Exploring the Role of the Ω-Loop in the Evolution of Ceftazidime Resistance in the PenA β-Lactamase from Burkholderia multivorans, an Important Cystic Fibrosis Pathogen

Abstract

The unwelcome evolution of resistance to the advanced generation cephalosporin antibiotic, ceftazidime is hindering the effective therapy of Burkholderia cepacia complex (BCC) infections. Regrettably, BCC organisms are highly resistant to most antibiotics, including polymyxins; ceftazidime and trimethoprim-sulfamethoxazole are the most effective treatment options. Unfortunately, resistance to ceftazidime is increasing and posing a health threat to populations susceptible to BCC infection. We found that up to 36% of 146 tested BCC clinical isolates were nonsusceptible to ceftazidime (MICs ≥ 8 μg/ml). To date, the biochemical basis for ceftazidime resistance in BCC is largely undefined. In this study, we investigated the role of the Ω-loop in mediating ceftazidime resistance in the PenA β-lactamase from Burkholderia multivorans, a species within the BCC. Single amino acid substitutions were engineered at selected positions (R164, T167, L169, and D179) in the PenA β-lactamase. Cell-based susceptibility testing revealed that 21 of 75 PenA variants engineered in this study were resistant to ceftazidime, with MICs of >8 μg/ml. Under steady-state conditions, each of the selected variants (R164S, T167G, L169A, and D179N) demonstrated a substrate preference for ceftazidime compared to wild-type PenA (32- to 320-fold difference). Notably, the L169A variant hydrolyzed ceftazidime significantly faster than PenA and possessed an ∼65-fold-lower apparent K_i (K_iapp) than that of PenA. To understand why these amino acid substitutions result in enhanced ceftazidime binding and/or turnover, we employed molecular dynamics simulation (MDS). The MDS suggested that the L169A variant starts with the most energetically favorable conformation (-28.1 kcal/mol), whereas PenA possessed the most unfavorable initial conformation (136.07 kcal/mol). In addition, we observed that the spatial arrangement of E166, N170, and the hydrolytic water molecules may be critical for enhanced ceftazidime hydrolysis by the L169A variant. Importantly, we found that two clinical isolates of B. multivorans possessed L169 amino acid substitutions (L169F and L169P) in PenA and were highly resistant to ceftazidime (MICs ≥ 512 μg/ml). In conclusion, substitutions in the Ω-loop alter the positioning of the hydrolytic machinery as well as allow for a larger opening of the active site to accommodate the bulky R1 and R2 side chains of ceftazidime, resulting in resistance. This analysis provides insights into the emerging phenotype of ceftazidime-resistant BCC and explains the evolution of amino acid substitutions in the Ω-loop of PenA of this significant clinical pathogen.

Keywords: Burkholderia; beta-lactamase; beta-lactams; ceftazidime.

FIG 1

(A) PenA crystal structure highlighting the Ω-loop (cyan) and active site. Residues E166 and N170 of the Ω-loop position a water molecule used during acylation and deacylation of β-lactams. Amino acids R164 and D179 form a salt bridge at the neck of the Ω-loop to maintain its shape. Substitutions at positions 164, 167, 169, and 179 in class A β-lactamases are found in clinical isolates and confer ceftazidime resistance. Hydrogen bonding interactions are depicted by gray dashed lines. (B) Amino acid sequences of the Ω-loops for six class A β-lactamases compared to PenA (white background). Single amino acid substitutions in Ω-loop residues that result in increased ceftazidime MICs or enhanced ceftazidime kinetics are shown by a yellow background; those that did not are shown by a pink background. All of the β-lactamases with substitutions listed were identified in clinical isolates and were not laboratory-generated strains (14–27, 45–50). (C) Chemical structure of ceftazidime with the R1 and R2 side chains highlighted in orange and purple, respectively.

FIG 2

(A) Percentage of strains susceptible (yellow) versus resistant (pink) to ceftazidime in a collection of 146 BCC clinical isolates from CF patients. (B) Ceftazidime MICs (bar graph) and anti-PenA immunoblotting results (bottom) for the R164X, T167X, L169X, and D179X variants. (Note that two immunoblots were conducted for each set of variants to fit all of the samples. In addition, the wild-type strain on each blot was placed in alphabetical order; red lines indicate where the immunoblots were spliced.) The blue bar represents the wild-type PenA in all of the bar graphs, the gray bar represents the variant that was purified for further analysis, and the green line represents the CLSI cutoff for ceftazidime resistance at 8 μg/ml.

FIG 3

(A) The interactions between the β-lactamase (E) and the β-lactam (S) were interpreted according to the scheme. Here, the formation of the noncovalent complex, E:S, is represented by the dissociation constant, K_s, which is equivalent to k_-1/k₁. k₂ is the first-order rate constant for the acylation step, or the formation of E-S complex. k₃ is the rate constant for the hydrolysis of the E-S acyl-enzyme and product (P) release. The Michaelis constant (K_m) is equivalent to K_s × (k₃/k₂ + k₃). (B) Progress curves of nitrocefin hydrolysis at absorbance at λ₄₈₂ for PenA and the Ω-loop variants. (C) Progress curves of ceftazidime hydrolysis at absorbance at λ₂₆₀ for PenA and the Ω-loop variants. (D to H) Progress curves showing the inhibition of nitrocefin hydrolysis at absorbance at λ₄₈₂ by increasing concentrations of ceftazidime as the inhibitor for PenA and the Ω-loop variants.

FIG 4

(A) Representation of the variation in energies for 50 Ω-loop conformations of PenA and all of the purified variants. Based on the LOOPER algorithm and energetic ranking after refinement and optimization, PenA's Ω-loop initial structure has a very unfavorable energy (136 kcal/mol). The L169A variant's Ω-loop has the most favorable (−28.1 kcal/mol) starting conformation for Ω-loop. (B) Dendrograms of PenA and L169A Ω-loop predicted structures. The cluster of the Ω-loop models according the pairwise main-chain RMS deviation (RMSD) of the loop region with predicted structures show that the starting structure (red line) for PenA, and the optimal structure (red arrow), are very different (from 3.3 Å to 1 Å) and energetically unfavorable. For the L169A β-lactamase variant, the best conformation (red arrow) and the initial one (red line) are very similar in terms of energy, and they belong to a unique cluster with a 3.5-Å RMSD for the main chains. (C) The apo-enzyme of PenA (yellow) and the L169A (mixed) showing the Ω-loop at the beginning of MDS (top) and at 100 ps and positioning of the deacylation water molecules (W1 for PenA [dark blue] and W2 for the L169A variant [cyan]). Hydrogen bonding interactions are indicated by dashed lines.

FIG 5

(A and B) Conformations of intact ceftazidime in PenA (A) and the L169A variant (B) Michaelis-Menten complexes. (C and D) Michaelis-Menten complexes of PenA (C) and the L169A variant (D) with ceftazidime with deacylation water molecule (W). Hydrogen bonding interactions are indicated by dashed lines.

FIG 6

Acyl-enzyme complex of PenA (A) and the L169A variant (B) with ceftazidime and PenA (C) and the L169A variant (D) with ceftazidime minus its R2 side chain (bottom). Hydrogen bonding interactions are indicated by dashed lines.