Fortune telling: metabolic markers of plant performance

Abstract

Background: In the last decade, metabolomics has emerged as a powerful diagnostic and predictive tool in many branches of science. Researchers in microbes, animal, food, medical and plant science have generated a large number of targeted or non-targeted metabolic profiles by using a vast array of analytical methods (GC-MS, LC-MS, ¹H-NMR….). Comprehensive analysis of such profiles using adapted statistical methods and modeling has opened up the possibility of using single or combinations of metabolites as markers. Metabolic markers have been proposed as proxy, diagnostic or predictors of key traits in a range of model species and accurate predictions of disease outbreak frequency, developmental stages, food sensory evaluation and crop yield have been obtained.

Aim of review: (i) To provide a definition of plant performance and metabolic markers, (ii) to highlight recent key applications involving metabolic markers as tools for monitoring or predicting plant performance, and (iii) to propose a workable and cost-efficient pipeline to generate and use metabolic markers with a special focus on plant breeding.

Key message: Using examples in other models and domains, the review proposes that metabolic markers are tending to complement and possibly replace traditional molecular markers in plant science as efficient estimators of performance.

Keywords: Breeding; Metabolic marker; Metabolomics; Plant performance; Prediction.

Fig. 1

Strategy combining phenotyping, metabotyping and modeling for selection in order to find a few performing genotypes from a full panel of genotypes for a given criterion. Metabolic marker may optimize cost and speed of the process by (A) “metabotyping” and precision phenotyping of a diversity subpanel in a series of representative environmental conditions, (B) using collected data to model genotype performance. The model would generate a workable combination of (C) adapted growth scenario, sampling procedure and a small cost-efficient set of metabolic markers which would be used for (D) validation on the full panel of genotypes or for a further selection program. For the purpose of estimating costs, we consider 1 sample per genotype as a pool of 5 plants

Fig. 2

Key milestones for improving and developing the use of metabolic markers