Pseudomonas aeruginosa ceftolozane-tazobactam resistance development requires multiple mutations leading to overexpression and structural modification of AmpC

Abstract

We compared the dynamics and mechanisms of resistance development to ceftazidime, meropenem, ciprofloxacin, and ceftolozane-tazobactam in wild-type (PAO1) and mutator (PAOMS, ΔmutS) P. aeruginosa. The strains were incubated for 24 h with 0.5 to 64× MICs of each antibiotic in triplicate experiments. The tubes from the highest antibiotic concentration showing growth were reinoculated in fresh medium containing concentrations up to 64× MIC for 7 consecutive days. The susceptibility profiles and resistance mechanisms were assessed in two isolated colonies from each step, antibiotic, and strain. Ceftolozane-tazobactam-resistant mutants were further characterized by whole-genome analysis through RNA sequencing (RNA-seq). The development of high-level resistance was fastest for ceftazidime, followed by meropenem and ciprofloxacin. None of the mutants selected with these antibiotics showed cross-resistance to ceftolozane-tazobactam. On the other hand, ceftolozane-tazobactam resistance development was much slower, and high-level resistance was observed for the mutator strain only. PAO1 derivatives that were moderately resistant (MICs, 4 to 8 μg/ml) to ceftolozane-tazobactam showed only 2 to 4 mutations, which determined global pleiotropic effects associated with a severe fitness cost. High-level-resistant (MICs, 32 to 128 μg/ml) PAOMS derivatives showed 45 to 53 mutations. Major changes in the global gene expression profiles were detected in all mutants, but only PAOMS mutants showed ampC overexpression, which was caused by dacB or ampR mutations. Moreover, all PAOMS mutants contained 1 to 4 mutations in the conserved residues of AmpC (F147L, Q157R, G183D, E247K, or V356I). Complementation studies revealed that these mutations greatly increased ceftolozane-tazobactam and ceftazidime MICs but reduced those of piperacillin-tazobactam and imipenem, compared to those in wild-type ampC. Therefore, the development of high-level resistance to ceftolozane-tazobactam appears to occur efficiently only in a P. aeruginosa mutator background, in which multiple mutations lead to overexpression and structural modifications of AmpC.

FIG 1

Dynamics of resistance development to ceftolozane-tazobactam and comparators. The modal values for three experiments are shown.

FIG 2

Numbers of genes showing modified expression in the studied ceftolozane-tazobactam resistant mutants compared with wild-type PAO1 in each of the 27 established functional categories. Negative numbers indicate genes with decreased expression, and positive numbers indicate genes with increased expression.

FIG 3

In vitro competition assays to assess the fitness cost associated with the development of resistance to ceftolozane-tazobactam (TOL-TAZ), ceftazidime (CAZ), meropenem (MER), and ciprofloxacin (CIP) in the three day 7 PAO1 and PAOMS mutants described in Table 1. Error bars indicate standard deviation.

FIG 4

Penicillin-binding protein (PBP) profiles of ceftolozane-tazobactam-resistant mutants obtained from PAO1 and PAOMS strains.