High frequency of mutations in the rpoB gene in rifampin-resistant clinical isolates of Mycobacterium tuberculosis from Singapore.

Rifampin (RIF) is a first-line antituberculosis drug. Resistance to RIF, in the majority of cases, has been associated with mutations within an 81-bp RIF resistance-determining region (RRDR) of the rpoB gene, which encodes the subunit of the RNA polymerase (8). RIF acts by binding to the subunit of the RNA polymerase, thus interfering with transcription and RNA elongation. Although targeted molecular analysis of the rpoB gene has been shown to be effective for detecting RIF resistance in over 90% of RIF-resistant strains from diverse geographical regions, there is only limited information from Southeast Asia (2, 3, 11, 12). Previous investigations on isoniazid-, pyrazinamide-, and ofloxacin-resistant Mycobacterium tuberculosis isolates from Singapore indicated that the mutation frequencies and spectra of the genes studied differed from those of genes from isolates from other geographical locations (4–6). To determine if targeted molecular screening of the rpoB gene may be useful in Singapore, we evaluated the frequency of mutations within the RRDR of rpoB for 51 RIF-resistant and 4 RIF-sensitive Mycobacterium tuberculosis isolates by direct sequencing of the entire RRDR. Consecutive RIF-resistant isolates were collected from August 1994 to December 1996 as previously described (4). The clinical data of this study population have been detailed elsewhere (1). Drug susceptibility testing was done by using the BACTEC 460 radiometric method (Becton Dickinson, Towson, Md.), and the test drug concentration was 2 g/ml. DNA extracted from bacterial colonies was used for PCR amplification of a 350-bp fragment of rpoB, as previously described (3). The PCR products were purified (Wizard PCR purification kit; Promega, Madison, Wis.) and directly sequenced by using the Applied Biosystems 377 DNA sequencer. Twelve different missense mutations involving codons 513, 514, 516, 522, 526, and 531 were identified in 49 RIF-resistant strains (Table 1). The most common alterations were a Ser3Leu substitution at codon 531, present in 25 (49.0%) isolates, and a His3Tyr substitution at codon 526 in 7 (13.7%) isolates. Mutations at codon 531 were observed in 28 (55%) of the isolates, at codon 526 in 12 (24%) isolates, and at codon 516 in 6 (12%) isolates. One isolate had mutations in both codons 514 and 516. A rare mutation at codon 513, resulting in a Gln3Glu substitution, was identified in one isolate. This mutation has only been reported once, in an isolate from Taiwan (7). Similarly, the deletion of codon 518 was observed in one isolate, and this aberration has been observed in only two other studies in Africa and Belgium (9, 10). No genetic alterations were detected in 2 (3.9%) of the 51 RIF-resistant isolates. Four RIF-sensitive isolates were included as controls, and no mutations were identified in these isolates. Genetic alterations in the RRDR were present at a high frequency of 96.1% in our Singaporean isolates, and the most commonly mutated residues were Ser-531, His-526, and Asp516, as has also been demonstrated by the majority of studies worldwide (2, 8, 13). In conclusion, this high rate of mutations within the RRDR of the rpoB gene suggests that targeted screening of the RRDR may be feasible for the determination of RIF resistance in clinical isolates of Mycobacterium tuberculosis from Singapore.