Evaluation of use of multiple locus variable number tandem repeat analysis for typing of Pseudomonas aeruginosa

Abstract number: P514

Turton J., Turton S., Yarde S., Kaufmann M., Pitt T.

Objectives: To establish and evaluate a method using Variable Number Tandem Repeat (VNTR) loci suitable for routine typing of Pseudomonas aeruginosa in a reference laboratory.

Methods: PCR amplification of up to 12 VNTR loci was carried out on a panel of isolates previously characterised by pulsed-field gel electrophoresis (PFGE) and on serotype reference strains. Repeat numbers at targets with small repeat units were determined using fluorescent forward primers and sizing on a sequencer. Repeat numbers at the remaining loci were determined by agarose gel electrophoresis. A scheme using eight loci was adopted and tested on a further 100 isolates, also typed by PFGE.

Results: The twelve loci initially used were reduced to eight, with no loss of discrimination, at least among the 40 isolates originally tested. Of the 28 PFGE types represented by the 77 isolates in the panel, 27 were successfully distinguished by their repeat numbers at loci 172, 211, 214, 217, 222, 207 and 209. The Liverpool and Midlands 1 strains, isolated from multiple patients with cystic fibrosis (CF), could be identified unambiguously by their characteristic VNTR profiles at the 7 loci. This was also the case for other CF associated strains. There was some variation in repeat numbers at two of the 7 loci among isolates of clone C from different patients, but, otherwise, with the odd exception only, repeat numbers were consistent among representatives of a single PFGE type. The method successfully distinguished all 17 serotype reference strains, none of the profiles of which matched one another, or any of the panel or routine isolates tested. Repeat numbers at the eighth locus (61) could provide discrimination within a PFGE type. Representatives from outbreaks, received within a short space of time, all shared the same number of repeats at this locus. In contrast, isolates of the Liverpool strain from CF patients showed variation in repeat number at this locus, even among isolates from the same centre. Agreement with PFGE was good for the further 100 isolates tested, but in two instances, the VNTR analysis failed to distinguish pairs of isolates that were distinct by PFGE, except at locus 61.

Conclusion: In the vast majority of cases, VNTR analysis at these eight loci provided discrimination at a level similar to that afforded by PFGE, with most strains being identified unambiguously. Results could be obtained within a day, leading to significant improvements in reporting times.