Systematic modulation of bacterial resource allocation by perturbing RNA polymerase availability via synthetic transcriptional switches

Abstract

Gene regulation and its interplay with physiological behaviors are the central topics of modern biology. Classical studies on gene regulation focus intensively on specific regulatory mechanisms of transcription. Nevertheless, the genome-wide impact of RNA polymerase (RNAP) availability on gene expression remains poorly understood. Here we developed two synthetic transcriptional switches to systematically titrate the expression of either ${\sigma ^A}$ (SigA, housekeeping sigma factor) or RpoBC (core enzyme) in Bacillus subtilis. Both systems effectively modulated cell growth, but with fundamentally distinct mechanisms. SigA limitation triggered significant resource reallocation, redirecting cellular investment from biosynthetic pathways to alternative cellular pathways, which could further facilitate the engineering of dynamic growth-bioproduction switch. In contrast, RpoBC depletion caused only weak changes of gene expression but induced ribosomal inactivation through blocking translation initiation. Notably, RpoBC depletion induced DNA damage response and increased the DNA damage sensitivity of bacteria, suggesting transcription-coupled repair as a critical survival mechanism. Our findings delineate two regulatory paradigms of resource allocation that are associated with the interplay between RNAP availability and bacterial physiological state, "abundance-based" and "activity-based" regulations. The orthogonal transcriptional switches serve as a powerful tool for dissecting the integrative role of RNAP in microbial physiology, offering meaningful implications for both fundamental studies of gene regulation and synthetic biology applications.

Graphical Abstract

Figure 1.

Synthetic transcriptional switches based on perturbation of RNAP availability. (A) Bacterial RNAP consist of core enzyme and sigma factors. Sigma factors contain the housekeeping sigma factor and other alternative factors, such as , , and . (B) Growth rates of B. subtilis cells during titration of the housekeeping sigma factor, . (C) Growth rates of B. subtilis cells during titration of the core enzyme components, RpoBC. (D) Growth rates of wild type, RpoBC- and SigA-titratable strains in LB and glucose + cAA media, respectively. Thirty micromolar IPTG was supplemented into the cultures of RpoBC- and SigA-titratable strains. (E) The quantitative mRNA and protein levels of sigA gene in SigA titratable strain under different inducer levels, measured by transcriptomic and proteomic analysis, respectively. (F) The quantitative mRNA and protein levels of rpoBC genes in RpoBC titratable strain under different inducer levels, measured by transcriptomic and proteomic analysis, respectively. The data of panels (E) and (F) were taken from the transcriptomic and proteomic analysis detailed in Figs 2 and 3 below. Error bars denote the standard deviations of at least three biological replicates.

Figure 2.

Global effect of perturbing RNAP availability on the transcriptome of exponentially growing B. subtilis. Cells were cultured in glucose + cAA medium. (A) Conditions for transcriptomic analysis including SigA- and RpoBC-titratable strains at three inducer levels, which correspond to three different growth rates, designated as “low,” “medium,” and “high” conditions. (B) The volcano analysis of the transcriptome data of B. subtilis during SigA and RpoBC limitations. (C) Pairwise scatter plots of the mRNA mass fractions of individual genes between different conditions, including RpoBC_high versus SigA_high, SigA_high versus SigA_low and RpoBC_high versus RpoBC_low. (D, E) Visualization of cellular transcriptome allocation for SigA- and RpoBC- titratable strains. Note that the term of “mitochondrial biogenesis” was based on KEGG categorization consisting of both prokaryotes and eukaryotes. The readers should thus treat these proteins here as translation factors. (F) Absolute and relative changes (Log₂ fold) in the mRNA fractions of various functional sectors for B. subtilis during SigA and RpoBC limitations, including “SigA_low” versus “SigA_high” conditions and “RpoBC_low” versus “RpoBC_high” conditions. Unlike panel (B), which shows the log₂fold changes of individual mRNA; log₂fold change of functional sectors is less pronounced here for two reasons: (i) the effect of one substantially changed genes in one functional group is often masked by other genes in the same group that undergo weaker or no changes. (ii) many functional groups such as ribosomes, translation-affiliated proteins, tRNA charging, glycolysis, and TCA cycle are highly abundant housekeeping genes groups and are thus difficult to undergo a large fold change as a whole. However, the absolute changes in the cellular abundances of those various housekeeping groups, such as ribosomes will have a profound effect on the bacterial resource allocation and growth physiology. (G–I) Scatter plots showing the correlation of the mRNA mass fractions of various functional sectors between different conditions. (J) The mRNA mass fraction of the total biosynthetic sector in B. subtilis during SigA and RpoBC limitations. (K) The mRNA mass fraction of various categories of genes involved in stress responsive processes in B. subtilis during SigA and RpoBC limitations. The sporulation pathway mainly belongs to SigE & SigF regulon while the general stress response belongs to SigB regulon. Error bars denote the standard deviations of three biological replicates. The transcriptome data in this work contains three biological replicates for each condition, except for the “medium” condition of RpoBC titration, which contains two biological replicates.

Figure 3.

Global effect of perturbing RNAP availability on the proteome of exponentially growing B. subtilis. Conditions are the same as transcriptomic analysis in Fig. 2, except that RpoBC_medium condition was not included in the proteomic study. (A–C) Scatter plots showing the correlation of the mass fraction of individual proteins between different conditions, including RpoBC_high versus SigA_high, SigA_high versus SigA_low, and RpoBC_high versus RpoBC_low. (D–F) Scatter plots showing the correlation of the proteome fractions of various functional sectors between different conditions. (G) Absolute changes in the proteome fractions of various functional sectors for B. subtilis during SigA and RpoBC limitations, including SigA_low condition versus SigA_high condition and RpoBC_low condition versus RpoBC_high condition. (H) The proteome fractions of the total biosynthetic sector in B. subtilis during SigA and RpoBC limitations. (I) The proteome fractions of various biosynthetic sectors in B. subtilis during SigA and RpoBC limitations. (J) The proteome fractions of various categories of genes involved in stress responsive processes in B. subtilis during SigA and RpoBC limitations. The proteome data in this work contains two biological replicates.

Figure 4.

Correlation between transcriptome and proteome for exponentially growing B. subtilis. (A) Correlation between mRNA fraction and proteome fraction for individual genes in SigA_high condition. (B) Correlation between mRNA fraction and proteome fraction for individual genes in SigA_low condition. (C) Correlation between mRNA fraction and proteome fraction for individual genes in RpoBC_high condition. (D) Correlation between mRNA and proteome fractions for individual genes in RpoBC_low condition. (E) Correlation between the relative changes of mRNA and protein levels for SigA_low condition versus SigA_high condition. (F) Correlation between the relative changes of mRNA and protein levels for RpoBC_low condition versus RpoBC_high condition.

Figure 5.

The effect of RNAP availability on the status of ribosomes in B. subtilis. (A–B) The mRNA and protein fractions of various core enzyme components of RNAP and sigma factors during SigA limitation. (C–D) The mRNA and protein fractions of various core enzyme components of RNAP and sigma factors during RpoBC limitation. (E) Schematics showing the status of RNAP inside B. subtilis during SigA and RpoBC limitations. At normal condition, cellular RNAP are associated with either or . Under SigA limitation, the core enzyme level is sustained and the fraction of -associated RNAP increase substantially due to shortage of . Instead, under RpoBC limitation, cells undergo a global shortage of RNAP. (F–G) The mRNA and protein levels of ribosomal proteins (r-protein) during SigA and RpoBC limitations. (H) Polysome profiling of B. subtilis under SigA- and RpoBC-limited conditions. (I) Schematics showing that limitations of SigA and RpoBC reduce the level or activity of ribosomes in B. subtilis, respectively.

Figure 6.

Growth-bioproduction switch based on SigA titratable systems. (A) Schematic illustration showing that SigA limitation triggers a resource re-allocation from biosynthetic pathways to other pathways mediated by alternative sigma factors, especially the SigO-regulated genes such as oxdC. (B) The protein and mRNA levels of oxdC and sigO genes under SigA limitation. (C) The expression levels of GFP driven by the oxdC promoter under SigA limitation. Visualization of GFP fluorescence of cultures under various IPTG concentrations was performed by a blue light illuminator. (D) The expression levels of LacZ driven by the oxdC promoter under SigA limitation. (E) The activities of several strong IPTG-inducible promoters in wild-type B. subtilis strain measured by LacZ reporter assay. (F) The maximal activities of oxdC promoter (panel D) and the three IPTG-inducible promoters (panel E). Error bars denote the standard deviations of three biological replicates.

Figure 7.

RpoBC limitation triggers DNA damage response of B. subtilis. (A) Relative mRNA levels of some typical DNA damage responsive genes during SigA and RpoBC limitations. (B) The total mass fraction of mRNA for genes shown in panel (A). (C–D) mRNA levels of recA, uvrB, and urvA genes during RpoBC and SigA limitations. (E, F) Flow cytometry analysis of the PrecA-gfp reporter in B. subtilis during RpoBC and SigA limitations. (G,H) Flow cytometry analysis of the PuvrBA-gfp reporter in B. subtilis during RpoBC and SigA limitations.

Figure 8.

RpoBC limitation sensitizes B. subtilis to DNA damage. (A) Either ultraviolet (UV) or mitomycin C (MMC) treatment was applied to impose DNA damage on B. subtilis. (B,C) The effect of RpoBC and SigA limitations on the viability of B. subtilis cells upon 200 ng/ml MMC and UV treatment (250 μW/cm²). (D,E) The effect of RpoBC limitation on the viability of uvrA-null B. subtilis cells upon MMC or UV treatment. Since uvrA-null cells are highly sensitive to UV treatment, cell viability was measured only at 3 min after the initiation of UV irradiation. Error bars denote the standard deviations of three biological replicates. From panels (B–D), the error bars have been shown for all the data points, most of them are within the size of the symbols in log scale plot.

Figure 9.

Illustration of the global effect of RNAP availability on gene expression and cellular economy of bacteria. To achieve fast growth, cells require balanced investment on metabolic proteins and ribosomes to support a high cellular biosynthetic flux. In the condition of SigA limitation (upper), the re-distribution of RNAP leads to proteome re-allocation from biosynthetic pathways to proteins of alternative functions such as stress response, resulting in a lower biosynthetic flux and growth rate. Conversely, under RpoBC limitation (bottom), the cellular abundances of various biosynthetic pathways (including metabolic proteins and ribosomes) change very weakly. Nevertheless, the reduction of core enzyme level results in the inactivation of a substantial fraction of ribosomes (shown in gray color) due to failure in translation initiation, which also results in lower biosynthetic flux and growth rate. Therefore, although the level of total biosynthetic pathway changes very weakly during RpoBC limitation, it supports a lower biosynthetic flux and is thus in a low activity state compared with normal condition (middle). Collectively, the effect of RpoBC limitation and SigA limitation are in two distinct paradigms, “activity-based regulation” and “abundance-based regulation,” respectively.