Identification and characterization of aortic valve mesenchymal progenitor cells with robust osteogenic calcification potential

Abstract

Advanced valvular lesions often contain ectopic mesenchymal tissues, which may be elaborated by an unidentified multipotent progenitor subpopulation within the valve interstitium. The identity, frequency, and differentiation potential of the putative progenitor subpopulation are unknown. The objectives of this study were to determine whether valve interstitial cells (VICs) contain a subpopulation of multipotent mesenchymal progenitor cells, to measure the frequencies of the mesenchymal progenitors and osteoprogenitors, and to characterize the osteoprogenitor subpopulation because of its potential role in calcific aortic valve disease. The multilineage potential of freshly isolated and subcultured porcine aortic VICs was tested in vitro. Progenitor frequencies and self-renewal capacity were determined by limiting dilution and colony-forming unit assays. VICs were inducible to osteogenic, adipogenic, chondrogenic, and myofibrogenic lineages. Osteogenic differentiation was also observed in situ in sclerotic porcine leaflets. Primary VICs had strikingly high frequencies of mesenchymal progenitors (48.0 +/- 5.7%) and osteoprogenitors (44.1 +/- 12.0%). High frequencies were maintained for up to six population doublings, but decreased after nine population doublings to 28.2 +/- 9.9% and 5.8 +/- 1.3%, for mesenchymal progenitors and osteoprogenitors, respectively. We further identified the putative osteoprogenitor subpopulation as morphologically distinct cells that occur at high frequency, self-renew, and elaborate bone matrix from single cells. These findings demonstrate that the aortic valve is rich in a mesenchyma l progenitor cell population that has strong potential to contribute to valve calcification.

Figure 1

VIC nodule formation in vitro is potentiated by osteogenic supplements. A: Phase contrast micrograph of a VIC nodule. B: After 1 week of culture, ALP-positive nodules formed most frequently in osteogenic medium (complete medium with 10 mmol/L β-glycerophosphate and 10 μg/ml ascorbic acid) supplemented with 10 mmol/L dexamethasone (Dex). Data are presented as mean ± SE for n = 3 independent samples. P < 0.05 by Fisher LSD. Additional experiments showed similar results. C: Cells associated with the nodules exhibited cuboidal, osteoblast-like morphologies. D: In contrast, cells in the non-nodule forming areas were elongated and fibroblastic. E: Cells beneath the nodules appeared viable and well-adhered to synthesized extracellular matrix.

Figure 2

Mineralized nodules are bone-like. A: Nodules stained positive for von Kossa, indicating the presence of calcium salts within the nodules. B: Transmission electron microscopy of the cross section of a nodule showing mineral closely apposed to intact, viable cells. M: mineral; C: cells. C: Electron diffraction patterns from mineral deposits showed between two to six rings with d-spacings consistent with those of hydroxyapatite. D and E: Consistent with this being an osteogenic process, the nodules stained positive for ALP activity (D) and osteocalcin (E; inset is negative control with no primary antibody added). F: Cells within the nodules did not take up the APOPercentage dye, indicating that they were not apoptotic (positive control is shown in the inset). G: Gene expression analysis by PCR showed positive expression of osteocalcin, osteonectin, and Runx2 in induced VICs. H: A large proportion of VICs within lesions in the fibrosa of sclerotic leaflets expressed Runx2 (inset is a healthy valve immunostained for Runx2).

Figure 3

VICs are capable of adipogenic, chondrogenic, and myofibrogenic differentiation. A: Primary VICs cultured two to three weeks in adipogenic medium demonstrated Oil Red O-positive lipid droplets in the cytoplasm and expressed adipogenic genes adipocyte fatty acid-binding protein 2 (aP2), lipoprotein lipase (LPL), and peroxisome proliferator-activated receptor γ2 (PPARγ2) (E). B: Primary VICs pelleted and grown in chondrogenic media for three weeks stained positive for alcian blue and expressed the chondrogenic transcription factor SRY-box containing gene 9 (Sox9) (F). C and D: Primary VICs plated on tissue culture plastic in complete medium for 2 days expressed a low level of monomeric α-SMA (C), but by day 6, most cells expressed high levels of α-SMA that was incorporated into stress fibers (D). E and F: PCR analysis of (E) adipogenic and (F) chondrogenic gene expression.

Figure 4

Mesenchymal progenitor frequencies in primary and subcultured VICs. A: The mesenchymal progenitor frequency of primary VICs was 48.0 ± 5.7% and was maintained at about 50% for up to six population doublings. After nine population doublings, the frequency dropped to 28.2 ± 9.9%. *P < 0.05 by Fisher’s LSD. B: The CFU-F frequency was independent of the plating density. Data in (A) and (B) are presented as mean ± SE for n = 3 independent experiments. Representative data from one experiment with primary VICs are shown in (B).

Figure 5

Osteoprogenitor frequencies in primary and subcultured VICs. A: Primary VICs had an osteoprogenitor frequency of 44.1 ± 12.0% that decreased to 5.8 ± 1.3% at passage three (∼nine population doublings). *P < 0.05 by Fisher’s LSD. B: The osteoprogenitor frequency was independent of the plating density. Data in (A) and (B) are presented as mean ± SE for n = 3 independent experiments. Representative data from one experiment with P3 VICs are shown in (B).

Figure 6

VICs contain a morphologically distinct subpopulation that is enriched for osteoprogenitors. A–D: Based on cell morphology, four major types of colonies were observed: (A) tight, swirling, fibroblast-like cells (S-type); (B) loosely-packed fibroblast-like cells (loose fibroblasts); (C) small, flat cells; and (D) large, flat cells. E: The relative proportions of each type of colony changed with subculture. In particular, the frequency of S-type colonies decreased significantly. This was accompanied by a significant increase in colonies made up of large, flat cells. The S-type colonies were enriched for osteoprogenitor cells (see text). Data are presented as mean ± SE for n = 3 independent experiments.