C. The Phosphodiesterase Family

Cyclic nucleotide phosphodiesterases are the intracellular enzymes responsible for the degradation of both cAMP and cGMP. Although intracellular levels of cGMP may also be reduced by the release of cGMP into extracellular space, this however accounts for very little of the total cGMP removed from the cell (Schini et. al., 1989). At present there are over 40 known isozymes of cyclic nucleotide specific phosphodiesterases found in various tissues (Beavo, et al., 1994). These isozymes have been divided into seven distinct families based on their biochemical characteristics (type I through VII.). As indicated in Table 1, each family subtype preferentially hydrolyzes a specific cyclic nucleotide substrate. This review will only focus upon the cGMP specific phosphodiesterase families that are found in vascular smooth muscle. Phosphodiesterase types I, II, III, IV, and V have been isolated in vascular smooth muscle, but it is thought that only phosphodiesterase types I, II, and V contribute to the hydrolysis of cGMP in vascular smooth muscle (Saeki and Saito, 1993).

It should be noted that no standardized nomenclature existed for the phosphodiesterase families until 1990 (Beavo and Reifsnyder, 1990). Before this time each phosphodiesterase isozyme was referred to by the order in which they were eluted from a DEAE column. To complicate matters further, differing extraction techniques yielded different phosphodiesterase isozymes. As a result much of the early literature is difficult to decipher in regards to the phosphodiesterase type being described.

Not only are the biochemical characteristics of these isozymes different, but comparison of the genes encoding each has confirmed the presence of multiple genes that encode for the different family members (Swinnen, et al., 1989)(Le Trong, et al., 1990)(Monaco, et al., 1994). Table 1 lists the number of known genes for each family. As can also been seen from Table 1, additional diversity is achieved by alternative splicing of many of the genes and by post-translational modification of the protein products. Due to the complexity of the phosphodiesterase families, it is difficult to discuss any single isozyme in any given family, but it is possible to discuss the general domain organization seen within all phosphodiesterase families. Although each phosphodiesterase subtype may be comprised of different subunits, all sequenced mammalian phosphodiesterase subunits display a high degree of homology within a conserved region of 270 amino acid located near the carboxyl terminus (Charbonneau, 1991). This region has been shown to contain the catalytic domain (McAllister-Lucas, et al., 1993).

In addition to the conserved catalytic domain, the cGMP-binding phosphodiesterase type II and V possess a second conserved region of approximately 340 residues, located upstream from the catalytic domain (Charbonneau, et al., 1990). Within this region there are seven conserved residues in which three possess charged side groups are thought to be important for substrate interaction (McAllister-Lucas, et al., 1993). This region has not been identified in any non cGMP-binding phosphodiesterase, as such this region is thought to comprise the cGMP binding domain. Type II phosphodiesterase (cGMP-stimulated) are comprised of two identical subunits to form a dimer structure. Each subunit has a mass of approximately 105 kD (Sonnenburg, et al., 1991). The binding of cGMP to each subunit is thought to be allosteric. Type II phosphodiesterase's peak activity is at submicromolar concentrations of cGMP. Cyclic GMP concentrations exceeding the submicromolar range have a mild inhibitory effect. Under in vivo conditions the degradative role played by this subtype may be transient as cGMP levels increase during activation. Type V phosphodiesterase (cGMP-binding) is a homodimer comprised of two 93 kD subunits (Thomas, et al., 1990). Although type V phosphodiesterase displays cGMP binding properties similar to those of type II, elevated levels of cGMP have not been shown to inhibit the degradative activity of type V phosphodiesterase. Due to its low Km for cGMP (0.5 µM), type V is thought to be the major phosphodiesterase isoform responsible for maintaining basal cGMP levels in vascular smooth muscle. The cGMP binding site for both type II and V seems to be the site targeted by a variety of cGMP specific phosphodiesterase inhibitory drugs.

The other major phosphodiesterase family responsible for the hydrolysis of cGMP in the vascular smooth muscle is the calmodulin dependent phosphodiesterase type I. Within the cGMP phosphodiesterase families, this subtype seems to be the most functionally diverse (Sonnenburg, et al., 1993) (Beavo, et al., 1994). One of the best studied type I phosphodiesterase is a 61 kD isoform. In this isoform there are two regions comprised of 80 residues located near the amino terminus that are required for calmodulin binding and activation (Florio, et al., 1994). Because type I phosphodiesterase preferential hydrolyze cAMP, it is likely that this subtype's cGMP hydrolytic activity is only active in vivo during EDRF stimulation when cGMP levels are high enough to stimulate its activity. When isolated, each phosphodiesterase subtype discussed above, possesses its own distinct kinetic profile. In vivo , each phosphodiesterase subtype contributes differently to the overall degradative activity of the tissue. Because different subtype isozymes are expressed in a tissue specific manner, each tissue type displays a degradative profile that is inherent in that specific tissue type. Therefore, this profile can be utilized when comparing the importance of each phosphodiesterase in governing overall cGMP levels in vivo.

The complexity of the phosphodiesterase families have produced major problems in studying these enzymes and their regulation. Despite the difficulties involved, a major precedent for dramatic hormonal regulation of cGMP phosphodiesterase has been shown in the retina (Stryer, 1986)(Stryer, 1991). In retinal rods and cones, the capture of a photon of light increases the rate of cGMP breakdown, this stimulus is transduced into decreased intracellular cGMP levels. Additional examples of complex hormonal regulation of cyclic nucleotide phosphodiesterases have also been shown by Dumas, et al., (1988) and Laugier, et al., (1988) in the quail oviduct. The knowledge that phosphodiesterases can be regulated in a complex fashion is an indication of the important role these enzymes may play in many cellular processes. Short-term regulation and long-term cellular modulation by different stimuli and maturation might be one function of phosphodiesterase (Conti, et al., 1991).

Additional forms of phosphodiesterase regulation which can be classified as non-hormonal have also been demonstrated. This form of regulation is on the molecular level and involves either phosphorylation of specific residues within the phosphodiesterase's catalytic domain, or differential expression of phosphodiesterase isozymes in different tissues. Specific phosphorylation by protein kinases of the Ser-120 residue within the calmodulin binding domain of type I phosphodiesterase, has been shown to reduce calmodulin's affinity for type I phosphodiesterase which results in a reduction of its activity (Florio, et al. 1994). Type II and V phosphodiesterases have also been demonstrated to be substrates for phosphorylation by cAMP dependent protein kinase (Beltman, et al., 1993)(Beavo, et al., 1994).

The most significant factor in the regulation of phosphodiesterases and their role in the cyclic nucleotide metabolism, is the fact that phosphodiesterase isozymes and families tend to be very tissue specific (Weishaar, et al., 1986)(Silver, et al., 1988)(Souness, et al., 1990)(Repaske, et al., 1993). As was previously discussed with soluble guanylate cyclase isozyme distribution, the heterogeneity of phosphodiesterase isozyme distribution certainly plays a role in cGMP regulation. In addition to normal phosphodiesterase tissue heterogeneity, evidence for differential expression has also been shown as a function of age. Not only does the degradative capacity of phosphodiesterase change with maturation, so does the synthetic capacity of soluble guanylate cyclase (Pearce, et al., 1994)(White and Pearce, 1996). Both the basal synthesis and degradation of cGMP are enhanced in newborn lambs vascular smooth muscle compared to adults. This may help explain why vascular resistance is lower in newborns than adults.

The heterogeneity in phosphodiesterase distribution, combined with the existence of drugs designed to inhibit specific phosphodiesterase types, has generated considerable interest for their clinical use (Polson, 1990)(Newby, et al., 1992)(Kulka and Tryba, 1994). In a traditional clinical setting, high blood pressure is treated by the use of beta-adrenergic blockers such as propranolol and atenolol. Unfortunately, vasodilatation induced in such a manner is systemic. While this has the desired effect of lowering blood pressure, it also effects systems which require higher blood pressures for optimum performance, such as the kidneys and lungs. The pharmacological development of phosphodiesterase inhibitors targeted at specific tissues, such as the vasculature, would be a valuable clinical tool. Specific phosphodiesterase inhibitors such as zaprinast have been shown to restore the vasorelaxant potency to nitroglycerin once tolerance has developed (Dundore, et al., 1993)(Pagani, et al., 1993). This possible therapeutic application has fueled a new interest in phosphodiesterase activity and its regulation.

NEXT