LPAR2 correlated with different prognosis and immune cell infiltration in head and neck squamous cell carcinoma and kidney renal clear cell carcinoma

Lysophosphatidic acid (LPA) and its receptors play a key role in regulating cancer progression. Upregulation of LPA receptor 2 (LPAR2) plays a role in carcinogenesis; however, the exact role of LPAR2 in tumors remains elusive. This study aims to explore the correlation between LPAR2 expression with tumor prognosis and immune infiltration in pan-cancers. The expression of LPAR2 in pan-cancers was analyzed using the Online Cancer Microarray Database (Oncomine), Tumor Immune Estimation Resource (TIMER), and UALCAN databases. The effects of LPAR2 on the clinical prognosis in pan-cancer were examined using the Kaplan–Meier plotter (KM plotter) as well as Gene Expression Profiling Interactive Analysis (GEPIA), UALCAN, and Human Protein Atlas (HPA) databases. Moreover, the R software program was applied for validation of expression and prognostic value of LPAR2 in tumor patients in the Cancer Genome Atlas (TCGA) dataset and the Gene Expression Omnibus (GEO) database. The relationship between the expression level of LPAR2 and the clinical and molecular criteria of head and neck squamous cell carcinoma (HNSC) and kidney renal clear cell carcinoma (KIRC) was analyzed using UALCAN, whereas the relationship between LPAR2 expression and prognosis in patients with HNSC and KIRC with different clinical characteristics was examined using the KM plotter. Furthermore, the correlation between LPAR2 expression and tumor immune infiltration was examined using TIMER. The correlation between LPAR2 expression and gene markers of tumor immune infiltrates was analyzed using TIMER and GEPIA. In addition, the cBioPortal for Cancer Genomics was used to calculate the mutations, methylations, and altered neighbor genes of LPAR2. The expression of LPAR2 was significantly correlated with the outcome of multiple types of cancer, especially HNSC and KIRC. Furthermore, high expression of LPAR2 was significantly associated with various immune markers in the immune cell subsets of HNSC and KIRC. High expression of LPAR2 plays significantly different prognostic roles in HNSC and KIRC possibly owing to its association with different immune markers. LPAR2 is correlated with tumor immune cell infiltration and is a valuable prognostic biomarker for HNSC and KIRC. However, further experiments are required to validate these findings.

In this study, we systematically investigated the expression of LPAR2 and its relationship with pan-cancer prognosis using the Oncomine, TIMER, UALCAN, GEPIA, KM plotter and HPA databases, as well as expression and survival analysis of LPAR2 in the TCGA and GEO data was validated by R software. Furthermore, we examined the relationship between LPAR2 expression and the clinical and molecular criteria of HNSC and KIRC using UALCAN. Subsequently, we investigated the relationship between LPAR2 expression and the prognosis of patients with HNSC and KIRC with different clinical characteristics using the KM plotter. In addition, we analyzed the correlation between LPAR2 and tumor-infiltrating immune cells in the microenvironment of pan-cancer using TIMER and GEPIA. Lastly, we used the cBioPortal for Cancer Genomics online tool to analyze the alterations, mutations, methylations, and pathways of LPAR2. Therefore, in this study, we demonstrated a potential mechanism of action of LPAR2, examined the prognostic role of LPAR2 in HNSC and KIRC, and identified LPAR2 as a key factor in regulating the immune microenvironment of HNSC and KIRC. The overall design and workflow of this study is presented in Fig. 1.

Assessment of LPAR2 expression in different cancers and normal tissues
On analyzing the mRNA expression levels of LPAR2 in pan-cancer and normal tissues using Oncomine, we found that LPAR2 expression was higher in bladder, brain and central nervous system (CNS), breast, colorectal, kidney, and lung cancers and lymphoma than in normal control tissues ( Fig. 2A). However, LPAR2 expression was lower in kidney cancer, leukemia, lung cancer, lymphoma and sarcoma tissues than in normal control tissues ( Fig. 2A). Table 1 summarizes the detailed findings of specific tumor types. Furthermore, we assessed differences in LPAR2 expression in pan-cancer using the TIMER databases and found that LPAR2 expression was significantly higher in bladder urothelial carcinoma (BLCA), breast invasive carcinoma (BRCA), cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), HNSC, KIRC, liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), prostate adenocarcinoma (PRAD), rectum adenocarcinoma (READ), stomach Keywords: Head and neck squamous cell carcinoma, Kidney renal clear cell carcinoma, Prognosis, LPAR2, Tumor immune infiltration Fig. 1 Analysis workflow of this research adenocarcinoma (STAD), and uterine corpus endometrial carcinoma (UCEC) than in the adjacent normal tissues (Fig. 2B). However, LPAR2 expression was significantly lower in kidney chromophobe (KICH) and thyroid carcinoma (THCA) than in the adjacent normal tissues (Fig. 2B). Subsequently, we examined LPAR2 expression using UALCAN and found that the mRNA expression levels of LPAR2 were significantly higher in BLCA, BRCA, cervical squamous cell carcinoma and endocervical adenocarcinoma (CECS), glioblastoma multiforme (GBM), HNSC, KIRC, kidney renal papillary cell carcinoma (KIRP), LIHC, LUAD, LUSC, PRAD, READ, STAD, and UCEC than in normal control tissues (Figs. 2C, 3). However, LPAR2 expression was significantly lower in KICH and THCA than in normal control tissues (Fig. 3). Differences in LPAR2 expression between tumors and normal adjacent tissue samples are demonstrated in Fig. 1C. To validate these results, we used R software to analyze expression of LPAR2 in pancancers via the TCGA databases (Fig. 4A). As a result, we observed the same trend in the expression of LPAR2 in pan-cancers (Fig. 4A).

Relationship between LPAR2 expression and prognosis in patients with cancer
We used KM plotter to determine the correlation between LPAR2 expression and the survival of patients with pan-cancer and those with normal tissues (Figure S1). A significant correlation was observed between LPAR2 expression and prognosis in BLCA, BRCA, CESC, HNSC, KIRC, STAD, THYM, and UCEC ( Fig. 4B (Fig. 4B, F, G, L, M, N). However, no significant correlation was observed between the mRNA expression levels of LPAR2 and the prognosis of other cancers ( Figure S1). Furthermore, we assessed the relationship between LPAR2 expression and the prognosis of multiple cancers using GEPIA ( Figure S2) and found that high mRNA expression of LPAR2 was associated with a worse prognosis in patients with KIRC (OS, HR = 2.1, P = 3.6e − 06; disease-free survival [DFS], HR = 1.9, P = 9e − 04), PRAD (OS, HR = 7.7, P = 0.024), and CHOL (DFS, HR = 2.6, P = 0.048) (Fig. 5A, C-E). Furthermore, high mRNA expression of LPAR2 was correlated with better OS in patients with HNSC (HR = 0.71, P = 0.012) and THYM (HR = 0.11, P = 0.013) (Fig. 5B, F). However, it was not significantly correlated with OS and DFS in patients with BRCA (OS, HR = 0.85, P = 0.49; DFS, HR = 0.74, P = 0.29) and other tumors ( Figure S2). In UALCAN databases, we found that higher expression of LPAR2 was associated with improved prognosis in patients with ACC (P = 0.00055), as well as with worse prognosis with KIRC (P < 0.0001) (Fig. 5G, I). Upregulation of LPAR2 might be correlated with good prognosis in HNSC patients, but this correlation was not statistically significant (Fig. 5H). Nevertheless, in UACLAN, no significant correlation was observed between LPAR2 expression and the prognosis of other cancers ( Figure S3).
Likewise, to validate these results, survival analysis of LPAR2 in pan-cancers of the TCGA databases was performed using the survival package via R software ( Figure  S4). Our results indicated that high expression level of LPAR2 was significantly associated with a worse OS in patients with ACC (HR = 2.   Taken together, the combination of OS, RFS, DFS and DMFS, and concern of bias, our findings illustrated the expression levels and prognostic value of LPAR2 in several types of cancers, especially HNSC and KIRC, suggesting that high LPAR2 expression plays significantly different prognostic roles in HNSC and KIRC. Thus, we performed LPAR2 expression analyses and survival analyses in HNSC and KIRC using GEO databases in the end. Results of differential expression analysis showed that LPAR2 expression was significantly higher in HNSC and KIRC than in normal control tissues in GSE30784, GSE31056, GSE53757 and GSE15641(P < 0.01) ( Fig. 6A-E). However, survival analysis of GSE686, GSE65858, GSE167573 and GSE22541 showed that no further significant correlations were found between LPAR2 expression and the prognosis of HNSC and KIRC (P > 0.05) ( Figure  S5A-D). These inconsistencies might be due to limited sample sizes of HNSC and KIRC in GEO databases and differences in data collection methods as well as biases in methods of adjustment. Therefore, much further experimental validation is needed to investigate the link between the expression of LPAR2 and prognosis in cancer patients with HNSC and KIRC as well as other kinds of cancers.

Relationship between protein expression of LPAR2 and prognosis in patients with HNSC and KIRC
After analyzing the mRNA expression of LPAR2 and its relationship with the prognosis of patients with HNSC and KIRC, we investigated the protein expression of LPAR2 and its correlation with the prognosis of patients with HNSC and KIRC using the HPA database. As demonstrated in Fig. 6 F-I, the protein expression of LPAR2 was moderate in HNSC and KIRC tissues and low in the corresponding normal tissues. Relevant clinical data was shown in Table S1. Furthermore, according to the data obtained from the HPA, the relationship between the protein expression of LPAR2 and prognosis was similar to that between the mRNA expression of LPAR2 and prognosis. Moreover, high protein expression of LPAR2 was associated with worse OS in patients with KIRC (P = 3.5e-9) but with improved OS in patients with HNSC (P = 0.0023) ( Fig. 6 J-K). The related clinical data were exhibited in Table S2 and Table S3.

Relationship between mRNA expression of LPAR2 and clinical characteristics of patients with HNSC and KIRC
Given that LPAR2 expression plays significantly different prognostic roles in HNSC and KIRC, we used UAL-CAN and TCGA to examine the relationship between LPAR2 expression and the clinical characteristics of patients with HNSC and KIRC. For the criterion of tumor stage, we found that LPAR2 expression was significantly higher in patients with stage 1-4 HNSC than in patients in the control group (P < 0.001) (Fig. 7D). For the criterion of race, the mRNA expression of LPAR2 was higher in the Caucasian and African-American patients with HNSC than in patients in the control group (P < 0.001); however, there was no significant difference in LPAR2 expression between the Asian patients with HNSC and those in the control group (P > 0.05) (Fig. 7C). In addition, LPAR2 expression was upregulated in both men and women with HNSC (P < 0.001) (Fig. 7B) in the age groups of 21-40 years (P < 0.001), 41-60 years (P < 0.001), 61-80 years, and 81-100 years (P < 0.001) (Fig. 7A). These findings suggested that the mRNA expression of LPAR2 was significantly higher in patients with HNSC than in patients in the control group (P < 0.01 and P < 0.001, respectively), irrespective of tumor grade, HPV expression status, nodal metastasis status, and mutation status ( Fig. 7E, F, G, H.).
In patients with KIRC, LPAR2 expression was upregulated in patients with tumor stages 3 and 4 (P < 0.001) (Fig. 8D). However, there was no significant difference in LPAR2 expression between patients with tumor stages 1-2 and those in the control group (P > 0.05) (Fig. 8D). Similar to HNSC, LPAR2 expression was significantly higher in the Caucasian and African-American patients with KIRC than in patients in the control group (P < 0.001); whereas there was no significant difference in LPAR2 expression between the Asian patients with KIRC and those in the control group (P > 0.05) (Fig. 8C). In addition, LPAR2 expression was upregulated in both men and women with KIRC (P < 0.001) (Fig. 8B). Meanwhile, we found that LPAR2 expression was upregulated in patients with KIRC in the age groups of 21-40 years (P < 0.05), 41-60 years (P < 0.01), and 61-80 years (P < 0.001) but not in the age group of 81-100 years (P > 0.05) (Fig. 8A). Our findings also suggested that the mRNA expression of LPAR2 was higher in patients with grade 3-4 KIRC than in patients in the control group (P < 0.001); nonetheless there was no significant difference between the mRNA expression of LPAR2 in patients with grade 1-2 KIRC and those in the control group (P > 0.05) (Fig. 8E). Furthermore, the mRNA expression of LPAR2 was higher in patients with node-positive KIRC than in patients with node-negative KIRC; however, it was higher in both node-positive and node-negative patients than in patients in the control group (P < 0.01 and P < 0.001, respectively) (Fig. 8F). These findings suggested that LPAR2 expression was associated with tumor stage, tumor grade, and lymph node metastasis in patients with KIRC, and with race in patients with HNSC and KIRC.  (Fig. 10N-O).

Relationship between mRNA expression of LPAR2 and prognosis in patients with HNSC and KIRC with different clinical characteristics
These results suggested that LPAR2 expression influenced the prognosis of patients with HNSC of high stage and grade. Upregulated expression of LPAR2 was beneficial to men with HNSC or patients with low LPAR2 mutation burden and was significantly associated with prognosis in White patients with HNSC and KIRC.

Association between LPAR2 expression and immune cell infiltration in HNSC and KIRC
Tumor-infiltrating lymphocytes are independent predictors of tumor stage, grade, and lymph node status in cancers [25,26]. Therefore, we used the TIMER database to analyze the relationship between LPAR2 expression and the degree of immune cell infiltration in HNSC and KIRC (Fig. 11) and found that LPAR2 expression was significantly correlated with tumor purity (R = 0.2, P = 7.74e-06), B cell infiltration (R = 0.217, P = 1.70e-05), and CD4 + T cell infiltration (R = 0.149, P = 1.07e-03) but not with the infiltration of CD8 + T cells, macrophages, neutrophils, and DCs in patients with HNSC (Fig. 11A). In patients with KIRC, LPAR2 expression was significantly correlated with tumor purity (R = -0.155, P = 8.49e-04), B cell infiltration (R = 0.168, P = 2.94e-04), CD4 + T cell infiltration (R = 0.242 P = 1.46e-07), neutrophil infiltration (R = 0.197, P = 2.09e-05), and DC infiltration (R = 0.141, P = 2.66e-03) (Fig. 11A) but not with the infiltration of CD8 + T cells and macrophages (Fig. 11A). We further analyzed the correlation between LPAR2 expressions and immune cell infiltration in patients with HNSC and KIRC by generating KM plots using the TIMER database. The results demonstrated that B-cell infiltration was significantly correlated with the prognosis of HNSC (P = 0.045) (Fig. 11B), and a significant correlation was observed between the mRNA expression of LPAR2 and prognosis in patients with KIRC (P < 0.001) (Fig. 11B). These results suggest that LPAR2 is important for regulating immune cell infiltration in HNSC and KIRC. Moreover, LPAR2 is more important for regulating tumor purity and the infiltration of B cells and CD4 + T

Relationship between LPAR2 and immune marker expression
Given that LPAR2 is important for regulating immune cell infiltration in HNSC and KIRC, we assessed the relationship between LPAR2 expression and immune cell infiltration based on the immunological markers of HNSC and KIRC using the TIMER and GEPIA databases. In addition, we evaluated the relationship between LPAR2 expression and several immunological marker subsets, including total T cells, B cells, CD8 + T cells, tumor-associated macrophages (TAMs), monocytes, M1 and M2 macrophages, natural killer (NK) cells, neutrophils, DCs, T follicular helper (Tfh) cells, type 1 T helper  Table 2). The results suggested that LPAR2 expression exhibited a significant correlation with the levels of most markers of B cells, M1 macrophages, Th2 cells, and Tfh cells in patients with HNSC (P < 0.0001, Table 2). Strikingly, in patients with HNSC, LPAR2 expression was closely associated with INOS of M1 macrophages, STAT5A of Th2 cells, and BCL6 of Tfh cells (P < 0.0001, Cor > 0.2, Table 2). In patients with KIRC, the mRNA expression of LPAR2 was closely correlated with the levels of most markers of total CD8 + T cells (CD8A and CD8B), T cells (CD3D, CD3E, and CD2), B cells (CD19 and CD79A), monocytes (CD86 and CD115), TAMs (CD68), M1 macrophages (IRF5), M2 macrophages (VSIG4), neutrophils (CD11b and CCR7), DCs, Th1 cells (STAT4, IFN-γ, TNF-α), Th2 cells (STAT5A), Tfh cells (BCL6), Tregs (FOXP3, CCR8, and TGF-β), and exhausted T cells (PD-1, CTLA4, and LAG3) (P < 0.0001, Cor > 0.2, Table 2). Furthermore, we assessed the relationship between the expression of LPAR2 and that of the aforementioned markers using GEPIA. The correlation between LPAR2 expression and these markers was   (Table 3). These findings suggested that LPAR2 was significantly correlated with infiltrating immune cells in HNSC and KIRC and played a significant role in the immune microenvironment of HNSC and KIRC.

Alterations, mutations, methylations, and frequently altered neighbor genes of LPAR2 in patients with HNSC and KIRC
We analyzed genetic alterations of LPAR2 using the cBi-oPortal for Cancer Genomics in the HNSC and KIRC (TCGA, Firehose Legacy) datasets. LPAR2 mutations and amplifications were found in 3 of 528 patients with HNSC but not in 537 patients with KIRC ( Fig. 12A-B).

Discussion
LPA, a growth factor-like phospholipid, is abundantly found in human tissues and fluids [22]. It participates in various biological functions, such as cell migration, cell proliferation, inflammation, angiogenesis, and survival [27,28]. LPA acts through G-protein-coupled LPA receptors, which are called LPARs [6,8]. LPAR2 belongs to the EDG family and contains 351 amino acids [22,29]. It is unique in the proximal region of the C-terminus and contains several putative palmitoylated cysteine residues and a dileucine motif [30]. A few studies have suggested that LPAR2 is associated with several cancers, such as breast [16,31,32], colon [20], ovarian [33], and stomach cancers [17]. These studies have reported that LPAR2 expression is important in cancer biology and may promote gene transcription and cell proliferation in the tumor microenvironment [17,34,35]. However, the mechanism of action of LPAR2 in tumors remains unclear.
In addition to traditional cancer treatment, cancer immunotherapy has emerged as an important therapy owing to its adequate efficacy and fewer side effects [36]. Nevertheless, immunotherapy has not been extensively investigated and effectively used to treat patients with HNSC and KIRC [37]. Given that immunotherapy mainly targets the tumor immune microenvironment, we analyzed the effects of LPAR2 on tumor prognosis and immune infiltration of HNSC and KIRC in this study.
We examined the mRNA and protein expression levels of LPAR2 in pan-cancer and the corresponding normal tissues using Oncomine, TIMER, UALCAN, and HPA databases, as well as validated by R software in TCGA and GEO databases. LPAR2 expression was evaluated in tumor and normal tissues in multiple cancer types (Figs. 2 and 4, Table 1). Differences in data collection methods and analytical approaches may  be attributed to the heterogeneity of LPAR2 expression among cancer types and databases. However, we consistently observed higher expression of LPAR2 in HNSC and KIRC across these databases. We used online tools, such as KM plotter, GEPIA2.0, UACLAN and HPA, and R software to examine the critical role of LPAR2 in predicting patient outcomes of multiple cancer types in TCGA and GEO databases. Our findings illustrated the expression levels and prognostic value of LPAR2 in several types of cancers, especially HNSC and KIRC ( Figures S1, 2, 3, 4). High LPAR2 expression was significantly correlated with a worse prognosis in KIRC. However, high LPAR2 expression was strongly correlated with improved prognosis in HNSC. These contradictory results suggested that LPAR2 acts as a tumor suppressor gene in HNSC and an oncogene in KIRC.
Given that LPAR2 expression plays significantly different prognostic roles in HNSC and KIRC, we used UALCAN and KM plotter to examine the relationship between the mRNA expression of LPAR2 and prognosis in patients with HNSC and KIRC with different clinical characteristics. The findings suggested that high LPAR2 expression was associated with advanced tumor stages, high tumor grades, and lymph node metastasis in patients with KIRC. Using KM plotter, we found that high LPAR2 expression was associated with improved prognosis in patients with HNSC with advanced tumor stages and high tumor grades. Meanwhile, high LPAR2 expression resulted in better prognosis in patients with Table 3 Correlation analysis between LPAR2 and relate genes and immune markers in GEPIA Cor, R value of Spearman's correlation; * P < 0.05; ** P < 0.01; *** P < 0.001  HNSC, which may be related to their mutational burden status. These results means that LPAR2 was involved in tumor development and progression of patients with HNSC or KIRC. Given that high LPAR2 expression affects prognosis related to clinical characteristics in HNSC and KIRC patients, we assessed the relationship between LPAR2 expression and the degree of immune cell infiltration using the TIMER database. Another important finding of this study was that LPAR2 expression was significantly associated with the infiltration of diverse immune cells in HNSC and KIRC. We found that LPAR2 expression had a positive correlation with tumor purity in HNSC and KIRC, the infiltration of B cells and CD4 + T cells in HNSC (Fig. 11A), and the infiltration of B cells, CD4 + T cells, neutrophils, and DCs in KIRC (Fig. 11A). These results suggest that LPAR2 is important for regulating immune cell infiltration in HNSC and KIRC, with particularly strong effects on tumor purity and infiltrating B cells, CD4 + T cells, neutrophils, and DCs.
Furthermore, to investigate the role of LPAR2 in the regulation of tumor immunology in HNSC and KIRC, we analyzed the relationship between LPAR2 expression and marker genes of immune cells. We found a significant positive correlation between LPAR2 expression and the markers of B cells (CD19 and CD79A), M1 macrophages (INOS and IRF5), neutrophils (CD11b), Th2 cells (STAT6 and STAT5A), Tfh cells (BCL6), and exhausted T cells (CTLA4) in HNSC (P < 0.01, Table 2). In addition, LPAR2 expression was strongly correlated with INOS of M1 macrophages, STAT5A of Th2 cells, and BCL6 of Tfh cells (P < 0.0001, Cor > 0.2, Table 2). These results indicate that LPAR2 promotes the polarization of macrophages to the M1 phenotype and regulates T cell responses. Furthermore, BCL6 recognizes DNA target sequences similar to those recognized by STAT5 [38]. Some studies have found that STAT5A inhibits cell invasion and metastasis in breast cancer [39]. LPAR2 may play a role in HNSC by interacting with STAT5A and BCL6 via the prolactin-JAK2-STAT5A pathway [38]; but further studies are warranted. In this study, LPAR2 expression was significantly correlated with most immune markers in KIRC, including CD3D and CD3E of total T cells; CD19 and CD79A of B cells; IRF5 of M1 macrophages; STAT5A of Th2 cells; FOXP3 and CCR8 of Treg cells; and PD-1, CTLA4, and LAG3 of exhausted T cells (P < 0.0001, Cor > 0.3, Table 2). In addition, the results indicate that LPAR2 activates Tregs and B cells, induces T cell exhaustion, and promotes Treg responses to suppress T cell-mediated immunity, thereby regulating T cell responses in KIRC. LPAR2 may promote the polarization of macrophages to the M1 phenotype via IRF5. Therefore, these findings collectively suggest that LPAR2 is a crucial factor for the recruitment and regulation of infiltrating immune cells in HNSC and KIRC.

Conclusion
LPAR2 plays significantly different prognostic roles in HNSC and KIRC might owing to its association with different immune markers. LPAR2 is important for governing immune cell infiltration, and is a valuable prognostic biomarker that may guide treatment in HNSC and KIRC. Nevertheless, further validation experiments are required.

Oncomine database analysis
The expression data of 715 genes were obtained from 86,733 samples and the mRNA expression levels of LPAR2 in pan-cancer were analyzed using the online cancer microarray database (Oncomine) (www. oncom ine. org). The Student's t-test was used to compare the mRNA expression of LPAR2 between normal and cancer samples. P-value was used to characterize significant differences. The fold change was 1.5, and the cut-off P-value was 0.0001.

TIMER database analysis
The Tumor Immune Estimation Resource (TIMER) (https:// cistr ome. shiny apps. io/ timer/) database comprises six tumor-infiltrating immune cell subsets [44], and the expression levels of six subsets are pre-calculated for 10,897 tumors across 32 cancer types from The Cancer Genome Atlas (TCGA). The database allows the analysis of gene expression and tumor immune infiltration (B cells, CD4 + T cells, CD8 + T cells, neutrophils, macrophages, and dendritic cells [DCs]) in various cancer types. In this study, TIMER was used to analyze the mRNA expression of LPAR2 in various cancer types and investigate the relationship between LPAR2 expression and the degree of infiltration of specific immune cell subsets. Furthermore, differences in the survival of patients with cancer based on gene expression or immune cell infiltration were examined using KM survival analysis. Lastly, the correlation between the expression of LPAR2 and that of specific immune markers was examined.  [45]. In this study, UALCAN was used to examine the mRNA expression level of LPAR2 in different cancer and normal samples using the TCGA data and investigate the relationship between LPAR2 expression and different clinical characteristics. In addition, the prognostic value of LPAR2 in pan-cancer and the relationship between LPAR2 expression and the prognosis of patients with different clinical characteristics were analysed.

KM plotter analysis
The KM plotter (http:// kmplot. com/ analy sis/) is an online database, which contains microarray gene expression data and survival information derived from the European Genome-Phenome Archive, Gene Expression Omnibus (GEO), and TCGA. It is used to assess the influence of multiple genes on the survival rate in 21 cancer types in a large number of samples [46]. In this study, the KM plotter was used to analyze the prognostic value of LPAR2 in pan-cancer and investigate the relationship between LPAR2 expression and the prognosis of patients with different clinical characteristics.

GEPIA2 database analysis
GEPIA (http:// gepia. cancer-pku. cn/ index. html) uses standard processing pipelines to analyze the RNAsequencing expression data of 8,587 normal samples and 9,736 tumors from the GTEx and TCGA projects [47]. GEPIA2 (http:// gepia2. cancer-pku. cn/# index) is an updated version of GEPIA [48]. In this study, GEPIA2 was used to examine the relationship between the mRNA expression of LPAR2 and pan-cancer prognosis as well as the relationship between the expression of LPAR2 and that of the markers of immune cell infiltration.

HPA database
The Human Protein Atlas (HPA) database (www. prote inatl as. org) was used to analyze the protein expression of LPAR2 in HNSC, KIRC, and normal tissues [49,50]. HPA provides access to the protein expression profiles of 32 human tissues and uses antibody profiling to accurately assess protein localization. In addition, it provides the measurements of RNA levels. In this study, HPA was used to visualize the representative immunohistochemical images of LPAR2 in HNSC, KIRC, and their corresponding normal tissues. In addition, the relationship between the protein expression level of LPAR2 and the prognosis of patients with HNSC and KIRC was examined.

TCGA and cBioPortal for cancer genomics
The cBioPortal for Cancer Genomics tool (http:// www. cbiop ortal. org) is used to analyze, visualize, and download cancer genomics datasets [51]. In this study, the cBioPortal for Cancer Genomics was used to download the HNSC and KIRC (TCGA, Firehose Legacy) datasets for LPAR2 analysis, which contained histopathological data of 528 patients with HNSC and 537 patients with KIRC. The genomic profiles were evaluated via the Genomic Identification of Significant Targets in Cancer (GISTIC) analysis and included the assessment of mutations, methylations, mRNA expression z-scores (RNA Seq V2 RSEM), protein expression z-scores (RPPA), and putative copy number alterations (CNAs). Co-expression was evaluated according to the instructions provided on cBioPortal.