Engineering an inducible gene expression system for Bacillus subtilis from a strong constitutive promoter and a theophylline-activated synthetic riboswitch

Abstract

Background: Synthetic riboswitches have been increasingly used to control and tune gene expression in diverse organisms. Although a set of theophylline-responsive riboswitches have been developed for bacteria, fully functional expression elements mediated by synthetic riboswitches in Bacillus subtilis are rarely used because of the host-dependent compatibility between the promoters and riboswitches.

Results: A novel genetic element composed of the promoter P43 and a theophylline-riboswitch was developed and characterized in B. subtilis. When combined with a P43 promoter (P43'-riboE1), the theophylline-riboswitch successfully switched the constitutive expression pattern of P43 to an induced pattern. The expression mediated by the novel element could be activated at the translational level by theophylline with a relatively high induction ratio. The induction ratios for P43'-riboE1 by 4-mM theophylline were elevated during the induction period. The level of induced expression was dependent on the theophylline dose. Correspondingly, the induction ratios gradually increased in parallel with the elevated dose of theophylline. Importantly, the induced expression level was higher than three other strong constitutive promoters including P_srfA, P_aprE, and the native P43. It was found that the distance between the SD sequence within the expression element and the start codon significantly influenced both the level of induced expression and the induction ratio. A 9-bp spacer was suitable for producing desirable expression level and induction ratio. Longer spacer reduced the activation efficiency. Importantly, the system successfully overexpressed β-glucuronidase at equal levels, and induction ratio was similar to that of GFP.

Conclusion: The constructed theophylline-inducible gene expression system has broad compatibility and robustness, which has great potential in over-production of pharmaceutical and industrial proteins and utilization in building more complex gene circuits.

Keywords: Bacillus subtilis; Controllable gene expression; RNA switch; Synthetic riboswitches; Theophylline.

Fig. 1

The combination of the novel theophylline-dependent riboswitch E1 and the P43 promoter in Bacillus subtilis. a Diagrams of the expression cassette driven by the native P43 promoter and the novel theophylline-dependent expression element (P43′-riboE1). The green fluorescent protein is represented by the green arrows, and the transcription start site is marked as TSS. The promoter fragment of the novel expression element, P43′, was produced by deletion of the entire sequence downstream of the native P43. The synthetic riboswitch E1 was placed immediately downstream of P43′, resulting in the novel expression element P43′-riboE1. The riboswitch was derived from the riboswitch E (previously reported by Topp et al.), in which a single A was inserted before the 5′ terminus of the original sequence. In the diagram, the inserted nucleotide is coloured red. The dash line denotes the difference between native P43 and the synthetic riboswitch E1 (riboE1). b The schematic diagram of mechanism of the novel genetic element. Theophylline is indicted in blue. c The growth curves of B. subtilis 168 harbouring pP43-gfp (BSG43) and pBSG11 BSG11). The recombinant B. subtilis strains were inoculated by pre-cultures with the initial OD₆₀₀ of 0.05. The cultures were sampled periodically to measure the cell density until the cell density began to decrease at 27 h. d Expression levels of GFP (y-axis) against time (x-axis) during the culture period. The dashed line denotes the induction by 4 mM theophylline. The GFP fluorescence was measured in triplicates and the data were shown in mean ± SD

Fig. 2

Characterization of induction of GFP from pBSG11 harbouring P43′-riboE1 element after induction by 4-mM theophylline. GFP induction is denoted by green histograms at different time points. The solid circles within in the histograms represent the level of induced GFP expression. Similarly, the open circles within the histograms are the corresponding basal expression level. The activation ratio was obtained by dividing the induced expression level by the corresponding basal level

Fig. 3

The activation of GFP expression by the P43′-riboE1 element displayed dose-dependency. a The expression levels of GFP induced by 1, 2, 4, or 8 mM theophylline were determined 24 h after induction, and were linearly correlated. The experiment was repeated independently in triplicates. b The activation ratios were estimated after induction of 1, 2, 4, or 8 mM theophylline. The ratios were elevated in response to the increased theophylline levels. The highest activation ratio was achieved by 8 mM theophylline, which caused an approximately sevenfolds increase. c SDS-PAGE analysis of with increasing amounts of theophylline. The black solid arrow indicates the GFP band. Twenty micrograms of total protein were loaded for each sample. The GFP fluorescence was measured in triplicates and the data were shown in mean ± SD

Fig. 4

Comparison of the induced expression level with P43′-riboE1 and three strong promoters from B. subtilis. a SDS-PAGE analysis of GFP controlled by the constitutive promoters P_srfA, P_aprE, P43 and by the theophylline-induced element P43′-riboE1. The BSG11 strain was activated by 8-mM theophylline for 24 h prior to sampling for SDS-PAGE. b Fluorescence intensity representing the relative expression level was driven by three constitutive promoters as well as by the P43′-riboE1 element after induction by 8-mM theophylline for total 31 and 24-h culture periods, respectively. The GFP fluorescence was measured in triplicates and the data were shown in mean ± SD

Fig. 5

Determination of the effect of length of spacer on the induced level of GFP expression. a Schematic diagram displays the gene structure of riboE1, and riboE1-15 harbouring 15-bp spacer between SD sequence and start codon was produced by insertion of PstI restriction site immediately downstream of the SD sequence. The SD sequences within riboE1 and the modified riboE1-15 are in the azure box. b SDS-PAGE analysis was carried out to detect the expression level of GFP driven by P43′-riboE1 and P43′-riboE1-15 after induction with 8 mM theophylline for both 12 and 24 h. The strains harbouring recombinant plasmids treated with 4% DMSO are designated as 0 mM (controls). Protein extracts from Bacillus subtilis 168, which were collected at 19 and 31-h culture (equal to the induced groups with 12 and 24-h induction, respectively), are used as the negative controls. The GFP stands for the purified GFP protein, which is served as the specific marker to indicate the corresponding position of heterologous GFP. c Measurement of the relative fluorescent units of GFP corresponding to the SDS-PAGE in b

Fig. 6

Validation of the theophylline-responsive gene expression system with gus reporter gene. a SDS-PAGE analysis showed the expression level of GUS in the absence and presence of theophylline. The induced expression level of GUS in the recombinant strain, BSGgus, was determined by treatment with 8 mM theophylline for 20 h culture, and the culture treated with 4% DMSO for the same time after induction was designated as controls (0 mM). The solid triangle pointed the overproduced GUS. Protein extracts from B. subtilis 168 sampled simultaneously with the induced recombinant strains was used to be the negative control. GUS, denoted the purified GUS protein, was used to be a specific marker to indicate the bands of GUS on SDS-PAGE. b Enzymatic assay for GUS in BSGgus after induction by 8 mM theophylline for 16 and 20 h. The induced activities of GUS and the levels for leakage are shown by solid and open circles, respectively. The histograms represent the corresponding induction ratios. The enzymatic activity assay was performed in triplicates and the data are presented in mean ± SD