TRIM28 inhibition as a synthetic lethal strategy for small cell lung cancers with NOTCH2-inactivation

The National Cancer Institute awarded Dana-Farber Cancer Institute, Inc. $609,130 on September 1, 2026, under the Cancer Treatment Research program (CFDA 93.395) to investigate TRIM28 inhibition as a synthetic lethal strategy for small cell lung cancers with NOTCH2-inactivation.

The research will use syngeneic immunocompetent mouse models of small cell lung cancer to test whether TRIM28 inhibition restores anti-tumor immune responses in NOTCH2-mutant tumors. The work builds on prior CRISPR/Cas9 screening that identified TRIM28 as a synthetic lethal interactor with NOTCH2 loss, showing that TRIM28 inactivation restores high type I interferon signaling selectively in NOTCH2-inactivated small cell lung cancer cells through activation of the STING/MAVS/TBK1 pathway. The project will examine whether this effect can be further exploited alone or in combination with immunotherapy agents such as anti-PD-L1 or tarlatamab.

Performance will take place in Boston, Massachusetts, with a period of performance from September 1, 2026, through August 31, 2031. The assistance type is a Project Grant.

Abstract

Small cell lung cancer (SCLC) remains one of the most treatment-refractory types of lung cancer. Despite its strong association with smoking and high TMB, the clinical impact of immune checkpoint blockade (ICB) has been marginal likely because most SCLC tumors downregulate MHC class I antigen presentation and expression of other genes required for immunogenicity and thus are incapable of being recognized by CD8 T cells. In line with this, recent work has identified a subset (~10%) of SCLC patients with long-term durable responses to ICB and high expression of MHC-I antigen presentation machinery was the most significant predictor for durable ICB responses in SCLC patients. ~25% of sclcs harbor LOF mutations in NOTCH1-4, while the vast majority of sclcs (60-70%) inherently repress NOTCH1-4 receptor expression. We performed crispr/CAS9 screens to identify synthetic lethal interactors with LOF NOTCH1 or NOTCH2 mutations (~15-20% of sclcs) and identified TRIM28 as a robust synthetic lethal interactor with NOTCH2 loss. The NOTCH2-TRIM28 synthetic lethal interaction validated in both human and mouse NOTCH2-isogenic SCLC cell lines. TRIM28 inactivation restored an "inflammatory" cell state with high type I interferon signaling selectively in NOTCH2-mutant SCLC cells. This "inflammatory" state causes apoptosis and is attenuated by inactivating sting or mavs viral sensing or by blocking TBK1 or JAK kinases with small molecule inhibitors. Therefore, we hypothesize that NOTCH2-TRIM28 synthetic lethality is due to their cooperative loss activating the sting/mavs/TBK1 type I interferon pathway, which will promote anti-tumor immune responses and hence can be further exploited in an immunocompetent host alone or when combined with immunotherapy (PD-L1 or tarlatamab). In aim 1, we will use syngeneic immunocompetent mouse models of SCLC to determine whether the synthetic lethal interaction between NOTCH2 and TRIM28 can be further exploited in an immunocompetent host alone or when combined with PD-L1 or tarlatamab. In aim 2, we will develop and validate small molecules to pharmacologically inhibit TRIM28 or its binding partners ZNF219 or SETDB1. This proposal is significant as IT will uncover the mechanistic basis and provide preclinical efficacy data to support targeting a novel synthetic lethal interaction as a new druggable therapeutic strategy for sclcs with LOF NOTCH2 mutations, which may extend to sclcs with low NOTCH2 expression. IT is innovative as we will develop new small molecule inhibitors to block TRIM28/krab-ZNF/SETDB1 function as a tractable therapeutic strategy for NOTCH2-mutant sclcs or SCLC with low NOTCH2 expression. Syngeneic immunocompetent mouse models of SCLC are required to determine whether targeting TRIM28 promotes anti-tumor immunity and enhances responses to immunotherapy in the context of a fully intact immune system. Current human models cannot fully recapitulate the tumor-immune interactions required to evaluate immune-mediated therapeutic efficacy.