SGM Journals
DIAGNOSTICS, TYPING AND IDENTIFICATION

PCR characterization and typing of Klebsiella pneumoniae using capsular type-specific, variable number tandem repeat and virulence gene targets

Jane F. Turton, Claire Perry, Suzanne Elgohari, Catherine V. Hampton

  • 1Laboratory of HealthCare Associated Infection, Centre for Infections, Health Protection Agency, London NW9 5EQ, UK
  • 2HealthCare Associated Infection and Antimicrobial Resistance Department, Centre for Infections, Health Protection Agency, London NW9 5EQ, UK
  • Correspondence
    Jane F. Turton
    jane.turton{at}hpa.org.uk

    • Journal of Medical Microbiology 2010; 59(5):541–547 · https://doi.org/10.1099/jmm.0.015198-0

    View at publisher PubMed

    Abstract

    A multiplex PCR is described which detects capsular types K1, K2, K5, K54 and K57, which are those most associated with invasive disease or pathogenicity, a further capsular type (K20), two putative virulence factors (rmpA and wcaG) and the 16S–23S internal transcribed spacer unit of Klebsiella pneumoniae, facilitating identification of this organism. wcaG encodes capsular fucose production and was associated with capsular types K1 and K54, but was also found in strains of other capsular types; 18 of the 543 isolates screened were PCR-positive for this gene. An eight-locus variable number tandem repeat (VNTR) scheme was designed, which provided discrimination at a level similar to that afforded by PFGE among a panel of 36 isolates representing 29 PFGE types. All isolates tested of the virulent K1 clone of CC23, associated with pyogenic liver abscesses, shared the same VNTR profile, which may be helpful in identifying this clone; such isolates were also PCR-positive for allS. These methods provide a rapid means of characterizing and typing isolates of this important agent of community-acquired and nosocomial infection.

    • The GenBank/EMBL/DDBJ accession numbers for the GDP-fucose synthetase gene (wcaG) sequences of NCTC 9136 and NCTC 9178 are GU325787 and GU325788, respectively.

    INTRODUCTION

    Klebsiella pneumoniae is an important nosocomial pathogen, most frequently causing pneumonia, and urinary tract, wound or blood infections (Podschun & Ullmann, 1998; Brisse et al., 2009). Nosocomial isolates are often associated with extended spectrum β-lactamases, including, recently, carbapenemases such as KPC and OXA-48 (Kitchel et al., 2009; Nordmann et al., 2009; Carrër et al., 2008; Gülmez et al., 2008), making treatment options limited. It is encapsulated, with 77 capsular types being recognized in an international scheme. The capsule is an important virulence factor, and some capsular types, particularly K1 and K2, but also K54 and K57 (Fang et al., 2007), are associated with a community-acquired invasive pyogenic liver abscess syndrome. In particular, the virulent clone of sequence type 23, referred to as CC23 (K1), or the K1 cluster, identified by MLST and PFGE, is implicated (Turton et al., 2007; Chung et al., 2008; Brisse et al., 2009). Other putative virulence factors have been described, including the plasmid-borne rmpA (regulator of mucoid phenotype A) (Nadasy et al., 2007; Fang et al., 2007; Yu et al., 2007, 2008), wcaG, which encodes capsular fucose synthesis, which may enhance the ability of the bacteria to evade phagocytosis by macrophages (Wu et al., 2008), alls, encoding the activator of the allantoin regulon, and a marker for CC23 (K1) (Chou et al., 2004; Brisse et al., 2009), and production of aerobactin (an iron siderophore) (Yu et al., 2007, 2008). Since K. pneumoniae subsp. pneumoniae is the most clinically important of the Klebsiella species, the ability to identify it rapidly is also helpful. Previously, we have used a multiplex PCR to detect K1, K2 and K5 capsular types using serotype-specific targets (Turton et al., 2008), but have now extended this to include further capsular types (K20, K54 and K57), putative virulence factors (rmpA and wcaG) and identification of K. pneumoniae subsp. pneumoniae based on the 16S–23S internal transcribed spacer (Liu et al., 2008), which we describe here. Primers for identification of Klebsiella oxytoca (Kovtunovych et al., 2003) may also be included.

    Since outbreaks due to this organism may occur in the nosocomial setting, typing is important to identify possible cases of transmission. This is usually performed by comparison of DNA fingerprints using PFGE, which provides an excellent method for this purpose. However, variable number tandem repeat (VNTR) analysis, in which isolates are characterized by their repeat numbers at multiple loci each with a variable number of tandem repeats, offers some advantages over PFGE. These loci are particularly suitable for epidemiological investigations, since they have a relatively high mutation rate, the addition or deletion of sequence units usually occurring as a result of slipped strand mispairing during replication (van Belkum, 1999; Yazdankhah & Lindstedt, 2007). Since the method is PCR-based, results can be obtained rapidly, and, as isolates are described by a series of numbers, profiles can be readily compared, both within and between laboratories. Here we describe eight VNTR loci for K. pneumoniae, which provide discrimination at the same level as that provided by PFGE. Use of the multiplex PCR coupled with the VNTR analysis provides a rapid means of characterization and typing of isolates.

    METHODS

    Bacterial isolates.

    The 77 capsular serotype reference strains and all isolates of Klebsiella referred to our laboratory for comparison by PFGE over an 18-month period (from January 2008 to June 2009), from hospitals in the UK, were subjected to the multiplex PCR. Twenty-eight of the clinical isolates were selected to provide a panel for testing the VNTR loci, which also included K1 isolates from Taiwan (TW3), Israel (IS1) and the UK (UK4) from a previous study (Turton et al., 2007) that belong to the K1 cluster (CC23), other strains of the K1 capsular type (NCTC 5054, AS4), a reference strain of capsular type K5 (NCTC 9660) belonging to the K5 cluster (Turton et al., 2008) and K2 and K54 capsular type reference strains (NCTC 5055 and NCTC 9174). Four of the isolates (isolates 14, 15, 16 and 17), representing distinct strains, produce OXA-48 (N. Woodford, personal communication). The panel contained an elevated proportion of isolates of the capsular types detected by the PCR; it also included a set of isolates (not of these capsular types) from a recent epidemiological investigation. Reference strains of Klebsiella species used were from the NCTC collection, unless stated otherwise. All isolates were also compared by PFGE of XbaI-digested DNA, as described previously (Turton et al., 2007). Isolates were described by number (1–30) and the referring hospitals, from which they were received, by letter (A–V) or (if from outside the UK) country.

    Multiplex PCR.

    Multiplex PCR was carried out using primers for nine targets, described in Table 1⇓, and a Qiagen Multiplex PCR kit, which includes 2× strength PCR mix specifically designed for multiplex reactions, with a built-in hot start. Primers PEH-C and PEH-D, described by Kovtunovych et al. (2003), for identification of K. oxytoca, giving an amplicon of 344 bp, may also be included (Fig. 1⇓). Reactions were carried out in 25 μl volumes with final concentrations/amounts of reagents of 1× Multiplex PCR mix, 5 pmol each primer and 2 μl crude DNA extract (prepared by suspending two to three colonies in PCR-quality water, vortexing and centrifugation; the supernatant was used in PCRs). PCR conditions consisted of an initial activation at 95 °C for 15 min, followed by 35 cycles of 94 °C for 30 s, 58 °C for 90 s and 72 °C for 90 s, and a final extension at 72 °C for 10 min. PCR products were separated by agarose gel electrophoresis in a 1.5 % (w/v) gel, or using a QIAxcel system (Qiagen). To gain better size differentiation by conventional gel electrophoresis, long gels (15 cm) were used.

    Figure image not available in archive
    Fig. 1.

    Results of multiplex PCR carried out on isolates of Klebsiella species, with (b) and without (a) inclusion of primers for detection of K. oxytoca. M, Size ladder (HyperLadder II). K1, K2, K5, K54, K57, K20, K16 and K58 reference strains (NCTC 5054, NCTC 5055, NCTC 9660, NCTC 9174, NCTC 9177, NCTC 9140, NCTC 9136 and NCTC 9178, respectively) are in lanes 1, 2, 3, 4, 5, 6, 9 and 10 respectively. NCTC 5050 (K. pneumoniae subsp. ozaenae), ATCC 13182 (K. oxytoca) and NCTC 9633 (K. pneumoniae subsp. pneumoniae) are in lanes 7, 8 and 11. Lane 12 is the negative (water) control. ITS, internal transcribed spacer.

    Table 1.

    Multiplex PCR primers

    Primers for identification of K. oxytoca (PEH-C and PEH-D), described by Kovtunovych et al. (2003), giving a PCR product of 344 bp, may also be included.

    Identification.

    The K.pneumoniae Pf/K.pneumoniae Pr1 primer pair included in the multiplex PCR is for the identification of K. pneumoniae (Liu et al., 2008). Both K. pneumoniae subsp. pneumoniae and subsp. ozaenae give a band with these primers (see Results and Discussion section); subsp. rhinoscleromatis does not. To distinguish between K. pneumoniae subsp. pneumoniae and K. pneumoniae subsp. ozaenae, Voges–Proskauer and malonate reactions were carried out. Most isolates of K. pneumoniae subsp. pneumoniae are positive in both tests, while the majority of isolates of K. pneumoniae subsp. ozaenae are negative in both (Hansen et al., 2004). Primers PEH-C and PEH-D, if included, provide specific identification of K. oxytoca (Kovtunovych et al., 2003).

    VNTR analysis.

    The genome sequence of K. pneumoniae subsp. pneumoniae MGH 78578 (GenBank accession no. NC_009648; CP000647) was put through a tandem repeat finder () and 17 potentially useful loci were selected from the 117 found. Primers were designed from the flanking sequences, with the aid of the design tool on the Sigma-Aldrich website (). PCRs were initially carried out on a panel of 23 isolates, representing 18 PFGE defined genotypes, and loci giving amplicons that varied in size among the set were identified (Table 2⇓). Reactions were carried out using 1× PCR buffer (containing 1.5 mM MgCl2) (Qiagen), 10 pmol each primer, 200 μM each dNTP, 3 μl crude DNA extract and 1.5 U Taq DNA polymerase in a total reaction volume of 25 μl. Thermocycler conditions were: 94 °C for 3 min, followed by 30 cycles of 94 °C for 30 s, 58 °C for 30 s, 72 °C for 45 s and a final extension at 72 °C for 10 min. Amplicons were separated in a 1.5 % (w/v) agarose gel and their size was estimated by comparison with a size ladder (HyperLadder II; BioLine). Since the sizes of the flanking sequences and repeat units were known, the number of repeats at each locus for each isolate could be determined and these were recorded in the order of loci A, E, H, J, K, D, I, L giving the VNTR profile. In some cases, amplicons were also sequenced on a Beckman-Coulter CEQ8000, following treatment with ExoSAP-IT (USB), according to the manufacturer's instructions, to verify the number of repeat units at each locus. Some wcaG amplicons from the multiplex PCR were also sequenced.

    Table 2.

    VNTR loci and primers used for their amplification

    Loci were identified from the complete genome sequence of K. pneumoniae MGH 78578 (accession no. CP000647).

    allS PCR.

    Detection of allS, associated with CC23 (K1), was carried out using the primers of Chou et al. (2004) in the same manner as for amplification of the VNTR loci, except that an annealing temperature of 50 °C was used.

    RESULTS AND DISCUSSION

    Reference strains of K1, K2, K5, K20, K54 and K57 were successfully detected in the multiplex PCR (Fig. 1⇑); reference strains of the other 71 capsular types failed to give these bands. The K1 reference strain (NCTC 5054) was also PCR-positive for rmpA and wcaG, the K2 reference strain (NCTC 5055) for rmpA, and the K54 reference strain (NCTC 9174) for wcaG. Some of the other reference strains were also PCR-positive for these elements (K3 and K6 for rmpA and K16 and K58 for wcaG). Most gave a band for the K. pneumoniae 16S–23S internal transcribed spacer region. K. oxytoca (ATCC 13182) was included in each run as a negative control (Fig. 1⇑) for this K. pneumoniae target. In common with the findings of Liu et al. (2008), we found that Klebsiella species/subspecies other than K. pneumoniae subsp. pneumoniae [K. oxytoca, K. pneumoniae subsp. rhinoscleromatis (NCTC 1936)] and related organisms [Raoultella planticola (ATCC 33531), Raoultella terrigena (ATCC 33257)] failed to give this band in the PCR, with the exception of K. pneumoniae subsp. ozaenae (NCTC 5050 and NCTC 5051); this could be distinguished from K. pneumoniae subsp. pneumoniae on the basis of its lack of reaction in malonate and Voges–Proskauer biochemical tests. Inclusion of the pehX primers provided specific detection of K. oxytoca (Fig. 1⇑). These findings were confirmed with a number of clinical and reference isolates, the former having previously been identified by biochemical testing. These included a further four isolates of K. pneumoniae subsp. ozaenae, the only subspecies not included in the original validation of Liu et al. (2008).

    There have been many reports linking rmpA with virulence, but wcaG has been less well studied. During an 18-month period between January 2008 and June 2009, we screened 543 isolates of Klebsiella species received by our laboratory for wcaG. Eighteen of these isolates were PCR-positive for wcaG with our primers, and it is noteworthy that many of them were from patients with invasive and serious infections (Table 3⇓). They included all the isolates of capsular types K1 and K54, and it is perhaps therefore not surprising that wcaG is associated with virulence. However, not all of the isolates were of these capsular types, with six being of type(s) not detected by the PCR. They may well be of capsular type(s) K16 or K58, since those reference strains (NCTC 9136, NCTC 9178) were also PCR-positive for wcaG; indeed two out of the three wcaG amplicons sequenced (from isolates 13 and 14) were identical to that found for the K16 type strain (GenBank accession no. GU325787); that of a K54 isolate (isolate 10) and of the K58 reference strain (GU325788) displayed sequence differences both from this and each other, the former matching that of the GDP-fucose synthetase gene of a K54 isolate in GenBank (accession no. AB289649). K16 and K58 capsular types have not been particularly associated with virulence. Nevertheless, two out of the three of these non-K1/K54 isolates for which clinical information was available were associated with serious disease (Table 3⇓). The 18 wcaG-positive isolates included three representatives of the K1 cluster and four isolates of outbreak strain E of capsular type K54 from hospital H (Fig. 2⇓); the remaining isolates represented distinct strains. Four isolates, including all the representatives of the K1 cluster, were also PCR-positive for rmpA; one also had blaOXA-48. These results suggest that wcaG may be a contributing factor to the virulence of these strains, with many having multiple factors that may enhance their pathogenicity. rmpA was similarly associated with multiple capsular types [K1, K2, K20, K54, K57 and type(s) not detected by the PCR]. Inclusion of these targets in the PCR may therefore identify potentially virulent strains that would otherwise be missed. However, isolates causing invasive disease are not confined to those with these elements. K2 was the most commonly found of the capsular types sought among submitted isolates (35/543). K1, K54 and K57 capsular types each accounted for only between 1 and 2 % of the total isolates. K5 isolates (3/543) were received only from thoroughbred horses having been submitted for screening prior to breeding programmes. Capsular type K20 appeared to be very rare; we found only one among our isolates.

    Figure image not available in archive
    Fig. 2.

    Dendrogram comparing PFGE profiles of 36 isolates, together with the results of multiplex PCR, VNTR and allS PCR, with the number of repeat units at each VNTR locus being given in the order A, E, H, J, K, D, I and L. Isolates 1, 2, TW3, IS1 and UK4 are representatives of the K1 cluster (CC23), and isolates 14, 15, 16 and 17 are PCR-positive for blaOXA-48. Most are clinical isolates from hospitals (A–T) in the UK. Isolates with designations of ‘unique’ each represent a distinct strain harboured by a single patient only. A dash (–) in the VNTR profile indicates that no amplicon was detected at that locus, IS indicates the likely presence of an insertion sequence, and + indicates an amplicon that is slightly larger than others of the same repeat number. Isolates TW3, IS1, 14 and AS4 were from Taiwan, Israel, Turkey and Australia, respectively. Two isolates gave amplicons that were much bigger than the expected size in the allS PCR, and were scored as – (large).

    Table 3.

    Description of isolates that were PCR-positive for wcaG for which the sending laboratory provided clinical information

    Isolate TW3, from Taiwan, was from a previous study (Turton et al., 2007); the remaining isolates were received by our laboratory between January 2008 and June 2009. Isolates with unique PFGE profiles represent distinct strains harboured by single patients only.

    Results on a panel of 36 isolates, representing 29 PFGE types, which were also subjected to VNTR analysis, are given in Fig. 2⇑. Eight (A, E, H, J, K, D, I and L) of the 17 VNTR loci tested gave amplicons that varied in size among an initial panel of 23 of these isolates, representing 18 PFGE types. In each case, the number of repeats was determined from the amplicon size and, in some cases, by sequencing. Strain MGH 78578, the genome sequence of which was searched to identify these loci, has repeat numbers of 4.8, 2.8, 7.3, 8.3, 2.3, 2.0, 2.0 and 4.5 at these loci, respectively, and we similarly found that isolates had amplicons that were slightly smaller or larger than expected on the basis of complete repeat units. Sequencing of PCR products revealed that they had the same incomplete repeat unit at the 3′-end as MGH 78578 for loci A, E, H, J, K and L; complete repeat units were found at loci D and I, which have only a 14 bp repeat. Repeat numbers were rounded up (e.g. 4.8 to 5) or down (e.g. 2.3 to 2) as appropriate, except for at locus L. In rare cases, more than one size of amplicon was found in the size range expected for a particular repeat number; in others, very large amplicons (∼2 kb) were produced, which we assumed to be due to the presence of insertion sequences, commonly found during VNTR analysis of Pseudomonas aeruginosa (Vu-Thien et al., 2007). These were designated by + (if slightly larger than expected for that repeat number) or IS, respectively. Some amplicons were similar in size to the flanking sequence, and the number of repeats was then scored as zero.

    Unique profiles were obtained for each type that had been distinguished by PFGE; representatives of the same type shared the same VNTR profile (e.g. isolates 3, 4 and 5 all shared a profile of 7, 8, 5, 1, −, 2, 2, 3.5) (Fig. 2⇑). Representatives of the CC23 (K1) cluster from the UK (isolates 1, 2 and UK4), Israel (isolate IS1) and Taiwan (isolate TW3) shared the same, characteristic VNTR profile (3, 5, 2, 7, 0, 1, 2, 3.5) (although sometimes isolates failed to give an amplicon at locus J), providing a rapid means of identifying isolates of this virulent clone; they were also all PCR-positive for allS. Other K1 strains not belonging to the K1 cluster (K1 reference strain; NCTC 5054) and isolate AS4 had distinct profiles and were PCR-negative for allS. Loci A, E, H and J were the most discriminatory, each having from five to seven alleles among our panel, whilst loci D and I were the least discriminatory, with only two alleles among our panel. Although five alleles were found at locus L, most (28/36) isolates had a repeat number of 3.5 at this locus, and it did not prove very helpful. Testing of further loci may well reveal others that provide better discrimination between isolates, and we will continue to screen for useful loci, using the genome sequences of K. pneumoniae 342 and NTUH-K2044, now available, as well as that of MGH 78578, to identify potential targets. On the basis of the Kp342 sequence (GenBank accession no. CP000964), introducing some ‘wobble’ into the reverse primer for locus H (using CTTTACCTGGCATGCKAACG) could improve the frequency of amplicons obtained at this locus, although it was mostly at locus K where isolates failed to amplify. Nevertheless, the present combination of loci did provide discrimination at a level similar to that provided by PFGE, at least among our panel. Determination of the number of repeat units at loci D and I, which have repeat units of 14 bp, could be done more conveniently by using fluorescently labelled forward primers in the PCRs and sizing on a sequencer. Coupled with the multiplex PCR, this method could provide a rapid and effective means of characterization of isolates of K. pneumoniae, confirming the identification, identifying important capsular types associated with particular disease manifestations, detecting putative virulence factors, and discriminating between strains. Known virulent strains, such as CC23 (K1), may be identified from their VNTR profiles. Such clones also have distinctive virulence gene profiles, with allS (encoding the activator of the allantoin regulon) being a marker for CC23 (K1) (Chou et al., 2004; Brisse et al., 2009) [although one non-K1 cluster isolate (isolate 27) was PCR-positive for allS in our tests], perhaps providing another potential candidate for inclusion in the multiplex. We hope that these PCR methods will prove helpful in providing useful information for clinicians in a timely manner.

    Acknowledgments

    We are grateful to colleagues in hospital laboratories for referring these isolates to us, and to Patrice Nordmann for sending a representative of OXA-48-producing clone A (isolate 14).

    References

    1. Brisse, S., Fevre, C., Passet, V., Issenhuth-Jeanjean, S., Tournebize, R., Diancourt, L. & Grimont, P. (2009). Virulent clones of Klebsiella pneumoniae: identification and evolutionary scenario based on genomic and phenotypic characterization. PLoS One 4, e4982
    2. Carrër, A., Poirel, L., Eraksoy, H., Cagatay, A. A., Badur, S. & Nordmann, P. (2008). Spread of OXA-48-positive carbapenem-resistant Klebsiella pneumoniae isolates in Istanbul, Turkey. Antimicrob Agents Chemother 52, 2950–2954.
    3. Chou, H.-C., Lee, C.-Z., Ma, L.-C., Fang, C.-T., Chang, S.-C. & Wang, J.-T. (2004). Isolation of a chromosomal region of Klebsiella pneumoniae associated with allantoin metabolism and liver infection. Infect Immun 72, 3783–3792.
    4. Chung, D. R., Lee, H. R., Lee, S. S., Kim, S. W., Chang, H.-H., Jung, S.-I., Oh, M.-D., Ko, K. S., Kang, C.-I. & other authors (2008). Evidence for clonal dissemination of the serotype K1 Klebsiella pneumoniae strain causing invasive liver abscesses in Korea. J Clin Microbiol 46, 4061–4063.
    5. Fang, C.-T., Chuang, Y.-P., Shun, C.-T., Chang, S.-C. & Wang, J.-T. (2004). A novel virulence gene in Klebsiella pneumoniae strains causing primary liver abscess and septic metastatic complications. J Exp Med 199, 697–705.
    6. Fang, C.-T., Lai, S.-Y., Yi, W.-C., Hsueh, P.-R., Liu, K.-L. & Chang, S.-C. (2007). Klebsiella pneumoniae genotype K1: an emerging pathogen that causes septic ocular or central nervous system complications from pyogenic liver abscess. Clin Infect Dis 45, 284–293.
    7. Gülmez, D., Woodford, N., Palepou, M. F., Mushtaq, S., Metan, G., Yakupogullari, Y., Kocagoz, S., Uzun, O., Hascelik, G. & Livermore, D. M. (2008). Carbapenem-resistant Escherichia coli and Klebsiella pneumoniae isolates from Turkey with OXA-48-like carbapenemases and outer membrane protein loss. Int J Antimicrob Agents 31, 523–526.
    8. Hansen, D. S., Aucken, H. M., Abiola, T. & Podschun, R. (2004). Recommended test panel for differentiation of Klebsiella species on the basis of a trilateral interlaboratory evaluation of 18 biochemical tests. J Clin Microbiol 42, 3665–3669.
    9. Kitchel, B., Rasheed, J. K., Patel, J. B., Srinivasan, A., Navon-Venezia, S., Carmeli, Y., Motakefi, A. & Giske, C. G. (2009). Molecular epidemiology of KPC-producing Klebsiella pneumoniae in the United States: clonal expansion of MLST sequence type 258. Antimicrob Agents Chemother 53, 3365–3370.
    10. Kovtunovych, G., Lytvynenko, T., Negrutska, V., Lar, O., Brisse, S. & Kozyrovska, N. (2003). Identification of Klebsiella oxytoca using a specific PCR assay targeting the polygalacturonase pehX gene. Res Microbiol 154, 587–592.
    11. Liu, Y., Liu, C., Zheng, W., Zhang, X., Yu, J., Gao, Q., Hou, Y. & Huang, X. (2008). PCR detection of Klebsiella pneumoniae in infant formula based on 16S–23S internal transcribed spacer. Int J Food Microbiol 125, 230–235.
    12. Nadasy, K. A., Domiati-Saad, R. & Tribble, M. A. (2007). Invasive Klebsiella pneumoniae syndrome in North America. Clin Infect Dis 45, e25–e28.
    13. Nordmann, P., Cuzon, G. & Naas, T. (2009). The real threat of Klebsiella pneumoniae carbapenemase-producing bacteria. Lancet Infect Dis 9, 228–236.
    14. Podschun, R. & Ullmann, U. (1998). Klebsiella spp. as nosocomial pathogens: epidemiology, taxonomy, typing methods, and pathogenicity factors. Clin Microbiol Rev 11, 589–603.
    15. Turton, J. F., Englender, H., Gabriel, S. N., Turton, S. E., Kaufmann, M. E. & Pitt, T. L. (2007). Genetically similar isolates of Klebsiella pneumoniae serotype K1 causing liver abscesses in three continents. J Med Microbiol 56, 593–597.
    16. Turton, J. F., Baklan, H., Siu, L. K., Kaufmann, M. E. & Pitt, T. L. (2008). Evaluation of a multiplex PCR for detection of serotypes K1, K2 and K5 in Klebsiella sp. and comparison of isolates within these serotypes. FEMS Microbiol Lett 284, 247–252.
    17. van Belkum, A. (1999). Short sequence repeats in microbial pathogenesis and evolution. Cell Mol Life Sci 56, 729–734.
    18. Vu-Thien, H., Corbineau, G., Hormigos, K., Fauroux, B., Corvol, H., Clément, A., Vergnaud, G. & Pourcel, C. (2007). Multiple-locus variable-number tandem-repeat analysis for longitudinal survey of sources of Pseudomonas aeruginosa infection in cystic fibrosis patients. J Clin Microbiol 45, 3175–3183.
    19. Wu, J.-H., Wu, A.-M., Tsai, C.-G., Chang, X.-Y., Tsai, S.-F. & Wu, T.-S. (2008). Contribution of fucose-containing capsules in Klebsiella pneumoniae to bacterial virulence in mice. Exp Biol Med (Maywood) 233, 64–70.
    20. Yazdankhah, S. P. & Lindstedt, B. A. (2007). Variable number tandem repeat typing of bacteria. Methods Mol Biol 396, 395–405.
    21. Yu, V. L., Hansen, D. S., Ko, W. C., Sagnimeni, A., Klugman, K. P., von Gottberg, A., Goossens, H., Wagener, M. M. & Benedi, V. J., International Klebsiella Study Group (2007). Virulence characteristics of Klebsiella and clinical manifestations of K. pneumoniae bloodstream infections. Emerg Infect Dis 13, 986–993.
    22. Yu, W.-L., Ko, W.-C., Cheng, K.-C., Lee, C.-C., Lai, C.-C. & Chuang, Y.-C. (2008). Comparison of prevalence of virulence factors for Klebsiella pneumoniae liver abscesses between isolates with capsular K1/K2 and non-K1/K2 serotypes. Diagn Microbiol Infect Dis 62, 1–6.