lutidine, 1.2-dimethylimidazole, and 1-methylimidazole were distilled ... 8.2) with m-cresol purple (pK, 8.3, 578 nm); 4-methyl imidazole (pK,. 7.8) with phenol red ...
Vol. 269, No. 27, Issue of July 8, pp. 17988-17992, 1994
THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.
Printed in U S A .
Comparison of Intra- and Intermolecular Proton Transfer in Human Carbonic Anhydrase11" (Received for publication, March 31, 1994)
Shinichi TaokaSQ,Chingkuang Tu$, Kurt A. Kistlea, and David N. Silverman$l/ From the $Department of Pharmacology and Therapeutics and Department of Biochemistry and Molecular Biology, University of Florida, Gainesville, Florida 32610 and the Wepartment of Chemistry, University of Pennsylvania, Philadelphia, Pennsylvania 19104
The catalysis of the hydration of CO, by human carbonic anhydrase I1 (HCA 11) includes the transfer of a proton from zinc-boundwater to histidine 64 utilizing a network of intervening hydrogen-bonded water molecules, then the proton is transferred to buffer in solution. We used stopped-flow spectrophotometry and "0 exchange between CO, and water measured by mass spectrometry to compare catalytic constants dependent on proton transfer in HCA I1 and in the mutant H64A HCA I1 containing the replacement Hisw-IAla. Maximal velocities and oxygen-18 exchange catalyzed by H64A HCAII showed that nearly all of the proton transfer with this mutant proceeded through the imidazole buffer. The followingparameters were verysimilar or identical in catalysis by H64A HCA I1 compared withcatalysis by wild-type HCA I1 both in the presence of large concenthe maximal rate of initrations of imidazole (100 m~): tial velocity and of exchange of "0 between CO, and water, solvent hydrogen isotope effects on the maximal velocity, and the dependence of these isotope effects on the atom fraction of deuterium in solvent water. These results indicate that the proton transfer involving the zinc-bound water in catalysis is not significantly affected by the difference betweenthe mobility of the free imidazole bufferand the side chain of His 64. Moreover, data for both the wild-type and mutant enzymes are consistent with proton transfer through intervening hydrogen-bonded water bridges in the active site. These features of the proton transfer are discussed in terms of a model in which the first proton transfer from the zincbound water to an adjacent water is rate limiting.
CO,
Hz0
+ EZnOH- s EZnHCO; EZnH,O
+ HCO;
(Eq' ')
The second stage comprises a series of proton transfer stepsby which the zinc-bound hydroxide is regenerated (Equation 21, including an intramolecular proton transfer from the zincbound water to another site on the enzyme and transfer from this site to buffer in solution (Steiner et al., 1975; Rowlett and Silverman, 1982). In the presence of ample buffer, the ratelimiting step for the maximum velocity is the intramolecular proton transfer between the zinc-bound water and a residue of the enzyme (Steiner et al., 1975; Lindskog, 1984). H ~ s ~ ~ - E Z ~+HB,a O H + H ~ s ~ ~ - E Z +~ O Bs HHis"-EZnOH- + BH'
(Eq. 2)
Here B is buffer in solution and H' to the left of His64designates protonation of this site, which acts as a shuttle to transfer protons between the zinc-bound water and buffer in solution (Tu et al., 1989a). There is now considerable evidence supporting the proton transfer function of His64in HCA 11. Silverman and Lindskog (1988) reviewed a number of experiments leadingindirectly to this hypothesis,includingsolvent hydrogen isotope effects (Steiner et al., 1975), product inhibition (Steiner et al., 19761, chemical modification (Tu et al., 1989b), andometal ion inhibition (Tu et al., 1981). The Ne2 of His64is 7.9 A from the zinc in HCA 11; this side chain extends out into the active site cavity and is separated from the metal by hydrogen-bonded water molecules observed in the crystal structure (Eriksson et al., 1988). The proton transfer most likely involves intervening water molecules since the side chain of His64 cannot extend close enough to the zinc-bound water for a direct transfer (Eriksson et al., 1988). The involvement of intervening wateris Human carbonic anhydrase I1 (HCA 11)' is a zinc-containing consistent with the solvent hydrogen isotope effect on the maxienzyme of molecular mass 30 kDa which is found predomi- mal velocity of hydration of CO, (Venkatasubban and Silvernantly insecretory tissues and red blood cells. The hydration of man, 1980). CO, catalyzed by carbonic anhydrase I1 occurs in two separate The proton transfer role of His64 has been shown by siteanddistinctstages.Thefirstisthe conversion ofCO, to specific mutagenesis in which His64 in HCA I1 has been reHCO, for which zinc-bound hydroxide is the main catalytic placed by Ala, a residue which cannot transfer a proton (Tuet unit (Silverman and Lindskog, 1988) (see Equation 1). al., 1989a; Forsman et al., 1988). Oxygen-18 exchange experiments carried out at chemical equilibrium in the absence of * This work was supported by Grant GM25154 from the National buffers demonstrated that proton transfer to the zinc-bound Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must there- hydroxide in H64A HCA I1was as much as 20-fold slower than in wild-type HCA I1 (Tu et al., 1989a). Other evidence for the fore be hereby marked "advertisement"in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. role of His64was the enhancement of activity of the less efficient 9 Current address: Dept. of Biochemistry, University of Nebraska, human CAI11by making the replacement Lys'j4 His; the Lincoln, NE 68583-0718. 1) To whom correspondence should beaddressed: Box 100267,Health resulting mutant had a pH profile for the maximal velocity of Center, University of Florida, Gainesville, FL 32610-0267.Tel.: 904- hydration of CO, consistent with proton transfer from the zinc392-3556;Fax: 904-392-9696. bound water to His64(Jewel1 et al., 1991). The abbreviationsused are: HCA, human carbonic anhydrase;H64A We have compared catalysis by HCA I1 and by the mutant H C A 11, the mutant of human carbonic anhydrase I1 containing the H64A HCA I1in thepresence of imidazole buffer. Several catareplacement His644Ala; Mops, 3-(N-morpholino)propanesulfonicacid; Taps, 3-[tris(hydroxymethyl)[methylaminopropanesulfonicacid; SHIE, lytic properties dependent on proton transfer were similar or identical: "0 exchange between CO, and water and maximal solvent hydrogen isotope effect.
~~~~
~
-
~
--j
17988
Proton fiansfer in Carbonic Anhydrase initial velocities, the solvent hydrogenisotope effect on the catalytic turnover kcat, and thedependence of this isotope effect on the deuterium content of water. These results indicate theextent to which the function of His64in proton transfer can be replaced by imidazole b a e r in solution, and canbe understood sein termsof a model in which the rate-limiting event in the quence of proton transfers is that from the zinc-bound water to the next adjacent water molecule forming a hydronium-like ion.
17989
'o~r"---i 00
2 . rc '
0
0
MATERIALSANDMETHODS Enzymes and Chemicals-Bacterial expression vectors containing the human CA I1 coding region derived from a cDNA clone2 were constructed with vectors of the PET series or their derivatives. Rosenberg et al. (1987) have described this class of expression vectors; a further description of their use in expression of carbonic anhydrase is given by Tanhauser et al. (1992). The site-specific mutation HisM Ala was inserted into the coding regionwith the method of Kunkel (1985),and the resulting mutant H64A HCAI1 was verifiedby DNA sequencing of the entire coding region. Both HCA II and the mutant H64A HCA I1 were purified by affinity chromatography (Khalifah et al., 1977) with purity estimated at greater than 98% by polyacrylamide gel electrophoresis. All carbonic anhydrase concentrations were estimated from the molar absorptivity of 5.48 x lo4 M" cm" at 280 nm (Nyman and Lindskog, 1964). Buffers and indicators were obtained from Sigma. The buffers 3,5lutidine, 1.2-dimethylimidazole, and 1-methylimidazolewere distilled under reduced pressure. Deuterium oxide(99.9atom % deuterium, Sigma) was distilled. Other reagents were used without further purification. Steady-state Measurements-Initial velocities of hydration of CO, were determined by stopped-flow spectrophotometry (Applied Photophysics model SF.17MV)measuring the rateof change of absorbance of a pH indicator (Khalifah, 1971; Rowlett and Silverman, 1982). Saturated solutions of CO, were made by bubbling CO, gas into distilled H,O or D,O at 25 "C. The concentrations of CO, in saturated solutions at 25 "C are 3.38 x 10.' M for H,O and 3.81 x lo-' M for D,O (Pocker and Bjorkquist, 1977).Aliquots were removed by allowing the solution to flow by gravity into a gas-tight syringe. Solutions for kinetic measurements were made by dilution of these saturated solutions with degassed H,O or D,O using two syringes. The buffer-indicator pairs (with the wavelengths observed) were: 3,5-lutidine (pK, 6.2) with chlorophenol red (pK, 6.3, 574 nm); imidazole (pK, 6.9) and 1-methylimidazole (pK, 7.2) with p-nitrophenol (pK, 7.1, 400 nm); 1,a-dimethyl imidazole (pK, 8.2) with m-cresol purple (pK, 8.3, 578 nm); 4-methyl imidazole (pK, 7.8) with phenol red (PK, 7.5,558 nm). Experiments were carried out at 25.0 f 0.2 "C with total ionic strength of solution maintained at 0.1 M using Na,SO,. Kinetic constants were estimated from initial velocities using ENZFITTER, a nonlinear least squaresanalysis (Leatherbarrow, 1987). All pH measurements are presented as uncorrected pH meter readings. This allows partial cancellation of the correction required of a pH meter reading to obtain pD and the change in pK, for almost all acids with pK, between 3 and 10. Oxygen-I8Exchange-The rate of the exchange of '*O between CO, and water and the rate of exchange offrom "C- to W-containing CO, were measured with a mass spectrometer (Extrel EXM-ZOO)using a membrane inlet (Silverman, 1982). These experiments give two rates in the catalytic pathway, R , is the catalyzed rate of interconversion of CO, and HCO, shown in Equation 3. --f
10'
5
6
7
8
9
pH
FIG.1. Variation with pH of the turnover numberk,,, for the hydration ofCO, catalyzed by wild-type HCA 11 (fizzed symbols) and H64A HCA I1 (open symbols) in the presence of 75 m~ 3,5lutidine (O), 100 m~ imidazole (A), 100 m~ l,f?-dimethyl imidazole (0) Temperature . was 25 "C and the totalionic strength of solution was maintained at 0.1 M by addition of Na2S0,.
tainties. The valuesof kc,&,, for H64A HCA I1 under the conditions described in the legend to Fig. 1can be described by a titration curve with an apparentpK, at 7.7 +- 0.1 and maximal at (5.7 2 0.6) x lo7M" s - I (data notshown). The value of kCa,jKm pH dependenceof k,,JKm for HCA I1is well studied with avalue of pK, 6.9 and a maximal value of k,,,jKm for hydration of 1.2 x 10' s-l given by Steiner et aZ. (1975) with similarvalues given by Khalifah (1971) and for bovine CA I1 by Pocker and Bjorkquist (1977). The lower values of k,,JKmand largervalues of pK, found in thisstudy, compared with these earliervalues, are due to inhibition by the largeconcentrations (75-100 mM) of buffers used in thiswork. Imidazole is a very weak inhibitor of HCA I1 and H64A HCA I1 (Tu et al., 1989a; see also Fig. 4). The ratiok,,/K, for hydration of CO, contains rate constants for the steps of catalysis up t o and including the departure of HCO;; that is,kCaJK,,, includes the stepsof Equation 1and does not include the inter- or intramolecular proton transfers of Equation 2. The identical results for kc,&,, in catalysis by HCA I1 and H64A HCA I1indicate that thereplacement His64+ Ala does not influence the steps of the interconversion of CO, and HCO,, as also concluded previously (Forsman et al., 1988; Tu et al., 1989a). The maximal values of kcatnear 1 x lo6 s-' for hydration were similar for catalysis by HCA I1 and H64A HCAI1 (Fig. 1).This is in agreement with the values of kc,, determined for HCA I1 (Khalifah, 1971; Steiner et al., 1975) and for bovine CA I1 (Pocker and Bjorkquist, 1977). The pH dependence of kc,, for the hydrationof CO, catalyzed by both the mutantH64A HCA I1 and the wild-type HCA I1 showed a pH dependence for each buffer determined by the fraction of total buffer in unprotonated form, the form which is capable of accepting a proton HCOO"0- + EZnH,O s EZn'80H- + CO, + H,O (Eq. 3) from the enzyme as in Equation 2. The solvent hydrogen isoRHzois the rate of the proton transfer-dependent release of l80-labe1ed tope effects (kcat),&Jl (where the subscripts0 and 1represent water from the enzyme, as shown in Equation 4. the atomfraction of deuterium insolvent) for hydration of CO, catalyzed by H64A HCAI1 as a function of the pK, of the buffer Hz0 a maximum at pK, 6.9 for the are given in Fig. 2. This ratio had EZnl*OH- + BH' * EZnlSOH,+ B e EZnH,O + H,'80 + B (Eq. 4) buffer imidazole for which (kcat)J(kcSJl = 3.6 f 0.4. The solvent are also given in Here BH' is a proton donor to the zinc-bound hydroxide; in these ex- hydrogen isotope effects (k,,JKm),,4k,,JKm), Fig. 2 and areclose to unity. In allcases, these are themaximal periments BH' is predominantly His64or imidazole buffer. steady-state constants measured at thehigh pHplateau region RESULTS for each buffer as in Fig. 1. Values of kJKm for the hydrationof CO, catalyzed by HCAII The steady-state turnover (kc,,), for hydration of CO, cataand H64A HCA I1 were identical, within experimental uncer- lyzed by H64A HCA 11 was measured as a function of n the atom fraction of deuterium insolvent. We observed that (Kc,,), D. A. Jewel1 and P. J. Laipis, unpublished results. had anexponential dependence on n (Fig. 31, whereas a plot of
17990
Proton Dansfer in Carbonic Anhydrase 4.01
0.oL
6.0
H64A HCA
i
T
7.0
8.0
I1
Wild-type HCA 11
1
Q K
FIG.2. Dependence on the buffer pK, of the solvent hydrogen isotope effects on k,, and k,,JK,,, for the hydration of CO, catalyzed by H64A HCA11. Filled symbols are (kcat)d(kCaJl and open symbols are (kCa~K,,,)d(kcaJKm),, where the subscripts are approximate valn n l ues of the atom fraction of deuterium in solvent water. Thefollowing 0 100 0 100 200 buffers were used: 3,5-lutidine(pK, 6.2); imidazole(pK, 6.9); I-methyl[Imidazole] mM [Imidazole] mM imidazole (pK, 7.2); 4-methylimidazole (pK, 7.8); and 1,2-dimethylimidazole (pK, 8.2). Temperature was25 "C, and the totalionic strength of FIG.4. Dependence of R , JEl (m) and RJ[El (0) on imidazole solution was maintainedat 0.1 M by addition of Na,SO,. Buffer concen- concentration for H64A HdA I1 (left)and HCA I1 (right). Here [El trations were 100 mM except for 3,5-lutidine which was 75 nux. In each is the total concentrationof enzyme, the pH was7.3, and the temperacase the pH of the measurement was in the plateau region of kc,, as ture was 10 "C. Total concentration of CO, and HCO; was 25 mM, and indicated in Fig.1;that is, at pH> pK,. Data aregiven as the mean and the total ionic strength of solution was maintained at 0.2 M by addition standard deviation from three tofive experiments. Thesolid line is a fit of Na,SO,. A least squares fit to a saturation curve gave a maximal of the Marcus rate theory to the data with parameters given in the text. value of RHzd[Elof (2.6 f 0.2) x IO5 for H64A HCA 11; the maximal value of 1.9 x lo5 s-l was found for wild-type HCA 11.
these two enzymes (see the legend to Fig. 4). The significant difference is that in the absence of buffer, HisM in HCA I1 provides the proton for donation to the zinc-bound hydroxide (Equation 4); with addition of imidazole, this rate isenhanced slightly (about 25% as shown in Fig. 4). For the mutant H64A is very low, HCA I1in theabsence of buffer, the value of RH2d[E1 but isenhanced greatly by imidazole (Fig. 4).Note in Fig. 4 the slight inhibitory effect of imidazole on R, for both enzymes. DISCUSSION
This study investigates the capacity of imidazole as buffer to replace the proton transfer function of His64in humancarbonic 0.0 0.2 0.4 0.6 0.8 1.0 anhydrase 11.We have approached this question by measuring rate constants and solvent hydrogen isotope effects for pron cesses dependent on proton transfer in humancarbonic anhyFIG.3. The turnover number (kcat),,for the hydration of CO, containing the replacement catalyzed byH64AHCA I1 plotted on a logarithmic ordinate drase I1 (HCA 11)and in the mutant against n, the atom fraction of deuterium in solvent water. His64 Ala (H64A HCA 11).In catalysis of CO, hydration by Measurements were made at 25 "C using 100 mM imidazole buffer in H64A HCA I1 in the presence of large concentrations of imidthe plateau region indicated in Fig. 1, uncorrected pH meter reading azole buffers, the maximal values of kc, are comparable with, 7.2-to 7.6; other conditions were as described in the legend to Fig. 1. Data are given as the mean and standard deviationfrom three to five but somewhat smaller than,kc,, for wild-type HCA 11(Fig. l),as had also been observed in previous studies (Forsman et al., experiments. 1988; Tu et al., 1989a). This feature is furtheramplified by the (kcat), on an arithmetic ordinate isclearly nonlinear and bulg- l80exchange data of Fig. 4 in which the saturable enhanceing down (not shown). These experiments were performed in ment of the rateof release of He180from the enzymes (Equation the high pHplateau region of kc,, using 100 mM imidazole at an 4) is compared. As demonstrated for intermolecular proton uncorrected pH meter reading of 7.2-7.6. Fig. 3 represents transfer inHCA I1 (Rowlett and Silverman, 1982; Pocker et al., values of (kc&/(kcaJ1 = 3.6 2 0.4 and anexponential dependence 1986) and intramolecular proton transfer in mutants of HCA on n to be compared with the results of Venkatasubban and I11 (Silverman et al., 1993), a majordeterminant of the magnifor proton transfer is the difference Silverman (1980) for bovine CA I1 at pH 8.2 using 50 mM 1,2- tude of these rate constants dimethylimidazole for which (kcat)d(kcat)l= 3.2 2 0.4 also with in pK, between the buffer as proton acceptor and theenzyme as proton donor, since these rates canbe described by linear free an exponential dependence on n. An additional view of the proton transfer between enzyme energy relationships such as the Brmsted plot. Fig. 1 shows of catalysis is maintainedfor H64A HCAI1 as and imidazole is provided in the "0 exchange experiments that this feature which yield RH?,,,the proton transfer-dependent rateof release in wild-type HCA 11. Very bulky and dipolar buffers such as of "0-labeled water from the active site observed at chemical Mops and Taps are inefficient proton acceptors in catalysis by equilibrium (Equation 4). Fig. 4 compares this rate asa func- H64A HCA I1 (4% of that for wild-type HCA I1 for 100 mM Mops tion of imidazole concentration for H64A HCA I1 and wild-type at pH 7.2, see Tu et al., 1989a1,presumably becausetheir strucby imidazole appears to be tures do not allow partial entry into the active site cavity. HCA 11. The enhancement of RHZO saturable for each enzyme and also has the feature that at the The solvent hydrogen isotope effect (SHIE) on kcatcatalyzed 3.6 (Fig. 21, and thedependence saturation level, the valuesof RH2,/[E]are nearly the samefor by H64A HCA11is as great as --j
Proton Anhydrase Tbansfer Carbonicin of this SHIE on the deuterium content of solvent water is exponential (Fig. 3). This SHIE on kc, is very close t o the values of 3.8 for HCA I1 (Steiner et al., 1975) and 3.3 for catalysis by bovine CA I1 (Pocker and Bjorkquist, 19771, and an exponential dependence of the SHIE of kcaton n was also found by Venkatasubban and Silverman (1980) for bovine CA I1 at pH 8.2 using 50 mM 1,2-dimethylimidazole.The rate-limiting step for kc,, is the proton transfer and the exponential dependence on n is consistent with proton transfer throughthe intervening hydrogen-bonded water molecules in theactive site (Venkatasubban and Silverman, 1980).The identical properties of the SHIE for wild-type and mutant suggest that imidazole buffer does not participate appreciably in direct proton transfer with the zincbound water, but like His64utilizes proton transfer throughthe hydrogen-bonded water network in the active site cavity. The SHIE near unity on kc,@,,, is consistent with many experiments whichshow thatthe interconversion of CO, and HCO; catalyzed by HCA 11, the first stage of catalysis as in Equation 1,proceeds without a rate contributing proton transfer (Simonsson et al., 1979; Silverman and Lindskog, 1988). A fit of Marcus rate theory, as applied to isotope effects by Kresge et dl. (1977), is indicated by the solid line in Fig. 2 for intermolecular proton transfer between buffers in solution and H64AHCA 11. This fit yields values of the intrinsic energy barrier AG*, = 0.6 2 0.3 kcaVmo1 and work terms wr- wp = 0.9 2 0.5 kcal/mol; these energies are very close t o those found for intramolecular proton transfer between His64 and the zincbound hydroxide in mutantsof HCA I11 (AGi,, = 1.3 * 0.3 kcaU mol and work terms w' - wp = 0.6 2 0.5 kcaVmol; Silverman et al., 1993). These results suggest that the energy barriers for the rate-limiting proton transfers are very similar for these carbonic anhydrases containing His64and the mutant H64A HCA I1 in the presence of imidazole buffer. As we have observed, the proton transfer involving the zincbound water in this catalysis is not greatly affected by the difference between the mobility of the free imidazole buffer and the relative immobility of the side chain of His64.Moreover, we believe it is unlikely that this proton transfer depends on a specific orientation of the imidazole ring as proton acceptor. The imidazole buffer in activating mutants of HCA I1 probably does not bind at the position of the side chain of His64in HCA 11, based on the observation of comparable values of k,, for wild-type HCA I1 and variants with lysine, glutamine, or alanine at position 64 (Forsman et al., 1988; this study used 1,2dimethylimidazole as buffer). It is interesting t o note that Krebs et al. (1991) showed that thecrystal structures of HCA I1 and T200S HCA I1 have different orientations of His64but have similar kinetic properties in thehydration of CO,. Such features of the proton transfer should be compared with the calculations of Liang and Lipscomb (1988) for HCA I1 who calculated the gas phase energy barriers for a sequential multiproton transfer between zinc-bound water and an ammonia molecule, with water molecules forming a bridge between them. The zinc was coordinated with three other ammonia molecules, and themolecular orientations were fully optimized for a hydrogen-bonded water bridge between the zinc and the terminal ammonia. The results, although qualitative, showed that theinitial proton transfer from the zinc-bound water to the first bridging water to form a zinc-bound hydroxide and hydronium ion is rate-limiting, with subsequent proton transfers between hydronium ions and adjacent water molecules having much smaller barrier heights. Calculations done without the terminal ammonia as acceptor showed the initial proton transfer t o be about the same barrier height, but subsequent transfers had significant barrier heights. Thus the presence of the terminal ammonia was the driving force necessary to eliminate
17991 mounting barriers after the initial proton transfer. In our experiments His64or imidazole buffer serves this function and may accept protons from many spatial orientations without disturbing the initialrate-determiningtransfer fromzincbound water to its nearest adjacent water molecule. The suggestion of the initial proton transfer as fully ratelimiting should be reconciled with the application of Marcus rate theory to proton transfer in mutantsof HCA I11containing His64(Silverman et al., 1993). These indicate a sizeable work function near 6kcaVmol describing an unfavorable equilibrium preceding the proton transfer in the hydration direction and near 10 kcaUmo1 in the dehydration direction; thermodynamic barriers of such magnitude could involve protein and solvent reorganization. These results suggest that although initial, rate-limiting proton transfer from zinc-bound water to an adjacent hydrogen-bonded water may be limiting, there is still considerable energy expended in orientation of side chains andor water to enhance the probability of completed proton transfer t o His64or imidazole rather than returnto reform the zinc-bound water. The calculations suggesting that the formation of a hydronium ion in the active site is therate-limiting event in catalysis provides an opportunity to understand the magnitude of the SHIE of 3.3 to 3.8 on kcat for hydration ofCO,by carbonic anhydrase (Steiner et al., 1975; Pocker and Bjorkquist, 1977; this work). The transition state for the intramolecular proton transfer with zinc-bound water as the donor is expected to be asymmetric with the transferring proton closer to the weaker base; that is, the hydronium ion is expected to be nearly completely formed in the transition state. Estimating the properties of the transition state using the fractionation factor +T = 0.69 for the hydronium ion (Schowen and Schowen, 1982) and +T = 0.79 for the aqueous ligand of the metal in its unprotonated form (determined using Co(I1)-substitutedbovine CA 11, Kassebaum and Silverman, 19891, wecan obtain an estimate of the overall isotope effect by use of the Gross-Butler equation (Schowen and Schowen, 1982): (kcat)% = (kcat)o(l- n + 0.79n)(l - n + 0.69n)3
(Eq. 5 )
This approach assumes that thefractionation factor of the zincbound water and water in theactive site cavity are both unity, which is thefractionation factor for water in bulk solvent. The result of (kcat)&at)l = 3.9 for n = 1.0 in Equation 5 is in good agreement with the observed value of 3.3-3.8. Such a calculation is also consistent with the exponential dependence of the SHIE on the atom fraction of deuterium in solvent as in Fig. 3. Acknowledgments-We thank Dr. A. Jerry Kresge for helpful comments and discussion. We aregratefulto Drs. Philip J. Laipisand Susan M. Tanhauser for their guidance in preparation of the site-specific mutant.
REFERENCES Eriksson, A. E., Kylsten, P. M., Jones, T.A., and Liljas, A. (1988)Proteins Struct. Funct. Genet. 4, 283-293 Forsman, C., Behravan, G., Jonsson, B.-H., Liang, 2.-W., Lindskog, S., Ren, X., Sandstrom, J., and Wallgren, K. (1988) FEBS LETZ? 229,360-362 Jewell, D. A., Tu, C., Paranawithana, S . R., Tanhauser, S . M., LoGrasso, P. V., Laipis, P. J., and Silverman, D. N. (1991) Biochemistry 30, 1484-1490 Kassebaum, J. W., and Silverman, D. N. (1989) J. Am. Chem. SOC. 111,2691-2696 Khalifah, R. G. (1971) J. Biol. Chem. 246, 2561-2573 Khalifah,R. G., Strader,D. J., Bryant, S . H., andGibson, S . M. (1977)Biochemistry 16,2241-2247 Krebs, J. F., Fierke, C. A., Alexander, R. S., and Christianson, D. W. (1991) Biochemistry 30,9153-9160 Kresge, A. J., Sagatys, D. S., and Chen, H. L. (1977) J. Am. Chem.Soc. 22, 7228-7233 Kunkel, T.(1985) Proc. Natl. Acad. Sci. (I. S. A. 82, 488-492 Leatherbarrow, R.J. (1987) Enzfitter: A NonlinearRegression Data Analysis Program for the IBM PC, Elsevier Biosoft, Cambridge Liang, J.-Y., and Lipscomb, W. N. (1988)Biochemistry 27, 8676-8682 Lindskog, S. (1984) J . Mol. Catal. 23, 357-368
17992
Proton Dansfer in Carbonic Anhydrase
Nyman, P. O., and Lindskog, S . (1964) Biochim. Biophys. Acta 86, 141-151 Pocker, Y., and Bjorkquist, D. W. (1977) Biochemistry 16,5698-5707 Pocker, Y.,Janjic, N., and Miao, C. H. (1986) in Zinc Enzymes (Bertini, I., Luchinat, C., Maret, W., and Zeppezauer, M., eds) pp. 341356, Birkhauser, Boston Rosenberg, A. H., Lade, B. N., Chui, D.-S., Lin, S.-W., Dunn, J. J., and Studier, F. W. (1987) Gene (Amst.) 66, 125-135 Rowlett, R. S., and Silverman, D. N.(1982) J . Am. Chem. SOC.104, 6737-6741 Schowen, K. B., and Schowen, R. L. (1982) Methods Enzymol. 87,551-606 Silverman, D. N. (1982) Methods in Enzymol. 87,732-752 Silverman, D. N., and Lindskog, S. (1988)Acct. Chem. Res. 21, 30-36 Silverman, D. N., Tu, C., Chen, X., Tanhauser, S. M., Kresge, J . A,, and Laipis, P. J . (1993) Biochemistry 32, 10757-10762
Simonsson, I., Jonsson, B.-H., and Lindskog, S. (1979) E m J. Biochem. 93, 409417 Steiner, H., Jonsson, B.-H., and Lindskog, S . (1975) Enr: J . Biochern. 59, 253-259 Steiner, H., Jonsson, B.-H., and Lindskog, S. (1976) FEBS Lett. 62, 16-20 Tanhauser, S. M., Jewell, D. A,, Tu, C. K., Silverman, D. N., Laipis, P. J. (1992) Gene (Amst.)117, 113-117 Tu, C., Wynns, G. C., and Silverman, D. N. (1981) J. Biol. Chem. 256,9466-9470 Tu, C., Silverman, D. N., Forsman, C., Jonsson, B. H., and Lindskog, S. (1989a) Biochemistry 28, 7913-7918 T u , C . K., Wynns, G. C., and Silverman, D. N. (1989b)J. Bid. Chem. 284,1238912393 Venkatasubban, K. S., and Silverman, D. N. (1980) Biochemistry 19,4984-4989
