Cell culture

HCmel cell lines were derived from Hgf-Cdk4R24C mice as previously described [6, 11]. BCmel4 was derived from the BrafV600E-Cdk4R24C model as previously described [12]. Melan-a cells were kindly provided by D. Bennett (London, United Kingdom). All HCmel cell lines, BCmel4 and Melan-a cells were cultured in RPMI 1640 medium (Life Technologies, Carlsbad, CA) supplemented with 10% fetal bovine serum (Biochrome, Berlin, Germany), 2 mM L-glutamine, 10 mM non-essential amino acids, 1 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 100 U/mL penicillin and 100 μg/mL streptomycin (all from Life Technologies, Carlsbad, CA) and 20 μM 2-mercaptoethanol (Sigma, St. Louis, MO) in a humidified incubator with 5% CO2 at 37 °C. For culturing of Melan-A cells additionally, 200 nM Phorbol 12-myristate 13-acetate (PMA) (Sigma) was supplemented into the medium. All cells were cultured at 37 °C and 5% CO2 in a humid environment.

For pharmacologic manipulation, cells were treated with recombinant mouse HGF protein (R&D, Minneapolis, MI), the MET inhibitor capmatinib (Cayman Chemical, Ann Arbor, MI), and the Gnaq-inhibitor FR900359 (FR) (kindly provided by E. Kostenis, Bonn, Germany) as described in the figure legends. Vehicle controls were performed with DMSO (Sigma) in a final concentration of 0.1%. For the retrovirus production, HEK293T (RRID: CVCL_0063) cells, obtained from the American Type Culture Collection (ATCC), were maintained in DMEM medium (Life Technologies) containing the same supplements. Cells were freshly thawed every 2 months. All cell lines used in our study were routinely tested for mycoplasma contamination via polymerase chain reaction (PCR) monthly. All experiments were performed with mycoplasma-free cells.

Cloning of pRP-mGnaqWT/mGnaqQ209L-T2A-mcherry plasmids and retroviral gene transfer

For generation of the retroviral overexpression constructs of GnaqWT and GnaqQ209L, the Gnaq cDNA was amplified from cDNA from HCmel3 cells, which are heterozygous for Gnaq209L using primers C1 and C2 (see Table S1). A fragment encoding T2A-mCherry was amplified via PCR using the plasmid pCAS9-mCherry-Frame +0 (Addgene #66939) containing mCherry as the template using the primers C3 and C4 (containing the T2A-sequence, see Table S1). The retroviral backbone pRP (Addgene #41841) was linearized using the restriction enzyme XhoI. The correct size of all fragments was verified via agarose gel electrophoresis with subsequent isolation of DNA using the NucleoSpin Gel and PCR clean-up kit (Macherey Nagel, Düren, Germany) according to manufacturer’s instructions. Subsequently, all fragments were assembled in a sequence-independent cloning reaction using the Gibson assembly master mix (New England Biolabs, Ipswich, MA) according to manufacturer’s recommendations.

HEK293T cells were transfected with the retroviral helper plasmids (gag-pol [Addgene plasmid #14888] and pCMV VSV-G [Addgene plasmid #14887], kindly provided by E. Latz, Bonn, Germany) and pRP-TagGFP2 by calcium phosphate transfection according to standard protocols. Retrovirus-containing supernatant was filtered using a 0.45 μM pore size syringe filter and added to target cells. Selection with 10 μg/mL puromycin (AppliChem, Darmstadt, Germany) was started 48 hours after transduction for 5 days. Successful transduction was confirmed by red fluorescence using a Zeiss AxioVert A1 microscope (Zeiss, Oberkochen, Germany).

Crystal violet assay

Cells were seeded at a density of 2 × 104 cells/well in biological triplicates in a 48-well plate and treated as indicated the next day. Subsequently, cells were incubated as described above. After this time, media was removed, and wells were washed with 1× phosphate-buffered saline (PBS) to remove dead cells. All wells were stained with 150 µL of 0.5% crystal violet in aqueous solution for 10 min, repetitively washed three times with water, and dried at room temperature for 48 hours. Images were acquired with an EOS 750D camera (Canon, Tokyo, Japan).

Immunoblot analysis

Whole-cell lysates were extracted from cultured cells using the M-PER mammalian protein reagent supplemented with protease and phosphatase inhibitors (all Thermo Fisher Scientific, Waltham, MA) according to the manufacturer’s protocol. Protein concentrations were quantified with a Pierce BCA Protein Assay Kit (Thermo Fisher Scientific) at 562 nm in a microplate reader (Tecan Group, Männedorf, Switzerland). Samples were prepared containing 2 μg protein, Roti-Load (Carl Roth, Karlsruhe, Germany), and nuclease-free water and denatured at 95 °C prior to loading. In the next step, samples were separated via 10% SDS-PAGE and transferred to a 0.45 μM polyvinylidene difluoride membrane (GE Healthcare, Boston, MA) by wet blotting (BioRad, Hercules, CA). After blocking with 5% milk (Carl Roth), blots were immunostained at 4 °C overnight. Bound antibodies were detected with horseradish peroxidase (HRP)-conjugated secondary antibodies and the SignalFire ECL Reagent (Cell Signaling, Danvers, MA) according to the manufacturer’s instructions. Chemiluminescence was visualized using an Octoplus QPLEX-Imager (NH DyeAGNOSTICS, Halle, Germany). Used antibodies were as follows: rabbit anti-mouse phospho-Met antibody (Cell Signaling, #3077), rabbit anti-mouse phospho-c-Met polyclonal antibody (Thermo Fisher Scientific, #44-888 G), mouse anti-mouse Met monoclonal antibody (Cell Signaling, #3127), mouse β-Actin monoclonal antibody (Santa Cruz, Dallas, TX, sc-47 778) and goat anti-rabbit IgG HRP-linked antibody (Cell Signaling, #7074 S).

Next-generation sequencing and data analysis

DNA was isolated from mouse tumor tissue or cultured cells using the NucleoSpin Tissue kit according to manufacturer’s instructions (Macherey Nagel). Target DNA regions for Gnaq/11 Exon 4 and 5, as well as Trp53 Exon 4–7 were amplified and supplemented with unique dual sequencing indices and adapters in a two-step PCR (Primer sequences see Table S1). Samples were multiplexed and sequenced on an Illumina MiSeq in paired-end mode, sequencing length of 150 bp and minimum depth of 100×. Samples were demultiplexed based on their unique sequencing indices, and quality control was performed separately for each sample using fastqc. Samples were aligned to the GRCm39 reference genome using bwa in mem-mode. Variant calling was performed with mutect2 without positional downsampling.

Analysis of published datasets

Met copy number and transcript data were obtained from the DepMap portal from the 2023Q2 release [13]. Whole-genome sequencing data from HCmel12 cells as published in [14], available under the accession SRP247646 were downloaded from the Sequence Read Archive as fastq files. Raw reads were aligned to the mm9 reference genome using BWA (version 0.7.17) using the BWA-MEM algorithm with default parameters. Duplicate marking was performed with GATK (version 4.4.0.0) using the command MarkDuplicatesSpark with default parameters. For variant calling, cnvkit (version 0.9.10) was used using the batch command using a “flat” reference of neutral copy number by providing the “n” flag without specifying a normal sample. Discrete copy number segmentation was performed with cnvkit segment using the cbs method and the “drop-low-coverage” flag. The absolute integer copy number was assigned using cnvkit call using the “clonal” method and assuming a purity of 95%.

Comparative genomic hybridization (CGH)

DNA was fluorescence-labeled using genomic DNA universal linkage system labeling kits (Agilent Technologies, Böblingen, Germany) and was hybridized using an Agilent mouse genome CGH 2x105k microarray (Agilent Technologies) according to the manufacturer’s instructions. The resolution of the CGH arrays was 2×105,000 oligonucleotides distributed genome-wide. The arrays were scanned using a DNA microarray scanner (Agilent Technologies), and the images were analyzed using Feature Extraction, version 10.5.1.1, and DNA Analytics software, version 4.0.85 (Agilent Technologies) based on the mouse genome build mm8.

Animal experiments

Hgf-Cdk4- and Cdk4 mice were taken from their own breeding [7, 15, 16]. Age- and sex-matched cohorts of mice were randomly allocated to the different experimental groups at the start of each experiment. Experiments were performed with 6-8-week-old mice. Induction of primary melanoma was performed with a single dose of 100 nmol DMBA dissolved in acetone applied to the shaved back skin as previously described [11]. UV irradiation was performed twice weekly with 4.5 kJ/m2 dose applied to the shaved back using a UV 302 L system (Waldmann, Villingen-Schwenningen, Germany) equipped with eight 100 W UV21 lamps (Phillips, Amsterdam, Netherlands) [6]. For tumor cell transplantation, cohorts of syngeneic Hgf-Cdk4- and Cdk4 mice were injected intracutaneously with 2 × 105 cells resuspended in 100 μL PBS (Life Technologies) into the right flank. Tumor growth was monitored by inspection and palpation. Tumor size was measured at least twice times weekly with a vernier caliper and recorded as the mean diameter of two perpendicular measurements. Mice were sacrificed when tumors exceeded 20 mm in diameter or when signs of illness were observed. All experiments were performed in groups of five or more mice and repeated independently three times. The sample size for animal experiments was determined based on our own previous experience of tumor engraftment and growth rates. No animals were excluded from the analysis. Tissue samples for histological analysis were fixed in formalin free Zinc-fixative (BD Bioscience) and subsequently prepared for sectioning and staining using routine histological techniques. Images were obtained using an AxioVert A1 Microscope (Zeiss). All experiments were performed in compliance with federal and international guidelines for animal experiments and with the approval of the responsible authorities (Landesverwaltungsamt Saxony-Anhalt, Germany, approval number: 42502-2-1556 UniMD).

Statistical analysis

Survival analyses were performed using Kaplan-Meier estimators with pairwise log-rank-tests between experimental cohorts. Tumor growth curves for individual tumors are additionally shown to demonstrate a similar variance between experimental cohorts. All bioinformatical and statistical analyses were performed using Python 3.9 and the packages biopython, pandas, scikit-allel, and lifelines.