Tautomer-Specific Deacylation and Ω-Loop Flexibility Explain the Carbapenem-Hydrolyzing Broad-Spectrum Activity of the KPC-2 β-Lactamase

KPC-2 (Klebsiella pneumoniae carbapenemase-2) is a globally disseminated serine-β-lactamase (SBL) responsible for extensive β-lactam antibiotic resistance in Gram-negative pathogens. SBLs inactivate β-lactams via a mechanism involving a hydrolytically labile covalent acyl-enzyme intermediate. Carbapenems, the most potent β-lactams, evade the activity of many SBLs by forming long-lived inhibitory acyl-enzymes; however, carbapenemases such as KPC-2 efficiently deacylate carbapenem acyl-enzymes. We present high-resolution (1.25–1.4 Å) crystal structures of KPC-2 acyl-enzymes with representative penicillins (ampicillin), cephalosporins (cefalothin), and carbapenems (imipenem, meropenem, and ertapenem) obtained utilizing an isosteric deacylation-deficient mutant (E166Q). The mobility of the Ω-loop (residues 165–170) negatively correlates with antibiotic turnover rates (kcat), highlighting the role of this region in positioning catalytic residues for efficient hydrolysis of different β-lactams. Carbapenem-derived acyl-enzyme structures reveal the predominance of the Δ1-(2R) imine rather than the Δ2 enamine tautomer. Quantum mechanics/molecular mechanics molecular dynamics simulations of KPC-2:meropenem acyl-enzyme deacylation used an adaptive string method to differentiate the reactivity of the two isomers. These identify the Δ1-(2R) isomer as having a significantly (7 kcal/mol) higher barrier than the Δ2 tautomer for the (rate-determining) formation of the tetrahedral deacylation intermediate. Deacylation is therefore likely to proceed predominantly from the Δ2, rather than the Δ1-(2R) acyl-enzyme, facilitated by tautomer-specific differences in hydrogen-bonding networks involving the carbapenem C-3 carboxylate and the deacylating water and stabilization by protonated N-4, accumulating a negative charge on the Δ2 enamine-derived oxyanion. Taken together, our data show how the flexible Ω-loop helps confer broad-spectrum activity upon KPC-2, while carbapenemase activity stems from efficient deacylation of the Δ2-enamine acyl-enzyme tautomer.


Contents
. Data collection and refinement statistics for KPC-2 E166Q crystal structures.   Figure S1. View of the KPC-2 apo-enzyme active site.

Supporting notes:
Supporting note 1 Supporting note 2 Tables   Table S1. Data collection and refinement statistics for KPC-2 E166Q crystal
Distance measurements between backbone amide atoms (Ser70:N and Thr237:N) and the C7 carbonyl of the meropenem-derived acyl-enzyme in both Δ1-(2R) and Δ2 tautomers over the triplicate 500 ns MD simulations revealed these distances (and subsequent likely hydrogen bonding interactions) remained relatively stable at around 3 Å (Figure S8). With the Δ2 meropenem-derived acyl-enzyme, the C7 carbonyl lost interaction with the oxyanion hole in 2 frames, representing a rotation of this region in a similar orientation to that seen in crystal structures of SHV-1 and TEM-1. [3][4] This movement may also link to recent studies of KPC-2 that indicated permissive and non-permissive states may be possible during carbapenem hydrolysis. 5 Extensive analysis of Molecular Dynamics trajectories reveals no large differences in behaviour of the KPC-2:meropenem acyl-enzyme between the Δ1-(2R) and Δ2 tautomers.
These results were expected due to the minute differences in protonation state of C3 and N4 of the carbapenem-derived scaffold (Figure 3).
From visual inspection of the MD trajectories, water molecules in the putative deacylating position were found to exchange frequently (between the 100ps and -5ns time scale within the simulations). However, this site remained highly occupied, with a water molecule almost always within H-bonding distance of Glu166 and Asn170, and roughly 3.5-4.5 Å from meropenem C-7.
Higher level computational approaches (combined Quantum mechanics/ Molecular mechanics, QM/MM) were therefore utilised to elucidate differences which may exist between the tautomers of the meropenem-derived acyl-enzymes. Hybrid QM/MM umbrella sampling simulations were used to investigate meropenem hydrolysis in the KPC-2 acyl-enzyme.
In a protocol that we have previously shown to be effective for modeling deacylation of carbapenemderived acyl-enzymes by class A β-lactamases 12 , we employ the SCC-DFTB2 [13][14][15] QM method in the AMBER16 simulation package, 16 using the ff14SB 17 MM forcefield for protein, the TIP3P-Ew 18 water model and the General AMBER Force Field (GAFF) 16 for meropenem-derived regions not in the QM region. The QM region (modelled with a charge of -2) comprises the water in the deacylating position (DW), Ser70, Glu166 and the common carbapenem scaffold. Conventional QM/MM umbrella sampling MD was performed along two distance-based reaction coordinates to simulate the deacylation reaction ( Figure S5). The first is the linear combination of distances between the distance of Glu166 OE2 and the nearest proton of the deacylating water and the distance between the same proton and the oxygen of the deacylating water, 1: rx = d(OεGlu166-HDW)d(HDW-ODW) which corresponds to proton transfer from DW to Glu166. The second is the distance from the newly formed hydroxide to the C8 carbonyl, 2: = d(C7meropenem-ODW) which represents the nucleophilic attack of DW on the carbonyl carbon ( Figure S5). 20 ps of QM/MM MD were performed for each simulation window, and simulations repeated in triplicate. 2D minimum free energy paths (MFEPs) were calculated as previously described 12 , using the weighted-histogram analysis method (WHAM); the highest point (saddle point) along the MFEP is taken as the transition state, giving the activation free energy, ∆ ‡ G calc Table S4). The adaptive string method (ASM) path calculation method was also used for MFEP calculation of the first step of carbapenem deacylation. Two collective variables were used with the ASM, with the first collective variable subtly differing to the first reaction coordinate used for conventional US: The distance between Glu166 OE2 and a proton of the deacylating water, 1: rx = d(OεGlu166-HDW) and the distance between the meropenem C7 and the oxygen of the newly formed hydroxide, 2: ry = d(C7meropenem-ODW). Each node along the string was sampled for 60 ps and the simulations repeated in triplicate.

Supporting note 2
We measured the Ser130-Oγ to N4 distance over ASM trajectories derived from QM/MM MD simulations of the respective acyl-enzymes, finding significant differences both across the whole trajectory and between the respective deacylation transition states ( Figure S16). Analysis of H-bond networks determined that Ser130 is likely positioned, by Lys73 and Lys234, to interact with either the meropenem C3 carboxylate or N4 over both the whole trajectory and the transition state frames. Indeed, further analysis of the orientation of Ser130 throughout the ASM trajectories also reveals more consistent hydrogen bonding to the Lys234 side chain amide (Lys234-NZ) in the ∆1-(2R) (imine) tautomer than in the ∆2 (enamine) (Figure S15). Thus, more stable electrostatic interactions in the ∆1-(2R) acyl-enzyme sustain Ser130-Oγ in a position close to either meropenem-O9 or meropenem N-4, whilst the relative instability of the ∆2 acyl-enzyme allows the Ser130 side chain to explore conformational space further from meropenem N-4. These differences directly result from differences between tautomers in the hydrogen bonding network around the Lys234 side chain ( Figure S17).
Specifically, Lys234-NZ makes interactions exclusive to each tautomer; hydrogen bonding predominantly to the Thr235 backbone oxygen in the ∆1-(2R), and to the Val127 backbone oxygen in the ∆2 configuration ( Figure S17). In consequence different orientations of the Lys234 side chain prevail in each tautomer, subsequently hindering its propensity to hydrogen bond to Ser130-Oγ in ∆2 but promoting it in ∆1-(2R), accounting for the difference between tautomers in the distance between Ser130-Oγ and meropenem N-4. Alongside the H-bond networks around T216, T235 (as described in results) our simulations reveal that the Δ1-(2R) meropenem acyl-enzyme makes more extensive hydrogen bonding interactions with KPC-2, and is consequently more stable, than the ∆2, resulting in an increased free energy barrier to tetrahedral intermediate formation, as evidenced by the 7.1 kcal/mol difference between tautomers identified by ASM calculations.