Dramatic Broadening of the Substrate Profile of the Aeromonas hydrophila CphA Metallo-β-lactamase by Site-directed Mutagenesis
2005; Elsevier BV; Volume: 280; Issue: 31 Linguagem: Inglês
10.1074/jbc.m414052200
ISSN1083-351X
AutoresCarine Bebrone, Christine Anne, Kris De Vriendt, Bart Devreese, Gian María Rossolini, Jozef Van Beeumen, J.-M. Frère, Moreno Galleni,
Tópico(s)Bacterial biofilms and quorum sensing
ResumoAmong class B β-lactamases, the subclass B2 CphA enzyme is characterized by a unique specificity profile. CphA efficiently hydrolyzes only carbapenems. In this work, we generated site-directed mutants that possess a strongly broadened activity spectrum when compared with the WT CphA. Strikingly, the N116H/N220G double mutant exhibits a substrate profile close to that observed for the broad spectrum subclass B1 enzymes. The double mutant is significantly activated by the binding of a second zinc ion under conditions where the WT enzyme is non-competitively inhibited by the same ion. Among class B β-lactamases, the subclass B2 CphA enzyme is characterized by a unique specificity profile. CphA efficiently hydrolyzes only carbapenems. In this work, we generated site-directed mutants that possess a strongly broadened activity spectrum when compared with the WT CphA. Strikingly, the N116H/N220G double mutant exhibits a substrate profile close to that observed for the broad spectrum subclass B1 enzymes. The double mutant is significantly activated by the binding of a second zinc ion under conditions where the WT enzyme is non-competitively inhibited by the same ion. In vivo, class B β-lactamases (1.Ambler R.P. Philos. Trans. R. Soc. Lond. B Biol. Sci. 1980; 289: 321-331Crossref PubMed Scopus (1346) Google Scholar) require one or two zinc ions as enzymatic cofactors. By efficiently catalyzing the hydrolysis of the β-lactam amide bond, these enzymes play a key role in bacterial resistance to this group of antibiotics. The metallo-β-lactamase family has been divided into three different subclasses, B1, B2, and B3, on the basis of sequence similarities (2.Galleni M. Lamotte-Brasseur J. Rossolini G.M. Spencer J. Dideberg O. Frère J.M. the Metallo-β-lactamases Working Group Antimicrob. Agents Chemother. 2001; 45: 660-663Crossref PubMed Scopus (323) Google Scholar, 3.Garau G. Garcia-Saez I. Bebrone C. Anne C. Mercuri P.S. Galleni M. Frère J.M. Dideberg O. Antimicrob. Agents Chemother. 2004; 48: 2347-2349Crossref PubMed Scopus (235) Google Scholar). The CphA metallo-β-lactamase produced by Aeromonas hydrophila belongs to subclass B2. It is characterized by a uniquely narrow specificity profile. CphA efficiently hydrolyzes only carbapenems and shows very poor activity against penicillins and cephalosporins, a behavior in contrast to that of metallo-β-lactamases of subclasses B1 and B3, which usually exhibit very broad activity spectra against nearly all β-lactam compounds, with the exception of monobactams (4.Segatore B. Massida O. Satta G. Setacci D. Amicosante G. Antimicrob. Agents Chemother. 1993; 37: 1324-1328Crossref PubMed Scopus (70) Google Scholar, 5.Felici A. Amicosante G. Oratore A. Strom R. Ledent P. Joris B. Fanuel L. Frère J.M. Biochem. J. 1993; 291: 151-155Crossref PubMed Scopus (174) Google Scholar). Moreover, in contrast to the BcII (Bacillus cereus) and CcrA (Bacteroides fragilis) enzymes belonging to the B1 subclass, and in general to most other metallo-β-lactamases, CphA exhibits a maximum activity as a mono-zinc enzyme. The presence of a Zn2+ ion in a second low affinity binding site non-competitively inhibits the enzyme with a Ki value of 46 μm at pH 6.5 (6.Hernandez-Valladares M. Felici A. Weber G. Adolph H.W. Zeppezauer M. Rossolini G.M. Amicosante G. Frère J.M. Galleni M. Biochemistry. 1997; 36: 11534-11541Crossref PubMed Scopus (174) Google Scholar). Recently, the structure of the mono-zinc CphA enzyme has been solved by x-ray crystallography (7.Garau G. Bebrone C. Anne C. Galleni M. Frère J.M. Dideberg O. J. Mol. Biol. 2005; 345: 785-795Crossref PubMed Scopus (181) Google Scholar). Similar to the known structures of metallo-β-lactamases of subclasses B1 (BcII (8.Carfi A. Duée E. Galleni M. Frère J.M. Dideberg O. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 313-323Crossref PubMed Scopus (0) Google Scholar), CcrA (9.Concha N.O. Rasmussen B.A. Bush K. Herzberg O. Structure. 1996; 4: 823-836Abstract Full Text Full Text PDF PubMed Scopus (343) Google Scholar), IMP-I (10.Concha N.O. Janson C.A. Rowling P. Pearson S. Cheever C.A. Clarke B.P. Lewis C. Galleni M. Frère J.M. Payne D.J. Bateson J.H. Abdel-Meguid S.S. Biochemistry. 2000; 39: 4288-4298Crossref PubMed Scopus (277) Google Scholar), BlaB (11.Garcia-Saez I. Hopkins J. Papamicael C. Franceschini N. Amicosante G. Rossoloni G.M. Galleni M. Frère J.M. Dideberg O. J. Biol. Chem. 2003; 278: 23868-23873Abstract Full Text Full Text PDF PubMed Scopus (122) Google Scholar)) and B3 (L1 (12.Ullah J.H. Walsh T.R. Taylor I.A. Emery D.C. Verma C.S. Gamblin S.J. Spencer J. J. Mol. Biol., 28. 1998; 4: 125-136Crossref Scopus (292) Google Scholar) and FEZ-1 (13.Garcia-Saez I. Mercuri P.S. Papamicael C. Kahn R. Frère J.M. Galleni M. Rossolini G.M. Dideberg O. J. Mol. Biol. 2003; 325: 651-660Crossref PubMed Scopus (115) Google Scholar)), the x-ray structure of CphA highlights an αββα sandwich with two central β-sheets and α-helices on the external faces. The active site is located at the bottom of the β-sheet core. In agreement with previous spectroscopic results (14.Hernandez-Valladares M. Kiefer M. Heinz U. Paul Soto R. Meyer-Klaucke W. Friederich Nolting H. Zeppezauer M. Galleni M. Frère J.M. Rossolini G.M. Amicosante G. Adolph H.W. FEBS Lett. 2000; 467: 221-225Crossref PubMed Scopus (50) Google Scholar, 15.Heinz U. Bauer R. Wommer S. Meyer-Klaucke W. Papamichaels C. Bateson J. Adolph H.W. J Biol. Chem. 2003; 278: 20659-20666Abstract Full Text Full Text PDF PubMed Scopus (73) Google Scholar) and site-directed mutagenesis studies (16.Vanhove M. Zakhem M. Devreese B. Franceschini N. Anne C. Bebrone C. Amicosante G. Rossolini G.M. Van Beeumen J. Frère J.M. Galleni M. Cell. Mol. Life Sci. 2003; 60: 2501-2509Crossref PubMed Scopus (31) Google Scholar), these structural data show that the sole Zn2+ ion resides in the Asp120-Cys221-His263 site of the A. hydrophila metallo-β-lactamase. In the di-zinc form of subclass B1, the zinc ions occupy both the His116, His118, and His196 and the Asp120, Cys221, and His263 sites (see Fig. 1). The histidine residue in position 116 in most metallo-β-lactamases is replaced by an asparagine residue in CphA (2.Galleni M. Lamotte-Brasseur J. Rossolini G.M. Spencer J. Dideberg O. Frère J.M. the Metallo-β-lactamases Working Group Antimicrob. Agents Chemother. 2001; 45: 660-663Crossref PubMed Scopus (323) Google Scholar, 17.Massida O. Rossolini G.M. Satta G. J. Bacteriol. 1991; 173: 4611-4617Crossref PubMed Google Scholar). This Asn-116 residue is not responsible for the narrow substrate profile of CphA, because the activity of the N116H mutant (where the three-histidine site found in most metallo-β-lactamases is recreated) against nitrocefin, benzyl-penicillin, and cephaloridine, although increased, remains rather low (16.Vanhove M. Zakhem M. Devreese B. Franceschini N. Anne C. Bebrone C. Amicosante G. Rossolini G.M. Van Beeumen J. Frère J.M. Galleni M. Cell. Mol. Life Sci. 2003; 60: 2501-2509Crossref PubMed Scopus (31) Google Scholar). Moreover, the KD2 values are similar for the N116S mutant and the wild-type enzyme, indicating that Asn116 does not participate in the binding of the second metal ion. Among the strictly conserved residues (Table I), Arg121 is also present in most subclass B1 enzymes, with the exception of CcrA and IMP-1 where the arginine is replaced by a cysteine and a serine, respectively. In subclass B3, position 121 is occupied by a histidine residue that becomes a ligand for the second zinc ion (12.Ullah J.H. Walsh T.R. Taylor I.A. Emery D.C. Verma C.S. Gamblin S.J. Spencer J. J. Mol. Biol., 28. 1998; 4: 125-136Crossref Scopus (292) Google Scholar, 13.Garcia-Saez I. Mercuri P.S. Papamicael C. Kahn R. Frère J.M. Galleni M. Rossolini G.M. Dideberg O. J. Mol. Biol. 2003; 325: 651-660Crossref PubMed Scopus (115) Google Scholar) (Fig. 1).Table IAmino acid residues that are strictly conserved in the known subclass B1, B2, and B3 zinc β-lactamasesSubclassPositions56737984103116117118120121123134142150caRepresents the position 150c in the amino acid sequence consensus of class B β-lactamases.183193195196197199203206217219220221232244263B1DHHDGTGHDVPLGGCGWHB2DVGGSNTHDRGATIGGAHTDVPLGNCGYHB3GLGGH/QADHADGGHTSHa Represents the position 150c in the amino acid sequence consensus of class B β-lactamases. Open table in a new tab Cysteine 221 is also present in subclasses B1 and B2. For CphA, its presence is essential to the interaction with the first zinc ion and for β-lactamase activity (3.Garau G. Garcia-Saez I. Bebrone C. Anne C. Mercuri P.S. Galleni M. Frère J.M. Dideberg O. Antimicrob. Agents Chemother. 2004; 48: 2347-2349Crossref PubMed Scopus (235) Google Scholar, 14.Hernandez-Valladares M. Kiefer M. Heinz U. Paul Soto R. Meyer-Klaucke W. Friederich Nolting H. Zeppezauer M. Galleni M. Frère J.M. Rossolini G.M. Amicosante G. Adolph H.W. FEBS Lett. 2000; 467: 221-225Crossref PubMed Scopus (50) Google Scholar, 15.Heinz U. Bauer R. Wommer S. Meyer-Klaucke W. Papamichaels C. Bateson J. Adolph H.W. J Biol. Chem. 2003; 278: 20659-20666Abstract Full Text Full Text PDF PubMed Scopus (73) Google Scholar, 16.Vanhove M. Zakhem M. Devreese B. Franceschini N. Anne C. Bebrone C. Amicosante G. Rossolini G.M. Van Beeumen J. Frère J.M. Galleni M. Cell. Mol. Life Sci. 2003; 60: 2501-2509Crossref PubMed Scopus (31) Google Scholar). In subclass B1, substitution of the cysteine side chain yields a poorly active mono-zinc enzyme but does yield a di-zinc form nearly as active as the WT 1The abbreviations used are: WT, wild type; CD, circular dichroism. 1The abbreviations used are: WT, wild type; CD, circular dichroism. β-lactamases (18.Paul-Soto R. Bauer R. Frère J.M. Galleni M. Meyer-Klaucke W. Nolting H. Rossolini G.M. de Seny D. Hernandez-Valladares M. Zeppezauer M. Adolph H.W. J. Biol. Chem. 1999; 274: 13242-13249Abstract Full Text Full Text PDF PubMed Scopus (120) Google Scholar). In subclass B3, cysteine 221 is replaced by a serine, which does not interact directly with the second zinc ion but with a water molecule located in the active site and may serve as a proton donor during the catalytic process (12.Ullah J.H. Walsh T.R. Taylor I.A. Emery D.C. Verma C.S. Gamblin S.J. Spencer J. J. Mol. Biol., 28. 1998; 4: 125-136Crossref Scopus (292) Google Scholar, 13.Garcia-Saez I. Mercuri P.S. Papamicael C. Kahn R. Frère J.M. Galleni M. Rossolini G.M. Dideberg O. J. Mol. Biol. 2003; 325: 651-660Crossref PubMed Scopus (115) Google Scholar). The analysis of Fig. 1 and Table I indicated that we could progressively, by site-directed mutagenesis, replace the metal binding site motif of subclass B2 metallo-β-lactamase (Asn116-X-His118-X-Asp120... His196... Gly219-Asn220-Cys221... His263) by those of subclass B1 (His116-X-His118-X-Asp120... His196... Gly219-Gly220-Cys221... His263) or subclass B3 (His116-X-His118-X-Asp120-His121..His196... His263), respectively. These mutations strikingly broadened the CphA activity spectrum. Some mutants efficiently catalyzed the hydrolysis of penicillins and cephalosporins in addition to carbapenems. Site-directed Mutagenesis—The Quick Change site-directed mutagenesis kit (Stratagene, La Jolla, CA) was used to introduce all the mutations. The primers designed for these experiments are listed in Table II. The N220G mutant was constructed using pET9a/CphA WT as template. The double mutant N116H/N220G was constructed using pET9a/CphA N220G as template. Triple mutants N116H/R121H/N220G and N116H/N220G/C221S were constructed from the N116H/N220G mutant. The quadruple mutant N116H/R121H/N220G/C221S was made by adding mutation C221S to the N116H/R121H/N220G triple mutant. Finally, the triple mutant N116H/R121H/C221S was obtained from the quadruple mutant by replacing Gly220 by the asparagine residue found in the wild-type enzyme.Table IIList of mutagenic primers used to generate mutantsPrimersaFor, forward primer; Rev: reverse primer.SequencesN116Hfor5′-CTGGAGGTGATCAACACCCACTACCACACCGACCG-3′N116Hrev5′-CGGTCGGTGTGGTAGTGGGTGTTGATCACCTCCAG-3′N220Gfor5′-CGAGCAGGTGCTCTATGGCGGCTGCATTCTCAAGGAG-3′N220Grev5′-CTCCTTGAGAATGCAGCCGCCATAGAGCACCTGCTCG-3′R121Hfor5′-CAACTACCACACCGACCACGCTGGCGGTAACGCC-3′R121Hrev5′-GGCGTTACCGCCAGCGTGGTCGGTGTGGTAGTTG-3′TmR121Hfor5′-CCACTACCACACCGACCACGCTGGCGGTAACGCC-3′TmR121Hrev5′-GGCGTTACCGCCAGCGTGGTCGGTGTGGTAGTGG-3′TmC221Sfor5′-GGTCCTCTATGGCGGCAGCATTCTCAAGGAGAAGC-3′TmC221Srev5′-GCTTCTCCTTGAGAATGCTGCCGCCATAGAGGACC-3′G220Nfor5′-CGAGCAGGTGCTCTATGGCAACAGCATTCTCAAGGAG-3′G220Nrev5′-CTCCTTGAGAATGCTGTTGCCATAGAGCACCTGCTCG-3′a For, forward primer; Rev: reverse primer. Open table in a new tab Protein Expression and Purification—The genes encoding wild-type CphA and the mutant proteins were cloned into pET9a using the BamHI and NdeI restriction sites. The different vectors were then introduced into the Escherichia coli strain BL21(DE3) pLysS Star (Invitrogen). Overexpression and purification of the mutant proteins were performed as described for the wild-type protein (6.Hernandez-Valladares M. Felici A. Weber G. Adolph H.W. Zeppezauer M. Rossolini G.M. Amicosante G. Frère J.M. Galleni M. Biochemistry. 1997; 36: 11534-11541Crossref PubMed Scopus (174) Google Scholar, 19.Hernandez-Valladares M. Galleni M. Frère J.M. Felici A. Perilli M. Franceschini N. Rossolini G.M. Oratore A. Amicosante G. Microbial. Drug Resistance. 1996; 2: 253-256Crossref PubMed Scopus (21) Google Scholar) with the following modifications. Bacteria expressing the wild-type protein were grown for 8 h at 37 °C in 2XYT medium (Oxoïd, Backingstoke, UK), whereas those expressing the mutants were grown for 30 h at 18 °C in the same medium. The temperature was decreased to 18 °C because all the mutant proteins were produced mainly in a non-soluble form when the cultures were grown at 37 °C. The N116H/R121H/N220G/C221S mutant could never be produced in a soluble form even at 18 °C. The N116H/N220G/C221S and N116H/R121H/C221S mutants turned out to be very unstable and could not be purified in sufficient quantities for detailed characterization. Determination of Kinetic Parameters—The hydrolysis of the antibiotics was monitored by following the absorbance variation resulting from the opening of the β-lactam ring, using a Uvikon 860 spectrophotometer equipped with thermostatically controlled cells and connected to a microcomputer via an RS232C serial interface. The wavelength and absorbance variations were those described by Matagne et al. (20.Matagne A. Misselyn-Bauduin A.M. Joris B. Th Erpicum Granier B. Frère J.M. Biochem. J. 1990; 265: 131-146Crossref PubMed Scopus (135) Google Scholar, 21.Matagne A. Lamotte-Brasseur J. Frère J.M. Eur. J. Biochem. 1993; 217: 61-67Crossref PubMed Scopus (18) Google Scholar). Cells with 0.2- to 1-cm path lengths were used, depending on the substrate concentrations. When the Km values of the studied enzymes were sufficiently high, the kcat and Km parameters were determined either under initial-rate conditions, using the Hanes linearization of the Henri-Michaelis-Menten equation, or by analyzing the complete hydrolysis time courses (22.De Meester F. Joris B. Reckinger G. Bellefroid-Bourguignon C. Frère J.M. Waley S.G. Biochem. Pharmacol. 1987; 36: 2393-2403Crossref PubMed Scopus (129) Google Scholar). Low and very high Km values were determined as Ki values using imipenem or nitrocefin as reporter substrates. In the cases of low Km values, the kcat values were obtained from the initial hydrolysis rates measured at saturating substrate concentrations, and in the cases of high Km values, kcat was directly derived from the kcat/Km ratio. All experiments were performed at 30 °C in 15 mm sodium cacodylate buffer, pH 6.5. Imipenem was from Merck Sharp and Dohme Research Laboratories (Rahway, NJ); benzylpenicillin, oxacillin, cephaloridine, and cefotaxime were from Sigma; and nitrocefin was from Unipath Oxoid (Basingstoke, UK). Enzymatic Measurement in the Presence of Increasing Concentrations of Zinc and Determination of KD2 —Apparent kcat and Km values were measured in the presence of increasing concentrations of zinc at 30 °C in 15 mm sodium cacodylate, pH 6.5. When binding of the second zinc ion resulted in a complete loss of activity, the data were analyzed Equation 1 (6.Hernandez-Valladares M. Felici A. Weber G. Adolph H.W. Zeppezauer M. Rossolini G.M. Amicosante G. Frère J.M. Galleni M. Biochemistry. 1997; 36: 11534-11541Crossref PubMed Scopus (174) Google Scholar), kcat/kcat(0)=[kD2/([Zn]+KD2)](Eq. 1) where kcat/kcat(0) is the ratio between kcat in the presence of zinc and kcat in the absence of added zinc, and KD2 is the dissociation constant for the second zinc ion. When binding of the second zinc ion resulted in an incomplete loss of activity or in an increase of activity, Equation 2 was used (16.Vanhove M. Zakhem M. Devreese B. Franceschini N. Anne C. Bebrone C. Amicosante G. Rossolini G.M. Van Beeumen J. Frère J.M. Galleni M. Cell. Mol. Life Sci. 2003; 60: 2501-2509Crossref PubMed Scopus (31) Google Scholar), kcat/kcat(0)=[(kD2+α[Zn])/([Zn]+KD2)](Eq. 2) where α represents the ratio of kcat at saturating zinc concentration over kcat in the absence of added zinc. Experimental data were fitted to Equations 1 or 2 by non-linear regression analysis with the help of KaleidaGraph for Windows (KaleidaGraph™, version 3.5, Synergy Software). Electron Spray Ionization-Mass Spectrometry and Metal Content Determination—Samples of wild-type and mutant CphA enzymes were equilibrated in 10 mm ammonium acetate buffer (pH 6.5) by centrifugal filtration prior to electron spray ionization-mass spectrometry. For each protein mass spectra were obtained under both denaturing and native conditions. Denaturation of the protein was performed by dilution of the protein in a 50:50 (v/v) water/acetonitrile mixture containing 0.1% formic acid to final concentrations of 2–5 pmol/μl. The native protein was diluted to 10–15 pmol/μl in 10 mm ammonium acetate. All mass spectra were acquired on a Q-TOF1 mass spectrometer (Micromass, UK) equipped with a nanoelectron spray ionization source using gold/lead-coated borosilicate needles purchased from Protana (Odense, Denmark). Capillary voltage was set at 1250 V and cone voltage at 40 and 60 V for the denatured protein and native protein, respectively. Acquisition time was 3–5 min across an m/z range of 400–3000. The mass spectra were processed with MassLynx version 3.1 software of Micromass. The instrument was calibrated using a mixture of myoglobin and trypsinogen. The zinc content of each protein was derived from the mass difference between the native and denatured protein. Determination of the Zinc Content Using ICP/MS—Proteins samples were dialyzed against 15 mm sodium cacodylate, pH 6.5. Protein concentrations were then determined by measuring the absorbance at 280 nm (ϵ280 = 38,000 m–1cm–1). Zinc contents were determined by inductively coupled plasma mass spectroscopy as previously described (6.Hernandez-Valladares M. Felici A. Weber G. Adolph H.W. Zeppezauer M. Rossolini G.M. Amicosante G. Frère J.M. Galleni M. Biochemistry. 1997; 36: 11534-11541Crossref PubMed Scopus (174) Google Scholar, 16.Vanhove M. Zakhem M. Devreese B. Franceschini N. Anne C. Bebrone C. Amicosante G. Rossolini G.M. Van Beeumen J. Frère J.M. Galleni M. Cell. Mol. Life Sci. 2003; 60: 2501-2509Crossref PubMed Scopus (31) Google Scholar). Stability toward Chaotropic Agents—The stability of the different proteins was studied by fluorescence. The WT and mutant enzymes (0.05 mg/ml) were incubated for 16 h in the presence of increasing urea concentrations (0–8 m) at 4 °C. Fluorescence emission spectra of the enzymes were recorded at 20 °C with a PerkinElmer Life Sciences LS50B luminescence spectrometer using excitation and emission wavelengths of 280 and 333 nm, respectively. Circular Dichroism—CD spectra of the different enzymes (0.5 mg/ml) were obtained using a Jasco J-810 spectropolarimeter. The spectra were scanned at 20 °C with 1-nm steps from 200 to 250 nm (far UV) and from 250 to 310 nm (near UV). From a B2 to a B1 Enzyme—The N220G and N116H/N220G mutants were expressed in E. coli BL21(DE3) pLysS Star and purified to homogeneity. The mass of the different proteins was verified by electrospray mass spectrometry. Within experimental errors, the mutants were found to exhibit the expected masses (25132.92 versus 25132 Da for N220G and 25154.34 versus 25155 for N116H/N220G, Table III).Table IIIMetal binding for wild-type CphA and mutantsProteinAssay conditionsElectrospray ionization-mass spectrometryICP/MS Zn2+ contentMass of proteinCalculatedMeasuredNo. of Zn2+ ionsDaCphA WTNative25,253.5311Denatured25,18925,189.62N116HNative25,274.4711Denatured25,21225,212.27N220GNative25,196.7111Denatured25,13225,132.92N116H/N220GNative25,218.7711Denatured25,15525,154.34R121HNative25,232.5 (40%)12 (expected: 1.6)25,294.5 (60%)2Denatured25,16925,169.28N116H/R121H/N220GNative25,198.32 (60%)11.7 (expected: 1.4)25,262.37 (40%)2Denatured25,13625,135.68 Open table in a new tab The CD spectra of the mutants in the far UV indicated the same α/β ratio as for the wild-type enzyme. CD spectra in near UV and fluorescence emission spectra of the N116H/N220G mutant suggest a small conformational change in the protein tertiary structure (not shown). The stability of the N220G mutant with urea as denaturating agent is similar to that of the WT enzyme. Transitions between native and denaturated states occurred near 3 m urea. The N116H/N220G mutant is a little less stable than the WT enzyme and the N220G mutant, because the transition already occurred between 2 and 3 m urea. The enzymes were stored in 15 mm sodium cacodylate, pH 6.5, and in the presence of a free zinc concentration lower than 0.4 μm. Under these conditions, the N220G and N116H/N220G proteins contained one zinc ion per molecule as reported for the WT β-lactamase (Table III). At a free zinc concentration of 100 μm, inductively coupled plasma mass spectroscopy results show that the WT enzyme and the N116H/N220G double mutant bind 2 zinc ions per molecule, whereas the N220G mutant binds 1.7 zinc ions per molecule. The activity of the two mutants was measured in the absence of added Zn2+ ( 1>2,000500Ampicillin<0.012,500aMeasured as Ki values.<4Nitrocefin0.0081,3006Cephaloridine<0.0066,000aMeasured as Ki values. 0.0002>1002N116HImipenem1501,400110,000Benzylpenicillin0.4910440Nitrocefin0.09332,700Cephaloridine0.24950250N220GImipenem390507,800,000Benzylpenicillin0.06530aMeasured as Ki values.120Nitrocefin0.00840020Cephaloridine0.035,500aMeasured as Ki values.5.5Cefotaxime0.0091,000aMeasured as Ki values.9N116H/N220GImipenem16 (8)185 (170)86,000 (47,000)Benzylpenicillin3.3 (20)150aMeasured as Ki values. (150)22,000 (130,000)Oxacillin1048021,000Ampicillin0.88001,000Nitrocefin0.7 (2)4 (3.7)175,000 (540,000)Cephaloridine0.9 (11)145 (145)6,200 (75,000)Cefotaxime0.3 (4)50 (50)6,000 (80,000)a Measured as Ki values. Open table in a new tab The behavior of N220G CphA is not very different from that of the wild-type enzyme (Table IV). By contrast, the substrate profile of the N116H/N220G mutant is completely different. It no longer behaves as a strict carbapenemase. It hydrolyzes benzylpenicillin, oxacillin, ampicillin, nitrocefin, cephaloridine, cefuroxime, and cefotaxime, in addition to imipenem. When compared with the WT enzyme, the kcat value of the latter decreased by 75-fold, but for all the other substrates, the Km values decreased and the kcat values increased (Table IV). Moreover, the N116H/N220G mutant efficiently catalyzes the hydrolysis of meropenem, biapenem, CENTA, cephalothin, cefaclor, and ceftriaxone (Table V).Table VInitial rates of hydrolysis for the N116H/N220G and some β-lactam antibioticsAntibioticsv0μmol min-1 mg-1Imipenem6.8Biapenem1.2Meropenem4Benzylpenicillin1.6Nitrocefin0.8CENTA0.6Cephalothin1.4Ceftriaxone0.4Cefaclor0.4 Open table in a new tab As mentioned before, the mono-zinc form of the wild-type enzyme (observed in the presence of contaminating zinc concentration, i.e. ∼0.4 μm) is active, and added zinc behaves as a non-competitive inhibitor resulting in negligible activity at high Zn2+ concentrations with a Ki (KD2) value of 46 μm (6.Hernandez-Valladares M. Felici A. Weber G. Adolph H.W. Zeppezauer M. Rossolini G.M. Amicosante G. Frère J.M. Galleni M. Biochemistry. 1997; 36: 11534-11541Crossref PubMed Scopus (174) Google Scholar). With imipenem, the behavior of N220G was similar to that of the WT, but the KD2 value increased to 86 ± 6 μm (see Fig. 2B). By contrast, and despite a decrease of KD2 to 5 ± 0.9 μm, the double mutant retained 50% of activity at saturating Zn2+ concentrations (Fig. 1C). In both cases, the addition of Zn2+ did not affect the Km values (Table VI). With the other substrates, and as already observed with the N116H mutant, the activity of the N116H/N220G double mutant increased up to 14-fold in the presence of added zinc (Fig. 1, D–G). Analysis of the experimental curves using Equation 2 yielded values of 3 ± 0.24, 9 ± 1.8, 7 ± 1, and 9 ± 1 μm for the dissociation constant of the second zinc ion with nitrocefin, benzylpenicillin, cephaloridine, and cefotaxime as substrates, respectively.Table VIInfluence of zinc concentrations on kcat and Km values for N116H/N220G and imipenem[Zinc]kcatKmμms-1μmImipenem<0.416185114.4190512.51901010.4160209170508.71601008.516010008170 Open table in a new tab In these cases too, only kcat was affected, Km being unaffected (Table VII). This behavior was also observed for N220G, at least for nitrocefin (the activities of this enzyme against benzylpenicillin, cephaloridine, and cefotaxime were too low to allow meaningful measurements), and KD2 obtained from this curve using Equation 2 was 110 ± 13 μm (Fig. 2B).Table VIIInfluence of zinc concentrations on kcat and Km values for N116H/N220G and nitrocefin, benzylpenicillin, cephaloridine, and cefotaximeAntibioticsZinckcatKmμms-1μmNitrocefin<0.40.6942.51.36351.523.5101.793502.003.41002.063.7Benzylpenicillin<0.43.3015010020150Cephaloridine<0.40.9014510011145Cefotaxime<0.40.3050100450 Open table in a new tab From a B2 to B3 enzyme—Only the R121H and N116H/R121H/N220G mutants could be produced in E. coli and purified to homogeneity in sufficient quantities. Masses were 25169 Da for R121H and 25136 Da for N116H/R121H/N220G, in agreement with the theoretical values (Table III). The far-UV CD spectra of the mutants are similar to that of the wild-type. The near-UV CD and fluorescence emission spectra of these mutants indicated small conformational modifications in the tertiary structure. The stabilities of the mutants in the presence of increasing concentrations of urea were a little lower than that of the WT. Transitions between native and denaturated states already occurred between 2 and 3 m. The activity of these two mutants was measured with imipenem, benzylpenicillin, nitrocefin, cephaloridine, and cefotaxime as substrates in the absence of added zinc, i.e. for a free zinc concentration below 0.4 μm (Table VIII). Under these conditions, a significant proportion of a di-zinc form was already present for both mutants, respectively, 60% for R121H and 40% for N116H/R121H/N220G, in contrast to the wild-type protein and to the N220G and N116H/N220G mutants, which were 100% mono-zinc under the same conditions (Table III). With imipenem, the R121H mutant was much less active than the wild-type, a loss of activity due to a 300-fold decrease of the kcat value. The Km value of benzylpenicillin significantly increased, whereas those of nitrocefin and cephaloridine decreased. The kcat/Km values increased significantly only for nitrocefin and cephaloridine but did so for different reasons, namely a decrease of Km for the former and an increase of kcat for the latter. The triple mutant exhibited a strong decrease of activity versus imipenem (due to a decrease of kcat) and, as for the N116H/N220G mutant, a significant increase versus all other substrates due to increased kcat values and, in the cases of nitrocefin and cephaloridine, significantly decreased Km values. However, in all cases, the kcat/Km values of the triple mutant remained well below those of N116H/N220G.Table VIIIKinetic parameters of the R121H and N116H/R121H/N220G mutantsEnzymeSubstratekcatKmkcat/Kms-1μmm-1 s-1WTImipenem1,2003403,500,000Benzylpenicillin0.03870aMeasured as Ki values.35Nitrocefin0.0081 3006Cephaloridine<0.0066,000aMeasured as Ki values. 0.0002> 1002R121HImipenem3.63809,500Benzylpenicillin0.1111,000aMeasured as Ki values.10Nitrocefin0.00310300Cephaloridine0.031,800aMeasured as Ki values.17Cefotaxime0.0017245aMeasured as Ki values.7N116H/R121H/N220GImipenem1.4 (16)63 (185)22,000 (86,000)Benzylpenicillin0.35 (3.3)1,500aMeasured as Ki values. (150)233 (22,000)Nitrocefin0.5 (0.7)85 (4)6,000 (175,000)Cephaloridine0.2 (0.9)250aMeasured as Ki values. (145)800 (6,200)Cefotaxime0.1 (0.3)120aMeasured as Ki values. (50)833 (6,000)a Measured as Ki values. Open table in a new tab The activity of the R121H mutant against imipenem decreased when the zinc concentration increased (Fig. 3A). Considering that in the absence of added zinc ([Zn2+] < 0.4 μm), 60% of the R121H enzyme is already in a di-zinc form, KD2 can be estimated to be below 1.0 μm. The activity of R121H with nitrocefin as substrate is independent of the zinc concentration (Fig. 3C). Surprisingly, the activity of the N116H/R121H/N220G mutant with imipenem and nitrocefin as substrates increased (1.9- and 2.5-fold, respectively) in the presence of added zinc (Fig. 2, B and D). Considering that, in the absence of added zinc ([Zn2+] < 0.4 μm), 40% of the N116H/R121H/N220G enzyme was already in a di-zinc form, KD2 can be estimated to be below 1.0 μm. From a B2 to B1 Metallo-β-lactamase—On the basis of sequence alignments, the B2 characteristic motif of CphA has been modified to mimic that found in broad spectrum B1 enzymes. The N116H mutant has already been studied by Vanhove et al. (16.Vanhove M. Zakhem M. Devreese B. Franceschini N. Anne C. Bebrone C. Amicosante G. Rossolini G.M. Van Beeumen J. Frère J.M. Galleni M. Cell. Mol. Life Sci. 2003; 60: 2501-2509Crossref PubMed Scopus (31) Google Scholar). Here, Asn220 was replaced by a glycine residue, and the N116H/N220G double mutant was constructed. Although the kcat/Km values of the N220G mutant were somewhat increased with benzylpenicillin, nitrocefin, cephaloridine, and cefotaxime when compared with the WT protein, its activity versus imipenem also slightly increased. These results indicate that the Asn220 side chain does not have a significant role in the catalytic process and does not influence the residues involved in this process. The mutation results in a higher KD2 value for the binding of the second zinc ion (Table IX). Titration of the apoenzyme 2K. De Vriendt, unpublished observation. of N220G by Zn2+ also indicates that binding of the second zinc ion is more difficult than with the WT enzyme. The three-dimensional structure of the mono-zinc N220G mutant shows that the increased backbone mobility due to the N220G mutation alters the ability of Cys221 to coordinate the zinc. As a consequence, the zinc ion occupies two sites, the wild-type site and a new site in which the zinc ion presents a tetrahedral geometry with the triad Asp120-Cys221-His263, the fourth ligand being Arg121 (7.Garau G. Bebrone C. Anne C. Galleni M. Frère J.M. Dideberg O. J. Mol. Biol. 2005; 345: 785-795Crossref PubMed Scopus (181) Google Scholar).Table IXKD2 values for the N220G and N116H/N220G mutantsMutantSubstrateKD2μmN220GImipenem86 ± 6Nitrocefin110 ± 10N116H/N220GImipenem5 ± 0.9Nitrocefin3 ± 0.24Benzylpenicillin9 ± 1.8Cephaloridine7 ± 1Cefotaxime9 ± 1 Open table in a new tab With imipenem as substrate, the di-zinc form of N220G is completely devoid of activity as observed for the wild-type CphA. In contrast, with nitrocefin as substrate, the di-zinc form of N220G is more active than the mono-zinc form, whereas there is no influence of added zinc on the activity of the WT enzyme versus nitrocefin. The N116H/N220G double mutant was not as active as the wild-type enzyme against imipenem, but several penicillins and cephalosporins were now found to be significantly hydrolyzed. Thus, recreating the characteristic motif of the B1 subclass clearly broadens the substrate profile. For this mutant, the kcat/Km values for benzylpenicillin, cephaloridine, and cefotaxime are of the same order as, or are only 1 order of magnitude lower than the kcat/Km value for imipenem, compared with 4–6 orders for the WT. The kcat/Km value for nitrocefin is even 2-fold higher, and this increases to 11-fold in the presence of added zinc (see below). With N116H/N220G and as already observed for N116H (16.Vanhove M. Zakhem M. Devreese B. Franceschini N. Anne C. Bebrone C. Amicosante G. Rossolini G.M. Van Beeumen J. Frère J.M. Galleni M. Cell. Mol. Life Sci. 2003; 60: 2501-2509Crossref PubMed Scopus (31) Google Scholar), the di-zinc form is more active against benzylpenicillin, cephaloridine, cefotaxime, and nitrocefin than the mono-zinc form, contrary to what happens for imipenem. This suggests that imipenem on the one hand, and these latter compounds on the other, are hydrolyzed via slightly different mechanisms. However, the hydrolysis of imipenem by the N116H/N220G mutant is not totally inhibited by the binding of the second zinc ion, which underscores a significant residual activity of the di-zinc form with this substrate. For the latter, the kcat/Km values for benzylpenicillin, nitrocefin, cephaloridine, and cefotaxime are of the same order as, or are 1 order of magnitude higher than the kcat/Km value for imipenem (Table IV), and the kcat/Km values for all the tested substrates are of the same order of magnitude as, or are only 1 order lower than those determined for the subclass B1 BcII enzyme (5.Felici A. Amicosante G. Oratore A. Strom R. Ledent P. Joris B. Fanuel L. Frère J.M. Biochem. J. 1993; 291: 151-155Crossref PubMed Scopus (174) Google Scholar), which is considered to exhibit a broad activity spectrum. In the N116H/N220G mutant, the affinity for the second metal ion (characterized by KD2) increased by 5- to 10-fold suggesting that the His116 side chain can interact with the zinc ion. This hypothesis was already proposed for the N116H mutant by Vanhove et al. (16.Vanhove M. Zakhem M. Devreese B. Franceschini N. Anne C. Bebrone C. Amicosante G. Rossolini G.M. Van Beeumen J. Frère J.M. Galleni M. Cell. Mol. Life Sci. 2003; 60: 2501-2509Crossref PubMed Scopus (31) Google Scholar) but must be confirmed by NMR and perturbed angular correlation experiments and by resolution of the three-dimensional structure of the N116H and N116H/N220G mutants. Unfortunately, no suitable crystals have been obtained yet. With both mutants, the zinc concentration did not modify the Km values, but kcat decreased for imipenem or increased for the other substrates. Such a "non-competitive" inhibition or activation suggests that the substrate does not modify the affinity for the second zinc, and vice versa. Accordingly, the KD2 values measured on the basis of the activation or inhibition curves were similar within the limits of experimental errors: 86 and 110 μm for N220G, and 3–9 μm for N116H/N220G (Table VIII). From a B2 to a B3 Metallo-β-lactamase—The Arg121 residue is conserved in the BcII and CphA enzymes. Dal Peraro et al. (23.Dal Peraro M. Vila A.J. Carloni P. J. Biol. Inorg. Chem. 2002; 7: 704-712Crossref PubMed Scopus (43) Google Scholar) have shown by quantum chemistry calculations that, in the mono-zinc form of BcII, Arg121 anchors the Asp120 side chain by forming a strong ionic bond, ultimately orienting the Zn(II)-bound hydroxide for nucleophilic attack of the antibiotic β-lactam ring. Moreover, Rasia et al. (24.Rasia R.M. Ceolin M. Vila A.J. Protein Sci. 2003; 12: 1538-1546Crossref PubMed Scopus (16) Google Scholar) have shown that the R121H mutation in BcII leads to poor positioning of Asp120, and thus the kcat values of this mutant are lower than those of WT BcII. Our kinetic results are consistent with a similar role for the Arg121 residue in CphA, because the kcat value for imipenem is strongly decreased with R121H. A similar conclusion can be reached by comparing the kinetic parameters of the triple mutant to those of N116H/N220G. The kcat values are decreased as in the R121H mutant of BcII. However, the triple mutant, which is much less active than N116H/N220G, conserved a relatively broad substrate spectrum with significant hydrolysis rates of benzylpenicillin and cephalosporins. Moreover, it was the only mutant for which the binding of the second zinc increased the activity versus imipenem. Unfortunately, it was not possible to construct proteins more similar to the B3 enzymes, because the mutants with the additional C221S mutation could not be produced or were too unstable to allow detailed studies. The mutants exhibit another characteristic of B3 enzymes, namely a strongly increased affinity for the second Zn2+ ion. Indeed, the KD2 value decreased to below 1 μm. Thus, the replacement of the positive Arg121 side chain by a neutral one enhances the affinity for Zn2+, as was also suggested for BcII (8.Carfi A. Duée E. Galleni M. Frère J.M. Dideberg O. Acta Crystallogr. Sect. D Biol. Crystallogr. 1998; 54: 313-323Crossref PubMed Scopus (0) Google Scholar, 25.Fabiane S.M. Sohi M.K. Wan T. Payne D.J. Bateson J.H. Mitchell T. Sutton B.J. Biochemistry. 1998; 37: 12404-12411Crossref PubMed Scopus (214) Google Scholar) and CcrA (9.Concha N.O. Rasmussen B.A. Bush K. Herzberg O. Structure. 1996; 4: 823-836Abstract Full Text Full Text PDF PubMed Scopus (343) Google Scholar). Rasia and Vila (24.Rasia R.M. Ceolin M. Vila A.J. Protein Sci. 2003; 12: 1538-1546Crossref PubMed Scopus (16) Google Scholar) have shown that the R121H mutation increases the affinity for the second Zn2+ ion in BcII and that the H121 side chain replaces His263 as a ligand of the second Zn2+. In contrast, the R121C mutation does not increase the affinity for the second Zn2+ in BcII (26.Rasia R.M. Vila A.J. Biochemistry. 2002; 41: 1853-1860Crossref PubMed Scopus (67) Google Scholar), and the C121R mutant of CcrA was isolated in the di-zinc form, although the removal of the second zinc appeared to be facilitated (27.Fast W. Wang Z. Benkovic S.J. Biochemistry. 2001; 40: 1640-1650Crossref PubMed Scopus (73) Google Scholar). All these data and our results indicate that the replacement of Arg121 by a neutral residue in CcrA, IMP-1, and the sub-class B3 enzymes increases the affinity for the second zinc. Indeed, no mono-Zn2+ form of sub-class B3 enzymes has ever been obtained. Starting with the narrow spectrum A. hydrophila B2 β-lactamase, we have obtained, after introducing two mutations, an enzyme with a much broader activity spectrum. The ratio between the kcat/Km for different antibiotics and imipenem calculated for different class B β-lactamases and the CphA mutants underscores the complete modification of the CphA activity. As the subclass B1 and B3 enzymes, the N116H/N220G mutant is able to hydrolyze with a similar efficiency carbapenems, penicillins, and cephalosporins (Table X).Table XValues of (kcat/Km antibiotic)/(kcat/Km imipenem) for the CphA WT, N116/N220G, and N116H/R121H/N220G CphA mutants, subclass B1 (BcII, IMP-1, and VIM-2), and subclass B3 (L1 and FEZ-1) enzymesAntibioticskcat/Km (antibiotic)/kcat/Km (imipenem)CphA (Subclass B2)Subclass B1Subclass B3WTN116H/N220G [Zn] ≥ 0.4 μmN116H/N220G [Zn] = 100 μmN116H/R121H/N220GBcIIIMP-1VIM-2L1FEZ-1Benzylpenicillin<0.0010.232.50.013.750.510.270.55Cephaloridin<0.0010.071.40.040.1620.70.130.08Cefotaxime<0.0010.071.50.0460.31.5312 Open table in a new tab
Referência(s)