Dependence of conformational stability on ... - Europe PMC

Proc. Natl. Acad. Sci. USA Vol. 84, pp. 4441-4444, July 1987 Biochemistry

Dependence of conformational stability on hydrophobicity of the amino acid residue in a series of variant proteins substituted at a unique position of tryptophan synthase a subunit (mutant protein/hydrophobic

scale/stabiity of a protein/site-directed mutagenesis)

KATSUHIDE YUTANI*t, KYOKO OGASAHARA*, TADAHIRO TsuJITA*, AND YOSHINOBU SUGINO§ *Institute for Protein Research, Osaka University, Yamadaoka, Suita, Osaka 565, Japan; and §Kansai Medical University, Uyamahigashi, Hirakata, Osaka 573, Japan

Communicated by Robert L. Baldwin, March 30, 1987 (received for review January 27, 1987)

To elucidate the role of individual amino acid ABSTRACT residues in stabilizing the conformation of a protein, we have constructed a series of variant a subunits of tryptophan synthase from Escherichia coli substituted by each of 20 amino acids at position 49, which is buried in the interior of the protein. The stabilities were quantitatively examined except for the mutant protein substituted by arginine, which was not obtained in enough quantity. The Gibbs energy of unfolding in water and the activation Gibbs energy of unfolding in 3 M guanidine hydrochloride for each protein were compared at pH 7.0 and pH 9.0. The Gibbs energy of unfolding in water at pH 7.0 varied from 0.72 to 1.92 times that of the wild-type protein by the substitutions, but the activation Gibbs energy of unfolding in 3 M guanidine hydrochloride varied only from 0.95 to 1.03 times that of the wild-type protein. Moreover, the stability of the protein substituted at this position, which is buried in the interior of the molecule, tended to increase linearly with increasing hydrophobicity of the substituted residue, unless the volume of the substituted residue was over a certain limit. We still cannot fully predict the three-dimensional structure of a protein on the basis of the amino acid sequence, although the sequence determines the three-dimensional structure of the protein under physiological conditions. On the other hand, recent advances in molecular biology have given us the ability to modify the gene of a protein virtually without limitations (1). This shows that if we can predict the threedimensional structure from the amino acid sequence, it will help greatly in designing a desired protein. To do so, however, more information is needed on the folding mechanism and stabilization of three-dimensional structure and their relationship to amino acid sequence. The study of the effect of single amino acid substitutions on conformational stability using mutant proteins is a fruitful approach to understanding the role of the amino acid sequence in protein folding and protein stability. The stabilities of mutant forms of tryptophan synthase ct subunit (2-6) and T4 phage lysozyme (7) have been studied quantitatively. In the case of tryptophan synthase a subunit from Escherichia coli, the conformational stabilities of wild-type and of six mutant proteins substituted at position 49, which is buried in the interior of the protein, have been compared. The results indicate that (i) the proteins easily become more stable or less stable to denaturant as a result of single amino acid substitutions and (ii) the stabilities of the proteins tend to increase linearly with hydrophobicity of the substituting residues (8). It is widely appreciated that the hydrophobic interaction is important in protein folding and protein stability (9-11). The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

However, the role of hydrophobic interaction in the conformational stability of a protein has yet to be quantitatively evaluated for each amino acid residue. The purpose of this article is to show how the hydrophobicity of the substituting residues at the same position in the interior of a protein correlates with the conformational stability of the protein. Recently, we have obtained all the mutants substituted by each of 20 amino acid residues at position 49 of tryptophan synthase a subunit from E. coli. In this paper, the values of unfolding Gibbs energy in water for these proteins were estimated from denaturation curves by guanidine hydrochloride (Gdn-HCl). The activation Gibbs energy of unfolding in 3 Gdn-HCl for each of these proteins was also estimated from denaturation kinetics. We will discuss the correlation between hydrophobicity of the substituting residues and the stability of these proteins.

MATERIALS AND METHODS The wild-type a subunits of tryptophan synthase of E. coli and six mutant proteins substituted by methionine, glutamine, valine, serine, leucine, and tyrosine in place of glutamic acid at position 49 were obtained as described (5). The other 13 mutant proteins substituted at the same position were newly obtained by site-directed mutagenesis using synthetic oligonucleotides. The single-stranded DNA template used for the mutagenesis was an M13mpll derivative containing the a-subunit gene isolated from an Hincd fragment (5632-6925) of tryptophan operon. A plasmid derived from plasmid pUC8 (Pharmacia) containing the tryptophan promoter introduced from pDR720 (Pharmacia) was used as the expression vector. Purified mutant proteins (20-100 mg) were obtained from 5 liters of broth, except for the mutant protein substituted by arginine. Purified a-subunit monomers gave a single peak on HPLC in Mono Q column (Pharmacia) in all cases. Gdn-HCl (specially prepared reagent grade from Nakarai Chemicals, Kyoto, Japan) was used without further purification. Other chemicals were reagent grade. Circular dichroism (CD) measurements were carried out with a Jasco J-500 recording spectropolarimeter equipped with a data processor for CD (Model DP-501). Stopped-flow denaturation reaction was followed by the same CD apparatus equipped with stopped-flow controller model SFC-5. Denaturation by Gdn-HCl was examined at pH 7.0 and pH 9.0 by following CD values at 222 nm as described (3, 12).

RESULTS CD Spectra of Mutant a Subunit. CD spectra examined in the region of 200 nm to 260 nm for the 18 mutant a subunits of tryptophan synthase substituted at position 49 were not Abbreviations: Gdn-HCl, guanidine hydrochloride; OMH, optimal

matching hydrophobicity(ies). tTo whom reprint requests should be addressed.

4442

Biochemistry: Yutani et al.

Proc. Natl. Acad. Sci. USA 84 (1987)

distinguishable from that of the wild-type a subunit (not shown). The data on two mutant proteins have already been reported (13). These results indicate that the backbone conformation of the mutant proteins is similar to that of the wild-type protein. Equilibrium Stability of the Wild-Type and Mutant Subunit. Fig. 1 shows typical denaturation curves of two mutant a subunits by Gdn HCl at pH 7.0, followed by CD values at 222 nm. As already reported (3), the substitutions at position 49 affected stabilities to Gdn HCl. These denaturation curves were biphasic, suggesting the presence of one stable intermediate state in the denaturation process. The tryptophan synthase a subunit of E. coli consists of two domains: a1, NH2-terminal, 188 residues; a2, COOH-terminal, 80 residues (14). Denaturation of this protein proceeds in two steps. In the first step, only the a2 domain unfolds, but the a1 domain remains native, resulting in a stable intermediate form. In the second step, the a1 domain also unfolds (12, 15). To obtain a quantitative measure of the stabilities of the mutant proteins, we estimated the Gibbs energy of unfolding from the Gdn'HCl denaturationr curves. This was possible because denaturation curves of the mutant proteins by Gdn'HCl were reversible. We assumed that there was one intermediate state in the denaturation process and that denaturation by Gdn HCl was represented by an equilibrium among a unique native form (N), an intermediate form (I), and an unfolded form (D) as shown below. a

N

-±

I

-±

D.

LLc

1.04 0

FIG. 1. Typical Gdn HCl denaturation curves of mutant a subunits of tryptophan synthase. CD values at 222 nm were measured after the sample was kept in 20 mM Potassium phosphate buffer (pH 7.0) containing 0.1 mM dithioerythritol and 0.1 ,mM EDTA at 254C for 30 min. The protein concentration was -0.03 mg/ml. Fapp was calculated according to Eq. 2. Solid lines are best-fit curves accQrding to Eq. 3. Pro-49 and Cys-49, mutant a subunit substituted by proline and cysteine at position 49, respectively.

Gdn HCl, respectively. Fapp following equation (3): Fapp

F.

=

[1

[pp]1

bn-

[2]

[0]

[Old'

where [0] represents the CD value at 222

nm

at a

given

concentration of Gdn HCl and [6]n and [Old represent the CD values at 222 nm in the absence and in the presence of 3.2 M

3

2

Gdn-HCI, M

[1]

From the Gdn HCl denaturation curve, we can calculate the apparent fraction of unfolding, Fapp, by the following equation:

1

_

KO,nd(l 1 +

+

Kond(l

can

be represented by the

+ 3a)"nnd aKoni(1 3a)An +

d

+

+

3a)"n-i [3]

Ko,ni(l + 3a)4

where Ko represents the equilibrium constant in the absence of denaturant (in water), An is the difference in the number of Gdn HCl molecules bound per protein molecule between any two states, a is the fraction of the total phange in CD value at 222 nm that occurs in going from N to I, and a is the activity of Gdn-HCl. Thus, in the above formula, Fapp is represented as a function of the activity of Gdn HCl. We used

Table 1. The values of unfolding Gibbs energy (AdG) in water and activation Gibbs energy of unfolding (A4G) in 3 M Gdn HCl for the 19 proteins substituted by each of the 19 amino acids at the same position of tryptophan synthase a subunit AdG in water,* AdGt in 3 M Gdn HCl, kcal/mol (251C) kcal/mol (250C) Residue at OMH pH 7.0 pH 9.0 pH 7.0 pH 9.0 position 49 scale* kcal/mol 7.1 ± 0.1 6.4 ± 0.1 18.24 ± 0.01 18.57 ± 0.06 -0.67 Gly 6.8 ± 0.2 18.08 ± 0.05 18.56 ± 0.09 0.5 Ala -0.40 8.5 ± 0.2 Val 9.4 ± 0.2 18.30 ± 0.17 12.0 ± 0.3 18.98 ± 0.05 1.5 0.91 16.8 ± 0.5 10.0 ± 0.3 18.62 ± 0.08 19.21 ± 0.06 Ile 1.25 12.2 ± 0.4 18.60 ± 0.12 Leu 15.0 ± 0.7 1.8 1.22 19.01 ± 0.09 6.9 ± 0.8 18,16 ± 0.01 18.77 ± 0.13 Pro 8.2 ± 0.3 -0.49 6.8 ± 0.5 17.99 ± 0.15 18.23 ± 0.02 1.67 2.3 8.8 ± 0.5 Tyr Phe 11.2 ± 0.2 8.3 ± 0.2 17.30 ± 0.04 17.95 + 0.05 2.5 1.92 5.7 ± 0.5 9.9 ± 0.5 16.87 ± 0.03 3.4 0.5 Trp 17.19 ± 0.05 His 10.1 ± 0.8 9.2 ± 0.4 18.49 ± 0.05 18.64 ± 0.05 0.5 -0.64 7.9 ± 0.5 7.5 ± 0.9 17.96 ± 0.12 Lys 18.36 ± 0.02 -0.67 8.2 ± 0.2 Asn 6.2 ± 0.1 17.47 ± 0.05 18.24 ± 0.02 -0.92 Gln 6.3 ± 0.3 8.5 ± 0.2 17.93 ± 0.04 18.51 ± 0.15 -0.91 8.5 ± 0.4 7.0 ± 0.1 17.36 ± 0.06 17.94 ± 0.08 Asp -1.31 Glu 8.8 ± 0.1 4.9 ± 0.3 18.16 ± 0.05 17.97 ± 0.06 -1.22 11.0 ± 0.1 8.3 ± 0.3 18.17 ± 0,05 Cys 0.17 17.84 ± 0.04 Met 13.3 ± 0.2 8.4 ± 0.3 18.11 ± 0.04 18.49 ± 0.02 1.02 1.3 Thr 8.8 ± 0.2 7.0 ± 0.8 17.54 ± 0.12 17.71 ± 0.03 0.4 -0.28 Ser 7.4 ± 0.2 8.0 ± 0.8 17.57 ± 0.04 18.06 ± 0.17 -0.55 -0.3 *Data of proteins that have Val, Tyr, Gln, Glu, Met, and Ser at position 49 are cited from ref. 4, and Leu data are from ref. 5. tData are from ref. 16. tData are from ref. 17.



4443

transition state of unfolding, AdGt from the rate constant of the denaturation, kf, by using the Eyring equation:

AdGt = RT in (kT/h) - RT in kf, Pro-49

[41

where k and h are the Boltzmann and Planck constants, respectively. We calculated the activation Gibbs energy of unfolding in 3 M Gdn HCl for 19 variant proteins of tryptophan synthase a subunits at pH 7.0 and pH 9.0, as shown in Table 1.

DISCUSSION 10

Time, sec

A globular protein is stabilized by hydrophobic residues in its interior. Position 49 of the a subunit of tryptophan synthase has been conjectured to be buried in the hydrophobic interior of the molecule, using the mutant protein (19, 20). Recently, x-ray crystallographic analysis of tryptophan synthase a2/32

FIG. 2. Denaturation by jumping from 0 M to 3 M Gdn-HCl of mutant a subunit of tryptophan synthase in 25 mM Tris HCl buffer (pH 7.0) at 250C, monitored at 222 nm by CD stopped-flow. Pro-49 or Thr-49, mutant a subunit substituted by proline or threonine at position 49, respectively. The protein concentration was -0.03 mg/ml. Each trace is the average of eight scans.

a computer program to give a least-squares fit of experimental data to obtain Komd In the present paper, the values of KO,nd were obtained at pH 7.0 and pH 9.0 for the mutant proteins substituted by glycine, alanine, isoleucine, proline, phenylalanine, tryptophan, histidine, lysine, asparagine, aspartic acid, cysteine, or threonine in place of glutamic acid at position 49 of the a subunit in addition to those already reported (3-5). The values of Gibbs energy of unfolding in water (AdG = -RT In Ko,nd) for the 19 proteins substituted at position 49 are presented in Table 1. The values of AdG in water from the equilibrium method using Eq. 3 have been directly confirmed by calorimetry for the wild-type and two mutant a subunits (18). Kinetic Stability of the Wild-Type and Mutant a Subunit. Fig. 2 shows the stopped-flow denaturation reaction of two mutant a subunits by jumping from 0 M to 3 M Gdn HCl at pH 7.0, monitored by CD values at 222 nm. As the data for the denaturation reaction of the proteins in 3 M Gdn HCl demonstrate single-phase kinetics, as shown in Fig. 3, we can calculate the activation Gibbs energy between the native and

A

151 '5 E V

O

10

0w

OY

-.1

5

I

I

0

2

1

Agr.

3

kcal/mol

B

l5 '5 E

Tv.

OF z

0H

10

/

O,

?E

OY

-1

5

-1

0 1 OMH scales

FIG. 4. Correlation between unfolding Gibbs -3'

4

Time,

energy

(AdG) in

water of wild-type and 18 mutant proteins and hydrophobicity of the substituting residues at position 49. Amino acids (designated by the single-letter code) correspond to the data of proteins that have the indicated amino acid at position 49. (A) At pH 7.0. As hydrophobic sec

FIG. 3. Semilogarithmic plots of denaturation by jumping from 0 M to 3 M Gdn HCl of mutant a subunits. Val-49, Met-49, or Ser-49, mutant a subunit substituted by valine, methionine, or serine at position 49, respectively. Fn, fraction of the native protein. The conditions were the same as in Fig. 2.

scales, Ag,, (16) is used. Straight line was obtained from least-squares fit of the 8 points shown by solid circles. The value of Agtr for glycine is 0 by definition since determining Agt, Nozaki and Tanford (16) subtract the transfer Gibbs energy for glycine from that for the amino acid in question. (B) At pH 7.0. As hydrophobic scales, OMH scales (17) are used. Straight line was obtained from least-squares fit of the 13 points shown by solid circles.

4444


complex of Salmonella typhimurium has yielded preliminary results (21). In this structure, the glutamic acid residue at position 49 is in the interior of the subunit of S. typhimurium, which has a sequence very similar to that of E. coli (22). Thus, it is natural that the conformational stability of the mutant proteins at position 49 is correlated with the hydrophobicity of the substituting residues. On the other hand, there have been different scales of hydrophobicity for amino acid residues. In this paper, we chose the Nozaki-Tanford solvent transfer scales (Agtr) obtained from the solubility of amino acid residues in organic solvents (16) and the SweetEisenberg optimal matching hydrophobicity (OMH) scales obtained by considering the observed frequency of amino acid replacements among related structures (17). In a previous paper (5), we correlated unfolding Gibbs energy in water (A1dGH20) of the wild-type tryptophan synthase and six mutant proteins, which are substituted by glutamine, methionine, valine, tyrosine, leucine, or serine at position 49 with Agtr of the substituting residue at pH 5.5, 7.0, and 9.0, indicating that their conformational stabilities increase linearly with Agtr of the substituting residues, except for tyrosine. Data on newly obtained mutants at pH 7.0 are included in Fig. 4A. The data on eight proteins that have leucine, methionine, valine, histidine, threonine, alanine, serine, or glycine at position 49 fit a straight line quite well, which was obtained from a least-squares fit of the eight points. The slope was 3.71. The outliers from the straight line in Fig. 4A are the mutant proteins substituted by the aromatic amino acids. The conformational stability of these proteins

did not increase in proportion to the increase in hydrophobicity of their residue. We notice that aromatic amino acids have larger residual volumes than the other amino acids that conform to the linear relationship. This suggests that the space available for position 49 residue in the interior of the molecule is just below the volume of these residues. At pH 9.0, a similar linearity among the eight proteins was found (not shown), but the slope was 1.90. The observation that the slopes are different at pH 7.0 and pH 9.0 and differ from 1.0 suggests that factors other than hydrophobicity must be important in determining this plot. We have previously discussed these differences in slope and suggested that an increase in hydrophobicity in the interior of a protein molecule produces extra electronic energy (5). Next we plotted AdG in water against OMH scales of hydrophobicities, as shown in Fig. 4B. The straight line in Fig. 4B was obtained by least-squares fit of the 13 proteins, which are shown by solid circles. The slope was 3.82. The two groups shown by open circles in the figure were outliers. One group consisted of the three proteins that have ionizable residues at pH 7.0-that is, glutamic acid, aspartic acid, and histidine. Glutamic acid residue at position 49 in the wild-type protein is not ionized at pH 7.0, since an apparent dissociation constant of this glutamic acid is 7.5 (5). The other two residues (aspartic acid and histidine at position 49) also might not be ionized at pH 7.0. If these residues remain un-ionized in abnormal surroundings, they may be left out of consideration. Another group of outliers was the mutant proteins substituted by aromatic residues. These results coincide with those of Fig. 4A. The ionized group would destabilize the protein conformation when a residue in the interior of a protein is substituted by an ionized residue. The yield during purification of the mutant protein substituted by arginine was extremely


low. This suggests that the arginine residue at position 49 of the mutant protein remains ionized at pH 7.0 and the stability is very low. The difference between the highest and the lowest values of the Gibbs energy at pH 7.0 among the 19 proteins examined was 12.53 kcal/mol for AdG in water, but only 1.43 kcal/mol for AdG* in 3 M Gdn HCl (1 cal = 4.184 J). AdG in water among the 18 mutant proteins varied from 0.72 to 1.92 times that of the wild-type protein by the substitutions, while AdGf in 3 M Gdn HCl varied only from 0.95 to 1.03 times that of the wild type. These results indicate that single amino acid substitutions affect equilibrium stability of the proteins in water more strongly than they affect kinetic stability in 3 M Gdn HCl. This work was supported in part by Grants-in-Aid for Scientific Research 58580112 and 61130001 from the Ministry of Education, Science and Culture of Japan. 1. Ulmer, K. M. (1983) Science 219, 666-671. 2. Yutani, K., Ogasahara, K., Sugino, Y. & Matsushiro, A. (1977) Nature (London) 267, 274-275. 3. Yutani, K., Ogasahara, K., Suzuki, M. & Sugino, Y. (1979) J. Biochem. (Tokyo) 85, 915-921. 4. Yutani, K., Ogasahara, K., Kimura, A. & Sugino, Y. (1982) J. Mol. Biol. 160, 387-390. 5. Yutani, K., Ogasahara, K., Aoki, K., Kakuno, T. & Sugino, Y. (1984) J. Biol. Chem. 259, 14076-14081. 6. Beasty, A. M., Hurle, M. R., Manz, J. T., Stackhouse, T., Onuffer, J. J. & Matthews, C. R. (1986) Biochemistry 25, 2965-2974. 7. Elwell, M. L. & Schellman, J. A. (1977) Biochim. Biophys. Acta 494, 367-383. 8. Yutani, K., Ogasahara, K. & Sugino, Y. (1985) Adv. Biophys. 20, 13-29. 9. Rose, G. D., Geselowitz, A. R., Lesser, G. J., Lee, R. H. & Zehfus, M. H. (1985) Science 229, 834-838. 10. Eisenberg, D. & McLachlan, A. D. (1986) Nature (London) 319, 199-203. 11. Baldwin, R. L. (1986) Proc. Natl. Acad. Sci. USA 83, 80698072. 12. Yutani, K., Ogasahara, K. & Sugino, Y. (1980) J. Mol. Biol. 144, 455-465. 13. Yutani, K., Ogasahara, K., Suzuki, M., Sugino, Y. & Matsushiro, A. (1978) in Biochemistry of Thermophily, ed. Friedman, S. M. (Academic, New York), pp. 233-249. 14. Higgins, W., Fairwell, T. & Miles, E. W. (1979) Biochemistry 18, 4827-4835. 15. Miles, E. W., Yutani, K. & Ogasahara, K. (1982) Biochemistry 21, 2586-2592. 16. Nozaki, Y. & Tanford, C. (1971) J. Biol. Chem. 246, 22112217. 17. Sweet, R. M. & Eisenberg, D. (1983) J. Mol. Biol. 171, 479-488. 18. Yutani, K., Khechinashvili, N. N., Lapshina, E. A., Privalov, P. L. & Sugino, Y. (1982) Int. J. Pept. Protein Res. 20, 331-336. 19. Yutani, K., Ogasahara, K., Suzuki, M. & Sugino, Y. (1980) J. Biochem. (Tokyo) 87, 117-121. 20. Ogasahara, K., Yutani, K., Suzuki, M., Sugino, Y., Nakanishi, M. & Tsuboi, M. (1980) J. Biochem. (Tokyo) 88, 1733-1738. 21. Hyde, C. C., Padlan, E. A., Davies, D. R., Ahmed, S. A. & Miles, E. W. (1987) Fed. Proc. Fed. Am. Soc. Exp. Biol. 46, in press. 22. Schneider, W. P., Nichols, B. P. & Yanofsky, C. (1981) Proc. Natl. Acad. Sci. USA 78, 2169-2173.