Abstract A059: MYC biomolecular condensates contribute to PDAC classical-to-basal transition and immune evasion

Abstract

MYC is one of the most ubiquitous amplified oncogenes in human cancers. Roughly 30% of pancreatic adenocarcinoma (PDAC) samples show MYC amplification, which is associated with increased aggressiveness and drug resistance. Our recent single cell analysis of 62 primary PDAC showed that MYC amplification promotes classical-to-basal transition and immune evasion. DNA-FISH, qPCR, and immunofluorescence experiments confirmed that MYC amplification upregulates basal-like genes such as TP63, while downregulates classical genes and immune regulatory genes, such as CDH1 and HLA, respectively. However, it is unclear how MYC distinctively regulates these genes oppositely beyond the traditional MYC-recognition sequences in the genome. Recently, MYCN, a c-MYC homolog, was identified to differentially regulate a subset of oncogenes and tumor suppressors through liquid-liquid phase separation (LLPS). The membrane-less biomolecular condensate formed through LLPS could serve as a biophysical barrier to enrich or exclude co-regulators, providing an extra layer of regulatory precision. C-MYC also possesses features reminiscent of other LLPS-prone proteins, including known domains promoting LLPS such as long intrinsically disordered regions; switch-like and dose dependent behavior, in which its levels do not increase linearly with increased copy numbers, but rather reach a threshold that enables its oncogenic phenotype. These features, together with our previous findings, led us to the hypothesis that MYC amplification results in higher nuclear protein levels above a critical concentration which leads to formation of MYC biomolecular condensates through LLPS. These then serve as transcription hubs by recruiting and sequestering co-regulators to drive PDAC classical-basal transition and immune evasion. Using live cell imaging and fluorescence recovery after photobleaching (FRAP), we show for the first time that the FP-tagged MYC protein forms liquid-like condensates in human PDAC cells, which recruit and sequester its binding partner MAX inside the condensates. MAX binding enhances MYC phase separation, resulting in increased condensate counts, but MAX did not phase separate alone, suggesting MYC is the driver protein in condensate formation and complex formation. Structural analysis and live-cell imaging identified two MYC truncation mutants that enhance or inhibit MYC condensate formation without affecting its DNA-binding domain. qPCR from PDAC cells expressing control, WT or individual mutants of MYC showed that LLPS is indispensable for MYC functions, as WT MYC specifically downregulated classical and immune genes such as GATA6 and HLA-A while upregulated basal-like genes such as TP63, whereas the LLPS-deficient mutant failed to significantly alter any genes, and the mutant with enhanced LLPS properties further exacerbated the effect. Given these findings, targeting MYC LLPS could be novel therapeutic strategy to be considered in MYC-amplified PDAC.