[Frontiers in Bioscience 3, d961-972, September 1, 1998]

Table of Conents
 Previous Section   Next Section


Carey J. Oliver and Shirish Shenolikar

Department of Pharmacology and Cancer Biology, Duke University Medical Center, Durham, North Carolina 27710

Received 8/14/98 Accepted 8/20/98


Reversible protein phosphorylation controls many physiological processes in both plant and animal cells with the phosphorylation of cellular proteins being defined by the protein kinases and phosphatases that act on them. Discovery of protein kinases directly activated by second messengers fostered the early view that kinases, and not phosphatases, were highly regulated in eukaryotic cells. However, analysis of cellular phosphatases soon made this concept untenable. For instance, the specific activity of non-receptor protein tyrosine phosphatases was found to far exceed that of the protein kinases which they opposed. This required that growth factors which initiate protein tyrosine phosphorylation to elicit a physiological response, must first overcome the barrier to signaling created by the phosphatases. This strongly hinted at the presence of cellular mechanisms to suppress phosphatase activity during times of kinase activation. These mechanisms would also provide for tremendous amplification of the physiological signals. Concomitant control of kinases and phosphatases provides the cell with the capacity to rapidly switch proteins from their phosphorylated to dephosphorylated state to meet differing physiological demands. This is perhaps best illustrated during the eukaryotic cell division cycle where decisions to proceed through different stages are made by the timely phosphorylation and dephosphorylation of specific cell cycle regulators. Thus, phosphorylation-dephosphorylation events act as switches or checkpoints that ensure that a cell has fulfilled the requirements to proceed to the next cell cycle stage. Errors in checkpoint control form the most prevalent basis for aberrant cell growth seen in human cancers and can carry dire developmental consequences for an organism.

A number of mechanisms have been analyzed with an eye on their potential for coordinating kinases and phosphatases. These include changes in the expression of protein phosphatases, their subcellular localization, phosphorylation of phosphatase catalytic and regulatory subunits and regulation by endogenous phosphatase inhibitors. Of these, hormonally-regulated protein inhibitors represent the best understood mechanism for regulating the major cellular protein serine/threonine phosphatases and for which there is also extensive physiological evidence. While endogenous inhibitors of protein tyrosine phosphatases have been reported, they remain largely uncharacterized (1). Therefore, in this review, we will limit our discussion to cellular proteins that regulate the major protein serine/threonine phosphatases in response to physiological stimuli. Analysis of these proteins has introduced us to several novel paradigms for cell signaling and has largely dismissed the notion of unregulated protein phosphatase activity.

One way in which the cell coordinates the activation of kinases with the inhibition of phosphatases is via endogenous phosphatase inhibitors which are activated by second messenger-regulated protein kinases. This is indeed the case for many proteins which function as endogenous phosphatase inhibitors only after they themselves have been phosphorylated. These phosphatase inhibitors provide a direct link between hormone-induced changes in second messenger levels and alterations in phosphatase activity that may themselves account for some aspects of the physiological response. By the activation of phosphatase inhibitors, the second messengers bolster or amplify the function of protein kinases, enhancing protein phosphorylation events. The coordination of kinases and phosphatases in this manner provides for a rapid onset of the physiological response and may also increase the size of the cellular response elicited by hormones.

Yet another mode of phosphatase regulation involves inhibitors that are active only in their dephosphorylated state. These proteins represent important mechanisms for maintaining latent pools of phosphatases that can be activated in response to certain hormones. In this regard, insulin and peptide hormones activate protein phosphatases that promote the dephosphorylation of enzymes involved in cell metabolism. Alternately, phosphatase inhibitors that are inactivated by phosphorylation may function as timers of the cellular response. In this scenario, hormones would trigger the activation of protein kinases that phosphorylate multiple proteins including the phosphatase inhibitors. The inactivation of the phosphatase inhibitors may lead to a delayed activation of phosphatases that terminate the cellular response. The presence of these phosphatase inhibitors and their phophorylation would then control the duration of a cellular response to hormones.

As most protein phosphatases are not dedicated to reverse the actions of specific protein kinases, changes in phosphatase activity likely have a broad impact on dephosphorylation and turnover of phosphoproteins which are substrates for many different kinases. Thus, hormones via the modulation of phosphatase inhibitors may control many different pathways, and phosphatase inhibitors become conduits for crosstalk between multiple signaling pathways or important devices for integrating and orchestrating the physiological response.

All the inhibitor proteins, discussed in this review (table 1), show remarkable specificity towards their target phosphatases. Being themselves phosphoproteins, it has been proposed that they function as psueodsubstrate inhibitors of the phosphatases. This then raises an intriguing issue of the structural features encoded within these phosphoproteins that make them highly selective and potent inhibitors, yet not substrates of their target phosphatases. In most cases, phosphorylation of these inhibitors is reversed by phosphatases other than the ones that they inhibit. This yields a novel regulatory cascade where one phosphatase, via the dephosphorylation of an inhibitor, modulates the function of another phosphatase. With the discovery of numerous phosphatase inhibitors and regulatory subunits that undergo reversible protein phosphorylation, there will undoubtedly be an increasing number of such protein phosphatase cascades. Finally, there is growing evidence of an intricate interplay between the inhibitors and other phosphatase regulatory subunits that controls enzyme activity so that the days of considering protein phosphatases as constitutive or unregulated enzymes are clearly over.

Table 1. Cellular phosphatase inhibitors

PP1 Inhibitors

IC50 (nM)

Mr (kDa)



PP1 Inhibitors

Inhibitor-1 (I-1)










Inhibitor-2 (I-2)




















PP2A Inhibitors

Inhibitor-1 (I1PP2A)





Inhibitor-2 (I2PP2A)





PP2B Inhibitors






2.1. Introduction to protein serine/threonine phosphatases

Given the focus of this review on protein serine/threonine phosphatase inhibitors, we need to briefly discuss the enzymes targeted by these proteins. For a complete description of these enzymes, the reader is referred to a number of excellent reviews (2, 3, 4). Prior to the advent of molecular cloning, protein serine/threonine phosphatases were classified by their biochemical properties (protein composition, in vitro substrate specificity, metal requirement and regulation by endogenous inhibitors) into two broad groups termed type 1 and type 2 phosphatases. The type-2 phosphatases were further subdivided into three groups, PP2A, PP2B and PP2C. Molecular cloning has now identified many protein serine/threonine phosphatases. With the exception of PP2C, the primary structures of these serine/threonine phosphatases show stretches of conserved amino acids that are the hallmarks of this enzyme family. PP1 (or type 1 phosphatase) and PP2A make up more than 90 % of the serine/threonine phosphatase activity in mammalian cells and are therefore the primary focus of this review.

2.1.1. Protein phosphatase-1

Protein Phosphatase 1 (PP1) regulates many biological processes including synaptic plasticity, cell cycle, gene transcription, and carbohydrate and lipid metabolism. Four mammalian isoforms of the PP1 catalytic subunit are generated from three genes. With the exception of PP1g2, which is predominantly expressed in testes, the other isoforms, PP1a, b and g1, are widely expressed in mammalian tissues. The functional importance of PP1 isoforms remains unclear. It has been speculated that they may associate with distinct regulatory subunits and thereby serve unique physiological functions but so far there has been little evidence to support this idea. PP1 activity is regulated by many hormones and growth factors. As the levels of the PP1 catalytic subunit do not change in response to physiological stimuli, hormonal regulation is thought to occur primarily through endogenous inhibitors and in some cases, through regulatory subunits (5). The largest number of phosphatase inhibitors thus far identified target PP1. These include Inhibitor-1 (I-1), Inhibitor-2 (I-2), dopmine- and cAMP-regulated phosphoprotein of Mr 32,000 (DARPP-32), nuclear inhibitor of PP1 (NIPP-1), C-kinase activated phosphatase inhibitor of Mr 17,000 (CPI17), and ribosomal inhibitor of PP1 (RIPP-1). In addition, a large number of PP1-binding proteins have been shown to inhibit the phosphorylase a phosphatase activity of PP1 in vitro. The precise role of these latter proteins in controlling PP1 activity in intact cells remains unknown, but some of these PP1-binding proteins may also turn out to be phosphatase inhibitors. Regardless, PP1 inhibitors are among the best understood phosphatase regulators and have set many of the prevailing paradigms for phosphatase regulation.

One of the paradigms for PP1 regulation operates during the control of the cAMP-response binding protein, CREB, which must be phosphorylated to be an active transcription factor. In many cells, this occurs in response to activation of PKA which directly phosphorylates CREB on serine133 (6). In other cells, such as neurons, elevation of intracellular calcium, either alone or in conjunction with phosphoinositide turnover, promotes CREB phosphorylation on serine133 (7). Experiments in cultured cells show that despite chronic activation of PKA by cell-permeable cAMP analogs (8) or treatment of cells with the calcium ionophore, ionomycin, and the PKC activator, phorbol ester (9), only transient phosphorylation of CREB and transcription of CRE-regulated genes occurs. This suggests that the CREB phosphatase, identified as a nuclear PP1 (8), is a dominant regulator of CREB function. This was confirmed by the pharmacological treatment of cells with the phosphatase inhibitor, okadaic acid, which leads to a robust and prolonged phosphorylation of CREB and may in part account for the cytotoxic effects of this toxin. Similarly, the expression of a constitutively active form of Inhibitor-1 (I-1), a PP1-specific inhibitor, promoted CREB phosphorylation and gene transcription, in the absence of elevated second messenger levels (10). These data point to the tremendous potential for signaling through PP1 to control CRE-mediated gene transcription and raises an intriguing possibility that a PP1 regulator, particularly one that inhibits PP1 activity, may be an important mediator of signal transduction in the nucleus.

The role of PP1 in the cell cycle also emphasizes the need for endogenous regulators. It has been established that the retinoblastoma gene product (RB) must be dephosphorylated as cells exit mitosis. This dephosphorylation activates RBís function as a growth suppressor that blocks cells in the G1 phase of the cell cycle (for review see 11). This was precisely what was shown by the studies of Berndt, et. al. (14). In these experiments, a potentially unregulated form of PP1 catalytic subunit by mutating a proposed phosphorylation site near the C-terminus was expressed in human cancer cells and blocked their entry into S phase. As PP1 levels do not change during the cell cycle, one presumes that cell cycle-dependent changes in PP1 activity must be mediated by either the covalent modification of PP1 catalytic subunit or by endogenous PP1 regulators. In this regard, a nuclear PP1-binding protein has been recently identified that may modulate PP1's activity against some cell cycle substrates (15).

In summary, considerable effort has been focused in recent years on the identification of endogenous PP1 regulators in an attempt to understand the physiological control of this protein phosphatase.

2.1.2. Protein phosphatase 2A

PP2A is also a major protein phosphatase in all eukaryotic cells and has a wide range of biological functions. These include the control of cell cycle, organization of cytoskeleton, transcription of immediate early genes, cholesterol and protein biosynthesis. PP2A is a heterotrimeric enzyme made up of a catalytic or C subunit, and two regulatory subunits, termed A and B. To date, cDNAs have been identified for two A subunits, two C subunits, and over twenty B subunits. This suggests the existence of numerous PP2A complexes in mammalian cells. It has been speculated that each PP2A complex serves distinct functions, although at this time, there is very little direct evidence to support this notion. PP2A regulation is further complicated by recent reports that A, B and C subunits all exist as phosphoproteins (3), but the physiological relevance of these findings remains to be established. The PP2AC subunit is also the major carboxymethylated protein in mammalian cells. This modification occurs at the C-terminal leucine309 and has been credited with various functions, including enzyme activation, assembly and localization to membranes (16). The identification of additional regulators, specifically PP2A inhibitors, is very much in its infancy. However, two PP2A inhibitors have recently been identified. Termed I1PP2A and I2PP2A, these proteins have been implicated in both physiological and pathophysiological regulation of PP2A (17).

2.1.3. Protein phosphatase 2B

PP2B, also termed calcineurin, is a calcium/calmodulin-activated protein serine/threonine phosphatase consisting of a catalytic A-subunit and a regulatory or calcium-binding B-subunit which makes this the only phosphatase directly regulated by second messengers (4). PP2B has a much narrower in vitro substrate specificity than either PP1 or PP2A. This is consistent with its specialized functions in the nervous system, T lymphocytes and other cells. Interestingly, the best known in vitro and in vivo substrates of PP2B are the PP1 inhibitors, I-1 and DARPP-32. Thus, PP2B controls PP1 activity and the two together form the first documented phosphatase cascade (18). PP2B has drawn much attention as the target of two clinically important immunosuppressive drugs, cyclosporin and FK506 (19). The complex of each drug with its cognate intracellular receptor, known as an immunophilin, binds to and inhibits the PP2B heterodimer. It has been speculated that the immunophilins may be physiological regulators of PP2B and the immunosuppressive drugs simply stabilize the formation of immunophilin/PP2B complexes to potentiate phosphatase inhibition (20). A more convincing candidate for an endogenous PP2B inhibitor is the newly discovered cain. This 240 kDa protein associates directly with the A or catalytic subunit and suppresses PP2B activity at micromolar concentrations (21). Several other PP2B binding proteins have been identified. Some of these like the A-kinase anchoring protein, AKAP-79, may also inhibit PP2B activity (22). PP2B associates with its substrate, the transcription factor NFAT (nuclear factor of activated T-cells). Recent studies show that the PP2B-binding domain of NFAT inhibits PP2B activity in vitro and in intact T-cells (23). Whether native NFAT functions as a substrate and a regulator of PP2B in mammalian cells remains to be determined.

2.1.4. Other protein serine/threonine phosphatases

A number of protein phosphatases have been identified by their homology to PP1 and PP2A. Some of these enzymes, such as PPX (or PP4) and PPV(or PP6), show intriguing and highly restricted subcellular localizations, pointing to specialized functions in cells. However, the study of these enzymes has lagged behind that of PP1, PP2A, and PP2B in large part due to problems with their expression and subsequent biochemical characterization. Thus, we know very little about the physiological functions and regulation of these enzymes in eukaryotic cells. The nomenclature for these newly discovered enzymes is also undergoing review and as such, they will not be discussed further in this review.