Robert J. Levy

The identification of the vitamin k-dependent bone protein osteocalcin as one of the γ-carboxyglutamic acid containing proteins present in calcified atherosclerotic plaque and mineralized heart valves☆

Calcium binding proteins containing gamma-carboxyglutamic acid (Gla) have previously been demonstrated to occur in calcified atherosclerotic plaque and calcified cardiac valves. Experiments were carried out to determine if one of the Gla-containing proteins in human cardiovascular calcifications is the vitamin K-dependent bone protein, osteocalcin. A radio-immunoassay for human osteocalcin was employed, and EDTA extractions of calcified atheromata, and aortic valves as well as relevant noncalcified material were analyzed. Tissue calcium levels were also determined, as were Gla levels as a measure of total vitamin K-dependent protein content. Osteocalcin was present at low levels in all calcified cardiovascular tissues (4.5-175.7 ng osteocalcin/mg protein) with trace levels or nondetectable amounts present in noncalcified tissue. Osteocalcin accounted for a small proportion of the total protein-bound Gla (0.01-0.05%). The relationship of osteocalcin to the other Gla-containing proteins of atherosclerotic plaque including atherocalcin, the principal extractable Gla-containing protein of calcified plaque, is discussed.

Atherocalcin, a γ-carboxyglutamic acid containing protein from atherosclerotic plaque

Abstract The specialized calcium binding amino acid, γ-carboxyglutamic acid (Gla) is quantitated in developing atherosclerotic plaque relative to progression of the disease, and a Gla-containing protein isolated from calcified atherosclerotic plaque is partially characterized. Low levels of Gla are found in fatty streak and fibrous plaque lesions, and a marked increase in Gla content occurs in calcified plaque. A unique Gla-containing protein is purified from 0.5M EDTA (pH 8.0) extracts of calcified plaque, named atherocalcin. The protein containing 19 Gla residues/1000 amino acids is 80,000 molecular weight, with a pI of 4.16 – 4.3 and is uniquely different from other known Gla-containing proteins. The implications of this work for the further understanding of the pathogenesis and therapy of atherosclerosis are discussed.

Porcine Bioprosthetic Valve Calcification in Bovine Left Ventricle-Aorta Shunts: Studies of the Deposition of Vitamin K-Dependent Proteins

Calcification of glutaraldehyde-preserved bioprosthetic cardiac valves represents a serious clinical problem. Previous work from this laboratory has established the presence in clinical bioprosthetic valve calcifications of vitamin K-dependent calcium-binding proteins, which contain the calcium-binding amino acid gamma-carboxyglutamic acid; no proteins containing gamma-carboxyglutamic acid are present in nonmineralized valves. The purpose of the present study was to examine a series of bovine circulatory bioprosthetic valve explants for calcification and proteins containing gamma-carboxyglutamic acid. Biochemical analyses of explanted bioprosthetic valves from calves demonstrated proteins with gamma-carboxyglutamic acid accumulating in calcified valves during both the onset and progression of valve calcification; calcium levels in the explanted calf bioprostheses were in the same range as those noted in clinical material. Accumulation of calcium and protein with gamma-carboxyglutamic acid occurred simultaneously and progressively, beginning 2 months after implantation. Small amounts of osteocalcin, the bone-derived protein containing gamma-carboxyglutamic acid, were present in both human and bovine bioprosthetic valve calcifications at comparable levels. No osteocalcin was detectable in non-mineralized valve tissue. Warfarin anticoagulant therapy did not prevent calcification or accumulation of protein with gamma-carboxyglutamic acid. It is concluded that proteins containing gamma-carboxyglutamic acid are involved in both the onset and progression of bioprosthetic valve calcification, and that conventional means of vitamin K antagonism do not alter this association or the course of bioprosthetic valve mineralization.

Immobilization of plasmid DNA on an anti-DNA antibody modified coronary stent for intravascular site-specific gene therapy

The aim of the present study was to investigate the incorporation of plasmid DNA (pDNA) onto a coronary stent by chemo‐immunoconjugation for achieving site‐specific gene delivery.

871. Surface Modification of Adenovectors with TAT- and Antp-Containing Polymers Protects Them from Neutralizing Antibodies and Increases Expression of Reporter Genes In Vitro and In Vivo

Top of pageAbstract nThe relative inefficiency of adenovirus (Ad)-based transduction vectors in cardiovascular tissues is related to the low expression of the CAR receptor on the surface of endothelial and smooth muscle cells as well as to the rapid neutralization of Ad by anti-Ad antibodies. Polymer modification of Ad surface potentially addresses these problems by masking Ad particles from neutralizing antibodies and providing a convenient platform for the conjugation of ligands possessing transduction-facilitating properties, such as TAT and Antp peptides. Our present studies investigated the hypothesis that by surface modifying Ad using photoreactive polymers to attach either TAT or Antp increased transduction could be achieved while reducing susceptibility to neutralizing antibodies. This was studied with reporter gene transduction with surface-modified Ad vectors in cell culture and in vivo, in rat carotid arteries. n nA series of fluorescently labeled polyallylamine (PAA)-based photochemically activatable (either via benzophenone, or azide chemistry) thiol reactive linear polymers were custom synthesized and conjugated with TAT-SH or Antp-SH. Ad-GFP or Ad-Luc were suspended in PBS x5 and admixed with either TAT-, Antp-derivatized or unmodified polymers to obtain the final polymer/Ad ratios ranging from 1:60 to 1:10 (w/w). Cellular uptake of vectors and ensuing transgene expression in rat aortic smooth muscle cell line (A10), primary bovine aortic endothelial cells (BAEC) and human alveolar epithelial cell line (A549) were examined by the fluorescence microscopy and fluorometry. Additionally, na|[iuml]|ve and surface- modified Ad vectors were exposed to escalating concentrations of neutralizing anti-knob antibodies (50 to 500 nM) prior to the cell infection. 6|[times]|108 pfu of either na|[iuml]|ve or polymeric TAT-conjugated Ad-Luc were locally delivered to the balloon-injured common carotid arteries of SD rats (n=6). The transgene expression was assessed 2 days after gene delivery by bioluminescence imaging. n nModel experiments have shown faster kinetics of photochemical polymer attachment to the capsid proteins with azide-compared to benzophenone-containing polymers. The surface modification of Ad-GFP with TAT- and Antp-derivatized polymers resulted in 3|[ndash]|10-fold increase of GFP expression in the studied cell types. 15 min incubation of na|[iuml]|ve Ad with neutralizing antibodies resulted in a 94% decrease of GFP expression in A10 cells, while the surface modification of Ad with TAT- and Antp-containing polymers bestowed protection against neutralizing antibodies in proportion with the extent of surface modification (59%, 14% and 6% inhibition for the polymer/Ad ratios of 1:60, 1:30 and 1:10, respectively). In the initial in vivo experiments local delivery of polymeric TAT- conjugated Ad resulted in 4.4-fold enhancement of luciferase activity in comparison with unmodified Ad. n nIn conclusion, light-induced covalent conjugation of TAT- and Antp-modified PAA-based polymers to Ad capsid proteins protects adenovectors from preformed neutralizing antibodies and significantly augments gene transduction both in vitro and in vivo.

An Ultra-Thin Polymer Coating for the Tethering of Adenoviral Vector to the Surface of Coronary Stents

Our group has previously demonstrated stent-based gene delivery with either viral or plasmid vectors. However, these previous studies utilized bulky PLGA or collagen stent coatings, known to cause inflammatory reactions in stented arteries. In the present experiments we successfully attached adenoviruses either directly, or via anti-adenovirus antibodies to the steel surface of stents using chemical coordination with biphosphonates. Disciplines Biochemical and Biomolecular Engineering | Chemical Engineering | Engineering | Materials Science and Engineering | Polymer Science Author(s) Ilia Fishbein, Ivan S. Alferiev, Gordon Sek-Yin Wong, John M. Vohs, Jeanne M. Connolly, Robert L. Wilensky, and Robert J. Levy This technical report is available at ScholarlyCommons: http://repository.upenn.edu/cbe_papers/173

Biomaterials-Associated Pathology of Cardiac Valve Prostheses: Clinical Explant Analysis and Studies of Tissue Valve Calcification

Studies of the pathology of explanted clinical and experimental valve specimens were done to determine the role of biomaterials in the failure of heart valve prostheses. From 7/80 through 12/84, 129 patients had valve prostheses removed at our hospital. Thrombosis/thromboembolism was the most frequent cause of the 39 mechanical valve failures (23%). Structural failure (poor durability) caused 12%. In contrast, with tissue-derived bioprostheses (90 patients), primary tissue degenerative failure was frequent (70%), and was usually associated with calcification (CALC). Infection caused 19% of all valve failures with a frequency not differing among valve types. Other causes of failure were not biomaterials-related. Thus, although frank materials degradation necessitates only a minority of mechanical valve removals, the suboptimal biologic properties of contemporary heart valve biomaterials contribute to some frequent complications (such as thrombosis/thromboembolism). Furthermore, degeneration due to mineralization is the overwhelming mode of failure of bioprosthetic valves. To understand and potentially eliminate bioprosthetic valve CALC, the kinetics, morphology and host and implant determinants of this process were studied using subcutaneous implants in rats. In this model, in which the pathologic features were similar to those of actual experimental and clinical valve replacements, CALC of glutaraldehyde-treated porcine aortic valve and bovine pericardium began within 48 hours following implantation; mineral levels were comparable to those in failed clinical valves after 84–126 days. Calcific deposits were initially associated with connective tissue cells, but later involved collagen. Systemically-administered ethane-hydroxy-diphosphonate (EHDP), a drug used to treat some forms of metabolic bone disease, inhibited CALC by 97%, but caused significant bone growth retardation. However, EHDP given by site-specific controlled-release largely mitigated CALC at only 1% of the total effective systemic dose, but without adverse effects.

137 STUDIES ON THE PATHOGENESES Of CALCIFIC AORTIC VALUE DISEASE: THE ROLE OF THE CALCIUM BINDING AMINO ACID δ CARBOXYGLUTAMIC ACID

δ Carboxyglutamic acid (Gla) is a recently discovered calcium binding amino acid, which has been shown to be present in the vitamin K-dependent clotting factors and in a new bone protein, osteocalcin. In prothrombin Gla synthesis occurs in a post-trans lational and vitamin K dependent enzymatic carboxylation of specific glutamic acid (Glu) residues. This reaction is inhibited by Dicumarol and other vitamin K antagonists with a net effect on clotting of anticoagulation. Furthermore, Gla has been shown to be present in metastatic and dystrophic calcifications, occurring in calcified atherosclerotic plaque, kidney stones, and subcutaneous nodules in dermatomyositis and scleroderma. The purpose of the present study was to determine if Gla is present in calcified aortic valve tissue from patients with congenital aortic stenosis and in cases of calcified aortic valve homografts and xenografts. Pathologic aortic valve tissue was obtained at cardiac surgery from 5 patients. Normal aortic valves were obtained from 2 patients at autopsy, and the Truncal valve from a patient with a calcified aortic valve homograft was also studied. Valve specimens were rinsed with saline and dried. One calcified aortic valve was extracted with .5M EDTA (pH8). The tissue and the EDTA extract were next subjected to 2N KOH hydrolysis, Dowex-1 chromotography, and automated amino acid quantitation of Gla. Gla was detected in all calcified aortic valve tissues. No Gla was found in the normal aortic valve tissue or in the Truncal valve. We conclude that the specific calcium binding amino acid, Gla, may be linked to the pathologic calcification of aortic valves. If Gla containing proteins can be related to the pathogenesis of calcific aortic valve disease, then vitamin K antagonism with Dicumarol or related compounds might retard the calcification process.

260 STUDIES ON A NEW COAGULATION PARAMETER: DECREASED URINARY X CARBOXYGLUTAMIC ACID EXCRETION WITH WARFARIN THERAPY

Warfarin anticoagulation has been previously shown to result from inhibition of post-translational synthesis of γ carboxyglutamic acid (Gla) from glutamic acid residues in the vitamin K-dependent clotting factors. This recently discovered amino acid (Gla) is synthesized in the liver and is known to be excreted in urine largely as the free amino acid. Furthermore, previous work indicates that urinary Gla excretion is chiefly derived from prothrombin breakdown and turnover. The purpose of the present study was to monitor urinary Gla excretion in order to detect any significant differences between normal and anticoagulated patients. Fourteen patients were studied. Seven were on stable Warfarin anticoagulation therapy and 6 were controls. One patient was studied postoperatively during initial anticoagulant treatment. Free urinary Gla was measured in 24 hour collections using 2N KOH hydrolysis and automated amino acid analysis. Normal subjects excreted 18-27 micromoles (μM of Gla)/24 hours (hrs) while anticoagulated patients (Prothrombin time range 20-30 sec. control = 12.0) excreted only 2.7-13 μM/24 hrs. Gla excretion in the patient begun on Warfarin dropped progressively during the first 10 days of therapy (128 μM/24 Hrs. to 22 μM/24 hrs.) while Prothrombin time stablized by day 6 (range 19.0 sec. to 22.9 sec) We conclude that urinary Gla excretion is reduced by Warfarin, and that monitoring urinary Gla may prove useful in regulating anticoagulation.