A versatile regulatory toolkit of arabinose-inducible artificial transcription factors for Enterobacteriaceae

The Gram-negative bacteria Salmonella enterica and Escherichia coli are important model organisms, powerful prokaryotic expression platforms for biotechnological applications, and pathogenic strains constitute major public health threats. To facilitate new approaches for research and biotechnological applications, we here develop a set of arabinose-inducible artificial transcription factors (ATFs) using CRISPR/dCas9 and Arabidopsis-derived DNA-binding proteins to control gene expression in E. coli and Salmonella over a wide inducer concentration range. The transcriptional output of the different ATFs, in particular when expressed in Salmonella rewired for arabinose catabolism, varies over a wide spectrum (up to 35-fold gene activation). As a proof-of-concept, we use the developed ATFs to engineer a Salmonella two-input biosensor strain, SALSOR 0.2 (SALmonella biosenSOR 0.2), which detects and quantifies alkaloid drugs through a measurable fluorescent output. Moreover, we use plant-derived ATFs to regulate β-carotene biosynthesis in E. coli, resulting in ~2.1-fold higher β-carotene production compared to expression of the biosynthesis pathway using a strong constitutive promoter.

Grey, non-inducing medium; green, inducing medium. M9 minimal medium (supplemented with 0.4% glycerol and 0.1% casamino acids) was used, and for induction, 0.05% arabinose was added to the medium. Data are expressed as the mean ± SD of the MFI obtained from three biological replicates. Two-sided t-test was performed. ns indicates no statistically significant difference from the T1, using a two-sided t-test. Abbreviations: a. u., arbitrary units; mRFP1, monomeric red fluorescent protein; MFI, mean fluorescent intensity. The full data are shown in Supplementary Data 12. ANAC102 and GRF9-derived ATFs were tested against the two (2x) and four (4x) copies of its binding site driving sfGFP reporter expression. Fluorescence output is measured in the absence (grey) and presence of 0.05% arabinose added to LB medium (light green) after 4h (induction).
PprsM, positive control. The MFI of each sample was calculated via FlowJo. Data are expressed as the mean ± SD of the MFI obtained from three independent colonies. Asterisks indicate a statistically significant difference from the non-inducing medium (two-sided t-test; * * * p ≤ 0.001).
Abbreviations Representative cultures of the constructed -carotene producing strains were shown on the left.
The -carotene absorbance at 450 nm was measured using the method reported by Lian et al 3 , and was divided to cdw. The data were normalized to that of the medium. Values represent the mean ± SD of three independent colonies in the presence of 0.2% arabinose. E. coli strains containing pK2151200 (WT/pK2151200) and pK2151201 (WT/pK2151201) were used as controls. Asterisks indicate a statistically significant difference (t-test; * * * p ≤ 0.001). "x" inside the columns represents the fold induction compared to pK2151201. Abbreviations: ATF, artificial transcription factor; BS, binding site; cdw, cell dry weight; JUB1, NAC TF J UNGBRUNNEN1; WT, wild-type. The full data are shown in Supplementary Data 20.

Construction of plasmids and strains
Construction of S. Typhimurium araBAD background strains EM12441 (ΔaraBAD::sfGFP): sfGFP together with a downstream terminator and 40-bp overhangs homologous to the regions up-and downstream of the araBAD operon was PCRamplified from the plasmid pEM8317 (lab collection) using primers DaraBAD-sfGFP-rev and DaraBAD-sfGFP-fv. The PCR product was introduced by electroporation into a (S. Typhimurium LT2) strain with a tetRA cassette replacing the araBAD operon and harboring the λ-RED-helper plasmid pKD46 (TH6706) 4 to facilitate λ red-mediated replacement of tetRA with sfGFP.
Successful recombinants were selected for tetracycline-sensitivity (TcS) using TcS plates 5 .
Subsequently, it was cloned into a PCR-amplified 1869 bp fragment of pUC19 (NEB, primer pair PUC19-fv/PUC19-rv) using the NEBuilder HiFi DNA assembly strategy to generate plasmid p106.
Next, the J106-targeting crRNA region of plasmid p106 was replaced with crRNA to target the J105 motif within the J1 region of the synthetic promoter 1 . To this end, the "p105" fragment was PCR-amplified (primer pair crRNA-J105-fv/crRNA-J105-rev on p106) and followed by NEBuilder HiFi DNA assembly-mediated circularisation of the PCR product, the plasmid p105 was generated. Subsequently, p105 was digested with AcII and AatII to remove a 690 bp fragment.
coli_pJUB1DBD" were transformed with pJUB2X. The positive clones were grown on LB supplemented with Kanamycin (25 g/ml) and Ampicillin (50 g/ml).