Revisão Acesso aberto Revisado por pares

Carbonic Anhydrase: New Insights for an Ancient Enzyme

2001; Elsevier BV; Volume: 276; Issue: 52 Linguagem: Inglês

10.1074/jbc.r100045200

ISSN

1083-351X

Autores

Brian C. Tripp, Kerry S. Smith, James G. Ferry,

Tópico(s)

Chemical Reactions and Mechanisms

Resumo

proton shuttle residue 2- (N-morpholino) propanesulfonic acid Carbonic anhydrase catalyzes the reversible hydration of CO2 (Equation 1).CO2+H2O⇄HCO3−+H+Equation 1 The first carbonic anhydrase was purified from erythrocytes in 1933 (1Meldrum N.U. Roughton F.J. J. Physiol. (Lond.). 1933; 80: 113-142Crossref Scopus (333) Google Scholar) followed by the characterization of several mammalian isozymes that dominated research on carbonic anhydrase until recently. Although it has been known since the 1940s that carbonic anhydrase is ubiquitous in plants (2Bradfield J.R.G. Nature. 1947; 159: 467-468Crossref PubMed Scopus (20) Google Scholar), where it is essential for CO2 fixation, relatively few studies had been reported. Until 1994, only five carbonic anhydrases had been purified from prokaryotes; however, a recent survey has established that the enzyme is widely distributed among phylogenetically and physiologically diverse prokaryotes, indicating a far greater role for this enzyme in nature than previously recognized (3Smith K.S. Jakubzick C. Whittam T.S. Ferry J.G. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 15184-15189Crossref PubMed Scopus (335) Google Scholar). The comparison of sequences and crystal structures of the mammalian and plant enzymes demonstrates that they evolved independently and have been designated the α- and β-class, respectively. An additional independently evolved γ-class was reported in 1994 (4Alber B.E. Ferry J.G. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6909-6913Crossref PubMed Scopus (203) Google Scholar) for which phylogenetic analyses predict an ancient origin (3Smith K.S. Jakubzick C. Whittam T.S. Ferry J.G. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 15184-15189Crossref PubMed Scopus (335) Google Scholar). This review discusses dramatic advances over the past 3 years regarding the structure and biological chemistry of carbonic anhydrases. The three classes have no significant sequence identity, and the overall folds (Fig. 1) underscore their independent origins (5Liljas A. Laurberg M. EMBO J. 2000; 1: 16-17Crossref Scopus (49) Google Scholar). Despite gross structural differences, the active sites of all three classes function with a single zinc atom essential for catalysis (6Lindskog S. Pharmacol. Ther. 1997; 74: 1-20Crossref PubMed Scopus (774) Google Scholar, 7Christianson D.W. Cox J.D. Annu. Rev. Biochem. 1999; 68: 33-57Crossref PubMed Scopus (328) Google Scholar). Kinetic studies indicate that all three classes employ a two-step isomechanism (8Northrop D.B. Simpson F.B. Arch. Biochem. Biophys. 1998; 353: 288-292Crossref Scopus (23) Google Scholar). The first step is the nucleophilic attack of a zinc-bound hydroxide ion on CO2(Equation 2). The second step is the regeneration of the active site by ionization of the zinc-bound water molecule and removal of a proton from the active site (Equation 3). In this step, the zinc ion acts as a Lewis acid to lower the pKa of the water molecule from ∼14 to 7.0.Zn2+­OH−+CO2⇄Zn2++HCO3−Equation 2 Zn2++H2O⇄H++Zn2+­OH−Equation 3 Most carbonic anhydrases have kcat values greater than 104 s−1, which requires an intermediate PSR1 (Equation4) to transfer the proton from the metal-bound water molecule to the external buffer, “B” (Equation 5).PSR+Zn2+­H2O⇄Zn2+­OH−+PSR­H+Equation 4 PSR­H++B⇄PSR+B­H+Equation 5 Proton transport from the active site is the rate-limiting step for enzymes with kcat > 104s−1. Thus, kcat is a reflection of the rate of proton transport (Equation 3), whereas the catalytic efficiency (kcat/Km) is more reflective of the hydration step (Equation 2) and is insensitive to the rate of proton transport. The following sections focus on recent results revealing specific properties of the three classes of carbonic anhydrase, which provide new structural and biochemical perspectives for this enzyme. The α-class is the best characterized with 11 isozymes identified in mammals. Several isozymes are implicated in various disease states for which treatment frequently involves the application of sulfonamides that inhibit carbonic anhydrase activity. Earnhardt et al. (9Earnhardt J.N. Qian M. Tu C. Lakkis M.M. Bergenhem N.C. Laipis P.J. Tashian R.E. Silverman D.N. Biochemistry. 1998; 37: 10837-10845Crossref PubMed Scopus (43) Google Scholar) summarize sulfonamide inhibition constants and maximal kcat andkcat/Km values for CO2 hydration by isozymes I–VII. Prokaryotic α-class enzymes are few compared with the other two classes. The recent characterization of mammalian and prokaryotic α-class enzymes has been reported and reviewed (10Smith K.S. Ferry J.G. FEMS Microbiol. Rev. 2000; 24: 335-366Crossref PubMed Google Scholar, 11Chirica L.C. Elleby B. Lindskog S. Biochim. Biophys. Acta. 2001; 1544: 55-63Crossref PubMed Scopus (55) Google Scholar, 12Hewett-Emmett D. Tashian R.E. Mol. Phylogenet. Evol. 1996; 5: 50-77Crossref PubMed Scopus (526) Google Scholar, 13Huang S. Xue Y. Sauer-Eriksson E. Chirica L. Lindskog S. Jonsson B.H. J. Mol. Biol. 1998; 283: 301-310Crossref PubMed Scopus (87) Google Scholar, 14Chirica L.C. Elleby B. Jonsson B.H. Lindskog S. Eur. J. Biochem. 1997; 244: 755-760Crossref PubMed Scopus (84) Google Scholar, 15Puskas L.G. Inui M. Zahn K. Yukawa H. Microbiology. 2000; 146: 2957-2966Crossref PubMed Scopus (26) Google Scholar). The α-class is by far the best studied with respect to the mechanism of catalysis. The reader is referred to excellent reviews of the literature prior to 1999 (6Lindskog S. Pharmacol. Ther. 1997; 74: 1-20Crossref PubMed Scopus (774) Google Scholar, 7Christianson D.W. Cox J.D. Annu. Rev. Biochem. 1999; 68: 33-57Crossref PubMed Scopus (328) Google Scholar). Recent advancements have focused on the rate-limiting proton transfer step. In several isozymes of the α-class, His-64 accepts a proton from active site water molecules that intervene between the zinc-bound water molecule and His-64. The His-64 PSR can be replaced with other residues that function as PSRs, a result consistent with the proposal that proton transfer occurs through different structures of intervening water chains (16Tu C. Qian M. Earnhardt J.N. Laipis P.J. Silverman D.N. Biophys. J. 1998; 74: 3182-3189Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar). Indeed, molecular dynamics simulations indicate that the number of intervening waters can vary from two to six (17Toba S. Colombo G. Merz K.M. J. Am. Chem. Soc. 1999; 121: 2290-2302Crossref Scopus (115) Google Scholar). Crystal structures of human CA-II show His-64 in either an “in” (toward zinc) or “out” position. This fluctuation of His-64 is postulated to facilitate proton transfer between active site waters and solvent water at the mouth of the active site cavity. Imidazole and imidazole derivatives mimic the PSR function of His-64 and rescue the H64A variant of CA-II that is 10-fold reduced inkcat. Crystal structures of the variant complexed with 4-methylimidazole show the rescue agent occupying the “out” position leading to the conclusion that this orientation of His-64 is important for proton transfer (18Duda D. Tu C. Qian M. Laipis P. Agbandje-McKenna M. Silverman D.N. McKenna R. Biochemistry. 2001; 40: 1741-1748Crossref PubMed Scopus (93) Google Scholar). On the other hand, aqueous phase molecular dynamics simulations of the wild-type enzyme in three protonation states indicate that His-64 primarily assumes the “in” orientation, a result leading the authors to suggest that fluctuations between the two orientations of this residue may have limited influence on proton transfer (17Toba S. Colombo G. Merz K.M. J. Am. Chem. Soc. 1999; 121: 2290-2302Crossref Scopus (115) Google Scholar). The rate of 18O exchange between the zinc-bound water molecule and solvent waters is used to determine the rate of intramolecular proton transfer. Fitting Marcus rate theory to the rate data (19Silverman D.N. Biochim. Biophys. Acta. 2000; 1458: 88-103Crossref PubMed Scopus (94) Google Scholar) requires a substantial adjustment in large work terms or thermodynamic components suggesting that intramolecular proton transfer involves a reorganization of the active site cavity. It is proposed that the reorganization includes waters not directly involved along the pathway; for example, movement of His-64 from the “out” to the “in” orientation involves breaking H-bonds between the side chain and water. Finally, ab initio studies of intramolecular proton transfer indicate that the donor-acceptor distance and the water chain motion are essential to the energetics (20Lu D.S. Voth G.A. J. Am. Chem. Soc. 1998; 120: 4006-4014Crossref Scopus (153) Google Scholar). Intermolecular proton transfer may involve more than a single PSR (21Earnhardt J.N. Qian M. Tu C. Laipis P.J. Silverman D.N. Biochemistry. 1998; 37: 7649-7655Crossref PubMed Scopus (13) Google Scholar). Site-specific replacement of Lys-91 and Tyr-131 near the mouth of the active site cavity of isozyme CA-VA produced variant enzymes compromised in kcat but notkcat/Km, indicating that these basic residues are PSRs. Moreover, kinetic analysis of a double variant suggested a cooperative behavior between the residues in facilitating proton transfer. The incorporation of a histidine analog by chemical modification of the Y131C variant resulted in enhanced proton transfer, a result that further supports the proposed PSR role for Tyr-131 (22Earnhardt J.N. Wright S.K. Qian M. Tu C. Laipis P.J. Viola R.E. Silverman D.N. Arch. Biochem. Biophys. 1999; 361: 264-270Crossref PubMed Scopus (18) Google Scholar). The α-class carbonic anhydrases are characterized by subpicomolar affinities for zinc, which have provided a system for investigating the fundamental properties of metal ion binding in metalloproteins. Recent studies have focused on structural features of the active site and the thermodynamics of solute association that influence metal binding specificity and avidity. The results indicate a role for hydrophobic core residues in human CA-II that are important for preorienting the histidine ligands in a geometry that favors zinc binding and destabilizes geometries that favor other metals (23Cox J.D. Hunt J.A. Compher K.M. Fierke C.A. Christianson D.W. Biochemistry. 2000; 39: 13687-13694Crossref PubMed Scopus (45) Google Scholar, 24Hunt J.A. Ahmed M. Fierke C.A. Biochemistry. 1999; 38: 9054-9062Crossref PubMed Scopus (103) Google Scholar). Calorimetric studies of CA-II and variants indicate that both desolvation of the metal ion and the binding site have major contributions to the overall thermodynamics, thus directing specificity of binding by optimizing desolvation (25DiTusa C.A. Christensen T. McCall K.A. Fierke C.A. Toone E.J. Biochemistry. 2001; 40: 5338-5344Crossref PubMed Scopus (75) Google Scholar, 26DiTusa C.A. McCall K.A. Christensen T. Mahapatro M. Fierke C.A. Toone E.J. Biochemistry. 2001; 40: 5345-5351Crossref PubMed Scopus (40) Google Scholar). The understanding of the β-class has lagged far behind that for the α-class; indeed, the first crystal structure for any β-class carbonic anhydrase was reported in 2000 (27Mitsuhashi S. Mizushima T. Yamashita E. Yamamoto M. Kumasaka T. Moriyama H. Ueki T. Miyachi S. Tsukihara T. J. Biol. Chem. 2000; 275: 5521-5526Abstract Full Text Full Text PDF PubMed Scopus (141) Google Scholar). Although initially thought to be composed solely of enzymes from plants, β-class carbonic anhydrases were recently isolated from a variety of algae (28Yagawa Y. Muto S. Miyachi S. Plant Cell Physiol. 1987; 28: 1253-1262Google Scholar, 29Hiltonen T. Karlsson J. Palmqvist K. Clarke A.K. Samuelsson G. Planta. 1995; 195: 345-351Crossref PubMed Scopus (28) Google Scholar, 30Eriksson M. Karlsson J. Ramazanov Z. Gardestrom P. Samuelsson G. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 12031-12034Crossref PubMed Scopus (134) Google Scholar) and found to be widely distributed in the Bacteria and Archaea domains (3Smith K.S. Jakubzick C. Whittam T.S. Ferry J.G. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 15184-15189Crossref PubMed Scopus (335) Google Scholar). The characterization of enzymes from the β-class reveals sharp differences from the other two classes. The α-class and γ-class enzymes are strictly monomers and trimers, respectively; however, members of the β-class are dimers, tetramers, hexamers, and octamers, which suggests a dimer as the basic building block (31Kimber M.S. Pai E.F. EMBO J. 2000; 19: 1407-1418Crossref PubMed Google Scholar). Furthermore, differences in secondary structure are evident from the crystal structures (Fig. 1). Finally, β-class crystal structures reveal that zinc is ligated by two conserved cysteines and one conserved histidine (27Mitsuhashi S. Mizushima T. Yamashita E. Yamamoto M. Kumasaka T. Moriyama H. Ueki T. Miyachi S. Tsukihara T. J. Biol. Chem. 2000; 275: 5521-5526Abstract Full Text Full Text PDF PubMed Scopus (141) Google Scholar, 31Kimber M.S. Pai E.F. EMBO J. 2000; 19: 1407-1418Crossref PubMed Google Scholar, 32Strop P. Smith K.S. Iverson T.M. Ferry J.G. Rees D.C. J. Biol. Chem. 2001; 276: 10299-10305Abstract Full Text Full Text PDF PubMed Scopus (112) Google Scholar). Phylogenetic analyses indicate that the β-class is more diverse than the other two classes (3Smith K.S. Jakubzick C. Whittam T.S. Ferry J.G. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 15184-15189Crossref PubMed Scopus (335) Google Scholar). Sequence alignment indicates that only 5 residues, the three zinc ligands plus an aspartate and an arginine, are completely conserved (33Smith K.S. Cosper N.J. Stalhandske C. Scott R.A. Ferry J.G. J. Bacteriol. 2000; 182: 6605-6613Crossref PubMed Scopus (43) Google Scholar). The plant sequences form two monophyletic clades representing dicotyledenous and monocotyledenous plants (3Smith K.S. Jakubzick C. Whittam T.S. Ferry J.G. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 15184-15189Crossref PubMed Scopus (335) Google Scholar). The remaining sequences are separated into five clades of which one is strongly supported by bootstrapping and appears distantly related to all other clades. This clade, represented by the enzyme “Cab” from the archaeon Methanobacterium thermoautotrophicum (32Strop P. Smith K.S. Iverson T.M. Ferry J.G. Rees D.C. J. Biol. Chem. 2001; 276: 10299-10305Abstract Full Text Full Text PDF PubMed Scopus (112) Google Scholar, 33Smith K.S. Cosper N.J. Stalhandske C. Scott R.A. Ferry J.G. J. Bacteriol. 2000; 182: 6605-6613Crossref PubMed Scopus (43) Google Scholar), is composed of sequences primarily from thermophiles in the Archaea and Gram-positive species in the Bacteria. This diversity is supported by recent crystal structures of enzymes from a red algae (Porphyridium purpureum, Fig. 1B) and pea (Pisum sativum, Fig. 1C), and prokaryotes from the Archaea (M. thermoautotrophicum, Fig.1D) and Bacteria (Escherichia coli, Fig.1E) domains (27Mitsuhashi S. Mizushima T. Yamashita E. Yamamoto M. Kumasaka T. Moriyama H. Ueki T. Miyachi S. Tsukihara T. J. Biol. Chem. 2000; 275: 5521-5526Abstract Full Text Full Text PDF PubMed Scopus (141) Google Scholar, 31Kimber M.S. Pai E.F. EMBO J. 2000; 19: 1407-1418Crossref PubMed Google Scholar, 32Strop P. Smith K.S. Iverson T.M. Ferry J.G. Rees D.C. J. Biol. Chem. 2001; 276: 10299-10305Abstract Full Text Full Text PDF PubMed Scopus (112) Google Scholar, 34Cronk J.D. Endrizzi J.A. Cronk M.R. O'Neill J.W. Zhang K.Y.J. Protein Sci. 2001; 10: 911-922Crossref PubMed Scopus (140) Google Scholar). The pea enzyme is a dimer of homodimers whereas the algal enzyme is a homodimer in which the monomer is composed of two internally repeated structures each with an active site. An overlay of the active sites of the P. sativum andM. thermoautotrophicum (Cab) enzymes shows near perfect alignments of the three zinc ligands and the β-class conserved aspartates and arginines; however, a water molecule is ligated to zinc only in Cab (32Strop P. Smith K.S. Iverson T.M. Ferry J.G. Rees D.C. J. Biol. Chem. 2001; 276: 10299-10305Abstract Full Text Full Text PDF PubMed Scopus (112) Google Scholar). An acetate molecule replaces a water molecule as the fourth zinc ligand in the P. sativum enzyme that was crystallized with acetate (31Kimber M.S. Pai E.F. EMBO J. 2000; 19: 1407-1418Crossref PubMed Google Scholar). Surprisingly, the fourth zinc ligand in both the P. purpureum and E. coli enzymes is the β-class conserved aspartate (27Mitsuhashi S. Mizushima T. Yamashita E. Yamamoto M. Kumasaka T. Moriyama H. Ueki T. Miyachi S. Tsukihara T. J. Biol. Chem. 2000; 275: 5521-5526Abstract Full Text Full Text PDF PubMed Scopus (141) Google Scholar, 34Cronk J.D. Endrizzi J.A. Cronk M.R. O'Neill J.W. Zhang K.Y.J. Protein Sci. 2001; 10: 911-922Crossref PubMed Scopus (140) Google Scholar). Residues Gln-151, Phe-179, and Tyr-205 of the P. sativum enzyme (31Kimber M.S. Pai E.F. EMBO J. 2000; 19: 1407-1418Crossref PubMed Google Scholar) are conserved among all the eukaryotic and bacterial enzymes in clades A–F; however, they are absent in the sequences of all other carbonic anhydrases that are in the same clade (clade G) as the M. thermoautotrophicumenzyme Cab. This observation led Kimber and Pai (31Kimber M.S. Pai E.F. EMBO J. 2000; 19: 1407-1418Crossref PubMed Google Scholar) to propose that the β-class is composed of two subclasses, the “plant type” (represented by the P. sativum enzyme) and the “Cab type” (represented by the M. thermoautotrophicum enzyme). Both the structural dissimilarities between the two subclasses and their varied responses to inhibitors (10Smith K.S. Ferry J.G. FEMS Microbiol. Rev. 2000; 24: 335-366Crossref PubMed Google Scholar) suggest differences in their mechanism. Kinetic analyses indicate a zinc hydroxide mechanism for the β-class (33Smith K.S. Cosper N.J. Stalhandske C. Scott R.A. Ferry J.G. J. Bacteriol. 2000; 182: 6605-6613Crossref PubMed Scopus (43) Google Scholar, 35Johansson I.M. Forsman C. Eur. J. Biochem. 1993; 218: 439-446Crossref PubMed Scopus (62) Google Scholar, 36Rowlett R.S. Chance M.R. Wirt M.D. Sidelinger D.E. Royal J.R. Woodroffe M. Wang Y.F. Saha R.P. Lam M.G. Biochemistry. 1994; 33: 13967-13976Crossref PubMed Scopus (60) Google Scholar, 37Johansson I.M. Forsman C. Eur. J. Biochem. 1994; 224: 901-907Crossref PubMed Scopus (29) Google Scholar). As is the case for the α-class, the zinc-bound acetate in the crystal structure of theP. sativum enzyme mimics the binding of bicarbonate in the active site (31Kimber M.S. Pai E.F. EMBO J. 2000; 19: 1407-1418Crossref PubMed Google Scholar). The zinc-bound oxygen of acetate hydrogen bonds with Asp-162 O-δ1, whereas the second oxygen hydrogen bonds to Gln-151 suggesting a role for these residues in catalysis. The bond between acetate and Asp-162 O-δ1 is identical to the hydrogen bond between the zinc-bound oxygen of acetate and Thr-199 O-δ1 of the α-class CA-II isozyme. Thr-199 O-δ1 functions to orient the zinc-bound hydroxide for nucleophilic attack on CO2. Superimposition of active sites also shows that the bond between Gln-151 and acetate in the P. sativum enzyme active site overlaps the hydrogen bond between the Thr-199 N and the second oxygen of bicarbonate. Thr-199 N is proposed to electrophilically activate CO2 by forming a hydrogen bond with CO2 (6Lindskog S. Pharmacol. Ther. 1997; 74: 1-20Crossref PubMed Scopus (774) Google Scholar). Therefore, Gln-151 and Asp-162 are thought to play the same roles as Thr-199 of the α-class enzymes (31Kimber M.S. Pai E.F. EMBO J. 2000; 19: 1407-1418Crossref PubMed Google Scholar). Asp-34 in the active site of Cab could function similarly to Asp-162 of the plant-type subclass (32Strop P. Smith K.S. Iverson T.M. Ferry J.G. Rees D.C. J. Biol. Chem. 2001; 276: 10299-10305Abstract Full Text Full Text PDF PubMed Scopus (112) Google Scholar). The route of proton transfer is not clear from the structures of theP. sativum and P. purpureum enzymes (27Mitsuhashi S. Mizushima T. Yamashita E. Yamamoto M. Kumasaka T. Moriyama H. Ueki T. Miyachi S. Tsukihara T. J. Biol. Chem. 2000; 275: 5521-5526Abstract Full Text Full Text PDF PubMed Scopus (141) Google Scholar, 31Kimber M.S. Pai E.F. EMBO J. 2000; 19: 1407-1418Crossref PubMed Google Scholar); however, in the structure of Cab, a HEPES buffer molecule is found ∼8 Å from the zinc (32Strop P. Smith K.S. Iverson T.M. Ferry J.G. Rees D.C. J. Biol. Chem. 2001; 276: 10299-10305Abstract Full Text Full Text PDF PubMed Scopus (112) Google Scholar) and within hydrogen bonding distance of the β-class conserved aspartate (Asp-34), which also forms a hydrogen bond with the zinc-bound water molecule. Thus, one possible pathway for proton transfer is from the zinc-bound water molecule to Asp-34 and then to the sulfate group of HEPES. Indeed, replacement of Asp-34 results in a 10-fold decrease in the kcat of Cab, and the D34A variant is chemically rescued by replacement of MOPS buffer with imidazole at pH 7.2. 2K. S. Smith, C. Ingram-Smith, and J. G. Ferry, submitted for publication. These results are consistent with a PSR role for the β-class conserved aspartate in at least the Cab-type subclass. The structures of the P. purpureum and E. colienzymes suggest an additional role for the β-class conserved aspartate, which is a fourth ligand to zinc (27Mitsuhashi S. Mizushima T. Yamashita E. Yamamoto M. Kumasaka T. Moriyama H. Ueki T. Miyachi S. Tsukihara T. J. Biol. Chem. 2000; 275: 5521-5526Abstract Full Text Full Text PDF PubMed Scopus (141) Google Scholar, 34Cronk J.D. Endrizzi J.A. Cronk M.R. O'Neill J.W. Zhang K.Y.J. Protein Sci. 2001; 10: 911-922Crossref PubMed Scopus (140) Google Scholar). The presence of a water molecule hydrogen-bonded to the conserved aspartates (Asp-151 and Asp-405) in the duplicated active sites of the P. purpureumenzyme leads the authors to propose a modified zinc hydroxide mechanism (27Mitsuhashi S. Mizushima T. Yamashita E. Yamamoto M. Kumasaka T. Moriyama H. Ueki T. Miyachi S. Tsukihara T. J. Biol. Chem. 2000; 275: 5521-5526Abstract Full Text Full Text PDF PubMed Scopus (141) Google Scholar) in which the aspartate functions as a base to abstract a proton from the bound water molecule yielding a nucleophilic hydroxide. The hydroxide moves toward and then binds to the zinc. When bound, the hydroxide attacks CO2 to generate the zinc-bound bicarbonate. The deprotonated aspartate binds zinc displacing bicarbonate. Finally, to regenerate the active site, a water molecule binds to the aspartate in each of the duplicated active sites. On the other hand, kinetic analysis of variants generated by replacement of the β-class conserved aspartate (Asp-34) in Cab2 shows that this residue is not essential for the CO2 hydration step of catalysis. The P. purpureum and E. coli enzymes are only active above neutral pH values, which prompted a second hypothetical model accounting for ligation of the β-class conserved aspartate to zinc. In this model, the aspartate ligand is exchanged with a water molecule above neutral pH values, thereby activating the enzyme (34Cronk J.D. Endrizzi J.A. Cronk M.R. O'Neill J.W. Zhang K.Y.J. Protein Sci. 2001; 10: 911-922Crossref PubMed Scopus (140) Google Scholar). A reorganization of residues in the active site coupled to the ligand exchange cannot be ruled out as an additional mechanism for activation; furthermore, it cannot be ruled out that a reorganization of the active site coupled to ligand exchange occurs repeatedly during a single catalytic turnover. Repeated ligand exchange during turnover would potentially allow the conserved aspartate to play a role in proton transfer as established for Cab. Because both the P. purpureum and E. coli carbonic anhydrases belong to the same phylogenetic clade, other enzymes from this clade may be expected to have a similar active site architecture and mechanism. The γ-class is thought to have evolved between 3.0 and 4.5 billion years ago (3Smith K.S. Jakubzick C. Whittam T.S. Ferry J.G. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 15184-15189Crossref PubMed Scopus (335) Google Scholar) and therefore precedes evolution of the α-class at 200–300 million years ago (12Hewett-Emmett D. Tashian R.E. Mol. Phylogenet. Evol. 1996; 5: 50-77Crossref PubMed Scopus (526) Google Scholar, 38Jiang W. Gupta D. Biochem. J. 1999; 344: 385-390Crossref PubMed Scopus (17) Google Scholar). The only γ-class enzyme characterized is “Cam” from the archaeonMethanosarcina thermophila (4Alber B.E. Ferry J.G. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6909-6913Crossref PubMed Scopus (203) Google Scholar). Cam is a homotrimer that adopts a left-handed parallel β-helical fold (Fig. 1F) (39Kisker C. Schindelin H. Alber B.E. Ferry J.G. Rees D.C. EMBO J. 1996; 15: 2323-2330Crossref PubMed Scopus (213) Google Scholar). Cam is heterologously produced in E. coli at high levels to yield a zinc enzyme (40Alber B.E. Ferry J.G. J. Bacteriol. 1996; 178: 3270-3274Crossref PubMed Google Scholar); however, iron- and cobalt-substituted forms exhibit greater CO2 hydration rates than the zinc enzyme (41Alber B.E. Colangelo C.M. Dong J. Stalhandske C.M. Baird T.T. Tu C. Fierke C.A. Silverman D.N. Scott R.A. Ferry J.G. Biochemistry. 1999; 38: 13119-13128Crossref PubMed Scopus (80) Google Scholar) 3B. C. Tripp and J. G. Ferry, unpublished results. ; thus, it is possible that Cam functions in M. thermophila using a different transition metal than zinc. Cam employs a metal hydroxide mechanism in catalysis with proton transport as the rate-limiting step (41Alber B.E. Colangelo C.M. Dong J. Stalhandske C.M. Baird T.T. Tu C. Fierke C.A. Silverman D.N. Scott R.A. Ferry J.G. Biochemistry. 1999; 38: 13119-13128Crossref PubMed Scopus (80) Google Scholar). Unlike many of the α-class enzymes, Cam does not exhibit esterase activity with p-nitrophenyl acetate as the substrate, and the inhibition by sulfonamides is low compared with the α-class (40Alber B.E. Ferry J.G. J. Bacteriol. 1996; 178: 3270-3274Crossref PubMed Google Scholar). The metal binding site consists of three histidine residues in a tetrahedral geometry similar to that of the monomeric α-class (39Kisker C. Schindelin H. Alber B.E. Ferry J.G. Rees D.C. EMBO J. 1996; 15: 2323-2330Crossref PubMed Scopus (213) Google Scholar); however, in Cam, two of the histidines are donated by one monomer (His-81, His-122) and the other from an adjacent monomer (His-117). High resolution crystal structures with bicarbonate bound to the active site have led to proposed roles for other active site residues (42Iverson T.M. Alber B.E. Kisker C. Ferry J.G. Rees D.C. Biochemistry. 2000; 39: 9222-9231Crossref PubMed Scopus (147) Google Scholar), which have been further investigated by site-directed mutational analysis. Solvent-accessible Gln-75 is located with the side chain 5 Å from the zinc and is structurally modeled with the carbonyl oxygen pointed toward the zinc and an amino group hydrogen-bonded to the carbonyl oxygen of the Asn-73 side chain (39Kisker C. Schindelin H. Alber B.E. Ferry J.G. Rees D.C. EMBO J. 1996; 15: 2323-2330Crossref PubMed Scopus (213) Google Scholar, 42Iverson T.M. Alber B.E. Kisker C. Ferry J.G. Rees D.C. Biochemistry. 2000; 39: 9222-9231Crossref PubMed Scopus (147) Google Scholar). The amino group of Asn-73 is in turn hydrogen-bonded to the side chain hydroxyl group of Ser-114. This hydrogen bond network indicates that the Gln-75 side chain is highly oriented with the carbonyl oxygen forming a hydrogen bond to one of two water molecules coordinated by the zinc (42Iverson T.M. Alber B.E. Kisker C. Ferry J.G. Rees D.C. Biochemistry. 2000; 39: 9222-9231Crossref PubMed Scopus (147) Google Scholar). Kinetic analyses of the Q75A variant indicate that Gln-75 is important for CO2 hydration activity. 4C. Brosius and J. G. Ferry, unpublished results. Thus, Gln-75 may function in analogy with Thr-199 in the α-class CA-II isozyme by hydrogen bonding with and orienting the zinc-bound hydroxide for attack on CO2. The carboxylate of Glu-62 resides 5 Å from zinc and has bicarbonate bound in the crystal structure suggesting a potential role in catalysis, although the binding of bicarbonate could be an off pathway event (42Iverson T.M. Alber B.E. Kisker C. Ferry J.G. Rees D.C. Biochemistry. 2000; 39: 9222-9231Crossref PubMed Scopus (147) Google Scholar). The role for Glu-62 was further investigated by kinetic analyses of variants in which Glu-62 was replaced with several different residues (43Tripp B.C. Ferry J.G. Biochemistry. 2000; 39: 9232-9240Crossref PubMed Scopus (63) Google Scholar). Only the E62D variant maintains wild-type activity whereas several other variants have lowkcat andkcat/Km values suggesting that the negative charge of Glu-62 is important for the CO2hydration step in catalysis, although the specific function is unknown. Glu-84 is adjacent to Glu-62 and assumes two different conformations (42Iverson T.M. Alber B.E. Kisker C. Ferry J.G. Rees D.C. Biochemistry. 2000; 39: 9222-9231Crossref PubMed Scopus (147) Google Scholar) in analogy with the PSR His-64 in the α-class CA-II isozyme. The replacement of Glu-84 in Cam yields variants with large decreases inkcat relative to wild type but only small changes in kcat/Km (43Tripp B.C. Ferry J.G. Biochemistry. 2000; 39: 9232-9240Crossref PubMed Scopus (63) Google Scholar). The same variants are rescued up to 46-fold in kcatwhen assayed in the presence of imidazole, results strongly indicating that Glu-84 functions as a PSR. Interestingly, bicarbonate can function as a proton donor in the dehydration direction of catalysis by Cam and the α-class human CA-II isozyme; however, it is not known if bicarbonate is essential for the proton transfer step in these enzymes (44Tu C. Tripp B.C. Ferry J.G. Silverman D.N. J. Am. Chem. Soc. 2001; 123: 5861-5866Crossref PubMed Scopus (42) Google Scholar). The guanido group of Arg-59 in Cam is located 6 Å from the zinc where it also partners in a salt bridge between Asp-61 and Asp-76 in adjacent monomers (39Kisker C. Schindelin H. Alber B.E. Ferry J.G. Rees D.C. EMBO J. 1996; 15: 2323-2330Crossref PubMed Scopus (213) Google Scholar). The Arg-59 is important for the association of monomers into the native trimer and is essential for the CO2hydration step in catalysis. 5Tripp, B. C., Tu, C., and Ferry, J. G., (2002)Biochemistry, in press. The specific catalytic function for Arg-59 is unknown; however, it is postulated that this residue may influence the pKaof the catalytic zinc-bound water molecule or bind bicarbonate as part of the product release cycle. In 1997, Francois Morel and co-workers (45Roberts S.B. Lane T.W. Morel F.M.M. J. Phycol. 1997; 33: 845-850Crossref Scopus (126) Google Scholar) reported the purification of a 27-kDa monomeric carbonic anhydrase, TWCA1, from the marine diatom Thalassiosira weissflogii (45Roberts S.B. Lane T.W. Morel F.M.M. J. Phycol. 1997; 33: 845-850Crossref Scopus (126) Google Scholar). The catalytic zinc was shown by x-ray absorption spectroscopy to be coordinated by three histidines and a water molecule (46Cox E.H. McLendon G.L. Morel F.M. Lane T.W. Prince R.C. Pickering I.J. George G.N. Biochemistry. 2000; 39: 12128-12130Crossref PubMed Scopus (111) Google Scholar), similar to the active sites of the α-class and γ-class carbonic anhydrases. Additionally, the near-edge spectra argue that the active site geometry is similar to that of α-class enzymes (46Cox E.H. McLendon G.L. Morel F.M. Lane T.W. Prince R.C. Pickering I.J. George G.N. Biochemistry. 2000; 39: 12128-12130Crossref PubMed Scopus (111) Google Scholar). Although no steady-state kinetics have been reported, the existence of a water molecule as a fourth ligand suggests that this enzyme may also follow the zinc hydroxide mechanism of the other three classes. Although TWCA1 has biochemical properties similar to the three known classes of carbonic anhydrase, the deduced sequence of the gene encoding TWCA1 revealed no significant identity to the three classes. In addition, our searches of the sequence data bases failed to identify open reading frames in the Archaea, Bacteria, or Eukarya domains with deduced sequence identity. Thus, TWCA1 is the prototype for a fourth class of carbonic anhydrase that we propose here to be designated the δ-class. During conditions in which the levels of TWCA1 are low in T. weissflogii, a 43-kDa cadmium-specific carbonic anhydrase is expressed (47Lane T.W. Morel F.M. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 4627-4631Crossref PubMed Scopus (559) Google Scholar). The sequence of the gene encoding this carbonic anhydrase has not yet been reported; thus, whether it represents a new class or belongs to a pre-existing class is not yet known. Undoubtedly, dramatic advances in both the physiology and biochemistry of carbonic anhydrases have been described in the past few years. The catalytic mechanisms for both the β- and γ-class have been further elucidated especially in the proton transfer pathway. It is expected that the recently solved structures of four β-class enzymes will result in significant progress in understanding the mechanism(s) of this class in the near future. Additional enzymes from both the β- and γ-class have been purified and characterized, broadening our knowledge of each class. The report of a fourth class represented by the T. weissfloggi enzyme suggests we may be only scratching the surface of carbonic anhydrase diversity.

Referência(s)