αPD-L1 bladder cancer (BC) immunotherapy is effective in <30% of cases. To address the large αPD-L1-unresponsive subset of patients, we tested αIL-2/IL-2 complexes (IL-2c) that block IL-2 from binding high-affinity IL-2Rα (CD25) for preferential IL-2Rβ (CD122) binding. Regulatory T cells (Tregs) capture IL-2 by CD25 whereas CD8+T, γδ T, and NK cells use CD122. We hypothesized that the tumor microenvironment, including local immune cells in primary versus metastatic BC, differentially affects immunotherapy responses. We used PD-L1+ mouse BC cell lines MB49 and MBT-2, for intravesical ([IVe] in bladder) and intravenous (IV) challenge studies of local versus metastatic BC. αPD-L1 or IL-2c alone reduced tumor burden and extended survival in IVe MB49 and MBT-2. Treg depletion using FOXP3DTR mice further enhanced IVe IL-2c effects, consistent with the known tumor-promoting role of Tregs in human BC. Using in vivo cell depletion approaches, we found that γδ T cells and NK cells, but not CD8+ T cells, were necessary for IL-2c efficacy in bladder. γδ T cells also reduced intratumoral Th17 cells that promote MB49 growth and are elevated in human BC. We confirmed γδ T cell effects in δ TCR KO mice, which abrogated IL-2c efficacy but not αPD-L1 efficacy. Neither αPD-L1 nor IL-2c alone treated metastatic MB49 and MBT-2 BC but the combination improved survival in both. These data are consistent with our recent findings in human BC patients in whom γδ T cell and NK cell cytotoxicity improved BCG immunotherapy. Thus, IL-2c is a promising novel BC immunotherapy that can improve bladder-specific immunity in primary BC. In metastatic BC, combination with αPD-L1 may also be a successful BC treatment strategy due to engagement of innate and adaptive immune responses.

αPD-L1 is an FDA-approved bladder cancer (BC) immunotherapy but is effective in ≤30% of cases. We tested conjugates of αIL-2 antibody + IL-2 (IL-2c) that block IL-2 from binding high-affinity IL-2Rα (CD25) for preferential IL-2Rβ(CD122) binding. CD25 and CD122 are preferred for IL-2 capture by regulatory T cells and anti-tumor effector T cells (Teff), respectively, allowing IL-2c to target Teff preferentially. Orthotopic, intravesical (in bladder) MB49 BC produces PD-L1+BC tumors in syngeneic BL6 mice. αPD-L1 or IL-2c treated intravesical, but not lung metastatic MB49 produced by intravenous injection. Still, in metastatic BC, αPD-L1 + IL-2c (combo) reduced lung metastases and extended survival. Preliminary data found combo treatment efficacy was better in lung versus bladder. In subcutaneous (SQ) B16 melanoma, combo was better than single agents, and increased CD8+CXCR5+TCF-1+Tim-3−PD-1+T stem cells (CXCR5+SC, see X. Zhang poster) vs. single agents. Neither αPD-L1 nor combo increased CXCR5+SC in MB49 in bladder or lung metastases, suggesting a novel treatment mechanism. To test the impact of tumor PD-L1 on treatment efficacy, we made PD-L1KOMB49, but it did not grow in bladder or SQ in wild type mice, but grew similar to control MB49 in immunodeficient NSG mice SQ, suggesting tumor PD-L1 microenvironment-specific effects. PD-L1KOB16 grew well SQ in wild type mice, excluding a PD-L1-specific defect. Thus, tumor PD-L1 differs in immune evasion in a tumor-dependent manner. Selective IL-2 targeting to CD122 improves αPD-L1 treatment of metastatic BC. Mechanisms differ from melanoma, which could be due to tumor, PD-L1, or microenvironment effects. We are assessing mechanisms and αPD-1 treatment effects (also FDA-approved for BC).

PD-1/PD-L1 provides a mechanism of immune escape, the blockage of which has reinvigorated interest in the treatment of urothelial cancer. We recently reported on cell-intrinsic PD-L1 effects in melanoma and ovarian cancer, and considered that effects in bladder cancer cells could differ based on distinct mutational landscapes. Using CRISPER/Cas9 methodology, we knockout PD-L1 in the murine bladder cancer cell line MB49 and the human bladder caner cell line RT. We show here that cell-intrinsic MB49 and RT4 cells express PD-L1 and PD-1 that each mediate cell-intrinsic signals. PD-L1 knockout has little effect on cells proliferation in vitro, but significant difference could be found in tumor size when we used MB49 Ctrl or MB49 PD-L1KO cells to challenge C57B16 mice. αPD-L1 and αPD-1 antibodies reduced bladder cancer cell proliferation in vitro demonstrating direct signaling effects. Bladder cancer cell-intrinsic PD-L1 promoted mTORC1 and tumor initiating cell generation similar to melanoma and ovarian cancer cells. By contrast to melanoma and ovarian cancer cells, bladder cancer cell PD-L1 promoted autophagy and had little effect on in vivo immune-independent growth. Base one the functional role, PD-L1 knockout increased proliferation inhibition induced by Rapamycin in both mouse and human bladder cancer cells. Furthermore, we found intrinsic PD-L1 suppress cytokine production, including CXCL10. Corresponding with increased CXCL10 secretion, more CXCR3+ NK cells could be attracted by PD-L1KO cancer cells. Overall, our findings further illustrated the intrinsic role of PD-L1 in bladder cancer development and found PD-1 expression on human bladder cancer cells for the first time.