View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

FEBS Letters 587 (2013) 666–672

journal homepage: www.FEBSLetters.org

PPIase independent -like function of recombinant human A during arginine kinase refolding ⇑ Xin-Chao Zhang a, Wei-Dong Wang a,b, Jin-Song Wang a, Ji-Cheng Pan a,

a College of Life Sciences, Hubei Normal University, Huangshi, Hubei 435002, PR China b State Key Laboratory of Virology, College of Life Sciences, Wuhan University, Wuhan, Hubei 430072, PR China

article info abstract

Article history: Whether Cyclophilin A (CyPA) functions as a foldase or a chaperone when assisting Received 17 December 2012 has long been argued. In this study, we engineered four variants of recombinant human Cyclophilin Revised 4 January 2013 A (rhCyPA), all of which were inactive to tetrapeptide substrate Suc-AAPF-pNA. However, these vari- Accepted 9 January 2013 ants were able to suppress aggregation during arginine kinase (AK) refolding as efficient as wild- Available online 31 January 2013 type rhCyPA, especially, variant Q63A had even more efficiency to suppress aggregation and improve Edited by Miguel De la Rosa reactivation yields of AK. These results indicate that rhCyPA have peptidyl–prolyl cis–trans isomer- ase (PPIase) independent chaperone-like activity during AK folding. In addition, results suggest that surface hydrophobicity of rhCyPA can suppress AK aggregation and binding to rhCyPA hydrophobic Keywords: Peptidyl–prolyl cis–trans active pocket is a prerequisite for chaperoning AK folding. Chaperone-like function Ó 2013 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. Arginine kinase Folding

1. Introduction was constructed and used to test the chaperone activity. Ou et al. observed that the inactive porcine kidney CypA, as well as an active The process of protein folding can be assisted by molecular PPIase, could improve the reactivation of denatured creatine kinase chaperones and foldases. The foldases include protein disulfide and prevent its aggregation during folding, which indicates that (PDIs) and cis–trans isomerases (PPIases); these en- CypA has a chaperone function [13]. Chakraborty et al. demon- zymes catalyse the rearrangement of disulfide bonds and the cis– strated that the N-terminal-truncated form (N22-88DEL) of trans isomerisation of peptidylprolyl bonds, respectively [1–3]. Leishmania donovani CyP18 can reactivate the activity of To date, the study of human CyPA mainly involves its role in viral the Leishmania donovani adenosine kinase through the disaggrega- infection and replication [4–6]. However, the chaperone function of tion of its oligomers [14] and prevent the aggregation of adenosine CyPA has attracted the interest of many investigators for several kinase both in vitro and vivo [15]. Recently, the controversy of decades. The single-domain CyPA has long been known to function whether CyPA has the ability to chaperone HCA II, which is the first as a PPIase and was proposed to function as a chaperone [7]. More substrate that is suggested to be chaperone-assisted by CypA, ap- recently, several multidomain PPIases have been found to act as pears to be definitely resolved [16]. However, to demonstrate that both PPIases and chaperones [8–10]. However, the chaperone func- this chaperone function is a general property of CyPA, it is very tion of CypA has long been argued because some researchers sug- important to show that CyPA has the ability to chaperone other pro- gest that the single-domain CyPA only acts as an isomerase teins as well. In this work, we show that another protein (AK) is during protein folding [11,12]. The focus of this controversial opin- chaperoned by rhCyPA. ion is whether CyPA play a PPIase or a chaperone role during pro- In previous research studies, we found that two equilibrium tein folding. To eliminate the interference of its PPIase activity intermediates accompanied the cis–trans prolyl isomerisation that during protein folding, a CyPA enzyme without PPIase activity occurs during the folding of AK, which was obtained from the mus- cle of shrimp (Fenneropenaeus chinensis) [17]. We have also inves- Abbreviations: ANS, 8-anilino-1-naphthalenesulfonate; CyPA and rhCyPA, cyclo- tigated the reactivation of denatured AK in the presence of small philin A and recombinant human cyclophilin A; CsA, cyclosporin A; PPIase, artificial molecular chaperones, such as Triton X-100 and Tween peptidyl–prolyl cis–trans isomerase; SDS–PAGE, sodium dodecyl sulfate–polyacryl- 20 [18]. Moreover, our investigation of the effect of rhCyPA during mide gel electrophoresis; Suc-AAPF-pNA, succinyl-Ala-Ala-Pro-Phe-p-nitroanilide; AK folding revealed that rhCyPA could form dimers and trimers TFE, trifluoroethanol ⇑ Corresponding author. Fax: +86 714 6513163. during the purification procedure [19]. To provide additional E-mail address: [email protected] (J.-C. Pan). evidence that demonstrate whether the single-domain CyPA acts

0014-5793/$36.00 Ó 2013 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.febslet.2013.01.028 X.-C. Zhang et al. / FEBS Letters 587 (2013) 666–672 667 as a chaperone during the folding of other proteins, we used AK (a in 1 mM HCl to a concentration of 10 mg/ml, and the Suc-AAPF- model protein that contains 13 residues; the folding of this pNA substrate was prepared and stocked in TFE with 470 mM LiCl protein involves prolyl cis–trans isomerism) as the substrate to at a concentration of 3.75 mM. For the PPIase inhibition assays, investigate the role of active and inactive PPIases, which were 1 mM of cyclosporin A (CsA) was dissolved in ethanol and engineered through the site-directed mutagenesis of the rhCyPA incubated with PPIase for at least 2 min prior to the initiation active sites, during the folding of AK. The results show that rhCyPA of the assay. The activity assay mixture consisted of 430 llof has a chaperone-like function that is independent of its PPIase 50 mM HEPES buffer (containing 100 mM NaCl, pH 8.0), 50 llof function. In addition, we observed that the surface hydrophobicity the a- stock solution and 10 ll PPIase (to a final of rhCyPA and its variants can suppress the aggregation of AK and concentrations of less than 10 nM; or PPIase-CsA at a PPIase- that the binding of AK to the hydrophobic of rhCyPA and to-CsA ratio of approximately 1:2). After a 5-min pre-incubation its variants is a prerequisite for the chaperoning of AK folding. of the mixture at 10 °C, the activity assay was initiated through the addition of 10 ll of the Suc-AAPF-pNA solution. The absor- 2. Materials and methods bance was recorded at 390 nm by an Ultrospec 4300 spectrometer every 2 s (the last time point tested was 4 min). 2.1. Materials Since the assay concentration of the substrate Suc-AAPF-pNA was much lower than Michaelis constant Km [21], the Michaelis– Suc-AAPF-pNA and trifluoroethanol (TFE) were purchased from Menten velocity equation can be simplified to V = Vmax [S]/Km = kcat Sigma. DEAE-Sepharose Fast Flow and Sephacryl S-100 were ob- [E][S]/Km = kobs[E], where kobs = kcat/Km[S]. Because [E] and [S] are tained from GE. All other chemicals were of analytical grade. known, the kobs value can be transformed to kcat/Km value, which The AK gene from shrimp (Fenneropenaeus chinensis) muscle can be used to characterize the catalytic activity. The kcat/Km values was cloned into the PET28a plasmid with a His tag and expressed of the rhCyPA variants were obtained using the same procedure. in Escherichia coli Rossta. The proteins were purified by Ni affinity chromatography. The concentrations of AK were determined 2.5. Folding experiments according to the method described by Virden et al. [20]. The AK protein (at a concentration of 150 lM) was denatured in 2.2. Preparation of wild-type and variants 3.0 M GuHCl for 1 h at room temperature. The refolding reaction of AK was then initiated by a 50-fold dilution (to a final concentration The wild-type rhCyPA, as well as the Q63A, H126A, F60A and of 3 lM) of the AK solution in 20 mM Tris–Cl (containing 10 mM R55A mutants, were engineered through inverse PCR and cloned b-mercaptoethanol, pH 7.8) at 20 °C in the absence or presence into the PQE-60 plasmid, which was subsequently transformed into of PPIase (final concentration of 6 lM) and the addition of CsA to the E. coli M15 bacterial strain. The bacterial colonies that were a concentration of 12 lM (when applicable). The optimal ratio of resistant to both kanamycin and ampicillin were sequenced to en- denatured AK and wild-type rhCyPA was 1:2. Under this condition, sure the fidelity of the . The cells were harvested by cen- the highest amount of aggregation suppression could be attained; trifugation at 5000 rpm and lysed by ultrasonication at 4 °Cina thus, this ratio was maintained in all of the folding experiments. buffer containing 50 mM Tris–Cl, 1 mM EDTA, 100 mM NaCl, The AK aggregation was followed by monitoring the turbidity at 1 mM phenylmethylsulfonyl fluoride (PMSF) pH 8.0 and 10 mM 400 nm using an Ultrospec 4300 spectrometer every 2 s for b-mercaptoethanol. After the cell debris was discarded by centrifu- 3 min. Immediately after the dilution, aliquots (20 ll) were col- gation (12000 rpm, 20 min, 4 °C), 40 ml of the supernatant was lected at suitable time intervals to determine the AK activity in loaded onto the Sephacryl S-100 column (2.6 100 cm). The mate- the presence and absence of PPIase, according to the protocol de- rial was then eluted with a buffer containing 20 mM Tris–Cl pH 7.8, scribed by Yu et al. [22]. To calculate the relative refolding yields, 1 mM EDTA, 1 mM PMSF and 10 mM b-mercaptoethanol. The frac- the activity of the native AK, which was treated in the same man- tions that contained rhCyPA were pooled and loaded onto a DEAE- ner as the refolding sample but without denaturation, was as- Sepharose column. The elution buffer was the same as that used for sumed to be 100%. All of the folding experiments were repeated the Sephacryl S-100 column, with the exception that it did not con- at least three times. The kinetic data were best fitted by a two- tain EDTA and PMSF. The flow-through fractions were analysed by exponential equation: SDS–PAGE, and the homogenous rhCyPA proteins were combined y ¼ A1ð1 expðk1tÞÞ þ A2ð1 expðk2tÞÞ and concentrated in a dialysis bag by polyethylene glycol (PEG). where y is the percent refolding yield, t denotes the time in min, A1 2.3. Fluorescence spectroscopy measurements and A2 denote the amplitudes of the first and second phases, respec- tively, k1 and k2 are the rate constants of the first and second The wild-type CyPA and the CyPA variants all had a concentra- phases, respectively. During the process, a dead time of 7 s was tion of 150 lM. For the spectroscopy experiments, the PPIase was set from the onset of the folding to the recording of the first data; diluted 10-fold. The fluorescence emission spectrum was mea- thus, any phase that is shorter than the dead time cannot be de- sured using a Hitachi F-4500 spectrofluorometer at 20 °C. tected by these UV measurements. The ANS fluorescence measurements were performed using slit widths for the excitation and emission of 5 and 10 nm, respec- 3. Results tively. The excitation was measured at a wavelength of 380 nm, the emission was recorded between 400 and 600 nm. 3.1. Variant Q63A has more exposed hydrophobic surface The intrinsic fluorescence measurements were performed using a slit width of 5 nm. After excitation at 295 nm, the emission spec- Fluorescence spectroscopy is often used to characterise the trum between 300 and 400 nm was monitored. changes that occur in macromolecular structures, such as proteins or nucleic acids [23,24]. Because ANS can bind with the hydropho- 2.4. Prolyl isomerase activity assay bic surface of proteins, which markedly enhances the intensity of the fluorescence [25], it can often be used to detect the solvent- An improved spectrophotometric assay [21] was used in this exposed states of proteins and other biomacromolecules with a experiment. The stock solution of a-chymotrypsin was dissolved hydrophobic surface. As shown in Fig. 1A, a higher amount of 668 X.-C. Zhang et al. / FEBS Letters 587 (2013) 666–672

ANS binds to the Q63A variant, which indicates that a larger hydro- phobic area is exposed in this variant than in the wild-type or the other variants. In addition, the intensity of the intrinsic fluores- cence of the Q63A variant is lower than that of the other variants and the wild-type (Fig. 1B), which suggests that the Trp in this var- iant is more efficiently quenched by the solvent. In agreement with the ANS data, our findings indicate that the Q63A caused the formation of a looser structure compared with the other vari- ants/wild-type.

3.2. Prolyl isomerase activity of the mutants retained about less than 0.2% of the wild-type rhCyPA

The argument of whether the single-domain CyPA has PPIase activity or chaperone activity during its assistance of the protein folding process began with the first report that demonstrated that CyPA functioned as a chaperone during the refolding of human car- bonic anhydrase II. To eliminate the effect of PPIase activity, four Fig. 2. Chymotrypsin-coupled PPIase activity assay with various concentrations of variants were constructed. In this work, we selected to experimen- wild-type rhCyPA. 1–6 represents 0, 0.25, 0.75, 1.50, 3.50, 7.5 nM wild-type, tally test three active site residues (R55, F60 and H126), which respectively. were shown to play an active role in the active site by Zydowsky et al. [26], and an additional active site residue (Q63), which was simulated through hybrid MM/QM by Li and Cui [27]. The results of the chymotrypsin-coupled PPIase activity assay for various amounts of the wild-type rhCyPA are shown in Fig. 2 and kcat/Km was calculated from the kobs-values from Fig. 3. The kcat/Km values of the variants were obtained following the same procedure.

The kcat/Km values of the wild-type and the variants are shown in Table 1. The wild-type rhCyPA has a kcat/Km value of 1.38 107 M1 s1 at 10 °C, which is close to the value reported in previous studies [21,26,28]. The kcat/Km values of the variants are all less than 0.2% of the kcat/Km value of the wild-type, which is in agreement with the data published by Zydowsky et al. [26]. Therefore, the four mutated residues, including Q63, are active-site residues, and the variants are completely inactive.

3.3. Variant Q63A can suppress aggregation and improve reactivation yield more efficiently than wild-type rhCyPA during AK folding

Fig. 3. Values of the kobs rate constants for different concentrations of the wild-type To compare the abilities of the wild-type and the variants to rhCyPA. prevent aggregation during the protein folding process, we used denatured AK, which contains 13 [29], as the substrate protein. Interestingly, as shown in Fig. 4, all of the turbidity curves when rhCyPA or its variants are added. Since rhCypA or its variants obtained in the presence or absence of rhCyPA (or its mutants) are do not significantly contribute to the absorbance at 400 nm, the very similar. However, the start and end points are apparently low- background turbidity that is induced by the PPIase protein can er than those obtained during the spontaneous folding that occurs be considered negligible. All of this evidence imply that the

Fig. 1. ANS fluorescence emission and intrinsic fluorescence emission spectra of variants and wild-type rhCyPA. (A) ANS fluorescence emission spectra, 1–5 represents wild- type, R55A, F60A, H126A and Q63A, respectively. (B) Intrinsic fluorescence emission spectra, 1–5 represents Q63A, wild-type, H126A, F60A, and R55A, respectively. X.-C. Zhang et al. / FEBS Letters 587 (2013) 666–672 669

Table 1 concentrations of the wild-type and the Q63A variant were 6 lM. Kinetic constants of the WT and its variants. The PPIase was incubated with CsA at a ratio of 1:2 for 12 h before 4 1 1 CyP kcat/Km (10 )s M it was added to the refolding system. To calculate the relative Wild-type 1380.00 refolding yields, the activity of the native AK, which was treated R55A 1.17 in the same manner as the refolding sample but without denatur- F60A 1.96 ation, was assumed to be 100%. In addition, the addition of the PPI- Q63A 1.10 ase inhibitor CsA to the folding system had negligible influence on H126A 2.27 the aggregation suppression (Fig. 5). Interestingly, as shown in Fig. 5, the kinetic curves of the aggregation suppression induced by the wild-type and the Q63A variant were almost identical to wild-type (and the variants) bind to the early-folding-intermedi- the respective curves obtained in the presence of CsA. In addition, ates of AK during the dead time and that this binding significantly because CsA is a competitive inhibitor [26] for CyPA, it must bind weakens the aggregation that occurs during the initial folding to the active pocket, which may block the access of the AK interme- steps. diates to the hydrophobic pocket of CyPA. These data altogether The folding kinetics (turbidity curves) of AK in the presence of indicate that the binding of the substrate to the active hydrophobic any of the inactive variants closely follow the kinetics observed pocket is not a prerequisite for the suppression of aggregation. in the presence of the active wild-type. However, compared with The reactivation kinetics of AK showed that the AK recovery the wild-type, the Q63A variant more efficiently suppresses AK activities of the Q63A variant and the wild-type rhCyPA were on aggregation (Fig. 4) and improves the AK reactivation yield average 14% and 10% higher, respectively, compared with the (Fig. 6). Coincidently, the global structure of the Q63A variant kinetics obtained in the absence of PPIase. However, once the ac- makes this variant more solvent-exposed than the other three vari- tive pocket was occupied by CsA, both the wild-type and the ants and the wild-type rhCyPA. It is likely that the increased Q63A variant lost their chaperone function during the folding of amount of exposed hydrophobic surface offers more accessibility AK. These results indicate that the reactivation of the AK folding for interaction with the folding intermediates of the AK folding intermediates must occur in the pocket, which is consistent with process; these results are in agreement with those reported by the data obtained in a previous study [30]. Chakraborty et al. [14]. 4. Discussion 3.4. Surface hydrophobicity of rhCyPA prevents the aggregation and hydrophobic pocket may contribute to the reactivation of AK Chaperones were originally described by their ability to prevent the aggregation of protein substrates by binding to the non-native Cyclosporin A (CsA), which is a high-affinity inhibitor that binds early folding intermediates and helping these molecules reach a to CyPA, was used to block the PPIase activity of CyPA. The functional conformation [31]. In fact, the prevention or suppression concentration of AK during the refolding was 3 lM, and the of aggregation and the enhancement of activity recovery during the

Fig. 4. Suppression of AK aggregation by wild-type rhCyPA and its variants during refolding. The AK protein, at an original concentration of 150 lM, was denatured in 3.0 M GuHCl for 1 h at room temperature. The refolding was initiated by a 50-fold dilution in 20 mM Tris–Cl buffer (pH 7.8, containing 10 mM b-mercaptoethanol) in the presence or absence of 6 lM of the wild-type rhCyPA or its variants at 20 °C. The squares, circles, and triangles represent the aggregation during AK refolding in the presence of 0 lM PPIase, 6 lM wild-type and 6 lM of the variants (A, R55A; B, F60A; C, Q63A; D, H126A), respectively. 670 X.-C. Zhang et al. / FEBS Letters 587 (2013) 666–672

site-directed mutagenesis, and the corresponding proteins were ex- pressed and purified to homogeneity by SDS–PAGE. The protease- coupled assay for the analysis of prolyl isomerase activity showed that the variants are almost inactive compared to the wild-type. The structures of the variants were characterised by ANS fluo- rescence spectrometry and intrinsic fluorescence spectrometry. The ANS fluorescence spectra and the intrinsic fluorescence spectra all indicate that the Q63A variant is more solvent-exposed than the other variants and the wild-type (Fig. 1). By comparing the ability of the variants and the wild-type to suppress aggregation during AK refolding, we found that the aggregation-suppressing ability of the variants is similar to that of the wild-type (Fig. 4). Interest- ingly, the Q63A variant was more efficient than the wild-type in the suppression of AK aggregation; since the Q63A variant has the loosest spatial structure, the higher efficiency of this variant in the suppression of protein aggregation may be due to the expo- sure of a higher amount of hydrophobic residues. Similar mecha- nisms have been proposed to explain why a truncated Leishmania Fig. 5. Suppression of AK aggregation by PPIase and CsA. During refolding, the final donovani CyPA is more efficient than the full-length CyPA in the concentrations of AK and PPIase were 3 and 6 lM, respectively. CsA was incubated with PPIase overnight to ensure adequate binding prior to the initiation of the reactivation of adenosine kinase [14]. folding process. The following conditions were assayed: CsA (right triangle), The results (Fig. 4) show that the slopes of the turbidity curves without wild-type rhCyPA or the variants (square), wild-type rhCyPA (circle), wild- obtained in the presence or absence of rhCypA (or its mutants) type rhCyPA + CsA (down triangle), Q63A (up triangle), Q63A + CsA (left triangle). have apparent absorption at 400 nm within the 7-s dead time; these results indicate that the folding intermediates emerged and rapidly formed considerable aggregates (Fig. 7). The curves re- corded after the dead time are equally steep, regardless of the pres- ence of rhCyPA. However, the aggregation is markedly suppressed within the dead time. All of these results imply that rhCyPA and its variants can only exert a chaperone effect by binding to the early intermediates of the AK folding process but cannot dissociate the formed aggregates (Fig. 7). This result is in agreement with the ac- tion of the chaperone GroE, which is only active during the early steps of protein folding and cannot rescue the insoluble aggregated proteins [32]. In addition, rhCyPA and its variants can block protein aggrega- tion during AK folding even in the presence of its inhibitor CsA (Fig. 5). However, once the active pocket binds to CsA, rhCyPA and its variants completely lose their ability to improve AK reacti- vation (Fig. 6). Although the molecular determinants of CyPA for CsA binding may not be identical to the molecular determinants for protein substrates [33], the CsA binding may induce a stereo- specific blockade that blocks the access of the AK intermediates Fig. 6. Time course of the reactivation of unfolded AK in the presence and absence to the hydrophobic pocket of rhCyPA. This evidence indicate that of the wild-type rhCyPA or its variants. The data were analysed according to the protocol described in Materials and methods (Section 2.5). To calculate the relative the hydrophobicity of rhCyPA and its variants may exert influence refolding yields, the activity of the native AK, which was treated in the same on the folding yields in two ways: the surface hydrophobicity pre- manner as the refolding sample but without denaturation, was assumed to be 100%. vents the aggregation that is caused by the hydrophobic interac- The following conditions were assayed: Q63A (circle), wild-type rhCyPA (up tion of AK proteins (Fig. 7), which is consistent with the data triangle), Q63A + CsA (down triangle), wild-type rhCyPA + CsA (left triangle), without wild-type rhCyPA or the variants (square), CsA (right triangle). During published in other reports [14,34], and the binding of the substrate refolding, the final concentrations of AK and CyPA were 3 and 6 lM, respectively. to the hydrophobic active pocket is a prerequisite of native folding (Fig. 7). Moreover, the Q63A variant improved the AK reactivation refolding of urea- or guanidine hydrochloride-denatured proteins yields more effectively than the wild-type CyPA. was regarded as the most important step of the chaperone- The folding kinetic curves showed that the aggregation reached facilitated protein folding process. Several chaperones were thus an equilibrium quickly (within 3 min) (Fig. 4) and that the activity reported, among which CyPA was considered to be one of the recovery was significantly delayed (Fig. 6). These results indicate candidates [7,13]. However, the original report on the chaperone that the AK refolding process involves the rapid production of an function of CyPA [7] was doubted because some researchers early folding intermediate (like-native) and a significantly slow believed that only the enzymatic activity of the PPIase played an rearrangement course (native) (Fig. 7). This result is in agreement important role during HCA II refolding [11]. A later report by Ou with the data published by Freskgard et al. [7]. et al. showed the efficiency of inactive CyPA in the chaperoning of Recently, Moparthi et al. [30] and Dorothy Kern’s group [35] re- the creatine kinase folding process. However, oxidization or other ported that rhCypA with active site mutations show similar kinet- indefinite factors that lead to the loss of PPIase activity during the ics compared to wild-type when catalyzing the folding of protein preservation of CyPA likely induces a stereospecific blockade that substrates. They suggested that the active sites mutants of rhCyPA, may affect the interaction of CyPA and the protein substrate. To although having no PPIase activity for the tetrapeptide substrate resolve these problems and provide more useful evidence that Suc-AAPF-pNA, still played a role in catalyzing proline isomeriza- support or disprove the chaperone activity of CyPA, four variants tion during protein folding. Because molecular chaperones typi- (Arg55A, F60A, Q63A and H126A) were constructed through cally increase the yield but not the rate of folding reactions [32], X.-C. Zhang et al. / FEBS Letters 587 (2013) 666–672 671

Fig. 7. Schematic representation of the AK folding process chaperoned by rhCyPA. A, aggregated state; D, denatured state; I1 and I2, folding intermediates; N, native state; P, proline. Through binding to the aggregation-prone intermediate I1, the rhCyPA surface hydrophobicity lowers the concentration of this intermediate and minimises the formation of the aggregates (A). After the slow rearrangement of the protein inside the hydrophobic pocket of rhCyPA is completed, the intermediate like-native (I2) is folded into its native state (N). the improved AK reactivation yields were mainly due to the rhCy- [9] Pirkl, F., Fischer, E., Modrow, S. and Buchner, J. (2001) Localization of the PA chaperone-like function but not its isomerisation activity. chaperone domain of FKBP52. J. Biol. Chem. 276, 37034–37041. [10] Mok, D., Allan, R.K., Carrello, A., Wangoo, K., Walkinshaw, M.D. and Ratajczak, To conclude, the rhCyPA has a chaperone-like function, which is T. (2006) The chaperone function of cyclophilin 40 maps to a cleft between the independent of its PPIase activity for the tetrapeptide substrate prolyl isomerase and tetratricopeptide repeat domains. FEBS Lett. 580, 2761– substrate Suc-AAPF-pNA, and the power of this function is associ- 2768. [11] Kern, G., Kern, D., Schmid, F.X. and Fischer, G. (1994) Reassessment of the ated with its surface hydrophobicity and the accessibility of its ac- putative chaperone function of prolyl-cis–trans-isomerases. FEBS Lett. 348, tive pocket. In addition, we showed that the folding process of AK, 145–148. which is a model protein that contains 13 proline residues and pro- [12] Fischer, G. and Schmid, F.X. (1999) In Molecular Chaperones and Folding Catalysts, Harwood Academic Publishers, Amsterdam. lyl cis–trans isomerism in the folding process, is chaperoned by [13] Ou, W.B., Luo, W., Park, Y.D. and Zhou, H.M. (2001) Chaperone-like activity of rhCyPA. This finding is essential for understanding the general peptidyl–prolyl cis–trans isomerase during creatine kinase refolding. Protein roles that CyPA plays in the cell. Sci. 10, 2346–2353. [14] Chakraborty, A., Das, I., Datta, R., Sen, B., Bhattacharyya, D., Mandal, C. and Datta, A.K. (2002) A single-domain cyclophilin from Leishmania donovani Acknowledgements reactivates soluble aggregates of adenosine kinase by isomerase-independent chaperone function. J. Biol. Chem. 277, 47451–47460. The present investigation was supported by Grant 30570412, [15] Chakraborty, A., Sen, B., Datta, R. and Datta, A.K. (2004) Isomerase- independent chaperone function of cyclophilin ensures aggregation 2008CDA066 from National Natural Science Foundation of China, prevention of adenosine kinase both in vitro and under in vivo conditions. Hubei Province’s Natural Science Foundation of China and Educa- Biochemistry 43, 11862–11872. tional Commission of Hubei Province of China (Q20122206). [16] Moparthi, S.B., Fristedt, R., Mishra, R., Almstedt, K., Karlsson, M., Hammarstrom, P. and Carlsson, U. (2010) Chaperone activity of Cyp18 through hydrophobic condensation that enables rescue of transient References misfolded molten globule intermediates. Biochemistry 49, 1137–1145. [17] Pan, J.C., Yu, Z., Su, X.Y., Sun, Y.Q., Rao, X.M. and Zhou, H.M. (2004) Unassisted [1] Gilbert, H.F. (1997) Protein disulfide isomerase and assisted protein folding. J. refolding of urea-denatured arginine kinase from shrimp Feneropenaeus Biol. Chem. 272, 29399–29402. chinensis: evidence for two equilibrium intermediates in the refolding [2] Ruddon, R.W. and Bedows, E. (1997) Assisted protein folding. J. Biol. Chem. pathway. Protein Sci. 13, 1892–1901. 272, 3125–3128. [18] Pan, J.C., Wang, J.S., Cheng, Y., Yu, Z., Rao, X.M. and Zhou, H.M. (2005) The role [3] Schmid, F.X. (1993) Prolyl isomerase: enzymatic catalysis of slow protein- of detergent in refolding of GdnHCl-denatured arginine kinase from shrimp folding reactions. Annu .Rev. Biophys. Biomol. Struct. 22, 123–142. Fenneropenaeus Chinensis: the solubilization of aggregate and refolding in [4] Ambrose, Z. et al. (2012) Human immunodeficiency virus type 1 capsid detergent solutions. Biochem. Cell Biol. 83, 140–146. mutation N74D alters cyclophilin a dependence and impairs macrophage [19] Zhang, X.C., Wang, W.D., Wang, J.S., Pan, J.C. and Zou, G.L. (2011) Evidences of infection. J. Virol. 86, 4708–4714. monomer, dimer and trimer of recombinant human Cyclophilin A. Protein [5] Hopkins, S., Bobardt, M., Chatterji, U., Garcia-Rivera, J.A., Lim, P. and Gallay, P.A. Pept. Lett. 18, 1188–1193. 0 (2012) The cyclophilin inhibitor SCY-635 disrupts hepatitis C virus NS5A- [20] Virden, R., Watts, D.C. and Baldwin, E. (1965) Adenosine 5 -triphosphate- cyclophilin A complexes. Antimicrob. Agents Chemother. 56, 3888–3897. arginine phosphotransferase from lobster muscle: purification and properties. [6] von Hahn, T. et al. (2012) Hepatocytes that express variants of cyclophilin a are Biochem. J. 94, 536–544. resistant to HCV infection and replication. Gastroenterology 143 (439–447), [21] Kofron, J.L., Kuzmic, P., Kishore, V., Colon-Bonilla, E. and Rich, D.H. (1991) e1. Determination of kinetic constants for peptidyl prolyl cis–trans isomerases by [7] Freskgard, P.O., Bergenhem, N., Jonsson, B.H., Svensson, M. and Carlsson, U. an improved spectrophotometric assay. Biochemistry 30, 6127–6134. (1992) Isomerase and chaperone activity of prolyl isomerase in the folding of [22] Yu, Z., Pan, J. and Zhou, H.M. (2002) A direct continuous pH- carbonic anhydrase. Science 258, 466–468. spectrophotometric assay for arginine kinase activity. Protein Pept. Lett. 9, [8] Mayr, C., Richter, K., Lilie, H. and Buchner, J. (2000) Cpr6 and Cpr7, two closely 545–552. related Hsp90-associated from Saccharomyces cerevisiae, differ [23] Steiner, R.F. and Winryb, I. (1971) Excited states of Proteins and Nucleic Acids, in their functional properties. J. Biol. Chem. 275, 34140–34146. Plenum Press, New York. 672 X.-C. Zhang et al. / FEBS Letters 587 (2013) 666–672

[24] Lakowics, J.R. (1986) Principles of Fluorescence Spectroscopy, Plenurn Press, 1.2 A resolution: mechanistic insights. Acta Crystallogr. D Biol. Crystallogr. 58, New York. 2009–2017. [25] Stryer, L. (1965) The interaction of a naphthalene dye with apomyoglobin and [30] Moparthi, S.B., Hammarstrom, P. and Carlsson, U. (2009) A non-essential role apohemoglobin. A fluorescent probe of non-polar binding sites. J. Mol. Biol. 13, for Arg 55 in cyclophilin18 for catalysis of proline isomerization during 482–495. protein folding. Protein Sci. 18, 475–479. [26] Zydowsky, L.D., Etzkorn, F.A., Chang, H.Y., Ferguson, S.B., Stolz, L.A., Ho, S.I. and [31] Buchner, J. (1996) Supervising the fold: functional principles of molecular Walsh, C.T. (1992) Active site mutants of human Cyclophilin A separate chaperones. FASEB J. 10, 10–19. peptidyl–prolyl isomerase activity from cyclosporin A binding and calcineurin [32] Hartl, F.U. (1996) Molecular chaperones in cellular protein folding. Nature inhibition. Protein Sci. 1, 1092–1099. 381, 571–579. [27] Li, G. and Cui, Q. (2003) What is so special about Arg 55 in the catalysis of [33] Davis, T.L. et al. (2010) Structural and biochemical characterization of the cyclophilin A? Insights from hybrid QM/MM simulations. J. Am. Chem. Soc. human cyclophilin family of peptidyl–prolyl isomerases. PLoS Biol. 8, 125, 15028–15038. e1000439. [28] Liu, J., Albers, M.W., Chen, C.M., Schreiber, S.L. and Walsh, C.T. (1990) Cloning, [34] Melkani, G.C., Zardeneta, G. and Mendoza, J.A. (2004) Oxidized GroEL can expression, and purification of human cyclophilin in Escherichia coli and function as a chaperonin. Front. Biosci. 9, 724–731. assessment of the catalytic role of cysteines by site-directed mutagenesis. [35] Bosco, D.A., Eisenmesser, E.Z., Clarkson, M.W., Wolf-Watz, M., Labeikovsky, W., Proc. Natl. Acad. Sci. USA 87, 2304–2308. Millet, O. and Kern, D. (2010) Dissecting the microscopic steps of the [29] Yousef, M.S., Fabiola, F., Gattis, J.L., Somasundaram, T. and Chapman, M.S. cyclophilin a enzymatic cycle on the biological HIV-1 capsid substrate by (2002) Refinement of the arginine kinase transition-state analogue complex at NMR. J. Mol. Biol. 403, 723–738.