The individual contributions of blaB, blaGOB and blaCME on MICs of β-lactams in Elizabethkingia anophelis

Abstract Background The blaB, blaGOB and blaCME genes are thought to confer β-lactam resistance to Elizabethkingia anophelis, based on experiments conducted primarily on Escherichia coli. Objectives To determine the individual contributions of β-lactamase genes to increased MICs in E. anophelis and to assess their impact on the in vivo efficacy of carbapenem therapy. Methods Scarless gene deletion of one or more β-lactamase gene(s) was performed in three clinical E. anophelis isolates. MICs were determined by broth microdilution. Hydrolytic activity and expressions of β-lactamase genes were measured by an enzymatic assay and quantitative RT–PCR, respectively. In vivo efficacy was determined using Galleria mellonella and murine thigh infection models. Results The presence of blaB resulted in >16-fold increases, while blaGOB caused 4–16-fold increases of carbapenem MICs. Hydrolysis of carbapenems was highest in lysates of blaB-positive strains, possibly due to the constitutionally higher expression of blaB. Imipenem was ineffective against blaB-positive isolates in vivo in terms of improvement of the survival of wax moth larvae and reduction of murine bacterial load. The deletion of blaB restored the efficacy of imipenem. The blaB gene was also responsible for a >4-fold increase of ampicillin/sulbactam and piperacillin/tazobactam MICs. The presence of blaCME, but not blaB or blaGOB, increased the MICs of ceftazidime and cefepime by 8–16- and 4–8-fold, respectively. Conclusions The constitutionally and highly expressed blaB gene in E. anophelis was responsible for increased MICs of carbapenems and led to their poor in vivo efficacy. blaCME increased the MICs of ceftazidime and cefepime.


Introduction
Nosocomial outbreaks of Elizabethkingia anophelis infections have been increasingly reported worldwide.E. anophelis isolates are highly resistant to multiple commonly used antibiotics, including β-lactams. 1 β-Lactam resistance in E. anophelis is assumed to result from two intrinsic MBLs (bla GOB and bla B ) and one ESBL (bla CME ).Most studies measured the contribution of these genes to β-lactam resistance in Escherichia coli (Table S1, available as Supplementary data at JAC Online), but not in E. anophelis.In this study, deletions of individual β-lactamase genes were performed in E. anophelis to determine their contributions to in vitro resistance to commonly used β-lactams and to assess their impact on in vivo efficacy of carbapenem therapy.

Bacterial isolates for scarless gene deletion
Three isolates (2018C07-210, 2002N07-090 and 2008N05-106) of different clonality from a nationwide surveillance programme (Taiwan Surveillance of Antimicrobial Resistance) were selected.2018C07-210 and 2002N07-090 both harboured bla B-1 , bla GOB-20 and bla CME-3 , while 2008N05-106 harboured bla B-3 , bla GOB-3 and bla CME-3 .Scarless gene deletion of bla GOB , bla B and/or bla CME was performed as described in our previous report. 2The suicide vector pUT-ermF, with an erm(F) gene encoding erythromycin resistance and a sacB gene for sucrose counter-selection was used for scarless deletion.The pUT-ermF containing the upstream and downstream regions of the target gene was amplified in E. coli S17-λpir, which was then conjugated with E. anophelis.Transconjugants were screened first in LB agar with 250 mg/L erythromycin and then in LB agar with 10% sucrose and without NaCl.Seven mutants were obtained from each parent strain (Table 1).Mutant and parent strains were subjected to broth microdilutions 3 of various β-lactams and levofloxacin (as a control) to determine changes of MICs.

β-Lactamase activity
Crude extracts of bacteria at stationary phase were prepared by disruption of bacteria by sonification (2 min at 2 s of sonication and 4 s of rest at 30% amplitude) with a Vibra-Cell VCX 600 (Sonics & Materials, Inc.) and centrifugation for 10 min at 15 000 g at 4°C. 4 The crude extracts of different strains with the same protein concentration were mixed with β-lactams in phosphate buffer (0.1 M, pH 7) with 50 mM ZnSO 4 at room temperature.The total protein concentration was measured by a BCA protein assay.The hydrolysis of antibiotics was monitored by SpectraMax ® M2 (Molecular Devices) for 10 min.The molar extinction coefficients (Δɛ) were as follows: Δɛ 297 = 8621 M −1 cm −1 for imipenem and Δɛ 297 = 5900 M −1 cm −1 for meropenem.

Real-time quantitative RT-PCR (qRT-PCR)
Expression levels of β-lactamase genes and 16S rRNA in E. anophelis strains were measured by qRT-PCR.Bacteria at mid-log phase were incubated in LB broth with or without imipenem (½ of MIC).The purification of RNA, removal of genomic DNA, and conversion to cDNA were conducted as described in a previous report. 5Samples were subjected to qRT-PCR with the reaction mixture containing Luminaris Color HiGreen qPCR Master Mix (Thermo Scientific).Cycling parameters were as follows: 1 cycle of 95°C for 10 min; 40 cycles of 95°C for 15 s, 60°C for 1 min; and 1 cycle of 95°C for 15 s, 60°C for 1 min, and 95°C for 15 s.Expression level results were normalized to the transcription levels of 16S rRNA.The relative expression level was calculated by dividing the expression level of the target β-lactamase gene by that of bla GOB in the WT strain without the addition of imipenem.

Galleria mellonella model
G. mellonella was reared at the National Health Research Institutes of Taiwan.][8] Briefly, a 10 μL aliquot of E. anophelis containing an inoculum of 10 4 cfu was injected into the haemocoel of each caterpillar via the most caudal left proleg.Imipenem (12.5 mg/kg) was given in 10 μL injections into another proleg.After injection, the caterpillars were incubated in plastic containers at 37°C.The number of dead caterpillars was recorded twice daily for 72 h.

Thigh infection model
Seven-week-old C57BL/6J mice were rendered neutropenic with cyclophosphamide before inoculation, following a previously described protocol. 9A 30 μL aliquot containing E. anophelis (10 7 cfu) was injected into each of the two rear thighs of each mouse.Imipenem (100 mg/kg/day) was administered 2 and 14 h after inoculation.The mice were sacrificed 24 h after inoculation.Bacterial counts were determined by plating the homogenized tissue in serial 10-fold dilutions on brain heart infusion agar.

Results and discussion
Table 1 illustrates that compared with isolates without any known β-lactamase gene (Δbla B Δbla GOB Δbla CME ), those carrying bla B exhibited >16-fold increases of imipenem and meropenem MICs; bla GOB resulted in only 8-16-and 4-8-fold increases, respectively.The bla B gene was also responsible for 32-and 8-32-fold increases of ampicillin/sulbactam and piperacillin/tazobactam MICs, respectively.However, bla GOB had no effect on the MICs of these two An enzymatic assay (Figure 1a) showed that the 10 min hydrolysis rates of imipenem and meropenem were highest in lysates of bla B -positive strains.The deletion of bla B almost eliminated hydrolytic activity.In contrast, bla GOB was only associated with limited hydrolysis of carbapenems.Imipenem did not prolong the survival of wax moth larvae infected with bla B -positive isolates, including WT or bla GOB -deleted strains.Furthermore, imipenem was effective among larvae infected with a bla B -deleted mutant (Figure 1b).A similar result was observed in the murine thigh infection model (Figure 1c).The reduction of bacterial load by imipenem was statistically significant in murine thighs infected by the bla B -deleted mutant, but not by bla B -positive isolates.
We hypothesized that the different resistance phenotypes of bla B -and bla GOB -positive isolates may result from dissimilar expression levels, and therefore determined their expression levels using qRT-PCR (Figure 1d).qRT-PCR showed a significantly higher expression of bla B than bla GOB in the WT strain.The deletion of one gene did not alter the expression of the other.In addition, the expressions of both bla B and bla GOB were unaffected by imipenem exposure.
The putative roles of bla B and bla GOB in increased carbapenem MICs were conflicting in the previous literature (Table S1); most studies showed that both bla B and bla GOB contributed equally to the increase of carbapenem MICs for E. coli.Our results of MIC testing, enzymatic activity assays and in vivo experiments evaluating E. anophelis showed a significant contribution of bla B to high carbapenem MICs that may mask the modest contribution of bla GOB ; this finding may be partly attributable to the higher expression level of bla B .Previous literature (Table S1) disclosed that ceftazidime susceptibility was reduced by all three β-lactamases and associated bla CME with increased aztreonam MICs.However, our study showed that bla CME was the only gene associated with high ceftazidime MICs and had no effect on aztreonam MICs (Table 1).The discrepancy between the results of our study and those of previous reports may be attributable to host factors (E. coli versus E. anophelis), the use of different subtypes of each β-lactamase gene 10 and dissimilar expression levels of β-lactamase genes pertinent to vectors or promoters used in the earlier studies.

Conclusions
The constitutional production of β-lactamase B in E. anophelis hydrolysed carbapenems and led to increased carbapenem MICs and a lack of in vivo efficacy of imipenem.The bla B gene was also responsible for increased MICs of ampicillin/sulbactam and piperacillin/tazobactam.The bla CME gene was associated with increased MICs of ceftazidime and cefepime.
The plasmid for scarless gene editing was a generous gift from Dr Kris Leung Kei Siu, National Health Research Institutes (NHRI).The isolates were from Taiwan Surveillance of Antimicrobial Resistance (TSAR).Resistance mechanism for Elizabethkingia

Figure 1 .
Figure1.Hydrolytic activity on carbapenems, in vivo efficacy of imipenem monotherapy, and relative expression levels of β-lactamase genes in clinical E. anophelis isolates and their gene-edited mutants.(a) Hydrolytic activity against imipenem and meropenem in 2018C07-210, its gene-edited mutants, and Klebsiella pneumoniae harbouring bla KPC (KP-bla KPC ).The y-axis indicates the amount of the carbapenem hydrolysed (nmol) per min and per mg of total protein.(b) Survival of G. mellonella given imipenem monotherapy using 2008N05-106 and its gene-edited mutants.(c) Bacterial loads in mice given imipenem monotherapy in a thigh infection model using 2008N05-106 and its gene-edited mutants.(d) Relative expression levels of β-lactamase genes in 2018C07-210 and its gene-edited mutants determined by qRT-PCR.The relative expression level was calculated by dividing the expression level of the target β-lactamase gene by that of bla GOB in the WT strain without addition of imipenem.IPM, imipenem; MEM, meropenem; ns, not significant; *P < 0.05.

Table 1 .
MICs of different antibiotics in three clinical strains of E. anophelis and their gene-edited mutants Isolates β-lactams and similar correlations between the presence of β-lactamase genes and increased MICs.The MICs of aztreonam and ceftazidime in isolates without bla CME , bla B and bla GOB remained high, indicating the presence of other mechanisms, i.e. low outer membrane permeability, reduced affinity of PBPs, unknown β-lactamases or overexpression of efflux pumps.