Cyclic Nucleotide Phosphodiesterases

Overview

Cyclic nucleotide phosphodiesterase activity plays an important role in cellular function. A variety of cellular activities are regulated through mechanisms controlling the level of cyclic AMP or cyclic GMP. These mechanisms include synthesis, degradation, efflux, and sequestration of cyclic nucleotides within the cell. Hydrolysis by phosphodiesterase is the unique mechanism for the degradation of the cyclic nucleotides. Despite their ubiquitous occurrence in biological systems, the cyclic nucleotides are not known to be substrates for any other enzymatic reaction.

The term “cyclic nucleotide phosphodiesterase” is meant to refer to a class of enzymes rather than slight variations in a single catalytic function. Several distinct hydrolytic activities have in common the same basic chemical reaction: the substrate is a nucleotide with a 3′,5′-phosphodiester and the product is a 5′-nucleotide. The phosphodiesterases isolated from various tissues differ in their substrate specificities, in their substrate affinities, and in their sensitivity to a variety of natural and artificial effectors. This variability has been observed even among the enzymes isolated from a single tissue. In fact, the presence of multiple forms of phosphodiesterase within a single tissue or cell type has been a consistent finding. In order to discuss this topic intelligently, a measure of order must be introduced. Accordingly, for the purposes of this review, we divide the commonly occurring phosphodiesterases into three broad categories: cyclic AMP phosphodiesterases, cyclic GMP phosphodiesterases, and cyclic nucleotide phosphodiesterases. This system of classification is based solely on the relative rates of hydrolysis of cyclic AMP and cyclic GMP at presumed physiological substrate conditions. The classification of a given enzyme refers only to its reported in vitro properties, and does not imply a judgment as to its true physiological role. This system of classification may not agree with that of all workers; almost all forms of phosphodiesterase will hydrolyze both cyclic AMP and cyclic GMP to some extent, and thus assays carried out at very high substrate concentrations will lead to different characterizations. Also, in much of the literature, the phosphodiesterases in a given tissue are described only in terms of a particular method of resolution. Such classifications do not lend themselves very readily to comparisons with work in other tissues using other methods. This is especially true of resolutions based on molecular size, which may give confusing results due to aggregation of similar enzymes to form enzymes with identical catalytic properties. Finally, there are certainly phosphodiesterases which do not fit into any of the above three categories and probably serve highly specific cell functions, for example, light-activated retinal cyclic GMP phosphodiesterase and cyclic CMP phosphodiesterase.

Each of the three categories of cyclic nucleotide phosphodiesterase, as defined above, is found in most mammalian tissues upon thorough examination, i.e., extracted by a number of methods, assayed in both soluble and particulate fractions, resolved by ion exchange chromatography, electrophoresis, or density gradient centrifugation, and assayed using both cyclic AMP and cyclic GMP over a range of substrate concentrations. The relative proportions of the three forms, however, vary greatly from organ to organ: in most brain tissues the cyclic GMP phosphodiesterase predominates; liver has a higher content of the low affinity, non-selective cyclic nucleotide phosphodiesterase, a lesser amount of the cyclic AMP phosphodiesterase, and very little of the cyclic GMP hydrolyzing enzyme; fat contains a relatively high proportion of the high affinity cyclic AMP phosphodiesterase. Even within cells, the distribution of the different enzyme forms is not uniform. While cytological localization is just beginning on mammalian phosphodiesterase, it appears, from work on homogenates, that the low affinity form is soluble and cytoplasmic, while the other two forms are particulate; cyclic GMP phosphodiesterase has been cytochemically visualized on the plasma membrane (at the synaptic junction of neurons), and cyclic AMP phosphodiesterase is found in microsomal preparations. The findings that the different forms of phosphodiesterase exhibit distinct catalytic and regulatory properties and different subcellular localizations suggest that they are indeed different enzymes, synthesized under separate genetic control and serving distinct physiological roles in regulating cellular activities. This is not a concept which is totally accepted; a number of investigators have obtained data which can be interpreted as indicating the intercon-vertability of the different cyclic nucleotide phosphodiesterases or the presence of a common subunit.

The question of the functions of the different forms of cyclic nucleotide phosphodiesterase must now be answered. A straight-forward pharmacological approach to this problem would be the first choice: introduce selective inhibitors of the different forms into the cells or organs of interest and observe which biological activities have been perturbed. This has not been successful. The best inhibitors of phosphodiesterase activity are methylxanthines and they are not selective to a great degree. Also cyclic nucleotides are so critical to cell function that the use of inhibitors leads to a great variety of direct and indirect changes. Another approach has been to analyze the phosphodiesterase activities extracted from cells which are functionally perturbed. Thus, insulin enhances the low K_m cyclic AMP phosphodiesterase of certain tissues; one role of this insulin-regulated enzyme may be in the regulation of carbohydrate and lipid metabolism. Similarly, the low K_m cyclic AMP phosphodiesterase is elevated in malignant cells. The enzyme might therefore be involved in regulating cell proliferation. Further evidence of discrete enzyme-function relationships comes from the localization of phosphodiesterase. The finding of high levels of cyclic GMP phosphodiesterase in neural tissues and the subcellular localization of this enzyme at postsynaptic membranes implicate this enzyme in some aspect of neurotransmission. Finally, the possibility exists that an enzyme considered to be hydrolytic is, in reality, a transferase utilizing the energy of the phosphate bond to carry out as yet undiscovered reactions.

To present the subject of cyclic nucleotide phosphodiesterase from a more historical point of view, one could cite the publications which mark significant milestones in research on the enzyme activity. First must come the original report of cyclic AMP hydrolysis (Sutherland and Rall 1958); then, after a gap of some years, the report of resolvable forms of phosphodiesterase (Thompson and Appleman 1971). Next, the three important findings of the regulation of the different forms of the enzyme by insulin (Senft et al. 1968; Loten and Sneyd 1970), by cyclic GMP (Beavo et al. 1970, 1971), and by calcium and a protein factor (Cheung 1970; Kakiuchi and Yamazaki 1970). Further developments to watch for might include immunological evidence for or against interconversion of various forms, cytochemical and immunocytochemical localization of different forms within the cell, development of highly specific inhibitors and, ultimately, the association of the various cyclic nucleotides with specific cellular activities.

In this review we will deal with a number of the subjects mentioned above. We will not discuss in any depth topics which have been recently reviewed: phosphodiesterase inhibitors (Chasin and Harris 1976), the methods of phosphodiesterase assay (Thompson et al. 1979 b), the calcium dependent regulator protein (Calmodulin) (Cheung and Storm, this volume). Recent general reviews dealing with these enzymes should also be mentioned (Wells and Hardman 1977; Amer and Kriegbaum 1975).