The thermostability of the recombinant α- and β-subunit homo-oligomers (α16mer and β16mer) and of natural chaperonins purified from cultured Thermococcus strain KS-1 cells was measured to understand the mechanism for the thermal acclimatization of T. KS-1. The β-subunit content of the natural chaperonin from cells grown at 90°C was higher than that at 80°C. The optimum temperature for ATPase activity of the natural chaperonins was 80–90°C, whereas that for α16mer and β16mer was 60°C and over 90°C respectively. Judging from the ATPase activity, β16mer was more thermostable than α16mer. The thermostabilities of the natural chaperonins were intermediate between α16mer and β16mer, whereas the natural chaperonin with a higher β-subunit content was more stable than that with a lower β-subunit content. Native polyacrylamide gel electrophoresis (PAGE) revealed that the chaperonin oligomers thermally dissociated to their ATPase-inactive monomers. The thermal denaturation process monitored by circular dichroism showed that the free β-subunit was more stable than the free α-subunit, and that the secondary structure of the chaperonin monomer in the oligomer was more stable than that in the free monomer. These results suggest that the structure of these subunits was stabilized in the oligomer, and that an increase in the β-subunit content conferred higher thermostability to the natural hetero-oligomeric chaperonin.
In the crowded protein environment in the cell, proteins require molecular chaperones for their proper folding. Chaperonin, one of the molecular chaperones, has a double-ring toroidal structure and is known to promote protein folding in vivo and in vitro in an ATP-dependent fashion (Ellis and van der Vies, 1991). It is classified into two distinct types that share a limited but significant sequence homology: group I chaperonin is found in bacteria and organella of eukaryotes and group II chaperonin is found in archaea and in the cytoplasm of eukaryotes (Kim et al., 1994; Kubota et al., 1995). The group I chaperonin, GroEL, of Escherichia coli is composed of a single species of a 60 kDa subunit, which forms a sevenfold, double-ring homo-oligomer and functionally co-operates with a heptamer of the co-chaperonin, GroES (Hartl, 1996; Fenton and Horwich, 1997; Sigler et al., 1998). No such co-chaperonin has been found for group II chaperonins.
The archaeal group II chaperonin, which consists of one to three kinds of homologous subunits, has an eight or ninefold rotationally symmetrical double-ring toroidal structure (Archibald et al., 1999; Gutsche et al., 1999). Most archaeal chaperonins, particularly those of Sulfolobus and Thermoplasma, have been reported to be composed of two distinct subunits, with about 60% identity, and form a hetero-oligomer with a fixed subunit composition (Knapp et al., 1994; Ditzel et al., 1998; Ellis et al., 1998). Recently, a gene for another chaperonin subunit, γ, was found in Sulfolobus genomes (Archibald et al., 1999). It has been reported that both the α- and β-subunits of chaperonin accumulated in response to a high growth temperature and/or heat shock in such archaea as Sulfolobus shibatae (Trent et al., 1991; Kagawa et al., 1995), Pyrodictium occultum (Phipps et al., 1991), Haloferax volcanii (Kuo et al., 1997) and Archaeoglobus fulgidus (Emmerhoff et al., 1998). The ratio of the α- to β-subunit is constant at different growth temperatures and their expression is thought to be co-regulated (Phipps et al., 1991; Trent et al., 1991; Kagawa et al., 1995). The natural chaperonin in these archaea has been reported to be a hetero-oligomer of fixed subunit ratio. However, the functional difference between these archaeal chaperonin subunits is still unknown. The chaperonin of Thermococcus strain KS-1 is composed of two distinct subunits, α and β (Yoshida et al., 1997, 2000). Their amino acid sequences are highly similar throughout the whole length, except for about 20 amino acid residues at the C-terminal end. Both of them form a toroidal, double-ring homo-oligomer with eightfold rotational symmetry. Despite their high sequence similarity (81% identity), their expression was different. The cellular content of the α-subunit was lower at higher temperatures, with the minimum at 90–93°C, whereas that of the β-subunit increased with increasing temperature up to 93°C (Yoshida et al., 2001). This result indicates that the gene expression of the α- and β-subunits of T. KS-1 chaperonin was differently regulated, and that only the β-subunit was a thermally inducible heat-shock protein. The T. KS-1 chaperonin formed in vivo a hetero-oligomer with variable subunit composition at different growth temperatures (Yoshida et al., 2001). This change in the subunit composition of T. KS-1 chaperonin is thought to be acclimatization to different temperatures. To examine this hypothesis, we studied the thermostability of the recombinant α- and β-chaperonin homo-oligomers and of natural chaperonins in Thermococcus strain KS-1.
Temperature dependence of the ATPase activity of the natural and recombinant chaperonin oligomers
The recombinant α- and β-homo-oligomers (α16mer and β16mer) showed their optimum temperatures for ATPase activity at 60°C and at more than 90°C respectively (Fig. 1). The optimum temperature for the ATPase activity of the natural chaperonins purified from cultured T. KS-1 cells grown at 80°C (nCPN80) and at 90°C (nCPN90) was about the same (80–90°C), being between the optima of the recombinant chaperonin homo-oligomers (Fig. 1). The specific activity of β16mer at its optimum temperature was higher than that of α16mer at its optimum temperature. At 50–90°C, the specific ATPase activity of the two natural chaperonins was about the same, but higher than that of α16mer and of β16mer.
Thermal stability of the recombinant and natural chaperonins
The thermal stability of the ATPase activity of the recombinant and natural chaperonin oligomers was measured at temperatures from 80–95°C (Table 1). At 80°C, the ATPase activity of α16mer and β16mer exponentially decreased at about the same rate (data not shown), being approximately 40% of the value before the heat treatment after 90 min of incubation at 80°C. At higher temperatures, the difference of their half-lives became obvious (Table 1), the β16mer being significantly more stable than α16mer at 85°C and higher. At 90°C, the half-life of the ATPase activity of nCPN90 was longer than those of α16mer and nCPN80 but was almost the same as that of β16mer (Table 1). These results indicate that β16mer was more thermostable than α16mer, and that the natural chaperonin with higher β-subunit content was more thermostable than that with lower β-subunit content.
Table 1. Half-life of the ATPase activity of recombinant and natural chaperonins.
Natural chaperonin oligomer
Incubation temperature [°C]
ND, not determined.
The recombinant α- and β-subunit homo-oligomers (α16mer and β16mer) and natural chaperonin oligomers, which had been purified from cultured cells grown at 80°C (nCPN80) and 90°C (nCPN90), were incubated at different temperatures for various periods of time in an ATPase reaction buffer without ATP. After the incubation, the reaction mixture was ice-chilled and the residual ATPase activity was measured at 80°C. Corresponding values were determined from best-fit curves.
Thermal dissociation of the T. KS-1 chaperonin oligomer to its monomer
To examine the thermal stability of the chaperonin oligomer, it was heat-treated and then analysed by native polyacrylamide gel electrophoresis (PAGE) (Fig. 2A). Native α16mer and β16mer migrated as multiple bands, and small amounts remained in the polyacrylamide gel well. These might have formed multimers of double-ring homo-oligomers. The chaperonin remaining in the well was thought to be aggregated chaperonin. The α16mer migrated a little faster than β16mer; after 30 min of incubation at temperatures from 80–95°C, the intensities of the multiple bands of α16mer and β16mer gradually decreased with increasing temperature. New bands corresponding to their respective monomers consistently appeared and the density gradually increased (Fig. 2A, arrowheads). These results suggest that α16mer and β16mer dissociated to the monomer. The purified, folded monomer of the α-subunit or β-subunit of Thermococcus chaperonin did not have detectable ATPase activity (data not shown). These data suggest that, at the higher temperatures, the chaperonin oligomer with ATPase activity dissociated to its ATPase-inactive monomer.
After incubating at 95°C for 30 min, the band corresponding to α16mer had completely disappeared, and the proteins seemed to aggregate in the well (Fig. 2A, *). This result suggests that at 95°C, α16mer dissociated to its monomer, unfolded and then aggregated. At 90°C, the dissociation of α16mer proceeded more rapidly than that of β16mer (data not shown). These results indicate that β16mer was more resistant to thermal dissociation than α16mer. The natural chaperonins, nCPN80 and nCPN90, were incubated at 90°C for 0–60 min to examine the thermal dissociation (Fig. 2B). They migrated as a single band with slightly different mobility. The nCPN80 contained a small amount of monomers at 0 min (Fig. 2B). To evalu-ate the dissociation of the oligomer to the monomer, the ratio of the band density of the monomer to that of the oligomer (monomer:oligomer ratio) was calculated from the data of incubation at 90°C. The broad band, which migrated a little faster than the α-subunit monomer, was thought to be its degradation product and was therefore taken into account. At this temperature, monomer/ oligomer ratio increased more rapidly in α16mer than that in β16mer (data not shown). The ratios for the natural chaperonins, nCPN80 and nCPN90, were similar to each other and were between the values for α16mer and β16mer (data not shown).
Thermal change in the circular dichroism of T. strain KS-1 chaperonin
The thermal unfolding process of the chaperonin in its oligomer and free monomer was analysed by far-UV circular dichroism (CD) at different temperatures. Although the temperature for 50% denaturation (Tm) could not be directly determined for any of the oligomers and monomers in the 30–100°C range (Fig. 3), Tm values were calculated from the best-fit denaturation curves by using a two-state denaturation model (Fig. 3; Table 2). CD at 222 nm for the chaperonin oligomers (α16mer and β16mer) changed sharply in the 95–105°C range. The values for the α- and β-subunit monomers changed gradually between 70 and 100°C, and between 80 and 100°C, respectively (Fig. 3, Table 2). The Tm value for the β-subunit monomer was higher than that of the α-subunit monomer, whereas the Tm values for the oligomers were about the same as each other, but higher than those of the monomers (Table 2). The standard enthalpy changes during the denaturation (ΔdnH) of the α- and β-subunit monomers, which were calculated by curve fitting, were 30 kcal mol−1 and 40 kcal mol−1 respectively (Table 2). These results indicate that the β-subunit monomer was more thermally stable than the α-subunit monomer. The CD change and Tm value of β16mer were almost the same as those of α16mer; however, the ΔdnH value of β16mer was higher than that of α16mer (Table 2). The ΔdnH value of the oligomer was also higher than that of the corresponding monomer (Table 2). These results indicate that the secondary structure of the monomers in the chaperonin oligomer was more stable than that of the free monomer, and that the β-subunit was more thermostable than the α-subunit.
Table 2. Thermodynamic parameter for the thermal unfolding of the chaperonin oligomer and monomer.
ΔdnH° and Tm values were calculated by fitting the curves to actual values assuming two-state denaturation.
ΔdnH° (kcal mol−1)
The optimal temperature for the ATPase activity of β16mer (>90°C) was higher than that for α16mer (approximately 60°C) (Fig. 1). However, the optimal temperatures (80–90°C) for the ATPase activity of the natural chaperonins of cells grown at 80°C and 90°C were about the same, being between those of α16mer and β16mer (Fig. 1). The optimum growth temperature for Thermococcus strain KS-1 is 85°C, with a growth temperature range between 60°C and 97°C (Hoaki et al., 1994). The subunit composition of natural T. KS-1 chaperonin changes with the growth temperature (Yoshida et al., 2001), and its subunit ratio (α/β) decreases with increasing growth temperature. The present data probably indicate that the optimum temperature for the ATPase activity of T. KS-1 natural chaperonin is relatively constant in the wide temperature range of their habitats. The specific ATPase activity of the natural chaperonin of T. KS-1 was significantly higher than that of α16mer and β16mer in the growth temperature range (Fig. 1). There may have been a synergistic effect between the subunits in the hetero-oligomer. The natural chaperonin of Pyrodictium occultum, which is a hetero-oligomer, has also been reported to have higher ATPase activity than those of the recombinant homo-oligomers (Minuth et al., 1998). A natural chaperonin with relatively high specific ATPase activity may have higher protein folding activity than a homo-oligomer with lower specific activity.
The β16mer was more thermally stable than α16mer in ATPase activity. The higher the β-subunit content, the higher was the thermal stability of the ATPase activity (Table 1). When α16mer and β16mer were incubated at a high temperature like 90°C, they were dissociated into their monomers (Fig. 2A). Although the archaeal chaperonin monomer from Methanococcus thermolithotrophicus exhibited no ATPase activity, it became ATPase active when it formed the oligomer (Furutani et al., 1998). The monomers of the Thermococcus chaperonin subunit also had no ATPase activity. The loss of ATPase is thought to have been caused by this dissociation. The dissociation of α16mer was faster than that of β16mer (Fig. 2A). Temperature-induced dissociation was also observed in the natural chaperonins (Fig. 2B). These results indicate that β16mer was more resistant to thermal dissociation than α16mer. The thermal denaturation process of the chaperonin homo-oligomers and subunit monomers was analysed by CD at 222 nm (Fig. 3). The standard enthalpy changes during the denaturation (ΔdnH°) indicates that the thermal stability of the subunit in the oligomer was higher than that of the free subunit monomers, and that the β-subunit was more thermally stable than the α-subunit. These subunits were stabilized by oligomerization, and the β-subunit was more adapted to higher temperatures than the α-subunit.
The eukaryotic group II chaperonin, CCT (chaperonin containing t-complex peptide 1), is an eightfold, symmetrical, double-ring hetero-oligomer with eight distinct subunits (Kubota et al., 1995). The mutants of yeast lacking any of the CCT subunits are lethal (Ursic and Culbertson, 1991; Chen et al., 1994; Li et al., 1994; Miklos et al., 1994; Vinh and Drubin, 1994). Unlike the group I chaperonins that can fold a board range of proteins, the in vivo substrates for CCT are known to be restricted to a relatively small number of cytoskeletal proteins, mainly actin and tubulin (Frydman et al., 1992; Gao et al., 1992; Yaffe et al., 1992). Binding of the substrate proteins to CCT is subunit-specific and depends on the geometry of the ring (Llorca et al., 1999, 2000). The amino acid sequence differences among their subunits are mainly in the apical domain, the substrate-binding domain (Kim et al., 1994). Each subunit of CCT may play a distinct role in the substrate inter-action (Kubota et al., 1994; Willison and Kubota, 1994). Whereas some methanogens and Pyrococcus species have a single species of subunit for the chaperonin, the chaperonins of most archaea consist of two homologous subunits (Archibald et al., 1999). The natural substrate of archaeal chaperonin is not yet known. In hyperthermophilic archaea, only a few molecular chaperones have been found in their genomes that may indicate that the substrate specificity of archaeal chaperonins is broad. The hetero-oligomer of dissimilar subunits may indicate that they have different functions or characteristics for adaptation to the environment.
The ATP dependence of protein folding mediated by archaeal chaperonins has been reported (Guagliardi et al., 1994; 1995; Yoshida et al., 1997; 2000; Furutani et al., 1998). The ATPase activity is the driving force for chaperonin-mediated protein folding. The chaperonins only showed ATPase activity when oligomers were formed (Furutani et al., 1998). This oligomerization is important for chaperonin activity. The thermal stability was different between the α- and β-subunits in Thermococcus. The present results clearly indicate that the hetero-oligomer with a high β-subunit content would be suitable for higher temperatures. A chaperonin with a high α-subunit content is probably adequate for lower temperatures. At high temperatures, like 90°C, the β-subunit may have higher protein-refolding activity than that of the α-subunit. At higher temperatures in vivo, the β-subunit content of na-tural chaperonins increased, and the resulting chaperonin may function more efficiently than those with a higher α-subunit composition. Thermococcus hyperthermophiles change the subunit composition of the chaperonin to acclimatize to their environment.
Expression and purification of the recombinant α- and β-subunit homo-oligomers of Thermococcus strain KS-1 chaperonin
The subunits of Thermococcus strain KS-1 chaperonin were expressed in Escherichia coli strain BL21 (DE3) cells with the expression vector, pK1Eα2 or pK1Eβ (Yoshida et al., 1997; 2000). They were grown aerobically overnight at 37°C in a 2× YT medium (1.6% bactotriptone, 1% yeast extract, 0.5% NaCl) supplemented with 100 μg ml−1 of ampicillin or 75 μg ml−1 of kanamycin. Both the recombinant α- and β-homo-oligomers (α16mer and β16mer) were purified as described previously (Yoshida et al., 2001). The purified recombinant homo-oligomers were stored at 4°C.
Purification of the natural chaperonin from Thermococcus strain KS-1 cells
Thermococcus strain KS-1 was grown anaerobically as described previously with slight modifications (Hoaki et al., 1994). The medium contained 75% natural sea water, 5 ml l−1 of Wolf’s trace element solution (Wolin et al., 1963), 1.89 g l−1 of yeast extract, 1.89 g l−1 of peptone, 1.35 g l−1 of cystine, approximately 0.05 g l−1 of S° and 0.6 mg l−1 of resazulin. Then, 50 ml of the overnight culture, grown at 90°C, was inoculated into several bottles containing 1000 ml of the same medium and incubated for 10 h at 90°C. Each culture was then transferred to 20 l of the same medium in 30 l fermenters and incubated overnight at 80°C or 90°C (Hoaki et al., 1994). After this cultivation, the culture was immediately cooled to room temperature by circulating water for approximately 10 min, and then pooled in plastic containers in an ice-chilled water bath, before the cells were harvested with a continuous centrifuge (Kokusan) at 10 000 g. The natural chaperonins were purified from Thermococcus strain KS-1 cells grown at 80°C or 90°C as described pre-viously (Yoshida et al., 2001). The ratios of β/α in these purified native chaperonins from the cells grown at 80°C and 90°C were, respectively, determined to be approximately 0.45 and 8.5 using an enzyme-linked immunosorbent assay (ELISA) (Yoshida et al., 2001). The purified natural oligomers were stored at 4°C.
Preparation of the chaperonin subunit monomers
To obtain the chaperonin monomer, α16mer and β16mer were denatured at 90°C for 60 min in a 25-mM Tris-HCl buffer (pH 8.1) containing 8 M guanidine HCl and 1 mM dithiothreitol to a final concentration of 5 mg ml−1. After denaturation, the protein solutions were diluted 25-fold with the 25 mM Tris-HCl buffer (pH 8.1) containing 300 mM KCl, and then incubated at room temperature for 2 h. The circular dichroism (CD) spectra of the monomers were the same as those of the corresponding chaperonin oligomers, indicating that their structures were intact. The monomer solutions were stored at 4°C.
Assay of ATPase activity
The ATPase activity was measured at 40–90°C in a reaction mixture containing a 25-mM Tris-HCl buffer (pH 7.2), 300 mM KCl, 1 mM MgCl2, 1 mM ATP and 50 ng μl−1 of chaperonin. The reaction mixture was preincubated at the assay temperature for 5 min before adding ATP to start the reaction. The reaction was terminated with perchloric acid at a final concentration of 2%. Produced Pi was measured by the malachite green method (Baykov et al., 1988; Geladopoulos et al., 1991), and the rate of spontaneous ATP hydrolysis at each temperature was measured and taken into account for assessing the ATPase activity.
Thermal stability of the chaperonin oligomers
To examine the thermal stability of the ATPase activity of chaperonin, chaperonin oligomers (final concentration of 0.5 μg μl−1) were incubated at 80–95°C for various periods in the ATPase reaction mixture without ATP. Aliquots were taken at appropriate intervals and chilled on ice, and then the residual ATPase activity was measured at 80°C as described in the method for the assay of ATPase activity. To examine the thermal dissociation of the recombinant homo-oligomer, the purified recombinant α- or β-subunit homo-oligomer (0.5 μg μl−1) was incubated in the ATPase reaction mixture without ATP at 80–95°C. After 30 min of incubation, the proteins in each 5 μl of the reaction mixture were analysed by polyacrylamide gel electrophoresis of a 3–6% linear gradient polyacrylamide gel without SDS (native PAGE). Each gel was stained with Coomassie brilliant blue R-250. For the time-course experiment on thermal dissociation, the natural chaperonins were incubated at 90°C. Aliquots were taken at adequate intervals and chilled on ice, and then the proteins were analysed by native PAGE. After Coomassie brilliant blue staining, protein bands were scanned with a flat-bed type scanner (GT-9500, EPSON) and analysed with MULTI-ANALYST software (Bio-Rad Laboratories).
Circular dichroism measurements
Far-UV CD was measured with a J-500C spectrometer (Jasco). A 1 cm path-length quartz cell was used, the temperature of the sample solution in the cell being maintained at 30–100°C with a Peritie type of programmable temperature controller. The in situ temperature was monitored with a thermocouple. The temperature scan rate was 1°C min−1. Thermococcus KS-1 chaperonin was dissolved at 50 μg ml−1 in a 25-mM Tris-HCl (pH 8.1) buffer containing 300 mM KCl. The thermal change of CD at 222 nm was monitored. The CD data for the unfolded chaperonin were measured for the same concentration of the chaperonin oligomer, which was incubated in the 25 mM Tris-HCl (pH 8.1) buffer containing 300 mM KCl and 8 M guanidine-HCl.
To indicate the degree of denaturation of the chaperonin, the CD data were normalized by those with and without 8 M guanidine-HCl at 30°C. The value of 1.0 corresponds to the denatured form by guanidine-HCl and 0 to the native form. The normalized data ([θ]) were analysed by assuming the two-state denaturation model:
where [θ]n and [θ]d are the extrapolated values for [θ] of the folded and unfolded chaperonin, respectively, and fd is the mole fraction of the unfolded chaperonin. It was assumed that [θ]n would be identical for the α monomer, β monomer, α16mer and β16mer (Fig. 3). The temperature dependence of [θ]n was calculated by the least squares method from [θ] of α16mer and β16mer at 30–80°C. The denaturation of chaperonin was not complete at 100°C, and [θ]d was unmeasurable. With the assumption that [θ]d was also identical for all the monomers and oligomers, [θ]d and the standard enthalpy change during denaturation (ΔdnH°) of α16mer and β16mer were simultaneously calculated by curve fitting with the semi-Newton method. ΔdnH° of the α- and β-subunit monomers was calculated by fitting [θ] to the model independently, with [θ]d and [θ]n-values calculated from [θ] of α16mer and β16mer.
The proteins were analysed by polyacrylamide gel electro-phoresis, on 3–6% gradient polyacrylamide gel without SDS (Laemmli, 1970). The protein concentration was measured by the Bradford method with a protein assay kit (Bio-Rad Laboratories), using bovine serum albumin (BSA) as the standard (Bradford, 1976).
We thank Ms. C. Ohkami and Ms. H. Iwabuchi for their technical assistance. This work was supported by grant-in-aid for scientific research on priority areas (11153206) from the Ministry of Education, Science, Sports and Culture of Japan and for the ‘Biodesign Research Program’ from RIKEN to M.Y. This study was conducted at the Marine Biotechnology Institute of Japan as part of The Basic Knowledge Creation and Development programme supported by the New Energy and Industrial Technology Development Organization of Japan.