A mechanistic study on different metal-substituted forms of the Metallo-beta-Lactamase NDM-1

PALACIOS, A.R.; LLARRULL, L.I.; VILA, A.J.

Chascomus

Congreso; 4th Latin American Meeting on Biological Inorganic Chemistry and 5th Argentinean Workshop of Bioinorganic Chemistry; 2014

Universidad de Buenos Aires; Society of Biological Inorganic Chemistry

INTRODUCTIONThe β-lactam antibiotic family represents the largest class of commercial antibiotics clinically used to treat Gram-negative and Gram-positive infections.1 β-Lactam antibiotics work by inhibiting cell wall synthesis. This class of antibiotics includes penicillin derivatives (penams), cephalosporins (cephems), monobactams and carbapenems. In particular, carbapenems comprise the latest generation of β-lactam antibiotics, used as last resort drugs to treat infections caused by resistant bacteria.The expression of β-lactamases is the most widely spread resistance mechanism displayed by bacteria. β-Lactamases are hydrolytic enzymes that selectively hydrolyze the four-membered β-lactam ring, rendering the antibiotic ineffective against their natural targets. There are four classes of β-lactamases: classes A, C, and D include enzymes with a serine in the active site, while class B enzymes are metal-dependent and hence called Metallo-β-Lactamases (MβLs). MβLs are divided into three subclasses, B1, B2, and B3.2 The class-B enzymes are of particular interest and concern given their ability to hydrolyze and provide resistance to virtually all classes of β-lactams antibiotics. Among them, NDM-1 is one of the most potent and widespread carbapenemases, raising an increasing clinical concern.3 NDM-1 belongs to subclass B1, is carried on a transmissible plasmid and has two Zn(II) ions in its active site. No clinical useful MβLs? inhibitor exists to date and unveiling the catalytic mechanism of MβLs is a prerequisite for the design of inhibitors or new β-lactam antibiotics.Here we report a mechanistic study by steady-state and pre-steady state kinetics on substrate hydrolysis by Zn(II), Co(II), and Cd(II)-NDM-1. EXPERIMENTAL METHODSNDM-1 was expressed as a soluble protein, with a deletion on the first 37 aminoacids, in the cytoplasm of E. coli cells, and purified to homogeneity. Metal-substituted NDM-1 was prepared by titration of the apoprotein with Co(II) or Cd(II). The apoprotein was prepared by dialysis of the purified holoprotein against two changes of 100 volumes of 10 mM HEPES, pH 7.5, 0.2 M NaCl and 20 mM EDTA over a 12-h period under stirring. EDTA was removed from the resulting apoenzyme solution by six dialysis steps against EDTA-free buffer, treated with Chelex 100.4 Metal incorporation to the apoenzyme was followed by UV-VIS spectroscopy and NMR, depending on the case. The kinetic parameters for the hydrolysis of different β-lactam antibiotics catalyzed by the different metal-substituted forms of NDM-1 under steady-state conditions were obtained by determination of the initial rate of reaction at different substrate concentrations. The plots of the dependence of the initial rates on substrate concentration were fitted to the Michaelis and Menten equation. The dissociation constants were estimated from competition experiments with the chromophoric chelators Mag-fura-2 and PAR.Quenching of the intrinsic fluorescence of NDM-1 upon substrate binding was monitored under pseudo-first order conditions in a Stopped-flow equipment.4 In addition, the reaction mechanism of hydrolysis of imipenem by Zn(II)-NDM-1 and Co(II)-NDM-1 was studied using a Stopped-Flow equipment coupled to a photodiode array. RESULTS AND DISCUSSIONSince the native Zn(II) ion present in MβLs is spectroscopically silent, and in order to follow the structural changes that occurred in the active site of NDM-1 during the hydrolysis reaction, we substituted the Zn(II) ions by Co(II) and Cd(II). The UV-Vis spectrum of Co(II)-substituted NDM-1 were identical to the ones previously reported for other B1 enzymes. Similar results were obtained by NMR for Cd(II)-NDM-1. The steady state parameters showed that the metal-substituted enzymes were active.For the cephalosporin substrates, substrate binding (followed in pre-steady-state studies of fluorescence quenching) was satisfactorily fitted to simple exponential functions. A simple reaction scheme was assumed for these substrates. A linear fit of the data points was therefore performed to calculate (k-1 + kcat) and k+1. Subtraction of the kcat value obtained from steady state experiments allowed us to calculate an estimate for k-1. The reaction rates for imipenem hydrolysis were obtained from a global analysis of fluorescence and absorption reaction traces at different substrate concentrations using the software DynaFit.5 The reaction mechanism for the hydrolysis of imipenem by Zn(II)-NDM-1 and Co(II)-NDM-1 was also studied. In the experiment performed with Zn(II)-NDM-1 we observed a spectral feature with a maximum at 390 nm that was not present in the absence of substrate and disappeared during turnover. This spectral feature is reminiscent of the spectrum of an intermediate observed upon imipenem hydrolysis by Zn(II)-BcII (the model B1 MβL).6 When the experiment was performed with Co(II)-NDM-1, we could detect the fast accumulation and subsequent consumption of a species with an absorption band at 405 nm, similar to the one characterized in Co(II)-BcII. The fact that the ligand field bands changed during the reaction indicated that the formation and consumption of that species is related to changes in the enzyme active site. CONCLUSIONOur substrate binding studies with Zn(II)-NDM-1 showed differences in the catalytic mechanisms of hydrolysis of different classes of β-lactam antibiotics. We showed the accumulation of a reaction intermediate upon hydrolysis of imipenem by Zn(II)- and Co(II)-NDM-1. We demonstrated that a negatively-charged reaction intermediate forms in the active site of this clinically relevant enzyme, upon hydrolysis of imipenem, as was the case for BcII. We propose that the chemical structure of this intermediate is a good starting point for the design of inhibitors that would target this clinically important enzyme. This information is also the cornerstone for the rational design of a common MβL inhibitor.REFERENCES1. Fisher J.F. et al., S. Chem. Rev. (2005), 105:395-424 2. Crowder M. W., Spencer J. and Vila A. J., Acc. Chem. Res. (2006), 39(10):721-728.3. Cannon B., Nature (2014), 509:S6-S8.4. Paul-Soto R. et al., FEBS Lett. (1998), 438(1-2):137-140. 5. Rasia R. M. and Vila A. J., J. Biol. Chem. (2004), 279(25):26046-26051.6. Tioni M. F. et al., J. Am. Chem. Soc. (2008), 282(42):30586-95.ACKNOWLEDGMENTSThe authors would like to thank CONICET, ANPCyT and the US NIH (1R01AI100560) for providing financial support to this project. ARP is recipient of a doctoral fellowship from CONICET. AJV and LIL are staff members from CONICET.