Modulators and effectors of anti-PD-1 responsiveness in cancer include T cell infiltration, an immune- suppressive microenvironment, tumor mutational burden, and tumor cell-intrinsic pathway (e.g., ?- catenin) alterations. Cell-surface PD-L1 level has been implicated as a baseline predictive marker of anti-PD-1 responsiveness, and cell-surface PD-L2 level may have predictive value independent of PD- L1. In melanoma, MAPK inhibitor therapy strongly induces PD-L1/L2 expression levels in tumor, stromal and immune cells, suggesting contributions to adaptive resistance. Induction of cell-surface PD-L1 is central to adaptive immune resistance and implicated as a mechanism of acquired anti-PD-1 resistance. Moreover, the regulation of surface PD-L1 protein stability has been implicated in tumor immune surveillance, where increased degradation augments tumor-specific T cell activity. We hypothesize that a better understanding of regulatory mechanisms controlling cell-surface PD- L1/L2 stability, clinical detection, and tumor cell-intrinsic pro-survival signaling could shed insights into melanoma immune evasion and therapeutic responsiveness. We will use proteomic approaches to dissect these processes and to explore the melanoma surface glycoproteome, the PD-L1/L2 interactome and the cytoplasmic signalosome in order to nominate mechanisms and/or markers of therapeutic responsiveness. We will interrogate iteratively clinical tumor samples and syngeneic mouse models of Braf, Nras and Nf1 mutant melanoma. These immune-competent models of melanoma are clinically relevant given their UV-induced high mutational burdens, dependence on CD8 T cells for therapeutic responses and capability for widespread metastases, including metastases to the brain. We will use cell-surface labeling of sialic acid-containing glycans to analyze the live cell surface glycoproteome, co-immunoprecipitation of PD-L1/L2 to enrich for interactomes, and APEX-based proteomic strategy to define in situ dynamic intracellular PD-L1/L2 neighborhood interactomes in response to PD-1 ligation or IFN? treatment. We will address what regulate PD-L1/L2 ubiquitination, recycling and degradation, how glycosylation affects membrane PD-L1 immunohistochemical detection, whether deglycosylation improves prediction of anti-PD-1 responses at baseline and on- treatment, and how novel cytoplasmic motifs of PD-L1/L2 mediate PD-1-dependent tumor cell pro- survival signaling. Using proximity ligation assays on clinical melanoma, we will test whether PD-L1/L2 tumor cell-intrinsic signaling reduces therapy efficacy. These proteomic approaches should advance our understanding of therapy response patterns in metastatic melanoma and nominate predictive biomarkers and combinatorial targets.
Mutation- and immune checkpoint-targeted therapies have improved survival of patients with metastatic melanoma, but drug resistance limits long-term survival. A common form of resistance is immunologic and may arise from the differential activities or regulations of PD-L1 and/or -L2 proteins on the surface of tumor cells. Various advanced proteomic techniqies will be employed to generate such knowledge, which can potentially help us select the most optimal therapies for specific patients as well as design new therapeutic strategies to reduce resistance.