Introduction
It is now almost 30 years since the discovery that prostaglandin E1 (PGE1) inhibits platelet responses to aggregating agents, together with finding that the
effects of this compound are mediated by adenosine 3′, 5′-cyclic monophosphate (cAMP)
initiated interest in the physiological and pharmacological regulation of platelet
function by other agents that increase platelet cAMP, as reviewed elsewhere.1 The most important agonists that stimulate cAMP formation in platelets have now been
identified as prostacyclin (PGI2), prostaglandin D2 (PGD2), and adenosine, which exert their effects through receptors of the serpentine or
seven transmembrane segment class (IP, DP and A2 receptors, respectively).2 The latter then stimulate cAMP formation by adenylyl cyclase via the GTP-dependent
activation of the G-protein, Gs (Fig. 1). In the classical view, cAMP exerts its effects solely by binding to the
RI and RII regulatory subunits of type I and type II cAMP-dependent protein kinases
(PKA). The catalytic subunits of the kinases then dissociate and phosphorylate selected
serine and threonine residues on target proteins that prevent or reverse platelet
activation.2 A crucial role is played by cAMP phosphodiesterases, which degrade cAMP to 5′-AMP,
thereby diminishing and terminating the effects of agonists that stimulate cAMP formation
(Fig. 1). In early studies, this was demonstrated by the ability of first-generation
inhibitors of cAMP phosphodiesterases, particularly the methylxanthines, to inhibit
platelet aggregation and potentiate the inhibitory effects of activators of platelet
adenylyl cyclase.1 Such studies provided the rationale for the subsequent development of more potent
and selective phosphodiesterase inhibitors as potential antithrombotic agents.
Interest in the role of guanosine 3′,5′-cyclic monophosphate (cGMP) in platelets closely
followed the discovery of the inhibitory action of cAMP. An early hypothesis that
cGMP might potentiate platelet aggregation was abandoned by 1978, after it was shown
that some inhibitors of platelet aggregation, such as nitroprusside (NP), also increased
platelet cGMP.1 It soon emerged that all nitrovasodilators release nitric oxide and activate soluble
guanylyl cyclase (GC) and that the cGMP formed stimulates cGMP-dependent protein kinases
(PKG) in many cells and tissues (Fig. 1), including vascular smooth muscle and platelets.3 The crucial physiological importance of this pathway was established with the identification
of endothelium-derived relaxing factor (EDRF) as nitric oxide.4 cGMP phosphodiesterases play an essential role by limiting increases in cellular
cGMP, and inhibition of these enzymes was found to potentiate the effects of nitric
oxide and nitric oxide donors on platelets and other cells.5
The ability of cAMP and cGMP to activate distinct protein kinases led to a persistent
view that these two cyclic nucleotides operate in parallel and independent ways to
inhibit platelet function, cAMP mediating the effects of agonists such as PGI2, and cGMP mediating the effects of nitric oxide.2,3 However, over the last 10 years, considerable evidence has accumulated to indicate
that this is not the case in platelets (or in many other cells) and that cross-talk
between the cAMP and cGMP systems may occur on at least two levels, affecting both
cyclic nucleotide phosphodiesterase (PDE) and protein kinase activities (Fig. 1).
One of the most significant of these interactions is through the effects of cGMP on
the hydrolysis of cAMP by PDEs. It is the purpose of this chapter to describe platelet
PDEs and to discuss how their individual characteristics and regulation may impact
platelet function and the design of useful antithrombotic agents. In addition, evidence
that both cGMP and cAMP may activate PKG and that these cyclic nucleotides may exert
effects in platelets that do not involve either PKA or PKG will be discussed briefly.
