(PQ3) NOVEL TUMOR INTRINSIC PD-L1 SIGNALS DIRECT TUMOR IMMUNE CELL INFILTRATION

We respond to PQ3 with our data showing that tumor PD-L1 (CD274, B7-H1) is a major regulator of tumorinflammatory infiltrates. Our preliminary data show that melanoma PD-L1 regulates TIL through severalpreviously unknown tumor-intrinsic and extrinsic mechanisms. We define novel effects of tumor intrinsic PD-L1signaling on tumor proliferation, sensitivity to immune killing, in vivo growth independent of anti-tumorimmunity, and regulation of mTOR signals. We identified intracellular PD-L1, including those whose surfaceexpression is low or negative, and identified interactions with tumor PD-1. We hypothesize that melanomaintrinsic PD-L1-driven signals, particularly mTOR signals, alter tumor progression and treatmentresponses. The research team is comprised of tumor immunotherapy, tumor immunology and PD-L1 expertsat UTHSCSA and Dartmouth. We focus on melanoma for scientific reasons and based on our expertise.Aim 1 Define how tumor PD-L1 alters tumor immune infiltrates and immunotherapy responses. We usecontrol versus PD-L1lo (shRNA) B16 in a novel model to study differential treatment outcomes by tumor PD-L1status. We generated PD-L1KO B16 by CRISPR for highly detailed follow up mechanistic studies, and to assessif PD-L1 null status differentially affects treatment versus PD-L1lo. Effects will also be tested in transplantedBrafV600E mutated D4M melanoma (PD-L1+) engineered to be PD-L1lo and PD-L1KO, in mice with inducedBrafV600E melanomas, and in syngeneic skin grafts of skin from Braf/Pten versus PD-L1KO Braf/Pten mice.Aim 2 Test tumor PD-L1-driven mTOR signal effects on TIL and immunotherapy responses. We will testPD-L1 KO, PD-1 KO and double KO melanoma cells for mTOR signals, TIL and treatment effects. Cells will beengineered for defects in mTORC1/2 for mechanistic studies, complemented with mTOR inhibitor treatments.We will use engineered tumors that express cytoplasm-only versus cell surface-only PD-L1, to define novel,intracellular PD-L1 signals. Constructs with mutations in known PD-1 signal sites will be engineered into thesetumors for a complete understanding of PD-L1/PD-1 interactions.Aim 3 Define cell-intrinsic PD-L1 effects in human melanoma. We use well-defined human melanoma linesthat are basal PD-L1+ and/or PD-1+ and/or BrafV600E mutated. We will use human vectors to knock down orknock out PD-L1, PD-1 and mTORC1/2 genes. In vitro assessments of effects on proliferation, responses tomTOR inhibitors, ?PD-L1 and ?PD-1 will be assessed. In vivo effects in NSG mice will be assessed. Primaryhuman melanoma lines will be studied to complement data from long-term lines.