Nucleophile specificity in α-chymotrypsin- and subtilisin-(Bacillus subtilis strain 72) catalyzed reactions

188

Biochimica et Biophysica Acta, 1160 (1992) 188- 192 © 1992 Elsevier Science Publishers B.V. All rights reserved 0167-4838/92/$05.00

BBAPRO 34327

Nucleophile specificity in a-chymotrypsin- and subtilisin(Bacillus subtilis strain 72) catalyzed reactions Mikhail Yu. Gololobov a, Tatyana L. Voyushina u, Valentin M. Stepanov b and Patrick Adlercreutz a a Department of Biotechnology, Chemical Center, University ofLund, Lund (Sweden) and b Institute for Genetics and Selection of Industrial Microorganisms, Moscow (Russia) (Received 3 February 1992)

Key words: a-Chymotrypsin; Subtilisin; Quantitative structure activity relationship; Nucleophile specificity; Substrate specificity

Nucleophilic properties of amino-acid amides were studied systematically in acyl-transfer reactions catalyzed by a-chymotrypsin and subti|isin from Bacillus subtilis strain 72 (subtilisin 72) using MaI-L-AIa-L-AIa-L-PheOMe as the acyl-group donor. In a-chymotrypsin-catalyzed reactions, the nucleophile reactivities increase in the following order: D-AIaNH 2 < GIyNH 2 < LAIaNH 2 < L-SerNH 2 < L-ThrNH 2 < L-HisNH 2 < L-ValNH 2 < L-LeuNH 2 < L-TrpNH 2 < L-MetNH 2 < L-NvaNH: < L-PheNH 2 < L-IIeNH 2 < L-TyrNH 2 < L-ArgNH 2. In reactions catalyzed by subtilisin 72, the reactivities increase as follows: L-LeuNH 2 < LIleNH 2 < L-ThrNH 2 < L-ArgNH 2 < L-TrpNH 2 < L-NvaNH 2 < L-ValNH 2 < L-MetNH2 < L-AIaNH 2 < L-SerNH2 < D-AIaNH : < GIyNH 2. In a-chymotrypsin-catalyzed reactions, hydrophobic interactions are entirely responsible for the differences between the reactivity of the nucleophiles for amides of all the amino-acids tested with the exception of D-A!aNH:, L-ArgNH 2 and L-TyrNH 2. In reactions catalyzed by subtilisin 72, amino-acid side-chain characteristics and the nucleophile reactivities are not related. The data obtained show the low selectivity of the S'1 subsite of subtilisin 72 and high specificity of this subsite in a-chymotrypsin.

Introduction Acyl-transfer reactions catalyzed by a-chymotrypsin and different subtilisins are widely used in mechanistic studies [1-12] and in preparative synthesis [13-20]. However, there is an evident lack of systematic data related to the specificity of nucleophiles in these reactions. The different experimental conditions used and the obvious influence of the nature of acyl groups on the partitioning of an acyl-enzyme between water and added nucleophile [10] hamper quantitative comparison. Qualitatively, the available data show that the S'1 subsite of a-chymotrypsin prefers bulky amino-acid residues [2,3,6,7,10,15,21-24]. This position in subtilisin

from Bacillus subtilis strain 72 (subtilisin 72) has no definite specificity but small amino-acid residues are preferred [16,25]. In this paper, we report systematic data on the reactivity of different amino-acid amides in acyl-transfer reactions catalyzed by a-chymotrypsin and subtilisin 72. A protected tripeptide methyl-ester (MalAlaAlaPheOMe) was used as the acyl-group donor. This comparatively long compound corresponds to the active site of both enzymes better than amino-acid derivatives [10,21,22,24,25]. The data obtained are discussed in terms of quantitative structure activity relationships (QSAR).

Materials and Methods

Correspondence to: P. Adlercreutz, Department of Biotechnology, Chemical Center, University of Lund, P.O. Box 124, 221 00 Lund, Sweden. Abbreviations: QSAR, quantitative structure activity relationships; E254, the molecular absorption coefficient at 254 nm; Mal-, maleyl; -OMe, methyl ester; Xaa, an amino-acid residue; subtilisin 72, subtilisin from Bacillus subtilis strain 72; DMSO, dimethyl sulfoxide; Nva, norvaline. The notation of enzyme binding subsites corresponds to the Schechter and Berger nomenclature [26]. If not otherwise stated, amino-acid residues are of the L-configuration.

Chemicals a-Chymotrypsin (EC 3.4.21.1) from bovine pancreas was obtained from Sigma and was used without further purification. A Ladizin enzyme plant (Ukraine) supplied us with the crude subtilisin from Bacillus subtilis strain 72. Purification was carried out with the published method [25]. MalAlaAlaPheOMe was synthesized as described elsewhere [27,28]. 5,5-Diethyl-barbituric acid (veronal) and acetic acid were from Merck,

189 DMSO was from J.T. Baker, acetonitrile (far UV) was purchased from Lab-Scan and trifluoroacetic acid (protein sequencing grade) from Sigma. If not otherwise stated, the chemicals were of analytical grade.

Measurements of nucleophile reactivities A stock solution (0.1 M) of MalAlaAlaPheOMe was prepared in DMSO. This stock solution (5 /zl) was added to 0.5 ml of 0.1 M veronal (pH 9.0) containing an appropriate concentration of the amino-acid amide and kept at 30°C for 10 min. The reaction was initiated by adding of 5 ml of the enzyme stock solution (about 0.1 mg/ml). Samples (100/xl) were collected at 5, 10, 20 and 45 min and added to 1 ml of DMSO. The resultant concentration of the components was determined by HPLC. In separate experiments, we made sure that the composition of the reaction mixture does not change after the addition of DMSO for at least three days. Veronal does not act as a nucleophile; only MalAIaAlaPheOH appeared in the progress of the reaction without the amino-acid amides. In a-chymotrypsin-catalyzed reactions, five or more different concentrations of each nucleophile were tested; they varied within an order of magnitude. In reactions catalyzed by subtilisin 72, the lower reactivity of the nucleophile usually did not allow us to vary nucleophile concentration considerably. Three concentrations of each nucleophile (having 2-3-fold difference) were tested. At every concentration of every nucleophile four experiments were usually carried out. The concentration of a nucleophile tested depended on the structure of this compound and the enzyme. In a-chymotrypsin-catalyzed reactions, the nucleophile concentrations were selected to give a [MalAlaAlaPheXaaNH 2]/[MalAIaAIaPheOH] ratio between 0.4 and 9. In reactions catalyzed by subtilisin 72, this ratio was between 0.1 and 1 due to the lower reactivity of the nucleophiles in these reactions. The nucleophile concentration when [MalAlaAlaPheXaaNH2] / [MalAlaAIaPheOH] is 1 (normally denoted as p) can be estimated using the formula: p = (55k~K,)/k4, the values of ( k ~ K , ) / k 4 can be taken from Tables I and II.

HPLC analysis Isocratic or gradient HPLC was carried out using a Shimadzu HPLC system (LC-6A pumps, SPD-6A UV detector, SCL-6A system controller, C-R4A chromatopac integrator) equipped with a 25 cm LiChrospher 100 RP-18 (5 m) column (Merck). The mobile phase contained different proportions of water and acetonitrile and in all cases 5% (v/v) of acetic (in isocratic chromatography) or trifluoroacetic acid (in gradient chromatography). Gradient elution was used when studying reactions with ArgNH2, SerNH2, PheNH 2 and TyrNH 2. UV detection was at 254 nm. For all reactions with the exception of acyl-transfer to

PheNH2, TyrNH 2 and TrpNH2, the molecular absorption coefficient (Ez54) of the peptide product was considered to be independent of the peptide structures and equal to e2s4 of MalAlaAlaPheOMe and MalAlaAlaPheOH. Amides of phenylalanine, tyrosine and tryptophane absorb light at 254 nm. In this case E254 of the peptide formed was considered to be the sum of the molecular absorption coefficients of MalAlaAlaPheOH and the amino-acid amides. Measurements were carried out using a Shimadzu UV-260 spectrophotometer. The following values of E254 were obtained: for MalAlaAlaPhePheNH2 e254 = 2300 M - i cm-1, for MalAlaAlaPheTyrNH 2 2600 M - l c m - 1, for MalAlaAlaPheTrpNH2 4800 M - ~cm- 1 and for MalAlaAlaPheOMe 2100 M-~cm -1. The latter value of e254 was supposed to be valid for the other peptides formed and MalAlaAlaPheOH.

Calculations Under the conditions used, hydrolysis of the peptide products was not detected. Therefore, the following scheme describes the processes (for details, see, for example, Refs. 6 and 29).

Pl E+S ~

Ks

)ES

k2 [

EA

k3[H20] ) E + P 2 +N

N~K~k~ t, I /

EAN

[H20] k4

~E + P

where E is the enzyme, S is the donor (MalAlaAlaPheOMe) of the acyl moiety (MalAlaAlaPhe) to be transferred to the added nucleophile N (amino-acid amide); ES is the complex of the enzyme with S, P denotes the peptide synthetic product (MalAlaAlaPheXaaNH2) , EA represents the acylenzyme intermediate, EAN is the complex of EA with the added nucleophile, P1 is methanol and P2 is MalAIaAIaPheOH. The meaning of the constants follows from the scheme. The height of the energy barrier between the acylenzyme and the synthetic product (P) is a measure of the nucleophile specificity of the enzyme. This height is proportional to k 4 / K n or, since we use the same acyl-group donor, to ka/(k~K,). With a significant excess of the nucleophile, the following relationship can be obtained: [N][A2IR 55 [A]

k;K. - k4

+

k~lN] k4

-

(1)

where A and A 2 denote the peak areas of the peptide P and the hydrolytic product P2, R represents E254 of the peptide P devided by E254 of MalAlaAlaPheOH. This equation was used in calculations.

190 TABLE I

TABLE II

Reactivity of amino-acid amides in ct-chymotrypsin-catalyzed reactions

Reactivity of amino-acid amides in reactions catalyzed by subtilisin from Bacillus subtilis strain 72

Conditions: pH 9.0 (0.1 M veronal), 30°C, 1% (v/v) DMSO, [MalAlaAlaPheOMe] 0 = 1 mM. Kinetic symbols correspond to the scheme; S.D., standard deviation. Amide

k 4 / ( k ~ K n)

S.D.

D-AIaNH 2 GIy-NH 2 AIaNH 2 SerNH 2 ThrNH 2 HisNH 2 ValNH 2 LeuNH 2 TrpNH 2 MetNH 2 NvaNH 2 PheNH 2 ILeNH 2 TyrNH 2 ArgNH 2

89 292 500 660 690 800 2 330 2 750 3 000 3 050 4 200 5 260 5 800 7 580 15 800

6 13 68 82 28 100 330 490 400 230 1300 850 900 900 6 800

Results

Conditions: pH 9.0 (0.1 M veronal), 30°C, 1% (v/v) DMSO, [MalAlaAlaPheOMe]0 = 1 mM. Kinetic symbols correspond to the scheme; S.D., standard deviation. For T r p N H 2, only two measurements were carried out; the data presented is the arithmetic mean of these values. Amide

k 4 / ( k ~ K n)

S.D.

LeuNH 2 IIeNH 2 ThrNH 2 TrpNH 2 NvaNH 2 ValNH 2 MetNH 2 AlaNH 2 SerNH 2 D-AlaNH 2 ArgNH 2 GIyNH 2

13 18 19 31 62.5 67 72 82 88 100 118 420

2.5 6.4 3.6 5.4 4 12 15 5 5 25 20

from water. The values used in this work were determined [31] according to the following equation:

Plotting our results according to Eqn. 1 showed that the reactivity of the nucleophiles was as a rule independent of the nucleophile concentration, i.e., k~--0. a-Chymotrypsin-catalyzed transfer of the acyl group to PheNH 2 and TyrNH 2 were the exceptions. In-depth analysis of the dependencies of the nucleophile reactivity on the nucleophile concentration for different acylgroup donor and nucleophiles is now in progress. The preliminary results show that the acyl-transfer reactions catalyzed by proteinases are more complex than it is usually assumed. The nucleophile reactivities obtained in this work are summarized in Tables I and II. The values for subtilisin-catalyzed reactions are normally lower than those for a-chymotrypsin-catalyzed reactions. For indepth understanding of the factors influencing nucleophile reactivity, we attempted to correlate our data with different characteristics of amino-acid side-chains available in the literature.

~side-chain = log P(A. . . . i. . . . . id amide) -- log P(Ac-glyci. . . . ide)

Amino acid

zr

V (A)

o'Hc (ppm)

Discussion

Ala Arg Gly His Ile Leu Met Nva Phe Ser Thr Trp Tyr Val

0.31 - 1.01 0.00 0.13 1.80 1.70 1.23 1.37 1.79 - 0.04 0.26 2.25 0.96 1.22

0.52 0.68 0.00 0.70 1.02 0.98 0.78 0.68 0.70 0.53 0.50 0.70 0.70 0.70

7.3 11.1 0.00 10.2 16.1 10.1 10.4 10.3 13.9 13.1 16.7 13.2 13.9 17.2

Numerous amino-acid side-chain parameters are described in the literature [30]. Each of them can be used in QSAR studies. In the discussion below, we shall try to correlate our results with the hydrophobicity (~-), the upsilon steric parameter (V) and the NMR ehemical shift of the a-carbon. We chose these parameters because they describe different effects of substituents, i.e., hydrophobic, steric and electronic. The hydrophobicity reflects the ability of active site of the enzyme to extract the given amino-acid residue

where P is the partition coefficient of the compound in the water-octanol two-phase system. The upsilon steric parameter is a function of the Bondi/van der Waals radii and can be held as a reliable measure of steric effects [32]. The NMR shift of the a-carbon (trHc) describes the electronic properties of the side-chain because it primarily reflects the shielding of the Ca-nucleus by the nearby electronic systems [33]. All these parameters are listed in Table III for convenience. TABLE III

Some parameters of amino-acid side chains [30] 7r, hydrophobicity; V, upsilon steric parameter, trHc, NMR chemical shift of a-carbon.

191 Plotting the data obtained shows that good correlation exists in one case only. Fig. 1 shows that in a-chymotrypsin-catalyzed reactions the nucleophile reactivities (here, log(k4/(k3K,))) correlate with the hydrophobicity of amino-acid side-chains for all reactions with the exception of acyl-transfer to ArgNH 2, DAlaNH 2. Moreover, experimental points for GIyNH 2, AIaNH z, SerNH 2, ThrNH 2, HisNH 2, ValNH 2, LeuNH2, MetNH2, NvaNH2, PheNH 2 and IIeNH 2 were around the straight line with a unit slope. The points for TyrNH 2 and TrpNH 2 were a little apart from this line but these deviations are not significant. Therefore, the data obtained show that hydrophobic interactions are entirely responsible for the difference between the reactivity of the different nucleophiles in a-chymotrypsin-catalyzed reactions for practically all amino-acid amides. Early, Schellenberger et al. [10] proposed this phenomenon, but insufficient data available at that time did not allow them to draw quantitative conclusions. The higher reactivity of ArgNH 2 is at all appearances the result of ionic interactions between two aspartic-acid residues (Asp-64 and Asp-35) and the positively charged side-chain of the arginine nucleophile [10]. Incorrect orientation of a methyl group in the complex of MalAlaAlaPhe-chymotrypsin with DAlaNH 2 would hamper productive binding, which is probably the reason for the low reactivity of the latter compound. Only a qualititative correlation can be made between log(k,a/(k~Kn)) and the upsilon steric parameter (V) (Fig. 2). This means that the size of the 10

&

>,

I1

°~

> o~

m 4) k.

4)

. ~u

>

O0

o

7'

o r-

5

0.0

|

!

0.4

0.8

1.2

v (A) Fig. 2. Correlation between the upsilon steric parameter and log( k 4 / ( k ~K, )) (nucleophile reactivity) for a-chymotrypsin-catalyzed reactions. Conditions: pH 9.0 (0.1 M veronal), 30°C, 1% (v/v) DMSO. Data were taken from Tables I and IlL

substituent affects the nucleophile reactivity indirectly, primarily by changing the hydrophobicity of the nudeophile molecule. Electronic properties (trH c) of the nucleophiles do not influence their reactivity. The molar refractivity (MR) is another popular parameter in QSAR studies of a-chymotrypsin-catalyzed reactions [34-36]. In this work, we did not find good correlation between the nucleophile reactivity of the amino-acid amides and MR. It is not suprising, since MR is a bulk parameter like V. No correlation between amino-acid side-chain parameters and nucleophile reactivities were found for acyl-transfer reactions catalyzed by subtilisin 72. We only became firmly convinced of the negative influence of the size of the substituent of the nucleophile molecule on transfer reactions. The high reactivity of D-AlaNH 2 (even higher than AIaNH 2) affirms the low selectivity of the S'~-subsite of subtilisin 72.

°~

rQ. 0 o

6

Conclusions

o

&

4

!

.5

0.5

!

2.5

hydrophobicity Fig. 1. Correlation between the hydrophobicity (~') and log(k4/ (k~Kn)) (nucleophile reactivity) for a-chymotrypsin-catalyzed reactions. Conditions: pH 9.0 (0.1 M veronal), 30°C, 1% (v/v) DMSO. Nucleophile reactivity of ArgNH 2 and D-AIaNH 2 is denoted as (zx). Data were taken from Tables I and III.

This work is an attempt towards a systematic study of structure-activity relationships of nucleophiles participating in acyl-transfer reactions catalyzed by achymotrypsin and subtilisin from Bacillus subtilis strain 72. The results show the importance of hydrophobic interactions in the Sl-subsite in a-chymotrypsin catalysis. Previously, a similar phenomenon was established for the Sl-subsite [37,38] of this enzyme. The size of the nucleophile molecule is not very important for reactivity. In reactions catalyzed by subtilisin 72, no definite correlation between the reactivity of the nu-

192 cleophiles and amino-acid side-chain parameters was found. These data support our published results on the specificity of this enzyme in hydrolytic reactions [25], which indicate that the S'l-subsite of subtilisin 72 is of broad selectivity.

Acknowledgements We wish to thank I.P. Morozova for purification of subtilisin 72, E.Yu. Terent'eva for her help in synthetic work, Scott Bloomer for linguistic advice and the Swedish Institute for the financial support of M.Yu. Gololobov.

References 1 Fastrez, J. and Fersht, A.R. (1973) Biochemistry 12, 2025-2034. 2 Fersht, A.R., Blow, D.M. and Fastrez, J. (1973) Biochemistry 12, 2035-2041. 3 Riechmann, L. and Kasche, V. (1984) Biochem. Biophys. Res. Commun. 120, 686-691. 4 Riechmann, L. and Kasche, V. (1985) Biochim. Biophys. Acta 830, 164-172. 5 Petkov, D.D. and Stoineva, I. (1984) Biochem. Biophys. Res. Commun. 118, 317-323. 6 Gololobov, M.Yu., Petrauskas, A., Pauliukonis, R., Koschke, V., Borisov, I.L. and Svedas, V. (1990) Biochim. Biophys. Acta 1041, 71-78. 7 Bratovanova, E.K. and Petkov, D.D. (1987) Biochem. J. 248, 957-960. 8 Clapes, P. Adlercreutz, P. and Mattiasson, B. (1990) Biotechnol. Appl. Biochemistry 12, 376-386. 9 Bratovanova, E.K., Stoineva, I.B. and Petkov, D.D. (1988) Tetrahedron 44, 3633-3637. 10 Schellenberger, V., Schellenberger, U., Mitin, Yu.V. and Jakubke, H.-D. (1990) Eur. J. Biochem. 187, 163-167. 11 Schellenberger, V., Kost, M., Jakubke, H.-D. and Aaviksaar, A. (199l) Biochim. Biophys. Acta. 1078, 1-7. 12 Kasche, V., Michaelis, G. and Galunsky, B. (1991) Biotechnol. Lett. 13, 75-80. 13 Kullmann, W. (1983) J. Protein Chem. 2, 289-301. 14 Aso, K. (1989) Agr. Biol. Chem. 53, 729-733. 15 Stoineva, I.B. and Petkov, D.D. (1985) FEBS Lett. 183, 103-106.

16 Voyushina, T.L., Lyublinskaya, L.A. and Stepanov, V.M. (1985) Bioorg. Khim. 11,738-744. 17 Barbas, C.F., III and Wong, C.-H. (1987) J. Chem. Soc. Chem. Commun. N8, 533-534. 18 Margolin, A., Tai, D.-F. and Klibanov, A, (1987) J. Am. Chem. Soc. 109, 7885-7887. 19 Riva, S., Chopineau, J., Kieboom, A.P.G. and Klibanov, A.M. (1988) J. Am. Chem. Soc. 110, 584-589. 20 Voyushina, T.L., Terent'eva, E.Yu., Pozdnev, V.F., Gaida, A,V., Gololobov, M.Yu., Lyublinskaya, L.A. and Stepanov, V.M. (1991) Bioorg. Khim. 17, 1066-1073. 21 Baumann, W.K., Bizzozero, S.A. and Dutler, H. (1970) FEBS Lett. 8, 257-260. 22 Bizzozero, S., Baumann, W.K. and Dutler, H. (1982) Eur. J. Biochem. 122, 251-258. 23 Hill, C.R. and Tomalin, G. (1983) Arch. Biochem. Biophys. 221, 324-328. 24 Schellenberger, V., Braune, K., Hofmann, H.-J. and Jakubke, H.D. (1991) Eur. J. Biochem. 199, 623-636. 25 Gololobov, M.Yu., Morozova, I.P., Voyushina, T.L., Timokhina, E.A. and Stepanov, V.M. (1992) Biochim. Biophys. Acta 1118, 267-276. 26 Schechter, I. and Berger, A. (1967) Biochem. Biophys. Res. Commun. 27, 157-162. 27 Gololobov, M.Yu., Kozlova, E.V., Borisov, I.L., Schellenberger, V., Schellenberger, U. and Jakubke, H.-D. (1992) Biotechnol. Bioeng. 40, 432-436. 28 Schellenberger, V., Schellenberger, U. and Jakubke, H.-D. (1988) Collect. Czech. Chem. Commun. 53, 2884-2889. 29 Gololobov, M.Yu., Borisov, I.L. and Svedas, V.K. (1989) J. Theor. Biol. 140, 193-204. 30 Fauchere, J.L. Charton, M., Kier, L.B., Verloop, A. and Pliska, V. (1988) Int. J. Peptide Protein Res. 32, 269-278. 31 Fauchere, J.L. and Pliska, V. (1983) Eur. J. Med. Chem. 18, 369-375. 32 Charton, M. (1981) J. Theor. Biol. 91, 115-123. 33 Fauchere, J.L. and Lauterwein, J. (1985) Quant. Struct. Act. Rel. 4, 11-13. 34 Clapes, P. and Adlercreutz, P. (1992) Biochim. Biophys. Acta 1118, 70-76. 35 Hansch, C., Grieco, C., Silipo, C., Smith, R.N. and Vittoria, A. (1977) J. Med. Chem. 20, 1420-1435. 36 Grieco, C., Hansch, C., Silipo, C., Smith, R.N., Vittoria, A. and Yamada, K. (1979) Arch. Biochem. Biophys. 194, 542-551. 37 Berezin, I.V. and Martinek, K. (1970) FEBS Lett. 8, 261-262. 38 Dorovska, V.N., Varfolomeev, S.D., Kazanskaya, N.F., Klyosov, A.A. and Martinek, K. (1972) FEBS Lett. 23, 122-124.