the of chemistry vol. 259, no. 20, of october 25. pp. 12563 … · 2001-09-03 · the journal of...

THE JOURNAL OF BIOLOGICAL CHEMISTRY Q 1984 by The American Society of Eiological Chemists, Inc.

Ca2+ Homeostasis in Unstimulated Platelets*

Vol. 259, No. 20, Issue of October 25. pp. 12563-12570,1984 Printed in U. S.A.

(Received for publication, April 16, 1984)

Lawrence F. Brass From the Hematology-Oncology Section, Department of Medicine and the Cancer Center, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania 19104

Unstimulated platelets maintain a low cytosolic free Ca2+ concentration and a steep plasma membrane Ca'+ gradient. The mechanisms that are required have not been completely defined. In the present studies, 4aCa'+ was used to examine the kinetics of Ca" exchange in intact uns t imula t~ platelets. Quint was used to measure the cytosolic free Ca'+ concentration. Under steady-state conditions, the maximum rate of Ca'+ exchange across the platelet plasma membrane, 2 pmol/ los platelets/min, was observed at extracellular free Ca" concentrations 20-fold less than in plasma. Two intrace~ular exchangeable Ca'+ pools were identified. The size of the more rapidly exchanging pool ( t l lz , 17 min) and the cytosolic free Ca'+ concentration were relatively unaffected by large changes in the extracellular Ca'+ concentration. In contrast, the size of the more slowly exchanging Ca'+ pool ftl,z, 300 min) varied with the extracellular Ca'+ concentration, which suggests that it is physically as well as kinetically distinct from the rapidly exchangeable Ca'+ pool. The locations of the Ca2+ pools were determined by differential per- meabilization of 46Ca2+-loaded platelets with digitonin. "'Ca'+ in the rapidly exchanging pool was released with lactate dehydrogenase, which suggests that it is located in the cytosol. "Ca'+ in the slowly exchanging pool was released with markers for both the dense tubular system and mitochondria, but inhibition of mitochondrial ca'+ uptake with carbonyl cyanide rn-chlo- rophenylhydrazone had no effect on the size of the slowly exchangeable Ca" pool or the cytosolic free Ca2+ concentration. In contrast, addition of metabolic inhibitors (KCN plus carbonyl cyanide m-chlorophen- ylhydrazone plus deoxyglucose) or trifluoperazine caused a decrease in the size of the slowly exchangeable Ca" pool and an increase in the cytosolic free Ca2+ Concentration. These observations suggest that Ca2+ homeostasis in unstimulated platelets is maintained by 1) limiting Ca'+ influx from plasma, 2) actively promoting Ca" efflux, and 3) sequestering Ca'+ within an internal site, which is most likely the dense tubular system and not m i t ~ h o n ~ i a .

Prior to activation by an agonist, the free Ca2+ concentration in the cytosol of human platelets is 0.1 MM, which is 10,000-fold less than the free ea2+ concentration in plasma

* This study was supported in part by National Institute of Health Grant HL29018. Portions of the study were presented at the Annual Meetings of the American Society of Hematology, San Francisco, CA in December 1983 and published in abstract form (Brass, L., and Belmonte, E. (1983) B W 62, Suppl. 1,251a). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertise- ment" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

~ _ _

(1-3). Since an increase in the cytosolic free Ca2+ concentration will cause platelet activation, mechanisms must exist to limit Ca2+ influx from plasma and actively remove Ca2+ from the cytosol (4-6). Many of the details of these processes are unknown. Studies performed with platelet membrane vesicles suggest that the dense tubular system, which is derived from megakaryocyte endoplasmic reticulum, is able to actively sequester Ca" (7-10). Although this ability has yet to be demonstrated in intact platelets, the suggestion has been made that the dense tubular system stores Ca2+ for subsequent release during platelet activation, a role that is analogous to sarcoplasmic reticulum (6, 32). Less is known about Ca2+ transport across the platelet plasma membrane. In other tissues efflux of cellular Ca2+ is most commonly mediated by either a Ca2+-Mg2+-ATPase or Na+/CaZ+ exchange. Although platelets do possess Ca2+-MF-ATPase activity, the available evidence suggests that this enzyme is present in the dense tubular system and the surface-connecting membrane system, but not the plasma membrane (10, 12). The role of Na2/Ca2+ exchange has not been investigated as extensively. However, there is evidence that the plasma membrane Na+ gradient affects platelet ~ n c t i o n and, possibly, Ca2+ transport (13,14).

At present, therefore, the mechanisms which maintain Caz+ homeostasis in platelets are incompletely defined. In order to better understand these processes, we have begun by consid- ering Ca2+ homeostasis as it exists in resting platelets. In the present studies, 45Caz+ was used to measure the kinetics of Ca2+ transport across the platelet plasma membrane and to identify exchangeable C&'+ pools within platelets. The fluo- rescent intracellular CaZ+ probe, quin2, was used to examine the inter-relationships between Ca2+ transport and the cytosolic free Ca2+ concentration. Data presented in the accompanying paper address the role of Na+/Ca2+ exchange in Ca2+ transport across the platelet plasma membrane (14). In order to focus upon the processes active in unstimulated platelets, all of the studies were performed under conditions designed to minimize platelet activation.

~ A T ~ I A L S AND METHODS

Preparation of Gel-filtered Platelets-Platelet-rich plasma was obtained from normal donors by differential centrifugation of blood anticoagulated with 13 mM sodium citrate and incubated for 30 min at room temperature with 1 mM aspirin (16). The platelets were isolated by gel filtration on Sepharose 2B (Pharmacia Fine Chemi- cals) using an elution buffer containing 137 mM NaCl, 2.7 mM KCI, 1 mM MgC12, 5.6 mM glucose, 1 mgfml bovine serum albumin, 3.3 mM NaHzPO,, and 4 mM HEPES,' pH 7.40. Gel-filtered platelets prepared in this manner aggregated normally in response to agonists such as ADP and epinephrine when stirred at 37 "C with Ca2+ and fibrinogen but did not aggregate "spontaneously" when stirred under

' The abbreviations used are: HEPES, 4-(2-hydroxyethyl)-I-piper- azineethanesulfonic acid; CCCP, carbonyl cyanide rn-chlorophen- ylhydrazone; PGL, prostaglandin Iz; EGTA, [ethylenebis(oxyethyl- enenitri1o)ltetraacetic acid.

12563

12564 Ca2+ Homeostasis in Platelets the same conditions without an agonist. For the studies performed with Gd3+ and La3+, phosphate was omitted from the elution buffer, and the buffer pH was decreased to 6.8 to avoid formation of insoluble lanthanide phosphate salts.

Studies of Caz+ Exehnge-Except where noted, all steps were performed at room temperature. Immediately after gel filtration, the platelets were incubated with CaClZ (0.01 to 1 mM) for 30 min before adding a tracer quantity (2 to 10 pCi/ml of platelets) of '6CaC1z (New England Nuclear). Afterwards, in the studies of '%a2+ influx, duplicate 0.2-mi aliquots of the platelet suspension were removed at intervals over 4 h. Each aliquot was diluted in a 5-ml wash buffer and filtered through 0.45-p Millipore HAWP filters (Millipore Corp., Bedford, MA) which were then rinsed with at least two additional 5- ml aliquots of wash buffer to remove extracellular and surface-bound '%?,a2+. Wash buffer was the same as gel-filtration buffer with 5 mM EGTA added and glucose and albumin omitted. Platelet-associated '%az+ retained by the filters was measured by scintillation counting using ACS I1 aqueous counting solution (Amersham Corp.).

The studies of "Ca" efflux began the same way, but the platelets were gel filtered a second time 2 h after the addition of %az+ to remove nonincorporated tracer. This process takes approximately 10 min and, except when noted, was performed using elution buffer to which CaC12 was added at the same final concentration as was present during the "Caz+ loading period. Immediately afterwards, the '%a2+- loaded platelets were diluted &fold with column elution buffer, and the subsequent efflux of intracellular %a2+ was followed for 3 h. At any time, the amount of radioactivity remaining inside the platelets was determined by filtering duplicate aliquots of platelet suspension through the Millipore filters.

Analysis of the Ca2+ Exchange Data-In both the influx and efflux studies, the accumulated data for the amount of platelet-associated 'Tal+ a t various times was analyzed as a sum of exponential terms using Equation 1 (efflux) or Equation 2 (influx),

'Taz+ (cpm/l@ platelets) = R;(e-KFt) (1)

rsCa2+ (cpm/l@ platelets) = R;(l - e-KFL) (2)

where Ri and Ki are constants related to the size and turnover of individual exchangeable Ca" pools within the platelet. The half-time of exchange (tl,z) of each pool is equal to ln(2)lKi. A good fit of the observed data was obtained using either two (efflux) or three (influx) exponential terms by using the nonlinear regression algorithm described by Koeppe and Hamann (17) adapted for use on a Hewlett- Packard HP-86 microcomputer. This algorithm determines the values for R; and K; which minimize the residual sum of squares shown in Equation 3,

L

j -1 i&lY where yj0 and yj are, respectively, the observed and calculated values for &Ca2+ uptake at each of the n times of observation. In the present studies, the residual sum of squares was typically <0.02. This analysis produced a set of estimated values for Ri and K;. The precision of these estimates was determined from the radex error (17) and by the method of support planes described by Duggleby (44). By either criterion, the standard errors in Ri and Ki in the present studies were typically.<lo%.

In the %az+ efflux studies performed at steady state, the extracellular Ca2+ concentration was- kept constant throughout the 'Taz+ loading period and the efflux period. The data were analyzed using an open series3-compartment model (18). In this model, extracellular T a Z + exchanges with two platelet-associated Ca2+ pools. The more slowly exchanging of these pools (Sz) communicates with the more- rapidly exchanging pool ( 4 ) but not directly with the extracellular space.

J 1 k o 1 Jz k1z Extracellular Ca2+ SI e SZ

klo k 2 1

where J1 and Jz represent Ca2+ flux between the pools. As used here, the term Caz' flux refers nonspecifically to Ca" movement. The term efflux refers specifically to Ca2+ movement out of the cell, and influx refers to Caz+ into the cell across the plasma membrane. At steady state, Ca2+ influx equals Ca2+ efflux and Ca2+ flux from pool 1 into pool 2 equals Ca2+ flux from pool 2 into pool 1. Ca2+ exchange refers

to the bidirectional movement of Ca2+ under steady-state conditions. k 1 0 , k lz , and lEl1 are the associated rate constants. A series model rather than a parallel model was chosen for data analysis because of the experience in other tissues that S2 is located in a membrane- enclosed intracellular space, such as endoplasmic reticulum, that does not communicate directly with the extracellular space (18). Values for J, S, and k were calculated using the equations published by Uchikawa and Borle (18). Because isotopic equilibrium was not reached during the period of observation an open system was used in the analysis and h1 was indeterminate. In some of the studies, '%a2+ efflux was observed in the presence of one or more potential inhibitors of Cas+ transport. In those studies, the inhibitors were added to the platelet suspension 15 min before the start of the *Ca2+ loading period and a second time immediately after the second gel filtration. Since in the presence of these inhibitors a steady state may not have been present, only the relative pool sizes were calculated.

Measurement of Free Ca2+ and Total Calcium-Extracellular free Ca" concentrations above 1 pM were measured with a calcium- specific electrode (Orion Research Corp., Cambridge, MA). Free Caz+ concentrations below 1 IM were buffered with EGTA and calculated from the association constant of EGTA for Ca2+ at pH 7.4 (16). For measurement of total platelet-associated calcium 5-ml aliquots of platelet suspension were sedimented through silicone oil by centrifugation for 2 min at 12,000 X g in a microcentrifuge. The platelet pellets were dissolved overnight at 50 "C in concentrated nitric acid. Total Caz+ in the platelet digest was measured by atomic absorption spectroscopy using a Varian 1200 spectrophotometer. Corrections were made for extracellular Caz+ in the platelet pellet by measuring the volume of trapped extracellular fluid with [SH]sorbitol (16).

Measurement of the Cytosolic Free Caz+ Concentmtwn-Platelet- rich plasma was incubated at room temperature or 37 "C with 50 p~ quina acetoxymethyl ester prepared as a 50 mM stock solution in dimethyl sulfoxide (quin2/AM, Calbiochem-Behring). After 60 to 90 min, extracellular quin2/AM was removed by gel filtering the platelet suspension. Fluorescence was measured at room temperature in a Perkin-Elmer model MPF-43A fluorescence spectrophotometer. The cytosolic free Caz+ concentration was calculated as described by Rink et al. (1).

Enzyme Assays-Lactate dehydrogenase activity was assayed by following the decrease in 340-nm absorbance of a reaction mixture that included sodium pyruvate (0.72 mM), NADH (0.24 mM), and HEPES (20 mM), pH 7.0. NADH dehydrogenase was assayed by the method of Hochstadt et al. (37). Glutamate dehydrogenase was assayed by the method of Schmidt (38).

Other Materials-Verapamil, nifedipine, CCCP, 2-deoxyglucose, and trifluoperazine were obtained from Sigma.

RESULTS

The Kinetics of Ca2+ Exchange Ca2+ exchange by intact unstimulated platelets was mea-

sured by observing either 'Ta2' influx or T a 2 + efflux under steady-state conditions. In the studies of "Ca2+ efflux, gel- filtered platelets in buffer containing 0.2 mM CaClz were loaded with '%a2+ and then gel filtered a second time to remove nonincorporated tracer. The subsequent efflux of &Ca2+ from the cells was followed for 3 h. In order to determine whether the total calcium content of the platelets changed under these conditions, platelet calcium was measured by atomic absorption spectroscopy immediately after gel filtration and again 4 h later. Immediately after gel filtration, the platelets contained 23.9 rt 3.6 nmol of calcium per los platelets (mean 2 S.E., n = 9), which is similar to published values (13, 19, 20). Four hours later this value was unchanged, 23.3 & 3.4 nmol/lOS platelets.

A representative study of V a 2 + efflux is shown in Fig. 1. Regression analysis of the data from 34 such studies defined a rapidly exchanging Ca2+ pool with a half-time of exchange (t1,2) of 17 min and a more slowly exchanging pool with a t 1 p

of 313 min (Table I). 46Ca2+ influx into platelets at steady state was studied in the same manner and analyzed by nonlinear regression based on Equation 2 (Fig. 1). Similar results were obtained. Both of the exchangeable Caz+ pools are intra-

Cu2+ Homeostasis in Plutekts 12565

0 . 8000 - 6000 -

0 60 120 180 240 MINUTES

FIG. 1. 4aCaa+ influx and efflux under steady-state conditions. Gel-filtered platelets were incubated for 30 min with 0.2 mM Ca2+ after which a tracer amount of "Ca2+ was added. A, efflux. After a 2-h loading period, the platelets were gel filtered a second time using buffer that contained 0.2 mM CaCL The subsequent efflux of %a2+ is expressed as a fraction of the amount of intracellular &Ca2+ that was present immediately after the second gel filtration. The solid line was generated by nonlinear regression analysis using Equation 1. B, influx. @Ca2+ influx was observed for 4 h after the addition of tracer. The solid line was generated by nonlinear regression analysis using Equation 2.

TABLE I Kinetics of eaz+ exchange in unstimulated platelets

"Ca2+ efflux was measured at steady state in buffer containing 0.2 mM CaC12. The data were analyzed using a 3-compartment series model for Ca2+ exchange that permitted characterization of two intracellular exchangeable Ca2+ pools within the platelet. The symbols are defined under "Materials and Methods." The results shown are the mean t S.E. ( n = 34).

Rapidly exchanging pool tlA Pool size (SI) Ca" flux (&) Rate constant (kt@)

Slowly exchanging pool tu Pool size (Sz) Ca2+ flux (Jz) Rate constant (kj2) Rate constant (k2 , )

52 f 3 pmol/108 platelets 17 2 1 min

1.53 k 0.09 pmol/l@ platelets/min 30.0 & 1.2 min" (X 1OOO)

313 f 13 min 183 f 12 pmol/108 platelets

0.57 f 0.04 pmol/l@ piateletslmin 11.2 f 0.5 min" (X 1OOO) 3.20 f 0.12 rnin" (X 1000)

cellular; surface-bound Ca", which exchanges with a t1/2 of (1 min (16) was removed by the washing procedure. This analysis does not imply that the two intracellular Ca2+ pools are homogenous or that other exchangeable Caz+ pools do not exist. This type of analysis tends to average together Ca2+ pools with similar t 1 / 2 and to ignore Ca2+ pools that exchange with tllz much greater than the period of observation. In the present case the close fit of the regression curve to the longest data points (Fig. 1) suggests that any additional slowly exchangeable Ca2+ pools would have tl12 >> 300 min. Similarly, in the efflux studies, a rapidly exchanging intracellular Ca2+ pool with t,,,2 cc 17 min might be missed because of the time (approximately 10 min) required to gel filter the platelets into tracer-free medium. However, in previous studies, there was no evidence for the existence of such a pool (16).

Using the steady-state efflux data, it was possible to cal-

culate the sizes of the two intracellular exchangeable Ca2+ pools and the rates of Ca2+ exchange between the pools and across the plasma membrane. The results are summarized in Table I. The more rapidly exchanging Ca2+ pool (SJ contained an average of 52 pmol/108platelets. The more slowly exchanging pool (S,) was larger, 183 pmol/108 platelets. Since total platelet Caz+ is approximately 24 nmol/108 platelets, less than 1% of platelet-associated Ca2+ was exchangeable during the time frame used for these studies. Although this fraction is smaller than has been reported in other tissues (29), it is consistent with the presence in platelets of storage granules containing large amounts of Ca2+ that would be expected to exchange slowly, if at all { 13).

The Effect of Changing the Extracellular ea2+ C~n~entratwn In order to characterize the Ca2+ transport processes fur-

ther, we examined the effect of varying the extracellular Ca2+ concentration from 10 pM to 1 mM on Ca2+ exchange rates, the sizes of the intracellular Ca2+ pools and the cytosolic free Ca2+ concentration.

Exchange Rates-The rates of Ca2+ exchange across the plasma membrane (J1) and into the slowly exchanging pool (Jz) were dependent upon the extracellular Ca2+ concentration. The kinetics were saturable with maximum values for both parameters reached at 50-100 FM Ca2' (Fig. 2 and Table 11).

A markedly different response to changes in extracellular Ca2+ was observed when 45Ca2+ efflux was examined under non-steady-state conditions. For these studies, platelets loaded with 45Ca2+ in buffer containing 0.2 mM CaC12 were transferred by gel filtration to tracer-free buffer containing either 0.2 mM CaCI, or 0.5 mM EGTA. The rate of 45Ca2+ efflux was unaffected by this abrupt change in the extracellular ea2+ concentration (Fig. 3), which suggests that the efflux rate may not be directly dependent on the extracellular Ca2+ concentration.

ea2+ Pool Size-The sizes of the two exchangeable Ca2+ pools were calculated from the steady-state 4sCa2+ efflux data. Over a 100-fold range of extracellular free Ca2+ concentra-

2.

FIG. 2. The effect of the extracellular eaz+ concentration on steady-state Cas+ exchange rates. A , steady-state '%a2+ efflux was measured as described in Fig. 1 using gel-Gltered ptatelets equi- librated with 0.01 to 1.0 mM CaC12. At each Ca2+ concentration, the rate of Ca2+ flux from the rapidly exchanging or cytosolic Ca*+ pool ( X ) (0) and between the pools (Jz) (a) was calculated. B, a double- reciprocal plot of the same data.

12566 Ca2+ Homeostasis in Ptatetets

tions, the size of the cytosolic Ca2+ pool increased only 5-fold (Fig. 4). Most of this increase occurred below 0.1 mM Ca2+. Above 0.1 mM Ca2+, the pool size was relatively unaffected by changes in extracellular Caz+. In contrast, the size of the more slowly exchanging Ca2+ pool continued to increase as the extracellular Ca" concentration increased, even beyond the

TABLE I1 The effect of the extracellular Caz+ concentration on Cu2+ exchange

rates The initial rate of Ca2+ influx was measured under steady-state

conditions 15 min after the addition of "Ca2+ as described in Fig. 5. Ca2+ flux across the plasma membrane ( J J and into the sequestered Ca" pool ( 5 2 ) was measured in steady-state '%a2+ efflux studies as described in Fig. 2. The values shown for the influx studies are mean _t S.E. ( n = 3). The mean values for the efflux studies are the mean and range obtained in two identical studies of each type.

Influx studies K, 22 2 10 /AM fJ1)

VRl*X 1.8 f 0.1 pmol/lO* p l a ~ l e ~ / m i n

K m 17rt5pM (JI) Vaax 1.4 It 0.1 pmol/108 platelets/min

K m 42+4pM ( J z ) VUlaX 0.59 rt 0.07 pmol/lOs platelets/min

Efflux studies

1 0 60 I20 I80

FIG. 3. The effect of the extracellular Ca" concentration on Ca2+ efflux under non-steady-state conditions. Gel-filtered platelets were incubated for 30 min with 0.2 mM CaC12, loaded with 45CaZ+ for 2 h, and then gel filtered into buffer containing either 0.2 mM CaC12 (0) or 0.5 mM EGTA (0). The subsequent efflux of intracellular 45Ca2+ is expressed as a fraction of the platelet-associated 45Ca2+ present immediately after gel filtration. The solid lines were generated by nonlinear regression analysis based upon Equation 1.

N I N UTES

FIG. 4. The effect of the extracellular Ca" concentration on the sizes of the exchangeable intracellular Caz+ pools. Steady-state Ca2+ efflux was measured at 0.01 to 1 mM Ca2+ as described in Fig. 2. From the observed data, the sizes of the rapidly exchanging (cytosolic) (0) and the slowly exchanging (sequestered) (Of Ca2+ pools were calculated.

point where the maximum rate of Ca2+ exchange into this pool was reached (Fig. 4).

Cytosolic Free Cu2+-These obse~ations suggested that the cytosolic free Ca" concentration might also be relatively unaffected by extracellular Ca2+. To test this hypothesis, the cytosolic free Ca2+ concentration was measured with the flu- orescent intracellular Ca" probe, quin2. Over the same range of extracellular Ca2+ concentrations in which the size of the cytosolic exchangeable ea2+ pool increased s-fold, the cytosolic free Ca2+ concentration increased only 50% (Table 111). Decreasing the extracellular free Ca2+ concentration further by adding EGTA had little additional effect on cytosolic free Ca2+. Since the rapidly exchangeable Ca2+ pool is located in the platelet cytosol (see below), these data suggest that the two exchangeable Ca2+ pools are physically as well as kinetically distinct.

M e ~ ~ r ~ ~ ~ t s of Cuz+ ~ n f l ~ x For these studies, the rate of Ca2+ influx into platelets was

measured by observing the initial rate of &Ca2+ uptake. &Ca2+ influx increased with the extracellular ea2+ concentration but was maximal at approximately 50 PM Ca", 20-fold less than the plasma-free Ca2+ concentration (Fig. 5 ) . Values for the K,,, and Jl(max) are shown in Table 11. As would be expected if the platelets actually are at steady state, the maximum value for J1 obtained in the influx studies is essentially the same as the maximum value for J , found in the efflux studies.

An Eadie-Scatchard plot of the influx rates is linear, suggesting that influx may occur through a single class of chan-

TABLE 111 The effect of the extracellular Ca2+ concentrution on the cytosolic free

Cu" concentration Platelet-rich plasma was incubated with 50 p~ quin2lAM for 90

min at 37 "C and then gel fiftered. Imme~ately afterwards, CaCb or EGTA was added to give the extracellular free Ca*+ concentrations that are shown. The cytosolic free Ca2+ concentration was determined from quin2 fluorescence immediately after a 3-h incubation at room temperature. The values shown are mean f S.E. ( R = 7).

Extracellular free Ca*+

800 flM 9 5 f 8 n M 10 pM 6 1 f 6 n M 1 nM 39+3nM

Intracellular free Ca"

0 6 ID 14 I8 INFLUX

1 I

0 50 100 [ co2+] free (PM)

FIG. 5. The effect of the extracellular Caa+ concentration on the initial rate of Caz+ influx. Gel-filtered platelets were incubated for 30 min with CaC12 (0.01-1.0 mM) after which a tracer amount of "Ca2+ was added. The initial rate of Ca2+ influx was determined by measuring platelet-associated 45Ca2+ 15 min after the addition of tracer. The free Ca2+ concentration in the platelet suspension was measured directly using a Ca2+-specific electrode. The inset shows an Eadie-Scatchard plot of the data. In this example, the K, for ca2+ was 12 PM with Jl(mar) equal to 1.75 pmol/l@ platelets/min.

Ca2+ Homeostasis in Platelets 12567

nels. In order to learn more about the nature of this channel, platelets were incubated with verapamil and nifedipine, representative members of two classes of smooth muscle Ca2+ channel blockers. Both drugs affect platelet activation (21- 24). Gel-filtered platelets were incubated with verapamil or nifedipine (10 nM to 100 phi) for 30 min prior to the addition of tracer. 46Ca2+ influx was observed for 4 h. Neither agent had a detectable effect (data not shown). Verapamil(100 p ~ ) also had no effect on the cytosolic free Ca" concentration (113 k 14 nM uersus 117 f 14 nM, mean l?r S.E., n = 6). In the presence of 10 p~ nifedipine, however, the cytosolic free Ca2+ concentration fell to 75 k 8 nM (mean k S.E., n = 6), a decrease that is small but statistically significant (P < 0.005).

Inhibition of Ca2+ Transport by Gd3+ and &x3+

The studies described up to this point show that under non- steady-state conditions Ca2+ influx and efflux rates differ in their response to changes in extracellular Ca2+. In order to examine this issue further, we examined the ability of two "calcium-like" cations, Gd3+ and La3+, to inhibit Ca2+ influx and efflux. At concentrations up to 100 p ~ , Gd3+ markedly inhibited Ca2+ influx but had no effect on Ca2+ efflux (Fig. 6). A Dixon plot of data obtained at several Ca2+ concentrations shows that the inhibition of Ca2+ influx by Gd3+ appears to be noncompetitive. Low concentrations of La3+ produced the same effects as Gd3+. Higher concentrations of La3+ (>0.5 mM), however, did inhibit Ca2+ efflux (data not shown) as has also been reported in calf platelets (8).

Locations of the E x & ~ ~ e Q ~ ~ ca2' Pook In other tissues in which Ca2+ exchange has been examined,

the more rapidly exchanging of the intracellular Ca2+ pools has proved to be located in the cytosol. The more slowly exchanging pool has been assigned to sites of Caz+ sequestra-

2w z -. -

0 0 40 80 120

p + ] W )

FIG. 6. The effect of extracellular GdS+ on Caa+ influx and efflux. A, CaZ+ influx (0) and efflux @I at 10 pM extracellular Ca2+ were measured in the presence of 0 to 100 PM GdClS. The results shown are the extents of influx and efflux measured at 4 and 3 h, respectively, and are expressed as a fraction of the results obtained in the absence of Gd3+. Each point is the mean +: S.E. of 3 determi- nations. B, the extent of Cas influx at 4 h was measured at 0 to 100 PM Gd3+ in the presence of 10 (U), 20 (01, and 50 #M (0) Caz+. The solid lines were obtained by linear regression. The Kt for G d 3 + is 151 W .

tion such as mitochondria or endoplasmic reticulum (23,331. In order to determine whether a similar distribution exists in platelets, gel-filtered platelets incubated with 45Ca2+ for 2 h were selectively permeabilized by brief exposure to digitonin. Digitonin was used as the detergent because it affects cellular membranes in an order determined by their cholesterol content (34-36). In general, plasma membranes are affected at much lower digitonin concentrations than either endoplasmic reticulum or mitochondria. Such differences in membrane cholesterol content and susceptibility to digitonin have also been found in platelets (10,39).

Fig. 7 shows the fraction of intracellular 45Ca2+ and various marker enzymes released from platelets by digitonin. In each case the amount of enzyme or 45Ca2+ released is expressed as a fraction of total platelet-associated activity determined by exposing platelets to 0.2% Triton X-100. An initial plateau of approximately 50% 45Ca2+ release was reached at 40 p~ digitonin. Lactate dehydrogenase, an enzyme present in the cytosol, was liberated by approximately the same concentrations of digitonin. Increasing the digitonin concentration above 500 p~ released an additional 20% of platelet 45Ca2+ and 80% of platelet NADH dehydrogenase and glutamate dehydrogenase. In platelets, NADH dehydrogenase is located in the dense tubular system (40). Glutamate dehydrogenase is located in the mitochondrial matrix (38). No further release of the enzymes or 45Ca2+ occurred when platelets were incubated with digitonin concentrations as high as 25 mM.

In nine such studies, 100 digitonin released 100% of platelet lactate dehydrogenase and 50 k 3% of total platelet 45Ca2+. For comparison, the fraction of intracellular &Ca2+ located in the rapidly exchangeable pool at the end of a 2-h loaing period can be calculated using the data in Table I. In 34 45Ca2+ efflux studies performed under steady-state conditions, the more rapidly exchanging pool contained an average of 52 pmol/108 platelets and had a tllz of 17 min. The slowly exchanging pool contained 183 pmol/108 platelets and had a t l l2 of 313 min. At steady state, these values apply to 45Ca2+ influx as well as 45Ca2+ efflux. Using Equation 2, the calculated fraction of total platelet 45Ca2+ located in the rapidly exchanging pool after a 2-h incubation with 45Ca2+ is 55%. The close agreement between this value and the fraction of 45Ca2+ re-

i f ii 20

n - 0 I 2 3

Log [Dqitonm](pM) FIG. 7. Release of marker enzymes and intracellular Wa2+

by digitonin. Gel-filtered platelets were incubated for 2 h with 0.2 mM ''CaC12 and then gel filtered a second time to remove nonincorporated &Cas+. Immediately afterwards, digitonin (1 g~ to 5 mM) and EGTA (1 mM) were added. After a 5-min incubation at 4 "C, the platelets were sedimented by cent r i f~a t ion a t 12,000 X g for 10 min, and &Ca2+ (01, lactate dehydrogenase (O), glutamate dehydrogenase (D), and NADH dehydrogenase (A) were measured in the supernatant. Each of the results shown is the mean of at least 3 studies. The values are expressed as a fraction of total platelet-associated activity determined by completely dissolving the platelets with 0.2% Triton x-100.

12568 ea2+ Homeostasis in Platelets

leased concurrently with lactate dehydrogenase suggests that the rapidly exchanging Ca2+ pool is located in the platelet cytosol.

The remaining 05Ca2+ that could be released from the platelets by digitonin appeared in the supernatant at the same digitonin concentrations that affected the permeability of mitochondria and the dense tubular system. Since there was only a small difference in the amounts of digitonin required to release NADH dehydrogenase and glutamate dehydrogenase, the relative contributions of these organelles cannot be determined by this method. Approximately 30% of the intracellular 45Ca2+ was not released at digitonin concentrations as high as 25 mM. Part of this 05Ca2+ may be accounted for by incomplete permeabili~tion of mitochondria or dense tu- bules; at most only 80% of the enzymes marking these organelles were liberated (Fig. 7).

Inhibitors of Cu2+ Transport Studies performed with isolated platelet membrane vesicles

suggest that cyclic AMP potentiates Ca" uptake into the dense tubular system (9). In order to examine this issue in intact cells, platelets were incubated with 50 nM PG12, which increases platelet cAMP levels and inhibits platelet aggregation (5). This concentration of PGI, was sufficient to completely inhibit platelet aggregation produced by 10 p~ ADP (not shown). After 30 min, 45Ca2+ was added. PGI, had no effect on the size of either of the exchangeable Ca2+ pools. PGI, also had no effect on the cytosolic free Ca2+ concentration measured 1 h after the addition of the inhibitor (Table IV).

ea2+ uptake into the dense tubular system is also believed to be accelerated by calmodulin (41, 46). '%a2+ efflux was measured in platelets preincubated with 20 p~ trifluoperazine, a calmodulin inhibitor. The size of the cytosolic Ca2+ pool increased from 35 to 45 pmol/lOs platelets. The size of the sequestered Ca" pool decreased by one-third. The cytosolic free Ca2+ concentration measured 1 h after the addition of trifluoperazine increased from 113 to 304 nM (Table IV).

In other tissues, Ca2+ accumulated by mitochondria in situ

TABLE IV The effect of inhibitors OR Ca2' homeostasis irs intact platelets

'%a2+ efflux was measured at 0.2 mM extracellular Ca2+ with each of the inhibitors present during both the '%a2+ loading period and the 44Ca2+ efflux period. The sizes of the cytosolic Ca2+ pool (S,) and the sequestered Ca2+ pool (S,) were calculated as described under "Materials and Methods." The values shown are the mean f S.E. expressed as picomoles per 108 platelets per min. "Free" refers to the cytosolic free ea2+ concentration measured with quin2 1 h after the addition of the inhibitors. The p values were calculated by paired t test. NS, not significant. TFP, trifluoperazine.

Control Inhibitor n P o 0 1

S2 1 5 3 2 21 97 +. 13 5 ~ ( 0 . 0 2 5 TFP (20 pM) S , 35 f 3 45 f 4 5 pCO.025

Free 113 f 14 304 f 16 6 p < 0.001

CCCP(10pM)+ SI 3 3 & 2 4 5 f 4 6 p(O.01 KCN (2 mM) Sz 152 f 13 141 f 31 6 NS

Free 1 1 3 f 14 1 2 3 f 16 6 NS

CCCP (10pM) + SI 5 9 f 4 73 f 6 6 ~ ( 0 . 0 5 KCN (2mM) + S2 2 6 0 f 43 1 9 6 f 2 6 6 pC0 .05 deoxyglucose Free 113 2 14 271 f 34 6 p<0.005 ( 10 mM)

PGlz (50 nM) Si 5 3 2 7 5 4 k 6 5 NS S2 2 1 4 2 32 1 7 5 f 19 5 NS Free 1 1 3 f 14 1 1 5 f 11 6 NS

can be released by uncoupling agents such as CCCP (35,42). In the present studies, gel-filtered platelets were preincubated with 10 pM CCCP and 2 mM KCN (to inhibit oxidative phosphorylation) prior to measurement of Ca" exchange. Addition of the inhibitors produced a small increase in the size of the cytosolic exchangeable Ca2+ pool that was similar to the increase seen with trifluoperazine. Unlike trifluoperazine, however, CCCP and KCN produced no change in the size of the sequestered pool or the cytosolic free Ca2+ concentration (Table IV).

Combining an inhibitor of glycolysis with mitochondrial inhibitors such as CCCP and KCN produces a sustained decrease in platelet ATP levels (43). (Ta2+ efflux was measured from platelets preincubated with 10 mM 2-deoxyglucose in addition to 10 p~ CCCP and 2 mM KCN. This combination of inhibitors produced a 20% decrease in the size of the slowly exchanging pool and a small increase in the size of the cytosolic pool and the cytosolic free Ca" concentration (Table IV). These results are consistent with an ATP requirement for both Ca2+ efflux and Ca2+ uptake into the sequestered Ca" pool. An energy requirement for Ca2+ homeostasis is also suggested by observations of "Ca2+ efflux at 4 "C; decreasing the temperature markedly inhibited "Ca2+ efflux. Increasing the temperature from room temperature to 37 "C had little effect (data not shown).

DISCUSSION

Prior to stimulation by an agonist, platelets maintain a low cytosolic free Ca" concentration and a steep plasma membrane Ca" concentration gradient. In other tissues, this is accomplished by limiting Caz+ influx and actively transport- ing Ca2+ from the cell. In addition, many tissues actively sequester cytosolic Caz+ within intracellular sites such as endoplasmic reticulum, sarcoplasmic reticulum, and mitochondria (32). Most of the published studies on Ca" transport in platelets have been performed either with membrane fragments (&10,25,26) or with platelets that have been exposed to an agonist (1, 16, 21,271. The present studies examine the mechanisms that maintain Ca2+ homeostasis in intact unstimulated platelets by measuring the kinetics of Caz+ exchange. The data show that exchangeable Ca2+ within platelets be- haves as if it were located in at least two physically distinct pools, one in the cytosol and one sequestered within an internal site, most likely the dense tubular system.

Control of Cu2+ Influx-Ca2+ influx was maximal at free Ca" concentrations 20-fold less than in plasma and inhibited by moderate concentrations of Gd3+ and La3+ (400 p ~ ) . Gd3+ competes with Caz+ for binding sites on the platelet surface but does not enter platelets or increase cAMP levels (16, 28). These results suggest, therefore, that CaZ+ influx occurs through a selective Ca2+ channel. The data also suggest that only a single class of channels is present, but the precise nature of this channel is unknown. Other investigators have shown smooth muscle Ca2+ channel blockers, such as verapamil and nifedipine, inhibit platelet aggregation (22, 231, secretion (23,24), and thromboxane production (23) and block an epinephrine-i~duced increase in *%a2+ uptake (21). Some of these effects, however, appear to be due to changes in receptor affinity rather than a specific effect of Ca" influx (22). We found that nifedipine, but not verapamil, produced a small decrease in the cytosolic free Caz+ concentration. The significance of this observation in terms of mechanisms of inhibition is not clear since at the same concentration neither nifedipine nor verapamil inhibited 45Ca2+ uptake by platelets. Whether either agent blocks the increase in Ca" exchange across the plasma membrane that occurs after platelet stim-

12569

dation, as has been suggested elsewhere (21), cannot be addressed by the present studies.

The Cytosolic Ca2+ Pool and Ca2+ Efflux-As in other tissues, the more rapidly exchanging Ca2+ pool in platelets appears to be located in the cytosol. The half-time of Ca2+ exchange by this pool is 17 min, similar to that seen in other tissues (29). All of the eCa2+ that has entered this pool after a 2-h incubation can be released by a concentration of digitonin that is similar to that which releases platelet lactate dehydrogenase. These data do not, however, imply that the cytosolic Ca2+ pool need be homogenous. Subsets of the pool that exchange with similar, but not necessarily identical, rates will be averaged together by this type of analysis (18). In fact, studies by Johnson et a1 (3) using aequorin suggest that inhomogeneities in the free Ca2+ concentration may exist within the cytosol.

Based upon the steady-state influx and efflux data, the maximum rate of Ca2+ exchange across the platelet plasma membrane is approximately 1-2 pmol/lO* platelets/min with a K , for extracellular Ca2+ of 20 I.IM (Table 11). Assuming that there is 0.4 mg of protein/l@ platelets and that platelets can be treated as spheres with a diameter of 1 p in order to calculate their surface area (undoubtedly an underestimate), the Ca2+ exchange rate is, at most, 4 pmol/min/mg of protein or 6 fmol/cm2/s. These values are 10-fold less than has been measured in a variety of other vertebrate tissues (29). The reason for this difference is unknown but may reflect the fact that platelets are fragments of megakaryocyte cytoplasm. Since platelets lack the capacity to replace damaged or senes- cent proteins, including transport proteins, the results obtained in the present studies reflect the average of a poten- tially heterogenous population. The maximum efflux rates may, therefore, underestimate the capabilities of a newly formed platelet.

Under steady-state conditions, the Ca2+ efflux rate varied with the extracellular Ca2+ concentration (Fig. 2). Under non- steady-state conditions, however, &CaZ+ efflux was unaffecixd by decreasing the extracellular Ca2+ concentration from 0.2 mM to 1 nM (Fig. 3). This phenomenon, which has also been seen in other tissues (30), may be more apparent than real. Since the size of the cytosolic exchangeable Ca" pool and the cytosolic free Caz+ concentration, but not the CaZ+ influx rate, are relatively unaffected by wide changes in the extracellular Ca2+ concentration, the efflux rate would be expected to vary with the extracellular Ca2+ concentration. Presumably, Ca2+ efflux varies with incremental changes in the cytosolic free Ca2+ concentration, acting to stabilize free Ca2+ at approximately 100 nM. The results of the non-steady-state experi- ments and the differential effect of Gd3+ on Ca2+ influx and efflux suggest that Ca2+ efflux is not directly linked to Ca2+ influx and does not occur by simple Ca2+/Ca2+ exchange.

In other tissues, Ca2+ efflux is mediated by either a Caz+- Me-ATPase or by Na2/Ca2+ exchange (29, 49, 50). Data presented in the accompanying paper show that changes in the plasma membrane Na+ gradient affect the kinetics of Ca2+ exchange in unstimulated platelets but not in the manner that would be expected if Na+/Ca2+ exchange were a major contributor to Ca2+ efflux (14). In addition, the cytosolic free Ca2+ concentration is unaffected by removing the Na+ gradient. Therefore, Na+/Ca2+ exchange seems unlikely to play a major role in unstimulated platelets. On the other hand, platelet lysates do contain CaZ'-MgZ+-ATPase activity, How- ever, histochemical studies have failed to demonstrate the enzyme in the platelet plasma membrane (12), and vesicles formed from platelet plasma membranes were unable to accumulate Ca2+ in the presence of ATP and oxalate (10).

Therefore, Ca2+-Me-ATPase-mediated transport has not been thought to account for Ca2+ efflux from platelets.

What, then, is the mechanism that mediates Ca2+ efflux from platelets? Some of its properties may be inferred from available data, including the studies presented here. 1) The process is undoubtedly energy-dependent, either directly or indirectly. 2) Ca2+ efflux may be stimulated by calmodulin; trifluoper~ine, which inhibits some c a l m ~ ~ n - d e p e n d e n t processes, increased the cytosolic free Ca2+ concentration and the size of the cytosolic exchangeable Caz+ pool. However, interpretation of this observation must be tempered by the fact that trifluoperazine, even at low concentrations, may also interfere with calmodulin-independent processes. 3) Ca2+ efflux is inhibited by high concentrations of La3+ ( H . 5 mM). 4) The Km for Ca2+ of the efflux mechanism must be approximately 0.1 PM, since this is the free Ca" concent~tion in the platelet cytosol. We recently reported that Ca" exchange is markedly abnormal in platelets of patients with Glanzmann's thrombasthenia (45). Platelets in this disorder lack membrane glycoproteins IIb and IIIa. Similar changes in Ca2+ transport could be produced in normal platelets by incubation under conditions that dissociate the IIb/IIIa complex. These observations may provide a new clue to the nature of the Ca2+ efflux mechanism.

The Sequestered Ca2+ Fool and Internal Ca2+ Transport- The dense tubular system is thought to play a role in platelets that is analogous to sarcoplasmic reticulum. This conclusion is based upon the observation that vesicles believed to be formed from the dense tubular system possess both Caz+- Mg2+-ATPase activity and the ability to sequester Ca" in the presence of ATP (9,10,25). The ATP-dependent Ca2+ uptake system has been partially characterized. Ca2+ uptake is half- maximal at approximately 0.1 pM (E) , which is similar to both the free Ca2+ concentration in the platelet cytosol and the K , for Caz+ of the isolated platelet Ca2+-Mg2+-ATPase (11). Some studies have also demonstrated enhanced Ca2+ uptake into membrane vesicles in the presence of CAMP and protein kinase (9) and inhibition of uptake by phenothiazines (46), suggesting a role for both CAMP and calmodulin.

In the present studies, it was possible to address this issue in intact platelets by observing the properties of the more slowly exchanging intracellular Ca2+ pool. Ca2+ exchange into this pool occurred with a t1f2 of several hundred minutes, similar to the results obtained in other tissues (29). The maximum rate of Ca" flux into the pool was approximately 0.6 pmol/l@ platelets/min and had a K, for extracellular Ca" of 42 FM (Table 11). Since the slowly exchanging pool was assumed in this analysis to exchange with the cytosolic ea2+ pool, but not directly with extracellular Ca2+, this Km presumably reflects the extracellular Ca2+ concentration at which the cytosolic Ca2+ concentration reaches the true Km for Ca2+ transport into the sequestered pool. Consistent with this interpretation, Dean and Sullivan (11) have reported that the K , of the isolated platelet Caz+-Mg2+-ATPase is approximately 0.1 pM, which is also the cytosolic free Caz+ concentration in platelets. In contrast to the cytosolic Ca" pool, the size of the slowly exchanging Ca2+ pool was markedly dependent on the extracellular Ca'+ concentration (Fig. 4). This property of the slowly exchanging pool has also been reported in kidney cells studied by the same techniques (31). The observation supports the concept that the slowly exchanging pool is physically, as well as kinetically, distinct from the rapidly exchanging pool and may possess the capacity to "buffer" the size of the rapidly exchanging pool by helping to remove Ca" from the cytosol.

Ass i~men t of the slowly exchanging pool to the dense

12570 Ca2+ Homeostasis in Platelets

tubular system is based upon concurrent release by digitonin of noncytosolic *%a2+ and NADH dehydrogenase, an enzyme associated with endoplasmic reticulum (37) and the dense tubular system (40). An alternative possibility, that part of the ea2+ was sequestered in platelet mitochondria, appears less likely for several reasons. An inhibitor of mitochondrial Ca2+ uptake (CCCP) had no effect on either the size of the slowly exchanging pool or the cytosolic free Ca2+ concentration unless an inhibitor of glycolysis was also present. By comparison, inhibition of mitochondrial ea2+ uptake in kidney cells reduces the size of the slowly exchanging eaz+ pool by >95% (31). This suggests that the reduction in the size of the slowly exchanging pool produced by combining CCCP, KCN, and 2-deoxyglucose is due to decreased platelet ATP levels rather than to selective inhibition of mitochondrial Caz+ uptake. In support of this conclusion, Steiner and Ta- teishi (13) have shown that very little of the 4sCa2+ that enters platelets is found in a fraction of the platelet homogenate that contains the mitochondria. Feinstein found no change in fluorescence when platelets loaded with chlortetracycline were incubated with agents that release mit~hondriai ea2+, such as carbonyl cyanide p-trifluoromethox~henylhydrazone or antimycin A (47). Chlortetracycline fluorescence reflects membrane-associated ea2+, including mitochondrial Ca2+. In contrast, Scharf and Luscher (48) have reported that isolated platelet mitochondria accumulate Ca2+ and suggested that this process contributes to platelet ea2+ homeostasis. They did not, however, address whether platelet mitochondria can accumulate ea2+ at the free ea2+ concentration normally present in the platelet cytosol. Even though Ca2+ may be present within the mitochondria, studies in other tissues suggest that significant ea2+ flux into mitochondria may not occur at 0.1 PM free eaz+ (32). The possibility that platelet mitochondria may have a greater role at the higher cytosolic Ca2+ concentrations found after platelet stimulation cannot be addressed by the present studies.

When platelets were incubated with inhibitors of ATP production, the size of the sequestered Ca" pool decreased and the free ea2+ concentration increased. This observation is consistent with earlier studies that have suggested a role for ATP in ea2+ uptake by the dense tubular system. Trifluo- perazine produced similar changes; however, the lack of com- plete specificity by trifluoperazine for calmodulin-dependent processes makes this observation consistent with, but not proof of, a role for calmodulin. In contrast, the size of the sequestered ea2+ pool and the cytosolic free ea2+ concentration were not affected by increasing platelet cAMP levels with PG12. This suggests that cAMP may not affect Ca" uptake into the dense tubular system in resting platelets.

The increase in the cytosolic free ea2+ concentration that occurs when platelets are activated is believed to be due in part to ea2+ release from the dense tubular system (6). This conclusion is supported by recent observations that the rise in cytosolic free ea2+ initiated by thrombin is only partially ablated by adding EGTA. Conversely, the absence of a rise in cytosolic free Ca" with other agonists in the presence of EGTA has been used as evidence that the Ca" increase is due exclusively to the net influx of extracellular Ca2+ (1). Although these points were not specifically addressed by the present studies, the data described in this manuscript suggest that agonist-induced changes in cytosolic free ea2+ observed in the presence of extracellular EGTA should be interpreted cautiously. Since low extracellular Ca" concentrations reduce the size of the dense tubular system Ca" pool, less sequestered Ca2+ is available for release into the cytosol.

In conclusion, the data from these studies suggest that Ca" homeostasis in resting platelets is maintained by limiting Ca" influx, by actively promoting ea2+ efflux, and by ea2+- ~ ~ + - A T P a ~ - d e ~ n d e ~ t uptake of Ca" into the dense tubular system. The precise identity of the ea2+ influx channel and the molecular basis for Ca2+ efflux across the plasma membrane remain to be determined.

Acknowledgments-I would like to thank Dr. Antonio Scarpa whose suggestions during the preparation of this manuscript were much appreciated. I also wish to thank Elizabeth Belmonte for her steadfast technical assistance and Janet L. Waxman for typing the manuscript.

~ ~ R E N C E S 1. Rink, T., Smith, S., and Tsien, R. (1982) FEBS Lett. 148,21-26 2. Purdon, D., Daniel, J., and Holmsen, H. (1983) CircLdation 66, Suppl. 2,

ma 3. Johnson, P. C., Cliveden, P., Smith, M., Lall, P., and Salzman, E. W. (1983)

4. Detweiler, T., Ciaro, I., and Feinman, R. (1978) Thromb. Haemstasis 40,

I Y Y

Blood 62, Su pl. 1,259a

m7-m 1 5. Feinstein, M., Rodan G. and Cutler, G. (1981) in Platelets in Bwlo and

Pathology (Gordon,'J. i., ed) Vol. 2,pp. 437472, Elsener/North-&land Biomedical Press, Amsterdam

6. Gerrard, J., Peterson, D., and White, J. (1981) in Platelets in Bhbgy an$ P a t h o h (Gordon, J. L., ed) Vol. 2, pp. 407-436, Elsevier/North-Holland Biomedical Press, Amsterdam

"_

7. White, J. (1972) Am. J. Pathol. 66,295-312 8. Robbiee, L., and Shepro, D. (1976) Biochim Eiophys. Acta 436,448459 9. Kiiser-Glanzmann. R.. Jakahovi. M.. Georee. J. N.. and Liischer. E. F.

(1977) Eiochim. 'Eiophys. Acta 466; 429-&0 ~, ~ ~

10. Menashi, S., Davis, C., and Crawford, N. (1982) FEBS Lett. 140,29&302 11. Dean, W., and Sullivan. D. (1982) J. Biol. Chem. 257.14390-14394 12. Cutler, L;, Rodan, G., .and Feinstein, M. (1978) Ehchim. Biophys. Acta

13. Steiner M., and Tateishi T. (1974) Eiochim. Biophys. Acta 367,232-246 14. Brass, f.,. (1984) J. Ehl. dhem. 259,12571-12575 15. Javors, M., Bowden, C., and Ross, D. (1982) Biachim. Biophys. Acta 691,

16. Brass, L. and Shattil, S. (1982) J. Ebl. Chem. 2 5 7 , 1 4 ~ 1 4 ~ 5 17. Koeppe, b., and Hamann, C. (1980) Comput. Programs Eiomed. 12,121-

18. Uchikawa, T., and Borle, A. (1978) Am. J. Physiol. 234, R29-R33 19. Murer, E., and Holme, R. (1970) Biochim. Eiophys. Acta 222, 197-205 20. La ea, B., Scrutton, M. C., Holmsen, H., Day, H. J., and Weiss, H. J. (1975)

21. Owen, N., Feinberg, H., and LeBreton, G. (1980) Am. J. Physiol. 239,

542,357-371

220-226

128

hod 46,119-130

U A R L U A R S l 22. Barnathan, E. S., Addonizio, V. P., and Shattil, S. J. (1982) Am. J. Physiol.

23. Ha~-P,-Boatwright, C., and Anllie, N. (1983) Thromb. Haemstasis 60,

**-"I ""

242, H19-H23

24. Zilberman, Y., Gutman, Y., and Koren, R. (1982) Bioehim, Ewphys. Acta

25. Robbiee, L., Shepro, D., and Belamarich, F. (1973) J. Gem Physiol. 61,

26. Detwiler, T., and Feinman, R. (1973) Biochemistry 12.282-289 27. Owen, N. E., and Le Breton, G. C. (1981) Am. J. PhyswL 241, H613-H619 28. Brass, L. F., and Shattil, S: J. (1984) J. Clin. Inuest. 73, 626-632 29. Borle, A. (1981) Reu. Physwl. Blochem. Phurmacol. 90,13-153 30. Kondo, S., and Schulz, I. (1976) J. Membr. Eiol. 29,185-203 31. Borle, A. (1972) J. Membr. Eiol. 10,45-66 32. Somlyo, A. P. (1984) Nature (Lomi.) 309,516-517 33. Williamson, J., Cooper, R., and Hoek, J. (1981) Eiochim. Ebphys. Acta

34. Ronnin , S., Heatley, G., and Martin, T. (1982) Proc. Natl. Acaci. Sei U.

35. Murphy, E., Coll, K., Rich, T. L., and Witiiamson, J. R. (1980) J. Bbi.

36. Meredith, M., and Reed, D. (1982) J. Ewl. Chem 257,3747-3753 37. Hochstadt, J., Quilan, D. C., Rader, R. L., Lin, C.-C., and Dowd, D. (1975)

38. Schmidt, E. (1974) in Methods of Enzymatic Analysis (Bergmeyer, H. U., Methods Membr. Eiol. 5,117-162

39. Akkerman, J. W. N., E z r i n k , R. H. M., Lips, J. P. M., and Christiaens, ed) pp. 650-656, Verl Chemie-Academic Press, New York

40. Menashi, S., Weintroub, H., and Crawford, N. (1981) J. Bid. Chern. 256, G. C. M. L. (1980) Er. J. Haematol. 44,291-300

41. LePeuch, C. J., LePeuch, D. A. M., Katz, S., Demaille, J. G., Hincke, M. 4095-4101

T., Bredoux, R., Enouf, J., Levy-Toledano, S., and Caen, J. (1983) Bioehim. Bwphys. Acta 731,456-464

42. Heytler, P. G., and Prichard, W. W. (1962) Eioehem. Eiophys. Res. Com- rnun. 7,272-275

43. Holmsen, H., and Robkin, L. (1979) Thromb. Haemostasis 42, 1460-1472 44. Duggleby, R. G. (1980) Eur. J. Biochem. 109,93-96 45. Brass L., and Belmonte, E. (1984) Clin. Res. 32,304a 46. Enouf, J., Lqy-Toledano, S., Bredoux, R., and Caen, J. P. (1983) Thromb.

47. Feinstein M. B. (1980) Biochem. Biophys. Res. Commun. 93,593-600 48. Scharf, R: E., and Liischer, E. F. (1979). Thromb. Haemstasis 42.82 49. Blaustein, M. P. (1974) Rev. Ph shl. Blochem. P+rmucol, 7 0 , 3 3 4 2 50. Stekboven, F. S., and Bontmng, 8. L. (1981) Physml. Reo. 61, l -76

01;5-nll

691,106-114

462-481

639,243-295

S. A. $9,6294-6298

Chem. 266,6600-6608

Haemstasls 50,77

the of chemistry vol. 259, no. 20, of october 25. pp. 12563 … · 2001-09-03 · the journal of...

Documents