A sigma factor toolbox for orthogonal gene expression in Escherichia coli

Abstract

Synthetic genetic sensors and circuits enable programmable control over timing and conditions of gene expression and, as a result, are increasingly incorporated into the control of complex and multi-gene pathways. Size and complexity of genetic circuits are growing, but stay limited by a shortage of regulatory parts that can be used without interference. Therefore, orthogonal expression and regulation systems are needed to minimize undesired crosstalk and allow for dynamic control of separate modules. This work presents a set of orthogonal expression systems for use in Escherichia coli based on heterologous sigma factors from Bacillus subtilis that recognize specific promoter sequences. Up to four of the analyzed sigma factors can be combined to function orthogonally between each other and toward the host. Additionally, the toolbox is expanded by creating promoter libraries for three sigma factors without loss of their orthogonal nature. As this set covers a wide range of transcription initiation frequencies, it enables tuning of multiple outputs of the circuit in response to different sensory signals in an orthogonal manner. This sigma factor toolbox constitutes an interesting expansion of the synthetic biology toolbox and may contribute to the assembly of more complex synthetic genetic systems in the future.

Figure 1.

The bacterial sigma factor is responsible for promoter selectivity through recognition of specific DNA sequences in the promoter region. (A) Sigma factors of the sig70-family are divided in different groups based on their protein domain composition, consisting of two to four domains (42). Subregions 4.2 and 2.4 are of utmost importance in contacting the −35 and −10 promoter elements, respectively. Subregion 3.0 can contact the extended −10 element. (NCR = non-coding region). (B) List of different sigma factors, relevant for this work, originating from Escherichia coli and Bacillus subtilis with their consensus promoter sequences, design (group) and function in the cell.

Figure 2.

Functionality and orthogonality of heterologous sigma factors and promoters in Escherichia coli. (A) Schematic view of core RNA polymerase consisting of 5 subunits (α₂ββ’ω)—given in different gray shades) that can bind the housekeeping sigma factor σ⁷⁰ (given in blue) to form a holoenzyme able to transcribe from a σ⁷⁰-specific promoter (also given in blue) in an E. coli cell. (B) Concept overview: an ideal orthogonal expression system combining several heterologous Bacillus subtilis sigma factors (namely sigma factor B, F and W given in yellow, red and green respectively) and cognate promoters introduced in the E. coli host. All sigma factors are able to bind the host core RNAP and form a holoenzyme that is able to recognize/transcribe specifically from its cognate promoter and not from the other promoters to yield a perfect orthogonal system without cross-talk. (C) Cross reactivity of sigma factor–promoter pairs measured in both exponential and stationary growth phase. Colors indicate activity (relative FL/OD) defined as the activity of the promoter-sigma pair divided by the activity of the reference (black square), a strong σ⁷⁰ promoter (P_high (12)) in wild-type E. coli MG1655. A gray square encloses the ECF sigma factors (M, W, X) in which some cross-talk is observed. Functionality of heterologous sigma factors from B. subtilis in E. coli strains bearing a cognate plasmid-borne promoter–reporter gene construct is shown by green rectangles. Data are the means of at least three biological replicates.

Figure 3.

Orthogonality of Bacillus subtilis promoters to native Escherichia coli sigma factors. Cross reactivity of sigma factor–promoter pairs measured in both exponential and stationary phase; and with and without addition of IPTG to induce expression of the E. coli sigma factors. Colors indicate activity (relative FL/OD) defined as the activity of the promoter–sigma pair divided by the activity of the reference, a σ⁷⁰ promoter (P_high (12)) in wild-type E. coli MG1655. B. subtilis promoters (P_M2 and P_X3) that were non-orthogonal toward E. coli (Figure 2C), and promoters with low TIF (<0.10) with their cognate sigma factor (Figure 2C, P_B3 and P_X2) were dismissed from this analysis.

Figure 4.

Promoter libraries setup and characterization. (A) Original promoters P_B2, P_F3 and P_W2 were selected to create three types of libraries by randomizing parts of the spacer sequence between the −10 and −35 conserved regions of the promoters. mKate2 expression is used as reporter to characterize library promoters and constitutively expressed sfGFP to correct for extrinsic factors. Vectors containing original promoters and their libraries were transformed in strains containing each of the cognate and non-cognate sigma factors and wild-type Escherichia coli. (B) All created strains were subjected to flow cytometry analysis. The presented data have been processed by asinh transformation of the mKate2/sfGFP ratio and corrected with the negative control (Supplementary Data Materials and Methods). Population means are displayed for non-cognate promoter–sigma factor pairs and whole library TIF distributions for cognate pairs. (C) Additionally, created σ⁷⁰ promoter libraries act as a reference for native E. coli promoter library distributions.

Figure 5.

Representative set of individual promoters for libraries L_B2, L_F3 and L_W2. A selection of 9 or 10 promoters was made for each library. The three sets were analyzed in presence of their cognate sigma factor, a non-cognate sigma factor or in absence of a heterologous sigma factor (wild-type MG1655). (A) For each promoter–sigma factor couple, the media blank corrected mKate to sfGFP ratio is calculated (Fluorescence_C). Depicted values are relative to the respective original promoter as reference (P_B2, P_F3 and P_W2, with cognate sigma factor, put equal to 1). (B) Activation ratios are calculated as the ratio of the activity of a promoter in presence of a heterologous sigma factor to the activity of that promoter in absence of a heterologous sigma factor (wild-type MG1655).