1GX

hIYRON

L.

BlTNDE!R

.INI)

BYRON it7. TURNQUEST

\.<)I. 7!J

[CONTRIBUTION FROM

THE

DEPARTMENT OF CIIEMISTRY, ILLINOIS INSTITUTE 01

General Basic Catalysis of Ester Hydrolysis and Its Relationship to Enzymatic Hydrolysis
BY MYRON L. BENDER A N D BYRON W. TURNQCEST~
RECEIVED AL-CUST 30, 1956 The hydrolyses of p-nitrophenyl acetate, 2,4-dinitrophenyl acetate, phenyl acetate and ethyl thiolacetate are catalyzed by a number of bases including pyridine, 3- and 4-picoline, trimethylamine, imidazole, N-methylimidazole, quinoline and acetate ion. The hydrolysis of p-nitrophenyl acetate meets the requirements of general basic catalysis since the rate is proportional t o the summation of catalytic constants times catalyst concentrations a t constant p H and ionic strength. The catalytic rate constants are related t o the basicity of the catalyzing amines of constant steric requirement in a manner similar to the Brfinsted catalysis law. The slope of this linear free energy relationship is 1.62, the highest recorded for a general basic catalysis. The mechanism of this general basic catalysis, involving not a rate-determining proton transfer but rather the addition of the base t o the substrate producing an unstable intermediate, offers a n explanation for the deviation of the slope from its usual limitation. Imidazole catalysis of the hydrolysis of various esters demonstrates that the rate of hydrolysis increases with the acidity of the alcoholic residue of the ester. It appears that general basic catalysis of ester hydrolysis is of importance only for esters containing an alcohol that is a reasonably strong acid ( p K , <11). Comparison of the imidazole- and 0-chymotrypsin-catalyzed hydrolyses of P-nitrophenyl acetate shows that the two types of hydrolysis differ by a factor of ca. lo5in the rate of formation of p-nitrophenol but are similar in the rate of formation of acetate from the acylimidazole and acyl-enzyme intermediates. The action of imidazole as a general basic catalyst offers only a partial ex. planation of t.he mecha.nism of enzymatic hydrolysis.

Introduction
The hydrogen and hydroxyl ion-catalyzed hydrolyses of esters have received a great deal of attention and the mechanisms of these reactions have been worked out in great detaiL3 Investigators have also considered the possibility of catalysis by acids and bases other than these two aquo species. General acid or basic catalysis of ester hydrolysis is of particular interest from the viewpoint of enzymatic hydrolysis; i t is likely that, for catalytic purposes, an enzyme uses general acids or bases on its surface rather than hydronium and/or hydroxyl ions in solution since the latter ions are present only in small concentration a t the 9s a t which enzymatic hydrolyses occur. It has been shown recently in this Laboratory that enzymatic hydrolysis does not resemble acidic or basic hydrolysis with respect to carbonyl oxygen exchange during ester hydrolysis4 or with respect to the effect of substituents on the rate of hydrolysis.6 On the other hand, it has been suggested that a Br@nstedbase, namely, imidazole, is a part of the catalytically active site of ol-chymotrypsin.6 In confirniation of this hypothesis, i t was shown that thiol esters such as acetylglutathione are cleaved by imidazole7 and that 9-nitrophenyl acetate is hydrolyzed by imidazole6 and by Sbenzoyl-L-histidine methyl ester.s Early investigations of general basic catalysis of ester hydrolysis were inconclusive. Dawson and co-workers studied the catalytic effects of acetic
(1) This investiaation was supported by reseach grant H-2416 of the National Institutes of Health. Paper V I of the series The Mechanism of Enzymatic Hydrolysis. (2) Eastman Kodak Co. research fellow. 1955-1956. (3) C. K. Ingold, Structure and Mechanism in Organic Chemist r y , Cornel1 Univ. Press, Ithaca, N . Y., 1953.pp. 752-782; J. Hine, Physical Organic Chemistry, McCraw-Hill Book Co., Inc., New York, N. Y., 1956,pp. 260-282. (4) M . L. Bender and K.C. Kemp, THIS J O U R N A L , 79, 116 (1957). (5) hl. . I Bender and B . W . Turnquest, ibid., 77,4271 (1955). (6) M. L. Bender and B. W , Turnquest, ibid., 79, 1652 (1957),and references therein. (7) E. R. Stadtman, The Mechanism of Enzyme Action. W. D. McElroy and B. Glass, eds., The Johns Hopkins Press, Baltimore, M d . , 1954,pp. 581-598. (8) B. 3. Hartley, A n u R~p1.r. on Progr Chpm. (Iheiiz . ? o f , T.ott * I o n ) , 51. 311 ( I R i l >

acid, acetate ion, chloroacetic acid and bisulfate However, ion on the hydrolysis of ethyl a ~ e t a t e . ~ the reported catalytic rate constants are minute (the catalytic constant for acetate ion was reported to be 0.25 X lo- l./mole sec. while that for hydroxide ion is 1.08 l./mole sec.) and their reality is questionable. These data have been critically discussed from the viewpoint of the primary salt effect and the effect of the nature of the buffer on the rate.l0V1l General catalysis by acetic acid and acetate ion, although of small magnitude, has been reported in the esterification of acetic acid in methanol.12 A consideration of the mechanism of the hydrolysis of carboxylic acid derivatives by hydrogen and hydroxyl ions points t o the direction in which general basic catalysis of ester hydrolysis may be sought, l 3 Wiberg14states that for the reaction

Y+

the initial addition depends on the nucleophilicity of Y but the partitioning of the addition compound to form reactant or product is related to the relative stabilities of Y and X. One would then expect general basic catalysis of ester hydrolysis when the anion formed from the ester is a t least of comparable stability (basicity) to the attacking base. Accordingly, in order to demonstrate general basic catalysis, esters were chosen in which the alcoholic portion of the ester was a much stronger acid than the ethanol in Dawsons original work. Among such esters are phenyl, negatively substituted phenyl and thiol esters.
(9) H. M. Dawson and W.Lowson, J. Chem. Soc., 2444 (1927); H. & I . Dawson and W. Lowson, ibid., 393 (1929); H . iM. Dawson, E. R . Pycock and E. Spivey, ibid., 291 (1933). (10) R. P. Bell, Acid-Base Catalysis, Oxford Univ. rem, London, 1941,p . 80. (11) G. Gorin, personal communication. (12) A. C . Rolfe and C . N. Hinshelwood, T r a n s . Faraday Soc., 3 0 , 935 (1934). . I .I,, Bender (13) M. I . Bender, THIS JOURNAL, 73, 1G26 (1951); 1 and R . D. Ginaer, i b i d . , 77, 348 (1955). ( 1 4 ) I<. B Wiberp, i b i d . , 7 7 , 2519 (10.55)

April 5, 1957 Experimental

BASIC CATALYSIS OF

ESTER HYDROLYSIS

1657

Materials .-2,4-Dinitrophenyl acetate was prepared from 2,4-dinitrophenol and acetic anhydride.l6 Two recrystallizations from Skellysolve B gave an almost colorless product, m.p. 71-71.5" (lit. m.p. 72"). The preparation of p-nitrophenyl acetate was described previously.6 Ethyl thiolacetate was prepared from ethanethiol and a ceyl chloride, b.p. 115-115.5" (lit. b.p. 115').'6 Phenyl acetate was an ~ Eastman Kodak Co. white label product, n Z 0 1.5025. Imidazole was described previously.6 In order to avoid turbidity (probably due to hydrocarbon contamination) of the solution when 4-picoline was mixed with water, the Eastman Kodak Co. white label product was distilled (b.p. 142-143'), converted to the hydrochloride and crystallized from ethanol-ether, reconverted to the free amine ~ with alkali and redistilled, b.p. 143-144', n Z o 1.5035 ~ 3-Picoline, a Reilly Coal Tar (lit. b.p. 145", n z o1.5029).'7 Co. product, was fractionated through a Todd distillation assembly packed with a monel spiral; the middle fraction was collected using a reflux ratio of seven-to-one, b.p. 144.l0, @ 2 2 . 5 ~ 1.5049 (lit. b.p. 143.8", n Z 6 D 1.5038).17 Fisher reagent grade pyridine was redistilled, b.p. 115:. Fisher reagent grade quinoline was redistilled, b.p. 236 Eastman Kodak Co. white label trimethylamine (as the hydrochloride) was used without further purification. NMethylimidazole was prepared from imidazole and methyl iodide, b.p. 196-198' (lit. b.p. 19S0).'8 The neutral equivalent was determined by titration of 40-mg. samples with 0.02 N H C l using a Beckman model H pH meter. Titration curves were drawn and the end-point taken as the inflection point of these curves; neut. equiv. found 80.7 and 81.0, calcd. 82.1. The ionization constant of N-methylimidazole was determined potentiometrically in the above manner. The pH of the half-neutralization point of the amine, used was found to be 7.00 and 6.98 for soluas the value of PK,, tions 0.02 A4 in total amine. Dedichenlg has reported a p K , of 7.34 for this amine a t ionic strengths of the same order of magnitude as above. J . T. Baker reagent grade acetic acid was redistilled, b.p. 118'. Commercial 1,4dioxane was purified as described previously.6 Kinetic Determinations.-The general procedure for the spectrophotometric determination of the kinetics using a Cary recording spectrophotometer has been given previously.6 The hydrolysis of p-nitrophenyl acetate was followed by measurement of the appearance of p-nitrophenolate ion a t 400 mp; the hydrolysis of 2,4-dinitrophenyl acetate was followed by measurement of the appearance of 2,4-dinitrophenolate ion a t 360 mp; the hydrolysis of phenyl acetate was followed by measurement of the appearance of phenol a t 272 mp and the hydrolysis of ethyl thiolacetate was followed by measurement of the disappearance of ester a t 245 m r . All species were followed a t or very near their absorption maxima except in the case of ethyl thiolacef imidazole interfered tate where the intense absorption o with measurements a t the absorption maximum of the ester. The specific rate constants were calculated from the spectrophotometric data by the method of Guggenheimso or by the use of the usual first-order equation. The average arithmetic deviation from the Guggenheim plots varied from less than 1% in most cases to 4 or 5 % in the worst cases (with 3-picoline, 4-picoline and pyridine as catalysts-this was due to a slight decrease in rate after the reaction was about 70% complete). The rate constants for the imidazole-catalyzed hydrolysis of ethyl thiolacetate were always larger in the first 15% of the reaction, reasonably constant through the middle range and smaller after about 60-70% completion. The attempted hydrolyses of ethyl acetate were performed in the following manner. The ester and catalyst were mixed a t 25.04'. Aliquots of the mixTure were withdrawn and titrated in the indicated manner. m-Aminobenzoic acid, anthranilic acid, benzoic acid, pyridine, manganous chloride and sodium thiosulfate solutions were titrated to pH 8 with standard alkali using phenolphthalein or a Beckman model H pH meter; nickelous chloride and cupric chloride solutions

were titrated to pH 4.0 with standard acid using the PH meter. The hydrolysis of acetylcholine bromide with imidazole was followed by periodic examination of the infrared spectrum in a Perkin-Elmer double beam infrared spectrophotometer. The solutions were prepared as follows: imidazole was dissolved in a solution of DCI in deuterium oxide (99.8%) t o give a solution containing equal amounts of free amine and deuteronated amine; to this solution was added a solution of ester in deuterium oxide. The concentrations were adjusted so that the final solution was 0.10 M in ester and 0.43 M in each imidazole species. Calcium fluoride infrared cells were used for the analysis of the deuterium oxide solutions. The hydrolysis of methyl p-nitrobenzoate with imidazole mas followed by measurement of the rate of appearance of pnitrobenzoate ion with the Cary recording spectrophotometer. The ester has a maximum a t 266 mp and p-nitrobenzoate ion has a maximum a t 272 m p . However, a t 290 mp the optical density of the latter is much greater than that of the former so that the increase in optical density a t this wave length can be used as a measure of the hydrolysis reaction.

Results Hydrolysis of p-Nitrophenyl Acetate by Tertiary Amines.-In Table I are listed the catalytic rate
TABLE I HYDROLYSIS OF p-NITROPHEKYL ACETATEBY TERTIARY THE AMINES~
ko X 102

Catalyst

pH

I./mole' ki X lo'. sec. sec-1

log k a

fiK? of

amineb

0.0531 -3.275 5.23' Pyridine 6 .35 ,261 -2.583 5.66d 3-Picoline 6 .37 .795 0.52 -2.100 6.05" 4-Picoline 6 .08 N-Methylimidazole 7 .05 31.7 -0.499 7.00' 1.45 -0.323 7.04' Imidazole 7 .15 46.9 Trimethyl9.87 12.0 -1.006 9.72' amine 8 .35 510 2.699 15.7 Hydroxide ion 3.64 Imidazole 7 .85 5% dioxane-water, 26.2'. Aqueous solution, 25'. c A . Gero and J. J. Markham, J . Org. Chem., 16, 18.55 (1951). d E . F. G. Herington, Disc. Furuday Soc., 9 , 26 (1950). Determined in this investigation; Dedichenlg reported 7.34. f A . H. M. Kirby and A. Neuberger, Biochem. J., 32, 1146 (1938). OH. S. Harned and R. -4. Robinson, THISJOURNAL, 50, 3154 (1928).

(15) J. J. Blanksma, C h e m . Weekblad, 6, 725 (1909). (16) P. N. Rylander and D. S. Tarbell, THIS JOURNAL, 72, 3021
(1950).
(17) (18) (19) (20) T . Eauchi, Btrll. C h e m . SOC. Japan, 3, 228 (1928). 0. Wallach, Ber., 16, 645 (1882). G.Dedichen, ibid., 39, 1840 (1906). E.A . Guzgenheim, P h i l . M a g . , (71 2, 538 (1920).

constants, kc, for the hydrolysis of p-nitrophenyl acetate by several heterocyclic bases and one aliphatic tertiary amine. These constants were obtained from the slopes of lines of a plot of kobs versus the free amine concentration, using the least squares method (Fig. 1). In Fig. 1 it is seen that for six catalysts the observed hydrolytic rate constants varied in a linear fashion with the buffer (catalyst) concentration a t constant hydrogen ion concentration and a t constant ionic strength. It has already been pointed out that the rate of hydrolysis of p-nitrophenyl acetate by imidazole was independent of primary salt effects6; doubling the ionic strength of the medium (from 0.117 to 0.245 M ) in the hydrolysis of p-nitrophenyl acetate with 4-picoline also had no effect on the specific rate constant. The rate of alkaline hydrolysis of 9-nitrophenyl acetate was estimated from the intercepts of Fig. 1. If it is assumed that k H ( H f ) = 0, the intercepts, Ki, are equal to ~ H , O koH(OH-). A plot of Ki versus (OH-) then yields KOH as the slope and

165s

MYRON

L. BENDER AND BYRON \Ti. TURNQIJI?ST

Vol.

T!)

0.004

(Free amine), mole/l. 0.012 0.020

to be 510 I./mole sec. from least squared data of Table I and K H ~ Owas estimated a t 5 X 10-5 sec.-I. Hydrolysis of 2,4-Dinitrophenyl Acetate, Phenyl Acetate and Ethyl Thiolacetate by Various Bases.The catalytic rate constants for the hydrolysis of these esters were obtained in the same manner as those for p-nitrophenyl acetate. The constants are listed in Table I1 and the plots from which they were obtained are given in Fig. 3.
(Free dniitie,, mole I

OOO,?
1

OOli)

0015

18.0

15.0
e

1.

H 12.0
9.0
i

0-

4
x
0.2 0.4 0.6 0.8 1.0 (Free amine), moles/l. Fig. 1.-Hydrolysis of P-nitrophenyl acetate with various amines a t 26.2": A, imidazole; B, N-methylimidazole; C, trimethylamine; D, 4-picoline; E, 3-picoline; 'F,pyridine. Top scale A, B and C; bottom scale D, E and F.

6.0

3 .O

the intercept. The data in Table I were used for the plot illustrated in Fig. 2. The secondorder rate constant for hydroxide ion was estimated

~ H , Oas

0.1 0.2 0.3 04 (Free amine), mole/l. Fig. 3.-Hydrolysis of various esters with imidazole and acetate ion. Imidazole catalysis: A, 2,4-dinitrophenyl acetate (top and left scales); B, ethyl thiolacetate (bottom and right scales); C, phenyl acetate (bottom and left scales). Acetate catalysis: D, 2,4-dinitrophenyl acetate (top scale X lo* and left scale X lo-]).

0.5 1.0 1.5 2.0 Hydroxide ion concentration X lo6, moles/l. Fig. 2 -The alkalitie hydrolysis of p-nitrophenyl acetate in ,5% dioxane-water a t 26.2".

Attempted Hydrolysis of Ethyl Acetate, Acetylcholine and Methyl $-Nitrobenzoate with Various Acids and Bases.-A study has been made of the hydrolysis of ethyl acetate by various acids and bases. In all cases given in Table I11 there was no measurable amount of reaction a t the indicated period of time. The extent of hydrolysis of acetylcholine bromide by imidazole in deuterium oxide solution was estimated from the diminution of the carbonyl peal; of the ester a t 1730 em.-'. In twenty days a t 25' the optical density fell from about 0.264 to 0.140 indicating approximately 40y0 hydrolysis. It was assumed that the deuterium oxide solution was near neutrality since pk', was 7.04 for imidazole in water a t 25". The half-life calculated for the hydrolysis of acetylcholine in water a t pH 7.0 is about 60 days ( k o ~ 1.20 l.,/mole sec. in water a t I n view of the uncertainty in the true concentration of OD- and in the value for the hydrolysis constant by deuteroxide ion, catalysis by
(21) J Batterworth, D 30 (1953)

Bley a n d G S Stone, Biochein J ,

IS,

April 5 , 1057

BASICCATALYSIS OF ESTER HYDROLYSIS

1650

THE HYDROLYSIS
Ester

O F 2,4-DIXITROPHENYL ACETATE, PHENYL

TABLE I1 ACETATEAND ETHYL THIOLACETATE BY VARIOUS BAS&

Catalyst

PH

log ko

of alcohol

log kOE

Imidazole Quinoline Acetate ion Phenyl acetate Imidazole Ethyl thiolacetate Imidazole p-Sitrophenyl acetate Imidazole 26.2'. * 56% acetone-water, 25', ref. 22. line, 67, dioxane-water. e 57, dioxane-water. 70, 1308 (1948).

2,4-Dinitrophenyl acetate

-2.196 9.89 -0.240b -3.770 10.5 - 1,386* -0.323 7.16 0.906b Determined a t only one concentration of quinoAqueous solution, 25'. Aqueous solution. 62Y0 acetone-water, J . R . Schaefgen, THIS JOURXAL,

6.12 5.00 5.30 7.49 7.13

649

0.829

4.02

15.0d 0. O5ge
,588' .0166'

TABLE I11 ATTEMPTEDHYDROLYSIS OF ETHYL ACETATE WITH VARIOUS ACIDSA N D BASES^

(Ester),

Time, months

proton. The concentration of these intermediates would, therefore, not be expected to attain an experimentally measurable value. Catalysis of hydrolysis by acetate ion should give an anhydride
0

as intermediate by analogy with the acyl derivatives of the amines.23 The rate constant for the uncatalyzed hydrolysis of acetic anhydride is givenz4 as 2.5 X sec.-l in aqueous solution a t 25'; the similarity between this value and the value for N-acetylimidazole indicates that these two intermediates are of comparable stability. I t is not surprising that the rates of hydrolytic cleavage are imidazole must be considered to be unproven; close since the latter is the nitrogen analog of if such hydrolysis does exist, i t must be quite small. the former. It was not possible, however, to obThe amount of hydrolysis of methyl p-nitro- tain physical evidence for the existence of the benzoate in imidazole solution was determined by anhydride. Since good first-order rate constants the increase in the optical density a t 290 mp. were obtained whether the intermediate is short For a solution 1.82 X M in ester, 0.601 M or long-lived, the concentration of the intermediate in total amine and 0.430 M in free amine a t p H must have no effect on the determination of these 7.50 in 5% dioxane-water a t 25", the optical constants for our conditions of excess catalyst. density increased from 1.050 to 1.190 in six days; This result substantiates the earlier kinetic arguthe increase found after this length of time repre- ments.6 sents about 25y0 hydrolysis. The half-life of the General Basic Catalysis.-For general basic hydrolysis of this ester by hydroxide ion a t 25' catalysis to be operative, the rate must be proporwas approximated to be about 17 days a t p H 7.5 tional to the summation of catalytic constants by extrapolation of the known rate constant in 56% times catalyst concentrations a t constant PH and acetone-water to 5% dioxane-water.22 It there- ionic strength. It has been demonstrated that the fore appears that the small amount of hydrolysis hydrolysis of p-nitrophenyl acetate fulfills this found for methyl p-nitrobenzoate in imidazole requirement. A corollary of general basic casolution is explainable on the basis of an alkaline talysis is that a linear relationship exists between the reaction. log of the rate constant and the log of the equilibDiscussion rium constant of the catalyzing base, the BrgnThe ~ data from the hydrolysis Mechanism of the Reaction.-Catalysis of ester sted r e l a t i ~ n s h i p . ~ hydrolysis by bases other than hydroxide ion is of p-nitrophenyl acetate with various tertiary believed to proceed according to the mechanism amines given in Table I was used to make such a outlined for imidazole-catalyzed hydrolysisa in- linear free energy plot according to the method of volving attack of the ester by the base giving an Br#nsted, in which log k, is plotted zersus P K a acylammonium (or acid anhydride) intermediate (Fig. 4). It can be seen that the relationship is followed which is subsequently hydrolyzed by water, yielding carboxylate ion and regenerating the catalyst. over three powers of ten in the rate. TrimethylOnly in the imidazole case is there direct evidence amine shows a large negative deviation which is for intermediate formatioa6 I n none of the other (23) Cf.L. P. H a m m e t t , "Physical Organic Chemistry," McGraw1940, p. 356 ammonium ion intermediates is there the possibility Hill Book Co.,Inc., New York, N. Y.. (24) A. C. D. Rivett and N. V. Sidnwick, J . Chem. Soc., 97, 732 of stabilization of the intermediate by loss of a
( 2 2 ) E. Tommila and C . N. Hinshelwood, J. Chcm. SOC., 1801

m-ilminobenzoic acid 0.038 0.037 1.0 Anthranilic acid ,025 ,024 1 . 0 Benzoic acid-sodium benzoate .017 (0.017) .017 1.0 .040 ( ,040) ,040 1 . 0 Pyridine-pyridinium ion Manganese chloride ,033 ,033 1 . 0 .015 .Ol5 0.25 Nckelous chloride .010 .025 .25 Nickelous chloride" Sodium thiosulfate ,025 ,025 .25 Cupric chlorideb ,015 ,013 .5 Cupric chloride' .025 ,025 .5 a pH 7.5, 0.060 M tris-( hydroxymethy1)-aminomethane buffer. pI 6.5, 0.079 M tris buffer. PH 7.7, 0.050 M Aqueous solution, 25.04'. tris buffer.

09

-I

OR8

+
k5

HzO

R'C-OH

jl

+ B + ROH

(1)

(19381

(1910). (25) 5. N.B r m s t e d and K. J. Pedetsea, Z . p h y s i k . Chem., 108, 185 (1924).

that of quinoline. These two bascs have pK, values of 4.75 and 4.94, respectively; yet the ratio of their catalytic constants is 173. The present results appear to indicate that attack a t a carbonyl carbon atom is governed in part by the charge 011 the attacking species, with tincharged reagents being more eflicient than negatively cliargcd reagents of the same basicity. It has been shown by Pedersen for general basic b catalysis involving protoil transfer that if the measured rate copstai1t, &,tis, is equal to the rate constant for proton transfer, the slope, a , of the Rr@iisted catalysis law plot must have the limits 0 < cy < 1.27f26 It has been pointed out that iii such a situation as a approaches ullity, generd basic catalysis cannot be detected due to the rniich greater reactivity of hydroxyl ion as a catalyst.: The question might arise as to why general basic catalysis is observed in the present instance when the slope is 1.G2. This phenomenon is explicable in terms of the large negative deviation in log k I I I I of the hydroxide ion catalysis. If the hydroxide ion catalysis were included in an approximate 6 8 10 linear relationship, the slope would be in the range PKA. of 0.37-0.47 which is the region in which general Fig. 4.-Linear free energy relationship in the hydrolysis basic catalysis is often observed.* If the relaof P-nitrophenyl acetate: a,pyridine; B, 4-picoline; C, tionship shown in Fig. 4 is a true Br@nstedrelaimidazole; D, S-methyliinid3zole; E, trimethylamine; F, tionship, the slope of 1.62 indicates that in this 3-picoline. reaction kobs cannot he equal to the rate constant for proton transfer. It has been postulated that expected because of the much greater steric re- kc in these ester hydrolyses does not involve proton quirements of this amine compared t o that of the transfer and is equal to the complex rate constant, heterocyclic bases. The slope of the line deter- k l k 8 / ( k 2 k3) (equation I ) . The complexity of kc mined by the method of least squares is 1.62; makes it difficult to assess the limits of a iu the this is the largest slope recorded for such a linear present instance. free energy relationship. It can be seen that these ester hydrolyses conFigure 4 gives information concerning the form to general basic catalysis if one extends the identity of the catalytic nitrogen atom of imidazole. concept of basic catalysis to include not only a rateA priori, one would expect the imino nitrogen to be determining proton transfer but also a rate-determore basic (and more nucleophilic) since imidazole mining introduction of base into the substrate is an analog of a vinylamine (containing nitrogen to form an unstable addition compound, the in place of carbon) and since vinylamines are accepted mechanism for the hydroxide ion-catprotonated on carbon atom two of the double alyzed hydrolysis of esters and the favored ineclibond. N-Methylimidazole shows no deviation anism for the amine- and acetate-catalyzed hyfrom the linear free energy relationship; a deviation drolyses of esters. would be expected if the attacking nitrogen conThe Br@nsted relationship has been applied tained a methyl group in place of hydrogen (com- in the past, for the most part, to reactions in which pare trimethylamine). The nitrogen participating proton transfer occurs in the catalytic step n s in the catalytic attack must then be the imino originally set forth by Br@tistecl. There are two nitrogen. The intermediate obtained from this reports in which this relationship has been applied amine as catalyst, N-acetyl-N-methylimidazolium to reactions involving nucleophilic attack on carbon. ion, should be much less stable than N-acetyliniid- One is the hydrolysis of anhydrides2gin which the azole since the former cannot stabilize itself by loss rate was proportional to the basicity of aniincs of a proton while the latter can. The ready ca- including pyridine, 3- and 4-picoline and isociEinotalysis by imidazole (and N-methylimidazole) as line with a slope of 0.925 but was not proportional well as their high basicities is due to the resonance to the basicity of sterically hindered amines such stabilization of their respective quaternary salts. as quinoline, 2,G-lutidine and 2-picoline. The Comparison of the rate constant for the alkaline second example of such a relation is the catalysis of hydrolysis of pnitrophenyl acetate given in Table I the hydrolysis of chloroacetate by general bases. with the data in Fig. 4 indicates that the hydroxide The rate was shown to follow the basicity of the ion has an extremely large negative deviation in log catalyst over a large range of basicity involving k (cn. 11 powers of ten). The same type of negative 31 bases with a slope of 0.203. These exatnples deviation exists for the acetate ion-catalyzed hy(28) K . J . Pedersen, J . P h y s . C h e w . 58, 581 (1934). drolysis of 2,4-dinitrophenyl acetate compared to (29) V. Gold a n d IZ C.. jrfierson, .I C h e n i . Soc., 14014 (195:{).

26s27

( 2 6 ) Reference 2 8 , p ? Y O . (27) K. P. Bell, Acid-Base C.~;itnlysis,Onford U n i v . Press. I.on d o n , 1041, Chaps 5 . 7

( 3 0 ) 11 h l . D ~ w s o n ,I < I < , P \ w c k x u d (1943). (31) (:. I; S m i t h , ; b i d . , 521 (1043)

(>,

I C Smith, ~ b ~ 517 d ,

April 5, 1057

BASICCATALYSIS OF ESTER HYDROLYSIS

1661

together with the .present results bring up the question of whether general basic catalysis must involve proton transfer in the rate-determining step. The function of a general basic catalyst appears to include not only a rate-determining proton transfer but also the introduction of a base into the substrate to form an unstable i r ~ t e r m e d i a t e , which ~~.~~ is essentially the situation in the two cases cited above as well as in the present research. Correlation of the Rate of Hydrolysis with the Acidity of the Alcoholic Residue.-An attempt was made to correlate the catalytic rate constants of imidazole-catalyzed hydrolysis with the p K a S of the alcoholic residue of the various esters. The data in Table I1 were used for the plot illustrated in Fig. 5. Although a linear relationship does not exist between the rate constant and the acidity of the alcohol, it is evident that the rate increases with the acid strength of the alcohol derived from the ester in accord with the mechanistic arguments presented earlier. For a series of esters consisting of alcoholic and acyl portions of constant steric requirement, a linear free energy relationship would in all probability be applicable as has been found in the alkaline hydrolysis of substituted benzyl acetates where the rates followed the Hammett relation with a reaction constant of 0.760.34 Figure 5 also shows a plot in which cancellation of the steric effects of the alcoholic group of the ester is attempted by plotting the log of the rate constants of the alkaline hydrolysis versus the log of the rate constants for imidazole catalysis. The data are given in Table 11. For three esters the logarithms of alkaline and imidazole catalysis are linearly related. This relationship is similar to those obtained by Taft35 who found that the linear free energy relationships in ester hydrolysis were independent of solvent and attacking base (hydroxide and methoxide ions). The relationship between alkaline and imidazole catalysis breaks down when an extension to methyl $-nitrobenzoate is attempted. The relationship predicts catalysis of this ester by imidazole equivalent to that for phenyl acetate; experimentally there is little if any imidazole-catalyzed hydrolysis of methyl p-nitrobenzoate. Changing steric reTABLE IV
RATECONSTANTS
FROM

10

I1

sz
1

1 2 3 -log k O H . Fig. 5.-The imidazole-catalyzed hydrolysis of various esters in 5% dioxane-water a t 26.2: curve I, bottom scale; curve 11, top scale; A, p-nitrophenyl acetate; B, phenyl acetate; C, ethyl thiolacetate; D, 2,4-dinitrophenyl acetate. -1
0

quirements can explain part of this failure but it is probable that there are other factors. General Basic Catalysis of Ester Hydrolysis as a Model System for Enzymatic Hydrolysis.--Since it has been postulated that the active catalytic site of a-chymotrypsin consists of a general base, namely, imidazole, it is interesting to compare the kinetics of the a-chymotrypsin-catalyzed and imidazole-catalyzed hydrolyses of p-nitrophenyl acetate. These data together with other hydrolyses of p-nitrophenyl acetate are shown in Table IV. In discussing the data in Table IV, the enzymecatalyzed hydrolysis of this ester will be assumed to proceed through the r n e c h a n i ~ r n ~ ~ - ~ ~
0

CHICOR

II

/I + En +-(CHICOR. . kz
0

ki

0
.Eli)

ka ---+

FOR THE FORMATION O F p-NITROPHENOL p-NITROPHENYL ACETATE BY VARIOUS CATALYSTS

Catalyst

k, l./mole see.

Solvent, 5 % aqueous s o h

Temp.,
OC.

CH3CEn
0

I1

+ OR-

(2)

Ref.

Insulin 0.58 Isopropyl DFP-a-chymotrypsin 0.47 Isopropyl a-Chymotrypsin >33 Isopropyl a-Chymotrypsin 3 Isopropyl Imidazole 0,475 Dioxane Hydroxide ion 51 1 Dioxane Sec.-. b This itivestigation.

alc. 25 alc. 25 alc. 25 alc. 18

36

CH3CEn

I1

kp + H20 -+

CHaCOH

/I

+ En + H

(3)

36
36 37

The fact that the reactions involving insulin and DFP-a-chymotrypsin have the same rate constant
(36) B . S. Hartley and B. A. Kilby, Biochem. J., 66, 288 (1954). (37) H. Gutfreund. Disc. F a v e d a y Soc., 20, 107 (1955); 11. Gutfreund, Advances in Catalysis, Val. IX, in press, Academic Press, Inc., New York, 1956; H. Gutfreund and J. M. Sturtevant, Biochem. J., 63, 656 (1956). In the last paper serine, and not imidazole, is postulated to participate in step k , , thus liberating p-nitrophenol. I. B.Wilson and F. Bergmann, J. Bioi. Chem.. 186,683 (1950). (38) ADDEDIN PROOF.-From the reported results of G. H. Dixon, W. J. Dreyer and H. Neurath, THIS JOURNAL, 78, 4810 (1956), and of Gutfreund and Sturtevantn with respect to the pH dependence of steps ks and kd, it appears that imidazole participates in step k r and not ka. This implies that step k , may consist of a two-step process involving an acyl-imidazole intermediate.

20.2

(32) J. W. Baker and E. Rothstein, Handbuch der Katalyse. Val. 2, G. M. Schwab, ed., J . Springer, Vienna, 1940, p. 51. (33) E. L King, Catalysis, Vol. 11, P . H. Emmett, ed., Reinhold Pub. Corp., New York, N . Y . ,1955, pp. 389-397. (34) Reference 23, p . 191. (35) R . W. Taft, J r . , THISJ O U R N A L ,7 4 , 2729 (1952).

as imidazole indicates that the catalytic species of these non-enzymatic proteins is the imidazole ring of a histidine residue. Even though p-nitrophenyl acetate does not possess the structural requirements of a typical substrate of a-chymotrypsin, the data of Table IV indicate that the ester is hydrolyzed by the catalytically active site of this enzyme since diisopropyl fluorophosphate inhibits this reaction as well as the reactions of typical substrates. Confirmation of this hypothesis has come from the isolation of acetyl-cr-chymotrypsiii from the reaction of o-nitrophenyl acetate and a-chyniotrypsin. This conipound possesses one acetyl group per niolecule of a-chymotrypsin and is enzymatically iiiactive; the acetyl group is rapidly hydrolyzed in neutral solution, regenerating the active c~11zy111e.~~ . I large difference in rate exists between the rate of formation of p-nitrophenol by a-chymotrypsin on the one hand and by imidazole or DFP-a-chymotrypsin on the other hand. The large difference in rate between active and DFP-inhibited enzyme is most easily explained by the reaction of the ester a t the active site in the former case and a t an inactive imidazole ring in the latter. It is significant that there are two histidine residues in a-chymotrypsin which can account for these two sites,4o For the rate of the former reaction, the rapid spectrophotometric data of Gutfreund are preferred to those of Hartley and Kilby since Gutfreund's rate constant applies to ka of reaction 2 and is presumably analogous to the rate constant for the imidazole-catalyzed reaction. The equivalence of the rate constants for imidazole and DFP-chymotrypsin catalysis and the much larger rate constant for a-chymotrypsin catalysis indicates that if the active site of achymotrypsin involves an imidazole ring3' the mechanism is certainly more complex than a simple displacement on ester by imidazole to give an acylenzyme intermediate followed by hydrolytic cleavage of the intermediate. This conclusion is strengthened by the fact that esters of aliphatic alcohols are readily hydrolyzed by a-chymotrypsin, but there is little or no catalysis by imidazole. If the secondary imidazole in DFP-a-chymotrypsin is equivalent to imidazole alone and is thus not sterically hindered, the imidazole in a-chymotrypsin must coiitain a unique cnr-ironnient to cause such a large increase in rate. In contrast to the large difference in the rate of liberation of p-nitrophenol with a-chymotrypsin and imidazole, the rate of formation of acetate from the acetyl-enzyme and N-acetylimidazole are reasonably close. The former was reported as IO-? sec.-l ,36 whereas the latter is found to be
(:{!I) A . K . Balls and 11. S . Wood, J . Bioi. C h c m . . 219, 245 (1956). GO) J. FI. Northrup. 11. Kunitz and R . M. Herriott, "Crystalline l i n z y m c s , " Columbia Utiiv. Press, New York, N. Y . , 1948, p. 26.

1.3 X l o p 3 set.-'. Siiice the. rate of foririation of acid is usually measured in enzymatic catalysis, differences in rate for various substrates would imply that specificity is important in the hydrolytic cleavage of the intermediate as well as in previous steps. It should be pointed out that S-acetylimidazole, while a relatively reactive intermediate, is still quite selective since it will react preferentially with more powerful nucleophiles such as phosphate ion and thiols rather than with water. Catalysis of ester hydrolysis by imidazole by 110 ineans explains all the phases of enzymatic liydrolysis. However, there are some aspects of these reactions which appear to be siniilar (i.e., formation of an acyl-catalyst intermediate and its rate of hydrolysis) so that it is at least reasonable to postulate imidazole as a part of the active center in a-chymotrypsin and to use the catalysis by imidazole as a starting point for thc constructiori of a more complete iiiodel for cu-cliyiiiotryl)silicatalyzed hydrolysis. Hydrolysis of esters by iriiidazole may possibly serve as a model system or the enzyme papain. Smith4' has presented evidence that the action of this enzyme is due to formation of a thiol ester of the enzyme (transesterification of the ester with the sulfhydryl group of the enzyme) and subsequent hydrolysis of the thiol ester. Since the hydrolysis of a thiol ester is more readily catalyzed by imidazole than the hydrolysis of the corresponding oxygenated ester, i t is entirely feasible that papain catalyzes hydrolysis through such a mechanism. It should be pointed out that whereas the hydroxide ion-catalyzed hydrolysis of ethyl thiolacetate is only slightly faster than that of ethyl acetate,'6 the iinidazole-catalyzed hydrolysis of the former is apparently infinitely faster than that of the latter. The use of thiol esters in biological reactions is by no nieans limited to hydrolytic reactions. The important acyl transfer and oxidative decarboxylation reactions involve thiolesters such as acetylcoenzyme and acetl-lglutathione; it is possible that these reactions occur through acy1a:nine intermediates.
ADDEDI N PRoon.-After submission of this and t h e preceding paper, it was found t h a t many aspects of this research were carried out 111dependently and simultaneously b y T. C. Rruice and G. 1, Schrnir, :i~rh. Biochcnz. B i o p h y s . , 63, 484 (1!25ti), and Tms J O U I I N A L , 79, 1tXG ( l S 5 7 ) . Substantial agreement exists between t h e results of these investigations. T h e authors gratefully acknowledge the wiiortunity of reading the nlanuscript nf Bruice and Schmir prior t o publication.

Acknowledgment.-The authors gratefully :tcknowledge valuable discussions with Drs. F. Bergmann, G. Gorin, 3. Gutfreund, D. E. Koshland, Jr., N. Kilpatrick, F. H. Westheimer ant1 K. B. n'iberg.
CHICAGO 10, ILLINOIS