María M. Tavío

Review on Bacterial Stress Topics

Abstract:  A complex network of regulatory systems ensures a coordinated and effective response to different types of stress that can act on a bacterium. Bacterial stress response generates changes that influence efflux system and virulence factor expression. Thus, partial or total loss of pathogenicity islands in uropathogenic Escherichia coli can be induced by SOS‐dependent or SOS‐independent pathways related to selection of quinolone‐resistant mutants. Likewise, hyperosmolarity and some chemicals, including fluoroquinolones, salicylate, nonantimicrobial medicaments like diazepam and anti‐inflammatory drugs are all able to induce an increased active efflux, cyclohexane tolerance, loss of porins, and decreased susceptibility to multiple antimicrobials in enterobacterial strains, suggesting that bacterial response to the stress caused by an increase in osmolarity might be linked to the development of the multidrug‐resistant phenotypes. Finally, a sudden downshift of the growth temperature (cold‐shock) triggers a drastic reprogramming of bacterial gene expression to allow cell survival under the new unfavorable conditions. The strategy developed by E. coli to reach this goal consists in the induction of a set of (cold‐shock) genes whose expression is regulated at both transcriptional and posttranscriptional levels.

Quorum-sensing regulator sdiA and marA overexpression is involved in in vitro-selected multidrug resistance of Escherichia coli

OBJECTIVES The role of sdiA in the acquisition of low-level multidrug resistance (MDR) was analysed and compared with that of marA and soxS in two Escherichia coli clinical isolates and two in vitro-selected mutants. METHODS The mutants were developed by growth in lomefloxacin and ceftazidime. The sdiA, marA, soxS, ftsI, tolC and acrB gene transcript levels were determined by RT-PCR. Analyses of 2,4-dinitrophenol susceptibility, the effect of an active efflux inhibitor on antibiotic and mitomycin C susceptibility, beta-lactamase hydrolytic activity, outer and inner membrane proteins and acrR gene sequencing were also performed. RESULTS Both mutants showed elevated marA and sdiA gene transcript levels, which were associated with increased susceptibility to 2,4-dinitrophenol; soxS overexpression was only seen in the mutant selected with ceftazidime. The two mutants showed MDR phenotypes in which ceftazidime, cefpirome and aztreonam MICs increased 4- to 128-fold, in addition to decreased susceptibility to quinolones, chloramphenicol and mitomycin C. The highest ceftazidime MIC in one of the mutants coincided with a frameshift mutation in acrR and the highest transcript level of ftsI (penicillin-binding protein 3), but not with a higher beta-lactamase activity. Likewise, active efflux associated with increased levels of acrB and tolC and decreased OmpF expression contributed to low-level MDR in both mutants. CONCLUSIONS marA and sdiA overexpression was a common feature of multidrug-resistant mutants selected by growth in lomefloxacin and ceftazidime. To our knowledge, this report is the first to describe in vitro selection with a fluoroquinolone or ceftazidime triggering sdiA overexpression in E. coli isolates.

Mechanisms involved in quinolone resistance in Mycoplasma mycoides subsp. capri

Mycoplasma mycoides subsp. capri is a causative agent of contagious agalactia in goats. In this study, M. mycoides subsp. capri mutants were selected for resistance to fluoroquinolones (norfloxacin, enrofloxacin and ciprofloxacin) by serial passes in broth with increasing concentrations of antibiotic. Mutations conferring cross-resistance to the three fluoroquinolones were found in the quinolone resistance determining regions of the four genes encoding DNA gyrase and topoisomerase IV. Different mutations in the DNA gyrase GyrA subunit suggest a different mechanism of inhibition between norfloxacin and the other tested fluoroquinolones. The presence of an adenosine triphosphate-dependent efflux system was suggested through the use of the inhibitor orthovanadate.

Decreased permeability and enhanced proton-dependent active efflux in the development of resistance to quinolones in Morganella morganii.

Five quinolone-resistant strains were developed from a clinical Morganella morganii isolate (M1 strain) which was susceptible to nalidixic acid and fluoroquinolones. All five in vitro selected mutants showed diminished norfloxacin accumulation and two of them also decreased the expression of the AgO in the outer membrane lipopolysaccharide with respect to their parent strain and to the M. morganii NCTC-235 type strain. Likewise, the M1 strain did not express a 37-38 kDa protein and accumulated less norfloxacin than the M. morganii NCTC-235 strain. The decreased norfloxacin uptake in the five mutants compared with the M. morganii NCTC-235 strain was due to an enhanced proton-dependent active efflux plus a pre-existent decreased expression of a 37-38 kDa protein in the parent strain.

Resistance to ceftazidime in Escherichia coli associated with AcrR, MarR and PBP3 mutations and overexpression of sdiA.

The mechanisms responsible for the increase in ceftazidime MIC in two Escherichia coli in vitro selected mutants, Caz/20-1 and Caz/20-2, were studied. OmpF loss and overexpression of acrB, acrD and acrF that were associated with acrR and marR mutations and sdiA overexpression, together with mutations A233T and I332V in FtSI (PBP3) resulted in ceftazidime resistance in Caz/20-2, multiplying by 128-fold the ceftazidime MIC in the parental clinical isolate PS/20. Absence of detectable β-lactamase hydrolytic activity in the crude extract of Caz/20-2 was observed, and coincided with Q191K and P209S mutations in AmpC and a nucleotide substitution at -28 in the ampC promoter, whereas β-lactamase hydrolytic activity in crude extracts of PS/20 and Caz/20-1 strains was detected. Nevertheless, a fourfold increase in ceftazidime MIC in Caz/20-1 compared with that in PS/20 was due to the increased transcript level of acrB derived from acrR mutation. The two Caz mutants and PS/20 showed the same mutations in AmpG and ParE.

In Vitro Selection and Characterization of Mutants in TEM-1-Producing Escherichia coli by Ceftazidime and Ceftibuten

Abstract The present work was undertaken to study the ability of ceftazidime and ceftibuten to select In Vitro resistant mutants of Escherichia coli HB101 harboring blaTEM-1 β-lactamase gene. Minimum inhibitory concentrations (MICs) of ceftazidime and ceftibuten were increased by a factor of 32, overcoming in the case of ceftazidime the breakpoint for clinical resistance. Outer membrane protein analysis and PCR for bla TEM alleles revealed that ceftazidime and ceftibuten select for different resistance mechanisms. Ceftazidime created mutants that encode an extended-spectrum β-lactamase (TEM-12) and exhibit decreased expression of OmpF. Ceftibuten was unable to select for extended-spectrum β-lactamase expressing mutants but reduced the expression of two porins, OmpC and OmpF. The stability of ceftibuten to hydrolysis and the difference in the structure of these β-lactam antibiotics could be responsible for the selection of different mechanisms of resistance.

Overexpression of the quorum-sensing regulator sdiA and soxS is involved in low-level multidrug resistance induced in Escherichia coli AG100 by haloperidol, diazepam and NaCl.

The authors would like to express their thanks to the investigaors, research staff and participants of this multicentre study. They cknowledge writing support from Dr Rajeeb Ghosh and Dr Payal hardwaj (Novartis Healthcare Pvt. Ltd.) in the preparation of this anuscript as well as Dr Marcin Mankowski (Chiron Corporation imited) in the preparation of the original study protocol. They lso acknowledge the support of Kamel Bouylout, Sabine Daugeat and Beate Kirchherr in trial logistics. List of investigators: Drs. retschmer, Bauerschmitz, Stege, Fierlbeck, Zouboulis, Staubach, ollnick, Dissemond, Welzel, Rodloff, Josten, Just, Kujath, Shekarriz, irth, Roger, Bressieux, Lucht, Crawford, Seaton, Lecha, Garciaechuz, Amado, Abril, Dabrowiecki, Checinski, Szyber, Rowinski, ziki, Jaworski, Sosnowski. Funding: This study was funded by a research grant from Novaris. Competing interests: SRQ and GF have received research grant rom Novartis for this study and have no other conflict of intersts. RAS has received research grant from Novartis for this study nd research funding for investigator-led studies (Novartis and fizer) and honoraria for speaking at symposia and advisory boards Novartis and Pfizer). JF is an employee of Novartis Pharma AG, and wns shares of Novartis Pharma AG. RLC is an employee of Novartis harma AG and has no other conflict of interests. Ethical approval: The study was approved by the institutional eview board or independent ethics committee of each participatng center.

Resolution of high-molecular-weight components in lipopolysaccharides of Escherichia coli, Morganella morganii, Citrobacter freundii and Citrobacter diversus strains with sodium dodecyl sulfate polyacrylamide gels

The use of 0.5% sodium dodecyl sulfate in polyacrylamide separation gels allowed the resolution in several bands of high-molecular-mass components in smooth lipopolysaccharide of bacterial outer membrane from Escherichia coli, Morganella morganii, Citrobacter freundii and Citrobacter diversus. With or without 0.1% SDS, however, such a result was not possible.

Four anti-inflammatory drugs induce tolC gene overexpression and repression of OmpF synthesis, enhance active efflux and decrease susceptibility to antibiotics in Escherichia coli AG100 strain.

To the Editor The expression of efflux systems is closely linked to bacterial stress response; regulators of this bacterial response trigger the expression of TolC that is the common outer membrane channel for most of the efflux systems identified in Escherichia coli (1). We previously reported the induction of multidrug resistance (MDR) phenotypes in E. coli AG100 by diazepam, haloperidol, and NaCl (2) such as it had been also described for salicylates including the anti-inflammatory drug acetyl salicylate (3). The present study assessed the induction of MDR phenotypes in the wild-type E. coli AG100 strain by four anti-inflammatory drugs, dexamethasone that is a steroidal anti-inflammatory drug (Merck, Madrid, Spain), and three non-steroidal antiinflammatory drugs, diclofenac (Novartis, Barcelona, Spain), ketorolac (Roche, Madrid, Spain), and methamizol (Boehringer Ingelheim, San Cugat del Vall es, Spain). Concentrations equal to 1/512, 1/256, 1/128, 1/ 64, 1/32, 1/16, 1/8, and 1/4 of minimum inhibitory concentrations (MICs) of diclofenac (4 mM), ketorolac (16 mM), methamizol (640 mM), and dexamethasone (5 mM) in AG100 were assayed for the induction of OmpF loss and increasing antimicrobial MICs and cyclohexane tolerance as described (2). AG112 (an OmpF-deficient strain) was used as control strain. The optimal induction of the cyclohexane tolerance and OmpF loss with the assayed anti-inflammatory drugs was obtained using onequarter of their MICs in AG100; therefore, only concentrations equal to one-quarter MIC of each assayed drug in the AG100 strain were tested to assess tolC expression level by reverse transcription of total RNA and PCR of cDNA (RT-PCR) and to measure norfloxacin uptake with and without 50 and 100 lM carbonyl cyanide m-chlorophenylhydrazone (CCCP) from Sigma (Madrid, Spain) as described (2). The levels of expression of tolC that were induced by NaCl (342 mM) and salicylate (5 mM) were used as controls. The band densities on gels were analyzed using ImageJ software (National Institutes of Health, Bethesda, MD, USA) . Inducing concentrations ≥1/32 MICs of diclofenac and ketorolac and ≥1/128 fold MICs of methamizol and dexamethasone, but not lower concentrations, decreased OmpF expression in AG100, acquiring the same OMP profile as the AG112 strain (Figs 1A, B and C). Likewise, 1/64–1/4 MIC of each assayed anti-inflammatory medicament increased 2–8 fold MIC of cefoxitin (Merck-Sharp & Dohme, Madrid, Spain), 2–16 fold MIC of nalidixic acid, 2–32 fold MIC of norfloxacin, 2–64 fold MIC of tetracycline, and 2–4 fold MIC of chloramphenicol in the AG100 strain. Norfloxacin, chloramphenicol, tetracycline, and nalidixic acid were purchased from Sigma. Nevertheless, 5 mM methamizol and 0.039 mM dexamethasone (concentrations equivalent to 1/128 of their MICs in the AG100 strain) only increased 2–4 fold MICs of the assayed antibiotics in AG100, concurring with the effect of 0.156 mM salicylate in AG100, whereas 0.031 mM diclofenac and 0.125 mM ketorolac did not have effect on AG100 susceptibility to antibiotics. Concentrations <1/128 MIC of each one of the four anti-inflammatory drugs in the AG100 strain did not change the susceptibility of the same strain to the tested antibiotics. The simultaneous OmpF loss and decrease in susceptibility to antibiotics that were induced in a concentrationdependent manner by the assayed drugs were also previously described under the inducer effect of diazepam, haloperidol, and NaCl. OmpF expression is dependent on changes in medium osmolarity and regulated by two-component systems as EnvZOmpR or CpxA-CpxR that result in ompF repression and ompC activation (4), although OmpC overexpression was not seen under induction with anti-inflammatory drugs (Figs 1A, B and C). Decreased norfloxacin uptake was observed when the AG100 strain was grown in the presence of any tested inducer, particularly dexamethasone induced the highest decreases in norfloxacin uptake in AG100 (Fig. 1B) such as it was described for 5 mM salicylate and NaCl (2). When 50 or 100 lM CCCP was added to dissipate the proton motive force, norfloxacin uptake increased 2–3-fold in the induced AG100 strain, as described for salicylate, diazepam, haloperidol, and NaCl (2). The increase in norfloxacin uptake by CCCP in MDR phenotypes induced by any tested anti-inflammatory drug was concomitant with an increase of 2–3.2fold in cyclohexane tolerance, in the first 3 h after cyclohexane addition, compared with that observed in the non-induced AG100 strain. Increases in cyclohexane tolerance 1.9-2.6-fold and 1.4 –2.1fold were also seen with concentrations of 1/8 MIC and 1/128–1/32 MIC of anti-inflammatory drugs, respectively. Cyclohexane tolerance is dependent on

Interaction of carbapenems and β-lactamase inhibitors towards CTX-M-15 and CTX-M-15G238C mutant

OBJECTIVES The aim of this study was to evaluate the role of residue 238 in CTX-M-15 and CTX-M-15G238C mutant with respect to carbapenems and various β-lactamase inhibitors. METHODS A CTX-M-15G238C laboratory mutant was generated by site-directed mutagenesis from CTX-M-15 enzyme by replacing glycine 238 with cysteine. Thiol titration and p-chloromercuribenzoate (PCMB) inactivation assays were used to ascertain the presence of a disulfide bridge in the active site of CTX-M-15G238C. Kinetic parameters were determined both for CTX-M-15 and CTX-M-15G238C enzymes by analysing either the complete hydrolysis time courses or under initial rate conditions. RESULTS In CTX-M-15G238C mutant, the two cysteines (C69 and C238) located in the enzyme active site were unable to form a disulfide bridge. CTX-M-15 and thermostable CTX-M-15G238C were used to study the kinetic interaction with carbapenems, which behaved as poor substrates for both enzymes. Meropenem and ertapenem acted as transient inactivators for CTX-M-15 and CTX-M-15G238C, and for these compounds the variation of kobs versus the inactivator concentration was linear. Imipenem behaved as a transient inactivator for CTX-M-15 and as an inactivator (with k+3=0) for CTX-M-15G238C. In any case, the k+2/K values for CTX-M-15G238C were higher than those for CTX-M-15. CONCLUSIONS Compared with CTX-M-15, CTX-M-15G238C mutant appears to have a more favourable conformation for carbapenem acylation and higher activity against cefotaxime, which could be due to the presence of free -SH groups in the enzyme active site.