Direct Observation and Automated Measurement of Stomatal Responses to Pseudomonas syringae pv. tomato DC3000 in Arabidopsis thaliana

Plant stomata plays a central role in the adaptation to stress conditions, so we need to measure stoma aperture to understand how plants respond to biotic and antibiotic stresses. However, stomatal aperture measurement is time-consuming and cumbersome. To measure stomatal aperture, we feel the epidermis and observe stomata under microscope.

Then we manually measure stomatal aperture, which takes a long time even though we’re experienced researchers. We challenged these barriers and developed a tool and a technique that can automatically measure stomatal aperture in intact Arabidopsis leaves. So we strongly believe that stomata imaging device and deep learning algorithm that we developed will facilitate functional analysis of stomata responses to various biotic and abiotic stresses because these technical advances significantly reduce time and human labor required for stomatal aperture measurement.

To begin, plant Arabidopsis thaliana seeds in soil. Grow them in a chamber with white fluorescent light. Next, inoculate a single colony of Pseudomonas syringae pv.

tomato DC3000 in five milliliters of King’s B liquid medium. Incubate the culture at 28 degrees Celsius with shaking at 200 revolutions per minute until the culture reaches the late logarithmic growth phase. Centrifuge the culture at 6, 000g for two minutes.

Then resuspend the bacterial pellet in one milliliter of sterile water. Next, pipette out the supernatant and resuspend the pellet in one milliliter of stomata-opening buffer. Measure the optical density of the suspension at 600 nanometers.

Dilute the bacterial suspension to an optical density of 0.2 with stomata-opening buffer containing 0.4%silicone surfactant. To inoculate the bacterial suspension, first place the Arabidopsis thaliana plants on a tray covered with a transparent lid under light for at least three hours. Remove the lid.

Then use an airbrush to spray the underside of the leaves with the bacterial suspension. Post-inoculation, incubate plants on a tray covered with a transparent lid to maintain a relative humidity of around 85%To begin, connect a portable stomatal imaging device to a computer equipped with image acquisition software. Use a piece of paper to carefully remove water droplets from the Pseudomonas syringae-inoculated leaves of the Arabidopsis thaliana plant.

Now open the top cover of the device. Place the leaf on the stage and shut the cover. Adjust the focus using the adjuster screw and click Save Image on the computer to capture the image.

To perform automated measurement of the stomatal aperture, open the Google Colaboratory notebook. Click on File followed by Save a Copy in Drive to make a local copy in Google Drive. When a new tab opens, close the tab of the original notebook.

Press the execute button under the environmental setup section in the notebook to import required libraries without unfolding the cell blocks. In the directory settings section, create three folders in Google Drive for analysis. Move the acquired images to the example result folder, categorizing them by treatment or sample for final graph generation.

Click on Download trained models to download the ONNX files of the trained models from Zenodo. Place the downloaded files in the model directory. Run the inference and measurement of stomatal aperture part of the notebook to quantify the stomatal aperture from individual images.

The resulting images with overlaid inference and the example result CSV file will be exported to the inference results directory. Execute the graph generation section to create a graph about the stomatal aperture ratio. Export this graph to the inference results directory.

A decrease in the stomatal aperture in PTO-inoculated plants compared with mock-inoculated plants was observed at one hour post-inoculation. At three hours post-infection, the stomatal aperture in both the PTO-inoculated plants and mock-inoculated plants were virtually the same. Automated measurement of the stomatal aperture took approximately five seconds to process one image, resulting in a 95%reduction in measurement time.