Renal cell carcinoma is one of the most deadly cancers in the urinary system, and its incidence is increasing year by year [11]. KIRC accounts for about 80% of all RCC histological subtypes and had the high mortality, invasiveness and metastasis rate [12]. Characteristic changes at the molecular level play a crucial role in the occurrence and metastasis of tumors, and had been considered as prognostic factors [13]. We found that the genes and tumor-infiltrating immune cells are differentially expressed between KIRC tumor tissue and normal tissue. In response to this finding, we constructed risk prediction models based on CeRNA network and tumor-infiltrating immune cells. Among them, we analyzed the correlation between CeRNA network and tumor-infiltrating immune cells, we inferred a potential mechanism of KIRC recurrence and that was lncRNA MALTA1 regulating Mast cells resting (MCs) and T cells follicular helper (TFH). Our findings could help clinical oncologists assess the prognosis and risk of recurrence of KIRC.
CeRNA network is that lncRNA compete for the miRNA recognition elements (MRE) to influences the mRNA regulation, thus, we speculate that this type of analysis could uncover molecular interactions and gene regulatory networks that have been missed by proteomic and conventional genomic methods [5]. Studies shown that 3’UTRs from coding genes may act as endogenous decoys, which binds to the MRE recognition sequence of miRNA through the principle of complementary base pairing and participates in post-transcriptional regulation of oncogenes and tumor suppressors [14], and lncRNA can also competitively bind the MRE sequence of miRNA to regulate the expression level of mRNA in tumor cells [15]. More importantly, lncRNA modulated mitochondrial membrane potential and enhanced the release of cytochrome C, indicating that apoptosis of KIRC induced by lncRNA may belong to mitochondria-mediated apoptosis [16], and lncRNA can also participate in the proliferation of cancer cells by promoting the expression level of KIRC distant metastasis related genes[17]. In this study, we obtain 3074 differentially expressed mRNA (1055 upregulated and 2019 downregulated), 359 differentially expressed lncRNA (71 upregulated and 280 downregulated), and 132 differentially expressed miRNA(70 upregulated and 62 downregulated) through bioinformatics analysis the RNA sequencing (RNA-seq) data of KIRC. We have constructed a CeRNA network that can prepare the prognosis of KIRC through the analysis of hypergeometric distribution test and correlation test, which contains 20 mRNAs, 7 lncRNAs and 14 miRNAs. And then ,we use the single factor COX analysis, lasso regression and multivariate Cox regression analysis to obtain 9 genes (RELT, MYO9B, KCNN4, SIX1, OTOGL, MALAT1, HSA-Mir-130b-3P, HSA-Mir-200b-3p, HSA-Mir-21-5p) in CeRNA that were significantly correlated with the overall survival rate of KIRC. We constructed a prognostic model using these risk genes and got the AUC values (for 1-year survival rates: 0.794, 3-year survival rates: 0.762, and 5-year survival rates: 0.782). After matching the 9 risk genes and the CeRNA network, it was found that hsa-miR-200-3p (miRNA), SIX1 (mRNA) and MALAT1 (lncRNA) were significantly correlated (P = 4.30E-6). SIX1 is a key transcription factor involved in the occurrence and development of a lot of tumor and its biological activity is regulated by miRNA [18, 19]. SIX1 may contribute to ovarian epithelial carcinogenesis by simultaneously increasing proliferation and decreasing apoptosis and imply that SIX1 may be an important target of ovarian cancer therapy response [19]. SIX1 can regulate the mitochondrial membrane potential by regulating the expression of the anti-apoptotic protein Bcl-2, and affect mitochondrial apoptosis via caspase-7, which suggests that SIX1 can be used as an effective target for the prognosis and treatment of gastric cancer [20]. However, there is few study had clearly shown the relationship between SIX1 and KIRC. Our results show that the decreased expression of miR-200 in KIRC may be involved in the formation and development of cancer by promoting the increased expression of SIX1.
Many previous studies have shown that the miR-200 family members of KIRC were significantly downregulated compared with normal kidney tissue. It is speculated that miR-200b as tumor biomarkers in renal tumor biopsies is feasible [21, 22]. Rola Saleeb et al. found that miR-200b was down-regulated in primary KIRC and further decreased in metastatic foci. More importantly, Kaplan-Meier survival curves indicate that miR-200b and miR-200c positive patients have significantly longer disease-free survival [23]. Therefore, the expression of miR-200 is negatively correlated with the progression of KIRC. In this study, we found that the HR value of miR-200 was less than 1 through single-factor COX, lasso regression and multi-factor COX analysis, indicating that miR-200 is a gene that protects renal clear cell carcinoma. This is consistent with the results of miR-200 family members mentioned in previous studies that inhibit the proliferation, migration, and invasion of cancer cells [24].
There is another important RNA in the significantly related network of hsa-miR-200-3p(miRNA)-SIX1(mRNA)-MALAT1(lncRNA), this is lncRNA MALAT1. Metastasis associated with lung adenocarcinoma transcript-1 (MALAT1) is one of the long non-coding RNA (lncRNA) associated with tumors which is composed of more than 8000 nucleotides and located on chromosome 11q13 [25]. MALAT1 was initially considered to be an effective prognostic factor for non-small cell lung cancer (NSCLC), tumors with high MALAT1 expression have a nearly five-fold increase in the risk of metastasis compared with tumors with low MALAT-1 expression [26]. Feng et al. proved that the direct binding between MALAT1 and miR-200a can promote the proliferation of lung cancer cells and lead to the development of gefitinib resistance [27]. In addition to NSCLC, MALAT1 is also present in other malignant tumors. In endometrial cancer, the expression of MALAT1 is negatively correlated with the expression of miR-200c, and its knockdown can promote the expression of miR-200c in cancer cells [28]. Studies have found that MALAT1 can also adjust the expression of miR-140 in prostate cancer cells to change the mRNA and protein expression levels of apoptosis inhibitor protein to promote the occurrence and development of tumors [29]. More importantly, some studies have shown that MALAT1 in KIRC promotes the growth of cancer cells by inhibiting cell apoptosis and accelerating the expression of specific proteins of epithelial-mesenchymal transition (EMT) [30]. In this study, we found that MALAT1 is a high-risk factor in KIRC through bioinformatics analysis, and its expression level is positively correlated with tumor formation and development, which is consistent with the role of MALAT1 in other tumors. Therefore, we speculate that hsa-miR-200-3p (miRNA) - SIX1 (mRNA) - MALAT1 (lncRNA) plays an important role in the development of renal clear cell carcinoma. However, how these genes affect the occurrence and development of KIRC is still unclear.
In the past few decades, more and more evidences show that the tumor-cell phenotype is determined by the intrinsic activity of cancer cells and the interaction of cells (especially tumor-infiltrating immune cells, TIICs) in the tumor microenvironment [31, 32]. Complex interactions between the immune cells in this microenvironment may promote cancer progression by inducing immune dysfunction in RCC patients [33]. Interestingly, lncRNA MALAT1 is an important inflammation regulator. In patients with systemic lupus erythematosus (SLE), lncRNA MALAT1 can be used as a CeRNA to interfere with the inhibitory effect of miRNA on the inflammatory factor IL-21, so the expression of MALAT1 in SLE patients is significantly increased [34]. During an asthma attack, MALAT1 activates airway inflammation and airway hyperresponsiveness of T cells by inducing the expression of miR-155 in T cells [35, 36]. However, it is not clear whether lncRNA MALAT1 is involved in the process of immune invasion, immune cell killing and immune escape in the tumor microenvironment of RCC.
In this study, we analyzed the correlation between gene expression and tumor-infiltrating immune cells to prove that MALAT1 and MCs, (R = -0.27, P = 8.95E-5), MALAT1 and TFH (R = 0.23, P = 9.85E-4) were significantly correlated. Therefore, Therefor, we speculate that MALAT1, the core molecule of the CeRNA network, participates in the occurrence of KIRC by regulating the content of tumor-infiltrating immune cells, such as TFH and MCs.
We first analyzed by using the CIBERSORT algorithm and found that the expression levels of resting MCs and TFH between KIRC and normal tissues are significantly different. Compared with other algorithms, CIBERSORT is the best algorithm to distinguish between resting and IgE-activated MCs. Our results suggested that most of the MCs in KIRC are in resting state. Studies have shown that MCs activated by IgE can prevent the occurrence of cancer [37]. However, there are high concentrations of polyamines in malignant tumor cells [38]. Polyamines oxidized by polyamine oxidase can prevent the activation of MCs by IgE, thereby inhibiting the tumor suppressor effect of resting MCs [39]. The resting MCs in KIRC can promote the immune escape of tumor cells, which is beneficial to tumors growth [40]. Therefore, when the content of resting MC is higher, it may promote the occurrence of tumors. We observed that the expression of MALAT1 increased in KIRC tumor tissues, and the content of resting MCs decreased observably. There is a negative correlation between MALAT1 and MCs. Above all, we speculate that MALAT1 may affect the generation of KIRC by the content of resting MCs.
TFH is an immune cell located in the peritumoral tertiary lymphoid structures (TLS), which belongs to a subset of CD4+ T cells [41]. TFH provides signals for the proliferation and activation of B cells by secretes a large number of chemokines in B cell [42]. Our results proved that the expression of lncRNA MALAT1 in the CeRNA network is positively correlated with the contents of TFH. The expression of MALAT1 in KIRC cells is increased, which suggested that the content of TFH in KIRC cells is higher than that in normal tissues. By using Multi-factor COX analysis, we found that the content of TFH in KIRC cells is ascended distinctly, which is consistent with our speculation. Therefore, we believed that TFH may be one of the maker cells of immune infiltration in KIRC.
In summary, this study constructed a KIRC related CeRNA network through the TCGA database, and selected risk genes through single factor COX analysis, lasso regression and multivariate Cox regression analysis to predict the survival rate of patients. We obtained a significant correlation network of hsa-miR-200-3p(miRNA)-SIX1(mRNA)-MALAT1(lncRNA) through correlation analysis. In order to explore how the genes in this network affect the formation of KIRC, this study also constructed another risk model based on tumor-infiltrating immune cells, and screened out risky tumor-infiltrating immune cells related to the occurrence and development of KIRC. We speculated that hsa-miR-200-3p and MALAT1 affect the internal mechanism of KIRC by regulating the expression of SIX1, and the relationship between MALAT1 and immune infiltration, so as to promote the self-management of KIRC patients.
Our research is a correlation study from multiple dimensions rather than a biological mechanism study. Based on the results of this research, we will verify our conclusions through biological experiments. In the future, we will combine more data and more experiments to explore the impact of KIRC molecular biology on CeRNA, as well as the relationship between tumor cells, TFH and resting MCs.