RESUMEN
BACKGROUND: There is a need for rapid, inexpensive methods for analysing a limited number of Mycobacterium tuberculosis strains. The ligation-mediated polymerase chain reaction (LM-PCR) method appears to be sufficiently discriminative and reproducible to be considered as a molecular tool for the initial evaluation of hospital outbreaks, laboratory cross-contamination, and family or small community transmission. OBJECTIVE: To develop a new LM-PCR method based on PCR amplification of the 5'-flanking region of insertion sequence (IS) 6110 consisting of SalI/PvuII digestion of chromosomal DNA, ligation of a SalI linker and differentiation of IS6110-carrying restriction fragments by suppression subtractive hybridisation. DESIGN: The fast ligation amplification polymorphism (FLAP) method was applied in the analysis of 62 M. tuberculosis clinical isolates and compared with IS6110-restriction fragment length polymorphism (RFLP) and mycobacterial interspersed repetitive units-variable number of tandem repeat (MIRU-VNTR) analyses of the same strains. RESULTS: The sensitivity of FLAP was estimated at 0.25 ng/l. FLAP yielded 32 patterns among the 62 M. tuberculosis strains compared to respectively 28 and 36 patterns obtained using MIRU-VNTR and IS6110-RFLP. Its Hunter-Gaston discriminatory index value (0.973) is similar to that of MIRU-VNTR (0.966) and IS6110-RFLP (0.971). The specificity of the FLAP patterns was also confirmed. CONCLUSION: FLAP proved highly discriminating, sensitive and specific and could be a valuable molecular tool, especially for analysing a limited number of M. tuberculosis strains.