Antibiotic susceptibility of clinical Burkholderia pseudomallei isolates in northeast Thailand during 2015-2018 and the genomic characterization of β-lactam-resistant isolates

Abstract

Melioidosis is an often fatal infection in tropical regions caused by an environmental bacterium, Burkholderia pseudomallei Current recommended melioidosis treatment requires intravenous β-lactam antibiotics such as ceftazidime (CAZ), meropenem (MEM) or amoxicillin-clavulanic acid (AMC) and oral trimethoprim-sulfamethoxazole. Emerging antibiotic resistance could lead to therapy failure and high mortality. We performed a prospective multicentre study in northeast Thailand during 2015-2018 to evaluate antibiotic susceptibility and characterize β-lactam resistance in clinical B. pseudomallei isolates. Collection of 1,317 B. pseudomallei isolates from patients with primary and relapse infections were evaluated for susceptibility to CAZ, imipenem (IPM), MEM and AMC. β-lactam resistant isolates were confirmed by broth microdilution method and characterized by whole genome sequence analysis, penA expression and β-lactamase activity. The resistant phenotype was verified via penA mutagenesis. All primary isolates were IPM-susceptible but we observed two CAZ-resistant and one CAZ-intermediate resistant isolates, two MEM-less susceptible isolates, one AMC-resistant and two AMC-intermediate resistant isolates. One of 13 relapse isolates was resistant to both CAZ and AMC. Two isolates were MEM-less susceptible. Strains DR10212A (primary) and DR50054E (relapse) were multi-drug resistant. Genomic and mutagenesis analyses supplemented with gene expression and β-lactamase analyses demonstrated that CAZ-resistant phenotype was caused by PenA variants: P167S (N=2) and penA amplification (N=1). Despite the high mortality rate in melioidosis, our study revealed that B. pseudomallei isolates had a low frequency of β-lactam resistance caused by penA alterations. Clinical data suggest that resistant variants may emerge in patients during antibiotic therapy and be associated with poor response to treatment.

FIG 1

Flow chart for identification and antibiotic susceptibility testing of Burkholderia pseudomallei at nine hospitals in northeast Thailand and at the Faculty of Tropical Medicine, Mahidol University (FTM).

FIG 2

Gene duplication and amplification (GDA) in clinical Burkholderia pseudomallei strain DR50054E. The 22-kb amplified regions in chromosome 2 of relapse strain DR50054E involve 20 genes from BPSS0944, BPS_RS23855 to BPSS0960, and BPS_RS23950 (red line) (see Table S3 in the supplemental material), including penA (purple line). The average sequence read coverage of this GDA region is approximately 6-fold higher than the average genome coverage.

FIG 3

Construction of penA mutants in Burkholderia pseudomallei strain DR10212A. (A) Construction of DR10212AΔpenA and DR10212ΔpenA::penA^K96243. penA is also known as BPSS0946 (BPS_RS23870) (pink). The gene located upstream of penA is BPSS0945 (BPS_RS23865) (green). The downstream genes are NCBI recently annotated hypothetical gene BPS_RS23875 (dark blue) and BPSS0948 (BPS_RS238800) (yellow). In step 1, penA encoding P167S (light blue) was knocked out from parental strain DR10212A and replaced with wild-type penA from K96243 in step 2 to produce DR10212AΔpenA::penA^K96243. Briefly, fragments containing the desired penA deletion and insertion sequences were synthesized. The plasmid vectors containing the cloned regions of homology allowed the exchange and integration of desired fragments into the chromosome by homologous recombination between cloned and chromosomal sequence. Multiple counterselection markers were applied using kanamycin and sucrose to finally obtain a desired clone. (B) The correct size of the products was confirmed by PCR using primer penA_muta (left) and penA1 (right). Lanes: M, 100-bp plus ladder; 1, DR10212AΔpenA; 2, DR10212AΔpenA::penA^K96243; 3, K96243; 4, sterile water. The gels were sliced for labeling purposes. (C) The MICs of CAZ for DR10212A (orange), DR10212AΔpenA (black), and DR10212AΔpenA::penA^K96243 (purple) were 128 μg/ml (R), 2 μg/ml (S), and 4 μg/ml (S), respectively. After complementation with wild-type penA from K96243, the MIC of MEM dropped from 8 μg/ml (DR10212A, LS) to 2 μg/ml (DR10212AΔpenA and DR10212AΔpenA::penA^K96243, S). The MICs of all isolates toward AMC and SXT remained unchanged, showing AMC-S and SXT-R phenotypes.

FIG 4

Growth curves of Burkholderia pseudomallei isolates cultured in LB broth.

FIG 5

Expression of penA and β-lactamase activity of clinical Burkholderia pseudomallei isolates and laboratory-generated penA mutants. (A) RT-PCR analysis was performed to compare the levels of penA expressed. The strains used were K96243 (CAZ-S), DR10212A (CAZ-R), DR80110A (CAZ-I), and primary-relapse pair DR50054A (CAZ-S) and DR50054E (CAZ-R). Briefly, bacteria cells were grown to mid-log phase in LB broth. Equal portions of the cultures remained untreated or were treated with a subinhibitory concentration of ceftazidime at 32 μg/ml. Total RNA extraction was conducted on bacterial cells at mid-log phase. The penA gene expression levels shown were after normalization with 16S rRNA. The error bars indicate the standard deviations between results of two biological replicates. (B) PenA activity for isolates DR10212A, DR80110A, DR50054A, DR50054E, and DR10212AΔpenA::penA^K96243 grown in LB broth. The enzyme activity was analyzed using nitrocefin as the reporter substrate. The data were analyzed based on two individual experiments conducted on separate days. The positive and negative controls were K96243 and DR10212AΔpenA, respectively. PenA β-lactamase activity units were derived by subtracting the activity observed with DR10212AΔpenA.