- Email: [email protected]
Whole-genome mapping for high-resolution genotyping of Pseudomonas aeruginosa
Journal of Microbiological Methods 106 (2014) 19–22
Contents lists available at ScienceDirect
Journal of Microbiological Methods journal homepage: www.elsevier.com/locate/jmicmeth
Note
Whole-genome mapping for high-resolution genotyping of Pseudomonas aeruginosa S.A. Boers a, R. Burggrave b, M. van Westreenen a, W.H.F. Goessens a, J.P. Hays a,⁎ a b
Department of Medical Microbiology and Infectious Diseases, Erasmus University Medical Centre, Rotterdam, The Netherlands Piext BV., Rosmalen, The Netherlands
a r t i c l e
i n f o
Article history: Received 7 June 2014 Received in revised form 20 July 2014 Accepted 20 July 2014 Available online 10 August 2014
a b s t r a c t A variety of molecular typing techniques have been developed to investigate the clonal relationship among bacterial isolates, including those associated with nosocomial infections. In this study, the authors evaluated whole-genome mapping as a tool to investigate the genetic relatedness between Pseudomonas aeruginosa isolates, including metallo beta-lactamase-positive outbreak isolates. © 2014 Elsevier B.V. All rights reserved.
Keywords: Pseudomonas aeruginosa Whole-genome mapping Molecular typing Nosocomial infection
Pseudomonas aeruginosa is an opportunistic pathogen that is a major cause of nosocomial infections, it being associated with significant morbidity and mortality in critically ill and immune-compromised patients (Lyczak et al., 2000). These infections are often hard to treat due to the intrinsic and acquired resistance of P. aeruginosa against many different types of antibiotics (Breidenstein et al., 2011; Maltezou, 2009). It is therefore extremely important to apply timely and correct infection prevention and control measures in hospitals and care centres in order to monitor and prevent the spread of these highly resistant microorganisms. A variety of molecular genotyping techniques have been developed to investigate the genetic relatedness among bacterial pathogens, generating information useful in helping trace, monitor, and prevent the source of nosocomial outbreaks of infection. A relatively new typing technique that has become available is whole-genome mapping (WGM), which generates high-resolution, ordered, restriction maps that span the entire genome of bacteria (Miller, 2013). Further, WGM has been previously utilized to investigate closely related isolates of Escherichia coli (Chen et al., 2006; Kotewicz et al., 2008) and Staphylococcus aureus (Bosch et al., 2013; Clarridge et al., 2013), though no information has been published regarding the use of WGM for genotyping P. aeruginosa isolates. Therefore, a study was conducted in order to evaluate the use of WGM to discriminate between
⁎ Corresponding author at: Erasmus University Medical Centre, Rotterdam, The Netherlands. Tel.: +31 107032177; fax: +31 107033875. E-mail address: [email protected] (J.P. Hays).
http://dx.doi.org/10.1016/j.mimet.2014.07.020 0167-7012/© 2014 Elsevier B.V. All rights reserved.
clinically relevant P. aeruginosa isolates, including nosocomial outbreakassociated isolates. A total of 32 P. aeruginosa isolates were selected for this study. Twenty of these isolates were clinical isolates collected between 2009 and 2011, comprising 19 isolates from The Netherlands (P2–P38) and a single isolate from Canada (P17). Eight of these isolates had been previously found to be metallo beta-lactamase (MBL)-positive by PCR (Pitout et al., 2005). The corresponding whole-genome maps were prepared by OpGen, Inc. (Gaithersburg, Maryland, USA) using the BamHI restriction enzyme, as previously described (Latreille et al., 2007). An extra twelve whole-genome maps were generated from publicly available whole-genome sequences of the following P. aeruginosa strains: 19BR (GenBank: AFXJ01000001), B136-33 (CP004061), c7747m (CP006728), DK2 (CP003149), LESB58 (FM209186), M18 (CP002496), MTB-1 (CP006853), NCGM2.S1 (AP012280), PA7 (CP000744), PAO1 (AE004091), RP73 (CP006245), and UCBPP-PA14 (CP000438). This is a convenient method for increasing the number of WGM isolate maps available for analysis and helps place newly WGM mapped bacterial isolates in the context of global isolates and outbreaks. The resulting in vitro and in silico P. aeruginosa whole-genome maps were aligned and compared using BioNumerics v7 software (Applied Maths, Sint-Martens-Latem, Belgium). Restriction fragments smaller than 3000 bp were excluded from the analysis and a relative tolerance of 15%, combined with an absolute tolerance of 3000 bp, was used to compensate for WGM resolution. All dendrograms were generated using UPGMA with a map distance of 5% (N95% similarity) as a cut-off point to define a WGM cluster, as reported previously (Bosch et al.,
20
S.A. Boers et al. / Journal of Microbiological Methods 106 (2014) 19–22
2013). Importantly, the BioNumerics v7 software allows two different approaches to be made when comparing whole-genome map similarities: 1) alignment-based comparison and 2) pattern-based comparison. The alignment-based approach uses a size tolerance algorithm for pairwise
comparison of the WGM fragments, whereas the pattern-based approach takes into account a neighbourhood of fragments, matching within a given size tolerance. To explore both methods, we applied them to our set of 32 whole-genome maps. As shown in Fig. 1, at the 95% similarity
Fig. 1. Differences observed between alignment-based and pattern-based algorithms when comparing WGM data derived from 32 P. aeruginosa strains. The alignment-based comparison (top) shows different clustering results for the in silico digested isolates PAO1, RP73, 19BR, c7747m and NCGM2.S1 (indicated in red) compared to the pattern-based comparison (bottom). Close inspection of the whole-genome maps reveals that chromosomal inversions among these strains are causing the clustering differences. Three clusters (I–III) can be identified using a similarity cut-off of 95%. The clustering of groups I–III isolates is similar using both alignment-based and pattern-based algorithms. * = metallo beta-lactamase PCR-positive isolates from Erasmus MC.
S.A. Boers et al. / Journal of Microbiological Methods 106 (2014) 19–22
level, both techniques generated three clusters. Cluster I consisted of eight isolates, P28, P29, P19, P15, P17, P13, P23 and P11, that all belonged to a previously reported outbreak at Erasmus MC (Van der Bij et al., 2011). Cluster II comprised the P14 and P20 isolates that are related to the previously reported and highly virulent UCBPP-PA14 strain. Information that may be useful to clinicians with respect to enhanced monitoring of patients infected with highly virulent P. aeruginosa strains. However, the two comparisons generated different clustering results dependant on the presence of genomic rearrangements. For example, a large chromosomal inversion (LCI) has been previously reported in the P. aeruginosa reference strain PAO1 (Lee et al., 2006), with close inspection of our whole-genome maps also revealing the presence of LCIs within in silico digested P. aeruginosa RP73, 19BR, c7747m and NCGM2.S1 genomes. The presence of these LCIs can lead to different interpretations of the genetic relatedness among P. aeruginosa isolates. To validate the use of WGM for P. aeruginosa, we compared our WGM clustering results with MLST results generated using the Highthroughput Multilocus Sequence Typing (HiMLST) technique (Boers et al., 2012). As shown in Fig. 2, both typing methods generated very similar clustering results that differed for only a single isolate, namely P. aeruginosa strain P33. However, this result reveals the limits of sensitivity of MLST typing, as isolate P33 appears to be identical to Cluster I MLST isolates, but possesses b95% similarity using WGM results, and is therefore not identical, but in fact closely related, to MLST Cluster I isolates. This type of information could have consequences for infection control outbreaks and programmes. The ordered structure of the whole-genome maps generated by WGM potentially allows for the rapid identification of specific, and diagnostic, genomic elements. To explore this possibility, we generated a multiple alignment to visualise the presence and absence of restriction fragments among the whole-genome maps. In order to simplify the analysis, the five in silico whole-genome maps that harboured LCIs, as
21
well as the in silico restriction map of the taxonomic outlier strain PA7 were excluded. The alignment of the remaining 26 whole-genome maps is presented in Fig. 3, highlighting some of the clusterdiscriminating fragments. As shown in Fig. 3, a single and unique discriminating fragment of ~61 kb in size could be identified among Cluster I WGM alignment isolates. This ~ 61 kb fragment could not be mapped to in silico WGM maps, including an additional 233 unfinished sequenced genomes (scaffolds/contigs) of P. aeruginosa in GenBank, and may potentially represent an unrecognised pathogenicity island in this cluster of isolates. Therefore, the use of WGM mapping allows rapid and easy visualization of cluster-specific elements that can be used as markers for high-throughput screening during ongoing nosocomial outbreaks of P. aeruginosa. Also, WGM offers the possibility of developing cluster-specific PCRs. In this short note, the authors demonstrate that WGM is an accurate and useful tool to discriminate P. aeruginosa isolates, which possesses superior discriminative power compared to traditional techniques for genotyping bacteria such as MLST. Further, the data can be reliably utilized by networks of researchers, and WGM provides useful information on the position and order of genome rearrangements and pathogenicity islands within the P. aeruginosa genome. However, the choice of restriction enzyme is likely important in enabling the standardisation of comparative WGM between research centres. Thus, whole-genome mapping may provide a suitable alternative to currently available bacterial genotyping techniques for clinicians and microbiologists interested in the comparative analysis of P. aeruginosa isolates in, for example, hospital outbreaks associated with this bacterial pathogen. The authors thank Trevor Wagner from OpGen, Inc. (Gaithersburg, Maryland, USA) for performing the WGM experiments and critical reading of the manuscript. This project was partially funded by an FP7 European Community grant (TEMPOtest-QC, HEALTH2009-241742).
Fig. 2. Dendrogram of 20 P. aeruginosa isolates based on concatenated MLST results. From this comparison, three clusters (I–III) can be identified with 100% similarity, which reveal high concordance with WGM clustering. * = metallo beta-lactamase PCR-positive isolates from Erasmus MC.
22
S.A. Boers et al. / Journal of Microbiological Methods 106 (2014) 19–22
Fig. 3. Multiple alignment of WGM data derived from 20 clinical P. aeruginosa strains and 6 in silico digested isolates. The whole-genome maps of isolates that possess similar chromosomal arrangements were compared and aligned to visualise the presence and absence of restriction fragments (above). A discriminating ~61 kb insert can be observed (bottom left) among Cluster I, as well as the closely related P33 isolate (92.5% similarity). * = metallo beta-lactamase PCR-positive isolates from Erasmus MC.
References Boers, S.A., van der Reijden, W.A., Jansen, R., 2012. High-throughput multilocus sequence typing: bringing molecular typing to the next level. PLoS One 7, e39630. Bosch, T.,Verkade, E.,van Luit, M.,Pot, B.,Vauterin, P., et al., 2013. High resolution typing by whole genome mapping enables discrimination of LA-MRSA (CC398) strains and identification of transmission events. PLoS One 8, e66493. Breidenstein, E.B.M., de la Fuente-Nunez, C., Hancock, R.E.W., 2011. Pseudomonas aeruginosa: all roads lead to resistance. Trends Microbiol. 19, 419–426. Chen, Q., Savarino, S.J., Venkatesan, M.M., 2006. Subtractive hybridization and optical mapping of the enterotoxigenic Escherichia coli H10407 chromosome: isolation of unique sequences and demonstration of significant similarity to the chromosome of E. coli K-12. Microbiology 152, 1041–1054. Clarridge 3rd, J.E.,Harrington, A.T.,Roberts, M.C.,Soge, O.O.,Maquelin, K., 2013. Impact of strain typing methods on assessment of relationship between paired nares and wound isolates of methicillin-resistant Staphylococcus aureus. J. Clin. Microbiol. 51, 224–231. Kotewicz, M.L., Mammel, M.K., LeClerc, J.E., Cebula, T.A., 2008. Optical mapping and 454 sequencing of Escherichia coli O157: H7 isolates linked to the US 2006 spinachassociated outbreak. Microbiology 154, 3518–3528.
Latreille, P., Norton, S., Goldman, B.S., Henkhaus, J., Miller, N., et al., 2007. Optical mapping as a routine tool for bacterial genome sequence finishing. BMC Genomics 8, 321. Lee, D.G., Urbach, J.M., Wu, G., Liberati, N.T., Feinbaum, R.L., et al., 2006. Genomic analysis reveals that Pseudomonas aeruginosa virulence is combinatorial. Genome Biol. 7. Lyczak, J.B., Cannon, C.L., Pier, G.B., 2000. Establishment of Pseudomonas aeruginosa infection: lessons from a versatile opportunist. Microbes Infect. 2, 1051–1060. Maltezou, H.C., 2009. Metallo-beta-lactamases in Gram-negative bacteria: introducing the era of pan-resistance? Int. J. Antimicrob. Agents 33 (405), e401–e407. Miller, J.M., 2013. Whole-genome mapping: a new paradigm in strain-typing technology. J. Clin. Microbiol. 51, 1066–1070. Pitout, J.D.D., Gregson, D.B., Poirel, L., McClure, J.A., Le, P., et al., 2005. Detection of Pseudomonas aeruginosa producing metallo-beta-lactamases in a large centralized laboratory. J. Clin. Microbiol. 43, 3129–3135. Van der Bij, A.K.,Van Mansfeld, R.,Peirano, G.,Goessens, W.H.,Severin, J.A., et al., 2011. First outbreak of VIM-2 metallo-beta-lactamase-producing Pseudomonas aeruginosa in The Netherlands: microbiology, epidemiology and clinical outcomes. Int. J. Antimicrob. Agents 37, 513–518.
Contents lists available at ScienceDirect
Journal of Microbiological Methods journal homepage: www.elsevier.com/locate/jmicmeth
Note
Whole-genome mapping for high-resolution genotyping of Pseudomonas aeruginosa S.A. Boers a, R. Burggrave b, M. van Westreenen a, W.H.F. Goessens a, J.P. Hays a,⁎ a b
Department of Medical Microbiology and Infectious Diseases, Erasmus University Medical Centre, Rotterdam, The Netherlands Piext BV., Rosmalen, The Netherlands
a r t i c l e
i n f o
Article history: Received 7 June 2014 Received in revised form 20 July 2014 Accepted 20 July 2014 Available online 10 August 2014
a b s t r a c t A variety of molecular typing techniques have been developed to investigate the clonal relationship among bacterial isolates, including those associated with nosocomial infections. In this study, the authors evaluated whole-genome mapping as a tool to investigate the genetic relatedness between Pseudomonas aeruginosa isolates, including metallo beta-lactamase-positive outbreak isolates. © 2014 Elsevier B.V. All rights reserved.
Keywords: Pseudomonas aeruginosa Whole-genome mapping Molecular typing Nosocomial infection
Pseudomonas aeruginosa is an opportunistic pathogen that is a major cause of nosocomial infections, it being associated with significant morbidity and mortality in critically ill and immune-compromised patients (Lyczak et al., 2000). These infections are often hard to treat due to the intrinsic and acquired resistance of P. aeruginosa against many different types of antibiotics (Breidenstein et al., 2011; Maltezou, 2009). It is therefore extremely important to apply timely and correct infection prevention and control measures in hospitals and care centres in order to monitor and prevent the spread of these highly resistant microorganisms. A variety of molecular genotyping techniques have been developed to investigate the genetic relatedness among bacterial pathogens, generating information useful in helping trace, monitor, and prevent the source of nosocomial outbreaks of infection. A relatively new typing technique that has become available is whole-genome mapping (WGM), which generates high-resolution, ordered, restriction maps that span the entire genome of bacteria (Miller, 2013). Further, WGM has been previously utilized to investigate closely related isolates of Escherichia coli (Chen et al., 2006; Kotewicz et al., 2008) and Staphylococcus aureus (Bosch et al., 2013; Clarridge et al., 2013), though no information has been published regarding the use of WGM for genotyping P. aeruginosa isolates. Therefore, a study was conducted in order to evaluate the use of WGM to discriminate between
⁎ Corresponding author at: Erasmus University Medical Centre, Rotterdam, The Netherlands. Tel.: +31 107032177; fax: +31 107033875. E-mail address: [email protected] (J.P. Hays).
http://dx.doi.org/10.1016/j.mimet.2014.07.020 0167-7012/© 2014 Elsevier B.V. All rights reserved.
clinically relevant P. aeruginosa isolates, including nosocomial outbreakassociated isolates. A total of 32 P. aeruginosa isolates were selected for this study. Twenty of these isolates were clinical isolates collected between 2009 and 2011, comprising 19 isolates from The Netherlands (P2–P38) and a single isolate from Canada (P17). Eight of these isolates had been previously found to be metallo beta-lactamase (MBL)-positive by PCR (Pitout et al., 2005). The corresponding whole-genome maps were prepared by OpGen, Inc. (Gaithersburg, Maryland, USA) using the BamHI restriction enzyme, as previously described (Latreille et al., 2007). An extra twelve whole-genome maps were generated from publicly available whole-genome sequences of the following P. aeruginosa strains: 19BR (GenBank: AFXJ01000001), B136-33 (CP004061), c7747m (CP006728), DK2 (CP003149), LESB58 (FM209186), M18 (CP002496), MTB-1 (CP006853), NCGM2.S1 (AP012280), PA7 (CP000744), PAO1 (AE004091), RP73 (CP006245), and UCBPP-PA14 (CP000438). This is a convenient method for increasing the number of WGM isolate maps available for analysis and helps place newly WGM mapped bacterial isolates in the context of global isolates and outbreaks. The resulting in vitro and in silico P. aeruginosa whole-genome maps were aligned and compared using BioNumerics v7 software (Applied Maths, Sint-Martens-Latem, Belgium). Restriction fragments smaller than 3000 bp were excluded from the analysis and a relative tolerance of 15%, combined with an absolute tolerance of 3000 bp, was used to compensate for WGM resolution. All dendrograms were generated using UPGMA with a map distance of 5% (N95% similarity) as a cut-off point to define a WGM cluster, as reported previously (Bosch et al.,
20
S.A. Boers et al. / Journal of Microbiological Methods 106 (2014) 19–22
2013). Importantly, the BioNumerics v7 software allows two different approaches to be made when comparing whole-genome map similarities: 1) alignment-based comparison and 2) pattern-based comparison. The alignment-based approach uses a size tolerance algorithm for pairwise
comparison of the WGM fragments, whereas the pattern-based approach takes into account a neighbourhood of fragments, matching within a given size tolerance. To explore both methods, we applied them to our set of 32 whole-genome maps. As shown in Fig. 1, at the 95% similarity
Fig. 1. Differences observed between alignment-based and pattern-based algorithms when comparing WGM data derived from 32 P. aeruginosa strains. The alignment-based comparison (top) shows different clustering results for the in silico digested isolates PAO1, RP73, 19BR, c7747m and NCGM2.S1 (indicated in red) compared to the pattern-based comparison (bottom). Close inspection of the whole-genome maps reveals that chromosomal inversions among these strains are causing the clustering differences. Three clusters (I–III) can be identified using a similarity cut-off of 95%. The clustering of groups I–III isolates is similar using both alignment-based and pattern-based algorithms. * = metallo beta-lactamase PCR-positive isolates from Erasmus MC.
S.A. Boers et al. / Journal of Microbiological Methods 106 (2014) 19–22
level, both techniques generated three clusters. Cluster I consisted of eight isolates, P28, P29, P19, P15, P17, P13, P23 and P11, that all belonged to a previously reported outbreak at Erasmus MC (Van der Bij et al., 2011). Cluster II comprised the P14 and P20 isolates that are related to the previously reported and highly virulent UCBPP-PA14 strain. Information that may be useful to clinicians with respect to enhanced monitoring of patients infected with highly virulent P. aeruginosa strains. However, the two comparisons generated different clustering results dependant on the presence of genomic rearrangements. For example, a large chromosomal inversion (LCI) has been previously reported in the P. aeruginosa reference strain PAO1 (Lee et al., 2006), with close inspection of our whole-genome maps also revealing the presence of LCIs within in silico digested P. aeruginosa RP73, 19BR, c7747m and NCGM2.S1 genomes. The presence of these LCIs can lead to different interpretations of the genetic relatedness among P. aeruginosa isolates. To validate the use of WGM for P. aeruginosa, we compared our WGM clustering results with MLST results generated using the Highthroughput Multilocus Sequence Typing (HiMLST) technique (Boers et al., 2012). As shown in Fig. 2, both typing methods generated very similar clustering results that differed for only a single isolate, namely P. aeruginosa strain P33. However, this result reveals the limits of sensitivity of MLST typing, as isolate P33 appears to be identical to Cluster I MLST isolates, but possesses b95% similarity using WGM results, and is therefore not identical, but in fact closely related, to MLST Cluster I isolates. This type of information could have consequences for infection control outbreaks and programmes. The ordered structure of the whole-genome maps generated by WGM potentially allows for the rapid identification of specific, and diagnostic, genomic elements. To explore this possibility, we generated a multiple alignment to visualise the presence and absence of restriction fragments among the whole-genome maps. In order to simplify the analysis, the five in silico whole-genome maps that harboured LCIs, as
21
well as the in silico restriction map of the taxonomic outlier strain PA7 were excluded. The alignment of the remaining 26 whole-genome maps is presented in Fig. 3, highlighting some of the clusterdiscriminating fragments. As shown in Fig. 3, a single and unique discriminating fragment of ~61 kb in size could be identified among Cluster I WGM alignment isolates. This ~ 61 kb fragment could not be mapped to in silico WGM maps, including an additional 233 unfinished sequenced genomes (scaffolds/contigs) of P. aeruginosa in GenBank, and may potentially represent an unrecognised pathogenicity island in this cluster of isolates. Therefore, the use of WGM mapping allows rapid and easy visualization of cluster-specific elements that can be used as markers for high-throughput screening during ongoing nosocomial outbreaks of P. aeruginosa. Also, WGM offers the possibility of developing cluster-specific PCRs. In this short note, the authors demonstrate that WGM is an accurate and useful tool to discriminate P. aeruginosa isolates, which possesses superior discriminative power compared to traditional techniques for genotyping bacteria such as MLST. Further, the data can be reliably utilized by networks of researchers, and WGM provides useful information on the position and order of genome rearrangements and pathogenicity islands within the P. aeruginosa genome. However, the choice of restriction enzyme is likely important in enabling the standardisation of comparative WGM between research centres. Thus, whole-genome mapping may provide a suitable alternative to currently available bacterial genotyping techniques for clinicians and microbiologists interested in the comparative analysis of P. aeruginosa isolates in, for example, hospital outbreaks associated with this bacterial pathogen. The authors thank Trevor Wagner from OpGen, Inc. (Gaithersburg, Maryland, USA) for performing the WGM experiments and critical reading of the manuscript. This project was partially funded by an FP7 European Community grant (TEMPOtest-QC, HEALTH2009-241742).
Fig. 2. Dendrogram of 20 P. aeruginosa isolates based on concatenated MLST results. From this comparison, three clusters (I–III) can be identified with 100% similarity, which reveal high concordance with WGM clustering. * = metallo beta-lactamase PCR-positive isolates from Erasmus MC.
22
S.A. Boers et al. / Journal of Microbiological Methods 106 (2014) 19–22
Fig. 3. Multiple alignment of WGM data derived from 20 clinical P. aeruginosa strains and 6 in silico digested isolates. The whole-genome maps of isolates that possess similar chromosomal arrangements were compared and aligned to visualise the presence and absence of restriction fragments (above). A discriminating ~61 kb insert can be observed (bottom left) among Cluster I, as well as the closely related P33 isolate (92.5% similarity). * = metallo beta-lactamase PCR-positive isolates from Erasmus MC.
References Boers, S.A., van der Reijden, W.A., Jansen, R., 2012. High-throughput multilocus sequence typing: bringing molecular typing to the next level. PLoS One 7, e39630. Bosch, T.,Verkade, E.,van Luit, M.,Pot, B.,Vauterin, P., et al., 2013. High resolution typing by whole genome mapping enables discrimination of LA-MRSA (CC398) strains and identification of transmission events. PLoS One 8, e66493. Breidenstein, E.B.M., de la Fuente-Nunez, C., Hancock, R.E.W., 2011. Pseudomonas aeruginosa: all roads lead to resistance. Trends Microbiol. 19, 419–426. Chen, Q., Savarino, S.J., Venkatesan, M.M., 2006. Subtractive hybridization and optical mapping of the enterotoxigenic Escherichia coli H10407 chromosome: isolation of unique sequences and demonstration of significant similarity to the chromosome of E. coli K-12. Microbiology 152, 1041–1054. Clarridge 3rd, J.E.,Harrington, A.T.,Roberts, M.C.,Soge, O.O.,Maquelin, K., 2013. Impact of strain typing methods on assessment of relationship between paired nares and wound isolates of methicillin-resistant Staphylococcus aureus. J. Clin. Microbiol. 51, 224–231. Kotewicz, M.L., Mammel, M.K., LeClerc, J.E., Cebula, T.A., 2008. Optical mapping and 454 sequencing of Escherichia coli O157: H7 isolates linked to the US 2006 spinachassociated outbreak. Microbiology 154, 3518–3528.
Latreille, P., Norton, S., Goldman, B.S., Henkhaus, J., Miller, N., et al., 2007. Optical mapping as a routine tool for bacterial genome sequence finishing. BMC Genomics 8, 321. Lee, D.G., Urbach, J.M., Wu, G., Liberati, N.T., Feinbaum, R.L., et al., 2006. Genomic analysis reveals that Pseudomonas aeruginosa virulence is combinatorial. Genome Biol. 7. Lyczak, J.B., Cannon, C.L., Pier, G.B., 2000. Establishment of Pseudomonas aeruginosa infection: lessons from a versatile opportunist. Microbes Infect. 2, 1051–1060. Maltezou, H.C., 2009. Metallo-beta-lactamases in Gram-negative bacteria: introducing the era of pan-resistance? Int. J. Antimicrob. Agents 33 (405), e401–e407. Miller, J.M., 2013. Whole-genome mapping: a new paradigm in strain-typing technology. J. Clin. Microbiol. 51, 1066–1070. Pitout, J.D.D., Gregson, D.B., Poirel, L., McClure, J.A., Le, P., et al., 2005. Detection of Pseudomonas aeruginosa producing metallo-beta-lactamases in a large centralized laboratory. J. Clin. Microbiol. 43, 3129–3135. Van der Bij, A.K.,Van Mansfeld, R.,Peirano, G.,Goessens, W.H.,Severin, J.A., et al., 2011. First outbreak of VIM-2 metallo-beta-lactamase-producing Pseudomonas aeruginosa in The Netherlands: microbiology, epidemiology and clinical outcomes. Int. J. Antimicrob. Agents 37, 513–518.