The effect of shear rate and fibrinogen concentration on adenosine diphosphate-induced aggregation of suspensions of washed human platelets in Poiseuille flow at 23 degrees C was studied using a previously described double infusion technique and resistive particle counter size analysis. Using suspensions of multiple-centrifuged and -washed cells in Tyrodes-albumin [3 x 10(5) microliters-1; (17)] with [fibrinogen] from 0 to 1.2 microM, the rate and extent of aggregation with 0.7 microM ADP in Tyrodes-albumin were measured over a range of mean transit times from 0.2 to 43 s, and at mean tube shear rates, G, = 41.9, 335 and 1,335 s-1. As measured by the decrease in singlet concentration, aggregation at 1.2 microM fibrinogen increased with increasing G up to 1,335 s-1, in contrast to that previously reported in citrated plasma, in which aggregation reached a maximum at G = 335 s-1. Without added fibrinogen, there was no aggregation at G = 41.9 s-1; at G = 335 s-1, there was significant aggregation but with an initial lag time, aggregation increasing further at G = 1,335 s-1. Without added fibrinogen, aggregation was abolished at all G upon incubation with the hexapeptide GRGDSP, but was almost unaffected by addition of an F(ab')2 fragment of an antibody to human fibrinogen. Aggregation in the absence of added fibrinogen was also observed at 37 degrees C. The activation of the multiple-washed platelets was tested using flow cytometry with the fluorescently labelled monoclonal antibodies FITC-PAC1 and FITC-9F9. It was shown that 57% of single cells in unactivated PRT expressed maximal GPIIb-IIIa fibrinogen receptors (MoAb PAC1) and 54% expressed pre-bound fibrinogen (MoAb 9F9), with further increases on ADP activation. However, incubation with GRGDSP and the F(ab')2 fragment did not inhibit the prebound fibrinogen. Moreover, relatively unactivated cells (8% expressing receptor, 14% prebound fibrinogen), prepared from acidified cPRP by single centrifugation with 50 nM of the stable prostacyclin derivative, ZK 36,374, and resuspension in Tyrodes-albumin at 5 x 10(4) microliters-1, aggregated with 2 and 5 microM ADP at G = 335 and 1,335 s-1 in the absence of added fibrinogen. We therefore postulate that a protein such as von Willebrand factor, secreted during platelet isolation or in flow at sufficiently high shear rates, may yield the observed shear-rate dependent aggregation without fibrinogen.