CircHIPK3 regulates cardiac fibroblast proliferation, migration and phenotypic switching through the miR-152-3p/TGF-β2 axis under hypoxia

Animals

C57BL/6J neonatal mice (1–3 days old) were obtained and housed at Zunyi Medical University (Zunyi, China) and maintained under SPF conditions. All mice were lived with their mothers and were breastfed. The mice were exposed to light for 12 h a day. All experimental animals were used for the isolation of cardiac fibroblast. According to the review of animal experiment Ethics Committee of Zunyi Medical University (Approval No. ZMUER2018-2-177), the experimental scheme followed the animal welfare and ethical principles and met the requirements of ethical norms.

Culture and treatment of cardiac fibroblasts

Primary CFs were isolated and cultured according to the previously described and slightly improved (Chowdhury et al., 2017). Ventricles were obtained from 1 to 3 days old C57BL/6J neonatal mice. CFs were isolated by enzymatic dissociation methods using 0.08% trypsin and 0.1% collagenase type II (Sigma–Aldrich, Shanghai, China) in D-Hank’s balanced salt solution at a 37 °C water bath for 1 h. The cells then were filtered with a 200 mesh screen. Cells were cultured in Dulbecco’s modified Eagle’s medium (Gibco, Dublin, Ireland) supplemented with L-glutamine, fetal bovine serum and penicillin-streptomycin. After 1.5 h, CFs were obtained by removing the cell suspension, and then fresh medium was added for further culture.

Cardiac fibroblasts were subjected to hypoxia. The logarithmic growth cells were incubated in a 94% N₂, 5% CO₂, and 1% O₂ gas mixture incubator (Thermo, Waltham, MA, USA) at 37 °C for 6, 12, 24, and 48 h. Cells viability was assessed by Cell Counting Kit-8 (CCK-8) assay.

Reverse transcription-quantitative PCR

Total cellular RNA was isolated using TRIzol reagent (Life Technologies, Carlsbad, CA, USA). For circRNA isolation, 4U RNase R (Geneseed, Guangzhou, China) was added to one μg total RNA and incubated at 37 °C for 30 min. cDNA synthesis and RT-qPCR analysis were implemented using PrimeScript RT Master Mix and TB Green Premix Ex Taq II kit (Takara, Dalian, China). To quantify the relative expression levels of miRNA, total RNA was reversed to first-strand cDNA with miRNA 1st Strand cDNA Synthesis Kit (Sangon Biotech, Shanghai, China). GAPDH and U6 primers were used as normalization control for circRNA and miRNA expression, respectively. Relative quantification of mRNA expression after normalization of the transcript amount to the endogenous control was calculated using the 2^−ΔΔCt method (Tabatabaeian & Hojati, 2013).

5-Ethynyl-2′-deoxyuridine incorporation assay

Cell proliferation was analyzed with Cell-Light 5-Ethynyl-2′-deoxyuridine (EdU) DNA Cell Proliferation Kit (RiboBio, Guangzhou, China) according to previously reported (Shan et al., 2017). Briefly, approximately 5 × 10⁶/ml logarithmic growth cells were cultured in confocal dish and placed in an incubator. CFs were labeled with 10 mM EdU working solution for 4 h. Subsequently, the working solution was discarded, and added 4% paraformaldehyde to fix the cells at room temperature. Apollo dye solution was labeled proliferating cells for 30 min in the dark, and then Hoechst 33342 solution was stained cell nuclei. Images were captured by fluorescence microscopy (Olympus, Tokyo, Japan).

Cell cycle assay

Assessment of cell proliferation was carried out using a Cell Cycle Analysis Kit (4A Biotech, Beijing, China) by FCM. Briefly, CFs were fixed with 95% anhydrous ethanol and kept at 4 °C before DNA staining with PI/RNaseA dye solution for 30 min away from light at 37 °C. Samples were acquired with a flow cytometer (BD Biosciences, San Jose, CA, USA).

Cell wound scratch assay

Cell migration was assessed using traditional cell wound scratch assay. Briefly, CFs were seeded in six-well plates. A cell-free zone was gently formed on a cell monolayer with a 200 μl plastic pipette tip when the cells reached ~80% confluence, then cells were washed with PBS to remove scraped cell debris. Images were acquired by microscopy. Cell migration was determined by the width of the gaps; the shorter the gaps, the higher the migration capacity.

Western blot analysis

Western blot analysis was performed as described by us (Wang et al., 2019b). Briefly, proteins was extracted from CFs with RIPA lysis buffer (Biosharp, Hefie, China) and quantified with a BCA protein assay reagent (Beyotime, Haimen, China). Equal of protein lysates was loaded into per lane for SDS-PAGE detection before transferred to PVDF membranes (Millipore, Billerica, MA, USA). Next, all membranes were blocked with 5% nonfat milk solution in TBST buffer for 1 h at room temperature and then incubated with primary antibodies on a shaker at 4 °C overnight. The primary anti-bodies: Anti-Collagen I (Cell Signaling Technology, 91144, dilution: 1:1,000), Anti-Collagen III (Abcam, ab7778, dilution: 1:7,000), Anti-α-SMA (Cell Signaling Technology, 19245, dilution: 1:1,000), Anti-p-Smad2 (Cell Signaling Technology, 18338, dilution: 1:1,000), Anti-p-Smad3 (Cell Signaling Technology, 9520, dilution: 1:1,000), Anti-TGF-β2 (Bioss, bs-20412R, dilution: 1:1,000) and Anti-β-actin Antibody (Abcam, ab8227, dilution: 1:1,000), were used. The next day, membranes were incubated with secondary antibody (Proteintech, SA00001-2, dilution: 1:5,000) for 1 h at room temperature. And then the membranes were detected with ECL chemiluminescent substrate (Biosharp, Hefie, China). β-actin acted as the internal standard.

Cell transfection

Cardiac fibroblasts were transfected with lentiviral constructs (HanBio, Shanghai, China). SiRNAs were provided by Sigma–Aldrich. In addition, transfection of miR-152-3p mimic or miR-152-3p inhibitor with the riboFECT CP Transfection Kit (RiboBio, Guangzhou, China) was performed. The efficiency and function assay were confirmed 48 h after transfection. Transfection was conducted according to the reagent instructions.

RNA fluorescent in situ hybridization assay

RNA fluorescent in situ hybridization (RNA FISH) assay was performed as described previously (Shan et al., 2017). Cy3-labeled circHIPK3 and FITC-labeled miR-152 probes were transfected into CFs in six-well plates. The fluorescent signals were detected with a Fluorescent in Situ Hybridization Kit according to the reagent guidelines. Cell nuclei were counterstained with DAPI. The images were acquired with a fluorescence microscope.

Dual luciferase reporter assay

Firstly, the pmirGLO luciferase reporter plasmid, which contained the potential circHIPK3 and TGF-β2 binding site sequences in miR-152-3p, were constructed (GeneCreate, Wuhan, China). HEK-293T cells were cotransfected with wild-type or mutated pmirGLO-circHIPK3 reporter plasmid and miR-152-3p mimics or negative control (NC). Similarly, the pmirGLO-TGF-β2 reporter vector was cotransfected with miR-152. Firefly luciferase activity was standardized to Renilla luciferase activity.

Statistical analysis

SPSS 20.0 and GraphPad Prism 8.0 software were used to assess data and for statistical analysis. Kolmogorov–Smirnov test detected that the data were normally distributed. The Student t test was used to analyze the differences between two groups, one-way analysis of variance (ANOVA) was suitable for multiple groups. After ANOVA analyses, LDS method was used for the homogeneity of variance，Tamhane’s T2 method was used for variances. The difference was statistically significant with P-value < 0.05.

CircHIPK3 is upregulated in hypoxia-induced cardiac fibroblasts, which affects cell proliferation, migration, and phenotypic modulation in vitro

CircBase retrieval revealed that the Hipk3 host gene might produce three circRNAs, named circHIPK3, in the mouse genome. One of them, circHIPK3 (mmu_circ_0001052), was highly conserved. Our previous research illustrated that circHIPK3 (mmu_circ_0001052) was upregulated in exosomes released from cardiomyocytes that had been exposed to hypoxia (Wang et al., 2019b). Moreover, circHIPK3 has been reported to be enriched in human fibroblasts (Jeck et al., 2013). This urged us to further investigate the function of circHIPK3 in cardiac fibrosis in CFs under hypoxia.

To investigate the effects of hypoxia on fibroblasts, we established a hypoxia model of CFs in vitro, and the CCK-8 assay displayed that cell viability significantly increased after hypoxia for 24 h (Fig. 1A). Next, CFs were subjected to hypoxia for 24 h or left untreated (normal). Our date showed that hypoxia led to significantly increased proliferation and migration of CFs (Figs. 1B–1G) and caused increased expression of markers of myofibroblasts, including α-SMA and collagen types I and III (Col I and Col III) (Figs. 1H and 1I). Accordingly, an immunofluorescence assay revealed that Col III was highly expressed in hypoxia-induced CFs (Figs. 1J–1O). As expected, RT-qPCR demonstrated that circHIPK3 was significantly upregulated in hypoxia-induced CFs (Fig. 1P). These results suggested that hypoxia induced CF proliferation, migration, and phenotypic switching and increased the expression of circHIPK3.

Hypoxia clearly enhanced circHIPK3 expression in CFs, so we further explored the effect of circHIPK3 in CFs. We performed gain- and loss-of-function analyses of circHIPK3, in which CFs were transfected with circHIPK3 and si-circHIPK3. After 48 h, the transfection efficiency was examined by RT-qPCR. The results showed that circHIPK3 transfection significantly upregulated circHIPK3 expression and that si-circHIPK3 transfection was downregulated its expression (Fig. 2A). CircHIPK3 overexpression significantly promoted proliferation and migration and increased the expression of α-SMA, Col I and Col III. In contrast, circHIPK3 silencing (si-circHIPK3) attenuated proliferation, migration, and α-SMA, Col I and Col III expression (Figs. 2B–2O). Our date confirmed that circHIPK3 can accelerate the phenotypic switch of CFs under hypoxia in vitro.

CircHIPK3 could bind to miR-152-3p

CircHIPK3 has been shown to be predominantly existed in the cytoplasm (Wang et al., 2019b). The RNA FISH assay also revealed that circHIPK3 was predominantly located in the cytoplasm of CFs (Figs. 3A–3C). CircRNAs in the cytoplasm are mainly involved in biological processes through sponge-like binding of miRNAs (Li et al., 2015; Xin et al., 2017). To determine whether circHIPK3 could sponge miRNAs, bioinformatics prediction using the StarBase database indicated that miR-152-3p was predicted to bind to circHIPK3. We further carried out a dual luciferase reporter assay in which RLuc-circHIPK3-WT or RLuc-circHIPK3-Mut was cotransfected with miR-152-3p mimic or NC mimic into HEK-293T cells. MiR-152-3p mimic transfection markedly reduced the luciferase activity of RLuc-circHIPK3-WT, but the luciferase activity of the mutant was not affected (Fig. 3D). RNA FISH assays revealed that circHIPK3 and miR-152-3p were colocalized in the cytoplasm of CFs (Figs. 3E–3H). However, circHIPK3 did not affect the expression of miR-152-3p (Figs. 3I and 3J).These results indicated that circHIPK3 could bind to miR-152-3p.

MiR-152-3p inhibits cell proliferation and phenotypic transformation by targeting TGF-β2

To explore the role of miR-152-3p in CFs, we examined its role in CFs under hypoxic conditions. The miR-152-3p mimic, miR-152-3p inhibitor, NC mimic, and NC inhibitor were transfected into CFs. The enforced expression of miR-152-3p in CFs using a mimic inhibited CF proliferation and the expression of Col I and Col III after hypoxia stimulation (Figs. 4A–4I). In contrast, miR-152-3p knockdown in CFs resulted in an augmentation of fibrosis, including cell proliferation and expression of the fibrotic markers Col I and Col III (Figs. 4J–4R). Moreover, we found that the miR-152-3p mimics significantly increased and the miR-152-3p inhibitor significantly decreased miR-152-3p expression in CFs (Fig. 4S).

MicroRNA play a vital biological role by regulating mRNA degradation through complete or incomplete pairing with the 3′ untranslated region of mRNA and then inhibiting gene expression (Boon & Dimmeler, 2015; Hodgkinson et al., 2015; Wang et al., 2018). We screened various fibrosis-associated genes Smad2, Smad3, TGF-β1, and TGF-β2. Then, we searched the TargetScan database to predict miR-152-3p targets and found that miR-152-3p has binding sites on Smad2 and TGF-β2, and we selected TGF-β2 as the target gene for our study based on the context++ score percentile. Ultimately, we identified TGF-β2 as a downstream target of miR-152-3p. Furthermore, TGF-β2 was confirmed as a target of miR-152-3p by the dual luciferase reporter assay (Fig. 4T). In addition, the miR-152-3p mimic markedly reduced the expression of TGF-β2 protein, while downregulation of miR-152-3p clearly increased TGF-β2 protein expression (Figs. 4U–4X). Thus, these results revealed that miR-152-3p participates in antagonizing the phenotypic transformation by targeting TGF-β2.

CircHIPK3 promotes phenotypic transformation via miR-152-3p/TGF-β2 in cardiac fibroblasts

To explore whether circHIPK3 was involved in miR-152-3p-mediated regulation of proliferation, migration and fibrosis of CFs in vitro, the cells were cotransfected with circHIPK3 and miR-152-3p. Our data illustrated that circHIPK3 silencing could suppress cell proliferation and migration. Furthermore, the combination of circHIPK3 silencing and miR-152-3p overexpression had a weaker effect under hypoxia than circHIPK3 knockdown alone (Figs. 5A–5J). To evaluate whether circHIPK3 exerted its functions by regulating TGF-β2 expression, we measured the effect of circHIPK3 by western blotting. The results demonstrated that circHIPK3 silencing attenuated the expression of TGF-β2 and the fibrosis markers Col I and Col III. Moreover, it decreased the expression of Smad2 and Smad3, which are downstream proteins of TGF-β2. MiR-152-3p overexpression promoted the inhibitory effect of circHIPK3 silencing in CFs (Figs. 5K and 5L). Collectively, these results indicated that circHIPK3 can accelerate fibrosis progression via the functional pathway of sponging miR-152-3p.