The lncRNA CCAT1 upregulates TGFβR1 via sponging miR-490-3p to promote TGFβ1-induced EMT of ovarian cancer cells

Ovarian cancer cells (SKOV3 and CaOV3) and 293T cell were purchased from ATCC and cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Hyclone) supplemented with 10% fetal bovine serum and 100 U/ml penicillin/streptomycin at 37 °C, 5% CO₂. TGFβ1 was purchased from R and D systems and was used to induce EMT in SKOV3 and CaOV3 (10 ng/ml) cells for the indicated time periods.

The TGFβR1 cDNA was subcloned into pCDNA3.1 vector which was transfected into cells using lipofectamine 2000 according to the instruction. For miR-490-3p mimics or miR-490-3p inhibitor transfection, we used LipofectamineVR LTX with PlusTM Reagent (Life Technologies) to transfect them into cells. All siRNAs, miR-490-3p mimics and miR-490-3p inhibitor were synthesized by GenePharma. The sequences are as follows:

miR-490-3p mimics: (sense) 5′-CAACCUGGAGGACUCCAUGCUC-3′; (antisense) 5′-GCAUGGAGUCCUCCAGGUUGUU-3′;

miR-490-3p inhibitor: 5′-CAGCAUGGAGUCCUCCAGGUUG-3′.

A cohort of 25 ovarian tumor tissues and adjacent normal ovarian tissue samples were obtained from patients aged 25–55 undergoing wedge biopsy of the ovaries or adnexectomy due to myoma or adenomyosis, between 2016.6 and 2017.5. No patients had received chemotherapy or radiotherapy prior to surgery. Consent from all patients were obtained. Ovarian cancer was validated by histological examination in all cases according to World Health Organization criteria. Ovarian cancer and normal ovarian tissue specimens excised surgically from patients were immediately snap-frozen and stored in liquid nitrogen until use. This experiment was approved by ethic committee of the 2nd Affiliated Hospital of Harbin Medical University, and the tissues were acquired with the consent of patients.

Migration of cells were measured by a wound healing assay in vitro. Briefly, 2 × 10⁵ SKOV3 or CaOV3 cells were seeded onto 6-well plates, with either sh-CCAT1 or sh-NC, and incubated in appropriate complete culture medium for 16 h under normoxic conditions at 37 °C. The monolayer was scratched and incubated in medium without FBS for 24 h. The wound width was measured after 24 h. Three different locations were visualized and photographed under inverted microscope.

Invasion assay was performed using chambers with 8.0-μm pore membranes (Millipore). Ovarian cancer cells (1 × 10⁵ cells) were resuspended in 200 µl of FBS-free medium, and then seeded into the top chamber with Matrigel-coated membrane. Next, 500 µl medium with 10% FBS was added to the bottom chamber as a chemoattractant. After 48 h of incubation, the invaded cells were fixed, stained with 0.005% crystal violet, and counted under the inverted microscope.

CCAT1 or TGFβR1 mutant was generated using site-directed mutagenesis. Then, the sequence of the CCAT1 or TGFβR1 was cloned into the firefly luciferase-expressing vector pGL3-luciferase plasmid. As for luciferase assay, the SKOV3 or CaOV3 cells were seeded for triplicates in 24-well plates at the day before transfection, and co-transfected with the CCAT1 or TGFβR1 reporter vector and miR-490-3p. Then, the cells were harvested and lysed, and the luciferase activities were assayed using the Dual-Luciferase Reporter System (Promega). Three independent experiments were performed.

The cells were harvested and washed with PBS buffer, then lysed by 1 × SDS loading buffer. The lysates were boiled at 100 °C for 5 min. The samples were centrifuged at 10,000 rpm for 1 min. Around 50 μg of total proteins was loaded onto SDS-PAGE gel and resolved. After that, the proteins were transferred to PVDF membrane at 300 mA for 1.5 h. The membrane was blocked with 5% non-fat milk in 1× TBST for 1 h at room temperature, the membrane was then incubated with primary antibodies at 4 °C overnight. The following day, the membrane was washed with 1× TBST for three times, 5 min each time. The membrane was incubated with secondary antibodies at room temperature for 1 h. Finally, the membrane was incubated with ECL solution and then exposed. The following antibodies were used: anti-TGFβR1 (cell signaling technology, USA), anti-E-cadherin (cell signaling technology, USA), anti-N-cadherin (cell signaling technology, USA), anti-Claudin (cell signaling technology, USA), anti-β-actin (Proteintech, USA), anti-MMP9 (Abcam, USA), anti-GAPDH (Proteintech, USA).

We extracted the RNA using Trizol method. Cells were lysed by Trizol buffer and then add chloroform to the mixture. The sample was centrifuged at 12,000 rpm for 10 min and transferred to new EP tube, mixed with equivalent volume of isopropanol, next, the resultant was centrifuged at 12,000 rpm for 10 min. Removing the supernatant and add 75% ethanol to wash the pellet and centrifuge. Finally, discard the ethanol and dry the pellet, use 20–30 μl Rnase-free H₂O to resolve the RNA.

For reverse transcription, about 1 μg of total RNA was used for reverse transcription according to manufacturer instruction (TAKARA PrimeScript Kit). The expression of miR-490-3p was quantified by TaqMan miRNA assays (Applied Biosystems, Foster City, CA, USA).

For real time PCR, we used SYBR as probe dye and detected the signal, the GAPDH and U6 were used as internal control. The following primers were used:

miR-490-3p-QPCR-F: 5′-CGCAACCTGGAGGACTCC-3′

miR-490-3p-QPCR-R: 5′-CGGCCCAGTGTTCAGACTAC-3′

TGFβR1-QPCR-F: 5′-GTGACAGATGGGCTCTGCTT-3′

TGFβR1-QPCR-R: 5′-AGGGCCAGTAGTTGGAAGTT-3′

E-cadherin-QPCR-F: 5′-TCACATCCTACACTGCCCAG-3;

E-cadherin-QPCR-R: 5′-AGTGTCCCTGTTCCAGTAGC-3′,

N-cadherin-QPCR-F: 5′-AGGGGACCTTTTCCTCAAGA-3′;

N-cadherin-QPCR-R: 5′-TCAAATGAAACCGGGCTATC-3′,

MMP9-QPCR-F: 5′-TTCCAAACCTTTGAGGGCGA-3′;

MMP9-QPCR-R:5′-CTGTACACGCGAGTGAAGGT-3′,

GAPDH-QPCR-F: 5′-AGCCCAAGATGCCCTTCAGT-3′;

GAPDH-QPCR-R: 5′-AGCCCAAGATGCCCTTCAGT-3′,

U6-QPCR-F: 5′-CTCGCTTCGGCAGCACA-3′;

U6-QPCR-R: 5′-AACGCTTCACGAATTTGCGT-3′.

Each experiment was performed for three times, all values were presented as mean ± SD, comparison of two groups were performed using the two-tailed unpaired student’s t-test. One-way ANOVA was used for comparison among multiple groups and multiple comparisons were further performed using post hoc Turkey test. *P < 0.05 were considered statistically significant (*P < 0.05, **P < 0.01, and ***P < 0.001).

To characterize the role of CCAT1 in TGFβ1-induced EMT of ovarian cancer cells, we first determined the expression level of CCAT1 in SKOV3 and CaOV3 cells when treated with 10 ng/ml TGFβ1 for 48 h. The result showed CCAT1 was upregulated by TGFβ1 (Fig. 1a). Next, we generated stable CCAT1-depleted SKOV3 and CaOV3 cells by shRNAs approach (Fig. 1b). And then we employed wound healing method to examine the migration of CCAT1-depleted ovarian cancer cells SKOV3 and CaOV3 in the presence of TGFβ1(10 ng/ml). The results showed that scramble (sh-NC) cells exhibited no migration difference compared to control, whereas CCAT1 knockdown (sh-CCAT1) cell significantly compromised migration (reduced by ~ 40–50%) of ovarian cancer cells induced by TGFβ1 in relative to sh-NC group (Fig. 1c, d). In addition, we sought to determine whether TGFβ1-induced invasion of ovarian cells were regulated by CCAT1 loss. Transwell matrix penetration assay demonstrated that knockdown of CCAT1 reduced the number of invasive ovarian cancer cells (by ~ 50%) in the presence of TGFβ1 compared to sh-NC group (Fig. 1e). To further investigate the mechanism underlying TGFβ1-induced invasion and migration of ovarian cancer cells regulated by CCAT1, we examined the expression of EMT-associated genes in control and sh-CCAT1 ovarian cells. RT-qPCR and western blot analyses showed that CCAT1 knockdown markedly attenuated TGFβ1-induced expression of vimentin, N-cadherin and MMP9, on the other hand, CCAT1 knockdown enhanced TGFβ1-induced expression of E-cadherin and Claudin (Fig. 1f, g). Taken together, these results revealed that CCAT1 loss led to remarkably attenuated EMT of ovarian cancer cells treated with TGFβ1.

In order to clarify the mechanism by which CCAT1 loss compromised TGFβ1-induced EMT, we hypothesized that TGFβ1 cognate receptor, TGFβR1, might be regulated by CCAT1. Interestingly, we observed that TGFβR1 mRNA level was diminished by about 50–70% in CCAT1-null cells compared to shNC cells (Fig. 2a). In agreement with this, protein level of TGFβR1 was also downregulated by CCAT1 depletion in both SKOV3 and CaOV3 cells (Fig. 2b). In sum, these data indicated that CCAT1 played its roles in TGFβ1-induced EMT of ovarian tumor through enhancing TGFβR1 expression.

It has been reported that miR-490-3p has been implicated in ovarian tumor invasion and metastasis [31]. We reasoned that CCAT1 might exert its role by regulating miR-490-3p in ovarian cancer cells. Bioinformatic analysis revealed that CCAT1 could directly target miR-490-3p by matching with sequence at 3′-terminus (Fig. 3a). Furthermore, luciferase reporter assay was used to confirm that CCAT1 directly interacted with miR-490-3p. We found that miR-490-3p overexpression only decreased wildtype CCAT1-fused luciferase activity, not CCAT1 mutant (Fig. 3b). Notably, miR-490-3p level was substantially upregulated by CCAT1 knockdown in SKOV3 and CaOV3 cells (Fig. 3c). Moreover, miR-490-3p mimics suppressed CCAT1 expression, and instead miR-490-3p inhibitor enhanced CCAT1 expression of SKOV3 and CaOV3 cells (Fig. 3d). RT-qPCR and western blot analyses showed that miR-490-3p overexpression augmented CCAT1 depletion-induced TGFβR1 downregulation, while miR-490-3p inhibitor greatly restored the expression level of TGFβR1 (Fig. 3e, f). Together, these results suggested that lncRNA CCAT1 loss downregulated TGFβR1 expression via directly targeting miR-490-3p.

As CCAT1 was shown to decrease miR-490-3p level, then upregulating TGFβR1 expression in SKOV3 and CaOV3 cells; therefore, we sought to assess whether miR-490-3p was essential for CCAT1-mediated tumor phenotypes of ovarian cancer cells. Wound healing assay showed that CCAT1 knockdown alone impaired (decreased by 52%) the ability of TGFβ1 to promote cells migration. Transfection of both miR-490-3p mimics and CCAT1 shRNA markedly attenuated (dropped by ~ 88%) migration of ovarian cancer cells relative to shCCAT1 alone, while co-transfection of miR-490-3p inhibitor and CCAT1 shRNA into cells exhibited more robust migration (increased by ~ 72%) than shCCAT1-expressing cells (Fig. 4a). Besides, we found that miR-490-3p mimics potentiated shCCAT1-inhibited invasiveness (the invasion cell number decreased by 60%) and instead miR-490-3p inhibitor attenuated (the invasived cell number increased by 61%) shCCAT1-inhibited invasiveness of SKOV3 and CaOV3 cells (Fig. 4b). As for EMT-associated markers, RT-qPCR and westernblot revealed that upregulation of E-cadherin and Claudin by CCAT1 loss was enhanced by miR-490-3p mimics and attenuated by miR-490-3p inhibitor in SKOV3 and CaOV3 cells, inverse in the expression of vimentin, N-cadherin and MMP9 (Fig. 4c, d). Collectively, these findings indicated that miR-490-3p was essential for TGFβ1-induced EMT of ovarian cancer cells regulated by CCAT1 depletion.

To further explore how miR-490-3p affected CCAT1 loss-induced TGFβR1 expression, we inferred that miR-490-3p probably targeted TGFβR1 and regulated its expression. The bioinformatic analyses revealed that miR-490-3p could target 3′-UTR of TGFβR1 mRNA (Fig. 5a). Luciferase reporter assay demonstrated that miR-490-3p mimics effectively inhibited the activity in wildtype 3′-UTR of TGFβR1 cells, not the mutant, suggesting miR-490-3p played its role via directly binding 3′-UTR of TGFβR1 (Fig. 5b).

To determine the biological function of miR-490-3p-induced TGFβR1 downregulation, we overexpressed miR-490-3p mimics alone or in combination with TGFβR1. The RT-qPCR and westernblot analyses showed TGFβR1 was reduced by miR-490-3p mimics overexpression; however, TGFβR1 level increased when the cells overexpressing miR-490-3p and TGFβR1 (Fig. 5c, d). Next, we observed that miR-490-3p overexpression greatly attenuated migration of ovarian cancer cells SKOV3 and CaOV3, whereas exogenous expression of miR-490-3p and TGFβR1 rescued TGFβ1-induced migration change of the cells (Fig. 5e, f). Similarly, miR-490-3p mimics caused remarkably decreased invasion of ovarian cancer cells compared to negative control, and transfection of miR-490-3p plus TGFβR1 could enhance the invasiveness of ovarian cancer cells comparable to control (Fig. 5f). Finally, we detected EMT-associated markers after overexpression of miR-490-3p and TGFβR1. Our results showed miR-490-3p inhibited TGFβ1-induced expression of vimentin, N-cadherin and MMP9, instead, upregulated TGFβ1-induced expression of E-cadherin and Claudin. More importantly, overexpression of TGFβR1 reverted miR-490-3p-mediated regulation of EMT-related genes in ovarian cancer cells (Fig. 5g).

To determine the clinical association of CCAT1 and miR-490-3p expression with progression of ovarian cancer, we examined CCAT1 and miR-490-3p expression level of ovarian tumors (n = 25) and adjacent normal tissues (n = 25) by RT-qPCR approach. CCAT1 level was higher (~ 2.6 folds) in tumors than in normal tissues; however, miR-490-3p level was lower (~ 65%) in tumors compared to normal tissues (Fig. 6a, b). In addition, the bioinformatic analysis showed negative association between CCAT1 and miR-490-3p expression in ovarian tumors (r² = 0.8579, p < 0.01) (Fig. 6c). To summarize, our data indicated that CCAT1, clinically, might promote ovarian cancer via inhibiting expression of miR-490-3p.