Partial least squares model and design of experiments toward the analysis of the metabolome of Jatropha gossypifolia leaves: Extraction and chromatographic fingerprint optimization. Issue 6 (18th February 2016)

• Record Type: Journal Article
• Authors: Pilon, Alan Cesar; Carnevale Neto, Fausto; Freire, Rafael Teixeira; Cardoso, Patrícia; Carneiro, Renato Lajarim; Da Silva Bolzani, Vanderlan; Castro‐Gamboa, Ian
• Is Part Of: Journal of separation science. Volume 39:Issue 6(2016)
• Page Start: 1023
• Page End: 1030
• Publication Date: 2016-02-18
• Subjects: Chromatographic fingerprint -- Design of experiments -- Extraction -- Jatropha gossypifolia -- Metabolomics; Separation (Technology) -- Periodicals; Chromatographic analysis -- Periodicals; 543.089
• DOI: 10.1002/jssc.201500892
• Languages: English
• ISSNs: 1615-9306

Abstract

Abstract : A major challenge in metabolomic studies is how to extract and analyze an entire metabolome. So far, no single method was able to clearly complete this task in an efficient and reproducible way. In this work we proposed a sequential strategy for the extraction and chromatographic separation of metabolites from leaves Jatropha gossypifolia using a design of experiments and partial least square model. The effect of 14 different solvents on extraction process was evaluated and an optimized separation condition on liquid chromatography was estimated considering mobile phase composition and analysis time. The initial conditions of extraction using methanol and separation in 30 min between 5 and 100% water/methanol (1:1 v/v) with 0.1% of acetic acid, 20 μL sample volume, 3.0 mL min −1 flow rate and 25°C column temperature led to 107 chromatographic peaks. After the optimization strategy using i ‐propanol/chloroform (1:1 v/v) for extraction, linear gradient elution of 60 min between 5 and 100% water/(acetonitrile/methanol 68:32 v/v with 0.1% of acetic acid), 30 μL sample volume, 2.0 mL min −1 flow rate, and 30°C column temperature, we detected 140 chromatographic peaks, 30.84% more peaks compared to initial method. This is a reliable strategy using a limited number of experiments for metabolomics protocols.