In this study, we found that dysregulation of LINC01671 was closely correlated with poor prognosis in CRC patients, and regulated YY1 to induce EMT status of CRC cells by sponging miR-141-3p/186-5p; additionally, cancer cells undergoing EMT secreted CCL2 to recruit M2-like TAM infiltration. LINC01671 played a crucial role in EMT-related TAM recruitment and alternative activation in CRC.
More and more emerging literatures have reported lncRNAs as essential diagnosis and therapeutic biomarkers in the tumorigenesis of human solid cancers during past decades(1, 26). LncRNA dysregulation in multiple tumors was also observed and involved in many molecular biology network interactions that modulate oncological behaviors(17, 26). However, the molecular mechanism of lncRNA-mediated carcinogenic behaviors by regulating crosstalk between cancer cells EMT and TAM in TME remains obscure. The full post-transcriptional length of LINC01671 is 1726 bp, which permits it to take an active part in many biological processes. Recently, four studies predicted the potential tumor-promoting efficacy of LINC01671 in lung cancer, gastric cancer, and breast cancer(27–30). But these papers were deficient in comprehensive and in-depth elucidations of its oncogenic role and underlying molecular mechanisms. Herein, we reported for the first time that the LINC01671 level was significantly upregulated and correlated with tumor size, tumor grade, LVI, TI, LNM and TNM stage, predicting unsatisfactory prognosis of CRC patients. Gain and loss-of-function assays demonstrated that LINC01671 exerted oncogenic effects by positively inducing EMT to affect tumorigenic behaviors, while LINC01671-mediated EMT of CRC modulated TAM recruitment and M2-like polarization, reprogrammed the tumor microenvironment, ultimately promoting CRC progression. All of these phenomena suggested that LINC01671 might be considered as a potential novel biomarker for CRC diagnosis or treatment or prognosis.
It is well acknowledged that lncRNA mainly exerts biological functions through various regulatory mechanisms such as competitive endogenous RNA (ceRNA), gene transcriptional, translational regulation and protein interactions(16, 17, 26, 31). The ceRNA hypothesis shows that lncRNA serves as a sponge of miRNA to abolish miRNA inhibition on its downstream target genes(31). For example, long non-protein coding RNA lnc-APUE was highly expressed in hepatocellular carcinoma samples, and interacted with miR-20b to prevent miR-20b from targeting and inhibiting E2F1 expression, thereby promoting cell proliferation and tumor growth(32). In lung cancer, lncRNA JPX upregulated Twist1 level by competitively sponging miR-33a-5p to facilitated tumor growth and metastasis(19). Analogously, LINC00680 functioned as a ceRNA to support esophageal squamous cell carcinoma progression via the miR-423-5p/PAK6 axis(33). In this work, we found that LINC01671 was primarily localized in the cytosol of CRC cells, and induced cancer cells EMT by regulating YY1. RNA pull-down and RIP assays further revealed that LINC01671 acted as a sponge for miR-141-3p and miR-186-5p, which were both proven to be downregulated and tumor suppressors in carcinomas(34, 35). MiRNA, as a post-transcriptional regulator, is involved in leading to a decrease of mRNA stability through directly combining with target sites of mRNA 3′-UTR(36). Consistently, RIP and luciferase reporter assay also confirmed miR-141-3p/186-5p-mediated degradation of YY1 mRNA and downregulation of YY1 protein. Moreover, the pro-tumor effects of LINC01671 or YY1 on EMT program and cell invasion were partially eliminated by ectopic miR-141-3p/186-5p expression in the rescue experiments. Therefore, all these data illustrated that LINC01671 served as a ceRNA of miR-141-3p and miR-186-5p to regulate YY1 expression, contributing to CRC EMT and progression.
YY1 possessed multiple functions as it acted as both a transcriptional activator and transcription repressor(7, 12). The overexpression of YY1 had been observed in various human cancers, such as liver carcinoma, gastric cancer, osteosarcoma, and CRC(10–12). High YY1 expression indicated the worsening state of malignant carcinomas and poor prognosis(11, 12). Recent studies had presented that YY1 participated in tumor EMT and supported cancer progression(7, 8). Mechanistically, YY1 primarily stimulated the expression of transcription factor Snail, which in turn caused cancer cells to undergo epithelial-mesenchymal transition by directly restraining metastasis-suppressor genes expression, such as claudins and E-cadherin, and elevating the level of metastasis-inducer genes, such as vimentin(7, 9). Han et al. also reported that YY1 induced EMT and metastasis of hepatocellular carcinoma by targeting quaking, which was closely related to EMT program through facilitating the generation of circRNAs(7). In addition, some studies showed that YY1 was involved in regulating TGF-β1 signaling-mediated EMT(8). In the present study, we also found that YY1 played a key role in LINC01671-induced EMT of CRC. Intriguingly, in the rescue assays of LINC01671-miR-141-3p/186-5p-YY1 regulatory network, YY1 overexpression significantly elevated LINC01671 expression, while the level of LINC01671 was reduced by knocking down YY1. Considering that YY1 is a transcription factor, we surmised that YY1 not only acted as the target of LINC01671/miR-141-3p/186-5p, but also transcriptionally upregulated LINC01671 in turn. Further bioinformatic predictions, as well as luciferase report and CHIP assays confirmed that YY1 directly bound to LINC01671 promoter to activate its transcription, indicating that LINC01671-miR-141-3p/186-5p-YY1 formed a positive feedback loop in CRC cells. Accordingly, previous literatures had also revealed that YY1 transcriptionally upregulated certain lncRNAs in cancer cells(37), but the effect of YY1 on LINC01671 was firstly reported in this study. Therefore, our results elucidated the essential role of LINC01671/YY1 positive feedback loop in CRC EMT.
In the tumor microenvironment, EMT-educated cancer cells have a stronger capacity to invade and migrate, and can remodel TME by self-reprogramming to gain a survival advantage(3). Studies had shown that the tumor sites rich in mesenchymal cancer cells were accompanied by more M2-like TAM infiltration(14, 38). However, the underlying mechanisms remain largely unknown. In this work, we found that LINC01671/YY1-mediaed EMT of CRC cells recruited TAM and induced M2-like polarization. Mechanistically, LINC01671/YY1 promoted CCL2 production to affect recruitment and M2-like polarization of TAM. Similarly, Tang and colleague had also confirmed the regulatory effect of YY1 on CCL2(13). CCL2, as a chemokine of attracting TAM infiltration, was also associated with TAM M2 polarization(14, 15). Intriguingly, when we tried to block CCL2 signaling, the LINC01671/YY1-mediated M2-like polarization was only partially attenuated, revealing that there were other factors influencing the pro-M2 polarization role of LINC01671/YY1, which remains to be demonstrated by further experiments. Unfortunately, our study did not elucidate the specific mechanism by which YY1 regulated CCL2 expression. But further bioinformatics prediction indicated that there existed potential binding sites for YY1 in the promoter region of CCL2 gene (data not shown), suggesting that YY1 might be involved in the regulation of CCL2 transcription. Further investigation is needed to illuminate the mechanism of YY1 inducing CCL2 production in CRC.
TAM has been diffusely considered as a favorable condition for cancer progression, including cell growth, EMT, metastasis, immune suppression and chemoresistance in TME(39). TAMs in TME are mainly divided into the tumor-suppressive M1-like and the tumor-promoting M2-like phenotype, and various research tended to M2-like (CD163high, IL-10high, TGF-β1high, and CCL18high) TAMs, which predicted poor prognosis in CRC(39, 40). In our study, macrophages derived from THP-1 cells became CRC-stimulated TAMs after incubation with the conditioned supernatant of transfected CRC cells. The resultant TAM displayed the features of M2-like macrophage, with CD163high, IL-10high, TGF-β1high, and CCL18high phenotypes. In TME, M2-like TAM could secrete numerous cytokines such as IL-10, TGF-β1, CCL18, CCL2, and MMPs, to inhibit anti-tumor immune reaction, induce EMT, stimulate angiogenesis, and remodel extracellular matrix(14, 38, 39, 41). The vital role of M2-like TAM in EMT of CRC was preliminarily determined(4). Reversely, our work demonstrated that EMT-programmed cancer cells produced high CCL2 level, sequentially leading to recruitment and M2-like polarization of TAM in vitro and in vivo. Silenced YY1 and inhibited CCL2 suppressed M2-like TAM recruitment and polarization. Therefore, these data demonstrated that EMT status of cancer cells could affect CCL2-related M2-like TAM chemotaxis and polarization. Recent studies mainly focused on the pro-tumor function of TAM, however, studies concerned about tumor EMT-mediated TAM polarization were relatively rare. Our results firstly clarified that LINC01671/YY1 regulated CCL2 expression while modulating cancer cells EMT, thereby driving TAM recruitment and polarization in TME.