- Email: [email protected]
Reactivation and refolding of rabbit muscle creatine kinase denatured in 2,2,2-trifluoroethanol solutions
Biochimica et Biophysica Acta 1545 (2001) 305^313 www.elsevier.com/locate/bba
Reactivation and refolding of rabbit muscle creatine kinase denatured in 2,2,2-tri£uoroethanol solutions Kai Huang a , Yong-Doo Park a , Zhi-Fang Cao a , Hai-Meng Zhou
a;b;
*
a
b
Department of Biological Science and Biotechnology, Tsinghua University, Beijing 100084, People's Republic of China National Laboratory of Biomacromolecules, Institute of Biophysics, Academia Sinica, Beijing 100084, People's Republic of China Received 3 July 2000; received in revised form 12 December 2000; accepted 12 December 2000
Abstract The unfolding and refolding of creatine kinase (ATP:creatine N-phosphotransferase (CK), EC 2.7.3.2) during denaturation and reactivation by trifluoroethanol (TFE) have been studied. Significant aggregation was observed when CK was denatured at TFE concentrations between 10% and 40% (v/v). 50% TFE (v/v) was used to study the denaturation and unfolding of CK. The activity loss of CK was a very quick process, as was the marked conformational changes during denaturation followed by fluorescence emission spectra and far-ultraviolet CD spectra. DTNB modification and size exclusion chromatography were used to find that CK dissociated and was in its monomer state after denaturation with 50% TFE. Reactivation and refolding were observed after 80-fold dilution of the denatured CK into 0.05 M Tris-HCl buffer, pH 8.0. The denatured CK recovered about 38% activity following a two phase course (k1 = 4.82 þ 0.41U1033 s31 , k2 = 0.60 þ 0.01U1033 s31 ). Intrinsic fluorescence maximum intensity changes showed that the refolding process also followed biphasic kinetics (k1 = 4.34 þ 0.27U1033 s31 , k2 = 0.76 þ 0.02U1033 s31 ) after dilution into the proper solutions. The far-ultraviolet CD spectra ellipticity changes at 222 nm during the refolding process also showed a two phase course (k1 = 4.50 þ 0.07U1033 s31 , k2 = 1.13 þ 0.05U1033 s31 ). Our results suggest that TFE can be used as a reversible denaturant like urea and GuHCl. The 50% TFE induced CK denaturation state, which was referred to as the `TFE state', and the partially refolded CK are compared with the molten globule state. The aggregation caused by TFE during denaturation is also discussed in this paper. ß 2001 Elsevier Science B.V. All rights reserved. Keywords: Creatine kinase; Denaturation; Reactivation; Refolding; Aggregation; Hydrophobic surface; Helix induction; Molten globule state; Tri£uoroethanol state
1. Introduction Rabbit MM-creatine kinase (CK) is a dimer, conAbbreviations: CK, creatine kinase; TFE, 2,2,2-tri£uoroethanol ; TFE state, 50% TFE induced CK denaturation state ; CD, circular dichroism; DTNB, 5,5P-dithiobis(2-nitrobenzoic acid); GuHCl, guanidine hydrochloride; ANS, 1-anilinonaphthalene-8sulfonate; SEC, size exclusion chromatography * Corresponding author, at address a. Fax: +86-10-6277-5505; E-mail: [email protected]
sisting of two identical subunits. The activity and conformational changes of CK in the presence of GuHCl and urea have been extensively studied [1^ 7] and upon diluting the denaturant under suitable conditions, both the conformation and the catalytic activity can be quantitatively recovered [8^11]. However, relatively few attempts have been made to study the denaturation and reactivation of CK in organic solutions such as alcohol. 2,2,2-Tri£uoroethanol (TFE)/H2 O mixtures, which
0167-4838 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 0 0 ) 0 0 2 9 3 - 4
BBAPRO 36359 24-1-01
306
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
were introduced by Goodman and Listowsky [12], are known to stabilize secondary structures such as K-helices and L-hairpins [13,14]. Peptides with sequences derived from proteins typically show little or no helix formation in water [15] but can be induced to form helices in TFE/H2 O mixtures [16^19]. TFE solutions have also been used to study partially folded states and equilibrium intermediates [20,21]. More recently, several groups have observed changes in the rates of folding and unfolding upon addition of TFE [22^27]. The denaturation and unfolding process of several proteins in TFE solutions have been studied previously [28^32]. Yang et al. [33] previously reported that CK was inactivated and underwent conformational changes in di¡erent concentrations of TFE. However, the present study shows signi¢cant aggregation of CK in TFE solutions with concentrations between 10% and 40% (v/v). Aggregation was not discovered at TFE concentrations lower than 5% (v/v) or higher than 50% (v/v). TFE can a¡ect several structural and physical properties of the system [34], such as weakening hydrophobic interactions, strengthening intramolecular hydrogen bonds, changing the dielectric constant of the solvent and acting as an osmolyte. TFE induced denaturation results in partly folded states [20,21,30], which may resemble in some features early intermediates during folding and might be prone to aggregation or misfolding [30,35]. During the denaturation of CK in TFE concentrations between 10% and 40%, one possible explanation for the observed aggregation is that the number of TFE molecules at these concentration ranges is not su¤cient to disrupt the hydrophobic core and/or any hydrophobic clusters of the induced partly unfolded intermediates. At concentrations larger than 50%, enough TFE molecules are present to prevent intermolecular condensation of the unfolding intermediates. At low concentrations, TFE does not signi¢cantly perturb the native structure of the protein [28] but does increase the helical content of the peptides in direct correlation with the propensity of the amino acid sequence to form helices [36]. Molten globule-like intermediates during alcohol induced denaturation have been observed for lysozyme [28,37] and myoglobin [38]. A clear distinction between the molten globule-like states and the TFE
states is important when considering the structural changes observed in the presence of TFE. This paper compares the observed TFE state and the molten globule state. We suggest that the TFE state induced by 50% TFE is not the molten globule state in the CK folding pathway. However, the partially refolded CK with high hydrophobic surface exposure, nativelike secondary structure and nearly the same compact structure as the native protein is very like the molten globule state in the CK folding process. The experimental results also suggest that TFE can be used as a kind of reversible denaturant like urea and guanidine hydrochloride (GuHCl) and the helical secondary structure recovery, the tertiary structure recovery and the activity reactivation are cooperative processes with similar kinetic constants. 2. Materials and methods The preparation of rabbit muscle creatine kinase and the assay of creatine kinase activity were as described by Yao et al. [39]. The ¢nal puri¢ed enzyme was homogeneous on polyacrylamide gel electrophoresis in the presence and absence of SDS. Tri£uoroethanol, 5,5-dithiobis(2-nitrobenzoic acid) (DTNB) and 1-anilinonaphthalene-8-sulfonate (ANS) were Sigma products, while creatine and ATP were obtained from Fluka. All other reagents were local products of analytical grade. The enzyme concentration was determined by measuring the absorbance at 280 nm and using the absorption coe¤cient A1% 1cm = 8.8 [40]. Enzyme denaturation was carried out in a solution containing 50% TFE (v/v) in 0.05 M Tris-HCl bu¡er, pH 8.0 for 1 h. Denatured CK modi¢cation was carried out with 10fold molar excess DTNB in 0.05 M Tris-HCl bu¡er, pH 8.0 for 1 h. Reactivation and refolding were carried out after 80-fold dilution of the denatured CK into the same bu¡er without TFE. Fluorescence spectrum measurements were made with a Hitachi 850 spectro£uorimeter. We used an excitation wavelength of 295 nm for the tryptophan £uorescence measurements. Circular dichroism (CD) spectra were recorded on a Jasco 725 spectropolarimeter with a path length of 2 mm. The CD intensity at 222 nm was used to calculate the amount of helical secondary structure. The aggregation measurements
BBAPRO 36359 24-1-01
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
307
were carried out with a Perkin Elmer Lambda Bio U/V spectrophotometer using the absorption wavelength 400 nm. Size exclusion chromatography was carried out with a Sephacryl s200 HR column on a Pharmacia FPLC apparatus. An excitation wavelength of 400 nm was used for ANS binding £uorescence measurements. The samples were incubated in 40 WM ANS for 30 min. All measurements were carried out at 25³C. 3. Results 3.1. Inactivation and unfolding of CK in the presence of 50% TFE (v/v) When CK was added to solutions containing different TFE concentrations in 0.05 M Tris-HCl bu¡er, pH 8.0, signi¢cant aggregation occurred at TFE concentrations between 10% and 40% (Fig. 1). At ¢rst, increasing the TFE concentration resulted in increased aggregation rate and extent. After reaching a maximum rate at a TFE concentration between 20% and 30%, the aggregation rate began to decrease
Fig. 1. Aggregation process of CK denatured in di¡erent concentrations of TFE. The enzyme (2.78 WM) was dissolved in 0.05 M Tris-HCl bu¡er at pH 8.0 containing TFE at di¡erent concentrations. The turbidity was followed using 400 nm light absorption. The TFE concentrations for curves 1^6 were 10, 20, 25, 30, 35 and 40%, respectively.
Fig. 2. Intrinsic £uorescence spectra of native CK, 50% TFE denatured CK and refolded CK. CK was dissolved in 0.05 M Tris-HCl bu¡er at pH 8.0 containing 50% TFE for 1 h. The denatured CK was then diluted 80-fold into the same bu¡er. The ¢nal enzyme concentration was 1.39 WM. 1, native enzyme; 2, 50% TFE denatured enzyme which was referred to as the `TFE state'; 3, refolded enzyme after dilution.
with only slight aggregation observed at 40% TFE. No aggregation was observed at TFE concentrations lower than 5% or higher than 50% (data not shown here). Since aggregation will blur the study of inactivation and unfolding induced by TFE, 50% TFE (v/v) was used to study the CK denaturation process. The intrinsic £uorescence emission spectrum changes of CK after denaturation in 50% TFE are shown in Fig. 2. The maximum £uorescence emission quickly red shifted to 345 nm in about 0.5 min accompanied by a decrease in emission intensity after adding TFE into the bu¡er containing native CK. The unfolding process was so quick that no further changes of the maximum emission and £uorescence intensity occurred after 1 min. The maximum emission of native CK is 333 nm (Fig. 2). The 50% TFE induced CK denaturation state, which was referred to as the `TFE state' [20,28,30,36,41^43], has somewhat lower £uorescence intensity than the native state, showing that the high TFE concentration may quench the tryptophan £uorescence. Secondary structural changes of CK during 50% TFE denaturation were studied using far-ultraviolet CD spectra. Fig. 3 shows the secondary structure changes of CK
BBAPRO 36359 24-1-01
308
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
Fig. 3. Far-ultraviolet CD spectra of CK during denaturation in 50% TFE at di¡erent time intervals. Experimental conditions were as for Fig. 1 except for the TFE concentration. The time intervals for curves 1^5 were 0.3, 0.72, 1.55, 5.45 and 25 min, respectively.
Fig. 4. Time course of relative helical structure changes of CK during denaturation in di¡erent concentrations of TFE. Experimental conditions were as for Fig. 2. The relative helical structure was calculated using the molecular average ellipticity at 222 nm obtained from Fig. 2. The TFE concentrations for curves 1 and 2 were 5% and 50%, respectively.
denatured in 50% TFE solutions at di¡erent time intervals. The molecular average ellipticity increased quickly to a maximum value in about 5.45 min. The marked intensity increase at 222 nm and 208 nm demonstrates that adding 50% TFE solvent increased the amount of helical secondary structure. Similar results were obtained by adding 5% TFE solvent (data not shown here). The molecular average ellipticity at 222 nm was measured at di¡erent time intervals during the denaturation of CK in 50% and 5% TFE solutions (Fig. 4). The amount of helical secondary structure in the TFE state is about 38% more than in native CK with only 6% more helical structure observed in CK denatured in 5% TFE solvent. Previous SEC results [55,56] showed that the DTNB-modi¢ed CK subunit in the denatured state (denatured by 3 M GuHCl and 6 M urea) is still present as a dissociated monomer when diluted into bu¡er not containing DTT, which shows that CK cannot be refolded and remains in a partially folded state. Fig. 5 (curve 3) shows the CK dissociated monomer state during its unfolding in 50% TFE solution.
Fig. 5. SEC elution pro¢les of native CK, 50% TFE denatured modi¢ed CK and refolded CK treated under various conditions. CK, 200 Wl 1.39 WM solution was loaded on a FPLC Sephacryl s200 HR column in 50 mM Tris-HCl bu¡er, pH 8.0. Curves 1^3 were native CK, refolded CK after 80-fold dilution and 50% TFE denatured modi¢ed CK, respectively. The 50% TFE denatured CK was modi¢ed by 10-fold molar excess DTNB in the same bu¡er for 1 h.
BBAPRO 36359 24-1-01
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
309
3.2. Reactivation and refolding of the TFE state When CK denatured in 50% TFE for 1 h was diluted 80-fold into a 0.05 M Tris-HCl bu¡er, pH 8.0, the enzyme activity recovered rapidly to a level of about 38% of its initial activity (Fig. 6). A semilogarithmic plot of the reactivation given in the inset shows that the reactivation followed a biphasic course with the kinetic constants k1 = 4.82 þ 0.41U1033 s31 and k2 = 0.60 þ 0.01U1033 s31 . The CK dimer structure is regained after dilution (Fig. 5) and the intrinsic £uorescence studies showed that the creatine kinase conformation recovered in conjunction with the reactivation. Fig. 7 shows the refolding process of the TFE state indicated by the intrinsic £uorescence maximum intensity measured at di¡erent time intervals with the excitation wavelength 295 nm. The emission intensity ¢rst increased quickly to about 222% of the native CK value and then slowly decreased with the gradual refolding of the molecule, accompanied by a blue shift of the emission maximum. That the dilution caused the emission intensity to very rapidly increase at ¢rst suggests that the TFE molecules may quench the
Fig. 6. Reactivation of 50% TFE denatured CK after 80-fold dilution. 50% TFE denatured CK was diluted into the bu¡er and the ¢nal enzyme concentration was 1.39 WM. The inset shows a semilogarithmic plot. b, experimental points; R, points obtained by subtracting the contribution of the slow phase from the data in curve (999).
Fig. 7. Kinetic course of £uorescence maximum intensity changes during refolding of the TFE state. Experimental conditions were as for Fig. 5. Fluorescence spectra were determined at suitable time intervals. The inset shows a semilogarithmic plot. b, experimental points; R, points obtained by subtracting the contribution of the slow phase from the data in curve (999).
tryptophan £uorescence and their removal is very rapid compared with the refolding process. We also suggest that the TFE molecules are exposed and that removing the bound TFE molecules is not the rate limiting step of the refolding. The partially refolded CK has a maximum emission at 337 nm and relatively higher intensity than the native CK which has a maximum emission at 333 nm (Fig. 2). The maximum emission of the partially refolded CK is close to that of the native enzyme suggesting that the partially refolded CK has a compact structure similar to the native CK, though its speci¢c tertiary structure is quite di¡erent from the native enzyme as shown by the hydrophobic exposure (Fig. 8). A semilogarithmic plot given in the inset (Fig. 7) shows that the kinetic constants of the biphasic refolding process were k1 = 4.34 þ 0.27U1033 s31 and k2 = 0.76 þ 0.02U1033 s31 . The partially refolded CK with about 38% of native activity agrees with the generally recognized opinion that conformational integrity is important for maintaining the enzyme activity and that slight changes at the active site can lead to complete inactivation of an enzyme.
BBAPRO 36359 24-1-01
310
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
Fig. 8. ANS binding £uorescence intensity of native CK, 50% TFE denatured CK and refolded CK. Experimental conditions were as for Fig. 5. The samples were incubated in 40 WM ANS for 30 min. 1, native enzyme; 2, refolded enzyme after dilution; 3, 50% TFE denatured enzyme.
The denatured CK secondary structure was also recovered after dilution (Fig. 9). The molecular average ellipticity at 222 nm representing the amount of helical secondary structure is plotted at di¡erent
Fig. 9. Far-ultraviolet CD spectra of native CK, 50% TFE denatured CK and refolded CK. Experimental conditions were as for Fig. 8. 1, native enzyme; 2, 50% TFE denatured enzyme which was referred to as the `TFE state'; 3, refolded enzyme after dilution.
Fig. 10. Kinetic course of relative helical structure changes during refolding of the TFE state. Experimental conditions were as for Fig. 5. The relative helical structure was calculated using the same method as in Fig. 4. The inset shows a semilogarithmic plot. b, experimental points; R, points obtained by subtracting the contribution of the slow phase from the data in curve (999).
times during the refolding process (Fig. 10). The very quick decrease of helical structure at the beginning of dilution corresponds to the rapid increase of the intrinsic £uorescence intensity shown in Fig. 7. It demonstrates that the TFE molecules are very easily removed and the removal process is more rapid than refolding. The changes of the helical secondary structure during refolding also follow a two phase course with the kinetic constants k1 = 4.50 þ 0.07U1033 s31 and k2 = 1.13 þ 0.05U1033 s31 (Fig. 10 insert). The similarity of the kinetic constants for reactivation, refolding and helical secondary structure recovery indicates that these three processes are related during the dilution of the denatured CK. The partially refolded CK has similar secondary structure content as the native CK (Fig. 9). The ANS binding £uorescence (Fig. 8) shows that the TFE state of CK contains a little more hydrophobic exposure than the native enzyme and the partially refolded CK contains much more hydrophobic exposure than the TFE state and the native enzyme. However, previous ANS binding £uorescence data (data not shown here) have shown that CK denatured in 6 M urea and 3 M GuHCl shows no hydrophobic surface exposure.
BBAPRO 36359 24-1-01
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
4. Discussion TFE is known to have two main e¡ects: decreasing the strength of hydrophobic interactions and increasing the intramolecular hydrogen bonds. Binary mixtures of water with alcohols like TFE denature the tertiary and the quaternary structures of proteins while enhancing their helicity [44^46]. Although the detailed mechanism for the disruption of polypeptide chains by TFE is still not clear, the observed e¡ects are consistent with a general phenomenological description of the action of TFE in terms of increasing the internal H-bonds and modifying the hydrophobic interactions. When CK was denatured by TFE, the solvent induced tertiary structure unfolding, dissociation, and a conformational change from the native structure to the predominantly K-helical TFE state (Figs. 2,3 and 5). The large helical content of the TFE state is evident in the most negative spectra shown in Fig. 3. An absence of stable tertiary structure is inferred in Fig. 2 from the marked red shift of the intrinsic £uorescence spectra. These properties may be compared with those of other proteins in the presence of TFE and similar solvents. For example, residues 17^30 of monellin in 50% ethanol were shown to undergo a transformation from L-structure to K-helix while an overall molten globule-like structure was retained [47]. Another case is egg white lysozyme in 70% (v/v) TFE which exhibited predominantly K-helical structure in the protein regions that were normally K-helical in the native state [48]. It had been shown that TFE induces helical conformation in only those portions of the sequences of proteins that are either helical in the native state or have a propensity to adopt helical conformations [16,49]. However, a previous study [50] of the e¡ects of TFE on an all L-sheet protein, cardiotoxin analogue I (CTX I) isolated from the venom of the Taiwan cobra, showed that the helix structure induced by TFE is `non-speci¢c'. A simple and often accepted explanation for the formation of the highly helical conformation proposed by Thomas and Dill [51] is that the hydrogen bonds between the alcohol-rich medium and the protein become weaker, making the hydrogen bonds within the polypeptide chain comparatively stronger. It was suggested that TFE weakens the hydrophobic interaction within protein chains, resulting in destabilization of their folded state.
311
Recently, the model for the formation and structure of protein aggregates was well reviewed that speci¢c intermolecular interactions between hydrophobic surfaces of structural subunits in partially folded intermediates are responsible for the aggregation [35]. In many cases, alcohol induced denaturation results in stabilization of extended helical rods with the exposed hydrophobic side chains, whereas polar amide groups are shielded from the solvent [36,51]. Based on these previous results and our data, we propose a possible explanation for the aggregation of CK denatured in TFE. Since TFE can induce hydrophobic clusters when stabilizing helical secondary structures, the hydrophobic clusters will increase with increasing TFE concentrations and will cause the partly unfolded states to aggregate. At TFE concentrations as low as 10%, the number of TFE molecules in the solution was not enough to o¡set the aggregation. At higher TFE concentrations greater than 50%, the number of TFE molecules was su¤cient to totally cover the hydrophobic clusters preventing intramolecular condensation. TFE have been routinely used as a membrane mimetic agent in NMR and CD spectroscopic experiments. However, the aggregation results in our study highlight the risk of using TFE at certain concentrations as membrane mimetics. CK contains eight SH groups of cysteine residues, two of which (Cys-282) have long been known to be more reactive than the others and are believed to be essential for enzyme catalysis from results of chemical modi¢cation studies and site directed mutagenesis [52,53]. Recent results showed that the reactive cysteine is located nearer to the ATP binding site than the creatine binding site and may play an important role not in the binding to the transition state analogue but in the conformational changes caused by the transition state analogue [54]. Wang et al. previously reported that, after being modi¢ed by thiol reagent in the denatured state, CK cannot be refolded and remained in a partially folded state and that three thiol groups in CK were essential for refolding [55]. After completely unfolding in 6 M urea, CK with all of its eight thiol groups modi¢ed by DTNB could be only partially refolded and remained in a monomeric state [56]. Based on the above studies, we used DTNB to modify the thiol groups of the TFE state. Measurements using size exclusion chro-
BBAPRO 36359 24-1-01
312
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
matography showed that CK dissociated during the denaturation process and that the TFE state was monomeric (Fig. 5). The TFE state found in our experiments is a di¡erent state from the native state and from the urea fully denatured state. The maximum emission wavelengths for the native CK, urea fully denatured CK and the high helical TFE state are 333 nm, 353 nm and 345 nm, respectively. The molten globule state, which has been observed in the folding of a number of proteins as both a kinetic and equilibrium intermediate [57,58], has the following characteristic conformational properties: it exhibits native-like secondary structure and it is nearly as compact as the native protein, but the speci¢c tertiary structure of the native state is almost completely lost. Recently, the TFE state has been compared with the molten globule intermediates produced in aqueous solutions [28,59]. These studies suggest that the TFE induced open helical structure, in which the interactions between helical segments are weak and many hydrophobic groups are exposed to the solvent, is distinct from the compact molten globule state observed in aqueous solutions, which is stabilized by the weak but signi¢cant interhelical hydrophobic interactions [60,61]. The TFE state in our present study does not possess the characteristics of the molten globule state. It has much more helical secondary structure than the molten globule state and is less compact than the typical molten globule state which is nearly as compact as the native protein from intrinsic £uorescence and far-ultraviolet CD spectra. Fig. 10 shows that the TFE state of CK at 50% TFE does not have signi¢cant ability to bind the hydrophobic dye, ANS, whereas signi¢cant dye binding is a characteristic property of the compact molten globule state [23,36] which has large amounts of hydrophobic surface exposure. From the results, we suggest that the high helical TFE state of CK is not the molten globule intermediate in the CK folding pathway but is a partially unfolded state which has more hydrophobic surface exposure than the native enzyme. On the other hand, the partially refolded CK contains the characteristics of the molten globule state. It exhibits native-like secondary structure (Fig. 9), it is nearly as compact as the native protein (Fig. 2), and it contains large amounts of hydrophobic surface exposure (Fig. 8) which indicates that the spe-
ci¢c tertiary structure of the partially refolded CK is quite di¡erent from the native protein. The results suggest that the conformation of the partially refolded CK resembles that of a molten globule state in the CK folding pathway. Figs. 2,5^7,9 and 10 show that TFE is a reversible denaturant and in the refolding process of TFE denatured CK, the helical secondary structure recovery, tertiary structure recovery and the activity reactivation are cooperative processes as indicated by the similar kinetic constants. The rapid decrease of ellipticity at 222 nm (Fig. 10) after dilution suggests that the interactions between helical segments in the TFE induced open helical structure may be very weak and they cannot be stabilized when there are few TFE molecules in the solution. The binding of TFE on the hydrophobic clusters is not tight and the removal of the bound TFE molecules is not a rate limiting step of the refolding process. The TFE denatured state can be as important as the urea or GuHCl denatured states if we are concerned about the protein folding process in vivo, because the hydrophobic environment for the protein during translocation through membranes in vivo is similar to the environment in the presence of TFE. Acknowledgements The present investigation was supported by National Key Basic Research Speci¢c Foundation of China No. G1999075607.
References [1] Q.Z. Yao, M. Tian, C.L. Tsou, Biochemistry 23 (1984) 2740^2744. [2] Q.Z. Yao, H.M. Zhou, L.X. Hou, C.L. Tsou, Sci. Sin. 25B (1982) 1296^1302. [3] F. Couthon, E. Clottes, C. Ebel, C. Vial, Eur. J. Biochm. 234 (1995) 160^170. [4] E. Clottes, C. Leydier, F. Couthon, O. Marcillat, V. Christian, Biochim. Biophys. Acta 1338 (1997) 37^46. [5] M. Gross, A. Lustig, T. Wallimanm, R. Further, Biochemistry 34 (1995) 10350^10357. [6] S.H. Grossman, J. Pyle, R.J. Steiner, Biochemistry 20 (1981) 6122^6128.
BBAPRO 36359 24-1-01
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313 [7] Q.Z. Yao, M. Tian, C.L. Tsou, Sci. Sin. 28B (1985) 484^493. [8] G.F. Bickersta¡, C. Paterson, N.C. Price, Biochim. Biophys. Acta 621 (1980) 305^314. [9] L.X. Hou, H.M. Zhou, Q.Z. Yao, C.L. Tsou, Acta Biochim. Biophys. Sin. 15 (1983) 393^397. [10] S.H. Grossman, Biochim. Biophys. Acta 785 (1984) 61^67. [11] H.M. Zhou, C.L. Tsou, Biochim. Biophys. Acta 869 (1986) 69^74. [12] M. Goodman, I. Listowsky, J. Am. Chem. Soc. 84 (1962) 3770^3771. [13] F. Blanco, G. Rivas, L. Serrano, Nat. Struct. Biol. 1 (1994) 584^590. [14] A. Jasano¡, A.R. Fersht, Biochemistry 33 (1994) 2129^2135. [15] V. Munoz, L. Serrano, Nat. Struct. Biol. 1 (1994) 399^409. [16] S. Segawa, T. Fukuno, K. Fujiwara, Y. Noda, Biopolymers 31 (1991) 497^509. [17] R.W. Storrs, D. Truckses, D.E. Wemmer, Biopolymers 32 (1992) 1695^1702. [18] D. Hamada, Y. Kuroda, T. Tanaka, Y. Goto, J. Mol. Biol. 254 (1995) 737^746. [19] A. Cammers-Goodwin, T.J. Allen, S.L. Oslick, K.F. McClure, J.H. Lee, D.S. Kemp, J. Am. Chem. Soc. 118 (1996) 3082^3090. [20] N. Schonbrunner, J. Wey, J. Engels, H. Georg, T. Kiefhaber, J. Mol. Biol. 260 (1996) 432^445. [21] D. Hamada, Y. Goto, J. Mol. Biol. 269 (1997) 479^487. [22] F. Chiti, N. Taddei, N. Nuland, F. Magherini, M. Stefani, G. Ramponi, C.M. Dobson, J. Mol. Biol. 283 (1998) 893^ 903. [23] F. Chiti, N. Taddei, P. Webster, D. Hamada, T. Fiaschi, G. Ramponi, C.M. Dobson, Nat. Struct. Biol. 6 (1999) 380^ 387. [24] A. Kentsis, T.R. Sosnick, Biochemistry 37 (1998) 14613^ 14622. [25] H. Lu, M. Buck, S.E. Radford, C.M. Dobson, J. Mol. Biol. 265 (1997) 112^117. [26] E. Zerovnik, R. Virden, R. Jerala, V. Turk, J.P. Waltho, Proteins 32 (1998) 296^303. [27] K.F. Fulton, E.R.G. Main, V. Daggett, S.E. Jackson, J. Mol. Biol. 291 (1999) 429^444. [28] M. Buck, S.E. Radford, C.M. Dobson, Biochemistry 32 (1993) 669^678. [29] L.O. Narhi, J.S. Philo, T. Li, M. Zhang, B. Samal, T. Arakawa, Biochemistry 35 (1996) 11447^11453. [30] K. Gast, D. Zirwer, M. Mu«ller-Frohne, G. Damaschun, Protein Sci. 8 (1999) 625^634. [31] Y.X. Zhang, S.L. Yan, H.M. Zhou, J. Protein Chem. 15 (1996) 631^637. [32] Q.X. Chen, R.Q. Zhang, P.Z. Yang, Y. Li, S.L. Chen, S. Li, Y. Yang, H.M. Zhou, Int. J. Biol. Macromol. 26 (1999) 103^ 107. [33] H.P. Yang, H.M. Zhou, Biochem. Mol. Biol. Int. 43 (1997) 1297^1304.
313
[34] E.R.G. Main, S.E. Jackson, Nat. Struct. Biol. 6 (1999) 831^ 835. [35] A.L. Fink, Fold. Des. 3 (1998) 9^23. [36] K. Shiraki, K. Nishikawa, Y. Goto, J. Mol. Biol. 245 (1995) 180^194. [37] S. Bhattacharjya, P. Balaram, Protein Sci. 6 (1997) 1065^ 1073. [38] J.R. Cort, N.H. Anderson, Biochem. Biophys. Res. Commun. 233 (1997) 687^691. [39] Q.Z. Yao, L.X. Hou, H.M. Zhou, C.L. Tsou, Sci. Sin. 25B (1982) 1186^1193. [40] L. Noda, S.A. Kuby, H. Larby, Methods Enzymol. 2 (1954) 605^610. [41] E. Zerovnik, R. Virden, R. Jerala, L. Kroon-Zitko, V. Turk, J.P. Waltho, Proteins 36 (1999) 205^216. [42] K. Gast, A.F. Cha¡otte, D. Zirwer, Y. Guillou, M. Mu«llerFrohne, C. Cadieux, M. Hodges, G. Damaschun, M. Goldberg, Protein Sci. 6 (1997) 2578^2588. [43] M. Hoshino, Y. Hagihara, D. Hamada, M. Kataoka, Y. Goto, FEBS Lett. 416 (1997) 72^76. [44] A. Miranker, S.E. Radfors, M. Karplus, C.M. Dobson, Nature 349 (1991) 633^636. [45] S.E. Radfors, C.M. Dobson, P.A. Evans, Nature 358 (1992) 302^307. [46] S. Barker, K.H. Mayo, J. Am. Chem. Soc. 113 (1991) 8201^ 8203. [47] P. Fan, C. Bracken, J. Baum, Biochemistry 32 (1993) 1573^ 1582. [48] M. Buck, H. Schwalbe, C.M. Dobson, Biochemistry 34 (1995) 13219^13232. [49] H.J. Dyson, G. Merutka, J.P. Waltho, R.A. Lerner, P.E. Wright, J. Mol. Biol. 226 (1992) 795^817. [50] A.I. Arunkumar, T.K.S. Kumar, C. Yu, Biochim. Biophys. Acta 1338 (1997) 69^76. [51] P.D. Thomas, K.A. Dill, Protein Sci. 2 (1993) 2050^2065. [52] A.R. Thomson, J.W. Eveleigh, B.J. Miles, Nature 203 (1964) 267^269. [53] T.A. Mahowald, Biochemistry 4 (1965) 732^740. [54] Y. Yang, H.M. Zhou, Biochim. Biophys. Acta 1338 (1998) 190^198. [55] H.R. Wang, J.H. Bai, S.Y. Zheng, Z.X. Wang, H.M. Zhou, Biochem. Biophys. Res. Commun. 221 (1996) 174^180. [56] Y. Yang, Y.D. Park, T.W. Yu, H.M. Zhou, Biochem. Biophys. Res. Commun. 259 (1999) 450^454. [57] K. Kuwajima, Proteins Struct. Funct. Genet. 6 (1989) 87^ 103. [58] O.B. Ptitsyn, Adv. Protein Chem. 47 (1995) 83^229. [59] A.T. Alexandrescu, Y.-L. Ng, C.M. Dobson, J. Mol. Biol. 235 (1994) 587^599. [60] Y. Hagihara, Y. Tan, Y. Goto, J. Mol. Biol. 237 (1994) 336^ 348. [61] I. Nishii, M. Kataoka, F. Tokunaga, Y. Goto, Biochemsitry 33 (1994) 4905^4909.
BBAPRO 36359 24-1-01
Reactivation and refolding of rabbit muscle creatine kinase denatured in 2,2,2-tri£uoroethanol solutions Kai Huang a , Yong-Doo Park a , Zhi-Fang Cao a , Hai-Meng Zhou
a;b;
*
a
b
Department of Biological Science and Biotechnology, Tsinghua University, Beijing 100084, People's Republic of China National Laboratory of Biomacromolecules, Institute of Biophysics, Academia Sinica, Beijing 100084, People's Republic of China Received 3 July 2000; received in revised form 12 December 2000; accepted 12 December 2000
Abstract The unfolding and refolding of creatine kinase (ATP:creatine N-phosphotransferase (CK), EC 2.7.3.2) during denaturation and reactivation by trifluoroethanol (TFE) have been studied. Significant aggregation was observed when CK was denatured at TFE concentrations between 10% and 40% (v/v). 50% TFE (v/v) was used to study the denaturation and unfolding of CK. The activity loss of CK was a very quick process, as was the marked conformational changes during denaturation followed by fluorescence emission spectra and far-ultraviolet CD spectra. DTNB modification and size exclusion chromatography were used to find that CK dissociated and was in its monomer state after denaturation with 50% TFE. Reactivation and refolding were observed after 80-fold dilution of the denatured CK into 0.05 M Tris-HCl buffer, pH 8.0. The denatured CK recovered about 38% activity following a two phase course (k1 = 4.82 þ 0.41U1033 s31 , k2 = 0.60 þ 0.01U1033 s31 ). Intrinsic fluorescence maximum intensity changes showed that the refolding process also followed biphasic kinetics (k1 = 4.34 þ 0.27U1033 s31 , k2 = 0.76 þ 0.02U1033 s31 ) after dilution into the proper solutions. The far-ultraviolet CD spectra ellipticity changes at 222 nm during the refolding process also showed a two phase course (k1 = 4.50 þ 0.07U1033 s31 , k2 = 1.13 þ 0.05U1033 s31 ). Our results suggest that TFE can be used as a reversible denaturant like urea and GuHCl. The 50% TFE induced CK denaturation state, which was referred to as the `TFE state', and the partially refolded CK are compared with the molten globule state. The aggregation caused by TFE during denaturation is also discussed in this paper. ß 2001 Elsevier Science B.V. All rights reserved. Keywords: Creatine kinase; Denaturation; Reactivation; Refolding; Aggregation; Hydrophobic surface; Helix induction; Molten globule state; Tri£uoroethanol state
1. Introduction Rabbit MM-creatine kinase (CK) is a dimer, conAbbreviations: CK, creatine kinase; TFE, 2,2,2-tri£uoroethanol ; TFE state, 50% TFE induced CK denaturation state ; CD, circular dichroism; DTNB, 5,5P-dithiobis(2-nitrobenzoic acid); GuHCl, guanidine hydrochloride; ANS, 1-anilinonaphthalene-8sulfonate; SEC, size exclusion chromatography * Corresponding author, at address a. Fax: +86-10-6277-5505; E-mail: [email protected]
sisting of two identical subunits. The activity and conformational changes of CK in the presence of GuHCl and urea have been extensively studied [1^ 7] and upon diluting the denaturant under suitable conditions, both the conformation and the catalytic activity can be quantitatively recovered [8^11]. However, relatively few attempts have been made to study the denaturation and reactivation of CK in organic solutions such as alcohol. 2,2,2-Tri£uoroethanol (TFE)/H2 O mixtures, which
0167-4838 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 0 0 ) 0 0 2 9 3 - 4
BBAPRO 36359 24-1-01
306
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
were introduced by Goodman and Listowsky [12], are known to stabilize secondary structures such as K-helices and L-hairpins [13,14]. Peptides with sequences derived from proteins typically show little or no helix formation in water [15] but can be induced to form helices in TFE/H2 O mixtures [16^19]. TFE solutions have also been used to study partially folded states and equilibrium intermediates [20,21]. More recently, several groups have observed changes in the rates of folding and unfolding upon addition of TFE [22^27]. The denaturation and unfolding process of several proteins in TFE solutions have been studied previously [28^32]. Yang et al. [33] previously reported that CK was inactivated and underwent conformational changes in di¡erent concentrations of TFE. However, the present study shows signi¢cant aggregation of CK in TFE solutions with concentrations between 10% and 40% (v/v). Aggregation was not discovered at TFE concentrations lower than 5% (v/v) or higher than 50% (v/v). TFE can a¡ect several structural and physical properties of the system [34], such as weakening hydrophobic interactions, strengthening intramolecular hydrogen bonds, changing the dielectric constant of the solvent and acting as an osmolyte. TFE induced denaturation results in partly folded states [20,21,30], which may resemble in some features early intermediates during folding and might be prone to aggregation or misfolding [30,35]. During the denaturation of CK in TFE concentrations between 10% and 40%, one possible explanation for the observed aggregation is that the number of TFE molecules at these concentration ranges is not su¤cient to disrupt the hydrophobic core and/or any hydrophobic clusters of the induced partly unfolded intermediates. At concentrations larger than 50%, enough TFE molecules are present to prevent intermolecular condensation of the unfolding intermediates. At low concentrations, TFE does not signi¢cantly perturb the native structure of the protein [28] but does increase the helical content of the peptides in direct correlation with the propensity of the amino acid sequence to form helices [36]. Molten globule-like intermediates during alcohol induced denaturation have been observed for lysozyme [28,37] and myoglobin [38]. A clear distinction between the molten globule-like states and the TFE
states is important when considering the structural changes observed in the presence of TFE. This paper compares the observed TFE state and the molten globule state. We suggest that the TFE state induced by 50% TFE is not the molten globule state in the CK folding pathway. However, the partially refolded CK with high hydrophobic surface exposure, nativelike secondary structure and nearly the same compact structure as the native protein is very like the molten globule state in the CK folding process. The experimental results also suggest that TFE can be used as a kind of reversible denaturant like urea and guanidine hydrochloride (GuHCl) and the helical secondary structure recovery, the tertiary structure recovery and the activity reactivation are cooperative processes with similar kinetic constants. 2. Materials and methods The preparation of rabbit muscle creatine kinase and the assay of creatine kinase activity were as described by Yao et al. [39]. The ¢nal puri¢ed enzyme was homogeneous on polyacrylamide gel electrophoresis in the presence and absence of SDS. Tri£uoroethanol, 5,5-dithiobis(2-nitrobenzoic acid) (DTNB) and 1-anilinonaphthalene-8-sulfonate (ANS) were Sigma products, while creatine and ATP were obtained from Fluka. All other reagents were local products of analytical grade. The enzyme concentration was determined by measuring the absorbance at 280 nm and using the absorption coe¤cient A1% 1cm = 8.8 [40]. Enzyme denaturation was carried out in a solution containing 50% TFE (v/v) in 0.05 M Tris-HCl bu¡er, pH 8.0 for 1 h. Denatured CK modi¢cation was carried out with 10fold molar excess DTNB in 0.05 M Tris-HCl bu¡er, pH 8.0 for 1 h. Reactivation and refolding were carried out after 80-fold dilution of the denatured CK into the same bu¡er without TFE. Fluorescence spectrum measurements were made with a Hitachi 850 spectro£uorimeter. We used an excitation wavelength of 295 nm for the tryptophan £uorescence measurements. Circular dichroism (CD) spectra were recorded on a Jasco 725 spectropolarimeter with a path length of 2 mm. The CD intensity at 222 nm was used to calculate the amount of helical secondary structure. The aggregation measurements
BBAPRO 36359 24-1-01
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
307
were carried out with a Perkin Elmer Lambda Bio U/V spectrophotometer using the absorption wavelength 400 nm. Size exclusion chromatography was carried out with a Sephacryl s200 HR column on a Pharmacia FPLC apparatus. An excitation wavelength of 400 nm was used for ANS binding £uorescence measurements. The samples were incubated in 40 WM ANS for 30 min. All measurements were carried out at 25³C. 3. Results 3.1. Inactivation and unfolding of CK in the presence of 50% TFE (v/v) When CK was added to solutions containing different TFE concentrations in 0.05 M Tris-HCl bu¡er, pH 8.0, signi¢cant aggregation occurred at TFE concentrations between 10% and 40% (Fig. 1). At ¢rst, increasing the TFE concentration resulted in increased aggregation rate and extent. After reaching a maximum rate at a TFE concentration between 20% and 30%, the aggregation rate began to decrease
Fig. 1. Aggregation process of CK denatured in di¡erent concentrations of TFE. The enzyme (2.78 WM) was dissolved in 0.05 M Tris-HCl bu¡er at pH 8.0 containing TFE at di¡erent concentrations. The turbidity was followed using 400 nm light absorption. The TFE concentrations for curves 1^6 were 10, 20, 25, 30, 35 and 40%, respectively.
Fig. 2. Intrinsic £uorescence spectra of native CK, 50% TFE denatured CK and refolded CK. CK was dissolved in 0.05 M Tris-HCl bu¡er at pH 8.0 containing 50% TFE for 1 h. The denatured CK was then diluted 80-fold into the same bu¡er. The ¢nal enzyme concentration was 1.39 WM. 1, native enzyme; 2, 50% TFE denatured enzyme which was referred to as the `TFE state'; 3, refolded enzyme after dilution.
with only slight aggregation observed at 40% TFE. No aggregation was observed at TFE concentrations lower than 5% or higher than 50% (data not shown here). Since aggregation will blur the study of inactivation and unfolding induced by TFE, 50% TFE (v/v) was used to study the CK denaturation process. The intrinsic £uorescence emission spectrum changes of CK after denaturation in 50% TFE are shown in Fig. 2. The maximum £uorescence emission quickly red shifted to 345 nm in about 0.5 min accompanied by a decrease in emission intensity after adding TFE into the bu¡er containing native CK. The unfolding process was so quick that no further changes of the maximum emission and £uorescence intensity occurred after 1 min. The maximum emission of native CK is 333 nm (Fig. 2). The 50% TFE induced CK denaturation state, which was referred to as the `TFE state' [20,28,30,36,41^43], has somewhat lower £uorescence intensity than the native state, showing that the high TFE concentration may quench the tryptophan £uorescence. Secondary structural changes of CK during 50% TFE denaturation were studied using far-ultraviolet CD spectra. Fig. 3 shows the secondary structure changes of CK
BBAPRO 36359 24-1-01
308
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
Fig. 3. Far-ultraviolet CD spectra of CK during denaturation in 50% TFE at di¡erent time intervals. Experimental conditions were as for Fig. 1 except for the TFE concentration. The time intervals for curves 1^5 were 0.3, 0.72, 1.55, 5.45 and 25 min, respectively.
Fig. 4. Time course of relative helical structure changes of CK during denaturation in di¡erent concentrations of TFE. Experimental conditions were as for Fig. 2. The relative helical structure was calculated using the molecular average ellipticity at 222 nm obtained from Fig. 2. The TFE concentrations for curves 1 and 2 were 5% and 50%, respectively.
denatured in 50% TFE solutions at di¡erent time intervals. The molecular average ellipticity increased quickly to a maximum value in about 5.45 min. The marked intensity increase at 222 nm and 208 nm demonstrates that adding 50% TFE solvent increased the amount of helical secondary structure. Similar results were obtained by adding 5% TFE solvent (data not shown here). The molecular average ellipticity at 222 nm was measured at di¡erent time intervals during the denaturation of CK in 50% and 5% TFE solutions (Fig. 4). The amount of helical secondary structure in the TFE state is about 38% more than in native CK with only 6% more helical structure observed in CK denatured in 5% TFE solvent. Previous SEC results [55,56] showed that the DTNB-modi¢ed CK subunit in the denatured state (denatured by 3 M GuHCl and 6 M urea) is still present as a dissociated monomer when diluted into bu¡er not containing DTT, which shows that CK cannot be refolded and remains in a partially folded state. Fig. 5 (curve 3) shows the CK dissociated monomer state during its unfolding in 50% TFE solution.
Fig. 5. SEC elution pro¢les of native CK, 50% TFE denatured modi¢ed CK and refolded CK treated under various conditions. CK, 200 Wl 1.39 WM solution was loaded on a FPLC Sephacryl s200 HR column in 50 mM Tris-HCl bu¡er, pH 8.0. Curves 1^3 were native CK, refolded CK after 80-fold dilution and 50% TFE denatured modi¢ed CK, respectively. The 50% TFE denatured CK was modi¢ed by 10-fold molar excess DTNB in the same bu¡er for 1 h.
BBAPRO 36359 24-1-01
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
309
3.2. Reactivation and refolding of the TFE state When CK denatured in 50% TFE for 1 h was diluted 80-fold into a 0.05 M Tris-HCl bu¡er, pH 8.0, the enzyme activity recovered rapidly to a level of about 38% of its initial activity (Fig. 6). A semilogarithmic plot of the reactivation given in the inset shows that the reactivation followed a biphasic course with the kinetic constants k1 = 4.82 þ 0.41U1033 s31 and k2 = 0.60 þ 0.01U1033 s31 . The CK dimer structure is regained after dilution (Fig. 5) and the intrinsic £uorescence studies showed that the creatine kinase conformation recovered in conjunction with the reactivation. Fig. 7 shows the refolding process of the TFE state indicated by the intrinsic £uorescence maximum intensity measured at di¡erent time intervals with the excitation wavelength 295 nm. The emission intensity ¢rst increased quickly to about 222% of the native CK value and then slowly decreased with the gradual refolding of the molecule, accompanied by a blue shift of the emission maximum. That the dilution caused the emission intensity to very rapidly increase at ¢rst suggests that the TFE molecules may quench the
Fig. 6. Reactivation of 50% TFE denatured CK after 80-fold dilution. 50% TFE denatured CK was diluted into the bu¡er and the ¢nal enzyme concentration was 1.39 WM. The inset shows a semilogarithmic plot. b, experimental points; R, points obtained by subtracting the contribution of the slow phase from the data in curve (999).
Fig. 7. Kinetic course of £uorescence maximum intensity changes during refolding of the TFE state. Experimental conditions were as for Fig. 5. Fluorescence spectra were determined at suitable time intervals. The inset shows a semilogarithmic plot. b, experimental points; R, points obtained by subtracting the contribution of the slow phase from the data in curve (999).
tryptophan £uorescence and their removal is very rapid compared with the refolding process. We also suggest that the TFE molecules are exposed and that removing the bound TFE molecules is not the rate limiting step of the refolding. The partially refolded CK has a maximum emission at 337 nm and relatively higher intensity than the native CK which has a maximum emission at 333 nm (Fig. 2). The maximum emission of the partially refolded CK is close to that of the native enzyme suggesting that the partially refolded CK has a compact structure similar to the native CK, though its speci¢c tertiary structure is quite di¡erent from the native enzyme as shown by the hydrophobic exposure (Fig. 8). A semilogarithmic plot given in the inset (Fig. 7) shows that the kinetic constants of the biphasic refolding process were k1 = 4.34 þ 0.27U1033 s31 and k2 = 0.76 þ 0.02U1033 s31 . The partially refolded CK with about 38% of native activity agrees with the generally recognized opinion that conformational integrity is important for maintaining the enzyme activity and that slight changes at the active site can lead to complete inactivation of an enzyme.
BBAPRO 36359 24-1-01
310
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
Fig. 8. ANS binding £uorescence intensity of native CK, 50% TFE denatured CK and refolded CK. Experimental conditions were as for Fig. 5. The samples were incubated in 40 WM ANS for 30 min. 1, native enzyme; 2, refolded enzyme after dilution; 3, 50% TFE denatured enzyme.
The denatured CK secondary structure was also recovered after dilution (Fig. 9). The molecular average ellipticity at 222 nm representing the amount of helical secondary structure is plotted at di¡erent
Fig. 9. Far-ultraviolet CD spectra of native CK, 50% TFE denatured CK and refolded CK. Experimental conditions were as for Fig. 8. 1, native enzyme; 2, 50% TFE denatured enzyme which was referred to as the `TFE state'; 3, refolded enzyme after dilution.
Fig. 10. Kinetic course of relative helical structure changes during refolding of the TFE state. Experimental conditions were as for Fig. 5. The relative helical structure was calculated using the same method as in Fig. 4. The inset shows a semilogarithmic plot. b, experimental points; R, points obtained by subtracting the contribution of the slow phase from the data in curve (999).
times during the refolding process (Fig. 10). The very quick decrease of helical structure at the beginning of dilution corresponds to the rapid increase of the intrinsic £uorescence intensity shown in Fig. 7. It demonstrates that the TFE molecules are very easily removed and the removal process is more rapid than refolding. The changes of the helical secondary structure during refolding also follow a two phase course with the kinetic constants k1 = 4.50 þ 0.07U1033 s31 and k2 = 1.13 þ 0.05U1033 s31 (Fig. 10 insert). The similarity of the kinetic constants for reactivation, refolding and helical secondary structure recovery indicates that these three processes are related during the dilution of the denatured CK. The partially refolded CK has similar secondary structure content as the native CK (Fig. 9). The ANS binding £uorescence (Fig. 8) shows that the TFE state of CK contains a little more hydrophobic exposure than the native enzyme and the partially refolded CK contains much more hydrophobic exposure than the TFE state and the native enzyme. However, previous ANS binding £uorescence data (data not shown here) have shown that CK denatured in 6 M urea and 3 M GuHCl shows no hydrophobic surface exposure.
BBAPRO 36359 24-1-01
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
4. Discussion TFE is known to have two main e¡ects: decreasing the strength of hydrophobic interactions and increasing the intramolecular hydrogen bonds. Binary mixtures of water with alcohols like TFE denature the tertiary and the quaternary structures of proteins while enhancing their helicity [44^46]. Although the detailed mechanism for the disruption of polypeptide chains by TFE is still not clear, the observed e¡ects are consistent with a general phenomenological description of the action of TFE in terms of increasing the internal H-bonds and modifying the hydrophobic interactions. When CK was denatured by TFE, the solvent induced tertiary structure unfolding, dissociation, and a conformational change from the native structure to the predominantly K-helical TFE state (Figs. 2,3 and 5). The large helical content of the TFE state is evident in the most negative spectra shown in Fig. 3. An absence of stable tertiary structure is inferred in Fig. 2 from the marked red shift of the intrinsic £uorescence spectra. These properties may be compared with those of other proteins in the presence of TFE and similar solvents. For example, residues 17^30 of monellin in 50% ethanol were shown to undergo a transformation from L-structure to K-helix while an overall molten globule-like structure was retained [47]. Another case is egg white lysozyme in 70% (v/v) TFE which exhibited predominantly K-helical structure in the protein regions that were normally K-helical in the native state [48]. It had been shown that TFE induces helical conformation in only those portions of the sequences of proteins that are either helical in the native state or have a propensity to adopt helical conformations [16,49]. However, a previous study [50] of the e¡ects of TFE on an all L-sheet protein, cardiotoxin analogue I (CTX I) isolated from the venom of the Taiwan cobra, showed that the helix structure induced by TFE is `non-speci¢c'. A simple and often accepted explanation for the formation of the highly helical conformation proposed by Thomas and Dill [51] is that the hydrogen bonds between the alcohol-rich medium and the protein become weaker, making the hydrogen bonds within the polypeptide chain comparatively stronger. It was suggested that TFE weakens the hydrophobic interaction within protein chains, resulting in destabilization of their folded state.
311
Recently, the model for the formation and structure of protein aggregates was well reviewed that speci¢c intermolecular interactions between hydrophobic surfaces of structural subunits in partially folded intermediates are responsible for the aggregation [35]. In many cases, alcohol induced denaturation results in stabilization of extended helical rods with the exposed hydrophobic side chains, whereas polar amide groups are shielded from the solvent [36,51]. Based on these previous results and our data, we propose a possible explanation for the aggregation of CK denatured in TFE. Since TFE can induce hydrophobic clusters when stabilizing helical secondary structures, the hydrophobic clusters will increase with increasing TFE concentrations and will cause the partly unfolded states to aggregate. At TFE concentrations as low as 10%, the number of TFE molecules in the solution was not enough to o¡set the aggregation. At higher TFE concentrations greater than 50%, the number of TFE molecules was su¤cient to totally cover the hydrophobic clusters preventing intramolecular condensation. TFE have been routinely used as a membrane mimetic agent in NMR and CD spectroscopic experiments. However, the aggregation results in our study highlight the risk of using TFE at certain concentrations as membrane mimetics. CK contains eight SH groups of cysteine residues, two of which (Cys-282) have long been known to be more reactive than the others and are believed to be essential for enzyme catalysis from results of chemical modi¢cation studies and site directed mutagenesis [52,53]. Recent results showed that the reactive cysteine is located nearer to the ATP binding site than the creatine binding site and may play an important role not in the binding to the transition state analogue but in the conformational changes caused by the transition state analogue [54]. Wang et al. previously reported that, after being modi¢ed by thiol reagent in the denatured state, CK cannot be refolded and remained in a partially folded state and that three thiol groups in CK were essential for refolding [55]. After completely unfolding in 6 M urea, CK with all of its eight thiol groups modi¢ed by DTNB could be only partially refolded and remained in a monomeric state [56]. Based on the above studies, we used DTNB to modify the thiol groups of the TFE state. Measurements using size exclusion chro-
BBAPRO 36359 24-1-01
312
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313
matography showed that CK dissociated during the denaturation process and that the TFE state was monomeric (Fig. 5). The TFE state found in our experiments is a di¡erent state from the native state and from the urea fully denatured state. The maximum emission wavelengths for the native CK, urea fully denatured CK and the high helical TFE state are 333 nm, 353 nm and 345 nm, respectively. The molten globule state, which has been observed in the folding of a number of proteins as both a kinetic and equilibrium intermediate [57,58], has the following characteristic conformational properties: it exhibits native-like secondary structure and it is nearly as compact as the native protein, but the speci¢c tertiary structure of the native state is almost completely lost. Recently, the TFE state has been compared with the molten globule intermediates produced in aqueous solutions [28,59]. These studies suggest that the TFE induced open helical structure, in which the interactions between helical segments are weak and many hydrophobic groups are exposed to the solvent, is distinct from the compact molten globule state observed in aqueous solutions, which is stabilized by the weak but signi¢cant interhelical hydrophobic interactions [60,61]. The TFE state in our present study does not possess the characteristics of the molten globule state. It has much more helical secondary structure than the molten globule state and is less compact than the typical molten globule state which is nearly as compact as the native protein from intrinsic £uorescence and far-ultraviolet CD spectra. Fig. 10 shows that the TFE state of CK at 50% TFE does not have signi¢cant ability to bind the hydrophobic dye, ANS, whereas signi¢cant dye binding is a characteristic property of the compact molten globule state [23,36] which has large amounts of hydrophobic surface exposure. From the results, we suggest that the high helical TFE state of CK is not the molten globule intermediate in the CK folding pathway but is a partially unfolded state which has more hydrophobic surface exposure than the native enzyme. On the other hand, the partially refolded CK contains the characteristics of the molten globule state. It exhibits native-like secondary structure (Fig. 9), it is nearly as compact as the native protein (Fig. 2), and it contains large amounts of hydrophobic surface exposure (Fig. 8) which indicates that the spe-
ci¢c tertiary structure of the partially refolded CK is quite di¡erent from the native protein. The results suggest that the conformation of the partially refolded CK resembles that of a molten globule state in the CK folding pathway. Figs. 2,5^7,9 and 10 show that TFE is a reversible denaturant and in the refolding process of TFE denatured CK, the helical secondary structure recovery, tertiary structure recovery and the activity reactivation are cooperative processes as indicated by the similar kinetic constants. The rapid decrease of ellipticity at 222 nm (Fig. 10) after dilution suggests that the interactions between helical segments in the TFE induced open helical structure may be very weak and they cannot be stabilized when there are few TFE molecules in the solution. The binding of TFE on the hydrophobic clusters is not tight and the removal of the bound TFE molecules is not a rate limiting step of the refolding process. The TFE denatured state can be as important as the urea or GuHCl denatured states if we are concerned about the protein folding process in vivo, because the hydrophobic environment for the protein during translocation through membranes in vivo is similar to the environment in the presence of TFE. Acknowledgements The present investigation was supported by National Key Basic Research Speci¢c Foundation of China No. G1999075607.
References [1] Q.Z. Yao, M. Tian, C.L. Tsou, Biochemistry 23 (1984) 2740^2744. [2] Q.Z. Yao, H.M. Zhou, L.X. Hou, C.L. Tsou, Sci. Sin. 25B (1982) 1296^1302. [3] F. Couthon, E. Clottes, C. Ebel, C. Vial, Eur. J. Biochm. 234 (1995) 160^170. [4] E. Clottes, C. Leydier, F. Couthon, O. Marcillat, V. Christian, Biochim. Biophys. Acta 1338 (1997) 37^46. [5] M. Gross, A. Lustig, T. Wallimanm, R. Further, Biochemistry 34 (1995) 10350^10357. [6] S.H. Grossman, J. Pyle, R.J. Steiner, Biochemistry 20 (1981) 6122^6128.
BBAPRO 36359 24-1-01
K. Huang et al. / Biochimica et Biophysica Acta 1545 (2001) 305^313 [7] Q.Z. Yao, M. Tian, C.L. Tsou, Sci. Sin. 28B (1985) 484^493. [8] G.F. Bickersta¡, C. Paterson, N.C. Price, Biochim. Biophys. Acta 621 (1980) 305^314. [9] L.X. Hou, H.M. Zhou, Q.Z. Yao, C.L. Tsou, Acta Biochim. Biophys. Sin. 15 (1983) 393^397. [10] S.H. Grossman, Biochim. Biophys. Acta 785 (1984) 61^67. [11] H.M. Zhou, C.L. Tsou, Biochim. Biophys. Acta 869 (1986) 69^74. [12] M. Goodman, I. Listowsky, J. Am. Chem. Soc. 84 (1962) 3770^3771. [13] F. Blanco, G. Rivas, L. Serrano, Nat. Struct. Biol. 1 (1994) 584^590. [14] A. Jasano¡, A.R. Fersht, Biochemistry 33 (1994) 2129^2135. [15] V. Munoz, L. Serrano, Nat. Struct. Biol. 1 (1994) 399^409. [16] S. Segawa, T. Fukuno, K. Fujiwara, Y. Noda, Biopolymers 31 (1991) 497^509. [17] R.W. Storrs, D. Truckses, D.E. Wemmer, Biopolymers 32 (1992) 1695^1702. [18] D. Hamada, Y. Kuroda, T. Tanaka, Y. Goto, J. Mol. Biol. 254 (1995) 737^746. [19] A. Cammers-Goodwin, T.J. Allen, S.L. Oslick, K.F. McClure, J.H. Lee, D.S. Kemp, J. Am. Chem. Soc. 118 (1996) 3082^3090. [20] N. Schonbrunner, J. Wey, J. Engels, H. Georg, T. Kiefhaber, J. Mol. Biol. 260 (1996) 432^445. [21] D. Hamada, Y. Goto, J. Mol. Biol. 269 (1997) 479^487. [22] F. Chiti, N. Taddei, N. Nuland, F. Magherini, M. Stefani, G. Ramponi, C.M. Dobson, J. Mol. Biol. 283 (1998) 893^ 903. [23] F. Chiti, N. Taddei, P. Webster, D. Hamada, T. Fiaschi, G. Ramponi, C.M. Dobson, Nat. Struct. Biol. 6 (1999) 380^ 387. [24] A. Kentsis, T.R. Sosnick, Biochemistry 37 (1998) 14613^ 14622. [25] H. Lu, M. Buck, S.E. Radford, C.M. Dobson, J. Mol. Biol. 265 (1997) 112^117. [26] E. Zerovnik, R. Virden, R. Jerala, V. Turk, J.P. Waltho, Proteins 32 (1998) 296^303. [27] K.F. Fulton, E.R.G. Main, V. Daggett, S.E. Jackson, J. Mol. Biol. 291 (1999) 429^444. [28] M. Buck, S.E. Radford, C.M. Dobson, Biochemistry 32 (1993) 669^678. [29] L.O. Narhi, J.S. Philo, T. Li, M. Zhang, B. Samal, T. Arakawa, Biochemistry 35 (1996) 11447^11453. [30] K. Gast, D. Zirwer, M. Mu«ller-Frohne, G. Damaschun, Protein Sci. 8 (1999) 625^634. [31] Y.X. Zhang, S.L. Yan, H.M. Zhou, J. Protein Chem. 15 (1996) 631^637. [32] Q.X. Chen, R.Q. Zhang, P.Z. Yang, Y. Li, S.L. Chen, S. Li, Y. Yang, H.M. Zhou, Int. J. Biol. Macromol. 26 (1999) 103^ 107. [33] H.P. Yang, H.M. Zhou, Biochem. Mol. Biol. Int. 43 (1997) 1297^1304.
313
[34] E.R.G. Main, S.E. Jackson, Nat. Struct. Biol. 6 (1999) 831^ 835. [35] A.L. Fink, Fold. Des. 3 (1998) 9^23. [36] K. Shiraki, K. Nishikawa, Y. Goto, J. Mol. Biol. 245 (1995) 180^194. [37] S. Bhattacharjya, P. Balaram, Protein Sci. 6 (1997) 1065^ 1073. [38] J.R. Cort, N.H. Anderson, Biochem. Biophys. Res. Commun. 233 (1997) 687^691. [39] Q.Z. Yao, L.X. Hou, H.M. Zhou, C.L. Tsou, Sci. Sin. 25B (1982) 1186^1193. [40] L. Noda, S.A. Kuby, H. Larby, Methods Enzymol. 2 (1954) 605^610. [41] E. Zerovnik, R. Virden, R. Jerala, L. Kroon-Zitko, V. Turk, J.P. Waltho, Proteins 36 (1999) 205^216. [42] K. Gast, A.F. Cha¡otte, D. Zirwer, Y. Guillou, M. Mu«llerFrohne, C. Cadieux, M. Hodges, G. Damaschun, M. Goldberg, Protein Sci. 6 (1997) 2578^2588. [43] M. Hoshino, Y. Hagihara, D. Hamada, M. Kataoka, Y. Goto, FEBS Lett. 416 (1997) 72^76. [44] A. Miranker, S.E. Radfors, M. Karplus, C.M. Dobson, Nature 349 (1991) 633^636. [45] S.E. Radfors, C.M. Dobson, P.A. Evans, Nature 358 (1992) 302^307. [46] S. Barker, K.H. Mayo, J. Am. Chem. Soc. 113 (1991) 8201^ 8203. [47] P. Fan, C. Bracken, J. Baum, Biochemistry 32 (1993) 1573^ 1582. [48] M. Buck, H. Schwalbe, C.M. Dobson, Biochemistry 34 (1995) 13219^13232. [49] H.J. Dyson, G. Merutka, J.P. Waltho, R.A. Lerner, P.E. Wright, J. Mol. Biol. 226 (1992) 795^817. [50] A.I. Arunkumar, T.K.S. Kumar, C. Yu, Biochim. Biophys. Acta 1338 (1997) 69^76. [51] P.D. Thomas, K.A. Dill, Protein Sci. 2 (1993) 2050^2065. [52] A.R. Thomson, J.W. Eveleigh, B.J. Miles, Nature 203 (1964) 267^269. [53] T.A. Mahowald, Biochemistry 4 (1965) 732^740. [54] Y. Yang, H.M. Zhou, Biochim. Biophys. Acta 1338 (1998) 190^198. [55] H.R. Wang, J.H. Bai, S.Y. Zheng, Z.X. Wang, H.M. Zhou, Biochem. Biophys. Res. Commun. 221 (1996) 174^180. [56] Y. Yang, Y.D. Park, T.W. Yu, H.M. Zhou, Biochem. Biophys. Res. Commun. 259 (1999) 450^454. [57] K. Kuwajima, Proteins Struct. Funct. Genet. 6 (1989) 87^ 103. [58] O.B. Ptitsyn, Adv. Protein Chem. 47 (1995) 83^229. [59] A.T. Alexandrescu, Y.-L. Ng, C.M. Dobson, J. Mol. Biol. 235 (1994) 587^599. [60] Y. Hagihara, Y. Tan, Y. Goto, J. Mol. Biol. 237 (1994) 336^ 348. [61] I. Nishii, M. Kataoka, F. Tokunaga, Y. Goto, Biochemsitry 33 (1994) 4905^4909.
BBAPRO 36359 24-1-01