Initiation of mineralization in bioprosthetic heart valves: Studies of alkaline phosphatase activity and its inhibition by AlCl3 or FeCl3 preincubations

Abstract

The principal cause of the clinical failure of bioprosthetic heart valves fabricated from glutaraldehyde‐pretreated porcine aortic valves is calcification. Other prostheses composed of tissue‐derived and polymeric biomaterials also are complicated by deposition of mineral. We have previously demonstrated that: (a) Failure due to calcification of clinical bioprosthetic valves can be simulated by either a large animal circulatory model or subdermal implants in rodents. (b) Calcification of bioprosthetic tissue has complex host, implant, and mechanical determinants. (c) The initial calcification event in the rat subdermal model is the mineral deposition in devitalized cells intrinsic to the bioprosthetic tissue within 48 to 72 h, followed later by collagen mineralization. (d) Initiation of bioprosthetic tissue mineralization, like that of physiological bone formation, has “matrix vesicles” as early nucleation sites. (e) Alkaline phosphatase (AP), an enzyme also associated with matrix vesicles involved in bone mineral nucleation, is present in both fresh and fixed bioprosthetic tissue at sites of initial mineralization. (f) Certain inhibitors of bioprosthetic tissue calcification (e.g., Al³⁺, Fe³⁺) are localized to the sites at which alkaline phosphatase is present. On the basis of these results, we hypothesize that alkaline phosphatase is a key element in the pathogenesis of mineralization of bioprosthetic tissue. In the present studies, we focused on the relationship of AP to early events in calcification, and the inhibition of both calcification and AP activity by FeCl₃ and AlCl₃ preincubations. Subdermal implants of glutaraldehyde pretreated bovine pericardium (GPBP) were done in 3‐week‐old rats. AP was characterized by enzymatic hydrolysis of paranitrophenyl phosphate (pnpp), and by histochemical studies. Calcification was evaluated chemically (by atomic adsorption spectroscopy) and morphologically (by light microscopy). The results of these studies are as follows: (a) Extractable AP activity is present in fresh but not glutaraldehyde‐pretreated bovine pericardial tissue. However, histochemical studies reveal active AP within the intrinsic devitalized cells of GPBP, despite extended glutaraldehyde incubation. (b) Extrinsic AP is rapidly adsorbed following implantation, with peak activity at 72 h (424 ± 67.2 nm pnpp/mg protein/min enzyme activity [units]), but markedly lesser amounts at 21 days (96.8 ± 3.9 units). (c) Simultaneously to the AP activity maximum, bulk calcification is initiated, with GPBP calcium levels rising from 1.2 ± 0.1 (unimplanted) to 2.4 ± 0.2 μg/mg at 72 h, to 55.6 ± 3.1 μg/mg at 21 days, despite a marked decline in AP activity at this later time. (d) Preincubation of GPBP in either FeCl₃ or AlCl₃, at concentrations (0.1 M) which inhibited GPBP calcification, significantly reduce AP activity. We conclude that endogenous AP activity is present but not extractable in unimplanted GPBP fixed for extended periods. However, concurrent with the time of the onset of GPBP calcification in the rat subdermal model, AP adsorbed following implantation rises sharply to a maximum, thereby augmenting intrinsic enzyme. Preincubations of GPBP in either AlCl₃ or FeCl₃ not only prevent calcification, but also result in reduction in AP activity. These results strongly suggest that AP is an important cofactor in the mechanism of bioprosthetic valve mineralization and may be a fruitful target for anticalcification treatments.