Flow cytometric detection of the redistribution of the glycoprotein Ib-IX complex on thrombin-stimulated platelets is dependent on the type of antibody conjugate used.

Flow cytometric measurement of changes in platelet surface membrane antigens is gaining acceptance as a method for assessing platelet activation status both in vitro and in clinical studies.' One effect of the strong agonist, thrombin, is the reciprocal upregulation of the glycoprotein (GP) IIb-IIIa complex from intracellular membrane stores, and the loss of around 50% of the GPIb-IX complex from the platelet surface. Michelson et al, in this Journal, have reported the detection of such changes, by flow cytometry, in washed platelets' and in platelets analysed in a whole blood assay.3 Loss of the GPIb-IX complex may be a useful marker of platelet changes in vivo. Reduced expression of GPIb-IX has been reported in blood emerging from bleeding time wounds' and after exercise! Whole blood flow cytometry, which analyzes diluted samples of unseparated blood, has advantages over fixed, washed platelet methods in that it is rapid, and it avoids manipulations that might introduce artefactual activation of the platelets, as for example during centrifugation or separation procedures.' Platelet response to agonists can be monitored in these assays, including the response to thrombin. Aggregation of the platelets can be prevented by addition of glycyl-L-prolyl-larginylL-proline-acetate peptide (GPRP) to inhibit cross-linking of fibrin.3 In such assays, directly conjugated antibodies are often used to minimize assay time and sample processing. When using a monoclonal antibody (MoAb) to GPIb (RFGP37), raised in this laboratorys and conjugated to fluorescein isothiocyanate (FlTC), we were surprised to find no reduction in GPIb-IX expression on thrombin-stimulated platelets, despite increases in GPIIb-IIIa expression and the appearance of the or-granule membrane antigen, P-selectin/GMP-140 (Fig 1). We had previously seen downregulation of GPIb in washed, thrombin-activated, fixed platelets using the same MoAb in an indirect flow cytometric assay. Epitope variability was not suspected. Michelson et al's studies clearly demonstrated that a number of MoAbs to different epitopes on the GPIb-IX complex gave identical res~l ts . '~~ Therefore, we looked at the methods used. Both Michelson and ourselves used essentially identical whole blood assays, but while we used an MoAb directly conjugated to FITC, Michelson's study employed biotinylated MoAbs, visualized with a streptavidin-phycoerythrin (ST-PE) conjugate. When we repeated our investigations with biotinylated RFGP37 and ST-PE, a decrease in GPIb was clearly seen, concomitant with an increase in the expression ofGPIIbIIIa and the neo-expression of GMP140 as illustrated in Fig 1. We postulate that the reason for the discrepancy in the results obtained with these different reagents lies in the location of the internalized GPIb-IX complexes that are lost from the platelet surface. Electron microscope studies have demonstrated that they become translocated to the surface connecting canalicular system (SCCS)? GPIb-IX complexes in the SCCS may be accessible to the MoAb-