Application of the optimal 24 MIRU-VNTR loci set for Mycobacterium tuberculosis strains improves the correlation between strain typing and epidemiological data.

Authors: Jason T. Evans (1), Beck Taylor (2), Desmond Estephane (1), Sarah Gardiner (1), E. Grace Smith (1), Peter M. Hawkey (1).

Affiliations:

(1) HPA West Midlands Laboratory, Heart of England NHS Foundation Trust, Bordesley Green East, Birmingham B9 5SS, UK.

(2)Heart of Birmingham Teaching Primary Care Trust, Bartholomew House, 142 Hagley Road, Edgbaston, Birmingham, B16 9PA.

Objectives: The utility of DNA fingerprinting of M. tuberculosis strains has been enhanced by MIRU-VNTR typing. An enhanced set of 24 loci that offers greater discrimination over the original 12 loci and aims to improve the concordance of strain typing data with epidemiological data has been published. The aim of this study was to evaluate the optimal MIRU-VNTR loci set in discriminating clusters defined by the original MIRU-VNTR loci that have varying levels of epidemiological links.

 

Methods: Six clusters containing 71 M. tuberculosis strains typed using the original MIRU-VNTR loci set were selected for further analysis by the optimal MIRU-VNTR loci set. Epidemiological links within each of the 6 clusters were investigated. The 6 clusters contained a range of varying levels of epidemiological links ranging from none found to definitive. MIRU-VNTR analysis was carried out by PCR and fragment sizing by agarose gel electrophoresis.

Results: Using the original set of MIRU-VNTR loci the degree of concordance between molecular data and epidemiological data was 61%. Analysis using 24 loci increased the level of concordance to 93%.

Conclusion: The optimal set of M. tuberculosis MIRU-VNTR loci greatly increases the concordance between strain typing and epidemiological data. This further enhances the utility of MIRU-VNTR loci in DNA fingerprinting of M. tuberculosis strains as clusters identified by the original MIRU-VNTR loci set but with no apparent epidemiological links are differentiated into epidemiological relevant sub-groups whereas epidemiological relevant clusters are not significantly split by the optimal MIRU-VNTR loci set.