Multiple-Locus Variable Number Tandem Repeat Analysis of Klebsiella pneumoniae : Comparison with Pulsed-Field Gel Electrophoresis

Abstract

Aim:

To assess whether multiple-locus variable-number tandem repeat analysis (MLVA) could replace pulsed-field gel electrophoresis (PFGE) for genotyping of Klebsiella pneumoniae, this study was conducted to compare the typeability, discriminatory power, and concordance of these methods.

Materials and Methods:

We used the nine variable number tandem repeat (VNTR) loci scheme to test its suitability for differentiating 114 ESBL-producing K. pneumoniae isolates collected from different clinical specimens in three hospitals of Tehran, Iran between April and December 2011. PFGE with XbaI was performed and the results were compared with those obtained by typing with MLVA.

Results:

MLVA and PFGE yielded 44 and 64 types, respectively. Simpson's Diversity Index of MLVA was 0.896 (a 95% confidence interval of 0.850–0.942) and of PFGE was 0.962 (a 95% confidence interval of 0.943–0.981). Congruence between PFGE and MLVA was low. The adjusted Wallace coefficient of PFGE to MLVA was 0.946; however, MLVA was less able to predict PFGE type (32.5%). A range of 2–12 alleles was identified at VNTR loci with Simpson's diversity values between 0.017 and 0.818.

Conclusion:

MLVA is a PCR-based typing method and is much easier and more rapid in comparison to PFGE. These data indicate that the MLVA typing scheme used in this study is discriminative and reliable for typing of K. pneumoniae. However, optimization of the VNTR markers is required to improve the discriminatory power of the method.

Introduction

Klebsiella pneumoniae is a nosocomial pathogen involved in important infections such as pneumonia, urinary tract, wound, and blood infections. Major epidemiological features of this organism include the propensity for nosocomial outbreaks, clonal dissemination, long-term persistence in the hospital setting, as well as multiple resistance to antimicrobial agents.¹ Production of extended-spectrum β-lactamases (ESBL) is one of the main mechanisms of β-lactam resistance in K. pneumoniae isolates.^2,3 Hospital outbreaks and the dissemination of ESBL-producing strains of K. pneumoniae have been reported worldwide, including Iran.^4–6

Since the number of outbreaks due to K. pneumoniae is currently increasing, typing is important to identify possible routes of transmission. Several molecular typing methods have been introduced and developed for investigation of the molecular epidemiology of K. pneumoniae infection. Pulsed-field gel electrophoresis (PFGE) has been used extensively, and it provides an excellent method for K. pneumoniae typing.¹ However, this method has several limitations; for example, it is time consuming and has poor reproducibility between laboratories.⁷ Multiple-locus variable-number tandem repeat analysis (MLVA) is a PCR-based typing method that involves the analysis of the copy numbers of tandem-repeat units of variable number tandem repeat (VNTR) loci found in the microbial genome, which requires only basic equipment and allows interlaboratory comparisons of results.⁷ An MLVA scheme for K. pneumoniae was introduced in 2012.^1,8 It analyzes nine VNTR loci that are amplified in PCR reactions, and products are analyzed by agarose gel electrophoresis. This is by far the most intensely used scheme for MLVA typing of K. pneumoniae.^8–13

In this study, we assess the validity of this MLVA scheme by comparing it with PFGE, as a well established genotyping method, for genotyping of 114 ESBL-producing K. pneumoniae strains isolated from different sources of three hospitals in Tehran over a 9-month time period.

Materials and Methods

Bacterial isolates and identification

A total of 200 consecutive nonduplicate K. pneumoniae isolates were collected from different clinical specimens in three hospitals of Tehran, Iran between April and December 2011, which were identified as K. pneumoniae using biochemical tests for identification of enterobacteriaceae. Hospital 1 is a 400- bed educational tertiary and referral poison treatment center in the southwest of Tehran with nearly annual average of 20,000 admissions and out-patient hospital visits. Hospital 2 is a 1,000-bed nonteaching tertiary care hospital belonging to The Social Security Organization, located in the northwest. Hospital 3 is a tertiary teaching hospital situated in the center of Tehran with 1,300 active beds, and it is regarded as the largest and the most advanced educational and medical center in Iran.

A PCR method was used to confirm the identification of K. pneumoniae subsp. pneumoniae, using specific primers for K. pneumoniae 16S–23S internal transcribed spacer region.¹ Pf: 5′-ATTTGAAGAGGTTGCAAACGAT-3′ and Pr1: 5′-TTCACTCTGAAGTTTTCTTGTGTTC-3′ (amplicon size: 130 bp). Genomic DNA was extracted by the freeze–thaw method and used as the template for PCR reactions.¹⁴ PCR conditions were as follows: Initial denaturation at 94°C for 5 min, followed by 35 cycles of 94°C for 1 min, 58°C for 1 min, and 72°C for 1 min. Final extension was at 72°C for 7 min. K. pneumoniae ATCC13883 was used as a positive control.

Combined disk method confirmed 114 (57%) of 200 isolates as ESBL producer. These 114 isolates were mostly isolated from tracheal secretions (63/114 isolates, 55.2%), urine (30/114 isolates, 26.3%), and wound (11/114 isolates, 9.6%).

Genotyping by MLVA

MLVA analysis of K. pneumoniae isolates was performed as described previously by PCR amplification of nine VNTR loci (A, E, H, J, K, D, N1, N2, and N4).^1,8 PCR reaction mixtures (25 μl) contained 3 μl template DNA, 10 pmol of each primer, 200 μM each dNTP, 1X PCR buffer, 1.5 mM MgCl2, and 1 U Taq DNA polymerase (Fermentas, Germany). Thermocycler conditions were as follows: initial denaturation at 94°C for 5 min, followed by 35 cycles of 94°C for 1 min, 52°C for 1 min, and 72°C for 1 min. Final extension was at 72°C for 7 min. Conditions were the same for all genes, except loci H, E, and J for which annealing temperature was 55°C. After PCR, the products were electrophoresed in a 1.5% (w/v) agarose gel in 0.5X TBE (Tris-Borate-EDTA) buffer and run at 8 V/cm for 5 hr. The gels were stained with ethidium bromide and photographed under UV light.

The size of the amplicons of each locus was estimated by comparison with a size ladder. Since the sizes of the flanking sequences and repeat units were known, the number of repeats for each locus was counted, and these were recorded in the order of loci A, E, H, J, K, D, N1, N2, and N4, giving the VNTR profile. Thus, for each isolate an allelic profile was generated. MLVA numerical profiles were entered into a related website (http://mlvaplus.net). Clustering of MLVA types was displayed in a minimum spanning tree (MST) based on categorical coefficient. A different MLVA type was given when one difference was observed at any VNTR. Clonal complexes (CCs) were defined as the groups of isolates for which the MLVA types differed at a maximum of two VNTR loci.¹⁵

Genotyping by PFGE

PFGE was performed using the protocol recommended by PulseNet with minor modifications.¹⁶ Agarose-embedded DNA was digested with 10 U of XbaI restriction enzyme (Jena Bioscience, Germany) for 6 hr at 37°C. The digested plugs were run in a 1% low melting agarose gel (Sigma) using a CHEF-DR II apparatus (Bio-Rad) with initial pulse time of 2.2 sec and final pulse time of 54.2 sec for 20 hr at 6 V. Salmonella serotype Braenderup H9812 was used as a molecular size marker.

PFGE patterns were analyzed and compared using GelCompar II software program, version 4.0 (Applied Maths, Belgium). A dendrogram was generated using the band-based Dice similarity coefficient, and the unweighted pair group method with arithmetic mean, with settings of 1.5% optimization and 1.5% band position tolerance. A PFGE genotype was defined as a PFGE pattern with one or more DNA bands different from the others.¹⁷

Statistical analysis

The discriminatory power of PFGE and MLVA was measured by Simpson's index of diversity, and 95% confidence intervals were calculated. The adjusted Wallace coefficient was calculated to evaluate the congruence between the two typing methods. This coefficient indicates the probability that two strains, which are assigned to the same type by one typing method, have a chance to be typed as identical by the other typing method.¹⁸ Calculations of the discriminatory power and concordance of the typing methods were done using the online tool, Comparing Partitions, located at http://Darwin.phyloviz.net/ComparingPartitions. The allelic diversity (discriminatory power) of each VNTR locus was calculated by Simpson's Diversity Index. Software (V-DICE) available at the Health Protection Agency bioinformatics tools website (www.hpa-bioinformatics.org.uk/cgi-bin/DICI/DICI.pl) was used to measure the variation of the number of repeats at each MLVA locus.¹⁹

Results

VNTR loci diversity

MLVA was performed on the 114 ESBL-producing clinical K. pneumoniae isolates. Analysis of the composition of the nine VNTR loci of the strains collection revealed that the number of alleles ranged from two for VNTR_N4 and VNTR_D to 12 for VNTR_J (Table 1). The diversity indices of the VNTR loci, measured by Simpson's Diversity Index, showed an average diversity of 0.6. The lowest diversity index was found for VNTR_N4 (0.017), as 99% (113/114) of the isolates had the same allele at this locus, and the highest diversity index was found for VNTR_J (0.818). The VNTR_E revealed lower diversity (0.671) in spite of having seven alleles, than N2 with five alleles and a diversity index of 0.724. Isolates failed to amplify mostly at locus K (6/114, 5.2%). Also, more than half of the isolates (68/114, 59.6%) had a repeat number of 1 at locus D with 2 alleles.

Table 1.

The Diversity Indices and Number of Repeats at Each MLVA Locus

Locus	Diversity index	Confidence interval	K
VNTR_J	0.818	0.781–0.855	12
VNTR_H	0.798	0.760–0.836	8
VNTR_K	0.732	0.696–0.768	6
VNTR_N2	0.724	0.690–0.757	5
VNTR_A	0.722	0.653–0.791	8
VNTR_E	0.671	0.610–0.732	7
VNTR_N1	0.525	0.432–0.618	5
VNTR_D	0.481	0.447–0.516	2
VNTR_N4	0.017	0.000–0.051	2

K, number of different repeats present at this locus; MLVA, multiple-locus variable-number tandem repeat analysis; VNTR, variable number tandem repeat.

Molecular typing of the isolates by PFGE and MLVA

Using any band difference in a pattern to designate MLVA types, 44 different MLVA types were identified among the 114 isolates (Fig. 1). Half of the isolates (57/114) were classified into four MLVA profiles (6,3,4.5,0,1,1,4,1,1–3,5,2,8,0.5,1,4,5,1–3,3,3,0,1,1,4,1,1–4,5,3,5.5,2,2,2,3,1). Allele string of 3,5,2,8,0.5,1,4,5,1 was the most commonly found MLVA type (MT7) representing 29.8% (34/114) of the isolates. If the size of some amplicons was similar to the flanking sequence, the number of repeat was then considered zero. Some alleles were reported as half size, which may result from the presence of intermediately sized repeat units or small deletions in the flanking sequence.

FIG. 1.

The minimum spanning tree based on nine-loci MLVA allelic profiles of 114 ESBL-producing Klebsiella pneumoniae strains. Each circle corresponds to a different MLVA type (MT). The size of the circle reflects the number of isolates with a particular genotype. The lines clustering circles show MLVA types that belong to the same complex. The circle shown by the arrow contained the largest number of strains (n = 34). MLVA complexes were assigned if two neighboring types did not differ in more than two VNTR loci. ESBL, extended-spectrum β-lactamases; MLVA, multiple-locus variable-number tandem repeat analysis.

To examine the relatedness among the identified MLVA types, an MST analysis was performed based on the categorical data (Fig. 1). Seven CCs were obtained. Three largest CCs identified were CC1, CC2, and CC4. CC1 consisting of 24 isolates was the largest clonal complex and 6 MLVA types (MT1–6) were identified in this clonal complex. All of CC1 strains, except one (23/24) were isolated from Hospitals 1 and 2. CC4 was composed of 13 isolates and 3 MLVA types (MT18–20) were identified in CC4 (MT20 was isolated from intensive care units [ICUs] of Hospitals 1 and 2). CC2 was composed of nine isolates and five MLVA types (MT26–30) were identified in this clonal complex. All members of this complex were isolated from Hospitals 2 and 3. Other complexes, CC3, CC5, CC6, and CC7 were composed of 5, 3, 2, and 2 isolates, respectively. The largest circle shown by the arrow in Figure 1 contains the largest number of strains (n = 34). All 34 isolates shared the same pattern (3,5,2,8,0.5,1,4,5,1; MT7) and were isolated from the ICU ward of Hospital 1 with the exception of one isolate, which was isolated from the emergency ward of Hospital 2.

PFGE analysis identified 64 distinct PFGE profiles among the 114 isolates. Pulsotypes P26 and P28 were the most common pulsotypes that appeared in 17 (14.9%) and 11 (9.6%) isolates, respectively. All isolates of pulsotypes P26, P27, and P28, as well as P29 were isolated from the ICU ward of Hospital 1, except one isolate which was obtained from the emergency ward of Hospital 2. Similar PFGE patterns (similarity above 90%) were seen between these isolates, and they shared identical VNTR profile (as 3,5,2,8,0.5,1,4,5,1; MT7). A similar PFGE pattern (similarity above 80%) was seen between isolates of P46-49 collected from NICU (seven isolates), women (one isolate), and neonatal (one isolate) wards, and they shared almost identical VNTR profiles differing by only one locus (locus A). Figure 2 shows the distribution of the PFGE types in the studied hospitals (strain numbers, hospital wards, VNTR profile, etc.).

FIG. 2.

Cluster analysis of 114 ESBL-producing isolates of K. pneumoniae based on PFGE-XbaI pattern using the Dice coefficient and UPGMA algorithm at a 1.5% tolerance and a 1.5% optimization. Characteristics of the strains, including strain numbers, ward, specimen, and MLVA allele string are indicated. The MLVA allele string is indicated in the order of A, E, H, J, K, D, N1, N2, and N4. A dash (–) in the VNTR profile denotes the absence of a PCR product at that locus. Trachea: Tracheal secretions, Pulmonary: Pulmonary secretions, Eye exu: Eye exudates, Peritonea: Peritoneal fluid. PFGE, pulsed-field gel electrophoresis; UPGMA, unweighted pair group method with arithmetic mean; VNTR, variable number tandem repeat.

Comparison of MLVA and PFGE

PFGE showed to be more discriminatory than MLVA; the discriminatory power of MLVA was 0.896 (Simpson's Diversity Index) with a 95% confidence interval of 0.850–0.942. The discriminatory power of PFGE showed a value of 0.962 with a 95% confidence interval of 0.943–0.981. To determine the congruence between the MLVA and PFGE methods, the adjusted Wallace coefficient was calculated. Overall, concordance between the methods was low: the chance that two isolates with the same PFGE type also shared the same MLVA type (MT) was 94.6%. By contrast, two strains with the same MLVA type had a 32.5% chance of having the same PFGE type. The MLVA profile of “3,5,2,8,0.5,1,4,5,1; MT7” was assigned to 34 strains, dividing into four PFGE profiles (P26, P27, P28, P29). Ten isolates with MLVA type 1 (MT1) were subdivided into eight different PFGE profiles. Furthermore, MLVA types MT18, MT4, MT36, MT26, and MT3 were each found for 4, 6, 2, 5, and 3 isolates that were discriminated into 4, 4, 2, 5, and 3 PFGE types, respectively; However, six isolates with PFGE type P46 showed three different MLVA profiles (MT1–3), varying in one locus (VNTR_A), and two isolates with PFGE type P59 showed two different MLVA types (MT26 and MT27) that differed in one locus (VNTR_J) only (Fig. 2).

Discussion

In this study, we compared the usefulness of PFGE and MLVA based on the previously reported VNTR loci^1,8 for typing of 114 ESBL-producing K. pneumoniae isolated from Iranian patients.

Several different genotyping methods have been used to study the epidemiology of K. pneumoniae.^1,20,21 Although for many years PFGE has been the method of choice because of its discriminatory ability,^22,23 the major disadvantage of this method is its poor reproducibility, making it unsuitable for comparison between laboratories.⁷ In contrast to PFGE, MLVA is rapid, cost-effective, and easy to interpret, which are general characteristics of PCR-based typing methods. Its reproducibility makes MLVA a good choice for epidemiological analyses.⁷ Turton et al. in 2010 were the first to describe an MLVA scheme for K. pneumoniae based on eight different VNTR loci (A, E, H, J, D, I, L, K).¹ Their results showed that some loci were not helpful, and therefore, in their next work in 2012,⁸ they used nine VNTR loci; six of the nine loci that were used, overlapped with loci included in the previously published MLVA. We performed MLVA based on these nine VNTR loci to compare the usefulness of this MLVA scheme with the PFGE in typing of 114 ESBL-producing K. pneumoniae isolates. The test population chosen for this study probably comprises the largest collection of strains ever used for such an analysis of this microorganism. In our study, calculation of the diversity indices for the MLVA loci showed that VNTR_J was the most polymorphic with 12 different alleles and the highest diversity index of 0.818. This finding supports those reported by Turton et al.¹ They also found that the VNTR_J was one of the most variable loci among the eight loci used in their MLVA scheme. VNTR_N4 was the least variable between our strains with two alleles and diversity index of 0.017. Obtaining such a low diversity index for VNTR_N4 suggests that this should be replaced with a more polymorphic VNTR. Also, more than half of our isolates (68/114, 59.6%) showed a repeat number of 1 at locus D with 2 alleles, and it was not found very helpful. Similar to Turton's study, our isolates failed to amplify mostly at locus K.

PFGE typing revealed a moderate genetic diversity with the identification of the 64 PFGE types among the 114 isolates. The PFGE demonstrated the clonal spread of the strains in the hospitals wards, especially in ICU and NICU, showing the vertical transmission of resistant isolates and low level of infection control strategies in these units. The PFGE pattern of pulsotype P29 detected in the ICU of Hospital 1 was also seen in a strain isolated from the emergency ward of Hospital 2, as both of them produced indistinguishable PFGE pattern. Also, the PFGE profile of P38 detected in the ICU of Hospital 2 was seen in a strain from ICU of Hospital 1 (Fig. 2). This may be due to the dissemination of ESBL-producing K. pneumoniae to other regions by interhospital transfer.

PFGE exhibited a higher discriminatory power than MLVA (0.962 versus 0.896; although the level of discrimination was not greatly different between them), and the overall concordance of the compared typing methods was low, most likely due to a greater discriminatory power of PFGE. Consequently, if a PFGE profile of an isolate is known, the MLVA type may be accurately predicted. If only the MLVA profile is known, however, the PFGE profile cannot be accurately predicted, since a MLVA profile may typically be corresponded to several PFGE profiles. For example, seven MLVA types (MT1, MT7, MT18, MT4, MT36, MT26, and MT3) were each found in isolates that were distinguishable by PFGE; but except for two pulsotypes P46 and P59, which were subdivided into three and two MLVA types, respectively, MLVA was not able to distinguish between strains with identical PFGE patterns. Our report is similar to those reported by Morris et al. they showed that PFGE was more discriminatory than MLVA.⁸ In their study on 13 K. pneumoniae producing KPC-2 carbapenemase, which were isolated during an outbreak in Ireland, isolates had identical MLVA profile but PFGE distinguished them into four profiles. It has been reported that the level of concordance between typing methods can vary and depends on the isolates collection. In fact, the MLVA validation study requires not only epidemiologically unrelated, but also closely-related isolates such as outbreak strains from hospitals, before drawing any conclusions about its discriminatory power.¹⁸ Although the 114 tested strains used in the current study were ESBL-producing isolates, they were collected from different specimens and wards of three hospitals.

In conclusion, our study showed that PFGE is more discriminatory than MVLA. Genotyping of K. pneumoniae is indispensable for monitoring of the spread of ESBL-producing strains, implementation of suitable infection control measures, and general epidemiology. Such objectives require development of a high-resolution genotyping tool with a fixed scheme.²⁴ MLVA is a PCR- based method and generates unambiguous data, in contrast to PFGE.²⁵ MLVA is useful as a screen tool for epidemiological analysis of an outbreak; however, optimization of the markers with more polymorphic VNTR loci is required to improve the discriminatory power of the method.

Footnotes

Acknowledgments

This study was supported by a grant from Tarbiat Modares University, Faculty of Medical Sciences, Tehran, Iran. The authors are grateful to all the staff of the microbiology laboratory at the three hospitals for collecting K. pneumoniae isolates used in this study. They would like to give special thanks to the pediatric infection research center personnel of Mofid Children Hospital.

Disclosure Statement

No competing financial interests exist.

References

Turton

J.F.

, Perry

, Elgohari

, and Hampton

C.V.

. 2010. PCR characterization and typing of Klebsiella pneumoniae using capsular type-specific, variable number tandem repeat and virulence gene targets. J. Med. Microbiol., 59(Pt 5):541–547.

Casella

, Rodriguez

M.M.

, Takahashi

J.T.

, Ghiglione

, Dropa

, Assuncao

, Nogueira

M.L.

, Lincopan

, Gutkind

, and Nogueira

M.C.

. 2015. Detection of blaCTX-M-type genes in complex class 1 integrons carried by Enterobacteriaceae isolated from retail chicken meat in Brazil. Int. J. Food Microbiol., 197:88–91.

Eckert

, Gautier

, and Arlet

. 2006. DNA sequence analysis of the genetic environment of various bla_CTX-M genes. J. Antimicrob. Chemother., 57:14–23.

Najar-Peerayeh

, Rostami

, Siadat

S.D.

, and Derakhshan

. 2014. High rate of aminoglycoside resistance in CTX-M-15 producing Klebsiella pneumoniae isolates in Tehran, Iran. Lab. Med., 45:231–237.

Derakhshan

, Najar-Peerayeh

, and Bakhshi

. 2016. Genotyping and characterization of CTX-M-15-producing Klebsiella pneumoniae isolated from an Iranian hospital. J. Chemother., 28:289–296.

Memariani

, Najar-Peerayeh

, Zahraei Salehi

, and Shokouhi Mostafavi

. 2015. Occurrence of SHV, TEM and CTX-M β-Lactamase genes among Enteropathogenic Escherichia coli strains isolated from children with diarrhea. Jundishapur. J. Microbiol., 8:e15620.

Sekse

, Sunde

, Lindstedt

B.A.

, Hopp

, Bruheim

, Cudjoe

K.S.

, Kvitle

, and Urdahl

A.M.

. 2011. Potentially human-pathogenic Escherichia coli O26 in Norwegian sheep flocks. Appl. Environ. Microbiol., 77:4949–4958.

Morris

, Boyle

, Morris

, Condon

, Delannoy-Vieillard

A.S.

, Power

, Khan

, Morris-Downes

, Finnegan

, Powell

, Monahan

, Burns

, O'Connell

, Boyle

, O'Gorman

, Humphreys

, Brisse

, Turton

, Woodford

, and Cormican

. 2012. Inter-hospital outbreak of Klebsiella pneumoniae producing KPC-2 carbapenemase in Ireland. J. Antimicrob. Chemother., 67:2367–2372.

Aqel

A.A.

, Meunier

, Alzoubi

H.M.

, Masalha

I.M.

, and Woodford

. 2014. Detection of CTX-M-type extended-spectrum beta-lactamases among Jordanian clinical isolates of Enterobacteriaceae. Scand. J. Infect. Dis., 46:155–157.

10.

Drew

R.J.

, Turton

J.F.

, Hill

R.L.

, Livermore

D.M.

, Woodford

, Paulus

, and Cunliffe

N.A.

. 2013. Emergence of carbapenem-resistant Enterobacteriaceae in a UK paediatric hospital. J. Hosp. Infect., 84:300–304.

11.

Hall

J.M.

, Corea

, Sanjeewani

H.D.

, and Inglis

T.J.

. 2014. Molecular mechanisms of beta-lactam resistance in carbapenemase-producing Klebsiella pneumoniae from Sri Lanka. J. Med. Microbiol., 63(Pt 8):1087–1092.

12.

Jain

, Hopkins

K.L.

, Turton

, Doumith

, Hill

, Loy

, Meunier

, Pike

, Livermore

D.M.

, and Woodford

. 2014. NDM carbapenemases in the United Kingdom: an analysis of the first 250 cases. J. Antimicrob. Chemother., 69:1777–1784.

13.

Skally

, Duffy

, Burns

, Doyle

, Foley

, Thomas

, Collins

, Smyth

, Turton

, and Humphreys

. 2014. What may be lurking in the hospital undergrowth? Inapparent cross-transmission of extended-spectrum beta-lactamase-producing Klebsiella pneumoniae. J. Hosp. Infect., 88:156–161.

14.

Kuske

C.R.

, Banton

K.L.

, Adorada

D.L.

, Stark

P.C.

, Hill

K.K.

, and Jackson

P.J.

. 1998. Small-scale DNA sample preparation method for field PCR detection of microbial cells and spores in soil. Appl. Environ. Microbiol., 64:2463–2472.

15.

Vu-Thien

, Corbineau

, Hormigos

, Fauroux

, Corvol

, Clement

, Vergnaud

, and Pourcel

. 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.

16.

Center for Disease Control and Prevention. 2009. The national molecular subtyping network for foodborne disease surveillance. One-day (24–28 h) standardized laboratory protocol for molecular subtyping of Escherichia coli O157:H7, Salmonella serotypes, Shigella sonnei, and Shigella flexneri by Pulsed Field Gel Electrophoresis (PFGE). CDC, Pulsenet.

17.

Tien

Y.Y.

, Wang

Y.W.

, Tung

S.K.

, Liang

S.Y.

, and Chiou

C.S.

. 2011. Comparison of multilocus variable-number tandem repeat analysis and pulsed-field gel electrophoresis in molecular subtyping of Salmonella enterica serovars Paratyphi A. Diagn. Microbiol. Infect. Dis., 69:1–6.

18.

Elberse

K.E.

, Nunes

, Sa-Leao

, van der Heide

H.G.

, and Schouls

L.M.

. 2011. Multiple-locus variable number tandem repeat analysis for Streptococcus pneumoniae: comparison with PFGE and MLST. PLoS One., 6:e19668.

19.

Zaluga

, Stragier

, Van Vaerenbergh

, Maes

, and De Vos

. 2013. Multilocus variable-number-tandem-repeats analysis (MLVA) distinguishes a clonal complex of Clavibacter michiganensis subsp. michiganensis strains isolated from recent outbreaks of bacterial wilt and canker in Belgium. BMC Microbiol. 13:126.

20.

Diancourt

, Passet

, Verhoef

, Grimont

P.A.

, and Brisse

. 2005. Multilocus sequence typing of Klebsiella pneumoniae nosocomial isolates. J. Clin. Microbiol., 43:4178–4182.

21.

van der Zee

, Steer

, Thijssen

, Nelson

, van't Veen

, and Buiting

. 2003. Use of multienzyme multiplex PCR amplified fragment length polymorphism typing in analysis of outbreaks of multiresistant Klebsiella pneumoniae in an intensive care unit. J. Clin. Microbiol., 41:798–802.

22.

Cookson

B.D.

, Robinson

D.A.

, Monk

A.B.

, Murchan

, Deplano

, de Ryck

, Struelens

M.J.

, Scheel

, Fussing

, Salmenlinna

, Vuopio-Varkila

, Cuny

, Witte

, Tassios

P.T.

, Legakis

N.J.

, van Leeuwen

, van Belkum

, Vindel

, Garaizar

, Haeggman

, Olsson-Liljequist

, Ransjo

, Muller-Premru

, Hryniewicz

, Rossney

, O'Connell

, Short

B.D.

, Thomas

, O'Hanlon

, and Enright

M.C.

. 2007. Evaluation of molecular typing methods in characterizing a European collection of epidemic methicillin-resistant Staphylococcus aureus strains: the HARMONY collection. J. Clin. Microbiol., 45:1830–1837.

23.

Schouls

L.M.

, Spalburg

E.C.

, van Luit

, Huijsdens

X.W.

, Pluister

G.N.

, van Santen-Verheuvel

M.G.

, van der Heide

H.G.

, Grundmann

, Heck

M.E.

, and de Neeling

A.J.

. 2009. Multiple-locus variable number tandem repeat analysis of Staphylococcus aureus: comparison with pulsed-field gel electrophoresis and spa-typing. PLoS One, 4:e5082.

24.

Brink

A.A.

, von Wintersdorff

C.J.

, van der Donk

C.F.

, Peeters

A.M.

, Beisser

P.S.

, Stobberingh

E.E.

, and Wolffs

P.F.

. 2014. Development and validation of a single-tube multiple-locus variable number tandem repeat analysis for Klebsiella pneumoniae. PLoS One, 9:e91209.

25.

Gierczynski

, Golubov

, Neubauer

, Pham

J.N.

, and Rakin

. 2007. Development of multiple-locus variable-number tandem-repeat analysis for Yersinia enterocolitica subsp. palearctica and its application to bioserogroup 4/O3 subtyping. J. Clin. Microbiol., 45:2508–2515.