Back

Metabolic-epigenetic coupling between leucine catabolism and glycolysis drives CDK4/6 inhibitor resistance

Oliphant, M. U.; Norton, K.; Kurmi, K.; Yang, H. H.; Joshi, S.; Lee, J.; Kozlova, N.; Muranen, T.; Clohessy, J. G.; Haigis, M. C.; Muthuswamy, S. K.

2026-09-24 cancer biology
10.64898/2026.09.23.752387 bioRxiv
Show abstract

Resistance to CDK4/6 inhibitors limits the durability of therapy for ER+ breast cancer. Despite the identification of mechanisms that regulate resistance, the metabolic adaptations that enable therapeutic escape remain poorly understood. Here, we identify a metabolic-epigenetic circuit that drives resistance by coordinately rewiring amino acid and glucose metabolism. CDK4/6 inhibitor-resistant ER+ tumor cells upregulate the leucine transporter SLC7A5, enhancing leucine uptake. SLC7A5 overexpression is sufficient to confer palbociclib resistance across ER+ cell lines, patient-derived organoids and xenografts. Stable isotope tracing in cell lines and in xenograft tumors revealed that leucine is catabolized through BCAT2 and HMGCL to increase acetyl-CoA levels, and elevated acetyl-CoA promotes H3K27 acetylation at the GLUT1 promoter, upregulating GLUT1 expression and glycolytic activity. Disrupting leucine transport, catabolism, or availability suppresses GLUT1 expression and restores therapeutic sensitivity in resistant models. In patients receiving palbociclib-based therapy, high SLC7A5 expression and coordinated SLC7A5-GLUT1 co-expression are associated with shorter progression-free survival. Together, these findings define a metabolic-epigenetic mechanism linking branched-chain amino acid catabolism to glycolysis and identify a biomarker-associated metabolic vulnerability in advanced ER+ breast cancer.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.