Aminoglycosides: Mechanisms of Action and Resistance - Springer Link

Chapter 14

Aminoglycosides: Mechanisms of Action and Resistance Maria L. Magalhães and John S. Blanchard

1 Antimicrobial Mechanism of Action Aminoglycosides are amongst the most important compounds used to treat serious nosocomial infections caused by aerobic, Gram-negative bacteria (1, 2). They are pseudo-polysaccharides containing amino sugars and can therefore be considered polycationic species for the purpose of understanding their biological interactions. Since they are highly positively charged at physiological pH values, they show high binding affinity for nucleic acids, especially for certain portions of the prokaryotic ribosomal RNA (rRNA). Different classes of aminoglycosides bind to different sites on rRNA, as will be discussed. Aminoglycoside uptake by bacterial cells has been shown to occur in three phases (3). The initial step involves electrostatic interactions between the antibiotic and the negatively charged lipopolysacharide (LPS) of the Gramnegative outer membrane (4). The polycationic antibiotic competitively displaces essential divalent cations (magnesium) that cross-bridge and stabilizes adjacent LPS molecules. Disruption of the outer membrane by this mechanism has been proposed to enhance permeability and initiate aminoglycoside uptake (4–6). Aminoglycoside transport across the cytoplasmic membrane involves an initial lag phase followed by a second phase in which the drug is rapidly taken up. Transport across the cytoplasmic membrane requires energy from the electron transport system in an oxygen-dependent process (3, 7–9). The intrinsic resistance of anaerobic bacteria to aminoglycosides can then be explained by the failure to transport the drug inside the cell. Once inside the cell, the drug binds to the 30S ribosomal subunit, at the Aminoacyl-tRNA (aa-tRNA) acceptor site (A) on the 16S ribosomal RNA (rRNA), affecting protein synthesis by induction of codon misreading and inhibition of translocation (10, 11). J.S. Blanchard ( ) Department of Biochemistry, Albert Einstein College of Medicine, Bronx, NY, USA [email protected]

Some aminoglycosides, like spectinomycin and kasugamycin, were found to have no effect on chain elongation (codon misreading) but block initiating ribosomes completely. Streptomycin and other aminoglycosides similarly block the initiation complex, but act later, decreasing the accuracy of translation (12). It is believed that fidelity of translation depends on two steps—an initial recognition between the codon on mRNA and the anticodon of the charged aa-tRNA, and a subsequent proofreading step. During the initial selection, the cognate codon is recognized, inducing GTP hydrolysis and the release of elongation factors from aa-tRNA (13). The aminoacyl end of aa-tRNA is free to move into the peptidyl transferase center on the 50S subunit, where peptide bond formation occurs (14). A similar sequence of events happen when a non-cognate codon is recognized. However, in such a case, following GTP hydrolysis and release of additional factors, non-cognate aa-tRNAs dissociate from the ribosome rather than enter the peptidyl transferase center, due to the lower stability of the codon-anticodon complex (13). Although the precise mechanism of aminoglycosideinduced miscoding is not completely understood, it has been shown that aminoglycosides enhance the binding stability of cognate aa-tRNAs to the small ribosomal subunit (15). It has been proposed that such stability enhancement would allow non-cognate tRNAs to enter the peptidyl transferase site, being incorporated into the nascent polypeptide chain. Recently, high-resolution crystal structures of the 30S ribosomal subunit (16, 17) as well as nuclear magnetic resonance (NMR) derived structures (18) of ribosomal constituents bound to aminoglycoside molecules have provided valuable information about the molecular mechanisms of aminoglycoside binding and action. The NMR structure of the complex between an A sitemimicking RNA molecule and the aminoglycoside paromomycin revealed how this aminoglycoside binds to the ribosome (18). Critical nucleotides for binding include A1492, U1495, as well as the C1407–G1494 and A1408–A1493 base pairs. These studies showed that the antibiotic binds in the major groove of the A-site in an L-shaped conformation. The

D.L. Mayers (ed.), Antimicrobial Drug Resistance, DOI 10.1007/978-1-59745-180-2_14, © Humana Press, a part of Springer Science+Business Media, LLC 2009

171

172

M.L. Magalhães and J.S. Blanchard

2-deoxystreptamine and 2,6-dideoxy-2,6-diamino-glucose rings contribute the most important intermolecular contacts. The N1 and N3 amino groups of the central deoxystreptamine ring, found in all typical aminoglycosides are required for specific binding to the 16S rRNA. A high-resolution crystal structure of the 30S subunit from Thermus thermophilus complexed with different antibiotics was reported in 2000, providing important insights into the molecular mechanisms of translation as well as the mode of action of aminoglycosides (17). In this work, Carter et al. proposed a model to address how typical aminoglycoside molecules increase the affinity of the aa-tRNA for the A-site. During translation, the selection of aa-tRNA occurs by formation of a mini-helix between the codon of mRNA and the anticodon of the cognate tRNA. They propose that when this tRNA-mRNA complex is formed, two adenines (A1492 and A1493) from 16S rRNA flip out from their intrahelical positions and form a hydrogen bonding-network with the 2′-OH groups on both sides of the codon-anticodon helix. The two adenines would sense the width of the minor groove, allowing for discrimination of distortions arising from mispairing. In the absence of any aminoglycoside molecule, some energy would be required to flip out these two adenine bases, but presumably this energy cost would be compensated by the formation of favorable interactions with the cognate aatRNA. By binding to the A site, the aminoglycoside stabilizes the flipped-out structure, thus reducing the energy cost of both cognate and non-cognate aa-tRNA binding and increasing aa-tRNA affinity for the A-site (17, 19). Therefore, typical aminoglycosides like paromomycin induce miscoding

HO HO

6' NH2 O 2' NH

2O

O

HO HO

NH2

6' 3 NH2 O

OH 1

NH2

HO H 2' H 2N

O

HO

2" OH

H N NH2 2

HO HN H O O C O H 3C HO HO HO

O

6'

O

O 2"

NH2

CH2OH OH

NH2 O

3 NH2 O HO 1 N O H CH2OH O OH HO 2" NH2 O

NH2 NH OH OH

NH2

Amikacin

Tobramycin

HO H 2N HO

O HO

CH3 OH

O

O

HN

H 3C

O

O

NH2

H 2N

O HO

O H N CH 3

Streptomycin

HO OH OH

3 NH2 NH2 1 O

HO

Ribostamycin

H 2N

by mimicking the conformational change in the 16S rRNA induced by a correct codon-anticodon pair. Indeed, it has been reported that aminoglycosides stabilize aa-tRNA binding about sixfold (15). In contrast, the rate of aminoglycoside-induced misreading ranges from 20- to 200-fold (20, 21) (depending on the codon and the antibiotic), suggesting the existence of additional mechanisms by which binding of aminoglycosides induces codon misreading (Fig. 1). The structure of the atypical aminoglycoside streptomycin bound to the 30S subunit has also been reported (17). The data revealed that the drug makes interactions with residues from four different domains of the 16S rRNA, including U14 in helix 1, C526 and G527 from helix 18, A913 and A914 from helix 27 and 28, respectively, and C1490 and G1491 from helix 44. It also makes contacts with K45 from protein S12. The reported data offers a structural rationale for the observed properties of streptomycin. It had previously been reported by Lodmell and Dahlberg (22) that there are two alternative base pairing schemes in E. coli rRNA during translation—one which leads to a ribosomal ambiguity (ram) conformation, with high affinity for tRNA which results in increased miscoding, and a second that leads to a restrictive state with low tRNA affinity—and the balance of these two states could be involved in the proofreading process (22, 23). The structural data from the streptomycin complex indicate that this aminoglycoside preferentially stabilizes the ram state (17), providing an explanation for the error-prone translation induced by this drug. By stabilizing the ram state, streptomycin would increase initial binding of non-cognate

OH O

CH3 NH

O O HO

NH2 HO

Apramycin

Fig. 1 Structures of typical (upper) and atypical (lower) Aminoglycosides

Spectinomycin

OH NH2

14 Aminoglycosides: Mechanisms of Action and Resistance

tRNAs as well as make the transition to the restrictive state more difficult, thereby affecting the proposed balance of such states and hence, proofreading. Although the mechanism of action at the translational level of aminoglycosides has been extensively clarified by the data above, the connection between protein misreading and bactericidal activity remains unclear. Most antibiotics that target ribosomes are bacteriostatic, while aminoglycosides are unique inhibitors of translation that cause “cidal” activity (24). In addition to codon misreading, early studies on streptomycin have revealed an additional effect: membrane damage. Several studies showed that the treatment of E. coli cells with streptomycin led to the loss of intracellular nucleotides (25), amino acids (26) and K+ (27). Later studies (12, 28, 29) have proposed that misreading would play an indirect, but essential and determinant role in the bactericidal action of aminoglycosides. The following model has been widely accepted: (1) small amounts of the antibiotic penetrate the cell, by a mechanism that is not completely understood, and binding to the A site in ribosomes that are actively elongating proteins cause a small degree of misreading; (2) the misread proteins are misfolded and are incorporated into the membrane, where it creates channels that permit a larger influx of antibiotic; (3) the intracellular antibiotic concentration rises and the drug is trapped inside the cell (30), resulting in the complete inhibition of protein synthesis, which causes bacterial death. The only difference in the sequence of the 16S rRNA between prokaryotes and eukaryotes is at position 1408, which is adenosine in all prokaryotic and eukaryotic mitochondrial sequences, but guanosine in cytoplasmatic eukaryotic sequences. The A1408–A1493 base pair in the bacterial ribosome creates a binding pocket for the primed ring that doesn’t occur in the eukaryotic structure, explaining the specificity of the drug for the bacterial target (18, 31).

2 Mechanism of Drug Resistance 2.1 Ribosomal Mutations Although target modification is a very common mechanism of bacterial drug resistance, clinical aminoglycoside resistance is generally not manifested by mutations of the ribosome. Most bacteria have multiple copies of the genes encoding rRNA, and thus to generate resistance, every copy of such gene would have to be mutated, and the probability of such occurrence is virtually nonexistent. Mycobacterium is the only genus that contains a single copy of the ribosomal operon (32) and, accordingly, is the only case in which clinical resistance due to ribosomal

173

mutations is relevant. High-level resistance to streptomycin in Mycobacterium tuberculosis has been reported to result from mutations in the genes encoding two components of the ribosome, the 16S rRNA (33–35) and the S12 protein (35, 36). The most frequently occurring mutation associated with streptomycin resistance in M. tuberculosis consists of point mutations in the ribosomal S12 protein encoded by gene rpsL. Mapping of these mutations revealed that all mutations occurred in highly conserved regions of the gene encoding one of the two critical lysines (K43 and K88) (35, 37, 38). Although structural studies have revealed that streptomycin makes direct contacts with S12 (17), mutations in this protein appear to affect streptomycin binding by perturbing the overall structure of 16S rRNA (39). In addition to streptomycin resistance, rpsL mutations can also lead to a different phenotype: streptomycin dependence. These mutations are associated with a hyper accurate phenotype, having extremely low translational error rates (38, 40). The mutations in the M. tuberculosis rrs gene, encoding the 16S rRNA, affect two highly conserved regions, the loop 530 and the region around nucleotide 912 resulting in decreased affinity for streptomycin (1, 2, 34, 36). Recently, it has also been observed that certain mutations in the conserved 530 stem-loop of 16S rRNA also results in streptomycin dependence phenotype (33).

2.2 16s rRNA Methylation Aminoglycoside-producing organisms have a range of defensive options available to avoid self-inactivation, including target modification and enzymatic inactivation of the drug. Members of the Actinomycetes produce inactive aminoglycosides, which are acetylated or phosphorylated molecules that are occasionally cleaved to produce the final active molecules (2, 41, 42). However, to further resist the secreted active compounds, many aminoglycosideproducing organisms also express rRNA methylases capable of modifying the 16S rRNA molecule at specific positions, thus preventing further binding of the drug (42, 43). A number of genes encoding S-adenosylmethionine (SAM)dependent methylases have been identified from several aminoglycoside producers (44–49), and such enzymes are members of the Agr family (for aminoglycoside resistance). In aminoglycoside-producing actinomycetes, methylation of residue G1405 has shown to result in high-level resistance to kanamycin and gentamicin (43, 45) while methylation of residue A1408 gave resistance to kanamycin, tobramycin, sisomycin and apramycin, but not gentamicin (43, 44, 50). Methylation of these nucleotides abolishes important intermolecular contacts between rRNA and the aminoglycoside molecule.

174

Until recently, ribosomal protection by methylation of 16S rRNA had been restricted to aminoglycoside-producing actinomycetes. Recently, however, several plasmid-encoded 16S rRNA methylases, encoded by the genes armA, rmtA, rmtB, and rmtC, have emerged in clinical isolates that exhibit a high level of resistance to numerous aminoglycosides (51–55). Analysis of the genetic environment of armA, rmtA, and rmtB genes imply that these resistance genes are present on mobile genetic elements carried by transferable large plasmids (54). Since these methylases can modify all copies of 16S rRNA, conferring a very high level of resistance against most clinically important aminoglycosides, the emergence of this type of resistance among clinically important microbial pathogens is of special concern for the future, as these genes can be easily disseminated.

2.3 Efﬂux-Mediated Resistance Aminoglycosides are vital components in the treatment of infections caused by the opportunistic pathogen Pseudomonas aeruginosa in cystic fibrosis patients. Such infections are difficult to treat because of intrinsic and acquired resistance caused by the expression of multidrug efflux systems in these organisms. Efflux systems are able to export an impressive variety of structurally unrelated molecules reducing the intracellular accumulation of antibiotics necessary for target inhibition. Intrinsic resistance is characterized by the constitutive expression of efflux pumps causing a natural low-level resistance to various antibiotics (56, 57). Mutations in the regulatory genes of the pumps, or induction of expression in the presence of a substrate, can lead to the overexpression of the formerly constitutive genes, causing high-level antibiotic resistance (2, 57). From a clinical perspective, the most relevant multidrug efflux systems are members of the resistance-nodulationdivision (RND) family (56). Several RND proteins have been shown to be involved in intrinsic aminoglycoside resistance in various Gram-negative pathogens (58–63). RND transporters use membrane proton motive force as energy source and are localized in the cytoplasmic membrane of Gramnegative bacteria. A membrane fusion protein (MFP) localized in the periplasmic space connects the RND transporter to the outer membrane pore (OMP) forming a continuous tripartite channel able to export substrates efficiently out of the cell (64, 65). Early studies have shown that intrinsic low resistance to aminoglycosides, tetracycline and erythromycin, in P. aeruginosa is mediated by the expression of the Mex (for multiple efflux) pumps, which are composed of a transmembrane protein (MexY), an outer membrane channel (OprM), and a periplasmic membrane fusion lipoprotein (MexX) (59, 60). The identity of the outer membrane channel of this tripartite


system remains under debate, but auxiliary OprM-like proteins such as OpmG and OpmI may also interact with MexX and MexY to form a tripartite functional pump (66). Later studies have proposed that the upregulation of amrAB genes, encoding a multidrug efflux system belonging to the NDR family in P. aeruginosa, also plays an important role in clinical resistance to aminoglycosides (61). Burkholderia pseudomallei is the causative agent of melioidosis, a disease that can be rapidly fatal if manifested in acute form. This Gram-negative bacillus is intrinsically resistant to a wide range of antimicrobial agents caused by the expression of efflux systems. Studies have also identified the presence of AmrAB-OprA multidrug efflux system specific for aminoglycoside and macrolide antibiotics (67). The E. coli genome contains several genes coding for RND transporters. The AcrD transporter was first identified based on the sequence similarity with the MexY sequence, and further shown to participate in the active effux of aminoglycoside molecules (63). Interestingly, recent studies have shown that AcrD not only captures aminoglycoside molecules from the cytoplasm, but also from the periplasmic space, followed by the active efflux of the drug out of the cell (68). In mycobacteria the majority of drug efflux pumps identified so far belong to the major facilitator superfamily (MFS). Efflux-mediated resistance to aminoglycoside and tetracycline have been recently described in Mycobacterium fortuitum by the expression of the tap gene (69), as well as in Mycobacterium bovis and Mycobacterium tuberculosis by the expression of the P55 gene (69, 70). Sequence analysis revealed 16 open reading frames encoding putative drug efflux pumps belonging to MFS class in M. tuberculosis (71). Such putative efflux pumps could account for streptomycin-resistant clinical isolates of M. tuberculosis that cannot be assigned to any other mechanism to date.

2.4 Enzymatic Drug Modiﬁcation The most common mechanism of aminoglycoside clinical resistance is the structural modification of the aminoglycoside molecule resulting from the action of intracellular bacterial enzymes that catalyze the covalent modification of specific amino or hydroxyl functions (Fig. 2). The chemically modified drug exhibits diminished binding to the A site of bacterial 16S rRNA, causing loss of antibacterial activity in resistant organisms that harbor these enzymes (72). Structural studies of aminoglycosides complexed to the 16S rRNA have highlighted the importance of several amino and hydroxyl groups for the proper binding of aminoglycoside molecules (17, 18). The N1 and N3


175

Phosphorylation

HO HO

NH2 O OH

MgADP

MgATP

HO

H2N O HO

NH2 O

NH2 O OH

O

APH(3')

O

NH2 OH

O

Kanamycin A

O P

OH

O

H2N O HO

NH2 O OH O

OH

NH2 OH OH

Adenylylation

HO HO

NH2 O OH

MgATP

H2N O HO

NH2 OH

O O

Kanamycin A

NH2 OH

O O P

MgPP Ade

O HO

OH

ANT(4')

H2N O HO

NH2 O OH O

NH2 OH OH

OH

O

Acetylation HO HO

O

NH2 O

NH2 O OH

CoASH

AcCoA

HO HO

H2N O HO

NH2 O OH O

Kanamycin A

H3C

AAC(6') NH2 OH

NH O OH

H2N O HO

NH2 O OH O

OH

NH2 OH OH

Fig. 2 Mechanisms of Enzyme-catalyzed Covalent Modification of Aminoglycosides

amino groups of the deoxystreptamine ring hydrogen-bond to nucleotides U1495 and G1494; the 3′ and 4′-hydroxyl groups of the primed ring contact A1493 and A1492 phosphates, respectively; the 2′-amino position forms an internal hydrogen bond with the doubly primed ring that is important for the correct positioning of the primed ring; and the amino and hydroxyl groups of the triply primed ring make electrostatic interactions with the phosphate backbone of several rRNA residues. Therefore, it is clear that modifications of these conserved or semi-conserved positions would lead to deleterious effects on the binding properties and thus the antibacterial activity of the drug. There are three classes of aminoglycoside modifying enzymes: aminoglycoside nucleotidyltransferases (ANTs), aminoglycoside phosphotransferases (APHs), and aminoglycoside acetyltransferases (AACs). They are divided into subtypes according to which position on the drug the modification occurs. For instance, the APH(3′) modifies the 3′-hydroxyl of susceptible aminoglycosides. The enzymes are further clas-

sified on the basis of the pattern of resistance designated by a Roman numeral and, in some cases, a letter designating a specific gene. Aminoglycoside-modifying enzymes can be either plasmid or chromosomally encoded, the former being associated with transposable elements, facilitating the rapid spread of the resistance phenotype not only within a given species but also among a large variety of bacterial species.

2.4.1 Aminoglycoside Adenylyltransferases Aminoglycoside adenylyltransferases catalyze the reaction between Mg-ATP and aminoglycoside molecules to form the O-adenylated aminoglycoside and the magnesium chelate of inorganic pyrophosphate. These enzymes adenylate hydroxyl groups on the positions 2″, 3″, 4′, 6, and 9 where the most relevant reactions, from a clinical perspective, are catalyzed by ANT(2″) and ANT(4′) (1, 2).

176

ANT(3″) is characterized by resistance to the atypical aminoglycosides streptomycin and spectinomycin, modifying the 3″-hydroxyl position of streptomycin and 9-hydroxyl group of spectinomycin (73). ANT(6) and ANT(9) adenylate 6-hydroxyl and 9-hydroxyl groups of streptomycin and spectinomycin, respectively, in Gram-positive organisms (74, 75). ANT(2″) was first identified in a clinical isolate of Klebsiella pneumoniea in 1971 (76), catalyzing the O-adenylylation of the 2″-hydroxyl group of 4,6-substituted aminoglycoside molecules. This enzyme causes resistance to multiple aminoglycosides, since it adenylylates a broad range of substrate molecules (77). Mechanistic studies have shown that the enzyme follows a Theorell-Chance kinetic mechanism in which the nucleotide binds first followed by the aminoglycoside, pyrophosphate is released prior to the nucleotidylated aminoglycoside, and turnover is controlled by the rate-limiting release of the final product (78). Further studies by Van Pelt et al. (79) have indicated that the nucleoside monophosphate is transferred directly to the hydroxyl group of the antibiotic in one step and that the reaction proceeds through inversion of the stereochemistry about the α-phosphorous. Recent substrate specificity studies have confirmed the importance of the 2′-substitution on 2″-O-adenylation, where molecules containing 2′-amino groups, instead of a 2′-hydroxyl, favor adenylation to occur (77, 80). The ANT(4′) kanamycin nucleotidyltransferase was originally isolated from clinical isolates of Staphylococcus aureus in 1976, adenylylating the 4″-hydroxyl group of kanamycin. The enzyme can utilize ATP, GTP, or UTP as the nucleotide substrate and can inactivate a wide range of aminoglycosides including kanamycins A, B, and C, gentamicin A, amikacin, tobramycin, and neomycins B and C (81). Recently, a chromosomally encoded 4′-O-nucleotidyltransferase from Bacillus clausii has been reported to cause resistance to kanamycin, tobramycin, and amikacin. The aadD2 gene was chromosomally located in all strains and was not transferable by conjugation, suggesting that aadD2 is specific to B. clausii (82).

2.4.2 Aminoglycoside Phosphotransferases The APH class of enzymes is the second-largest group of aminoglycoside-modifying enzymes. These enzymes catalyze the transfer of the γ-phosphoryl group from ATP to hydroxyl groups on aminoglycoside molecules. As a consequence, favorable electrostatic interactions that formerly existed between the hydroxyl group and specific residues on rRNA are abolished, resulting in the poor


binding of the drug to its target—the ribosome. The majority of these enzymes belong to the APH(3′) subfamily, which is also the most widespread among pathogenic organisms (83). The aph(3′)-IIIa gene is found primarily in Grampositive cocci and confer resistance to a wide range of aminoglycoside antibiotics, including kanamycin, amikacin, neomycin, and butirosin (83). The three-dimensional structure of the corresponding APH(3′)-IIIa has been solved to 2.2 Å and has been shown to have significant structural similarity to eukaryotic serine/threonine and tyrosine protein kinases (EPK) (84). In addition to structural similarities, APH(3)′-IIIa is inhibited by specific EPK inhibitors (85), and is able to phosphorylate several EPK substrates (86). Recent evidence has shown that Ser/ Thr kinases are not exclusive to eukaryotes and many aminoglycoside-producing organisms have been shown to encode eukaryotic-like kinases (87). Based on such evidence an attractive possible origin for APHs is that an ancestral bacterial protein kinase also provided a means of protection against the toxic effects of aminoglycosides in producing organisms, and then diverged, during evolution, for detoxification purposes. APH(3′)-IIIa operates by a Theorell-Chance mechanism, where ATP binds prior to the aminoglycoside; the modified drug is the first product to leave, followed by the rate-limiting dissociation of ADP (88). In Gram-positive organisms, the expression of a bifunctional enzyme 6′-N-acetyltransferase and 2″-O-phosphotransferase is responsible for high-level resistance to most aminoglycosides currently used in clinical practice (89). Both activities can be separately expressed and the kinetic properties of the bifunctional enzyme do not differ from its monofunctional counterparts (90). Streptomycin resistance due to aminoglycoside phosphotransferases is the result of two classes of enzymes, the APH(3″) and the APH(6) (91). Both enzymes are found in the streptomycin producer Streptomyces griseus and the aph (6)-encoding gene is clustered with streptomycin biosynthetic genes. The reason for such redundancy in aminoglycoside self-defense is not known at the present. APH(4) and APH(9) are responsible for resistance to hygromycin and spectinomycin, respectively, by phosphorylation of the 4- and 9-hydroxyl positions on the respective aminoglycoside molecules (87).

2.4.3 Aminoglycoside Acetyltransferases Aminoglycoside acetyltransferases are the largest group of aminoglycoside-modifying enzymes and catalyze the acetylCoA-dependent N-acetylation of amino groups of typical


aminoglycoside molecules. This class of enzymes include four major subclasses, which modify the amino groups of positions 1 and 3 of the central deoxystreptamine ring as well as the 2′ and 6′ amino groups of the 2,6-dideoxy-2,6diamino-glucose ring (2, 50, 92, 93). The first aminoglycoside-modifying enzyme reported in bacteria was kanamycin 6′-N-acetyltransferase IV, first identified in 1965 by Okamoto and Suzuki (94). This enzyme was the second example (after the discovery of penicillinase) of a bacterial enzyme causing antibiotic resistance by drug inactivation or modification. AAC(6′)-IV was the subject of the development of new kinetic diagnostics of enzymatic mechanisms by Radika and Northrop (95) who used these methods to establish that AAC(6′)-IV follows a rapid equilibrium random kinetic mechanism (96). A chromosomally encoded aminoglycoside 6′-N-acetyltransferase (AAC(6′)-Iy) has been identified in clinical isolates of aminoglycoside-resistant Salmonella enterica (97). The aac(6′)-Iy gene was located at the end of a long operon in sensitive strains, however, a massive 60 kbp deletion placed the constitutive nmp promoter directly upstream of the gene, resulting in the observed resistance phenotype. The deduced AAC(6′)-Iy sequence of 145 amino acids showed significant primary sequence homology with the Gcn5-related N-acetyltransferases (GNAT) superfamily. This is an enormous superfamily of enzymes (>10,000 identified to date from published sequenced genomes), whose members show sequence homology to the histone acetyltransferases (HAT) (98). To date, over three dozen members of the GNAT family have been structurally characterized, revealing a structurally conserved fold. The kinetic characterization of AAC(6′)-Iy has shown that the enzyme presents narrow acyl-donor specificity, but very broad specificity with respect to aminoglycosides containing a 6′-amino functionality. Both substrates must bind to the enzyme before catalysis occurs, and the order of substrate binding was proposed to be random (99).The structural characterization of this enzyme in 2004, confirmed that AAC(6′)-Iy is a member of the GNAT superfamily and revealed strong structural similarities with the Sacharomyces cerevisiae Hpa2-encoded histone acetyltransferase (100). The authors also demonstrated that AAC(6′)-Iy catalyzes acetylation of eukaryotic histone proteins. Such structural and catalytic similarities suggest that bacterial aminoglycoside acetyltransferases and eukaryotic histone acetyltransferases may be evolutionarily linked. The aacA29b gene was identified from a multi-drug resistant clinical isolate of Pseudomonas aeruginosa, exhibiting high-level resistance to various aminoglycosides. On the basis of amino acid sequence homology, it was proposed that this gene encoded a 6′-N-acetyltransferase. Surprisingly, this enzyme was found to confer amino-

177

glycoside resistance not by acetylating the drug, but by sequestering aminoglycoside molecules as a result of tight binding, thus preventing the molecule from reaching its target: the ribosome (101). As previously discussed, high-level aminoglycoside resistance, in E. faecalis, is often due to the plasmidmediated expression of the bifunctional AAC(6′)-APH(2″) (89). In E. faecium, intrinsic resistance is mediated by the expression of the chromosomally encoded aac(6′)-Ii gene, conferring low-level resistance to aminoglycosides (102). Kinetic studies have shown that AAC(6′)-Ii follows an ordered Bi-Bi mechanism, in which acetyl-CoA binds first to the enzyme followed by the aminoglycoside (103). Chemistry is not rate-limiting, as evidenced by very small solvent isotope effects and large dependence of the maximum velocity on the solvent micro viscosity, arguing that a physical step, probably product dissociation governs the overall rate of catalysis (103). The molecular mechanism of this enzyme was investigated by mutagenesis studies and the role of several potential catalytic residues on the active site of the Enterococcal AAC(6′)-Ii were explored (104). These studies indicate that Glu72 is critical for the proper positioning and orientation of aminoglycoside substrates in the active site. In addition, the amide NH group of Leu 76 is implicated in important interactions with acetyl-CoA and transition state stabilization. The three-dimensional structure of the E. faecium AAC(6′)-Ii was solved at 2.7 Å resolution, revealing a compact GNAT fold (105). In a very recent report, Robicsek et al. identified a variant of gene aac(6′)-Ib in clinical isolates of Gram-negative bacteria that has acquired the ability to modify fluoroquinolones (106). This enzyme was shown to reduce the activity of ciprofloxacin by N-acetylation of the secondary amino nitrogen of its piperazinyl substituent. The acquisition of this additional substrate activity by an aminoglycoside acetyltransferase represents a notable adaptation that justifies considerable future concern. AAC(2′) is a class of aminoglycoside acetyltransferases with significantly more restricted occurrence in bacteria. All aac(2′) genes reported so far are chromosomally encoded and universally present in mycobacteria, where the physiological role is not understood (107). The aac(2′)-Ic gene of M. tuberculosis was cloned and expressed in E. coli and the purified enzyme acetylated all aminoglycoside substrates tested in vitro. Dead-end inhibition studies as well as alternative substrate diagnostic studies supported an ordered sequential mechanism with a degree of randomness, where binding of acyl-CoA is preferred followed by the aminoglycoside. The enzyme is able to perform both N-acetyl as well as O-acetyl transfer (108). The aac(2′) genes are not responsible for clinical resistance in mycobacteria.

178

The AAC(3) family of aminoglycoside acetyltransferases regioselectively modify the 3-amino group of the deoxy-streptamine ring, and at present this family includes five major types: I–V, based on the pattern of aminoglycoside resistance that they confer. As previously discussed, the 3-amino group is found in all aminoglycosides and is required for specific binding of these molecules to the A site of the rRNA. Acetylation at this position would disrupt crucial interactions required for specific binding, resulting in poor binding to the ribosome. The AAC(3) I and II isoenzymes preferentially modify the gentamicin group of aminoglycosides (109, 110). Initial velocity, product, dead-end, and substrate inhibition studies have reveled that this enzyme follows a random Bi-Bi kinetic mechanism where both substrates must bind to the enzyme active site before catalysis can occur (111). AAC(3)-III enzymes catalyze the covalent acetylation of a wide variety of aminoglycosides including gentamycin, tobramycin and neomycin (112). The AAC(3)-IV enzyme was the first aminoglycosidemodifying enzyme identified as capable of modifying the novel aminoglycoside used for veterinary use, apramycin (113). This enzyme was originally found in E. coli and S. typhimurium animal isolates (113, 114) but was quickly identified in human clinical isolates from hospitalized patients (115), representing a serious concern given the activity of this enzyme with essentially all therapeutically useful aminoglycosides (116). The enzyme from E. coli has been recently kinetically characterized, revealing the broadest aminoglycoside specificity range of all AAC(3) enzymes (116). Dead-end inhibition and isothermal titration calorimetry (ITC) experiments reveled that the enzyme follows a sequential, random, Bi-Bi kinetic mechanism. Substrate specificity studies showed that acylation at the 1-N position sterically interfere with 3-N acetylation. Similar results have been observed with other AAC(3) enzymes, including AAC(3)-III and AAC(3)-I. Sequence alignment studies indicate that this enzyme is not a member of the GNAT superfamily, but currently no structural data have been reported to confirm such findings. The last member of the AAC(3) class of enzymes identified to date, was AAC(3)-V, isolated from a clinical isolate of Pseudomonas aeruginosa resistant to kanamycin, gentamicin, tobramycin, and sisomycin (117). The only member of this class of enzymes to be structurally characterized to date is the Serratia marascens AAC(3)-I (118). The monomer fold was typical of the GNAT superfamily, with the characteristic central antiparallel β-sheet containing two amino-terminal helices on one side of the sheet and the two carboxy-terminal helices on the other.


3 Mechanism of the Spread of Resistance In general, the process by which bacteria become resistant to antibiotics occurs either by mutations or by horizontal gene transfer; in which one bacterium transfers genetic material to another, either of the same or different genus. The rapid spread of drug resistance among pathogenic bacteria is usually attributed to horizontal gene transfer since the development of antimicrobial resistance by mutational changes is a relatively slow process (119, 120). The natural history of the emergence of bacterial resistance has been proposed to involve gene transfer from antibiotic-producing soil organisms to Gram-positive bacteria, and then to Gram-negative bacteria (120). Many of the genes that mediate resistance are found on transferable plasmids or on transposons that can be disseminated among various bacteria. Transposons are mobile pieces of DNA that can insert themselves into various locations on the bacterial chromosome, as well as move into plasmids or bacteriophage DNA (121). Three mechanisms of gene transfer in bacteria have been identified: transformation, which involves the uptake and incorporation of naked DNA; conjugation, which depends on cell–cell contact to transfer DNA elements; and transduction whereby the host DNA is encapsulated into a bacteriophage that acts as the vector for its injection into a recipient cell (122). The rapid dissemination of aminoglycoside resistance among pathogenic organisms has been largely attributed to conjugation of plasmids and non-replicative transposons among bacteria (119, 120, 123, 124). A clinical example of the ongoing importance of conjugative plasmid transfer on resistance to aminoglycosides is the shocking case of untreatable and fatal neonatal septicemia mediated by Klebsiella pneumonia EK105, which carries a mobile plasmid encoding resistance to amikacin, ampicillin chloramphenicol, kanamycin, streptomycin, tobramycin, netilmicin, oxacillin, gentamicin, and mezlocillin (125). Although aminoglycosides are not first-line therapy for staphylococcal infections, the recent increase in nosocomial infections caused by aminoglycoside-resistant strains is worrisome because it is often associated with resistance to drugs commonly used to treat staphylococcal infections (126). In addition, aminoglycoside resistance plasmids can reside in avirulent Staphylococcus epidermides strains present in skin flora of ill patients, being a reservoir that can be further transferred to virulent strains via conjugative transfer (127, 128). Recent studies have shown that 80% of methicillin-resistant S. aureus (MRSA) infections showed resistance to multiple aminoglycosides including gentamycin, tobramycin, kanamycin, amikacin, astromicin, and arbekacin, where 56% of such cases carried a transferable plasmid encoding a bifunctional


aminoglycoside-modifying enzyme AAC(6′)-APH(2″) (129). The gene aac(6′)-aph(2″) is present in the Tn 4100-like transposon which is inserted in both the R plasmid and the chromosome of aminoglycoside-resistant isolates (89). The worldwide-disseminated armA gene confers highlevel resistance to essentially all clinically important aminoglycosides by methylation of the 16S rRNA. Recent studies have shown that armA gene is part of the functional transposon Tn-1548 together with an ant(3″)(9) gene (130, 131). The reported data suggest that armA gene is spread by conjugation followed by transposition. This combination accounts for the worldwide dissemination of aminoglycoside resistance by 16S rRNA methylases in pathogenic organisms. The fact that bacteria produce a remarkable array of tools to overcome the toxic effects of antimicrobials is already alarming. But the fact that such genetic information is located in mobile DNA elements, which can be easily and rapidly disseminated between most diverse bacteria, is particularly worrisome. Increased incidence of multi-drug resistant bacteria and rising evidence of resistance transfer from one organism to another may lead to increasing emergence of nosocomial pathogens for which there is no antibiotic solution.

179

susceptible to the action of many aminoglycoside-modifying enzymes, and therefore is especially effective in the treatment of bacteria resistant to other aminoglycosides (24, 50, 135). After the success of amikacin in circumventing drug inactivation by modifying enzymes, other 1-N substituted derivatives, like isepamicin and arbekacin, were synthesized. In these derivatives, the 1-amino substitution protects against modification at 2″-hydroxyl and 3-amino positions, most likely by steric hindrance. This valuable feature explains the broad success and utility of the 1-amino substituted derivatives in situations of resistance to kanamycin, gentamicin, or tobramycin. These compounds, however, are still largely susceptible to ANT(4′) enzymes (81). Dibekacin (a 3′,4′-dideoxykanaymcin B derivative) was rationally designed to circumvent inactivation by the APH(3′) and ANT(4′) enzymes. Further modification of this drug by addition of a 4-amino-2-hydroxybutyryl group on the 1-amino group produced arbekacin. Arbekacin is particularly successful against MRSA, and has been used in Japan since 1990 (135, 136). However, strains of S. aureus resistant to arbekacin were recently isolated, where a mutation in the aac(6′)-aph(2″) gene permits arbekacin acetylation at the 4″ position (137).

4 Cross-Resistance References Aminoglycosides are often combined with a β-lactam drug in the treatment of infections caused by staphylococcal, enterococcal, and streptococcal strains (132). Resistance to β-lactams is usually caused by expression of β-lactamases, which are enzymes capable of hydrolyzing the β-lactam ring of penicillins, cephalosporins and related antimicrobial drugs, rendering them inactive. Since the report of the first β-lactamase-producing organism in 1983, β-lactam resistance is often associated with high-level resistance to aminoglycosides. In fact, genes encoding β-lactamases are usually carried on transferable plasmids that often also contain aminoglycoside resistance genes (119, 121, 133, 134). The resulting cross-resistance can make serious enteroccocal infections, such as endocarditis, extremely difficult to treat. Alarmingly, strains of E. faecium resistant to all known antibiotics have emerged as lethal pathogens in intensive care units in hospitals across the United States (24).

5 Alternative Agents Amikacin is a semi-synthetic derivative prepared from kanamycin A by acylation of the 1-amino group of the 2-deoxystreptamine ring with 2-hydroxy-4-aminobutyric acid. Because of this structural modification, amikacin is less

1. Kotra, L. P., Haddad, J., and Mobashery, S. (2000) Antimicrob Agents Chemother 44, 3249–3256 2. Magnet, S., and Blanchard, J. S. (2005) Chem Rev 105, 477–498 3. Schlessinger, D. (1988) Clin Microbiol Rev 1, 54–59 4. Hancock, R. E. (1984) Annu Rev Microbiol 38, 237–264 5. Hancock, R. E., Farmer, S. W., Li, Z. S., and Poole, K. (1991) Antimicrob Agents Chemother 35, 1309–1314 6. Hancock, R. E., Raffle, V. J., and Nicas, T. I. (1981) Antimicrob Agents Chemother 19, 777–785 7. Bryan, L. E., Kowand, S. K., and Van Den Elzen, H. M. (1979) Antimicrob Agents Chemother 15, 7–13 8. Miller, M. H., Edberg, S. C., Mandel, L. J., Behar, C. F., and Steigbigel, N. H. (1980) Antimicrob Agents Chemother 18, 722–729 9. Mates, S. M., Patel, L., Kaback, H. R., and Miller, M. H. (1983) Antimicrob Agents Chemother 23, 526–530 10. Davies, J., Gorini, L., and Davis, B. D. (1965) Mol Pharmacol 1, 93–106 11. Davies, J., Jones, D. S., and Khorana, H. G. (1966) J Mol Biol 18, 48–57 12. Davis, B. D. (1987) Microbiol Rev 51, 341–350 13. Green, R., and Noller, H. F. (1997) Annu Rev Biochem 66, 679–716 14. Moazed, D., and Noller, H. F. (1989) Nature 342, 142–148 15. Karimi, R., and Ehrenberg, M. (1994) Eur J Biochem 226, 355–360 16. Vicens, Q., and Westhof, E. (2001) Structure 9, 647–658 17. Carter, A. P., Clemons, W. M., Brodersen, D. E., Morgan-Warren, R. J., Wimberly, B. T., and Ramakrishnan, V. (2000) Nature 407, 340–348 18. Fourmy, D., Recht, M. I., Blanchard, S. C., and Puglisi, J. D. (1996) Science 274, 1367–1371

180 19. Fourmy, D., Yoshizawa, S., and Puglisi, J. D. (1998) J Mol Biol 277, 333–345 20. Davies, J., and Davis, B. D. (1968) J Biol Chem 243, 3312–3316 21. Bilgin, N., and Ehrenberg, M. (1994) J Mol Biol 235, 813–824 22. Lodmell, J. S., and Dahlberg, A. E. (1997) Science 277, 1262–1267 23. Puglisi, J. D., Blanchard, S. C., and Green, R. (2000) Nat Struct Biol 7, 855–861 24. Chambers, H. F., and Sande, M. A. (1995) in Goodman & Gilman’s The Pharmacological Basis of Therapeutics (Ruddon, R. W., Ed.) pp. 1103–1123, McGraw-Hill, New York 25. Roth, H., Amos, H., and Davis, B. D. (1960) Biochim Biophys Acta 37, 398–405 26. Anand, N., and Davis, B. D. (1960) Nature 185, 22–23 27. Dubin, D. T., and Davis, B. D. (1961) Biochim Biophys Acta 52, 400–402 28. Davis, B. D., Chen, L. L., and Tai, P. C. (1986) Proc Natl Acad Sci U S A 83, 6164–6168 29. Busse, H. J., Wostmann, C., and Bakker, E. P. (1992) J Gen Microbiol 138, 551–561 30. Nichols, W. W., and Young, S. N. (1985) Biochem J 228, 505–512 31. Lynch, S. R., and Puglisi, J. D. (2001) J Mol Biol 306, 1037–1058 32. Bercovier, H., Kafri, O., and Sela, S. (1986) Biochem Biophys Res Commun 136, 1136–1141 33. Honore, N., Marchal, G., and Cole, S. T. (1995) Antimicrob Agents Chemother 39, 769–770 34. Meier, A., Kirschner, P., Bange, F. C., Vogel, U., and Bottger, E. C. (1994) Antimicrob Agents Chemother 38, 228–233 35. Honore, N., and Cole, S. T. (1994) Antimicrob Agents Chemother 38, 238–242 36. Finken, M., Kirschner, P., Meier, A., Wrede, A., and Bottger, E. C. (1993) Mol Microbiol 9, 1239–1246 37. Blanchard, J. S. (1996) Annu Rev Biochem 65, 215–239 38. Toivonen, J. M., Boocock, M. R., and Jacobs, H. T. (1999) Mol Microbiol 31, 1735–1746 39. Allen, P. N., and Noller, H. F. (1989) J Mol Biol 208, 457–468 40. Bjorkman, J., Samuelsson, P., Andersson, D. I., and Hughes, D. (1999) Mol Microbiol 31, 53–58 41. Lacalle, R. A., Tercero, J. A., Vara, J., and Jimenez, A. (1993) J Bacteriol 175, 7474–7478 42. Cundliffe, E. (1989) Annu Rev Microbiol 43, 207–233 43. Beauclerk, A. A., and Cundliffe, E. (1987) J Mol Biol 193, 661–671 44. Skeggs, P. A., Thompson, J., and Cundliffe, E. (1985) Mol Gen Genet 200, 415–421 45. Thompson, J., Skeggs, P. A., and Cundliffe, E. (1985) Mol Gen Genet 201, 168–173 46. Kelemen, G. H., Cundliffe, E., and Financsek, I. (1991) Gene 98, 53–60 47. Holmes, D. J., and Cundliffe, E. (1991) Mol Gen Genet 229, 229–237 48. Holmes, D. J., Drocourt, D., Tiraby, G., and Cundliffe, E. (1991) Gene 102, 19–26 49. Ohta, T., and Hasegawa, M. (1993) J Antibiot (Tokyo) 46, 511–517 50. Mingeot-Leclercq, M. P., Glupczynski, Y., and Tulkens, P. M. (1999) Antimicrob Agents Chemother 43, 727–737 51. Yan, J. J., Wu, J. J., Ko, W. C., Tsai, S. H., Chuang, C. L., Wu, H. M., Lu, Y. J., and Li, J. D. (2004) J Antimicrob Chemother 54, 1007–1012 52. Yokoyama, K., Doi, Y., Yamane, K., Kurokawa, H., Shibata, N., Shibayama, K., Yagi, T., Kato, H., and Arakawa, Y. (2003) Lancet 362, 1888–1893 53. Doi, Y., Yokoyama, K., Yamane, K., Wachino, J., Shibata, N., Yagi, T., Shibayama, K., Kato, H., and Arakawa, Y. (2004) Antimicrob Agents Chemother 48, 491–496

M.L. Magalhães and J.S. Blanchard 54. Yamane, K., Wachino, J., Doi, Y., Kurokawa, H., and Arakawa, Y. (2005) Emerg Infect Dis 11, 951–953 55. Wachino, J., Yamane, K., Shibayama, K., Kurokawa, H., Shibata, N., Suzuki, S., Doi, Y., Kimura, K., Ike, Y., and Arakawa, Y. (2006) Antimicrob Agents Chemother 50, 178–184 56. Poole, K. (2004) Clin Microbiol Infect 10, 12–26 57. Poole, K. (2005) J Antimicrob Chemother 56, 20–51 58. Islam, S., Jalal, S., and Wretlind, B. (2004) Clin Microbiol Infect 10, 877–883 59. Hocquet, D., Vogne, C., El Garch, F., Vejux, A., Gotoh, N., Lee, A., Lomovskaya, O., and Plesiat, P. (2003) Antimicrob Agents Chemother 47, 1371–1375 60. Aires, J. R., Kohler, T., Nikaido, H., and Plesiat, P. (1999) Antimicrob Agents Chemother 43, 2624–2628 61. Westbrock-Wadman, S., Sherman, D. R., Hickey, M. J., Coulter, S. N., Zhu, Y. Q., Warrener, P., Nguyen, L. Y., Shawar, R. M., Folger, K. R., and Stover, C. K. (1999) Antimicrob Agents Chemother 43, 2975–2983 62. Magnet, S., Courvalin, P., and Lambert, T. (2001) Antimicrob Agents Chemother 45, 3375–3380 63. Rosenberg, E. Y., Ma, D., and Nikaido, H. (2000) J Bacteriol 182, 1754–1756 64. Murakami, S., Nakashima, R., Yamashita, E., and Yamaguchi, A. (2002) Nature 419, 587–593 65. Zgurskaya, H. I., and Nikaido, H. (2000) Mol Microbiol 37, 219–225 66. Jo, J. T., Brinkman, F. S., and Hancock, R. E. (2003) Antimicrob Agents Chemother 47, 1101–1111 67. Moore, R. A., DeShazer, D., Reckseidler, S., Weissman, A., and Woods, D. E. (1999) Antimicrob Agents Chemother 43, 465–470 68. Aires, J. R., and Nikaido, H. (2005) J Bacteriol 187, 1923–1929 69. Ainsa, J. A., Blokpoel, M. C., Otal, I., Young, D. B., De Smet, K. A., and Martin, C. (1998) J Bacteriol 180, 5836–5843 70. Silva, P. E., Bigi, F., de la Paz Santangelo, M., Romano, M. I., Martin, C., Cataldi, A., and Ainsa, J. A. (2001) Antimicrob Agents Chemother 45, 800–804 71. De Rossi, E., Arrigo, P., Bellinzoni, M., Silva, P. A., Martin, C., Ainsa, J. A., Guglierame, P., and Riccardi, G. (2002) Mol Med 8, 714–724 72. Llano-Sotelo, B., Azucena, E. F., Jr., Kotra, L. P., Mobashery, S., and Chow, C. S. (2002) Chem Biol 9, 455–463 73. Hollingshead, S., and Vapnek, D. (1985) Plasmid 13, 17–30 74. Ounissi, H., Derlot, E., Carlier, C., and Courvalin, P. (1990) Antimicrob Agents Chemother 34, 2164–2168 75. Murphy, E. (1985) Mol Gen Genet 200, 33–39 76. Benveniste, R., and Davies, J. (1971) FEBS Lett 14, 293–296 77. Gates, C. A., and Northrop, D. B. (1988) Biochemistry 27, 3820–3825 78. Gates, C. A., and Northrop, D. B. (1988) Biochemistry 27, 3826–3833 79. Van Pelt, J. E., Iyengar, R., and Frey, P. A. (1986) J Biol Chem 261, 15995–15999 80. Wright, E., and Serpersu, E. H. (2005) Biochemistry 44, 11581–11591 81. Le Goffic, F., Baca, B., Soussy, C. J., Dublanchet, A., and Duval, J. (1976) Ann Microbiol (Paris) 127, 391–399 82. Bozdogan, B., Galopin, S., Gerbaud, G., Courvalin, P., and Leclercq, R. (2003) Antimicrob Agents Chemother 47, 1343–1346 83. Kim, C., and Mobashery, S. (2005) Bioorg Chem 33, 149–158 84. Hon, W. C., McKay, G. A., Thompson, P. R., Sweet, R. M., Yang, D. S., Wright, G. D., and Berghuis, A. M. (1997) Cell 89, 887–895 85. Daigle, D. M., McKay, G. A., and Wright, G. D. (1997) J Biol Chem 272, 24755–24758 86. Daigle, D. M., McKay, G. A., Thompson, P. R., and Wright, G. D. (1999) Chem Biol 6, 11–18

14 Aminoglycosides: Mechanisms of Action and Resistance 87. Wright, G. D., and Thompson, P. R. (1999) Front Biosci 4, D9–D21 88. McKay, G. A., and Wright, G. D. (1995) J Biol Chem 270, 24686–24692 89. Culebras, E., and Martinez, J. L. (1999) Front Biosci 4, D1–D8 90. Martel, A., Masson, M., Moreau, N., and Le Goffic, F. (1983) Eur J Biochem 133, 515–521 91. Heinzel, P., Werbitzky, O., Distler, J., and Piepersberg, W. (1988) Arch Microbiol 150, 184–192 92. Azucena, E., and Mobashery, S. (2001) Drug Resist Updat 4, 106–117 93. Wright, G. D. (1999) Curr Opin Microbiol 2, 499–503 94. Okamoto, S., and Suzuki, Y. (1965) Nature 208, 1301–1303 95. Radika, K., and Northrop, D. (1984) Anal Biochem 141, 413–417 96. Radika, K., and Northrop, D. B. (1984) J Biol Chem 259, 12543–12546 97. Magnet, S., Courvalin, P., and Lambert, T. (1999) J Bacteriol 181, 6650–6655 98. Vetting, M. W., LP, S. D. C., Yu, M., Hegde, S. S., Magnet, S., Roderick, S. L., and Blanchard, J. S. (2005) Arch Biochem Biophys 433, 212–226 99. Magnet, S., Lambert, T., Courvalin, P., and Blanchard, J. S. (2001) Biochemistry 40, 3700–3709 100. Vetting, M. W., Magnet, S., Nieves, E., Roderick, S. L., and Blanchard, J. S. (2004) Chem Biol 11, 565–573 101. Magnet, S., Smith, T. A., Zheng, R., Nordmann, P., and Blanchard, J. S. (2003) Antimicrob Agents Chemother 47, 1577–1583 102. Costa, Y., Galimand, M., Leclercq, R., Duval, J., and Courvalin, P. (1993) Antimicrob Agents Chemother 37, 1896–1903 103. Draker, K. A., Northrop, D. B., and Wright, G. D. (2003) Biochemistry 42, 6565–6574 104. Draker, K. A., and Wright, G. D. (2004) Biochemistry 43, 446–454 105. Wybenga-Groot, L. E., Draker, K., Wright, G. D., and Berghuis, A. M. (1999) Structure 7, 497–507 106. Robicsek, A., Strahilevitz, J., Jacoby, G. A., Macielag, M., Abbanat, D., Park, C. H., Bush, K., and Hooper, D. C. (2006) Nat Med 12, 83–88 107. Ainsa, J. A., Perez, E., Pelicic, V., Berthet, F. X., Gicquel, B., and Martin, C. (1997) Mol Microbiol 24, 431–441 108. Hegde, S. S., Javid-Majd, F., and Blanchard, J. S. (2001) J Biol Chem 276, 45876–45881 109. Le Goffic, F., Martel, A., and Witchitz, J. (1974) Antimicrob Agents Chemother 6, 680–684 110. Williams, J. W., and Northrop, D. B. (1976) Biochemistry 15, 125–131 111. Williams, J. W., and Northrop, D. B. (1978) J Biol Chem 253, 5902–5907 112. Biddlecome, S., Haas, M., Davies, J., Miller, G. H., Rane, D. F., and Daniels, P. J. (1976) Antimicrob Agents Chemother 9, 951–955

181 113. Davies, J., and O’Connor, S. (1978) Antimicrob Agents Chemother 14, 69–72 114. Chaslus-Dancla, E., Martel, J. L., Carlier, C., Lafont, J. P., and Courvalin, P. (1986) Antimicrob Agents Chemother 29, 239–243 115. Chaslus-Dancla, E., Pohl, P., Meurisse, M., Marin, M., and Lafont, J. P. (1991) Antimicrob Agents Chemother 35, 590–593 116. Magalhaes, M. L., and Blanchard, J. S. (2005) Biochemistry 44, 16275–16283 117. Coombe, R. G., and George, A. M. (1982) Biochemistry 21, 871–875 118. Wolf, E., Vassilev, A., Makino, Y., Sali, A., Nakatani, Y., and Burley, S. K. (1998) Cell 94, 439–449 119. Waters, V. L. (1999) Front Biosci 4, D433–456 120. Courvalin, P. (1994) Antimicrob Agents Chemother 38, 1447–1451 121. Gold, H. S., and Moellering, R. C., Jr. (1996) N Engl J Med 335, 1445–1453 122. Davison, J. (1999) Plasmid 42, 73–91 123. Dzidic, S., and Bedekovic, V. (2003) Acta Pharmacol Sin 24, 519–526 124. Feizabadi, M. M., Asadi, S., Zohari, M., Gharavi, S., and Etemadi, G. (2004) Can J Microbiol 50, 869–872 125. Tolmasky, M. E., Roberts, M., Woloj, M., and Crosa, J. H. (1986) Antimicrob Agents Chemother 30, 315–320 126. Allen, J. R., Hightower, A. W., Martin, S. M., and Dixon, R. E. (1981) Am J Med 70, 389–392 127. Forbes, B. A., and Schaberg, D. R. (1983) J Bacteriol 153, 627–634 128. Archer, G. L., and Johnston, J. L. (1983) Antimicrob Agents Chemother 24, 70–77 129. Udou, T. (2004) Am J Infect Control 32, 215–219 130. Galimand, M., Sabtcheva, S., Courvalin, P., and Lambert, T. (2005) Antimicrob Agents Chemother 49, 2949–2953 131. Gonzalez-Zorn, B., Catalan, A., Escudero, J. A., Dominguez, L., Teshager, T., Porrero, C., and Moreno, M. A. (2005) J Antimicrob Chemother 56, 583–585 132. Gonzalez, L. S., 3rd, and Spencer, J. P. (1998) Am Fam Physician 58, 1811–1820 133. Hodel-Christian, S. L., and Murray, B. E. (1990) Antimicrob Agents Chemother 34, 1278–1280 134. Krogstad, D. J., Korfhagen, T. R., Moellering, R. C., Jr., Wennersten, C., and Swartz, M. N. (1978) J Clin Invest 61, 1645–1653 135. Smith, C. A., and Baker, E. N. (2002) Curr Drug Targets Infect Disord 2, 143–160 136. Yoshikawa, Y., Morikawa, K., Nonaka, M., and Torii, I. (2004) J Infect Chemother 10, 268–273 137. Fujimura, S., Tokue, Y., Takahashi, H., Kobayashi, T., Gomi, K., Abe, T., Nukiwa, T., and Watanabe, A. (2000) FEMS Microbiol Lett 190, 299–303