T cell-intrinsic role for Nod2 in protection against Th17-mediated uveitis

Abstract

Mutations in nucleotide-binding oligomerization domain-containing protein 2 (NOD2) cause Blau syndrome, an inflammatory disorder characterized by uveitis. The antimicrobial functions of Nod2 are well-established, yet the cellular mechanisms by which dysregulated Nod2 causes uveitis remain unknown. Here, we report a non-conventional, T cell-intrinsic function for Nod2 in suppression of Th17 immunity and experimental uveitis. Reconstitution of lymphopenic hosts with Nod2^-/- CD4⁺ T cells or retina-specific autoreactive CD4⁺ T cells lacking Nod2 reveals a T cell-autonomous, Rip2-independent mechanism for Nod2 in uveitis. In naive animals, Nod2 operates downstream of TCR ligation to suppress activation of memory CD4⁺ T cells that associate with an autoreactive-like profile involving IL-17 and Ccr7. Interestingly, CD4⁺ T cells from two Blau syndrome patients show elevated IL-17 and increased CCR7. Our data define Nod2 as a T cell-intrinsic rheostat of Th17 immunity, and open new avenues for T cell-based therapies for Nod2-associated disorders such as Blau syndrome.

Fig. 1. Nod2 plays an unexpected protective role in uveitis.

Six to ten week old WT and Nod2^−/− mice were immunized and assessed weekly thereafter for disease. a Representative fundus images of the posterior pole at day 21 post-immunization, and (b) clinical grading of uveitis based on fundoscopy. Data in b is combined from 6 independent experiments with an n = 12 mice/group for day 7 and day 17, n = 24 (WT) and 19 (Nod2^−/−) mice/group for day 10, n = 44 (WT) and 28 (Nod2^−/−) mice/group for day 14, and n = 56 (WT) and 54 (Nod2^−/−) day 21; *p = 0.0177 (14 day), *p = 0.0093 (day 17), and ***p < 0.0001 (day 21). c Representative epifluorescence images and (d) quantification of blood-retinal barrier integrity 10 days post-immunization. Data is combined from 2 independent experiments where n = 9 (WT adjuvant and Nod2^−/− IRBP) or 8 (WT IRBP and Nod2^−/− adjuvant) mice; ***p < 0.0001. e Uveitis was scored by histopathology 21 days post-immunization. Data was combined from 6 independent experiments where n = 38 (WT) and 39 (Nod2^−/−) mice; ***p < 0.0001. f Representative H&E-stained sections of the anterior segment (top) and posterior segment (bottom) of eyes day 21 post immunization. Pathological features indicated: retinal folding (arrowhead), granulomatous formation (asterisk), vasculitis (arrow), and severe photoreceptor damage (bracket). I, iris; CB, ciliary body; V, vitreous; R, retina. g Apoptosis within the retina (14 days post immunization) was determined by a TUNEL-based assay. Data was combined from 2 independent experiments where n = 8 (WT) and 7 (Nod2^−/−) mice; ***p = 0.0003. h Leukocyte infiltration in the aqueous of the anterior segment (AS) and vitreous of the posterior segment (PS) were enumerated. Data are combined from 6 independent experiments where n = 35 (WT AS, WT PS), 39 (Nod2^−/−AS), 42 (Nod2^−/− PS) mice; **p = 0.0011 (AS) and ***p = 0.0005 (PS). Data are mean ± SEM (b, e, h) with dots representing individual values (e, h) or box-whisker plots with median, 25–75th percentile and min–max range (d, g). Source data are provided as a Source Data file.

Fig. 2. Nod2 influences retinal antigen-specific cellular responses.

a The proportion of ocular CD45⁺ cells of WT and Nod2^−/− mice 20 days post-immunization was determined by flow cytometry. Dashed line denotes baseline level for adjuvant control eyes. b Frequency of macrophages (CD45⁺CD11b⁺F4/80⁺), neutrophils (CD45⁺CD11b⁺GR-1⁺), dendritic cells (CD45⁺CD11c⁺), and T cells (CD45⁺CD3⁺CD4⁺) are summarized as stacked bars. c Live T cell populations of gated CD45⁺ cells. Data (a–c) are representative of 4 independent experiments each with n = 4 mice/group; *p = 0.0286 (a, c). d Uveitis was evaluated in immunized WT or Nod2^−/− mice that were depleted of CD4⁺ T cells with α-CD4 mAb or Isotype control Ab (IC Ab). Data are representative of 2 independent experiments with n = 12 mice/group; ***p < 0.0001 (WT IC vs. WT Ab, Nod2^−/− IC vs. Nod2^−/− Ab, WT IC vs. Nod2^−/− IC), and *p = 0.0373 (WT Ab vs. Nod2^−/− Ab). e–g Pinealitis was assessed in WT and Nod2^−/− mice 14 days post-immunization. e Pinealitis scores from histopathological evaluation, and (f) representative H&E-stained sections (circle denotes dense lymphocytic infiltration). Scores in e were combined from 2 independent experiments (shown left to right with n = 4, 6, 4, 5); *p = 0.0433 (WT IRBP vs. Nod2^−/− IRBP), *p = 0.0159 (Nod2^−/− adjuvant vs. Nod2^−/− IRBP). g Frequency of CD4⁺ and CD8⁺ T cells in pineal glands was determined by flow cytometry. Data are combined from 3 independent experiments, each having 5 pooled pineal glands/genotype/treatment. For CD4⁺ cells: *p = 0.0111 (WT IRBP vs. Nod2^−/− IRBP) and *p = 0.0081 (WT adjuvant vs. Nod2^−/− adjuvant)^. For CD8⁺ cells: *p = 0.0122 (WT adjuvant vs. Nod2^−/− adjuvant⁾. h BM chimeric mice were assessed for EAU by histopathology 20 days post-immunization. Data are combined from 3 independent experiments (shown left to right with n = 31,31,30,28); **p = 0.0030 and ***p < 0.0001. Data are box-whisker plots showing median, 25th–75th percentile and min–max range (a, c–e, h), floating bars showing median with min–max range (g), or mean (b). Data were analyzed with two-tailed Mann–Whitney U test (a–e, h) or unpaired, two-tailed Student’s t-test (g). Source data are provided as a Source Data file.

Fig. 3. Nod2 negatively regulates hyper-pathogenic Th17 cells.

Splenocytes from immunized WT and Nod2^−/− mice were stimulated in vitro with IRBP_1–20 (a–e). a [³H] thymidine incorporation after 72 h. Data are combined from 3 independent experiments, n = 12 mice/group; ***p < 0.0001. b Gene expression of CD4⁺ T cells 18 h was assessed by qPCR. Heatmap indicates expression relative to IRBP_1–20-stimulated CD4⁺ T cells from adjuvant-control mice. c Cytokine production after 48 h was assessed by ELISA. Data are combined from 4 independent experiments, n = 16 mice/group. d Frequencies of IL-17⁺ or IFNγ⁺ of CD4⁺Thy1.2⁺ T cells was determined by flow cytometry. Data are combined from 2 independent experiments, n = 5 (unstimulated) or 6 (IRPB-stimulated) mice/group. IL-17: **p = 0.0043 (WT unstimulated vs. WT peptide, Nod2^−/− unstimulated vs. Nod2^−/− peptide^), **p = 0.0022 (WT peptide vs. Nod2^−/− peptide^). IFNγ: **p = 0.0043 (WT unstimulated vs. WT peptide), **p = 0.0025 (Nod2^−/− unstimulated vs. Nod2^−/− peptide⁾. e Proportion of Th17 cells (IL-17⁺CD4⁺Thy1.2⁺) co-producing cytokines. Data are combined from 2 independent experiments, n = 6 mice/group; **p = 0.0022. f–h Ocular cells from immunized mice were stimulated in vitro with IRBP_1–20 for 18 h. Data are combined from 3 independent experiments, n = 6 mice pooled/genotype. f Representative contour plot showing IL-17⁺CD4⁺ T cells. g IL-17⁺ cells from gated live CD4⁺Thy1.2⁺ T cells. ^**p = 0.0013 (%), ***p = 0.0001 (#). h CD4⁺Thy1.2⁺ T cells co^-expressing TNFα, IFNγ, or Foxp3. i–k The effect of cytokine neutralization on uveitis was evaluated clinically and histopathologically 21 days post-immunization. Data are combined from 3 independent experiments, n = 18 (WT IC, WT αIL-17), 19 (WT αIFNγ, Nod2^−/− IC), 17 (Nod2^−/− αIL-1, or 23 (Nod2^−/− αIFNγ) mice. i Representative images of fundi and H&E-stained retinal sections 21 days post-immunization. j Uveitis scores from histopathology. *p = 0.0095 (WT IC vs. WT IL-17 Ab), ***p < 0.0001 (Nod2^−/− IC vs. Nod2^−/− IL-17 Ab, WT IC vs. Nod2^−/− IC). k Uveitis scores (from j) were normalized to the IC Ab group within each genotype (baseline = 0); **p = 0.0035, ***p < 0.0001. Data are box-whisker plots showing median with 25th–75th percentile and min–max range (a, d–e, j, k), floating bars showing median with min–max range (g, h), or mean ± SEM (dots represent individual values) (c). Data were analyzed by two-tailed Mann–Whitney U test (a, d–e, h, j) or unpaired, two-tailed Student’s t-test (c, g, k). Source data are provided as a Source Data file.

Fig. 4. Nod2 suppresses uveitis through a non-conventional cellular mechanism.

a–b CD4⁺ T cells from IRBP-immunized donor WT or Nod2^−/− mice were adoptively transferred into naïve WT recipients. a Clinical uveitis scores and incidence 20 days post-transfer. Data are from 6 independent experiments with n = 27 (WT) and 28 (Nod2^−/−) mice; *p = 0.0336. b Representative photos of WT recipients of WT or Nod2^−/− CD4⁺ T cells 20 days post-transfer. c IL-17 and IFN-γ were measured by ELISA following 18 h stimulation with IRBP_1–20 (20 µg/ml) of co-cultured APCs (isolated from naïve WT or Nod2^−/− mice) and CD4⁺ T cells (isolated from WT or Nod2^−/− immunized mice). Control APCs without the addition of T cells (“none”) were stimulated with IRBP. Data are combined from 4 independent experiments. For each experiment, APCs or T cell populations were isolated from n = 3 pooled mice/genotype; *p = 0.0286 for comparisons with both APC conditions. d–e Rag1^−/− or Nod2^−/−Rag1^−/−mice were reconstituted with WT CD4⁺ T cells and immunized with IRBP 24 h later. d Clinical uveitis was evaluated weekly. Data is combined from 3 independent experiments. For Rag1^−/− recipients n = 16 (7 days), 20 (14 days), 18 (21 and 28 days), 16 (35 days). For Nod2^−/−Rag1^−/− recipients n = 20 (7 days), 22 (14, 21, and 28 days), 20 (35 days). e Splenocytes were stimulated 18 h in vitro with IRBP_1–20 and IL-17 in culture supernatants was measured by ELISA. Data are combined from 3 independent experiments with n = 12 (Rag1^−/− media), 18 (Nod2^−/−Rag1^−/− media), 13 (Rag1^−/− IRBP), 15 (Nod2^−/−Rag1^−/− IRBP) mice; ***p = 0.0001. Data are mean ± SEM (a, d) with dots representing individual values (a) or box-whisker plots with median, 25–75th percentile and min–max range (c, e). Statistical differences between groups were calculated using unpaired, two tailed Student’s t-test (a), two-tailed Mann–Whitney U test (c, d, e). Source data are provided as a Source Data file.

Fig. 5. Nod2 inhibition of Th17-mediated uveitis is T cell-intrinsic.

Rag1^−/− mice were reconstituted with CD4⁺ T cells from naïve WT or Nod2^−/− mice and immunized 1 day later (a–f). a Clinical uveitis was evaluated weekly by fundoscopy after immunization. Data is combined from 6 independent experiments and (from left to right), n = 43, 40, 38, 24, 17 (WT) and n = 30, 46, 46, 28, 12 (Nod2^−/−) mice; *p = 0.0275 (14 day), *p = 0.0160 (21 day), *p = 0.0216 (28 day), *p = 0.0232 (35 day). b Images of fundus and H&E-stained retinal sections 35 days post-immunization. Splenocytes harvested 35 days post-immunization were analyzed by flow cytometry (c–e) or ELISA (f). c Frequencies of effector memory (CD62L^loCD44^mid), central memory (CD62L^loCD44^hi), and naïve CD4⁺ T cells (CD62L^hiCD44^lo). Data are combined from 2 independent experiments for n = 7 (WT) and 9 (Nod2^−/−); *p = 0.0418. d Cell surface expression of CD62L, CD69, and CD44 on CD4⁺ T cells, and (e) frequency and total number of CD4⁺ T cells expressing IL-23R recovered from spleens of IRBP-immunized Rag1^−/− recipients. Data are combined from 2 independently performed experiments with n = 5 (WT) and n = 4 (Nod2^-/-) mice; *p = 0.0159 (% and #). f IL-17 production was measured by ELISA following 18 h in vitro stimulation with IRBP_1–20. Data are combined from 2 independent experiments. For WT, n = 10 (media) and 9 (peptide) and for Nod2^−/−, n = 11(media) and 10 (peptide) mice; **p = 0.0030. (g-h) Rag1^−/− mice were reconstituted with thymocytes derived from naïve WT or Nod2^−/− mice and immunized with IRBP 8 days later. g Clinical uveitis was evaluated weekly by fundoscopy. Data is combined from 2 independent experiments. For WT, n = 14 (7, 14, 21, 35 day), 13 (28 day) and for Nod2^−/−, n = 12 mice; *p = 0.0331 (14 day), ***p < 0.0001 (21 day), ***p = 0.0002 (28 day), *p = 0.0347 (35 day). h IL-17 production was measured with ELISA following 18 h in vitro stimulation of splenocytes with IRBP_1–20. Data are combined from 2 independent experiments with n = 9 (WT media) and 8 (all other conditions) mice; *p = 0.0281. Data are mean ± SEM (a, g), box-whisker plots showing median, 25–75th percentile and min-max range (c, f), or box-whisker plots showing median, 10–90th percentile and min-max (e, h). Statistical differences between groups were calculated using two-tailed Mann–Whitney U test. Source data are provided as a Source Data file.

Fig. 6. Nod2 functions independent of Rip2-signaling in T cells.

a–b EAU was induced in WT and Rip2^−/− mice. Data are combined from 3 independent experiments. For WT, n = 10 (10 and 19 day) and 24 (13 day). For Rip2^−/−, n = 9 (10 day), 21 (13 day), 12 (19 day). a Clinical uveitis was evaluated weekly by fundoscopy; *p = 0.0320. b Representative fundus photographs 13 days post-immunization. c–d Rag1^−/− mice were reconstituted with thymocytes from naïve WT or Rip2^−/− mice and immunized with IRBP 8 days later. Data are combined from 2 independent experiments with n = 10 (WT recipients) and n = 18 (Rip2^−/− recipients). c Clinical uveitis was evaluated by fundoscopy 28 days post-immunization; *p = 0.0074. d Representative fundus photographs are shown. Data are mean ± SEM (a) or box-whisker plots showing median, 25–75th percentile, and min–max range (with dots representing individual values) (c). Statistical differences between groups were calculated using two-tailed Mann–Whitney U test (a, c). Source data are provided as a Source Data file.

Fig. 7. Nod2 controls Th17 immunity in naïve mice and Blau syndrome.

a Nod2 expression in IRBP_1–20–stimulated CD4⁺ T cells of naïve, adjuvant-only, or immunized WT mice relative to unstimulated CD4⁺ T cells from naïve WT mice was measured by qPCR. Controls include IFNγ-stimulated WT BMDM and Nod2^−/− BMDM cells. Data are representative of 2 independent experiments, each with n = 5 spleens pooled/genotype. b–c Rag2^−/− hosts reconstituted with CD4⁺ T cells from naïve R161M or Nod2^−/−R161M mice were assessed for uveitis 7 days post-transfer. Data are combined from 3 independent experiments with n = 12 mice/group. b Clinical uveitis was scored, *p = 0.0500. c Representative fundus and histopathology photographs. d–g T cell responses in naïve WT and Nod2^−/− (C57BL/6 J) mice. d Naïve CD4⁺ T cells (CD62L^hiCD44^-) from WT and Nod2^−/− mice were differentiated under Th17-polarizing conditions and frequency of IL-17-producing CD4⁺ T cells (CD4⁺CD8^-Thy1.2⁺IL-17⁺) was determined by flow cytometry. Data are combined from 2 independent experiments for n = 8 mice/genotype. e Frequency of memory (CD62L^-CD44⁺) CD4⁺ T cells from naïve WT or Nod2^−/− mice expressing indicated factors was determined by flow cytometry 5 days post-stimulation with anti-CD3/anti-CD28 mAb. Data is representative of 3 independently performed experiments, each with n = 4 mice/genotype; *p = 0.0286 for all indicated comparisons. f–g Memory CD4⁺ T cells from naive WT or Nod2^−/− mice stimulated 16 h with anti-CD3/anti-CD28 mAb were analyzed by multiplex-qPCR. Transcripts are relative to unstimulated control memory CD4⁺ T cells of WT mice (horizontal line), with n = 3 mice/condition. h–i PBMCs of two healthy controls or two Blau syndrome patients were stimulated with anti-CD3/anti-CD28 mAb for 5 days then with PMA/ionomycin for 4 h. Experiment was repeated twice. h Frequency of IL-17-producing CD4⁺ T cells (CD3⁺CD4⁺CD19^-IL-17⁺)(*p = 0.0055) and representative contour plot. i Representative histogram showing MFI of CCR7 by CD4⁺ T cells (CCR7⁺CD3⁺CD4⁺CD19^-). Data are box-whisker plots showing median with 25–75th percentile and min–max range (b, d), box-whisker plots showing median with 10–90th percentile (e), or floating bars showing median and min–max range (f, g). Dots represent individual values (b, h). Statistical differences were calculated using two-tailed Mann–Whitney U test (d, e) or unpaired, two-tailed Student’s t-test (b, f–h). Source data are provided as a Source Data file.⁺.