Specific Microbiota-Induced Intestinal Th17 Differentiation Requires MHC II but not GALT and Mesenteric Lymph Nodes

Abstract

Interleukin (IL)-17 expressing CD4+ T lymphocytes (Th17 cells) naturally reside in the intestine where specific cytokines and microbiota, such as segmented filamentous bacteria (SFB), promote their differentiation. Intestinal Th17 cells are believed to initially differentiate in the GALT and/or mesenteric lymph nodes (mLN) upon antigen encounter and subsequently home to the lamina propria (LP) where they mediate effector functions. However, whether GALT and/or mLN are required for intestinal Th17 differentiation, and how microbiota containing SFB regulate antigen-specific intestinal Th17 cells remain poorly defined. Here we observed that naïve CD4+ T cells were abundant in the intestinal LP prior to weaning and that the accumulation of Th17 cells in response to microbiota containing SFB occurred in the absence of lymphotoxin (LT)-dependent lymphoid structures and the spleen. Furthermore, the differentiation of intestinal Th17 cells in the presence of microbiota containing SFB was dependent on MHC II expression by CD11c+ cells. Lastly, the differentiation of antigen-specific Th17 cells required both the presence of cognate antigen and microbiota containing SFB. These findings suggest that microbiota containing SFB create an intestinal milieu that may induce antigen-specific Th17 differentiation against food and/or bacterial antigens directly in the intestinal LP.

Introduction

CD4+ T lymphocytes constitute a principal component of the adaptive immune system that functions together with innate immune cells to afford host protection against infection and tissue damage (1). The ability of CD4+ T cells to effectively respond to an extensive array of bacteria, viruses, helminthes, and other microbes is the result of a broad T cell receptor (TCR) repertoire and the capacity to differentiate into specific effector subsets. Among the best studied of these effector subsets are Th1 cells, which secrete IFN-γ and provide protection against intracellular pathogens, and Th2 cells, which produce IL-4 in response to extracellular bacteria and parasites (2). More recently, Th17 cells were identified as a population of CD4+ T cells distinct from classical Th1 or Th2 cells (3). Th17 cells were shown to be distinct from classical Th1 or Th2 cells and are defined by expression of their hallmark cytokine IL-17A (referred to as IL-17) and can also produce IL-17F, IL-21 and IL-22. Regulated IL-17 production by CD4+ T cells aids in the clearance of extracellular pathogens and fungi in part due its role in recruiting and activating neutrophils. Uncontrolled Th17 responses, however, can lead to pathological tissue damage and have been implicated in numerous infectious, autoimmune, and inflammatory diseases in mice and humans (4).

The in vitro differentiation and expansion of naïve CD4+ T cells along the Th17 lineage is dependent on TCR signaling in the presence of key cytokines including TGF-β1, IL-6, and IL-1β along with the downstream transcription factors of STAT3, IRF4, BATF and RORγt (5). IL-23 does not appear to be required for initial Th17 differentiation, but IL-23 receptor is expressed by developing Th17 cells and IL-23 can stimulate further differentiation, expansion and survival of Th17 cells (6). In vivo, many of the effects of Th17 cells are linked to IL-23 and specific blockade of the p19 subunit of IL-23 ameliorates experimental autoimmune encephelomyelitis, collagen induced arthritis, and colitis (7). Thus, IL-23 and Th17 cells are considered attractive targets for treatment of several autoimmune and inflammatory diseases.

While Th17 cells are induced during infectious and pathological states, they are constitutively present at mucosal surfaces, especially in the intestinal LP (8). The development of intestinal Th17 cells is dependent on the gut microbiota as mice treated with antibiotics from birth and germ-free mice are deficient in these cells (9). Interestingly, SFB are spore-forming, Gram-positive commensal bacteria that adhere tightly to intestinal epithelial cells (IECs) and robustly induce intestinal Th17 cells in mice (10, 11). Since ex vivo culture conditions for SFB have not been defined, many investigations of microbiota-induced Th17 responses have relied upon differences in the SFB status of mice from different vendors. In particular, B6 mice from Jackson Laboratory are void of SFB and consequently harbor a paucity of intestinal Th17 cells, while those from Taconic Laboratory are colonized by SFB and have an appreciable population of intestinal Th17 cells (12). Additionally, horizontal transmission of microbiota containing SFB from Taconic-derived mice to Jackson-derived mice is sufficient to induce Th17 cells in the latter. Although SFB can induce intestinal Th17 cells in vivo, the role for specific signaling pathways regulating this intestinal Th17 development are still being defined and appears to be independent of MyD88 and Trif but may be amplified by ATP (9) and/or serum amyloid A (12).

Currently, numerous aspects of intestinal Th17 differentiation remain poorly defined. Following SFB adherence to epithelium, it is believed that CD11b+ LP dendritic cells (DCs) induce Th17 cells (13–17), however the requirements for GALT and mLN along with MHC II-mediated antigen presentation requires further investigation. Utilizing mice deficient in LT-dependent lymphoid structures, we demonstrate that colonization by microbiota containing SFB induced intestinal Th17 cell differentiation independent of mLN and GALT, however, MHC II expression by CD11c+ cells and cognate antigen were required—indicating that intestinal Th17 cell differentiation may occur in situ in the intestinal LP. These findings suggest that microbiota containing SFB create an intestinal milieu that may induce antigen-specific Th17 differentiation against food and/or bacterial antigens directly in the intestinal LP.

Materials and Methods

Mice

Age- and sex-matched C57BL/6 (B6), B6.129S2-Lta^tm1Dch/J (Lta^−/−), SPLx Lta^−/− and B6.129S2-H2^dlAb1-Ea/J (MHC II^Δ/Δ), B6.Cg-Tg (Itgax-cre)1-1Reiz/J (CD11c-cre), B6. SJL-Ptprc^aPepc^b/BoyJ (CD45.1) and B6.PL-Thy1^a/CyJ (Thy1.1), and B6.Cg-Tg(TcraTcrb)425Cbn/J (OTII) mice were purchased from The Jackson Laboratory (JAX). B6 (Lta^+/+) and Lta^−/− mice were purchased from Jackson Labs in order to specifically ensure that mice were SFB-free. Immediately upon arrival at Emory University Lta^+/+ and Lta^−/− mice were cohoused. All SFB-containing microbiota transfer studies using Lta^+/+ and Lta^−/− mice were initiated 2 weeks after arrival from JAX in order to avoid unintentional colonization by SFB in our animal facility and to allow for equilibration of any differences in microbiota between Lta^+/+ and Lta^−/− mice during cohousing. Immediately prior to the introduction of SFB-containing microbiota, all Lta^+/+ and Lta^−/− mice were verified to be void of SFB (as determined by qPCR detection for SFB DNA in cecal contents and fecal pellets). These cohousing measures were also taken for SPLx Lta^−/−, MHC II^Δ/Δ, and relevant JAX B6 control mice. Co-housed, age-matched Ltbr^+/+ and Ltbr^−/− littermates were provided by R.D. Newberry and analyses of these mice were performed on-site at Washington University-St. Louis. MHC II^FF mice (provided by P.A. Koni) and CD11c-cre mice were crossed to generate MHC II^ΔDC mice. MHC II^FF litter- and cage-mates were used as controls. B6 mice purchased from Taconic were utilized as donors of SFB-containing intestinal microbiota for cecal content transfer experiments. Mice were maintained under specific pathogen-free conditions and animal protocols were reviewed and approved by the Institute Animal Care and Use Committee of Emory University and Georgia State University.

Antibodies and reagents

The following antibodies were purchased from eBioscience: IFNγ (XMG1.2), CD90.1 (H1S51), CD69 (H1.2F3), CD45RB (C363.16A), CD45.1 (A20), Vα2 (B20.1), IL-17A (eBio17B7), CD8α (eBioT4/11.8), CD25 (PC61.5), CD3ε (eBio500A2), and RORγ(t)-PE (B2D). Antibodies purchased from BD Biosciences were: TCRβ (H57-597), Vβ5 (MR9-4), CCR6 (140706), IL-17A (TC11-18H10), Vα2 (B20.1), and CD4 (RM4-5). Dead cells were identified using the fixable Aqua dead cell staining kit (Invitrogen). The following biotin-conjugated antibodies (eBioscience) were used for negative selection in conjunction with anti-biotin and anti-APC microbeads (Miltenyi Biotec): CD8α (53-6.7), Ly-6G (RB6-8C5), F4/80 (BM8), TER-119 (TER-119), CD11b (M1/70), NK1.1 (PK136), CD11c (N418), CD19 (eBio1D3). Isolation of LP cells and flow cytometry was performed as previously described (18).

Preparation and gavage of cecal contents containing SFB

The cecal contents from Taconic B6 mice were resuspended in 5 ml of sterile PBS, passed through a 100 µm cell strainer, and 150 ul of the suspension was gavaged into recipient mice twice with 3 hr between each gavage. All cecal content suspensions were verified to contain SFB via qPCR (12). Recipient mice were utilized two weeks after arrival from the JAX and verified to be void of SFB prior to gavage and positive for SFB post-gavage of SFB-containing cecal contents as assessed by qPCR analysis of fresh fecal pellets.

In vivo Th17 differentiation

Naïve CD4+ T cells were enriched via negative selection utilizing magnetic-activated cell sorting to deplete cells expressing: CD25, CD19, CD11b, CD11c, NK1.1, F4/80, Ly-6G, CD8α, and Ter119 on MACS LS columns with anti-biotin and anti-APC microbeads (Miltenyi Biotec). Frequency of CD4+IL-17A+ T cells (<1%) was verified using flow cytometry on the LSR II (BD). 5 × 10⁶ cells were injected i.v. into CD45.2 congenic hosts. Recipients were gavaged with cecal contents from Taconic mice 24 hours later. After 10 days, recipients were harvested for assessment of intestinal Th17 differentiation.

For co-transfer experiments, naive Thy1.1+ OT-II cells were mixed at a 1:1 ratio with CD45.1+ cells and a total of 10⁷ cells were injected i.v. After 24 hours, mice were gavaged with SFB-containing cecal contents from Taconic mice and/or fed albumin from chicken egg white (Sigma-Aldrich) in the drinking water (15 mg/ml) for 10 days.

Statistics

Statistical analyses were performed with Prism software (Graph-Pad Software) using the Student’s t test. Error bars represent SEM as indicated and p values equal to or less than 0.05 were considered statistically significant while p values greater than 0.05 were considered not statistically significant (N.S.).

Results

Naïve CD4+ T cells are present in the intestinal LP independent of LT-dependent lymphoid structures

The mLN are specialized secondary lymphoid organs (SLO) that drain the intestine, and as such are a site where LP DCs can migrate to present antigens to naive CD4+ T cells (19). These observations suggest that mLN may be the primary site for naïve CD4+ T cell priming and differentiation into Th17 cells that home to the intestinal LP. Hence, an enrichment of Th17 cells in the mLN would be expected relative to other SLO that do not drain the intestine. To investigate if the mLN are indeed enriched for Th17 cells, a comparative analysis was conducted to assess the proportion of these cells induced by SFB-containing microbiota in the spleen (Spl), peripheral lymph nodes (pLN), mLN, small intestine (SI) LP, and large intestine (LI) LP of JAX B6 mice gavaged SFB-containing cecal contents (SFB⁺ CC). The frequency of Th17 cells in the Spl and pLN was similar to that of the mLN, with all being less than 0.5% of total CD4+ T cells (Supplementary Fig. 1). Additionally, the frequencies of Th17 cells were significantly higher in the SI LP and LI LP (comprised 14% and 11% of total CD4+ T cells respectively), relative to those observed in the Spl, pLN, and mLN (Supplementary Fig. 1). These findings demonstrate that Th17 cells are not significantly enriched in the mLN compared to other SLO that do not drain the intestine and that the enrichment of Th17 cells in the intestinal LP may be due to CD4+ T cell priming and differentiate in situ.

In order to begin investigating whether CD4+ T cells may be primed and differentiate into Th17 cells within the intestine, we first determined whether naïve CD4+ T cells are present in the intestinal LP during development. Further, we used mice void of SLO—including the lymph nodes, Peyer’s patches, and isolated lymphoid follicles—as a result LT signaling deficiency (Lta^−/−) to investigate the differentiation of Th17 cells in the absence of mLN and other SLO. Ontogeny studies were conducted to characterize the proportion of naïve CD4+ T cells in the SI LP and LI LP of B6 (Lta^+/+) and Lta^−/− mice, and the Spl was used to provide a comparison between the intestine and a peripheral lymphoid organ. In the Spl of Lta^+/+ and Lta^−/− mice, ~70% of the CD4+ T cells were characterized as naïve (CD45RB^hi Foxp3−) independent of age (Fig. 1A, 1B). These CD45RB^hi Foxp3− CD4+ T cells were further verified as naïve due to their lack of expression for the activation and memory markers CD25, CD44, and CD69 (data not shown). Interestingly, ~80% of the SI LP and LI LP CD4+ T cells in both Lta^+/+ and Lta^−/− mice were CD45RB^hi Foxp3− at 1 week of age and this frequency decreased to ~60% in the SI LP and ~40% in the LI LP at 3 weeks and remained at ~20% in both the SI LP and LI LP into adulthood (Fig. 1C). Additionally, the absolute cell numbers for cell subsets were not altered and were similar at the various points (data not shown). Taken together, our results demonstrate that appreciable numbers of naïve CD4+ T cells are present in the intestine and do not require mLN or other LT-dependent lymphoid structures for their accumulation at this site.

FIGURE 1.

Naïve CD4+ T cells are present in the intestinal LP independent of LT-dependent lymphoid structures. Ontogeny of naïve CD45RB^hi Foxp3− CD4+ T cells in the spleen (Spl) and intestinal lamina propria (LP) was investigated utilizing flow cytometry. Representative FACS plots of cells pre-gated on TCRβ and CD4 and assessed for the expression of CD45RB and Foxp3 in Lta^+/+ (A) and Lta^−/− mice (B). CD45RB^hi Foxp3− CD4+ T cells were further verified to be negative for the memory markers of CD44 and CD25 (data not shown). (C) Frequencies of CD45RB^hi Foxp3− CD4+ T cells in the Spl, small intestine (SI) LP, and large intestine (LI) LP of Lta^+/+ and Lta^−/− mice during development into adulthood. Samples for week 1 and 2 were pooled for each age group due to small size of the organs. Data are representative of at least two independent experiments with three to eight mice per age group.

Intestinal Th17 differentiation takes place in the absence of the GALT, mLN, and other LT-dependent lymphoid structures

To examine the requirements for GALT, mLN and other LT-dependent lymphoid structures in intestinal Th17 differentiation, Lta^+/+ and Lta^−/− from JAX, which had undetectable levels of SFB DNA (data not shown), were gavaged vehicle (PBS) alone or SFB⁺ CC isolated from Taconic B6 mice, and intestinal Th17 differentiation of recipient mice was assessed 10 days post-gavage. Interestingly, colonization of Lta^+/+ and Lta^−/− mice with SFB⁺ CC induced robust differentiation of intestinal Th17 cells, increasing their frequency to ~15% in the SI LP and ~10% in the LI LP (Fig. 2A, 2B). Similar trends were observed for absolute cell numbers (Fig. 2C). The levels of SFB in Lta^+/+ and Lta^−/− mice at day 10 post-gavage were comparable to TAC B6 based on quantitation of SFB, and the induction of intestinal Th17 cells by SFB-containing microbiota was negligible in the IEL compartment (data not shown). Furthermore, these intestinal Th17 cells were confirmed to be bona fide Th17 cells as they expressed the nuclear orphan receptor RORγt, which is both necessary and sufficient for the Th17 program, as well as the chemokine receptor, CCR6 (Fig. 2D). Additionally, these intestinal Th17 cells were negative for IL-10 based on flow cytometry (data not shown; positive control included) and therefore do not appear to be regulatory Th17 cells (20). To address the possibility that the Spl may be a site for microbiota-driven intestinal Th17 differentiation in the absence of the GALT, mLN and other LT-dependent lymphoid structures, similar experiments were performed in splenectomized (Splx) Lta^−/− mice as well as, lymphotoxin β receptor deficient (Ltbr^−/−) mice, an additional model of SLO deficiency. Both Splx Lta^−/− and Ltbr^−/− mice yielded similar results as observed in Ltα^−/− mice (Fig. 3), and these findings confirm that intestinal Th17 differentiation induced by SFB-containing microbiota does not require the Spl, mLN or other LT-dependent lymphoid structures and may occur directly in the intestinal LP.

FIGURE 2.

Intestinal Th17 differentiation driven by SFB-containing microbiota takes place in the absence of the GALT, mLN, and other LT-dependent lymphoid structures. JAX Lta^+/+ and JAX Lta^−/− mice void of SFB were gavaged PBS or SFB-containing cecal contents (SFB⁺ CC) and intestinal Th17 differentiation was assessed 10 days later. (A) Representative FACS plots of intestinal Th17 cell frequencies in Lta^+/+ and Lta^−/− mice on day 10 post-gavage. (B) Comparison of intestinal Th17 cell frequencies and numbers (C) between Lta^+/+ and Lta^−/− mice. Data are representative of at least two independent experiments with three to four mice per group for A and six to nine mice per group for B. (D) Expression of RORγt and CCR6 by intestinal Th17 cells induced by SFB-containing microbiota. Bolded histograms are pre-gated on IL-17A+ CD4+ T cells while histograms not bolded are pre-gated on IL-17A− CD4+ T cells. Data in D are representative of two independent experiments with four mice per group. Error bars represent SEM. *, p ≤ 0.05; not statistically significant (N.S.), p > 0.05 using a Student’s t test.

FIGURE 3.

Intestinal Th17 differentiation in additional models of SLO deficiency. (A) Representative FACS plots of intestinal Th17 cells driven by SFB-containing microbiota for JAX Lta^+/+ and JAX Splx Lta^−/− mice. Comparison of intestinal Th17 cell frequencies (B) and numbers (C) for Lta^+/+ and Splx Lta^−/− mice. (D) Representative FACS plots of intestinal Th17 cells in Ltbr^+/+ and Ltbr^−/− mice. Comparison of intestinal Th17 cell frequency (E) and number (F) in Ltbr^+/+ and Ltbr^−/− mice. Data are representative of two independent experiments with four mice per group. Error bars represent SEM. *, p ≤ 0.05; not statistically significant (N.S.), p > 0.05 using a Student’s t test.

MHC II is required for intestinal Th17 differentiation induced by SFB-containing microbiota

The differentiation of intestinal Th17 cells is promoted by SFB-containing microbiota along with specific cytokines/factors that are secreted by CD11b+CD103+ LP DCs (12). Whether LP DCs or other antigen presenting cells are promoting intestinal Th17 differentiation via presentation of antigen(s) on MHC II, or via other biological functions, remains unclear. Hence, we employed a naïve CD4+ T cell and specific microbiota transfer system to evaluate intestinal Th17 differentiation in B6 (MHC II^+/+) and MHC II-deficient (MHC II^Δ/Δ) JAX mice. To do so, naïve polyclonal CD4+ T cells were enriched from the Spl and peripheral lymph nodes of CD45.1 mice (purity >99% IL-17A− CD4+ T cells) and adoptively transferred into CD45.2+ MHC II^+/+ and MHC II^Δ/Δ mice that were void of SFB. One day later, recipients were gavaged SFB⁺ CC, and Th17 differentiation was assessed amongst both the host and donor CD4+ T cells 10 days post-gavage. This experimental system enabled us to study the role of MHC II in modulating the ability of adoptively transferred naïve CD4+ T cells to differentiate into Th17 cells upon conditioning with SFB-containing microbiota in a defined timeframe. As expected, a paucity of CD4+ T cells were observed in MHC II^Δ/Δ mice since MHC II is essential for the proper development and survival of CD4+ T cells (Supplementary Fig. 2A–C; (21)). Interestingly, intestinal Th17 differentiation induced by SFB-containing microbiota was significantly attenuated in MHC II^Δ/Δ mice based on frequency and cell number for host and donor (Fig. 4A, 4B) lamina propria lymphocytes (LPL), relative to MHC II^+/+ mice. Approximately 30-fold reduction in the proportion of Th17 cells was observed for host (Fig. 4B, top left panel) and donor (Fig. 4B, top right panel) intestinal Th17 cells in MHC II^Δ/Δ mice relative to MHC II^+/+ mice, while for absolute cell numbers, >70- and >16-fold reductions were observed in host (Fig. 4B, bottom left panel) and donor LPL (Fig. 4B, bottom right panel), respectively. The abrogation of intestinal Th17 induction in MHC II^Δ/Δ mice by SFB-containing microbiota was not due to impaired survival of the donor CD4+ T cells since the number of donor CD4+ LPL were similar on day 10 post-gavage (Fig. 4C). The deficiency in MHC II also impacted the differentiation of intestinal Foxp3+ Treg (Fig. 4D, left panel) and Th1 cells (Fig. 4D, right panel). Collectively, these results establish that intestinal Th17 differentiation induced by SFB-containing microbiota is dependent upon MHC II.

FIGURE 4.

MHC II is required for intestinal Th17 differentiation induced by SFB-containing microbiota. (A–D) Naïve polyclonal CD4+CD25− T cells enriched from CD45.1 mice were adoptively transferred into JAX CD45.2 MHC II^+/+ and JAX CD45.2 MHC II^Δ/Δ mice on day −1. Recipients were gavaged SFB⁺ CC on the next day and intestinal Th17 differentiation was assessed amongst lamina propria lymphocytes (LPL) in the small intestine on day 10. (A) Representative FACS plots of microbiota-induced Th17 cells of host and donor SI LPL in MHC II^+/+ and MHC II^Δ/Δ mice. (B) Comparison of microbiota-induced Th17 differentiation for host LPL and donor LPL frequency (upper panels) and cell number (lower panels). (C) Number of donor CD45.1 CD4+ T cells amongst LPL in MHC II^+/+ and MHC II^Δ/Δ mice. (D) Frequency of Foxp3+ Treg (left panel) and Th1 (right panel) cells amongst donor CD4+ LPL in MHC II^+/+ and MHC II^Δ/Δ mice. Data are representative of four mice per group from two independent experiments. Error bars represent SEM. *, p ≤ 0.05; not statistically significant (N.S.), p > 0.05 using a Student’s t test.

To investigate the cell lineage for which MHC II expression is required, we examined mice specifically lacking MHC II on CD11c-expressing cells (MHC II^ΔDC). Both MHC II^ΔDC and MHC II^FF mice were littermate controls and co-housed in the same cage. MHC II was verified to be absent on CD11c+CD103+ LP DCs isolated from MHC II^ΔDC mice (Supplementary Fig. 3A) and the loss of MHC II on CD11c+CD103+ LP DCs in MHC II^ΔDC mice did not affect their abundance as the frequency and cell number were similar (Supplementary Fig. 3B; data not shown). Intestinal Th17 development examined 10 days post-gavage of SFB⁺ CC demonstrated significantly less Th17 cells in intestinal LP of MHC II^ΔDC mice in comparison to MHC II^FF mice (5% versus 15% of CD4+ T cells, respectively), and this was specific for Th17 cells since intestinal Foxp3+ Treg and Th1 cells were not significantly affected (Fig. 5A, 5B). In addition, deletion of MHC II on CD11c+ cells did not dramatically affect pro-inflammatory cytokine expression (Supplementary Fig. 4) nor impair CD4+ T cell accumulation and abundance in the intestinal LP (Fig. 5C; data not shown). Altogether, these data highlight the importance of MHC II on CD11c+ cells and suggests that DCs may be specialized in providing antigenic stimulation to promote the development of Th17 cells in the presence of SFB-containing microbiota.

FIGURE 5.

MHC II expression on CD11c+ cells is important for intestinal Th17 differentiation induced by SFB-containing microbiota. Litter- and cage-mate MHC II-floxed (MHC II^FF) mice and CD11c-cre MHC II^FF (MHC II^ΔDC) mice were gavaged with SFB⁺ CC and CD4+ T cell differentiation amongst the LPL of the small intestine was assessed on day 10 post-gavage. Representative FACS plots (A) and corresponding bar graphs (B) of Th17 (top panels), Foxp3+ Treg (middle panels), and Th1 (bottom panels) cells amongst LPL for MHC II^ΔDC and MHC II^FF mice. (C) Accumulation of donor CD45.1 CD4+ T cells in the SI LP of MHC II^ΔDC and MHC II^FF mice on day 10 post-gavage. Data are representative of two independent experiments with four to five mice per group. *, p ≤ 0.05; Not statistically significant (N.S.), p > 0.05 using a Student’s t test.

Cognate antigen promotes intestinal Th17 differentiation in the presence of SFB-containing microbiota

Our previous data demonstrates that Th17 differentiation induced by SFB-containing microbiota is MHC II-dependent (Fig. 4, Fig. 5), however, specific SFB-derived antigens that may induce intestinal Th17 cells have not yet been defined. Thus, we investigated whether a model food antigen is sufficient using an antigen-specific CD4+ T cell transfer system. Naïve CD45.1+ CD4+ T cells and Thy1.1+ OT-II cells were enriched from Spl and pLN, respectively, and were mixed at a 1:1 ratio followed by adoptive transfer into CD45.2⁺ JAX B6 recipients void of SFB (data not shown). The purity of donor cells was verified to be >99% IL-17A−CD4+ T cells (data not shown). One day later, recipients were gavaged SFB⁺ CC and/or fed the cognate antigen for OT-II cells, chicken ovalbumin (OVA), for 10 days in the drinking water. On day 10, mice were euthanized and intestinal Th17 differentiation was assessed amongst host and donor (both CD45.1+ and Thy1.1+ OT-II) CD4+ T cell populations (Fig. 6A). As expected, host CD4+ LPL (CD45.1−Thy1.1−) differentiated into Th17 cells following gavage of SFB⁺ CC (Fig. 6B, 6C). The donor CD45.1+Thy1.1−CD4+ LPL responded similarly to the host LPL (CD45.1−Thy1.1−) in robustly differentiating to Th17 cells following gavage of SFB⁺ CC, relative to mice that were not gavaged (Fig. 6B, 6D). Importantly, OT-II LPL (CD45.1−Thy1.1+Vα2+Vβ5+) differentiated into Th17 cells comparable to host CD4+ LPL (CD45.1− Thy1.1−) and donor CD45.1+Thy1.1−CD4+ LPL only in mice given SFB⁺ CC and OVA (Fig. 6B, 6E). In the context of OVA without SFB⁺ CC, a small proportion of the OT-II cells differentiated into Th17 cells. With the absence of cognate antigen, mice gavaged SFB⁺ CC yielded a paucity of donor OT-II cells and the corresponding Th17 cells were negligible (data not shown). Overall, both cognate antigen and specific microbiota are required for robust intestinal Th17 differentiation and the cognate antigen does not have to be of bacterial (SFB) origin.

FIGURE 6.

Intestinal Th17 differentiation induced by SFB-containing microbiota is dependent on antigenic stimulation and the conditioned intestinal microenvironment. (A) Naïve CD4+CD25− cells were enriched from the Spl and LN of CD45.1 and OT-II mice, respectively, and adoptively transferred at a 1:1 ratio into JAX CD45.2 B6 mice void of SFB on day −1. On day 0, mice were gavaged SFB⁺ CC and/or fed chicken ovalbumin (OVA) in the drinking water for 10 days. On day 10, mice were euthanized, SI LPL were isolated, and intestinal Th17 differentiation amongst CD4+ T cells of the host (CD45.1−Thy1.1−) and donor CD45.1+Thy1.1− and donor CD45.1−Thy1.1+ OT-II (Vα2+Vβ5+) cells was assessed. (B) Representative FACS plots for host and donor SI LPL evaluated for IL-17A expression in the three different conditions of: OVA only, SFB⁺ CC only, or OVA and SFB⁺ CC. FACS plots were pre-gated on TCRβ+CD4+ cells. Comparison of intestinal Th17 induction relative to OVA only group for host (C), donor CD45.1+ (D), and donor Thy1.1+ OT-II LPL (E). Data are representative of two independent experiments with three to four mice per group. Error bars represent SEM. *, p ≤ 0.05; not statistically significant (N.S.), p > 0.05 using a Student’s t test.

Discussion

Here, we demonstrate that the accumulation of naïve CD4+ T cells in the intestine and the development of intestinal Th17 cells in response to microbiota containing SFB did not require the Spl, mLN, and other LT-dependent lymphoid structures. Furthermore, using a CD4+ T cell transfer system, intestinal Th17 differentiation was shown to require MHC II expression by CD11c+ cells and could be induced by a model food antigen. These results suggest that specific components of the microbiota are important in conditioning the local intestinal milieu to facilitate the differentiation of Th17 cells upon antigenic stimulation by LP DCs in situ.

The intestine is unique among organs in that it harbors large numbers of Th17 and Foxp3+ regulatory T cells in the steady-state (4). The presence of these T cell subsets in the intestinal LP is profoundly influenced by specific components of the microbiota (11, 12, 22, 23), and their metabolites (22, 24) since these bacteria interact directly with intestinal epithelial cells (25) and may be sampled by underlying LP DCs and macrophages (26). The site of this steady-state CD4+ T cell differentiation has been assumed to be in the mLN based upon several key observations: 1) naïve CD4+ T cells primarily traffic through secondary lymphoid tissues and not the intestine, 2) Peyer’s patch (PP) and LP DCs migrate to the mLN where they present antigens to naïve T cells resulting in their expansion and induction of gut homing molecules (27–30), and 3) delivery of soluble antigen via the oral route induces Foxp3+ T cell differentiation in the mLN (31–33). Importantly, none of these observations are inconsistent with CD4+ T cell differentiation taking place directly within the intestinal LP. Additionally, previous reports demonstrated that PP and colonic patches are dispensable for the differentiation of intestinal Th17 cells (9).

While CD4+ T cells primarily traffic through the Spl and lymph nodes, we show that they are abundant in the intestinal LP both before and after weaning. In fact, naïve T cells can enter various non-lymphoid organs as part of a normal migratory pathway (34). While intestinal DCs do migrate via the afferent lymphatics to the mLN (35), this process is dramatically augmented by inflammatory stimuli (36) suggesting that in the steady-state only a fraction of DCs migrate to the mLN, while the majority remains in the LP. Macrophages are also abundant in the intestinal LP and their migration to the mLN is regulated by the microbiota (37, 38). Thus, the intestinal LP contains all of the necessary requirements for CD4+ T cell priming and differentiation: abundant numbers of naïve CD4+ T cells, MHC II bearing DCs and macrophages, and a microbiota-induced local milieu.

While our data demonstrate that GALT and mLN are not required for SFB-induced intestinal Th17 differentiation, they do not imply that these lymphoid structures play no role in this process. In fact, LT-dependent lymphoid structures have been reported to influence steady-state intestinal Th17 cells in mice lacking SFB (39) and our data are consistent with these observations since Lta^−/− mice from JAX that are void of SFB demonstrated a steady-state reduction in intestinal Th17 cells. These data may be explained by the enhanced skewing of intestinal CD4+ T cells toward the Th1 subset in Lta^−/− mice (data not shown). Thus, the GALT/mLN and intestinal LP may make unique and perhaps overlapping contributions to intestinal Th17 differentiation depending on whether they are “naturally-derived” (40, 41) or induced in response to specific components of the microbiota.

Currently, the antigenic specificity of Th17 cells that reside in the intestinal LP at steady-state remains undefined (42). The requirement for specific components of the microbiota indicates that the TCR repertoire of these cells may be reactive to bacterial antigens (12). Cytokines and other factors within the intestinal tissue may also promote Th17 differentiation upon priming of naïve CD4+ T cells by microbial-, self-, and/or food-derived antigens. The reduced intestinal Th17 development we observed in MHC II^Δ/Δ mice colonized with SFB strongly suggests that antigenic stimulation of naïve CD4+ T cells is essential for intestinal Th17 differentiation, and conditioning by specific components of the microbiota alone is insufficient to drive this process. Furthermore, Th17 differentiation of naïve OT-II cells within the intestinal LP of mice was only observed when both OVA and SFB-containing microbiota were present. These findings are consistent with in vitro studies demonstrating that both TCR stimulation and specific cytokines are required for Th17 differentiation (6, 43). While intestinal Th17 cells may directly respond to SFB-derived antigens, the overall antigen reactivity of intestinal Th17 cells is clearly not limited to SFB since OVA, a food antigen, is sufficient. While SFB has been shown to promote the development of Th17 cells, the role of other bacteria in the SFB-containing microbiota that may influence intestinal Th17 cell differentiation cannot be excluded in our SFB⁺ CC transfer system. Thus, the antigen specificity of intestinal Th17 cells may encompass reactivity to select bacteria, food- and/or self-antigens. Further, investigations into the TCR specificity of “natural” and induced Th17 cells are clearly warranted. In summary, our findings highlight several previously unappreciated aspects of specific microbiota-induced Th17 differentiation and suggests that the intestinal LP may be an important site for this process.

Abbreviations used in this article

JAX

Jackson Laboratory

LI

large intestine

LP

lamina propria

LPL

lamina propria lymphocytes

LT

lymphotoxin

mLN

mesenteric lymph nodes

OVA

ovalbumin

pLN

peripheral lymph nodes

PP

Peyer's patches

SFB

segmented filamentous bacteria

SFB⁺ CC

SFB-containing cecal contents

SI

small intestine

Spl

spleen