A protein that stabilizes cancer-suppressing mRNAs turns out to be a key brake on breast cancer metastasis. A paper from Core Investigator @genophoria and lab in @ScienceAdvances traces the mechanism to RNA decay and an RNA-binding protein most studies have overlooked.

Gene expression in tumors is typically attributed to genetic and epigenetic changes. But when the team profiled RNA decay rates across six breast cancer cell lines, transcription alone left a large share of expression variability unexplained. That gap pointed to mRNA stability.

To map the regulators behind it, the team developed GreyHound, a deep learning model that predicts transcript stability from RNA sequence features and RNA-binding protein (RBP) expression simultaneously across cell types.

The top predicted regulator was RBMS3. CLIP-seq confirmed direct binding at AUA-rich 3'UTR elements; reporter assays showed those sites are necessary and sufficient; SLAM-seq confirmed the effect is post-transcriptional, not transcriptional.

Functionally, RBMS3 loss increased metastatic lung colonization in two independent xenograft models. An in vivo CRISPRi screen then identified TXNIP, a tumor suppressor, as the key effector. Epistasis experiments confirmed RBMS3 acts upstream of TXNIP to suppress metastasis.

The Goodarzi Lab is hiring a joint postdoctoral fellow. See Arc open roles: