Targeting Epigenetic Heterogeneity to Improve Lung Cancer Immunotherapy Response

Eight of every ten patients with squamous lung cancer fail initial therapy1. Immunotherapies, particularly those that inhibit the PD1/PD-L1 interaction and drive T cells to recognize and kill tumor cells, have shown striking responses in a small subset of late stage, treatment refractory squamous lung tumors2-5. However, the majority of patients do not have a lasting response. Notwithstanding these promising clinical results, a need still exists to find ways to elicit better immunotherapy responses for more patients. PD-L1 expression appears to correlate with response, as does a three gene 'T effector cell' signature including PD-L1, CXCL9 and IFNy expression2-5. One way to influence gene expression within the tumor cells and within the immune microenvironment is with inhibitors of epigenetic complexes, including the Polycomb Repressive Complex 2 (PRC2). A combinatorial approach of primary therapies with epigenetic inhibitors has proven successful in many preclinical models and is just starting to be explored clinically6-9. Consistent with literature from developing lung10,11, I recently observed that inhibition of the PRC2 methyltransferase EZH2 drove a basal-like phenotype in lung cancers in vitro and in vivo12. Part of the basal transcriptional program is expression of the immune evasion molecule PD-L113,14. Furthermore, I previously demonstrated that basal-like tumor cells were enriched for tumor propagation in murine model of squamous cell carcinoma12,13. Together these findings suggest that squamous lung tumor propagating cells (TPCs) can be enriched by EZH2 inhibition and subsequently targeted by immunotherapy. Furthermore, my team and others recently identified neutrophils as the predominant immune cells in squamous lung cancers13,15,16, and evidence suggests that EZH2 inhibition will also target these cells17,18. Based on these findings, my central hypothesis is that EZH2 inhibitors will boost immunotherapy response in squamous tumors, both by increasing the PD-L1 expressing TPC population and by depleting immunosuppressive tumor associated neutrophils. The overarching goal of the proposed study is to validate combining EZH2 inhibition with PD1/PD-L1 targeted immunotherapy and learn the molecular mechanisms when the treatment is successful as well as when it is not successful. Three specific aims will test this hypothesis: Specific Aim 1: to determine if EZH2 inhibition changes the cellular hierarchies within squamous cell carcinomas to favor more PD-L1 expressing tumor propagating cells. I will accomplish this by using both mouse models and human patient derived organoid cultures, and serial orthotopic transplantation after EZH2 inhibitor treatment. I will also test if non-TPC cells can de-differentiate and continue to fuel tumor growth by sorting for presence or absence of the putative TPC cell surface marker nerve growth factor receptor (NGFR). The epigenetic and transcriptional consequences of EZH2 inhibition on squamous lung tumor cells will be assessed by ChIP-seq and RNA-seq. Specific Aim 2: to demonstrate the impact of EZH2 inhibitors on the tumor microenvironment of squamous lung cancer. Specifically, I will focus on the tumor associated neutrophils (TANs). I will compare TANs from placebo treated mice to those from EZH2 inhibitor treated mice for abundance, migration capacity and ability to suppress T cells. I will also use an EZH2 conditional knock-out mouse model to further characterize how EZH2 loss affects TANs. Again ChIP-seq and RNA-seq will be used to dissect the molecular changes driven by EZH2 inhibition in the neutrophil populations. Specific Aim 3: to test the combination of EZH2 inhibition and anti-PD1 in autochthonous mouse models and characterize clinically relevant changes in the tumors. I will treat immune-competent squamous lung cancer bearing mice with the EZH2 inhibitor GSK126 or EPZ-6438 (Tazemetostat)19, and the immunotherapy PD-1 antibody and follow tumor growth by magnetic resonance imaging. I will characterize tumor phenotypes in responders and non-responders and examine tumors that develop acquired resistance to this therapeutic drug combination. I have extensively characterized several mouse models that faithfully recapitulate the histology, gene expression and microenvironmental aspects of human squamous lung cancers12,13. I also have extensive experience using EZH2 inhibitors alone and in combinations with other therapies6,20. Completion of these aims will solidify the efficacy of a promising therapeutic combination and uncover mechanisms by which tumor hierarchies and microenvironments are changed by EZH2 inhibitors in squamous lung cancers. Given that one arm of a Phase 1/2 clinical trial combining EZH2 inhibition with anti-PDL1 just began recruiting late stage non- small cell lung cancer patients, learning the phenotypes and mechanisms of responders and non-responders will be extremely timely for any Phase 2/3 trials that ensue.