Hypermutable Haemophilus influenzae with mutations in mutS are found in cystic fibrosis sputum

Michael E. Watson, Jr1,2, Jane L. Burns3 and Arnold L. Smith1

1 Seattle Biomedical Research Institute, 307 Westlake Ave N, Suite 500, Seattle, WA 98109, USA
2 Department of Molecular Microbiology and Immunology, School of Medicine, University of Missouri-Columbia, Columbia, MO 65212, USA
3 Division of Infectious Diseases, Children's Hospital and Regional Medical Center, 4800 Sand Point Way, Seattle, WA 98105, USA

Correspondence
Arnold L. Smith
arnold.smith{at}sbri.org


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Hypermutable bacterial pathogens exist at surprisingly high prevalence and benefit bacterial populations by promoting adaptation to selective environments, including resistance to antibiotics. Five hundred Haemophilus influenzae isolates were screened for an increased frequency of mutation to resistance to rifampicin, nalidixic acid and spectinomycin: of the 14 hypermutable isolates identified, 12 were isolated from cystic fibrosis (CF) sputum. Analysis by enterobacterial repetitive intergenic consensus (ERIC)-PCR and ribotyping identified eight distinct genetic fingerprints. The hypermutable phenotype of seven of the eight unique isolates was associated with polymorphisms in conserved sites of mutS. Four of the mutant mutS alleles were cloned and failed to complement the mutator phenotype of a mutS : : TSTE mutant of H. influenzae strain Rd KW20. Antibiotic susceptibility testing of the hypermutators identified one {beta}-lactamase-negative ampicillin-resistant (BLNAR) isolate with two isolates producing {beta}-lactamase. Six isolates from the same patient with CF, with the same genetic fingerprint, were clonal by multilocus sequence typing (MLST). In this clone, there was an evolution to higher MIC values for the antibiotics administered to the patient during the period in which the strains were isolated. Hypermutable H. influenzae with mutations in mutS are prevalent, particularly in the CF lung environment, and may be selected for and maintained by antibiotic pressure.


Abbreviations: BLNAR, {beta}-lactamase-negative ampicillin-resistant; CF, cystic fibrosis; ERIC, enterobacterial repetitive intergenic consensus; MLST, multilocus sequence typing; MMR, methyl-directed mismatch repair; PBP, penicillin-binding protein

The GenBank/EMBL/DDBJ accession numbers for the sequences reported in this paper are AY568590, AY568591, AY568592, AY568593, AY568594, AY568595, AY568596 and AY568597.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Haemophilus influenzae is a human-adapted organism existing as a common commensal of the upper respiratory mucosa, and as a pathogen causing otitis media, sinusitis, bronchitis and community-acquired pneumonia. Chronic infection with H. influenzae can occur in patients with a lower airway disease, such as chronic obstructive pulmonary disease (COPD), bronchiechtasis and cystic fibrosis (CF) (Tang et al., 2001). H. influenzae persists within the human respiratory tract by generating diversity through mutational mechanisms that produce antigenic and phase variation. The resulting phenotypic changes facilitate escape from immune detection and permit adaptation to a changing environment (Foxwell et al., 1998; Moxon et al., 1994). An unresolved issue is the prevalence and the contribution to pathogenesis of so-called mutator, or hypermutable, strains of H. influenzae.

Hypermutable isolates of bacterial pathogens have been described among natural populations of Escherichia coli, Salmonella enterica, Pseudomonas aeruginosa, Neisseria meningitidis and Staphylococcus aureus (Bucci et al., 1999; Denamur et al., 2002; LeClerc et al., 1996; Matic et al., 1997; Oliver et al., 2000, 2002; Prunier et al., 2003; Richardson et al., 2002). The majority of mutators among natural bacterial populations have been found to have defects in one or more components of the methyl-directed mismatch repair (MMR) system, particularly mutS (LeClerc et al., 1996; Li et al., 2003; Matic et al., 1997; Oliver et al., 2002). The MMR system is responsible for correcting errors in newly replicated DNA and prevents recombination between homologous DNA sequences (Matic et al., 1995; Modrich & Lahue, 1996; Rayssiguier et al., 1989). Mutants in MMR genes produce a global mutator phenotype with up to 1000-fold increased mutation rates, increased recombination between non-identical DNA sequences and increased rates of frameshift mutations, particularly in microsatellite regions with runs of repetitive sequences (Cox, 1976; Glickman & Radman, 1980; Matic et al., 1995; Schaaper & Dunn, 1987; Strauss, 1999).

The prevalence of global mutators within some populations was estimated to be about 1 % of natural pathogenic and commensal isolates of E. coli and Sal. enterica, and considerably higher within specific conditions (i.e. up to 20 % of P. aeruginosa isolated from the sputum of patients with CF) (LeClerc et al., 1996; Matic et al., 1997; Oliver et al., 2000). It is thought that the reason mutators are not more prevalent is due to the decreased fitness of excess deleterious mutations associated with higher mutation frequencies (Dawson, 1999). However, when a favourable mutation, such as antibiotic resistance, is acquired and selected for, the hypermutable phenotype can stabilize and be propagated by association with the selective mutation (hitch-hiking) (Chao & Cox, 1983; Giraud et al., 2002; Taddei et al., 1997a). This has been demonstrated in a gnotobiotic mouse model in which antibiotic treatments not only selected for antibiotic-resistant bacteria, but also for MMR-defective mutators (Giraud et al., 2002). Bacteria with high mutation frequencies are predicted to have a higher probability of survival in changing and unpredictable environments and have greater ability to adapt to new conditions, including antibiotic pressure (Giraud et al., 2002; Ishii et al., 1989; LeClerc et al., 1996; Leigh, 1970; Sniegowski et al., 1997; Taddei et al., 1997b). Therefore, mutator bacteria are clinically important as they can promote the emergence of strains resistant to one or more antibiotics which increases hospitalization costs as well as patient morbidity and mortality (Carmeli et al., 1999). A positive correlation has been found between high mutation frequencies and antibiotic resistance in P. aeruginosa isolates from CF patients receiving long-term antibiotic therapy (Oliver et al., 2000).

The present study defines the prevalence of global hypermutators within a panel of H. influenzae isolated from a variety of patients, and commensal isolates. We identified several hypermutable strains, with the majority of those being isolated from CF sputum. The hypermutable phenotype of these isolates was associated with polymorphisms in conserved regions of mutS. In addition, six hypermutable isolates were discovered to be clonal, originating from the same CF patient collected over a period of about 11 months. These clonal isolates were further characterized to examine the effect of the hypermutable phenotype on antibiotic resistance over time.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and growth conditions.
H. influenzae were grown on chocolate agar [per litre: 36 g Difco GC medium with agar (Becton Dickinson), 10 g haemoglobin, 10 ml isovitalex and 5 units bacitracin ml–1] or supplemented brain heart infusion (sBHI) broth or agar [per litre: 37 g BHI media±15 g Bacto agar (Remel) supplemented with 10 µg {beta}-NAD ml–1, 10 µg haem ml–1, 10 µg histidine ml–1 and 5 units bacitracin ml–1]. Strains were incubated at 37 °C in room air. Antibiotic concentrations used for H. influenzae were (ml–1) 25 µg rifampicin, 14 µg nalidixic acid, 20 µg spectinomycin, 2 µg chloramphenicol and 30 µg ribostamycin. Chemicals were purchased from Sigma unless otherwise specified. In some experiments bacteria were washed and resuspended in phosphate-buffered saline (pH 7·0) with 0·1 % gelatin (PBSg).

Five hundred H. influenzae isolates, both commensal and from a variety of diseases (Table 1) were screened for the hypermutable phenotype as described below. Isolates were obtained from 450 different patients, mostly from the area of Seattle, WA, USA, with 25 patients contributing two or more isolates over time. All isolates with the hypermutable phenotype on screening were confirmed as H. influenzae by Gram staining, oxidase testing and the requirement for X (haemin) and V ({beta}-NAD+) factors for aerobic growth. {beta}-Lactamase production was detected with the nitrocefin dry slide test (Becton Dickinson). Encapsulation of strains was assessed by PCR for the presence of the capsule export locus (bexA) and, if present, PCR for the capsular gene cluster using primers specific for capsular types a–f (Falla et al., 1994). Strains Rd KW20 (R652) (Wilcox & Smith, 1975) and Rd KW20 mutS : : TSTE (R3544) (Watson et al., 2004) were used as negative and positive controls, respectively, for determination of mutation frequency.


View this table:
[in this window]
[in a new window]
 
Table 1. Percentage composition of clinical sources of screened H. influenzae isolates

Five hundred isolates from the collection of one of the authors (A. L. S.) were chosen to produce a diverse panel of sources for investigation of the prevalence of hypermutable strains. Isolates were obtained from 450 different clinical patients, mostly from the Seattle area, with 25 patients contributing two or more isolates over time.

 
DNA manipulations, PCR and sequencing.
Bacterial genomic DNA was prepared using the DNeasy Tissue kit from Qiagen. All PCR amplicons were partially purified using the Qiaprep PCR Purification kit (Qiagen). All restriction and modifying enzymes were purchased from New England Biolabs. PCR was performed using the High Fidelity Polymerase Blend (Roche) or the Fail Safe Amplification System (Epicentre). Oligonucleotide primers were purchased from Integrated DNA Technologies (Coralville, IA, USA). Southern blotting to nylon membranes was performed with the Turboblotter kit (Midwest Scientific) with probes labelled using the DIG High Prime DNA labelling kit (Roche) and chemiluminescent detection according to manufacturer's recommendations. DNA sequencing was performed by the DNA core facility at Seattle Biomedical Research Institute. Primer sequences and reaction conditions for multilocus sequence typing (MLST) specific for H. influenzae were obtained from www.mlst.net (Meats et al., 2003).

Determination of mutation frequency.
All 500 isolates were initially screened by assessment of the mutation frequency to rifampicin resistance and were considered potential hypermutators if they yielded >=10 rifampicin-resistant c.f.u. per 109 viable cells. This value was chosen as it is approximately 10 times the mutation frequency of strain Rd KW20 and is similar to the mutation frequency of a constructed mutS : : TSTE mutant of Rd KW20 (R3544) (Watson et al., 2004). Potential mutators were confirmed with additional measurement of mutation frequency to rifampicin, nalidixic acid and spectinomycin resistance assessed on sBHI agar plates. Fluctuation assays (Luria & Delbruck, 1943) were performed by inoculating 10 ml sBHI broth in a 125 ml flask to achieve an initial density of 107 c.f.u. ml–1 of H. influenzae and growing for 8–10 h shaking at 200 r.p.m. in 37 °C in room air. For unclear reasons, H. influenzae requires a relatively high inoculum density to achieve growth in sBHI liquid medium. The culture density after growth was approximately 1010 c.f.u. ml–1. Cultures were pelleted and resuspended in 500 µl PBSg. From these concentrated cultures 100 µl was spread in duplicate onto sBHI plates containing rifampicin, nalidixic acid or spectinomycin to determine the number of drug-resistant mutants per ml of culture. Another 100 µl of the same culture was serially diluted in PBSg and plated in duplicate on sBHI plates without antibiotics to determine the viable cell count per ml of concentrated culture. Growth was allowed for 48 h at 37 °C in room air prior to quantification. Mutation frequency was calculated as the number of drug-resistant cells per ml of culture divided by the total number of viable cells per ml of culture. Because the mutation frequency fluctuates between individual cultures, at least four independent measurements were sampled in duplicate for each strain with the median mutation frequency reported. Using the median frequency method of Drake (1991) an estimate was made of the mutation rate, defined as antibiotic-resistant bacteria produced per cell per division. It should be noted that due to the relatively high inoculum density required to achieve growth in broth media, some experiments could possibly have contained pre-existing mutants in strains with the highest mutation rate which would tend to lead to an overestimation of the mutation frequency. The combined effect of multiple replicates and reporting the median mutation frequency rather than the mean was used to minimize the effect of these artificial ‘jackpots’.

Enterobacterial repetitive intergenic consensus (ERIC)-PCR fingerprinting.
Random amplification PCR analysis was performed using primers ERIC 1 or ERIC 2. Primer sequences and reaction conditions were as described by Gomez-De-Leon et al. (2000). A negative control containing all reaction ingredients except for template DNA was run in parallel to each experiment and results were analysed only if the negative control failed to yield amplified DNA. The amplification products (8·5 µl) were separated by 2·0 % agarose gel electrophoresis at 5 V cm–1 for 5–7 h with subsequent staining by ethidium bromide. Amplification reactions were repeated a minimum of three times to verify the reproducibility of the ERIC-PCR banding profiles.

Ribotyping of bacterial isolates.
Genomic DNA was prepared for each isolate and 10 µg DNA was digested with 10 units EcoRI. Digested DNA was separated by 0·8 % agarose gel electrophoresis and subsequently transferred to nylon membranes. Membranes were blotted with a DIG-labelled probe for H. influenzae 16S rRNA genes (rrnABCDEF) generated from the 740 bp PCR product of primers rrnB F and rrnB R. Hybridization was detected by chemiluminescence using an alkaline phosphatase antibody against digoxigenin.

MLST.
MLST was performed for H. influenzae as described by Meats et al. (2003). The sequences were aligned and allele assignments were produced at the MLST website (www.mlst.net). Novel allele sequences identified in this work were submitted to the H. influenzae MLST database and assigned numbers by the database curator, Emma Meats.

Antibiotic susceptibility testing.
The close-interval MIC was determined using agar dilution and a Steer's replicator (Steers et al., 1959). Plates of sBHI with antibiotic concentrations of 0, 0·125, 0·25, 0·5, 1, 2, 4, 8, 16, 32 and 64 µg amoxycillin, ampicillin, cefaclor, ceftriaxone, cephalexin, chloramphenicol, ciprofloxacin, tetracycline, ticarcillin and tobramycin ml–1 were prepared. Bacteria were harvested from fresh overnight growth at 37 °C in room air on chocolate agar and resuspended in PBSg to an OD600 of 0·2, which corresponds to approximately 1x108 c.f.u. ml–1. This suspension was diluted 1 : 10 in PBSg and 150 µl was loaded into the wells of the multi-pin Steer's replicator; the multi-pin head was dipped into the wells and then gently applied to the surface of the agar plates. The pins deposit a volume of approximately 10 µl to the plate surface from each well which results in 105 c.f.u. delivered per spot. After drying for 30 min, the plates were incubated at 37 °C for 36 h in room air before reading. Inocula with less than 10 % growth relative to the antibiotic-free plate were considered negative for growth at that antibiotic concentration (Syriopoulou et al., 1979). H. influenzae strain ATCC 49766 was used as a control for ampicillin susceptibility testing with MICs interpreted using the recommendations in NCCLS standard M100-S11 (NCCLS, 2001). Antibiotic susceptibility tests were repeated three times and the mean MIC value reported.

Sequencing of mutS alleles from hypermutable isolates.
Genomic DNA from each strain was prepared to sequence mutS using overlapping primer sets listed in Table 2. Each primer set yields a product of about 500–1000 bp and the products were sequenced in both forward and reverse directions with the same primers used for amplification. PCR products were amplified using the Expand High Fidelity Polymerase (Roche) for maximal accuracy and were subsequently visualized by 0·8 % agarose gel electrophoresis, purified and sequenced at the SBRI DNA core facility. Sequences were aligned with the Bioedit Sequence Alignment Editor for Windows which is available online as freeware at www.mbio.ncsu.edu/BioEdit/bioedit.html. The mutS sequences from H. influenzae strains Rd KW20 (Fleischmann et al., 1995), Strain-12 (R2846; unpublished data), INT-1 (R2866; unpublished data) and E. coli strain K-12 (Blattner et al., 1997) were used as a template for general alignment of mutS sequences. Multisequence alignments were also performed using CLUSTAL W available at http://clustalw.genome.ad.jp/ (Thompson et al., 1994).


View this table:
[in this window]
[in a new window]
 
Table 2. Oligonucleotides used for PCR amplification and sequencing

 
Cloning of mutS alleles for complementation analysis.
The mutS genes were cloned for complementation analysis. To clone mutS, primers XbaI MutS F and XbaI MutS R were used to amplify mutS from genomic DNA of H. influenzae strains without tnaCAB upstream of mutS, and primers MutS Front F and XbaI MutS R were used for strains with tnaCAB. These primer sets produce 3 kb amplicons with XbaI sites at both the 5' and 3' ends, and these sites were used to ligate the XbaI-digested mutS sequences into XbaI-digested pSU2718, an E. coli to H. influenzae shuttle vector (Martinez et al., 1988). Following transformation and confirmation of cloning in E. coli DH5{alpha}, plasmids were purified and resuspended in Nanopure water for electroporation into H. influenzae strain R3544 (Rd KW20 mutS : : TSTE) with selection on chocolate agar plates supplemented with 2 µg chloramphenicol ml–1. Electroporation of plasmids into H. influenzae was performed as described by Mitchell et al. (1991). Strain R3544 transformed with plasmids encoding various mutS alleles was tested for complementation of the mutator phenotype by measuring mutation frequency to rifampicin resistance (25 µg ml–1) on sBHI plates as described above.

Methods for statistical analysis.
Statistical analyses were performed using the statistical functions of Microsoft Excel 1997 or by VassarStats, a website for Statistical Computation by Richard Lowry http://faculty.vassar.edu/lowry/VassarStats.html. The non-parametric, two-tailed Mann–Whitney test was used to test the significance of mutation frequencies, with P<0·05 indicating statistically significant differences.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Screening and detection of hypermutable H. influenzae isolates
Five hundred isolates from the collection of one of the authors (A. L. S.) were selected to obtain a panel from diverse patient sources (Table 1). The majority of the isolates were previously identified by the clinical microbiology service as non-typable; 42 isolates from blood or CSF were previously determined to be type b. The commensal isolates were isolated during screening of children at well-child physician visits to define the prevalence of antibiotic resistance. Using a single measurement of spontaneous mutation frequency to rifampicin resistance as an indicator, 27 of 500 isolates (5·4 %) were initially scored as possible hypermutators and were further tested by additional fluctuation tests to rifampicin, nalidixic acid and spectinomycin resistance. Fourteen of the 500 isolates (2·8 %) were confirmed as having a hypermutable phenotype and results of mutation frequency testing are shown in Table 3. The mutation rates to rifampicin and nalidixic acid resistance of strain Rd KW20 mutS : : TSTE (R3544) are about eightfold greater than the mutation frequency of the parent strain Rd KW20 (R652) due to insertional inactivation of the MMR protein mutS (Watson et al., 2004). These mutation rates are comparable to previously reported mutation rates to rifampicin and nalidixic acid resistance for strains Rd KW20 and Rd KW20 {Delta}mutS (Bayliss et al., 2002, 2004). The mutation rates of the natural hypermutable isolates were variable and ranged from about 5- to 380-fold greater than Rd KW20 (R652). These mutation frequency values are consistent with a previous report describing H. influenzae mutation frequencies to rifampicin resistance ranging from 10–9 to 10–6 (Mendelman et al., 1982). Some isolates produced higher mutation frequencies to one antibiotic compared to the others. Two isolates, C2394 and R465, were hypermutable to rifampicin and nalidixic acid resistance, but had a much lower frequency of spontaneous mutation to spectinomycin resistance. Thirteen of the 14 (92·9 %) hypermutable isolates were non-typable, defined as lacking the capsule export locus bexA by PCR. One isolate of the 14 (7·1 %), C2202, contained bexA, and further testing using PCR primers specific for capsular serotypes a–f (Falla et al., 1994) identified it as a type b encapsulated strain. Twelve of the 14 (85·7 %) hypermutable isolates were CF sputum isolates, which was 8·3 % of the total 145 CF sputum isolates tested. One mutator was an invasive blood isolate (C2202) and one was a conjunctivitis isolate (C2475).


View this table:
[in this window]
[in a new window]
 
Table 3. Mutation rates of H. influenzae isolates

 
Genetic fingerprinting of hypermutable isolates
To genetically differentiate the hypermutable isolates we used two independent methods, ERIC-PCR and ribotyping. ERIC-PCR is a method for fingerprinting strains utilizing randomly primed PCR with primers for ERIC sequences. These primers non-specifically hybridize to sequences within the H. influenzae genome and produce PCR product banding patterns that have been previously used to discriminate H. influenzae isolates (Gomez-De-Leon et al., 2000; Pettigrew et al., 2002; van Belkum et al., 1994). We found PCR using only one primer per reaction provided more distinctive and reproducible amplification profiles. Fig. 1 shows ERIC-PCR fingerprints for the 14 hypermutable isolates. Using the ERIC-1 primer, eight distinct PCR banding profiles were produced with amplicon sizes ranging from 300 to 1800 bp. Six isolates (C2408, C2561, C2562, C2687, C2692 and C2696) produced identical amplicon patterns. Two other isolates (C2621 and C2677) had identical patterns distinct from all other strains. PCR amplification with the ERIC-2 primer produced different amplicon patterns, but still differentiated the 14 isolates into eight distinct profiles with the same isolates showing identical banding profiles as seen with the ERIC-1 primer. In addition, ribotyping of EcoRI restricted genomic DNA was performed for the hypermutable isolates using a DIG-labelled probe for H. influenzae 16S rRNA (rrnABCDEF). Restriction with EcoRI produced only three distinct ribotypes with fragment sizes ranging from 300 to 6000 bp (data not shown); ribotyping with EcoRI is known to be limited in its discriminatory ability among non-typable isolates (Sharma et al., 2002). Six isolates (C2408, C2561, C2562, C2687, C2692 and C2696) produced an identical ribotype pattern, distinct from all other strains. Strain R465 produced a unique ribotype, and the ribotypes of the remaining strains were indistinguishable. A review of medical records, with approval of the Children's Hospital and Regional Medical Center Institutional Review Board, revealed that these six isolates with identical ERIC-PCR and ribotype patterns were all isolated over a period of about 11 months from the same patient; a male child with CF. MLST (Maiden et al., 1998; Meats et al., 2003) of adk, atpG, frdB, fucK, mdh, pgi and recA, confirmed these six strains were clonal, providing strong evidence that hypermutable strains C2408, C2561, C2562, C2687, C2692 and C2696 evolved from a common ancestor (allele designations of strain C2696 were archived in the H. influenzae MLST database). From these results it was concluded that eight genetically distinct hypermutable strains were isolated.



View larger version (87K):
[in this window]
[in a new window]
 
Fig. 1. ERIC-PCR fingerprinting of hypermutable isolates. Genomic DNA from each strain was randomly PCR-amplified with either the ERIC-1 (left panel) or ERIC-2 (right panel) primer. A negative control with no template DNA added was included in each experiment. Results were visualized by 2 % agarose gel electrophoresis and subsequent staining with ethidium bromide. A 100 bp ladder up to 1 kb is included on the left of each image; the 100 and 200 bp bands are not shown.

 
Sequencing of mutS in hypermutable isolates
The majority of natural hypermutable isolates described in other bacterial pathogens have been found to have point mutations or deletions in all or part of the mutS gene which inactivates the MMR system (LeClerc et al., 1996; Li et al., 2003; Matic et al., 1997; Oliver et al., 2000). The mutS alleles from each of the eight distinct members of the hypermutable panel were sequenced in an attempt to identify mutations in regions previously shown to be essential for mutS function in other organisms. Strain C2696 was sequenced as a representative of the clonal CF mutator strains determined by MLST. Strain C2677 was sequenced as a representative of C2677 and C2621; these strains were found to be identical by ERIC-PCR and ribotyping, and were both isolated from a separate child with CF. For comparison to alleles of mutS from non-hypermutable strains, the mutS sequences of H. influenzae Rd KW20 (Fleischmann et al., 1995), Strain 12 (R2846; unpublished data) and INT-1 (R2866; unpublished data), were included as well as mutS from E. coli K-12 ( Blattner et al., 1997). The online version of this paper (at http://mic.sgmjournals.org) contains a supplementary figure showing the alignment of the mutS sequences described here. DNA sequences of the mutS ORF from most strains were 2586 bp in length, with the exceptions of strain C2696, which was 2583 bp in length, and C2372, which was 2585 bp in length. DNA sequences of H. influenzae mutS shared a minimum identity of 94 % with each other, but were considerably polymorphic with a great number of both synonymous and non-synonymous nucleotide substitutions. In general, the predicted amino acid sequences of H. influenzae MutS proteins were about 99 % similar to each other, indicating the majority of polymorphisms in the DNA sequences were conservative mutations. No single mutation was identified in the predicted MutS sequence that was capable of distinguishing high-level mutators from moderate-level mutators. The predicted MutS protein of strain C2372 is only about 78 % similar to the other strains. Examination of the DNA sequence of this strain identified a deletion of thymine 1911 which results in a frameshift mutation and a premature TAA stop codon being introduced after amino acid position 636. This truncates the MutS protein and removes 225 aa from the C-terminus, a region previously shown in E. coli MutS to be necessary for homodimerization and interaction with MutL (Wu & Marinus, 1999). The predicted MutS protein sequence of strain C2696 lacks a conserved histidine residue at position 74. Examination of the DNA sequence of this strain identified a deletion of three nucleotides after nucleotide 219, which removes the codon for histidine, but keeps the remainder of the codons in the proper reading frame. Deletion of the N-terminal 300 aa in E. coli MutS eliminates DNA binding but preserves interaction with downstream proteins (Wu & Marinus, 1999). Histidine 74 is within the DNA binding region and may be required for DNA binding activity. The predicted MutS protein sequence of strain C2394 has two mutations in conserved amino acid positions: R198G and Q628P. Arginine 198 is located in the region required for DNA binding in E. coli MutS, and glutamine 628 is located near the P-loop motif required for nucleoside triphosphate binding and ATPase activity (Wu & Marinus, 1999). Mutations in the E. coli P-loop region are associated with a dominant negative mutator phenotype, presumably because mismatches are bound by defective MutS in an irreversible manner preventing further repair (Wu & Marinus, 1994). The predicted MutS protein sequence of strain C2475 has a mutation in a conserved residue: R563C. Deletion of this region has been shown to eliminate homodimerization in E. coli MutS and to eliminate DNA binding in MutS of Thermus thermophilus (Tachiki et al., 1998). The predicted MutS protein sequence of strain C2382 has a mutation, T615P, in the conserved P-loop motif region required for ATPase activity. The predicted MutS protein sequence of strain R465 has two mutations in conserved amino acid positions: S670L and G743S. Both of these mutations occur in regions previously shown to be necessary in E. coli for homodimerization and interaction with MutL (Wu & Marinus, 1999). The predicted MutS protein sequence of strain C2677 also has a mutation in the conserved residue G743S.

The region upstream of mutS was PCR-amplified using the MutS promoter F-R primer pair or the MutS Front F and the MutS promoter R primer pair. The PCR products of strains C2677 and C2382 were of equal size (about 530 bp) to the Rd KW20 PCR product. The PCR products of strains C2696, C2372, C2394, C2202, C2475 and R465, were each about 3·7 kb. Partial sequencing of these large amplicons identified the presence of the tryptophanase operon (tnaCAB); these strains were all indole-positive (data not shown). This operon has been previously shown to be widespread (70–75 %) among commensal respiratory isolates but more prevalent (94–100 %) in disease isolates (Martin et al., 1998). All strains examined were nearly identical in their DNA sequence in the first 70 bp upstream of mutS. Further upstream sequences shared much less identity due to the presence of the tnaCAB operon in six of the strains. The hypermutable strain C2382 has a deletion of five conserved nucleotides about 30 bp upstream of the mutS ATG codon within the putative promoter region.

mutS alleles are defective in trans
To determine if mutS alleles sequenced in this study were defective, complementation analysis was used: we cloned mutS alleles into the plasmid pSU2718 and then tested the ability of these plasmids to suppress the mutator phenotype of Rd KW20 mutS : : TSTE (R3544). Table 4 lists the mutation frequency to rifampicin resistance for the Rd KW20 mutS : : TSTE (R3544) mutator strain transformed with various plasmid constructs. Strain R3544 has a mutation rate about eightfold greater than the parental strain R652; the addition of the pSU2718 vector by itself did not significantly alter the R3544 mutation frequency (two-tailed Mann–Whitney test, P=0·144). Strain R3544 carrying the pMW058 plasmid with the R652 wild-type mutS allele had a significantly reduced mutation frequency (P<0·0005). Strain R3544 with either pMW059, pMW060 or pMW064 did not have a significantly reduced mutation frequency compared to R3544 alone (P=0·168 for each). Strain R3544 with the pMW065 plasmid had an intermediate mutation frequency that was significantly reduced from the R3544 parent strain (P<0·001), but remained significantly greater than R3544 with plasmid pMW058 (P<0·01).


View this table:
[in this window]
[in a new window]
 
Table 4. Complementation analysis of mutS alleles

 
Detection of {beta}-lactamase-negative ampicillin-resistant (BLNAR) hypermutable isolates
Bacterial strains with elevated mutation rates have an increased frequency of antibiotic resistance due to mutations in the target for the antibiotic or modifications in antibiotic permeability. In H. influenzae, resistance to {beta}-lactam antibiotics, such as ampicillin, occurs commonly through production of plasmid-mediated TEM-1 (Vega et al., 1976) or ROB-1 (Medeiros et al., 1986) {beta}-lactamases and less commonly by mutations in penicillin-binding proteins (PBPs) resulting in a lower affinity for {beta}-lactams or a reduction in the drug permeability (Mendelman et al., 1990a, 1984). BLNAR isolates are often resistant to first- and second-generation cephalosporins (Mendelman et al., 1990b; Powell & Williams, 1988). The hypermutable panel was tested for the presence of BLNAR isolates. All of the hypermutable isolates were screened for {beta}-lactamase production using the nitrocefin dry slide test: two isolates, C2475 and C2382, produced the enzyme, while in the remaining 12 it was not detectable. All isolates were tested by the agar dilution method for antibiotic susceptibility to ampicillin, amoxycillin, cefaclor and ceftriaxone (Table 5). The two isolates producing {beta}-lactamase exhibited high-level resistance (MIC>64 µg ml–1) to ampicillin and amoxycillin, and both were resistant to cefaclor (MIC>=32 µg ml–1). All isolates were susceptible to ceftriaxone, (MIC=0·125 µg ml–1). One isolate, R465, lacked detectable {beta}-lactamase activity and was found by agar dilution to be resistant to ampicillin and amoxycillin (MIC=16 µg ml–1), and had high-level resistance to cefaclor (MIC>64 µg ml–1); this isolate was considered to be BLNAR. It was found to have non-conservative point mutations in PBP3, conferring resistance to several {beta}-lactam antibiotics (Valentine and Smith, unpublished data). Several isolates were resistant to cefaclor (MIC>=32) but susceptible to ampicillin or amoxycillin (MIC<=1).


View this table:
[in this window]
[in a new window]
 
Table 5. Antibiotic susceptibility testing for H. influenzae hypermutators

MIC values are presented in µg ml–1. Data shown are the mean values of three independent measurements of MIC using the agar dilution method.

 
Evolution of antibiotic resistance in a clonal hypermutable strain
Six hypermutable isolates of clonal origin were isolated over about 11 months from the same child with CF (C2408, C2561, C2562, C2687, C2692 and C2696). A review of medical records indicated that this patient received ticarcillin/clavulinic acid, tobramycin and cephalexin throughout the 11 month period that these isolates were collected. We hypothesized that the persistent hypermutable H. influenzae strain isolated from this patient was selected for and maintained over time by constant antibiotic pressure. The antibiotic susceptibilities of these clonal hypermutable isolates were examined to a panel of additional antibiotics (Table 6). No differences in antibiotic susceptibilities were found for cephalexin (all resistant), ciprofloxacin (all susceptible) and tetracycline (all susceptible). The MIC doubled for chloramphenicol from 0·5 to 1·0 µg ml–1 for the three isolates C2687, C2692 and C2696. The MIC for ticarcillin quadrupled from 1·0 µg ml–1 for C2408, to 4·0 µg ml–1 for C2687, C2692 and C2696. Tobramycin resistance increased from an MIC of 4·0 µg ml–1 for the first isolate C2408, then increased to 64 µg ml–1 for C2561, then decreased to 32·0 µg ml–1 for C2562 and stabilized at 16·0 µg ml–1 for the remaining isolates C2687, C2692 and C2696. In addition, the antibiotic susceptibilities for ampicillin, amoxycillin and cefaclor also doubled for the three isolates C2687, C2692 and C2696 (Table 5). Although these changes in susceptibility are modest, the close-interval MIC testing demonstrates how a natural hypermutable strain may evolve antibiotic resistance over time.


View this table:
[in this window]
[in a new window]
 
Table 6. Additional antibiotic susceptibilities* for clonal CF hypermutators

MIC values are presented in µg ml–1. Data shown are the mean values of three independent measurements of MIC using the agar dilution method.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Hypermutability among natural isolates of bacterial pathogens has received increasing attention relatively recently; it seems an emerging theme is that many, if not all, bacterial species studied in detail will be found to have a relatively high proportion with increased mutation frequencies. We have found hypermutable isolates within a diverse panel of H. influenzae. Of 500 isolates tested, 14 (2·8 %) were hypermutable and subsequent genetic analysis identified eight unique hypermutable strains, with six of the eight (75·0 %) unique mutators being CF sputum isolates. These percentages are comparable to the original description (LeClerc et al., 1996) of hypermutable E. coli and Sal. enterica at frequencies of greater than 1 % in natural populations and reflects the increase in hypermutable isolates recoverable from specific disease conditions. Sputum from CF patients has previously yielded a high percentage of hypermutable isolates of P. aeruginosa and Sta. aureus (Matic et al., 1997; Oliver et al., 2000; Prunier et al., 2003). The high frequency of mutators in CF sputum found in our study and by others suggests that the hypermutable phenotype is advantageous in the CF lung environment, where repeated exposure to antibiotic therapy may select for and maintain strains with defects in DNA repair. Our screen may have missed a small number of additional hypermutable isolates due to poor growth of some H. influenzae clinical isolates under laboratory conditions. In addition, the clonality of the sample population as a whole was largely unknown, and as a result our percentage of hypermutable isolates in this population is an estimate and may actually be greater.

Hypermutable bacteria commonly have defects in genes involved with DNA proofreading, such as mutS of the MMR system (LeClerc et al., 1996; Li et al., 2003; Oliver et al., 2002). As a result of lesions that inactivate the MMR system, the global mutation rate increases (up to 1000-fold greater in E. coli) and the barrier to genetic recombination between divergent DNA sequences is lowered (Matic et al., 1995; Rayssiguier et al., 1989). Mutants of mutS are more abundant than isolates with mutations in other MMR genes in natural populations of E. coli (LeClerc et al., 1996; Matic et al., 1997). In contrast, in the laboratory environment mutants of other E. coli MMR genes, including mutL, are readily isolated, suggesting certain mutations may be counterselected against in nature due to deleterious effects that are not compensated by advantageous randomly generated mutations (Denamur et al., 2000). We have found several mutations in mutS alleles from our hypermutator panel that were in regions associated with defective MutS activity in E. coli. Four of these mutS alleles were cloned and these failed to completely complement the mutator phenotype of strain Rd KW20 mutS : : TSTE (R3544), suggesting that these mutS alleles are defective and are likely to be responsible for a mutator phenotype. The mutS allele of strain C2394 in plasmid pMW065 partially complemented the R3544 mutator phenotype (Table 4). C2394 is a moderate-level mutator with at least two mutations at conserved positions, and even though its MutS is defective, it still seems to retain some functions capable of partial complementation. In H. influenzae, mutants of mutS may be particularly common as mutS is an isolated ORF; in contrast, mutL and mutH are located within operons with several other genes involved in DNA metabolism, protein synthesis and cell envelope functions, and therefore only mutations which do not produce polar effects are more likely to be selected, potentially limiting the frequency of mutL and mutH mutations in nature. A similar explanation has been previously offered to account for the overrepresentation of natural E. coli mutS mutators compared to mutL mutators, with mutL in E. coli being part of an operon subject to polar mutations (Denamur et al., 2000). In addition, in H. influenzae point mutations inactivating MMR activity may be preferred over gene deletions as the remaining DNA sequence following a point mutation may facilitate genetic recombination at a later time to reacquire a functional mutS allele from the environment by lateral transfer. However, it should be noted that no evidence exists to suggest deletions inhibit lateral transfer. The mutS region of E. coli is highly polymorphic and phylogenetic analysis between mutS alleles and their respective genomes suggests that mutS may have been transferred by horizontal exchange multiple times throughout E. coli populations to rescue defective mutS alleles (Brown et al., 2001). The mutS sequences among our isolates were polymorphic and further comparison of these sequences to other housekeeping genes is required to determine if a similar mechanism is likely to occur in H. influenzae.

Antibiotic resistance has been previously correlated with the presence of the mutator phenotype in P. aeruginosa and Sta. aureus isolates from CF patients receiving long-term antibiotic therapy (Oliver et al., 2000; Prunier et al., 2003). Among our panel of mutators we found one BLNAR CF sputum isolate (R465) out of 12 (8·3 %) non-{beta}-lactamase-producing mutators, and seven of 12 (58·3 %) non-{beta}-lactamase-producing mutator isolates were resistant to the second-generation cephalosporin cefaclor (MIC>=32 µg ml–1). H. influenzae BLNAR isolates are relatively rare in the United States with estimates ranging from <0·1 to 2·5 % of all isolates investigated (Doern et al., 1997). The higher prevalence of hypermutable H. influenzae in patients receiving long-term antibiotic therapy, as in CF, may select and increase the frequency of BLNAR isolates recovered from those populations.

We have been able to identify six hypermutable CF sputum isolates from the same patient over a period of about 11 months that were determined by three independent genetic methods to have at least originated from the same ancestral strain, providing us with the ability to track the effects of the hypermutable phenotype on patterns of antibiotic resistance. Examination of antibiotic susceptibilities for these isolates with close-interval MIC testing showed increases in resistance for multiple antibiotics for some strains, mostly to the antibiotics the patient received as treatment for chronic respiratory infection. The significance of the MIC for tobramycin being unstable in these isolates is unknown, but may indicate that although these isolates are clonal in origin they may not be direct descendants of each other. These isolates may have originated from the same hypermutable ancestral strain and then underwent parallel evolution in different microenvironments of the CF lung. Parallel evolution of serial E. coli isolates with different antibiotic susceptibilities has been described (Low et al., 2001). We believe these increases in resistance demonstrate the ability of a persistent hypermutable strain to evolve antimicrobial resistance over time and illustrate the difficulty of eradicating pathogens in diseases such as CF where small variations in antibiotic accessibility may create a microenvironment where resistant bacterial strains will emerge (Pennington, 1981).

In summary, hypermutable clinical isolates of H. influenzae with mutations in mutS exist at relatively high frequencies, particularly in certain environments such as CF sputum. These isolates are capable of causing persistent infections and are likely to be selected for and maintained within natural populations by the benefits associated with an increased mutation frequency and a relaxed barrier to genetic recombination, producing isolates with new traits including antibiotic resistance.


   ACKNOWLEDGEMENTS
 
This work was supported by the National Institutes of Health grant AI44002.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bayliss, C. D., van de Ven, T. & Moxon, E. R. (2002). Mutations in polI but not mutSLH destabilize Haemophilus influenzae tetranucleotide repeats. EMBO 21, 1465–1476.[Abstract/Free Full Text]

Bayliss, C. D., Sweetman, W. A. & Moxon, E. R. (2004). Mutations in Haemophilus influenzae mismatch repair genes increase mutation rates of dinucleotide repeat tracts but not dinucleotide repeat-driven pilin phase variation rates. J Bacteriol 186, 2928–2935.[Abstract/Free Full Text]

Blattner, F. R., Plunkett, G., 3rd, Bloch, C. A. & 14 other authors (1997). The complete genome sequence of Escherichia coli K-12. Science 277, 1453–1474.[Abstract/Free Full Text]

Brown, E. W., LeClerc, J. E., Li, B., Payne, W. L. & Cebula, T. A. (2001). Phylogenetic evidence for horizontal transfer of mutS alleles among naturally occurring Escherichia coli strains. J Bacteriol 183, 1631–1644.[Abstract/Free Full Text]

Bucci, C., Lavitola, A., Salvatore, P., Del Giudice, L., Massardo, D. R., Bruni, C. B. & Alifano, P. (1999). Hypermutation in pathogenic bacteria: frequent phase variation in meningococci is a phenotypic trait of a specialized mutator biotype. Mol Cell 3, 435–445.[Medline]

Carmeli, Y., Troillet, N., Karchmer, A. W. & Samore, M. H. (1999). Health and economic outcomes of antibiotic resistance in Pseudomonas aeruginosa. Arch Intern Med 159, 1127–1132.[Abstract/Free Full Text]

Chao, L. & Cox, E. C. (1983). Competition between high and low mutating strains of Escherichia coli. Evolution 37, 125–134.

Cox, E. C. (1976). Bacterial mutator genes and the control of spontaneous mutation. Annu Rev Genet 10, 135–156.[CrossRef][Medline]

Dawson, K. J. (1999). The dynamics of infinitesimally rare alleles, applied to the evolution of mutation rates and the expression of deleterious mutations. Theor Popul Biol 55, 1–22.[CrossRef][Medline]

Denamur, E., Lecointre, G., Darlu, P. & 9 other authors (2000). Evolutionary implications of the frequent horizontal transfer of mismatch repair genes. Cell 103, 711–721.[Medline]

Denamur, E., Bonacorsi, S., Giraud, A. & 8 other authors (2002). High frequency of mutator strains among human uropathogenic Escherichia coli isolates. J Bacteriol 184, 605–609.[Abstract/Free Full Text]

Doern, G. V., Brueggemann, A. B., Pierce, G., Holley, H. P., Jr & Rauch, A. (1997). Antibiotic resistance among clinical isolates of Haemophilus influenzae in the United States in 1994 and 1995 and detection of beta-lactamase-positive strains resistant to amoxicillin-clavulanate: results of a national multicenter surveillance study. Antimicrob Agents Chemother 41, 292–297.[Abstract]

Drake, J. W. (1991). A constant rate of spontaneous mutation in DNA-based microbes. Proc Natl Acad Sci U S A 88, 7160–7164.[Abstract]

Falla, T. J., Crook, D. W., Brophy, L. N., Maskell, D., Kroll, J. S. & Moxon, E. R. (1994). PCR for capsular typing of Haemophilus influenzae. J Clin Microbiol 32, 2382–2386.[Abstract]

Fleischmann, R. D., Adams, M. D., White, O. & 37 other authors (1995). Whole-genome random sequencing and assembly of Haemophilus influenzae Rd. Science 269, 496–512.[Medline]

Foxwell, A. R., Kyd, J. M. & Cripps, A. W. (1998). Nontypeable Haemophilus influenzae: pathogenesis and prevention. Microb Mol Biol Rev 62, 294–308.[Abstract/Free Full Text]

Giraud, A., Matic, I., Radman, M., Fons, M. & Taddei, F. (2002). Mutator bacteria as a risk factor in treatment of infectious diseases. Antimicrob Agents Chemother 46, 863–865.[Abstract/Free Full Text]

Glickman, B. W. & Radman, M. (1980). Escherichia coli mutator mutants deficient in methylation-instructed DNA mismatch correction. Proc Natl Acad Sci U S A 77, 1063–1067.[Abstract]

Gomez-De-Leon, P., Santos, J. I., Caballero, J., Gomez, D., Espinosa, L. E., Moreno, I., Pinero, D. & Cravioto, A. (2000). Genomic variability of Haemophilus influenzae isolated from Mexican children determined by using enterobacterial repetitive intergenic consensus sequences and PCR. J Clin Microbiol 38, 2504–2511.[Abstract/Free Full Text]

Ishii, K., Matsuda, H., Iwasa, Y. & Sasaki, A. (1989). Evolutionarily stable mutation rate in a periodically changing environment. Genetics 121, 163–174.[Abstract/Free Full Text]

LeClerc, J. E., Li, B., Payne, W. L. & Cebula, T. A. (1996). High mutation frequencies among Escherichia coli and Salmonella pathogens. Science 274, 1208–1211.[Abstract/Free Full Text]

Leigh, E. G. (1970). Natural selection and mutability. Am Nat 104, 301–305.[CrossRef]

Li, B., Tsui, H. C., LeClerc, J. E., Dey, M., Winkler, M. E. & Cebula, T. A. (2003). Molecular analysis of mutS expression and mutation in natural isolates of pathogenic Escherichia coli. Microbiology 149, 1323–1331.[CrossRef][Medline]

Low, A. S., MacKenzie, F. M., Gould, I. M. & Booth, I. R. (2001). Protected environments allow parallel evolution of a bacterial pathogen in a patient subjected to long-term antibiotic therapy. Mol Microbiol 42, 619–630.[CrossRef][Medline]

Luria, S. & Delbruck, M. (1943). Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28, 491.[Free Full Text]

Maiden, M. C., Bygraves, J. A., Feil, E. & 10 other authors (1998). Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms. Proc Natl Acad Sci U S A 95, 3140–3145.[Abstract/Free Full Text]

Martin, K., Morlin, G., Smith, A., Nordyke, A., Eisenstark, A. & Golomb, M. (1998). The tryptophanase gene cluster of Haemophilus influenzae type b: evidence for horizontal gene transfer. J Bacteriol 180, 107–118.[Abstract/Free Full Text]

Martinez, E., Bartolome, B. & de la Cruz, F. (1988). pACYC184-derived cloning vectors containing the multiple cloning site and lacZ alpha reporter gene of pUC8/9 and pUC18/19 plasmids. Gene 68, 159–162.[CrossRef][Medline]

Matic, I., Rayssiguier, C. & Radman, M. (1995). Interspecies gene exchange in bacteria: the role of SOS and mismatch repair systems in evolution of species. Cell 80, 507–515.[Medline]

Matic, I., Radman, M., Taddei, F., Picard, B., Doit, C., Bingen, E., Denamur, E. & Elion, J. (1997). Highly variable mutation rates in commensal and pathogenic Escherichia coli. Science 277, 1833–1834.[Free Full Text]

Meats, E., Feil, E. J., Stringer, S., Cody, A. J., Goldstein, R., Kroll, J. S., Popovic, T. & Spratt, B. G. (2003). Characterization of encapsulated and noncapsulated Haemophilus influenzae and determination of phylogenetic relationships by multilocus sequence typing. J Clin Microbiol 41, 1623–1636.[Abstract/Free Full Text]

Medeiros, A. A., Levesque, R. & Jacoby, G. A. (1986). An animal source for the ROB-1 beta-lactamase of Haemophilus influenzae type b. Antimicrob Agents Chemother 29, 212–215.[Medline]

Mendelman, P. M., Roberts, M. C. & Smith, A. L. (1982). Mutation frequency of Haemophilus influenzae to rifampin resistance. Antimicrob Agents Chemother 22, 531–533.[Medline]

Mendelman, P. M., Chaffin, D. O., Stull, T. L., Rubens, C. E., Mack, K. D. & Smith, A. L. (1984). Characterization of non-beta-lactamase-mediated ampicillin resistance in Haemophilus influenzae. Antimicrob Agents Chemother 26, 235–244.[Medline]

Mendelman, P. M., Chaffin, D. O. & Kalaitzoglou, G. (1990a). Penicillin-binding proteins and ampicillin resistance in Haemophilus influenzae. J Antimicrob Chemother 25, 525–534.[Abstract]

Mendelman, P. M., Chaffin, D. O., Krilov, L. R., Kalaitzoglou, G., Serfass, D. A., Onay, O., Wiley, E. A., Overturf, G. D. & Rubin, L. G. (1990b). Cefuroxime treatment failure of nontypable Haemophilus influenzae meningitis associated with alteration of penicillin-binding proteins. J Infect Dis 162, 1118–1123.[Medline]

Mitchell, M. A., Skowronek, K., Kauc, L. & Goodgal, S. H. (1991). Electroporation of Haemophilus influenzae is effective for transformation of plasmid but not chromosomal DNA. Nucleic Acids Res 19, 3625–3628.[Abstract]

Modrich, P. & Lahue, R. (1996). Mismatch repair in replication fidelity, genetic recombination, and cancer biology. Annu Rev Biochem 65, 101–133.[CrossRef][Medline]

Moxon, E. R., Rainey, P. B., Nowak, M. A. & Lenski, R. E. (1994). Adaptive evolution of highly mutable loci in pathogenic bacteria. Curr Biol 4, 24–33.[Medline]

NCCLS (2001). Performance Standards for Antimicrobial Susceptibility Testing. Approved standard M100-S11. Wayne, PA: National Committee for Clinical Laboratory Standards.

Oliver, A., Canton, R., Campo, P., Baquero, F. & Blazquez, J. (2000). High frequency of hypermutable Pseudomonas aeruginosa in cystic fibrosis lung infection. Science 288, 1251–1254.[Abstract/Free Full Text]

Oliver, A., Baquero, F. & Blazquez, J. (2002). The mismatch repair system (mutS, mutL and uvrD genes) in Pseudomonas aeruginosa: molecular characterization of naturally occurring mutants. Mol Microbiol 43, 1641–1650.[CrossRef][Medline]

Pennington, J. E. (1981). Penetration of antibiotics into respiratory secretions. Rev Infect Dis 3, 67–73.[Medline]

Pettigrew, M. M., Foxman, B., Ecevit, Z., Marrs, C. F. & Gilsdorf, J. (2002). Use of pulsed-field gel electrophoresis, enterobacterial repetitive intergenic consensus typing, and automated ribotyping to assess genomic variability among strains of nontypeable Haemophilus influenzae. J Clin Microbiol 40, 660–662.[Abstract/Free Full Text]

Powell, M. & Williams, J. D. (1988). In-vitro activity of cefaclor, cephalexin and ampicillin against 2458 clinical isolates of Haemophilus influenzae. J Antimicrob Chemother 21, 27–31.[Abstract]

Prunier, A. L., Malbruny, B., Laurans, M., Brouard, J., Duhamel, J. F. & Leclercq, R. (2003). High rate of macrolide resistance in Staphylococcus aureus strains from patients with cystic fibrosis reveals high proportions of hypermutable strains. J Infect Dis 187, 1709–1716.[CrossRef][Medline]

Rayssiguier, C., Thaler, D. S. & Radman, M. (1989). The barrier to recombination between Escherichia coli and Salmonella typhimurium is disrupted in mismatch-repair mutants. Nature 342, 396–401.[CrossRef][Medline]

Richardson, A. R., Yu, Z., Popovic, T. & Stojiljkovic, I. (2002). Mutator clones of Neisseria meningitidis in epidemic serogroup A disease. Proc Natl Acad Sci U S A 99, 6103–6107.[Abstract/Free Full Text]

Schaaper, R. M. & Dunn, R. L. (1987). Spectra of spontaneous mutations in Escherichia coli strains defective in mismatch correction: the nature of in vivo DNA replication errors. Proc Natl Acad Sci U S A 84, 6220–6224.[Abstract]

Sharma, A., Kaur, R., Ganguly, N. K., Singh, P. D. & Chakraborti, A. (2002). Subtype distribution of Haemophilus influenzae isolates from north India. J Med Microbiol 51, 399–404.[Abstract/Free Full Text]

Sniegowski, P. D., Gerrish, P. J. & Lenski, R. E. (1997). Evolution of high mutation rates in experimental populations of E. coli. Nature 387, 703–705.[CrossRef][Medline]

Steers, E. E., Foltz, L., Graves, B. S. & Riden, J. (1959). An inocula replicating apparatus for routine testing of bacterial susceptibility of antibiotics. Antibiot Chemother 9, 307–311.

Strauss, B. S. (1999). Frameshift mutation, microsatellites and mismatch repair. Mutat Res 437, 195–203.[CrossRef][Medline]

Syriopoulou, V. P., Scheifele, D. W., Sack, C. M. & Smith, A. L. (1979). Effect of inoculum size on the susceptibility of Haemophilus influenzae b to beta-lactam antibiotics. Antimicrob Agents Chemother 16, 510–513.[Medline]

Tachiki, H., Kato, R., Masui, R., Hasegawa, K., Itakura, H., Fukuyama, K. & Kuramitsu, S. (1998). Domain organization and functional analysis of Thermus thermophilus MutS protein. Nucleic Acids Res 26, 4153–4159.[Abstract/Free Full Text]

Taddei, F., Radman, M., Maynard-Smith, J., Toupance, B., Gouyon, P. H. & Godelle, B. (1997a). Role of mutator alleles in adaptive evolution. Nature 19, 700–702.

Taddei, F., Vulic, M., Radman, M. & Matic, I. (1997b). Genetic variability and adaptation to stress. In Environmental Stress, Adaptation, and Evolution, pp. 271–290. Edited by R. Bijlsma & V. Loeschcke. Basel, Switzerland: Birkhauser.

Tang, C. M., Hood, D. W. & Moxon, E. R. (2001). Pathogenesis of Haemophilus influenzae Infections. In Principles of Bacterial Pathogenesis, pp. 675–716. Edited by E. A. Groisman. St Louis, MO: Academic Press.

Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 4673–4680.[Abstract]

van Belkum, A., Duim, B., Regelink, A., Moller, L., Quint, W. & van Alphen, L. (1994). Genomic DNA fingerprinting of clinical Haemophilus influenzae isolates by polymerase chain reaction amplification: comparison with major outer-membrane protein and restriction fragment length polymorphism analysis. J Med Microbiol 41, 63–68.[Abstract]

Vega, R., Sadoff, H. L. & Patterson, M. J. (1976). Mechanisms of ampicillin resistance in Haemophilus influenzae type B. Antimicrob Agents Chemother 9, 164–168.[Medline]

Watson, M. E., Jr, Jarisch, J. & Smith, A. L. (2004). Inactivation of deoxyadenosine methyltransferase (dam) attenuates Haemophilus influenzae virulence. Mol Microbiol 53, 651–664.[CrossRef][Medline]

Wilcox, K. & Smith, H. (1975). Isolation and characterization of mutants of Haemophilus influenzae deficient in an adenosine 5'-triphosphate-dependent deoxyribonuclease activity. J Bacteriol 122, 443–453.[Medline]

Wu, T. H. & Marinus, M. G. (1994). Dominant negative mutator mutations in the mutS gene of Escherichia coli. J Bacteriol 176, 5393–5400.[Abstract]

Wu, T. H. & Marinus, M. G. (1999). Deletion mutation analysis of the mutS gene in Escherichia coli. J Biol Chem 274, 5948–5952.[Abstract/Free Full Text]

Received 10 April 2004; revised 17 June 2004; accepted 30 June 2004.



This Article
Abstract
Full Text (PDF)
Supplementary sequence data
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Watson, M. E.
Articles by Smith, A. L.
Articles citing this Article
PubMed
PubMed Citation
Articles by Watson, M. E., Jr
Articles by Smith, A. L.
Agricola
Articles by Watson, M. E.
Articles by Smith, A. L.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2004 Society for General Microbiology.