Development of Combination Strategies for Focal Adhesion Kinase Inhibition in Diffuse Gastric Cancer

Abstract

Purpose: Diffuse gastric cancer (DGC) is an aggressive and frequently lethal subtype of gastric cancer. Because DGC often lacks genomic aberrations that indicate clear candidate therapeutic targets, it has been challenging to develop targeted therapies for this gastric cancer subtype. Our previous study highlighted the contribution of focal adhesion kinase (FAK) in the tumorigenesis of DGC and the potential efficacy of small-molecule FAK inhibitors. However, drug resistance to monotherapy often hinders the efficacy of treatment.

Experimental design: We generated a genome-scale library of open reading frames (ORF) in the DGC model of Cdh1-/-RHOAY42C/+ organoids to identify candidate mechanisms of resistance to FAK inhibition. Compensatory activated pathways were also detected following treatment with FAK inhibitors. Candidates were investigated by cotargeting in vitro and in vivo experiments using DGC.

Results: We found that cyclin-dependent kinase 6 (CDK6) promoted FAK inhibitor resistance in ORF screen. In addition, FAK inhibitor treatment in DGC models led to compensatory MAPK pathway activation. Small-molecule CDK4/6 inhibitors or MAPK inhibitors effectively enhanced FAK inhibitor efficacy in vitro and in vivo.

Conclusions: Our data suggest that FAK inhibitors combined with MAPK inhibitors or CDK4/6 inhibitors warrant further testing in clinical trials for DGC.

Figure 1.. Genome-scale lentiviral ORFeome library screen identifies drivers of FAK inhibitor resistance in DGC.

(A) Schematic description of the genome-scale ORFeome library screen. (B, C) Scatter plots presenting the z-scores of average log₂(fold change) for defactinib vs. ETP (y-axis) and for DMSO vs. ETP (x-axis) (B); PF-573228 vs. ETP (y-axis) and DMSO vs. ETP (x-axis) (C) in Cdh1^−/−RHOA^Y42C/+ organoids. Z-scores of DMSO vs. ETP < 3 nominate genes not associated with enhanced growth in the DMSO, whereas z-scores of defactinib or PF-573228 vs. ETP ≥ 3 nominate genes associated with resistance to defactinib or PF-573228. Genes with z-scores < 3 for DMSO and ≥ 3 for defactinib or PF-573228 were nominated as candidate genes conferring resistance and classified as significant ORFs. (D) Scatter plots presenting the z-scores of log₂(fold change) for defactinib vs. ETP (x-axis) and for PF-573228 vs. ETP (y-axis) in Cdh1^−/−RHOA^Y42C/+ organoids. (E) Immunoblot analysis to validate CDK6 overexpression in SNU668. (F) In vitro proliferation of control or CDK6 overexpressed SNU668 treated with 2.5 μM defactinib for indicated days. The results are the representative of three independent experiments, each done in quadruplicate. Data are mean ± S.D. **P<0.01, two-way ANOVA test. (G) Representative images of control or CDK6 overexpressed SNU668 treated with DMSO or 2.5 μM defactinib. Scale bar, 100 μm.

Figure 2.. CDK4/6 inhibitor enhanced the efficacy of FAK inhibitors.

(A) Immunoblot analysis of genes involved in FAK pathway and cell-cycle pathway in Cdh1^−/−RHOA^Y42C/+ organoids treated with 1 μM defactinib (def), 0.5 μM palbociclib (palbo), the combination or with DMSO control. Protein lysates were collected after drug treatment for 24 hours and 48 hours. Immunoblots from one representative experiment (n=2) are shown. (B) The frequency of G0/G1 cells of Cdh1^−/−RHOA^Y42C/+ organoids treated with 24 hours of 1 μM defactinib, 0.5 μM palbociclib, the combination or with DMSO control. Following treatment, the cells were harvested, stained with propidium iodide, and assayed with flow cytometry (n=3). Data are mean ± S.D, *P<0.05, two-way ANOVA test. (C) Representative phase contrast images (left) and HE staining images (right) of Cdh1^−/−RHOA^Y42C/+ organoids treated with 48 hours of 1 μM defactinib, 0.5 μM palbociclib, the combination or with DMSO control. Scale bar, 50 μm. (D) The plots show the relative cell amount of Cdh1^−/RHOA^Y42C/+ organoids treated with 48 hours of 1 μM defactinib, 0.5 μM palbociclib, the combination or with DMSO control (n=3). Cell amount in each group was normalized to that in DMSO control group. Data are mean ± S.D, **P<0.01, ***P<0.001, two-way ANOVA test. (E) Immunoblot analysis of genes involved in FAK pathway and cell-cycle pathway in NUGC4 (left) and SNU668 (right) treated with 2.5 μM defactinib, 0.5 μM palbociclib, the combination or with DMSO control. Protein lysates were collected after drug treatment for 24 hours and 48 hours. Immunoblots from one representative experiment (n=2) are shown. (F) The frequency of G0/G1 cells of NUGC4 and SNU668 treated with 24 hours and 48 hours of 2.5 μM defactinib, 0.5 μM palbociclib, the combination or with DMSO control (n=3). Data are mean ± SD. *P<0.05, ***P<0.001, ****P<0.0001, two-way ANOVA test. (G) Images show representative results of colony formation assays of NUGC4 (left) and SNU668 (right). Cells were cultured in 6-well plates and treated with DMSO, indicated concentration of defactinib or palbociclib alone or together for 7 to 10 days, and then fixed and stained with crystal violet solution. All experiments were repeated at least twice. (H) Top: growth curve for NUGC4 xenograft tumors (n=6–10) treated with vehicle control, VS-4718 (50 mg/kg, bid), palbociclib (50 mg/kg, qd) or the combination. Treatment began on day1. Data are mean ± SEM. *P<0.05, **P<0.01, two-way ANOVA test. Bottom: waterfall plot showing the tumor volume change (at day 22) relative to baseline volume (at day 1). Each bar represents one xenograft tumor. (I) Representative images of Ki-67, TUNEL and pERK1/2 staining of the NUGC4 xenograft tumors. Scale bar = 100 μm.

Figure 3.. FAK inhibitors induce MAPK activation in DGC organoids and cell lines.

(A) Heat map representation of selected antibodies from Digiwest analysis of Cdh1^−/−RHOA^Y42C/+ organoids treated with DMSO, 2.5 μM defactinib or 2.5 μM PF573228 for 48 hours. Log₂fold change of the target signal in treatment group vs DMSO group in each replicate were used in this heatmap. (B, C, D) Immunoblot analysis of Cdh1^−/−RHOA^Y42C/+ organoids (B) and NUGC4, SNU668 cell lines (C, D) treated with defactinib or PF-573228 for indicated time. Immunoblots from one representative experiment (n=2) are shown.

Figure 4.. MAPK inhibitor and FAK inhibitor have synergistic effect in DGC models.

(A) Immunoblot analysis of Cdh1^−/−RHOA^Y42C/+ organoids treated with defactinib, VS-6766, the combination or DMSO control for 24 hours. Immunoblots from one representative experiment (n=2) are shown. (B) Representative phase contrast images (left) and HE staining images (right) of Cdh1^−/−RHOA^Y42C/+ organoids treated for 48 hours with 2.5 μM defactinib, 0.5 μM VS-6766, the combination or with DMSO control. Scale bar, 50 μm. (C) Cdh1^−/−RHOA^Y42C/+ organoids were treated with defactinib (0.25 μM to 4 μM) or VS-6766 (0.125 μM to 2 μM) alone or together for 3 days. Viability in the treatment groups was normalized to DMSO control. The inhibition rate was shown (left). Analysis of synergistic effect in defactinib and VS6766 combination was performed by SynergyFinder using Zero Interaction Potency (ZIP) model. The inhibition rate was used to calculate ZIP synergy score. The box indicates the most synergistic area (right). Representative of two independent experiments were shown. (D) Immunoblot analysis of BL62 (left) and DE66 (right) treated with defactinib, VS-6766, the combination or DMSO control for 5 days. Immunoblots from one representative experiment (n=2) are shown. (E) Representative phase contrast images of BL62 (top) and DE66 (bottom) treated for 5 days with 2 μM defactinib, 1 μM VS-6766, the combination or with DMSO control. Scale bar, 50 μm. (F) Immunoblot analysis of NUGC4 (left) and SNU668 (right) treated with defactinib, VS-6766, the combination or DMSO control for 48 hours. Immunoblots from one representative experiment (n=2) are shown. (G) Images show representative results of colony formation assays of NUGC4 (left) and SNU668 (right). Cells were cultured in 6-well plates and treated with DMSO, indicated concentration of defactinib or VS-6766 alone or together for 7 to 10 days, and then fixed and stained with crystal violet solution. All experiments were repeated for at least twice.

Figure 5.. CDK4/6 or MEK inhibition improves FAK inhibition response in DGC xenografts.

(A) Schematic description of in vivo experiment of SNU668 xenograft. We sacrificed one mouse in each group to obtain the tumor for IHC staining after the treatment of seven days. All mice were sacrificed at the endpoint, and the tumors were also collected. (B) Growth curve for SNU668 xenograft tumors (n=10–12) treated with vehicle control, VS-4718 (50 mg/kg, bid), palbociclib (50 mg/kg, qd), VS-6766 (0.3 mg/kg, qd), VS-4718+Palbociclib combination or VS-4718+VS-6766 combination. Treatment began on day1. Data are mean ± SEM. Statistical analysis was performed using data before day 22, ***P<0.001, ****P<0.0001, two-way ANOVA test. (C) Representative images of Ki-67, TUNEL and pERK1/2 staining of the SNU668 xenograft tumors collected at treatment day 7. Scale bar=100 μm.