hide
Free keywords:
confocal microscopy, fungal pathogens, machine learning
Abstract:
Laser scanning confocal microscopy’s ability to generate high-contrast 3D images has become essential to studying plant-fungal interactions. Techniques such as visualization of native fluorescence, fluorescent protein tagging of microbes, GFP/RFP-fusion proteins, and fluorescent labelling of plant and fungal proteins have been widely used to aid in these investigations. Use of fluorescent proteins have several pitfalls including variability of expression in planta and the requirement of gene transformation. Here we used the unlabeled pathogens Parastagonospora nodorum, Pyrenophora teres f. teres, and Cercospora beticola infecting wheat, barley, and sugar beet respectively, to show the utility of a staining and imaging technique that uses propidium iodide (PI), which stains RNA and DNA, and wheat germ agglutinin labeled with fluorescein isothiocyanate (WGA-FITC), which stains chitin, to visualize fungal colonization of plants. This method relies on the use of KOH to remove the cutin layer of the leaf, increasing its permeability. This permeability allows the staining solution to penetrate and efficiently bind to its targets, resulting in a consistent visualization of cellular structures. We have also used this staining technique in conjunction with machine learning to analyze fungal volume, which indicates the fitness of the pathogen in planta, as well as quantifying nuclear breakdown, an early indicator of programmed cell death (PCD). This technique is simple to use, robust, consistent across host species, and can be applied to any plant-fungal interaction. Therefore, this technique can be used to characterize model systems as well as non-model interactions where transformation is not routine.
