Pan-cancer analysis of C3AR1 mRNA expression levels
First, we compared the difference of C3AR1 between ovarian cancer and other human tumors and normal tissues. Increased expression of C3AR1 is higher in esophageal carcinoma (ESCA), glioblastoma multiforme (GBM), kidney renal clear cell carcinoma (KIRC), kidney renal papillary cell carcinoma (KIRP), acute myeloid leukemia (LAML), brain lower grade glioma (LGG), ovarian serous cystadenocarcinoma (OC), pancreatic adenocarcinoma (PAAD), skin cutaneous melanoma (SKCM), stomach adenocarcinoma (STAD), and testicular germ cell tumors (TGCT) than normal tissues. However, C3AR1 mRNA expression levels in adrenocortical carcinoma (ACC), lymphoid neoplasm diffuse large B-cell lymphoma (DLBC), lung squamous cell carcinoma (LUSC), and thymoma (THYM) are lower than normal tissues (Fig. 1A).
Increased C3AR1 is associated with poor prognosis of OC patients
We investigated mRNA and protein expression of C3AR1 in the OC and normal cohorts in the GEPIA, CPTAC and HPA datasets. Results showed that C3AR1 mRNA in OC patients was about twice that of the normal group (Fig. 1B). In addition, compared with non-metastatic cells ES-2 and OV56, metastatic cells Hey-A8, ONCO-DG-1, and 59M have higher C3AR1 mRNA levels (Figure S1). To further verify the results, C3AR1 protein levels in OC and normal samples were analyzed. As shown in Fig. 1C and 1D, C3AR1 protein levels in tumor sample tissues were higher than that in normal tissues, which were mainly distributed in the membrane and cytoplasm of cancer cells. Immunohistochemical results from our cohort showed that C3AR1 expression was higher in the ovarian cancer group than in the normal group (Figure S1A).In the TCGA OC HumanMethylation450K cohort, transcription levels of C3AR1 in OC patients were negatively correlated with its DNA methylation levels (cg09238677, r=-0.186, p < 0.05; cg23205183, r=-0.112, p < 0.05.) (Fig. 1E, F). We further evaluated the prognostic potential of C3AR1 in OC. The overall survival (HR = 1.26, 95%CI (1.09–1.46), P = 0.0021), disease-free survival (HR = 1.35, 95%CI (1.18–1.55), P = 1.6E-05), post progression survival (HR = 1.29, 95%CI (1.07–1.56), P = 0.0081) (Fig. 1G-I) were worse in OC patients with higher C3AR1 expression levels (Fig. 1G-I).
Increased C3AR1 correlated with malignant clinicopathological features and poor prognosis of patients with OC
Relationships between C3AR1 and clinicopathological features were analyzed in 557 OC patients in the TCGA. Results showed that elevated C3AR1 was significantly related to multiple malignant characteristics, including tumor grade, stage, days to new tumor event, lymph node metastasis, cancer status, and survival status (Table 1). However, little differences of C3AR1 were observrd in different ages and anatomic neoplasm subdivision. Crucially, C3AR1 is positively correlated with 18 tumor metastasis related genes including FN1, SNAI2, and ZEB2 (Figure S2).
Table S1
Correlations between C3AR1 and clinicopathological parameters of ovarian cancer patients in the TCGA (n = 557).
Catagory | | H group | L group | P-value |
Age | < 60 | 152 | 148 | 0.76 |
| > 60 | 127 | 130 | |
Anatomic neoplasm subdivision | single | 68 | 76 | 0.35 |
| bilateral | 197 | 183 | |
Stage | III-IV | 265 | 244 | 0.0004 |
| I-II | 11 | 34 | |
Days to new tumor event | > 1 year | 97 | 132 | < 0.0001 |
| < 1 year | 76 | 39 | |
Lymphatic invasion | Y | 91 | 44 | 0.0013 |
| N | 35 | 43 | |
Neoplasm histologic grade | G3 | 238 | 211 | 0.02 |
| G2 | 33 | 50 | |
New neoplasm event | recurrence | 142 | 123 | 0.09 |
| progression | 9 | 16 | |
Cancer status | tumor free | 44 | 71 | 0.0043 |
| with tumor | 201 | 175 | |
Vital status | deceased | 123 | 105 | 0.0047 |
| living | 155 | 172 | |
Additional radiation therapy | Y | 22 | 8 | 0.03 |
| N | 139 | 125 | |
H high C3AR1 expression; L low C3AR1 expression; Y yes; N no.
Screening and functional enrichment analysis of C3AR1 co-expressed genes in OC
co-expressed genes of C3AR1 were analyzed by R package and visualized through the ggplot package. 11967 co-expressed genes were identified (8123 positive correlations, 3844 negative correlations, P < 0.05)). Top 9 genes with the strongest correlation with C3AR1 are presented (Fig. 2A-I), including LAPTM5 (cor = 0.96, P = 1.49E-212), CD53 (cor = 0.96, P = 6.88E-212), CYBB (cor = 0.96, P = 2.04E-210), LAIR1 (cor = 0.96, P = 6.81E-200), MS4A6A (cor = 0.95, P = 8.03E-190), HAVCR2 (cor = 0.95, P = 7.12 E-185), CD86 (cor = 0.94, P = 8.6E-182), FCCR2A (cor = 0.94, P = 3.62E-178) and FCCR2C (cor = 0.94, P = 3.62E-178). Co-expressed genes with coefficient |cor|> 0.6, P < 0.05 were analyzed by GO (gene ontology) and KEGG pathway enrichment analysis and enriched in 945 biological processes (GO-BP), 80 cell components (GO-CC), 68 molecular functions (GO-MF) and 61 KEGG signaling pathways. GO analysis showed that these genes are mainly involved in T cell, leukocyte and lymphocyte activation and cytokine and chemokine activity (Fig. 2K-M). KEGG pathway analysis showed that co-expressed genes are mainly involved in cytokine-cytokine receptor interaction, tuberculosis, and chemokine signaling pathways (Fig. 2J).
Identification and functional enrichment analysis of differentially expressed genes in patients with different C3AR1 transcription levels in OC
To further characterize the molecular function of C3AR1, patients with ovarian cancer in TCGA were devided into high and low C3AR1 groups. 415 DEGs (402 up-regulated and 13 down-regulated) were presented in the volcano plot (Fig. 3A). Distinct different gene expression patterns between the two groups were visualized in the cluster plot (Fig. 3B). KEGG analysis was performed regarding DEGs, and 69 up-regulated and 2 down-regulated pathways were found, among which cytokine-cytokine receptor interaction (FDR = 4.17E-24, P = 4.18E-24), the hematopoietic cell lineage (FDR = 1.08E-23, P = 4.18E-24) and the chemokine signaling pathway (FDR = 9.25E-16, P = 3.94E-17) are the most significantly enriched (Fig. 3C-D). Gene ontology analysis revealed 1135 up-regulated biological processes and 156 down-regulated biological processes, among which T cell activation (FDR = 2.49E-41, P = 7.02E-45) and leukocyte cell-cell adhesion (FDR = 8.32E-44, P = 1.47E-40) and leukocyte proliferation (FDR = 5.18E-39, P = 6.12E-36) are the most significantly enriched (Fig. 3E-F). We also applied the ESTIMATE algorithm to calculate stromal score, immune score, and ESTIMATE score. The higher C3AR1 levels is associated with higher the immune score, stromal score and ESTIMATE score (Figure S3). The enrichment analysis of DEGs and co-expressed genes of C3AR1 are highly consistent, indicating that C3AR1 may be related to tumor immune cells or immune regulation.
Correlation analysis between C3AR1 and tumor immune cell infiltration
Correlation between C3AR1 expression in OC and immune cell infiltration were evaluated and found that C3AR1 is positively correlated with the six types of immune cells. According to correlation index, they are neotrophil (r = 0.647, P = 2.83E-58), dendritic cell (r = 0.606, P = 2.23E-49), macrophage (r = 0.503, P = 4.36E-32), CD8 + T cells (r = 0.453, P = 1.31E-32), CD4 + T cells (r = 0.327, P = 1.94E-13) cells and B cells (r = 0.266, P = 3.13E-09) (Fig. 4A). MCP counter analysis showed that patients have higher C3AR1 tend to have more tumor immune cell infiltration (Figure S4), including T cells, cytotoxic lymphocytes, B lineage, NK cells, monocytic lineage, myeloid dendritic cells, neutrophils, endothelial cells and fibroblasts (P < 0.01).
We further evaluated the association between C3AR1 and various immune cell markers in OC cohorts of TIMER, GEPIA and TCGA database (Table 2). Results show that C3AR1 has the strongest correlation with M2 macrophages (CD163, MRC1, CD209), tumor-associated macrophages (CCL2, CD86, CD68) and monocyte immune markers, including (CD14, CD33, ITGAX) (Fig. 4B-D). In addition, we evaluated the interaction between C3AR1 and 22 immune checkpoint genes. As shown in Fig. 4E, the mRNA of 8 immune checkpoints (CD274, CTLA4, HAVCR2, LAG3, PDCD1, PDCD1LG2, TIGIT, SIGIEC15) increased significantly in the C3AR1 high group. Therefore, these evidences indicate that C3AR1 plays an important role in immune tolerance and immune escape of OC cells.
Table 2
Correlation of C3AR1 with immune cell biomarkers in TIMER, GEPIA, and TCGA databases.
Description | Gene markers | TIMER | | GEPIA | | TCGA | |
| | Cor | p | Cor | p | Cor | p |
B cell | CD19 | 0.10 | 9.79E-02 | -0.02 | 0.68 | 0.07 | 1.98E-01 |
| MS4A1 | 0.55 | 5.07E-25 | 0.19 | 8.70E-05 | 0.35 | 4.91E-12 |
| CD79A | 0.42 | 2.13E-14 | 0.22 | 4.00E-06 | 0.37 | 7.88E-14 |
CD8 + T Cell | CD8A | 0.63 | 0.00E + 00 | 0.50 | 0.00E + 00 | 0.63 | 3.70E-42 |
| CD8B | 0.48 | 0.00E + 00 | -0.01 | 0.98 | 0.33 | 7.64E-11 |
| IL2RA | 0.76 | 0.00E + 00 | 0.68 | 0.00E + 00 | 0.77 | 3.41E-74 |
Tfh | CXCR3 | 0.69 | 0.00E + 00 | 0.51 | 0.00E + 00 | 0.70 | 1.33E-56 |
| CXCR5 | 0.42 | 1.83E-14 | -0.06 | 0.22 | 0.27 | 9.07E-08 |
| ICOS | 0.65 | 4.04E-38 | 0.70 | 6.40E-67 | 0.64 | 5.78E-45 |
Th1 | IL12RB1 | 0.83 | 2.03E-79 | 0.85 | 7.40E-122 | 0.79 | 3.67E-80 |
| CCR1 | 0.92 | 0.00E + 00 | 0.89 | 1.50E-146 | 0.88 | 1.11E-119 |
| CCR5 | 0.88 | 2.03E-100 | 0.89 | 4.10E-143 | 0.86 | 6.12E-111 |
Th2 | CCR4 | 0.69 | 4.34E-44 | 0.70 | 3.40E-65 | 0.64 | 1.43E-44 |
| CCR8 | 0.51 | 3.61E-21 | 0.54 | 5.50E-33 | 0.33 | 3.63E-11 |
| HAVCR1 | 0.18 | 1.66E-03 | 0.31 | 7.50E-11 | -0.01 | 9.15E-01 |
Th17 | IL21R | 0.79 | 5.63E-67 | 0.82 | 6.50E-103 | 0.75 | 1.52E-69 |
| IL23R | 0.07 | 2.24E-01 | 0.15 | 1.70E-03 | 0.10 | 5.05E-02 |
| CCR6 | 0.44 | 0.00E + 00 | 0.56 | 3.10E-37 | 0.39 | 5.93E-15 |
Treg | FOXP3 | 0.61 | 0.00E + 00 | 0.64 | 4.30E-50 | 0.62 | 1.59E-40 |
| NT5E | 0.47 | 0.00E + 00 | 0.51 | 4.00E-29 | 0.47 | 2.95E-22 |
| IL7R | 0.69 | 0.00E + 00 | 0.71 | 1.50E-67 | 0.67 | 7.68E-51 |
T cell exhaustion | PDCD1 | 0.54 | 2.20E-24 | 0.56 | 8.70E-37 | 0.53 | 5.05E-28 |
| CTLA4 | 0.65 | 2.05E-38 | 0.69 | 1.10E-60 | 0.63 | 3.44E-43 |
| LAG3 | 0.58 | 0.00E + 00 | 0.52 | 7.20E-31 | 0.55 | 1.21E-30 |
M1 Macrophage | NOS2 | 0.06 | 2.99E-01 | 0.23 | 1.40E-06 | 0.07 | 1.61E-01 |
| IRF5 | 0.42 | 4.76E-15 | 0.47 | 2.30E-24 | 0.48 | 1.33E-22 |
| PTGS2 | 0.23 | 3.98E-05 | 0.30 | 2.60E-10 | 0.24 | 2.34E-06 |
M2 Macrophage | CD163 | 0.90 | 0.00E + 00 | 0.82 | 1.40E-103 | 0.91 | 3.06E-146 |
| MRC1 | 0.73 | 0.00E + 00 | 0.78 | 7.10E-88 | 0.72 | 2.55E-61 |
| CD209 | 0.61 | 0.00E + 00 | 0.70 | 2.00E-64 | 0.62 | 4.07E-41 |
TAM | CCL2 | 0.61 | 0.00E + 00 | 0.59 | 4.20E-42 | 0.63 | 5.97E-42 |
| CD86 | 0.95 | 0.00E + 00 | 0.95 | 1.20E-223 | 0.95 | 2.34E-191 |
| CD68 | 0.94 | 0.00E + 00 | 0.93 | 9.80E-181 | 0.72 | 4.95E-62 |
Monocyte | CD14 | 0.93 | 0.00E + 00 | 0.93 | 7.90E-193 | 0.94 | 2.90E-170 |
| CD33 | 0.91 | 0.00E + 00 | 0.88 | 1.00E-138 | 0.88 | 3.49E-125 |
| ITGAX | 0.85 | 0.00E + 00 | 0.84 | 8.60E-117 | 0.85 | 2.15E-103 |
Natural killer cell | B3GAT1 | 0.08 | 1.87E-01 | 0.20 | 4.70E-05 | -0.04 | 4.96E-01 |
| KIR3DL1 | 0.33 | 2.49E-09 | 0.45 | 8.20E-23 | 0.35 | 4.27E-12 |
| CD7 | 0.61 | 0.00E + 00 | 0.63 | 1.30E-47 | 0.60 | 1.76E-38 |
Neutrophil | FCGR3A | 0.92 | 4.84E-126 | 0.94 | 6.20E-195 | 0.94 | 4.66E-170 |
| CD55 | 0.20 | 4.58E-02 | 0.34 | 6.10E-13 | 0.28 | 4.80E-08 |
| ITGAM | 0.89 | 0.00E + 00 | 0.92 | 8.10E-171 | 0.91 | 8.75E-142 |
Dendritic cell | CD1C | 0.53 | 6.36E-23 | 0.57 | 2.40E-37 | 0.50 | 6.45E-25 |
| THBD | 0.55 | 0.00E + 00 | 0.64 | 2.40E-50 | 0.57 | 2.69E-34 |
| NRP1 | 0.51 | 3.54E-21 | 0.54 | 1.10E-33 | 0.53 | 7.75E-29 |
Cor correlation coefficient; TAM, tumor-associated macrophages; Th1/2, T helper type 1/2 cells; Treg, regulatory T cells; P, P-value;
Enrichment analysis of co-expressed genes and DEGs indicated that C3AR1 is involved in the densely mediated signaling pathways of chemokines, interleukins, and interferons. Therefore, we evaluated the correlation between C3AR1 and chemokines, interleukins and interferons in the ovarian cancer microenvironment in the TCGA database. As shown in Fig. 5, C3AR1 has strong correlations with 24 chemokines (Fig. 5A) including CCR1 and CCR5, 26 interleukins (Fig. 5B) including IL2RA, IL10RA and IL21R, and six interferons (Fig. 5C) including IDO1 and INFG. All these factors have been shown to be important immune regulators.
Correlation between C3AR1 and m6A modifiers in OC
M6A-related genes are widely involved in multiple processes such as ovarian cancer metastasis, chemotherapy resistance, and tumor microenvironment regulation. Analysis of TCGA found that C3AR1 co-expressed with multiple m6A-related genes, including HNRNPC, IGF2BP2, RBMX and ZC3H13, which were significantly positively correlated (Figure S5A, P < 0.05). Genes significantly negatively correlated with C3AR1 include ALKBH5, IGFBP3, METL14, RBM15, WTAP, YTHDC2 and YTHDF3 (P < 0.05). In addition, these 9 methylation related genes were differentially expressed in C3AR1 high expression and low expression groups (RBMX, IGFBP2, WTAP, VIRMA, YTHDF3, METTL14, FTO, HNRNPC and YTHDC2) (Figure S6). Kaplan-Meier curve shows FTO (HR = 1.26, 95%CI 1.1–1.44, P = 0.001), IGFBP3 (HR = 1.25, 95%CI 1.08–1.44, P = 0.0031), WTAP (HR = 1.39, 95%CI The high expression of 1.21–1.59, P = 1.5E-06) and YTHDF3 (HR = 1.18, 95%CI 1.03–1.34, P = 0.015) are closely related to the poor prognosis of LUAD (Figure S5B-E). These results indicate that C3AR1 ultimately affects the progression and prognosis of ovarian cancer by participating in the m6A modification of OC, especially with WTAP.
C3AR1 promotes the proliferation of ovarian cancer cells
The expression level of C3AR1 in SKOV3 cells was low, therefore we overexpressed C3AR1 to conduct subsequent functional experiments. mRNA and protein levels of C3AR1 were significantly increased after plasmid transfection (Fig. 6B-D), and the mRNA overexpression efficiency was the highest when the plasmid was 2ug/ml. Therefore, 2ug/ml plasmid was selected for subsequent study. The cell proliferation rate of the control group, empty vector plasmid group and C3AR1 plasmid group was detected by EdU assay. It was found that C3AR1 overexpression significantly promoted the proliferation of SKOV3 cells (Figure.6E).