Acquired resistance to BRAF inhibitors mediated by a RAF kinase switch in melanoma can be overcome by cotargeting MEK and IGF-1R/PI3K

Abstract

BRAF is an attractive target for melanoma drug development. However, resistance to BRAF inhibitors is a significant clinical challenge. We describe a model of resistance to BRAF inhibitors developed by chronic treatment of BRAF(V)⁶⁰⁰(E) melanoma cells with the BRAF inhibitor SB-590885; these cells are cross-resistant to other BRAF-selective inhibitors. Resistance involves flexible switching among the three RAF isoforms, underscoring the ability of melanoma cells to adapt to pharmacological challenges. IGF-1R/PI3K signaling was enhanced in resistant melanomas, and combined treatment with IGF-1R/PI3K and MEK inhibitors induced death of BRAF inhibitor-resistant cells. Increased IGF-1R and pAKT levels in a post-relapse human tumor sample are consistent with a role for IGF-1R/PI3K-dependent survival in the development of resistance to BRAF inhibitors.

Copyright © 2010 Elsevier Inc. All rights reserved.

Figures

Figure 1. BRAFV600E mutant melanomas chronically treated with BRAF inhibitors develop drug resistance

(A) Schematic representation of generation of SB-590885 (885) resistant cells. The resistant cells are indicated by the name of the parental cell line followed by R. (B–C) Sensitivity to BRAF inhibition of parental (blue) and 885 chronically treated melanoma cells (red) was assessed by MTT assays. Relative growth (RG) was calculated as the ratio of treated to untreated cells at each dose for each replicate. Data are represented as mean ± SEM (n=7). (B) At all doses less than 10 μM, RG was significantly lower for 451Lu cells (p<0.05). (C) At all doses RG was significantly lower for Mel1617 cells (p<0.05). (D) Cells were treated with DMSO or 1 μM 885 for 24h, stained with propidium iodide and analyzed for cell cycle progression by flow cytometry. Response to treatment between the two cell lines was not significantly different (p>0.05). The percentage of cells in G0/G1 and SubG1 are shown. Representative cell cycle plots from one experiment are shown. (E–F) Sensitivity of 451Lu (E) and Mel1617 (F) parental and the corresponding 885-resistant cells to PLX4720 was assessed by MTT assays as in (B). Cells were treated with the indicated concentrations of PLX4720 (nM). Data represent mean of three independent experiments ± SEM. Parental and resistant cells were significantly different (p<0.05) at doses > 1 μM. See also Figure S1.

Figure 2. Growth properties of SB-590885-resistant melanoma cells

(A) Cells were treated with 1 μM 885 for 6 days, fixed at the indicated days, stained with crystal violet and photographed. Cell number was determined relative to day 0. Data represents mean ± SEM (n=3). P<0.05 when comparing 451Lu + 885 with 451Lu + DMSO or 451Lu-R +/− 885 at day 6; p>0.05 when comparing 451Lu-R +/− 885 with 451Lu. Scale bar, 800 μm (B) 451Lu parental and resistant cells were grown in soft agar for 12 days +/− 1 μM 885. Anchorage-independent growth was assessed by counting individual colonies using ImagePro-Plus software in triplicate and normalizing to vehicle controls for each condition. Data represent mean ± SEM (n=3), p<0.05 DMSO-treated vs. 885-treated parental cells, and p>0.05 for parental DMSO-treated vs. resistant cells. Scale bar, 10 mm. (C) 451Lu (left panels) and 451Lu-R (right panels) collagen-embedded spheroids were treated with the indicated concentrations of 885. Spheroids were stained with calcein-AM and imaged with a confocal microscope. Relative viability was assessed based on the % of cells remaining after treatment and morphological appearance. Scale bar, 150 μm.

Figure 3. Abrogation of ERK activity in SB-590885-resistant melanomas requires inhibition of all three RAF isoforms and leads to cell cycle arrest

(A) 451Lu and 451Lu-R cells were treated with increasing concentrations of 885. Cells lysates were analyzed by immunoblotting. (B) 451Lu and 451Lu-R cells were infected with either a control [C] lentiviral shRNA or 3 different clones targeting BRAF (1, 3 or 4). Cell lysates were analyzed by immunoblotting. (C) 451Lu-R cells were infected with lentiviral shRNA directed against CRAF or GFP. Infected cells were treated with 1 μM 885 (+) or left untreated (−) for 24h. Cell lysates were analyzed by immunoblotting with antibodies against CRAF, pERK, and total ERK (loading control). (D) 451Lu-R cells were infected with lentiviral shRNA against ARAF or GFP. Infected cells were treated with 1 μM 885 or DMSO for 24h. Cells were harvested, lysed and analyzed by immunoblotting with antibodies against ARAF, BRAF, CRAF, pERK, and total ERK (loading control). (E) 451Lu–R cells were sequentially infected with lentiviral shRNA directed against CRAF followed by ARAF. Infected and control cells were then treated with 1 μM 885 for 24h (+) or left untreated (−). Cell lysates were analyzed by immunoblotting. (F) After lentiviral infection, cells were treated with DMSO or 1 μM 885 for 72h. Cells were harvested, fixed, stained with propidium iodide and analyzed by flow cytometry. The percentage of cells in G0/G1 is indicated for each condition. See also Figure S2.

Figure 4. The MEK inhibitor GSK1120212 prevents ERK activation and proliferation in both SB-590885 sensitive and resistant cell lines

(A) Chemical structure of GSK1120212 (212). (B) 451Lu and 451Lu-R cells were treated with 1 μM 885, 10 μM UO126 (UO) or increasing concentrations of 212 (nM) for 24h. Cell lysates were analyzed by immunoblotting. (C–D) Sensitivity to the MEK inhibitor 212 was assessed by MTT assays as in Figure 1. Data represent means ± SEM (n=7). (E) Parental and 885-resistant cells were treated with 212 for 72 h, stained with propidium iodide, and analyzed by flow cytometry. Data represent mean ± SEM (n=3); (*) p<0.05 when compared to DMSO-treated cells. (F) 451Lu and 451Lu-R collagen-embedded spheroids were treated with the indicated concentrations of 212 for 72 hours. Cells were imaged with a confocal microscope. Scale bar, 150 μm. See also Figure S3 and Table S2.

Figure 5. Enhanced IGF-1R in cells chronically treated with BRAF inhibitors

(A) 451Lu and 451Lu-R cells were treated with 1 μM 885 or DMSO (D) for 24h. Whole-cell lysates were incubated on RTK antibody arrays. Each RTK antibody is spotted in duplicate. Positive RTK dots are circled in red and indicated by a number; the corresponding RTKs are listed next to the arrays. (B) Sensitivity to the IGF-1R inhibitor PPP (μM) or the c-Met inhibitor PHA (μM) was assessed in 451Lu-R and Mel1617-R by MTT assays as in Figure 1. Data represent means ± SEM (n=7) (C) IGF-1R surface expression was assed by indirect immunoflourescence in parental (black) and resistant (red) melanomas treated with 885 (1 μM) for 20h. Dotted lines denote control rabbit antibody for the corresponding parental or resistant cells. Numbers on the top right indicate percent positive surface expression of IGF-1R in parental (black) and resistant (red) cells. (D) Expression and phosphorylation of IGF-1R was assessed in parental (P) and resistant (R) melanomas treated with 885 (1 μM) for 20h. Cell lysates were analyzed by immunoblotting with the indicated antibodies. See also Figure S4.

Figure 6. IGF-R mediates PI3K signaling in BRAF-inhibitor resistant cells

(A) Mel1617-R cells were treated with increasing concentrations of PPP (μM) as single agent or in combination with 0.1 μM 212. The effect of IGF-1R inhibition on MAPK, AKT, and Bcl-2 family proteins was assessed by immunoblotting. (B) Mel1617-R cells were infected with adenoviruses encoding dominant negative (dn) IGF-1R at 10 or 100 MOI, or LacZ as a negative control. Infected cells were treated 48h post infection with 0.1 μM 212 or left untreated. Cells lysates were analyzed by immunoblotting; cl-PARP, cleaved PARP. (C) Parental Mel1617 cells were infected in serum-free medium with adenoviruses encoding IGF-1 at 10, 100 or 500 MOI. Infected cells were serum starved for 48 h and then treated with 1 μM 885 or left untreated. Cells lysates were analyzed by immunoblotting. Doted lines indicate where blot was cut to remove an empty lane. See also Figure S5.

Figure 7. Co-inhibition of IGF-1R/PI3K and MEK induces cytotoxicity in BRAF-inhibitor resistant cells

(A) Cell cycle profiles of 451Lu-R cells treated with DMSO, 1 μM 212, 1 μM PPP, or a combination of both inhibitors for 72h. Percentage of cells in G0/G1 and SubG1 are shown. (B) 451Lu-R cells were treated with DMSO, 212 (1 μM), PPP (1 μM) or both inhibitors at the same concentrations for 72h. Cells were collected and apoptosis was assessed by Annexin-V staining. Numbers in each quadrant indicate percentage of cells. Representative results of 2 independent experiments are shown. (C) Collagen-embedded Mel1617-R spheroids were treated with DMSO, 10 μM 885, 1 μM 212, 1 μM 458 or 1 μM 212 + 1 μM 458 for 72 hours. Cells were imaged with a confocal microscope. Scale bar, 150 μm. See also Figure S6 and Tables S3–4.

Figure 8. Increased IGF-1R expression and phosphorylation of AKT in relapsed patient samples

Paired tumor samples (pt 1, Table S5) taken before treatment (pre-treatment, sub-cutaneous back lesion) and drug resistant (post-relapse, small bowel) were analyzed for IGF-1R expression or phospho-AKT by immunohistochemistry. Low magnification representative images are shown on the left (scale bar, 500 μm) and higher magnification images are on shown on the right (scale bar, 50 μm). See also Figure S7 and Table S5.