Introduction

The Arabian camel (Camelus dromedaries) is one of few mammals which can survive under conditions of intense heat and dryness. Surrounding temperatures have large effects on the biochemistry and physiology of all organisms at the molecular level. The native structure of proteins is stabilized by different types of relatively weak non-covalent interaction, for example hydrogen bonding, and hydrophobic, ionic, and dipole–dipole interactions (Dill 1990). At high temperatures, however, the vibration of atoms in proteins increases and breaks the weak non-covalent bonds (Vogt et al. 1997), resulting in denaturation and loss of protein function. Proteins are synthesized in a highly crowded cytosolic environment (>200 mg/ml proteins); at such a high level of molecular crowding, incorrect interactions may lead to misfolding and aggregation and/or proteolysis of newly synthesized proteins (Brown 1991; Tavaria et al. 1996).

Heat shock proteins are of crucial importance in the folding and stability of almost all proteins (Liu et al. 2010) and are usually expressed under extreme environmental conditions, for example biotic and abiotic stresses, including high temperatures (Santoro 2000). It is fair to assume that the Arabian camel that lives under extreme hot and dry climatic conditions must face extra challenges with regard to protein folding and functioning.

The key function of heat shock proteins in living cells and organisms is to provide stability to protein homeostasis. The Hsp70 class of chaperones enables an organism to adapt to a variety of environmental, developmental, and physiological stresses (Mayer and Bukau 2005). In E. coli, it has been observed that deletion of Hsp70 chaperone results in aggregation of more than 300 different proteins (Deuerling et al. 2003). In eukaryotes, the Hsp70 chaperones make up 1–2 % of total protein (Zylicz et al. 2001) and assist in the folding of thousands of cellular proteins (Mayer and Bukau 2005). Hsp70 chaperones actively participate in several other cellular processes, including transportation of proteins across membranes (Borges and Ramos 2005; Walter and Buchner 2002), cardiovascular diseases (Benjamin and McMillan 1998), immunity (Nishikawa et al. 2008), and apoptosis (Beere et al. 2000). Hsp70 chaperones are a fertile area for biomedical research and have been targeted for development of therapeutic proteins (Zylicz et al. 2001).

HSPA6, also known as HSP70B’, is a member of Hsp70 family of molecular chaperones. Members of Hsp70 are crucially involved in maintaining cellular proteostasis (Heldens et al. 2010). Under unstressed conditions, expression of HSPA6 cannot be detected (Leung et al. 1990; Parsian et al. 2000). Under stressed conditions, however, its level increases substantially, and it has been used as biomarker in several cellular stress studies (Wada et al. 2005; Wang et al. 2005). Structurally, Hsp70 chaperones are highly conserved, monomeric, two-domain proteins. The larger N-terminal domain (~45 kDa) is known as the nucleotide binding domain (NBD) whereas the smaller C-terminal domain (25 kDa) is the substrate-binding domain (SBD); the domains are connected by a short, flexible, hydrophobic linker (Flaherty et al. 1990). Chaperones of Hsp70 members have been studied extensively but their mechanism of action is not completely understood (Bukau and Horwich 1998; Morano 2007). It has been suggested that the Hsp70 chaperones temporarily bind to hydrophobic residues of the proteins under stress, prevent partially-denatured proteins from aggregation, and assist them in refolding. Results obtained so far indicate that NBD and SBD interact in an allosteric regulatory fashion. Crucial residues involved in the allosteric regulation have been identified (Rist et al. 2006; Vogel et al. 2006). Binding of ATP to the NBD can induce conformational change in the SBD which leads to release of the substrate. In contrast, binding of the substrate to the SBD enhances the ATPase activity of the NBD (Buchberger et al. 1995; Moro et al. 2003; Revington et al. 2005; Swain et al. 2007).

In this study we report overexpression of cHSPA6 in E. coli, and its subsequent purification. Studies of nucleotide-binding by use of fluorescence-quenching methods revealed high affinity (K D ~50 nM). Far-UV CD measurements in the nucleotide-bound state revealed small changes of secondary structure. Structural and thermodynamic properties were characterized by use of dynamic multimode spectroscopy. Similar to DnaK of E. coli (EcDnaK) and human Hsc70, cHSPA6 also unfolds via three thermal transitions. Interestingly, in the nucleotide-bound state, the conformational state of cHSPA6 is more similar to that of human Hsc70 than to that of EcDnaK.

Materials and methods

Materials

The vector pET15-cHSPA6 encoding Arabian camel HSPA6, characterization of which has been reported elsewhere (Elrobh et al. 2011), was kindly provided by Dr Elrobh (Department of Biochemistry, King Saud University, KSA). E. coli BL21 (DE3) pLysS was obtained from Life Technologies. Benzonase was purchased from Sigma and chicken egg lysozyme was purchased from USB Corporation. IPTG and ampicillin were obtained from Biobasic. Ni–NTA resin, Superdex 75 prepacked columns, and low molecular weight markers were purchased from Amersham Biosciences. Other chemicals were reagent grade.

cHSPA6 expression and purification

cHSPA6 fused with a hexa-histidine tag at the N-terminus was cloned on pET15 under strong T7 promoter. For protein expression, pET15-cHSPA6 plasmid was transformed into E. coli BL21 (DE3) pLysS. To prepare biomass, 1 % overnight culture was transferred to 1 l LBamp (200 μg/ml) in 6-l conical flasks. cHSPA6 expression was induced by adding 25 μM IPTG when growth reached the mid-exponential phase (OD600 = 0.6). After post-induction incubation for 5 h, cells were harvested at 5,000 rpm for 15 min. The biomass (2.5 g) was resuspended in 15 ml lysis buffer (50 mM Tris, 200 μM PMSF, 3 mM DTT, 2 mM MgCl2, and 300 μg/ml lysozyme, pH 7.5). The biomass was homogenized in a mechanical homogenizer to a uniform slurry then 1 μl benzonase (250 U/μl) was added to the slurry and the mixture was incubated at room temperature for 1 h. The slurry was then subjected to mild sonication (Soniprep 150; MSE UK), twice, each time for 10 s, at 5 μm amplitude and 4 °C. Next, the slurry was centrifuged for 30 min at 13,000 rpm at 4 °C. Supernatant was collected and filtered through a 0.45 μm filter. Sodium chloride was added to the filtrate to achieve a final concentration of 500 mM. Further imidazole was added to the solution to bring the final concentration to 10 mM. The supernatant was passed through a 1-ml HisTrap column which had previously been equilibrated with 20 ml equilibration buffer (50 mM Tris, 10 mM imidazole, and 500 mM NaCl, pH 7.5). The column was connected to a fast protein liquid chromatography (FPLC) system. After washing with equilibration buffer until the absorbance at 280 nm reached baseline level, the His-tagged cHSPA6 was eluted with a gradient of 0–50 % with elution buffer (50 mM Tris, 500 mM, imidazole, and 500 mM NaCl, pH 7.5) at 0.5 ml/min. The purity of cHSPA6 in different fractions was checked on 12 % SDS gel. The fractions containing cHSPA6 were combined and loaded on to a Superdex 75 26/60 column pre-equilibrated with 25 mM Tris and 250 mM NaCl, pH 7.5. The purity of the fractions was analyzed on 12 % SDS gel. The buffer of the purified cHSPA6 was exchanged in 4 cycles with 20 mM sodium phosphate buffer containing 50 mM NaCl at pH 7.4, by use of Centricon (MWCO 10,000 Da). The permeate from the fourth cycle was used as reference control in spectroscopic studies.

Sequence alignment and prediction of the 3D structure of cHSPA6

cHSPA6 (accession number HQ214118.1), human HSP70B’ (BAD97082.1), and bovine HSPA6 (XP_589747.1) were used for multiple sequence alignment by ClustalW on MAFFT multiple sequence alignment program. Output is reported in accordance with the conserved sequence. The amino acid sequence of cHSPA6 (accession number HQ214118.1) was used to predict 3D structure by use of the ITASSER server based on multiple-threading alignments by LOMETS and iterative TASSER assembly simulations (Roy et al. 2010).

UV–visible spectroscopy

UV–visible spectra of cHSPA6 were recorded from 250 to 700 nm at 20 °C, by use of an Ultrospec 2,100 Pro UV–visible spectrophotometer (Amersham Pharmacia Biotech). A quartz cell with 10-mm path length was used. Protein concentration was calculated by use of a cHSPA6 molar extinction coefficient of 34,840.

Fluorescence spectroscopy

Fluorescence of 2 μM cHSPA6, at 20 °C, was recorded by use of a Jasco FP-750 spectrofluorimeter with a 10-mm path length quartz cell; the excitation and emission slit widths were 5 nm. The emission spectra of native and 6 M GdnHCl-denatured cHSPA6 were measured by excitation at 280 nm and recording emission between 300 and 400 nm. Conformational transitions of cHSPA6 induced by different concentrations of ATP and ADP were recorded by excitation at 280 nm and recording emission between 295 and 400 nm. The fluorescence signal at 335 nm was used to calculate percentage quenching by use of the equation:

$$\% {\text{ quenching}} = \left( {F_{0} - F} \right) \times 100/F_{0}$$

where F 0 and F represent fluorescence intensity in the absence and presence of quencher, respectively.

ANS (8-anilino-1-naphthalenesulfonic acid) binding was measured, at 20 °C, by use of a Jasco FP-750 spectrofluorimeter with a 10-mm path length cuvette; the excitation and emission slit widths were 5 nm. Samples were excited at 375 nm and emission was recorded between 450 and 510 nm. The cuvette containing 100 μM ANS was pre-equilibrated with 2 μM cHSPA6 before nucleotide binding was started. Twofold serially diluted ATP or ADP solutions (2 μl) were inoculated and equilibrated before each measurement. Final dilution was less than 3 %. Changes in cHSPA6 hydrophobic surfaces as a result of the conformational change induced by nucleotide binding were estimated by ANS binding.

Circular dichroism

Circular dichroism measurements were performed, at 20 °C, with a Chirascan-Plus CD spectrophotometer (Applied Photophysics, Leatherhead, UK) coupled with Peltier temperature controller. The CD instrument was calibrated with (1S)-(+)-10-camphorsulfonic acid. To measure far-UV spectra, the monochromator was set at 260 nm with 1 nm bandwidth and data were collected at 0.5 s per point. Five scans of the CD spectra of 3 μM cHSPA6 in 10 mM sodium phosphate buffer, pH 7.5, containing 50 mM NaCl were performed between 200 and 260 nm in the presence and absence of 0.66 mM nucleotides (ATP and ADP). Air baseline and buffer background were subtracted. The data were analyzed by use of Chirascan’s Pro-data viewer. The path length was 1 mm and CD results were expressed in millidegrees (mdeg). Far-UV CD spectra were deconvoluted by use of the CDNN software package (version 2.1) provided by Applied Photophysics.

Dynamic multi-mode spectroscopy was performed with a Chirascan-Plus spectrophotometer, in accordance with the manufacturer’s instructions. Briefly, temperature-dependent conformational changes were measured for 3 μM cHSPA6 in 10 mM sodium phosphate buffer, pH 7.5, containing 50 mM NaCl. The CD spectra were recorded between 200 and 250 nm. The temperature was increased gradually from 20 to 94 °C, at a precise rate of 1 °/min. Thermal transition data were analyzed by use of Chirascan’s Global 3 software.

Results and discussion

cHSPA6 structure

The (His)6-tag was fused at the N-terminus of cHSPA6 and a highly specific protease cleavage site (TEV protease) was placed between the (His)6-tag and cHSPA6 to enable removal of the (His)6-tag after purification (Fig. 1a). The recombinant cHSPA6 fusion structure was 1983 nucleotides in length, corresponding to 661 amino acids or 72.7 kDa. The modeled 3D structure of cHSPA6 contained two domains (Fig. 1b). The N-terminal domain containing NBD has an α-helix and β-sheet structure whereas the C-terminal domain containing SBD has an α-helix-rich structure with a 24 amino acid-long stretch of an unstructured flexible C-terminus. The amino acid sequence of cHSPA6, when aligned with human HSP70B’ (BAD97082.1) or bovine HSPA6 (XP_589747.1) had 93 and 94 % sequence homology, respectively. The proteins had two conserved tryptophans and 14 conserved tyrosines (Tyr547 was an exception). One tryptophan residue is located in each domain whereas tyrosine residues were well distributed across the sequence.

Fig. 1
figure 1

Design of the cHSPA6 fusion protein. a Schematic diagram of the His-tagged cHSPA6 fusion protein. Amino acid numbering starts with the N-terminus of the His-tag (1–10), followed by the highly specific TEV protease cleavage site (11–18) and cHSPA6 (19–661). b 3D structure of cHSPA6 modeled by use of multiple-threading alignments by LOMETS. N-terminal NBD is shown in magenta and the C-terminal SBD is shown in green. The last 24 residues of SBD have a highly flexible, structureless conformation, shown in orange

Full size image
Fig. 2
figure 2

Purification of cHSPA6. Ni–NTA eluate was purified by size exclusion chromatography to obtain the monomeric form of cHSPA6. The purity of the Superdex 75 eluate was analyzed by Coomassie Blue staining. Lane 1, low molecular weight marker; lane 2, Ni–NTA eluate; lane 3, blank; lane 4, gel filtration eluate

Full size image

Expression and purification recombinant cHSPA6

The cHSPA6 was expressed by E. coli BL21(DE3) pLysS and subsequently purified, as described in the “Materials and methods” section. E. coli BL21(DE3) pLysS strain is lysogenic for the λ-DE3 cassette which expresses T7 RNA polymerase under the control of the IPTG-inducible lac-UV5 promoter. It also contains the pLysS plasmid to suppress the leakiness of cHSPA6 (Moffatt and Studier 1987). To obtain monomeric cHSPA6 species from higher-order structures and aggregated forms, the Ni–NTA elute was subjected to size-exclusion chromatography. Because it is well known that the HSP70 family of proteins have a tendency to dimerize and oligomerize, especially upon storage at low ionic strength (Borges and Ramos 2006; Schonfeld and Behlke 1998), a fresh preparation of cHSPA6 was made for spectroscopic studies. Highly pure cHSPA6 was obtained after size-exclusion ch romatography (Fig. 2). After further purification, 9 mg highly purified cHSPA6 was obtained, corresponding to 3.6 mg per gram wet biomass.

Fig. 3
figure 3

The UV–visible spectrum of purified cHSPA6 contained a single peak indicative of the ligand-free form of the protein

Full size image

UV–visible spectroscopy confirmed production of pure protein

Recombinant cHSPA6 contains two tryptophan residues (W92 and W582) which are located in NBD and SBD, respectively. The cHSPA6 also contains 16 tyrosines and 25 phenylalanine residues which are distributed throughout the primary structure. The UV–visible spectrum in the range 250–700 nm contained a single peak at 275 nm because of the dominance of tyrosine absorption (Fig. 3). The UV–visible spectrum is characteristic of soluble protein with no additional bound chromospheres.

Fig. 4
figure 4

Amino acid sequence alignment. Camel HSPA6 (accession number HQ214118.1), human HSP70B’ (BAD97082.1), and bovine HSPA6 (XP_589747.1) were aligned by use of MAFFT multiple sequence alignment software. Conserved tryptophans are indicated with hash symbols and all the tyrosines were conserved, indicated with asterisk symbols, except Tyr 547

Full size image

Fluorescence spectroscopy-indicated nucleotide binding

Recombinant cHSPA6 is a tyrosine-rich protein (Fig. 4). The fluorescence spectrum (excitation at 280 nm) was indicative of complex emission, with distinct peaks at 310 and 334 nm. This is probably because of the presence of 16 tyrosines and two tryptophans. The emission spectra of proteins containing both tryptophan and tyrosine residues are usually dominated by tryptophan fluorescence, even for proteins with a high tyrosine content, for example human serum albumin (one Trp and seventeen Tyr residues) (Teale 1960). Occasionally, however, for example the 33-kDa protein from spinach photosystem II (contains one Trp and eight Tyr residues), wheat alpha-purothionin (zero Trp and five Tyr), and crambin (zero Trp and two Tyr residues), the tyrosine emission spectrum has been observed (Prendergast et al. 1984; Ruan et al. 2002). Therefore, native cHSPA6 may belong to latter group of proteins, because its emission spectrum contained peaks corresponding to tyrosine and tryptophan emission (Fig. 5). When cHSPA6 was denatured, the 334 nm peak was red shifted to 346 nm, because of the environment-sensitive nature of tryptophan. (Edelhoch 1967; Rao et al. 1981). Moreover, when cHSPA6 was excited at 295 nm, the amplitude of the peak at 310 nm was substantially reduced (data not shown), further supporting the conclusion that the 310 nm peak corresponded to tyrosine emission whereas the 334 nm peak corresponded to tryptophan emission.

Fig. 5
figure 5

Intrinsic fluorescence spectra of native and denatured cHSPA6. a Excitation spectra of cHSPA6 in the native (solid line) and denatured (broken line) states recorded between 250 and 300 nm, Em 340 nm. Ex maximum was at 275 nm, because of the high tyrosine content. b Emission spectra measured at 280 nm, which excites both tryptophan and tyrosine residues. Two peaks are present: peak 1 is tyrosine emission spectra and peak 2 is tryptophan emission. The solid line is the emission spectrum of cHSPA6 in the native state and the broken line is that of the denatured state. When cHSPA6 was denatured, λ max of the second peak was red-shifted 14 nm because of the polarity-sensitive nature of tryptophan emission

Full size image

The results in Fig. 6a indicate that ATP and ADP quenched cHSPA6 total fluorescence. The tryptophan (Trp92) of cHSPA6 located in the NBD must be sensitive to the conformational changes induced by ATP or ADP binding. In previous studies, ATP binding to HSP70 member proteins resulted in fluorescence quenching (Banecki et al. 1992; Theyssen et al. 1996). Reported K D of mammalian Hsc70 ranges from 12 nM to 9.5 µM for ATP and from 18 nM to 1.6 µM for ADP. The dissociation constant of prokaryotic (E. coli, T. thermophilus) DnaK ranges from 1 to 47 nM for ATP and from 47 to 130 nM for ADP (Borges et al. 2005; Gao et al. 1994; Ha and McKay 1994; Huang et al. 1993; Klostermeier et al. 1998; Palleros et al. 1991; Russell et al. 1998; Theyssen et al. 1996). The precise affinity of Hsp70 for nucleotides varies, because of large differences among ATP release rate for members of the Hsp70 family (Mayer et al. 2001; Wegele et al. 2004). Our results show cHSPA6 has similar affinity for ADP and ATP, in agreement with those of Gao et al. (1994). Higher affinity for ATP (Ha and McKay 1994; Theyssen et al. 1996) or for ADP (Palleros et al. 1991) have also been reported. This different affinity is probably because of the different functional diversity of Hsp70 members and the different techniques used for investigation.

Fig. 6
figure 6

Nucleotide-induced change in the conformation of cHSPA6. a Intrinsic fluorescence quenching of 2 μM cHSPA6 in response to increasing concentration of ATP (closed triangles) and ADP (closed squares). Emission spectra were obtained by exciting at 280 nm, and emission at λmax 334 nm was used for calculating fluorescence quenching. b ANS fluorescence measurement in the presence of increasing concentrations of ATP (closed triangles) and ADP (closed squares). 2 μM cHSPA6 was mixed with 100 μM final ANS concentration. Changes in surface hydrophobicity as a result of increasing nucleotide concentrations were monitored by exciting at 375 nm. The insets in both figures show the fluorescence quenching and ANS fluorescence in response to nM concentrations of nucleotides

Full size image

Variations in surface hydrophobicity of cHSPA6 as a result of binding of ATP or ADP to the NBD were determined by use of an extrinsic fluorescent probe. The fluorescent dye 8-anilino-1-naphthalenesulfonic acid (ANS) is widely used to measure exposure of surface hydrophobicity during the course of conformational change (Walmsley et al. 1994). As shown in Fig. 6b, ATP and ADP binding both lead to changes in conformation, as is apparent from the loss of the ANS signal because of the decrease in surface hydrophobicity. Thus fluorescence spectroscopy reveals that ATP and ADP binding in the nanomolar range lead to conformational changes in cHSPA6.

Circular dichroism-characterized nucleotide-induced changes in the conformation of cHSPA6

cHSPA6 is an α-helix-rich structural protein, as demonstrated by its far-UV CD spectra (Figs. 1b, 7). Our results showed that ATP binding led to an increase in α-helix structure whereas in the ADP-bound state the α-helix structure was less than in the unbound state. One possibility is that the conformation of Hsp70 switches between the ATP-bound open conformation and the ADP-bound closed state. When ATP binds, Hsp70 assumes a substrate-binding-site-open conformation. After hydrolysis of ATP to ADP, Hsp70 undergoes structural changes in which the substrate-binding site is in the closed state and the substrates are trapped (Freeman et al. 1995; Wu et al. 2001). In-vitro experiments have shown that Hsp70 proteins have affinity for misfolded proteins and small peptides. They also have weak ATPase activity. Binding of ATP induces conformational changes and results in release of the peptides (Flynn et al. 1989; Gragerov et al. 1994; Mayer et al. 2001; Palleros et al. 1993).

Fig. 7
figure 7

Far-UV CD spectra of cHSPA6 in the presence of the nucleotides. Changes in the secondary structure induced by nucleotide binding were monitored by far-UV CD. The solid line is the spectrum of 3 μM cHSPA6 in the absence of nucleotide; the dotted line is the spectrum in the presence of 0.66 mM ATP; the dashed line is the spectrum in the presence of 0.66 mM ADP

Full size image

The α-helix content of the secondary structure in cHSPA6 in the presence or absence of nucleotides was estimated from far-UV CD spectra by use of CDNN CD software (Table 1) (Bohm et al. 1992). Binding of ATP resulted in an increase in α-helix content (Table 1, Fig. 7). Similar conformational changes in the presence of ATP have also been observed for EcDnaK by use of infrared spectroscopy and time-resolved infrared difference spectroscopy (Banecki et al. 1992; Moro et al. 2006). Interestingly, ATP binding with EcDnaK leads to quenching and conformational changes but ADP binding neither quenches EcDnaK fluorescence nor induces conformational change. However, human Hsc70 undergoes conformational changes when bound to ADP (Wu et al. 1985). ATP induces conformational changes in both human Hsc70 and EcDnaK. However, ADP binding leads to conformational change of human Hsc70 but not of EcDnaK. Such changes might lead to functional diversity in the chaperoning system (Borges et al. 2005; Cheetham and Caplan 1998; Fan et al. 2004).

Table 1 Alpha-helix content of cHSPA6 in the presence and absence of nucleotides, calculated from 200 to 260 nm CD spectra by use of CDNN
Full size table

Dynamic multimode spectroscopy

Dynamic multimode spectroscopy is a newly developed, information-rich experimental technique which furnishes structural and thermodynamic data. To calculate the thermodynamics of unfolding and the mid-points of unfolding transitions, CD spectra between 200 and 250 nm of cHSPA6 in the nucleotide-free state were recorded as a function of temperature. Figure 8a shows the protein underwent three well-defined thermal transitions as a function of temperature, suggesting four folding states. The CD spectra indicated gradual conversion of the alpha-helix rich structure into a β-sheet dominant structure as the temperature was increased (Fig. 8b). The relative concentrations of the different folding states as a function of temperature are shown in Fig. 8c. The melting points calculated by use of Global 3 analysis software were 42.3 ± 0.2, 61.3 ± 0.1, and 81.2 ± 0.2 °C. DSC study of DnaK revealed three unfolding transitions in the nucleotide-free state, with T m values close to 45.2, 58, and 73.3 °C (Montgomery et al. 1999). Thermally induced unfolding of human Hsc70 also occurred via three transitions, with T m of 46, 68, and 82 °C (Palleros et al. 1991); as a result of unfolding of the NBD and SBD and collapse of the remaining structures in the N and C-terminal domains (Montgomery et al. 1999). The NBD domain of cHSPA6 is slightly less stable than those of the bacterial DnaK and human Hsc70 counterparts, whereas the stability of cHSPA6 SBD is between those of DnaK and Hsc70. A 3D model of the thermal transitions in cHSPA6 was generated by use of Global 3 analysis software (Fig. 8d). Three transitions as a function of temperature are clearly apparent at low wavelengths.

Fig. 8
figure 8

Far-UV CD as a function of temperature and wavelength. a Temperature-dependent conformation changes at different wavelengths. CD values in mdeg units at different wavelengths in the far-UV region were plotted as a function of temperature. Global 3 analysis software, especially developed by Applied Photophysics for dynamic multimode spectroscopy, was used for determination of the thermal transitions. The wavelengths are indicated on the right Y axis. Three thermal transitions are clearly seen at wavelengths between 204 and 225 nm. b Calculated far-UV CD spectra of the contributing species. At low temperature cHSPA6 adopts an α-helix conformation. The plots show the CD spectra of the species if they were the sole contributors on completion of a thermal transition. With increasing temperature, cHSPA6 loses the α-helix conformation, and unfolding species 2 and 3 are the more beta-strand-rich structures. c Calculated concentration profiles of the folding species. The graph shows the temperature-dependent disappearance and appearance of the species and their relative concentrations. d Calculated temperature–wavelength and CD surface. The 3D model was calculated by use of Global 3 software by using CD values obtained at increasing temperatures in the far-UV region

Full size image

Conclusion

We have produced cHSPA6 by recombinant techniques and purified it to homogeneity in quantities sufficient to perform structural and thermodynamic studies. We observed via intrinsic fluorescence quenching that cHSPA6 had similar affinity for ATP or ADP. In ATP or ADP-bound cHSPA6 the hydrophobic surface was reduced. Far-UV CD measurements showed that cHSPA6 gains α-helix structure in the ATP-bound state but loses α-helix structure in the ADP-bound state. A similar conformational change in the nucleotide bound state has also been observed for mammalian Hsp70. The cHSPA6 unfolds via three thermal transitions in a manner similar to that observed for E. coli and for humans.