Terpenoid metabolism in wild-type and transgenic Arabidopsis plants

Abstract

Volatile components, such as terpenoids, are emitted from aerial parts of plants and play a major role in the interaction between plants and their environment. Analysis of the composition and emission pattern of volatiles in the model plant Arabidopsis showed that a range of volatile components are released, primarily from flowers. Most of the volatiles detected were monoterpenes and sesquiterpenes, which in contrast to other volatiles showed a diurnal emission pattern. The active terpenoid metabolism in wild-type Arabidopsis provoked us to conduct an additional set of experiments in which transgenic Arabidopsis overexpressing two different terpene synthases were generated. Leaves of transgenic plants constitutively expressing a dual linalool/nerolidol synthase in the plastids (FaNES1) produced linalool and its glycosylated and hydroxylated derivatives. The sum of glycosylated components was in some of the transgenic lines up to 40- to 60-fold higher than the sum of the corresponding free alcohols. Surprisingly, we also detected the production and emission of nerolidol, albeit at a low level, suggesting that a small pool of its precursor farnesyl diphosphate is present in the plastids. Transgenic lines with strong transgene expression showed growth retardation, possibly as a result of the depletion of isoprenoid precursors in the plastids. In dual-choice assays with Myzus persicae, the FaNES1-expressing lines significantly repelled the aphids. Overexpression of a typical cytosolic sesquiterpene synthase resulted in the production of only trace amounts of the expected sesquiterpene, suggesting tight control of the cytosolic pool of farnesyl diphosphate, the precursor for sesquiterpenoid biosynthesis. This study further demonstrates the value of Arabidopsis for studies of the biosynthesis and ecological role of terpenoids and provides new insights into their metabolism in wild-type and transgenic plants.

Figure 1.

Detection of Arabidopsis Headspace Volatiles by SPME, Tenax Trapping/Elution, and Tenax Trapping/Thermodesorption. (A) The SPME method permits the detection of volatiles from only two fully expanded, detached rosette leaves of 4-week-old Arabidopsis plants. Using an autosampler that automatically exposes a fiber to the sample headspace and introduces it into the injection port of the GC-MS apparatus, 40 Arabidopsis lines could be screened during an overnight run. (B) The Tenax system for the detection of headspace volatiles from intact Arabidopsis plants. Air is drawn through the inlet through a Tenax matrix, and the headspace volatiles emitted by the plant are absorbed at the outlet on a second Tenax matrix. Volatiles are eluted from the Tenax using pentane:diethyl ether (4:1) and are analyzed by GC-MS. In this study, headspace sampling was conducted for 24 h. (C) Tenax trapping for the detection of headspace volatiles from intact Arabidopsis plants coupled to an automatic thermodesorption system. Filtered air is introduced to the glass container headspace, volatiles are trapped on Tenax, and after trapping, volatiles are released by thermodesorption and cryofocused before their introduction into the GC-MS apparatus. The sampling and GC-MS analysis were performed automatically, so that detailed emission patterns were obtained during a certain period of time. (D) Typical chromatogram obtained after performing the analysis described in (C) and sampling headspace volatiles emitted by two flowering Arabidopsis plants. A subset of the volatiles are depicted: peak 1, linalool; peak 2, nonanal; peak 3, decanal; peak 4, impurity; peak 5, β-caryophyllene; peak 6, impurity; peak 7, thujopsene; peak 8, α-humulene; peak 9, α-farnesene; peak 10, β-bisabolene; peak 11, (Z)-nerolidol; and peak 12, (E)-nerolidol.

Figure 2.

Diurnal Emission of Endogenous Terpenoids and the Heterologous Linalool in Flowering Arabidopsis Plants. Volatiles trapped on Tenax were analyzed using thermodesorption and injection into the GC-MS apparatus. Sampling was performed every hour during 48 h with two flowering FaNES1-expressing plants. (A) Three endogenous monoterpenes showing a diurnal pattern of emission. (B) Three endogenous sesquiterpenes showing a diurnal pattern of emission. (C) Diurnal emission pattern of linalool produced by the expression of FaNES1.

Figure 3.

Transformation of Arabidopsis Plants with a Plastid-Targeted FaNES1. (A) Fusion of the 61–amino acid (aa) N terminus encoded by the wild strawberry FvNES1 gene (A. Aharoni and H.J. Bouwmeester, unpublished data) to the coding region of the cultivated species FaNES1 gene fused to the GFP. L/MID is a conserved motif in a large number of monoterpene synthases. The RR and L/MID are conserved motifs in a large number of monoterpene synthases. (B) Testing the GFP localization of the pTAR-GFP construct shown in (A). RpoT;3 is a positive control for targeting to plastids (Hedtke et al., 1999), whereas pOL-LT-GFP (full name, pOL-LT-GFP-L64T65; a gift from Ian Small) is the basic vector used to generate the pTAR-GFP construct in which GFP is localized to the cytosol and the nucleoplasm. GFP fluorescence was detected in plastids (p) with both pTAR-GFP and RpoT;3 and in the cytosol (c) with pOL-LT-GFP, as indicted by the arrows. CA, chlorophyll autofluorescence detected in channel 1; GFP, green fluorescent protein detected in channel 2; CA/GFP, an overlay of the images obtained from both channels. (C) Scheme of the fusion construct harbored by the pTAR-NES binary vector that was used to transform Arabidopsis plants.

Figure 4.

Simultaneous Emission of Linalool and Nerolidol by FaNES1-Expressing Arabidopsis Plants. Headspace Tenax trapping and subsequent GC-MS analysis show emission of linalool and nerolidol from vegetative parts of transgenic Arabidopsis and their absence in the headspace profile of the wild-type plant.

Figure 5.

Enantioselective Multidimensional GC-MS Analysis of Linalool Isolated from Leaves of FaNES1-Expressing Arabidopsis Plants. The top and middle chromatograms show the S- and R-linalool standards, respectively, and the bottom chromatogram shows linalool isolated from transgenic Arabidopsis.

Figure 6.

Linalool, Linalool Derivatives, and Nerolidol Produced by Leaves of Wild-Type and FaNES1-Expressing Arabidopsis Plants. Free (A) and glycosidically bound (B) terpenoids of Arabidopsis rosette leaves were isolated by solid-phase extraction, subjected to GC-MS analysis, and quantified. The plants are progeny (T2) of the selfed Tr-9 primary transformant (T1). FW, fresh weight; Nt, not transformed; Tr, transgenic.

Figure 7.

Linalool and Its Derivatives Produced by FaNES1-Expressing Arabidopsis Plants. The assumed glycosylation sites in linalool and its derivatives produced by leaves of transgenic Arabidopsis and potato plants are indicated. Potato plants were transformed with the same construct used to express FaNES1 in Arabidopsis (A. Aharoni and M.A. Jongsma, unpublished data).

Figure 8.

FaNES1-Expressing Arabidopsis Plants Show Growth Retardation. T3 progeny of selfed T2 transgenic FaNES1 lines are shown at top (each column of the growth tray is labeled with the T2 line number). The T2 lines originated from either the Tr-9 or Tr-26 primary transformant (T1). In the table, 30-day-old plants were scored as either small (Sm; inflorescence stems of <5 cm), intermediate (Int; inflorescence stems of 5 to 16 cm), or normal (No; inflorescence stems of >16 cm). The percentage of lines showing one of the three phenotypes (Sm, Int, or No) among the total number of T3 plants examined per T2 line (total) is presented. At bottom left, a closer observation of two Tr-9 selfed progeny lines (left and middle) and a wild type-plant (right) of the same age is shown. The graph at bottom right shows the correlation between the size of FaNES1-expressing Arabidopsis plants and the production of free and glycosylated linalool, linalool derivatives, and nerolidol. Free and glycosidically bound terpenoids of Arabidopsis rosette leaves were isolated by solid-phase extraction, subjected to GC-MS analysis, and quantified. The plants are progeny (T2) of the selfed Tr-9 and Tr-26 primary transformants (T1). FW, fresh weight; Nt, not transformed; Tr, transgenic.

Figure 9.

The Behavior of M. persicae Aphids Is Influenced by Volatiles Produced in Transgenic Plants Expressing the FaNES1 Gene. The graph shows the preference of M. persicae aphids (represented by percentage) for detached leaves of either wild-type plants (open bars) or transgenic FaNES1-expressing plants (closed bars) in a dual-choice assay. The calculated P values of the one-sided Wilcoxon signed-rank test (Hollander and Wolfe, 1973) showed significant evidence for repellence by the transgenic lines at the last three time points (*, P < 0.05; **, P < 0.01).

Figure 10.

Headspace Measurements Using SPME of Detached Rosette Leaves of Transgenic Arabidopsis Plants Overexpressing the Germacrene A Synthase Gene from Chicory. Trace amounts of β-elemene (the thermal rearrangement product of germacrene A; de Kraker et al., 1998) was detected in the headspace of the transgenic lines (Ger-28 and Ger-35) but not in wild-type control Arabidopsis. The chromatogram shows m/z 93, a typical m/z value for many monoterpenes and sesquiterpenes, but β-elemene was identified using the full mass spectrum (de Kraker et al., 1998).