SUMMARY. Radiation therapy (RT) and immunotherapy with PD-1 inhibitors are used in early and relapsed head and neck squamous cell cancers (HNSCC), respectively; RT failures are common, and most patients do not respond to PD-1 inhibition. Both impact the tumor microenvironment (TME), but studying the TME is limited by the availability of models with human immune cells. PI3K/mTOR signaling regulates protein synthesis through transcription factors such as Myc or SOX2, that dictate growth, invasion, and drug resistance in HNSCC. SVC112 is a fully synthetic small molecule that inhibits protein synthesis at the elongation step by inhibiting eEF2. SVC112 had greater effect on cancer over non-cancer cells, and was more potent and selective than homoharringtonin (HHT), an FDA-approved translation elongation inhibitor. Cancer cells had higher eEF2 than non-cancer cells, and the most susceptible strain had the highest eEF2 expression. SVC112 depleted SOX2, Myc, and Cyclin D1 in HNSCC cells at concentrations that had minimal effect on the rest of the proteome. SVC112 decreased spheres in vitro, reduced tumor growth in vivo in patient-derived xenografts (PDX), and induced tumor regression when given with RT in three out of four PDX models. SVC112 also reduced the E6 oncoprotein in human papillomavirus-driven cells. Lastly, SVC112 decreased PD-1 ligand (PD-L1) resulting in decreased PD-L1:PD-1 interactions leading to increased sphere T cell invasion, while having no effect on T cell fitness or function. Thus, SVC112 can target multiple key pathways (SOX2, Myc, PD-L1), and can influence the TME to reverse immune evasion. To test SVC112’s therapeutic potential in HNSCC, several unique tools will be used including 1) spheres containing cancer cells and T cells (iSpheres) that enable studying TME interactions and immunotherapy in vitro, 2) syngeneic mouse models of HNSCC, and 3) humanized mouse (HM) models, developed by the PI, that enable studying TME interactions and immunotherapy in vivo. The hypothesis that SVC112’s discriminating effect is due to selective depletion of key proteins will be tested by ribosome profiling (to identify mRNA targets) and proteomics analysis (to identify proteins targets). Then, we will test the mechanism of SVC112 synergy with RT, and if the modulation of the TME by SVC112 will further enhance RT efficacy. Lastly, the hypothesis that reduction of PD-L1 by SVC112 in HNSCC will alter the tumor-immune interaction will be tested by defining the effect of SVC112 on the expression of proteins critical for the TME, T cell tumor infiltration, and cancer-immune crosstalk. The effect of SVC112 on T cell fitness and function will be examined, aiding SVC112 translation. We will assess SVC112 and PD-1 inhibition in iSphere and HM models, and explore their efficacy when both are combined. Lastly, we will test the predictive value of eEF2 in SVC112 susceptibility, and we will assess eEF2 expression in human tumors, to enable identifying a compani...