Optimization of a recombinant von Willebrand factor fragment as an antagonist of the platelet glycoprotein Ib receptor.

The binding of von Willebrand factor (vWF) to platelet glycoprotein (GP) Ib receptor is one of the initial events in thrombus formation. Previous studies have shown that RG12986, a reduced and alkylated recombinant fragment of vWF (Ser445-Val733), can inhibit binding of native vWF to GP Ib and offers potential as an anti-thrombotic agent. We have now evaluated a series of deletion mutants of RG12986 and found that reduced and alkylated rvWF508-704 is close to the minimal sequence with optimal RG12986-like activity (IC50 for inhibition of GP Ib-dependent platelet aggregation in the absence of modulators: 0.022 microM +/- 0.01, n = 3) and that it too binds directly to GP Ib. Under in vitro conditions, with no exogenous modulators present and in the absence of shear stress, oxidized rvWF508-704 (containing a disulfide bond between Cys508 and Cys659) is approximately 5-fold less active than reduced and alkylated rvWF508-704; the two fragments, however, display comparable activity in the presence of the modulator botrocetin. The smaller rvWF508-704 fragment offers distinct advantages over RG 12986. In particular, removal of non-active NH2 and COOH terminal sequences may reduce the risk of antigenicity and may contribute to rendering the molecule mostly monomeric in solution, as opposed to the monomer-dimer equilibrium previously described for RG12986.

Production and functional characterization of a recombinant fragment of von Willebrand factor (vWF): an antagonist to platelet receptor GP Ib.

We expressed a recombinant peptide fragment (Ser445-Val733) of human von Willebrand factor (vWF), containing the binding domain for the platelet receptor of GP Ib, in E. coli. This 33 kD peptide blocks binding of the intact vWF molecule to GP Ib in the presence of modulators. Thus, it offers potential as an antithrombotic agent. High level expression was achieved in a plasmid construct driven by the bacteriophage T7 promoter. The peptide was solubilized from inclusion bodies in strong chaotrope, then reduced and alkylated. Following purification, formulation at pH 3.5, and lyophilization, the reconstituted experimental product (RG 12986) exists as an equilibrium of monomer and dimer species. When formulated above pH 5.0, soluble aggregates are formed; these solutions have less bioactivity than RG 12986. Interestingly, the non-aggregated state of RG 12986 remains conserved following dilution and incubation with platelet-poor plasma. The overall purification/low pH formulation strategies may be applicable to other E. coli recombinant proteins having a tendency to aggregate following removal of chaotrope near physiologic pH when in a concentrated format.

Uptake of low density lipoprotein, albumin, and water by deendothelialized in vitro minipig aorta.

The purpose of this research was to study the effect of pressure on arterial hydration in vitro and the effect of pressure and flow (stirred reagent) on the in vitro transport of 125I-albumin and 125I-LDL into deendothelialized minipig aortas over a 24-hour period. It was found that the arterial hydration (fractional mass of water) was 0.740 +/- 0.0043 SEM for control tissue; after 24 hours this rose to 0.745 +/- 0.0038 for 0 mm Hg, 0.752 +/- 0.0046 for 100 mm Hg, and 0.755 +/- 0.0065 for 200 mm Hg. In the transport studies, the following effects were found. The reagent radioactivity concentration and composition did not change with pressure or stirring over the 24-hour period. For unpressurized tissue, the 24-hour normalized uptake [uptake (M mg cm-2) divided by reagent concentration (co mg cm-3)] of albumin was (4.86 +/- 0.43 cm) X 10(-3) from stirred and (5.46 +/- 0.36 cm) X 10(-3) from nonstirred reagent; that of LDL was (0.31 +/- 0.02 cm) X 10(-3) from stirred and (0.37 +/- 0.02 cm) X 10(-3) from nonstirred reagent. Pressurization (100 mm Hg) of the tissue increased albumin uptake by 52% from stirred and by 125% from nonstirred reagent and the LDL uptake by 52% from stirred and 241% from nonstirred reagent. Pressure increased the intimal surface concentration of albumin and LDL at the nonstirred, but not at stirred, interfaces. The electrophoretic properties of the intimal surface fluid showed only minor differences from those of the bulk reagent. These data demonstrate that pressure causes a slight, but significant, increase in arterial hydration and that radiolabeled albumin and LDL appear to be sieved by the superficial intimal layers of the deendothelialized porcine aorta under the in vitro conditions of this study.

Effect of serum and stirring on diffusive 125I-albumin and Evans blue dye uptake.

The diffusive in vitro uptake of homologous 125I-albumin (MA, nmol.cm-2) and Evans blue dye (EBD) (ME, nmol.cm-2) by the deendothelialized canine aorta from serum and from a simple albumin solution with and without EBD and with and without vigorous stirring was measured in 18 preparations. The results show that 1) MA and ME were significantly smaller from serum than from a simple albumin solution, 2) vigorous stirring of the liquid phase caused a slight decrease (approximately 5%) in MA and increase (approximately 9%) in ME, 3) MA was not influenced by the presence of EBD, and 4) at least 90% of the radioactivity in the tissue was free 125I-albumin with an electrophoretic mobility identical to its nonlabeled cohort molecules and albumin in the original reagent. These observations confirm the identity of the tissue radioactivity with the labeled protein in the reagent, show that less than 10% of the labeled protein is irreversibly bound in the tissue, indicate that significant concentration gradients do not occur in the reagent phase, and indicate that albumin appears to interact with other plasma components in the reagent phase.

Cholesterol transfer from normal and atherogenic low density lipoproteins to Mycoplasma membranes.

The purpose of this study was to determine whether the free cholesterol of hypercholesterolemic low density lipoprotein from cholesterol-fed nonhuman primates has a greater potential for surface transfer to cell membranes than does the free cholesterol of normal low density lipoprotein. The low density lipoproteins were isolated from normal and hypercholesterolemic rhesus and cynomolgus monkeys, incubated with membranes from Acholeplasma laidlawii, a mycoplasma species devoid of cholesterol in its membranes, and the mass transfer of free cholesterol determined by measuring membrane cholesterol content. Since these membranes neither synthesize nor esterify cholesterol, nor degrade the protein or cholesterol ester moieties of low density lipoprotein, they are an ideal model with which to study differences in the cholesterol transfer potential of low density lipoprotein independent of the uptake of the intact low density lipoprotein particle. When added at an equivalent particle concentration, there was greater enrichment of membranes with free cholesterol from hypercholesterolemic low density lipoprotein. Hypercholesterolemic low density lipoprotein, however, contains more cholesterol per particle than normal low density lipoprotein; yet calculations on the basis of equivalent free cholesterol content showed no difference in either the rate or extent of free cholesterol transfer from normal or hypercholesterolemic low density lipoprotein. This was true for the transfer of at least 90% of the free cholesterol from both lipoproteins. These studies indicate that, even though there are marked differences in the cholesterol composition of normal and hypercholesterolemic low density lipoproteins, this does not result in a greater chemical potential for surface transfer of free cholesterol. Consequently, if a difference in the surface transfer of free cholesterol is responsible for the enhanced ability of hypercholesterolemic low density lipoprotein to promote cellular cholesterol accumulation and, perhaps, also atherosclerosis, it must be the result of differences in the interaction to the hypercholesterolemic low density lipoprotein with the more complicated mammalian cell membranes, rather than differences in the chemical potential for cholesterol transfer.

Metabolism by cells in culture of low-density lipoproteins of abnormal composition from non-human primates with diet-induced hypercholesterolemia.

Diet-induced hypercholesterolemia in non-human primates results in the production of a low-density lipoprotein (LDL) of abnormal size and composition. This LDL from hypercholesterolemic monkeys has been shown to be more atherogenic than the same amount of LDL from normocholesterolemic animals. Previous studies have demonstrated that hypercholesterolemic LDL is approximately twice as effective as normal LDL in stimulating cholesterol accumulation and esterification in arterial smooth muscle cells in culture. The purpose of the present study was determine whether this effect was secondary to differences in metabolism of the normal and hypercholesterolemic LDL. for this, the metabolism of 125I-labeled normal and hypercholesterolemic LDL from rhesus and cynomolgus monkeys was compared in several lines of skin fibroblasts and smooth muscle cells. Both normal and hypercholesterolemic LDL bound with high affinity to the same cell surface receptor. However, the affinity for binding of hypercholesterolemic LDL was about twice that of normal LDL (apparent dissociation constant for binding, Kd, was 2.63 micrograms protein/ml and 4.35 micrograms protein/ml, respectively). Conversely, only about 50% as many particles of hypercholesterolemic were able to bind to the receptor, compared with normal LDL. Those cells with the greatest capacity to metabolize LD generally accumulated the most cholesterol with either hypercholesterolemic or normal LDL. In all cell lines, nearly twice as much cholesterol accumulated in cells incubated with hypercholesterolemic LDL compared with normal LDL, and this differential could not be explained by differences in metabolism of the two lipoproteins, suggesting that some cholesterol entered the cells independent of the uptake of the intact LDL molecule. LDL receptors appear necessary for this to occur, since no difference in cholesterol accumulation was observed in cells genetically deficient in LDL receptors.