Tackling microbial threats in agriculture with integrative imaging and computational approaches

Abstract

Pathogens and pests are one of the major threats to agricultural productivity worldwide. For decades, targeted resistance breeding was used to create crop cultivars that resist pathogens and environmental stress while retaining yields. The often decade-long process of crossing, selection, and field trials to create a new cultivar is challenged by the rapid rise of pathogens overcoming resistance. Similarly, antimicrobial compounds can rapidly lose efficacy due to resistance evolution. Here, we review three major areas where computational, imaging and experimental approaches are revolutionizing the management of pathogen damage on crops. Recognizing and scoring plant diseases have dramatically improved through high-throughput imaging techniques applicable both under well-controlled greenhouse conditions and directly in the field. However, computer vision of complex disease phenotypes will require significant improvements. In parallel, experimental setups similar to high-throughput drug discovery screens make it possible to screen thousands of pathogen strains for variation in resistance and other relevant phenotypic traits. Confocal microscopy and fluorescence can capture rich phenotypic information across pathogen genotypes. Through genome-wide association mapping approaches, phenotypic data helps to unravel the genetic architecture of stress- and virulence-related traits accelerating resistance breeding. Finally, joint, large-scale screenings of trait variation in crops and pathogens can yield fundamental insights into how pathogens face trade-offs in the adaptation to resistant crop varieties. We discuss how future implementations of such innovative approaches in breeding and pathogen screening can lead to more durable disease control.

Keywords: Agriculture; Genome-wide association mapping; High-throughput phenotyping; Image analysis; Plant pathogens; Sustainability.

Fig. 1

The biology of plant-pathogens interactions. A) A wide range of insects feed on leaves as herbivores. B) Longitudinal root section and rhizosphere. Underground interactions can have significant effects on the stem phenotype and on the entire plant. Filamentous fungi and bacteria can have beneficial effects on the plant by facilitating the acquisition of phosphorus and nitrogen or detrimental effects when confronted by root pathogens. C) Dorsiventral section of a leaf colonized by different pathogenic filamentous fungi. From the left: fungi of the Fusarium, Aspergillus, Magnaporthe and Podosphaera genus. D) Filamentous fungi and oomycetes causing leaf infections. Filamentous pathogens can penetrate into the mesophyll through stomata. Once inside the mesophyll, pathogens colonize the entire tissue within days or weeks. Depending on the lifestyle (biotrophic or necrotrophic), pathogens secrete small proteins (called effectors) to manipulate plant cells. Plants detecting effectors (blue receptors) can mount a hypersensitive response (HR) leading to an autophagy-like cell death prevents the spread of the pathogen. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

High-throughput phenotyping techniques for plants and pathogens. A) Light in the visible spectrum can be used to detect changes in color and morphology of infected plant tissue. Infrared and short-wave infrared enable to record changes in water content, leaf thickness and photosynthetic efficiency. Long-wave infrared allows assessments of plant surface temperatures. Hyperspectral sensors capture multiple images across the range of 300–2500 nm. B) Imaging systems assess absorption, transmission, or reflectance characteristics of the incident electromagnetic radiation interacting with the plant surface. Diseased plant tissue often differs in reflectance compared to healthy tissue. Image analysis algorithms define contrasts between diseased and non-diseased leaf areas. C) Spatial scales of plant phenotyping approaches. D) Screening of pathogen populations can be performed in liquid cultures or on solid media. The most common experiments monitor growth rates by assessing culture densities over time or estimate the dose–response curves when exposed to antimicrobial compounds. Co-cultures of multiple microbes may be analyzed using two distinct emission/excitation pairs specific for each the co-cultured species.

Fig. 3

A comprehensive framework for determining the genetic basis of crop-pathogen interactions. Genetically diverse pathogen populations and crop cultivars from different geographies form the basis of the screening. Genome sequencing enables to conduct joint genome-wide association studies (GWAS) to determine the genetic architecture of virulence (in pathogens) and resistance (in crops). Global populations of both pathogen and crop will capture most relevant genetic variation. Beyond virulence, pathogen populations can be screened for loci underlying pesticide resistance, thermal adaptation and metabolite production (e.g. melanin). Identifying genetic correlations among pathogen traits facilitates the identification of pleiotropic genes governing trade-offs. Some illustrations were provided by biorender.com according to their usage conditions.