Interleukin-37 promotes colitis-associated carcinogenesis via SIGIRR-mediated cytotoxic T cells dysfunction

Abstract

Interleukin-37b (hereafter called IL-37) was identified as fundamental inhibitor of natural and acquired immunity. The molecular mechanism and function of IL-37 in colorectal cancer (CRC) has been elusive. Here, we found that IL-37 transgenic (IL-37tg) mice were highly susceptible to colitis-associated colorectal cancer (CAC) and suffered from dramatically increased tumor burdens in colon. Nevertheless, IL-37 is dispensable for intestinal mutagenesis, and CRC cell proliferation, apoptosis, and migration. Notably, IL-37 dampened protective cytotoxic T cell-mediated immunity in CAC and B16-OVA models. CD8⁺ T cell dysfunction is defined by reduced retention and activation as well as failure to proliferate and produce cytotoxic cytokines in IL-37tg mice, enabling tumor evasion of immune surveillance. The dysfunction led by IL-37 antagonizes IL-18-induced proliferation and effector function of CD8⁺ T cells, which was dependent on SIGIRR (single immunoglobulin interleukin-1 receptor-related protein). Finally, we observed that IL-37 levels were significantly increased in CRC patients, and positively correlated with serum CRC biomarker CEA levels, but negatively correlated with the CD8⁺ T cell infiltration in CRC patients. Our findings highlight the role of IL-37 in harnessing antitumor immunity by inactivation of cytotoxic T cells and establish a new defined inhibitory factor IL-37/SIGIRR in cancer-immunity cycle as therapeutic targets in CRC.

Fig. 1

The colon epithelium of IL-37-transgenic mice exhibits balanced homeostasis. a Representative H&E-stained sections (left, scale bar: 50 μm) of the distal colon of IL-37tg mice and WT mice, and quantification of the crypt length (right) is shown. At least 15 well-oriented crypts were measured on slides from each mouse, n = 5/group. b Representative alcian blue and PAS staining of distal colon sections obtained from IL-37tg mice and WT mice (left, scale bar: 50 μm). Statistical analysis of enumeration of goblet cells per crypt (right), at least 15 well-oriented crypts were measured on slides from each mouse, n = 5/group. c 50 mg/kg BrdU was administrated to IL-37tg mice and WT mice by intraperitoneal injection. The colon sections were harvested and stained for BrdU-positive cells at 24 h after BrdU injection. Representative BrdU-stained sections are shown (left, scale bar: 50 μm). Statistical analysis of BrdU-positive cells per crypt (right), at least 15 well-oriented crypts were counted on slides from each mouse, n = 5/group. d DNA fragmentation in the colon of IL-37tg mice and WT mice was detected by in situ TUNEL assay. Apoptotic cells were stained brown and nuclei were stained blue by hemotoxylin, at least 15 Well-oriented crypts were counted on slides from each mouse, scale bar: 50 μm, n = 5/group. e Immunoblotting was performed to detect CDK4 and Bcl-xL in colon epithelial cells from IL-37tg and WT mice. β-Actin was used as a loading control. f qRT-PCR analysis of indicated genes from the colons of IL-37tg mice and WT mice, n = 5/group. g The same amount of colon tissue from IL-37tg mice and WT mice was cut into small pieces and incubated in serum-free RPMI medium for 24 h. Secreted chemokines and cytokines in the medium were measured by ELISA assay, n = 5/group. h qRT-PCR analysis was performed for each indicated gene in colon tissues of IL-37tg mice and WT mice. n = 5/group. All data are presented as mean ± SD. Statistics analyzed by Two-tailed Student’s T-test. NS, not significant

Fig. 2

Enhanced tumorigenesis in IL-37tg mice. a Details of the azoxymethane/dextran sodium sulfate (AOM/DSS) treatment used for the induction of inflammation-associated colorectal cancer. IL37tg mice and WT mice were injected with AOM on day 0, and were then given a 2% DSS solution during three 6-day cycles as described in experimental procedures. b qRT-PCR analysis of IL-37 expression in the colon of IL-37tg mice after treatment with or without AOM/DSS, n = 5/group. c The body weight loss of IL-37tg mice and WT mice following injected with AOM on day 0 and administered 3 rounds of 2% DSS in drinking water, n = 8/group. d Representative image of the distal colon at day 70 after AOM/DSS administration, tumor development in the colon was determined. e Quantification of the number of tumors in the colon at day 70 after AOM/DSS administration, n = 8/group. f Colon weight was determined in IL-37tg mice and WT mice at day 70 after AOM/DSS administration, n = 8/group. g Colon tissue sections of IL-37tg mice and WT mice by H&E staining at day 70 after AOM/DSS administration. Scale bar, 100 μm. h, i Total histological scores (h) and scores for different parameters (i) of IL-37tg mice and WT mice at day 70 after AOM/DSS administration, n = 8/group. j–l Representative Ki67 (j), p-Stat3 (k), Cleaved Caspase3 (l) immunohistochemistry of colon in IL-37tg mice and WT mice at day 70 after AOM/DSS administration, scale bars: 100 μm. Positive cells were quantified by counting the stained dots, n = 8/group. All data are presented as mean ± SD. Statistics analyzed by Two-tailed Student’s T-test. *P < 0.05; **P < 0.01; ***P < 0.001

Fig. 3

IL-37 inhibited the activation of CD8⁺ cytotoxic T cells in AOM/DSS-induced CRC mouse model. a IL-37tg mice and WT mice were treated with AOM/DSS for 70 days, absolute cell numbers of CD4⁺ T cells and CD8⁺ T cells in the tumor-bearing colon tissue as assessed by flow cytometry. A total of 200 000 live colonic cells were acquired to normalise the baselines for all samples, n = 5/group. b Quantification of intracellular FACS analysis of CD4⁺ T cells (Th1), IL-17A-producing CD4⁺ T cells (Th17), and Foxp3⁺ CD4⁺ T cells (Treg) in the colon of IL-37tg mice and WT mice which were treated with AOM/DSS for 70 days, n = 5/group. c, d IL-37tg mice and WT mice were treated with AOM/DSS for 70 days, absolute cell numbers of CD19⁺ cells (B cells) and NK1.1⁺ cells (NK cells), CD11B⁺Gr-1⁺ (MDSCs), CD11B⁺F4/80⁺ cell (Mφ cells) and CD11C⁺MHC-II⁺ cell (DCs) in the colon as assessed by flow cytometry. A total of 200,000 live colonic cells were acquired to normalise the baselines for all samples, n = 5/group. e, f IL-37tg mice and WT mice were treated with AOM/DSS for 70 days, representative and quantification of FACS analysis of CD8⁺ T cells expressing proliferation marker Ki67, apoptosis marker Annexin V, integrin associated with intestinal T cell retention CD103, activation marker (CD44, CD62L or CD69), and effector molecules (CD107 and IFN-γ) in the tumor-bearing colon tissue (e) and mesenteric lymph nodes (f), n = 5/group. All data are presented as mean ± SD. Statistics analyzed by Two-tailed Student’s T-test. *P < 0.05; **P < 0.01; ***P < 0.001. NS, not significant

Fig. 4

IL-37 enhanced tumorigenesis via CD8⁺ T cell inactivation. a–c IL-37tg mice and WT mice were subcutaneous injection of 5 × 10⁵ B16-OVA melanoma cells. a Xenograft tumor growth curve of different groups. n = 5/group. b, c The percentage of CD8⁺ T cells was measured 15 days after tumor inoculation. Quantification (b) and representative (c) of FACS analysis of CD8⁺ T cells and Tetramer-SIINFEKL CD8⁺ T cells, CD8⁺ T cells expressing activation marker CD69 and effector molecules IFN-γ. n = 5/group. d Xenograft tumor growth curve of different treatment groups. The mice were subcutaneous injection of 5 × 10⁵ B16-OVA melanoma cells, subsequently intraperitoneal injection of 250 μg CD8-neutralizing antibody or IgG2b isotype control twice a week, B16-OVA tumor growth was monitored. n = 6/group. e, f IL-37tg mice and WT mice were treated with or without AOM/DSS for 70 days, simultaneously intraperitoneal injection of 250 μg CD8-neutralizing antibody or IgG2b isotype control twice a week. e Quantification of the number of tumors in the colon, n = 5/group. f Colon weight was determined, n = 5/group. The data are presented as mean ± SD. Statistics analyzed by Two-way ANOVA analysis of variance with Turkey’s post hoc test (a, d), Two-tailed Student’s T-test (b), and One-way ANOVA analysis of variance with Turkey’s post hoc test (e, f). **P < 0.01; ***P < 0.001. NS, not significant

Fig. 5

IL-37 inhibited the CD8⁺ cytotoxic T cells–mediated antitumor immunity that was depended on IL-18. a IL-37tg mice and WT mice were treated with or without AOM/DSS for 70 days, simultaneously intraperitoneal injection of 50 μg IL-18–neutralizing antibody or IgG2a isotype control twice a week. Quantification of the number of tumors in the colon, n = 5/group. b Xenograft tumor growth curve of different treatment groups. The mice were subcutaneous injection of 2 × 10⁵ B16-OVA melanoma cells, subsequently intravenous injection of 100 ng recombinant IL-37 every day, and intraperitoneal injection of 50 μg IL-18–neutralizing antibody or IgG2a isotype control twice a week, B16-OVA tumor growth was monitored. n = 5/group. c IL-37tg mice and WT mice were treated with or without AOM/DSS for 70 days, simultaneously intraperitoneal injection of 50 μg IL-18–neutralizing antibody or IgG2a isotype control twice a week, absolute cell numbers of CD8⁺ T cells in the tumor-bearing colon tissue as assessed by flow cytometry. A total of 200,000 live colonic cells were acquired to normalise the baselines for all samples, n = 5/group. d, e The mice were subcutaneous injection of 2 × 10⁵ B16-OVA melanoma cells, subsequently intravenous injection of 100 ng recombinant IL-37 every day, and intraperitoneal injection of 50 μg IL-18–neutralizing antibody or IgG2A isotype control twice a week. The percentage of CD8⁺ T cells was measured 15 days after tumor inoculation. Representative and quantification of FACS analysis of CD8⁺ T cells and Tetramer-SIINFEKL CD8⁺ T cells, CD8⁺ T cells expressing activation marker effector molecules IFN-γ in the tumor tissue (d) and tumor-draining lymph node (e). n = 5/group. The data are presented as mean ± SD. Statistics analyzed by One-way ANOVA analysis of variance with Turkey’s post hoc test (a, c–e). Two-way ANOVA analysis of variance with Turkey’s post hoc test (b). *P < 0.05; **P < 0.01; ***P < 0.001. NS, not significant

Fig. 6

IL-37 limited IL-12/18 induce the cytotoxic activity of CD8⁺ T cells. a Isolated mouse naive CD8⁺ T cells were labeled with 5 μM CFSE and pulsed with or without CD3/CD28 in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 100 ng/mL IL-37b for 72 h. Proliferation was determined by the CFSE dilution assay. Representative histograms of CFSE dilution (left) and statistical analysis of proliferating cells (right). Numbers in the histogram plots represent the percentage of proliferating cells, n = 3/group. b ELISA analysis of IFN-γ levels in culture medium from isolated mouse naive CD8⁺ T cells treated with or without CD3/CD28 in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 100 ng/mL IL-37b for 72 h. n = 3/group. c Isolated mouse naive CD8⁺ T cells were treated with or without CD3/CD28 in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 100 ng/mL IL-37b for 72 h. Representative and quantification of FACS analysis of CD107⁺CD8⁺ cells and IFN-γ⁺CD8⁺ cells, n = 3/group. d Isolated mouse naive OT-I CD8⁺ cells were labeled with 5 μM CFSE and pulsed with or without 5 μg/mL OVA257-264 peptides in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 100 ng/mL IL-37b for 72 h. Proliferation was determined by the CFSE dilution assay. Representative histograms of CFSE dilution (left) and statistical analysis of proliferating cells (right). Numbers in the histogram plots represent the percentage of proliferating cells, n = 3/group. e ELISA analysis of IFN-γ levels in culture medium from isolated mouse naive CD8⁺ T cells treated with or without 5 μg/mL OVA257-264 peptides in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 100 ng/mL IL-37b for 72 h. n = 3/group. f Isolated mouse naive CD8⁺ T cells were treated with or without 5 μg/mL OVA257-264 peptides in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 100 ng/mL IL-37b for 72 h. Representative and quantification of FACS analysis of CD107⁺CD8⁺ cells and IFN-γ⁺CD8⁺ cells. n = 3/group. All data are presented as mean ± SD. Statistics analyzed by One-way ANOVA analysis of variance with Turkey’s post hoc test. **P < 0.01; ***P < 0.001

Fig. 7

IL-37 limited IL-12/IL-18–induced functional activities of CD8⁺ T cells via SIGIRR. a Isolated mouse naive CD8⁺ T cells were transfected for 72 h with the SMARTpool siRNA reagent against SIGIRR or with a control Accell nontargeting siRNA, transfected CD8⁺ T cells were labeled with 5 μM CFSE and pulsed with or without CD3/CD28 in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 100 ng/mL IL-37b for 72 h. Proliferation was determined by the CFSE dilution assay. Statistical analysis of proliferating cells, numbers in the histogram plots represent the percentage of proliferating cells, n = 3/group. b–d SIGIRR siRNA or Ctrl siRNA transfected CD8⁺ T cells were treated with or without CD3/CD28 in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 100 ng/mL IL-37b for 72 h. Quantification of FACS analysis of CD107⁺CD8⁺ cells (b) and IFN-γ⁺CD8⁺ cells (c), ELISA analysis of IFN-γ levels in culture medium (d). n = 3/group. e Isolated mouse naive CD8⁺ OT-I cells were transfected for 72 h with the SMARTpool siRNA reagent against SIGIRR or with a control Accell nontargeting siRNA, transfected CD8⁺ T cells were labeled with 5 μM CFSE and pulsed with or without 5 μg/mL OVA257-264 peptides in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 100 ng/mL IL-37b for 72 h. Proliferation was determined by the CFSE dilution assay. Statistical analysis of proliferating cells, numbers in the histogram plots represent the percentage of proliferating cells, n = 3/group. f–h SIGIRR siRNA or Ctrl siRNA transfected CD8⁺ T cells were treated with or without 5 μg/mL OVA257-264 peptides in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 100 ng/mL IL-37b for 72 h. Quantification of FACS analysis of CD107⁺CD8⁺ cells (f) and IFN-γ⁺CD8⁺ cells (g), ELISA analysis of IFN-γ levels in the culture medium (h), n = 3/group. i Naive CD8⁺ T cells were isolated from the spleen of IL-37tg mice and WT mice, isolated CD8⁺ T cells were treated with CD3/CD28 in the presence or absence of 1 ng/mL IL-12, 100 ng/mL IL-18, 10 µg/ml anti-SIGIRR antibody, 10 µg/ml Ctrl antibody for 48 h. ELISA analysis of IFN-γ levels in the culture medium. n = 3/group. j Naive CD8⁺ T cells were isolated from the spleen of IL-37tg mice and WT mice, isolated CD8⁺ T cells were treated with CD3/CD28 in the presence or absence of 5 ng/ml TGF-β, 3 μM SIS3 for 48 h. ELISA analysis of IFN-γ levels in the culture medium. n = 3/group. k–l Naive CD8⁺ T cells were isolated from the spleen of IL-37tg mice and WT mice, isolated CD8⁺ T cells were stimulated with CD3/CD28, 1 ng/mL IL-12, and 100 ng/mL IL-18 presence 10 µg/ml anti-SIGIRR antibody or 10 µg/ml Ctrl antibody for 48 h. k Quantification of FACS analysis of IRAK4, and JNK phosphorylation in CD8⁺ T cells. MFI, mean fluorescent intensity. n = 3/group. l Immunoblotting was performed to detect phosphorylation of IRAK4 and JNK in CD8⁺ T cells. GAPDH was used as a loading control. All data are presented as mean ± SD. Statistics analyzed by One-way ANOVA analysis of variance with Turkey’s post hoc test (a–h). Statistics analyzed by Two-tailed Student’s T-test (i–k). *P < 0.05; **P < 0.01; ***P < 0.001. NS, not significant

Fig. 8

IL-37 levels were negatively correlated with CD8⁺ cytotoxic T cells in patients with CRC. a The expression distribution of IL-37 in CRC tissues (n = 620) and adjacent normal tissues (n = 51) from TCGA database. b Serum levels of IL-37 were measured by ELISA, IL-37 levels in CRC patients (n = 32) compared with normal controls (n = 21). c A significant correlation was found between the serum levels of IL-37 and CEA in CRC patients (n = 32). d The score distribution of CD8⁺ T cells in CRC tissues (n = 620) and adjacent normal tissues (n = 51) from TCGA database. e Kaplan–Meier survival curves comparing the high (red) and low (blue) expression of CD8 in colorectal cancer, the data available from the human protein atlas datasets (v 20.0.proteinatlas.org, https://www.proteinatlas.org/ENSG00000153563-CD8A/pathology/colorectal+cancer). f Representative pictures for low and high CD8⁺ cells infiltration in colorectal cancer tumors of CRC patients, scale bar: 50 μm. Quantification of the infiltration percentage of CD8⁺ T cells in colorectal cancer tumors of CRC patients (n = 32). g A significant correlation was found between the serum levels of IL-37 and CD8⁺ cells in colorectal cancer tumors in CRC patients (n = 32). The data are presented as mean ± SD, statistics analyzed by Two-tailed Student’s T-test, *P < 0.05. (a, b, d). The r and p values obtained with Spearman’s rank correlation test (c, g)

Fig. 9

Schematic representation for the functional mechanism of IL-37 in CRC. IL-37 requires SIGIRR to interrupts the IL-18 drives MyD88–IRAK4–JNK signaling in CD8⁺ CTLs, subsequently leads to suppression of proliferation and cytotoxic activity of CD8⁺ CTLs, IFN-γ production are decreased. Dysfunctional CD8⁺ CTLs were deprived of immune surveillance for tumor, enhancing tumorigenesis of colorectal cancer.