Lipiarmycin (Lprm) (8), a macrocyclic antibiotic also known as tiacumicin (2), is currently under development under the name of OPT-80 (Optimer Pharmaceuticals, Inc., San Diego, Calif.) as a narrow-spectrum antibacterial agent to treat Clostridium difficile-associated diarrhea (1, 7). Lprm is a transcription inhibitor, but unlike rifampin and streptolydigin, it preferentially inhibits holoenzyme transcription at a much greater rate than it inhibits transcription of the core enzyme (5, 6). Genetic mapping experiments in Bacillus subtilis indicate that mutants are located between loci determining rifampin resistance and streptolydigin resistance (6). To precisely identify the positions of the mutations, we selected lipiarmycin-resistant colonies and sequenced the domains of the RNA polymerase located between these two loci. Lprm was produced and purified by fermentation of Actinoplanes deccanensis (DSMZ 43806) by the method of Talpaert et al. (8). Mutant strains were isolated as spontaneous variants of Bacillus subtilis CIP 52.62 that were able to form colonies on nutrient agar plates containing Lprm (40 μg/liter). At this concentration the frequency of resistant colonies was less than 10−7. Among the 10 resistant colonies selected, 8 exhibited an increased MIC for several classes of antibiotics and were likely to be permeability mutants (Table (Table1).1). Two mutants (mut1 and mut2) were highly resistant to Lprm. TABLE 1. Antibiotic resistance and transcription activity of Lprm-resistant Bacillus subtilis To correlate the resistance phenotypes with the activity of Lprm on the RNA polymerase, we tested enzymes from the wild-type and mutant colonies in an in vitro transcription assay by the method of Sonenshein et al. (6). The parent cells (CIP 52.62) and mutant cells specifically resistant to Lprm (mut1 and mut2) showed the same pattern of sensitivity to rifampin. In contrast, the transcription activity of the wild-type cells was more strongly affected by the addition of Lprm than was the activity of either mut1 or mut2. This confirmed the link between resistance to Lprm resistance and transcription. mut3 to mut10 strains were equally sensitive to rifampin and Lprm, suggesting that the resistance was not due to a mutation in the polymerase. After purification of the genomic DNA, we PCR amplified and sequenced the regions located between the loci determining rifampin and streptolydigin resistance. The following primers were used: 5′-1285CGTGTGGTTCGTGAGAGAATGT1306-3′, 5′-3257TAAGCTTCAAGTGCCCAAACCT3236-3′, and 5′-2524CTTGTTGGTAAAGTAACGCCTA2545-3′ for rpoB and 5′-1163CGTTTCGCACTCTTAATGTTGTG1141-3′, 5′-593CACAAGGACAACGCCGTAC611-3′, and 5′-2804GTTAACTGTGTACCAGGCTCACC2782-3′ for rpoC. A point mutation resulting in the substitution of the R326 of rpoC by L (CG[T/C]TT) was found in mut1 and mut2 in a highly conserved region; no mutation was detected in mut3 to mut10. When the PCR product harboring the R326L mutation was transfected into B. subtilis CIP 52.62, the frequency of lipiarmycin-resistant bacteria was 100-fold higher than the frequency observed for bacteria with the wild-type fragment. By analogy with the three-dimensional structure of the Thermus aquaticus enzyme (3), R326 is located in proximity to region 3.2 of σ, which in turn occupies the same space as the exiting RNA transcript (3). This could delineate a new binding site for transcription inhibitors.