Protein phosphatase 5 in signal transduction

Evidence that protein phosphatase 5 functions to negatively modulate the maturation of the Hsp90-dependent heme-regulated eIF2alpha kinase

The maturation and activation of newly synthesized molecules of the heme-regulated inhibitor of protein synthesis (HRI) in reticulocytes require their functional interaction with Hsp90. In this report, we demonstrate that protein phosphatase 5 (PP5), a previously documented component of the Hsp90 chaperone machine, is physically associated with HRI maturation intermediates. The interaction of PP5 with HRI is mediated through Hsp90, as mutants of PP5 that do not bind Hsp90 do not interact with HRI. PP5 was also present in Hsp90 heterocomplexes with another Hsp90 cohort, p50(cdc37), and expression of newly synthesized HRI enhanced the amount of p50(cdc37) associated with Hsp90/PP5-HRI heterocomplexes. The functional significance of the interaction of PP5 with Hsp90-HRI heterocomplexes was examined by characterizing the effects of compounds that impact PP5 activity in vitro. The protein phosphatase inhibitors okadaic acid and nodularin enhanced the kinase activity of HRI when applied during HRI maturation/activation, while the PP5 activators arachidonic and linoleic acid repressed HRI activity when applied during HRI maturation/activation. However, application of these compounds after HRI's "transformation" to an Hsp90-independent form did not similarly impact HRI's kinase activity. Furthermore, the Hsp90 inhibitor geldanamycin negated the effects of phosphatase inhibitors on HRI maturation/activation. The finding that PP5 downregulates an Hsp90-dependent process supports models for regulated Hsp90 function and describes a novel potential substrate for PP5 function in vivo.

Aurora-B associated protein phosphatases as negative regulators of kinase activation

The human serine/threonine kinase Aurora-B is structurally related to the protein kinase Ipl1p from S cerevisiae and aurora from Drosophila melanogaster, which are key regulators of mitosis. The present study shows that human Aurora-B is activated by okadaic acid and forms complexes with the protein serine/threonine phosphatase type 1 (PP1) or PP2A, but not with PP5. These data identified Aurora-B associated protein phosphatases as negative regulators of kinase activation. We then used a series of substrates based on a histone H3 phosphorylation site (residues 5-15) to determine the substrate specificity of human Aurora-B. We found that this enzyme is an arginine-directed kinase that can phosphorylate histone H3 at serines 10 and 28 in vitro, suggesting that human Aurora-B is a mitotic histone H3 kinase.

Identification of potential physiological activators of protein phosphatase 5

The protein serine/threonine phosphatase designated PP5 has little basal activity, and physiological activators of the enzyme have never been identified. Purified PP5 can, however, be activated by partial proteolysis or by the binding of supraphysiological concentrations of polyunsaturated long-chain fatty acids to its tetratricopeptide repeat (TPR) domain. To test whether activation of PP5 by polyunsaturated but not saturated fatty acids was an artifact of the lower solubility of saturated fatty acids, the effects of fatty acyl-CoA esters were examined. Saturated and unsaturated long-chain fatty acids are both freely water-soluble when esterified to CoA. Long-chain fatty acyl-CoA esters activated PP5 at physiological concentrations, with the saturated compounds being more effective. We investigated the effects of chain length and of the CoA moiety on PP5 activation. Chains of 16 carbons or more were required for optimal activation, with no activation observed below 10 carbons. On the basis of competition studies using acetyl-CoA, the function of the CoA moiety appeared to be to increase solubility of the fatty acyl moiety rather than to interact with a specific binding site. These data suggested that long-chain fatty acid-CoA esters might be physiological activators of PP5 and point to a potential link between fatty acid metabolism and signal transduction via this enzyme. Because heat shock protein 90 is also known to bind to the TPR domain of PP5 via its C-terminal domain (C90), we investigated its effect on PP5 activity. C90 activated the enzyme approximately 10-fold. Thus, we have identified two potential physiological activators of PP5.

Signaling by protein phosphatases in the nucleus

The nucleus contains a large variety of protein phosphatases, which function in key processes such as cell-cycle progression, replication, transcription and RNA processing. Here, we review the pleiotropic action of nuclear protein phosphatases and focus in particular on the underlying signaling strategies. It appears that nuclear protein phosphatases can both mediate and antagonize signaling by protein kinases, sometimes as part of feedback loops. Some protein phosphatases shuttle between the cytoplasm and the nucleus, which enables them to act as signal transducers between both compartments. An emerging theme is the contribution of protein phosphatases to cycles of protein phosphorylation and dephosphorylation that steer the assembly and firing of molecular machines in the nucleus.

Novel protein phosphatases in yeast

During the last decade several novel yeast genes encoding proteins related to the PPP family of Ser/Thr protein phosphatases have been discovered and their functional characterization initiated. Most of these novel phosphatases display intriguing structural features and/or are involved in a number of important functions, such as cell cycle regulation, protein synthesis and maintenance of cellular integrity. While in some cases these genes appear to be restricted to fungi, in others similar proteins can be found in higher eukaryotes. This review will summarize the latest advances in our understanding about how these phosphatases are regulated and fulfil their functions in the yeast cell.

A novel tetratricopeptide repeat (TPR) containing PP5 serine/threonine protein phosphatase in the malaria parasite, Plasmodium falciparum

BACKGROUND

The malarial parasite, Plasmodium falciparum (Pf), is responsible for nearly 2 million deaths worldwide. However, the mechanisms of cellular signaling in the parasite remain largely unknown. Recent discovery of a few protein kinases and phosphatases point to a thriving reversible phosphorylation system in the parasite, although their function and regulation need to be determined.

RESULTS

We provide biochemical and sequence evidence for a protein serine/threonine phosphatase type PP5 in Plasmodium falciparum, and named it PfPP5. The 594-amino acid polypeptide was encoded by a 1785 nucleotide long intronless gene in the parasite. The recombinant protein, expressed in bacteria, was indistinguishable from native PfPP5. Sequencing comparison indicated that the extra-long N-terminus of PfPP5 outside the catalytic core contained four tetratricopeptide repeats (TPRs), compared to three such repeats in other PP5 phosphatases. The PfPP5 N-terminus was required for stimulation of the phosphatase activity by polyunsaturated fatty acids. Co-immunoprecipitation demonstrated an interaction between native PfPP5 and Pf heat shock protein 90 (hsp90). PfPP5 was expressed in all the asexual erythrocytic stages of the parasite, and was moderately sensitive to okadaic acid.

CONCLUSIONS

This is the first example of a TPR-domain protein in the Apicomplexa family of parasites. Since TPR domains play important roles in protein-protein interaction, especially relevant to the regulation of PP5 phosphatases, PfPP5 is destined to have a definitive role in parasitic growth and signaling pathways. This is exemplified by the interaction between PfPP5 and the cognate chaperone hsp90.

Identification of amino acids in the tetratricopeptide repeat and C-terminal domains of protein phosphatase 5 involved in autoinhibition and lipid activation

Protein phosphatase 5 (PP5) exhibits low basal activity due to the autoinhibitory properties of its N-terminal and C-terminal domains but can be activated approximately 40-fold in vitro by polyunsaturated fatty acids. To identify residues involved in regulating PP5 activity, we performed scanning mutagenesis of its N-terminal tetratricopeptide repeat (TPR) domain and deletion mutagenesis of its C-terminal domain. Mutating residues in a groove of the TPR domain that binds to heat shock protein 90 had no effect on basal phosphatase activity. Mutation of Glu-76, however, whose side chain projects away from this groove, resulted in a 10-fold elevation of basal activity without affecting arachidonic acid-stimulated activity. Thus, the interface of the TPR domain involved in PP5 autoinhibition appears to be different from that involved in heat shock protein 90 binding. We also observed a 10-fold elevation of basal phosphatase activity upon removing the C-terminal 13 amino acids of PP5, with a concomitant 50% decrease in arachidonic acid-stimulated activity. These two effects were accounted for by two distinct amino acid deletions: deleting the four C-terminal residues (496-499) of PP5 had no effect on its activity, but removing Gln-495 elevated basal activity 10-fold. Removal of a further three amino acids had no additional effect, but deleting Asn-491 resulted in a 50% reduction in arachidonic acid-stimulated activity. Thus, Glu-76 in the TPR domain and Gln-495 at the C-terminus were implicated in maintaining the low basal activity of PP5. While the TPR domain alone has been thought to mediate fatty acid activation of PP5, our data suggest that Asn-491, near its C-terminus, may also be involved in this process.