Detection of rpoB mutations in Mycobacterium tuberculosis with LightCycler technology.

In a recent article, Edwards et al. evaluated three biprobes to detect mutations in rifampin-resistant isolates of Mycobacterium tuberculosis (1). The authors claimed that their approach allowed the detection of all mutations in the 46 isolates studied. We read their paper with interest, because we have used LightCycler PCR to detect resistance to rifampin and isoniazid in M. tuberculosis (3). However, there are aspects of their work that warrant an explanation. The first and major issue concerns biprobe C, which spans codon 526. The authors observed three melting peaks with this probe, one for the wild-type strains (61°C), one for the strains with mutations at codon 526 (49°C), and one for the strains with mutations at codon 531 (66°C). The third melting peak was confirmed by testing 27 strains. Given that biprobe C spans codons 523 to 529, it can hardly be used to detect a mutation at codon 531 (see GenBank accession number L27989 for rpoB). The third peak may, however, correspond to an unreported mutation within codons spanned by the probe (i.e., 523 to 529). The authors should also explain how it is possible that the melting temperature for this unreported mutation is 5°C higher than that for the wild-type. The only valid explanation is that the sequence of the probe they used did not perfectly match the wild-type sequence. In that case, the mutant would indeed have been a better match and have therefore had a higher melting temperature. However, according to the published sequence (5), the perfect match occurs in fact between the probe and the wild-type sequence. Yet another point that requires clarification is the reported melting peaks of biprobes B and C. Edwards et al. reported a melting temperature for the mutants at a certain codon, independent of the type of mutation. In our experience, distinguishing a shift in the melting peak when there is a different single substitution at the same codon is difficult, though some authors have been able to detect different mutations at the same position due to differences in melting temperature (2, 4). In strains with a double mutation, and the author reported several of them, a shift in the melting temperature ought to be observed.