Stabilization of integrin-linked kinase by binding to Hsp90

Integrin-linked kinase (ILK) is a serine/threonine kinase that interacts with the cytoplasmic domain of beta-integrins and growth factor receptors in response to extracellular signals. It is a key molecule in cell adhesion, proliferation, and cell survival. We found that treating cells with specific inhibitors of the heat shock protein 90 (Hsp90) caused rapid cell detachment. Screening the responsible proteins revealed a decreased amount of ILK in Hsp90 inhibitor-treated cells. ILK was identified as a new Hsp90 client protein because it formed a complex with Hsp90 and Cdc37, and binding was suppressed by Hsp90 inhibitors. Experiments with a series of ILK-deletion mutants revealed that the amino acid residues 377-406 were required for Hsp90 binding. Dissociation of ILK from Hsp90 shortened its half-life by promoting proteasome-dependent degradation. These results indicate that Hsp90 plays an important role in the stability of ILK in cells.     

Definition of protein kinase sequence motifs that trigger high affinity binding of Hsp90 and Cdc37

Hsp90 cooperates with its co-chaperone Cdc37 to provide obligatory support to numerous protein kinases involved in the regulation of cellular signal transduction pathways. In this report, the crystal structure of the Src family tyrosine kinase Lck was used to guide the creation of kinase constructs to determine features recognized by Hsp90 and its "kinase-specific" co-chaperone Cdc37. Two parameters were assayed: the ability and extent to which the constructs bound to Hsp90 and Cdc37, and the ability of the constructs to trigger salt-resistant high affinity complexes with Hsp90 and Cdc37 independent of the presence of molybdate. Although Hsp90 interacted with both the N-terminal and C-terminal lobes (NL and CL, respectively) of the catalytic domains of the kinases, the lobes themselves were not sufficient to trigger the high affinity binding of Hsp90. Only constructs containing a complete N- or C-terminal lobe and part of the adjacent lobe bound to Hsp90 and Cdc37 in salt-stable complexes independent of molybdate. The two minimum constructs that bound Hsp90 and Cdc37 contained the alpha-C-helix and the beta4- and beta5-strands of the NL through to end of the CL and the NL through to the alpha-E-helix and the amino acids that cap the helix. Cdc37 interacted with only the NL and minimally required the alpha-C-helix and beta4- and beta5-strands of this lobe of Lck. The results indicate that the high affinity binding activity of Hsp90 is triggered through its interaction with adjacent subdomain structures of kinase catalytic domains. Furthermore, the alpha-C-helix and part of its adjoining loop connection to the beta4-strand appear to be the primary determinants recognized by Cdc37.     

Tumor suppressor Tsc1 is a new Hsp90 co‐chaperone that facilitates folding of kinase and non‐kinase clients

The tumor suppressors Tsc1 and Tsc2 form the tuberous sclerosis complex (TSC), a regulator of mTOR activity. Tsc1 stabilizes Tsc2; however, the precise mechanism involved remains elusive. The molecular chaperone heat‐shock protein 90 (Hsp90) is an essential component of the cellular homeostatic machinery in eukaryotes. Here, we show that Tsc1 is a new co‐chaperone for Hsp90 that inhibits its ATPase activity. The C‐terminal domain of Tsc1 (998–1,164 aa) forms a homodimer and binds to both protomers of the Hsp90 middle domain. This ensures inhibition of both subunits of the Hsp90 dimer and prevents the activating co‐chaperone Aha1 from binding the middle domain of Hsp90. Conversely, phosphorylation of Aha1‐Y223 increases its affinity for Hsp90 and displaces Tsc1, thereby providing a mechanism for equilibrium between binding of these two co‐chaperones to Hsp90. Our findings establish an active role for Tsc1 as a facilitator of Hsp90‐mediated folding of kinase and non‐kinase clients—including Tsc2—thereby preventing their ubiquitination and proteasomal degradation.     

Intracellular dynamics of the Hsp90 co-chaperone p23 is dictated by Hsp90

p23 is a component of the Hsp90 molecular chaperone machine. It binds and stabilizes the ATP-bound dimeric form of Hsp90. Since Hsp90 binds protein substrates in the ATP conformation, p23 has been proposed to stabilize Hsp90-substrate complexes. In addition, p23 can also function as a molecular chaperone by itself and even possesses an unrelated enzymatic activity. Whether it fulfills the latter functions in cells while bound to Hsp90 remains unknown and is difficult to extrapolate from cell-free biochemical experiments. Using the "fluorescence recovery after photobleaching" (FRAP) technology, I have examined the dynamics of human p23, expressed as a fusion protein with the green fluorescent protein (GFP), in living human HeLa cells. GFP-p23 is distributed throughout the cell, and its mobility is identical in the cytoplasm and in the nucleus. When the Hsp90 interaction is disrupted either with the Hsp90 inhibitor geldanamycin or by introduction of point mutations into p23, the mobility of p23 is greatly accelerated. Under these conditions, its intracellular movement may be diffusion-controlled. In contrast, when wild-type p23 is able to bind Hsp90, a more complex FRAP behavior is observed, suggesting that it is quantitatively bound in Hsp90 complexes undergoing a multitude of other interactions.     

Sensitivity of mature Erbb2 to geldanamycin is conferred by its kinase domain and is mediated by the chaperone protein Hsp90

ErbB receptors are a family of ligand-activated tyrosine kinases that play a central role in proliferation, differentiation, and oncogenesis. ErbB2 is overexpressed in >25% of breast and ovarian cancers and is correlated with poor prognosis. Although ErbB2 and ErbB1 are highly homologous, they respond quite differently to geldanamycin (GA), an antibiotic that is a specific inhibitor of the chaperone protein Hsp90. Thus, although both mature and nascent ErbB2 proteins are down-regulated by GA, only nascent ErbB1 is sensitive to the drug. To reveal the underlying mechanism behind these divergent responses, we made a chimeric receptor (ErbB1/2) composed of the extracellular and transmembrane domains of ErbB1 and the intracellular domain of ErbB2. The ErbB1/2 protein is functional since its kinase activity was stimulated by epidermal growth factor. The sensitivity of ErbB1/2 to GA was similar to that of ErbB2 and unlike that of ErbB1, indicating that the intracellular domain of the chimera confers GA sensitivity. This finding also suggests that the GA sensitivity of mature ErbB2 depends on cytosolic Hsp90, rather than Grp94, a homolog of Hsp90 that is restricted to the lumen of the endoplasmic reticulum, although both chaperones bind to and are inhibited by GA. Lack of Grp94 involvement in mediating ErbB2 sensitivity to GA is further suggested by the fact that a GA derivative with low affinity for Grp94 efficiently depleted ErbB2 protein in treated cells. To localize the specific region of ErbB2 that confers GA sensitivity, we made truncated receptors with progressive deletions of the cytoplasmic domain and tested the GA sensitivity of these molecules. We found that ErbB2 constructs containing an intact kinase domain retained GA sensitivity, whereas those lacking the kinase domain (ErbB2/DK) lost responsiveness to GA completely. Hsp90 co-immunoprecipitated with all ErbB2 constructs that were sensitive to GA, but not with ErbB2/DK or ErbB1. Both tyrosine-phosphorylated and non-phosphorylated ErbB2 proteins were similarly sensitive to GA, as was a kinase-dead ErbB2 mutant. These data suggest that Hsp90 uniquely stabilizes ErbB2 via interaction with its kinase domain and that GA stimulates ErbB2 degradation secondary to disruption of ErbB2/Hsp90 association.     

Control of estrogen receptor ligand binding by Hsp90

The molecular chaperone Hsp90 interacts with unliganded steroid hormone receptors and regulates their activity. We have analyzed the function of yeast and mammalian Hsp90 in regulating the ability of the human estrogen receptor (ER) to bind ligands in vivo and in vitro. Using the yeast system, we show that the ER expressed in several different hsp82 mutant strains binds reduced amounts of the synthetic estrogen diethylstilbestrol compared to the wild type. This defect in hormone binding occurs without any significant change in the steady state levels of ER protein. To analyze the role of mammalian Hsp90, we synthesized the human ER in rabbit reticulocyte lysates containing geldanamycin, an Hsp90 inhibitor. At low concentrations of geldanamycin we observed reduced levels of hormone binding by the ER. At higher concentrations, we found reduced synthesis of the receptor. These data indicate that Hsp90 functions to maintain the ER in a high affinity hormone-binding conformation.     

Exposure of protein kinase motifs that trigger binding of Hsp90 and Cdc37

Hsp90 and its co-chaperone Cdc37 are required for the activity of numerous eukaryotic protein kinases. c-Jun N-terminal kinases (JNKs) appear to be Hsp90-independent kinases, as their activity is unaffected by Hsp90 inhibition. It is currently unknown why some protein kinases are Hsp90- and Cdc37-dependent for their function, while others are not. Therefore, we investigated what structural motifs within JNKs confer or defer Hsp90 and Cdc37 interaction. Both Hsp90 and Cdc37 recognized structural features that were exposed or destabilized upon deletion of JNK1alpha1's N-terminal non-catalytic structural motif, while only Hsp90 bound JNK when its C-terminal non-catalytic structural motif was deleted. Mutations in JNK's activation loop that are known to constitutively activate or inactivate its kinase activity had no effect on JNK's lack of interaction with Hsp90 and Cdc37. Our findings suggest a model in which Hsp90 and Cdc37 each recognize distinct features within the catalytic domains of kinases.     

Nuclear translocation of the phosphoprotein Hop (Hsp70/Hsp90 organizing protein) occurs under heat shock, and its proposed nuclear localization signal is involved in Hsp90 binding

The Hsp70-Hsp90 complex is implicated in the folding and regulation of numerous signaling proteins, and Hop, the Hsp70-Hsp90 Organizing Protein, facilitates the association of this multichaperone machinery. Phosphatase treatment of mouse cell extracts reduced the number of Hop isoforms compared to untreated extracts, providing the first direct evidence that Hop was phosphorylated in vivo. Furthermore, surface plasmon resonance (SPR) spectroscopy showed that a cdc2 kinase phosphorylation mimic of Hop had reduced affinity for Hsp90 binding. Hop was predominantly cytoplasmic, but translocated to the nucleus in response to heat shock. A putative bipartite nuclear localization signal (NLS) has been identified within the Hsp90-binding domain of Hop. Although substitution of residues within the major arm of this proposed NLS abolished Hop-Hsp90 interaction as determined by SPR, this was not sufficient to prevent the nuclear accumulation of Hop under leptomycin-B treatment and heat shock conditions. These results showed for the first time that the subcellular localization of Hop was stress regulated and that the major arm of the putative NLS was not directly important for nuclear translocation but was critical for Hop-Hsp90 association in vitro. We propose a model in which the association of Hop with Hsp90 and the phosphorylated status of Hop both play a role in the mechanism of nucleo-cytoplasmic shuttling of Hop.     

Nox5 stability and superoxide production is regulated by C-terminal binding of Hsp90 and CO-chaperones

Heat shock protein 90 (Hsp90) is a molecular chaperone that orchestrates the folding and stability of proteins that regulate cellular signaling, proliferation and inflammation. We have previously shown that Hsp90 controls the production of reactive oxygen species by modulating the activity of Noxes1–3 and 5, but not Nox4. The goal of the current study was to define the regions on Nox5 that bind Hsp90 and determine how Hsp90 regulates enzyme activity. In isolated enzyme activity assays, we found that Hsp90 inhibitors selectively decrease superoxide, but not hydrogen peroxide, production. The addition of Hsp90 alone only modestly increases Nox5 enzyme activity but in combination with the co-chaperones, Hsp70, HOP, Hsp40, and p23 it robustly stimulated superoxide, but not hydrogen peroxide, production. Proximity ligation assays reveal that Nox5 and Hsp90 interact in intact cells. In cell lysates using a co-IP approach, Hsp90 binds to Nox5 but not Nox4, and the degree of binding can be influenced by calcium-dependent stimuli. Inhibition of Hsp90 induced the degradation of full length, catalytically inactive and a C-terminal fragment (aa398–719) of Nox5. In contrast, inhibition of Hsp90 did not affect the expression levels of N-terminal fragments (aa1–550) suggesting that Hsp90 binding maintains the stability of C-terminal regions. In Co-IP assays, Hsp90 was bound only to the C-terminal region of Nox5. Further refinement using deletion analysis revealed that the region between aa490-550 mediates Hsp90 binding. Converse mapping experiments show that the C-terminal region of Nox5 bound to the M domain of Hsp90 (aa310–529). In addition to Hsp90, Nox5 bound other components of the foldosome including co-chaperones Hsp70, HOP, p23 and Hsp40. Silencing of HOP, Hsp40 and p23 reduced Nox5-dependent superoxide. In contrast, increased expression of Hsp70 decreased Nox5 activity whereas a mutant of Hsp70 failed to do so. Inhibition of Hsp90 results in the loss of higher molecular weight complexes of Nox5 and decreased interaction between monomers. Collectively these results show that the C-terminal region of Nox5 binds to the M domain of Hsp90 and that the binding of Hsp90 and select co-chaperones facilitate oligomerization and the efficient production of superoxide.     

Thermodynamics of aryl-dihydroxyphenyl-thiadiazole binding to human Hsp90

The design of specific inhibitors against the Hsp90 chaperone and other enzyme relies on the detailed and correct understanding of both the thermodynamics of inhibitor binding and the structural features of the protein-inhibitor complex. Here we present a detailed thermodynamic study of binding of aryl-dihydroxyphenyl-thiadiazole inhibitor series to recombinant human Hsp90 alpha isozyme. The inhibitors are highly potent, with the intrinsic K(d) approximately equal to 1 nM as determined by isothermal titration calorimetry (ITC) and thermal shift assay (TSA). Dissection of protonation contributions yielded the intrinsic thermodynamic parameters of binding, such as enthalpy, entropy, Gibbs free energy, and the heat capacity. The differences in binding thermodynamic parameters between the series of inhibitors revealed contributions of the functional groups, thus providing insight into molecular reasons for improved or diminished binding efficiency. The inhibitor binding to Hsp90 alpha primarily depended on a large favorable enthalpic contribution combined with the smaller favorable entropic contribution, thus suggesting that their binding was both enthalpically and entropically optimized. The enthalpy-entropy compensation phenomenon was highly evident when comparing the inhibitor binding enthalpies and entropies. This study illustrates how detailed thermodynamic analysis helps to understand energetic reasons for the binding efficiency and develop more potent inhibitors that could be applied for therapeutic use as Hsp90 inhibitors.     

Requirement of Hsp90 for centrosomal function reflects its regulation of Polo kinase stability

We have previously shown that the molecular chaperone heat shock protein 90 (Hsp90) is required to ensure proper centrosome function in Drosophila and vertebrate cells. This observation led to the hypothesis that this chaperone could be required for the stability of one or more centrosomal proteins. We have found that one of these is Polo, a protein kinase known to regulate several aspects of cell division including centrosome maturation and function. Inhibition of Hsp90 results in the inactivation of Polo kinase activity. It also leads to a loss in the ability of cytoplasmic extracts to complement the failure of salt-stripped preparations of centrosomes to nucleate microtubules. This effect can be rescued upon addition of active recombinant POLO: We also show that Polo and Hsp90 are part of a complex and conclude that stabilization of Polo is one of the mechanisms by which Hsp90 contributes to the maintenance of functional centrosomes.     

Regulation of the Src family kinase Lck by Hsp90 and ubiquitination

Regulation of the Src-related tyrosine kinase Lck is crucial to the outcome of T-cell receptor (TCR) stimulation. It was previously shown that the stability of the constitutively active mutant LckY505F is controlled by Hsp90 (M. J. Bijlmakers and M. Marsh, Mol. Biol. Cell. 11:1585-1595, 2000). Here we establish that following TCR stimulation, endogenous activated Lck in T cells is also degraded in the presence of the Hsp90 inhibitor geldanamycin. Using Lck constructs expressed in COS-7 cells, we show that the presence of activating Lck mutations results not only in the enhanced dependence on Hsp90 but also in enhanced ubiquitination of Lck. Although both processes were induced by mutations Y505F and W97A that release the SH2 and SH3 inhibitory intramolecular interactions, respectively, neither process required Lck kinase activity or activation-dependent phosphorylation at serines 42 and 59 or tyrosine 394. By binding to the ATP-binding site, the Src family inhibitor PP2 reduced ubiquitination and overcame the need for Hsp90 monitoring of active Lck. We conclude that the levels of active Lck are influenced by two opposing processes, targeting for degradation by ubiquitination and rescue from degradation by Hsp90 monitoring. Based on the PP2 result, we propose that activation-induced conformational changes of the Lck kinase domain instigate both regulatory processes.     

p50(cdc37) is a nonexclusive Hsp90 cohort which participates intimately in Hsp90-mediated folding of immature kinase molecules

Hsp90 and p50(cdc37) provide a poorly understood biochemical function essential to certain protein kinases, and recent models describe p50(cdc37) as an exclusive hsp90 cohort which links hsp90 machinery to client kinases. We describe here the recovery of p50(cdc37) in immunoadsorptions directed against the hsp90 cohorts FKBP52, cyp40, p60HOP, hsp70, and p23. Additionally, monoclonal antibodies against FKBP52 coadsorb maturation intermediates of the hsp90-dependent kinases p56(lck) and HRI, and the presence of these maturation intermediates significantly increases the representation of p50(cdc37) and hsp90 on FKPB52 machinery. Although the native heterocomplex between hsp90 and p50(cdc37) is salt-labile, their dynamic interactions with kinase substrates produce kinase-chaperone heterocomplexes which are highly salt-resistant. The hsp90 inhibitor geldanamycin does not directly disrupt the native association of hsp90 with p50(cdc37) per se, but does result in the formation of salt-labile hsp90-kinase heterocomplexes which lack the p50(cdc37) cohort. We conclude that p50(cdc37) does not simply serve as a passive structural bridge between hsp90 and its kinase substrates; instead, p50(cdc37) is a nonexclusive hsp90 cohort which responds to hsp90's nucleotide-regulated conformational switching during the generation of high-affinity interactions within the hsp90-kinase-p50(cdc37) heterocomplex.     

Rab-alphaGDI activity is regulated by a Hsp90 chaperone complex

The Rab-specific alphaGDP-dissociation inhibitor (alphaGDI) regulates the recycling of Rab GTPases. We have now identified a novel alphaGDI complex from synaptic membranes that contains three chaperone components: Hsp90, Hsc70 and cysteine string protein (CSP). We find that the alphaGDI-chaperone complex is dissociated in response to Ca(2+)-induced neurotransmitter release, that chaperone complex dissociation is sensitive to the Hsp90 inhibitor geldanamycin (GA) and that GA inhibits the ability of alphaGDI to recycle Rab3A during neurotransmitter release. We propose that alphaGDI interacts with a specialized membrane-associated Rab recycling Hsp90 chaperone system on the vesicle membrane to coordinate the Ca(2+)-dependent events triggering Rab-GTP hydrolysis with retrieval of Rab-GDP to the cytosol.     

High affinity estrogen binding by rabbit liver

Macromolecular binding components for [³H]estradiol-17β are present to cytosol prepared from rabbit liver. When cytosol from sexually mature male liver was incubated with [³H]estradiol and analyzed for binding on low ionic strength sucrose gradients, two peaks of binding activity were detected. One peak had a sedimentation coefficient of 4–5 S and the other had a sedimentation coefficient of 8–9 S. The two components differed from each other regarding steroid specicity and various physiocochemical parameters. [³H]-estradiol binding to the 4–5 S component was not inhibited by estrogens, 5α-dihydrotestosterone, progesterone or cortisol. Binding to this component did not appera to be saturable and lavel was rapidly stripped from it by cahrcoal. Estradiol bindng to the 8–9 S component was estrogen specific, saturable and of high affinity. The specific binder dissociates on high ionic strength sucrose gradients and sediments as a 4–5 S moiety. The specific binding protein has a K_d of 3.05 · 10⁻¹⁰ M and a dissociation half-time of 33 h and there are 35.2 fmol of binding sites/mg cytosol protein. Estrogen binders are also present in liver cytosol from sexually mature female and sexually immature male rabbits. During prolonged incbuation of [³H]estradiol with mature male liver cytosol at 0–5°C polar metabolites of estradiol are produced.     

High affinity estrogen binding by rabbit liver

Macromolecular binding components for [3H]estradiol-17beta are present to cytosol prepared from rabbit liver. When cytosol from sexually mature male liver was incubated with [3H]estradiol and analyzed for binding on low ionic strength sucrose gradients, two peaks of binding activity were detected. One peak had a sedimentation coefficient of 4--5 S and the other had a sedimentation coefficient of 8--9 S. The two components differed from each other regarding steroid specificity and various physiocochemical parameters. [3H]estradiol binding to the 4--5 S component was not inhibited by estrogens, 5alpha-dihydrotestosterone, progesterone or cortisol. Binding to this component did not appear to be saturable and label was rapidly stripped from it by charcoal. Estradiol binding to the 8--9 S component was estrogen specific, saturable and of high affinity. The specific binder dissociates on high ionic strength sucrose gradients and sediments as a 4--5 S moiety. The specific binding protein has a Kd of 3.05 . 10(-10) M and a dissociation half-time of 33 h and there are 35.2 fmol of binding sites/mg cytosol protein. Estrogen binders are also present in liver cytosol from sexually mature female and sexually immature male rabbits. During prolonged incubation of [3H]estradiol with mature male liver cytosol at 0--5 degrees C polar metabolites of estradiol are produced.     

The C-terminal half of Hsp90 is responsible for its cytoplasmic localization.

With some exceptions, research so far has shown heat shock protein (Hsp) 90 to be a cytoplasmic protein. Here, we studied the sequence determinants which dictate the subcellular localization of Hsp90. By constructing hybrid molecules between a nuclear protein, progesterone receptor (PR), and parts of Hsp90, we demonstrated that the C-terminal but not the N-terminal half of Hsp90 can prevent nuclear translocation of the PR. Studies with an antibody raised against a region which contains the major nuclear localization signal (NLS) of the PR suggest that the inhibition of nuclear localization is not due to steric hindrance of the NLS of the PR by Hsp90 sequences in hybrid molecules. In order to characterize further the cytoplasmic anchoring of Hsp90 we constructed four chimeric molecules between the C-terminal half of Hsp90 and estrogen receptor (ER) with different numbers of nuclear localization protosignals (proto-NLS). When the C-terminal half of Hsp90 was fused with ER containing no or one proto-NLS, the hybrid molecule was located exclusively in the cytoplasm. When the nuclear translocation signal was strengthened by adding two or three protosignals, the hybrid molecule was exclusively nuclear. These results suggest that the C-terminal half of Hsp90 contains a sequence which is responsible for the cytoplasmic localization of the protein. Further deletions of the molecule suggested that the cytoplasmic anchoring signal is located between amino acids 333 and 664.     

High affinity binding of the mastoparans by calmodulin

Calmodulin exhibits high affinity, calcium-dependent binding of the mastoparans — a group of cytoactive tetradecapeptides. The dissociation constants for the peptide-calmodulin complexes determined in 0.20 N KCl, 1.0 mM CaCl₂, pH 7.3 are ～0.3 nM for mastoparan, ～0.9 nM for mastoparan X, and ～3.5 nM for $\text{Polistes}$ mastoparan. The dissociation constant for the mastoparan-calmodulin complex is the smallest known for any calmodulin binding protein or peptide, suggesting that some type of peptide-calmodulin interaction could be physiologically significant.