Search

Amyloid precursor protein (A betaPP) processing results in generation of amyloid beta peptide (A beta) which deposits in the brain parenchyma and cerebrovasculature of patients with Alzheimer's disease (AD). Evidence that the vascular deposits derive in part from A betaPP fragments originating from activated platelets includes findings that individuals who have had multiple small strokes have a higher prevalence of AD compared to individuals who have taken anti-platelet drugs. Thus, determination of whether platelet A betaPP fragments are capable of traversing the blood-brain barrier (BBB) is critical. We have established that activated platelets from patients with AD retain more surface transmembrane-bound A betaPP (mA betaPP) than control platelets. We report here that this mA betaPP can be cleaved to A beta-containing fragments which pass through a novel BBB model system. This model utilizes human BBB endothelial cells (BEC) isolated from brains of patients with AD. These BEC, after exposure to activated platelets which have been surface-labeled with fluorescein and express surface-retained mA betaPP, cleave fluorescein-tagged surface proteins, including mA betaPP, resulting in passage to the BEC layer The data confirm that BEC contribute to processing of platelet-derived mA betaPP and show that the processing yields A beta containing fragments which could potentially contribute to cerebrovascular A beta deposition.

We report here the discovery of two novel human platelet and megakaryocytic DAMI cell enzymes that have beta-secretase-like activity. These activities could potentially effect cleavage of the amyloid precursor protein (APP) at the beta-amyloid peptide N-terminus, by an EC 3.4.24.15-like metalloprotease, and the N terminus-1 position, by a serine protease. Thus both enzymes may generate the amyloidogenic beta-peptide. Studies of intact and Triton X-100-lysed DAMI cells, as well as intact versus subcellular fractions of platelets, demonstrate the presence of these proteolytic activities. The resting platelet has (1) a surface serine protease, demonstrated by its ability to cleave a beta-secretase substrate and by its inhibitor sensitivity; and (2) a metalloprotease, recognized by an antibody to EC 3.4.24.15, which resides intracellularly in the alpha-granule membrane, is translocated to the surface on activation, and shows beta-secretase-like activity by cleaving the same substrate. This metalloprotease can also cleave recombinant APP to a potentially amyloidogenic fragment. Surface metalloprotease was identified in DAMI cells by flow cytometry and Western blotting with a specific anti-EC 3.4.24.15 monoclonal antibody, while activity was identified by using two beta-secretase substrates. This article is the first to document two previously unknown endoproteinases with beta-secretase-like activity in platelets and DAMI cells. These proteases are capable of effecting cleavage of APP and could therefore contribute to Abeta deposition in the cerebrovasculature.

Neutropenia is considered a significant risk factor for invasive aspergillosis but is almost always associated with concurrent thrombocytopenia. Studies determined that platelets, like neutrophils, attached to cell walls of the invasive hyphal form of Aspergillus fumigatus. Organisms were damaged as shown by loss of cell wall integrity in scanning laser confocal microscopy and release of defined hyphal surface glycoproteins. Rapid expression appearance of surface antigen CD63 and release of markers of platelet degranulation confirmed activation during attachment to hyphae. Optimal platelet activation required opsonization of hyphae with fresh or heat-inactivated whole plasma. These effects of opsonization with whole plasma could not be duplicated by pooled human serum, immunoglobulin G, or fibrinogen, whether used separately or combined. Thus, platelets in the presence of whole plasma have the potential to play an important role in normal host defenses against invasive aspergillosis.

Platelets, when released as anuclear cells by their precursor megakaryocytes, already carry soluble proteolytic fragments of the amyloid precursor protein (APP) within their alpha-granules and intact APP in the alpha-granule membranes. In response to activation signals elicited by physiologic stimuli such as thrombin, platelets release their granules' soluble contents and translocate granule membrane-bound proteins to the plasma membrane. Because platelets carry >90% of the circulation's APP, activated platelets have been implicated as origins of the beta-amyloid peptide fragment of APP (A beta), whose deposition in the cerebrovasculature is characteristic of Alzheimer's disease. We have therefore studied the APP contents and proteolytic processing in resting DAMI human megakaryocytic cells, along with the consequences of the activation of these cells by thrombin, comparing the results in each case to those with human platelets. Resting and PMA-differentiated DAMI cell contents were examined by Western blotting, immunoprecipitation, or metabolic labeling with sulfur 35-labeled methionine during culture, while plasma membrane-bound APP was evaluated by flow cytometry. Activation was followed by changes in cytoplasmic calcium concentration ((Ca++)in) and in membrane potential. Like platelets, DAMI cells exhibited a thrombin dose-dependent delta(Ca++)in, and membrane potential change; in contrast to the surface of a platelet, the surface of an agranular resting DAMI cell expresses granule-membrane proteins (APP and CD63) that appear on platelets only after activation. DAMI cell culture with 35S-labeled methionine confirmed that megakaryocytes synthesize large amounts of APP, of slightly higher molecular weight, and degrade their APP extensively before platelets are formed.

Upon activation, platelet alpha-granules' soluble contents are secreted and membrane-bound contents are translocated to the plasma membrane. Membrane-bound proteins include the beta-amyloid precursor protein (APP) from which the beta-amyloid (A beta) deposits found surrounding the cerebrovasculature of patients with Alzheimer's Disease (AD) may originate. We show here that activated platelets from AD patients exhibit less APP processing, retain more of the protein on their surface, and secrete less as soluble fragments than do controls. Surface labeling demonstrated that there is little APP or CD62 on the surface of resting platelets. Upon activation, control platelets exhibited more of both proteins on their surface, while advanced AD patients exhibited similar amounts of CD62 as controls, but retained significantly more surface APP. AD platelets secreted similar amounts of most soluble alpha-granule contents as controls, but less APP fragments. Together these results suggest a processing defect that may account for greater deposition of A beta-containing products in the vasculature to which activated platelets adhere.

Using three experimental approaches, we have addressed the questions of whether the presence of saturably bound thrombin plays a role in potentiating the activation of platelet phospholipase C (PLC) and/or accumulation of the 3-phosphorylated phosphoinositides (3-PPI), i.e. phosphatidylinositol 3,4-bisphosphate and phosphatidylinositol 3,4,5-trisphosphate, and whether the generation of tethered ligand (Vu, T-K.H., Hung, D. T., Wheaton, V. I., and Coughlin, S. R. (1991) Cell 64, 1057-1068) by thrombin can account fully for thrombin's proteolytic effects in activating platelets, as gauged by the above parameters. We have 1) measured PLC activation or 3-PPI after we have exposed platelets to thrombin for various periods and either blocked thrombin's proteolytic activity without interrupting its binding or blocked both binding and proteolytic activity of thrombin; 2) attempted to potentiate 3-PPI accumulation, using combinations of protein kinase C stimulation, Ca2+ elevation, and saturating but proteolytically inactive thrombins; and 3) compared the activation of platelets by thrombin with activation by the "thrombin" receptor-directed peptide, SFLLRNPNDKYEPF (SFLL; a portion of the tethered ligand created by thrombin's proteolytic activity), and examined the effect of thrombin on this latter activation. We conclude that the initial and sustained effects of thrombin in stimulating PLC and the accumulation of 3-PPI are completely attributable to thrombin's proteolytic activity. Further, thrombin's effects in promoting these responses can be accounted for by the actions of SFLL peptide, and by implication, formation of tethered ligand.

Human platelets exhibit an extremely rapid increase in cytoplasmic Ca2+ concentrations ((Ca2+]in) and a dose-dependent cytoplasmic pH change ((pH]in) upon thrombin stimulation. A cytoplasmic alkalinization, maximal by 60 s, is preceded by a very rapid acidification, which is masked by the alkalinization when saturating thrombin doses are used. Using the pH probe 2',7'-bis-(carboxyethyl)-5(6)-carboxyfluorescein we report here the kinetics of simultaneous cytoplasmic pH and Ca2+ changes in thrombin-stimulated platelets, measured in single cells by flow cytometry. This permits analysis of the responding subpopulation. Maximal thrombin stimulation (greater than or equal to 4.5 nM) induces a dose-dependent increase in pHin from approximately 7.0 to 7.30 and a maximal [Ca2+]in transient of up to 800 nM. The Ca2+ transient coincides temporally with the rapid initial acidification, while the alkalinization is maximal considerably later. The Ca2+ transients occur maximally in each responding cell, but occur only in a subpopulation of the platelets at subsaturating (less than 4.5 nM) thrombin doses; in contrast, the dose-dependent cytoplasmic acidification appears to occur uniformly in all platelets. The rapid increase in [Ca2+]in is not dependent on the alkalinization, and the former occurs maximally in amiloride treated, Na+/H+ exchange inhibited human platelets. These results indicate that the acidification and the rise in [Ca2+]in may be interrelated, whereas the cytoplasmic alkalinization (maximal considerably later than either the acidification or the [Ca2+]in rise) may be independent of these earlier, temporally correlated increases in H+ and Ca2+ concentrations.

The response of human platelets to stimulation by a specific aggregant such as thrombin has been postulated to proceed sequentially via induction of response at the membrane, followed by execution of shape change, secretion, and aggregation of the platelets. We have shown earlier that the platelet response includes a depolarization of the membrane which starts within less than 5 s and is thrombin-dose-dependent up to 4.5 nM alpha-thrombin. This depolarization may be measured by the distribution of either a fluorescent or a tritium-labeled lipophilic cation. We present here an adaptation of techniques for intracellular pH measurements to the human platelet. These show that stimulation with thrombin also induces a rapid change in the platelet transmembrane pH gradient as measured using either a weak base or a fluorescein derivative as a probe. The pH gradient undergoes a time-dependent and thrombin-dose-dependent change which parallels that exhibited by the membrane potential and by serotonin secretion.

alpha-Thrombin induces a dose-dependent rapid transient increase in platelet cytosolic Ca2+ levels, coming solely from intracellular stores, since EGTA has no effect. In contrast, the post-stimulation equilibrium [Ca2+]in depends upon an influx from the extracellular milieu, and is lower in the presence of EGTA. We measured the Ca2+ transient (with Indo-1, 1-[2-amino-5-(6-carboxyindol-2-yl)-phenoxy]-2-(2'-amino-5'-methylp henoxy)- ethane-N,N,N',N'-tetraacetic acid), cytosolic alkalinization (with BCECF, 2',7-bis-(2-carboxyethyl)-5(and 6)-carboxyfluorescein), membrane depolarization (with diS-C3-(5), 3,3'-dipropylthiodi-carbocyanide iodide), and degranulation (by beta-glucuronidase release) induced in washed human platelets by 9 nM thrombin in the absence or presence of extracellular or intracellular Ca2+ chelating agents (EGTA and BAPTA, 5,5'-dimethyl-bis-(o-aminophenoxy)-ethane-N,N,N',N'-tetraacetic acid, respectively). Platelets loaded simultaneously with 2 microM Indo-1 and 15 microM BAPTA (each as the acetoxymethyl ester) before addition of thrombin exhibited no cytoplasmic Ca2+ transient or alkalinization, no depolarization or degranulation. Replenishment of such cells with extracellular CaCl2 restored resting [Ca2+]in. Upon stimulation with 9 nM thrombin these replenished platelets exhibited no Ca2+ transient, and a slow gradual increase in [Ca2+]in from extracellular stores, a slow alkalinization and depolarization, and partial degranulation, all abolished by extracellular EGTA. Thus thrombin-induced platelet activation exhibits a biphasic Ca2+ requirement: the initial transient increase in [Ca2+]in comes from intracellular stores only, while the later steps of depolarization, alkalinization, and degranulation can proceed, albeit more slowly, if only extracellular Ca2+ is available.

A noncovalent fluorescent probe that responded to changes in transmembrane potential was used to study the response of washed human platelets to aggregating agents. Concentration-dependent changes in the fluorescence were observed in response to ADP and to thrombin. No such changes were observed in response to collagen fibrils. Thus there was an indication that platelet membrane potential changed in response to aggregating stimuli, supporting the hypothesis that the mechanisms of platelet aggregation resembled the mechanisms of other systems that show stimulus-response coupling (e.g., muscle, adrenal chromaffin cells). The different responses to specific agents indicate that the agents may trigger platelet aggregation through different mechanisms.

Thrombin stimulation of human platelets initiates a membrane depolarization attributable to a Na+ influx into, and an alkalinization of, the cytoplasm, both of which follow a similar rapid time scale and thrombin-dose dependence. These responses precede secretion of the contents of the dense granules (serotonin) and, after 1 minute, of lysosomes (beta-glucuronidase). We have evaluated these parameters in the presence of 2H2O in order to determine if the Na+ influx and H+ efflux are sequential or simultaneous. NMR evidence indicates that 2H2O equilibration in rapid, and virtually complete within the 3 min prestimulation platelet equilibration period. In response to an 0.05 U/ml addition of thrombin, the rate of depolarization is 70-80% slower in 2H2O than in H2O. The time to reach maximal depolarization is 5 to 10 seconds longer in 2H2O, the extent of depolarization 60% inhibited, and the pH change 85% inhibited. The serotonin secretion is unaltered, while the beta-glucuronidase secretion is 130-180% enhanced. Dimethylamiloride inhibits the Na+ influx and the pH change completely. These results suggest that the Na+ and H+ fluxes across the plasma membrane are interdependent but neither simultaneous nor electroneutral. Furthermore, granule secretion, previously shown by us to be independent of the existent Na+ gradient, depends on the cytoplasmic K+ and H+ concentrations.

Human platelets stimulated by epinephrine undergo enhanced turnover of phosphatidylinositol 4,5-bisphosphate, accumulate inositol trisphosphate, diacylglycerol, and phosphatidic acid, and phosphorylate a 47-kDa protein. All of these phenomena indicate stimulation of phospholipase C. These responses are blocked completely by inhibitors of alpha 2-adrenergic receptors (yohimbine), cyclooxygenase (aspirin or indomethacin), phospholipase A [2-(p-amylcinnamoyl)amino-4-chlorobenzoic acid (ONO-RS-082)], Na+/H+ exchange [ethylisopropylamiloride (EIPA)], fibrinogen binding to glycoprotein IIb/IIIa (antibody A2A9), Ca2+/Mg+ binding (EDTA), or removal of fibrinogen. Epinephrine evokes (i) an increased turnover of ester-linked arachidonic acid in aspirin treated platelets that is inhibited by ONO-RS-082, EDTA, yohimbine, or the absence of fibrinogen and (ii) a rapid cytoplasmic alkalinization that is inhibited partially by blockage of cyclooxygenase activity and completely by A2A9 or EIPA. In contrast, when incubated with subaggregatory concentrations of the prostaglandin H2/thromboxane A2 analogue [(15S)-hydroxy-11 alpha,9 alpha-(epoxymethano)prosta-5,13-dienoic acid (U46619) and epinephrine, aspirin-treated platelets show a potentiation of phospholipase C activation that is unaffected by the above inhibitors. We propose that epinephrine, in promoting exposure of glycoprotein IIb/IIIa sites for fibrinogen binding, leads to a cytoplasmic alkalinization, which, in conjunction with local shifts in Ca2+, promotes low-level activation of phospholipase A. The resulting free arachidonic acid is converted to cyclooxygenase products, which, potentiated by epinephrine, activate phospholipase C. This further amplifies the initial stimulatory response.

Investigations in this laboratory have demonstrated that thrombin induces dose-dependent changes in the transmembrane electrical potential of gel-filtered human platelets. This change is monitored with the fluorescent lipophilic cation, 3,3'-dipropylthiodicarbocyanine (diS-C3-(5], whose rapid release from the platelet (maximal within 30 s) correlates with a rapid, dose-dependent influx of sodium, a depolarization, and an increase in the intracellular pH. There is also a later release of this probe, detectable only 60 s after activation by thrombin. It is shown that this latter probe release is also thrombin dose dependent, and correlates in time course and extent with the secretion of beta-glucuronidase from the platelet's lysosomal granules, implying that it corresponds to probe sequestered in these granules in the resting platelet. Such a conclusion is corroborated by the fact that both the thrombin-induced secondary release of diS-C3-(5) and the secretion of the lysosomal enzyme, beta-glucuronidase, are inhibitable to the same extent by pretreatment of the probe-equilibrated platelets with valinomycin, a K+ ionophore, are partially inhibited to a comparable extent when thrombin is removed from the platelet membrane by an excess of hirudin within 15 s of activation, and are unaffected by amiloride, a Na+ blocking agent. We suggest therefore that some of the membrane potential probe diS-C3-(5) is accumulated by the platelet lysosomal granules and is secreted when the platelets are stimulated by the high doses of thrombin which induce lysosomal enzyme secretion. This secondary dye release is linearly proportional to, and can be used as a continuous and quantitative indicator of, the thrombin-induced lysosomal enzyme secretion by human platelets.

Thrombin-induced stimulation of human platelets is accompanied by a dramatic increase in cytoplasmic calcium concentrations followed by a slow decrease. These changes are very rapid, are maximal by 10-15 s, and can be detected with probes such as Indo-1. Suspension studies using spectrofluorometry, which reflect a value which is the average of 3 x 10(7) cells per ml, indicate a thrombin dose-dependent increase in cytoplasmic calcium at doses up to 0.025 units per ml. We show here, using flow cytometry, that at less than half-saturating thrombin doses only subpopulations of platelets rather than the entire sample are responding. The extent of these responses, however, still depends on thrombin concentration. When the thrombin doses are between half and fully saturating, one subpopulation responds fully (i.e., its extent of increase in cytoplasmic calcium concentration, [Ca++]in is 100% of that seen at saturating thrombin concentrations) while the remaining platelets respond partially or not at all. There is thus evidence of positive cooperativity leading to disproportionate thrombin receptor occupancy on different subpopulations when platelets are subjected to subsaturating doses of thrombin. The existence of responding subpopulations may explain how the reported multiple stimulations of the same suspension of platelets at low thrombin doses occur.

The effects of alpha-tocopherol (vitamin E) deficiency on membrane properties of platelets were studied to determine if vitamin E has a measureable stabilizing role in biological membranes. Three groups of rats and three of mice were studied: two groups consisted of Fisher strain rats and one of Sprague-Dawley rats fed a Draper corn oil diet with and without high levels of supplementary vitamin E. The mice were two groups of BALB/c animals maintained on an 8% hydrogenated coconut oil diet, and one group of CBA/J mice on an 8% lard diet, in each case either deficient in or supplemented with vitamin E. The relative content of fatty acids obtained from both rat platelets and erythrocytes was unchanged by vitamin E deficiency. Depletion of vitamin E had no effect on the degree of fluorescence polarization of 1,6-diphenyl-1,3,5,-hexatriene-labeled rat platelets. No changes in hematocrit values were seen in any of the studies. The platelet count of only the vitamin E-deficient Sprague-Dawley rats was elevated with respect to vitamin E-supplemented counterparts; the others remained constant. Platelet reactivities, as measured by ADP-and thrombin-induced platelet aggregation and by the thrombin-induced changes in platelet transmembrane potential, were unaffected by vitamin E deficiency in all three groups of rats. Our results indicate that a membrane stabilizing effect of vitamin E on rat platelet or erythrocyte membrane fatty acids or on platelet response to external stimuli could not be demonstrated, nor was elevation in platelet count a general phenomenon associated with vitamin E deficiency.-Whitin, J. C., R. K. Gordon, L. M. Corwin, and E. R. Simons. The effect of vitamin E deficiency on some platelet membrane properties.

alpha-Thrombin has previously been shown to bind to specific, saturable glycoproteins on the platelet surface. Modification of the thrombin active site with tosyllysyl chloromethyl ketone (TosLysCH2Cl) does not alter thrombin's binding characteristics. Interaction of alpha-thrombin with high-affinity binding sites (KD = 10(-9) M) initiates the platelet response which involves proteolytic hydrolysis of this glycoprotein. Although TosLysCH2Cl--thrombin binds to and competes for the same sites as alpha-thrombin, it cannot induce platelet stimulation because it is enzymatically inactive. In this study, we describe the preparation and application of photoreactive tritium-labeled thrombin analogues. The alpha-thrombin derivative retains its platelet-stimulating and enzymatic activities and, upon photoactivation, covalently binds to specific platelet membrane components. When freshly washed human platelets are exposed to less than saturation doses (less than or equal to 2 nM) of the thrombin derivatives in the dark and photoactivated, a single labeled complex is detected. The same experiment with greater than saturating doses (greater than or equal to 20 nM) of the thrombin derivative yields a similar complex as well as two additional ones. Molecular weight estimates of these thrombin-bound complexes were obtained by gel filtration and NaDodSO4--polyacrylamide gel electrophoresis. The low dose (high affinity) complex with TosLysCH2Cl--thrombin has an approximate molecular weight of 200 000, while that with active alpha-thrombin is smaller, approximately 120 000, due to enzymatic cleavage. The additional complexes detected with the high thrombin dose had estimated molecular weights of 400 000 and 46 000, respectively, and appeared to be the same for TosLysCH2Cl--thrombin and for the alpha-thrombin coupled platelets. These isolated complexes appear to correspond to the two previously detected populations of thrombin binding sites on the platelet.

Thrombin causes a dose-dependent depolarization of the transmembrane potential of normal human platelets which can be continuously measured by the fluorescent probe, 3,3'-dipropylthiodicarbocyanine, whose distribution across the plasma membrane has been shown to be dependent upon the membrane potential. The dose-dependent depolarization of the platelet's negative membrane potential by thrombin is in large part due to a rapid uptake of sodium. Both the membrane potential change and the rapid sodium influx can be inhibited by a fast acting analog of amiloride, a sodium channel blocker, while valinomycin, a potassium ionophore, has no effect on the potential change nor on the sodium uptake, suggesting that the transmembrane potassium gradient is not important in the thrombin-induced depolarization. Neither the secretion of serotonin nor that of lysosomal enzymes nor the secondary release of the fluorescent probe which correlates with the lysosomal enzyme secretion occur if treatment with valinomycin precedes activation by thrombin. It is thus apparent that: 1) the change in the membrane potential induced by thrombin is directly dependent upon the transmembrane sodium gradient and is primarily due to a dose-dependent sodium uptake by the platelets; and 2) the thrombin-induced secretory processes are dependent upon maintenance of the transmembrane potassium gradients.

Removal of calcium from thrombospondin with EDTA results in a decrease in the intensity of the negative CD peak between 200 and 250 nm. Quantitative analysis of the CD spectrum of thrombospondin indicates that thrombospondin contains approximately 11% alpha-helix, 43% beta-sheet, and 46% random coil in the presence of calcium and that a small change in secondary structure may occur upon removal of calcium with EDTA. When the change in the CD at 220 nm is measured as a function of calcium concentration, a sigmoidal curve with a transition midpoint of 120 microM is obtained, indicating that the binding is cooperative. Analysis of Hill plots of the data revealed a Hill coefficient of 12.3. Calcium was also found to affect the peptide pattern produced by limited tryptic digestion of thrombospondin, with some portions of the molecule being resistant to trypsin in the presence of calcium. When the change in quantity of a 65,000-dalton tryptic fragment was measured as a function of calcium concentration, a sigmoidal curve was again obtained. The midpoint of this transition is achieved at a free calcium concentration of 45 microM at 0 degrees C and 103 microM at 25 degrees C. These data indicate that thrombospondin contains at least 12 binding sites for calcium and that cooperative interactions between sites are associated with a conformational change in the thrombospondin molecule.

Noncovalently bound fluorescent probes have been used to study changes in the platelet which may occur during platelet aggregation. Platelets were exposed to either N-phenyl-naphthylamine (NPN) or 8-anilino-1-naphthalene-sulfonic acid (ANS). Both dyes were bound by the platelet, and platelet aggregation by collagen or thrombin was unaffected by the presence of the label. No change in fluorescence intensity or wavelength of maximum intensity was observed during platelet aggregation.

The grey platelet syndrome is a rare inherited disorder characterized by a marked decrease or absence of alpha-granules and of platelet-specific alpha-granule proteins. By utilizing platelets from two patients with this syndrome, we here demonstrate that the initial response of human platelets to alpha-thrombin does not require the presence of alpha-granules nor the effective release of their constituents. Furthermore, these platelets respond to thrombin with a normal, dose-dependent membrane potential change, and a normal secondary release of diS-C3-(5) thought to be released in parallel with beta-glucuronidase from the lysosomal granules. These results give new insight into the initial steps in the thrombin response of normal platelets.

Thrombin causes a dose-dependent cytoplasmic alkalinization of normal human platelets. The pH probe 2',7'-bis(carboxyethyl)-5(6)-carboxyfluorescein (BCECF) permits an easier and more accurate quantitation of the kinetics of this change previously measured with 9-aminoacridine and 6-carboxyfluorescein (6-CF). We report here a modification of the previously published 6-CF technique and confirm the thrombin-induced cytoplasmic pH change in the human platelet using a second fluorescein derivative, BCECF. Maximal thrombin stimulation raises the resting cytoplasmic pH of the human platelet from 7.0 to 7.25.