Uveal melanoma (UM) is the most common intraocular cancer in adults and can develop to the liver [1, 2]. For patients with metastatic UM, the 1-year survival rate is 20%, the 5-year survival rate is less than 5%, and the median overall survival rate is only 5–7 months [1]. The miRNA-mRNA regulatory network has not been established in patients with UVM metastasis, so we constructed a miRNA-mRNA regulatory network in this study, hoping to improve the targeted therapy for patients with UVM metastasis.
Through the analysis of 10 hub target mRNAs and Key mRNAs obtained by combining TCGA database, it was found that most of them were related to cancer metastasis and survival. In addition, the expression of some key mRNAs and corresponding DNA methylation are closely related to the survival of UVM. These findings suggest that the specific miRNA and mRNA functional analyses contribute to identifying new predictive biomarkers and therapeutic targets for UVM patients in clinical settings.
Our results indicated that some of 22 miRNAs play roles in cancer proliferation and metastasis. Specifically, it has been proved that miR-149 plays an important role in cancer metastasis [32, 33]. In nasopharyngeal cancer and melanoma, miR-149 is upreg-ulated. In our study miR-149 was upregulated in UVM metastatic tissues, consistent with the above findings. Similarly, a clinical observation also revealed an association between high expressed miR-342 and improved long-term survival of metastatic pa-tients [34]. miR-6819-3p may enhance the expression levels of actin γ1 and inhibit the expression levels of Toll-like receptor 3 in alcohol-associated hepatocellular carcinoma tissues, which plays an important role in the occurrence and development of hepato-cellular carcinoma [35]. Studies have shown that miR-6881-3p and ANXA2 promote autophagy, proliferation and invasion of epithelial ovarian cancer by interacting with circEEF2 (has-circ-0048559) [36]. Thus, miR-149, miR-342, miR-6819 and miR-6881 may represent new potential therapeutic targets in UVM metastasis, and in-depth study of the role of miRNAs and their regulated mRNAs in cancer metastasis has positive sig-nificance for the treatment of patients with UVM metastasis. In the present study, we predicted the target mRNAs of these miRNAs, performed GO and pathway enrichment analyses of the target mRNAs. Notably, we revealed some novel UVM metasta-sis-related pathways such as MAPK, endocytosis, spliceosome and Wnt signaling pathways.
In this study, we found that HNRNPA1, SRSF1, TRA2B, TIAL1, FUS, FN1, SFPQ, and HNRNPU in 10 hub target mRNAs play an important role in cancer metastasis. For in-stance, hnRNPA1 can regulate the gene expression and translation of several key play-ers related to tumorigenesis and cancer progression [37]. It is reported that SUMOy-lation of HNRNPA1 at the lysine 113 residue, which mediated recognition of ELNAT1 by the endosomal sorting complex required for transport (ESCRT) and facilitated its packaging into EVs, can promote lymph angiogenesis and LN metastasis in bladder cancer [38]. Importantly, FN1 was associated not only with cancer metastasis, but also with survival in patients with UVM metastasis. Sun et al. found that transglutaminase 2 (TGM2) activates FN1 via nuclear factor kappa B (NFκb) to promote papillary thyroid cancer invasion and metastasis [39]. Liang et al. found that promoting transi-tion-associated lncRNA (PTAL) positively regulates the expression of FN1 through sponging of miR-101 and promotes ovarian cancer cell metastasis by regulating epithe-lial-mesenchymal transition [40]. In addition, a previous study predicted that FN1 is also associated with survival in gastric cancer patients [41]. In the present study, we also found that FN1 is closely correlated with survival of the patients with UVM, indicating that FN1 may be a good biomarker of the patients with metastatic UVM.
In the present study, we identified 14 key mRNAs in combination with TCGA. We found that FOXD3, SPDEF and CA12 in 14 key mRNAs play an important role in cancer metastasis, and CA12 is also associated with survival in patients with metastatic UVM. A study has found that down regulation of FOXD3 could promote the growth, migration, invasion and angiogenesis of neuroblastoma cells [42]. Foxd3 can inhibit the progression and metastasis of colorectal cancer by regulating UBA2 induced by Mir-133a [43]. Meanwhile, a transcriptional analysis of several genes related to tumor metastasis, in-vasion and epithelial-mesenchymal transition (EMT) showed that SPDEF expression could selectively down-regulate MMP9 and MMP13 in prostate cancer cells, playing an inhibitory role in tumor metastasis [44]. Moreover, in hepatocellular Cancer cell lines, knockdown of SPDEF increased tumor growth and metastasis, and the increase in growth and metastasis may be due to elevated expression of the EMT driver SLUG [45]. In this study, we found that FOXD3 and SPDEF was down-regulated in metastatic UVM tissues, indicating that both FOXD3 and SPDEF were associated with metastatic UVM. A study has revealed that tumor cytokines from hepatocellular carcinoma samples could induce the upregulation of CA12 in tumor-infiltrating TAMs through the HIF1α pathway, thereby inducing the protumor phenotype that promotes tumor growth and metastasis [46]. An investigation has confirmed that high expression of CA12 is asso-ciated with good prognosis in breast cancer [46]. In this study, we also found that CA12 was up-regulated in UVM tissues. Therefore, FOXD3, CA12 and SPDEF are expected to become new biomarkers and therapeutic targets for UVM metastasis.
Variation in DNA methylation has been widely observed in many human diseases, in-cluding autoimmune disease [47] and metabolic disorders [48], thus, DNA methylation level has also been recognized as a hallmark of cancers [49]. Currently, DNA methyla-tion level has rapidly emerged as a diagnostic and prognostic signature in many cancers because methylation level can also reflect survival outcomes. Studies have shown that TDRD10 methylation was a promising diagnostic and prognostic marker for breast cancer as well as others [50]. Epigenome analysis showed that COL11A2 was a new pan cancer differential methylation cue signal in the 70kb intergenic region upstream of the SASH1 tumor suppressor gene [51]. The relationship between NYNRIN methylation and cancer has not been studied. In the current study, we also found that the levels of TDRD10, COL11A2 and NYNRIN in key mRNAs were closely related to their methyla-tion level. Although we could not find the relevant research on the relationship be-tween these mRNAs and the patients with metastatic UVM in the PubMed database, we still speculate that the methylation level of the key mRNAs is an important factor which can affect the survival rate of UVM patients.
Nowadays, with the popularization of the concept of precision medicine, the treatment methods for individual and individual differences are increasingly valued by scientific researchers and medical investigators. Therefore, it is necessary to find new biomarkers and therapeutic targets for the patients based on the differences between the individu-als. On the basis of this new concept and the TCGA platform, we have constructed a new miRNA-mRNA network that includes the above mentioned hub mRNAs and key mRNAs. We confirm that the construction of this network has important guiding sig-nificance for the prediction of metastasis in patients with UVM and the treatment op-tions.