HES5 Activates Long Noncoding RNA UCA1 to Induce Colorectal Cancer Progression by Modulating miR-185/NOTCH3 Signaling

Colorectal cancer (CRC) is one of the most common diagnosed cancers around the world. The poor prognosis and high fatality caused by metastasis are still the challenges for clinical treatment. Therefore, it is promising to clarify the detailed molecular mechanism of CRC metastasis. Accumulating evidences indicate that long noncoding RNAs (lncRNAs) play important roles in cancer progression including CRC. In this study, the function of lncRNA UCA1 was investigated. UCA1 was confirmed to be highly expressed in colorectal cancer. Moreover, the UCA1 expression level was positively related to tumor stages. Silencing UCA1 showed inhibitory effect on cell proliferation and metastasis. Both UCA1 and NOTCH3 were validated as direct targets of miR-185. Silencing UCA1 repressed NOTCH3 expression through the miR-185 sponge. NOTCH3 was found to be highly expressed in CRC patients and positively related to UCA1 expression. Furthermore, HES5 was verified as a transcription factor of UCA1, which induced UCA1 expression. In conclusion, UCA1 is a direct target of HES5. UCA1 promotes CRC metastasis through regulating the miR-185/NOTCH3 axis.


Introduction
Colorectal cancer (CRC) is one of the most commonly diagnosed cancers around the world [1]. In year 2018, nearly 1.8 million new CRC cases were diagnosed and over 800,000 deaths occurred due to metastasis [2]. Although the treatments for CRC have been improved recently, the poor prognosis and high fatality are still the challenges because of metastasis. Therefore, it is promising to clarify the detailed molecular mechanisms of CRC metastasis and provide potential therapeutic targets. Accumulating evidences indicate that long noncoding RNAs (lncRNAs) play important roles in cancer progression including CRC [3]. Thus, our study will investigate the function of lncRNA in CRC metastasis.
lncRNA is a type of noncoding RNA with more than 200 nucleotides lacking protein coding ability [4]. lncRNA exhibits multiple functions including chromatin modification, gene transcription, and posttranscription regulation [5]. The lncRNA urothelial cancer-associated 1 (UCA1) was first identified from bladder cancer cell line BLZ-211. It locates at chromosome 19p13.12 with around 2280 nucleotides [6]. UCA1 functions as a tumorigenic lncRNA in different cancers such as bladder cancer [7], breast cancer [8], lung cancer [9], melanoma [10], and colorectal cancer [11]. Furthermore, UCA1 is considered as a biomarker indicating poor prognosis of patients, in which metastasis and drug resistance occur [12]. Previous studies have shown that UCA1 promotes CRC metastasis through different molecular mechanisms including the lncRNA-miRNA-mRNA regulatory axis. Nevertheless, it is still imperfect of the whole landscape of UCA1 mechanisms involving different miRNAs in CRC metastasis. Therefore, our study is aimed at demonstrating the detailed molecular mechanism of UCA1 related to miRNA and transcription regulation, which will contribute in and enrich the UCA1 regulation mechanism in CRC metastasis.
In lncRNA regulatory function, the competitive endogenous RNA (ceRNA) network is critical in describing the lncRNA mechanism [5]. In CRC tumorigenesis and progression, the ceRNA network exhibits key roles, in which the lncRNA/miRNA/mRNA axis is a highlight. Previous studies clarified that UCA1 is involved in ceRNA regulations. In ovarian cancer, UCA1 targets miR-143 directly through 3 ′ UTR binding, which causes the expression changes of FOSL2 [13]. Furthermore, UCA1 promotes prostate cancer progression through miR-143 targeting and forms a ceRNA relation with MYO6 [11]. Therefore, UCA1-mediated ceRNA networks are potential targets in preventing cancer progression. In this study, we validated the UCA1/miR-185/NOTCH3 regulatory axis in CRC.
2.1.1. Vectors and Transfection. The HES5 CDS region was obtained through PCR and was cloned into pcDNA3.1 (+) vector (pcD-HES5). The inserted sequence was verified by Sanger sequencing. In transfection, cells were seeded into 6well plates and cultivated overnight to over 75% confluence. Vectors, miRNAs, or siRNAs were transfected into cells by Lipofectamine RNAiMax (Thermo Fisher, USA) according to the provided instruction. In transfection, the final concentrations of vectors and RNAs were 1 μg/ml and 25 nM.

Quantitative
Real-Time PCR (Q-PCR). The High Pure RNA Isolation Kit (Roche, Switzerland) was used to extract total RNAs of treated cells. In reverse transcription, 500 ng total RNA was used to obtain cDNA. Q-PCR was conducted with SYBR® Premix Ex Taq™ (RR420A; TaKaRa) in the Roche LightCycler 480 system. The results were normalized to GAPDH and calculated by using the 2 −ΔΔCt value. The primers used in q-PCR analysis were listed in Table 1. 2.1.3. Western Blot. Total protein was isolated from cells after different treatments by using RIPA lysis buffer. Protein concentrations of each group were quantified with BCA kit based on instructions. Total 30 μg protein was loaded on 10% sodium dodecyl sulfate polyacrylamide gel (SDS-PAGE). The electrophoresis conditions have a constant voltage of 120 V for 120 min. Then, protein was transferred to polyvinylidene difluoride membrane from gel (300 mA, 90 min). Membranes were blocked by 5% milk in 1x TBST and then incubated with primary antibodies CDH1 (1 : 1000, Life, USA), VIM (1 : 1000, Santacruz, USA), NOTCH3 (1 : 1000, Abcam, USA), and β-actin (1 : 2000, Abcam, USA) overnight at 4°C. Next, secondary antibody in 5% milk was incubated with membranes at room temperature for 1 h. The LI-COR Odyssey Imaging System was used for blot imaging.

Migration Assay.
For migration ability measurement, wound healing assay was employed. Cells were seeded into 6-well plates and cultivated overnight. Cell surface wound was generated by scratching cells with a 10 μl tip. Cell debris was removed by prewarmed PBS and then cultivated for 1 h. Wound images were taken after 1 h cultivation and recorded as 0 h timepoint. Cells were continuously cultivated for 24 h, and then images were taken, recorded as a terminal timepoint. The migration rate was calculated with the difference of distances between two timepoints and normalized to the control group. The relative migration ability was calculated by comparing the migration rate of each group to the control group.
2.1.5. Transwell Assay. To detect cell invasion ability, transwell assay was conducted. After different treatments, cells were seeded into the upper chamber with 8 μm membrane coated with Matrigel at the density of 5 × 10 4 cells/chamber. Fresh media without FBS were filled in upper chambers while media with 10% FBS were filled in lower chambers. After 48 h cultivation, chambers were fixed with methanol for 15 min and then stained with 0.5% crystal violet for     2.1.8. ChIP-qChIP (Q-ChIP) Analysis. DLD-1 cells were used to confirm the transcription factor target. The crosslinking process and chromatin immunoprecipitation were performed according to the instruction provided in the QuikChIP(TM) Kit (Novus, USA). Q-PCR was conducted followed by ChIP. The HES5 antibody (ab19411, Abcam, USA) used for q-ChIP was diluted at 1 : 100. AchR was served as a negative control according to the previous study [20]. The primers used for q-ChIP were listed in Table 2.
2.1.9. Statistics. SPSS 21.0 was used to calculate all the values (means ± standard error of the mean). Statistical analyses were analyzed with Student's t-test in a two-group comparison. The one-way analysis of variance (ANOVA) test was used to verify the significance among experimental groups. The statistical significance was p < 0:05.

Results
3.1. Long Noncoding RNA UCA1 Highly Expressed in Colorectal Cancer. By analyzing the TCGA-COAD datasets, UCA1 was found to be highly expressed in samples with tumor burden compared to that in the normal group (Figure 1(a)). Furthermore, we analyzed three independent CRC cohorts (GSE21510, GSE37364, and GSE39582) obtained from GEO database. Results indicated that UCA1 was also highly expressed in CRC tumor tissues compared to normal samples (Figures 1(b)-1(d)). In addition, UCA1 expression was positively correlated to CRC advanced stages (Figure 1(e)). Next, we tested several CRC cell lines to validate UCA1 expression. Q-PCR analysis revealed that UCA1 expression was higher in CRC cell lines, especially in SW480 cell. Therefore, UCA1 is a tumorigenic lncRNA promoting CRC progression potentially.

Silencing UCA1 Inhibited Colorectal Cancer Cell
Proliferation and Metastasis. Since UCA1 had a significantly higher level in CRC cell lines, the function of UCA1 was studied. In SW480 cell, UCA1 was knocked down by the designed specific siRNA pool (Figure 2(a)). MTT assay revealed that UCA1 knocking down remarkably reduced proliferation rates of SW480 and DLD-1 cells (Figure 2(b)). Next, several epithelial-mesenchymal transition (EMT) markers were tested in SW480 and DLD-1 cells when UCA1 was silenced. Both q-PCR and Western blot results indicated that epithelial marker CDH1 (E-cadherin) was induced while mesenchymal marker VIM was inhibited by UCA1 siRNA (Figures 2(c) and 2(d)), suggesting that UCA1 promotes the EMT process in CRC cells. In addition, the migration ability of SW480 and DLD-1 cells was repressed by UCA1 siRNA compared to that of the negative control group (Figure 2(e)).
The similar results were also observed in Boden chamber transwell assay, which illustrated that UCA1 silencing caused a less invasive ability in CRC cells (Figure 2(f)). The above results suggest that UCA1 promotes proliferation and metastasis of CRC cells.

Gastroenterology Research and Practice
transfected into SW480 cells to achieve ectopic expression (Figure 3(b)). Ectopic expression of miR-185 significantly reduced the UCA1 level compared to that of the negative control group, validated by q-PCR analysis (Figure 3(c)). Meanwhile, q-PCR and Western blot analysis showed that CDH1 was induced but VIM was suppressed by ectopically expressed miR-185 (Figures 3(d) and 3(e)). Furthermore, the basal level of miR-185 was also observed to be downregulated in the indicated CRC cell lines (Figure 3(f)), suggesting that miR-185 is a tumor-suppressor miRNA. Dual luciferase assay validated that UCA1 was a direct target of miR-185 (Figure 3(g)). Interestingly, silencing UCA1 induced miR-185 expression, in which ectopic UCA1 expression repressed miR-185 expression (Figure 3(h)), which implied that UCA1 functions as a sponge of miR-185. In summary, UCA1 is a direct target of miR-185. UCA1 is a sponge of miR-185, which affects miR-185 expression.

UCA1 Modulated the miR-185/NOTCH3 Regulatory
Axis. Since the ceRNA network is important for lncRNA function and regulation, here, we also investigated the potential gene which is possibly regulated by miR-185. Bioinformatics analysis revealed that NOTCH3 was a potential direct target of miR-185 (Figure 4(a)). In TCGA-COAD datasets, NOTCH3 was highly expressed in tumor tissues compared to those of the normal group (Figure 4(b)). Highly expressed NOTCH3 indicated advanced tumor stages and a   (Figures 4(c) and 4(d)). Ectopic miR-185 expression repressed NOTCH3 mRNA and protein levels in SW480 cell (Figures 4(e) and 4(f)). The direct target between miR-185 and NOTCH3 was validated by dual luciferase assay since miR-185 inhibited the luciferase activity of the pGL3 reporter vector containing the NOTCH3 3 ′ UTR wild-type sequence (Figure 4(g)). However, silencing NOTCH3 slightly changes the miR-185 expression level (Figure 4(h)), suggesting that NOTCH3 is under a one-way regulation by miR-185. Furthermore, silencing UCA1 in SW480 caused a lower NOTCH3 expression level compared to that of the negative control group, while miR-185  (Figures 4(i) and 4(j)), indicating that UCA1 regulates NOTCH3 via miR-185 mediation. Indeed, both migration ability and invasion ability of SW480 cells were reversed by miR-185 antagomir when UCA1 was silenced (Figures 4(k) and 4(l)), implying that UCA1 promotes SW480 cell metastasis through regulating the miR-185/NOTCH3 target. Collectively, UCA1 promotes CRC metastasis by upregulating NOTCH3, which is mediated by miR-185 (Figure 4(m)).

HES5 Was Validated as a Transcription Factor of UCA1.
To further clarify UCA1 regulation in CRC, the potential transcription factor was investigated. Here, the 2000 bp promoter region of UCA1 was obtained from the UCSC genome browser and scanned in JASPAR to discover potential transcription factors. As shown in Figure 5(a), HES5 was predicted as a putative transcription factor of UCA1 by binding with the indicated typical motif. From the analysis in TCGA-COAD datasets, HES5 expression was positively correlated to UCA1 expression in CRC ( Figure 5(b)). In HEK-293T cells, overexpression of HES5 induced the UCA1 level (Figures 5(c) and 5(d)). To confirm the target between HES5 and UCA1, we performed dual-luciferase assay on the pGL vector containing the UCA1 promoter region mutated following the indicated strategy ( Figure 5(e)). The mutant binding motif of UCA1 exhibited a lower luciferase activity compared to that of the wild-type group, suggesting that HES5 regulates UCA1 expression through activating transcription ( Figure 5(f)). In addition, q-ChIP analysis demonstrated that the UCA1 promoter region was enriched by HES5 precipitation, validating the direct target between UCA1 and HES5 ( Figure 5(g)). Furthermore, overexpression HES5 repressed miR-185 and induced NOTCH3 expression in HEK-293T cells ( Figure 5(h)). In summary, UCA1 is a direct target of HES5. NOTCH3 is regulated by HES5 through the HES5/UCA1/mIR-185 axis.

Discussion
Colorectal cancer is the third most commonly diagnosed cancer around the world, which caused over 800000 deaths in 2018 [2]. One of the challenges in CRC clinical treatment is metastasis which results in poor prognosis and high recurrence rate. Recent studies indicate that lncRNA exerts regulatory functions in CRC progression. In previous studies, UCA1 has been identified to serve as a tumorigenic lncRNA in different cancer types such as bladder cancer, breast cancer, lung cancer, melanoma, and colorectal cancer [6][7][8][9][10][11]. Furthermore, UCA1 is considered as a biomarker indicating a poor prognosis, metastasis, and drug resistance [21]. Here, UCA1 was confirmed to be highly expressed in colorectal cancer. Moreover, the UCA1 expression level was positively correlated to tumor stages. Silencing UCA1 showed an inhibitory effect on cell proliferation and metastasis. Therefore, UCA1 is a oncogenic lncRNA promoting CRC progression.
In lncRNA regulation, the competitive endogenous RNA (ceRNA) network is critical in describing the lncRNA function mechanisms [5]. In current research, NOTCH3, the member of the NOTCH signaling pathway, was validated as a direct target of miR-185. In CRC, NOTCH3 has been identified to drive tumor progression. Increased NOTCH3 was observed in CRC and associated with accelerated tumor growth [29]. Here, NOTCH3 shared the same miR-185 target with UCA1. In addition, NOTCH3 was indued by UCA1 through miR-185 mediation. Therefore, CRC is promoted by the UCA1/miR-185/NOTCH3 regulatory axis. Besides the ceRNA regulatory network, we also performed analysis on the potential transcription factor of UCA1. Previous studies have identified that UCA1 is regulated by several transcription factors such as Ets-2, C/EBPα, TAZ/YAP/TEAD, HIF-1α, SATB1, and CAPERα/TBX3 [14][15][16][17][18][19]. In this study, HES5 was characterized as a transcription factor of UCA1. In the UCA1 promoter region, the binding motif of TGGCAGGTGCCT was confirmed to be critical for HES5 transcription ability. Furthermore, NOTCH3 was also induced by HES5 via the HES5/UCA1/ miR-185/NOTCH3 regulatory axis. Therefore, UCA1 plays a central role in mediating HES5, miR-185, and NOTCH3 function in CRC progression.
In conclusion, UCA1 promotes CRC metastasis and proliferation. UCA1 is a direct target of HES5. Both UCA1 and NOTCH3 are direct targets of miR-185, through which NOTCH3 is induced by UCA1, summarized in Figure 5(i).

Data Availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.