miR-4454 up-regulated by HPV16 E6/E7 promotes invasion and migration by targeting ABHD2/NUDT21 in cervical cancer

Abstract The abnormal expression of HPV16 E6/E7 activates oncogenes and/or inactivates tumor suppressor genes, resulting in the selective growth and malignant transformation of cancer cells. miR-4454 was selected by sequencing due to its abnormal high expression in HPV16 E6/E7 positive CaSki cell compared with HPV16 E6/E7 negative C33A cell. Overexpression of miR-4454 enhances cervical cancer cell invasion and migration. ABHD2 and NUDT21 are identified as a target gene of miR-4454.The effects of ABHD2 and NUDT21 on migration and invasion of CaSki and C33A cells were determined. The dual luciferase and RT-qPCR assays confirmed that miR-4454 might regulate its targets ABHD2 and NUDT21 to promote the proliferation, invasion and migration, whereas, inhibit the apoptosis in CaSki and C33A cells.


Introduction
The incidence of cervical cancer ranks the second among all female tumors with nearly 500,000 women worldwide diagnosed with cervical cancer every year, nearly half of whom ultimately die of the disease [1,2]. It has been confirmed that cervical cancer is associated with high-risk human papillomavirus (HPV) infection, with HPV16 and HPV18 being the most common infectious agents found in 99.7% of cases of cervical squamous cell carcinoma and 94-100% of cases of cervical adenocarcinoma [1,[3][4][5][6]. China's morbidity and mortality associated with cervical cancer account for nearly one-quarter of the world's total, representing a serious threat to women's health [1].
In HPV-positive cervical cancers, all of the malignant cells contain at least one copy of the viral genome with transcriptional activity [7,8]. Recent studies have shown that non-coding microRNAs (miRNAs) are also regulated by E6 and E7 at the transcriptional and post-transcriptional level [6,[9][10][11][12][13].
In the present study, we investigated the molecular mechanism of HPV16 E6/E7-regulating miRNA in cervical cancer progress. We used HPV16 E6/E7-positive human cervical cancer cell line CaSki as the major cell model. The aim of the present study was to elucidate the biological functions and potential target of miR-4454 in cervical cancer cells and provide evidence that miR-4454 may serve a potential candidate for the clinical treatment for cervical cancer.
The differentially expressed gene was selected by sequencing siRNA sequences are synthesized by Sangon Biotech (Shanghai) Co., Ltd. sh-E6: F:5´-GGAGCGACCCAGAAAGUUATT-3´,R:5´-UAACUUUCUGGGUCGCUCCTT-3´. Then, the interfering shRNA was screened and transfected into CaSki cell. Finally, CaSki cell and siRNA cell were sent to the personalbio (Shanghai, China) for sequencing and screening for differential miRNA. Sample processing: after total RNA was extracted, the RNA quality was detected. Then, the small RNA library was constructed by IlluminatheTruSeq Small RNA Sample prep Kit. The enriched library was amplified by PCR and purified by gel electrophoresis. The library quality was checked by Agilent High Sensitivity DNA Kit using Agilent 2100 Bioanalyzer. After testing the qualified library, the library was quantified by Quant-iT Picogreen dsDNA Assay Kit. Finally, single-stranded library was used as template to bridge PCR amplification, and primers were annealed and sequenced. Data analysis: the original data are de-joined and filtered, and the filtered sequence is de-reprocessed. The deleted small RNA sequence is annotated on the basis of the reference genome and the corresponding annotation abundance is given. After statistical analysis, the differentially expressed miRNAs were clustered and predicted, and the predicted target genes were enriched and analyzed.

Transwell assay
A transwell assay was performed to investigate the migration and invasion of C33A and CaSki cells. The cells were harvested for 24 h before cell suspension, and then the cells were digested to be used as signal cells at a 5 × 10 5 /ml concentration. The cells were suspended using a serum-free medium containing bovine serum albumin, and 100 μl of the cell suspension was added to the transwell chamber. The 24-well plate chamber was filled with 600 μl of medium per well containing 10% FBS, and then the plate was incubated for 24 h at 37 • C. After incubation, tweezers were used to carefully remove the chamber. The upper chamber fluid was drained and approximately 800 μl of methanol was added to each well for fixation for 30 min at room temperature. The chamber was taken out and dyed with 1 ml of 0.5% Crystal Violet solution for 30 min. The cells were observed using a microscope (TS100-F, Nikon, Tokyo, Japan). The experimental protocol for the invasion assay of C33A and CaSki cells was similar to that of the migration assay, except for the addition of Matrigel (BD, NJ, U.S.A.) and the serum-free medium diluted at a 1:8 ratio at 4 • C. The plate was coated with the upper chamber surface of the bottom membrane in the transwell membrane. The Matrigel was polymerized to a gel at 37 • C for 30 min.

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium (MTT) assay
The logarithmic-phase cells from cell subculture were plated in 96-well plates with 180 μl (5 × 10 3 cells) per well, and the plates were incubated at 37 • C overnight. The effect of miR-4454, ABHD2, and NUDT21 on the proliferation of HPV16-positive CaSki cells and HPV16-negative C33A cells was detected at 24, 48 and 72 h. After the addition of 20 μl MTT (Beyotime, Shanghai, China) solution (5 mg/ml) to the cell culture plates of different groups, the plates were incubated for 4 h. After the supernatant was removed, 150 μl of dimethyl sulfoxide (Beyotime) solution was added to each well and mixed, and the absorbance value of each well was measured at 490 nm.

Flow cytometry assay
After the C33A and CaSki cells of the different treatment groups were collected, they were centrifuged at 1000 ×g for 5 min, and then suspended by phosphate-buffered saline (PBS) and counted. Approximately 2 × 10 6 cells/ml were suspended with binding buffer, and then centrifuged at 300 × g for 10 min [14]. Subsequently, 5 μl of Annexin V-FITC (Solarbio, Beijing, China) was added to 100 μl of the cell suspension and mixed at room temperature for 10 min, followed by the addition of 5 μl propidium iodide and mixed at room temperature for 5 min. Finally, the PBS was added to reach a constant volume (500 μl), and cell apoptosis was detected by flow cytometry within 1 h.

Reverse transcription-quantitative PCR (RT-qPCR) assay
CaSki and C33A cells were harvested and the total RNA was extracted using Trizol reagent (Ambion, TX, U.S.A.). The total RNA was reverse-transcribed to cDNA by PrimeScript™ RT Master Mix (TaKaRa, Dalian, China) according to the manufacturer's instructions. The mRNA was quantified by SYBR ® Premix Ex Taq™ II kit (TaKaRa, Dalian, China). The following primers were designed by Nanjing Genscript Biological Technology Co., Ltd.

Statistical analysis
All data were analyzed by the data analysis software SPSS19.0 (SPSS, IL, U.S.A.), and are presented as the mean + − standard deviation. Data were analyzed by one-way analysis of variance followed by Dunnett's or Duncan's test. A P-value less than 0.05 was considered statistically significant.

miR-4454 targets ABHD2 and NUDT21
The interactions between miR-4454 and ABHD2/NUDT21 were evaluated in 293 cells by a dual luciferase assay ( Figure 2E,H). In the ABHD2 wild-type (WT) group, R/F was significantly increased in the miR-4454-transfected cells compared with the control (P<0.001) and miR-4454 negative control (NC)-transfected cells (P<0.01), while there was no significant difference in ABHD2 mutated (Mut) group (P<0.001). Similarly, in the NUDT21 WT group, F/R was significantly decreased in the miR-4454-transfected cells compared with the control (P<0.01) and miR-4454 NC-transfected cells (P<0.01), while there was no significant difference in NUDT21 mutated (Mut) group (P<0.001). These results suggested that miR-4454 targets both ABHD2 and NUDT21. To validate these results, an RT-qPCR assay and Western blot were used to detect the expression of ABHD2 and NUDT21 mRNA levels and protein levels in CaSki and C33A cells ( Figure 2F

ABHD2 and NUDT21 suppresses the invasion and migration of HPV16 E6/E7-positive CaSki cells
Since we demonstrated that ABHD2 and NUDT21 are targets of miR-4454, we next explored the effects of ABHD2 and NUDT21 on the invasion and migration of HPV16-positive cells (CaSki) and HPV16-negative cells (C33A). The RT-qPCR assay confirmed the successful construction of pre-ABHD2, sh-ABHD2, pre-NUDT21, and sh-NUDT21 plasmids. The expression of ABHD2 was up-regulated in pre-ABHD2 group and was down-regulated in sh-ABHD2 group ( Figure 3A). In addition, the expression of NUDT21 was up-regulated in pre-NUDT21 group and down-regulated in sh-NUDT21 group ( Figure 3B). The invasion and migration of the transfected cells were then detected by a transwell assay. Our results showed that pre-ABHD2 markedly decreased the invasion ( Figure 3C 05 and P<0.01, respectively). These data indicated that ABHD2 can decrease the invasion and migration of HPV16 cells (C33A and CaSki), or in other words, a lack of ABHD2 can increase the invasion and migration of HPV16 cells (C33A and CaSki). Moreover, compared with C33A cells, the invasion and migration were promoted to a greater extent in CaSki cells (P<0.05). Similar results were obtained for pre-NUDT21 and sh-NUDT21 (Figure 4). Overall, these results suggest that HPV16 E6/E7 can affect cell invasion

miR-4454 regulates invasion and migration in HPV16 cells through functional target ABHD2/NUDT21 in vitro
The invasion and migration of the transfected cells were detected by a transwell assay. Our results showed that transfected miR-4454 mimics and miR-4454 mimics with sh-ABHD2/NUDT21 markedly promote the invasion ( Figure

Discussion
High-risk HPV16 E6/E7 can regulate the tumor suppressor p53 and pRb, respectively [15][16][17]. This activation consequently promotes the expression of c-Myc, and cell cycle protein A and E (CyclinA/E), which is beneficial to the synthesis of DNA and induces the progression of cells from G1 to S, thus promoting cell proliferation [17]. miR-4454 is involved in promoting the inflammation, catabolism, and cell death activity of chondrocytes in spinal osteoarthritis (facet joints) [18]. A previous study showed that miR-4454 was overexpressed in the bladder tumor, WBCs, and urine [19]. In tumor necrosis factor-alpha-stimulated human cervical cancer HeLa cells, miR-4454 was identified as an NF-κB-targeting miRNA, with roles in the anti-inflammatory, apoptosis, and other physiological processes of cervical cancer cells [19,20]. They suggested that miR-4454 may affect cancer cell migration, invasion, proliferation, and apoptosis. In the present study, we determined specific high expression of miR-4454 in HPV16 E6/E7 through a gene chip screening process. Therefore, we focused on the effects of miR-4454 and its downstream targets determined by querying the miRNAs target database on cervical cancer cells. Our data are in line with previous studies showing that miR-4454 is overexpressed in tumors, and is highly enriched in the HPV16 E6/E7 positive cervical cancer cell. Moreover, miR-4454 promoted the proliferation, invasion, and migration, and inhibited the apoptosis of CaSki cell compared with C33A cell.
The subsequent study was interested in the target of miR-4454 to further understand its effect on the progression of cervical cancer. ABHD2 and NUDT21 were selected as candidate targets of miR-4454 based on information from online databases. Yun et al. [21] described a novel mechanism by which the serine hydrolase ABHD2 regulates calcium transfer from the endoplasmic reticulum to the mitochondria to affect cell death (anoikis) resistance, which is related to malignant phenotypes and is a hallmark of cancer. Yamanoi et al. [22] showed that down-regulation of ABHD2 could promote a malignant phenotype and was associated with a poor prognosis for women with serous ovarian cancer. In addition, Obinata et al. [23] showed that the expression of ABHD2 was up-regulated by androgen in LNCaP and VCaP cells. Disruption of ABHD2 expression also promotes cell proliferation and migration of LNCaP cells. According to these results, ABHD2 was proposed as a novel target for the diagnosis and treatment of prostate cancer.
In our study, we demonstrated that ABHD2 is a target of carcinogenic miR-4454 and its expression is negatively correlated to that of miR-4454. Therefore, we supposed that suppression of ABHD2 would have a carcinogenic effect. The invasion and migration of HPV16 E6/E7-positive CaSki cell were increased to a greater extent under ABHD2 inhibition compared with the effects observed in HPV16-negative C33A cells, consistent with the previous study in prostate cancer [23]. The other study showed that knockout of NUDT21 promoted the proliferation, metastasis, and tumorigenesis of HCC cells, whereas enhanced the expression of NUDT21 had the opposite effect [24]. Our data also suggested that NUDT21 is negatively correlated to miR-4454, and knockdown of NUDT21 increased the invasion and migration of HPV16 E6/E7-positive CaSki cell compared with HPV16-negative C33A cell, indicating an effect in promoting the development of human cervical cancer.
In conclusion, miR-4454 was overexpressed in HPV16 positive cervical cancer cells, and HPV16 E6/E7 may affect cancer cell invasion and migration via miR-4454/ABHD2/NUDT21 axis in cervical cancer cells. In follow-up experiments, we will further explore the specific roles of ABHD2/NUDT21 in the development of HPV16-positive cervical cancer and its related mechanisms. This work is expected to provide new insights into the mechanism by which HPV16 E6/E7 regulate ABHD2/NUDT21 through miR-4454 to better understand the malignant transformation of cervical cancer.

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