BCG-Induced Tumor Immunity Requires Tumor-Intrinsic CIITA Independent of MHC-II

Abstract

For decades, BCG immunotherapy has been the standard of care for non-muscle-invasive bladder cancer. Despite this clinical experience, the mechanism by which BCG stimulates tumor-eliminating immunity is unclear, and there is still a need for more accurate prediction of clinical outcomes in advance of treatment initiation. We have shown that BCG stimulates tumor-specific T-cell immunity that requires tumor cell expression of the IFNγ receptor (IFNGR); however, the downstream components of IFNGR signaling responsible for responsiveness to BCG are unknown. Here, we demonstrate that the IFNγ-driven, tumor cell intrinsic expression of the class II transactivator CIITA is required for activation of a tumor-specific CD4 T-cell response and BCG-induced tumor immunity. Despite the established role for CIITA in controlling MHC-II antigen presentation machinery, the requirement for CIITA is independent of MHC-II and associated genes. Rather, we find that CIITA is required for a broader tumor-intrinsic transcriptional program linked to critical pathways of tumor immunity via mechanisms that remain to be determined. Tumor cell intrinsic expression of CIITA is not required for a response to immunotherapy targeting programmed cell death protein 1 (PD-1), suggesting that different modalities of immunotherapy for bladder cancer could be employed based on tumor-intrinsic characteristics.

Figure 1:. MHC-II expression in murine and human bladder cancer cells

A. MB49 cells were treated with 100 μg/mL IFNγ for 72 hours and expression of CD80 and CD86 was determined by flow cytometry. Expression of CD80 with and without IFNγ is shown on the left and a representative flow plot is show on the right. B. Experimental schematic. Mice underwent implantation of MB49-YFP bladder tumors and received intravesical BCG or PBS on Days 2, 9, and 16. On Days 7, 14, and 21 mice from each group were euthanized, and tumor single-cell suspensions were stained and evaluated by flow cytometry. C. Percent MHC-II+ among live, YFP+ cells in BCG- and PBS-treated bladders. Representative flow plots from each time point are shown on the right. D. The indicated human bladder cancer cell lines were treated with 100 μg/mL IFNγ for 72 hours and the percent of cells expressing MHC-II was determined by flow cytometry. E. MHC-II immunohistochemical staining of four human non-muscle-invasive bladder cancer specimens. Data is representative of two independent experiments. Error bars represent average ± SD.

Figure 2:. Expression of CIITA by bladder cancer cells is required for the efficacy of intravesical BCG therapy

A. MB49 or MB49_CIITAKO cells were treated with 100 μg/mL IFNγ for 72 hours. Percent of cells expressing MHC-II was determined by flow cytometry. Error bars represent average ± SD. P-values were derived by one-way ANOVA with Bonferroni’s multiple comparisons. Data is representative of three independent experiments. B. MB49 or MB49_CIITAKO cells were treated with 100 μg/mL IFNγ for 72 hours. Percent of cells expressing MHC-I (left panel) or PD-L1 (right panel) was determined by flow cytometry. Error bars represent average ± SD. P-values were derived by one-way ANOVA with Bonferroni’s multiple comparisons. Data is representative of three independent experiments. C. Mice underwent implantation of MB49 or MB49_CIITAKO bladder tumors and received five weekly treatments of intravesical BCG or PBS starting on Day 2. There were 26 mice in each of the BCG groups and 24 mice in each of the PBS groups. Survival curves are shown. P-values derived by log-rank test. Data is representative of three independent experiments. D. Mice (n=15 in each group) underwent implantation of MB49, MB49_CIITAKO#1 (clone used in the other experiments), MB49_CIITAKO#2, or MB49_CIITAKO#3 bladder tumors and received five weekly treatments of intravesical BCG starting on Day 2. Survival curves are shown. P-values derived by log-rank test. E. Experimental schematic. Mice were implanted with MB49 bladder tumors and given 5 weekly BCG treatments to generate a cohort of survivor mice. 6 weeks after the final BCG treatment, 2×10⁵ MB49 or MB49_CIITAKO cells were injected subcutaneously in the flank. Tumor area was determined 2 weeks later. F. MB49 or MB49_CIITAKO tumor areas of mice described in E. Data are representative of two independent experiments. Error bars represent mean ± SD. P-values were derived by one-way ANOVA with Bonferroni’s multiple comparisons. *, p < 0.05; **, p < 0.005; ***, p < 0.0005. ns – non-significant

Figure 3:. MHC-II antigen presentation is dispensable for BCG-induced tumor elimination

A. Mice underwent implantation of MB49 or MB49_MHCIIKO bladder tumors and received five treatments with weekly intravesical BCG or PBS starting on Day 2. Survival curves are shown. P values derived by log-rank test. Data is representative of three independent experiments. B. Experimental schematic. Mice were implanted with MB49 bladder tumors and given 5 weekly BCG treatments to generate a cohort of survivor mice. 6 weeks after the final BCG treatment, 2×10⁵ MB49_MHCIIKO cells were injected subcutaneously in the flank. Tumor area was determined 2 weeks later. C. MB49 or MB49_MHCIIKO tumor areas of mice described in B. Data are representative of two independent experiments. Error bars represent mean ± SD. P-values were derived by Student’s t-test. Data is representative of two independent experiments. D. Mice (n=15 in each group) underwent implantation of MB49_H2-DMKO bladder tumors and received five treatments with weekly intravesical BCG or PBS starting on Day 2. Survival curves are shown. P-values derived by log-rank test. Data is representative of two independent experiments. E. Mice (n=15 in each group) underwent implantation of MB49 or MB49_CD74KO bladder tumors and received five treatments with weekly intravesical BCG or PBS starting on Day 2. Survival curves are shown. P-values derived by log-rank test. Data is representative of two independent experiments. *, p < 0.05; **, p < 0.005; ***, p < 0.0005. ns – non-significant.

Figure 4:. Dynamics of tumor elimination in mice with mixed wild type and CIITA-deficient tumors

A. Experimental schematic. Mice underwent intravesical implantation of a mixture of MB49-YFP and MB49_CIITAKO-mCherry cells (25,000 cells of each type) and received intravesical BCG or PBS on Days 2, 9, and 16. On Days 14 and 21, mice from each group were euthanized, and tumor single-cell suspensions were evaluated by flow cytometry. After excluding dead cells and CD45+ cells, the ratios of YFP+ to mCherry+ cells were determined. B. Representative Day 14 flow plots of tumors from mice implanted with mixed MB49-YFP and MB49_CIITAKO-mCherry tumors and treated with BCG (left plot) or PBS (right plot). The percentage of MB49-YFP and MB49_CIITAKO-mCherry is indicated. C. Weight of bladders from pre-tumor implantation and BCG- and PBS-treated tumor-bearing mice on Days 14 and 21. Error bars represent mean ± SD. D. Ratio of YFP+ to mCherry+ cells in PBS-treated mice on Days 14 and 21. Error bars represent mean ± SD. Data is representative of two independent experiments. E. Ratio of YFP+ to mCherry+ cells in BCG-treated mice on Days 14 and 21. Error bars represent mean ± SD. Data is representative of two independent experiments.

Figure 5:. CIITA is required for the tumor-specific but not BCG-specific CD4 T-cell response

A. Experimental schematic. Mice underwent implantation of MB49 or MB49_CIITAKO bladder tumors and received intravesical BCG or PBS on Days 2, 9, and 16. On Days 17 and 21, tumors were removed and analyzed by flow cytometry. B. CD4+FOXP3− percent of live CD45+ cells on Days 17 and 21. C. CD8+ percent of live CD45+ cells on Days 17 and 21. D. CD44+CD62L− and PD-1+ percent of live CD4+FOXP3− cells on Day 17. E. CD44+CD62L− and PD-1+ percent of live CD8+ cells on Day 17. F. Experimental schematic. Mice underwent implantation of MB49^OVA or MB49_CIITAKO^OVA bladder tumors and received intravesical BCG on Days 2, 9, and 16. On Day 7, 1×10⁶ BCG-specific P25 (CD45.1) and OVA-specific OT-II (CD90.1) CD4 T cells were transferred via retro-orbital injection. On Day 20, bladder-draining lymph nodes were removed and analyzed by flow cytometry. G. CD4+FOXP3− percent of live CD45+ cells (left panel), OT-II (CD45.1-CD90.1+) percent of live CD4+ cells (central panel), and P25 (CD45.1+CD90.1−) percent of live CD4+ cells (right panel). H. CD44+CD62L− and PD-1+ percent of endogenous CD4 (CD45.1-CD90.1−), OT-II (CD45.1-CD90.1+), and P25 (CD45.1+CD90.1−) T cells. I. IFNγ percent positive and median fluorescent intensity (MFI) of endogenous (CD45.1-CD90.1), OT-II (CD45.1-CD90.1+), and P25 (CD45.1+CD90.1−) CD4 cells following ex vivo stimulation with PMA and ionomycin. Data is representative of two independent experiments. P-values were derived by one-way ANOVA with Bonferroni’s multiple comparisons. Error bars represent mean ± SD. *, p < 0.05; **, p < 0.005; ***, p < 0.0005

Figure 6:. RNA sequencing of PBS- or BCG-treated MB49 versus MB49_CIITAKO bladder tumors

A. Experimental schematic. Mice underwent implantation of MB49-YFP or MB49_CIITAKO-YFP bladder tumors and received intravesical BCG or PBS on Days 2, 9, and 16. On Day 21, tumors were removed, single cell suspensions were prepared, and cells were stained for viability and CD45 expression. Live CD45− YFP+ cells were sorted and analyzed by RNA sequencing. B. Tree plot using Pearson correlation clustering specimens by transcriptional profile, with indicated BCG treatment status and tumor genotype. C. Volcano plot displaying the comparison of wild-type and CIITA-deficient specimens. Significant differentially expressed genes (FDR-corrected P ≤ 0.01) with at least 2-fold (CIITA versus wild-type) increased or decreased expression are highlighted. D. Heat map showing the 57 top differentially expressed genes (by fold-change) between wild-type and CIITA-deficient specimens. E. Enriched hallmark gene set pathways in wild-type versus CIITA-deficient BCG-treated specimens (left panel) and in CIITA-deficient versus wild-type BCG-treated specimens (right panel). X-axis represents the negative log of FDR-corrected q values, F. Enriched hallmark gene set pathways in wild-type BCG- versus PBS-treated tumors (left panel) and in CIITA-deficient BCG- versus PBS-treated tumors (right panel). X-axis represents negative log of FDR-corrected q values,

Figure 7:. Expression of CIITA by tumor cells is not required for response to treatment with PD-1 blocking antibodies

A. Experimental schematic. Mice underwent implantation of MB49 or MB49_CIITAKO bladder tumors and received intraperitoneal injection of PD-1 blocking antibody or PBS on Days 3, 6, 9, and 12. B. Survival curve of mice (n=10 in each group) from A. P-values derived by log-rank test. C. B16 or B16_CIITAKO cells were treated with 100 μg/mL IFNγ for 72 hours. Percent MHC-II+ cells was determined by flow cytometry. Error bars represent average ± SD. P-values were derived by one-way ANOVA with Bonferroni’s multiple comparisons. D. B16 or B16_CIITAKO cells were treated with 100 μg/mL IFNγ for 72 hours. Percent of cells expressing MHC-I (right panel) or PD-L1 (left panel) was determined by flow cytometry. Error bars represent average ± SD. P-values were derived by one-way ANOVA with Bonferroni’s multiple comparisons. Representative histograms of MHC-I expression under the different conditions are shown on the right. E. Experimental schematic. Mice underwent implantation of intradermal B16 or B16_CIITAKO in the flank and received intraperitoneal injection of PD-1 blocking antibody or PBS on Days 3, 6, 9, and 12. Tumor area was determined 2 weeks later. F. B16 or B16_CIITAKO tumor areas of mice (n=5 in each group) described in E. Data are representative of two independent experiments. Error bars represent mean ± SEM. *, p < 0.05; **, p < 0.005; ***, p < 0.0005. ns – non-significant.