Analysis the IRF4 expression level in cancers
The expression levels of IRF4 in tumors and normal tissues of different cancers were analyzed from Ocomine database. As shown in Fig. 1a, the IRF4 expression was higher in brain and central nervous system (CNS) cancer, kindey cancer, lymphoma, melanoma and myeloma cancers, while lower expression of IRF4 was presented in breast, colorectal, gastric, leukemia and sarcoma cancers, compared to the corresponding normal tissues. To further verify the expression levels of IRF4 in tumors, the RNA-seq data of TCGA was used from TIMER database. The expression levels of IRF4 between tumor and adjacent normal tissues were shown in Fig. 1b. IRF4 expression was significantly higher in HNSC (head and neck cancer), KIRC (kidney renal clear cell carcinoma), LUAD (lung adenocarcinoma), LUSC (lung adenocarcinoma). However, IRF4 expression was significantly lower in BLCA (bladder urothelial carcinoma), BRCA (breast invasive carcinoma), CODA (colon adenocarcinoma), KICH (kidney chromophobe), LIHC (liver hepatocellular carcinoma) and READ (rectum adenocarcinoma).
Prognostic value of IRF4 in Cancers
The prognostic value of IRF4 in cancer patients was assessed by PrognoScan database. IRF4 expression influences the prognosis of 6 type cancers patients (Table S1), especially, breast, colorectal and lung cancers (Fig. 2). GSE2034, GSE4922-GPL96 and GSE2990 breast cancer cohorts shown that high expression of IRF4 was related with better DMFS, DFS and relapse free survival (RFS). Similarly, high expression of IRF4 associated with a better prognosis were also detected in colorectal (GSE17536 and GSE17537) and lung cancer patients (jacob-00182-HLM, GSE31210 and GSE8894).
Kaplan-Meier plotter database was used to evaluate the IRF4 prognostic value based on Affymetrix microarrays. It’s shown that higher expression of IRF4 was related with better prognosis in breast cancer (OS, HR = 0.69, 95%CI = 0.55–0.86, logrank P = 0.00072; RFS, HR = 0.68, 95%CI = 0.61–0.76, logrank P = 2.9e-12). Base the above results, it’s suggested that increased expression of IRF4 was associated with better prognosis in breast and colorectal cancer.
Correlation analysis between IRF4 and clinical characteristic of breast cancer patient
The relationship of IRF4 and clinical characteristic of breast cancer patient including intrinsic subtype, grade and TP53 status was investigated by Kaplan Meier plotter database. Expression of IRF4 shown a better OS and RFS in breast cancer patient (Table 1). High expression of IRF4 was correlated with better OS in basal subtype (n = 241, HR = 0.47, 95%CI = 0.28–0.78, P = 0.0027) and grade 3 (n = 503, HR = 0.55, 95%CI = 0.39–1.27, P = 0.00042). Interestingly, IRF4 significantly influences RFS of all subtypes and grade 2 and 3 of breast cancer. Therefore, IRF4 can impact the prognosis of breast cancer patient with subtypes and grades.
Table 1
Correlation of IRF4 and clinical prognosis in breast cancer with clinical characteristic by Kaplan-Meier plotter.
Clinical characteristic | OS(n = 1402) | RFS(n = 3955) |
| n | HR | P value | n | HR | P value |
subtype | | | | | | |
basal | 241 | 0.47(0.28–0.78) | 0.0027 | 618 | 0.47(0.36–0.62) | 1e-08 |
Luminal A | 611 | 0.73(0.51–1.04) | 0.081 | 1933 | 0.73(0.62–0.87) | 0.00033 |
Luminal B | 433 | 0.84(0.58–1.22) | 0.35 | 1149 | 0.58(0.48–0.7) | 2.5e-08 |
HER2+ | 117 | 0.64(0.33–1.23) | 0.18 | 251 | 0.49(0.33–0.74) | 0.00041 |
Grade | | | | | | |
1 | 161 | 0.76(0.3–1.88) | 0.55 | 345 | 1.18(0.7–1.98) | 0.53 |
2 | 387 | 0.84(0.55–1.3) | 0.43 | 901 | 0.76(0.6–0.97) | 0.027 |
3 | 503 | 0.55(0.39–0.77) | 0.00042 | 903 | 0.75(0.6–0.93) | 0.01 |
HR Hazard ratio; OS overall survive; RFS relapse free survival; The bold values mean statistically significant. |
Immune infiltration correlation analysis between IRF4 and BRCA/COAD
We assessed the correlations of IRF4 expression with immune infiltration levels in BRCA and colon adenocarcinoma (COAD) by TIMER database (Fig. 3). IRF4 is relevant with higher tumor purity in BRCA and its subtypes as well as COAD. IRF4 expression has significant correlations with infiltrating levels of B cell (r = 0.542, P = 1.28e − 75), CD8 + T cell (r = 0.517, P = 6.39e − 68), CD4 + T cell (r = 0.582, P = 1.97e − 88), neutrophils (r = 0.56, P = 1.08e − 79), DCs (r = 0.624, P = 7.86e − 104) in BRCA. In addition, there were positive correlation with infiltration levels of B cell (r = 0.435, P = 4.66e − 20), CD8 + T cell (r = 0.264, P = 6.36e − 08), CD4 + T cell (r = 0.502, P = 4.78e − 27), macrophages (r = 0.327, P = 1.13e − 11), neutrophils (r = 0.517, P = 9.27e − 29), DCs (r = 0.552, P = 1.97e − 33) in COAD. Therefore, these results suggest that IRF4 may play an important role in immune infiltration in BRCA and COAD.
Correlation analysis between IRF4 and immune cells markers in BRCA and COAD
Given the positive correlation of immune infiltration with BRCA and COAD, the correlation between IRF4 and immune cells markers was analyzed by TIMER and GEPIA database. Immune gene markers of B cells, CD8 + T cells, T cells (general), M1 and M2 macrophages, neutrophils, DCs, Th1/2/17 cells, Tfh cells, Tregs, T cell exhaustion, NKs, monocytes and TAM in BRCA and COAD were assessed (Table 2). After adjusted by purity, the results revealed the IRF4 was significantly correlated with most immune marker sets of various immune cells. We found that the expression levels of most marker of B cells, CD8 + T cells, T cells (general), DCs, Th1, T cell exhaustion and monocytes have significant correlations with IRF4 expression in BRCA and COAD. The significant correlations between IRF4 and B cells (CD19, CD79A), CD8 + T cells (CD8A), T cell (general) (CD3D, CD3E, CD2), and monocytes (CD86, CD115) are presented in Fig. 4A-E. In addition, we also analyzed the above immune markers of B cells, CD8 + T cells, T cells (general), and monocytes by the GEPIA database (Table 3) and the most results were consistent with the TIMER database, except CD115 in COAD.
Table 2
Correlation analysis between IRF4 and immune cells markers by TIMER.
Description | Gene markers | BRCA | COAD |
None | Purity | None | Purity |
Cor | P | Cor | P | Cor | P | Cor | P |
B cell | CD19 | 0.807 | 4.47E-253 | 0.741 | 5.80–178 | 0.746 | 1.13E-82 | 0.71 | 1.79E-63 |
| CD79A | 0.981 | 0 | 0.85 | 3.03E-278 | 0.87 | 7.27E-143 | 0.84 | 2.10E-109 |
CD8 + T cell | CD8A | 0.797 | 1.17E-242 | 0.733 | 6.70E-168 | 0.63 | 5.61E-52 | 0.568 | 5.46E-36 |
| CD8B | 0.726 | 8.64E-181 | 0.635 | 8.80E-114 | 0.35 | 5.57E-15 | 0.279 | 1.02E-09 |
T cell(general) | CD3D | 0.816 | 4.30E-263 | 0.743 | 7.19E-175 | 0.716 | 4.78E-73 | 0.665 | 4.18E-53 |
| CD3E | 0.828 | 1.80E-278 | 0.76 | 1.64E-187 | 0.765 | 2.38E-89 | 0.72 | 3.54E-60 |
| CD2 | 0.835 | 1.74E-286 | 0.774 | 6.03E-199 | 0.765 | 3.73E-89 | 0.72 | 3.43E-66 |
M1 Macrophage | INOS | 0.039 | 0.192 | -0.02 | 0.954 | 0.288 | 3.47E-10 | 0.263 | 7.16E-08 |
| IRF5 | 0.35 | 5.44E-33 | 0.285 | 4.86E-20 | 0.288 | 3.17E-10 | 0.278 | 1.22E-08 |
| COX2 | 0.369 | 9.28E-37 | 0.223 | 1.12E-12 | 0.295 | 1.16E-10 | 0.24 | 9.80E-07 |
M2 Macrophage | CD163 | 0.502 | 2.18E-71 | 0.427 | 2.98E-45 | 0.583 | 4.57E-43 | 0.514 | 8.71E-29 |
| VSIG4 | 0.34 | 3.51E-31 | 0.219 | 3.00E-12 | 0.522 | 2.53E-33 | 0.44 | 1.25E-20 |
| MS4A4A | 0.574 | 2.40E-97 | 0.468 | 2.46E-55 | 0.583 | 4.31E-43 | 0.51 | 3.31E-28 |
Neutrophils | CD66b | 0.039 | 0.199 | 0.064 | 0.043 | 0.083 | 0.0778 | 0.14 | 0.00465 |
| CD11b | 0.422 | 9.10E-49 | 0.31 | 1.42E-23 | 0.514 | 2.59E-32 | 0.438 | 1.94E-20 |
| CCR7 | 0.774 | 4.25E-220 | 0.682 | 5.01E-137 | 0.758 | 1.21E-86 | 0.714 | 1.44E-64 |
Dendritic cell | HLA-DPB1 | 0.648 | 5.84E-132 | 0.509 | 1.66E-66 | 0.638 | 1.14E-59 | 0.566 | 1.03E-35 |
| HLA-DQB1 | 0.558 | 3.81E-91 | 0.45 | 1.26E-50 | 0.475 | 3.84E-27 | 0.407 | 1.36E-17 |
| HLA-DRA | 0.696 | 2.08E-160 | 0.599 | 1.04E-97 | 0.612 | 1.84E-48 | 0.545 | 9.26E-33 |
| HLA-DPA1 | 0.655 | 1.21E-135 | 0.542 | 3.90E-77 | 0.648 | 5.48E-56 | 0.586 | 8.59E-39 |
| BDCA-1(CD1C) | 0.586 | 1.43E-102 | 0.421 | 4.69E-44 | 0.562 | 1.87E-39 | 0.499 | 5.30–27 |
| BDCA-4(NRP1) | 0.246 | 1.22E-16 | 0.087 | 0.0058 | 0.457 | 5.66E-25 | 0.359 | 8.25E-14 |
| CD11c(ITGAX) | 0.578 | 6.31E-99 | 0.473 | 1.18E-56 | 0.582 | 7.85E-43 | 0.502 | 2.98E-27 |
Th1 | T-bet(TBX21) | 0.823 | 1.81E-272 | 0.759 | 2.25E-187 | 0.659 | 2.27E-58 | 0.615 | 1.37E-43 |
| STAT4 | 0.777 | 7.35E-223 | 0.684 | 8.08E-138 | 0.669 | 1.17E-60 | 0.615 | 1.11E-43 |
| STAT1 | 0.516 | 6.15E-76 | 0.509 | 9.45E-67 | 0.474 | 4.54E-27 | 0.418 | 1.32E-18 |
| INF-γ(IFNG) | 0.703 | 1.90E-164 | 0.644 | 2.34E-117 | 0.458 | 3.86E-25 | 0.435 | 3.48E-20 |
| TNF-α(TNF) | 0.377 | 2.10E-38 | 0.335 | 2.01E-27 | 0.526 | 5.68E-34 | 0.478 | 1.36E-24 |
Th2 | GATA3 | -0.341 | 2.00E-31 | -0.244 | 6.24E-15 | 0.592 | 1.06E-44 | 0.537 | 1.17E-31 |
| STAT6 | 0.1 | 0.00087 | 0.048 | 1.29E-01 | 0.217 | 2.70E-06 | 0.263 | 7.64E-08 |
| STAT5A | 0.334 | 3.77E-30 | 0.194 | 6.90E-10 | 0.381 | 2.68E-27 | 0.351 | 3.37E-12 |
| IL13 | 0.319 | 2.15E-27 | 0.269 | 6.34E-18 | 0.304 | 2.97E-11 | 0.25 | 3.52E-07 |
Tfh | BCL6 | 0.106 | 4.11E-04 | 0.064 | 0.0425 | 0.406 | 1.43E-19 | 0.323 | 2.56E-11 |
| IL21 | 0.519 | 9.43E-77 | 0.472 | 2.78E-56 | 0.358 | 2.90E-15 | 0.358 | 1.00E-13 |
Th17 | STAT3 | 0.11 | 0.000246 | 0.084 | 0.00841 | 0.459 | 3.15E-25 | 0.419 | 1.02E-18 |
| IL17A | 0.309 | 8.67E-26 | 0.234 | 8.47E-15 | 0.33 | 3.99E-13 | 0.364 | 3.59E-14 |
Treg | FOXP3 | 0.721 | 7.68E-177 | 0.66 | 1.71E-125 | 0.68 | 2.26E-63 | 0.621 | 1.32E-44 |
| CCR8 | 0.639 | 3.05E-127 | 0.602 | 3.98E-99 | 0.6 | 4.50E-46 | 0.554 | 4.26E-34 |
| STAT5B | 0.145 | 1.36E-06 | 0.086 | 0.00648 | 0.237 | 2.92E-07 | 0.236 | 1.44E-06 |
| TGFβ(TGFB1) | 0.298 | 4.78E-24 | 0.108 | 0.000624 | 0.517 | 1.04E-32 | 0.413 | 4.01E-18 |
T cell exhaustion | PD-1(PDCD1) | 0.761 | 1.63E-208 | 0.682 | 9.21E-137 | 0.627 | 2.12E-51 | 0.574 | 6.10E-37 |
| CTLA4 | 0.77 | 9.32E-217 | 0.715 | 1.24E-156 | 0.634 | 8.56E + 53 | 0.584 | 1.62E-38 |
| LAG3 | 0.597 | 2.32E-107 | 0.549 | 1.77E-79 | 0.616 | 2.90E-49 | 0.559 | 9.04E-35 |
| TIM-3(HAVCR2) | 0.514 | 3.63E-75 | 0.417 | 5.01E-43 | 0.572 | 3.28E-41 | 0.495 | 1.91E-26 |
| GZMB | 0.735 | 8.16E-188 | 0.661 | 5.98E-126 | 0.24 | 2.03E-07 | 0.216 | 1.09E-05 |
Natural killer cell | KIR2DL1 | 0.442 | 1.08E-53 | 0.384 | 3.44E-36 | 0.257 | 2.51E-08 | 0.19 | 0.000115 |
| KIR2DL3 | 0.471 | 1.06E-61 | 0.404 | 2.95E-40 | 0.25 | 6.18E-08 | 0.175 | 3.79E-04 |
| KIR2DL4 | 0.566 | 2.85E-94 | 0.521 | 3.56E-70 | 0.383 | 1.84E-17 | 0.309 | 2.02E-10 |
| KIR3DL1 | 0.511 | 2.38E-74 | 0.441 | 1.36E-48 | 0.289 | 2.77E-10 | 0.195 | 7.86E-05 |
| KIR3DL2 | 0.568 | 5.20E-95 | 0.492 | 7.94E-62 | 0.393 | 2.11E-06 | 0.324 | 2.39E-11 |
| KIR3DL3 | 0.304 | 5.05E-25 | 0.266 | 1.48E-17 | 0.219 | 2.31E-06 | 0.195 | 7.63E-05 |
| KIR2DS4 | 0.408 | 2.56E-45 | 0.331 | 6.72E-27 | 0.255 | 3.14E-08 | 0.231 | 2.69E-06 |
Monocyte | CD86 | 0.611 | 1.68E-113 | 0.526 | 7.84E-72 | 0.614 | 9.05E-49 | 0.536 | 1.35E-31 |
| CD115(CSF1R) | 0.505 | 2.20E-72 | 0.363 | 2.08E-32 | 0.654 | 3.09E-57 | 0.602 | 1.96E-41 |
TAM | CCL2 | 0.54 | 3.63E-84 | 0.433 | 1.17E-46 | 0.48 | 1.00E-27 | 0.408 | 1.07E-17 |
| CD68 | 0.511 | 4.34E-74 | 0.417 | 3.74E-43 | 0.466 | 4.09E-26 | 0.401 | 4.14E-17 |
| IL10 | 0.566 | 5.16E-94 | 0.474 | 7.26E-57 | 0.581 | 1.02E-42 | 0.545 | 9.37E-33 |
BRCA breast cancer; COAD colon adenocarcinoma; Cor R value of Spearman’s correlation; None correlation without adjustment; Purity correlation adjusted by purity; Th T helper cell; Tfh Follicular helper T cell; Treg regulatory T cell; TAM tumor-associated macrophage. |
Table 3
Correlation analysis between IRF4 and immune cells markers by GEPIA.
Description | Gene markers | BRCA | COAD |
Tumor | Normal | Tumor | Normal |
R | P | R | P | R | P | R | P |
B cell | CD19 | 0.78 | 2.20E-227 | 0.34 | 0.00021 | 0.72 | 3.00E-45 | 0.39 | 0.12 |
| CD79A | 0.82 | 3.70E-100 | 0.41 | 9.30E-06 | 0.87 | 3.80E-86 | 0.76 | 1.50E-06 |
CD8 + T cell | CD8A | 0.75 | 5.60E-74 | 0.27 | 0.0042 | 0.68 | 1.40E-38 | 0.18 | 0.26 |
T cell | CD3D | 0.77 | 6.50E-80 | 0.25 | 0.0068 | 0.76 | 3.30E-53 | 0.27 | 0.089 |
| CD3E | 0.87 | 1.20E-122 | 0.31 | 8.00E-04 | 0.82 | 1.50E-69 | 0.38 | 0.015 |
| CD2 | 0.86 | 3.90E-121 | 0.39 | 2.40E-05 | 0.83 | 7.40E-70 | 0.38 | 0.012 |
Monocyte | CD86 | 0.61 | 1.60E-109 | 0.15 | 0.11 | 0.71 | 1.90E-43 | 0.41 | 0.0075 |
| CD115(CFS1R) | 0.5 | 1.00E-70 | 0.26 | 0.0053 | 0.75 | 8.50E-51 | 0.62 | 1.80E-05 |
BRCA breast cancer; COAD colon adenocarcinoma; R value of Spearman’s correlation; Tumor correlation analysis in tumor tissue of TCGA; Normal correlation analysis in normal tissue of TCGA. |
GO function and pathway enrichment analysis
199 potential target genes were finally obtained from GTRD according to the selection criteria. GO functional and Reactome pathway enrichment analysis were performed on IRF4 and the above target genes. The top 10 items of GO and Reactome pathway enrichment were presented in bar charts (Fig. 5). The positively enriched GO items contained the top 10 positive BP, such as positive regulation of interleukin-2 biosynthetic process, definitive hematopoiesis, regulation of antigen receptor-mediated signaling pathway in BP category; interleukin-6 receptor binding, oxidoreductase activity, oxidizing metal ions, NAD or NADP as acceptor in MF category. The Reactome pathway was positively enriched in beta-oxidation of very long chain fatty acids, cross-presentation of particulate exogenous antigens (phagosomes), MAPK1 (ERK2) activation and MAPK3 (ERK1) activation. These results suggest that IRF4 may be involved in many important biological processes by regulating various genes.