The present study was approved by the Ethics Committee of the First Affiliated Hospital of Nanjing Medical University, and written informed consent was obtained from all patients prior to the study. Aortic leaflets were collected from patients receiving aortic replacement between January 2016 and December 2018 at Jiangsu Province Hospital. Echocardiography was performed to confirm the morphology and function of the leaflets prior to surgery. A total of 15 BAV samples were examined in the present study, all of which presented type 1 BAV leaflets, with fusion of the right and left coronary leaflets. A total of 15 degenerative TAVs were selected according to the echocardiographic results and intra surgical inspections. All leaflets were sectioned into two parts. The larger part was flash frozen in liquid nitrogen and stored for RNA or protein extraction, while the smaller part was fixed in 10% formalin and embedded in paraffin for subsequent experimentation.

Agilent Human miRNA Microarray Slide (version 21.0; 8×60K; Design ID: 070156; Agilent Technologies, Inc.) was used to profile miRNA expression in 8 aortic valve tissue samples from patients with BAV or TAV. Total RNA was quantified using a NanoDrop ND-2000 spectrophotometer (Thermo Fisher Scientific, Inc.) and the RNA integrity was assessed using an Agilent Bioanalyzer 2100 (Agilent Technologies, Inc.). The sample labeling, microarray hybridization and washing were performed based on the manufacturer's standard protocol. Briefly, total RNA was dephosphorylated, denaturated and labeled with Cyanine-3-CTP. After purification the labeled RNAs were hybridized onto the microarray. After washing, the arrays were scanned using the Agilent Scanner G2505C (Agilent Technologies, Inc.).

Feature Extraction software (version 10.7.1.1; Agilent Technologies, Inc.) was used to analyze array images to get raw data. Subsequently, Genespring software (version 13.1; Agilent Technologies, Inc.) was employed to finish the basic analysis with the raw data. To begin with, the raw data was normalized with the quantile algorithm. The probes that at least 100.0 percent of samples in any 1 condition out of 2 conditions have flags in ‘Detected’ were chosen for further data analysis. Differentially expressed miRNAs were then identified through fold change as well as P-value calculated using t-test. The threshold set for up- and down-regulated genes was a fold change>=1.5 and a P-value<=0.05. Target genes of differentially expressed miRNAs were the intersection predicted with 3 databases [(TargetScan (www.targetscan.org), microRNAorg (www.microrna.org) and PITA (www.pictar.org)].

miRs or mRNAs were extracted from the tissues obtained from surgical operations or cultured VICs using the miRNeasy Mini kit (Qiagen, Inc.) according to the manufacturer's instructions, and reverse transcribed into cDNA using the Takara PrimeScript™ RT master mix (Takara Bio, Inc.) according to the manufacturer's instructions. qPCR was subsequently performed using the 7900HT Fast Real-Time PCR System (Thermo Fisher Scientific, Inc.) according to manufacturer's protocol. The primer sequences used for qPCR were purchased from Realgene Bio-Technologies, Inc. and are presented in Table I. The following thermal cycling conditions were used for the qPCR: Initial denaturation at 95°C for 10 min; then 40 cycles of 95°C for 15 sec and 60°C for 1 min. Relative expression levels were measured using the 2^−ΔΔCq method (18) and normalized to the internal reference gene GAPDH. U6 small nuclear RNA was used as internal control to normalize miRNA expression. All experiments were performed in triplicate.

The tissue samples were fixed with 10% formalin at room temperature for 24 h, embedded in paraffin and cut into 20-µm thick sections for Von Kossa and Sirius Red staining and semi-quantitative analysis. Tissues were also used for immunohistochemical staining for CREBBP, collagen I, Runt-related transcription factor 2 (Runx 2), matrix metalloproteinase (MMP)-2 and MMP-9. In brief, sections were washed in Tris-buffered saline/0.1% Tween-20 and subsequently blocked in 5% milk at room temperature for 2 h. Tissue sections were incubated with the primary antibodies (all 1:1,000) listed in Table II at 4°C overnight. Sections were washed in PBS prior to incubation with biotinylated secondary antibody (1:200; cat. no. 21537; Merck KGaA) at room temperature for 30 min. The sections were subsequently treated with Vectastain ABC reagent (Vector Laboratories, Inc.) and 3,3′-diaminobenzidine, prior to counterstaining with hematoxylin at room temperature for 2 min. A light optical microscope was used for image capture (magnification, ×10 or ×200) using visible light source without any filters. Image-Pro plus software (version 6.0; Media Cybernetics, Inc.) was used for signal quantification. In brief, the positive area was automatically selected according to criteria that contained a specific Hue-Saturation-Intensity (HIS) parameter adjusted to the different staining method. Subsequently, the optical density (OD) of the positive areas and the whole image were calculated automatically. The data were extracted and the OD of per unit area was calculated and presented as a semi-quantitative indicator.

All procedures involving the use of animals were approved by the Institutional Animal Care and Use Committee of Nanjing Medical University. Porcine aortic VICs were harvested from porcine hearts, which were supplied by the Laboratory Animal Center of Nanjing Medical University and isolated by collagenase digestion as previously described (19). In brief, pigs at 3–4 months old were sedated using a cocktail of ketamine hydrochloride (20 mg/kg, intramuscular injection) and xylazine (2 mg/kg, intramuscular injection). Following adequate sedation (~15 min), pigs were euthanized by overdose of pentobarbital (156 mg/kg body weight) administered intravenously and their hearts were harvested immediately. Aortic valves were excised with precision to exclude non-leaflet tissue. The excised leaflets were carefully scraped on the aortic and ventricular aspects to remove the endothelial layer. Tissues samples were subsequently sectioned into ~2×2 mm explants, which were cultured in DMEM (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 1% penicillin-streptomycin (Beyotime Institute of Biotechnology), and 15% fetal bovine serum (ScienCell Research Laboratories, Inc.), in 5% CO₂ at 37°C. Cells that reached ~100% confluence were extracted from the explants for subculture and those of the 2nd or 3rd generations were used for further analyses.

An antisense-based strategy was adapted to block miRNA function in cultured cells. Anti-miR-330 2′-O-methyl oligoribonucleotides (Guangzhou RiboBio Co., Ltd.) were synthesized to sequence-specifically inactivate miR-330 and thus to release the suppression on target mRNAs without changing its expression level (20). A total of 7.5×10⁴ cells/cm² VICs were transfected with 5 nmol/l miR-330-3p mimic or inhibitor (micrONTM miR mimic/inhibitor/nc_Standard; cat. nos. miR10000751-1-5 and miR20000751-1-5; Guangzhou RiboBio Co., Ltd.) using Lipofectamine™ 3000 reagent (Invitrogen; Thermo Fisher Scientific, Inc.). The transfected cells were incubated in DMEM supplemented with 15% FBS for 6 h and subsequently were incubated in DMEM supplemented with 1% penicillin-streptomycin and 15% FBS. At 48 h post-transfection, cells were used for subsequent experiments.

VICs were also transfected with miR-659 mimic, miR-663 mimic, or miR-146 inhibitor (Guangzhou RiboBio Co., Ltd.) using the aforementioned protocol.

To prepare the calcification medium, 1 mmol of CaCl₂ and 1 mmol of Na₅P₃O₁₀ were dissolved in 50 ml of water to produce reagent one and reagent two, respectively. Subsequently, 500 µl reagent one, 375 µl reagent two and 4 ml of 8% FBS were added to 45 ml DMEM. VICs were cultured with or without pro-osteogenic medium for 48 h post-transfection, and subsequently stained with Alizarin Red S Staining (Beyotime Institute of Biotechnology) according to the manufacturer's protocol. Briefly, cells were washed with Ca²⁺-free PBS three times, fixed with 4% paraformaldehyde for 10 min at room temperature and then fixed with 95% ethanol for 20 min at room temperature. Subsequently, cells were stained with 1% alizarin red solution (pH 4.2; cat. no. ST1078; Beyotime Institute of Biotechnology) for 1 min at room temperature to visualize matrix calcium deposition. The solution was washed out with distilled water and the stained cells were photographed using a digital camera. The images were then processed and analyzed using Image-Pro Plus software (version 6.0, Media Cybernetics, Inc.).

Transfected cells were harvested and used for preparation of protein extracts. Cells were lysed using a cell lysis kit (Nanjing KeyGen Biotech Co., Ltd.). The bicinchoninic acid method was used for quantification of total protein of the samples. Equal amounts of protein extracts (30 µg) were separated by 10% SDS-PAGE and transferred to PVDF membranes. Following blocking with TBS-Tween-20 solution containing 5% BSA (cat. no. B600036; Sangon Biotech Co., Ltd.) for 2 h at room temperature, the membranes were incubated with the following primary antibodies: Rabbit polyclonal antibody against Runx 2 (1:1,000; cat. no. 12556; Cell Signaling Technology, Inc.), rabbit polyclonal antibody against BMP2 (1:1,000; cat. no. 14933; Abcam), rabbit polyclonal antibody against CREBBP (1:1,000; cat. no. 2832; Abcam), rabbit polyclonal antibody against collagen I (1:1,000; cat. no. WL0088; Wanleibio Co., Ltd.), and mouse monoclonal antibody against GAPDH (1:1,000; cat. no. 8245; Abcam) overnight at 4°C. The membranes were then probed using either a goat anti-rabbit or a goat anti-mouse horseradish peroxidase-conjugated secondary antibody (1:5,000; cat. nos. ab150077 and ab205719; Abcam) for 2 h at room temperature. The blots were developed with an Omni-ECL™ Enhanced Pico Light Chemiluminescence kit (Epizyme Co.) and exposed on a ChemiDoc MP imaging system (Bio-Rad Laboratories, Inc.). Band density was quantified using ImageJ software (version 1.8.0; National Institutes of Health) and normalized to GAPDH.

Specific culture medium was prepared by mixing DMEM, 10% FBS (ScienCell Research Laboratories, Inc.), 1% nonessential amino acids, 1% L-glutamine and 1% penicillin with streptomycin. 293T cells (National Infrastructure of Cell Line Resource) were seeded into 12-well plates (5×10⁴ cells/cm²) and cultured in specific medium at 37°C in 5% CO₂.

The pmirGLO dual luciferase miR target expression vector (pmirGLO), containing both firefly and Renilla luciferase genes was purchased from Promega Corporation. Human CREBBP 3′-untraslated region (3′-UTR), including the predicted binding site of miR-330-3p, was amplified via RT-PCR and inserted into the 3′-UTR downstream of the firefly luciferase gene in the pmirGLO vector (pmirGLO-UTR) using XbaI and SacI restriction sites. The QuikChange XL Site-Directed Mutagenesis kit (Agilent Technologies, Inc.) was used to construct the mutant miR-330-3p-binding site vector (pmirGLO-UTR-MUT) according to the manufacturer's protocol. Restriction enzyme digestion and sequencing was performed to validate the constructs. After cellular transfection for 36 h, luciferase activity was assessed with Lipofectamine™ 3000 reagent (Invitrogen; Thermo Fisher Scientific, Inc.) using a Dual-Glo luciferase assay system (Promega Corporation). Firefly luciferase activity was normalized to Renilla luciferase activity.

The design and use of engineered Cas9 complex (pLenti-CMV-NLS-dCas9-VP64-2A-Puro; cat. no. H7281; Heyuan Kangning Medical Company Biotechnology Limited Company) and efficient single guide RNA (gRNA; plenti-U6-gRNA-2×(wt+f6)MS2-CMV-MCP-P65-HSF1-IRES-blasticidin; cat. no. H7284; Heyuan Kangning Medical Company Biotechnology Limited Company) to induce CREBBP transcriptional activation was performed according to previously published protocols (21,22). CCTop, a CRISPR/Cas9 target online predictor (cctop.cos.uni-heidelberg.de) was used to design the gRNAs used in the present study. The sequences of both control and CREBBP gRNAs are as follows: 5′-GCACTACCAGAGCTAACTCA-3′ (control gRNA) and 5′-CCACTTAATGAATTCGCTCG-3′ (CREBBP gRNA). The gRNA primers were annealed and cloned into gRNA (MS2)-plasmids via the BbsI sites. All CRISPR constructs were purchased from Heyuan Biotech Company (cat. nos. 61422, 61423 and 61424). The CREBBP promoter-luciferase construct was generated by synthesizing the DNA fragment corresponding to the CREBBP promoter region from GenScript, and sub-cloned into the pGL4.20 vector (Heyuan Biotech Company).

Statistical analysis was performed using SPSS software (version 19.0; IBM Corp.). Univariate analyses were performed using the χ² test for categorical variables, and Student's t-test or one-way ANOVA followed by Dunnett's post hoc were performed for continuous variables. Pearson's correlation analysis was used to investigate the relationship between two quantitative, continuous variables. Data are presented as the mean ± SD. P<0.05 was considered to indicate a statistically significant difference.

Clinical characteristics of the patients involved in the present study are listed in Table III. Patients with BAV were significantly younger compared with patients with TAV. No significant differences were observed between the two groups in sex and other risk factors concerning cardiovascular calcification. Both groups presented similar preoperative echocardiography results, which measured the left ventricular ejection fraction, aortic valve area and mean trans valvular gradient. Based on the echocardiography results, patients with a mean aortic valve area <0.75 cm² were diagnosed with severe AS.

Valve tissues from patients with AS, with either BAV (n=4) or TAV (n=4) were used for miR profiling. miRs that were commonly upregulated or downregulated in patients with BAV compared with patients with TAV were selected, including 11 upregulated and eight downregulated genes, respectively (Table IV). The heat map of the differentially expressed miRs are presented in Fig. 1A.

miR sequencing results of the patients' valve tissues were validated via RT-qPCR analysis. miR-330-3p, miR-659 and miR-663 expression levels in the BAV leaflets were significantly increased compared with the degenerative TAV leaflets by 3.72-fold, 5.04-fold and 4.22-fold, respectively (Fig. 1B-D); however, miR-146 expression was significantly decreased by −11.56-fold (Fig. 1E). No significant differences were observed in the other 15 miRs (Fig. S1).

VICs were successfully transfected with miR-659 or miR-663 mimics or miR-146 inhibitor as the expression of these miRNAs were significantly increased (miR-659 and miR-663) or decreased (miR-146) following transfection (Fig. S2A, E and I). However, levels of calcification-associated factors Runx 2 and BMP2 and the fibrosis-associated factor collagen I in miR-659 mimic- and miR-663 mimic-transfected cells did not show any significant difference compared with controls (Fig. S2B-D, F-H). In addition, mRNA expression levels of Runx 2 and BMP2 were significantly decreased in miR-146 inhibitor-transfected cells compared with controls (Fig. S2J and K). As the simulation of tissue expression alterations to the three microRNAs in cultured VICs failed to demonstrate similar calcification-associated protein alterations to those observed in BAV tissue samples, further experimentation on miR-659, miR-663 and miR-146 was ceased.

To investigate whether miR-330-3p plays a role in cellular calcification, VICs transfected with miR-330 mimic or inhibitor were incubated under pro-osteogenic conditions (Fig. 2A). After 48 h, both under pro-osteogenic conditions, a significantly higher degree of calcification was observed in the miR-330-3p mimic group compared with the miR-330-3p negative control group, while cells transfected with miR-330 inhibitor demonstrated a significant decrease of calcification compared with the miR-330-3p negative control group (Fig. 2A). Higher collagen I mRNA expression was observed in VICs treated with miR-330-3p mimic compared with the control group (Fig. 2B), indicating that the fibrosis progress was accelerated by miR-330, thus fibrosis is considered a pro-calcification progress. Concurrently, calcification-associated pathways were investigated at the mRNA level. VICs exhibited a 2.88-fold increase in BMP2 mRNA expression in the miR-330 mimic group compared with the control group (Fig. 2B). Similar alteration patterns were demonstrated in the expression of the downstream pro-osteogenic transcription factor, Runx 2 (7.95-fold increase vs. control; Fig. 2B). Alternatively, VICs transfected with miR-330 inhibitor demonstrated a significant decrease in mRNA expression of collagen I, BMP2 and Runx 2 compared with their respective control groups (Fig. 2C).

Furthermore, the protein expression of collagen I, BMP2 and Runx 2 significantly increased in cells transfected with miR-330-3p mimic compared with the control groups following pro-calcification culture, indicating that the calcification progress was accelerated in these cells (Fig. 3A and B). VICs transfected with miR-330-3p inhibitor showed decreased protein expression levels of BMP2 and Runx 2 compared with the control group; however, protein expression level of collagen 1 increased (Fig. 3A and C). Western blots from different experiments are shown in Fig. S3A and B. Taken together, these results suggested that miR-330-3p might facilitate the calcification progress in cultured VICs.

Based on online bioinformatics analysis using TargetScan software (www.targetscan.org, version 7.0) and relevant studies investigating vascular calcification (23,24), CREBBP was identified as a potential target gene of miR-330-3p. Immunohistochemical staining of CREBBP was performed in human aortic valve tissue samples, which demonstrated significantly increased CREBBP protein expression in BAV leaflet tissues compared with TAV leaflets (Fig. 4A and B). CREBBP mRNA expression in human leaflets was also measured, and the results demonstrated a −2.08-fold decrease of CREBBP mRNA expression in BAV leaflets compared with TAV tissues (Fig. 4C). Pearson's correlation analysis on these human tissue samples indicated that the CREBBP protein expression levels negatively correlated with miR-330-3p expression (r=−0.897; P<0.05; Fig. 4D).

Computational analysis demonstrated that the seed sequence of hsa-miR-330-3p was complementary to the 56–62 nt of the 3′-UTR in CREBBP mRNA, which is a potential binding site for miR-330-3p directly targeting CREBBP (Fig. 4E). Dual-luciferase reporter assay was subsequently performed using 293T cells transfected with wild-type or point mutation vectors of the 3′-UTR region of CREBBP at the predicted binding site. Luciferase activity was significantly decreased in cells treated with miR-330-3p mimic in the wild type group compared with the negative control group, while there was no significant difference in luciferase activity between the mutated groups (Fig. 4F), indicating that miR-330-3p directly binds to the CREBBP 3′-UTR.

Furthermore, cultured VICs treated with miR-330 mimic for 36 h exhibited significantly lower CREBBP protein expression by −1.88-fold compared with the control group, while treatment with miR-330-3p inhibitor significantly increased CREBBP protein expression by 1.55-fold (Fig. 4G-I). Western blots from repeat experiments are shown in Fig. S3A and B. CREBBP mRNA expression in cultured VICs showed the same patterns to protein expression (Fig. 4J and K).

It was further investigated whether the calcification promotion effect of miR-330-3p depends on the downregulation of CREBBP. Since the coding sequence of CREBBP gene is too long, CREBBP protein cannot be overexpressed by the conventional construction of plasmids containing its cDNA sequence. Hence, endogenous overexpression was performed using the transcriptional activation method. CREBBP transcriptional activation was induced by CRISPR activation (CRISPRa) strategy, which exploits the targeting mechanism of an sgRNA to direct the complex to a promoter or enhancer of a gene. In the present study, CRISPRa targeted the CREBBP promoter region and significantly increased the expression of endogenous CREBBP in VICs compared with the control group (Fig. 5A and B). CREBBP mRNA expression was also significantly increased in the CREBBP sgRNA group compared with sgRNA negative control group in the presence of miR-330-3p mimic (Fig. 5C). mRNA (Fig. 5D) and protein (Fig. 5E and F; S3C and D) expression levels of Runx 2 and BMP2 significantly decreased following overexpression of CREBBP in the presence of miR-330-3p mimic, indicating that the presence of CREBBP may alleviate the degree of calcification in the high miR-330-3p environment.

To validate that upregulated microRNA-330-3p promotes calcification in patients with BAV, the positive areas of pro-calcification proteins in aortic valve samples from patients with BAV and TAV were examined. Von Kossa staining demonstrated similar calcium deposits between TAV and BAV leaflet slices (Fig. 6E). Furthermore, immunohistochemical staining of Runx 2, collagen 1, MMP 2 and MMP 9 demonstrated significantly higher positive areas in BAV samples compared with TAV samples (Fig. 6A-D). These alterations suggest a poor disorder of the interstitial function of VICs in patients with BAV compared with patients with TAV. Furthermore, Sirius Red staining demonstrated a higher concentration of elastin in the BAV group compared with the TAV group, which is prone to calcification in a number of disease processes (Fig. 6F).