Identification of 2-keto-3-deoxy-D-xylonate and 2-oxo-4-hydroxybutyrate as natural and artificial effectors of the transcription factors XynR and YjhI, respectively, and application to the development of biosensors for the bioproduction of 2,4-dihydroxybutyric acid in E. coli

Identification of 2-keto-3-deoxy-D-xylonate and 2-oxo-4-hydroxybutyrate as natural and artificial effectors of the transcription factors XynR and YjhI, respectively, and application to the development of biosensors for the bioproduction of 2,4-dihydroxybutyric acid in E. coli | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Identification of 2-keto-3-deoxy-D-xylonate and 2-oxo-4-hydroxybutyrate as natural and artificial effectors of the transcription factors XynR and YjhI, respectively, and application to the development of biosensors for the bioproduction of 2,4-dihydroxybutyric acid in E. coli Thibault MALFOY, Ceren ALKIM, Julie FREDONNET, Juan LAJARIN-HERNANDEZ, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8830407/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 01 May, 2026 Read the published version in Applied Microbiology and Biotechnology → Version 1 posted You are reading this latest preprint version Abstract In E. coli , D-xylonate catabolism is carried out by two distinct operons, namely yagEFG and yjhIHG , present on two distinct cryptic phages in the genome of this bacterium. These two operons are controlled by the transcription factors XynR and YjhI, the former acting as a repressor of yagEFG while the latter acts as an activator of yjhIHG . Although D-xylonate is known to induce these two operons, the effective inducer remains unknown to date. Through the construction of biosensors based on XynR and YjhI using syfp2 as a reporter gene, we showed that the true natural effector of the two D-xylonate catabolic operons is 2-keto-3-deoxy-D-xylonate, which is formed by dehydration of D-xylonate catalyzed by the dehydratases encoded by yagF and yjhG . Building on the finding that these two operons were also upregulated in E. coli strains producing the non-natural platform molecule 2,4-dihydroxybutyric acid, we discovered that both XynR- and YjhI-based biosensors were responsive to the non-natural molecule 2-oxo-4-hydroxybutyric acid, harboring comparable characteristic performances in term of response threshold, sensitivity, cooperativity and dynamic response as the natural effector 2-keto-3-deoxy-D-xylonate. Given that the enzymatic steps involved in the production of this non-natural metabolite from C2 and C5/C6 carbons, catalyzed respectively by threonate dehydratase and homoserine transaminase, constitute bottlenecks in the synthetic pathways for the production of 2,4-dihydroxybutyric acid, we showed that the transcription factors XynR and YjhI can be repurposed as biosensors to select more active variants of these enzymes, thereby improving the production of this platform molecule from carbon sources. biosensor DNA-binding transcription regulator D-xylonate 2-oxo-4-hydroxybutyric acid 2 4-dihydroxybutyric acid genetic circuits synthetic pathways Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 INTRODUCTION The Dahms pathway enables the assimilation of xylose into glycoaldehyde and pyruvate involving four enzymatic steps (Dahms 1974 ). The first reaction in this pathway is the oxidation of xylose to D-xylonate by a xylose dehydrogenase. While this enzyme does not exist in E. coli , the remaining enzymatic steps are encoded by two distinct operons, namely yagEFGH and yjhIGHF that are present on two cryptic phages, CP4-6 and KpL12-prophages, respectively, in the genome of this bacterium. Genetic and biochemical data have shown that yagF and yihG encode a xylonate dehydratase that converts D-xylonate to 2-keto-3-deoxy-D-xylonate (KDX). This enzymatic step is followed by an aldolic cleavage of KDX into glycoaldehyde and pyruvate by a pyruvate-dependent aldolase encoded by yagE and yjhH . On the other hand, D-xylonate uptake is assumed to be taken up by a transporter encoded by yagG and YjhF (reviewed in (Banares et al. 2021 ), see Figure S1 ). It has been reported that these two operons are induced by D-xylonate through the action of a transcription factor apparently specific for each operon. Shimada et al . (Shimada et al. 2017 ) showed that YagI, renamed XynR, which is also encoded by the CP4-6 prophage acts as a repressor of the yag operon, whereas Banares et al. (Banares et al. 2019 ) demonstrated that YjhI, which is present on KpL12-phages-like elements, very likely works as an activator of the yjh operon, which also includes its autoinduction. However, it remained elusive whether D-xylonate is the direct effector of these two transcription factors. Two types of experimental data suggested that the effector of XynR and YjhI is likely not D-xylonate. On the one hand, the concentration of D-xylonate required to release the binding of XynR to DNA or to activate YjhI-dependent genes is in the range of 10 mM to 20 mM (Banares et al. 2019 ; Shimada et al. 2017 ), which is very high with respect to micro to millimolar concentration of most of metabolite effectors (Hanko et al. 2020 ). On the other hand, transcriptomic analyses of E. coli challenged with the non-natural platform molecule 2,4-dihydroxybutyric acid (2,4-DHB) triggered a strong upregulation of the D-xylonate catabolic genes (our unpublished data). Since metabolic pathways designed to produce 2,4-DHB are different to the catabolic pathway of D-xylonate, these data suggested that the effector of XynR and YjhI could be a metabolite derived from D-xylonate and that this metabolite may be close to 2,4-DHB or its proximal intermediate 2-oxo-4-hydroxybutyrate (OHB). To answer these questions, we designed a biosensor consisting of a reporter gene encoding a fluorescent protein dependent on the promoter of a gene under the control of XynR or YjhI. With this tool, and combined with mutants defective for D-xylonate catabolic pathway, we could on the one hand investigate the true effector of these two transcription factors and on the other hand, examine whether these TFs are responsive to 2,4-DHB or an intermediate in the production pathway of this molecule, in order to repurpose them as biosensors tool for improving the flux production of this non-natural platform molecule. The motivation for developing such biosensors sensitive to either 2,4-DHB or its proximal intermediate, 2-oxo-4-hydroxybutyrate (OHB), stems from the still low performance in terms of yield and productivity of the E. coli strains equipped with synthetic pathways designed for the bioproduction of this platform molecule (2,4-DHB) from glucose (Walther et al. 2018 ; Walther et al. 2017 ) or from C1/C2 carbon source (Frazão et al. 2023 ). Enzymes in these non-native pathways have been engineered to catalyse reaction on non-natural substrates, leading to non-natural intermediate OHB, which is produced either from L-homoserine by the action of promiscuous transaminases (Bouzon et al. 2017 ; Walther et al. 2018 ) or from D-threonate by D-threonate dehydratase from Herbaspirillum huttiense . Clearly, these enzymes are rate limiting in the production pathways of 2,4-DHB because they harbor a catalytic efficiency (k cat /K M ) at least 100-fold lower than most of the other enzymes needed for the bioproduction. Thus, directed enzyme evolution is required to optimise their catalytic efficiency, which relies on a high throughput screening method. Transcription factor-based metabolite biosensors would be well appropriate as long as OHB or 2,4-DHB can be recognized as a target of a transcription factor already existing in E. coli . Since nature has developed a wide range of sensors using natural endogenous metabolites that enable bacteria to adapt to their environment, the use of TF-based biosensors becomes challenging when it comes to optimizing a purely synthetic pathway whose intermediates and/or end product are completely unknown to the microorganism. In this work, we demonstrated that the natural effector, and therefore the potential ligand, of XynR and YjhI is the intermediate metabolite KDX. We validated that both transcription factors can be repurposed as biosensors to establish high throughput screening strategy aiming at improving the flux production of the non-natural platform molecule 2,4-DHB. MATERIAL & METHODS Chemicals and reagents All chemicals and solvents were purchased from Sigma-Aldrich unless otherwise stated. The 2,4–dihydroxybutyric acid (racemic form, ammonium salt) was a kind gift from Adisseo and was > 95% pure as determined by HPLC. The 2-oxo-4-hydroxybutyrate (OHB) was synthesized by incubating 125 mM D-homoserine with 4.5 U/mL porcine kidney D-amino acid oxidase, 4000 U/mL beef liver catalase in 100 mM Tris-HCl buffer (pH 8) for at least 2 h at 37°C. These enzymes were purchased from Sigma-Aldrich. The reaction was filtered through AmiconTM Ultra filters (10 kDa threshold, Millipore). OHB was quantified by NMR using a Bruker Avance III HD 800 MHz spectrometer equipped with a cryogenically cooled 5 mm QCI-P (H/P-C/N/D) quadruple resonance probe. The spectra were acquired and reprocessed with Bruker Topspin 4.1 software, and were calibrated for reference to the frequency of TSP (TrimethylSilylPropanoic acid) on the 1H spectrum at 0 ppm. The solution, giving a yield of conversion of D-homoserine ranging from 77 to 88%, was stored at -80°C in small aliquots, which were used only once after thawing. Restriction endonucleases, Phusion polymerase, restriction enzymes, HiT4 DNA ligase, and HiFi DNA assembly were purchased from New England Biolabs and used according to instructions of the manufacturer. Moreover, In-Fusion® HD Cloning Kit was purchased from TaKaRa-Clontech. DNA plasmid isolation was performed using GeneJET Plasmid Miniprep Kit (Thermo Scientific). DNA extraction from agarose gel was carried out using the GeneJET Gel Extraction Kit (Thermo Scientific). DNA sequencing was carried out by Eurofins SAS (Ebersberg, Germany). Strains construction and growth conditions The strain E. coli K-12 substr. MG1655 (ATCC 47076) was used as the parental strain for all constructions in this study. The strain MGΔ7 was constructed by deleting the 6 genes, namely eda, yfaU, garL, yagE, yjjH and dgoA encoding pyruvate-dependent aldolase by the phage transduction method adapted from Miller (Miller 1992 ), whereas mhpE was deleted using CRISPR-Cas9 method developed by Jiang et al (Jiang et al. 2015a ). In transduction procedure, donor strains were taken from the Keio collection (Baba et al. 2006 ), which contains mutants where a single gene is deleted and replaced by a kanamycin resistance cassette flanked by FRT sites. This method allowed the construction of simple, double and etc. mutants, since all target genes were separated by at least 100 kb. Likewise, MG1655 was deleted of its membrane-associated lactate oxidoreductase encoded by lldD , dld and ykfEGF by transduction giving rise to MGΔLO, as described elsewhere (Malfoy et al. 2024 ). Positive clones were selected on LB agar plates containing kanamycin (50 µg.mL − 1 ) and verified by PCR analysis. The Kan cassette was removed from the genome by expressing FLP recombinase from the pCP20 plasmid (Cherepanov and Wackernagel 1995 ) and correct excision of the cassette was verified by PCR using locus specific primers ( Table S1 ). Other strain deletions, such as MG1655 ∆xynR , MG1655 ∆yjhI , MG1655 ∆xynR ∆yjhI etc., as reported in Table 1 were also carried out by P1 phage transduction according to (Miller 1992 ). Strain MGΔ4 defective in the aspartate-homoserine pathway ( ∆thrA , ∆metL , ∆asd, ∆lysC ) was created as reported elsewhere (Malfoy et al. 2024 ). To obtain the mutant strain MGDT which was deleted for four transaminases ( alaC , tyrB , ybdL , ilvE ) and asd encoding the aspartate semi-aldehyde dehydrogenase, the transaminases genes were consecutively deleted in MG1655 by transduction (Thomason et al. 2007 ). Then, asd gene was deleted in the obtained mutant MG1655 ∆alaC::FRT ∆tyrB::FRT ∆ybdL::FRT ∆ilvE::FRT by CRISPR-Cas9 method. This mutant was first transformed with pCas9. Guide RNA was expressed from a pTargetF-asd (Table S1 ). Overnight cultures grown at 30°C were diluted to an OD₆₀₀ of 0.05 in LB containing kanamycin and arabinose (10 mM) and grown to OD₆₀₀ ≈ 0.6. Cells were chilled on ice for 15 min, harvested at 5,000 × g for 5 min at 4°C, washed twice with ice-cold sterile water and once with 10% glycerol, and resuspended in 100 µL 10% glycerol. Competent cells were electroporated according to (Wang et al. 2009 ) with 100 ng pTarget and 400 ng donor DNA (Eurofins Genomics) using a Gene Pulser Xcell (Bio-Rad) at 2.5 kV in a 2-mm cuvette. Cells were recovered in SOC medium for 1–2 h at 30°C and plated on LB agar containing kanamycin and spectinomycin. Genome edits were verified by PCR and sequencing, and pTarget and pCas9 were sequentially cured from positive clones. Table 1 Strains used in this study Strain name Description -genotype source MG1655 F- lambda- ilvG- rfb-50 rph-1 ATCC n°47046 NEB®5α fhuA2 Δ(argF-lacZ)U169 phoA glnV44 Φ80 Δ(lacZ)M15 gyrA96 recA1 relA1 endA1 thi-1 hsdR17 NEB competent E. coli BL21 (DE3) fhuA2 [lon] ompT gal (λ DE3) [dcm] ΔhsdS λ DE3 = λ sBamHIo ΔEcoRI-B int::(lacI::PlacUV5::T7 gene1) i21 Δnin5 NEB BW25113 Δ(araD-araB)567 ΔlacZ4787(::rrnB-3) λ- rph-1 Δ(rhaD-rhaB)568 hsdR514 CGSC#7636 (Baba et al. 2006 ) MGΔ7 MG1655 ΔyagE::FRT ΔyjhH::FRT Δeda::FRT ΔdgoA::FRT ΔyfaU::FRT ΔgarL::FRT ΔmhpE This study MGΔLO MG1655 Δdld::FRT ΔykgEFG::FRT ΔlldD::FRT This study MGDT MG1655 Δybdl ΔtyrB ΔilvE ΔalaC This study MGΔ4 MG1655 ΔlysC ΔthrA Δasd ΔmetL This study MGΔF MG1655 ΔyagF ::FRT This study MGΔG MG1655 ΔyghG ::FRT This study MGΔFG MG1655 ΔyagF ::FRT ΔyjhG ::FRT This study MGΔX MG1655 ΔxynR ::FRT This study MGΔJ MG1655 ΔyjhI ::FRT This study MGΔXJ MG1655 ΔxynR ΔyjhI This study MGΔ7 ΔXJ MG1655 ΔxynR ::FRT ΔyjhI ::FRT ΔyagE::FRT ΔyjhH::FRT Δeda::FRT ΔdgoA::FRT ΔyfaU::FRT ΔgarL::FRT ΔmhpE This study MGDTDA MG1655 Δybdl ΔtyrB ΔilvE ΔalaC Dasd This study Unless otherwise stated, E. coli strains were cultivated in Lysogeny Broth (LB) medium at 37°C on a rotary shaker running at 200 rpm. The antibiotics kanamycin, spectinomicine or chloramphenicol were added when required at concentrations of 50, 100 and 25 mg.L − 1 , respectively. Construction of the metabolite-based transcription factor biosensors Plasmids used in this study are reported in Table 2 while primers to make the various constructs can be downloaded from Table S3 in Supplementary data. Scheme of the constructed biosensor plasmids are reported in Figure S2 . To generate pBS0, a 200 bp fragment starting to -200 bp to -1 bp before ATG codon of yagE was amplified from MG1655 genomic DNA using primers TM7 and TM8, adding XhoI and HindIII restriction sites on 5’ and 3’ ends, respectively. The PCR product was ligated with HiT4 DNA ligase into gel purified XhoI-HindIII digested pREP22 vector from (Frazao et al. 2018 ) that carries syfp2 gene encoding the super yellow fluorescent protein (Kremers et al. 2006 ), which also carries the optimised RBS already present on the pREP22 plasmid between the promoting region of yagE and the syfp2 gene. To yield pBS2, the xynR gene was then inserted into pBS0 as follows. The coding sequence of this gene (759bp) along with 190 bp upstream the ATG codon was amplified from MG1655 strain using primers TM4 and TM11. Meanwhile, T7 terminator was amplified from pET-28a with TM5 and TM14. These primers allowed insertion of flanking regions corresponding to linearized pBS1 carried out with TM12 and TM13 primers. These three fragments were gel purified and incubated with HiFi DNA assembly mix to generate pBS2 plasmid, which was validated by sequencing after amplification in NEB-5α competent E. coli cells and purification. To generate pBS3, primers TM71 and TM146 were used to amplify the syfP2-xynR cassette of pBS2 plasmid whereas medium copy vector pZA33 was amplified with primers JFR64 and JFR65. Both fragments were gel-purified and ligated using NEBuilder HiFi DNA assembly. Validation of pBS3 was done by sequencing using primers JFR66/JF67 and TM10/TM16. To obtain pBS5, lldD gene from the genome of MG1655 was amplified with primers TM67/68, λ tL3 terminator of Cas9 was amplified from pCas plasmid (Jiang et al. 2015a ) using primers TM69 and TM70 and pBS2 using primers TM71 and TM72. All these fragments were gel purified and assembled with HiFi DNA assembly, to form pBS5. In addition, a DNA fragment that contains pJ23106 promoter (ordered from Eurofins genomics, France) and Shine-Dalgarno sequence was obtained by mixing primers TM63 and TM64 at 70°C, followed by slow decrease of temperature and then amplified with primers TM65 and TM66. After purification of the different fragments on agarose gel, they were assembled by the NEBuilder HiFi DNA Assembly (Biolabs) to generate pBS5, which was validated by sequencing using primers TM9/TM10/TM15/TM16 after amplification in NEB-5α competent E. coli cells and purification. Table 2 Plasmids constructed and used in this study Plasmid Short-Description source pET28a Novagen pZA33 Clm R , p15A ori ; promoter P A1 lacO1 Expressys pREP22 pZE13 derivative, promoter P yqhD hybrid: RBS O1 :syfp2 (Frazao et al. 2018 ) pCP20 Amp R , Clm R , pSC101 ori (Cherepanov and Wackernagel 1995 ) pBS0 pZE13 derivative, promoter P yagE : RBS O1 ::syfp2 This study pBS2 pZE13 derivative, promoter P yagE : RBS O1 ::syfp2, promoter P xynr :RBS xynr : xynR This study pBS3 pZA33 derivative, promoter P yagE : RBS O1 ::syfp2, promoter P xynr :RBS xynr : xynR This study pBS5 pZE13 derivative, promoter P yagE : RBS O1 ::syfp2, promoter P xynr :RBS xynr : xynR; promoter BBa_J23106:lldD This study pBS7 pZE13 derivative, promoter P yjhi : RBS O1 ::syfp2, This study pBS8 pZE13 derivative, promoter P yjh : RBS O1 ::syfp2, promoter P yjhI :RBS yjhI : yjhI This study pBS6 pZE13 derivative, promoter P yagE : RBS O1 ::syfp2, promoter P yjhI :RBS yjhI : yjhI This study pBS9 pZE13 derivative, promoter P yjhI : RBS O1 ::syfp2, promoter P xynr :RBS xynR : xynR This study pETM_empty pET28a with T5 promoter_empty This study pETM_alaC pET28a with T5 promoter carrying wild type alaC This study pETM_alaC** pET28a with T5 promoter carrying mutated alaC expressing AlaC A142P Y275D variant This study In-Fusion assembly protocol was used to construct the plasmids pBS0, pBS7, pBS8, pBS6 and pBS9 (see scheme of these plasmids in Figure S2 in supplementary material). Whenever plasmids were used as templates for PCR amplification, the PCR products were treated with DpnI (New England Biolabs) to eliminate the template plasmid. Plasmid pBS3 was used as the template to generate pBS0, pBS6, and pBS9. Biosensor pBS0 was obtained by first removing the transcriptional factor xynR and the reporter gene syfp2 with the yagE promoter from pBS3 by PCR (primers 3135 + 3136) and then reinsertion of the SYFP2 cassette as follows. The syfp2 gene with the yagE promoter and its corresponding T7 terminator was amplified by PCR (primers 3131 + 3132). After PCR purification, a second amplification was performed to create 15 bp homology regions with the first template DNA. Following the In-Fusion assembly reaction, each plasmid construct was transformed into competent E. coli NEB-5α cells according to the manufacturer’s instructions. Transformed cells were plated on LB agar supplemented with chloramphenicol (25 µg/mL final concentration). Colonies were initially screened by PCR, and plasmids were extracted from positive clones. Two confirmed constructs were sent for full plasmid sequencing (Eurofins Genomics). These procedures were repeated whenever new plasmid constructs were obtained. Plasmid pBS6 was obtained by replacing the transcriptional factor XynR with YjhI in pBS3. For this purpose, chromosomal DNA from E. coli MG1655 was used to amplify yjhI , including 199 bp upstream of the ORF to cover its native promoter. A T7 terminator was added 49 bp downstream of yjhI during the same PCR reaction (primers 3130 + 3137). The purified product was then subjected to a second PCR to add 15 bp homology regions (primers 3128 + 3129). In parallel, pBS3 was used to amplify the plasmid backbone lacking xynR and its promoter region (primers 3126 + 3127). The plasmid backbone and the amplified yjhI fragment were assembled using the In-Fusion reaction to generate biosensor pBS6. Plasmids pBS8 and pBS9 were constructed using the same approach, starting from pBS6 and pBS3, respectively. In pBS6, the yagE promoter of the reporter gene was replaced with the yjhI promoter to generate pBS8. Similarly, replacement of the yagE promoter with the yjhI promoter in pBS3 yielded pBS9. Finally, pBS7 was constructed by overlap PCR using pBS8 as the template and overlapping primers 3154 + 3155, which removed the transcriptional factor yjhI and its promoter. The primer 3154_FW can anneal to both the upstream and downstream regions of P jjhI + yjhI and its corresponding T7 terminator, respectively, in order to eliminate P jjhI + yjhI . The reverse primer 3155 was designed to contain a 50-bp homology region corresponding to the terminal sequence of primer 3154_FW. PCR was performed using pBS8 as the template with these primers. Following DpnI treatment and PCR clean-up, the product was transformed into E. coli NEB-5α cells. After PCR verification, positive clones were sent for whole-plasmid sequencing. The correct plasmids were kept at -20°C for further experiments. For the construction of pBios-JF2 through JF6, pBS3 was used as the backbone. The primers seq-R and seq-F were used on pBS3 to yield after a DNA fragment without the promoter sequence of xynR . Then, the insertion of new promoter sequences bearing a strong RBS 01 calculated according to Salis et al. (Salis 2011 ) was carried out by assembling this linearized fragment with PCR amplified synthetic fragments ( see Table S2 ) obtained with primer seq-xynR_F on the 5’ end and one of the 5 other primers listed in Table S2 on 3’-end using the NEBuilder HiFi DNA Assembly (Biolabs). All the constructs were validated by sequencing. Construction of plasmids for screening assays of OHB-producer enzymes. Plasmid pZS2-aspC-kdgT- Hh. araD (Frazão et al. 2023 ) was used as a template for reverse PCR using primers TM250 & TM251 for deleting aspC . Resulting plasmid, pZS2-kdgT- Hh. araD only carried genes kdgT ( E. coli ) and araD ( Herbaspirillum hutiense ) coding respectively a D-threonate transporter and a D-arabinonate dehydratase, previously characterised as D-threonate dehydratase for in vivo production of OHB (Frazão et al. 2023 )). A control plasmid was also obtained from the latter by reverse PCR using primers TM252 & TM253 in order to remove araD gene. The T7-inducible promoter of alaC gene and its mutant alaC** , encoding the variant AlaC A142P Y275D (Bouzon et al. 2017 ), initially cloned in pET28a were replaced by T5 promoter to allow IPTG induction in MG1655 strain. To this end, pET28a was amplified using the primers TM233 and TM234, while T5 promoter was amplified from the PCA24N vector from the ASKA library (Kitagawa et al. 2005 ) using primers TM242 and TM243. Both fragments displayed a floating tail of 10 nt on each side that are homologous to the ends of the other amplicon. The T5 promoter was cloned using the HiFi DNA Assembly Master Mix from NEB, following the manufacturer guidelines. The resulting vector was named pETM28, carrying the T5 promoter upstream of the coding sequence. Wild type alaC and mutant alaC** were amplified using the primers 917, bearing a Sac I site and 926, bearing a Hind III site. pET28M and the inserts were digested with Sac I and Hind III, gel purified and ligated using the HiT4 DNA ligase (NEB), following the supplier recommendations. The constructed plasmids pET28M_alaC and pET28M_alaC** were validated by sequencing. In vivo fluorescence experiments Unless otherwise stated, bacterial culture for in vivo fluorescence experiments were performed as follows. At day 1, a preculture of the bacterial strain was carried out in 2 to 5 ml LB medium with appropriate antibiotic in 50 ml falcon tube at 37°C in rotary shaker set at 200 rpm. The next day, the cultures were diluted in 5 ml of either LB or in M9-MOPS mineral medium containing 4 g/L xylose with the appropriate antibiotics in 50 ml falcon tubes at an initial OD 600 of 0.05. After 3 h and half for LB or 6 h for M9/MOPS, the cultures were collected by centrifugation at 3250 g for 15 minutes at room temperature, washed once with 5 ml PBS buffer, centrifugated again and then resuspended, unless otherwise stated, in 5 ml of PBS. Three hundred µl of the culture were delivered in 48-well microtiter plates and each culture was made in triplicate with independent bacterial clones. After 5 min incubation at 37°C, metabolites/effectors as indicated in corresponding figures were added. Growth at 600 nm and fluorescence of the SYFP2 protein were measured over a time period of 16 h either using a spectrofluorometer Clario Star from GmbH Labtech (excitation and emission wavelength set at 515 nm and 527 nm, respectively) or with a Biotek Synergy HTX from Agilent/Thermo Scientific using excitation filter at 485 and emission filter at 525 nm. The fluorescence expressed in arbitrary unit (AU) corresponds to the absolute fluorescence divided by OD measured at 600 nm. Fluorescence was also determined by flow cytometry using a BD Accuri™ C6 Plus Personal Flow Cytometer (BD Biosciences) with excitation set at 488 nm. Forward-scatter characteristics (FSC) and side-scatter characteristics (SSC) were detected as small-angle and large-angle scatters of the 488 nm laser, respectively. SYFP2 fluorescence was detected using a 530/30 nm (channel FL1) band-pass filter set. A total of 100,000 events were recorded per sample, and electronic gating was applied on the densest subset of cells based on forward- versus side-scatter height. The same gate was used to estimate median levels of SYFP2 fluorescence. The protocol was slightly modified for the screen of transaminases as the gene encoding these enzymes were cloned in a plasmid requiring their induction by IPTG. Therefore, after overnight culture in LB with appropriate antibiotics, the cells were reinoculated in LB with the appropriate antibiotic at OD 600 0.05, cultivated for another 2 h, after which 0.5 mM IPTG was added and growth in an incubator set at 37°C and 200 rpm was continued for 4 h. Cells were collected by centrifugation, washed once with PBS and resuspended in PBS as above. For threonate dehydratase no IPTG was added since the expression of the gene was constitutive and after overnight culture in LB, the culture was washed twice with PBS, resuspended in PBS at OD 600 ≈1.0 to which was added D-threonate (10 mM), OHB (1 mM) or water. The fluorescence/growth was monitored as described above using the Biotek Synergy HTX fluorometer or by flow cytometer. Data processing and statistical analysis For each data sets, absolute fluorescence intensity and OD 600 were recorded and the relative fluorescence in AU was obtained by dividing absolute value to OD. All experiments were done at least with three biological triplicates. Excel tools were used to calculate the mean, standard deviation and covariance (CV). Determination of the response threshold of the biosensor to its metabolites was obtained by fitting a nonlinear regression curve of the fluorescence intensity versus the concentration of the substrate using Solver tool in Excel or Graph Pad Prism V10.6.1 RESULTS Construction of the XynR and YjhI-based biosensor and investigation of their responses to D-xylonate. According to previous works, the transcription factor encoded by xynR acts as a repressor of D-xylonate metabolic genes present in the cryptic CP4-6 prophage whereas YjhI transcription factor is an activator of similar genes that belong to the KpLE12 phage-like elements ( Figure S1 ). To confirm these data, we build four types of biosensors on a medium copy plasmid, using syfp2 encoding the super yellow fluorescent protein as the readout (Kremers et al. 2006 ). For the two first biosensors carried on pBS0 and pBS7, syfp2 was under the control of a XynR or YjhI-dependent promoter, while the others (pBS3 and pBS8) carry also xynR or yjhI driven by its own promoter (see Figure S2 in Supplementary data). These plasmids were inserted by transformation into E. coli MG1655 strain, and fluorescence was measured over time in response to the addition of 1 and 20 mM D-xylonate. As shown in Fig. 1 , addition of this sugar acid triggered an increase in fluorescence intensity with both pBS0 and pBS3. However, the basal fluorescence in pBS3 was 200 times lower than that measured with pBS0, and the response time at both D-xylonate concentration was faster with pBS0 than with pBS3. These data are consistent with the repressive effect exerted by XynR, present in greater quantities in strain carrying pBS3 than pBS0 since in the latter, only xynR is expressed from its genomic copy. Unlike XynR, the YjhI-dependent fluorescence response upon addition of D-xylonate was significantly enhanced with pBS8 compared to pBS7. This result is consistent with the function of YjhI, working as an activating transcription factor (Banares et al. 2019 ). Also, the induction factor was higher with pBS8 than pBS7, even at low (1 mM) D-xylonate, which can be explained by the auto-activation of yjhI , caused by the binding of YjhI to its own promoter (Banares et al. 2019 ). In conclusion, the XynR-based biosensor harbors the typical repressed-repressor architecture whereas the activated-activator architecture characterizes the YjhI-based biosensor (Mannan et al. 2017 ). In addition, for both transcription factor-based biosensor, the fold induction was much higher at 20 than 1 mM D-xylonate, suggesting that XynR and YjhI exhibited a relatively weak affinity to D-xylonate. The metabolite effector of XynR and YjhI is the intermediate 2-keto-3-deoxy-D-xylonate (KDX) D-xylonate has been reported as the inducer of the XynR and YjhI-regulated operons but whether D-xylonate was the direct effector sensed by these TFs remained an open question (Banares et al. 2019 ; Shimada et al. 2017 ). The use of the XynR and YjhI-based biosensors developed above could be useful to address this question, which is all the more relevant in light of the fluorescence kinetics obtained in response to D-xylonate, as the increase of fluorescence was preceded by a lag period of few minutes to one hour (Fig. 1 ) . This lag period could be due either to the time needed to release all XynR-molecules that bind to yagE promoter, to the time for sufficient autoactivation of YjhI or to the time required to produce and accumulate a metabolite from D-xylonate. To solve this issue, we deleted yagF and yjhG which code for a D-xylonate dehydratase that catalyses the dehydration of D-xylonate into 2-keto-3-dexoy-D-xylonate (KDX) (see Figure S1 ) . Remarkably, XynR-and YjhI-based biosensors expressed in a mutant defective in both genes were no longer responsive to D-xylonate (Fig. 2 ). Furthermore, this experiment showed that the D-xylonate dehydratase encoded by yagF was more active than that encoded by yjhG , as the fluorescence response in the yagF mutant expressing the XynR-based biosensor was more reduced than in the yjhG mutant, and was even abolished in the case of YjhI-based biosensor in response to D-xylonate (Fig. 2 B). Based on these data, we investigated the effect of KDX as a potential metabolite effector of XynR and YjhI transcription factors. This molecule was purchased from Biosynth. Ltd and was validated by LC-MS in our laboratory as about ≈ 90% pure (data not shown). As shown in Fig. 3 A, the addition of 1 mM KDX to MG1655 carrying XynR-based biosensor on pBS3 resulted in a fluorescence response that was faster and higher than after addition of 1 mM D-xylonate. Moreover, the increase in fluorescence was more potent in the mutant strain MGΔ7 which is deleted for yagE and yjhH encoding promiscuous aldolases reported to cleave KDX into pyruvate and glycolaldehyde (Bhaskar et al. 2011 ; Liu et al. 2013 ). With the YjhI-based biosensor carried by pBS8, the response to KDX and 1 mM D-xylonate was very weak, similar as the one reported above (Fig. 1 B), but a 4 to 5-fold induction was measured in the aldolase-deficient MGΔ7 mutant, which can be explained by the absence of KDX degradation as well as by metabolization of D-xylonate into KDX in this mutant (Fig. 3 B). Altogether, these data support the notion that the direct effector of XynR and YjhI transcription factor is the intermediate KDX and not D-xylonate. The transcription factors XynR and YjhI are responsive to the non-natural molecule 2-oxo-4-hydroxybutyrate (OHB) A transcriptomic analysis of an E. coli MG1655 engineered for the production of 2,4-DHB by fermentation from glucose (Walther et al. 2018 ) revealed a strong upregulation of D-xylonate catabolic genes from the CP4-6 and KpLE2 -phage elements during the production phase of this non-natural metabolite, with notably yagE and yagF being increased by 13.6 and 9.1-fold, respectively (see data in Table S4 ). Interestingly, yjhI was also upregulated by about 8-fold, which was accompanied by an increased expression, albeit weak, of yjhH and yjhG . Based on these data, we asked whether 2,4-DHB or OHB could mimic the effect of KDX to activate the XynR- and/or YjhI-based biosensor. Results of this experiment are reported in Fig. 4 . DHB was used at 100 mM because its action on the D-xylonate dependent operons was expected at high concentration whereas OHB was used at 1 mM because intracellular concentration of this metabolite was estimated in the millimolar range (our unpublished data). It can be seen that both 2,4-DHB and OHB caused a comparable and significant increase in fluorescence intensity over time with both biosensors. Also, the dynamic response expressed as fold induction was very similar for response to these unnatural effectors on the two TFs-based biosensors. However, the time response to OHB was clearly faster than that of 2,4-DHB, suggesting that the effect of the latter molecule was not direct but could result from its metabolization. We recently reported that 2,4-DHB can be oxidized into OHB by membrane-associated lactate oxidoreductases encoded by lldD , dld and ykgEFG (Pinchuk et al. 2009 ), as a first step of its assimilation by E. coli (Malfoy et al. 2024 ). We therefore repeated this experiment using the MGΔLO strain deficient in these lactate oxidoreductases. As shown in Fig. 4 C and 4 D, the increase of fluorescence signal upon addition of 2,4-DHB was totally abolished in this mutant expressing the biosensor carried on pBS3 or pBS8. In contrast, the absence of lactate oxidoreductase had no effect on the response of these biosensors to OHB, indicating that this molecule is likely a direct effector of XynR and YjhI. No crosstalk in the response of XynR and YjhI to KDX and OHB Since XynR and YjhI were responsive to the presence of KDX and OHB, we wanted to verify whether a cross-talk between these two TFs could take place. In other words, could it be possible that XynR affected the expression of genes that belong to the YjhI-controlled genes operon and vice versa? To answer this question, we constructed the plasmid pBS9 bearing xynR under its own promoter while syfP2 reporter gene was under yjhI promoter. Conversely, plasmid pBS6 that carried yjhI under its own promoter with syfp2 under yagE promoter was built (see Figure S2 for the scheme of these constructions). The reference strain MG1655, single mutant and double mutants defective in xynR and yjhI were transformed with these two plasmids and challenged with 1 mM OHB or 20 mM D-xylonate, which was used instead of KDX due to the limited availability of this compound. This experiment showed that the fluorescence induction in response to OHB and D-xylonate in MG ΔxynRΔyjhI mutant strain showed barely no significant difference as compared to the control. This result supported the notion that each transcription factor controls only the genes of its own operon, and therefore indicated an absence of cross-talk in the control of these operons by D-xylonate. The 3 to 6-fold induction observed in response to D-xylonate and OHB with pBS9 in both MG1655 and MGΔxynR can be ascribed to the action of yjhI present in the genome, which triggered expression of syfp2 gene reporter that is under the yjhI promoter in this plasmid. Conversely, the weak fold increase in response to OHB and D-xylonate in MG1655 or MGDyjhI strains carrying pBS6 was likely due to the low level of XynR protein, as the gene is expressed from the chromosomal copy, and hence its low abundance resulted in an already high expression of syfp2 gene that is weakly enhanced by addition of OHB or D-xylonate. This behavior resembled that found with MG1655 expressing pBS0 (see Fig. 1 ). Characteristic performances of the XynR and YjhI-based biosensor to KDX and OHB As clearly outlined in the reviews by F. Zhang and colleagues (Hartline et al. 2021 ; Mannan et al. 2017 ), several criteria can be used to evaluate the biosensor performances which include the response threshold ( i.e. K 0.5 ), defined as the concentration of metabolite required for 50% of the maximal expression or induction, the detection range and the dynamic range corresponding to the maximal fold increase relative to the baseline. To evaluate these parameters for XynR-and YjhI-based sensors to OHB and KDX, MGΔ7 strain deleted for the so far seven identified genes encoding pyruvate-dependent aldolase (He et al. 2020 ) was used to prevent the metabolization of KDX into pyruvate and glycoladehyde (for KDX) and OHB into pyruvate and formaldehyde (Banares et al. 2021 ; Malfoy et al. 2024 ). Results reported in Fig. 6 showed that the XynR-based biosensor turned to be more sensitive to OHB than KDX (K 0.5 ≈ 0.27 vs 1.0 mM, Table 3 ), while a comparable K 0.5 in the range of 0.3–0.4 mM was obtained for both metabolites with the YjhI-based biosensor (Table 3 ). Furthermore, the cooperative behavior of the two TF-based biosensors expressed by the Hill number (n H ) was significantly different between KDX and OHB. While the cooperativity of the XynR-based biosensor was low for KDX (nH ≈ 1.13), it was two times higher for OHB. The opposite behavior was observed for the YjhI-based biosensor, for which the cooperativity was roughly two time higher for KDX (n H ≈2.7) than for OHB (see Table 3 ). These results suggested that although both effectors may interact with the same protein, they differentially modulate cooperative interactions within the protein. The dynamic range (DR), expressed as the fold induction of fluorescence relative to baseline, for the natural (KDX) and artificial (OHB) effectors can be also extrapolated from the dose-response curves. They revealed that the dynamic range of the YjhI-based biosensor for OHB and KDX was about 2-times better than that of XynR-based biosensor ( i.e. in the range of 50-fold versus 25; see Table 3 ). Table 3 Characteristic performance of XynR and YjhI transcription factors towards their metabolic effectors Effector XynR* YjhI* n H K 0.5 DR n H K 0.5 DR OHB 2.05 ± 0.7 0.27 ± 0.05 25 ± 5.0 1.25 ± 0.25 0.36 ± 0.07 48 ± 6.0 KDX 1.13 ± 0.3 1.0 ± 0.04 25 ± 5.0 2.7 ± 0.8 0.32 ± 0.04 52 ± 8.0 D-xylonate 2.1 ± 0.7 4.1 ± 0.7 nd 2.4 ± 0.4 4.0 ± 0.5 nd *Values are the mean ± SD of four biological replicates. n H means Hill number, DR means dynamic range Although KDX is the direct effector of XynR and YjhI transcription factors, the dose-response curves to D-xylonate were also performed, showing that both TFs exhibited similar characteristic performances with a response threshold in the range of 5 mM, and a dynamic range of about 30 ± 5.0 ( Figure S3 in Supplementary data). Development of a OHB biosensor From the results presented above, it turned out that both transcription factors XynR and YjhI could be exploited for the design of an OHB-responsive biosensor. However, to ensure the practical use of these biosensors, we had to verify their specificity to this unnatural metabolite in relation to other metabolites present in bacterial cells, particularly organic and amino acids. These metabolites were tested at a pretty large concentration (10 mM), and the fluorescence signal that can be triggered by these compounds was compared to that obtained upon addition of 1 mM of OHB. As shown in Fig. 7 A, most of the organic acids had minor to no effect on the XynR-based biosensor (Note: the time course response of these metabolites are reported in Figure S4 , supplementary data ) . However, compounds that, like OHB, have a ketone function at C2, such as pyruvate, α-ketoglutarate, and aspartate caused a 3 to 4-fold induction of fluorescence. A 2 to 3-fold induction was also recorded with organic acids having an α hydroxyl function, such as D and L-lactate or 2-hydroxybutyrate. Less obvious was the finding of a 4-fold induction caused by 10 mM fumarate. On the other hand, the basal fluorescence of the biosensor was lower than the control in the presence of L-malate or glycolate, which both also harbor an α-hydroxyl group. Same effect was found with glycolaldehyde. How these molecules caused such a reduction of the fluorescence is so far unclear but nevertheless, these effects can be considered negligible overall, as they occurred at concentrations well above those experienced in cells (Bennett et al. 2009 ). Similar results were obtained with the YjhI-based biosensor, with notably a stronger inhibition of the signal fluorescence by L-malate, glycolate and glycolaldehyde and an even higher fold induction in response to fumarate ( Figure S5 in suppl data). We also investigated the effects of sugar acids homologous to D-xylonate as they can be used as carbon substrate. Results of this experiment clearly indicated that XynR-based biosensor was responsive neither to the enantiomer (L-xylonate) nor the stereoisomer (L-arabinonate) of D-xylonate (Fig. 7 B ) . The lack of effect of these sugar acids is consistent with the fact that these compounds cannot be metabolized by E. coli (Ren et al. 2022 ). Conversely, the 4-fold increase in response to D-gluconate could be explained either by its dehydration into 2-keto-3-deoxy-D-gluconate catalysed by YagF/YagG dehydratase or to the possibility that gluconate can be converted into 2-keto-3-deoxy-6-phosphogluconate (KDPG) through the Entner-Doudoroff pathway (Eisenberg and Dobrogosz 1967 ). We solved this issue by showing that a mutant lacking the dehydratase did not exhibit any fluorescence signal in response to D-gluconate (data not shown). D-glucuronate and D-threonate were both inefficient as expected since the former one can only be metabolized via the Entner-Doudoroff providing D-glucuronate isomerase encoded by uxaC (Portalier et al. 1980 ) is functional, which needs the presence of this sugar acid to be expressed (Mandrand-Berthelot et al. 2004 ), whereas only L-threonate is reported to be metabolized by E. coli (Frazão et al. 2023 ). Similar results were obtained with the YjhI-based biosensor (see Figure S5B in suppl. Data). Beside specificity, other relevant properties of a metabolite responsive transcription factor exploited as a biosensor should be to exhibit a wide range of effector concentrations and a large dynamic response characterized by high fold induction of fluorescence (response) relative to the background. These properties can be tuned at first glance through promoter engineering of either the transcription factor and/or the reporter gene syfp2 . Starting from pBS3 as the reference plasmid bearing the XynR-based biosensor, we generated a series of plasmids in which either the native RBS in the XynR-promoter was replaced by a synthetically optimised RBS 01 according to RBS calculator (Salis 2011 ) or the native promoter of XynR was replaced with iGEM promoters of decreasing strength, while retaining the same RBS 01 in these synthetic promoters. These different constructs did not result in a better dynamic range to OHB as compared to the original construct in pBS3 (see Figure S6 in suppl. data). It is clear that other modifications, such as engineering the transcription factor XynR to broaden the detection range or sensitivity to OHB, should be considered, but this work is beyond the scope of this study. Use of the XynR- and YjhI-based biosensor to screen for OHB producing enzymes. The compound OHB is the most proximal intermediate in the production of 2,4-DHB, a chemical platform for the production of several added-value products (Francois 2023 ). The synthesis of OHB can be obtained by transamination of L-homoserine with pyruvate or α-ketoglutarate as the co-substrate (Bouzon et al. 2017 ; Walther et al. 2018 ). However, the catalytic efficiency of this enzyme is approximately 100 to 1000 times lower than that of other enzymes in these pathways. Improving the activity of these enzymes is therefore mandatory in order to increase the flux in the pathway and consequently increase the rate of 2,4-DHB production. In that frame, a biosensor able to monitor the formation of the product of the reaction catalysed by the enzyme, such as OHB in our case, could be an appropriate tool for high throughput screening of more active transaminases. We thus decided to validate the use of the XynR/YjhI-based biosensor to screen for transaminase activity using the variant of AlaC A142P Y245D as a positive control since this variant turned out to be the most active to produce OHB from L-homoserine in the presence of pyruvate or a-ketoglutarate, albeit at a still low catalytic efficiency of ≈ 500 M − 1 s − 1 (Bouzon et al. 2017 ). To validate the screen, we used the strain MGDT which was deleted for tyrB , ilvE, ybdL and alaC to get rid of most of the alternate transaminases exhibiting even a weak activity on L -homoserine (Walther et al. 2018 ; Walther et al. 2017 ). It was also deleted of asd to avoid endogenous production of L-homoserine. At first glance, we tested the XynR-biosensor through co-transformation of this strain with pBS3 and either pETM (empty), used as control, pETM carrying alaC (coding for wild type AlaC) and pETM carrying alaC** (coding for the variant AlaC A142P Y245D ). Since these genes were under the control of an IPTG-inducible promoter, part of the culture was treated for 4 h with 0.5 mM IPTG before addition of the various metabolites. It turned out that the results obtained were inconsistent, with an apparent interference of IPTG on the expression of pBS3 since the fluorescence response to OHB was strongly impaired in cells treated with IPTG (data not shown). We therefore tested the YjhI-based biosensor following the same experimental procedure. As reported in Fig. 8 , the fold induction of fluorescence of IPTG-untreated cultures in response to homoserine, pyruvate or α-ketoglutarate was relatively weak and always lower than in response to 1 mM OHB use as control. One could however notice a statistically significant (about 8-fold) fold induction of fluorescence relative to the control in the strain carrying pETM-alaC**. This result indicated that even in the absence of IPTG, the genes in these plasmids could be expressed and moreover these data supported the fact that AlaC A142P Y250D variant was more active than AlaC to produce OHB. Also, the weak but significant increase in fluorescence in strain expressing the empty plasmid in response to homoserine alone or with pyruvate or α-ketoglutarate could be ascribed to endogenous transaminases such as the one encoded by aspC which has been reported to have a weak OHB producing activity (Walther et al. 2018 ). More importantly, the treatment of the culture with IPTG prior addition of the various metabolites clearly validated the biosensor tool since a 40-fold induction of fluorescence was recorded after addition of L-homoserine with either pyruvate or α-ketoglutarate in strain expressing the AlaC A142P Y250D variant, whereas only a 8-10-fold induction was measured with the strain expressing the wild type enzyme. Of note, a 40-fold induction was found to be close to the maximal dynamic response of OHB (Table 3 ). Furthermore, pre-treatment with IPTG of bacterial cultures carrying either an empty plasmid, pETM-alaC, or pETM-alaC** caused a fold induction of fluorescence in response to homoserine, pyruvate, or α-ketoglutarate alone, which was almost equivalent to that obtained after adding 1 mM OHB. Part of the explanation may lie in the action of other transaminases, particularly that encoded by aspC , but also in the fact that the presence of IPTG may have caused a metabolic rearrangement, providing precursors for OHB synthesis. We then tested the D-threonate dehydratase activity which was derived from the promiscuous D-arabinonate dehydratase of Herbaspirillum huttiense ( HharaD ) (Watanabe et al. 2019 ). This enzyme was shown to synthetize OHB by dehydration of D-threonate in a conceived synthetic pathway starting from ethylene glycol and ending with the formation of 2,4-DHB (Frazão et al. 2023 ). For this validation, MG1655 strain was co-transformed with pBS3 (XynR-based biosensor) and either pZS2 carrying HharaD with kdgT encoding a transporter of Cupriavidus necator reported to facilitate the import of D-threonate, or with pZS2 carrying only kgdT . Remarkably, the addition of D-threonate to bacterial cells expressing HharaD encoding this threonate dehydratase resulted in a significant increase in fluorescence signal that was equivalent to that obtained after the addition of 1 mM OHB, whereas the fluorescence of cells lacking this gene remained at a basal level after the addition of the sugar acid (Fig. 9 ). Of note, the assay was not carried out with the YjhI-based biosensor due to the scarcity of D-threonate. Use of the XynR-based biosensor to screen for 2,4-DHB transporters Since 2,4-DHB is a non-natural molecule for E. coli , we sought whether the OHB-biosensor could be used to screen for a transporter of this molecule taking into account that OHB can be produced by the oxidation of 2,4-DHB by the membrane-associated lactate oxidoreductase (Malfoy et al. 2024 ). To ensure that this reaction would not be rate limiting in the screen, we inserted lldD encoding one of the three lactate oxidoreductases present in E. coli into the pBS3 plasmid under the constitutive synthetic pJ23106 promoter yielding pBS5 (see Figure S2 ). We validated this construct by measuring the activity on 2,4-DHB in lysates of cells that have been transformed with pBS5 and found a 10-times higher activity on this compound than in cells carrying pBS3 (data not shown). Then, around 200 E. coli mutants of the Keio collection (Baba et al. 2006 ) deleted for known and putative transporters of organic and amino acids were transformed with pBS5 and cultivated in M9 with 0.4% D-xylose in the presence of either 10 or 100 mM 2,4-DHB. Fluorescence was monitored by flow cytometry after 4, 8 and 24 h of culture. Results collected after 24 h showed that there were several genes belonging to amino acids and organic acids transporters that exhibited significant reduction of the fluorescence and thus potentially impaired the import of 2,4-DHB (Fig. 10 ). In particular, two out of the 30 retained candidates causing > 50% reduction of fluorescence encoded a multidrug efflux pump, 11 encoded amino acids transporters, 6 encoded organic acids transporters and 5 encoded uncharacterized transporters ( Table S2 ). When the screen was carried out with 100 mM 2,4-DHB, only two candidates of those identified at 10 mM were found, namely proX which encodes glycine betaine ABC transporter and yfdC , which encodes a yet uncharacterized transporter (Table S2 ). This large collection of transporters suggested a seemingly unspecific import of 2,4-DHB. To get more insight about this screening assay, we choose to delete ydfC , acrD , glnH , kdgT and ygbN into the strain MGΔLO and confirmed our original data with the mutant from the Keio collection (data not shown). DISCUSSION The construction of biosensors dependent on the transcription factors XynR and YjhI enabled us to resolve two major problems concerning the induction of the D-xylonate catabolic pathway in E. coli. First, the natural effector of these two transcription factors that induce the catabolic genes present on the yag and yjh operons is not D-xylonate, but the intermediate KDX, which is formed by the dehydration of D-xylonate by D-xylonate dehydratase encoded by yagF and yjhG . Moreover, our results suggested that in vivo , YagF is more active than YjhgG. This higher activity could be explained either by a higher expression of yagF than yjhG , or by a higher affinity of YagF for D-xylonate. To date, only kinetic data have been reported for the D-xylonate dehydratase YjhG, showing a K M of 4.88 mM and a relative weak k cat /K M of 66 M − 1 s − 1 (Jiang et al. 2015b ). Interestingly, this value is in the range of the concentration of D-xylonate leading to 50% of the maximal fluorescence induction of XynR-and YjhI based biosensor. On the other hand, we have shown that there is no interference in the control of the yagEFG and yjhIHG operons, as each transcription factor regulates its own operon without interfering with the other's operon. The other part of the work was to examine whether XynR and/or YjhI are also responsive to the non-natural metabolites 2,4-DHB and/or OHB to repurpose them as biosensor tools to improve the synthetic pathways of the platform molecule 2,4-DHB (Francois 2023 ). Our data showed that OHB can effectively play the same role as KDX on these two transcription factors, most likely because, except for the fact that the latter is a 5-carbon molecule, OHB and KDX share the same chemical structure. Based on data with other organic and amino acids used to test the specificity of Xynr and YjhI to OHB and KDX, it can be proposed that the minimal structure recognized by these TFs required the functional carboxyl group, a keto function on the carbon α and a hydroxyl group on the carbon g. Also, the fact that XynR and YjhI are responsive to these two effectors is due to their 3D-structure perfectly overlapping, according to the Alphafold-2 prediction (see Figure S8 in in supplementary data). However, there is an additional β-sheet between aa 54 and 56, as well as an extended short α-helix at the C-ter of the YjhI protein. These small differences could be sufficient to explain why YjhI is an activator and not a repressor like XynR, but more detailed structure-function will be required to ascertain this suggestion. Although the protein structure of these two transcription factors is almost identical, their kinetic characteristics, namely K 0.5 and the Hill number n H , with respect to OHB and KDX were found different. The transcription factor XynR exhibited an affinity for OHB that is 3 to 4 times better than for KDX, while the opposite was observed for YjhI. Altogether, the finding that XynR and YjhI are structurally similar while the former is acting as a repressor of transcription and the latter as an activator and that they exhibit kinetic differences to their ligands may require more detailed structure-function analysis, which is beyond the scope of this work. In previous works, we developed 4 different synthetic pathways leading to the production of 2,4-DHB from C1 to C6-renewable carbon sources (Frazão et al. 2023 ; Walther et al. 2018 ; Walther et al. 2017 ). This synthon can be considered a unique bio-based platform molecule capable of giving rise to several value-added molecules, including a hydroxylated derivative of methionine, 1,3-propanediol, 3-hydroxypropionic acid, 1,2,4-butanetriol and biopolymers (Francois 2023 ; Pascouau et al. 2023a ; Pascouau et al. 2023b ). Two of these four pathways, which have proven to be the most suitable for industrial application, rely on the formation of the intermediate OHB by promiscuous enzymes whose catalytic efficiency is far too low, significantly hampering the productivity and yield of 2,4-DHB production. An OHB-sensitive biosensor could be an appropriate method, which combined with flow cytometry coupled with cell sorting (FACS), would enable the selection of enzyme variants with higher catalytic activity. The XynR-based biosensor can be optimally designed for use in this context, as it features a repressed-repressor architecture that enables positive selection, i.e. the signal response is triggered only when the target metabolite is produced. Our data aligned in part with these requirements. We validated the potential value of this OHB-sensitive biosensor as a screening tool for threonate dehydratase, the rate limiting enzyme in the synthetic pathway yielding to DHB from ethylene glycol as the carbon source (Frazão et al. 2023 ). However, it was not possible to use this XynR-based biosensor as a screening method for homoserine transaminase, which catalyzes the rate-limiting step in the formation of DHB from glucose via the aspartate-homoserine pathway (Walther et al. 2018 ), whereas the YjhI-based biosensor, which has an activator-activated architecture, proved to be functional. It is difficult to find a simple explanation for this result, but one possible reason is that the screening requires the presence of two plasmids in the bacteria, one carrying the biosensor and the other carrying the gene encoding the transaminase. For the latter, gene expression is dependent on the IPTG-inducible promoter. We observed strong interference on the XynR-based biosensor upon incubation with IPTG as witnessed by a weak increase of fluorescence in response to OHB. Finally, we showed that this biosensor could be useful to identify putative DHB transporter, using the single-gene knockout mutants of E. coli (Baba et al. 2006 ). Notably, ygbN was identified in this screen, which we previously reported as a transporter of the D-form of 2,4-DHB (Malfoy et al. 2024 ). It is worth noticing that the OHB-biosensor reported in this work is totally different from the OHB biosensor (termed HOB biosensor) developed by Schann et al . (Schann et al. 2024 ). The latter is a growth-based sensor whose actual sensor is formaldehyde that is produced via promiscuous aldolases and transaminases to be assimilated through homoserine in a strain that is defective in the natural aspartate -homoserine pathway. Declarations Conflicts of interest The authors declare no commercial or financial conflict of interest. Author contribution statement T.M. and J.M.F. designed the study. T.M., C.A., J.F., J. L-H., and J.M.F. conducted the experiments. T.M. and J.M.F. drafted the manuscript, which was reviewed and improved by all authors before a final version written by J.M.F. and approved by all authors. Funding This work was supported by French Research National Agency (ANR) grant ANR-CE43-0008-01 (PolyDHB) and grant ANR-21-COBI-0003-01 (SYNBIOMET) to J.M.F. Author Contribution T.M. and J.M.F. designed the study. T.M., C.A., J.F., J. L-H., and J.M.F. conducted the experiments. T.M. and J.M.F. drafted the manuscript, which was reviewed and improved by all authors before a final version written by J.M.F. and approved by all authors. Acknowledgements We are grateful To Dr C. Frazao & Prof Th. Walther from TU Dresden, Germany, for the kind gift of pZS2-kdgT_ Hh .araD plasmids together with a few mg of D-threonate. References Baba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, Datsenko KA, Tomita M, Wanner BL, Mori H (2006) Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2:20060008. 10.1038/msb4100050 Banares AB, Nisola GM, Valdehuesa KNG, Lee WK, Chung WJ (2021) Understanding D-xylonic acid accumulation: a cornerstone for better metabolic engineering approaches. Appl Microbiol Biotechnol 105(13):5309–5324. 10.1007/s00253-021-11410-y Banares AB, Valdehuesa KNG, Ramos KRM, Nisola GM, Lee WK, Chung WJ (2019) Discovering a novel D-xylonate-responsive promoter: the PyjhI-driven genetic switch towards better 1,2,4-butanetriol production. Appl Microbiol Biotechnol 103(19):8063–8074. 10.1007/s00253-019-10073-0 Bennett BD, Kimball EH, Gao M, Osterhout R, Van Dien SJ, Rabinowitz JD (2009) Absolute metabolite concentrations and implied enzyme active site occupancy in Escherichia coli. Nat Chem Biol 5(8):593–599 :nchembio.186 [pii];10.1038/nchembio.186 [doi] Bhaskar V, Kumar M, Manicka S, Tripathi S, Venkatraman A, Krishnaswamy S (2011) Identification of biochemical and putative biological role of a xenolog from Escherichia coli using structural analysis. Proteins 79(4):1132–1142. 10.1002/prot.22949 Bouzon M, Perret A, Loreau O, Delmas V, Perchat N, Weissenbach J, Taran F, Marliere P (2017) A Synthetic Alternative to Canonical One-Carbon Metabolism. ACS Synth Biol 6(8):1520–1533. 10.1021/acssynbio.7b00029 Cherepanov PP, Wackernagel W (1995) Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant. Gene 158(1):9–14 Dahms AS (1974) 3-Deoxy-D-pentulosonic acid aldolase and its role in a new pathway of D-xylose degradation. Biochem Biophys Res Commun 60(4):1433–1439 Eisenberg RC, Dobrogosz WJ (1967) Gluconate metabolism in Escherichia coli. J Bacteriol 93(3):941–949. 10.1128/jb.93.3.941-949.1967 Francois JM (2023) Progress advances in the production of bio-sourced methionine and its hydroxyl analogues. Biotechnol Adv 69:108259. 10.1016/j.biotechadv.2023.108259 Frazão CJR, Wagner N, Rabe K, Walther T (2023) Construction of a synthetic metabolic pathway for biosynthesis of 2,4-dihydroxybutyric acid from ethylene glycol. Nat Commun 14(1):1931. 10.1038/s41467-023-37558-x Frazao CR, Maton V, Francois JM, Walther T (2018) Development of a Metabolite Sensor for High-Throughput Detection of Aldehydes in Escherichia Coli. Front Bioeng Biotechnol 6:118. 10.3389/fbioe.2018.00118 Hanko EKR, Paiva AC, Jonczyk M, Abbott M, Minton NP, Malys N (2020) A genome-wide approach for identification and characterisation of metabolite-inducible systems. Nat Commun 11(1):1213. 10.1038/s41467-020-14941-6 Hartline CJ, Schmitz AC, Han Y, Zhang F (2021) Dynamic control in metabolic engineering: Theories, tools, and applications. Metab Eng 63:126–140. 10.1016/j.ymben.2020.08.015 He H, Hoper R, Dodenhoft M, Marliere P, Bar-Even A (2020) An optimized methanol assimilation pathway relying on promiscuous formaldehyde-condensing aldolases in E. coli. Metab Eng 60:1–13. 10.1016/j.ymben.2020.03.002 Jiang Y, Chen B, Duan C, Sun B, Yang J, Yang S (2015a) Multigene editing in the Escherichia coli genome via the CRISPR-Cas9 system. Appl Environ Microbiol 81(7):2506–2514. 10.1128/AEM.04023-14 Jiang Y, Liu W, Cheng T, Cao Y, Zhang R, Xian M (2015b) Characterization of D-xylonate dehydratase YjhG from Escherichia coli. Bioengineered 6(4):227–232. 10.1080/21655979.2015.1040208 Kitagawa M, Ara T, Arifuzzaman M, Ioka-Nakamichi T, Inamoto E, Toyonaga H, Mori H (2005) Complete set of ORF clones of Escherichia coli ASKA library (a complete set of E. coli K-12 ORF archive): unique resources for biological research. DNA Res 12(5):291–299. 10.1093/dnares/dsi012 Kremers GJ, Goedhart J, van Munster EB, Gadella TW Jr (2006) Cyan and yellow super fluorescent proteins with improved brightness, protein folding, and FRET Forster radius. Biochemistry 45(21):6570–6580. 10.1021/bi0516273 Liu H, Ramos KR, Valdehuesa KN, Nisola GM, Lee WK, Chung WJ (2013) Biosynthesis of ethylene glycol in Escherichia coli. Appl Microbiol Biotechnol 97(8):3409–3417. 10.1007/s00253-012-4618-7 Malfoy T, Alkim C, Barthe M, Fredonnet J, François JM (2024) Enzymatic promiscuity and underground reactions accounted for the capability of Escherichia coli to use the non-natural chemical synthon 2,4-dihydroxybutyric acid as a carbon source for growth. Microbiol Res 288:127888. https://doi.org/10.1016/j.micres.2024.127888 Mandrand-Berthelot MA, Condemine G, Hugouvieux-Cotte-Pattat N (2004) Catabolism of Hexuronides, Hexuronates, Aldonates, and Aldarates. EcoSal Plus 1(1). 10.1128/ecosalplus.3.4.2 Mannan AA, Liu D, Zhang F, Oyarzun DA (2017) Fundamental Design Principles for Transcription-Factor-Based Metabolite Biosensors. ACS Synth Biol 6(10):1851–1859. 10.1021/acssynbio.7b00172 Miller JH (1992) A Short Course in Bacterial Genetics: A Laboratory Manual and Handbook for Escherichia coli and Related Bacteria. Plainview, NY Pascouau C, Méreau R, Grau E, Wirotius A-L, Carlotti S, Cramail H, Peruch F (2023a) Ring-Opening Copolymerization of α-Hydroxy-γ-butyrolactone and ε-Caprolactone. Toward the Metal-Free Synthesis of Functional Polyesters. ACS Appl Polym Mater 5(8):6685–6694. 10.1021/acsapm.3c01303 Pascouau C, Wirotius A-L, Carlotti S, Cramail H, Peruch F (2023b) Functional polyesters via ring-opening copolymerization of α–hydroxy–γ–butyrolactone and ε-caprolactone: La[N(SiMe3)2]3 as an efficient coordination-insertion catalyst. Eur Polymer J 185:111793. https://doi.org/10.1016/j.eurpolymj.2022.111793 Pinchuk GE, Rodionov DA, Yang C, Li X, Osterman AL, Dervyn E, Geydebrekht OV, Reed SB, Romine MF, Collart FR, Scott JH, Fredrickson JK, Beliaev AS (2009) Genomic reconstruction of Shewanella oneidensis MR-1 metabolism reveals a previously uncharacterized machinery for lactate utilization. Proc Natl Acad Sci U S A 106(8):2874–2879. 10.1073/pnas.0806798106 Portalier R, Robert-Baudouy J, Stoeber F (1980) Regulation of Escherichia coli K-12 hexuronate system genes: exu regulon. J Bacteriol 143(3):1095–1107. 10.1128/jb.143.3.1095-1107.1980 Ren Y, Eronen V, Blomster Andberg M, Koivula A, Hakulinen N (2022) Structure and function of aldopentose catabolism enzymes involved in oxidative non-phosphorylative pathways. Biotechnol Biofuels Bioprod 15(1):147. 10.1186/s13068-022-02252-5 Salis HM (2011) The ribosome binding site calculator. Methods in enzymology 498:19–42 doi:B978-0-12-385120-8.00002-4 [pii];10.1016/B978-0-12-385120-8.00002–4 [doi] Schann K, Bakker J, Boinot M, Kuschel P, He H, Nattermann M, Paczia N, Erb T, Bar-Even A, Wenk S (2024) Design, construction and optimization of formaldehyde growth biosensors with broad application in biotechnology. Microb Biotechnol 17(7):e14527. 10.1111/1751-7915.14527 Shimada T, Momiyama E, Yamanaka Y, Watanabe H, Yamamoto K, Ishihama A (2017) Regulatory role of XynR (YagI) in catabolism of xylonate in Escherichia coli K-12. FEMS Microbiol Lett 364(22). 10.1093/femsle/fnx220 Thomason LC, Costantino N, Court DL (2007) E. coli genome manipulation by P1 transduction. Curr Protoc Mol Biol Chap 1:Unit 1 17. 10.1002/0471142727.mb0117s79 Walther T, Calvayrac F, Malbert Y, Alkim C, Dressaire C, Cordier H, Francois JM (2018) Construction of a synthetic metabolic pathway for the production of 2,4-dihydroxybutyric acid from homoserine. Metab Eng 45:237–245. 10.1016/j.ymben.2017.12.005 Walther T, Topham CM, Irague R, Auriol C, Baylac A, Cordier H, Dressaire C, Lozano-Huguet L, Tarrat N, Martineau N, Stodel M, Malbert Y, Maestracci M, Huet R, Andre I, Remaud-Simeon M, Francois JM (2017) Construction of a synthetic metabolic pathway for biosynthesis of the non-natural methionine precursor 2,4-dihydroxybutyric acid. Nat Commun 8:15828 :ncomms15828 [pii];10.1038/ncomms15828 [doi] Wang HH, Isaacs FJ, Carr PA, Sun ZZ, Xu G, Forest CR, Church GM (2009) Programming cells by multiplex genome engineering and accelerated evolution. Nature 460(7257):894–898. 10.1038/nature08187 Watanabe S, Fukumori F, Watanabe Y (2019) Substrate and metabolic promiscuities of d-altronate dehydratase family proteins involved in non-phosphorylative d-arabinose, sugar acid, l-galactose and l-fucose pathways from bacteria. Mol Microbiol 112(1):147–165. 10.1111/mmi.14259 Additional Declarations No competing interests reported. Supplementary Files OHbiosensorSupplementaryInformation.docx Cite Share Download PDF Status: Published Journal Publication published 01 May, 2026 Read the published version in Applied Microbiology and Biotechnology → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8830407","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":590988117,"identity":"7a1d9567-c4d9-4a47-8e41-1d468f9d9290","order_by":0,"name":"Thibault MALFOY","email":"","orcid":"","institution":"UMR INSA -CNRS5504 and UMR INSA-INRAE 792","correspondingAuthor":false,"prefix":"","firstName":"Thibault","middleName":"","lastName":"MALFOY","suffix":""},{"id":590988118,"identity":"0a9b9641-75d9-44c1-b8be-ee20c1242d66","order_by":1,"name":"Ceren ALKIM","email":"","orcid":"","institution":"UMR INSA -CNRS5504 and UMR INSA-INRAE 792","correspondingAuthor":false,"prefix":"","firstName":"Ceren","middleName":"","lastName":"ALKIM","suffix":""},{"id":590988119,"identity":"31a1fcab-92f1-4e2e-8442-b71717bdcba7","order_by":2,"name":"Julie FREDONNET","email":"","orcid":"","institution":"UMS INRAE-INSA-CNRS","correspondingAuthor":false,"prefix":"","firstName":"Julie","middleName":"","lastName":"FREDONNET","suffix":""},{"id":590988120,"identity":"943ea46e-f200-410d-8e0e-e919a1cf3688","order_by":3,"name":"Juan LAJARIN-HERNANDEZ","email":"","orcid":"","institution":"UMR INSA -CNRS5504 and UMR INSA-INRAE 792","correspondingAuthor":false,"prefix":"","firstName":"Juan","middleName":"","lastName":"LAJARIN-HERNANDEZ","suffix":""},{"id":590988121,"identity":"d37fd182-1e4d-42a8-b495-34dac4923e35","order_by":4,"name":"Jean Marie FRANCOIS","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA9ElEQVRIiWNgGAWjYBAC9gYQycMgxwYT4YNQcji18ByAaDGGaElgYIDqNSaghYEhsYF4LdKHnz1gkNmW3ifd/vAD4w+bxDb2HuMPHxgM8nFq4UszN2DguZ3bJnPGWIIhIS2xjeeMmeQMBgPLBhxa7HkYzCTAWiRyGKT/JBxOBDLMmHkY/hjgtIWH/RtISzqbRPrjHwwQLcaf/zAY4NHCA7YlgU0iAciAaDGQZsCvpQykxRDkHguGtDTjNp5jZZI9Bvi0sG+TYOy5LS8/I/3xDQYbG9l+9ubNH35U4NYCAsx/ezDE8GoAgR+EFIyCUTAKRsGIBgBms0WEgC5KkgAAAABJRU5ErkJggg==","orcid":"","institution":"UMR INSA -CNRS5504 and UMR INSA-INRAE 792","correspondingAuthor":true,"prefix":"","firstName":"Jean","middleName":"Marie","lastName":"FRANCOIS","suffix":""}],"badges":[],"createdAt":"2026-02-09 12:25:45","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8830407/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8830407/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00253-026-13840-y","type":"published","date":"2026-05-01T15:57:52+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":102766273,"identity":"d2b989e3-bf09-419c-9d12-484b4082b3e6","added_by":"auto","created_at":"2026-02-16 11:26:49","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":14178494,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTime course response of XynR and YjhI-based biosensors to D-xylonate\u003c/strong\u003e. \u003cem\u003eE. coli\u003c/em\u003e MG1655 strain was transformed with pBS0, pBS3, pBS7 or pBS8 (see Table 1 \u0026amp; Figure S1 for description of these plasmids). The fluorescence emitted by SYFP2 protein after addition of 1 mM or 20 mM D-xylonate (Xyl1, Xyl20, co means addition of equivalent volume of water) was recorded over 15 h in spectrofluorometer together with OD\u003csub\u003e600\u003c/sub\u003e as described in Material \u0026amp; Methods. Fluorescence intensity was normalized to OD, given arbitrary units (AU). The fold induction (FI) by D-xylonate at 12 h of incubation was calculated by dividing the fluorescence (AU) in response to D-xylonate to that of the control. The data are the mean ± SD (shown by bars on the curves and histograms) of 3 biological replicates.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/e87941f37e2f97a86f469087.png"},{"id":102766287,"identity":"35ea978c-c3ad-4e68-938c-89bfff1907e2","added_by":"auto","created_at":"2026-02-16 11:26:56","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":14569091,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLack of response of XynR and YjhI-based biosensors to D-xylonate in mutants defective in D-xylonate dehydratase. \u003c/strong\u003eSame experiment as in Figure 1 except that \u003cem\u003eE. coli\u003c/em\u003e MG1655 and mutant defective in the D-xylonate dehydratase encoded by \u003cem\u003eyagF\u003c/em\u003e and \u003cem\u003eyjhG \u003c/em\u003ewere transformed with pBS3 or pBS8. The concentration of\u003cstrong\u003e \u003c/strong\u003eD-xylonate was 20 mM. The data reported are the mean ± SD (shown by bar of the curves) of 3 biological replicates.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/711fc34f748a2f3672f219ed.png"},{"id":102766299,"identity":"92cb26a9-c4da-4531-b9a1-9f6fe9d26862","added_by":"auto","created_at":"2026-02-16 11:26:59","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":14886741,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe intermediate 2-keto-3-deoxy-D-xylonate is the metabolite effector of XynR and YjhI transcription factors. \u003c/strong\u003eSame procedure as in Figure 1 except that wild type MG1655 and mutant defective in major pyruvate-dependent aldolase (MGD7) was transformed with pBS3 or pBS8. D-xylonate (Xyl) and 2-keto-3-deoxy-D-xylonate (KDX) were used at 1 mM. The fold induction (FI) given at 12 h of incubation was calculated as in Figure 1. Data are the mean ± SD (shown by bars on the curves and histogram) of three biological replicates.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/04f9f185f4b5ca93773f3954.png"},{"id":102766271,"identity":"7c03c72a-b8d7-40b5-b45b-aec7e6c3cef0","added_by":"auto","created_at":"2026-02-16 11:26:49","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":14746548,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eXynR and YjhI are responsive to 2-oxo-4-hydroxybutyrate and not to 2,4-dihydroxybutyrate. \u003c/strong\u003eSame procedure as in Figure 1 except that wild type MG1655 and mutant defective in membrane-associated lactate dehydrogenase encoded by \u003cem\u003elldD\u003c/em\u003e,\u003cem\u003e dlD\u003c/em\u003e and \u003cem\u003eykgEFG\u003c/em\u003e (MGDLO) was transformed with pBS3 and pBS8. 2,4-dihydroxybutyrate (2,4-DHB) and 2-oxo-4-hydroxybutyrate (OHB) were added at 100 and 1 mM, respectively, and the control was done with 100 mM NH\u003csub\u003e4\u003c/sub\u003eCl as NH\u003csub\u003e4\u003c/sub\u003e\u003csup\u003e+\u003c/sup\u003e is the counterion of 2,4-DHB. The fold induction (FI) given at 12 h of incubation was calculated as in Figure 1. Data are the mean ± SD (shown by bars on the curves and histogram) of three biological replicates.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/dbe238aee4072b243100f2af.png"},{"id":102766281,"identity":"87b7bc5d-79ee-4b2f-88bd-8d6d19e05f8b","added_by":"auto","created_at":"2026-02-16 11:26:50","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":4300203,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eNo cross-talk between XynR- and YjhI-based biosensors in response to OHB and D-xylonate\u003c/strong\u003e. Same procedure as in Figure 1 except that strain defective in either \u003cem\u003exynR\u003c/em\u003e, \u003cem\u003eyjhI\u003c/em\u003e or both \u003cem\u003exynR\u003c/em\u003e and \u003cem\u003eyjhI\u003c/em\u003e were used together with the reference MG1655. Plasmids pBS6 and pBS9 (see description in Figure S2) were inserted in these strains by transformation. The response of the biosensors expressed from pBS6 and pBS9 were measured in the absence (co) or the presence of 1 mM OHB or 20 mM D-xylonate. The fold induction reported in the figure was determined from fluorescence after 12 h of incubation at 37°C. Data are the mean ± SD (shown by bars on the curves and histogram) of three biological replicates. Statistical significance was determined by Student's \u003cem\u003et\u003c/em\u003e-test (*\u003cem\u003ep \u003c/em\u003e\u0026lt; 0.05; **\u003cem\u003ep\u003c/em\u003e\u0026lt; 0.01; ***\u003cem\u003ep\u003c/em\u003e\u0026lt;0.001)\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/35c96b851f73bbe8a2588dc5.png"},{"id":102766278,"identity":"874dd3ae-0da0-41ea-8ccf-f1a6cf9bfe8c","added_by":"auto","created_at":"2026-02-16 11:26:49","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":7804414,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDose-response curves of XynR and YjhI-based biosensors to 2-keto-3-deoxy-D-xylonate (KDX) and 2-oxo-4-hydroxybutyrate (OHB). \u003c/strong\u003eThe strain used for KDX and OHB dose-response was MGD7 transformed with pBS3 and pBS8. The fluorescence was measured by spectrofluorometer over 15 h and values at 12 h after addition of the effectors were used for dose-response analysis, using Graph Pad Prism for data treatment and graph representation. Data are the mean ± SD (shown by bars on the curves) of four biological replicates, each containing 2 technical replicates. Dashed blue lines was obtained from least squares regression smoothing with Graph Pad Prism tool v10.2\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/b6d1dc79ff79cdd1e7835524.png"},{"id":102766274,"identity":"9d638108-14a9-4b1a-b0ea-8a07af4e5bd0","added_by":"auto","created_at":"2026-02-16 11:26:49","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":13203488,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe XynR-based biosensor is specific to OHB. \u003c/strong\u003eIn (A) is reported the fluorescence measured after 12 h incubation of the strain MGD7 transformed with pBS3 in the presence of various organic and amino acids added at 10 mM except for OHB (1 mM). (B) is the same as (A) except that different sugar acids were added at 20 mM. Dashed line represented the induction fold of control which is 1.0. Data are the mean ± SD (shown by bar on the histogram) of three biological replicates.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/8e71fef8e4995d05cb52ee64.png"},{"id":102766276,"identity":"87cc58c0-038f-4505-a6d2-7f636afd4de7","added_by":"auto","created_at":"2026-02-16 11:26:49","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":10797368,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUse of the YjhI-based biosensor to screen for transaminase in OHB formation. \u003c/strong\u003eThe strain MGDT (Table 1) was co-transformed with pBS8 and with pETM-empty (grey histograms), pETM expressing wild type \u003cem\u003eAlaC\u003c/em\u003e (pETM_alaC; orange histograms) or a variant AlaC\u003csup\u003eA142P Y245D\u003c/sup\u003e (pETM_alaC**; green histograms). After pre-culture in LB with appropriate antibiotic, IPTG was added to part of the culture when reached OD\u003csub\u003e600\u003c/sub\u003e 0.6 and incubated for another 4 h before collecting and resuspension in PBS buffer as described in Material and Methods. After addition of the various compounds at 5mM, except OHB, which was used as the positive control and added at 1 mM, the fluorescence was monitored for 16 h. Data reported are the fold induction determined after 12 h of incubation at 37°C and value reported is the mean ± SD (shown by bars on the histogram) of four biological replicates. Abbreviation: OHB = 2-oxo-4-hydroxybutyrate; Pyr = pyruvate; HMS = L-homoserine; a-KG = alpha ketoglutarate\u003c/p\u003e","description":"","filename":"Figure8.png","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/17a6ba2782e3ae4b6b70eafe.png"},{"id":102766297,"identity":"943067a9-a080-4370-991c-ad2657943e45","added_by":"auto","created_at":"2026-02-16 11:26:57","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":3546952,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUse of the XynR-based biosensor to screen for threonate dehydratase catalysing the formation of OHB. \u003c/strong\u003eThe strain MGD\u003cem\u003e7\u003c/em\u003e was co-transformed with pBS3 and pZS2-kdgT or pZS2-kdgT \u003cem\u003eHh\u003c/em\u003earaD which encodes the threonate dehydratase from \u003cem\u003eHerbaspirillum huttiense\u003c/em\u003e. Induction of these genes and incubation of the cells with D-threonate (10 mM) are described in Material \u0026amp; Methods. The fluorescence was determined by flow cytometry after 12 h incubation. Data are the mean ± SD (shown by bar of the histogram) of two biological replicates, each containing 3 technical replicates.\u003c/p\u003e","description":"","filename":"Figure9.png","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/4a80a1b445864832cc32137b.png"},{"id":102766272,"identity":"24c00a85-c5c6-4d06-a08c-305a705f4045","added_by":"auto","created_at":"2026-02-16 11:26:49","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":2993548,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eScreening of the Keio mutants of \u003c/strong\u003e\u003cem\u003e\u003cstrong\u003eE. coli\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e for 2,4-DHB transporter using XynR transcription factor biosensor. \u003c/strong\u003eThe screen of the Keio collection from Baba \u003cem\u003eet al\u003c/em\u003e. (Baba et al. 2006) was restricted to about 200 hundred mutants with deletion in a gene encoding a known or putative organic acid or amino acid transporter. They were transformed with pBS5. The fluorescence was determined by flow cytometry after 24 h of incubation in the presence of 10 or 100 mM 2,4-DHB. The fluorescence intensity was determined on 100,000 individual cells (events) for each mutant strain and divided by the fluorescence intensity measured in the reference BW25113. The data were then expressed as the fluorescence ratio and the bars on the histograms showed the SD from three independent experiments.\u003c/p\u003e","description":"","filename":"Figure10.png","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/bd33345b651a2482f0aa2b28.png"},{"id":102766258,"identity":"0918a29d-d650-40fe-be30-4e0c035dc3ef","added_by":"auto","created_at":"2026-02-16 11:26:46","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1274286,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/299b8c53-f60d-4e60-b558-55565923ad9e.pdf"},{"id":102766275,"identity":"a5e9b26c-8887-4a83-a0f5-0e4cc6aaac7c","added_by":"auto","created_at":"2026-02-16 11:26:49","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":1333478,"visible":true,"origin":"","legend":"","description":"","filename":"OHbiosensorSupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-8830407/v1/040f274f019dd18e9fb74ebc.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Identification of 2-keto-3-deoxy-D-xylonate and 2-oxo-4-hydroxybutyrate as natural and artificial effectors of the transcription factors XynR and YjhI, respectively, and application to the development of biosensors for the bioproduction of 2,4-dihydroxybutyric acid in E. coli","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eThe Dahms pathway enables the assimilation of xylose into glycoaldehyde and pyruvate involving four enzymatic steps (Dahms \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e1974\u003c/span\u003e). The first reaction in this pathway is the oxidation of xylose to D-xylonate by a xylose dehydrogenase. While this enzyme does not exist in \u003cem\u003eE. coli\u003c/em\u003e, the remaining enzymatic steps are encoded by two distinct operons, namely \u003cem\u003eyagEFGH\u003c/em\u003e and \u003cem\u003eyjhIGHF\u003c/em\u003e that are present on two cryptic phages, CP4-6 and KpL12-prophages, respectively, in the genome of this bacterium. Genetic and biochemical data have shown that \u003cem\u003eyagF\u003c/em\u003e and \u003cem\u003eyihG\u003c/em\u003e encode a xylonate dehydratase that converts D-xylonate to 2-keto-3-deoxy-D-xylonate (KDX). This enzymatic step is followed by an aldolic cleavage of KDX into glycoaldehyde and pyruvate by a pyruvate-dependent aldolase encoded by \u003cem\u003eyagE\u003c/em\u003e and \u003cem\u003eyjhH\u003c/em\u003e. On the other hand, D-xylonate uptake is assumed to be taken up by a transporter encoded by \u003cem\u003eyagG\u003c/em\u003e and \u003cem\u003eYjhF\u003c/em\u003e (reviewed in (Banares et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), see Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). It has been reported that these two operons are induced by D-xylonate through the action of a transcription factor apparently specific for each operon. Shimada \u003cem\u003eet al\u003c/em\u003e. (Shimada et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) showed that YagI, renamed XynR, which is also encoded by the CP4-6 prophage acts as a repressor of the \u003cem\u003eyag\u003c/em\u003e operon, whereas Banares \u003cem\u003eet al.\u003c/em\u003e (Banares et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) demonstrated that YjhI, which is present on KpL12-phages-like elements, very likely works as an activator of the \u003cem\u003eyjh\u003c/em\u003e operon, which also includes its autoinduction. However, it remained elusive whether D-xylonate is the direct effector of these two transcription factors. Two types of experimental data suggested that the effector of XynR and YjhI is likely not D-xylonate. On the one hand, the concentration of D-xylonate required to release the binding of XynR to DNA or to activate YjhI-dependent genes is in the range of 10 mM to 20 mM (Banares et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Shimada et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), which is very high with respect to micro to millimolar concentration of most of metabolite effectors (Hanko et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). On the other hand, transcriptomic analyses of \u003cem\u003eE. coli\u003c/em\u003e challenged with the non-natural platform molecule 2,4-dihydroxybutyric acid (2,4-DHB) triggered a strong upregulation of the D-xylonate catabolic genes (our unpublished data). Since metabolic pathways designed to produce 2,4-DHB are different to the catabolic pathway of D-xylonate, these data suggested that the effector of XynR and YjhI could be a metabolite derived from D-xylonate and that this metabolite may be close to 2,4-DHB or its proximal intermediate 2-oxo-4-hydroxybutyrate (OHB). To answer these questions, we designed a biosensor consisting of a reporter gene encoding a fluorescent protein dependent on the promoter of a gene under the control of XynR or YjhI. With this tool, and combined with mutants defective for D-xylonate catabolic pathway, we could on the one hand investigate the true effector of these two transcription factors and on the other hand, examine whether these TFs are responsive to 2,4-DHB or an intermediate in the production pathway of this molecule, in order to repurpose them as biosensors tool for improving the flux production of this non-natural platform molecule.\u003c/p\u003e \u003cp\u003eThe motivation for developing such biosensors sensitive to either 2,4-DHB or its proximal intermediate, 2-oxo-4-hydroxybutyrate (OHB), stems from the still low performance in terms of yield and productivity of the E. coli strains equipped with synthetic pathways designed for the bioproduction of this platform molecule (2,4-DHB) from glucose (Walther et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Walther et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) or from C1/C2 carbon source (Fraz\u0026atilde;o et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Enzymes in these non-native pathways have been engineered to catalyse reaction on non-natural substrates, leading to non-natural intermediate OHB, which is produced either from L-homoserine by the action of promiscuous transaminases (Bouzon et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Walther et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) or from D-threonate by D-threonate dehydratase from \u003cem\u003eHerbaspirillum huttiense\u003c/em\u003e. Clearly, these enzymes are rate limiting in the production pathways of 2,4-DHB because they harbor a catalytic efficiency (k\u003csub\u003ecat\u003c/sub\u003e/K\u003csub\u003eM\u003c/sub\u003e) at least 100-fold lower than most of the other enzymes needed for the bioproduction. Thus, directed enzyme evolution is required to optimise their catalytic efficiency, which relies on a high throughput screening method. Transcription factor-based metabolite biosensors would be well appropriate as long as OHB or 2,4-DHB can be recognized as a target of a transcription factor already existing in \u003cem\u003eE. coli\u003c/em\u003e. Since nature has developed a wide range of sensors using natural endogenous metabolites that enable bacteria to adapt to their environment, the use of TF-based biosensors becomes challenging when it comes to optimizing a purely synthetic pathway whose intermediates and/or end product are completely unknown to the microorganism. In this work, we demonstrated that the natural effector, and therefore the potential ligand, of XynR and YjhI is the intermediate metabolite KDX. We validated that both transcription factors can be repurposed as biosensors to establish high throughput screening strategy aiming at improving the flux production of the non-natural platform molecule 2,4-DHB.\u003c/p\u003e"},{"header":"MATERIAL \u0026 METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eChemicals and reagents\u003c/h2\u003e \u003cp\u003eAll chemicals and solvents were purchased from Sigma-Aldrich unless otherwise stated. The 2,4\u0026ndash;dihydroxybutyric acid (racemic form, ammonium salt) was a kind gift from Adisseo and was \u0026gt;\u0026thinsp;95% pure as determined by HPLC. The 2-oxo-4-hydroxybutyrate (OHB) was synthesized by incubating 125 mM D-homoserine with 4.5 U/mL porcine kidney D-amino acid oxidase, 4000 U/mL beef liver catalase in 100 mM Tris-HCl buffer (pH 8) for at least 2 h at 37\u0026deg;C. These enzymes were purchased from Sigma-Aldrich. The reaction was filtered through AmiconTM Ultra filters (10 kDa threshold, Millipore). OHB was quantified by NMR using a Bruker Avance III HD 800 MHz spectrometer equipped with a cryogenically cooled 5 mm QCI-P (H/P-C/N/D) quadruple resonance probe. The spectra were acquired and reprocessed with Bruker Topspin 4.1 software, and were calibrated for reference to the frequency of TSP (TrimethylSilylPropanoic acid) on the 1H spectrum at 0 ppm. The solution, giving a yield of conversion of D-homoserine ranging from 77 to 88%, was stored at -80\u0026deg;C in small aliquots, which were used only once after thawing.\u003c/p\u003e \u003cp\u003eRestriction endonucleases, Phusion polymerase, restriction enzymes, HiT4 DNA ligase, and HiFi DNA assembly were purchased from New England Biolabs and used according to instructions of the manufacturer. Moreover, In-Fusion\u0026reg; HD Cloning Kit was purchased from TaKaRa-Clontech. DNA plasmid isolation was performed using GeneJET Plasmid Miniprep Kit (Thermo Scientific). DNA extraction from agarose gel was carried out using the GeneJET Gel Extraction Kit (Thermo Scientific). DNA sequencing was carried out by Eurofins SAS (Ebersberg, Germany).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eStrains construction and growth conditions\u003c/h3\u003e\n\u003cp\u003eThe strain \u003cem\u003eE. coli\u003c/em\u003e K-12 substr. MG1655 (ATCC 47076) was used as the parental strain for all constructions in this study. The strain MGΔ7 was constructed by deleting the 6 genes, namely \u003cem\u003eeda, yfaU, garL, yagE, yjjH\u003c/em\u003e and \u003cem\u003edgoA\u003c/em\u003e encoding pyruvate-dependent aldolase by the phage transduction method adapted from Miller (Miller \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1992\u003c/span\u003e), whereas \u003cem\u003emhpE\u003c/em\u003e was deleted using CRISPR-Cas9 method developed by Jiang et al (Jiang et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2015a\u003c/span\u003e). In transduction procedure, donor strains were taken from the Keio collection (Baba et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), which contains mutants where a single gene is deleted and replaced by a kanamycin resistance cassette flanked by FRT sites. This method allowed the construction of simple, double and etc. mutants, since all target genes were separated by at least 100 kb. Likewise, MG1655 was deleted of its membrane-associated lactate oxidoreductase encoded by \u003cem\u003elldD\u003c/em\u003e, \u003cem\u003edld\u003c/em\u003e and \u003cem\u003eykfEGF\u003c/em\u003e by transduction giving rise to MGΔLO, as described elsewhere (Malfoy et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Positive clones were selected on LB agar plates containing kanamycin (50 \u0026micro;g.mL\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and verified by PCR analysis. The Kan cassette was removed from the genome by expressing FLP recombinase from the pCP20 plasmid (Cherepanov and Wackernagel \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e1995\u003c/span\u003e) and correct excision of the cassette was verified by PCR using locus specific primers (\u003cb\u003eTable \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003c/b\u003e). Other strain deletions, such as MG1655 \u003cem\u003e∆xynR\u003c/em\u003e, MG1655 \u003cem\u003e∆yjhI\u003c/em\u003e, MG1655 \u003cem\u003e∆xynR ∆yjhI\u003c/em\u003e etc., as reported in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e were also carried out by P1 phage transduction according to (Miller \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e1992\u003c/span\u003e). Strain MGΔ4 defective in the aspartate-homoserine pathway (\u003cem\u003e∆thrA\u003c/em\u003e, \u003cem\u003e∆metL\u003c/em\u003e, \u003cem\u003e∆asd, ∆lysC\u003c/em\u003e) was created as reported elsewhere (Malfoy et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). To obtain the mutant strain MGDT which was deleted for four transaminases (\u003cem\u003ealaC\u003c/em\u003e, \u003cem\u003etyrB\u003c/em\u003e, \u003cem\u003eybdL\u003c/em\u003e, \u003cem\u003eilvE\u003c/em\u003e) and \u003cem\u003easd\u003c/em\u003e encoding the aspartate semi-aldehyde dehydrogenase, the transaminases genes were consecutively deleted in MG1655 by transduction (Thomason et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Then, \u003cem\u003easd\u003c/em\u003e gene was deleted in the obtained mutant MG1655 \u003cem\u003e∆alaC::FRT ∆tyrB::FRT ∆ybdL::FRT ∆ilvE::FRT\u003c/em\u003e by CRISPR-Cas9 method. This mutant was first transformed with pCas9. Guide RNA was expressed from a pTargetF-asd (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Overnight cultures grown at 30\u0026deg;C were diluted to an OD₆₀₀ of 0.05 in LB containing kanamycin and arabinose (10 mM) and grown to OD₆₀₀ \u0026asymp; 0.6. Cells were chilled on ice for 15 min, harvested at 5,000 \u0026times; \u003cem\u003eg\u003c/em\u003e for 5 min at 4\u0026deg;C, washed twice with ice-cold sterile water and once with 10% glycerol, and resuspended in 100 \u0026micro;L 10% glycerol. Competent cells were electroporated according to (Wang et al. \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) with 100 ng pTarget and 400 ng donor DNA (Eurofins Genomics) using a Gene Pulser Xcell (Bio-Rad) at 2.5 kV in a 2-mm cuvette. Cells were recovered in SOC medium for 1\u0026ndash;2 h at 30\u0026deg;C and plated on LB agar containing kanamycin and spectinomycin. Genome edits were verified by PCR and sequencing, and pTarget and pCas9 were sequentially cured from positive clones.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eStrains used in this study\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eStrain name\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDescription -genotype\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003esource\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMG1655\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eF- lambda- \u003cem\u003eilvG- rfb-50 rph-1\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eATCC n\u0026deg;47046\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNEB\u0026reg;5α\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003efhuA2 Δ(argF-lacZ)U169 phoA glnV44 Φ80 Δ(lacZ)M15 gyrA96 recA1 relA1 endA1 thi-1 hsdR17\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNEB\u003c/p\u003e \u003cp\u003ecompetent E. coli\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBL21 (DE3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003efhuA2 [lon] ompT gal (λ DE3) [dcm] ΔhsdS\u003c/em\u003e\u003c/p\u003e \u003cp\u003e\u003cem\u003eλ DE3\u0026thinsp;=\u0026thinsp;λ sBamHIo ΔEcoRI-B int::(lacI::PlacUV5::T7 gene1) i21 Δnin5\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNEB\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBW25113\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cem\u003eΔ(araD-araB)567 ΔlacZ4787(::rrnB-3) λ- rph-1 Δ(rhaD-rhaB)568 hsdR514\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCGSC#7636 (Baba et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2006\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGΔ7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔyagE::FRT ΔyjhH::FRT Δeda::FRT ΔdgoA::FRT ΔyfaU::FRT ΔgarL::FRT ΔmhpE\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGΔLO\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔdld::FRT ΔykgEFG::FRT ΔlldD::FRT\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGDT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔybdl ΔtyrB ΔilvE ΔalaC\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGΔ4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔlysC ΔthrA Δasd ΔmetL\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGΔF\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔyagF\u0026nbsp;::FRT\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGΔG\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔyghG\u0026nbsp;::FRT\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGΔFG\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔyagF\u0026nbsp;::FRT ΔyjhG\u0026nbsp;::FRT\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGΔX\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔxynR\u0026nbsp;::FRT\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGΔJ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔyjhI\u0026nbsp;::FRT\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGΔXJ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔxynR ΔyjhI\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGΔ7 ΔXJ\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔxynR\u0026nbsp;::FRT ΔyjhI\u0026nbsp;::FRT ΔyagE::FRT ΔyjhH::FRT Δeda::FRT ΔdgoA::FRT ΔyfaU::FRT ΔgarL::FRT ΔmhpE\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMGDTDA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMG1655 \u003cem\u003eΔybdl ΔtyrB ΔilvE ΔalaC Dasd\u003c/em\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eUnless otherwise stated, \u003cem\u003eE. coli\u003c/em\u003e strains were cultivated in Lysogeny Broth (LB) medium at 37\u0026deg;C on a rotary shaker running at 200 rpm. The antibiotics kanamycin, spectinomicine or chloramphenicol were added when required at concentrations of 50, 100 and 25 mg.L\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, respectively.\u003c/p\u003e\n\u003ch3\u003eConstruction of the metabolite-based transcription factor biosensors\u003c/h3\u003e\n\u003cp\u003ePlasmids used in this study are reported in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e while primers to make the various constructs can be downloaded from Table S3 in Supplementary data. Scheme of the constructed biosensor plasmids are reported in \u003cb\u003eFigure S2\u003c/b\u003e. To generate pBS0, a 200 bp fragment starting to -200 bp to -1 bp before ATG codon of \u003cem\u003eyagE\u003c/em\u003e was amplified from MG1655 genomic DNA using primers TM7 and TM8, adding \u003cem\u003eXhoI\u003c/em\u003e and \u003cem\u003eHindIII\u003c/em\u003e restriction sites on 5\u0026rsquo; and 3\u0026rsquo; ends, respectively. The PCR product was ligated with HiT4 DNA ligase into gel purified XhoI-HindIII digested pREP22 vector from (Frazao et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) that carries \u003cem\u003esyfp2\u003c/em\u003e gene encoding the super yellow fluorescent protein (Kremers et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), which also carries the optimised RBS already present on the pREP22 plasmid between the promoting region of \u003cem\u003eyagE\u003c/em\u003e and the \u003cem\u003esyfp2\u003c/em\u003e gene. To yield pBS2, the \u003cem\u003exynR\u003c/em\u003e gene was then inserted into pBS0 as follows. The coding sequence of this gene (759bp) along with 190 bp upstream the ATG codon was amplified from MG1655 strain using primers TM4 and TM11. Meanwhile, T7 terminator was amplified from pET-28a with TM5 and TM14. These primers allowed insertion of flanking regions corresponding to linearized pBS1 carried out with TM12 and TM13 primers. These three fragments were gel purified and incubated with HiFi DNA assembly mix to generate pBS2 plasmid, which was validated by sequencing after amplification in NEB-5α competent \u003cem\u003eE. coli\u003c/em\u003e cells and purification. To generate pBS3, primers TM71 and TM146 were used to amplify the \u003cem\u003esyfP2-xynR\u003c/em\u003e cassette of pBS2 plasmid whereas medium copy vector pZA33 was amplified with primers JFR64 and JFR65. Both fragments were gel-purified and ligated using NEBuilder HiFi DNA assembly. Validation of pBS3 was done by sequencing using primers JFR66/JF67 and TM10/TM16. To obtain pBS5, \u003cem\u003elldD\u003c/em\u003e gene from the genome of MG1655 was amplified with primers TM67/68, λ tL3 terminator of Cas9 was amplified from pCas plasmid (Jiang et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2015a\u003c/span\u003e) using primers TM69 and TM70 and pBS2 using primers TM71 and TM72. All these fragments were gel purified and assembled with HiFi DNA assembly, to form pBS5. In addition, a DNA fragment that contains pJ23106 promoter (ordered from Eurofins genomics, France) and Shine-Dalgarno sequence was obtained by mixing primers TM63 and TM64 at 70\u0026deg;C, followed by slow decrease of temperature and then amplified with primers TM65 and TM66. After purification of the different fragments on agarose gel, they were assembled by the NEBuilder HiFi DNA Assembly (Biolabs) to generate pBS5, which was validated by sequencing using primers TM9/TM10/TM15/TM16 after amplification in NEB-5α competent \u003cem\u003eE. coli\u003c/em\u003e cells and purification.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePlasmids constructed and used in this study\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePlasmid\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eShort-Description\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c6\" namest=\"c3\"\u003e \u003cp\u003esource\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"1\" nameend=\"c7\" namest=\"c7\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epET28a\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003eNovagen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c7\" namest=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epZA33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eClm\u003csup\u003eR\u003c/sup\u003e, p15A ori\u0026nbsp;; promoter P\u003csub\u003eA1\u003c/sub\u003e lacO1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eExpressys\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epREP22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epZE13 derivative, promoter P\u003csub\u003eyqhD\u003c/sub\u003e hybrid: RBS\u003csub\u003eO1\u003c/sub\u003e:syfp2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003e(Frazao et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2018\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epCP20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAmp\u003csup\u003eR\u003c/sup\u003e, Clm\u003csup\u003eR\u003c/sup\u003e, pSC101 ori\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003e(Cherepanov and Wackernagel \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e1995\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epBS0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epZE13 derivative, promoter P\u003csub\u003eyagE\u003c/sub\u003e\u0026nbsp;: RBS\u003csub\u003eO1\u003c/sub\u003e::syfp2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epBS2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epZE13 derivative, promoter P\u003csub\u003eyagE\u003c/sub\u003e\u0026nbsp;: RBS\u003csub\u003eO1\u003c/sub\u003e::syfp2, promoter P\u003csub\u003exynr\u003c/sub\u003e:RBS\u003csub\u003exynr\u003c/sub\u003e: xynR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epBS3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epZA33 derivative, promoter P\u003csub\u003eyagE\u003c/sub\u003e\u0026nbsp;: RBS\u003csub\u003eO1\u003c/sub\u003e::syfp2, promoter P\u003csub\u003exynr\u003c/sub\u003e:RBS\u003csub\u003exynr\u003c/sub\u003e: xynR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epBS5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epZE13 derivative, promoter P\u003csub\u003eyagE\u003c/sub\u003e\u0026nbsp;: RBS\u003csub\u003eO1\u003c/sub\u003e::syfp2, promoter P\u003csub\u003exynr\u003c/sub\u003e:RBS\u003csub\u003exynr\u003c/sub\u003e: xynR; promoter BBa_J23106:lldD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epBS7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epZE13 derivative, promoter P\u003csub\u003eyjhi\u003c/sub\u003e\u0026nbsp;: RBS\u003csub\u003eO1\u003c/sub\u003e::syfp2,\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epBS8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epZE13 derivative, promoter P\u003csub\u003eyjh\u003c/sub\u003e\u0026nbsp;: RBS\u003csub\u003eO1\u003c/sub\u003e::syfp2, promoter P\u003csub\u003eyjhI\u003c/sub\u003e:RBS\u003csub\u003eyjhI\u003c/sub\u003e: yjhI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epBS6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epZE13 derivative, promoter P\u003csub\u003eyagE\u003c/sub\u003e\u0026nbsp;: RBS\u003csub\u003eO1\u003c/sub\u003e::syfp2, promoter P\u003csub\u003eyjhI\u003c/sub\u003e:RBS\u003csub\u003eyjhI\u003c/sub\u003e: yjhI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epBS9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epZE13 derivative, promoter P\u003csub\u003eyjhI\u003c/sub\u003e : RBS\u003csub\u003eO1\u003c/sub\u003e::syfp2, promoter P\u003csub\u003exynr\u003c/sub\u003e:RBS\u003csub\u003exynR\u003c/sub\u003e: xynR\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epETM_empty\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epET28a with T5 promoter_empty\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epETM_alaC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epET28a with T5 promoter carrying wild type alaC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003epETM_alaC**\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003epET28a with T5 promoter carrying mutated alaC expressing AlaC\u003csup\u003eA142P Y275D\u003c/sup\u003e variant\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eThis study\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eIn-Fusion assembly protocol was used to construct the plasmids pBS0, pBS7, pBS8, pBS6 and pBS9 (see scheme of these plasmids in Figure S2 in supplementary material). Whenever plasmids were used as templates for PCR amplification, the PCR products were treated with DpnI (New England Biolabs) to eliminate the template plasmid. Plasmid pBS3 was used as the template to generate pBS0, pBS6, and pBS9. Biosensor pBS0 was obtained by first removing the transcriptional factor \u003cem\u003exynR\u003c/em\u003e and the reporter gene \u003cem\u003esyfp2\u003c/em\u003e with the \u003cem\u003eyagE\u003c/em\u003e promoter from pBS3 by PCR (primers 3135\u0026thinsp;+\u0026thinsp;3136) and then reinsertion of the SYFP2 cassette as follows. The \u003cem\u003esyfp2\u003c/em\u003e gene with the \u003cem\u003eyagE\u003c/em\u003e promoter and its corresponding T7 terminator was amplified by PCR (primers 3131\u0026thinsp;+\u0026thinsp;3132). After PCR purification, a second amplification was performed to create 15 bp homology regions with the first template DNA. Following the In-Fusion assembly reaction, each plasmid construct was transformed into competent \u003cem\u003eE. coli\u003c/em\u003e NEB-5α cells according to the manufacturer\u0026rsquo;s instructions. Transformed cells were plated on LB agar supplemented with chloramphenicol (25 \u0026micro;g/mL final concentration). Colonies were initially screened by PCR, and plasmids were extracted from positive clones. Two confirmed constructs were sent for full plasmid sequencing (Eurofins Genomics). These procedures were repeated whenever new plasmid constructs were obtained. Plasmid pBS6 was obtained by replacing the transcriptional factor XynR with YjhI in pBS3. For this purpose, chromosomal DNA from \u003cem\u003eE. coli\u003c/em\u003e MG1655 was used to amplify \u003cem\u003eyjhI\u003c/em\u003e, including 199 bp upstream of the ORF to cover its native promoter. A T7 terminator was added 49 bp downstream of \u003cem\u003eyjhI\u003c/em\u003e during the same PCR reaction (primers 3130\u0026thinsp;+\u0026thinsp;3137). The purified product was then subjected to a second PCR to add 15 bp homology regions (primers 3128\u0026thinsp;+\u0026thinsp;3129). In parallel, pBS3 was used to amplify the plasmid backbone lacking \u003cem\u003exynR\u003c/em\u003e and its promoter region (primers 3126\u0026thinsp;+\u0026thinsp;3127). The plasmid backbone and the amplified \u003cem\u003eyjhI\u003c/em\u003e fragment were assembled using the In-Fusion reaction to generate biosensor pBS6. Plasmids pBS8 and pBS9 were constructed using the same approach, starting from pBS6 and pBS3, respectively. In pBS6, the \u003cem\u003eyagE\u003c/em\u003e promoter of the reporter gene was replaced with the \u003cem\u003eyjhI\u003c/em\u003e promoter to generate pBS8. Similarly, replacement of the \u003cem\u003eyagE\u003c/em\u003e promoter with the \u003cem\u003eyjhI\u003c/em\u003e promoter in pBS3 yielded pBS9. Finally, pBS7 was constructed by overlap PCR using pBS8 as the template and overlapping primers 3154\u0026thinsp;+\u0026thinsp;3155, which removed the transcriptional factor \u003cem\u003eyjhI\u003c/em\u003e and its promoter. The primer 3154_FW can anneal to both the upstream and downstream regions of P\u003csub\u003e\u003cem\u003ejjhI\u003c/em\u003e\u003c/sub\u003e+\u003cem\u003eyjhI\u003c/em\u003e and its corresponding T7 terminator, respectively, in order to eliminate P\u003csub\u003e\u003cem\u003ejjhI\u003c/em\u003e\u003c/sub\u003e+\u003cem\u003eyjhI\u003c/em\u003e. The reverse primer 3155 was designed to contain a 50-bp homology region corresponding to the terminal sequence of primer 3154_FW. PCR was performed using pBS8 as the template with these primers. Following DpnI treatment and PCR clean-up, the product was transformed into \u003cem\u003eE. coli\u003c/em\u003e NEB-5α cells. After PCR verification, positive clones were sent for whole-plasmid sequencing. The correct plasmids were kept at -20\u0026deg;C for further experiments.\u003c/p\u003e \u003cp\u003eFor the construction of pBios-JF2 through JF6, pBS3 was used as the backbone. The primers seq-R and seq-F were used on pBS3 to yield after a DNA fragment without the promoter sequence of \u003cem\u003exynR\u003c/em\u003e. Then, the insertion of new promoter sequences bearing a strong RBS\u003csub\u003e01\u003c/sub\u003e calculated according to Salis \u003cem\u003eet al.\u003c/em\u003e (Salis \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) was carried out by assembling this linearized fragment with PCR amplified synthetic fragments (\u003cb\u003esee Table S2\u003c/b\u003e) obtained with primer seq-xynR_F on the 5\u0026rsquo; end and one of the 5 other primers listed in \u003cb\u003eTable S2\u003c/b\u003e on 3\u0026rsquo;-end using the NEBuilder HiFi DNA Assembly (Biolabs). All the constructs were validated by sequencing.\u003c/p\u003e \u003cp\u003e \u003cb\u003eConstruction of plasmids for screening assays of OHB-producer enzymes.\u003c/b\u003e \u003c/p\u003e \u003cp\u003ePlasmid pZS2-aspC-kdgT-\u003cem\u003eHh.\u003c/em\u003earaD (Fraz\u0026atilde;o et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) was used as a template for reverse PCR using primers TM250 \u0026amp; TM251 for deleting \u003cem\u003easpC\u003c/em\u003e. Resulting plasmid, pZS2-kdgT-\u003cem\u003eHh.\u003c/em\u003earaD only carried genes \u003cem\u003ekdgT\u003c/em\u003e (\u003cem\u003eE. coli\u003c/em\u003e) and \u003cem\u003earaD\u003c/em\u003e (\u003cem\u003eHerbaspirillum hutiense\u003c/em\u003e) coding respectively a D-threonate transporter and a D-arabinonate dehydratase, previously characterised as D-threonate dehydratase for \u003cem\u003ein vivo\u003c/em\u003e production of OHB (Fraz\u0026atilde;o et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)). A control plasmid was also obtained from the latter by reverse PCR using primers TM252 \u0026amp; TM253 in order to remove \u003cem\u003earaD\u003c/em\u003e gene. The T7-inducible promoter of \u003cem\u003ealaC\u003c/em\u003e gene and its mutant \u003cem\u003ealaC**\u003c/em\u003e, encoding the variant AlaC\u003csup\u003eA142P Y275D\u003c/sup\u003e (Bouzon et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), initially cloned in pET28a were replaced by T5 promoter to allow IPTG induction in MG1655 strain. To this end, pET28a was amplified using the primers TM233 and TM234, while T5 promoter was amplified from the PCA24N vector from the ASKA library (Kitagawa et al. \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2005\u003c/span\u003e) using primers TM242 and TM243. Both fragments displayed a floating tail of 10 nt on each side that are homologous to the ends of the other amplicon. The T5 promoter was cloned using the HiFi DNA Assembly Master Mix from NEB, following the manufacturer guidelines. The resulting vector was named pETM28, carrying the T5 promoter upstream of the coding sequence. Wild type \u003cem\u003ealaC\u003c/em\u003e and mutant \u003cem\u003ealaC**\u003c/em\u003e were amplified using the primers 917, bearing a \u003cem\u003eSac\u003c/em\u003eI site and 926, bearing a \u003cem\u003eHind\u003c/em\u003eIII site. pET28M and the inserts were digested with \u003cem\u003eSac\u003c/em\u003eI and \u003cem\u003eHind\u003c/em\u003eIII, gel purified and ligated using the HiT4 DNA ligase (NEB), following the supplier recommendations. The constructed plasmids pET28M_alaC and pET28M_alaC** were validated by sequencing.\u003c/p\u003e\n\u003ch3\u003eIn vivo fluorescence experiments\u003c/h3\u003e\n\u003cp\u003eUnless otherwise stated, bacterial culture for \u003cem\u003ein vivo\u003c/em\u003e fluorescence experiments were performed as follows. At day 1, a preculture of the bacterial strain was carried out in 2 to 5 ml LB medium with appropriate antibiotic in 50 ml falcon tube at 37\u0026deg;C in rotary shaker set at 200 rpm. The next day, the cultures were diluted in 5 ml of either LB or in M9-MOPS mineral medium containing 4 g/L xylose with the appropriate antibiotics in 50 ml falcon tubes at an initial OD\u003csub\u003e600\u003c/sub\u003e of 0.05. After 3 h and half for LB or 6 h for M9/MOPS, the cultures were collected by centrifugation at 3250 g for 15 minutes at room temperature, washed once with 5 ml PBS buffer, centrifugated again and then resuspended, unless otherwise stated, in 5 ml of PBS. Three hundred \u0026micro;l of the culture were delivered in 48-well microtiter plates and each culture was made in triplicate with independent bacterial clones. After 5 min incubation at 37\u0026deg;C, metabolites/effectors as indicated in corresponding figures were added. Growth at 600 nm and fluorescence of the SYFP2 protein were measured over a time period of 16 h either using a spectrofluorometer Clario Star from GmbH Labtech (excitation and emission wavelength set at 515 nm and 527 nm, respectively) or with a Biotek Synergy HTX from Agilent/Thermo Scientific using excitation filter at 485 and emission filter at 525 nm. The fluorescence expressed in arbitrary unit (AU) corresponds to the absolute fluorescence divided by OD measured at 600 nm. Fluorescence was also determined by flow cytometry using a BD Accuri\u0026trade; C6 Plus Personal Flow Cytometer (BD Biosciences) with excitation set at 488 nm. Forward-scatter characteristics (FSC) and side-scatter characteristics (SSC) were detected as small-angle and large-angle scatters of the 488 nm laser, respectively. SYFP2 fluorescence was detected using a 530/30 nm (channel FL1) band-pass filter set. A total of 100,000 events were recorded per sample, and electronic gating was applied on the densest subset of cells based on forward- versus side-scatter height. The same gate was used to estimate median levels of SYFP2 fluorescence.\u003c/p\u003e \u003cp\u003eThe protocol was slightly modified for the screen of transaminases as the gene encoding these enzymes were cloned in a plasmid requiring their induction by IPTG. Therefore, after overnight culture in LB with appropriate antibiotics, the cells were reinoculated in LB with the appropriate antibiotic at OD\u003csub\u003e600\u003c/sub\u003e 0.05, cultivated for another 2 h, after which 0.5 mM IPTG was added and growth in an incubator set at 37\u0026deg;C and 200 rpm was continued for 4 h. Cells were collected by centrifugation, washed once with PBS and resuspended in PBS as above. For threonate dehydratase no IPTG was added since the expression of the gene was constitutive and after overnight culture in LB, the culture was washed twice with PBS, resuspended in PBS at OD\u003csub\u003e600\u003c/sub\u003e \u0026asymp;1.0 to which was added D-threonate (10 mM), OHB (1 mM) or water. The fluorescence/growth was monitored as described above using the Biotek Synergy HTX fluorometer or by flow cytometer.\u003c/p\u003e\n\u003ch3\u003eData processing and statistical analysis\u003c/h3\u003e\n\u003cp\u003eFor each data sets, absolute fluorescence intensity and OD\u003csub\u003e600\u003c/sub\u003e were recorded and the relative fluorescence in AU was obtained by dividing absolute value to OD. All experiments were done at least with three biological triplicates. Excel tools were used to calculate the mean, standard deviation and covariance (CV). Determination of the response threshold of the biosensor to its metabolites was obtained by fitting a nonlinear regression curve of the fluorescence intensity versus the concentration of the substrate using Solver tool in Excel or Graph Pad Prism V10.6.1\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cp\u003e \u003cb\u003eConstruction of the XynR and YjhI-based biosensor and investigation of their responses to D-xylonate.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eAccording to previous works, the transcription factor encoded by \u003cem\u003exynR\u003c/em\u003e acts as a repressor of D-xylonate metabolic genes present in the cryptic CP4-6 prophage whereas YjhI transcription factor is an activator of similar genes that belong to the KpLE12 phage-like elements (\u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003c/b\u003e). To confirm these data, we build four types of biosensors on a medium copy plasmid, using \u003cem\u003esyfp2\u003c/em\u003e encoding the super yellow fluorescent protein as the readout (Kremers et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). For the two first biosensors carried on pBS0 and pBS7, \u003cem\u003esyfp2\u003c/em\u003e was under the control of a XynR or YjhI-dependent promoter, while the others (pBS3 and pBS8) carry also \u003cem\u003exynR\u003c/em\u003e or \u003cem\u003eyjhI\u003c/em\u003e driven by its own promoter (see \u003cb\u003eFigure S2\u003c/b\u003e in Supplementary data). These plasmids were inserted by transformation into \u003cem\u003eE. coli\u003c/em\u003e MG1655 strain, and fluorescence was measured over time in response to the addition of 1 and 20 mM D-xylonate. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, addition of this sugar acid triggered an increase in fluorescence intensity with both pBS0 and pBS3. However, the basal fluorescence in pBS3 was 200 times lower than that measured with pBS0, and the response time at both D-xylonate concentration was faster with pBS0 than with pBS3. These data are consistent with the repressive effect exerted by XynR, present in greater quantities in strain carrying pBS3 than pBS0 since in the latter, only \u003cem\u003exynR\u003c/em\u003e is expressed from its genomic copy. Unlike XynR, the YjhI-dependent fluorescence response upon addition of D-xylonate was significantly enhanced with pBS8 compared to pBS7. This result is consistent with the function of YjhI, working as an activating transcription factor (Banares et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Also, the induction factor was higher with pBS8 than pBS7, even at low (1 mM) D-xylonate, which can be explained by the auto-activation of \u003cem\u003eyjhI\u003c/em\u003e, caused by the binding of YjhI to its own promoter (Banares et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). In conclusion, the XynR-based biosensor harbors the typical repressed-repressor architecture whereas the activated-activator architecture characterizes the YjhI-based biosensor (Mannan et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). In addition, for both transcription factor-based biosensor, the fold induction was much higher at 20 than 1 mM D-xylonate, suggesting that XynR and YjhI exhibited a relatively weak affinity to D-xylonate.\u003c/p\u003e\n\u003ch3\u003eThe metabolite effector of XynR and YjhI is the intermediate 2-keto-3-deoxy-D-xylonate (KDX)\u003c/h3\u003e\n\u003cp\u003eD-xylonate has been reported as the inducer of the XynR and YjhI-regulated operons but whether D-xylonate was the direct effector sensed by these TFs remained an open question (Banares et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Shimada et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The use of the XynR and YjhI-based biosensors developed above could be useful to address this question, which is all the more relevant in light of the fluorescence kinetics obtained in response to D-xylonate, as the increase of fluorescence was preceded by a lag period of few minutes to one hour (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u003cb\u003e)\u003c/b\u003e. This lag period could be due either to the time needed to release all XynR-molecules that bind to \u003cem\u003eyagE\u003c/em\u003e promoter, to the time for sufficient autoactivation of YjhI or to the time required to produce and accumulate a metabolite from D-xylonate. To solve this issue, we deleted \u003cem\u003eyagF\u003c/em\u003e and \u003cem\u003eyjhG\u003c/em\u003e which code for a D-xylonate dehydratase that catalyses the dehydration of D-xylonate into 2-keto-3-dexoy-D-xylonate (KDX) (see \u003cb\u003eFigure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e)\u003c/b\u003e. Remarkably, XynR-and YjhI-based biosensors expressed in a mutant defective in both genes were no longer responsive to D-xylonate (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). Furthermore, this experiment showed that the D-xylonate dehydratase encoded by \u003cem\u003eyagF\u003c/em\u003e was more active than that encoded by \u003cem\u003eyjhG\u003c/em\u003e, as the fluorescence response in the \u003cem\u003eyagF\u003c/em\u003e mutant expressing the XynR-based biosensor was more reduced than in the \u003cem\u003eyjhG\u003c/em\u003e mutant, and was even abolished in the case of YjhI-based biosensor in response to D-xylonate (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003eBased on these data, we investigated the effect of KDX as a potential metabolite effector of XynR and YjhI transcription factors. This molecule was purchased from Biosynth. Ltd and was validated by LC-MS in our laboratory as about\u0026thinsp;\u0026asymp;\u0026thinsp;90% pure (data not shown). As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA, the addition of 1 mM KDX to MG1655 carrying XynR-based biosensor on pBS3 resulted in a fluorescence response that was faster and higher than after addition of 1 mM D-xylonate. Moreover, the increase in fluorescence was more potent in the mutant strain MGΔ7 which is deleted for \u003cem\u003eyagE\u003c/em\u003e and \u003cem\u003eyjhH\u003c/em\u003e encoding promiscuous aldolases reported to cleave KDX into pyruvate and glycolaldehyde (Bhaskar et al. \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Liu et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). With the YjhI-based biosensor carried by pBS8, the response to KDX and 1 mM D-xylonate was very weak, similar as the one reported above (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB), but a 4 to 5-fold induction was measured in the aldolase-deficient MGΔ7 mutant, which can be explained by the absence of KDX degradation as well as by metabolization of D-xylonate into KDX in this mutant (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). Altogether, these data support the notion that the direct effector of XynR and YjhI transcription factor is the intermediate KDX and not D-xylonate.\u003c/p\u003e\n\u003ch3\u003eThe transcription factors XynR and YjhI are responsive to the non-natural molecule 2-oxo-4-hydroxybutyrate (OHB)\u003c/h3\u003e\n\u003cp\u003eA transcriptomic analysis of an \u003cem\u003eE. coli\u003c/em\u003e MG1655 engineered for the production of 2,4-DHB by fermentation from glucose (Walther et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) revealed a strong upregulation of D-xylonate catabolic genes from the CP4-6 and KpLE2 -phage elements during the production phase of this non-natural metabolite, with notably \u003cem\u003eyagE\u003c/em\u003e and \u003cem\u003eyagF\u003c/em\u003e being increased by 13.6 and 9.1-fold, respectively (see data in \u003cb\u003eTable S4\u003c/b\u003e). Interestingly, \u003cem\u003eyjhI\u003c/em\u003e was also upregulated by about 8-fold, which was accompanied by an increased expression, albeit weak, of \u003cem\u003eyjhH\u003c/em\u003e and \u003cem\u003eyjhG\u003c/em\u003e. Based on these data, we asked whether 2,4-DHB or OHB could mimic the effect of KDX to activate the XynR- and/or YjhI-based biosensor. Results of this experiment are reported in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e. DHB was used at 100 mM because its action on the D-xylonate dependent operons was expected at high concentration whereas OHB was used at 1 mM because intracellular concentration of this metabolite was estimated in the millimolar range (our unpublished data). It can be seen that both 2,4-DHB and OHB caused a comparable and significant increase in fluorescence intensity over time with both biosensors. Also, the dynamic response expressed as fold induction was very similar for response to these unnatural effectors on the two TFs-based biosensors. However, the time response to OHB was clearly faster than that of 2,4-DHB, suggesting that the effect of the latter molecule was not direct but could result from its metabolization. We recently reported that 2,4-DHB can be oxidized into OHB by membrane-associated lactate oxidoreductases encoded by \u003cem\u003elldD\u003c/em\u003e, \u003cem\u003edld\u003c/em\u003e and \u003cem\u003eykgEFG\u003c/em\u003e (Pinchuk et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2009\u003c/span\u003e), as a first step of its assimilation by \u003cem\u003eE. coli\u003c/em\u003e (Malfoy et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). We therefore repeated this experiment using the MGΔLO strain deficient in these lactate oxidoreductases. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD, the increase of fluorescence signal upon addition of 2,4-DHB was totally abolished in this mutant expressing the biosensor carried on pBS3 or pBS8. In contrast, the absence of lactate oxidoreductase had no effect on the response of these biosensors to OHB, indicating that this molecule is likely a direct effector of XynR and YjhI.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eNo crosstalk in the response of XynR and YjhI to KDX and OHB\u003c/h2\u003e \u003cp\u003eSince XynR and YjhI were responsive to the presence of KDX and OHB, we wanted to verify whether a cross-talk between these two TFs could take place. In other words, could it be possible that XynR affected the expression of genes that belong to the YjhI-controlled genes operon and vice versa? To answer this question, we constructed the plasmid pBS9 bearing \u003cem\u003exynR\u003c/em\u003e under its own promoter while \u003cem\u003esyfP2\u003c/em\u003e reporter gene was under \u003cem\u003eyjhI\u003c/em\u003e promoter. Conversely, plasmid pBS6 that carried \u003cem\u003eyjhI\u003c/em\u003e under its own promoter with \u003cem\u003esyfp2\u003c/em\u003e under \u003cem\u003eyagE\u003c/em\u003e promoter was built (see \u003cb\u003eFigure S2\u003c/b\u003e for the scheme of these constructions). The reference strain MG1655, single mutant and double mutants defective in \u003cem\u003exynR\u003c/em\u003e and \u003cem\u003eyjhI\u003c/em\u003e were transformed with these two plasmids and challenged with 1 mM OHB or 20 mM D-xylonate, which was used instead of KDX due to the limited availability of this compound. This experiment showed that the fluorescence induction in response to OHB and D-xylonate in MG\u003cem\u003eΔxynRΔyjhI\u003c/em\u003e mutant strain showed barely no significant difference as compared to the control. This result supported the notion that each transcription factor controls only the genes of its own operon, and therefore indicated an absence of cross-talk in the control of these operons by D-xylonate. The 3 to 6-fold induction observed in response to D-xylonate and OHB with pBS9 in both MG1655 and MGΔxynR can be ascribed to the action of \u003cem\u003eyjhI\u003c/em\u003e present in the genome, which triggered expression of \u003cem\u003esyfp2\u003c/em\u003e gene reporter that is under the \u003cem\u003eyjhI\u003c/em\u003e promoter in this plasmid. Conversely, the weak fold increase in response to OHB and D-xylonate in MG1655 or MGDyjhI strains carrying pBS6 was likely due to the low level of XynR protein, as the gene is expressed from the chromosomal copy, and hence its low abundance resulted in an already high expression of \u003cem\u003esyfp2\u003c/em\u003e gene that is weakly enhanced by addition of OHB or D-xylonate. This behavior resembled that found with MG1655 expressing pBS0 (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eCharacteristic performances of the XynR and YjhI-based biosensor to KDX and OHB\u003c/h2\u003e \u003cp\u003eAs clearly outlined in the reviews by F. Zhang and colleagues (Hartline et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Mannan et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), several criteria can be used to evaluate the biosensor performances which include the response threshold (\u003cem\u003ei.e.\u003c/em\u003e K\u003csub\u003e0.5\u003c/sub\u003e), defined as the concentration of metabolite required for 50% of the maximal expression or induction, the detection range and the dynamic range corresponding to the maximal fold increase relative to the baseline. To evaluate these parameters for XynR-and YjhI-based sensors to OHB and KDX, MGΔ7 strain deleted for the so far seven identified genes encoding pyruvate-dependent aldolase (He et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) was used to prevent the metabolization of KDX into pyruvate and glycoladehyde (for KDX) and OHB into pyruvate and formaldehyde (Banares et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Malfoy et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Results reported in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e showed that the XynR-based biosensor turned to be more sensitive to OHB than KDX (K\u003csub\u003e0.5\u003c/sub\u003e \u0026asymp; 0.27 vs 1.0 mM, Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e), while a comparable K\u003csub\u003e0.5\u003c/sub\u003e in the range of 0.3\u0026ndash;0.4 mM was obtained for both metabolites with the YjhI-based biosensor (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Furthermore, the cooperative behavior of the two TF-based biosensors expressed by the Hill number (n\u003csup\u003eH\u003c/sup\u003e) was significantly different between KDX and OHB. While the cooperativity of the XynR-based biosensor was low for KDX (nH\u0026thinsp;\u0026asymp;\u0026thinsp;1.13), it was two times higher for OHB. The opposite behavior was observed for the YjhI-based biosensor, for which the cooperativity was roughly two time higher for KDX (n\u003csup\u003eH\u003c/sup\u003e \u0026asymp;2.7) than for OHB (see Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). These results suggested that although both effectors may interact with the same protein, they differentially modulate cooperative interactions within the protein. The dynamic range (DR), expressed as the fold induction of fluorescence relative to baseline, for the natural (KDX) and artificial (OHB) effectors can be also extrapolated from the dose-response curves. They revealed that the dynamic range of the YjhI-based biosensor for OHB and KDX was about 2-times better than that of XynR-based biosensor (\u003cem\u003ei.e.\u003c/em\u003e in the range of 50-fold versus 25; see Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCharacteristic performance of XynR and YjhI transcription factors towards their metabolic effectors\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eEffector\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e \u003cp\u003eXynR*\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c7\" namest=\"c5\"\u003e \u003cp\u003eYjhI*\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003en\u003csup\u003eH\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eK\u003csub\u003e0.5\u003c/sub\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eDR\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003en\u003csup\u003eH\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eK\u003csub\u003e0.5\u003c/sub\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eDR\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOHB\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.05\u0026thinsp;\u0026plusmn;\u0026thinsp;0.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.27\u0026thinsp;\u0026plusmn;\u0026thinsp;0.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e25\u0026thinsp;\u0026plusmn;\u0026thinsp;5.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.25\u0026thinsp;\u0026plusmn;\u0026thinsp;0.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.36\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e48\u0026thinsp;\u0026plusmn;\u0026thinsp;6.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eKDX\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.13\u0026thinsp;\u0026plusmn;\u0026thinsp;0.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.0\u0026thinsp;\u0026plusmn;\u0026thinsp;0.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e25\u0026thinsp;\u0026plusmn;\u0026thinsp;5.0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.7\u0026thinsp;\u0026plusmn;\u0026thinsp;0.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.32\u0026thinsp;\u0026plusmn;\u0026thinsp;0.04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e52\u0026thinsp;\u0026plusmn;\u0026thinsp;8.0\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eD-xylonate\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.1\u0026thinsp;\u0026plusmn;\u0026thinsp;0.7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003end\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.4\u0026thinsp;\u0026plusmn;\u0026thinsp;0.4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e4.0\u0026thinsp;\u0026plusmn;\u0026thinsp;0.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003end\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"7\"\u003e*Values are the mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD of four biological replicates. n\u003csup\u003eH\u003c/sup\u003e means Hill number, DR means dynamic range\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eAlthough KDX is the direct effector of XynR and YjhI transcription factors, the dose-response curves to D-xylonate were also performed, showing that both TFs exhibited similar characteristic performances with a response threshold in the range of 5 mM, and a dynamic range of about 30\u0026thinsp;\u0026plusmn;\u0026thinsp;5.0 (\u003cb\u003eFigure S3\u003c/b\u003e in Supplementary data).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eDevelopment of a OHB biosensor\u003c/h2\u003e \u003cp\u003eFrom the results presented above, it turned out that both transcription factors XynR and YjhI could be exploited for the design of an OHB-responsive biosensor. However, to ensure the practical use of these biosensors, we had to verify their specificity to this unnatural metabolite in relation to other metabolites present in bacterial cells, particularly organic and amino acids. These metabolites were tested at a pretty large concentration (10 mM), and the fluorescence signal that can be triggered by these compounds was compared to that obtained upon addition of 1 mM of OHB. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eA, most of the organic acids had minor to no effect on the XynR-based biosensor (Note: the time course response of these metabolites are reported in \u003cb\u003eFigure S4\u003c/b\u003e, supplementary data\u003cb\u003e)\u003c/b\u003e. However, compounds that, like OHB, have a ketone function at C2, such as pyruvate, α-ketoglutarate, and aspartate caused a 3 to 4-fold induction of fluorescence. A 2 to 3-fold induction was also recorded with organic acids having an α hydroxyl function, such as D and L-lactate or 2-hydroxybutyrate. Less obvious was the finding of a 4-fold induction caused by 10 mM fumarate. On the other hand, the basal fluorescence of the biosensor was lower than the control in the presence of L-malate or glycolate, which both also harbor an α-hydroxyl group. Same effect was found with glycolaldehyde. How these molecules caused such a reduction of the fluorescence is so far unclear but nevertheless, these effects can be considered negligible overall, as they occurred at concentrations well above those experienced in cells (Bennett et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). Similar results were obtained with the YjhI-based biosensor, with notably a stronger inhibition of the signal fluorescence by L-malate, glycolate and glycolaldehyde and an even higher fold induction in response to fumarate (\u003cb\u003eFigure S5\u003c/b\u003e in suppl data). We also investigated the effects of sugar acids homologous to D-xylonate as they can be used as carbon substrate. Results of this experiment clearly indicated that XynR-based biosensor was responsive neither to the enantiomer (L-xylonate) nor the stereoisomer (L-arabinonate) of D-xylonate (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB\u003cb\u003e)\u003c/b\u003e. The lack of effect of these sugar acids is consistent with the fact that these compounds cannot be metabolized by \u003cem\u003eE. coli\u003c/em\u003e (Ren et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Conversely, the 4-fold increase in response to D-gluconate could be explained either by its dehydration into 2-keto-3-deoxy-D-gluconate catalysed by YagF/YagG dehydratase or to the possibility that gluconate can be converted into 2-keto-3-deoxy-6-phosphogluconate (KDPG) through the Entner-Doudoroff pathway (Eisenberg and Dobrogosz \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e1967\u003c/span\u003e). We solved this issue by showing that a mutant lacking the dehydratase did not exhibit any fluorescence signal in response to D-gluconate (data not shown). D-glucuronate and D-threonate were both inefficient as expected since the former one can only be metabolized via the Entner-Doudoroff providing D-glucuronate isomerase encoded by \u003cem\u003euxaC\u003c/em\u003e (Portalier et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e1980\u003c/span\u003e) is functional, which needs the presence of this sugar acid to be expressed (Mandrand-Berthelot et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), whereas only L-threonate is reported to be metabolized by \u003cem\u003eE. coli\u003c/em\u003e (Fraz\u0026atilde;o et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Similar results were obtained with the YjhI-based biosensor (see \u003cb\u003eFigure S5B\u003c/b\u003e in suppl. Data).\u003c/p\u003e \u003cp\u003eBeside specificity, other relevant properties of a metabolite responsive transcription factor exploited as a biosensor should be to exhibit a wide range of effector concentrations and a large dynamic response characterized by high fold induction of fluorescence (response) relative to the background. These properties can be tuned at first glance through promoter engineering of either the transcription factor and/or the reporter gene \u003cem\u003esyfp2\u003c/em\u003e. Starting from pBS3 as the reference plasmid bearing the XynR-based biosensor, we generated a series of plasmids in which either the native RBS in the XynR-promoter was replaced by a synthetically optimised RBS\u003csub\u003e01\u003c/sub\u003e according to RBS calculator (Salis \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) or the native promoter of XynR was replaced with iGEM promoters of decreasing strength, while retaining the same RBS\u003csub\u003e01\u003c/sub\u003e in these synthetic promoters. These different constructs did not result in a better dynamic range to OHB as compared to the original construct in pBS3 (see \u003cb\u003eFigure S6\u003c/b\u003e in suppl. data). It is clear that other modifications, such as engineering the transcription factor XynR to broaden the detection range or sensitivity to OHB, should be considered, but this work is beyond the scope of this study.\u003c/p\u003e \u003cp\u003e \u003cb\u003eUse of the XynR- and YjhI-based biosensor to screen for OHB producing enzymes.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe compound OHB is the most proximal intermediate in the production of 2,4-DHB, a chemical platform for the production of several added-value products (Francois \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The synthesis of OHB can be obtained by transamination of L-homoserine with pyruvate or α-ketoglutarate as the co-substrate (Bouzon et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Walther et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). However, the catalytic efficiency of this enzyme is approximately 100 to 1000 times lower than that of other enzymes in these pathways. Improving the activity of these enzymes is therefore mandatory in order to increase the flux in the pathway and consequently increase the rate of 2,4-DHB production. In that frame, a biosensor able to monitor the formation of the product of the reaction catalysed by the enzyme, such as OHB in our case, could be an appropriate tool for high throughput screening of more active transaminases. We thus decided to validate the use of the XynR/YjhI-based biosensor to screen for transaminase activity using the variant of AlaC\u003csup\u003eA142P Y245D\u003c/sup\u003e as a positive control since this variant turned out to be the most active to produce OHB from L-homoserine in the presence of pyruvate or a-ketoglutarate, albeit at a still low catalytic efficiency of \u0026asymp;\u0026thinsp;500 M\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Bouzon et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). To validate the screen, we used the strain MGDT which was deleted for \u003cem\u003etyrB\u003c/em\u003e, \u003cem\u003eilvE, ybdL\u003c/em\u003e and \u003cem\u003ealaC\u003c/em\u003e to get rid of most of the alternate transaminases exhibiting even a weak activity on L -homoserine (Walther et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Walther et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). It was also deleted of \u003cem\u003easd\u003c/em\u003e to avoid endogenous production of L-homoserine. At first glance, we tested the XynR-biosensor through co-transformation of this strain with pBS3 and either pETM (empty), used as control, pETM carrying \u003cem\u003ealaC\u003c/em\u003e (coding for wild type AlaC) and pETM carrying \u003cem\u003ealaC**\u003c/em\u003e (coding for the variant AlaC\u003csup\u003eA142P Y245D\u003c/sup\u003e). Since these genes were under the control of an IPTG-inducible promoter, part of the culture was treated for 4 h with 0.5 mM IPTG before addition of the various metabolites. It turned out that the results obtained were inconsistent, with an apparent interference of IPTG on the expression of pBS3 since the fluorescence response to OHB was strongly impaired in cells treated with IPTG (data not shown). We therefore tested the YjhI-based biosensor following the same experimental procedure. As reported in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e, the fold induction of fluorescence of IPTG-untreated cultures in response to homoserine, pyruvate or α-ketoglutarate was relatively weak and always lower than in response to 1 mM OHB use as control. One could however notice a statistically significant (about 8-fold) fold induction of fluorescence relative to the control in the strain carrying pETM-alaC**. This result indicated that even in the absence of IPTG, the genes in these plasmids could be expressed and moreover these data supported the fact that AlaC\u003csup\u003eA142P Y250D\u003c/sup\u003e variant was more active than AlaC to produce OHB. Also, the weak but significant increase in fluorescence in strain expressing the empty plasmid in response to homoserine alone or with pyruvate or α-ketoglutarate could be ascribed to endogenous transaminases such as the one encoded by \u003cem\u003easpC\u003c/em\u003e which has been reported to have a weak OHB producing activity (Walther et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). More importantly, the treatment of the culture with IPTG prior addition of the various metabolites clearly validated the biosensor tool since a 40-fold induction of fluorescence was recorded after addition of L-homoserine with either pyruvate or α-ketoglutarate in strain expressing the AlaC\u003csup\u003eA142P Y250D\u003c/sup\u003e variant, whereas only a 8-10-fold induction was measured with the strain expressing the wild type enzyme. Of note, a 40-fold induction was found to be close to the maximal dynamic response of OHB (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Furthermore, pre-treatment with IPTG of bacterial cultures carrying either an empty plasmid, pETM-alaC, or pETM-alaC** caused a fold induction of fluorescence in response to homoserine, pyruvate, or α-ketoglutarate alone, which was almost equivalent to that obtained after adding 1 mM OHB. Part of the explanation may lie in the action of other transaminases, particularly that encoded by \u003cem\u003easpC\u003c/em\u003e, but also in the fact that the presence of IPTG may have caused a metabolic rearrangement, providing precursors for OHB synthesis.\u003c/p\u003e \u003cp\u003eWe then tested the D-threonate dehydratase activity which was derived from the promiscuous D-arabinonate dehydratase of \u003cem\u003eHerbaspirillum huttiense\u003c/em\u003e (\u003cem\u003eHharaD\u003c/em\u003e) (Watanabe et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). This enzyme was shown to synthetize OHB by dehydration of D-threonate in a conceived synthetic pathway starting from ethylene glycol and ending with the formation of 2,4-DHB (Fraz\u0026atilde;o et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). For this validation, MG1655 strain was co-transformed with pBS3 (XynR-based biosensor) and either pZS2 carrying \u003cem\u003eHharaD\u003c/em\u003e with \u003cem\u003ekdgT\u003c/em\u003e encoding a transporter of \u003cem\u003eCupriavidus necator\u003c/em\u003e reported to facilitate the import of D-threonate, or with pZS2 carrying only \u003cem\u003ekgdT\u003c/em\u003e. Remarkably, the addition of D-threonate to bacterial cells expressing \u003cem\u003eHharaD\u003c/em\u003e encoding this threonate dehydratase resulted in a significant increase in fluorescence signal that was equivalent to that obtained after the addition of 1 mM OHB, whereas the fluorescence of cells lacking this gene remained at a basal level after the addition of the sugar acid (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e). Of note, the assay was not carried out with the YjhI-based biosensor due to the scarcity of D-threonate.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eUse of the XynR-based biosensor to screen for 2,4-DHB transporters\u003c/h2\u003e \u003cp\u003eSince 2,4-DHB is a non-natural molecule for \u003cem\u003eE. coli\u003c/em\u003e, we sought whether the OHB-biosensor could be used to screen for a transporter of this molecule taking into account that OHB can be produced by the oxidation of 2,4-DHB by the membrane-associated lactate oxidoreductase (Malfoy et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). To ensure that this reaction would not be rate limiting in the screen, we inserted \u003cem\u003elldD\u003c/em\u003e encoding one of the three lactate oxidoreductases present in \u003cem\u003eE. coli\u003c/em\u003e into the pBS3 plasmid under the constitutive synthetic pJ23106 promoter yielding pBS5 (see \u003cb\u003eFigure S2\u003c/b\u003e). We validated this construct by measuring the activity on 2,4-DHB in lysates of cells that have been transformed with pBS5 and found a 10-times higher activity on this compound than in cells carrying pBS3 (data not shown). Then, around 200 \u003cem\u003eE. coli\u003c/em\u003e mutants of the Keio collection (Baba et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2006\u003c/span\u003e) deleted for known and putative transporters of organic and amino acids were transformed with pBS5 and cultivated in M9 with 0.4% D-xylose in the presence of either 10 or 100 mM 2,4-DHB. Fluorescence was monitored by flow cytometry after 4, 8 and 24 h of culture. Results collected after 24 h showed that there were several genes belonging to amino acids and organic acids transporters that exhibited significant reduction of the fluorescence and thus potentially impaired the import of 2,4-DHB (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e). In particular, two out of the 30 retained candidates causing\u0026thinsp;\u0026gt;\u0026thinsp;50% reduction of fluorescence encoded a multidrug efflux pump, 11 encoded amino acids transporters, 6 encoded organic acids transporters and 5 encoded uncharacterized transporters (\u003cb\u003eTable S2\u003c/b\u003e). When the screen was carried out with 100 mM 2,4-DHB, only two candidates of those identified at 10 mM were found, namely \u003cem\u003eproX\u003c/em\u003e which encodes glycine betaine ABC transporter and \u003cem\u003eyfdC\u003c/em\u003e, which encodes a yet uncharacterized transporter \u003cb\u003e(Table S2\u003c/b\u003e). This large collection of transporters suggested a seemingly unspecific import of 2,4-DHB. To get more insight about this screening assay, we choose to delete \u003cem\u003eydfC\u003c/em\u003e, \u003cem\u003eacrD\u003c/em\u003e, \u003cem\u003eglnH\u003c/em\u003e, \u003cem\u003ekdgT\u003c/em\u003e and \u003cem\u003eygbN\u003c/em\u003e into the strain MGΔLO and confirmed our original data with the mutant from the Keio collection (data not shown).\u003c/p\u003e \u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThe construction of biosensors dependent on the transcription factors XynR and YjhI enabled us to resolve two major problems concerning the induction of the D-xylonate catabolic pathway in E. coli. First, the natural effector of these two transcription factors that induce the catabolic genes present on the \u003cem\u003eyag\u003c/em\u003e and \u003cem\u003eyjh\u003c/em\u003e operons is not D-xylonate, but the intermediate KDX, which is formed by the dehydration of D-xylonate by D-xylonate dehydratase encoded by \u003cem\u003eyagF\u003c/em\u003e and \u003cem\u003eyjhG\u003c/em\u003e. Moreover, our results suggested that \u003cem\u003ein vivo\u003c/em\u003e, YagF is more active than YjhgG. This higher activity could be explained either by a higher expression of \u003cem\u003eyagF\u003c/em\u003e than \u003cem\u003eyjhG\u003c/em\u003e, or by a higher affinity of YagF for D-xylonate. To date, only kinetic data have been reported for the D-xylonate dehydratase YjhG, showing a K\u003csub\u003eM\u003c/sub\u003e of 4.88 mM and a relative weak k\u003csub\u003ecat\u003c/sub\u003e/K\u003csub\u003eM\u003c/sub\u003e of 66 M\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e s\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e (Jiang et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2015b\u003c/span\u003e). Interestingly, this value is in the range of the concentration of D-xylonate leading to 50% of the maximal fluorescence induction of XynR-and YjhI based biosensor. On the other hand, we have shown that there is no interference in the control of the yagEFG and yjhIHG operons, as each transcription factor regulates its own operon without interfering with the other's operon.\u003c/p\u003e \u003cp\u003eThe other part of the work was to examine whether XynR and/or YjhI are also responsive to the non-natural metabolites 2,4-DHB and/or OHB to repurpose them as biosensor tools to improve the synthetic pathways of the platform molecule 2,4-DHB (Francois \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Our data showed that OHB can effectively play the same role as KDX on these two transcription factors, most likely because, except for the fact that the latter is a 5-carbon molecule, OHB and KDX share the same chemical structure. Based on data with other organic and amino acids used to test the specificity of Xynr and YjhI to OHB and KDX, it can be proposed that the minimal structure recognized by these TFs required the functional carboxyl group, a keto function on the carbon α and a hydroxyl group on the carbon g. Also, the fact that XynR and YjhI are responsive to these two effectors is due to their 3D-structure perfectly overlapping, according to the Alphafold-2 prediction (see \u003cb\u003eFigure S8\u003c/b\u003e in in supplementary data). However, there is an additional β-sheet between aa 54 and 56, as well as an extended short α-helix at the C-ter of the YjhI protein. These small differences could be sufficient to explain why YjhI is an activator and not a repressor like XynR, but more detailed structure-function will be required to ascertain this suggestion. Although the protein structure of these two transcription factors is almost identical, their kinetic characteristics, namely K\u003csub\u003e0.5\u003c/sub\u003e and the Hill number n\u003csup\u003eH\u003c/sup\u003e, with respect to OHB and KDX were found different. The transcription factor XynR exhibited an affinity for OHB that is 3 to 4 times better than for KDX, while the opposite was observed for YjhI. Altogether, the finding that XynR and YjhI are structurally similar while the former is acting as a repressor of transcription and the latter as an activator and that they exhibit kinetic differences to their ligands may require more detailed structure-function analysis, which is beyond the scope of this work.\u003c/p\u003e \u003cp\u003eIn previous works, we developed 4 different synthetic pathways leading to the production of 2,4-DHB from C1 to C6-renewable carbon sources (Fraz\u0026atilde;o et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Walther et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Walther et al. \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). This synthon can be considered a unique bio-based platform molecule capable of giving rise to several value-added molecules, including a hydroxylated derivative of methionine, 1,3-propanediol, 3-hydroxypropionic acid, 1,2,4-butanetriol and biopolymers (Francois \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Pascouau et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2023a\u003c/span\u003e; Pascouau et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2023b\u003c/span\u003e). Two of these four pathways, which have proven to be the most suitable for industrial application, rely on the formation of the intermediate OHB by promiscuous enzymes whose catalytic efficiency is far too low, significantly hampering the productivity and yield of 2,4-DHB production. An OHB-sensitive biosensor could be an appropriate method, which combined with flow cytometry coupled with cell sorting (FACS), would enable the selection of enzyme variants with higher catalytic activity. The XynR-based biosensor can be optimally designed for use in this context, as it features a repressed-repressor architecture that enables positive selection, \u003cem\u003ei.e.\u003c/em\u003e the signal response is triggered only when the target metabolite is produced. Our data aligned in part with these requirements. We validated the potential value of this OHB-sensitive biosensor as a screening tool for threonate dehydratase, the rate limiting enzyme in the synthetic pathway yielding to DHB from ethylene glycol as the carbon source (Fraz\u0026atilde;o et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). However, it was not possible to use this XynR-based biosensor as a screening method for homoserine transaminase, which catalyzes the rate-limiting step in the formation of DHB from glucose via the aspartate-homoserine pathway (Walther et al. \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), whereas the YjhI-based biosensor, which has an activator-activated architecture, proved to be functional. It is difficult to find a simple explanation for this result, but one possible reason is that the screening requires the presence of two plasmids in the bacteria, one carrying the biosensor and the other carrying the gene encoding the transaminase. For the latter, gene expression is dependent on the IPTG-inducible promoter. We observed strong interference on the XynR-based biosensor upon incubation with IPTG as witnessed by a weak increase of fluorescence in response to OHB. Finally, we showed that this biosensor could be useful to identify putative DHB transporter, using the single-gene knockout mutants of \u003cem\u003eE. coli\u003c/em\u003e (Baba et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Notably, \u003cem\u003eygbN\u003c/em\u003e was identified in this screen, which we previously reported as a transporter of the D-form of 2,4-DHB (Malfoy et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). It is worth noticing that the OHB-biosensor reported in this work is totally different from the OHB biosensor (termed HOB biosensor) developed by Schann \u003cem\u003eet al\u003c/em\u003e. (Schann et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The latter is a growth-based sensor whose actual sensor is formaldehyde that is produced via promiscuous aldolases and transaminases to be assimilated through homoserine in a strain that is defective in the natural aspartate -homoserine pathway.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eConflicts of interest\u003c/h2\u003e \u003cp\u003eThe authors declare no commercial or financial conflict of interest.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eAuthor contribution statement\u003c/h2\u003e \u003cp\u003eT.M. and J.M.F. designed the study. T.M., C.A., J.F., J. L-H., and J.M.F. conducted the experiments. T.M. and J.M.F. drafted the manuscript, which was reviewed and improved by all authors before a final version written by J.M.F. and approved by all authors.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis work was supported by French Research National Agency (ANR) grant ANR-CE43-0008-01 (PolyDHB) and grant ANR-21-COBI-0003-01 (SYNBIOMET) to J.M.F.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eT.M. and J.M.F. designed the study. T.M., C.A., J.F., J. L-H., and J.M.F. conducted the experiments. T.M. and J.M.F. drafted the manuscript, which was reviewed and improved by all authors before a final version written by J.M.F. and approved by all authors.\u003c/p\u003e\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eWe are grateful To Dr C. Frazao \u0026amp; Prof Th. Walther from TU Dresden, Germany, for the kind gift of pZS2-kdgT_\u003cem\u003eHh\u003c/em\u003e.araD plasmids together with a few mg of D-threonate.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBaba T, Ara T, Hasegawa M, Takai Y, Okumura Y, Baba M, Datsenko KA, Tomita M, Wanner BL, Mori H (2006) Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol 2:20060008. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/msb4100050\u003c/span\u003e\u003cspan address=\"10.1038/msb4100050\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBanares AB, Nisola GM, Valdehuesa KNG, Lee WK, Chung WJ (2021) Understanding D-xylonic acid accumulation: a cornerstone for better metabolic engineering approaches. Appl Microbiol Biotechnol 105(13):5309\u0026ndash;5324. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00253-021-11410-y\u003c/span\u003e\u003cspan address=\"10.1007/s00253-021-11410-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBanares AB, Valdehuesa KNG, Ramos KRM, Nisola GM, Lee WK, Chung WJ (2019) Discovering a novel D-xylonate-responsive promoter: the PyjhI-driven genetic switch towards better 1,2,4-butanetriol production. Appl Microbiol Biotechnol 103(19):8063\u0026ndash;8074. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00253-019-10073-0\u003c/span\u003e\u003cspan address=\"10.1007/s00253-019-10073-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBennett BD, Kimball EH, Gao M, Osterhout R, Van Dien SJ, Rabinowitz JD (2009) Absolute metabolite concentrations and implied enzyme active site occupancy in Escherichia coli. Nat Chem Biol 5(8):593\u0026ndash;599 :nchembio.186 [pii];10.1038/nchembio.186 [doi]\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBhaskar V, Kumar M, Manicka S, Tripathi S, Venkatraman A, Krishnaswamy S (2011) Identification of biochemical and putative biological role of a xenolog from Escherichia coli using structural analysis. Proteins 79(4):1132\u0026ndash;1142. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/prot.22949\u003c/span\u003e\u003cspan address=\"10.1002/prot.22949\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBouzon M, Perret A, Loreau O, Delmas V, Perchat N, Weissenbach J, Taran F, Marliere P (2017) A Synthetic Alternative to Canonical One-Carbon Metabolism. ACS Synth Biol 6(8):1520\u0026ndash;1533. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acssynbio.7b00029\u003c/span\u003e\u003cspan address=\"10.1021/acssynbio.7b00029\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCherepanov PP, Wackernagel W (1995) Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant. Gene 158(1):9\u0026ndash;14\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDahms AS (1974) 3-Deoxy-D-pentulosonic acid aldolase and its role in a new pathway of D-xylose degradation. Biochem Biophys Res Commun 60(4):1433\u0026ndash;1439\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEisenberg RC, Dobrogosz WJ (1967) Gluconate metabolism in Escherichia coli. J Bacteriol 93(3):941\u0026ndash;949. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1128/jb.93.3.941-949.1967\u003c/span\u003e\u003cspan address=\"10.1128/jb.93.3.941-949.1967\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrancois JM (2023) Progress advances in the production of bio-sourced methionine and its hydroxyl analogues. Biotechnol Adv 69:108259. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.biotechadv.2023.108259\u003c/span\u003e\u003cspan address=\"10.1016/j.biotechadv.2023.108259\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFraz\u0026atilde;o CJR, Wagner N, Rabe K, Walther T (2023) Construction of a synthetic metabolic pathway for biosynthesis of 2,4-dihydroxybutyric acid from ethylene glycol. Nat Commun 14(1):1931. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-023-37558-x\u003c/span\u003e\u003cspan address=\"10.1038/s41467-023-37558-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrazao CR, Maton V, Francois JM, Walther T (2018) Development of a Metabolite Sensor for High-Throughput Detection of Aldehydes in Escherichia Coli. Front Bioeng Biotechnol 6:118. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fbioe.2018.00118\u003c/span\u003e\u003cspan address=\"10.3389/fbioe.2018.00118\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanko EKR, Paiva AC, Jonczyk M, Abbott M, Minton NP, Malys N (2020) A genome-wide approach for identification and characterisation of metabolite-inducible systems. Nat Commun 11(1):1213. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-020-14941-6\u003c/span\u003e\u003cspan address=\"10.1038/s41467-020-14941-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHartline CJ, Schmitz AC, Han Y, Zhang F (2021) Dynamic control in metabolic engineering: Theories, tools, and applications. Metab Eng 63:126\u0026ndash;140. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ymben.2020.08.015\u003c/span\u003e\u003cspan address=\"10.1016/j.ymben.2020.08.015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe H, Hoper R, Dodenhoft M, Marliere P, Bar-Even A (2020) An optimized methanol assimilation pathway relying on promiscuous formaldehyde-condensing aldolases in E. coli. Metab Eng 60:1\u0026ndash;13. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ymben.2020.03.002\u003c/span\u003e\u003cspan address=\"10.1016/j.ymben.2020.03.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang Y, Chen B, Duan C, Sun B, Yang J, Yang S (2015a) Multigene editing in the Escherichia coli genome via the CRISPR-Cas9 system. Appl Environ Microbiol 81(7):2506\u0026ndash;2514. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1128/AEM.04023-14\u003c/span\u003e\u003cspan address=\"10.1128/AEM.04023-14\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang Y, Liu W, Cheng T, Cao Y, Zhang R, Xian M (2015b) Characterization of D-xylonate dehydratase YjhG from Escherichia coli. Bioengineered 6(4):227\u0026ndash;232. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/21655979.2015.1040208\u003c/span\u003e\u003cspan address=\"10.1080/21655979.2015.1040208\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKitagawa M, Ara T, Arifuzzaman M, Ioka-Nakamichi T, Inamoto E, Toyonaga H, Mori H (2005) Complete set of ORF clones of Escherichia coli ASKA library (a complete set of E. coli K-12 ORF archive): unique resources for biological research. DNA Res 12(5):291\u0026ndash;299. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/dnares/dsi012\u003c/span\u003e\u003cspan address=\"10.1093/dnares/dsi012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKremers GJ, Goedhart J, van Munster EB, Gadella TW Jr (2006) Cyan and yellow super fluorescent proteins with improved brightness, protein folding, and FRET Forster radius. Biochemistry 45(21):6570\u0026ndash;6580. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/bi0516273\u003c/span\u003e\u003cspan address=\"10.1021/bi0516273\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu H, Ramos KR, Valdehuesa KN, Nisola GM, Lee WK, Chung WJ (2013) Biosynthesis of ethylene glycol in Escherichia coli. Appl Microbiol Biotechnol 97(8):3409\u0026ndash;3417. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00253-012-4618-7\u003c/span\u003e\u003cspan address=\"10.1007/s00253-012-4618-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalfoy T, Alkim C, Barthe M, Fredonnet J, Fran\u0026ccedil;ois JM (2024) Enzymatic promiscuity and underground reactions accounted for the capability of Escherichia coli to use the non-natural chemical synthon 2,4-dihydroxybutyric acid as a carbon source for growth. Microbiol Res 288:127888. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.micres.2024.127888\u003c/span\u003e\u003cspan address=\"10.1016/j.micres.2024.127888\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMandrand-Berthelot MA, Condemine G, Hugouvieux-Cotte-Pattat N (2004) Catabolism of Hexuronides, Hexuronates, Aldonates, and Aldarates. EcoSal Plus 1(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1128/ecosalplus.3.4.2\u003c/span\u003e\u003cspan address=\"10.1128/ecosalplus.3.4.2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMannan AA, Liu D, Zhang F, Oyarzun DA (2017) Fundamental Design Principles for Transcription-Factor-Based Metabolite Biosensors. ACS Synth Biol 6(10):1851\u0026ndash;1859. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acssynbio.7b00172\u003c/span\u003e\u003cspan address=\"10.1021/acssynbio.7b00172\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMiller JH (1992) A Short Course in Bacterial Genetics: A Laboratory Manual and Handbook for Escherichia coli and Related Bacteria. Plainview, NY\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePascouau C, M\u0026eacute;reau R, Grau E, Wirotius A-L, Carlotti S, Cramail H, Peruch F (2023a) Ring-Opening Copolymerization of α-Hydroxy-γ-butyrolactone and ε-Caprolactone. Toward the Metal-Free Synthesis of Functional Polyesters. ACS Appl Polym Mater 5(8):6685\u0026ndash;6694. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acsapm.3c01303\u003c/span\u003e\u003cspan address=\"10.1021/acsapm.3c01303\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePascouau C, Wirotius A-L, Carlotti S, Cramail H, Peruch F (2023b) Functional polyesters via ring-opening copolymerization of α\u0026ndash;hydroxy\u0026ndash;γ\u0026ndash;butyrolactone and ε-caprolactone: La[N(SiMe3)2]3 as an efficient coordination-insertion catalyst. Eur Polymer J 185:111793. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.eurpolymj.2022.111793\u003c/span\u003e\u003cspan address=\"10.1016/j.eurpolymj.2022.111793\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePinchuk GE, Rodionov DA, Yang C, Li X, Osterman AL, Dervyn E, Geydebrekht OV, Reed SB, Romine MF, Collart FR, Scott JH, Fredrickson JK, Beliaev AS (2009) Genomic reconstruction of Shewanella oneidensis MR-1 metabolism reveals a previously uncharacterized machinery for lactate utilization. Proc Natl Acad Sci U S A 106(8):2874\u0026ndash;2879. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1073/pnas.0806798106\u003c/span\u003e\u003cspan address=\"10.1073/pnas.0806798106\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePortalier R, Robert-Baudouy J, Stoeber F (1980) Regulation of Escherichia coli K-12 hexuronate system genes: exu regulon. J Bacteriol 143(3):1095\u0026ndash;1107. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1128/jb.143.3.1095-1107.1980\u003c/span\u003e\u003cspan address=\"10.1128/jb.143.3.1095-1107.1980\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRen Y, Eronen V, Blomster Andberg M, Koivula A, Hakulinen N (2022) Structure and function of aldopentose catabolism enzymes involved in oxidative non-phosphorylative pathways. Biotechnol Biofuels Bioprod 15(1):147. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13068-022-02252-5\u003c/span\u003e\u003cspan address=\"10.1186/s13068-022-02252-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSalis HM (2011) The ribosome binding site calculator. Methods in enzymology 498:19\u0026ndash;42 doi:B978-0-12-385120-8.00002-4 [pii];10.1016/B978-0-12-385120-8.00002\u0026ndash;4 [doi]\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchann K, Bakker J, Boinot M, Kuschel P, He H, Nattermann M, Paczia N, Erb T, Bar-Even A, Wenk S (2024) Design, construction and optimization of formaldehyde growth biosensors with broad application in biotechnology. Microb Biotechnol 17(7):e14527. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/1751-7915.14527\u003c/span\u003e\u003cspan address=\"10.1111/1751-7915.14527\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShimada T, Momiyama E, Yamanaka Y, Watanabe H, Yamamoto K, Ishihama A (2017) Regulatory role of XynR (YagI) in catabolism of xylonate in Escherichia coli K-12. FEMS Microbiol Lett 364(22). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/femsle/fnx220\u003c/span\u003e\u003cspan address=\"10.1093/femsle/fnx220\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThomason LC, Costantino N, Court DL (2007) E. coli genome manipulation by P1 transduction. Curr Protoc Mol Biol Chap 1:Unit 1 17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/0471142727.mb0117s79\u003c/span\u003e\u003cspan address=\"10.1002/0471142727.mb0117s79\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalther T, Calvayrac F, Malbert Y, Alkim C, Dressaire C, Cordier H, Francois JM (2018) Construction of a synthetic metabolic pathway for the production of 2,4-dihydroxybutyric acid from homoserine. Metab Eng 45:237\u0026ndash;245. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ymben.2017.12.005\u003c/span\u003e\u003cspan address=\"10.1016/j.ymben.2017.12.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalther T, Topham CM, Irague R, Auriol C, Baylac A, Cordier H, Dressaire C, Lozano-Huguet L, Tarrat N, Martineau N, Stodel M, Malbert Y, Maestracci M, Huet R, Andre I, Remaud-Simeon M, Francois JM (2017) Construction of a synthetic metabolic pathway for biosynthesis of the non-natural methionine precursor 2,4-dihydroxybutyric acid. Nat Commun 8:15828 :ncomms15828 [pii];10.1038/ncomms15828 [doi]\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang HH, Isaacs FJ, Carr PA, Sun ZZ, Xu G, Forest CR, Church GM (2009) Programming cells by multiplex genome engineering and accelerated evolution. Nature 460(7257):894\u0026ndash;898. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/nature08187\u003c/span\u003e\u003cspan address=\"10.1038/nature08187\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWatanabe S, Fukumori F, Watanabe Y (2019) Substrate and metabolic promiscuities of d-altronate dehydratase family proteins involved in non-phosphorylative d-arabinose, sugar acid, l-galactose and l-fucose pathways from bacteria. Mol Microbiol 112(1):147\u0026ndash;165. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/mmi.14259\u003c/span\u003e\u003cspan address=\"10.1111/mmi.14259\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"biosensor, DNA-binding transcription regulator, D-xylonate, 2-oxo-4-hydroxybutyric acid, 2,4-dihydroxybutyric acid, genetic circuits, synthetic pathways","lastPublishedDoi":"10.21203/rs.3.rs-8830407/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8830407/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn \u003cem\u003eE. coli\u003c/em\u003e, D-xylonate catabolism is carried out by two distinct operons, namely \u003cem\u003eyagEFG\u003c/em\u003e and \u003cem\u003eyjhIHG\u003c/em\u003e, present on two distinct cryptic phages in the genome of this bacterium. These two operons are controlled by the transcription factors XynR and YjhI, the former acting as a repressor of \u003cem\u003eyagEFG\u003c/em\u003e while the latter acts as an activator of \u003cem\u003eyjhIHG\u003c/em\u003e. Although D-xylonate is known to induce these two operons, the effective inducer remains unknown to date. Through the construction of biosensors based on XynR and YjhI using \u003cem\u003esyfp2\u003c/em\u003e as a reporter gene, we showed that the true natural effector of the two D-xylonate catabolic operons is 2-keto-3-deoxy-D-xylonate, which is formed by dehydration of D-xylonate catalyzed by the dehydratases encoded by \u003cem\u003eyagF\u003c/em\u003e and \u003cem\u003eyjhG\u003c/em\u003e. Building on the finding that these two operons were also upregulated in \u003cem\u003eE. coli\u003c/em\u003e strains producing the non-natural platform molecule 2,4-dihydroxybutyric acid, we discovered that both XynR- and YjhI-based biosensors were responsive to the non-natural molecule 2-oxo-4-hydroxybutyric acid, harboring comparable characteristic performances in term of response threshold, sensitivity, cooperativity and dynamic response as the natural effector 2-keto-3-deoxy-D-xylonate. Given that the enzymatic steps involved in the production of this non-natural metabolite from C2 and C5/C6 carbons, catalyzed respectively by threonate dehydratase and homoserine transaminase, constitute bottlenecks in the synthetic pathways for the production of 2,4-dihydroxybutyric acid, we showed that the transcription factors XynR and YjhI can be repurposed as biosensors to select more active variants of these enzymes, thereby improving the production of this platform molecule from carbon sources.\u003c/p\u003e","manuscriptTitle":"Identification of 2-keto-3-deoxy-D-xylonate and 2-oxo-4-hydroxybutyrate as natural and artificial effectors of the transcription factors XynR and YjhI, respectively, and application to the development of biosensors for the bioproduction of 2,4-dihydroxybutyric acid in E. coli","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-16 11:26:36","doi":"10.21203/rs.3.rs-8830407/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a0750ca5-90a6-4d78-ae69-1dff327e6236","owner":[],"postedDate":"February 16th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-04T16:07:10+00:00","versionOfRecord":{"articleIdentity":"rs-8830407","link":"https://doi.org/10.1007/s00253-026-13840-y","journal":{"identity":"applied-microbiology-and-biotechnology","isVorOnly":false,"title":"Applied Microbiology and Biotechnology"},"publishedOn":"2026-05-01 15:57:52","publishedOnDateReadable":"May 1st, 2026"},"versionCreatedAt":"2026-02-16 11:26:36","video":"","vorDoi":"10.1007/s00253-026-13840-y","vorDoiUrl":"https://doi.org/10.1007/s00253-026-13840-y","workflowStages":[]},"version":"v1","identity":"rs-8830407","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8830407","identity":"rs-8830407","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}