The human melanoma cell line A375 was purchased from the European Collection of Cell Cultures (Salisbury, UK) and cultured in BioWhittaker™ RPMI-1640 medium (Lonza, Verviers, Belgium) supplemented with 10% fetal bovine serum (FBS; Sigma-Aldrich, St. Louis, MO, USA), 2 mM BioWhittaker™ L-glutamine (Lonza), and 50 μg/ml BioWhittaker™ gentamicin (Lonza) (hereafter referred to as complete medium, CM). The A375R subline with acquired resistance to dabrafenib was generated by growing the parental A375 cells in gradually increasing concentrations of dabrafenib (from 1 nM up to 1.5 μM) over a period of 4 months. After selection, A375R cells were maintained in CM supplemented with 1.5 μM dabrafenib.

Dabrafenib (GSK2118436A) and GSK2126458A were dissolved in dimethyl sulfoxide (DMSO; Sigma-Aldrich) at a final concentration of 1.92 and 1 mM, respectively. Both drugs were stored as stock solutions at −80°C and diluted in CM just before use.

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was purchased from Sigma-Aldrich, dissolved at a concentration of 5 mg/ml in Gibco™ phosphate-buffered saline (PBS; Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) and stored at 4°C.

For western blot analysis, the following rabbit polyclonal antibodies were used: anti-human ERK1/2 (GTX17942; GeneTex, Irvine, CA, USA), anti-human phospho-ERK1/2 (Thr185/Tyr187) (44-680G; Invitrogen), anti-human β-tubulin (sc-9104; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), anti-vascular endothelial growth factor receptor-2 (VEGFR-2) (sc-6251; Santa Cruz Biotechnology), anti-human AKT (9272; Cell Signaling Technology, Inc., Beverly, MA, USA) and anti-human phospho-AKT (Ser473) (9271; Cell Signaling Technology).

Reagents for sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) were all purchased from Bio-Rad Laboratories, Inc. (Hercules, CA, USA).

DNA extraction from A375 and A375R cell lines was performed using the DNeasy Blood & Tissue kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer’s protocol, and sheared with a Bioruptor sonicator (Diagenode, Liege, Belgium). DNA libraries were prepared and captured using the SureSelect Human All Exon kit (50 Mb; Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer’s instructions. DNA libraries were sequenced by Genomix4life S.r.l., (Baronissi, Salerno, Italy) on Illumina NextSeq 500 platform (Illumina, San Diego, CA, USA) with 100-bp paired-end reads.

Sequence reads were mapped to the reference human genome assembly hg19 using Burrows-Wheeler Aligner (BWA-MEM) version 0.7.12 (24). PCR duplicates were removed using Picard 1.130 (http://broadinstitute.github.io/picard). The Genome Analysis Toolkit (GATK) version 3.4 was used for realignment of insertions and deletions, quality recalibration and variant calling (25). Then, to create a sub-selection of high-quality single nucleotide variations (SNVs), a series of filtering options was applied, including minimum read depth of 8x, SNV call quality score above 30, and non-reference allele frequency above 20%. Variants were annotated with information from RefGenes using wANNOVAR web-based tool (26).

A375 and A375R cells were suspended in CM, seeded (50 μl/well) into BD Falcon™ 96-well plates (BD Biosciences, Bedford, MA, USA) and allowed to adhere at 37°C in a 5% CO₂ atmosphere for 18 h. Graded amounts of dabrafenib or GSK2126458A were then added to the cells in 50 μl of CM. As a control, A375 and A375R cells were treated with DMSO alone. The plates were incubated at 37°C for five days and cell proliferation was then evaluated by the MTT assay, as previously described (27). Three replica wells were used for each group. Drug concentration producing 50% inhibition of cell growth (i.e. IC₅₀), was calculated on the regression line in which absorbance values at 595 nm were plotted against the logarithm of drug concentration. In a different set of experiments, A375R cells, plated as described above, were incubated with DMSO alone, 100 nM dabrafenib, 20 nM GSK2126458A or a combination of the two drugs for 5 days and then assayed for proliferation using the MTT assay.

A375 and A375R cells were suspended in CM, seeded into BD Falcon™ 10-cm-dishes (BD Biosciences) and allowed to adhere at 37°C in a 5% CO₂ atmosphere for 18 h. Thereafter, dabrafenib (100 nM) or DMSO alone was added to the cultures and the dishes incubated at 37°C for additional 6 or 48 h. In a different set of experiments, A375R cells, plated as described above, were incubated with DMSO alone, 100 nM dabrafenib, 20 nM GSK2126458A or a combination of the two drugs for 48 h. At the end of the incubation period, melanoma cells were processed for western blot analysis or tested for their ability to invade the extracellular matrix (ECM) in vitro.

For analysis of ERK1/2, phosphorylated ERK1/2, AKT and phosphorylated AKT expression, control and drug-treated cells were recovered from culture, washed and total cellular extracts were prepared as described previously (28). For analysis of VEGFR-2 expression, melanoma cells were left in the culture dishes, washed, lysed directly in SDS sample buffer (50 mM Tris-HCl pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue and 10% glycerol) and then boiled for 5 min. Proteins per sample (15 μg) were run on a 6% (VEGFR-2) or 12% SDS-polyacrylamide gels, transferred to nitrocellulose membranes (Amersham Biosciences, Buckinghamshire, UK) and blocked with 5% non-fat milk in Tris-buffered saline supplemented with 0.1% Tween-20 (TBST) for 1 h at room temperature. The membranes were then incubated in the same solution, or in TBST containing 5% bovine serum albumin (BSA), overnight at 4°C with primary antibodies at the following dilutions: anti-ERK1/2, anti-phospho-ERK1/2, anti-AKT and anti-β-tubulin 1:1,000; anti-phospho-AKT 1:500; anti-VEGFR-2 1:200. The anti-β-tubulin antibody was used as an internal standard for loading. Immunodetection was carried out using appropriate horseradish peroxidase-linked secondary antibodies and enhanced chemiluminescence (ECL) or ECL Prime (phospho-AKT) detection reagents (Amersham Biosciences). Where indicated, film was scanned on a GS-710 Calibrated Imaging Densitometer and analyzed by means of Quantity One Software Version 4.1.1 (Bio-Rad Laboratories).

This assay was performed as previously described (29). Briefly, control and drug-treated melanoma cells were removed from culture, washed, suspended in invasion medium (1 μg/ml heparin/0.1% BSA in RPMI-1640) and loaded (2×10⁵ cells) into the upper compartment of Boyden chambers equipped with 8-μm pore diameter polycarbonate filters (Nuclepore, Whatman Inc., Clifton, NJ, USA) coated with 20 μg of BD Matrigel™ Basement Membrane Matrix (BD Biosciences, hereafter referred to as Matrigel). Invasion medium or, where indicated, invasion medium containing 20 ng/ml VEGF-A (ImmunoTools, Friesoythe, Germany) was added to the lower compartment of the chambers. For each experimental condition, three Boyden chambers were set up. After incubation of the Boyden chambers at 37°C in a 5% CO₂ atmosphere for 4 h, the filters were removed from the chambers and the cells were fixed in ethanol for 5 min and stained in 0.5% crystal violet for 15 min. The cells from the upper surface of the filter were removed by wiping with a cotton swab and the migrated cells, attached to the lower surface of the filters, were counted under the microscope. Twelve microscopic fields (magnification, x200), randomly selected on triplicate filters, were scored for each experimental condition.

In a set of experiments, invasion assays were performed in the presence of the anti-human VEGF-A monoclonal antibody bevacizumab (5 μg/ml) (Avastin; Roche, Welwyn Garden City, UK) or the corresponding control IgG₁ immunoglobulins (MAB002; R&D Systems, Minneapolis, MN, USA). A375R cells were pre-incubated with the antibodies for 30 min at room temperature in a rotating wheel. The cells were then loaded in the Boyden chambers without removing the antibodies.

A375 and A375R spheroids were generated as described by Naber et al (30) with minor modifications. Briefly, melanoma cells were suspended (4×10⁴ cells/ml) in CM containing 0.24% methylcellulose (Sigma-Aldrich) and seeded (100 μl/well) into non-adhesive, round-bottom 96-well plates (Corning Inc., Corning, NY, USA). Plates were centrifuged at 1250 x g for 90 min and then incubated at 37°C in a 5% CO₂ atmosphere for 18 h to allow the cells to organize into three-dimensional spheroids (one spheroid in each well). Spheroids were then harvested and individually embedded between two layers of Matrigel as previously described (31) with minor modifications. Briefly, 50 μl of Matrigel diluted in RPMI-1640 medium (4 μg/ml), were added to the wells of a flat-bottom 96-well plate, that was then incubated at 37°C for 15 min to let the Matrigel solidify. Melanoma spheroids, suspended in 100 μl of Matrigel diluted in invasion medium (4 μg/ml), were then seeded in the Matrigel-precoated plate (one spheroid/well, five replicates). After a 30-min incubation at 37°C to allow Matrigel polymerization, spheroids were overlaid with 100 μl of invasion medium and the plate incubated at 37°C for 48 h. At the end of incubation, images of the invading spheroids were taken using a Leica inverted microscope and a Canon digital camera PowerShot G5.

Oligonucleotide siRNA targeting AKT1 (CUCACAGCCCUGAAGUACU) (siAKT) and the corresponding scramble oligonucleotide (GAUCCUAUAUUCGGUUAGU) (siCTRL) to be used as a control were purchased from Sigma-Proligo (The Woodlands, TX, USA).

A375R cells were suspended in CM without antibiotics, seeded into BD Falcon™ 6-well plates (BD Biosciences), allowed to adhere at 37°C in a 5% CO₂ atmosphere for 18 h, and then transfected with 10 nM siAKT or siCTRL using Lipofectamine™ RNAiMAX reagent (Invitrogen) according to the manufacturer’s protocol. Twenty-four hours after transfection, the cells were incubated with 100 nM dabrafenib or DMSO alone. After 48 h of culture, the cells were collected and processed for western blot analysis or tested for ability to invade the ECM.

A375 and A375R cells were suspended in CM, seeded into BD Falcon 10 cm-dishes (BD Biosciences) and allowed to adhere at 37°C in a 5% CO₂ atmosphere for 18 h. Thereafter, A375 cells were incubated with DMSO alone or 100 nM dabrafenib, whereas A375R were incubated with DMSO alone, 100 nM dabrafenib, 20 nM GSK2126458A or a combination of the two drugs. After 48 h of drug exposure, culture supernatants were collected, centrifuged at 600 x g for 10 min to remove cells in suspension and debris, and frozen at −20°C until use. Cells were detached with a solution of 1.5 mM EDTA in PBS and counted to determine the total number of cells in the culture. Quantification of the VEGF-A amount in the culture supernatants was performed as previously described (32). For MMP-9 evaluation, culture supernatants were concentrated at least 5-fold in Centriplus concentrators (Amicon, Beverly, MA, USA). The amount of active MMP-9 in the culture supernatants was then determined using the Quantikine ELISA Human MMP-9 immunoassay (R&D Systems) according to the manufacturer’s protocol, and calibrated against a standard curve. The amount of VEGF-A and MMP-9 was normalized to the number of total cells counted in each culture at the time of supernatant collection.

Statistical significance among different values of drug IC₅₀, different number of ECM invading cells and different levels of VEGF-A and MMP-9 secretion, was assessed using two-sided Student’s t-test analysis.

To generate a cellular model of acquired resistance to dabrafenib, the human melanoma cell line A375, which harbors the BRAF^V600E mutation and is highly sensitive to the growth suppressive effects of BRAF inhibitors, was cultured with gradually increasing concentrations of dabrafenib until a subline (i.e. A375R) able to growth in the presence of 1.5 μM of the drug was obtained. This concentration of dabrafenib corresponds to the peak plasma concentration reported for patients treated with the drug at the recommended dose of 150 mg BID (33).

A375 and A375R cells were initially analyzed for sensitivity to the antiproliferative effects of dabrafenib, and for expression of total and phosphorylated ERK1/2, either under basal conditions or following a 6-h exposure to 100 nM of the drug. MTT assays, performed after five days of cell culture with graded concentrations of dabrafenib, revealed that the A375R subline was 10,000-fold more resistant to the growth inhibitory effect of this drug than the parental cell line (Fig. 1A). A375R cells also expressed higher basal levels of phosphorylated ERK1/2 as compared with A375 cells (Fig. 1B). Moreover, while dabrafenib, as expected, markedly downregulated the level of phosphorylated ERK1/2 in A375 cells, it further increased it in the drug-resistant subline (Fig. 1B).

Previous studies have shown that melanoma cells with acquired resistance to the BRAFi vemurafenib, PLX4720 or SB590885 possess a more invasive phenotype (18,20–22), whereas no data are available regarding the metastatic potential of dabrafenib-resistant cells. Therefore, we evaluated the ability of A375 and A375R cells to spontaneously invade the ECM using the Boyden chamber assay. The results illustrated in Fig. 1C show that the dabrafenib-resistant A375R cells were ~2-fold more invasive than the dabrafenib-sensitive A375 cells.

The invasive behavior of A375 and A375R cells was also investigated using an invasion assay based on Matrigel-embedded spheroids, that mimics the three-dimensional tumor architecture and incorporates components of the ECM (34). Consistent with results obtained in Boyden chamber assays A375R spheroids invaded into Matrigel to a higher extent as compared with parental A375 cells (Fig. 1D).

Finally, to investigate whether reactivation of the MAPK/ERK pathway in A375R cells could be dependent on mutations in HRAS, KRAS, NRAS, MAP2K1, MAP2K2 or on NF1 loss, whole exome sequencing was performed in both A375 and A375R cell lines. No mutations in the three RAS genes or in MAP2K1 and MAP2K2, and no loss of NF1 were detected in A375 and A375R cells.

We previously showed that an autocrine loop sustained by the interaction of VEGF-A with VEGFR-2 promoted melanoma cell ability to migrate and invade the ECM in vitro (35). Recently, Liu et al (36) demonstrated that PLX4720 inhibited VEGF production in a melanoma cell line and xenografts, and that downregulation of VEGF expression occurred in the tumor of melanoma patients treated with PLX4720. Furthermore, Beazley-Long et al (37) reported that vemurafenib decreased VEGF-A expression in BRAF-mutant and drug-sensitive melanoma cells, whereas it increased this cytokine in BRAF-wild-type melanoma cells. Therefore, to investigate the effects of dabrafenib on the metastatic potential of melanoma cells, A375 and A375R cells were treated with 100 nM of the drug (or with DMSO as vehicle control) for 48 h and then analyzed for the ability to invade the ECM under basal conditions and in response to exogenously added VEGF-A, as well as for VEGF-A secretion and VEGFR-2 expression.

The results illustrated in Fig. 2A show that both cell lines responded to VEGF-A with an increase of ECM invasion, and that not only spontaneous but also VEGF-A-induced invasive capacity was higher in the A375R cell line as compared with the parental counterpart. Dabrafenib significantly inhibited spontaneous and VEGF-A-induced ECM invasion in A375 cells. In contrast, A375R cell invasiveness was markedly stimulated by the drug in both conditions.

VEGF-A secretion in the culture supernatants of A375 and A375R cells, and VEGFR-2 expression in the two cell lines were evaluated by specific ELISA and western blot analysis, respectively. We found that basal secretion of VEGF-A was higher in A375R cells as compared with parental A375 cells (Fig. 2B). Moreover, VEGF-A secretion was markedly impaired in dabrafenib-treated A375 cells, whereas it was significantly increased in drug-treated A375R cells. Exposure to dabrafenib caused a reduction of VEGFR-2 expression in A375 cells but it did not substantially affect the receptor levels in A375R cells (Fig. 2C). Of note, the effects of 1.5 μM dabrafenib on A375R invasiveness, VEGF-A secretion and VEGFR-2 expression were comparable to those produced by 100 nM of the drug (data not shown).

The findings that A375R cells were more invasive and secreted higher levels of VEGF-A than A375 cells, and responded to dabrafenib with a further increase of ECM invasion and VEGF-A secretion, prompted us to investigate whether inhibition of VEGF-A binding to its cognate RTKs could impair spontaneous and dabrafenib-induced invasiveness of A375R cells. To this end, the cells were treated with 100 nM dabrafenib or DMSO alone for 48 h, and then assayed for ECM invasion in the absence or in the presence of the anti-VEGF-A monoclonal antibody bevacizumab (38,39), or of a control antibody. As illustrated in Fig. 3, bevacizumab did not affect ECM invasion of DMSO-treated cells, whereas it abrogated dabrafenib-induced stimulation of A375R cell invasiveness.

VEGF-A binding to VEGFR-2 can activate a number of downstream signaling pathways, including the MAPK/ERK and PI3K/AKT cascades (40,41). Moreover, previous studies (42) demonstrated that VEGF-A overproduction and concomitant expression of its receptors promote melanoma cell growth and survival through the activation of the MAPK/ERK and PI3K/AKT signaling pathways. We therefore investigated whether, in addition to reactivation of the MAKP/ERK signaling in the presence of dabrafenib, A375R cells also showed hyperactivation of the PI3K/AKT pathway. Western blot analysis of total and phosphorylated AKT in A375 and A375R cells, showed that the drug-resistant subline expressed higher basal levels of phosphorylated AKT as compared with the parental cell line (Fig. 4A). Moreover, in both cell lines exposure to 100 nM dabrafenib for 48 h enhanced the levels of phosphorylated AKT (Fig. 4A). Exome sequencing performed in A375 and A375R cells did not reveal mutations in the AKT1, AKT3, PIK3CA, PIK3CG, PIK3R2 and PHLPP1 genes, or PTEN loss.

To establish whether hyperactivation of the PI3K/AKT pathway could underlie the increased invasive capacity of A375R cells, AKT1 was silenced in these cells using specific siRNAs. The results illustrated in Fig. 4B show that total AKT levels were efficiently reduced in the siAKT-transfected cells exposed to DMSO or to 100 nM dabrafenib for 48 h. Notably, DMSO-treated and dabrafenib-treated siAKT-transfected cells also showed a marked reduction of phospho-AKT levels. This finding indicates that in A375R cells, basal and dabrafenib-induced phosphorylation of AKT mainly involve the AKT1 isoform. As observed for bevacizumab, AKT1 silencing failed to reduce basal invasiveness of A375R cells, whereas it completely abrogated upregulation of ECM invasion induced by dabrafenib (Fig. 4C).

Based on the key role played by MAPK/ERK signaling reactivation and PI3K/AKT/mTOR signaling upregulation in acquired resistance to BRAFi, several therapeutic approaches co-targeting the MAPK/ERK and the PI3K/AKT/mTOR pathways in patients with BRAF-mutant melanoma are underway to improve the efficacy and durability of clinical response (12; https://www.clinicaltrials.gov/).

We therefore investigated whether co-treatment of A375R cells with the dual PI3K/mTOR inhibitor GSK2126458A could antagonize the stimulating effects exerted by dabrafenib on ECM invasion and VEGF-A secretion. Preliminary experiments were also performed to evaluate A375R cell sensitivity to the growth inhibitory effect of GSK2126458A and to select the drug concentration to be used in combination with dabrafenib.

In agreement with previous findings (43), MTT proliferation assays, performed after five days of cell exposure to graded concentrations of GSK2126458A, revealed that the dabrafenib-resistant subline was responsive to the PI3K/mTOR inhibitor, displaying a drug IC₅₀ value comparable to that of the parental cells (Fig. 5A).

The clinically achievable concentration of 20 nM GSK2126458A was then selected for the subsequent experiments. This drug concentration inhibited A375R proliferation by ~70%, either alone or in combination with dabrafenib (Fig. 5B). Moreover, a 48 h-exposure to 20 nM GSK2126458A completely abrogated AKT phosphorylation in both DMSO-treated and dabrafenib-treated A375R cells (Fig. 5C) without affecting cell viability (data not shown).

A375R cells were exposed to DMSO, 20nMGSK2126458A, 100 nM dabrafenib or a combination of the two drugs for 48 h and then tested for ECM invasion and VEGF-A secretion. Invasiveness of A375R cells was not substantially affected by GSK2126458A alone (Fig. 6A). However, the PI3K/mTOR inhibitor abrogated dabrafenib-induced stimulation of ECM invasion (Fig. 6A). In agreement with the invasion assays, VEGF-A secretion was not significantly altered in A375R exposed to GSK2126458A alone (Fig. 6B). On the other hand, the PI3K/mTOR inhibitor was able to completely counteract the upregulation of VEGF-A release caused by the BRAFi (Fig. 6B).

MMPs, a large family of proteolytic enzymes able to degrade the ECM, play a critical role in many physiological processes and are also implicated in tumor angiogenesis and metastatic spreading of cancer cells (reviewed in refs. 44,45). In melanoma, increased expression of MMP-1, MMP-2 and MMP-9 has been associated with an invasive phenotype (46). Moreover, VEGF-A has been reported to upregulate the expression of certain MMPs, including MMP-9, in endothelial cells and ovarian cancer cells (47,48).

We, therefore, determined MMP-9 levels in culture supernatants of A375 and A375R cells, treated with DMSO or 100 nM dabrafenib for 48 h. In A375R cells, MMP-9 secretion was also evaluated after treatment with 20 nM GSK2126458A, alone or in combination with dabrafenib.

The results illustrated in Fig. 7 show that A375 cells secreted low amounts of MMP-9, which were not affected by dabrafenib treatment. A375R displayed levels of MMP-9 secretion 2-fold higher than those detected in A375 cells. Moreover, in the resistant subline the release of MMP-9 was significantly stimulated by dabrafenib. Similarly to the results obtained with VEGF-A, GSK2126458A alone did not alter basal MMP-9 secretion in A375R cells, but abrogated the upregulation caused by dabrafenib treatment (Fig. 7).