Two human 90-kDa heat shock proteins are phosphorylated in vivo at conserved serines that are phosphorylated in vitro by casein kinase II

Amino-terminal protein sequence analysis revealed that exponentially growing human HeLa cells at 37 degrees C express two closely related 90-kDa "heat shock" proteins (hsp 90) in nearly equal amounts. Both hsp 90s begin with proline; the initial methionine residue is removed. The alpha protein contains a 9-amino acid segment, TQTQDQPME, from residues 4 to 12, that is replaced by a 4-amino acid segment, VHHG, in the beta form. The purified hsp 90 mixture contains 2 mol of phosphate/mol of polypeptide. Both hsp 90 proteins are phosphorylated at two homologous sites. For the alpha protein, these sites correspond to serine 231 and serine 263. A 5-amino acid segment, ESEDK, between the two phosphorylation sites is absent from the beta protein. The sequence between phosphorylation sites of both hsp 90s is predicted to have alpha helical structure. Dephosphorylated hsp 90 is phosphorylated at both sites by casein kinase II from HeLa cells, calf thymus, or rabbit reticulocytes; no other hsp 90 residues were phosphorylated by casein kinase II in vitro.     

Possible involvement of the 90-kDa heat shock protein in the regulation of protein synthesis

The heme-sensitive eukaryotic initiation factor (eIF)-2 alpha kinase regulates translational activity in reticulocytes by phosphorylation of the smallest subunit of eukaryotic peptide initiation factor 2, eIF-2. Highly purified preparations of the kinase contain an abundant 90-kDa polypeptide which appears to modulate the activity of the enzyme. The physical properties and structural characteristics of the reticulocyte 90-kDa peptide are similar to those of the 90-kDa heat shock protein (hsp 90) from HeLa and other mammalian cells. The reticulocyte and HeLa cell proteins are shown to be immunologically cross-reactive. A direct comparison of the two proteins by one-dimensional peptide mapping of large peptides generated by limited proteolysis and by reversed-phase high performance liquid chromatography analysis of tryptic peptides indicates that they represent the same protein species. Like the 90-kDa reticulocyte protein, HeLa cell hsp 90 causes increased eIF-2 alpha phosphorylation by the heme-sensitive kinase and is a potent inhibitor of protein synthesis in the reticulocyte lysate system. A potential mechanism for the latter inhibition is inferred. These results implicate hsp 90 in the regulation of protein synthesis via its interaction with and perhaps regulation of the heme-sensitive kinase and phosphorylation of eIF-2 alpha.     

Brassica napus hsp90 can autophosphorylate and phosphorylate other protein substrates

A Brassica napus cDNA encoding the 90 kDa heat shock protein, hsp90, was modified to add 6 histidines at the C-terminus and expressed in insect cells to prepare a recombinant histidine-tagged hsp90. The recombinant protein was purified over Ni²⁺-NTA agarose columns and its identity was confirmed by Western blotting, using a plant hsp90-specific antiserum. Incubation of purified hsp90 with [-³²P] ATP in the presence of Mn²⁺ resulted in its autophosphorylation on serine residues. The purified hsp90 could also phosphorylate other protein substrates such as histones and casein in the presence of Mn²⁺. Analysis of phosphorylated casein revealed that serine residues are phosphorylated by hsp90. This is the first demonstration that a cytosolic hsp90 homolog can phosphorylate other protein substrates.     

The phosphorylation state of the reticulocyte 90-kDa heat shock protein affects its ability to increase phosphorylation of peptide initiation factor 2 alpha subunit by the heme-sensitive kinase

The rabbit reticulocyte Mr 90,000 protein associated with the heme-sensitive eIF-2 alpha kinase has been identified previously as the mammalian heat shock protein of this size class (hsp 90). Purified reticulocyte hsp 90 when added exogenously to the kinase increases its activity. This stimulatory effect is abolished after incubation of hsp 90 with a highly purified type 1 phosphoprotein phosphatase isolated from reticulocytes. Phosphorylation of dephosphorylated hsp 90 by casein kinase II but not by cAMP-dependent protein kinase restores the biological activity of hsp 90 to stimulate eIF-2 alpha phosphorylation.     

The 90-kilodalton peptide of the heme-regulated eIF-2 alpha kinase has sequence similarity with the 90-kilodalton heat shock protein

Highly purified preparations of the heme-controlled eIF-2 alpha (eukaryotic peptide initiation factor 2 alpha subunit) kinase of rabbit reticulocytes contain an abundant 90-kilodalton (kDa) peptide that is immunologically cross-reactive with spectrin and that modulates the activity of the enzyme [Kudlicki, W., Fullilove, S., Read, R., Kramer, G., & Hardesty, B. (1987) J. Biol. Chem. 262, 9695-9701]. The amino-terminal sequence of the 90-kDa protein has a high degree of similarity with the known amino-terminal sequences of the Drosophila 83-kDa heat shock protein (20 out of 22 residues) and with other related heat shock proteins. The amino acid sequence of a tryptic phosphopeptide isolated by high-performance liquid chromatography from the eIF-2 alpha kinase associated 90-kDa protein after phosphorylation by casein kinase II is shown to be identical with a 14 amino acid segment of the known sequence of the Drosophila 83-kDa heat shock protein. Results of hydrodynamic studies indicate a highly elongated structure for the reticulocyte protein, characteristic of a structural protein. Additional structural similarities between the eukaryotic heat shock proteins, the reticulocyte eIF-2 alpha kinase associated 90-kDa peptide, and spectrin are discussed.     

Identification of a cytoplasmic protein kinase regulated by IL-1 that phosphorylates the small heat shock protein, hsp27

IL-1 increases phosphorylation of the small heat shock protein (hsp27) in intact cells. This change was also shown both by introducing [gamma-32P]ATP and Mg2+ into MRC-5 fibroblasts permeabilized by LPC after stimulation by IL-1, and by adding the labeled ATP and Mg2+ to cell extracts. Hsp27 phosphorylated in permeabilized cells or cell extracts was shown by 2D electrophoresis to comprise the three forms seen in metabolically labeled cells, suggesting that the physiologically relevant kinase was acting on the substrate in vitro. Mixing of extracts of resting and IL-1-stimulated cells revealed that stimulated cells contained increased levels of kinase activity that phosphorylated substrate hsp27 in the extracts of resting cells. Existence of the activated kinase was confirmed by showing that extracts of IL-1-stimulated cells phosphorylated purified homogeneous hsp27 at a greater rate than those of resting cells. The kinase activity was maximal in cells stimulated with IL-1 for 5 to 10 min, but had declined to the resting level after stimulation for 40 min. Membrane and cytosolic fractions prepared from cell homogenates both contained hsp27 kinase, but the IL-1-dependent increase was associated with the cytosolic fraction. TNF-stimulated cells also contained increased hsp27 kinase activity in the cytosol. The evidence suggests that the cytosolic hsp27 kinase is responsible for the changes in hsp27 phosphorylation induced by the cytokines in intact cells.     

The hsp90-related protein TRAP1 is a mitochondrial protein with distinct functional properties

The hsp90 family of molecular chaperones was expanded recently due to the cloning of TRAP1 and hsp75 by yeast two-hybrid screens. Careful analysis of the human TRAP1 and hsp75 sequences revealed that they are identical, and we have cloned a similar protein from Drosophila. Immunofluorescence data show that human TRAP1 is localized to mitochondria. This mitochondrial localization is supported by the existence of mitochondrial localization sequences in the amino termini of both the human and Drosophila proteins. Due to the striking homology of TRAP1 to hsp90, we tested the ability of TRAP1 to function as an hsp90-like chaperone. TRAP1 did not form stable complexes with the classic hsp90 co-chaperones p23 and Hop (p60). Consistent with these observations, TRAP1 had no effect on the hsp90-dependent reconstitution of hormone binding to the progesterone receptor in vitro, nor could it substitute for hsp90 to promote maturation of the receptor to its hormone-binding state. However, TRAP1 is sufficiently conserved with hsp90 such that it bound ATP, and this binding was sensitive to the hsp90 inhibitor geldanamycin. In addition, TRAP1 exhibited ATPase activity that was inhibited by both geldanamycin and radicicol. Thus, TRAP1 has functions that are distinct from those of hsp90.     

Cyclic AMP-dependent protein kinase (PKA) and protein kinase C phosphorylate sites in the amino acid sequence corresponding to the M3/M4 cytoplasmic domain of alpha4 neuronal nicotinic receptor subunits

To determine whether alpha4 subunits of alpha4beta2 neuronal nicotinic receptors are phosphorylated within the M3/M4 intracellular region by cyclic AMP-dependent protein kinase A (PKA) or protein kinase C (PKC), immunoprecipitated receptors from Xenopus oocytes and a fusion protein corresponding to the M3/M4 cytoplasmic domain of alpha4 (alpha4(336-597)) were incubated with ATP and either PKA or PKC. Both alpha4 and alpha4(336-597) were phosphorylated by PKA and PKC, providing the first direct biochemical evidence that the M3/M4 cytoplasmic domain of neuronal nicotinic receptor alpha4 subunits is phosphorylated by both kinases. When the immunoprecipitated receptors and the alpha4(336-597) fusion protein were phosphorylated and the labeled proteins subjected to phosphoamino acid analysis, results indicated that alpha4 and alpha4(336-597) were phosphorylated on the same amino acid residues by each kinase. Furthermore, PKA phosphorylated serines exclusively, whereas PKC phosphorylated both serines and threonines. To determine whether Ser(368) was a substrate for both kinases, a peptide corresponding to amino acids 356-371 was synthesized (alpha4(356-371)) and incubated with ATP and the kinases. The phosphorylation of alpha4(356-371) by both PKA and PKC was saturable with K(m)s of 15.3 +/- 3.3 microM and 160.8 +/- 26.8 microM, respectively, suggesting that Ser(368) was a better substrate for PKA than PKC.     

Generation of constitutively active p90 ribosomal S6 kinase in vivo. Implications for the mitogen-activated protein kinase-activated protein kinase family.

p90 ribosomal S6 kinases (RSKs), containing two distinct kinase catalytic domains, are phosphorylated and activated by extracellular signal-regulated kinase (ERK). The amino-terminal kinase domain (NTD) of RSK phosphorylates exogenous substrates, whereas the carboxyl-terminal kinase domain (CTD) autophosphorylates Ser-386. A conserved putative autoinhibitory alpha helix is present in the carboxyl-terminal tail of the RSK isozymes ((697)HLVKGAMAATYSALNR(712) of RSK2). Here, we demonstrate that truncation (Delta alpha) or mutation (Y707A) of this helix in RSK2 resulted in constitutive activation of the CTD. In vivo, both mutants enhanced basal Ser-386 autophosphorylation by the CTD above that of wild type (WT). The enhanced Ser-386 autophosphorylation was attributed to disinhibition of the CTD because a CTD dead mutation (K451A) eliminated Ser-386 autophosphorylation even in conjunction with Delta alpha and Y707A. Constitutive activity of the CTD appears to enhance NTD activity even in the absence of ERK phosphorylation because basal phosphorylation of S6 peptide by Delta alpha and Y707A was approximately 4-fold above that of WT. A RSK phosphorylation motif antibody detected a 140-kDa protein (pp140) that was phosphorylated upon epidermal growth factor or insulin treatment. Ectopic expression of Delta alpha or Y707A resulted in increased basal phosphorylation of pp140 compared with that of WT, presenting the possibility that pp140 is a novel RSK substrate. Thus, it is clear that the CTD regulates NTD activity in vivo as well as in vitro.     

Identification of the phosphorylation sites of the murine small heat shock protein hsp25.

Native phosphorylated mouse small heat shock protein hsp25 from Ehrlich ascites tumor cells was isolated and the in vivo phosphorylation sites of the protein were determined. Furthermore, native hsp25 was phosphorylated by the endogenous kinase(s) in a cell-free system as well as recombinant hsp25 was phosphorylated in vitro by protein kinase C and catalytic subunit of cAMP-dependent protein kinase. The two major phosphorylation sites of native and recombinant hsp25 were determined as Ser-15 and Ser-86. There are no differences in the hsp25 phosphorylation sites phosphorylated by the protein kinase C, the catalytic subunit of cAMP-dependent protein kinase and the unknown intracellular kinase(s). The serine residues identified exist in all known small mammalian stress proteins and are located in the conserved kinase recognition sequence Arg-X-X-Ser.     

Analysis of the native forms of the 90 kDa heat shock protein (hsp90) in plant cytosolic extracts

A polyclonal antibody, R2, was raised against a fusion protein consisting of a portion of plant hsp90 fused to the trpE protein of Escherichia coli. This antibody was found to be specific towards plant hsp90, showing little or no cross-reactivity with mouse and human hsp90 proteins. The R2 antibody identified an 83 kDa protein as the hsp90 homologue in cytosolic extracts of several dicot and monocot plants. Two-dimensional gel electrophoresis indicated that at least two different isoforms of hsp90 are expressed in Brassica napus seedlings. An examination of the native state of hsp90 by non-denaturing gel electrophoresis showed that this protein exists as a monomer, dimer and as a high-molecular-mass complex of ca. 680 kDa in cell extracts of spinach cotyledons and leaves, B. napus seedlings and wheat germ. Native gel analysis and cross-linking studies of purified hsp90 showed that plant hsp90 exists predominantly as a monomer. When 35S-labelled B. napus cytosolic extracts were immunoprecipitated with the R2 antiserum, hsp90 and two additional proteins with approximate molecular masses of 49 and 45 kDa were detected in the immunoprecipitates. These results are consistent with the idea that hsp90:protein heterocomplexes exist in plant cells.     

Regulation of the atrial natriuretic peptide receptor by heat shock protein 90 complexes

Heat shock protein 90 (hsp90) is a chaperone required for the proper folding and trafficking of many proteins involved in signal transduction. We tested whether hsp90 plays a role as a chaperone for GC-A, the membrane guanylate cyclase that acts as a receptor for atrial natriuretic peptide (ANP). When cultured cells expressing recombinant GC-A were treated with geldanamycin, an inhibitor of hsp90 function, the ANP-stimulated production of cyclic GMP was inhibited. This suggested that hsp90 was required for GC-A processing and/or stability. A physical association between hsp90 and GC-A was demonstrated in coimmunoprecipitation experiments. Treatment with geldanamycin disrupted this association and led to the accumulation of complexes containing GC-A and heat shock protein 70 (hsp70). Protein folding pathways involving hsp70 and hsp90 include several pathway-specific co-chaperones. Complexes between GC-A and hsp90 contained the co-chaperone p50(cdc37), typically found associated with protein kinase.hsp90 heterocomplexes. GC-A immunoprecipitates did not contain detectable amounts of Hop, FKBP51, FKBP52, PP5, or p23, all co-chaperones found in hsp90 complexes with other signaling proteins. The association of hsp90 and p50(cdc37) with GC-A was dependent on the kinase homology domain of this receptor but not on its ANP-binding, transmembrane, or guanylate cyclase domains. The data suggest that GC-A is regulated by hsp90 complexes similar to those involved in the maturation of protein kinases.     

Identification of heat shock protein 90-associated 84-kDa phosphoprotein

In eukaryotic cells, HSP90 is associated with several protein kinases and regulates their activities. HSP90 was also reported to possess an autophosphorylase activity. In this study, we examined in vitro autophosphorylation of HSP90, which was purified from chick muscle. We show that HSP90 was not phosphorylated in vitro, but an 84-kDa protein (p84) was highly phosphorylated. P84 was neither HSP90 nor its degradative product, as it was not detected by an antibody (BF4) specific to HSP90 in denaturing immunoprecipitation and Western blot analysis. Phosphorylation of a protein similar to p84 was also detected with purified human brain and HeLa HSP90, indicating that p84 is present in many different types of cells. P84 appeared to exist as large complexes, as determined by HPLC and native gel electrophoresis. Native immunoprecipitation using anti-HSP90 (BF4)-conjugated Affi-gel revealed that this phosphoprotein is specifically associated with HSP90. The interaction of p84 and HSP90 was not affected by p84 phosphorylation. In addition, p84 phosphorylation was prevented by the presence of divalent cations such as Mg(2+) and Mn(2+). In contrast, p84 phosphorylation was significantly activated by addition of exogenous Ca(2+) between 100 and 500 microM, although it was blocked by higher concentrations (>1 mM) of Ca(2+). HSP90, but not p84, could be phosphorylated by casein kinase II. Finally, p84 phosphorylation was specifically prevented by hemin, but not by other kinase inhibitors, indicating that p84 phosphorylation may be regulated by heme-regulated protein kinase.     

Effects of geldanamycin, a heat-shock protein 90-binding agent, on T cell function and T cell nonreceptor protein tyrosine kinases

The benzoquinoid ansamycins geldanamycin (GA), herbimycin, and their derivatives are emerging as novel therapeutic agents that act by inhibiting the 90-kDa heat-shock protein hsp90. We report that GA inhibits the proliferation of mitogen-activated T cells. GA is actively toxic to both resting and activated T cells; activated T cells appear to be especially vulnerable. The mechanism by which GA acts is reflected by its effects on an essential hsp90-dependent protein, the T cell-specific nonreceptor tyrosine kinase lck. GA treatment depletes lck levels in cultured T cells by a kinetically slow dose-dependent process. Pulse-chase analyses indicate that GA induces the very rapid degradation of newly synthesized lck molecules. GA also induces a slower degradation of mature lck populations. These results correlate with global losses in protein tyrosine kinase activity and an inability to respond to TCR stimuli, but the activity of mature lck is not immediately compromised. Although the specific proteasome inhibitor lactacystin provides marginal protection against GA-induced lck depletion, proteasome inhibition also induces changes in lck detergent solubility independent of GA application. There is no other evidence for the involvement of the proteosome. Lysosome inhibition provides quantitatively superior protection against degradation. These results indicate that pharmacologic inhibition of hsp90 chaperone function may represent a novel immunosuppressant strategy, and elaborate on the appropriate context in which to interpret losses of lck as a reporter for the pharmacology of GA in whole organisms.