Activity of Thiolsubtilisin in Organic Solvents - American Chemical

Sudipta Chatterjeet and Alan J. Russell*q+J. Department of Chemical Engineering, University of Pittsburgh, Pittsburgh, Pennsylvania 15261, and Center...
0 downloads 0 Views 355KB Size
Biotechnoi. hog. 1992, 8, 256-258

256

Activity of Thiolsubtilisin in Organic Solvents Sudipta Chatterjeet and Alan J. Russell*q+J Department of Chemical Engineering, University of Pittsburgh, Pittsburgh, Pennsylvania 15261, and Center for Biotechnology and Bioengineering, University of Pittsburgh, Pittsburgh, Pennsylvania 15261

In this paper, we compare the activities of subtilisin and thiolsubtilisin in anhydrous organic solvents and in mixtures of water and miscible organic media. The information obtained indicates that changing from one organic solvent to another does not significantly affect the stability of the transition state for the reaction. Thus, since the transition state is charged, it is likely that the active site of the enzyme is electrostatically protected from bulk solvent. Further, since the activities of subtilisin and thiolsubtilisin are different in the solvents used, it is clear that the active site serine is a catalytically important residue in both water and organic solvents.

The use of proteases and lipases in essentially anhydrous organic solvents is now commonplace (Zaks and Russell, 1988; Klibanov, 1989; Dordick, 1989). These enzymes catalyze the alcoholysis of a variety of esters (Zaks and Klibanov, 1985). In order to compare the activity and specificity of these enzymes in different solvents, it is necessary to show first that the mechanism for the alcoholysis reaction (often referred to as transesterification) does not change with the solvents used. It has been shown that these enzymes obey typical Michaelis-Menten kinetics (Zaks and Klibanov, 1985), that protein structure is similar in some aqueous and organic environments (Clark et al., 1989;Burke et al., 19891,that the Hammett constant for subtilisin (the most characterized enzyme which functions in nonaqueous media) is the same in aqueous and organic solvents (Kanerva and Klibanov, 19891, and that the kinetic isotope effect is similar in both aqueous and nonaqueous environments (Adams et al., 1991). We report data concerning the activity of thiolsubtilisin, an active site modified variant of the native enzyme,in organic solvents. Subtilisin Carlsberg, a serine protease secreted by Bacillus licheniformis, is an ideal model system for mechanistic nonaqueous enzymology (Alvaro and Russell, 1991; Wells and Estell, 1988). In aqueous solution, the enzyme catalyzes the hydrolysis of amino acid esters via an acylenzymemechanism (Fersht, 1985a). There has been much indirect evidence that an equivalent mechanism holds for the enzyme when it is suspended in organic solvents (Kanervaand Klibanov, 1989;Adams et al., 1991). However,an acylenzymehas never been isolated in organic solvents, and for most substrates saturation of the enzyme cannot be achieved because of extremely high Michaelis constants (Km)and substrate solubility limitations (Zaks and Klibanov, 1988). Thus, turnover numbers (kcat) and Km’sfor the ester substrate cannot be determined, and there is still a question of whether the enzyme mechanism is altered when enzymes are placed in nonaqueous environments [the generally accepted rules for the proof of a proposed enzyme mechanism have been presented previously (Fersht, 1985b)l. Of particular concern is that the specificity constant (kcat/Km) for subtilisin-catalyzed

* Address correspondence to this author a t the following address: 1235Benedum Hall,Department of Chemical Engineering,University of Pittsburgh, Pittsburgh, PA 15261. + Department of Chemical Engineering. t Center for Biotechnology and Bioengineering. 8756-7938/92/3008-0256$03.00/0

alcoholysis of N-protected amino acid esters in organic solvents is almost identical to that for the hydrolysis reaction of mutant subtilisins (in aqueous solution) which cannot utilize the acylenzyme mechanism. The k,aJKm for a subtilisin variant not containing a nucleophile at position 221 (the mutant enzyme functions via transitionstate stabilization alone) is 8.4 X s-l M-l. This is a reduction in the aqueous catalytic efficiency of 1.7 X lo6, which is strikingly similar to the difference in activity between subtilisin-catalyzed hydrolysis and transesterification of typical amino acid esters in aqueous and organic environments, respectively. It should be noted that the removal of the hydroxyl group, in this case by protein engineering, does not significantly alter Km for the substrate (N-succinyl-L-alanyl-L-alanyl-L-prolyl-Lphenylalanyl-p-nitroanilide). The difference between a thiol and a hydroxyl group at position 221 also has a negligible effect on substrate binding (Carter and Wells, 1988). We have synthesized a variant of subtilisin Carlsberg, thiolsubtilisin, in which the active site serine residue is replaced by a thiol-containing cysteine using an adaptation from a previously reported method (Polgar, 1976). Yields of >60% were achieved. The purity of the preparation was checked by measuring thiol content (Ellman, 1959). There was no contamination with native subtilisin. The resulting enzyme has now been analyzed in organic solvents. Previous studies have shown that the overall structures of the thiolsubtilisin and subtilisin are equivalent in aqueous solution (Tsai and Bender, 1979). Kinetic studies on thiolsubtilisin, in water, indicate that an acylenzyme is formed by attack of the thiol group upon the carbonyl of the substrate (Philipp et al., 1979). The broad specificity of subtilisin is impaired when the hydroxyl is replaced by the larger thiol group, but the overall mechanism is not changed (Tsai, 1977). The difference in the activity of subtilisin and thiolsubtilisin results from changes in rates of deacylation and acylation, not from interference with substrate binding. Before a comparison of the activity of subtilisin and thiolsubtilisin in organic solvents, it is important to demonstrate that in water the addition of an organic solvent will not affect thiolsubtilisin and subtilisin differently. That is, the ratios (kcat)subJ(kcat)thiol, ( K m h u b d Wmhhiol, and ( k c a J K m ) s u b J ( k c a J K m h i o 1 will only be altered upon addition of solvent if the kinetic and structural changes caused by the presence of an organic solvent are

0 1992 American Chemical Society and American InstWute of Chemical Engineers

257

Blotechnol. Prog., 1992, Vol. 8, No. 3

Table I. Effect of Solvent on Subtilisin- (subt) and Thiolsubtilisin- (thiol) Catalyzed Hydrolysis of (pNitropheny1) butyrates enzyme subtilisin thiolsubtilisin

5 10 30

kcat,min-l 137 f 1.65 4.66 f 0.23

29.4 f 3 30.4 3 30.1 3

* *

Km, mM 0.082 f 0.008 0.058 f 0.005

*

1.40 0.1 1.86 f 0.2 2.20 f 0.2

20.7 f 2 16.6 f 3 13.7 f 4

Assays were performed spectrophotometrically, following the initialrelease ofp-nitrophenol at412 nm, at 25 "C in 0.1 M phosphate buffer, pH 7.8. Acetonitrile was added to the buffers to give the desired concentration of acetonitrile and substrate (0.03 mM to 1 mM), and the reaction was initiated with the addition of enzyme (0.08 pM subtilisin or 0.4 pM thiolsubtilisin).

not equivalent in both enzymes. We tested the effect of solvent on the subtilisin-catalyzed hydrolysis of @-nitrophenyl)butyrate, a substrate which has been used previously in the analysis of thiolsubtilisin (Philipp et al., 1979). Since the ratios for (kcat)subJ(kcat)thiol reported in Table I do not change significantly with an increasing concentration of acetonitrile, thiolsubtilisin and subtilisin must respond to the addition of solvent in water identically, in terms of rates of acylation and deacylation. Also, as mentioned previously, the substrate binding pockets of subtilisin and thiolsubtilisin are known to behave distinctly, thus the difference in (Km)subt/(Km)thiol Values [which results in altered (kcaJKm)~~bJ(kcaJKm)thiolvalues1 at varying concentrations of acetonitrile is to be expected. It should be noted that assays of the activity of thiolsubtilisin are subject to greater error than those for subtilisin: rates of reaction with thiolsubtilisin are 50100 times slower than with subtilisin. If subtilisin suspended in organic solvents did not utilize the acylenzyme mechanism described above, then since neither enzyme would utilize Ser221, one would expect little difference in the activities of thiolsubtilisin and the native enzyme in the transesterification reaction. Indeed, the removal of noncatalytic side chains which do not participate in substrate binding (thiolsubtilisin actually binds these substrates better than subtilisin in water) is unlikely to effect the activity of enzymes (Alvaro and Russell, 1991). As shown in Table 11, thiolsubtilisin is far less efficient than subtilisin in all organic solvents tested, and with both substrates utilized. The water content of both enzymes was identical as measured by Karl-Fischer titration. It should be noted that the enzyme preparations were prepared from identical pH's and salt concentrations. The bimolecular rate constant, kcaJKm,has the same meaning for both hydrolysis and alcoholysis (via an acylenzyme intermediate), and therefore the values for (kcaJ Kmlthiol and (kcaJKm)subt in both water and organic solvents can be compared directly. Interestingly, for each organic solvent tested the ratio (kcat/Km)subJ (kcaJKm)thiol is not altered significantly, and it remains between 5 and 6. Given the results presented in Table I, the decreased catalytic efficiency of thiolsubtilisin relative to that of subtilisin in nonaqueous media is not the result of thiolsubtilisin being preferentially inactivated by the solvent. Rather, this is evidence that the active site serine hydroxyl group of subtilisin is taking part in the catalytic mechanism of alcoholysis. The data do not indicate directly that the acylenzymeis formed, although when these data are added to the mounting evidence described above, it seems inconceivable that the mechanism of the enzyme changes when transferred from water to nonaqueous media. It is

Table 11. Effect of Solvent on Subtilisin- and Thiolsubtilisin-Catalyzed Alcoholysis of (pNitropheny1)butyrate (pNPB) and N-Acetylphenylalanine Methyl Ester (Ac-Phe-OMe).

water hexane hexane (+0.1% H20) dioxane acetonitrile hexane

30 (pNPB) 5.8 (pNPB) 5.3 (pNPB) 6.7 (pNPB) 5.3 (pNPB) 68 (Ac-Phe-OMe)

4.62 2.39 2.26 2.58 2.26 5.73

Kinetic constants for the transesterification reaction between pNPB or Ac-Phe-OMe and ethanol were determined as reported previously (Zaksand Klibanov, 1988). Typically, both enzymes (0.05 mg/mL) were lyophilized from 0.01 M phosphate buffer, pH 7.8,0.2 mM ethylenediaminetetraacetic acid, and then suspended in dried organic solvents (0.2mg of subtilisin/mL, or 0.4mgof thiolsubtilisin/ mL) at 30 "C. The production of ethyl butyrate, p-nitrophenol, or N-acetylphenylalanine ethyl ester was followed with a gas chromatograph. AAG* calculations were performed aa described previously (Wilkinson et al., 1983). Values for individual specificity constants, Vm/Km,were obtained by measuring initial rates at 3-5 substrate concentrations. Since the concentration of enzyme cancels out when comparing subtilisin and thiolsubtilisin VJK, values are compared, The actual (k,J the ratios reported are (k,JKm)s"bJ(k,JKm)thiol. (k,JKm)he. values for subtilisin and thiolsubtilisin were 14.7 and 16.0, respectively.

also interesting that the ratio (kcaJKm),ubJ(kcaJKm)thiol is significantly different in water and the organic solvents tested. As mentioned previously, there is some difference in the binding sites of thiolsubtilisin and subtilisin. As such, it is not surprising that in the presence of organic solvents the value of this ratio changes. Indeed, the experiment described above for solubilized enzymes in the presence of acetonitrile already demonstrates this trend. It is known that the addition of small amounts of water (