Diabetes Volume 63, August 2014 2821

Camilla Hartmann Friis Hansen,1 Łukasz Krych,2 Karsten Buschard,3 Stine B. Metzdorff,1 Christine Nellemann,4 Lars H. Hansen,5 Dennis S. Nielsen,2 Hanne Frøkiær,1 Søren Skov,1 and Axel K. Hansen1

A Maternal Gluten-Free Diet Reduces Inflammation and Diabetes Incidence in the Offspring of NOD Mice

Diabetes 2014;63:2821–2832 | DOI: 10.2337/db13-1612

Early-life interventions in the intestinal environment gluten-free (GF) diet decreased the incidence of diabetes have previously been shown to influence diabetes in- from 64% to 15% when nonobese diabetic (NOD) mice cidence. We therefore hypothesized that a gluten-free were fed a GF diet after weaning (1), and eating a GF diet

(GF) diet, known to decrease the incidence of type 1 decreased the incidence of diabetes to just 6% in the off- PATHOPHYSIOLOGY diabetes, would protect against the development of spring in two generations, which indicates that the inter- diabetes when fed only during the pregnancy and lactation play between gut antigens and immune pathways leading period. Pregnant nonobese diabetic (NOD) mice were to diabetes is particularly important in the preweaning fed a GF or standard diet until all pups were weaned to period when insulitis starts to progress (2). a standard diet. The early-life GF environment dramat- Accumulating evidence suggests that gut immune ically decreased the incidence of diabetes and insulitis. reactivity is skewed in human and murine diabetic Gut microbiota analysis by 16S rRNA gene sequencing patients. Studies in young human patients with T1D revealed a pronounced difference between both moth- have demonstrated increased numbers of interferon-g ers and their offspring on different diets, characterized – – Akkermansia (IFN-g) producing, interleukin (IL)-1a producing, and by increased numbers of , Proteobacteria, – and TM7 in the GF diet group. In addition, pancreatic IL-4 producing cells in the small intestinal lamina propria, fl forkhead box P3 regulatory T cells were increased in re ecting T1D preceded by intestinal immune activation GF-fed offspring, as were M2 macrophage gene markers (3). Similarly in NOD mice, a diabetes-promoting diet fl and tight junction–related genes in the gut, while intesti- induced proin ammatory cytokines IFN-g and tumor ne- nal gene expression of proinflammatory cytokines was crosis factor-a in the small intestinal lamina propria (4), reduced. An increased proportion of T cells in the pan- and an antidiabetogenic diet decreased the high numbers + + creas expressing the mucosal integrin a4b7 suggests of CD11b CD11c dendritic cells (DCs) found in the colon that the mechanism involves increased trafficking of lamina propria (5). Under germ-free conditions, reduced gut-primed immune cells to the pancreas. In conclusion, expression of forkhead box P3 (FoxP3) in the ileum, colon, a GF diet during fetal and early postnatal life reduces and the draining lymph node was associated with acceler- the incidence of diabetes. The mechanism may involve ated development of insulitis in NOD mice (6), and, likewise changes in gut microbiota and shifts to a less proinflam- in humans, Badami et al. (7) found that jejunal biopsy matory immunological milieu in the gut and pancreas. samples from T1D patients showed reduced frequency of 2 CD4+CD25+FoxP3+CD127 regulatory T cells (Tregs). The Gluten has previously been shown to affect the devel- link between the gut and pancreas has also been empha- opment of type 1 diabetes (T1D) in animal models. A sized in studies demonstrating that pancreatic islet T cells

1Department of Veterinary Disease Biology, Faculty of Health and Medical Sci- 5Department of Biology, Faculty of Science, University of Copenhagen, Copenha- ences, University of Copenhagen, Frederiksberg, Denmark gen, Denmark 2 Department of Food Science, Faculty of Science, University of Copenhagen, Corresponding author: Camilla Hartmann Friis Hansen, [email protected]. Frederiksberg, Denmark Received 18 October 2013 and accepted 27 March 2014. 3Bartholin Institute, Rigshospitalet, Copenhagen, Denmark 4Division of Toxicology and Risk Assessment, National Food Institute, Technical © 2014 by the American Diabetes Association. Readers may use this article as University of Denmark, Søborg, Denmark long as the work is properly cited, the use is educational and not for profit, and the work is not altered. 2822 Diabetes-Protective Diet Across Two Generations Diabetes Volume 63, August 2014 express gut homing receptor a4b7 integrin, which recog- at 4 weeks of age to the STD diet. Ten pups were killed nizes mucosal addressin cell adhesion molecule-1 in the from each group at 4 weeks of age, and 10 mice from each pancreas (8,9). group were killed at 10 weeks of age. The remaining 30 Failure in immune tolerance leading to pancreatic mice in each group were killed when a diagnosis of di- b-cell depletion has been suggested to be regulated in abetes was made or at 30 weeks of age, when the study part by gliadin-induced intestinal enteropathy and innate ended. Measurements of tail blood glucose levels were immune responses (10,11). However, altered gut micro- made twice a week from 10 weeks of age, and a mouse biota, previously demonstrated in GF-fed versus gluten- was considered to be diabetic when blood glucose levels fed mice (12), might also contribute to modify intestinal exceeded 12 mmol/L on 2 consecutive days. The body inflammation and development of autoimmune diabetes. weight of all offspring was monitored once a week. In support of this, impaired oral tolerance to intestinal Histology microbes was demonstrated in NOD mice (5), and the Hematoxylin-eosin–stained pancreas sections were evalu- fi impact of microbes has also been veri ed in germ-free ated for insulitis score in a blinded fashion by two per- (13), antibiotic-treated (14,15), and probiotic-treated di- sons. Lymphocytic infiltration was graded as follows: 0, abetes-prone rodent models (16). It seems reasonable to no infiltration; 1, intact islets but with few mononuclear assume that gluten and certain microbes have a synergistic cells surrounding the islets; 2, peri-insulitis; 3, islet in- effect on the development of T1D, as was also recently filtration ,50%; and 4, islet infiltration .50%. Twenty- fi suggested by Patrick et al. (17). Cytokine pro les of gut- five islets were scored for each nondiabetic mouse killed at associated lymphoid tissue have revealed a strong associa- 10 and 30 weeks of age. tion between intestinal IFN-g production and the incidence of diabetes, especially in several gluten intervention stud- Gut Microbiota ies (4,11,17–20). Also, type 1 T-helper cells proliferated Feces samples aseptically obtained from the mothers specifically in the mesenteric lymph node (MLN) in re- during pregnancy and from the offspring at 4 and 10 sponse to wheat protein antigens (19). A GF diet was weeks of age when they were killed were analyzed by PCR fi furthermore shown to reverse this shift in gut homeosta- ampli cation of the V3 region of the 16S rRNA gene foll- sis toward an anti-inflammatory state with more trans- owed by denaturing gradient gel electrophoresis (DGGE) fi forming growth factor-b (TGF-b)–producing T cells (18). as described previously (21). The resulting DGGE pro les As early-life interventions in the intestinal environ- were analyzed using BioNumerics version 4.5 (Applied ment can influence the incidence of diabetes, we hypoth- Maths, Sint-Martens-Latem, Belgium). The composition esized that a GF diet exclusively fed to mice during of the prokaryotic community of feces samples from gestation and lactation would be sufficient to protect the the mothers and the 4-week-old pups was determined offspring from the development of diabetes even though using tag-encoded 454/FLX Titanium (Roche) pyrose- they were weaned to a standard gluten-containing (STD) quencing of the V3 and V4 regions of the 16S rRNA diet. We hypothesized that the dietary protective effect gene by the National High Throughput DNA Sequen- would be partly mediated by a shift in the gut microbiota, cing Centre, University of Copenhagen, Copenhagen, and that this shift is of imperative importance in the first Denmark (22), and was analyzed as described by Krych period of life during which the immune system develops. et al. (23). An open source software package, Quantitative Insight Into Microbial Ecology (QIIME version 1.7.0) was RESEARCH DESIGN AND METHODS used to analyze the pyrosequencing data (National Cen- ter for Biotechnology Information database accession The experiment was performed in accordance with the #PRJNA215143). Principal coordinate analysis (PCoA) Council of Europe Convention European Treaty Series was made using the jackknife_beta_diversity.py workflow 123 on the Protection of Vertebrate Animals used for (the –e value: 2,000 sequences). The PCoA plot including Experimental and Other Scientific Purposes, and the bacterial taxa was drawn out with the make_3d_plots.py Danish Animal Experimentation Act (LBK 1306 from 23 script based on the summary information of bacterial November 2007). The study was approved by the Animal phyla, and the differences in the taxa relative distribution Experiments Inspectorate, Ministry of Justice, Denmark. between categories were tested with Metastats (24) Animals and Diet independently for both the phylum-level and the genus- NOD/BomTac mice (Taconic, Hudson, NY) were fed ad level summarized taxa. The P value was calculated based libitum either a GF modified Altromin diet or an STD on 1,000 permutations. Altromin diet (Altromin, Lage, Germany), as described by Cell Isolation and Flow Cytometry Funda et al. (1). The two groups were mated separately, Single-cell suspensions from spleen, MLN, and pancreatic and their female offspring were group-housed (five mice/ lymph node (PLN) isolated from 4- and 10-week-old cage) in our barrier-protected rodent facility (Faculty of offspring immediately upon their being killed, and flow Health and Medical Sciences, University of Copenhagen, cytometric analyses of CD11b+CD11c+ DCs and T-cell Frederiksberg, Denmark) under standard conditions in populations, including FoxP3+ Tregs, were performed as open cages without filter lids. All offspring were weaned previously described (22). All antibodies were purchased diabetes.diabetesjournals.org Hansen and Associates 2823 from eBiosciences (San Diego, CA). The analyses were per- formed using an Accuri C6 flow cytometer (Accuri Cytom- eters Inc., Ann Arbor, MI).

Quantitative PCR Immediately after the mice were killed, 1-cm fragments of the ileum and colon were placed in RNAlater (Ambion, Austin, TX), after all luminal content was scraped out of the gut. Homogenization, RNA isolation with MagMAX- 96 RNA Isolation Kit (Ambion), and cDNA synthesis using the High-Capacity cDNA Reverse Transcriptase Kit (Ap- plied Biosystems, Foster City, CA) were performed as described previously (25). An inventoried TaqMan Mouse Immune Array (Appled Biosystems) containing 90 TaqMan gene expression assays of immune-related genes was used to investigate ileal samples isolated at weaning, which were analyzed as described previously (26). Actinb, Ocln, Tjp1, Cldn8, Cldn15, Muc1,andMuc2 TaqMan gene expression assays (Applied Biosystems) were used for quantitative PCR (qPCR) analyses on ileum and colon cDNA isolated at 4 weeks of age, and the data were analyzed as described previously (25). cDNA samples from the ileum and colon collected at weaning were further analyzed for the presence of Akkermansia muciniphila,whichwasquantified in dupli- cate using the 7500 Fast Real-time PCR System (Applied Biosystems), as previously described (27). Statistical Analysis — GraphPad Prism version 5.02 (GraphPad Software, San Figure 1 A: Cumulative diabetes incidence in female NOD off- spring of mice fed an STD diet (n = 37, blue) or a GF diet (n = 36, Diego, CA) was used for statistical analysis, and P values red). All offspring were weaned to an STD diet at 4 weeks of age and ,0.05 were considered to be significant. Cumulative di- were diagnosed as diabetic and killed when blood glucose levels abetes incidence was calculated using the Kaplan-Meier exceeded 12 mmol/L on 2 consecutive days. Comparisons of the fi two survival curves were tested by log-rank test, and the P value is estimation, whereas statistical signi cance was evaluated shown. B: Average insulitis score for offspring of STD and GF diet– by the log-rank test. Other differences were estimated fed NOD mice at 10 weeks of age that were weaned to an STD diet. by two-tailed t test or one-way ANOVA test with Tukey C: Percentage of islets with a given score in 10-week-old offspring post test. of STD diet–fed (n = 9) and GF diet–fed (n = 8) NOD mice. D: Average insulitis score for nondiabetic offspring of STD and GF – RESULTS diet fed NOD mice at 30 weeks of age. E: Percentage of islets with a given score in 30-week-old nondiabetic offspring of STD A Maternal GF Diet Attenuates Diabetes in the diet–fed (n = 8) and GF diet–fed (n = 10) NOD mice. White, no Offspring infiltration; light gray, few mononuclear cells infiltrated; gray, peri- < fi > fi In this study, it was demonstrated that feeding a GF diet insulitis; dark gray, 50% islet in ltration; black, 50% islet in l- tration. Error bars represent the SEM. *P < 0.05, **P < 0.01. to pregnant NOD mice significantly reduced the cumu- lative diabetes incidence (P , 0.01) and increased the onset time (P , 0.01) in their offspring, even though all pups were weaned to an STD diet at 4 weeks of age (Fig. GF Diet Leads to a Gut Microbiota Enriched in 1A). The diabetes incidence at 210 days was 51% (n =37) Verrucomicrobia, Proteobacteria, and TM7 in Dams in the offspring of STD diet–fedmice,and22%(n =36) and Offspring in the offspring of GF diet–fed mice. Histological evalu- Gut microbiota analysis by DGGE demonstrated a differ- ation of insulitis in pancreatic sections from nondiabetic ence in the fecal gut microbiota between the two groups offspring revealed a significantly lower insulitis score in of pregnant NOD mice (Fig. 2A). ANOVA based on the offspring of GF diet–fedmicecomparedwiththeoff- first (X), second (Y), and third (Z) principal component spring of STD diet–fed mice at both 10 weeks (P , (PC) revealed a significant difference in PC2 values (P , 0.05; Fig. 1B and C) and 30 weeks of age (P , 0.05; 0.05), and a tendency to cluster in PC1 (P = 0.07) and PC3 Fig. 1D and E). No significant difference in body weight values (P = 0.08). The separate clustering on the PC anal- gain was observed between the two groups of nondia- ysis plot was also evident in their offspring at weaning, betic NOD mice within the observational period (data confirming that parental microbiomes altered by diet are not shown). inheritable (28); significant differences in PC1 values (P , 2824 Diabetes-Protective Diet Across Two Generations Diabetes Volume 63, August 2014

Figure 2—A: PC analysis plot based on DGGE profiles of 16S rRNA gene PCR–derived amplicons of feces samples collected from NOD offspring of mice fed an STD diet (n = 16, dark blue) or a GF diet (n = 16, red) at 4 weeks of age. STD diet–fed mothers (n = 6, light blue) and GF diet–fed mothers (n = 5, yellow) are also illustrated. B: PC analysis plot based on DGGE of feces collected from the two groups of offspring at 10 weeks of age, which were weaned to an STD diet at 4 weeks of age. ANOVA based on the X, Y, and Z PC analysis was used to compare the groups. C: PCoA plot of 16S rRNA gene tag–encoded pyrosequencing reads based on the weighted UniFrac distance matrix showing clustering of the two groups of offspring. The gray spherical coordinates indicating bacterial phyla are plotted as a weighted average of sample coordinates. The size of each sphere is proportional to the mean relative abundance among all plotted samples. D: The diabetes.diabetesjournals.org Hansen and Associates 2825

Table 1—Phyla relative distribution in feces samples collected from 4-week-old offspring of STD diet–fed and GF diet–fed NOD mice Taxa STD diet mean abundance GF diet mean abundance P value* q value* Cyanobacteria 1.276 0.001 0.001 0.009 Verrucomicrobia 0.686 7.811 0.002 0.016 Proteobacteria 0.176 0.969 0.001 0.016 TM7 0.012 6.062 0.001 0.009 Firmicutes 54.510 49.030 0.501 1.000 Bacteroidetes 42.545 35.747 0.362 1.000 Deferribacteres 0.544 0.055 0.029 0.139 Tenericutes 0.057 0.007 0.195 0.669 Actinobacteria 0.005 0.000 0.086 0.407 *Values were calculated with Metastats using 1,000 permutations.

0.05) and in PC3 values (P , 0.001) were evident between that the difference observed between the groups was ma- the two groups of offspring. However, how much of this inly due to a significantly expanded representation of the difference in microbiota is due to vertical transfer from bacterial phyla Verrucomicrobia, TM7, and Proteobacteria mothers to pups or to early ingestion of the GF diet by the in the mothers that were eating a GF diet and their pups pups is not known. Gut microbiota analysis of feces from compared with STD diet–fed mice (Fig. 2D). In addition, offspring at 10 weeks of age, 6 weeks after weaning to the the phylum Cyanobacteria was found in approximately STD diet, revealed no difference between the two groups half of the offspring of STD diet–fed mice, but in none of offspring (Fig. 2B). The influence of the diet and the of the offspring of GF diet–fed mice. The annotation of gut microbiota of the mothers on the gut microbiota of reads within the Verrucomicrobia represented one spe- the offspring was thus not permanent. cies, A. muciniphila, whereas the genus Proteus was respon- The differences in the fecal microbial composition sible for the difference evident in Proteobacteria phyla between the two groups of NOD mice and their offspring (Table 2). at 4 weeks of age were further corroborated by tag- qPCR analysis of Akkermansia in the ileum mucosal encoded 16S rRNA gene–based pyrosequencing. The raw layer at weaning revealed a higher abundance in the off- number of reads generated from all 39 fecal samples spring of GF diet–fed mice compared with STD diet–fed scored 1,332,137. Sequences that met all requirements mice (Fig. 2E; P , 0.001). A similar tendency was also of the quality control (minimum length 300 bp, quality evident in the colonic mucus layer (Fig. 2F; P = 0.06). score $25) and were free from chimeric reads yielded An Early GF Environment Increases Anti-Inflammatory 848,346, providing an average of 21,752 sequences per Immune Cells and Intestinally Primed T Cells in PLN sample (minimum 1,178 sequences, maximum 88,674 Both intestinal and pancreatic DCs have been reported to sequences, SD = 17,103 sequences), with a mean sequence play important roles in T1D, and especially the CD11b+ length of 458 bp (minimum 300 bp, maximum 470 bp). subset has been associated with both pathogenic and tol- One sample was discarded because of the low number of erogenic immunity to pancreatic islets. In this study, the reads (,1,000 reads). PCoA based on weighted UniFrac CD11b+CD11c+ DCs were less abundant among pancreatic distance metrics showed a clear separation of the two (P , 0.01), intestinal (P , 0.001), and systemic (P , 0.05) categories comprising GF diet–fed NOD mice and their lymphocyte populations in the offspring of GF diet–fed offspring, and STD diet–fed NOD mice with their off- mice compared with STD diet–fed offspring at the time spring. The proportion of the cumulative information de- of weaning (Fig. 3A). No differences in the overall propor- scribing the variance using the first two PCs reached 60% tion of CD11c+ DCs were observed, and no differences were (Fig. 2C). observed between the groups at 10 weeks of age. The most abundant phyla in both categories were The proportions of Tregs (CD4+FoxP3+) in the MLN Firmicutes and Bacteroidetes that constituted 50% and and spleen were similar in the offspring of GF diet–fed 40%, respectively (Table 1). Metastats analysis revealed and STD diet–fed mice, both at weaning and at 10 weeks

relative distribution of bacterial phyla across samples. Bar charts represent the relative abundance of the seven major phyla across samples. Taxonomy assignation was based on the Greengenes reference database (version 12_10). E: qPCR analysis of A. muciniphila in ileum samples from 4-week-old NOD offspring of mice fed an STD diet (n = 8) or a GF diet (n = 8). F: qPCR analysis of A. muciniphila in ileum samples from 4-week-old NOD offspring of mice fed an STD diet (n = 9) or a GF diet (n = 9). The relative distribution of A. muciniphila within all bacteria is shown for each sample. All samples analyzed by qPCR were quantified in duplicate. 2826 ibtsPoetv itArs w Generations Two Across Diet Diabetes-Protective

Table 2—Significant differences in taxa among 4-week-old offspring of STD diet–fed and GF diet–fed NOD mice STD diet mean GF diet mean Phylum Class Order Family Genus abundance abundance P value* q value* Firmicutes Clostridia Unclassified Unclassified Unclassified 0.460 1.818 0.001 0.011 Firmicutes Clostridia Unknown Unknown Unknown 0.128 1.616 0.001 0.011 Firmicutes Clostridia Clostridiales Unclassified Unclassified 0.096 0.358 0.001 0.011 Firmicutes Clostridia Clostridiales Halobacteriaceae Dehalobacterium 0.125 0.614 0.001 0.011 Firmicutes Clostridia Clostridiales Lachnospiraceae Roseburia 0.000 0.104 0.001 0.011 Firmicutes Clostridia Clostridiales Lachnospiraceae [Ruminococcus] 0.153 0.594 0.001 0.011 Firmicutes Clostridia Clostridiales Ruminococcaceae Unclassified 0.225 0.950 0.001 0.011 Firmicutes Clostridia Clostridiales Ruminococcaceae Unknown 0.289 1.446 0.003 0.026 Firmicutes Clostridia Clostridiales Ruminococcaceae Anaerotruncus 0.048 0.181 0.001 0.011 Firmicutes Clostridia Clostridiales Ruminococcaceae Oscillospira 5.844 11.871 0.006 0.049 Proteobacteria Gammaproteobacteria Enterobacteriales Enterobacteriaceae Proteus 0.000 0.187 0.002 0.020 TM7 TM7-3 CW040 F16 Unknown 0.012 6.057 0.001 0.011 Verrucomicrobia Verrucomicrobiae Verrucomicrobiales Verrucomicrobiaceae Akkermansia 0.686 7.809 0.003 0.026 Bacteroidetes Bacteroidia Bacteroidales [Paraprevotellaceae] [Prevotella] 1.918 0.012 0.001 0.011 Cyanobacteria 4C0d-2 YS2 Unknown Unknown 1.255 0.000 0.001 0.011 Firmicutes Bacilli Lactobacillales Lactobacillaceae Unknown 0.437 0.025 0.001 0.011 Firmicutes Clostridia Clostridiales Unknown Unknown 1.517 0.167 0.002 0.020

Firmicutes Clostridia Clostridiales Clostridiaceae Unclassified 0.964 0.001 0.001 0.011 Diabetes Brackets indicate suggested but not verified names. Unclassified, inability to assign given operational taxonomic unit (OTU) into single taxonomic group; Unknown, absence of information about given OTU in the database. *Values were calculated with Metastats using 1,000 permutations. oue6,Ags 2014 August 63, Volume diabetes.diabetesjournals.org Hansen and Associates 2827

Figure 3—Flow cytometric analysis of cells isolated from PLN, MLN, and spleen. Percentages of CD11b+CD11c+ DCs (A) and FoxP3+CD4+ Tregs (B) isolated from 4-week-old offspring from GF diet–fed (n = 8) or STD diet–fed (n = 8) NOD mice. C and D: Representative flow cytometric dot plots illustrating the percentages of CD11b+CD11c+ DCs in MLN after gating on CD11c+ cells. E and F: Representative flow cytometric dot plots illustrating the percentages of FoxP3+CD4+ Tregs in PLN after gating on CD4+ T cells. Percentages of CD8+ T cells (G), CD4+ T cells (H), a4b7+CD8+ T cells (I), and a4b7+CD4+ T cells (J) isolated from 4-week-old offspring from GF diet–fed (n = 8) or STD diet– fed (n = 8) NOD mice are illustrated as indicated. K–N: Representative flow cytometric dot plots illustrating the percentages of a4b7+CD8+ and a4b7+CD4+ T cells in PLN after gating on CD8+ and CD4+ T cells, respectively, as indicated. Error bars represent the SEM. *P < 0.05, **P < 0.01, ***P < 0.001. of age, but, interestingly, a significantly higher proportion T cells in the MLN, PLN, and spleen were not different (P , 0.001) of these cells was evident in the PLN in the between the two groups either at weaning or at 10 weeks offspring of GF diet–fed mice compared with offspring of of age, except for a higher proportion (P , 0.001) of CD8 STD diet–fed mice at weaning, indicating a more anti-in- T cells in the PLN at weaning in the offspring of GF diet– flammatory local immune system due to an early GF en- fed mice compared with offspring of STD diet–fed mice vironment (Fig. 3B). The proportions of CD4 and CD8 (Fig. 3G and H). Furthermore, these CD8 T cells were 2828 Diabetes-Protective Diet Across Two Generations Diabetes Volume 63, August 2014 marked with gut-homing receptor a4b7 integrin, which is regulatory immunological activity, genes implicated in induced when T cells are activated in the intestinal envi- immune cell migration, microbial recognition, and re- ronment (Fig. 3I). Also on the CD4 T cells in PLN a higher sponse, as well as T-cell activation and signaling in the gut proportion of cells (P , 0.01) was marked with a4b7in were analyzed at weaning. Thirty-one of the 90 gene the offspring of GF diet–fed mice compared with off- expressions analyzed were significantly altered between spring of STD diet–fed mice (Fig. 3J). However, this dif- the two groups of offspring (Table 3). Of importance, the ference in a4b7 integrin was evident only at weaning and early GF environment caused significantly less expression not at 10 weeks of age. of several inflammatory mediators of insulitis such as Intestinal Gene Expression Is Skewed Toward an Ifng, Il12b, and Il18 genes, and Prf1 and Gzmb genes for Anti-Inflammatory Phenotype in Offspring of GF the cytolytic enzymes perforin-1 and granzyme B, respec- Diet–Fed Mice tively. Interestingly, the Cd68 macrophage marker, the To further explore how the early-life GF environ- Hmox1 gene for the heme oxygenase-1 enzyme, and the ment affected gut homeostasis and the establishment of Stat6 gene involved in exerting IL-4 were more

Table 3—Relative gene expression of immune-related genes analyzed in an inventoried TaqMan Mouse Immune Array in ileal samples collected from 4-week-old offspring of GF diet–fed and STD diet–fed NOD mice RQ mean* Gene Gene name [protein name] STD diet GF diet P value† Signaling C3 Complement component 3 [C3] 1.07 6 0.14 2.06 6 0.33 ,0.01 Col4a5 Procollagen, type IV, a5 [COL4a5] 1.06 6 0.16 2.63 6 0.66 ,0.05 Edn1 Endothelin 1 [EDN1] 1.28 6 0.39 6.65 6 1.12 ,0.001 Ikbkb Inhibitor of kB kinase b [IKKb] 1.03 6 0.09 1.51 6 0.17 ,0.05 Nfkb1 Nuclear factor of k light chain gene enhancer in 1.02 6 0.07 1.37 6 0.12 ,0.05 B-cells 1 [NFkB1] Smad3 MAD homolog 3 (Drosophila) [SMAD3] 1.03 6 0.09 1.67 6 0.11 ,0.001 Smad7 MAD homolog 7 (Drosophila) [SMAD7] 1.03 6 0.09 1.64 6 0.16 ,0.01 Ski Sloan-Kettering viral oncogene homolog [SKI] 1.01 6 0.05 1.79 6 0.12 ,0.0001 Socs1 Suppressor of cytokine signaling 1 [SOCS1] 1.03 6 0.10 0.51 6 0.11 ,0.01 Socs2 Suppressor of cytokine signaling 2 [SOCS2] 1.02 6 0.07 1.89 6 0.17 ,0.001 Stat6 Signal transducer and activator of 1.02 6 0.07 1.66 6 0.11 ,0.001 transcription 6 [STAT6] Cytokines/cytokine receptors Fas Fas (TNF receptor superfamily member) [FAS] 1.02 6 0.07 1.50 6 0.16 ,0.05 Ifng IFN-g 3.13 6 1.63 0.29 6 0.22 ,0.01 Il10 IL-10 1.25 6 0.35 0.00 6 0.00 ,0.0001 Il12b IL-12b 3.42 6 1.55 0.85 6 0.80 ,0.05 Il15 IL-15 1.06 6 0.15 3.05 6 0.45 ,0.0001 Il17 IL-17 1.25 6 0.33 0.33 6 0.32 ,0.001 Il18 IL-18 1.03 6 0.09 0.20 6 0.02 ,0.0001 Chemokine/chemokine receptors Ccr4 Chemokine (C-C motif) receptor 4 [CCR4] 1.72 6 0.37 0.15 6 0.04 ,0.01 Cxcl11 Chemokine (C-X-C motif) ligand 11 [I-TAC] 1.03 6 0.11 3.18 6 0.56 ,0.0001 Cell surface receptors Cd34 CD34 antigen [CD34] 1.12 6 0.22 2.53 6 0.46 ,0.01 Cd38 CD38 antigen [CD38] 1.01 6 0.06 0.52 6 0.06 ,0.01 Cd3e CD3 antigen, ´ polypeptide [CD3e] 1.16 6 0.19 0.57 6 0.12 ,0.05 Cd68 CD68 antigen [CD68] 1.04 6 0.11 1.56 6 0.20 ,0.05 Cd8a CD8 antigen, a chain [CD8a] 1.19 6 0.23 0.49 6 0.10 ,0.05 Lrp2 LDL receptor–related protein 2 [LRP2] 1.39 6 0.29 4.16 6 0.67 ,0.01 Tfrc Transferrin receptor [TRFC] 1.02 6 0.08 1.38 6 0.11 ,0.05 Enzymes Gzmb Granzyme B [CTLA1] 1.55 6 0.40 0.19 6 0.09 ,0.01 Hmox1 Heme oxygenase (decycling) 1 [HMOX1] 1.09 6 0.19 2.41 6 0.48 ,0.05 Nos2 Nitric oxide synthase 2, inducible, macrophage [NOS2] 1.05 6 0.12 0.15 6 0.04 ,0.0001 Prf1 Perforin 1 (pore-forming protein) [PRF1] 1.77 6 0.47 0.44 6 0.20 ,0.05 *Relative quantification was calculated by the comparative Ct method, where the expression of each gene is first normalized to the expression of Actb. Comparative gene expression is calculated for the mean control group, and fold change (RQ) values are obtained, with fold change = 1 for mean control. RQ values for genes more highly expressed compared with controls are marked in bold. †Only genes that were significantly expressed differently (P , 0.05) between the two groups are included. Statistical analysis was performed on dCt values [Ct(target) 2 Ct(reference)]. diabetes.diabetesjournals.org Hansen and Associates 2829 profoundly expressed in the offspring of GF diet–fed mice compared with STD diet–fed offspring, and they are all characteristics of immunosuppressive M2 macrophages, whereas the expression of the Nos2 (iNos) gene marker of M1 macrophages was low. However, gene expressions of intestinal cytokines Il10 and Il17, which have previ- ously been associated with diabetes protection, were lower in these mice similar to Socs1, an inhibitor of IFN-g signaling and Smad7 and Ski, which are inhibitors of the regulatory cytokine TGF-b signaling. Expressions of T-cell marker genes such as Cd3e, Cd8a, and Cd38 were also downregulated in the ileum, including Ccr4 expression, which has been demonstrated on patho- genic autoimmune T cells in NOD mice. Supportive of the flow cytometry results, the CD34 gene expression, which encodes a key molecule in T-cell trafficking to lymph nodes, was higher in the offspring of GF diet–fed mice compared with STD diet–fed offspring. Expressions of Intestinal Tight Junction and Mucus- Related Genes Are Elevated in the Gut at Weaning To investigate whether the more anti-inflammatory gene expression profile in the gut was associated with an improved intestinal barrier at weaning, gene expressions of tight junction and mucus components were analyzed in Figure 4—Relative gene expression of Ocln (A), Tjp1 (B), Cldn8 (C), the ileum and colon. The following tight junction Cldn15 (D), Muc1 (E), and Muc2 (F) in ileum and colon from 4-week- component genes have been shown to be regulated in old offspring of GF diet–fed or STD diet–fed mice. Data were nor- association with improved intestinal permeability assay malized to Actb and then to the mean control group, which was fi fi – results (29). Gene expressions of Ocln, which encodes de ned as 1. All samples were quanti ed in duplicate (n =8 9 per group). Error bars represent the SEM. *P < 0.05, **P < 0.01. occludin (P , 0.01; Fig. 4A), Tjp1, which encodes tight junction protein 1 (P , 0.01; Fig. 4B), and Cldn15, which encodes claudin-15 (P , 0.05; Fig. 4D), were elevated in ilea from the offspring of GF diet–fed mice compared with insulitis than the nondiabetic STD mice, which indicates those of STD diet–fed mice. Ocln (P , 0.05) gene expres- that the GF group is not just delayed in diabetes develop- sion was also elevated in the colon together with a tendency ment, but is as far from diabetic as the nondiabetic con- for higher expression of Muc1, which encodes a protein trol mice that usually do not develop diabetes; however, that represents membrane-associated mucin (P =0.06; a longer observational period would be necessary to fully Fig. 4E). Conversely, colonic Muc2 expression representing clarify whether the mice are completely protected against secreted mucin was lower (P , 0.05; Fig. 4F). diabetes. In agreement with this result, the ability of a low- DISCUSSION protein diet to modify diabetes incidence was reported to The importance of gluten for the development of auto- cause a significant drop in diabetes incidence from 86% in immune diabetes was previously demonstrated in both control NOD mice to 63% in NOD mice when given only NOD mice and BB rats, and it has become clear that during pregnancy and lactation (30). In contrast, wheat disease-modulating interventions, including dietary or and barley protein deprivation only until weaning was microbial antigen treatments, in these animal models demonstrated not to be sufficient to significantly delay are particularly imperative in early life. To clarify whether diabetes development. However, in the gestational and the effect of gluten could be prevented exclusively by preweaning period this diet was also supplemented with limiting its exposure in the postnatal period, a two- fish oil and Vit D3, which in the same study (31) were generation approach was used. Most importantly, the demonstrated to abrogate the protective effect of a wheat diabetes incidence was significantly reduced in the off- and barley protein-free diet. Interestingly, it was also spring of GF diet–fed NOD mice even though they were shown that accelerated diabetes was not completely weaned to an STD diet. Thus, a GF diet during gestation reconstituted by supplementing the wheat and barley pro- and lactation was protective later in life. The fact that we, tein-free diet with gliadin, which indicates that other di- despite the low incidence of T1D in our facility, see a sig- etary or microbial antigens also mediate the protective nificant difference in the GF group further substantiates effect. For example, the gut microbiota in GF diet–fed the strong effect of this diet. The 30-week-old nondiabetic mice, which may not be altered by a pure gliadin supple- offspring of GF diet–fed mice had a lower incidence of ment, is different from that in gluten-fed mice (12). A 2830 Diabetes-Protective Diet Across Two Generations Diabetes Volume 63, August 2014 recent article (17) proposed cereal dietary antigens as a protected against the development of T1D later in life. stronger T1D inducer than microbes. The authors Thus, its immune regulatory effect on, for example, Tregs reported a similar protective effect on diabetes incidence at weaning in the prediabetic stage seems to have a long- in BB rats that were fed a low-antigen hydrolyzed casein lasting impact on the capability of the immune system to (HC) diet in both germ-free and specific pathogen–free protect against autoimmune attack on the b-cells. (SPF) conditions. However, the HC diet was more pro- Interestingly, an early article (39) demonstrated that tective in the germ-free than in the SPF condition, also adoptively transferred T cells from NOD mice on an HC indicating a diabetes-promoting effect of the microbes. diet were unable to change the incidence of diabetes and This was further supported by a low b-cell mass only in were presented with similar T-cell receptor–mediated pro- the cereal-fed BB rats compared with the germ-free and liferative responses compared with controls. Considering HC-fed BB rats. this, it was hypothesized that immune regulatory mecha- In the current study, a distinct bacterial profile nisms in the pancreatic environment at weaning down- enriched in especially Akkermansia, TM7, and Proteobac- regulate otherwise fully functional diabetogenic T-cell teria was evident in both NOD mice fed a GF diet and in response in GF diet–fed NOD mice. In the current study, their offspring at weaning. These taxonomic groups were this hypothesis was supported by increased proportions also previously associated with protection against the de- of FoxP3+ Tregs in PLN and fewer CD11b+ DCs at wean- velopment of autoimmune diabetes in NOD mice only ing, which also previously were modulated by an antidia- when present before weaning (22), which is interesting betogenic diet (5). A similar anti-inflammatory immune as Akkermansia has been demonstrated to modulate host profile has been observed in BALB/c mice receiving a GF immune responses in monocolonized mice (32). In addi- diet in both pancreatic and gut-associated lymphoid tissue tion, taxonomic differences between the gut microbiomes (18). The changes in the immune system were only found of healthy and diabetic children were characterized by at weaning and not at 10 weeks of age. Thus, it seems that mucin-degrading Akkermansia, which was more abundant changes in immunity later in life when insulitis is more in control subjects than in case patients (33). As gluten progressive are not as critical as in postnatal life for the has potential irritating effects in the small intestine where development of autoimmune diabetes, during which the it is degraded, this might affect mucus production and insulitis process begins. This further indicates that the GF barrier function. By leaving out gluten from the diet, diet, possibly through a change in the gut microbiota, may the mucus production may increase, which in turn can have delayed the development of the adaptive immunity, lead to an increase in the presence and metabolic activity but that the mice, independently of whether they develop of specific bacterial strains, not least Akkermansia, which T1D or not, eventually develop a mature immune system. previously has been shown to grow on mucus proteins Lower intestinal gene expression of proinflammatory (34). Minor fold-change differences were seen in tight cytokines and higher expression of anti-inflammatory M2 junction component gene expressions, which indicated macrophage markers, together with increased gut Cd3e, that Akkermansia might be associated with an improved Cd8a expression were found in the offspring of GF diet– intestinal barrier in both the ileum and colon, although fed mice. It is interesting that these findings were also this link is purely speculative. The restoration of impaired seen in SPF BB rats fed an HC diet but not to the same intestinal barrier and alleviated signs of gut epithelial extent as in the germ-free HC-fed BB rats (17). Thus, it irritation such as colonic crypt hyperplasia have also pre- required the presence of microbes. M2 macrophages were viously been shown in response to antidiabetogenic diets furthermore found to be associated with the microbiota- in T1D rodent models (5,35). dependent sex difference observed in T1D development in Not much is known about the role of TM7 in mam- NOD mice (40). The intestinal alterations found in the malian health and disease, but in humans it is mainly offspring of GF diet–fed mice may therefore, also in this associated with the oral cavity, where it has been associated experiment be a consequence of the altered microbiota in with periodontitis (36,37). Furthermore, Kuehbacher et al. early life rather than the presence of cereal antigens. Al- (38) suggested that TM7 members are involved in the though these data cannot be causally linked to the pan- ethology of Crohn disease, although the mechanism creas, it was demonstrated that an early GF environment remains unknown. A detailed examination of the influ- increased the trafficking of T cells with a mucosal pheno- ence of the microbes from GF diet–fed NOD mice by type to PLN. It would be interesting to further investigate transferring the microbiota to germ-free mice would be the regulatory properties of these T cells because a high informative. It is striking that the effect on the microbial intestinal IL-15 level, as seen in the GF group, promotes composition is only present as long as the mice are fed intestinal epithelial cell activation of noncytotoxic CD8 a GF diet. This indicates that the effect of the microbiota T cells with suppressor function (41). These cells have is dependent on the diet and that this effect is especially nondetectable granzyme B, which was also expressed sig- important during the development of the immune sys- nificantly lower in the GF group. This is further consistent tem. Even more striking is the fact that even though with the prevalence of M2 anti-inflammatory macro- the microbiota reverses after the introduction of gluten phages as these and Tregs seem to have a mutual ability in the diet at weaning, the mice born by GF mothers are to promote the differentiation of one another, at least diabetes.diabetesjournals.org Hansen and Associates 2831 partly through the TGF-b pathway (42,43). It is possible 9. Yang XD, Sytwu HK, McDevitt HO, Michie SA. Involvement of beta 7 integrin that any effect of bacterial or dietary antigens on pancre- and mucosal addressin cell adhesion molecule-1 (MAdCAM-1) in the de- atic immune homeostasis would be mediated in part by velopment of diabetes in obese diabetic mice. Diabetes 1997;46:1542–1547 the migration of these immune cells that are activated in 10. Malaisse WJ, Courtois P, Scott FW. Insulin-dependent diabetes and gut dysfunction: the BB rat model. Horm Metab Res 2004;36:585–594 the tolerogenic gut environment. 11. Maurano F, Mazzarella G, Luongo D, et al. Small intestinal enteropathy in Whether the early ingestion of cereal antigens before non-obese diabetic mice fed a diet containing wheat. Diabetologia 2005;48: weaning or the altered microbiota exerts its effect sepa- 931–937 rately or synergistically is not known. Most importantly, 12. Hansen AK, Ling F, Kaas A, Funda DP, Farlov H, Buschard K. Diabetes the early GF environment clearly attenuated diabetes preventive gluten-free diet decreases the number of caecal bacteria in non-obese development in the NOD mice even though they were diabetic mice. Diabetes Metab Res Rev 2006;22:220–225 weaned to a gluten-containing diet, but the changes in gut 13. Wen L, Ley RE, Volchkov PY, et al. Innate immunity and intestinal microbiota microbiota and immune system were no longer evident in the development of Type 1 diabetes. Nature 2008;455:1109–1113 later in life, from which we can conclude that the time and 14. Brugman S, Klatter FA, Visser JTJ, et al. Antibiotic treatment partially protects the diet before weaning are of imperative importance for against type 1 diabetes in the Bio-Breeding diabetes-prone rat. Is the gut flora – protection against diabetes development. involved in the development of type 1 diabetes? Diabetologia 2006;49:2105 2108 15. Buschard K, Pedersen C, Hansen SV, Hageman I, Aaen K, Bendtzen K. Anti- diabetogenic effect of fusidic acid in diabetes prone BB rats. Autoimmunity 1992; 14:101–104 Funding. This work was carried out as part of the 3G Center—Gut, Grain & 16. Calcinaro F, Dionisi S, Marinaro M, et al. Oral probiotic administration in- Greens; the 3G Center is supported by the Danish Council for Strategic Research. duces interleukin-10 production and prevents spontaneous autoimmune diabetes This work was further funded by CHANCE (Chemometric Analysis Centre at the in the non-obese diabetic mouse. Diabetologia 2005;48:1565–1575 University of Copenhagen), the Center for Applied Laboratory Animal Research, 17. Patrick C, Wang GS, Lefebvre DE, et al. Promotion of autoimmune diabetes and the Beckett Foundation. by cereal diet in the presence or absence of microbes associated with gut im- Duality of Interest. No potential conflicts of interest relevant to this article mune activation, regulatory imbalance, and altered cathelicidin antimicrobial were reported. peptide. Diabetes 2013;62:2036–2047 Author Contributions. C.H.F.H. conceived and designed the study, an- 18. Antvorskov JC, Fundova P, Buschard K, Funda DP. Dietary gluten alters the alyzed histology, performed flow cytometry, contributed to the discussion, and balance of pro-inflammatory and anti-inflammatory cytokines in T cells of BALB/c wrote the manuscript. Ł.K. analyzed the data, contributed to the discussion, and mice. Immunology 2013;138:23–33 wrote the manuscript. K.B. analyzed histology and contributed to the discussion. 19. Chakir H, Lefebvre DE, Wang H, Caraher E, Scott FW. Wheat protein-induced S.B.M., C.N., and H.F. were involved in the analysis of gene expression data and proinflammatory T helper 1 bias in mesenteric lymph nodes of young diabetes- contributed to the discussion. L.H.H. performed pyrosequencing and contributed prone rats. Diabetologia 2005;48:1576–1584 to the discussion. D.S.N. analyzed the data and contributed to the discussion. 20. Hoorfar J, Buschard K, Dagnaes-Hansen F. Prophylactic nutritional modi- S.S. performed flow cytometry and contributed to the discussion. A.K.H. con- fication of the incidence of diabetes in autoimmune non-obese diabetic (NOD) ceived and designed the study, contributed to the discussion, and wrote the mice. Br J Nutr 1993;69:597–607 manuscript. C.H.F.H. is the guarantor of this work and, as such, had full access 21. Hufeldt MR, Nielsen DS, Vogensen FK, Midtvedt T, Hansen AK. Variation in to all the data in the study and takes responsibility for the integrity of the data and the gut microbiota of laboratory mice is related to both genetic and environmental the accuracy of the data analysis. factors. Comp Med 2010;60:336–347 References 22. Hansen CH, Krych L, Nielsen DS, et al. Early life treatment with vancomycin propagates Akkermansia muciniphila and reduces diabetes incidence in the NOD 1. Funda DP, Kaas A, Bock T, Tlaskalová-Hogenová H, Buschard K. Gluten-free mouse. Diabetologia 2012;55:2285–2294 diet prevents diabetes in NOD mice. Diabetes Metab Res Rev 1999;15:323–327 23. Krych L, Hansen CH, Hansen AK, van den Berg FW, Nielsen DS. Quantitatively 2. Funda DP, Kaas A, Tlaskalová-Hogenová H, Buschard K. Gluten-free but different, yet qualitatively alike: a meta-analysis of the mouse core gut microbiome also gluten-enriched (gluten+) diet prevent diabetes in NOD mice; the gluten with a view towards the human gut microbiome. PLoS One 2013;8:e62578 – enigma in type 1 diabetes. Diabetes Metab Res Rev 2008;24:59 63 24. White JR, Nagarajan N, Pop M. Statistical methods for detecting differen- 3. Westerholm-Ormio M, Vaarala O, Pihkala P, Ilonen J, Savilahti E. Immu- tially abundant features in clinical metagenomic samples. PLoS Comput Biol nologic activity in the small intestinal mucosa of pediatric patients with type 1 2009;5:e1000352 – diabetes. Diabetes 2003;52:2287 2295 25. Hansen CH, Frøkiær H, Christensen AG, et al. Dietary xylooligosaccharide 4. Flohé SB, Wasmuth HE, Kerad JB, et al. A wheat-based, diabetes-promoting diet downregulates IFN-g and the low-grade inflammatory cytokine IL-1b systemi- – induces a Th1-type cytokine bias in the gut of NOD mice. Cytokine 2003;21:149 154 cally in mice. J Nutr 2013;143:533–540 5. Alam C, Valkonen S, Palagani V, Jalava J, Eerola E, Hänninen A. In- 26. Fink LN, Metzdorff SB, Zeuthen LH, et al. Establishment of tolerance to flammatory tendencies and overproduction of IL-17 in the colon of young NOD commensal bacteria requires a complex microbiota and is accompanied by mice are counteracted with diet change. Diabetes 2010;59:2237–2246 decreased intestinal chemokine expression. Am J Physiol Gastrointest Liver 6. Alam C, Bittoun E, Bhagwat D, et al. Effects of a germ-free environment on Physiol 2012;302:G55–G65 gut immune regulation and diabetes progression in non-obese diabetic (NOD) 27. Hansen CH, Holm TL, Krych L, et al. Gut microbiota regulates NKG2D ligand mice. Diabetologia 2011;54:1398–1406 expression on intestinal epithelial cells. Eur J Immunol 2013;43:447–457 7. Badami E, Sorini C, Coccia M, et al. Defective differentiation of regulatory 28. Myles IA, Fontecilla NM, Janelsins BM, Vithayathil PJ, Segre JA, Datta SK. FoxP3+ T cells by small-intestinal dendritic cells in patients with type 1 diabetes. Parental dietary fat intake alters offspring microbiome and immunity. J Immunol Diabetes 2011;60:2120–2124 2013;191:3200–3209 8. Hänninen A, Salmi M, Simell O, Jalkanen S. Mucosa-associated (beta 29. Hsiao EY, McBride SW, Hsien S, et al. Microbiota modulate behavioral and 7-integrinhigh) lymphocytes accumulate early in the pancreas of NOD mice and physiological abnormalities associated with neurodevelopmental disorders. Cell show aberrant recirculation behavior. Diabetes 1996;45:1173–1180 2013;155:1451–1463 2832 Diabetes-Protective Diet Across Two Generations Diabetes Volume 63, August 2014

30. Chamson-Reig A, Arany EJ, Summers K, Hill DJ. A low protein diet in early 37. Segata N, Haake SK, Mannon P, et al. Composition of the adult digestive life delays the onset of diabetes in the non-obese diabetic mouse. J Endocrinol tract bacterial microbiome based on seven mouth surfaces, tonsils, throat and 2009;201:231–239 stool samples. Genome Biol 2012;13:R42 31. Schmid S, Koczwara K, Schwinghammer S, Lampasona V, Ziegler AG, 38. Kuehbacher T, Rehman A, Lepage P, et al. Intestinal TM7 bacterial phy- Bonifacio E. Delayed exposure to wheat and barley proteins reduces diabetes logenies in active inflammatory bowel disease. J Med Microbiol 2008;57:1569– – incidence in non-obese diabetic mice. Clin Immunol 2004;111:108 118 1576 32. Derrien M, Van Baarlen P, Hooiveld G, Norin E, Müller M, de Vos WM. Mod- 39. Hermitte L, Atlan-Gepner C, Payan MJ, Mehelleb M, Vialettes B. Dietary ulation of mucosal immune response, tolerance, and proliferation in mice colonized protection against diabetes in NOD mice: lack of a major change in the immune by the mucin-degrader Akkermansia muciniphila. Front Microbiol 2011;2:166 system. Diabete Metab 1995;21:261–268 33. Brown CT, Davis-Richardson AG, Giongo A, et al. Gut microbiome meta- 40. Yurkovetskiy L, Burrows M, Khan AA, et al. Gender bias in autoimmunity is genomics analysis suggests a functional model for the development of autoim- influenced by microbiota. Immunity 2013;39:400–412 munity for type 1 diabetes. PLoS One 2011;6:e25792 41. Roda G, Jianyu X, Park MS, et al. Characterizing CEACAM5 interaction with 34. Derrien M, Collado MC, Ben-Amor K, Salminen S, de Vos WM. The Mucin CD8a and CD1d in intestinal homeostasis. Mucosal Immunol 2014;7:615–624 degrader Akkermansia muciniphila is an abundant resident of the human in- 42. Denning TL, Wang YC, Patel SR, Williams IR, Pulendran B. Lamina propria testinal tract. Appl Environ Microbiol 2008;74:1646–1648 35. Visser JT, Lammers K, Hoogendijk A, et al. Restoration of impaired intestinal macrophages and dendritic cells differentially induce regulatory and interleukin barrier function by the hydrolysed casein diet contributes to the prevention of type 1 17-producing T cell responses. Nat Immunol 2007;8:1086–1094 diabetes in the diabetes-prone BioBreeding rat. Diabetologia 2010;53:2621–2628 43. Tiemessen MM, Jagger AL, Evans HG, van Herwijnen MJ, John S, Taams 36. Brinig MM, Lepp PW, Ouverney CC, Armitage GC, Relman DA. Prevalence of LS. CD4+CD25+Foxp3+ regulatory T cells induce alternative activation of bacteria of division TM7 in human subgingival plaque and their association with human monocytes/macrophages. Proc Natl Acad Sci U S A 2007;104:19446– disease. Appl Environ Microbiol 2003;69:1687–1694 19451