Pan-cancer screening of chemokine receptors identify prognostic value of XCR1
As chemokines and their receptors interaction determines immune cell infiltration in the TME, there is a great probability that chemokine family molecules play an important guiding role in the diagnosis and prognosis of cancers. To identify such molecules, we used the TCGA database and analyzed the relationship of patients’ survival with mRNA expression of each chemokine receptors in several common malignancies (Fig. 1A). Among all the chemokine receptors, only XCR1 was closely related with the hazard ratio in most types of the cancers.
Further cox proportional hazards analysis showed that XCR1 was a low-risk gene in SKCM(skin cutaneous melanoma, p < 0.001), LIHC(liver hepatocellular carcinoma, p < 0.001), HNSC(head and neck squamous cell carcinoma, p < 0.001), LUAD(lung adenocarcinoma, p = 0.003), ESAD (esophageal adenocarcinoma, p = 0.005), KIRC (kidney renal clear cell carcinoma, p = 0.012), BRCA (breast invasive carcinoma, p = 0.019), SARC(sarcoma, p = 0.021), CESC (cervical squamous cell carcinoma and endocervical adenocarcinoma, p = 0.034), and ESCA (esophageal carcinoma, p = 0.044) (Fig. 1B). Furthermore, XCR1 wasn’t a significant high-risk gene in any type of cancer, suggesting the universal protective role of XCR1 in human cancer.
Higher XCR1 expression related with better patient survival
Then we analyzed the relationship between XCR1 expression and the overall survival (OS)/disease specific survival (DSS) in tumor patients (32). Kaplan-Meier survival analysis showed that high-expression of XCR1 tends to be associated with a better prognosis of OS in all these above types of tumors (Fig. 2A). Moreover, analysis of DSS data also revealed the association between low XCR1 expression and poor prognosis in patients with SKCM(HR = 0.54(0.41–0.7), p < 0.001), LIHC(HR = 0.42(0.26–0.67), p < 0.001), HNSC(HR = 0.58(0.41–0.82), p = 0.002), CESC (HR = 0.45(0.26–0.79), p = 0.005) (Fig. 2B).
Then, diagnostic ROC (receiver operating characteristic) analysis was performed to assess the accuracy of prognostic efficacy with XCR1. The AUC value was greater than 0.7 in COAD (Colon Adenocarcinoma, AUC = 0.849), UCEC (uterine corpus endometrial carcinoma, AUC = 0.793), and KIRC (AUC = 0.728). It was also greater than 0.6, less than 0.7 in LUSC (lung squamous cell, AUC = 0.696), LUAD (AUC = 0.671), THCA (thyroid carcinoma, AUC = 0.649), LIHC (AUC = 0.649), HNSC (AUC = 0.638), KICH (kidney chromophobe, AUC = 0.637), BLCA (bladder urothelial carcinoma, AUC = 0.609), and OSCC (oral squamous cell carcinoma, AUC = 0.602). These data revealed a good sensitivity and specificity of XCR1 in the prognosis of survival in pan-cancer (Fig. 3A-K).
XCR1 was decreased in Human Cancer
We then analyzed the expression of XCR1 among different tumors in TCGA database with TIMER2.0 (Fig. 4A). It has been noted that XCR1 was expressed in a relatively low level both in tumor and adjacent normal tissue, according to the low TPM in all types of tumors. However, compared with adjacent normal tissue, an even lower expression of XCR1 was observed in most cancer (Kruskal-Wallis test P < 0.05), such as LIHC, LUAD, LUSC, READ (rectum adenocarcinoma), THCA, UCEC, HNSC (Fig. 4B). Particularly, in KIRC XCR1 expression was higher than it in nearby normal tissues. The change of XCR1 expression level in tumors indicates that it may be a protective factor related to the occurrence and development of tumor.
Low Expression of XCR1 in advanced Cancer Stages
Next, we investigated the clinical features of XCR1 according to the pathologic stages of the patients in the TCGA project. We found that in most common cancer types, XCR1 expression levels was significantly lower in almost all stages, which including COAD, LIHC, LUAD, LUSC and THCA (Fig. 4C). In contrast, its expression was elevated in all stages in KIRC, suggesting a specific effect of XCR1 in KIRC. Furthermore, significant differences were found between different tumor advance stages in all these types of cancer: T stage (the size of the primary tumor), N stage (lymph node metastases), M stage (distal metastatic conditions) (Fig. 5). Among above cancers, we further explored the multifactorial cox regression analysis of XCR1. The results showed that XCR1 was associated with a good survival prognosis (OS/DSS/PFI) and positively correlated with the T stage of clinical staging. However, there was no significant difference in other factors such as gender, age, BMI, etc. (Fig. 5, Table S1). It suggests that the low expression of XCR1 is closely related to the occurrence, poor development and spread of cancer cells.
Mutation Features of XCR1 in Pan-Cancer
In order to further elucidate the mutational characteristics and biological functions of XCR1 in tumor occurrence and progression, we investigated the genetic alteration status of XCR1 in pan-cancer based on the cBioPortal database(29, 33). Missense and truncating are the most important types of XCR1 gene mutations (Fig. 6A). XCR1 mutations are less frequent and have no direct impact on the overall survival of their patients (Fig. 6B). In addition, the missense and truncating mutation of XCR1 were the main two types of genetic alterations: Y14Tfs*33 truncating alteration was the high-frequency mutation, which was detected in 3 cases of STAD (Stomach adenocarcinoma), 1 case of COAD, and 1 case of UCEC; A33T/S alteration was an important site mutation, which was detected in cases of COAD, STAD, OV (Ovarian serous cystadenocarcinoma), and UCEC (Fig. 6C), respectively. Furthermore, the genetic alteration analysis of the alteration frequency with XCR1 was > 4%, and the primary types were deletion and missense, such as DLBC (Lymphoid Neoplasm Diffuse Large B-cell Lymphoma), KIRC, and SKCM etc. (Fig. 6D).
XCR1 Expression Correlation with Immune Infiltration of CD8+ T Cell
Although above results supported prognostic implications of XCR1 in different cancers, its potential role warranted additional investigations. Immune cells play a crucial role in the immune microenvironment and can affect the prognosis of patients with cancer. We then investigated relationship between XCR1 expression and immune cell infiltration levels in diverse cancer types. XCR1 expression level was significantly correlated with immune cells infiltration in most types of cancer, especially CD4+ T cells, CD8+ T cells, neutrophils, macrophage and DCs (Fig. 7A, left). The expression of XCR1 was confirmed to significantly associated with CD8+ T cells and DCs infiltration with XCELL algorithm (Fig. 7A, right). No significant correlation was found between XCR1 and infiltration of NK cells.
We continued to analyze the effect of XCR1 expression on CD8+ T cells infiltration in different tumors and plotted the corresponding scatter plots. However, the expression of XCR1 was mainly positively correlated with CD8+ T cells in most cancers (Fig. 7B). It was concluded that in pan-cancer, XCR1 expression can promote infiltration of DCs and CD8+ T cells which indicate a better survival expectation and potential sensitivity to immunotherapy.
CXCL9-CXCR3 axis were involved in immune infiltration by XCR1
To further study the molecular mechanism of XCR1 in tumorigenesis and development, we decided to screen out XCR1-binding proteins for protein-protein interaction network analysis. Using the STRING online tool, we obtained 10 XCR1-binding proteins, including members which were supported by experimental evidence or prediction (Fig. 8A). The directly acting molecules verified were XCL1, XCL2, CCL3, CCL8, CCL19, CCL24, and CXCL9. And then we also conducted gene co-expression analyses to explore the relationships between XCR1 and associated genes.
To compare the common genes between related genes and interacting genes of XCR1, an intersection analysis of top 100 related genes and top 20 interacting proteins was conducted shown by an interactive Venn diagram. We obtained the only one common molecules, named CXCL9 (Fig. 8B). Then, we analyzed the molecular correlation between XCR1 and CXCL9 in different tumors. It can be found that XCR1 and CXCL9 do have a significant positive correlation in a variety of tumors, including THCA, SKCM, SARC, KIRC, TGCT (Testicular Germ Cell Tumors), BRCA, PCPG (Pheochromocytoma and Paraganglioma) KIRP, OV (Ovarian Serous Cystadenocarcinoma), ACC (Adrenocortical Carcinoma), PAAD (Pancreatic Adenocarcinoma), and LIHC (Fig. 8C). Other related genes like CLEC9A, BATF3 were characteristic genes expressed in XCR1+ DCs(34, 35) or CD8+ T cells(36). Thus, all these data strongly support the conclusion that XCR1 expression significantly correlated with DCs and CD8+ T cell infiltration in the TME.
In addition, we performed Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis and Gene Ontology (GO) enrichment analysis based on SKCM and LIHC (Fig. 8D-E). The KEGG data indicated that XCR1 was involved in tumor development process through the “cytokine-cytokine receptor interaction” and “chemokine signaling pathway”. We also found that the gene set “cell adhesion molecules” was also enriched in LIHC, which also confirms previous reports that the expression of XCR1 is associated with tumor cells metastasis and spread in the later stages of cancers(23). The GO enrichment analysis suggested that these genes were mainly related to “immune response-activating cell surface receptor signaling pathway”, “humoral immune response”, “lymphocyte mediated immunity”, “adaptive immune response based on somatic recombination of immune receptors built from immunoglobulin superfamily domains”, “immunoglobin complex” and “antigen binding”, strongly indicating its key role in tumor immunity in the microenvironment.
XCR1 mainly expressed in DCs and malignant cells
To discriminate the cell types that express XCR1, we analyzed the distribution of XCR1 in different types of cancer. We analyzed the distribution of XCR1 in the immune cells from three different datasets NSCLC-007-02-1A, UCEC-024-01-1A, and NSCLC_GES127465 (Fig. 9A-C). Cell clustering and gene expression profile confirmed that XCR1 was mainly expressed in a subset of DCs and malignant cells. These results may indicate that XCR1 not only plays an important role in recruiting immune cell infiltration in tumor microenvironments but may also relates to the biological functions of tumor cells.