Activation of the Aryl Hydrocarbon Receptor by 10-Cl-BBQ Prevents Insulitis and Effector T Cell Development Independently of Foxp3⁺ Regulatory T Cells in Nonobese Diabetic Mice Free

Targeting T cells is a promising therapeutic strategy for the prevention or treatment of autoimmune diseases as a way to improve efficacy and minimize the toxicity that is associated with most currently used therapeutics. In this regard, the aryl hydrocarbon receptor (AhR) represents a potential new drug target as a ligand-activated transcription factor that directly alters T cell differentiation without cytotoxicity (1–3). Studies with 2,3,7,8-tetrachlordibenzo-p-dioxin (TCDD), the most potent AhR ligand, have shown that AhR signaling during the early stages of CD4⁺ T cell activation results in premature cessation of effector T cell (Teff) differentiation in association with the induction of Tr1-like regulatory T cells (Tregs) (3–5). More recently, both direct and indirect (via tolerogenic dendritic cells [DCs]) induction of Foxp3⁺ Tregs has been associated with AhR-mediated immune suppression (2, 6, 7). Treatment of mice with TCDD has been shown to prevent or ameliorate several different types of autoimmune and allergic diseases supporting the AhR as a therapeutic target (6, 8–11).

Although studies with TCDD highlight the therapeutic potential of the AhR pathway, TCDD itself has unfavorable pharmacological properties specifically related to its high lipid solubility and resistance to metabolism, which lead to a half-life of ∼7 y in humans (12). Additionally, the notoriety of TCDD as an environmental toxicant would likely limit its acceptance for pharmacologic uses. To identify other high-affinity AhR ligands that exhibit more favorable pharmacokinetic properties than TCDD, we used small-molecule screening to discover a novel immunosuppressive AhR ligand, 10-chloro-7H-benzimidazo[2,1-a]benzo[de]iso-quinolin-7-one (10-Cl-BBQ) (13). This nontoxic compound was shown to have an in vivo half-life of approximately 2 h but could be administered to mice in a repeated dosing scheme to maintain AhR activation over time, similar to the prolonged AhR activation induced by TCDD. Using an acute graft-versus-host model, we found that 10-Cl-BBQ, similar to TCDD, directly targeted the AhR in CD4⁺ T cells to induce an AhR-dependent Treg phenotype, suppress the allogeneic CTL response, and ameliorate the symptoms of graft-versus-host disease (GVHD) (13).

In the present study, we investigated the potential for 10-Cl-BBQ to suppress the development of type 1 diabetes (T1D), a T cell–mediated autoimmune disease in which islet-specific T cells destroy the insulin-producing β cells in the pancreas, resulting in hyperglycemia (14). Multiple mechanisms have been ascribed to influencing the development of T1D, including defective control of aberrant immune responses by Tregs (15, 16). In the NOD mouse model for T1D, the spontaneous and progressive destruction of β cells is suppressed by treatment with TCDD, and hyperglycemia is completely prevented (9). Suppression of T1D by TCDD was also associated with an increase in frequency of Foxp3⁺ Tregs, in concordance with numerous other studies that have shown a correlation between autoimmune disease suppression by AhR ligands and increased frequency of Foxp3⁺ Tregs (2, 7, 8, 10, 17–20). However, a direct role for Foxp3⁺ Tregs as a primary mechanism of AhR-mediated immune suppression has not been clearly established.

In this study, we demonstrate the therapeutic efficacy of 10-Cl-BBQ in T1D as evidenced by the almost complete prevention of the development of insulitis in NOD mice with chronic treatment. Furthermore, using NOD.Foxp3^DTR mice, we show that 10-Cl-BBQ suppresses insulitis in the absence of Foxp3⁺ Tregs and that inhibition of Teff development does not rely on Foxp3⁺ Tregs. These results directly challenge the widely supported proposition that induction of Foxp3⁺ Tregs is the primary mechanism by which AhR ligands mediate suppression of immune-mediated diseases.

NOD/LtJ mice were originally purchased from The Jackson Laboratory (Bar Harbor, ME). NOD.AhR^−/− mice were generated by backcrossing B6.129-AHR^tm1Bra/J onto the NOD/LtJ background for 13 generations. The Ahr locus is present on chromosome 12 (15–16.5 cM), and the nearby genes have not been linked to the diabetic process based on annotated protein homology, proposed functions, or analyses of previously published mutations. NOD.Foxp3^DTR mice (21) were obtained from the Juvenile Diabetes Research Foundation Center for Immune Tolerance in Diabetes at Harvard University. Mice were bred and maintained under specific pathogen-free conditions at Oregon State University facilities. All experimental procedures using animals were approved by the Institutional Animal Care and Use Committee at Oregon State University.

10-Cl-BBQ was dissolved in DMSO/Cremophor/Peceol (30/20/50%) and delivered by oral gavage. For the time-course experiments, mice were treated with 10-Cl-BBQ starting at 7 wk of age and insulitis was evaluated at 12, 15, and 20 wk of age. The dosing strategy was continually optimized from the first 12-wk experiment (alternating 60 and 30 mg/kg, daily), to the 15-wk experiment (60 mg/kg every other day), to the 20-wk experiment (60 mg/kg, three times per week) to minimize the number of doses of 10-Cl-BBQ that needed to be administered yet would maintain activation of AhR as evidenced by Cyp1a1 expression. We established 60 mg/kg treatment with 10-Cl-BBQ by oral gavage three times per week as the optimal dosing schedule. Twelve- and 15-wk experiments were repeated with this optimized dosing schedule for subsequent studies. As a positive control, mice were treated with TCDD (50 μg/kg loading dose, followed by 15 μg/kg every other week) by oral gavage.

RNA was isolated from liver using RNeasy columns (Qiagen), and cDNA was synthesized using a SuperArray ReactionReady first-strand cDNA synthesis kit. PCR reactions were performed on an ABI Prism 7500 real-time PCR system (Applied Biosystems) using SYBR Green/ROX Mastermix (SABiosciences). Cyp1a1 levels were normalized to Actb using primers from SABiosciences.

Blood glucose was measured weekly using a Glucocard Vital glucose meter (Arkray USA, Minneapolis, MN). A blood glucose level >200 mg/dl for 2 consecutive days was considered diabetic. Insulitis was scored on sequential H&E-stained pancreas sections separated by 200 μm. At least 50 islets per pancreas were scored as no infiltration, peri-infiltration, <50% infiltration, or >50% infiltration.

At the indicated time points, mice were sacrificed, pancreatic lymph nodes (PLN) and pancreata were excised, and single-cell suspensions were prepared. FcRs were blocked with rat IgG (Jackson ImmunoResearch Laboratories, West Grove, PA) and the cells were stained with the following Abs from eBioscience (San Diego, CA): CD4 (RM4), CD8 (53-6.7), CD19 (1D3e), Foxp3 (FJK-16s), neuropilin 1 (Nrp1; 3DS304M), CD25 (PC61.5), IL-17 (eBio17B7), T-bet (eBio4B10), and retinoic acid–related orphan receptor (ROR) γt (AFKJS-9). CD45 (30-F11) and IFN-γ (XMG1.2) were from BD Biosciences. For intracellular staining, cells were fixed and permeabilized using Foxp3 fixation/permeabilization buffer (eBioscience). For cytokine staining, cells were stimulated with PMA, ionomycin, brefeldin A, and monensin (eBioscience) for 4 h in culture. Data were acquired using an FC-500 flow cytometer (Beckman Coulter, Brea, CA) and compensated and analyzed using FlowJo software (Tree Star, Ashland, OR). Fluorescence minus one controls were used for setting gates for analysis.

For ELISAs, splenocytes were stimulated with PMA/ionomycin (eBioscience) for 4 h. IFN-γ and IL-17 were measured in culture supernatants according to the manufacturer’s protocol (eBioscience).

Foxp3⁺ cells were depleted from NOD.Foxp3^DTR mice as previously described (21) with some modifications. Briefly, 4- to 6-wk-old NOD.Foxp3^DTR mice were injected i.p. with 500 ng diphtheria toxin (DT; Sigma-Aldrich, St. Louis, MO) on days 0, 1, 3, 5, and 7 and sacrificed on day 9.

Data are presented as mean of biological replicates. For statistical comparisons of degree of islet infiltration between groups, χ² analyses were performed. A Student t test was used for analysis of all other data. A p value ≤0.05 was considered significant.

Chronic treatment of NOD mice with the prototypic AhR ligand, TCDD, was previously shown to prevent insulitis and resulting hyperglycemia as long as activation of the AhR was maintained (9). We hypothesized that the efficacy of TCDD could be recapitulated with the novel, rapidly metabolized compound 10-Cl-BBQ (chemical structure shown in Fig. 1A) if the dosing schedule ensured continued activation of AhR. The dose of 10-Cl-BBQ was calculated empirically by determining a concentration that would induce equivalent Cyp1a1 expression as a body burden of 25 μg/kg TCDD, the therapeutic dose in NOD mice. Because NOD mice express the low responder AhR^d allele, they require a higher dose of 10-Cl-BBQ to achieve comparable activation of AhR when compared with AhR^b-expressing C57BL/6 mice (13). Oral treatment with 60 mg/kg 10-Cl-BBQ induced commensurate expression of Cyp1a1 as the therapeutic dose of TCDD in NOD mice (Fig. 1B).

A time course was conducted to evaluate the capacity of 10-Cl-BBQ to prevent islet infiltration at both early and later stages of disease progression. Treatment was initiated at 7 wk of age with 60 mg/kg 10-Cl-BBQ administered by oral gavage on Monday, Wednesday, and Friday established as an optimal dosing schedule (Fig. 1C). This regimen allowed for the maintenance of AhR activation while minimizing complications from frequent oral administration of the vehicle carrier, and it resulted in significant suppression of diabetes incidence through 20 wk of age (Fig. 1D). At 12 wk of age, 65 ± 3.4% of islets in the pancreas of vehicle-treated mice were infiltrated with inflammatory cells with 27.9 ± 1.2 and 37.5 ± 3.1 showing <50% and >50% of the islets infiltrated, respectively. In contrast, treatment with 10-Cl-BBQ successfully prevented infiltration, with 92 ± 2.0% of islets free from infiltration and the remaining 8% showing only minimal infiltration (Fig. 1E, 1F). Similarly, at 20 wk of age when control mice had more severe infiltration with 28.8 ± 4.5 and 56.8 ± 7.8 <50% and >50% of the islet infiltrated, respectively, 10-Cl-BBQ significantly suppressed infiltration in 73.1 ± 6.2% of islets (Fig. 1G). To determine the role of AhR in mediating the effects of 10-Cl-BBQ in suppressing T1D, we generated AhR-deficient NOD mice. AhR expression was required for 10-Cl-BBQ to induce its therapeutic effects, as the same treatment in AhR-deficient NOD mice had no effect on insulitis (Fig. 1H). At the doses used, treatment with 10-Cl-BBQ was more effective at preventing islet infiltration than was TCDD (Fig. 1F, 1G). Because the extent of islet infiltration is directly linked to development of elevated blood glucose and overt diabetes, islet infiltration is useful as a reliable surrogate biomarker for more prolonged prevention studies. The strong suppression of islet infiltration in 10-Cl-BBQ–treated mice clearly demonstrates its therapeutic potential. The absence of clinical signs of toxicity associated with prolonged treatment with 10-Cl-BBQ through 20 wk of age also support the therapeutic potential of this novel AhR ligand (Supplemental Table I).

Tregs are capable of suppressing islet infiltration in NOD mice (21, 22), and suppression of diabetes in TCDD-treated NOD mice was associated with an increase in Foxp3⁺ Tregs at 30 wk of age in the PLN (9). To determine whether suppression of insulitis by 10-Cl-BBQ was associated with increased Foxp3⁺ Tregs, the leukocyte composition of the pancreas and PLN was assessed at different time points during disease progression. In the pancreas at 12 wk of age, the total number of CD45⁺ cells was significantly decreased in 10-Cl-BBQ–treated mice (Fig. 2A), correlating with the reduction of islet infiltration induced by 10-Cl-BBQ treatment. 10-Cl-BBQ treatment did not alter the proportion of CD19⁺ B cells or CD8⁺ T cells in the pancreas (Fig. 2B, 2C), but it significantly reduced the percentage of infiltrating CD4⁺ T cells (Fig. 2D). Additionally, the percentage of pancreatic CD4⁺ T cells that expressed Foxp3 was increased by treatment (Fig. 2E). However, as a consequence of the decrease in total CD45⁺ cells and in the percentage of CD4⁺ cells, the total number of Foxp3⁺ cells in the pancreas was unaltered (Fig. 2F).

As the percentage of Foxp3⁺ cells was increased in the pancreas of 10-Cl-BBQ–treated mice at 12 wk of age, this population was further studied over time in the PLN at 12, 15, and 20 wk of age. At 15 and 20 wk of age, there was a small but significant increase in the percentage of CD4⁺ cells that were Foxp3⁺ cells (Fig. 3A, 3B); this trend was not found for the total number of Foxp3⁺ cells in the PLN (Fig. 3C). Instead, at 12 wk of age, the total number of CD4⁺Foxp3⁺ cells was lower in 10-Cl-BBQ mice, owing to the reduction in the proportion of CD4⁺ cells at this time point (Fig. 3C, 3D). Similar findings were found in the PLN of TCDD-treated mice and in the spleen of 10-Cl-BBQ or TCDD mice, with small increases in the frequency of CD4⁺ cells that were Foxp3⁺ (Supplemental Fig. 1).

To further characterize the Foxp3⁺ Tregs that are increased by AhR activation, coexpression with the high-affinity IL-2 receptor, CD25, and Nrp1 was assessed in the PLN at 20 wk of age. The increased percentage of Foxp3⁺ cells in the PLN were CD25⁺ (Fig. 4A) and Nrp1⁻ (Fig. 4B) in 10-Cl-BBQ–treated mice. These results suggest that AhR activation is not expanding the population of thymically derived Tregs but instead is inducing or predominantly maintaining peripherally derived Tregs (23).

Treatment of NOD mice with 10-Cl-BBQ reduced the number of CD4⁺ T cells in the pancreas, lymph node, and spleen at 20 wk of age, suggesting a reduction in Th1 and/or Th17 effector cells, which have been implicated in T1D pathogenesis (24). AhR activation by both 10-Cl-BBQ and TCDD suppressed IL-17 production by splenic CD4⁺ cells as measured by intracellular staining (Fig. 5A) and ELISA (Fig. 5C). In contrast, IFN-γ production was not altered by AhR activation (Fig. 5B, 5D). These results suggest that AhR activation preferentially inhibits the development of Th17 cells in T1D.

10-Cl-BBQ treatment resulted in an increased percentage of Foxp3⁺ cells in the pancreas and PLN of NOD mice, consistent with the effect of activation of AhR in other autoimmune disease models. However, the functional significance of the AhR-induced Foxp3⁺ Tregs in suppressing Teffs and disease has not been clearly demonstrated. To directly determine whether Foxp3⁺ Tregs are functionally required for suppression of insulitis by AhR ligands, NOD.Foxp3^DTR mice were used to specifically ablate Foxp3⁺ cells upon administration of DT (21). Starting at 4 wk of age, NOD.Foxp3^DTR mice were injected with DT on days 0, 1, 3, 5, and 7 and sacrificed on day 9. NOD.Foxp3^WT mice treated with DT, and NOD.Foxp3^DTR mice treated with vehicle were used as controls. Concurrently, AhR ligand–treated mice were orally dosed with 60 mg/kg 10-Cl-BBQ (days −1, 1, 3, 5, and 7) or TCDD (day −1) relative to DT treatment on day 0 (Fig. 6A). Two days following the initiation of DT treatment, ∼80% of Foxp3⁺ cells were depleted (Fig. 6B). By day 9, 50% of Foxp3⁺ cells were depleted, consistent with the reported rebound in Foxp3⁺ cells following DT administration (21). 10-Cl-BBQ did not alter the depletion efficacy of DT (Fig. 6C). Ablation of Foxp3⁺ cells led to rapid and extensive infiltration in the pancreas with 83.5 ± 16.5% of islets in vehicle-treated mice showing infiltration (30.5 ± 23.5 and 52.4 ± 38.8 with <50% and >50% of the islet infiltrated, respectively). Strikingly, 10-Cl-BBQ treatment prevented infiltration in 73.2 ± 25.2% of islets (Fig. 6D), an effect recapitulated by TCDD treatment (Fig. 6E). Thus, AhR activation was not only able to suppress infiltration in the absence of Foxp3⁺ cells, but it also compensated for the lack of Foxp3⁺ cells to suppress infiltration in their absence.

Because Foxp3⁺ cells were not required for 10-Cl-BBQ-mediated suppression of insulitis, an alternative mechanism by which 10-Cl-BBQ might mediate suppression is through induction of a Foxp3⁻CD25⁺ Tregs. In an acute GVHD model (3, 25), and in in vitro studies (26), AhR activation has been shown to induce a Tr1-like Treg, characterized by CD25 expression in conjunction with CTLA4, GITR, and IL-10. Additionally, AhR activation has been recently shown to transdifferentiate Th17 cells into Tr1 cells in vitro (27). Although the percentage of CD25⁺ cells increased in the PLN following depletion of Foxp3⁺ cells, 10-Cl-BBQ treatment did not alter the percentage of CD25⁺Foxp3⁻CD4⁺ cells in comparison with vehicle-treated mice (Fig. 7A). Additionally, CD4⁺ cells expressing the Tr1 markers CD49b⁺ and Lag3⁺ (28) were not increased by 10-Cl-BBQ treatment (Fig. 7B). These results do not support an AhR-mediated induction of Tr1 cells in the NOD model.

Nrp1 expression on Foxp3⁻ cells has been associated with an activated Teff phenotype (29). Nrp1 expression was assessed on CD4⁺Foxp3⁻ cells to determine whether Nrp1 was altered by AhR activation in the NOD Foxp3.DTR model. DT treatment, and subsequent islet infiltration, resulted in an increase in CD4⁺Nrp1⁺Foxp3⁻ cells from 4.3 ± 0.6% in nondepleted mice to 12.6 ± 2.7% in DT-treated mice. This increase in Nrp1⁺ cells was not observed in mice treated with 10-Cl-BBQ or TCDD (Fig. 8A, 8B), suggesting that these cells are associated with the development of a pathogenic T cell population. To test this hypothesis, the percentage of Nrp1⁺Foxp3⁻ cells was plotted versus the percentage of highly infiltrated islets for each vehicle-treated mouse injected with DT; the percentage of CD4⁺Nrp1⁺Foxp3⁻ cells was positively and significantly correlated with disease severity (Fig. 8C). CD4⁺Nrp1⁺Foxp3⁻ cells were also reduced in the PLN of NOD wild-type mice treated with 10-Cl-BBQ. Likewise, CD4⁺Nrp1⁺Foxp3⁻ cells were decreased, albeit nonsignificantly (p = 0.09), in TCDD-treated mice (Fig. 8D), reflecting a lower efficacy of TCDD to suppress islet infiltration in comparison with treatment with 10-Cl-BBQ (Fig. 8D, 8E). Similar findings were observed in the spleen (Supplemental Fig. 2). In contrast to CD4⁺Nrp1⁺Foxp3⁻ cells, the percentage of CD4⁺Foxp3⁺ cells did not correlate with the degree of islet infiltration (Fig. 8E).

The detrimental role of Th17 cells in exacerbating T1D has been widely established (24, 30–33). To further address whether AhR activation was inhibiting the induction of a pathogenic Teff population, coexpression of RORγt, the transcription factor for Th17 cells, with CD4⁺Nrp1⁺Foxp3⁻ cells was assessed. In the CD4⁺Foxp3⁻ population, all RORγt⁺ cells coexpressed Nrp1 (Fig. 8F). Additionally, and consistent with the AhR-mediated reduction in CD4⁺Nrp1⁺Foxp3⁻ cells, AhR activation by TCDD reduced the percentage of CD4⁺ cells that were RORγt⁺ (Fig. 8G). TCDD treatment also reduced the percentage of CD4⁺ cells that were T-bet⁺ in mice that were depleted of Foxp3⁺ cells (Supplemental Fig. 3), although CD4⁺IFN-γ⁺ cells were unaltered by AhR activation in Foxp3 replete mice. Taken together, these data demonstrate that AhR activation mediates the suppression of pathogenic Th17 cells in NOD mice and this occurs independently of Foxp3⁺ Tregs.

The AhR represents an unexploited target for treatment of immune-mediated diseases. Activated by exogenous compounds, AhR-driven suppression of Th1-, Th2-, and Th17-mediated immune responses has been extensively described (4, 6, 34). Most of our knowledge on the immunosuppressive effects of AhR activation have been derived from studies using TCDD, a potent ligand that is not suitable for human therapy due to its propensity to bioaccumulate in the body. We previously reported results from a small molecule screen to identify high-affinity AhR ligands with a more favorable pharmacokinetic profile than the metabolism-resistant TCDD. 10-Cl-BBQ was identified as a noncytotoxic compound with the ability to activate the AhR at nanomolar concentrations, and with repeated dosing could maintain AhR activation despite a serum half-life of ∼2 h. Furthermore, treatment with 10-Cl-BBQ led to similar in vivo immunological outcomes as did the therapeutic dose of TCDD, inducing AhR-dependent CD4⁺CD25⁺ Tregs (3) as effectively as did TCDD and dose-dependent increases in the expression of known AhR-associated genes in an acute parent-into-F₁ graft-versus-host model (13). In the present study, we examined the utility of 10-Cl-BBQ to maintain AhR activation in a chronic disease model. We found that repeated dosing with 10-Cl-BBQ effectively suppressed islet infiltration in NOD mice, and treatment from 7 to 20 wk of age was not associated with clinical measures of toxicity, emphasizing 10-Cl-BBQ as a therapeutic candidate.

The mechanism by which AhR activation leads to its profound immunosuppressive effects is not fully understood. In our group, experiments in an acute GVHD model using TCDD highlight the importance of AhR signaling during early stages of T cell activation. For suppression of the CTL response, AhR activation must be initiated within the first 3 d following transfer of donor cells into the F₁ host. This model further highlighted CD4⁺ T cells as a critical target for AhR activation; AhR expression was required in donor CD4⁺ T cells in order for TCDD to induce its suppressive effects (5). In other models, DCs have been shown to play a role in AhR-driven immune suppression. For example, using an experimental autoimmune encephalitis model with Rag1 or CD11c conditional AhR knockout mice showed that the absence of AhR in either population leads to a decrease in AhR-mediated immune suppression (35). Additional studies have described the role of both CD4⁺ T cells and DCs as targets for AhR (2, 7, 36).

Many studies have demonstrated that AhR activation in either T cells or tolerogenic DCs leads to an increase in Foxp3⁺ Tregs (2, 7, 9, 10, 17–20, 37). However, there is recent debate over whether Foxp3⁺ Tregs are required for the AhR-mediated suppression of Teff responses (35, 37). There is also mixed evidence for AhR activation increasing the total number of Foxp3⁺ cells or just increasing their frequency alone (8, 18, 20, 37), making it unclear whether improving the Treg/Teff balance is critical for immunosuppression. Alternatively, the reduction in the Teff population alone may be sufficient to minimize pathology. Several reports tried to tease apart the role of Tregs in the AhR suppressive response using anti-CD25–neutralizing Abs to indirectly target Foxp3⁺ cells. For example, in the peanut allergy mouse model, where TCDD treatment led to an increase in the percentage but not total numbers of splenic CD25⁺Foxp3⁺ cells, anti-CD25 treatment reduced CD25⁺Foxp3⁺ cells but did not affect CD25⁻Foxp3⁺ cells. This treatment partially abolished some, but not all, of TCDD’s suppressive effects (8). Similarly, studies in an autoimmune uveitis model showed that anti-CD25 treatment partially reversed TCDD-mediated reduction of disease score (10). These results suggest that CD25⁺ cells play a role in AhR-mediated immune suppression; however, because anti-CD25 may have off-target effects, it is difficult to definitively implicate Foxp3⁺ cells in the observed effects. In the present study, we found an increase in the percentage but not total numbers of Foxp3⁺ cells following AhR activation, as a result of a decrease in the number of CD4⁺ cells in both the pancreas and PLN. Critically, specific depletion of Foxp3⁺ Tregs in NOD.Foxp3^DTR mice demonstrated that Foxp3⁺ Tregs are not required for AhR-mediated suppression of insulitis. This conclusion was supported by a lack of correlation between the percentage of Foxp3⁺ Tregs and the severity of islet infiltration in vehicle-treated NOD mice. Collectively, these results indicate that a reduction in Teff number is the driving force of AhR-mediated immune suppression, rather than an improved Treg/Teff balance.

Support for the hypothesis that the reduction in Teff number is the primary event in AhR-driven suppression of insulitis comes from the analysis of CD4⁺Foxp3⁻ cells. Nrp1 expression on CD4⁺Foxp3⁻ cells strongly correlated with the degree of islet infiltration in NOD mice. Furthermore, the coexpression of Nrp1with RORγt, the transcription factor for disease-inducing Th17 cells (30, 38) in CD4⁺Foxp3⁻ cells, strengthens the idea that these cells represent a pathogenic effector population. Nrp1⁺Foxp3⁻ cells have previously been described as activated/memory T cells that were greatly increased in TGF-βRII–deficient mice with autoinflammatory disease (29). Nrp1 is a receptor for TGF-β (39), which in the context of IL-6 leads to Th17 differentiation (40). As a result of the proinflammatory microenvironment in NOD mice (41), Th17 differentiation is consistent with the coexpression of RORγt with Nrp1 on CD4⁺Foxp3⁻ cells. Regardless of whether Nrp1 expression on CD4⁺Foxp3⁻ cells in the PLN is functionally involved in the development of pathogenic T cells, CD4⁺Nrp1⁺Foxp3⁻ cells may have utility as a biomarker for the extent of islet infiltration in the pancreas. In fact, both 10-Cl-BBQ and TCDD treatments reduced the frequency of Nrp1⁺ CD4⁺Foxp3⁻ cells, which directly correlated with reduced islet infiltration. Functionally, Nrp1 is expressed in the immunological synapse and has been implicated in productive signaling between T cells and DCs (42). A decrease in Nrp1 expression following treatment with AhR ligands could result in altered crosstalk between T cells and DCs, impairing T cell differentiation and precluding Teff accumulation.

T1D is a chronic disease, and it has recently been shown that islet inflammatory lesions are dynamic with constant influx of new T cells entering the islet (43). Because early T cell activation events are targeted by AhR signaling (4), there is continuous opportunity for AhR ligands to interfere with the activation/differentiation of new autoantigen-specific T cells. The possibility that 10-Cl-BBQ treatment will be able to halt β cell destruction during the “honeymoon” period in humans following early diagnosis of T1D is of particular interest (44).

Collectively, to our knowledge, this study is the first to demonstrate that specifically depleting Foxp3⁺ cells does not abrogate immune suppression by AhR ligands. These findings directly challenge the dogma that Foxp3⁺ Tregs are responsible for the therapeutic effects of AhR ligands in immune-mediated diseases. Instead, the reduction in Teff appears to be sufficient for the prevention of disease. AhR activation acts to compensate for the depletion of Foxp3⁺ Tregs, and thus AhR ligands might be useful as an alternate treatment strategy in diseases characterized by defective Treg numbers or functionality. Our results highlight 10-Cl-BBQ as an effective and nontoxic AhR ligand for the therapeutic intervention of immune-mediated diseases.

We thank Samuel Bradford for FACS support, Brian Dolan and Patrick Videaup for helpful discussions, and Shanthakumar Tyavanagimatt and Prasad Kopparapu for formulation of 10-Cl-BBQ.

The authors have no financial conflicts of interest.