Regulation of rRNA transcription correlates with nucleoside triphosphate sensing

Abstract

We have previously shown that the activity of the Escherichia coli rRNA promoter rrnB P1 in vitro depends on the concentration of the initiating nucleotide, ATP, and can respond to changes in ATP pools in vivo. We have proposed that this nucleoside triphosphate (NTP) sensing might contribute to regulation of rRNA transcription. To test this model, we have measured the ATP requirements for transcription from 11 different rrnB P1 core promoter mutants in vitro and compared them with the regulatory responses of the same promoters in vivo. The seven rrnB P1 variants that required much lower ATP concentrations than the wild-type promoter for efficient transcription in vitro were defective for response to growth rate changes in vivo (growth rate-dependent regulation). In contrast, the four variants requiring high ATP concentrations in vitro (like the wild-type promoter) were regulated with the growth rate in vivo. We also observed a correlation between NTP sensing in vitro and the response of the promoters in vivo to deletion of the fis gene (an example of homeostatic control), although this relationship was not as tight as for growth rate-dependent regulation. We conclude that the kinetic features responsible for the high ATP concentration dependence of the rrnB P1 promoter in vitro are responsible, at least in part, for the promoter's regulation in vivo, consistent with the model in which rrnB P1 promoter activity can be regulated by changes in NTP pools in vivo (or by hypothetical factors that work at the same kinetic steps that make the promoter sensitive to NTPs).

FIG. 1

DNA sequences in the core promoter region of wild-type and mutant rrnB P1 promoters. Numbers above the wild-type sequence refer to promoter positions with respect to the transcription initiation site. The −10 and −35 hexamers are underlined and in bold in the wild-type promoter and underlined in mutant promoters. Mutations are indicated in bold uppercase letters and underlined. The rrnB P1 promoters analyzed extended from −66 to +9.

FIG. 2

Effect of ATP concentration on in vitro transcription of wild-type and mutant rrnB P1 promoters. Transcription was normalized to the highest level for each promoter. The wild-type regression fit is in bold. (A) Mutants that required at least threefold lower ATP concentrations for half-maximal transcription than wild-type rrnB P1. (B) Mutants that required ATP concentrations similar to or higher than that required by the wild type. Note the different x-axis scale than in panel A.

FIG. 3

Growth rate-dependent control (GRDC) of wild-type and mutant rrnB P1 promoters. Each panel includes data points from three independent experiments for each promoter. The wild-type regression fit is in bold. Note that all of the panels have different ordinate scales. The top panels display the actual β-galactosidase activities versus the growth rate (doublings per hour). The bottom panels display the promoter activities normalized at the lowest growth rate in order to facilitate visualization of defects in regulation (see text). (A and C) Mutants that are defective for growth rate-dependent regulation. (B and D) Mutants with activities that increase with growth rate similarly to that of the wild type.

FIG. 4

Summary of regulation of wild-type (wt) and mutant rrnB P1 promoters. Promoter mutants are listed in the same order for each panel, from the promoter with the lowest [ATP]_1/2max to that with the highest. The bars for the wild-type promoter are black, those for the seven low-NTP, growth rate regulation-defective mutants are light gray, and those for the four high-NTP, growth rate regulated mutants are dark gray. The bars for the promoter mutants in panel C are not separated into classes because distinctions are somewhat ambiguous in this assay. (A) The mean [ATP]_1/2max values were determined from at least two independent experiments. Variation was less than 20%, except for C-19T (27%). (B) Average fold increase with growth rate was calculated for the wild type and mutants (except C-17T) by dividing the β-galactosidase activity in LB (μ = ∼1.4) by the activity in M9 glycerol (μ = ∼0.33). The averages and standard deviations (less than 14%) were calculated for three independent experiments. The activity of the C-17T promoter is too close to the background in M9 glycerol for accurate assessment, so its fold increase was estimated by dividing the activity in LB by the activity in M9 glucose (μ = 0.58). C-17T increased approximately twofold more than the wild-type promoter in this growth rate range (standard deviation = 26%). Therefore, its fold increase with growth rate may be an underestimate. (C) Feedback derepression in fis mutant strains. fis/wild-type promoter activity ratio is reported as in Table 2.