Integration of GC-MS Based Non-Targeted Metabolic Profiling with Headspace Solid Phase Microextraction Enhances the Understanding of Volatile Differentiation in Tobacco Leaves from North Carolina, India and Brazil

In this report, gas chromatography-mass spectrometry (GC-MS) based non-targeted metabolomics is used to develop appropriate headspace solid phase microextractions (HS-SPME) to enhance the understanding of volatile complexity of flue-cured tobacco leaves. Non-targeted metabolic profiling of GC-MS shows that the extraction condition of HS-SPME at 100?C for 30 min provides a better metabolite profile than other extraction conditions tested. GC-MS and principal component analyses (PCA) show that among five types of fibers tested, 100 μm polydimethylsiloxane (PMDS), 65 μm polydimethylsiloxane/divinylbenzene (PMDS/DVB) and 75 μm carboxen/polydimethylsiloxane (CAR/ PMS) provide a better reproducible metabolite profile. Based on an appropriate PDMS extraction condition optimized, we use GC-MS analysis and PCA to compare metabolite profiles in flue-cured leaves of tobacco plants grown in North Carolina, India and Brazil, respectively. The resulting data of PCA show that the global metabolic profiles in North Carolina samples are separated from those in Brazil and India samples, two groups of which are characterized by a partially overlapped pattern. Several peaks that were differentially accumulated in samples were annotated to known metabolites by deconvolution analysis, such as norsolanadione, solavetivone and rishtin. Norsolanadione is detected only in Brazil samples. Solavetivone is detected in samples of India and Brazil but not in those of North Carolina. Rishtin is detected in samples of North Carolina and India but not in Brazil samples. These data indicate that not only can a non-targeted metabolic profiling approach enhance the understanding of volatile complexity, but also can identify marker volatile metabolites in tobacco leaves produced in different growth regions.

A Genome-Wide Scenario of Terpene Pathways in Self-pollinated Artemisia annua

Scenarios of genes to metabolites in Artemisia annua remain uninvestigated. Here, we report the use of an integrated approach combining metabolomics, transcriptomics, and gene function analyses to charac- terize gene-to-terpene and terpene pathway scenarios in a self-pollinating variety of this species. Eightyeight metabolites including 22 sesquiterpenes （e.g., artemisinin）, 26 monoterpenes, two triterpenes, one diterpene and 38 other non-polar metabolites were identified from 14 tissues. These metabolites were differentially produced by leaves and flowers at lower to higher positions. Sequences from cDNA libraries of six tissues were assembled into 18 871 contigs and genome-wide gene expression profiles in tissues were strongly associated with developmental stages and spatial specificities. Sequence mining identified 47 genes that mapped to the artemisinin, non-amorphadiene sesquiterpene, monoterpene, triterpene, 2-C- methyl-D-erythritol 4-phosphate and mevalonate pathways. Pearson correlation analysis resulted in network integration that characterized significant correlations of gene-to-gene expression patterns and gene expression-to-metabolite levels in six tissues simultaneously. More importantly, manipulations of amorpha-4,11-diene synthase gene expression not only affected the activity of this pathway toward artemisinin, artemisinic acid, and arteannuin b but also altered non-amorphadiene sesquiterpene and genome-wide volatile profiles. Such gene-to-terpene landscapes associated with different tissues are fundamental to the metabolic engineering of artemisinin.