Expressions of HLA Class II Genes in Cutaneous Melanoma Were Associated with Clinical Outcome: Bioinformatics Approaches and Systematic Analysis of Public Microarray and RNA-Seq Datasets

Major histocompatibility complex (MHC) class II molecules, encoded by human leukocyte antigen (HLA) class II genes, play important roles in antigen presentation and initiation of immune responses. However, the correlation between HLA class II gene expression level and patient survival and disease progression in cutaneous melanoma is still under investigation. In the present study, we analyzed microarray and RNA-Seq data of cutaneous melanoma from The Cancer Genome Atlas (TCGA) using different bioinformatics tools. Survival analysis revealed higher expression level of HLA class II genes in cutaneous melanoma, especially HLA-DP and -DR, was significantly associated with better overall survival. Furthermore, the expressions of HLA class II genes were most closely associated with survival in cutaneous melanoma as compared with other cancer types. The expression of HLA class II co-expressed genes, which were found to associate with antigen processing, immune response, and inflammatory response, was also positively associated with overall survival in cutaneous melanoma. Therefore, the results indicated that increased HLA class II expression may contribute to enhanced anti-tumor immunity and related inflammatory response via presenting tumor antigens to the immune system. The expression pattern of HLA class II genes may serve as a prognostic biomarker and therapeutic targets in cutaneous melanoma.


Introduction
Cutaneous melanoma, a malignant tumor that originates from melanocytes, is an aggressive neoplasm with high metastatic potential that accounts for most of skin cancer related deaths. The algorithm for staging and management of malignant melanoma had been well established. However, reliable biomarkers for prediction of clinical outcome in patients with malignant melanoma are still under investigation. Since changes in protein expression patterns in melanoma may be associated with tumor progression, angiogenesis and metastasis [1,2], investigating the tumor-associated

Gene Expression Analysis
The expression levels of HLA class II genes in normal skin tissue and tumor samples were evaluated using the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia. cancer-pku.cn/) [18] which was established using microarray and RNA-Seq data from The Cancer Genome Atlas (TCGA) [19] and Genotype-Tissue Expression (GTEx) [20]. A total of 471 cutaneous melanoma patients with 480 clinical samples were included in the TCGA PanCanAtlas dataset (https://gdc.cancer.gov/about-data/publications/pancanatlas). Moreover, the expression of HLA class II genes in different sample types of cutaneous melanoma was evaluated by the UALCAN database (http://ualcan.path.uab.edu) [21]. The mRNA expression of HLA class II molecules in clinical cancer tissues and cell lines of cutaneous melanoma was analyzed by the Metabolic gEne RApid Visualizer (MERAV) [22].

Co-Expression Gene Analysis
To explore the HLA class II co-expressed gene, we downloaded the public microarray data of cutaneous melanoma via the web-based microarray database Oncomine (www.oncomine.org) [23] and cBioPortal [24,25]. Three cutaneous melanoma datasets were selected, including TCGA melanoma, Diagnostics 2019, 9,59 3 of 19 Harlin melanoma [26] and Bogunovic melanoma [27]. Co-expressed genes with respect to HLA-DPA1 and -DRA expression in cutaneous melanoma tissues were analyzed. The first dataset (TCGA provisional dataset, cutaneous melanoma) was composed of 109 primary melanoma samples and 371 metastatic melanoma samples. The second dataset (Harlin melanoma) was composed of 44 metastatic melanoma samples, 5 melanoma cell lines, and 3 normal melanocyte primary cell cultures. The third dataset (Bogunovic melanoma) was composed of 44 metastatic melanoma samples from 38 patients.

Survival Analysis
The Kaplan-Meier plot generation by the online databases Gene Expression Profiling Interactive Analysis (GEPIA) and UALCAN was performed to explore the association of 15 HLA class II genes and clinical outcome of cutaneous melanoma. Overall survival rate and hazard ratio estimation with respect to expression of HLA class II genes in cutaneous melanoma were also generated by GEPIA. In GEPIA database analysis, the patient samples were split into two groups (low vs. high expression). In UALCAN database analysis, the patient samples were split into three groups (low, medium and high expression). The Kaplan-Meier plot generation by UALCAN was also performed to explore the association between genes co-expressed with HLA class II genes and clinical outcome in patients affected by cutaneous melanoma.

OncoLnc Tool
OncoLnc (http://www.oncolnc.org) is a newly available resource for Cox coefficients and linking TCGA survival data to mRNA, miRNA or lncRNA expression [28]. Cox coefficients of HLA class II genes and their co-expressed genes were generated to explore the association of gene expressions and survival rate in cutaneous melanoma. Moreover, Cox coefficients of HLA class II genes were compared with other cancer types.

Gene Ontology Enrichment Analysis
Gene ontology enrichment analysis were conducted to examine HLA-DPA1 and/or HLA-DRA co-expressed genes by using the Database for Annotation, Visualization and Integrated Discovery (DAVID; http://david.abcc.ncifcrf.gov/) [29]. The categories of biological process, cellular component and molecular function were selected, and all options were set as defaults. The data listed in the table were those terms with p-value < 0.05.

Statistical Analysis
The hazard ratio and p-value in survival analysis were generated by GEPIA and UALCAN. The Cox coefficient and p-value in Cox regression analysis were generated by OncoLnc. The fold enrichment and p-value in gene otology enrichment analysis were generated by DAVID. The p-value < 0.05 was considered statistically significant difference in all analysis.

The Expressions of HLA Class II Genes Were Up-Regulated in Cutaneous Melanoma
To clarify the expression pattern of HLA class II genes, the expressions of 15 HLA class II genes in skin cutaneous melanoma from TCGA samples (N = 461) were compared with normal skin tissue from TCGA (N = 1) and Genotype-Tissue Expression (GTEx) (N = 557) samples using the online Gene Expression Profiling Interactive Analysis (GEPIA) database. The expressions of most HLA class II genes were significantly up-regulated in cutaneous melanoma, except HLA-DQB2, which was significantly down-regulated in cutaneous melanoma compared to normal skin tissues ( Figure 1). In addition, the expression levels of these HLA class II genes were observed to have wide range of distribution among cutaneous melanoma samples.
We next investigated the changes in HLA class II mRNA expression based on cutaneous melanoma stages (Stage I, II, III, and IV according to American Joint Committee on Cancer (AJCC) TNM staging) using the UALCAN web resource. The expression pattern of 15 HLA class II genes in each stage of cutaneous melanoma was displayed as heatmap in Figure 2A. A detailed analysis of expression levels between stages was also obtained from the UALCAN database. The results indicated the expression levels of HLA-DPB1, -DPB2, -DQA1, -DQB1, -DRA, -DMA and -DMB were significantly higher in Stage I than in Stage II (Stage I versus Stage II, HLA-DPB1: p = 2.68 × 10 −2 , HLA-DPB2: p = 2.70 × 10 −2 , HLA-DQA1: p = 1.32 × 10 −2 , HLA-DQB1: p = 9.46 × 10 −3 , HLA-DRA: p = 5.22 × 10 −3 , HLA-DMA: p = 3.92 × 10 −2 , HLA-DMB: p = 3.23 × 10 −4 ), while the expression levels of other HLA class II genes were not significantly different between Stage 1 and Stage 2. However, there was no statistical significance between other stages (p > 0.05), although there was a tendency of gradual increase of HLA class II gene expressions from Stage 2 to Stage 4 ( Figure 2B). We next investigated the changes in HLA class II mRNA expression based on cutaneous melanoma stages (Stage I, II, III, and IV according to American Joint Committee on Cancer (AJCC) TNM staging) using the UALCAN web resource. The expression pattern of 15 HLA class II genes in each stage of cutaneous melanoma was displayed as heatmap in Figure 2A. A detailed analysis of expression levels between stages was also obtained from the UALCAN database. The results indicated the expression levels of HLA-DPB1, -DPB2, -DQA1, -DQB1, -DRA, -DMA and -DMB were significantly higher in Stage I than in Stage II (Stage I versus Stage II, HLA-DPB1: p = 2.68 × 10 −2 , HLA-DPB2: p = 2.70 × 10 −2 , HLA-DQA1: p = 1.32 × 10 −2 , HLA-DQB1: p = 9.46 × 10 −3 , HLA-DRA: p = 5.22 × 10 −3 , HLA-DMA: p = 3.92 × 10 −2 , HLA-DMB: p = 3.23 × 10 −4 ), while the expression levels of other HLA class II genes were not significantly different between Stage 1 and Stage 2. However, there was no statistical significance between other stages (p > 0.05), although there was a tendency of gradual increase of HLA class II gene expressions from Stage 2 to Stage 4 ( Figure 2B).  We next investigated the changes in HLA class II mRNA expression based on cutaneous melanoma stages (Stage I, II, III, and IV according to American Joint Committee on Cancer (AJCC) TNM staging) using the UALCAN web resource. The expression pattern of 15 HLA class II genes in each stage of cutaneous melanoma was displayed as heatmap in Figure 2A. A detailed analysis of expression levels between stages was also obtained from the UALCAN database.

High HLA Class II mRNA Expression Levels Were Associated with a Good Prognosis in Cutaneous Melanoma Patients
The survival analysis of 15 HLA class II genes in cutaneous melanoma by online databases UALCAN revealed patients with cutaneous melanoma in high expression group of all HLA class II genes had longer survival as compared with patients in low/medium expression group (Figure 3). To verify this finding, the survival analysis and hazard ratio estimation of the expression levels of 15 HLA class II genes in cutaneous melanoma were performed using online bioinformatics database GEPIA. The overall survival between high and low expression levels of 15 HLA class II genes were displayed as survival plot in Figure 4, and the relative risk of patients with high compared to low HLA class II gene expression levels was indicated in hazard ratio ( Table 1). The expression fold changes between high and low expression groups were quantitated from the RNA-seq data downloaded from the cBioPortal database. Hazard ratios of <1.0 indicated patients with high HLA class II gene expressions had better overall survival.

High HLA Class II mRNA Expression Levels Were Associated with a Good Prognosis in Cutaneous Melanoma Patients
The survival analysis of 15 HLA class II genes in cutaneous melanoma by online databases UALCAN revealed patients with cutaneous melanoma in high expression group of all HLA class II genes had longer survival as compared with patients in low/medium expression group (Figure 3). To verify this finding, the survival analysis and hazard ratio estimation of the expression levels of 15 HLA class II genes in cutaneous melanoma were performed using online bioinformatics database GEPIA. The overall survival between high and low expression levels of 15 HLA class II genes were displayed as survival plot in Figure 4, and the relative risk of patients with high compared to low HLA class II gene expression levels was indicated in hazard ratio ( Table 1). The expression fold changes between high and low expression groups were quantitated from the RNA-seq data downloaded from the cBioPortal database. Hazard ratios of < 1.0 indicated patients with high HLA class II gene expressions had better overall survival.

The Expressions of HLA Class II Genes in Cutaneous Melanoma Were Most Closely Correlated with Patient Survival among 21 Common Cancer Types in TCGA Database
To investigate the effect of HLA class II gene expressions on survival among various cancer types, including cutaneous melanoma, the mRNA levels of HLA class II genes were examined using the bioinformatics tool OncoLnc, and the Cox regression results and the p-value ranks for the 15 HLA class II genes in cutaneous melanoma were listed in Table 2. A negative Cox coefficient indicated higher gene expression reduced the risk of death. Similar to results obtained from UALCAN and

The Expressions of HLA Class II Genes in Cutaneous Melanoma Were Most Closely Correlated with Patient Survival among 21 Common Cancer Types in TCGA Database
To investigate the effect of HLA class II gene expressions on survival among various cancer types, including cutaneous melanoma, the mRNA levels of HLA class II genes were examined using the bioinformatics tool OncoLnc, and the Cox regression results and the p-value ranks for the 15 HLA class II genes in cutaneous melanoma were listed in Table 2. A negative Cox coefficient indicated higher gene expression reduced the risk of death. Similar to results obtained from UALCAN and GEPIA databases, the expressions of HLA class II genes in cutaneous melanoma was highly correlated to patients' overall survival. Among the 15 HLA class II genes, HLA-DP and -DR genes were found to have higher expression levels and p-value rankings in association with survival. Furthermore, the expressions of HLA class II genes in cutaneous melanoma were found more closely associated with survival in cutaneous melanoma as compared to other cancer types. As shown in Table 3, HLA-DPA1 is the 80th gene most closely correlated with survival in cutaneous melanoma, although the expression level of HLA-DPA1 in cutaneous melanoma was not the highest among 21 common cancer types. As shown in Table 4, HLA-DRA is the 46th most closely correlated gene in cutaneous melanoma. For both HLA-DPA1 and HLA-DRA, the rank of correlations is highest in cutaneous melanoma, followed by lung adenocarcinoma, sarcoma and breast invasive carcinoma. Instead of focusing on p-values, it was suggested to focus more on the rank of the correlations, and also on the Cox coefficients in analysis by OncoLnc tool. The Cox coefficient and p-value are from the gene term in precomputed multivariate Cox regressions. The false discovery rate (FDR) correction is performed per cancer analysis per data type. The rank is also performed per cancer per data type.

Top 11 Genes Co-Expressed with HLA-DPA1 and HLA-DRA in Clinical Cutaneous Melanoma Samples
The expression patterns of HLA class II genes in cutaneous melanoma from primary tumor samples and tumor cell lines were obtained from Oncomine database [23], and we found the expression patterns of HLA class II genes were different between cell lines and primary tumor samples. Specifically, as compared with in vitro cell lines of cutaneous melanoma, HLA class II gene expressions in resected tissues of melanoma tumors were generally higher, particularly HLA-DMA, -DMB, -DPA1, -DPB1, -DQA1, -DQB1, -DRA and -DRB1, as displayed in Figure 5. The phenomenon suggested that the tumor microenvironment may closely regulate the expression of these genes. Thus, we further analyzed the co-expressed genes of HLA class II genes to investigate the biological changes related to aberrant expression of HLA class II in cutaneous melanoma tissues. Using the Oncomine database and cBioPortal for public microarray data of cutaneous melanoma, we selected three datasets composed of melanoma tumor samples to analyze the genes co-expressed with HLA class II genes, specifically HLA-DPA1 and HLA-DRA. In the analysis of HLA-DPA1 co-expressed gene, 14 genes were consistently identified as top 100 genes by co-expression score in the three datasets of cutaneous melanoma; in the analysis of HLA-DRA co-expressed gene, 13 genes were consistently identified as top 100 genes by co-expression score in the three datasets of cutaneous melanoma ( Figure 6). The genes consistently identified to be closely co-expressed to HLA class II genes were listed in Table 5, including 11 genes (APOL3, CD74, CTSS, CXCR3, C1QA, C1QB, ITGB2, LAPTM5, NCF1C, SLAMF8 and TNFRSF1B) other than HLA class II genes. gene, 14 genes were consistently identified as top 100 genes by co-expression score in the three datasets of cutaneous melanoma; in the analysis of HLA-DRA co-expressed gene, 13 genes were consistently identified as top 100 genes by co-expression score in the three datasets of cutaneous melanoma ( Figure 6). The genes consistently identified to be closely co-expressed to HLA class II genes were listed in Table 5, including 11 genes (APOL3, CD74, CTSS, CXCR3, C1QA, C1QB, ITGB2, LAPTM5, NCF1C, SLAMF8 and TNFRSF1B) other than HLA class II genes.   datasets of cutaneous melanoma; in the analysis of HLA-DRA co-expressed gene, 13 genes were consistently identified as top 100 genes by co-expression score in the three datasets of cutaneous melanoma ( Figure 6). The genes consistently identified to be closely co-expressed to HLA class II genes were listed in Table 5, including 11 genes (APOL3, CD74, CTSS, CXCR3, C1QA, C1QB, ITGB2, LAPTM5, NCF1C, SLAMF8 and TNFRSF1B) other than HLA class II genes.    To identify the mechanisms underlying the expression of HLA class II genes and its co-expressed genes, the gene ontology enrichment analysis was conducted using Database for Annotation, Visualization and Integrated Discovery (DAVID) bioinformatics tool. Overall, 10 biological processes, 18 cellular constituents, and 4 molecular function terms were significantly enriched (p < 0.05, Table 6). Based on the results of the analysis, biological functions related to antigen processing and presentation and immune response were most functionally enriched, with CTSS, CD74 and TNFRSF1B involved in these biological functions, in addition to HLA class II related genes. In addition, CD74 was the only related gene classified in the MHC class II protein complex binding, other than HLA class II related genes. Noticeably, four genes (APOL3, TNFRSF1B, ITGB2 and CXCR3) were linked to the biological process term "inflammatory response" (GO:0006954). These co-expressed genes were input into Search Tool for the Retrieval of Interacting Genes (STRING) database [30] and OmicsNet [31] for interaction network visualization. In the network produced by STRING database, CD74 and CTSS were clustered with HLA class II related genes ( Figure 7A), while in the major subnetwork obtained from OmicsNet based on InnateDB database, CD74 was the molecule in close conjunction to HLA-DPA1 and -DRA ( Figure 7B). Overall, the functional enrichment analysis results suggested that the cluster of genes co-expressed with HLA-DPA1 and -DRA may promote anti-tumor immune response through antigen processing and presentation via MHC class II. To evaluate the significance of these 11 co-expressed genes in clinical outcome, the association of their expression levels with survival was assessed using the Kaplan-Meier plotter tool in the    To evaluate the significance of these 11 co-expressed genes in clinical outcome, the association of their expression levels with survival was assessed using the Kaplan-Meier plotter tool in the UALCAN web resource. The survival analysis showed higher expression level of all these 11 genes predicted a better survival in cutaneous melanoma ( Figure 8). Correlation between expression level of HLA class II co-expressed genes and survival in cutaneous melanoma was also examined using the bioinformatics tool OncoLnc. Similar to the results from UALCAN database, Cox coefficients of these genes were all negative, which indicated that higher expression of these genes reduce the risk of death. APOL3, CD74, C1QA and C1QB were found to rank highest in association with survival among all co-expressed genes ( Table 7). of HLA class II co-expressed genes and survival in cutaneous melanoma was also examined using the bioinformatics tool OncoLnc. Similar to the results from UALCAN database, Cox coefficients of these genes were all negative, which indicated that higher expression of these genes reduce the risk of death. APOL3, CD74, C1QA and C1QB were found to rank highest in association with survival among all co-expressed genes ( Table 7).

Discussion
The present study demonstrated the novel findings that mRNA expressions of HLA class II genes, especially HLA-DP and -DR, were significantly increased in clinical cutaneous melanoma samples and positively associated with overall survival, while multivariate Cox regression analysis revealed that the HLA class II gene expressions were most closely associated with survival in cutaneous melanoma compared to other cancer types. Another novel finding in the present study is

Discussion
The present study demonstrated the novel findings that mRNA expressions of HLA class II genes, especially HLA-DP and -DR, were significantly increased in clinical cutaneous melanoma samples and positively associated with overall survival, while multivariate Cox regression analysis revealed that the HLA class II gene expressions were most closely associated with survival in cutaneous melanoma compared to other cancer types. Another novel finding in the present study is the identification of HLA class II co-expressed genes, including APOL3, CD74, CTSS, CXCR3, C1QA, C1QB, ITGB2, LAPTM5, NCF1C, SLAMF8 and TNFRSF1B. These co-expressed genes were associated with antigen processing and presentation, immune response and inflammatory response. The expression levels of these co-expressed genes were positively associated with overall survival in patients with cutaneous melanoma. To the best of our knowledge, this is the first study to comprehensively analyze the microarray and RNA-Seq data of HLA class II genes and their co-expressed genes in cutaneous melanoma datasets with largest sample size.
The significant association of HLA class II gene expressions with overall survival found in this study was consistent with the results obtained in large B-cell lymphoma [32], colorectal cancer [33][34][35], breast cancer [36,37], ovarian cancer [38], and laryngeal squamous cell carcinoma [39], as well as in metastatic melanoma [40]. In metastatic melanoma, positive staining for MHC-II expression in Stage III and IV cutaneous melanoma correlated with longer overall survival [40]. These studies suggested the functional role of HLA class II genes in tumor suppression. However, the available information for prognostic value of HLA class II antigens among different cancer types remained limited. In the present study, we found the rank of significance between HLA class II genes and patient overall survival was highest in cutaneous melanoma, suggesting that HLA class II genes possess greatest effect on overall survival of patients with cutaneous melanoma than other cancer types. This finding was compatible with the previous observation of special immunobiology in cutaneous melanoma [41,42]. Interestingly, cutaneous melanoma has been considered as a 'model disease' to investigate tumor immunobiology, to unveil the molecular basis underlying the interactions between neoplastic cells and host's immune system, and ultimately to set up new bio-immunotherapeutic approaches [5]. However, the findings in this study were not consistent with previous study in metastatic melanoma patients, in which a low expression of HLA class II genes on neoplastic cells was associated with longer survival [43]. In that study, only 48 locoregional metastatic tumors from 39 patients were analyzed. The number of samples may be too small to draw conclusions and account for the discrepancy among these studies.
Previously, some studies focused on the change in HLA class II gene expressions during progression of cutaneous melanoma and underlying regulatory mechanisms. Here, we demonstrated that the expression levels of HLA-DP, -DQ, -DR, and -DM were significantly lower in Stage II than in Stage I cutaneous melanoma, while there was a tendency toward gradual increase in HLA class II gene expressions from Stage II to Stage IV ( Figure 2). This finding was similar to the previous observation of dynamic expression of MHC during melanoma progression, and the potential mechanisms for the overexpression of MHC molecules in earlier stage were associated with the increased rates of DNA copy number gains in primary melanoma [10]. In metastatic melanoma, the master regulator of MHC genes, Class II Major Histocompatibility Complex Transactivator (CIITA), was significantly downregulated compared to vertical growth phase melanomas, which was considered to account for the discrepancy between expression level and DNA copy number [10,44,45].
In this study, survival analysis showed higher expression level of HLA class II co-expressed genes, especially APOL3, CD74, C1QA and C1QB, were associated with better prognosis in cutaneous melanoma. Surprisingly, APOL3, which encoded Apolipoprotein L3 [46] was found to be the 58th most closely correlated gene with patient survival in cutaneous melanoma, which was as high as HLA class II genes. The high expression of APOL3 was reported to predict worse clinical outcome in patients with acute myeloid leukemia [47]. However, no study related to the functional role of APOL3 in cutaneous melanoma has been reported. Recent studies have emphasized the association between metabolic syndrome and increased cancer risks, including skin cancer [48,49]. Non-alcoholic fatty liver disease (NAFLD), related to obesity and metabolic syndrome, has been proposed an added risk factor for extra-hepatic cancers [49][50][51], including melanoma (standardized incidence ratio of 2.4) [49]. The evidence suggests the importance of lipid metabolism in cancer development. However, whether APOL3 directly bridge the relation between cutaneous melanoma and metabolic syndrome cannot be clarified from the results of the present study.
CD74, along with CTSS and HLA class II genes, were found involved in antigen processing and presenting via MHC-II by gene ontology enrichment analysis, while interaction network analysis identified CD74 to be in close conjunction to HLA-DPA1 and -DRA. The Cox regression analysis also indicated CD74 as one of the higher-ranking genes positively correlated to overall survival in patients with melanoma. CD74 is associated with development of specific adaptive immune response, and higher CD74 expression correlated with improved recurrence-free and overall survival [52], regulated by interferon-gamma [53]. In brain metastasis, CD74 is also a potent positive prognostic marker for survival [54]. Blockade of CD74 related immunosuppressive signaling may restore anti-tumor immune response in metastatic melanoma, suggesting the crucial role of CD74 in immunotherapy [55].
C1QA and C1QB encode complement C1q subunit A and B, which are the initiator of the classical complement pathway and can induce dendritic cell maturation to generate Th1 type response [56,57]. C1q appears to have pro-tumorigenic and anti-tumorigenic role in cancer, depending on the context of the disease [58]. The anti-cancer role of C1q has been reported in prostate cancer [59], as well as in syngeneic murine model of melanoma [60]. LAPTM5 (lysosomal-associated protein transmembrane 5) is one of the top genes significantly associated with longer survival in metastatic melanoma [61], and ITGB2 (integrin beta 2) co-expression with HLA-DR participates in signal transduction, cell adhesion, and motility in neoplastic melanocytes [5,62]. However, the prognostic value of ITGB2 in cutaneous melanoma has not yet been reported. The other co-expression genes, including TNFRSF1B, CXCR3, SLAMF8 and NCF1, appeared to be involved in apoptosis and inflammatory response [63,64], lymphocyte activation and T cell trafficking [65]. Taken together, the HLA class II co-expressed genes may represent the anti-tumor immune response through antigen processing as well as inflammatory response in tumor microenvironment. However, the causality of expression of HLA class II genes and their co-expressed genes in cutaneous melanoma still need to be verified in further studies.
The limitation of the current study is that the results were derived based on public cancer data repository, therefore, we validated the gene expression patterns and the results of survival analysis in the different databases available. To further investigate the prognostic value of HLA class II genes in cutaneous melanoma, collection of clinical samples with longitudinal follow up is necessary.

Conclusions
The expression levels of HLA class II in cutaneous melanoma may reflect the tumor neo-antigen presentation by tumor specific MHC-II molecules and, therefore, correlate with clinical outcome. The results of the present study indicate that the expression pattern of HLA class II and their co-expressed gene may serve as prognostic biomarkers in prediction of overall survival. A schematic summary is shown in Figure 9. Therapeutic strategies aiming to increase HLA class II expression in cutaneous melanoma may prolong overall survival, especially in combination with other immunotherapy strategies. Future studies focusing on the mechanisms involved in the regulation of expression and activity of HLA class II genes may provide further insights into the treatment of cutaneous melanoma. Public Databases: The GEPIA [18] is an interactive web server for analyzing the RNA sequencing expression data of tumors and normal samples from the TCGA and the GTEx projects to provide differential expression analysis, survival analysis, and correlation analysis. The UALCAN [21] is an interactive web resource for analyzing cancer genomics data and these resources allow researchers to gather valuable information and data about the genes of interest. The MERAV [22] is a web-based tool for whole genome analysis which query a database comprising of about 4300 microarrays, representing human gene expression in normal tissues, cancer cell lines and primary tumors, and offers advantages in comparing gene expression among different tissue types as well as between normal and cancer cells and generating heatmaps. The Oncomine Platform [23] is a webbased microarray database that provides peer-reviewed analysis methods and contains a powerful set of analysis functions that compute gene expression signatures and automatically extracts biological insights from the data. The cBioPortal [24,25] provides visualization, analysis and download of large-scale cancer genomics datasets. The OncoLnc [28] contains survival data from patients of 21 cancer types, and is a tool for interactively exploring survival correlations and downloading clinical data coupled with mRNA expression data.  Public Databases: The GEPIA [18] is an interactive web server for analyzing the RNA sequencing expression data of tumors and normal samples from the TCGA and the GTEx projects to provide differential expression analysis, survival analysis, and correlation analysis. The UALCAN [21] is an interactive web resource for analyzing cancer genomics data and these resources allow researchers to gather valuable information and data about the genes of interest. The MERAV [22] is a web-based tool for whole genome analysis which query a database comprising of about 4300 microarrays, representing human gene expression in normal tissues, cancer cell lines and primary tumors, and offers advantages in comparing gene expression among different tissue types as well as between normal and cancer cells and generating heatmaps. The Oncomine Platform [23] is a web-based microarray database that provides peer-reviewed analysis methods and contains a powerful set of analysis functions that compute gene expression signatures and automatically extracts biological insights from the data. The cBioPortal [24,25] provides visualization, analysis and download of large-scale cancer genomics datasets. The OncoLnc [28] contains survival data from patients of 21 cancer types, and is a tool for interactively exploring survival correlations and downloading clinical data coupled with mRNA expression data.