Characterizing Genetic Circuit Components in E. coli towards a Campylobacter jejuni Biosensor

Campylobacter jejuni is responsible for most cases of bacterial gastroenteritis (food poisoning) in the United Kingdom. The most common routes of transmission are by contact with raw poultry. Current detection systems for the pathogen are time-consuming, expensive or inaccessible for everyday users. In this article we propose a cheaper and faster system for detection of C. jejuni using a synthetic biology approach. We aimed to detect C. jejuni by the presence of xylulose, an uncommon bacterial capsular saccharide. We characterized two sugar-based regulatory systems that displayed potential to act as tools for detection of xylulose. Using a two-plasmid reporter system in Escherichia coli, we investigated the regulatory protein component (MtlR) of the mannitol operon from Pseudomonas fluorescens. Our findings suggest that the promoter of mtlE is activated by MtlR in the presence of a variety of sugar inducer molecules, and may exhibit cross-activity with a native regulator of E. coli. Additionally, we engineered the L-arabinose transcriptional activator (AraC) of E. coli for altered ligand specificity. We performed site-specific saturation mutagenesis to generate AraC variants with altered effector specificity, with an aim to generate a mutant activated by xylulose. We characterized several mutant AraC variants which have lost the ability to respond specifically to the native L-arabinose effector. We promote this technique as a powerful tool for future iGEM teams to create regulatory circuits activated by novel small molecule ligands.


Introduction
Campylobacter jejuni is a Gram-negative, microaerophilic, corkscrew-shaped bacteria which has been implicated as being one of the most common causes of human gastroenteritis worldwide [1][2] [3]. Infection with C. jejuni causes common symptoms such as diarrhoea, abdominal pain, fever, headache, nausea and vomiting [3]. C. jejuni is harboured by poultry, though it has been reported in other meat products, raw milk, and in untreated drinking water [3]. The high prevalence of C. jejuni makes it an interesting target for synthetic biology-based solutions.
Traditional methods for detection of C. jejuni include culture-based techniques, which are relatively cheap to perform and require less training than other methods [4].
However, they are incredibly time and labour intensive and therefore, in recent years, a move has been made towards use of rapid detection testing [5]. Such techniques include enzyme immunoassay and lateral flow systems, which require only one to two hours to give a result [6]. However, use of these methods requires highly trained employees, and so detection of C. jejuni in an industrial or agricultural setting would require outsourcing to specialists. In addition, a comparison of three rapid detection systems demonstrated a high number of false negative results, which is a drawback when considering detection of C. jejuni to reduce the incidence of disease outbreaks [7].
To reduce incidences of food poisoning by C. jejuni, and to improve upon current methods of detection, we decided to create a biosensor that was able to detect the presence of C. jejuni quickly and accurately. We envisaged a two-part biosensor that would specifically require two sensory inputs associated with C. jejuni to report a reliable result. The first molecule we identified as a marker for C. jejuni was autoinducer-2 (AI-2) [9]. AI-2 is a secreted quorum-sensing molecule. However, many varied gram-positive and gram-negative bacterial species sense their population density and surrounding bacterial environment using this molecule [10]. On the other hand, this ubiquity meant that AI-2 gene regulation was well characterised, with prior iGEM teams having worked on the E. coli AI-2 quorum sensing regulatory system [11] [12]. Therefore, we searched for another marker with greater specificity to C. jejuni to use in tandem with AI-2.
Xylulose is a rare sugar found incorporated within the polysaccharide capsule of C.
jejuni [8]. The presence of xylulose is uncommon in bacterial polysaccharide capsules [8]. Additionally, the glyosidic bonds which incorporate xylulose were found to be extremely acid-labile [8], providing a possibility to release the molecule from the capsule and allow for whole-cell based detection. For the detection of xylulose two possible avenues were explored. One exploits the mannitol metabolism operon of Pseudomonas fluorescens [13]. It has been previously reported that xylulose acts as a direct inducer of the regulatory protein MtlR, activating transcription from the p mtlE promoter [14]. To investigate its suitability for biosensor, we characterized the p mtlE /MtlR regulatory system in E. coli.
Use of a gene regulatory system outside the context of the native organism may be problematic. For this reason, we aimed to construct an alternative xylulose sensor, by exploiting components of the L-arabinose operon, native to E. coli [15]. Several previous studies have shown that its regulatory protein, AraC, can be engineered to activate transcription in response to non-native small molecules [16][17] [18]. Sitesaturation mutagenesis of residues positioned within the ligand-binding pocket of AraC (Fig 1), coupled with fluorescence-based cell sorting, allowed isolation of AraC variants with altered effector specificity [16]. Based on these findings we aimed to use multiple site-saturation mutagenesis and fluorescence-based screening to generate mutant AraC responsive specifically to xylulose.

Materials and Methods
Standard molecular biology techniques.
All protocols used during this work, including a standardised set for routine laboratory techniques, are detailed in S3 File. All ligation reactions described herein were transformed into DH5α commercial chemically competent E. coli cells, plated on Lagar containing the required antibiotics, and grown at 37°C overnight. All plasmid constructs herein were verified by diagnostic restriction digest and DNA sequencing prior to use or further subcloning.
Plasmid design for expression of p mtlE /MtlR regulatory system All plasmids were constructed using BioBrick Standard Assembly. The biobrick p tet promoter (R0040) followed by a medium-strength ribosome binding site (RBS) B0032 was ordered for synthesis by Integrated DNA Technologies (IDT) as complementary oligos, that were annealed and ligated into a pSB1C3 plasmid backbone. mtlR coding sequence was ordered as a gBlock Gene Fragment by IDT and inserted behind the R0040+B0032 promoter/RBS part in pSB1C3. This produced the regulatory plasmid BBa_K2442202 (Fig 2).
The sequence of the p mtlE promoter with its native RBS was obtained from Liu et al.  Plasmid design for expression of p BAD /AraC regulatory system Fragments containing parts R0011 (LacI-regulated promoter) upstream of ribosomebinding site (RBS) B0032 were synthesised by IDT as oligonucleotides. Wild type araC sequence as described by Miyada et al. (1980) [19] was amplified from BBa_I0500 using primers araC_BBPre_F and araC_BBSuf_R (Table 1) to introduce BioBrick prefix and suffix to both ends of WT araC. The PCR product was subsequently inserted downstream of R0011 promoter and B0032 RBS in pSB1C3 backbone to generate the regulatory plasmid BB_K2442104 (Fig 3).
Minimal p BAD promoter was synthesised by IDT as a gBlock (see S1 Fig for full sequence). The fragment was ligated into pSB3k3, upstream of part BBa_I13500 (containing B0034 RBS and GFP). This resulted in the final reporter plasmid BBa_K2442102 (Fig 3).

Mutant AraC Library Construction
Site sirected mutagenesis by PCR was performed according to the protocol in S3 File.
QIAQuick PCR purification of the product was performed according to the Qiagen protocol (S3 File). The purified PCR product was ligated into pSB1C3 backbone, downstream of R0011 promoter and B0032 RBS. Ligation reaction was ethanolprecipitated, and transformed into electrocompetent DH5α by electroporation. This resulted in the araC mutant library. Colonies carrying the mutant library were washed off transformation plates with 1.5ml double distilled H 2 O, (ddH 2 O) then transferred to a 1.5ml tube. The cell pellet was centrifuged at room temperature, the supernatant was then discarded and pellet resuspended in 1.5ml ddH 2 O. This centrifugationresuspension step was repeated 3 times to remove agar plate debris. Plasmid DNA was extracted using the Qiagen Plasmid MiniPrep Kit with 2 of each PB buffer and PE buffer wash steps to ensure maximum extract purity. Sequencing primer VF2 was then added and sample sent for sequencing.

Characterization of the p mtlE /MtlR system activity in E. coli
To characterize activity of the p mtlE /MtlR in E. coli, we studied the levels of GFP fluorescence using a 96 well plate in the FLUOstar Omega plate reader (BMG Labtech). Each strain was grown to saturation in an overnight culture of LB with appropriate antibiotics for the relevant plasmids (chloramphenicol and kanamycin).
Then the culture was diluted 1:100 with fresh LB+antibiotic and placed into a black bottomed 96 well plate. All readings were performed in the plate reader at 37°C shaking at 200 RPM. GFP fluorescence was measured (excitation at 485nm and emission at 530nm) every hour for 8 hours.

Mutant Screening
Plasmid DNA containing the araC mutant libraries were transformed into DS941 strain E. coli carrying the reporter plasmid K2442102 (in pSB3K3 vector). Transformants were plated on LB agar medium containing chloramphenicol plus kanamycin (to select for K2442102 and K2442104) plus one of the tested inducers: arabinose, xylose or decanal. Concentrations were 40mM for arabinose and xylose and 2mM for decanal.
As a control test, transformants were also plated on LB medium containing chloramphenicol and kanamycin only. Fluorescence images of the conditional transformation plates were obtained using a GE-Healthcare Typhoon FLA-9500 laser scanner. Excitation was recorded at 473nm, emission was recorded using 520-540nm filter. Colonies which exhibited fluorescence were observed as dark colonies on the scan. Out of those which appeared dark on xylose or decanal plates, 200 were replica short-streaked onto plates containing xylose, arabinose, decanal, or no additive. Each colony of interest was picked and then immediately short-streaked onto each new condition plate using the same toothpick, into the same position using grids. This method of replica plating ensures the plates can later be aligned and short streaks of the same origin directly compared between the plates. After overnight incubation at 37°C, fluorescence scans of each plate were again obtained using the laser scanner.
Liquid culture fluorescence assay for AraC mutants Colonies of interest were inoculated into L-broth containing chloramphenicol and kanamycin. Each liquid culture was grown overnight at 37°C, shaking at 225 rpm. The following day the culture was diluted 1:100 into fresh L-broth containing the above antibiotics plus one of the following inducer conditions: of each culture to be tested was placed in a well of a clear-based 6-well plate, then incubated at 37°C shaking at 300 rpm for 12 hours in a BMG FLUOstar Omega fluorescence plate reader. GFP fluorescence of each culture was measured at 1 hour time intervals for the duration of the culture experiment, using excitation wavelength 485nm and emission wavelength 530nm. Cell growth was simultaneously tracked by measuring optical density at 600nm.

E. coli
From the literature, we identified a regulatory system that responds to xylulose -the mannitol-inducible promoter from P. fluorescens and its regulatory protein MtlR [14].
To utilize these parts in our dual-input biosensor, their activity needed to be characterized in E. coli. Two constructs were assembled: the regulatory plasmid with a constitutively active p tet promoter driving expression of the MtlR protein, and the reporter plasmid containing p mtlE promoter regulating expression of GFP (Fig 1).
To test whether MtlR can activate p mtlE in E. coli, we measured GFP fluorescence in E. coli carrying the reporter plasmid alone, or both the regulatory and reporter plasmids. The experiment was done in presence of 6 structurally similar sugars (ribose, fructose, xylose, mannitol, arabinose and sorbitol) [14], to investigate substrate specificity of MtlR. Xylulose itself was not tested due to budgetary constraints. Fluorescence levels were compared to basal fluorescence levels of DH5α cells, not expressing either plasmid. Cells expressing the reporter plasmid alone showed higher levels of fluorescence than empty cells, in presence of all sugars tested (Fig 4 and S1 File). This suggests that P mtlE may interact with native E. coli proteins.
GFP fluorescence levels further increased in cells expressing both MtlR and the reporter plasmid (Fig 4 and S1 File). This supports previous findings that MtlR functions as an activator of p mtlE in E. coli. Although we were unable to test xylulose, we found that p mtlE was induced in presence of a number of structurally similar sugars, showing highest response to ribose and sorbitol (Fig 4 and S1 File). We conclude that p mtlE promoter functions in E. coli, and MtlR acts as its activator. However, the P. fluorescens p mtlE promoter is not strictly regulated by MtlR when expressed in E. coli. Split regulatory components of the L-arabinose operon from E. coli are functional and can be used for construction of L-arabinose-inducible systems As the P. fluorescens MtlR regulatory system lacked specificity in E. coli, we aimed to utilize components of the L-arabinose operon from E. coli to generate a new tightly controlled xylulose-regulatory system. We chose to mutagenize the AraC protein to change its effector specificity, based on previous reports of successful engineering of AraC to respond to non-native inducers [16][17] [18]. The p BAD promoter is regulated by the AraC transcriptional regulator, which drives expression from p BAD only in presence of L-arabinose [15]. In nature the p BAD promoter overlaps with the araC coding region [20]. To allow for mutagenesis, we split araC from p BAD . The minimal p BAD was designed to retain all the sites required for AraC binding. The start codon of araC within p BAD has been changed from ATGàAGT (S1 Fig).
To test activity of minimal p BAD , we expressed regulatory and reporter plasmids in AraC-negative E. coli strain DS941 and plated the cells on LB medium. GFP fluorescence measurements demonstrated that AraC expressed from a separate plasmid can induce expression from minimal p BAD upon binding of arabinose, as cells exhibited fluorescence only in presence of arabinose (Fig 5 and S2 File). GFP fluorescence measurements revealed that minimal p BAD is inducible by L-arabinose 300-fold (Fig 5 and S2 File), showing improvement over previously characterized p BAD parts. The minimal p BAD promoter is tightly regulated by AraC expressed either from our regulatory plasmid, or from bacterial chromosome. Site-directed mutagenesis of the AraC protein provides a tool to develop biosensors responsive to non-native molecules We performed site-directed saturation mutagenesis of the araC gene, targeting four amino acids within the L-arabinose binding pocket of AraC. The protocol successfully generated a mutant library of ~24,000 AraC variants. Sequencing confirmed that NNS mutations were introduced at correct codon positions (residues 8, 24, 80 and 82; Fig   6). E. coli DS941 transformed with the mutant library and reporter plasmids were screened for GFP fluorescence in presence of 3 different inducer molecules that were available to us, to identify colonies with altered AraC/p BAD activity. Due to prohibitive cost, we were unable to test xylulose. Nevertheless, 4 colonies displaying altered expression patterns were identified and subsequently characterized (Fig 7 and S2 File). Two AraC mutants constitutively activated p BAD , without the requirement for Larabinose. The other two had lost the ability to respond to arabinose. Potentially, the latter two variants could be responsive to a yet unidentified compound.

Discussion
Due to lack of cheap and rapid detection methods for C. jejuni, we designed a biosensor responsive to a biomarker specific to the pathogen -xylulose. In our study we characterized activity of a mannitol-responsive regulator MtlR in E. coli. Liu et al.
(2015) [14] previously reported that mannitol and xylulose act as direct inducers of MtlR to activate transcription from the p mtlE promoter. Our results suggest that this system is functional when expressed in E. coli, with MtlR activating transcription from p mtlE . However, reporter expression was also induced independently of MtlR. This was achieved in presence of a variety of sugars structurally similar to mannitol and xylulose, including arabinose, fructose, xylose, ribose and sorbitol. Contradictory to how the system works in P. fluorescens [14], the strongest reporter expression was achieved in presence of sorbitol and ribose, rather than mannitol. Although the p mtlE was previously found to be activated by sorbitol independently of MtlR [14], no response to ribose has yet been observed. We suggest that other, yet unidentified proteins naturally found in E. coli act as transcriptional activators of p mtlE . Although we aimed to generate a xylulose-regulated system, we were unable to test our constructs in presence of xylulose due to its prohibitive cost. Nevertheless, the sugar would be required to screen for AraC mutants responsive to xylulose. We found that xylulose isomerase enzyme is capable of converting the cheaper sugar, xylose, into xylulose [21] [22]. We have considered development of a suitable expression plasmid, which in the future could be used to overexpress and purify the enzyme for production of xylulose in subsequent experiments (for full description see http://2017.igem.org/Team:Glasgow/XyluloseBiosynthesis).
Although xylulose is rarely found in bacterial capsules [8], potential contamination of the tested area by xylulose from other sources could lead to false positive results from our detector. For this reason, we designed our biosensor to detect two sensory inputs.
Apart from xylulose, we identified autoinducer-2 (AI-2) as another marker for C. jejuni [9]. AI-2 is a secreted quorum sensing molecule. In future development of the biosensor components, the detectors for both xylulose and autoinducer-2 would form two components of an AND gate that will ensure a positive result is given only when both xylulose and autoinducer-2 are present, improving specificity and accuracy of the detector (for full description see http://2017.igem.org/Team:Glasgow/ANDGate).
To increase efficiency of the whole-cell based biosensor, we went through several design iterations to create a device that would house the engineered bacteria to make the use of the biosensor simple and easy (for full description see

Conclusions
We conclude that the AraC mutagenesis protocol was successful at generating AraC variants with altered effector specificity. A larger scale mutant screen could result in identification of a xylulose-inducible AraC variant. For increased specificity, the xylulose-regulated p BAD /AraC system could be combined with an AI-2-sensing construct within an AND gate, and transformed into a host bacterium to produce a dual-input biosensor tool for rapid detection of C. jejuni.