Twin Study Indicates Loss of Interaction Between Microbiota and Mucosa of Patients With Ulcerative Colitis

Institute of Clinical Molecular Biology, Christian-Albrechts University-Kiel, Kiel, Germany.
Gastroenterology (Impact Factor: 16.72). 04/2011; 141(1):227-36. DOI: 10.1053/j.gastro.2011.04.011
Source: PubMed


Interactions between genetic and environmental factors are believed to be involved in onset and initiation of inflammatory bowel disease. We analyzed the interaction between gastrointestinal mucosal microbiota and host genes in twin pairs discordant for ulcerative colitis (UC) to study the functional interaction between microbiota and mucosal epithelium.
Biopsy were collected from sigmoid colon of UC patients and their healthy twins (discordant twin pairs) and from twins without UC. Microbiota profiles were determined from analysis of 16S ribosomal DNA libraries; messenger RNA profiles were determined by microarray analysis.
Patients with UC had dysbiotic microbiota, characterized by less bacterial diversity and more Actinobacteria and Proteobacteria than that of their healthy siblings; healthy siblings from discordant twins had more bacteria from the Lachnospiraceae and Ruminococcaceae families than twins who were both healthy. In twins who were both healthy, 34 mucosal transcripts correlated with bacterial genera, whereas only 25 and 11 correlated with bacteria genera in healthy individuals and their twins with UC, respectively. Transcripts related to oxidative and immune responses were differentially expressed between patients with UC and their healthy twins.
The transcriptional profile of the mucosa appears to interact with the colonic microbiota; this interaction appears to be lost in colon of patients with UC. Bacterial functions, such as butyrate production, might affect mucosal gene expression. Patients with UC had different gene expression profiles and lower levels of biodiversity than their healthy twins, as well as unusual aerobic bacteria. Patients with UC had lower percentages of potentially protective bacterial species than their healthy twins.


Available from: Limas Kupcinskas
Twin Study Indicates Loss of Interaction Between Microbiota and
Mucosa of Patients With Ulcerative Colitis
*Institute of Clinical Molecular Biology and
Department of Internal Medicine I, Christian-Albrechts University - Kiel, Kiel, Germany;
Jouy-en-Josas, France;
Institute for Digestive Research and
Department of Gastroenterology, Lithuanian University of Health Sciences, 44307 Kaunas, Lithuania;
Department of Internal Medicine, Asklepios Westklinikum Hamburg, Hamburg, Germany
BACKGROUND & AIMS: Interactions between genetic
and environmental factors are believed to be involved in
onset and initiation of inflammatory bowel disease. We
analyzed the interaction between gastrointestinal mucosal
microbiota and host genes in twin pairs discordant for
ulcerative colitis (UC) to study the functional interaction
between microbiota and mucosal epithelium. METH-
Biopsy were collected from sigmoid colon of UC patients
and their healthy twins (discordant twin pairs) and from twins
without UC. Microbiota profiles were determined from analysis
of 16S ribosomal DNA libraries; messenger RNA profiles were
determined by microarray analysis.
RESULTS: Patients with
UC had dysbiotic microbiota, characterized by less bacte-
rial diversity and more Actinobacteria and Proteobacteria
than that of their healthy siblings; healthy siblings from
discordant twins had more bacteria from the Lachno-
spiraceae and Ruminococcaceae families than twins who
were both healthy. In twins who were both healthy, 34
mucosal transcripts correlated with bacterial genera,
whereas only 25 and 11 correlated with bacteria genera in
healthy individuals and their twins with UC, respectively.
Transcripts related to oxidative and immune responses
were differentially expressed between patients with UC
and their healthy twins. CONCLUSIONS: The tran-
scriptional profile of the mucosa appears to interact
with the colonic microbiota; this interaction appears
to be lost in colon of patients with UC. Bacterial
functions, such as butyrate production, might affect
mucosal gene expression. Patients with UC had differ-
ent gene expression profiles and lower levels of biodi-
versity than their healthy twins, as well as unusual
aerobic bacteria. Patients with UC had lower percent-
ages of potentially protective bacterial species than
their healthy twins.
Keywords: Dysbiosis; Crohn Disease; Transcript; Inflam-
mation; Microbiome.
hronic inflammatory bowel diseases (IBDs), that is,
Crohn’s disease (CD) and ulcerative colitis (UC), are
characterized by dysbiosis of the intestinal microbiota.
Dysbiosis, an altered composition of the commensal bac-
terial populations, is discussed as a major factor in disease
pathogenesis that interacts with genetic susceptibility and
leads to the dysregulation of the immune response to
bacterial antigens observed in IBD.
Loss of natural
intestinal diversity and a shift of bacterial composition
toward a more deleterious profile might reflect the net
effect of environmental influences over the past decades
leading to the dramatic increase in the incidence of IBD in
the industrialized world.
Familial aggregation suggests a genetic predisposition
to IBD, with 5% to 20% of patients with IBD having a
family history of the disease.
However, a lower monozy-
gotic twin concordance rate for UC than for CD suggests
a smaller contribution of genetic factors in UC.
atic studies have given insights into the genetic architec-
ture of both IBDs. More than 40 disease genes and loci
have been identified and point to cytokine-driven immune
dysregulation (eg, the interleukin [IL]-23 pathway), innate
immunity, autophagy, and other factors important for
integrity of the intestinal epithelial cell.
Genes involved in the IL-23 pathway (IL-23R, IL-12B,
STAT3) and NKX2-3, DLG5 genes are associated with
both CD and UC. Recently, 2 single nucleotide polymor-
phisms on the multidrug resistance 1 gene (MDR1) have
been observed in association with UC.
Another single
nucleotide polymorphism on the same gene, rs3789243,
was found to be associated with pancolitis in patients
with UC.
Very recently, a risk locus on IL17REL was also
described in UC.
In contrast, only a few systematic analyses are available
that describe the gut microbiota dysbiosis in this disease.
Using fingerprinting methods, several groups observed a
decreased diversity in gut microbiota of patients with UC
compared with healthy controls at both mucosal
fecal levels.
Sokol et al applied in situ hybridization
coupled to flow cytometry to analyze fecal microbiota of
patients with IBD.
They highlighted a high percentage
of uncommon bacteria in patients with UC, with 40% of
total bacteria not detected by their probes designed
to target major groups of gut commensal microbiota.
Garrett et al showed a link between the expression
pattern of host genes and gut bacterial communities in a
Abbreviations used in this paper: FDR, false discovery rate; IL, inter-
leukin; PCA, principal component analysis; rDNA, ribosomal DNA.
© 2011 by the AGA Institute
GASTROENTEROLOGY 2011;141:227–236
Page 1
TRUC (Tbet
, Rag
, Ulcerative Colitis) mouse model
of colitis.
TRUC mice lack the T-bet transcription factor,
which is a negative regulator of the transcription of tumor
necrosis factor
. These TRUC mice spontaneously de-
velop a colitis that resembles UC. This colitis is vertically
transmissible to progeny through their intestinal micro-
biota. Moreover, they showed that healthy wild-type mice
that were fostered by TRUC mice developed colitis, which
strongly supports the hypothesis of the involvement of
environmental factors, such as the gut microbiota, in the
onset of UC.
Analysis of human twin pairs provides the unique op-
portunity to discriminate between the contribution of
genetic and environmental factors to phenotypic variance.
We have chosen UC because the phenotype is mimicked in
animal models that were investigated for the interaction
between epithelial genomics and gut microbiota and to
reach a higher homogeneity in the clinical phenotype. The
present study uses systematic molecular technologies to
describe gut dysbiosis in UC at both mucosal microbiota
and host level. It also provides insight into the genetic
determination of dysbiosis compared with healthy twins.
Patients and Methods
Biopsy specimens were sampled from the sigmoid
colon of 62 volunteers: 11 healthy dizygotic Lithuanian
twin pairs (Hli - aD/bD), 7 healthy monozygotic Lithua-
nian twin pairs (Hli - aM/bM), 8 monozygotic German
twin pairs discordant for UC (Hu/UC), and 10 healthy
unrelated German volunteers as a control cohort (Hger).
Clinical data of the UC discordant twins are shown in
Table 1. All twins recruited for this study were tested for
monozygosity or dizygosity as previously reported.
Female to male ratio was 1.3:1 (Supplementary Table 1).
Mucosal Microbiota Composition and
Diversity Assessment
Following total DNA extraction from biopsy
specimens, 16S ribosomal DNA (rDNA) amplification,
and cloning as previously described,
53 clone libraries
were sequenced (ABI PRISM; Applied Biosystems, Foster
City, CA) containing 144 clones each (except the pooled
German control cohort: 840 clones). After chimeras check
(Mallard software, http://www.bioinformatics-toolkit
.org/Mallard, Cardiff, UK), sequences were analyzed with
the RapidOTU pipeline
(CLUSTALW alignment, DNA-
dist program, DOTUR software,
.edu/micro/schloss/software/dotur.html). For phyloge-
netic affiliation, sequences were subjected to National
Center Biotechnology Information blast analyses and the
Seqmatch and Classifier programs at RDPII
Database Project, release 9.58). Bacterial phylotypes (or
operational taxonomic units) were defined as groups of
sequences sharing at least 98% of similarity and phyloge-
netically affiliated to their closest relative bacterial species.
The 2145 operational taxonomic unit representative se-
quences have been submitted to GenBank database under
the accession numbers HM805116 to HM807260.
Estimates of phylotypes richness and similarity indices
were calculated according to the bias-corrected Chao1
estimator and Sørensen similarity index, respectively. Phy-
lotypes richness is a measure of biodiversity and consists
of a count of the number of different species (or phylo-
types) in a given ecosystem. High species richness for a
given ecosystem means a high level of redundancy in its
function, which further denotes the ability of the ecosys-
tem to withstand natural disturbances. Principal compo-
nent and clustering analyses were performed to map each
individual’s microbiota based on their overall bacterial
species composition and to assess similarities between
individual’s microbiota and further define relevant clus-
ters, respectively (R packages ade4 and pvclust). Principal
component analyses with the different clinical factors as
instrumental variables (interclass principal component
analyses) were computed and statistically assessed by a
Monte Carlo rank test to observe the net effect of the
different factors on the scattering of the microbiota of
Table 1. Clinical Data for UC Discordant Twin Pairs
status Gender Age (y)
Status CAI
Duration of
the disease Location Treatment Surgery
Hu1 H M 52 NS na
UC1 UC M 52 NS 4 15 Pancolitis 5-ASA/Steroids 0
Hu3 H F 26 NS na
UC3 UC F 26 NS 2 5 Left-sided 5-ASA 0
Hu4 H M 42 S na
UC4 UC M 42 NS 3 14 None na 1
Hu5 H F 18 NS na
UC5 UC F 18 NS 3 3 Proctitis 5-ASA 0
Hu6 H F 21 NS na
UC6 UC F 21 NS 7 4 Pancolitis 5-ASA/Azathioprin 0
Hu7 H M 23 NS na
UC7 UC M 23 NS 7 4 Left-sided 5-ASA/Steroids 0
Hu8 H F 49 NS na
UC8 UC F 49 NS 6 6 Left-sided 5-ASA/Azathioprin 0
Hu9 H M 21 NS na
UC9 UC M 21 NS 7 5 Left-sided 5-ASA/Steroids 0
CAI, Colitis Activity Index; na, non applicable; Hu, Healthy sibling; UC, ulcerative colitis sibling; NS, Never smoked; S, Smoker; ASA, mesalamine.
Page 2
different individuals (Supplementary Materials and Meth-
Mucosal Eukaryotic Expression Profiles and
Correlations With Microbiota Profiles
Total RNA was extracted from biopsy specimens,
processed as previously described,
and hybridized to
Affymetrix UG 133 Plus 2.0 arrays (Santa Clara, CA). Data
were normalized using GCRMA (R, Bioconductor), and
signals that were not present in at least 80% of the sam-
ples were excluded from further analysis. The experimen-
tal and analytical part of the microarray analysis was
submitted according to MIAME standards to Gene Ex-
pression Omnibus (; se-
ries GSE7821 for healthy Lithuanian twins [submitted
earlier, samples GSM189751 to GSM189790] and
GSE22619 for discordant UC twin pairs). Differences be-
tween experimental groups (here: Hu vs UC) were assessed
using the Mann–Whitney U test, while P values were
corrected for multiple testing using the Benjamini–Hoch-
berg method.
Additionally, we performed a Westfall and
Young permutation to determine the false discovery rate
(FDR) of the fold changes (K 5000 permutations).
Genes with corrected P values .05 and an FDR 5% were
considered potentially disease relevant.
As a quantitative measure of host-microbiome cross
talk, we decided to assess the relationship between the
host gene expression and microbiota profiles. Statistical
dependence between these 2 variables was assessed using
Spearman rank correlations by comparing potentially dis-
ease-relevant transcripts and differentially represented
bacterial genera. Confounding effects of unequal group
sizes were avoided by selecting 7 individuals from each of
the 3 groups (unrelated healthy individuals, healthy indi-
viduals from discordant twin pairs, and diseased individ-
uals from discordant twin pairs). To control for false-
positive correlations, a Westfall and Young permutation
was used to determine the FDR using K 5000 permu-
tations for each correlation pair.
Correlations with an
FDR 5% were considered significant.
Heritability of the Human Gut Microbiota
Of the 8328 bacterial 16S rDNA sequences ob-
tained, 21% were chimeras or showed too low quality. A
total of 6554 sequences were further analyzed.
Microbiota composition similarity was significantly
higher in monozygotic than in dizygotic twins or unre-
lated individuals (Figure 1). Paired similarity indices
within dizygotic twin pairs (average Sørensen similarity
index Cs 0.12) were slightly higher than in unrelated
individuals (Cs 0.08). For monozygotic twins, gut mi-
crobiota Cs indices were similar whatever the clinical
status, healthy or UC discordant pairs (Cs 0.24), with a
higher variability in UC discordant pairs.
Total richness of Hu and UC microbial communities
was not significantly different (Figure 2). Chao1 estimated
a richness of 577 phylotypes within the Hu community
and 446 phylotypes in UC. Because the confidence inter-
vals overlap, one cannot reject the null hypothesis at the
significance level of .05 that there was no difference be-
tween the richness of Hu and UC communities. Interest-
ingly, unaffected siblings from UC discordant twin pairs
showed lower bacterial phylotype richness than did
healthy volunteers from both Germany (Chao1 estimates
of 1538 phylotypes) and Lithuania (Chao1 estimates of
1235 phylotypes). Because confidence intervals were not
overlapping, this observation can be considered as
strongly reliable.
Mucosal Microbiota Dysbiosis in UC
At a general level of dominant phyla repartition,
unaffected siblings from UC discordant twin pairs were
more similar to healthy individuals than to their siblings
(Figure 3). However, these unaffected siblings had signif-
icantly lower percentages of bacteria from the Proteobac-
teria phylum (P .045) than healthy individuals and
Figure 1. Sørensen similarity index based on qualitative analysis of
phylotypes. (A) Phylotypes were compared between individuals 2 by 2
(or between siblings from each twin pair) within each clinical group (ie,
aM01 vs bM01). Unrelated healthy individuals (n 22); Hli-DZ, healthy
dizygotic twins (n 11 pairs); Hli-MZ, healthy monozygotic twins (n 7
pairs); MZ-uc, monozygotic twins discordant for UC (n 8 pairs). P, P
value using an unpaired Student t test; P value in black, comparison with
Hli-DZ; P value in red, comparison with unrelated individuals. (B) Sø-
rensen similarity index based on qualitative analysis of phylotypes. Phy-
lotypes were pooled for half of the overall pairs for each clinical group and
compared with the other half (ie, aM vs bM).
Page 3
more Actinobacteria (P .013). Even though less repre-
sented in microbiota from patients with UC, distribution
of Firmicutes was not significantly different between un-
affected and affected siblings from UC discordant pairs
using a paired Student t test, whereas Bacteroidetes were
significantly different (P .046). The lower proportion of
Bacteroidetes in UC was mainly due to bacteria from the
Prevotellaceae family (average 12.7% of total clones in Hu
vs 0.7% in UC).
Disparate Profiles of UC Microbiota
Principal component analysis (PCA) and clustering
using Ward’s minimum variance method
of the phylo-
types repartition between unaffected (Hu) and patients
with UC from UC discordant twin pairs are presented in
Figure 4A and highlighted 3 different clusters. Cluster 1
had a probability value of 0.977 (standard error 0.014)
and grouped only patients with UC (UC4, UC5, UC6, and
UC8). Cluster 2 represented a second “UC cluster” (prob-
ability value 0.948; standard error 0.027), with UC1
and UC7 grouped together with an unaffected sibling
(Hu1) (Supplementary Figure 1). Lastly, cluster 3 repre-
sented mainly the unaffected siblings but also UC3 and
UC9. Impact of disease duration, CAI, age, sex, or location
of the disease on this clustering was further analyzed
(interclass correlation) and resulted in no significant cor-
relations. Only the treatment and mostly the use of cor-
ticosteroids or azathioprine in combination with mesala-
mine led to a similar clustering (Figure 4C). This effect
was significant using a Monte Carlo test (P .023), based
on 999 replicates.
Microbiota of patients with UC belonging to cluster 1
was characterized by high levels of Actinobacteria, mostly
Rhodococcus genus, and a lower proportion of both Bacte-
Figure 2. Estimated bacterial diversity in the 4 clinical groups using the Chao1 richness estimator. Chao1 is a nonparametric estimator for species
richness, taking into account the observed phylotypes number and the number of singletons and doubletons within the population. Light-colored
areas represent the 95% confidence intervals computed with DOTUR. H-ger, healthy German individuals (n 10); H-lith, healthy Lithuanian
individuals (n 10; randomly chosen samples to estimate phylotypes richness); Hu, healthy siblings from the UC discordant twin pairs (n 8); UC,
UC siblings from the UC discordant twin pairs (n 8).
Figure 3. Phyla repartition for each clinical group. Clinical groups are as
defined in the legend to Figure 2. Box plots represent 25th percentile,
median, and 75th percentile. Whiskers represent minimum and maxi-
mum values. Empty circles represent values of outliers. As pooled sam-
ples from 10 individuals were analyzed within the healthy German cohort,
only means were computed that are as follows: Firmicutes, 42.94%;
Bacteroidetes, 41.98%; Proteobacteria, 12.30%; and Actinobacteria,
Page 4
roides and Prevotella genera compared with their healthy
siblings. In our study, the main observed species was
related to Rhodococcus erythropolis (Table 2). It was the most
prevalent species, being present in all 8 patients with UC
at different percentages (0.9% to 65.69%), with a median
of 29.56%. This genus represented more than 50% of the
microbiota of patients with UC from cluster 1. Phylo-
types related to this species were also prevalent in
unaffected twins from UC discordant pairs, detected in
6 of 8 individuals but with a lower median value of
12.55% (0%–27.19%). Patients with UC from cluster 2
showed low percentages of Actinobacteria but high
proportions of Proteobacteria instead, mainly Entero-
bacteriaceae (Shigella/Escherichia). Microbiota of patients
in cluster 3 did not show any specific dysbiotic profile
but higher percentages of Ruminococcus obeum.
Potentially Protective Species in Unaffected
Twins From UC Discordant Twin Pairs
Several phylotypes from the Ruminococcaceae and
Lachnospiraceae families were overrepresented in unaf-
fected Hu twins compared with their UC siblings and the
healthy twins (Table 2). These phylotypes were also pres-
ent in patients with UC from cluster 3 at levels close to
their unaffected siblings. A phylotype related to butyrate-
producing bacterium A2-194 was observed in 3 of 8 Hu
individuals with an average percentage of 4.07%. It was
also detectable in one fourth of the healthy Hli twins
(1.11%). Butyrate-producing bacterium SS2/1 showed
similar percentages in Hu and Hli twins (2.96 and 2.67%,
respectively) but was detectable in half of the Hu twins
and only 22% of the Hli individuals. Bacterium mpn-
isolate group 18, a close relative of Blautia wexlerae, was
more frequent in Hu twins (50%) than in Hli twins (16.7%)
and represented also a higher percentage in Hu twins
(2.18%) than in Hli twins (0.95%).
Clustering of Eukaryotic Gene Expression
After excluding all transcripts that were not ex-
pressed in at least 80% of all samples, the remaining
21,747 of 54,675 transcripts were subjected to further
analysis, of which 63 were selected as potentially disease
relevant (present in at least 80% of the discordant samples,
Figure 4. PCA of microbiota phylotypes and host transcripts within UC discordant twin pairs. (A) PCA and clustering using Ward’s minimum variance
method with correlation distances analysis of the phylotypes repartition between Hu (white circles, healthy siblings) and UC (gray circles, UC siblings)
patients. These 2 components explain 49.73% of the variability. Statistical clustering using Ward’s minimum variance method is represented by ellipses on
the PCA. (B) PCA of the gene expression based on 361 differentially expressed genes between Hu and UC. Hu (white circles), healthy siblings from the UC
discordant twin pairs (n 8); UC (gray circles), UC-affected siblings from the UC discordant twin pairs (n 8). These 2 components explain 63.52% of the
variability. (C) Impact of medical treatment on phylotype composition; interclass analysis of discordant twin pairs. PCA with medical treatment as a clinical
factor was performed based on the phylotype abundance. Individuals (gray circles, UC siblings; white circles, unaffected siblings from UC discordant pairs)
were clustered (gray ellipses) and the center of gravity computed for each class. P value of the link between medical treatment and phylotype abundance
was assessed using a Monte Carlo test (999 replicates). P .023. AZA, azathioprine; 5-ASA, mesalamine.
Page 5
corrected P .05, FDR 5%). PCA analysis of eukaryotic
gene expression profiles of mucosa from both Hu and UC
twins resulted in a pattern of clusters similar to the
microbiota analysis (Figure 4B). Three of the 4 patients
with UC who clustered together when analyzing the mi-
crobiota (UC4, UC6, UC8) also clustered together when
computing the host expression profiles and together with
UC1 and UC7. Patient UC9 still clustered with the healthy
siblings. UC5, whose microbiota clustered within the UC
group, showed expression profiles close to the healthy
siblings, whereas UC3 was intermediate.
Two transcripts were strongly overexpressed in patients
with UC as compared with their healthy siblings (Supple-
mentary Table 2): Olfactomedin 4 (OLFM4, also called
hGC-1 or GW112; 58-fold; P .004) and regenerating
islet-derived family, member 4 (REG4; 29-fold; P .041).
Moreover, an important set of genes involved in extra-
cellular matrix was significantly overexpressed in pa-
tients with UC: collagen types 1, 4, 6, 12; fibronectin;
matrix metallopeptidase 12; tenascin C; and actin
gamma 2. On the other hand, metallothioneins, crea-
tine kinase, brain (CKB), 3-hydroxy-3-methylglutaryl-
coenzyme A synthase 2 (mitochondrial, HMGCS2), ace-
toacetyl CoA thiolase (ACAT1), and resistin-like beta
(RETNLB) were significantly down-regulated in UC
compared with Hu.
Table 2. Differential Representation of Phylotypes (OTUs) for Each of the Discordant Twin Pairs as Classified in Clusters
Closest relative of the OTU Sim.
UC discordant twin pairs
pair 4 pair 5 pair 6 pair 8
Over represented in UC
In UC Cluster 1
Rhodococcus erythropolis 1 44.83 38.64 42.67 61.76
Bradyrhizobium japonicum 1 0 1.90 0.28 3.92
Stenotrophomonas maltophilia 1 2.13 2.56 1.94 3.92
Stenotrophomonas maltophilia 0.99 1.06 0 1.38 0.98
Variovorax sp. 1 0 0.95 1.10 0.98
Ralstonia detusculanense 0.97 1.06 0.95 0 0.98
pair 1 pair 7
In UC Cluster 2
Escherichia coli 1 27.58 34.19
Bacteroides vulgatus 1 12.50 3.62
Ralstonia pickettii 0.94 3.60 3.42
Rhodococcus erythropolis 0.99 0.90 4.06
Clostridium lactatifermentans 0.90 3.60 0.85
Clostridium leptum 0.81 0.90 1.71
Eubacterium plautii 0.78 0.90 0.85
Bacteroides dorei 0.97 0.90 0.85
pair 3 pair 9
In UC Cluster 3
Ruminococcus obeum 1 5.48 6.56
Under represented in UC pair 4 pair 5 pair 6 pair 8
In UC Cluster 1
Bacteroides vulgatus 1 7.37 7.69 1.64 1.80
butyrate-producing bacterium A2-194 0.98 10.53 0 0.82 0.90
butyrate-producing bacterium SS2/1 0.98 2.11 2.56 0 2.46
Dorea longicatena 0.94 1.05 1.28 0.82 3.60
Ruminococcus obeum 0.98 1.05 2.56 0 2.70
bacterium mpn-isolate group 18 0.98 0 2.56 0.82 1.80
pair 1 pair 7
In UC Cluster 2
Ruminococcus torques 0.93 1.83 5.45
Bacteroides caccae 1 0.92 0.90
pair 3 pair 9
In UC Cluster 3
Rhodococcus erythropolis 1 23.19 10.88
Bilophila wadsworthia 0.95 1.75 0.85
Clusters are as defined in Figure 4A and Supplementary Figure 1 based on the statistical ward’s minimum variance method.
Data is presented as delta of percentage of over- or under-representation in UC.
Sim, Hit score similarity of the phylotype with its closest relative isolate as assessed by RDP seqmatch.
Page 6
Loss of Relevant Bacteria-Host Cross Talk in
Patients With UC and Healthy Siblings
Sharing Identical Genetic Background
On the overall 54,675 transcripts and 134 bacterial
genera, a correlation analysis was performed on a subset
of 63 potentially disease-relevant transcripts and 34 po-
tentially disease-relevant bacterial genera, aiming to deter-
mine bacteria-host cross talk in the context of UC. As a
result, we identified loss of cross talk as a prominent
pattern; the presented 36 host genes were significantly
correlated with 20 bacterial genera within the twins (Fig-
ure 5). Most of the transcripts were functionally associ-
ated with cell adhesion/proliferation/differentiation,
signal transduction, immune/inflammatory response,
transcription regulation, metal ion binding, and oxida-
tion reduction. A significantly lower number of correla-
tions were observed in affected (UC, 11 transcripts and 5
genera) and unaffected twins (Hu, 25 transcripts and 12
genera) from UC discordant pairs than in healthy Lithu-
anian twins (Hli, 34 transcripts and 20 genera) (FDR
5%). The individual correlations are presented in Sup-
plementary Table 3.
Even though genetics, immune response, and mi-
crobiota are currently described as the 3 main factors
involved in the development and maintenance of IBD,
description of the way they interact with each other re-
mains challenging. In the present study, we were able to
show for the first time correlations between microbiota
composition and host gene expression based on in vivo
data. The number of correlations was strongly decreased
at UC patients’ mucosal interface (10 host transcript/
bacterial genera pairs, compared with 95 transcript/bac-
terial genera pairs in healthy individuals, Hli), which may
indicate a loss of cross talk between host and microbiota
Figure 5. Correlations between microbiota and gene expression. Transcripts (second column) are organized as Gene Ontology terms for biological
processes (first column). Transcripts with no gene symbol are listed with their public reference; a star indicates that this transcript represents the same
gene as the previous transcript. Bacterial genera (second row) are classified by phyla (first row). Hli, healthy unrelated individuals from the
monozygotic cohort; Hu, healthy individuals from discordant twin pairs (UC); UC, diseased individuals from discordant twin pairs (UC); F, Firmicutes;
B, Bacteroidetes; P, Proteobacteria; A, Actinobacteria. Blue squares represent positive correlations (r 0.45), and yellow squares represent negative
correlations (r ⬍⫺0.45).
Page 7
in the context of colitis. More surprisingly, even though a
large number of the correlations observed at the mucosal
interface in unaffected siblings from UC discordant pairs
were also detected in healthy Lithuanians, their number
was gradually decreased to the level of the affected sib-
lings. This key observation is in line with a study from
Bodger et al where they assessed the mucosal expression
of the oncofetal carbohydrate antigen TF and nuclear
B by immunohistochemistry in unaffected mo-
nozygotic twins of patients with IBD.
They showed that
activated nuclear factor
B was present in the surface
epithelium of mucosal biopsy specimens from 13 of 14
unaffected IBD twins but in only 6 of 22 histologically
normal controls. This occurred in both UC and CD twins.
The investigators hypothesized that epithelial nuclear fac-
B activation in the unaffected twins may occur as a
result of interaction between the epithelium and intralu-
minal component(s). They further postulated that some
additional trigger would still be required to induce signif-
icant inflammation.
In the present work, differentially overrepresented bac-
teria in UC patients’ microbiota related mostly to poten-
tially pathogenic aerobic genera (Rhodococcus,
Escherichia, and Stenotrophomonas
). Together with the
increased level of transcripts related to the extracellular
matrix in UC, this would suggest that presence of these
bacteria might be due to a defect in the barrier function of
the epithelium in UC. Indeed, in the colonic mucosa of
patients with UC, tenascin C was shown to indicate tissue
repair and mucosal concentrations were correlated with
local disease activity.
Matrix metalloproteinase levels are
increased in inflamed mucosa, and matrix metalloprotei-
nases are known to degrade most of the macromolecules
making up the extracellular matrix. Matrix metalloprotei-
nases are also increased in Helicobacter pylori–induced gastric
Our experimental setup does not allow determin-
ing whether the presence of these specific bacterial species is
a cause or a consequence of the barrier defect but indicates
a potential functional link. In this scenario, these bacteria
might, in turn, contribute to the establishment of a vicious
circle sustaining the inflammatory process.
Our work nonetheless strongly suggests that some bac-
terial species or components, rather than representing a
triggering factor in patients with UC, might protect the
unaffected siblings from chronic inflammation. Several
butyrate-producing bacteria were especially more abun-
dant and more prevalent in Hu than in UC but also than
in healthy individuals and butyrate has been shown to
inhibit inflammatory response.
Moreover, defective co-
lonic epithelial oxidation of butyrate in UC has been
implicated in its pathogenesis.
One of the main enzymes
involved in this oxidation is the mitochondrial acetoacetyl
CoA thiolase (ACAT1), with activity decreased in the co-
lonic mucosa of patients with UC
and also decreased in
patients with UC in the present study. In addition, pa-
tients with UC had lower expression levels of 3-hydroxy-
3methylglutaryl CoA Synthase 2 (mHMGCS2), a key con-
trol site of ketogenesis from butyrate, which has also been
described as lower in germ-free than in conventional
These data suggest that the intestinal microbiota,
through butyrate production, could control the expres-
sion of colonic mHMGCoA synthase.
Another bacterial species detected as more abundant in
unaffected siblings as compared with patients with UC,
namely Faecalibacterium prausnitzii, has already been de-
scribed as exhibiting anti-inflammatory properties
being less present in colitis microbiota.
In the present
study, abundance of F prausnitzii was correlated with de-
creased mucosal expression of MAP3K8 (also called Tpl2/
Cot) in unaffected siblings from UC discordant pairs.
MAP3K8 is a serine-threonine kinase critical in innate
immunity, linking Toll-like receptors to tumor necrosis
factor production through its activation of ERK in re-
sponse to bacterial lipopolysaccharide or interleukin-1
ligation. STAT4 directly binds to MAP3K8 and, in a pos-
itive feedback loop, MAP3K8 promotes the expression of
T-bet and STAT4 itself.
Moreover, pharmacologic in-
hibitors of MAP3K8 efficiently block inflammatory re-
sponses in human monocytes.
OLFM4 and REG4 transcripts, strongly up-regulated in
the mucosa from patients with UC, have been recently
described as highly sensitive biomarkers for gastric can-
OLFM4 was also observed in colon cancers in early
and was up-regulated in crypt epithelium of in-
flamed colonic mucosa
together with Reg1
. Chronic
inflammation in patients with UC is known to increase
the prevalence of colorectal cancer.
However, consider-
ing that none of the patients included in this study had
developed colon cancer since sampling and that these
genes were overexpressed in all patients, they could indi-
cate events at an early stage, where chronic inflammation
has not yet promoted tumor development.
On the contrary, metallothionein gene expression was
significantly reduced in patients with UC as compared
with their unaffected siblings, which was previously ob-
served in intestinal inflammation.
These proteins are
induced by a large variety of stimuli (eg, oxidative stress,
glucocorticoids) and provide protection against oxidative
stress. Metallothionein messenger RNA expression was
also found to be down-regulated in azathioprine-treated
CaCo2 cells, suggesting that azathioprine might down-
regulate metallothionein gene expression.
However, in
the present work, patients with UC being treated with
azathioprine did not show lower expression of metallo-
thionein than patients being treated with corticosteroids,
while our observations indicate a potentially protective
anti-inflammatory role of metallothionein in the mucosa
of unaffected siblings from UC discordant pairs, which
was recently discussed by Inoue et al.
Similar to any study in a clinical setting monitoring
disease-associated processes in complex tissue samples,
the presented result is exposed to various sources of vari-
ation. Changes in cellular composition as well as altered
bacterial colonization are likely to be one major source of
variation. However, besides these limitations, primary tis-
sue offers the unique opportunity to monitor molecular
Page 8
mechanisms of UC pathophysiology while reflecting the
interaction of various cell types.
Finally, the genetic background seemed to strongly in-
fluence the human gut microbiota composition and di-
versity. Microbiota of monozygotic twins was much more
similar than the one of unrelated individuals. Moreover,
the observation that the microbiota of dizygotic twins was
less similar than of monozygotic twins, even though
raised together, suggests a major impact of genetics when
early environmental factors are similar. Surprisingly, the
similarity of microbiota was still high in monozygotic
twins discordant for UC. Regardless of the direction of
influence between the host transcriptome and the micro-
biome, these similarities within twin pairs support the
hypothesis of heritable elements, which have been shown
for transcriptional profiles of twins.
Nonetheless, the
main objective of the presented study is to identify altered
host-microbiome cross talk, while the analytical setup
does not emphasize zygosity effects. However, the micro-
biota of unaffected siblings from UC twin pairs exhibited
higher percentages of potentially protective bacteria,
which might play a protective role in a deleterious con-
text. Bacterial diversity was decreased in patients with UC
as compared with healthy individuals, which has already
been observed in CD.
Strikingly, unaffected siblings
from UC discordant pairs also showed a lower bacterial
biodiversity than healthy individuals, further supporting
the heritability concept. The microbiota dysbiosis phe-
nomenon described in CD and to a lesser extent in UC
was also observed in this study. An important increase in
potentially pathogenic aerobic bacteria was observed in
the gut microbiota of patients with UC, which could
account at least in part for the 40% of bacteria that were
not targeted by using usual gut commensal bacteria
probes in the study by Sokol et al.
In conclusion, the presented data document for the
first time how the interaction between the gut microbiota
and its host is altered in the context of UC. Together with
a reduced complexity in the mucosal microbiota, the ob-
served loss of cross talk between host gene expression and
bacterial profiles indicates that these key elements, which
can influence disease manifestation and progression,
should not be treated as independent events. Moreover,
our twin study suggests that this interaction is partially
genetically controlled. Further research to understand the
communication between host genetics, host transcrip-
tomics, and gut microbiota, representing 3 major compo-
nents in UC, may finally lead to the development of novel
diagnostic and therapeutic concepts.
Supplementary Material
Note: To access the supplementary material
accompanying this article, visit the online version of
Gastroenterology at, and at doi:
1. Ott SJ, Musfeldt M, Wenderoth DF, et al. Reduction in diversity of
the colonic mucosa associated bacterial microflora in patients
with active inflammatory bowel disease. Gut 2004;53:685– 693.
2. Seksik P, Lepage P, de la Cochetiere MF, et al. Search for local-
ized dysbiosis in Crohn’s disease ulcerations by temporal temper-
ature gradient gel electrophoresis of 16S rRNA. J Clin Microbiol
2005;43:4654 4658.
3. Sokol H, Seksik P, Rigottier-Gois L, et al. Specificities of the fecal
microbiota in inflammatory bowel disease. Inflamm Bowel Dis
2006;12:106 –111.
4. Tamboli CP, Neut C, Desreumaux P, et al. Dysbiosis as a prereq-
uisite for IBD. Gut 2004;53:1057.
5. Colombel JF, Vernier-Massouille G, Cortot A, et al. Epidemiology
and risk factors of inflammatory bowel diseases. Bull Acad Natl
Med 2007;191:1105–1118; discussion 1118 –1123.
6. Colombel JF, Grandbastien B, Gower-Rousseau C, et al. Clinical
characteristics of Crohn’s disease in 72 families. Gastroenterol-
ogy 1996;111:604 607.
7. Spehlmann ME, Begun AZ, Burghardt J, et al. Epidemiology of
inflammatory bowel disease in a German twin cohort: results of a
nationwide study. Inflamm Bowel Dis 2008;14:968 –976.
8. Craddock N, Hurles ME, Cardin N, et al. Genome-wide association
study of CNVs in 16,000 cases of eight common diseases and
3,000 shared controls. Nature 2010;464:713–720.
9. Franke A, Balschun T, Karlsen TH, et al. Replication of signals
from recent studies of Crohn’s disease identifies previously un-
known disease loci for ulcerative colitis. Nat Genet 2008;40:713–
10. Franke A, Balschun T, Karlsen TH, et al. Sequence variants in
IL10, ARPC2 and multiple other loci contribute to ulcerative colitis
susceptibility. Nat Genet 2008;40:1319 –1323.
11. Franke A, Balschun T, Sina C, et al. Genome-wide association
study for ulcerative colitis identifies risk loci at 7q22 and 22q13
(IL17REL). Nat Genet 2010;42:292–294.
12. Imielinski M, Baldassano RN, Griffiths A, et al. Common variants
at five new loci associated with early-onset inflammatory bowel
disease. Nat Genet 2009;41:1335–1340.
13. Kaser A, Lee AH, Franke A, et al. XBP1 links ER stress to intestinal
inflammation and confers genetic risk for human inflammatory
bowel disease. Cell 2008;134:743–756.
14. Annese V, Valvano MR, Palmieri O, et al. Multidrug resistance 1
gene in inflammatory bowel disease: a meta-analysis. World J
Gastroenterol 2006;12:3636 –3644.
15. Huebner C, Browning BL, Petermann I, et al. Genetic analysis of
MDR1 and inflammatory bowel disease reveals protective effect of
heterozygous variants for ulcerative colitis. Inflamm Bowel Dis
2009;15:1784 –1793.
16. Nishikawa J, Kudo T, Sakata S, et al. Diversity of mucosa-associ-
ated microbiota in active and inactive ulcerative colitis. Scand J
Gastroenterol 2009;44:180 –186.
17. Ott SJ, Plamondon S, Hart A, et al. Dynamics of the mucosa-
associated flora in ulcerative colitis patients during remission and
clinical relapse. J Clin Microbiol 2008;46:3510 –3513.
18. Martinez C, Antolin M, Santos J, et al. Unstable composition of the
fecal microbiota in ulcerative colitis during clinical remission. Am J
Gastroenterol 2008;103:643– 648.
19. Garrett WS, Lord GM, Punit S, et al. Communicable ulcerative
colitis induced by T-bet deficiency in the innate immune system.
Cell 2007;131:33– 45.
20. Barbaro A, Cormaci P. DNA analysis from mixed biological mate-
rials. Forensic Sci Int 2004;146(Suppl):S123–S125.
21. von Wurmb-Schwark N, Malyusz V, Simeoni E, et al. Possible
pitfalls in motherless paternity analysis with related putative fa-
thers. Forensic Sci Int 2006;159:92–97.
22. Tap J, Mondot S, Levenez F, et al. Towards the human intestinal
microbiota phylogenetic core. Environ Microbiol 2009;11:2574
Page 9
23. Cole JR, Wang Q, Cardenas E, et al. The Ribosomal Database
Project: improved alignments and new tools for rRNA analysis.
Nucleic Acids Res 2009;37:D141–145.
24. Mah N, Thelin A, Lu T, et al. A comparison of oligonucleotide and
cDNA-based microarray systems. Physiol Genomics 2004;16:
25. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a
practical and powerful approach to multiple testing. J Roy Statist
Soc Ser B 1995;57:289 –300.
26. Westfall PH, Young S. Resampling-based multiple testing. Wiley
series in probability and statistics. 1st ed. New York, NY: Wiley-
Interscience, 1993.
27. Ward RF, Werner SL. Analysis of variance of the composition of a
migmatite. Science 1963;140:978 –979.
28. Bodger K, Halfvarson J, Dodson AR, et al. Altered colonic glyco-
protein expression in unaffected monozygotic twins of inflamma-
tory bowel disease patients. Gut 2006;55:973–977.
29. Baba H, Nada T, Ohkusu K, et al. First case of bloodstream
infection caused by Rhodococcus erythropolis. J Clin Microbiol
30. Roy M, Sidhom S, Kerr KG, et al. Case report: Rhodococcus
erythropolis osteomyelitis in the toe. Clin Orthop Relat Res 2009;
467:3029 –3031.
31. Hellmig S, Ott S, Musfeldt M, et al. Life-threatening chronic enter-
itis due to colonization of the small bowel with Stenotrophomonas
maltophilia. Gastroenterology 2005;129:706 –712.
32. Falagas ME, Kastoris AC, Vouloumanou EK, et al. Attributable
mortality of Stenotrophomonas maltophilia infections: a system-
atic review of the literature. Future Microbiol 2009;4:1103–1109.
33. Riedl S, Tandara A, Reinshagen M, et al. Serum tenascin-C is an
indicator of inflammatory bowel disease activity. Int J Colorectal
Dis 2001;16:285–291.
34. Hellmig S, Ott S, Rosenstiel P, et al. Genetic variants in matrix
metalloproteinase genes are associated with development of gastric
ulcer in H. pylori infection. Am J Gastroenterol 2006;101:29 –35.
35. Bergin PJ, Anders E, Sicheng W, et al. Increased production of
matrix metalloproteinases in Helicobacter pylori-associated hu-
man gastritis. Helicobacter 2004;9:201–210.
36. Segain JP, Raingeard de la Bletiere D, Bourreille A, et al. Butyrate
inhibits inflammatory responses through NFkappaB inhibition: im-
plications for Crohn’s disease. Gut 2000;47:397– 403.
37. De Preter V, Bulteel V, Suenaert P, et al. Pouchitis, similar to active
ulcerative colitis, is associated with impaired butyrate oxidation by
intestinal mucosa. Inflamm Bowel Dis 2009;15:335–340.
38. Santhanam S, Venkatraman A, Ramakrishna BS. Impairment of
mitochondrial acetoacetyl CoA thiolase activity in the colonic mu-
cosa of patients with ulcerative colitis. Gut 2007;56:1543–1549.
39. Cherbuy C, Darcy-Vrillon B, Morel MT, et al. Effect of germfree
state on the capacities of isolated rat colonocytes to metabolize
n-butyrate, glucose, and glutamine. Gastroenterology 1995;109:
1890 –1899.
40. Cherbuy C, Andrieux C, Honvo-Houeto E, et al. Expression of
mitochondrial HMGCoA synthase and glutaminase in the colonic
mucosa is modulated by bacterial species. Eur J Biochem 2004;
41. Sokol H, Pigneur B, Watterlot L, et al. Faecalibacterium prausnitzii
is an anti-inflammatory commensal bacterium identified by gut
microbiota analysis of Crohn disease patients. Proc Natl Acad Sci
U S A 2008;105:16731–16736.
42. Sokol H, Seksik P, Furet JP, et al. Low counts of Faecalibacterium
prausnitzii in colitis microbiota. Inflamm Bowel Dis 2009;15:
43. Watford WT, Hissong BD, Durant LR, et al. Tpl2 kinase regulates
T cell interferon-gamma production and host resistance to Toxo-
plasma gondii. J Exp Med 2008;205:2803–2812.
44. Symons A, Beinke S, Ley SC. MAP kinase kinase kinases and
innate immunity. Trends Immunol 2006;27:40 48.
45. Hall JP, Kurdi Y, Hsu S, et al. Pharmacologic inhibition of tpl2
blocks inflammatory responses in primary human monocytes, sy-
noviocytes, and blood. J Biol Chem 2007;282:33295–33304.
46. Oue N, Sentani K, Noguchi T, et al. Serum olfactomedin 4
(GW112, hGC-1) in combination with Reg IV is a highly sensitive
biomarker for gastric cancer patients. Int J Cancer 2009;125:
47. Koshida S, Kobayashi D, Moriai R, et al. Specific overexpression
of OLFM4(GW112/HGC-1) mRNA in colon, breast and lung cancer
tissues detected using quantitative analysis. Cancer Sci 2007;
48. Shinozaki S, Nakamura T, Iimura M, et al. Upregulation of Reg
1alpha and GW112 in the epithelium of inflamed colonic mucosa.
Gut 2001;48:623– 629.
49. Munkholm P. Review article: the incidence and prevalence of
colorectal cancer in inflammatory bowel disease. Aliment Pharma-
col Ther 2003;18 Suppl 2:1–5.
50. Kruidenier L, Kuiper I, Van Duijn W, et al. Imbalanced secondary
mucosal antioxidant response in inflammatory bowel disease.
J Pathol 2003;201:17–27.
51. Dooley TP, Curto EV, Reddy SP, et al. Regulation of gene expres-
sion in inflammatory bowel disease and correlation with IBD
drugs: screening by DNA microarrays. Inflamm Bowel Dis 2004;
52. Inoue K, Takano H, Shimada A, et al. Metallothionein as an
anti-inflammatory mediator. Mediators Inflamm 2009;2009:
53. Costello CM, Mah N, Hasler R, et al. Dissection of the inflamma-
tory bowel disease transcriptome using genome-wide cDNA mi-
croarrays. PLoS Med 2005;2:e199.
54. Hasler R, Begun A, Freitag-Wolf S, et al. Genetic control of global
gene expression levels in the intestinal mucosa: a human twin
study. Physiol Genomics 2009;38:73–79.
55. Manichanh C, Rigottier-Gois L, Bonnaud E, et al. Reduced diversity
of faecal microbiota in Crohn’s disease revealed by a metag-
enomic approach. Gut 2006;55:205–211.
Received September 13, 2010. Accepted April 8, 2011.
Reprint requests
Address requests for reprints to: Stefan Schreiber, MD, PhD,
Institute for Clinical Molecular Biology and Clinic for General Internal
Medicine, I Medical Department, University Hospital Schleswig-
Holstein, Campus Kiel, Schittenhelmstr 12, 24105 Kiel, Germany.
e-mail: s.schreiber@; fax: (49) 0 431 597-1842.
The authors thank Dorina Oelsner, Tanja Kaaksteen, Ulrike
Panknin, Meike Barche, and Fabienne Béguet for technical
assistance; Marion Leclerc, Stanislas Mondot, and Julien Tap for
helpful discussion; and Nicole von Wurmb-Schwark for performing
the zygosity testing of the twins.
Microarray data were submitted according to MIAME standards to
Gene Expression Omnibus series GSE7821 for healthy Lithuanian
twins (GSM189751 to GSM189790) and GSE22619 for discordant
UC twin pairs.
Representative bacterial sequences were submitted to GenBank
database under accession numbers HM805116 to HM807260.
P.L., R.H., and M.E.S. contributed equally to this study.
Conflicts of interest
The authors disclose no conflicts.
Supported by the National German Genome Network, the SFB 415
Project Z1, the Excellence Cluster Inflammation at Interfaces, and
the MFG Educative Science, Hamburg, Germany.
Page 10
  • Source
    • "There is a growing belief that a healthy gut microfloral population can positively influence immune responses such that the individual might be protected from the development of inflammatory pathologies such as ulcerative colitis, arthritis or even asthma, with a growing number of studies linking general wellbeing and healthy gut function with highly diverse intestinal microbiota and intake of fibre in the diet [1,[91][92][93]. Bacteria belonging to the Bacteroidetes and Firmicutes phyla have been found to produce high levels of the SCFAs C2–C4 which are the main agonists for FFAR2. "
    [Show abstract] [Hide abstract] ABSTRACT: There are significant numbers of nutrient sensing G protein-coupled receptors (GPCRs) that can be found in cells of the immune system and in tissues that are involved in metabolic function, such as the pancreas or the intestinal epithelium. The family of free fatty acid receptors (FFAR1-4, GPR84), plus a few other metabolite sensing receptors (GPR109A, GPR91, GPR35) have been for this reason the focus of studies linking the effects of nutrients with immunological responses. A number of the beneficial anti-inflammatory effects credited to dietary fats such as omega-3 fatty acids are attributed to their actions on FFAR4.This might play an important protective role in the development of obesity, insulin resistance or asthma. The role of the short-chain fatty acids resulting from fermentation of fibre by the intestinal microbiota in regulating acute inflammatory responses is also discussed. Finally we assess the therapeutic potential of this family of receptors to treat pathologies where inflammation is a major factor such as type 2 diabetes, whether by the use of novel synthetic molecules or by the modulation of the individual’s diet.
    Full-text · Article · Mar 2016 · Biochemical Pharmacology
  • Source
    • "Similarly, the role of gut bacteria in IBD development is well documented in animal models of IBD [122,123] . While no specific bacteria have been identified as a causative pathogen for IBD, general dysbiosis [119,124125126 as well as an overall decrease in microbial diversity [22,127] has been observed in IBD patients compared to healthy controls. In addition, UC patients had lower microbial diversity during active disease periods compared to periods of disease remission [47,127] , demonstrating the fluidity of the microbial composition even within the same individual. "
    Preview · Article · Jan 2016 · World Journal of Gastroenterology
  • Source
    • "tion. For example , shifts in diet can produce rapid and pronounced alterations in the microbiota ( David et al. , 2014a , b ). In addition to diet , stressors such as antibiotic treatment , enteric infection , and colitis can induce remodeling of the gut microbiota ( Garrett et al . , 2007 ; Stecher et al . , 2010 ; Dethlefsen and Relman , 2011 ; Lepage et al . , 2011 ) . The insensitivity of a community to disturbance is referred to as its stability . Stability can be measured both in terms of resistance , the degree to which a community is unresponsive to a change in conditions , as well as the resilience , the tendency for a community to return to its original state following a disturbance ( Shade"
    [Show abstract] [Hide abstract] ABSTRACT: Acute colitis causes alterations in the intestinal microbiota, but the microbiota is thought to recover after such events. Extreme microbiota alterations are characteristic of human chronic inflammatory bowel diseases, although alterations reported in different studies are divergent and sometimes even contradictory. To better understand the impact of periodic disturbances on the intestinal microbiota and its compositional difference between acute and relapsing colitis, we investigated the beginnings of recurrent inflammation using the dextran sodium sulfate (DSS) mouse model of chemically induced colitis. Using bacterial 16S rRNA gene-targeted pyrosequencing as well as quantitative fluorescence in situ hybridization, we profiled the intestinal and stool microbiota of mice over the course of three rounds of DSS-induced colitis and recovery. We found that characteristic inflammation-associated microbiota could be detected in recovery-phase mice. Successive inflammation episodes further drove the microbiota into an increasingly altered composition post-inflammation, and signatures of colitis history were detectable in the microbiota more sensitively than by pathology analysis. Bacterial indicators of murine colitis history were identified in intestinal and stool samples, with a high degree of consistency between both sample types. Stool may therefore be a promising non-invasive source of bacterial biomarkers that are highly sensitive to inflammation state and history.
    Full-text · Article · Dec 2015 · Frontiers in Microbiology
Show more