JAC Advance Access originally published online on February 1, 2008
Journal of Antimicrobial Chemotherapy 2008 61(3):509-514; doi:10.1093/jac/dkm523
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Original research |
Characterization of extended-spectrum β-lactamase-producing isolates of Haemophilus parainfluenzae
1 School of Human Life Sciences, University of Tasmania, Launceston, Tasmania 7250, Australia 2 University of Pretoria, Pretoria, South Africa 3 Dalhousie University, Halifax, Canada 4 Queen Elizabeth II Health Sciences Centre, Halifax, Canada
* Corresponding author. Tel: +61-3-63-243323; Fax: +61-3-63-243658; E-mail: stephen.tristram{at}utas.edu.au
Received 9 September 2007; returned 28 November 2007; revised 20 November 2007; accepted 10 December 2007
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
Objectives: To characterize the β-lactam resistance mechanisms of two clinical isolates of cefotaxime-resistant Haemophilus parainfluenzae recovered from patients in South Africa.
Methods: The relatedness of isolates and plasmids was assessed using PFGE and restriction enzyme analysis, respectively. Plasmid-mediated and chromosomally integrated blaTEM genes and ftsI genes were sequenced, and the plasmid-mediated blaTEM-15 was used to transform a range of control organisms.
Results: The two isolates were found to be unique according to PFGE, but had an identical 3.7 kb plasmid encoding a TEM-15 β-lactamase. Both isolates also had substitutions in penicillin binding protein 3 (PBP3) consistent with substitutions known to exist in β-lactamase-negative ampicillin-resistant (BLNAR) strains of Haemophilus influenzae. The cefotaxime MICs for control strains of H. influenzae, H. parainfluenzae and BLNAR H. influenzae transformed with the plasmid-mediated blaTEM-15 were 1.0, 1.0 and 4.0 mg/L, respectively, compared with 16.0 and 8.0 mg/L, respectively, for the two parent H. parainfluenzae.
Conclusions: The high-level cefotaxime resistance in the H. parainfluenzae isolates was due to a combination of a plasmid-mediated TEM-15 extended-spectrum β-lactamase with altered PBP3 probably contributing. Other contributing resistance mechanisms could not be excluded.
Keywords: plasmid , ftsI , TEM , ESBL , PBP3 , promoter , penicillin binding protein
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
Haemophilus influenzae and Haemophilus parainfluenzae are normal inhabitants of the human upper respiratory tract, with H. influenzae being a common opportunistic pathogen and H. parainfluenzae an infrequent and less well-recognized opportunistic pathogen.1 Ampicillin resistance mediated by TEM-1 β-lactamase is common in both species2,3 and is associated with two types of plasmids. Small non-conjugative plasmids of
4 kb generally code only for the β-lactamase, whereas larger plasmids of 40 kb often encode multiple antibiotic resistance genes in addition to the β-lactamase and are usually chromosomally integrated.2 β-Lactamase-negative ampicillin-resistant (BLNAR) strains of H. influenzae with altered penicillin binding protein 3 (PBP3) and associated resistance or reduced susceptibility to β-lactam antibiotics are now widespread, but remain of low prevalence worldwide.4 Although some of these strains have greatly increased cefotaxime MICs, of up to 2.0 mg/L, strains resistant to extended-spectrum cephalosporins have not yet been described.4 Little is known about PBP3 in H. parainfluenzae and whether it is associated with resistance to β-lactams in this organism.5
Given the high prevalence of TEM-1 β-lactamase-positive strains of H. influenzae and H. parainfluenzae, and the widespread use of cephalosporins, it is surprising that TEM-derived extended-spectrum β-lactamases (ESBLs) and associated extended-spectrum cephalosporin resistance have not emerged as they have in Enterobacteriaceae. In studies where cloned TEM-derived ESBLs were expressed in recombinant strains of H. influenzae, the cefotaxime MICs only increased as high as 0.5 mg/L, however, when the ESBLs were expressed in strains of H. influenzae with altered PBP3, the strains were resistant, with cefotaxime MICs as high as 8.0 mg/L.6–8
Against this background, this study details the characterization of two cefotaxime-resistant clinical isolates of H. parainfluenzae recovered from patients in South Africa.9
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
Plasmids and strains
H. parainfluenzae SF2 was isolated in March 1999 from pus collected during bronchoscopy on a chronically malnourished 6-year-old female. The patient was diagnosed with community-acquired pneumonia and had failed treatment with amoxicillin/clavulanate. H. parainfluenzae SF3 was isolated in January 2000 from the sputum of a 54-year-old female presenting with an acute exacerbation of her chronic obstructive pulmonary disease. The patients were not related, came from geographically distinct informal settlements in South Africa and were lost to follow-up. These clinical isolates along with other strains and plasmids used in this study are described in Table 1.
|
Culture media
Chocolate agar was supplemented with antibiotics or other additives as described, and supplemented brain heart infusion (sBHI) broth contained 15 mg/L haematin and NAD, and 2% (v/v) Vitox (Oxoid, Australia).
Initial characterization of isolates
H. parainfluenzae SF2 and SF3 were identified by Gram’s stain, catalase production and standard X and V factor requirements. PFGE was performed on the isolates as previously described using ApaI and SmaI (Promega, Australia) enzymes individually.10
Susceptibility testing to ampicillin, amoxicillin/clavulanate, cefotaxime, tetracycline, trimethoprim/sulfamethoxazole, rifampicin and ciprofloxacin was performed using CLSI disc diffusion.8 MICs of cefotaxime and cefotaxime/clavulanate were determined using Etest® (AB Biodisk, Sweden). The presence of ESBLs was detected using the amoxicillin/clavulanate pre-diffusion disc test previously established for ESBL screening of H. influenzae.11
Plasmids pSF2 and pSF3 were extracted from H. parainfluenzae SF2 and SF3, respectively, using the Qiagen Hispeed Maxi Kit (Qiagen, Australia) and digested individually and as a triple digest by restriction enzymes PstI, AvaI and SacI (Promega).
Crude preparations of β-lactamases were made from overnight cultures on chocolate agar using four freeze/thaw cycles in 0.2 M sodium acetate (pH 5.5), focused on pH 5–8 Ready Gel (Bio-Rad, Australia) pre-cast IEF gels, stained with nitrocefin (Oxoid) and compared with broad-range (pI 4.45–9.6) IEF standards.
Plasmid transformation studies
Plasmid pSF2 was used to transform H. influenzae Rd, H. influenzae Rd BLNAR5 and H. parainfluenzae SS using electroporation as previously described.12 Transformants were selected on chocolate agar with 8 mg/L ampicillin. β-Lactamase production was confirmed using nitrocefin hydrolysis and ESBL was presumptively identified using the amoxicillin/clavulanate pre-diffusion disc test and then confirmed with sequencing.
Curing of plasmids from H. parainfluenzae SF2 and SF3 was attempted by methods described elsewhere,13–15 in order to examine the relative effect of non-plasmid-mediated mechanisms on cefotaxime resistance. Strains were grown overnight in sBHI with 2 mg/L ethidium bromide, 20 mg/L acridine orange or 0.2 mg/L mitomycin C and then plated on chocolate agar. One hundred colonies were chosen at random and replica-plated onto a series of chocolate agar plates with 0, 0.5, 1.0, 2.0 and 5.0 mg/L cefotaxime. Colonies that showed reduced resistance to cefotaxime relative to control inocula of untreated H. parainfluenzae SF2 and SF3 were further investigated for possible loss of plasmid (and plasmid-mediated ESBL) by the amoxicillin/clavulanate pre-diffusion disc test.
Nalidixic-acid-resistant mutants of H. influenzae Rd and H. parainfluenzae SS (for use as recipients in conjugation) were produced by multistep selection, where first-step mutants were selected by inoculating 100 µL of a 1.0 McFarland suspension of organism onto chocolate agar with 2 mg/L nalidixic acid. Second-step mutants were selected similarly using an inoculation of first-step mutants onto chocolate agar containing 8 mg/L nalidixic acid.
The ability of H. parainfluenzae SF2 and SF3 to conjugally transfer pSF2 and pSF3, respectively, to H. influenzae RdNalR and H. parainfluenzae SSNalR was assessed using the membrane mating technique as previously described,3 using chocolate agar with both 8 mg/L ampicillin and 8 mg/L nalidixic acid to screen for recipients.
Plasmid, extracted from H. parainfluenzae SF2, SF3 and strains transformed with pSF2, was used as template to amplify the plasmid blaTEM by PCR using primers A and B (Table 2), and subsequently sequenced as previously described.16
Sequencing of the chromosomal blaTEM required a different strategy because of interference from the plasmid-mediated blaTEM (which was subsequently identified as blaTEM-15). Forward (15X2F) and reverse (15X2R) primers (Table 2) were designed with a mismatch at the 3' end with respect to a G512(566)A substitution (see blaTEM nucleotide numbering) in the plasmid blaTEM-15 relative to blaTEM-1. An additional mismatch at the penultimate base of the 3' end was also introduced to further destabilize primer binding to plasmid blaTEM-15. Gradient analysis was used to optimize the annealing temperature such that PCR with primers A and 15X2R, and primers B and 15X2F would amplify control blaTEM-1 but not plasmid-mediated blaTEM-15. Primers A and 15X2R, and B and 15X2F were used to amplify 524 and 585 bp fragments from either side of bp 512(566) of the chromosomal blaTEM using a boiled cell suspension of H. parainfluenzae SF2 and SF3 as template. These fragments were then sequenced as above.
|
ftsI sequencing
Initial attempts to amplify the ftsI genes from H. parainfluenzae SF2 and SF3 using the primers described by Ubukata et al.12 for H. influenzae were unsuccessful, so an approach specific for H. parainfluenzae was designed. The sequence of the H. parainfluenzae ftsI gene was determined by aligning the H. influenzae Rd (ATCC 51907) ftsI gene sequence (GenBank accession number NC00097) with the genome sequence of H. parainfluenzae strain T3T1 (Sanger Institute).17 Primers ftsIparaF and ftsIparaR (Table 2) were then designed to amplify a 1038 bp fragment encoding the PBP3 transpeptidase domain, which was amplified and sequenced as previously described for H. influenzae.12
Attempts were made to transform recipients H. parainfluenzae SS and H. influenzae Rd with ftsI genes from H. parainfluenzae SF2 and SF3 using a variety of approaches.
Chromosomal DNA was extracted from H. parainfluenzae SF2 and SF3 using the DNeasy kit (Qiagen) and transformation was attempted using the M-IV method.18 In addition, an open reading frame encompassing the entire ftsI gene was amplified by PCR using primers transparaF and transparaR (Table 2) and used for transformation by electroporation. Both the PCR and electroporation were performed as previously described.12 Finally, H. parainfluenzae SF2 and SF3 were used as donors, and H. influenzae Rd NalR and H. parainfluenzae SS NalR as recipients, and incubated together in various donor/recipient pairs in sBHI broth for 2 h as described by Takahata et al.,5 to test for in vitro horizontal ftsI gene transfer.
For all methods, chocolate agar with 0.5 mg/L ampicillin or 0.1 mg/L cefotaxime were both used to screen for transformants, and in the case of the in vitro horizontal ftsI gene transfer method, both the cefotaxime and ampicillin plates were additionally supplemented with 8.0 mg/L nalidixic acid.
The nucleotides of the blaTEM-1 genes in this study are numbered according to Sutcliffe, without taking deletions or insertions in the promoter region into account to enable comparisons between particular nucleotides in these genes and published sequences to be easily made. Where given, the numbers in parentheses represent the actual consecutive numerical position of the nucleotides, taking deletions and insertions into account.
Nucleotide sequence accession numbers
The complete DNA sequences including promoter regions for blaTEM-15 and blaTEM-1, and the partial ftsI gene sequence as determined in this study appear in the DDBJ/EMBL/GenBank nucleotide sequence databases under accession numbers AM849805 [GenBank] , AM849806 [GenBank] and AM849807 [GenBank] , respectively.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
Initial characterization
Identification of both isolates of H. parainfluenzae was confirmed by a requirement for V factor only. PFGE results (data not shown) indicate that H. parainfluenzae SF2 and SF3 were unique strains. Both were resistant to ampicillin, amoxicillin/clavulanate, cefotaxime, tetracycline, trimethoprim/sulfamethoxazole and rifampicin, susceptible to ciprofloxacin and positive for an ESBL by the amoxicillin/clavulanate pre-diffusion disc test. Isolate SF2 was resistant to chloramphenicol. MICs are presented in Table 3 for selected antibiotics.
|
Plasmid characterization
Both pSF2 and pSF3 gave identical restriction digest patterns, yielding a single band of 3.7 kb when digested with AvaI, PstI or SacI, and three bands of 2.4, 0.7 and 0.6 kb as a triple digest.
H. parainfluenzae SF2 and SF3 displayed bands at pI 5.4 and 6.0, whereas H. influenzae Rd-pSF2, H. parainfluenzae SS-pSF2 and H. influenzae BLNAR5-pSF2 only produced bands at pI 6.0. These results suggest that plasmid pSF2 contains only a single blaTEM gene.
Strains transformed with pSF2 were nitrocefin hydrolysis positive, produced a β-lactamase of pI 6.0, were positive for an ESBL by the amoxicillin/clavulanate pre-diffusion disc test and had increased cefotaxime MICs (Table 3) relative to their parent strains. None of the ‘non-β-lactamase’-mediated resistances observed in H. parainfluenzae SF2 and SF3 were observed in any of the recipients of pSF2, indicating that these resistances were not encoded on pSF2.
The pSF2- and pSF3-mediated blaTEM gene was identified as a blaTEM-15, with G512(566)A and G914(968)A nucleotide substitutions predicting the amino acid substitutions E102K and G236S of TEM-15 (amino acids numbered according to Sutcliffe), which is consistent with the observed pI of 6.0. A 54 bp insertion sequence in the promoter region is consistent with the recently described blaTEM Prpt promoter.19
The chromosomal blaTEM genes in H. parainfluenzae SF2 and SF3 were blaTEM-1. In H. parainfluenzae SF2, a 17 bp deletion of nucleotides 31–47 was detected in the promoter region and this appears to generate an additional promoter (Figure 1). This deletion was not detected in H. parainfluenzae SF3 which had the C32T substitution associated with the Pa/Pb promoter.
|
ftsI sequencing and deduced amino acid sequences
Alignment of nucleotide and amino acid sequences for ftsI and PBP for H. influenzae Rd and H. parainfluenzae T3T1 (Sanger Institute) showed 78% and 83% homology, respectively. Nucleotides 931–1707 and deduced amino acids 311–569 from H. parainfluenzae SF2 and SF3 were aligned against those from the reference H. parainfluenzae T3T1 and amino acid substitutions A343V, S385T and N526H were detected in both H. parainfluenzae SF2 and SF3.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
The restriction analysis of pSF2 and pSF3 indicated that the plasmids were identical, so only pSF2 was subjected to further study. The plasmid analysis, susceptibility testing, sequencing of the blaTEM genes and IEF analysis of H. parainfluenzae SF2 and SF3 and the control strains transformed with pSF2 indicate that blaTEM-15 is located on the small 3.7 kb plasmid, and that blaTEM-1 is probably located on a chromosomally integrated plasmid. Of significant interest is the finding of the 54 bp insertion associated with the Prpt promoter in the blaTEM-15 genes of pSF2 and pSF3. This promoter has only been described in H. influenzae19 and suggests that the blaTEM-15 gene has evolved from a blaTEM-1 within Haemophilus rather than being acquired by transposition from Enterobacteriaceae. Also of significant interest is the 17 bp deletion in the promoter region of the blaTEM-1 of H. parainfluenzae SF2. This deletion brings together a putative pre-existing but redundant –35 region of TTGAAG with a putative –10 region of TACGAT (consensus sequence is TATAAT), which replaces the TACGCC (if P3) or TACGCT (if Pa/Pb) without the deletion (Figure 1). It is proposed that this new promoter be referred to as Pdel17. Promoter modifications appear to be common in the blaTEM genes of Haemophilus, with Prpt (54 bp insertion) and Pdel (135 bp deletion) having already been described.19,20 The small 3.7 kb blaTEM-15-encoding plasmids were shown to be identical by restriction analysis, but are too small to be conjugative.14 They may have been mobilized by the larger chromosomally integrated conjugative plasmids,14 thought to encode the blaTEM-1 genes detected in both isolates. If this was the case, then the 17 bp deletion in the promoter region of blaTEM-1 in strain SF2 must have occurred after conjugation as it was not detected in the blaTEM-1 of strain SF3. This hypothesis remains speculative as it was not possible in this study to conjugally transfer any of the plasmids and blaTEM genes to control recipients. A similar event to that proposed above has been reported, where a plasmid-mediated blaTEM gene was conjugally transferred between different strains of Escherichia coli in an infant’s gut, and the blaTEM gene in the donor strain subsequently mutated to produce a stronger promoter.21
When control strains of H. influenzae and H. parainfluenzae were transformed with blaTEM-15 on pSF2, the cefotaxime MICs were only moderately increased from 0.012 to 1.0 mg/L compared with 8–16 mg/L in the H. parainfluenzae isolates, which were also amoxicillin/clavulanate resistant. These observations prompted speculation that an additional β-lactam resistance mechanism, possibly altered PBP3, may have been present in the H. parainfluenzae isolates.
The findings of altered PBP3 in H. parainfluenzae SF2 and SF3 are consistent with both the resistance to amoxicillin/clavulanate,10,12 and the cefotaxime MICs that were higher than those predicted if due solely to the production of a TEM-derived ESBL.6,7
The S385T and the N526H substitutions are probably significant findings. Substitutions at these positions are common in BLNAR strains of H. influenzae and are associated with the higher level of cefotaxime resistance in the range observed in those strains.4,12 It is interesting to note that in the numerous studies that have characterized the amino acid substitutions in BLNAR strains of H. influenzae, the substitutions at any given amino acid position are consistent.4 The S385T substitution in this study correlates with the S385T substitutions seen in BLNAR strains of H. influenzae, but the N526H substitution in the H. parainfluenzae isolates is consistently N526K in H. influenzae.4 It seems likely that a histidine substitution would have a similar impact on protein structure and binding affinity of PBP3 to lysine, as both amino acids are similarly polar and basic in nature. BLNAR strains with substitutions only at positions 385 and 526 are uncommon in H. influenzae, but have been reported.4,22 The recent finding by Takahata et al.5 that horizontal transfer of the ftsI gene can occur in both an inter- and intra-species manner in H. influenzae is relevant to this study. It is possible that the H. parainfluenzae SF2 and SF3 have undergone ftsI gene transfer rather than both isolates undergoing independent mutation with the same resultant mutations.
Attempts to examine the effect of the PBP3 substitutions in the H. parainfluenzae isolates in isolation from the plasmid-mediated TEM-15, either by curing them of the plasmid or introducing the mutated ftsI genes into control strains, were unsuccessful. Plasmid curing is notoriously difficult in Haemophilus spp.,13,14 and in this case possibly more difficult, because attempts to select potential plasmid-free colonies were complicated by the presence of a chromosomal β-lactamase and residual reduced cefotaxime susceptibility due to the altered PBP3 in non-cured colonies.
It is not known why none of the attempts to transfer the ftsI genes was successful and in the absence of studies on transformants, the role of the ftsI mutations in β-lactam resistance (particularly to cefotaxime) in the H. parainfluenzae isolates is not proven and further work is required. However, the high degree of nucleotide (ftsI) and protein (PBP3) sequence homology between H. parainfluenzae and H. influenzae support the hypothesis.
In the absence of H. parainfluenzae ftsI transformants, the use of a BLNAR strain of H. influenzae as a recipient of pSF2 offers an opportunity to examine the degree of additional resistance produced when the TEM-15 is expressed from a background of altered PBP3. The H. influenzae BLNAR5 strain used in this study as a recipient of pSF2 has previously been characterized as having N526K and S385T (in addition to other substitutions) and the cefotaxime MIC of 1.0 mg/L was raised to 4.0 mg/L when transformed with pSF2 encoding blaTEM-15. This does not approach the cefotaxime MICs of 16.0 and 8.0 mg/L, respectively, for H. parainfluenzae SF2 and SF3. This may indicate other specific resistance mechanisms not identified in this study, such as mutations in the acrR gene, which regulates the AcrAB efflux pump,23 or alterations in PBPs other than PBP3,24 or simply reflect differences in expression and/or membrane permeability between the wild-type H. parainfluenzae SF2 and SF3 and the transformed Rd strain. Neither the acrR gene nor genes encoding other PBPs were examined in this study.
In conclusion, two H. parainfluenzae isolated in South Africa with high cefotaxime MICs have been characterized as having a plasmid-mediated TEM-15 ESBL and altered PBP3. This is the first description of an ESBL in any Haemophilus sp., and the first description of altered PBP3 in H. parainfluenzae, although the role of this latter mechanism was not conclusively demonstrated. There is potential for the plasmid-mediated ESBL to be transferred to H. influenzae, and if this occurred in BLNAR strains, which are becoming increasingly common, currently existing decreased susceptibility to cefotaxime could become complete resistance.8
|
Funding |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
This work was funded by the Clifford Craig Medical Research Trust, Launceston, Tasmania, and, in part, by the Queen Elizabeth II Health Sciences Centre Research Fund.
|
Transparency declarations |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
None to declare.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionFundingTransparency declarationsReferences |
|---|
1 Frankard J, Rodriguez-Villalobos H, Struelens MJ, et al. Haemophilus parainfluenzae: an underdiagnosed pathogen of biliary tract infections? Eur J Clin Microbiol Infect Dis (2004) 23:46–8.[CrossRef][Web of Science][Medline]
2
Leaves NI, Dimopoulou I, Hayes I, et al. Epidemiological studies of large resistance plasmids in Haemophilus. J Antimicrob Chemother (2000) 45:599–604.
3
Scheifele DW, Fussell SJ, Roberts MC. Characterization of ampicillin resistant Haemophilus influenzae. Antimicrob Agents Chemother (1982) 21:734–9.
4
Tristram SG, Jacobs MR, Appelbaum PC. Antimicrobial resistance in Haemophilus influenzae. Clin Microbiol Rev (2007) 20:368–89.
5
Takahata S, Ida T, Senju N, et al. Horizontal gene transfer of ftsI, the gene encoding penicillin binding protein 3 in Haemophilus influenzae. Antimicrob Agents Chemother (2007) 51:1589–95.
6
Tristram SG. Effect of extended spectrum β-lactamases on the susceptibility of Haemophilus influenzae to cephalosporins. J Antimicrob Chemother (2003) 51:39–43.
7
Bozdogan B, Tristram SG, Appelbaum PC. Combination of altered PBPs and expression of cloned extended spectrum β-lactamases confers cefotaxime resistance in Haemophilus influenzae. J Antimicrob Chemother (2006) 57:747–9.
8 Clinical and Laboratory Standards Institute. Performance Standards for Antimicrobial Susceptibility Testing: Sixteenth Informational Supplement M100-S16 (2007) Wayne, PA, USA: CLSI.
9 Pitout M, Macdonald K, Musgrave H, et al. Characterisation of extended spectrum β-lactamase (ESBL) activity in Haemophilus parainfluenzae. In: Abstracts of the Forty-second Interscience Conference on Antimicrobial Agents and Chemotherapy, San Diego, CA, 2002. Washington, DC, USA: American Society for Microbiology. Abstract C2-645, p. 96.
10
Dabernat H, Delmas C, Seguy M, et al. Diversity of β-lactam resistance conferring amino acid substitutions in penicillin binding protein 3 of Haemophilus influenzae. Antimicrob Agents Chemother (2002) 46:2208–18.
11
Tristram SG, Bozdogan B, Appelbaum PC. Disc diffusion-based screening tests for extended-spectrum β-lactamases in Haemophilus influenzae. J Antimicrob Chemother (2005) 55:570–3.
12
Ubukata K, Shibasaki Y, Yamamoto K, et al. Association of amino acid substitutions in penicillin-binding protein 3 with β-lactam resistance in β-lactamase-negative ampicillin-resistant Haemophilus influenzae. Antimicrob Agents Chemother (2001) 45:1693–9.
13
Elwell L, Saunders J, Richmond M, et al. Relationships among some R plasmids found in Haemophilus influenzae. J Bacteriol (1977) 131:356–62.
14 Saunders JR, Elwell LP, Falkow S, et al. β-Lactamases and R-plasmids of Haemophilus influenzae. Scand J Infect Dis (1978) 13(Suppl):16–22.
15
Windsor HM, Gromkova RC, Koornhof HJ. Plasmid mediated NAD independence in Haemophilus parainfluenzae. J Gen Microbiol (1991) 137:2415–21.
16
Chanal C, Poupart M, Sirot D, et al. Nucleotide sequences of CAZ-2, CAZ-6, and CAZ-7 β-lactamase genes. Antimicrob Agents Chemother (1992) 36:1817–20.
17 Sanger Institute 2007. Haemophilus parainfluenzae sequencing project. http://www.sanger.ac.uk/Projects/H_parainfluenzae/ (21 August 2007, date last accessed).
18 Barcak GJ, Chandler MS, Redfield RJ, et al. Genetic systems in Haemophilus influenzae. Methods Enzymol (1991) 204:321–42.[Web of Science][Medline]
19
Tristram SG, Nichols S. A multiplex PCR for β-lactamase genes of Haemophilus influenzae and description of a new blaTEM promoter variant. J Antimicrob Chemother (2006) 58:183–5.
20
Tristram SG, Hawes R, Souprounov J. Variation in selected regions of blaTEM genes and promoters in Haemophilus influenzae. J Antimicrob Chemother (2005) 56:481–4.
21
Karami N, Martner A, Enne V, et al. Transfer of an ampicillin resistance gene between two Escherichia coli strains in the bowel microbiota of an infant treated with antibiotics. J Antimicrob Chemother (2007) 60:1142–5.
22
Hasegawa K, Chiba N, Koboyashi R, et al. Rapidly increasing prevalence of β-lactamase-nonproducing, ampicillin-resistant Haemophilus influenzae type b in patients with meningitis. Antimicrob Agents Chemother (2004) 48:1509–14.
23
Kaczmarek FS, Gootz TD, Dib-Hajj F, et al. Genetic and molecular characterization of β-lactamase negative ampicillin-resistant Haemophilus influenzae with unusually high resistance to ampicillin. Antimicrob Agents Chemother (2004) 48:1630–9.
24
Fluit AC, Florijn A, Verhoef J, et al. Susceptibility of European β-lactamase-positive and -negative Haemophilus influenzae isolates from the periods 1997/1998 and 2002/2003. J Antimicrob Chemother (2005) 56:133–8.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||