Intrinsic ADP-ATP Exchange Activity Is a Novel Function of the Molecular Chaperone, Hsp70*

Hsp70 is a multifunctional molecular chaperone whose interactions with protein substrates are regulated by ATP hydrolysis and ADP-ATP exchange. We show here that, in addition to ATPase activity, purified Hsp70 free from nucleoside-diphosphate (NDP) kinase exhibits intrinsic ADP-ATP exchange activity. The rate constants for ATP hydrolysis and ATP synthesis were in a similar range at the optimum pH of 7.5–8.5 in the presence of 5 mm ATP and 0.5 mm ADP. Hsp70 exhibited a considerably strict preference for ATP as a phosphate donor, and a biased substrate specificity, unlike NDP kinase; ADP, UDP, CDP > dTDP, dCDP > GDP, dGDP. During the reaction, Hsp70 formed an acid-labile autophosphorylated intermediate, and nucleoside diphosphate-dependent dephosphorylation of the latter then occurred. These properties of Hsp70 are not identical but similar to those of NDP kinase, but are not similar to those of adenylate kinase and ATP synthase.

of the nucleotide exchange reaction of Hsp70 in the cytosol. In this paper, we first report intrinsic ADP-ATP exchange activity as a novel function of Hsp70. The characteristics of the enzyme activity are similar to those of nucleoside-diphosphate (NDP 1 ) kinases, but not to those of adenylate kinase and ATP synthase.
Purification of Hsp70 -Because Hsp70 purchased from Sigma exhibits greater than 95% purity but there is still a possibility of contaminating proteins with affinity for ATP, such as NDP kinase (17), the preparation of Hsp70 was further purified to homogeneity by HPLC on a Mono Q anion-exchange column and then on a gel permeation TSK-Gel G3000SW column. Briefly, commercially available Hsp70 (600 g) in Buffer A (25 mM Tris-HCl, pH 7.2, 0.1 mM EDTA, and 0.5 mM dithiothreitol) was applied to a Mono Q column and then eluted with a narrow salt gradient of 0 -0.15 M KCl in Buffer A as shown in Fig. 1, which is able to separate conventional NDP kinase from Hsp70 (17). Hsp70 was eluted with approximately 0.1 M KCl and separated from contaminating proteins. Conventional NDP kinase, eluted with approximately 0.04 and 0.07 M KCl, was not detected among the contaminants in the commercial products of Hsp70 on that chromatography. The fractions of Hsp70 were concentrated and then subjected to HPLC on a double-linked TSK-Gel G3000 SW column (7.5 ϫ 600 mm each) with 50 mM ammonium formate buffer, pH 5.5, 0.1 mM EDTA, and 0.5 mM dithiothreitol. The final preparation exhibited a purity of greater than 99% and was judged to be free from conventional NDP kinase on silver-stained SDS-PAGE.
Assaying of ATP Hydrolysis and ATP Synthesis-Although almost all the data for Hsp70 ATPase reported were obtained by measurement of inorganic phosphate liberated from ATP, we found that Hsp70 exhibits the reverse activity, i.e. ADP-ATP exchange, and thus the amount of inorganic phosphate in the reaction is not reflected by the ATPase activity. In this regard, we analyzed the ATPase activity by measuring the conversion of [ 14 C]ATP to [ 14 C]ADP. The ATP synthesis activity was analyzed by measuring the conversion of [ 14 C]ADP to [ 14 C]ATP as described (18,19). The reactions were carried out at 37°C for 30 min in 100 mM Hepes-KOH buffer, pH 8, containing 5 mM ATP, 0.5 mM ADP, and 6 mM MgCl 2 with 0 -20 pmol of Hsp70 (monomeric form), and 0.05 Ci of [8-14 C]ATP and 0.02 Ci of [8-14 C]ADP for the assaying of ATPase and ATP synthesis activities, respectively, in a total volume of 10 l. After incubation, the reaction mixtures were immediately spotted onto a polyethyleneimine-cellulose TLC plate (Macherey-Nagel). ADP and ATP were separated by ascending chromatography with 1 M formic acid containing 0.7 M LiCl, and the radioactivity in the resolved spots was quantitated with a Bio Imaging Analyzer BAS 1500 (Fuji Photo Film Co., Ltd., Japan).
Nucleotide Specificity as a Phosphate Acceptor of Hsp70 -The nucleotide specificity of Hsp70 was examined at 37°C for 30 min in 100 mM Hepes-KOH buffer, pH 8, containing 1 Ci of [␥-32 P]ATP and 6 mM MgCl 2 with 10 pmol of Hsp70 and 0.5 mM of various ribo-or deoxyribonucleoside diphosphates as phosphate acceptors in a 10-l reaction mixture. After incubation, the nucleotides were resolved by polyethyl-eneimine-cellulose TLC with 0.75 M KH 2 PO 4 , pH 3.5, followed by quantification with an imaging analyzer.
Detection of an Autophosphorylated Hsp70 Intermediate-Autophosphorylation and CDP-dependent dephosphorylation of Hsp70 were analyzed using 10 g of Hsp70, 10 Ci of [␥-32 P]ATP (6000 Ci/mmol), 100 M ATP, and 6 mM MgCl 2 in 100 mM Hepes-KOH buffer, pH 8, in the absence and presence of 5 mM CDP, respectively, in a total volume of 10 l. After incubation at 37°C for 2 h, both reactions were quenched by the further addition of 10 mM EDTA, and half of each sample was then treated with the traditional (pH 6.8) SDS sample buffer without boiling and subjected to 15% SDS-PAGE. After electrophoresis, the gel was dried without acid fixation and analyzed with an imaging analyzer. The remaining samples were analyzed as to the acid and base stability of the phosphorylated Hsp70 as described (20). These samples were then analyzed with an imaging analyzer.

Identification of ADP-ATP Exchange Activity of Hsp70 -We
found the purified Hsp70 free from NDP kinase exhibited ADP-ATP exchange activity, in addition to ATPase activity. As shown in Fig. 2A, Hsp70 exhibited weak ATPase activity in a dose-dependent manner, but no activity was observed for the control protein, BSA. Although ADP is a product inhibitor of common ATPase, a limited amount of ADP, i.e. concentrations between 0.1 and 1.0 mM, in the reaction mixture stimulated the ATPase activity of Hsp70 by 2-3.7 fold, as shown in Fig. 2A. Inhibition by ADP was then observed at concentrations in excess of 1 mM (data not shown). AMP had no effect on this activity.
The reverse reaction of ATP hydrolysis, i.e. ATP synthesis activity, of Hsp70 was analyzed (Fig. 2B). Surprisingly, Hsp70 catalyzed ATP synthesis at a slow but linear rate in the presence of substrate ADP but not in the presence of AMP. The ATP synthesis activity of Hsp70 gave saturation curves with ADP as a phosphate acceptor and with ATP as a phosphate donor. The reaction of ATP synthesis, as catalyzed by Hsp70, was entirely different from that of adenylate kinase, which catalyzes the reversible ATP-dependent synthesis of ADP from AMP, as shown in Fig. 3C. Furthermore, the reaction was different from that of conventional ATP synthase, which catalyzes ATP synthesis from ADP and P i . When ATP was replaced with 5 mM P i , Hsp70 failed to produce ATP from ADP (Fig. 2B). The conversion of ADP to ATP, i.e. ADP-ATP exchange, catalyzed by Hsp70 was similar to the reaction catalyzed by NDP kinase, which catalyzes the nucleoside triphosphate-dependent synthesis of ribo-and deoxyribonucleoside triphosphates from the corresponding diphosphates (21)(22)(23). No ATP synthesis activity was observed for the control protein, BSA (Fig. 2B).
From the kinetic results for ATP hydrolysis and for ATP synthesis, the K m value of ATP for the ATP hydrolysis in the presence of 0.5 mM ADP and the K m value of ADP for the ATP synthesis of Hsp70 were calculated to be 4.3 mM and 0.31 mM, respectively (data not shown). The specific activities of Hsp70 for ATP synthesis at pH 7.4, measured under the conditions in  1-8) and deoxyribonucleoside diphosphates (lanes 9 -16), in 100 mM Hepes-KOH buffer, pH 8, in a total volume of 10 l. After incubation at 37°C for 30 min, samples were analyzed by TLC, followed by autoradiography with an imaging analyzer. C, the adenylate kinase activity of Hsp70 (lane 2) and the activity of adenylate kinase from chicken muscle (lane 3) were assayed as described (17,19) (15,24), which were determined by measuring the release of inorganic phosphate from ATP. Since Hsp70 exhibits both the activities of ATP hydrolysis and ATP synthesis, the kinetic constants for Hsp70 ATPase determined by measurement of inorganic phosphate in the reaction are not correct. Furthermore, the K m value of ATP for the ATPase activity of Hsp70 in our experiments is much higher than the values reported in the range of 0.7-1.4 M (15, 24) and is in a range similar to that reported for NDP kinase (25), because the K m value was determined with 0.5 mM ADP in our experiments. Since both the K m values are in a range similar to the concentrations of ATP (Ϸ5 mM) and ADP (Ϸ0.5 mM) in the cytosol, changes in substrate availability would be expected to significantly affect the in vivo enzyme reaction.
The pH optima for Hsp70 ATPase and ATP synthesis were determined using 100 mM Mes buffer (pH 5.5-6), Hepes buffer (pH 7-8), Ches buffer (pH 9), and Chaps buffer (pH 10). Without ADP, the pH optimum for ATPase was found to be 7, consistent with a previous report (26). The optimum pH of the stimulated Hsp70 ATPase in the presence of 0.5 mM ADP, however, shifted to 7.5-9. The pH optimum for Hsp70 ATP synthesis was 7-9, which overlapped that of the stimulated ATPase. It is noteworthy that, at pH 7.5-8.5, the rate constants for ATP hydrolysis and ATP synthesis of Hsp70 were similar in the presence of 5 mM ATP and 0.5 mM ADP in the reaction mixture. The ATP synthesis activities were slightly higher below pH 7.5 and lower above pH 8.5 than the ATP hydrolysis activities.
Hsp70 Functions as a NDP Kinase-like Enzyme-NDP kinase is known to use any nucleoside triphosphate as a phosphate donor with similar efficiency. Hsp70, however, utilized ATP most efficiently, with the other ribo-and deoxyribonucleoside triphosphates being utilized at rates 16 -33% of the rate of ATP utilization (data not shown). NDP kinase transfers the terminal phosphate of a nucleoside triphosphate to a nucleoside diphosphate. To determine whether this mechanism is the same for Hsp70, ATP analogs, such as ATP␥S and AMP-PNP, were added together with ATP to the reaction mixture, and then the transfer of the ␥-phosphate from ATP to ADP was analyzed. This transfer by both NDP kinase and Hsp70 was competitively inhibited by ϳ60 -50% in the presence of equal concentrations of ATP␥S and ATP and was completely inhibited by 5-fold excess concentrations of these analogs over those of ATP (data not shown).
Next we examined the specificity of each nucleotide as an acceptor of the transfer of the [␥-32 P]phosphate of ATP, as catalyzed by Hsp70. As shown in Fig. 3, A and B, NDP kinase converted all ribo-and deoxyribonucleoside diphosphates to the corresponding nucleoside triphosphates with almost similar efficiency, consistent with previous reports (25,27,28). Hsp70, however, exhibited a biased substrate specificity, i.e. UDP, CDP Ͼ dTDP, dCDP Ͼ GDP, dGDP. Under these assay conditions, the conversion of ADP and dADP to ATP and dATP, respectively, could not be analyzed, because the newly formed products overlapped the phosphate donor, [␥-32 P]ATP. Nevertheless, it was confirmed that the rates of conversion by Hsp70 of [ 14 C]ADP and [ 14 C]CDP to the corresponding nucleoside triphosphates exhibited almost similar efficiency (data not shown).
Since NDP kinase requires Mg 2ϩ or Mn 2ϩ for its activity, the effects of divalent cations on the ATP hydrolysis and ATP synthesis activities of Hsp70 were analyzed. Both activities of Hsp70 are almost equally stimulated by 6 mM Mg 2ϩ , Mn 2ϩ , Co 2ϩ , and Ni 2ϩ (data not shown).

Formation of an Autophosphorylated Hsp70
Intermediate and Its CDP-dependent Dephosphorylation-We examined whether or not the intrinsic ADP-ATP exchange activity of Hsp70 is due to a contaminant, since its activity is similar to that of NDP kinase. This possibility is unlikely for the following reasons. Firstly, the commercially available Hsp70 was further purified by HPLC on a Mono Q anion-exchange and then a gel permeation column to remove possible contaminants with affinity for ATP, such as NDP kinase. The purified Hsp70 was Ͼ99% pure and was confirmed to be free from conventional NDP kinase. This was also confirmed by the fact that no immunoreactivity of the Hsp70 preparation against anti-NDP kinase-A antibodies was observed (Fig. 4A). Secondly, the nucleotide specificity for the activity of ATP synthesis by Hsp70 was different from that reported for NDP kinase. Thirdly, an autophosphorylated intermediate was observed in the protein band of Hsp70, as described below, as in the case of the autophosphorylated intermediate of conventional 16-kDa NDP kinase.
NDP kinase autophosphorylates the active site histidine of the intermediate in the process of the catalytic phosphate transfer reaction (21)(22)(23). Hsp70 gave a single autophosphorylated intermediate band corresponding to a protein with a molecular mass of 70 kDa in the absence of the nucleoside diphosphate, CDP, in the reaction mixture, which contained 10 Ci of [␥-32 P]ATP, 100 M ATP, and 6 mM MgCl 2 (Fig. 4B, lane  1), and no formation of CTP was observed (Fig. 4C, lane 1). In the presence of 5 mM CDP in the reaction mixture, however, the radioactivity of the phosphorylated intermediate decreased significantly with the concomitant formation of 32 P-labeled CTP from CDP (Fig. 4B, lane 2, and Fig. 4C, lane 2, respectively). Furthermore, a significant decrease in the radioactivity of the autophosphorylated intermediate was also observed on the addition of 5 mM CDP to the reaction mixture after the first reaction of the formation of an autophosphorylated intermediate (data not shown). These results indicate that Hsp70 catalyzes the transfer of the ␥-phosphate group from ATP to CDP, and that this transfer involves a phosphoenzyme intermediate. These properties are similar to those of the conventional NDP kinase reported. To characterize the phosphorylation of Hsp70, autophosphorylated Hsp70 was subjected to acid, neutral, and basic treatments as described (20,22), which allow the evaluation of a high energy phosphate on a histidine residue in NDP kinase and histidine protein kinase. The phosphorylated intermediate of Hsp70 was stable as to alkaline treatment, but labile as to acid treatment, with a significant decrease in the autophosphorylation level, as compared with neutral treatment, as shown in Fig. 4B, lanes 1, 3, and 4, respectively. That phosphorylation at serine, threonine, and tyrosine is stable as to acid treatment suggests that the acid-labile and alkali-stable phosphorylated residue(s) in Hsp70 may be a basic amino acid residue, although the phosphorylated residue has not yet been identified.

DISCUSSION
When we analyzed the ATPase activity of Hsp70, we observed a discrepancy in the ATPase activity determined with two different methods; the ATPase activity determined as the conversion of [ 14 C]ATP to [ 14 C]ADP was higher than that determined as the release of P i from ATP. In turn, we obtained Hsp70 which is free from contaminants with affinity to ATP, such as conventional NDP kinase, and measured the activities of ATP hydrolysis as well as ATP synthesis of Hsp70 and found that Hsp70 also exhibits the activity of ADP-ATP exchange.
A mechanism for Hsp70-protein/peptide interaction and ATP hydrolysis has been proposed (3,14). The first step in the mechanism involves the binding of a substrate protein/peptide to an Hsp70⅐ADP complex, resulting in a conformational change which causes acceleration of ADP-ATP exchange in the presence of ATP (step 2). The binding of ATP causes a conformational change which triggers substrate release from the complex (step 3). Finally, ATP is hydrolyzed to ADP to afford an Hsp70⅐ADP complex that can then participate in a new cycle of binding (step 4). The mechanisms underlying ADP-ATP exchange (step 2) and ATP hydrolysis (step 4) are best understood for the bacterial Hsp70 homolog, DnaK, with its cofactors, DnaJ and GrpE, which accelerate the ATP-dependent cycle of substrate binding and release (3,17,29). Interaction with DnaJ stimulates the hydrolysis of ATP by DnaK, resulting in the formation of a stable ternary complex of an unfolded substrate polypeptide/protein, DnaJ and DnaK. The ADP-ATP exchange factor, GrpE, promotes the release of ADP from DnaK, which is rate-limiting in the cycle, following dissociation of DnaJ and subsequent ATP binding and substrate dissociation from DnaK (13,30). Although the prokaryotic and eukaryotic Hsp70 systems have similar functional properties, a GrpE-like nucleotide exchange factor has not been found in the eukaryotic cytosol (13). On the basis of our present data, we propose that the intrinsic ADP-ATP exchange activity of Hsp70 accelerates the ATP-dependent reaction cycle in protein/peptide folding. To our knowledge, this is the first report of intrinsic ADP-ATP exchange activity of Hsp70. Determination of the precise role of the activity in the function of Hsp70 awaits further biochemical and genetic analysis.