Retinoblastoma-Binding Protein 5 Regulates H3K4 Methylation Modification to Inhibit the Proliferation of Melanoma Cells by Inactivating the Wnt/β-Catenin and Epithelial-Mesenchymal Transition Pathways

Histone 3 lysine 4 methylation (H3K4me), especially histone 3 lysine 4 trimethylation (H3K4me3), is one of the most extensively studied patterns of histone modification and plays crucial roles in many biological processes. However, as a part of H3K4 methyltransferase that participates in H3K4 methylation and transcriptional regulation, retinoblastoma-binding protein 5 (RBBP5) has not been well studied in melanoma. The present study sought to explore RBBP5-mediated H3K4 histone modification and the potential mechanisms in melanoma. RBBP5 expression in melanoma and nevi specimens was detected by immunohistochemistry. Western blotting was performed for three pairs of melanoma cancer tissues and nevi tissues. In vitro and in vivo assays were used to investigate the function of RBBP5. The molecular mechanism was determined using RT-qPCR, western blotting, ChIP assays, and Co-IP assays. Our study showed that RBBP5 was significantly downregulated in melanoma tissue and cells compared with nevi tissues and normal epithelia cells (P < 0.05). Reducing RBBP5 in human melanoma cells leads to H3K4me3 downregulation and promotes cell proliferation, migration, and invasion. On the one hand, we verified that WSB2 was an upstream gene of RBBP5-mediated H3K4 modification, which could directly bind to RBBP5 and negatively regulate its expression. On the other hand, we also confirmed that p16 (a cancer suppressor gene) was a downstream target of H3K4me3, the promoter of which can directly bind to H3K4me3. Mechanistically, our data revealed that RBBP5 inactivated the Wnt/β-catenin and epithelial-mesenchymal transition (EMT) pathways (P < 0.05), leading to melanoma suppression. Histone methylation is rising as an important factor affecting tumorigenicity and tumor progression. Our findings verified the significance of RBBP5-mediated H3K4 modification in melanoma and the potential regulatory mechanisms of melanoma proliferation and growth, suggesting that RBBP5 is a potential therapeutic target for the treatment of melanoma.


Introduction
Melanoma is the most lethal type [1] of skin tumor and originates from the malignant translation of melanocytes [2]. Melanocytes are derived from neuroectodermal cells and are widely distributed throughout the body. As a result, melanoma can occur ubiquitously, including skin, acra, mucosa, rectum, and uvea [3,4]. In 2022, there are estimated 99,780 new skin melamoma patients and 7,650 deaths occurred in the United States [5]. Although targeted and immune-based therapeutic strategies have improved the survival rate of melanoma in recent years, 60-70% of melanoma patients still have no response to immune checkpoint inhibitors due to toxicity, drug resistance or other reasons [6]. In patients with metastatic melanoma, the 5-year survival rate is only approximately 23% [7]. Terefore, exploring the molecular mechanisms and discovering new strategies for melanoma are essential to improve the outcomes of melanoma.
Retinoblastoma-binding protein 5 (RBBP5), which is also known as SWD1 or RBQ3, is a protein coding gene and encodes a nuclear protein that belongs to a highly conserved subfamily of water displacement (WD)-repeat proteins. Te encoded protein binds directly to retinoblastoma protein, which regulates cell proliferation. RBBP5, WDR5, ASH2L, and DPY30 are components of the WRAD complex, which is an H3K4 methyltransferase and participates in H3K4 methylation and transcriptional regulation [8][9][10][11][12]. Te H3K4 methyltransferase family has similar structures, the SET domain for catalysis and the WRAD complex for regulation. Te methyltransferase activity of the SET domain itself is weak, which has high methyltransferase activity only after binding with the WRAD complex [8][9][10][11]. Terefore, the deletion of RBBP5 may lead to a decrease in H3K4 methyltransferase activity and the degree of H3K4 methylation. In addition, previous studies have shown that RBBP5 is associated with prostate carcinoma [13], hepatocellular carcinoma [14], multiple myeloma [15], gliomas [16], etc.
Histone modifcations can change the structure and afect gene transcription. H3K4me, especially H3K4me3, is a well-known epigenetic mark of histone modifcation. In recent years, H3K4me3 has been found to play a crucial role in many cancers, including breast carcinoma [17], cervical cancer [18], renal carcinoma [19,20], colon cancer [21], ovarian cancer [22], and hepatocellular carcinoma [23]. H3K4me3 is associated not only with transcriptional activation but also with suppressed gene expression [24,25]. However, the roles of H3K4me3 and RBBP5 in melanoma remain unknown.
In the present study, we found that RBBP5 was a tumor suppressor factor in human melanoma,which was downregulated in melanoma tissues and cells. Functional studies showed that overexpression of RBBP5 could enhance the expression of H3K4me3 and inhibit the proliferation, migration and invasion of melanoma cells. We explored the signalling pathway related to RBBP5 and found that RBBP5 could afect the progression of melanoma through multiple signalling pathways, including Wnt/β-catenin and epithelial-mesenchymal transition (EMT). In addition, H3K4me3 could directly bind with the promoter of the tumor suppressor gene p16 and increase its expression. Our study suggested that RBBP5 plays an important role in tumor progression in melanoma and might be a potential therapeutic target for the treatment of melanoma. Hospital of Kunming Medical University. All cells were cultured with Dulbecco's Modifed Eagle Medium (DMEM, Gibco, c11995500BT; Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS, BI, 04-001-1ACS) and 1% penicillin/streptomycin (NCM, C125C5) at 37°C in an incubator with 5% CO 2 .

Tissue Specimens.
Tree pairs of nevi, melanoma, and adjacent tissues were obtained from patients at the Tird Afliated Hospital of Kunming Medical University and stored at −80°C refrigerator. In addition, 106 cases of parafn-embedded melanoma tissues, 100 cases of parafnembedded adjacent specimens, and 23 cases of parafnembedded nevi tissues were also obtained from the Tird Afliated Hospital of Kunming Medical University. None of the patients were treated with chemotherapy or radiotherapy before surgery. Tis research was approved by the Ethics Committee of Kunming Medical University.

Cell
Counting Kit-8 (CCK8) Assay. A CCK8 assay was performed to measure cell proliferation ability. A total of 2000 cells were seeded into a 96-well plate and incubated at 37°C and 5% CO 2 . Ten, 100 µl of serum-free medium containing 10 µl CCK8 reagent (APE × Bio, K1018) was added to each well every day. After incubation at 37°C and 5% CO 2 for 2 h, the absorbance was measured at 450 nm.
2.6. Wound Healing Assay. Cells were seeded into 6-well culture plates at a density of 1 × 10 6 cells, until they reached 90%. Linear wounds were created using 100 µl of tips, and cells were cultured for 24 h and 48 h in serum-free medium at 37°C and 5% CO 2 . Images were taken with a microscope, and the wound healing ability was examined by calculating the wound area.

Migration and Invasion
Assays. Te invasion/migration ability of the cells was measured in a Boyden chamber (Corning, 3422) with or without Matrigel (Corning, 356234). A total of 1 × 10 5 A375 cells and 3 × 10 4 A2058 cells were suspended in 200 µl of serum-free medium and added to the upper chamber, and 600 µl of complete medium containing 20% fetal bovine serum (FBS) was added to the lower chamber. After incubation for 24 h at 37°C and 5% CO 2 in an incubator, the cells on the top surface of the membrane were removed, and the cells that had migrated or invaded to the lower surface of the membrane were fxed with methyl alcohol for 15 min and stained with 0.2% crystal violet for 30 min. Te cells were photographed with a microscope, and the number of cells was counted by Image J.

Colony Formation Assay.
For the colony formation assay, 500 cells were seeded in a 6-well plate, and the medium was changed every 3 days. Te study was terminated when colony formation was visible to the naked eye. After washing the cells three times with PBS, the cells were fxed with methyl alcohol for 15 min and stained with 0.2% crystal violet for 30 min. Visible colonies were photographed and counted with Image J.

Chromatin Immunoprecipitation (ChIP) Assay.
ChIP was performed using a Simple ChIP ® Enzymatic Chromatin IP Kit (magnetic beads) (CST, 9003) according to the manufacturer's instructions. In brief, cells were seeded into a 15 cm culture dish until the density reached 90%. To crosslink the protein with DNA, 540 µl 37% formaldehyde was added to each 15 cm dish, which contained 20 ml of medium (making the fnal concentration of formaldehyde 1%), and the dish was rotated and incubated at room temperature for 10 minutes. Ten, 2 ml of 10× glycine was added, and the dish was slightly vortexed and incubated at room temperature for 5 minutes, and it was washed with 1x PBS two times. Ten, 2 mL of frozen 1× PBS + 200× PIC was added, and then the dish was centrifuged at 4°C and 2000× g for 5 minutes. Next, nuclear preparation and chromatin digestion were performed, and fragmented chromatin was obtained by nuclease and sonication. Chromatin immunoprecipitation was performed, and sufcient 1× chip bufer was prepared to dilute the digested chromatin. Ten, 10 µl rabbit anti-histone H3 (CST, 4620) was added as a technical positive control, 2 µl normal rabbit immunoglobulin G (IgG) (CST, 2729) was added as a negative control, and 2 µl rabbit antiH3K4me3 antibody (Abcam, ab8580) was added. Te immunoprecipitation (IP) samples were rotated and incubated at 4°C overnight. Ten, 30 μl of Protein G Magnetic Beads (CST, 9006) was added to each IP reaction, rotated, and incubated for 2 hours at 4°C. After reverse cross-linking and DNA purifcation, immunoprecipitated DNA was quantifed by real-time PCR using SimpleChIP ® Universal qPCR Master Mix (CST, 88989) with primers for p16 binding sites in the p16 promoter (forward primer: 5′-AGC ACTCGCTCACGGCGTC-3′ and reverse primer: 5′-CTG TCCCTCAAATCCTCTGGA-3′) and RPL30 (CST, 7014). Fold enrichment was calculated based on the threshold cycle 2.11. Co-immunoprecipitation (Co-IP). After washing three times with PBS, cells and tissue samples were lysed on ice with cell lysis bufer, which was preadded with protease and phosphatase inhibitors. After 30 min, the samples were centrifuged at 4°C and 12000 rpm for 10 min. A total of 100 µl of protein lysate was used as an input control (positive control). Ten, 2 µl WSB2 antibody was added to 200 µl protein lysate, 2 µl normal rabbit IgG antibody was added to another 200 µl protein lysate as a negative control, and the sample was rotated and incubated for 24 hours at 4°C. Ten, 400 µl phosphate-bufered saline with Tween PBST (1× PBS + 0.5% Tween-20, pH 7.4) was used to wash the Protein A/G Magnetic Beads (MCE, HY-K0202), and the beads were separated by a magnetic rack. Te above step was repeated two times. Te antigen antibody mixture was added to the pretreated beads and incubated at 4°C for 3 hours to obtain the antigen-antibody-bead mixture. Te antigen-antibody-bead mixture was adsorbed with a magnetic rack, and then 5x sodium dodecyl sulfate (SDS) loading bufer (including the input control) was added and boiled at 95°C for 10 min. Western blotting was used to verify the interaction between proteins.

Xenograft Mouse Model.
Te animal studies were approved by the Animal Ethics Committee of Kunming Medical University. Six-to eight-week-old female BALB/c nude mice were purchased from SPF Biotechnology Co. Ltd. (Beijing, China). For the preliminary experiment, ten mice were randomly divided into two groups (n � 5 per group), and 2 × 10 6 RBBP5 overexpression or vector A375 cells were injected subcutaneously into each mouse. In a follow-up experiment, thirty mice were randomly divided into four groups. Te frst two groups had ten mice in each group, and 2 × 10 6 RBBP5 knockdown or vector A375 cells were injected subcutaneously. Te latter two groups had 5 mice in each group, and 2 × 10 6 RBBP5 overexpression or vector A375 cells were injected subcutaneously. Te volume of the tumor was measured by the equation (L × W 2 )/2. At the end of the experiment, the tumors were quantifed.

Flow Cytometry.
For the cell cycle assay, 1 × 10 6 cells were seeded into a 6-well plate for 24 h and then harvested and washed twice with PBS. A total of 1 ml of 75% cold ethanol was used to fx the cells, and a Cell Cycle and Apoptosis Analysis Kit (Beyotime, C1052) was used to stain the cells for 30 min in a 37°C dark and warm bath. Te cells were run on a FACS Calibur fow cytometer (BD Biosciences, Haryana, India). Data were analyzed with FlowJo software.

Statistical Analysis.
Statistical analysis of all data was performed with SPSS 20.0 and GraphPad Prism 8.0 software. Data are presented as mean ± standard deviation (SD). All experiments were performed at least three times. Student's ttest was used to compare diferences between two groups. One-way analysis of variance (ANOVA) was used to compare diferences among multiple groups. p < 0.05 was considered statistically signifcant. Signifcance is presented as follows: * p < 0.05, * * p < 0.01, and * * * p < 0.001.

WSB2 Was Bound Directly to RBBP5 and Negatively
Regulated It. Te research results of our group showed that WSB2 was signifcantly increased in melanoma and promoted the proliferation and migration of melanoma cells [26]. Search Tool for the Retrieval of Interacting Genes/ Proteins (STRING) (https://cn.string-db.org/), an online database, was used to predict the downstream targets of WSB2, and the results showed that WSB2 was bound to the RBBBP5 protein (Figure 1(a)). Co-IP was used to confrm the direct protein association between WSB2 and RBBP5 ( Figure 1(f )). In addition, we found that WSB2 knockdown efciently enhanced RBBP5 and H3K3me3 expression in melanoma cells (Figures 1(b)-1(e)). To further explore the correlation between WSB2 and RBBP5, we performed immunohistochemical staining on 104 melanoma parafn sections ( Figure 1(g)). Te results were classifed as negative or positive (Figures 1(h) and 1(j)). Positive WSB2 staining was observed mainly in the cytoplasm and partly in the  Western blotting was used to explore the changes in RBBP5 and H3K4me3 protein after WSB2 knockdown. (f ) Co-IP and western blotting were performed to evaluate the interaction between WSB2 and RBBP5. Data are represented as mean ± SD of three independent experiments. * p < 0.05, * * p < 0.01, and * * * p < 0.001; ns, no signifcance. (g, h) Immunohistochemical staining of RBBP5 and WSB2 in 104 cases of melanoma tumor tissues. Representative images are shown. Scale bar: 50 µm. Te correlation between RBBP5 and WSB2 was calculated (i, j). n � 104, Pearson chi-square test.   (g, h) Te relative score of RBBP5 was quantifed by the grade of staining intensity and the percentage of positively stained cells. A375 and A2058 cells were transfected with empty vector (shRBBP5-vector) or shRNAs targeting RBBP5 (shRBBP5), and the relative mRNA level of RBBP5 was measured by qRT-PCR. GAPDH was used as an internal control (j). Te protein expression of RBBP5 was detected by western blotting. GAPDH was used as a loading control (k, l). A375 and A2058 cells were transfected with RBBP5 empty vector (RBBP5-vector) or RBBP5 overexpression lentivirus (RBBP5), and transfection efciency was verifed by q-PCR (i) and western blotting (m, n). GAPDH was used as a loading control. Data are represented as mean ± SD of three independent experiments. * p < 0.05, * * p < 0.01, and * * * p < 0.001; ns, no signifcance. nucleus. Positive RBBP5 was observed mainly in the cytoplasm and nucleus. Te IHC score showed that RBBP5 expression was negatively correlated with WSB2 expression (Figure 1(i)). Tese experiments verifed that WSB2 could bind with RBBP5 and then negatively regulate RBBP5 and H3K4me3 protein expression.

RBBP5 Was Downregulated in Melanoma.
To explore the potential role of RBBP5 in melanoma, the expression of RBBP5 in melanoma was investigated by WB, IHC and qRT-PCR. First, WB results showed that the expression of RBBP5 was signifcantly decreased in melanoma tissue compared with nevi and adjacent tissue (p < 0.05, Figures 2(a) and  2(b)). Second, IHC was performed in 106 cases of melanoma tumor tissues, 100 cases of adjacent tissues, and 23 cases of nevi tissues to identify the potential diference among different groups. Positive RBBP5 staining was detected in the nucleus and cytoplasm. Te percentage of positive RBBP5 staining was 12.3% (13/106) in tumor tissues, 24% (24/100) in adjacent tissues, and 21.8% (5/23) in nevi tissues. Moreover, the average RBBP5 IHC score of tumor tissues was lower than the average RBBP5 IHC score of adjacent tissues and nevi tissues (Figures 2(f )-2(h)). Tird, the qRT-PCR and WB results also showed that the expression of RBBP5 was signifcantly downregulated in two melanoma cell lines (A375 and A2058) compared with a human immortalized keratinocyte (HaCaT) cell line (Figures 2(c)-2(e)). Together, our study showed that the expression of RBBP5 was signifcantly downregulated in melanoma tissue and cells. To explore the function of RBBP5 in melanoma, we constructed stable cell lines with RBBP5 knockdown and overexpression in A375 and A2058 cells (Figures 2(i)-2(n)).

RBBP5 Positively Regulated the Expression of H3K4me3
and P16. According to our previous study, we found that the expression of RBBP5 and H3K4me3 was notably downregulated after WSB2 knockdown. To explore the function of RBBP5 in melanoma, we constructed stable cell lines with RBBP5 knockdown and overexpression in A375 and A2058 cells (Figures 2(i)-2(n)). Our results indicated that the expression of H3K4me3 and p16 was signifcantly inhibited after RBBP5 knockdown (Figures 3(a)-3(c)). In contrast, the expression of H3K4me3 and p16 was signifcantly increased after RBBP5 was overexpressed (Figures 3(d)-3(f )). Because H3K4me3 can bind directly to the promoter of genes and mediate transcriptional activation, we further investigated the interaction of H3K4me3 and p16 by chromatin immunoprecipitation (ChIP). We observed that H3K4me3 can bind to the promoter of p16 and then upregulate p16 expression (Figure 3(g)). Together, our study showed that RBBP5 could upregulate the level of H3K4me3, which could bind to the promoter of p16 and activate transcription of p16.

RBBP5 Downregulated the Activity of the Wnt/β-Catenin and EMT Signalling Pathways.
A previous study showed that WSB2 could activate the Wnt/β-catenin signalling pathway and promote the proliferation of melanoma cells. To explore the molecular mechanism of RBBP5 in melanoma cells, we analyzed the expression of β-catenin and c-myc, two marked target genes of the Wnt/β-catenin signalling pathway, in A375 and A2058 cells. We observed that upon RBBP5 knockdown in A375 and A2058 cells, the expression of β-catenin and c-myc was upregulated, as shown by western blotting (Figures 6(a)-6(d)). We also observed that two wellknown proteins of EMT, N-cadherin and MMP-7, were upregulated, and E-cadherin was downregulated in RBBP5 knockdown A375 and A2058 cells (Figures 6(a)-6(d)). Te results were contrary to RBBP5-overexpressing cell lines A375 and A2058 (Figures 6(e)-6(h)). Together, our study demonstrated that RBBP5 could inactivate the Wnt/ β-catenin signalling pathway and inhibit EMT.

RBBP5 Inhibits the Proliferation of Melanoma In Vivo.
To investigate the impact of RBBP5 on the tumorigenic capacity of melanoma cells in vivo, we subcutaneously injected RBBP5-overexpressing and RBBP5-silenced A375 cells into female nude mice. We observed that RBBP5 knockdown increased the subcutaneous xenograft volume and tumor weight compared with the control group ( Figures  7(a)-7(c)). Tere was no signifcant diference in body weight of mice (Supplementary Figure S3). In contrast, overexpression of RBBP5 in A375 cells resulted in a marked decrease in the tumor-initiating ability and tumor weight (Figures 8(a)-8(c)). In addition, the western blot results showed that the expression of the RBBP5 downstream proteins H3K4me3 and p16 was decreased in the RBBP5 knockdown group (Figure 7(e), Supplementary Figure S1). Furthermore, the expression ofβ-catenin and c-myc in the Wnt/β-catenin signalling pathway was increased in the RBBP5 knockdown group in xenogeneic tumor tissues of nude mice, and N-cadherin and MMP-7 in EMT also increased (Figure 7(d), Supplementary Figure S1). We obtained the opposite results in the RBBP5-overexpressing group of xenogeneic tumor tissues (Figures 8(d) and 8(e), Supplementary Figure S2). Tese results were in agreement with our previous cell experiments. Immunohistochemical staining of xenogeneic tumor tissues showed that RBBP5 knockdown signifcantly increased the expression of Ncadherin and decreased the expression of H3K4me3 (Figures 7(f )-7(i)), which was opposite to the RBBP5 overexpression group (Figures 8(f )-8(i)). Together, our research suggested that RBBP5 could suppress the tumor growth and EMT of melanoma cells in vivo.

Discussion
RBBP5 is a protein coding gene. In embryonic stem (ES) cells, RBBP5 plays a crucial role in diferentiation potential, particularly along the neural lineage, regulating gene induction and H3 "Lys-4" methylation at key developmental loci, including those mediated by retinoic acid (by similarity). Components or associated components of some histone methyltransferase complexes regulate transcription through recruitment of those complexes to gene promoters [27]. As part of the MLL1/MLL complex, RBBP5 is involved in mono-, di-, and trimethylation at "Lys-4" of histone H3 [28]. Histone H3 "Lys-4" methylation represents a specifc tag for epigenetic transcriptional activation [28]. In association with ASH2L and WDR5, RBBP5 stimulates the histone methyltransferase activities of KMT2A, KMT2B, KMT2C, KMT2D, SETD1A, and SETD1B [29,30]. Studies have shown that RBBP5 is associated with many cancers [13][14][15][16], such as prostate cancer [13], hepatocellular carcinoma [14], and gliomas [16]. Tese results indicate that the expression of RBBP5 was upregulated and played a crucial role in the progression of tumors. Liu et al. reported that although RBBP5 was highly expressed in multiple myeloma cells and associated with cell proliferation, RBBP5 knockdown resulted in an increased adhesion rate of multiple myeloma cells [15]. Liu et al. also showed that multiple myeloma cells become less sensitive to chemotherapy druginduced apoptosis when RBBP5 expression is knocked down [15]. Tese results indicate that RBBP5 may play an important anticancer role. However, the role of RBBP5 in melanoma cancer remains unclear. In our study (Figures  2(a)-2(h)), we verifed that the expression of RBBP5 was downregulated in melanoma tumor tissues and melanoma cells; furthermore, in our IHC results, the IHC score of positive RBBP5 staining in melanoma tumors was lower than the IHC score in adjacent tissues and nevi tissues. Tese results were inconsistent with previous reports. To explore the function and mechanism of RBBP5 in melanoma, we constructed stable melanoma cell lines A375 and A2058 by knocking out and overexpressing RBBP5 (Figures 2(i)-2(n)). We revealed that knockdown of RBBP5 in melanoma cells promoted proliferation, clone formation, invasion, and migration ( Figure 4). However, overexpression of RBBP5 inhibited all of the above abilities (Figures 5(a)-5(j)). We also found that RBBP5 overexpression inhibited the cell cycle transition from the G0/G1 phase to the S phase ( Figures 5(k)-5(m)). Tus, our study clearly proved that RBBP5 is a tumor progression inhibitor in melanoma.
WSB2 (WD repeat and SOCS box containing protein 2), as an E3 ubiquitination enzyme, has been demonstrated to play a key role in the proliferation of melanoma [26]. We used the STRING website (https://cn.string-db.org/) to predict proteins that could interact with WSB2. We found and demonstrated that WSB2 can bind with RBBP5 and downregulate the expression of RBBP5 (Figure 1). Tus, we considered that RBBP5, as a downstream factor, may play a key role in WSB2-mediated ubiquitination degradation and may be used as a substrate. Terefore, further research and verifcation are needed in the follow-up.
H3K4me is a well-known histone modifcation mediated by the SET 1 complex, which is stimulated by the WRAD (the components are RBBP5, WDR5, ASH2L, and DPY30) complex [31][32][33]. Furthermore, Li et al. reported that human RBBP5-ASH2L heterodimer is the major structural unit that interacts with and activates MLL family [33]. High levels of mono-, di-, and trimethylation of H3K4 (H3K4me1, H3K4me2, and H3K4me3) are detected as the promoters of genes and are associated with the transcriptional activity of genes [34][35][36]. Chen et al. frst reported that broad H3K4me3 that was specifcally enriched at the transcription start site could elongate the transcription of tumor suppressor genes and increase their activity [24]. In 2017, the results of the two studies were consistent with the results of previous studies; simultaneously, the mechanism was also described [37,38]. Two studies of renal carcinoma showed that lower levels of H3K4me3 were correlated with advanced stage, distant metastasis, and a shorter period of progression-free survival [19,20]. Han et al. reported that higher levels of H3K4me3 could improve the overall survival of ovarian cancer and inhibit the proliferation of ovarian cancer cells and could be a potential prognostic factor [22]. Our study conclusion agreed with previous studies showing that overexpression of RBBP5 can enhance the level of H3K4me3 and inhibit the progression of melanoma in vivo and in vitro. In contrast, other previous studies verifed that the expression of H3K4me3 was high and associated with poor prognosis in cervical cancer [18], hepatocellular carcinoma [23], and colon cancer [21]. Te discrepancy between these studies may be due to the broad H3K4me3 domain activating cell-type-specifc or disease-specifc genes [24,[39][40][41].
H3K4me3 was used to identify novel tumor suppressors because tumor suppressor genes (TP53, PTEN, GPX3, and SPRY2) were associated with broad H3K4me3 peaks [24]. Yang et al. showed that inhibiting the demethylation of H3K4me3 resulted in increased trimethylation of H3K4 and increased recruitment of H3K4me3 to the promoters of p16 and p27, thus increasing their transcription in breast cancer [17]. Tis conclusion is in accordance with two previous studies on lung cancer and gastric cancer [42,43]. In our study, we demonstrated that H3K4me3 can directly 8 Journal of Oncology bind to the promoter of p16 (Figure 3(g)). Terefore, knockdown of RBBP5 expression can decrease the trimethylation level of H3K4, thus resulting in decreased transcription of the tumor suppressor gene p16 (Figures 3(a)-3(c)). In contrast, overexpression of RBBP5 had the opposite efect (Figures 3(e)-3(g)). Our fndings reveal a novel link between RBBP5, H3K4me3, and p16 and melanoma progression. We further elucidated that the downstream signalling pathway correlated with RBBP5 tumor suppressor gene    function in melanoma and verifed that RBBP5-mediated tumor inhibition could be, at least partly, due to inhibition of Wnt/β-catenin signalling, which activation is important for Tcell exclusion, resistance to antiPD-L1/antiCTLA4 therapy, and progression of melanoma [26,44,45]. RBBP5 knockdown induced Wnt/β-catenin signalling activation in A375 and A2058 cells, as evidenced by western blotting, and the expression of Wnt/β-catenin signalling representative genes (β-catenin, c-myc) was upregulated. In our study, we also found that RBBP5 knockdown induced EMT, and the expression of N-cadherin and MMP-7 was upregulated, while the expression of E-cadherin was downregulated ( Figure 6). Puisieux et al. [46] reported that EMT could be induced by Wnt/β-catenin to stimulate several EMT-related factors, such as Snail, ZB3, and E47. Because the occurrence and development of cancer is a complex process, a variety of signalling pathways are involved and interact. Te mechanism of the interaction of Wnt/β-catenin and EMT should be further explored. Te epithelial-mesenchymal transition (EMT) describes a reversible switch from an epithelial-like to a mesenchymallike phenotype [47][48][49][50]. EMT was crucial for the invasion and metastasis of tumor, which has been suggested as a driving role in the acquisition of a metastatic phenotype [51][52][53][54]. A determinant hallmark of EMT is the presence of the cadherin switch. Te epithelial to neural cadherin switch has been reported by previous studies, which was accomplished by the downregulation of the protein E-cadherin and upregulation of N-cadherin [55][56][57]. In our study, we observed that partial-EMT-related markers N-cadherin and MMP-7 were upregulated, and E-cadherin was downregulated after RBBP5 knockdown in melanoma cells, which agree with previous research about cadherin switch.
EMT-related markers include N-cadherin, E-cadherin, vimentin, and Snail, which have been documented in lots of cancers, such as breast [58], colorectal [59], oral [60], lung [61], pancreatic [62], melanoma [51,[63][64][65], and liver cancers [66]. EMT-related markers were associated with tumor initiation, invasion, metastasis, and resistance to therapy [47,[67][68][69]. Several previous studies have reported that the membranous E-cadherin/β-catenin complexes were associated with tumor progression and poor survival [70,71]. Luo et al. reported that the high expression of nuclear vimentin and cytoplasmic E-cadherin was signifcantly associated with worse outcome of nasopharyngeal carcinoma [47]. Deeb et al. also demonstrated that cytoplasmic high staining of E-cadherin was associated with shorter survival in lung cancers [72]. Vimentin is another mesenchymal marker for EMT, and several reports have demonstrated that vimentin is overexpressed in cancers, such as gastric cancer [73], breast cancer [74], and prostate cancer [75], which is associated with invasive phenotype and poor prognosis. Te prognostic efect of N-cadherin was also contradictory in some studies. Bachmann reported that Ncadherin expression has no correlation with the clinical outcome in melanoma patients [63], but Kreizenbeck et al. showed that high expression of N-cadherin was associated with better overall survival [64]. On the contrary, Pieniazek found that high expression of N-cadherin was associated with shorter overall survival [65]. Lade-Keller et al. reported that when a "switch profle" (E-cadherin switches to Ncadherin expression) is triggered, it is signifcantly associated with poor survival and distant metastasis-free survival in melanoma patients [76]. Despite the diferent EMT states and mechanisms regulating cell in melanoma reported, there are still many questions that are unclear. Combining the previous research and our research, we can fnd that EMTrelated markers are closely linked with the prognosis of many types of cancer, which indicate that further studying new therapeutic strategies related to the control mechanism of EMT can help us prevent melanoma progression and metastasis, providing a new promising therapeutic method in the future.

Conclusions
Altogether, our study demonstrated the potential ability of RBBP5 to act as a tumor suppressor to inhibit the progression of melanoma by inhibiting the Wnt/β-catenin signalling pathways and EMT. Furthermore, our research provided strong evidence that RBBP5 can increase the expression of H3K4me3, which could promote the transcription of p16 by directly binding to the promoter of p16 and inhibiting the progression of melanoma ( Figure 9). Tus, our fndings suggest that RBBP5 might be a potential therapeutic target to reverse melanoma progression.

Data Availability
All data generated or analyzed during this study are included within the article. Te datasets used during the present study are available from the corresponding author on reasonable request.

Ethical Approval
Te animal experiments were approved by Animal Ethics Committee of Kunming Medical University and were performed in accordance with established guidelines. Te use of human melanoma carcinoma and nevi tissues was reviewed and approved by the Medical Ethics Committee of Kunming Medical University. Te samples were retrospectively acquired from the Tird Afliated Hospital of Kunming Medical University. Te use of patient samples conformed to the Declaration of Helsinki.

Consent
All patients provided informed written consent.

Conflicts of Interest
Te authors declare that there are no conficts of interest.

Authors' Contributions
ZY was responsible for conceptualization, data curation, methodology, formal analysis, investigation, writing, and visualization. YJ was responsible for data curation, methodology, formal analysis, investigation, and visualization. SW was responsible for methodology and resources. YZ was responsible for data collection, methodology, visualization, and resources. WF was responsible for methodology, visualization, and resources. XW was responsible for methodology and resources. LH, XS, and XY were responsible for data curation and validation. HY was responsible for the design of the study, conceptualization, supervision, project administration, resources, funding acquisition, and editing. YZ was responsible for conceptualization, methodology, formal analysis, investigation, writing, and visualization. ZY and YJ contributed equally to this work. All authors have read and approved the fnal manuscript.