Investigating a novel function for phosphoserine aminotransferase 1 (PSAT1) in epidermal growth factor receptor (EGFR)-mediated lung tumorigenesis.

Abstract

Phosphoserine aminotransferase 1 (PSAT1) catalyzes the second enzymatic step within the serine synthetic pathway (SSP) and its expression is elevated in numerous human cancers, including non-small cell lung cancer (NSCLC). Epidermal growth factor receptor (EGFR) mutant NSCLC is characterized by activating mutations within its tyrosine kinase domain and accounts for 17% of lung adenocarcinomas. Although elevated SSP activity has been observed in EGFR-mutant lung cancer cells, the involvement of PSAT1 in EGFR-mediated oncogenesis is still unclear. Here, we explore a putative non-canonical function for PSAT1 using biochemical approaches to elucidate unknown interacting proteins and genomic RNA-seq profiling to identify cellular processes impacted by PSAT1. We further determined the cellular phenotypes affected by PSAT1 loss, which were verified by experimental rescue studies, including metabolite supplementation and restoration of protein expression/localization. Initially, we identified PKM2 as a novel PSAT1 associating protein. Although PSAT1 selectively induced the pyruvate kinase (PK) activity of recombinant PKM2, its loss in NSCLC cells did not alter cellular PK activity or expression of PKM2. However, fractionation studies revealed that PSAT1 localized to the nucleus and was required for EGFR-mediated nuclear PKM2 translocation. Phenotypically, PSAT1 loss led to a defect in EGFR-activated cell motility, which was partially restored by a nuclear expression of an acetyl-mimetic PKM2 mutant, but not wild-type vi PKM2 or metabolite supplementation. To get insight into cellular mechanisms downstream of PSAT1 activity, we conducted RNA-seq profiling. Consistent with the reported function of PSAT1, E2F targets and nucleotide metabolism genes were decreased upon PSAT1 silencing. Accordingly, the anchorage-independent growth was impacted by PSAT1 silencing and rescued by metabolite supplementation, but not by nuclear PKM2 expression. The correlation between decreased expression of actin-related genes and F-actin formation upon PSAT1 silencing suggested a role for PSAT1 in actin cytoskeleton rearrangements. Furthermore, identified PSAT1-associated gene signatures were predictive towards survival outcomes in EGFR-mutant NSCLC. Together, our data suggest multiple roles for PSAT1 in promoting EGFR-mutant NSCLC involving not only canonical SSP activity but also a non-canonical nuclear function through mediating protein localization. These findings have laid the foundation for future studies to fully define PSAT1’s response under EGFR-activation.