Background
Microbes in the body are tenfold more than the number of human cells, and encode 100 times more unique genes than human genomes. Intestinal microbiota are in close contact with mucosal immune system, affecting host physiology and maintaining homeostasis [
1‐
3]. The incidence of Crohn’s disease (CD) has been reported to be increasing around the world, and while it is responsible for inducing chronic intestinal inflammation in a genetically susceptible host, the underlying mechanism is thought to result from an inappropriate and ongoing activation of the mucosal immune system driven by several stimuli such as intestinal microbiota and various environmental factors [
4].
CD is relatively more prevalent in western population, and numerous genetic susceptibility loci have been identified [
5‐
7]. However, well-known CD susceptible genes such as
NOD2 and
ATG16L1 did not show replication results in Asian population, suggesting a different genetic background [
8,
9]. Although the genetic susceptibility loci differ between Asian and western population, the incidence of CD in Asian population is increasing with similar immunologic phenomena [
10]. Therefore, it can be presumed that environmental factors, especially intestinal commensal bacteria, besides genetic factors, play a fundamental role in the development of CD. In fact, it is well known that the intestinal microbial community of western CD patients demonstrates dysbiosis different from healthy population [
11,
12]. However, there is no published study demonstrating intestinal microbial profiles of Korean CD patients using high throughput sequencing methods.
In the present study, we examined and compared the fecal and mucosal microbial community of Korean CD patients and healthy controls (HC) by applying a next-generation sequencing method after isolation of microbial DNA.
Discussion
Our deep sequencing study showed that the intestinal microbial community in Korean CD patients differed from that of HC. The dysbiosis patterns in Korean CD patients, including decreased microbial diversity, increased relative abundances of Proteobacteria and Fusobacteria and decreased proportions of Firmicutes and Bacteroidetes, were similar to those of western CD patients. In addition, the intestinal microbial community structure of mucosal tissues samples differed from that of fecal samples. Finally, we demonstrated that the intestinal microbial community in CD patients was affected by clinical disease activity, disease location and behavior, and therapeutic medication such as anti-TNF agent.
Intestinal microbiota maintain the homeostasis of our body by keeping close contact with the host [
18]. CD affects intestinal commensals in genetically susceptible hosts through various environmental factors, and excessive stimuli of the immune system ultimately leads to full-blown disease [
19]. Compared with western population, Korean CD patients have distinct epidemiologic characteristics such as male predominance, frequent peri-anal lesions, and distinctive ileo-colonic manifestations, and the susceptibility genes commonly seen in western countries are rarely detected in Korean CD patients [
9,
20]. However, since
NOD2 and
ATG16L1 mutations, which are closely related with bacterial handling, are observed in western CD population [
5], and although not the same mutation,
ATG16L2 mutation related with regulation of bacteria-related signal was observed in Korean CD patients [
9], it can be presumed that intestinal microbiota play an important role in the development of CD [
21].
In the present study, Korean CD patients demonstrated alterations of intestinal microbial community structure differed from that of HC. The microbial diversity was decreased and the relative proportions of
Proteobacteria and
Fusobacteria were increased with concomitant decrease of relative abundances of
Firmicutes and
Bacteroidetes in CD patients compared to HC, especially in stool samples. These results are consistent with those of previous studies from western population [
22,
23], suggesting an essential role of the intestinal microbial community in the development of CD throughout the world.
It has been well known that the relative proportions of
Proteobacteria including adherent invasive
E. coli and other numerous bacteria are increased in the mucosa of inflammatory bowel diseases (IBD) patients [
24,
25]. In this study,
Gammaproteobacteria class accounted for most of increase in the proportion of
Proteobacteria phylum. And,
Escherichia/Shigella genus were predominantly observed in both mucosal tissues and fecal samples of CD patients compared to HC, agreeing that
E. coli strains play an important role in inducing chronic inflammation of intestinal tract.
On the other hand, the relative abundance of
Clostridia class of
Firmicutes phylum were decreased in both fecal and mucosal tissue samples of CD patients compared to HC.
Firmicutes has been known to synthesize important short chain fatty acids, mainly acetate, propionate and butyrate, via fermentation of ingested dietary fiber within the intestine, which have profound effects on gut health as an energy source and inflammatory modulator [
26]. Since intestinal microorganisms are closely related to food [
11], it is interesting to know that decrease of relative proportion of
Firmicutes was consistently observed in both Korean and western CD patients, even though the diet pattern of Korean patients is different from that of western counterpart.
Clostridia class has also been reported to induce colonic regulatory T cells, which have a central role in the suppression of intestinal inflammation [
27]. In addition, we showed that the relative proportions of
Faecalibacterium and
Lachnospiraceae were decreased in the fecal samples of CD patients.
Faecalibacterium prausnitzii, an important protective commensal bacteria, has been known to be decreased in CD patients [
28].
In the present study, we compared the intestinal microbial community between fecal samples and mucosal tissue samples in HC. In contrast to predominant proportions of
Firmicutes and
Bacteroidetes in fecal samples, the proportion of
Proteobacteria was markedly increased in mucosal tissue samples with concomitant decrease of
Firmicutes and
Bacteroidetes. In addition, PCoA showed certain clustering pattern between fecal and mucosal tissue samples. Analysis of the microbial community between feces and mucosal tissues in CD patients also showed different bacterial community structures. These data are in accordance with previous studies demonstrating a considerable difference in composition between the mucosal and fecal samples [
29,
30]. Considering mucosal associated bacterial profiles might have more important role in IBD than luminal bacteria, it is important to know that fecal microbial analysis does not fully reflect mucosal bacterial community structure [
31].
It is well known that the intestinal microbial community of CD patients demonstrates dysbiosis different from healthy population [
11,
12]. However, whether the changes of intestinal microbial composition in CD are primary or secondary events has not been determined yet. Transmission of colitis with fecal transplant from diseased donor to wildtype recipient, association of dysbiosis with genetic polymorphisms, and induction of immune mediated colitis through mono-association with commensal bacteria suggest dysbiosis as a primary cause of IBD. Meanwhile, similar dysbiotic patterns in nonspecific intestinal inflammation or infection, increased mucosal associated bacteria in patients with infectious colitis, and reversal of dysbiotic pattern with steroid treatment in IBD patients suggest secondary events of dysbiosis [
32]. In our study, certain intestinal microbial population was related to clinical disease activity in CD patients. The relative abundances of
Gammaproteobacteria and
Fusobacteria were higher in both fecal and mucosal tissues of active-stage CD patients than inactive-stage CD patients. These results agree with several previous reports demonstrating disparities of intestinal microbial community between active versus quiescent IBD [
33,
34]. We also demonstrated that the intestinal microbial composition in CD patients was related to disease location and behavior. Moreover, anti-TNF (infliximab) treatment affected a change in the intestinal microbial community structure in CD patients. These our results suggest dysbiosis as a secondary change of intestinal inflammation in CD.
This is the first study showing the unique change of intestinal microbiota community after infliximab treatment. 5-aminosalicylic acid (5-ASA), which is used for the treatment of IBD, activates peroxisome proliferator-activated receptor-r and affects luminal pH and intestinal mucosal pro-inflammatory cytokines. Although 5-ASA has previously been shown to alter the mucosal microbiota composition [
29,
35], these effects of 5-ASA were not seen in our study because most of the enrolled patients had been taking 5-ASA. In contrast, a definitive clustering trend of microbial clades and differences of relative bacterial proportion were observed between infliximab-treated group and infliximab-untreated group. Interestingly, the relative abundance of
Gammaproteobacteria was increased in both fecal and mucosal tissue samples of the infliximab-treated group. However, the differences of relative bacterial abundances between two groups were not significantly different. In addition, we did not investigate the sequential changes of intestinal microbial composition in same patients before and after infliximab treatment. Considering that the proportion of
Proteobacteria increases as the severity of inflammation worsens, it is possible that the patients in the infliximab-treated group have more severe intestinal inflammation. Further studies are warranted to confirm potential influence of infliximab treatment on intestinal microbial community in CD patients.
In this study, the relative abundance of
Fusobacterium genus was increased in both fecal and mucosal tissue samples of CD patients compared to HC. In addition, the proportion of
Fusobacteria class were higher in CD patients with Montreal classification L3 lesion, which involves both small and large intestines. Moreover, the relative proportion of
Fusobacteria class was decreased in the infliximab-treated group.
Fusobacteria has recently become the bacterial strain of concern because of its relationship with IBD and colorectal cancer [
36,
37].
Fusobacterium nucleatum is associated with chronic inflammation of the oral cavity, and it is commonly cultured from the intestinal mucosa of IBD patients. Because of its invasiveness, it is closely related with intestinal inflammation. Further studies using genetically susceptible mice monoassociated with
Fusobacteria species will be needed to confirm whether this bacterial strain plays an important role in the development of chronic colitis as a pathobiont.
Our study faces several limitations. First of all, analysis at the species level could not be performed due to the technical limitations of metagenomic sequencing, and the extraordinary cost of whole genome sequencing. Secondly, most of our sequencing data regarding the differences of relative abundances of specific bacteria between groups could not reach statistically significant because of small sample size and relatively high variability of microbial composition in each group. Further study with a large population is required to confirm our data regarding such bacterial profiles of feces and mucosal tissues in Korean CD patients. Thirdly, in the present study, the effects of age on the intestinal microbial composition were minimized by enrolling Korean CD patients whose age was within the limits of CD prevalence and an age-matched HC group.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
CSE and MJK participated in the design of the study; participated in data acquisition; performed data analysis and interpretation; wrote the manuscript. DSH conceived the study; participated in the design of the study; performed critical revision of manuscript for intellectual content; performed the study supervision. ARL participated in data acquisition. DIP and SKY participated in data acquisition; participated in data analysis and interpretation. YSK and JFK participated in data analysis and interpretation. All authors read and approved the final manuscript.