One of our project aims was to develop an endophytic chassis organism. To ensure that we were working with a suitable organism, we needed to determine whether or not our selected organism, Pseudomonas sp. strain CT 364, was able to colonise roots as both a wild type and a transformant. To do this it was necessary to develop root colonisation protocols, explore methods to detect the endophyte within plant roots and ensure the transformed endophyte is not detrimental to plant growth.

The first step was to see if Pseudomonas sp. could be used to colonise the roots of Arabidsopsis thaliana. We looked into methods that involved root wounding, and the subsequent submergence of wounded roots into liquid endophyte cultures. However, engagement with GrowUp Urban Farms led us to consider alternative routes. This stakeholder engagement brought two issues to our attention: firstly, root wounding methods (as we initially intended) are not accessible to farmers and it was suggested that an alternative method of seed coating may be more appropriate. The second issue raised was that, from a commercial point of view, there is a concern that the genetically modified  Pseudomonas sp.  would be present in the leaves which may be harvested and eaten.

To address these concerns the team ensured to investigate the presence of the Pseudomonas sp. in both the root and stem / leaves. Furthermore, the team shifted from using a wounding inoculation method to a seed-coating method.

The first set of results came from seed-coating experiments, where 96 Arabidopsis thaliana seeds were surface sterilised and coated in a liquid culture of Pseudomonas sp. wild type. Once germinated, 8 seedlings were selected and surface sterilised before being cut at the root and leaves for plating on typtone-soya agar (TSA) plates. Wild type Pseudomonas sp. was re-isolated in pure culture from all of these samples and no Pseudomonas sp. was re-isolated from a set of 8 control Arabidopsis thaliana seedlings that had not been treated (Figure 1). These results showed that seed-coating was an appropriate inoculation method which makes our project more accessible for commercial use.

Figure 1. Pseudomonas sp. on a TSA plate that has been re-isolated from Arabidopsis thaliana seedlings which have been inoculated by the seed-coating method suggested by GrowUp Urban Farms (left), and a plate showing no Pseudomonas sp. isolates amongst the wild endophytes (Right).

The team decided that the most valuable way to assess the endophytic relationship between Pseudomonas sp. and its host plant was to use microscopy. The team sought expertise from Dr Vasilios Andriotis who advised bright field and fluorescence microscopy techniques.

Using seedlings inoculated with wild type Pseudomonas sp. as a positive control, initial microscopy used only water to mount the samples onto the microscope slide. This method was unsuitable for visualising bacteria in the root as bacteria were indistinguishable from the plant (Figure 8). Our Pseudomonas sp. strain produces a fluorescent siderophore causing green fluorescence so we attempted to view the sample through a GFP filter, this was unsuccessful as the auto-fluorescence of plant tissue was too great to see the bacteria. To overcome this issue, 4′,6-diamidine-2′-phenylindole dihydrochloride (DAPI) stain was used. DAPI is a highly-specific cell permeable fluorescent probe for double-stranded DNA (absorption 358 nm; emission 461 nm) – we used this to stain nucleic acids in the bacteria and visualise Pseudomonas sp. in our samples. DAPI was added to distilled water at a concentration of 1 microlitre per millilitre DAPI, this was used to mount the sample and a DAPI filter was used to visualise the sample. Using a 100X immersion objective, bacteria were visible and this revealed a biofilm was present on the surface of the root (Figure 3), however, it was unclear if the bacteria were inside the root. To minimise the risk of non-specific root-microbe interactions, seedlings were washed at least three times in distilled water before imaging. Examination of the cleaned seedlings was challenged by the fact that root hair cells had wrapped around the main root during the washing process meaning that it was unclear if bacteria were inside the main root or merely on the surface.

To overcome the issue of root hair cells wrapping, the team used 15 % glycerol in place of water for mounting samples. This was highly successful as the viscosity of the glycerol caused root hair cells to spread and revealed clear signs of colonisation. Pseudomonas sp. was present in the intercellular spaces along both the root and hypocotyl. These bacteria were still motile and we could see them moving in real time.

Figure 2. Bright field microscopy of a DAPI stained Arabidopsis thaliana. Wild type Pseudomonas sp. is visible in the intercellular spaces.The seedling was imaged under bright field (BF) and under fluorescence (Fluoro) simultaneously, and photographed with a monochrome camera. The image is an overlay of the two channels (BF and Fluoro). DAPI fluorescence is shown in green

Figure 3. Fluorescence microscopy with a DAPI filter, Pseudomonas sp. is visible on the surface of root hair cells.

We were able to take images of the bright field microscopy with and without a DAPI filter (Figure 4 and 5 respectively) before merging the images to provide a clear visual representation of how the bacteria live within the plant (Figure 6).

Figure 4. Bright field microscopy showing a Pseudomonas sp. inoculated Arabidopsis thaliana root.

Figure 5. This is the same image as Figure 4 however this time a DAPI fluorescence filter is used, showing Pseudomonas sp. in the tissue.

Figure 6. A Z-stack was made from the sample in Figure 4 and 5 under both bright field and fluorescence microscopy. This Z-stack was then used to make a maximum projection image, combining the two.

As we had a positive control in the form of wild type Pseudomonas sp. the team decided a to use E. coli DH5α as a negative control since it is not known to be a plant coloniser. A selection of seedlings inoculated by the seed-coating method were again selected for microscopy. Microscopy revealed that though E. coli was present on the root surface in small numbers, there was no sign of an endophytic relationship or a clear pattern in bacterial cell accumulation like that of Pseudomonas sp., seen clearly in the difference between Figure 5 and Figure 10. These results were very valuable and showed us what our transformant should and should not look like when examined.

Figure 7. A maximum projection image combining bright field and fluorescence microscopy, showing the root of our negative control Arabidopsis thaliana where E. coli DH5α was visible on the surface of the root but not inside the root as an endophyte.

Figure 8.Bright field microscopy showing the root surface of our negative control Arabidopsis thaliana.

Figure 9. Microscopy with DAPI fluorescence filter showing the root surface of our negative control Arabidopsis thaliana

Figure 10. A maximum projection image combining bright field and fluorescence microscopy, E. coli are visible on the root surface of Arabidopsis thaliana seedling but there is no evidence of an an endophytic relationship or any pattern to bacterial cell accumulation.

A selection of Pseudomonas sp. transformant-inoculated seedlings were taken for microscopy, again seedlings were washed with distilled water and mounted on a microscope slide with the same solution of DAPI and glycerol. Endophytic Pseudomonas sp. were visible inside the root and hypocotyl of both Arabidopsis thaliana and Eruca sativa seedlings showing that transformation had not altered the bacteria’s ability to colonise. This showed that our endophytic chassis was a plant coloniser and works as expected.

Figure 11. A maximum projection image, combining fluorescence and bright field microscopy of an Arabidopsis thaliana seedling root, showing transformed Pseudomonas sp. living as an endophyte in the intercellular spaces.

Figure 12. Bright field microscopy of an Arabidopsis thaliana seedling root inoculated with Pseudomonas sp. transformant.

Figure 13. A Microscope image with DAPI fluorescence filter of an Arabidopsis thaliana seedling root inoculated with Pseudomonas sp. showing the bacterium.

Figure 14. A maximum projection image made from Z-stacks allowing Figure 12 and Figure 13 combination, showing transformed Pseudomonas sp. in the intercellular spaces of an Arabidopsis thaliana seedling.

Germination Experiment

In order to assess the effect of our transformed Pseudomonas sp. on seed germination and growth, an experiment was devised to compare the effects of wild type Pseudomonas sp. to a transformant containing a gentamicin resistance cassette (Addgene plasmid #79813). 22 seeds were coated in the wild type Pseudomonas sp. and another 22 in the transformant. These seeds were the plated onto water agar supplemented with gentamicin at a concentration of 70 µg/ml (1) and allowed to germinate on the windowsill for 2 weeks.

Figure 15. Wild type Pseudomonas sp. inoculated (left) and transformant Pseudomonas sp. inoculated (right) Arabidopsis thaliana seedlings after 2 weeks of growth.

Figure 16. Wild type Pseudomonas sp. inoculated (left) and transformant Pseudomonas sp. inoculated (right) Arabidopsis thaliana seedlings after 2 weeks of growth.

Figure 17. Wild type Pseudomonas sp. inoculated (left) and transformant Pseudomonas sp. inoculated (near side) Arabidopsis thaliana seedlings after 2 weeks of growth.

After 2 weeks of growth, there were no obvious signs of impairment to growth between the two treatments with no discolouration or noticeable size difference. In terms of germination, 21/22 seeds germinated when coated with wild-type Pseudomonas sp. and 20/22 germinated when coated in transformed Pseudomonas sp., indicating no detriment to germination and growth.

Conclusions

The results of our root colonisation experiments clearly demonstrate that our proposed endophytic chassis, Pseudomonas sp.,  is able to colonise both the roots and leaves of  Arabidopsis thaliana seedlings, following a seed coating treatment. Furthermore, a Pseudomonas sp. transformant, transformed with a plasmid conferring gentamicin resistance, was also used to successfully colonise Arabidopsis thaliana seedlings using the same method, with no negative impacts on seed germination or plant growth.

Microscopy analysis was used to successfully visualise both wild type and transformed Pseudomonas sp. which had colonised Arabidopsis thaliana seedling roots. We also observed that the bacterium colonises not only the root but also the hypocotyl and is able to form a biofilm on the root surface, another level to the plant-microbe association.

Finally, our germination assays demonstrated that the germination rate of Arabidopsis thaliana seeds which had undergone a seed coating treatment with transformed Pseudomonas sp. did not differ to the germination rate of seeds which had undergone a seed coating treatment with the wild type strain. This provides further evidence that the use of an engineered bacterial endophyte is suitable for use in our proposed system without negatively impacting plant growth.