An IS6110-targeting fluorescent amplified fragment length polymorphism DNA fingerprinting alternative to IS6110 RFLP: a pilot study

Abstract number: 1733_1493

Thorne N., Evans J., Smith E.G., Hawkey P., Gharbia S., Arnold C.

Objectives: A rapid, simple, highly-discriminatory DNA fingerprinting methodology that produces data that can be easily interpreted, compared and transported is the ultimate goal for M. tuberculosis infection control professionals and epidemiologists. The current `gold standard', IS6110 RFLP, is far from meeting all these requirements; however its discriminatory power has not been matched by a rapid, alternative method that also analyses the IS6110 sequence.

Methods: A panel of 78 blinded isolates from TB cases that were either epidemiologically linked or distinct were selected for analysis by IS6110 RFLP and FAFLP. A pilot study for a novel Fluorescent Amplified Fragment Length Polymorphism (FAFLP) approach to M. tuberculosis DNA fingerprinting targeting the variable IS6110 marker was tested for specificity and reproducibility and compared with IS6110 RFLP typing results on the panel of isolates. IS6110 RFLP was carried out using the internationally standardised published method.

Results: IS6110 RFLP defined the 78 isolates into nine clusters containing 2–32 isolates, with a total of 63 clustered isolates and 15 unique isolates. There were seven low-copy number isolates included in which four clustered into two, each containing a pair of isolates. Comparison of in silico analysis with strain H37Rv found FAFLP detected fragments to be highly specific in vitro. The reproducibility of FAFLP fragment profiles for five repeated digestions with strain H37R was 100%. FAFLP and RFLP clustering results were found to be highly congruent. FAFLP clustering was highly similar to RFLP discrimination of the 78 isolates as unique or clustered strains. FAFLP allocated 97% of RFLP-clustered isolates to the same eight clusters as RFLP. FAFLP failed to cluster two single-copy strains that had been clustered by RFLP, however clustering based on a single IS6110 copy is not reliable and secondary fingerprinting is necessary.

Conclusion: The discriminatory power exhibited by FALP in this initial collection of isolates indicates that it would be a rapid and specific screening alternative to IS6110 RFLP. The accuracy of fragment sizing by FAFLP has allowed common fragments to be defined that are likely hot-spots for IS6110 transposition. FAFLP will enable rapid screening of strains for epidemiological investigations and initiate further research into the frequency, conservation and consequence of IS6110 hotspots and specific transposition sites.