Kinetic Analysis of Multiple Proton Shuttles in the Active Site of Human Carbonic Anhydrase*

Transcription

1 THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 277, No. 41, Issue of Octoer 11, pp , y The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Kinetic Analysis of Multiple Proton Shuttles in the Active Site of Human Caronic Anhydrase* Received for pulication, June 11, 2002, and in revised form, July 31, 2002 Pulished, JBC Papers in Press, August 8, 2002, DOI /jc.M Chingkuang Tu, Minzhang Qian, Haiqian An, Nina R. Wadhwa, David Duda, Craig Yoshioka, Yashash Pathak, Roert McKenna, Philip J. Laipis, and David N. Silverman From the Departments of Pharmacology and Therapeutics and Biochemistry and Molecular Biology, University of Florida College of Medicine, Gainesville, Florida We have prepared a site-specific mutant of human caronic anhydrase (HCA) II with histidine residues at positions 7 and 64 in the active site cavity. Using a different isozyme, we have placed histidine residues in HCA III at positions 64 and 67 and in another mutant at positions 64 and 7. Each of these histidine residues can act as a proton transfer group in catalysis when it is the only nonliganding histidine in the active site cavity, except His 7 in HCA III. Using an 18 O exchange method to measure rate constants for intramolecular proton transfer, we have found that inserting two histidine residues into the active site cavity of either isozyme II or III of caronic anhydrase results in rates of proton transfer to the zinc-ound hydroxide that are antagonistic or suppressive with respect to the corresponding single mutants. The crystal structure of Y7H HCA II, which contains oth His 7 and His 64 within the active site cavity, shows the conformation of the side chain of His 64 moved from its position in the wild type and hydrogen-onded through an intervening water molecule with the side chain of His 7. This suggests a cause of decreased proton transfer in catalysis. The caronic anhydrases in the class include the mammalian isozymes and are all zinc-containing monomeric enzymes, generally with molecular masses near 30 kda. These isozymes of caronic anhydrase catalyze the dehydration/hydration of HCO 3 /CO 2 y a two-stage or ping-pong mechanism in which the first step (Equation 1) is the reaction of icaronate with the enzyme containing zinc-ound water resulting in the formation of CO 2 and leaving zinc-ound hydroxide at the active site (1, 2). HCO 3 EZnH 2 O ^ CO 2 EZnOH H 2 O (Eq. 1) EZnOH BH ^ EZnH 2 O B (Eq. 2) The second stage is the regeneration of the zinc-ound water through proton transfer (1, 2). BH represents a proton donor that is a residue of the enzyme itself or uffer in solution. In the caronic anhydrases, the maximal velocities of catalysis and the rates of exchange of 18 O etween CO 2 and water are limited y the intramolecular proton transfers indicated in Equation 2. This was initially determined y oservation of a solvent deuterium isotope effect of 3.8 on k cat for hydration catalyzed y human caronic anhydrase (HCA) 1 II, among the most efficient isozymes in the class of caronic anhydrases (3). The solvent hydrogen isotope effect was 2.4 for 18 O exchange catalyzed y HCA III, the least efficient of the class (4). Susequent results including ph profiles, uffer activation, and computer simulations for these isozymes support the rate-limiting nature of proton transfer in these isozymes (4 7). For HCA II, the catalytic turnover is near 10 6 s 1. For this isozyme the predominant shuttle residue has een identified as His 64 (3, 5), the side chain of which extends into the active site cavity without apparent interactions with other residues of the protein (8, 9). The imidazole ring of His 64 is aout 7.5 Å from the zinc when this side chain is in the in conformation pointing toward the metal (8). This is too far for direct proton transfer to the zinc-ound hydroxide requiring proton transfer through intervening hydrogen-onded water ridges. Water molecules that are apparently hydrogen-onded have een oserved in the active site cavity in crystal structures of HCA II (8 10), suggesting possile proton transfer pathways (11, 12). Caronic anhydrase is a good model system for investigating properties of proton transfer in a protein environment; in many cases we know the identity of the proton donor and acceptor, and the rate constant for proton transfer can e easily determined ecause it is often rate-determining in catalysis. Moreover, the range of rate constants for proton transfer is rather large, from 10 3 s 1 for isozyme III to 10 6 s 1 for intramolecular proton transfer in HCA II. Recent discussion has focused on the features of the enzyme that influence this rate of intramolecular proton transfer and why it cannot e made faster (11, 12). The presence of multiple proton shuttle pathways is a possiility in many systems under current investigation such as cytochrome c oxidase (13), glycinamide rionucleotide transformylase (14), acteriorhodopsin (15), and the acterial photosynthetic reaction center (16), in which the proton transfer pathway involves many amino acid side chains and water molecules. In the case of HCA II and III, we have enzymes for which there is considerale previous work on proton transfer, one in which the effect of multiple proton shuttle groups can e readily assessed. We investigate here the kinetic properties of caronic anhydrases in the case that a single active site accommodates two proton shuttle residues. To investigate this possiility we have con- * This work was supported y Grant GM25154 from the National Institutes of Health. The costs of pulication of this article were defrayed in part y the payment of page charges. This article must therefore e herey marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The atomic coordinates and structure factors (code 1LZV) have een deposited in the Protein Data Bank, Research Collaoratory for Structural Bioinformatics, Rutgers University, New Brunswick, NJ ( To whom correspondence should e addressed: Box Health Center, University of Florida College of Medicine, Gainesville, FL Tel.: ; Fax: The areviation used is: HCA, human caronic anhydrase This paper is availale on line at

2 Proton Transfer in Caronic Anhydrase Resolution shells Å TABLE I X-ray diffraction data statistics No. of unique reflections % data a R sym a % data I/I( ) 2.0. R sym hkl i I i (hkl) I i (hkl) / hkl i I i (hkl), where I i (hkl) is the average intensity over symmetry equivalent reflections. TABLE II Model refinement statistics a R cryst R free No. of residues No. of atoms 2093 No. of H 2 O molecules 55 Root mean square deviation for ond lengths (Å) Root mean square deviation for angles (deg) Ramachandran statistics (%) Most favored regions 85.1 Allowed regions 14.0 Generously allowed regions 0.9 Disallowed regions 0.0 B-factors (Å 2 ) Average main chain atoms 21.3 Average side chain atoms 22.0 Solvent 25.6 a R cryst ( F P(os) F P(calc) / F P(os) ). R free was calculated using 5% of data excluded from refinement. structed mutants of HCA II and HCA III each containing two histidine residues; each histidine residue, with one exception, acts as a proton shuttle in the asence of the second. We first comment on the proton transfer capacity of histidine residues at single sites in caronic anhydrase and then discuss the data with two histidine residues in the cavity. MATERIALS AND METHODS Enzymes Human caronic anhydrases II and III and mutants were prepared using acterial expression vectors optimized for site-specific mutagenesis and protein synthesis (17). The vectors were derived from the T7 expression vectors of Studier et al. (18) and contained a acteriophage f1 origin of replication for production of single-stranded DNA. Both single-site and cassette mutants were prepared with these vectors, and protein expression was in the range of 5 30 mg/liter. All of the mutations were confirmed y DNA sequencing of the expression vectors. Wild type and mutant forms of HCA II were purified y affinity gel chromatography (19); wild type and mutant forms of HCA III were purified y gel filtration and ion exchange chromatographies (20). The resulting enzymes were greater than 95% pure determined y gel electrophoresis. The concentrations of wild type and mutants of HCA II were determined y active site titrations using the tight inding inhiitor ethoxzolamide. Concentrations of wild type and mutants of HCA III were determined from the extinction coefficient M 1 cm 1 at 280 nm (21). Oxygen-18 Exchange We used mass spectrometry to measure the catalyzed and uncatalyzed rates of exchange of 18 O from species of CO 2 into water and the rates of exchange of 18 O etween 12 C-containing and 13 C-containing species of CO 2 at chemical equilirium. Because this method is carried out at chemical equilirium, ph control is not a significant issue, and the experiments are performed with solutions containing no added uffers. This is an advantage in measuring intramolecular proton transfer ecause there is not interference y intermolecular proton transfer etween the enzyme and the uffer. Equations 3 and 4 demonstrate the catalytic pathway for the exchange of 18 O from icaronate to water. In Equation 4, B is a uffer in solution and/or an amino acid side chain in the enzyme. HOCO 18 O EZnH 2 O ^ EZn 18 OH CO 2 H 2 O (Eq. 3) H 2 O EZn 18 OH BH ^ EZn 18 OH 2 B - 0 EZnH 2 O H 18 2 O B (Eq. 4) Two rates in the catalytic pathway can e determined y this method; oth result from solution of the kinetic equations including singly and multiply laeled CO 2 species (22). The first is R 1, the rate of interconversion of CO 2 and HCO 3 at chemical equilirium. Equation 5 expresses the sustrate dependence of R 1. R 1 / E k ex cat S / K S eff S (Eq. 5) Here [E] is the total enzyme concentration, k ex cat, is a rate constant for maximal HCO 3 to CO 2 interconversion, [S] is the sustrate concentration of HCO 3 and/or CO 2, and K S eff is an apparent sustrate inding constant (23). This equation can e used to determine the values of k ex cat /K S eff when applied to the data for varying sustrate concentration or to determine k ex cat /K S eff directly from R 1 when [S] K S eff. In oth theory FIG. 1.The ph dependence of R H2O /[E] (s 1 ) catalyzed y wild type HCA II ( ), Y7H HCA II (Œ), Y7H/H64A HCA II (E), and H64A HCA II ( ). The values were determined y 18 O exchange at 25 C using solutions containing no uffers; ionic strength of solution was maintained at 0.2 M y addition of Na 2 SO 4. The lines are least squares fits of Equation 6 to the data yielding the values of pk a of the donor and acceptor groups and the values of k B given in Tale III. The data for H64A HCA II is from Qian et al. (30). and practice, k ex CO2 cat /K eff is equivalent to k cat /K m for CO 2 hydration as measured y steady state methods (22, 23). Also determined y this method is R H2O, the rate of release from the enzyme of water laeled with 18 O (Eq. 4). A proton donated from a donor group BH converts the zinc-ound hydroxide to zinc ound water, which readily exchanges with unlaeled water. The 18 O lael is greatly diluted into the solvent water. The value of R H2O can e interpreted in terms of the rate constant from a predominant donor group to the zinc-ound hydroxide according to Equation 6 (7), in which k B is the rate constant for proton transfer to the zinc-ound hydroxide, K a donor is the ionization constant for the donor group, and K a ZnH2O is the noninteracting ionization constant of the zinc-ound water molecule. R H2O / E k B / 1 K a donor / H 1 H /K a ZnH2O (Eq. 6) An Extrel EMX-200 mass spectrometer and a memrane inlet permeale to dissolved gases was used to measure the rate of distriution of 18 O (22). The experiments were carried out in the asence of uffers, which were not needed to maintain ph ecause these experiments were carried out at chemical equilirium. Crystallography Crystals of Y7H HCA II were produced y the hanging drop method (24), using a 10 mg/ml protein solution in 50 mm Tris-HCl, ph 7.8. The crystallization drops were otained y mixing 5 l of protein solution with 5 l of precipitant solution. Each drop was equilirated against 1 ml of precipitant solution, consisting of M (NH 4 ) 2 SO 4,50mM Tris-HCl, ph 7.8. The crystals were grown at 4 C, appeared within 3 days, and grew to full size within a week. X-ray diffraction data for Y7H HCA II were collected using an R- AXIS IV image plate system with Osmic mirrors and a Rigaku HU-H3R CU rotating anode operating at 50 kv and 100 ma. The data were collected using a 0.3-mm collimator with a detector to crystal

3 38872 Proton Transfer in Caronic Anhydrase TABLE III Maximal values of k cat /K m for hydration of CO 2 and rate constants k B for proton transfer to zinc-ound hydroxide in human caronic anhydrase II and mutants determined y rates of exchange of 18 O etween CO 2 and water Rate constants k B were otained y a nonlinear least squares fit of Equation 6 to the data for 18 O exchange (R H2O /[E]). Experimental conditions were as descried in the legend to Fig. 1. Variant of HCA II (Fig. 4a) k B pk a (donor) a pk a (ZnH 2 O) a k cat /K m pk a (ZnH 2 O) ms 1 M 1 s 1 Wild type Y7H Y7H/H64A c H64A 10 d d e a These values of pk a were determined from the ph profiles of R H2O /[E] in Fig. 1. These values of k cat /K m and pk a were determined from the ph profile of k cat /K m for hydration as determined y 18 O exchange. c The high ph component of the ell-shaped curve was not well determined, and the pk a of the donor is estimated as a lower limit. d The ph profile was too complex to assign values of pk a. e The data showed evidence of more than one pk a. 5.8 FIG. 2.Stereo view of the active site of wild type HCA II (A) and Y7H HCA II (B). Shown are the positions of the relevant amino acids and solvent. The side chains of His 7 /Tyr 7, His 64, and Trp 5 are depicted as stick drawings, and Zn 2 and solvent are yellow and red spheres, respectively. The main chain atoms of the model are represented y a gray coil. A2 F o F c electron density map (green) contoured at 2 for the side chains of His 7 /Tyr 7 and His 64 show the quality of the map. The panels were created using BOBSCRIPT (44) and RASTER3D (45). distance of 150 mm and the 2 fixed at 0. The data were collected at 300 K using three crystals yielding 121 of data. A total of 53,414 reflections were measured to a maximum resolution of 2.3 Å. All of the frames were collected using a 1 oscillation angle with an exposure time of 180 s/frame. The crystals of Y7H HCA II elonged to the space group P2 1 with unit cell dimensions a 42.9, 41.6, and c 73.0 Å with The data were merged to a set of 9586 independent reflections (85.0% complete) with DENZO and SCALEPACK (25), resulting in an R sym of 15.1% (Tale I). The structure of H64A HCA II (Protein Data Bank code 1G0F) (10) including the zinc ion, mercury ion, all water molecules, and with Tyr 7 mutated to Ala was used as the starting phasing model for Y7H HCA II using the software package Crystallography and NMR System (CNS) (26). Having Ala residues at positions 7 and 64 removed any phase ias in the initial map of the positions of His 7 and His 64. The initial structure was phased to 2.3 Å resolution. After one cycle of rigid ody refinement, geometry restrained positional refinement 2 F o F c and F o F c Fourier maps were calculated. Using the graphics program O, version 7 (27), showed the position of the zinc ion and side chains of His 7 and His 64. The zinc ion and side chains were assigned to the model, which was then further simulated, annealed, and refined y heating to 3000 K and gradual cooling, followed y temperature factor refinement with CNS (26). The model was then further refined y an iterative process with computer graphics-assisted molecular modeling prior to the next cycle of refinement. Solvent molecules were placed into the model using FIG. 3. The ph dependence of R H2O /[E] (ms 1 ) catalyzed y wild type HCA III ( ), K64H HCA III ( ), R67H HCA III ( ), and the doule mutant K64H/R67H HCA III (E). The values were determined y 18 O exchange at 25 C using solutions containing no uffers; ionic strength of solution was maintained at 0.2 M y addition of Na 2 SO 4. The lines are least squares fits of Equation 6 to the data yielding the values of pk a of the donor and acceptor groups and the values of k B given in Tale IV. the automated water picking algorithm from CNS (26). Convergence of refinement was deemed when no new water molecules could e placed and refined into the model. Tale II gives a full description of the refinement statistics of the final Y7H HCA II model. The coordinates and structure factors have een deposited in the Protein Data Bank with code 1LZV. RESULTS Human Caronic Anhydrase II We prepared the mutant Y7H/H64A HCA II containing His 7 as the single (nonliganding) histidine residue in the active site cavity. There are no other apparent proton shuttle residues. The rate constant R H2O /[E] for catalysis y this mutant is presented in Fig. 1. R H2O /[E] is a rate constant for release from the enzyme of 18 O-laeled water, which is determined y proton transfer from a donor residue to the zinc-ound hydroxide as descried in Equation 4 (5, 7, 22). The ell-shaped ph profile for R H2O /[E] reflects the values of pk a of the proton donor and acceptor involved and can e descried y Equation 6 for Y7H/H64A as well as for wild type HCA II, also shown in Fig. 1. These procedures give a fitted value of the rate constant k B for proton transfer from a donor group to the zinc-ound hydroxide as well as information on values of the pk a of the proton donor(s) and the zinc-ound water molecule. The values resulting from a fit of these data to Equation 6 are presented in Tale III. We have confirmed the value of pk a of the zinc-ound water otained in these fits from independent information, the pk a determined from the ph

4 Proton Transfer in Caronic Anhydrase TABLE IV Maximal values of k cat /K m for hydration of CO 2 and rate constants k B for proton transfer to zinc-ound hydroxide in human caronic anhydrase III and mutants determined y rates of exchange of 18 O etween CO 2 and water catalyzed y these enzymes Rate constants k B were otained y a nonlinear least squares fit of Equation 6 to the data for 18 O exchange (R H2O /[E]). Experimental conditions were as descried in the legend to Fig. 3. Variant of HCA III k B pk a (donor) a pk a (ZnH 2 O) a k cat /K m pk a (ZnH 2 O) ms 1 M 1 s 1 Variants of Fig. 4 Wild type c K64H d,e not determined f 5 R67H g K64H/R67H h Variants of Fig. 4d F198L i K64H/F198L e Y7H/F198L Y7H/K64H/F198L Variants of Fig. 4c (see wild type and K64H aove) Y7H Y7H/K64H , 8.6 h a These values of pk a were determined from the ph profiles of R H2O /[E] such as shown in Fig. 3. These values of k cat /K m and pk a were determined from the ph profile of k cat /K m for hydration as determined y 18 O exchange. c In Ref. 30. d In Ref. 4. e In Ref. 7. f In Ref. 43. g In Ref. 29. h These data fit two pk a values. i In Ref. 42. profile of k cat /K m for hydration (also given in Tale III). The data for wild type HCA II are in good agreement, with values previously reported from steady state data with a rate constant for intramolecular proton transfer k B at 800 ms 1 (Tale III) (2, 5, 28). The same procedure for Y7H/H64A gives a value of k B at 120 ms 1. In Tale III, these are compared with the value of k B near 10 ms 1 for H64A HCA II, which has no histidine in the active site cavity (Fig. 1). We have also prepared the mutant Y7H HCA II, which has two histidine residues in the active site cavity, His 7 and His 64. The ph profile for R H2O /[E] is again ell-shaped (Fig. 1) with the fitted values from Equation 6 given in Tale III. For this variant containing two His residues in the active site cavity the value of k B is 130 ms 1 (Tale III). Analysis of 2 F o F c and F o F c electron density maps of Y7H HCA II show the positions of oth His 7 and His 64 and the presence of a water molecule (305) forming a hydrogenonded ridge etween them (Fig. 2B). This water molecule 305 occupies a position 2.83 Å from N 1 of His 7 and 2.71 Å from N 1 of His 64 (Fig. 2B). Superposition of the Y7H HCA II model with wild type HCA II (crystallized under the same conditions as Y7H HCA II; results not shown) revealed an equivalent water molecule (426) in the wild type HCA II structure (Fig. 2A) 1.9 Å from the position of water molecule (305) in Y7H HCA II (Fig. 2B). There was a conformational change of His 64 in Y7H HCA II compared with wild type. In the wild type HCA II model, His 64 is in a conformation corresponding to the out position (8); for His 64 in the Y7H HCA II model, the 1 torsion angle is 55 (49 in wild type), and the 2 torsion angle is 62 ( 86 in wild type). Further analysis of the superposition of Y7H HCA II and wild type HCA II models showed that His 7 (Y7H HCA II) and Tyr 7 (wild type HCA II) occupy a similar spatial position and that there is no significant difference in the zincound solvent molecule (301). Three other water molecules (313, 314, and 333) in wild type HCA II that extend etween the zinc-ound solvent molecule and His 64 are not seen in the Y7H HCA II structure (Fig. 2). The values of k B for intramolecular proton transfer in the mutants of Tale III vary over a range of aout 80-fold; however, the maximal values of k cat /K m for hydration of CO 2 vary y at most 2-fold. The ratio k cat /K m contains rate constants for the conversion of CO 2 into icaronate in the first stage of catalysis (Equation 1). These oservations indicate that the mutations we have made do not alter the structure of the enzyme in a manner that causes large changes in activity of this first stage of catalysis. Human Caronic Anhydrase III Wild type HCA III does not have a histidine proton donor in the active site cavity (residue 64 is lysine) and has values of R H2O /[E] independent of ph in the range of ph 5 8 with a value of k B near 3 ms 1 (Fig. 3 and Tale IV). However, effective proton transfer in the catalysis has een oserved with the mutant K64H HCA III (4) and the mutant R67H HCA III (29). In each of these cases the rate constant k B was enhanced at least 7-fold with respect to the wild type HCA III (Tale IV). Fig. 3 shows the ell-shaped ph profiles for R H2O /[E] catalyzed y the doule mutant K64H/R67H HCA III and the single mutant R67H HCA III; the data for K64H are from Jewell et al. (4) and show the increase in R H2O /[E] as ph is decreased consistent with proton donation y His 64. These data are readily fit y Equation 6 for a single prominent proton donor with data given in Tale IV. We also investigated the potential for His 7 in HCA III to act as a proton donor in catalysis. The single mutant Y7H HCA III showed only a small increase in k B over wild type (Tale IV), and the ph profile was rather featureless, suggesting that this small increment is not caused y proton transfer from His 7 (data not shown). The additional replacement of Y7H/F198L in HCA III greatly affected proton transfer (Tale IV). Residue 198 has its side chain in the active site cavity rather close to the zinc; in HCA II this residue is Leu. The ph profile of R H2O /[E] for Y7H/F198L HCA III was ell-shaped, and its fit y Equation 6 suggests enhanced proton transfer through His 7 with k B near 35 ms 1, a value 3-fold greater than k B for F198L HCA III (Tale IV). We prepared two additional mutants containing two histidine residues in the active site cavity of HCA III: Y7H/K64H and Y7H/K64H/F198L. For each enzyme, we carried out the

5 38874 Proton Transfer in Caronic Anhydrase 18 O exchange experiment and fit the ph profiles with Equation 6. The results for all of the mutants are presented in Tale IV. The ph profiles for k cat /K m for the hydration of CO 2 catalyzed y HCA III and mutants were determined y 18 O exchange with maximal values and apparent pk a of the zinc-ound water molecule given in the last two columns of Tale IV. The value of pk a 5.3 for wild type HCA III was reported y Qian et al. (30) ased on data for k cat /K m and attriuted to the zinc-ound water. The values of the pk a of the zinc-ound water estimated in this manner were in general agreement with the values otained from the ph profiles of R H2O /[E] (Tale IV). In the case of mutants of HCA III, there is enhancement of k cat /K m caused y introduction of His at residues 7 and 67 and Leu at residue 198. To some extent this can e attriuted to the increase in the pk a of the zinc-ound water, which then renders the zinc-ound hydroxide more nucleophilic in the CO 2 hydration step (Equation 1). Chen et al. (31) and LoGrasso et al. (32) have discussed the effect on catalysis of the mutation at residue 198. Two of these mutants, Y7H/F198L and Y7H/K64H/F198L, were somewhat unstale at ph 6 and showed evidence of decay during the course of the 18 O exchange experiments. DISCUSSION In this study we have two histidine residues in the active site of HCA II at residues 7 and 64. Using a different isozyme, we have placed two histidine residues in HCA III at residues 64 and 67 and, in a different experiment, at residues 64 and 7. Each of these histidine residues can act as a proton transfer group when it is the only nonliganding histidine in the active site cavity, except His 7 in HCA III. We investigate the properties in catalysis y variants of caronic anhydrase with two proton shuttle sites within the active site cavity and along the proton transfer pathway extending from the zinc-ound solvent molecule to solution. Human Caronic Anhydrase II We have presented evidence in Tale III and Fig. 1 that His 7 in HCA II can act as a proton donor to the zinc-ound hydroxide during catalysis. The rate constant for this intramolecular proton transfer k B is descried y Equation 6. The maximum value of k B for Y7H/ H64A HCA II from Fig. 1 is near 120 ms 1 (Tale III), which means that His 7 in Y7H/H64A HCA II appears to e only aout 15% as efficient in proton transfer as His 64 in wild type HCA II, for which k B is 800 ms 1 (Tale III). Tyr 7 in HCA II has its phenolic hydroxyl 7 Å from the zinc and extends into the active site cavity; moreover, the hydroxyl of Tyr 7 participates in hydrogen onding with the water structure in the active site that also forms hydrogen onds with the zinc-ound water (9). Hence, this residue is placed within the active site cavity such that proton transfer involving His 7 is readily explained. The replacement of Tyr 7 in HCA II y Phe resulted in marginally reduced activity, indicating that Tyr 7 does not have a significant role in proton transfer (33), proaly a result of its asic pk a. His 7 occurs naturally in murine caronic anhydrase V and has no significant proton transfer role in catalysis (34). We used a doule mutant cycle (35) to compare our estimates of the rate constants for proton transfer k B for catalysis y the mutants of HCA II containing oth His 7 and His 64. His 7 and His 64 in Y7H HCA II have an antagonistic interaction when values of k B and the changes in free energy calculated from them are compared (Fig. 4a). In Fig. 4a we consider two single mutations, each of which introduces a proton shuttle in the enzyme. These alter the activation arriers to catalysis, descried y G 1 and G 2 when acting separately ( G RTln[k B (mutant2) /k B (mutant1) ). The change in the activation arrier for the doule mutant is G 1 2. The value of G 1 2 in Fig. 4a for Y7H HCA II ( 1.4 kcal/mol) is greater than G 1 for the variant resulting in the most change ( 2.6 kcal/mol). This FIG. 4.Comparisons of the rate constants for intramolecular proton transfer for HCA II and mutants (a) and HCA III and mutants ( d). The values of k B (ms 1 ) appear eneath each designated mutant, and the values adjacent to the arrows are the free energy changes ( G in kcal/mol) for the arriers to proton transfer for the corresponding mutations calculated from k B ( G RTln[k B (mutant2) / k B (mutant1) ]. The experimental conditions are as given in the legend to Fig. 1 for the variants of HCA II and in the legend to Fig. 3 for the variants of HCA III. is an antagonism in the classification of doule mutant interactions as descried y Mildvan et al. (Ref. 36; however, ecause we have oserved increases in enzyme activity, the signs of the changes in free energy arriers G corresponding to antagonistic and synergistic interactions are reversed compared with the notation of Mildvan et al.). That is, in the presence of His 7 the proton shuttle efficiency of His 64 is reduced. The crystal structure of Y7H HCA II suggests the source of this result. It may arise from a hydrogen onding interaction etween the neary residues His 7 and His 64 that hinders the moility of the side chains; that is, oth His 7 and His 64 are hydrogen-onded to water molecule 305 shown in Fig. 2B. The

6 Proton Transfer in Caronic Anhydrase orientation of the side chain of His 64 is consideraly altered in this structure of Y7H HCA II compared with wild type (Fig. 2A). One of the suggested requirements of a functional proton shuttle is flexiility of its side chain (37); comparison of these structures shows that the moility of His 64 may e altered in such a manner that a new predominant conformation has the side chain of His 64 in a position not effective for proton transfer. Another possiility is that His 7 alters the hydrogen-onded water network in the active site cavity and therey reduces proton transfer rates, in a manner found y Jackman et al. (38) for variants of human caronic anhydrase II containing replacements of Ala 65. Human Caronic Anhydrase III HCA III is unique among the class of caronic anhydrases ecause it has the lowest catalytic activity in the hydration/dehydration of CO 2 /HCO 3.It has a catalytic turnover in the hydration of CO 2 k cat, which is aout 10 ms 1 at its greatest (4), smaller y at least a factor of 10 when compared with the other caronic anhydrases in this class. One reason for this smaller catalytic turnover for caronic anhydrase III is that it lacks a proton shuttle residue in the active site. HCA III has Lys 64, which does not function as an efficient proton shuttle (4, 7). The ackone conformation of ovine (39), rat (40), and human (41) caronic anhydrase III are very similar to that of HCA II with root mean square deviations of less than 1 Å, with the similarity strong especially in the active site region (39). The catalytic activity of HCA III has een enhanced y replacements of Lys 64 with His (4) and Arg 67 with His (29); in oth cases there was convincing evidence that the newly introduced His was functioning as a proton shuttle (Tale IV). These residues are located in the active site cleft with their carons oth aout 9.5 Å from the zinc. Both side chains Lys 64 and Arg 67 extend into the active site cavity and have few interactions with other residues in the crystal structures of caronic anhydrase III. Catalysis of 18 O exchange y the doule mutant K64H/R67H HCA III is dominated y a proton donor group or groups of pk a near 7.3 and a proton acceptor group, the zinc-ound hydroxide, the conjugate acid of which has a pk a value near 5.1 (Fig. 3 and Tale IV). There is little evidence from the 18 O exchange data for a significant donor group with pk a near 5.6 in K64H/ R67H, even though such a group was oserved in the single mutant R67H (Fig. 3 and Tale IV). It is not surprising that the doule mutant has values of pk a that are altered for His 64 and His 67 ; it is not possile with these data to determine which histidine side chain is predominant in proton transfer or whether oth contriute. Our est estimates of the rate constants for intramolecular proton transfer k B from proton shuttles to the zinc-ound hydroxide (Equation 6) show a consideraly larger value for the single mutant R67H than for either K64H or K64H/R67H (Tale IV). 2 Fig. 4 shows an antagonistic interaction etween His 64 and His 67 in proton transfer. It is notale that over a considerale range of ph in Fig. 3, the rate of release from the enzyme of H 18 2 O(R H2O /[E] of Fig. 3) for the doule mutant K64H/R67H HCA III is greater than those of either of the corresponding single mutants. Two factors contriute to this result. First, the donor group in K64H/R67H hasapk a that is more asic (pk a 7.3) than the donor group in R67H (pk a 5.6). This roadens the ph profile of R H2O /[E] for K64H/R67H in comparison with that of R67H. Second, the rate constant for proton transfer k B is aout 2-fold greater for K64H/R67H than for K64H (Tale IV). Unlike the insertion of histidine at residue 7 in H64A HCA 2 This is proaly related, at least in part, to the value of pk a etween the acceptor and donor groups, which is close to 0 for R67H. Proton donation from His 67 to the zinc-ound hydroxide has less of a thermodynamic arrier than for His in K64H or K64H/R67H (Tale IV). II, the insertion of histidine in HCA III does not show convincing evidence supporting proton transfer. The most significant effect of the replacement Tyr 7 to His in HCA III appears to e a sustantial increase in the pk a of the zinc-ound water molecule from 5.3 to near 6.7 (Tale IV). There was no comparale shift in pk a with this replacement in HCA II (Tale III). This certainly is a clue to the different active site environments of caronic anhydrase II and III, although the distance from the phenolic hydroxyl of Tyr 7 to the zinc is nearly identical at 7 Å in each (9, 39). There is a small increase in k B, the rate constant for proton donation to the zinc-ound hydroxide, in Y7H HCA III compared with wild type (Tale IV), ut this may result from the increased asicity of the zinc-ound water in Y7H compared with wild type rather than to introduction of a new proton shuttle residue. It is not surprising then to find that the changes made in k B y introduction of His 7, which is proaly not a proton shuttle in HCA III, and His 64, which is a proton shuttle in mutants of HCA III (4, 7), are simply additive (Fig. 4c). That is, the replacement of Tyr 7 y His alters the pk a of the zinc-ound water y a mechanism that is separate and independent of the proton transfer steps involving His 64. We have repeated these experiments using a mutant containing the replacement Phe Leu. Leu 198 is found in the most efficient caronic anhydrases including HCA II, and the replacement of Phe 198 with Leu in HCA III enhances its catalytic activity and makes it more like caronic anhydrase II (Tale IV and Ref. 42). The mutant Y7H/F198L HCA III is 10-fold greater than wild type (3-fold greater than F198L) in the rate constant for proton transfer k B (Tale IV), and we conclude that His 7 in Y7H/F198L HCA III is functioning as a proton shuttle. (It is noted here that k cat /K m for this doule mutant is also consideraly increased, as shown in Tale IV.) It is interesting to note that when these two replacements, Tyr 7 to His and Lys 64 to His, are made in the mutant also containing Leu 198, then the doule mutant cycle (Fig. 4d) shows a partially additive interaction etween sites 7 and 64: G 1 G 1 2 G 1 G 2 (36). Conclusions In two separate cases, the insertion of two proton donor groups in the active site cavity of caronic anhydrase (His 7 and His 64 in HCA II and His 64 and His 67 in HCA III) results in an antagonistic interaction in intramolecular proton transfer with respect to the corresponding single mutants. The crystal structure of Y7H HCA II suggests interactions that orient the proton donating side chain of His 64 away from the active site and possily restricts the moility of the side chain of His 64 and/or alters the water structure through which proton transfer proceeds. Although there is a formal antagonism in rate constants for proton transfer caused y the presence of the proton shuttles in the active site of caronic anhydrase, there is a kinetic advantage in specific ranges of ph for the mutant of HCA III containing two histidines. Moreover, these studies show that residue 7 is a site from which efficient proton transfer can occur; His 7 in the doule mutant Y7H/ H64A HCA II transfers protons at a rate of 120 ms 1. Hence, these studies show no significant advantage in catalysis and mostly a suppression of proton transfer when two histidine residues are located within the active site cavity of caronic anhydrase. It should e pointed out that the cases of two histidine residues we investigated are not positioned for a sequential proton transfer through a long channel ut are located at sites roughly equidistant from their proton acceptor, the zinc-ound hydroxide. REFERENCES 1. Christianson, D. W., and Fierke, C. A. (1996) Acc. Chem. Res. 29, Lindskog, S. (1997) Pharmacol. Ther. 74, Steiner, H., Jonsson, B.-H., and Lindskog, S. (1975) Eur. J. Biochem. 59,

7 38876 Proton Transfer in Caronic Anhydrase 4. Jewell, D. A., Tu, C. K., Paranawithana, S. R., Tanhauser, S. M., LoGrasso, P. V., Laipis, P. J., and Silverman, D. N. (1991) Biochemistry 30, Tu, C. K., Silverman, D. N., Forsman, C., Jonsson, B. H., and Lindskog, S. (1989) Biochemistry, 28, Tu, C. K., Sanyal, G., Wynns, G. C., and Silverman, D. N. (1983) J. Biol. Chem. 258, Silverman, D. N., Tu., C. K., Chen, X., Tanhauser, S. M., Kresge, A. J., and Laipis, P. J. (1993) Biochemisty 32, Nair, S. K., and Christianson, D. W. (1991) J. Am. Chem. Soc. 113, Eriksson, A. E., Jones, T. A., and Liljas, A. (1988) Proteins Struct. Funct. Genet. 4, Duda, D., Tu, C. K., Qian, M. Z., Laipis, P. J., Agandje-McKenna, M., Silverman, D. N., McKenna, R. (2001) Biochemistry 40, Lu, D., and Voth, G. A. (1998) J. Am. Chem. Soc. 120, Kresge, A. J., and Silverman, D. N. (1999) Methods Enzymol. 308, Mitchell, D. M., Fetter, J. R., Mills, D. A., Adelroth, P., Pressler, M. A., Kim, Y., Aasa, R., Brzezinski, P., Malmstrom, B. G., Alen, J. O., Bacock, G. T., Ferguson-Miller, S., and Gennis, R. B. (1996) Biochemistry 35, Warren, M. S., Marolewski, A. E., and Benkovic, S. J. (1996) Biochemistry 35, Luecke, H., Schoert, B., Richter, H.-T., Cartailler, J.-P., and Lanyi, J. K. (1999) Science 286, Ädelroth, P., Paddock, M. L., Tehrani, A., Beatty, J. T., Feher, G., and Okamura, M. Y. (2001) Biochemistry 40, Tanhauser, S. M., Jewell, D. A., Tu, C. K., Silverman, D. N., and Laipis, P. J. (1992) Gene (Amst.) 117, Studier, F. W., Rosenerg, A. H., Dunn, J. J., and Duendorf, J. W. (1990) Methods Enzymol. 185, Khalifah, R. G., Strader, D. J., Bryant, S. H., and Gison, S. M. (1977) Biochemistry 16, Tu, C. K., Thomas, H. G., Wynns, G. C., and Silverman, D. N. (1986) J. Biol. Chem. 261, Engerg, P., Millqvist, E., Pohl, G, and Lindskog, S. (1985) Arch. Biochem. Biophys. 241, Silverman, D. N. (1982) Methods Enzymol. 87, Simonsson, I., Jonsson, B.-H., and Lindskog, S. (1979) Eur. J. Biochem. 93, McPherson, A. (1982) Preparation and Analysis of Protein Crystals, John Wiley & Sons, Inc., New York 25. Otwinowski, Z., and Minor, W. (2000) Denzo and Scalepack, International Tales for Crystallography, Volume F, pp Brünger, A. T., Adams, P. D., Clore, G. M., DeLano, W. L., Gross, P., Grosse- Kunstleve, R. W., Jiang, J. S., Kuszewski, J., Nilges, M., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T., Warren, G. L. (1998) Acta Crystallogr. Sect. D Biol. Crystallogr. 54, Jones T. A., Zou J. Y., Cowan S. W., Kjeldgaard M Acta Crystallogr. Sect. A. 47, Khalifah, R. G. (1971) J. Biol. Chem. 246, Ren, X., Tu, C. K., Laipis, P. J., and Silverman, D. N. (1995) Biochemistry 34, Qian, M., Tu, C. K., Earnhardt, J. N., Laipis, P. J., and Silverman, D. N. (1997) Biochemistry 36, Chen, X., Tu, C. K., LoGrasso, P. V., Laipis, P. J., and Silverman, D. N. (1993) Biochemistry, 32, LoGrasso, P. V., Tu, C. K., Chen, X., Taoka, S., Laipis, P. J., and Silverman, D. N. (1993) Biochemistry, 32, Liang, Z., Xue, Y., Jonsson, B.-H., and Lindskog, S. (1993) Eur. J. Biochem. 211, Heck, R. W., Tanhauser, S. M., Manda, R., Tu, C. K., Laipis, P. J., and Silverman, D. N. (1994) J. Biol. Chem. 269, Carter, P. J., Winter, G., Wilkinson, A. J., and Fersht, A. R. (1984) Cell 38, Mildvan, A. S., Weer, D. J., and Kuliopulos, A. (1992) Arch. Biochem. Biophys. 294, Jude, K. M., Wright, K., Tu, C. K., Silverman, D. N., Viola, R. E., and Christianson, D. W. (2002) Biochemistry 41, Jackman, J. E., Merz, K. M., and Fierke, C. A. (1996) Biochemistry 35, Eriksson, A. E., and Liljas, A. (1993) Proteins Struct. Funct. Genet. 16, Mallis, R. J., Poland, B. W., Chatterjee, T. K., Fisher, R. A., Darmawan, S., Honzatko, R. B., and Thomas, J. (2000) FEBS Lett. 482, Duda, D. M., Yoshioka, C., Govindasamy, L., An, H., Tu, C. K., Silverman, D. N., and McKenna, R. (2002) Acta Crystallogr. Sect. D Biol. Crystallogr. 58, LoGrasso, P. V., Tu, C. K., Jewell, D. A., Wynns, G. C., Laipis, P. J., and Silverman, D. N. (1991) Biochemistry 30, Tu, C. K., Chen, X., LoGrasso, P. V., Jewell, D. A., Laipis, P. J., and Silverman, D. N. (1994) J. Biol. Chem. 269, Bacon, D., and Anderson, W. F. (1988) J. Mol. Graphics 6, Merritt, E. A., and Murphy, M. E. P. (1995) Acta Crystallogr. Sect. D Biol. Crystallogr. 50,