Materials availability
Materials associated with this manuscript will be available upon request. Some plasmid constructs will be subjected to an MTA due to the backbone use restrictions.
Reagents
4-Hydroxytamoxifen (4OHTx) was obtained from Sigma (St. Louis, MO, USA). Primary antibodies against Melan-A (MA5-15237; IHC 1:100), KIF14 (PA5-87769; WB 1:1000; IHC 1:100), p16Ink4a (MA5-17142; IHC 1:100), S100B (#710363; IHC 1:100), CHTF18 (A301-883A; IHC 1:50), BNC1 (PA5-85984; IHC 1:200) and p21CIP1/WAF1 (MS-891-P1; WB 1:1000) were purchased from Thermo Scientific (Waltham, MA, USA). The primary antibody against BNC1 (ARP33283; WB 1:1000) was obtained from Aviva Biosciences (San Diego, CA, USA). The primary antibody against NCKAP5 (STJ196143; WB 1:1000) was obtained from St. John’s laboratory (London, UK). Primary antibodies against PARP1 (#9542; WB 1:1000), Caspase 6 (#9762; WB 1:1000), Cyclin-D1 (#2926; WB 1:1000), p-RbS795 (#9301; WB 1:1000), p-RbS807/811 (#9308; WB 1:1000) and p27KIP1 (#2552; WB 1:1000) were purchased from Cell Signaling (Danvers, MA, USA). Primary antibodies against CHTF18 (sc-374632; WB 1:500), PHGDH (sc-100317; IHC 1:20; WB 1:500), and PFK1 (sc-67028; WB 1:1000) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Primary antibodies against GAPDH (60004-1; WB 1:10,000), UPP1 (14186-1-AP; WB 1:000; IHC 1:50), PRKG2 (55138-1-AP; WB 1:1000; IHC 1:50), and RB1 (10048-2-Ig; WB 1:1000) were obtained from Proteintech (Rosemont, IL, USA). Primary antibodies against NCKAP5 (HPA034639; IHC 1:100), Cyclin-B1 (05-373; WB 1:1000), and β-Actin (A3854; WB 1:10,000) were obtained from Merck (Darmstadt, Germany). Primary antibodies against Ki67 (ab16667; IHC 1:200) and Cyclin-E (ab7959; WB 1:1000) were purchased from Abcam (Cambridge, UK). The primary antibody against BRAFV600E (#760-5095) was obtained from Ventana Medical Systems. Primary antibodies against TRP1 (Pep1) and TRP2 (Dct; Pep8) were obtained from Dr. Vincent Hearing, National Institutes of Health, and each was used at a 1:2000 dilution [67]. Fluorescent secondary antibodies were purchased from Thermo Fisher Scientific.
Mouse model
The Tyr::CreERT2,BrafCA strain [10], and the Tyr::CreERT2;BrafCA;Lkb1F/F strain [68] were previously described. Both sexes were used for experiments. All mice were maintained on a mixed C57BL/6, FVB, and 129 background. All mice were cared for and maintained in accordance with animal welfare regulations under a protocol approved by the Institutional Animal Care and Use Committee of Vall d’Hebron Institut de Recerca (VHIR) and Biomedical Research Park of Barcelona (PRBB).
Melanoma initiation in vivo
Mice were treated topically on postnatal days 2.5 and 3.5 with 100 µl of an acetone solution containing 100 mg/ml 4OHTx. Neonatal mice were irradiated on postnatal day 3.5, as previously described [15]. The exposure time was 15 min for a total dose of 9.6 kJ/m2 UVB (280–330 nm) using TL 40w −12 RS lamps. Melanoma-free survival was analyzed by the Kaplan‒Meier analysis.
Primary melanocyte culture
For both models (Tyr::CreERT2;BrafCA/+ generated by crossing Tyr::CreERT2;BrafCA/CA X WT (obtained by crossing BrafCA/+X BrafCA/+) and Tyr::CreERT2;BrafCA/+;Lkb1F/F generated by crossing Tyr::CreERT2;Lkb1F/F X Tyr::CreERT2;BrafCA/CA;Lkb1F/F), dorsal skin samples were collected from neonatal mice (three pups per litter, with at least three litters per genotype) following a protocol adapted from [12]. In brief, the skin was removed from 1- or 2-days-old pups, sterilized by immersion in 70% ethanol for 5 s, and immediately washed with PBS. The skin was then incubated with 0.05% trypsin-EDTA (Gibco, Waltham, MA, USA) at 37 °C in a 5% CO2 incubator for 1–2 h. Next, the epidermis was separated from the dermis using sterile forceps and mechanically digested with a scalpel. Melanocytes were then cultured in selective medium consisting of Dulbecco’s modified Eagle’s medium (DMEM)-F12 (Biowest, Riverside, MO, USA) supplemented with 5% fetal bovine serum (FBS; Gibco), 200 pM cholera toxin (Sigma), 200 nM phorbol esters (12-O-tetradecanoylphorbol-13-acetate; TPA) (Sigma), 2 mM L-glutamine (Gibco), 10 U/ml penicillin and 10 μg/ml streptomycin (Gibco).
Cell lines and treatments
Normal human epidermal melanocytes (NHEMs; C-12400) and human neonatal melanocytes (MEL-F-NEO cells) were purchased from PromoCell (Heidelberg, Germany) and Zenbio (Durham, NC, USA), respectively, and cultured following the manufacturers’ recommendations. UACC903 (CVCL_4052) cells were a gift from J. Trent (Tgen, Phoenix, AZ, USA). A375 (CRL-1619), SKMel28 (HTB-72), and G361 (CRL-1424) cells were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). IMR90 cells were obtained from M. Abad (VHIO, Barcelona, Spain). Patient-derived cell lines, including MMLN9, MMLN10, MMLN14, MMLN16, MMLN23, and MMLN24, were derived from cells obtained from patients after tumor surgery [69, 70]. All the samples were obtained and used with informed consent from the patients and approval from the Vall d’Hebron Hospital Ethical and Clinical Research Committee (CEIC) (PR(AG)115/2013). Patient-derived cells were cultured in DMEM (Biowest) supplemented with 20% FBS (Gibco), 2 mM L-glutamine (Gibco), 100 U/ml penicillin, 100 μg/ml streptomycin (Gibco), and 5 μg/ml Plasmocin (InvivoGen, Toulouse, France). A375 and G361 cells were grown in DMEM. UACC903 cells were cultured in RPMI 1640 medium (Biowest). SKMel28 and IMR90 (CCL-186) cells were cultured in Eagle’s minimum essential medium (EMEM) (ATCC). Unless otherwise specified, all media were supplemented with 10% FBS, 2 mM L-glutamine, 100 U/ml penicillin, 100 μg/m streptomycin, and 5 μg/ml Plasmocin. All cells were maintained at 37 °C in a 5% CO2 incubator. 5-Benzylacyclouridine (BAU) was obtained from MedChemExpress (Monmouth Junction, NJ, USA) and used at the indicated concentrations. Isopropyl-beta-D-1-thiogalactopyranoside (IPTG) was purchased from Gold Biotechnology (St. Louis, MO, USA) and used at 500 µM.
Mouse model-derived melanocytes were treated with 4OHTx for in vitro recombination and exposed to UVR. Tamoxifen was added at a concentration of 1 μM for 48 h. For UVR exposure, a Stratagene UV Crosslinker 1800/2400 was used. After the removal of the medium and the plate lid, cells at 60–70% confluence were irradiated with 30 J/m2 UVB. For combined treatments, cells were irradiated 48 h after 4OHTx treatment. After treatment, fresh medium was added, and the cells were incubated for 27 days prior to RNA isolation. During the incubation period, the culture medium was replaced every 2–3 days.
Human melanoma samples
All pseudonymized human melanoma samples were provided by the Vall d’Hebron Research Institute under the National Research Ethics Service approved study number PR(AG)59-2009. Informed consent was provided by all patients.
Molecular cloning
The pLenti-IRES-GFP plasmid was obtained from Stephan Tenbaum (Vall d’Hebron Institute of Oncology VHIO). The human UPP1, CHTF18, and BNC1 sequences were subcloned from pCMV-SPORT6-UPP1, pCMV-SPORT6-CHTF18, and pLOC-BNC1, respectively (Dharmacon, Cambridge, UK), to construct pLenti-rtTA2-UPP1-IRES-GFP, pLenti-rtTA2-CHTF18-IRES-GFP, and pLenti-rtTA2-BNC1-IRES-GFP. All short hairpin RNAs (shRNAs) against PHGDH, UPP1, PRKG2, KIF14, NCKAP5, CHTF18, and BNC1 were purchased from Sigma (Table S11). The backbone vectors of these plasmids were pLKO.1. for constitutive expression plasmids and pLKO-IPTG-3xLac0 for inducible expression plasmids. The BRAFV600E sequence was obtained from pCR4-BRAFV600E and was subcloned and inserted into pLTPC-hPGK-rtTA2-p2A-mCherry-TRE, which was kindly provided by the Hector Palmer laboratory.
Lentiviral transduction: overexpression and shRNA-mediated gene knockdown
To produce lentiviruses, three million HEK293T cells were seeded in p100 plates (Sarstedt) in 8 ml of DMEM supplemented with 10% FBS, 2 mM L-glutamine, 100 U/ml penicillin-100 μg/ml streptomycin, and 5 µg/ml Plasmocin. One hour before transfection, the medium was replaced with DMEM supplemented with 10% FBS (complement inactivated) and chloroquine (Sigma) at a final concentration of 15 μM. Then, 40 µg of total DNA (20 μg of lentiviral vector, 15 μg of pPAX2, and 5 μg of pMD2.G) was added to 150 mM sodium chloride (NaCl) to a final volume of 800 μl. Then, 200 μl of polyethylenimine (PEI) transfection reagent (1 mg/ml) (Sigma) was added to the DNA mixture, and the mixture was incubated at room temperature for 15 min. The mixture was added to cells, and the cells were incubated overnight at 37 °C. The medium was then replaced with DMEM supplemented with 2% FBS (complement inactivated) and 5 mM sodium butyrate (Sigma). After 24 h, the virus-containing supernatant was collected, and fresh medium was added to the cells. The collected supernatant was centrifuged for 5 min at 1000 rpm and passed through 0.2 µm filters (Sarstedt). The cells were incubated for another 24 h, after which the collection procedure was repeated. For cell infection, the processed supernatant containing either 4 µg/ml (MEL-F-NEO) or 8 µg/ml (UACC903, SKMel28, A375, and G361) polybrene (SCBT) was added to the cells. The infection efficacy was assessed after the addition of doxycycline (1 µg/ml) (Sigma) through detection of green (GFP) and/or red (mCherry) signals under a fluorescence microscope.
Chromogenic assay for β-galactosidase activity
Senescence-associated β-galactosidase (SA-βgal) activity was measured following the protocol of Debacq-Chainiaux et al. (Debacq-Chainiaux et al. 2009). In brief, melanocytes were isolated from the skin of neonatal mice and seeded directly on a 24-well plate (Sarstedt) at 50% confluence. After attachment, the melanocytes were treated with 4OHTx (1 µM) and/or exposed to UVR (30 J/m2) and incubated for 7 days. Before performing the assay, the cells were rinsed twice with PBS and fixed with PBS containing 2% formaldehyde-0.2% glutaraldehyde (Sigma) for 5 min at room temperature. After two washes with PBS, the cells were stained with a freshly prepared 5-bromo-4-chloro-3-indolyl-β-D-galactoside (X-Gal) staining solution containing 40 mM citric acid/Na buffer, 5 mM potassium ferrocyanide, 150 mM NaCl, 2 mM MgCl2, and 1 mg/ml X-Gal, all of which were purchased from Sigma. The cells were incubated overnight at 37 °C in the dark in a CO2-free incubator. β-Galactosidase activity was also determined in nonmalignant melanocytes overexpressing BRAFV600E alone or in combination with UPP1, CHTF18, or BNC1. For determination of the β-galactosidase activity of these cells, we followed the same procedure described above.
Immunocytochemical (IC), immunofluorescence (IF), and immunohistochemical (IHC) analyses
For IC analysis, mouse-derived melanocytes were seeded directly on coverslips in a 24-well plate (Sarstedt) at 50% confluence, incubated overnight, and, when indicated, subjected to treatment the next day. Cell staining was performed as previously described (Andreu-Pérez et al. 2011). For IHC analysis, the entire dorsal skin of the animals was collected (four animals per genotype and condition). For IF and IHC analyses, 4 μm sections of formalin-fixed, paraffin-embedded tumor samples were stained according to the antibody manufacturer’s protocol. Antibody binding was visualized with the UltraViewTM Universal DAB Detection Kit (Ventana Medical Systems) by using either horseradish peroxidase-conjugated or fluorophore-conjugated secondary antibodies. Staining was performed either manually or on an automated immunostainer (Beckmarck XT, Ventana Medical Systems, Roche, Tucson, AZ, USA). For the manually processed samples, antigen retrieval was performed using target retrieval solution (pH 6.0; Agilent, Santa Clara, CA, USA). Hematoxylin and eosin (H&E) staining was performed to evaluate the structure of the tissue. The samples were scanned (panoramic slide digital scanner) and evaluated by two independent pathologists (using 3DHistech software).
Protein isolation and western blotting
Cells were lysed in protein lysis buffer (150 mM NaCl, 50 mM Tris, 1 mM EDTA, and 1% NP-40) supplemented with protease and phosphatase inhibitor cocktails (Millipore, Burlington, MA, USA). Then, equal amounts of protein were separated by SDS‒PAGE and transferred to a PVDF membrane. Immunoblotting was performed as previously described [69]. The membrane was incubated at 4 °C overnight with primary antibodies and was then incubated for 1 h at room temperature with secondary antibodies. β–Actin and GAPDH were used as loading controls.
RNA Isolation, quantification, and quality control
For RNA purification, a Direct-Zol RNA Kit (Zymo Research, Irvine, CA, USA) was used. The isolation procedure was performed according to the manufacturer’s recommendations. The isolated RNA was analyzed with a 2100 Bioanalyzer (Agilent) to confirm the quantity and integrity of the sample. Only high-quality samples were used for subsequent analyses.
Microarray analysis
The mouse transcriptome was analyzed using a Clariom S Mouse Array (Affymetrix Santa Clara, CA, USA) (three biological replicates per condition). The raw data generated from the array were processed using the open-source software R Studio. Intensity values were converted into gene expression values using the robust multiarray average (RMA) algorithm, which consists of background correction, logarithmic transformation, quantile normalization, and probe normalization steps, via the Bioconductor package oligo. Identification of differentially expressed genes (DEGs) was carried out using the limma package. A moderated t-test was applied for each comparison based on the empirical Bayes method. Then, we generated lists of upregulated (log2FC > 0.265) and downregulated (log2FC < −0.265) genes comparing Control, B and B + UVB groups (Table S4) and B, B + UVB and B;L groups (Table S6). For functional analysis of the identified DEGs, Metascape (metascape.org) (Zhou et al., 2019) and the Molecular Signatures Database (MsigDB; gsea-msigdb.org) (Subramanian, Tamayo et al., 2005; Liberzon et al., 2011) were used. Heatmaps were generated using Morpheus (software.broadinstitute.org/morpheus). Hierarchical cluster analysis was performed based on the average pairwise distance calculated by Pearson correlation analysis.
Quantitative reverse transcription (RT)‒PCR
A total of 250 ng of purified RNA was reverse transcribed to cDNA using the SuperScriptTM III First-Strand Synthesis Kit (Thermo Fisher). Quantitative PCR analysis was performed using SYBRTM Select Master Mix (Applied Biosystems, Waltham, MA, USA), and the reactions were carried out in 384-well plates using a 7900 HT system (Applied Biosystems). The primers used were designed using BiSearch software (Arányi T et al., 2006), and the sequences are listed in Table S12. The ΔCt method was used to calculate relative expression values, which were normalized to the expression of the housekeeping genes Actb, Gapdh, and Vim. The data are presented as the mean ± SD of triplicate samples from three independent experiments.
Single-cell RNA sequencing (scRNA-seq) and data analysis
Human neonatal melanocytes (MEL-F-NEO) were simultaneously transduced with two different lentiviral constructs: pLenti-hPGK-rtTA2-p2A-mCherry-TRE-BRAFV600E and pLenti-hPGK-rtTA2-TRE-UPP1/CHTF18/BNC1-IRES-GFP. The expression of BRAFV600E, UPP1, CHTF18, and BNC1 was induced by the addition of doxycycline (1 µM) to the culture medium. After 27 days of induction, cells exhibiting red fluorescence (expressing BRAFV600E) and cells exhibiting both red and green fluorescence (expressing BRAFV600E and UPP1, CHTF18, or BNC1) were sorted by flow cytometry using Aurora CS spectral technology (Cytek).
The raw sequencing data (fastq format) were processed using 10× Genomics Cell Ranger 7.0.0 to align the reads to the GRCh38 human genome. After obtaining the count matrices, low-quality cells were filtered out. Cells with fewer than 1000 or more than 35,000 unique molecular identifiers (UMIs), fewer than 500 or more than 7000 detected genes, or an abundance of mitochondrial DNA greater than 20% of the transcriptome were excluded from the downstream analysis.
BRAF and pooled samples were combined for joint analysis. We merged both count matrices and, using the Seurat package (version 4 [71] i), we normalized the data, determined the top 3000 highly variable genes, and scaled the data. We also performed principal component analysis (PCA). We then ran Harmony ([72] ii) to remove the influence of the dataset of origin from the embedding and downstream analysis. We obtained the Harmony-corrected PCA embedding and used the top 25 corrected principal components to generate the uniform manifold approximation and projection (UMAP) plot for visualization. To cluster the cells, we first generated the 20 nearest neighbors graph and performed clustering using the original Louvain algorithm with a resolution of 0.3.
To characterize the clusters, we used several melanocyte signatures that we computed across cells using the UCell package ([73] iii). Additionally, we performed classical marker gene analysis for each cluster. All these steps were carried out using R statistical software ((version 4.3.1, R Core Team 2023)4).
In vivo tumor growth
Six-week-old athymic Foxn1nu/nu nude mice were purchased from Janvier (Le Genest, Pays de la Loire, France). Both male and female mice were used for this study. A total of 5 × 106 A375 (BRAFV600E) cells were subcutaneously injected into the flanks of the mice. When the tumors reached a volume between 50–100 mm3, mice with similarly sized tumors were randomized into treatment cohorts (n = 6 per group) and IPTG was administered via the drinking water (10 mM) or intraperitoneally (i.p.) (20 mM, two injections per week) beginning at the initiation of the experiment. The mice in the control groups were treated with vehicle (PBS). Tumor growth was monitored every two days for 28 days. Tumor volume was calculated as L × w × h, where “L” is the major diameter, “w” is the minor diameter, and “h” is the third axis of the tumor mass. The maximal tumor volume permitted by the Vall d’Hebron Institute of Research Ethics Committee is 1500 mm3. Tumor samples were collected at the end of the experiment for further analysis. The results are presented as the mean tumor volumes ± SDs.
Statistical analysis
Microarray RNA expression and single-cell RNA-seq analysis were described in their respective sections. Statistical analyses were performed in GraphPad Prism 9.0 (GraphPad Software, Inc.) using 2-tailed Student’s t-test or one-way ANOVA to evaluate the significance of differences between the groups.
Senescence signature establishment
For senescence signature establishment, seven different recent publications (Marthandan et al. [31]; Hernández-Segura et al. [28]; Wiley et al. [32]; Hernandez-Segura et al. [29]; Zirkel et al. [33]; Lee and Schmitt et al. [30]; and Casella et al. [27]), including articles and reviews, were used. Studies were selected according to their relevance in the field and included the responses of different cell types to different senescence inducers.
