Identification and Validation of Novel Chromosomal Integration and Expression Loci in Escherichia coli Flagellar Region 1

Escherichia coli is used as a chassis for a number of Synthetic Biology applications. The lack of suitable chromosomal integration and expression loci is among the main hurdles of the E. coli engineering efforts. We identified and validated chromosomal integration and expression target sites within E. coli K12 MG1655 flagellar region 1. We analyzed five open reading frames of the flagellar region 1, flgA, flgF, flgG, flgI, and flgJ, that are well-conserved among commonly-used E. coli strains, such as MG1655, W3110, DH10B and BL21-DE3. The efficiency of the integration into the E. coli chromosome and the expression of the introduced genetic circuit at the investigated loci varied significantly. The integrations did not have a negative impact on growth; however, they completely abolished motility. From the investigated E. coli K12 MG1655 flagellar region 1, flgA and flgG are the most suitable chromosomal integration and expression loci.


Introduction
Escherichia coli K-12 is a model Gram-negative bacterium used as an intermediate and final destination chassis for a number of biotechnological applications and Synthetic Biology devices [1][2][3][4][5]. Integration and expression of genetic circuits from the E. coli chromosome is preferable to their introduction into the genome on plasmids due to the lower metabolic burden and no need for selective pressure [6]. Chromosomal integration strategies differ in their speed and flexibility. Recently developed methods include 'the one-step clonetegration' that relies on the bacteriophage integrase-mediated site-specific recombination between phage attachment (att) sites [7]. Besides transposons, phages and phage-derived elements [8,9], Red recombinase of the bacteriophage λ [10] is often used to engineer E. coli. The recently developed Red recombinase-based strategies include those based on the yeast mitochondrial homing endonuclease I-SceI [9], knock-in/knock-out (KIKO) vectors [11], and plasmid pSB1K3(FRTK) [12].
The lack of suitable chromosomal integration loci is among the main hurdles of the current E. coli engineering efforts. Traditionally phage attachment (att) sites have been used as integration targets [7]; however, att sequences are not present in all commonly-used E. coli strains. Integration target sites should be well-conserved, non-essential, well-characterized and highly expressed. Previously we have analyzed four genes of the E. coli K12 MG1655 flagellar gene region 3a and identified one locus that fulfills all these criteria and supports high efficiency integration [12]. The identification and characterization of the alternative integration sites is crucial for the development of a robust synthetic biology toolkit. Some applications require integrations of a number of genetic circuits into the chromosome (f.i. measuring synergistic effect of mutltiple genetic circuits on the phenotype or integration of the repressor and other genetic circuit(s) regulated by this repressor). Thus identification and validation of other suitable chromosomal integration and expression target sites would significantly help progress. Here we analyse open reading frames of the E. coli flagellar region 1. This region of the E. coli chromosome shows high probability of being occupied by RNA polymerase [13,14] and is therefore potentially suitable for integration and expression of genetic circuits. We integrated a genetic circuit harbouring the thermosensitive λ repressor [15] into five well-conserved open reading frames of the flagellar region 1. We propose flgA and flgG as the most suitable chromosomal integration and expression loci of the E. coli flagellar region 1.

Materials and Methods
Bacterial strains, plasmids, and growth conditions Bacterial strains and plasmids used in this analysis are listed in Table 1. Escherichia coli strains were routinely grown in Luria-Bertani (LB) medium. When needed, LB broth was supplemented with kanamycin (50 μg/ml) or ampicillin (100 μg/ml). Liquid cultures were grown in LB broth on a rotatory shaker at 200 r.p.m. Solid cultures were grown on agar plates for 24 hours. Depending on the requirements, both liquid and solid cultures were grown at 30°C, 37°C or 42°C.

Recombinant DNA methodology
PCR amplifications of DNA fragments were performed using Dream Taq master mix kit (Thermo Scientific) or Phusion DNA polymerase (Thermo Scientific) in 50 μl volumes according to the supplier's instructions. Oligonucleotide primers used in this analysis were synthesized by Integrated DNA Technologies (IDT). Plasmid isolations and gel extractions were carried out using Qiaprep Spin Miniprep kit (Qiagen) and Qiaquick Gel Extraction kit (Qiagen), respectively, according to the manufacturer's instructions. Sequencing of DNA fragments was performed by Source Bioscience (Cambridge, UK). DNA fragments were assembled using Plasmids pCP20 FLP recombinase helper plasmid [10] pKM208 Red recombinase controlled by lacZ [33] pSB1A1(GFP) Amp R , GFP under the control of λ promoter [12] pSB1K3(FRTKr) Kan R , λ repressor [12] the modified Gibson Isothermal Assembly method [16] as described previously (http://www. srcf.ucam.org/~wac26/gibson/index.html) [12]. The modified Gibson Isothermal Assembly reaction was incubated at 50°C for 60 min prior to E. coli transformation.

E. coli chromosomal integration
Electro-competent and chemically-competent E. coli were prepared by the modified Miller and Nickoloff [17] and Hannah methods [18], respectively. Integrations into the E. coli chromosome were performed by method described previously [12]. Briefly, DNA fragment composed of the thermosensitive λ repressor construct, kanamycin and FRT sites flanked by sequences homologous to the E. coli chromosome target genes was amplified by PCR and gel-purified. E. coli strain K12 MG1655 was transformed with pKM208 harboring resistance to ampicillin and selected on ampicillin plates at 30°C. Overnight culture (1:100 dilution) of E. coli strain K12 MG1655 with pKM208 was inoculated into LB with ampicillin and grown at 30°C to OD 600 of 0.2. 1 mM IPTG was added at this time point and the bacterial culture was grown to OD 600 of 0.4-0.6. Cells were washed twice with 10% (v/v) glycerol and resuspended in 100 μl of 10% glycerol per 100 ml of starting culture. Purified DNA fragment was introduced into electrocompetent cells by electroporation on a Bio-Rad micropulser. Transformants were selected on kanamycin plates at 37°C. Temperature sensitive pKM208 was cured out by incubation at 42°C. Successful integration into the E. coli was confirmed by PCR using the flanking primers.
Quantitative measurement of GFP fluorescence E. coli cultures were grown overnight and normalized to OD 600 of 0.05. 200 μl of these were transferred into flat-bottomed black 96 well plates (Greiner BioOne, UK). The plates were incubated in a Fluostar Omega fluorimeter (BMG Labtech, UK) at 30°C for 3 hours followed by 10 hours at 42°C for the measurement of the GFP fluorescence with an automatically repeated protocol. Excitation filter 485-12, emission filter EM520, gain 1400, double orbital shaking at 200 rpm, and cycle time 60 min were used for the GFP fluorescence measurement.

Plate reader measurement of the growth rate
The overnight E. coli cultures were normalized to OD 600 of 0.05. The total volume of 200 μl was transferred into flat-bottomed clear 96 well plates (Sterilin Sero-Well, UK). The plates were incubated in a microplate reader (Fluostar Omega) (BMG Labtech) for 24 hours at 37°C and 30°C with an automatically repeated protocol. The following parameters were used for the absorbance measurement (600 nm absorbance filter, cycle time 60 min, double orbital shaking at 500 rpm).

RNA isolation and RT-PCR
Isolate II RNA Mini Kit (Bioline) was used to isolate the total RNA from 10 9 E. coli cells grown into mid-exponential phase (OD 600 = 0.7

Motility assay
Motility assays were performed on motility agar plates. 13 cm motility agar plates composed of 100 ml motility agar (5 g NaCl, 10 g tryptone, 0.25% Bacto-Agar (Difco)) were let to set overnight and pre-warmed at 37°C before inoculation of bacteria. Overnight bacterial cultures (2 μl of the OD600 of 1.0) were spotted into the middle of the motility plates and incubated for 4-6 hours at 37°C.

Flagellar region 1 as target for chromosomal integration
The main hurdle of E. coli engineering is the shortage of suitable loci that support high efficiency integration and expression of integrated genetic circuits [12]. Previously we showed that fliT is the most suitable integration site from the four analysed open reading frames of the E. coli K12 MG1655 flagellar region 3a [12]. The identification and validation of the alternative integration sites is crucial, particularly for applications requiring integrations of multiple genetic circuits into the chromosome. Here we investigate flagellar region 1 as potential target for chromosomal integrations. ChIP-seq data [13,14] indicate high probability of the flagellar region 1 being occupied by RNA polymerase (Fig. 1), suggesting that genetic circuits integrated here would be highly expressed. Chromosomal integration sites should be well-conserved. flgA and flgI encode proteins for the flagellar basal-body periplasmic P ring assembly [21][22][23][24]. flgF and flgG encode flagellar components of the cell-proximal and the cell-distal part of basal body, respectively [25]. flgJ encodes B-N-acetylglucosaminidase that is essential for the bacterial flagella biogenesis [26]. The location of flgA, flgF, flgG, flgI, and flgJ in the E. coli K12 MG1655 flagellar region 1 are shown in Fig. 1. We used our streamlined Red recombinasebased chromosomal integration protocol [12] to integrate a genetic circuit (Repr-ts-1) consisting of the thermosensitive λ repressor expressed using a strong constitutive promoter, a RBS and a terminator taken from the iGEM Parts Registry into the investigated target genes flgA, flgF, flgG, flgI, and flgJ of the E. coli flagellar region 1.

Chromosomal integrations into flgA and flgG occur with the highest frequency
Chromosomal integration efficiency differs significantly between integration sites, suggesting that to find the most suitable loci, the integration efficiencies have to be determined experimentally [12]. Genetic circuit Repr-ts-1 was integrated into five targeted open reading frames of the E. coli flagellar region 1: flgA (flgAi), flgF (flgFi), flgG (flgGi), flgI (flgIi), and flgJ (flgJi). Integration into the chromosome was verified by diagnostic PCR with flanking primers (S1 Fig.) and DNA sequencing. Primer sequences homologous to the target genes that were used to amplify Repr-ts-1 and the diagnostic flanking primers are listed in S1 Table. The integration efficiencies for the five targeted integration loci differed significantly. The integration of Repr-ts-1 into flgA and flgG occured with the highest efficiency (Fig. 2). The integrations into both flgA and flgG generated two times and four times more recombinants per μg DNA than integrations into flgF/ flgI, and flgJ, respectively. As flgF, flgG, flgI, and flgJ are part of one operon, while flgA is located in another operon, the observed differences in the integration efficiencies are not operon-specific. Notably, integrations into the analyzed genes of the E. coli flagellar region 1 occured with the lower efficiency than integrations into the best locus of the previously analyzed region 3a, fliT [12]. Integrations into fliT yielded approximately twice the number of recombinants than integrations into flgA and flgG (Fig. 2).  (Fig. 3). This is in contrast with the previously analysed flagellar region 3a where integrations into two of the four analysed loci had only a minor effect on motility [12]. As integrations should not inhibit growth, the target loci can not be located within essential genes [1,5,[27][28][29]. The growth rates of the engineered strains harboring Repr-ts-1 integrations (flgAi, flgFi, flgGi, flgIi, and flgJi) and E. coli K12 MG1655 wild type were measured with the microplate reader (Fluostar Omega). Integrations into all five target loci did not have a negative impact on the growth at both the permissive (30°C) and the restrictive (37°C) conditions for the Repr-ts-1-bourne thermosensitive repressor (Fig. 4). Some of the engineered strains (flgAi, flgFi, flgGi, flgIi) grew to a higher final density than the wild type strain at permissive temperatures, while the growth of flgJi and the wild type strain was similar at both conditions (Fig. 4). This result is similar to that obtained for flagellar region 3a where strains with integrations grew to a higher density than the wild type strain [12]. The abolished motility and good growth (particularly of flgAi, flgFi, flgGi, flgIi) confirm the suitability of the tested loci for chromosomal integrations. The high probability of the E. coli K12 MG1655 flagellar region 1 being occupied by RNA polymerase ( Fig. 1) suggests that genetic circuits integrated here would be strongly expressed. We verified this by measuring relative expression of flgA, flgF, flgG, flgI, and flgJ by RT-PCR (Fig. 5A). The relative expression of the two integration target loci, flgF and flgG, was higher (2-3 fold) than the mean expression of the housekeeping genes arcA and rpoD [30,31]. The relative expression of flgI and flgJ was lower, while the relative expression of flgA was not different from the mean expression of the housekeeping genes (Fig. 5A). RT-PCR was also employed to asses the relative expression of the genetic circuit integrated into flgA, flgF, flgG, flgI, and flgJ. Repr-ts-1 was expressed strongly in all five target loci of the E. coli flagellar region 1 (Fig. 5B).
The highest expression of Repr-ts-1 was detected when integrated into flgA and flgF. The expression at these sites was 4-5 fold higher than the mean expression of the housekeeping genes (Fig. 5B). Notably, the expression of Repr-ts-1 was significantly higher (3-4 fold) at the best locus of the previously analyzed flagellar region 3a, fliT [12] (Fig. 5B). This could be due to the higher probability of the E. coli K12 MG1655 flagellar region 3a being occupied by RNA polymerase [12][13][14]. To assess the expression of the integrated genetic circuit we measured the repressive activity of the Repr-ts-1-bourne thermosensitive repressor on the GFP expression (Fig. 6). We introduced plasmid pSB1A1(GFP) into E. coli harboring chromosomal integrations flgAi, flgFi, flgGi, flgIi, and flgJi. pSB1A1(GFP) harbours GFP-encoding gene  Table. doi:10.1371/journal.pone.0123007.g004 downstream of the pR promoter that is controlled by the Repr-ts-1-bourne repressor. The repressive activity of the integrated genetic circuit was quantified by measuring GFP expression over time with the multiwell plate fuorimeter (Fluostar Omega). At 30°C GFP was not expressed, while increased temperature (42°C) led to GFP expression (Fig. 6).

Conclusions
The E. coli flagellar genes are suitable targets for integration of genetic circuits. Previous investigation of the E. coli K12 MG1655 flagellar region 3a led to the identification of one locus that supports both high integration efficiency and strong expression of the integrated genetic circuit [12]. The identification and validation of other suitable loci would significantly aid progress. The aim of this analysis was to identify and validate suitable integration and expression sites within E. coli flagellar region 1, which has good transcriptional profile and high probability of being occupied by RNA polymerase. We integrated genetic circuit Repr-ts-1 into five open reading frames of the flagellar region 1 (flgA, flgF, flgG, flgI, and flgJ) that are well-conserved among commonly-used E. coli strains. The five targeted genes of the flagellar region 1 varied significantly in the efficiency of integration and expression of the integrated genetic circuit. The integrations did not have a negative impact on growth; however, they all abolished motility. This analysis shows that flgA and flgG are the most suitable chromosomal integration sites of the investigated E. coli K12 MG1655 flagellar region 1 due to high efficiency of integration and strong relative expression. Expression of the integrated genetic circuit at flgF is higher than expression at flgG; however, the integration efficiency at flgF is only half of that measured at flgA and flgG. In conclusion, the presented study of the E. coli flagellar region 1 contributes two novel sites for the robust integration and expression of synthetic gene circuits.