定义KRAS突变癌症中的KRAS和ERK依赖性转录组

Recent US Food and Drug Administration approval of direct inhibitors of G12C (Gly12→Cys) mutations in the formerly “undruggable” KRAS marks an important milestone in cancer drug discovery, and inhibitors of more prevalent KRAS mutations [G12D/V (Gly12→Asp or Val), and others] are now in clinical evaluation. However, notably few patients respond initially, and most of those individuals relapse quickly. Defining genetic markers and drivers of primary and treatment-associated acquired resistance to KRAS inhibitors will be essential to achieve broader and more durable responses.

A major molecular output of aberrant KRAS activation involves systemwide deregulation of gene transcription. Despite numerous efforts to establish KRAS-associated gene transcription signatures, present signatures show notably limited overlap, likely reflecting divergent experimental strategies and cancer models. In this work, we sought to define a comprehensive KRAS-dependent transcriptional signature that detects target inhibition in KRAS-mutant cancer patients treated with KRAS mutation–selective inhibitors.

Most of the previous KRAS signatures, including the present gold-standard Hallmark KRAS signaling gene sets, profiled gene expression changes caused by persistent steady-state expression of mutant KRAS. Instead, we applied RNA sequencing (RNA-seq) to determine transcriptional changes caused by acute KRAS suppression in endogenously KRAS-mutant pancreatic ductal adenocarcinoma (PDAC) cell lines, thereby limiting the confounding effects of compensation for the loss of KRAS signaling. In contrast to the Hallmark, our KRAS gene signature was strongly enriched in changes in response to pharmacologic inhibition of mutant KRAS in KRAS-mutant PDAC cell lines and tumors, as well as in lung and colorectal xenograft tumors. Thus, the KRAS-regulated transcriptome may be broadly applicable. Despite the plethora of validated and putative KRAS effectors, we found that RAF-MEK-ERK mitogen-activated protein kinase (MAPK) effector signaling alone, but not the PI3K-AKT-mTORC1 pathway, showed that sufficient to support mutant KRAS-dependent PDAC growth. Consistent with this, the KRAS-regulated transcriptome was driven largely through ERK MAPK activity. Pathway analyses showed that our merged KRAS- and ERK-dependent gene signature composed of 278 up-regulated genes was highly enriched in cell cycle processes. A comparison of the ERK-regulated transcriptome and total proteome showed that ~80% of the regulation of protein expression changes was at the level of gene transcription. Another subset was at the level of posttranscriptional mechanisms, including ERK phosphorylation and modulation of the anaphase promoting complex/cyclosome (APC/C), which is involved in cell cycle regulation. Finally, our KRAS-ERK gene signature accurately detected KRAS-ERK target inhibition, and that inhibition correlated generally with clinical responses in KRAS-mutant cancer patients treated with KRAS or ERK inhibitors. An accurate portrait of the molecular output that mirrors aberrant KRAS signaling in cancer patients will further elucidate mechanistically how KRAS drives cancer.

Our study established a KRAS-ERK regulated gene signature that detected KRAS-ERK inhibition in KRAS-mutant cancer patients. Coupled with our ERK-dependent total proteome and phosphoproteome signatures, it revealed that aberrant KRAS signaling drives cancer growth through the regulation of cell cycle progression at multiple levels. The KRAS-ERK transcriptome may define molecular markers for primary and acquired resistance in patients treated with KRAS-ERK MAPK– targeted therapies.