A DnaA-dependent riboswitch for transcription attenuation of the his operon

Abstract

Transcription attenuation in response to the availability of a specific amino acid is believed to be controlled by alternative configurations of RNA secondary structures that lead to the arrest of translation or the release of the arrested ribosome from the leader mRNA molecule. In this study, we first report a possible example of the DnaA-dependent riboswitch for transcription attenuation in Escherichia coli. We show that (i) DnaA regulates the transcription of the structural genes but not that of the leader hisL gene; (ii) DnaA might bind to rDnaA boxes present in the HisL-SL RNA, and subsequently attenuate the transcription of the operon; (iii) the HisL-SL RNA and rDnaA boxes are phylogenetically conserved and evolutionarily important; and (iv) the translating ribosome is required for deattenuation of the his operon, whereas tRNA^His strengthens attenuation. This mechanism seems to be phylogenetically conserved in Gram-negative bacteria and evolutionarily important.

Keywords: DnaA‐dependent; Escherichia coli; his operon; riboswitch; transcription attenuation.

Figure 1

The stem‐loop attenuator region is required for transcription attenuation of the his operon and bestows a fitness advantage. (A) Deletion of the hisL‐SL region greatly increases his operon transcription. Wild‐type (wt) and mutant (∆hisL or ∆his‐SL) MOR1378 cells were exponentially grown in ABTGcasa medium at 37°C. The hisL, hisG, hisI, and lacZ genes are as indicated. Black arrows represent the hisL promoter (hisLp) and SL represents the stem‐loop region. The relative β‐galactosidase activity in the mutants was compared with that in wt cells as described in the Materials and Methods section. (B) his operon transcription is significantly enhanced in the dnaA _A345S mutant in a hisL‐SL‐dependent manner. wt and mutant (dnaA _A345S, ∆hisL, ∆his‐SL, ∆hisL dnaA _A345S, and ∆hisL‐SL dnaA _A345S) MOR1378 cells were exponentially grown and the relative β‐galactosidase activity in dnaA _A345S mutant cells was compared with that in control cells as described in the Materials and Methods section. The genotype of each strain is as shown. (C) Excess DnaA represses his operon transcription in a hisL‐SL sequence‐dependent manner. The relative β‐galactosidase activity was compared between wt and mutant (ΔhisL or Δhis‐SL) MOR1378 cells, in the presence of pdnaA116 (DnaA overproducer) or control vector (pLEX5BA) as described in (A). The genotype of each strain is as shown. (D, E) DnaA regulates the transcription of the structural hisG gene but not that of the leader hisL gene. (D) wt, dnaA _A345S, wt/control vector, and wt/pdnaA116 (DnaA overproducer) MOR1378 cells were exponentially grown in ABTGcasa medium at 37°C. The HisL and HisG RNA levels in total RNA (1 μg/lane) from each strain were analyzed by northern blot using digoxin‐labeled hisL and hisG probes, respectively. (E) The HisL or HisG RNA level in the same total RNA presented in (D) was determined by RT‐qPCR as described in the Materials and Methods section. The level of each RNA was normalized using the C _t value corresponding to the Escherichia coli rplO gene (internal reference) and calculated relative to that of wt cells. The values are the average of three independent experiments and the standard errors are as presented; ***p < 0.001, **p < 0.01, *p < 0.05, ns, not significant, one‐way analysis of variance.

Figure 2

Transcription attenuation of the his operon depends on both DnaA and the hisL‐SL sequence. The hisL promoter (hisLp), hisLp‐hisL, or hisLp‐hisL‐SL fragment was inserted in front of the promoterless lacZ gene in the pTAC3953 plasmid as described in the Materials and Methods section, yielding the phisLp, phisLp‐hisL, or phisLp‐hisL‐SL plasmids (A), respectively. In each case, lacZ expression was under the control of the hisLp, hisLp‐hisL, or hisLp‐hisL‐SL regulatory element. The relative β‐galactosidase activity in these plasmids was measured in exponentially growing wild‐type (wt) MC4100 cells (B), dnaA _A345S MC4100 cells (C), wt CM735 cells (D), and ∆dnaA CM735 cells (E) as mentioned in the legend for Figure 1. The genotype of each strain and the standard deviations are as indicated; **p < 0.01, *p < 0.05, ns, not significant, one‐way analysis of variance.

Figure 3

The DnaA protein binds to HisL‐SL RNA and the absence of rDnaA boxes 1, 2, and 3 compromises DnaA binding and subsequently also transcription attenuation. (A, B) The DnaA protein protects the HisL‐SL RNA and tRNA^His. (A) The gapA gene and the hisL‐SL fragment were PCR‐amplified using cDNA reverse‐transcribed from DnaA‐RIP RNA as a template and detected on 1% agarose gels. Lanes 1 and 4 were loaded with PCR fragments from total RNA as positive controls; lanes 3 and 6 were loaded with PCR fragments from immunoprecipitated RNA (immno–RNA–DnaA) without RNase A treatment; and lanes 2 and 5 were loaded with PCR fragments from RNA treated with RNase A. The gapA gene was used as a negative control as it does not contain a DnaA box. Fragment sizes are as indicated. (B) tRNA^His was detected in the immunoprecipitation assay as described in the Materials and Methods section. Lanes 1 and 4 were loaded with PCR fragments from total RNA; lanes 2 and 5 were loaded with PCR fragments from immunoprecipitated RNA (immno–RNA–DnaA) without RNase A treatment; and lanes 3 and 6 were loaded with PCR fragments from RNA treated with RNase A. The DL2000 DNA ladder was used as the size marker (M); the fragment sizes are as indicated in the figure. (C, D) Mutations in the rDnaA boxes of HisL‐SL RNA affect its binding affinity for DnaA and, consequently, transcription attenuation. (C) Deleted nucleotides and point mutations and their positions in the rDnaA boxes in the phisLp‐hisL‐SL plasmid are as shown. The sequences of rDnaA boxes 1–4 are boxed; mismatches compared with the consensus sequence are in red. (D) Complementary DNA from exponentially growing MC4100 ΔhisL‐SL dnaA‐flag cells carrying the phisLp‐hisL‐SL, pU49G, pG127A, pU128C, pG129A, p∆48‐54, p∆128‐134, or p∆187‐195 plasmid was obtained as described in the Materials and Methods section. The relative binding affinity of DnaA for HisL‐SL RNA was determined as enrichment of the DnaA‐protected RNA molecule by real‐time quantitative‐PCR (RT‐qPCR), as mentioned in the Materials and Methods section. The values were normalized to that of the rplO reference gene and the binding affinity of DnaA for the mutated HisL‐SL RNA was calculated relative to that of wild‐type HisL‐SL RNA on the phisLp‐hisL‐SL plasmid (red bars). Values are averages with the standard error of three individual experiments with two technical replicates. β‐Galactosidase activity from the mutated plasmids in MC4100 cells was measured relative to that in the phisLp‐hisL‐SL plasmid (blue bars). Values are the averages with the standard deviation of three individual experiments; ***p < 0.001, **p < 0.01, *p < 0.05, one‐way analysis of variance (ANOVA).

Figure 4

Excess tRNA^His and the absence of the translating ribosome strengthen transcription attenuation of the his operon. (A) Excess tRNA^His strengthens his operon transcription attenuation. β‐Galactosidase activity in exponentially growing wild‐type (wt) and derived MOR1378 cells was measured in the presence or absence of excess tRNA^His (the phisR plasmid, a tRNA^His overproducer) as mentioned in the legend for Figure 1. β‐Galactosidase activity was calculated relative to wt MOR1378 cells expressing the control vector (pACYC177). The values are averages with the standard deviation of three individual experiments; ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05, ns, not significant, one‐way analysis of variance (ANOVA). (B, C) The absence of the translating ribosome also strengthens translation attenuation of the his operon. (B) Mutations in the phisLp‐hisL‐SL plasmid were introduced by site‐directed mutagenesis, yielding plasmids phisL‐SD_mut, phisL‐TTG, or phisL‐TAA. (C) Relative β‐galactosidase activity from these plasmids was measured in exponentially growing wt and dnaA _A345S MC4100 cells as mentioned in the legend for Figure 1. The values are averages with the standard deviation of three individual experiments; ****p <0.0001, ***p < 0.001, **p < 0.01, *p < 0.05, one‐way ANOVA. (D) A model of the DnaA‐dependent riboswitch for transcription attenuation of the his operon. The translating ribosome (green) is arrested at DnaA (pink)‐bound rDnaA box 1/2, allowing the formation of the AB, CD, and EF (attenuator) stem‐loops, as indicated, resulting in transcription attenuation. The binding of DnaA to rDnaA box 3 prevents the formation of the alternative DE stem‐loop structure. tRNA^His (blue) forms part of the DnaA/rDnaA box complex. When the DnaA protein is released from rDnaA boxes 1, 2, and 3, the translating ribosome covers the A region of HisL‐SL RNA while translating the hisL mRNA, allowing the formation of the BC and DE stem‐loop structures; consequently, the transcription machinery reads through the downstream structural genes. Arrows indicate the orientation of transcription.