Vemurafenib resistance reprograms melanoma cells towards glutamine dependence

Abstract

Background: (V600) BRAF mutations drive approximately 50% of metastatic melanoma which can be therapeutically targeted by BRAF inhibitors (BRAFi) and, based on resistance mechanisms, the combination of BRAF and MEK inhibitors (BRAFi + MEKi). Although the combination therapy has been shown to provide superior clinical benefits, acquired resistance is still prevalent and limits the overall survival benefits. Recent work has shown that oncogenic changes can lead to alterations in tumor cell metabolism rendering cells addicted to nutrients, such as the amino acid glutamine. Here, we evaluated whether melanoma cells with acquired resistance display glutamine dependence and whether glutamine metabolism can be a potential molecular target to treat resistant cells.

Methods: Isogenic BRAFi sensitive parental (V600) BRAF mutant melanoma cell lines and resistant (derived by chronic treatment with vemurafenib) sub-lines were used to assess differences in the glutamine uptake and sensitivity to glutamine deprivation. To evaluate a broader range of resistance mechanisms, isogenic pairs where the sub-lines were resistant to BRAFi + MEKi were also studied. Since resistant cells demonstrated increased sensitivity to glutamine deficiency, we used glutaminase inhibitors BPTES [bis-2-(5 phenylacetamido-1, 2, 4-thiadiazol-2-yl) ethyl sulfide] and L-L-DON (6-Diazo-5-oxo-L-norleucine) to treat MAPK pathway inhibitor (MAPKi) resistant cell populations both in vitro and in vivo.

Results: We demonstrated that MAPKi-acquired resistant cells uptook greater amounts of glutamine and have increased sensitivity to glutamine deprivation than their MAPKi-sensitive counterparts. In addition, it was found that both BPTES and L-DON were more effective at decreasing cell survival of MAPKi-resistant sub-lines than parental cell populations in vitro. We also showed that mutant NRAS was critical for glutamine addiction in mutant NRAS driven resistance. When tested in vivo, we found that xenografts derived from resistant cells were more sensitive to BPTES or L-DON treatment than those derived from parental cells.

Conclusion: Our study is a proof-of-concept for the potential of targeting glutamine metabolism as an alternative strategy to suppress acquired MAPKi-resistance in melanoma.

Figure 1

Vemurafenib resistant cells uptake and use glutamine at a higher rate than vemurafenib sensitive cells. Melanoma M249 and M229 vemurafenib sensitive (parental) and single drug resistant (SDR) cell lines were plated to be 60% confluent and medium was changed after overnight incubation. Medium only was used as a control. Cells were then cultured for 24 h. Medium was then taken from growing cells and measurements were obtained of glutamine uptake, NH₄ ⁺ production, glucose uptake, and lactate production using the Nova Bioprofiler 100 plus as described in materials and methods. Cells were counted and results are expressed as a glutamine uptake per cell (mmol/L) and NH₄ ⁺ production per cell (mmol/L) and b glucose uptake per cell (mmol/L) and lactate production per cell (mmol/L) and are representative of the average (±standard deviation) of triplicate experiments.

Figure 2

Vemurafenib resistant cells are more sensitive to glutamine deprivation. a Melanoma M249 and b M229 single drug resistant (SDR) cells were cultured in the presence of medium with or without glutamine (Gln) or glucose (Gluc). After 24 h representative images and surviving fractions (survival %) compared to control cells in complete medium are shown using DIMSCAN, a microcomputer fluorescence-based cytotoxicity assay (n = 6, **p < 0.01). Western blot analysis was also used to assess levels of cleaved PARP as an indicator of apoptosis in both c M249 and d M229 parental and SDR cell lines. e Apoptosis was also assessed using flow cytometry of M249 parental and resistance cells stained with AnnexinV and DAPI. M249 parental and SDR cells were cultured in the presence of medium with and without glutamine for 24 h. Representative dot blots and percentages of AnnexinV positive cells are shown. f Melanoma double drug resistant (DDR) cell line M249 DDR5 was cultured in the presence of medium with and without glutamine (Gln) or glucose (Gluc). After 24 h representative images and surviving fractions (survival %) compared to control cells in complete medium are shown using DIMSCAN, a microcomputer fluorescence-based cytotoxicity assay (n = 6, ***p < 0.001). g Apoptosis was also assessed using flow cytometry of M249 parental and DDR cells stained with AnnexinV and DAPI. M249 parental and DDR cells were cultured in the presence of medium with and without glutamine for 24 h. Representative dot blots and percentages of AnnexinV positive cells are shown.

Figure 3

Vemurafenib resistant cells are more sensitive to glutaminase inhibitors. a M249 and M229 parental single drug resistant (SDR) cells were cultured in the presence of medium control (DMSO) or 10 μM BPTES for 48 h. SDR cell lines were cultured in combination of vemurafenib (1 μM) and BPTES. Surviving cell fraction percentage (survival %) compared to vehicle control (DMSO) was assessed using DIMSCAN technology (n = 6, ***p < 0.001, *p < 0.05). b M249 and M229 parental and single drug resistant cells were cultured in the presence of vehicle control (water) or 10 μM L-DON for 48 h. SDR cell lines received a combination of vemurafenib (1 μM) and L-DON. Surviving cell fraction percentage (survival %) compared to vehicle control (DMSO) was assessed using DIMSCAN technology (n = 6, ***p < 0.001, *p < 0.05). c M249 parental and double drug resistant (DDR) cells were cultured in the presence of medium control (DMSO) or 10 μM BPTES for 48 h. DDR cell lines were cultured in combination with vemurafenib (1 μM), selumetinib (1 μM), and BPTES. Surviving cell fraction percentage compared to vehicle control (DMSO) (survival %) was assessed using DIMSCAN technology (n = 6, ***p < 0.001). d M249 single (SDR) and double drug (DDR) resistant lines were cultured in the presence or absence of inhibitor (1 μM vemurafenib for SDR and 1 μM vemurafenib and 1 μM selumentib for DDR) and subsequently treated with 10 μM BPTES or deprived of glutamine (Gln) for 48 h. Surviving cell fraction percentage compared to vehicle control (DMSO) (survival %) was assessed using DIMSCAN technology.

Figure 4

Knock down of NRAS in vemurafenib resistant cells reduces sensitivity to glutaminase inhibitor. Melanoma M249 resistant cells were infected with shNRAS or shControl lentiviral particles and selected for puromycin resistance after 48 h. a Quantitative PCR was used to measure relative mRNA expression of NRAS in shControl and shNRAS infected cells. b shNRAS or shControl cells were cultured in the presence of 10 μM BPTES for 48 h or with vehicle control (DMSO). Surviving cell fraction percentage (survival %) compared to vehicle control (DMSO) was determined using DIMSCAN analysis (n = 6, *p < 0.05).

Figure 5

Vemurafenib resistant melanoma tumors are sensitive to glutaminase inhibitor treatment in vivo. a Nod Scid Gamma (NSG) mice were injected with 5 × 10⁶ of M249 parental or single drug resistant cells subcutaneously on the right flank. When tumor size reached an average of 100 mm³ tumor cell volume, treatment with 15 mg/kg of BPTES or vehicle control (DMSO) was provided every other day through intraperitoneal injection. Measurements were taken for tumor length and width. Tumor volume (mm³) was calculated by multiplying (length × width × width)/2. Graph represents mean tumor volume ± SD. b Nude mice were injected with 2 × 10⁶ of M249 resistant cells subcutaneously on right flank. When tumor size reached an average of 100 mm³ tumor cell volume, treatment with 20 mg/kg of L-DON or vehicle control (water) was provided twice a week through intraperitoneal injection. Graph represents mean tumor volume ± SD.