 |  |
| ||||
Copyright © 2003, American Society of Plant Biologists (E)-β-Ocimene and Myrcene Synthase Genes of Floral Scent Biosynthesis in Snapdragon Function and Expression of Three Terpene Synthase Genes of a New Terpene Synthase Subfamily 1To whom correspondence should be addressed. E-mail dudareva/at/hort.purdue.edu; fax 1-765-494-0391 2These authors contributed equally to this work. 3Current address: Biotechnology Laboratory, University of British Columbia, Vancouver, BC, Canada V6T 1Z3. Received February 3, 2003; Accepted March 9, 2003.  This article has been cited by other articles in PMC. | |||||||||||||||||||
Abstract Snapdragon flowers emit two monoterpene olefins, myrcene and (E)-β-ocimene, derived from geranyl diphosphate, in ad-dition to a major phenylpropanoid floral scent component, methylbenzoate. Emission of these monoterpenes is regulated developmentally and follows diurnal rhythms controlled by a circadian clock. Using a functional genomics approach, we have isolated and characterized three closely related cDNAs from a snapdragon petal-specific library that encode two myrcene synthases (ama1e20 and ama0c15) and an (E)-β-ocimene synthase (ama0a23). Although the two myrcene synthases are almost identical (98%), except for the N-terminal 13 amino acids, and are catalytically active, yielding a single monoterpene product, myrcene, only ama0c15 is expressed at a high level in flowers and contributes to floral myrcene emission. (E)-β-Ocimene synthase is highly similar to snapdragon myrcene synthases (92% amino acid identity) and produces predominantly (E)-β-ocimene (97% of total monoterpene olefin product) with small amounts of (Z)-β-ocimene and myrcene. These newly isolated snapdragon monoterpene synthases, together with Arabidopsis AtTPS14 (At1g61680), define a new subfamily of the terpene synthase (TPS) family designated the Tps-g group. Members of this new Tps-g group lack the RRx8W motif, which is a characteristic feature of the Tps-d and Tps-b monoterpene synthases, suggesting that the reaction mechanism of Tps-g monoterpene synthase product formation does not proceed via an RR-dependent isomerization of geranyl diphosphate to 3S-linalyl diphosphate, as shown previously for limonene cyclase. Analyses of tissue-specific, developmental, and rhythmic expression of these monoterpene synthase genes in snapdragon flowers revealed coordinated regulation of phenylpropanoid and isoprenoid scent production. | |||||||||||||||||||
INTRODUCTION Monoterpenes, the C10 members of the terpenoid family of natural products, are common constituents of floral scents that function as volatile cues to attract insects or other pollinators (Knudsen et al., 1993; Dobson, 1994; Dudareva and Pichersky, 2000). These floral odors may have evolved from protective functions of monoterpenes in vegetative and reproductive organs of angiosperms and gymnosperms (Bohlmann et al., 2000a). At present, information about monoterpene biosynthesis and their regulation comes mostly from studies with vegetative tissues, in which they serve as compounds in direct and indirect defense against herbivores and as protection against pathogens. A large group of terpene synthase (TPS) genes and enzymes has been characterized in vegetative tissues of several angiosperms and gymnosperms (Bohlmann et al., 1998; Davis and Croteau, 2000; Aubourg et al., 2002). However, to date, only two terpene synthase genes of floral scent biosynthesis have been reported. Linalool synthase, a monoterpene synthase that catalyzes the formation of an acyclic monoterpene alcohol, linalool, was the first to be isolated and characterized in Clarkia breweri flowers (Pichersky et al., 1995; Dudareva et al., 1996), and a sesquiterpene synthase gene for the formation of germacrene D was characterized recently in Rosa hybrida petals (Guterman et al., 2002). Several features of snapdragon flowers involved in the attraction of bee pollinators, such as the composite floral scent of phenylpropanoids and isoprenoids, the concentration of scent production and emission in the lobes of a zygomorphic flower, and the rhythmic nature of emission, make it a very attractive system in which to decipher the biochemical and molecular genetic regulation of floral scent biology (Dudareva et al., 2000; Kolosova et al., 2001a, 2001b). The major components in the fragrance of snapdragon flowers (cv Maryland True Pink) are the aromatic ester methylbenzoate, which is derived from the phenylpropanoid pathway (~60% of total floral scent output), and two monoterpene olefins, myrcene and (E)-β-ocimene (~8 and 20% of total floral scent output, respectively), both of which are derived from geranyl diphosphate (Dudareva et al., 2000). Investigation of the regulation of methylbenzoate emission showed that the major factors controlling its production and emission during flower development include the regulation of expression of the benzoic acid carboxyl methyltransferase (BAMT) gene, which encodes the final enzyme in methylbenzoate biosynthesis, and the level of supplied substrate (benzoic acid) for the reaction (Dudareva et al., 2000). Moreover, it was shown that the level of substrate also plays a major role in the regulation of the circadian emission of methylbenzoate in diurnally (snapdragon) and nocturnally (petunia cv Mitchell and Nicotiana suaveolens) emitting flowers (Kolosova et al., 2001a). In contrast to recent progress in our understanding of the regulation of methylbenzoate formation and emission in snapdragon flowers (Dudareva et al., 2000; Kolosova et al., 2001a, 2001b), very little is known about the enzymes and genes responsible for the biosynthesis of floral monoterpenes and the molecular mechanisms that control the pathway flux to these volatile compounds. Here, we present a detailed analysis of the production of the volatile monoterpenes myrcene and (E)-β-ocimene in snapdragon flowers. Developmental and rhythmic emission of these monoterpene compounds was determined, as was emission in the absence of environmental cues (in continuous dark and continuous light), to evaluate if a circadian, endogenous clock controls their emission rhythmicity. Because many monoterpene synthases catalyze the formation of multiple products (Bohlmann et al., 1998; Wise and Croteau, 1999; Davis and Croteau, 2000), we also analyzed whether myrcene and ocimene biosynthesis requires specialized monoterpene synthases for each component or if a multifunctional synthase similar to myrcene/(E)-β-ocimene synthase from Arabidopsis (Bohlmann et al., 2000b) is involved. Three monoterpene synthases have been isolated from snapdragon flowers using a functional genomics approach, and biochemical characterization of the enzymes encoded by these three cDNAs showed that they represent two myrcene synthases and an (E)-β-ocimene synthase, enzymes responsible for the biosynthesis of two major monoterpene compounds of snapdragon floral scent. Analyses of the tissue-specific, developmental, and rhythmic expression of monoterpene synthase genes revealed coordinated regulation of phenylpropanoid and isoprenoid scent production in snapdragon flowers. | |||||||||||||||||||
RESULTS Developmental and Rhythmic Emission of Myrcene and (E)-β-Ocimene from Snapdragon Flowers We showed previously that snapdragon floral scent is dominated by (E)-β-ocimene, myrcene, and methylbenzoate (Dudareva et al., 2000). To determine variations in the emission of myrcene and (E)-β-ocimene during flower development, we performed time-course headspace collections from single, living flowers at 24-h intervals followed by gas chromatography–mass spectrometry (GC-MS) analysis. Headspace collections began with mature flower buds at 1 day before opening and ended 12 days later. Unopened flowers (buds) and 1-day-old flowers released little, if any, detectable myrcene and (E)-β-ocimene (Figure 1A). Monoterpene emission increased strongly on the second day after anthesis [50 and 39% of the maximum peak levels for myrcene and (E)-β-ocimene, respectively]. Emission of myrcene peaked on days 5 to 7 and remained relatively high for the next 2 to 3 days, whereas emission of (E)-β-ocimene peaked sharply on days 6 to 7 and declined gradually thereafter. Although the emission of (E)-β-ocimene by snapdragon flowers was four times higher than that of myrcene, with maximum levels at 31.9 and 8.3 μg per flower per 24 h on day 6 after anthesis, respectively, there was no difference in the emission of these monoterpenes in old (11- and 12-day-old) flowers (Figure 1A).
To determine diurnal fluctuations in the emission of myrcene and (E)-β-ocimene in snapdragon flowers, floral volatile compounds were collected from 3-day-old flowers at seven time points over a period of 48 h of two normal light/dark cycles (12 h of light/12 h of dark). This experiment revealed that, as with methylbenzoate (Kolosova et al., 2001a), emission of the two monoterpenes follows diurnal cycles, with the highest emission rate between 11 am and 6 pm (Figure 1B). To determine whether this diurnal rhythmicity of myrcene and (E)-β-ocimene emission is controlled by irradiation level or by a circadian clock, plants were grown in a 12-h-light/12-h-dark period at a constant temperature of 25°C and then transferred to conditions of either continuous light or continuous dark. Floral volatiles were collected at seven time points during each 24-h interval for 120 h (5 days) (Figures 2A and 2B). Flowers used for headspace analysis were 3 days old at the beginning of the time-course experiment. Results obtained show that the rhythmic emission of monoterpene compounds displayed a “free-running” cycle in the absence of environmental cues (in continuous dark or continuous light), indicating the circadian nature of diurnal rhythmicity (Figure 2).
Cloning and Sequence Characterization of Snapdragon Monoterpene Synthases Formation of the two major monoterpene floral scent components, myrcene and (E)-β-ocimene, could involve specialized monoterpene synthases for each of the two compounds or a multifunctional synthase. As part of an ongoing effort to isolate genes involved in floral scent production in snapdragon flowers, we randomly sequenced 792 clones (ESTs) from a cDNA library constructed from mRNA isolated from petals (upper and lower lobes) of 1- to 5-day-old flowers, the parts of the flower highly specialized for floral scent biosynthesis (Dudareva et al., 2000). A search of this EST database for potential monoterpene synthases revealed three unique cDNA clones (designated ama0c15, ama0a23, and ama1e20) with overall high sequence similarity (Figure 3) to angiosperm monoterpene synthases of the Tps-b cluster of the TPS gene family of higher plants (Bohlmann et al., 1998; Aubourg et al., 2002) as well as with linalool synthase from Clarkia (Dudareva et al., 1996) and gymnosperm monoterpene synthases of the Tps-d cluster (Bohlmann et al., 1999).
The ama0c15 EST represented a full-length cDNA clone of 1901 nucleotides. This clone had a very short 5′ untranslated region (only 6 nucleotides) and encoded an open reading frame of 1743 nucleotides corresponding to a predicted protein of 581 amino acids with a calculated molecular weight of 66,701 and a pI of 6.04 (Figure 4). The other EST cDNA clones, ama0a23 and ama1e20, were found to be truncated at their 5′ ends. Two methods, 5′ rapid amplification of cDNA ends (RACE) and cDNA library screening, were used to recover corresponding full-length clones for these ESTs. Initial 5′ RACE based on truncated ama1e20 resulted in the isolation of an independent cDNA clone identical to ama0c15. However, screening of the cDNA library using the truncated ama1e20 EST probe yielded a full-length cDNA clone corresponding to ama1e20 of 2412 nucleotides, with an open reading frame of 1752 nucleotides that encoded a predicted protein of 584 amino acids (Figure 4) with a calculated molecular weight of 67,349 and a pI of 6.11. The proteins encoded by ama0c15 and ama1e20 were 98% identical to each other, and the differences were found only within the N-terminal 13 amino acids. In addition, the ama0c15 protein was three amino acids shorter than ama1e20. The third cDNA clone, ama0a23, was nearly full length, with 1844 bp lacking the ATG start codon and the first codon of its open reading frame lining up with the position of codon 10 in ama1e20. Using 5′ RACE, a 1902-bp full-length cDNA sequence of ama0a23 was recovered that encoded an open reading frame of 1737 nucleotides corresponding to a protein of 579 amino acids (Figure 4) with a calculated molecular weight of 66,724 and a pI of 6.15. This cDNA clone has 92% amino acid identity to ama1e20 and ama0c15.
Despite overall sequence similarity to other monoterpene synthases, the deduced proteins of the snapdragon monoterpene synthase cDNAs ama0c15, ama0a23, and ama1e20 are quite different from the known monoterpene synthases of angiosperms (Tps-b group; 22 to 27% amino acid identity) and gymnosperms (Tps-d group; 22 to 27% amino acid identity). Strikingly, the predicted snapdragon proteins do not contain the RRx8W motif (Figure 4), which is a key feature of all monoterpene synthases of the Tps-b and Tps-d groups characterized to date (Bohlmann et al., 1998; Aubourg et al., 2002). Functional Identification and Characterization of ama0c15, ama0a23, and ama1e20 For functional characterization of the TPS proteins encoded by ama0c15, ama0a23, and ama1e20, the cDNAs were subcloned into the pET-TOPO D100 expression vector, and the encoded proteins then were expressed separately in Escherichia coli BL21-CodonPlus(DE3). Extracts of isopropylthio-β-galactoside–induced bacteria were assayed for monoterpene synthase, sesquiterpene synthase, and diterpene synthase activities using the corresponding prenyl diphosphate substrates geranyl diphosphate (GPP), farnesyl diphosphate, and geranylgeranyl diphosphate. Enzymatic formation of terpene olefins was observed only with GPP as a substrate, not with farnesyl diphosphate or geranylgeranyl diphosphate, identifying all three genes as monoterpene synthases. No oxygenated monoterpene products were detected. Extracts prepared from E. coli (same strain) transformed with pET-TOPO D100 lacking a cDNA insert and heat-denatured enzyme preparations served as controls for monoterpene formation independent of snapdragon TPS. These extracts did not yield any detectable amounts of monoterpene product.Products formed by recombinant snapdragon monoterpene synthases with GPP as a substrate were analyzed by GC-MS. The two very similar proteins encoded by ama1e20 and ama0c15 each yielded a single monoterpene product of the same retention time. This product was identified as myrcene based on a comparison of retention times and mass spectra with those of an authentic myrcene standard (Figure 5). Protein encoded by ama0a23 also formed a single major product with GPP that was identified based on retention time and mass spectral data as (E)-β-ocimene (97% of total monoterpene olefin product), along with two minor products identified as (Z)-β-ocimene (2%) and myrcene (1%) (Figure 6). Together, these three monoterpene synthase genes, ama1e20, ama0c15, and ama0a23, account for the major monoterpene compounds of floral scent in snapdragon.
To test the presence of corresponding monoterpene synthase activities in snapdragon floral tissue, cell-free extracts were prepared from upper and lower petal lobes of 5-day-old flowers and assayed with GPP as a substrate. Cell-free extracts produced two monoterpenes, (E)-β-ocimene and myrcene, constituting 80 and 20% of enzyme products, respectively, along with minor amounts of (Z)-β-ocimene (Figure 7).
Tissue-Specific, Developmental, and Rhythmic Gene Expression of Monoterpene Synthases To determine if these three monoterpene synthase genes are expressed with spatial and temporal patterns correlated with floral scent emission, we analyzed monoterpene synthase gene expression in snapdragon flower tissues. Total RNA was isolated from leaves and different floral parts of 3-day-old snapdragon flowers (sepals, pistils, stamens, and different regions of the corolla: the upper petal lobes, the lower petal lobes, and the tube) and used in RNA gel blot analysis. Because of the high sequence similarity between ama0a23, ama0c15, and ama1e20 clones, a single monoterpene synthase probe derived from ama1e20 was used initially for all RNA gel blot analyses. The highest level of monoterpene synthase gene expression was found in the upper and lower lobes of petals (Figure 8A), the parts of the flower that were shown previously to be primarily responsible for scent production and emission in snapdragon (Dudareva et al., 2000). A very low level of transcripts was detected in the tube and stamens. No detectable signals were found in pistils, sepals, and leaf tissues.
Because the signals detected in this RNA gel blot experiment could represent a sum of transcripts for all three isolated monoterpene synthases, the contribution of each gene to the total expression level was analyzed further by reverse transcriptase (RT)–PCR using gene-specific primers. When total RNA isolated from upper and lower petal lobes was used in RT-PCR, different levels of expression were found for ama0a23, ama0c15, and ama1e20 (Figure 8B). Although ama0a23 [(E)-β-ocimene synthase] and ama0c15 (myrcene synthase) genes were expressed at similarly high levels, expression of ama1e20 (myrcene synthase) was very low and detectable only after overexposure of the autoradiographs presented in Figure 8B. According to quantitative RT-PCR analysis, 60% of the TPS hybridization signal obtained by RNA gel blot analysis (Figures 8A and 9) was attributed to the expression of ama0c15 and 40% of the signal was attributed to ama0a23. This ratio of gene expression was stable during flower development and during a light/dark cycle. ama1e20 did not contribute significantly to monoterpene synthase expression.
To examine the steady state levels of monoterpene synthase mRNA in upper and lower petal lobes during flower development, total RNA was extracted from the petal lobes of mature flower buds at 1 day before anthesis and from open flowers at 1 to 12 days after anthesis (Figure 9A). A plot showing the changes in the relative amounts of transcripts for monoterpene synthases during the life span of the flower, normalized to equal amounts of 18S rRNA, is presented in Figure 9B. Monoterpene synthase mRNA was detected first in mature flower buds, and its level increased until it peaked on day 4 after anthesis (Figures 9A and 9B). Over the next 3 days (days 5 to 7), mRNA levels declined sharply by 40% and decreased slowly thereafter (Figures 9A and 9B). Next, we determined whether the oscillations in myrcene and (E)-β-ocimene emissions during a light/dark cycle (Figure 1B) were the function of rhythmic regulation of the corresponding gene expression. Accumulation of mRNA was examined in upper and lower petal lobes at nine time points during a 27-h interval (Figure 9C). The mRNA levels were quantified and normalized for any loading differences, and the relative values are shown in Figure 9D. We found only slight variations in monoterpene synthase mRNA levels during the daily light/dark cycle. Levels of transcripts for monoterpene synthases increased during the light period, and their maximum accumulation was detected at ~3 pm; they then declined at least twofold during the dark period (12 am), showing a weak diurnal oscillation pattern. The rhythmic pattern of mRNA accumulation was retained under continuous dark (Figure 9E), suggesting that the cyclic expression of monoterpene synthases is under circadian control. Although changes in transcript levels may not directly determine protein levels or enzyme activities due to possible post-transcriptional, post-translational, or enzyme-regulatory mechanisms, the positive correlation between transcript levels and volatile emission suggests that changes in transcript level are an important determinant of scent production. | |||||||||||||||||||
DISCUSSION Monoterpene Synthases Involved in Snapdragon Floral Scent Biosynthesis Monoterpene compounds, such as (E)-β-ocimene and myrcene, contribute significantly to the floral odors of numerous plant species (Knudsen et al., 1993). (E)-β-Ocimene constitutes 87% of the scent of the orchid Laelia anceps (Kaiser, 1993), whereas the scent of Brugmansia × candida, a member of the Solanaceae, contains as much as 52% (E)-β-ocimene (Kite and Leon, 1995). In snapdragon, the monoterpene fraction of the floral scent bouquet is dominated by (E)-β-ocimene and myrcene, which account for 20 and 8% of total floral volatiles, respectively (Dudareva et al., 2000). Developmental and diurnal rhythmic emissions of these two acyclic monoterpenes follow similar patterns, with (E)-β-ocimene approximately twofold to fourfold more abundant than myrcene (Figure 1). Therefore, it is possible that the coordination of the formation and emission of myrcene and (E)-β-ocimene is the result of the action of a single multiproduct monoterpene synthase. A myrcene/ocimene synthase (AtTPS10 [At2g24210]) has been characterized in Arabidopsis that converts geranyl diphosphate into myrcene (56% of total hydrocarbon product), (E)-β-ocimene (20%), and small amounts of cyclic monoterpenes (each <5%) (Bohlmann et al., 2000b). On the other hand, specialized monoterpene synthases with a single major product also are known. For example, Arabidopsis AtTPS03 (At4g16740) produces mainly (E)-β-ocimene (Fäldt et al., 2003a), whereas the only product of myrcene synthase from grand fir is myrcene (Bohlmann et al., 1997).As part of an ongoing EST-based effort to discover genes of floral scent in snapdragon, we isolated three new monoterpene synthase genes and identified them using a biochemical, functional approach as myrcene (ama1e20 and ama0c15) and (E)-β-ocimene (ama0a23) synthases. The products of these synthases are the two dominant isoprenoids of snapdragon floral fragrance. These three monoterpene synthases are highly similar to each other and encode proteins of 66 to 67 kD with a pI value near 6.0, which is similar to that of other known monoterpene synthases (Bohlmann et al., 1998; Wise and Croteau, 1999). ama1e20 and ama0c15 deduced proteins are almost identical except for the N-terminal 13 amino acids (Figure 4). Two independent methods (5′ RACE and cDNA library screening) verified that these clones are not cDNA cloning artifacts of the same gene but represent different genes or allelic variations of the same gene. Despite the differences within the N-terminal amino acids, ama1e20 and ama0c15 both encode functional myrcene synthases. This finding supports earlier data that N-terminal regions of up to 60 to 70 amino acids of monoterpene synthases are not critical for enzyme catalytic function (Williams et al., 1998; Bohlmann et al., 1999). ama0a23 [(E)-β-ocimene synthase] also is highly similar (92% identical) to ama1e20 and ama0c15, yet it is specialized for a different biochemical function. This small family of three monoterpene synthases from snapdragon demonstrates that the function of TPS ESTs cannot be predicted based on sequence. The snapdragon TPS genes are most likely derived from a common ancestor via gene duplication and functional radiation, similar to the TPS genes in Arabidopsis and conifers (Bohlmann et al., 1999; Aubourg et al., 2002). The precise structural features that determine the product specificity of monoterpene synthases remain to be deciphered, and such knowledge will contribute to our understanding of the adaptive evolution of floral scent phenotypes. Interestingly, one of the myrcene synthase genes, ama1e20, is not expressed at significant levels in snapdragon petal tissues, indicating that only ama0c15 (myrcene synthase) and ama0a23 [(E)-β-ocimene synthase] contribute to terpene synthase transcript accumulation and, correspondingly, to floral myrcene and (E)-β-ocimene emission (Figures 8 and 9). The myrcene synthase ama0c15 contributes 1.5 times more (60%) than the (E)-β-ocimene synthase ama0a23 (40%) to the total monoterpene transcript levels in upper and lower petal lobes of snapdragon flowers (Figure 8B). However, snapdragon flowers emit more (E)-β-ocimene relative to myrcene (20 and 8% of total floral volatiles, respectively), and monoterpene synthase enzyme activities form a mixture of monoterpenes that is high in (E)-β-ocimene (~80% of native enzyme product) and low in myrcene (~20%), along with minor amounts of (Z)-β-ocimene (Figure 7). Together, these data imply that the snapdragon monoterpene floral scent profile may be regulated at translational and/or post-translational levels of the corresponding monoterpene synthases. Different rates of protein synthesis and proteolytic turnover and/or differences in protein modifications or the kinetic properties of the two monoterpene synthases could influence the ratio of (E)-β-ocimene and myrcene emitted from floral tissues. Secondary modification of monoterpene olefins (e.g., oxidation/glycosylation) or sequestration also could contribute to the monoterpene emission profile. Indeed, transgenic expression of linalool synthase in petunia flowers resulted in the accumulation of monoterpenol conjugates, leading to a weak emission of linalool (Lücker et al., 2001). Snapdragon Monoterpene Synthases Lack the Conserved RRx8W Motif An interesting feature of all three snapdragon monoterpene synthases is the lack of the RRx8W motif (Figure 4), which is characteristic of other monoterpene synthases of the angiosperm Tps-b group and the gymnosperm Tps-d group (Bohlmann et al., 1998, 1999; Aubourg et al., 2002). In the formation of the cyclic monoterpene (−)-4S-limonene catalyzed by Mentha spicata limonene synthase, the double-Arg element of the RRx8W motif is involved in the diphosphate migration step to yield 3S-linalyl diphosphate (LPP) (Williams et al., 1998). LPP is the proposed essential, enzyme-bound intermediate for monoterpene cyclization (Wise and Croteau, 1999). However, the formation of acyclic myrcene and ocimene may not require the isomerization of GPP to an LPP intermediate but could be formed directly from GPP after cleavage of the diphosphate group without covalent readdition of the diphosphate moiety (Bohlmann et al., 1999). The absence of any cyclic monoterpenes in the product fraction of snapdragon myrcene and (E)-β-ocimene synthases could be attributable to the lack of an intact RRx8W motif. The RRx8W motif also is missing in two other monoterpene synthases of known function, Clarkia LIS (Dudareva et al., 1996) and Arabidopsis AtTPS14 (At1g61680) (Aubourg et al., 2002; Chen et al., 2003). Clarkia LIS and Arabidopsis AtTPS14 both encode synthases specialized for the formation of the acyclic monoterpene alcohol linalool from GPP (Dudareva et al., 1996; Chen et al., 2003).Snapdragon Monoterpene Synthases Are Members of a New Tps-g Subfamily Recent phylogenetic analysis of the deduced amino acid sequences of isolated plant terpenoid synthases revealed six TPS gene subfamilies, designated Tps-a through Tps-f (Bohlmann et al., 1998). The three newly isolated snapdragon monoterpene synthases do not cluster with functional homologs of the Tps-b family of angiosperm monoterpene synthases and the Tps-d family of conifer monoterpene synthases (Bohlmann et al., 1998; Aubourg et al., 2002), despite sharing overall sequence similarity (Figure 10). Moreover, among all functionally identified plant TPS, the snapdragon ocimene and myrcene synthases closely resemble Arabidopsis AtTPS14 (At1g61680), a TPS that previously represented an isolated branch of the plant TPS family (Aubourg et al., 2002). Together, the snapdragon monoterpene synthases ama0c15, ama0a23, and ama1e20 and Arabidopsis AtTPS14 (At1g61680) define a new subfamily of the TPS family designated the Tps-g group (Figure 10).
In addition to overall high sequence identity within this new subfamily and low sequence relatedness with the TPS of other subfamilies, monoterpene synthases of this new Tps-g group lack the RRx8W motif, which is a characteristic feature of the Tps-d and Tps-b monoterpene synthases. All four functionally characterized members of the Tps-g group produce acyclic monoterpenes, and surprisingly, all of them, including Arabidopsis AtTPS14 (Chen et al., 2003), contribute to floral volatiles. Members of the Tps-f subfamily (Clarkia linalool synthase and Arabidopsis AtTPS04 [At1g61120; unknown function]) and members of the Tps-e and Tps-c groups (kaurene synthases and copalyl diphosphate synthase, respectively) contain an additional 200–amino acid motif of unknown function that is missing in the newly identified Tps-g members (Bohlmann et al., 1998; Trapp and Croteau, 2001; Aubourg et al., 2002). Moreover, this new Tps-g subfamily is clearly separated, by sequence and function, from the Tps-a group, which contains angiosperm sesquiterpene and diterpene synthases (Figure 10). Coordinated Circadian Rhythmicity, Tissue Specificity, and Developmental Control of Isoprenoids and Phenylpropanoids in Snapdragon Flowers Previously, we showed that the main sites of the biosynthesis of methylbenzoate, a major phenylpropanoid floral scent component of snapdragon, are the upper and lower petal lobes, which make almost equal contributions to the whole flower fragrance (Dudareva et al., 2000). Here, we have shown that the production and emission of the other two main fragrance compounds, the monoterpenes myrcene and (E)-β-ocimene, also occur in the upper and lower petal lobes, where high levels of monoterpene synthase gene expression have been found (Figure 8A). Emissions of these monoterpene compounds follow very similar diurnal rhythms controlled by a circadian clock (Figures 1B and 2) that match the rhythmic emission of methylbenzoate (Kolosova et al., 2001a), indicating that two apparently separate metabolic pathways are regulated coordinately by a circadian clock to yield diurnal emission of a complex mixture of floral fragrance derived from the isoprenoid and phenylpropanoid pathways.Similar to BAMT mRNA, monoterpene synthase mRNA levels oscillate, with a maximum at ~3 pm before declining at least twofold during the dark period (12 am) (Figures 9C and 9D), showing strong correlation with corresponding monoterpene emission (Figure 1B). Rhythmic fluctuations in mRNA levels continue in constant dark (Figure 9E), suggesting that, like BAMT mRNA, a free-running internal circadian clock regulates the monoterpene synthase mRNA abundance. Moreover, like BAMT mRNA, the steady state mRNA levels of monoterpene synthases in petals are regulated developmentally, peaking on day 4 after anthesis (Figures 9A and 9B), 1 to 2 days ahead of maximum emission (Figure 1A). These results indicate that the emission of phenylpropanoid and isoprenoid volatiles in snapdragon is under similar or identical developmental and diurnal control, as was shown in this study for monoterpenes and previously for methylbenzoate (Dudareva et al., 2000; Kolosova et al., 2001a). Overall, these findings suggest that emission of a composite floral scent bouquet in snapdragon is regulated upstream of individual metabolic pathways, thus coordinating the emission of phenylpropanoids and isoprenoids. This regulation includes the coordinate expression of genes that encode enzymes involved in the final steps of scent biosynthesis (Figure 9) (Dudareva et al., 2000; Kolosova et al., 2001a). However, the level of scent biosynthetic enzymes is not the only limiting factor, and the target for the regulation of scent production might include the level of supplied substrate in the cell. Although the involvement of benzoic acid in the regulation of rhythmic and developmental methylbenzoate emission has been shown (Dudareva et al., 2000; Kolosova et al., 2001a), the role of GPP in the regulation of monoterpene emission remains to be determined. The biochemical characterization of myrcene and (E)-β-ocimene synthases also will provide further information regarding the competition of these two enzymes for the same substrate. The possibility that the level of emitted monoterpenes is regulated at the translational/post-translational level or by the modification of monoterpenes cannot be excluded. | |||||||||||||||||||
METHODS Plant Material and Headspace Analysis of Floral Volatiles Snapdragon (Antirrhinum majus) cv Maryland True Pink (Ball Seed, West Chicago, IL) was grown under standard greenhouse conditions as described previously (Dudareva et al., 2000). For the headspace collections of floral volatiles, plants were placed in growth chambers (model E8; Conviron, Asheville, NC) with a 25/20°C (light/dark) temperature cycle, a RH of 40%, and a 12-h photoperiod with light intensity of 350 μmol·m−2·s−1 for a normal day/night cycle and 100 μmol·m−2·s−1 for the experiments in continuous light. Emitted volatiles were determined by headspace analysis as described previously (Raguso and Pichersky, 1995; Raguso and Pellmyr, 1998) at 24-h intervals for developmental emission and at seven time points during a 24-h period for rhythmic emission. Trapped floral scent compounds were analyzed by gas chromatography–mass spectrometry (GC-MS) with a FinniganMAT GCQ instrument (Thermoquest, San Jose, CA) and by GC with an Agilent 6890 series gas chromatograph (Agilent Technologies, Palo Alto, CA) as described previously (Dudareva et al., 2000; Kolosova et al., 2001a). Quantification was based on flame ionization detector peak areas and the internal standard. Ten microliters of a 0.03% (v/v) solution of toluene in hexane was added to each sample as an internal standard. Specific correction factors were developed from the myrcene and (E)-β-ocimene (Aldrich, Milwaukee, WI) standard curves and applied to peak areas of monoterpenes in samples.Substrates, Standards, and Reagents 1-3H-Geranyl diphosphate (GPP) (3.7 kBq/mol), 1-3H-farnesyl diphosphate (3.7 kBq/mol), and 1-3H-geranylgeranyl diphosphate (1.2 kBq/mol) were from American Radiolabeled Chemicals (St. Louis, MO). Unlabeled GPP (1 mg/mL) was from Echelon Research Laboratories (Salt Lake City, UT). Terpenoid standards, other chemicals, and reagents were purchased from Sigma Chemical Co., Fluka Chemical Co., or Aldrich Chemical Co. unless noted otherwise.EST Sequencing, Rapid Amplification of cDNA Ends, and cDNA Library Screening A λ-ZAPII cDNA library was constructed from mRNA isolated from upper and lower petal lobes of 1- to 5-day-old snapdragon flowers according to the manufacturer's protocol (Stratagene, La Jolla, CA) (Dudareva et al., 2000). The titer of the unamplified library was 1.1 × 106 plaque-forming units. This primary library was amplified, and the amplified library had a titer of 1.5 × 1010 plaque-forming units. Mass excision of pBluescript SK− phagemids (Stratagene) from the amplified library resulted in a stock that was used for plating and random colony picking for DNA sequencing. The inserts were partially sequenced from one end using T3 primer. To isolate full-length cDNA clones, a cDNA library was screened with ama1e20 probe under high-stringency conditions, as described by Dudareva et al. (1996). Rapid amplification of cDNA ends was performed using the First Choice RLM-RACE system (Ambion, Austin, TX) according to the manufacturer's instructions.Sequence Analysis Sequence analysis was performed using programs from the Lasergene DNA-Star Package version 5, CLUSTAL X (http://www-igbmc.u-strasbg.fr/BioInfo/ClustalX) and Genedoc (http://www.psc.edu/biomed/genedoc). Sequence relatedness was analyzed using CLUSTAL X and the neighbor-joining method, and the unrooted tree was visualized using Treedraw (http://taxonomy.zoology.gla.ac.uk/rod/treeview.html; Page, 1996).Functional Expression of Monoterpene Synthases in Escherichia coli and Enzyme Assays Subcloning and functional expression of ama1e20, ama0c15, and ama0a23 in E. coli followed a previously published procedure (Fäldt et al., 2003b). The cDNAs were subcloned into the pET100/D-TOPO expression vector (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. The ama1e20 cDNA was amplified by PCR using forward primer 5′-CACCATGATCTATATTTGGATCTGCTTTTATCTCCA-3′ in combination with reverse primer 5′-TGTTTAATCCGACAACATAAACTTGAC-3′; ama0c15 insert was amplified using primers 5′-CACCATGGCATTTTGCATTAGCTAC-3′ and 5′-TTAATCCGACAACATAAA-CTTGAC-3′; and ama0a23 insert was amplified using primers 5′-CACCCTCTTAGGTGCAGTGCTTCC-3′ and 5′-TTAATCCGACAACATAAA-CTTGAC-3′. PCR was performed in volumes of 100 μL containing 20 mM Tris-HCl, pH 8.8, 10 mM KCl, 10 mM (NH4)2SO4, 2 mM MgSO4, 0.1% Triton X-100, 10 μg of BSA, 200 μM of each deoxynucleotide triphosphate, 0.1 μM of each primer, 2.5 units of high-fidelity Turbo Pfu polymerase (Stratagene), and 2 ng of template DNA. After a denaturing step at 95°C for 2 min, 30 cycles of amplification were performed using the following temperature program in an MJ PTC100 thermocycler (MJ Research, Watertown, MA): 30 s at 95°C, 30 s at 59°C, 2 min at 72°C, followed by a 10-min final extension at 72°C. PCR products were cloned into the pET100/D-TOPO vector, and recombinant plasmids were transformed into E. coli TOP10 F′ cells according to the protocol for ligation and transformation (Invitrogen). Plasmids pET-ama1e20, pET-ama0c15, and pET-ama0a23 were identified by PCR and purified, inserts were sequenced, and the plasmids were transformed for expression in E. coli BL21-CodonPlus(DE3) (Stratagene).For functional expression, bacterial E. coli BL21-CodonPlus(DE3) strains transformed with pET-ama1e20, pET-ama0c15, or pET-ama0a23 were grown to A600 = 0.5 at 37°C in 5 mL of Luria-Bertani medium supplemented with 20 μg/mL ampicillin. Cultures then were induced by the addition of isopropyl-1-thio-β-d-galactopyranoside to a final concentration of 1 mM and grown for another 12 h at 20°C. Cells were harvested by centrifugation, resuspended in 1 mL of monoterpene synthase buffer (25 mM Hepes, pH 7.2, 10 mM MnCl2, 10% glycerol, and 5 mM DTT), and disrupted by sonication using a Branson Sonifier 250 (Branson Ultrasonic Corp., Danbury, CT) at constant power (5 W) for 10 s. Lysates were cleared by centrifugation, and the resulting supernatants containing the soluble enzyme were assayed for monoterpene, sesquiterpene, and diterpene synthase activity using the appropriate radiolabeled substrates or nonlabeled substrates as described previously (Bohlmann et al., 1997; Peters et al., 2000; Fäldt et al., 2003b). Enzyme Assays of Snapdragon Native Floral Monoterpene Synthases Flowers were ground in a chilled mortar with 10 volumes of extraction buffer (50 mM 3-[N-morpholino]-2-hydroxypropanesulfonic acid, pH 6.9, 5 mM DTT, 5 mM Na2S2O5, 1% [w/v] polyvinylpyrrolidone-40, and 10% glycerol). Extracts were shaken gently for 30 min on ice followed by centrifugation (15,000g for 30 min at 4°C). The supernatant was passed through a cell strainer to remove any large particles and frozen at −80°C. To assay monoterpene activity, the protein extract was thawed in water (22°C) and desalted by filter centrifugation (30K Macrosep; Pall Corp., Ann Arbor, MI). Proteins were resuspended in assay buffer (as described for recombinant enzyme assays) and incubated for 2 h with 250 μM GPP. Products of native enzyme assays were purified as described above for recombinant enzymes.Monoterpene Product Identification by GC-MS Monoterpene products formed by recombinant or native enzymes were identified by GC-MS analysis on an Agilent 6890 Series GC System coupled to an Agilent 5973 Network Mass Selective Detector (70 eV) using an HP 5 (Agilent Technologies) capillary column (both of which have the following dimensions: 0.25 mm i.d. × 30 m with 25-μm film). The injector was operated splitless at a temperature of 200°C with a column flow of 1 mL He/min. The following temperature program was used: initial temperature of 40°C (2-min hold), increase to 160°C at 3°C/min, 10°C/min ramp to 200°C, followed by a 20°C/min ramp to 300°C (3-min hold). Product identifications were verified using a DB-WAX capillary column (J&W Scientific, Palo Alto, CA) using the same program described above but with a final temperature of 260°C. Monoterpenes were identified by comparing mass spectra using Chemstation software (Agilent Technologies) and Wiley125 MS-library (Wiley Publishing, Hoboken, NJ) and by comparing retention times and mass spectra with those of authentic standards.RNA Isolation and RNA Gel Blot Analysis Total RNA was isolated and analyzed as described previously (Dudareva et al., 1996, 1998, 2000; Wang et al., 1997; Kolosova et al., 2001a) from floral tissues and petals at different stages of flower development, nine time points during a daily light/dark cycle, and eight time points during continuous dark conditions. A 1.9-kb EcoRI fragment containing the coding region of the monoterpene synthase genes derived from ama1e20 was used as a probe in RNA gel blot analysis. Five micrograms of total RNA was loaded in each lane. Hybridization signals were quantified using a Storm 860 phosphor screen (Molecular Dynamics, Sunnyvale, CA), and mRNA transcript levels were normalized to rRNA levels to overcome error in RNA quantitation by spectrophotometry. Multiple independent RNA gel blots were prepared and analyzed for each time point, and a representative time-course profile is presented.Reverse Transcriptase–PCR Quantitative reverse transcriptase (RT)–PCR was performed as described previously (Gang et al., 2002) with some modifications. One microgram of total RNA was used to make cDNA using random hexamer primers with the Advantage RT-for-PCR kit according to the manufacturer's instructions (Clontech, Palo Alto, CA). To analyze the contribution of ama1e20 and ama0c15 in the expression pattern, two forward gene-specific primers were designed for PCR. These primers amplified products of 800 bp in the 5′ regions of the coding sequences and were as follows: for ama1e20, 5′-TATTTGGATCTGCTTTTATCTCC-3′, and for ama0c15, 5′-TTGCATTAGCTACGTAGGTGC-3′. The contribution of ama0c15 and ama0a23 then was examined using two other forward gene-specific primers, 5′-CATAACACGAGTAAACTACATCG-3′ for ama0c15 and 5′-TATTTCATAACACGAGTAAACATG-3′ for ama0a23, which amplified products of 740 bp. For each PCR, the same reverse conserved primer, 5′-TACTTGCAAATCCATCAGGGC-3′, was used. Reactions were performed in duplicate for 30 cycles with an annealing temperature of 52°C.Analysis of products obtained by PCR with different amounts of cDNA (5, 10, 15, and 20 μL) showed that product increased linearly. Fifteen microliters of cDNA was chosen as an optimal level for further PCR. To ensure that an equal amount of RNA was used for all samples and that RT reactions were equally effective, PCR was performed with the 18S rRNA gene-specific primers 5′-GATAAAAGGTCGACACGGGCTCTGC-3′ (forward) and 5′-AACGGAATTAACCAGACAAATCGCTCC-3′ (reverse). For rDNA, PCR was performed for 15 cycles at an annealing temperature of 60°C. The amplified products were run on 1.2% agarose gels, blotted onto nitrocellulose membranes (Trans-Blot Transfer Medium; Bio-Rad, Hercules, CA), and hybridized with the corresponding DNA probe, a 1.9-kb EcoRI fragment of ama1e20 or a 1-kb BamHI-EcoRI fragment of rDNA. Quantitation was performed using a Storm 860 PhosphorImager and ImageQuant software (Molecular Dynamics). Upon request, all novel materials described in the article will be made available in a timely manner for noncommercial research purposes. Accession Numbers The GenBank accession numbers for the sequences mentioned in this article are as follows: ama0c15, AY195608; ama1e20, AY195609; ama0a23, AY195607; AtTPS03 [At4g16740; (E)-β-ocimene synthase], AY151086; and AtTPS10 (At2g24210; myrcene/ocimene synthase), AF178535.1. | |||||||||||||||||||
Acknowledgments We thank Gen-ichiro Arimura for assistance and Rodney Croteau for helpful comments. This work was supported by grants to N.D. from the National Science Foundation (Grant MCB-0212802), the Fred Gloeckner Foundation, and the American Floral Endowment and by grants to J.B. from the Natural Sciences and Engineering Research Council of Canada, the Canadian Foundation for Innovation, and British Columbia Knowledge Development Funds. J.F. received fellowships from the Bengt Lundqvist Minne Foundation (Sweden) and the Swedish Foundation for International Cooperation in Research and Higher Education. D.M. is a recipient of a Walter C. Koerner fellowship of the University of British Columbia. This article is contribution 16,971 from the Purdue University Agricultural Experimental Station. | |||||||||||||||||||
Notes Article, publication date, and citation information can be found at www.plantcell.org/cgi/doi/10.1105/tpc.011015. | |||||||||||||||||||
References
| |||||||||||||||||||
