Biothimlta el Biophysica Acta, 1078 (1991) 1-7 ,,~ 1991 ElsevierScience Publishers B.V. 0167-4838/91/$03.50 ADONIS 016748389100180F
BBAPRO 33904
Electrostatic effects in the o~-chymotrypsin-catalyzed acyl transfer. I. Influence of different inorganic salts Volker Schellenberger i, Marika Kosk 2, H a n s - D i e t e r Jakubke ~ and Aavo Aaviksaar 2 t Deparonent of Biochemistry. Biosciences Dwision, LeipzJg Unwer~it~', Lelpctg (F.R.G.) and 2 Laboralory ofBit;~rgunic Chcmtst,% Institute of Chemical Physics and Biophysics of the Estonian Aeader,~v of Soen~es. l'ullinn (U.SS.R.)
(Received22 May 1991)) (Revisedmanusciipt received28 November 19901
Key words: a-Chymotrypsin:Acyl transfer: Charged nucleophile: Electrostatic interaction; Sail el'feet: Cation binding
We investigated the deacylation of two acyl-a-chymotrypsins by added nucleophiles. The nucleophile binding sile of the enzyme shows a strong preference for positively charged compounds. Most of our data can be explained by direct electrostatic interaction between the ionic nucleophiles and two negatively charged residues which are located close to the active site of the enzyme molecule. The influence of inorganic sal~ on the affl transfer includes the following effeots: (I) reduction of electrostatic interactions between the aeyl-enzyme and the nucleophile by addition of salts; (2) binding of divalent cations to the nueleophile binding site of the acyl-enzyme leading to a si~ificantly changed specificity.; and (3) linear dependence of the activity coefficients of the added nucleophiles on salt concentration.
Introduction Electrostatic interactions are of great importance in enzyme catalysis and in the interaction of proteins with other molecules in general [1-4]. Since the protein is a low dielectric medium [5], and the solvent is a high dielectric medium, the protein surface constitutes a dielectric interface which makes the system extremely complicated. A number of computer modeling approaches have been developed to predict the intensity of electrostatic interactions on protein surfaces [6,7], but the results are far from satisfactory. The presence of inorganic salts in solution affects charge-charge interactions on a protein surface as coulombic screening of electrostatic potentials arises from counterions in the vicinity of the protein. The salt effects may be complicated by specific interactions of individual ion species with the protein. Experiments
Abbreviations: Bz. henzoyl; Mal, maleyl-(3-carboxyacryloyl);OMe, methyl ester; NH 2, amide; symbols: E enzyme; S, acyl donor ester; N, added nucleophile; EA. acyl enzyme; EAt '+, acyt-enzymc-eation complex; FAN, acyl-eazyme-nucleuphilecomplex; EAI~+ N, acyl-enzyme-cation-nuclcophilecomplex; Pl, leavinggroup of the acyl donor
ester; P2, hydrolysisproduct.Ps. amlnolysisproduct. Correspondence:H-D.Jakubke. Departmentof Biochemistry,Biosci. encesDivision,LeipzigUniversity,Talstr. 33, Leipzig,O-7010,F.R.G.
which allow us to evaluate the influence of individual charges on catalysis can lead to a better understanding of electrostatic interactions on protein surfaces [2]. Studies on a-chymotrypsin-catalyzed acyl transfer to added nucleophiles have shown that the nucleophile efficiency is strongly dependent on its charge [8-10]. H-Arg-NH 2 was found to be an extremely good nucleophil¢ whereas dipeptides with a free negatively charged carboxyl group were of low efficiency. These results led us to the conclusion that electrostatic interactions must play an important role in the deacylation of acyl-ct-chymotrypsin by charged nucleophiles. For proteinase-catalyzed acyl transfer reactions it is widely accepted that nucleophiles interact with the S'subsites of the enzyme (binding site notation according to Schechter and Berger [11]) in the same way as the leaving group in peptide hydrolysis [12,13]. According to crystallographic data [14-161, two negatively charged residues, Asp-35 and Asp-64, are located in the range of the S'-subsites of n-chymotrypsin. The electrostatic field of these groups could be the reason for the preference of a-chymotrypsin for positively charged nudeophiles. In an earlier paper it has been shown that the addition of KC1 to the reaction medium improves the nucleophilic behaviour of dipeptides towards acyl-achymotrypsins by more than one order of magnitude [9]. In the present work we compare the influence of various inorganic salts on the tt-chymotrypsin-catalyzed
a%l transfer to the negatively charged H-Ala-Ala-OH and to the positively charged H.Arg-NH:. The effect of a negatively charged group in the acyl part on the efficiency o1 the nucleophiles was also determined. Materials and Methods
Chemicals Bovine a-chymotrypsin (EC 3.4.21.1) (three times crystallized, research grade, lot 24046) was purchased from Serva (Heidelberg, F.R.G.). Bz-Tyr-OMe and HAla-Ala-OH were from Reanal (Hungary). Mal-TyrOMe and H-Arg-NH 2 were from our collection. Inorganic salts were analytical grade from Reakhim (U.S.S.R.).
Ao,I transfer reactions Enzymatic acyl transfer reactions were performed in a pH-stat, (TIT 80, Radiometer, Denmark) to avoid the use of buffers. The reactions with H-Arg-NH, were performed in 3 ml, all other reactions in 1 ml reaction vols. The reaction mixtures containing the appropriate salt. nucleophile and substrate ester were prepared from stock solutions of the components. 60 mM substrate ester stock solutions in DMSO were used. The final substrate concentrations were between 0.03 and 0.3 mM, i.e. the DMSO concentration in the reaction mixture was 0.5% or lower. The concentration of nucleophile was chosen depending on the partition constant in the range of 0.3 Papp to 2 Papp,but not higher than 150 mM. The mixtures were thermostated at 25 °C and the reaction was started by addition of 1-5 #1 a-chymotrypsin stock solution, a-Chymotrypsin was dissolved in 1 mM HCL The concentration of the enzyme stock solution was 10 mg/ml for the reactions with Mal-Tyr-OMe and 0.5 mg/ml for the i,,,,,,d~,ns wii.h Bz-Tyr-OMe. In order to prevent secondary hydrolysis of the peptide product, enzyme concentrations providing approx. 80% substrate ester conversion within the reaction time were u~d. After 1-3 min reaction time 100 ~tl of the reaction mixture were withdrawn and analyzed by HPLC.
uct ratios. The standard deviation of the repeated experiments was less than 3% if performed on the same day and less than 5,% ~f performed on different days. The slightly reduced long-term reproducibility seems to be caused by small changes of the column features over time. The reproducibility was also slightly reduced for experiments in which very small product amounts had to be quantified, i.e. when the p constants were higher than 300 raM. In all cases, the salt effects discussed in the paper exceeded the error limit by at least one order of magnitude. Results
Serine proteinase-catalyzed acyl transfer reactions are usually described by Scheme l, i.e. the added nuclcophile forms a complex with the acyl-enzyme that ca t undergo aminolysis. The ratio between aminolysis and hydrolysis of the acyl-enzyme reflects the S'-subsite specificity of the enzyme I121. As an efficiency' parameter for the added nucleophile in acyl transfer reactions we introduced the partition constant p [17] which is defined according to Eqn. 1: t'~
INI
t'H
p
(1)
p gives the nucleophile concentration for which va = an. If the nucleophite is used in a great excess over the aeyl donor the product ratio [P3I/IP21 is constant over the whole reaction time and equal to the ratio va/v n. Thus, p can be calculated from the product ratio according to Eqn. 2: p = INIIP~I/LP3]
(2)
For the model reactions under investigation p did not depend on the nucleophile concentration. The only exception we observed for reactions at high concentrations of H-Arg-NH 2 and [salt] < 0.1 M. This effect became significant at product ratios [P,I/IP2I > 5. It may result from hydrolysis of the acyl-enzyme-
ItPLC analyses Analyses were performed using a Series 8800 gradient system (DuPont Instruments. U.S.A.) connected with a computing integrator SP4100 (Spectra-Physics, U.S.A.) and a 4.6 × 150 mm Zorbax ODS column. Absorbance was monitored at 250 nm. Since the hydrolysis product and the synthesis product contain the same chromophoric system, abso~gtion coefficients were assumed to be equal. Product analyses were performed under isocratic regime using various mixtures of 0.1% phosphoric acid and ethanol or acetonitrile, respectively. Repeated runs of a number of reactions demonstrated the good reproducibility of the determined prod-
E+~ KS
E+S~-~
~2
ES ~
EA
FAN
E+P~ Sche~e I. Serineproteinase-catalyzedacyltransferreaction.
nucleophile complex EAN or from unspecific mtcractions between the nuuieophile and the acyl-enzyme. In all further experiments we determined the partition constant p at nucleophile concentrations leadlng to product ratios [P3]/[P2] not higher than :. The ratio t'a/% shows a linear dependence on nuclecphile concentration under such conditions and the partition constare is given as: p k~gx/k~
(3)
=
In order to compare the salt effects in four model reactions, Fig. 1 shows a plot of Iog(p~pp) versus ionic strength, it can be seen that all three salts reduce the nueleophilic efficiency of H-Arg-NH2 whereas the efficiency of H-Ala-Ala-OH in a-chymotrypsin deacylation is improved. This effect is in agreement with our assumption that the aminolysis of acyl-a-chymotrypsin by charged nucleophiles is mainly affected by a negative electrostatic potential in the nucleophile-binding-site of the enzyme. The influence of salts, however, cannot be described by changes in the activity coefficients of the reactants in terms of the Debye-Hi~clde theory [18] because the individual salts affect the acyl transfer to a different extent.
In accordance with the literature [19 22] sah~ have small or no influence on the rate-limiting deacylamm step in the chymutryptic hydrolysis of specific ester substrates. We found that KCI (0.2 3 M} or CaCI, (0.1-2 M) did not afk'ct k,~t for the nydroiysis of Bz-Tyr-OMe while kc, , for the hydrolysis of Mal-TyrOMe slightly increased from 140 s t at 0.1 M KCI to 160 s-~ at 3.2 M KC1. This is much less than the observed changes in the p values. Thus. the changes in the acyl transfer efficiency can be attributed to the influence of salt on the acyl-enzyme aminolysis. Our data allow no decision whether the salts affect k+ or KN. The a-amino groups of the nuclcophiles are involved in protonation equilibria, and only the nonprotonated species are effective in deacylation. We found pK~ values of 8.22 at 0.01 M KC1 and 8.41 at 3 M KCI fc,r H-AIa-AIa-OH, and 721 at 0.01 M KCI and 7.87 at 3 M KCI for H-Arg-NH 2. Consequently, both nucleophiles are almost completely deprotonized at pH 9.0 and the concentration of the nonprotonated nucleophile can be consider~ to be independent upon the salt concentration.
DeaGlation by H-Arg-NH,_ Influence of monovalent cations
Identification of the salt dependent reaction step The partition constant represents the quotient betwcen the rate constants of hydrolysis and aminolysis of the acyl-enzyme, in order to identify the salt-dependent reaction we investigated the k:netics of the acyl donor hydrolysis.
.Bz-T),r-
BBAPRO 33904
Electrostatic effects in the o~-chymotrypsin-catalyzed acyl transfer. I. Influence of different inorganic salts Volker Schellenberger i, Marika Kosk 2, H a n s - D i e t e r Jakubke ~ and Aavo Aaviksaar 2 t Deparonent of Biochemistry. Biosciences Dwision, LeipzJg Unwer~it~', Lelpctg (F.R.G.) and 2 Laboralory ofBit;~rgunic Chcmtst,% Institute of Chemical Physics and Biophysics of the Estonian Aeader,~v of Soen~es. l'ullinn (U.SS.R.)
(Received22 May 1991)) (Revisedmanusciipt received28 November 19901
Key words: a-Chymotrypsin:Acyl transfer: Charged nucleophile: Electrostatic interaction; Sail el'feet: Cation binding
We investigated the deacylation of two acyl-a-chymotrypsins by added nucleophiles. The nucleophile binding sile of the enzyme shows a strong preference for positively charged compounds. Most of our data can be explained by direct electrostatic interaction between the ionic nucleophiles and two negatively charged residues which are located close to the active site of the enzyme molecule. The influence of inorganic sal~ on the affl transfer includes the following effeots: (I) reduction of electrostatic interactions between the aeyl-enzyme and the nucleophile by addition of salts; (2) binding of divalent cations to the nueleophile binding site of the acyl-enzyme leading to a si~ificantly changed specificity.; and (3) linear dependence of the activity coefficients of the added nucleophiles on salt concentration.
Introduction Electrostatic interactions are of great importance in enzyme catalysis and in the interaction of proteins with other molecules in general [1-4]. Since the protein is a low dielectric medium [5], and the solvent is a high dielectric medium, the protein surface constitutes a dielectric interface which makes the system extremely complicated. A number of computer modeling approaches have been developed to predict the intensity of electrostatic interactions on protein surfaces [6,7], but the results are far from satisfactory. The presence of inorganic salts in solution affects charge-charge interactions on a protein surface as coulombic screening of electrostatic potentials arises from counterions in the vicinity of the protein. The salt effects may be complicated by specific interactions of individual ion species with the protein. Experiments
Abbreviations: Bz. henzoyl; Mal, maleyl-(3-carboxyacryloyl);OMe, methyl ester; NH 2, amide; symbols: E enzyme; S, acyl donor ester; N, added nucleophile; EA. acyl enzyme; EAt '+, acyt-enzymc-eation complex; FAN, acyl-eazyme-nucleuphilecomplex; EAI~+ N, acyl-enzyme-cation-nuclcophilecomplex; Pl, leavinggroup of the acyl donor
ester; P2, hydrolysisproduct.Ps. amlnolysisproduct. Correspondence:H-D.Jakubke. Departmentof Biochemistry,Biosci. encesDivision,LeipzigUniversity,Talstr. 33, Leipzig,O-7010,F.R.G.
which allow us to evaluate the influence of individual charges on catalysis can lead to a better understanding of electrostatic interactions on protein surfaces [2]. Studies on a-chymotrypsin-catalyzed acyl transfer to added nucleophiles have shown that the nucleophile efficiency is strongly dependent on its charge [8-10]. H-Arg-NH 2 was found to be an extremely good nucleophil¢ whereas dipeptides with a free negatively charged carboxyl group were of low efficiency. These results led us to the conclusion that electrostatic interactions must play an important role in the deacylation of acyl-ct-chymotrypsin by charged nucleophiles. For proteinase-catalyzed acyl transfer reactions it is widely accepted that nucleophiles interact with the S'subsites of the enzyme (binding site notation according to Schechter and Berger [11]) in the same way as the leaving group in peptide hydrolysis [12,13]. According to crystallographic data [14-161, two negatively charged residues, Asp-35 and Asp-64, are located in the range of the S'-subsites of n-chymotrypsin. The electrostatic field of these groups could be the reason for the preference of a-chymotrypsin for positively charged nudeophiles. In an earlier paper it has been shown that the addition of KC1 to the reaction medium improves the nucleophilic behaviour of dipeptides towards acyl-achymotrypsins by more than one order of magnitude [9]. In the present work we compare the influence of various inorganic salts on the tt-chymotrypsin-catalyzed
a%l transfer to the negatively charged H-Ala-Ala-OH and to the positively charged H.Arg-NH:. The effect of a negatively charged group in the acyl part on the efficiency o1 the nucleophiles was also determined. Materials and Methods
Chemicals Bovine a-chymotrypsin (EC 3.4.21.1) (three times crystallized, research grade, lot 24046) was purchased from Serva (Heidelberg, F.R.G.). Bz-Tyr-OMe and HAla-Ala-OH were from Reanal (Hungary). Mal-TyrOMe and H-Arg-NH 2 were from our collection. Inorganic salts were analytical grade from Reakhim (U.S.S.R.).
Ao,I transfer reactions Enzymatic acyl transfer reactions were performed in a pH-stat, (TIT 80, Radiometer, Denmark) to avoid the use of buffers. The reactions with H-Arg-NH, were performed in 3 ml, all other reactions in 1 ml reaction vols. The reaction mixtures containing the appropriate salt. nucleophile and substrate ester were prepared from stock solutions of the components. 60 mM substrate ester stock solutions in DMSO were used. The final substrate concentrations were between 0.03 and 0.3 mM, i.e. the DMSO concentration in the reaction mixture was 0.5% or lower. The concentration of nucleophile was chosen depending on the partition constant in the range of 0.3 Papp to 2 Papp,but not higher than 150 mM. The mixtures were thermostated at 25 °C and the reaction was started by addition of 1-5 #1 a-chymotrypsin stock solution, a-Chymotrypsin was dissolved in 1 mM HCL The concentration of the enzyme stock solution was 10 mg/ml for the reactions with Mal-Tyr-OMe and 0.5 mg/ml for the i,,,,,,d~,ns wii.h Bz-Tyr-OMe. In order to prevent secondary hydrolysis of the peptide product, enzyme concentrations providing approx. 80% substrate ester conversion within the reaction time were u~d. After 1-3 min reaction time 100 ~tl of the reaction mixture were withdrawn and analyzed by HPLC.
uct ratios. The standard deviation of the repeated experiments was less than 3% if performed on the same day and less than 5,% ~f performed on different days. The slightly reduced long-term reproducibility seems to be caused by small changes of the column features over time. The reproducibility was also slightly reduced for experiments in which very small product amounts had to be quantified, i.e. when the p constants were higher than 300 raM. In all cases, the salt effects discussed in the paper exceeded the error limit by at least one order of magnitude. Results
Serine proteinase-catalyzed acyl transfer reactions are usually described by Scheme l, i.e. the added nuclcophile forms a complex with the acyl-enzyme that ca t undergo aminolysis. The ratio between aminolysis and hydrolysis of the acyl-enzyme reflects the S'-subsite specificity of the enzyme I121. As an efficiency' parameter for the added nucleophile in acyl transfer reactions we introduced the partition constant p [17] which is defined according to Eqn. 1: t'~
INI
t'H
p
(1)
p gives the nucleophile concentration for which va = an. If the nucleophite is used in a great excess over the aeyl donor the product ratio [P3I/IP21 is constant over the whole reaction time and equal to the ratio va/v n. Thus, p can be calculated from the product ratio according to Eqn. 2: p = INIIP~I/LP3]
(2)
For the model reactions under investigation p did not depend on the nucleophile concentration. The only exception we observed for reactions at high concentrations of H-Arg-NH 2 and [salt] < 0.1 M. This effect became significant at product ratios [P,I/IP2I > 5. It may result from hydrolysis of the acyl-enzyme-
ItPLC analyses Analyses were performed using a Series 8800 gradient system (DuPont Instruments. U.S.A.) connected with a computing integrator SP4100 (Spectra-Physics, U.S.A.) and a 4.6 × 150 mm Zorbax ODS column. Absorbance was monitored at 250 nm. Since the hydrolysis product and the synthesis product contain the same chromophoric system, abso~gtion coefficients were assumed to be equal. Product analyses were performed under isocratic regime using various mixtures of 0.1% phosphoric acid and ethanol or acetonitrile, respectively. Repeated runs of a number of reactions demonstrated the good reproducibility of the determined prod-
E+~ KS
E+S~-~
~2
ES ~
EA
FAN
E+P~ Sche~e I. Serineproteinase-catalyzedacyltransferreaction.
nucleophile complex EAN or from unspecific mtcractions between the nuuieophile and the acyl-enzyme. In all further experiments we determined the partition constant p at nucleophile concentrations leadlng to product ratios [P3]/[P2] not higher than :. The ratio t'a/% shows a linear dependence on nuclecphile concentration under such conditions and the partition constare is given as: p k~gx/k~
(3)
=
In order to compare the salt effects in four model reactions, Fig. 1 shows a plot of Iog(p~pp) versus ionic strength, it can be seen that all three salts reduce the nueleophilic efficiency of H-Arg-NH2 whereas the efficiency of H-Ala-Ala-OH in a-chymotrypsin deacylation is improved. This effect is in agreement with our assumption that the aminolysis of acyl-a-chymotrypsin by charged nucleophiles is mainly affected by a negative electrostatic potential in the nucleophile-binding-site of the enzyme. The influence of salts, however, cannot be described by changes in the activity coefficients of the reactants in terms of the Debye-Hi~clde theory [18] because the individual salts affect the acyl transfer to a different extent.
In accordance with the literature [19 22] sah~ have small or no influence on the rate-limiting deacylamm step in the chymutryptic hydrolysis of specific ester substrates. We found that KCI (0.2 3 M} or CaCI, (0.1-2 M) did not afk'ct k,~t for the nydroiysis of Bz-Tyr-OMe while kc, , for the hydrolysis of Mal-TyrOMe slightly increased from 140 s t at 0.1 M KCI to 160 s-~ at 3.2 M KC1. This is much less than the observed changes in the p values. Thus. the changes in the acyl transfer efficiency can be attributed to the influence of salt on the acyl-enzyme aminolysis. Our data allow no decision whether the salts affect k+ or KN. The a-amino groups of the nuclcophiles are involved in protonation equilibria, and only the nonprotonated species are effective in deacylation. We found pK~ values of 8.22 at 0.01 M KC1 and 8.41 at 3 M KCI fc,r H-AIa-AIa-OH, and 721 at 0.01 M KCI and 7.87 at 3 M KCI for H-Arg-NH 2. Consequently, both nucleophiles are almost completely deprotonized at pH 9.0 and the concentration of the nonprotonated nucleophile can be consider~ to be independent upon the salt concentration.
DeaGlation by H-Arg-NH,_ Influence of monovalent cations
Identification of the salt dependent reaction step The partition constant represents the quotient betwcen the rate constants of hydrolysis and aminolysis of the acyl-enzyme, in order to identify the salt-dependent reaction we investigated the k:netics of the acyl donor hydrolysis.
.Bz-T),r-
