion properties, competition against Bruce SJ, Hardy N, Sansone SA, Antti H, Jonsson P, Daykin C, Navar-
enteropathogens and modulation of IL-8 production. Int J ange M, Beger RD, Verheij ER, Amberg A, Baunsgaard D, Cantor GH,
Food Microbiol 2008, 125:286-292. Lehman-McKeeman L, Earll M, Wold S, Johansson E, Haselden JN,
44. Roselli M, Finamore A, Nuccitelli S, Carnevali P, Brigidi P, Vitali B, Kramer K, Thomas C, Lindberg J, Schuppe-Koistinen I, Wilson ID,
Nobili F, Rami R, Garaguso I, Mengheri E: Prevention of TNBS- Reily MD, Robertson DG, Senn H, Krotzky A, Kochhar S, Powell J,
induced colitis by different Lactobacillus and Bifidobacterium Ouderaa F van der, Plumb R, Schaefer H, Spraul M: Summary rec-
strains is associated with an expansion of gammadeltaT and ommendations for standardization and reporting of meta-
regulatory T cells of intestinal intraepithelial lymphocytes. bolic analyses. Nat Biotechnol 2005, 23:833-838.
Inflamm Bowel Dis 2009, 15:1526-1536.
45. Saito Y, Sakamoto M, Takizawa S, Benno Y: Monitoring the cell
number and viability of Lactobacillus helveticus GCL1001 in

Page 13 of 13
(page number not for citation purposes)