HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 282, Issue 48, 35386-35395, November 30, 2007

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 282/48/35386    most recent M703844200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Wayne, N. Articles by Bolon, D. N. Search for Related Content PubMed PubMed Citation Articles by Wayne, N. Articles by Bolon, D. N. Social Bookmarking                      What's this?

Dimerization of Hsp90 Is Required for in Vivo Function

DESIGN AND ANALYSIS OF MONOMERS AND DIMERS^*

From the Department of Biochemistry and Molecular Pharmacology, University of Massachusetts Medical School, Worcester, Massachusetts 01605

Received for publication, May 9, 2007 , and in revised form, September 25, 2007.

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Heat shock protein 90 (Hsp90) plays a central role in signal transduction and has emerged as a promising target for anti-cancer therapeutics, but its molecular mechanism is poorly understood. At physiological concentration, Hsp90 predominantly forms dimers, but the function of full-length monomers in cells is not clear. Hsp90 contains three domains: the N-terminal and middle domains contribute directly to ATP binding and hydrolysis and the C domain mediates dimerization. To study the function of Hsp90 monomers, we used a single-chain strategy that duplicated the C-terminal dimerization domain. This novel monomerization strategy had the dual effect of stabilizing the C domain to denaturation and hindering intermolecular association of the ATPase domain. The resulting construct was predominantly monomeric at physiological concentration and did not function to support yeast viability as the sole Hsp90. The monomeric construct was also defective at ATP hydrolysis and the activation of a kinase and steroid receptor substrate in yeast cells. The ability to support yeast growth was rescued by the addition of a coiled-coil dimerization domain, indicating that the parental single-chain construct is functionally defective because it is monomeric.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Among heat shock proteins, Hsp90 is unusual because it is not required for the proper folding of most cellular proteins (1) and instead is disproportionately linked to a select group of proteins required for receiving, transducing, and responding to environmental signals. The list of Hsp90 substrates continues to grow and includes over 40 kinases (2) and many steroid hormone receptors (3). Biochemical studies demonstrate that Hsp90 along with a handful of co-chaperones is required for many hormone receptors (including glucocorticoid, androgen, estrogen, and progesterone) to bind steroid ligand (3, 4). V-src, the transforming tyrosine kinase from Rous sarcoma virus, was the first kinase to be identified as an Hsp90 substrate and is one of the most extensively studied. The transforming phenotype of v-src can be reversed with polyketide benzoquinone ansamycins including geldanamycin (GA),² which directly bind to Hsp90 (5) and blocks the association of v-src and Hsp90 (6). Stimulated by these observations, Hsp90 has emerged as a promising molecular target for anti-cancer therapeutics (7). Recently, Hsp90 activation of the Chk1 kinase has been reconstituted in vitro and has been shown to require a set of co-chaperones that partially differ from those required for hormone receptors (8). As with hormone receptors, the molecular mechanism of Hsp90 activation of v-src and other kinases remains to be determined.

How does Hsp90 dimerization contribute to its function in cells? Structurally, Hsp90 contains an N-terminal domain that binds nucleotide, a middle domain, and a C-terminal dimerization domain (Fig. 1). Truncations of Hsp90 that lack the C domain are largely monomeric (9) and do not support yeast viability (10). These results do not demonstrate that dimerization of Hsp90 is essential because the C domain may contribute other functional properties in addition to dimerization. The deletion of a loop from the C domain (582-601) that is largely solvent exposed (11) prevents yeast viability (10), suggesting that dimerization is not the only essential role of the C domain. In addition, a point mutation (A587T) in this same loop impairs the chaperoning of glucocorticoid receptor (GR) in yeast (12). Because Hsp90 activity requires co-chaperones and Hsp90 binds promiscuously to about 10% of the yeast proteome (2), it is important to study Hsp90 function in cells where all endogenous binding partners are present. To rigorously test how dimerization affects Hsp90 function in cells requires a full-length Hsp90 monomer.

The ATPase activity of Hsp90 is required for cellular function, although the molecular mechanism that couples ATP hydrolysis to substrate activation is poorly understood. Consistent with its role in signal transduction, Hsp90 is required for viability in all eukaryotes that have been tested (yeast, flies, and worms) (13-15). Structural and biochemical studies demonstrate that Hsp90 is a "split" ATPase with catalytic amino acids from both the N (Glu³³) and M (Arg³⁸⁰) domains contributing to hydrolysis (11, 16, 17). Mutation of either of these catalytic amino acids impairs ATPase activity and does not support yeast viability as the sole Hsp90. Hsp90 point mutations that perturb the ATPase activity from 2 to 400% of the wild-type level can support yeast growth under non-stressful conditions (18). Thus, cell viability can tolerate a wide range of Hsp90 ATPase levels.

Dimerization of Hsp90 has been implicated in ATP hydrolysis. NM constructs of yeast Hsp90 lacking the C-terminal dimerization domain have concentration-dependent specific ATPase activity (9). These results indicate that the N and/or M domains can mediate weak dimerization that leads to increased ATPase activity. Consistent with this analysis, the disulfide cross-linking of truncated Hsp90 constructs increases their ATPase activity (19). These studies raise the possibility that full-length Hsp90 monomers may be inactive for ATP hydrolysis. Studies with heterodimers made up of one full-length Hsp90 subunit and one C-domain, indicate that full-length/C-domain Hsp90 heteromers have about 30% the ATPase activity of wild-type (9), well within the 2 to 400% range previously observed to support yeast viability (18). These observations have stimulated the current work to determine whether Hsp90 monomers are functional in cells.

We have used a single-chain strategy to design a full-length NMCC Hsp90 monomer (Fig. 1) and analyzed its function in yeast cells. To our knowledge, this is a novel approach to making full-length monomers and relative to traditional strategies that mutagenize the dimer interface, has the advantage of stabilizing the dimerization domain to denaturation. We find that NMCC monomers do not function to support yeast growth, the activation of a kinase nor a hormone receptor substrate in yeast. Appending a coiled-coil dimerization motif to NMCC rescues dimerization and function, indicating that the parental NMCC construct is defective only because dimerization is impaired. Whereas Hsp90 monomers do not support yeast viability, wild-type Hsp90 monomers form transiently, on the same time scale as ATP hydrolysis, suggesting that they may play a role in the Hsp90 chaperone cycle. NMCC Hsp90 will provide a tool to further explore the biochemical properties of Hsp90 monomers and their role in the chaperone cycle.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Construction of Hsp90 Variants—All Hsp90 constructs used in these studies were generated from the yeast HSP82 gene. To aid in protein purification and Western blot detection, a His₆ tag (GGHHHHHHGGH) was appended to the N terminus of Hsp90 constructs. Constructs were cloned into pACYC for bacterial expression and into p414GPD for expression in yeast. The C-terminal domain of Hsp90 contains amino acids 542-709. Single-chain constructs of the C-terminal domain were constructed with amino acids 537-680 from HSP82 followed by the Gly-rich linkers outlined in Fig. 2 then by amino acids 542-709 from HSP82. The NMCC construct contains amino acids 1-680 from HSP82 followed by the 15-amino acid linker in Fig. 2, then amino acids 542-709 from HSP82. The NMCCcoil construct contains a coiled-coil sequence (GGGTSSVKELEDKNEELLSEIAHLKNEVARLKKLVGERTG) inserted after amino acid 678 of the second C domain of NMCC.

Protein Production—All Hsp90 constructs were expressed in BLR(DE3) cells induced with 1 mM isopropyl 1-thio-β-D-galactopyranoside at 30 °C for 5 h. Cells were harvested by centrifugation and resuspended in Wash Buffer (50 mM potassium phosphate, pH 8, 300 mM potassium chloride, 20 mM imidazole). Cell lysis was accomplished by treatment with lysozyme and sonication. After pelleting cell debris, lysates were incubated with Ni²⁺-nitrilotriacetic acid-agarose (Qiagen). After rinsing the nickel resin extensively with Wash Buffer, Hsp90 protein was competitively eluted with Elution Buffer (200 mM imidazole, pH 7.5). EDTA was added to 10 mM to eluates that were subsequently dialyzed into Buffer A (20 mM potassium phosphate, pH 6.8, 1 mM EDTA). Protein samples were further purified by anion exchange chromatography using a Q Sepharose HP column (GE Healthcare) eluted with a linear gradient from 0 to 500 mM potassium chloride in Buffer A. As a final purification, protein samples were subjected to size exclusion chromatography using a Sephacryl S300 column (GE Healthcare) in Buffer B (20 mM potassium phosphate, pH 6.8, 1 mM EDTA, 100 mM potassium chloride). Proteins were concentrated in Buffer B to 10 mg/ml using Amicon Ultra concentrators (Millipore). Protein concentrations were determined spectroscopically using extinction coefficients (M^-1 cm^-1) based on amino acid composition using the program Sednterp (Amgen) at 280 nm: WT (54,050), NMCC and NMCCcoil (64,860), C domain (10,810), and C-linker-C constructs (21,620).

Circular Dichroism—CD measurements were acquired on a Jasco A-810 spectropolarimeter equipped with a Peltier temperature control unit and an autotitrator. C domain spectra were acquired in a 1-mm path length cuvette at a subunit concentration of 20 µM in 20 mM potassium phosphate, pH 7, at 25 °C. All equilibrium urea titrations were performed in Buffer B at 25 °C. Protein denaturation was followed by monitoring the loss of ellipticity at 222 nm. Measurements at a subunit concentration of 2 µM were made in a cuvette with a 1-cm path length using an autotitrator with a mixing time of 600 s (determined to be more than three times greater than the time constant for unfolding at the C_m). For urea titrations at a subunit concentration of 20 µM, samples were manually mixed, allowed to equilibrate for 30 min, and measurements made in a 1-mm cuvette. Both the pre- and post-transition regions were fit to linear equations and used to replot the data as the fraction unfolded. The fraction unfolded plots were fit to two-state models with or without dimerization as appropriate and previously described (20, 21) using the program Kaleidograph (Synergy Software).

Analytical Ultracentrifugation—Equilibrium experiments were performed at 20 °C on a Beckman XLI instrument using absorbance optics and a Ti60 rotor. Absorbance profiles were taken at 12-h intervals, and overlapping profiles were used as a criterion for equilibration. For the C domain constructs, samples were analyzed at a subunit concentration of 50 µM in Buffer A with a rotor speed of 15,000 rpm and an equilibration time of 36 h. For full-length constructs, samples were analyzed at a subunit concentration of 12 µM in Buffer B with a rotor speed of 8,000 rpm and an equilibration time of 36 h. Absorbance profiles were fit to a single species model as previously described (22) using the equation,

(Eq.1)

where c₂(x) is the concentration at a radial distance of x, x₀ is a reference point, M is the molecular weight, V is the partial specific volume, is the buffer density, is the angular velocity, R is the universal gas constant, and T is the temperature. Data were fit in Kaleidograph using buffer density and V-bar values based on the buffer and amino acid composition and determined using the Sednterp program.

Analytical Size Exclusion Chromatography—100-µl samples ranging in subunit concentration from 5 to 50 µM were analyzed using Buffer B and a Superdex 200 column (GE Healthcare). Absorbance at 280 nm was used to monitor the elution profile.

Enzymatically Coupled ATPase Assay—ATP hydrolysis was enzymatically linked to NADH oxidation that was monitored spectroscopically (23). Hsp90 catalyzed ATP hydrolysis to generate ADP. Pyruvate kinase was used to convert ADP and phosphenolpyruvate to ATP and pyruvate. Lactose dehydrogenase was used to convert pyruvate and NADH to lactate and NAD with a corresponding drop in absorbance at 340 nm (₃₄₀ = 6220 M^-1 cm^-1). ATPase assays were performed at 37 °C. Both Hsp90 protein samples and ATPase components were pre-warmed, then mixed. Using a Bio50 Spectrophotometer equipped with a Peltier temperature control unit (Cary) and a 1-cm path length cuvette, absorbance at 340 nm was measured at 15-s intervals (avoids photobleaching of NADH) for 10 min. The final concentration of ATPase components was 20 mM Tris, pH 7.5, 5 mM MgCl₂, 100 mM KCl, 1 mM ATP (Sigma), 0.17 mM NADH (Sigma), 0.67 mM phosphoenolpyruvate (Sigma), 0.01 mg/ml pyruvate kinase (Sigma), and 0.02 mg/ml lactose dehydrogenase (EMD Biosciences). The rate of NADH oxidation with 0.3 mM ADP was more than an order of magnitude greater than the highest rate observed with Hsp90 demonstrating that ATPase components are not rate-limiting. Rates were determined by fitting the change in absorbance versus time to a linear model, converting absorbance units to the amount of NADH oxidized and normalizing to the concentration of Hsp90.

Fluorescent GA Binding—BODIPY-GA was synthesized as described (24). BODIPY-GA concentration was determined by absorbance in methanol using an extinction coefficient of 80,000 M^-1 cm^-1 at 506 nm (25) and was in close agreement with the weighed mass of BODIPY-GA used. 3 µM BODIPY-GA was preincubated in assay buffer (20 mM Tris, pH 7.5, 5 mM MgCl₂, 100 mM KCl, 0.01% Nonidet P-40) with 10 mM dithiothreitol for 4 h at room temperature to convert GA to the high-affinity hydroquinone form (26). Samples were prepared in assay buffer with 5 mM dithiothreitol containing 1 µM reduced BODIPY-GA and wild-type or NMCC Hsp90 ranging from 0.1 to 3.3 µM. These concentrations of Hsp90 and BODIPY-GA are both well above the 5 nM K_d determined for human Hsp90 and BODIPY-GA (26). Samples were equilibrated for 10 min at room temperature and fluorescence anisotropy measurements were made in a 0.3-cm path length cuvette on a PTI QM-4SE spectrofluorometer with excitation set to 488 nm and emission at 510 nm.

Sole Hsp90 in Yeast—The haploid Saccharomyces cerevisiae strains iG170Da (1) and ECU82a (12) are derivatives of W303 in which both endogenous Hsp90 genes, HSP82 and HSC82, are knocked out. In strain iG170Da, the temperature-sensitive G170D mutant of HSP82 is chromosomally integrated and expressed from a GPD promoter. In strain ECU82a, wild-type HSC82 is constitutively expressed from pKAT6, a URA3 marked 2-µm (hi copy) plasmid that is amenable to negative selection. To test the ability of our Hsp90 constructs to support yeast viability, we generated them in p414GPD (27), a TRP marked CEN plasmid with a strong constitutive promoter. Lithium acetate was used to introduce our Hsp90 constructs into iG170Da and ECU82a and plating in the absence of tryptophan was used to select for transformants. Transformants were grown in liquid media lacking tryptophan to an A₆₀₀ of 0.7, serially diluted, and plated under permissive or non-permissive conditions. All samples were grown and plated in synthetic media lacking tryptophan and with 2% dextrose as a sugar source. For strain iG170Da, 25 °C was permissive and 37 °C was non-permissive. Strain ECU82a was grown at 25 °C, in the absence or presence of 5-fluoroorotic acid, which selects for loss of the pKAT6 plasmid. To determine the expression level of our Hsp90 constructs in ECU82a, liquid cultures in minimal media were grown to a cell density of 5 x 10⁷ cells/ml determined with a hemacytometer. Cells were lysed by vortexing with glass beads and resuspension in SDS. Proteins from the lysis of 10⁷ cells were separated by SDS-PAGE and the expression level quantified by Western blotting against the His₆ epitope tag. A standard curve was generated using purified Hsp90 added to lysates from cells that did not express any epitope-tagged Hsp90.

V-src Assay—Plasmid Y316v-srcv5 was generated from Y316v-src (12) and contains v-src with a C-terminal v5 epitope tag (GKPIPNPLLGLDST) under a galactose-regulated promoter on a URA3-marked CEN plasmid. Lithium acetate was used to introduce Y316v-srcv5 into iG170Da cells and plating in the absence of uracil was used to select for transformants. To these cells we introduced our Hsp90 constructs on p414GPD plasmids. On plates, cells were grown with dextrose as the sugar source to prevent v-src expression. Liquid cultures were grown in synthetic raffinose media without uracil and tryptophan at 25 °C to a cell density of about 5 x 10⁶ cells/ml (determined with a hemocytometer). Cells were pelleted and resuspended in media with either 2% raffinose or galactose as the sugar source pre-warmed to 38 °C (to inactivate G170D Hsp90). Cells were grown in a shaking incubator at 35 °C for 6 h. Cell pellets were collected by centrifugation, washed once with water, and frozen at -80 °C. The frozen cell pellets were lysed by vortexing with glass beads in Src Lysis Buffer (50 mM Tris, pH 7.5, 5 mM EDTA, 0.2 mM sodium orthovanadate to inhibit dephosphorylation, 1 mM phenylmethylsulfonyl fluoride) followed by addition of SDS to 2%. Protein concentration in these SDS lysates was assessed using the BCA assay (Pierce). Samples (2 µg of protein) were subject to SDS-PAGE and phosphotyrosine levels quantified by Western blot analysis with antibody 4G10 (Upstate) in the presence of 0.1% Tween 20. The level of v-src was quantified by Western blot analysis (20 µg of protein/lane) with -v5 antibody (Invitrogen).

GR Assay—P2A/GRGZ (12) is a 2-µm ADE2 plasmid that contains rat glucocorticoid receptor expressed from the constitutive GPD promoter and the β-galactosidase reporter under the control of three glucocorticoid response elements. We introduced P2A/GRGZ together with our p414GPD Hsp90 constructs into iG170Da cells. Cells were grown in synthetic dextrose media lacking tryptophan and adenine at 25 °C to a cell density of about 5 x 10⁶ cells/ml. Cells were pelleted by centrifugation and resuspended in media prewarmed to 38 °C (to inactivate G170D Hsp90). 15-ml cultures were grown in a shaking incubator at 35 °C for 5 min. Deoxycorticosterone dissolved in ethanol was added to final concentrations of 0, 0.08, 0.4, 2, 10, or 50 µM (final concentration of ethanol was 0.1% in all cases). Cells were grown for a further 60 min at 35 °C and collected by centrifugation. After washing once with water, cell pellets were frozen at -80 °C. Cells were lysed by vortexing with glass beads in βGal Lysis Buffer (100 mM potassium phosphate, pH 7.3, 2 mM magnesium acetate, 1 mM phenylmethylsulfonyl fluoride). After removing cell debris, protein concentration in the lysates was determined using the BCA assay (Pierce). A total of 10 µg of lysate was reacted for 15 min in a total volume of 80 µl of 1 mg/ml o-nitrophenyl β-D-galactoside (Sigma) in βGal Lysis Buffer. The reaction was stopped by adding 80 µl of 1 M sodium carbonate. Reporter activity was quantified by monitoring the absorbance at 420 nm. GR assays were repeated three times starting from fresh yeast colonies.

View larger version (20K): [in this window] [in a new window]   FIGURE 1. Design strategy. A, architecture of the Hsp90 gene and the dimeric protein structure. Hsp90 contains three domains referred to as N (N-terminal), M (middle), and C (C-terminal). In the structural representation, these domains are colored blue for N, green for M, and red for C with lighter colors distinguishing the dimer subunit pictured on the left. B, the stability to urea denaturation of the C domain is concentration dependent. At the low protein concentrations required to populate monomers, the stability of the C domain would be marginal. C, NMCC strategy to design a full-length Hsp90 monomer with increased C domain stability.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION CONCLUSIONS REFERENCES

Isolated Single-chain C Domain Is Stable and Monomeric—With the aim of generating a single-chain C domain that is both monomeric and stable, we generated CC constructs separated by variable length glycine-rich linkers (Fig. 2A). Based on the crystal structure of Hsp90, the flexible linkers need to span a distance of 27 Å. Whereas 8 amino acids in a fully extended conformation can span this distance, previous single-chain studies have found that a greater number of amino acids is required to avoid strain and maximize stability (29). We made glycine-rich linkers of 11, 15, and 19 amino acids based on a proteolytically stable loop in bacteriophage gene-3 protein that had been used in a previous single-chain study (30). All three single-chain constructs (C11C, C15C, and C19C) expressed to high levels in Escherichia coli, were readily purified, highly soluble, and had CD spectra similar to the wild-type C domain (Fig. S1). At a subunit concentration of 2 µM, we observe that C11C, C15C, and C19C all have increased stability to urea denaturation relative to wild type (Fig. 2B). These results indicate that the effective concentration caused by the covalent single-chain linkages is greater than the concentration of wild-type protein (2 µM).

View larger version (16K): [in this window] [in a new window]   FIGURE 2. Single-chain C-domains are increased in stability and monomeric. A, design of single-chain C-domain constructs with glycine-rich linkers to span the domains. In the ribbon diagram, spheres indicate the N terminus and C terminus of this domain. 27 Å is the distance that the linker needs to span to link the C terminus of one domain to the N terminus of the next domain. B, all three single-chain constructs are increased in stability to urea denaturation compared with the wild-type C-domain. C, equilibrium analytical ultracentrifugation at 30 µM protein concentration demonstrates that C15C is monomeric. The theoretical distribution expected for a dimer is shown for comparison.

Based on its design, the C15C domain should be monomeric when intramolecular C domain interactions outcompete intermolecular association. At very high concentrations it should be possible for intermolecular C15C association to occur. We can use the concentration-dependent stability of the wild-type C domain to estimate the protein concentration that would be required for intermolecular association to energetically equal intramolecular interactions. Based on the concentration-dependent stability of wild-type C domain (Fig. 1B), we estimate that a concentration of 2 mM would be required to achieve the same C_m to urea denaturation as C15C. Thus, 2 mM represents the effective concentration brought about by the covalent linkage of C domains. Effective concentrations in the millimolar range have been reported for single-chain constructs of similar length (29). From this analysis, we predict that at a concentration of 2 mM the C15C construct will equally partition between monomers and dimers, but that at physiological Hsp90 concentration (10 µM) C15C will be almost entirely monomeric. This analysis is supported by our observation that at 30 µM, C15C is monomeric as determined by analytical ultracentrifugation (Fig. 2C).

Full-length NMCC Monomer Design Is Deficient for ATP Hydrolysis—To explore the function of full-length Hsp90 monomers, we generated an NMCC construct (Fig. 1C) with the 15-amino acid linker between the two C domains. The single-chain C domain in NMCC should hinder dimerization, resulting in a molecule with single N and M domains, the location of ATP hydrolysis as well as numerous interactions with co-chaperones and substrates (11, 17, 31-33). Equilibrium analytical ultracentrifugation was used to assess the oligomeric state of NMCC (Fig. 3A). The radial distribution of NMCC at 12 µM during centrifugation fits well to a single molecular mass species with a mass of 103 kDa, similar to the mass of a monomer (99 kDa). The residuals from the single species fit are randomly distributed indicating that higher order oligomers are not detected. We find that under identical centrifugation and concentration (12 µM) conditions, wild-type Hsp90 is dimeric (Fig. S2). These results demonstrate that the single-chain C domain effectively blocks dimerization in the full-length NMCC construct.

How does blocking Hsp90 dimerization effect ATPase activity in vitro? Using an enzymatically coupled ATPase assay, we observed that NMCC was severely impaired for ATPase activity relative to wild-type Hsp90 (Fig. 3B). The ATPase activity of NMCC was inhibited by GA, a small-molecule inhibitor with high specificity for Hsp90. These results indicate that the ATPase domain in NMCC can bind to both ATP and GA. To determine the fraction of NMCC molecules capable of binding to GA, we performed a stoichiometric binding experiment (Fig. 3C). A concentration of fluorescently labeled BODIPY-GA of 1 µM was used and Hsp90 concentrations of WT and NMCC were increased to determine the ratio of Hsp90 to GA where binding was saturated. For both wild-type and NMCC Hsp90, binding saturates at about 1.4 Hsp90 subunits to 1 GA. Taking into account uncertainty in concentration determinations and possible non-fluorescent contaminants in the BODIPY-GA preparation, these experiments indicate that both wild-type and NMCC are capable of high-affinity binding of 1 molecule of GA per Hsp90 subunit. These results indicate that the ATP binding site of NMCC is structured similarly to wild-type Hsp90. Consistent with the engineering strategy of NMCC, the single-chain C domain does not destroy the ATP or GA binding function of the N domain.

View larger version (22K): [in this window] [in a new window]   FIGURE 3. Full-length NMCC construct is predominantly monomeric and is defective for ATP hydrolysis. A, the distribution of 12 µM NMCC in equilibrium analytical ultracentrifugation fits well to a single species with the molecular weight of a monomer. The theoretical distribution expected for a dimer is shown for comparison. B, at 6 µM protein concentration NMCC is more than 10-fold slower at ATP hydrolysis than wild-type Hsp90 and can be inhibited with GA, a specific inhibitor for the ATP binding site of Hsp90. C, stoichiometry binding experiment indicates that binding of Hsp90 to BODIPY-GA saturates at a ratio of 1 GA molecule per 1 Hsp90 subunit for both wild-type and NMCC Hsp90. D, the specific ATPase activity of NMCC Hsp90 is concentration dependent.

The linear increase in specific activity of NMCC indicates that these experiments do not saturate the dimer species over this concentration range, preventing the direct determination of the dimer dissociation constant. Assuming that the NMCC dimers have the ATPase activity of wild-type dimers, we estimate that NMCC has a dimer dissociation constant of 200 µM under the conditions of the ATPase assay. This result is consistent with interactions between the N and M domains driving weak association of NMCC. Consistent with this interpretation, NM constructs show concentration-dependent ATPase activity with similar apparent dimer dissociation constants (9). In addition, this dissociation constant for NMCC is about 10-fold tighter than the dissociation constant we estimate for the single-chain C domain alone that corresponds to a free energy of NM association of about 5 kJ/mol. Based on the ATPase activity at low NMCC concentration, we estimate the upper limit for monomer activity as 0.05 min^-1, or 3% of the rate of wild-type Hsp90.

Function of NMCC in Vivo—Hsp90 is an essential gene in eukaryotes (13-15). Interestingly, Hsp90 point mutants with in vitro ATPase levels less than 2% of wild-type can function to support yeast viability (12, 18). Therefore, we tested the ability of NMCC to support viability as the sole Hsp90 in yeast. Both NMCC and wild-type Hsp90 with N-terminal His₆ epitope tags were constitutively expressed from a plasmid system previously shown to accumulate Hsp90 to near wild-type levels (34). These plasmids were introduced into two yeast strains: one whose other source of Hsp90 was a temperature-sensitive allele, and one whose other source of Hsp90 was encoded on a URA3 plasmid that can be swapped out using 5-fluoroorotic acid, which is converted to the toxic fluoroorotidine monophosphate when the URA3 gene is expressed. Under conditions where active dimeric wild-type Hsp90 is present, expression of NMCC did not compromise cell growth in either strain (Fig. S3). Under conditions that select for NMCC as the sole active Hsp90, both strains failed to grow (Fig. 4).

The failure to support growth could be due to a low expression level of NMCC. To test this possibility, we analyzed expression levels by Western blotting cell lysates for the quantity of His₆ epitope tag (Fig. 4C). Expression levels of His₆-tagged wild-type and NMCC Hsp90 is similar. In addition, the SDS mobility of NMCC indicates that it is full-length and that degradation products do not accumulate. Thus the glycine-rich linker between the C domains is proteolytically stable in yeast. By comparing the Western signal to a standard curve of purified His₆ Hsp90 in cell lysate (supplemental Fig. S3), we estimate that His₆ wild-type Hsp90 accumulates to 130,000 copies per cell and NMCC accumulates to 110,000 copies per cell. Assuming an average cell volume of 40 fl (35), and excluding nuclear volume (36), we estimate the cytoplasmic concentrations as 7 µM for NMCC and 8 µM for wild-type. Wild-type Hsp90 is predominantly dimeric at these concentrations and NMCC is predominantly monomeric. Together, these results indicate that monomeric NMCC cannot function as the sole Hsp90 in vivo.

View larger version (18K): [in this window] [in a new window]   FIGURE 4. NMCC as the sole Hsp90 does not support yeast viability. A, NMCC does not rescue growth of a Hsp90ts strain at 37 °C. B, NMCC does not rescue growth at 23 °C in the absence of other Hsp90 sources. C, Western blot analysis of cell lysates demonstrates that the NMCC protein accumulates to similar levels as wild-type Hsp90.

View larger version (19K): [in this window] [in a new window]   FIGURE 5. In yeast cells, NMCC does not aid in the activation of v-src or GR. A, Western blots to detect v-src levels and phosphotyrosine levels in yeast with no active Hsp90, WT Hsp90, or NMCC Hsp90. In yeast with NMCC and all other Hsp90 inactive, v-src activity is compromised relative to cells with wild-type Hsp90. B, quantification of phosphotyrosine and v-src levels. C, reporter activity from a glucocorticoid response element driven promoter is defective with NMCC compared with wild-type Hsp90. DOCS, deoxycorticosterone.

V-src is a promiscuous tyrosine kinase. Yeast have very low background levels of phosphotyrosine and expression of v-src causes a dramatic and Hsp90-dependent increase in phosphotyrosine (38). When v-src is induced from a galactose-inducible promoter in the presence of wild-type Hsp90, Western blotting with an anti-phosphotyrosine antibody shows that many yeast proteins are phosphorylated by v-src (Fig. 5A). In contrast, when v-src is induced with NMCC Hsp90, the v-src kinase accumulates, but proteins with phosphotyrosine do not accumulate indicating that v-src is inactive (Fig. 5, A and B). NMCC Hsp90 does not activate v-src in vivo.

In yeast, GR responds to steroid agonists including deoxycorticosterone to enhance expression from promoters with glucocorticoid response elements (39). When yeast expressing GR are stimulated with deoxycorticosterone, wild-type Hsp90 aids in the hormone-induced expression of β-galactosidase from a glucocorticoid response element containing promoter (Fig. 5B). In contrast, NMCC does not increase reporter expression relative to cells with no active Hsp90. NMCC Hsp90 does not activate GR in vivo.

NMCC Function Is Rescued by Addition of a Coiled-coil Dimerization Motif—The lack of function observed for NMCC Hsp90 could be due to lack of dimerization or some other unintended consequence of the engineering strategy. For example, function of Hsp90 may require transient dissociation of the C domain and NMCC may be non-functional because it slows C domain dissociation. To address these concerns, we appended a coiled-coil dimerization domain to generate NMCCcoil (Fig. 6A). The coiled-coil should induce this construct to form dimers. If NMCC is functionally defective because it is monomeric, then NMCCcoil should rescue both dimerization and function. In size exclusion chromatography NMCCcoil has a radius of gyration consistent with the dominant solution form being dimeric, with a small amount of tetramer also observed (Fig. 6B).

Is NMCCcoil functionally active? It is about 70% as active as wild-type Hsp90 for ATPase activity in vitro (Fig. 6C). In addition, NMCCcoil enables yeast growth as the sole source of Hsp90 (Fig. 6D). NMCCcoil expresses to about one-third the level of wild-type Hsp90 (supplemental Fig. S3). Consistent with previous findings that decreased Hsp90 levels result in reduced growth at elevated temperatures (13), yeast with NMCCcoil are temperature sensitive (data not shown). Our analysis of NMCCcoil demonstrates that the coiled-coil domain rescues dimerization, in vitro ATPase activity, and the ability to serve as the sole Hsp90 in yeast. Because coiled-coil-induced dimerization rescues function in NMCCcoil, we conclude that the parental NMCC construct is defective because it is monomeric.

View larger version (27K): [in this window] [in a new window]   FIGURE 6. Dimerization of NMCCcoil rescues in vitro and in vivo function. A, architecture and putative structural model of the NMCCcoil dimer. B, size exclusion chromatography indicates that appending the coiled-coil domain transforms the parental NMCC monomer into a predominantly dimeric NMCCcoil. C, ATPase activity of NMCCcoil is rescued compared with NMCC. D, as the sole Hsp90 source, NMCCcoil supports yeast viability. FOA, 5-fluoroorotic acid.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION CONCLUSIONS REFERENCES

We find that NMCC monomers have an in vitro ATPase activity that is at most 3% the level of wild-type Hsp90. This level of ATPase activity is at the extreme low end of the spectrum of ATPase activities that have been observed in Hsp90 point mutants that support yeast growth (18). Of note, studies on mixtures of full-length and C domain constructs show that full-length/C-domain heterodimers have an ATPase activity 30% that of wild-type homodimers (9), well within the range of ATPase activity required for in vivo activity. Why does the NMCC monomer have lower in vitro ATPase activity compared with full-length/C-domain heterodimers? The ATPase activity measured for NMCC monomers was determined at low NMCC concentrations where monomers dominate. The ATPase activity for full-length/C-domain heterodimers was determined by titrating increasing concentrations of C-domain and assuming equal dimerization probability for all species. This last assumption was based on the observation that both full-length and C-domain constructs have similar dimerization constants in the absence of nucleotide. Under the conditions of the ATPase assay nucleotide promoted dimerization of the NM domains may favor full-length homodimers over full-length/C-domain heterodimers. In this case the activity of full-length/C-domain heterodimers would be an overestimate.

Human Hsp90 has high sequence conservation with yeast Hsp90 (58% sequence identity), but different ATPase properties (41, 42). The in vitro ATPase activity of human Hsp90 is about 10% that of yeast. A monomeric NM construct of human Hsp90 is reduced in activity by 10-fold relative to the full-length, suggesting that dimerization is important for efficient ATP hydrolysis in human Hsp90 as well as in yeast. However, human NM constructs do not exhibit concentration-dependent ATPase activity at protein concentrations up to 100 µM. Thus the activity of human NM Hsp90 monomers appears to be about 1% the rate of full-length yeast Hsp90. With the data in this paper we cannot accurately determine the level of ATPase activity of yeast NMCC monomers, but find that it is less than 3% of full-length yeast Hsp90 and it may be similar to monomers of human Hsp90.

Two possible explanations for the human NM results are that either human Hsp90 does not require NM dimerization for efficient ATPase hydrolysis, or the dimerization constant of NM is weaker in human Hsp90 than in yeast. We note that there are two amino acid changes at the NM dimer interface (V23F and L378I, yeast to human) that may reduce the affinity of this interface in human Hsp90. Also, human NM constructs have greatly reduced ATPase activity compared with full-length, indicating that dimerization contributes to efficient ATPase activity. Alternatively, it is possible that human NM constructs have reduced activity because of removal of the C domain. The monomer design strategy used in this paper could be used to differentiate these possibilities by generating human monomers with intact C domains. Lastly, human Hsp90 rescues yeast with endogenous Hsp90 knocked out (39), indicating that human Hsp90 functions with yeast co-chaperones and substrates, some of which bind in a nucleotide-dependent manner to Hsp90 (3, 43). These lines of reasoning suggest that human Hsp90 functions by the same basic mechanism as yeast Hsp90, but that the affinity between the NM domains is stronger in yeast.

Changes in the in vitro ATPase activity of Hsp90 correlate loosely with in vivo function. Point mutations that destroy in vitro ATPase activity are non-functional in yeast (44, 45), but mutations that alter ATPase activity from 2 to 400% of wild-type levels are tolerable under non-stressful conditions (18). The in vivo ATPase activity of Hsp90 mutants may be buffered through binding to co-chaperones. Hundreds of yeast proteins interact with Hsp90 (2), and at least three have been shown to impact the ATPase rate of Hsp90 (46-48). Modulating the level of one of these co-chaperones in human cells has been shown to dramatically alter the activity level of an Hsp90 substrate (49). One of the unmet challenges in the study of Hsp90 is to understand how ATPase hydrolysis contributes to substrate activation.

NMCC will be a useful model for analyzing the binding properties of monomers to co-chaperones and substrates and the potential role of monomers in the Hsp90 conformational cycle. At cytoplasmic concentrations, Hsp90 is predominantly dimeric. However, Hsp90 subunits exchange rapidly, on the same time scale as ATP hydrolysis, suggesting that monomers of Hsp90 form transiently during the activation of substrates (9). In addition to transient formation of Hsp90 monomers, the equilibrium population of monomeric Hsp90 in the cytosol is 1000 copies/cell (based on expression level and the observed dimer dissociation constant) compared with 130,000 copies of Hsp90 dimers. Some substrates or co-chaperones may preferentially interact with Hsp90 monomers during part of the Hsp90 chaperone cycle. In addition, the binding of Hsp90 to transcription factors and other proteins with nuclear localization signals causes Hsp90 to traffic to the nucleus where it can accumulate at low abundance (50). The low concentration of Hsp90 in the nucleus should favor dimer dissociation, and perhaps Hsp90 monomers have a function in the nucleus. NMCC Hsp90 provides a tool to study full-length Hsp90 monomers both in vitro and inside cells.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
We find that the in vitro ATPase activity of NMCC is at best at the extreme low end of the range observed to support yeast viability and that NMCC does not support yeast viability. These observations indicate that the cellular environment is not sufficient to rescue the chaperone function of monomeric Hsp90. In addition, NMCC does not increase the activation of v-src nor GR in cells, indicating that Hsp90 dimerization is required to chaperone these specific substrates in yeast.

	FOOTNOTES

 
^* This work was supported by a research grant from the Worcester Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The on-line version of this article (available at http://www.jbc.org) contains supplemental Figs. S1-S4.

¹ To whom correspondence should be addressed: 364 Plantation St., LRB 922, Worcester, MA 01605. Fax: 508-856-6464; E-mail: Dan.Bolon{at}umassmed.edu.

² The abbreviations used are: GA, geldanamycin; C_m, concentration midpoint; GPD, glyceraldehyde 3-phosphate dehydrogenase; GR, glucocorticoid receptor; AMP-PNP, adenosine 5'-(β,-imino)triphosphate; WT, wild-type.

	ACKNOWLEDGMENTS

 
We thank S. Miller and A. Bhunia for advice and aid with the synthesis of BODIPY-GA and B. Middlebrook for aid with cloning. We also thank N. Rhind and K. Knight for fruitful discussions. Yeast strains ECU82a and iG170Da as well as plasmids Y316v-src and P2A/GRGZ were generously provided by S. Lindquist. Analytical ultracentrifugation was performed at the University of Massachusetts Medical School Ultracentrifugation Facility with the aid of K. Crowley. C. R. Matthews generously provided CD time. We thank C. S. Sevier and L. Cowen for technical advice regarding yeast genetics.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

1. Nathan, D. F., Vos, M. H., and Lindquist, S. (1997) Proc. Natl. Acad. Sci. U. S. A. 94, 12949-12956[Abstract/Free Full Text]
2. Zhao, R., Davey, M., Hsu, Y. C., Kaplanek, P., Tong, A., Parsons, A. B., Krogan, N., Cagney, G., Mai, D., Greenblatt, J., Boone, C., Emili, A., and Houry, W. A. (2005) Cell 120, 715-727[CrossRef][Medline] [Order article via Infotrieve]
3. Pratt, W. B., and Toft, D. O. (1997) Endocr. Rev. 18, 306-360[Abstract/Free Full Text]
4. Smith, D. F., Stensgard, B. A., Welch, W. J., and Toft, D. O. (1992) J. Biol. Chem. 267, 1350-1356[Abstract/Free Full Text]
5. Uehara, Y., Hori, M., Takeuchi, T., and Umezawa, H. (1986) Mol. Cell. Biol. 6, 2198-2206[Abstract/Free Full Text]
6. Whitesell, L., Mimnaugh, E. G., De Costa, B., Myers, C. E., and Neckers, L. M. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 8324-8328[Abstract/Free Full Text]
7. Whitesell, L., and Lindquist, S. L. (2005) Nat. Rev. Cancer 5, 761-772[CrossRef][Medline] [Order article via Infotrieve]
8. Arlander, S. J., Felts, S. J., Wagner, J. M., Stensgard, B., Toft, D. O., and Karnitz, L. M. (2006) J. Biol. Chem. 281, 2989-2998[Abstract/Free Full Text]
9. Richter, K., Muschler, P., Hainzl, O., and Buchner, J. (2001) J. Biol. Chem. 276, 33689-33696[Abstract/Free Full Text]
10. Louvion, J. F., Warth, R., and Picard, D. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 13937-13942[Abstract/Free Full Text]
11. Ali, M. M., Roe, S. M., Vaughan, C. K., Meyer, P., Panaretou, B., Piper, P. W., Prodromou, C., and Pearl, L. H. (2006) Nature 440, 1013-1017[CrossRef][Medline] [Order article via Infotrieve]
12. Nathan, D. F., and Lindquist, S. (1995) Mol. Cell. Biol. 15, 3917-3925[Abstract]
13. Borkovich, K. A., Farrelly, F. W., Finkelstein, D. B., Taulien, J., and Lindquist, S. (1989) Mol. Cell. Biol. 9, 3919-3930[Abstract/Free Full Text]
14. Birnby, D. A., Link, E. M., Vowels, J. J., Tian, H., Colacurcio, P. L., and Thomas, J. H. (2000) Genetics 155, 85-104[Abstract/Free Full Text]
15. Yue, L., Karr, T. L., Nathan, D. F., Swift, H., Srinivasan, S., and Lindquist, S. (1999) Genetics 151, 1065-1079[Abstract/Free Full Text]
16. Chadli, A., Bouhouche, I., Sullivan, W., Stensgard, B., McMahon, N., Catelli, M. G., and Toft, D. O. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 12524-12529[Abstract/Free Full Text]
17. Meyer, P., Prodromou, C., Hu, B., Vaughan, C., Roe, S. M., Panaretou, B., Piper, P. W., and Pearl, L. H. (2003) Mol. Cell 11, 647-658[CrossRef][Medline] [Order article via Infotrieve]
18. Prodromou, C., Panaretou, B., Chohan, S., Siligardi, G., O'Brien, R., Ladbury, J. E., Roe, S. M., Piper, P. W., and Pearl, L. H. (2000) EMBO J. 19, 4383-4392[CrossRef][Medline] [Order article via Infotrieve]
19. Wegele, H., Muschler, P., Bunck, M., Reinstein, J., and Buchner, J. (2003) J. Biol. Chem. 278, 39303-39310[Abstract/Free Full Text]
20. Bowie, J. U., and Sauer, R. T. (1989) Biochemistry 28, 7139-7143[CrossRef][Medline] [Order article via Infotrieve]
21. Santoro, M. M., and Bolen, D. W. (1988) Biochemistry 27, 8063-8068[CrossRef][Medline] [Order article via Infotrieve]
22. Cantor, C. R., and Schimmel, P. R. (1980) Biophysical Chemistry, W.H. Freeman and Co., New York
23. Norby, J. G. (1988) Methods Enzymol. 156, 116-119[Medline] [Order article via Infotrieve]
24. Llauger-Bufi, L., Felts, S. J., Huezo, H., Rosen, N., and Chiosis, G. (2003) Bioorg. Med. Chem. Lett. 13, 3975-3978[CrossRef][Medline] [Order article via Infotrieve]
25. Lakowicz, J. R. (1999) Principles of Fluorescence Spectroscopy, Second Ed., Kluwer Academic/Plenum Publishers, Boston, MA
26. Maroney, A. C., Marugan, J. J., Mezzasalma, T. M., Barnakov, A. N., Garrabrant, T. A., Weaner, L. E., Jones, W. J., Barnakova, L. A., Koblish, H. K., Todd, M. J., Masucci, J. A., Deckman, I. C., Galemmo, R. A., Jr., and Johnson, D. L. (2006) Biochemistry 45, 5678-5685[CrossRef][Medline] [Order article via Infotrieve]
27. Mumberg, D., Muller, R., and Funk, M. (1995) Gene (Amst.) 156, 119-122[CrossRef][Medline] [Order article via Infotrieve]
28. Bornberg-Bauer, E., and Chan, H. S. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 10689-10694[Abstract/Free Full Text]
29. Robinson, C. R., and Sauer, R. T. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 5929-5934[Abstract/Free Full Text]
30. Martin, A., Baker, T. A., and Sauer, R. T. (2005) Nature 437, 1115-1120[CrossRef][Medline] [Order article via Infotrieve]
31. Prodromou, C., Roe, S. M., O'Brien, R., Ladbury, J. E., Piper, P. W., and Pearl, L. H. (1997) Cell 90, 65-75[CrossRef][Medline] [Order article via Infotrieve]
32. Vaughan, C. K., Gohlke, U., Sobott, F., Good, V. M., Ali, M. M., Prodromou, C., Robinson, C. V., Saibil, H. R., and Pearl, L. H. (2006) Mol. Cell 23, 697-707[CrossRef][Medline] [Order article via Infotrieve]
33. Plescia, J., Salz, W., Xia, F., Pennati, M., Zaffaroni, N., Daidone, M. G., Meli, M., Dohi, T., Fortugno, P., Nefedova, Y., Gabrilovich, D. I., Colombo, G., and Altieri, D. C. (2005) Cancer Cell 7, 457-468[CrossRef][Medline] [Order article via Infotrieve]
34. Chang, H. C., and Lindquist, S. (1994) J. Biol. Chem. 269, 24983-24988[Abstract/Free Full Text]
35. Johnston, G. C., Ehrhardt, C. W., Lorincz, A., and Carter, B. L. (1979) J. Bacteriol. 137, 1-5[Abstract/Free Full Text]
36. Larabell, C. A., and Le Gros, M. A. (2004) Mol. Biol. Cell 15, 957-962[Abstract/Free Full Text]
37. Smith, D. F., Schowalter, D. B., Kost, S. L., and Toft, D. O. (1990) Mol. Endocrinol. 4, 1704-1711[Abstract]
38. Xu, Y., and Lindquist, S. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 7074-7078[Abstract/Free Full Text]
39. Picard, D., Khursheed, B., Garabedian, M. J., Fortin, M. G., Lindquist, S., and Yamamoto, K. R. (1990) Nature 348, 166-168[CrossRef][Medline] [Order article via Infotrieve]
40. Prodromou, C., Roe, S. M., Piper, P. W., and Pearl, L. H. (1997) Nat. Struct. Biol. 4, 477-482[CrossRef][Medline] [Order article via Infotrieve]
41. McLaughlin, S. H., Ventouras, L. A., Lobbezoo, B., and Jackson, S. E. (2004) J. Mol. Biol. 344, 813-826[CrossRef][Medline] [Order article via Infotrieve]
42. McLaughlin, S. H., Smith, H. W., and Jackson, S. E. (2002) J. Mol. Biol. 315, 787-798[CrossRef][Medline] [Order article via Infotrieve]
43. McLaughlin, S. H., Sobott, F., Yao, Z. P., Zhang, W., Nielsen, P. R., Grossmann, J. G., Laue, E. D., Robinson, C. V., and Jackson, S. E. (2006) J. Mol. Biol. 356, 746-758[CrossRef][Medline] [Order article via Infotrieve]
44. Panaretou, B., Prodromou, C., Roe, S. M., O'Brien, R., Ladbury, J. E., Piper, P. W., and Pearl, L. H. (1998) EMBO J. 17, 4829-4836[CrossRef][Medline] [Order article via Infotrieve]
45. Obermann, W. M., Sondermann, H., Russo, A. A., Pavletich, N. P., and Hartl, F. U. (1998) J. Cell Biol. 143, 901-910[Abstract/Free Full Text]
46. Richter, K., Walter, S., and Buchner, J. (2004) J. Mol. Biol. 342, 1403-1413[CrossRef][Medline] [Order article via Infotrieve]
47. Panaretou, B., Siligardi, G., Meyer, P., Maloney, A., Sullivan, J. K., Singh, S., Millson, S. H., Clarke, P. A., Naaby-Hansen, S., Stein, R., Cramer, R., Mollapour, M., Workman, P., Piper, P. W., Pearl, L. H., and Prodromou, C. (2002) Mol. Cell 10, 1307-1318[CrossRef][Medline] [Order article via Infotrieve]
48. Prodromou, C., Siligardi, G., O'Brien, R., Woolfson, D. N., Regan, L., Panaretou, B., Ladbury, J. E., Piper, P. W., and Pearl, L. H. (1999) EMBO J. 18, 754-762[CrossRef][Medline] [Order article via Infotrieve]
49. Wang, X., Venable, J., LaPointe, P., Hutt, D. M., Koulov, A. V., Coppinger, J., Gurkan, C., Kellner, W., Matteson, J., Plutner, H., Riordan, J. R., Kelly, J. W., Yates, J. R., 3rd, and Balch, W. E. (2006) Cell 127, 803-815[CrossRef][Medline] [Order article via Infotrieve]
50. Kang, K. I., Devin, J., Cadepond, F., Jibard, N., Guiochon-Mantel, A., Baulieu, E. E., and Catelli, M. G. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 340-344[Abstract/Free Full Text]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles: