All animal work was approved by The Roslin Institute's and the University of Edinburgh's Protocols and Ethics Committees. The animals were maintained in accordance with UK Home Office guidelines under the regulations of the Animal (Scientific Procedures) Act 1986.

Sheep primary aortic VICs were harvested from a 4-year-old Scottish mule sheep (generated from a Bluefaced Leicester sire and Scottish Blackface dam cross; Dryden Farm, Midlothian, UK). Rat aortic VICs were isolated from aortic valve leaflets dissected from the hearts of eight 5-week-old male Sprague Dawley rats as previously described (22). Sheep and rat valve leaflets were digested in 0.6 mg/ml collagenase Type II (Worthington, New Jersey USA) for 30 min and washed in Hanks' Balanced Salt Solution (HBSS; Life Technologies, Paisley, UK) to remove valve endothelial cells. The leaflets were subsequently digested with 0.6 mg/ml collagenase Type II for a further 1 h to release the VICs. Cells were pelleted at 300 × g for 5 min, before resuspension in growth medium consisting of Dulbecco's Modified Eagle Medium: Nutrient Mixture F-12 (DMEM/F12; Life Technologies) supplemented with 10% heat-inactivated foetal bovine serum (FBS; Life Technologies) and 1% gentamicin (Life Technologies), and cultured at 37°C in a humidified atmosphere of 95% air/5% CO₂ and grown for four passages.

Immortalised cell lines were established by Capital Biosciences (Gaithersburg, Maryland, USA) from the primary sheep and rat VICs through transduction with recombinant lentivirus encoding Simian virus (SV40) large and small T antigens (sheep), or large T antigen only (rat). The non-clonal cell lines were derived from multiple founder cells. Following continuous culture to 10 passages, transgene expression was confirmed by real-time quantitative PCR (RT-qPCR) for the expression of SV40 large T antigen (Capital Biosciences). The resulting cell lines were designated SAVIC (sheep; SVIC-SVTta) and RAVIC (rat; RVIC-SV40T).

SAVIC and RAVIC cells were seeded in growth media in multi-well plates at a density of 1.11×10⁴ cells/cm². Calcification was induced as described previously (22,25). Cells were grown to 80% confluence (Day 0), before treating with control (1.05 mM Ca/0.95 mM Pi) or test media: 1.5 to 3.6 mM calcium (Ca) and/or 1.5 to 2.5 mM phosphate (Pi). CaCl₂ and Na₂HPO₄/NaH₂PO₄ (Sigma-Aldrich, Dorset, UK) were used to supplement ionic calcium and phosphate in the media. To study the effect of calcification inhibitors and bisphosphonates on VIC calcification, SAVICs were exposed to inorganic pyrophosphate (PPi) and etidronate (both 0.1 mM; Sigma-Aldrich). Cells were incubated for up to 7 days in a humidified atmosphere of 95% air/5% CO₂, and the medium was changed every second/third day.

Calcium deposition was quantified based on a previously described method (26,27). Cells were washed twice with phosphate buffered saline (PBS) and decalcified with 0.6 M HCl at room temperature for 2 h. Free calcium was determined colorimetrically by a stable interaction with phenolsulphonethalein, using a commercially available kit (Randox Laboratories Ltd., County Antrim, UK), and corrected for total protein concentration (Bio-Rad Laboratories Ltd., Hemel Hempstead, UK) following solubilisation with 0.1 M NaOH/0.1% SDS. Absorbances were measured using a Synergy HT microplate reader (BioTek, Swindon UK) at 570 nm (calcium) and at 690 nm (protein).

To confirm retention of mesenchymal phenotype, cell monolayers cultured on glass coverslips were fixed with 4% paraformaldehyde (PFA) and washed with phosphate-buffered saline (PBS). Fixed cells were permeabilised with 0.3% Triton X-100 (Sigma-Aldrich) and incubated with rabbit polyclonal anti-α-smooth muscle actin (α-SMA; catalogue #ab5694; 1:100; Abcam, Cambridge, UK), mouse monoclonal anti-vimentin (catalogue #V6384; 1:900; Sigma-Aldrich) or rabbit polyclonal anti-cluster of differentiation 31 (CD31; 1:900; Abcam, Cambridge, UK) overnight at 4°C. After washing, cells were incubated with Alexa Fluor^® 488 donkey-anti-rabbit antibody (catalogue #A-21,206; 1:250; ThermoFisher Scientific) or Alexa Fluor^® 594 goat-anti-mouse antibody (catalogue #A-110,055; 1:250; ThermoFisher Scientific) for 1 h in the dark. Glass coverslips were mounted onto slides with Prolong Gold Anti-Fade Reagent containing DAPI (Life Technologies). Slides were examined using a Leica DMLB fluorescence microscope (Leica Geosystems, Milton Keynes, UK). In place of the primary antibody, control cells were incubated with non-immune mouse or rabbit IgG (2 µg IgG/ml, Sigma-Aldrich).

RNA extraction was performed using the RNeasy Mini kit (Qiagen, West Sussex, UK), according to the manufacturer's instructions. RNA abundance was quantified, RNA was reverse transcribed, and the expression of selected genes were quantified via RT-qPCR employing the SYBR green detection method (PrecisionPLUS mastermix, Primerdesign Ltd, Southampton, UK), measured on a Stratagene Mx3000P (Agilent Technologies, Stockport, UK), as previously reported (28,29). Sheep primers for Runt-related transcription factor 2 (RUNX2; Forward 5′-CTCCTCCATCCATCCACTCC-3′; Reverse 5′-CAGAGGCAGAAGTCAGAGGT-3′) and Matrix Gla protein (MGP; Forward 5′-ACAACAGAGATGGAGAGCGA-3′; Reverse 5′-CGGAAATAACGGTCGTAGGC-3′) were designed via Primer3 (http://primer3.ut.ee/) to span exon-exon junctions, and obtained from Invitrogen (Paisley, UK). Primers for sheep glyceraldehyde 3-phosphate dehydrogenase (GAPDH; sequences not disclosed), tyrosine 3-monooxygenase (YWHAZ; sequences not disclosed), and sodium-dependent phosphate transporter 1 (PiT1; also known as SLC20A1; Forward 5′-ACATCTTGAACGCCGCTA-3′; Reverse 5′-AGTAGCAGCAATAGCAGTGGTA-3′) were obtained from Primerdesign Ltd. Sheep expression data were normalised against the geometric mean of GAPDH and YWHAZ. Rat primers for Gapdh, Mgp, and Runx2 were obtained from Qiagen (sequences not disclosed; QuantiTect primers, Qiagen). Rat Pit1 primers were acquired from Primerdesign Ltd. (sequences not disclosed). Rat expression data were normalised against Gapdh. The ΔΔCq method was used to analyse relative gene expression (30).

All statistical analyses were performed using Minitab 17 (Minitab Inc., Coventry, UK). General Linear Model (GLM) analysis incorporating pairwise comparisons and the Student's t-test were used to assess the data. Data are presented as mean ± standard error of the mean (SEM). P<0.05 was considered to be statistically significant, and P-values are represented as: *P<0.05; **P<0.01; ***P<0.001.

Cells showed positive immunohistochemical staining for vimentin (Fig. 1A) and α-SMA (Fig. 1B), in agreement with previous reports of primary VIC cultures (31,32). Cells were negative for the endothelial marker CD31 (Fig. 1C).

Initial studies were undertaken to determine whether the calcification of SAVICs could be induced when cultured in the presence of calcifying medium containing Ca and/or Pi (Fig. 2A). Ca potently induced the calcification of SAVICs from 2.7 mM (1.9 fold; P<0.001; n=6; Fig. 2A) whereas Pi treatment alone had no effect (Fig. 2A). The treatment of VICs with Ca and Pi together had a synergistic effect on VIC calcification from 2.7 mM Ca/2.0 mM Pi (22.2 fold; P<0.001; n=6; Fig. 2B).

Next, gene expression studies were undertaken in calcifying SAVICs to investigate the expression profile of key genes associated with vascular calcification. Expression of the gene encoding the master osteoblastic transcription factor, RUNX2, was significantly increased from 4.5 mM Ca (1.3 fold; P<0.05; n=6; Fig. 3A) compared to control culture conditions. Expression of sodium-dependent phosphate transporter 1 gene (PiT1) was significantly increased at 5.4 mM Ca (1.2 fold; P<0.05; n=6; Fig. 3B). In contrast, a decrease in the expression of the calcification inhibitor Matrix Gla Protein gene (MGP) was noted from 3.6 mM Ca (1.3 fold; P<0.01; n=6; Fig. 3C).

Further experiments investigated whether the calcification of SAVICs could be reduced using recognised inhibitors of calcification.

As CAVD is a consequence of hydroxyapatite formation and deposition in the aortic valve, we examined the effects of PPi (0.01-1 mM; Fig. 4A) and the bisphosphonate etidronate (0.01-1 mM; Fig. 4B), both established inhibitors of hydroxyapatite formation (33–35) on SAVIC calcification. A significant decrease in calcium deposition was observed at 1 mM PPi (1.8 fold; P<0.01; n=6; Fig. 4A). Additionally, following exposure to 1 mM etidronate, a significant reduction in calcium deposition (3.2 fold; n=6; P<0.01) was observed, confirming the inhibitory effect of this bisphosphonate on valve calcification in vitro (Fig. 4B).

We further corroborated the application of immortalised VICs as an in vitro model of aortic valve calcification through the generation of a rat immortalised VIC cell line, RAVIC. A presented in Fig. 5, RAVICs also showed positive immunohistochemical staining for vimentin (Fig. 5A) and α-SMA (Fig. 5B), and were negative for the endothelial marker CD31 (data not shown).

Analogous to our findings with the SAVICs, Ca markedly induced the calcification of RAVICs from 2.7 mM (4.6 fold; P<0.01; n=6; Fig. 5E) and Pi treatment alone had no effect (Fig. 5E). Furthermore, the treatment of RAVICs with Ca and Pi together had a synergistic effect on cell calcification (2.7 mM Ca/2.0 mM Pi; 82.2 fold; P<0.001; n=6; Fig. 5F). 5.4 mM Ca treatment induced a significant increase in the mRNA expression of Runx2 and Pit1 (1.5 fold; P<0.01, and 4.7 fold; P<0.001, respectively; Fig. 5G and H), with a concomitant reduction in Mgp expression (3.2 fold; P<0.001; Fig. 5I).