Introductory Review Cellscience Reviews Vol 2 No.2 ISSN 1742-8130

Protein Phosphatase Inhibitor-1, An Amplifier of cAMP signals

Suzanne Sikes & Shirish Shenolikar*

I. Introduction

Early studies (1) which noted changes in intracellular cyclic AMP levels in response to hormones established the second messenger hypothesis, which states that circulating hormones associate with cell surface receptors on target cells to activate membrane transducers or effectors, such as adenylyl cyclase, and increase the levels of intracellular cAMP. Cyclic AMP in turn functions as an intracellular mediator or “second messenger” to transduce signals that elicit the cellular response to hormones. The primary intracellular receptor for cyclic AMP was later identified as a protein kinase (2). This enzyme, now commonly termed PKA, phosphorylates a variety of cellular proteins on serines and threonines to induce a conformational change that modulates the biochemical properties or cellular functions of these PKA substrates. A key feature of this regulatory paradigm is that the protein phosphorylation events are transient and reversible. Thus, the dephosphorylation of these phosphoproteins resets the cell for subsequent stimulation by hormones. Studies over the last four decades have clearly established that reversible protein phosphorylation, most of which occur on serine and threonine residues, is a major mechanism by which extracellular stimuli regulate the physiology of all eukaryotic cells.

Protein phosphatase-1 (PP1) accounts for 50-90% of cellular protein serine/threonine phosphatase activity. Thus, in mammalian tissues, PP1 potentially regulates more than 100,000 protein phosphorylation events. Early work by Lee and coworkers (3) showed that hormones not only activated PKA, but also modulated phosphorylase phosphatase activity, later shown to be primarily PP1, in rabbit skeletal muscle. These studies suggested that coordinate activation of PKA and inhibition of PP1, possibly via the generation of endogenous phosphatase inhibitor(s), significantly amplified the hormone signals and provided more effective control of the size and duration of the hormone response.

Glinsman and colleagues (4) analyzed hormonally stimulated skeletal muscle tissue for the existence of phosphatase inhibitors and isolated two thermostable proteins, termed inhibitor-1 (I-1) and inhibitor-2 (I-2), that potently inhibited skeletal muscle phosphorylase phosphatase. I-1 was only an effective inhibitor of the muscle phosphorylase phosphatase activity after it had been phosphorylated on a specific threonine by PKA (4,5). Remarkably, the PKA-phosphorylated I-1 and I-2 selectively and potently inhibit PP1 activity in all eukaryotic cell extracts (6), and thus became key experimental tools by which PP1 activity was differentiated from other “type-2” protein serine/threonine phosphatases. As phospho-I-1 inhibits PP1 activity against a wide variety of phosphoprotein substrates, the PKA activation of I-1 also functions to impose cAMP regulation over proteins that are phosphorylated by protein kinases other than PKA. In this regard, regulation of I-1 activity is pivotal to the hormonal control of cellular functions mediated by these phosphoproteins.

II. Structure and Regulation of Inhibitor-1

I-1(7) was first purified to homogeneity from rabbit skeletal muscle and following PKA phosphorylation, nanomolar concentrations of this phosphoprotein were shown to inhibit the phosphorylase phosphatase activity in rabbit skeletal muscle extracts. By comparison, the unphosphorylated I-1, even at micromolar concentrations, failed to inhibit the same phosphatase activity. Interestingly, the analysis of direct real-time binding suggests that both phosphorylated and unphosphorylated I-1 bind the isolated PP1 catalytic subunit with similar micromolar affinity so that in contrast to the greater than 10,000-fold change in PP1 inhibitory activity, PKA phosphorylation only increases I-1’s affinity for PP1 by 2-4 –fold (8). This paradox can be explained by the finding that phospho-I-1 displays a mixed competitive mode of inhibition of PP1 catalytic subunits. These data suggest that the pivotal interaction of I-1 with PP1 occurs at a non-catalytic site and this in turn may facilitate the interaction between the phosphorylated threonine-35 and the PP1 catalytic site to inhibit enzyme activity.

Solutions of purified I-1 are resistant to boiling, treatment with trichloroacetic acid and a variety of organic solvents (7). In contrast, I-1 is readily inactivated by incubation with proteases, such as trypsin. Surprisingly, digestion of I-1 with CNBr yielded a N-terminal fragment that was equipotent as the intact WT I-1 protein as a PP1 inhibitor (9). Together, these studies established that I-1 was a largely disordered protein with the key PP1 inhibitory domain localized within the N-terminal 50 amino acids that encompassed the threonine-35 phosphorylated by PKA. During the elucidation of the complete primary sequence (171 amino acids) of rabbit skeletal muscle I-1 (8), a number of active and inactive I-1 phosphopeptides were identified. These studies confirmed that threonine-35 phosphorylation was by itself insufficient to inhibit PP1 activity at nanomolar concentrations and suggested that an N-terminal sequence, 9KIQF13, were also required for enzyme inhibition. The human I-1 shares greater 90% sequence identity with frog, mouse, rat, rabbit, dog and frog I-1 (Figure 1), within the N-terminal 50 amino acids, emphasizing the functional importance of this domain in PP1 inhibition. By contrast, the I-1 C-terminus shows much less conservation of its primary structure, suggesting that it plays a lesser role in regulating PP1 functions in different species.

Figure 1 . Comparison of Primary Structure of Inhibitor-1. Panel A shows a schematic of the structure of the three I-1 isoforms expressed in mammalian tissues (40). They share a N-terminal domain (amino acids 1-61, colored yellow) that contains the PP1 KIQF binding site and threonine-35 phosphorylated by PKA. I-1α is generated by splicing of exon 5 and results in the elimination of C-terminal amino acids 83-134. By contrast, I-1β is produced by the removal of both exons 4 and 5, resulting in frame-shift that produces a unique C-terminus (colored blue). Panel B shows the alignment of predicted primary sequences of human (GI:2499742), chimpanzee (LOC456191), dog (GI: 622900317), rabbit (GI:124834), mouse (GI:10946734), rat (GI:112225) and frog, Xenopus laevis (GI:32442464 or 49116739) I-1 proteins. An unnamed protein product (GI:345296381) is predicted to be the human I-1β, differing from human WT I-1 by the presence of a unique sequence C-terminal to amino acid 61. The peptide sequence predicted from the analysis of chimpanzee gene (NCBI Unigene) suggests that this most likely represents the I-1α polypeptide lacking internal residues 83-137.

Greengard and colleagues (10) identified a structural homologue of I-1 in mammalian brain. This protein termed DARPP-32 (Dopamine and cAMP-regulated phosphoprotein of apparent Mr 32,000) is a larger polypeptide (204 amino acids) than I-1 but shares extensive sequence homology only within in the N-terminal 40 amino acids of I-1. However, like I-1, DARPP-32 becomes a potent and selective inhibitor of PP1 activity after it has been stoichiometrically phosphorylated on the conserved threonine-34 by PKA. Following the cloning of cDNAs encoding I-1 (11) and DARPP-32 (12), detailed structure-function studies were undertaken with recombinant I-1 (8,13) and DARPP-32 (14) proteins. These studies firmly established the dual requirement of the KIQF sequence and the phosphorylated threonine for potent PP1 inhibition by both phosphoproteins. Interestingly, I-1 and DARPP-32 are both expressed in some tissues, such as regions of the brain (15,16), kidney (17) and pancreatic β-cells (18). This raises the intriguing possibility that in addition to transducing cAMP signals, the two proteins may recruit additional different physiological signals to regulate PP1 activity. In this regard, DARPP-32 and I-1 show virtually no sequence homology with their C-termini, where DARPP-32 shows a lengthy stretch of acidic residues.

I-1 as purified from rabbit skeletal muscle is extensively phosphorylated on serine-67 (19). As the purified protein is totally unphosphorylated at threonine-35 and lacks PP1 inhibitory activity, I-1 phosphorylation at serine-67 clearly plays a role other than directly regulating PP1 activity. However, more recent studies (20,21) suggested that a proline-directed Cyclin-dependent kinase 5, Cdk5, phosphorylates I-1 at serine-67 in nervous tissues and results in PP1 inhibition via a PKA-independent mechanism. While this finding remains controversial (20), I-1 phosphorylation at serine-67 also attenuates its subsequent phosphorylation at threonine-T35 by PKA. Thus, the interplay of these two protein kinases may combine to control I-1 function during hormonal control of PP1 activity in some mammalian tissues. By comparison, DARPP-32 is not only phosphorylated by Cdk5 (22) but also by casein kinase-I (23,24) and casein kinase-II (25) within its C-terminal sequence. Some or all of these phosphorylations appear to regulate the phosphorylation state of DARPP-32 at threonine-34 (24,25), possibly by modulating its dephosphorylation by type-2 protein phosphatases. Thus, multiple kinases combine to regulate the PP1 inhibitory activity of DARPP-32.

PP1 does not exist in cells as a free catalytic subunit but is bound to one of up 60 different regulatory subunits (26,27) that function to localize PP1 at specific subcellular sites and/or focus its activity to dephosphorylate a subset of cellular substrates. The structure-function analysis of several PP1 regulators has highlighted the presence of a conserved R/KI/VXF motif that is essential for PP1 binding. Co-crystallization of recombinant PP1 catalytic subunit with a synthetic peptide encoding the PP1-binding motif from GM, the skeletal muscle glycogen-targeting subunit, shows the association of this peptide in an extended manner within a surface hydrophobic pocket located diagonally opposite to the catalytic site on the PP1 catalytic subunit (28). This hydrophobic pocket is occupied by the PP1-binding motif in a very similar manner even in the context of a larger polypeptide representing a major portion of the myosin-targeting subunit (29). As many PP1 regulators dock at this common site, they potentially compete with I-1 for PP1 binding and cellular PP1 complexes containing some regulators are indeed resistant to inhibition by I-1. Thus, cells have evolved at least two other mechanisms for regulating these PP1 complexes. Some PP1 regulators, like GM, are phosphorylated by PKA at serines within or adjacent to the PP1-binding motif. This results in their reduced affinity for the PP1 catalytic subunit. Thus, PKA-mediated phosphorylation coordinates the disassembly of some PP1 complexes, which, by itself, may attenuate their physiological functions but with the parallel phosphorylation and activation of I-1 (Figure 2, Model 1). Thus, these PP1 complexes may be even more rapidly or completely inhibited following the hormone stimulus. Yet other PP1 regulators, like GADD34 (30), bind both PP1 and I-1 (13) and thereby, circumvent the requirement for I-1 binding at the hydrophobic pocket (Figure, Model 2). In this manner, the heterotrimeric complex of PP1/GADD34/I-1 may mediate the cAMP inhibition of protein translation (31). In this regard, a number of I-1 binding proteins have been identified (13) but their role in modulating I-1’s function as a PP1 inhibitor has not been investigated.

Figure 2. Signaling Pathways Regulating I-1 Function in Mammalian Tissues. The association of circulating hormone (red) with a seven transmembrane GPCR (G-protein-coupled receptor, colored green) results in the dissociation of the heterotrimeric GTP-binding protein, Gs (light blue). The αs subunit binds the effector molecule, adenylyl cyclase (blue) and enhances its activity to elevate intracellular cAMP. Cyclic AMP activates PKA and promotes the phosphorylation of inactive I-1 (green) to generate a potent PP1 inhibitor (red). I-1 in turn targets PP1 complexes, localized in distinct subcellular compartments such as the actin cytoskeleton (red lines) or endoplasmic reticulum (yellow). Complexes containing regulators ( R ), like GM and neurabin, at or near the PP1 binding site (Model 1). This results in the dissociation of PP1 catalytic subunit (orange), which is then more effectively inhibited by the activated I-1. Alternately, other PP1 complex may either facilitate I-1 binding, as described for GADD34, or may not compete with I-1 for binding at the surface hydrophobic pocket and thus, are permissive to regulation by I-1 (Model 2). PKA may also phosphorylate cell surface calcium channels to elicit an influx of calcium, which may activate PP2B (or calcineurin) to dephosphorylate I-1 and reset the above regulatory mechanism.

Use of I-1 to selectively inhibit PP1 activity has facilitated the identification of a number of I-1-resistant type-2 protein serine/threonine phosphatases. Of these, PP2B (also known as calcineurin), a calcium/calmodulin-regulated protein serine/threonine phosphatase and PP2A are very effective I-1 phosphatases in vitro and may reverse the activity of I-1 following a hormone stimulus (Figure 2). The dephosphorylation of I-1 by PP2B may also represent a focal point for antagonism between the second messengers, cAMP and Ca²⁺, as has been noted in some mammalian tissues.

III. Inhibitor-1 Expression in Mammalian Tissues

I-1 is widely expressed in mammalian tissues with the highest levels of the protein seen in skeletal and smooth muscle, kidney and brain (32). Northern analyses revealed two I-1 mRNAs (1.8 and 0.7 kb) in all rat tissues (11). In contrast to the high levels seen skeletal muscle, brain and kidney, much lower levels of I-1 mRNA were noted in heart, lung and no I-1 mRNA was detected in liver. Several studies have also highlighted the absence of I-1 protein in rat and mouse liver although the protein was readily detected in livers from other species, including human, sheep and pig (16,32). While the I-1 mRNA could not be detected in rat liver by Northern hybridization, the more sensitive RT-PCR established the presence of I-1 mRNA in this tissue, albeit at much lower levels than all other rat tissues (approximately 12 copies per liver cell compared to 425 copies per neuron and >2500 copies per muscle cell, 33).

In the brains of both rodents and primates, I-1 is highly enriched in the cortex and striatum, with little I-1 expression seen in cerebellar purkinje cells, thalamus, and brain stem. Given its proposed role in regulating PP1 activity during long-term potentiation (34-36), prior studies have questioned the presence of I-1 protein in the rat hippocampus, specifically the CA1 region. In particular, immunocytochemical studies in rat (15,37), and rhesus monkey (15) demonstrated high levels of I-1 in the dentate gyrus, but observed only trace I-1 expression in CA1 neurons. By comparison, immunoprecipitation/immunoblotting studies clearly identified I-1 protein in rat hippocampal CA1 neurons (38) and in situ hybridization (39) demonstrated the presence of I-1 mRNA in adult rat hippocampus in both CA1 and CA3 hippocampal neurons. In fact, Sakagami et al. (39) noted that while I-1 mRNA expression in the adult rat cortex was relatively laminar that of PP1 mRNA was more diffuse, suggesting a restricted role for I-1 in regulating PP1 activity in specific regions of the cortex. To this point, PP1 mRNA is strongly expressed in brain stem nuclei where only trace amounts of I-1 mRNA are seen and conversely, I-1 mRNA is highly abundant in the arachnoid membrane where PP1 mRNA is poorly expressed. These findings emphasize the tremendous variability in the expression of the I-1 regulatory mechanism in controlling PP1 functions in different tissues.

Both I-1 mRNA (39) and protein (16) are observed in the rat brain only after birth, suggesting developmental regulation of the rat I-1 gene. However, studies in the mouse have observed I-1 expression shortly after blastocyst formation and suggested that I-1 may function as a marker for the formation of the mesothelium in the developing mouse embryo (17). The expression of I-1 protein in the developing rodent brain precedes that of DARPP-32 and is detected as early as E9 (embryonic day 9) stage in the mouse and E15 in rat. Interestingly, the levels of I-1 rise postnatally, peaking at P7 (postnatal day 7). Subsequently, I-1 protein levels in cerebellum, hippocampus, and cortex steadily decline until approximately 3 weeks of age. In the hippocampus, the decline in I-1 is transient and recovers by 6-10 weeks of age. By comparison, cortical and cerebellar I-1 mRNA levels in mouse are relatively constant at P7, P14 and adult tissues (40). Most remarkable postnatal decline in I-1 mRNA is seen in the heart and testes of the developing mouse so that I-1 mRNA is virtually undetectable by RT-PCR in the adult testes. This suggests that the I-1 gene may be developmentally regulated with different temporal profiles in different species and different tissues. This results in the wide range of expression of I-1 mRNA and protein observed in tissues of the adult animal.

FISH (fluorescence in-situ hybridization) using a BAC clone of human I-1 gene and Southern analysis of mouse/human hybrid cells lines suggested the presence of two distinct I-1 genes (In addition, a cDNA encoding an unidentified polypeptide – Figure 1- showed extensive sequence homology with the N-terminal of human I-1 and predicted to function as a PP1 inhibitor. This may point to yet another gene encoding an I-1-like protein), one being on human chromosome 12 and another on chromosome 11 (S. Shenolikar - unpublished observations). However, the principal I-1 mRNA seen in most mammalian tissues appears to the product of the human gene located at 12q13.2. Consistent with this, disruption of the homologous mouse I-1 gene (on chromosome 15) results in complete loss of I-1 mRNA and protein in all tissues (35). This and other characteristics of this gene suggest that the second I-1 gene is most likely a pseudogene and is not actively transcribed.

Recent studies (41) suggested that the human I-1 mRNA is alternatively spliced to generate three distinct mRNAs (termed I-1WT, I-1α, and I-1ß). While the largest mRNA encodes full-length or WT I-1, splicing that results in the deletion of exon 5 yields the I-1α mRNA, encoding a polypeptide that lacks the C-terminal amino acids, 83-134. By comparison, the I-1ß mRNA lacks both exons 4 and 5 and results in a frameshift at residue 61 to create I-1 polypeptide with a unique 71 amino acid C-terminus lacking any homology with WT or I-1α. Semi-quantitative PCR showed that in addition to WT I-1 mRNA, the I-1α and I-1ß mRNAs were expressed in all mouse tissues, although levels of the two splice variants were significantly lower than WT I-1 mRNA. As with WT I-1, neither splice variant was detected in mouse liver. Surprisingly, I-1ß mRNA levels increased postnatally in the heart of the developing mice while the levels of WT I-1 and I-1α mRNAs steadily declined in this tissue. The functional significance of I-1 alternative splicing is still unclear but biochemical studies suggested the I-1α and I-1β proteins possessed reduced activity as PP1 inhibitors. Moreover, lacking critical C-terminal sequences that bind GADD34, I-1α and I-1β are unable to regulate the activity of the PP1/GADD34 complex and regulate eIF2α dephosphorylation. In this regard, the temporal and tissue-specific regulation of I-1 mRNA splicing may control the expression of I-1 isoforms in mammalian tissues. Finally, emerging evidence suggests that I-1 mRNA expression may also be influenced by disease state (discussed below). Thus, the myocardium of failing human hearts displayed reduced I-1 content and increased PP1 activity (42,43), which contributed the heart failure.

IV. Inhibitor-1 Transduces cAMP Signals to Regulate Cardiac Contractility

β-adrenergic agonists enhance cAMP signaling to increase cardiac contractility. This is achieved at least in part by the increased phosphorylation of several PKA substrates including troponin-1, phospholambam, ryanodine receptor and voltage-gated calcium channels (42,43). While cardiac muscle contained less I-1 protein than other tissues, cell fractionation studies suggested that greater than 78% of cardiac I-1 was present in the sarcoplasmic reticulum (43). This somewhat surprising finding overturned a widely-held view that I-1 was a largely cytosolic protein. The findings also suggested that I-1 in cardiac tissue might primarily be responsible for the regulation of PP1 substrates in sarcoplasmic reticulum (SR). In this regard, phosphorylation of the SR protein, phospholamban, is closely associated with increased cardiac contractility (44) and mutations in the human phospholamban gene (45,46) result in cardiomyopathy and heart failure. Interestingly, activity of the SR-associated PP1, the principal phospholamban phosphatase, is increased in human and animal models of heart failure. Thus, recent studies that defined PP1 as a negative regulator of cardiac function (47),demonstrated that cardiac function was impaired in the I-1 null mouse. Similarly, failing human hearts show a marked reduction in their content of I-1 protein and show increased PP1 activity (42). This suggests that I-1 plays a key role in suppressing PP1 activity and regulating contractile function in the human heart. Remarkably, a recent study that reported a significant decrease in both total and phosphorylated I-1 in failing human heart, demonstrated that introduction of an activated I-1 protein in cardiomyocytes from the failing human heart largely redressed their contractile response to β-agonists. Moreover, expression of the activated I-1 in mice also enhanced cardiac function (48) and protected the transgenic mouse hearts from insults that induce heart failure. Together, these studies firmly placed I-1 as a critical regulator of the physiological and pathophysiological pathways that control cardiac function. The above studies showed that I-1 expression preferentially increased phospholamban phosphorylation at serine-16 and threonine-17 and other cardiac phosphoproteins, such as troponin-C and ryanodine receptor, were essentially unchanged. Elevation of diastolic calcium activates PP2B in the cardiac muscle, which dephosphorylates and inactivates I-1. This increases cardiac PP1 activity and promotes the dephosphorylation of phospholamban. This in turn inhibits SR calcium transport and cardiac contractility (El-Armouche, 2004). These data strongly point a key role for the PP1/I-1 mechanism in regulating cardiac function.

Interestingly, recent studies showed that PKCα may also phosphorylate and activate I-1 in the mouse heart (49). This raises the possibility that I-1 may also play a role in amplifying PKC signaling in the mammalian heart.

V. Inhibitor-1 Transduces cAMP Signals to Regulate Hippocampal Plasticity

Learning and memory is associated with the modulation of kinase-phosphatase signaling in the postsynaptic neuron to control the plasticity of the mammalian synapse. Long-term potentiation (LTP) and long-term depression (LTD) in the CA1 region of hippocampus are linked with lasting changes in kinase and phosphatase activities following an increase in intracellular calcium induced by the excitatory neurotransmitter, glutamate, acting on NMDA receptors. While the activation of CaMKII, PKA, PKC and ERK, has been linked with the induction of LTP (50-53), PP1 and PP2B regulate the reversal of phosphorylations catalyzed by these kinases to elicit LTD.

In hippocampus, PP1 activity associated with the cytoskeletal compartment known as the postsynaptic density that underlies the postsynaptic membrane rapidly dephosphorylates CaMKII (threonine-286). PKA activity is essential to extend the duration of CaMKII phosphorylation and elicit LTP. Electrophysiological studies suggested that influx of calcium through the NMDA receptors not only activates CaMKII but also the Ca²⁺/calmodulin-activated adenylyl cyclase present in the hippocampal synapse. This increases synaptic cAMP levels and activates PKA. I-1 is among the identified postsynaptic substrates of PKA (34,36). The increased phosphorylation of I-1 and concomitant inhibition of postsynaptic PP1 activity prolongs CaMKII activity and induces LTP.

In contrast, LTD is experimentally induced by low frequency stimuli, which result modest increases in synaptic calcium than those seen following LTP-inducing stimuli. The smaller rise in intracellular calcium preferentially activates the calcium/calmodulin-activated phosphatase, PP2B, which has a 100-fold higher affinity for calcium/calmodulin than CaMKII. Treatment of hippocampal slices with PP2B inhibitors, FK506 and cyclosporin A, inhibits LTD induction (54), demonstrating the requirement for this phosphatase in regulating synaptic plasticity. Experimental evidence suggests that I-1 is a key substrate of PPB in the hippocampal synapse as treatment with cell-permeable cAMP analogs to activate endogenous I-1 or intracellular injection of an active thiophosphorylated I-1, mimics the PP2B inhibitors and inhibits LTD. These studies demonstrated the existence of a postsynaptic phosphatase cascade consisting of PP2B, which by dephosphorylating I-1, increases PP1 activity and dephosphorylates substrates such as CaMKII, Stargazin (55,56) and the GluR1 subunit of the AMPA receptor (57) to elicit LTD. PP1-dependent dephosphorylation of GluR1 at Ser845 has been suggested to promote the intracellular retention of AMPAR following endocytosis (58-60) and thus, reduce the sensitivity of the postsynaptic neurons to the excitatory neurotransmitter, glutamate, and induce LTD. Consistent with this model, PP1 is highly concentrated in the hippocampal synapse and specifically the postsynaptic density. Recent studies identified the neuronal actin-binding proteins, neurabin-1 and neurabin-II/spinophilin as two key PP1 regulators that target this enzyme to the postsynaptic density (61,62). Remarkably, LTD-inducing stimuli increase the translocation or concentration of PP1 in spines of postsynaptic neurons. Furthermore, disruption of synaptic PP1 complexes by loading neurons with competing PP1-binding peptides, including that derived from I-1 containing the KIQF motif, inhibits LTD (63). The synaptic neurabin/PP1 complex has also been implicated in the enhanced trafficking of the AMPA GluR1 subunit to the postsynaptic membrane (64). Moreover, PKA activation promotes the phosphorylation of neurabin at a serine immediately adjacent to the KIKF PP1-binding motif (62,65), and results in decreased affinity of neurabin for PP1 (Figure 2, Model 1). Thus, as discussed above for GM, PKA activation coordinates the phosphorylation of neurabin with activation of I-1 to regulate AMPA receptor trafficking and LTD.

In the face of above data, analysis of LTP in CA1 hippocampal region in the I-1 null mouse was surprisingly normal. This suggests the presence of redundant PKA signaling mechanisms, such as PKA-mediated disruption of the PP1/neurabin complex, which may be sufficient to sustain LTP in CA1 neurons. On the other hand, LTP in the I-1 knockout mouse was elicited using only a single stimulation paradigm, which may not have induced PKA activation of I-1 and the potential contribution of I-1 in induction of some forms of LTP may have been overlooked. In contrast to hippocampus, LTP was significantly impaired in CA3 and dentate gyrus of the I-1 null mouse. These brain regions expressed higher levels of I-1 and may lack some of the redundant mechanisms for cAMP signaling (Allen and Greengard, 2000).

Finally, aging in rodents is associated with increased PP1 activity in the hippocampus and inability to store memory following repeated training (66). Expression of an activated I-1 in the mouse brain cortex suppressed PP1 activity and the aging transgenic mice retained full capacity to store spatial memory. This suggested that the I-1/PP1 mechanism not only controls synaptic plasticity in the hippocampal slice but also regulates the acquisition and storage of long-term memory in the intact animal.

VI. Conclusion

Since the discovery of I-1 as a hormonally regulated PP1 inhibitor in skeletal muscle, studies over the last two decades have established that I-1 functions to inhibit specific PP1 complexes in many mammalian tissues and thus, provide selective control of physiological events in response to cAMP. Above, we have reviewed the role of I-1 in regulating cardiac contractility, neuronal plasticity and yet other studies suggest that I-1 may regulate glycogen metabolism in skeletal muscle (47). In all these cases, I-1 functions to amplify cAMP signaling and control the size and duration of the physiological response.

In addition to the well-characterized activation by PKA, I-1 is also phosphorylated by PKC, PKG, Cdk5 and possibly other kinases. The role of these phosphorylations in controlling I-1 function remains to be investigated. Studies of I-1 have provided the foundation for the discovery of several additional PP1 inhibitors, whose activity is also regulated by reversible phosphorylation (67). Thus, activation of the PHI family of PP1 inhibitors by PKC enhance PKC signaling and regulates smooth muscle contraction (68). In general, the parallel or coordinate control of kinases and phosphatase by regulated PP1 inhibitors provides a very robust mechanism to exercise tight control over the cellular response to hormones and other physiological stimuli. While considerable progress has been made in understanding the biochemical mechanisms that regulate I-1 function, the role of numerous I-1-binding proteins and mechanisms that regulate I-1 expression are still poorly understood. In this regard, continued studies of I-1 structure and function should provide an important conceptual framework for understanding of the physiological functions of this and other PP1 regulators.

Acknowledgements

Work in this laboratory is supported by NIH grants DK52054 and NS41063 (to SS).

References

1. Sutherland, E. W. (1972) Science 177(47), 401-408

2. Krebs, E. G. (1989) JAMA 262(13), 1815-1818

3. Paris, H., Ganapathi, M. K., Silberman, S. R., Aylward, J. H., and Lee, E. Y. (1984) J Biol Chem 259(12), 7510-7518

4. Huang, F. L., and Glinsmann, W. H. (1976) Eur J Biochem 70(2), 419-426

5. Huang, F. L., and Glinsmann, W. (1976) FEBS Lett 62(3), 326-329

6. Ingebritsen, T. S., and Cohen, P. (1983) Science 221(4608), 331-338

7. Nimmo, G. A., and Cohen, P. (1978) Eur J Biochem 87(2), 341-351

8. Endo, S., Zhou, X., Connor, J., Wang, B., and Shenolikar, S. (1996) Biochemistry 35(16), 5220-5228

9. Shenolikar, S., Foulkes, J. G., and Cohen, P. (1978) Biochem Soc Trans 6(5), 935-937

10. Williams, K. R., Hemmings, H. C., Jr., LoPresti, M. B., Konigsberg, W. H., and Greengard, P. (1986) J Biol Chem 261(4), 1890-1903

11. Elbrecht, A., DiRenzo, J., Smith, R. G., and Shenolikar, S. (1990) J Biol Chem 265(23), 13415-13418

12. Kurihara, T., Lewis, R. M., Eisler, J., and Greengard, P. (1988) J Neurosci 8(2), 508-517

13. Connor, J. H., Weiser, D. C., Li, S., Hallenbeck, J. M., and Shenolikar, S. (2001) Mol Cell Biol 21(20), 6841-6850

14. Hemmings, H. C., Jr., Nairn, A. C., Elliott, J. I., and Greengard, P. (1990) J Biol Chem 265(33), 20369-20376

15. Barbas, H., Gustafson, E. L., and Greengard, P. (1993) J Comp Neurol 334(1), 1-18

16. Hemmings, H. C., Jr., Girault, J. A., Nairn, A. C., Bertuzzi, G., and Greengard, P. (1992) J Neurochem 59(3), 1053-1061

17. McLaren, L., Boyle, S., Mason, J. O., and Bard, J. B. (2000) Mech Dev 96(2), 237-241

18. Lilja, L., Meister, B., Berggren, P. O., and Bark, C. (2005) Biochem Biophys Res Commun 329(2), 673-677

19. Aitken, A., Bilham, T., and Cohen, P. (1982) Eur J Biochem 126(2), 235-246

20. Bibb, J. A., Nishi, A., O'Callaghan, J. P., Ule, J., Lan, M., Snyder, G. L., Horiuchi, A., Saito, T., Hisanaga, S., Czernik, A. J., Nairn, A. C., and Greengard, P. (2001) J Biol Chem 276(17), 14490-14497

21. Huang, K. X., and Paudel, H. K. (2000) Proc Natl Acad Sci U S A 97(11), 5824-5829

22. Bibb, J. A., Snyder, G. L., Nishi, A., Yan, Z., Meijer, L., Fienberg, A. A., Tsai, L. H., Kwon, Y. T., Girault, J. A., Czernik, A. J., Huganir, R. L., Hemmings, H. C., Jr., Nairn, A. C., and Greengard, P. (1999) Nature 402(6762), 669-671

23. Desdouits, F., Cohen, D., Nairn, A. C., Greengard, P., and Girault, J. A. (1995) J Biol Chem 270(15), 8772-8778

24. Desdouits, F., Siciliano, J. C., Greengard, P., and Girault, J. A. (1995) Proc Natl Acad Sci U S A 92(7), 2682-2685

25. Girault, J. A., Hemmings, H. C., Jr., Williams, K. R., Nairn, A. C., and Greengard, P. (1989) J Biol Chem 264(36), 21748-21759

26. Ceulemans, H., Stalmans, W., and Bollen, M. (2002) Bioessays 24(4), 371-381

27. Cohen, P. T. (2002) J Cell Sci 115(Pt 2), 241-256

28. Egloff, M. P., Johnson, D. F., Moorhead, G., Cohen, P. T., Cohen, P., and Barford, D. (1997) EMBO J 16(8), 1876-1887

29. Terrak, M., Kerff, F., Langsetmo, K., Tao, T., and Dominguez, R. (2004) Nature 429(6993), 780-784

30. He, B., Gross, M., and Roizman, B. (1998) J Biol Chem 273(33), 20737-20743

31. Brush, M. H., Weiser, D. C., and Shenolikar, S. (2003) Mol Cell Biol 23(4), 1292-1303

32. MacDougall, L. K., Campbell, D. G., Hubbard, M. J., and Cohen, P. (1989) Biochim Biophys Acta 1010(2), 218-226

33. Aleem, E. A., Flohr, T., Hunziker, A., Mayer, D., Bannasch, P., and Thielmann, H. W. (2001) Mol Cell Biochem 217(1-2), 1-12

34. Brown, G. P., Blitzer, R. D., Connor, J. H., Wong, T., Shenolikar, S., Iyengar, R., and Landau, E. M. (2000) J Neurosci 20(21), 7880-7887

35. Allen, P. B., Hvalby, O., Jensen, V., Errington, M. L., Ramsay, M., Chaudhry, F. A., Bliss, T. V., Storm-Mathisen, J., Morris, R. G., Andersen, P., and Greengard, P. (2000) J Neurosci 20(10), 3537-3543

36. Blitzer, R. D., Connor, J. H., Brown, G. P., Wong, T., Shenolikar, S., Iyengar, R., and Landau, E. M. (1998) Science 280(5371), 1940-1942

37. Gustafson, E. L., Girault, J. A., Hemmings, H. C., Jr., Nairn, A. C., and Greengard, P. (1991) J Comp Neurol 310(2), 170-188

38. Lowenstein, P. R., Shering, A. F., MacDougall, L. K., and Cohen, P. (1995) Brain Res 676(1), 80-92

39. Sakagami, H., Ebina, K., and Kondo, H. (1994) Brain Res Mol Brain Res 25(1-2), 7-18

40. Weiser, D. C., Sikes, S., Li, S., and Shenolikar, S. (2004) J Biol Chem 279(47), 48904-48914

41. Liu, H. T., Lin, T. H., Kuo, H. C., Chen, Y. C., Tsay, H. J., Jeng, H. H., Tsai, P. C., Shie, F. K., Chen, J. H., and Huang, H. B. (2002) Biochem Biophys Res Commun 291(5), 1293-1296

42. El-Armouche, A., Pamminger, T., Ditz, D., Zolk, O., and Eschenhagen, T. (2004) Cardiovasc Res 61(1), 87-93

43. Gupta, R. C., Mishra, S., Rastogi, S., Imai, M., Habib, O., and Sabbah, H. N. (2003) Am J Physiol Heart Circ Physiol 285(6), H2373-2381

44. El-Armouche, A., Rau, T., Zolk, O., Ditz, D., Pamminger, T., Zimmermann, W. H., Jackel, E., Harding, S. E., Boknik, P., Neumann, J., and Eschenhagen, T. (2003) FASEB J 17(3), 437-439

45. Haghighi, K., Kolokathis, F., Pater, L., Lynch, R. A., Asahi, M., Gramolini, A. O., Fan, G. C., Tsiapras, D., Hahn, H. S., Adamopoulos, S., Liggett, S. B., Dorn, G. W., 2nd, MacLennan, D. H., Kremastinos, D. T., and Kranias, E. G. (2003) J Clin Invest 111(6), 869-876

46. Song, Q., Schmidt, A. G., Hahn, H. S., Carr, A. N., Frank, B., Pater, L., Gerst, M., Young, K., Hoit, B. D., McConnell, B. K., Haghighi, K., Seidman, C. E., Seidman, J. G., Dorn, G. W., 2nd, and Kranias, E. G. (2003) J Clin Invest 111(6), 859-867

47. Carr, A. N., Schmidt, A. G., Suzuki, Y., del Monte, F., Sato, Y., Lanner, C., Breeden, K., Jing, S. L., Allen, P. B., Greengard, P., Yatani, A., Hoit, B. D., Grupp, I. L., Hajjar, R. J., DePaoli-Roach, A. A., and Kranias, E. G. (2002) Mol Cell Biol 22(12), 4124-4135

48. Pathak, A., del Monte, F., Zhao, W., Schultz, J. E., Lorenz, J. N., Bodi, I., Weiser, D., Hahn, H., Carr, A. N., Syed, F., Mavila, N., Jha, L., Qian, J., Marreez, Y., Chen, G., McGraw, D. W., Heist, E. K., Guerrero, J. L., DePaoli-Roach, A. A., Hajjar, R. J., and Kranias, E. G. (2005) Circ Res 96(7), 756-766

49. Braz, J. C., Gregory, K., Pathak, A., Zhao, W., Sahin, B., Klevitsky, R., Kimball, T. F., Lorenz, J. N., Nairn, A. C., Liggett, S. B., Bodi, I., Wang, S., Schwartz, A., Lakatta, E. G., DePaoli-Roach, A. A., Robbins, J., Hewett, T. E., Bibb, J. A., Westfall, M. V., Kranias, E. G., and Molkentin, J. D. (2004) Nat Med 10(3), 248-254

50. Klann, E., Chen, S. J., and Sweatt, J. D. (1993) Proc Natl Acad Sci U S A 90(18), 8337-8341

51. Kanterewicz, B. I., Urban, N. N., McMahon, D. B., Norman, E. D., Giffen, L. J., Favata, M. F., Scherle, P. A., Trzskos, J. M., Barrionuevo, G., and Klann, E. (2000) J Neurosci 20(9), 3057-3066

52. Yasuda, H., Barth, A. L., Stellwagen, D., and Malenka, R. C. (2003) Nature Neurosci 6(1), 15-16

53. Malinow, R., Schulman, H., and Tsien, R. W. (1989) Science 245(4920), 862-866

54. Mulkey, R. M., Endo, S., Shenolikar, S., and Malenka, R. C. (1994) Nature 369(6480), 486-488

55. Tomita, S., Adesnik, H., Sekiguchi, M., Zhang, W., Wada, K., Howe, J. R., Nicoll, R. A., and Bredt, D. S. (2005) Nature 435(7045), 1052-1058

56. Tomita, S., Stein, V., Stocker, T. J., Nicoll, R. A., and Bredt, D. S. (2005) Neuron 45(2), 269-277

57. Huang, C. C., Liang, Y. C., and Hsu, K. S. (2001) J Biol Chem 276(51), 48108-48117

58. Lee, H. K., Barbarosie, M., Kameyama, K., Bear, M. F., and Huganir, R. L. (2000) Nature 405(6789), 955-959

59. Lee, H. K., Kameyama, K., Huganir, R. L., and Bear, M. F. (1998) Neuron 21(5), 1151-1162

60. Ehlers, M. D. (2000) Neuron 28(2), 511-525

61. Allen, P. B., Ouimet, C. C., and Greengard, P. (1997) Proc Natl Acad Sci U S A 94(18), 9956-9961

62. Oliver, C. J., Terry-Lorenzo, R. T., Elliott, E., Bloomer, W. A., Li, S., Brautigan, D. L., Colbran, R. J., and Shenolikar, S. (2002) Mol Cell Biol 22(13), 4690-4701

63. Morishita, W., Connor, J. H., Xia, H., Quinlan, E. M., Shenolikar, S., and Malenka, R. C. (2001) Neuron 32(6), 1133-1148

64. Terry-Lorenzo, R. T., Roadcap, D. W., Otsuka, T., Blanpied, T. A., Zamorano, P. L., Garner, C. C., Shenolikar, S., and Ehlers, M. D. (2005) Mol Biol Cell 16(5), 2349-2362

65. McAvoy, T., Allen, P. B., Obaishi, H., Nakanishi, H., Takai, Y., Greengard, P., Nairn, A. C., and Hemmings, H. C., Jr. (1999) Biochemistry 38(39), 12943-12949

66. Genoux, D., Haditsch, U., Knobloch, M., Michalon, A., Storm, D., and Mansuy, I. M. (2002) Nature 418(6901), 970-975

67. Oliver, C. J., and Shenolikar, S. (1998) Frontiers Biosci 3, D961-972

68. Eto, M., Senba, S., Morita, F., and Yazawa, M. (1997) FEBS Lett 410(2-3), 356-360