ROS production induced by BRAF inhibitor treatment rewires metabolic processes affecting cell growth of melanoma cells

Abstract

Background: Most melanoma patients with BRAF^V600E positive tumors respond well to a combination of BRAF kinase and MEK inhibitors. However, some patients are intrinsically resistant while the majority of patients eventually develop drug resistance to the treatment. For patients insufficiently responding to BRAF and MEK inhibitors, there is an ongoing need for new treatment targets. Cellular metabolism is such a promising new target line: mutant BRAF^V600E has been shown to affect the metabolism.

Methods: Time course experiments and a series of western blots were performed in a panel of BRAF^V600E and BRAF^WT/NRAS^mut human melanoma cells, which were incubated with BRAF and MEK1 kinase inhibitors. siRNA approaches were used to investigate the metabolic players involved. Reactive oxygen species (ROS) were measured by confocal microscopy and AZD7545, an inhibitor targeting PDKs (pyruvate dehydrogenase kinase) was tested.

Results: We show that inhibition of the RAS/RAF/MEK/ERK pathway induces phosphorylation of the pyruvate dehydrogenase PDH-E1α subunit in BRAF^V600E and in BRAF^WT/NRAS^mut harboring cells. Inhibition of BRAF, MEK1 and siRNA knock-down of ERK1/2 mediated phosphorylation of PDH. siRNA-mediated knock-down of all PDKs or the use of DCA (a pan-PDK inhibitor) abolished PDH-E1α phosphorylation. BRAF inhibitor treatment also induced the upregulation of ROS, concomitantly with the induction of PDH phosphorylation. Suppression of ROS by MitoQ suppressed PDH-E1α phosphorylation, strongly suggesting that ROS mediate the activation of PDKs. Interestingly, the inhibition of PDK1 with AZD7545 specifically suppressed growth of BRAF-mutant and BRAF inhibitor resistant melanoma cells.

Conclusions: In BRAF^V600E and BRAF^WT/NRAS^mut melanoma cells, the increased production of ROS upon inhibition of the RAS/RAF/MEK/ERK pathway, is responsible for activating PDKs, which in turn phosphorylate and inactivate PDH. As part of a possible salvage pathway, the tricarboxylic acid cycle is inhibited leading to reduced oxidative metabolism and reduced ROS levels. We show that inhibition of PDKs by AZD7545 leads to growth suppression of BRAF-mutated and -inhibitor resistant melanoma cells. Thus small molecule PDK inhibitors such as AZD7545, might be promising drugs for combination treatment in melanoma patients with activating RAS/RAF/MEK/ERK pathway mutations (50% BRAF, 25% NRAS^mut, 11.9% NF1^mut).

Keywords: BRAF inhibitors; MEK inhibitors; Melanoma; Metabolism; NRAS; PDK inhibitors; Pyruvate dehydrogenase; Pyruvate dehydrogenase kinases; RAS/RAF/MEK/ERK pathway; Reactive oxygen species.

Fig. 1

BRAF inhibitors induce upregulation of PDH-E1α phosphorylation in BRAF^V600E melanoma cells but not in wild-type BRAF melanoma cells. Western blot analysis of A375, IGR37 and 501 Mel cells (BRAF^V600E) (a) and MelJuso, IPC298 and SKMel30 cells (BRAF^WT/NRAS^mut) (b) treated with 1 μM of PLX4032 and 100 nM of GSK2118432 for the indicated time points. α-Tubulin was used as loading control; representative blots of three biological replicates are shown. c Quantification of pPDH and PDK1 levels for A375, IGR37 and 501Mel, normalized to the untreated control. Error bars represent the standard deviation of three biological replicates. Statistical significance was determined using one-way ANOVA coupled with Dunnett’s multiple comparisons tests. *p > 0.05, **p > 0.01, ***p > 0.001

Fig. 2

Knock-down of ERK1/2 by siRNAs and treatment with a MEK1 inhibitor induces pS₃₀₀PDH. a A375 cells were transfected with siRNA against ERK1/2 (50 nM each siRNA) or a scrambled control (100 nM) for 72 h. 48 h prior to collection, the cells were incubated with either PLX4032 (1 μM) or Trametinib (3 nM) or both drugs. One representative of three biological replicates is shown. b Western blot analysis of A375 (BRAF^V600E), MelJuso, IPC298 and SKMel30 cells (NRAS^mut) treated with 5 nM of Trametinib for the indicated time points. α-Tubulin was used as loading control; representative blots of three biological replicates are shown. 1: Untransfected; 2: scrambled; 3: ERK1/2 siRNA

Fig. 3

PLX4032 treatment induces upregulation of PDK mRNA in BRAF^V600E melanoma cells. Quantitative RT-PCR of PDKs and HIF-1α mRNA. The fold change was calculated relative to untreated controls. Error bars represent the standard deviation of three biological replicates. PDK2 was not detectable in tested cell lines (Cq ≥ 30) while PDK4 (Cq ≥ 30) was not detectable in IGR37 cells only. Error bars represent the standard deviation of three biological replicates. Statistical significance was determined in comparison to the untreated control using paired Student’s t-tests. *p > 0.05, **p > 0.01, ***p > 0.001

Fig. 4

PDKs are crucial for PDH phosphorylation. A375 cells were transfected with siRNA against all four PDK isoforms (PDK1–4; 25 nM siRNA for each isoform) or 100 nM of scrambled control; after 24 h cells were incubated with PLX4032 (1 μM) for another 24 h. a qPCR analysis of PDK1–4 mRNA levels to show the efficiency of the siRNA knock-down of each siRNA and (b) corresponding western blot analysis. c Western blot analysis of A375 treated with PLX4032 (1 μM for 24 h), DCA (20 mM for 48 h) or both drugs. Error bars represent the standard deviation of three biological replicates. For each western blot experiment, one representative of three biological replicates is shown. Statistical significance was determined using paired Student’s t-tests. *p > 0.05, **p > 0.01, ***p > 0.001. 1: Untransfected; 2: scrambled; 3: PDKs siRNA

Fig. 5

PLX4032 and GSK2118436 induce upregulation of AMPK phosphorylation in BRAF^V600E melanoma cells. Western blot analysis of A375, IGR37 and 501 Mel cells (BRAF^V600E) treated with 1 μM of PLX4032 and 100 nM of GSK2118432 for the indicated time points. α-Tubulin was used as loading control, and one representative blot is shown

Fig. 6

ROS are induced after PLX4032 treatment of melanoma cells. a A375, 501 Mel or IGR37 cells were transiently transfected with plasmids encoding Grx1-roGFP2 or mito-roGFP2. Two days post transfection the cells were either subjected to a 1 h pretreatment with mitoQ (150 nM), a treatment with PLX4032 (3 h for A375; 10 h for 501Mel or (IGR37) or a combination thereof. Following this, the redox status was measured in the cytosol or mitochondria. Data were normalized to untreated cells, followed by one-way ANOVA coupled with Dunnett’s multiple comparisons tests. Representative graphs of two biological replicates are shown. b Western blot analysis of A375, IGR37 and 501Mel pre-treated for 1 h with mitoQ (150 nM) followed by 24 h of 1 μM PLX4032 treatment. Results are shown for one representative of three biological replicates. c Western blot analysis of A375, IGR37 and 501 Mel treated with hydrogen peroxide (H₂O₂; 100 μM) for the indicated time points. Results are shown for one representative of three biological replicates. *p > 0.05, **p > 0.01, ***p > 0.001

Fig. 7

AZD7545 mediated growth suppression of BRAF^V600E mutant and NRAS^mut melanoma cells. Twenty-four hours after plating, both A375 and IGR37 cell lines (BRAF^V600E) (a) and SKMel30, IPC298 and MelJuso cell lines (NRAS^mut) (b) were treated with 10 μM of AZD7545. The plates were imaged using an IncuCyte ZOOM live cell microscope (Essen BioScience) and images were taken every 3 h for a total of 90 h (BRAF^V600E) and 120 h (NRAS^mut). Results are shown for one representative of three biological replicates

Fig. 8

Combination of AZD7545 and PLX4032 more efficiently suppresses melanoma growth compared to each compound alone. a Represenative experiment of A375 melanoma cells expressing iRFP treated either with 1 μM of PLX4032 or with 1 μM of PLX4032 in combination with 10 μM AZD7545 for 3 weeks. The intensity of red fluorescence was quantified and the bar diagram represents three biological replicates with their standard deviation. b Spheroid cultures of A375 melanoma cells were treated with DMSO control, with 1 μM of PLX4032 or with 1 μM of PLX4032 in combination with 10 μM AZD7545. After 3 days sphere diameters were measured and represented as bar diagrams. Error bars represent the standard deviation of a minimum of four technical replicates of one representative experiment of three biological replicates. c Twenty-four hours after plating, BRAFi-resistant A375 melanoma cell (A375-R) were stimulated with 10 μM of AZD7545. The plates were imaged using an IncuCyte ZOOM live cell microscope (Essen BioScience) and images were taken every 3 h for a total of 90 h. Results are shown for one representative of three biological replicates. Statistical significance was determined using paired Student’s t-tests. *p > 0.05, **p > 0.01, ***p > 0.001

Fig. 9

Model summarising metabolic changes in sentitive versus resistant melanoma cells upon BRAF inhibition. a BRAF^V600E, in addition to its well known effects on the cell cycle and anti-apoptosis, leads to upregulation of glucose uptake and LDH-A, promoting aerobic glycolysis and cell growth. Moreover, activation of BRAF, leads to the induction of genes involved in antioxidant defense. b Upon treatment with BRAF inibitors, or other MEK/ERK pathway inhibitors, glycolysis is inhibited. Changes in metabolic fluxes through the TCA cycle together with the inhibition of the transcription of antioxidant genes, induce ROS. The increasing levels of ROS activate PDKs, which in turn inactivate PDH, thus reducing pyruvate use in the TCA cycle, which in a negative regulatory loop, inhibit further ROS production. c In BRAF inhibitor resistance, the MAPK pathway is reactivated by compensatory mechanisms and as a consequence glycolysis and antioxidant-defence genes are reactivated. Glycolysis produces ATP, nucleotides (via the pentose phosphate pathway (PPP)), NADPH (for antioxidant defense via PPP) and amino acids. The cells become addicted to an oxidative metabolism with glutamine (Gln) feeding anaplerotically into the TCA cycle to produce NADH, amino acids, ATP (via Oxphos), fatty acids (via AcetylCoA) and ROS. These ROS can be kept in check by NADPH, which is produced by the PPP and by antioxidant gene transcription downstream of the RAS/RAF/MEK/ERK pathway. At need, PDKs are activated by ROS and PDH is inhibited. This prevents pyruvate from entering the TCA and ROS to reach toxic levels. d In BRAF inhibitor resistance of melanoma, pharmacological inhibition of PDKs releases the brake on pyruvate entry into the TCA and causes unchecked pyruvate use, associated with an increase in ROS production. This increase in ROS tilts the balance towards cell death. Thus potent nonomolar PDK inhibitors could efficiently reduce the viability of BRAFi-resistant cells or might prevent or delay BRAFi resistance