Inflammatory bowel disease (IBD) is an expanding autoimmune disease afflicting millions that remains difficult to treat due to the accumulation of multiple immunological changes. BT-11 is an investigational new drug for IBD that is orally active, gut restricted, and targets the lanthionine synthetase C-like 2 immunometabolic pathway. CD25+ FOXP3+ CD4+ T cells are increased locally within the colon of BT-11–treated mice in Citrobacter rodentium and IL-10−/− mouse models of colitis. The maintained efficacy of BT-11 in the absence of IL-10 combined with the loss of efficacy when direct cell–cell interactions are prevented suggest that the regulatory T cell (Treg)–related elements of suppression are cell contact–mediated. When PD-1 is inhibited, both in vitro and in vivo, the efficacy of BT-11 is reduced, validating this assertion. The depletion of CD25+ cells in vivo abrogated the retention of therapeutic efficacy postdiscontinuation of treatment, indicating that Tregs are implicated in the maintenance of tolerance mediated by BT-11. Furthermore, the involvement of CD25 suggested a role of BT-11 in IL-2 signaling. Cotreatment with BT-11 and IL-2 greatly enhances the differentiation of CD25+ FOXP3+ cells from naive CD4+ T cells relative to either alone. BT-11 enhances phosphorylation of STAT5, providing a direct linkage to the regulation of FOXP3 transcription. Notably, when STAT5 is inhibited, the effects of BT-11 on the differentiation of Tregs are blocked. BT-11 effectively enhances the IL-2/STAT5 signaling axis to induce the differentiation and stability of CD25+ FOXP3+ cells in the gastrointestinal mucosa to support immunoregulation and immunological tolerance in IBD.
Inflammatory bowel disease (IBD) is a significant public health problem in the United States and worldwide. The complex etiology of IBD includes an interplay between genetics, bacteria, and environmental factors (1). Immune dysfunction, through overproduction of TNF-α and other inflammatory products, causes intestinal epithelial cells to die, resulting in epithelial erosion and disruptions of the gut lining (2). Once there is weakness in the epithelial barrier, bacteria are able to translocate and penetrate the tissue. Intestinal immune cells are activated to kill the bacteria, causing inflammation, and collaterally damage and kill sensitized epithelial cells (3). The additional loss of epithelial cells further decreases barrier integrity, and a self-perpetuating cycle occurs that leads to chronic inflammation. Over 3 million people in America and 5 million worldwide suffer from IBD. This widespread and debilitating illness results in decreased quality of life and significant health care–related costs (4). Although current treatments for IBD have improved (5, 6), these therapies remain modestly successful, with adverse side effects including immunosuppression, increased risk for infection, malignancies, and death (7). Thus, there is an unmet clinical need for safer and more effective IBD therapeutics for long-term care of Crohn disease (CD) and ulcerative colitis (UC) patients.
Although the involvement of the immune system in IBD is multifaceted, CD4+ T cells are among the most commonly cited cell type responsible for the initiation, progression, and chronicity of inflammation. In the healthy intestine, these subtypes are able to exist in homeostasis to allow inflammation, through Th1 and Th17 cells, that can be quickly controlled and resolved, through regulatory T cells (Tregs), once the inflammation-causing agent is removed. However, in IBD patients, this process is disrupted, allowing excessive and dysregulated production of inflammatory mediators, with IFN-γ and TNF-α paramount among these, and recruitment of neutrophils and monocytes to the intestinal lamina propria (8). Certain animal models of IBD clearly show a dysfunction within the Treg fraction of immune cells, but the role in patients may be more complex (9). Recent evidence suggests that CD4+ Th cells and FOXP3+ Tregs in IBD patients are more likely to coexpress cytokines and transcription factors associated with effector subtypes (10). This coexpression of inflammatory markers in Tregs suggests that these cells have an intermediate phenotype rather than a purely regulatory role, thus diminishing their overall anti-inflammatory effect. Targeting this mechanism to restore the suppressive function of Tregs may result in successful long-term therapeutics that provide sustained mucosal healing in IBD.
This strategy of increased generation and stability of Tregs is one of the proposed mechanisms of BT-11, a novel lanthionine synthetase C-like 2 (LANCL2)–targeting immunometabolic therapeutic (11). BT-11 is an orally active, gut restricted, small-molecule first-in-class therapeutic that is stable in the gastrointestinal (GI) tract and induces immunological changes when dosed orally. Through the LANCL2 pathway, we have previously elucidated that BT-11 modulates late-stage glycolysis in immune cells to bias cellular metabolism toward a profile of increased oxidative phosphorylation and decrease lactate production associated with regulatory function (11). BT-11 was discovered through innovative medical chemistry approaches from a family of 3 billion novel chemical entities (12). The decision to target the LANCL2 pathway stemmed from a thorough investigation into the role of LANCL2 in inflammation, metabolic disorders, and autoimmune diseases (13–16). The effects of LANCL2 activation have been linked to secondary messenger signaling (17), cellular metabolism (18, 19), cell survival (20, 21), and immune effects (13, 14). BT-11 has also been confirmed to modulate cytokine profiles of PBMCs and LPMCs from CD and UC patients, in line with the in vivo effects observed in animal models (11). In addition to efficacy, BT-11 has also been identified to possess a benign safety profile up to tested limit doses of 1000 mg/kg and unique pharmacokinetics with high levels of localization to the GI tract when dosed orally (22). With these characteristics, BT-11 is an ideal oral, gut-restricted, with limited systemic exposure, safe and effective therapeutic first-in-class drug candidate for the treatment of IBD.
In these studies, we aim to validate the critical importance of regulatory CD4+ T cells for the therapeutic efficacy of BT-11 in IBD. Based on consistently observed upregulation of Tregs in five mouse models of IBD, this subtype is clearly influenced by oral BT-11 treatment. Using in vitro differentiation studies, we examine the mechanisms of action of BT-11 directly on the signaling pathways of this cell subset. Combining these targeted cellular and molecular studies with in vivo validation, this study investigates for the first time, to our knowledge, the immunoregulatory mechanisms by which oral treatment with BT-11 provides protection from IBD by inducing and maintaining lasting immune tolerance in the GI tract through immunoregulatory mechanisms mediated by Tregs.
Materials and Methods
Mouse models of colitis
C57BL/6 wild-type (WT), IL-10−/−, and Mdr1a−/− mice were housed in standard 12-h on-off light cycle with ad libitum access to food and water. All procedures were approved by Institutional Animal Care and Use Committee. For the Citrobacter rodentium model of colitis, fresh C. rodentium (strain DBS100) was isolated in log-phase growth and suspended in a Luria-Bertani broth to provide an oral inoculum of 1 × 109 CFU/mouse. Mice were scored daily and feces was collected on day 7 to confirm successful challenge and validate that no differences in initial bacterial load were induced by BT-11 treatment. C. rodentium presence was tracked by PCR of the espB gene in feces and cecal contents (23). BT-11 treatment was initiated at day 1 postinfection. Mice were euthanized at d postinfection (dpi) 11 and 25. In PD1 neutralization experiments, mice were administered anti-PD1 Ab (50 μg/mouse, clone RMP1-14) or isotype control on days 1, 5, and 9 postinfection by injection. For the IL-10−/− model, mice began monitoring and treatment at 8 wk of age. Treatment was either continued daily throughout the 11-wk monitoring period or stopped after 7 wk of treatment. Mice were scored twice per week and were euthanized after 11 wk of treatment. In CD25 neutralization experiments, mice were administered anti-CD25 Ab (100 μg/mouse, clone PC-61.5.3) or isotype control after 7 wk of treatment. For the Mdr1a−/− model, mice began treatment and monitoring at 6 wk of age and continued through 10 wk of age, at which point euthanasia occurred. BT-11 (8 mg/kg) was delivered in a 0.5% methylcellulose suspension by oral gavage. Disease activity was assessed as a cumulative score (0–4) encompassing presence of diarrhea; rectal bleeding and inflammation; and hunchback position, activity, and general appearance. All experiments were conducted with gender- and litter-matched groupings.
Ex vivo culture
Mesenteric lymph nodes were excised, crushed, purified of RBCs by hypotonic lysis, and filtered in sterile conditions at day 25 postinfection in vehicle- and BT-11–treated mice challenged with C. rodentium. Soluble C. rodentium Ag was prepared from inoculum by formaldehyde inactivation followed by homogenization and centrifugation-based removal of insoluble Ag and remaining whole cells. Ag was diluted to a concentration of 100 ng/ml for stimulation. Cells were stimulated with Ag for 48 h in RPMI 1640 supplemented with FBS, HEPES, and penicillin/streptomycin at 37°C. No BT-11 was added to culture media during stimulation. After 48 h, cells were collected and processed for flow cytometry analysis.
In vitro Treg differentiation
Naive CD4+ T cells were isolated from the spleens of mice by magnetic sorting. Isolated cells were incubated in anti-CD3/anti-CD28–coated 96-well plates in IMDM media supplemented with FBS, HEPES, penicillin/streptomycin and l-glutamine. Differentiation media contained 10 nM all–trans-retinoic acid and 5 ng/ml TGF-β. Additional experiments were conducted comparing differentiation in standard Treg differentiation media with the addition of 10 ng/ml IL-2 or IL-12. Cells were incubated with vehicle, 10, or 100 nM BT-11 in differentiation media for 48 h prior to assay or coculture. Prior to assay, cells were stimulated with PMA and ionomycin for 6 h.
Treg and CD4+ T cell coculture
Tregs were differentiated in vitro as described above, and CD4+ T cells were isolated fresh from the spleen on the day of coculture start. Cell types were mixed in a 1:1 ratio with 200,000 cells per well. Tregs were thoroughly washed prior to coculture to remove any residual BT-11. Cells were incubated for 1 h, then stimulated with PMA and ionomycin for 6 h prior to collection of culture supernatant. Supernatant was assayed for concentrations of IFN-γ, TNF-α, and IL-10 by cytometric bead array. In Transwell experiments, CD4+ T cells were placed in the lower chamber of the well and Tregs were placed above the 0.4-μm–pore Transwell membrane. Concentrations of cytokines were validated to be equal above and below membrane. In neutralization experiments, anti-Lag3 (1 μg/ml, clone C9B7W) or anti-PD1 (1 μg/ml, clone RMP1-14) were added to each well at the same time as cells.
Colonic lamina propria lymphocytes and cultured cells were plated in 96-well plates (6 × 105 cells/well) and processed for immunophenotyping by flow cytometry as previously described. Briefly, cells were incubated with fluorochrome-conjugated Abs to extracellular markers: CD45, CD4, CD3, CD25, PD1, CD11b, CD11c, CD64, F4/80, CD103, Gr-1, MHCII, PDL2, CD19, CD80, and CD8. Samples needing a secondary staining were incubated with secondary Abs or streptavidin-conjugated fluorochrome. The samples were then fixed and permeabilized. Cells were incubated with Abs to intracellular markers: Tbet, IFN-γ, IL-10, FOXP3, IL-17, IL-21, RORγT, Helios, TNF-α. Data were acquired with a BD FACSCelesta flow cytometer and analyzed using FACSDiva software (BD Pharmingen).
Total RNA was isolated from mouse colon using a Qiagen RNA isolation mini kit. cDNA was generated from each sample using the iScript cDNA synthesis kit. Standards were produced through a PCR on the cDNA with Taq DNA polymerase from Invitrogen. The amplicon was purified using the Mini-Elute PCR purification kit from Qiagen. Expression levels were obtained through quantitative real-time PCR on a Bio-Rad CFX 96 Thermal Cycler using the Bio-Rad SYBR Green Supermix. For analysis, the starting amount of cDNA was compared with that of β-actin, as a control.
H&E-stained colonic sections were prepared from portions of colons collected into 10% buffered formalin and embedded in paraffin. Slides were examined by a board-certified veterinary pathologist via an Olympus microscope, and images were collected with Image-Pro software. Samples were scored (0–4) for leukocytic infiltration, epithelial erosion, and mucosal thickening in a blinded manner.
Treg protein extracts were isolated by cellular lysis in T-PER containing protease and phosphatase inhibitors. Extracts were quantified with the Pierce BCA protein assay to standardize protein concentrations. Standardized samples were run on 10% gels, transferred to nitrocellulose membranes, and incubated with p-STAT5a, p-FOXO1, or STAT5a Abs overnight at 4°C. Secondary Abs were anti-rabbit–HRP and anti-mouse HRP for 1 h at room temperature, upon which the membrane was visualized using ImageLab software. Band intensities were normalized to total protein content per lane or nonphosphorylated protein content.
Data are expressed as mean and SEM. Parametric data were analyzed using ANOVA, followed by the Scheffe multiple comparisons test. ANOVA was performed using the general linear model procedure of SAS (SAS Institute, Cary, NC). A 2 × 2 factorial arrangement comparing genotype and treatment was used. Statistical significance was determined at p < 0.05.
BT-11 promotes the downregulation of intestinal inflammation in C. rodentium colitis
Mice were infected with C. rodentium and treated daily with BT-11 (8 mg/kg, oral) or vehicle beginning on day 1 postinfection. Mice treated with BT-11 had significantly lower disease activity scores, beginning on day 5 postinfection and continuing through the peak of disease activity and into recovery (Fig. 1A). Colonic lamina propria leukocytes were isolated at the peak of disease activity, dpi 11, in a subset of mice. BT-11 treatment resulted in an increase of regulatory cells highlighted by CD25+ FOXP3+ IL-10+ regulatory CD4+ T cells (Fig. 1B) and CD103+ dendritic cells (Fig. 1C). Furthermore, at dpi 11, both Th1 and Th17 CD4+ subsets (Fig. 1D, 1E) as well as inflammatory myeloid cells, such as neutrophils (Fig. 1F) and F4/80hi TNF-producing macrophages (Fig. 1G), were suppressed by BT-11 treatment. BT-11 treatment reduced leukocytic infiltration upon histopathological examination of colon (Supplemental Fig. 1). However, BT-11 did not impair recognition of C. rodentium Ag or the generation of a memory response. After clearance of C. rodentium at dpi 25 (Supplemental Fig. 1), mesenteric lymph nodes were collected from vehicle and BT-11–treated mice and stimulated ex vivo with C. rodentium Ag. After stimulation, samples isolated from BT-11–treated mice resulted in an increase in PDL2+ CD80+ memory B cells (Fig. 1H). Additionally, CD4+ T cell responses were modulated with an increase in FOXP3+ IL-10+ cells (Fig. 1I) and a decrease in IFN-γ+ cells (Fig. 1J). No significant differences in cell types existed in the absence of stimulation. Thus, BT-11 treatment imparts lasting immunological tolerance in the C. rodentium model of colitis, in which immune effects induced by BT-11 are evident in cells after removal of BT-11 stimulation ex vivo.
BT-11 remains effective in the absence of IL-10
Although the C. rodentium model was suitable for testing the theory of immune tolerance due to the presence of a specific known Ag, a separate chronic model was needed to test if immune effects in vivo could be retained after discontinuation of BT-11 treatment. To validate the lasting immune tolerance induced by BT-11 treatment, we discontinued treatment in a chronic, spontaneous IL-10−/− model of colitis. BT-11 treatment was discontinued in a single group once the group average reached a score of 1, at week 7 of treatment. Minimal, nonsignificant separation occurred between groups receiving BT-11 treatment and those that had discontinued treatment for 3 wk after discontinuation (Fig. 2A). Moderate separation persisted through the end of the study between these two groups, although the discontinuation group still had suppressed disease activity relative to vehicle-treated controls. At necropsy, colons were collected for gene expression and flow cytometry analysis. BT-11 treatment resulted in lower expression of Tnf (Fig. 2B) and Ifng (Fig. 2C) in the colon, whereas discontinuation of BT-11 retained significant suppression of the two markers relative to the vehicle-treated group, albeit at a slightly lower magnitude. Cellularly, BT-11 resulted in increased proportions of PD1+ FOXP3+ (Fig. 2D) and CD25+ FOXP3+ Helios+ (Fig. 2E) CD4+ Tregs but did not increase CD25+ FOXP3+ Helios− CD4+ T cells (Fig. 2F) relative to vehicle-treated control. The discontinuation of BT-11 resulted in an intermediate level of PD1+ FOXP3+ cells and a significantly elevated CD25+ FOXP3+ Helios+ subtype. Similarly, both the BT-11 group and discontinuation group had lower Th1 (Fig. 2G) and IL-21+ (Fig. 2I) CD4+ subtypes, but only the BT-11 group had a significant decrease in F4/80hi TNF-producing macrophages (Fig. 2H). In addition to changes in cellular composition of lamina propria leukocytes, BT-11 also reduced histological lesions upon examination of colons in both BT-11 and discontinued BT-11 groups, highlighted by significant reductions in mucosal thickening and leukocytic infiltration, with minimal epithelial erosion observed across all groups (Fig. 2J–O).
Depletion of CD25+ cells reduces the retention of tolerance after discontinuing BT-11 treatment in IL-10−/− colitis
CD25+ Tregs were identified as a likely cell type contributing to the mechanisms of tolerance of BT-11 in vivo after observation of increases of CD25+ Tregs across multiple models of colitis. To confirm, we depleted CD25+ cells in vivo through administration of a CD25-neutralizing Ab at the time of discontinuation of treatment in the IL-10−/− model. Disease activity scores indicated that the administration of anti-CD25 quickly destroyed the retention of BT-11 efficacy postdiscontinuation but did not largely impact vehicle or groups that continued receiving BT-11 after anti-CD25 administration (Fig. 3A). As observed earlier, the discontinuation of BT-11 did not abrogate changes in the immune cell composition in the colonic lamina propria (Fig. 3B–E). In inflammatory cell types, Th1 (Fig. 3B) and neutrophils (Fig. 3E), the administration of anti-CD25 Abs diminished the retained effects of BT-11 in the discontinuation group, whereas the Th17 population (Fig. 3C) was numerically increased. Within the discontinuation group, BT-11 effects on CD25+ FOXP3+ cells (Fig. 3D) was not regained 4 wk after CD25 neutralization despite sufficient time to re-establish, as evidenced by significant increase within the BT-11 continuation group. Gating strategies are presented in Supplemental Fig. 2. Additionally, colonic expression of Tnf followed similar trends in all groups (Fig. 3F). Immune changes were supported by changes in histological scores (Fig. 3G–I). CD25 neutralization significantly increased leukocytic infiltration and mucosal thickening within the discontinuation group without inducing mean changes in the BT-11 continuation group. The effects of CD25 depletion on the retention of BT-11 efficacy was confirmed within the Mdr1a−/− model (Supplemental Fig. 3).
BT-11 synergizes with IL-2 to induce differentiation of CD25+ FOXP3+ regulatory CD4+ T cells
With the importance of CD25+ FOXP3+ regulatory CD4+ T cells to the efficacy of BT-11, we aimed to determine the direct effect of BT-11 on their differentiation and ability to retain phenotype in inflammatory conditions. Naive CD4+ T cells were differentiated into Tregs in vitro in the presence or absence of IL-2. BT-11 treatment (100 nM) significantly increased the establishment of a CD25+ FOXP3+ subtype in the absence of IL-2, a difference that was further accentuated by the addition of IL-2 (Fig. 4A). At concentrations as low as 10 nM, BT-11 induces significantly more CD25+ FOXP3+ cells in the presence of IL-2. Additionally, only low levels of a mixed CD25+ Tbet+ subtype were observed under these differentiation conditions, which were not statistically altered by BT-11 (Fig. 4B). Of note, a slight numerical increase occurred in vehicle-treated controls with the addition of IL-2 that was absent in the presence of BT-11. Meanwhile, in the opposite condition, the presence and absence of IL-12, BT-11 retains significantly higher levels of CD25+ FOXP3+ cells that are suppressed in vehicle-treated samples (Fig. 4C). The addition of IL-12 also induced an increase in CD25+ Tbet+ cells in all groups, although BT-11 provided a dose-dependent protection against this mixed subset (Fig. 4D). The direct differentiation of naive CD4+ T cells into effector subsets, Th1 and Th17, was significantly affected by BT-11 treatment only at higher concentrations, >100 nM (Supplemental Fig. 4).
BT-11 establishes changes in immune response through cell contact–mediated suppression by Tregs
Given the efficacy of BT-11 in an IL-10−/− model of colitis, we sought to determine the mechanism of immunoregulation through CD4+ Tregs. To validate that IL-10 was not a dominant driver of this immunoregulation, we used a coassay of in vitro differentiated Tregs with CD4+ T cells. Tregs were differentiated from WT and IL-10−/− naive CD4+ T cells in the presence of BT-11 (100 nM) or vehicle and washed, then cultured with matching WT CD4+ samples in the absence of BT-11. Although IL-10−/− Tregs had a lesser suppressive effect on IFN-γ (Fig. 5A) and TNF-α (Fig. 5B) production, in general, BT-11–treated Tregs still possessed greater suppressive capacity relative to vehicle-treated Tregs. WT Tregs treated with BT-11 did produce greater concentrations of IL-10 (Fig. 5C), but the BT-11–treated IL-10−/− Tregs did not induce greater IL-10 production from the WT CD4+ T cells. Based on these findings, we hypothesized that the greater suppressive effect of BT-11–treated Tregs was due to cell contact–mediated mechanisms. We conducted a similar coassay of CD4+ T cells with BT-11–treated Tregs using a Transwell system. In Transwells, BT-11–treated Tregs fail to suppress IFN-γ (Fig. 5D) and TNF-α (Fig. 5E) production yet retain increased expression of IL-10 (Fig. 5F). Furthermore, a coculture in standard wells using anti-Lag3 and anti-PD1 Abs was used. Anti-PD1 abrogated BT-11–treated Treg-induced suppression of IFN-γ (Fig. 5G) and TNF-α (Fig. 5H) production, whereas anti-Lag3 abrogated effects only in IFN-γ production. Neither Ab altered the production of IL-10 (Fig. 5I). Additionally, neither BT-11 directly or BT-11–treated Tregs induced greater apoptosis in CD4+ T cells (Supplemental Fig. 4), although previous data support a decrease in proliferation in CD4+ T cells cultured with BT-11–treated Tregs (11).
Loss of contact-mediated suppression abrogates efficacy of BT-11 in C. rodentium colitis
To confirm the role of contact-mediated suppression in vivo, we used the C. rodentium model of colitis in combination with administration of anti-PD1 Abs. Administration of anti-PD1 diminished differences in disease activity induced by BT-11 treatment (Fig. 6A). Anti-PD1 significantly changed the composition of colonic lamina propria immune cells in BT-11–treated mice. The significant decreases of Th1 (Fig. 6D), neutrophil (Fig. 6E), and Th17 (Fig. 6F) populations, as well as the increase in CD25+ FOXP3+ Helios+ (Fig. 6B) cells, were abrogated with the administration of anti-PD1. The proportion of CD25+ FOXP3+ Helios− CD4+ T cells was not significantly increased in the BT-11–treated group, although anti-PD1 did provide a moderate numerical decrease relative to those given control Ab (Fig. 6C). Histologically, only mice receiving BT-11 and isotype control Ab had observable differences in leukocytic infiltration, including into the submucosa, and mucosal thickening combined with mild reduction of epithelial erosion (Fig. 6G–M). The administration of anti-PD1 Ab abrogated histology differences. In addition to the IL-10−/− and C. rodentium models, BT-11 preferentially induces increases in CD25+ FOXP3+ Helios+ cells over Helios− counterparts within the regulatory CD4+ T cell subfamily in the Mdr1a−/− model of colitis (Supplemental Fig. 3).
STAT5 signaling mediates the BT-11 effects on CD25+ regulatory cell differentiation
To identify signaling pathways modulated by BT-11 in vivo, we isolated colonic CD4+ T cells from vehicle and BT-11–treated Mdr1a−/− mice at presentation of colitis, at 10 wk of age. The Mdr1a−/− model was used for the expression profile of CD4+ T cells, as it would be least likely to be influenced by model artifacts, such as a potential shift in profile to elicit pathogen clearance in the C. rodentium model or the lack of IL-10 causing compensatory shifts in expression in Tregs in the IL-10−/− model. Additionally, the Mdr1a−/− model may have greater potential to represent human IBD as opposed to the others due to translational implications of the gene in human disease. In CD4+ T cells, oral BT-11 treatment resulted in significantly higher expression of Stat5a (Fig. 7A) and Foxo1 (Fig. 7B), two members of the IL-2 signaling pathway in both the whole fraction of CD4+ T cells and Tregs. The relative expression across both BT-11– and vehicle-treated groups was increased within the Treg fraction, but similar magnitudes of difference were observed. Meanwhile, expression of Pten (Fig. 7C) was significantly increased in Tregs but not CD4+ T cells, and Phlpp1 (Fig. 7D) was slightly, but nonsignificantly, increased in either population. In vitro, STAT5a is phosphorylated in a greater ratio in BT-11–treated samples in control Treg differentiating media and in the presence of either IL-2 or IL-12 (Fig. 7E). FOXO1 is similarly affected in both control and IL-2–containing media but not in IL-12–containing media (Fig. 7F). Cells were also differentiated in the presence of inhibitors for PTEN (SF1670) or STAT5 (STAT5i). In both Treg media containing IL-2 (Fig. 7G, 7H) or IL-12 (Fig. 7I, 7J), the addition of STAT5i prevented the effects of BT-11 on CD25+ FOXP3+ and CD25+ Tbet+ cells. In contrast, SF1670 only prevented effects of BT-11 on CD25+ Tbet+ cells in IL-2–containing media (Fig. 7H).
BT-11 is a novel oral, gut-restricted, small-molecule therapeutic for IBD, with limited systemic exposure that targets the LANCL2 pathway. BT-11 is safe and effective (22, 24) with a no observed adverse effect level >1000 mg/kg in 3-mo good laboratory practice toxicology studies in rats and dogs and is more effective than current IBD therapeutics in nonclinical models (11). The U.S. Food and Drug Administration has approved BT-11 as an investigational new drug (IND) for CD (IND 128490) and UC (IND 138071), it has successfully completed phase 1 human clinical testing, and it will enter phase II clinical testing in UC and CD in 2019. Previously, the therapeutic efficacy of BT-11 has been identified to be dependent on its LANCL2-targeting ability through immunometabolic mechanisms (11, 12). In this article, we further defined the cellular and molecular mechanisms of BT-11 in the colonic mucosa. We provide molecular evidence that BT-11 enhances IL-2 signaling through an increase in STAT5a phosphorylation, allowing the stable differentiation of CD25+ FOXP3+ regulatory CD4+ T cells, a subtype important in the induction and maintenance of immunological tolerance in models of IBD.
In line with metabolic changes in late-stage glycolysis, highlighted by decreased lactate production and increased oxidative phosphorylation (11), BT-11 also supports regulatory CD4+ T cell differentiation and stability through unique interactions with the IL-2 pathway. In combination with IL-2, BT-11 enhances the phosphorylation of STAT5, one of the three main signaling axes downstream of IL-2 interactions (25). Together with FOXO1, this upregulation of STAT5 activity is a direct regulator of FOXP3 transcription (26, 27). The presence of both BT-11 and IL-2 amplify the actions of either alone. Although 10 nM BT-11 is unable to induce significant changes in CD25+ FOXP3+ cells alone, the addition of IL-2 provides a significant increase. This suggests that although higher concentrations are able to promote independent effects, the IL-2/STAT5 signaling axis could be critical to the in vivo therapeutic efficacy of BT-11 in IBD. These effects on STAT5 may also further define the metabolic signaling pathway of BT-11 in CD4+ T cells. Together with Akt, STAT5 has been shown to increase glucose uptake that may fuel the oxidative phosphorylation demands of Tregs (28). Importantly, this increase would be in agreement with our previously described downstream metabolic effects of BT-11 on continued Eno1 engagement in glycolysis, increased pyruvate dehydrogenase activity, and altered calcium signaling (11).
Although LANCL2 has previously been shown to impact Akt in liver cells (21), which is downstream of the PI3K pathway, a key element of the IL-2 signaling axes, the activation of LANCL2 by BT-11 does not stimulate an increase in Akt phosphorylation in the presence or absence of IL-2 in CD4+ T cells. Importantly, the regulation of Akt by LANCL2 is a result of direct interaction between mTORC2 and LANCL2. In contrast, the binding of LANCL2 to its natural ligand, abscisic acid, may promote the translocation of LANCL2 away from its inner cellular membrane–anchored position (29). Given the lack of Akt effect in CD4+ T cells and the translocation of LANCL2 when ligand bound, BT-11 binding may not increase mTORC2 activity due to a lack of direct interaction following ligand-induced relocation of the LANCL2 protein or simply a shift in conformation that prevents this direct interaction. With the increase in STAT5 phosphorylation, this may indicate that LANCL2 could function as a switch, acting either as a docking-based enhancer of phosphorylation for Akt or STAT5, depending on its binding by BT-11. Slight increases in both Phlpp1 and Pten expression may also contribute to attenuating any residue LANCL2-Akt interaction through direct dephosphorylation of Akt. Meanwhile, the downstream upregulation of FOXO1 phosphorylation supports the Akt-independent role of BT-11 in CD4+ T cells and is an additional mechanism for an increased expression of FOXP3.
The ability to induce a therapeutic benefit in IBD through activation of regulatory CD4+ T cells has been questioned because of the lack of phenotypic differences in the behavior of regulatory cells isolated from peripheral blood and the absence of a decrease of regulatory cells locally in the intestine when IBD patients are compared with non-IBD controls. In inflammatory conditions, such as those induced by the Th1-promoting cytokine, IL-12, FOXP3+ Tregs lose expression of CD25 on the path to adopting an intermediate or effector-biased phenotype, as a proposed means of losing high-affinity IL-2 family binding (30). These intermediate and weakly suppressive Tregs may contribute to the impaired ability to control inflammation in autoimmune disease. Indeed, recently emergent evidence supports that locally within the intestinal lamina propria, IBD patients present with greater amounts of effector-biased plasticity in FOXP3+ Tregs (31). Through BT-11 treatment, we show the ability to rescue a regulatory phenotype. When differentiated in the presence of IL-12, BT-11 treatment prevents the decrease of CD25+ FOXP3+ cells as well as the increase in CD25+ Tbet+ cells compared with vehicle treatment. When STAT5 is inhibited, BT-11’s ability to induce these changes is reduced. In combination with increased expression of stability-based markers (11), this STAT5 mechanism of BT-11 treatment may allow greater numbers of FOXP3+ Tregs to exhibit their intended regulatory functionality in the suppression of local inflammation in IBD.
In addition to acute changes in regulatory cell signaling and phenotype, BT-11 establishes long-lasting effects on the maintenance of immune tolerance and promotion of tolerogenic memory responses in the GI tract. In vivo, the activation of STAT5 signals, and more simply IL-2 expression, by BT-11 may be responsible for the enhanced development of thymic-like Tregs and the ability of Tregs to survive when they migrate to the periphery (32). Notably, BT-11 treatment preferentially increases Helios+ Tregs, which are closer in line with a natural Treg phenotype as opposed to an induced Treg state. The expression of Helios in Tregs has been shown to increase survival, fitness, and regulatory function (33). In vitro, the suppressive mechanisms stimulated by BT-11 are dependent on cell-to-cell contact as opposed to cytokine-mediated mechanisms. When contact is prevented by Transwell membrane, suppressive effects on IFN-γ and TNF-α production of untreated CD4+ T cells are lost. This hypothesis is also supported by the finding that the loss of IL-10 does not largely impact the increased suppressive function of BT-11–treated Tregs or the overall efficacy in the IL-10−/− model of colitis. Furthermore, the neutralization of PD-1 by Abs reduces the efficacy of BT-11 both in vitro and in vivo.
In these nonclinical models of IBD, the primary driver of BT-11 efficacy is consistently identified to be tied to the drug’s ability to increase both Treg populations and functionality. It has previously been reported that Tregs help to establish Th17 responses against C. rodentium in mice (34). Although there are potential parallels relating to local IL-2 engagement by elevated Tregs, our data support that BT-11 is capable of reducing inflammation without impairing the ability to attenuate this pathogenic bacterial population. In the absence of treatment with immunoregulatory drugs, the connectivity of Treg and Th17 responses may be beneficial in pathogen responses. The Tregs described in Wang, 2014, may be those of intermediate phenotype resulting from Treg/Th17 plasticity that are coregulated by FOXP3 and RORγT and may enhance Th17 function as a result. However, treatment with BT-11 favors the maintenance and the loss of regulatory function, and prevents the acquisition of intermediate inflammatory functions, which may help to explain the disconnect in Treg and Th17 responses in comparison with previously reported changes in C. rodentium–induced host responses. Mice that had received BT-11 treatment retain efficacy postdiscontinuation based on histological and immune measures. Depletion of CD25+ Tregs prevents this maintenance of therapeutic efficacy highlighted by increases in leukocytic infiltration, colonic Th1 and neutrophil levels, and colonic Tnf expression. Whereas this provides definitive proof for the importance of CD25+ Tregs on the maintenance of immune tolerance with oral BT-11 treatment, it does not preclude BT-11 from having direct effects on other aspects of intestinal homeostasis such as epithelial barrier integrity, myeloid cells, the recruitment of neutrophils, or Treg-independent suppression of effector CD4+ T cells. The ability of BT-11 to elicit these immunoregulatory mechanisms in addition to its effects on Tregs may positively impact its success in human IBD, in which the contributing factors are both numerous and diverse.
Among these additional mechanisms may be a role in shaping immune memory and germinal center interactions. Upon ex vivo Ag stimulation, MLN cells from BT-11–treated mice displayed a greater percentage of PDL2+ CD80+ B cells despite lesser inflammation in vivo during the course of disease. Previously, this cell type has been defined as producing little to no germinal center B cells but undergoing a fast isotype switch to generate secondary responses (35). Combined with previous data showing lesser lymphoid aggregates with BT-11 in vivo, this may indicate a yet unexplored arm of BT-11’s mechanism in adaptive immunity. The expression of PDL2 in germinal centers can decrease T cell activation and stem local inflammation (36). When treating IBD patients, with established inflammatory germinal centers, lymphoid aggregates, and tertiary lymphoid structures that are reacting to a wide variety of both pathogenic and commensal bacteria, this selective gut-restricted modulation by BT-11 may help to reset responses and restore tolerance to dietary, microbial, and self-antigens.
BT-11 restores and maintains immunological tolerance in the GI mucosa through activation of LANCL2 and support of FOXP3+ regulatory CD4+ T cells. The enhanced Treg differentiation and STAT5 phosphorylation with IL-2 stimulation provides direct signaling and transcriptional mechanisms for increased CD25+ FOXP3+ regulatory CD4+ T cells in the colonic lamina propria in models of IBD. The depletion of these cells or the inhibition of their contact-mediated suppression limit the retention of BT-11 efficacy following discontinuation of treatment. This manuscript further elucidates the regulatory arm of the novel dual mechanism of BT-11, with direct anti-inflammatory and proregulatory effects, which is absent in all current therapeutics for autoimmune diseases and the majority of those in development.
The online version of this article contains supplemental material.
Abbreviations used in this article:
inflammatory bowel disease
investigational new drug
lanthionine synthetase C-like 2
regulatory T cell
All authors are employees of Landos Biopharma. J.B.-R. is a shareholder of Landos Biopharma.