Patients diagnosed with CD (n = 51) by endoscopic, histological and radiological criteria were recruited for the study for blood or biopsy collection. Healthy subjects (n = 6) without any known underlying acute or chronic pathological condition served as control blood donors. Epithelial crypts were obtained from surgical resection specimens from non-IBD individuals (n = 5) undergoing surgery for colorectal cancer; a segment of healthy mucosa was collected at least 10 cm from the margin of the affected area. Supplementary Tables S1, S2 show the clinical and demographic characteristics from non-IBD subjects and CD patients. This study was carried out in accordance with the recommendations of ethics committees at the Hospital Clínic de Barcelona, Hospital Mutua de Terrassa and Hospital Universitari de Bellvitge-IDIBELL with written informed consent from all subjects. All subjects gave written informed consent in accordance with the Declaration of Helsinki. The protocol was approved by ethics committees at the Hospital Clínic de Barcelona, Hospital Mutua de Terrassa and Hospital Universitari de Bellvitge-IDIBELL.

The RORγt inhibitor BI119 (Boehringer Ingelheim Pharmaceuticals Inc., Ridgefield, CT, USA) was discovered by screening a small-molecule compound library. BI119 strongly bound to the human RORγ ligand-binding domain (LBD) and was active in an RORγ LBD reporter assay (Kd for RORγ LBD– 65 nM; IC50 for RORγ LBD reporter assay 260 nM). The compound showed high selectivity toward RORγt as demonstrated by a lack of significant activity against RORα (IC50 > 10 μM) and RORβ (IC50 > than 6 μM).

Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized peripheral blood by Ficoll (Sigma-Aldrich, Madrid, Spain) gradient centrifugation. Cells were cultured in X-VIVO 15 medium (Bio Whittaker, Lonza, Belgium) supplemented with 2% inactivated AB human serum (Sigma-Aldrich) for 7 days. PBMCs were cultured with the microbial commensal proteins FrvX (Prometheus Laboratories Inc., San Diego, CA, USA) and YidX (exonBio, San Diego, CA, USA) at 2 μg/ml. An unstimulated condition was used as negative control. Heat-killed Candida albicans (C. albicans) was kindly provided by the Microbiology Department, Hospital Clínic-IDIBAPS, Barcelona, Spain and was used at 1 colony-forming unit (CFU): 1 PBMC as a positive control for IL-17 producing T cells. Recombinant interleukin (IL)-2 (20 UI/ml) (R&D systems, Minneapolis, MN, USA) was added to the culture on day 3. For RORγt blocking experiments, PBMCs were cultured in the presence of BI119 at 1 μM or dimethyl sulfoxide (DMSO) (vehicle control, 1:10,000). On the seventh day, supernatants were centrifuged and stored at −20°C until assayed. PBMCs were washed with cold PBS, re-suspended in 600 μL of buffer RLT (Qiagen, Hilden, Germany) and stored at −80°C until RNA extraction.

Non-IBD intestinal epithelial crypts were isolated from intestinal tissue as previously described (21). For short-term crypt culture, 40 isolated crypts/25 μl Matrigel (BD Biosciences) were plated and cultured in either complete crypt culture medium or in medium containing supernatants from activated sorted antigen-specific CD4⁺ T cells (more detailed information is provided in Supplementary Materials and Methods). Antigen-specific T-cell supernatants were obtained from re-stimulating sorted cells with corresponding antigen (FrvX or YidX) treated with or without BI119. After overnight culture of crypts, RNA was extracted.

Intestinal biopsies (4–6 per patient) were obtained from inflamed areas (defined by the presence of ulcers) of the colon or ileum from CD patients. Biopsies were washed twice in RPMI 1640 medium (Lonza, MD, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Biosera, France), 100 U/ml penicillin, 100 U/ml streptomycin and 250 ng/ml amphotericin B (Lonza), 10 μg/ml gentamicin sulfate (Lonza) and 1.5 mM Hepes (Lonza). Whole biopsies were divided in two wells and cultured in the presence of BI119 at 1 μM or vehicle control (DMSO, 1:10,000) at 37°C in humidified atmosphere containing 5% CO₂ incubator for 18 h. Total RNA was isolated and transcriptional analysis was performed.

Total RNA was isolated using RNeasy mini kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Purity and integrity of the total RNA were assessed using the 2100 Bioanalyzer (Agilent, Germany) and quantified with a NanoDrop spectrophotometer (Nanodrop Technologies, DE, USA). Only samples with an RNA integrity number (RIN) greater than 7.0 were used.

Total RNA (250 ng) was transcribed to complementary DNA using a reverse transcriptase (High Capacity cDNA Archive RT kit, Thermo Fisher Scientific, Waltham, MA, USA). Quantitative real-time PCR (qPCR) was performed in an ABI PRISM 7500 Fast RT-PCR System (Applied Biosystems) using predesigned TaqMan Assays (Applied Biosystems). ACTB was used as a reference gene and arbitrary units (AU) were calculated relative to ACTB.

This study was carried out in accordance with the recommendations of Boehringer Ingelheim's Institutional Animal Care and Use Committee.

CB6F1 female mice served as cell donors and female CB.17 severe-combined immunodeficient (SCID) mice as recipients (Jackson Labs, Sacramento, CA). Mice were allowed a minimum of 2 weeks for acclimatization in specific pathogen-free conditions with 12-h light/dark cycle. All mice were used at 8–10 weeks of age with access to food and water provided ad libitum. CB6F1 mice were humanely sacrificed and spleens collected on ice. Following homogenization of spleens and red blood cells lysis, CD4⁺ cells were enriched from pooled spleen cells using a commercially available negative selection kit (Stemcell Technologies, Vancouver, BC) following the manufacturer's protocol. CD4⁺ enriched cells were stained with Alexa Flour 488 conjugated anti-CD4 clone RM4-5, Alexa Flour 647 conjugated anti-CD45RB clone C363-16A and PE conjugated anti-CD25 clone PC61 (Thermo Fisher Scientific). CD4⁺CD45RB^high cells were sorted on a FACS Aria II (BD Biocsience). CD4⁺CD45RB^low cells were collected separately. Groups of recipient CB.17 SCID mice were injected intraperitoneally with (5 × 10⁵) purified CD4⁺CD45RB^high donor lymphocytes in 200 μL PBS. A separate group of recipients was injected with the same quantity of non-colitogenic CD4⁺CD45RB^low lymphocytes as a control. A group of CB.17 SCID normal mice was also used as a control. All mice were weighed and observed weekly for clinical signs of illness including piloerection, hunched posture, decreased skin turgor and eye crusting. Diseased animals were humanely sacrificed for analysis at 4 weeks post-transfer or when body weight loss exceeded 20% of starting weight.

BI119 was suspended in MC/tween solution (0.5% methylcellulose, 0.015% polysorbate 80) using a dounce homogenizer until a uniform suspension was formed. Recipient mice that received CD4⁺CD45RB^high cells were administered either MC/tween or 100 mg/kg BI119 by oral gavage immediately prior to cell transfer and then twice daily until study termination.

Statistical analysis was performed using Bioconductor tools in R (V. 3.4.2). In experiments performed with human samples, differences between continuous variables were tested with nonparametric test (unpaired or paired Mann-Whitney-Wilcoxon test). Error bars show the mean and standard error of the mean (SEM). P-values were adjusted by false discovery rate (fdr) and were considered statistically significant when equal or less than 0.05. Regarding mice experiments, significance testing was performed by one-way ANOVA and Sidak's multiple comparisons test. For the total sum histopathology score, the Kruskal Wallis test and Dunn's multiple comparison test was performed.

In order to evaluate the efficacy of a small molecule antagonist for RORγt, we stimulated human PBMCs to secrete measurable amounts of IL-17 in response to microbial antigens. Here we used the fungi C.albicans as a positive control because it is a strong inducer of Th17 responses. In addition, we used two different Escherichia coli-derived proteins: FrvX, which was selected based on reported increased sera reactivity in CD (22), and YidX, which we previously showed as capable of inducing exacerbated Th17 responses in CD patients (4). Supplementary Figure 1 shows production of IL-17, IFNγ, and IL-5 by stimulated PBMCs from both non-IBD donors (n = 6) and CD patients (n = 6) (Supplementary Table 1, Patient group 1). YidX and FrvX induced the production of significantly higher concentrations of IL-17, but not Th1 and Th2-related cytokines, in CD patients compared to non-IBD controls. In contrast, as expected, C. albicans induced similar increases in IL-17 levels in controls and CD patients compared to unstimulated conditions.

Using this system, we next tested the effect of a specific RORγt inhibitor (BI119) on the expression of a number of gene transcripts and proteins. PBMCs from an independent group of CD patients (n = 12; Supplementary Table 1, Patient group 2) were cultured alone or stimulated with C.albicans, FrvX, and YidX in the presence of BI119 (1 μM) or vehicle control. BI119 significantly reduced transcription of the Th17-related genes RORC, IL17A, IL17F, IL22, IL26, and IL23R under all conditions studied, while it did not alter RORA expression (Figure 1A and Supplementary Figure 2A). IL-17A (Figure 1A) and IL-22 (Supplementary Figure 2B) protein concentrations in culture supernatants from stimulated PBMCs were also significantly reduced by BI119 compared to vehicle-treated conditions. Remarkably, the Th1 or Th2 master regulators (TBX21 and GATA3 respectively), or their main effector cytokines (IFN-γ and IL-5), were not regulated by BI119 under any of the conditions studied (Figures 1B,C). In contrast, BI119 induced a small but significant (paired analysis, adjusted p < 0.05) increase in both FOXP3 transcription and IL-10 production in microbial-stimulated PBMCs from CD patients (Figure 1D). These data demonstrate that BI119 strongly and specifically blocks Th17-related genes induced by microbial stimulation of CD PBMCs, though it does not interfere with Th1 and Th2 responses and up-regulates a Treg expression profile.

It has been described that IL-17 acts on epithelial cells by enhancing the expression of neutrophil and Th17 recruiting chemokines such as CXCL1, CXCL8 and CCL20, while it represses the Th1-attracting cytokine CXCL10 (4). Here we tested whether treating microbial activated CD4⁺ T cells with BI119 could reduce their inflammatory effect on the epithelial layer. We focused only on FrvX and YidX because their relevant responses in CD patients (4).

In order to do this, CFSE-labeled PBMCs from CD patients (Supplementary Table 1, Patient group 3) were stimulated with FrvX (n = 8) or Yidx (n = 7). Responding CFSE⁻CD4⁺ T cells were sorted and re-stimulated with their cognate antigen in the presence of BI119 or vehicle control (Figure 2A). After 7 days supernatants from these cultures were collected and added to whole colonic crypts from non-IBD surgical specimens (n = 5). Supernatants from both FrvX and YidX-specific CD4⁺ T cells induced a marked up-regulation of CXCL1, CXCL8, CCL20, and CXCL10 transcription on intestinal crypts (Figures 2B–E). Remarkably, supernatants from FrvX and YidX-specific CD4⁺ T cells treated with BI119 showed a significantly lower induction of CXCL1, CXCL8, and CCL20 by intestinal crypts (Figures 2B–D). In contrast, CXCL10 expression was either increased or unchanged under the same conditions (Figure 2E). Our results collectively indicate that the specific inhibition of RORγt on T cells modulates their pro-inflammatory effects on the epithelial lining.

Thus far we have shown that BI119 effectively inhibits RORγt-dependent gene transcription and protein secretion on circulating PBMCs responding to microbial antigen stimulation, specifically on CD4⁺ T cells. Expression of RORγt, however, is not limited to these lymphocyte populations. Indeed, intestinal mucosa cell types poorly represented in peripheral blood (i.e., intraepithelial lymphocytes and ILC3 cells) have been shown to express the Th17 transcriptional regulator.

Therefore, we investigated the effect of RORγt inhibition in intestinal tissue explants (endoscopic biopsies) obtained from the inflamed involved mucosa of CD patients (n = 18; Supplementary Table 2, Patient group 4) to assess the effect of BI119 on direct as well as indirect targets in tissue. Based on the differential cellular composition and transcriptional profiles between colonic and ileal mucosa, we analyzed separately the effects of RORγt inhibition on both intestinal locations. Samples were taken from inflamed colonic (n = 10) and ileal (n = 8) CD segments with comparable endoscopic disease severity based on their segmental CD endoscopic index of severity (CDEIS) score (Supplementary Table 2). Despite comparable disease severity, we found higher basal expression of IL17A and S100A8 in ileal samples compared to colonic ones (Supplementary Figure 3). In contrast, expression of other genes regulated in inflammation as CXCL1, CXCL8, IFNG, IL6 and IL10 was comparable between colonic and ileal CD samples, suggesting that the differences in IL17A and S100A8 expression may potentially reflect differences in disease location rather than degree of inflammation. The tissue specific marker DEFA5 was, as expected, also significantly over-expressed in ileal compared to colonic CD.

Despite these variances in basal expression, treatment with BI119 significantly reduced transcription of IL17A, IL17F, and IL26 in both the colon (Figure 3A) and ileum (Figure 3B) of CD patients with active inflammation. Remarkably IL22 was not affected in either location, in striking difference to our results in PBMCs (Supplementary Figure 2). Other genes that are significantly regulated in CD mucosa compared to non-IBD controls (data not show) such as IL23R, CSF2, CXCL1, CXCL8, IFNG, S100A8, IL6, IL10, and DEFA5 were also significantly changed by BI119 in ileal CD, whereas the effect on colonic biopsies did not reach statistical significance in most cases. These results suggest that BI119 can modulate some of the Th17-related genes, as well as inflammation-related genes, while preserving IL22 expression in inflamed samples from active CD. Moreover, they indicate that ileal, compared to colonic, Th17 responses may be more amenable to modulation by RORγt antagonists.

In contrast to the experiments shown in Figure 1 using microbial antigen-expanded CD4⁺ T cells from PBMCs, BI119 did not significantly regulate the expression of RORC in tissue explants. Noticeably, the RORC gene can encode for 4 different isoforms, 2 of which give rise to the proteins RORγ and RORγt. The fact that RORγ can be expressed by most cells in the mucosa, including epithelial cells (data not shown), may explain the differences in RORC modulation between PBMCs and tissue explants. As the TaqMan assay used here does not differentiate between the two isoforms, we cannot accurately measure changes specifically related to RORγt using whole biopsy tissue.

Finally, the efficacy of BI119 in vivo was assessed in a murine colitis model induced by transfer of CD4⁺CD45RB^high T cells into T- and B-cell-deficient CB.17 SCID mice. Mice that received CD4⁺CD45RB^high cells were orally administered either MC/tween (vehicle control) (n = 11) or 100 mg/kg BI119 (n = 12) twice a day for 28 days. Mice that were injected with a non-colitogenic CD4⁺CD45RB^low population (n = 9) and untreated CB.17 SCID mice (n = 9) were used as controls. Transfer of CD4⁺CD45RB^high T cells into immune-depleted mice induced a significant decrease in body weight, and a significant increase in macroscopic (colon weight-to-length ratio), histologic, fecal (lipocalin), and circulating (plasmatic sCD14) markers of inflammation at day 28 compared to both control groups (Figures 4A–E). Oral administration of BI119 significantly reduced both fecal lipocalin and circulating sCD14 compared to vehicle treatment (Figure 4B). A significant decrease in mucosal thickness due to a decrease in crypt hyperplasia was also observed in mice receiving BI119 (Figure 4C). Furthermore, administration of BI119 significantly reversed clinical and macroscopic signs of inflammation, while histologic improvement did not reach statistical significance (p = 0.18 vs. vehicle group) despite a clear amelioration on epithelial changes, mucosal inflammation and gland loss in compound-treated mice (Figures 4D,E and Supplementary Table 3).

We next looked at the molecular mechanisms involved in response to BI119 in this animal model by measuring the transcription of key inflammatory genes, including RORγt-dependent targets in the intestinal mucosa. As expected, colitis induced by CD4⁺CD45RB^high T cells in immune-depleted mice was associated with a significant increase in Th17 genes, IFNG, IL10 as well as S100A8 (Figure 5) and S100A9 (data not shown). In agreement with the clinical, macroscopic and histologic protective effect of RORγt inhibition, oral BI119 treatment significantly reduced the expression of Th17-related genes such as IL17A, IL17F, and IL22, as well as IFNG and calprotectin (Figure 5). Remarkably, BI119 despite controlling all the other inflammation-related genes, did not significantly reduce transcription of IL10 in the colon. Besides the influence on IL10, these in vivo experiments show a clear effect of oral BI119 administration in preventing T-cell driven murine experimental colitis.

ErR, ML, JW, MP, JZ, CH, RH, MT, DS, and GN are Boehringer Ingelheim Pharmaceuticals Inc. employees. AS has received consultancy fees and grant money from Boehringer Ingelheim Pharmaceuticals Inc. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.