Fecal microbiota transplantation via capsules relieves ulcerative colitis by improving gut microbiota and serum metabolites

doi:10.21203/rs.3.rs-1606914/v1

Download PDF

Article

Fecal microbiota transplantation via capsules relieves ulcerative colitis by improving gut microbiota and serum metabolites

https://doi.org/10.21203/rs.3.rs-1606914/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

BACKGROUND & AIMS:

Fecal microbiota transplantation (FMT) is a therapy that can induce clinical remission in ulcerative colitis. FMT through enema or colonoscopy brings pain to patients. In this study, we aimed to assess the efficiency of FMT via capsules in UC and determine the specific factors related to the response to clinical remission.

METHODS

We conducted a single-blind trial in Zhongshan Hospital of Xiamen University, China. 21 patients were randomized to accept FMT, and 4 were assigned to the placebo. Patients in the FMT group received capsules filled with healthy donor fecal material and those in the placebo group received capsules filled with glycerol. We performed 16s ribosomal RNA gene sequencing and shotgun metagenomics in fecal samples and analyzed serum samples for metabolome features.

RESULTS

FMT via capsules could receive favorable levels of therapeutic success in UC. 3 FMTs in 1 week induced clinical remission in 57.1% (12 of 21) and clinical response in 76.2% (16 of 21). Gut bacterial richness after FMT was increased in the patients who achieved remission. The increased relative abundance of Alistipes sp, O. splanchnicus, and F. prausnitzii was the strongest predictor of clinical remission. Patients in remission after FMT had enrichment of Alistipes sp, O. splanchnicus and had increased levels of indolelactic acid. Patients who did not achieve remission had enrichment of Escherichia coli and Klebsiella and increased levels of 12,13-DiHOME and lipopolysaccharide biosynthesis.

CONCLUSIONS

We had demonstrated the efficacy of FMT via capsules in active UC. We found that FMT could promote gut microbiome reconstruction, reduce the invasion of harmful bacteria and the synthesis of LPS, and accelerate the metabolism of indoleacetic acid by increasing the abundance of specific bacteria to alleviate intestinal inflammation. These findings may raise a new way in the design of FMT therapy for UC. ClinicalTrails.gov (NCT03426683).

Fecal microbiota transplantation (FMT)

Ulcerative colitis

Capsules

Microbiome

Metabolism

Inflammatory bowel disease (IBD), mainly defined as either ulcerative colitis (UC) or Crohn’s disease (CD), is a chronic and relapsing intestinal inflammation involving various factors and an abnormal immune response[1]. Ulcerative colitis (UC) is a continuous inflammation of the colonic mucosa that usually violates the rectum and gradually spreads to the entire colon. The inflammation caused by an abnormal response from the intestinal mucosal immune system plays an important role in the pathogenesis of UC. The therapy-related options include 5-ASA, corticosteroids, and immunosuppressive agents; these drugs cause only a periodical remission and are likely to result in a relapse[2]. Therefore, new effective treatments need to be explored. Given the pathophysiological role of microbes in inflammatory bowel disease, the manipulation of microbiota for the treatment of IBD could represent an alternative therapeutic approach[3, 4].

For decades, fecal microbiota transplantation (FMT), the administration of a fecal suspension obtained from healthy donors into a patient’s gastrointestinal tract, has been successfully used for bacteriotherapy in CDI (Clostridioides difficile) infections [5]. Recently, an increasing number of subjects have used FMT as a treatment for IBD, especially UC, and FMT has demonstrated variable efficacy against active UC. It appears to be more effective than other treatment methods involving antibiotics, probiotics, and prebiotics[4], and is associated with improved gastrointestinal symptoms, and reduced diarrhea and gut microbiota alterations[6–8]. In a large multi-center, randomized, double-blinded clinical trial in Australia, steroid-free clinical remission was achieved in 32% (12 of 38) of participants receiving low-density donor FMT prepared anaerobically after 1 week of treatment[6]. Another clinical trial involving UC patients reported steroid-free clinical remission in 18 (44%) of 41 patients who underwent colonoscopic infusions once and received enemas 5 days per week for 8 weeks[9]. However, Rossen et al. found out that there was no difference between patients receiving dFMT (donor stool) and aFMT (own stool), using a nasoduodenal infusion of FMT[10]. Multiple modes of FMT are available to clinicians, including colonoscopy, nasoduodenal tube, enemas, and oral capsules[11–15]. A systematic review and meta-analysis of the CDI cure rate data showed that FMT, when performed with colonoscopy, was superior to an enema, and comparable to treatment with capsules[16]. Another randomized clinical trial demonstrated similar results and showed that results obtained with FMT via oral capsules were not inferior to those observed when delivery was achieved by colonoscopy, for preventing recurrent infection over 12 weeks among adults with RCDI[17]. Additionally, the use of fresh or frozen samples may have contributed to these outcomes[18]. There is no doubt that high-quality donor stool processing, and suitable patients and infusion methods are important to successfully performing FMT treatment. Although the efficacy of FMT in UC has been assessed in numerous randomized controlled trials, larger clinical studies are required to conclusively determine its efficacy.

Here, we investigated the longitudinal dynamics of the gut bacteriome in UC patients receiving three doses of FMT in a week. The study aimed to identify the efficiency of FMT via capsule administration and determine the functional factors associated with clinical remission.

Methods

Study Design and Patient Selection

We conducted a single-center, randomized, single-blinded clinical trial by enrolling 25 patients with active UC between June 2016 and June 2019 at Zhongshan Hospital, affiliated with Xiamen University. Of these, 21 were randomized to the FMT group, and 4 were included in the placebo group. All participants were aged 20–68 years, with a total Mayo score of 4–12 points, and provided written informed consent. The total Mayo score, a component representative of the clinical symptoms (stool frequency and rectal bleeding), endoscopic assessment, and physician rating, ranged from 0 to 12. A stable dose of mesalazine was used for patients but not for the control. All the patients were at a stage where the treatment regimen needed to be switched. However, a maintenance dose of mesalazine was required during FMT treatment. Blood was collected for laboratory examination, liver function, ESR, C-reactive protein, and TORCH analyses. Stool was collected for routine and microscopic, C difficile toxin mRNA, CMV DNA, and microbiological analyses. Some exclusion criteria included the detection of bloodstream or gastrointestinal infections, use of antibiotics or probiotics one month prior to enrollment, preparation for pregnancy, and a history of other diseases affecting the gut microbiota, such as metabolic syndromes and autoimmunity.

Donor Selection and Stool Processing

Donors were sought via an advertisement at Xiamen University to undertake a screening questionnaire, including a lifestyle and medical history interview. Potential donors sequentially underwent a medical examination, followed by blood and fecal tests, excluding infectious risks. Nine donors, all students, were selected to provide stool samples, and they underwent medical examination every 3 months. More than 100 g of fresh stool was collected per person in a sterile plastic container. After adding 300 mL of saline to samples, extraction was undertaken using a semi-automatic extraction instrument (Treatgut TG-01, Anjiezhishan Company, China). The supernatant was collected after centrifugation and the precipitate was weighed. Each stool batch consisted of 30 g precipitate, which was resuspended with glycerol. Finally, well-mixed samples were included in capsules (DrCaps,19504907). Each treatment batch of about 32 capsules provided treatment for one participant. On the other hand, placebo capsules only contained glycerol.

Interventions

All patients were administered a dose in the morning. Before taking capsules orally, patients were required to fast overnight for at least 8 h. Treatment course: FMT treatment consisted of a total of one session and was provided 3 times a week. The total weight of stool administered over the 3 FMT procedures was 90 g. Patient stool samples were collected at weeks 0, 4, and 12 for microbiome analysis, and blood samples were collected for laboratory examination. At the baseline (week 0) and week 12, a colonoscopy was needed to assess the colon mucosa. The unblinding process was performed only in case of symptom exacerbation during follow-up. This study was approved by Zhongshan Hospital Affiliated with Xiamen University, and registered with ClinicalTrails.gov (NCT03426683).

Outcomes

Clinical response defined as by a ≥ 3 point reduction in the total Mayo score or a reduction of ≥ 30% at week 12. Clinical remission defined as the total Mayo score ≤ 2 at week 12. Endoscopic remission should result in mucosal healing and a Mayo score ≤ 1.

Sample Collection, DNA Extraction, and Amplicon Sequencing

Fresh fecal samples were collected and immediately frozen in a freezer at -80°C until DNA extraction. According to the manufacturer's protocol, total fecal DNA was extracted from 0.25 g of each of the homogenized samples using the QIAamp PowerFecal DNA Kit (QIAGEN, DE). The DNA yield and quality were then checked with a Multiskan™ GO spectrophotometer (Thermo Fisher Scientific, US).

To identify bacterial communities, 16S rRNA gene V3-V4 amplicon libraries were generated in a 20 µL reaction system containing 10 µL KAPA HiFi HotStart ReadyMix (KAPA Biosystems, USA), 2 µL DNA (60 ng), and 1 µL of each of the barcoded primers (10 µM). The forward and reverse primers used were 5´-CCTACGGGNBGCASCAG-3´ and 5´-GGACTACNVGGGTWTCTAAT-3´, respectively. The cycling conditions were 95 ℃ for 3 min, 30 cycles at 95 ℃ for 20 s, 60 ℃ for 30 s, and 72 ℃ for 30 s, and finally at 72 ℃ for 10 min. In accordance with the AxyPrep™ PCR Cleanup Kit (Axygen, USA) protocol, we purified the PCR product of each sample and assessed their concentrations using Qubit 3.0 (Thermo Fisher Scientific, US). A total of 60 ng of each sample was pooled to produce a sample library, and then sequenced using a 250 bp paired-end sequencing protocol on a HiSeq 2500 platform (Illumina, San Diego, CA, USA) at Xiamen Treatgut Biotechnology Co., Ltd. Raw sequences were deposited in the NCBI Sequence Read Archive under the accession number PRJNA672846.

Serum Sample Collection and Metabolomics

Serum samples were collected and immediately frozen in a freezer at -80°C until metabolite extraction. Total metabolite content was extracted from 0.05 g of each of the serum samples, followed by LC-MS/MS analyses, performed using the UHPLC system (1290, Agilent Technologies) with a UPLC HSS T3 column (2.1 mm × 100 mm, 1.8 µm) coupled to the Q Exactive system (Orbitrap MS, Thermo). MS raw data were filtered using the following criteria: 1) metabolites presenting with less than 50% of all samples in a group were removed; 2) missing values were replaced by half of the minimum value found by default in the dataset. Peaks were annotated against an in-house MS/MS database constructed from HMDB, Metlin, and Mona. Finally, normalization to the relative level of abundance for each metabolite was performed.

Microbial Bioinformatics Analyses

Raw V3-V4 sequencing reads were assembled to high-quality reads using FLASH [19] with parameters such as -M 200 and -x 0.15. Primers were further removed using Cutadapt and the generated reads were chimera-checked and clustered to generate OTUs exhibiting 97% similarity using USEARCH[20]. For bacterial taxonomy, the representative sequences of OTUs were classified against the Silva database [21] using the RDP Classifier at a confidence level of 50%[22]. The OTU table was re-sampled to 19295 reads/sample for downstream analyses. The raw shotgun sequencing reads were filtered using Trimmomatic[23] to remove sequencing adaptors, low-quality and low-complexity reads, and kneaddata was used (https://huttenhower.sph.harvard.edu/kneaddata) to remove contaminants from the human host. The remainder of the reads were classified using Kraken2[24] to determine the taxonomical composition and analyzed using Humann3[25] to perform functional annotation.

Statistical Analyses and Visualization

Microbial α-diversity indexes, including richness (Observed and Chao1), Shannon diversity (Shannon), and Pielou's evenness (Evenness) were computed using the vegan package[26]. Microbial β-diversity was assessed based on the Euclidean distance and significances were tested using PERMANOVA with 9999 permutations using adonis in the vegan package. Metabolomic profiles were analyzed via partial least-squares-discriminant analysis (PLS-DA) using the mixOmics package. The significance of the diversity indexes, individual taxa, and metabolites among groups was determined using a nonparametric Kruskal-Wallis rank-sum test and Benjamini-Hochberg corrections, with the agricolae package[27]. Effect size analysis was performed according to the method described by Wang et al.[28]. Finally, results were visualized using the custom R script, mainly based on R packages ggplot2[29] or VennDiagram[30]. These analyses were performed using R v3.3.2[31].

Gut Bacterial Dysbiosis in UC Patients

To verify gut bacterial dysbiosis in UC patients, fecal samples obtained from 25 UC patients and 25 healthy subjects (HS) were analyzed via 16S V3-V4 sequencing. Bacterial alpha-diversity was measured according to Observed richness (P < 0.05) and Shannon (P < 0.05) indexes; the results showed that it was significantly lower in UC patients than in HS (Fig. 1A), while Pielow’s evenness for UC patients also showed a reducing trend (P = 0.07). Besides, the Venn diagram displayed a large number (1107) of exclusive OTUs in UC, compared to only 439 exclusive OTUs in HS (Fig. 1B). The analysis of beta-diversity via PCA revealed that overall, the gut bacterial communities of patients with UC were clearly (PERMANOVA, F = 4.131, P < 0.001) clustered from the HS group (Fig. 1C). This pattern was significantly associated with a higher abundance of Escherichia_Shigella, Veillonella, Enterobacter, and Collinsella, and a lower level of Alistipes and Akkermansia in UC patients (Figs. 1C and S1C). As expected, the dominant bacterial phylotypes in the intestinal tract of both UC and HS groups belonged to the phyla Bacteroidetes (Bacteroides, Prevotella_9), Firmicutes (Phascolarctobacterium, Veillonella), and Proteobacteria (Escherichia_Shigella), and a smaller proportion belonged to the Actinobacteria and Fusobacteria phyla (Figs. 1D, S1B). The average levels of Proteobacteria and Actinobacteria were significantly higher in UC patients, while that of Bacteroidetes was lower (Fig S1A).

Beneficial Outcomes of FMT

Between June 2016 and June 2019, 25 patients were recruited and assessed for eligibility. Of these, 21 were randomized to accept FMT, and 4 were assigned to the placebo group (Fig. 2A). Baseline characteristics of patients and measures of disease activity showed that there were no significant differences between the 2 treatment groups (Table 1). Participants in the FMT group were followed up for 12 weeks, while patients in the placebo group were withdrawn at 4 weeks after treatment due to worsening colitis (Fig. 2B). The mean total Mayo score of the FMT group significantly (P < 0.001) decreased in the FMT group at week 12 (Fig. 2D), while the partial Mayo score (not able to access the endoscopic subscore) of the placebo group significantly (p < 0.01) increased at the endpoint at week 4. The primary endpoint of clinical remission was achieved in 12 (57.1%) out of 21 patients from the FMT group; of these, 10 (47.6%) achieved endoscopic remission (Fig. 2D, Table 1). Clinical response was observed in up to 76.2% (16/21) of the patients (Fig S2). All the patients in the placebo group failed to respond at week 4 and had to be provided corticosteroids or immunosuppressive agents. Interestingly, some of the patients who received FMT could get relief from clinical symptoms at week 4. Satisfactorily, no serious adverse events occurred after the procedure, though increased bowel movement or capsule shells were observed in the stool of a few patients.

Alterations in Gut Microbiota after FMT

Bacterial diversity was assessed by determining the richness (Observed and Chao1), and Shannon and Pielow’s evenness. Both richness indices were significantly higher in donor samples than in baseline samples of patients allocated to the FMT or placebo groups. The bacterial richness of samples in the FMT group increased significantly and was comparable to that of the donor samples at week 1 after FMT, but reduced at week 4 and week 12. In contrast, no obvious transformations were observed in the placebo group (Fig. 3A). Shannon diversity also increased at weeks 1 and 4 after FMT. Principal component analysis revealed the obvious shifts in overall microbial communities of patients undergoing FMT towards the profiles observed in donor samples (Fig. 3B). Notably, the profiles of patients undergoing FMT changed in such a manner that Prevotellaceae was observed to become dominant at the family level (Fig. 3C). In addition, the enrichment of Prevotella_9 and depletion of Bacteroide and Escherichia_Shigella were observed from the top 25 genera; these alterations were not observed in the placebo group (Fig S3). Compared with the baseline, Prevotella_9, Alloprevotella, and Odoribacter were significantly increased in the FMT group, while Bacteroides, Veillonella, and Enterococcus were significantly decreased (Figs. 3D-E). The shifts in microbial profiles of patients undergoing FMT were more remarkable at the OTU (Fig S4) and species levels, as shown by metagenomics analysis (Fig S5). No obvious alterations in diversity indices, overall communities, or profiles were observed in the placebo group (Fig S4).

Enrichment of Good Bacteria Associated with the Clinical Remission to FMT Accompanied the Decreased Purported Pathobionts

UC patients who achieved clinical remission exhibited efficient and durable changes in their microbiota after FMT. Gut bacterial richness during follow-up after FMT was increased and was comparable with that of donors in the remission group (Rm), while the richness indices at week 4 and week 12 after FMT were significantly (P < 0.05) lower than those of donors in the non-remission group (NRm), though no significant differences were observed between the Rm and NRm groups at each of the time points (Fig. 4A). The increasing trends in alpha diversity after FMT were more remarkable in datasets sequenced via metagenomics (Fig S6). Interestingly, the Shannon and evenness values were marginally (0.05 < P < 0.1) lower, while the richness was relatively higher in Rm than in NRm at week 12, which implies that some taxa from FMT were effectively grafted and maintained their growth advantage in the Rm group. Overall, gut bacterial communities of patients achieving remission were clustered more closely to donor samples since 1 week after FMT, while NRm samples became closer to that of donors until week 12 (Fig. 4B, Fig S6). Compared with the baseline, significant changes in bacterial communities were detected at both week 1 (PERMANOVA, F = 1.7603, P = 0.0384) and week 4 (PERMANOVA, F = 1.926, P = 0.034) in the Rm group, but not in the NRm group in both 16S rRNA and metagenomic datasets (Table S1). More importantly, there were significant differences between Rm and NRm at week 4 (Fig. 4B and Table S1). These results raise the possibility that it could be important to evaluate the microbial changes associated with remission at 4 weeks after FMT.

To identify the taxa associated with clinical remission, the differences in genera between each time point for Rm or NRm after FMT and the baseline were analyzed by the paired rank-sum test. Butyricimonas and genera in Ruminococcacae were exclusively increased at week 1 after FMT in the Rm group (not in the NRm group). More notably, well-known beneficial bacteria (Faecalibacter and Butyricimonas) were exclusively enriched, and harmful taxa (Escherichia_Shigella and Enterobacter) were exclusively decreased only in the Rm group at week 4 (Fig. 4C). Correspondingly, Prevotella_9 and Alloprevotella were specifically increased in the NRm group at week 12. These findings imply that clinical remission after FMT was associated with the eradication of harmful bacteria and enrichment of beneficial bacteria. [Ruminococcus]_gnavus_group was found to be significantly higher in the NRm group than in the Rm group at the baseline, week 1, and week 4 after FMT (Fig S7). Pedioccus, Enterococcus, and Moganella showed similar trends (Fig S7). These taxa might be potential biomarkers used for the prediction of clinical outcomes before FMT.

Shot-gun Metagenomics Confirmed Signatures Associated with Clinical Remission

To further reveal the gut microbial signatures associated with clinical remission, the microbial data of shot-gun metagenomics analyses between W0_Rm and W4_Rm were analyzed using the paired Wilcoxon test, along with that between W0_NRm and W4_NRm. A total of 5466 species were annotated using kraken2 in W0_Rm and W4_Rm, of which the relative abundance of 937 species was significantly different, specifically between W0_Rm and W4_Rm (Table S2). Among the top 20 shifted species, the increased relative abundance of Alistipes (Alistipes sp 3BBH6, 5CBH24, 5CPEGH6, and Alistipes shahii), Odoribacter splanchnicus, and Faecalibacterium prausnitzii and decreased relative abundance of Escherichia coli, Enterobacter hormaechei, and Citrobacter freundii complex sp CFNIH3 were the strongest predictors of clinical remission (Fig. 5A). In terms of microbial functional pathways, a total of 21 KEGG modules (Table S3) and 98 MetaCyc pathways (Table S4) were identified to be specifically associated with clinical remission. Acetoacetate + acetyl-CoA (M00036), gluconeogenesis to fructose-6P (M00003), adenosine ribonucleotide de novo biosynthesis (PWY-7219), and thiamin formation (PWY-7357) were uniquely increased in W4_Rm. Lipopolysaccharide biosynthesis-KDO2-lipid A (M00060), and lipid IVA biosynthesis (NAGLIPASYN-PWY), and the superpathways for N-acetylneuraminate degradation (P441-PWY) were uniquely decreased in W4_Rm, as compared to those in W0_Rm (Figs. 5B and C). Consistently, taxa that mainly contributed to lipopolysaccharide biosynthesis pathways included Escherichia coli and Klebsiella (Table S5), and their abundance was also decreased in the W4_Rm group.

Metabolomic Profiling of Patients Achieving Remission

Global metabolomics analysis of serum samples was performed to reveal metabolic alterations and metabolites associated with FMT and clinical remission. As visualized in the PLS_DA ordination, the metabolomic profiles of patients were shifted after FMT therapy in both positive and negative ion models (Figs. 6A and S8A). In particular, the metabolomic profiles of patients showed a large variation at week 0 and became more similar to each other with time after FMT (Fig. 6A). Overall, 378 metabolites detected in either the positive or the negative ion model were significantly different between the baseline and after FMT (Table S6). Among these, 24 metabolites with variable importance in projection (VIP) score > 1 were structurally identified in databases. These metabolites included L-tryptophan, 12,13-DiHOME, and PGH2-EA (Fig. 6B). There were significant differences in metabolomic profiles between week 0 and week 4 after FMT in both positive (PERMANOVA, F = 2.375, P = 0.047) and negative (PERMANOVA, F = 2.086, P = 0.041) ion models. In terms of clinical remission, metabolomic changes in patients from week 0 to week 4 in the Rm group were relatively larger than the changes from week 0 to week 4 in the NRm group (Figs. 6C and S8B). Specifically, significant differences (P < 0.05, VIP > 1) were observed during the structural identification of 14 metabolites between W0_Rm and W4_Rm (Fig. 6D), and these included increased levels of indolelactic acid, PGH2-EA, and isohyodeoxycholic acid. By contrast, 35 metabolites were significantly different (P < 0.05, VIP > 1) only between W0_NRm and W4_NRm (Fig S9), including L-tryptophan, 12,13-DiHOME, and oleoyl ethyl amide. Furthermore, metabolomic profiles of W4_Rm were significantly (PERMANOVA, F = 3.171, P = 0.024) different from W4_NRm. Notably, a significantly higher abundance of indolelactic acid was observed in W4_Rm, compared to that in W4_NRm (Fig. 6E, Table S7), which was consistent with the differences between W4_Rm and W0_Rm, while a significantly lower relative abundance of 12,13-DiHOME was observed in W4_Rm, compared to that in W4_NRm (Fig. 6E, Table S7), and in W0_NRm, compared to that in W4_NRm (Fig S9).

Inter-omic Correlations between Gut Microbes and Metabolites

Procrustes analysis demonstrated the strong cooperative relationships between the gut microbiota and KEGG modules (M² = 0.5005, P = 0.001, Fig S10A), MetaCyc pathways (M² = 0.44, P = 0.001, Fig S10B), and serum metabolites (M² = 0.939, P = 0.024; Fig. 7A). Permutational multivariate analysis of variance (PERMANOVA)-based effect size analysis was performed to understand how each set of gut microbiomes affected metabolic profiles. The effect sizes of microbial species, MetaCyc pathways, and KEGG modules accounted for 53.7%, 22.3%, 23.3% of metabolome variance in the positive model, and 62.3%, 19.9%, 17.8% of metabolome variance in the negative model, respectively (Fig. 7B). In order to couple the alterations in the gut microbiome and serum metabolism that were associated with clinical remission, the Pearson correlation values between the top 20 gut microbial species (Fig. 5A), MetaCyc pathways (Fig. 5B), KEGG modules (Fig. 5C), and serum metabolites (Fig. 6D) that were uniquely significantly different between W0_Rm and W4_Rm were determined. A total of 256 significant (P < 0.05) shotgun-metagenomics-based gut microbial features and metabolite correlations were identified (Fig. 7C), while only 10 significant correlations were identified between genera from the 16S V4 dataset and metabolites (Fig S11). Four major clusters of correlations were visualized via network analysis, one of which was characterized by increased levels of beneficial metabolites (POS133_PGH2-EA, NEG151_LysoPA(18:1(9Z)/0:0)) and microbial species (Alistipes sp., Odoribacter splanchnicus, etc.), while another was characterized by decreased levels of metabolites (POS490_Pentacyclo[9.7.0.01,3.03,8.012,16]octadec-5-en-4-one, POS18 0_Phe-Asp-Trp-Asn, etc.), harmful species (Escherichia coli, Klebsiella sp. PO552, Salmonella enterica, etc.), and functional pathways (PWY-5347: superpathway of L-methionine biosynthesis (transsulfuration), FASYN-INITIAL-PWY: superpathway of fatty acid biosynthesis initiation (E. coli), M00348: glutathione transport system, M00227: glutamine transport system). The correlation of positions between Faecalibacterium prausnitzii, NEG101_Isohyodeoxycholic acid, and POS329_ PC(17:0/17:2(9Z,12Z)) was observed in both metagenomics and 16S V4 datasets. Notably, we observed a significant decrease in the KEGG module M00060 and the pathway of KDO2-lipid A biosynthesis from UDP-N-acetylglucosamine to LPS in patients achieving remission (W4_Rm), compared to that observed with W0_Rm. This module consisted of several genes (K00677, K02536, K03269, K00748, K02560, etc.), and the expression of most of these genes were significantly decreased in W4_Rm, as compared to W0_Rm and W4_NRm. Notably, the abundance of strains that were stratified to encode these genes by Humann3 was decreased in the W4_Rm group as well, and these included Escherichia coli, Klebsiella pneumoniae, Haemophilus parainfluenzae (Figs. 5A and 7D). These results suggest that the depletion of strains such as E. coli and Klebsiella after FMT may contribute to the lowered production of LPS and better therapeutic effects in the remission group. Moreover, the concentration of indolelactic acid was higher in W4_Rm than in W0_Rm and W4_NRm (Fig. 6E), which was found to be consistent with the higher abundance of the tryptophanase gene (K01167) and the taxa Alistipes sp. 3BBH6 and Odoribacter splanchnicus, which convert tryptophan to indolelactic acid.

In this study, we explored the clinical effect of low-intensity FMTs by administering capsules containing fecal material obtained from donors to patients with active UC and conducted comprehensive analyses of gut microbiota and serum metabolites, to identify specific taxonomic, functional, and metabolite changes that were associated with therapeutic success. A total of 3 FMTs in 1 week induced clinical remission in 57.1% (12 of 21) and clinical response in 76.2% (16 of 21) of patients at week 12 after the last FMT, while the Mayo scores of all patients in the placebo group increased at week 4 after the last FMT. Gut microbiomes and serum metabolites of patients were significantly altered by FMT, and it might be suitable to evaluate the changes associated with clinical outcomes at 4 weeks after FMT. Improvements in microbial richness, depletion of harmful microbes (Escherichia coli, Enterobacter hormaechei, Salmonella enterica), pathways (LPS biosynthesis), and pro-inflammatory metabolites (12,13-DiHOME), enrichment of beneficial microbes (Faecalibacterium prausnitzii, Alistipes sp), pathways (acetyl-CoA biosynthesis), and anti-inflammatory metabolites (indolelactic acid) as well as their interactions were identified to be associated with the therapeutic effects of FMT.

The key features distinguishing this study from previous studies on FMT in UC patients were the low intensity of microbiota transplantation using capsules and high efficiency of clinical remission. Previously, colonic transendoscopic tubes or enemas were the main methods used for the delivery of fecal microbiota, and the treatment cycle lasted for a relatively long period. For example, Paramsothy et al. first achieved a single colonoscopic delivery of FMT to the right colon and then performed enemas 5 days per week for 8 weeks[9]. Samuel et al. demonstrated the efficacy of FMT by performing enemas twice a week[6] and Zhang et al. used a single fresh FMT through the mid-gut[32]. Pai et al. reported on a colonoscopic infusion at the baseline, followed by the oral administration of capsules two times per week for 6 weeks in patients with ages between 3 and 17 years[33]. All invasive operations were inevitably risky and resulted in more costs to patients[34]. In our study, all participants received FMT 3 times via capsules in one week, with no serious side effects, and reported that it was more comfortable than a colonoscopy or enema. The clinical remission rate was noteworthy, and higher than values reported previously (37.0% [95% CI 28.8–45.9]) in quasi-experimental or RCT studies[35]. The results showed that the abdominal pain score, diarrhea score, bloody stool score, intestinal mucosal lesions, and Mayo score of patients decreased significantly after FMT. Generally, if clinical symptoms were relieved in 4 weeks after FMT, most of the patients could be in a stationary phase. The longer the time needed to obtain relief from symptoms, the worse the prognosis. We also observed remarkable and specific differences in gut microbiota and metabolites between the Rm group and NRm group at week 4 after FMT. Notably, previous studies also investigated changes in the microbiota at 4 weeks after FMT, and results show consistent trends[6, 36, 37]. We could probably assess whether patients would benefit from FMT at 4 weeks after FMT. If not, another therapy needs to be provided to such patients as soon as possible.

We observed that microbial richness was significantly lower in patients with UC than in healthy controls or donors, and was improved to levels comparable to those of donors only in patients who achieved clinical remission at week 12 after FMT. Improvement in gut microbial richness after FMT has been reported in previous clinical studies on UC[6, 36, 38], and was suggested as an important indication for success in FMT[6, 39–41]. Moreover, the higher gut microbial richness of donors was reported to be associated with remission in UC patients[42]. Overall, the bacterial community structure was shifted to be more closed to donor samples. Specifically, levels of Prevotella_9 were notably increased to be equal to those of the dominant genus after FMT in our 16S rRNA gene dataset, accompanied by the decrease in levels of Bacteroides. Since both of these genera belong to Bacteroidales, it is unclear why this replacement occurred[43], but the enrichment of Preovella_9 after FMT was frequently reported in other studies on FMT[36, 44]. Other important changes after FMT in the bacterial profile included the eradication of Veillonella and Enterococcus, which are well-known harmful bacteria, and the enrichment of Alloprevotella and Odoribacter, which can produce short-chain fatty acids[45, 46] that benefit human health.

Microbes and microbial functional pathways associated with clinical remission were further analyzed based on 16S rRNA gene and metagenomic datasets. Compared with the baseline, the increase in the relative abundance of many microbes, including Faecalibacterim (F. prausnitzii), Butyricimonas, Alistipes (A. sp 3BBH6, 5CBH24, 5CPEGH6, and A. shahii), O. splanchnicus, and Christensenellaceae R-7, was uniquely associated with clinical remission after FMT. Among these, Faecalibacterim, Butyricimonas, and Alistipes are well-known SCFA-producing bacteria[45] and anti-inflammatory commensal bacteria that play a role in IBD[47], obesity[48], appendectomy[49], and depression[50]. Interestingly, a recently published clinical study reported that O. splanchnicus was a key strain that could promote both metabolism (producing SCFA) and mucosal immunity (inducing regulatory T cells), and has protective effects against ulcerative colitis[51]. The family Christensenellaceae, reported to be widespread and associated with human health, was depleted in individuals with CD, UC, and IBS[52]. A decrease in certain taxa, including Escherichia_Shigella (E. coli), Enterobacter (E. hormaechei), and Citrobacter freundii was also exclusively linked with clinical remission; this was consistent with the results of previous reports, which showed that they acted as pathogens that caused inflammatory bowel disease, diarrhea, and other infections[53–55]. Microbial functions, including acetoacetate + acetyl-CoA, gluconeogenesis to fructose-6P, and thiamin formation, were specifically improved. Importantly, the relative abundance of lipopolysaccharide biosynthesis-KDO2-lipid A, lipid IVA biosynthesis, and the superpathways for N-acetylneuraminate degradation were exclusively decreased during clinical remission after FMT, which coincided with the changes in the levels of Escherichia coli, Klebsiella, and Salmonella enterica, which contribute to the formation of LPS[56, 57]. As a major endotoxin that strongly stimulates host innate immunity, LPS from enteric opportunistic bacteria has been implicated in IBD, autoimmune diseases, and metabolic disorders[58, 59]; therefore, the prevention of the accumulation of LPS provides protection against the development of severe colitis[57].

Serum metabolite profiles were significantly different between the baseline and week 4 after FMT. Among these profiles, elevated levels of indolelactic acid, PGH2-EA, and isohyodeoxycholic acid were associated with a favorable treatment output, while an increased level of 12,13-DiHOME was associated with worse outcomes. PGH2-EA, a prostaglandin, showed anti-inflammatory properties in ulcerative colitis[60]. The 12,13-diHOME produced by bacterial epoxide hydrolase genes altered the expression of PPARγ-regulated genes, reduced anti-inflammatory cytokine secretion and the number of regulatory T cells, increased inflammation, impeded immune tolerance, and was involved in the development of atopy, eczema, or asthma[61]. The elevation in the levels of tryptophan biosynthesis by FMT might be another important factor contributing to therapeutic success, and this tryptophan is further metabolized by gut microbes into indolelactic acid, indole acetic acid and indoxyl sulfuric acid, in patients achieving clinical remission[62]. These metabolites can act as signals that activate the innate immunity of intestinal mucosa and induce a rapid inflammatory response[63]. Indolelactic acid modulates ex vivo immune responses of human CD4 + T cells and monocytes in a dose-dependent manner by acting as an agonist of both the AhR and hydroxycarboxylic acid receptor 3[64].

In general, we demonstrated the efficacy of FMT in patients with mild to severe UC through well-prepared donor stool capsules administered orally three times a week. FMT can accelerate the metabolism of tryptophan by increasing the abundance of Alipipes sp. 3BBH6, Odoribacter splanchnicus, and other bacteria, and enable the production of indole lactic acid and other products that alleviate intestinal inflammation.

Funding

This project was supported by the National Natural Science Foundation of China (81770558 and 81800517). This project was also sponsored by the Xiamen Key Programs of Medical Health (3502Z20204007), Science and Technology Planning Project of Fujian Province (2019J01554), Fujian Provincial Natural Science Foundation (2020J05286 and 2021J011329) and the Xiamen Priority Programs of Medical Health (3502Z20199172 and 3502Z20209026).

Acknowledgments

The authors thank lab members and clinicians of the department of gastroenterology in Zhongshan Hospital of Xiamen University for thoughtful comments on the manuscript and for helping to manage the patients. We also thank TopEdit (www.topeditsci.com) for its linguistic assistance during the preparation of this manuscript.

Loftus, E.V., Jr., Clinical epidemiology of inflammatory bowel disease: Incidence, prevalence, and environmental influences. Gastroenterology, 2004. 126(6): p. 1504–17.
Jeong, D.Y., et al., Induction and maintenance treatment of inflammatory bowel disease: A comprehensive review. Autoimmun Rev, 2019. 18(5): p. 439–454.
Kostic, A.D., R.J. Xavier, and D. Gevers, The microbiome in inflammatory bowel disease: current status and the future ahead. Gastroenterology, 2014. 146(6): p. 1489–99.
Shanahan, F. and E.M. Quigley, Manipulation of the microbiota for treatment of IBS and IBD-challenges and controversies. Gastroenterology, 2014. 146(6): p. 1554–63.
van Nood, E., et al., Duodenal infusion of donor feces for recurrent Clostridium difficile. N Engl J Med, 2013. 368(5): p. 407–15.
Costello, S.P., et al., Effect of Fecal Microbiota Transplantation on 8-Week Remission in Patients With Ulcerative Colitis: A Randomized Clinical Trial. JAMA, 2019. 321(2): p. 156–164.
Fuentes, S., et al., Microbial shifts and signatures of long-term remission in ulcerative colitis after faecal microbiota transplantation. ISME J, 2017. 11(8): p. 1877–1889.
Moayyedi, P., et al., Fecal Microbiota Transplantation Induces Remission in Patients With Active Ulcerative Colitis in a Randomized Controlled Trial. Gastroenterology, 2015. 149(1): p. 102–109 e6.
Paramsothy, S., et al., Multidonor intensive faecal microbiota transplantation for active ulcerative colitis: a randomised placebo-controlled trial. Lancet, 2017. 389(10075): p. 1218–1228.
Rossen, N.G., et al., Findings From a Randomized Controlled Trial of Fecal Transplantation for Patients With Ulcerative Colitis. Gastroenterology, 2015. 149(1): p. 110–118 e4.
Steube, A., et al., Long-term Multidonor Faecal Microbiota Transfer by Oral Capsules for Active Ulcerative Colitis. J Crohns Colitis, 2019. 13(11): p. 1480–1481.
Hirsch, B.E., et al., Effectiveness of fecal-derived microbiota transfer using orally administered capsules for recurrent Clostridium difficile infection. BMC Infect Dis, 2015. 15: p. 191.
Eiseman, B., et al., Fecal enema as an adjunct in the treatment of pseudomembranous enterocolitis. Surgery, 1958. 44(5): p. 854–9.
Cold, F., et al., Multidonor FMT capsules improve symptoms and decrease fecal calprotectin in ulcerative colitis patients while treated - an open-label pilot study. Scand J Gastroenterol, 2019. 54(3): p. 289–296.
Brandt, L.J., et al., Long-term follow-up of colonoscopic fecal microbiota transplant for recurrent Clostridium difficile infection. Am J Gastroenterol, 2012. 107(7): p. 1079–87.
Ramai, D., et al., Fecal Microbiota Transplantation (FMT) with Colonoscopy Is Superior to Enema and Nasogastric Tube While Comparable to Capsule for the Treatment of Recurrent Clostridioides difficile Infection: A Systematic Review and Meta-Analysis. Dig Dis Sci, 2021. 66(2): p. 369–380.
Kao, D., et al., Effect of Oral Capsule- vs Colonoscopy-Delivered Fecal Microbiota Transplantation on Recurrent Clostridium difficile Infection: A Randomized Clinical Trial. JAMA, 2017. 318(20): p. 1985–1993.
Jiang, Z.D., et al., Randomised clinical trial: faecal microbiota transplantation for recurrent Clostridum difficile infection - fresh, or frozen, or lyophilised microbiota from a small pool of healthy donors delivered by colonoscopy. Aliment Pharmacol Ther, 2017. 45(7): p. 899–908.
Magoč, T. and S.L. Salzberg, FLASH: fast length adjustment of short reads to improve genome assemblies. Bioinformatics (Oxford, England), 2011. 27(21): p. 2957–2963.
Edgar, R.C., UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nat Methods, 2013. 10(10): p. 996–8.
Henderson, G., et al., Improved taxonomic assignment of rumen bacterial 16S rRNA sequences using a revised SILVA taxonomic framework. PeerJ, 2019. 7: p. e6496.
Wang, Q., et al., Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Applied and Environmental Microbiology, 2007. 73(16): p. 5261–5267.
Bolger, A.M., M. Lohse, and B. Usadel, Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics, 2014. 30(15): p. 2114–2120.
Wood, D.E., J. Lu, and B. Langmead, Improved metagenomic analysis with Kraken 2. Genome Biology, 2019. 20(1): p. 257.
Franzosa, E.A., et al., Species-level functional profiling of metagenomes and metatranscriptomes. Nature methods, 2018. 15(11): p. 962–968.
Oksanen, J., et al., vegan: community ecology package. R package, 2015.
Mendiburu, F.d., agricolae: Statistical Procedures for Agricultural Research. 2015.
Wang, X., et al., Aberrant gut microbiota alters host metabolome and impacts renal failure in humans and rodents. Gut, 2020. 69(12): p. 2131–2142.
Wickham, H., W. Chang, and RStudio, ggplot2: Create elegant data visualisations using the grammar of graphics. 2017, New York: Springer.
Chen, H., VennDiagram: Generate high-resolution Venn and Euler plots. 2014.
R Core Team, R: a language and environment for statistical computing. 2017, Vienna, Austria: R Foundation for Statistical Computing.
Zhang, T., et al., Short-Term Surveillance of Cytokines and C-Reactive Protein Cannot Predict Efficacy of Fecal Microbiota Transplantation for Ulcerative Colitis. PLoS One, 2016. 11(6): p. e0158227.
Pai, N., et al., Protocol for a double-blind, randomised, placebo-controlled pilot study for assessing the feasibility and efficacy of faecal microbiota transplant in a paediatric Crohn's disease population: PediCRaFT Trial. BMJ Open, 2019. 9(11): p. e030120.
Kaplan, G.G., The global burden of IBD: from 2015 to 2025. Nat Rev Gastroenterol Hepatol, 2015. 12(12): p. 720–7.
Caldeira, L.d.F., et al., Fecal microbiota transplantation in inflammatory bowel disease patients: A systematic review and meta-analysis. PLoS One, 2020. 15(9): p. e0238910.
Paramsothy, S., et al., Multidonor intensive faecal microbiota transplantation for active ulcerative colitis: a randomised placebo-controlled trial. Lancet, 2017. 389(10075): p. 1218–1228.
Chu, N.D., et al., Dynamic colonization of microbes and their functions after fecal microbiota transplantation for inflammatory bowel disease. mBio, 2021. 12(4): p. e00975-21.
Matsuoka, K., Fecal microbiota transplantation for ulcerative colitis. Immunological medicine, 2020: p. 1–5.
Papanicolas, L.E., et al., Improving risk-benefit in faecal transplantation through microbiome screening. Trends Microbiol, 2020. 28(5): p. 331–339.
Ng, S.C., et al., Scientific frontiers in faecal microbiota transplantation: joint document of Asia-Pacific Association of Gastroenterology (APAGE) and Asia-Pacific Society for Digestive Endoscopy (APSDE). Gut, 2020. 69(1): p. 83–91.
Vermeire, S., et al., Donor species richness determines faecal microbiota transplantation success in inflammatory bowel disease. J Crohns Colitis, 2016. 10(4): p. 387–394.
Kump, P., et al., The taxonomic composition of the donor intestinal microbiota is a major factor influencing the efficacy of faecal microbiota transplantation in therapy refractory ulcerative colitis. Aliment Pharm Ther, 2018. 47(1): p. 67–77.
Ley, R.E., Prevotella in the gut: choose carefully. Nature Reviews Gastroenterology & Hepatology, 2016. 13(2): p. 69–70.
Zhu, F., et al., Effects of Different Treatment of Fecal Microbiota Transplantation Techniques on Treatment of Ulcerative Colitis in Rats. Frontiers in Microbiology, 2021. 12(1833).
Vital, M., A. Karch, and D.H. Pieper, Colonic butyrate-producing communities in humans: an overview using omics data. mSystems, 2017. 2(6): p. e00130-17.
Zagato, E., et al., Endogenous murine microbiota member Faecalibaculum rodentium and its human homologue protect from intestinal tumour growth. Nature microbiology, 2020. 5(3): p. 511–524.
Schirmer, M., et al., Dynamics of metatranscription in the inflammatory bowel disease gut microbiome. Nature microbiology, 2018. 3(3): p. 337–346.
Barathikannan, K., et al., Gut Microbiome Modulation Based on Probiotic Application for Anti-Obesity: A Review on Efficacy and Validation. Microorganisms, 2019. 7(10): p. 456.
Cai, S., et al., Appendectomy is associated with alteration of human gut bacterial and fungal communities. Frontiers in Microbiology, 2021. 12(2624): p. 724980.
Yang, C., et al., Possible role of the gut microbiota–brain axis in the antidepressant effects of (R)-ketamine in a social defeat stress model. Translational Psychiatry, 2017. 7(12): p. 1294.
Lima, S., et al., Transferable Immunoglobulin A–Coated Odoribacter splanchnicus in Responders to Fecal Microbiota Transplantation for Ulcerative Colitis Limits Colonic Inflammation. Gastroenterology, 2021.
Waters, J.L. and R.E. Ley, The human gut bacteria Christensenellaceae are widespread, heritable, and associated with health. BMC biology, 2019. 17(1): p. 83–83.
Darfeuille-Michaud, A., et al., High prevalence of adherent-invasive Escherichia coli associated with ileal mucosa in Crohn’s disease. Gastroenterology, 2004. 127(2): p. 412–421.
Kitamoto, S., et al., The Intermucosal Connection between the Mouth and Gut in Commensal Pathobiont-Driven Colitis. Cell, 2020. 182(2): p. 447–462.e14.
Liu, L., et al., Genetic Diversity, Multidrug Resistance, and Virulence of Citrobacter freundii From Diarrheal Patients and Healthy Individuals. Frontiers in Cellular and Infection Microbiology, 2018. 8(233).
Rezania, S., et al., Extraction, Purification and Characterization of Lipopolysaccharide from Escherichia coli and Salmonella typhi. Avicenna journal of medical biotechnology, 2011. 3(1): p. 3–9.
Yang, W.H., et al., Neu3 neuraminidase induction triggers intestinal inflammation and colitis in a model of recurrent human food-poisoning. Proceedings of the National Academy of Sciences, 2021. 118(29): p. e2100937118.
Whitfield, C., D.M. Williams, and S.D. Kelly, Lipopolysaccharide O-antigens—bacterial glycans made to measure. Journal of Biological Chemistry, 2020. 295(31): p. 10593–10609.
Vila, A.V., et al., Gut microbiota composition and functional changes in inflammatory bowel disease and irritable bowel syndrome. Science Translational Medicine, 2018. 10(472): p. eaap8914.
Decara, J., et al., Peroxisome Proliferator-Activated Receptors: Experimental Targeting for the Treatment of Inflammatory Bowel Diseases. Frontiers in pharmacology, 2020. 11: p. 730–730.
Levan, S.R., et al., Elevated faecal 12,13-diHOME concentration in neonates at high risk for asthma is produced by gut bacteria and impedes immune tolerance. Nature microbiology, 2019. 4(11): p. 1851–1861.
Zhang, J., et al., Metabolites of microbiota response to tryptophan and intestinal mucosal immunity: A therapeutic target to control intestinal inflammation. Medicinal Research Reviews, 2021. 41(2): p. 1061–1088.
Dodd, D., et al., A gut bacterial pathway metabolizes aromatic amino acids into nine circulating metabolites. Nature, 2017. 551(7682): p. 648–652.
Laursen, M.F., et al., Bifidobacterium species associated with breastfeeding produce aromatic lactic acids in the infant gut. Nature microbiology, 2021. 6(11): p. 1367–1382.

Table 1 is available in the Supplemental Files section.

(Not answered)

TableS1.xlsx
Table S1
TableS2.xlsx
Table S2
TableS3.xlsx
Table S3
TableS4.xlsx
Table S4
TableS5.xlsx
Table S5
TableS6.xlsx
Table S6
TableS7.xlsx
Table S7
Table1.pdf
Table 1
SupplementaryMaterials.pdf
Supplementary Materials

Download PDF

Version 1

posted

You are reading this latest preprint version

Fecal microbiota transplantation via capsules relieves ulcerative colitis by improving gut microbiota and serum metabolites

Status:

Version 1

Abstract

BACKGROUND & AIMS:

METHODS

RESULTS

CONCLUSIONS

Figures

Introduction

Methods

Study Design and Patient Selection

Donor Selection and Stool Processing

Interventions

Outcomes

Sample Collection, DNA Extraction, and Amplicon Sequencing

Serum Sample Collection and Metabolomics

Microbial Bioinformatics Analyses

Statistical Analyses and Visualization

Results

Gut Bacterial Dysbiosis in UC Patients

Beneficial Outcomes of FMT

Alterations in Gut Microbiota after FMT

Shot-gun Metagenomics Confirmed Signatures Associated with Clinical Remission

Metabolomic Profiling of Patients Achieving Remission

Inter-omic Correlations between Gut Microbes and Metabolites

Discussion

Declarations

Funding

Acknowledgments

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1