Transient increase in intracellular pH during Dictyostelium differentiation.

The intracellular pH (pHi) of Dictyostelium discoideum amebae has been determined using the pH-dependent fluorescence of intracellularly trapped fluorescein (Thomas, J. A., R. N. Buschbaum, A. Zimiak, and E. Racker, Biochemistry, 18:2210-2218). The pHi of cells measured 45-60 min after initiation of differentiation was between 6.2 and 6.3. At approximately 2 h into differentiation cells underwent a transient intracellular alkalinization during which the pHi rose to 7.13 (+/- 0.3, n = 4), after which the pHi returned to approximately the original value (6.2-6.4). Cells that were removed from growth medium but were incubated in differentiation medium containing 3% dextrose did not exhibit this transient increase in pHi. The alkalinization event can also be prevented from occurring by differentiation in Na+-free solutions or by the addition of amiloride to sodium-containing buffer solutions, suggesting that the alkalinization is sodium dependent. When the alkalinization was prevented by amiloride treatment, cells did not progress normally into differentiation. This increase in pHi was initiated by the cells 2 h after removal from nutrient medium and it could be inhibited by several treatments that had been observed to delay the differentiation program, suggesting that it plays a major role in the initiation of the developmental program of this organism.

Calmodulin confers calcium sensitivity on ciliary dynein ATPase.

Extraction of demembranated cilia of Tetrahymena by Tris-EDTA (denoted by the suffix E) yields 14S-E and 30S-E dyneins with ATPase activities that are slightly increased by Ca++. This effect is moderately potentiated when bovine brain calmodulin is added to the assay mixture. Extraction with 0.5 M KCl (denoted by the suffix K) yeilds a 14S-K dynein with a low basal ATPase activity in the presence of Ca++. Subsequent addition of calmodulin causes marked activation (up to 10-fold) of ATPase activity. Although 14S-K and 14S-E dyneins have Ca++-dependent ATPase activities that differ markedly in the degree of activation, the concentration of calmodulin required for half-maximal saturation is similar for both, approximately 0.1 microM. Both 30S-K and 30S-E dyneins, however, require approximately 0.7 microM bovine brain calmodulin to reach half-maximal activation of their Ca++-dependent ATPase activities. Tetrahymena calmodulin is as effective as bovine brain calmodulin in activating 30S dynein , but may be slightly less effective than the brain calmodulin in activating 14S dynein. Rabbit skeletal muscle troponin C also activates the Ca++-dependent ATPase activity of 30S dynein and, to a lesser extent, that of 14S dynein, but in both cases is less effective than calmodulin. The interaction of calmodulin with dynein that results in ATPase activation is largely complete in less than 1 min, and is prevented by the presence of low concentrations of ATP. Adenylyl imidodiphosphate can partially prevent activation of dynein ATPase by calmodulin plus Ca++, but at much higher concentrations than required for prevention by ATP. beta, gamma-methyl-adenosine triphosphate appears not to prevent this activation. The presence of Ca++-dependent calmodulin-binding sites on 14S and 30S dyneins was demonstrated by the Ca++-dependent retention of the dyneins on a calmodulin-Sepharose-4B column. Gel electrophoresis of 14S dynein that had been purified by the affinity-chromatography procedure showed that presence of two major and one minor high molecular weight components. Similar analysis of 30S dynein purified by this procedure also revealed on major and one minor high molecular weight components that were different from the major components of 14S dynein. Ca++-dependent binding sites for calmodulin were shown to be present on axonemes that had been extracted twice with Tris-EDTA or with 0.5 M KCl by the use of 35S-labeled Tetrahymena calmodulin. It is concluded that the 14S and 30S dyneins of Tetrahymena contain Ca++-dependent binding sites for calmodulin and the calmodulin mediates the Ca++-regulation of the dynein ATPases of Tetrahymena cilia.

Identification of a platelet membrane glycoprotein as a falciparum malaria sequestration receptor.

Infections with the human malaria parasite Plasmodium falciparum are characterized by sequestration of erythrocytes infected with mature forms of the parasite. Sequestration of infected erythrocytes appears to be critical for survival of the parasite and to mediate immunopathological abnormalities in severe malaria. A leukocyte differentiation antigen (CD36) was previously suggested to have a role in sequestration of malaria-infected erythrocytes. CD36 was purified from platelets, where it is known as GPIV, and was shown to be a receptor for binding of infected erythrocytes. Infected erythrocytes adhered to CD36 immobilized on plastic; purified CD36 exhibited saturable, specific binding to infected erythrocytes; and purified CD36 or antibodies to CD36 inhibited and reversed binding of infected erythrocytes to cultured endothelial cells and melanoma cells in vitro. The portion of the CD36 molecule that reverses cytoadherence may be useful therapeutically for rapid reversal of sequestration in cerebral malaria.

Reduced thrombin binding and aggregation in Bernard-Soulier platelets.

Platelets from two patients with Bernard-Soulier disease showed a reduction in their ability to bind human thrombin. Thrombin binding studies in the high affinity range showed 1,500 sites for the Bernard-Soulier platelets as against 4,000 for normal controls. However, the dissociation constant was the same for both normals and patients (4.4 nM) indicating identical affinity for thrombin at the available sites. In the low affinity range, the Bernard-Soulier platelets showed 8,800 thrombin binding sites as against 24,000 for the controls, but again with identical values of Kd (37 nM). In addition, platelets from these Bernard-Soulier patients showed a decreased rate of aggregation with thrombin at both optimal (300 mU/ml) and suboptimal (60 and 120 mU/ml) thrombin concentrations. The decreased amount of thrombin which can bind to Bernard-Soulier platelets and the decrease in thrombin-induced aggregation may partly explain the hemostatic defect in these patients. In addition, the identical ratios of high affinity and low affinity binding sites in normals and in patients (0.37 and 0.36, and 0.36, respectively) supports the idea of a single class of binding sites for thrombin on the platelet surface.

Cell surface sialylation of two human tumor cell lines and its correlation with their platelet-activating activity.

The relationship between cell surface sialylation and platelet-activating activity was studied in two tumor cell lines of human origin, the SKNMC neuroblastoma line and the U87MG glioblastoma line. Their platelet-activating activity was evaluated in two different experimental systems, one that measures platelet aggregation and the other that quantifies platelet thrombus formation on vascular subendothelium under flow conditions. Our results demonstrate that, for the SKNMC line, the loss of 30% of surface sialic acid induced a significant reduction in its platelet-activating capacity. Upon recultivation desialylated SKNMC cells did not regenerate surface sialic acid and did not restore their initial values of platelet aggregation and platelet thrombus formation. Conversely, removal of 35% sialic acid from the surface of U87MG cells did not affect their pattern of platelet activation in either system tested. These results demonstrate that there is a correlation between cell surface sialylation and the capacity of SKNMC cells to activate platelets. The lack of effect of desialylation on U87MG-induced platelet activation indicates that different surface components may be the modulators of the interactions of these tumor cells with platelets. Our results support the hypothesis that heterologous mechanisms regulate platelet-tumor cell interactions and that tumor cell sialic acid may be only one of the aspects involved in such interactions.

Inhibitory effects of dipyridamole on growth, nucleoside incorporation, and platelet-activating capability in the U87MG and SKNMC human tumor cell lines.

The effects of dipyridamole on tumor cell function were examined in cultures of two lines of human origin, the SKNMC neuroblastoma line that activates platelets by a mechanism which is dependent on the release of adenosine 5'-diphosphate and the U87MG glioblastoma line that induces platelet activation by the generation of thrombin. Cells grown in the presence of dipyridamole at 1 microM showed greater than 80% inhibition of uptake of adenosine, thymidine, and uridine with both lines. At 5 microM tumor cell growth was inhibited by 70% (U87MG) and 90% (SKNMC) but without concomitant cytotoxicity as determined by clonogenic assay (50% inhibitory concentration approximately 20 microM). At 10 microM dipyridamole cyclic adenosine 3':5'-monophosphate levels increased 150% with both cell lines but no changes above baseline values were seen at 2.5 microM. The two cell lines showed different responses to being cultured in the presence of dipyridamole in terms of their ability to subsequently activate platelets. U87MG cells cultured in 10 microM dipyridamole showed a doubling of the lag time as compared with cells grown in the absence of dipyridamole but with full aggregation; with SKNMC cells the aggregation rate was reduced and cells grown in 10 microM dipyridamole showed no reversible first wave, a 5-fold increase in lag time and a 75% inhibition in total aggregation. Since therapeutic doses of dipyridamole result in plasma concentrations of approximately 3.5 microM these results suggest that potential antimetastatic effects of dipyridamole could be direct arising from inhibition of important steps in tumor cell metabolism or indirect by suppressing one or more of the mechanisms involved in the ability of tumor cells to activate platelets.

Synergistic effects of lectins in the interaction of thrombin with human platelets.

Several lectins have been studied for their effects on the interaction of thrombin with human platelets. Wheat germ agglutinin, concanavalin A and Ricinus communis lectin increased the number of high affinity sites for diisopropylphosphothrombin on washed platelets from 3000 to about 12 000 but the binding affinities were unchanged (Kd approx 4 nM). Two other lectins, Lens culinaris and Bandieria simplicifolia, were without effect. (2) Using formalinized platelets to avoid possible complications of the platelet release reaction, wheat germ agglutinin showed a marked increase (5-fold) in the binding of active thrombin, peanut agglutinin had no effect while Ricinus communis and :Bandieria simplicifolia showed marginal increases (2-fold). Thrombin binding was decreased to about one quarter with Lens culinaris, Phaseolus vulgaris and concanavalin A. (3) Wheat germ agglutinin caused a synergistic increase of platelet aggregation at low concentrations of thrombin (12.5 mU/ml) and ADP (1 microM), both in the absence and presence of added fibrinogen, but had no effect on ristocetin-induced aggregation.

Effects of N-acetylglucosamine and platelet inhibitors on the synergistic interaction of platelets and aggregating agents in the presence of wheat germ agglutinin.

Low concentrations of wheat germ agglutinin (4 micrograms/ml) have been shown to act synergistically to induce platelet aggregation with epinephrine, collagen, arachidonate and ionophore A23187. Aggregation ceased on the addition of the haptenic sugar N-acetylglucosamine at any time following the onset of aggregation with these agonists and a small degree of disaggregation was observed during the reversible first wave with the biphasic aggregating agents epinephrine and ADP. Cyclooxygenase inhibitors such as indomethacin and aspirin blocked the second wave of aggregation with the biphasic aggregating agents epinephrine and ADP but a synergistic response continued to be shown with the first wave in the presence of these inhibitors. Release of [14C]serotonin and the mobilization of [3H]arachidonate by epinephrine and collagen were markedly stimulated in the presence of wheat germ agglutinin but there was no increase of either radiolabel in the case of ADP. Platelet shape change, but not aggregation, occurred with low levels of wheat germ agglutinin and the synergistic response with ADP, collagen or ionophore A23187 occurred without further shape change. Wheat germ agglutinin did not affect the basal or stimulated levels of cyclic AMP. The membrane fluidity of platelets was not affected by the lectin or by thrombin as shown by the lack of change in fluorescence polarization with diphenylhexatriene. It is suggested that the binding of wheat germ agglutinin to the platelet surface induces platelet activation by mechanisms similar to those of other agonists and that it may affect the distribution of membrane-bound Ca2+ by a reversible perturbation of the platelet membrane.

Characterization of type III intermediate filament regulatory protein target epitopes: S-100 (beta and/or alpha) binds the N-terminal head domain; annexin II2-p11(2) binds the rod domain.

We have investigated the interaction of S-100 proteins (beta and/or alpha) and annexin II2-p11(2) with glial fibrillary acidic protein (GFAP) and desmin to have further information on the mechanisms whereby S-100 proteins and annexin II2-p11(2) affect assembly/disassembly of GFAP and desmin intermediate filaments (IFs). Analyses were conducted on either native IF subunits, GFAP or desmin rod domain, or headless GFAP or desmin. Our data indicate that: (i) S-100 proteins bind to GFAP and desmin N-terminal head domain; (ii) annexin II2-p11(2) binds to GFAP rod domain; (iii) annexin II2-p11(2) does not interact with desmin nor affects desmin assembly. The present data suggest that the ability of S-100 proteins to inhibit GFAP and desmin assemblies and to promote the disassembly of preformed GFAP and desmin IFs depends on occupation of a site on the N-terminal head domain of these IF subunit. It is known that the N-terminal head domain is critical for the progression from the stage of GFAP and desmin dimers/tetramers to that of large oligomers. On the other hand, the ability of annexin II2-p11(2) to stimulate GFAP assembly under conditions where this latter is normally hampered (e.g., at alkaline pH values) might depend on annexin II2-p11(2)-induced changes in the structure of GFAP rod domain, possibly as a consequence of charge modifications. By contrast, the inability of annexin II2-p11(2) to bind to desmin would depend on desmin resistance to charge modifications.

Adenine nucleotide binding and photoincorporation in Glanzmann's thrombasthenia platelets.

Adenosine 5'-(1-thiotriphosphate) (ATP alpha S) binds to about 25,000 high affinity sites in platelets (Kd approximately 3 nM), competes fully in inhibiting the binding of ADP and, despite the absence of a specific photoactivatable substituent, is directly photoincorporated into a specific 18 kDa domain beginning at Tyr-198 in the alpha chain of glycoprotein IIb (GPIIb alpha) following ultraviolet irradiation of fresh unfixed platelets (Greco et al. (1991) J. Biol. Chem. 266, 13627-13633). 8-azido ATP has now been shown to have similar binding parameters (Kd 8 nM, 20,000 sites/platelet) but, in this case, photoincorporation occurred equally in GPIIb and GPIIIa. To determine the possible function of GPIIb alpha in ADP-induced activation, platelets were isolated from two Glanzmann's thrombasthenia patients whose platelets contain approximately 6% of normal levels of GPIIb. ADP and ATP alpha S bound to intact, formaldehyde-fixed Glanzmann's platelets at high affinity sites with dissociation constants of approximately 30 nM and approximately 2 nM, respectively. Both nucleotides also bound to low affinity sites with dissociation constants of approximately 2 microM: these values are similar to those obtained with control platelets. ATP alpha S antagonized the shape ADP-induced shape change response of Glanzmann's platelets (EC50 5 microM) indicating that it bound to the P2T (ADP) receptor. However, photoincorporation was low (approximately 7% of control) similar to their content of GPIIb alpha. These results show that ADP binding and photoincorporation are occurring at different sites on the platelet surface but suggest that the ADP binding site may be located in proximity to GPIIb alpha.

Characterization of human platelet UDPglucose-collagen glucosyltransferase using a new rapid assay.

A rapid and specific assay has been developed for UDPglucose-collagen glucosyltransferase (UDPglucose: 5-hydroxylysine-collagen glucosyltransferase, EC 2.4.1.66) using galactosylhydroxylysine (Gal-Hyl) as acceptor. Studies with intact human platelets and isolated plasma membranes indicated that about 5--10% of the total activity was surface bound and the rest was of cytoplasmic origin. The two forms of the enzyme had similar broad pH optima (6.5--8.0), Km values for UDPglucose (5 muM) and Gal-Hyl (approx. 4 mM) and for optimal manganese concentrations (25 mM). The soluble form of the enzyme was purified 80-fold. The reaction mechanism was determined as being rapid equilibrium random BiBi + dead end complex or ordered BiBi with UDPglucose being the first substrate to bind. Using Gal-Hyl bound in purified alpha 1 chain of chick skin collagen, a Km value three orders of magnitude less (2 muM) was found than for free Gal-Hyl and the manganese requirement decreased to 2 mM. These results suggest that the binding to the enzyme of Gal-Hyl in the collagen molecule is enhanced by the presence of the protein portion so that the enzyme may be capable of recognizing not only the carbohydrate side chains but also the primary structure of collagen.

Phencyclidine inhibits epinephrine-stimulated platelet aggregation independently of high affinity N-methyl-D-aspartate (NMDA)-type glutamatereceptors.

The psychotomimetic analgesic phencyclidine (PCP), which binds to a high affinity site on the neuronal N-methyl-D-aspartate (NMDA)-sensitive glutamate receptor, has previously been found to bind to platelets with high affinity and to specifically delay the onset of epinephrine-stimulated platelet aggregation (Jamieson et al. (1992) Biochem. J. 285, 35-39). We have now shown that the rank order of binding affinities of 14 synthetic PCP analogs at the high affinity binding site on platelets does not parallel the rank order of their affinities in binding to rat brain membranes, indicating that the high affinity PCP binding sites in platelets is distinct from the neuronal NMDA receptor. The order of potency of six of these analogs in delaying the onset of epinephrine-stimulated platelet aggregation also did not parallel the rank order of their binding affinities for platelet or brain binding sites. These data indicate that the ability of PCP analogs to inhibit epinephrine-stimulated aggregation is not related to their ability to bind to the high affinity platelet PCP binding site. Furthermore, (+)MK-801, which binds to the same high affinity binding site in neurons as does PCP, failed to inhibit epinephrine-stimulated platelet aggregation, further suggesting that the site at which PCP acts in platelets is not related to the NMDA-type glutamate receptor. Further studies showed that 5-HT2 receptors and effects on platelet secretion are not involved in PCP-mediated inhibition of epinephrine-induced platelet aggregation.

The distribution of collagen:glucosyltransferase in human blood cells and plasma.

Collagen:glucosyltransferase (UDP-glucose:5-hydroxylysine-collagen glucosyltransferase, EC 2.4.1.66) present in platelets, plasma, granulocytes and lymphocytes has been compared in order to determine whether the platelet enzyme has unique properties or distribution which would support a possible role in platelet-collagen interaction. The enzyme was purified 5400-fold from human plasma and 4400 from human platelets. The two enzymes were similar in terms of Km values for reacting with galactosylhydroxylysine (2.75 mM) and UDPglucose (7.4 microM), optimal Mn2+ concentration (10--15 mM) and pH optimum (7.0). The enzyme was not detectable in red cells. As in platelets, the enzyme was detected in membrane-bound and soluble forms in lymphocytes and granulocytes. Identical mobilities were obtained after elution following polyacrylamide gel electrophoresis of the enzymes from plasma, platelets, granulocytes and lymphocytes. These studies do not support a unique role for the collagen:glucosyltransferase of platelets in platelet-collagen interaction.

S-100 (alpha and beta) binding peptide (TRTK-12) blocks S-100/GFAP interaction: identification of a putative S-100 target epitope within the head domain of GFAP.

Alignment of previously characterized S-100 (alpha and beta)-binding peptides (J. Biol. Chem. 270, 14651-14658) has enabled the identification of a putative S-100 target epitope within the head domain of glial fibrillary acidic protein (GFAP). The capacity of a known peptide inhibitor of S-100 protein (TRTK-12), homologous to this region, to perturb the interaction of S-100 (alpha and beta) and GFAP (J. Biol. Chem 268, 12669-12674) was investigated. Fluorescence spectrophotometry and chemical cross-linking analyses determined TRTK-12 to disrupt S-100:GFAP interaction in a dose- and Ca(2+_dependent manner. TRTK-12 also inhibited S-100's ability to block GFAP assembly and to mediate disassembly of preformed glial filaments. Each of these events was strictly dependent upon the presence of calcium and inhibitory peptide, maximal inhibition occurring at a concentration of TRTK-12 equivalent to the molar amount of S-100 monomer present. Together with our recent report demonstrating TRTK-12 also blocks the interaction of S-100 protein with the actin capping protein, CapZ, these results suggest TRTK-12 functions as a pleiotropic inhibitor of S-100 function. Availability of a functional inhibitor of S-100 will assist the further characterization of S-100 protein function in vitro and in vivo. Moreover, this report provides additional evidence supportive of a role for S-100 as a multi-faceted regulator of cytoskeletal integrity.

Identification of the amino acids comprising a surface-exposed epitope within the nucleotide-binding domain of the Na+,K(+)-ATPase using a random peptide library.

Monoclonal antibodies that bind native protein can generate considerable information about structure/function relationships, but identification of their epitopes can be problematic. Previously, monoclonal antibody M8-P1-A3 has been shown to bind to the catalytic (alpha) subunit of the Na+,K(+)-ATPase holoenzyme and the synthetic peptide sequence 496-HLLVMK*GAPER-506, which includes Lys 501 (K*), the major site for fluorescein-5'-isothiocyanate labeling of the Na+,K(+)-ATPase. This sequence region of alpha is proposed to comprise a portion of the enzyme's ATP binding domain (Taylor, W. R. & Green, N. W., 1989, Eur. J. Biochem. 179, 241-248). In this study we have determined M8-P1-A3's ability to recognize the alpha-subunit or homologous E1E2-ATPase proteins from different species and tissues in order to deduce the antibody's epitope. In addition the bacteriophage random peptide or "epitope" library, recently developed by Scott and Smith (1990, Science 249, 386-390) and Devlin et al. (Devlin, J. J., Panganiban, L. C., & Devlin, P. E., 1990, Science 249, 404-406), has served as a convenient technique to confirm the species-specificity mapping data and to determine the exact amino acid requirements for antibody binding. The M8-P1-A3 epitope was found to consist of the five amino acid 494-PRHLL-498 sequence stretch of alpha, with residues PRxLx being critical for antibody recognition.

Facile synthesis of 2-[(3-aminopropyl)thio]adenosine 5'-diphosphate: a key intermediate for the synthesis of molecular probes of adenosine 5'-diphosphate function.

Adenosine 5'-diphosphate (ADP) and, uniquely, its C-2 derivatized analogues are able to induce platelet activation. We here report the synthesis of 2-[(3-aminopropyl)thio]-ADP from ADP itself via 1,N6-etheno-ADP. 2-[(3-Aminopropyl)thio]-ADP induced platelet aggregation with a potency about one-seventh that of ADP itself and should prove a useful intermediate in the synthesis of other probes of platelet function.

Pathophysiology of platelet thrombin receptors.

The evidence is reviewed and a model presented for two distinct receptors being involved in platelet activation induced by alpha-thrombin: a high affinity thrombin receptor constituting approximately 50 copies of a supercomplexed form of GPIb coupled to phospholipase A2 and a moderate affinity receptor constituting approximately 2000 copies of the proteolytically activated, G protein-coupled seven transmembrane domain receptor coupled to phospholipase C. Reasons for the failure of certain studies to detect this role for GPIb are discussed.

Structure and function of platelet glycocalicin.

Present knowledge of the structure and function of platelet glycocalicin is reviewed. Glycocalicin (M 150,000) is a glycoprotein component of the outer surface of intact platelets which is released in soluble form following platelet homogenization. Glycocalicin has been purified and shown to inhibit platelet aggregation induced by thrombin or by ristocetin. Thrombin binding activity is associated with the peptide "tail" of the molecule (Mr 45,000), the macroglycopeptide portion (Mr 120,000) being without effect. Glycocalicin and membrane-bound glycoprotein I have been shown to be functionally and immunologically identical. Studies with platelets modified by chymotrypsin, and with platelets from patients with Bernard-Soulier disease and an ill-defined bleeding abnormality show that the amount of thrombin bound is proportional to the total amount of glycocalicin and glycoprotein I present. These results support the concept of a single class of binding site for thrombin in platelets.

A resolution of reported discrepancies in the characteristics of platelet glycoproteins IV (GPIV) and IIIb (GPIIIb).

The terms glycoprotein IV (GPIV) and glycoprotein IIIb (GPIIIb) have been used interchangeably and reports in the literature have indicated this glycoprotein as having a molecular weight variously described as either 88,000 or 97,000, a fast anodal mobility on crossed electrophoresis and either 13 or less than 1 methionine residues on amino acid analysis of the purified glycoprotein. To resolve these discrepancies, we have evaluated the characteristics of GPIV both in whole platelets and after isolation. These studies have shown that the term GPIV defines a protease-resistant platelet surface glycoprotein with Mr 88,330 +/- 2,240 which is immunologically identical with the CD36 differentiation antigen, which migrates with a relatively slow anodal mobility on crossed immunoelectrophoresis and which contains approximately 13 methionine residues per mole.

Evaluation of the binding to fixed platelets of agonists and antagonists of ADP-induced aggregation.

Steady state binding of eleven different ADP analogues to formaldehyde-fixed platelets has been determined in a competitive binding assay using 3H-ADP. The compounds tested were the inactive analogues L-ADP and L-ATP; the agonists 2-chloroadenosine 5'-diphosphate, adenosine 5'-O-(2-thiodiphosphate) and the diasteroisomeric pair Sp-adenosine 5'-(1-thiodiphosphate) (Sp-ADP-alpha-S) and Rp-adenosine 5'-(1-thiodiphosphate) (Rp-ADP-alpha-S); and the antagonists adenosine 5'-O-thiomonophosphate, 2-chloroadenosine 5'-O-thiomonophosphate, 2-choloroadenosine 5'-triphosphate, and the diastereoisomeric pair 5'-(1-thiotriphosphate) (Sp-ATP-alpha-S) and RP-adenosine 5'-(1-thiotriphosphate) (Rp-ATP-alpha-S). All compounds tested competed at the high affinity binding sites for ADP previously identified (Blood 1988; 71: 110-6) but in some cases competition could not be demonstrated at the low affinity sites because of the high nucleotide concentrations required. As a group, C2-substituted analogues bound less strongly (Ki greater than 2 micro M) than did the analogues without substituents in the purine ring (Ki less than 0.7 microM). With the pair of diastereoisomeric agonists Sp-ADP-alpha-S and Rp-ADP-alpha-S the Ki values at the high affinity site (210 nM and 560 nM) were of the same relative magnitude and in the same direction as their reported potencies as agonists (Ki 4 microM and 20 microM). With the diastereoisomeric antagonists Sp-ATP-alpha-S and Rp-ATP-alpha-S a similar relationship was seen between affinity (17 nM and 156 nM) and inhibitory potency (Ki 4 microM and 20 microM).(ABSTRACT TRUNCATED AT 250 WORDS)