Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

MALDI-TOF Mass Spectrometry for Multilocus Sequence Typing of Escherichia coli Reveals Diversity among Isolates Carrying blaCMY-2-Like Genes

  • Kaitlin A. Tagg,

    Affiliation Centre for Infectious Diseases and Microbiology, The Westmead Institute for Medical Research, The University of Sydney and Westmead Hospital, Westmead, New South Wales, Australia

    ⨯
  • Andrew N. Ginn,

    Affiliation Centre for Infectious Diseases and Microbiology, The Westmead Institute for Medical Research, The University of Sydney and Westmead Hospital, Westmead, New South Wales, Australia

    ⨯
  • Sally R. Partridge,

    Affiliation Centre for Infectious Diseases and Microbiology, The Westmead Institute for Medical Research, The University of Sydney and Westmead Hospital, Westmead, New South Wales, Australia

    ⨯
  • Jonathan R. Iredell

    jonathan.iredell@sydney.edu.au

    Affiliation Centre for Infectious Diseases and Microbiology, The Westmead Institute for Medical Research, The University of Sydney and Westmead Hospital, Westmead, New South Wales, Australia

    ⨯

Abstract

Effective surveillance and management of pathogenic Escherichia coli relies on robust and reproducible typing methods such as multilocus sequence typing (MLST). Typing of E. coli by MLST enables tracking of pathogenic clones that are known to carry virulence factors or spread resistance, such as the globally-prevalent ST131 lineage. Standard MLST for E. coli requires sequencing of seven alleles, or a whole genome, and can take several days. Here, we have developed and validated a nucleic-acid-based MALDI-TOF mass spectrometry (MS) method for MLST as a rapid alternative to sequencing that requires minimal operator expertise. Identification of alleles was 99.6% concordant with sequencing. We employed MLST by MALDI-TOF MS to investigate diversity among 62 E. coli isolates from Sydney, Australia, carrying a blaCMY-2-like gene on an IncI1 plasmid to determine whether any dominant clonal lineages are associated with the spread of this globally-disseminated resistance gene. Thirty-four known sequence types were identified, including lineages associated with human disease, animal and environmental sources. This suggests that the dissemination of blaCMY-2-like-genes is more complex than the simple spread of successful pathogenic clones. E. coli MLST by MALDI-TOF MS, employed here for the first time, can be utilised as an automated tool for large-scale population analyses or for targeted screening for known high-risk clones in a diagnostic setting.

Citation: Tagg KA, Ginn AN, Partridge SR, Iredell JR (2015) MALDI-TOF Mass Spectrometry for Multilocus Sequence Typing of Escherichia coli Reveals Diversity among Isolates Carrying blaCMY-2-Like Genes. PLoS ONE 10(11): e0143446. https://doi.org/10.1371/journal.pone.0143446

Editor: Jose Luis Balcazar, Catalan Institute for Water Research (ICRA), SPAIN

Received: September 16, 2015; Accepted: November 4, 2015; Published: November 20, 2015

Copyright: © 2015 Tagg et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited

Data Availability: All relevant data are within the paper and its Supporting Information. Related data can be made available by contacting the Corresponding Author.

Funding: KAT was supported by an Australian Postgraduate Award from the University of Sydney and a Westmead Millennium Institute Top-Up Scholarship. ANG was supported by G1046886, JRI was supported by PF1002076 and this work was supported by G1001021 and G512396, all from the National Health and Medical Research Council of Australia. Illumina sequencing was funded by a grant from the Australian Society for Antimicrobials.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Escherichia coli is an important commensal and pathogenic organism and a leading cause of urinary tract infections, diarrhoeal diseases and bacteraemia worldwide [1]. Effective infection control and management of these organisms requires reliable and robust subtyping schemes. Fine-scale pulsed-field gel electrophoresis (PFGE) is considered the ‘gold standard’ for tracking local outbreaks, as it has the power to detect variation between closely related organisms [2]. Typing based on multiple-locus variable number tandem repeat analysis (MLVA) [3, 4] is more rapid and reproducible than PFGE and is increasingly being adopted [5]. However, surveillance for organisms of particular lineages that may not necessarily be part of a specific outbreak requires typing schemes that can elucidate more distant evolutionary relationships such as multilocus sequence typing (MLST).

MLST utilises genetic polymorphisms within well-conserved chromosomal ‘housekeeping’ genes to define the general relatedness of bacterial strains. Each of these loci (seven for E. coli) may have several hundred variants that are given different allele numbers. The combination of all seven allele numbers defines the sequence type (ST) of a strain [6]. Reported STs are collated in an online MLST database (http://mlst.warwick.ac.uk/mlst/dbs/Ecoli is the most widely used for E. coli) that is updated regularly to include new allelic variants and STs. STs can be further grouped into ST complexes (STC) (clonal complexes) that contain STs that vary at only one or two alleles and are therefore deemed to be related lineages. Typing by MLST has enabled the recognition of important E. coli strain types (‘clones’) that are more virulent or likely to spread resistance, including the globally-successful ST131 clonal lineage [7, 8].

MLST by traditional Sanger sequencing is simple and highly reproducible but is relatively labour-intensive for larger sets of isolates. Whole-genome sequencing (WGS) allows extraction of MLST data with relative ease but sample preparation time and processing procedures mean that results are often not available for several weeks. Nucleic-acid-based MALDI-TOF mass spectrometry (MALDI-TOF MS) is a rapid method that has been successfully employed for MLST of bacterial pathogens, including Neisseria meningitidis and Streptococcus pneumoniae [9–11]. This method involves initial PCR amplification followed by in vitro transcription and base-specific cleavage to generate fragments small enough for detection by MS, which are then compared to a reference library, similar to protein-based MALDI-TOF MS methods [12].

Nucleic-acid-based MALDI-TOF MS is increasingly being utilised for characterisation of pathogenic organisms including single nucleotide polymorphism (SNP) detection in Staphylococcus aureus [13], identification of human papilloma virus subtypes [14] and monitoring of hepatitis B virus quasispecies during treatment [15]. The high level of automation ensures reproducibility and minimises costs as well as reducing the need for specialist technical expertise. The rapid availability of MALDI-TOF MS results can inform short-term infection control strategies and over time will enable the accumulation of valuable epidemiological information, both of which are important for improving public health outcomes.

Here we describe the development and validation of a MALDI-TOF MS method for MLST of E. coli. We employed this method to investigate clonal diversity among E. coli strains from Sydney, Australia, carrying the plasmid-borne AmpC β-lactamase (pAmpC) gene blaCMY-2 and variants [16]. blaCMY-2 is the most common pAmpC gene worldwide, particularly in E. coli, and although found on several distinct plasmid lineages (IncI1, F, A/C, K and B/O) [17–19], it is predominantly associated with IncI1 plasmids in Australia [20, 21].

Materials and Methods

Bacterial isolates

A set of 33 E. coli isolates with STs determined from Illumina MiSeq sequences using the MLST finder database (https://cge.cbs.dtu.dk/services/MLST), or by conventional Sanger sequencing (http://mlst.warwick.ac.uk/mlst/dbs/Ecoli) [22, 23], was used for validation of MLST by MALDI-TOF MS and included 10 examples of ST131 and representatives of 14 other STs. Crude bacterial lysates, prepared by suspending a loopful of colonies in 1 ml sterile distilled water and boiling for 10–20 min, were used as templates for all PCRs. Seventy-two E. coli isolates identified as carrying both a blaCMY-2-like gene and an IncI1 plasmid in previous surveys (Feb 2005-April 2014 [21, 24]) were investigated to identify those in which the blaCMY-2-like gene is located on the IncI1 plasmid. Primers linking the blaCMY-2-like gene to the IncI1 backbone [25] and/or S1 nuclease digestion/PFGE and hybridisation with IncI1 and blaCMY-2 probes (primers listed in S1 Table), as described previously [26], were used to confirm the genetic location of the blaCMY-2-like gene.

Reference libraries for MLST

Existing E. coli MLST primers are located in conserved regions flanking a smaller gene fragment used as the MLST allele. As the whole amplicon is cleaved and analysed by MALDI-TOF MS, allele libraries must include the primer sequences and the ‘intervening regions’ between the primers and the MLST allele. Reference libraries for each MLST locus were compiled from the current list of allelic variants available from the E. coli MLST website (http://mlst.warwick.ac.uk/mlst/dbs/Ecoli/Downloads_HTML; accessed January 2015). A consensus ‘intervening’ sequence for each allele was generated from ~100 E. coli sequences available in GenBank, similar to the method described previously [10].

MLST by MALDI-TOF MS

The seven target genes, adk, fumC, gyrB, icd, mdh, purA and recA, were amplified with primers based on published forward and reverse MLST primers (http://mlst.warwick.ac.uk/mlst/dbs/Ecoli/documents/primersColi_html) but with recognition sequences for either T7 (forward primer) or SP6 RNA polymerase (reverse primer) included (S1 Table). PCRs were performed in a final volume of 10 μl (per allele) with 0.02 U HotStarTaq® DNA polymerase (Qiagen, Doncaster, Australia), 1 x PCR buffer, 200 nM each dNTP, 200 nM of each of the forward and reverse primers and 2 μl of crude lysate as template (1:10 dilution). PCR conditions were: activation at 95°C for 15 min; 45 cycles of 95°C for 20 s, 60°C for 30 s, 72°C for 60 s; final extension at 72°C for three min. Amplicons were treated with shrimp alkaline phosphatase (SAP) (Agena Bioscience, San Diego, USA) to remove unincorporated dNTPs according to the manufacturer’s instructions.

Following SAP treatment, simultaneous reverse-transcription and base-specific cleavage reactions were performed according to the MassCLEAVE protocol (Agena Bioscience) [9]. In brief, each amplicon was subjected to four separate in vitro transcription and base-specific cleavage reactions, resulting in fragments cleaved in both forward and reverse directions at either T- (U-) or C- nucleotides [12]. Products of each reaction were desalted using anion-exchange resin and 8–12 nl dispensed onto a 96-well matrix-coated SpectroCHIP using the MassARRAY™ RS1000 Nanodispenser and analysed on a MassARRAY™ Analyzer 4 using SpectroAcquire 4.0 (Agena Bioscience). Resulting spectra were analysed using iSEQ software 1.0 (Agena Bioscience), which compares sample spectra to the spectral library generated from in silico cleavage of the allelic variants in the reference libraries. Allele numbers are assigned automatically based on the ‘best match’, calculated by an iterative process of analysing spectral features such as missing and additional peaks to produce a list of allelic variants ranked by ‘confidence score’ [9]. Samples with more than one best match (denoted with an asterisk by the iSEQ software) or with a confidence score ≤0.9 were manually reviewed (comparison of the top three highest ranked matches) for spectral features that supported assignment of a single best match. DNA sequencing was conducted if manual review was inconclusive. Allele combinations for each sample were entered into the MLST database for ST assignment. The repeatability of MLST by MALDI-TOF MS was evaluated using 10 of the isolates used for validation, representing diverse STs, repeated in triplicate.

Discriminatory power of three versus seven alleles

Simpson’s index of diversity (D) is a measure of diversity in a population of n individuals and can be used as an indicator of the discriminatory power of different typing methods [27, 28]. We calculated D when all seven alleles were used and compared it to a three-allele approach, to determine whether the discriminatory power of MLST is affected when the number of alleles used to measure clonal diversity is reduced. The set of E. coli isolates carrying blaCMY-2 on an IncI1 plasmid was used to calculate D and confidence intervals (CI) were calculated as previously described [28].

Results

Evaluation of MLST by MALDI-TOF MS

Thirty-three E. coli isolates of known ST were used to evaluate the MALDI-TOF MS MLST method. MALDI-TOF MS assigned the expected allele number at all seven loci for 32/33 validation isolates and at six loci for the remaining isolate, corresponding to 230/231 correctly assigned alleles (99.6%). The single discrepant allele was assigned icd-306 instead of the expected icd-16. These two alleles differ by only one SNP and have almost identical spectral features (Table 1).

thumbnail
Download:
Table 1. Examples of discrepancies between MLST by MALDI-TOF MS and sequencing.

https://doi.org/10.1371/journal.pone.0143446.t001

Repeatability

Ten isolates representing diverse STs were used to assess repeatability, but one allele from one sample failed MS repeatedly, generating a total of 69 alleles repeated in triplicate. Only four of the 69 alleles (two fumC, two gyrB) were not assigned the same number in all three replicates. In all cases, the software had assigned either multiple top matches or a single top match with a confidence score below 0.9, but manual review was unable to resolve a single best match. Repeatability was therefore calculated as 94.2% (65/69) overall, with discrepancies in fumC or gyrB only.

Refinement of references libraries

In three samples recA was assigned recA-325 by MALDI-TOF MS but sequencing indicated that recA-2 was present. Manual inspection revealed that the reason for this discrepancy was a SNP in the left-hand intervening region that was not represented in the consensus sequence as it was present in only 20% of E. coli sequences analysed. The recA library was updated to include an additional recA-2 entry with the identified SNP in the intervening region and the three discrepant samples were reanalysed. All three recA alleles were then correctly identified as recA-2, highlighting the need for iterative improvement of the reference libraries. A similar update was made to the fumC reference library where an additional fumC-4 entry with a modified intervening region resulted in confident assignment of fumC-4 for two samples that were initially assigned multiple best matches (fumC-4 and -315). Two additional mdh entries (mdh-8 and -11) with modified intervening regions improved the power of the software to discriminate between mdh-8 and -283 and mdh-11 and -264, since these pairs of allele variants have nearly identical spectra.

MLST of E. coli carrying IncI1-blaCMY-2 plasmids

PCR and/or S1 nuclease digestion/PFGE and Southern hybridisation were used to determine the genetic location of the blaCMY-2-like gene in 72 E. coli isolates identified as carrying both a blaCMY-2-like gene and an IncI1 plasmid in previous surveys [21, 24]. The blaCMY-2-like gene was confirmed to be located on the IncI1 plasmid in 62 of these isolates (S2 Fig). MLST by MALDI-TOF MS was used to investigate the clonality of these 62 isolates after refinements were made to the reference libraries, as discussed above. As a secondary validation, 10 isolates representing different STs to those in the initial validation set were also subjected to DNA sequencing as well as MLST by MALDI-TOF MS. The allele numbers obtained from sequencing were 100% concordant with those assigned by MALDI-TOF MS for all 10 isolates.

The remaining 52 isolates were subjected to MLST by MALDI-TOF MS only. Forty-five were assigned a known ST and seven were assigned a combination of alleles that did not correspond to a known ST, prompting manual inspection to confirm the assignment of each allele. Any alleles that were not confidently assigned the same number upon manual review were sequenced. Three of the seven isolates were subsequently assigned to a known ST after a single allele from each isolate was found to be misassigned by MALDI-TOF MS (one fumC, one icd and one mdh–Table 1). Two isolates were confirmed to have new STs (one with a new combination of alleles and one with a new gyrB variant) and two were found to be mixed at one or more loci, representing either heterogeneity at these loci or mixed samples containing more than one clone. These two mixed isolates were excluded from further analysis.

In total, 58/60 isolates were assigned to a known ST, corresponding to 34 different STs. Thirteen STs were represented more than once, but no ST was represented more than six times (Table 2). The most prevalent STC was STC38 (9/60) followed by STC10 and STC23 (both 5/60), and STC648 (3/60), but no single STC appeared to cluster temporally.

thumbnail
Download:
Table 2. Sequence types of E. coli strains carrying IncI1-blaCMY-2 plasmids.

https://doi.org/10.1371/journal.pone.0143446.t002

Discriminatory power of three versus seven alleles

Simpson’s index of diversity (D) was used to assess the utility of three alleles instead of seven to measure clonal diversity in the set of isolates carrying a blaCMY-2-like gene on an IncI1 plasmid (n = 60, excluding two mixed samples) [27, 28]. We selected adk, fumC and recA as they have previously been shown to be the most discriminatory alleles individually for determining ST [29] and the iSEQ simulation software can confidently differentiate the allelic variants of adk, fumC and recA (S2 Table). The discriminatory power of three alleles (D = 0.959, CI = 0.934–0.983) was not substantially different to that of all seven alleles (D = 0.975, CI = 0.960–0.989) (confidence intervals overlap) when used to measure the clonal diversity of isolates in this set.

Discussion

Rapid and reproducible typing of E. coli is essential for effective infection control and for surveillance of known virulent and/or resistant subtypes. Here we showed that full E. coli MLST by MALDI-TOF MS is over 99% concordant with conventional MLST by DNA sequencing. Results are obtainable within 12 hours and the running costs for MALDI-TOF MS (approx. AUD$90/isolate) are comparable to bidirectional Sanger sequencing (approx. AUD$95/isolate) and currently lower than WGS (approx. AUD$155/isolate using Illumina MiSeq technology). While an initial investment in specialist equipment is required (approx. AUD$250,000), the MALDI-TOF MS has proven a valuable tool for a variety of diagnostic applications [13–15].

The utility of MLST by MALDI-TOF MS lies in its speed, high-throughput capacity and automated analysis that requires minimal operator expertise, enabling both large-scale population analyses and targeted screening for known pathogenic lineages in a diagnostic setting, with simple boiled lysates as the template. A simplified three-allele approach (adk, fumC, recA) would further reduce the cost (approx. AUD$40/isolate) and processing time per sample and in the population investigated here, the discriminatory power of three alleles was not substantially different to that of seven alleles, suggesting that this approach may suffice in some settings as a measure of population diversity.

The limitations of this method include a small error rate of less than 1% but errors can be detected if an isolate is assigned an unknown combination of alleles, prompting manual review and confirmatory sequencing. With iterative refinement of the reference libraries, including modification of the intervening consensus sequences and periodic updates to include new allelic variants, the error rate is expected to be progressively reduced. Furthermore, the incorrect allele number assignments that did occur were only one SNP different to the correct allele number and thus the sample was still assigned to the correct STC. Manual inspection was required for approximately 20% of alleles (either prompted by the software or with a confidence score of ≤0.9), but this process is relatively quick and by ensuring that adequate amounts of DNA (5 ng/ μl) are used in the cleavage reactions and 8–12 nl of cleaved product is dispensed onto the SpectroCHIP, the need for manual review may be reduced. Finally, new SNPs can be identified using this method [9] but must be confirmed by sequencing as new alleles can only be submitted to the MLST website if the isolate is sequenced by WGS methods.

Using MALDI-TOF MS MLST, we identified a high level of clonal diversity (D = 0.975) among E. coli strains from Sydney, Australia carrying a blaCMY-2-like gene on an IncI1 plasmid. Some STs identified here have been reported in association with blaCMY-2-like genes previously, such as ST131 and members of STC10 and STC38, and are associated with human disease [30–33], while other STs have been reported in isolates of animal or environmental origin (http://mlst.warwick.ac.uk/mlst/dbs/Ecoli/GetTableInfo_html). This pattern of polyclonality has been reported among E. coli carrying blaCMY-2 in other parts of the world [30, 31, 34] and reflects an interesting difference in the dynamics of spread compared with the more prominent blaCTX-M-15, which appears to have spread predominantly via its association with IncF plasmids in ST131 [35].

In conclusion, E. coli MLST by MALDI-TOF MS is a rapid alternative to DNA sequencing with minimal operator expertise required and when implemented in a clinical setting has the capacity to identify the spread of high-risk pathogenic lineages. We identified substantial clonal diversity among E. coli carrying blaCMY-2 on IncI1 plasmids from Sydney, Australia, implicating horizontal gene transfer rather than clonal expansion as an important dynamic in the dissemination of this resistance gene.

Supporting Information

S1 Fig. Workflow for the assignment of Escherichia coli sequence types by MALDI-TOF MS.

https://doi.org/10.1371/journal.pone.0143446.s001

(DOCX)

S2 Fig. S1 nuclease digestion pulsed-field gel electrophoresis (PFGE) and Southern hybridisation.

https://doi.org/10.1371/journal.pone.0143446.s002

(DOCX)

S1 Table. Primers used in this study.

https://doi.org/10.1371/journal.pone.0143446.s003

(DOCX)

S2 Table. Simulation results for all seven Escherichia coli MLST alleles.

https://doi.org/10.1371/journal.pone.0143446.s004

(DOCX)

S3 Table. gyrB allele pairs that differ by 10,290 Da peak.

https://doi.org/10.1371/journal.pone.0143446.s005

(DOCX)

Acknowledgments

We thank Mark Smith for technical assistance. KAT was supported by an Australian Postgraduate Award from the University of Sydney and a Westmead Millennium Institute Top-Up Scholarship. ANG was supported by G1046886, JI was supported by PF1002076 and this work was supported by G1001021 and G512396, all from the National Health and Medical Research Council of Australia. Illumina sequencing was funded by a grant from the Australian Society for Antimicrobials.

Author Contributions

Conceived and designed the experiments: KAT ANG SRP JRI. Performed the experiments: KAT ANG. Analyzed the data: KAT ANG. Contributed reagents/materials/analysis tools: SRP JRI. Wrote the paper: KAT ANG SRP JRI.

References

  1. 1. Kaper JB, Nataro JP, Mobley HLT. Pathogenic Escherichia coli. Nat Rev Microbiol. 2004;2:123–40. pmid:15040260
  2. 2. Goering RV. Pulsed field gel electrophoresis: a review of application and interpretation in the molecular epidemiology of infectious disease. Infect Genet Evol. 2010;10:866–75. pmid:20692376
  3. 3. Lindstedt BA, Brandal LT, Aas L, Vardund T, Kapperud G. Study of polymorphic variable-number of tandem repeats loci in the ECOR collection and in a set of pathogenic Escherichia coli and Shigella isolates for use in a genotyping assay. J Microbiol Methods. 2007;69:197–205. pmid:17291612
  4. 4. Løbersli I, Haugum K, Lindstedt BA. Rapid and high resolution genotyping of all Escherichia coli serotypes using 10 genomic repeat-containing loci. J Microbiol Methods. 2012;88:134–9. pmid:22088357
  5. 5. Naseer U, Olsson-Liljequist BE, Woodford N, Dhanji H, Cantón R, Sundsfjord A, et al. Multi-locus variable number of tandem repeat analysis for rapid and accurate typing of virulent multidrug resistant Escherichia coli clones. PLoS ONE. 2012;7:e41232. pmid:22859970
  6. 6. Wirth T, Falush D, Lan R, Colles F, Mensa P, Wieler LH, et al. Sex and virulence in Escherichia coli: an evolutionary perspective. Mol Microbiol. 2006;60:1136–51. pmid:16689791
  7. 7. Woodford N, Turton JF, Livermore DM. Multiresistant Gram-negative bacteria: the role of high-risk clones in the dissemination of antibiotic resistance. FEMS Microbiol Rev. 2011;35:736–55. pmid:21303394
  8. 8. Riley LW. Pandemic lineages of extraintestinal pathogenic Escherichia coli. Clin Microbiol Infect. 2014;20:380–90. pmid:24766445
  9. 9. Honisch C, Chen Y, Mortimer C, Arnold C, Schmidt O, van den Boom D, et al. Automated comparative sequence analysis by base-specific cleavage and mass spectrometry for nucleic acid-based microbial typing. PNAS. 2007;104:10649–54. pmid:17563374
  10. 10. Dunne EM, Ong EK, Moser RJ, Siba PM, Phuanukoonnon S, Greenhill AR, et al. Multilocus sequence typing of Streptococcus pneumoniae by use of mass spectrometry. J Clin Microbiol. 2011;49:3756–60. pmid:21880964
  11. 11. Groves MD, Jordan D, Chapman TA, Jassim RA. Multilocus sequence typing of Australian Streptococcus suis type 2 by MALDI-TOF mass spectrometry analysis of PCR amplicons. Vet Microbiol. 2015;177:394–7. pmid:25862280
  12. 12. Stanssens P, Zabeau M, Meersseman G, Remes G, Gansemans Y, Storm N, et al. High-throughput MALDI-TOF discovery of genomic sequence polymorphisms. Genome Res. 2004;14:126–33. pmid:14707174
  13. 13. Syrmis MW, Moser RJ, Whiley DM, Vaska V, Coombs GW, Nissen MD, et al. Comparison of a multiplexed MassARRAY system with real-time allele-specific PCR technology for genotyping of methicillin-resistant Staphylococcus aureus. Clin Microbiol Infect. 2011;17:1804–10. pmid:21595795
  14. 14. Söderlund-Strand A, Dillner J, Carlson J. High-throughput genotyping of oncogenic human papilloma viruses with MALDI-TOF mass spectrometry. Clin Chem. 2008;54:86–92. pmid:17981923
  15. 15. Rybicka M, Stalke P, Bielawski KP. Dynamics of hepatitis B virus quasispecies heterogeneity in association with nucleos(t)ide analogue treatment determined by MALDI-TOF MS. Clin Microbiol Infect. 2015;21:288.e1-4. pmid:25658547
  16. 16. Jacoby GA. AmpC β-lactamases. Clin Microbiol Rev. 2009;22:161–82. pmid:19136439
  17. 17. Börjesson S, Jernberg C, Brolund A, Edquist P, Finn M, Landén A, et al. Characterisation of plasmid-mediated AmpC-producing E. coli from Swedish broilers and association with human clinical isolates. Clin Microbiol Infect. 2013;19:E309–11. pmid:23581796
  18. 18. Mataseje LF, Baudry PJ, Zhanel GG, Morck DW, Read RR, Louie M, et al. Comparison of CMY-2 plasmids isolated from human, animal, and environmental Escherichia coli and Salmonella spp. from Canada. Diagn Microbiol Infect Dis. 2010;67:387–91. pmid:20638610
  19. 19. Voets GM, Fluit AC, Scharringa J, Schapendonk C, van den Munckhof T, Leverstein-van Hall MA, et al. Identical plasmid AmpC β-lactamase genes and plasmid types in E. coli isolates from patients and poultry meat in the Netherlands. Int J Food Microbiol. 2013;167:359–62. pmid:24184615
  20. 20. Sidjabat HE, Seah KY, Coleman L, Sartor A, Derrington P, Heney C, et al. Expansive spread of IncI1 plasmids carrying blaCMY-2 amongst Escherichia coli. Int J Antimicrob Agents. 2014;44:203–8. pmid:25052868
  21. 21. Tagg KA, Ginn AN, Jiang X, Ellem J, Partridge SR, Iredell JR. Distribution of acquired AmpC β-lactamase genes in Sydney, Australia. Diagn Microbiol Infect Dis. 2015;83:56–8. pmid:26099646
  22. 22. Kamruzzaman M, Shoma S, Bari SMN, Ginn AN, Wiklendt AM, Partridge SR, et al. Genetic diversity and antibiotic resistance in Escherichia coli from environmental surface water in Dhaka City, Bangladesh. Diagn Microbiol Infect Dis. 2013;76:222–6. pmid:23541688
  23. 23. Partridge SR, Zong Z, Iredell JR. Recombination in IS26 and Tn2 in the evolution of multiresistance regions carrying blaCTX-M-15 on conjugative IncF plasmids from Escherichia coli. Antimicrob Agents Chemother. 2011;55:4971–8. pmid:21859935
  24. 24. Ginn AN, Zong Z, Wiklendt AM, Thomas LC, Merlino J, Gottlieb T, et al. Limited diversity in the gene pool allows prediction of third-generation cephalosporin and aminoglycoside resistance in Escherichia coli and Klebsiella pneumoniae. Int J Antimicrob Agents. 2013;42:19–26. pmid:23706544
  25. 25. Tagg KA, Iredell JR, Partridge SR. Complete sequencing of IncI1 sequence type 2 plasmid pJIE512b indicates mobilization of blaCMY-2 from an IncA/C plasmid. Antimicrob Agents Chemother. 2014;58:4949–52. pmid:24890591
  26. 26. Partridge SR, Ginn AN, Wiklendt AM, Ellem J, Wong JSJ, Ingram P, et al. Emergence of blaKPC carbapenemase genes in Australia. Int J Antimicrob Agents. 2015;45:130–6. pmid:25465526
  27. 27. Hunter PR, Gaston MA. Numerical index of the discriminatory ability of typing systems: an application of Simpson's index of diversity. J Clin Microbiol. 1988;26:2465–6. pmid:3069867
  28. 28. Grundmann H, Hori S, Tanner G. Determining confidence intervals when measuring genetic diversity and the discriminatory abilities of typing methods for microorganisms. J Clin Microbiol. 2001;39:4190–2. pmid:11682558
  29. 29. Weissman SJ, Johnson JR, Tchesnokova V, Billig M, Dykhuizen D, Riddell K, et al. High-resolution two-locus clonal typing of extraintestinal pathogenic Escherichia coli. Appl Environ Microbiol. 2012;78:1353–60. pmid:22226951
  30. 30. Weissman SJ, Adler A, Qin X, Zerr DM. Emergence of extended-spectrum β-lactam resistance among Escherichia coli at a US academic children's hospital is clonal at the sequence type level for CTX-M-15, but not for CMY-2. Int J Antimicrob Agents. 2013;41:414–20. pmid:23434250
  31. 31. Naseer U, Haldorsen B, Simonsen GS, Sundsfjord A. Sporadic occurence of CMY-2-producing multidrug-resistant Escherichia coli of ST-complexes 38 and 448, and ST131 in Norway. Clin Microbiol Rev. 2010;16:171–8.
  32. 32. Oteo J, Cercenado E, Cuevas O, Bautista V, Delgado-Iribarren A, Orden B, et al. AmpC β-lactamases in Escherichia coli: emergence of CMY-2-producing virulent phylogroup D isolates belonging mainly to STs 57, 115, 354, 393, and 420, and phylogroup B2 isolates belonging to the international clone O25b-ST131. Diagn Microbiol Infect Dis. 2010;67:270–6. pmid:20462723
  33. 33. Manges AR, Harel J, Masson L, Edens TJ, Portt A, Reid-Smith RJ, et al. Multilocus sequence typing and virulence gene profiles associated with Escherichia coli from human and animal sources. Foodborne Pathog Dis. 2015;12:302–10. pmid:25774654
  34. 34. Matsumura Y, Nagao M, Iguchi M, Yagi T, Komori T, Fujita N, et al. Molecular and clinical characterization of plasmid-mediated AmpC β-lactamase-producing Escherichia coli bacteraemia: a comparison with extended-spectrum β-lactamase-producing and non-resistant E. coli bacteraemia. Clin Microbiol Infect. 2013;19:161–8. pmid:22332968
  35. 35. Cantón R, González-Alba JM, Galán JC. CTX-M enzymes: origin and diffusion. Front Microbiol. 2012;3:110. pmid:22485109