Refinement
?Resolution range (A) no. of reflections Rwork/Rfree r.m.s.d.b ?bond length (A) bond angles (deg) ?average B-factors (A2) protein, inhib., wat., others Ramachandran plot statisctics (%) core regions additional allowed reg. generously allowed reg. disallowed regions Numbers in parentheses refer to the highest resolution shell. rmsd, root-mean-square deviation. focus X-ray generator with Saturn 944+ CCD X-ray detector. Three hundred degrees of data for this SHV-1: penem 1 dataset was collected with 0.5u Dphi oscilliation steps at a single ?wavelength (Cu Ka wavelength of 1.542 A). The data were processed with both anomalous reflections kept separate and as well by merging them. The anomalous scaled dataset was subsequently used to generate an anomalous difference Fourier map using the FFT program [17] of CCP4 suite to identify the positions of sulfur atoms of the penem 1 intermediate. X-ray diffraction data for the SHV-1:SA1-204 complex was collected at the Brookhaven National Synchrotron Light Source beamline X-29. Both the penem 1 and SA1-204 data sets were processed using HKL2000 [18]. Data for the SHV-1:SA-3-53 complex was collected at ALS beamline 4.2.2. and processed using D*trek [19]. Structure determinations were carried out using isomorphous replacement starting with chain A of the isomorphous crystal structure of SHV1 b-lactamase (PDB 1VM1) [20]. Crystallographic refinement was carried out using REFMAC [21] and model building was done using COOT [22]. After initial refinement, density in the active site revealed a covalent intermediate attached to the S70 side chain. The PRODRG2 server [23] was used to obtain the parameter and topology files for the covalently attached inhibitor intermediates observed in the electron density maps (Figure 2). Crystallographic refinement was monitored using the program DDQ [24] and the final model quality was assessed using PROCHECK [25]. Data collection and refinement statistics are summarized in Table 1.Minimum Inhibitory Concentrations (MICs)
Chemical synthesis for the compounds used have been previously described for SA1-204 [4], SA3-53 [8], and penem 1 [3]. MICs were performed as reported earlier [5]. We compared the combination of piperacilin/SA1-204 to i) piperacillin/ tazobactam (one of the most widely used b-lactam b-lactamse inhibitor combination in hospitals), ii) piperacillin/penem 1, and iii) piperacillin/LN1-255 [5]. Each of these b-lactam b-lactamase inhibitor combinations was compared to piperacillin alone. MICs were also performed against E. coli DH10B strains that possessed variants of SHV-1 that contained amino acid substitutions which confer resistance to commercially available b-lactamase inhibitors and advanced generation cephalosporins. The strains were constructed as described.
Kinetics
Kinetics for SA1-204 were performed by measuring Km (Ki) as a direct competition reaction as previously described [5].Figure 3. Penem 1 in SHV-1 active site. (A) Stereo view of penem 1 interactions in SHV-1 active site. Hydrogen bonds are shown as dashed lines. (B) Stereo view of superpositioning of penem 1:SHV-1 (black) and penem 2:SHV-1 structures (grey). (C) Stereo view of superpositioning of penem 1:SHV-1 (black) and penem 2:GC1 b-lactamase (grey). Results and Discussion SHV-1: Penem 1 Structure?The SHV-1: penem 1 structure was resolved at 1.84 A. The initial unbiased Fo-Fc map of the penem 1 soaked SHV-1 crystal revealed electron density emanating from catalytic S70 residue (Figure 2A). Based on the shape of the density and the suspected reaction mechanism (Figure 1B), a 7-membered-ring acyl-intermediate was modeled. All 5 different possible species (intermediates 7?1, Figure 1) were considered, yet after careful analysis of the stereochemistry and torsion angles, we concluded that penem 1 is in the R-isomer conformation (species 10). This interpretation was aided by the measured torsion angles of 4u, defined by N4-C3C-S, and 33u, defined by N4-C5-C6-C7. These two torsion angles are relative close to 0u indicating close to planarity due to the likelysp2 hybridization that we hypothesize to be present for these bonds representing species 10. The dihedral angle around the bond involving atom C7 is 67u (defined by atoms C5-C6-C7-C). To further aid in the interpretation of the density, the anomalous signals of S atoms further confirmed positions of the two sulphur atoms of the modeled species 10 (Figure 2B). The acyl-enzyme intermediate for penem 1 is in agreement with a previous electrospray ionization mass spectroscopy study, although the precise enantiomer could not be identified from the mass alone [13]. In addition to penem 1, the refinement also included a HEPES buffer molecule and one intact and one partial Cymal-6 molecule. 243 water molecules were examined carefully and added in the refinement including the deacylation water near residues E166 and N170.
The final R/Rfree is 18.7/23.4% and residues with phi-psi angles in the disallowed region of the Ramanchandran plot are not present (Table 1). We collected a number of additional datasets with penem 1 at different soaking time points ranging from 7 minutes to 21 hours and all reveal the same single R-configuration at the C7 stereo center of the acyl intermediate (data not shown). The intermediate representing species 10 has a few direct hydrogen bonding interactions with the enzyme active site and has a number of additional van der Waals interactions (Figure 3A). The b-lactam carbonyl group is positioned in the oxyanion hole forming hydrogen-bonds with backbone nitrogen atoms of S70 and A237 (Figure 2A and 3A). In addition, the NH-thiazepine atom is ?hydrogen bonded to the carbonyl atom of S130 (3.2 A) and the nitrogen of the bicyclic R1 side chain of penem 1 is hydrogen bonding to N132. As a result of forming the 7 membered heterocycle, the C3 carboxylate group points outwards and has ?2.9?.2 A distance water-mediated interactions with D104 and S106. The heterocyclic substitution lies above the polar side chain of N170 and also points toward the bulk solvent. We emphasize that the C3 carboxylate of penem 1 is not in the vicinity of R244, a residue previously been shown to be important and affect the Km (Ki) of penem 1 about 100-fold, when substituted [13]. In the preacylation steps residue R244, either directly or indirectly, has a role in binding the carboxyl moiety of substrates and inhibitors in the active site of b-lactamases [26]. The discrepancy with our covalently bound structure suggests that the Km (Ki) value is more dependent on the initial interactions of the inhibitors than on interactions made by a later inhibitory intermediate such as observed crystallographically. Note that none of the previous blactamase:penem complexes have the penem carboxyl moiety interacting with R244 or with an equivalent arginine residue at that position (these structures will be discussed below).
