Cross-talk between EphA2 and BRaf/CRaf is a key determinant of response to Dasatinib

Abstract

Purpose: EphA2 is an attractive therapeutic target because of its diverse roles in cancer growth and progression. Dasatinib is a multikinase inhibitor that targets EphA2 and other kinases. However, reliable predictive markers and a better understanding of the mechanisms of response to this agent are needed.

Experimental design: The effects of dasatinib on human uterine cancer cell lines were examined using a series of in vitro experiments, including MTT, Western blot analysis, and plasmid transfection. In vivo, an orthotopic mouse model of uterine cancer was utilized to identify the biologic effects of dasatinib. Molecular markers for response prediction and the mechanisms relevant to response to dasatinib were identified by using reverse phase protein array (RPPA), immunoprecipitation, and double immunofluorescence staining.

Results: We show that high levels of CAV-1, EphA2 phosphorylation at S897, and the status of PTEN are key determinants of dasatinib response in uterine carcinoma. A set of markers essential for dasatinib response was also identified and includes CRaf, pCRaf(S338), pMAPK(T202/Y204) (mitogen-activated protein kinase [MAPK] pathway), pS6(S240/244), p70S6k(T389) (mTOR pathway), and pAKT(S473). A novel mechanism for response was discovered whereby high expression level of CAV-1 at the plasma membrane disrupts the BRaf/CRaf heterodimer and thus inhibits the activation of MAPK pathway during dasatinib treatment.

Conclusions: Our in vitro and in vivo results provide a new understanding of EphA2 targeting by dasatinib and identify key predictors of therapeutic response. These findings have implications for ongoing dasatinib-based clinical trials.

Figure 1. In vitro effects of dasatinib in uterine cancer cell lines

(a) Expression level of EphA2 and CAV-1 and expression of wild-type or mutant RAS and PTEN for each cell line. −, No expression; +, low expression; +++, high expression (left). Median inhibitory concentration (IC₅₀) after treatment with dasatinib (middle). Cell viability after treatment with dasatinib at 0, 10, 50, 100, 150, 1,000, 1,500, or 10,000 nM for 72 hours (right). Data represent means of triplicate measurements with error bars to represent SEM. (b) Immunoblot analysis of proteins associated with the Src/FAK/EphA2 pathway in cells treated with (+) or without (−) dasatinib at 100 nM for 16 hours (left). Quantification of band intensity relative to actin intensity is shown graphically. Black bars, no treatment. White bars, treatment (right). (c) Immunoprecipitation (IP) and Western blot(WB) analysis of tyrosine-phosphorylated EphA2 in HEC1-A, Ishikawa, and SKUT-2 cells treated with (+) or without (−) dasatinib at 100 nM for 16 hours. (d) Effect of wild-type (WT) EphA2 and pEphA2^S897 status on the sensitivity of SKUT-2 cells to dasatinib at 0, 10, 50, 100, 150, 1,000, 1,500, or 10,000 nM for 72 hours. Cell viability assay was performed with SKUT-2 cells stably transfected with Myc-DDK-tagged plasmid with WT EphA2, inactivated pEphA2^S897, or constitutively activated pEphA2^S897 (left). Median inhibitory concentration (IC₅₀) after treatment with dasatinib(right, top) at 100 nM for 16 hours. Western blotresults showing status of EphA2 and pEphA2^S897 in SKUT-2 cells after transfection. Anti-Myc antibody was used as a marker for efficiency of transfection. β-Actin was used as a loading control(right, bottom). EV, empty vector.

Figure 2. Expression of multiple proteins associated With Src/FAK/EphA2 and RAS/RAF/MAPK signaling pathways as detected with reverse-phase protein array

(a) Heatmap of molecules whose expression significantly (p<0.05) differed before and after treatment with dasatinib (100 nM), paclitaxel (5 nM), or both in SPEC-2 cells (top) and SKUT-2 cells (bottom) cells for 16 hours. (b and c) Normalized expression levels in SPEC-2 cells (b) and SKUT-2 cells (c) treated with (red) or without (blue) dasatinib. Data represent means of triplicate measurements.

Figure 3. Effect of casatinib treatment on protein expression of pS6^S240/244, CAV-1, BRaf and CRaf in uterine cancer cells

(a and b) Western blot (WB) analysis of the expression levels of pS6^S240/244 and CAV-1 (a) and BRaf and CRaf (b) in uterine cancer cells treated with (+) or without (−) dasatinib at 100 nM for 16 hours (top). Densitometry was performed to objectively assess potential differences (bottom). (c) Immunoprecipitation (IP) and WB analysis of the interaction of CAV-1, BRaf and EphA2 in SKUT-2 cells treated with (+) or without (−) dasatinib for 8 hours. (d) Immunoprecipitation and WB analysis of the interaction of CRaf and BRaf in Ishikawa, SKUT-2, and HEC1-A cells treated with (+) or without (−) dasatinib for 8 hours.

Figure 4. Effect of dasatinib on direct binding between EphA2 and CAV-1, EphA2 and BRaf, or CAV-1 and BRaf in uterine cancer cells

(a) Double Immunofluorescence (IF) staining analysis of distribution of CAV-1 (red)/BRaf (green) (left), CAV-1 (red)/EphA2 (green) and of EphA2 (red)/BRaf (green) in SKUT-2 cells (right), before and after dasatinib treatment for 8 hours. Nucleus staining by Sytox (blue). (b) Dynamic effect of dasatinib on distribution of BRaf (green) and CAV-1 (red) in SKUT-2 cells for the indicated time periods (left). IF staining analysis of the effect of CAV-1 silencing on distribution of BRaf (green) and CAV-1 (red) in SKUT-2 (right panel, left) or SPEC2 cells treated with or without dasatinib for 8 hours (right panel, right). Nucleus staining by Sytox (blue). (c) MTT analysis of dasatinib on SKUT-2 cells transfected with CAV-1 siRNA or control siRNA (bottom). Western blotresults showing the effect of CAV-1 silencing with CAV-1 siRNA in SKUT-2 cells (top). (d) PLA analysis of distribution of CAV-1 and BRaf (top, left) or EphA2 (bottom, left) in SKUT-2 and HEC1-A cells grown on small chambers, before and after dasatinib treatment for 8 hours. Each individual blob (red) represents CAV-1 in close proximity with BRaf or EphA2. Statitical analysis of the signals using the Blobfinder V3.2 software (right).

Figure 5. In vivo effects of therapy with dasatinib and paclitaxel in uterine cancer models

(a and b) In vivo effect of dasatinib (15 mg/kg oral, daily), paclitaxel (100 μg in 200 μL of PBS intraperitoneally, weekly), or both in an EphA2-positive model (SKUT-2) (a) and an EphA2-negative model (SPEC-2) (b). Error bars indicate SEM. *p<0.05, **p<0.01, and ***p<0.001. (c and d) Immunohistochemical staining showing the effect of dasatinib, paclitaxel, or both on uterine cancer cells angiogenesis (CD31), proliferation (Ki67), and apoptosis (caspase 3) in the SKUT-2 model (c) and the SPEC-2 model (d). Immunohistochemical staining of pEphA2^S897 expression in SKUT-2 cells (c, bottom). *p<0.05, **p<0.01, and ***p<0.001 compared with control group. Original magnification ×100 or ×200.

Figure 6. Expression of predictive markers in uterine cancer patients

(a) Representative pictures of immunohistochemical staining of CAV-1, pAKT^S473, pEphA2^Y594, and pEphA2^S897 expression in normal ueterine samples and uterine cancer samples. Original magnification ×200. (b) Cartoon for EphA2 signaling pathway and its interaction with CAV-1 and BRaf in dasatinib-sensitive uterine cancer cells. Dasatinib drives BRaf and EphA2 to high level CAV-1 at the plasma membrane, resulting in disruption of the BRaf/CRaf heterodimer and thus downregulating MAPK signaling.