HLA Class II Polymorphisms Modulate Gut Microbiota and Experimental Autoimmune Encephalomyelitis Phenotype

Shailesh K. Shahi, Soham Ali, Camille M. Jaime, Natalya V. Guseva and Ashutosh K. Mangalam Downloaded from

ImmunoHorizons 2021, 5 (8) 627-646 doi: https://doi.org/10.4049/immunohorizons.2100024 http://www.immunohorizons.org/content/5/8/627 http://www.immunohorizons.org/ This information is current as of September 26, 2021. http://www.immunohorizons.org/

Supplementary http://www.immunohorizons.org/content/suppl/2021/08/11/immunohorizon Material s.2100024.DCSupplemental References This article cites 98 articles, 14 of which you can access for free at: http://www.immunohorizons.org/content/5/8/627.full#ref-list-1 by guest on September 26, 2021 Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: by guest on September 26, 2021 http://www.immunohorizons.org/alerts

ImmunoHorizons is an open access journal published by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 All rights reserved. ISSN 2573-7732. RESEARCH ARTICLE

Adaptive Immunity

HLA Class II Polymorphisms Modulate Gut Microbiota and Experimental Autoimmune Encephalomyelitis Phenotype

Shailesh K. Shahi,*,1 Soham Ali,*,†,1 Camille M. Jaime,* Natalya V. Guseva,* and Ashutosh K. Mangalam*,‡,§ *Department of Pathology, University of Iowa, Iowa City, IA; †Roy J. and Lucille A. Carver College of Medicine, University of Iowa, Iowa City, IA; ‡Graduate § Program in Immunology, University of Iowa, Iowa City, IA; and Graduate Program in Molecular Medicine, University of Iowa, Iowa City, IA Downloaded from

ABSTRACT

Multiple sclerosis (MS) is an autoimmune disease of the CNS in which the interaction between genetic and environmental factors plays http://www.immunohorizons.org/ an important role in disease pathogenesis. Although environmental factors account for 70% of disease risk, the exact environmental factors associated with MS are unknown. Recently, gut microbiota has emerged as a potential missing environmental factor linked with the pathobiology of MS. Yet, how genetic factors, such as HLA class II gene(s), interact with gut microbiota and influence MS is unclear. In the current study, we investigated whether HLA class II genes that regulate experimental autoimmune encephalomyelitis (EAE) and MS susceptibility also influence gut microbiota. Previously, we have shown that HLA-DR3 transgenic mice lacking endogenous mouse class II genes (AE-KO) were susceptible to myelin proteolipid protein (91–110)–induced EAE, an animal model of MS, whereas AE-KO.HLA-DQ8 transgenic mice were resistant. Surprisingly, HLA-DR3.DQ8 double transgenic mice showed higher disease prevalence and severity compared with HLA-DR3 mice. Gut microbiota analysis showed that HLA-DR3, HLA-DQ8, and HLA- DR3.DQ8 double transgenic mice microbiota are compositionally different from AE-KO mice. Within HLA class II transgenic mice, the microbiota of HLA-DQ8 mice were more similar to HLA-DR3.DQ8 than HLA-DR3. As the presence of DQ8 on an HLA-DR3 background by guest on September 26, 2021 increases disease severity, our data suggests that HLA-DQ8–specific microbiota may contribute to disease severity in HLA-DR3.DQ8 mice. Altogether, our study provides evidence that the HLA-DR and -DQ genes linked to specific gut microbiota contribute to EAE susceptibility or resistance in a transgenic animal model of MS. ImmunoHorizons, 2021, 5: 627–646.

INTRODUCTION T cell immune response to a number of myelin Ags, including proteolipid protein (PLP), and myelin oligodendrocyte glyco- Multiple sclerosis (MS) is a chronic, autoimmune inflammatory protein (1, 2). The cause of MS is poorly understood, but collec- demyelinating disease of the CNS resulting from aberrant CD4 tive evidence suggests that the interaction of both genetic and

Received for publication March 16, 2021. Accepted for publication July 20, 2021. Address correspondence and reprint requests to: Dr. Ashutosh K. Mangalam, Department of Pathology, University of Iowa, 25 S. Grand Avenue, 1080A Medical Laboratory, Iowa City, IA 52246. E-mail address: [email protected] ORCIDs: 0000-0002-8506-7655 (S.K.S.); 0000-0002-6797-8842 (C.M.J.); 0000-0002-9926-2531 (A.K.M.). 1S.K.S. and S.A. contributed equally. The sequences presented in this article have been submitted to the Sequence Read Archives (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA738522/) under accession number PRJNA738522. This work was supported by a National Multiple Sclerosis Society grant (RG 5138A1/1T), a National Institutes of Health, National Institute of Allergy and Infectious Diseases grant (1R01AI137075-01), a National Institute of Environmental Health Sciences career development award (P30 ES005605), a Carver Trust Medical Research Initiative Grant, and a University of Iowa Environmental Health Sciences Research Center pilot grant. S.A. was supported by the Roy J. and Lucille A. Carver College of Medicine Emory Warner Fellowship, which provides medical students the opportunity to take a full year out of their medical school curriculum to work in a laboratory in the University of Iowa Department of Pathology. C.M.J. was supported by a scholarship from the Iowa Biosciences Academy and a National Institutes of Health Initiative for Maximizing Student Development R25 training grant (5R25GM058939). Abbreviations used in this article: AE-KO, MHC class II gene knockout; DSS, dextran sulfate sodium; EAE, experimental autoimmune encephalomyelitis; FDR, false discovery rate; GxE, gene–environment interaction; IBD, inflammatory bowel disease; LDA, linear discriminant analysis; LEfSe, LDA effect size; MS, multiple sclerosis; padj, adjusted p value; PCA, principal component for analysis; PLP, proteolipid protein; SCFA, short-chain fatty acid. The online version of this article contains supplemental material. This article is distributed under the terms of the CC BY 4.0 Unported license. Copyright © 2021 The Authors https://doi.org/10.4049/immunohorizons.2100024 627

ImmunoHorizons is published by The American Association of Immunologists, Inc. 628 HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA ImmunoHorizons environmental factors plays an important role (3). Furthermore, Although prior studies have reported a role of MHC genes recent studies showed a greater consensus on the contribution of in gut microbiota dysbiosis (30–32), the importance of HLA- gene–environment interactions (GxE) in the pathogenesis of MS DQ8 gene in modulating HLA-DR3 microbiome has not been (4–6). Genetic factors account for 30% of disease risk as deter- studied previously. In the current study, we investigated mined from studies of identical twins (7). In addition, environ- whether changes in HLA class II molecules from HLA-DQ8 to mental factors account for 70% of disease risk (8). However, how HLA-DR3 influences gut microbiota. Additionally, we ask genetic and environmental factors are linked with a predisposi- whether the HLA-DQ8 gene can modulate gut microbiota of tion to induce MS or protect from it is unknown. HLA-DR3 transgenic mice and modulate disease severity. We Among all the genetic factors associated with MS suscepti- found that the gut microbiota of MHC class II genes knockout bility, the strongest association has been found with MHC (AE-KO) mice differ from HLA-DR3, HLA-DQ8, and HLA- genes. MHC genes are among the most polymorphic genes in DR3.DQ8 double transgenic mice. We also further observed vertebrates (9), also known in humans as the HLA genes. Previ- that the gut microbiota of HLA-DR3 mice are different from ous studies have reported that individuals with HLA-DR2/ that of HLA-DQ8. Furthermore, our results showed that HLA-DQ6, HLA-DR3/HLA-DQ2, and HLA-DR4/HLA-DQ8 although HLA-DQ8 confers resistance to EAE, its shared Downloaded from class II haplotypes have an increased frequency of MS (10, 11). microbiota with HLA-DR3/HLA-DQ8 double transgenic mice Although a direct role of HLA-DR alleles in MS has been eluci- increase disease severity. Therefore, our study demonstrates an dated, the contribution of HLA-DQ alleles in disease pathogen- important role of the HLA-DQ8 and HLA-DR3 genes in shap- esis is not known. Human studies have shown that HLA-DQ ing gut microbiota and the HLA-DQ8 gene’s influence on the alleles may play a modulatory role in MS progression (12, 13). gut microbiota of HLA-DR3 transgenic mice. Using single transgenic mice expressing HLA class II genes, we http://www.immunohorizons.org/ showed that although both HLA-DR3 and HLA-DQ8 were able MATERIALS AND METHODS to recognize and mount a CD4 T cell response to PLP91–110, only HLA-DR3 transgenic mice were susceptible to myelin Generation of single and double transgenic mice and PLP91–110–induced experimental autoimmune encephalomyelitis (EAE), whereas HLA-DQ8 (DQB1*0302) transgenic mice were breeding scheme resistant (14, 15). The CD4 T cells from HLA-DR3/HLA-DQ8 AE-KO mice. AE-KO mice were originally generated by Mathis double transgenic mice also recognized and proliferated to and Benoist (33). Briefly, AE-KO mice lack all four conventional

PLP91–110 peptide. However, the HLA-DR3/HLA-DQ8 double MHC class II genes (Aa,Ab,Ea,andEb) because of a large transgenic mice showed higher disease prevalence and severity (80 kb) deletion of the entire mouse class II region. Embryonic (14, 15), suggesting that the disease-resistant HLA-DQ8 allele stem cells with deleted MHC class II locus were injected into by guest on September 26, 2021 may synergize with HLA-DR3 to modulate the disease severity C57BL/6 blastocysts following standard procedures (34). Chi- and progression. We have also shown that HLA-DQ8–induced meras were then crossed with C57BL/6 mice to generate AE- IL-17playsaroleinmodulatingdiseaseHLA-DR3.DQ8trans- KO mice, which were then interbred to generate homozygous genic mice. As gut microbiota has been shown to play a central AE-KO mice and maintained in homozygous conditions for a role in the development of IL-17–producing CD4 T cells (16), it prolonged period by interbreeding. is possible that HLA-DQ8 modulate disease through its influ- ence on gut microbiota. HLA-DQ8. AE-KO transgenic mice. Generation of DQ8 trans- Although environmental factors account for 70% of MS risk, genic mice on B10 background was achieved as follows. Briefly, the identification of an exact environmental factor associated cosmids H11A (30-kb DNA) and Â10A (38-kb DNA fragment), with MS remains unidentified. We and others have shown that which contain the DQA*0301 and DQB*0302 genes, respectively, the gut microbiota are an important environmental factor linked were microinjected into (CBA/J. B10.M) F2 embryos, as previ- with the etiopathogenesis of MS (17–22). The consensus of these ously described (35). Transgene-positive founders were identified studies is that MS patients exhibit gut dysbiosis, i.e., a change in by Southern blot analysis of tail DNA and subsequently mated to the composition of the gut microbiota community characterized B10.M mice. The HLA-DQ8 transgenes were crossed with AE-KO by an increase in harmful bacteria (pathobionts) and a decrease mice to generate HLA-DQ8.AE-KO transgenic mice (36). Animals in beneficial bacteria. A healthy gut microbiota helps to maintain from F1 cross were genotyped by PCR to select mice positive for the healthy state of the host through multiple ways, including DQ8 genes and negative for endogenous MHC class II genes. food metabolism, maintenance of intestinal barrier, energy Mice with the correct genotype were intercrossed and maintained homeostasis, inhibition of colonization by pathogenic organisms, in homozygous conditions by prolonged interbreeding. regulation of host physiology, and shaping of immune responses (23–26). Change in gut microbiota because of various important AE-KO. HLA-DR3 transgenic mice. Generation of transgenic endogenous (genetic factors) and exogenous factors (diets, antibi- mice expressing HLA-DR3 (DRB1*0301) (gift from Dr. Gunter otics and other drugs, lifestyle, and smoking) results in alterations Hammerling, Heidelberg, Germany) has been described previously in its metabolites’ synthesis and perturbation of normal homeo- (37). Briefly, 6 kb NdeI fragment of an HLA-DRA genomic con- stasis, even leading to intestinal and systemic disorders (27–29). struct and a 24 kb ClaI  SalI fragment containing the HLA-DRB

https://doi.org/10.4049/immunohorizons.2100024 ImmunoHorizons HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA 629 gene of DRBI*0301 were coinjected into fertilized eggs from Disease induction and scoring (C57B1/6  DBA/2)F1donorsmatedwithC57BL/6jmales,as AE-KO, HLA-DQ8, HLA-DR3 and HLA-DR3.DQ8 transgenic described previously (37). Transgene-positive founders were identi- mice (8 to 12 wk old) were immunized s.c. in both flanks using fied by Southern blot analysis of tail DNA and subsequently mated 25 mgofPLP91–110 peptides (GenScript, Piscataway, NJ) that to B10.M mice. The HLA-DQ8 transgenes were crossed with AE- were emulsified in CFA containing Mycobacterium tuberculosis KO mice to generate HLA-DR3.AE-KO transgenic mice. Mice H37Ra (100 mg/mouse; Becton Dickinson, Sparks, MD). Pertus- from F1 cross were genotyped by PCR to select for mice positive sis toxin (Sigma, St. Louis, MO; 80 ng) was administered i.p. at for DR3 genes and negative for endogenous MHC class II genes. days 0 and 2 postimmunization. Mice were observed daily for Mice with the correct genotype were intercrossed and maintained clinical disease up to day 19. Disease severity was scored in homozygous conditions by interbreeding. according to the standard 0–5 scoring system described previ- ously (42). Briefly, this scoring is 0, normal; 1, loss of tail tonic- DR3.DQ8 transgenic mice. The homozygous HLA-DQ8.AE-KO ity; 2, hind limb weakness; 3, hind limb paralysis; 4, complete mice were mated with the homozygous HLA.DR3.AE-KO mice hind limb paralysis and forelimb paralysis or weakness; and 5, to obtain the HLA-DR3.DQ8.AE-KO double transgenic mice

moribund/death. Downloaded from lines (14). All the HLA-DR3.DQ8.AE-KO mice used in the study were from F1 cross between HLA-DR3 and HLA-DQ8 trans- RESULTS genic mice. Double transgenic mice were genotyped to select pups positive for both HLA-DR3 and HLA-DQ8 genes. To analyze the effect of HLA polymorphism on EAE, we HLA transgenic as well as control AE-KO mice strains were immunizedAE-KO,HLA-DR3,HLA-DQ8,andHLA-DR3.DQ8 maintained in the animal facility at the University of Iowa. http://www.immunohorizons.org/ mice with PLP91–110 peptide (43). The HLA-DR3 mice began Eight- to twelve-wk-old male and female mice from HLA class showing clinical signs of EAE disease on day 11 postimmuniza- II transgenic strains or AE-KO control were used in this study. tion, with a steadily increasing average EAE score that reached For simplicity, these mice strains were referred as HLA-DR3, around 2.8 6 0.4 by day 19 (Fig. 1A). In contrast, no disease HLA-DQ8, HLA-DR3.DQ8, and AE-KO mice throughout the was observed in mice either lacking endogenous MHC class II manuscript. All experiments were approved by the Institutional (AE-KO) or expressing HLA-DQ8. However, the HLA- Animal Care and Use Committee at the University of Iowa, and DR3.DQ8 double transgenic mice began showing symptoms at animals were maintained in accordance with National Institutes day 9 postimmunization, and disease severity worsened quickly, of Health and institutional guidelines. reaching around 4.3 6 0.4byday19(Fig.1A).Fig.1Bfurther demonstrates the significant severity of disease, showing a by guest on September 26, 2021 mean cumulative EAE score over three times higher in the Microbiome analysis HLA-DR3.DQ8 group compared with the HLA-DR3 group. Mouse fecal samples were collected from naive HLA-DR3, To characterize the relationship between the HLA polymor- HLA-DQ8, HLA-DR3.DQ8 transgenic mice, and AE-KO mice phisms and microbiome composition, DNA was extracted from groups. Microbial DNA extraction, 16S amplicon (V3–V4 fecal samples of AE-KO (n 5 16; 8 male and 8 female), HLA- region), and sequencing were done as described previously DQ8 (n 5 12; 5 male and 7 female), HLA-DR3 (n 5 18; 9 male (38). Raw 16S sequence data were processed by R script dada2 and9female),andHLA-DR3.DQ8(n 5 15; 7 male and 8 female) to generate amplicon sequence variants, which were then mice. The V3–V4 region of 16S rRNA was then amplified and assigned taxonomies using a naive Bayesian classifier with the sequenced, and the sequence data were processed through the Silva database as a reference. Functional profiling of these bac- open access, R-based dada2 pipeline to perform sequence qual- terial communities was then performed using PICRUSt2 (39) to ity checks, trimming, merging of paired reads, alignment, and generate pathway abundance tables for the samples. taxonomy assignment of the aligned sequences. The resultant Analysis of these data were then conducted using online median read depth was 38085, ranging from 5146 to 99,124. microbiome tools METAGENassist (40) and MicrobiomeAna- lyst(41),aswellasin-housescripts.MicrobiomeAnalystwas Distinct gut microbiota among AE-KO and HLA class II used to generate relative abundance bar graphs as well as to transgenic mice perform linear discriminant analysis (LDA) effect size (LEfSe) a-Diversity analysis for genus richness (Chao1 index) showed analysis. METAGENassist was used for ordination of the data significantly decreased richness in AE-KO mice versus the and generation of principal components for analysis (PCA). Dif- other three groups (p values < 0.03) (Fig. 2A). There was no ferential abundance analysis for taxonomic and pathway data significant difference in a-diversity among HLA-DQ8, HLA- were performed using nonparametric tests (Wilcoxon signed- DR3, and HLA-DR3.DQ8. Euclidean distance–based b-diver- rank test for two groups and Kruskal–Wallis test for $3 sity analysis at the genus level between AE-KO and each of groups)andfalsediscoveryrate(FDR)adjustedformultiple the other groups demonstrated a clear separation between comparisons using the Benjamini–Hochberg algorithm. AE-KO and both HLA-DR3 and HLA-DR3.DQ8 (Fig. 2B),

https://doi.org/10.4049/immunohorizons.2100024 630 HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA ImmunoHorizons

HLA class II polymorphism influences gut microbiota profile Numerous genera were depleted in AE-KO mice compared with both HLA-DQ8 and HLA-DR3, most of which belong to the Firmicutes phylum: Acetatifactor, Anaeroplasma, Ileibacte- rium, Colidextribacter, Intestinimonas, and Lachnospiraceae UCG-006. Additionally, Wolinella of the Campilobacterota phy- lum and Rikenella of the Bacteroidota (more often referred to as Bacteroidetes) phylum were depleted in AE-KO. In the Desulfobacterota phylum, whereas Desulfovibrio was depleted in AE-KO, Bilophila was significantly overabundant compared with the other two groups (Fig. 3, Table I). Although numerous bacteria were differentially abundant in AE-KO, we wanted to identify specific bacteria that could best Downloaded from discriminatebetweenAE-KO,HLA-DQ8,andHLA-DR3.Todo this, we performed a random forest algorithm to classify the samplesaseitherAE-KO,HLA-DQ8,orHLA-DR3usingthe genus abundance data. Important bacteria were then identified based on how much the random forest classification accuracy

decreased when that bacterium was removed from the feature http://www.immunohorizons.org/ set (Fig. 2D). Using this approach, Rikenella and Wolinella were identified as the top two important features in identifying the AE-KO genotype, with mean decrease accuracy of 6 and 4%, respectively. In an effort to highlight the bacteria most likely to have biological relevance in the differences between AE-KO, HLA-DQ8, and HLA-DR3 mice, LEfSe was performed. The effect sizes, or how well the abundance of each bacteria separated the groups, of the 15 most significant genera are

shown in Fig. 2E. This analysis highlights Bacteroides, Par- by guest on September 26, 2021 abacteroides, Helicobacter,andEscherichia as important FIGURE 1. PLP91–110–induced EAE in AE-KO, HLA-DQ8, HLA-DR3, features for distinguishing AE-KO from the other two and HLA-DR3.DQ8 transgenic mice. groups. Each of these genera are overabundant in AE-KO; (A) Daily average EAE scores of AE-KO, HLA-DQ8, HLA-DR3, and HLA- DR3.DQ8 mice 19 d after immunization. The p values were determined although Parabacteroides was only significantly overabun- by multiple t tests. (B) Averaged cumulative EAE score for both groups dant in AE-KO compared with HLA-DQ8, Helicobacter and of mice across all 19 d. The p value was determined by Mann–Whitney Escherichia were only significantly overabundant compared unpaired U test (B). with HLA-DR3. whereasAE-KOandHLA-DQ8exhibitedsomecomposi- Distinct gut microbiota in single versus double transgenic tional similarity. mice A Kruskal–Wallis test was performed at all taxonomic Next, microbial composition in HLA-DQ8, HLA-DR3, and levels from phylum to genus, restricted to taxa with relative HLA-DR3.DQ8 were characterized to highlight differ- abundance >0.002. At a FDR of 5%, 33 differentially abun- ences and similarities between single and double trans- dant taxa were identified (Fig. 2C, Table I). AE-KO mice genic mice. First, b-diversity analysis using Euclidean- had a higher overall abundance of Bacteroidota (also known based distance at the genus level was performed on all as Bacteroidetes; the Silva database uses the Genome Tax- four groups, showing HLA-DQ8, HLA-DR3, and HLA- onomy Database naming convention) when compared with DR3.DQ8 clustering together, separate from AE-KO, even HLA-DQ8 and HLA-DR3 mice. The AE-KO group also when stratified by sex (Fig. 4A, Supplemental Fig. 1). exhibited higher abundance of the phyla Campilobacterota Removing AE-KO from this analysis shows that HLA-DR3. andDeferribacterotawhencomparedwithHLA-DR3,but DQ8 is compositionally distinct from HLA-DR3 but very there was no difference when compared with HLA-DQ8. similar to HLA-DQ8 (Fig. 4B). The Patescibacteria phylum was found in lower abundance The Kruskal–Wallis test was performed to identify taxa that in AE-KO mice when compared with HLA-DR3 mice but were differentially abundant between HLA-DQ8, HLA-DR3, not with HLA-DQ8 mice (Table I). and HLA-DR3.DQ8 groups at all taxonomic levels from phylum

https://doi.org/10.4049/immunohorizons.2100024 ImmunoHorizons HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA 631 Downloaded from http://www.immunohorizons.org/ by guest on September 26, 2021

FIGURE 2. Distinct gut microbiota among AE-KO and HLA transgenic mice. (A) Box plot showing unfiltered genus richness (Chao1 index) of AE-KO, HLA-DQ8, HLA-DR3, and HLA-DR3·DQ8 groups. Overall p value: 0.0049. (B) Pairwise principal coordinate analysis of b-diversity using Euclidean distance at the genus level. Distinct separation between AE-KO and both HLA-DR3 and HLA-DR3·DQ8, whereas there is overlap between AE-KO and HLA-DQ8. PC1 and PC2 are the first and second principal components, respectively, and represent the most and second-most amount of variation in the bacterial abundance data between all the samples. (C) Square root scaled bar plots of relative abundances of significant taxa at each taxonomic level with relative abundance >0.002. (D) Top 15 important features selected by random for- est. Removing Rikenella from the feature set available to the model led to 16% decrease in classification accuracy. Removing Desulfovibrio led to around a 10% decrease in accuracy. (E) Top 15 significant genera selected by LEfSe analysis, LDA score reflecting their effect sizes, and the heat map on the right depicting whether they were high, medium, or low in AE-KO, HLA-DQ8, and HLA-DR3 groups from left to right.

https://doi.org/10.4049/immunohorizons.2100024 632 HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA ImmunoHorizons

TABLE I. Significant taxa between AE-KO, HLA-DQ8, and HLA-DR3 at the phylum through genus levels Mean (Standard Error) Relative Abundance Taxa AE-KO HLA-DQ8 HLA-DR3 p Value q Value Phylum Bacteroidota 5.66 Â 10À1 (3.32 Â 10À2) 4.82 Â 10À1 (3.53 Â 4.22 Â 10À1 (3.21 Â 1.21 Â 10À2 1.93 Â 10À2 (Bacteroides) 10À2) 10À2) Campilobacterota 8.82 Â 10À2 (1.55 Â 10À2) 8.15 Â 10À2 (1.29 Â 4.45 Â 10À2 (1.05 Â 3.45 Â 10À2 4.59 Â 10À2 10À2) 10À2) Deferribacterota 2.45 Â 10À2 (6.33 Â 10À3) 1.96 Â 10À2 (6.40 Â 5.65 Â 10À3 (2.33 Â 1.31 Â 10À3 2.62 Â 10À3 10À3) 10À3) Desulfobacterota 1.83 Â 10À3 (4.62 Â 10À4) 1.32 Â 10À3 (4.10 Â 9.07 Â 10À3 (2.39 Â 1.94 Â 10À4 6.17 Â 10À4 10À4) 10À3) Firmicutes 2.44 Â 10À1 (2.12 Â 10À2) 3.59 Â 10À1 (3.99 Â 4.50 Â 10À1 (3.43 Â 2.31 Â 10À4 6.17 Â 10À4 10À2) 10À2) Patescibacteria 7.58 Â 10À3 (1.93 Â 10À3) 1.59 Â 10À2 (2.55 Â 3.06 Â 10À2 (3.38 Â 2.07 Â 10À5 1.66 Â 10À4 10À3) 10À3) Proteobacteria 6.11 Â 10À2 (1.15 Â 10À2) 3.03 Â 10À2 (8.53 Â 2.46 Â 10À2 (4.55 Â 4.76 Â 10À2 5.44 Â 10À2 Downloaded from 10À3) 10À3) Class Bacilli 1.12 Â 10À1 (1.50 Â 10À2) 2.07 Â 10À1 (3.55 Â 3.09 Â 10À1 (4.19 Â 1.07 Â 10À3 3.26 Â 10À3 10À2) 10À2) Bacteroidia 5.66 Â 10À1 (3.32 Â 10À2) 4.82 Â 10À1 (3.53 Â 4.22 Â 10À1 (3.21 Â 1.21 Â 10À2 2.01 Â 10À2 10À2) 10À2) http://www.immunohorizons.org/ Campylobacteria 8.82 Â 10À2 (1.55 Â 10À2) 8.15 Â 10À2 (1.29 Â 4.45 Â 10À2 (1.05 Â 3.45 Â 10À2 4.92 Â 10À2 10À2) 10À2) Coriobacteriia 2.90 Â 10À3 (6.10 Â 10À4) 2.57 Â 10À3 (4.55 Â 7.15 Â 10À3 (1.23 Â 1.63 Â 10À3 3.26 Â 10À3 10À4) 10À3) Deferribacteres 2.45 Â 10À2 (6.33 Â 10À3) 1.96 Â 10À2 (6.40 Â 5.65 Â 10À3 (2.33 Â 1.31 Â 10À3 3.26 Â 10À3 10À3) 10À3) Desulfovibrionia 1.83 Â 10À3 (4.62 Â 10À4) 1.32 Â 10À3 (4.10 Â 9.07 Â 10À3 (2.39 Â 1.94 Â 10À4 9.71 Â 10À4 10À4) 10À3) Gammaproteobacteria 6.11 Â 10À2 (1.15 Â 10À2) 3.03 Â 10À2 (8.53 Â 2.46 Â 10À2 (4.55 Â 4.76 Â 10À2 5.95 Â 10À2 10À3) 10À3) Saccharimonadia 7.58 Â 10À3 (1.93 Â 10À3) 1.59 Â 10À2 (2.55 Â 3.06 Â 10À2 (3.38 Â 2.07 Â 10À5 2.07 Â 10À4 10À3) 10À3) by guest on September 26, 2021 Order Acholeplasmatales 4.47 Â 10À4 (4.47 Â 10À4) 2.54 Â 10À2 (5.61 Â 8.05 Â 10À3 (2.20 Â 2.22 Â 10À7 4.00 Â 10À6 10À3) 10À3) Bacteroidales 5.66 Â 10À1 (3.32 Â 10À2) 4.82 Â 10À1 (3.53 Â 4.22 Â 10À1 (3.21 Â 1.21 Â 10À2 2.17 Â 10À2 10À2) 10À2) Clostridiales 2.86 Â 10À3 (6.17 Â 10À4) 1.05 Â 10À3 (4.84 Â 2.09 Â 10À3 (7.96 Â 2.54 Â 10À2 4.15 Â 10À2 10À4) 10À4) Coriobacterales 2.90 Â 10À3 (6.10 Â 10À4) 2.57 Â 10À3 (4.55 Â 7.15 Â 10À3 (1.23 Â 1.63 Â 10À3 4.19 Â 10À3 10À4) 10À3) Deferribacterales 2.45 Â 10À2 (6.33 Â 10À3) 1.96 Â 10À2 (6.40 Â 5.65 Â 10À3 (2.33 Â 1.31 Â 10À3 3.92 Â 10À3 10À3) 10À3) Desulfovibrionales 1.83 Â 10À3 (4.62 Â 10À4) 1.32 Â 10À3 (4.10 Â 9.07 Â 10À3 (2.39 Â 1.94 Â 10À4 8.74 Â 10À4 10À4) 10À3) Enterobacterales 4.17 Â 10À2 (9.63 Â 10À3) 1.40 Â 10À2 (7.07 Â 1.13 Â 10À4 (4.85 Â 6.41 Â 10À7 5.77 Â 10À6 10À3) 10À5) Lactobacillales 6.74 Â 10À2 (1.13 Â 10À2) 1.29 Â 10À1 (3.17 Â 2.21 Â 10À1 (3.60 Â 1.24 Â 10À3 3.92 Â 10À3 10À2) 10À2) Oscillospirales 1.53 Â 10À2 (2.30 Â 10À3) 3.00 Â 10À2 (6.35 Â 2.56 Â 10À2 (3.53 Â 1.02 Â 10À2 2.17 Â 10À2 10À3) 10À3) Peptococcales 3.12 Â 10À4 (9.98 Â 10À5) 3.25 Â 10À4 (6.94 Â 1.12 Â 10À4 (4.46 Â 1.09 Â 10À2 2.17 Â 10À2 10À5) 10À5) Saccharimonadales 7.58 Â 10À3 (1.93 Â 10À3) 1.59 Â 10À2 (2.55 Â 3.06 Â 10À2 (3.38 Â 2.07 Â 10À5 1.24 Â 10À4 10À3) 10À3) Family Acholeplasmataceae 4.47 Â 10À4 (4.47 Â 10À4) 2.54 Â 10À2 (5.61 Â 8.05 Â 10À3 (2.20 Â 2.22 Â 10À7 6.22 Â 10À6 10À3) 10À3) Deferribacteraceae 2.45 Â 10À2 (6.33 Â 10À3) 1.96 Â 10À2 (6.40 Â 5.65 Â 10À3 (2.33 Â 1.31 Â 10À3 5.23 Â 10À3 10À3) 10À3)

https://doi.org/10.4049/immunohorizons.2100024 ImmunoHorizons HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA 633

TABLE I. continued

Mean (Standard Error) Relative Abundance Taxa AE-KO HLA-DQ8 HLA-DR3 p Value q Value Desulfovibrionaceae 1.83 Â 10À3 (4.62 Â 10À4) 1.32 Â 10À3 (4.10 Â 9.07 Â 10À3 (2.39 Â 1.94 Â 10À4 1.09 Â 10À3 10À4) 10À3) Eggerthellaceae 1.86 Â 10À3 (3.42 Â 10À4) 2.28 Â 10À3 (3.94 Â 6.94 Â 10À3 (1.17 Â 1.16 Â 10À4 8.09 Â 10À4 10À4) 10À3) Enterobacteriaceae 4.17 Â 10À2 (9.63 Â 10À3) 1.40 Â 10À2 (7.07 Â 1.13 Â 10À4 (4.85 Â 6.41 Â 10À7 8.97 Â 10À6 10À3) 10À5) Lactobacillaceae 6.70 Â 10À2 (1.13 Â 10À2) 1.27 Â 10À1 (3.17 Â 2.20 Â 10À1 (3.59 Â 1.24 Â 10À3 5.23 Â 10À3 10À2) 10À2) Peptococcaceae 3.12 Â 10À4 (9.98 Â 10À5) 3.25 Â 10À4 (6.94 Â 1.12 Â 10À4 (4.46 Â 1.09 Â 10À2 3.38 Â 10À2 10À5) 10À5) Prevotellaceae 1.00 Â 10À1 (1.08 Â 10À2) 1.39 Â 10À1 (2.76 Â 6.28 Â 10À2 (8.58 Â 9.70 Â 10À3 3.38 Â 10À2 10À2) 10À3) À À À À À Saccharimonadaceae 7.58E-3 (1.93 Â 10 3) 1.59 Â 10 2 (2.55 Â 3.06 Â 10 2 (3.38 Â 2.07 Â 10 5 1.94 Â 10 4 Downloaded from 10À3) 10À3) Genus Lachnospiraceae: 5.71 Â 10À5 (4.48 Â 10À5) 1.73 Â 10À3 (5.41 Â 1.41 Â 10À3 (6.47 Â 1.65 Â 10À4 8.27 Â 10À4 Acetatifactor 10À4) 10À4) Rikenellaceae: Alistipes 4.89 Â 10À2 (1.09 Â 10À2) 6.44 Â 10À2 (7.74 Â 2.89 Â 10À2 (4.92 Â 1.50 Â 10À2 3.29 Â 10À2 10À3) 10À3) Prevotellaceae: 8.45 Â 10À2 (1.04 Â 10À2) 1.38 Â 10À1 (2.77 Â 4.85 Â 10À2 (7.01 Â 2.13 Â 10À3 6.88 Â 10À3 http://www.immunohorizons.org/ Alloprevotella 10À2) 10À3) Bacteroidetes: 2.67 Â 10À1 (3.56 Â 10À2) 1.46 Â 10À1 (2.82 Â 1.83 Â 10À1 (1.67 Â 2.84 Â 10À2 4.88 Â 10À2 Bacteroides 10À2) 10À2) Proteobacteria: Bilophila 1.77 Â 10À3 (4.45 Â 10À4) 2.96 Â 10À4 (1.00 Â 3.28 Â 10À4 (1.38 Â 2.83 Â 10À3 8.19 Â 10À3 10À4) 10À4) Clostridiaceae: C. 2.86 Â 10À3 (6.17 Â 10À4) 1.05 Â 10À3 (4.84 Â 2.09 Â 10À3 (7.96 Â 2.54 Â 10À2 4.88 Â 10À2 arthromitus 10À4) 10À4) Eubacteriales: 4.47 Â 10À3 (8.08 Â 10À4) 1.14 Â 10À2 (2.07 Â 1.12 Â 10À2 (1.42 Â 2.40 Â 10À4 1.10 Â 10À3 Colidextribacter 10À3) 10À3) Proteobacteria: 6.95 Â 10À5 (3.34 Â 10À5) 1.03 Â 10À3 (3.63 Â 8.75 Â 10À3 (2.40 Â 6.33 Â 10À8 1.74 Â 10À6 Desulfovibrio 10À4) 10À3) À À À À À À Proteobacteria: 4.17 Â 10 2 (9.63 Â 10 3) 1.40 Â 10 2 (7.07 Â 1.13 Â 10 4 (4.85 Â 6.41 Â 10 7 8.81 Â 10 6 by guest on September 26, 2021 Escherichia 10À3) 10À5) Helicobacteraceae: 8.82 Â 10À2 (1.55 Â 10À2) 8.00 Â 10À2 (1.27 Â 4.31 Â 10À2 (1.02 Â 2.70 Â 10À2 4.88 Â 10À2 Helicobacter 10À2) 10À2) Firmicutes: 7.95 Â 10À4 (4.61 Â 10À4) 1.61 Â 10À2 (5.81 Â 1.17 Â 10À2 (4.10 Â 1.64 Â 10À3 5.63 Â 10À3 Ileibacterium 10À3) 10À3) Lachnospiraceae: 1.22 Â 10À2 (2.33 Â 10À3) 1.43 Â 10À2 (1.86 Â 2.78 Â 10À2 (4.73 Â 2.77 Â 10À2 4.88 Â 10À2 Lachnoclostridium 10À3) 10À3) Lachnospiraceae: 2.26 Â 10À3 (1.06 Â 10À3) 4.56 Â 10À3 (1.44 Â 6.38 Â 10À3 (2.01 Â 6.59 Â 10À3 1.65 Â 10À2 Lachnospiraceae 10À3) 10À3) UCG 006 Lactobacillaceae: 6.70 Â 10À2 (1.13 Â 10À2) 1.27 Â 10À1 (3.17 Â 2.20 Â 10À1 (3.59 Â 1.24 Â 10À3 4.80 Â 10À3 Lactobacillus 10À2) 10À2) Deferribacteraceae: 2.45 Â 10À2 (6.33 Â 10À3) 1.96 Â 10À2 (6.40 Â 5.65 Â 10À3 (2.33 Â 1.31 Â 10À3 4.80 Â 10À3 Mucispirillum 10À3) 10À3) Porphyromonadaceae: 1.67 Â 10À2 (3.68 Â 10À3) 2.93 Â 10À2 (5.67 Â 1.12 Â 10À2 (1.62 Â 2.67 Â 10À2 4.88 Â 10À2 Odoribacter 10À3) 10À3) Bacteroidetes: 8.89 Â 10À2 (8.27 Â 10À3) 5.38 Â 10À2 (7.64 Â 7.11 Â 10À2 (1.00 Â 2.57 Â 10À2 4.88 Â 10À2 Parabacteroides 10À3) 10À2) Bacteroidetes: 2.20 Â 10À4 (1.83 Â 10À4) 6.49 Â 10À3 (9.47 Â 1.49 Â 10À2 (1.98 Â 1.82 Â 10À8 1.00 Â 10À6 Rikenella 10À4) 10À3) Proteobacteria: 1.29 Â 10À5 (8.96 Â 10À6) 1.50 Â 10À3 (3.16 Â 1.41 Â 10À3 (3.54 Â 2.76 Â 10À6 2.53 Â 10À5 Wolinella 10À4) 10À4) Kruskal–Wallis test with Benjamini–Hochberg adjustment was performed on sample-scaled counts at a significance level of 0.05 to identify differentially abundant taxa. The q values are the padj. to genus. The p values were adjusted using an FDR of 5%, and DQ8 group exhibited an overabundance of Alistipes but the 28 resulting taxa with relative abundance >0.006 are decreased abundance of Prevotellaceae UCG-001 at the genus showninFig.4C.WithintheBacteroides phylum, the HLA- level. The HLA-DQ8 group also had a higher abundance of

https://doi.org/10.4049/immunohorizons.2100024 634 HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA ImmunoHorizons Downloaded from http://www.immunohorizons.org/ by guest on September 26, 2021

FIGURE 3. Differentially abundant bacteria in AE-KO, HLA-DQ8, and HLA-DR3 mice. Colors indicate relative abundance increasing in value from gray to red. Bacteria were grouped into three groups based on k-means clustering.

Desulfovibrio (Desulfobacterota) and Rikenella (Bacteroidota) Campilobacterota, Campylobacteria, and Proteobacteria and (Figs.4C,5,6;TableII). depletion of the phylum Patescibacteria, mirroring the The phyla Campilobacterota and Proteobacteria were sig- changes opposite to those observed in the HLA-DR3 group. nificantly underrepresented in HLA-DR3, whereas Patescibac- At the genus level, however, most of the significant genera teria were overabundant. Within the Proteobacteria phylum, belong to the Firmicutes phylum. Within this phylum, Helicobacter was depleted in HLA-DR3 compared with the Streptococcus and Marvinbryantia are depleted in HLA- other two groups. Several genera within the Firmicutes class DR3.DQ8 compared with the other two groups, whereas were depleted in HLA-DR3 (Butyricicoccus, Romboutsia,and Roseburia, Paludicola,andClostridium sensu stricto 1were Turicibacter)whereasIntestinimonas was overabundant. Within overabundant(Figs.4C,5,8;TableII). the Bacteroidota phylum, both Alloprevotella and Odoribacter We again used a random forest model to identify the were significantly depleted. Additionally, Mucispirillum of the most discriminatory bacteria between HLA-DQ8, HLA-DR3, Deferribacterota phylum was significantly depleted. Other than and HLA-DR3.DQ8. Turicibacter, Roseburia,andCandidatus Intestinimonas, the other genus significantly overabundant in saccharimonas were the most important features in separat- HLA-DR3 was Enterorhabdus of the Actinobacteriota phylum ing these groups, all with around a 4% mean decrease accu- (Figs. 4C, 5, 7; Table II). racy, respectively (Fig. 4D). Further analysis with LEfSe Interestingly, at the phylum level, the HLA-DR3.DQ8 highlights four HLA-DR3–depleted genera with the greatest group demonstrated significant overabundance of potential biological relevance, Alloprevotella, Helicobacter,

https://doi.org/10.4049/immunohorizons.2100024 ImmunoHorizons HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA 635 Downloaded from http://www.immunohorizons.org/ by guest on September 26, 2021

FIGURE 4. Distinct microbiota in single and double transgenic mice. (A) PCA plot comparing b-diversity of all four groups at the genus level demonstrating separation of AE-KO from the other three groups. (B)PCA plot comparing Euclidean distance–based b-diversity of DQ8, DR3, and DR3.DQ8 at the genus level. (C) Square root scaled bar (Continued)

https://doi.org/10.4049/immunohorizons.2100024 636 HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA ImmunoHorizons Downloaded from http://www.immunohorizons.org/ by guest on September 26, 2021

FIGURE 5. Differentially abundant bacteria among HLA-DQ8, HLA-DR3, and HLA-DR3.DQ8 mice. Colors indicate relative abundance increasing in value from gray to red. Bacteria were grouped into three groups based on k-means clustering.

Turicibacter,andMucispirillum, with effect sizes greater conducted from pathway profiles generated using PICRUSt2. than four (Fig. 4E). Bacteroides and Parabacteroides were Numerous pathways were significantly differentially abundant also identified as having a high effect size, although they in the three groups (Supplemental Table I), including four are significantly overabundant in HLA-DR3.DQ8 only when pathways related to short-chain fatty acids (SCFAs) (Fig. 9). compared with HLA-DR3 (Fig 4E). Three of these pathways involved SCFA production and were increasedinbothHLA-DQ8andHLA-DR3.DQ8compared with HLA-DR3: TCA cycle VII (acetate producers) (adjusted p À3 Altered short-chain fatty acid metabolism and value [padj] 5 4.3 Â 10 ), L-lysine fermentation to acetate and methanogenesis in HLA-DR3.DQ8 mice butyrate (padj 5 0.024), and succinate fermentation to butyrate À4 To examine potential mechanisms behind different disease phe- (padj 5 2.3 Â 10 ) (Fig. 9). The fourth SCFA-related pathway, À4 notypes, functional analysis of these microbiomes was methanogenesis from acetate (padj 5 2.0 Â 10 ), was plots of relative abundances of significant taxa at each taxonomic level with relative abundance >0.006. (D) Top 15 important features selected by random forest. Removing Turicibacter from the feature set available to the model leads to 11% decrease in accuracy. (E) Top 15 significant genera selected by LEfSe analysis. LDA score reflects their effect sizes, and the heat map on the right depicts whether they were high, medium, or low in HLA-DQ8, HLA-DR3, and HLA-DR3.DQ8 groups from left to right.

https://doi.org/10.4049/immunohorizons.2100024 ImmunoHorizons HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA 637

single (HLA-DR3 or DQ8) or double HLA class II genes (HLA- DR3.DQ8) showed more distinct microbiota than AE-KO mice. Furthermore, clustering of the HLA-DQ8 microbiome with HLA-DR3.DQ8 rather than HLA-DR3 and severe disease in HLA-DR3.DQ8 mice points toward an important role of gut microbiota in disease modulation in EAE-susceptible HLA-DR3 mice. GxE have been proposed to play a significant role in dis- ease susceptibility versus resistance (5), and our data suggest that such an interaction between HLA gene polymorphisms and gut microbiota contributes to disease outcome in MS. The association of MHC class II genes with the risk of developing autoimmune diseases has been well described. MHC class II molecules play a central role in the maturation of T cells and thus can affect the autoreactivity of an individual’s T cell population (44). Indeed, multiple HLA genes in humans Downloaded from such as HLA-DR2, HLA-DR3, HLA-DR4, HLA-DQ6, and HLA- DQ8 have been associated with MS susceptibility (11, 45). We have previously used transgenic mice expressing MS-linked HLA class II molecules to authenticate the importance of HLA polymorphisms in the pathobiology of MS (14, 15, 46–48). FIGURE 6. Differentially abundant bacteria in HLA-DQ8 mice. http://www.immunohorizons.org/ Abundance values are sum scaled to one million. The three lines repre- Examining the relationship between HLA genotype and the sent the lower quartile, median, and upper quartile from lowest to microbiome, the current study showed that HLA class II mole- highest, respectively. Significance labels: *p < 0.05, **0.05 > p > 0.01, cules influence the composition of the gut microbiome. Specifi- ***0.01 > p > 0.001, ****p < 0.001. cally, we showed that the microbiome composition of AE-KO mice is distinct from transgenic mice expressing human HLA classIImolecules(HLA-DQ8,HLA-DR3,andHLA-DR3·DQ8 significantly more abundant in HLA-DR3.DQ8 compared with mice). Additionally, the microbiome of AE-KO mice exhibits both HLA-DR3 and HLA-DQ8 (Fig. 9). There were also several lower genus richness than HLA-DQ8, HLA-DR3, and HLA- pathways predicted by PICRUSt2 to be relatively more abun- DR3.DQ8 mice, suggesting an important role of MHC class II molecule in the selection of gut microbiota. In the absence of dant in the HLA-DQ8 group compared with both HLA-DR3 by guest on September 26, 2021 and HLA-DR3.DQ8 (Supplemental Table I), and these are MHC class II molecule, mice are unable to maintain a diverse involved in the synthesis of widely prevalent proteins and bio- microbiome, allowing select bacteria to flourish at the cost of logical compounds such as bacteriochlorophyll (pwy-5531 and bacterial richness. pwy-7159) and polyamines (polyamsyn-pwy) and the degrada- Our data also suggests that MHC class II polymorphisms tion of nucleotides (pwy-6608 and pwy-6353). Overall, these might influence the selection of specific bacteria in the gut. results demonstrate an association of HLA class II polymor- Previously, using HLA class II transgenic mice, we have vali- dated the importance of HLA class II genes in susceptibility phisms with gut bacterial functions, including several SCFA versus resistance. Specifically, we found that HLA-DR3 mice metabolicpathwaysaswellasmethaneproduction. are susceptible to disease, and HLA-DQ8 mice are resistant to disease. Interestingly, double transgenic mice expressing both DISCUSSION HLA-DR3 and HLA-DQ8 developed more severe disease, sug- gesting that the disease-resistant HLA-DQ8 gene also has a Both genetic as well as environmental factors have been linked modulatory effect on disease severity (14, 46, 47). Thus, we with a risk of autoimmune diseases. In MS, genetic factors hypothesized that the gut microbiome might play an important account for only 30% of disease risk, whereas the remaining role in disease susceptibility, resistance, and/or severity in the 70% is attributed to environmental factors. Development of dis- context of the HLA-DR3 or -DQ8 molecule. EAE phenotypes ease only in a subset of individuals with susceptible HLA class can be broadly described by three characteristics, disease sus- II genes points toward an important role of GxE in disease sus- ceptibility or resistance and disease severity. As HLA-DR3 and ceptibility versus protection. Given that the gut microbiome has HLA-DR3.DQ8 mice were susceptible for disease, they might recently emerged as a potential environmental factor linked be selecting for bacteria with a potential role in disease suscep- with MS risk, it is imperative to analyze its interplay with tibility. Although HLA-DQ8 mice were resistant to disease genetic factors, especially HLA genotypes. Our study shows development, the presence of HLA-DQ8 in a disease-suscepti- that HLA class has a direct influence on the composition of gut ble background can worsen the disease. Thus, we argued that microbiota as mice lacking MHC class II showed lower bacte- HLA-DQ8 may be selecting for bacteria that protect from the rial diversity. Additionally, transgenic mice expressing either development of disease but also selecting for bacteria that can

https://doi.org/10.4049/immunohorizons.2100024 638 HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA ImmunoHorizons

TABLE II. Significant taxa at between HLA-DQ8, HLA-DR3, and HLA-DR3.DQ8 at the phylum through genus ranks Mean (Standard Error) Relative Abundance Taxa HLA-DQ8 HLA-DR3 HLA-DR3.DQ8 p Value q Value Phylum Campilobacterota 8.04 Â 10À2 (1.28 Â 10À2) 4.47 Â 10À2 (1.04 Â 1.26 Â 10À1 (1.87 Â 9.95 Â 10À4 1.99 Â 10À3 10À2) 10À2) Deferribacterota 1.93 Â 10À2 (6.35 Â 10À3) 5.65 Â 10À3 (2.32 Â 3.00 Â 10À2 (6.99 Â 4.32 Â 10À4 1.15 Â 10À3 10À3) 10À3) Desulfobacterota 1.31 Â 10À3 (4.10 Â 10À4) 9.15E-3 (2.39 Â 10À3) 4.35 Â 10À3 (1.21 Â 3.91 Â 10À4 1.15 Â 10À3 10À3) Patescibacteria 1.57 Â 10À2 (2.54 Â 10À3) 3.08 Â 10À2 (3.41 Â 6.35 Â 10À3 (8.80 Â 1.29 Â 10À6 1.03 Â 10À5 10À3) 10À4) Proteobacteria 1.45 Â 10À2 (5.56 Â 10À3) 2.46 Â 10À2 (4.57 Â 3.24 Â 10À2 (3.20 Â 5.95 Â 10À3 9.52 Â 10À3 10À3) 10À3) Class Campylobacteria 8.04 Â 10À2 (1.28 Â 10À2) 4.47 Â 10À2 (1.04 Â 1.26 Â 10À1 (1.87 Â 9.95 Â 10À4 1.99 Â 10À3 10À2) 10À2) Downloaded from Coriobacteriia 2.22 Â 10À3 (3.69 Â 10À4) 7.01 Â 10À3 (1.18 Â 2.25 Â 10À3 (3.47 Â 3.67 Â 10À4 1.08 Â 10À3 10À3) 10À4) Deferribacteres 1.93 Â 10À2 (6.35 Â 10À3) 5.65 Â 10À3 (2.32 Â 3.00 Â 10À2 (6.99 Â 4.32 Â 10À4 1.08 Â 10À3 10À3) 10À3) Desulfovibrionia 1.31 Â 10À3 (4.10 Â 10À4) 9.15 Â 10À3 (2.39 Â 4.35 Â 10À3 (1.21 Â 3.91 Â 10À4 1.08 Â 10À3 10À3) 10À3) http://www.immunohorizons.org/ Gammaproteobacteria 1.45 Â 10À2 (5.56 Â 10À3) 2.46 Â 10À2 (4.57 Â 3.24 Â 10À2 (3.20 Â 5.95 Â 10À3 9.92 Â 10À3 10À3) 10À3) Saccharimonadia 1.57 Â 10À2 (2.54 Â 10À3) 3.08 Â 10À2 (3.41 Â 6.35 Â 10À3 (8.80 Â 1.29 Â 10À6 1.29 Â 10À5 10À3) 10À4) Order Acholeplasmatales 2.55 Â 10À2 (5.69 Â 10À3) 8.12 Â 10À3 (2.21 Â 1.14 Â 10À2 (1.46 Â 5.24 Â 10À3 1.11 Â 10À2 10À3) 10À3) Burkholderiales 1.45 Â 10À2 (5.56 Â 10À3) 2.46 Â 10À2 (4.57 Â 3.24 Â 10À2 (3.20 Â 5.95 Â 10À3 1.12 Â 10À2 10À3) 10À3) Campylobacterales 8.04 Â 10À2 (1.28 Â 10À2) 4.47 Â 10À2 (1.04 Â 1.26 Â 10À1 (1.87 Â 9.95 Â 10À4 2.42 Â 10À3 10À2) 10À2) Clostridiales 4.99 Â 10À3 (1.94 Â 10À3) 2.50 Â 10À3 (8.39 Â 1.81 Â 10À2 (3.53 Â 5.47 Â 10À5 3.10 Â 10À4 by guest on September 26, 2021 10À4) 10À3) Coriobacteriales 2.22 Â 10À3 (3.69 Â 10À4) 7.01 Â 10À3 (1.18 Â 2.25 Â 10À3 (3.47 Â 3.67 Â 10À4 1.22 Â 10À3 10À3) 10À4) Deferribacterales 1.93 Â 10À2 (6.35 Â 10À3) 5.65 Â 10À3 (2.32 Â 3.00 Â 10À2 (6.99 Â 4.32 Â 10À4 1.22 Â 10À3 10À3) 10À3) Desulfovibrionales 1.31 Â 10À3 (4.10 Â 10À4) 9.15 Â 10À3 (2.39 Â 4.35 Â 10À3 (1.21 Â 3.91 Â 10À4 1.22 Â 10À3 10À3) 10À3) Peptococcales 3.19 Â 10À4 (6.51 Â 10À5) 1.13 Â 10À4 (4.46 Â 2.95 Â 10À4 (8.24 Â 8.71 Â 10À3 1.48 Â 10À2 10À5) 10À5) Peptostreptococcales_ 3.40 Â 10À3 (1.22 Â 10À3) 7.74 Â 10À4 (1.68 Â 1.02 Â 10À2 (2.45 Â 3.47 Â 10À6 2.95 Â 10À5 Tissierellales 10À4) 10À3) Saccharimonadales 1.57 Â 10À2 (2.54 Â 10À3) 3.08 Â 10À2 (3.41 Â 6.35E-3 (8.80 Â 1.29 Â 10À6 2.19 Â 10À5 10À3) 10À4) Family Acholeplasmataceae 2.55 Â 10À2 (5.69 Â 10À3) 8.12 Â 10À3 (2.21 Â 1.14 Â 10À2 (1.46 Â 5.24 Â 10À3 1.28 Â 10À2 10À3) 10À3) Bacteroidaceae 1.47 Â 10À1 (3.04 Â 10À2) 1.83 Â 10À1 (1.65 Â 1.14 Â 10À1 (9.09 Â 1.85 Â 10À2 3.24 Â 10À2 10À2) 10À3) Butyricicoccaceae 2.33 Â 10À3 (6.39 Â 10À4) 1.37 Â 10À3 (3.02 Â 3.24 Â 10À3 (4.03 Â 5.97 Â 10À3 1.28 Â 10À2 10À4) 10À4) Clostridiaceae 4.99 Â 10À3 (1.94 Â 10À3) 2.50 Â 10À3 (8.39 Â 1.81 Â 10À2 (3.53 Â 5.47 Â 10À5 5.10 Â 10À4 10À4) 10À3) Deferribacteraceae 1.93 Â 10À2 (6.35 Â 10À3) 5.65 Â 10À3 (2.32 Â 3.00 Â 10À2 (6.99 Â 4.32 Â 10À4 2.01 Â 10À3 10À3) 10À3) Desulfovibrionaceae 1.31 Â 10À3 (4.10 Â 10À4) 9.15 Â 10À3 (2.39 Â 4.35 Â 10À3 (1.21 Â 3.91 Â 10À4 2.01 Â 10À3 10À3) 10À3)

https://doi.org/10.4049/immunohorizons.2100024 ImmunoHorizons HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA 639

TABLE II. continued

Mean (Standard Error) Relative Abundance Taxa HLA-DQ8 HLA-DR3 HLA-DR3.DQ8 p Value q Value Eggerthellaceae 2.22 Â 10À3 (3.69 Â 10À4) 7.01 Â 10À3 (1.18 Â 2.25 Â 10À3 (3.47 Â 3.67 Â 10À4 2.01 Â 10À3 10À3) 10À4) Helicobacteraceae 8.04 Â 10À2 (1.28 Â 10À2) 4.47 Â 10À2 (1.04 Â 1.26 Â 10À1 (1.87 Â 9.95 Â 10À4 3.48 Â 10À3 10À2) 10À2) Marinifilaceae 2.89 Â 10À2 (5.44 Â 10À3) 1.13 Â 10À2 (1.66 Â 2.73 Â 10À2 (4.44 Â 8.73 Â 10À4 3.48 Â 10À3 10À3) 10À3) Peptococcaceae 3.19 Â 10À4 (6.51 Â 10À5) 1.13 Â 10À4 (4.46 Â 2.95 Â 10À4 (8.24 Â 8.71 Â 10À3 1.63 Â 10À2 10À5) 10À5) Peptostreptococcaceae 3.20 Â 10À3 (1.21 Â 10À3) 3.28 Â 10À4 (1.46 Â 1.00 Â 10À2 (2.44 Â 6.94 Â 10À7 1.80 Â 10À5 10À4) 10À3) Prevotellaceae 1.38 Â 10À1 (2.75 Â 10À2) 6.32 Â 10À2 (8.57 Â 1.12 Â 10À1 (9.97 Â 2.58 Â 10À3 8.02 Â 10À3 10À3) 10À3) À À À À À À Saccharimonadaceae 1.57 Â 10 2 (2.54 Â 10 3) 3.08 Â 10 2 (3.41 Â 6.35 Â 10 3 (8.80 Â 1.29 Â 10 6 1.80 Â 10 5 Downloaded from 10À3) 10À4) Streptococcaceae 1.25 Â 10À3 (4.37 Â 10À4) 4.93 Â 10À4 (9.65 Â 1.72 Â 10À4 (5.96 Â 6.39 Â 10À3 1.28 Â 10À2 10À5) 10À5) Sutterellaceae 1.45 Â 10À2 (5.56 Â 10À3) 2.46 Â 10À2 (4.57 Â 3.24 Â 10À2 (3.20 Â 5.95 Â 10À3 1.28 Â 10À2 10À3) 10À3) Tannerellaceae 5.37 Â 10À2 (7.87 Â 10À3) 7.14 Â 10À2 (9.97 Â 3.24 Â 10À2 (7.21 Â 4.83 Â 10À3 1.28 Â 10À2 10À3) 10À3) http://www.immunohorizons.org/ Genus Rikenellaceae: Alistipes 6.39 Â 10À2 (7.60 Â 10À3) 2.91 Â 10À2 (4.90 Â 2.96 Â 10À2 (3.61 Â 1.04 Â 10À3 3.29 Â 10À3 10À3) 10À3) Prevotellaceae: 1.37 Â 10À1 (2.76 Â 10À2) 4.88 Â 10À2 (7.02 Â 1.01 Â 10À1 (1.01 Â 2.88 Â 10À4 1.83 Â 10À3 Alloprevotella 10À3) 10À2) Anaeroplasmataceae: 2.55 Â 10À2 (5.69 Â 10À3) 8.12 Â 10À3 (2.21 Â 1.14 Â 10À2 (1.46 Â 5.24 Â 10À3 1.30 Â 10À2 Anaeroplasma 10À3) 10À3) Bacteroidetes: Bacteroides 1.47 Â 10À1 (3.04 Â 10À2) 1.83 Â 10À1 (1.65 Â 1.14 Â 10À1 (9.09 Â 1.85 Â 10À2 3.77 Â 10À2 10À2) 10À3) Clostridiaceae: 1.63 Â 10À3 (6.27 Â 10À4) 2.05 Â 10À4 (8.03 Â 1.21 Â 10À3 (2.59 Â 6.34 Â 10À5 5.17 Â 10À4 Butyricicoccus 10À5) 10À4) À À À À À À C. saccharimonas 1.57 Â 10 2 (2.54 Â 10 3) 3.08 Â 10 2 (3.41 Â 6.35 Â 10 3 (8.80 Â 1.29 Â 10 5 3.67 Â 10 5 by guest on September 26, 2021 10À3) 10À4) Clostridiaceae: C. sensu 3.96 Â 10À3 (1.96 Â 10À3) 3.87 Â 10À4 (3.21 Â 1.33 Â 10À2 (3.48 Â 1.05 Â 10À5 1.50 Â 10À4 stricto 1 10À4) 10À3) Proteobacteria: 1.02 Â 10À3 (3.64 Â 10À4) 8.82 Â 10À3 (2.41 Â 4.20 Â 10À3 (1.23 Â 3.22 Â 10À4 1.84 Â 10À3 Desulfovibrio 10À3) 10À3) Eggerthellaceae: 1.44 Â 10À3 (2.58 Â 10À4) 3.71 Â 10À3 (7.55 Â 1.57 Â 10À3 (2.20 Â 3.79 Â 10À3 1.03 Â 10À2 Enterorhabdus 10À4) 10À4) Helicobacteraceae: 7.90 Â 10À2 (1.25 Â 10À2) 4.33 Â 10À2 (1.02 Â 1.24 Â 10À1 (1.84 Â 1.04 Â 10À3 3.29 Â 10À3 Helicobacter 10À2) 10À2) Eubacteriales: 8.09 Â 10À4 (1.34 Â 10À4) 2.27 Â 10À3 (3.48 Â 8.77 Â 10À4 (9.55 Â 1.80 Â 10À3 5.39 Â 10À3 Intestinimonas 10À4) 10À5) Lachnospiraceae: 2.69 Â 10À3 (8.47 Â 10À4) 6.32 Â 10À3 (1.55 Â 6.30 Â 10À4 (2.91 Â 1.75 Â 10À5 2.00 Â 10À4 Marvinbryantia 10À3) 10À4) Deferribacteraceae: 1.93 Â 10À2 (6.35 Â 10À3) 5.65 Â 10À3 (2.32 Â 3.00 Â 10À2 (6.99 Â 4.32 Â 10À4 2.06 Â 10À3 Mucispirillum 10À3) 10À3) Porphyromonadaceae: 2.89 Â 10À2 (5.44 Â 10À3) 1.13 Â 10À2 (1.66 Â 2.73 Â 10À2 (4.44 Â 8.73 Â 10À4 3.11 Â 10À3 Odoribacter 10À3) 10À3) Oscillospiraceae: Paludicola 1.04 Â 10À4 (2.56 Â 10À5) 1.21 Â 10À4 (5.75 Â 3.00 Â 10À4 (7.30 Â 1.40 Â 10À2 2.96 Â 10À2 10À5) 10À5) Bacteroidetes: 5.37 Â 10À2 (7.87 Â 10À3) 7.14 Â 10À2 (9.97 Â 3.24 Â 10À2 (7.21 Â 4.83 Â 10À33 1.25 Â 10À2 Parabacteroides 10À3) 10À3) Sutterellaceae: Parasutterella 1.45 Â 10À2 (5.56 Â 10À3) 2.46 Â 10À2 (4.57 Â 3.24 Â 10À2 (3.20 Â 5.95 Â 10À3 1.41 Â 10À2 10À3) 10À3) Peptococcaceae: 3.19 Â 10À4 (6.51 Â 10À5) 1.13 Â 10À4 (4.46 Â 2.95 Â 10À4 (8.24 Â 8.71 Â 10À3 1.91 Â 10À2 Peptococcus 10À5) 10À5) Prevotellaceae UCG 1 1.14 Â 10À3 (9.72 Â 10À4) 1.44 Â 10À2 (3.02 Â 1.11 Â 10À2 (2.07 Â 4.46 Â 10À4 2.06 Â 10À3 10À3) 10À3)

https://doi.org/10.4049/immunohorizons.2100024 640 HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA ImmunoHorizons

TABLE II. continued

Mean (Standard Error) Relative Abundance Taxa HLA-DQ8 HLA-DR3 HLA-DR3.DQ8 p Value q Value Bacteroidetes: Rikenella 6.41  10À3 (9.18  10À4) 1.50  10À2 (1.97  1.45  10À2 (1.21  1.18  10À4 8.39  10À4 10À3) 10À3) Peptostreptococcaceae: 3.20  10À3 (1.21  10À3) 3.28  10À4 (1.46  1.00  10À2 (2.44  6.94  10À7 3.67  10À5 Romboutsia 10À4) 10À3) Lachnospiraceae: Roseburia 2.55  10À3 (5.37  10À4) 4.21  10À3 (1.60  9.73  10À3 (1.56  4.69  10À4 2.06  10À3 10À3) 10À3) Streptococcaceae: 1.25  10À3 (4.37  10À4) 4.93  10À4 (9.65  1.72  10À4 (5.96  6.39  10À3 1.46  10À2 Streptococcus 10À5) 10À5) Erysipelotrichaceae: 1.73  10À2 (6.46  10À3) 2.00  10À4 (2.00  9.91  10À3 (2.25  2.13  10À6 4.05  10À5 Turicibacter 10À4) 10À3) Kruskal–Wallis test with Benjamini–Hochberg adjustment was performed on sample-scaled counts at a significance level of 0.05 to identify differentially abundant taxa. The q values are the padj. Downloaded from exacerbate existing disease. The distinct microbiome profile and intestinal damage in dextran sulfate sodium (DSS)–induced between HLA-DQ8 and HLA-DR3 bolsters the evidence that colitis mice (50, 51). In one study comparing the microbiome of interaction between host genetics and microbiome composition diabetic db/db mice with nondiabetic db/m mice, Rikenella was mayplayaprotectiveroleinMS. decreased in the gut microbiota of diabetic mice (52). Further- Desulfovibrio, Rikenella, and an uncultured genus of the Pre- more, this depletion was reversed upon oral administration of votellaceae family showed relatively higher abundances in dis- resveratrol, a compound that is shown to alleviate symptoms of http://www.immunohorizons.org/ ease-susceptible HLA-DR3 and HLA-DR3.DQ8 transgenic mice diabetic nephropathy. Another study found a decrease in Rike- but were reduced in disease-resistant HLA-DQ8 mice. Desulfo- nella in inflammatory bowel disease (IBD) patients compared vibrio has previously been shown to be associated with the MS with controls and other gastrointestinal diseases (53). The disease state (49), and other studies have shown that the abun- depletionofanunculturedPrevotellaceae family member is dance of Desulfovibrio is positively associated with symptoms interesting because Prevotella, another member of that family, by guest on September 26, 2021

FIGURE 7. Differentially abundant bacte- ria in HLA-DR3. Abundance values are sum scaled to one million. The three lines represent the lower quartile, median, and upper quartile from lowest to highest, respectively. Sig- nificance labels: *p < 0.05, **0.05 > p > 0.01, ***0.01 > p > 0.001, ****p < 0.001.

https://doi.org/10.4049/immunohorizons.2100024 ImmunoHorizons HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA 641

havemeasuredthepresenceofH. pylori using levels of serum anti–H. pylori Abs. Findings thus far suggest that H. pylori infection serves a protective role against the development of MS, and the level of seropositivity was also negatively associ- ated with disease severity (56, 57). However, studies focused on the abundance of the Helicobacter genus in the gut microbiome of patients with inflammatory diseases have demonstrated an increase in Helicobacter associated with EAE and IBD and a decrease in butyrate-producing bacteria (55, 58). In fact, H. hepaticus and H. bilis have been comparable to the colitis- inducing chemical DSS in inducing IBD in mouse models (59, 60), and H. bilis abundance was associated with progression from IBD to colorectal cancer, suggesting a possible correlation with the severity of inflammation or duration of disease (61). Similarly, Mucispirillum has been consistently associated with Downloaded from theprevalenceandseverityofDSS-inducedcolitis(50,62).The relative increase of these bacteria in both HLA-DR3.DQ8 and HLA-DQ8 suggests that the HLA-DQ8 molecule may select for bacteria with context-dependent roles disease exacerbation. Similarly, Butyricicoccus has been inversely correlated with MS http://www.immunohorizons.org/ symptoms and generally depleted in various inflammatory states including IBD (63, 64). Butyricicoccus is a known buty- rate-producer in the human gut and is thus thought to be a generally beneficial bacteria, but one study suggests that these beneficial effects vary between individuals as a result of overall microbiome composition (65). Alloprevotella, a close relative of Prevotella, was also signifi- cantly depleted in HLA-DR3 compared with the other two FIGURE 8. Differentially abundant bacteria in HLA-DR3.DQ8. groups. Multiple studies have shown Prevotella to be depleted Abundance values are sum scaled to one million. The three lines repre- in MS patients and increased after treatment with disease-mod- by guest on September 26, 2021 sent the lower quartile, median, and upper quartile from lowest to ifying drugs (18, 19). Similarly, in EAE, administration of P. his - highest, respectively. Significance labels: *p < 0.05, **p < 0.01, ticola has shown significant improvements in EAE (42, 66). ****p < 0.001. Similar findings have been shown in patients with MS, and Pre- votella has also been found to be negatively correlated with dis- ease severity (18, 19). The relationship between Alloprevotella has been characterized numerous times as being a member of a andMSismuchlessstudied,andthusfar,studieshaveshown healthy microbiome (19, 54). It is possible that these two Prevo- varying associations of this genus with inflammatory diseases tellaceae members may compete over shared resources, and a (67, 68). Romboutsia, Odoribacter,andTuricibacter were all also high abundance of Prevotella confers protection via suppression relatively decreased in HLA-DR3. Although these bacteria have of this uncultured genus. The depletion of these three genera not been studied with relation to EAE or MS, they have been suggest a potential role in the conferring susceptibility to EAE, associated with various inflammatory diseases (69, 70). in which HLA-DQ8 may protect against EAE through suppres- Given the current findings in literature and our findings in sion of these genera or the HLA-DR3 molecule may promote this study, although bacteria such as Helicobacter and Mucispir- their growth in the other two groups. illum are depleted in HLA-DR3 mice that experience moderate In the disease-resistant HLA-DQ8 group, there was a higher disease (Fig. 1), their overabundance in mice with severe dis- relative abundance of Alistipes compared with the HLA-DR3 ease (HLA-DR3.DQ8) suggest they may be correlated with dis- and HLA-DR3.DQ8 groups, and this genus has previously been ease severity. The similar abundance in HLA-DR3.DQ8 and found in decreased abundance in EAE mice (55) as well as dia- HLA-DQ8 mice further suggests that these bacteria may be betic mice (52). This suggests a potential protective role of Alis- positively selected for by the HLA-DQ8 phenotype and thus tipes in certain inflammatory conditions including EAE. may play a specific role in modulating disease severity but not Meanwhile, the HLA-DR3 mice had relatively lower abundance prevalence of the disease itself. of Helicobacter and Mucispirillum. Because of the clinical There have been limited studies connecting Enterorhabdus importance of the H. pylori in gastritis and peptic ulcer disease, and Intestinimonas, the two genera overabundant in HLA-DR3 many studies have focused on this species’ role in autoimmune mice, with autoimmune or inflammatory diseases. Two studies diseases as well, including MS and EAE. Most of these studies have isolated species of Enterorhabdus from the gut of colitis

https://doi.org/10.4049/immunohorizons.2100024 642 HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA ImmunoHorizons

FIGURE 9. SCFA metabolism and meth- ane production differ between HLA- DQ8, HLA-DR3, and HLA-DR3.DQ8 mice. Abundance values are sum scaled to one Downloaded from million. The three lines represent lower quartile, median, and upper quartile from lowest to highest, respectively. Signifi- cance labels: *p < 0.05, **0.05 > p > 0.01, ***0.01 > p > 0.001, ****p < 0.001. http://www.immunohorizons.org/ by guest on September 26, 2021 mouse models (71, 72), and a few other studies have found an demonstrated that a strain of the Erysipelotrichaceae family intro- association between Enterorhabdus and oxidative stress in mice duced into the gut of germ-free mice enhances the Th17 (73, 74). Similarly, there are very limited studies examining the response, and a strain of Lactobacillus reuteri may present pepti- relationship between Streptococcus, Roseburia, C. saccharimonas, des that mimic myelin oligodendrocyte glycoprotein (80). Addi- Paludicola,orC. sensu stricto 1 with autoimmune diseases. A few tionally, H. hepaticus (increased in HLA-DR3.DQ8 mice at genus studies have shown Streptococcus to be overabundant in patients level) promotes the development of IL-17– and IFN-g–producing with MS and other inflammatory diseases (75, 76). Roseburia has CD4Tcellsasamechanismofcausingcolitisinmice(55,81). been shown to be depleted in DSS-induced colitis and rheuma- Other species of Helicobacter, specifically H. hepaticus, can toid arthritis, and one study found that administration of R. induce severe IBD disease through IL-12 and IL-23 production intestinalis flagellin protein decreased the colitis-associated (82). Similarly, a few studies have shown Mucispirillum associ- disease-associated index (77). The epistatic interaction between ated with increased Th1 responses and the proinflammatory HLA-DR3 and HLA-DQ8 could be selecting for bacteria that cytokine MCP-1 in the pathogenesis of DSS-induced colitis in may worsen existing disease, such as Roseburia, and suppressing mice (62, 83). Additionally, Desulfovibrio is a significant contribu- bacteria that may be protecting or ameliorating disease, such as tor to hydrogen sulfide production in the gut, enabling mucosal Streptococcus. inflammation (84, 85), and has also been shown to induce a BecausewehavepreviouslyobservedthatHLA-DQ8caused Th17 response in germ-free mice (86). Thus, HLA-DQ8–selected an increase in disease severity through the production of IL-17 gut bacteria can modulate disease through the induction of and GM-CSF (14), we reasoned that gut microbiota can play an proinflammatory Th1 and Th17 pathways. Understanding how important role in the disease-modulating effect of HLA-DQ8 as HLA polymorphisms affect the abundance of immunomodula- the latter has been shown to regulate levels of Th17 cells in tory bacteria in the gut could shed light on potential targets for mice. A rapidly growing body of literature is elucidating the affecting disease pathogenesis and progression. mechanistic connections between the gut microbiome and auto- Functional profiling with PICRUSt2 revealed differential immune diseases such as MS (78, 79). Another recent study abundance of four pathways related to SCFA metabolism.

https://doi.org/10.4049/immunohorizons.2100024 ImmunoHorizons HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA 643

Interestingly, the pathways involved in generating acetate and DISCLOSURES butyrate had higher relative abundance not only in HLA-DQ8 but HLA-DR3.DQ8 as well. Butyrate is generally regarded as a A.K.M. is an inventor of the use of P. histicola for treatment of protective metabolite in inflammatory diseases (87–90). autoimmune disease, and the patent is owned by Mayo Clinic Although serum acetate has been associated with higher (Rochester, MN). The technology has been licensed by the Expanded Disability Status Scale scores in MS (91), it can be Mayo Clinic to Evelo Biosciences. A.K.M. received royalties further metabolized to butyrate by various gut bacteria (92–95). from Mayo Clinic (paid by Evelo Biosciences). The other However, the HLA-DR3.DQ8 group in this study possessed a authors have no financial conflicts of interest. uniquely high relative abundance of a pathway that uses acetate to generate methane, suggesting a potential shift toward higher REFERENCES gut methane-to-butyrate ratios compared with HLA-DQ8 and HLA-DR3. Previous studies have demonstrated increased meth- 1. Sospedra, M., and R. Martin. 2005. Immunology of multiple sclero- ane production in a subset of MS patients using breath tests sis. Annu. Rev. Immunol. 23: 683–747. 2. Steinman, L. 1995. Multiple sclerosis. Presenting an odd autoanti-

(18). Elevated gut methane has also been shown to increase Downloaded from gen. Nature. 375: 739–740. colonic transit time (96, 97) and thus posited to increase nutri- 3. Waubant, E., R. Lucas, E. Mowry, J. Graves, T. Olsson, L. Alfreds- ent absorption and heighten risk for obesity (98). Obesity is a son, and A. Langer-Gould. 2019. Environmental and genetic risk known driver of systemic inflammation and has also been factors for MS: an integrated review. Ann. Clin. Transl. Neurol. 6: linked with higher Expanded Disability Status Scale scores in 1905–1922. 4. International Multiple Sclerosis Genetics Consortium. Electronic MS patients (99). These findings together raise the possibility address: [email protected]; International Multiple Sclerosis http://www.immunohorizons.org/ that epistatic interaction between HLA class II alleles such as Genetics. 2018. Low-frequency and rare-coding variation contrib- HLA-DR3 and HLA-DQ8 could contribute to MS susceptibility utes to multiple sclerosis risk. Cell. 175: 1679–1687.e7. and severity via modulation of gut bacterial metabolic path- 5. Montgomery, T. L., A. Kunstner,€ J. J. Kennedy, Q. Fang, L. Asarian, ways. Potential relationships between inflammation and the R. Culp-Hill, A. D’Alessandro, C. Teuscher, H. Busch, and D. N. Kre- mentsov. 2020. Interactions between host genetics and gut micro- pathways uniquely increased in HLA-DQ8 have yet to be biota determine susceptibility to CNS autoimmunity. Proc. Natl. elucidated. Acad. Sci. USA. 117: 27516–27527. There are some limitations to our findings on differences in 6. Olsson, T., L. F. Barcellos, and L. Alfredsson. 2017. Interactions the microbiome among different strains. As AE-KO, HLA-DR3, between genetic, lifestyle and environmental risk factors for multi- ple sclerosis. Nat. Rev. Neurol. 13: 25–36. and HLA-DQ8 mice were maintained through inbreeding for a 7. Willer, C. J., D. A. Dyment, N. J. Risch, A. D. Sadovnick, and G. C. prolonged period, there is a possibility that the observed differ-

Ebers; Canadian Collaborative Study Group. 2003. Twin concor- by guest on September 26, 2021 ences in microbiota among these genetically modified strains dance and sibling recurrence rates in multiple sclerosis. Proc. Natl. were due to so-called strain founder effect (100, 101). Previ- Acad. Sci. USA. 100: 12877–12882. ously, Goodrich et al. (101) showed that although different 8. Ramagopalan, S. V., A. E. Handel, G. Giovannoni, S. Rutherford Siegel, G. C. Ebers, and G. Chaplin. 2011. Relationship of UV expo- TLR-deficient mice strains and their respective wild-type litter- sure to prevalence of multiple sclerosis in England. Neurology. 76: mate controls harbored distinct microbiota and post–antibiotic 1410–1414. treatment, all strains had similar microbial community. Addi- 9. Maccari, G., J. Robinson, K. Ballingall, L. A. Guethlein, U. Grimholt, tionally, microbiota of mice within the same litter were more J.Kaufman,C.S.Ho,N.G.deGroot,P.Flicek,R.E.Bontrop,etal. 2017. IPD-MHC 2.0: an improved inter-species database for the study similar than mice from different litters. Based on these results, of the major histocompatibility complex. Nucleic Acids Res. 45(D1): the study concluded that maternal transmission of the micro- D860–D864. biome is dominant over TLR signaling. However, all HLA- 10. Zivadinov, R., L. Uxa, A. Bratina, A. Bosco, B. Srinivasaraghavan, A. DR3.DQ8 transgenic mice used in the study were from F1 cross Minagar, M. Ukmar, S. Benedetto, and M. Zorzon. 2007. HLA- between HLA-DR3 to HLA-DQ8 mice. Thus, future studies DRB1*1501, -DQB1*0301, -DQB1*0302, -DQB1*0602, and -DQB1*0603 alleles are associated with more severe disease out- exploring the role of antibiotics on gut microbiota among dif- come on MRI in patients with multiple sclerosis. Int. Rev. Neuro- ferent HLA transgenic lines or rederivation of these transgenic biol. 79: 521–535. lines in a germ-free environment will be able to help in deter- 11. Dyment,D.A.,B.M.Herrera,M.Z.Cader,C.J.Willer,M.R.Lincoln, mining the precise role of strain founder effect on the composi- A. D. Sadovnick, N. Risch, and G. C. Ebers. 2005. Complex interactions tion of gut microbiota. among MHC haplotypes in multiple sclerosis: susceptibility and resis- tance. Hum. Mol. Genet. 14: 2019–2026. Previously, we have shown that HLA class II allele such as 12. Amirzargar, A., J. Mytilineos, A. Yousefipour, S. Farjadian, S. HLA-DR3 and HLA-DQ8 can modulate disease susceptibility, Scherer, G. Opelz, and A. Ghaderi. 1998. HLA class II (DRB1, DQA1 severity, and resistance through the modulation of pro- and and DQB1) associated genetic susceptibility in Iranian multiple anti-inflammatory cytokines. Future studies on dissecting the sclerosis (MS) patients. Eur. J. Immunogenet. 25: 297–301. 13. Marrosu, M. G., F. Muntoni, M. R. Murru, G. Costa, M. P. interaction between gut microbiota, immune system, cytokines, Pischedda, M. Pirastu, S. Sotgiu, G. Rosati, and C. Cianchetti. 1992. and disease phenotypes will help in better understanding the HLA-DQB1 genotype in Sardinian multiple sclerosis: evidence for a role of gut microbiome in the pathobiology of MS. key role of DQB1 *0201 and *0302 alleles. Neurology. 42: 883–886.

https://doi.org/10.4049/immunohorizons.2100024 644 HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA ImmunoHorizons

14. Mangalam, A., D. Luckey, E. Basal, M. Jackson, M. Smart, M. and human b2-microglobulin affect the gut microbiota of trans- Rodriguez, and C. David. 2009. HLA-DQ8 (DQB1*0302)-restricted genic rats. PLoS One. 9: e105684. Th17 cells exacerbate experimental autoimmune encephalomyelitis 33. Madsen, L., N. Labrecque, J. Engberg, A. Dierich, A. Svejgaard, C. in HLA-DR3-transgenic mice. J. Immunol. 182: 5131–5139. Benoist, D. Mathis, and L. Fugger. 1999. Mice lacking all conven- 15. Das, P., K. M. Drescher, A. Geluk, D. S. Bradley, M. Rodriguez, and tional MHC class II genes. Proc. Natl. Acad. Sci. USA. 96: C. S. David. 2000. Complementation between specific HLA-DR 10338–10343. and HLA-DQ genes in transgenic mice determines susceptibility to 34. Cardell, S., M. Merkenschlager, H. Bodmer, S. Chan, D. Cosgrove, experimental autoimmune encephalomyelitis. Hum. Immunol. 61: C. Benoist, and D. Mathis. 1994. The immune system of mice lack- 279–289. ing conventional MHC class II molecules. Adv. Immunol. 55: 16. Ivanov, I. I., K. Atarashi, N. Manel, E. L. Brodie, T. Shima, U. Karaoz, 423–440. D. Wei, K. C. Goldfarb, C. A. Santee, S. V. Lynch, et al. 2009. Induc- 35. Nabozny,G.H.,J.M.Baisch,S.Cheng,D.Cosgrove,M.M.Griffiths, tion of intestinal Th17 cells by segmented filamentous bacteria. Cell. H. S. Luthra, and C. S. David. 1996. HLA-DQ8 transgenic mice are 139: 485–498. highly susceptible to collagen-induced arthritis: a novel model for 17. Colpitts, S. L., E. J. Kasper, A. Keever, C. Liljenberg, T. Kirby, K. human polyarthritis. J. Exp. Med. 183: 27–37. Magori, L. H. Kasper, and J. Ochoa-Reparaz. 2017. A bidirectional 36. Abraham, R. S., Y. C. Kudva, S. B. Wilson, J. L. Strominger, and association between the gut microbiota and CNS disease in a C. S. David. 2000. Co-expression of HLA DR3 and DQ8 results in

biphasic murine model of multiple sclerosis. Gut Microbes. 8: the development of spontaneous insulitis and loss of tolerance to Downloaded from 561–573. GAD65 in transgenic mice. Diabetes. 49: 548–554. 18. Jangi, S., R. Gandhi, L. M. Cox, N. Li, F. von Glehn, R. Yan, B. Patel, 37. Strauss, G., D. A. Vignali, G. Schonrich,€ and G. J. H€ammerling. M. A. Mazzola, S. Liu, B. L. Glanz, et al. 2016. Alterations of the 1994. Negative and positive selection by HLA-DR3(DRw17) mole- human gut microbiome in multiple sclerosis. Nat. Commun. 7: cules in transgenic mice. Immunogenetics. 40: 104–108. 12015. 38. Shahi, S. K., K. Zarei, N. V. Guseva, and A. K. Mangalam. 2019. 19. Chen, J., N. Chia, K. R. Kalari, J. Z. Yao, M. Novotna, M. M. Paz Microbiota analysis using two-step PCR and next-generation 16S

Soldan, D. H. Luckey, E. V. Marietta, P. R. Jeraldo, X. Chen, et al. rRNA gene sequencing. J.Vis. Exp. (152): 10.3791/59980. http://www.immunohorizons.org/ 2016. Multiple sclerosis patients have a distinct gut microbiota 39. Douglas, G. M., V. J. Maffei, J. R. Zaneveld, S. N. Yurgel, J. R. compared to healthy controls. Sci. Rep. 6: 28484. Brown, C. M. Taylor, C. Huttenhower, and M. G. I. Langille. 2020. 20. Tremlett, H., D. W. Fadrosh, A. A. Faruqi, J. Hart, S. Roalstad, J. PICRUSt2 for prediction of metagenome functions. Nat. Biotech- Graves, S. Lynch, and E. Waubant; US Network of Pediatric MS nol. 38: 685–688. Centers. 2016. Gut microbiota composition and relapse risk in 40. Arndt,D.,J.Xia,Y.Liu,Y.Zhou,A.C.Guo,J.A.Cruz,I.Sinelnikov, pediatric MS: a pilot study. J. Neurol. Sci. 363: 153–157. K. Budwill, C. L. Nesbø, and D. S. Wishart. 2012. METAGENassist: a 21. Miyake, S., S. Kim, W. Suda, K. Oshima, M. Nakamura, T. Matsuoka, comprehensive web server for comparative metagenomics. Nucleic N. Chihara, A. Tomita, W. Sato, S. W. Kim, et al. 2015. Dysbiosis in Acids Res. 40(W1): W88–W95. the gut microbiota of patients with multiple sclerosis, with a striking 41. Dhariwal, A., J. Chong, S. Habib, I. L. King, L. B. Agellon, and J. depletion of species belonging to clostridia XIVa and IV clusters. Xia. 2017. MicrobiomeAnalyst: a web-based tool for comprehensive PLoS One. 10: e0137429. statistical, visual and meta-analysis of microbiome data. Nucleic

22. Ochoa-Reparaz, J., and L. H. Kasper. 2014. Gut microbiome and Acids Res. 45(W1): W180–W188. by guest on September 26, 2021 the risk factors in central nervous system autoimmunity. FEBS 42. Mangalam, A., S. K. Shahi, D. Luckey, M. Karau, E. Marietta, N. Lett. 588: 4214–4222. Luo, R. S. Choung, J. Ju, R. Sompallae, K. Gibson-Corley, et al. 23. Zhang, K., M. W. Hornef, and A. Dupont. 2015. The intestinal epi- 2017. Human gut-derived commensal bacteria suppress CNS thelium as guardian of gut barrier integrity. Cell. Microbiol. 17: inflammatory and demyelinating disease. Cell Rep. 20: 1269–1277. 1561–1569. 43. Mangalam, A. K., D. Luckey, S. Giri, M. Smart, L. R. Pease, M. 24. Rescigno, M. 2011. The intestinal epithelial barrier in the control Rodriguez, and C. S. David. 2012. Two discreet subsets of CD8 T of homeostasis and immunity. Trends Immunol. 32: 256–264. cells modulate PLP(91-110) induced experimental autoimmune 25. Jarchum, I., and E. G. Pamer. 2011. Regulation of innate and adap- encephalomyelitis in HLA-DR3 transgenic mice. J. Autoimmun. 38: tive immunity by the commensal microbiota. Curr. Opin. Immunol. 344–353. 23: 353–360. 44. Unanue, E. R., V. Turk, and J. Neefjes. 2016. Variations in MHC 26. Blacher, E., M. Levy, E. Tatirovsky, and E. Elinav. 2017. Micro- class II antigen processing and presentation in health and disease. biome-modulated metabolites at the interface of host immunity. J. Annu. Rev. Immunol. 34: 265–297. Immunol. 198: 572–580. 45. Barcellos, L. F., J. R. Oksenberg, A. B. Begovich, E. R. Martin, S. 27. Forbes, J. D., G. Van Domselaar, and C. N. Bernstein. 2016. The gut Schmidt, E. Vittinghoff, D. S. Goodin, D. Pelletier, R. R. Lincoln, P. microbiota in immune-mediated inflammatory diseases. Front. Bucher, et al; Multiple Sclerosis Genetics Group. 2003. HLA-DR2 Microbiol. 7: 1081. dose effect on susceptibility to multiple sclerosis and influence on 28. Kåhrstrom,€ C. T., N. Pariente, and U. Weiss. 2016. Intestinal disease course. Am. J. Hum. Genet. 72: 710–716. microbiota in health and disease. Nature. 535: 47. 46. Mangalam, A., D. Luckey, E. Basal, M. Behrens, M. Rodriguez, and 29. Kinross, J. M., A. W. Darzi, and J. K. Nicholson. 2011. Gut micro- C. David. 2008. HLA-DQ6 (DQB1*0601)-restricted T cells protect biome-host interactions in health and disease. Genome Med. 3: 14. against experimental autoimmune encephalomyelitis in HLA- 30. Khan, M. A. W., W. Z. Stephens, A. D. Mohammed, J. L. Round, DR3.DQ6 double-transgenic mice by generating anti-inflammatory and J. L. Kubinak. 2019. Does MHC heterozygosity influence IFN-gamma. J. Immunol. 180: 7747–7756. microbiota form and function? PLoS One. 14: e0215946. 47. Luckey, D., D. Bastakoty, and A. K. Mangalam. 2011. Role of HLA 31. Kubinak, J. L., W. Z. Stephens, R. Soto, C. Petersen, T. Chiaro, L. class II genes in susceptibility and resistance to multiple sclerosis: Gogokhia, R. Bell, N. J. Ajami, J. F. Petrosino, L. Morrison, et al. 2015. studies using HLA transgenic mice. J. Autoimmun. 37: 122–128. MHC variation sculpts individualized microbial communities that 48. Mangalam, A. K., V. Taneja, and C. S. David. 2013. HLA class II control susceptibility to enteric infection. Nat. Commun. 6: 8642. molecules influence susceptibility versus protection in inflamma- 32. Lin, P., M. Bach, M. Asquith, A. Y. Lee, L. Akileswaran, P. Stauffer, tory diseases by determining the cytokine profile. J. Immunol. 190: S. Davin, Y. Pan, E. D. Cambronne, M. Dorris, et al. 2014. HLA-B27 513–518.

https://doi.org/10.4049/immunohorizons.2100024 ImmunoHorizons HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA 645

49. Zhang, C., M. Zhang, S. Wang, R. Han, Y. Cao, W. Hua, Y. Mao, X. 64. Facchin, S., N. Vitulo, M. Calgaro, A. Buda, C. Romualdi, D. Pohl, Zhang, X. Pang, C. Wei, et al. 2010. Interactions between gut B. Perini, G. Lorenzon, C. Marinelli, R. D’Inca, et al. 2020. Micro- microbiota, host genetics and diet relevant to development of met- biota changes induced by microencapsulated sodium butyrate in abolic syndromes in mice. [Published erratum appears in 2010 patients with inflammatory bowel disease. Neurogastroenterol. ISME J. 4:312–313.] ISME J. 4: 232–241. Motil. 32: e13914. 50. Song, H., W. Wang, B. Shen, H. Jia, Z. Hou, P. Chen, and Y. Sun. 65. Geirnaert, A., J. Wang, M. Tinck, A. Steyaert, P. Van den Abbeele, 2018. Pretreatment with probiotic Bifico ameliorates colitis-associ- V. Eeckhaut, R. Vilchez-Vargas, G. Falony, D. Laukens, M. De Vos, ated cancer in mice: Transcriptome and gut flora profiling. Cancer et al. 2015. Interindividual differences in response to treatment Sci. 109: 666–677. with butyrate-producing Butyricicoccus pullicaecorum 25-3T 51. Sun, J., H. Chen, J. Kan, Y. Gou, J. Liu, X. Zhang, X. Wu, S. Tang, studied in an in vitro gut model. FEMS Microbiol. Ecol. 91: fiv054. R. Sun, C. Qian, et al. 2020. Anti-inflammatory properties and gut 66.Shahi,S.K.,S.N.Jensen,A.C.Murra,N.Tang,H.Guo,K.N. microbiota modulation of an alkali-soluble polysaccharide from Gibson-Corley,J.Zhang,N.J.Karandikar,J.A.Murray,and purple sweet potato in DSS-induced colitis mice. Int. J. Biol. Mac- A. K. Mangalam. 2020. Human commensal Prevotella histicola romol. 153: 708–722. ameliorates disease as effectively as interferon-beta in the experi- 52. Cai, T. T., X. L. Ye, R. R. Li, H. Chen, Y. Y. Wang, H. J. Yong, M. L. mental autoimmune encephalomyelitis. Front. Immunol. 11: Pan, W. Lu, Y. Tang, H. Miao, et al. 2020. Resveratrol modulates 578648.

the gut microbiota and inflammation to protect against diabetic 67. Ren, W., S. Chen, L. Zhang, G. Liu, T. Hussain, X. Hao, J. Yin, J. Downloaded from nephropathy in mice. Front. Pharmacol. 11: 1249. Duan, B. Tan, G. Wu, et al. 2016. Interferon tau affects mouse 53. Lopetuso, L. R., V. Petito, C. Graziani, E. Schiavoni, F. Paroni Sterbini, intestinal microbiota and expression of IL-17. Mediators Inflamm. A. Poscia, E. Gaetani, F. Franceschi, G. Cammarota, M. Sanguinetti, 2016: 2839232. et al. 2018. Gut microbiota in health, diverticular disease, irritable 68. Sun, Y., Q. Chen, P. Lin, R. Xu, D. He, W. Ji, Y. Bian, Y. Shen, Q. Li, bowel syndrome, and inflammatory bowel diseases: time for micro- C. Liu, et al. 2019. Characteristics of gut microbiota in patients bial marker of gastrointestinal disorders. Dig. Dis. 36: 56–65. with rheumatoid arthritis in Shanghai, China. Front. Cell. Infect. 54. Shahi, S. K., S. N. Freedman, A. C. Murra, K. Zarei, R. Sompallae, Microbiol. 9: 369. http://www.immunohorizons.org/ K. N. Gibson-Corley, N. J. Karandikar, J. A. Murray, and A. K. 69. He, H., H. Xu, J. Xu, H. Zhao, Q. Lin, Y. Zhou, and Y. Nie. 2020. Mangalam. 2019. Prevotella histicola, a human gut commensal, is Sodium butyrate ameliorates gut microbiota dysbiosis in lupus-like as potent as COPAXONEVR in an animal model of multiple sclero- mice. Front. Nutr. 7: 604283. sis. Front. Immunol. 10: 462. 70. El Mouzan, M. I., H. S. Winter, A. A. Assiri, K. S. Korolev, A. A. Al 55. Li, X. L., B. Zhang, M. J. Sun, C. C. Bao, B. Y. Yuan, Q. F. Xie, L. J. Sarkhy, S. E. Dowd, M. A. Al Mofarreh, and R. Menon. 2018. Wang, and M. X. Wang. 2019. Mechanism of gut microbiota and Microbiota profile in new-onset pediatric Crohn’s disease: data Axl/SOCS3 in experimental autoimmune encephalomyelitis. Bio- from a non-Western population. Gut Pathog. 10: 49. sci. Rep. 39: BSR20190228. 71. Clavel, T., C. Charrier, A. Braune, M. Wenning, M. Blaut, and D. 56. Yao, G., P. Wang, X. D. Luo, T. M. Yu, R. A. Harris, and X. M. Haller. 2009. Isolation of bacteria from the ileal mucosa of Zhang. 2016. Meta-analysis of association between Helicobacter TNFdeltaARE mice and description of Enterorhabdus mucosicola pylori infection and multiple sclerosis. Neurosci. Lett. 620: 1–7. gen. nov., sp. nov. Int. J. Syst. Evol. Microbiol. 59: 1805–1812. by guest on September 26, 2021 57. Ranjbar, R., S. Karampoor, and F. A. Jalilian. 2019. The protective 72. Clavel, T., W. Duck, C. Charrier, M. Wenning, C. Elson, and D. effect of Helicobacter pylori infection on the susceptibility of mul- Haller. 2010. Enterorhabdus caecimuris sp. nov., a member of the tiple sclerosis. J. Neuroimmunol. 337: 577069. family Coriobacteriaceae isolated from a mouse model of sponta- 58. Yu, W., X. Su, W. Chen, X. Tian, K. Zhang, G. Guo, L. Zhou, T. neous colitis, and emended description of the genus Enterorhab- Zeng, and B. Han. 2019. Three types of gut bacteria collaborating dus Clavel et al. 2009. Int. J. Syst. Evol. Microbiol. 60: 1527–1531. to improve Kui Jie’an enema treat DSS-induced colitis in mice. 73. Song, R., J. Yao, Q. Shi, and R. Wei. 2018. Nanocomposite of half- Biomed. Pharmacother. 113: 108751. fin anchovy hydrolysates/zinc oxide nanoparticles exhibits actual 59. Willis, C. R., A. Seamons, J. Maxwell, P. M. Treuting, L. Nelson, G. non-toxicity and regulates intestinal microbiota, short-chain fatty Chen, S. Phelps, C. L. Smith, T. Brabb, B. M. Iritani, and L. Mag- acids production and oxidative status in mice. Mar. Drugs. 16: 23. gio-Price. 2012. Interleukin-7 receptor blockade suppresses adap- 74. Feng, Y. L., G. Cao, D. Q. Chen, N. D. Vaziri, L. Chen, J. Zhang, M. tive and innate inflammatory responses in experimental colitis. J. Wang, Y. Guo, and Y. Y. Zhao. 2019. Microbiome-metabolomics Inflamm. (Lond.). 9: 39. reveals gut microbiota associated with glycine-conjugated metabo- 60. Han, Q., T. J. Y. Kono, C. G. Knutson, N. M. Parry, C. L. Seiler, lites and polyamine metabolism in chronic kidney disease. Cell. J. G. Fox, S. R. Tannenbaum, and N. Y. Tretyakova. 2020. Multi- Mol. Life Sci. 76: 4961–4978. omics characterization of inflammatory bowel disease-induced 75. Zeng, Q., X. Junli Gong, C. Liu, X. Chen, H. Sun, Y. Li, C. Zhou, Y. hyperplasia/dysplasia in the Rag2-/-/Il10-/- mouse model. Int. J. Cui, Y. Wang, Yang, et al. 2019. Gut dysbiosis and lack of short Mol. Sci. 22: 364. chain fatty acids in a Chinese cohort of patients with multiple scle- 61. Peng, W., H. Li, Y. Xu, L. Yan, Z. Tang, A. Hossein Mohseni, S. rosis. Neurochem. Int. 129: 104468. Taghinezhad-S, X. Tang, and X. Fu. 2020. Association of Helico- 76. Schepici, G., S. Silvestro, P. Bramanti, and E. Mazzon. 2019. The bacter bilis infection with the development of colorectal cancer. gut microbiota in multiple sclerosis: an overview of clinical trials. Nutr. Cancer. DOI: 10.1080/01635581.2020.1862253. Cell Transplant. 28: 1507–1527. 62. Caruso, R., T. Mathes, E. C. Martens, N. Kamada, A. Nusrat, N. 77. Wu, X., S. Pan, W. Luo, Z. Shen, X. Meng, M. Xiao, B. Tan, K. Nie, Inohara, and G. Nunez.~ 2019. A specific gene-microbe interaction T. Tong, and X. Wang. 2020. Roseburia intestinalis-derived flagel- drives the development of Crohn’s disease-like colitis in mice. Sci. lin ameliorates colitis by targeting miR-223-3p-mediated activation Immunol. DOI: 10.1126/sciimmunol.aaw4341. of NLRP3 inflammasome and pyroptosis. Mol. Med. Rep. 22: 63. Reynders, T., L. Devolder, M. Valles-Colomer, A. Van Remoortel, 2695–2704. M. Joossens, J. De Keyser, G. Nagels, M. D’hooghe, and J. Raes. 78. Shahi, S. K., S. N. Freedman, and A. K. Mangalam. 2017. Gut micro- 2020. Gut microbiome variation is associated to Multiple Sclerosis biome in multiple sclerosis: The players involved and the roles phenotypic subtypes. Ann. Clin. Transl. Neurol. 7: 406–419. they play. Gut Microbes. 8: 607–615.

https://doi.org/10.4049/immunohorizons.2100024 646 HLA CLASS II POLYMORPHISM AND GUT MICROBIOTA ImmunoHorizons

79. Marietta, E., A. K. Mangalam, V. Taneja, and J. A. Murray. 2020. 91. Perez-Perez, S., M. I. Domínguez-Mozo, A. Alonso-Gomez, S. Intestinal dysbiosis in, and enteral bacterial therapies for, systemic Medina, N. Villarrubia, J. I. Fernandez-Velasco, M. A. García-Mar- autoimmune diseases. Front. Immunol. 11: 573079. tínez, E. García-Calvo, H. Estevez, L. Costa-Frossard, et al. 2020. 80. Miyauchi, E., S. W. Kim, W. Suda, M. Kawasumi, S. Onawa, N. Acetate correlates with disability and immune response in multiple Taguchi-Atarashi, H. Morita, T. D. Taylor, M. Hattori, and H. sclerosis. PeerJ 8: e10220. Ohno. 2020. Gut microorganisms act together to exacerbate 92. Duncan, S. H., A. Barcenilla, C. S. Stewart, S. E. Pryde, and H. J. inflammation in spinal cords. Nature. 585: 102–106. Flint. 2002. Acetate utilization and butyryl coenzyme A (CoA):ace- 81.Morrison,P.J.,D.Bending,L.A.Fouser,J.F.Wright,B.Stockinger, tate-CoA transferase in butyrate-producing bacteria from the A. Cooke, and M. C. Kullberg. 2013. Th17-cell plasticity in Helico- human large intestine. Appl. Environ. Microbiol. 68: 5186–5190. bacter hepaticus-induced intestinal inflammation. Mucosal Immunol. 93. Diez-Gonzalez, F., D. R. Bond, E. Jennings, and J. B. Russell. 1999. 6: 1143–1156. Alternative schemes of butyrate production in Butyrivibrio fibri- 82. Kullberg, M. C., D. Jankovic, C. G. Feng, S. Hue, P. L. Gorelick, solvens and their relationship to acetate utilization, lactate produc- B. S. McKenzie, D. J. Cua, F. Powrie, A. W. Cheever, K. J. Maloy, tion, and phylogeny. Arch. Microbiol. 171: 324–330. and A. Sher. 2006. IL-23 plays a key role in Helicobacter hepati- 94. den Besten, G., K. van Eunen, A. K. Groen, K. Venema, D. J. Reijngoud, cus-induced T cell-dependent colitis. J. Exp. Med. 203: 2485–2494. and B. M. Bakker. 2013. The role of short-chain fatty acids in the inter- 83. Shintouo, C. M., T. Mets, D. Beckwee, I. Bautmans, S. M. Gho- play between diet, gut microbiota, and host energy metabolism. J. Lipid gomu, J. Souopgui, L. Leemans, H. D. Meriki, and R. Njemini. Res. 54: 2325–2340. 2020. Is inflammageing influenced by the microbiota in the aged 95. Duncan, S. H., G. Holtrop, G. E. Lobley, A. G. Calder, C. S. Stewart, Downloaded from gut? A systematic review. Exp. Gerontol. 141: 111079. and H. J. Flint. 2004. Contribution of acetate to butyrate formation 84. Guo, J., X. Han, J. Zhan, Y. You, and W. Huang. 2018. Vanillin alle- by human faecal bacteria. Br. J. Nutr. 91: 915–923. viates high fat diet-induced obesity and improves the gut micro- 96. Jahng, J., I. S. Jung, E. J. Choi, J. L. Conklin, and H. Park. 2012. biota composition. Front. Microbiol. 9: 2733. The effects of methane and hydrogen gases produced by enteric 85. Kushkevych, I., D. Dordevic, and M. Vítezova. 2020. Possible syn- bacteria on ileal motility and colonic transit time. Neurogastroen- ergy effect of hydrogen sulfide and acetate produced by sulfate- terol Motil. 24: 185–190, e92. http://www.immunohorizons.org/ reducing bacteria on inflammatory bowel disease development. J. 97. Pimentel, M., H. C. Lin, P. Enayati, B. van den Burg, H. R. Lee, Adv. Res. 27: 71–78. J. H. Chen, S. Park, Y. Kong, and J. Conklin. 2006. Methane, a gas 86. Figliuolo, V. R., L. M. Dos Santos, A. Abalo, H. Nanini, A. Santos, produced by enteric bacteria, slows intestinal transit and augments N. M. Brittes, C. Bernardazzi, H. S. P. de Souza, L. Q. Vieira, R. small intestinal contractile activity. Am. J. Physiol. Gastrointest. Coutinho-Silva, and C. M. L. M. Coutinho. 2017. Sulfate-reducing Liver Physiol. 290: G1089–G1095. bacteria stimulate gut immune responses and contribute to inflam- 98. Basseri, R. J., B. Basseri, M. Pimentel, K. Chong, A. Youdim, K. mation in experimental colitis. Life Sci. 189: 29–38. Low, L. Hwang, E. Soffer, C. Chang, and R. Mathur. 2012. Intesti- 87. Archer, S., S. Meng, J. Wu, J. Johnson, R. Tang, and R. Hodin. nal methane production in obese individuals is associated with a 1998. Butyrate inhibits colon carcinoma cell growth through two higher body mass index. Gastroenterol. Hepatol. (N. Y.). 8: 22–28. distinct pathways. Surgery 124: 248–253. 99. Stampanoni Bassi, M., E. Iezzi, F. Buttari, L. Gilio, I. Simonelli, F. 88. Christl, S. U., H. D. Eisner, G. Dusel, H. Kasper, and W. Scheppach. Carbone, T. Micillo, V. De Rosa, F. Sica, R. Furlan, et al. 2020. Obe- 1996. Antagonistic effects of sulfide and butyrate on proliferation sity worsens central inflammation and disability in multiple sclero- by guest on September 26, 2021 of colonic mucosa: a potential role for these agents in the patho- sis. Mult. Scler. 26: 1237–1246. genesis of ulcerative colitis. Dig. Dis. Sci. 41: 2477–2481. 100. Ubeda, C., L. Lipuma, A. Gobourne, A. Viale, I. Leiner, M. Equinda, 89. Csordas, A. 1996. Butyrate, aspirin and colorectal cancer. Eur. J. R. Khanin, and E. G. Pamer. 2012. Familial transmission rather than Cancer Prev. 5: 221–231. defective innate immunity shapes the distinct intestinal microbiota 90. Inan,M.S.,R.J.Rasoulpour,L.Yin,A.K.Hubbard,D.W.Rosenberg, of TLR-deficient mice. J. Exp. Med. 209: 1445–1456. and C. Giardina. 2000. The luminal short-chain fatty acid butyrate 101. Goodrich, J. K., S. C. Di Rienzi, A. C. Poole, O. Koren, W. A. Walters, modulates NF-kappaB activity in a human colonic epithelial cell line. J.G.Caporaso,R.Knight,andR.E.Ley.2014.Conductingamicro- Gastroenterology. 118: 724–734. biome study. Cell. 158: 250–262.

https://doi.org/10.4049/immunohorizons.2100024

Supplemental Figure 1. Beta-diversity clustering of all four mouse groups stratified by sex | PCA plot comparing beta-diversity of HLA-DR3, HLA-DQ8, HLA-DR3.DQ8 transgenic mice and AE-KO control mice at the genus level demonstrated similar microbiota compositions in both sexes within each group. Axis 1 and 2 are the first and second principal components, respectively, and represent the most and second-most amount of variation in the bacterial abundance data between all the samples.