Portal de Periódicos da CAPES

Overexpression and Mechanistic Analysis of Chromosomally Encoded Aminoglycoside 2′-N-Acetyltransferase (AAC(2′)-Ic) fromMycobacterium tuberculosis

2001; Elsevier BV; Volume: 276; Issue: 49 Linguagem: Inglês

10.1074/jbc.m108810200

1083-351X

Subray S. Hegde, Farah Javid-Majd, John S. Blanchard,

Antibiotic Resistance in Bacteria

The chromosomally encoded aminoglycosideN-acetyltransferase, AAC(2′)-Ic, of Mycobacterium tuberculosis has a yet unidentified physiological function. Theaac(2′)-Ic gene was cloned and expressed inEscherichia coli, and AAC(2′)-Ic was purified. Recombinant AAC(2′)-Ic was a soluble protein of 20,000 Da and acetylated all aminoglycosides substrates tested in vitro, including therapeutically important antibiotics. Acetyl-CoA was the preferred acyl donor. The enzyme, in addition to acetylating aminoglycosides containing 2′-amino substituents, also acetylated kanamycin A and amikacin that contain a 2′-hydroxyl substituent, although with lower activity, indicating the capacity of the enzyme to perform bothN-acetyl and O-acetyl transfer. The enzyme exhibited "substrate activation" with many aminoglycoside substrates while exhibiting Michaelis-Menten kinetics with others. Kinetic studies supported a random kinetic mechanism for AAC(2′)-Ic. Comparison of the kinetic parameters of different aminoglycosides suggested that their hexopyranosyl residues and, to a lesser extent, the central aminocyclitol residue carry the major determinants of substrate affinity. The chromosomally encoded aminoglycosideN-acetyltransferase, AAC(2′)-Ic, of Mycobacterium tuberculosis has a yet unidentified physiological function. Theaac(2′)-Ic gene was cloned and expressed inEscherichia coli, and AAC(2′)-Ic was purified. Recombinant AAC(2′)-Ic was a soluble protein of 20,000 Da and acetylated all aminoglycosides substrates tested in vitro, including therapeutically important antibiotics. Acetyl-CoA was the preferred acyl donor. The enzyme, in addition to acetylating aminoglycosides containing 2′-amino substituents, also acetylated kanamycin A and amikacin that contain a 2′-hydroxyl substituent, although with lower activity, indicating the capacity of the enzyme to perform bothN-acetyl and O-acetyl transfer. The enzyme exhibited "substrate activation" with many aminoglycoside substrates while exhibiting Michaelis-Menten kinetics with others. Kinetic studies supported a random kinetic mechanism for AAC(2′)-Ic. Comparison of the kinetic parameters of different aminoglycosides suggested that their hexopyranosyl residues and, to a lesser extent, the central aminocyclitol residue carry the major determinants of substrate affinity. aminoglycosideN-acetyltransferase polymerase chain reaction Aminoglycosides are natural or semisynthetic, broad spectrum bactericidal compounds consisting of a central six-membered amino cyclitol ring linked to two or more deoxy-amino sugars by glycosidic bonds. Aminoglycosides were among the first antibiotics discovered and are clinically used in the treatment of severe Gram-positive and Gram-negative bacterial infections. Their mechanism of action is the inhibition of bacterial protein synthesis by binding to the small ribosomal subunit, and additionally, they are also capable of directly disrupting the integrity of the outer membrane of Gram-negative bacteria by displacing divalent cations bridging adjacent lipopolysaccharide molecules (1Davis B.D. Microbiol. Rev. 1987; 51: 341-350Crossref PubMed Google Scholar, 2Rather P.N. Drug Resist. Updates. 1998; 1: 285-291Crossref PubMed Scopus (19) Google Scholar, 3Mingeot-Leclercq M.-P. Glupczynski Y. Tulkens P.M. Antimicrob. Agents Chemother. 1999; 43: 727-737Crossref PubMed Google Scholar, 4Hancock R.E.W. J. Antimicrob. Chemother. 1981; 8: 249-276Crossref PubMed Scopus (141) Google Scholar). The emergence of growing numbers of resistant strains has somewhat reduced the utility of aminoglycosides in empiric therapies (5Murray B.E. J. Infect. Dis. 1991; 163: 1185-1194Crossref Scopus (142) Google Scholar). Resistance to aminoglycosides can arise by three distinct mechanisms: (i) alteration of ribosomal RNA and small ribosomal subunit protein binding sites, (ii) decreased uptake and/or accumulation of the drug in bacteria due to changes in the permeability of the outer membrane or active efflux, and most commonly (iii) enzymatic detoxification of the drug (6Moore R.A. DeShazer D. Reckseidler S. Weissman A. Woods D.E. Antimicrob. Agents Chemother. 1999; 43: 465-470Crossref PubMed Google Scholar, 7Taber H.W. Mueller J.P. Miller P.F. Arrow A.S. Microbiol. Rev. 1987; 51: 439-457Crossref PubMed Google Scholar, 8Davies J. Wright G.D. Trends Microbiol. 1997; 5: 234-240Abstract Full Text PDF PubMed Scopus (333) Google Scholar, 9Shaw K.J. Rather P.N. Hare R.S. Miller G.H. Microbiol. Rev. 1993; 57: 138-163Crossref PubMed Google Scholar, 10Miller G.H. Sabatelli F.J. Naples L. Hare R.S. Shaw K.J. J. Chemother. 1995; 7: 31-44PubMed Google Scholar, 11Wright G.D. Curr. Opin. Microbiol. 1999; 2: 499-503Crossref PubMed Scopus (202) Google Scholar, 12Azucena E. Mobashery S. Drug Resist. Updates. 2001; 4: 106-117Crossref PubMed Scopus (101) Google Scholar). Three different types of enzymatic modification of the drug includeO-phosphorylation, O-nucleotidylation, andN-acetylation. Expression of genes encoding these enzymes is either by acquisition of foreign DNA (plasmid-borne) or by overexpression of a housekeeping gene with modifying activity (13Courvalin P. Carlier C. Collatz E. J. Bacteriol. 1980; 143: 541-551Crossref PubMed Google Scholar, 14Rather P.N. Orosz E. Shaw K.J. Hare R. Miller G. J. Bacteriol. 1993; 175: 6492-6498Crossref PubMed Scopus (67) Google Scholar, 15Rudant E. Bouvet P. Courvalin P. Lambert T. Syst. Appl. Microbiol. 1999; 22: 59-67Crossref PubMed Scopus (15) Google Scholar). Aminoglycoside N-acetyltransferases (AACs)1 use acetyl-coenzyme A (acetyl-CoA) as the acetyl group donor and are grouped into four classes: AAC(1), AAC(2′), AAC(3), and AAC(6′) based on the site of their regioselective acetylation of the aminoglycoside. Members of the AAC(6′) class are the most common in nature, and there have been several kinetic and mechanistic studies reported on them (16Radhika K. Northop D.B. Biochemistry. 1984; 23: 5118-5122Crossref PubMed Scopus (17) Google Scholar, 17Radhika K. Northop D.B. J. Biol. Chem. 1984; 258: 12543-12546Abstract Full Text PDF Google Scholar, 18Wright G.D. Ladak P. Antimicrob. Agents Chemother. 1997; 41: 956-960Crossref PubMed Google Scholar, 19Magnet S. Lambert T. Courvalin P. Blanchard J.S. Biochemistry. 2001; 40: 3700-3709Crossref PubMed Scopus (45) Google Scholar). The only other enzyme subjected to a rigorous kinetic analysis has been gentamicin acetyltransferase, AAC(3)-I (20Williams J.W. Northop D.B. J. Biol. Chem. 1978; 17: 5908-5914Abstract Full Text PDF Google Scholar, 21Williams J.W. Northop D.B. J. Biol. Chem. 1978; 17: 5902-5907Abstract Full Text PDF Google Scholar). AAC(2′) is another class of aminoglycosideN-acetyltransferases with more restricted occurrence in bacteria. All aac(2′) genes reported so far are chromosomally encoded and have been shown to be universally present in all mycobacteria (22Ainsa J.A. Perez E. Pelicic V. Berthet F. Gicquel B. Martin C. Mol. Microbiol. 1997; 24: 431-441Crossref PubMed Scopus (82) Google Scholar). The AAC(2′)-Ia from Providencia stuartii has been implicated in O-acetylation of peptidoglycan (23Payie K.G. Clarke A.J. J. Bacteriol. 1997; 179: 4106-4114Crossref PubMed Google Scholar), while the physiological role of the mycobacterial enzymes is not clearly understood. The aac(2′)-Ib fromMycobacterium fortuitum has been cloned inMycobacterium smegmatis but was neither expressed, purified, nor subjected to kinetic analysis. The P. stuartiiAAC(2′)-Ia has been expressed in Escherichia coli, and a preliminary kinetic characterization has very recently been reported (24Franklin K. Clarke A.J. Antimicrob. Agents Chemother. 2001; 45: 2238-2244Crossref PubMed Scopus (33) Google Scholar). None of the aac(2′) genes appear to be responsible for clinical resistance of aminoglycoside antibiotics, and interactions between the 2′-substituent of clinically effective aminoglycosides (-OH or -NH2) and the ribosomal site of action are not known and may differ among aminoglycoside structural classes. In the present paper we report the cloning, overexpression, purification, substrate specificity, kinetic mechanism, and solvent kinetic isotope effects of AAC(2′)-Ic from Mycobacterium tuberculosis. All chemicals, coenzyme A derivatives, and aminoglycoside antibiotics were purchased from Sigma-Aldrich Chemical Co. Enzymes used in molecular biology studies were supplied by New England Biolabs or Promega. DNA purification kits and plasmid pET-23a(+) were from Novagen. E. coli strains XL10-Gold and BL21(DE3) were obtained from Stratagene. The open reading frame of the aac(2′)-Ic gene (GenBankTMaccession no. U72714) was amplified from M. tuberculosisH37Rv genomic DNA by standard PCR techniques using the oligonucleotides AS1 (5′-ATTCCATATGCACA CCCAGGTACACACGG-3′) and AA1 (5′-GCGGAATTCTTACCAGACGTCGCCCGC-3′) containing the underlined NdeI and EcoRI restriction sites shown, respectively. The PCR fragment was cloned into pET-23a(+) and expressed in the E. coli strain BL21(DE3). For shake flask growth, 1 liter of Luria broth medium supplemented with carbenicillin (50 μg/ml) was inoculated with 10 ml of an overnight culture and incubated at 37 °C. The culture was grown to midlog phase (A 600 ∼ 1.0), cooled to 20 °C for 15 min, induced with 0.2 mm isopropyl thio-β-d- galactosidase, and further incubated for 8–10 h at 20 °C. All purification procedures were carried out at 4 °C. The cells were collected by centrifugation at 10,000 rpm and resuspended in buffer A (25 mm triethanolamine, pH 7.8, 50 mm NaCl, 1 mm EDTA, 1 mm dithiothreitol) containing protease inhibitors (Roche Molecular Biochemicals), lysozyme (0.2 μg/ml), and DNase I (1 μg/ml) and stirred for 20 min. The cells were then lysed by passage through a French press (1200 p.s.i.), and cell debris was removed by centrifugation at 16,000 rpm for 40 min. Solid ammonium sulfate was then added to the supernatant to a final concentration of 0.85 m. The mixture was stirred for 1 h and centrifuged at 20,000 rpm. The supernatant was further treated with solid ammonium sulfate to a final concentration of 1.6m. After stirring for 2 h, the pellet, separated by centrifugation at 20,000 rpm, was resuspended in buffer B (25 mm triethanolamine containing 1 mmdithiothreitol, 1 mm EDTA, 50 mm NaCl, and 0.5m (NH4)2SO4). After extensive dialysis, the precipitate was removed by centrifugation at 20,000 rpm for 10 min. The clear supernatant was then loaded onto a phenyl-Sepharose (Amersham Pharmacia Biotech) column pre-equilibrated with buffer B. The bound proteins were eluted with a linear 0.5–0m ammonium sulfate gradient at a flow rate of 1 ml/min. The active fractions were pooled, solid ammonium sulfate was added to a final concentration of 0.5 m, and the solution was rechromatographed on phenyl-Sepharose as above. The enzyme after this step was found to be >90% pure as assessed by SDS-polyacrylamide gel electrophoresis. The enzyme preparation was stored in 50% glycerol at −20 °C for long term storage and retained full activity for 8–10 months. Protein concentrations were estimated by the Bio-Rad protein assay method using bovine serum albumin as a standard. Commercial kanamycin A containing small amounts of kanamycin B (5%) was purified as follows. Kanamycin A (500 mg) in 50 mm phosphate buffer, pH 7.2, was loaded onto a CM-Sephadex column (5 × 1.5 cm) pre-equilibrated with the same buffer. The column was washed with 50 ml of equilibration buffer followed by a linear 0–0.6 m NaCl gradient. Bound antibiotic eluted as two peaks at ∼0.4 and 0.55 m NaCl (Authentic Kanamycin B sulfate eluted as a single peak at 0.6m NaCl under these conditions.). Kanamycin A-containing fractions, eluting at 0.4 m NaCl, were detected and quantitated using a phenol-sulfuric acid method, concentrated by lyophilization, desalted on a Sephadex G-10 column (1.5 × 50 cm) using Milli-Q water, and used in kinetic studies. Reaction rates were measured spectrophotometrically by following the increase in absorbance at 324 nm due to the reaction between the free sulfhydryl group of CoASH, generated by the enzyme-catalyzed aminoglycoside acylating activity, and 4,4′-dithiopyridine. The reaction was monitored continuously on a UVIKON 943 spectrophotometer, and enzyme activities were calculated using a molar absorption coefficient of 19,800m−1 cm−1. Assay mixtures contained 100 mm potassium phosphate, pH 7.0, 0.2 mm 4,4′-dithiopyridine, in addition to substrates or inhibitors in a volume of 1 ml. Reactions were initiated by the addition of enzyme and followed at 30 °C for 1–2 min. Initial velocity kinetic data were fitted using the programs of Cleland (25Cleland W.W. Methods Enzymol. 1979; 63: 103-138Crossref PubMed Scopus (1929) Google Scholar). Equation 1 was used to fit simple substrate saturation kinetics, where the concentration of one substrate was varied at a fixed, saturating (>20 × K m) concentration of the other substrate. Equation 2 was used to fit initial velocity patterns, where the concentrations of both substrates were varied. Equations 3 and 4were used to fit competitive and noncompetitive inhibition data, respectively. v=V/(Ka+A)Equation 1 v=VAB/(KaB+KbA+KiaKb+AB)Equation 2 v=VA/[Ka(1+I/Kis)+A]Equation 3 v=VA/[Ka(1+I/Kis)+A(1+I/Kii)]Equation 4 In Equations Equation 1, Equation 2, Equation 3, Equation 4, v is the measured reaction velocity, V is the maximal velocity, A andB are the concentrations of substrates (aminoglycoside and acyl-CoA), K a and K b are the corresponding Michaelis-Menten constants, Kia is the inhibition constant for substrate A, I is the concentration of the inhibitor, and Kis andKii are the slope and intercept inhibition constant for the inhibitor, respectively. Equation 5 was used to fit kinetic plots in which "substrate activation" was apparent from Lineweaver-Burk plots, where A is the concentration of the variable substrate. The constants B, C, andD are collections of rate constants that depend on the kinetic mechanism and the resulting rate equation (for model and rate equation, see Ref. 19Magnet S. Lambert T. Courvalin P. Blanchard J.S. Biochemistry. 2001; 40: 3700-3709Crossref PubMed Scopus (45) Google Scholar).v=V(A2+DA)/(A2+BA+C)Equation 5 The solvent kinetic isotope effects on V andV/K were determined by measuring the initial velocities using saturating concentrations of acyl-CoA and varying concentrations of aminoglycoside in H2O or ∼98% D2O as solvent. Solvent deuterium isotope effects were calculated from the equation:v=VA/[Ka(1+IEV/K)+A(1+IEV)Equation 6 where, I represents the fraction of isotope, andEVK and EV are isotope effects on V/K and V − 1, respectively. PCR amplification of the aac(2′)-Ic gene using primers AS1 and AA1 and theM. tuberculosis H37Rv genomic DNA template yielded a single fragment of the expected length. Cloning and overexpression of the PCR product resulted in an expressed protein product with an apparent molecular weight, by SDS-polyacrylamide gel electrophoresis, in agreement with the weight of 20,038 deduced from the nucleotide sequence. DNA sequencing of the cloned fragment confirmed the absence of any mutations introduced during PCR amplification. The two-step purification procedure yielded greater than 90% pure protein. Recombinant AAC(2′)-Ic was stable in the presence of 0.45 m ammonium sulfate and catalyzed aminoglycoside acyltransferase activity in vitro. Any attempt to remove ammonium sulfate resulted in the precipitation of the protein, and expression of the enzyme with either N- or C-terminal His6affinity tags yielded insoluble protein products. Initial velocities were determined spectrophotometrically, at pH 7.0, at 8–12 different concentrations of each substrate. The data were plotted as double-reciprocal plots and fitted using the programs of Cleland (25Cleland W.W. Methods Enzymol. 1979; 63: 103-138Crossref PubMed Scopus (1929) Google Scholar) as well as by nonlinear, least square curve fitting using Sigmaplot (version 3.0). Kinetic constants for acetyl-CoA and other CoA derivatives, determined at a saturating concentration of sisomicin, are presented in Table I. The steady-state kinetic parameters for various aminoglycosides (Scheme FS1) at saturating concentrations of acetyl-CoA and propionyl-CoA are also summarized in Table I. The structures of the aminoglycosides used are shown in Fig.1. Aminoglycoside substrates can be divided into two groups on the basis of their kinetic behavior. Seven of 13 antibiotics, using acetyl-CoA (8 of 10 using propionyl-CoA) as the second substrate, exhibited linear reciprocal plots at low substrate concentrations but rapidly increasing velocities at higher antibiotic concentrations (Fig. 2). This phenomenon, termed substrate activation, has previously been observed with other aminoglycoside acetyltransferases (19Magnet S. Lambert T. Courvalin P. Blanchard J.S. Biochemistry. 2001; 40: 3700-3709Crossref PubMed Scopus (45) Google Scholar, 21Williams J.W. Northop D.B. J. Biol. Chem. 1978; 17: 5902-5907Abstract Full Text PDF Google Scholar). For substrates exhibiting substrate activation, two pairs of fitted parameters are included in Table I. V 2 was 1.5–6 times higher than V 1, K 2 was 6–1500 times higher than K 1, andV/K 1 was 3–300 times higher thanV/K 2 at saturating concentrations of acetyl-CoA, while these same parameters were 2–8, 20–140, and 25–70 times higher, respectively, at saturating concentrations of propionyl-CoA. The enzyme did not exhibit any activity with lividomycin A, kanamycin A, and amikacin when propionyl-CoA was the acyl donor. Amikacin and kanamycin A, both of which contained 2′-hydroxyl substituents, were extremely poor, but bona fide, acetyltransferase substrates for the AAC(2′)-Ic enzyme.Table IKinetic parameters for coenzyme A derivatives and aminoglycosidesSubstrateK mk catV/KK m 1-aValues at high substrate concentrations.k cat1-aValues at high substrate concentrations.1-bSum of velocities at low and high concentrations.V/Kμmmin −110 5μmmin −110 5A Acetyl-CoA1-cMeasured at 5-cm path length.0.8 ± 0.1220 ± 102750 n-Propionyl-CoA3.0 ± 0.318 ± 0.460 Malonyl-CoA120 ± 18155 ± 813 Me-malonyl-CoA74 ± 1113 ± 0.61.8 n-Hexanoyl-CoA75 ± 63.9 ± 0.080.52 n-Butyryl-CoA188 ± 263.3 ± 0.020.18 Isovaleryl-CoA275 ± 243.8 ± 0.020.13 Gluteryl-CoANA Acetoacetyl-CoANAB Sisomicin1-dSubstrate exhibiting substrate activation.3.4 ± 0.3258 ± 6.37601450 ± 1380330 ± 242.3 Netilmicin1-dSubstrate exhibiting substrate activation.9.8 ± 0.6454 ± 9.54631780 ± 995625 ± 333.5 Tobramycin1-dSubstrate exhibiting substrate activation.2.5 ± 0.591 ± 43643730 ± 790140 ± 140.4 Ribostamycin2.5 ± 0.369 ± 1.1276 Kanamycin B1.4 ± 0.0532 ± 0.3229 Gentamicin1-dSubstrate exhibiting substrate activation.22 ± 0.7364 ± 4.61653230 ± 1530505 ± 311.6 Dibekacin1-dSubstrate exhibiting substrate activation.22 ± 3.9354 ± 31161940 ± 420650 ± 467.0 Neomycin B19 ± 2120 ± 363 Paromomycin1-dSubstrate exhibiting substrate activation.86 ± 18144 ± 19173050 ± 1150380 ± 471.2 Butirosin1-dSubstrate exhibiting substrate activation.29 ± 1141 ± 7.2141530 ± 300250 ± 101.6 Lividomycin A1600 ± 96186 ± 51.2 Kanamycin A320 ± 2915.2 ± 0.40.48 Amikacin968 ± 1487.8 ± 1.20.08C Ribostamycin1-dSubstrate exhibiting substrate activation.8.3 ± 1.89.5 ± 0.811.4390 ± 4035 ± 1.40.90 Kanamycin B1-dSubstrate exhibiting substrate activation.16 ± 2.79 ± 0.65.62310 ± 81017 ± 2.20.07 Neomycin B1-dSubstrate exhibiting substrate activation.16 ± 3.37.0 ± 1.64.4310 ± 11018 ± 2.40.57 Sisomicin1-dSubstrate exhibiting substrate activation.39 ± 9.016 ± 1.94.12270 ± 83057 ± 4.60.25 Tobramycin1-dSubstrate exhibiting substrate activation.20 ± 4.77.8 ± 1.23.91360 ± 31042 ± 2.90.31 Netilmicin1-dSubstrate exhibiting substrate activation.46 ± 8.717 ± 1.73.71870 ± 38064 ± 3.90.34 Dibekacin1-dSubstrate exhibiting substrate activation.43 ± 1211 ± 1.52.61980 ± 60042 ± 2.50.21 Gentamicin1-dSubstrate exhibiting substrate activation.21 ± 124.2 ± 0.82.0610 ± 19034 ± 2.10.56 Butirosin429 ± 5411 ± 0.50.25 Paromomycin498 ± 4312 ± 0.30.24 Lividomycin ANA Kanamycin ANA AmikacinNAValues in A were determined at a fixed, saturating concentration of sisomicin. Values in B and C were determined at fixed saturating concentrations of acetyl-CoA and propionyl-CoA, respectively. NA, no acyl transfer activity was detected.1-a Values at high substrate concentrations.1-b Sum of velocities at low and high concentrations.1-c Measured at 5-cm path length.1-d Substrate exhibiting substrate activation. Open table in a new tab Figure 1Structures of 4,5-disubstituted (A) and 4,6-disubstituted (B) aminoglycosides used in the study. a and b, hexopyranosyl and 2-deoxystreptamine residues, respectively.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Figure 2Double-reciprocal plots of initial velocitiesversus [ribostamycin] obtained at a fixed, saturating concentration (100 μm) of propionyl-CoA in H 2 O (▴) and 98% D 2 O (●). The symbols are experimentally determined values, while the smooth curve is a fit of the data to Equation 5.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Values in A were determined at a fixed, saturating concentration of sisomicin. Values in B and C were determined at fixed saturating concentrations of acetyl-CoA and propionyl-CoA, respectively. NA, no acyl transfer activity was detected. The initial velocity pattern was determined using sisomicin and acetyl-CoA at five different concentrations of acetyl-CoA and four different concentrations of sisomicin. The resultant double-reciprocal plot was intersecting (data not shown), indicating a sequential kinetic mechanism. Alternative aminoglycoside kinetics were determined using propionyl-CoA as the variable substrate at fixed, saturating concentrations of eight different aminoglycosides (Table II). The values of V/Kfor propionyl-CoA varied modestly (over a 7-fold range) with this change being the result of changes in both V andK m values for propionyl-CoA. Alternate acyl coenzyme A kinetics were determined using sisomicin as the variable substrate at a fixed, saturating concentrations of five different coenzyme A derivatives (Table II). The values of V/K for sisomicin varied over a 2500-fold range with this change being the result of changes in both V and K values for sisomicin (Table II). Dead-end inhibition experiments were carried out using desulfocoenzyme A versus either acetyl-CoA or sisomicin. Desulfocoenzyme A exhibited linear, competitive inhibitionversus acetyl-CoA and linear, noncompetitive inhibitionversus sisomicin (data not shown).Table IIAlternate substrate kinetic parametersK mk catV/K propionyl-CoAμmmin −110 5A Tobramycin1.34 ± 0.1130.4 ± 0.523 Kanamycin B0.71 ± 0.038.91 ± 0.113 Ribostamycin2.73 ± 0.1123.4 ± 0.28.6 Dibekacin8.02 ± 0.8454.2 ± 1.66.8 Netilmicin4.40 ± 0.4128.9 ± 0.76.6 Neomycin B2.33 ± 0.1114.5 ± 0.26.2 Sisomicin3.01 ± 0.3517.6 ± 0.45.9 Gentamicin5.68 ± 0.9219.0 ± 0.73.3K mk catV/K sisomicinμmmin −110 5B Acetyl-CoA3.4 ± 0.3258 ± 6.3760 Malonyl-CoA8.7 ± 1.087 ± 2.5100 Propionyl-CoA39 ± 9.016 ± 1.94.1 Me-malonyl-CoA51 ± 1114 ± 1.02.8 n-Hexanoyl-CoA175 ± 125.9 ± 0.10.3Kinetic parameters for propionyl-CoA at fixed, saturating concentrations of the indicated aminoglycosides (A) and sisomicin at fixed, saturating concentration of the indicated coenzyme A derivatives (B). Open table in a new tab Kinetic parameters for propionyl-CoA at fixed, saturating concentrations of the indicated aminoglycosides (A) and sisomicin at fixed, saturating concentration of the indicated coenzyme A derivatives (B). Solvent kinetic isotope effects on acyl transfer were determined by measuring initial velocities at pH 7.0 in both H2O and 98% D2O. Reactions were performed at fixed, saturating concentrations of either acetyl-CoA or propionyl-CoA and at varying aminoglycoside concentrations. Two different aminoglycosides, ribostamycin and paromomycin, exhibiting relatively low and high K m values, respectively, using both acyl coenzymes A were used. The solvent kinetic isotope effects on V were 1.7 ± 0.02 and 2.1 ± 0.08 for ribostamycin and paromomycin, respectively, when acetyl-CoA was the acyl donor (Fig.3). A value forD2O V of 1.49 ± 0.05 was obtained using paromomycin and propionyl-CoA, and a value forD2O V of 1.66 ± 0.09 was obtained using only the high concentration ribostamycin data and propionyl-CoA (Fig. 2). The solvent kinetic isotope effects on V/K were identical for paromomycin using either acetyl-CoA or propionyl-CoA (0.98 ± 0.1 and 1.03 ± 0.1, respectively) and for ribostamycin using acetyl-CoA (1.15 ± 0.07). Using just the data obtained using propionyl-CoA at low ribostamycin concentrations (5–100 μm), the V and V/K solvent kinetic isotope effects were 1.5 ± 0.1 and 1.5 ± 0.2, respectively (Fig. 3). The latter is significantly greater than one, the value for the V/K solvent kinetic isotope effect observed in all other cases. From the determined steady-state kinetic parameters (Table I), particularly the relative V/K values, acetyl-CoA was the strongly preferred acyl donor. The M. tuberculosisAAC(2′)-Ic exhibited a comparable k cat value but a 10-fold lower K m value for acetyl-CoA compared with the AAC(2′)-Ia from P. stuartii (24Franklin K. Clarke A.J. Antimicrob. Agents Chemother. 2001; 45: 2238-2244Crossref PubMed Scopus (33) Google Scholar). Propionyl-CoA and malonyl-CoA exhibited significant acyl transferase activity, while other coenzyme A derivatives were either extremely poor substrates or did not demonstrate any activity. The drastic decrease in theV/K value observed by the addition of a methylene group to acetyl-CoA (propionyl-CoA) is primarily a k cateffect, whereas the decrease observed by the addition of a carboxyl group to acetyl-CoA (malonyl-CoA) is primarily a K m effect. Methylmalonyl-CoA, having both an additional methylene and carboxyl group, exhibited an intermediate K m value but a low k cat value, and acetoacetyl-CoA did not have any measurable acyl transfer activity. This acyl donor selectivity and differential affinity for CoA derivatives suggest a rigid, sterically restricted acyl coenzyme binding site that influences both binding and catalysis. Such severe acyl donor selectivity has also been reported for the AAC(6′) from Salmonella enterica (19Magnet S. Lambert T. Courvalin P. Blanchard J.S. Biochemistry. 2001; 40: 3700-3709Crossref PubMed Scopus (45) Google Scholar) in contrast to the AAC(6′) enzymes from E. coli andEnterococcus faecium (16Radhika K. Northop D.B. Biochemistry. 1984; 23: 5118-5122Crossref PubMed Scopus (17) Google Scholar, 18Wright G.D. Ladak P. Antimicrob. Agents Chemother. 1997; 41: 956-960Crossref PubMed Google Scholar). The AAC(2′)-Ic exhibits a very broad specificity with respect to aminoglycosides. All aminoglycosides having a 2′-amino substituent were substrates. Kanamycin A and amikacin (a semisynthetic derivative of kanamycin A) contain hydroxyl groups at the 2′ position but exhibited measurable activity, consistent with the unique ability of the enzyme to perform both O- and N-acyl transfer to the 2′ position. The substrate specificity, as evaluated by the V/Kvalues of substrates, varied over a 1000-fold range, much higher than reported for AAC(2′)-Ia. Among the 4,5-disubstituted aminoglycosides, ribostamycin was the most efficient substrate, whereas itsN 1-derivative, butirosin, exhibited 20- and 45-fold lower V/K values using acetyl-CoA and propionyl-CoA, respectively, as the acyl donors. Paromomycin having a 6′-hydroxyl substituent compared with neomycin B exhibited a 4-fold lowerV/K value for acetylation (18-fold lower for propionylation) than neomycin B, indicating the importance of the 6′-amino group in substrate recognition. Lividomycin A, the largest aminoglycoside with a 6′-hydroxyl substituent, exhibited 50-fold lower specificity for acetylation than Neomycin B and was the poorest 4,5-disubstituted aminoglycoside substrate examined. 4,6-Disubstituted aminoglycosides were generally better substrates for AAC(2′)-Ic. Sisomicin was a better substrate than itsN 1-ethyl derivative, netilmicin, at low substrate concentrations, whereas netilmicin exhibited stronger substrate activation. In contrast to ribostamycin and butirosin, sisomicin and netilmicin exhibited identical specificity for propionylation and only slightly lowered (1.6-fold) V/Kvalues. These data suggest that either the amino group at the N1 position may not be crucial for substrate recognition or that the relatively smaller ethyl substituent at the N1 position of netilmicin does not cause as much steric hindrance compared with the bulkyN 1-hydroxybutyramide moiety of butirosin. Tobramycin and kanamycin B (identical except that the former lacks a 3′-hydroxyl group) had comparable K m values, but the additional removal of the 4′-hydroxyl group, as in dibekacin, resulted in a 10-fold increase in the K m value, implying a role for the 4′-hydroxyl in substrate recognition. Dibekacin and gentamicin (a mixture of C1, C1a, and C2 used in this study), which have similar substitution patterns on the 6-deoxy-6-aminopyranose (primed) and central aminocyclitol (deoxystreptamine) rings but differ with respect to substitution around the 6"-ring, displayed identical kinetic parameters for acetylation. In contrast to the reported aminoglycoside specificity of the P. stuartii AAC(2′)-Ia, the M. tuberculosis AAC(2′)-Ic exhibits a much larger difference in the kinetic parameters for aminoglycoside substrates. While we observe differences in V/K values of almost 1000-fold using acetyl-CoA as the acyl donor, differences of only 10-fold were observed for the 2′-N-acetyltransferase from Providencia. Some of this difference can be attributed to the larger number of aminoglycosides tested in the current study, including the poor 2′-hydroxyl-substituted substrates, kanamycin A and amikacin. However, the direct comparison between the specificity exhibited between the two enzymes for the same substrate reveals some interesting differences. For example, gentamicin is a 4-fold better substrate for the AAC(2′)-Ia enzyme than netilmicin, while netilmicin is a 3-fold better substrate than gentamicin for the AAC(2′)-Ic enzyme. Butirosin exhibits the sameV/K value as kanamycin B for the AAC(2′)-Ia enzyme but a 16-fold lower V/K value than kanamycin B for the AAC(2′)-Ic enzyme. Despite the complex substrate specificity of both the AAC(2′)-Ic and AAC(2′)-Ia enzymes, the data suggest that the most important determinants of substrate recognition are present on the primed 6-deoxy-6-aminopyranosyl ring, where acetylation occurs, as previously suggested for AAC(2′)-Ia (24Franklin K. Clarke A.J. Antimicrob. Agents Chemother. 2001; 45: 2238-2244Crossref PubMed Scopus (33) Google Scholar). Many of the aminoglycosides exhibited nonlinear double-reciprocal plots, referred to as substrate activation, which have also been observed for AAC(6′)-Iy from S. enterica (19Magnet S. Lambert T. Courvalin P. Blanchard J.S. Biochemistry. 2001; 40: 3700-3709Crossref PubMed Scopus (45) Google Scholar) and gentamicin acetyltransferase (21Williams J.W. Northop D.B. J. Biol. Chem. 1978; 17: 5902-5907Abstract Full Text PDF Google Scholar). Kanamycin B, neomycin B, and ribostamycin did not show any substrate activation for acetylation but did for propionylation, whereas paromomycin and butirosin did not exhibit substrate activation upon propionylation but did for acetylation. None of the acyl-CoAs displayed substrate activation. There was no single structural feature or substitution that distinguishes between linear and nonlinear behavior of aminoglycoside substrates. Such aminoglycoside-dependent shifts from linear to nonlinear kinetics have been proposed to be the result of changes in kinetic mechanism (21Williams J.W. Northop D.B. J. Biol. Chem. 1978; 17: 5902-5907Abstract Full Text PDF Google Scholar) or to substrate-dependent subunit interactions in a dimeric enzyme (19Magnet S. Lambert T. Courvalin P. Blanchard J.S. Biochemistry. 2001; 40: 3700-3709Crossref PubMed Scopus (45) Google Scholar). The intersecting initial velocity plots obtained for sisomicin and acetyl-CoA suggests a sequential kinetic mechanism where both substrates must be bound to the enzyme for catalysis to occur. The sequential kinetic mechanism appears to be universally used by aminoglycoside N-acetyltransferases of all classes studied to date (3, 2′, and 6′), although there is no reason, a priori, why a ping-pong kinetic mechanism could not be used. Dead-end inhibition studies were performed to distinguish between the ordered and random addition of substrates. Desulfo-CoA exhibited linear, competitive inhibition versus acetyl-CoA and linear, noncompetitive inhibition versus sisomicin. These data are compatible with the ordered addition of acetyl-CoA followed by sisomicin. Alternate aminoglycoside kinetics is another useful approach to discriminate between various kinetic mechanisms and has been used earlier to determine the kinetic mechanisms of kanamycin 6′-N-acetyltransferase and AAC(6′)-Iy (17Radhika K. Northop D.B. J. Biol. Chem. 1984; 258: 12543-12546Abstract Full Text PDF Google Scholar, 19Magnet S. Lambert T. Courvalin P. Blanchard J.S. Biochemistry. 2001; 40: 3700-3709Crossref PubMed Scopus (45) Google Scholar). We determined the V/K value of propionyl-CoA at saturating concentrations of eight different aminoglycoside substrates. The determined values of V/K depend on the identity of aminoglycoside used with the range of values varying some 7-fold. TheV/K value for sisomicin was highly dependent on the identity of the acyl-CoA substrate used with the range varying over 2000-fold. For an ordered sequential mechanism, the V/K value of the first substrate to bind should not be influenced by the identity of the second substrate, but the V/K value of the second substrate to bind will be substantially influenced by the identity of the first substrate. In the present case, there is only a modest effect on theV/K value of propionyl-CoA on the identity of the aminoglycoside substrate and a much larger effect on the V/Kvalue for sisomicin on the identity of the acyl-CoA substrate. This suggests that the acyl-CoA preferentially binds to the free enzyme in agreement with the dead-end inhibition data. However, the modest change in the V/K value of propionyl-CoA with aminoglycoside identity requires that there be a degree of randomness in substrate binding with aminoglycoside substrates being able to bind to both enzyme-acyl-CoA and free enzyme. The independent evaluation of aminoglycoside binding to the free enzyme is being pursued. Studies of solvent kinetic isotope effects can be a useful tool in the determination of rate-limiting chemical steps and the kinetic mechanism, especially to distinguish between steady-state random and rapid equilibrium random mechanisms. In a rapid equilibrium random kinetic mechanism, the rates of substrate binding and product dissociation are very fast relative to the catalytic step, and equivalent isotope effects must be observed on V andV/K. This is not a requirement for the steady-state random mechanism. Solvent kinetic isotope effects were determined at pH 7.0, a region where small changes in pH(pD) did not have any effect on kinetic parameters (data not shown). Solvent kinetic effects were measured for ribostamycin and paromomycin (Fig. 3) at fixed, saturating concentrations of either acetyl-CoA or propionyl-CoA. The unitary values of the solvent kinetic isotope effects on V/K suggest that both ribostamycin and paromomycin are kinetically "sticky" and that no slow, solvent isotopically sensitive step occurs between aminoglycoside binding and the first irreversible step, generally assumed to be the release of first product. The only conditions in which the V/K solvent kinetic isotope effect was statistically greater than 1.0 were using ribostamycin at low concentrations and propionyl-CoA. We are presently examining other substrate pairs to see whether this is unique to this substrate pair or a general phenomenon that might have diagnostic utility. The larger values of the solvent kinetic isotope effects on V must reflect the effects of solvent isotopic substitution on the release of the second product or the conformational changes that allow this to occur. The solvent kinetic isotope effects observed for the AAC(2′)-Ic (1.5–2.1) may be compared with the much smaller effects of 1.3 reported for the S. enterica AAC(6′)-Iy (19Magnet S. Lambert T. Courvalin P. Blanchard J.S. Biochemistry. 2001; 40: 3700-3709Crossref PubMed Scopus (45) Google Scholar). The AAC(2′) family of aminoglycoside acetyltransferases uniquely exhibit the catalytic ability to perform either N-acetyl orO-acetyl transfer to various substrates. In the present study, we have established that the homogeneous M. tuberculosis AAC(2′)-Ic can use both kanamycin A and amikacin as substrates, thus demonstrating the first example of O-acetyl transfer to aminoglycosides by an aminoglycoside acetyltransferase. In contrast, in the case of the well studied 6′-N-acetyltransferases, all aminoglycosides bearing 6′-hydroxyl substituents are inhibitors and exhibit no detectable activity (18Wright G.D. Ladak P. Antimicrob. Agents Chemother. 1997; 41: 956-960Crossref PubMed Google Scholar, 19Magnet S. Lambert T. Courvalin P. Blanchard J.S. Biochemistry. 2001; 40: 3700-3709Crossref PubMed Scopus (45) Google Scholar). In the case of the related P. stuartiiAAC(2′)-Ia enzyme, the enzyme has been shown to perform theO-acetylation of peptidoglycan intermediates (23Payie K.G. Clarke A.J. J. Bacteriol. 1997; 179: 4106-4114Crossref PubMed Google Scholar), and very recently, the overexpressed and homogeneous enzyme has been shown to be capable of acetylation of aminoglycosides; however, aminoglycosides bearing a 2′-hydroxyl substituent, e.g. kanamycin A and amikacin, were not tested in this study (24Franklin K. Clarke A.J. Antimicrob. Agents Chemother. 2001; 45: 2238-2244Crossref PubMed Scopus (33) Google Scholar). These data are consistent with a physiological role in peptidoglycan acetylation for theProvidencia enzyme and by extension to the M. tuberculosis enzyme. In support of this idea, clinical resistance to aminoglycosides in M. tuberculosis has been shown to be due to single base substitutions in the 16 S rRNA gene or the gene encoding the S12 ribosomal protein, resulting in single amino acid substitutions (26Blanchard J.S. Annu. Rev. Biochem. 1996; 65: 215-239Crossref PubMed Scopus (234) Google Scholar). Finally, aac(2′) genes are universally present in mycobacterial species, and the deduced amino acid sequences are highly homologous (63–79%), yet their presence does not correlate with resistance to aminoglycosides (22Ainsa J.A. Perez E. Pelicic V. Berthet F. Gicquel B. Martin C. Mol. Microbiol. 1997; 24: 431-441Crossref PubMed Scopus (82) Google Scholar). These data are consistent with the theory that AAC(2′) gene products have a yet unidentified physiological role, and aminoglycoside acetylation is only a fortuitous secondary activity. Studies to determine the true substrate or substrates for the M. tuberculosis enzyme are being pursued.