*3.3. Impact of FMT on Gene Enrichment and Correlations with Secondary Bile Acid Production*

As some of the bacterial species enriched by FMT are implicated in bile acid metabolism, we next determined the effect of FMT on bacterial bile acid metabolic gene copy number. Starting with a broad overview of the impact of FMT on gut microbial gene abundance, we performed pathway analysis using *omePath* of the metagenomic data (Figure 3A). FMTenriched genes involved in cell proteolysis pathways. Taking a closer look at bile acid metabolism, FMT did not impact gene abundance for most known gut bacterial bile acid metabolic genes, except for a reduction in *BaiB* and *BaiE*. These changes were noted at 4 weeks and not at baseline in the FMT group compared to the placebo. FMT did not impact the gene abundance of other genes involved in bile acid 7-α-dehydroxylation, including *BaiCD*, *BaiA2*, *BaiF*, and *BaiH*. Further work using metatranscriptomics is warranted to determine the impact of FMT on bacterial bile acid metabolic gene expression.

To identify candidate bacteria involved in gut bacterial bile acid metabolism, we assessed correlations between bacterial species abundance and bile acid profile, with a focus on bile acid products of gut bacterial metabolism, namely unconjugated bile acids, and the secondary bile acids, DCA and LCA. Bile acid levels were measured in the same samples used for metagenomics analysis, as previously described [10]. The impact of FMT on bile acid levels in this sample set has been previously reported [10]. We focused on bacterial species that were positively correlated with bile acid sub-types that are produced, at least in part, through interactions with the gut microbiota with a *p*-value less than or equal to 0.08. Bacterial species that met these criteria are presented in Figure 3B. *Phocaeicola dorei* and *Bacteroides ovatus* were positively correlated with unconjugated chenodeoxycholic acid (CDCA) (*<sup>p</sup>* = 7.47 × <sup>10</sup>−<sup>45</sup> and 2.50 × <sup>10</sup><sup>−</sup>8). *Bifidobacterium adolescentis* and *Collinsella aerofaciens* were positively correlated with the production of DCA (specifically, the glycine-conjugated form, GDCA) (*p* = 0.0123 and 0.0634). The same strain of *B. ovatus* was positively correlated with unconjugated cholic acid (CA) (*p* = 0.0317). Lastly, *Faecalibacterium prausnitzii* was positively correlated with LCA (*p* = 0.0634). These data point to a potential role for *Phocaeicola dorei* and *B. ovatus* in bile acid deconjugation. Further, these data suggest that *Bif. adolescentis*, *C. aerofaciens,* and *F. prausnitzii* may play a role in the conversion of primary to secondary bile acids, which requires further functional validation.

**Figure 2.** Impact of FMT on gut microbial composition. (**A**) Relative abundance of each bacterial phylum. Relative abundance of *Paraprevotella* (**B**) and *Longibaculum* (**C**), *Clostridium hylemonae* (**D**), and *Desulfovibrio fairfieldensis* (**E**) in fecal samples from placebo vs. FMT at baseline and after 4 weeks of treatment. Data presented as mean ± SEM. \* *p* < 0.05.

**Figure 3.** Association between bile acids and bacterial species. (**A**) Assessment of pathways enriched by FMT relative to baseline (**left**) and relative to placebo control after 4 weeks of treatment (**right**). (**B**) Bacterial species and strains that are correlated with gut bacterial-derived bile acids. \* *p* < 0.2, \*\* *p* < 0.05, \*\*\* *p* < 0.001.

### **4. Discussion**

In the present study, we utilized a fecal sample set from patients receiving FMT or placebo that exhibited alterations in gut bacterial bile acid metabolism to improve our understanding of the gut bacterial species involved in gut bacterial bile acid metabolism and how these pathways are dynamically regulated by FMT. Using metagenomics, we identified an enrichment of *Paraprevotella*, *Longibaculum*, *Desulfovibrio fairfieldensis*, and *Clostridium hylemonae* in response to FMT. Furthermore, through the assessment of correlations between fecal bile acid levels and bacterial species relative abundances, we identified *Bifidobacterium adolescentis*, *Bacteroides ovatus*, *Faecalibacterium prausnitzi*, and *Phocaeicola dorei* as potentially contributing to gut bacterial bile acid metabolism. Further work is needed to better understand secondary bile acid metabolism, its roles in metabolic disease, and how it can be manipulated through FMT.

The effect of FMT to enrich *Paraprevotella, Longibaculum*, *C. hylemonae*, and *D. fairfieldensis* may have contributed to the effect of FMT to enhance gut microbial bile acid metabolism and/or slow the development of glucose intolerance. For example, members of the genera *Clostridium* are the predominant human intestinal species thought to perform 7-α-dehydroxylation of primary bile acids [25]. Additionally, *C. hylemonae* has been shown to convert CA into DCA in vitro [26]. Furthermore, *Paraprevotella* abundance was significantly increased after FMT in individuals with functional constipation whose changes in fecal microbiome compositions were measured before and after FMT. This increase in *Paraprevotella* abundance correlated with improved relief of clinical symptoms measured by three different clinical scales for constipation, suggesting *Paraprevotella* could improve metabolic dysregulation through gastric motility [27]. The role of *Longibaculum* in host metabolic health is poorly defined; however, dietary fiber supplementation has been shown to enrich for *Longibaculum* [28]. *D. fairfieldensis* is a Gram-negative anaerobic bacillus that has been implicated in bile acid metabolism. Interestingly, *D. fairfieldensis* bacteremia was found to be associated with choledocholithiasis in a case report [29]. Furthermore, a recent study reports that *Desulfovibrionales* are enriched in patients with cholelithiasis. Further, the administration of *Desulfovibrionales* to mice with antibiotic-induced depletion of the gut microbiome increased secondary bile acid production [30]. *Desulfovibrionales* can reduce

taurine into H2S, which has been suggested to facilitate 7-α-dehydroxylation [31]. Together, these data suggest that *Desulfovibrionales*, and in particular *D. fairfieldensis*, may play a role in gut bile acid metabolism. Further, these data highlight the potentially important cooperative interactions among bacteria that facilitate gut microbial bile acid metabolism.

In this study, we identified five bacteria that were positively correlated with gut microbiome-derived bile acids. Specifically, *Bifidobacterium adolescentis* and *Collinsella aerofaciens* were positively correlated with DCA. *Bacteroides ovatus* was positively correlated with unconjugated CDCA and CA. *Phocaeicola dorei* was positively correlated with unconjugated CDCA, and *Faecalibacterium prausnitzii* was positively correlated with LCA. Thus, our data suggest that *Phocaeicola dorei* and *Bacteroides ovatus* may perform bile acid deconjugation. Consistent with this, previous work reports that several *Bacteroides* strains, including strains of *B. ovatus*, express BSH [32]. Whether *Phocaeicola dorei* can perform bile acid deconjugation is unknown and requires further testing. Interestingly, a previous study identified a correlation between *Phocaeicola dorei*, also named *Bacteroides dorei*, and the risk of developing type 1 diabetes [33], suggesting a potential metabolic role for this species. The bacteria that were positively correlated with secondary bile acids (*Bifidobacterium adolescentis* and *Faecalibacterium prausnitzii*) are known to have BSH functions [34,35]. *Faecalibacterium prausnitzii* has been connected to anti-inflammatory effects and improvement of intestinal barrier function [36,37]. A role for *Collinsella aerofaciens* in the production of DCA has not been previously tested. However, *Collinsella aerofaciens,* previously known as *Eubacterium aerofaciens*, was found to have NADP-dependent 7-β-hydroxysteriod dehydrogenase activity, which is necessary for the production of hydrophilic secondary bile acids such as ursodeoxycholic acid [38].

The advantages of this study include the application of metagenomics to the analysis of the gut microbiome of individuals receiving FMT or placebo control. Additionally, the results from this secondary analysis are from individuals with obesity such that bacteria identified from this specific population can better inform FMT for the treatment of obesity and metabolic disease. Limitations of this study include the small sample size. While bile acid gene abundance was studied, metatranscriptomics analysis is needed to assess the impact of FMT on gene expression. Further work is needed to functionally validate bacteria identified as potentially contributing to the effects of FMT on gut bacterial bile acid metabolism. Together, these data demonstrate that FMT can alter the composition of bile acids and bacterial communities in the gut microbiome.

**Author Contributions:** Investigation: J.-M.B., T.D., C.L., Z.M., J.R.M., K.A.C., A.R., B.H.M., J.R.A. and B.P.C. Conceptualization, Resources: J.R.A. and B.P.C. Methodology: J.-M.B., T.D., C.L., Z.M., J.R.M., K.A.C., A.R., B.H.M., J.R.A. and B.P.C. Formal Analysis: J.-M.B., T.D., C.L., Z.M., J.R.M., K.A.C., A.R. and B.P.C. Writing—Original Draft Preparation: J.-M.B. Writing—Review and Editing: J.-M.B., T.D., C.L., Z.M., J.R.M., C.C.T., K.A.C., A.R., B.H.M., J.R.A. and B.P.C. Supervision, Funding Acquisition: K.A.C., A.R. and B.P.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** Research reported in this publication was supported by the National Center for Complementary and Integrative Health of the National Institutes of Health under Award number R21AT010956 to B.P.C. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Metabolomics studies were performed at the MRC-NIHR National Phenome Centre at Imperial College London; this center receives financial support from the Medical Research Council (MRC) and National Institute of Health Research (NIHR) (grant number MC\_PC\_12025). B.H.M. is the recipient of an NIHR Academic Clinical Lectureship (CL-2019-21-002) and was formerly in receipt of an MRC Clinical Research Training Fellowship (MR/R000875/1). The Division of Digestive Diseases and MRC-NIHR National Phenome Centre at Imperial College London receive financial and infrastructure support from the NIHR Imperial Biomedical Research Centre (BRC) based at Imperial College Healthcare NHS Trust and Imperial College London.

**Institutional Review Board Statement:** The study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Institutional Review Board at the Brigham and Women's Hospital, and all patients provided written informed consent before participation (NCT02741518). In addition, Food and Drug Administration approval via an investigational new drug application (16936, 2016) was obtained. All authors had access to the study data and approved the final manuscript.

**Informed Consent Statement:** All patients provided written informed consent before participation.

**Data Availability Statement:** Sequence data generated by this project are deposited in NCBI Sequence Read Archive (SRA) and associated with BioProject PRJNA904790.

**Acknowledgments:** We thank Castle Raley and the George Washington Genomics Core for sample processing and sequencing for metagenomics.

**Conflicts of Interest:** J.-M.B., C.L., Z.M., J.R.M., K.A.C., A.R. and B.P.C. have no relevant conflict of interest to declare. J.R.A. consults for and has research support from Finch Therapeutics Group, Janssen, Pfizer, Abbvie, Iterative Sopes, Seres Therapeutics, Ferring, Merck, Bristol Myer Squibb and has research support from Pfizer and Merck. T.D. has research support from AMPEL BioSolutions. B.H.M. has received consultancy fees from Finch Therapeutics Group and Ferring Pharmaceuticals. C.C.T. consults for Apollo Endosurgery, Boston Scientific, Medtronic, Enterasense Ltd., EnVision Endoscopy, Fractyl, Fujifilm, GI Dynamics, GI Windows, Lumendi, Olympus, USGI Medical, Xenter, Endoquest Robotics. He has received Research Support from Apollo Endosurgery, Boston Scientific, ERBE, Fujifilm, GI Dynamics, Lumedi Olympus, USGI Medical. He serves on Advisory Boards for Fractyl, Fujifilm, USGI Medical, Xenter and Endoquest Robotics. He is a founder, board member and receives ownership interest from Enterasense Ltd., EnVision Endoscopy and GI Windows, He is on a speakers bureau for Boston Scientific, Fujifilm and Olympus. He receives royalty payments from GI Windows, EndoSim and Enterasense Ltd.
