Targeting ARF1-IQGAP1 interaction to suppress colorectal cancer metastasis and vemurafenib resistance

Abstract

Introduction: Acquired resistance to BRAF inhibitor vemurafenib is frequently observed in metastatic colorectal cancer (CRC), and it is a thorny issue that results in treatment failure. As adaptive responses for vemurafenib treatment, a series of cellular bypasses are response for the adaptive feedback reactivation of ERK signaling, which warrant further investigation.

Objectives: We identified ARF1 (ADP-ribosylation factor 1) as a novel regulator of both vemurafenib resistance and cancer metastasis, its molecular mechanism and potential inhibitor were investigated in this study.

Methods: DIA-based quantitative proteomics and RNA-seq were performed to systematic analyze the profiling of vemurafenib-resistant RKO cells (RKO-VR) and highly invasive RKO cells (RKO-I8), respectively. Co‑immunoprecipitation assay was performed to detect the interaction of ARF1 and IQGAP1 (IQ-domain GTPase activating protein 1). An ELISA-based drug screen system on FDA-approved drug library was established to screen the compounds against the interaction of ARF1-IQGAP1.The biological functions of ARF1 and LY2835219 were determined by transwell, western blotting, Annexin V-FITC/PI staining and in vivo experimental metastasis assays.

Results: We found that ARF1 strongly interacted with IQGAP1 to activate ERK signaling in VR and I8 CRC cells. Deletion of IQGAP1 or inactivation of ARF1 (ARF-T48S) restored the invasive ability induced by ARF1. As ARF1-IQGAP1 interaction is essential for ERK activation, we screened LY2835219 as novel inhibitor of ARF1-IQGAP1 interaction, which inactivated ERK signaling and suppressed CRC metastasis and vemurafenib-resistance in vitro and in vivo with no observed side effect. Furthermore, LY2835219 in combined treatment with vemurafenib exerted significantly inhibitory effect on ARF1-mediated cancer metastasis than used independently.

Conclusion: This study uncovers that ARF1-IQGAP1 interaction-mediated ERK signaling reactivation is critical for vemurafenib resistance and cancer metastasis, and that LY2835219 is a promising therapeutic agent for CRC both as a single agent and in combination with vemurafenib.

Keywords: ARF1-IQGAP1 interaction; Colorectal cancer; LY2835219; Metastasis; Vemurafenib resistance.

Graphical abstract

Fig. 1

ARF1-mediated ERK signaling activation links to vemurafenib resistance and cancer metastasis. (A) Experimental scheme showing the establishment of vemurafenib resistant CRC cell line (RKO-VR) and their invasive ability was compared. The number of invaded cells is shown in the bar chart (B). Scale bar, 100 μm. (C) Transwell invasion assay comparing the invasive ability of RKO and RKO-VR cells with or without vemurafenib treatment (10 μM). Scale bar, 100 μm. (D) Diagram depicting the screening of highly invasive cell lines (RKO-I8 and HCT116-I8). Transwell invasion assay comparing the invasive ability of I8 cells with their corresponding parental cells (E). Scale bar, 100 μm. (F, G) GSEA analysis showing the MAPK signaling pathway was enriched in the DEPs/DEGs of VR and I8 cells. (H) Venn diagram illustrating the overlap of upregulated DEPs/DEGs of RKO-VR and RKO-I8, identifying ARF1 regulating both vemurafenib resistance and cancer invasion. (I) The mRNA level of ARF1 in RKO-VR cells, I8 cells and their corresponding parental cells were analyzed by qRT-PCR. (J, K) Western blotting of the ARF1 and p-ERK levels in HCT116-I8, RKO-I8 and RKO-VR cells compared with their corresponding parental cells. The expression of ARF1 in triplicates was quantified and normalized to corresponding Actin. Bars, SD; **, P < 0.01; ***, P < 0.001; ns, no significant difference.

Fig. 2

ARF1 promotes the vemurafenib resistance and metastasis of CRC. HCT116-con/ARF1, RKO-con/ARF1 (A) and DLD1-shcon/shARF1, HT29-shcon/shARF1 (B) were determined by transwell matrigel invasion assays. Scale bar, 100 μm. The number of invaded cells is shown in the bar chart. (C) Representative bioluminescent images monitoring the cancer metastasis of NCG mice received intravenous injection of indicated cells (HCT116-luc-con or HCT116-luc-ARF1), and the intensity of bioluminescence was quantified. The lungs were harvested for imaging and H&E staining (D). (E, F) Comparison of the invasive ability of RKO-con/ARF1 and HCT116-con/ARF1 treated with or without vemurafenib (10 μM). Scale bar, 100 μm. (G) Western blotting measurement of the protein level of ARF1 in RKO-VR treated with siRNAs. (H) Transwell matrigel invasion assay determining the inhibitory effect of vemurafenib (10 μM) on invasion ability of RKO-VR cells with or without knockdown of ARF1. (I) Western blotting of the protein level of p-ERK and ERK in ARF1-overexpression CRC cells treated with or without vemurafenib (10 μM). (G) Western blotting of the protein level of p-ERK and ERK in ARF1-kncokdown CRC cells treated with or without vemurafenib (10 μM). Scale bar, 100 μm. Bars, SD; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, no significant difference.

Fig. 3

ARF1 promotes CRC invasion and vemurafenib resistance by activating ERK signaling. (A) Transwell matrigel invasion assay was examined in CRC cells with or without stably expressing ARF1, followed by the treatment of U0126 (10 μM) as indicated. Scale bar, 100 μm. (B) Transwell matrigel invasion assay of CRC cells with indicated treatments. Scale bar, 100 μm. (C) Co-immunoprecipitation assay was performed in HCT116/HCT116-I8 and RKO/RKO-VR cells for analyzing the interaction of ARF1-IQGAP1 by immunoprecipitated IQGAP1. (D) Transwell matrigel invasion assay was performed in HCT116-sgcon/sgIQGAP1 and RKO-sgcon/sgIQGAP1 with or without overexpressing ARF1. Scale bar, 100 μm. (E) Transwell assays were performed to determine invasion of CRC cells with or without overexpressing ARF1-T48S mutant. Scale bar, 100 μm. Bars, SD; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, no significant difference.

Fig. 4

LY2839219 disrupts the interaction of ARF1-IQGAP1. (A) Experimental scheme of ELISA-based drug screen system using FDA-approved drug library. (B) The absorbance in wells with different drugs treatment. nc, negative control. (C) The structure of LY2839219. (D) Co-immunoprecipitation assay was examined in CRC cells treated with LY2839219 or DMSO for detecting the interaction of ARF1-IQGAP1. (E) Co-precipitation of endogenous ARF1 with IQGAP1 from HCT116 and RKO-VR cells treated with increasing concentrations of LY2839219 (0–3 μM). (F) Western blotting measurement of the protein level of IQGAP1, ARF1, p-ERK and ERK expression in HCT116 and RKO treated with elevating concentration of LY2839219.

Fig. 5

LY2839219 suppresses CRC metastasis in vivo and in vitro. (A) Transwell matrigel invasion assay of HCT116 and RKO cells treated with elevating concentration of LY2839219. Scale bar, 100 μm. (B) The invasive ability of HCT116-sgcon/HCT116-sgIQGAP1 and RKO-sgcon/RKO-sgIQGAP1 treated with or without LY2839219 (2 μM). Scale bar, 100 μm. (C) Invasive ability of ARF1-overexpressing CRC cells treated with or without LY2839219 (2 μM) were compared with control cells. Scale bar, 100 μm. (D) Transwell assay comparing the invasion of RKO-VR cells treated with LY2839219 (2–3 μM) and DMSO. Scale bar, 100 μm. (E) Representative bioluminescent images monitoring the cancer metastasis in NCG mice received intravenous injection of HCT116-Luc-ARF1 cells treated with or without LY2839219 (5 mg/kg), and the intensity of bioluminescence was quantified. (F) The lungs were harvested for imaging and H&E staining. Scale bar, 100 μm. Bars, SD; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, no significant difference.

Fig. 6

Combination of LY2839219 and vemurafenib significantly inhibits ARF1-mediated cancer metastasis. (A) Transwell matrigel invasion assay was used in HCT116-ARF1 and RKO-ARF1 cells treated with or without the combined use of LY2839219 and vemurafenib. Scale bar, 100 μm. (B) Combined treatment of LY2839219 with vemurafenib suppressed CRC metastasis in vivo. Upper panel, diagram depicting the drugs treatment. Down panel, cancer metastasis in NCG mice was detected by bioluminescent images and quantified. The lung and kidney were harvested for imaging and H&E staining of the metastatic nodes (C). (D) ALT and AST were analyzed in mice serum. Bars, SD; *, P < 0.05; **, P < 0.01; ***, P < 0.001.