Roles of Transcriptional and Translational Control Mechanisms in Regulation of Ribosomal Protein Synthesis in Escherichia coli

Abstract

Bacterial ribosome biogenesis is tightly regulated to match nutritional conditions and to prevent formation of defective ribosomal particles. In Escherichia coli, most ribosomal protein (r-protein) synthesis is coordinated with rRNA synthesis by a translational feedback mechanism: when r-proteins exceed rRNAs, specific r-proteins bind to their own mRNAs and inhibit expression of the operon. It was recently discovered that the second messenger nucleotide guanosine tetra and pentaphosphate (ppGpp), which directly regulates rRNA promoters, is also capable of regulating many r-protein promoters. To examine the relative contributions of the translational and transcriptional control mechanisms to the regulation of r-protein synthesis, we devised a reporter system that enabled us to genetically separate the cis-acting sequences responsible for the two mechanisms and to quantify their relative contributions to regulation under the same conditions. We show that the synthesis of r-proteins from the S20 and S10 operons is regulated by ppGpp following shifts in nutritional conditions, but most of the effect of ppGpp required the 5' region of the r-protein mRNA containing the target site for translational feedback regulation and not the promoter. These results suggest that most regulation of the S20 and S10 operons by ppGpp following nutritional shifts is indirect and occurs in response to changes in rRNA synthesis. In contrast, we found that the promoters for the S20 operon were regulated during outgrowth, likely in response to increasing nucleoside triphosphate (NTP) levels. Thus, r-protein synthesis is dynamic, with different mechanisms acting at different times.IMPORTANCE Bacterial cells have evolved complex and seemingly redundant strategies to regulate many high-energy-consuming processes. In E. coli, synthesis of ribosomal components is tightly regulated with respect to nutritional conditions by mechanisms that act at both the transcription and translation steps. In this work, we conclude that NTP and ppGpp concentrations can regulate synthesis of ribosomal proteins, but most of the effect of ppGpp is indirect as a consequence of translational feedback in response to changes in rRNA levels. Our results illustrate how effects of seemingly redundant regulatory mechanisms can be separated in time and that even when multiple mechanisms act concurrently their contributions are not necessarily equivalent.

Keywords: autogenous control; ppGpp; regulation of promoter activity; ribosome synthesis; stringent response; translational feedback regulation.

FIG 1

Reporters for measuring regulation of S20 r-protein synthesis. (A) A diagram of the regulatory region in the S20 r-protein operon is shown above schematics of the reporter constructs. Reporters 1 to 8 contain a fusion of the first 54 nucleotides (nt) of the rpsT ORF, coding for the first 18 amino acids (aa) of S20 (S20 [1–18]), in frame with an ORF encoding an SBP-lacZ fusion (S20 [1–18]-SBP-lacZ). We chose to fuse 54 nt (18 codons) of the rpsT ORF, instead of 18 nt (6 codons) like previous rpsT leader-lacZ fusions (25), to allow formation of the two hairpins that had been previously characterized in this region because they may have a role in translational feedback (26). The SBP (streptavidin-binding peptide)-lacZ fusion is described in the supplemental material. Constructs 1, 3, and 9 contain the wild-type rpsT leader sequence starting at the rpsT P2 transcription start site. Constructs 2, 4, and 6 to 8 contain a trp-lacZ leader instead of the rpsT leader, as well as an S20 (1–18)-SBP-lacZ fusion in which mutations were introduced into the S20 translation initiation region to eliminate regulation through translational feedback (MUT S20 [1–18]; details in the supplemental material). Constructs 6, 7, and 8 contain the rpsT P1, rpsT P2, and rrnB P1 promoters, respectively, fused to the trp-lacZ leader. Constructs 3 to 5, 9, and 10 contained the lacUV5 promoter (48), a control promoter that is not regulated by ppGpp (7). Construct 5 contains the lacUV5 promoter fused to the rpsT mRNA leader that corresponds to the leader transcribed from rpsT P1. Red asterisks represent 7 point mutations in the −35 and −10 elements of the rpsT P2 promoter designed to eliminate rpsT P2 activity (details in the supplemental material). The PlacUV5 S20 (1–18)-lacZ and PlacUV5_lacZ reporters (constructs 9 and 10, respectively) contain the WT lacZ ORF and were used for size comparison to the S20 (1–18)-SBP-lacZ reporter protein. Numbers below the lines refer to positions relative to the rpsT P1 transcription start site. (B) Representative protein gels showing only the reporter protein products expressed from the constructs in panel A. A construct encoding SBP-lacZ but without a promoter was used to measure background levels of reporter protein expression (lane 11). The background strain used for reporter constructs, VH1000, is the “No reporter” control (lane 12).

FIG 2

S20 synthesis during induction of ppGpp synthesis. (A) ppGpp levels were measured as described in Materials and Methods. Error bars represent the range (n = 2). Data are from chromatogram shown in Fig. 5B, lanes 1 to 5. Relative increases in ppGpp concentration (y axis) are likely an underestimate because the signal at time zero is close to background. (B) The reporter protein band from the strains containing the rpsT P1+P2_rpsT leader and PlacUV5_lacZ leader constructs (constructs 1 and 4 in Fig. 1A), and either the active or inactive RelA' plasmid, was quantified by phosphorimaging, corrected for background, and normalized to the reporter band at time zero for each strain. Error bars represent the range (n = 2). Numbers in parenthesis in the labels refer to the numbers of the reporter constructs shown in Fig. 1A. (C) Representative gel illustrating ³⁵S-pulse-labeled protein profiles following ppGpp induction as described in the text and Materials and Methods. Arrow shows the position of the reporter band.

FIG 3

Regulation of S20 synthesis by ppGpp requires the rpsT leader. (A) ³⁵S reporter protein bands from a representative gel are shown from rpsT reporters following induction of ppGpp. Relative synthesis plots at the right are from multiple experiments. (B) Reporter protein bands and quantification from multiple experiments as in panel A during transition into stationary phase. Cells were grown in MOPS medium (44) supplemented with 19 aa (no methionine) and glucose as the carbon source. Pulse-labeled samples were taken every hour starting at an OD₆₀₀ of 0.2 (see Materials and Methods). (C) Same as panel A but with the long rpsT leader reporter (Fig. 1A, construct 5). Error bars in each graph indicate the range (n = 2). Numbers in parenthesis in the labels refer to the numbers of the reporter constructs shown in Fig. 1A.

FIG 4

Regulation of rpsT promoters. (A) ³⁵S reporter protein bands from representative gel and quantification from multiple experiments showing changes in reporter synthesis from transcriptional fusions (Fig. 1A, constructs 4 and 6 to 8) following induction of ppGpp in vivo. Promoter endpoints of the transcriptional fusions are PlacUV5 (−59/+1), rpsT P1 (−100/+2), rpsT P2 (−90/+2), and rrnB P1 (−61/+1), where +1 is the transcription start site. Error bars represent the range (n = 2). (B) Promoter activity during outgrowth was measured by RT-qPCR and was initiated by a 1:10 dilution of a stationary-phase culture into fresh medium. Error bars represent the range (n = 2). (C) Promoter activity measured by in vitro transcription at different CTP concentrations. Promoter fragments with the indicated endpoints relative to the transcription start site (rpsT P2, −89/+50; rrnB P1 + 1C, −66/+9; and PlacUV5, −59/+38) were cloned into pRLG770, resulting in plasmids pRLG9237, pRLG3735, and pRLG2222, respectively (for details, see Table S1 in the supplemental material). Error bars represent the SD (n = 3). Numbers in parenthesis in the labels refer to the numbers of the reporter constructs shown in Fig. 1A.

FIG 5

DksA is required for ppGpp-dependent regulation of S20 synthesis. (A) To the left is a representative gel showing ³⁵S reporter bands from the PlacUV5_rpsT leader and PlacUV5_lacZ leader constructs at times after ppGpp induction in either a WT or ΔdksA strain background. To the right is the quantification of reporter protein synthesis from multiple experiments (n = 2). Numbers in parenthesis in the labels refer to the numbers of the reporter constructs shown in Fig. 1A. Error bars represent the range (n = 2). (B) Thin-layer chromatogram shows measurement of ppGpp levels following induction of relA' expression from pALS13 in the strains used in panel A (n = 2). ppGpp indicates the region of the plate where ppGpp and pppGpp run together.

FIG 6

Regulation of the S10 operon is independent of the rpsJ promoter. (A) Schematic diagrams of the 5′ region of the S10 operon and the rpsJ reporters. Numbering below diagrams is relative to the PrpsJ transcription start site (49). Reporters containing the rpsJ leader are N-terminal fusions of the rpsJ leader and the first 30 nt of the rpsJ ORF, which code for the first 10 aa of S10 (S10 [1–10]) to the MUT S20 (1–18)-SBP-lacZ reporter (described in the supplemental material). The lacUV5_lacZ leader construct is the same one shown in Fig. 1A. (B) ³⁵S reporter protein bands from representative gel and quantification from multiple experiments following induction of ppGpp. Error bars represent the standard deviation (n = 3).