Catalytic Properties of Mouse Carbonic Anhydrase V*

A cDNA encoding the mouse carbonic anhydrase V gene was isolated by reverse transcription and polymerase chain reaction from BALB/c mouse liver mRNA. Vectors containing the full coding sequence as well as two different NH,-terminal truncated genes expressed enzymatically active protein in Escherichia coli. The carbonic anhydrase V produced by a vector containing the full coding sequence, which includes a possible NH,-terminal mitochondrial targeting signal, was proteolyti- cally processed by E. coli and contained several amino-terminal ends. The two NH,-terminal truncated vectors deleted, respectively, 1) the 29-amino acid putative targeting sequence and 2) 51 amino acids, yielding a pro- tein equivalent to a carbonic anhydrase (CA) V isolated from mouse liver mitochondria; and both vectors produced homogeneous protein fractions. These latter two forms of CA V had identical steady-state constants for the hydration of CO,, with maximal values of k,,JK, at 3 x lo7 M-’ s-l and It,, at 3 x 10‘ s-l with an apparent pK, for catalysis of 7.4 determined from k,,JK,. CO, CO, into controlled CO, known. Dilutions done between syringes with a gas-tight connection. Final concentrations of GO, were varied from 0.5 to 17 mM. The buffer-indicator pairs, their pK, values, and the wavelength observed were as follows: Mes (pK, = 6.1) red (pK, = 6.3), 574 nm; Mops (pK, = 7.2) with p-nitro- (pK, = 7.1), 400 nm; Hepes (pK, = 7.5) with phenol red (pK, = 7.5), 557 nm; Taps (pK, 8.4) with m-cresol purple (pK, 8.3),578 nm; and Ches (pK, = 9.3) and thymol blue (pK, = 8.91, 590 nm. initial case

catalysis of 7.4 determined from k,,JK,. In catalytic properties, mouse CA V is closest to CA I; however, in inhibition by acetazolamide, ethoxzolamide, and cyanate, CA V is very similar to CA 11. Mouse CA V has a tyrosine at position 64, where the highly active isozyme I1 has histidine serving as a proton shuttle in the catalytic pathway. Investigation of a site-specific mutant of CA V containing the replacement TyrM + His showed that the unique kinetic properties of CAV are not due to the presence of tyrosine at position 64.
Carbonic anhydrase (CA)' catalyzes the reversible hydration of CO, to form bicarbonate and a proton. A carbonic anhydrase activity associated with mitochondria was noted in early reports (Datta and Shepard, 1959;Maren and Ellison, 1967). Many subsequent studies (reviewed by Dodgson (1991)) have suggested that a mitochondrial carbonic anhydrase activity is used in the metabolic pathways of gluconeogenesis and ureagenesis, both of which require bicarbonate as substrate for enzymes compartmentalized within the mitochondria.
Although many of these studies postulated a unique mitochondrial form of carbonic anhydrase, the first evidence for a specific mitochondrial isozyme came from characterization of a *This work was supported by Grant GM25154 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 To whom correspondence should be addressed: Box 5-267 Health Center, University of Florida College of Medicine, Gainesville, FL 32610-0267. Tel.: 904-392-3556;Fax: 904-392-9696. The abbreviations used are: CA, carbonic anhydrase; HCA, human CA; PCR, polymerase chain reaction; Mes, 2-(N-morpholino)ethanesulfonic acid; Mops, 3-(N-morpholino)propanesulfonic acid; Taps, 3-[tris(hydroxymethyl)methyllaminopropanesu~fonic acid; Ches, 24cy-clohexy1amino)ethanesulfonic acid. protein purified from guinea pig liver mitochondria by Hewett-Emmett et al. (1986). They reported an amino acid sequence of 24 residues with distinct similarities to other carbonic anhydrases (this isoform was subsequently termed CA V). More recently, Amor-Gueret and Levi-Strauss (1990), while screening a mouse liver cDNA library for mRNA clones containing a mouse B2 repeat, isolated a cDNA that encoded a protein with strong sequence similarities to a carbonic anhydrase, Messenger RNA was detectable only in the liver when seven tissues were tested by Northern blotting and hybridization (Amor-Gueret and Levi-Straws, 1990). Examination of this cDNA sequence and comparison with the partial sequence of Hewett-Emmett et al. (1986) suggested that the cDNA encodes the mouse homolog of the guinea pig mitochondrial carbonic anhydrase (CA V) with the addition of a potential mitochondrial leader sequence.
We used the cDNA sequence of Amor-Gueret and Levi-Strauss (1990) and reverse transcription-PCR to isolate and clone a full-length CA V sequence from mouse liver RNA. Nagao et al. (1993) have similarly cloned a mouse CA V cDNA and used it to isolate a human CA V cDNA. We have expressed the entire mouse CAV coding sequence in a procaryotic system (Tanhauser et al., 1992). We also expressed two shorter forms of the enzyme truncated at the amino terminus, one corresponding to a protein lacking the putative mitochondrial leader sequence and the other corresponding to the amino terminus reported for guinea pig CA V (Hewett-Emmett et al., 1986). All of these expressed proteins were catalytically active in the hydration of CO,, with identical steady-state constants. CAV had appreciable CO, hydration activity; it was determined to be very similar to carbonic anhydrase I in terms of the maximal values of kc,, and k,,JK,,, and -20% as active as CA 11. CA V, however, is nearly identical to CA I1 in inhibition by acetazolamide, ethoxzolamide, and cyanate. Using a site-specific mutant, we found that the kinetic properties of CA V are influenced to only a minor extent by the tyrosine at position 64. This is in contrast to histidine 64 in CA 11, which has been shown to play a major role in proton transfer during catalysis (Tu et al., 1989).

MATERIALS AND METHODS
Cloning and Expression of Mouse CA V-The mouse CA V coding sequence was reverse-transcribed from BALB/cJ mouse liver RNA using an antisense 3'-end primer (nucleotides 1030-1049; a BamHI site was added to the 5'-end) based upon the cDNAsequence o f h o r -G u e r e t and Levi-Strauss (1990). The products of first strand synthesis were amplified using the polymerase chain reaction and a 5'-sense primer (nucleotides 105-123; an EcoRI site was added to the 5'-end). A major PCR product of 959 base pairs was obtained. It was inserted into a Bluescribe vector (Stratagene) that had been cut with SmaI and Ttailed (Marchuk et al., 1991). The cloned products were identified as mouse CA V by restriction site analysis and partial DNA sequencing. A full-length coding sequence, CA Va, as well as two deletion mutants, mouse CA Vb and mouse CA Vc, were synthesized by PCR from the initial PCR-derived clones using the original 3'-oligonucleotide and -30 -2 0  1. NH,-terminal amino acid sequence of mouse CAV. The numbering is based upon the sequence for CA 11. The sequence for CA Va is the deduced amino acid sequence from the nucleotide sequence of Amor-Gueret and Levi-Strauss (1990). CA Vb is a truncated protein with the Vc is a truncated protein with the same approximate length as carbonic anhydrase V isolated from mouse liver mitochondria. Sequences are from putative mitochondrial leader removed; it incidentally has the same approximate length as the nonmitochondrial isozymes CA I, 11, and 111. CA Fraser and Curtis (1986) for mouse CAI, Ventaet al. (1985) for mouse CA 11, and Tweedie and Edwards (1989) for mouse CAIII. Conserved residues are shown in boldface type. The daggers designate the most abundant amino-terminal residues found in several preparations of CAVa; and T h r g were the major species. The asterisks designate 3 amino acids similar to a motif found in mitochondrial signal sequence cleavage sites described by Hendrick et al. (1989).
three new 5'-primers, which added a new NdeI restriction site and Met start codon at the positions shown in Fig. 1. The NdeI and BamHI restriction sites allowed the constructs to be inserted into the pET31 T7 expression vector (Tanhauser et al., 1992). These clones were transformed into Escherichia coli BL21 (DE31 pLysS, a strain optimized for T7 RNA polymerase expression, and the active carbonic anhydrases were expressed. The mutant Y64H CA V was constructed using a mutating oligonucleotide (Kunkel, 1985) and verified by DNA sequencing.
Purification and NH,-terminal Sequencing-Frozen BL21 (DE3) pLysS cells expressing one of the mouse CA V constructs were thawed in a solution of 100 mM Tris, 200 mM Na,SO,, 1 m~ mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, and 2 VM leupeptin. This solution was stirred at 4 "C in the presence of 0.1 mg/ml deoxyribonuclease I for 1 h to allow the cells to lyse and to degrade the bacterial DNA. Cellular debris was removed from solution by centrifugation a t 23,000 x g for 1 h. Mouse recombinant CAV (CA Va, Vb, or Vc) was purified from the resulting supernatant by affinity chromatography usingp-aminomethylbenzenesulfonamide coupled to agarose beads (Khalifah et al., 1977). The purity of the isolated enzymes was determined by electrophoresis on a 10% polyacrylamide gel stained with Coomassie Blue (see Fig. 2). These enzymes, also separated on a 10% SDS-polyacrylamide gel, were transferred to a Pro-Blot membrane (Applied Biosystems) and amino-terminally sequenced on a n Applied Biosystems Automated Gas-phase Sequencer (Model 473A) at the Interdisciplinary Center for Biotechnology Research at this University.
Measurement of Catalysis a t Steady State-Initial velocities of CO, hydration were measured by stopped-flow spectrophotometry (Applied Photophysics Model SF.17MV) following the change in absorbance of a pH indicator at 25 "C (Khalifah, 1971). Solutions of CO, were prepared by bubbling CO, into water under controlled conditions for which the concentration of saturated CO, is known. Dilutions were done between syringes with a gas-tight connection. Final concentrations of GO, were varied from 0.5 to 17 mM. The buffer-indicator pairs, their pK, values, and the wavelength observed were as follows: Mes (pK, = 6.1) with chlorophenol red (pK, = 6.3), 574 nm; Mops (pK, = 7.2) with p-nitrophenol (pK, = 7.1), 400 nm; Hepes (pK, = 7.5) with phenol red (pK, = 7.5), 557 nm; Taps (pK, = 8.4) with m-cresol purple (pK, = 8.3),578 nm; and Ches (pK, = 9.3) and thymol blue (pK, = 8.91, 590 nm. Solutions were maintained at a constant total ionic strength of 0.2 M by addition of the appropriate amount of Na,SO,. The mean initial rate in each case was determined from at least four reaction traces comprising the initial 5-10% of the reaction. Uncatalyzed rates were subtracted, and determination of the kinetic constants kc,, and k,,JK,,, was by nonlinear least-squares methods (Enzfitter, Elsevier-Biosoft), Oxygen-18 Exchange Kinetics-The rate of exchange of "0 between CO, and water and the rate of exchange of from 12C0, to 13C0, caused by the transitory labeling of the active site of carbonic anhydrase were determined using the method described by Silverman (1982). Both of these exchanges were observed at chemical equilibrium. The reaction solution was in contact with a membrane permeable to gases; CO, passing across the membrane entered a mass spectrometer (Extrel EXM-200) providing a continuous measure of isotopic content of CO,. Experiments were performed without added buffers. Solutions were maintained at a constant total ionic strength of 0.2 M by addition of the appropriate amount of Na,SO,.
The "0 method is useful because it measures the rate of interconversion of CO, and HCO; a t chemical equilibrium, R,, as shown in Equation 1.
where k g t is a rate constant for maximal interconversion of GO, and HCO,, is an apparent substrate binding constant, and [SI is the concentration of CO, or HCO, or both (Simonsson et al., 1979). Values of kg& for the enzymes were determined by nonlinear least-squares fit of the above expression for R , to the data for varying substrate concentration or by measurement of R, a t values of [SI much smaller than K:;f. I n theory and in practice, kz;JeF is equal to k,,JKm for the hydration of CO, obtained by steady-state methods (Simonsson et al., 1979;Silverman, 1982). Values of K, were obtained from "0 exchange experiments and were determined from the least-squares fit of R , a t various inhibitor concentrations to the expression for mutual depletion of free enzyme and inhibitor (Segel, 1975). Since the total substrate concentration ([COJ + [HCO;] = 25 mM) is much less thanK:;,(>100 mM) for these experiments, we cannot differentiate between competitive and noncompetitive modes of inhibition.

RESULTS
Cloning and Expression of Mouse CA V-Reverse transcription of BALB/c liver mRNA followed by PCR amplification yielded several independent clones very similar to that reported by Amor-Gueret and Levi-Strauss (1990); the BALB/c sequence is incomplete, but contains several silent nucleotide changes from the B1O.HTT sequence. Sequences coding for three forms of mouse CAV were inserted into the T7 expression vector pET31 (Tanhauser et al., 1992); their amino termini are shown in Fig. 1. CAVa is a full-length cDNA sequence including the 29 amino acids at the NH, terminus, which could represent a mitochondrial targeting sequence. This amino-terminal sequence has no counterpart in CAI, 11, and 111. CAVb and Vc are shorter isoforms of CA V; in addition, we synthesized a mutant of CA Vc, replacing with His (Y64H CA Vc). Levels of CA production from these vectors ranged from 1 to 4 mg/liter of culture. Electrophoresis of the purified enzymes on SDS-polyacrylamide gels showed that mouse CA Vb and Vc and Y64H CA Vc were >95% pure and had the expected molecular masses of 31 and 28 kDa (Fig. 2). The expressed enzymes CA Va, Vb, and Vc were amino-terminally sequenced. Analysis of several preparations of CA Va indicated that it was a mixed fraction with 18-51 residues of the expected NH,-terminal sequence removed by E. coli proteases, even in the presence of several protease inhibitors during lysis and purification. CA Vb had the expected amino-terminal residues; CA Vc had the initial methionine removed (Fig. 1).
The concentrations of CA Vb and Vc were determined by titration of the catalyzed "0 exchange activity with the tightbinding inhibitor ethoxzolamide. The molar absorptivities of CAVb and Vc were determined to be 5.4 x lo4 and 4.8 x lo4 cm", respectively, at 280 nm; the concentrations ofY64H CAVc were determined from the latter value.
Measurement of Catalytic Activity-The catalytic constant kc&, for CO, hydration catalyzed by CAVc (the shortest form; see Fig. 1) was determined by stopped-flow spectrophotometry and by "0 exchange measured by mass spectrometry. Over the range of pH 5-9, k,,JKm determined from '*O exchange could be described as dependent on the unprotonated state of a single ionizable group with pK, = 7.4 2 0.1 and a maximum of (3.5 * 0.1) x 10' M-' s-I (Fig. 3). The results obtained by stopped-flow spectrophotometry were pK, = 7.8 2 0.1 and k,,JK,,, = (3.2 2 0.3) x lo7 s-I. The values of kc,JKn, and apparent pK, obtained by stopped-flow spectrophotometry and "0 exchange for the longer form, CAVb (see Fig. 11, were identical to those obtained for CA Vc. The longest form, CA Va, was subject to proteolysis, and we were not able to isolate sufficient quantities in pure form for kinetic analysis. However, the catalytic activity of the mixture was not appreciably different from that of the two purified shorter forms, suggesting that the amino-terminal segments in the longer forms do not affect catalysis.
The values of kc,, for the hydration of CO, for the two forms CA Vb and Vc showed a monotonic increase with pH. In each case, measurement of a maximal value of kc,, was hampered by the instability of the enzyme above pH 9. For CA Vb, there was only a hint of reaching a maximum, with values of kc,, near 3.0 x lo5 s-' at pH 9.5. Extended work with CAVc gave the clearest indication of the maximal value of kc,, a t (3.2 0.3) x lo5 s" and an apparent pK, of 9.2 2 0.2 (Fig. 4). The steady-state constants for CAV and comparisons with isozymes I, 11, and I11 are given in Table I. Comparison of the inhibition of these enzymes by the sulfonamides acetazolamide and ethoxzolamide, as well as by cyanate, are given in Table 11. The results from the catalysis of the hydrolysis of 4-nitrophenyl acetate by CAVc at 25 "C gave kc,JKn, = 151 2 10 M" s-' and an apparent pK, of 6.7 2 0.1. This activity is very low compared with the esterase activity of CA I1 under comparable conditions (kcaJKn, as great as lo3 M-* s-') (Simonsson and Lindskog, 1982;Pocker and Stone, 1967).
Activity of Y64H C A Vc-The values of kc,JKm for the hydration of CO, catalyzed byY64H CAVc were very similar to those for CA Vc (Fig. 3), with maximal values of kc,JK,, and apparent pK, determined by "0 exchange given in the legend to Fig. 3. Values of kc,JKm determined by stopped-flow spectrophotometry were lower by as much as %fold compared with those in Fig. 3 in the range of pH 7-8.0; this decrease may be due to inhibition by buffer. The values of kc,, for Y64H CA Vc were identical to those for CAVc at pH >8; at a lower range of pH, the values of kc,, for Y64H CA Vc were enhanced compared with those for CA Vc (Fig. 4). DISCUSSION We have cloned, expressed, and purified three types of mouse carbonic anhydrase V (forms Va, Vb, and Vc; Fig. 1) that represent possible isoforms of this enzyme in the cytoplasm and mitochondria. CAVa is a potentially full-length protein product based on the complete cDNA sequence (Amor-Gueret and Levi-Strauss, 1990). Electrophoretic analysis and amino-terminal sequencing (Figs. 1 and 2) show that this protein is a mixture of a t least four species, with loss of between 18 and 51 amino  4. kc,, catalytic constants for hydration of CO, catalyzed by carbonic anhydrase Vc (0) and mutant YMH CA Vc (A). Stopped-flow measurements a t 25 "C used solutions containing the following buffers at 25 mM: Mes, pH 5.8-6.9; Mops, pH 6.6-7.3; Hepes, pH 7.1-7.9; Taps, pH 8.1-8.7; and Ches, pH ~8 . 9 .
The solid line (for CAVc) is a least-squares fit to a single ionization with pK, = 9.2 2 0.2 and a maximal value ofh,,, = (3.2 2 0.3) x IO5 s". The dotted line (forY64H CA Vc) describes two ionizations with pK, = 9.3 and 6.2 and maxima of 3.8 x 1 0 and 6 x IO3 s-I, respectively. " These values are for the two forms of CA V designated b and c in Fig. 1. acids. The initial 29 amino acid residues of mouse CAVa shown in Fig. 1 are presumably part of a mitochondrial targeting sequence and would play no role in catalysis (see below). The amino-terminal region has the structural properties of a mito- inhibition of isozymes of carbonic anhydrase measured by "0 exchange at p H 7.2-7.5 and 25 "C The inhibition data were obtained using a least-squares fit to the expression for mutual depletion of inhibitor and free enzyme (Segel, 1975)  These data for HCA I are taken from Sanyal et al. (1982) and were cyanate. These values are for CA Vc, the shortest form of CA V (Fig. 1).
chondrial targeting sequence (Hartl et al., 1989). It is rich in positively charged residues and deficient in acidic residues and could possibly form an amphipathic helix. The shorter form of carbonic anhydrase (CA Vb) (Fig. 1) corresponds to the protein produced by cleavage at a consensus mitochondrial processing site with an arginine at position -2 (Hartl et al., 1989). Nagao et al. (1993) have reported that a recombinant human CA V expressed in COS cells was cleaved at an analogous site.
The mouse CAV sequence also contains 3 residues (shown by asterisks in Fig. 1) similar to a conserved amino acid motif found in mitochondrial leader peptides that undergo a two-step cleavage during import (Hendrick et al., 1989); this predicts a cleavage that corresponds to CA Vc in Fig. 1. This motif in mouse CA V has a proline where the consensus sequence of Hendrick et al. (1989) shows a hydrophobic residue, and the glutamates at positions 5 and 6 are inconsistent with many mitochondrial leaders. Cleavages a t positions analogous to CA Vc were reported in purified guinea pig, rat, and human CA V proteins (Hewett-Emmett et al., 1986;Ohliger et al., 1993;Nagao et al., 1993). The observation that many cleavages can occur within the first 50 amino acids of CA V in our bacterial system indicates that this entire amino-terminal region may be very susceptible t o proteolysis. We have purified CA V from mouse liver mitochondria and found a predominant fraction of protein with an amino terminus identical to that of CA Vc in Fig. 1. ' Nagao et al. (1993) reported that a full-length human CA V cDNA expressed in COS cells produced a weakly active carbonic anhydrase. A portion of this protein was present in a mitochondrial fraction in processed form. Processed forms were also present in cytoplasm fractions. Nagao et al. isolated the processed form from total COS cell homogenates; it had an amino terminus that began at a position homologous to position 3 in Fig. 1; a still shorter form was extracted in the absence of protease inhibitors.
We have found that the two shorter forms of CA V (forms Vb and Vc; Fig. 1) have the same catalytic properties in the hydration of CO,. This is consistent with the report that a deletion mutant of HCA I1 lacking 16 residues at the NH, terminus is nearly fully a~t i v e .~ This segment of CA I1 (as well as of CA I and 111) contains Tyr7, the side chain of which extends into the active-site cavity and is a component of the hydrogen-bonded array of side chains and water molecules found in the activesite cavity of the crystal structure (Eriksson et al., 1988) respectively. It seems clear that the absence of residue 7 and the NH,-terminal segment in which it is found do not contribute significantly to the catalytic pathway. This is further emphasized by the result from kinetic analysis that mixtures of the longer forms of CA V, obtained from proteolytic products of CAVa such as shown in Fig. 2, had catalytic activity equivalent to that of CA Vb and Vc.
The maximal values of the kinetic constant k,,JK, for the hydration of CO, for CA Vb and Vc were 3 x lo7 M -~ s-'. The pH profile of k,,JKm appeared to be dependent on the basic form of a single ionizable group with apparent pK, near 7.4 (Fig. 3). This pK, has been shown to represent the ionization of the zinc-bound water (Simonsson and Lindskog, 19821, which in isozyme I1 is close to 7, although it is dependent on experimental conditions and the ionization state of the nearby His64 (Simonsson and Lindskog, 1982). Thus, isozyme V has an apparent pK, for this zinc-bound water close to that of isozyme 11. A comparison of maximal values of k,,JKm for the hydration of CO, catalyzed by several isozymes is shown in Table I. A significant difference in catalysis by isozymes I and I1 compared with isozyme V was found in kc,, for hydration. For CAI and 11, kc,, is described by a titration curve with a pK, near 7 reaching a maximum at high pH (Khalifah, 1971;Steiner et al., 1975). In CA 11, this behavior is a reflection of the pK, of histidine 64, the proton acceptor in the intramolecular proton shuttle that transfers a proton to solution and regenerates the zinc-bound hydroxide (Steiner et al., 1975;Silverman and Lindskog, 1988). In mouse CA Vb and Vc, we observed an increase in kcat for hydration with an increase in pH and an apparent pK, of 9.2 (Fig. 4). However, the maximal values of kc,, for the hydration of CO, by mouse CA Vb and Vc are near 3 x lo5 s-l, a very sizable value. Together, these data suggest the presence of a proton acceptor in the active-site cavity with a pK, near 9. There could be more than one proton acceptor, of course, but there does not appear to be a predominant proton acceptor in the initial 51 residues of CAVa ( Fig. 1) since the deletion of this segment leaves kcat relatively unchanged.
One prominent difference in the active-site cavities of CA V and I1 is at position 64. The histidine at this position in human CA I1 has been shown to act as a proton shuttle, accepting protons from the zinc-bound water and transferring them to solution ( T u et al., 1989). The presence of His64 in CA I1 is responsible for the apparent pK, near 7 in kc, for the hydration of CO, ( T u et al., 1989;Steiner et al., 1975). CAV has a tyrosine at position 64 that is a poor proton shuttle at pH 7, but is a more efficient shuttle at pH near the pK, of the phenolic hydroxyl. To assess the effects of these residues a t position 64, we replaced T y r 6 4 with His in CA Vc (Y64H CA Vc).
The mutant Y64H CA Vc shows very similar k,,JK, values compared with CA Vc (Fig. 3). This constant, kcaJKm, contains rate constants for the steps of the interconversion of CO, and HCO, (Equation 2). CO, + EZnOH-3 EZnHCO; e EZnH,O + HCO; (Eq. 2) Thus, the replacement T y r 6 4 + His has only small effects on the conversion of CO, to HCO, at the zinc; one noticeable difference is the small deviation from a single ionization of k,,JK, for Y64H CA Vc at pH <7, which may be due to ionization of His64 in this pH range.
The steady-state constant kcat in catalysis by carbonic anhydrase is determined by the transfer of protons from the zincbound water to buffers in solution, as in Equation 3 (Silverman and Lindskog, 1988 Here, X is a residue of the enzyme capable of accepting a proton

Mouse Carbonic Anhydrase
V from the zinc-bound water and transferring it to buffer B in solution. The most prominent acceptor in Y64H CA Vc has a pK, near 9.2; in fact, kcat values for this mutant and CA Vc are identical at pH >8 (Fig. 4). Thus, it can be concluded that this proton acceptor is not T y r 6 4 , but another residue near the zinc.
At pH <7, the values of kc,, for this mutant exceed those for CA Vc by as much as 5-fold and show a second pK, near 6.2 with a maximum near lo4 s-' (Fig. 4). This suggests a possible proton acceptor role of His64 in Y64H CA Vc; His64 is a proton acceptor in HCA I1 (Tu et al., 1989) and in K64H HCA I11 (Jewell et al., 1991). However, in HCA 11, the presence of His64 supports proton transfer to give a maximal value of kcat near lo6 s-'; in Y64H CA Vc, this maximal value appears to be closer to lo4 s-'. Thus, the replacement Tyr'j4 + His in CA Vc does not result in keat values similar to those for CA 11, and the presence of TyP4 is not a major factor in the unique kinetic properties of CA V.
A comparison of the maximal values of the steady-state kinetic constants for the hydration of CO, appears in Table I. The maximal steady-state parameters for CA V resemble most closely those for CA I. It is useful to note that previous reports of human CA V as a low activity enzyme (Nagao et al., 1993; their value is not a specific activity) and of a turnover number of 24,000 s-' for rat CAV (Dodgson and Cherian, 1989) refer to pH near 7.4; at more alkaline pH, the activity is appreciable and not a low activity carbonic anhydrase. This may be significant since intramitochondrial pH is expected to be alkaline. The inhibition of CA Vc by two prominent sulfonamides and cyanate is very similar to that of CA I1 (Table 11). Thus, it will not be possible to differentiate between CA I1 and V in physiological function using these sulfonamides.