Modular control of Pancreatic ductal adenocarcinoma (PDAC) progression by the substrate interaction domains of ERK2

Background: PDAC is a lethal disease with 5-year survival rate of just 13%, in desperate need of new therapeutic interventions. The primary driver of PDAC pathogenesis is the KRAS; however, current therapeutics targeting the activities of components of the KRAS/MAPK pathway fail to provide lasting benefit. We hypothesize that this results from active-site directed drugs attenuating the function of both substrate binding domains of ERK, which perform opposing roles in cancer pathogenesis. In myeloproliferative neoplasms, the ERK2-D domain promotes, and the ERK2-DBP domain impedes progression. Consequently, attenuating RAS/MAPK activity is akin to pressing both the gas and brake pedals on an automobile. Here we investigated the possibility that ERK signaling may also have cancer promoting and opposing roles in PDAC. Methods: Here, we seek to determine which ERK family member (ERK1 or ERK2) is the primary driver of PDAC pathogenesis and if the ERK D and DBP substrate domains play opposing roles. To do so, we utilize the KPC-PDAC mouse model and cross it to mice deficient in ERK1, ERK2, or an ERK2-DBP mutant to assess the impact on PDAC pathogenesis. We also utilized an ERK2 allelic series of isogenic KPC mouse cell lines (FC1245) to study the impact of ERK2-D and DBP mutants on pathogenesis in vitro and upon transplantation into mice. Results: We found that while ERK1 loss had no impact on PDAC pathogenesis, ablation of ERK2 markedly attenuated PDAC progression in KPC mice, indicating that ERK2 is primarily responsible for progression. Moreover, we found that inactivation of the ERK2-DBP domain was sufficient to delay PDAC progression to the same degree as ablation of ERK2. Attenuation of PDAC progression by ERK2-DBP mutation was associated with reduced expression and nuclear localization of c-Myc, a critical driver of PDAC progression. In addition to the cell autonomous effects of the ERK2-DBP mutation on PDAC, we also found differences in tumor-infiltrating immune cells, including fewer phenotypic myeloid derived suppressor cells and reduced expression of Tim3 on infiltrating T cells. The DBP mutation also attenuated colony formation in vitro, however, this was not observed for the ERK2-D domain, suggesting distinct functions in pathogenesis by the two ERK2 substrate domains. To investigate the therapeutic implications, we tested the efficacy of ERK2-DBP domain inhibitors to attenuate PDAC growth in vitro and in initial studies, determined that PDAC growth was substantially inhibited by DBP domain selective inhibitors. Conclusion: ERK2 DBP is the domain responsible for PDAC progression and its inactivation delays tumor development. We are now using generative modeling to identify DBP inhibitors that have superior efficacy and ability to co-modulate the cell-autonomous effects on PDAC growth and the cell extrinsic effects on the immune system, alone and in combination with immune interventions.