Mutator Bacteria as a Risk Factor in Treatment of Infectious Diseases

We show in a gnotobiotic mouse model that, in addition to directselection of antibiotic-resistant bacteria, some antibiotictreatments also select for mutator alleles. Because of thesemutator alleles' high mutation rates, the initial treatmentfailure increases the probability of failures in subsequenttreatments with other drugs.

INTRODUCTION

Top

Introduction

The emergence of antibiotic resistance during therapy can increasethe rate of secondary bacteremia, hospitalization costs, andmortality (3). Bacterial resistance can be acquired either bychromosomal mutations or by horizontal transfer of plasmid-borneresistance genes. In sensitive, strictly clonal bacterial populations(plasmid bearing or not), the generation of antibiotic resistancedepends on the rate of emergence of resistant mutants, i.e.,on the bacterial mutation rate (8, 9). A correlation betweenhigh mutation rate and antibiotic resistance has been reportedin the case of Pseudomonas aeruginosa isolated from the lungsof cystic fibrosis patients (11). We used a model of bacterialcolonization of germfree mice to assess in vivo the risk ofemergence of antibiotic-resistant mutants due to mutator bacteriain the course of antibiotic treatment.

Animals and bacteria.

Top

Animals and Bacteria.

Adult germfree mice (C3H/He/Jax./Jouy-en-Josas) were rearedin isolators (La Calhène, Vélizy, France). Inoculationwas performed per os with 0.5 ml of an inoculum containing wild-typestrain Escherichia coli MG1655 (2) or its isogenic mutator variant(6) that bears a mutation in the mutS gene that inactivatesthe methyl-directed mismatch repair (MMR) system (5) and thatleads to a 100-fold increase in the mutation rate.

Antibiotic treatment design and bacterial count.

Top

Antibiotic treatment design and...

On the first day of treatment, 0.5 ml of an antibiotic(s) solutionwas administered per os to mice. Their drinking water was thensupplemented with the antibiotic(s) until the end of the experiment.Concentrations were 30 µg/ml for fosfomycin (FOF), 80µg/ml for spectinomycin (SPT), 100 µg/ml for streptomycin(STR), and 500 µg/ml each for nalidixic acid (NAL) andampicillin (AMP). The evolution of the bacterial populationwas monitored before and during the treatment by plating dilutionsof feces on Luria-Bertani plates supplemented with antibiotics,when needed (STR at 100 µg/ml, NAL at 40 µg/ml,SPT at 80 µg/ml, FOF at 30 µg/ml, AMP at 500 µg/ml,and rifampin at 100 µg/ml). Because rifampin was not usedin the treatments, the frequency of spontaneously occurringrifampin-resistant mutants [f(Rif^r)] reflects the bacterialmutation rate (6).

Emerging mutator characterization.

Top

Emerging mutator...

f(Rif^r) was obtained in three independent measurements for atleast seven clones randomly isolated from fecal populationsfrom each mouse. Plasmids (19) carrying the wild-type allelesof the different genes of the MMR system were used to electrotransformone clone from each population exhibiting a high mutation rate.Transformants were tested for a reduction in f(Rif^r).

We inoculated the guts of germfree mice with wild-type strainE. coli MG1655 (2) or its isogenic mutator variant (6). Suchmutator bacteria have been found at high frequencies among differentpathologies (7, 10, 11, 13). A few days after inoculation, thebacterial population size reached 10¹⁰ bacteria per gram offeces, the maximum population size reached when E. coli is inoculatedinto axenic mice. In this animal model, the bacterial populationis clonal and readily reaches a high density (6), mimickingthe infection conditions of normally sterile sites (e.g., urinarytract, the lungs, and surgical sites). After a delay of 13 daysto allow the establishment of the inoculated strain, the bacterialpopulation was challenged with various antibiotic treatments(Table 1). The elimination (reduction below the detection limit,e.g., 10³ bacteria per of feces) of the bacterial populationfrom the gastrointestinal tract (decontamination) or the maintenanceof an antibiotic-resistant bacterial population was monitoredas an indicator of the efficiency of the treatment.

View this table:
[in this window]
[in a new window]
TABLE 1. Impact of bacterial mutation rate on the result of antibiotic therapies^a

The administration of FOF and SPT at a 1-day interval (FOF thenSPT) successfully eliminated the bacteria in four of the sixmice inoculated with the wild-type strain. The treatment wasunsuccessful in the other two mice (mice M1 and M2; Table 1),in which bacteria resistant to both antibiotics reached themaximum population size soon after the beginning of the treatment.The same protocol failed to decontaminate the digestive tractof each of the six mice inoculated with the mutator strain (Table1).

When mice received FOF and SPT simultaneously, the bacterialpopulation was eliminated in 10 of 12 mice inoculated with thewild-type strain (Table 1). In the other two mice (mice M3 andM4), the bacterial population sizes were only transiently diminishedby the treatment. The same protocol failed in 10 of 13 miceinoculated with the mutator strain (Table 1). However, the simultaneoususe of FOF, SPT, and STR efficiently eliminated the mutatorpopulations from all of the mice (Table 1). Interestingly, thesingle use of AMP at a high concentration efficiently eliminatedthe mutator bacteria (Table 1), presumably because, in the strainused, no simple mutational events can generate resistance toa high AMP concentration. However, it was previously shown invitro that high-level resistance could be obtained in a mutatorbackground if the AMP concentration gradually increases (17).

Clones were isolated from the fecal populations of mice M1,M2, M3, and M4, in which the treatments failed, to measure f(Rif^r).The clones from mouse M1 did not show a significant differencein f(Rif^r) from that of the wild-type ancestor (P = 0.3 by thet test). The clones from mice M2, M3, and M4 had significantlyhigher f(Rif^r) values (on average, 200-, 200-, and 400-foldincreases, respectively; P $<=$ 0.008 by the t test), suggestingthat strains with high mutation rates (emerging mutators [EMs])were selected from the original populations with low mutationrates. The high mutation rates of one EM clone from mouse M3and one EM clone from mouse M4 were lowered when the bacteriawere transformed with a plasmid bearing a wild-type mutS allele,suggesting that these EMs carry a defect in the MMR system.

One clone isolated from the wild-type population that had survivedthe treatment with FOF and SPT at a 1-day interval in mouseM1 and one EM clone isolated from the population that survivedthe same treatment in mouse M2 were each inoculated separatelyinto a new group of six germfree mice. Animals were then treatedwith STR and NAL. This treatment failed to decontaminate allmice inoculated with the EM strain isolated from mouse M2, whereasit was successful against the nonmutator bacterial populationsoriginating from mouse M1 (Table 1). In mice inoculated withthe genetically constructed mutator strain, the same treatmentfailed in seven of nine mice (Table 1).

Our results show that some antibiotic treatments can selectfor mutator bacteria present at low frequencies among all wild-typepopulations. Actually, by selecting for a resistance allele,the antibiotic selective pressure also selected for a mutatorallele as the mechanism that generated the resistance. Moreover,mutators could also facilitate the modification of the activesites of detoxification enzymes to shift the resistance fromresistance to a low dose to resistance to a high dose (17) andextend their resistance spectra (12). It could also rapidlyaccumulate compensatory mutations that limit the cost to thebacteria associated with the resistance alleles (1). Mutatorbacteria can be considered risk markers for antibiotic therapy.If corroborated by epidemiological data, our results would suggestthat, in the case of a first therapeutic failure and if timeallows, a diagnostic assay for the presence of mutators shoulddetermine the next therapeutic step. If a drug is availablefor which mutational events can very rarely generate antibioticresistance, such as AMP at a high concentration used againstthe bacteria in the present study, then it is the first choicefor use against the mutators. Otherwise, the use of a combinationtherapy seems to be the best alternative. As the selection ofmutator bacteria is favored by several bacterial and environmentalfactors (18), some conditions (e.g., a large bacterial population)allow the selection of such strains more than others. In thesecases, if possible, therapies should be initiated directly witha protocol that limits the risk of selection of mutators. Forexample, the use of antibiotics that inhibit a single enzymeshould probably be set aside, even when used in combinationtherapies, as we did in the present work.

Due to their increased genetic adaptability, the rate of selectionof mutator variants among populations undergoing multiple adaptivesteps should be increased (6, 8, 14, 16, 18). For this reason,some pathology might allow the emergence and fixation of mutatoralleles more frequently than others (4). In the case of chronicinfections, the size of the infecting population combined withthe duration of infection, the reiterated challenges imposedby antibiotic therapies, and the host's immune response shouldprovide conditions prone to the fixation of mutator alleles.Numerous mutator E. coli, Salmonella enterica, and P. aeruginosaisolates from patients and other natural environments have defectiveMMR systems (6, 7, 10, 11), similar to the emerging mutatorstrains isolated in the present work. Given the abundance ofmutators among bacterial pathogens, viruses, and tumors (15),the concept that one failed therapy is a potential risk factorfor the next therapy might be relevant to other therapeuticstrategies.

ACKNOWLEDGMENTS

This work was supported by the Ligue contre le Cancer, Associationpour la Recherche contre le Cancer, the Programme Environnementet Santé (MATE), the Programme de Recherche Fondamentaleen Microbiologie et Maladies Infectieuses et Parasitaires (MENRT),and Fondation pour la Recherche Médicale.

We are grateful to A. Clara and E. Morin for technical helpand to D. Brégeon, E. Denamur, M. S. Fox, M. P. Francino,A. Gomez, S. Delmas, L. Le Chat, F. W. Stahl, E. Stewart, O.Tenaillon, and anonymous referees for comments on the manuscript.

FOOTNOTES

* Corresponding author. Mailing address: E9916, Institut National de la Santé et de la Recherche Médicale, Faculté de Médecine "Necker-Enfants Malades," Université René Descartes Paris V, 156 rue de Vaugirard, 75015 Paris, France. Phone: 33 (0) 1 40 61 53 23. Fax: 33 (0) 1 40 61 53 24. E-mail: taddei{at}necker.fr.

Dedicated to the memory of J. Bjorkman.

REFERENCES

Top