The response to PAK1 inhibitor IPA3 distinguishes between cancer cells with mutations in BRAF and Ras oncogenes

Abstract

While new drugs aimed at BRAF-mutated cancers are entering clinical practice, cells and tumors with activating Ras mutations are relatively resistant to those and quite a few other anti-cancer agents. This inspires the effort to reverse this resistance or to uncover new vulnerabilities in such resistant cancers. IPA3 has been originally identified as a small molecule inhibitor of p21-activated protein kinase 1 (PAK1), a candidate therapeutic target in human malignancies. We have tested a battery of melanoma and colon carcinoma cell lines that carry mutations in BRAF, NRAS and KRAS genes and have observed that those with NRAS and KRAS mutations are more sensitive to killing by IPA3. Genetic manipulations suggest that the differential response depends not just on these oncogenes, but also on additional events that were co-selected during tumor evolution. Furthermore, sublethal doses of IPA3 or ectopic expression of dominant-negative PAK1 sensitized Ras-mutated cells to GDC-0897 and AZD6244, which otherwise have reduced efficiency against cells with activated Ras. Dominant-negative PAK1 also reduced the growth of NRAS-mutated cells in confluent cultures, but, unlike IPA3, caused no significant toxicity. Although it remains to be proven that all the effects of IPA3 are exclusively due to inhibition of PAK1, our findings point to the existence of selective vulnerabilities, which are associated with Ras mutations and could be useful for better understanding and treatment of a large subset of tumors.

Figure 1. Differential sensitivity of BRAF- and RAS-mutated cells to IPA3

Indicated cell lines (mutated oncogenes are shown in parentheses) were plated at the same density (30000 cells/well) in 12-well plates and next day treated with various concentrations of IPA3. Two days later, images of the remaining cells were taken (A), and the plates were fixed and the cells were quantified (B and C) by methylene blue staining/extraction method as described in Methods. The number of remaining viable cells is displayed relative to the number of cells in parallel untreated cultures of the same cell line. Each data point was collected in triplicates, and the standard deviations are denoted by error-bars.

Figure 2. The effects of IPA3 and GDC-0897 on NRAS-mutated melanoma cell lines

SK-MEL-103 (A) and SK-MEL-147 (B), transduced either with a BRAF-V600E – expressing construct (“BRAF”) or with the corresponding empty vector (“vector”), were plated at 30000 cells/well in 12-well plates and treated next day with the indicated doses of IPA3 and GDC-0897. 72 hours later, the remaining viable cells were fixed and quantified as in Figure 1.

Figure 3. The effect of NRAS on response of A375 cells to IPA3 and GDC-0897

Melanoma cell lines A375 (BRAF-mutated) was transduced with a construct expressing activated NRAS (“NRAS”) or the respective vector control (“vector”). The cells were plated in 12-well plates (30000/cells per well) and treated next day with indicated doses of GDC-0897 (A) or IPA3 (B). Three days later, the remaining viable cells were quantified as in Figure 1.

Figure 4. The effect of dominant-negative PAK1 on response of A375 cells to GDC-0897, AZD6244 and IPA3

Melanoma cell lines SK-MEL-103 (NRAS-mutated) was transduced with a construct expressing dominant-negative PAK1 (“PAK1-K299R”) or the respective vector control (“vector”). The cells were plated in 24-well plates (30000/cells per well) and treated with indicated doses of GDC-0897 (A), AZD6244 (B) or IPA3 (C). Three days later, the remaining viable cells were quantified as in Figure 1.

Figure 5. The effect of dominant-negative PAK1 on growth of SK-MEL-103 cells in confluent cultures

A. SK-MEL-103 cells transduced with a construct expressing dominant-negative PAK1-K299R (“dnPAK”) or the respective vector control (“vector”) were cultured for two days after reaching confluence. Three wells of each variant were fixed and stained with Hoechst 33342, and readings were taken from 25 points on each of the wells. The average value (in arbitrary units) of all the readings for each cell line is shown with standard deviation. B-C. The indicated cell lines were cultured for six more days after reaching confluence, fixed and stained with Hoechst 33342 and photographed under a fluorescent microscope. Note the change from uniform multilayer growth to individual foci.