From these results, we can conclude that the effect of mutant 8R on transcription is exclusively due to the alteration in the −35 box, whereas the downstream mutation does not contribute to the ability of the RNAP to bind the bmrA promoter. Most probably, the upstream mutation improves the initial binding check details of the RNAP. In vitro transcription experiments were

carried out using B. subtilis RNAP and wild type and the three mutated template DNAs covering the bmrA promoter and a region downstream from the transcription start site. Figure 4 shows the formation of a visible band only in lanes 2 (MW) and 4 (MM), which is in accordance with the data obtained by real-time PCR on the amount of mRNA in the wild type and double mutant strain as well as the results of the lacZ reporter gene assays. Furthermore, the in vitro transcription data substantiate the RO4929097 results of the EMSA. To confirm that the increased levels of bmrA mRNA correspond to an increase in the corresponding protein level, membrane protein fractions were prepared from wild type and double mutant 8R and separated on a 12% sodium dodecyl sulfate-polyacrylamide gel. As shown in Fig. 5, a new band of ≈64 kDa is visible in the mutant fraction that is hardly detectable in the wild-type extract from B. subtilis 168. Elution of the band, its digestion with trypsin and subsequent

analysis confirmed that this band consists of BmrA. A mutant strain of B. subtilis 168 containing only the single mutation in the −35 box of the bmrA promoter designated B. subtilis YH2M grew only in the presence of 3 μM CmC, but not at 4 μM CmC, in contrast to the fragment containing both mutations that transformed B. subtilis to resistance against 5 μM CmC. A fragment comprising just the +6 mutation was used to transform B. subtilis 168 and B. subtilis YH2M. The resulting double transformant containing the −35 and the +6 mutation grew in the presence of 5 μM CmC. Transformation of B. subtilis 168 with this fragment did not yield any transformants growing in the presence GPX6 of >1 μM CmC. In vitro

studies using EMSA and transcription experiments showed no influence of this +6 mutation on the promoter activity. These data show that the stepwise increase in CmC resistance during mutant selection is due to the cumulative effect of two mutations in the promoter region. Apparently, both mutations cooperate to yield the 5 μM CmC resistance found in the double mutant 8R. All constructs were proven by sequencing PCR fragments obtained from their genomic DNA. Because the results of the lacZ reporter gene fusions, EMSAs and in vitro transcription indicated that only the upstream mutation in the −35 box affected RNAP binding, and hence, the total amount of bmrA mRNA, we can now draw the conclusion that the downstream mutation in the noncoding region of bmrA is responsible for the stabilization of bmrA mRNA.