Defining the condensate landscape of fusion oncoproteins.

Abstract

Fusion oncoproteins (FOs) occur in ∼17% of cancers and are often drivers of oncogenesis through one of two molecular mechanisms. Many FOs promote oncogenesis by driving aberrant gene expression, while others drive aberrant cellular signaling. Notably, several FOs have recently been shown to promote their oncogenic mechanisms through formation of aberrant biomolecular condensates. To better understand the relationship between fusion oncoproteins and condensate formation, we compiled a database of 4,560 FO sequences (termed FOdb) as a resource for comprehensive studies. We then transfected HeLa cells with mEGFP-tagged forms of 166 FOs selected from the FOdb and found that 58% formed condensates. These studies show that FOs that formed condensates have distinct physicochemical features from those that displayed diffuse localization. We find that patterns of these sequence-derived features also correlate with sub-cellular localization of the FOs. Further, using conserved domain annotation, we found that FOs localized to the nucleus were predominantly associated with regulation of gene expression functional terms, while those localized to the cytoplasm are associated with cell signaling terms. We used this information to develop a machine learning model that can use a FOs sequence to accurately predict if it will form biomolecular condensates in cells. This model predicts 59% of the untested patient-derived FOs in the FOdb form cellular condensates, many of which have physicochemical features indicative of gene regulation and nuclear localization. Conversely, a substantial portion of the FOs predicted to have diffuse localization have sequence features that suggest cytoplasmic localization and cell signaling functions. We conclude that most FOs that drive aberrant gene transcription operate in part through condensate formation, and those that change cell signaling do so without forming condensates.