br Fig Survival curves for EC
Fig. 3. Survival curves for EC patients in diﬀerent subgroup.
Fig. 4. The nomogram to predict 5-year OS for EC patients.
demonstrated the potential of the 6-lncRNA signature as an in-dependent biomarker for predicting OS in patients with colorectal cancer (Fan and Liu, 2018). An article of a six-miRNA prognostic model in EC had been published, but it only downloaded 348 patient data from TCGA database in 2016 (now the data has been updated in Feb. 2018) and did not validate the model in the testing set (Wang et al., 2018). Therefore, its results are not as stable and accurate as those of the 4-miRNA prognostic model.
Through our analysis, we proposed that hsa-mir-4758, hsa-mir-876, may reduce the survival rate of EC patients by negatively regulating cancer suppressor genes, while hsa-mir-142, hsa-mir-190b may increase the survival rate of EC patients by negatively regulating breast cancer oncogenes. This is an inference derived from the miRNA coe ﬃcients calculated by multivariate Cox regression analyses, which is consistent with its molecular mechanisms that have been reported. Consistent with the above results, miR142 was reported as a prognostic marker for EC patients, and it demonstrated that patients with high Liproxstatin-1 of miR142 had a longer PFS (Jayaraman et al., 2017). Similarly, it had a similar role as a tumor suppressor in breast cancer and non-small cell lung cancer (NSCLC) (Schwickert et al., 2015; Xiao and Liu, 2015). The miRNA-142 can target PIK3CA and HMGB1, thereby inhibiting the development of tumors (Wang et al., 2017). The tumor-suppressing eﬀect of miR-142 was also reflected in the ability to target the Frizzled 7 (Fzd7) receptor to inhibit the Wnt pathway in cervical cancer, and miR-142 regulated Smad3 thereby inhibiting the TGFβ-signaling pathway in NSCLC (Shrestha et al., 2017). In both HCC and lung cancer, miR-876 expression levels have been reported to correlate with patient prognosis. HCC patients with low miR-876 expression have significantly poorer OS and disease-free survival (DFS). Meanwhile, miR-876 inhibits EMT progression, migration and invasion of HCC cells by inhibiting BCORL1 (Xu et al., 2018). And miRNA-876 inhibits lung cancer EMT through targeting BMP-4 (Bao et al., 2017). Moreover, miR-190b has been reported as a tumor suppressor that can negatively regulate PTEN and Bcl-2, thereby inhibiting cell proliferation and induces apoptosis (Monteiro et al., 2015; Kang et al., 2017). One study found that miR-4758 was up-regulated in gangliogliomas and could identify gang-liogliomas from dysembryoplastic neuroepithelial tumors.
In order to further evaluate the function of the four miRNAs, the results of GO analysis and Pathway analysis showed that the function of their target genes was closely related to cancer. The results of the pathway analysis presented that not only miR-142 can regulate Wnt and TGF-b pathways, miR-190b and miR-876 were also likely to Gene 697 (2019) 86–93
of stem cells
Table 4 Pathway analysis of miR-142, miR-190b, KEGG pathway Apoptosis(hsa04210) mTOR signaling pathway(hsa04150) Wnt signaling pathway(hsa04310)
TGF-beta signaling pathway(hsa04350) p53 signaling pathway(hsa04115) FoxO signaling pathway(hsa04068)
Endometrial cancer(hsa05213) Colorectal cancer(hsa05210)
AMPK signaling pathway(hsa04152)
Signaling pathways regulating pluripotency (hsa04550)
Fig. 5. GO analysis and pathway analysis of three miRNAs.
regulate these two pathways. Each of the 10 pathways in Table 4 was closely related to the occurrence and development of cancer, and all three miRNAs participated simultaneously, indicating that these three miRNAs can play a cooperative regulatory role. Interestingly, pathway analysis indicated that miR-142, miR-190b, and miR-876 may be po-tentially involved in endometrial cancer-related pathways. This re-inforces the rationality of the 4-miRNA biomarker established in this study as a prognostic model for endometrial cancer.