Multiple-locus variable number tandem repeat analysis of Staphylococcus aureus: comparison with pulsed-field gel electrophoresis and spa-typing

Abstract

Background: Molecular typing of methicillin-resistant Staphylococcus aureus (MRSA) is required to study the routes and rates of transmission of this pathogen. Currently available typing techniques are either resource-intensive or have limited discriminatory ability. Multiple-locus variable number tandem repeat analysis (MLVA) may provide an alternative high throughput molecular typing tool with high epidemiological resolution.

Methodology/principal findings: A new MLVA scheme for S. aureus was validated using 1681 S. aureus isolates collected from Dutch patients and 100 isolates from pigs. MLVA using 8 tandem repeat loci was performed in 2 multiplex PCRs and the fluorescently labeled PCR products were accurately sized on an automated DNA sequencer. The assessed number of repeats was used to create MLVA profiles consisting of strings of 8 integers that were used for categorical clustering. MLVA yielded 511 types that clustered into 11 distinct MLVA complexes which appeared to coincide with MLST clonal complexes. MLVA was at least as discriminatory as PFGE and twice as discriminatory as spa-sequence typing. There was considerable congruence between MLVA, spa-sequence typing and PFGE, at the MLVA complex level with group separation values of 95.1% and 89.2%. MLVA could not discriminate between pig-related MRSA strains isolated from humans and pigs, corroborating the high degree of relationship. MLVA was also superior in the grouping of MRSA isolates previously assigned to temporal-spatial clusters with indistinguishable SpaTypes, demonstrating its enhanced epidemiological usefulness.

Conclusions: The MLVA described in this study is a high throughput, relatively low cost genotyping method for S. aureus that yields discrete and unambiguous data that can be used to assign biological meaningful genotypes and complexes and can be used for interlaboratory comparisons in network accessible databases. Results suggest that MLVA offsets the disadvantages of other high discriminatory typing approaches and represents a promising tool for hospital, national and international molecular epidemiology.

Figure 1. Schematic map of the genome of S. aureus strain N315, indicating the positions of the VNTR loci used in MLVA.

Figure 2. Minimum spanning tree of the 1681 S. aureus isolates typed by MLVA.

Clustering of MLVA profiles was done using a categorical coefficient. In the minimum spanning tree the MLVA types are displayed as circles. The size of each circle indicates the number of isolates with this particular type. Thick solid lines connect types that differ in a single VNTR locus and a thin solid connects types that differ in 2 VNTR loci. The color of the halo surrounding the MLVA types denote types that belong to the same complex. MLVA complexes were assigned if 2 neighboring types did not differ in more than 1 VNTR locus and if at least 5 types fulfilled this criterion. MLVA complexes are also indicated in characters e.g. MC8 denotes MLVA complex 8.

Figure 3. Minimum spanning tree of 882 S. aureus isolates typed by spa-sequence typing.

The SpaTypes are displayed as circles. For description of the symbols, lines etc. see legend of figure 2. Clustering was performed using the Bionumerics spa-plugin with a conservative alignment setting of 100% and 0% maximum duplication length. Halos surrounding SpaTypes indicate related SpaTypes.

Figure 4. Minimum spanning tree of 1304 S. aureus isolates typed by PFGE.

The PFGE types are displayed as circles. Clustering was performed using the Pearson correlation coefficient and a 2% optimization setting. The resulting similarity matrix was used to construct a minimum spanning tree using a 10% similarity bin size. Related PFGE grouped and this is denoted by the double circle (compacted complexes). The MLVA complex is denoted as the color of the circles (see figure 2) and is also indicated in characters.

Figure 5. Discriminatory power of MLVA, spa-sequence typing and PFGE for S. aureus.

Diversity is plotted as the number of types versus the percentage of isolates. To obtain the plot the list of typing results was first sorted by frequency and the percentage of isolates versus the number of types was determined. The plot only shows the results for the 529 S. aureus isolates that were analyzed by all 3 methods.

Figure 6. Concordance of the grouping obtained by MLVA and by MLST.

A minimum spanning tree was constructed based on MLVA data of 40 isolates for which MLST had already been performed. The circles represent the MLVA types and the numbers in the circles represent the STs. The numbers displayed between types denote the number of VNTR loci that differ between these types. The colors represent the MLST clonal complexes. The white circles are STs that have not been assigned to a particular clonal complex. Only single locus or double locus MLVA variants are connected by lines. The size of each circle indicates the number of isolates with this particular MLVA type.

Figure 7. In silico analysis of 14 S. aureus strains for which the genome sequence has been determined.

Strains were clustered on basis of their whole genome sequence, MLST profile, MLVA profile and spa-sequence type, respectively.

Figure 8. Minimum spanning tree of 295 pig MRSA isolates typed by MLVA.

Each circle represents an MLVA type. The orange circles or sectors of circles denote types obtained from 195 pig MRSA strains isolated from humans. The MLVA profiles of 100 pig MRSA strains isolated from pigs are indicated in light blue. The numbers in the circles denote the SpaTypes and their frequency within the 2 dominant MLVA types.