Integrin-Src-YAP1 signaling mediates the melanoma acquired resistance to MAPK and PI3K/mTOR dual targeted therapy

Abstract

Activation of PI3K/AKT pathway is one of the most recurrent resistant mechanisms for BRAF-targeted therapy, and the combination of MAPK and PI3K/AKT inhibitors becomes one of the most promising regimens for BRAF-targeted relapsed melanoma patients. Although the potent drug efficacy was observed in preclinical experiments and early clinical trials, the dual-drug resistance is inevitable observed. In this study, we systematically explored the mechanisms of dual-drug resistance to MAPKi and PI3K/mTORi in melanoma. With transcriptomic dissection of dual-drug resistant models, we identified that the drug tolerance was mediated by ECM-integrins α3β1 and α11β1 signaling. Upon binding ECM, the integrins activated downstream kinase Src rather than FAK, WNT, or TGFβ. Knockdown of integrins α3, α11, and β1 significantly inhibited the proliferation of dual-drug resistant sublines while with trivial effects on parental cells. Although Src inhibition suppressed the phosphorylation of AKT, c-JUN, and p38, none of inhibitors targeting these kinases reversed the dual-drug resistance in model cells. Notably, Src inhibitor promoted the phosphorylations of LATS1 and YAP1, subsequently, re-localized YAP1 from nucleus to cytosol facilitating further degradation. Both small molecule inhibitors and shRNAs targeting YAP1 or Src overcame the MAPKi and PI3K/mTORi dual-drug resistance. In conclusion, our data not only illuminated an integrin-Src-YAP1 pathway mediated MAPKi and PI3K/mTORi dual-drug resistant mechanism but also provided a potential combinatorial regimen for the drug-relapsed melanoma patients.

Keywords: Integrin-Src-YAP1 axis; MAPK; Melanoma; PI3K/mTOR; Resistance; Targeted therapy.

Fig. 1

Chronic MAPK and PI3K dual-inhibition lead to acquired drug resistance. a Relative drugs exposure time to achieve resistance to MEKi+PI3K/mTORi in WM2664 and SKMEL28. b Drug naïve cells were chronically treated with increasing concentration of MEK inhibitor AZD6244 and PI3K/mTOR inhibitor BEZ235. c Phase-contrast images showing morphological changes in WM2664 and SKMEL28 parental and resistant cell lines (left), immunofluorescence staining for visualizing cell boundaries by fluorescence microscope (middle), immunofluorescence staining for cytoskeleton by confocal (right), (scale bar = 100 μm). d Survival curves of parental and dual-drug resistant cell lines titrated with the AZD6244 and BEZ235 or combine for 72 h. Results are shown relative to DMSO-treated controls (mean ± SEM, n = 5; dashed line, 50% inhibition). e Long-term colony formation assays of melanoma isogenic pairs. Parental and resistant clones were treated with indicated concentration of AZD6244 and BEZ235 for 12–14 days and then stained with 0.05% crystal violet to assay viability. The image is representative of three biological replicates. f Parental and DPR sublines were treated with DMSO or AZD6244 (1 μM) + BEZ235 (0.1 μM) for 48 h. Cells were collected and apoptosis was assessed via Annexin V-FITC staining. Quantification of the percentage of apoptosis cells (right). g Cell cycle analysis were assessed propidium iodide staining in parental and dual-drug resistant sublines treated with DMSO or AZD6244 (1 μM) plus BEZ235 (0.1 μM) for 24 or 48 h

Fig. 2

Compensatory signaling supported the survival of melanoma cells with MAPK and PI3K dual inhibition. a Dose-dependent suppression of MAPK and PI3K pathways by AZD6244 (0, 0.1, 1, 10 μM) combined with BEZ235 (0, 0.01, 0.1, 1 μM) in WM2664 and SKMEL28 parental and DPRs (left). Phosphorylation levels of these proteins in parental or DPRs treated with AZD6244 (1 μM) and BEZ235 (0.1 μM) for indicated durations (h) (right). b IHC analysis of indicated proteins in tumors grown with or without treatment of AZD6244 and BEZ235 (scale bar = 100 μm). Image is representative of five independent experiments

Fig. 3

Transcriptome profiling suggested that ECM receptor pathways were enrichment in resistant cell lines. a PCA analysis of RNA-seq profiles in WM2664 and SKMEL28 parental, DPRs, and single drug resistance (SDR). Each dot represents one sample. b Heat map of all expressed genes in SKMEL28 and WM2664 DPR, SDR, and parental cell lines. Gene-expression variant was calculated by Z-score. c Venn diagram of differential expression genes (DEGs) in all resistant cell lines. d KEGG enrichment analysis of DEGs in DPR resistant cell lines (p < 0.05, top 5). e Expression profile of DEGs in the ECM pathway of WM2664 DPR or SKMEL28 DPR, colors of outer circles present z-score of gene expression, and colors of inter circle means p-value of Wald statistical test between DPR resistance and Parental cell groups. f STRING analysis identified the interaction of differential proteins in the ECM pathway; the size of points indicates the node degree of genes

Fig. 4

Transcriptome profiling suggested that ECM receptor pathways were enrichment in resistant cell lines. a mRNA expression of integrins in SKMEL28 and WM2664 parental and DPRs. Significance was determined by Wilcoxon signed-rank test, ∗p < 0.05. b Western blot showing protein levels of integrins α1, α3, α11 and β1 in WM2664 and SKMEL28 parental and DPRs. c Representative images of IHC analysis with integrin α3, α11 and β1 in tumors grown with or without dual inhibitors (scale bar = 100 μm). Image is representative of five independent experiments. d Clonogenic assays of parental and DPRs engineered control (vector) or integrins-targeting shRNAs. Results are shown for one representative of three independent experiments. e Quantification of (d), ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001

Fig. 5

The inhibition of Src suppressed the growth of dual-drug resistant cells. a SKMEL28 DPR and WM2664 DPR were transduced with shRNA targeting nontarget (vector) or shRNA targeting against integrin α3, α11 and β1. Cell lysates were made for immunoblot analysis with antibodies indicated. β-ACTIN is as loading control. b Survival curves of SKMEL28 DPR and WM2664 DPR titrated with dasatinib for 72 h. Results are shown relative to DMSO-treated controls (mean ± SEM, n = 5; dashed line, 50% inhibition). c Long-term colony assays of resistant sub-lines treated with dasatinib as indicated concentration. d Quantification of (c) by Image J, ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001. e Clonogenic assays of DPRs engineered control (vector) or Src-targeting shRNA treated with 1 μM AZD6244 and 0.1 μM BEZ235. f Survival curves of resistant sublines of SKMEL28 and WM2664 titrated with AZD6244 + BEZ235 with or without PF-562271 for 72 h. Results are shown relative to DMSO-treated controls (mean ± SEM, n = 5; dashed line, 50% inhibition). g Long-term colony assays of SKMEL28 DPR, WM2664 TPR and WM2664 DPR treated with PF-562271. h Quantification of (g) by Image J, ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001

Fig. 6

Src-YAP1 mediated the resistance to MAPK and PI3K/mTOR dual inhibitors. a Western blots showing indicated proteins levels in SKMEL28 DPR and WM2664 DPR treated with or without dasatinib (1 μM) for 1 h. b-e Immunofluorescence micrographs of YAP1 localization in SKMEL28 DPR (b) and WM2664 DPR (d) treated with dasatinib for indicated durations, scare bar = 100 μm. c, quantification of b, ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001. e, quantification of d, ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001. f TPM values of YAP1 targets (CTGF and CYR61) in SKMEL28 and WM2664 isogenic pairs, ∗p < 0.05. g mRNA expression of YAP1 targets (CTGF and CYR61) measured by q-PCR in SKMEL28 and WM2664 parental and DPRs, ∗p < 0.05, ∗∗p < 0.01, or ∗∗∗p < 0.001. h Western blots showing YAP1 targets protein levels in SKMEL28 and WM2664 parental and dual resistant sublines. i Representative images of IHC analysis with indicated proteins in tumors grown with or without AZD6244 plus BEZ235, scale bar = 100 μm. Image is representative of five independent experiments. j Clonogenic assays of DPRs were performed 10–14 days after treated with YAP1 inhibitor as indicated concentration. k Clonogenic assays of DPRs engineered control (vector) or YAP1-targeting shRNA treated with 1 μM AZD6244 and 0.1 μM BEZ235

Fig. 7

A proposed model depicting the role of integrin/Src/YAP1 axis in MAPK and PI3K/mTOR dual-inhibitor resistance. a In the sensitive cell lines, the MAPK or PI3K/mTOR pathway is continuously activated due to aberrant mutation or amplification, which leads to abnormal growth of melanoma cell. The combination of MAPK and PI3K/mTOR targeted therapy effectively induced melanoma cell death and inhibited the tumor growth. b In the MAPKi and PI3K/mTORi dual-resistant cell lines, the activation of integrin/Src/YAP1 signaling renders tolerance to MAPK and PI3K/mTOR dual inhibitors and inhibition of integrin or Src or YAP1 is capable to restore melanoma cell sensitivity to MAPKi and PI3K/mTORi, which leads to tumor growth inhibition