Root colonisation experiments indicated that Pseudomonas sp. strain CT 364 was successfully able to colonise the roots of Arabidopsis. The next step was to develop Pseudomonas sp. as an endophytic chassis, so that roots could be colonised by a Pseudomonas sp. strain that had been genetically engineered to express gene(s) of interest. This involved three main objectives: identifying suitable selectable markers, establishing a transformation protocol for Pseudomonas sp., and the design, assembly and characterisation of a new genetic part.

Pseudomonas sp. (CT 364) was obtained from DSMZ, Germany (DSM No. 25356). The strain arrived freeze-dried and was revived according to the protocol recommended by DSMZ.As Pseudomonas sp. is a soil bacterium it has an optimum growth temperature of 28 °C. Pure cultures were obtained following revival after overnight incubation (Figure 1).

The first step in the chassis development was to identify antibiotics that Pseudomonas sp. was susceptible to. This would indicate which selectable markers were suitable for use in transformation procedures. Five antibiotics were tested, chloramphenicol, kanamycin, gentamicin, carbenicillin and streptomycin. Kanamycin, gentamicin and streptomycin are all aminoglycosides with similar but independent modes of action interfering with protein synthesis via ribosomal binding. Chloramphenicol prevents protein chain elongation and carbenicillin interferes with cell wall synthesis. Many Pseudomonas species are known to be resistant to chloramphenicol but this antibiotic was tested to determine compatibility with pSB1C3 (1). Carbenicillin and the aminoglycosides have been identified as active against many Pseudomonas species (2 and 3).

Antibiotic screening on tryptone soy agar (TSA) showed Pseudomonas sp. to be resistant to chloramphenicol, kanamycin and carbenicillin. Antibiotic concentrations of 50 and 100 µg/ml were tested with two replicates for each concentration. Lawns formed on agar containing 100 µg/ml of each antibiotic (Figures 2, 3 & 4) after 24 hours incubation at 28 °C.

Screening found Pseudomonas sp. to be susceptible to both streptomycin (Figure 5) and gentamicin (Figure 6). Antibiotic concentrations of 50 and 100 µg/ml were again tested with two replicates for each concentration. No colony forming units (CFUs) were visible on tryptone soy agar containing 50 µg/ml of either antibiotic inoculated with Pseudomonas sp. after 24 hours incubation at 28 °C.

After identifying which antibiotics were active against Pseudomonas sp. The next step was to identify working concentrations of these antibiotics to be used when selecting transformants. This was done by carrying out minimum inhibitory concentration (MIC) experiments where growth was tested against a range of antibiotic concentrations. To minimise variation between replicates MIC experiments were automated using an Opentrons OT-2 pipetting robot.

The results of MIC experiments showed a clear dose response between antibiotic concentration and growth of Pseudomonas sp. for both streptomycin and gentamicin (Figures 7 and 8). Gentamicin was found to be the more effective antibiotic with a concentration of 1.5 µg/ml sufficient to prevent growth. A minimum concentration was clear for gentamicin with a significant drop in growth rate between concentrations of 1.0 µg/ml and 1.5 µg/ml. The minimum inhibitory concentration for streptomycin was less clear, with a small drop in growth rate for each increase in concentration. A working concentration of 6.0 µg/ml was determined for streptomycin. A slight increase in absorbance was observed in the positive control for both antibiotics. This is likely due to release of compounds by bacterial cells upon death.

After identifying the MIC in liquid culture, the MICs in agar were determined for gentamicin as this was the antibiotic being taken forward for use in transformation. Concentrations of 2.0 µg/ml, 4.0 µg/ml (Figure 9) and 6.0 µg/ml (Figure 10) were tested as the working concentration was expected to be slightly higher than the 1.5 µg/ml MIC in liquid culture due to lower availability of antibiotic in agar. It was necessary to establish a working concentration in agar as cells would be plated onto agar following transformation. Gentamicin was selected as the working antibiotic as it is known to stunt Arabidopsis growth (4). Treatment of Arabidopsis with a gentamicin resistant transformant could prevent stunting of growth by breaking down gentamicin in the rhizosphere (5). Demonstrating that a plant colonising microbe can be used to change the physiology of the plant.

Agar assays showed that concentrations of 2.0 µg/ml and 4.0 µg/ml of antibiotic were insufficient to prevent growth of Pseudomonas sp. on agar. A lawn formed on agar containing 2.0 µg/ml and some colonies formed on agar containing 4.0 µg/ml (Figure 9). A gentamicin concentration of 6.0 µg/ml was sufficient to prevent growth of Pseudomonas sp. as no colonies formed at this concentration (Figure 10).

Once working concentrations had been determined for both antibiotics, transformation of Pseudomonas sp. could be attempted. Addgene plasmid number 79813 was selected for transformation protocol optimisation due to its Pseudomonas origin of replication and gentamicin resistance gene. The function of this plasmid was as a helper plasmid for conjugation. This plasmid could therefore also be used to optimise conjugation protocols for Pseudomonas sp. if transformation of larger plasmids is required. The plasmid was supplied by Addgene in E. coli and a Qiagen QIAprep Spin Miniprep Kit was used to extract the plasmid from overnight liquid cultures. Plasmid extractions were divided into multiple aliquots and a vacuum concentrator was used to remove water from some samples resulting in three solutions with varying DNA concentrations (Table1).

Electroporation was chosen as the first transformation method to be tested on Pseudomonas sp. primarily because it was the most utilised transformation method for Pseudomonas in available literature. It was also advantageous due to the quick procedure for preparing electrocompetent cells allowing high throughput experimentation. Heat-shock and conjugation were identified as potential other transformations methods should electroporation fail.

Four electroporation experiments were carried out, one for each DNA solution and a negative control where water was added in place of a DNA solution. Electrocompetent cells were prepared by carrying out five washes with 300 mM. Electroporation reaction mixes consisted of 100 µl of competent cells with 2 µl of plasmid DNA solution.

Gentamicin resistant colonies were obtained for each of the DNA solutions with solution 1 containing the highest DNA concentration yielding the highest number of CFUs (Figure 11) and solution 3 containing the lowest DNA concentration yielding the fewest colonies (Figure 13). Testing with higher DNA concentrations would be required to determine if more CFUs can be obtained. No colonies formed where the negative control had been plated into selective media (Figure 14). Colony forming units obtained from each miniprep are displayed in Table 2.

Table 2. Colonies formed from electrocompetent Pseudomonas sp. inoculated with solutions containing Addgene plasmid #79813

An additional MIC experiment was conducted to characterise gentamicin resistance of the transformant (Figure 15). This experiment was carried out on a 96-well plate using the OT-2 robot. More replicates were used than previously as fewer concentrations were being tested leaving more space on the plate, as a result standard error was reduced compared to previous MICs. A low concentration of 5.0 µg/ml was tested to determine if the transformant was significantly more resistant than the wild-type that is eliminated by a working concentration of 1.5 µg/ml. A 10 times concentration of 50 µg/ml was also tested to determine if there was an upper limit to resistance. The transformant was resistant to gentamicin at the highest concentration tested (50 µg/ml) whereas the wild type was unable to grow at the lowest concentration tested (5.0 µg/ml).

A composite part containing the antibiotic resistance gene aadA (BBa_K125520), an Anderson promoter (BBa_J23119), a strong RBS (BBa_B0034) and a native terminator (BBa_K2797001) new to the registry was designed (Figure 16). This part conferred resistance to streptomycin, the second antibiotic found to be active against Pseudomonas sp. The part was synthesised as a gBlock by IDT with compatible ends for insertion into the pSB1C3 backbone using Gibson assembly. This part was assembled into pSB1C3 for characterisation and submission to the registry. The composite part was inserted into the backbone using a NEBuilder Hi-Fi Gibson assembly kit. A vector:insert molar ratio of 1:2 was used; the reaction mix is shown in Table 3. A positive control assembly reaction was also carried out using a standard NEBuilder control DNA mix containing a backbone and two gBlocks. This was mixed with NEBuilder Hi-Fi assembly mix in a 1:1 ratio.

Following assembly, chemically competent E. coli cells were transformed by heat shock using each reaction mix. CFUs were obtained for both the streptomycin resistance Gibson assembly reaction mix (Figure 17) and the positive control (Figure 18). No colonies formed on agar inoculated with the negative control (Figure 19).

Following transformation, colonies were re-streaked onto agar containing either streptomycin (50 µg/ml) (Figure 20) or chloramphenicol (25 µg/ml) (Figure 21). This indicated that transformants containing the correctly assembled plasmid (BBa_K2797002 inserted into the pSB1C3 backbone) were resistant to both antibiotics. As the BBa_K2797002 and the pSB1C3 backbone confer resistance to streptomycin and chloramphenicol respectively.

Streptomycin resistance of the transformed E. coli was compared to the wild type (Figure 22). The transformant was resistant to streptomycin at the highest concentration tested (64 µg/ml). The wild-type E. coli showed greater resistance to streptomycin than Pseudomonas sp. with residual growth in wells containing both 16 µg/ml and 32 µg/ml of streptomyin.

Pseudomonas sp. was found to be susceptible to two antibiotics, streptomycin and gentamicin. Working concentrations were determined for both of these antibiotics using solid and liquid media. These data firstly aided the selection of a suitable plasmid for use in transformation protocols, and also helped to inform the design of a new part. The selected plasmid (Addgene plasmid 79813) was used in successful transformations of Pseudomonas sp. using electroporation, demonstrating that Pseudomonas sp. is genetically tractable and offers potential as an endophytic chassis. Finally, a new streptomycin resistance part, conferring strong resistance to streptomycin in E. coli, was designed, built and characterised.

The Pseudomonas sp. transformation protocol could be optimised by implementing the automated transformation optimisation methods established in our Measurement work. These methods have proved effective for the optimisation of E. coli transformation protocols, so we anticipate that the same methodologies could be applied to improving transformation efficiency in Pseudomonas sp. These optimised protocols could also be used to examine promoter strength in Pseudomonas sp. using the improved measurement devices constructed by the team. The new streptomycin resistance part (BBa_K2797002) could also be introduced into Pseudomonas sp. using optimised transformation procedures and used for selection of Pseudomonas sp. transformants.