3.1 Defining STAT family in STAD and normal tissue STAD.
To explore the function of STAT family in STAD, we first check their expression with Oncomine. And seven members of STAT family were identified in mammal (Fig. 1). Based on the data of Oncomine, the mRNA level of STAT1/3/5A/5B was significantly increased in STAD tissues (Table 1). A total of two datasets suggested that STAT1 was upregulated in gastric intestinal type adenocarcinoma (P = 6.96E-15) and gastric mixed adenocarcinoma (P = 1.34E-04), with a fold Change of 2.703, and 2.449, respectively (17). As for STAT3 expression, three datasets revealed upregulation of STAT3 STAD (18). According to the dataset of Mariarosaria, STAT5A (fold change = 2.563, P = 4.53E-04) and STAT5B (fold change = 2.89, P = 5.59E-045) were upregulated in Gastric Mixed Adenocarcinoma compared with normal gastric tissue.
Table 1
The mRNA levels of STAT family in STAD.
TLR | Type | Fold Change | P value | t-test | Reference |
STAT1 | Gastric Intestinal Type Adenocarcinoma Gastric Mixed Adenocarcinoma | 2.703 2.449 | 6.96E-15 1.34E-04 | 9.751 5.291 | PMID:12925757 PMID:12925757 |
STAT3 | Gastric Mixed Adenocarcinoma Diffuse Gastric Adenocarcinoma Gastric Intestinal Type Adenocarcinoma | 2.190 2.096 2.252 | 6.45E-06 4.08E-04 2.26E-10 | 7.834 5.117 7.653 | PMID:19081245 PMID:19081245 PMID:19081245 |
STAT5A | Gastric Mixed Adenocarcinoma | 2.563 | 4.53E-04 | 5.012 | PMID:19081245 |
STAT5B | Gastric Mixed Adenocarcinoma | 2.895 | 5.59E-04 | 5.077 | PMID:19081245 |
UALCAN was used to further detect STAT family expression in STAD. As a consequence, the TCGA STAD dataset suggested that the level of STAT1(P < 1E-12), STAT2 (P = 1.62E-12), STAT3 (P = 1.78E-08), STAT4 (P = 1.90E-07), STAT5A(P = 7.50E-08) and STAT6 (P = 0.0084) were increased in STAD (Fig. 2A). However, no significance of STAT5B expression was obtained between tumor tissues and normal tissues.
We then analyzed the correlation between STAT family expression and pathological stage as well as methylation. As shown in Fig. 2B, the expression of STAT2, STAT4, and STAT6 were significantly correlated with pathological stage. More specifically, with the development of STAD, the expression of STAT2, STAT4, and STAT6 gradually increased (Fig. 2B). Moreover, methylation analysis revealed that methylation could downregulate STAT family expression, except STAT4 (Fig. 2C)
3.2 The prognostic value of STAT family in STAD.
We then evaluated the prognostic value of STAT family in STAD. In overall survival, we revealed that STAD patients with high level of STAT1(HR = 0.7, 95%CI:0.59–0.83, P = 5E-05) and low level of STAT5A(HR = 1.25, 95%CI:1.06–1.48, P = 0.0094), STAT5B(HR = 1.54, 95%CI:1.3–1.83, P = 6.8E-07), and STAT6(HR = 1.28, 95%CI:1.08–1.52, P = 0.0042) (Fig. 3A). Moreover, low STAT1(HR = 0.6, 95%CI:0.48–0.75, P = 4.9E-06) level and high level of STAT4(HR = 1.36, 95%CI:1.09–1.7, P = 0.0056), STAT5A(HR = 1.53, 95%CI:1.22–1.91, P = 0.00016), STAT5B(HR = 2.37, 95%CI:1.88–2.99, P = 4.2E-14), and STAT6(HR = 1.36, 95%CI:1.09–1.69, P = 0.0063) were significantly associated with a worse PPS (Fig. 3B). As for PF analysis, the data suggested a worse PF in STAD patients with low STAT1(HR = 0.71, 95%CI:0.58–0.87, P = 0.00091) level and high STAT5B (HR = 1.61, 95%CI:1.31–1.97, P = 4.4E-06) level. (Fig. 3C)Therefore, STAT1/5A/5B/6 may functioned as biomarkers for the prognosis of STAD patients.
3.3 STAT family associates with the cancer hallmarks in STAD.
In order to evaluate the potential effects of disruption of STAT family in STAD patients, we performed genetic alteration, cancer-related pathway and STAT family. We found that STAT1, STAT2, STAT3, STAT4, STAT5A, STAT5B, STAT6 were altered in 8%, 6%, 6%, 7%, 11%, 8%, 11% of the queried STAD samples, respectively (Fig. 4A). Moreover, amplification and mRNA high were the most common alteration forms (Fig. 4A). Interestingly, these genetic alterations could affect the disease-free survival (Fig. 4B, P = 0.0117) but not overall survival (Fig. 4C, P = 0.306). we also analysis the role of STAT family in famous cancer-related pathway activity, which revealed that STAT family were involved in the activation of apoptosis pathway, EMT pathway, and hormone ER pathways (Fig. 4D). Besides, STAT family were involved in the inhibition of cell cycle pathway, and DNA damage response pathways (Fig. 4D).
3.4 Enrichment analysis of STAT family in STAD.
In order to further clarify the function of STAT family in STAD, we performed enrichment analysis of STAT family in STAD. GO enrichment analysis revealed that the functions of STAT family in STAD were mainly associated with JAK-STAT cascade, interleukin-9-mediated signaling pathway, and interleukin-21-mediated signaling pathway (Fig. 5A-5B). Enrichment analysis of KEGG pathway suggested that the significant association between STAT family and hepatitis B, Th1 and Th2 cell differentiation as well as Th17 cell differentiation (Fig. 5C-5D). MCODE was extracted and revealed the involvement of STAT family in JAK-STAT cascade, STAT cascade and cellular response to interleukin-9 (Fig. 5E-5F).
3.5 Drug sensitivity analysis of STAT family in STAD.
Previous results suggested that STAT family may play an important part in the tumorigenesis and progression of STAD, and some of STAT family may act as the therapeutic targets in STAD. An important way to develop drug therapy targets for cancer is to evaluate the association between genes and existing drug targets. Thus, we analyzed the correlation of STAT family expression and 481 small molecules or drugs from Therapeutics Response Portal (CTRP) as well as 265 small molecules or drugs from Genomics of Drug Sensitivity in Cancer (GDSC). Based on the results of GDSC, sensitivity (negative correlation) is associated with the expression of STAT5A and STAT5B (Fig. 6). Interestingly, similar results were obtained for CTRP. STAT5A and STAT5B were related with the drug sensitivity (negative correlation) (Fig. 7). The results may suggest that STAT5A and STAT5B is potential biomarkers for drug screening.
3.6 Immune infiltration of STAT5A expression in STAD.
Increasing evidences demonstrated that JAK/STAT pathway is crucial in the regulation of the immune response, and some of STAT family could act as the immune checkpoint inhibitor for various diseases, including cancers (5, 7, 19). Above result suggested STAT5A played an important part in the tumorigenesis and progression of STAD and acted as a potential biomarker for drug screening in STAD. Therefore, STAT5A were selected for further analysis its potential as the immune checkpoint inhibitor for STAD. We first evaluated the correlation between STAT5A level and immune cell infiltration, which revealed that STAT5A level showed significant correlation with the abundance of 5 immune cells (all P < 0.001, Fig. 8A), including CD8 + T cells (Cor = 0.419), CD4 + T cells(Cor = 0.35), Macrophage (Cor = 0.354), Neutrphils (Cor = 0.436) and Dendritic cells(Cor = 0.543). Moreover, somatic copy number alterations of STAT5A could significantly inhibit immune cell infiltration (Fig. 8B). In order to further clarify the significant role in immune infiltration, we also evaluated the correlation between STAT5A level and immune biomarkers, which have been widely reported (6, 20–22). As expected, strong correlations were obtained between STAT5A level and immune biomarkers in STAD (Table 2). We revealed positive correlation between STAT5A expression and the level of all the biomarkers of CD8 + T cell (CD8A, CD8B), T cell (CD3D, CD3E, CD2), B cell (CD19, CD79A), Monocyte (CD86, CD115), TAM (CCL2, CD68, IL10). Most of the biomarkers of M1 Macrophage, M2 Macrophage, Neutrophils and Natural killer cell were positively associated STAT5A expression. Moreover, the expression of all the biomarkers of Dendritic cell (HLA-DPB1, HLA-DQB1, HLA-DRA, HLA-DPA1, CD1C, NRP1, ITGAX), Th1 (TBX21, STAT4, STAT1, IFNG, TNF), Th2(GATA3, STAT6, IL13) and Tfh (BCL6, IL21) presented positive correlations with STAT5A level in STAD. Those STAD patients with high STAT5A level were also present with high level of FOXP3, CCR8, STAT5B, TGFB1, PD-1, CTLA4, LAG3, TIM-3, and GZM). These evidences demonstrated STAT5A as immune checkpoint inhibitor for STAD immunological therapy.
Table 2
The association between STAT5A and immune biomarkers in STAD.
Immune cells | Biomarkers | Correlation | P-value |
CD8 + T cell | CD8A CD8B | 0.479 0.259 | *** *** |
T cell (general) | CD3D CD3E CD2 | 0.424 0.454 0.455 | *** *** *** |
B cell | CD19 CD79A | 0.356 0.362 | *** *** |
Monocyte | CD86 CD115(CSF1R) | 0.416 0.488 | *** *** |
TAM | CCL2 CD68 IL10 | 0.228 0.296 0.411 | *** *** *** |
M1 Macrophage | INOS (NOS2) IRF5 COX2(PTGS2) | 0.104 0.257 0.044 | * *** 0.392 |
M2 Macrophage | CD163 VSIG4 MS4A4A | 0.452 0.376 0.452 | *** *** *** |
Neutrophils | CD66b (CEACAM8) CD11b (ITGAM) CCR7 | -0.041 0.548 0.42 | 0.429 *** *** |
Natural killer cell | KIR2DL1 KIR2DL3 KIR2DL4 KIR3DL1 KIR3DL2 KIR3DL3 KIR2DS4 | 0.147 0.129 0.186 0.202 0.285 0.064 0.094 | ** * *** ** *** 0.214 0.0674 |
Dendritic cell | HLA-DPB1 HLA-DQB1 HLA-DRA HLA-DPA1 BDCA-1(CD1C) BDCA-4(NRP1) CD11c (ITGAX) | 0.454 0.349 0.45 0.463 0.356 0.306 0.466 | *** *** *** *** *** *** *** |
Th1 | T-bet (TBX21) STAT4 STAT1 IFN-g (IFNG) TNF-a (TNF) | 0.497 0.452 0.226 0.267 0.165 | *** *** *** *** ** |
Th2 | GATA3 STAT6 STAT5A IL13 | 0.321 0.251 - 0.138 | *** *** - *** |
Tfh | BCL6 IL21 | 0.235 0.27 | *** *** |
Th17 | STAT3 IL17A | 0.43 -0.106 | *** 0.04 |
Treg | FOXP3 CCR8 STAT5B TGFb (TGFB1) | 0.419 0.481 0.533 0.357 | *** *** *** *** |
T cell exhaustion | PD-1 (PDCD1) CTLA4 LAG3 TIM-3 (HAVCR2) GZMB | 0.448 0.308 0.364 0.467 0.227 | *** *** *** *** *** |