A systems biology approach has enabled the identification of several novel KRAS-dependent signaling pathways and proteins in cancer cell lines.
Although activating KRAS mutations occur frequently in cancer, targeting this oncogene has proved difficult. Three groups have now identified pathways that have not been previously linked to KRAS — all of which may provide pharmacologically tractable anticancer targets — and that are crucial in cancer cells dependent on activated KRAS.
Steve Elledge and colleagues conducted a genome-wide RNA interference (RNAi) screen in an isogenic pair of DLD-1 colorectal cancer cell lines: one that carried an endogenous activated KRAS allele (KRASWT/G13D cells) and one in which this allele was disrupted (KRASWT/- cells). By finding short hairpin RNAs (shRNAs) that were selectively depleted over time in KRASWT/G13D cells compared with KRASWT/- cells, and refining their initial list using a second isogenic pair of HCT116 colorectal cancer cells, they identified a functionally diverse list of 50 genes that might have synthetic lethal interactions with activated KRAS. This list included many genes involved in mitotic regulation, such as Polo-like kinase 1 (PLK1) and genes that encode anaphase-promoting complex/cyclosome (APC/C) subunits. Cells expressing KRASG13D were sensitive to PLK1 inhibition in vitro and in mouse xenograft models. The cells were also sensitive to inhibition of the ubiquitin ligase APC/C in vitro. As APC/C ultimately requires proteasome activity, the authors showed that the proteasome inhibitors MG132 and bortezomib were synthetic lethal with activated KRAS. Furthermore, non-small-cell lung cancer (NSCLC) cell lines were more sensitive to APC/C subunit knockdown if they harboured Ras mutations. Is this pathway relevant in vivo? Analysis of Ras-associated gene signatures in human lung tumour samples indicated that those tumours with a positive Ras signature that also had a signature suggestive of decreased APC/C activity correlated with increased patient survival.
Gary Gilliland, Bill Hahn and colleagues performed a similar RNAi screen across eight cancer cell lines (four that expressed a KRAS mutant and four that were KRAS wild type) and found that shRNAs targeting the kinase STK3 3 impaired cell viability and proliferation only in cells that were dependent on mutant KRAS. This was confirmed in several other cell types (25 in total), including both haematopoietic and epithelial cancer cell lines, and STK33 shRNA reduced tumour formation by 4 different KRAS-dependent epithelial cancer cell lines in immunocompromised mice. Interestingly, DLD-1 cells, as used in the studies above, were also sensitive to STK33 knockdown. Forcing KRAS dependence by expressing KRASG13D in two KRAS-independent acute myeloid leukaemia cell lines also led to sensitivity to STK33. What does STK33 do? The authors examined the activation of several signalling pathways in cells with STK33 suppression. Loss of STK33 decreased phosphorylation and activity of the S6K1 kinase and downstream phosphorylation of the proapoptotic BH3-only protein BAD, leading to apoptosis.
Jeff Settleman and colleagues examined the effect of KRAS depletion in several lung and pancreatic cancer cell lines. By comparing KRAS-dependent cells (in which KRAS knockdown induced apoptosis) with KRAS-independent cells, they noted that KRAS-dependent cells tended to have an epithelial morphology, whereas KRAS-independent cells appeared to be more mesenchymal. Induction of an epithelial–mesenchymal transition (EMT) in KRAS-dependent H358 NSCLC cells allowed these cells to become KRAS independent, and conversely, reversal of EMT in KRAS-independent cells rendered these cells KRAS dependent. The authors then derived a gene expression signature of KRAS dependency, which was also associated with well-differentiated primary lung tumours. From this signature, they found three pharmacologically tractable targets that were upregulated in KRAS-dependent cells: the kinase SYK, the RON receptor tyrosine kinase (encoded by MST1R) and 6 integrin (encoded by ITGB6). shRNAs against SYK and ITGB6 led to EMT and apoptosis in two KRAS-dependent cell lines, but not two KRAS-independent cell lines, whereas MST1R knockdown was context specific, inducing apoptosis in only one KRAS-dependent cell line. KRAS-dependent cells were also more sensitive to a pharmacological SYK inhibitor.
Sarah Seton-Rogers
References
Luo, J. et al. A genome-wide RNAi screen identifies multiple synthetic lethal interactions with the Ras oncogene. Cell137, 835–848 (2009) | Article | PubMed |
Scholl, C. et al. Synthetic lethal interaction between oncogenic KRAS dependency and STK33 suppression in human cancer cells. Cell137, 821–833 (2009) | Article | PubMed |
Singh, A. et al. A gene expression signature associated with "K-Ras addiction" reveals regulators of EMT and tumor cell survival. Cancer Cell15, 489–500 (2009) | Article | PubMed |
6 integrin (encoded by 