Crystal structure of the IMP-1 metallo beta-lactamase from Pseudomonas aeruginosa and its complex with a mercaptocarboxylate inhibitor: binding determinants of a potent, broad-spectrum inhibitor

Metallo beta-lactamase enzymes confer antibiotic resistance to bacteria by catalyzing the hydrolysis of beta-lactam antibiotics. This relatively new form of resistance is spreading unchallenged as there is a current lack of potent and selective inhibitors of metallo beta-lactamases. Reported here are the crystal structures of the native IMP-1 metallo beta-lactamase from Pseudomonas aeruginosa and its complex with a mercaptocarboxylate inhibitor, 2-[5-(1-tetrazolylmethyl)thien-3-yl]-N-[2-(mercaptomethyl)-4 -(phenylb utyrylglycine)]. The structures were determined by molecular replacement, and refined to 3.1 A (native) and 2.0 A (complex) resolution. Binding of the inhibitor in the active site induces a conformational change that results in closing of the flap and transforms the active site groove into a tunnel-shaped cavity enclosing 83% of the solvent accessible surface area of the inhibitor. The inhibitor binds in the active site through interactions with residues that are conserved among metallo beta-lactamases; the inhibitor's carboxylate group interacts with Lys161, and the main chain amide nitrogen of Asn167. In the "oxyanion hole", the amide carbonyl oxygen of the inhibitor interacts through a water molecule with the side chain of Asn167, the inhibitor's thiolate bridges the two Zn(II) ions in the active site displacing the bridging water, and the phenylbutyryl side chain binds in a hydrophobic pocket (S1) at the base of the flap. The flap is displaced 2.9 A compared to the unbound structure, allowing Trp28 to interact edge-to-face with the inhibitor's thiophene ring. The similarities between this inhibitor and the beta-lactam substrates suggest a mode of substrate binding and the role of the conserved residues in the active site. It appears that the metallo beta-lactamases bind their substrates by establishing a subset of binding interactions near the catalytic center with conserved characteristic chemical groups of the beta-lactam substrates. These interactions are complemented by additional nonspecific binding between the more variable groups in the substrates and the flexible flap. This unique mode of binding of the mercaptocarboxylate inhibitor in the enzyme active site provides a binding model for metallo beta-lactamase inhibition with utility for future drug design.

beta-Lactamase epidemiology and the utility of established and novel beta-lactamase inhibitors

beta-Lactamase inhibitor:beta-lactam combinations remain one of the most successful strategies for the treatment of bacterial infections. Over the last 20 years the number and diversity of serine and metallo active site beta-lactamases has increased dramatically. This review highlights some of the new additions to the beta-lactamase arena and discusses how the commercially available beta-lactamase inhibitors are keeping pace with the changing epidemiology of beta-lactamases. In addition, we survey the progress with the design of novel inhibitors of serine and metallo-beta-lactamases. Focus is given to the recent advances in the design of metallo-beta-lactamase inhibitors as these enzymes pose a serious emerging threat to the use of all beta-lactam based therapies.

Crystal structure of the zinc-dependent beta-lactamase from Bacillus cereus at 1.9 A resolution: binuclear active site with features of a mononuclear enzyme

The structure of the zinc-dependent beta-lactamase II from Bacillus cereus has been determined at 1.9 A resolution in a crystal form with two molecules in the asymmetric unit and 400 waters (space group P3121; Rcryst = 20.8%). The active site contains two zinc ions: Zn1 is tightly coordinated by His86, His88, and His149, while Zn2 is loosely coordinated by Asp90, Cys168, and His210. A water molecule (W1) lies between the two zinc ions but is significantly closer to Zn1 and at a distance of only 1.9 A is effectively a hydroxide moiety and a potential, preactivated nucleophile. In fact, Asp90 bridges W1 to Zn2, and its location is thus distinct from that of the bridging water molecules in the binuclear zinc peptidases or other binuclear zinc hydrolases. Modeling of penicillin, cephalosporin, and carbapenem binding shows that all are readily accommodated within the shallow active site cleft of the enzyme, and the Zn1-bound hydroxide is ideally located for nucleophilic attack at the beta-lactam carbonyl. This enzyme also functions with only one zinc ion present. The Zn1-Zn2 distances differ in the two independent molecules in the crystal (3.9 and 4.4 A), yet the Zn1-W1 distances are both 1.9 A, arguing against involvement of Zn2 in W1 activation. The role of Zn2 is unclear, but the B. cereus enzyme may be an evolutionary intermediate between the mono- and bizinc metallo-beta-lactamases. The broad specificity of this enzyme, together with the increasing prevalence of zinc-dependent metallo-beta-lactamases, poses a real clinical threat, and this structure provides a basis for understanding its mechanism and designing inhibitors.

Nucleotide and amino acid sequences of the metallo-beta-lactamase, ImiS, from Aeromonas veronii bv. sobria

The Aeromonas veronii bv. sobria metallo-beta-lactamase gene, imiS, was cloned. The imiS open reading frame extends for 762 bp and encodes a protein of 254 amino acids with a secreted modified protein of 227 amino acids and a predicted pI of 8.1. To confirm the predicted sequence, purified ImiS was digested and the resulting peptides were identified, yielding an identical sequence for ImiS, with 98% identity to CphA. Both possessed the putative active-site sequence Asn-Tyr-His-Thr-Asp at positions 88 to 92, which is unique to the Aeromonas metallo-beta-lactamases.

Inhibition of metallo-beta-lactamases by a series of thiol ester derivatives of mercaptophenylacetic acid

A series of mercaptophenylacetic acid thiol esters bearing a phenyl substituent adjacent to the carboxylic acid function has been shown to be inhibitors of metallo-beta-lactamases. The inhibition of the Bacteroides fragilis CfiA and Bacillus cereus II metallo-beta-lactamases was Zn2- dependent, greater inhibition being observed at 1 microM ZnSO4 than at 100 microM ZnSO4. Despite this Zn2+ dependency, isothermal titration calorimetry studies illustrated that representative compounds had no detectable affinity for Zn2+ (K > 1 mM). This indicates that their mode of inhibition was not by chelation of the active site Zn2+. Greatest potency was observed against the Stenotrophomonas maltophilia L1 metallo-beta-lactamase with I50 values of between < 1.95 microM and 6 microM and SB-217843 exhibited a similar level of inhibition of this enzyme at 1 and 100 microM Zn2+ (I50 values 5 and 6 microM, respectively). Inhibition of B. cereus II metallo-beta-lactamase by SB-218018 and SB-217782 was competitive with Ki values of 185 microM and 1500 microM, respectively. Therefore, these compounds are specific inhibitors of metallo-beta-lactamases and provide further probes of the active sites of these enzymes.

Inhibition of metallo-beta-lactamases by a series of mercaptoacetic acid thiol ester derivatives

A series of mercaptoacetic acid thiol esters have been identified as metallo-beta-lactamase inhibitors. Electrospray mass spectrometry (ESMS) has shown that irreversible inhibition of the Bacillus cereus II metallo-beta-lactamase by SB214751, SB214752, and SB213079 was concomitant with a 90-Da increase in mass of the enzyme. Tryptic digestion of the B. cereus II inhibited with SB214751 illustrated that the peptide fragment, containing the only cysteine of the enzyme, had undergone a mass increment of 90 Da. It was further demonstrated that B. cereus II hydrolyzed this type of compound across the thiol ester bond to yield mercaptoacetic acid. Mercaptoacetic acid is the only molecular fragment common to SB214751, SB214752, and SB213079, and free mercaptoacetic acid does not bind covalently to B. cereus II. Therefore, it is concluded that these compounds inhibit B. cereus II by the mechanism-based delivery of mercaptoacetic acid, forming a disulfide linkage with the active sites cysteine (predicted mass shift = +90 Da) under the aerobic conditions of the assay. The different thiol esters examined had a broad range of potencies against the metallo-beta-lactamases tested. For example SB214751, SB214752, and SB213079 all had 50% inhibitory concentrations of < 10 and > 1,000 microM for the Stenotrophomonas maltophilia L-1 and Bacteroides fragilis CfiA enzymes, respectively. SB216968 was particularly active against the Aeromonas hydrophila CphA metallo-beta-lactamase and was found to be an uncompetitive inhibitor of this enzyme (Ki = 3.9 microM), whereas it exhibited irreversible inhibition of the L-1 enzyme. These observations with this series of compounds have revealed subtle differences between the active sites of different metallo-beta-lactamases. Finally, a novel application for isothermal titration calorimetry for assessing the zinc chelating activity of candidate inhibitors is also presented.

Phosphonamidate analogues of dipeptides with carboxypeptidase A and beta-lactamase-inhibitory activity: elucidation of the mechanism of beta-lactamase inhibition by electrospray mass spectrometry

A series of phosphonamidate compounds with different P1' amino acid residues have been shown to be irreversible inactivators of the serine beta-lactamase from Enterobacter cloacae P99. The efficiency of inhibition (based on k2/K values) of P99 by these derivatives, ordered in decreasing potency, is: beta-phenyl-beta-Ala > L-Phe > beta-Ala > Gly > D-Phe > D-Pro > D-thiazolidine. The D- and L-Phe compounds also inhibit carboxypeptidase A. The proline and thiazolidine derivatives were phosphonamidate methyl esters, whereas the others were salts of diacids. Electrospray mass spectrometry showed that equimolar mixtures of the P99 enzyme with each of the following derivatives, Gly, D-Phe, L-Phe, beta-Ala and beta-phenyl-beta-Ala, effected efficient adduct formation (70-95% of enzyme modified), illustrating the particularly active nature of some of these compounds. All the primary amino acid derivatives gave a similar mass increment, which suggests the displacement of the variable P1' part of the molecule. This observation provides evidence that the compounds phosphonylate the active-site serine, with the phosphonamidate bond as the scissile bond and the amino acid as the leaving group. The thiazolidine derivative (phosphonamidate methyl ester) also appeared to work by the same mechanism. The comparable proline derivatives caused lower than expected mass shifts of 227-229, and therefore it is proposed that with these compounds both the amino acid and the phosphonamidate ester methoxy group were displaced at the phosphorus atom during the inhibition process. Therefore, electrospray mass spectrometry has provided both a measure of potency and a rationale for the mechanism of inhibition of P99 by these compounds.

Rapid identification of metallo- and serine beta-lactamases

Simple methods to detect, identify, and differentiate metallo- and serine beta-lactamases were developed and used to differentiate enzymes produced by 17 clinical isolates of Xanthomonas maltophilia. All isolates exhibited beta-lactamase activity, and in 16 strains this was induced by imipenem. All but one isolate hydrolyzed imipenem (and meropenem), and in all cases this activity was inhibited by 1 mM EDTA. The metallo- and serine beta-lactamases in the cell extracts were distinguished on isoelectric focusing (IEF) gels by using the following procedures. (i) Cell lysates were preincubated with 83 mM EDTA prior to IEF and subsequent visualization with nitrocefin, and (ii) after IEF, the gels were overlaid with either 1 mM zinc sulfate or 100 microM BRL 42715 before staining with nitrocefin. Bands of beta-lactamase activity which were removed by BRL 42715 but unaffected by EDTA or zinc sulfate were categorized as serine beta-lactamases. Bands which were unaffected by BRL 42715 but inhibited by EDTA or enhanced by zinc sulfate were classified as metallo-beta-lactamases. By using this approach, seven metallo-beta-lactamases were differentiated with pI values of 4.8 (two strains), 5.5 (four strains), 5.7 (one strain), 6.0 (one strain), 6.4 (four strains), 6.6 (one strain), and 6.8 (three strains). The metallo-beta-lactamase band with a pI of 6.4 aligned with the recently characterized metallo-beta-lactamase from X. maltophilia 511. Heterogeneity was also observed for the serine beta-lactamases: 14 isolates elaborated serine beta-lactamase activity which focused with major bands with at least eight different pIs. The remaining three strains produced serine beta-lactamases which focused with five distinct bands with pIs of 6.4, 6.2, 5.7, 5.5, and 5.2. We conclude that X. maltophilia produces many types of metallo- and serine beta-lactamases distinguishable by these new methods and that the previously reported L-1 and L-2 enzymes are not solely representative of the beta-lactamases produced by this species.