No abstract

GMP-140 is a 140-kD granule membrane protein, found in the alpha granules of platelets and the Weibel-Palade bodies of endothelial cells, that is surface expressed on cell activation and mediates neutrophil attachment. Cloning data for GMP-140 from an endothelial library predict a soluble form of the protein, the transcription message for which is also found in platelets. In this study, we report the detection by enzyme-linked immunosorbent assay of soluble GMP-140 in plasma centrifuged for 3 h at 100,000 g (to remove platelet microparticles) and confirm its identity by purification from plasma. Plasma concentrations were found to be 0.251 +/- 0.043 micrograms/ml (means +/- SD, n = 10) in normal male controls and 0.175 +/- 0.063 micrograms/ml (means +/- SD, n = 10) in normal female controls. The purified protein had an identical molecular mass (nonreduced) to platelet membrane GMP-140 (approximately 3 kD lower, reduced) and was immunoblotted by polyclonal anti-GMP-140, and the anti-GMP-140 monoclonal antibodies AK4 and AK6. Analytical gel filtration studies indicated that the plasma GMP-140 eluted as a monomer whereas detergent-free, platelet membrane GMP-140 eluted as a tetramer consistent with plasma GMP-140 lacking a transmembrane domain. Purified plasma GMP-140 bound to the same neutrophil receptor as the membrane-bound form, and when immobilized on plastic, bound neutrophils equivalently to immobilized platelet membrane GMP-140. Since it has been shown that fluid-phase GMP-140 is antiinflammatory and downregulates CD18-dependent neutrophil adhesion and respiratory burst, its presence in plasma may be of major importance in preventing the inadvertent activation of neutrophils in the circulation.

Human platelet glycoprotein Ib (GP Ib) i s a major integral membrane protein that has been identified as the platelet-binding site mediating the factor VIII/von Willebrand-factor-dependent adhesion of platelets to vascular subendothelium. Recent evidence suggests that G P Ib is normally coinplexed with another platelet membrane protein, GP IX. In this study, human platelet plasma membranes were selectively solubilized with a buffer containing 0.1% (v/v) Triton X-100. The GPIb complex (GPIb plus GPIX) was purified to homogeneity in = 30% yield by immunoaffinity chromatography of the membrane extract using the anti-(glycoprotein Ib complex) murine monoclonal antibody, WM 23, coupled to agarose. G P Ib and GP IX were subsequently isolated as purified components by immunoaffinity chromatography of the GP Ib complex using a second anti-(glycoprotein Ib complex) monoclonal antibody, FMC 25, coupled to agarose. As assessed by dodecyl sulphate/ polyacrylamide gel electrophoresis, purified GP Ib was identical to the molecule on intact platelets and had an apparent relative molecular mass of 170000 under nonreducing conditions and 135000 (cr subunit) and 25000 (p subunit) under reducing conditions. GP IX had an apparent M , of 22000 under both nonreducing and reducing conditions. Purified Gb Ib complex and GP Ib inhibited the ristocetin-mediated, human factor VIII/von Willebrand-factor-dependent and bovine factor VIII/von Willebrand-Factor-dependent agglutination of washed human platelets suggesting the proteins had been isolated in functionally active form. G P Ib, had a similar amino acid composition to that previously reported for its proteolytic degradation product, glycocalicin. The amino acid compositions of GP Ib, and GP IX were similar but showed marked differences in the levels of glutamic acid, alanine, histidine and arginine. The N-termini of GP Ib, and G P 1X were blocked; G P Ib, had the N-terminal sequence, Ile-Pro-Ala-Pro-. On crossed immunoelectrophoresis, both GP Ib and GP IX were found to occur in the same immunoprecipitin arc(s) whether the platelets had been solubilized in the absence or presence of the calcium-dependent protease inhibitor, leupeptin. Binding studies in platelet-rich plasma indicated a similar number of binding sites (2 I: SD) for three anti-(glycoprotein Ib complex) monoclonal antibodies: AN 51, epitope on GP Ib, (22000 I: 2700, n = 3), WM 23, epitope on G P Ib, (21 000 3400, n = 3), FMC 25, epitope on GP IX (20100 f 2700, n = 3), and FMC 25 (Fab')2 (27100 800, n = 2). The combined evidence suggests that GP Ib is normally bound to GP IX and that the stoichioinetry of this complex is 1 : 1.The initial event in hemostasis in response to vascular injury involves the adhesion of platelets to the exposed vascular subendothelium. The adhesion reaction appears to be dependent on three components [I]: a receptor on platelets which on the basis of a variety of evidence has been defined as the platelet membrane protein, glycoprotein Ib (GP Ib); a plasma component, factor VIII/von Willebrand fac...

Interaction of von Willebrand factor (vWF) with its platelet receptor only occurs in vitro in the presence of a modulator such as ristocetin. We have recently confirmed that the human platelet membrane glycoprotein (GP) Ib-IX complex is the receptor involved in the ristocetin-dependent binding of vWF by reconstitution with the purified components [Berndt, M.C., Du, X., & Booth, W.J. (1988) Biochemistry 27, 633-640]. We have now developed a similar solid-phase reconstitution assay using an alternate modulator, botrocetin, for the competitive analysis of functional domains in both vWF and the GP Ib-IX complex. Botrocetin was purified from Bothrops jararaca venom by ammonium sulfate fractionation and subsequent DEAE-cellulose and hydroxylapatite chromatography. The purified protein was a 25-kilodalton (kDa) disulfide-linked dimer with apparent subunit molecular weights of 14,000 and 14,500. Binding studies with immobilized botrocetin demonstrated that botrocetin bound to vWF and to a 52/48-kDa region of vWF that contains the GP Ib binding domain, but not to glycocalicin, a proteolytic fragment of GP Ib that contains the vWF binding site. Binding of 125I-labeled vWF to GP Ib-IX complex coated beads and to platelets was strictly botrocetin-dependent with half-maximal binding at a botrocetin concentration of congruent to 0.27 microM. Botrocetin-dependent binding of vWF was specific, saturable, and comparable to that observed with ristocetin. An anti-vWF monoclonal antibody, 3F8, inhibited ristocetin- but not botrocetin-dependent binding of vWF, suggesting the presence of distinct ristocetin and botrocetin modulator sites on vWF. The botrocetin reconstitution assay was at least an order of magnitude more sensitive than the corresponding ristocetin assay for the competitive analysis of functional domains on both vWF and the GP Ib-IX complex and has confirmed the localization of the vWF-binding domain to the 45-kDa N-terminal region of GP Ib.