Integrating Transcriptome Analysis with Targeted Gene and Metabolite Profiling of Specialized Tissue/Cell Types: Lessons from Sweet Basil
In the sweet basil (Ocimum basilicum) variety “EMX,” leaf trichomes accumulate very high levels of the phenylpropene methylchavicol in the peltate glandular trichomes. To identify the genes involved in methylchavicol biosynthesis, cell-specific EST libraries from these trichomes were constructed and mined for sequences bearing homology to O-methyltransferases (Gang et al. 2002). One sequence bearing homology to Type I methyltransferases (MT) was identified and confirmed to catalyze the O-methylation of chavicol to form methylchavicol. The gene ObCVOMT was also shown to be highly expressed in the trichome of methylchavicol-accumulating varieties and devoid in a nonaccumulating variety. In the variety “SW,” leaf trichomes accumulate very high levels of the phenylpropene eugenol with no detectable methylchavicol. Mining the glandular trichome EST libraries of this variety for putative oxidoreductases enabled the discovery of isoeugenol synthase (ObEGSl), a PIP family of NADPH-dependent reductases, responsible for the formation of eugenol from coniferyl acetate (Koeduka et al. 2006).
Comparative Transcriptomics Analysis and Targeted Gene and Metabolite Profiling: Harnessing Within- and Between-Species Genetic Variation in Floricultural and Agricultural Crop Plants
The Phalaenopsis orchids are iconic house plants. However, modern-day varieties have been selectively breed to provide stunning, colorful, and long-lasting floral displays at the expense of the complete loss of floral scent production. Phalaenopsis bellina and P. aphrodite are two wild species with contrasting floral scent profiles. Flowers of P. bellina are highly scented and are dominated by the monoterpenes linalool, geraniol, and their derivatives, while P. aphrodite flowers are scentless to humans (Chuang et al. 2018).
To understand the molecular mechanisms underpinning this contrasting scent phenotype, RNA- seq transcriptomes of four flower developmental stages representing the onset, increase, peak, and decline of monoterpene emission were compared with that of P. aphrodite flowers at the floral bud and full-blossom stages (Chuang et al. 2018). One gene encoding geranyl diphosphate synthase (PbGDPS) and three predicted monoterpene synthases (PbTPS5, 9, and 10) were DE (i.e., highly upregulated in P. bellina and lowly expressed/absent in P. aphrodite). Mining of candidate TFs that share similar DE pattern revealed eight candidate TFs, of which five (i.e., PbbHLH4, PbbHLH6, PbbZIP4, PbERFl, and PbNACl) were later confirmed to transactivate the promoters of GDPS, TPS5, and TPS10 in dual-luciferase assays. Furthermore, all five TFs were able to induce the biosynthesis of floral terpenes (between 10- and 950-fold) in the scentless P. aphrodite when overexpressed in transient expression experiments.
Another example of the use of the comparative transcriptomic analysis in elucidating terpene biosynthesis comes from the rose (Rosa). The rose is perhaps best known as an iconic ornamental garden plant and as a cut flower, but is also an important floriculture crop that provides essential oils for the fragrance, food, and cosmetic industries. A typical rose scent is dominated by various monoterpene alcohols and 2-phenylethanol. In a breakthrough discovery, a novel terpene synthase- independent route to monoterpenes was recently reported (Magnard et al. 2015). The discovery was made by comparing the volatile, transcriptome, and gene expression of several rose varieties with distinct scent profiles. An unexpected enzyme that belongs to the Nudix hydrolase family (RhNUDXl) was prioritized and later confirmed to be involved in geraniol biosynthesis.
Expression of RhNUDXl was also shown to be tissue (petal)-specific, to increase concomitantly with scent emissions as the flower develops to maturity, and to be positively correlated with the occurrence of monoterpene alcohols (e.g., geraniol, nerol, citronellol) in ten varieties with contrasting scent profiles. Stable and transient transformation in two other heavily scented rose varieties (i.e., R. chinensis cv. Old Blush and R. hybrida The McCartney Rose) with RNA interference constructs unequivocally demonstrated that downregulation of RhNUDXl expression significantly reduced the total monoterpene content in petal tissues (due to decrease in geraniol derivatives) compared to the controls. All other volatile classes assayed were not affected. Further biochemical assays show that RhNUDXl uses GPP as the substrate, and hydrolyses one phosphate group to form geranyl monophosphate. Thus, formation of geraniol in rose petals must require a yet unidentified second phosphatase.
