- Email: [email protected]
Unfolding and refolding of the reduced constant fragment of the immunoglobulin light chain
J. Mol. Biol.
Unfolding
(1982)
156, 911-926
and Refolding of the Reduced Constant Fragment of the Immunoglobulin Light Chain Kinetic
Role of the Intrachain YUJI GOTO AND Kozo
Disulfide
Bond
HAMACUCHI
Department of Biology, Faculty of Science Osaka University, Toyonaka, Osaka 560, Japan (Received
3 November
1981)
The constant fragment of the immunoglobulin light chain whose intrachain disulfide bond is reduced (reduced CL fragment) assumes a conformation very similar to the intact CL fragment (Goto 6 Hamaguchi, 1979). The kinetics of reversible unfolding and refolding of the reduced CL fragment by guanidine hydrochloride at pH 7.5 and 25°C were studied and were compared with those of the intact CL fragment described in the accompanying paper (Goto & Hamaguchi. 1982). Tryptophyl fluorescence, far-ultraviolet circular dichroism, and reactivity of the SH groups toward 5,5’-dithiobis+nitrobenzoic acid) were used to follow the kinetics. The results obtained were thoroughly explained on the basis of the threespecies mechanism, U, +Uz--N, where U, and U, are slow-folding and fastfolding species, respectively, of unfolded protein and N is native protein. The rate constants of interconversion between U, and U, and the rate constant for the process from N to UZ were very similar to the respective values for the intact C, fragment. Only the rate constant for the process from U2 to S was greatly different between the intact and reduced CL fragments: the rate constant for the reduced C, fragment was about 100 times smaller than that for the intact CL fragment. These results indicated that the slow isomerization of the unfolded molecule is independent of the presence of the disulfide bond and that the kinetic role of the intrachain disulfide bond is to accelerate the folding process. This kinetic role in the folding of the CL fragment was explainable only in terms of the decreased entropy in the unfolded state of the intact CL fragment due to the presence of the disulfide bond.
1. Introduction l&h domain of the immunoglobulin molecule contains only one intrachain disulfide bond buried in the interior of the molecule and the loop formed by the disulfide bond consists of about 60 residues of the total of about 100 amino acid residues (Beale & Feinstein, 1976; Amzel & Poljak, 1979). To clarify the role of the disulfide bond in the folding of the immunoglobulin domain will thus provide a basis for understanding the role of the intrachain disulfide bond in the folding of the protein. Based on this idea, we have carried out a series of studies using the constant (C,) fragment of the immunoglobulin light chain and the CL fragment
91”
I’.GOTO
AND
K.
HAMAGI’C’HI
whose intrachain disulfide bond is reduced (reduced C’,, fragment) (Goto et al., 197!): Goto 8: Hamaguchi. 1979,1981.1982). We first compared the conformation and stability between t,he intact (‘L fragment, and the reduced (‘L fragment?. and found that the reduced (“L fragment assumes a conformation very similar to the intact CYLfragment in the absence of denat,urant. but the stability of the former is lower than that of the latter (Goto CY Hamaguchi. 1979). The decreased stability on reduction of the disulfide bond is almost entirely explained in terms of the larger entropy of the reduced CL fragment compared with the entropy of the intact CL fragment in the denatured state. We then studied the formation of the intrachain disulfide bond from the reduced (‘L, fragment in the presence of glutathione (Goto CyrHamaguchi, 1981). We showed that the unfolding of the molecule to expose the two SH groups is necessary to form the disulfide bond in the reduced CL fragment and estimated the rates of unfolding and refolding of the molecule using the reaction rates of the SH groups with glutathiom,. UY describe the kinetics of the reversible In the accompanying paper. denaturation of the intact CL fragment by GuHCl. The kinetics inside and abo\rta the unfolding transition zone were consistently explained on the basis of the three species mechanism, but the refolding kinetics below t’he transition zone indicated the formation of an intermediate containing non-native proline residues. In the present paper. we describe the kinetics of unfolding and refolding of the reduced (‘[, fragment. Because the conformations of the intact and reduced CL fragments are very similar and the C, fragment contains only one intrachain disulfide bond. we expected that, the comparison of the kinetics of mlfolding and refolding of both proteins would give us useful and distinct information of t,he kinet#ic rcile of thr intrachain disulfide bond in the folding of the CL fragment.
2. Materials (a)
and Methods Matrrials
The C’, f’ragment of Bence Jones protein Kiag (type A) was obtained h.v digestion with papain as described (Goto & Hamaguchi, 1979). The reduced CL fragment was prepared by reduction of the intrachain disulfide bond of the CL fragment with 20 mwdithiothreitol in the presence of 4 wGuH(‘I at pH 8 and then separated from the residual reagents on a column of Sephadex (:-2.5 equilibrated with an acetate lmffer at pH 5 (Qoto h Hamaguchi, 1979). GuH(‘I (specially puritied grade), D’I’IYB, and other reagents were obtained from h-akarai Chemicals (‘0. and were used without further purification.
(1)) Kin&
mrasurrmptrts
All measurements were carried out in 035 wTris EDTA. and a given concentration of GuHCI at degassed. For measurements of fast kinetics a used. For measurements of slow kinetics manual were observed either by tryptophyl fluorescence t Ahbreriations c,.d.. c~iwular~
used: dkhroism
reduced : 1)TSK.
HCI pH 7.5 Union mixing or cd.
(“L fragment. CqL fragment .~..S’-dithiol)is-(Z-llitrol,cllzoic
buffer containing 0.15 IV-KU, 1 mMand 25°C’. 811 the buffers used were Giken stopped-flow apparatus was was made in a cell and the kinetics The details of the measurements and
whose intrachain a&l): (:uH(‘I.
disultide guanidinr
bond is reduced: h~tiroc~hlolid~,.
FOLDING
KINETICS
OF REDr(‘ED
dat’a analysis are described in the accompanying was obsrrv~~d after each kinetic measurement.
P(L FRAGMENT
paper. So formation
of thr disulfidc
!) I3 bond
(c) Titration of SH groups with DTSR The reactivities of the two SH groups of the reduced CL fragment toward DTNB at various concentrations of UuHCl were used to follow the protein unfolding reaction. To a 2.5 ml solution of the reduced CL fragment in 0.05 iv-Tris.HCl buffer containing 015 M-KU, 1 rnMEDTA. and a given concentration of GuHCl at pH 7.5 was added @5 ml of freshly prepared 3 rnM-DTNB solution in the same buffer containing the same concentration of GuHCl, and its absorbance at 412 nm was measured against a freshly prepared reagent blank with a Hitachi model 323 spectrophotometer. All the buffers used were degassed. The molar tlstincbtion efficient of reduced DTNR was assumed to be 13.600 M-l cm-’ at 114 nm range of 0 to 0.7 ~(iuHC!l where the ((&tithing 8 Davidson, 1972) in the concentration prt~srnt rxperimcnts were rarried out. (d) Protein cvncentratiot/ Th(x concentration of the reduced CL fragment was determined spectrophotometrically using a value of Ai)m = 14.5 at 280 nm (Goto & Hamaguchi, 1979). pH measurement lvith a l%adiomet,er l’HJl2Bc meter at 25.C. (e)
pH was measured
3. Results Previously we reported the reversible equilibrium of the unfolding of the reduced (‘L fragment by GuH(‘1 as measured by c.d. at 218 nm (Goto & Hamaguchi. 1979). The fluorescence spectrum of the reduced CL fragment has a maximum at’ 330 nm and the intensity is much higher compared with the intensity for the intact CiL fragment. This is due to the quenching effect of the intrachain disulfide bond on the tryptophyl fluorescence being removed on reduction of the disulfide bond. Figure 1 shows the denaturation curve of the reduced C, fragment as measured by t’ryptophyl fluorescence using the stopped-flow apparatus. The unfolding transition was reversible and had a midpoint at 0.4 M-GuHCl. This unfolding curve was the same as those obtained by measuring fluorescences at 350 and 400 nm using the Hitachi fluorescence spectrophotometer (not shown) and was also the same as that obtained by c.d. at 218 nm (Goto & Hamaguchi. 1979). Because the change in the fluorescence at around 360 nm upon unfolding of the reduced CL fragment was small. the kinetic measurements made by manual mixing were mainly done using t)he fluorescence at 400 nm, where a large increase was observed upon unfolding.
(b)
[Trrfolditzg
kinetics
of
the reducrd
C,, fragment
The kinetics of unfolding of the reduced CL fragment were measured by fluorescence at 400 nm using manual mixing below 0.8 M-GuHCl and by fluorescence at wavelengths longer than 330 nm using stopped-flow mixing above 0.7 M-GuHCl. The overall process of the unfolding follows first-order kinetics above
914
Y.GOTO
0
ASI)
I
K.
H,4MAGIT(‘HI
2 3 [Gu HCll ( M)
4
PIG:. 1, Equilibrium and kinetic data for the denaturation of the reduced CvL fragment by (:uH(‘I. pH 7.5. 25°C. Measurements were carried out by fluorescence using a stopped-flow apparatus at a protein concentration of 092 mg/ml. The initial concentrations of GuHCl for unfolding (0) and refolding (0) equilibria were 0 M and 4 ~1. respectively. The initial points of the s1ow and fast phases of unfolding kinetics are shown by A and A. respectively. They were obtained by extrapolating the respective phases to time zero. Initial conditions for unfolding kinetics: 0 M-GuHCl. protein concentration. 0.04 mg/ml. The continuous line indicates the theoretical curve for the unfolding transition constructed using the equation proposed by Tanford (1970) (see Got,rl & Hamaguchi. 1979).
1.5 M-GuH(‘I. exponential
Below 12 w-GuHCl, decay terms :
however,
F(t)-F(rX,)=F,exp(-hlt)+Fzexp(-Xzt).
the kinetics
were
described
by two
(1)
where F(t) is the fluorescence at time t, F( CO) is the fluorescence after equilibration, /\1 and h, are the apparent rate constants of t’he slow and fast phases. respectively. and F, and F, are the amplitudes of the respective phases. We denote the amplitudes of the slow and fast phases relative to the total change as ml1 and x2. respectively. where x1 -t- t2 = 1. Figure 2 shows the dependence on GuHCl concentration of the apparent rate constants (h, and h2) of the slow and fast phases and the amplitude (a2) of the fast phase. The initial points of the fast, and slow phases. which were obtained b) extrapolation to time zero of the respective phases. are shown in Figure 1. As can be seen. the initial points of the fast phase fell on a line expected for the dependence on GuHCl concentration of the fluorescence of native protein. This shows that the total unfolding process was seen in these experiments. The apparent rate constants obtained using manual mixing agreed with those of the slow phases obtained using stopped-flow mixing. As shown in Figure 2. the apparent rate constant (AI) of the slow phase increased with an increase in GuHCl concentration and tended to level off above 1 M-(:uH(‘l. The apparent rate constant (h2) of the fast phase. though we determined it onI> above 07 M-GuHCl. increased with an increase in GuHCI concentration. The
FOLDING
KINETICS
OF
REDUCED
C, FRAGMENT
915
(b)
0
I
I 2 [Gu HCll
I 3
I 4
(M)
FIN:. 2. The dependence on GuHCl concentration of (a) the apparent rate constants (A, and h2) and (b) the relative amplitude (1s) of the fast phase for unfolding and refolding kinetics of the reduced CL fragment, pH 75, 25°C. (0) From unfolding kinetics obtained by stopped-flow fluorescence measurements. The conditions are as described in the legend to Fig. 1. (a) From unfolding kinetics measured by fluorescence at 400 nm using manual mixing. Initial conditions: protein concentration, 1 mg/ml; 0 M-GuHCI. Final conditions: protein concentration, 902 mg/ml and the indicated concentration of GuHCl. (A) From refolding kinetics measured by fluorescence at 409 nm using manual mixing. Initial conditions: protein concentration, 1 mg/ml; 4 M-GuHCl. Final conditions: protein concentration, 902 mg/ml; the indicated concentration of GuHCl. (0) From refolding kinetics measured by c.d. at 218 nm. Initial conditions: protein concentration, 7 mg/ml; 4 M-GuHCI. Final conditions: protein concentration. 92 mg/ml; 91 M-GuHCI. ,( x ) The apparent rate constant of the isomerization process of Uz to U, in mechanism 1 after unfolding measured by double-jump experiments. The conditions are as described in the legend to Fig. 3. The dots below 0.7 M-GuHCI show the values of A, calculated using the microscopic rate constants in Fig. 5 (see the text). The open square shows the value of AZ calculated using the result of the double-jump experiment (see the text). The continuous line in (b) indicates the value of a2 calculated using K1, = 10 and the value of fN obtained from the equilibrium measurement (see the text).
amplitude (Q) of the fast phase increased from about @2 at @7 M-GuHCl to 1.0 above 1.5 M-GuHCl. The values of h, and its dependence on GuHCl concentration were very similar to those for the intact CL fragment, except that the curve for the reduced CL fragment shifts by about 1 M to a lower concentration region of GuHCl (see Fig. 3(a) of the accompanying paper). As explained for the unfolding of the 31
916
Y.GOTO
AND
K. HAMAGUCHI
Frc. 3. A double-jump experiment measuring the process of U, to U, in mechanism 1 after unfolding, pH 75, 25°C. Formation of the U, species was monitored by taking samples at the indicated times of unfolding and measuring the refolding kinetics by fluorescence at 400 nm (0) or at 350 nm (0). All the mixing procedures were done manually. The ordinate shows the fluorescence change for the slow phase observed in the refolding process relative to the maximum fluorescence change obtained at infinite time in the unfolding conditions. Initial conditions: 0 M-GuHCl; protein concentration, 2.5 mg/ml. Unfolding conditions: 4 M-GuHCI ; protein concentration, 1.2 mg/ml. Refolding conditions : @07 M-GuHCl ; protein concentration, O-02 mg/ml. The continuous line indicates the theoretical curve with a rate constant of 3 x lo-* s-l.
intact C, fragment in the accompanying paper, the results of Figure 2 suggests that the unfolding kinetics of the reduced CL fragment also follows the three-species mechanism : U +,+
(mechanism
1)
where N is native protein, U, and U2 are the two forms of unfolded protein, and krz, k,,, kz3 and k,, are the rate constants for the respective processes. In the accompanying paper, we have demonstrated the presence of two forms of the intact CL fragment by the double-jump method (Brandts et al., 1975; Nall et al., 1978). This method was also used here to confirm whether two forms are present in the unfolded molecule of the reduced CL fragment. The result is shown in Figure 3. The reduced CL fragment was first denatured by 4 M-GuHCl, and then refolding was initiated at 096 M at various times in the denaturing conditions. Unfolding of the reduced CL fragment in 4 M-GuHCl was complete within one second. As shown in Figure 3, the amplitude of the slow phase of refolding was dependent on the time of exposure of the protein to 4 M-GuHCl. This indicates that there exist the fastrefolding species (U,) and slow-refolding species (U,) in the unfolded reduced CL molecule and that the former converts slowly to the latter in 4 M-GuHCl. The apparent rate constant for the interconversion between U, and U, was estimated to be 3~1O-~s-‘, which is very similar to that for the intact CL fragment. As shown in Figure 3, however, the slow phase did not disappear and amounted to 20% of the total amplitude even at time zero. This will be discussedlater. As in the case of the intact CL fragment, the rate of the interconversion estimated from the double-jump experiments agreed well with the values of h, obtained from the
FOLDING
KINETICS
OF
REDUCED
C, FRAGMENT
917
unfolding experiments at high concentrations of GuHCl. These results obtained by the double-jump experiments strongly suggest that the unfolding kinetics of the reduced C, fragment follow the three-species mechanism, in which the equilibrium constant Kzl ( = k,,/k12) is independent of GuHCl concentration.
(c) Refolding
kinetics
of the reduced
C,, fragment
Refolding kinetics of the reduced CL fragment from 41v1-GuHCl measured by fluorescencesat 330,350 and 400 nm were all the same. The kinetics were very slow and the observable process (more than 90% of the total change) followed first-order kinetics. The amplitudes of the fast phasesat 400 nm, estimated assuming that the kinetics consist of two phases, are included in Figure 2(b). The apparent rate constants determined by fluorescence at 400 nm are shown in Figure Z(a). They agreed well with the apparent rate constants of the slow phasesdetermined by the unfolding experiments. The agreement of the kinetic parameters obtained from unfolding kinetics with those from refolding kinetics establishes the kinetic reversibility of the folding transition of the reduced CL fragment. The kinetics of refolding of the reduced CL fragment at 91 M-GuHCl from 2 MGuHCl were also studied by circular dichroism at 218 and 230 nm. The refolding was very slow and the apparent rate constant was found to be 3 x 10-j s-r. which is independent of the wavelength used and is the same as the rate constant determined by fluorescence measurement (Fig. 2(a)). The amplitude of the slow phase determined by c.d. at 218 and 230 nm was about 90% of the total change of the c.d. estimated by extending to lower GuHCl concentrations the linear portion corresponding to the unfolded state in the transition curve. This relative amplitude was the same as that estimated by fluorescence measurement (Fig. 2(b)). The agreement of the refolding kinetics obtained by fluorescence and cd. measurements suggests that there is no intermediate in the refolding process. This is consistent with the three-species mechanism. The large amplitude of the slow phase at very low concentrations of GuHCl and the low values of A, at very low concentrations of GuHCl compared with the rate constant of the isomerization from UZ to U, obtained by the double-jump experiments suggest that the proportion of U, to the total unfolded molecule is large.
(d) Titration
of SH groups
qf the
reduced
C,‘ fragment
Reduced CL fragment has two SH groups. These SH groups are buried in the interior of the molecule and are inaccessible to SH reagents (Goto & Hamaguchi, 1979,1981). Previously we studied the formation of the intrachain disulfide bond from the reduced CL fragment in the presence of glutathione and showed that the reaction of the disulfide formation is explained in terms of the two-state model, in which the SH groups can react freely with the reagent in the unfolded molecule but not in the native molecule (Goto & Hamaguchi, 1981). We also studied” the reactivities of the SH groups toward DTNB (Goto & Hamaguchi, 1979) and showed that the time-course of the reaction consists of a rapid phase and a slow phase and the rate constant (1 x 10m3s-r) of the latter phase is comparable with the rate
918
Y.GOTO
1 0
AND
I I
K.
I
2 [Gu
HAMAGI’CHI
I
I
I
3
4
5
HO.1 (MI
FIG. 4. The dependence on GuHCl concentration of the apparent determined by SH titration of the reduced CL fragment by DTNH. reduced CL fragment at a protein concentration of @06 mg/ml were reaction was monitored by the absorbance change at 412 nm. The shown in Fig. 2.
rate constant (0) for the slow phase pH 7.5. 25°C. The SH groups of the titrated with 0% mM-DTNB and the triangles represent the values of hZ
(6~1O-~s-’ ) of the unfolding estimated from the kinetics of the disulfide formation. On the basis of these findings it was expected that the reactivities of the SH groups can be used for studying the conformational change of the reduced CL fragment. We studied the reactivities of the SH groups of the reduced CL fragment toward DTNB at various concentrations of GuHCl. Titration of the SH groups with DTNB at various concentrations was first examined in the absence of denaturant. Between 0.1 and 25 m&f-DTNB, all the titration processes gave the same kinetics ; 5% of the total SH groups first reacted rapidly within the dead time of the measurement, (20 s) and the remaining 95% reacted slowly with a rate constant of 1 x 1O-3 s-‘. This shows that the intramolecular process of the reduced CL fragment is ratelimiting and not the reaction of the exposed SH groups with DTNB. The titration of the SH groups with 05 mM-DTNB was carried out at various concentrations of GuHCl. All the titration processes consisted of two phases but the rate constant and the amplitude of the slow phase changed markedly depending on GuHCl concentration. At the highest concentration of GuHCl used (0.7 M), the titration of the SH groups was examined by changing DTNB concentration three times, but the kinetics was independent of DTNB concentration, indicating that even at 0.7 M-GuHCl the intramolecular process of the reduced CL fragment is ratelimiting in the titration of the SH groups. The dependence on GuHCl concentration of the rate constant of the slow phase and the amplitude of the fast phase are shown in Figures 4 and 5(b), respectively. The rate constant of the slow phase increased markedly with increasing GuHCl concentration. These rate constants are on a line obtained by extending the values of hz determined at higher concentrations of GuHCl. The amplitude of the fast phase increased sharply with increasing GuHCl concentration (Fig. 5(b)). This change is the same as the unfolding
FOLDING
KINETICS
OF
REDUCED
I
I
CL FRAGMENT
919
I
(b)
0
I
2
3
CGu HCII
(M)
4
5
FIN:. 5. The dependence on GuHCl concentration of (a) the microscopic rate constants estimated based on mechanism 1 for folding of the reduced CL fragment and (b) the transition curve for unfolding obtained from kinetic measurements. pH 7.5. 25°C. (a) (0) k,,; (a) k,,; (A) kzl; (A) k,*. (b) (0) The amplitude of the fast phase determined by SH titration of the reduced CL fragment by DTNB. The conditions are as described in the legend to Fig. 4. (0) The fraction of the unfolded species calculated using the 4 microscopic rate constants. The continuous line indicates the transition curve for unfolding obtained from equilibrium measurement.
transition curve obtained by fluorescence or c.d. measurement. These results are explained in terms of mechanism 1. Because the reaction of the exposed SH groups with DTNB is very fast in the presence of excess DTNB (Whitesides et al., 1977), the amplitude of the fast phase should correspond to the proportion of the total amount of the unfolded species (U, +U,) and the amplitude of the slow phase should correspond to the proportion of the folded species (N). Therefore the change with GuHCl concentration in the amplitude of the fast phase should represent the unfolding transition of the molecule. The titration of the buried SH groups of the native molecule is thus rate-limited by the unfolding process and the rate of titration is equal to the rate of the unfolding of the reduced CL molecule (ks2). The value of A2 at a high GuHCl concentration represents the value of kS2 (see Discussion), and the rate constants obtained by SH titration are on a line extending
920
Y.GOTOANDK.HAMAGUCHI
the values of h2 at higher concentrations rate constant obtained by SH titration
of GuHCl. This confirms further that the represents the value of IL,, in mechanism 1.
4. Discussion (a) Folding
mechanism
of the reduced
CJragment
We have described the kinetics of unfolding and refolding of the reduced CL fragment as measured by tryptophyl fluorescence, cd. and by reactivities of the SH groups toward DTNB. All the kinetic data obtained here are found to be consistent with the three-species mechanism, U, --UZ+N. As described in the accompanying paper, an intermediate is formed during the refolding process of the intact CL fragment below the transition zone but not inside the zone. In the refolding of the reduced CL fragment, however, we did not observe anv intermediate. As shown in Figure 1. the unfolding transition of the reduced Ci. fragment begins at a very low GuHCl concentration and thus we could study the refolding only in the transition zone. If we could have studied the refolding of the reduced C, fragment far below the transition zone, we might have found an intermediate during the refolding process. The rate constants in mechanism 1 can be determined essentially the same way as described in the accompanying paper. The value of (lCr2+k2r) for the interconversion between IT, and U2 was determined to be 3 x 10m2 s-’ by the double-jump experiment. If the value of A, at a very low GuHCl concentration is assumed to be equal to k, 2, the value of K,, (= k,,/k,,) is estimated to be about 10. Since the unfolding transition of the reduced CL fragment begins at a very low GuHCl concentration, it is uncertain whether the value of A, at a low GuHCl concentration is exactly the same as the value of k,,. However, we tentatively assumed the value of K,, to be 10. This value is the same as the value of K21 for the intact fragment. The value of k32 was directly obtained from the titration of the SH groups with DTNB (Fig. 4). The value of k,, is obtained by the relation: 4,
= k,,(l
+K21)(1
-fd/fn.
(2)
using the values off, (the proportion of the total unfolded molecule) determined by the titration of the SH groups with DTNB (Fig. 5(b)). Although only the values of hi are available below 67 M-GuHCl, the values of k,, and k,, can be calculated with equations (3) and (4) of the accompanying paper and the known values of k,,, k32, fD and A, and the assumed value of K,, Between 0.7 and 1.2 M-GuHCl. the values of A,, A2 and a2 are available and thus the four microscopic rate constants in mechanism 1 can be determined with equations (3) to (6) of the accompanying paper. In the region far above the transition zone, the value of A2 is equal to k,,, because k23 should be small. The four microscopic rate constants thus obtained at various concentrations of GuHCl are shown in Figure 5(a). In Figure 5(b) the values offn in the region of 0.7 to 1.2 M-GuHCl, calculated using the kinetic parameters. are shown. It can be seen that they fall on the equilibrium curve obtained by fluorescence measurement. With increasing concentration of GuHCl, k2, decreases and k,, increases sharply. This is expected for the rates of conformational transition of proteins. The values of
FOLDING
KINETICS
OF
REDUCED
CL FRAGMENT
92 1
kZ3 and k,, are the same at 0.8 M-GuHCl, which is higher by 64 M-GuHCl than the apparent midpoint of the equilibrium transition. This difference is due to the presence of the two forms of the unfolded species and is discussed for the intact CL fragment in the accompanying paper. The rates of the isomerization process (k,, and k,,) are independent of GuHCl concentration. These results confirm the validity of the above assumption that K,, equals to 10 and is independent of GuHCl concentration. The values of hz below 96 M-GuHCl, where the values of hz could not be measured directly by our procedure, can be calculated using the above microscopic rate constants (see equation (2) of the accompanying paper). The values of h2 thus calculated are given in Figure 2. The calculated hz values agreed well with the value obtained experimentally at 0.7 M-GuHCl and increase with decreasing GuHCl concentration. The dependence of h2 on GuHCl concentration has a minimum at’ about 0.8 M-GuHCl. The appearance of the minimum is expected from mechanism 1 (see the accompanying paper). The continuous line in Figure 2(b)shows the values of a2 expected under the limiting conditions where h2/h, $ 1 (see equation (9) of the accompanying paper). The observed values of a2 inside the transition zone are smaller than these values in the limiting conditions. A similar observation has been reported for ribonuclease A (Brandts et al., 1975 ; Hagerman & Baldwin, 1976). Hagerman (1977) has discussed the amplitude of the fast phase in mechanism 1. which is expressed as a function of the ratio of the two apparent rate constants (h,/h,) at fixed values offN and K,,, and shown that under the conditions where the ratio of AZ/h1 is not so large, a2 is expected to be smaller than the value of a2 under the limiting conditions. As can be seen from Figure 2(a), the separation of h2 and h, at about 1 M-GuHCl is actually not so large. In the case of the intact CL fragment. h2 is well separated from h, at all the concentrations of GuHCl studied and the observed values of or2, except below the transition zone where an intermediate emerges. are nearly equal to the values of a2 under the limiting conditions. In the double-jump experiments (Fig. 3), the slow phase of refolding did not disappear and amounted to about 2OyA of the total change at time zero. The relation between the relative amplitude and the ratio of the two apparent rate constants (/\2/h,) is extended to the analysis of the results obtained by the doublejump experiments. The relative amplitude of the fast phase at time zero obtained by the double-jump experiment is expressed as a function of h2/h, by: (3) where In equation (3), fi is the proportion of the native molecule at the start of unfolding and f; is the proportion of the native molecule when refolding is complete under specified conditions. This equation can be applied to the case where the unfolding of N to U2 is complete and refolding is initiated prior to conversion from U2 to U,. This equation predicts that, in the refolding conditions where fk is nearly equal to j”& cy2 approaches 1 when the ratio of h,/h, is extremely high and l2
022
Y.GOTO
ASI)
Ii.
HAMAGI-(‘HI
hecomes smaller than 1, i.e. the slow phase does not disappear is not so large compared with h,. We analyzed the result of the double-jump experiment est,imated t’he apparent rate constant of the fast phase conditions where the value of h, was difficult to determine R2r = 10. f& = 995. fi = 090 and l\2 = 04 in equation (3). obtained for X,/h,. When a value of 3 x 1V3 s- ‘. which was the refolding experiments. is used for h,, we obtain a value of hz. This value agrees quite well with the value obtained rate constants (Fig. 2).
(b) tlolr
of the intrnchuin
disulfide
at time zero when A2 using equation (3) and (h2) in the refolding experimentally. ITsing a value of about 50 is obtained directly from of 0.15 s- ’ as the value using the microscopic
baud
We no\v obtained the four microscopic rate constants in the three-species mechanism for the folding kinetics of the reduced (‘L fragment. As reported in the accompanying paper, we also obtained the microscopic rate constants for the intact CL fragment under the conditions where the kinetics follow the three-species mechanism. We compare the rate constants for the reduced CYLfragment (Fig. 5) with those for the intact CL fragment (Fig. 9 of the accompanying paper) and consider the role of the intrachain disulfide bond in the folding of the CL fragment. The rate constants of the isomerization in the unfolded state (kIZ and k,,) are the same for both proteins and the slow-refolding species amounts to about 90% of the total unfolded species for both proteins. This indicates that the two forms of the unfolding species are formed irrespective of the presence of the intrachain disulfide bond. The two forms of the unfolded species tnust, have been formed by the isomerization of proline residues, and the isomerization process for the reduced CL fragment is the same as that for the intact C, fragment. The role of a particular prolinr residue in the folding of the intact CL fragment is discussed in detail in the accompanying paper. The rate constant (kj2) of unfolding of the reduced (:, fragment is only slightly greater than that, of the intact C, fragment. Only the rate constant (kz3) of folding from the fast-folding species to native protein differs greatly between the two proteins. Although the rate constants kz3 for both proteins were obtained in different concentration ranges of GuHCl because of their different stabilities. t)he marked difference of k,, is apparent. When we compare t,he values of k,, at 1 .O MGuH(‘1. where the rate constants for both proteins were determined experimentally. the rate constant for the intact, (1, fragment is about 100 t,imes greater than that, for the reduced rL fragment. The large increase in the rate constant of refolding by the presence of the intrachain disulfide bond indicates thus importance of the disulfide bond in the direct folding transition of the CL fragment As reported previously (Goto & Hamaguchi. 1979), the stability of the (I,> fragment decreases upon reduction of the intrachain disulfide bond. This is due mostly to the decreased rate constant (kz3) from I’, to K for the reduced VL fragment. Knowledge of the process from U2 to N is important if we are to understand the folding of the C, fragment and we consider t,his process in detail. W’e estimated the free energy cohange of denaturation of the CL fragment using a
FOLDING
KINETICS
OF
REDUCED
923
C, FRAGMENT
two-state approximation for the equilibrium transition (Goto & Hamaguchi, 1979). As described in the accompanying paper, however, when two forms of the unfolded species (U, and U,) are present, the apparent equilibrium constant (K,,,) of native protein to the total unfolded protein may be expressed by: K a,,,, = (W,)+(U,)V(pu’)
= K32(l+K21).
(4)
where K32
= P’,)/(N)
= k&m
Thus the apparent free energy change of denaturation ( - RT In K,,,) is smaller by RT In (1+ Kzl) than the true free energy change (-RT In K32). The term RT in (1 + K21) is calculated to be 1.4 kcal mol-’ using K,, = 10 at 25°C for the intact and reduced CL fragments. We first constructed the free-energy diagrams of the folding and unfolding of the intact and reduced CL fragments at 1.0 M-GuHCl, where the rate constants of unfolding and refolding (k23 and k32) were determined experimentally for both proteins. The transition state free energy, A@. is calculated using the relation: A# = - RT In (kh/k,T),
(-5)
where k is the rate constant, h is Planck’s constant, kB is Boltzmann’s constant, T is t,he temperature in Kelvin, and R is the gas constant. Figure 6(a) shows the profiles at 25°C constructed by setting the native states for both proteins at the same energy level. The free energy change of denaturation estimated using the rate constants of unfolding and refolding is 2.0 kcal mol- ’ for the intact C, fragment and - 1.2 kcal mol - ’ for the reduced CL fragment. These values are in good agreement with those calculated using equation (4) and the equilibrium transition data (Goto & Hamaguchi. 1979).
u--.-. (a)
20
1
lntoct
CL
(b) Intact CL
Reduced
19.4
Reduced
21.8
CL
CL
21.5
19.1
2.0
_-------_-
N
L;-
-1.2
N
N
PI<:. 6. Free-energy profiles for the unfolding and refolding of the intact and reduced Cl,. fragments in the presence of 1 M-GuHCl and (b) in the absence of GuHCI. pH 7.5. BR’C. X and 11 represent native and denatured states. respectively. The energy lrvels of the native states of the intact reduced C, fragments were set to zero (see the text).
(a) the and
924
Y.GOTO
AND
K.
HAMAGUCHI
Figure 6(b) shows the free-energy profiles in the absence of GuHCl. The profile for the reduced CL fragment can be constructed directly using the values of k, 3 and Ic,, Because the values of kZ3 and k32 for the intact CL fragment in the absence of GuHCl were not determined, we constructed the profile assuming that the difference in ks2 in the absence of GuHCl between the intact CL fragment and reduced CL fragment is the same as that in the presence of 1.0 RI-GuHCl, and using the true free energy change of denaturation in the absence of GuHCl, which is calculated with equation (4). and the equilibrium transition data reported previously (Goto & Hamaguchi, 1979). The rate constant of refolding of the intact CL fragment is estimated to be about 80 s-l from this energy profile. When the free energies for the intact and reduced CL fragments are assumed to be the same in the native state, the energy levels of the transition states for both proteins are very similar and only the energy levels in the denatured state are different between the two proteins. Previously we showed that the decrease in the stability of the (1, fragment on reduction of the disulfide bond is consistently explained in terms of the increase in the entropy of the denatured state (Goto Br Hamaguchi, 1979). Provided that this is the case, the free-energy profile shows that the change in the refolding rate is due entirely to the change in the entropy of the denatured state. By the presence of the disulfide bond, the energy level of the denatured state is increased by about 4 kcal mol-‘, which is due to the decreased entropy of the denatured state. This results in an increase of about 100 times in the refolding rate. The other cont’ribution of the disulfide bond to the folding kinetics of the CL fragment. if any, may be small compared with the entropic effect, and we may thus conclude that the kinetic role of the disulfide bond in the folding of the C, fragment is largely entropic. This may be applicable to the folding of other proteins with several disulfide bonds. The energy levels of the denatured states are different between the intact and reduced VL fragment only because of the different, entropies of the denatured proteins but the energy levels of the transition states are the same for both proteins. This means that the presence of the disulfide bond does not affect the ratelimiting step for the conformational transition of the CL fragment. Several mechanisms have been proposed for protein folding (see, for instance; Baldwin. 1978.1980). The problem as to the rate-limiting step in the folding pathway is still controversial. The results obtained here suggest t,he location of the rate-limit,ing step when the folding process is measured by the compactness of the molecule. If the rate-limiting step of the folding of the C, fragment is located near the denatured state. the transition states of the t,wo proteins may be in different energy levels because the loop formed by t,he disulfide bond is large and thus the entropies of the two proteins in the early stage of folding may be different. If we assume that the intact and reduced CL fragments fold through the same folding pathway, the same energy level for the transition states may be realized only when a compact conformation, for which there is no difference in the conformational entropy between the intact and reduced CL fragments. is formed at a later stage of folding. The unfolding rates for the intact and reduced CL fragments are expected to be the same for the transition state with such a compact conformation. On t’he basis of his detailed kinetic data for the reduction and oxidation of the
FOLDING
KINETIC8
OF REDUCED
C, FRAGMENT
925
three disulfide bonds in bovine pancreatic trypsin inhibitor, Creighton (1977,1978,1980u) has constructed the energy profile of the entire intramolecular transition in the unfolding and refolding of the protein. He has shown that the transition is very co-operative and its rate-limiting step is very close to the native state. The kinetics of formation of the disulfide bonds in ribonuclease A obtained by a similar method (Creighton, 1979) and the results of unfolding and refolding of several proteins obtained by urea gradient gel electrophoresis (Creighton. 19806) are also consistent with this idea. Segawa et al. (1973) studied the kinetics of thermal unfolding of lysozyme and suggested that in the transition state the molecule folds tightly, though many intramolecular hydrogen bonds are broken. Although a further study is necessary to define the rate-limiting step in the folding of the CL fragment, our results are consistent with these suggestions. On the basis of the kinetic data for the disulfide formation in the reduced CL fragment, we estimated the rates of unfolding and refolding of the reduced CL fragment in the absence of denaturant (Goto & Hamaguchi, 1981). Whereas the unfolding rate obtained (6 x 10T4 s-l) is in fair agreement with that obtained in the present experiments (1 x lo- 3 s-l ), the refolding rate estimated from the disulfide formation (3 x 10m2s-l) is considerably smaller than the value (918 s-l) estimated here. In the analysis of the kinetic data for the disulfide formation, we used a twostate approximation instead of the three-species mechanism. Therefore, in addition to the true rate of the refolding of U2 to N, the rate of the slow isomerization processof U, to U, is involved in the rate of refolding estimated from the kinetics of the disulfide formation. On the other hand, the same process of the unfolding of the native molecule can be seen both in the kinetics of the disulfide formation and in the kinetics described here, and thus both kinetic data give the same rate constant of unfolding. This work was supported by grants for scientific research from the Ministry Science, and Culture of Japan and a grant from the HoanshaFoundation.
of Education,
REFERENCES Amzel, L. M. & Poljak, R. L. (1979). Annu. Rev. Biochem. 48, 961-997. Baldwin, R. L. (1978). Trends Biochem. Sci. 3, 66-68. Baldwin, R. L. (1980). In Protein Folding (Jaenicke, R., ed.), pp. 369-384, Elsevier/North Holland, Amsterdam. Beale, D. & Feinstein, A. (1976). Quart. Rev. Biophys. 9, 135180. Brandts, H. F., Halvorson, H. R. & Brennan, M. (1975). Biochemistry, 14, 4953-4963. Creighton, T. E. (1977). J. Mol. Biol. 113, 295-312. Creighton, T. E. (1978). Prog. Biophys. Mol. Biot. 33, 231-297. Creighton, T. E. (1979). J. Mot. Biot. 129, 411-431. Creighton, T. E. (198Ou). In Protein Folding (Jaenicke, R., ed.), pp. 427-441, Elsevier/North Holland, Amsterdam. Creighton, T. E. (1980b). J. Mot. Biol. 137, 61-80. Gething, M. J. H. & Davidson, D. E. (1972). Eur. J. B&hem. 39, 352-353. Goto, Y. & Hamaguchi, K. (1979). J. Biochem. 86, 1433-1441. Goto, Y. & Hamaguchi, K. (1981). J. Mol. Biol. 146, 321-340. Goto, Y. & Hamaguchi, K. (1982). J. Mol. Biol. 156. W-910. Goto, Y., Azuma, T. & Hamaguchi, K. (1979). J. Biochem. 85, 1427-1438. Hagerman, P. J. (1973). Biopotymers, 16, 731-747.
9%
Y.GOTO ANI) K.HAMAGCTCHI
Hagerman, P. J. & Baldwin, R. L. (1976). Biochemistry, 15, 1462-1473. Nail, B. T.. Garel, J.-R. & Baldwin, R. L. (1978). J. Mol. Biol. 118, 317-330. Segawa, 8., Husimi, Y. & Wada, A. (1973). Biopolymers, 12. 2521-2537. Tanford, C. (1970). Advun. Protein Ghem. 24, l-95. Whitesides, G. M., Lilburn, J. E. & Szajewski, R. P. (1977). J. Ory. Chem. 42, 332-338 Edited
by S. Brenner
Unfolding
(1982)
156, 911-926
and Refolding of the Reduced Constant Fragment of the Immunoglobulin Light Chain Kinetic
Role of the Intrachain YUJI GOTO AND Kozo
Disulfide
Bond
HAMACUCHI
Department of Biology, Faculty of Science Osaka University, Toyonaka, Osaka 560, Japan (Received
3 November
1981)
The constant fragment of the immunoglobulin light chain whose intrachain disulfide bond is reduced (reduced CL fragment) assumes a conformation very similar to the intact CL fragment (Goto 6 Hamaguchi, 1979). The kinetics of reversible unfolding and refolding of the reduced CL fragment by guanidine hydrochloride at pH 7.5 and 25°C were studied and were compared with those of the intact CL fragment described in the accompanying paper (Goto & Hamaguchi. 1982). Tryptophyl fluorescence, far-ultraviolet circular dichroism, and reactivity of the SH groups toward 5,5’-dithiobis+nitrobenzoic acid) were used to follow the kinetics. The results obtained were thoroughly explained on the basis of the threespecies mechanism, U, +Uz--N, where U, and U, are slow-folding and fastfolding species, respectively, of unfolded protein and N is native protein. The rate constants of interconversion between U, and U, and the rate constant for the process from N to UZ were very similar to the respective values for the intact C, fragment. Only the rate constant for the process from U2 to S was greatly different between the intact and reduced CL fragments: the rate constant for the reduced C, fragment was about 100 times smaller than that for the intact CL fragment. These results indicated that the slow isomerization of the unfolded molecule is independent of the presence of the disulfide bond and that the kinetic role of the intrachain disulfide bond is to accelerate the folding process. This kinetic role in the folding of the CL fragment was explainable only in terms of the decreased entropy in the unfolded state of the intact CL fragment due to the presence of the disulfide bond.
1. Introduction l&h domain of the immunoglobulin molecule contains only one intrachain disulfide bond buried in the interior of the molecule and the loop formed by the disulfide bond consists of about 60 residues of the total of about 100 amino acid residues (Beale & Feinstein, 1976; Amzel & Poljak, 1979). To clarify the role of the disulfide bond in the folding of the immunoglobulin domain will thus provide a basis for understanding the role of the intrachain disulfide bond in the folding of the protein. Based on this idea, we have carried out a series of studies using the constant (C,) fragment of the immunoglobulin light chain and the CL fragment
91”
I’.GOTO
AND
K.
HAMAGI’C’HI
whose intrachain disulfide bond is reduced (reduced C’,, fragment) (Goto et al., 197!): Goto 8: Hamaguchi. 1979,1981.1982). We first compared the conformation and stability between t,he intact (‘L fragment, and the reduced (‘L fragment?. and found that the reduced (“L fragment assumes a conformation very similar to the intact CYLfragment in the absence of denat,urant. but the stability of the former is lower than that of the latter (Goto CY Hamaguchi. 1979). The decreased stability on reduction of the disulfide bond is almost entirely explained in terms of the larger entropy of the reduced CL fragment compared with the entropy of the intact CL fragment in the denatured state. We then studied the formation of the intrachain disulfide bond from the reduced (‘L, fragment in the presence of glutathione (Goto CyrHamaguchi, 1981). We showed that the unfolding of the molecule to expose the two SH groups is necessary to form the disulfide bond in the reduced CL fragment and estimated the rates of unfolding and refolding of the molecule using the reaction rates of the SH groups with glutathiom,. UY describe the kinetics of the reversible In the accompanying paper. denaturation of the intact CL fragment by GuHCl. The kinetics inside and abo\rta the unfolding transition zone were consistently explained on the basis of the three species mechanism, but the refolding kinetics below t’he transition zone indicated the formation of an intermediate containing non-native proline residues. In the present paper. we describe the kinetics of unfolding and refolding of the reduced (‘[, fragment. Because the conformations of the intact and reduced CL fragments are very similar and the C, fragment contains only one intrachain disulfide bond. we expected that, the comparison of the kinetics of mlfolding and refolding of both proteins would give us useful and distinct information of t,he kinet#ic rcile of thr intrachain disulfide bond in the folding of the CL fragment.
2. Materials (a)
and Methods Matrrials
The C’, f’ragment of Bence Jones protein Kiag (type A) was obtained h.v digestion with papain as described (Goto & Hamaguchi, 1979). The reduced CL fragment was prepared by reduction of the intrachain disulfide bond of the CL fragment with 20 mwdithiothreitol in the presence of 4 wGuH(‘I at pH 8 and then separated from the residual reagents on a column of Sephadex (:-2.5 equilibrated with an acetate lmffer at pH 5 (Qoto h Hamaguchi, 1979). GuH(‘I (specially puritied grade), D’I’IYB, and other reagents were obtained from h-akarai Chemicals (‘0. and were used without further purification.
(1)) Kin&
mrasurrmptrts
All measurements were carried out in 035 wTris EDTA. and a given concentration of GuHCI at degassed. For measurements of fast kinetics a used. For measurements of slow kinetics manual were observed either by tryptophyl fluorescence t Ahbreriations c,.d.. c~iwular~
used: dkhroism
reduced : 1)TSK.
HCI pH 7.5 Union mixing or cd.
(“L fragment. CqL fragment .~..S’-dithiol)is-(Z-llitrol,cllzoic
buffer containing 0.15 IV-KU, 1 mMand 25°C’. 811 the buffers used were Giken stopped-flow apparatus was was made in a cell and the kinetics The details of the measurements and
whose intrachain a&l): (:uH(‘I.
disultide guanidinr
bond is reduced: h~tiroc~hlolid~,.
FOLDING
KINETICS
OF REDr(‘ED
dat’a analysis are described in the accompanying was obsrrv~~d after each kinetic measurement.
P(L FRAGMENT
paper. So formation
of thr disulfidc
!) I3 bond
(c) Titration of SH groups with DTSR The reactivities of the two SH groups of the reduced CL fragment toward DTNB at various concentrations of UuHCl were used to follow the protein unfolding reaction. To a 2.5 ml solution of the reduced CL fragment in 0.05 iv-Tris.HCl buffer containing 015 M-KU, 1 rnMEDTA. and a given concentration of GuHCl at pH 7.5 was added @5 ml of freshly prepared 3 rnM-DTNB solution in the same buffer containing the same concentration of GuHCl, and its absorbance at 412 nm was measured against a freshly prepared reagent blank with a Hitachi model 323 spectrophotometer. All the buffers used were degassed. The molar tlstincbtion efficient of reduced DTNR was assumed to be 13.600 M-l cm-’ at 114 nm range of 0 to 0.7 ~(iuHC!l where the ((&tithing 8 Davidson, 1972) in the concentration prt~srnt rxperimcnts were rarried out. (d) Protein cvncentratiot/ Th(x concentration of the reduced CL fragment was determined spectrophotometrically using a value of Ai)m = 14.5 at 280 nm (Goto & Hamaguchi, 1979). pH measurement lvith a l%adiomet,er l’HJl2Bc meter at 25.C. (e)
pH was measured
3. Results Previously we reported the reversible equilibrium of the unfolding of the reduced (‘L fragment by GuH(‘1 as measured by c.d. at 218 nm (Goto & Hamaguchi. 1979). The fluorescence spectrum of the reduced CL fragment has a maximum at’ 330 nm and the intensity is much higher compared with the intensity for the intact CiL fragment. This is due to the quenching effect of the intrachain disulfide bond on the tryptophyl fluorescence being removed on reduction of the disulfide bond. Figure 1 shows the denaturation curve of the reduced C, fragment as measured by t’ryptophyl fluorescence using the stopped-flow apparatus. The unfolding transition was reversible and had a midpoint at 0.4 M-GuHCl. This unfolding curve was the same as those obtained by measuring fluorescences at 350 and 400 nm using the Hitachi fluorescence spectrophotometer (not shown) and was also the same as that obtained by c.d. at 218 nm (Goto & Hamaguchi. 1979). Because the change in the fluorescence at around 360 nm upon unfolding of the reduced CL fragment was small. the kinetic measurements made by manual mixing were mainly done using t)he fluorescence at 400 nm, where a large increase was observed upon unfolding.
(b)
[Trrfolditzg
kinetics
of
the reducrd
C,, fragment
The kinetics of unfolding of the reduced CL fragment were measured by fluorescence at 400 nm using manual mixing below 0.8 M-GuHCl and by fluorescence at wavelengths longer than 330 nm using stopped-flow mixing above 0.7 M-GuHCl. The overall process of the unfolding follows first-order kinetics above
914
Y.GOTO
0
ASI)
I
K.
H,4MAGIT(‘HI
2 3 [Gu HCll ( M)
4
PIG:. 1, Equilibrium and kinetic data for the denaturation of the reduced CvL fragment by (:uH(‘I. pH 7.5. 25°C. Measurements were carried out by fluorescence using a stopped-flow apparatus at a protein concentration of 092 mg/ml. The initial concentrations of GuHCl for unfolding (0) and refolding (0) equilibria were 0 M and 4 ~1. respectively. The initial points of the s1ow and fast phases of unfolding kinetics are shown by A and A. respectively. They were obtained by extrapolating the respective phases to time zero. Initial conditions for unfolding kinetics: 0 M-GuHCl. protein concentration. 0.04 mg/ml. The continuous line indicates the theoretical curve for the unfolding transition constructed using the equation proposed by Tanford (1970) (see Got,rl & Hamaguchi. 1979).
1.5 M-GuH(‘I. exponential
Below 12 w-GuHCl, decay terms :
however,
F(t)-F(rX,)=F,exp(-hlt)+Fzexp(-Xzt).
the kinetics
were
described
by two
(1)
where F(t) is the fluorescence at time t, F( CO) is the fluorescence after equilibration, /\1 and h, are the apparent rate constants of t’he slow and fast phases. respectively. and F, and F, are the amplitudes of the respective phases. We denote the amplitudes of the slow and fast phases relative to the total change as ml1 and x2. respectively. where x1 -t- t2 = 1. Figure 2 shows the dependence on GuHCl concentration of the apparent rate constants (h, and h2) of the slow and fast phases and the amplitude (a2) of the fast phase. The initial points of the fast, and slow phases. which were obtained b) extrapolation to time zero of the respective phases. are shown in Figure 1. As can be seen. the initial points of the fast phase fell on a line expected for the dependence on GuHCl concentration of the fluorescence of native protein. This shows that the total unfolding process was seen in these experiments. The apparent rate constants obtained using manual mixing agreed with those of the slow phases obtained using stopped-flow mixing. As shown in Figure 2. the apparent rate constant (AI) of the slow phase increased with an increase in GuHCl concentration and tended to level off above 1 M-(:uH(‘l. The apparent rate constant (h2) of the fast phase. though we determined it onI> above 07 M-GuHCl. increased with an increase in GuHCI concentration. The
FOLDING
KINETICS
OF
REDUCED
C, FRAGMENT
915
(b)
0
I
I 2 [Gu HCll
I 3
I 4
(M)
FIN:. 2. The dependence on GuHCl concentration of (a) the apparent rate constants (A, and h2) and (b) the relative amplitude (1s) of the fast phase for unfolding and refolding kinetics of the reduced CL fragment, pH 75, 25°C. (0) From unfolding kinetics obtained by stopped-flow fluorescence measurements. The conditions are as described in the legend to Fig. 1. (a) From unfolding kinetics measured by fluorescence at 400 nm using manual mixing. Initial conditions: protein concentration, 1 mg/ml; 0 M-GuHCI. Final conditions: protein concentration, 902 mg/ml and the indicated concentration of GuHCl. (A) From refolding kinetics measured by fluorescence at 409 nm using manual mixing. Initial conditions: protein concentration, 1 mg/ml; 4 M-GuHCl. Final conditions: protein concentration, 902 mg/ml; the indicated concentration of GuHCl. (0) From refolding kinetics measured by c.d. at 218 nm. Initial conditions: protein concentration, 7 mg/ml; 4 M-GuHCI. Final conditions: protein concentration. 92 mg/ml; 91 M-GuHCI. ,( x ) The apparent rate constant of the isomerization process of Uz to U, in mechanism 1 after unfolding measured by double-jump experiments. The conditions are as described in the legend to Fig. 3. The dots below 0.7 M-GuHCI show the values of A, calculated using the microscopic rate constants in Fig. 5 (see the text). The open square shows the value of AZ calculated using the result of the double-jump experiment (see the text). The continuous line in (b) indicates the value of a2 calculated using K1, = 10 and the value of fN obtained from the equilibrium measurement (see the text).
amplitude (Q) of the fast phase increased from about @2 at @7 M-GuHCl to 1.0 above 1.5 M-GuHCl. The values of h, and its dependence on GuHCl concentration were very similar to those for the intact CL fragment, except that the curve for the reduced CL fragment shifts by about 1 M to a lower concentration region of GuHCl (see Fig. 3(a) of the accompanying paper). As explained for the unfolding of the 31
916
Y.GOTO
AND
K. HAMAGUCHI
Frc. 3. A double-jump experiment measuring the process of U, to U, in mechanism 1 after unfolding, pH 75, 25°C. Formation of the U, species was monitored by taking samples at the indicated times of unfolding and measuring the refolding kinetics by fluorescence at 400 nm (0) or at 350 nm (0). All the mixing procedures were done manually. The ordinate shows the fluorescence change for the slow phase observed in the refolding process relative to the maximum fluorescence change obtained at infinite time in the unfolding conditions. Initial conditions: 0 M-GuHCl; protein concentration, 2.5 mg/ml. Unfolding conditions: 4 M-GuHCI ; protein concentration, 1.2 mg/ml. Refolding conditions : @07 M-GuHCl ; protein concentration, O-02 mg/ml. The continuous line indicates the theoretical curve with a rate constant of 3 x lo-* s-l.
intact C, fragment in the accompanying paper, the results of Figure 2 suggests that the unfolding kinetics of the reduced CL fragment also follows the three-species mechanism : U +,+
(mechanism
1)
where N is native protein, U, and U2 are the two forms of unfolded protein, and krz, k,,, kz3 and k,, are the rate constants for the respective processes. In the accompanying paper, we have demonstrated the presence of two forms of the intact CL fragment by the double-jump method (Brandts et al., 1975; Nall et al., 1978). This method was also used here to confirm whether two forms are present in the unfolded molecule of the reduced CL fragment. The result is shown in Figure 3. The reduced CL fragment was first denatured by 4 M-GuHCl, and then refolding was initiated at 096 M at various times in the denaturing conditions. Unfolding of the reduced CL fragment in 4 M-GuHCl was complete within one second. As shown in Figure 3, the amplitude of the slow phase of refolding was dependent on the time of exposure of the protein to 4 M-GuHCl. This indicates that there exist the fastrefolding species (U,) and slow-refolding species (U,) in the unfolded reduced CL molecule and that the former converts slowly to the latter in 4 M-GuHCl. The apparent rate constant for the interconversion between U, and U, was estimated to be 3~1O-~s-‘, which is very similar to that for the intact CL fragment. As shown in Figure 3, however, the slow phase did not disappear and amounted to 20% of the total amplitude even at time zero. This will be discussedlater. As in the case of the intact CL fragment, the rate of the interconversion estimated from the double-jump experiments agreed well with the values of h, obtained from the
FOLDING
KINETICS
OF
REDUCED
C, FRAGMENT
917
unfolding experiments at high concentrations of GuHCl. These results obtained by the double-jump experiments strongly suggest that the unfolding kinetics of the reduced C, fragment follow the three-species mechanism, in which the equilibrium constant Kzl ( = k,,/k12) is independent of GuHCl concentration.
(c) Refolding
kinetics
of the reduced
C,, fragment
Refolding kinetics of the reduced CL fragment from 41v1-GuHCl measured by fluorescencesat 330,350 and 400 nm were all the same. The kinetics were very slow and the observable process (more than 90% of the total change) followed first-order kinetics. The amplitudes of the fast phasesat 400 nm, estimated assuming that the kinetics consist of two phases, are included in Figure 2(b). The apparent rate constants determined by fluorescence at 400 nm are shown in Figure Z(a). They agreed well with the apparent rate constants of the slow phasesdetermined by the unfolding experiments. The agreement of the kinetic parameters obtained from unfolding kinetics with those from refolding kinetics establishes the kinetic reversibility of the folding transition of the reduced CL fragment. The kinetics of refolding of the reduced CL fragment at 91 M-GuHCl from 2 MGuHCl were also studied by circular dichroism at 218 and 230 nm. The refolding was very slow and the apparent rate constant was found to be 3 x 10-j s-r. which is independent of the wavelength used and is the same as the rate constant determined by fluorescence measurement (Fig. 2(a)). The amplitude of the slow phase determined by c.d. at 218 and 230 nm was about 90% of the total change of the c.d. estimated by extending to lower GuHCl concentrations the linear portion corresponding to the unfolded state in the transition curve. This relative amplitude was the same as that estimated by fluorescence measurement (Fig. 2(b)). The agreement of the refolding kinetics obtained by fluorescence and cd. measurements suggests that there is no intermediate in the refolding process. This is consistent with the three-species mechanism. The large amplitude of the slow phase at very low concentrations of GuHCl and the low values of A, at very low concentrations of GuHCl compared with the rate constant of the isomerization from UZ to U, obtained by the double-jump experiments suggest that the proportion of U, to the total unfolded molecule is large.
(d) Titration
of SH groups
qf the
reduced
C,‘ fragment
Reduced CL fragment has two SH groups. These SH groups are buried in the interior of the molecule and are inaccessible to SH reagents (Goto & Hamaguchi, 1979,1981). Previously we studied the formation of the intrachain disulfide bond from the reduced CL fragment in the presence of glutathione and showed that the reaction of the disulfide formation is explained in terms of the two-state model, in which the SH groups can react freely with the reagent in the unfolded molecule but not in the native molecule (Goto & Hamaguchi, 1981). We also studied” the reactivities of the SH groups toward DTNB (Goto & Hamaguchi, 1979) and showed that the time-course of the reaction consists of a rapid phase and a slow phase and the rate constant (1 x 10m3s-r) of the latter phase is comparable with the rate
918
Y.GOTO
1 0
AND
I I
K.
I
2 [Gu
HAMAGI’CHI
I
I
I
3
4
5
HO.1 (MI
FIG. 4. The dependence on GuHCl concentration of the apparent determined by SH titration of the reduced CL fragment by DTNH. reduced CL fragment at a protein concentration of @06 mg/ml were reaction was monitored by the absorbance change at 412 nm. The shown in Fig. 2.
rate constant (0) for the slow phase pH 7.5. 25°C. The SH groups of the titrated with 0% mM-DTNB and the triangles represent the values of hZ
(6~1O-~s-’ ) of the unfolding estimated from the kinetics of the disulfide formation. On the basis of these findings it was expected that the reactivities of the SH groups can be used for studying the conformational change of the reduced CL fragment. We studied the reactivities of the SH groups of the reduced CL fragment toward DTNB at various concentrations of GuHCl. Titration of the SH groups with DTNB at various concentrations was first examined in the absence of denaturant. Between 0.1 and 25 m&f-DTNB, all the titration processes gave the same kinetics ; 5% of the total SH groups first reacted rapidly within the dead time of the measurement, (20 s) and the remaining 95% reacted slowly with a rate constant of 1 x 1O-3 s-‘. This shows that the intramolecular process of the reduced CL fragment is ratelimiting and not the reaction of the exposed SH groups with DTNB. The titration of the SH groups with 05 mM-DTNB was carried out at various concentrations of GuHCl. All the titration processes consisted of two phases but the rate constant and the amplitude of the slow phase changed markedly depending on GuHCl concentration. At the highest concentration of GuHCl used (0.7 M), the titration of the SH groups was examined by changing DTNB concentration three times, but the kinetics was independent of DTNB concentration, indicating that even at 0.7 M-GuHCl the intramolecular process of the reduced CL fragment is ratelimiting in the titration of the SH groups. The dependence on GuHCl concentration of the rate constant of the slow phase and the amplitude of the fast phase are shown in Figures 4 and 5(b), respectively. The rate constant of the slow phase increased markedly with increasing GuHCl concentration. These rate constants are on a line obtained by extending the values of hz determined at higher concentrations of GuHCl. The amplitude of the fast phase increased sharply with increasing GuHCl concentration (Fig. 5(b)). This change is the same as the unfolding
FOLDING
KINETICS
OF
REDUCED
I
I
CL FRAGMENT
919
I
(b)
0
I
2
3
CGu HCII
(M)
4
5
FIN:. 5. The dependence on GuHCl concentration of (a) the microscopic rate constants estimated based on mechanism 1 for folding of the reduced CL fragment and (b) the transition curve for unfolding obtained from kinetic measurements. pH 7.5. 25°C. (a) (0) k,,; (a) k,,; (A) kzl; (A) k,*. (b) (0) The amplitude of the fast phase determined by SH titration of the reduced CL fragment by DTNB. The conditions are as described in the legend to Fig. 4. (0) The fraction of the unfolded species calculated using the 4 microscopic rate constants. The continuous line indicates the transition curve for unfolding obtained from equilibrium measurement.
transition curve obtained by fluorescence or c.d. measurement. These results are explained in terms of mechanism 1. Because the reaction of the exposed SH groups with DTNB is very fast in the presence of excess DTNB (Whitesides et al., 1977), the amplitude of the fast phase should correspond to the proportion of the total amount of the unfolded species (U, +U,) and the amplitude of the slow phase should correspond to the proportion of the folded species (N). Therefore the change with GuHCl concentration in the amplitude of the fast phase should represent the unfolding transition of the molecule. The titration of the buried SH groups of the native molecule is thus rate-limited by the unfolding process and the rate of titration is equal to the rate of the unfolding of the reduced CL molecule (ks2). The value of A2 at a high GuHCl concentration represents the value of kS2 (see Discussion), and the rate constants obtained by SH titration are on a line extending
920
Y.GOTOANDK.HAMAGUCHI
the values of h2 at higher concentrations rate constant obtained by SH titration
of GuHCl. This confirms further that the represents the value of IL,, in mechanism 1.
4. Discussion (a) Folding
mechanism
of the reduced
CJragment
We have described the kinetics of unfolding and refolding of the reduced CL fragment as measured by tryptophyl fluorescence, cd. and by reactivities of the SH groups toward DTNB. All the kinetic data obtained here are found to be consistent with the three-species mechanism, U, --UZ+N. As described in the accompanying paper, an intermediate is formed during the refolding process of the intact CL fragment below the transition zone but not inside the zone. In the refolding of the reduced CL fragment, however, we did not observe anv intermediate. As shown in Figure 1. the unfolding transition of the reduced Ci. fragment begins at a very low GuHCl concentration and thus we could study the refolding only in the transition zone. If we could have studied the refolding of the reduced C, fragment far below the transition zone, we might have found an intermediate during the refolding process. The rate constants in mechanism 1 can be determined essentially the same way as described in the accompanying paper. The value of (lCr2+k2r) for the interconversion between IT, and U2 was determined to be 3 x 10m2 s-’ by the double-jump experiment. If the value of A, at a very low GuHCl concentration is assumed to be equal to k, 2, the value of K,, (= k,,/k,,) is estimated to be about 10. Since the unfolding transition of the reduced CL fragment begins at a very low GuHCl concentration, it is uncertain whether the value of A, at a low GuHCl concentration is exactly the same as the value of k,,. However, we tentatively assumed the value of K,, to be 10. This value is the same as the value of K21 for the intact fragment. The value of k32 was directly obtained from the titration of the SH groups with DTNB (Fig. 4). The value of k,, is obtained by the relation: 4,
= k,,(l
+K21)(1
-fd/fn.
(2)
using the values off, (the proportion of the total unfolded molecule) determined by the titration of the SH groups with DTNB (Fig. 5(b)). Although only the values of hi are available below 67 M-GuHCl, the values of k,, and k,, can be calculated with equations (3) and (4) of the accompanying paper and the known values of k,,, k32, fD and A, and the assumed value of K,, Between 0.7 and 1.2 M-GuHCl. the values of A,, A2 and a2 are available and thus the four microscopic rate constants in mechanism 1 can be determined with equations (3) to (6) of the accompanying paper. In the region far above the transition zone, the value of A2 is equal to k,,, because k23 should be small. The four microscopic rate constants thus obtained at various concentrations of GuHCl are shown in Figure 5(a). In Figure 5(b) the values offn in the region of 0.7 to 1.2 M-GuHCl, calculated using the kinetic parameters. are shown. It can be seen that they fall on the equilibrium curve obtained by fluorescence measurement. With increasing concentration of GuHCl, k2, decreases and k,, increases sharply. This is expected for the rates of conformational transition of proteins. The values of
FOLDING
KINETICS
OF
REDUCED
CL FRAGMENT
92 1
kZ3 and k,, are the same at 0.8 M-GuHCl, which is higher by 64 M-GuHCl than the apparent midpoint of the equilibrium transition. This difference is due to the presence of the two forms of the unfolded species and is discussed for the intact CL fragment in the accompanying paper. The rates of the isomerization process (k,, and k,,) are independent of GuHCl concentration. These results confirm the validity of the above assumption that K,, equals to 10 and is independent of GuHCl concentration. The values of hz below 96 M-GuHCl, where the values of hz could not be measured directly by our procedure, can be calculated using the above microscopic rate constants (see equation (2) of the accompanying paper). The values of h2 thus calculated are given in Figure 2. The calculated hz values agreed well with the value obtained experimentally at 0.7 M-GuHCl and increase with decreasing GuHCl concentration. The dependence of h2 on GuHCl concentration has a minimum at’ about 0.8 M-GuHCl. The appearance of the minimum is expected from mechanism 1 (see the accompanying paper). The continuous line in Figure 2(b)shows the values of a2 expected under the limiting conditions where h2/h, $ 1 (see equation (9) of the accompanying paper). The observed values of a2 inside the transition zone are smaller than these values in the limiting conditions. A similar observation has been reported for ribonuclease A (Brandts et al., 1975 ; Hagerman & Baldwin, 1976). Hagerman (1977) has discussed the amplitude of the fast phase in mechanism 1. which is expressed as a function of the ratio of the two apparent rate constants (h,/h,) at fixed values offN and K,,, and shown that under the conditions where the ratio of AZ/h1 is not so large, a2 is expected to be smaller than the value of a2 under the limiting conditions. As can be seen from Figure 2(a), the separation of h2 and h, at about 1 M-GuHCl is actually not so large. In the case of the intact CL fragment. h2 is well separated from h, at all the concentrations of GuHCl studied and the observed values of or2, except below the transition zone where an intermediate emerges. are nearly equal to the values of a2 under the limiting conditions. In the double-jump experiments (Fig. 3), the slow phase of refolding did not disappear and amounted to about 2OyA of the total change at time zero. The relation between the relative amplitude and the ratio of the two apparent rate constants (/\2/h,) is extended to the analysis of the results obtained by the doublejump experiments. The relative amplitude of the fast phase at time zero obtained by the double-jump experiment is expressed as a function of h2/h, by: (3) where In equation (3), fi is the proportion of the native molecule at the start of unfolding and f; is the proportion of the native molecule when refolding is complete under specified conditions. This equation can be applied to the case where the unfolding of N to U2 is complete and refolding is initiated prior to conversion from U2 to U,. This equation predicts that, in the refolding conditions where fk is nearly equal to j”& cy2 approaches 1 when the ratio of h,/h, is extremely high and l2
022
Y.GOTO
ASI)
Ii.
HAMAGI-(‘HI
hecomes smaller than 1, i.e. the slow phase does not disappear is not so large compared with h,. We analyzed the result of the double-jump experiment est,imated t’he apparent rate constant of the fast phase conditions where the value of h, was difficult to determine R2r = 10. f& = 995. fi = 090 and l\2 = 04 in equation (3). obtained for X,/h,. When a value of 3 x 1V3 s- ‘. which was the refolding experiments. is used for h,, we obtain a value of hz. This value agrees quite well with the value obtained rate constants (Fig. 2).
(b) tlolr
of the intrnchuin
disulfide
at time zero when A2 using equation (3) and (h2) in the refolding experimentally. ITsing a value of about 50 is obtained directly from of 0.15 s- ’ as the value using the microscopic
baud
We no\v obtained the four microscopic rate constants in the three-species mechanism for the folding kinetics of the reduced (‘L fragment. As reported in the accompanying paper, we also obtained the microscopic rate constants for the intact CL fragment under the conditions where the kinetics follow the three-species mechanism. We compare the rate constants for the reduced CYLfragment (Fig. 5) with those for the intact CL fragment (Fig. 9 of the accompanying paper) and consider the role of the intrachain disulfide bond in the folding of the CL fragment. The rate constants of the isomerization in the unfolded state (kIZ and k,,) are the same for both proteins and the slow-refolding species amounts to about 90% of the total unfolded species for both proteins. This indicates that the two forms of the unfolding species are formed irrespective of the presence of the intrachain disulfide bond. The two forms of the unfolded species tnust, have been formed by the isomerization of proline residues, and the isomerization process for the reduced CL fragment is the same as that for the intact C, fragment. The role of a particular prolinr residue in the folding of the intact CL fragment is discussed in detail in the accompanying paper. The rate constant (kj2) of unfolding of the reduced (:, fragment is only slightly greater than that, of the intact C, fragment. Only the rate constant (kz3) of folding from the fast-folding species to native protein differs greatly between the two proteins. Although the rate constants kz3 for both proteins were obtained in different concentration ranges of GuHCl because of their different stabilities. t)he marked difference of k,, is apparent. When we compare t,he values of k,, at 1 .O MGuH(‘1. where the rate constants for both proteins were determined experimentally. the rate constant for the intact, (1, fragment is about 100 t,imes greater than that, for the reduced rL fragment. The large increase in the rate constant of refolding by the presence of the intrachain disulfide bond indicates thus importance of the disulfide bond in the direct folding transition of the CL fragment As reported previously (Goto & Hamaguchi. 1979), the stability of the (I,> fragment decreases upon reduction of the intrachain disulfide bond. This is due mostly to the decreased rate constant (kz3) from I’, to K for the reduced VL fragment. Knowledge of the process from U2 to N is important if we are to understand the folding of the C, fragment and we consider t,his process in detail. W’e estimated the free energy cohange of denaturation of the CL fragment using a
FOLDING
KINETICS
OF
REDUCED
923
C, FRAGMENT
two-state approximation for the equilibrium transition (Goto & Hamaguchi, 1979). As described in the accompanying paper, however, when two forms of the unfolded species (U, and U,) are present, the apparent equilibrium constant (K,,,) of native protein to the total unfolded protein may be expressed by: K a,,,, = (W,)+(U,)V(pu’)
= K32(l+K21).
(4)
where K32
= P’,)/(N)
= k&m
Thus the apparent free energy change of denaturation ( - RT In K,,,) is smaller by RT In (1+ Kzl) than the true free energy change (-RT In K32). The term RT in (1 + K21) is calculated to be 1.4 kcal mol-’ using K,, = 10 at 25°C for the intact and reduced CL fragments. We first constructed the free-energy diagrams of the folding and unfolding of the intact and reduced CL fragments at 1.0 M-GuHCl, where the rate constants of unfolding and refolding (k23 and k32) were determined experimentally for both proteins. The transition state free energy, A@. is calculated using the relation: A# = - RT In (kh/k,T),
(-5)
where k is the rate constant, h is Planck’s constant, kB is Boltzmann’s constant, T is t,he temperature in Kelvin, and R is the gas constant. Figure 6(a) shows the profiles at 25°C constructed by setting the native states for both proteins at the same energy level. The free energy change of denaturation estimated using the rate constants of unfolding and refolding is 2.0 kcal mol- ’ for the intact C, fragment and - 1.2 kcal mol - ’ for the reduced CL fragment. These values are in good agreement with those calculated using equation (4) and the equilibrium transition data (Goto & Hamaguchi. 1979).
u--.-. (a)
20
1
lntoct
CL
(b) Intact CL
Reduced
19.4
Reduced
21.8
CL
CL
21.5
19.1
2.0
_-------_-
N
L;-
-1.2
N
N
PI<:. 6. Free-energy profiles for the unfolding and refolding of the intact and reduced Cl,. fragments in the presence of 1 M-GuHCl and (b) in the absence of GuHCI. pH 7.5. BR’C. X and 11 represent native and denatured states. respectively. The energy lrvels of the native states of the intact reduced C, fragments were set to zero (see the text).
(a) the and
924
Y.GOTO
AND
K.
HAMAGUCHI
Figure 6(b) shows the free-energy profiles in the absence of GuHCl. The profile for the reduced CL fragment can be constructed directly using the values of k, 3 and Ic,, Because the values of kZ3 and k32 for the intact CL fragment in the absence of GuHCl were not determined, we constructed the profile assuming that the difference in ks2 in the absence of GuHCl between the intact CL fragment and reduced CL fragment is the same as that in the presence of 1.0 RI-GuHCl, and using the true free energy change of denaturation in the absence of GuHCl, which is calculated with equation (4). and the equilibrium transition data reported previously (Goto & Hamaguchi, 1979). The rate constant of refolding of the intact CL fragment is estimated to be about 80 s-l from this energy profile. When the free energies for the intact and reduced CL fragments are assumed to be the same in the native state, the energy levels of the transition states for both proteins are very similar and only the energy levels in the denatured state are different between the two proteins. Previously we showed that the decrease in the stability of the (1, fragment on reduction of the disulfide bond is consistently explained in terms of the increase in the entropy of the denatured state (Goto Br Hamaguchi, 1979). Provided that this is the case, the free-energy profile shows that the change in the refolding rate is due entirely to the change in the entropy of the denatured state. By the presence of the disulfide bond, the energy level of the denatured state is increased by about 4 kcal mol-‘, which is due to the decreased entropy of the denatured state. This results in an increase of about 100 times in the refolding rate. The other cont’ribution of the disulfide bond to the folding kinetics of the CL fragment. if any, may be small compared with the entropic effect, and we may thus conclude that the kinetic role of the disulfide bond in the folding of the C, fragment is largely entropic. This may be applicable to the folding of other proteins with several disulfide bonds. The energy levels of the denatured states are different between the intact and reduced VL fragment only because of the different, entropies of the denatured proteins but the energy levels of the transition states are the same for both proteins. This means that the presence of the disulfide bond does not affect the ratelimiting step for the conformational transition of the CL fragment. Several mechanisms have been proposed for protein folding (see, for instance; Baldwin. 1978.1980). The problem as to the rate-limiting step in the folding pathway is still controversial. The results obtained here suggest t,he location of the rate-limit,ing step when the folding process is measured by the compactness of the molecule. If the rate-limiting step of the folding of the C, fragment is located near the denatured state. the transition states of the t,wo proteins may be in different energy levels because the loop formed by t,he disulfide bond is large and thus the entropies of the two proteins in the early stage of folding may be different. If we assume that the intact and reduced CL fragments fold through the same folding pathway, the same energy level for the transition states may be realized only when a compact conformation, for which there is no difference in the conformational entropy between the intact and reduced CL fragments. is formed at a later stage of folding. The unfolding rates for the intact and reduced CL fragments are expected to be the same for the transition state with such a compact conformation. On t’he basis of his detailed kinetic data for the reduction and oxidation of the
FOLDING
KINETIC8
OF REDUCED
C, FRAGMENT
925
three disulfide bonds in bovine pancreatic trypsin inhibitor, Creighton (1977,1978,1980u) has constructed the energy profile of the entire intramolecular transition in the unfolding and refolding of the protein. He has shown that the transition is very co-operative and its rate-limiting step is very close to the native state. The kinetics of formation of the disulfide bonds in ribonuclease A obtained by a similar method (Creighton, 1979) and the results of unfolding and refolding of several proteins obtained by urea gradient gel electrophoresis (Creighton. 19806) are also consistent with this idea. Segawa et al. (1973) studied the kinetics of thermal unfolding of lysozyme and suggested that in the transition state the molecule folds tightly, though many intramolecular hydrogen bonds are broken. Although a further study is necessary to define the rate-limiting step in the folding of the CL fragment, our results are consistent with these suggestions. On the basis of the kinetic data for the disulfide formation in the reduced CL fragment, we estimated the rates of unfolding and refolding of the reduced CL fragment in the absence of denaturant (Goto & Hamaguchi, 1981). Whereas the unfolding rate obtained (6 x 10T4 s-l) is in fair agreement with that obtained in the present experiments (1 x lo- 3 s-l ), the refolding rate estimated from the disulfide formation (3 x 10m2s-l) is considerably smaller than the value (918 s-l) estimated here. In the analysis of the kinetic data for the disulfide formation, we used a twostate approximation instead of the three-species mechanism. Therefore, in addition to the true rate of the refolding of U2 to N, the rate of the slow isomerization processof U, to U, is involved in the rate of refolding estimated from the kinetics of the disulfide formation. On the other hand, the same process of the unfolding of the native molecule can be seen both in the kinetics of the disulfide formation and in the kinetics described here, and thus both kinetic data give the same rate constant of unfolding. This work was supported by grants for scientific research from the Ministry Science, and Culture of Japan and a grant from the HoanshaFoundation.
of Education,
REFERENCES Amzel, L. M. & Poljak, R. L. (1979). Annu. Rev. Biochem. 48, 961-997. Baldwin, R. L. (1978). Trends Biochem. Sci. 3, 66-68. Baldwin, R. L. (1980). In Protein Folding (Jaenicke, R., ed.), pp. 369-384, Elsevier/North Holland, Amsterdam. Beale, D. & Feinstein, A. (1976). Quart. Rev. Biophys. 9, 135180. Brandts, H. F., Halvorson, H. R. & Brennan, M. (1975). Biochemistry, 14, 4953-4963. Creighton, T. E. (1977). J. Mol. Biol. 113, 295-312. Creighton, T. E. (1978). Prog. Biophys. Mol. Biot. 33, 231-297. Creighton, T. E. (1979). J. Mot. Biot. 129, 411-431. Creighton, T. E. (198Ou). In Protein Folding (Jaenicke, R., ed.), pp. 427-441, Elsevier/North Holland, Amsterdam. Creighton, T. E. (1980b). J. Mot. Biol. 137, 61-80. Gething, M. J. H. & Davidson, D. E. (1972). Eur. J. B&hem. 39, 352-353. Goto, Y. & Hamaguchi, K. (1979). J. Biochem. 86, 1433-1441. Goto, Y. & Hamaguchi, K. (1981). J. Mol. Biol. 146, 321-340. Goto, Y. & Hamaguchi, K. (1982). J. Mol. Biol. 156. W-910. Goto, Y., Azuma, T. & Hamaguchi, K. (1979). J. Biochem. 85, 1427-1438. Hagerman, P. J. (1973). Biopotymers, 16, 731-747.
9%
Y.GOTO ANI) K.HAMAGCTCHI
Hagerman, P. J. & Baldwin, R. L. (1976). Biochemistry, 15, 1462-1473. Nail, B. T.. Garel, J.-R. & Baldwin, R. L. (1978). J. Mol. Biol. 118, 317-330. Segawa, 8., Husimi, Y. & Wada, A. (1973). Biopolymers, 12. 2521-2537. Tanford, C. (1970). Advun. Protein Ghem. 24, l-95. Whitesides, G. M., Lilburn, J. E. & Szajewski, R. P. (1977). J. Ory. Chem. 42, 332-338 Edited
by S. Brenner