E3 ubiquitin ligase PARK2, an inhibitor of melanoma cell growth, is repressed by the oncogenic ERK1/2-ELK1 transcriptional axis

Malignant melanoma, the most aggressive form of skin cancer, is characterized by high prevalence of BRAF/NRAS mutations and hyperactivation of extracellular signal-regulated kinase 1 and 2 (ERK1/2), mitogen-activated protein kinases (MAPK), leading to uncontrolled melanoma growth. Efficacy of current targeted therapies against mutant BRAF or MEK1/2 have been hindered by existence of innate or development of acquired resistance. Therefore, a better understanding of the mechanisms controlled by MAPK pathway driving melanogenesis will help develop new treatment approaches targeting this oncogenic cascade. Here, we identify E3 ubiquitin ligase PARK2 as a direct target of ELK1, a known transcriptional effector of MAPK signaling in melanoma cells. We show that pharmacological inhibition of BRAF-V600E or ERK1/2 in melanoma cells increases PARK2 expression. PARK2 overexpression reduces melanoma cell growth in vitro and in vivo and induces apoptosis. Conversely, its genetic silencing increases melanoma cell proliferation and reduces cell death. Further, we demonstrate that ELK1 is required by the BRAF-ERK1/2 pathway to repress PARK2 expression and promoter activity in melanoma cells. Clinically, PARK2 is highly expressed in wild type BRAF and NRAS melanomas, but it is expressed at low levels in melanomas carrying BRAF/NRAS mutations. Overall, our data provide new insights into the tumor suppressive role of PARK2 in malignant melanoma and uncover a novel mechanism for the negative regulation of PARK2 via the ERK1/2-ELK1 axis. These findings suggest that reactivation of PARK2 may be a promising therapeutic approach to counteract melanoma growth.


Abstract
Malignant melanoma, the most aggressive form of skin cancer, is characterized by high prevalence of BRAF/NRAS mutations and hyperactivation of extracellular signal-regulated kinase 1 and 2 (ERK1/2), mitogen-activated protein kinases (MAPK), leading to uncontrolled melanoma growth. Efficacy of current targeted therapies against mutant BRAF or MEK1/2 have been hindered by existence of innate or development of acquired resistance. Therefore, a better understanding of the mechanisms controlled by MAPK pathway driving melanogenesis will help develop new treatment approaches targeting this oncogenic cascade. Here, we identify E3 ubiquitin ligase PARK2 as a direct target of ELK1, a known transcriptional effector of MAPK signaling in melanoma cells. We show that pharmacological inhibition of BRAF-V600E or ERK1/2 in melanoma cells increases PARK2 expression.
PARK2 overexpression reduces melanoma cell growth in vitro and in vivo and induces apoptosis. Conversely, its genetic silencing increases melanoma cell proliferation and reduces cell death. Further, we demonstrate that ELK1 is required by the BRAF-ERK1/2 pathway to repress PARK2 expression and promoter activity in melanoma cells. Clinically, PARK2 is highly expressed in wild type BRAF and NRAS melanomas, but it is expressed at low levels in melanomas carrying BRAF/NRAS mutations. Overall, our data provide new insights into the tumor suppressive role of PARK2 in malignant melanoma and uncover a novel mechanism for the negative regulation of PARK2 via the ERK1/2-ELK1 axis. These findings suggest that reactivation of PARK2 may be a promising therapeutic approach to counteract melanoma growth.

Introduction
Malignant melanoma is the most aggressive form of skin cancer. The most prevalent genetic alterations in melanoma are mutually exclusive mutations in BRAF and NRAS, which occur in nearly 50% and 25% of melanoma patients, respectively (1). These mutations result in hyperactivation of the mitogen-activated protein kinases (MAPK) extracellular signal-regulated kinase 1 and 2 (ERK1/2), and consequent uncontrolled melanoma growth. These terminal kinases of the cascade catalyze the phosphorylation, mainly at Ser/Thr-Pro residues, of hundreds of cytoplasmic and nuclear substrates, including regulatory molecules and transcription factors (2). Among the ERK1/2 nuclear targets is the transcription factor ELK1, a member of ETS (E twenty-six) oncogene family of transcription factors, which is directly phosphorylated by ERK1/2 on multi-sites in its transactivation domain (3)(4)(5). The aberrant activation of the RAS-RAF-MEK1/2-ERK1/2-ELK1 signaling pathway has provided the basis for efficient targeted therapy with specific inhibitors of mutant BRAF and MEK in melanoma. However, the presence of innate and development of acquired resistance have hindered the long-term clinical benefits of these treatments. Therefore, a better understanding of the mechanisms controlled by MAPK signaling will help develop efficient treatment approaches targeting this pathway in dismal skin cancer.
Here, we identified PARK2 as a novel target of the oncogenic ERK1/2-ELK1 pathway and we provided insights into the role of PARK2 in melanoma. The E3 ubiquitin ligase PARK2 has been shown to act as a tumor suppressor in several contexts (6). PARK2 loss of heterozygosity and copy number loss have been observed in human cancers, including melanoma (7). In addition, PARK2 inactivating mutations are associated with increased risk of melanoma (8). Consistent with its tumor suppressive role, PARK2 ectopic expression has been shown to reduce cell proliferation in several types of cancer (9)(10)(11)(12)(13)(14)(15)(16)(17)(18). Our findings define a novel mechanism through which the MAPK pathway controls melanoma cell growth through the suppression of PARK2 in an ELK1-dependent manner, and thus will contribute develop new treatment approaches targeting this oncogenic cascade.

PARK2 is repressed by the RAS-RAF-MEK1/2-ERK1/2 signaling in melanoma cells
Western blot (WB) and quantitative real-time PCR (qPCR) analyses in melanoma cell lines showed that PARK2 expression is lower in cells harboring BRAF V600E or NRAS Q61R mutations (SK-Mel-2, SK-Mel-5, SK-Mel-28, A375, 501-Mel) compared to those with wild type (wt) BRAF or NRAS (M51, SSM2c, SK-Mel-197). Only SK-Mel-197 cells express high level of pERK1/2 and low levels of PARK2, although do not carry mutations in BRAF or NRAS ( Figure 1A and B). Further analysis showed reduced PARK2 mRNA expression in melanoma cell lines compared to normal human epidermal melanocytes (NHEM) ( Figure 1B). In support of the biological relevance of this finding, analysis of The Cancer Genome Atlas (TCGA) Melanoma cohort shows higher PARK2 expression in wt compared to mutant BRAS/NRAS metastatic melanomas (p=0.0027) ( Figure 1C). In agreement with these data, analysis of publicly available transcriptomic data sets (GDS1375) showed that PARK2 mRNA was expressed at higher level in human nevi (n=18) compared to malignant melanomas (n=45) (p<0.05) ( Figure 1D). In addition, there is a trend toward improved overall survival (OS) with increased PARK2 expression in metastatic disease, although not statistically significant (Supplementary Figure 1). At the cellular level, WB analysis indicated a cytosolic localization of PARK2 in melanoma cells ( Figure  1E). Immunofluorescence confirmed these data and showed a co-localization of PARK2 with the protein COXIV, a mitochondrial marker ( Figure  1F). Altogether, our data indicate that PARK2 expression is downregulated in human melanomas compared to nevi and that metastatic melanomas carrying wt BRAF/NRAS show higher expression of PARK2 compared to those with mutant BRAF/NRAS.
To investigate the effect of the RAS-RAF-MEK1/2-ERK1/2 signaling on PARK2 expression, melanoma cells harboring BRAF V600E (A375, SK-Mel-5, SK-Mel-28 and 501-Mel) were treated with specific inhibitors of the BRAF-MEK1/2-ERK1/2 cascade. Treatment with the BRAF-V600E inhibitor Vemurafenib (19,20) led to a timedependent increase in PARK2 protein levels ( Figure 1G; Supplementary Figure 2A). Likewise, treatment with SCH-772984, an ERK1/2 inhibitor (21), consistently increased PARK2 protein levels in all four melanoma cell lines ( Figure 1H; Supplementary Figure 2B). To further clarify whether modulation of PARK2 by the BRAF-MEK1/2-ERK1/2 signaling was exerted also at transcriptional level, qPCR analysis of PARK2 mRNA was performed after inhibition of BRAF-V600E or ERK1/2. Expression of PARK2 mRNA was drastically increased upon treatment with Vemurafenib or SCH-772984 ( Figure 1I). The efficiency of these inhibitors was confirmed by strong downregulation of phosphorylated ERK1/2 level ( Figure 1G and H; Supplementary Figure 2) and of Cyclin D1 ( Figure 1J), an established mitogenic target of mutant RAS signaling. Consistent with these results, transient overexpression of BRAF-V600E in SSM2c melanoma cells and in HEK-293T cells, which harbor wt BRAF and NRAS, led to a reduction of PARK2 both at mRNA and protein levels ( Figure  1K-M). Furthermore, qPCR analysis in HEK-293T cells treated with EGF, which induces phosphorylation of ERK1/2, shows an increase in CyclinD1 and c-Fos levels (Supplementary Figure  3A and B) and a time dependent decrease of PARK2 level (Supplementary Figure 3C). Altogether, these data indicate that in melanoma cells the RAS-RAF-MEK1/2-ERK1/2 pathway negatively regulates expression of PARK2.

The transcription factor ELK1 is required by RAS-RAF-MEK1/2-ERK1/2 pathway to repress PARK2 expression
Bioinformatic analysis identified a putative ELK1 binding sites (BS) within the human PARK2 promoter (obtained from the UCSC Genome Browser assembly ID:hg38) near the Transcription Start Site (TSS) (Figure 2A; Supplementary Figure  4). ELK1, a major downstream effectors of ERK1/2, is a member of the ETS (E-twenty-six) oncogene family of transcription factors involved in many biological processes, such as cell growth, differentiation and survival, inflammation and cancer (3)(4)(5)(22)(23)(24). Therefore, we investigated whether ELK1 might be a mediator of the RAS-RAF-MEK1/2-ERK1/2 cascade in modulating PARK2 expression. Treatment of A375 and SK-Mel-5 melanoma cells with the ERK1/2 inhibitor SCH-772984 drastically reduced ELK1 phosphorylation at Serine 383 in both cell lines ( Figure 2B). To investigate whether ELK1 directly binds PARK2 promoter, we performed chromatin immunoprecipitation assay using two different ELK1 specific antibodies. Analysis of the immunoprecipitated DNA by qPCR showed that endogenous ELK1 binds to PARK2 promoter ( Figure 2C). This result was confirmed also upon ELK1 overexpression in HEK-293T cells (Supplementary Figure 5A) and upon ELK1 silencing in A375 melanoma cells (Supplementary Figure 5B). To confirm the ability of ELK1 to regulate PARK2 expression, the PARK2 promoter (-635 to +98 from the TSS) containing the putative ELK1 BS (CCGGAAA) was cloned into a luciferase reporter. Ectopic expression of ELK1 in A375 and SK-Mel-197 melanoma cells showed a strong reduction of luciferase activity ( Figure 2D). Mutation of the ELK1 binding element strongly reduces the inhibition of PARK2 transactivation by ELK1 overexpression in both cell lines ( Figure  2D). Consistent with the negative regulation of PARK2 by ELK1, silencing of ELK1 with two specific independent shRNAs (LV-shELK1-1, LV-shELK1-2) increased PARK2 mRNA and protein levels in A375 and SK-Mel-5 melanoma cells (Supplementary Figure  5C-F), whereas overexpression of ELK1 in HEK-293T cells had the opposite effect (Supplementary Figure 5G and H). In support of the relevance of the negative regulation of PARK2 expression by ELK1, a statistically significant negative correlation was found between ELK1 and PARK2 expression in the TCGA melanoma cohort (n=481) ( Figure 2E).
To confirm the involvement of ELK1 downstream of BRAF-MERK1/2-ERK1/2 in regulating PARK2 expression, we overexpressed BRAF-V600E in ELK1-silenced cells. ELK1 depletion enhanced PARK2 mRNA and protein levels, whereas BRAF-V600E overexpression reduced PARK2 expression, as expected. Interestingly, silencing of ELK1 in presence of BRAF-V600E overexpression rescued the reduction of PARK2 expression elicited by BRAFV-600E ( Figure 2F and G). Consistent with these results, genetic silencing of ELK1 strongly increased PARK2 promoter activity, even when it was co-expressed with BRAF-V600E ( Figure 2H). In addition, PARK2 transactivation induced by the ERK1/2 inhibitor SCH-772984 was reverted by ELK1 overexpression ( Figure 2I). The negative modulation of PARK2 by ELK1 was also confirmed in glioblastoma (U87MG) and breast cancer (MCF7) cell lines. Indeed, in both cell types PARK2 mRNA and protein levels were decreased upon ELK1 expression (Supplementary Figure 6). Altogether, our data indicate that in melanoma cells PARK2 expression is negatively regulated by the RAS-RAF-MEK1/2-ERK1/2 signaling through the transcription factor ELK1, a new repressor of PARK2 transcription. Further, our results suggest that the regulation of PARK2 by ELK1 may take place in other cancer types beyond melanoma, and it might be a general mechanism to restrain PARK2 function in cancer cells.

PARK2 reduces melanoma growth in vitro and in vivo
To evaluate the effect of restoring PARK2 expression in melanoma, we overexpressed it in four melanoma cell lines having low level of PARK2 (A375, SK-Mel-5, SK-Mel-28 and 501-Mel) using a retroviral vector encoding full length PARK2 (PARK2). Stable overexpression of PARK2, confirmed at protein level ( Figure Figure 7D and E). Next, we tested whether ectopic PARK2 expression might affect the ability to form colonies in soft agar. We found that the number of colonies formed by PARK2-expressing melanoma cells was reduced compared to those in the control melanoma cells in A375 and SK-Mel-5 ( Figure 3F and G). Further, we demonstrated that the reduced growth of melanoma cells expressing PARK2 was due to an increase in apoptosis. FACS-based Annexin V/7AAD analysis showed an increase of the fraction of apoptotic cells in PARK2 overexpressing cell lines ( Figure 3H and I; Supplementary Figure 7F and G). Further analysis showed reduced expression of the anti-apoptotic factors BCL-2 and BCL-XL at protein level upon PARK2 overexpression ( Figure 3J; Supplementary Figure 7H). qPCR showed also an increase of the pro-apoptotic factor PIG3 in PARK2overexpressing cells ( Figure 3K and L). PARK2 has been shown to negatively regulate the AKT pathway, a known survival pathway in melanoma (9,25). Our data confirmed the reduction of AKT activation (phosphorylation of Ser473) upon PARK2 overexpression in SK-Mel-5 and, to a lesser extent, in A375 and SK-Mel-28 melanoma cells (Supplementary Figure 7I).
To further investigate the role of PARK2 in melanoma, PARK2 was silenced in two patientderived melanoma cell lines (SSM2c and M51), which expressed the highest levels of PARK2, and in a commercial melanoma cell line (SK-Mel-197), which expressed low PARK2 levels, using two different short interference RNAs (shRNA) specific for PARK2 (LV-shPARK2-1, LV-shPARK2-2). Western blot analysis showed a strong reduction of PARK2 protein level in cells transduced with both shRNAs ( Figure 8E). Altogether, our data indicate that PARK2 reduces melanoma cell growth promoting apoptosis.
To investigate whether PARK2 might affect melanoma xenograft growth in vivo, A375 cells stably transduced with pBABE or PARK2 were subcutaneously injected into the flanks of athymic nude mice and tumor growth was monitored over time. Ectopic expression of PARK2 reduced by 60% the size of melanoma xenografts compared with pBABE control ( Figure  5A and B). Western blot in dissected tumors confirmed PARK2 overexpression with a 20-30 fold increase compared to controls ( Figure 5C) and a drastic decrease of BCL-2 protein level ( Figure  5C), consistent with in vitro tumor cell growth experiments. Induction of apoptosis was confirmed at the molecular level in PARK2-overexpressing xenografts by increased BAX/BCL-2 ratio ( Figure  5D), an indicator of apoptosis (26,27). PARK2 overexpression in xenografts was also confirmed by immunohistochemistry ( Figure 5E). The degree of reduction of melanoma growth upon PARK2 overexpression in vivo was greater than the decrease of melanoma cell growth observed in vitro, suggesting a potential role of the tumor microenvironment. Altogether, these results indicate that PARK2 represses melanoma cell growth in vitro and in vivo, further confirming the tumor suppressive role of PARK2 in melanoma.
Finally, we addressed whether silencing of PARK2 would revert the reduction of melanoma cell proliferation induced by inhibition of BRAF-V600E. Vemurafenib treatment reduced melanoma cell growth in a dose dependent manner in SK-Mel-28 cells transduced with LV-c, as expected. On the other hand, vemurafenib had a very minor effect in reducing melanoma cell growth in absence of PARK2 ( Fig. 6A and B). For instance, treatment with Vemurafenib at 50 nM reduced growth of LVc-transduced cells by 26%, but only by 3.6% in PARK2-silenced cells (Fig. 6C). These results indicate that PARK2 depletion partially rescues the effect of BRAF-V600E inhibition, suggesting that PARK2 is important to mediate the effects of BRAF-ERK1/2 activation in melanoma.

Discussion
The RAS-RAF-MEK1/2-ERK1/2 pathway is a complex signaling network that integrates numerous upstream stimuli to modulate several cellular processes, including cell growth, proliferation and survival. Aberrant activation of this signaling pathway occurs in the majority of malignant melanomas (1). In this study, we provide the evidence that PARK2 is negatively modulated by the RAS-RAF-MEK1/2-ERK1/2 signaling via the transcription factor ELK1. In addition, we provide new evidence of the tumor suppressive role of PARK2 in melanoma, where PARK2 restrains melanoma cell growth downstream of ERK1/2 through the induction of cellular apoptosis.
Our data highlight a previously unexplored mechanism of PARK2 regulation by the RAS-RAF-MEK1/2-ERK1/2 signaling through the transcription factor ELK1, which belongs to the ETS family. ELK1 is the best studied ETS member and it is directly phosphorylated and activated by ERK1/2 (3,4,24), functioning as both activator or repressor of transcription (28). Like all ETS proteins, ELK1 binds the conserved core motif (GGAA/T) embedded in a larger 10 bp consensus sequence that determines the specific recognition of target sites in different genes (29,30). Our findings indicate that ELK1 represses PARK2 expression, through binding to a consensus sequence (CCGGAAA) within the proximal promoter of PARK2. The biological relevance of this regulation is supported by the negative correlation between the expression of PARK2 and ELK1 in a cohort of 481 human melanoma samples. Ultimately, the negative regulation of PARK2 by the ERK1/2-ELK1 axis leads to an increase of proliferation and tumor growth. Consistently, our findings indicate that inhibition of BRAF-V600E or ERK1/2 reduces ELK1 phosphorylation and, as such, ELK1 cannot longer repress PARK2 transcription, with consequent tumor growth arrest and increased cellular apoptosis ( Figure 6D).
PARK2 is a RBR type E3 ubiquitin ligase, which mediates degradation of several substrates through the ubiquitin-proteasome system (9,25,31). A number of studies in the last few years have shown that PARK2 is involved in protein turnover, stress response, mitochondria homeostasis, genomic stability, metabolism and many other cellular processes regulating cell growth and survival (6). Mutations in PARK2 gene have been originally associated with the pathogenesis of autosomal recessive juvenile Parkinson's disease (32,33) and a wide spectrum of brain disorders (34)(35)(36). Although the link between PARK2 and cancer susceptibility is not clear, PARK2 deletion, copy number alteration, mutations and altered mRNA/protein expression have been found in several types of cancer, such as glioblastoma, breast, ovarian, lung and colorectal. In particular, in glioblastoma PARK2 activation correlates inversely with disease progression and patient survival (12). Interestingly, recent evidence shows a link between PARK2 somatic mutations in melanoma and Parkinson's disease (37,38).
A recent study proposed that PARK2inactivating mutations increase the risk of melanoma and that restoration of PARK2 expression in PARK2-deficient melanoma cell lines reduces colony formation (8). However, another report suggested an oncogenic role of PARK2 in melanoma (39). Our findings provide several lines of evidence supporting the tumor suppressive role of PARK2 in human melanoma. First, re-expression of PARK2 in different melanoma cell lines expressing mutated BRAF strongly reduces proliferation in vitro and melanoma xenograft growth in vivo. Second, PARK2 genetic silencing enhances melanoma cell growth and colony formation. Third, PARK2 expression is downregulated in human melanomas compared to nevi or normal melanocytes. Altogether our data suggest that PARK2 loss-offunction may cooperate with BRAF mutations/amplifications during the early phase of melanoma progression. Following this hypothesis, a previous report suggested an association between alterations in BRAF gene (BRAF-V600E mutation and BRAF amplification) and PARK2 copy loss in primary melanomas (7).
The molecular mechanism by which PARK2 exerts its tumor suppressive function in melanoma is, in part, through the induction of cellular apoptosis, as revealed by the strong upregulation of the anti-apoptotic factors BCL-2 and BCL-XL in PARK2-silenced melanoma cells and their downregulation upon PARK2 overexpression in vitro and in vivo. Our findings are consistent with a recent report showing that PARK2 directly binds to and ubiquitinates BCL-XL (40) and other members of the BCL-2 family, such as MCL1 (41). There is evidence in other cancer types that PARK2 targets both Cyclin D1 and E1 for degradation (42), and it interacts with both β-catenin and EGFR to promote their ubiquitination in glioblastoma (9). In addition, PARK2 negatively regulates the PI3K/AKT pathway and PARK2 depletion promotes PTEN inactivation by S-nitrosylation and ubiquitination (25). However, in melanoma cells we did not observe any effect of PARK2 on Cyclin D and E, β-catenin nor PTEN. Nevertheless, our results confirm a role of PARK2 in the negative modulation of AKT pathway (9,25,43), evident upon PARK2 depletion. These data suggest that in melanoma PARK2 explicates anti-proliferative effects mainly by regulating programmed cell death, unlike in glioblastoma and colon cancer where PARK2 is mostly involved in controlling cell cycle progression.
In conclusion, our study uncovers a novel mechanism of negative regulation of PARK2 by the ERK1/2-ELK1 transcriptional axis, suggesting that reactivation of PARK2 in cancer cells might have a potential therapeutic effect. Our data suggest that inhibitors of the BRAF-MEK1/2-ERK1/2 cascade could be useful to induce PARK2 expression. In addition, X-ray data combined with computational modeling have contributed to establish a complete structural model of human PARK2 (44,45) providing the basis for targeted drug design to identify small-molecule activators of this E3 ubiquitin ligase. The complete understanding of PARK2's activation and the ability to improve target specificity will be key determinants in future drug discovery efforts to reactivate PARK2 function.  (Table 1) (46,47). HEK-293T and all melanoma cell lines were grown in DMEM (Euroclone, Milan, Italy) supplemented with 10% fetal bovine serum (FBS), 1% Penicillin-streptomycin, 2 mM L-Glutamine (Lonza, Basel, Switzerland). Human melanoma samples were obtained after approved protocols by the Ethics Committee. Mycoplasma was periodically tested by PCR upon thawing of a new batch of cells and cultures were renewed every month. Transduced cells were selected with puromycin (Invivogen, San Diego, California, USA) at 1-2 µg/ml for 72 hrs.

RNA isolation and quantitative real-time PCR
Total RNA was isolated from cells using Trizol Reagent (Thermo Fisher Scientific) and treated with DNAse I (Sigma-Aldrich), to remove genomic contamination. cDNA was obtained using the High-Capacity RNA-to-cDNA™ Kit (Thermo Fisher Scientific). Quantitative real-time PCR (qPCR) were carried out at 60°C using Sso Advanced Universal SybrGreen supermix (BioRad) in a Rotorgene-Q (Qiagen, Hilden, Germany). Primer sequences are listed in Table 2.

Luciferase reporter assay
by guest on November 6, 2020 http://www.jbc.org/ Downloaded from PARK2 promoter-luciferase reporter was used in combination with Renilla luciferase pRL-TK reporter vector (Promega) in a ratio 10:1, to normalize luciferase activities; pGL3Basic vector (Promega) was used to equal DNA amounts. Luminescence was measured using the Dual-Glo Luciferase Assay System (Promega) and the GloMax 20/20 Luminometer (Promega).

Flow cytometry analysis
For analysis of apoptosis, Annexin V-PE/7-AAD staining was used to detect cells in early-or in lateapoptosis (Becton Dickinson, Franklin Lakes, NJ, USA) after exposure to serum-deprived conditions for 48 hrs. Cytometric analysis was performed with CytoFLEX S (Beckman Coulter, Brea, CA, USA).

Xenograft experiments
A375 melanoma cells transduced with pBABE or PARK2 retroviruses were resuspended in Matrigel (BD Biosciences, Franklin Lakes, NJ, USA)/DMEM (1/1 ratio) and subcutaneously injected (10.000 cells/injection) into both lateral flanks of adult female athymic nude mice (n=8 per group) (CD-1 Nude Mice) (Charles River Laboratories Italy, Milan, Italy). Subcutaneous tumor size was measured 3 times a week by a caliper, and tumor volumes were calculated using the formula V = W 2 × L × 0.5, where W represent the tumor width and L the length. Animals were monitored daily, housed in specific pathogen free conditions and the experiment was approved by the Italian Ministry of Health in accordance with the Italian guidelines and regulations.

Bioinformatical analysis
PARK2 expression in nevi and melanoma samples was analyzed using publicly available microarray data set (GDS1375) (49), from Gene Expression Omnibus (GEO) profiled on Affymetrix U133 platforms. The University of California Stata Cruz Xena platform was used to analyze correlation, transcriptomic and survival data from The Cancer Genome Atlas (TCGA) Melanoma (SKCM) cohort of 17 data sets (50). This curated survival data from the Pan-cancer Atlas manuscript highlighting four types of curated survival endpoints of recommended use including overall survival (51). Gene expression profile was measured using the Illumina HiSeq 2000 RNA sequencing platform by the University of North Carolina TCGA genome characterization center. Level 3 data was downloaded from TCGA data coordination center and gene-level transcription estimates were shown as log2(+1) transformed RSEM normalized count. Genes were mapped onto the human genome coordinates using UCSC Xena HUGO probeMap.

Statistical analysis
Data are presented as mean ± s.d. or ± s.e.m. from at least three independent experiments. P values were calculated using two-tailed Student's t-test (two groups) or ANOVA (more than two groups; multiple comparison using Bonferroni's correction). Value of p<0.05 was considered statistically significant. Correlation between the expression of PARK2 and ELK1 using the TCGA melanoma cohort (n=481). Pearson's correlation test was used to analyze the correlation between PARK2 and ELK1 expression.
Data availability: All the described data are contained within this manuscript.
VM and BS conceived and designed the study; VM, LM, AA, SP, RMC, MEFZ generated, collected and analyzed the data; VM and BS wrote the manuscript; BS assembled figures; all authors revised and approved the final version of the manuscript.

Funding and additional information
This work was supported by funding obtained from Institute for Cancer Research, Prevention and Clinical Network (ISPRO). VM and LM were supported by a postdoctoral fellowship from Italian Association for Cancer Research (AIRC, project n. 19580 and n. 22644 respectively).        On the left, BRAF-MEK1/2-ERK1/2 signaling induces phosphorylation and nuclear translocation of ELK1, which binds PARK2 proximal promoter acting as negative transcriptional regulator of PARK2. On the right, inhibition of BRAF-V600E or ERK1/2 reduces ELK1 phosphorylation with consequent induction of PARK2 transcription and suppression of tumor growth. Genetic silencing of ELK1 resembles the effect of BRAF-V600E or ERK1/2 inhibition on PARK2 expression. Vemurafenib is a BRAF-V600E inhibitor, whereas SCH-772984 is a ERK1/2 inhibitor.
by guest on November 6, 2020