AUTHOR=Budzinski Ilara Gabriela Frasson , de Moraes Fabricio Edgar , Cataldi Thais Regiani , Franceschini Lívia Maria , Labate Carlos Alberto TITLE=Network Analyses and Data Integration of Proteomics and Metabolomics From Leaves of Two Contrasting Varieties of Sugarcane in Response to Drought JOURNAL=Frontiers in Plant Science VOLUME=10 YEAR=2019 URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2019.01524 DOI=10.3389/fpls.2019.01524 ISSN=1664-462X ABSTRACT=

Uncovering the molecular mechanisms involved in the responses of crops to drought is crucial to understand and enhance drought tolerance mechanisms. Sugarcane (Saccharum spp.) is an important commercial crop cultivated mainly in tropical and subtropical areas for sucrose and ethanol production. Usually, drought tolerance has been investigated by single omics analysis (e.g. global transcripts identification). Here we combine label-free quantitative proteomics and metabolomics data (GC-TOF-MS), using a network-based approach, to understand how two contrasting commercial varieties of sugarcane, CTC15 (tolerant) and SP90-3414 (susceptible), adjust their leaf metabolism in response to drought. To this aim, we propose the utilization of regularized canonical correlation analysis (rCCA), which is a modification of classical CCA, and explores the linear relationships between two datasets of quantitative variables from the same experimental units, with a threshold set to 0.99. Light curves revealed that after 4 days of drought, the susceptible variety had its photosynthetic capacity already significantly reduced, while the tolerant variety did not show major reduction. Upon 12 days of drought, photosynthesis in the susceptible plants was completely reduced, while the tolerant variety was at a third of its rate under control conditions. Network analysis of proteins and metabolites revealed that different biological process had a stronger impact in each variety (e.g. translation in CTC15, generation of precursor metabolites, response to stress and energy in SP90-3414). Our results provide a reference data set and demonstrate that rCCA can be a powerful tool to infer experimentally metabolite-protein or protein-metabolite associations to understand plant biology.