Non-Invasive, Bioluminescence-Based Visualisation and Quantification of Bacterial Infections in Arabidopsis Over Time.

Abstract

Plant-pathogenic bacteria colonise their hosts using various strategies, exploiting both natural openings and wounds in leaves and roots. The vascular pathogen Xanthomonas campestris pv. campestris (Xcc) enters its host through hydathodes, organs at the leaf margin involved in guttation. Subsequently, Xcc breaches the hydathode-xylem barrier and progresses into the xylem vessels causing systemic disease. To elucidate the mechanisms that underpin the different stages of an Xcc infection, a need exists to image bacterial progression in planta in a non-invasive manner. Here, we describe a phenotyping setup and Python image analysis pipeline for capturing 16 independent Xcc infections in Arabidopsis thaliana plants in parallel over time. The setup combines an RGB camera for imaging disease symptoms and an ultrasensitive CCD camera for monitoring bacterial progression inside leaves using bioluminescence. The method reliably quantified bacterial growth in planta for two bacterial species, that is, vascular Xcc and the mesophyll pathogen Pseudomonas syringae pv. tomato (Pst). The camera resolution allowed Xcc imaging already in the hydathodes, yielding reproducible data for the first stages prior to the systemic infection. Data obtained through the image analysis pipeline was robust and validated findings from other bioluminescence imaging methods, while requiring fewer samples. Moreover, bioluminescence was reliably detected within 5 min, offering a significant time advantage over our previously reported method with light-sensitive films. Thus, this method is suitable to quantify the resistance level of a large number of Arabidopsis thaliana accessions and mutant lines to different bacterial strains in a non-invasive manner for phenotypic screenings.

Plant phenotyping relevance

植物感染を非侵襲的に画像化・定量するフェノタイピング装置とPython解析パイプラインを開発し、複数の細菌感染で検証しているため、方法が中心的です。

Code and data availability

The paper's Python image analysis pipeline (Digital phenotyper) for quantifying bioluminescent bacterial infection in Arabidopsis is explicitly and publicly deposited by the authors on GitHub.