Proc. Natl. Acad. Sci. USA
Vol. 76, No. 2, pp. 557-560, February 1979
Chemistry
Do cleavages of amides by serine proteases occur through a stepwise pathway involving tetrahedral intermediates? (proton transfer/basicity/non-stepwise mechanism)
MAKOTO KOMIYAMA AND MYRON L. BENDER Departments of Biochemistry and Chemistry, Northwestern University, Evanston, Illinois 60201
Contributed by Myron L. Bender, November 20, 1978
ABSTRACT
The mechanism of the serine protease-cata-
1yzed cleavage of amides (acylation) was examined in terms of the basicity of the functional groups participating in the catalysis. It is proposed that the reaction does not proceed through a stepwise pathway, as opposed to the cleavage of esters and anilides, which start with general base-catalyzed formation of the tetrahedral intermediate followed b its general acid-catalyzed breakdown. Instead, the proton abstracted from the hydroxyl group of the serine by the imidazolyl group of the histidine is donated to the nitrogen atom of the leaving group of the amide before the bond between the carbonyl carbon atom of the amide and the attacking serine oxygen atom is completed. Reactions proceed by a SN2-like reaction through the cooperation of acid catalysis by the imidazolyl cation and nucleophilic attack by the serine. The mechanisms of the enzymatic hy-
drolyses of anilides and esters proceed through a discrete tetrahedral intermediate, but the enzymatic hydrolyses of amides probably do not.
It has been widely accepted that reactions by serine proteases involve general base catalysis by the imidazolyl group of histidine (1, 2). In most accepted mechanisms, a tetrahedral intermediate between the hydroxyl group of the serine of serine proteases and amide (or ester) is formed in acylation, the breakdown of which to acyl-enzyme and amine (or alcohol) is catalyzed by acid catalysis (probably) by the imidazolyl cation of the histidine. In the deacylation the tetrahedral intermediate is formed from acyl-enzymes and water. The formation of tetrahedral intermediates is well documented in the serine protease-catalyzed hydrolyses of esters (3, 4) and anilides (5, 6). In the a-chymotrypsin-catalyzed hydrolyses of esters, an oxygen ester and the corresponding thiol ester showed identical rate constants of the acylation, indicating a rate-determining formation of an intermediate followed by its fast breakdown (3, 4). Furthermore, stopped-flow spectrophotometry of the serine protease-catalyzed hydrolyses of anilides detected the buildup of tetrahedral intermediates (5, 6). In the serine protease-catalyzed hydrolyses of the most important substrates, proteins (amides), however, evidence for a tetrahedral intermediate is scant. Analogy with enzymatic hydrolyses of esters and anilides and inference from the results of nonenzymatic hydrolyses of amides (7) assume important roles in the arguments favorable for this intermediate. The anomalous behavior of the pKa values governing kcat and Km (8-10) seems inconclusive, though they have been used as evidence for the existence of tetrahedral intermediates. Similar arguments were made in the enzymatic hydrolyses of anilides (11, 12). However, the anomaly probably comes from some interactions of leaving groups with the active site imidazole or leaving group specificity sites (13) or both. Furthermore, niThe 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. 557
trogen kinetic isotope effects do not definitely indicate the existence of a tetrahedral intermediate (14). In this report, proton transfer in the serine protease-catalyzed hydrolyses of amides is examined in terms of the basicity of the functional groups participating in the catalysis. It is suggested that the enzymatic cleavage of amides (acylation) may not proceed through a stepwise pathway starting with general base-catalyzed breakdown of 2 as shown in Scheme 1. Instead,
the proton abstracted from the hydroxyl group of the serine by the imidazolyl group of the histidine is donated to the nitrogen atom of the leaving group of the amide before the bond between the carbonyl carbon atom of the amide and the attacking serine oxygen atom is completed (see 3). Thus, the process is more like a SN2 reaction. This is opposed to nonenzymatic hydrolysis, during which both amides and esters proceed through tetrahedral intermediates on attack by hydroxide ion. General base
catalysis
NH
D.
c=O I
HI
1 R2
I
NH
General acid
catalysis
-
-
Acyl-enzyme
,cIoe
I
fil 2
Scheme 1. First, the fate of a proton abstracted from the hydroxyl group of serine by the imidazolyl group of histidine in serine protease-amide complexes (see 1) will be described. This proton abstraction undoubtedly increases the nucleophilicity of the oxygen atom of the serine, promoting the formation of the bond between this oxygen atom and the carbonyl carbon atom of the amide. This is usual general base catalysis and might result in the formation of the tetrahedral intermediate. However, one quite important factor is neglected in this argument. Partial bond-making between the serine oxygen atom and the amide carbonyl carbon atom of 3 reduces the double-bond character of the nitrogen atom-carbonyl carbon atom bond of the amide. Consequently, the basicity of the nitrogen atom of the amide markedly increases with bondmaking between the serine oxygen atom and the carbonyl
558
Chemistry: Komiyama and Bender
Proc. Natl. Acad. Sci. USA 76 (1979)
carbon atom of the amide because of changes due to both resonance and inductive effects. The pKa of the nitrogen atom of the amide in the initial state (pKa0) can be estimated by Eq. 1 (15): PKa0 = -18.6 + 1.04 pKaamine, [1] in which pKaamine represents the pKa of the amine itself (of the amide). Thus, for a typical amine of pKaamine = 10, pKao is estimated to be -8.2. On the other hand, the pKa of the nitrogen atom of the tetrahedral intermediate 2 (assuming it is formed), should have a pKa similar to that of the amine (16). Thus, there should be a change from -8.2 to +10 during the reaction. Therefore, the basicity of the nitrogen atom of the amide gets closer to that of the imidazolyl group of the histidine (pKa 7) as the reaction proceeds, and the basicity of the former exceeds the basicity of the latter somewhere along the reaction coordinate. The rate constant of proton transfer from donors to acceptors increases as the basicity of the acceptor increases (17). Therefore, proton transfer from the imidazolyl group of histidine to the nitrogen atom of the amide becomes faster as the reaction (the bond-making between the oxygen atom of serine and the carbonyl carbon atom of the amide) proceeds. For example, the rate constant of the proton transfer from imidazolyl cation to 55 M water (pKa of H30+ = -1.7) is 1.5 X 103 sec-1 (18). Thus, the rate constant of the proton transfer from the imidazolyl cation of histidine to the amide nitrogen atom can be 2.7 X 102 sec-1, when the pKa of the amide nitrogen atom is decreased from pKao to -1.7 as the reaction proceeds.* Here, the effective concentration of the amide, functioning as intramolecular proton acceptor, is taken as 10 M, which is the average value of the ratio between corresponding intermolecular and intramolecular catalysts and is a very conservative value (21). Furthermore, this proton transfer can be much faster since the histidine moiety has an orientation suitable for hydrogen bonding with the leaving group (22). Still, the estimated rate constant of the proton transfer (2.7 X 102 sec'1) is larger than the rate constants of the cleavage of amides catalyzed by a-chymotrypsin by 3-4 orders of magnitude.t Of course, the rate of the proton transfer becomes much larger when it occurs at a later stage of reaction, where the pKa of the amide nitrogen atom can become much larger than -1.7. Hydrogen bonding between serine and histidine might retard proton transfer from histidine to the amide nitrogen atom. Internal hydrogen bonding in a proton donor decreases the rate constant of proton transfer to a proton acceptor considerably (17). However, the magnitude of this effect on the present proton transfer should be small, since this proton transfer takes place after hydrogen bonding between the serine and the histidine is weakened by proton abstraction by the histidine. A pKa of -1.7 of the nitrogen atom of the amide (partly attacked by the serine oxygen atom) corresponds to a species that has about 64% contribution of 1 and 36% of 2. Thus, this pKa can probably be attained at a reaction coordinate before the transition state of the formation of the tetrahedral intermediate, 2. The above arguments show that the rate constant of proton transfer from the imidazolyl cation of histidine to the nitrogen atom of the amide can be much larger than that of enzymatic
cleavage of the amide, when the pKa of the amide (the nitrogen atom) is sufficiently high (for example, -1.7 or even lower).
I---C7061 3
-
There are arguments about whether His-57 or Asp-102 has a pKa of around 7, and experimental results contradict one another (refs. 19 and 20; W. W. Bachovchin and J. D. Roberts, unpublished results). However, the following arguments are the same as long as the pKa of the His-Asp pair is around 7. t The rate constant of the a-chymotrypsin-catalyzed cleavage of Nacetyl-L-phenylalanine amide (the acylation step) is 4.76 X 10-2 sec*
at 25°C
(23).
4
It is hypothesized here that the proton abstracted from serine by histidine can be transferred to the hydrogen bond between histidine and the nitrogen atom of the amide, 4. In 4, the proton is located largely on the N1 nitrogen atom of histidine. This conversion from 3 to 4 should be much faster than the proton transfer from histidine to the nitrogen atom of the amide and can take place at an earlier stage. Thus, this conversion should not be rate-determining in the enzymatic hydrolyses of amides. The "charge-relay system" can enhance this conversion through an increase of electron density on the NI nitrogen atom of histidine and through the coupling of two proton transfers. After 4 is formed, proton donation from histidine to the nitrogen atom of the amide and nucleophilic attack by serine at the carbonyl carbon atom can take place almost simultaneously. The present mechanism is different from the "pretransition state protonation mechanism" proposed by Wang and Parker (24), and criticized by others (9, 25, 26), in that proton transfer takes place only after the bond between the serine and the carbonyl carbon atom is partially formed. The reaction through 4 is more favorable for the following cleavage of the amide bond than that through 2, since proton donation assists nucleophilic attack by serine and vice versa. Thus, the reaction intermediate (ES2) observed by the stopped-flow method in the a-chymotrypsin-catalyzed hydrolysis of furylacryloyl-L-tryptophan amide by Hess et al. (27) might be 4 rather than 2. (See the schematic free energy diagram in Fig. 1.) In the reaction without proton donation from histidine to the nitrogen atom of the amide (which involves the formation of a tetrahedral intermediate), however, all of the activation energy [compensation of the resonance energy of the amide bond, which is about 20 kcal/mole (28), the energy for the conversion of the carbonyl carbon from sp2 to Sp3, and so on] must be provided only by the nucleophilic attack by serine. No acid catalysis can assist here. Furthermore, once the tetrahedral intermediate is formed, its nitrogen atom should be immediately protonated, since its PKa is higher than that of the solutions under biological conditions (29). The formation of such an unstable intermediate is unlikely. According to the results of x-ray crystallography, the proton acceptor of the leaving group takes a nearly ideal position for hydrogen bonding with the imidazolyl group of histidine in the
Chemistry: Komiyama and Bender
Proc. Natl. Acad. Sci. USA 76 (1979)
4n CDa)
a1) U.
1
Reaction coordinate
FIG. 1. Schematic energy diagram of the serine protease-catalyzed cleavage of amide (the acylation). Numbers refer to the species Mechanism proposed in this communication; (see text). mechanism involving tetrahedral intermediate (2). -,
subtilisin-boronic acid complex (a tetrahedral intermediate analog) (22, 30). This fact indicates that the nitrogen atom of the amide is located close to the Ni nitrogen atom of histidine in 3. Thus, conversion from 3 to 4 followed by proton donation from histidine to the nitrogen atom of the amide is reasonable from the steric point of view. However, formation of the tetrahedral intermediate (2) in the neighborhood (almost hydrogen-bonding distance) of the imidazolyl cation is not physically reasonable, since the pKa of the nitrogen atom of 2 (around 10) is larger than that of the imidazolyl cation of histidine (around 7). The proposed mechanism of enzymatic hydrolyses of amides, which proceeds through 4 instead of 2, is not inconsistent with the results of enzymatic hydrolyses of esters and anilides showing the existence of tetrahedral intermediates (3-6). In the hydrolyses of esters and anilides (in all cases para-nitro-substituted anilides were used), the pKa of the leaving groups are so small that proton donation from histidine to the leaving groups can hardly occur before the tetrahedral intermediates are completely formed. The present mechanism satisfactorily explains why acid catalysis by the imidazolyl group of histidine can be very effective in enzymatic reaction (ref. 31 and refs. therein). In this mechanism, the proton of the serine can be smoothly transferred to the "amine" of the substrate (amides) via the imidazolyl group of the histidine. If, on the other hand, the imidazolyl cation of histidine and
559
the tetrahedral intermediate 2 were formed and the tetrahedral intermediate had a lifetime long enough to be called an intermediate, the imidazolyl cation should easily lose its proton to the surrounding water during the lifetime of the tetrahedral intermediate. Thus, no effective acid catalysis by imidazole could be expected. This argument is supported by the results of general base-catalyzed hydrolyses of trifluoroacetanilides by imidazole (e.g., the proton abstracted from water by imidazole for general base catalysis was easily dissipated in the surrounding water during the lifetime of the tetrahedral intermediate rather than donated to the leaving aniline moiety); thus, a second water molecule must function as acid catalyst instead of imidazolyl cation (unpublished data). Although the present mechanism of the enzymatic hydrolyses of amides is heretical in that it suggests a different mechanism for amides and esters, it is consistent with very little charge development on the attacking nitrogen atom in the transition state in the aminolysis of acyl-enzymes, the reverse reaction (32, 33), considering microscopic reversibility. Furthermore, the present mechanism agrees well with the results of D20 solvent isotope experiments (23) and those of the nitrogen isotope experiments (14, 34), which showed rate-determining proton transfer and rate-determining C-N bond cleavage in the transition state.
Satterthwait and Jencks (29) proposed another possible mechanism of enzymatic amide hydrolyses which involves rapid movement of the imidazolyl group of histidine between a position suitable for hydrogen bonding with serine and another position suitable for hydrogen bonding with the amide nitrogen atom (mechanism 1 of the six mechanisms proposed by them). However, this mechanism is not likely, as discussed below. Table 1 shows a comparison of the rate constants of the general base-catalyzed hydrolyses of esters by imidazole and its derivatives with those of the alkaline hydrolyses in model reactions. The rate constants due to intramolecular general base catalyses by imidazole are comparable to those of alkaline hydrolyses at pH 8.0, the pH of maximum enzymatic rate. Here, the ratio of the rate constant catalyzed by an intramolecular general base catalyst to that catalyzed by an intermolecular general base catalyst was taken as 1.0 M, which is reasonable for a general base catalyst (35). The application of this result to the hydrolyses of acyl-enzymes shows that the rate of the hydrolysis of acyl-enzyme due to intramolecular general base catalysis by the imidazolyl group of histidine is comparable to that due to alkaline hydrolysis. No role of the imidazolyl cation of histidine is taken into consideration here, although catalysis by a combination of hydroxide ion and imidazolyl cation is kinetically indistinguishable from- general base catalysis by
Table 1. Comparison of rate constants of general base-catalyzed hydrolyses of esters by imidazole (or N-methylimidazole) (him) with those of alkaline hydrolyses (kOH) in model reactions
kim,
1
kIm,*
kOH,
10-6
kOHqt
(1
-
kim)/
M-1 sec1 sec-1 M-1 sec1 sec1 (10-6 kOH) Ethyl acetatel 1.3 X 10-7 1.3 X 10-7 1.1 X 10-1 1.1 X 10-7 1.2 Trifluoroethyl acetate§ 4.2 X 10-6 4.2 X 10-6 3.2 3.2 X 10-6 1.3 Ethyl chloroacetateT 6.0 X 10-5 6.0 X i0-5 15 1.5 X 10-5 4.0 Ethyl chloroacetatell 2.3 X i0-5 2.3 X 10-5 16 1.6 X 10-5 1.4 * Rate constant due to intramolecular general base catalysis by imidazole. Here, the ratio of intramolecular reaction rate to intermolecular reaction rate was taken as 1.0 M (rather than 10 M for nucleophile catalyses), since this value is smaller in general base catalyses than in nucleophile catalyses because of loose transition state (35). t Rate of alkaline hydrolyses at pH 8.0. $ Catalyzed by imidazole (36). § Catalyzed by N-methylimidazole (36). Catalyzed by 2-benzimidazolylacetate, which has both imidazolyl group and carboxylate ion in the same molecule (37). Catalyzed by N-methylimidazole (37). Ester
-
560
Chemistry: Komiyama and Bender
neutral imidazole. The imidazolyl cation can cause some additional acceleration by either proton donation or electrostatic effect. Thus, the effective concentration of the imidazolyl group of histidine as an intramolecular general base catalyst must be very much larger (for example, 103-106 M) than the 1 M observed in model reactions (35) in order that the rate constant of intramolecular general base catalysis by the imidazolyl group of histidine be much larger than that of alkaline hydrolysis. This can be attained only by strict freezing of the rotation of the imidazolyl ring. It is quite unlikely that the imidazolyl ring, which has considerable rigidity, exhibits fast movement, as proposed by Satterthwait and Jencks (29). In conclusion, it is proposed that the serine protease-catalyzed cleavage of amides (acylation) does not proceed via the stepwise mechanism of the general base-catalyzed formation of the tetrahedral intermediates followed by their general acid-catalyzed breakdown. Instead, the proton abstracted from serine by histidine is donated to the nitrogen atom of the amide. Reaction proceeds by the cooperation of acid catalysis by the imidazolyl cation and nucleophilic attack by serine, close to an SN2-like reaction. In the enzymatic hydrolyses of esters and anilides, proton transfer from histidine to the leaving groups takes place only after the bond between the oxygen atom of serine and the carbonyl carbon of the substrate is almost completed. Thus, the mechanisms of the enzymatic hydrolyses of amides, anilides, and esters are consistent with each other but can differ in timing. This work was supported by grants from the National Science Foundation (CHE76-14283), and Merck, Sharp and Dohme. 1. Bender, M. L. & Killheffer, J. V. (1973) Crit. Rev. Biochem. 1, 149-199. 2. Blow, D. M. (1976) Acc. Chem. Res. 9, 145-152. 3. Frankfater, A. & Kezdy, F. J. (1971) J. Am. Chem. Soc. 93, 4039-4043. 4. Hirohara, H., Bender, M. L. & Stark, R. S. (1974) Proc. Natl. Acad. Sci. USA 71, 1643-1647. 5. Hunkapiller, M. W., Forgac, M. D. & Richards, J. H. (1976)
Biochemistry 15,5581-5588.
6. Petkov, D. D. (1978) Biochim. Biophys. Acta 523,538-541. 7. Johnson, S. L. (1967) Adv. Phys. Org. Chem. 5,237-330. 8. Fersht, A. R. & Requena, Y. (1971) J. Am. Chem. Soc.' 93, 7079-7087. 9. Fersht, A. R. (1972) J. Am. Chem. Soc. 94,293-294.. 10. Lucas, E. C., Caplow, M. & Bush, K. J. (1973) J. Am. Chem. Soc.
95,2670-2673.
Proc. Natl. Acad. Sci. USA 76 (1979) 11. Inagami, T., Patchornik, A. & York, S. S. (1969) J. Biochem. 65, 809-819. 12. Caplow, M. (1969) J. Am. Chem. Soc. 91,3639-3645. 13. Philipp, M., Pollack, R. M. & Bender, M. L. (1973) Proc. Natl. Acad. Sci. USA 70,517-520. 14. O'Leary, M. H. & Kluetz, M. D. (1972) J. Am. Chem. Soc. 94, 3585-3589. 15 Fersht, A. R. (1971) J. Am. Chem. Soc. 93,3504-3514. 16. Satterthwait, A. C. & Jencks, W. P. (1974) J. Am. Chem. Soc. 96, 7031-7044. 17. Eigen, M. (1964) Angew. Chem. Int. Ed. Engl. 3, 1-19. 18. Eigen, M., Hammes, G. G. & Kustin, K. (1960) J. Am. Chem. Soc.
82,3482-3483. 19. Hunkapiller, M. W., Smallcombe, S. H., Whitaker, D. Ri. & Richards; J. H. (1973) Biochemistry 12, 4732-4743. 20. Koeppe II, R. E. & Stroud, R. M. (1976) Biochemistry 15, 3450-3458. 21. Bender, M. L. (1971) Homogeneous Catalysis from Protons to Proteins (Wiley, New York). 22. Matthews, D. M., Alden, R. A., Birktoft, J. J., Freer, S. T. & Kraut, J. (1975) J. Biol. Chem. 250, 7120-7126. 23. Bender, M. L., Clement, G. E., Kezdy, F. J. & Heck, H. D'A. (1964) J. Am. Chem. Soc. 86,3680-3690. 24. Wang, J. H. & Parker, L. (1967) Proc. Natl. Acad. Sci. USA 58, 2451-2454. 25. Williams, A. (1970) Biochemistry 9,3383-3390. 26. Lucas, E. C. & Caplow, M. (1972) J. Am. Chem. Soc. 94,960963. 27. Hess, G. P., McConn, J., Ku, E. & McConkey, G. (1970) Phil. Trans. R. Soc. London Ser. B: 257, 89-104. 28. Pauling, L. (1960) The Nature of the Chemical Bond (Cornell Univ. Press, Ithaca, NY), 3rd Ed. 29. Satterthwait, A. C. & Jencks, W. P. (1974) J. Am. Chem. Soc. 96, 7018-7031. 30. Matthews, D. A., Alden, R. A., Birktoft, J. J., Freer, S. T. & Kraut, J. (1977) J. Biol. Chem. 252, 8875-8883. 31. Komiyama, M. & Bender, M. L. (1978) J. Am. Chem. Soc. 100, 5977-5978. 32. Inward, P. W. & Jencks, W. P. (1965) J. Biol. Chem. 240, 1986-1996. 33. Zeeberg, B. & Caplow, M. (1973) J. Biol. Chem. 248, 58875891. 34. O'Leary, M. H. & Kluetz, M. D. (1970) J. Am. Chem. Soc. 92, 6089-6090. 35. Page, M. I. (1973) Chem. Soc. Rev. 2,295-323. 36. Kirsch, J. F. & Jencks, W. P. (1964) J. Am. Chem. Soc. 86, 837-846. 37. Komiyama, M. & Bender, M. L. (1976) Bioorg. Chem. 6,1320.
Vol. 76, No. 2, pp. 557-560, February 1979
Chemistry
Do cleavages of amides by serine proteases occur through a stepwise pathway involving tetrahedral intermediates? (proton transfer/basicity/non-stepwise mechanism)
MAKOTO KOMIYAMA AND MYRON L. BENDER Departments of Biochemistry and Chemistry, Northwestern University, Evanston, Illinois 60201
Contributed by Myron L. Bender, November 20, 1978
ABSTRACT
The mechanism of the serine protease-cata-
1yzed cleavage of amides (acylation) was examined in terms of the basicity of the functional groups participating in the catalysis. It is proposed that the reaction does not proceed through a stepwise pathway, as opposed to the cleavage of esters and anilides, which start with general base-catalyzed formation of the tetrahedral intermediate followed b its general acid-catalyzed breakdown. Instead, the proton abstracted from the hydroxyl group of the serine by the imidazolyl group of the histidine is donated to the nitrogen atom of the leaving group of the amide before the bond between the carbonyl carbon atom of the amide and the attacking serine oxygen atom is completed. Reactions proceed by a SN2-like reaction through the cooperation of acid catalysis by the imidazolyl cation and nucleophilic attack by the serine. The mechanisms of the enzymatic hy-
drolyses of anilides and esters proceed through a discrete tetrahedral intermediate, but the enzymatic hydrolyses of amides probably do not.
It has been widely accepted that reactions by serine proteases involve general base catalysis by the imidazolyl group of histidine (1, 2). In most accepted mechanisms, a tetrahedral intermediate between the hydroxyl group of the serine of serine proteases and amide (or ester) is formed in acylation, the breakdown of which to acyl-enzyme and amine (or alcohol) is catalyzed by acid catalysis (probably) by the imidazolyl cation of the histidine. In the deacylation the tetrahedral intermediate is formed from acyl-enzymes and water. The formation of tetrahedral intermediates is well documented in the serine protease-catalyzed hydrolyses of esters (3, 4) and anilides (5, 6). In the a-chymotrypsin-catalyzed hydrolyses of esters, an oxygen ester and the corresponding thiol ester showed identical rate constants of the acylation, indicating a rate-determining formation of an intermediate followed by its fast breakdown (3, 4). Furthermore, stopped-flow spectrophotometry of the serine protease-catalyzed hydrolyses of anilides detected the buildup of tetrahedral intermediates (5, 6). In the serine protease-catalyzed hydrolyses of the most important substrates, proteins (amides), however, evidence for a tetrahedral intermediate is scant. Analogy with enzymatic hydrolyses of esters and anilides and inference from the results of nonenzymatic hydrolyses of amides (7) assume important roles in the arguments favorable for this intermediate. The anomalous behavior of the pKa values governing kcat and Km (8-10) seems inconclusive, though they have been used as evidence for the existence of tetrahedral intermediates. Similar arguments were made in the enzymatic hydrolyses of anilides (11, 12). However, the anomaly probably comes from some interactions of leaving groups with the active site imidazole or leaving group specificity sites (13) or both. Furthermore, niThe 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. 557
trogen kinetic isotope effects do not definitely indicate the existence of a tetrahedral intermediate (14). In this report, proton transfer in the serine protease-catalyzed hydrolyses of amides is examined in terms of the basicity of the functional groups participating in the catalysis. It is suggested that the enzymatic cleavage of amides (acylation) may not proceed through a stepwise pathway starting with general base-catalyzed breakdown of 2 as shown in Scheme 1. Instead,
the proton abstracted from the hydroxyl group of the serine by the imidazolyl group of the histidine is donated to the nitrogen atom of the leaving group of the amide before the bond between the carbonyl carbon atom of the amide and the attacking serine oxygen atom is completed (see 3). Thus, the process is more like a SN2 reaction. This is opposed to nonenzymatic hydrolysis, during which both amides and esters proceed through tetrahedral intermediates on attack by hydroxide ion. General base
catalysis
NH
D.
c=O I
HI
1 R2
I
NH
General acid
catalysis
-
-
Acyl-enzyme
,cIoe
I
fil 2
Scheme 1. First, the fate of a proton abstracted from the hydroxyl group of serine by the imidazolyl group of histidine in serine protease-amide complexes (see 1) will be described. This proton abstraction undoubtedly increases the nucleophilicity of the oxygen atom of the serine, promoting the formation of the bond between this oxygen atom and the carbonyl carbon atom of the amide. This is usual general base catalysis and might result in the formation of the tetrahedral intermediate. However, one quite important factor is neglected in this argument. Partial bond-making between the serine oxygen atom and the amide carbonyl carbon atom of 3 reduces the double-bond character of the nitrogen atom-carbonyl carbon atom bond of the amide. Consequently, the basicity of the nitrogen atom of the amide markedly increases with bondmaking between the serine oxygen atom and the carbonyl
558
Chemistry: Komiyama and Bender
Proc. Natl. Acad. Sci. USA 76 (1979)
carbon atom of the amide because of changes due to both resonance and inductive effects. The pKa of the nitrogen atom of the amide in the initial state (pKa0) can be estimated by Eq. 1 (15): PKa0 = -18.6 + 1.04 pKaamine, [1] in which pKaamine represents the pKa of the amine itself (of the amide). Thus, for a typical amine of pKaamine = 10, pKao is estimated to be -8.2. On the other hand, the pKa of the nitrogen atom of the tetrahedral intermediate 2 (assuming it is formed), should have a pKa similar to that of the amine (16). Thus, there should be a change from -8.2 to +10 during the reaction. Therefore, the basicity of the nitrogen atom of the amide gets closer to that of the imidazolyl group of the histidine (pKa 7) as the reaction proceeds, and the basicity of the former exceeds the basicity of the latter somewhere along the reaction coordinate. The rate constant of proton transfer from donors to acceptors increases as the basicity of the acceptor increases (17). Therefore, proton transfer from the imidazolyl group of histidine to the nitrogen atom of the amide becomes faster as the reaction (the bond-making between the oxygen atom of serine and the carbonyl carbon atom of the amide) proceeds. For example, the rate constant of the proton transfer from imidazolyl cation to 55 M water (pKa of H30+ = -1.7) is 1.5 X 103 sec-1 (18). Thus, the rate constant of the proton transfer from the imidazolyl cation of histidine to the amide nitrogen atom can be 2.7 X 102 sec-1, when the pKa of the amide nitrogen atom is decreased from pKao to -1.7 as the reaction proceeds.* Here, the effective concentration of the amide, functioning as intramolecular proton acceptor, is taken as 10 M, which is the average value of the ratio between corresponding intermolecular and intramolecular catalysts and is a very conservative value (21). Furthermore, this proton transfer can be much faster since the histidine moiety has an orientation suitable for hydrogen bonding with the leaving group (22). Still, the estimated rate constant of the proton transfer (2.7 X 102 sec'1) is larger than the rate constants of the cleavage of amides catalyzed by a-chymotrypsin by 3-4 orders of magnitude.t Of course, the rate of the proton transfer becomes much larger when it occurs at a later stage of reaction, where the pKa of the amide nitrogen atom can become much larger than -1.7. Hydrogen bonding between serine and histidine might retard proton transfer from histidine to the amide nitrogen atom. Internal hydrogen bonding in a proton donor decreases the rate constant of proton transfer to a proton acceptor considerably (17). However, the magnitude of this effect on the present proton transfer should be small, since this proton transfer takes place after hydrogen bonding between the serine and the histidine is weakened by proton abstraction by the histidine. A pKa of -1.7 of the nitrogen atom of the amide (partly attacked by the serine oxygen atom) corresponds to a species that has about 64% contribution of 1 and 36% of 2. Thus, this pKa can probably be attained at a reaction coordinate before the transition state of the formation of the tetrahedral intermediate, 2. The above arguments show that the rate constant of proton transfer from the imidazolyl cation of histidine to the nitrogen atom of the amide can be much larger than that of enzymatic
cleavage of the amide, when the pKa of the amide (the nitrogen atom) is sufficiently high (for example, -1.7 or even lower).
I---C7061 3
-
There are arguments about whether His-57 or Asp-102 has a pKa of around 7, and experimental results contradict one another (refs. 19 and 20; W. W. Bachovchin and J. D. Roberts, unpublished results). However, the following arguments are the same as long as the pKa of the His-Asp pair is around 7. t The rate constant of the a-chymotrypsin-catalyzed cleavage of Nacetyl-L-phenylalanine amide (the acylation step) is 4.76 X 10-2 sec*
at 25°C
(23).
4
It is hypothesized here that the proton abstracted from serine by histidine can be transferred to the hydrogen bond between histidine and the nitrogen atom of the amide, 4. In 4, the proton is located largely on the N1 nitrogen atom of histidine. This conversion from 3 to 4 should be much faster than the proton transfer from histidine to the nitrogen atom of the amide and can take place at an earlier stage. Thus, this conversion should not be rate-determining in the enzymatic hydrolyses of amides. The "charge-relay system" can enhance this conversion through an increase of electron density on the NI nitrogen atom of histidine and through the coupling of two proton transfers. After 4 is formed, proton donation from histidine to the nitrogen atom of the amide and nucleophilic attack by serine at the carbonyl carbon atom can take place almost simultaneously. The present mechanism is different from the "pretransition state protonation mechanism" proposed by Wang and Parker (24), and criticized by others (9, 25, 26), in that proton transfer takes place only after the bond between the serine and the carbonyl carbon atom is partially formed. The reaction through 4 is more favorable for the following cleavage of the amide bond than that through 2, since proton donation assists nucleophilic attack by serine and vice versa. Thus, the reaction intermediate (ES2) observed by the stopped-flow method in the a-chymotrypsin-catalyzed hydrolysis of furylacryloyl-L-tryptophan amide by Hess et al. (27) might be 4 rather than 2. (See the schematic free energy diagram in Fig. 1.) In the reaction without proton donation from histidine to the nitrogen atom of the amide (which involves the formation of a tetrahedral intermediate), however, all of the activation energy [compensation of the resonance energy of the amide bond, which is about 20 kcal/mole (28), the energy for the conversion of the carbonyl carbon from sp2 to Sp3, and so on] must be provided only by the nucleophilic attack by serine. No acid catalysis can assist here. Furthermore, once the tetrahedral intermediate is formed, its nitrogen atom should be immediately protonated, since its PKa is higher than that of the solutions under biological conditions (29). The formation of such an unstable intermediate is unlikely. According to the results of x-ray crystallography, the proton acceptor of the leaving group takes a nearly ideal position for hydrogen bonding with the imidazolyl group of histidine in the
Chemistry: Komiyama and Bender
Proc. Natl. Acad. Sci. USA 76 (1979)
4n CDa)
a1) U.
1
Reaction coordinate
FIG. 1. Schematic energy diagram of the serine protease-catalyzed cleavage of amide (the acylation). Numbers refer to the species Mechanism proposed in this communication; (see text). mechanism involving tetrahedral intermediate (2). -,
subtilisin-boronic acid complex (a tetrahedral intermediate analog) (22, 30). This fact indicates that the nitrogen atom of the amide is located close to the Ni nitrogen atom of histidine in 3. Thus, conversion from 3 to 4 followed by proton donation from histidine to the nitrogen atom of the amide is reasonable from the steric point of view. However, formation of the tetrahedral intermediate (2) in the neighborhood (almost hydrogen-bonding distance) of the imidazolyl cation is not physically reasonable, since the pKa of the nitrogen atom of 2 (around 10) is larger than that of the imidazolyl cation of histidine (around 7). The proposed mechanism of enzymatic hydrolyses of amides, which proceeds through 4 instead of 2, is not inconsistent with the results of enzymatic hydrolyses of esters and anilides showing the existence of tetrahedral intermediates (3-6). In the hydrolyses of esters and anilides (in all cases para-nitro-substituted anilides were used), the pKa of the leaving groups are so small that proton donation from histidine to the leaving groups can hardly occur before the tetrahedral intermediates are completely formed. The present mechanism satisfactorily explains why acid catalysis by the imidazolyl group of histidine can be very effective in enzymatic reaction (ref. 31 and refs. therein). In this mechanism, the proton of the serine can be smoothly transferred to the "amine" of the substrate (amides) via the imidazolyl group of the histidine. If, on the other hand, the imidazolyl cation of histidine and
559
the tetrahedral intermediate 2 were formed and the tetrahedral intermediate had a lifetime long enough to be called an intermediate, the imidazolyl cation should easily lose its proton to the surrounding water during the lifetime of the tetrahedral intermediate. Thus, no effective acid catalysis by imidazole could be expected. This argument is supported by the results of general base-catalyzed hydrolyses of trifluoroacetanilides by imidazole (e.g., the proton abstracted from water by imidazole for general base catalysis was easily dissipated in the surrounding water during the lifetime of the tetrahedral intermediate rather than donated to the leaving aniline moiety); thus, a second water molecule must function as acid catalyst instead of imidazolyl cation (unpublished data). Although the present mechanism of the enzymatic hydrolyses of amides is heretical in that it suggests a different mechanism for amides and esters, it is consistent with very little charge development on the attacking nitrogen atom in the transition state in the aminolysis of acyl-enzymes, the reverse reaction (32, 33), considering microscopic reversibility. Furthermore, the present mechanism agrees well with the results of D20 solvent isotope experiments (23) and those of the nitrogen isotope experiments (14, 34), which showed rate-determining proton transfer and rate-determining C-N bond cleavage in the transition state.
Satterthwait and Jencks (29) proposed another possible mechanism of enzymatic amide hydrolyses which involves rapid movement of the imidazolyl group of histidine between a position suitable for hydrogen bonding with serine and another position suitable for hydrogen bonding with the amide nitrogen atom (mechanism 1 of the six mechanisms proposed by them). However, this mechanism is not likely, as discussed below. Table 1 shows a comparison of the rate constants of the general base-catalyzed hydrolyses of esters by imidazole and its derivatives with those of the alkaline hydrolyses in model reactions. The rate constants due to intramolecular general base catalyses by imidazole are comparable to those of alkaline hydrolyses at pH 8.0, the pH of maximum enzymatic rate. Here, the ratio of the rate constant catalyzed by an intramolecular general base catalyst to that catalyzed by an intermolecular general base catalyst was taken as 1.0 M, which is reasonable for a general base catalyst (35). The application of this result to the hydrolyses of acyl-enzymes shows that the rate of the hydrolysis of acyl-enzyme due to intramolecular general base catalysis by the imidazolyl group of histidine is comparable to that due to alkaline hydrolysis. No role of the imidazolyl cation of histidine is taken into consideration here, although catalysis by a combination of hydroxide ion and imidazolyl cation is kinetically indistinguishable from- general base catalysis by
Table 1. Comparison of rate constants of general base-catalyzed hydrolyses of esters by imidazole (or N-methylimidazole) (him) with those of alkaline hydrolyses (kOH) in model reactions
kim,
1
kIm,*
kOH,
10-6
kOHqt
(1
-
kim)/
M-1 sec1 sec-1 M-1 sec1 sec1 (10-6 kOH) Ethyl acetatel 1.3 X 10-7 1.3 X 10-7 1.1 X 10-1 1.1 X 10-7 1.2 Trifluoroethyl acetate§ 4.2 X 10-6 4.2 X 10-6 3.2 3.2 X 10-6 1.3 Ethyl chloroacetateT 6.0 X 10-5 6.0 X i0-5 15 1.5 X 10-5 4.0 Ethyl chloroacetatell 2.3 X i0-5 2.3 X 10-5 16 1.6 X 10-5 1.4 * Rate constant due to intramolecular general base catalysis by imidazole. Here, the ratio of intramolecular reaction rate to intermolecular reaction rate was taken as 1.0 M (rather than 10 M for nucleophile catalyses), since this value is smaller in general base catalyses than in nucleophile catalyses because of loose transition state (35). t Rate of alkaline hydrolyses at pH 8.0. $ Catalyzed by imidazole (36). § Catalyzed by N-methylimidazole (36). Catalyzed by 2-benzimidazolylacetate, which has both imidazolyl group and carboxylate ion in the same molecule (37). Catalyzed by N-methylimidazole (37). Ester
-
560
Chemistry: Komiyama and Bender
neutral imidazole. The imidazolyl cation can cause some additional acceleration by either proton donation or electrostatic effect. Thus, the effective concentration of the imidazolyl group of histidine as an intramolecular general base catalyst must be very much larger (for example, 103-106 M) than the 1 M observed in model reactions (35) in order that the rate constant of intramolecular general base catalysis by the imidazolyl group of histidine be much larger than that of alkaline hydrolysis. This can be attained only by strict freezing of the rotation of the imidazolyl ring. It is quite unlikely that the imidazolyl ring, which has considerable rigidity, exhibits fast movement, as proposed by Satterthwait and Jencks (29). In conclusion, it is proposed that the serine protease-catalyzed cleavage of amides (acylation) does not proceed via the stepwise mechanism of the general base-catalyzed formation of the tetrahedral intermediates followed by their general acid-catalyzed breakdown. Instead, the proton abstracted from serine by histidine is donated to the nitrogen atom of the amide. Reaction proceeds by the cooperation of acid catalysis by the imidazolyl cation and nucleophilic attack by serine, close to an SN2-like reaction. In the enzymatic hydrolyses of esters and anilides, proton transfer from histidine to the leaving groups takes place only after the bond between the oxygen atom of serine and the carbonyl carbon of the substrate is almost completed. Thus, the mechanisms of the enzymatic hydrolyses of amides, anilides, and esters are consistent with each other but can differ in timing. This work was supported by grants from the National Science Foundation (CHE76-14283), and Merck, Sharp and Dohme. 1. Bender, M. L. & Killheffer, J. V. (1973) Crit. Rev. Biochem. 1, 149-199. 2. Blow, D. M. (1976) Acc. Chem. Res. 9, 145-152. 3. Frankfater, A. & Kezdy, F. J. (1971) J. Am. Chem. Soc. 93, 4039-4043. 4. Hirohara, H., Bender, M. L. & Stark, R. S. (1974) Proc. Natl. Acad. Sci. USA 71, 1643-1647. 5. Hunkapiller, M. W., Forgac, M. D. & Richards, J. H. (1976)
Biochemistry 15,5581-5588.
6. Petkov, D. D. (1978) Biochim. Biophys. Acta 523,538-541. 7. Johnson, S. L. (1967) Adv. Phys. Org. Chem. 5,237-330. 8. Fersht, A. R. & Requena, Y. (1971) J. Am. Chem. Soc.' 93, 7079-7087. 9. Fersht, A. R. (1972) J. Am. Chem. Soc. 94,293-294.. 10. Lucas, E. C., Caplow, M. & Bush, K. J. (1973) J. Am. Chem. Soc.
95,2670-2673.
Proc. Natl. Acad. Sci. USA 76 (1979) 11. Inagami, T., Patchornik, A. & York, S. S. (1969) J. Biochem. 65, 809-819. 12. Caplow, M. (1969) J. Am. Chem. Soc. 91,3639-3645. 13. Philipp, M., Pollack, R. M. & Bender, M. L. (1973) Proc. Natl. Acad. Sci. USA 70,517-520. 14. O'Leary, M. H. & Kluetz, M. D. (1972) J. Am. Chem. Soc. 94, 3585-3589. 15 Fersht, A. R. (1971) J. Am. Chem. Soc. 93,3504-3514. 16. Satterthwait, A. C. & Jencks, W. P. (1974) J. Am. Chem. Soc. 96, 7031-7044. 17. Eigen, M. (1964) Angew. Chem. Int. Ed. Engl. 3, 1-19. 18. Eigen, M., Hammes, G. G. & Kustin, K. (1960) J. Am. Chem. Soc.
82,3482-3483. 19. Hunkapiller, M. W., Smallcombe, S. H., Whitaker, D. Ri. & Richards; J. H. (1973) Biochemistry 12, 4732-4743. 20. Koeppe II, R. E. & Stroud, R. M. (1976) Biochemistry 15, 3450-3458. 21. Bender, M. L. (1971) Homogeneous Catalysis from Protons to Proteins (Wiley, New York). 22. Matthews, D. M., Alden, R. A., Birktoft, J. J., Freer, S. T. & Kraut, J. (1975) J. Biol. Chem. 250, 7120-7126. 23. Bender, M. L., Clement, G. E., Kezdy, F. J. & Heck, H. D'A. (1964) J. Am. Chem. Soc. 86,3680-3690. 24. Wang, J. H. & Parker, L. (1967) Proc. Natl. Acad. Sci. USA 58, 2451-2454. 25. Williams, A. (1970) Biochemistry 9,3383-3390. 26. Lucas, E. C. & Caplow, M. (1972) J. Am. Chem. Soc. 94,960963. 27. Hess, G. P., McConn, J., Ku, E. & McConkey, G. (1970) Phil. Trans. R. Soc. London Ser. B: 257, 89-104. 28. Pauling, L. (1960) The Nature of the Chemical Bond (Cornell Univ. Press, Ithaca, NY), 3rd Ed. 29. Satterthwait, A. C. & Jencks, W. P. (1974) J. Am. Chem. Soc. 96, 7018-7031. 30. Matthews, D. A., Alden, R. A., Birktoft, J. J., Freer, S. T. & Kraut, J. (1977) J. Biol. Chem. 252, 8875-8883. 31. Komiyama, M. & Bender, M. L. (1978) J. Am. Chem. Soc. 100, 5977-5978. 32. Inward, P. W. & Jencks, W. P. (1965) J. Biol. Chem. 240, 1986-1996. 33. Zeeberg, B. & Caplow, M. (1973) J. Biol. Chem. 248, 58875891. 34. O'Leary, M. H. & Kluetz, M. D. (1970) J. Am. Chem. Soc. 92, 6089-6090. 35. Page, M. I. (1973) Chem. Soc. Rev. 2,295-323. 36. Kirsch, J. F. & Jencks, W. P. (1964) J. Am. Chem. Soc. 86, 837-846. 37. Komiyama, M. & Bender, M. L. (1976) Bioorg. Chem. 6,1320.
