Plant Phenotype Demarcation Using Nontargeted LC-MS and GC-MS Metabolite Profiling

Abstract

The characterization of the metabolome is a critical aspect in basic research and plant breeding. In this work, the Putative application of metabolomics for phenotyping closely related genotypes has been tested. Crude extracts were profiled by LC-MS and GC-MS, and mass data extraction was performed with XCMS software. Result validation was achieved with principal component analysis (PCA). The ability of the profiling methodologies to discriminate plant genotypes was assessed after hierarchical clustering analysis (HCA). Cluster robustness was assessed by a multiscale bootstrap resampling method. A better performance of LC-MS profiling over GC-MS was evidenced in terms of phenotype demarcation after PCA and HCA, Citrus demarcation was similarly achieved independently of the environmental conditions used to grow plants. In addition, when all different locations were pooled in a single experimental design, it was still possible to differentiate the three closely related genotypes. The presented methodology provides a fast and nontargeted workflow as a powerful tool to discriminate related plant phenotypes. The novelty of the technique relies on the use of mass signals as markers for phenotype demarcation independent of putative metabolite identities and the relatively simple analytical strategy that can be applicable to a wide range of plant matrices with no previous optimization.