One of the biggest hurdles for the development of metabolism-targeted therapies is to identify the responsive tumor subsets. However, the metabolic vulnerabilities for most human cancers remain unclear. Establishing the link between metabolic signatures and the oncogenic alterations of receptor tyrosine kinases (RTK), the most well-defined cancer genotypes, may precisely direct metabolic intervention to a broad patient population. By integrating metabolomics and transcriptomics, we herein show that oncogenic RTK activation causes distinct metabolic preference. Specifically, EGFR activation branches glycolysis to the serine synthesis for nucleotide biosynthesis and redox homeostasis, whereas FGFR activation recycles lactate to fuel oxidative phosphorylation for energy generation. Genetic alterations of EGFR and FGFR stratify the responsive tumors to pharmacological inhibitors that target serine synthesis and lactate fluxes, respectively. Together, this study provides the molecular link between cancer genotypes and metabolic dependency, providing basis for patient stratification in metabolism-targeted therapies.

It remains a big challenge to develop HDAC inhibitors effective for solid tumors. Previous studies have suggested that the feedback activation of JAK-STAT3 pathway represents a key mechanism leading to resistance to HDAC inhibitors in breast cancer, suggesting the therapeutic promise of JAK/HDAC dual inhibitors. In this work, we discovered a series of pyrrolo [2,3-d]pyrimidine-based derivatives as potent JAK and HDAC dual inhibitors. Especially, compounds 15d and 15h potently inhibited JAK1/2/3 and HDAC1/6 and displayed antiproliferative and proapoptotic activities in triple-negative breast cancer cell lines. Besides, compounds 15d and 15h also diminished the activation of LIFR-JAK-STAT signaling triggered by tumorassociated fibroblasts, which suggests that these compounds could potentially overcome the drug resistance caused by the tumor microenvironment. More importantly, compound 15d effectively inhibited the tumor growth in MDA-MB-231 xenograft tumor model. Overall, this work provides valuable leads and novel antitumor mechanisms for the treatment of the SAHA-resistant triple-negative breast cancers.

Highlights d A proteomic and phosphoproteomic landscape of 43 KRAS mutant cancer cell lines d Multi-omic analysis reveals three subsets with distinct characteristics d Compensatory phosphoprotein biomarker analysis identifies new drug combinations d DOT1L and SHP2 inhibition synergistically treats a subset of KRAS mutant cancer