All studies were conducted in accordance with the University of Texas Health Science Center at San Antonio Animal Care and Use guidelines and biosafety guidelines and approved by Institutional Animal Care and Use Committee (protocol #20170020AR) and Biosafety Committee. Six to twelve-week-old male and female mice were used for experiments. C57BL/6 (wild-type, WT) mice were originally obtained from The Jackson Laboratory (Bar Harbor) and bred at the University of Texas Health Science Center at San Antonio. IL-23^-/- and IL-22^-/- mice were described previously (27, 28). All mice used in this study were on C57BL/6 background and maintained under specific pathogen free conditions. Age/sex matched controls or littermate controls were used for all experiments.

To block IL-10Rα in vivo, mice were injected i.p. with 350 μg of clone 1B1.3A (BioXCell, West Lebanon, NH) 12–16 h prior to infection and again 1, 4, and 7 days after infection. Isotype mIgG1clone HRPN (BioXCell, West Lebanon, NH) or PBS (Phosphate-buffered saline) was used as control.

Campylobacter jejuni NCTC 11168 inocula were streaked onto Mueller Hinton (MH) agar and grown at 37 °C under microaerophilic conditions using anaerojars and Oxoid CampyGen sackets (Thermo Scientific) for 48 h. Several colonies were picked from the initial growth and subcultured on MH agar for 24–28 h. Mice were pre-treated with an antibiotic cocktail for seven days in the drinking water as previously described (29). One day after the removal of antibiotics, mice were inoculated with a single dose of 1–5 × 10⁹ CFU of C. jejuni or medium by oral gavage. Mice were weighed daily following infection and euthanized if more than 20% of initial body weight was lost.

Ten days after infection, mice were euthanized using CO₂ inhalation and death was ensured by performing cervical dislocation. The severity of colitis was assessed by hematoxylin and eosin (H&E) staining and colonic mass-to-length ratio. H&E stained sections were evaluated for characteristics of inflammation and images were taken with a Zeiss Axiophot2 microscope (Thornwood, NY). For histopathologic scoring, crypt architecture, mucosal thickening, cellular infiltrates, hemorrhage, goblet cell depletion and tissue damage were evaluated ( Supplementary Table 1 ).

10 μg of IL-23 p19-p40-expressing plasmid (p-IL-23) (30) or control plasmid (pRK) in 1.7 ml of TransIT-EE Hydrodynamic Delivery Solution (Mirus Bio) were intravenously injected in the tail vein on the second day after C. jejuni inoculation.

The distal half of the colon tissue was cut longitudinally and washed three times in sterile PBS. The tissue was weighed, homogenized in MH media, serially diluted and plated on MH agar (Muller-Hinton agar II Agar, BD). Plates were incubated at 42 °C for 48 h before counting.

Lamina propria leukocytes were isolated as previously described (31). Briefly, the cecum and colon were cut open and rinsed twice in PBS to remove fecal material. Colon pieces were incubated in complete medium (RPMI-40 supplemented with 3% FBS, 1 mM penicillin-streptomycin) containing 2 mM EDTA for 20 min at 37 °C with slow rotation (100 r.p.m.) to remove epithelial cells. Remaining tissue pieces were digested in serum-free RPMI-40 containing 200 μg/ml Liberase TM (Roche) and 0.05% DNase I (Sigma) for 40 min at 37 °C with shaking at 100 r.p.m. The digested tissue was passed through a mesh screen, washed with PBS containing 3% FBS and separated by 80/40% Percoll gradient (GE Healthcare).

The cells were preincubated for 20 min with anti-CD16/32 (2.4G2, Fc block from BioXCell) and stained for viability with LIVE/DEAD^® (Invitrogen, Carlsbad, CA) for 20 min on ice. Cells were surface stained for flow cytometry with the combination of the following antibodies (eBioscience or Biolegend): Anti-CD45 (104), anti-CD3 (145-2C11), anti-Thy1.2 (30-H12), anti-B220 (RA3-6B2), anti-Ter-119 (TER119), anti-Gr-1 (RB6-8C5), anti-CD11c (N418), anti-CD5 (53-7.3), anti-Eomes (Dan11mag), anti-T-bet (4B10), anti-RORγt (Q31-378, BD Pharmingen), anti-IFNγ (XMG1.2), anti-IL-17a (TC11-18H10.1), anti-Ly6G (1A8), anti-CD11b (M1/70), anti-CD64 (X54-5/7.1), anti-CD103 (2E7), anti-MHCII (M5/114.15.2). For intracellular cytokine staining was cells (1 x 10⁶ cells) were restimulated for 4 hours at 37°C with PMA (50 ng/ml) and ionomycin (500 nM) in the presence of brefeldin A (5 μg/ml) (all from Sigma-Aldrich, St. Louis MO) and fixed and permeabilized using fixation and permeabilization solution (eBiosciences). All cells were analyzed on either a FACS Celesta or LSR II (BD Biosciences) and analyzed with FlowJo software (FlowJo LLC).

RNA from colon was isolated using E.Z.N.A. Total RNA kit I (Omega Bio-tek). cDNA synthesis and real-time RT-PCR was performed as described previously (32) using Power SYBR Green master mix (Applied Biosystems). Relative mRNA expression of the target gene was determined using the comparative 2^−ΔΔCt method. Primers used are listed in Supplementary Table 2 .

Statistical analysis was performed using the Mann & Whitney test or two-way analysis of variance (ANOVA) when appropriate. Data are presented as means ± S.E.M. with p < 0.05 being considered significant. All statistical tests were performed in GraphPad Prism 8.0 program. ns, not significant; *p < 0.05, **p < 0.01, ***p < 0.001.

IL-23 is known to participate in both protective and pathogenic responses in the gut (17, 18, 33). To define the role of IL-23 in pathogenesis of C. jejuni-induced intestinal inflammation, we used a previously described animal model of Campylobacteriosis using mice with IL-10R blockade which mimics pathophysiological features of human disease (2, 34). To facilitate C. jejuni colonization, IL-23^-/- and WT control mice were pre-treated with broad spectrum antibiotics, as described (34) and orally infected with C. jejuni ( Figure 1A ). C. jejuni infection induced severe intestinal inflammation in aIL-10Rα antibody-treated WT mice, as demonstrated by significant weight loss ( Figure 1B ), colon thickening and increased colonic mass-to-length ratio ( Figure 1C ). Histological analysis revealed inflammatory lesions in both cecum and colon of infected WT mice ( Figures 1E, F ). In stark contrast, IL-23^-/- mice did not show significant body weight loss and colon thickening ( Figures 1B, C ). Although the IL-23-deficiency did not affect C. jejuni colonization of the intestine ( Figure 1D ), the histological evaluation of colons revealed reduced intestinal inflammation with minor immune cell infiltration, epithelial damage and crypt hyperplasia compared to WT mice ( Figures 1E, F ). Accordingly, although WT and IL-23^-/- mice without aIL-10Rα treatment displayed similar C. jejuni colonization ( Supplementary Figures 1B, C ), intestinal pathology was reduced in IL-23 deficient mice ( Supplementray Figure 1A ). Together, these observations indicate that IL-23 drives pathology in C. jejuni infected mice.

Campylobacter jejuni-induced intestinal inflammation is associated with infiltration of myeloid cells such as neutrophils, monocytes, macrophages and dendritic cells (DCs) to the large intestine (29, 35). Given that C. jejuni infected IL-23^-/- mice did not develop significant intestinal pathology, we next asked if the accumulation of myeloid populations were impaired in these mice. Evaluation of colonic myeloid cell populations demonstrated fewer frequency of neutrophils, eosinophils (identified by a negative gating, Supplementary Figure 2A ), macrophages and monocytes among CD45⁺ immune cells compared to WT mice on day 10 after infection ( Figures 2A, B, D–G, I, J ). By contrast, there was no significant difference in DCs and B cells between IL-23^-/- and WT mice ( Figures 2C, H ). These results demonstrate that IL-23-driven response is essential for the accumulation of neutrophils, eosinophils, macrophages and monocytes in the large intestine during C. jejuni induced colitis.

Innate immune cells such as neutrophils, macrophages and dendritic cells upon activation can produce multiple cytokines that can activate and facilitate recruitment of circulating lymphocytes and inflammatory myeloid cells to the sites of inflammation. Previously, it was shown that expression of proinflammatory cytokines such as TNF, IL-1β, IL-12, and IL-6 was increased in the colon of C. jejuni-infected IL-10^-/- mice (29, 35, 36). To further examine the role of IL-23 in C. jejuni induced inflammatory response, we assessed the expression of these cytokines in the colons of infected mice at day 10 post infection. As shown in Figure 3 , the mRNA levels of IL-12p35 and IL-12p40 subunits were unchanged in IL-23^-/- mice compared to WT mice, indicating that IL-23 deficiency is not associated with impaired expression of IL-12. In contrast, IL-6, IL-1β, IL-23R, and TNF were downregulated in IL-23^-/- mice ( Figures 3A–C, F ). As IL-23 is one of the major cytokines regulating IL-22 expression in inflammatory conditions (30, 37), we next measured IL-22 levels. Indeed, IL-22 expression was significantly reduced in IL-23^-/- mice after C. jejuni infection ( Figure 3G ).

These data indicate that IL-23 regulates accumulation of myeloid cells in the gut during C. jejuni-induced colitis. Moreover, IL-23 deficiency diminishes expression of proinflammatory cytokines, most likely due to reduced infiltration of myeloid cells to the intestine.

Numeral studies have demonstrated that IL-22 can regulate intestinal inflammation by promoting epithelial cell repair and production of anti-microbial peptides (38). IL-22 can also promote intestinal pathology in several models of intestinal inflammation driven by CD40 (19), Toxoplasma gondii (39) and in a model of spontaneous colitis caused by IL-10 deficiency (40). IL-23 promotes IL-22 production mainly by T cells and ILCs during mucosal bacterial infection (41, 42). We found that colonic mRNA expression of IL-22 was significantly increased in WT mice after C. jejuni infection ( Figure 3G ), consistent with previous studies (35, 43). To further examine the role of IL-22 in C. jejuni-induced intestinal pathology, we infected aIL-10Rα-blocked IL-22^-/- mice according to the same protocol ( Figure 1A ) and assessed the development of disease. Surprisingly, there was no significant difference in body weight change, colonic mass-to-length ratio and bacterial burdens in IL-22^-/- mice compared to WT mice ( Figures 4A–C ). Similarly, no significant difference in histopathological score was observed between infected WT and IL-22^-/- mice ( Figures 4D, E ). Together these results indicate that IL-22 is not essential for C. jejuni-induced intestinal pathology in this model.

C. jejuni-induced colitis in IL-10^-/- mice is characterized by mixed Th1 and Th17 responses (35, 43). Recently it has been shown that both IFNγ and IL-17, signature cytokines of Th1 and Th17 responses, respectively, contribute to C. jejuni-induced pathology (35, 43). Depletion of either IFNγ or IL-17 or both during C. jejuni infection significantly decreased inflammation and pathology (35, 44). We showed that colonic IL-17 expression was decreased in IL-23^-/- mice compared to WT mice ( Figure 5A ). IL-23 can drive IL-17 production by both CD4⁺ T cells and innate lymphoid cells (ILCs) (45).

To determine which of these cells contribute to IL-23-dependent IL-17 production in C. jejuni infection, we isolated cells from colonic lamina propria of IL-23^-/- mice. Cells were stimulated with PMA and ionomycin in the presence of Brefeldin A and IL-17 production by T cells and ILCs was analyzed by flow cytometry ( Supplementary Figures 2B, C ). Our analysis revealed that the total number and frequency of IL-17 producing ILCs were reduced in IL-23^-/- mice ( Figures 5B, C ), however only total number of IL-17⁺ T cells was significantly reduced ( Figure 5D ). These data suggest that both ILCs and T cells contribute to IL-23 dependent IL-17-production during C. jejuni infection.

The major producers of IFNγ during C. jejuni infection are T cells, ILCs and NK cells (35, 44). Our data suggest that IL-23^-/- mice display reduced IFNγ expression compared to WT mice ( Figure 6A ). We next asked if the production of IFNγ by these cells was affected by IL-23 deficiency during C. jejuni infection. We found a marked reduction of IFNγ-producing ILCs but not NK or T cells in the colon of IL-23^-/- mice on day 10 after C. jejuni infection ( Figures 6B, C ). aIL-10Rα antibody treatment in uninfected mice did not affect the number of IFNγ-producing ILCs and T cells in the colon ( Supplementary Figures 3A, B ), in agreement with published report (44). In contrast to colon, we did not find difference in IFNγ production by ILCs or T cells in MLN between WT and IL-23^-/- infected mice ( Supplementary Figures 3C–F ). These data suggest that IL-23 contributes to IFNγ production in the colon during C. jejuni infection. To further define if IL-23 regulates IFNγ-production in another gut bacterial infection model, we infected IL-23^-/- mice with Citrobacter rodentium and analyzed IFNγ-expression on day 5 p.i. We did not detect any difference in IFNγ-expression between IL-23^-/- and control mice (data not shown), in agreement with protective role of IL-23 in this infection (25, 46).

Previous studies demonstrated that IFNγ can be produced by group 1 ILCs characterized by expression of signature transcription factor T-bet and by group 3 ILCs characterized by RORγt expression (47, 48). As total cell number of IFNγ⁺ ILCs was selectively reduced in IL-23 deficient mice, we next asked which subtypes of ILCs were affected. As shown in Figures 6D–F , both frequency and total cell number of IFNγ-producing ILC1s and ILC3s were significantly reduced in IL-23^-/- mice after C. jejuni infection, demonstrating that IL-23 is important to promote IFNγ-production by these ILCs during C. jejuni-induced colitis.

To further define if IL-23 contributes to C. jejuni-induced intestinal pathology, we used hydrodynamic injection of a plasmid encoding both p19 and p40 IL-23 subunits (p-IL-23) (30). We found that p-IL-23 treated mice lost significantly more weight and displayed increased levels of IFNγ, IL-17 and IL-22 in the colon ( Supplementary Figure 4 ). These data further support the pathogenic role of IL-23 in C. jejuni-induced intestinal pathology.

Collectively, these data indicate that IL-23 signaling contributes to C. jejuni-induced intestinal inflammation by driving IL-17 and IFNγ responses by T cells and ILCs.