the biochemical pharmacology of the human platelet …epubs.surrey.ac.uk/851396/1/11012644.pdf ·...
TRANSCRIPT
THE BIOCHEMICAL PHARMACOLOGY OF THE HUMAN
PLATELET ADP RECEPTOR
By
David A. HALL, BSc. (Hons)
A thesis submitted in accordance with the requirements of the University of Surrey for the
Degree of Doctor of Philosophy.
Receptors and Cellular Regulation Research Group, November 1993.
School of Biological Sciences,
University of Surrey,
Guildford,
Surrey. GU2 5XH.
ProQuest Number: 11012644
All rights reserved
INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a com p le te manuscript and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
uestProQuest 11012644
Published by ProQuest LLC(2018). Copyright of the Dissertation is held by the Author.
All rights reserved.This work is protected against unauthorized copying under Title 17, United States C ode
Microform Edition © ProQuest LLC.
ProQuest LLC.789 East Eisenhower Parkway
P.O. Box 1346 Ann Arbor, Ml 48106- 1346
Particles of raw inspiration sleet through the universe all
the time. Every once in a while one of them hits a receptive
mind, which then invents DNA or the flute sonata form or a
way of making light bulbs wear out in half the time. But
most of them miss...
'Wyrd Sisters’, Terry Pratchet.
ACKNOWLEDGEMENT
I would like to thank the Medical Research Council and Ciba-Geigy Pharmaceuticals for
financial support and Prof. L. J. King for the opportunity to work in his laboratories. I
would also like to thank Prof. R. Crow for the use of the simulated ward for blood taking
and Dr. J. Wright for teaching me how to take blood.
Thank you to Dr. Susanna Hourani for her help and guidance through the three years of
my PhD at Surrey and for her sometimes brutal yet always encouraging criticisms of this
thesis during its preparation. Thanks also to Dr. Ian (’Leader of the Pack') Kitchen for his
advice and for being a source of constant amusement to us all. Thank you to Dr. Ian
Matthews and to my previous supervisors at Ciba for their help during the time I spent in
Horsham.
I would like to thank everyone in the lab (Tracy, Lara, Julie, Cheryl, Fiona, Yushe et al.)
for making my time at the Surrey so enjoyable. Special thanks to Julie for giving up her
front room for the two months when most of this thesis was written and for her taxi
service' to the binders in Reading.
Finally, I would like to thank my parents for providing me with constant and unwavering
support, without which I would never have come this far.
1
SUMMARY
The effects of a number of analogues of ADP and ATP were determined on increases in
platelet cytosolic calcium concentration ([Ca2+]j). ADP, 2-methyIthio-ADP, adenosine
5'-0-( 1 -thiodiphosphate) and adenosine 5’-0-(2-thiodiphosphate) each induced increases
in [Ca2+]j in a manner similar to their reported ability to induce platelet aggregation. The
effects of these compounds were antagonized by ATP (a selective ADP receptor
antagonist) and in the case of ADP this antagonism was shown to be competitive by
Schild analysis. 2-chloro-ATP, adenosine 5'-0-( 1 -thiotriphosphate), adenosine
5’-0-(2-thiotriphosphate), P1,P5-diadenosine pentaphosphate and adenylyl 5’-(p,y-
methylene)diphosphonate all inhibited ADP-induced increases in [Ca2+]; in an apparently
competitive manner. The pA2 values obtained for these compounds and for ATP as
inhibitors of increases in [Ca2+]; showed a good correlation with those obtained for these
compounds as inhibitors of aggregation. Adenosine 5,-(a,p-methylene)triphosphonate and
UTP were only weak inhibitors of ADP-induced increases in [Ca2+]; whilst
2-methylthio-ATP non-competitively inhibited the effect of ADP on platelet [Ca2+]j and
these results are also consistent with the effects of these compounds on aggregation. This
strong correlation between the effects of analogues of adenine nucleotides on platelet
aggregation and on increases in [Ca2+]f confirms the causal relationship between these
two actions of ADP and shows that they are mediated by the same receptor.
The effects of suramin, an antagonist of P2X and P2Y receptors, were investigated on a
number of platelet responses. In washed platelets suramin acted as an apparently
competitive inhibitor of platelet aggregation causing parallel rightward shifts in the log
concentration-response curve with no depression of the maximal response. However, the
2
slope of the Schild plot was around 2 suggesting that the effect of suramin was not simply
competitive. The value of suramin was 4.62. Suramin was also able to inhibit
non-competitively aggregation induced by U46619 and by 5-hydroxytryptamine (5-HT).
Suramin inhibited ADP-induced increases in [Ca2+]j in a similar manner to aggregation
giving a Schild plot slope of ~2 and a pA2 value of 4.63. Again suramin was also able to
inhibit non-competitively increases in [Ca2+]j induced by 5-HT. Suramin also inhibited the
effect of ADP on prostaglandin Ej (PGE^-stimulated adenylate cyclase, however, in this
case the slope of the Schild plot was not significantly different from unity suggesting that
the effect was competitive. The pA2 value of suramin was 5.09. Suramin was also able to
inhibit the stimulation of adenylate cyclase by PGEr These results show that suramin is
an antagonist of the platelet ADP receptor, as it is of P2X and P2Y receptors, although it
also has non-specific inhibitory effects.
The effect of varying the extracellular divalent cation concentration was determined on
platelet responses to ADP. There was a small but significant leftward shift in the log
concentration-response curve to ADP in the presence of EGTA compared to that in the
presence of both calcium and magnesium. This leftward shift was no longer significant
when the log concentration-response curves were plotted against ADP3' rather than
against total ADP concentration. There was also a marked increase in the pA value of
ATP in the presence of EGTA compared to the presence of calcium and magnesium.
However, when the pA^ value was determined for ATP4* there was no longer a significant
difference in the pA2 under these different conditions. Similar results were obtained when
the inhibition of PGE,-stimulated adenylate cyclase was compared in the presence of
calcium and magnesium and in the presence of EGTA. These results suggest that, in
common with other P2-purinoceptors, the platelet ADP receptor recognises the
uncomplexed forms of adenine nucleotides.
In studies on the effect of divalent cation concentration on ADP-induced inhibition of
adenylate cyclase, ATP not only inhibited the effect of ADP but also appeared to enhance
the accumulation of cAMP induced by PGEj. However, ATP was unable to induce an
increase in adenylate cyclase activity in the absence of PGEj and apyrase also enhanced
the accumulation of cAMP induced by PGEj. This suggests that the effect of ATP was in
fact due to inhibition of the effect of endogenous ADP released by platelets during
preparation and storage rather than a true stimulation of adenylate cyclase.
In binding studies, platelets showed both high and low affinity binding sites for
[35S]-Sp-ATPaS. There were -9000 high affinity sites per platelet with a KD value of
0.21 pM and -24,000 low affinity sites with a KD value of 32 pM. This binding was not
displaced by selective ligands at adenosine, or 5-HT2 receptors and was not displaced
by FSBA but was displaced by a number of competitive ADP receptor antagonists. The
low affinity binding was not displaced by suramin. However, the ability of some of these
ligands to displace bound ATPaS was not consistent with their effects in functional
studies suggesting that neither of these binding sites is the ADP receptor.
4
CONTENTS
Page
ACKNOWLEDGEMENT 1
SUMMARY 2
LIST OF TABLES AND FIGURES 9
LIST OF PUBLICATIONS 13
CHAPTER 1 - GENERAL INTRODUCTION 14
1.1 Structure & Function of Platelets 15
1.2 Purinoceptors 19
1.3 Metabolism of Extracellular Adenine Nucleotides by Platelets 22
1.4 Platelet Responses to ADP 23
1.4.1 Morphological Responses 23
1.4.2 Biochemical Responses 26
1.5 Measurement of Platelet Responses 35
1.6 Structure-Activity Relationships of the ADP Receptor 36
1.6.1 Agonists 36
1.6.2 Antagonists 39
1.6.3 Multiple ADP Receptors? 41
1.7 Aims of this Thesis 46
CHAPTER 2 - STUDIES ON INCREASES IN INTRACELLULAR 47
CALCIUM CONCENTRATION
5
2.1 Introduction 48
2.2 Materials and Methods 50
2.2.1 Materials 50
2.2.2 Platelet Aggregation Studies 51
2.2.3 Loading of Platelets with Fura-2 51
2.2.4 Calcium Measurement 52
2.2.5 HPLC Separation of Nucleotides 53
2.2.6 Nucleotide Purification 54
2.2.7 Nucleotide Synthesis 56
2.2.7.1 Synthesis of Sp-ADPaS 56
2.2.7.2 Synthesis of Sp-ATPaS 57
2.2.7.3 Synthesis of Sp-ATPf3S 58
2.3 Results 59
2.4 Discussion 68
CHAPTER 3 - STUDIES WITH SURAMIN 75
3.1 Introduction 76
3.2 Materials and Methods 77
3.2.1 Materials 77
3.2.2 Platelet Aggregation Studies 78
3.2.3 Calcium Measurement 79
3.2.4 Measurement of Inhibition of PGEj-stimulated Adenylate Cyclase 81
6
3.3 Results 82
3.4 Discussion 94
CHAPTER 4 - EFFECTS OF EXTRACELLULAR DIVALENT CATION 102
CONCENTRATION
4.1 Introduction 103
4.2 Materials and Methods 105
4.2.1 Materials 105
4.2.2 Measurement of Platelet Shape Change 105
4.2.3 Measurement of Inhibition of PGEj-stimulated Adenylate Cyclase 107
4.2.4 Determination of Nucleotide Release from Platelets 107
4.2.5 Calculation of Free ADP3* and ATP4* Concentrations 108
4.3 Results 109
4.4 Discussion 115
CHAPTER 5 - BINDING STUDIES 125
5.1 Introduction 126
5.2 Materials and Methods 131
5.2.1 Materials 131
5.2.2 Preparation of Washed Platelets 132
5.2.3 Time Course of ATPaS Binding 132
5.2.4 Displacement Studies 133
5.2.5 HPLC Studies 135
7
5.3 Results 135
5.4 Discussion 144
CHAPTER 6 - GENERAL DISCUSSION 151
6.1 Suggestions for Future Work 158
REFERENCES 161
8
LIST OF FIGURES AND TABLES
Figure 1.1
Figure 1.2
Figure 1.3
Figure 1.4
Table 2.1
Figure 2.1
Figure 2.2
Figure 2.3
Figure 2.4
Figure 2.5
Page
A diagrammatic representation of a section through a human platelet 17
showing the major structural features.
A summary of the major processes thought to be associated with the 33
increase in cytosolic calcium concentration induced by the ADP
receptor.
The structures of some ADP analogues and 5’-/>-fluorosulphonyl- 37
benzoyladenosine.
The structures of some ATP analogues. 42
A comparison of the retention times of a number of adenine 55
nucleotide analogues in the two HPLC systems used in this study.
A typical fluorimeter trace from dual wavelength calcium 60
measurement.
The effect of ATP on increases in cytosolic calcium concentration in 61
human platelets induced by ADP.
Increases in platelet cytosolic calcium concentration induced by ADP 63
alone or in the presence of Cl-ATP, ATPaS, ATPpS or Ap5A.
Increases in platelet cytosolic calcium concentration induced by ADP 64
alone or in the presence of AMPPCP, AMPCPP, UTP or MeSATP.
Platelet aggregation induced by ADP alone or in the presence of 65
ATPpS or ATPaS.
9
Figure 2.6
Figure 2.7
Figure 2.8
Figure 3.1
Figure 3.2
Figure 3.3
Figure 3.4
Figure 3.5
Figure 3.6
Figure 3.7
Increases in platelet intracellular calcium concentration induced by 66
ADP, MeSADP or MeSADP in the presence of ATP.
Increases in platelet cytosolic calcium concentration induced by 67
ADPaS or ADPPS and the effect of ATP on these increases.
Comparison of the abilities of a number of ATP analogues to inhibit 72
responses to ADP.
The structure of suramin. 77
A typical fluorimeter trace from single wavelength calcium 80
measurement.
Aggregation of human platelets in plasma induced by ADP alone or 84
in the presence of suramin alone, ATP alone or both suramin and
ATP.
The effect of suramin on washed platelet aggregation induced by 85
ADP.
Schild plots obtained when washed platelets were preincubated with 86
suramin for 0,10 or 40 minutes.
Aggregation of washed platelets induced by U46619 or 5-HT (in the 88
presence of 100 pM adrenaline) alone or in the presence of suramin
or ATP.
Effects of suramin on increases in platelet cytosolic calcium 89
concentration ([Ca2+]j) induced by ADP.
10
Figure 3.8 Effects of suramin on increases in platelet cytosolic calcium 90
concentration induced by 5-HT.
Figure 3.9 Effects of suramin on the inhibition of PGEj-stimulated cAMP 91
accumulation by ADP.
Figure 3.10 Inhibition of PGEr stimulated cAMP accumulation by adrenaline in 92
the presence or absence of suramin.
Figure 3.11 The accumulation of cAMP in human platelets induced by PGEj in 93
the presence or absence of suramin.
Table 4.1 The pD2 values for the induction of platelet shape change by ADP 112
and ADP3* and the pA2 values for its inhibition by ATP and ATP4*.
Figure 4.1 Shape change induced by ADP alone or in the presence of ATP, in 110
the presence of both magnesium and calcium, magnesium alone or
calcium alone.
Figure 4.2 Shape change induced by ADP alone or in the presence of ATP , in 111
the absence of added divalent cations or in the presence of EGTA. A
comparison of the data presented in figures 4.1 and 4.2a & b.
Figure 4.3 Inhibition of PGEj-stimulated cAMP accumulation by ADP alone or 113
in the presence of ATP, in the presence of both magnesium and
calcium or in the presence of EGTA.
Figure 4.4 The data presented in figure 4.2c shown in terms of ADP3* 116
concentration.
11
Figure 4.5
Figure 4.6
Table 5.1
Figure 5.1
Figure 5.2
Figure 5.3
Figure 5.4
Figure 5.5
Figure 5.6
Figure 5.7
Figure 6.1
The data presented in figure 4.3 shown in terms of ADP3* 117
concentration.
Shape change induced by ADP in the presence of both magnesium 120
and calcium or in the presence of EGTA calculated as % of the
maximal response.
The K; values for ADP receptor antagonists for the displacement of 140
[35S]-ATPaS from high affinity and low affinity binding sites on
human platelets.
Binding of ATPaS to platelets using a saturation protocol. 134
Kinetics of the binding of [35S]-ATPaS to washed human platelets. 137
Homologous ATPaS displacement isotherm. 138
Displacement of [35S]-ATPaS by a number of ADP receptor 139
antagonists.
Displacement of [35S]-ATPaS by dimethylformamide, FSB A, 142
ketanserin, yohimbine, UTP, CTP and GTP.
HPLC trace showing the release of ATP from platelets under the 143
conditions of a binding experiment.
Comparison of the abilities of a number of ATP analogues to displace 146
bound ATPaS from human platelets and to inhibit platelet responses
to ADP.
A comparison of the ability of ADP to induce shape change, increases 160
in [Ca2+]i and aggregation of washed human platelets.
12
LIST OF PUBLICATIONS
The publications arising from this thesis are listed below:
HOURANI, S. M. O., HALL, D. A. & NIEMAN, C. J. (1992) Effects of the
P2-purinoceptor antagonist, suramin, on human platelet aggregation induced by adenosine
5-diphosphate. Br. J. Pharmacol 105,453-457.
HALL, D. A. & HOURANI, S. M. O. (1992) Antagonism by suramin of increases in
intracellular calcium in human platelets induced by ADP. Int. J. Purine Pyrimidine Res.
3, 81 (Abstract).
HALL, D. A. & HOURANI, S. M. O. (1993) Effects of analogues of adenine nucleotides
on increases in intracellular calcium mediated by P2T-purinoceptors on human blood
platelets. Br. J. Pharmacol. 108, 728-733.
HALL, D. A. & HOURANI, S. M. O. (1994) Effects of suramin on increases in cytosolic
calcium and on inhibition of adenylate cyclase induced by adenosine 5-diphosphae in
human platelets. Biochem. Pharmacol., in press.
13
CHAPTER 1
GENERAL INTRODUCTION
14
1.1 Structure and Function of Platelets
Platelets are small, anucleate cells derived from bone marrow megakaryocytes. Under
normal circumstances they circulate freely in the blood showing no interaction with the
surrounding vessel walls (Roth, 1992). If, however, the vessel is damaged, platelets
adhere to the exposed subendothelium, via a specific interaction with collagen (Marcus &
Safier, 1993; Roth, 1992; Siess, 1989). Fibrillar collagen is an extremely potent stimulus
to platelets and causes them to release a number of soluble mediators, such as ADP
(which is also released from cells when damaged), thromboxane (TXA2) and
5-hydroxytryptamine (5-HT) (Hourani & Cusack, 1991; Siess, 1989). These substances
act as a positive feedback system activating circulating platelets and causing them to
adhere to the platelets already recruited to the developing thrombus. Other proteins
released from activated platelets help to stabilize the platelet aggregate by strengthening
interactions between the recruited platelets. Platelet aggregates also form a surface upon
which the clotting cascade can be activated (Ehrman et al, 1978) resulting in the
formation of thrombin which is another potent platelet activator and is also the enzyme
which converts fibrinogen into fibrin, a process which further stabilizes the developing
thrombus. Thus the exposure of the subendothelium in damaged blood vessels initiates a
series of events which results in the formation of a tight platelet plug at the site of the
injury which prevents further and potentially fatal blood loss from occurring. ADP is well
known for participating in aggregation induced by a number of other agonists in vitro
(Hourani & Cusack, 1991; Siess, 1989) and evidence is now accumulating which suggests
that ADP also plays an important role in vivo during thrombogenesis (Cattaneo et ah,
1991; Cattaneo et a l, 1992; Emms & Lewis, 1986; Hirata et al., 1993; Kim et al., 1992;
McClure et al., 1988).
Resting platelets are discoid (lentiform) in shape and about 3 jum in diameter (a
diagrammatic representation of a platelet is shown in figure 1.1). This shape is
maintained by a circumferential ring of microtubules located just below the plasma
membrane (Behnke, 1970). Disruption of this microtubule ring e.g. by chilling the
platelets or treating them with colchicine results in the loss of the discoid morphology,
the platelets becoming spherical (White & Krivit, 1967; White, 1968a). This change in
platelet shape is also seen in the early stages of the response of platelets to various
stimuli. Platelets also contain a dense cytoskeleton of actin filaments which undergoes
extensive changes during platelet shape change.
Platelets contain few mitochondria, but possess many glycogen granules which act as a
store of glucose which is used to supply the increased energy demand during platelet
activation (Crawford & Scrutton, 1987). They also contain numerous cytoplasmic storage
granules (White, 1987). These granules can be classified into three different types, dense
granules, a-granules and lysosomes, depending on their appearance under the electron
microscope, the nature of their contents and the time course of their release from the
platelet during activation. Dense granules appear as electron dense bodies with a
characteristic 'bull’s eye' appearance under the electron microscope (White, 1987). Their
contents seem to be the most readily released by platelet stimuli (Akkerman et a l , 1982).
These granules contain large quantities of the adenine nucleotides ADP and ATP and
biogenic amines (predominantly 5-hydroxytryptamine in the case of human platelets) as
well as inorganic phosphate and pyrophosphate in a macromolecular complex with
calcium and magnesium (Fukami et a l , 1980; Martin et a l, 1974; Meyers et a l, 1982;
Plescher et a l, 1971; Skaer et a l, 1974). Dense granules also contain smaller quantities
16
Microtubules
Open Canalicular System
Dense Tubular System
lycogen
Dense Granule
a-Granule/Lysosome
Mitochondrion
Figure 1.1 A diagrammatic representation of a section through a human platelet showing the major structural features.
17
of guanine nucleotides (Holmsen, 1985). Both ADP and 5-HT are platelet aggregating
agents and thus constitute a positive feedback mechanism helping to recruit further
platelets to a developing thrombus. Along with ATP these substances also have potent
effects on the tone of the surrounding vascular smooth muscle, causing relaxation in the
presence of endothelium or contraction if the muscle is exposed to these mediators
directly (Vanhoutte, 1988). Alpha granules are protein storage granules which contain a
wide variety of substances including adhesive proteins (e.g. Von Willebrand factor and
thrombospondin), coagulation factors (e.g. factor V) and cell mitogens (e.g
platelet-derived growth factor), as well as fibrinogen, an essential factor in platelet
aggregation (for a recent review on a-granules see Harrison & Cramer, 1993). Lysosomes
are morphologically indistinguishable from a-granules, however, they can be identified
as a distinct subset of these granules using cytochemical staining techniques (Bentfield &
Bainton, 1975). Lysosomal acid hydrolases are only releasable by very strong platelet
stimuli such as high concentrations of thrombin or collagen and are not released upon
stimulation with ADP (Kaplan et a l, 1979). Indeed, platelet agonists can be classified by
their abilities to induce the release of platelet granule contents: weak agonists (e.g. 5-HT
& ADP) do not cause the release of granule contents at physiological calcium
concentrations; intermediate agonists (TXA^ vasopressin & platelet-activating factor)
induce the release of dense and a-granule contents but not those of lysosomes; strong
platelet agonists (thrombin & collagen) can induce the release of all three classes of
secretory granules although lysosomal contents are only released at high concentrations
(Siess, 1989).
Platelets contain two closely associated membrane systems, the dense tubular system and
18
the open canalicular system. The dense tubular system is wholly cytoplasmic and is
derived from the endoplasmic reticulum of the parent megakaryocyte (White, 1987). It is
probably the site within the platelet of the agonist-sensitive calcium store (Skaer et al.,
1974) and possesses an active uptake system (Ca2+-ATPase) which allows it to
accumulate calcium from the cytoplasm, a process which is important in the maintenance
of the resting cytosolic calcium concentration (Cutler et al. 1978; Kaser-Glanzmann et
al., 1977). The open canalicular system is formed from extensive invaginations of the
platelet plasma membrane and appears to come into close association with elements of
the dense tubular system (White, 1972). It is thought to act as a conduit for the contents
of the cytoplasmic granules when they are released (White, 1973). During platelet
activation the cytoplasmic granules are concentrated at the centre of the cell due to a
constriction of the circumferential microtubule ring (White, 1968b). The storage granules
then discharge their contents into the extracellular medium by fusing with elements of the
open canalicular system (White, 1973; 1974a). In common with the dense tubular system
the open canalicular system also possesses a Ca2+-ATPase activity which participates in
the maintenance and restoration of the resting cytoplasmic calcium concentration by
extruding calcium across the plasma membrane into the extracellular medium (Cutler et
al, 1978; Rink & Sage, 1987).
1.2 Purinoceptors
Extracellular adenosine and adenine nucleotides are known to have biological effects on
a wide range of tissues via specific cell surface receptors known as purinoceptors.
Purinoceptors were initially subclassified as P} and P2 purinoceptors which respond to
either adenosine (P^ or ADP and ATP (P2), AMP has little if any activity at either
19
subclass of purinoceptor. The effects of adenosine at Pj-purinoceptors were antagonized
by substituted xanthines (e.g. caffeine and theophylline) whilst these compounds had no
effect at P2-purinoceptors (see Bumstock, 1978).
The Pj purinoceptors have now been further subclassified into A, and A receptors. At Aj
receptors the N6-substituted adenosine analogue N6-cyclopentyladenosine (CPA), is more
potent than the 5’ substituted analogue 5'-N-ethylcarboxamidoadenosine (NECA), whilst
at A2 receptors NECA is more potent than CPA. There are also a number of competitive
antagonists of adenosine receptors which show very high selectivity for Aj over
receptors (e.g. 8-cyclopentyl-l,3-dipropylxanthine), although there are as yet no highly
selective \ antagonists. A x receptors occur presynaptically on neurones where they
inhibit neurotransmitter release and occur on certain types of smooth muscle where they
cause contraction (Bumstock, 1989). A2 receptors also occur on various types of smooth
muscle and this class of adenosine receptor also occurs on platelets. In both cases these
receptors are inhibitory causing relaxation of smooth muscle or preventing platelet
aggregation by stimulating adenylate cyclase (Bumstock, 1989; Hourani & Cusack,
1991). There is now evidence that A2 receptors can be further subclassified into A^ and
A^ receptors (Bruns, 1986). The platelet A^ receptor seems to be of the A^ subclass
(Dionisotti et al., 1992). Molecular biological studies suggest that a novel third subclass
of adenosine receptor (A3) also exists (Zhou et a l , 1992). Unlike other Pj receptors this
A3 receptor appears to be insensitive to antagonism by substituted xanthines (Zhou et a l ,
1992). For a recent review of adenosine receptor classification see Collis & Hourani
(1993).
The P2 purinoceptors have been subclassified into at least 5 subclasses, P2X, P2Y, P2Z, P2U
20
and P2T on the basis of structure activity relationships of these receptors for various
adenine nucleotide analogues (Bumstock & Kennedy, 1985; Gordon, 1986; O'Connor et
al, 1991). Some of these receptors have now been cloned (Lustig et al., 1993; Webb et
al, 1993). There are, however, no selective, competitive antagonists for the subtypes of
P2 receptor at present which hampers the production of a definitive system of
sub-classification. A large number of ATP analogues have been synthesized and tested for
their effects at these various receptor subtypes (Cusack & Hourani, 1990). However,
classification of these receptors can be achieved by using only a small number of these
analogues, namely the methylenephosphonate derivatives adenosine 5'-(a,p-methylene)-
triphosphonate (AMPCPP) and adenylyl 5'-(P,y-methylene)diphosphonate (AMPPCP),
the 2-substituted analogue 2-methylthioadenosine 5-triphosphate (MeSATP) and in the
case of the more recently proposed subtypes the pyrimidine nucleotide uridine
5-triphosphate (UTP). The P2X receptor is excitatory and occurs on various smooth
muscles and sensory neurones (Gordon, 1986). At this receptor AMPCPP and AMPPCP
are more potent than the parent nucleotide, ATP, which is equipotent with MeSATP
(Bumstock & Kennedy, 1985). This structure activity relationship is effectively reversed
at P2Y receptors where MeSATP is more potent than ATP whilst the
methylenephosphonate analogues are less potent as agonists at this receptor subtype
(Bumstock & Kennedy, 1985). P2Y receptors are generally inhibitory and occur on a wide
variety of cell types including smooth muscle, endothelial cells and hepatocytes (Gordon,
1986). The P2U receptor has the structure activity relationship UTP « ATP > MeSATP and
shows similar functional effects to the P2Y receptor (O'Connor et a l, 1991). The P2Z
receptor is found on a number of immune cells, e.g. mast cells, and mediates an
ATP-induced cell permeabilization (Gordon, 1986). The agonist at the P2Z receptor seems
21
to be the tetrabasic form of ATP, ATP4-, rather than the divalent cation complex
(MgATP2- & CaATP2-) which is the predominant form in physiological solutions
(Cockcroft & Gomperts, 1979). Methylenephosphonate analogues of ATP are virtually
inactive at this receptor, whilst MeSATP is approximately equipotent (Tatham et al.,
1988). Unlike the P2X and P2Y receptors at which ADP is approximately equipotent with
ATP as an agonist, ADP has no activity at P^ receptors. The P2T receptor is found only on
platelets (and megakaryocytes) where it mediates aggregation. The natural agonist for this
receptor is ADP, ATP acting at this receptor as a competitive antagonist (Macfarlane &
Mills, 1975). The stracture-activity relationships of the P2T receptor will be discussed in
section 1.6.
1.3 Metabolism of Extracellular Adenine Nucleotides by Platelets
All tissues which have so far been studied possess enzymes which can metabolize
extracellular adenine nucleotides (Meghji, 1993; Olsson & Pearson, 1990). In most
tissues ATP and ADP are degraded by sequential dephosphorylation by
'ecto-nucleotidase' enzymes to yield adenosine (Meghji, 1993; Olsson & Pearson, 1990).
This adenosine can then be taken up by the tissue (where it can undergo further
modification by cytoplasmic enzymes) or can be further degraded to inosine by
extracellular adenosine deaminase (Meghji, 1993). Plasma also possesses enzyme
activities capable of sequentially dephosphoiylating these nucleotides (Chambers et al.,
1968). However, although platelets have been reported to possess an 'ecto-ATPase'
activity (Chambers et al., 1968), the major enzyme activity associated with the platelet
surface appears to be nucleoside diphosphokinase (NDPK), which converts ADP to ATP
without the concomitant formation of AMP, (Guccione et al, 1971; Packham et al.,
22
1974). The previously reported ATPase activity may therefore be a consequence of the
NDPK using the added ATP as a ’high-energy phosphate’ donor. This is also consistent
with the observed enhancement of platelet NDPK activity in the presence of exogenously
added ATP (Lips et ah, 1980).
1.4 Platelet Responses to ADP
1.4.1 Morphological Responses
For recent reviews of platelet activation by ADP see Siess, 1989; Gachet & Cazenave,
1991; Hourani & Cusack, 1991. The initial observable response of platelets in suspension
on stimulation with ADP is a change in shape from discoid to spherical with the
extension of pseudopodia (Behnke, 1970; White, 1974b). During shape change the
cytoplasmic granules are concentrated in the centre of the cell within a ’web-like' mesh of
microtubules (White, 1968b). This mesh of microtubules seems to be formed by the
translocation of the marginal bundle of microtubules from periphery of the cell towards
the centre (White, 1968b). Shape change is also accompanied by marked changes in the
arrangement of actin in the platelet cytoskeleton which, in activated platelets, forms long
filaments extending from the centre of the cell into the pseudopodia whilst forming a
mesh throughout the rest of the cell (Hartwig, 1992; Nachmias, 1980; White, 1984).
There is also a decrease in the amount of G-actin within the platelet during shape change
(Spangenberg et ah, 1992) suggesting an increase in the total amount of filamentous actin
within the activated platelet as well as the change in its organization. Indeed, it has been
suggested that actin polymerization may be the driving force for pseudopodium extension
during shape change (Cohen et ah, 1978).
23
Following shape change, if the concentration of ADP is sufficient, the platelets begin to
aggregate. Pseudopodium formation during shape change seems to be important in this
process (possibly by increasing the likelihood of contact between platelets) as shape
changed platelets have been reported to be preferentially included into platelet aggregates
(Milton & Frojmovic, 1984). Also the presence of low concentrations of cytochalasin B,
which inhibits platelet shape change, reduces the rate of primary aggregation induced by
ADP in citrated plasma (White, 1971) supporting this suggestion. This aggregation is
reversible and dependent upon the presence of fibrinogen and at least micromolar levels
of extracellular calcium (Solum & Stormorken, 1965; Zucker & Zaccardi, 1964).
Exposure of platelets to ADP results in the exposure of fibrinogen binding sites on the
platelet surface (Bennett & Vilaire, 1979; Marguerie et a l, 1979; Mustard et al., 1978;
Peerschke et al., 1980). These binding sites appear to be located in the glycoprotein
nb-ffla (Gp Ilb-EIa) complex since the platelets of patients with the inherited bleeding
disorder Glanzmann’s thrombasthenia lack glycoproteins lib and Ilia (Phillips & Agin,
1977) and fail to aggregate or to express fibrinogen binding sites in response to ADP
(Bennett Sc Vilaire, 1979; Lee et al., 1981; Mustard et al., 1979) or indeed to a number of
other agonists (George Sc Nurden, 1987; Hardistry Sc Caen, 1987; Siess, 1989). A
Ca2+-dependent fibrinogen binding site has also been demonstrated in the complex
formed from isolated glycoproteins Hb and Ilia (Nachman Sc Leung, 1982). Fibrinogen is
a symmetrical molecule and is thought to mediate aggregation by forming cross-bridges
between adjacent platelets via these binding sites. Fibrinogen can be immunolocalized
between platelets in sections of ADP-induced platelet aggregates (Hensler et al., 1992;
Suzuki et al., 1988). Also a recent study has shown that fibrinogen exhibits a bell-shaped
dose-response relationship for aggregation with the peak of the aggregation curve
24
occurring around the KD of fibrinogen for the fibrinogen receptor determined in parallel
binding studies (Landolfi et a l, 1991). This suggests that fibrinogen supports optimal
aggregation at a concentration which would provide one fibrinogen molecule for every
two fibrinogen receptors, which is consistent with its suggested 'bridge-forming’ role.
In the presence of physiological concentrations of extracellular calcium (~1 mM)
platelets undergo only a single, reversible phase of aggregation which is usually referred
to as primary aggregation. However, under conditions where the extracellular calcium
concentration is artificially reduced, for example in plasma anticoagulated with citrate,
ADP causes the platelets to undergo two phases of aggregation. Primary aggregation
occurs as described above, although this response appears to be weaker than that induced
at more physiological calcium concentrations (Packham et al., 1989). Primary
aggregation is then followed by a second and irreversible phase of aggregation
(MacMillan, 1966). This secondary aggregation is associated with the production of
prostaglandin endoperoxides and TXA2 and the release of platelet dense and a-granule
contents. The irreversibility of secondary aggregation is thought to be due to
thrombospondin released from the platelet a-granules binding to and stabilising the
inter-platelet fibrinogen bridges (Lawler, 1986). Secondary aggregation and the release of
granule contents are inhibited by cyclo-oxygenase inhibitors (Charo et a l, 1977; Mustard
et a l, 1975) suggesting that thromboxane production occurs prior to release and that it is
the thromboxane acting at its own cell surface receptor which causes release and thus
renders aggregation irreversible. Unstirred platelets or platelets whose fibrinogen
receptors have been blocked with a monoclonal antibody to Gp Hb-IIIa show severely
impaired thromboxane production and do not release their granule contents (Balduini et
25
a l , 1988). However, ADP-stimulated platelets do release their granule contents when
they are brought into close contact by centrifugation (Massini & Liischer, 1971)
suggesting that it is contact between platelets which acts as the stimulus for thromboxane
production under conditions of reduced extracellular calcium concentration. It is not yet
clear how platelet-platelet contact causes arachidonate release in the presence of reduced
extracellular calcium. It should be noted that in the effective absence of calcium, e.g. in
plasma anticoagulated with EDTA, platelets do not aggregate as micromolar
concentrations of calcium are required for the association of glycoproteins lib and Ilia
(Peerschke, 1985). Thus in the absence of calcium the fibrinogen receptor cannot form as
the Gp Ilb-IIIa complex cannot form. Platelets do change shape under these conditions,
however, suggesting that extracellular divalent cations are not essential for the binding of
ADP to its receptor or for its agonist action.
1.4.2 Biochemical Responses
Biochemically, platelet activation induced by ADP can be correlated with three events: an
increase in the cytosolic calcium concentration ([Ca2+];), inhibition of stimulated
adenylate cyclase activity and the stimulation of Na+/H+-exchange. Na+/H+-exchange does
not seem to be an important event in the induction of primary aggregation as primary
aggregation to ADP is not altered in Na+-free medium (where Na+ was replaced
iso-osmotically with N-methyl-D-glucosamine) or in the presence of concentrations of the
drug amiloride (or its analogues) which are selective for the inhibition of
Na+/H+-exchange (Connolly & Limbird, 1983; Funder et a l , 1988; Sweatt et a l , 1985).
Na+/H+-exchange is important for secondary aggregation, however, as this phase of
ADP-induced aggregation is absent under the above conditions. These and other studies
26
suggest that Na+/H+-exchange is necessary for the activation of phospholipase A2 and the
subsequent release of arachidonate metabolites which are required for platelet secondary
aggregation (Funder et al., 1988; Siffert et al., 1990; Sweatt et al., 1985; Sweatt et al.,
1986a,b).
Agents which raise platelet cAMP levels (e.g. prostaglandins \ and EL and adenosine) are
well known as inhibitors of platelet function (see for example Hourani & Cusack, 1991).
ADP inhibits this stimulated adenylate cyclase activity (Cole et al., 1971; Haslam, 1973)
presumably via a G; dependent mechanism (Aktories & Jakobs, 1985). This inhibitoiy
activity of ADP is, however, not sufficient in itself to explain its ability to induce platelet
aggregation since agents which selectively inhibit adenylate cyclase, such as
2',5’-dideoxyadenosine or 9-(tetrahydro-2-furyl)-adenine (SQ22536), do not induce
platelet aggregation and do not potentiate aggregation induced by aggregating agents
when adenylate cyclase activity is not stimulated above basal (Haslam et al., 1978). Also
some agonists, such as vasopressin, cause aggregation without inhibiting adenylate
cyclase activity, although this does make its effects far more susceptible to inhibition by
PGEj than those of ADP (Haslam et al., 1978). This ability of ADP to inhibit stimulated
adenylate cyclase activity will, however, act to reduce the effects of inhibitory platelet
agonists which may be present in the physiological situation and may play a role in the
balancing of the response of platelets to the pro- and anti-aggregatory signals to which
they are exposed within the circulation.
The response of platelets which is thought to initiate aggregation directly is an increase in
platelet [Ca2+];. It remains unclear, however, how this increase in [Ca2+]j is elicited. ThereV
is good evidence that ADP can induce both influx of calcium across the plasma
27
membrane and release of calcium into the cytoplasm from intracellular stores and this
will be discussed later in this section. However, only a few groups have looked at the
effect of ADP on human platelet inositol lipid metabolism or inositol trisphosphate
formation, which would be the prime candidate for the second messenger which
mobilized calcium from intracellular stores, and there is some confusion as to its effects.
Daniel et a l (1986) showed a significant (~2 fold) increase in the level of inositol
1,4,5-trisphosphate in 32P-labelled, aspirinated, fura-2-loaded human platelets upon
stimulation with ADP. Also Heemskerk et al (1993) have shown a transient increase in
the level of inositol trisphosphate induced by ADP in washed aspirinated-platelets
resuspended in a buffer containing lithium, apyrase and a physiological concentration of
extracellular calcium. However, a number of other authors detect no such changes. Fisher
et al (1985) showed no significant changes in the levels of phosphoinositides,
phosphatidic acid or inositol phosphates in [3H]-inositol labelled, aspirinated platelets and
Sweatt et a l (1986a) showed that indomethacin significantly reduced (>90%) inositol
trisphosphate accumulation in platelets stimulated with ADP, although it did not
completely abolish this effect. Also MacIntyre et al (1985) showed that whilst ADP was
able to induce marked increases in platelet intracellular calcium concentration as
measured in quin-2-loaded platelets, it did not induce a measurable increase in the levels
of [32P]-phosphatidic acid in [32P]-phosphate-loaded platelets. This was in contrast to
other calcium mobilizing agonists tested in this study. In a recent extensive study Vickers
et a l (1990) showed that, in the presence of fibrinogen and a physiological concentration
of extracellular calcium, ADP did not cause a significant increase in the level of inositol
trisphosphate present in washed platelets. However, these authors did observe a slight but
significant decrease in the amount of phosphatidylinositol mono- and bis-phosphates
28
under the same conditions. Raha et al (1993) reported that ADP was able to induce
transient increases in inositol trisphosphate levels during the first 4 sec of ADP
stimulation using quenched flow techniques. However, shear forces in the experimental
system were able to induce increases in inositol trisphosphate of similar magnitude and
duration in control platelets making the significance of these results difficult to interpret.
Using a more indirect approach Packham et al (1993) have suggested that phospholipase
C does not play an important role in aggregation induced by ADP. These authors showed
that in comparison with thrombin, an agonist which is known to activate phospholipase
C, ADP-induced aggregation was insensitive to the effects of inhibition of diacylglycerol
kinase (which should potentiate aggregation by prolonging the effect of endogenously
released diacylglycerol) or to inhibition of protein kinase C by staurosporine both of
which suggest that this arm of the PI pathway is only poorly activated by ADP. Thus
activation of phospholipase C by ADP in human platelets seems to be a rather variable
effect and would seem to play a relatively minor role in the response to this agonist.
In contrast to the confusion over the effects of ADP on platelet phosphoinositide
metabolism, calcium mobilization induced by ADP presents a more coherent picture,
although again this phenomenon shows some interesting differences from calcium
mobilization induced by other aggregating agents. It has been known for some time that
substances, including ADP, which can cause an increase in platelet [Ca2+]i do so in part
by inducing influx of calcium across the plasma membrane from the extracellular
medium. For example, when calcium mobilization is measured in the presence of
divalent cation chelators (EGTA) to remove extracellular calcium, the increase in [Ca2+];
is reduced compared to that in the presence of extracellular calcium (Hallam & Rink,
1985a). Of course, the fact that ADP is still able to cause an increase in [Ca2+]i in the
29
absence of extracellular calcium does suggest that it is also able to mobilize calcium from
the platelet's intracellular stores. Also, if quin-2-loaded platelets are stimulated in the
presence of extracellular Mn2+, which quenches the fluorescence of calcium indicators
such as quin-2 (Hesketh et a l, 1983), the intracellular dye signal shows rapid quenching
suggesting the opening of divalent cation channels in the plasma membrane (Hallam &
Rink, 1985b). However, when the sub-second kinetics of calcium mobilization induced
by ADP or other aggregating agents were studied using stopped flow techniques (Sage &
Rink, 1986; 1987; Sage et al., 1989) ADP was shown to cause its increase in [Ca2+]j much
more rapidly than other agonists. In these studies maximally active concentrations of
ADP were able to induce an increase in [Ca2+]; without a measurable delay (<10 ms) in
the presence of extracellular calcium whilst with other agonists this response was always
delayed by at least 200 ms. When these authors measured the kinetics of calcium
mobilization in the absence of extracellular calcium or in the presence of Ni2+ (which
blocks stimulated divalent cation influx across the plasma membrane, Hallam & Rink,
1985b) the response to ADP was also delayed by -200 ms, showing that extracellular
calcium is required for this rapid response to ADP. Interestingly, whilst the release of
calcium from intracellular stores by ADP (i.e. that component of the increase which was
present in the absence of extracellular calcium) was blocked by increased levels of
cAMP, the influx of calcium across the plasma membrane was not. The increase in
[Ca2+]; induced by thrombin was inhibited by cAMP in a similar manner in both the
presence and absence of extracellular calcium. In the absence of extracellular calcium the
effect of cAMP on ADP-induced and thrombin-induced calcium mobilization was very
similar suggesting that this aspect of ADP-induced calcium mobilization may occur by a
similar mechanism to that of aggregating agents known to stimulate phospholipase C.
30
More recently Sage et al (1990) have shown that at 17°C calcium mobilization by ADP
can be resolved into two distinct phases: a rapid phase which is dependent on the
presence of extracellular calcium and which occurs without measurable delay, even at
this reduced temperature, and a second phase delayed by -1.4 seconds at 17°C (and by
about 200 ms at 37°C). From the rapid onset of the response and its insensitivity to
changes in temperature it was suggested that the initial phase of calcium influx was due
to the opening of a plasma membrane ion-channel closely coupled to the ADP receptor.
Mahaut-Smith et al (1990; 1992) have now demonstrated the presence of ADP-gated
cation channels in the platelet plasma membrane using patch clamping techniques both
on isolated membrane patches and in whole cells. The delaying of the second phase of the
transient, however, is consistent with the need to produce some form of diffusible second
messenger. Interestingly, the release of calcium from intracellular stores (delayed phase)
occurred with a shorter delay in the presence of extracellular calcium than in its absence
(an observation which is also true of thrombin, Sage et a l , 1989) suggesting that
extracellular calcium may by some mechanism modulate the release from intracellular
stores. There is now some evidence that store-regulated calcium influx may involve
tyrosine phosphorylation, as ADP (and thrombin) evoked calcium influx, but not
mobilization of calcium from intracellular stores, has recently been shown to be inhibited
by the tyrosine kinase inhibitors genistein and methyl-2,5-hydroxycinnamate, but not by
an inactive analogue of genistein (Sargeant et a l , 1993a, b).
It should also be pointed out that in the above studies at 17°C, ADP also induced a
biphasic influx of Mn2+ across the plasma membrane one phase of which occurred
without measurable delay and is presumably due to the activation of the ADP-gated
cation channel. The other, however, had similar kinetics to the phase of calcium
31
mobilization associated with store release and was blocked by increased cellular levels of
cAMP. This was suggested to represent a mechanism by which intracellular calcium
stores are refilled from the extracellular medium without the ion entering the cytosol and
which may be controlled by the state of filling of the intracellular stores. This entry
process appears to involve cytochrome-P450 as it is blocked by a number of agents which
are said to inhibit microsomal cytochrome-P450 activity selectively, e.g. the imidazole
antimycotic econazole, (Alonso et a l, 1991; Sargeant et a l, 1992).
Fluorescence microscopic studies on single fura-2-loaded platelets have revealed that
ADP (and, indeed, thrombin) induced oscillating increases in intracellular calcium
concentration (Heemskerk et a l , 1992, 1993; Nishio et al., 1992) in common with a
number of other systems (e.g. Wakui et a l , 1990) and it has been suggested that these
oscillations are mediated by a calcium-induced calcium release mechanism (Heemskerk
et a l, 1993). However, platelets do not appear to contain the caffeine sensitive calcium
storage pool which is thought to be involved in calcium oscillation and calcium-induced
calcium release in other cell types (Foskett et a l, 1991a, b; Heemskerk et a l, 1993;
Wakui et a l, 1990) and it has been suggested that a thapsigargin (a Ca2+-ATPase
inhibitor) sensitive calcium store may assume this function in its stead. (Heemskerk et a l,
1993). The current model concerning the mechanisms by which ADP induces an increase
in [Ca2+]j are summarized in figure 1.2.
Associated with the activation of platelets by physiological agonists is the
phosphorylation of two proteins, one of 20 kDa and one of 47 kDa (Siess, 1989). The 20
kDa protein appears to be identical with the 20 kDa light chain of myosin (Daniel et a l,
1981). The phosphorylation of myosin light chain is performed by a platelet
32
ADP
Nil+ Ca2+
" plasma membrane
econazole
secondmessenger
d p 3?) \
cAMP
calciumstore ATP
ADP + Pi
ATPThaps
calciumsensitive
store
ADP + R
Thaps
Figure 1.2 A summary of the major processes thought to be associated with calcium mobilization induced by the ADP receptor. Abbreviations: R - the ADP receptor; AC - adenylate cyclase; Thaps - thapsigargin.
33
Ca2Vcalmodulin-dependent myosin light chain kinase (Dabrowska & Hartshome, 1978;
Hathaway & Adelstein, 1979) and is the major phosphorylation event associated with an
increase in [Ca2+];. Phosphorylation of myosin light chain causes myosin to associate with
cytoskeletal structures via an association with actin filaments (Fox & Phillips, 1982) and
stimulates ATP-dependent contractile activity in platelet actomyosin complexes
(Lebowitz & Cooke, 1978). Myosin phosphorylation has been suggested to be involved in
the initiation (but not maintenance) of platelet shape change as ADP has been shown to
induce the transient phosphorylation of myosin light chain with a similar time course to
shape change (Daniel et al., 1984). This phosphorylation was not inhibited by
indomethacin indicating that it was not due to released thromboxane but was inhibited by
ATP, a competitive ADP antagonist (see later) indicating that the effect was directly
mediated by the ADP receptor. Haslam & Lynham (1977) reported that ADP did not
induce myosin phosphorylation, however, closer inspection of this paper suggests that
this may be because the phosphorylation was measured at a time when it was later shown
by Daniel et al. (1984) to have returned to control levels. The platelet 47 kDa protein
appears to be the major substrate for protein kinase C (Siess, 1989), although its precise
function is unknown. ADP has not consistently been reported to induce 47 kDa protein
phosphorylation (Carty et a l, 1987; Daniel et al., 1984; Packham et a l, 1993) and in the
study by Carty et al., who showed phosphorylation, the effect of ADP was sensitive to
indomethacin suggesting an effect of released prostanoids. This lack of ADP-induced 47
kDa protein phosphorylation again suggests that ADP has little or no effect on platelet
phospholipase C activity.
34
1.5 Measurement of Platelet Responses
Possibly the simplest method of measuring platelet activation is platelet aggregometiy, a
method invented by Bom (1962a,b). In this method a beam of light is passed through a
warmed, stirred sample of platelet suspension and the change in light transmission
through the sample, relative to a platelet-free control, is measured. Shape change is seen
as an initial small decrease in light transmission (increase in turbidity) accompanied by a
decrease in the level of noise on the trace (due to the more homogeneous light scattering
properties of spherical rather than discoid particles). Aggregation is then seen as an
increase in light transmittance as the effective number of light scattering particles in the
suspension decreases. This method may be used to measure aggregation) in I either
platelet-rich plasma or of platelets suspended in buffer (washed platelets) in which case
the addition of exogenous fibrinogen is usually necessary. Platelet aggregometers can also
be used to measure shape change, in this case the platelets are treated in such a way that
aggregation does not occur, e.g. by using EDTA to remove the extracellular calcium or,
for washed platelets, by performing the assays in the presence of millimolar
concentrations of calcium in the absence of added fibrinogen, although this second
method is less commonly employed. Platelet responses may also be measured more
directly by measuring either the inhibition of stimulated adenylate cyclase (in cases where
the agonist does this) or by measuring the initial calcium transient induced by the agonist.
Cyclase assays are performed by pre-labelling the platelets with radiolabelled adenine
then extracting the labelled cAMP following stimulation (e.g. Haslam & Rosson, 1975).
Relatively simple methods for the routine measurement of increases in platelet [Ca2+];
have only recently become available with the advent of cell permeant fluorescent
indicators e.g. fura-2 (Giynkiewycz et al., 1985) or its predecessor quin-2 (Tsien, 1980).
1.6 Structure-Activity Relationships of the ADP Receptor
1.6.1 Agonists
Numerous analogues of ADP have been synthesized and tested for their activity as
aggregating agents. A selection of the more interesting of these compounds is shown in
figure 1.3. The receptor shows extreme sensitivity to modifications of the ADP molecule.
Most modifications of the purine ring cause a decrease in agonist potency. Other naturally
occurring nucleoside diphosphates (GDP, IDP, CDP and UDP) are effectively inactive as
aggregating agents (Gaarder et a l , 1961). Modifications to the N1, C6, or C8 positions also
result in loss of activity (Gaarder & Laland, 1964; Stone et a l , 1976), although
8-bromo-ADP has recently been reported as an antagonist of ADP-induced aggregation
(Jefferson et al., 1988). However, substitutions at the 2-position are tolerated and even
some very large substituents such as spin labels can be added without apparent loss of
potency (Robey et al., 1979). Indeed, some 2-substituted ADP analogues, notably
2-methylthio-, 2-chloro- and 2-azido-ADP are markedly more potent than the parent
nucleotide (Cusack & Bom, 1977; Gough et a l, 1972; Maguire & Michal, 1968).
Replacement of the D-ribose with its unnatural enantiomer L-ribose results in L-ADP
which is devoid of agonist activity (Cusack et a l, 1979). Other modifications to the
ribose moiety such as removal or inversion of the ring hydroxyls (as in 2-deoxy-ADP,
3-deoxy-ADP and adenine-9-p-D-arabinofuranoside 5f-diphosphate) or periodate
cleavage of the ring to yield the corresponding linear dialdehyde results in analogues with
markedly reduced, if any, aggregating activity (Gaarder et a l, 1961; Gaarder & Laland,
1964; Pearce et al, 1978; Ragatz et a l, 1977).
Modifications of the 5-diphosphate chain also adversely effect agonist activity. Addition
36
NH; NH;
O OII II
O* O'
HO OH
ADP
soNH'
O Oii ii
■ o ' MO ' o
<"XbL^ n - ^ n* S
HO OH
2-MeS-ADP
O OII II
S I O* y y
HO OH
ADP-p-S
O oII II. p ^ p
HO OH
2-Chloro-ADPNH'
O S II II
• o ' ^ S r pO' r ° ^ X r
HO OH
Sp-ADP-a-S
o oII II
O - o *
HO OH
8-Bromo-ADP
HO OH
5-p-fluorosulphonylbenzoyladenosine
Figure 13 The structures of some ADP analogues and 5 -p-fluorosulphonylbenzoyl- adenosine.
37
or removal of phosphate to make ATP or AMP result/ in competitive antagonistsL
(Macfarlane & Mills, 1975; Packham et al., 1972). However, AMP is a very weak
competitive antagonist and its effects in vitro are thought primarily to be due to its
degradation to adenosine (Packham et al., 1972) which will then inhibit platelet
aggregation via its own receptor (see section 1.2). Replacement, of a bridging oxygen with
methylene (to give adenosine 5?-(a,P-methylene)diphosphonate (AMPCP) and
homo-ADP) results in analogues with only weak, if any, aggregating activity (Horak &
Barton, 1974; Cusack et al., 1985) indeed AMPCP has been reported to be an apparently
non-competitive antagonist of ADP-induced aggregation (Cusack & Pettey, 1988). This
effect does not appear simply to be due to a reduction in polarity of these analogues
relative to ADP, as suggested by Horak & Barton, as analogues of AMPCP which are
isopolar with ADP also show reduced agonist activity. a,p-imido-ADP (AMPNP) is
reported to be a partial agonist at the ADP receptor, whilst a,p-difluoromethylene-ADP
and a,p-dichloromethylene-ADP were both reported to be inhibitors of ADP-induced
aggregation (Cusack & Pettey, 1988), although the effect of the difluoro- analogue
appeared to be non-competitive. This may imply an additional role of steric effects, as the
geometry of the P-O-P bond in ADP will be somewhat different from that of the P-N-P or
P-C-P bonds in these analogues, by analogy with similar ATP analogues (Yount et a l,
1971). Replacement of ionized oxygen with ionized sulphur on either the inner or
terminal phosphate groups, to give the analogues adenosine 5-0-(l-thiodiphosphate)
(ADPaS) and adenosine 5'-0-(2-thiodiphosphate) (ADPpS), results in partial agonists of
aggregation with intrinsic activities of about 0.75 (Cusack & Hourani, 1981a,b). ADPaS
occurs as a pair of diastereoisomers, the receptor showing slight (~5 fold) selectivity for
the Sp-isomer (Cusack & Hourani, 1981b). Replacement of a terminal hydroxyl with
fluorine results in an analogue with only weak aggregating activity (Cusack et al., 1985).
Far fewer analogues have been tested for their ability to inhibit stimulated adenylate
cyclase activity. ADP and 2-chloro-ADP are approximately equipotent at inhibiting
adenylate cyclase and inducing aggregation (Cusack & Hourani, 1982c). However,
2-methylthio-ADP and 2-azido-ADP are far more potent as inhibitors of adenylate cyclase
than they are as aggregating agents (Cusack & Hourani, 1982c; Macfarlane et a l , 1982;
1983), whereas the phosphorothioate analogues of ADP show greater efficacy in inducing
aggregation than they do in inhibiting adenylate cyclase. ADPPS is a partial agonist of the
effect of ADP on adenylate cyclase with an intrinsic activity of 0.5 (Cusack & Hourani,
1981a). However, although ADPaS is a partial agonist of aggregation, it acts as a pure
competitive antagonist of the inhibition by ADP of stimulated adenylate cyclase. Again
the ADP receptor showed ~5 fold stereoselectivity for the Sp-isomer (Cusack & Hourani,
1981b). As with aggregation AMPCP has no agonist activity for the inhibition of
adenylate cyclase (Cusack & Pettey, 1988). The dihalomethylenephosphonate analogues
have effects on this response which are similar to their effects on aggregation (Cusack &
Pettey, 1988).
1.6.2 Antagonists
A number of analogues of AMP and ATP have been tested as inhibitors of ADP-induced
aggregation although in many cases competition has been assumed rather than being
more rigorously established and the antagonist potency has not been fully evaluated
(Gough et al., 1969; Gough et al., 1978; Kikugawa et al., 1973; Michal et al., 1969;
Pearce et al., 1978). Several ADP analogues have now been reported as antagonists of
39
aggregation as mentioned above, i.e. 8-bromo-ADP and AMPCP and its difluoro- and
dichloro- analogues. A number of oc,o>diadenosine polyphosphates and more recently a
number of methylenephosphonate and phosphorothioate analogues of P1,P4-diadenosine
tetraphosphate (Ap4A) have also been reported to competitively inhibit ADP-induced
aggregation (Harrison et al., 1975; Zamecnik et al., 1992). Of the antagonists whose
competitive behaviour has been established, ATP has a reported Kj value of -20 juM
against aggregation (Macfarlane & Mills, 1975). The most potent antagonist so far
described is the SP-isomer of adenosine 5'-0-(l-thiotriphosphate) (ATPaS) which has a
reported KB value of 4 pM against aggregation and 4.6 j l i M against inhibition of adenylate
cyclase (Cusack & Hourani, 1982b). Those compounds which have also been tested as
antagonists of the cyclase effects of ADP have K; values which are similar to those
obtained against aggregation (Cusack & Hourani, 1982b). Indeed, from this study,
although only a limited number of ATP analogues was used, it would appear that the
structural requirements of antagonists are largely the same as those of agonists, i.e.
2-substituted analogues are more potent than the parent nucleotide whilst the methylene
phosphonate analogues are less so. The phosphorothioate analogues form an exception to
this as the ADP analogues are weaker aggregating agents than ADP whilst ATPaS is a
rather more potent antagonist than ATP. Another group of antagonists which should be
mentioned here are the 2-alkylthio- analogues of ATP and AMP. These molecules are
selective but non-competitive inhibitors of ADP-induced aggregation (Cusack & Hourani,
1982a). Furthermore, these analogues could not completely abolish the effect of a
maximally effective concentration of ADP (for aggregation) causing a maximum of only
-50% inhibition (Cusack & Hourani, 1982a). Interestingly, 2-methylthio-adenylyl
5'-(p,y-methylene)diphosphonate (MeSAMPPCP) whilst only inhibiting maximal
aggregation by -50% was able to completely inhibit the effect of ADP on adenylate
cyclase (Cusack et al., 1985; Hourani et al., 1986). Schild analysis of this effect of
MeSAMPPCP gave a slope of 1.12 which suggests that this effect was competitive. This
suggests that the inhibition of adenylate cyclase by ADP, or perhaps an effect associated
with activation of the same G-protein (e.g. production of a calcium mobilising second
messenger) may be important for the full effect of ADP on aggregation. In support of this
is the observation that whilst MeSAMPPCP inhibited aggregation induced by ADPpS, a
partial agonist of adenylate cyclase inhibition, in a manner similar to its inhibition of
ADP-induced aggregation, it was not able to inhibit aggregation induced by ADPaS, an
analogue which does not inhibit adenylate cyclase but which antagonizes this effect
(Cusack et al., 1985; Hourani et al., 1986). The structures of a selection of ATP
analogues is shown in figure 1.4.
1.6.3 Multiple ADP Receptors?
The lack of correlation between the abilities of the 2-substituted and phosphorothioate
analogues of ADP to inhibit adenylate cyclase and to induce aggregation has led to the
suggestion that these effects are mediated by two different receptors (Colman et a l,
1980; Macfarlane et al., 1982; 1983). Support for this suggestion has been claimed from
observations with the affinity reagent 5'-p-fluorosulphonylbenzoyladenosine (FSBA) (the
structure of which is shown in figure 1.3) which inhibits ADP-induced shape change and
aggregation but does not affect the inhibition by ADP of adenylate cyclase. [3H]-FSBA
labels a 100 kDa protein in the platelet plasma membrane which has been called
’aggregin’ and proposed to be the ADP receptor which mediates aggregation (Colman,
1992). However, the incorporation of 3H from labelled FSBA into platelet membranes
41
O 0 0D o "
ATP
NH,
? 9 o
HO b Hot< * to 'y r
ho' J dh
NH,
" i6N ^ c i
2-Chloro-ATPNH2
- ° " o
N '
HO OH
2-Methylthio-ATP
O O sp £ «•p .p._,p. 0 <N
0'V '0'£'0tt>0 — 'T V
HO OH
Sp-ATP-a-S
NH,
,N-*s n ^
f? 0 0.n J a 110 I
o-
NH,
NH,
HO OH
AMPPCP
o s oII II II \
o- o- o- V iHO OH
9 P OIIu »■O'f'-O'f
NH
*>
0* 0*
N ,^ J l Jov
N
• N
HO OH
AMPCPP
o oS 11
Sp-ATP-fl-s
0 - P = o
0 ' | | n0 " | n- 0 ' ^ N / \ ^ N
Ap5A
Figure X.4 The structures of some ATP analogues.
42
is only inhibited at very high concentrations of ADP and ATP (10 mM) and is also
blocked by similar concentrations of adenosine, 2-chloroadenosine, GDP and UDP
(Bennett et al., 1978) none of which are ADP receptor ligands. The incorporation of
FSBA into platelet membrane proteins is not therefore consistent with the known
pharmacology of the ADP receptor. Also a high concentration (10 mM) of FSBA does not
inhibit the binding of [32P]-MeSADP, a potent ADP agonist, to platelets (Mills et al.,
1985). Due to its relatively high potency as an inhibitor of adenylate cyclase compared to
its ability to induce aggregation P2P]-MeSADP has been claimed to be a selective ligand
for the ADP receptor which mediates inhibition of adenylate cyclase (Macfarlane et al.,
1983). However, since MeSADP is also a potent aggregating agent and binding studies
using this ligand showed no evidence for two binding sites (Scatchard plots being linear)
(Macfarlane et al., 1983) it must be assumed that MeSADP induces both of its effects via
a single class of receptor, which is not blocked by FSBA. It has also been shown that
FSBA does not inhibit the mobilization of calcium induced by ADP (Rao & Kowalski,
1987), a response which is thought to be the direct trigger of aggregation. The
observations that FSBA cannot inhibit either of the second messenger effects induced by
ADP, that it does not inhibit the binding to platelets of a potent aggregating agent and that
the incorporation of FSBA can be blocked by compounds which do not interact with the
ADP receptor (e.g. adenosine) would suggest that FSBA does not, in fact, interact with
the ADP receptor and that it may exert its inhibitory effects by interacting with a site
further along the pathway which leads to aggregation. Interesting, in the light of the above
suggestion, is the observation that FSBA can also inhibit shape change and aggregation
induced by a number of other aggregating agents, e.g. TXA^ collagen, thrombin and
adrenaline (for a review see Colman, 1990). This was taken by these authors to imply a
43
central role for ADP in the effects of other aggregating agents. However, primary
aggregation induced by adrenaline, for example, occurs prior to the release of ADP
(Charo et al., 1977) and is not blocked by ATP (a selective ADP receptor antagonist)
(Macfarlane & Mills, 1975), strongly suggesting that this response is independent of
ADP. Thus the ability of FSBA to inhibit responses to these agents can equally well be
explained by a non-selective inhibition by FSBA of a common event which occurs in
response to all of these agonists. Interestingly, the molecular weight of aggregin (the
protein labelled by [3H]-FSBA), 100 kDa, is the same as that of glycoprotein Hla which
forms part of the agonist-inducible fibrinogen receptor which is essential for aggregation
in response to most aggregating agents. The ability of FSBA to modify glycoprotein Ilia
would, therefore, explain quite well its non-specific ability to inhibit platelet aggregation.
However, Colman et al (1988) have shown quite convincingly that aggregin is a distinct
polypeptide from Gp Ilia making this suggestion untenable but still not confirming
aggregin as an ADP receptor.
Another possible explanation for these differences in potency is that ADP acts at a single
class of receptor which is coupled via two different effector mechanims (Cusack &
Hourani, 1982b). Different regions of the ADP molecule may be involved in activating
these different effector mechanisms and structural analogues of ADP would not, then,
necessarily be expected to exhibit the same potency in inducing these effects. Support for
this second hypothesis comes from a detailed study of the effects of a number of ADP
receptor antagonists (Cusack & Hourani, 1982b). All of these antagonists were shown to
be competitive by Schild analysis and there was a good correlation between the pAs
values obtained for these antagonists as inhibitors of aggregation and as inhibitors of the
effects of ADP on adenylate cyclase. Also binding studies using the agonists 2-azido-ADP
44
and MeSADP revealed only a single class of binding site (Macfarlane et al., 1982; 1983).
As both of these compounds are potent aggregating agents as well as inhibitors of
adenylate cyclase this must suggest that the effects of ADP are mediated by only a single
class of receptor. In a recent study using a novel photoaffinity reagent
2-azidophenylethylthio-ADP, which was shown to be both an aggregating agent and an
inhibitor of stimulated adenylate cyclase activity, only a single 43 kDa protein was
labelled (Cristalli & Mills, 1993). In contrast, Jefferson et al. (1988) have shown two
binding sites for [3H]-ADP on formaldehyde fixed platelets. However, the process of
formaldehyde fixation causes extensive cross linking of proteins throughout the cell,
including cell membrane proteins such as receptors. Indeed, these authors state that the
fixation was performed in order to inactivate ADP metabolising enzymes which are a
complicating factor in binding studies with adenine nucleotides and their analogues in
unfixed platelets. If fixation is able to denature the ADP binding sites of enzymes such
that they no longer possess catalytic activity, it is not unreasonable to assume that the
ADP receptor binding site may be similarly denatured. This, therefore, casts doubt as to
whether either of the binding sites detected in this study would represent a functional
ADP receptor in native platelets. Also, the structure-activity relationships of these two
binding sites were not consistent with those obtained in functional studies further
complicating the acceptance of either of these sites as an ADP receptor. Two recent
reports suggest the presence of an adenine nucleotide binding site on the Gp Ilb-IIIa
complex. Greco et al. (1991) used [35S]-ATPaS as a photoaffinity reagent (activating the
purine ring by exciting it with light at 254 nm) and labelled a single 120 kDa protein
which they showed to be identical to the a-chain of Gp lib using several techniques.
Mayinger & Gawaz (1992) used the photoaffinity analogue 8-azido-ATP labelled with 32P
45
on the y-phosphate and labelled two membrane glycoproteins which they showed to be
glycoproteins lib and Ilia by Western blotting with specific monoclonal antibodies. This
labelling was blocked in the presence of 1 mM ATP. However, this binding site is
unlikely to represent an ADP receptor as Greco et al. (1992) subsequently showed that
this binding site could also be photo-labelled by 35S labelled guanosine
5'-0-(l-thiotriphosphate) (GTPaS) and uridine 5-0-(l-thiotriphosphate) (UTPaS) which
were both shown in this study to be only very weak inhibitors of responses to ADP. Also,
platelets from thrombasthenic patients are still able to respond to ADP but lack these two
glycoproteins.
Thus, the weight of current evidence would seem to support the hypothesis that only a
single class of ADP receptor is responsible for the various responses of platelets to ADP,
the observed lack of correlation between the potencies of some compounds as agonists of
these responses simply being a reflection of differing efficacy of these compounds in
activating the signal transducing mechanisms associated with this ADP receptor. This
ADP receptor would appear to be distinct from the Gp Ilb-IIIa complex and from
aggregin the 100 kDa protein labelled by FSBA but may be identical with the 43 kDa
protein which is photoaffinity labelled by 2-azidophenylethylthio-ADP.
1.7 Aim of this Thesis
The aim of this thesis was to further characterize the pharmacology of the human platelet
ADP receptor, both to address the question of whether there are one or two receptors on
human platelets mediating the response to ADP and to compare the ADP receptor with
other P2-purinoceptors subtypes.
46
CHAPTER 2
STUDIES ON INCREASES IN
CYTOSOLIC CALCIUM
CONCENTRATION
47
2.1 Introduction
The second messenger response of platelets which is thought to be most important in the
induction of shape change and aggregation by ADP is an increase in cytosolic calcium
concentration (see section 1.4 of the General Introduction). The kinetics of this increase
have been extensively studied, and there is good evidence that the increase in intracellular
calcium concentration induced by ADP involves both the influx of extracellular calcium
via an ADP-gated cation-channel and the release of calcium from intracellular stores
(Hallam & Rink, 1985a, b; Mahaut-Smith et a l, 1990; 1992; Sage & Rink, 1986; 1987;
Sage et a l, 1989; 1990). The kinetics of mobilization of calcium from intracellular stores
are consistent with the production of a second messenger, however the identity of this
messenger has yet to be unequivocally determined (see section 1.4.2). Influx of
extracellular calcium is sensitive to inhibition by tyrosine kinase inhibitors suggesting
that the ADP-gated ion-channel or store regulated influx may be activated by tyrosine
phosphorylation (Sargeant et a l , 1993b). The phosphorylation of a 130 kDa protein was
consistently inhibited by the tyrosine kinase inhibitors used in this study and may
therefore represent this channel or a subunit thereof. However, apart from a brief
statement by Drummond and MacIntyre (1987) that ATP inhibits ADP-induced increases
in [Ca2+]j (which is consistent with the known pharmacology of the ADP receptor), very
little is known about the structure-activity relationship of the ADP receptor which
induces this response. Thus the effects of ADP, ATP and a number of adenine nucleotide
analogues whose effects on ADP-induced aggregation and inhibition adenylate cyclase
have previously been characterised were determined. A brief summary of the known
effects of the compounds used in the present study is given below (for further details see
section 1.6 of the General Introduction).
MeSADP is a potent aggregating agent and inhibitor of adenylate cyclase being about 5
fold more potent than ADP as an aggregating agent (Macfarlane et al., 1983) and about
200 fold more potent than ADP as an inhibitor of adenylate cyclase (Macfarlane et al,
1983). Sp-ADPocS and ADPpS are both partial agonists of aggregation having intrinsic
activities of 0.75 (Cusack & Hourani, 1981a, b). ADPpS is also a partial agonist of
adenylate cyclase inhibition, although in this case its intrinsic activity is only 0.5, whilst
ADPaS acts as a pure antagonist of this response. Of the known competitive antagonists
of ADP-induced aggregation and inhibition of adenylate cyclase (Cusack & Hourani,
1982b) Sp-ATPaS, Cl-ATP, AMPPCP and Ap5A were tested for their ability to inhibit
ADP-induced increases in [Ca2+]j. The effects of the compounds MeSATP, AMPCPP and
UTP were also determined. MeSATP has been shown to be a selective but
non-competitive inhibitor of ADP-induced aggregation (Cusack & Hourani, 1982a),
however a close structural analogue, MeSAMPPCP, appeared to be a competitive
inhibitor of the effects of ADP on adenylate cyclase (Hourani et al., 1986; see section
1.5.2 of the General Introduction for further discussion). AMPCPP has been reported to
have no effect on ADP-induced aggregation (Lips et al., 1980), its effects on inhibition by
ADP of adenylate cyclase have not been reported. There has been increasing interest
recently in the effect of UTP in some tissues which appear to possess atypical P2Y
receptors (O'Connor et al., 1991) and it was, therefore, of interest to determine the effect
of this compound at the ADP receptor. The effects of SP-ATPpS were determined both on
calcium mobilization and ADP-induced aggregation. The effects of ATPpS on
aggregation have not previously been studied and were, therefore, of interest in further
characterising the pharmacology of this response as well as for comparing it with calcium
mobilization.
49
Some of the results presented here have been published (Hall & Hourani, 1993), and a
copy of this paper is included at the back of this thesis.
2.2 Materials and Methods
2.2.1 Materials
ADP, ATP, AMPCPP, AMPPCP, adenosine 5-phosphorothioate (AMPS), bovine serum
albumin (BSA), creatine kinase (type 1 from rabbit muscle), creatine phosphate,
diethylaminoethyl (DEAE)-sephadex A25, Pl,P5-diadenosine pentaphosphate (Ap5A),
[ethylene-bis(oxyethylenenitrilo)] tetraacetic acid (EGTA), fura-2-AM, myokinase
(chicken muscle), myosin (rabbit muscle), phosphoenolpyruvate, prostacyclin, pyruvate
kinase (type HI from rabbit muscle), tetrabutylammonium hydrogen sulphate,
tris(hydroxymethyl)aminomethane (tris), triton X-100 and UTP were obtained from
Sigma Chemicals Company. ADPPS was obtained from Boehringer Mannheim. ATPaS
was obtained from Boehringer Mannheim and purified as described in section 2.2.6 or
synthesised enzymatically as described in section 22.12. MeSATP, MeSADP and
2-chloro-ATP (Cl-ATP) were obtained from Research Biochemicals Incorporated and
purified as described in section 2.2.6. ADPaS and ATPpS were synthesised
enzymatically as described in sections 2.2.7.1 and 22 .13 , respectively. ADP, ATP,
AMPPCP, Ap5A and UTP were obtained as the sodium salt, other nucleotides were
obtained as the lithium salt. HPLC solvents were obtained from Fisons Scientific
Apparatus. Prostacyclin was dissolved at 100 pg/ml in 10 mM NaOH. Fura-2-AM was
dissolved at 5 mM in dimethylsulphoxide. Both were stored frozen at -20°C. All other
drugs were dissolved in water. Nucleotides were stored frozen in solution.
50
2.2.2 Platelet Aggregation Studies
Blood (9 volumes) from healthy human volunteers, who said they had not taken any drugs
for the previous 10 days, was drawn into 1 volume of 3.8 % tri-sodium citrate,
centrifuged at 260 g for 20 min and the supernatant platelet-rich plasma (PRP) was
removed. Platelet-poor plasma was obtained by centrifuging PRP in a microfuge for 1
min. Aliquots (300 jul) of PRP were taken and incubated at 37°C for 3 min prior to
transfer to a Chrono-Log Lumi-Aggrometer. Aggregation was quantified as the maximal
rate of increase in light transmission (in arbitraiy units per min) through the stirred
sample, relative to platelet-poor plasma, on addition of agonist. The PRP was maintained
at 37°C within the aggregometer and was stirred at 1000 rpm. Agonists were added in 10
pi of water. As adenine nucleotides are known to be rapidly degraded in plasma (Maguire
& Satchell, 1981) antagonists, where used, were added simultaneously with the agonist to
avoid possible complicating effects of pharmacologically active degradation products.
2.2.3 Loading of Platelets with Fura-2
Platelets were loaded with fura-2 by a modification of the method of Rink & Sage (1987).
Blood (6 volumes) was drawn into 1 volume of acid-citrate-dextrose anticoagulant (25 g/1
tri-sodium citrate dihydrate, 15 g/1 citric acid monohydrate, 20 g/1 glucose) and
centrifuged for 20 min at 260 g. The supernatant PRP was removed and incubated at
37°C for 45 min in the presence of 4 pM fura-2-AM. The fura-2-loaded platelets were
then collected by centrifugation at 680 g for 20 min, in the presence of 1 pM prostacyclin
(to prevent aggregation), and resuspended at 1 x 108/ml in HEPES-saline (145 mM NaCl,
5 mM KC1,1 mM MgC^, 10 mM N-2-hydroxyethylpiperazine-N-2-ethanesulphonic acid
51
(HEPES), 10 mM glucose and 2 mg/ml BSA, adjusted to pH 7.4 with 1 M NaOH).
Aliquots of platelet suspension (0.7 ml) were incubated at 37°C in siliconised glass
aggregometer tubes for 3 min in the presence of 1 mM CaCl2 and transferred to a
Perkin-Elmer LS50 luminescence spectrometer for estimation of cytosolic calcium
concentration.
2.2.4 Calcium Measurement
The fura-2-loaded platelet suspension was maintained at 37°C and stirred whilst it was
excited alternately with light at 340 ran and 380 nm. The emitted light intensity was
measured at 510 nm. The ratio, R, of the emitted light intensity upon excitation at these
two wavelengths was determined and used to calculate the cytosolic calcium
concentration. The maximal intensity ratio (Rmax) was determined following lysis of the
platelets with 0.1 % Triton X-100 and the minimum intensity ratio (R ^) was determined
by subsequent addition of 20 mM tris followed by 10 mM EGTA. These measurements
were used to calibrate the fluorescence signal by the Perkin-Elmer Intracellular
Biochemistry programme which calculates cytosolic calcium concentrations from the
sample fluorescence intensity ratios using the following formula (taken from
Grynkiewicz et a l, 1985):
[Ca2+]i = Kd ((R - Rmin) / ( R ^ - R)) SFB
where, KD is the apparent dissociation constant of the fura-2.calcium complex (= 225
nM) and SFB is the ratio of the fluorescence intensities of free fura-2 and its calcium
complex when excited at 380 nm.
52
Agonists were added in 14 pi of water and antagonists, where used, were added
simultaneously with the agonist. Responses were quantified as the maximal change in
cytosolic calcium concentration following addition of the agonist. Log
concentration-response curves were obtained at least three times with blood from
different donors and the results pooled. EC50 values were obtained by regression of the
linear parts of the pooled log concentration-response curves and dose ratios for
antagonists were calculated from the EC50 values in the presence and absence of
antagonist. Where Schild analysis was not performed the apparent KD values for
antagonists were obtained from the dose ratios using the formula KD = concentration of
antagonist / (dose ratio - 1) and pA2 values were calculated as -log10KD. Statistical
significance was determined by Student’s t-test.
2.2.5 HPLC Separation of Nucleotides
To determine the purity of the adenine nucleotides used in this study nucleotides wered
separated by ion-paired reverse-phase HPLC on a Techsphere 50DS column (15 cm x 3.9
mm id). The injection volume was 50 pi. For 2-substituted nucleotides the mobile phase
consisted of two buffers: 100 mM KHjPO^ 8 mM tetrabutylammonium hydrogen
sulphate, pH 6.0 (buffer A) and a mixture of buffer A and acetonitrile (60:40, v/v) (buffer
B). The column was eluted at a flow rate of 1.3 ml/min with the following gradient: 2.5
min at 100 % A, 2.5 min up to 20 % B, 5 min up to 40 % B, 3 min up to 100 % B. The
column was held at 100 % B for a further 7 min and then returned to 100 % A over 5 min.
The column was maintained in 100 % buffer A for a further 15 min prior to injection of
the next sample. Nucleotides were detected photometrically at the appropriate
wavelength (see section 2.2.6, Nucleotide Separation).
53
For the phosphorothioate analogues buffer A was 100 mM KEyPC , pH 7.0 containing 8
mM tetrabutylammonium hydrogen sulphate and buffer B was a mixture of buffer A and
methanol (70:30, v/v). The column was eluted at a flow rate of 1.3 ml/min with the
following gradient: 2.5 min at 100 % A, 7.5 min up to 100 % B, hold at 100 % B for 10
min. The column was returned to 100 % A over 5 min and maintained at 100 % A for a
further 10 min prior to injection of the next sample.
Unsubstituted nucleotides could be resolved using either of these HPLC systems,
although the system which used acetonitrile in buffer B was routinely used. Retention
times for the nucleotides used in this study and some of their degradation products in
these HPLC systems are shown in table 2.1.
2.2.6 Nucleotide Purification
Where necessary nucleotides were purified by ion-exchange chromatography on a 30 x
1.5 cm DEAE-sephadex column, eluting with a 2 1 linear gradient of 1 - 800 mM
triethylammonium bicarbonate, collecting 12-13 ml samples. The fractions containing the
nucleotides were identified spectrophotometrically at the maximally absorbing
wavelength. This corresponds to 273 nm for 2-methylthio-substituted adenine
nucleotides, 266 nm for 2-chloro-substituted adenine nucleotides and 259 nm for
nucleotides with unmodified adenine moieties. Fractions containing the required
nucleotides were pooled and concentrated with removal of the buffer salt as follows. The
pooled aqueous fractions were rotaiy evaporated to diyness in a 1 1 vessel at a
temperature not greater than 30°C. Under these conditions the triethylammonium
bicarbonate dissociates into its molecular components, triethylamine, carbon dioxide and
water, all of which are evaporated from the sample. Residual triethylammonium
54
Nucleotide
Retention time / min
Buffer B
40% Acetonitrile
Buffer B
30% Methanol
Adenosine 8.9 8.7
AMP 9.3 6.7
ADP 11.5 11.2
ATP 13.6 12.3
AMPS 10.9 10.5
ADPccS 13.6 12.9
ATPaS 15.8 13.6
ADPPS 13.2 12.1
ATPpS 13.5 13.0
MeSADP 15.8 n.d.
MeSATP 16.4 n.d.
Cl-ATP 15.5 n.d.
Ap5A n.d. 14.4
AMPCPP 12.6 n.d.
AMPPCP 12.6 n.d.
Table 2.1 A comparison of the retention times of various adenine nucleotide analogues in the two HPLC systems used in this study (n.d. = not determined).
bicarbonate and water were removed from the sample by twice dissolving the residue in
methanol (50 ml) and rotary evaporating. The residue was then transferred to a 25 ml
vessel in 5 ml of methanol and evaporated to dryness along with a second 5 ml methanol
wash of the original vessel. Residual methanol was removed by dissolving the sample in
water (5 ml) and evaporating to dryness. Finally the nucleotide was dissolved to a known
concentration (determined photometrically using published extinction coefficients) in
water. This method of purification yields the nucleotide as the triethylammonium salt.
2.2.7 Nucleotide Synthesis
2.2.7.1 Synthesis ofSp-ADPaS
Initially attempts were made to synthesize SP-ADPaS from AMPS using myokinase (MK)
and ATP. MK has been reported to stereospecifically phosphorylate AMPS to Sp-ADPaS
(Eckstein & Goody, 1976; Sheu & Frey, 1977). Creatine kinase (CPK) and creatine
phosphate (CrP) were included in the reaction mixture to regenerate ATP. CPK
specifically phosphorylates the Rp-isomer of ADPaS (Eckstein & Goody, 1976) and
should not, therefore, phosphorylate the desired product of the reaction, SP-ADPaS. The
reactions which it was assumed would occur were:
myokinaseAMPS + ATP-----------------» Sp-ADPaS + ADP
creatineADP + CrP-----------------> ATP + creatine
kinase
However, when the reaction was carried out the final product was not Sp-ADPaS but
SP-ATPaS. Presumably the MK was able to further phosphorylate any ADPaS which was
56
formed, i.e. Sp-ADPaS + ADP > Sp-ATPaS + AMP. Thus, this system could not be
used to synthesize SP-ADPaS, however, it was used in initial syntheses of SP-ATPaS (see
section 2.2.7.2).
Sp-ADPaS was subsequently synthesised by the enzymatic dephosphorylation of
Sp-ATPaS with myosin (Jaffe & Cohn, 1978). The reaction mixture contained: 100 mM
HEPES, pH 7.4, 30 mM KC1, 3 mM CaCl2, 3.2 mM ATPaS, 0.5 mg/ml myosin, 5 %
glycerol. The reaction was carried out at 37°C in a total volume of 10 ml and was
followed by HPLC (see section 2.2.5) taking 10 pi samples every 35-40 min. The reaction
was complete within 35 min and was terminated by addition of 1 ml of 50 %
trichloroacetic acid (TCA). The precipitated protein was removed by filtration and the
filtrate was neutralised with NaOH and applied to an ion-exchange column (see section
2.2.6). The resulting ADPaS was at least 95 % pure by HPLC.
22.12 Synthesis of SP-ATPaS
The Sp-isomer of ATPaS was initially synthesised using the coupled reactions myokinase
and CPK using ATP and CrP as the phosphoryl group donors (see section 2.2.7.1). In
subsequent syntheses pyruvate kinase (PK), which is specific for the phosphorylation of
Sp-ADPaS, and phosphoenolpyravate (PEP) were used instead of CPK and CrP. The
overall reaction in this case being:
MK/PK/ATPAMPS + 2 PEP--------------------> Sp-ATPaS + 2 pyruvate
For the synthesis of SP-ATPaS using creatine kinase the reaction mixture contained: 50
mM tris.HCl, pH 8.0, 3 mM MgCl2, 6.7 mM AMPS, 7 mM ATP, 10 mM creatine
57
phosphate, 50 units/ml myokinase and 20 units/ml creatine kinase. The reaction was
carried out at 37 °C in a total volume of 10 ml and was followed by HPLC (see section
2.2.5) taking 10 pi samples every 35-40 min. After 4 hours the reaction reached
equilibrium and was applied to an ion-exchange column to separate the nucleotides (see
section 2.2.6). The resulting ATPaS was at least 95 % pure by HPLC and was a substrate
for myokinase confirming it as the Sp-isomer (Eckstein & Goody, 1976).
For the synthesis of Sp-ATPaS using pyruvate kinase the reaction mixture contained: 50
mM tris.HCl, pH 8.0, 3 mM MgCl2, 100 mM KC1, 6.7 mM AMPS, 0.2 mM ATP, 15 mM
phosphoenolpyruvate, 50 units/ml myokinase and 20 units/ml pyruvate kinase. The
reaction was carried out at 37°C in a total volume of 10 ml and was followed by HPLC
(see section 2.2.5) taking 10 pi samples every 35-40 min. After 5 hours approximately 50
% of the AMPS had been converted to ATPaS and the reaction was terminated by the
addition of 1 ml of 50 % TCA. The precipitated protein was removed by filtration and the
filtrate neutralised with NaOH. The ATPaS was separated by ion-exchange
chromatography (see section 2.2.6). The Sp-ATPaS produced was at least 97 % pure by
HPLC.
2.2.7.3 Synthesis of Sp-ATPpS
ATPpS was synthesised by the phosphorylation of ADPpS with pyruvate kinase. This
reaction is specific for the formation of the SP-isomer (Eckstein & Goody, 1976; Jaffe &
Cohn, 1978). The reaction mixture was of the following composition: 50 mM tris.HCl,
pH 8.0, 100 mM KC1, 3 mM MgC^, 3.5 mM ADPpS, 4 mM phosphoenolpyruvate, 15
units/ml pyruvate kinase. The reaction was carried out at 37 °C in a total volume of 10 ml
58
and was followed by HPLC. After 3 hours 20 min the reaction was terminated by
applying the reaction mixture to an ion-exchange column. The resulting ATPpS was 95
% pure by HPLC. This ATPpS was heated at 60°C to inactivate contaminating enzyme
(subsequent nucleotide syntheses were terminated by addition of TCA to remove the
enzymes before they were applied to the ion-exchange columns). The precipitate was
removed by filtration. This heat treatment left the ATPpS contaminated by approximately
15 % AMP. This quantity of contaminating AMP was not sufficient at the concentrations
of ATPpS used in this study to have any significant antagonistic effect (Packham et a l,
1972). For ATPpS the concentrations given in the text refer to the concentration of
ATPpS present not the total nucleotide concentration, due to this relatively large impurity
content.
2.3 Results
A typical trace of the changes in fluorescence intensity ratio of a sample of fura-2-loaded
platelet suspension on addition of ADP is shown in figure 2.1a along with a trace
showing the determination of Rmin and R ^ . Figure 2.1b shows the corresponding graph of
the changes in cytosolic calcium concentration calculated from the fluorescence intensity
ratios.
ADP induced concentration-dependent increases in platelet cytosolic calcium
concentration, with an EC50 value of 0.8 juM. This effect was inhibited by ATP which
caused parallel rightward shifts in the ADP log concentration-response curve (figure
2.2a). Schild analysis of this inhibition by ATP resulted in a slope of 0.95 ± 0.07 (not
significantly different from unity, p>0.5) and a pA2 value of 5.01 (corresponding to a KD
59
aOC
cwGO
4>Oc<Dco<0s -OE
Tris, EGTA
6
Triton X-100
4
2
ADP
00 20 6040
Time / sec
300
G 200
100
ADP
40200
Time / sec
Figure 2.1 A typical fluorimeter trace from dual wavelength calcium measurement, (a) A fluorimeter trace showing the change in fluorescence intensity ratio of a sample of fura-2-loaded platelets on addition of ADP (100 pM) (solid line) and a calibration trace (dashed line), (b) The data shown in (a) converted to calcium concentrations.
60
+CNu
o00032aO
250
200
150
100
100 300100.1 1
ADP concentration / jiM
03
<DcoO3
W)O
2
1
04 3■5
log [ATP]
Figure 2.2 The effect of ATP on increases in cytosolic calcium concentration ([Ca2+]j) in human platelets induced by ADP. (a) Effect of ADP alone (•) or in the presence of ATP at 100 pM (■), 200 pM (A), 400 pM or 800 pM (▼). Each point is the mean of at least 3 determinations, and the vertical bars show the s.e. mean, (b) The Schild plot of the data presented in (a). For abbreviations, see text.
61
value of 9.8 pM) (figure 2.2b). Although only one concentration was tested, the
analogues Cl-ATP (30 pM), ATPaS (10 pM), ATPPS (8.5 pM), Ap5A (200 pM) and
AMPPCP (100 pM) (figure 2.3a-d and figure 2.4a) also inhibited ADP-induced increases
in [Ca2+],. in an apparently competitive manner, i.e. they caused parallel, rightward shifts
of the log concentration-response curve with no significant depression of the maximal
response. However, little inhibition was observed in the presence of AMPCPP (100 pM)
or UTP (100 pM) (figure 2.4b and c). The pA values for these antagonists, derived from
the dose-ratios, were 5.60 for Cl-ATP, 5.92 for ATPaS, 5.96 for ATPpS, 5.10 for Ap5A
and 4.27 for AMPPCP, whilst those for AMPCPP and UTP were both less than 3.82.
ATPpS also appeared to act as a competitive antagonist of ADP-induced aggregation
with an apparent affinity similar to that of ATPaS, giving a against ADP-induced
aggregation of 5.46 (figure 2.5). MeSATP (50 pM) also inhibited ADP-induced calcium
mobilization, but this inhibition was not competitive and could not be overcome by a high
concentration of ADP (300 pM) (figure 2.4d).
MeSADP also induced increases in cytosolic calcium concentrations and was about 20
times more potent than ADP, with an EC50 value of 41 nM. This effect of MeSADP was
inhibited by ATP (100 pM) in an apparently competitive manner (ATP causing parallel,
rightward shifts in the log concentration-response curve with no apparent decrease in the
maximal response), giving a pA for ATP of 4.66 (figure 2.6). ADPaS and ADPpS each
caused increases in cytosolic calcium concentration, but in each case achieved only about
60 % of the maximum effect of ADP even at high concentration (100 or 300 pM) (figure
2.7a and b). ADPaS was approximately 8 fold more potent than ADPpS at causing these
increases having an EC50 value of -0.5 pM compared to ~4 pM for ADPpS. The effects
62
a b200-1
150-
100 -
5 0 -
0-10.1 1 10 100+<N
250 n
200 -
150-
100-
5 0 -
0-J0.1 1 10 100
.9
200 n
150-
100 -
5 0 -
0-J0.1 1 10 100
8 250-1o£
200 -
150-
100 -
5 0 -
0.1 1 10 100
ADP concentration / pMFigure 2.3 Increases in platelet cytosolic calcium concentration ([Ca2+];) induced by ADP alone (•) or in the presence (■) of (a) Cl-ATP (30 pM), (b) ATPaS (10 pM), (c) ATPpS (8.5 pM) or (d) Ap5A (200 pM). Each point is the mean of at least 3 determinations, and the vertical bars show the s.e. mean. For abbreviations, see text.
63
200 n
100 -
50-
0.1 1 10 100
200-1
150-
100 -
50-
§C3 0 J+CN 0.1 1 10 100
S 250-1a
200-
150-
100-
50-
0 J
0.1 1 10 100
200-
100 -
50-
0.1 1 10 100
ADP concentration / jiM
Figure 2.4 Increases in platelet cytosolic calcium concentration ([Ca2+]j) induced by ADP alone (•) or in the presence (■) of (a) AMPPCP (100 pM), (b) AMPCPP (100 pM), (c) UTP (100 pM) or (d) MeSATP (50 pM). Each point is the mean of at least three determinations, and the vertical bars show the s.e. mean. For abbreviations, see text.
64
.a a<3Q*COa
VhC«t-l
■eosv.a o -*—> 03 toO&b£)OJ)03
«4HO£S
73
xcd
30 n
20 -
10 -
0.1 10 100 300
ADP concentration / jiM
Figure 2.5 Platelet aggregation induced by ADP alone (•) or in the presence of ATPpS (17 pM) (■) or ATPaS (20 pM) (▲). Each point is the mean of three determinations and the vertical bars show the s.e. mean. For abbreviations, see text.
65
250-
sa 200 -
•»—<i— i+CNC3y. iso-a
1 100-
5
50-
0.01 0.1 10 1001
Agonist concentration / jliM
Figure 2.6 Increases in platelet cytosolic calcium concentration ([Ca2+]s) induced by ADP (□), MeSADP (O) or MeSADP in the presence of ATP (100 pM) (•). Each point is the mean of at least three determinations, and vertical bars show the s.e. mean. For abbreviations, see text.
66
200 -
100 -
5 0 -
0-110 1000.1 1+
u.9<DcoI 150O►3
100 -
5 0 -
0-J0.1 1 10 100
250-,
200
1 50-
100 -
5 0 -
0 -J10 1000.1 1
150 n
100 -
5 0 -
0-10.1 1 10 100
Agonist concentration / jiM
Figure 2.7 Increases in platelet [Ca2+]j induced by ADP (•) or by (a) ADPaS (■) or (b) ADPpS (A). Effect of (c) ADPaS (■,□) or (d) ADPpS (A ,A) on platelet [Ca2 in the absence (closed symbols) or presence (open symbols) of ATP (100 juM). Each point is the mean of at least three determinations, and vertical bars show the s.e. mean. For abbreviations, see text.
67
of ADPaS and ADPpS were inhibited by ATP (100 juM), giving apparent pA values for
ATP of 5.55 and 5.32 respectively (figure 2.7c and d).
2.4 Discussion
These results represent the first detailed investigation into the structure-activity
relationships of the ADP receptor which mediates increases in cytosolic calcium
concentration in human platelets. As expected the log concentration-response curve to
ADP itself, giving an EC50 value of 0.8 pM, is similar to that usually found in for the
induction of aggregation, in washed platelets or in PRP, and it is also similar to that found
in previous studies on the concentration-dependence of the effect of ADP on platelet
cytosolic calcium concentration (Drummond & MacIntyre, 1987; Sage & Rink, 1987).
MeSADP was -20 fold more potent than ADP at inducing increases in platelet cytosolic
calcium concentration. This values lies between its reported potency ratios for the
induction of aggregation or the inhibition of adenylate cyclase of -5 and -200,
respectively (Macfarlane et a l , 1983) but is somewhat closer to that reported for
aggregation. The effect of MeSADP on cytosolic calcium concentration, like that on
aggregation and stimulated adenylate cyclase activity (Cusack & Hourani, 1982c), was
competitively inhibited by ATP with a pA value similar to that shown by ATP at
inhibiting the effects of ADP, indicating that MeSADP is acting via the ADP receptor.
The phosphorothioate analogues of ADP, ADPaS and ADPpS, each achieved lower
maximal increases in cytosolic calcium concentration than did ADP, with ADPaS being
more potent than ADPpS. The effects of these two analogues were also inhibited by ATP
with an appropriate pAs value, indicating that these compounds are also acting via the
ADP receptor, presumably as partial agonists. Their effects on platelet cytosolic calcium
concentration, therefore, exactly mimic their effects on platelet aggregation, where they
act as partial agonists, with ADPaS being more potent than ADPpS. However, at least in
the case of ADPaS, the effect differs from that on stimulated adenylate cyclase activity
where ADPaS acts as an antagonist (Cusack & Hourani, 1981a,b). Overall, therefore, the
ability of agonists to cause increases in platelet cytosolic calcium concentration seems to
agree more closely with their ability to induce aggregation than their ability to inhibit
stimulated adenylate cyclase activity supporting the idea that calcium mobilization is the
most important signal for ADP-induced aggregation. The lack of correlation between the
ability of these agonists to induce increases in cytosolic calcium concentration and to
inhibit adenylate cyclase also suggests that the inhibition of adenylate cyclase is not
simply a consequence of this increase in calcium concentration but is a separately
mediated event.
Of the analogues tested as antagonists ATP, Cl-ATP, ATPaS, ATPpS, Ap5A and
AMPPCP caused apparently competitive inhibition of the effect of ADP-induced calcium
mobilization, although apart from ATP only one concentration of each antagonist was
tested. AMPCPP and UTP were essentially inactive as inhibitors. The weak action of
AMPCPP is in agreement with its reported lack of inhibitory activity on ADP-induced
aggregation (Lips et al., 1980), and confirms the sensitivity of the ADP receptor to
modifications of the phosphate chain. The lack of inhibitory effect of UTP is also not
unexpected as, although UTP itself has not been tested as an inhibitor of ADP-induced
aggregation, the corresponding diphosphate, UDP, is 500 - 1000 fold less potent as an
aggregating agent than ADP (Chambers et al., 1968), and the phosphorothioate analogue
of UTP, UTPaS, is much less potent than ATPaS as an inhibitor of ADP-induced shape
69
change and aggregation, with 700 - 800 joM being required to fully inhibit aggregation
induced by 5 juM ADP (Greco et al., 1992). In this study [3H]-ADP binding was displaced
by ATPaS, GTPaS and CTPaS with veiy similar K; values and by UTPaS only 2 - 1 0
fold more weakly and this was taken to imply that the ADP receptor shows little
specificity for the adenine base. However, in the same paper ATPaS was said to be at
least 40 - 80 fold more effective than these other nucleotide analogues at inhibiting
functional responses to ADP which is consistent with previous reports (see section 1.6 of
the General Introduction) and directly contradicts their assertion of low structural
specificity. In view of the discrepancy between these binding affinities and the weak
inhibitory effect of GTPaS, CTPaS and UTPaS on functional responses to ADP these
binding sites probably do not represent functional ADP receptors. Thus, although there is
increasing interest at the moment in the possibility that there may be nucleotide receptors
which recognise both ATP and UTP (O'Connor et a l, 1991), and there has even been a
suggestion that there may be specific pyrimidoceptors (Siefert & Schultz, 1989), the
platelet ADP receptor appears to be a true purinoceptor, with a high degree of specificity
towards the adenine base (see Haslam & Cusack, 1981; Hourani & Cusack, 1991 for
review).
MeSATP inhibited ADP-induced increases in platelet cytosolic calcium concentration but
in a non-competitive manner, and this is veiy similar to its effects on aggregation, as in
each case the concentration-response curve to ADP in the presence of the inhibitor
reached a plateau at about 60 - 70 % of the maximal effect in control experiments.
Several 2-alkylthio analogues of ATP and AMP, including MeSATP, have been shown to
be non-competitive but specific inhibitors of ADP-induced aggregation (these compounds
70
showing no ability to inhibit aggregation induced by 11,9-epoxymethanoprostaglandin
a thromboxane agonist), although their mechanism of action is unknown (Cusack &
Hourani, 1982a; Hourani et al., 1986). Interestingly, MeSAMPPCP appeared to be a
competitive antagonist of the effects of ADP on adenylate cyclase (Hourani et a l, 1986).
Thus, in common with the agonists tested, the effects of MeSATP on ADP-induced
increases in cytosolic calcium concentration agree well with its effects on ADP-induced
aggregation, whilst the agreement between its effects on ADP-induced increases in
cytosolic calcium concentration and those of MeSAMPPCP on inhibition by ADP of
adenylate cyclase is veiy poor, again suggesting that these two second signalling
responses to ADP are independent events.
For the six apparently competitive antagonists, the pA2 values obtained against
ADP-induced increases in cytosolic calcium concentration in this study were in very good
agreement with the pA2 values for the inhibition of ADP-induced aggregation reported by
Cusack & Hourani (1982b) (or in the case of ATPPS also determined in this study)
(correlation coefficient 0.976, p<0.01) (figure 2.8a). This confirms the conclusion that the
above two events are causally related and that they are mediated by the same receptor and
casts further doubt on the suggestion that FSBA can be used to label a platelet ADP
receptor (Colman, 1992). FSBA has been reported to have no inhibitory effects on
ADP-induced increases in cytosolic calcium concentration (Rao & Kowalski, 1987) an
observation which was taken by these authors to imply that aggregation and shape change
were mediated by a receptor which is distinct from that which induces increases in
cytosolic calcium concentration. However, in the light of the results of the present study
which show an extremely strong relationship between aggregation and increases in
platelet calcium and the serious doubt which already exists over the specificity of the
71
a<N
4 5 6
pA2 against aggregation
o S 3
°*3 >* c °•H <d
4-> - J i C/2 CO
.9CO Ch 60 <L> cO T 3
CO
6 n
5 -
4 65
pA2 against increases in [Ca2+]i pA2 against aggregation
Figure 2.8 Comparison of the abilities of a number ATP analogues to inhibit responses to ADP. (a) Comparison of the pA2 values against aggregation and increases in cytosolic calcium concentration ([Ca2+];). (b) Comparison of the pAj values against increases in [Ca2+]i and inhibition of adenylate cyclase, (c) Comparison of the published pA values against aggregation and inhibition of adenylate cyclase for the compounds shown in (b). The lines show the least squares regression fit to the data points. Compounds were: ATP (•), ATPaS (■), Cl-ATP (A), Ap5A (♦), AMPPCP (▼) and ATP|3S 00- For abbreviations, see text.
72
effects of FSBA (see section 1.6.3 of the General Introduction) it can only be concluded
that the protein labelled by FSBA (’aggregiri) is not an ADP receptor but represents a
factor which is essential for the more distal events associated with platelet activation.
For the five antagonists reported here whose effects on inhibition by ADP of adenylate
cyclase are known (the effects of ATPpS on this response have not been studied), the
correlation between the reported pA2 values determined against inhibition of adenylate
cyclase (Cusack & Hourani, 1982b) and those obtained in this study against ADP-induced
increases in cytosolic calcium concentration was less good and did not quite reach
significance at the 5 % level (correlation coefficient 0.826, 0.10>p>0.05) (figure 2.8b).
However, if the published pAs values of these same 5 antagonists against aggregation and
inhibition of adenylate cyclase (Cusack & Hourani, 1982b) are compared, the correlation
is also not quite significant at the 5 % level (correlation coefficient 0.847, 0.10>p>0.05)
(figure 2.8c), whilst when the 3 other antagonists used in the study of Cusack & Hourani
are included the correlation is highly significant (p<0.01) (see Cusack & Hourani,
1982b). This suggests that a larger number of antagonists would need to be studied before
the significance of the correlation between these two parameters (antagonism of
inhibition of adenylate cyclase and increases in cytosolic calcium concentration) can
properly be evaluated.
In conclusion, these results show that there is an extremely good agreement between the
effects of a number of ADP receptor agonists and antagonists on platelet cytosolic
calcium concentration and platelet aggregation. This provides strong evidence that these
two responses are mediated by the same receptor and provides further support for the idea
that an increase in intracellular calcium concentration is the major signal for platelet
73
aggregation induced by ADP.
CHAPTER 3
STUDIES WITH SURAMIN
75
3.1 Introduction
It has recently been reported that the trypanocidal drug suramin (the structure of which is
shown in figure 3.1) is a selective, competitive antagonist at P2X and P2Y receptors on
vascular and visceral smooth muscle preparations, although it does not distinguish
between these two receptor subtypes having a pA2 value of around 5 in each case (Den
Hertog et al., 1989a, b; Dunn & Blakeley, 1988; Hoyle et al., 1990; Leff et al., 1990; Von
Kiigelgen et al, 1990). Suramin also selectively inhibits the effects of ATP on PC 12
phaeochromocytoma cells with a similar potency, the reported pA2 value being 4.5 (Inoue
et a/., 1991; Nakazawa et al., 1990). As well as antagonizing P2 -purinoceptors, suramin is
also known to inhibit a number of other proteins with nucleotide binding sites, for
example yeast hexokinase (Wills & Wormall, 1950), erythrocyte membrane
Na+/K+-ATPase (Fortes et al., 1973), firefly luciferase (Fortes et al., 1973), vacuole-type
H+-ATPases (Calcaterra et al., 1988; Moriyama & Nelson, 1988), smooth muscle
ectonucleotidases (Hourani & Chown, 1989), protein kinase C (Mahoney et al., 1990),
mitochondrial adenine nucleotide exchanger (Calcaterra et a l, 1988), the GTPase activity
associated with G; (Butler et al., 1988) and various polynucleotide synthesising enzymes
(Broder et al., 1985; Offensperger et al., 1988; Ono et al., 1988). As the ADP receptor is
classified as a P2-purinoceptor and acts as an ATP binding site (ATP being an antagonist
at this receptor; Macfarlane & Mill, 1975), it was of interest to determine the effects of
suramin on platelet responses to ADP. The effects of this compound on these various
responses may also help to further resolve the debate as to whether there are one or two
ADP receptors on platelets (see section 1.6.3 of the General Introduction) responsible for
mediating these effects.
76
Some of the data presented here have already been published (Hourani et al., 1992) and
some have been reported as an abstract (Hall & Hourani, 1992), and copies of these
papers are included at the back of this thesis.
Figure 3.1 The structure of suramin.
3.2 Materials and Methods
3.2.1 Materials
ADP, ATP, adrenaline, bacitracin, BSA, 11 a,9a-epoxymethanoprostaglandin H2
(U46619), fibrinogen (fraction 1 from human plasma, essentially plasminogen free),
fura-2-AM, 5-hydroxytryptamine creatinine sulphate complex (5-HT),
isobutylmethylxanthine (IBMX), prostaglandin E, (PGE,), prostacyclin and triton X-100
were from Sigma Chemical Company. AG50W-X8 [TP] was obtained from Bio Rad.
[2,8-3 H]-cyclic AMP, ammonium salt (41.5 Ci/mmol) in 50 % aqueous ethanol and
[U-1 4 C]-adenine (270 mCi/mmol) in 2 % aqueous ethanol were from Amersham
International, pic. Suramin was a generous gift from Bayer, UK. All other chemicals were
AnalaR grade from BDH. Prostacyclin was dissolved at 100 pg/ml in 10 mM NaOH,
PGE, was dissolved at 1 mM in 50 % aqueous ethanol, U46619 was initially dissolved at
30 mM in absolute ethanol then diluted to 10 mM with distilled water and fura-2-AM was
77
dissolved at 5 mM in dimethylsulphoxide. These solutions were stored at -20°C. IBMX
was dissolved in HEPES-saline and all other drugs were dissolved in water. Adrenaline,
bacitracin, fibrinogen, 5-HT, IBMX and suramin were made up fresh each day while the
nucleotides were stored frozen.
3.2.2 Platelet Aggregation Studies
Venous blood was drawn from healthy human volunteers into one sixth of the volume of
Acid-Citrate-Dextrose anticoagulant (25 g/1 tri-sodium citrate dihydrate, 15 g/1 citric acid
monohydrate, 20 g/1 glucose) and centrifuged at 260 g for 20 min. Volunteers denied
taking drugs for 10 days before the experiment. The PRP was removed and the platelets
isolated by centrifugation at 680 g for 20 min in the presence of prostacyclin (1 juM) (to
prevent activation). The supernatant plasma was discarded and the platelets resuspended
at a density of 1 x 108/ml in HEPES-saline. Aggregation and shape change were followed
photometrically (Bom, 1962a, b) with a Chrono-Log Lumi-Aggrometer, and aggregation
was quantified as the maximal rate of change in light transmission (expressed as arbitrary
units per min) through a stirred sample (0.5 ml) at 37°C on addition of agonist. Human
fibrinogen (0.3 mg/ml) was added to all samples 20 sec before addition of agonist, and
CaCl2 (1 mM) was added at the same time as suramin or at least 3 min before addition of
the agonist. Suramin was added either simultaneously with the agonist or preincubated
with the platelet suspension for various times at 37°C. The agonists used were ADP,
U46619 and 5-HT. In the case of 5-HT, adrenaline (100 juM) was added simultaneously
to potentiate its effects and result in a measurable aggregation response. Studies on the
aggregation of platelets in plasma were performed as described in section 2 .2 .2 .
78
EC5 0 values were obtained by regression analysis of the linear portion of the log
concentration-response curves and dose ratios were calculated from the EC5 0 values
obtained in the presence and absence of suramin. Where Schild analysis was not
performed the apparent values for antagonists were calculated using the formula
Kd = concentration of antagonist / (dose ratio - 1) and p ^ values were calculated as
-log1 0 KD. Statistical significance was determined using Student’s t-test.
3.2.3 Calcium Measurement
Platelets were loaded with fura-2 as described in section 2.2.3. Aliquots of platelet
suspension (0.7 ml; 1 x 108 platelets/ml) were incubated in siliconised glass aggregometer
tubes at 37°C for 3 min in the presence of 1 mM CaCl2 and transferred to a Perkin-Elmer
LS50 luminescence spectrometer for estimation of cytosolic calcium concentration. The
stirred suspension was maintained at 37°C and excited with light at 380 nm. The emitted
light intensity was measured at 510 nm. The fluorescence intensity at a saturating
concentration of calcium ( F J and in the absence of calcium (Fabs) were determined by
lysing the platelets with 0.1 % triton X-100 (F^J followed by the sequential addition of
20 mM tris and 10 mM EGTA. The cytosolic calcium concentration was then calculated
using the equation [Ca2+] = KD (F-Fabs) / (F^-F) (Grynkiewicz et a l , 1985), where KD is
the dissociation constant of the fura-2.calcium complex (225 nM). Suramin has a broad
ultra-violet absorption spectrum and absorbs wavelengths up to 370 nm, it was therefore
not possible to use the more usual dual wavelength method of calcium measurement with
fura-2 as this requires excitation at around 340 nm. At 380 nm an increase in calcium
concentration corresponds to a decrease in the fluorescence intensity of the dye (see
figure 3.2). Agonists were added in 14 jul of water and suramin, where used, was added
79
<Z)*23
S-,aS—!
su<w
CDOcooc/3<DJ-.oJ3
E
120 ADP
90
60
Triton X-100
Tris, EGTA30
0 10 20 30
Time / sec
Figure 3.2 A typical fluorimeter trace from single wavelength calcium measurement. The trace shows the change in fluorescence intensity of a sample of fura-2-loaded platelets on addition of ADP (100 pM) (solid line) and a corresponding calibration spectrum (dashed line).
80
simultaneously with the agonist. The response was quantified as the maximal change in
cytosolic calcium concentration following addition of the agonist. The agonists used were
ADP and 5-HT0adrenaline was not added with the 5-HT as a reliable response was
obtained in its absence.
Log concentration-response curves were obtained at least three times using blood from
different donors and the results were pooled. EC5 0 values were obtained by linear
regression of the linear parts of the pooled log concentration-response curves and dose
ratios were calculated from the EC5 0 values in the presence and absence of suramin.
3.2.4 Measurement of Inhibition of PGE}-stimulated Adenylate Cyclase
PRP (obtained as in section 3.2.2) was incubated with 4 pM [1 4 C]-adenine for 45 min at
37°C. The platelets were then collected by centrifugation at 680 g for 20 min in the
presence of prostacyclin (1 pM) (to prevent activation) and resuspended at 1.1 x 108/ml in
HEPES-saline. Platelet cyclic AMP (cAMP) levels were determined by a modification of
the method of Haslam & Rosson (1975). Aliquots (450 pi) of platelet suspension, which
had been incubated for 3 min at 37°C in the presence of 1 mM CaCl2, were treated with
solutions (50 pi) of ADP or adrenaline which contained PGEj (10 pM), to stimulate
adenylate cyclase, IBMX (1 mM), to inhibit phosphodiesterase, and an appropriate
concentration of suramin. After a further 30 second incubation at 37°C the reaction was
terminated by addition of ice cold perchloric acid (100 pi; 3 M) containing [3 H]-cAMP
(-20,000 dpm) to estimate recovery. After at least 30 min on ice the samples were
centrifuged for 4 min in a microfuge and an aliquot (450 pi) of the supernatant applied to
columns of AG50W-X8 [H+] (1.3 ml) to remove the adenine and adenosine. The cAMP
81
peak was eluted with 1 mM KH^PC (pH 7.5) and the eluate was twice treated with
nascent barium sulphate (by addition of 0.3 ml of 0.25 M zinc sulphate and 0.3 ml of 0.25
M barium hydroxide) and centrifuged, to remove the adenine nucleotides other than
cAMP. An aliquot ( 5 - 6 ml) of the supernatant was lyophilised and the 3H and 14C
estimated by scintillation counting. 14C dpm were corrected for the recovery of
[3 H]-cAMP, and the baseline level of [1 4 C]-cAMP (i.e. that in the presence of IBMX
alone) was subtracted. Inhibition of adenylate cyclase was expressed as a percentage
relative to the level of [1 4 C]-cAMP in platelets stimulated with PGEj alone. Log
concentration-response curves were obtained in platelets from at least three different
donors in triplicate and pooled. The EC5 0 values were calculated by linear regression of
the linear portion of the pooled log concentration-response curves. Dose ratios were
calculated from these EC5 0 values.
Concentration-response curves to PGE} were performed in a similar manner to above, but
in the absence of ADP or adrenaline, and the results were expressed in dpm as
[1 4 C]-cAMP present in the platelet suspension, corrected for the baseline effects of
IBMX. These experiments were carried in the presence of a constant concentration of
ethanol corresponding to that present at the highest concentration of PGEj assayed.
3.3 Results
In citrated human plasma, ADP induced platelet aggregation with an EC5 0 value of 1.1
pM, and this aggregation was inhibited in an apparently competitive manner by
simultaneous addition of ATP (40 pM), with a pA> value of 5.00. Suramin (100 juM)
added simultaneously with ADP had no effect on ADP-induced aggregation or on the
82
inhibition of this aggregation by ATP (figure 3.3a). Preincubation with suramin (100 pM)
for 40 min at 37°C did not inhibit ADP-induced aggregation (figure 3.3b) even in the
presence of the non-selective peptidase inhibitor bacitracin (2 mg/ml) (figure 3.3c).
Suramin alone did not cause shape change or aggregation at concentrations up to 1 mM
(data not shown).
In washed platelets, ADP induced aggregation with an EC5 0 value of 4.3 pM, and suramin
added simultaneously with ADP caused dose-dependent, parallel, rightward shifts of the
log concentration-response curve (figure 3.4a). Schild analysis of these data gave a slope
of 1.82 ± 0.21, which was significantly greater than unity (p<0.05), and an apparent pA2
value (the negative log of the concentration causing a dose-ratio 2) of 4.62 (figure 3.4b).
Under these conditions ATP also caused dose related, parallel, rightward shifts of the log
concentration-response curve for ADP, but gave a Schild plot slope of 0.96 ±0.16 and a
pA2 value of 4.74 (figure 3.4b). Incubation of washed platelets with suramin for 10 or 40
min at 37°C before addition of ADP also yielded Schild plots with slopes significantly
greater than unity (2 . 0 1 ± 0.16 and 2.06 ± 0 .0 1 , respectively), and with increasing time of
incubation there was a reduction in the potency of suramin (apparent pA2 values of 4.43
and 4.05 after 10 and 40 min, respectively) (figure 3.5). For this reason all subsequent
studies using suramin were performed with suramin added simultaneously with the
agonist.
Washed platelet aggregation induced by U46619 or by 5-HT (in the presence of 100 juM
adrenaline) was also inhibited by suramin at concentrations of 200 or 400 pM, but this
inhibition was not competitive and suramin caused a marked reduction in the maximal
response to these agonists. ATP (100 pM) also reduced the maximal response to U46619,
83
0.1 1 10 100
a. S o .tS.S60 c<D wtrv ^w) a>&p p,^ C0
^ .tso<D -*-<s
Is —. § ££3 03 03 'w '
a>>V-i
2
10 100
3 0 -
10 1000.1 1
ADP concentration / jllM
Figure 3.3 Aggregation of human platelets in plasma induced by ADP alone (•) or in the presence of suramin (100 pM) (O), ATP (40 pM) (■) or both ATP (40 pM) an<i suramin (100 pM) (□). (a) Antagonists added simultaneously with ADP. (b) Platelets preincubated with suramin for 40 min before addition of ADP. (c) Platelets incubated in the presence of bacitracin (2 mg/ml) and suramin or vehicle for 40 min before addition of ADP. Each point is the mean of 3 determinations, and vertical bars show the s.e. mean.
84
2
G
1_ J
G1*4
0.1 1 10 100
ADP concentration / jiM
b2 n
ocd
5 3-4
log [antagonist]
Figure 3.4 The effect of suramin on washed platelet aggregation induced by ADP. (a) Effect of ADP alone (O) (n = 10) or in the presence of suramin at 50 pM (•) (n = 10), 100 pM (■) (n = 10), 150 pM (A) (n = 3), 200 pM ^ ) (n = 10) or 400 pM (▼) (n = 6 ). Each point is the mean of 3 - 10 determinations (as shown in the brackets), and vertical bars show the s.e. mean, (b) Schild plot of the data presented in (a) (•), and an equivalent Schild plot for the antagonism of ADP-induced washed platelet aggregation by ATP (■).
85
e<DcoO
3W)o
1
0
-5 -4 -3
log [suramin]
Figure 3.5 Schild plots obtained when washed platelets were preincubated for 0 (•) (as shown in figure 3.4), 10 p ) or 40 min (▲) before addition of ADP.
86
the effect being similar to 200 pM suramin, but did not affect aggregation induced by
5-HT (figure 3.6).
ADP also induced concentration-dependent increases in platelet cytosolic calcium
concentration with an EC5 0 value of 1.0 pM. This effect of ADP was inhibited by suramin
which caused dose-dependent, rightward shifts of the log-concentration response curve to
ADP (figure 3.7a). Schild analysis of these data gave a slope of 2.3 ± 0.3, which was
significantly greater than unity (p<0.05) and an apparent p ^ value of 4.63 (figure 3.7b).
Suramin (100 - 400 pM) was also able to inhibit increases in platelet cytosolic calcium
concentration induced by 5-HT although this inhibition was not competitive as suramin
caused a decrease in the maximal response to this agonist and the effect was limited as
200 and 400 juM suramin caused similar levels of inhibition (figure 3.8). Suramin alone
did not affect the fluorescence of the platelet suspension and therefore had no effect on
the basal cytosolic calcium concentration at any concentration tested.
PGE^stimulated cAMP accumulation was concentration-dependently inhibited by ADP,
giving an EC5 0 value of 0.47 pM. Suramin inhibited this effect of ADP in a
concentration-dependent manner (figure 3.9a) and Schild analysis of this inhibition
resulted in a slope of 1.00 ± 0.10 and a value of 5.09 (figure 3.9b). Suramin had no
effect on the inhibition of stimulated adenylate cyclase by adrenaline (figure 3.10), but
was able to inhibit accumulation of cAMP induced by sub-maximal concentrations of
PGEj, causing a 4 fold shift to the right of the log concentration-response curve (figure
3.11). Suramin alone (400 pM) had no significant effect on the basal level of cAMP in
platelets (289 ± 34 dpm (n = 7) in the absence of suramin, 292 ±41 dpm (n = 3) in its
presence) or on the effect of PGE} (1 pM) in the experiments measuring inhibition by
a
.9£<3Dh0 24—1
• cS3
03t-i-t—I
*£wc ocdtoOeW)tooC3
<+-tO<D4->cdVh
73a*ctf
4 0 - ,
3 0 -
1 0 -
0 -J100100.1 1
10
0 -J1001010.1
Agonist concentration / jiM
Figure 3.6 Aggregation of washed platelets induced by (a) U46619 or (b) 5-HT (in the presence of 1 0 0 pM adrenaline) alone (•) or in the presence of suramin ( 2 0 0 pM (■) or 400 juM (A)) or ATP (100 pM) (#). Each point is the mean of 3 determinations, and vertical bars show the s.e. mean.
88
a
I 150-
50-
100 3000.1 101
ADP concentration / jiM
b
ioaS-iooT3
3■5 ■4
log [antagonist]Figure 3.7 Effects of suramin on increases in platelet cytosolic calcium concentration ([Ca2+];) induced by ADP. (a) Increases in platelet [Ca2+]; induced by ADP in the absence (•) or presence of suramin at 50 pM (■), 80 pM (A), 130 pM ($) or 200 pM (f). Each point is the mean of three separate determinations on platelets from different donors, and vertical bars show the s.e. mean, (b) Schild plot of the data presented in (a) (•) and the Schild plot of inhibition by ATP of this response (taken from figure 2.2) (dashed line).
89
100-I
+
u
<D!ZJcd£O5
5 0 -
0 J ,--------------------------1--------------------------1------------ 10.1 1 10 30
5-HT concentration / jiM
Figure 3.8 Effects of suramin on increases in platelet cytosolic calcium concentration ([Ca2 +]j) induced by 5-HT. Increases in [Ca2+]j induced by 5-HT in the absence (•) or presence of suramin at 100 pM (■), 200 pM (A) or 400 pM (♦). Each point is the mean of three determinations on blood from different donors, and vertical bars show the s.e. mean.
90
a
Bett3.§ <L> £3 coco J3W £
<D
100 n
oOh«4H J-S
>>CJ<D
-20 J
ADP concentration / jiM
«sS-4
<DCOO3OS)o
2
1
0
5 ■4 3
log [suramin]
Figure 3.9 Effects of suramin on the inhibition of PGE,-stimulated cAMP accumulation by ADP. (a) Inhibition of PGEj-stimulated cAMP accumulation by ADP in the absence (•) or presence of suramin at 50 pM (■), 100 pM (▲), 200 pM ($) or 400 pM (▼). Each point is the mean of at least 3 determinations on platelets from different donors, and vertical bars show the s.e. mean, (b) Schild plot of the data presented in (a).
91
100 n
80 -
<Dooc3o
U60 -
0 J0.01 1010.1
Adrenaline concentration / jiM
Figure 3.10 Inhibition of PGEr stimulated cAMP accumulation by adrenaline in the absence (•) or presence (♦) of suramin (400 pM). Each point is the mean of at least 3 determinations on platelets from different donors, and vertical bars show the s.e. mean.
92
800“ i
y 400
PGEj concentration / jiM
Figure 3.11 A representative graph showing the accumulation of cAMP in human platelets induced by PGEj in the absence (•) or presence (■) of suramin (400 juM). Each point is the mean of triplicate determinations, and vertical bars show the s.e. mean.
93
ADP or adrenaline (933 ± 128 dpm (n = 7) in the absence of suramin, 897 ±117 dpm
(n = 7) in its presence).
3.4 Discussion
In this study suramin was able to antagonize ADP-induced aggregation of human platelets
in buffer but had no effect on platelets in plasma. The lack of effect of suramin in plasma
was not likely to be due to degradation of the antagonist by plasma enzymes, as varying
the incubation time between 0 and 40 min even in the presence of the peptidase inhibitor
bacitracin did not affect the results. Indeed, in pharmacokinetic studies in vivo suramin
has been shown not to be significantly metabolized, however it did bind non-specifically
and with high capacity to plasma proteins, with greater than 99 % of the drug being
protein-bound at equilibrium (Collins et al., 1986). It is likely therefore that the lack of
effect of suramin on platelets in plasma is due to its rapid binding to plasma proteins.
In washed platelets, suramin antagonized the aggregation of human platelets induced by
ADP in an apparently competitive manner, causing parallel, dose-dependent, rightward
shifts of the log concentration-response curve to ADP without decreasing the maximal
response (where full log concentration-response curves were obtained). However, Schild
analysis of this inhibition revealed that suramin was not acting as a pure competitive
antagonist as the slope of the Schild plot was around 2. In a detailed study of the
antagonism by suramin of the effects of ATP on P2x-purinoceptors in the rabbit ear artery,
Leff et al. (1990) also reported a Schild plot slope significantly greater than unity and
attributed this to slow equilibration of suramin with the receptors, as increasing the
incubation times for low concentrations of suramin reduced the Schild plot slope to unity.
From an analysis of the time-dependence of the inhibitory effects of suramin, Leff et al
94
(1990) concluded that very long incubation times, up to 220 min, were necessary to reach
equilibrium at low concentrations. However, in platelets slow equilibration is unlikely to
be the explanation for the steep Schild plot slope as increasing the incubation time from 0
to 40 min (incubation times longer than 40 min could not be used because responses to
ADP could not be reliably obtained) did not affect the slope of the Schild plot but merely
reduced the potency of the antagonist, possibly due to slow binding of suramin to the
BSA which was routinely included in the assay buffer. In addition, in a stirred suspension
of platelets access of suramin to the receptors is likely to be more rapid and complete
than in a smooth muscle preparation, and long incubation times are not necessary to
obtain Schild plot slopes of unity with other antagonists in PRP (Cusack & Hourani,
1982b).
Suramin also caused a concentration-dependent inhibition of ADP-induced increases in
platelet cytosolic calcium concentration and again the slope of the Schild plot for this
inhibition was significantly greater than unity. Interestingly, the Schild plot slope
obtained for the inhibition by suramin of this response, 2.3 ± 0.3, was very similar to that
obtained for its inhibition of platelet aggregation, 1.82 ± 0 .2 1 , as was the apparent pAs
value, these being 4.63 and 4.62, respectively. In chapter 2, it was shown that there is a
very strong correlation between the ability of a number of ADP receptor antagonists to
inhibit ADP-induced aggregation and their ability to inhibit ADP-induced increases in
cytosolic calcium concentration. However, the compounds used in this study were all
analogues of ATP and were therefore all structurally quite similar to ADP. Suramin is
structurally unrelated to these compounds and yet still shows an extremely good
agreement between its effects on these two responses to ADP. This observation further
strengthens the evidence for the suggestion, made in Chapter 2, that these two effects of
ADP are mediated by the same receptor.
Suramin also acted as an antagonist of the inhibition by ADP of PGEj-stimulated
adenylate cyclase activity. However, in this case Schild analysis resulted in a slope of
unity, suggesting that the effect was competitive, the pA value being 5.09 (corresponding
to an apparent KD value of 8 .1 juM). Suramin (400 jliM) did not reduce the ability of
adrenaline (which was used here as a control agonist in preference to 5-HT which has
little effect on platelet adenylate cyclase activity) to inhibit stimulated adenylate cyclase
activity. However, suramin did inhibit the accumulation of cAMP induced by PGEj,
although at the concentration of PGEj used in the inhibition assays (1 juM, which is close
to being maximally effective) the effect of 400 pM suramin was insignificant (when
measured in parallel control assays in these experiments). The inhibition of the effect of
PGEj by suramin appears to be competitive, in that the log-concentration curves are
parallel and the maximal response is not decreased. However, suramin was somewhat
less potent at inhibiting the effect of PGEj than responses to ADP, 400 pM suramin
causing only a 4 fold shift in the log concentration-response curve to PGEj. This
concentration of suramin caused an approximately 1 0 0 fold shift in the log
concentration-response curve for ADP-induced aggregation and an approximately 40 fold
shift in the log concentration-response curve for inhibition by ADP of stimulated cyclase
activity. Interestingly, both ADP and PGEj are negatively charged in solution at
physiological pH and this may be taken to imply the presence of positively charged
regions in their respective receptor binding sites to pair with these negative charges. In
early studies with suramin as an enzyme inhibitor (Wills & Wormall, 1950) it was
suggested that this compound has its inhibitory effects by binding to positively charged
96
regions of these proteins. As ADP is a more highly charged molecule than PGEj
(possessing a diphosphate chain compared to a single carboxylate group) this binding of
suramin to positive charges could be used to explain the relative potencies of suramin
against these two agonists, the ADP receptor contains a larger positively charged region
and would therefore bind suramin more strongly. These data may therefore suggest that
suramin acts as an antagonist at the platelet P2T receptor, as it does at smooth muscle P2X
and P2Y receptors, by charge pairing with positively charged amino acid residues in or
near to the nucleotide binding site. However, it also has various non-specific inhibitory
effects (it can also non-competitively inhibit responses to 5-HT and U46619, this will be
discussed further below) possibly due to its ability to bind to positively charged regions of
other proteins on the platelet surface.
In the cases of ADP-induced aggregation and increases in cytosolic calcium
concentration, inhibition by suramin resulted in Schild plot slopes significantly greater
than unity. Possible explanations of a Schild plot slope greater than unity are that the
antagonist is not in equilibrium with the receptor population, that more than one
molecule of antagonist is binding to the receptor, that it is acting non-competitively, that
the antagonist is acting on a heterogeneous population of receptors or that it is having
multiple effects on the response (Kenakin, 1993). As discussed earlier, the antagonist not
being at equilibrium with the receptors is unlikely to be the explanation of the steep
Schild plot slope in this study as increasing the incubation time from 0 to 40 min did not
decrease the slope of the Schild plot, as would be expected if this were the case. A Schild
plot slope near 2 could imply that on average 2 molecules of suramin must bind to the
receptor to cause its inhibitory effect, but although from our results it is not possible to
draw firm conclusions about the stoichiometry of the interaction, there is no evidence that
any other ADP antagonist exhibits multiple binding interactions (Cusack & Hourani,
1982b). However, the symmetrical ('dimeric') nature of the suramin molecule may confer
different binding characteristics from those of previously tested ADP antagonists, which
were all analogues of AMP and ATP. The third possibility, that suramin is acting as a
non-competitive antagonist, cannot be ruled out, however the log concentration-response
curves to ADP were shifted to the right with no apparent depression of the maximal
response, at least at concentrations of suramin for which full concentration-response
curves could be obtained, suggesting against this.
The fourth possibility, that there are multiple ADP receptors, has been suggested by
Colman and co-workers (1980; see also Colman, 1992), although this suggestion (as has
already been pointed out in section 1.5.3 of the General Introduction) is not consistent
with results from antagonist studies (Cusack & Hourani, 1982b). In any case, of the two
receptors proposed in this model only one is thought to mediate aggregation, the other
being proposed to mediate the inhibition of adenylate cyclase. From the results of the
study presented here, if there are indeed two ADP receptors on platelets then both are
responsible for mediating aggregation (and indeed for mediating increases in cytosolic
calcium concentration), although as suramin gave a Schild plot slope of unity for
inhibition of the effects of ADP in adenylate cyclase this may imply that only one of these
putative receptors mediates this response. This could suggest that the ion-channel
mediated (calcium influx) and G-protein mediated (mobilization of intracellular calcium
and inhibition of adenylate cyclase) responses to ADP are in fact caused by distinct
receptors. This would also explain the apparent non-competitive inhibitory effects of the
2-alkylthio-analogues of ATP and AMP on aggregation and increases in intracellular
calcium concentration. From the results described for MeSAMPPCP on inhibition of
98
adenylate cyclase (Hourani et al., 1986) these compounds appear to behave as
competitive antagonists of the putative G-protein coupled receptor. If, however, this class
of compounds do not inhibit the putative ion-channel coupled receptor, it would explain
the observation that these compounds cannot completely inhibit aggregation (Cusack &
Hourani, 1982a) but inhibit this response by a maximum of only about 50 %. Further this
could explain the lack of inhibitory effect of MeSAMPPCP on responses to ADPaS.
ADPaS is a partial agonist of aggregation (Cusack & Hourani, 1981b) and increases in
cytosolic calcium concentration (see Chapter 2 of this thesis), whilst it is an antagonist of
the effects of ADP on adenylate cyclase (Cusack & Hourani, 1981b). As in this model
MeSAMPPCP only acts at the putative G-protein coupled receptor, at which ADPaS is
also an antagonist, but not at the putative ion-channel coupled receptor, at which ADPaS
is an agonist, MeSAMPPCP would not be expected to show any inhibition of responses to
this compound. This hypothesis could be tested using the analogues ADPaS and
MeSATP. If ADPaS is indeed only an agonist at the putative ion-channel coupled
receptor it should not induce increases in intracellular calcium concentration when this
response is measured in the absence of extracellular calcium. Also, if MeSATP is a
competitive antagonist of the putative G-protein coupled receptor then in the absence of
extracellular calcium this antagonist should show competitive antagonism of
ADP-induced increases in intracellular calcium concentration. However, again this model
is not consistent with the data obtained with other antagonists which showed no
suggestion of multiple receptor sites mediating aggregation and which showed a good
correlation between the ability of these antagonists to inhibit the effects of ADP on
aggregation and adenylate cyclase (Cusack & Hourani, 1982b).
99
However, suramin clearly did have non-specific effects in addition to its antagonism of
ADP, as it non-competitively inhibited aggregation induced by U46619 and 5-HT, which
act at thromboxane ('TP') and 5-HT2 receptors, respectively, on platelets (Hourani &
Cusack, 1991). Also suramin non-competitively inhibits the effects of 5-HT on platelet
cytosolic calcium levels. Given its avid binding to protein molecules it is possible that, at
high concentrations, suramin may interact with platelet surface proteins and in some way
interfere with the pathways by which platelets increase their cytosolic calcium
concentration in response to aggregating agents. The fifth explanation, that suramin is
having multiple effects on the response to ADP is, therefore, the most likely. Although
ATP at high concentrations also non-competitively inhibited aggregation induced by
U46619 this is probably due to its inhibition of the effects of released ADP, as ATP did
not inhibit aggregation induced by 5-HT which is a weaker aggregating agent and does
not induce the release of stored nucleotides from platelets. Indeed to achieve aggregations
to 5-HT it was necessary to potentiate its effects with a fixed concentration of adrenaline,
as either of these agonists alone did not induce reliable aggregations.
Although the pA2 values for suramin obtained against ADP-induced aggregation and
increases in intracellular calcium in this study have to be treated with caution because of
the steep Schild plot slopes and consequent uncertainty as to the mechanism of action of
suramin, these pA2 values and the pA2 value obtained against inhibition by ADP of
stimulated adenylate cyclase activity do correspond well with those obtained in other
studies. As an antagonist of the effects of ADP on adenylate cyclase, where suramin
appeared to behave competitively, the pA2 value was 5.09 which is close to the values
ranging from 4.5 to 5.4 which have been reported for inhibition of the effects of ATP on
smooth muscle preparations and phaeochromocytoma cells (Hoyle et ah, 1990; Inoue et
100
a l, 1991; Leff et al., 1990; Von Kugelgen et a l, 1990). The apparent pA values obtained
for suramin against ADP-induced platelet aggregation and increases in cytosolic calcium
concentration of 4.62 and 4.63, respectively, are also well within this range. In those
cases in which no pA2 value has been reported, the concentrations of suramin which have
shown an inhibitory effect against ATP are generally in the micromolar range, and are
also therefore consistent with these values (Den Hertog et a l , 1989a, b; Hoiting et a l ,
1990). The similarity of the pA2 values found for the inhibition of platelet responses to
ADP and for antagonism of ATP on other tissues, suggests that the ADP receptor is
indeed similar to P2-purinoceptors elsewhere, and that suramin is unable to distinguish
between the sub-types of P2 -purinoceptor. The concentrations of suramin required for
antagonism at P2-purinoceptors are similar to the IC5 0 or K; values reported for inhibition
by suramin of the various enzymes which bind purine nucleotides, which are also
generally in the micromolar range (Wills & Wormall, 1950; Fortes et a l , 1973; Butler et
al, 1988; Calcaterra et a l, 1988; Moriyama & Nelson, 1988; Ono et a l, 1988; Mahoney
et a l, 1990), suggesting that these binding sites may be structurally related.
In conclusion, the results presented here show that suramin is an antagonist of platelet
responses to ADP, although it is not effective in plasma, it is not specific for ADP and in
the case of aggregation and increases in cytosolic calcium concentration it is not simply
competitive. The pA2 values are consistent with the values found for P2-purinoceptors
elsewhere, showing that although it is unique in that ATP is an antagonist rather than an
agonist, the platelet ADP receptor is indeed a type of P2 -purinoceptor.
101
CHAPTER 4
EFFECTS OF EXTRACELLULAR
DIVALENT CATION
CONCENTRATION
102
4.1 Introduction
In physiological buffers, and indeed in biological fluids e.g. plasma, ATP exists as a
mixture of several species including divalent cation complexes, Mg. ATP2' and Ca.ATP2',
and the tetrabasic form, ATP4*. It is now widely accepted that the P2Z receptor responds to
this tetrabasic form of ATP rather than any of the other species of ATP found in solution
(Cockcroft & Gomperts, 1979). Evidence is also accumulating that the preferred agonist
at other ATP receptors is this tetrabasic form of ATP rather than, for example, the
magnesium complex which acts as a substrate of ecto-ATPases (Pearson et a l, 1980) and
is well known as the substrate of kinase enzymes. For example it has been suggested that
ATP4* is the agonist species at the P2X receptor on the guinea-pig vas deferens (Fedan et
a l, 1990) and that ATP4* and UTP4* are the agonists at the P2Y and P2U receptors present
on aortic endothelial cells (Motte et a l, 1993). It would seem therefore that a
characteristic of ATP receptors is that they require the uncomplexed form of ATP and do
not recognise the divalent cation complexes, whilst the opposite is true of ATP
phosphotransferase enzymes. This may have implications as to the conformational
requirements of these proteins for the triphosphate chain of ATP. There are presumably
conformations of the triphosphate chain which are not accessible when it is bound to
magnesium, one of these conformations is therefore likely to be the ligand for the
P2 -purinoceptors. Similarly, there are conformations of the triphosphate chain which are
more favourable in the presence of magnesium and it is one of these conformations which
acts as the substrate for the phosphotransferases (Eckstein, 1985; Frey, 1992).
Although platelets do not aggregate in the absence of divalent cations, due to the
dissociation of the Gp Eb-IIIa complex (Peerschke, 1985), they are still able to undergo a
103
shape change response. This would suggest that extracellular divalent cations are not
required for the binding of ADP to the ADP receptor or for the production of this
response. Also, 2-azido-ADP and MeSADP have both been shown to bind to platelets in
the absence of divalent cations (Macfarlane et a l , 1982; 1983) further suggesting that
extracellular divalent cations are not required for the binding of ADP to its receptor.
However, the divalent cation requirement of the response of platelets to ADP has never
been studied in detail and it is not known whether the ADP receptor responds to all ADP
species or simply to the uncomplexed form of ADP, ADP3*. Similarly it is not known
which form (or forms) of ATP (divalent cation complexes, ATP4*) acts as the antagonist
at this receptor. In this study the effects of varying the extracellular concentration of
magnesium and calcium was determined on the ability of ADP to activate platelets. Two
responses of platelets to ADP were studied, shape change and inhibition of adenylate
cyclase, both of which occur in the absence of extracellular calcium. ADP-induced
platelet shape change, however, is thought to be triggered by the increase in cytosolic
calcium concentration induced by this agonist (see section 1.4 of the General
Introduction). This increase in cytosolic calcium concentration is, in part, caused by
influx of calcium into the platelet from the extracellular medium (Hallam & Rink, 1985a;
Sage & Rink, 1986) and removal of extracellular calcium may therefore attenuate this
response. The effect of extracellular divalent cation concentration was, therefore, also
determined on the ability of ADP to inhibit PGE^stimulated adenylate cyclase activity, as
this response does not appear to be due to the increase in cytosolic calcium concentration,
but seems to occur via an independent, G-protein mediated pathway (see Chapter 2 of this
thesis).
As changes in the potency of an agonist may reflect changes in both the affinity of the
104
agonist for the receptor and its efficacy in inducing the response, the effects of varying
the extracellular divalent cation concentration were also determined on the ability of ATP
to antagonize these ADP receptor mediated responses. The effects of a competitive
antagonist are only dependent upon its affinity for the receptor and thus the effects of
divalent cation concentration on antagonism of the ADP receptor by ATP may help to
clarify the effects of these manipulations on responses to ADP.
4.2 Materials and Methods
4.2.1 Materials
ADP, ATP, BSA, IBMX, PGEl 5 prostacyclin and tetrabutylammonium hydrogen sulphate
were obtained from Sigma. AG50W-X8 [H+] was from Bio Rad. [2,8-3 H]cAMP (41.5
Ci/mmol) in 50 % aqueous ethanol and [U-1 4 C]-adenine (270 mCi/mmol) in 2 % ethanol
were from Amersham International. HPLC solvents were from Fisons, all other chemicals
were AnalaR grade from BDH. PGEj was dissolved at 1 mM in 50 % aqueous ethanol
and prostacyclin was dissolved at 100 pg/ml in 10 mM NaOH. Both compounds were
stored in solution at -20°C. IBMX was dissolved in nominally divalent cation-free
HEPES-saline and nucleotides were dissolved in water. IBMX was made up freshly each
day whilst nucleotides were stored frozen.
4.2.2 Measurement of Platelet Shape Change
Blood was drawn from healthy human volunteers into one sixth of the volume of
Acid-Citrate-Dextrose anticoagulant and centrifuged at 260 g for 20 min to obtain PRP.
Donors denied taking drugs for 10 days before the experiment. Platelets were obtained by
105
centrifuging the PRP at 680 g for 20 min in the presence of prostacyclin (1 juM). The
supernatant plasma was discarded and the platelets were suspended in 1 0 ml of nominally
divalent cation-free HEPES-saline (145 mM NaCl, 5 mM KC1, 10 mM glucose, 10
HEPES, 2 mg/ml BSA adjusted to pH 7.4 with NaOH 1M). The platelets were again
collected by centrifugation at 680 g for 20 min in the presence of prostacyclin (1 pM) and
resuspended at 1 x 108/ml in nominally divalent cation-free HEPES-saline. The second
centrifugation step was employed to help to remove residual plasma proteins which may
allow the platelets to aggregate during shape change measurement.
Aliquots (0.3 ml) of platelet suspension were incubated in siliconised glass aggregometer
tubes at 37°C for 3 min in the presence of 1 mM MgCl2, 1 mM CaCl2, both 1 mM MgCl2
and 1 mM CaCl2, 1 mM EGTA or an equivalent volume of water and transferred to a
Chrono-Log Lumi-Aggrometer. Shape change was quantified as the maximal decrease (in
arbitrary units) in light transmission through the stirred sample at 37°C on addition of the
agonist, relative to a 50 % dilution of the platelet suspension (i.e. 5 x 107 platelets/ml).
Diluted platelet suspension was used as the reference rather than HEPES-saline to
magnify the responses. Under the above conditions, i.e. in the absence of fibrinogen (and
other plasma proteins), aggregation in response to ADP is negligible as measured in the
aggregometer.
ADP was added in 6 pi of water and ATP, where used, were added simultaneously with
the ADP. Log concentration-response curves were obtained in triplicate in blood obtained
from at least 3 donors. EC5 0 values were obtained by regression analysis of the linear
portion of the triplicate concentration-response curves and dose ratios for antagonists
were determined from the EC5 0 values in the presence and absence of the ATP. The
106
apparent KD values for ATP were obtained from the dose ratios using the formula
Kd = concentration of ATP / (dose ratio - 1) and pA^ values were calculated as -log1 0 KD.
Statistical significance was determined by analysis of variance or, when only two sets of
conditions were compared, Students t-test.
4.2.3 Measurement of Inhibition of PGEj-stimulated Adenylate Cyclase
Inhibition of PGEj-stimulated adenylate cyclase activity was measured as described in
section 3.2.4 with the following modifications. Platelets were suspended at 1.1 x 108/ml
in nominally divalent cation-free HEPES-saline which was supplemented with either
MgCl2 (1 mM) and CaCl2 (1 mM) or EGTA (1 mM) during the 3 min preincubation at
37°C before addition of agonist solutions. Log concentration-response curves were
obtained in triplicate in blood from at least three donors and presented as [1 4 C]-cAMP
dpm in the sample corrected for recovery. EC5 0 values were obtained by regression
analysis of the linear portion of these concentration-response curves. Dose ratios for ATP
were determined from the EC5 0 values in the presence and absence of ATP and apparent
Kd values were calculated using the formula KD = concentration of ATP / (dose ratio - 1).
Statistical significance was determined using Student's t-test.
4.2.4 Determination of Nucleotide Release from Platelets
Platelets were resuspended at 1 x 108/ml in nominally divalent cation-free HEPES-saline
and aliquots (1.8 ml) were incubated at 37°C in the presence of either 1 mM MgCl2 and 1
mM CaCl2 or 1 mM EGTA. After 3.5 min, aliquots (0.5 ml) were removed and layered
onto a mixture of silicone oils (4 parts Dow Coming 704 EU diffusion pump fluid and
1 part Dow Coming 200/0.65 cs silicone fluid) (200 pi). The platelets were separated
107
from the suspending medium by centrifugation in a swing out microfuge for 1 min and
aliquots (300 - 400 pi) of the supernatant were removed and stored at -20°C before HPLC
analysis. The concentration of ADP released from the platelets was determined by HPLC
(using methanol as the solvent in buffer B) as described in section 2.2.5.
4.2.5 Calculation of Free ADP3- and ATP4' Concentrations
In solutions containing only a single divalent cation species (i.e. Mg2+ or Ca2+ alone) the
free ADP3* or ATP4* concentration was calculated using the method of Dahlquist &
Diamant (1974) by solving the following quadratic equation:
KJAXP ] 2 + [AXP]/Km[M]t - Km[AXP]t + 1 ) - [AXP]T = 0
where,
M represents either Ca2+ or Mg2+,
AXP represents either ADP or ATP,
Km is the stability constant of the M.AXP complex (logKM = 2.80 for Ca.ADP,
3.30 for Mg.ADP, 3.94 for Ca.ATP and 4.28 for Mg.ATP: Taqui Kahn & Martell,
1966; 1967),
[AXP]f is the concentration of free ADP3- or ATP 4%
[AXP]t is the total concentration of AXP, and
[M]t is the total concentration of M.
In solutions containing both 1 mM calcium and 1 mM magnesium the [AXP]f was
approximated as the mean of the [AXP]f calculated in the presence of 2 mM calcium and
in the presence/2 mM magnesium. No adjustment was made to the free divalent cationK
concentration when determining [AXP]f in solutions containing both ADP and ATP, as
108
the change due to the presence of the other nucleotide would not exceed - 1 0 % even at
the highest concentrations used (i.e. in solutions containing 100 pM ATP and 100 pM
ADP).
4.3 Results
ADP induced platelet shape change in a concentration-dependent manner under all
conditions tested (figures 4.1 & 4.2). The pD2 values (-log1 0 EC50) obtained for ADP were
not significantly different (p<0.05) under the various conditions tested (the pD2 values
under these various conditions are shown in table 4.1). However, there did appear to be a
slight decrease in the maximal response to ADP in the absence of added divalent cations
and in the presence of EGTA (figure 4.2c). This shape change was inhibited by ATP in all
cases (figures 4.1 & 4.2a & b), although this inhibition by ATP was significantly greater
in the presence of EGTA and in the absence of added divalent cations than in the
presence of divalent cations (p<0.05) (the pA2 values for ATP under these conditions are
also shown in table 4.1). This increase in pA2 value shows an 18.2 fold increase in the
affinity (apparent KD) of ATP for the ADP receptor in the presence of EGTA compared to
the presence of both calcium and magnesium.
ADP also inhibited PGEr induced accumulation of cAMP by platelets both in the
presence of divalent cations and in the presence of 1 mM EGTA, however, in this case
there was a significant leftward shift in the log concentration-response curve in the
absence of divalent cations (figure 4.3). The pD2 values were 6.48 ± 0.10 in the presence
of divalent cations (EC5 0 = 0.33 pM) and 7.21 ± 0.24 in the presence of EGTA
(EC50= 0.07 pM) (p<0.05). Again ATP inhibited the response to ADP and this inhibition
109
00 4-» • <—(
GG
cd
•ecd
<DW)
43u
0 )f tcd
CO
4 0 n
3 0 -
20 -
10 -
0
0.01 0.1 10 100
CO-4->GG
2■w
■ecdoW)Gcd
6<Df tcd43
CO
4 0 n
3 0 -
20 -
o - i ,------- ,------ 40.01 0.1 1 10 100
c4 0 n
3 0 -
20 -
10-
0.01 0.1 1 10 100
ADP concentration / jiM
Figure 4.1 Shape change induced by ADP alone (filled symbols) or in the presence of ATP ( 1 0 0 jliM ) (open symbols), in the presence of (a) 1 mM MgCl2 and 1 mM CaCl2
(• ,o ) , (b) 1 mM CaCl2 alone (■,□) or (c) 1 mM MgCl2 alone (A ,A). Each point is the mean of 3 triplicate determinations on platelets from different donors, and vertical bars show the s.e. mean.
110
cs
cd-t—>• pH•ewOW)
xsUocdX3
CO
40 n
30-
10 -
0.01 0.1 1 10 100
40 n
30-
10 -
0 J i 1 r0.01 0.1 1 10 100
30-
20 -
10 -
1000.01 0.1 1 10
ADP concentration / jiM
Figure 4.2 Shape change induced by ADP alone (filled symbols) or in the presence of ATP ( 1 0 0 j jM ) (open symbols) (a) in the absence of added divalent cations (# ,0 ) or (b) in the presence of 1 mM EGTA (^V ). Each point is the mean of 3 triplicate determinations, and vertical bars show the s.e. mean, (c) A comparison of the data shown in figures 4.1 and 4.2a & b. Symbols are as previously stated, error bars have been omitted for clarity.
I l l
Conditions pD2 for ADP pA for ATP pD2 for ADP3* pA2 for ATP4*
1 mMMgCLj + 1 mM CaCl2 6.42 ± 0.07 4.88 ±0.11 6.97 ±0.06 6.31 ±0.06
1 mM CaCl2 6.67 ±0.12 5.42 ±0.16 6.90 ±0.11 6.38 ±0.16
1 mMMgCl2 6.59 ±0.14 5.36 ±0.16 6.99 ±0.16 6.72 ±0.16
No added divalent cations 6.55 ±0.10 6.13 ±0.20 6.70 ±0.10 6.31 ±0.14
1 mM EGTA 6.60 ± 0.05 6.14 ±0.14 6.71 ±0.10 6.37 ±0.16
Table 4.1 The pD2 values for the induction of platelet shape change by ADP and ADP3' and the pA values for its inhibition by ATP and ATP4-. Values are the mean ± s.e. mean of three determinations on platelets from different donors.
2500-1
<01p— 1U
2000-
^ 1500H
1000-
500-
0 J rH0 0.1 10 100
ADP concentration / jiM
Figure 4.3 Inhibition of PGEr stimulated cAMP accumulation by ADP alone ( • ,▼) or in the presence of ATP (100 pM) (0 ,V). Experiments were performed in the presence of 1 mM CaCI2 and 1 mM MgCl2 (# , 0 ) or in the presence of 1 mM EGTA (▼,V). Unconnected symbols show the levels of cAMP in platelets incubated in the absence (• ,□ ) or presence of PGEj (#,Y) in the presence of divalent cations (• ,■ ) or EGTA (▼,□). The data are representative of 3 similar experiments, each point is the mean of triplicate or quadruplicate (controls) determinations, and vertical bars show the s.e. mean.
113
was significantly greater in the absence of divalent cations than in their presence, the pA
values being 6.59 ± 0.26 and 5.24 ± 0.05, respectively (p<0.01). This gives an
approximately 22 fold increase in the apparent affinity of ATP for the receptor in the
presence of EGTA.
Interestingly, along with its inhibition of the effect of ADP, ATP also seemed to enhance
the ability of PGEj to stimulate adenylate cyclase in the presence of EGTA. In the
presence of divalent cations the accumulation of cAMP induced by PGEj was similar in
the presence or absence of ATP (100 juM) (5578.8 ± 54 dpm (n = 4) in its absence
5522.5 ± 193 dpm (n = 3) in its presence). However, incubation of platelets with 0.5
units/ml apyrase (a concentration which completely inhibited aggregation induced by
1 pM ADP in PRP) could also enhance the PGEr stimulated accumulation of cAMP. In
the presence of EGTA both ATP (100 pM) and apyrase caused a slight increase in the
cAMP accumulation induced by PGEj. The levels of cAMP were 4854.9 ±195 dpm
(n = 3) for PGEj alone, 5318.8 ± 354.1 dpm (n = 3) in the presence of apyrase and
5216.7 ± 101 dpm in the presence of ATP, however, neither of these increases was
significant. Similarly, in the presence of divalent cations there was a slight, but not quite
significant, increase in the amount of cAMP accumulated in the presence of apyrase
(6076.5 ± 285.4 dpm (n = 3)) compared with its absence (5578 ± 54 dpm (n = 4)). Also,
in the presence of EGTA ATP (100 pM) caused no significant accumulation of cAMP in
the absence of PGEj, the basal level of cAMP in platelets was 562 ± 136.4 dpm (n = 4) in
the absence of ATP and 547.9 ±110 dpm (n = 3) in presence of ATP (100 pM). In HPLC
studies, platelets released significantly more ADP, 1.39 ± 0.11 pM (n = 3), in the
presence of EGTA than in the presence of divalent cations, 1.12 ± 0.11 pM ADP (n = 3),
114
(p<0.05).
Figures 4.4 and 4.5 show the data presented so far in this study in terms of the
concentration of ADP3' present rather than in terms of the total ADP concentration.
Again, in the shape change experiments ADP shows no significant difference in potency
in inducing the response in the presence or absence of divalent cations (see table 4.1). In
the experiments on the inhibition of adenylate cyclase when the ADP concentration is
presented in terms of free ADP3' the pD2 values are no longer significantly different, the
values being 6.98 ± 0.10 and 7.21 ± 0.24 in the presence of calcium and magnesium or
EGTA, respectively. Also when the pA2 values for inhibition of these responses by ATP
are determined using the concentration of free ATP4* present rather than the total ATP
concentration these values are no longer significantly different when measured for either
response. The pA2 values for ATP4- from the shape change data are summarised in table
4.1, from the adenylate cyclase inhibition experiments the p ^ values for ATP4' were
6.64 ± 0.03 in the presence of divalent cations and 6.59 ± 0.26 in the presence of EGTA.
4.4 Discussion
These data suggest that the platelet ADP receptor does not recognise all of the species of
ADP present in solution as agonists but that only ADP3', the free ionic form of ADP, is
able to act as an agonist at this receptor. Similarly it would appear that it is ATP4* which
acts as the antagonist at this receptor rather than other ATP species.
In the presence of 1 mM calcium and 1 mM magnesium the concentration of ATP4'
present in solution is approximately 4 % of the total ATP concentration. This is,
therefore, approximately 25 fold less than the concentration present in the absence of
115
0.001 0.01 0.1 1 10 100
ADP3- concentration/ |iM
Figure 4.4 This figure shows the data presented in figure 4.2c but in terms of the ADP3* concentration. Symbols are as previously stated, curves in the presence of ATP and error bars have been omitted for clarity.
116
[14C
]-cA
MP/
dpm
2500
2000
1500-
1000-
500-
°o.doi 0.01 0.1 1 10 100ADP3- concentration / jiM
Figure 4.5 This figure shows the data presented in figure 4.3 in terms of ADP3' concentration. Symbols are as previously described.
117
these divalent cations, which has been assumed to be 100 % of the total ATP present
(although, of course, this is an over-estimate as, for example, the second ionizable
hydroxyl of the y-phosphate will be protonated to some extent at physiological pH). Thus
if ATP4-, rather than all forms of ATP, were the active antagonist at the ADP receptor a
25 fold increase in the affinity of ATP would be expected in the absence of divalent
cations. The observed increases in affinity of ATP for the ADP receptor in the absence of
divalent cations, of 18.2 and 22.3 in the shape change and adenylate cyclase inhibition
experiments respectively, compare very well, therefore, with this predicted value.
In a solution of ADP containing 1 mM of these two divalent cations, however,
approximately 30 % of the ADP is present as ADP3-, and therefore only a 3.3 fold
leftward shift in the log concentration-response curve to ADP would be expected in the
absence of divalent cations. In the experiments measuring inhibition of PGEr stimulated
adenylate cyclase by ADP, there was indeed a small, but significant, leftward shift in the
log concentration-response curve, giving a potency ratio of 5.5, which is also close to that
predicted from the change in the concentration of ADP3- present in the solution in the
absence and presence of divalent cations.
In the case of shape change the EC5 0 values for ADP in the presence and absence of
divalent cations were not significantly different and the observed potency ratio of 1.5 is
only about half of the expected value of 3.3. However, the log concentration-response
curves in the presence and absence of divalent cations are somewhat differently shaped.
The curves in the absence of divalent cations show a lower maximal response and are
rather flatter than those in their presence, suggesting that the removal of extracellular
divalent cations interferes with this response in some way. Interestingly, it is only in the
118
absence of both divalent cations that this depression of the maximal response is seen
rather than simply in the absence of calcium, suggesting that calcium influx is not
required for shape change but that mobilization of intracellular calcium is sufficient to
produce a maximal response. This observation also suggests that some extracellular
factor which requires divalent cations (either calcium or magnesium) for functional
integrity is necessary to produce a maximal shape change in response to ADP. An
obvious candidate for this role would be glycoprotein Ilb-IHa which is known to require
divalent cations for its association and which is essential for aggregation in response to
ADP.
Another possible explanation for the decreased maximal response in the absence of
divalent cations is desensitization of the ADP receptor by the endogenous ADP released
by the platelets when they are kept in buffer, there being more ADP released in the
presence of EGTA than in the presence of divalent cations. However, if this were due to
desensitization, it might be expected that the depression of the maximal response would
be accompanied by a rightward shift in the concentration-response curve, and even if this
rightward shift were not seen the curves should be superimposed when expressed as
percentages of the maximal response. When expressed in this way the curves were not
superimposed (figure 4.6a), indeed the curve in the presence of EGTA was shifted to the
left (see below), suggesting that desensitization is not a satisfactory explanation for this
reduction in the maximal response.
The differing maxima of the log concentration-response curve for induction of shape
change makes the comparison of the EC5 0 values from these curves unreliable as an index
of changing potency. However, when the shape change data in the presence of divalent
119
oC0aoOhCO£13
sB
g<L>W>s
uoo.cdJO
CO
lOO-i
50 -
100100.1 1
ADP concentration / jiM
100 n
5 0 -
0 i r r0.001 0.01 0.1 10 1001
ADP3- concentration / jiM
Figure 4.6 Shape change induced by ADP in the presence of 1 mM CaCl2 and 1 mM MgCl2 (•) or 1 mM EGTA (▼) calculated as % maximal response, (a) Data presented against the total ADP concentration, (b) Data presented in terms of the concentration of ADP3-. Each point is the mean of 3 triplicate determinations in blood from different donors, vertical bars show the s.e. mean.
120
cations and in the presence of EGTA are normalized to percentages of the response to
30 pM ADP (figure 4.6) the curves are comparable and a significant difference between
the pD2 values for ADP is seen under these two sets of conditions, the pD2 values being
6.43 ± 0.06 (EC5 0 - 0.37 pM) in the presence of divalent cations and 6 . 8 8 ± 0.03
(EC5 0 = 0.14 pM) in the presence of EGTA (p<0.01). There is, therefore, a 2.6 fold
increase in the potency of ADP in the absence of divalent cations which again agrees very
well with the ratio of 3.3 expected if ADP3' is in fact the agonist. Indeed when the data
are presented in terms of the concentration of ADP3- the log concentration-response
curves are almost superimposed (figure 4.6b) and there is no significant difference
between the pD2 values, which are 6.96 ± 0.07 and 6 . 8 8 ± 0.03, respectively, further
suggesting that the ADP receptor is in fact an ADP3- receptor.
In the results from the studies on the inhibition by ADP of PGEj-stimulated adenylate
cyclase it was noticeable that, in the absence of divalent cations, ATP not only inhibited
the effects of ADP but also appeared to enhance the ability of PGEj to stimulate
adenylate cyclase. There has recently been a suggestion that ATP, as well as acting as a
competitive antagonist at the ADP receptor, is able to stimulate platelet adenylate cyclase
via a P2X-like receptor (Soslau & Parker, 1989; Soslau et al., 1993). However, washed
platelets are known to release adenine nucleotides into the suspending medium (Mustard
et al., 1989) and this apparent stimulation of adenylate cyclase by ATP may, therefore,
represent the reversal of the inhibitory effects of released ADP rather than a separate
effect of ATP mediated by a distinct purinoceptor subtype on these cells. In the present
study washed platelets did indeed release ADP and, in fact, released slightly more ADP in
the presence of EGTA than they did in the presence of divalent cations. Thus in the
121
absence of divalent cations, when the apparent stimulation of adenylate cyclase by ATP is
most pronounced, platelets not only release more ADP, which will cause a greater basal
level of inhibition of this enzyme, but a greater proportion of this ADP will be in the
active form (ADP3*) and a greater proportion of the added ATP will be present as the
active antagonist (ATP4*). Thus if the apparent stimulation of adenylate cyclase by ATP is
actually due to its inhibition of the effects of released ADP, the ATP would be expected
to show the most marked effect in the presence of EGTA, as indeed it does.
When platelets were exposed to PGEj in the presence of apyrase, which will breakdown
the released ADP to AMP, the increase in cAMP was slightly greater than that in the
absence of apyrase. Although the increase with either apyrase or ATP did not quite reach
significance at the 5 % level there was a trend in the effect of both towards a potentiation
of the effect of PGEj. Thus, the breakdown of released ADP has the same effect as the
presence of ATP on the response to PGEj suggesting that this effect of ATP is simply due
to reversal of the inhibition by released ADP of the response to PGEj. More importantly,
when platelets were exposed to ATP in the absence of PGEj (in the presence of EGTA,
when ATP had its greatest apparent stimulatory effect) the basal level of cAMP as not
significantly different from that in the presence of IBMX alone, further suggesting that
the effect of ATP is not to increase cAMP itself but simply to antagonize competitively
the effects of released ADP.
In the study by Soslau et a l (1993) it was suggested that the ability of ATP to inhibit
aggregation by U46619 and collagen was due to this apparent ability of ATP to increase
platelet cAMP levels. However, both of these aggregating agents are known to cause the
release of ADP from platelet dense-granules (Siess, 1989), this released ADP potentiating
122
the response to the primary agonist. The inhibition of aggregation induced by U46619 and
collagen is, therefore, attributable to the ability of ATP to inhibit the effects of released
ADP. The putative ATP receptor in this study was apparently desensitized by
pre-incubation in the presence of AMPCPP for up to 120 min, the ability of ATP to
inhibit aggregation induced by U46619 and collagen decreasing with increasing time of
incubation. The inhibition of ADP-induced aggregation by ATP remained relatively
constant over this period of time. However, platelets become refractory to ADP when
stored in the absence of ADP scavenging systems (e.g. apyrase or creatine
phosphate/creatine kinase) due to the release of ADP, whilst they are still able to respond
to other aggregating agents. Because of this gradual reduction in the platelet response to
ADP, the relative contribution that released ADP will make to the response to these other
aggregating agents will also decrease with time. Thus ATP would be expected to show a
reduced level of non-competitive inhibition of the response to other agonists simply due
to the reduced contribution to the response made by released ADP. The results of this
study are therefore entirely explicable in terms of previously characterised responses of
platelets and a separate ATP receptor need not be postulated. Also these authors did not
show the effect of this incubation period on their control responses thus it was not
possible to see if there was a reduction in the responses to U46619 or collagen simply due
to the incubation period.
One important technical criticism of these studies (Soslau & Parker, 1989; Soslau et a l,
1993) is that in their studies on adenylate cyclase they claim to have terminated their
assays by addition of 4 mM EDTA (which will simply chelate extracellular divalent
cations) followed by centrifugation to sediment the platelets. However, the results of the
study presented here show that adenylate cyclase activity is not inhibited by chelation of
123
extracellular divalent cations. This must cast further doubt on the results from these
studies.
In conclusion, the data presented in this chapter suggest that the platelet ADP receptor is
in fact an ADP3’ receptor, binding only this form of ADP as its agonist. Similarly, the
receptor only appears to recognise ATP4- as an antagonist. Thus in common with other
P2-purinoceptors the platelet ADP receptor appears to recognise the free ionic form of its
ligands rather than the divalent cation complexes. ATP also appears to potentiate the
accumulation of cAMP by platelets induced by PGEj. However, this potentiation is
probably due to the reversal of the inhibitory effects of ADP released by platelets when
suspended in buffer rather than to any independent effect of ATP.
124
CHAPTER 5
BINDING STUDIES
125
5.1 Introduction
A number of studies have been performed on the binding of radiolabelled ADP to human
platelets and these have been carried out under a wide range of experimental conditions,
for example on intact platelets both in citrated plasma and in buffer and on platelet
membrane preparations (e.g. Bom, 1965; Salzman et a!., 1966; Nachman & Ferris, 1974;
Bom & Feinberg, 1975; Bauvois et al., 1980; Lips et a l, 1980). These authors reported
receptor densities ranging from 30,000 to 100,000 sites per platelet, with affinities in the
low micromolar range (for review see Haslam & Cusack, 1981). However, these studies
suffered from a number of technical difficulties, for example, rapid metabolism of the
radioligand to other species (ATP, AMP and adenosine) which will have different binding
constants and in the case of adenosine will bind to a distinct receptor site and be taken up
into the platelet cytosol. Platelets suspended in citrated plasma release ADP from their
dense granules in response to exogenous ADP causing isotopic dilution. Also, the
platelets will change shape in response to the ADP (and indeed aggregate in citrated
plasma) potentially making the estimation of the non-specific binding difficult due to
changes in the extracellular space when centrifugation assays are employed.
Interpretation of the data from binding studies on platelet membranes will also be
complicated by the exposure of irrelevant nucleotide binding sites on the inner leaflet of
the plasma membrane and on other intracellular membranes (see Haslam & Cusack,
1981; Siess, 1989; Hourani & Cusack, 1991).
The most convincing studies so far performed on intact, functional platelets used the
2-substituted ADP analogues 2-azido-ADP (N3 -ADP) and MeSADP labelled with 32P on
the (3-phosphate (Macfarlane et al., 1982; 1983). These studies were performed on
126
platelets suspended in EDTA-plasma and on washed platelets suspended in calcium-free
buffer. These authors showed that both of these analogues bound to a single class of
binding site, N3-ADP binding to approximately 500 sites per platelet with an affinity of
134 nM in plasma and 17.5 nM in buffer and MeSADP binding to approximately 750
sites per platelet with an affinity of 14.7 nM in plasma and 5.7 nM in buffer. The binding
of both of these analogues was displaced by ADP and ATP with Kj values in the low
micromolar range, by AMP at least 100 fold more weakly than ADP or ATP and by IDP
and GDP at least 20 fold more weakly. These values are consistent with the effects of
these compounds in functional studies (Macfarlane & Mills, 1975; Packham et a l, 1972;
Packham et al., 1974). The binding of MeSADP was displaced by N3 -ADP. In a later
study by Cristalli and Mills (1993) the binding of MeSADP was also shown to be
displaced by ATPaS and ADPpS. ATPaS was equipotent with ATP which is consistent
with the effects of these two compounds on the inhibition by ADP of adenylate cyclase
(Cusack & Hourani, 1982b) and indeed the K; values obtained by Cristalli & Mills
(around 4pM) correspond very well with the KB values obtained by Cusack & Hourani
(4.6 pM for ATPaS and 6.2 pM for ATP). ADPpS was a slightly better displacer of
MeSADP than ADP, although as these compounds are both agonists it is difficult to relate
differences in their affinity obtained from binding studies to differences in potency
obtained in functional studies.
In the study by Macfarlane et al. (1983), the binding of MeSADP was also displaced by
the sulphydryl reagent -mercuribenzene sulphonate (pMBS), a compound which blocks
the ability of ADP to inhibit adenylate cyclase but which does not inhibit shape change
induced by this compound (Mills & Macfarlane, 1977; Macfarlane & Mills, 1981), but
127
was not displaced by FSBA (Mills et a l , 1985) which inhibits shape change and
aggregation but not the effects of ADP on adenylate cyclase (Colman, 1992). This was
taken by these authors as evidence that these two compounds bind to the ADP receptor
which causes inhibition of adenylate cyclase. However, Rao & Kowalski (1987) showed
that whilst pMBS had no effect on ADP-induced shape change, it did inhibit increases in
platelet cytosolic calcium concentration. However, even at 1 mM, pMBS did not
completely inhibit increases in cytosolic calcium concentration induced by ADP and the
increase in cytosolic calcium concentration which occurred in its presence may still have
been sufficient to induce an unaltered shape change. Sensitivity to pMBS but not FSBA
does not, therefore, imply that the MeSADP binding site on platelets only mediates the
inhibition of adenylate cyclase as pMBS can inhibit both of the second messenger effects
of ADP (inhibition of adenylate cyclase and increases in cytosolic calcium concentration)
and there is some doubt whether FSBA labels a functional ADP receptor at all. Also both
N3-ADP and MeSADP are potent aggregating agents as well as inhibitors of adenylate
cyclase, thus the binding of both of these compounds to only a single site on the platelet
surface would suggest that there is in fact only a single receptor responsible for these two
responses to ADP.
More recently a number of binding studies have been performed on formaldehyde-fixed
platelets (Jefferson et a l, 1988; Agarwal et a l, 1989; Greco et a l, 1991; 1992) using
[3 H]-ADP and the antagonist [3 5 S]-ATPaS as radioligands. Formaldehyde-fixed platelets
were used to remove the complications of rapid metabolism of the radioligand by
platelets and isotopic dilution by released ADP (Jefferson et a l, 1988), however, as these
authors themselves state the major drawback of the use of fixed platelets is that the
fixation process will cause chemical modifications of the binding sites which may affect
128
ligand accessibility. Indeed, if enzymes on the platelet surface are modified in such a way
that they no longer possess catalytic activity it is not unreasonable to assume that similar
marked changes will occur in the binding sites of other nucleotide-binding proteins
treated in this manner including the ADP receptor. These binding studies show both high
and low affinity binding sites for ADP and ATPaS which are mutually accessible to these
ligands, platelets showing around 30,000 high affinity sites and 300,000 low affinity sites
with dissociation constants of 30 nM and 3 j l iM for ADP and 3 nM and 200 nM for
ATPaS, respectively. However, the structure-activity relationships for these binding sites
are not entirely consistent with the known pharmacology of the ADP receptor, for
example guanosine and pyrimidine nucleotide analogues bind to these binding sites with
similar affinities to ATPaS whilst being at least 50 fold less potent than this compound as
inhibitors of ADP induced shape change, and indeed the results of the two most recent of
these reports (Greco et a l , 1991; 1992) show that these binding sites are on the
glycoprotein Ilb-IIIa complex. In the 1991 study ATPaS was used in an attempt to
covalently label the ADP receptor. A single 120 kDa polypeptide was labelled and this
was shown to be identical to glycoprotein lib. This labelling was greatly reduced in the
presence of EDTA, which dissociates the Gp Ilb-HIa complex, suggesting that the
nucleotide binding site is formed within the Gp Ilb-IIIa complex. This is supported by the
observation that the photoaffinity reagent 8 -azido-ATP labelled both glycoprotein lib and
m a (Mayinger & Gawaz, 1992). These ADP/ATPaS binding sites are unlikely to
represent the ADP receptor, however, as individuals with Glanzmanris thrombasthenia,
whose platelets lack both of these glycoproteins (Phillips & Agin, 1977), are still able to
respond to ADP (e.g. Zucker et a l 1966; Powling & Hardistry, 1985).
129
A number of other attempts have been made to identify the ADP receptor by covalent
labelling. An attempt was made to use N3-ADP as a photoaffinity reagent for the ADP
receptor and this molecule did label a number of sites on the platelet surface, however,
this labelling was not prevented by ADP (Macfarlane et a l , 1982) suggesting that none of
these proteins represent the ADP receptor. More recently the photoaffinity reagent
2-azidophenylethylthio-ADP was used in an attempt to label the ADP receptor identified
in binding studies with MeSADP and N3-ADP and this compound was found to label a
single 43 kDa polypeptide (Cristalli 8c Mills, 1993). This labelling was prevented by a
number of ADP receptor ligands (both agonists and antagonists) in a manner which was
reasonably consistent with the pharmacology of the ADP receptor although there were
some discrepancies, e.g. ATP was - 1 0 fold better at preventing this labelling than
ATPaS, whilst ATPaS is at least 5 fold better than ATP as an antagonist of responses to
ADP (Cusack & Hourani, 1982b; Chapter 2 of this thesis). This polypeptide is probably
the best candidate so far for the ADP receptor.
A 61 kDa ADP binding protein was extracted from freeze-thawed platelet membranes by
washing with isotonic saline (Adler & Handin, 1979). This protein bound ADP with a
dissociation constant of 0.38 pM and bound ADP was displaced by ADP and ATP with
similar potencies, at least 1000 fold more weakly by AMP and not at all by adenosine
which is consistent with the pharmacology of the ADP receptor. However, it seems
unlikely that a receptor protein would be so easily extracted from the plasma membrane
and, indeed, no further work on this protein has been published. FSBA labels the 100 kDa
protein 'aggregin' (Colman, 1992) but as discussed in the General Introduction this is also
unlikely to represent an ADP receptor.
130
In this study, the binding of [3 5 S]-Sp-ATPaS to intact, washed platelets was investigated
along with the ability of a number of ATP analogues, which are known to act as ADP
receptor antagonists, to displace this binding. Antagonists were used in this study to avoid
the complicating effects associated with the use of agonists (see earlier). Also as the
compounds used in this study were all ATP analogues and the predominant enzyme
activity associated with platelets is nucleoside diphosphokinase (Guccione et a l , 1971),
which phosphorylates ADP to ATP, breakdown of these compounds should not present a
problem in this study. An ATP regenerating system (pyruvate kinase and
phosphoenolpyruvate) was included in the binding experiments to guard further against
ligand breakdown.
5.2 Materials and Methods
5.2.1 Materials
AMPCPP, AMPPCP, Ap5 A, ATP, BSA, CTP, FSBA, GTP, phosphoenolpyruvate,
prostacyclin, pyruvate kinase (type III from rabbit muscle), tetrabutylammonium
hydrogen sulphate and UTP were from Sigma. Cl-ATP, ketanserin, MeSATP, NECA and
yohimbine were obtained from Research Biochemicals Incorporated. Unlabeled ATPaS
was obtained initially from Boehringer Mannheim or subsequently from Calbiochem (for
reasons of cost). [3 5 S]-SP-ATPaS (>1000 Ci/mmol) in 10 mM tricine buffer, pH 7.6,
containing 1 mM dithiothreitol was obtained from NEN. For use in time course
experiments the specific activity was adjusted to 2 Ci/mmol (on the reference date) with
unlabelled ATPaS. For use in displacement experiments the specific activity was
adjusted to 10 Ci/mmol (on the reference date) with unlabelled ATPaS. HPLC solvents
131
and scintillation fluid, Optiphase, were from Fisons Scientific Apparatus. Suramin was a
generous gift from Bayer, U.K. All other chemicals were AnalaR grade from BDH.
Where necessary nucleotides were purified by ion-exchange chromatography as described
in section 2.2.6. Prostacyclin was dissolved at 100 pg/ml in 10 mM NaOH and stored
frozen in aliquots, FSBA was dissolved at 50 mM in dimethylformamide and diluted to
10 mM in water and other drugs were dissolved in water. FSBA, ketanserin, suramin and
yohimbine were made up freshly each day, nucleotides were stored frozen in solution.
5.2.2 Preparation of Washed Platelets
Blood was drawn from healthy human volunteers into one sixth of the volume of ACD
anticoagulant. Donors denied taking aspirin for 10 days before the experiment. PRP was
obtained by centrifuging the blood at 260 g for 20 min. The platelets were isolated from
the PRP by centrifugation (680 g for 20 min) in the presence of prostacyclin (1 pM) and
resuspended in 2 ml of HEPES-saline (see section 2.2.3) containing phosphoenolpyruvate
(100 pM) and pyruvate kinase (5 units/ml) and an aliquot (10 pi) was diluted 50 fold in
1 % ammonium oxalate for counting on a haemocytometer. The platelet suspension was
then made up to between 30 and 60 ml with HEPES-saline containing
phosphoenolpyruvate (100 pM) and pyruvate kinase (5 units/ml). This resulted in a
platelet count between 1.5 and 3.0 x 108 /ml.
5.2.3 Time Course of ATPaS Binding
For association time courses, platelets were incubated at room temperature for various
times in the presence of 1 pM [3 5 S]-ATPaS (2 Ci/mmol). Non-specific binding was
determined in parallel experiments in the presence of 100 pM unlabelled ATPaS. At
132
appropriate times triplicate aliquots (0.5 ml) of platelet suspension were removed and
bound radioactivity was determined as described in section 5.2.5 below.
For dissociation time courses, platelets were incubated at room temperature in the
presence of 1 pM [3 5 S]-ATPaS for 40 min at which point 100 pM unlabelled ATPaS was
added to displace the bound radioligand. Parallel estimates of the non-specific binding
were performed by adding 100 pM unlabelled ATPaS simultaneously with the
radioligand. Bound radioactivity was determined as described in section 5.2.5 below.
5.2.4 Displacement Studies
Aliquots (1.8 ml) of platelet suspension were incubated at room temperature in the
presence of 60 nM [3 5 S]-ATPaS (10 Ci/mmol) and various concentrations of displacing
ligand. After 40 min triplicate 0.5 ml aliquots of suspension were removed and the bound
radioactivity quantified as described in section 5.2.5 below.
Due to the high concentrations of radioligand which would be necessary to saturate the
receptor population and the correspondingly high concentrations of displacing ligand
which would be necessary to define the non-specific binding, saturation experiments
could not be used to determine the binding parameters of ATPaS. The results of an
attempted saturation experiment are shown in figure 5.1. The dissociation constants of
ATPaS and number of ATPaS binding sites were, therefore, determined by Scatchard
analysis of homologous displacement curves. Heterologous displacement curves were
analysed by weighted non-linear curve fitting algorithms using the LIGAND program.
133
s
>oao
*T3C3
'OcoPQ
6 0 0 0 0 -i
5 0 0 0 0 -
3 0 0 0 0 -
20000-
10000-
0 10 20 3 0
[35s]-ATPaS concentration / jjM
Figure 5.1 The binding of ATPaS to platelets using a saturation protocol. The figure shows a representative data for the total (■), non-specific (•) and specific (4) ATPaS binding to washed platelets. Values are the mean of triplicate determinations and the vertical bars show the s.e. mean. Non-specific binding was determined in the presence of 1 mM ATP.
134
5.2.5 Determination of Bound Radioactivity
Aliquots (0.5 ml) of platelet suspension were layered onto silicone oil (4 parts Dow
Coming 704 EU diffusion pump fluid and 1 part Dow Coming 200/0.65 cs silicone fluid)
and centrifuged for 1 min in a swing out microfuge to separate the platelets from the
suspending medium. An aliquot (10 pi) of the supernatant buffer was transferred to a
scintillation vial for counting. The remaining buffer and silicone oil were removed and
the pellets washed three times by gentle immersion in 0.9 % NaCl (to avoid resuspending
the platelets). The tubes were inverted to allow any residual saline to drain. After ~30 min
the tip of the tube, containing the pellet, was cut off and transferred to a scintillation vial.
The bound radioactivity was extracted over night in 5 ml of Optiphase and 0.4 ml of 1 %
triton X-100 after an initial vortex mixing. The tubes were then vortex mixed and the
radioactivity present quantified by scintillation counting. The dpm present in the
supernatant were used to calculate the specific activity of the radioligand used in the
experiment.
5.2.6 HPLC analysis
HPLC analysis of aliquots of the supernatant from the binding assays (performed in the
absence of radioligand) to determine the release of adenine nucleotides from platelets and
the degradation of adenine nucleotide analogues by platelets were performed as described
in Chapter 2 section 2.2.5.
5.3 Results
Binding studies were performed in the absence of added calcium (but in the presence of
135
magnesium) as preliminary studies showed that the binding was stable for a longer period
in its absence (see figure 5.2). The binding of [3 5 S]-ATPaS was essentially at equilibrium
after 30 min incubation and remained stable for at least another 30 min under the assay
conditions (figure 5.2b). Thus an incubation time of 40 min was chosen for the
displacement assays to ensure binding equilibrium had been reached. This binding was
rapidly reversible returning to the level of non-specific binding within 1 0 min of addition
of a large excess of unlabelled ligand (figure 5.2c).
Homologous displacement studies showed that the binding of ATPaS was best described
by a two site binding model (figure 5.3). The level of non-specific binding in the
displacement assays never exceeded 50 % of the total radioactivity bound and never
exceeded 0.08 % of the total radioactivity added to an assay. There were 1668 ±316
(n = 4) high affinity sites with a dissociation constant of 0.21 pM (pKD = 6.69 ± 0.08) and
24184 ± 9988 (n = 4) low affinity sites with a dissociation constant of 32 pM (pKD = 4.50
± 0.28). Binding to the high affinity site was displaced by ATP, MeSATP, Cl-ATP,
AMPCPP, AMPPCP, Ap5A and suramin (figure 5.4) and the values for this
displacement are summarised in table 5.1. ATP, MeSATP & Cl-ATP were also able to
displace ATPaS from the low affinity binding sites and the Kj values for this
displacement are again shown in table 5.1. In the case of AMPCPP, AMPPCP and Ap5 A,
it was not possible to define a K; at the low affinity site as the affinities of these
compounds at this site were too low. Consequently the affinity at the high affinity site for
these compounds was defined using a one site displacement model. The displacement of
[3 5 S]-ATPaS binding by suramin followed a much steeper curve than that of the other
displacing ligands and appeared to reach a plateau at a level greater than that of
136
a2000 -i
1000 -
3 010 20
sO h
W).9.9PQo
<+HooOh
0 0
2000 n
1000 -
0 20 4 0 6 0
c
2000
1000 -
150 5 10
Time/min
Figure 5.2 Kinetics of the binding of [3 5 S]-ATPaS to washed human platelets, (a) Time course of the binding of [3 5 S]-ATPaS to platelets in the presence of 1 mM Ca2+. (b) Time course of the binding of [3 5 S]-ATPaS to human platelets in the absence of added calcium, (c) Time course of the displacement of bound [3 5 S]-ATPaS from platelet by addition of an excess of unlabelled ATPaS. The data are representative of three similar experiments.
137
\- 9
HOO
Unlabelled ATPocS / \iM
b300
200 -<L>
100 -
0 1 3 42
Bound ATPaS / pmol
Figure 5.3 (a) Homologous ATPaS displacement isotherm (oo on the abscissa represents the level of non-specific binding). The data are representative of 5 similar experiments. Values are the mean of triplicate determinations and vertical bars show the s.e. mean, (b) Scatchard plot of the data presented in (a).
3Son
.3T 3.3PQao
10000-
5000-
0 -«0 0.01 0.1 10 100
138
100 -
5 0 -
0 J i------1------ 1------ r 2 * !------ 10.01 0.1 1 10 100 1000
oo<DQ*co
100 -
5 0 -
0 - J1000
100 -
5 0 -
0 | | | | i i0.01 0.1 1 10 100 1000
100 -
5 0 -
0 | | | | | i0.01 0.1 1 10 100 1000
Displacing ligand concentration / jiM
Figure 5.4 Displacement of [3 5 S]-ATPaS by a number of ADP receptor antagonists, (a) ATPaS (•) and Ap5A (o ), (b) ATP (■) and suramin (□), (c) Cl-ATP (A) and MeSATP (A), and (d) AMPCPP (▼) and AMPPCP (v ). Values are the mean of three separate determinations using blood from different donors and vertical bars show the s.e. mean.
139
Ligand n PKH KH(pM) PKL KL(pM)
ATP 5 6.47 ±0.12 0.34 4.07 ±0.31 85.1
MeSATP 3 6.70 ±0.26 0 . 2 4.76 ±0.40 17.6
Cl-ATP 3 7.58 ±0.11 0.03 5.67 ±0.20 2 . 1
AMPCPP 3 5.41 ± 0.08 3.9 n.d. n.d.
AMPPCP 3 3.90 ±0.20 126 n.d. n.d.
ApsA 3 4.01 ±0.11 97 n.d. n.d.
Suramin 3 4.42 ±0.08* 63.6* - -
*Due to the steep displacement curve for suramin pIC5 0 and IC^ values are shown.
Table 5.1 The K* values of ADP receptor antagonists for the displacement of [3 5 S]-ATPaS from high affinity (KH) and low affinity (KL) binding sites on human platelets. Values are the mean ± s.e. mean of three separate determinations on blood from different donors. (n.d. = not determined, - = did not displace).
140
non-specific binding suggesting that it could not displace the radioligand from the low
affinity binding site. Due to this steep displacement curve IC5 0 rather than K; values are
quoted for suramin in table 5.1 as satisfactory analysis of these data by LIGAND to give
values was not possible. The binding of ATPaS was not displaced by NECA (100
pM), ketanserin (1 pM), yohimbine (10 pM), FSBA (100 pM) (relative to an appropriate
solvent control) or UTP (100 pM), but 46.4 ± 1.6 % of the specifically bound radioligand
was displaced by GTP (100 pM) and 11 ±3.7 % was displaced by CTP (100 pM) (see
figure 5.5).
HPLC studies showed that none of the adenine nucleotide analogues used in this study
were detectably broken down under the conditions of these binding assays at 100 pM
(results not shown). However, adenine nucleotides were released by the platelets and this
resulted in approximately 2 pM ATP being present during the assays (figure 5.6). The
amount of ATP present in a given experiment appeared to be constant throughout its
duration as the total binding measured in control assays performed at the beginning and
end of a binding experiment were not significantly different. The presence of this ATP
should not affect the apparent affinity of the displacing ligands determined in these
studies, however, it will result in a marked underestimation of the number of the high
affinity sites available to the radioligand. Using the binding constants obtained here the
presence of 2 pM ATP would result in a 5.5 fold under estimation of the number of high
affinity binding sites suggesting that there may in fact be -9000 high affinity sites. The
estimation of the number of low affinity sites should not be affected due to the much
lower affinity of ATP at this site.
141
Figure 5.5 Displacement of [3 5 S]-ATPaS by dimethylformamide (DMF) (0.5 %), FSBA (100 pM), NECA (100 pM), ketanserin (1 pM), yohimbine (10 pM), UTP (100 pM), CTP (100 pM) and GTP (100 pM). Values are the mean of three separate determinations using blood from different donors and vertical bars show the s.e. mean. (** = p<0.05, ++ = not significantly different from control in the presence of DMF).
142
a*H<
C /5
C /5
Figure 5.6 (a) An HPLC trace showing the ATP released from platelets into the suspending medium under the conditions of a binding experiment, the radioligand being omitted. No ADP peak is seen due to the presence of pyruvate kinase and phosphoenolpyruvate in the medium, (b) HPLC chromatogram showing adenosine, AMP, ADP and ATP standards at a concentration of 2 pM.
143
5.4 Discussion
These results show that ATPaS binds to intact platelets in a reversible and saturable
manner. Scatchard plots of homologous ATPaS displacement isotherms were non-linear
and analysis of these data showed two binding sites, a high affinity binding site with a
dissociation constant of 0.21 pM and a low affinity binding site with a dissociation
constant of 32 pM. Of these two binding sites the high affinity binding site is most likely
to represent an ADP receptor as the affinity of the low affinity site for ATPaS is too low
in comparison with that obtained from functional studies which ranges between 1 and
4.6 pM (Cusack & Hourani, 1982b; Chapter 2 of this thesis). The affinity of ATPaS for
the high affinity site is somewhat higher than that obtained in the functional studies,
however, this may be due to more complete equilibration of the ligand with the receptors
in these binding studies. In the experiments presented here the ATPaS was incubated
with the platelets for 40 min to determine the binding constants whilst in the functional
studies it was added simultaneously with agonist to avoid problems associated with its
breakdown which would result in the formation of compounds with differing
pharmacological properties. Thus in the functional studies the ligand may not have
reached binding equilibrium with the receptor population (although Schild plots with unit
slopes may suggest that it was) and this would result in an underestimation of the affinity
of ATPaS for the receptor in the functional studies. There were -9000 high affinity sites
per platelet when the presence of released ATP was taken into account and this number is
somewhat greater than the 500 - 1200 MeSADP binding sites which have previously been
reported (Macfarlane et al, 1983; Mills et a l, 1985; Cristalli & Mills, 1993). The number
of low affinity binding sites, 24,148 =fc 9,988, is, however, comparable with the number of
144
high affinity ATPaS binding sites found on fixed platelets (31,000; Greco et a l , 1991)
which appear to be associated with Gp Ilb-IIIa and indeed is within the range of the
number of fibrinogen binding sites found on platelets (16,000 - 80,000; Siess, 1989).
The binding of ATPaS was not displaced by high concentrations of NECA (100 juM),
yohimbine (10 pM) or ketanserin (1 pM) showing that it does not represent binding at
adenosine, a 2 or 5-HT2 receptors. The binding was also not displaced by FSBA (100 pM)
suggesting that neither of these binding sites represents 'aggregin' the 100 kDa membrane
protein labelled by this compound (Colman, 1992). Also the binding was only weakly
displaced by other naturally occurring nucleotides, the 46 % displacement achieved by
100 pM GTP being similar in extent to that achieved by 100 pM AMPPCP or Ap5A
(which had Kj values around 100 pM, see table 5.1) and suggesting therefore that GTP
binds at least 300 fold less strongly than ATP to this binding site.
The high affinity binding of ATPaS to platelets was displaced by a number of ADP
receptor antagonists. The correlations between the pK; values obtained in this study and
the pAj values for these analogues in functional studies are shown in figure 5.7. The order
of potency of the ATP analogues for the displacement of high affinity ATPaS binding
was Cl-ATP > MeSATP = ATPaS « ATP > AMPCPP > AMPPCP « Ap5 A, which is most
similar to the potency order of these analogues as antagonists of the inhibition by ADP of
adenylate cyclase, which is Cl-ATP > ATPaS « ATP > Ap5A > AMPPCP (Cusack &
Hourani, 1982b). Indeed a significant (p<0.05) correlation between the pK{ values for
displacing ATPaS and the pA2 values for these analogues as inhibitors of the effects of
ADP on adenylate cyclase was obtained. However, the slope obtained from regression
analysis of these data was -2.5 suggesting that there is not a simple linear relationship
145
7 -
pA2 against Aggregation
'M w) *3 .s^ ■§ S w^ co g bIp!a.
b
8 1
7 -
6 -
5 -
4 - 4 -
pA2 against Increases in [Ca2*^ pA2 against Inhibition of Adenylate Cyclase
Figure 5.7 Comparison of the abilities of a number ATP analogues to displace bound ATPaS from platelets and to inhibit responses to ADP. (a) Comparison with the published pAj values of these compounds against aggregation (Cusack & Hourani, 1982b). (b) Comparison with the pA2 values of these compounds against increases in [Ca2+];. (c) Comparison with the published values of these compounds against inhibition of adenylate cyclase (Cusack & Hourani, 1982b). The lines show the least squares regression fit to the data points. Compounds were: ATP (•), ATPaS (■), Cl-ATP (A), Ap5A (♦) and AMPPCP (▼). For abbreviations, see text.
146
between these parameters. The correlation between these pKj values and the pA, values
for inhibition of aggregation and increases in cytosolic calcium concentration was not
significant. The ability of MeSATP to displace this ATPaS binding is indicative of a
competitive interaction by MeSATP at this ATPaS binding site. Interestingly, whilst
MeSATP and its analogues appear to act as non-competitive antagonists of ADP-induced
aggregation (Cusack & Hourani, 1982a) and increases in cytosolic calcium concentration
(Chapter 2 of this thesis), MeSAMPPCP appears to be a competitive antagonist of the
effects of ADP on adenylate cyclase (Hourani et al., 1986) again showing the
structure-activity relationship of this binding to be more similar to that of inhibition of
adenylate cyclase by ADP. The similar affinity of ATP and ATPaS for the high affinity
binding site is consistent with the effect of these compounds on the binding of MeSADP
to platelets (Cristalli & Mills, 1993). However, the affinities of these compounds
obtained in the present study are somewhat higher than those found by Cristalli & Mills
(1993), in that ATP and ATPaS displaced MeSADP with K; values of around 4 pM
whilst in the present study the Kj values were 0.34 and 0.21 pM, respectively. This
discrepancy may again be due to the rather longer incubation times used in the present
study compared to that used in the MeSADP binding studies (40 min compared to 5 - 10
min). However, along with the rather larger number of high affinity ATPaS binding sites
this discrepancy between the affinities of ATP and ATPaS at these two binding sites may
imply that they are in fact distinct sites. Of course it is possible that in the studies on
MeSADP binding, which were performed in plasma or in buffer in the presence of
EDTA, there was also some ADP released by the platelets and this may have affected the
binding constants determined in this study. However, saturation protocols were used to
determine the binding constants in the MeSADP binding studies, and so the added
MeSADP would have been able to compete with any released ADP and would therefore
still have saturated the entire receptor population at a sufficiently high concentration. The
determination of the receptor density should not, therefore, have been affected. As the
effects of the other antagonists used in this study have not been determined on the
binding of MeSADP to platelets it is not possible to compare these binding sites further.
The number of low affinity ATPaS binding sites is similar to the number of Gp Ilb-IIIa
complexes on platelets (see Siess, 1989) and is also similar to the number of high affinity
ATPaS (and ADP) binding sites found on fixed platelets (Jefferson et a l , 1988; Agarwal
et a l , 1989; Greco et a l, 1991). However, where K; values could be determined in the
present study they are generally higher (i.e. the compounds have lower affinity) than
those found in the studies on fixed platelets and the potency order of these compounds
are different. Whilst on fixed platelets the potency order is ATPaS > ATP > Cl-ATP, that
obtained at the low affinity site in the present study is Cl-ATP > ATPaS > ATP. Whilst it
is possible that these differences may suggest that these are also distinct binding sites and
that the low affinity site described here is in fact not the binding site present on Gp
Ilb-IIIa, it may also reflect changes in the structure of this binding site which may occur
during the fixation process.
Suramin did not appear to displace ATPaS from the low affinity binding site and the
displacement curve for suramin was somewhat steeper than that of the other compounds
used in this study. This steep displacement curve may suggest a degree of positive
cooperativity in the interaction of suramin with this binding site. This could explain the
observed steep Schild plots for the inhibition by suramin of ADP-induced aggregation
and increases in cytosolic calcium concentration as more than one molecule of suramin
148
would have to bind to the ADP receptor in order for a cooperative interaction to be
observed and this is indeed one explanation for a non-unit Schild plot slope. However, the
slope of the Schild plot for inhibition by suramin of the effects of ADP on adenylate
cyclase was not significantly different from unity suggesting a simple competitive
interaction. This could be interpreted as showing that aggregation and inhibition of
adenylate cyclase are indeed mediated by two distinct ADP receptors at which suramin
displays a different binding interaction and that this ATPaS binding site represents an
ADP receptor which induces increases in cytosolic calcium concentration and
aggregation. However, the potency order for displacement by the ATP analogues was
more similar to that for ADP-induced inhibition of adenylate cyclase than the other
responses suggesting that this interpretation is not correct.
The affinities of ATP, Cl-ATP and AMPCPP (as well as that of ATPaS) are somewhat
greater than those reported from functional studies (Macfarlane & Mills, 1975; Cusack &
Hourani, 1982b; Lips et a l, 1980; Chapter 2 of this thesis). However, as for ATPaS this
may be due to the rather longer incubation times used in the binding assays, 40 min
compared to up to 1 min in the functional studies. In the case of AMPPCP and Ap5 A, the
Kj values were rather greater than expected being approximately 100 juM for both
compounds. From functional studies Ap5A would be expected to be equipotent with or
slightly less potent than ATP and AMPPCP should be if anything slightly more potent
than AMPCPP. The relatively low affinity and similar K; values for these compounds may
suggest that they were perhaps being broken down during the assay to a compound which
has veiy little affinity for the ADP receptor. However, at 100 jliM none of the nucleotides
used in this study showed any evidence of being broken down, thus breakdown cannot
149
explain the results obtained with AMPPCP and Ap5 A. The most likely explanation for
these discrepancies would be that the ATPaS is binding to a binding site distinct from the
ADP receptor. The inability of NECA and FSBA to displace this binding is good evidence
that this site does not represent an adenosine receptor or 'aggregin'. A possible candidate
could be nucleoside diphosphokinase (NDPK). This enzyme does show a similar
preference for the naturally occurring nucleotide triphosphates as substrates as the
ATPaS binding site described here does for the displacement of bound ATPaS i.e.
ATP > GTP > CTP (Packham et al., 1974). However, the structure-activity relationships
of this enzyme for the ATP analogues described here are not known and would have to be
determined before this binding site could be confirmed as NDPK.
These results, therefore, show that ATPaS binds to two binding sites on intact washed
platelets, however, neither of these binding sites has properties which are entirely
consistent with those of the ADP receptor. The low affinity site does not bind the ATP
analogues studies with great enough affinity to represent the ADP receptor and is not
displaced by suramin, whilst at the high affinity site the structure-activity relationship is
not really consistent with the functional studies on the ADP receptor.
150
CHAPTER 6
GENERAL DISCUSSION
151
This study has used a number of measures of platelet activation to investigate further the
pharmacology of the human platelet ADP receptor. Studies on the effects of a number of
ADP and ATP analogues on ADP-induced increases in platelet cytosolic calcium
concentration showed that the structure-activity relationship of this response is very
similar to that previously determined for these compounds on ADP-induced aggregation.
This close correlation was further demonstrated in the studies on suramin, a compound
which is structurally unrelated to ADP, which also showed very similar effects on these
two responses to ADP. These observations provide further evidence in support of the
view that this increase in platelet cytosolic calcium concentration is the major signal for
platelet aggregation induced by this agonist. These results also cast further doubt on the
use of the affinity reagent FSBA as a label of a platelet ADP receptor. FSBA has been
shown to inhibit shape change and aggregation induced by ADP, but not to inhibit the
effect of ADP on stimulated adenylate cyclase and this has been taken to imply that shape
change and aggregation are mediated by a receptor distinct from that which causes the
inhibition of stimulated adenylate cyclase (e.g. see Colman, 1992). However, FSBA was
also unable to inhibit increases in platelet cytosolic calcium concentration induced by
ADP (Rao & Kowalski, 1987) and this was taken by these authors to imply that increases
in cytosolic calcium concentration induced by ADP may be induced by another receptor
distinct from that inducing shape change and aggregation. This increase in cytosolic
calcium concentration appears to be the major signal for aggregation induced by ADP
and the studies presented here provide strong evidence that these two responses are
mediated by the same receptor. Thus if FSBA does not inhibit the increase in cytosolic
calcium concentration induced by ADP it cannot be acting at the ADP receptor and must
therefore inhibit ADP-induced aggregation and shape change by a different mechanism.
152
Another piece of evidence which has been taken to support the 'two receptor hypothesis'
was the observation that the non-penetrating sulphydryl reagent pMBS inhibits the effects
of ADP on stimulated adenylate cyclase but does not inhibit platelet shape change
induced by this agonist (Mills & Macfarlane, 1977; Macfarlane & Mills, 1981). However,
in the study by Rao & Kowalski (1987) these authors showed that whilst pMBS did not
inhibit shape change induced by ADP it was able to inhibit increases in cytosolic calcium
concentration induced by this compound, although this inhibition did not reach 1 0 0 % at
the highest concentration of pMBS used (1 mM). In this study 1 mM pMBS inhibited
-90 % of the increase in intracellular calcium induced by 25 juM ADP. However, platelet
shape change is the most sensitive response of platelets to the presence of ADP
suggesting that only a small increase in platelet cytosolic calcium concentration is
required to induce this response (see figure 6.1). Thus, the lack of effect of pMBS on
platelet shape change may simply reflect the fact that in the presence of 1 mM pMBS the
increase in cytosolic calcium concentration induced by ADP is still sufficient to induce
an unaltered shape change. It is also possible that pMBS could inhibit the increase in
platelet cytosolic calcium concentration by modifying a membrane protein other than the
ADP receptor, for example the store regulated calcium channel (Sage et a l, 1990) in
which case it would not affect ADP-induced mobilization of calcium from intracellular
stores. The shape change induced by ADP in EDTA-anticoagulated plasma, which was
used in the study by Rao & Kowalski, is only dependent on this mobilization of calcium
from intracellular stores and would therefore not be affected by agents which only inhibit
the influx of extracellular calcium.
The correlation between the effects of the ADP and ATP analogues on increases in
cytosolic calcium concentration and inhibition of adenylate cyclase was less good,
153
however. As with aggregation the effects of the four ADP analogues tested here on
increases in cytosolic calcium concentration were not in good agreement with their
effects on adenylate cyclase. However, as with the discrepancy between the effects of
these compounds on aggregation and adenylate cyclase, this is explicable in terms of
differences in the efficacies of these compounds at the same receptor for inducing
responses via different effector mechanisms rather than in terms of two receptors
mediating the two responses (e.g. Hourani & Cusack, 1991). When the pA2 values for the
antagonists which appeared to act competitively in this study were compared to those
obtained for their inhibition of the effect of ADP on adenylate cyclase the correlation did
not reach significance at the 5 % level, possibly suggesting that these two effects of ADP
are indeed mediated by separate receptors. However, in my study only 5 compounds were
used to make this comparison compared to the 8 compounds used by Cusack & Hourani
(1982b) to compare the inhibition by ATP analogues of aggregation and the effect of
ADP on adenylate cyclase. Indeed, if the same 5 compounds were used to make the
comparison between aggregation and effects on adenylate cyclase this too is not
significant at the 5 % level. Thus the effects of a larger number of antagonists need to be
studied before the significance of this correlation can be properly evaluated.
The P2-purinoceptor antagonist suramin was also shown to inhibit platelet responses to
ADP and the pA values obtained against each response was similar providing further
evidence in favour of the hypothesis that only a single class of ADP receptor mediates
increases in cytosolic calcium, aggregation and inhibition of adenylate cyclase. The pA2
values obtained for suramin in this study were also similar to those obtained for the
inhibition by suramin of responses to ATP mediated by other P2-purinoceptors (Hoyle et
al, 1990; Inoue et a l, 1991; Leff et a l, 1990; Von Kiigelgen et a l, 1990) showing that
154
suramin cannot distinguish between the various subtypes of P2 -purinoceptor. In the cases
of aggregation and increases in platelet cytosolic calcium concentration this inhibition by
suramin was not simply competitive, however, as the slope of the Schild plot was
significantly greater than unity. Suramin was also able to inhibit non-competitively
aggregation induced by U46619 and by 5-HT (in the presence of adrenaline) and
increases in platelet cytosolic calcium concentration induced by 5-HT showing that
suramin also has non-specific effects on these two platelet responses. As suramin is a
highly charged molecule which does not cross the plasma membrane (Fortes et al., 1973)
this non-specific effect is likely to occur at the cell surface and could be due to inhibition
of the ion-channels involved in calcium influx. A non-specific effect of suramin on
platelet calcium influx pathways would explain both the steep Schild plots obtained for
aggregation and increases in cytosolic calcium and the unit Schild plot obtained for
inhibition of the effect of ADP on adenylate cyclase. If suramin were blocking the influx
of extracellular calcium as well as having a competitive inhibitory effect at the ADP
receptor this would steepen the slope of the Schild plot for inhibition of aggregation and
increases in cytosolic calcium concentration, as these responses are dependent on this
influx of calcium. However, it would not affect the slope of the Schild plot for the
inhibition of the effect of ADP on adenylate cyclase as this response is not dependent on
an increase in intracellular calcium concentration (see Chapter 2). It might also be
expected that such a non-competitive effect of suramin would decrease the maximal
response to ADP, however this would depend on the relative ability of suramin to
produce these inhibitory effects. The hypothesis that suramin is able to inhibit
non-specifically agonist induced calcium influx could be tested by studying the effect of
suramin on increases in platelet cytosolic calcium concentration in the absence of
155
extracellular calcium. This would effectively negate the contribution to the inhibition that
any effects of suramin on calcium influx may have by removing this influx.
Studies on the effects of varying the extracellular calcium concentration on platelet
responses to ADP and their inhibition by ATP suggested that, in common with other
P2-purinoceptors (Cockcroft & Gomperts, 1979; Fedan et a l, 1990; Motte et a l, 1993),
the platelet ADP receptor is in fact a receptor for ADP3* and that the active antagonist is
A tp4-. Thus it would appear that P2-purinoceptors possess a binding site for ATP4* rather
than for Mg.ATP which is the ligand required by phosphotransferase enzymes. The
magnesium ion seems to be required by the phosphotransferase enzymes to produce the
correct conformation of the triphosphate chain for the phosphoryl transfer reaction to
occur (Eckstein, 1985; Frey, 1992). Presumably therefore the purinoceptors require a
conformation of the phosphate chain which is not accessible to the magnesium complex.
These results have important implications for the work in the rest of this thesis,
particularly the work on ADP-induced increases in cytosolic calcium concentration and
the binding studies and the comparison of these results with the results of previous
studies. The various adenine nucleotide analogues used in these studies have different
affinities for divalent cations than the parent nucleotide, for example AMPPCP has ~3
fold greater affinity for magnesium than ATP (Yount et a l, 1971) whilst
phosphorothioate analogues generally have lower affinities for magnesium than the
parent nucleotide (Pecoraro et a l, 1984). Thus, in the presence of divalent cations the
observed effects of these compounds will depend upon the stability constants of their
divalent cation complexes as this will affect the concentration of the analogue present in
the uncomplexed form. Consequently to gain a true comparison between the effects of
156
these compounds on the various platelet responses to ADP it is necessary to measure the
responses in the absence of divalent cations or at least to take into account their affinities
for these cations. As platelet aggregation cannot be studied in the absence of divalent
cations a knowledge of the relative affinities of the nucleotides studied for divalent
cations would appear to be essential for a truly meaningful comparison of their effects to
be made. Importantly, the studies on increases in intracellular calcium presented in this
thesis were performed in the presence of 1 mM calcium and 1 mM magnesium, whilst the
binding studies were performed in the absence of calcium. The absence of calcium will
alter the observed affinity of these various compounds to different extents depending on
their affinities for this metal ion. Similarly previous functional studies (e.g. Cusack &
Hourani, 1981a, b; 1982a, b, c) were performed in citrated platelet-rich plasma, this not
only contains enzymes capable of degrading adenine nucleotide analogues but also
contains a lower concentration of divalent cations than that present in my studies on
increases in cytosolic calcium concentration.
[3 5 S]-ATPaS was used as a radioligand in an attempt to determine the number and
binding parameters of the platelet ADP receptor. Scatchard plots for this binding were
non-linear and analysis of the data using the LIGAND program showed two classes of
binding site. However, the neither of these binding sites had properties which were fully
consistent with those of the ADP receptor determined in functional studies. In addition
the release of adenine nucleotides by the platelets resulted in a marked underestimation
of the number of high affinity binding sites for the radiolabel. Thus it would seem that
washed platelets are not an ideal system for studying the binding of adenine nucleotides
to platelets. However, it may be possible to improve this binding protocol (see below).
157
6.1 Suggestions for Further Work
In light of the relatively poor correlation between the pA , values of ADP receptor
antagonists on increases in intracellular calcium concentration and on inhibition of
adenylate cyclase this is an obvious area for further study. This could be done either by
determining the effects of a greater number of antagonists as used in the study by Cusack
& Hourani (1982b) on increases in cytosolic calcium concentration or by determining the
effects of AMPCPP and ATPpS on inhibition of adenylate cyclase. The compounds
which were used in the study of Cusack & Hourani which were not used in the studies
presented here were Rp-ATPaS, 2-chloroadenosine 5'-phosphorothioate and adenosine
5'-0-(3-fluorotriphosphate). As these three compounds are not commercially available
and cannot be synthesized enzymatically the most convenient approach would probably
be to study the effects of ATPJ3S and AMPCPP on inhibition of adenylate cyclase.
From the studies of Sage & co-workers on the kinetics of increases in cytosolic calcium
induced by ADP (see section 1.4.2 of the General Introduction) and from my work on
suramin (see Chapter 3) it is possible that there are two ADP receptors mediating
increases in platelet cytosolic calcium concentration, a ligand-gated ion-channel
mediating influx and a G-protein coupled receptor causing the release of calcium from
intracellular stores. It would therefore be interesting to study the structure-activity
relationships associated with increases in platelet cytosolic calcium concentration in the
absence of extracellular calcium. This would allow the mobilization of intracellular
calcium to be studied in the absence of influx, and would allow further studies into the
effects of extracellular divalent cations on responses to ADP. It would also be interesting
determine the effects of pMBS on the response to ADP under these conditions to see if
158
this compound does show selective effects on calcium influx over the release of calcium
from intracellular stores. It may be more difficult to study influx in the absence of release
from intracellular stores, however, Sage and co-workers have shown that the initial influx
of calcium induced by ADP is not inhibited by cAMP (Sage & Rink, 1987). It may
therefore be possible to study this influx alone in the presence of PGEj or adenosine at
concentrations which inhibit the increase in cytosolic calcium concentration in the
presence of EGTA. As stated earlier in this thesis (see the Discussion of Chapter 3)
studying the effects of MeSATP and ADPaS on these two components of the increase in
cytosolic calcium concentration separately may also be able to shed light on their
mechanisms of action.
It may also be possible to overcome some of the drawbacks of the binding studies
described here if the binding were studied in plasma anticoagulated with EDTA as was
done by Macfarlane et al (1982; 1983). The use of an antagonist ligand will remove the
possible complicating effects (e.g. difficulty in estimating the extracellular space)
associated with platelet shape change which will occur when ADP analogues are used as
radioligands. In the presence of EDTA the concentration of magnesium is extremely low
and is not sufficient for the activity of nucleotidases thus removing the problems
associated with nucleotide metabolism. The removal of magnesium may also prevent the
binding of the NDPK as this enzyme requires the magnesium complex of ATP as a
substrate. Also, in the presence of EDTA glycoprotein Ilb-IIIa will be dissociated and the
nucleotide binding site on this molecule may therefore not be accessible. Indeed in the
presence of EDTA photolabelling of glycoprotein Ilb-IIIa by ATPaS was markedly
reduced (Greco et a l , 1991). Finally, in the absence of divalent cations the adenine
159
nucleotides will be present predominantly as the uncomplexed species which seems to be
the form recognised by the ADP receptor. It would still of course be necessary to
determine whether the platelets were releasing adenine nucleotides under these
conditions.
pa oC ZA• 3 p
S
2 2<aZAPOo.CO&<+-!o
p<DOPOUPLh
Q<
100-1
5 0 -
0-110001001010.10.01
ADP concentration / jiM
Figure 6.1 A comparison of the ability of ADP to induced shape change (•) (n = 3), increases in cytosolic calcium concentration (■) (n = 34) and aggregation (A) (n = 10) of washed human platelets. The data are presented as a percentage of the response to the highest concentration of ADP used (30 pM for shape change, 100 pM for increases in cytosolic calcium concentration and 300 pM for aggregation). Values are the mean of at least 3 separate determinations (see brackets) and vertical bars show the s.e. mean.
160
REFERENCES
161
ADLER, J. R. & HANDIN, R. I. (1979) Solubilization and characterization of a platelet membrane ADP-binding protein. J. Biol. Chem. 254, 3866-3872.
AGARWAL, A. K., TANDON, N. N., GRECO, N. J., CUSACK, N. J. & JAMIESON, G. A. (1989) Evaluation of the binding to fixed human platelets of agonists and antagonists of ADP-induced aggregation. Thrombos. Haemostas. 62,1102-1106.
AKKERMAN, J. W. N., GORTER, G. & KLOPROGGE, E. (1982) Kinetic analysis of a-granule secretion by platelets. A methodological report. Thrombos. Res. 27, 59-64.
AKTORJDES, K. & JAKOBS, K. H. (1985) Regulation of platelet cyclic AMP formation. In The Platelets: Physiology and Pharmacology (Longenecker, G., ed), pp. 271-288, Academic Press, Florida.
ALONSO, M. T., ALVAREZ, J., MONTERO, M., SANCHEZ, A. & GARCIA-SANCHO, J. (1991) Agonist-induced Ca2+ influx into human platelets is secondary to the emptying of intracellular Ca2+ stores. Biochem. J. 280, 783-789.
BALDUINI, C. L., BERTOLINO, G., NORIS, P., SININGAGLIA, F., BISIO, A. & TORTI, M. (1988) Interrelation of platelet aggregation, release reaction and thromboxane A production. Biochem. Biophys. Res. Commun. 156, 823-829.
BAUVOIS, B., LEGRAND, C. & CAEN, J. P. (1980) Interaction of adenosine and adenylnucleotides with the human platelet membrane. Further characterization of the ADP binding sites. Haemostasis 9, 92-104.
BEHNKE, O. (1970) Microtubules in disc shaped blood cells. Int. Rev. Exp. Pathol. 9,1-92.
BENNETT, J. S., COLMAN, R. F. & COLMAN, R. W. (1978) Identification of adenine nucleotide binding proteins in human platelet membranes by affinity labelling with 5-/?-fluorosuIfonylbenzoyl adenosine. J. Biol. Chem. 253, 7346-7354.
BENNETT, J. S. & VILAIRE, G. (1979) Exposure of platelet fibrinogen receptors by ADP and epinephrine. J. Clin. Invest. 64,1393-1401.
BENTFIELD, M. E. & BAINTON, D. E. (1975) Cytochemical localization of lysosomal enzymes in rat megakaryocytes and platelets. J. Clin. Invest. 56,1635-1649.
BORN, G. V. R. (1962a) Quantitative investigations into the aggregation of blood platelets. J. Physiol. 162, 67-68 (Abstract).
BORN, G. V. R. (1962b) Aggregation of blood platelets by adenosine diphosphate and its reversal. Nature 194,927-929.
BORN, G. V. R. (1965) Uptake of adenosine and adenosine diphosphate by human blood platelets. Nature 206, 1121-1122.
162
BORN, G. V. R. & FEINBERG, H. (1975) Binding of adenosine diphosphate to intact human platelets. J. Physiol 251, 803-816.
BRODER, S. YARCHOAN, R., COLLINS, J. M., LANE, H. C., MARKHAM, P. D., KLECKER, R. W., REDFIELD, R. R., MITSUYA, H., HOTH, D. F., GELMANN, E., GROOPMEN, J. E., RESNICK, L., GALLO, R. C., MYERS, C. E. & FAUCI, A. S. (1985) Effects of suramin on HTLV-III/LAV infection presenting Kaposi's sarcoma or AIDS-related complex: clinical pharmacology and suppression of vims replication in vivo. Lancet ii, 627-630.
BRUNS, R. F., LU, G. H. & PUGSLEY, T. A. (1986) Characterisation of the A receptor labelled by [3 H]-NECA in rat striatal membranes. Molec. Pharmacol 29, 331-346.
BURNSTOCK, G. (1978) A basis for distinguishing two types of purinergic receptor. In Cell and Membrane Receptors for Drugs and Hormones: A Multidisciplinary Approach (Bolis, L. & Straub, R. W., eds), pp. 107-118, Raven Press, New York.
BURNSTOCK, G. (1989) Purine receptors. In Adenosine Receptors in the Nervous System (Ribeiro, J. A., ed), pp. 1-14, Taylor & Francis, London.
BURNSTOCK, G. & KENNEDY, C. (1985) Is there a basis for distinguishing two types of P2 -purinoceptor? Gen. Pharmacol 16,433-440.
BUTLER, S. J., KELLY, C. H. McKENZIE, F. R., GUILD, S. B., WAKELAM, M. J. O. MILLIGAN, G. (1988) Differential effects of suramin on the coupling of receptors to individual species of pertussis-toxin-sensitive guanine-nucleotide-binding proteins. Biochem. J. 251,201-205.
CALCATERRA, N. B., VICARIO, L. R. & ROVERI, O. A. (1988) Inhibition by suramin of mitochondrial ATP synthesis. Biochem. Pharmacol 37, 2521-2527.
CARTY, D. J., FREAS, D. L. & GEAR, A. R. L. (1987) ADP causes subsecond changes in protein phosphorylation of platelets. Blood 70, 511-515.
CATTANEO, M., AKKAWAT, B., LECCHI, A., CIMMINIELLO, C., CAPITANIO, A. M. & MANNUCCI, P. M. (1991) Ticlopidine selectively inhibits human platelet responses to adenosine diphosphate. Thrombos. Haemostas. 6 6 , 694-699.
CATTANEO, M., LECCHI, A., RANDI, A. M., McGREGOR, J. L. & MANNUCCI, P. M. (1992) Identification of a new congenital defect of platelet function characterized by severe impairment of platelet responses to adenosine diphosphate. Blood 80,2787-2796.
CHAMBERS, D. A., SALZMAN, E. W. & NERI, L. L. (1968) Platelet nucleotide metabolism and platelet aggregation. Exp. Biol. Med. 3,62-70.
CHARO, I. F., FEINMAN, R. D. & DETWILER, T. C. (1977) Interrelations of platelet aggregation and secretion. J. Clin. Invest. 60, 866-873.
163
COCKCROFT, S. & GOMPERTS, B. D. (1979) Activation and inhibition of calcium-dependent histamine secretion by ATP ions applied to rat mast cells. J. Physiol. 296, 229-243.
COHEN, I., GERRARD, J. M., BERGMAN, R. N. & WHITE, J. G. (1978) The role of contractile filaments in platelet activation. In Protides o f Biological Fluids (Peeters, H., ed), pp. 555-566, Pergamon, Oxford.
COLE, B., ROBISON, A. & HARTMANN, R. C. (1971) Studies on the role of cyclic AMP in platelet function. Ann. N. Y. Acad. Sci. 185,477-487.
COLLINS, J. M., KLECKER, R. W., YARCHOAN, R., LANE, H. C., FAUCI, A. S., REDFIELD, R. R., BRODER, S. & MYERS, C. E. (1986) Clinical pharmacokinetics of suramin in patients with HTLV-IH/LAV infections. J. Clin. Pharmacol. 26,22-26.
COLLIS, M. G. & HOURANI, S. M. O. (1993) Adenosine receptor sub-types. Trends Pharmacol. Sci. 14, 360-366.
COLMAN, R. W. (1990) Aggregin: a platelet ADP receptor that mediates activation. FASEBJ. 4, 1425-1435.
COLMAN, R. W. (1992) Platelet ADP receptors stimulating shape change and inhibiting adenylate cyclase. News Pharmacol. Sci. 7, 274-278.
COLMAN, R. W., FIGURES, W. R., COLMAN, R. F., MORINELLI, T. A., NIEWIAROWSKI, S. & MILLS, D. C. B. (1980) Identification of two distinct adenosine diphosphate receptors in human platelets. Trans. Assoc. Am. Physicians 93,305-316.
COLMAN, R. W. FIGURES, W. R., WU, Q.-X., CHUNG, S. Y., MORINELLI, T. A., TUSZYNSKI, G. P., COLMAN, R. F. & NIEWIAROWSKI, S. (1988) Distinction between glycoprotein Ilia and the 100-kDa membrane protein (aggregin) mediating ADP-induced platelet activation. Arch. Biochem. Biophys. 262,298-306.
CONNOLLY, T. M. & LIMBIRD, L. E. (1983) Removal of extraplatelet Na+ eliminates indomethacin-sensitive secretion from human platelets stimulated by epinephrine, ADP and thrombin. Proc. Natl. Acad. Sci. USA 80, 5320-5324.
CRAWFORD, N. & SCRUTTON, M. C. (1987) Biochemistry of blood platelets. In Haemostasis and Thrombosis (Bloom, A. L. Sc Thomas, D. P., eds), pp. 47-77, Churchill Livingstone, Edinburgh.
CRISTALLI, G. Sc MILLS, D. C. B. (1993) Identification of a receptor for ADP on blood platelets by photoaffinity labelling. Biochem. J. 291, 875-881.
CUSACK, N. J. Sc BORN, G. V. R. (1977) Effects of photolysable 2-azido analogues of adenosine, AMP and ADP on human platelets. Proc. R. Soc. Lond. B 197, 515-520.
164
CUSACK, N. J., HICKMAN, M. E. & BORN, G. V. R. (1979) Effects of D- and L- enantiomers of adenosine, AMP and ADP and their 2-chloro- and 2-azido- analogues on human platelets. Proc. R. Soc. Lond. B 206, 139-144.
CUSACK, N. J. & HOURANI, S. M. O. (1981a) Partial agonist behaviour of adenosine 5'-0-(2-thiodiphosphate) on human platelets. Br. J. Pharmacol 73,405-408.
CUSACK, N. J. & HOURANI, S. M. O. (1981b) Effects of RP and SP diastereoisomers of adenosine 5’-0-(l-thiodiphosphate) on human platelets. Br. J. Pharmacol. 73,409-412.
CUSACK, N. J. & HOURANI, S. M. O. (1982a) Specific but non-competitive inhibition by 2-alkylthio analogues of adenosine 5-monophosphate and adenosine 5 -triphosphate of human platelet aggregation induced by adenosine 5'-diphosphate. Br. J. Pharmacol. 75, 397-400.
CUSACK, N. J. & HOURANI, S. M. O. (1982b) Adenosine 5-diphosphate antagonists and human platelets: no evidence that aggregation and inhibition of adenylate cyclase are mediated by different receptors. Br. J. Pharmacol. 76,221-227.
CUSACK, N. J. & HOURANI, S. M. O. (1982c) Competitive inhibition by adenosine 5-triphosphate of the actions on human platelets of 2-chloroadenosine 5-diphosphate, 2-azidoadenosine 5-diphosphate and 2-methylthioadenosine 5-diphosphate. Br. J. Pharmacol. 77,329-333.
CUSACK, N. J. & HOURANI, S. M. O. (1990) Subtypes of P2 -purinoceptors: studies using analogues ATP. Ann. N. Y. Acad. Sci. 603,172-181.
CUSACK, N. J., HOURANI, S. M. O & WELFORD, L. A. (1985) Characterisation of ADP receptors. Adv. Exp. Med. Biol. 192,29-39.
CUSACK, N. J. & PETTEY, C. J. (1988) Effects of isopolar, isosteric phosphonate analogues of adenosine 5-diphosphate (ADP) on human platelets. In Adenosine and Adenine Nucleotides: Physiology & Pharmacology (Paton. D. M., ed), p. 287, Taylor & Francis, London.
CUTLER, L., RODAN, G. & FEINSTEIN, M. B. (1978) Cytochemical localization of adenylate cyclase and of calcium ion, magnesium ion-activated ATPases in the dense tubular system of human blood platelets. Biochim. Biophys. Acta 542,357-371.
DABROWSKA, R. & HARTSHORNE, D. J. (1978) A Ca2+ and modulator-dependent myosin light chain kinase from non-muscle cells. Biochem. Biophys. Res. Commun. 85, 1352-1359.
DAHLQUIST, R. & DIAMANT, B. (1974) Interaction of ATP and calcium on the rat mast cell: effect on histamine release. Acta Pharmacol. Toxicol. 34, 368-384.
DANIEL, J. L., DANGELMAIER, C. A., SELAK, M. & SMITH, J. B. (1986) ADP stimulates TP3 formation in human platelets. FEBSLett. 206,299-303.
165
DANIEL, J. L., MOLISH, I. R. & HOLMSEN, H. (1981) Myosin phosphorylation in intact platelets. J. Biol. Chem. 256, 7510-7515.
DANIEL, J. L., MOLISH, I. R., RIGMAIDEN, M. & STEWART, G. (1984) Evidence for a role of myosin phosphorylation in the initiation of the platelet shape change response. J. Biol. Chem. 259, 9826-9831.
DEN HERTOG, A., NELEMANS, A. & VAN DEN AKKER, J. (1989a) The inhibitory action of suramin on the P2-purinoceptor in smooth muscle cells of guinea-pig taenia caeci. Eur. J. Pharmacol. 166, 531-534.
DEN HERTOG, A., VAN DEN AKKER, J. & NELEMANS, A. (1989b) Suramin and inhibitory junction potential in taenia caeci of the guinea-pig. Eur. J. Pharmacol 173, 207-209.
DIONISOTTI, S., ZOCCHI, C., VARANI, K., BOREA, P. A. & ONGINI, E. (1992) Effects of adenosine derivatives in human and rabbit platelet aggregation. Correlation of adenosine receptor affinities and antiaggregatory activity. Naunyn-Scmiedeberg's Arch. Pharmacol. 346, 673-676.
DRUMMOND, A. H. & MACINTYRE, D. E. (1987) Platelet inositol lipid metabolism and calcium influx. In Platelets in Biology and Pathology III. (MacIntyre, D. E. & Gordon, J. L., eds), pp. 373-431, Elsevier, Amsterdam.
DUNN, P. M. & BLAKELEY, A. G. H. (1988) Suramin: a reversible P2-purinoceptor antagonist in the mouse vas deferens. Br. J. Pharmacol. 93,243-245.
ECKSTEIN, F. (1985) Nucleoside Phosphorothioates. Annu. Rev. Biochem. 54,367-402.
ECKSTEIN, F. & GOODY, R.S. (1976) Synthesis and properties of diastereoisomers of adenosine 5-(0-l-thiotriphosphate) and adenosine 5'-(0-2-thiotriphosphate). Biochemistry 15,1685-1691.
EHRMANN, M., TOTH, E. & FROJMOVIC, M. (1978) A platelet procoagulant activity associated with platelet shape change. J. Lab. Clin. Med. 92,393-401.
EMMS, H. & LEWIS, G. P. (1986) The roles of prostaglandin endoperoxides, thromboxane A2 and adenosine diphosphate in collagen-induced aggregation in man and the rat. Br. J. Pharmacol. 87,109-115.
FEDAN, J. S., DAGIRMANJIAN, J. P., ATTFEILD, M. D. & CHEDECKEL, E. W.(1990) Evidence that the P2X purinoceptor of the smooth muscle of the guinea pig vas deferens is an ATP4- receptor. J. Pharmacol. Exp. Ther. 255,46-51.
FISHER, G. J., BAKSHIAN, S. & BALDASSARE, J. J. (1985) Activation of human platelets by ADP causes a rapid rise in cytosolic free calcium without hydrolysis of phosphatidylinositol-4,5-bisphosphate. Biochem. Biophys. Res. Commun. 129, 958-964.
166
FORTES, P. A. G., ELLORY, J. C. & LEW, V. L. (1973) Suramin: a potent ATPase inhibitor which acts on the inside of the sodium pump. Biochim. Biophys. Acta 318, 262-277.
FOSKETT, J. K., ROIFMAN, C. M. & WONG, D. (1991a) Activation of calcium oscillations by thapsigargin in parotid acinar cells. J. Biol Chem. 266, 2778-2782.
FOSKETT, J. K. & WONG, D. (1991b) Free cytoplasmic Ca2+ concentration oscillations in thapsigargin-treated parotid acinar cells are caffeine- and ryanodine-sensitive. J. Biol Chem. 266, 14535-14538.
FOX, J. E. B. & PHILLIPS, D. R. (1982) Role of phosphorylation in mediating the association of myosin with cytoskeletal structures of human platelets. J. Biol. Chem. 257, 4120-4126.
FREY, P. A. (1992) Nucleotidyltransferases and phosphotransferases: stereochemistry and covalent intermediates. In The Enzymes, Vol. 20, pp. 141-186, Academic Press, New York.
FUKAMI, M. H., DANGLEMAIER, C. A., BAUER, J. S. & HOLMSEN, H. (1980) Secretion, subcellular localization and metabolic status of inorganic pyrophosphate in human platelets. Biochem. J. 192, 99-105.
FUNDER, J., HERSHCO, L., ROTHSTEIN, A. & LIVNE, A. (1988) Na+/H+ exchange and aggregation of human platelets activated by ADP: the exchange is not required for aggregation. Biochem. Biophys. Acta 938,425-433.
GAARDER, A., JONSEN, J., LALAND, S., HELLEM, A. & OWREN, P. A. (1961) Adenosine diphosphate in red cells as a factor in the adhesiveness of human blood platelets. Nature 192, 531-532.
GAARDER, J. & LALAND, S. (1964) Hypothesis for the aggregation of platelets by nucleotides. Nature 202, 909-910.
GACHET, C. & CAZENAVE, J. P. (1991) ADP induced blood platelet activation: a review. Nouv. Rev. Fr. Haematol. 33,347-358.
GEORGE, J. N. & NURDEN, A. T. (1987) Inherited disorders of the platelet membrane: Glanzmann’s thrombasthenia and Bemard-Souiier syndrome. In Hemostasis and Thrombosis (Colman, R. W., Hirsh, J., Marder, V. J. & Salzman, E. W., eds), pp. 726-749, Lippincott, London.
GORDON, J. L. (1986) Extracellular ATP: effects, sources and fate. Biochem. J. 233, 309-319.
GOUGH, G., MAGUIRE, M. H. & MICHAL, F. (1969) 2-chloroadenosine 5'-phosphate and 2-chloroadenosine 5'-diphosphate, pharmacologically active nucleotide analogs. J. Med. Chem. 12,494-498.
167
GOUGH, G., MAGUIRE, H. & PENGLIS, F. (1972) Analogues of adenosine 5’-diphosphate - new platelet aggregators. Influence of the purine ring and phosphate chain on the platelet-aggregating potency of adenosine 5'-diphosphate. Molec. Pharmacol 8,170-177.
GOUGH, G. R., NOBBS, D. M., MIDDLETON, J. C., PENGLIS-CAREDES, F. & MAGUIRE, M. H. (1978) New inhibitors of platelet aggregation. 5’-phosphate, 5'-phosphorothioate and 5-O-sulfamoyl derivatives of 2-substituted adenosine analogues. J. Med Chem. 21, 520-525.
GRECO, N. J., YAMAMOTO, N., JACKSON, B. W., TANDON, N. N., MOOS, M. & JAMIESON, G. A. (1991) Identification of a nucleotide-binding site on glycoprotein lib. Relationship to ADP-induced platelet activation. J. Biol Chem. 266,13627-13633.
GRECO, N. J., TANDON, N. N., JACKSON, B. W. & JAMIESON, G. A. (1992) Low structural specificity for nucleoside triphosphates as antagonists of ADP-induced platelet activation. J. Biol. Chem. 267, 2966-2970.
GRYNKIEWICZ, G., POENIE, M. & TSIEN, R. Y. (1985) A new generation of Ca2+ indicators with greatly improved fluorescence properties. J. Biol. Chem. 260, 3440-3450.
GUCCIONE, M. A., PACKHAM, M. A., KINLOUGH-RATHBONE, R. L. & MUSTARD, J. F. (1971) Reactions of I4 C-ADP and 1 4 C-ATP with washed platelets from rabbits. Blood 37, 542-555.
HALLAM, T. J. & RINK, T. J. (1985a) Responses to adenosine diphosphate in human platelets loaded with the fluorescent calcium indicator Quin2. J. Physiol. 368,131-146.
HALLAM, T. J. & RINK, T. J. (1985b) Agonists stimulate divalent cation channels in the plasma membrane of human platelets. FEBSLett. 186,175-179.
HARDISTRY, & CAEN, J. P. (1987) Disorders of platelet function. In Haemostasis and Thrombosis (Bloom, A. L. & Thomas, D. P., eds), pp. 365-392, Churchill Livingstone, Edinburgh.
HARRISON, M. J., BROSSMER, R. & GOODY, R. S. (1975) Inhibition of platelet aggregation and the platelet release reaction by a,© diadenosine polyphosphates. FEBS Lett. 54, 57-60.
HARRISON, P. & CRAMER, E. M. (1993) Platelet a-granules. Blood Reviews 7,52-62.
HARTWIG, J. H. (1992) Mechanisms of actin rearrangements mediating platelet activation. J. Cell Biol. 118,1421-1442.
HASLAM, R. J. (1973) Interactions of the pharmacological receptors of blood platelets with adenylate cyclase. Series Haematologica 6,333-350.
168
HASLAM, R. J. & CUSACK, N. J. (1981) Blood platelet receptors for ADP and for adenosine. In Purinergic Receptors: Receptors and Recognition Series B, (Bumstock, G., ed), vol. 12, pp. 221-285, Chapman & Hall, London.
HASLAM, R. J., DAVIDSON, M. M. L. & DESJARDINS, J. V. (1978) Inhibition of adenylate cyclase by adenosine analogues in preparations of broken and intact human platelets. Evidence for the unidirectional control of platelet function by cyclic AMP. Biochem. J. 176, 83-95.
HASLAM, R. J. & LYNHAM, J. A. (1977) Relationship between phosphorylation of blood platelet proteins and secretion of platelet granule constituents. I. Effects of different aggregating agents. Biochem. Biophys. Res. Commun. 77, 714-722.
HASLAM, R. J. & ROSSON, G. M. (1975) Effects of adenosine on levels of adenosine cyclic 3 ’,5 '-monophosphate in human blood platelets in relation to adenosine incorporation and platelet aggregation. Molec. Pharmacol. 11, 528-544.
HATHAWAY, D. R. & ADELSTEIN, R. S. (1979) Human platelet myosin light chain kinase requires the calcium-binding protein calmodulin for activity. Proc. Natl. Acad. Sci. 16, 1653-1657.
HEEMSKERK, J. W. M., HOYLAND, J., MASON, J. & SAGE, S. O. (1992) Spiking in cytosolic calcium concentration in single fibrinogen-bound fura-2 -loaded human platelets. Biochem. J. 283,379-383.
HEEMSKERK, J. W. M., VIS, P., FEIJGE, M. A. R , HOYLAND, J., MASON, W. T. & SAGE, S. O. (1993) Roles of phospholipase C and Ca2+-ATPase in calcium responses of single, fibrinogen-bound platelets. J. Biol. Chem. 268, 356-363.
HENSLER, M. E., FROJMOVIC, M., TAYLOR, R. G., HANTGAN, R. R. & LEWIS, J. C. (1992) Platelet morphological changes and fibrinogen receptor localization. Initial responses in ADP-activated human platelets. Am. J. Pathol. 141, 707-719.
HESKETH, T. R , SMITH, G. A., MOORE, J. P., TAYLOR, M. V. & METCALF, J. C. (1983) Free cytoplasmic calcium concentration and the mitogenic stimulation of lymphocytes. J. Biol. Chem. 258,4876^4882.
HIRATA, Y., TAKIGUCHI, Y., WADA, K., MATSUNO, H., UMEMURA, K., UEMATSU, T. & NAKASfUMA, M. (1993) Roles of platelet-activating factor, thromboxane A^ ADP and thrombin in thrombogenesis in the guinea pig. Eur. J. Pharmacol. 231,421-425.
HOITING, B. MOLLEMAN, A., NELEMANS, A. & DEN HERTOG, A. (1990) P2 -purinoceptor-activated membrane currents and inositol tetrakisphosphate formation are blocked by suramin. Eur. J. Pharmacol. 181,127-131.
HOLMSEN, H. (1985) Nucleotide metabolism of platelets. Annu. Rev. Biochem. 47, 677-690.
169
HORAk, H. & BARTON, P. G. (1974) Effects of a,P-methylene-adenosine- 5 -diphosphate on blood platelet aggregation. Biochim. Biophys. Acta 373,471-480.
HOURANI, S. M. O. & CHOWN, J.A. (1989) The effects of some possible inhibitors of ectonucleotidases on the breakdown and pharmacological effects of ATP in the guinea-pig urinary bladder. Gen. Pharmacol. 20,413-416.
HOURANI, S. M. O. & CUSACK, N. J. (1991) Pharmacological receptors on blood platelets. Pharmacol Rev. 43,243-298.
HOURANI, S. M. O., WELFORD, L. A. & CUSACK, N. J. (1986) 2-MeS-AMP-PCP and human platelets: implications for the role of adenylate cyclase in ADP-induced aggregation? Br. J. Pharmacol. 87, 84P (Abstract).
HOYLE, C. H. V., KNIGHT, G. E. & BURNSTOCK, G. (1990) Suramin antagonizes responses to P2-purinoceptor agonists and purinergic nerve stimulation in the guinea-pig urinary bladder and taenia coli. Br. J. Pharmacol. 99, 617-621.
INOUE, K., NAKAZAWA, K., OHARA-IMAIZUMI, M., ONAMA, T., FUJIMORI, K. & TAKANAKA, A. (1991) Selective and competitive antagonism by suramin of ATP-stimulated catecholamine secretion from PC 12 phaeochromocytoma cells. Br. J. Pharmacol. 102, 581-584.
JAFFE, E. K. & COHN, M. (1978) 31P nuclear magnetic resonance spectra of the thiophosphate analogues of adenine nucleotides; effects of pH and Mg2+ binding. Biochemistry 17, 652-657.
JARVIS, M. F., SCHULTS, R., HUTCHISON, A. J., DO, U. H., SILLS, M. A. & WILLIAMS, M. (1989) [3H]-CGS21680, a selective A^receptor ligand directly labels Aj-receptors in rat brain. J. Pharmacol. Exp. Ther. 251, 888-893.
JEFFERSON, J. R., HARMON, J. T. & JAMIESON, G. A. (1988) Identification of high affinity (Kd 0.35 pmol/L) and low affinity (Kd 7.9 pmol/L) platelet binding sites for ADP and competition by ADP analogues. Blood 71,110-116.
KAPLAN, K. L., BROEKMAN, M. J., CHERNOFF, A., LESZNIK, G. R. & DRILLINGS, M. (1979) Platelet a-granule proteins: studies on release and subcellular localization. Blood S3, 604-618.
KASER-GLANZMANN, R. GERBER, E. & LUSCHER, E. F. (1978) Regulation of the intracellular calcium level in human blood platelets: cyclic adenosine3’,5-monophosphate dependent phosphorylation of a 22,000 dalton component in isolated Ca2+-accumulating vesicles. Biochim. Biophys. Acta 558, 344-347.
KENAKIN, T. (1993) Pharmacologic Analysis o f Drug Receptor Interaction. 2nd Edition, Raven Press, New York.
170
KIKUGAWA, K., SUEHIRO, H. & ICHINO, M. (1973) Platelet aggregation inhibitors. 7. S-substituted2-thioadenosine 5-monophosphates. J. Med. Chem, 16, 1389-1391.
KIM, B. K., ZAMECNIK, P., TAYLOR, G., GUO, M. J. 8c BLACKBURN, M. (1992) Antithrombotic effect of p,p’-monochloromethylene diadenosine 5',5'"-P1 ,P4- tetraphosphate. Proc. Natl. Acad. Sci. USA 89, 11056-11058.
LANDOLFI, R., DE CRISTOFARO, R., DE CANDIA, E., ROCCA, B & BIZZI, B.(1991) Effect of fibrinogen concentration on the velocity of platelet aggregation. Blood 78,377-381.
LAWLER, J. (1986) The structural and functional properties of thrombospondin. Blood 67,1197-1209.
LEBOWITZ, E. A. 8c COOKE, R. (1978) Contractile properties of actomyosin from human blood platelets. J. Biol. Chem. 253, 5443-5447.
LEE, H., NURDEN, A. T., THOMAIDIS, A. & CAEN, J. P. (1981) Relationship between fibrinogen binding and the platelet glycoprotein deficiencies in Glanzmann’s Thrombasthenia type I and type II. Br. J. Haematol. 48,47-57.
LEFF, P. WOOD, B. E. 8c O’CONNOR, S. E. (1990) Suramin is a slowly-equilibrating but competitive antagonist at P2X-receptors in the rabbit isolated ear artery. Br. J. Pharmacol. 101,645-649.
LIPS, J. P. M., SIXMA, J. J. & SCfflPHORST, M. E. (1980) Binding of adenosine diphosphate to human blood platelets and to isolated blood platelet membranes. Biochim. Biophys. Acta 628,451-467.
LUSTIG, K. D., SHIAU, A. K., BRAKE, A. J. & JULIUS, D. (1993) Expression cloning of an ATP receptor from mouse neuroblastoma cells. Proc. Natl. Acad. Sci. USA 90, 5113-5117.
MACFARLANE, D. E. 8c MILLS, D. C. B. (1975) The effects of ATP on platelets: evidence against the central role of released ADP in primary aggregation. Blood 46, 309-320.
MACFARLANE, D. E. & MILLS, D. C. B. (1981) Inhibition by ADP of prostaglandin induced accumulation of cyclic AMP in intact human platelets. J. Cyclic Nulcleotide Res. 7,1-11.
MACFARLANE, D. E., MOLLS, D. C. B. & SRIVASTRAVA, P. C. (1982) Binding of 2-azidoadenosine [p-3 2 P]-diphosphate to the receptor on intact human blood platelets which inhibits adenylate cyclase. Biochemistry 21, 544-549.
171
MACFARLANE, D. E., SRIVASTRAVA, P. C. & MILLS, D. C. B. (1983) 2-methylthioadenosine [p-3 2 P]-diphosphate. An agonist and radioligand for the receptor that inhibits the accumulation of cyclic AMP in intact blood platelets. J. Clin. Invest. 71, 420-428.
MACINTYRE, D. E., POLLOCK, W. K., SHAW, A. M., BUSHFIELD, M., MACMILLAN, L. J. & McNICOL, A. (1985) Agonist-induced inositol phospholipid metabolism and Ca++ flux in human platelet activation. Adv. Exp. Biol. Med. 192, 127-144.
MACMILLAN, M. C. (1966) Secondary clumping effect in human citrated platelet-rich plasma produced by adenosine diphosphate and adrenaline. Nature 211,140-144.
MAGUIRE, M. H. & MICHAL, F. (1968) Powerful new aggregator of blood platelets -2-chloroadenosine 5'-diphosphate. Nature 217, 571-573.
MAGUIRE, M. H. & SATCHELL, D. G. (1981) Purinergic receptors in vascular smooth muscle. In Purinergic Receptors: Receptors and Recognition Series B, (Bumstock, G., ed), vol. 12, pp. 47-92, Chapman & Hall, London.
MAHAUT-SMITH, M. P., SAGE, S. O. & RINK, T. J. (1990) Receptor-activated single channels in intact human platelets. J. Biol. Chem. 265,10479-10483.
MAHAUT-SMITH, M. P., SAGE, S. O. & RINK, T. J. (1992) Rapid ADP-evoked currents in human platelets recorded with the nystatin permeabilized patch technique. J. Biol. Chem. 267,3060-3065.
MAHONEY, C. W., AZZI, A, & HUANG, K.-P. (1990) Effects of suramin, an anti-human immunodeficiency virus reverse transcriptase agent, on protein kinase C. J. Biol. Chem. 265, 5424-5428.
MARCUS, A. J. & SAFIER, L. B. (1993) Thromboregulation: multicellular modulation of platelet reactivity in haemostasis and thrombosis. FASEB J. 7, 516-522.
MARGUERIE, G. A., PLOW, E. F. & EDGINGTON, T. S. (1979) Human platelets possess an inducible and saturable receptor specific for fibrinogen. J. Biol. Chem. 254, 5357-5363.
MARTIN, J. H., CARSON, F. L. & RACE, G. J. (1974) Calcium-containing platelet granules. J. Cell Biol. 60,775-777.
MASSINI, P. & LUSCHER, E. F. (1971) The induction of the release reaction in human blood platelets by close cell contact. Thrombos. Diathes. Haemorrh. 25, 13-20.
MAYINGER, P. & GAWAZ, M. (1992) Photoaffinity labelling of integrin ocIIbp3 (glycoprotein Ilb-IIIa) on intact platelets with 8 -azido-[y-3 2 P]ATP. Biochim. Biophys. Acta 1137, 77-81.
172
McCLURE, M. O., KAKKAR, A., CUSACK, N. J. & BORN, G. V. R. (1988) Evidence for the dependence of arterial haemostasis on ADP. Proc. R. Soa Lond. B 234,255-262.
MEGHJI, P. (1993) Storage, release, uptake and inactivation of purines. Drug Development Res. 28,214-219.
MEYERS, K. M., HOLMSEN, H. & SEACHORD, C. L. (1982) Comparative study of platelet dense granule constituents. Am. J. Physiol. 243, R454-R461.
MICHAL, F., MAGUIRE, M. H. & GOUGH, G. (1969) 2-methylthioadenosine- 5-phosphate: a specific inhibitor of platelet aggregation. Nature 222,1073-1074.
MILLS , D. C. B .,FIGURES, W. R., SCEARCE, L. M., STEWART, G. J. COLMAN, R. F. & COLMAN, R.W. (1985) Two mechanisms for inhibition of ADP-induced platelet shape change by 5-/?-fluorosulfonlybenzoyladenosine. Conversion to adenosine, and covalent modification at an ADP binding site distinct from that which inhibits adenylate cyclase. J. Biol. Chem. 260, 8078-8083.
MILLS, D. C. B. & MACFARLANE, D. E. (1977) An attempt to define a platelet ADP receptor with 2 0 3 Hg-/>-mercuribenzene sulphonate (MBS) (Abstract). Thrombos. Haemostas. 38, 82A.
MILTON, J. G. & FROJMOVIC, M. M. (1984) Adrenaline and adenosine diphosphate-induced platelet aggregation require shape change. Importance of pseudopods. J. Lab. Clin. Med. 104, 805-815.
MORIYAMA, Y. & NELSON, N. (1988) Inhibition of vacuolar H+-ATPases by fusidic acid and suramin. FEBS Lett. 234, 383-386.
MOTTE, S., PIROTTON, S. & BOEYNAEMS, J. M. (1993) Evidence that the form of ATP uncomplexed with divalent cations is the ligand of the P2y and nucleotide/P2u receptors on aortic endothelial cells. Br. J. Pharmacol. 109, 967-971.
MUSTARD , J. F., KINLOUGH-RATHBONE, R. L. & PACKHAM, M. A. (1989) Isolation of human platelets from plasma by centrifugation and washing. Methods Enzymol. 169, 3-11.
MUSTARD, J. F., KINLOUGH-RATHBONE, R. L., PACKHAM, M. A., PERRY, D. W., HARFENIST, E. J. & PAI, K. R. M. (1979) Comparison of fibrinogen associated with normal and thrombasthenic platelets on exposure to ADP or chymotrypsin. Blood 54, 987-993.
MUSTARD, J. F. & PACKHAM, M. A. (1977) Normal and abnormal haemostasis. Br. Med. Bull. 33, 187-193.
MUSTARD, J. F., PACKHAM, M. A., KINLOUGH-RATHBONE, R. L., PERRY, D. W. & REGOECZI, E. (1978) Fibrinogen and ADP-induced aggregation. Blood 52,453-466.
173
MUSTARD, J. F., PERRY, D. W., KINLOUGH-RATHBONE, R. L. & PACKHAM, M. A. (1975) Factors responsible for ADP-induced release reaction of human platelets. Am. J; Physiol 228,1757-1765.
NACHMAN, R. L. & FERRIS, B. (1974) Binding of adenosine diphosphate by isolated membranes from human platelets. J. Biol. Chem. 249, 704-710.
NACHMAN, R. L. & LEUNG, L. L. (1982) Complex formation of platelet membrane glycoproteins lib and 331a with fibrinogen. J. Clin. Invest. 69,263-269.
NACHMIAS, V. T. (1980) Cytoskeleton of human platelets at rest and after spreading. J. Cell Biol. 86,795-802.
NAKAZAWA, K., FUJIMORI, K., TAKANAKA, A. & JNOUE, K. (1990) Reversible and selective antagonism by suramin of ATP-activated inward current in PC 12 phaeochromocytoma cells. Br. J. Pharmacol. 101,224-226.
NISfflO, H., IKEGAMI, Y., NAKATA, Y. & SEGAWA, T. (1992) Fluorescence digital image analysis of thrombin and ADP induced rise in intracellular Ca2+ concentration of single blood platelets. Neurochem. Int. 21, 75-81.
O'CONNOR, S. E., DAINTY, I. A. & LEFF, P. (1991) Further subclassification of ATP receptors based on agonist studies. Trends Pharmacol. Sci. 12,137-141.
OFFENSPERGER, W.B., WALTER, E., OFFENSPERGER, S., ZESCHNIGK, C., BLUM, H. E. & GEROK, W. (1988) Duck hepatitis B virus: DNA polymerase and reverse transcriptase activities of replicative complexes isolated from liver and their inhibition in vitro. Virology 164,48-54.
OLSSON, R. A. & PEARSON, J. D. (1990) Cardiovascular purinoceptors. Physiol. Rev. 70, 761-845.
ONO, K., NAKANE, H. FUKASHIMA, M. (1988) Differential inhibition of various deoxyribonucleic and ribonucleic acid polymerases by suramin. Eur. J. Biochem. 172, 349-353.
PACKHAM, M. A., BRYANT, N. L., GICCIONE, M. A., KINLOUGH-RATHBONE, R.L. & MUSTARD, J. F. (1989) Effect of the concentration of Ca2+ in the suspending medium on the responses of human and rabbit platelets to aggregating agents. Thrombos. Haemostas. 62, 968-976.
PACKHAM, M. A., GUCCIONE, M. A., PERRY, D. W., KINLOUGH-RATHBONE, R. L. & MUSTARD, J. F. (1972) AMP inhibition of reactions of ADP with washed platelets from humans and rabbits. Am. J. Physiol. 223,419-424.
PACKHAM, M. A., GUCCIONE, M. A., PERRY, D. W. & MUSTARD, J. F (1974) Interactions of nucleoside di- and triphosphates with rabbit platelets. Am J. Physiol. 227, 1143-1148.
174
PACKHAM, M. A., LIVNE, A.-A., RUBEN, D. H. & RAND, M. L. (1993) Activation of phospholipase C and protein kinase C has little involvement in ADP-induced primary aggregation of human platelets: effects of diacylglycerols, the diacylglycerol kinase inhibitor R59022, staurosporine and okadaic acid. Biochem. J. 290, 849-856.
PEARCE, P. H., WRIGHT, J. M., EGAN, C. M. & SCRUTTON, M. C. (1978) Interaction of human blood platelets with the 2',3-dialdehyde and 2',3-dialcohol derivatives of adenosine 5-diphosphate and adenosine 5’-triphosphate. Eur. J. Biochem. 8 8 , 543-554.
PEARSON, J. D., CARLETON, J. S. & GORDON, J. L. (1980) Metabloism of adenine nucleotides by ectoenzymes of vascular endothelial and smooth muscle cells in culture. Biochem. J. 190,421-429.
PECORARO, V. L., HERMES, J. D. & CLELAND, W. W. (1984) Stability constants of Mg2+ and Cd2+ complexes of adenine nucleotides and thionucleotides and rate constants for the formation and dissociation of MgATP and MgADP. Biochemistry 23, 5262-5271.
PEERSCHKE, E. I. B. (1985) The platelet fibrinogen receptor. Semin. Hematol. 22, 241-259.
PEERSCHKE, E. I., ZUCKER, M. B., GRANT, R. A., EGAN, J. J. & JOHNSON, M. M. (1980) Correlation between fibrinogen binding to human platelets and platelet aggregability. Blood. 55, 841-847.
PHILLIPS, D. R. & AGIN, P. P. (1977) Platelet membrane defects in Glanzmann’s thrombasthenia. Evidence for decreased amounts of two major glycoproteins. J. Clin. Invest. 60, 535-545
PLESCHER, A., DA PRADA, M., BERNEIS, K. H. & TRANZER, J. P. (1971) New aspects on the storage of 5-hydroxytryptamine in blood platelets. Experientia 27, 993-1120.
POWLING, M. J. & HARDISTRY, R. M. (1985) Glycoprotein Hb-ffla complex and Ca2+ influx into stimulated platelets. Blood 6 6 , 731-734.
RAGATZ, B. H., IATRIDIS, P. G., IATRIDIS, S. G., MARKIDOU, S. G., ASTERITA, M. F. & GADAROWSKI, J. (1977) Interactions of the ADP analogs, 2-O-methyl ADP and adenine 9-P-D-arabinofuranoside 5'-diphosphate and catecholamines with human platelets. Thrombos. Haemostas. 38,209 (Abstract).
RAHA, S., JONES, G. D. & GEAR, A. R. (1993) Sub-second oscillations of inositol 1,4,5-trisphosphate and inositol 1,3,4,5-tetrakisphosphate during platelet activation by ADP and thrombin: lack of correlation with calcium kinetics. Biochem. J. 292,643-646.
175
RAO, A. K. & KOWALSKI, M. A (1987) ADP-induced platelet shape change and mobilization of cytoplasmic ionised calcium are mediated by distinct binding sites on platelets: 5'-p-fluorosulfonylbenzoyladenosine is a weak platelet agonist. Blood 70, 751-756.
RINK, T. J. & SAGE, S. O. (1987) Stimulated calcium efflux from fura-2-loaded human platelets. J. Physiol 393, 513-524.
ROBEY, F. A., JAMIESON, G. A. & HUNT, J. B. (1979) Synthesis and use of a new spin-labelled analogue of ADP with platelet-aggregating activity. J. Biol Chem. 254, 1114-1118.
ROTH, G. J. (1992) Platelets and blood vessels: the adhesion event. Immunology Today 13, 100-105.
SAGE, S. O., MERRITT, J. E., HALLAM, T. J. & RINK, T. J. (1989) Receptor-mediated calcium entry in fura-2-loaded human platelets stimulated with ADP and thrombin. Biochem. J. 258, 923-926.
SAGE, S. O., REAST, R. & RINK, T. J. (1990) ADP evokes biphasic Ca2+ influx in fura-2-loaded human platelets. Evidence for Ca2+ entry regulated by the intracellular Ca2+ store. Biochem J. 265, 675-680.
SAGE, S. O. & RINK, T. J. (1986) Kinetic differences between thrombin-induced and ADP-induced calcium influx and release from internal stores in fura-2-loaded human platelets. Biochem. Biophys. Res. Commun. 136, 1124-1129.
SAGE, S. O. & RINK, T. J. (1987) The kinetics of changes in intracellular calcium concentration in fura-2-loaded human platelets. J. Biol Chem. 262,16364-16369.
SALZMAN, E. W., CHAMBERS, D. A. & NERI, L. L. (1966) Incorporation of labelled nucleotides and aggregation of human blood platelets. Thrombos. Diathes. Haemorrh. 15, 52-68.
SARGEANT, P., CLARKSON, W. D., SAGE, S. O. & HEEMSKERK, J. W. M. (1992) Calcium influx evoked by Ca2+ store depletion in human platelets is more susceptible to cytochrome P-450 inhibitors than receptor-mediated calcium entry. Cell Calcium 13, 553-564.
SARGEANT, P., FARNDALE, R.W. & SAGE, S. O. (1993a) The tyrosine kinase inhibitors methyl 2,5-dihydroxycinnamate and genistein reduce thrombin-evoked tyrosine phosphorylation and Ca2+ entry in human platelets. FEBS Lett. 315,242-246.
SARGEANT, P., FARNDALE, R. W. & SAGE, S. O. (1993b) ADP- and thapsigargin- evoked Ca2+ entry and protein-tyrosine phosphorylation are inhibited by the tyrosine kinase inhibitors genistein and methyl-2 ,5-dihydroxycinnamate in fura-2 -loaded human platelets. J. Biol Chem. 268,18151-18156.
176
SHEU, K.-W., R. & FREY, P. A. (1977) Enzymatic and 31P nuclear magnetic resonance study of adenylate kinase-catalysed stereospecific phosphorylation of adenosine 5'-phosphorothioate. J. Biol. Chem. 252,4445-4448.
SDEFERT, R. & SCHULTZ, G. (1989) Involvement of pyrimidoceptors in the regulation of cell functions by uridine and uracil nucleotides. Trends Pharmacol Sci. 10, 365-369.
SEESS, W. (1989) Molecular mechanisms of platelet activation. Physiol Rev. 69, 58-178.
SIFFERT, W., SIFFERT, G., SCHEID, P & AKKERMAN, J. W. N. (1990) Na+/H+ exchange modulates Ca2+ mobilization in human platelets stimulated by ADP and thromboxane mimetic U46619. J. Biol Chem. 264, 719-725.
SKAER, R. J., PETERS, P. D. & EMMINES, J. P. (1974) The localization of calcium and phosphorus in human platelets. J. Cell Sci. 15, 679-692.
SOLUM, N. O. & STORMORKEN, H. (1965) Influence of fibrinogen on the aggregation of washed human blood platelets induced by adenosine diphosphate, thrombin, collagen and adrenaline. Scand. J. Lab. Clin. Invest. 17 (suppl 84), 170-182.
SOSLAU, G., BRODSKY, I. & PARKER, J. (1993) Occupancy of P2 purinoceptors with unique properties modulates the function of human platelets. Biochim. Biophys. Acta 1177,199-207.
SOSLAU, G. & PARKER, J. (1989) Modulation of platelet function by extracellular adenosine triphosphate. Blood 74, 984-993.
SPANGENBERG, P., HEPINSTALL, S., CRAWFORD, J. & TILL, U. (1992) Measurements of the G-/F-actin equilibrium in ADP-stimulated human platelets. Acta Histochem. 41, SI81-186.
STONE, J. V., SINGH, R. K., HORAK, H. & BARTON, P. G. (1976) Sulfhydryl analogues of adenosine diphosphate: chemical synthesis and activity asplatelet-aggregating agents. Can. J. Biochem. 54, 529-533.
SUZUKI, H., KINLOUGH-RATHBONE, R. L., PACKHAM, M. A., TANOUE, K , YAMAZAKI, H. & MUSTARD, J. F. (1988) Immunocytochemical localization of fibrinogen on washed human platelets. Lack of requirement for fibrinogen during adenosine diphosphate-induced responses and enhanced fibrinogen binding in a medium with low calcium levels. Blood 71, 850-860.
SWEATT, J. D., JOHNSON, S. L., CRAGOE, E. J. & LIMBIRD, L. E. (1985) Inhibitors of Na+/H+ exchange block stimulus-provoked arachidonic acid release in human platelets. Selective effects on platelet activation by epinephrine, ADP and lower concentrations of thrombin. J. Biol Chem. 260, 12910-12919.
177
SWEATT, J. D., BLAIR, I. A., CRAGOE, E. J. & LIMBIRD, L. E. (1986a) Inhibitors of Na+/H+ exchange block epinephrine- and ADP-induced stimulation of platelet phospholipase C by blockade of arachidonic acid release at a prior step. J. Biol. Chem. 261, 8660-8666.
SWEATT, J. D., CONNOLLY, T. M., CRAGOE, E. J. & LIMBIRD, L. E. (1986b) Evidence that Na+/H+ exchange regulates receptor-mediated phospholipase A2 activation in human platelets. J. Biol. Chem. 261, 8667-8673.
TAQUI KAHN, M. M. & MARTELL, A. E. (1966) Thermodynamic quantities associated with the interaction of adenosine triphosphate with metal ions. J. Am. Chem. Soc. 88, 668-671.
TAQUI KAHN, M. M. & MARTELL, A. E. (1967) Thermodynamic quantities associated with the interaction of adenosinediphosphoric and adenosinemonophosphoric acids with metal ions. J. Am. Chem. Soc. 89, 5585-5590.
TATHAM, P. E. R., CUSACK, N. J. & GOMPERTS, B. D. (1988) Characterisation of the ATP4- receptor that mediated permeabilisation of rat mast cells. Eur. J. Pharmacol. 147,13-21.
TSIEN. R. Y. (1980) New calcium indicators and buffers with high selectivity against magnesium and protons: design, synthesis and properties of prototype structures. Biochemistry 19,2396-2404.
VANHOUUTTE, P. M. (1988) Platelets, endothelium and blood vessel wall. Experientia 44, 105-109.
VICKERS, J. D., KINLOUGH-RATHBONE, R. L., PACKHAM, M. A. & MUSTARD, J. F. (1990) Inositol phospholipid metabolism in human platelets stimulated by ADP. Eur. J. Biochem. 193,521-528.
VON KUGELGEN, I., BULTMANN, R. & STARKE, K. (1990) Interaction of adenine nucleotides, UTP and suramin in the mouse vas deferens: suramin-sensitive and suramin-insensitive components in the contractile effect of ATP. Naunyn- Schmiedebergs Arch. Pharmacol. 342, 198-205.
WAKUI, M., OSIPCHUK, Y. V. & PETERSEN, O. H. (1990) Receptor-activated cytoplasmic Ca2+ spiking mediated by inositol trisphosphate is due to Ca2+-induced Ca2+ release. Cell 63,1025-1032.
WEBB, T. E., SIMON, J., KRISHEK, B. J., BATESON, A. N., SMART, T. G., KING, B. F., BURNSTOCK, G. & BARNARD, E. A. (1993) Cloning and functional expression of a brain G-protein-coupled ATP receptor. FEBS Lett. 324,2119-225.
WHITE. J. G. (1968a) Effects of colchicine and Vinca alkaloids on human platelets. 1. Influence on platelet microtubules and contractile function. Am. J. Pathol. 53,281-291.
178
WHITE, J. G. (1968b) Fine structural alterations induced in platelets by adenosine diphosphate. Blood 31, 604-622.
WHITE, J. G. (1971) Platelet microtubules and microfilaments: effects of cytochalasin B on structure and function. In Platelet Aggregation (Caen, J., ed), pp. 15-52, Masson, Paris.
WHITE, J. G. (1972) Interaction of membrane systems in blood platelets. Am. J. Pathol 66, 295-312.
WHITE, J. G. (1973) Identification of platelet secretion in the electron microscope. Series Haematologica 6,429-459.
WHITE, J. G. (1974a) Electron microscopic studies of platelet secretion. Progress Haemostas. Thrombos. 2,49-98.
WHITE, J. G. (1974b) Shape change. Thrombos. Diathes. Haemorrh 60,159-171.
WHITE, J. G. (1984) Arrangements of actin filaments in the cytoskeleton of human platelets. Am. J. Pathol. 117,207-217.
WHITE, J. G. (1987) Platelet ultrastructure. In Haemostasis and Thrombosis (Bloom, A. L. & Thomas, D. P., eds), pp. 20-46, Churchill Livingstone, Edinburgh.
WHITE, J. G. & KRIVTT, W. (1967) An ultrastructural basis for the shape changes induced by chilling. Blood 30, 625-635.
WILLS, E. D. & WORMALL, A. (1950) Studies on suramin, the actions of the drug on some enzymes. Biochem. J. 47,158-170.
YOUNT, R. G., BABCOCK, D., BALLANTYNE, W. & OJALA, D. (1971) Adenylyl imidodiphosphate, an adenosine triphosphate analog containing a P-N-P linkage. Biochemistry 10,2484-2489.
ZAMECNIK, P. C., KIM, B. GAO, M. J., TAYLOR, G. & BLACKBURN, G. M. (1992) Analogues of diadenosine 5',5"f-P1,P4-tetraphosphate (Ap4A) as potential anti-platelet- aggregation agents. Proc. Natl. Acad. Sci. USA 89, 2370-2373.
ZHOU, Q.-Y., LI, C., OLAH, M. E., JOHNSON, R. A., STILES, G. L. & CIVELLI, O. (1993) Molecular cloning and characterization of an adenosine receptor: The A3 adenosine receptor. Proc. Natl. Acad. Sci. USA 89, 7432-7436.
ZUCKER, M. B., PART, J. H. & HILGARTNER, M. W. (1966) Platelet function in a patient with Thrombasthenia. Blood 28, 524-534.
ZUCKER, M. B. & ZACCARDI, J. B. (1964) Platelet shape change induced by adenosine diphosphate and prevented by adenosine monophosphate. Fed. Proc. 23,299 (Abstract).
179
M A T E R I A L R E D A C T E D A T R E Q U E S T O F UNI VE RSI TY