Improved differentiation of Mycobacterium tuberculosis isolates in the north of England using additional variable number tandem repeat loci

Abstract

Mycobacterial interspersed repetitive unit-variable number tandem repeat (MIRU-VNTR) genotyping of over 3300 Mycobacterium tuberculosis isolates from the north of England has identified large clusters of strains which share common profiles. However, many apparent clusters identified when typed using the existing 15 loci lack clear epidemiological links. This study seeks to discover whether or not six additional VNTR loci can increase the discriminatory power of the existing MIRU-VNTR 15-loci technique. Two hundred and six M. tuberculosis isolates were genotyped, including 57 isolates from 20 epidemiologically linked clusters and 149 from unlinked patients belonging to six large MIRU-VNTR-defined clusters. The discriminatory power of the six additional loci was high (Hunter Gaston Discriminatory Index [HGDI]: 0.952). Five of the six loci were highly discriminative (h>0.6); however, locus 2401 was less discriminative (h=0.5). The additional VNTR loci were able to subtype all six unlinked common MIRU-VNTR clusters into 56 subclusters, significantly differentiating unrelated strains in a set previously incorrectly clustered using 15 MIRU-VNTR loci. The largest cluster size was 14 (9.3%) when typed using the six additional VNTR loci, compared to 30 (20%) when typed using the original 15 MIRU-VNTR loci. The same loci were also found to be stable as a result of their inability to subdivide any of the epidemiologically linked clusters. This study has demonstrated that expanding the MIRU-VNTR panel beyond the 15 previously used loci significantly increases the discriminatory power of the technique and thus provides a valuable tool in the epidemiological monitoring of this disease.