Gut-derived Enterococcus faecium from ulcerative colitis patients promotes colitis in a genetically susceptible mouse host
Recent metagenomic analyses have revealed dysbiosis of the gut microbiota of ulcerative colitis (UC) patients. However, the impacts of this dysbiosis are not fully understood, particularly at the strain level.
We perform whole-genome shotgun sequencing of fecal DNA extracts from 13 healthy donors and 16 UC and 8 Crohn’s disease (CD) patients. The microbiota of UC and CD patients is taxonomically and functionally divergent from that of healthy donors, with E. faecium being the most differentially abundant species between the two microbial communities. Transplantation of feces from UC or CD patients into Il10−/− mice promotes pathological inflammation and cytokine expression in the mouse colon, although distinct cytokine expression profiles are observed between UC and CD. Unlike isolates derived from healthy donors, E. faecium isolates from the feces of UC patients, along with E. faecium strain ATCC 19434, promotes colitis and colonic cytokine expression. Inflammatory E. faecium strains, including ATCC 19434 and a UC-derived strain, cluster separately from commercially available probiotic strains based on whole-genome shotgun sequencing analysis. The presence of E. faecium in fecal samples is associated with large disease extent and the need for multiple medications in UC patients.
E. faecium strains derived from UC patients display an inflammatory genotype that causes colitis.
KeywordsInflammatory bowel disease Crohn’s disease Microbiota Metagenome
The hundreds of trillions of microbes in the gastrointestinal tract contribute to many host physiological processes including nutrient acquisition and development of the gut immune system, while dysbiosis of the microbiota can contribute to the development of several diseases . The commensal gut microbiota is closely related to the pathogenesis of inflammatory bowel diseases (IBD), including ulcerative colitis (UC) and Crohn’s disease (CD) . A significant amount of information on the role of the gut microbiota in IBD patients has been accumulated based on clinical studies, metagenome analyses, and animal experiments. The recent development of high-throughput metagenome sequencing techniques and analysis methods has uncovered dysbiosis of the microbiota of IBD patients, with a decrease in microbial diversity compared with healthy individuals [3, 4, 5]. In addition, the gut community of IBD patients shows an increased prevalence of the phylum Proteobacteria , including inflammatory species Escherichia coli , and a decrease in the phylum Firmicutes , including anti-inflammatory species Faecalibacterium prausnitzii [4, 7]. While metagenomic analysis can reveal an association between dysbiosis and disease, animal studies can demonstrate a causative association between specific bacteria and the pathogenesis of colitis. According to previous mouse studies, F. prausnitzii , Bacteroides fragilis , and Clostridia species  attenuate intestinal inflammation via various mechanisms, including induction of IL-10 and regulatory T cells in the colon, whereas Escherichia coli  and Enterococcus faecalis [11, 12] promote colitis in mouse models.
Although previous studies have built up a significant body of information, questions remain about the causal relationship between human IBD pathogenesis and microbiota dysbiosis. Firstly, because of the complexity of the gut microbiota and the multifactorial nature of IBD, it is unclear exactly which bacterial species in the dysbiotic IBD microbiota are responsible for the pathogenesis of colitis. Because few reports reproduce the microbial composition of the dysbiotic microbiota in colitic mouse models, it is unclear whether decreases or increases in a single anti-inflammatory or pro-inflammatory species are truly responsible for IBD pathogenesis. Secondly, the specific genotypes of bacterial species putatively responsible for IBD pathogenesis are yet to be elucidated. Metagenomic analysis of the gut microbiota, particularly 16S ribosomal RNA (rRNA) sequencing-based approaches, cannot identify the composition of the microbiota at the strain level. Therefore, it is difficult to judge based only on metagenomic analysis whether the decreased or increased prevalence of a species in the IBD microbiota has a pathogenic effect.
In the present study, to answer these questions, we attempted to find a link between human metagenome data and phenotype in a mouse model of colitis. We first analyzed whole-genome shotgun sequencing data corresponding to fecal metagenomes obtained from patients with IBD. Feces from the same subjects was then transplanted into Il10−/− mice to assess the causal relationship between dysbiosis of the microbiota and intestinal inflammation. Finally, the genotype of a bacterial species possibly responsible for inflammation, Enterococcus faecium, was identified by sequencing multiple strains isolated from UC patients. Thus, we clearly showed a causal relationship between UC-derived bacterial strains and colonic inflammation.
The gut microbiota of IBD patients is taxonomically and functionally divergent from that of healthy donors
To investigate taxonomic and functional differences in gut microbiota between patients with IBD and healthy controls, fecal samples were collected from 13 healthy donors (HD), 16 UC patients, and 8 CD patients. The CD patients were younger and had higher serum C-reactive protein levels than the HD, although no difference was observed between HD and UC patients in other baseline characteristics (Additional file 1: Table S1). The extent of disease in UC patients was either left-sided colitis or pancolitis (left-sided colitis, 68.8%; pancolitis, 31.3%; Additional file 1: Table S2). Fifteen of the 16 UC patients were being treated with medications, including mesalazine, corticosteroids, tacrolimus, azathioprine, mercaptopurine, and TNF antagonist, and both active and inactive colitis patients were represented in the UC cohort according to total Mayo scores (average Mayo score, 3.25; Additional file 1: Table S2). All the CD patients were being treated with medications, including mesalazine, corticosteroids, azathioprine, and TNF antagonist, and in most cases, the disease was inactive based on the Crohn’s disease activity index (average CDAI, 71.1; Additional file 1: Table S3).
The gut microbiota of UC and CD patients cause colitis in Il10−/− mice
To explore a causal relationship between the gut microbiota and colitis, the fecal samples used in the sequencing analysis were transplanted into Il10−/− colitis-susceptible mice. Fecal microbiota samples from each of the subjects were transplanted into at least two microbiota-depleted Il10−/− mice, and colonic inflammation was evaluated on day 28 post-transplantation by pathology and gene expression analyses. Although longitudinal analysis of the fecal microbiota following fecal transplantation revealed that the composition of the human gut community was not completely restored in the transplanted mice (Additional file 2: Figure S1a and b), the composition of the murine microbiota was uniquely dependent on the human subjects’ microbiota and remained stably segregated from the composition in other mice up to 28 days post-transplantation (Additional file 2: Figure S1c, P = 0.001, PERMANOVA). Effects of cage difference (Additional file 2: Figure S1d) and repeats of transplantation (Additional file 2: Figure S1e) were modest, and PCoA plots were significantly differentially segregated by differences between the donor feces (Additional file 2: Figure S1d, P = 0.001; Additional file 2: Figure S1e, P = 0.029: PERMANOVA). Therefore, we evaluated the effects of these unique human-derived microbial communities on the severity of colitis in the mouse model.
The abundance of Enterococcus in the fecal microbiota of mice is associated with inflammation in the colon
E. faecium causes colitis in Il10−/− mice
Subject-derived E. faecium strains induce different pathological changes and cytokine expression profiles in the colon
Consistent with these findings, strain HD26a did not increase the expression of inflammatory cytokines in colon tissues compared with HD55 microbiota transplantation (Fig. 6d). Interestingly, colon tissue cytokine expression profiles differed following inoculation with IB18a or IB51a. IB51a induced higher levels of expression of Il1b and Il12b compared with the HD55 microbiota, while IB18a only induced higher expression of Il12b (Fig. 6d). Commercially available E. faecium strain SF68, which is used as a probiotic drug for animals , was also tested by inoculation into microbiota-depleted Il10−/− mice. SF68 did not cause pathological inflammation and increases of colonic cytokine expression, whereas UC-derived strain IB44a caused pathological colitis and increased expression of Tnf, Il12b, and Il17a (Additional file 2: Figure S7). Next, UC-derived strain IB51a was inoculated into germ-free Il10−/− mice to see whether a single strain was sufficient to induce colitis. Inoculation with IB51a induced colonization with 100 times more E. faecium than transplantation of IBD51 feces that was the origin of IB51a (Additional file 2: Figure S6 g). Both monocolonization with IB51a and transplantation of IBD51 feces into germ-free mice induced more severe colitis than transplantation of HD55 feces (Fig. 6e). Strain IB51a increased expression of Tnf, Il1b, Il12b, and Il17a in the colon tissue, but IBD51 feces only increased expression of Il23a compared with HD55 feces (Fig. 6f). The cellular source of IL-17 in colonic lamina propria was both Th17 cells and CD3-negative cells (Additional file 2: Figure S8a and b). TNF and IL-6 were mainly produced by CD11c−MHC-class-II+ cells as well as smaller proportion of CD11c+MHC-class-II+ cells (Additional file 2: Figure S8c to f).
Presence of E. faecium in the gut is associated with disease extent and the requirement for combination therapy in UC patients
A causal link between specific bacterial species and the pathogenesis of IBD remains controversial because of the multifactorial causes of IBD and difficulties surrounding reproducing the disease in mouse models. In this study, we attempted to identify a causal relationship between specific bacterial species and IBD by conducting a metagenomic analysis of colitis-susceptible Il10−/− mice transplanted with fecal microbiota from 37 human subjects. The fecal microbiota from both UC and CD patients caused pathological inflammation in the colon tissues of experimental animals, while microbiota from HD rarely caused severe colitis (Fig. 2c–e). The microbiota from UC patients induced expression of the Il6–Il17a axis in the colon, whereas the microbiota from CD patients induced expression of Il23a (Fig. 3a). Enterococcus was differentially abundant in the microbiota of UC patients compared with the HD group (Fig. 1a), which was replicated in the mice transplanted with the UC microbiota (Fig. 4a and Additional file 2: Figure S5). E. faecium strain ATCC 19434 (Fig. 5c–e), along with strains isolated from UC patients (Fig. 6c, d), caused pathological inflammation and upregulation of cytokine expression in the colon. The genotypes of 10 analyzed E. faecium strains were different and could be separated into two major clusters: one containing two probiotic strains and the other containing pro-inflammatory strain ATCC 19434 (Fig. 7a). Thus, these findings implicate E. faecium strains with a particular genotype in colonic inflammation in genetically susceptible hosts.
The metagenomic analysis carried out in this study revealed obvious differences in the composition of the microbiota from UC or CD patients compared with that from HD (Fig. 1c). Several of the observed differences between the IBD patients and HD were consistent with previous reports, such as the decreased abundance of F. prausnitzii [7, 17] and Eubacterium rectale  in the microbiota of UC patients, and the increased abundance of Escherichia species, including Escherichia coli , in the CD microbiota. According to previous reports, enterococci are more abundant in the fecal matter [5, 18] and mucosa  of CD patients compared with healthy controls. In this analysis, E. faecium was more abundant in the UC microbiota, but not the CD microbiota, compared with HD (Fig. 1a). Although the results of metagenomic analyses are often affected by large interindividual differences arising from differences in factors such as geographic location, diet, and antibiotic use, the overall results of the compositional difference between HD and IBD in the current study are fairly consistent with previous reports. Because most of the patients enrolled in this study were being treated for their disorder, it is unclear whether the observed changes in microbiota are associated with the disease onset. Longitudinal metagenomic analysis from a naïve state to a treated state for IBD patients is therefore necessary to clarify the association between disease onset and specific bacterial species.
Inflammatory cytokines are crucial to the pathogenesis of IBD . TNF is a central driver of inflammation in mucosal tissues, as evidenced by the anti-inflammatory effects of a neutralizing antibody against TNF that is currently used worldwide for the treatment of both UC and CD [21, 22]. IL-6 produced by lamina propria myeloid cells prevents apoptosis of T cells and activates macrophages . Antibody-mediated blockade of IL-6 signaling resulted in clinically beneficial responses in a subgroup of CD patients in a clinical trial . IL-6 is also necessary for differentiation of Th17 cells. IL-17A produced by Th17 cells in the lamina propria is abundant in the mucosa of both UC and CD patients . Although neutralization of IL-17A was not effective in the relief of colitis in a mouse model  or in a clinical study of CD patients , another Th17-type cytokine, IL-21, also has pro-inflammatory effects and may be a suitable therapeutic target . IL-23 is mainly produced by antigen-presenting cells to stabilize Th17 cells, but also activates macrophages. A neutralizing antibody against IL-12p40, a subunit of IL-23, is now used for the treatment of CD patients worldwide . Interestingly, in this study, cytokine expression profiles in colon tissues of Il10−/− mice were dependent on the composition of the transplanted microbiota. The UC microbiota tended to increase the expression of the Il6–Il17a axis, while the CD microbiota increased Il23a expression in the colon (Fig. 3a). In addition, cytokine profiles were dependent on the genotypes of the E. faecium strains in the Il10−/− mice (Fig. 6d). Strain IB51a increased expression of Tnf, Il1b, Il12b, and Il17a in gnotobiotic conditions, but original IBD51 feces only increased Il23a expression (Fig. 6f). Even in the same host background, differences in the bacterial community composition might affect the specific cytokine expression profiles of the colon tissues. This study also showed distinct differences in bacterial species composition between the UC and CD microbiota and the HD microbiota (Fig. 1a, b), as previously reported . Similar to the current study, differences in Klebsiella pneumoniae genotype induced different inflammation profiles in the colon of Il10−/− mice . Therefore, accumulating information regarding the relationship between the expression of intestinal cytokines and the composition of the gut microbiota, as well as the genetic background of the host, may help in the development of individualized treatment protocols to appropriately neutralize various cytokines.
Il10−/− mice, which lack anti-inflammatory cytokine IL-10 and show spontaneous colitis in a gut microbiota-dependent manner , were used in this study. Because IL10 gene polymorphisms contribute to UC  and CD  susceptibility in humans, the Il10−/− mouse model is a good approximation of the genetic susceptibility of IBD patients. However, multiple alleles are usually involved in the pathogenesis of human IBD , and not all IBD patients have an IL10 polymorphism. Therefore, further investigation using other mouse models of IBD is necessary to validate the results of the current study, although mouse models of genetic susceptibility to colitis are scarce. Microbiota-dependent patterns of cytokine production also need to be confirmed in other mouse models. A limitation of this study was that transplantation of human feces into mice after depletion of native mouse microbiota with antibiotics resulted in incomplete restoration of the human microbiota (Additional file 2: Figure S1). Thus, we were able to study effects of subject-unique microbiota by fecal transplantation, but the microbiota were not the same as the original microbiota of the donor human. Despite this limitation of the mouse experiments in this study, the presence of E. faecium was associated with increased disease extent and the requirement for multiple medications in UC patients with a non-specific genetic background. Taken together, the results of the current study suggest that E. faecium may promote colonic inflammation in UC patients. To validate the inflammatory effects of E. faecium in IBD patients, well-designed clinical studies using a larger UC population are necessary.
Probiotic E. faecium strains NCIMB 11181  and SF68 [14, 32] have been used for many years to maintain animal health; however, the detailed molecular mechanisms underlying their probiotic effects have not been fully elucidated. In comparison, the virulence factors of E. faecium have been extensively studied because of the significant health concern caused by vancomycin-resistant E. faecium strains in hospitalized patients . Unlike streptococci and staphylococci, most enterococci do not produce pro-inflammatory toxins; instead, they possess many genes encoding adhesion proteins that mediate adherence to host tissues . Acm is one such adhesion protein produced by clinical E. faecium strains. Acm binds collagen in host tissues, and deletion of acm results in attenuation of E. faecium-mediated endocarditis in an animal model of disease . Surface protein Esp in E. faecium appears to have been acquired as part of a pathogenicity island, and promotes biofilm formation and urinary tract infection in mice . Adhesion and biofilm formation are important properties involved in the translocation of luminal bacteria in colon tissues . The genetic analysis of 10 E. faecium strains in the current study identified that pathways for secondary bile acid biosynthesis were abundant in the pro-inflammatory cluster of E. faecium strains. Because particular types of bile acids control biofilm formation of Enterococcus , polarization of host macrophages, generation of oxidative stress , and expansion of pathobionts in the colon , the ability of “inflammatory” E. faecium to alter the composition of bile acids in the intestinal environment may be a key feature of pathogenesis of E. faecium in colitis. Abundance of the pathway of lipoic acid metabolism was also different between the inflammatory and the probiotic clusters. Lipoic acids are considered as anti-oxidants and effective to relieve colitis in pre-clinical animal studies . Because E. faecium strains in the inflammatory cluster possessed ability to generate more ROS than the probiotic cluster, ROS-related pathogenic effect may be involved in the colitis induced by inflammatory E. faecium. In several fecal transplantation, E. faecium caused colitis in Il10−/− mice with low bacterial load (Additional file 2: Figure S5). Ability of inflammatory E. faecium in generating metabolites related to bile acids and ROS might account for a reason why low bacterial load of the strains promoted colitis.
Metagenomic analysis of the fecal microbiota of IBD patients and transplantation of feces from the corresponding patients into genetically susceptible animals confirmed a causal relationship between inflammatory E. faecium strains and colitis. Because the cytokine expression profile in the colon may partly be dependent on the gut microbiome, characterizing the composition of a patient’s gut microbiota may lead to personalized antibody therapy for IBD patients in the future.
Clinical study population
This study aimed to examine fecal microbiota of IBD patients who visited Kanazawa University Hospital in various stages including both active and remission stages. Between February 2014 and March 2015, 60 IBD patients enrolled in the study. Among them, 55 subjects received colonoscopy to evaluate disease status. To investigate the composition and metagenomes of the gut microbiota of the different patient groups, fecal samples were collected from patients immediately after enrollment and stored at − 80 °C. Among the 55 subjects, 16 UC patients and 8 CD patients provided us with good and sufficient feces immediately after enrollment, and these feces were used for the study. Feces that were too watery or that were left at room temperature for > 3 h were not used. Patients received both written and oral information before consenting to participate in this study. Healthy individuals were first tested to confirm that they met the following inclusion criteria: body mass index < 25 kg/m2, normal blood pressure, normal serum cholesterol, normal blood glucose and hemoglobin A1c levels, normal serum aspartate transaminase and alanine aminotransferase levels, no anemia, no fatty liver noted by ultrasonography, and no past history of cancer. In total, 13 individuals were enrolled as HD. Immediately after consent, feces were collected from the HD and stored at − 80 °C. This study was approved by the Ethics Committee of Kanazawa University (approval number 2012-109).
Male and female Il10−/− mice with a C57BL/6j background were purchased from Jackson Laboratories (Bar Harbor, ME, USA). Mice were housed in individually ventilated racks in an animal facility with access to autoclaved food and water ad libitum. Stringent husbandry techniques, including a strictly enforced order of cage handling and scrupulous attention to environmental sanitization, were followed to prevent contamination. Germ-free Il10−/− mice with a C57BL/6j background were generated by Japan CLEA (Tokyo, Japan). Germ-free or gnotobiotic mice were housed in germ-free isolators (ICM, Tsukuba, Japan) in the animal facility of Kanazawa University. All animal procedures were performed in accordance with the Guidelines on the Care and Use of Laboratory Animals issued by Kanazawa University and were approved by the Ethical Committee for Animal Experiments of Kanazawa University (approval number 143267).
Fecal transplantation and bacterial inoculation
Frozen fecal samples were thawed and resuspended in a 10 times volume (w/v) of reduced phosphate-buffered saline (PBS) containing 0.1% resazurin (w/v) and 0.05% l-cysteine-HCl (Sigma-Aldrich, St. Louis, MO, USA) under anaerobic conditions (80% N2, 10% H2, 10% CO2) in an anaerobic chamber (RUSKINN, Bridgend, UK), before being filtered through a 100-μm pore size nylon membrane filter. Depletion of the gut microbiota was performed as previously described , with slight modification. Briefly, 8- to 12-week-old C57BL6 mice were treated with antibiotics, including doripenem hydrate (0.25 g/L; Shionogi, Japan) and vancomycin hydrochloride (0.5 g/L; Shionogi), via water for 2 weeks prior to experimentation to deplete the gut microbiota. Fecal suspensions were then inoculated into the microbiota-depleted mice by oral gavage (100 μL) and rectally (100 μL) using a flexible plastic tube on days 1 and 2 after cessation of antibiotic treatment. For bacterial inoculation experiments, E. faecium strains ATCC 19434 (ATCC, Manassas, VA, USA), NCIMB 11181 (NCIMB, Aberdeen, UK), SF68 (also known as NCIMB 10415; PURINA, Largo, FL, USA), or human feces-derived strains were prepared at a concentration of 2.5 × 108 colony-forming units (CFU)/100 μL in PBS. Suspensions were then inoculated into microbiota-depleted mice as described for the inoculation of fecal suspension. The microbiota-depleted mice were housed individually in separate isolators following inoculation with fecal suspension or specific bacterial strains.
Whole-genome shotgun and 16S rRNA amplicon sequencing analysis of fecal and bacterial DNA extracts
Microbial DNA was extracted from the stored fecal samples using a PowerFecal DNA Isolation Kit according to the manufacturer’s instructions (MO BIO, Carlsbad, CA, USA), although 0.1-mm glass beads (MO BIO) rather than 0.7-mm garnet beads were used to homogenize feces.
For whole-genome shotgun sequencing, DNA fragmentation was performed using a Nextera DNA Library Prep Kit (Illumina, San Diego, CA, USA) with incubation at 55 °C for 5 min. Following product clean-up using a DNA Clean & Concentrator-5 Kit (ZYMO, Irvine, CA, USA), indexing PCR was performed using a Nextera Index Kit (Illumina), with an initial cycle of 72 °C for 3 min and 98 °C for 30 s, followed by 5 cycles of 98 °C for 10 s, 63 °C for 30 s, and 72 °C for 3 min. PCR products were purified from oligo DNA contaminants using AMPure XP beads (Beckman Coulter, Fullerton, CA, USA) and then quantified using a 2100 Bioanalyzer and High Sensitivity DNA Kit (Agilent Technologies, Santa Clara, CA, USA). The libraries were pooled, and sequencing was performed using the MiSeq system (Illumina) with a MiSeq Reagent Kit V3 (600 cycles; Illumina).
Preprocessing of the acquired sequences was performed as follows. After trimming of bases with low-quality scores from the ends of the acquired sequences, the sequences were filtered at a Q-score cutoff of 20 using FASTX Toolkit . Paired-end joining was performed using MacQIIME v1.9.1 . After mapping of the resultant sequences against UCSC human reference genome hg19 using bowtie2 version 2.2.4 , human genome sequences were removed using SAMtools-1.2 . Finally, PCR duplicates were removed using PRINSEQ version 0.20.4 . The average number of reads per sample before and after quality control was 1,544,241 ± 687,438 and 1,106,822 ± 560,821 (mean ± SD), respectively.
The resultant FASTA files were used for taxonomic analysis of the fecal samples using MetaPhlAn2 version 2.0.0 . Metabolic pathways indicated by the sequences were identified using HUMAnN2  version 0.1.9. A KEGG version 56 database was generated using DIAMOND version 0.7.5  and used for HUMAnN2 analysis. After normalization of the abundance to 1 million reads, comparisons between groups were performed by LDA using the linear discriminative analysis effect size (LEfSe) tool . The Bray-Curtis dissimilarities among the gut microbiota communities were calculated based on taxonomic or metabolic pathway data, and the resultant distances were visualized by PCoA. Statistical differences between communities were tested using PERMANOVA with MacQIIME version 1.9.1.
Whole-genome shotgun sequencing of individual E. faecium strains was performed as described for fecal metagenome analysis. Assembly of the obtained paired-end FASTQ files was performed using SPAdes version 3.12.0 . The resultant scaffold FASTA files were used to annotate genomic features in PROKKA version 1.12 . Hierarchical clustering of 1683 genes from the 10 E. faecium strains annotated by PROKKA was performed using an average linkage method in Genesis version 1.7.6 . KEGG-based metabolic pathways indicated by the sequences were also identified using HUMAnN2 version 0.1.9, and comparisons between “inflammatory” and “probiotic” clusters were performed by LEfSe.
For 16S rRNA gene sequencing analysis, PCR amplicons were prepared using primers targeting the V3–V4 region of the gene, with Illumina adapter sequences:
Forward 5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG-3′; Reverse 5′-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC-3′.
PCR reactions were performed using KAPA HiFi HotStart Ready Mix (KAPA Biosystems, Wilmington, MA, USA) with an initial step at 95 °C for 3 min, followed by 25 cycles of 95 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s, with a final elongation at 72 °C for 5 min. Following purification, indexing PCR was performed using a Nextera XT Index Kit (Illumina) with an initial step at 95 °C for 3 min, followed by 8 cycles of 95 °C for 30 s, 55 °C for 30 s, and 72 °C for 30 s, with a final elongation step at 72 °C for 5 min. Libraries were pooled and mixed with the PhiX Control Library (Illumina) before being sequenced using the MiSeq system with a MiSeq Reagent Kit V3 (600 cycles). Downstream processing of the sequences was performed using MacQIIME version 1.9.1 . After paired-end sequence joining and trimming of low-quality reads, the resulting sequences had lengths 442–464 bp. The average number of reads per sample after quality control was 96,311 ± 35,047 (mean ± SD). Operational taxonomic units (OTUs) were assigned using UCLUST. For OTU analysis, sequences were clustered, and then, those with > 97% similarity were binned into the same OTU. Taxonomic assignment of representative sequences from each OTU was performed using RDP Classifier using the Greengenes reference database  clustered at 97% identity. Genus-level taxonomy was summarized and used for subsequent analyses. Following the generation of a phylogenetic tree, unweighted UniFrac distances for the gut microbiota communities were calculated, with the resultant distances visualized by PCoA.
All sequencing data and metadata were archived in the NCBI Sequence Read Archive under BioProject numbers PRJNA511372 and 511382.
Total RNA extraction and cDNA synthesis were performed as previously described . Briefly, colon tissue samples were fixed in RNAlater RNA Stabilization Reagent (Qiagen, Hilden, Germany) and stored at − 80 °C until processing. Total RNA was isolated from the specimens using an RNeasy Mini Kit (Qiagen) according to the manufacturer’s instructions. A High Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA, USA) was used to reverse transcribe 500 ng of total RNA into first-strand cDNA. Quantitative PCR was carried out using a real-time 7900HT Sequence Detection System (Applied Biosystems). TaqMan probes used for real-time PCR were as follows: Tnf (Mm00443260_g1), Il1b (Mm00434228_m1), Il6 (Mm00446190_m1), Il12b (Mm01288989_m1), Il17a (Mm00439618_m1), and Il23a (Mm00518984_m1) (Applied Biosystems). To control for variations in the amount of DNA available for PCR, target gene expression in each sample was normalized relative to the expression of endogenous control gene β-actin (Applied Biosystems) using the ΔΔCt method. PCR amplification was performed with an initial cycle of 50 °C for 2 min and 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min.
Detection of E. faecium and E. faecalis in fecal samples was performed as previously described . Briefly, 2 ng of fecal DNA was amplified with SYBR Green (Qiagen) and primers Efm12 or Efl6  to detect E. faecium or E. faecalis, respectively, by reaction at 94 °C for 10 min, followed by 40 cycles of 94 °C for 15 s and 60 °C for 1 min. To draw a standard curve, DNA extracted from E. faecium strain ATCC 19434 or E. faecalis strain V583 (ATCC) was used, and copy number was calculated.
Colon samples were fixed in 10% buffered formalin using a “Swiss roll” technique . Fixed tissues were embedded in paraffin and stained with hematoxylin and eosin. Hematoxylin and eosin-stained sections were used for histological scoring. Tissue sections were coded to eliminate pathologist bias and scored as previously described , with slight modification. Briefly, the proximal colon, distal colon, and rectum from each mouse were scored based on the severity of mucosal epithelial changes, degree of inflammation, and extent of pathology (Additional file 1: Table S7). The segment score was calculated by summing the severity scores: [segment score = mucosal score + inflammation score + extent of segment affected in any manner (extent 1) + extent of segment affected at level 3 or 4 in M or I score (extent 2)], where the maximum segment score was 15. The total score for each mouse was calculated by summing the scores from the individual segments (maximum total score of 45). Subsequently, the mean total scores were calculated for each group.
E. faecium strains were proliferated in Brain Heart Infusion (BHI) medium (Eiken Chemical Co., Tokyo, Japan). To isolate E. faecium strains from human feces, fecal samples were streaked onto ECS agar plates (Eiken Chemical Co.). Single colonies were picked from the selection agar and cultivated on EF agar plates (Nissui, Japan), upon which E. faecalis forms purple colonies and E. faecium produces orange colonies. Individual orange colonies were picked and proliferated in BHI broth. Species identification was carried out using PCR-based methods as previously described [53, 55]. Finally, Enterococcus species were confirmed by whole-genome shotgun sequencing as described in the “Whole-genome shotgun and 16S rRNA amplicon sequencing analysis of fecal and bacterial DNA extracts”section. Fecal suspension was cultured on EF agar plates to obtain CFU counts, and colonies were confirmed as E. faecium by PCR-based methods.
Intracellular cytokine staining by flow cytometry
Single cell suspension was obtained from the colonic lamina propria. The colon tissues were cut into 1-cm pieces and incubated with 5 mM ethylenediaminetetraacetic acid (Invitrogen) in Hanks’ balanced salt solution at 37 °C for 20 min, followed by washing with HBSS three times. Next, the tissue was minced with scissors and transferred into conical tubes containing 10 mL RPMI 1640 medium (Thermo Fisher, Waltham, MA, USA) and 0.5 mg/mL collagenase (Sigma-Aldrich, St. Louis, MO, USA) and incubated at 150 rpm and 37 °C for 30 min. The single cell suspension was collected and passed through 100-μm and 40-μm cell strainers and washed twice with RPMI 1640. For analysis of IL-17A expression, the single colonic lamina propria cells were restimulated in complete RPMI 1640 with 5 ng/mL phorbol 12-myristate 13-acetate and 500 ng/mL ionomycin in the presence of 2 μM monensin (Sigma-Aldrich) for 3 h at 37 °C. For analysis of IL-6 and TNF expression, the cells were incubated in complete RPMI 1640 in the presence of 2 μM monensin for 3 h at 37 °C. Dead cells were excluded from all analyses using a LIVE/DEAD Fixable Aqua Dead Cell Stain Kit (Thermo Fisher). For intracellular cytokine staining, cells were fixed with IC Fixation Buffer (Thermo Fisher). Fluorescent dye-conjugated antibodies were used to stain CD45, CD3 (BD Biosciences, Franklin Lakes, NJ, USA), CD4, CD11c, MHC-class II, IL-17, TNF, and IL-6 (BioLegend, San Diego, CA, USA). All data were acquired on a FACS Aria II flow cytometer (BD Biosciences) and analyzed using FlowJo V.10 software (FlowJo, Ashland, OR, USA).
E. faecium strains were tested for hydrolase activity against glyco-conjugated bile acid as previously described . Overnight MRS broth cultures were streaked on MRS agar supplemented with 2 mM GDCA (Sigma; G3258). The plate was then incubated for 48 h at 37 °C. BSH activity was detectable when deoxycholic acid precipitated in the agar medium around a colony.
ROS measurement in culture supernatant
OxiSelect In vitro ROS/RNS assay kit (green fluorescence) (Cell Biolabs, cat. STA-347) was used to measure the ROS/RNS (reactive nitrogen species) in culture supernatant as previously described  with modification. After E. faecium strains were cultured in BHI medium overnight, 50 μL of the supernatants were mixed with 50 μL of catalyst (provided in the kit) to accelerate the oxidative reaction. Following 5 min incubation at room temperature, 100 μL of DCFH-DiOxyQ probe solution was added to the mixture to measure the total free radical population. DCFH probe can react with free radical molecules that are representative of both ROS and RNS. The samples were incubated at room temperature for 30 min and read with a fluorescence plate reader at Ex/Em = 480/530 nm. The standard curve of H2O2 was used to semi-quantify the free radical content in the culture supernatant samples. Then, relative ROS level was determined.
Statistical analyses of taxonomic and KEGG pathway comparisons between microbiota communities were performed using the LEfSe tool . Statistical differences between two values were analyzed using a Mann-Whitney U test or the Kruskal-Wallis test followed by Dunn’s test for correction of multiple comparisons in GraphPad Prism 7 (GraphPad Software, San Diego, CA, USA). Statistical differences between treatments or in the percentages of disease extent were analyzed by χ2 test in GraphPad Prism 7. A P value < 0.05 was considered statistically significant. Linear regression coefficients between pathology scores, percentage body weight changes, and Tnf expression levels were calculated using GraphPad Prism 7. Spearman’s rank correlation was analyzed using GraphPad Prism 7.
The authors thank Dr. Souichiro Sasaki (Division of Molecular Bioregulation, Cancer Research Institute, Kanazawa University) for advice regarding animal experiments, Ms. Akiko Katayama (Kanazawa University) for assistance with histological analysis, Ms. Maki Wakabayashi (Kanazawa University) for animal care, and Ms. Masami Kitaoka (Kanazawa University) for collecting fecal samples. We also thank Tamsin Sheen, PhD, from Edanz Group (www.edanzediting.com/ac) for editing a draft of this manuscript.
The review history is available as Additional file 5.
Peer review information
Kevin Pang was the primary editor on this article and managed its editorial process and peer review in collaboration with the rest of the editorial team.
JS and NI contributed equally to this work. NI and EM conceived and designed this study. JS and NI acquired, analyzed, and interpreted the data from animal experiments. NI, KK, TK, and YSh collected the human samples. JS, NI, ZW, MY, and YF acquired and analyzed the bacterial experiment data. NI acquired, analyzed, and interpreted the data from sequencing experiments. AS analyzed the flow cytometry data. JS and NI drafted the manuscript. TarY, YSa, MH, TatY, EM, and SK critically revised the manuscript for important intellectual content. EM and SK supervised this study. All authors read and approved the final manuscript.
Ethics approval and consent to participate
The clinical study was approved by the Ethics Committee of Kanazawa University (approval number 2012-109). All animal experiments were performed in accordance with the Guidelines on the Care and Use of Laboratory Animals issued by Kanazawa University and were approved by the Ethical Committee for Animal Experiments of Kanazawa University (approval number 143267). All experimental methods were in accordance with Helsinki Declaration.
Consent for publication
All participants provided written informed consent for publication of study results of the collected biomaterials paired with anonymized information on clinical data.
The authors declare that they have no competing interests.
- 7.Machiels K, Joossens M, Sabino J, De Preter V, Arijs I, Eeckhaut V, Ballet V, Claes K, Van Immerseel F, Verbeke K, et al. A decrease of the butyrate-producing species Roseburia hominis and Faecalibacterium prausnitzii defines dysbiosis in patients with ulcerative colitis. Gut. 2014;63:1275–83.PubMedCrossRefGoogle Scholar
- 8.Sokol H, Pigneur B, Watterlot L, Lakhdari O, Bermudez-Humaran LG, Gratadoux JJ, Blugeon S, Bridonneau C, Furet JP, Corthier G, 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;105:16731–6.PubMedPubMedCentralCrossRefGoogle Scholar
- 14.Bednorz C, Guenther S, Oelgeschlager K, Kinnemann B, Pieper R, Hartmann S, Tedin K, Semmler T, Neumann K, Schierack P, et al. Feeding the probiotic Enterococcus faecium strain NCIMB 10415 to piglets specifically reduces the number of Escherichia coli pathotypes that adhere to the gut mucosa. Appl Environ Microbiol. 2013;79:7896–904.PubMedPubMedCentralCrossRefGoogle Scholar
- 23.Atreya R, Mudter J, Finotto S, Mullberg J, Jostock T, Wirtz S, Schutz M, Bartsch B, Holtmann M, Becker C, et al. Blockade of interleukin 6 trans signaling suppresses T-cell resistance against apoptosis in chronic intestinal inflammation: evidence in crohn disease and experimental colitis in vivo. Nat Med. 2000;6:583–8.PubMedCrossRefGoogle Scholar
- 26.Hueber W, Sands BE, Lewitzky S, Vandemeulebroecke M, Reinisch W, Higgins PD, Wehkamp J, Feagan BG, Yao MD, Karczewski M, et al. Secukinumab, a human anti-IL-17A monoclonal antibody, for moderate to severe Crohn’s disease: unexpected results of a randomised, double-blind placebo-controlled trial. Gut. 2012;61:1693–700.PubMedPubMedCentralCrossRefGoogle Scholar
- 36.McKenney PT, Yan J, Vaubourgeix J, Becattini S, Lampen N, Motzer A, Larson PJ, Dannaoui D, Fujisawa S, Xavier JB, Pamer EG. Intestinal bile acids induce a morphotype switch in vancomycin-resistant Enterococcus that facilitates intestinal colonization. Cell Host Microbe. 2019;25:695–705 e695.PubMedCrossRefGoogle Scholar
- 52.Iida N, Nakamoto Y, Baba T, Nakagawa H, Mizukoshi E, Naito M, Mukaida N, Kaneko S. Antitumor effect after radiofrequency ablation of murine hepatoma is augmented by an active variant of CC Chemokine ligand 3/macrophage inflammatory protein-1alpha. Cancer Res. 2010;70:6556–65.PubMedCrossRefGoogle Scholar
- 55.Ryu H, Henson M, Elk M, Toledo-Hernandez C, Griffith J, Blackwood D, Noble R, Gourmelon M, Glassmeyer S, Santo Domingo JW. Development of quantitative PCR assays targeting the 16S rRNA genes of Enterococcus spp. and their application to the identification of enterococcus species in environmental samples. Appl Environ Microbiol. 2013;79:196–204.PubMedPubMedCentralCrossRefGoogle Scholar
- 58.Iida N. IBD_WGS; WGS data of fecal microbiota of IBD patients and healthy individuals. NCBI Sequence Read Archive. 2018 https://www.ncbi.nlm.nih.gov/bioproject/PRJNA511372/. Accessed 21 Dec 2018.
- 59.Iida N. IBD_mouse_16SrRNA; 16SrRNA amplicon sequencing of fecal DNA of IBD mouse model. NCBI Sequence Read Archive. 2018 https://www.ncbi.nlm.nih.gov/bioproject/PRJNA511382/. Accessed 21 Dec 2018.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.