CXC Chemokines as Therapeutic Targets and Prognostic Biomarkers in Skin Cutaneous Melanoma Microenvironment

Background Skin Cutaneous Melanoma (SKCM) is a tumor of the epidermal melanocytes induced by gene activation or mutation. It is the result of the interaction between genetic, constitutional, and environmental factors. SKCM is highly aggressive and is the most threatening skin tumor. The incidence of the disease is increasing year by year, and it is the main cause of death in skin tumors around the world. CXC chemokines in the tumor microenvironment can regulate the transport of immune cells and the activity of tumor cells, thus playing an anti-tumor immunological role and affecting the prognosis of patients. However, the expression level of CXC chemokine in SKCM and its effect on prognosis are still unclear. Method Oncomine, UALCAN, GEPIA, STRING, GeneMANIA, cBioPortal, TIMER, TRRUST, DAVID 6.8, and Metascape were applied in our research. Result The transcription of CXCL1, CXCL5, CXCL8, CXCL9, CXCL10, and CXCL13 in SKCM tissues were significantly higher than those in normal tissues. The pathological stage of SKCM patients is closely related to the expression of CXCL4, CXCL9, CXCL10, CXCL11, CXCL12, and CXCL13. The prognosis of SKCM patients with low transcription levels of CXCL4, CXCL9, CXCL10, CXCL11, and CXCL13 is better. The differential expression of CXC chemokines is mainly associated with inflammatory response, immune response, and cytokine mediated signaling pathways. Our data indicate that the key transcription factors of CXC chemokines are RELA, NF-κB1 and SP1. The targets of CXC chemokines are mainly LCK, LYN, SYK, MAPK2, MAPK12, and ART. The relationship between CXC chemokine expression and immune cell infiltration in SKCM was closed. Conclusions Our research provides a basis for screening SKCM biomarkers, predicting prognosis, and choosing immunotherapy.


INTRODUCTION
Skin Cutaneous Melanoma is a common skin tumor caused by abnormal hyperproliferation of melanocytes (1). Its incidence varies with race, region, and age. The incidence of white people is much higher than that of black people, with the incidence of white people living in Queensland, Australia reaching 17/100,000 (2). In recent decades, the incidence of melanoma has continued to increase worldwide. It is the fastest-growing malignant disease in men and second only to lung cancer in women, with annual growth rates of 3-5% (3). Melanoma ranks 5th and 6th in malignant diseases in males and females, respectively, second only to adult leukemia in terms of risk of death (4). The median age of onset was 45-55 years old. Risk factors for melanoma include clear family history, melanoma history, multiple atypical or dysplastic nevi, and congenital genetic mutations (5). Sun exposure may also promote the development of melanoma (5). The association between molecular biological indicators and tumor prognosis has aroused great attention of researchers lately. Studies have revealed the relationship between the expression of S-100, Vimentin, and other proteins and tumor metastasis and prognosis (6). However, more effective therapeutic targets and more sensitive markers related to prognosis are still to be developed.
Chemokines are a superfamily composed of small molecules of cytokine-like proteins (7). CXC chemokines are an important member of this family (8). They occupy a major position in inflammation and damage repair and are closely linked with the occurrence, development, invasion, and metastasis of tumors (9). The function of CXC chemokines in tumors is deep association with the existence of ELR domains (10). CXC chemokines containing ELR domains are mainly concerned with the growth, proliferation, metastasis, and angiogenesis regulation of tumor cells. The diversity of functions of CXC chemokines without ELR domains in tumors may be related to the difference of ELR-CXC chemokine binding receptors. Therefore, we therefore speculate that CXC chemokines may also occupy and major position in various biological processes of SKCM.
In SKCM there are many updated reviews on molecular factors important for biology, drug targeting, and prognosis (BRAF, MEK, PD-1 et. al) (11)(12)(13). However, few studies have focused on the mechanism of CXC chemokine in SKCM and its potential value. Studies have reported the expression and function of some members of CXC chemokines in SKCM. However, so far it is still unknown which CXC can be used as a biomarker for prognosis and a target for immunotherapy for SKCM. With the continuous update and iteration of a variety of biological detection methods, it is possible to comprehensively analyze CXC chemokines.
Hence, the focus of this research is to find biological targets from CXC that can be used as the diagnosis and treatment of SKCM. In addition, we hope to discover the specific molecular mechanism of CXC affecting prognosis of SKCM through bioinformatics analysis.

MATERIALS AND METHODS
ONCOMINE ONCOMINE (http://www.oncomine.org) is a great tumor gene chip database, which contains multiple functions include finding outliers, predicting co-expressed genes, and analyzing gene expression differences (14). It can also be classified according to clinical information such as tumor stage, grade, and tissue type, and can also be used to search for possible diagnostic biomarkers and therapeutic targets. In this study, figure was exported to show the expression difference of CXC chemokines in SKCM through analysis from this website. The expression differences of CXC chemokines in SKCM were analyzed by a Student t-test. ONCOMINE contains 715 datasets and 86,733 samples. UALCAN UALCAN (http://ualcan.path.uab.edu/index.html) can effectively analyze and mine cancer data on-line (15). The tumor-related data of the TCGA database is the basis of data mining on this website. The function of UALCAN includes biomarker identification, expression profile analysis, survival analysis, etc. In our study, the SKCM data set was import into UALCAN Expression Analysis module to evaluate the CXC chemokines expression. Student t-test was used to analyzed SKCM data.
GEPIA GERIA (http://gepia.cancer-pku.cn/index.html) is a newly developed interactive web server for analyzing the RNA sequencing expression data of 9,736 tumors and 8,587 normal samples from the TCGA and the GTEx projects, using a standard processing pipeline (16). It fills the gap of big data information in cancer genomics. GEPIA analyzed RNA sequencing expression data from 9,736 tumors and 8,587 normal samples from the TCGA and GTEx projects. The expression data of TCGA and GTEx are calculated under the same pipeline, which can be directly analyzed in a very comprehensive way. The database is an open public database. In this study, we used GEPIA's "Single Gene Analysis" to analyze the differential expression of CXC chemokines in tumor tissues and normal tissues, as well as pathological stage Analysis and prognosis Analysis. "Multiple Gene Comparison" was used for the polygenic comparative analysis of CXC chemokines in the "SKCM" data set. Student t-test was used to explore SKCM data. STRING STRING (https://string-db.org/) is a database that searches for known interactions between proteins and predicts their interactions (17). The database, available for 2,031 species, contains 9.6 million proteins and 13.8 million protein-protein interactions. It contains experimental data, PubMed abstracts, and other database data, as well as results predicted by bioinformatics methods. In this study, STRING was used for PPI analysis of CXC chemokines and peripheral interaction gene.

GeneMANIA
The function of GeneMANIA (http://www.genemania.org) is to predict the interactions between proteins, including Predicted, Physical Interactions, co-location, co-expression, Pathway, Shared protein domains, and Genetic Interactions (18). GeneMANIA has almost 2,277 networks that collectively contain nearly 600 million interactions covering almost 163,599 genes. cBioPortal cBioPortal (www.cbioportal.org) integrates data from 126 tumor genome studies (19). These included large tumor research projects such as TCGA and ICGC, which included data from 28,000 samples, and some samples with phenotypic information such as clinical prognosis. Genetic variation, gene network, and co-expression of CXC chemokines in SKCM were analyzed by cBioPortal. TIMER TIMER (https://cistrome.shinyapps.io/timer/) using a deconvolution algorithm from the gene expression profile (TIICs) concluded that tumor-infiltrating immune cells in abundance (20). Gene expression data of 10,897 TCGA samples from 32 cancer types were reanalyzed to estimate the abundance of six TIIC subsets (B cells, CD4+T cells, CD8+T cells, macrophages, neutrophils, and dendritic cells) to establish a link between tumor immunity and genomic data. The web server provides the abundance of immune infiltrates estimated by multiple immune-deconvolution methods and allows users to dynamically generate high-quality graphics to fully explore the immunological, clinical, and genomic characteristics of tumors.
TRRUST TRRUST (https://www.grnpedia.org/trrust/) is a record of transcription factors regulating the relationship database (21). TRRUST contains 8,444 and 6,552 TF-target regulatory relationships of 800 human TFs and 828 mouse TFs (21). It includes not only the corresponding target genes of transcription factors but also the regulatory relationship between transcription factors. At present, the database only stores regulatory information related to humans and mice, and these regulatory relations are sorted out from the literature through text mining. DAVID 6.8 DAVID 6.8 integrates biological data and analysis tools to provide systematic and comprehensive biological functional annotation information for the large-scale gene or protein lists (hundreds of gene or protein ID lists) to help users extract biological information from DAVID 6.8 (https://david.ncifcrf. gov/summary.jsp) (22). GO enrichment analysis and KEGG analysis on CXC chemokines and 50 closely related genes were performed by applying DAVID.

METASCAPE
Metascape (http://metascape.org) integrates more than 40 bioinformatics databases (23). It not only includes bio-pathway enrichment analysis, protein interaction network structure analysis, and abundant gene annotation functions but also presents the results in high-quality graphical language that biologists can easily understand. The enrichment analysis of CXC chemokines and highly interacting genes were verified by the Express Analysis module.

QRT-PCR ANALYSIS
10 SKCM tissues and 12 normal tissues were enrolled in our research to verify the CXC chemokines expression. We obtained consent of patients and ethical approval from the Xiangya hospital ethics committee. All procedures were executed by the ethics guidelines and regulations. We extracted total RNA of the tissue by using Trizol reagent (Invitrogen, CA, USA). qRT-PCR was performed with SYBR ® Green dye (TaKaRa, Shiga, Japan). The primer sequences of relative genes are provided in Supplementary Table 1.

CXC Chemokines in SKCM Patients
The expression differences of 16 CXC chemokines between SKCM patients and controls were analyzed in the Oncomine database ( Figure 1 and Table 1). Analysis results showed that the transcription levels of CXCL1, CXCL5, CXCL8, CXCL9, CXCL10, and CXCL13 in SKCM patients were significantly increased compared with normal skin tissues. Risker et al. confirmed that the transcription level of CXCL1, CXCL5, CXCL8, and CXCL13 in SKCM patients were significantly higher than those in normal skin tissues (24). Haqq et al. also proved that the transcription of CXCL1, CXCL9, and CXCL10 in melanoma patients is significantly higher than that in normal skin tissues with a P-value of 0.005, 6.75E-4, and 2.52E-4 (25). Talatov et al. also confirmed that the transcription level of CXCL8 was significantly increased with a P-value of 0.004 (26). There was no significant difference in the transcriptional levels of SKCM patients and normal people among CXCL2, CXCL3, CXCL4, CXCL6, CXCL7, CXCL11, CXCL12, CXCL14, CXCL16, and CXCL17.

CXC Chemokines Affect Prognostic of SKCM Patients
We evaluated the worth of different CXC chemokines in SKCM clinical outcomes by using GEPIA. No obvious correlation between the level of CXC chemokine family transcription and disease-free survival time in SKCM patients was found ( Figure  5). The worth of CXC chemokines with different expressions in the total survival of SKCM patients was assessed ( Figure 6). The results showed that low expression of CXCL4 (p = 0.0028), CXCL9 (P = 0.00024), CXCL10 (P = 2.8E-05), CXCL11 (P =1.9e-05), and CXCL13 (P = 2.7e-05) was remarkably related to longer overall survival in SKCM.

Biological Process Analysis of CXC Chemokines
The enrichment analysis of CXC chemokines and its closely interacting neighboring genes is performed by DAVID 6.8 and Metascape. Figure 8A shows the first 10 items using DAVID 6.8 for the functional enrichment analysis. In the BP category, inflammatory response, immune response, cytokine-mediated signaling pathway, leukocyte chemotaxis, positive regulation of ERK1 and ERK2, and cellular response to interferon-g, interleukin-1, and tumor necrosis factor were associated with SKCM oncogenesis and progression. In the CC category, the extracellular space, external side of the plasma membrane, extracellular region, cell, and Pseudopodium are the five functional enrichment projects. In the MF category, the CXC chemokines and its adjacent genes are mostly enriched in CXC chemokine receptor binding, C-C chemokine receptor activity, C-X-C chemokine receptor activity, CXCR Chemokine receptor binding, MAP kinase activity, cytokine activity, heparin-binding, and growth factor activity. KEGG pathway analysis was also performed, and the results showed the chemokine signaling pathway, cytokine receptor interaction, TNF signaling pathway, toll-like receptor signaling pathway, Leishmaniasis, Intestinal Immune Network for IgA production, Influenza, and NOD-like receptor signaling Pathway, Chagas disease, inflammatory disease, tuberculosis, T cell signaling pathway are closely related to tumor formation ( Figure 8B).
We also used Metascape for enrichment analysis. The functions of CXC chemokines and their adjacent genes are mostly focused on   Figures 9A, B). In order to explore the interaction between 16 CXC chemokines and its closely related gene, we analyzed the related data using the PPI network and mCODE components (Figures 9C, D). The result showed that the CXC chemokines biological functions are mainly associated with chemokine receptors bind chemokines, chemokine-mediated signaling pathways, cellular response to chemokine, IL-7 signaling pathway, AGE-RAGE signaling pathway, and TNF signaling pathway ( Figure 9E).

The Association of CXC Chemokines With Immune Cell Infiltration
Clinical outcome is related to inflammatory response and immune cell infiltration in SKCM patients, in which CXC chemokines are also involved. Timer database was applied to comprehensively quest whether CXC chemokines expression is associated with immune infiltration in SKCM patients (Figrues 10A-P and Table 4).

DISCUSSION
Chemokines are tiny protein molecules whose main function is associated with inflammation and immunity (27). Chemokines have been divided into five main subfamilies: CXC, CC, CX3C, XC, and CX (28). CXC chemokines occupy a vital position in the chemokine family. They have been classified into ELR + CXC and ELR − CXC according to the motif with or without Glu leu-ARG (ELR) (29). Chemokine receptors are G-protein-coupled transmembrane receptors that mediate the function of chemokines and contain 7 transmembrane regions (30,31). CXC chemokines and their receptors are expressed in a variety of cells and play an important role in the generation, differentiation, development regulation, immune response regulation of immune cells, and bone marrow hematopoietic cells (32). They can promote angiogenesis, tumor cell proliferation, survival, and dissemination. They also participate in the process of organ-specific metastasis of malignant tumors and play an important role in the occurrence and development of tumors (33,34). However, there are few studies on the potential value of CXC's clinical diagnosis and treatment guidance in SKCM. Therefore, exploring the pathological and molecular mechanisms of CXC chemokines in SKCM is basis for clinical diagnosis and treatment.
To probe CXC chemokines and its relationship with the pathological stage in SKCM patients, mRNA analyses were performed for the 16 CXC chemokines using ONCOMINE and GEPIA. Among the 16 CXC chemokines, six chemokines (CXCL1, CXCL5, CXCL8, CXCL9, CXCL10, and CXCL13) were identified as being expression difference between SKCM patient tissues and normal tissues. ONCOMINE database analysis results showed that the expression levels of these six CXC chemokines in SKCM tissues were significantly higher than that in normal tissues. Further validation by clinical samples, the mRNA level of CXCL1, CXCL5, CXCL8, CXCL9, CXCL10, and CXCL13 was considerably adjusted. One research has pointed out that CXCL1 down-regulation leads to reduced colorectal cell viability, invasion, and proliferation. In vivo, knockdown of CXCL1 resulted in the prevention of tumor growth in nude mice (35). CXCL1 promotes tumor growth through VEGF pathway activation and is associated with inferior survival in gastric cancer (36). Therefore, the high expression of CXCL1 might promote melanoma cell viability, invasion, and proliferation. Neutrophils may be the cause of elevated CXC5 expression in primary melanoma (37). The upregulation of CXCL8 and its receptors CXCR1 and CXCR2 may be one of the reasons for the development of SKCM. CXCL8 promotes tumor angiogenesis by maintaining the proliferation and activity of endothelial cells, and CXCR1 and CXCR2 are the basis of this function (38). CXCL9 promotes the metastasis of melanoma cells through its combination with CXCR3 to enhance the permeability of tumor blood vessels (39). High expression of CXCL10 will inhibit the immune response mediated by T cells, leading to an increase in tumor growth rate (40). The CXCL13: CXCR5 axis is an fundamental regulatory component in the biological process of SKCM (41). Overall, these findings match the results of our data mining. Besides, GEPIA was used to explore whether there is a connection between CXC with the survival rate in SKCM. The increased expression of six CXC chemokines (CXCL4, CXCL9, CXCL10, CXCL11, CXCL12, and CXCL13) was related to tumor progression. The SKCM patients with low expression of them had better overall survival. Previous studies have demonstrated that CXC chemokines may also play an important role in the progression of SKCM (42,43).
The expression differences of CXC chemokines between primary SKCM and metastatic SKCM were explored by using UALCAN database. The result showed that CXCL1 and CXCL7 were higher in the primary SKCM than in the metastatic SKCM. One research indicates that primary melanoma cells might down-regulate the invasion activity of metastatic melanoma   (44). Therefore, we speculated that the high expression of CXCL1 in primary SKCM can inhibit tumor metastasis. So far, no studies on CXCL7 in primary and metastatic SKCM have been found. The studies of other tumors showed that CXCL7 was closely related to tumor metastasis, and the high expression of CXCL7 would cause tumor metastasis. This contradicts the results of our analysis. There are two possible reasons for considering UALCAN database analysis results. The first reason is that the amount of data collected by the UALCAN database is insufficient, so there is a deviation in the analysis results. The second reason is that CXCL7 expression may be higher in primary SKCM than in metastatic SKCM, but there is no specific mechanism to explain. In the future study, samples will be collected for verification and related mechanism exploration. As suggested by the results of the DAVID analysis, the CXC chemokines and 50 most closely interacting genes were enriched for the terms inflammatory response and immune response in the "biological process" category, the term extracellular space and external side of the plasma membrane in the "cellular component" category, chemokine activity and chemokine receptor binding in the "molecular function" category. The KEGG pathway analysis showed the enrichment of 50 most closely interacting genes for the terms of the chemokine signaling pathway, cytokine-cytokine receptor interaction, and TNF signaling pathway. The enrichment analysis results of Metascape were similar with DAVID. This is following the results reported before, which have shown that CXC chemokines signaling pathways occupy a mayor position in multiple biological activities of tumors (45,46). A large number of clinical and epidemiological studies have shown that 15-20% of a malignant tumor is caused by infection and inflammation of the controllability, such as inflammatory bowel disease associated with colon cancer, chronic hepatitis B virus infection can lead to liver cancer, Helicobacter pylori infection was significantly associated with gastric cancer, EB virus infection can cause nasopharyngeal carcinoma, the human papillomavirus infection can cause cervical cancer or Burkitt lymphoma, etc. (47). Chronic inflammation is involved in the pathogenesis, development, invasion, and metastasis of malignant tumors (48). These results indicate that CXC chemokines regulate the progress of SKCM by regulating inflammation and immune responses.
TRRUST database was used to explore the transcription factor targets of CXC chemokines with different expressions. The analysis results showed that the transcription factors which play an important in regulating CXC chemokine are RELA, NFKB1, and SP1. NF-KB pathway was regulated by Phosphorylated RELA which is related to the course of tumor, and inflammation-related diseases (49). RELA has been shown to play a key role in mediating cancer-induced senescence in precancerous lesions (50). NF-kB1 inhibits the occurrence and development of a variety of cancers by reducing the overexpression of the NF-kB signaling pathway (51). Some research has showed that Sp1 is overexpressed in cancer cells and contributes to the formation of cancer (52). Inhibiting the expression level of Sp1 can significantly inhibit the proliferation      of human malignant melanoma cells (53). The above results confirmed that RELA, NF-KB, and Sp1 play an important role in CXC chemokines regulation of SKCM. LinkedOmics database was applied to explore the kinase target networks of CXC chemokines in SKCM. The result suggested that LCK, LYN, FYN, MAPK1, MAPK3, and CSNK1D may be targets for differential expression of CXC chemokines. These kinases influence cancer formation and progress by modulating cancer cell migration, invasion, and apoptosis. The results of our observation indicated that these kinases were the important regulator of CXC chemokines in SKCM. CXC chemokines can mediate the migration and localization of immune cells (54,55). Timer database was applied to analyze the association between CXC chemokines and immune cells. There is growing confirmation showed that immune cell infiltration may influence cancer development and recurrence and is an important determinant of immunotherapy response and clinical outcome (56,57). The results showed that CXC chemokines and immune cell infiltration levels were positively correlated in SKCM. Macrophage, CXCL2, CXCL14, and CXCL17 were closely related to the clinical outcome of SKCM patients. These strongly confirmed the positive correlation between CXC chemokines and immune infiltration in SKCM. Thus, it is likely that CXC chemokines not only as being diagnostic biomarkers but also affect the immune condition in SKCM.
There are some limitations in our research. Analysis of transcription can indicate immune condition but not integrated status. Also, vivo and vitro research should be conducted to verify our results. Our findings will offer new understanding to support the program of new immunotherapy, assist doctors select efficacious drugs and prognostic biomarkers for SKCM patients and identify biomarkers to precisely estimate prognosis.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.