Nov 25, 2018 - 0 1984 by The American Society of Biological Chemista, Inc. vol. 259, No. .... 800 V (constant) until the pH gradient was fully developed and the.
THEJOURNAL OF BIOLOGICAL CHEMISTRY
vol. 259, No.22, Issue of November 25, pp. 14076-14081 1984 Printed in &s.A.
0 1984 by The American Society of Biological Chemista, Inc.
Effect of Amino Acid Residues on Conformational Stability in Eight Mutant Proteins Variously Substitutedat a Unique Position of the Tryptophan Synthase a-Subunit" (Received for publication, May 21,1984)
Katsuhide YutaniS, Kyoko Ogasahara, KunioAoki, Tomisaburo Kakuno, and Yoshinobu Suginoj From the Institute for Protein Research, Osaka University, YamadaokaSuita, Osaka 565, Japan andSKansai Medical University, Uyamahigashi, Hirakata Osaka 573,Japan
To elucidate the role of individual amino acid residues in stabilizing the conformation of a protein, the stabilities of wild-type tryptophan synthase a-subunit from Escherichia coli and seven mutant proteins substituted by single amino acid residues at position 49, which is buried in the interior of the protein, were compared. The mutant proteins have Gln, Met, Val, Tyr, Leu, Ser, or Lys in place of Glu in the wild-type protein. The dissociation constant, pK, of the Glu residue at position 49 for the wild-type protein was determined to be 7.5 from atitration curve obtained by comparison of two-dimensional isoelectric focusing electrophoresis of the wild-type and mutant proteins. Our results indicate that 1) the conformational stabilities of the proteins studied increase linearly with hydrophobicity of the substituting residues (except Tyr), with the coefficient of this linear dependence being 2.0, 3.4, or 1.3 at pH 5.5, 7.0, or 9.0, respectively; and 2) Lys or Glu at position 49 serve as a destabilizing factor when ionized.
dine hydrochloride at different pH values. They have been found to be correlated with the hydrophobicity of the substituting amino acid residue (3-6). However, whether Glu at position 49 is protonated or not at the examined pH remained unknown, so that the hydrophobic parameter for Glu could not be estimated. Consequently, a correlation between the hydrophobicity of the substituting amino acid residue and conformational stability of the protein could not be made. In this paper, the pK for the Glu residue at position 49 is determined from a titration curve obtained by two-dimensional isoelectric focusing electrophoresis of a mixture of the wild-type and mutantproteins. The correlation of conformational stability with the natureof the substitutingamino acid residue willbe discussed on the basis of studies of eight proteins, including our new results on mutant proteins substituted by Leu and Lys, in addition to the older data on six of the proteins (5). MATERIALS AND METHODS
The a-subunitof tryptophan synthasewas isolated from the PW11, trpA3, trpAl1, trpA33, trpA88, Ser49, Lys49, and Leu49 strains of E. coli. The peak I and I1 proteins from PWll were used as the wildtype and N-terminally formylated a-subunit, respectively (9).' The The conformation of a native protein in a given environ- Ser49, ku49, and Lys49 strains were obtained from an amber mument is determined by the amino acid sequence (1).Therefore, tant at position 49 (10) by lysogenizing with a specialized transthe three-dimensional structure of a protein should be pre- ducing phage X carrying the S U I + suppressor gene, h?'psul+h, the sy+ dictable in theory onthe basis of the amino acid sequence. In suppressor gene, hps~~+cI857h,andthe s y + suppressor gene, practice, however, the prediction of three-dimensional struc- X:'u-ampsuP+, respectively, all of which were provided by Dr. Ozeki (11). The trpA3, trpA11, trpA33,and trpA88 strains (12) were donated ture from a given amino acid sequence has yet to be realized. by Drs. Yanofsky (Stanford University) and Crawford (Iowa UniverMore information is needed on folding mechanisms and on sity). The trpA3,trpA11,trpA33,trpA88, Ser49, Leu49, or Lys49 the stabilization of the three-dimensional structure, and their strains have Val, Gln, Met, Tyr, Ser, Leu, or Lys, respectively, in relation to the amino acid sequence. One approach to under- place of Glu at position 49 of the a-subunit. The purifications were standing therole of the aminoacid sequence in protein folding performed as described (13). Purified a-subunit monomers gave a and protein stability is to study the effects of single amino single band on isoelectric focusing in all cases. Guanidine hydrochloride, specially prepared reagent-grade from Nakarai Chemicals, acid substitutions on conformational stability using mutant Kyoto, was used without further purification. Other chemicals were proteins (2-8). reagent-grade. On the basis of studies of the wild-type tryptophan synthase Protein concentrations of the wild-type and mutant proteins were 5: 4.8 for a-subunit (EC 18.104.22.168) of Escherichia coli andits many estimated from absorbance at 278.5 nm, assuming mutant proteins, we could suggest that proteins may easily Tyr49 and 4.4 for the other proteins (14). Acid-base titrations of protein samples were carried out under become more or less stable as a result of single amino acid humidified nitrogen in a water-jacketed cell at 10 "C with an Orion substitutions (2). Conformational stabilities for the wild-type Model 701A pH meter. In each experiment, 4.0 mlof the protein and the mutantproteins substituted by Gln, Ser, Met, Val, or solution, which had been dialyzed against 0.15 M KC1, were placed in Tyr in place of Glu at position 49 have been estimated by the the titration cell thermostatted at 10 "C. Then either HCI (0.004-0.5 calorimetric method or by anaIysis of denaturation inguani- N) or KOH (0.004-0.5 N) was added in portions of 15 pl from a calibrated Gilson pipette. The pH was measured after the solution *This work was supported in part by Grant 58580112 from the The abbreviations used are: G1u49 and formylated protein, the Ministry of Education, Science, and Culture of Japan to K. Y . The costs of publication of this article were defrayed in part by the wild-type and N-terminally formylated tryptophan synthase a-subpayment of page charges. This article must therefore be hereby unit, respectively; Va149, Gln49, Met49, Tyr49,Ser49, Leu49, or Lys49 marked "advertisement" in accordance with 18 U.S.C. Section 1734 mutant proteins substituted by Val, Gln, Met, Tyr, Ser, Leu, or Lys in place of Glu at position 49 of the wild-type a-subunit, respectively; solely to indicate this fact. GdnHC1, guanidine hydrochloride. $ To whom correspondence should be sent.
Comparison of Stability in
Eight Mutant Proteins
800 V (constant) until the pH gradient was fully developed and the system was in a steady state. Then, the cathodic and anodic regions
were cut away, and new electrode strips were overlaid onto thegel a t right angles to those of the first run. A filter paper strip of about 0.7mm width was overlaid perpendicular to the preformed pH gradient a t the appropriate position. Approximately 40 pl of sample (2.5 mg/ ml) were applied to the stripwith a microsyringe, and the stripwas taken off a few minutes later. The second run perpendicular to the pHgradient was performed for 15-18 min a t lo00 V (constant). During electrophoresis the temperature was maintained constant. Before staining, the pH was measured on the gel surface by using surface electrode (LKB 2117-111). CD measurements were carried out with a Jasco 5-500 recording spectropolarimeter equipped with a data processor for CD (Model DP-501). Denaturation by GdnHCl was examined a t three different pH values for Leu49 by following CD values a t 222 nm as described (3).
* i PH
FIG.1. Acid-base titration curve for tryptophan synthase a-subunit (wild type)in 0.15 M KCl. 2 represents the average net number of hydrogen ions bound to or released from a protein molecule. Temperature was 10.0 "C, and protein concentration was 0.11 mM. was stirred for 3 to 5 min on a magnetic stirrer. Similar titration was performed on 4.0 mlof the solvent as a blank (0.15 M KCI). The titration curve of the protein was obtained by subtraction. The two-dimensional isoelectric focusingelectrophoresis technique (15) was performed by using an LKB 2217 Ultrophor apparatus (26 X 11 cm, 0.5-mm thick) asfollows. An ultrathin gel was polymerized to contain 5.5% acrylamide and 2% Ampholine by casting onto a polyacrylamide gel-bond film. The concentration of acrylamide was 3.5% when electrophoresis was performed in 8 M urea. The cathodic or anodic strips were soaked in 2% ethylenediamine or 1% phosphoric acid before use, respectively. The first dimension was run for 3 h a t
Acid-base Titration of Tryptophan Synthase a-SubunitFig. 1 shows the results of the titration of Glu49 in 0.15 M KC1 at 10 "C. The ordinate 2 represents the average net number of hydrogen ions bound to or released from a protein molecule. The isoionic point (2 = 0) of this protein was found to be 5.2, by finding a point such that the amount of acid needed to bring the solution to sufficiently low pH from the isoionic pointis equivalent to the total number of basic groups, whichis found to be 29 from the amino acid sequence. The value obtained here agreed with the value of 5.3, previously obtained from isoelectric focusing (9). This indicates that titration data are reliable, although a back titration could not be obtained because solutions of the protein became turbid
r I S
FIG. 2. Titration curves by electrophoresis perpendicular to preformed pH gradient. Protein mixture in a, b, and c contains Glu49 + Va149, Glu49 + Ser49, and G 1 d 9 + formylnted protein, respectively. That in d contains Glu49 + Val49 in the presence of 8 M urea. a, b, and c were done a t 15 "C and d a t 25 'C. S represents the start line of each sample. The anode was at theupper end in each gel. pH was measured on the gel surface a t 3cm distance from the sample line a t intervals a t 1 cm. The length of each slab gel was about 23 cm.
Comparison of Stability in Eight Mutant Proteins
;0.2 rd +J
a 0 a
PH FIG. 3. The difference between mobilities of two proteins in two-dimensional electrophoresis. a,and b were obtained from the data of Fig. 2, a and c, respectively. Distance moved was read by a digitizer (Mutoh Bit Pad One). The midpoint of the curve in a or b represents the apparent pK for the Glu residue at position 49 or the a-amino group of Glu.49, respectively.
after exposure to extreme acidic or basic pH. As the 2 value was found to be about -9 at pH8.5 from Fig. 1, the difference in mobility on electrophoresis at this pHof this protein and
a mutant protein substituted by a single amino acid residue that does not dissociate at this pH would amount to 11%(1/ 9). This suggested the possibility that the dissociation constant of an ionizable residue near this pH would be obtained from titration curves by two-dimensional isoelectric focusing electrophoresis of a mixture of two proteins differing by a single amino acid residue. Isoelectric Focusing Electrophoresis--In order to obtain the dissociation constant of the carboxyl group of the Glu at position 49 of Glu49, electrophoresis perpendicular to a stationary pH gradient (pH 3.5-10) was performed for a mixture of Glu49 and Val49 (Fig. 2 4 . The isoionic points of both proteins were 5.3 and agreed with that reported (9). In this case, the two curves were separated and parallel above pH 8, whereas they coincided below pH 7. The component that moved more rapidly at pH values above 8 in Fig. 2a was identified a Glu49 by comparison to experiments with Glu49 or Val49 alone. The distances moved from the start line in Fig. 2a were 3.5 and 3.85 cm at pH8.5. The ratio (3.5/3.85 = 0.91) agreed with that (8/9 = 0.89) predicted from the acidbase titration curve for the netcharge difference of one. These results suggest that Glu49 has one more negative charge than Val49 above pH 8, but theextra charge in Glu49 is protonated below pH 7. The difference in distance moved between the two curves in Fig. 2a is shown in Fig. 3a. The midpoint can be taken to be the apparent dissociation constant, pKapp,for the ionizable group in question. The pKappwas estimated to be 7.47 (average of three experiments). In spite of the large difference from the usual pK of Glu in proteins,it seems most natural toconclude that thispKapp represents the dissociation constant of the Glu at position 49. To confirm this value, we performed similar experiments with another mutant protein. Fig. 2b shows the titration curve of a mixture of Glu49 and Ser49. The two curves also coincide below pH 7. In thiscase, pKappof Glu at position 49 was calculated to be 7.49 (average of two experiments). This is in excellent agreement with that obtained by using Val49, in spiteof the considerable difference in the substituted residues. Furthermore (data not shown), the titration curves of Val49 and Ser49, both of which have no ionizable groups at position 49, were completely superpos-
FIG. 4. Correlation betweenthe Gibbs energy change of unfolding in water of the wild-type and mutant proteins and the hydrophobicity of the substituting residues at position 49. The dataexcept for Led9 are taken from Ref. 5. Age represents the value of the Gibbs energy change (21) for transfer from ethanol to water of amino acid residues. The value for ionized Glu residue is taken from Levitt's estimation (22). Top, at pH 9.0 and 25 "C,middle, at pH 7.0 and 25 "C;bottom, at pH5.5 and 3 "C. a, transition from the native to the completely unfolded form, AC3'; b, transition from the native to the intermediate form, AG?'.
: . c
pH 5.5 51
Hydrophobc parameter dgtr (kcal I mol
Comparison of Stability inMutant Eight Proteins
FIG. 5. Effect of pH on stabilities of the wild-type and mutant proteins.a, change of CD values at 222 nm of Glu49 (01, GIN29 (A), Met49 (e),Vu149 (W), Ser49 (O), and Tyr49 (A)as functions of pH in the absence of denaturant. In the acid region (below pH 7), CD at 222 nm and pH were measured at 3 "C after the samples were kept 2 h at 3 "C. In the alkaline region, CD at 222 nm and pHwere measured at 25 "C after 30 min at 25 "C. pH was adjusted by addition of HCI or KOH. All experiments were done in 0.15 M KC1 containing 0.1 mM EDTA and 0.1 mM dithioerythritol. CD measurements were performed at protein concentrations of 0.02 to 0.04 mg/ml using a cell with 10-mm path length. The CD value at pH7 is taken asunity. Results for Glu49 and Gln.49 have been reported (4).b, residual activity of Glu49 (0)and Lys49 (e)as a function of pH. After the sample was incubated in a solution containing 0.1 mM EDTA and 0.1 mM dithioerythritol for 30 min at 53 "C and the respective pH, activities were measured by the consumption of indole after reaction at 37 "C and pH8.0 for 20 min in the presence of pyridoxal phosphate, serine, and the&subunit of tryptophan synthase (23).
able. These results strongly suggest that the difference between the two curves in Fig. 2 a o r b is not due to theionization of a third group influenced by the residue at position 49, but to ionization of Glu at position 49 itself. The Glu pK.,, value is very far from that reported for Glu residues in other proteins (16). If the electrophoresis were performed in 8 M urea, the pK,,, value should be close to normal. Fig. 2d shows the result of experiments on Glu49 and Val49 in 8 M urea. The two curves are separate from the alkaline region down to pH 5. The high pK value of Glu-49 suggests that thisresidue is not completely exposedto solvent. This agrees with our previous conclusion that position 49 is in the interior of the molecule (14, 17). In a mixture of Glu49 and formylatedprotein, the two curves should be separate below pH 7, where a charge difference arises as the &-aminogroup is protonated. This is indeed the case, as demonstrated in Fig. 2c. The isoionic points of the formylated protein and Glu49 were 5.1 and 5.3, respectively, and again they agree with those reported (9). The pK.,,of the a-amino group for Glu49 was calculated to be 7.10 from
the difference between the two curves (Fig.3b), in agreement with the normal value reported for a-amino groups (16). These results indicate that thepK values of ionic groupsin proteins can be obtained from titration curves by isoelectric focusing electrophoresis when two homologous proteins differing only at a single position are available. Correlation between Gibbs Energy Change of Unfolding and Hydrophobicity of Substituted Residues-Thermodynamic parameters for unfolding of a-subunit of tryptophan synthase have already been reported at three pH values for the wildtype and five mutant proteins substituted at position 49 (5). In this study, we also provide data for the mutant protein substituted by Leu at the same position. The value of the Gibbs energy change on unfolding for Leu49 was calculated from GdnHCl denaturation curves by following CD valuesat 222 nm as described previously (3). The denaturation curves were biphasic as reported for the wild-type and mutant proteins (3). These proteins consist of two domains: a-1, the Nterminal 188 residues; and a-2, the C-terminal 80 residues (18).The denaturation by GdnHCl proceeds in two-steps. In the first step,only the a-2 domain unfolds and thea-1domain remains native, and in the second step, the a-1 domain also unfolds (3-5, 19, 20). In the transition from the native to the completely denatured state, the value of the Gibbs energy change for unfolding, AGHy,for Leu49 in water was calculated to be8.6,15.0, and 12.2 kcal/mol at pH 5.5, 7.0, and 9.0, respectively. In thefirst stepof denaturation, the value of the Gibbs energy change, AG?', for Leu49 in water was 2.9, 5.8, and 5.1 kcallmol at pH 5.5, 7.0, and 9.0, respectively. The Gibbs energy changes of unfolding in the transition from the native to the completely denatured state for the seven proteins are shown in Fig. 4a as a function of the transfer Gibbs energy change from ethanol to water, kt,, of the substituting amino acid residues at position 49. From the preceding section, the Glu-49 can be assumed to be protonated at pH 5.5 and 7, but ionized at pH 9. Consequently, for A g c r of Glu at pH 5.5 or 7, a value (0.55 kcal/mol) for protonated Glu was used (21), and at pH 9, a value (-2.5 kcal/mol) for ionized Glu residue was used (22). The values of AG9' for the seven proteins were linearly correlated with the hydrophobicity of the substituting residues at the three pH value, except for Tyr49. Straight lines in the figure were drawn by the least-squares method from the data points obtained, excluding Tyr49. The slopes (dAG2O/dAgtr)at pH 5.5, 7, and 9 were 2.0, 3.4, and 1.3, respectively. The deviation of AGBO from the straightline obtained by least squares was 0.40,0.38, and 0.76 kcal/mol at pH 5.5,7, and 9, respectively. In the first step of denaturation,the difference of the Gibbsenergy change, AG?', among the seven proteins was not marked, as shown in Fig. 4b, suggesting that the stabilities of the a-2 domains in all the proteins were similar. Acid or Alkaline Denaturation of the Wild-type and Mutant Proteins-As described above, the Gibbs energy change of unfolding changed remarkably at three different pH values for each protein. Therefore, the pH dependence of stability for these proteins was examined by CD measurements at 222 nm as shown in Fig. 5a. In the acid region, the transition curves of these proteins (except Gln49) coincided with each other. In the alkaline region, the midpoint of the transition curve for Glu49 was at pH 10.9, and it was the most labile of the six proteins compared, in contrast to the results in the acid region. The instability to alkali of Glu49 relative to the other proteins corresponded to its instability in GdnHCl at pH 9 (Fig. 4a). We could not obtain enough purified Lys49 to measure the CD. Therefore, activities after heating at 53 "C for 30 min
Comparison of Stability inMutant Eight Proteins
were measured for Lys49 and Glu49 as a function of pH, although this measured irreversible denaturation. Lys49 was more labile in the acid region and more stable in the alkaline region than Glu49.
where P is the molar polarizability value of an amino acid residue, p the density of the material, and M its molecular weight. The dielectric constant generally decreased with increase in hydrophobicity in amino acidresiduesexamined here. This suggests the possibility that an increase in hydroDISCUSSION phobicity in the interior of a protein molecule produces extra p K of Glu at Position 49-Electrophoresis of a protein electrostatic energy. The plots of Tyr49 at three pH values in Fig. 4a are far sample perpendicular to a preformed pH gradient in agel slab has been proposed as an effective preliminary procedure to below the straight lines, indicating that the conformational obtain optimal conditions for ion-exchange chromatography stability for Tyr49did not increase in proportion to the of proteins (24). We applied this method to obtain the pK of increase in hydrophobicity due to the Tyr residue. This may a charged amino acid in a protein by the useof mutant be due to an effect of the phenolic OH or the large volume of proteins substituted by a neutral amino acid at the same Tyr counteracting the effect of hydrophobicity. Effect of an Ionizable Group on Conformational Stabilityposition. To confirm the titration curves we (i) compared them with the acid-base titration ofGlu49; (ii) conducted It has been reported that the far-UV CD spectra of Glu49 do electrophoresis of different mutant proteins substituted at the not change between pH 7 and9 (4), indicating that the same position; (iii) performed electrophoresis in a preformed backbone structure of the protein is not affected by pH change pH gradient of Glu49 and Val49 in the presence of 8 M urea; in this range. This indicates that the conformation of Glu49 and (iu)applied the method to a basic residue (a-amino group remains native, evenwhenGlu-49 is ionized at pH 9.As of Glu49). All procedures gave reasonable results. This indi- reported previously (6, 26), Glu49 shows pH dependence in cates that the apparent pK value of Glu at position 49 ob- the thermal denaturation between pH 7 and 9, in contrast to the mutant proteins substituted with a neutral amino acids tained by this method is reliable. The pKappvalue obtained (7.5) is very far from those so far at position 49. This difference can be attributed to the dissoreported (16), in agreement with our previous result that ciation of Glu-49. We have discussed the possibility that this extra carboxyl position 49 ofthis protein is buried in thehydrophobic interior ion might serve as adestabilizing factor in the alkaline region, of the molecule. Effect of Hydrophobicity of the Substituted Residue on Con- either by interacting electrostatically with other residues with formational Stability-A globular protein is stabilized by hy- negative charges or by being bound with water (4). But from drophobic residues inits interior. The magnitude of the the dataof Fig. 4a(top), the instability of Glu49 at pH 9seems "hydrophobicity" of amino acid residues has been estimated to be caused by a decrease in hydrophobicity caused by ionifrom the Gibbs energy change for transfer from ethanol to zation: the value of A g t r for Glu residues is 0.55 and -2.5 kcal/ water (21). The conformational stability of Glu49, Gln49, and mol in the nonionized (21) and ionized forms (22), respecMet49 has been reported to be correlated with the hydropho- tively. On the other hand, the instability ofLys49 in the acid bicity of the substituting amino acid residues (3). Fig. 4a also shows that thestabilities of the proteins are linearly depend- region (Fig. 5b) can be compared with the instability of Glu49 ent upon the hydrophobicity of the substituting amino acid in the alkaline region. Ionization of Lys at position 49 in residues at three pH values in the transition from the native Lys49 may serve as a destabilizing factor in the acid region. to thecompletely denatured state.Although calculation of the It may also beexplained by the change in hydrophobicity on Gibbsenergy change obtained from GdnHCl denaturation ionization: the values of Ag,, for Lys are 1.5 and -3.0 kcal/ curves needed someassumptions, some values usedhere have mol for the nonionized (21) and ionized forms (22), respecbeen directly confirmed by calorimetry (6). At pH 7 ,the Gibbs tively. energy change (AG3O) on unfolding was about three times Acknowledgments-We thank Drs. T. Horio, N. Ui, K. Hamaguchi, the increase in Gibbs transfer energy change (Ag,) of the substituting amino acid residue at position 49. But the slopes and T. Takagi for their valuable comments. Mutant strains were kindly supplied by Drs. C. Yanofsky (Stanford University) and I. P. of the increases (dAG3O/dAgtr)at pH 5.5 and 9 were smaller Crawford (Iowa University), and the X phage carryingthe suppressor than that at pH7. gene by Dr. H. Ozeki. Thanks are also due to the Crystallographic As shown in Fig. 4b, in the first step of denaturation, i.e. Research Center of the Institute for Protein Research, Osaka Unithe transition from the native to the intermediate state in versity, through which an ACOS S850 computer was made available whichonly the a-2 domain is unfolded, the difference of to us. A G F ~ Oamong the examined proteins was not marked. This REFERENCES indicates that the difference in the stabilities among these 1. Anfinsen. C. B.. and Scherarra. H. A. (1975) . . Adu. Protein Chem. proteins mainly depends on the difference in the stability of 29,205-300 the a-1 domain in which position 49 is located. 2. Yutani, K., Ogasahara, K., Sugino, Y., and Matsushiro, A. (1977) Why is the hydrophobicity at pH 7 more effective in staNature ( L o n d . ) 267,274-275 bilizing a protein than that at pH5.5 or 9? Stabilization of a 3. Yutani, K., Ogasahara, K., Suzuki, M., and Sugino, Y.(1979) J. protein conformation caused by hydrophobic interactions Biochem. (To~Yo) 85,915-921 alone can not directly depend on pH. Thus, we are faced with 4. Yutani, K., Ogasahara, K., and Sugino, Y. (1980) J. Mol. Biol. 144,455-465 a case of electrostatic interactions modified by hydrophobic5. Yutani, K., Ogasahara, K., Kimura, A., and Sugino, Y. (1982) J. ity. Electrostatic interactions are inversely proportional to Mol. Biol. 1 6 0 , 387-390 dielectric constants in the medium; in this case, this means 6. Yutani, K., Khechinashvili, N. N., Lapshina, E. A., Privalov, P. that theelectrostatic interaction depends upon dielectric conL., and Sugino, Y. (1982) Int. J. Pept. Protein Res. 2 0 , 331stants in the interior. The dielectric constants (€) of amino 336 acid residues may be calculated from polarizabilities by the 7. Grutter,M. G., Hawkes, R. B., and Matthews, B. W. (1979) Nature (Lond.)2 7 7 , 667-668 following equation (25), - ,
-1 "_ t
8. Schellman, J. A., Lindorfer, M., Hawkes, R., and Grutter, M.
(1981) Biopolymers 2 0 , 1989-1999 9. Sugino, Y., Tsunasawa, S., Yutani, K., Ogasahara, K., and Suzuki,
Comparison of Mutant Eight Proteins Stability in M. (1980) J . Biochem. (Tokyo) 87, 351-354 10. Drapeau, G. R., Brammar, W. J., and Yanofsky, C. (1968) J. Mol. Biol. 35,357-367 11. Ozeki, H., Inokuchi, H., Yamano, F., Kodaira, M., Sakano, H., Ikemura, T., and Shimura,Y. (1980) in Transfer RNA:Biological Aspects (Abelson, S.D., and Shimmel, P. R.,e&) pp. 341362, Cold Spring Laboratory Press, New York 12. Yanofsky, C . , and Horn, V. (1972) J. Bwl. Chem. 247, 44944498 13. Tsunasawa, S., Yutani, K., Ogasahara, K., Taketani, M., Yasuoka, N., Kakudo, M., and Sugino, Y. (1983) J. Biochem. (Tokyo) 47, 1393-1395 14. Ogasahara, K., Yutani, S., Suzuki, M., Sugino, Y., Nakanishi, M., and Tsuboi, M. (1980) J. Biockm. (Tokyo)88,1733-1738 15. Righetti, P. G., Krishnamoorthy, R., Gianazza, E., and Labie, D. (1978) J . Chromatogr. 1 6 6 , 455-460 16. Nozaki, Y., and Tanford, C. (1967) Methods Enzymol. 11, 715734
14081 17. Yutani, K., Ogasahara, K., Suzuki, M., and Sugino, Y. (1980) J . Biochem. (Tokyo) 87,117-121 18. Higgins, W., Fairwell, T., and Miles, E. W. (1979) Biochemistry 18,4827-4835 19. Miles, E. W., Yutani, K., and Ogasahara, K. (1982) Biochemistry 2 1.2586-2592 20. Iwahashi, H.,Yutani, K., Ogasahara, K., Tsunasawa, S., Kyogoku, Y., and Sugino, Y. (1983) Biochim. Biophys. Acta 744, 189-192 21. Tanford, C. (1962) J. Am. Chem. SOC.84,4240-4247 22. Levitt, M.(1976) J. Mol. Biol. 104, 59-107 23. Smith, H. O., and Yanofsky, C. (1962) Methods Enzymol. 5 , 794806 24. Haff, L. A., Fagerstam, L. G., and Barry, A. R. (1983) J . Chromatogr. 266, 409-425 25. Pethig, R. (1979) in Dielectric and Electronic Properties of Biological Materiuls, John Wiley & Sons, New York 26. Ogasahara, K., Yutani, K., Suzuki, M., and Sugino, Y. (1984) Int. J. Peptide Protein Res. 24, 147-154