Aromatic l-amino acid decarboxylases (AADCs) are key enzymes operating at the interface between primary and secondary metabolism. The Arabidopsis thaliana genome contains two genes, At2g20340 and At4g28680, encoding pyridoxal 5′-phosphate-dependent AADCs with high homology to the recently identified Petunia hybrida phenylacetaldehyde synthase involved in floral scent production. The At4g28680 gene product was recently biochemically characterized as an l-tyrosine decarboxylase (AtTYDC), whereas the function of the other gene product remains unknown. The biochemical and functional characterization of the At2g20340 gene product revealed that it is an aromatic aldehyde synthase (AtAAS), which catalyzes the conversion of phenylalanine and 3,4-dihydroxy-l-phenylalanine to phenylacetaldehyde and dopaldehyde, respectively. AtAAS knock-down and transgenic AtAAS RNA interference (RNAi) lines show significant reduction in phenylacetaldehyde levels and an increase in phenylalanine, indicating that AtAAS is responsible for phenylacetaldehyde formation in planta. In A. thaliana ecotype Columbia (Col-0), AtAAS expression was highest in leaves, and was induced by methyl jasmonate treatment and wounding. Pieris rapae larvae feeding on Col-0 leaves resulted in increased phenylacetaldehyde emission, suggesting that the emitted aldehyde has a defensive activity against attacking herbivores. In the ecotypes Sei-0 and Di-G, which emit phenylacetaldehyde as a predominant flower volatile, the highest expression of AtAAS was found in flowers and RNAi AtAAS silencing led to a reduction of phenylacetaldehyde formation in this organ. In contrast to ecotype Col-0, no phenylacetaldehyde accumulation was observed in Sei-0 upon wounding, suggesting that AtAAS and subsequently phenylacetaldehyde contribute to pollinator attraction in this ecotype.
Aromatic l-amino acid decarboxylases (AADCs) catalyze the pyridoxal-5′-phosphate (PLP)-dependent decarboxylation of aromatic l-amino acids to form their corresponding amines, and operate at the interface between primary and secondary metabolism (Facchini et al., 2000). AADCs are widespread in nature, and have been identified in plants, mammals and insects. These enzymes share extensive amino acid identity across different taxonomic lineages, and belong to group II of amino acid decarboxylases (Sandmeier et al., 1994). Whereas mammalian and insect AADCs accept a broad range of aromatic l-amino acids, plant enzymes exhibit strict substrate specificity toward l-amino acids, with either an indole or phenol side chain, but not both. To date, l-tyrosine decarboxylase (TYDC) and l-tryptophan decarboxylase (TDC) have been isolated and characterized from a variety of plant species (Gügler et al., 1988; De Luca et al., 1989; Kawalleck et al., 1993; Facchini and De Luca, 1994, 1995; López-Meyer and Nessler, 1997; Yamazaki et al., 2003; Kang et al., 2007). The products of these enzymes, tyramine and tryptamine, are precursors of numerous plant secondary metabolites, including alkaloids (Facchini et al., 2000) and serotonin (5-hydroxytryptamine) (Schröder et al., 1999; Kang et al., 2007, 2008), as well as hydroxycinnamic acid-conjugated amides, the deposition of which in the cell wall is believed to contribute to plant defense by creating a physical barrier against pathogens (Facchini et al., 2002).
Biochemical characterization of plant TYDCs revealed that they use l-tyrosine (Tyr) and 3,4-dihydroxy-l-phenylalanine (l-Dopa) as substrates, but are inactive with l-phenylalanine (Phe) (summarized in Facchini et al., 2000). The only exception known to date is tomato LeAADC, which can catalyze the conversion of Phe to 2-phenylethylamine, but still prefers Tyr at low substrate concentrations (Tieman et al., 2006). Another enzyme that shares extensive amino acid identity (approximately 65%) with plant TYDCs and TDCs, is Petunia hybrida phenylacetaldehyde synthase (PAAS), which displays strict substrate specificity for Phe and catalyzes an unusual, combined decarboxylation-amine oxidation reaction, leading to the formation of phenylacetaldehyde (PHA) from Phe (Kaminaga et al., 2006).
To determine whether similar enzymes with PAAS activity exist in other plant species, we searched the Arabidopsis thaliana genome for PAAS homologs. This search revealed two genes, At2g20340 and At4g28680, which belong to group II of amino acid decarboxylases. A recent biochemical analysis showed that At4g28680 encodes a TYDC (Lehmann and Pollmann, 2009), whereas the function of the other gene product remains unknown. Here, we report the biochemical and functional characterization of the At2g20340 gene product and show that it is a functional aromatic aldehyde synthase (AAS). AtAAS expression is induced in leaves of A. thaliana Col-0 upon treatment with methyl jasmonate and wounding. This suggests that the product of the AtAAS enzyme, PHA, emitted upon wounding and herbivory, is involved in plant defense. In contrast to ecotype Col-0, AtAAS is responsible for the PHA produced in flowers of A. thaliana ecotypes Sei-0 and Di-G, which emit this compound, and thus it may contribute to pollinator attraction.
Identification of a putative AADC in Arabidopsis thaliana
A search of the A. thaliana genome for genes encoding proteins with homology to the recently identified P. hybrida PAAS (Kaminaga et al., 2006) revealed two genes, At2g20340 and At4g28680, that encode proteins of 490 and 536 amino acids (54.4 and 59.6 kDa), respectively, with high sequence similarity to each other (82%) as well as to P. hybrida PAAS (75 and 73%, respectively). Both proteins belong to group II PLP-dependent amino acid decarboxylases, which include histidine, serine, glutamate and aromatic amino acid decarboxylases (Figure 1a). Although these proteins display highest homology to known plant TYDCs (70–76% similarities), TDCs (64–73% similarities) and PAASs (73–75% similarities), they form a separate subgroup within plant AADCs (Figure 1a). Recently it was shown that At4g28680 encodes AtTYDC (Lehmann and Pollmann, 2009); however, sequence comparison and phylogenetic clustering were not predictive of the specific function of the second putative A. thaliana AADC encoded by At2g20340.
To determine the At2g20340 expression profile and compare it with that of AtTYDC, a quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR) with gene-specific primers was performed using total RNA isolated from rosette and cauline leaves, stems, flowers and roots of A. thaliana (ecotype Col-0). Although the highest transcript level for At2g20340 was found in rosette leaves, the gene was expressed in all tissues analyzed (Figure 1b), in contrast to AtTYDC, which exhibited flower-specific expression (Figure 1c).
Biochemical characterization of Arabidopsis thaliana AADC encoded by At2g20340
To determine the enzymatic activity of the putative A. thaliana AADC, the coding region of At2g20340 was subcloned into the expression vector pET-28a, which contains an N-terminal hexahistidine tag, and expressed in Escherichia coli. The affinity-purified recombinant protein was evaluated for its ability to catalyze the PLP-dependent decarboxylation of potential substrates, including L-tryptophan, Tyr, Phe and l-Dopa. Whereas no activity was detected with tryptophan and Tyr, the At2g20340 encoded enzyme converted Phe and l-Dopa to PHA and dopaldehyde, respectively, as determined by GC-MS and HPLC/GC-MS analyses (Figures 2 and S1, respectively). Thus, this enzyme was designated as A. thaliana aromatic aldehyde synthase (AtAAS). In addition to aldehydes, AtAAS also produced NH4+ and CO2. In contrast to the P. hybrida PAAS-catalyzed reaction (Kaminaga et al., 2006), the reaction catalyzed by AtAAS using Phe as a substrate neither resulted in H2O2 formation nor inhibition upon the addition of H2O2. Kinetic characterization of the purified recombinant AtAAS revealed apparent Km values of 4.12 ± 0.39 mm (mean ± SE, n = 3) and 0.55 ± 0.10 mm (mean ± SE, n = 3) for Phe and l-Dopa, respectively (Table 1). The apparent catalytic efficiency (kcat/Km ratio) of AtAAS with l-Dopa was almost 16-fold greater than that with Phe (Table 1). When Phe was used as a substrate, AtAAS exhibited a broad pH optimum ranging from pH 7.0 to 8.5. However, reactions with l-Dopa were performed at pH 7.5, as l-Dopa is highly unstable at basic pH (Sherald et al., 1973). The molecular mass of active AtAAS determined by gel filtration chromatography was approximately 127 kDa, suggesting that, similar to many plant TDCs and TYDCs (Facchini et al., 2000), and in contrast to the tetrameric P. hybrida PAAS (Kaminaga et al., 2006), the native AtAAS enzyme is a homodimer consisting of two identical 54 kDa subunits.
All values represent means ± SE (n = 3). Km, Michaelis constant; Vmax, maximal velocity; kcat, turnover number.
4.12 ± 0.39
449.6 ± 3.3
0.049 ± 0.0004
0.012 ± 0.001
0.55 ± 0.10
956.3 ± 72.1
0.104 ± 0.008
0.19 ± 0.02
Analysis of AtAAS function in planta
To investigate the in vivo function of AtAAS we screened various T-DNA insertion collections for AtAAS mutants and identified one SALK line (N672304/aas; Alonso et al., 2003) carrying a T-DNA insertion within the 3′ untranslated region (UTR) of the gene (Figure 3a). Sequence analysis of the T-DNA flanking regions revealed that the insertion is located in exon 13 of the gene, 113 bp downstream of the stop codon, and that a fragment of 31 bp adjacent to the insertion site was deleted (Figure 3a). Homozygous aas plants were obtained, as confirmed by genomic PCR analysis with gene-specific primers flanking the insertion site, which failed to amplify the respective gene region (Figure 3b). RT-PCR analysis with two sets of gene-specific primers located at the 5′ and 3′ ends of the gene (the latter flanks the insertion; Figure 3a) revealed that AtAAS transcripts accumulated in this line, but were lacking its authentic 3′ end (Figure 3c). The AtAAS mRNA level in leaves was reduced by 55% in this insertion line relative to the wild type (Figure 3d), as determined by qRT-PCR.
As no other AtAAS mutants could be identified, we generated transgenic A. thaliana plants with reduced AtAAS mRNA levels via an RNAi approach. The RNAi construct was generated using a 238 bp fragment within the 3′ end of the AtAAS coding region and expressed in A. thaliana (ecotype Col-0) under the control of the cauliflower mosaic virus 35S promoter. Fifteen independent transgenic plants were obtained, three of which (lines 12, 14 and 15) exhibited 76–80% reduction in AtAAS expression in leaves, relative to the wild type (Figure 3d). Under normal growth conditions A. thaliana (ecotype Col-0) leaves emit very low levels of PHA (Figure 4b). However, microarray analysis had indicated that the expression of AtAAS (At2g20340), encoding the enzyme that can potentially be responsible for PHA formation in planta, is upregulated by both wounding and jasmonic acid in leaves (Yan et al., 2007). Indeed, treatment with methyl jasmonate and mechanical wounding resulted in an approximately threefold increase in AtAAS transcript levels relative to untreated control leaves (Figure 4a). Wounding also increased PHA emission by approximately fivefold relative to untreated plants (Figure 4b). Analysis of PHA emission from wounded leaves of the aas mutant and RNAi lines revealed that it was reduced by 53 and 45% (on average), respectively, relative to that in the wild type (Figure 3e). As a result, Phe levels in the aas mutant and RNAi lines were increased by 1.3- and approximately 1.8-fold (on average), respectively, when compared with the wild type control (Figure 3f). Moreover, the internal PHA pool in flower tissue of the aas mutant and RNAi lines detected upon methyl jasmonate treatment was reduced by approximately 47% relative to that in the wild type (Figure S2). In summary, these results suggest that AtAAS is indeed responsible for PHA formation in planta.
Different functions of AtAAS in Arabidopsis thaliana ecotypes Col-0, Sei-0 and Di-G
As methyl jasmonate and mechanical wounding are known to simulate insect or pathogen attack, and induce production of volatiles involved in plant defense (Martin et al., 2003; Phillips et al., 2007; Qualley and Dudareva, 2008), feeding experiments with larvae of the biting-chewing herbivore Pieris rapae on 4-week-old A. thaliana ecotype Col-0 plants were performed, with subsequent analysis of emitted leaf volatiles and leaf tissue consumption. Similar to mechanical wounding, caterpillar feeding induced PHA emission from leaves by approximately threefold relative to control plants (Figure 4b). These results suggest that emitted PHA, the product of AtAAS, is involved in plant defense against herbivores. As leaf consumption by P. rapae caterpillars were not statistically different in wild type and aas mutant plants (1031.3 ± 282.5 and 716.3 ± 264.8 mm2 in the wild type and the aas mutant, respectively), PHA is likely to be involved in indirect rather than direct plant defense against herbivores.
Phenylacetaldehyde (PHA) is a common scent component produced by flowers of many plant species (Knudsen et al., 1993). However, it was not detected in flowers of A. thaliana ecotype Col-0 (Figure 5a), which emit low levels of monoterpene and sesquiterpene compounds (Aharoni et al., 2003; Chen et al., 2003; Tholl et al., 2005). Screening of 37 different A. thaliana accessions (Tholl et al., 2005) for the emission of phenylpropanoid compounds identified two ecotypes, Sei-0 and Di-G, in which flowers emit PHA (Figure 5a). qRT-PCR analysis revealed that, in contrast to ecotype Col-0, in Sei-0 and Di-G flowers the AtAAS transcript levels were almost 10-fold higher than in leaves, and exceeded the levels in Col-0 flowers by 15–20-fold (Figure 5b), suggesting that AtAAS is involved in PHA production in the flowers of these A. thaliana ecotypes. This was further supported by RNAi mediated downregulation of AtAAS expression in the Sei-0 ecotype, where a reduction of AtAAS transcripts by 36–45% (shown in two lines out of seven; Figure 5c) led to a 24% reduction in flower internal PHA pools (Figure 5d). Interestingly, in contrast to the Col-0 ecotype (Figure 4), no increase in AtAAS mRNA levels and PHA emission was observed in ecotype Sei-0 leaves in response to mechanical wounding (Figure 5e and f), suggesting that PHA is probably not involved in plant defense in this ecotype.
tydc knock-out mutants are deficient in tyramine synthesis
To gain insight into the in planta function of the other A. thaliana PAAS homolog, AtTYDC, which exclusively catalyzes decarboxylation of l-tyrosine to tyramine in vitro (Lehmann and Pollmann, 2009), we identified two SALK lines (N590725/tydc1 and N671930/tydc2; Alonso et al., 2003) with T-DNA insertions in the AtTYDC gene. Sequence analysis of the T-DNA flanking regions revealed that the insertion in tydc1 is located in the third exon of the gene (amino acid position 129 of AtTYDC), and causes an additional 3 bp deletion, whereas in tydc2 the T-DNA insertion causes a 622 bp deletion from the 3′ end of exon 3 to the intron between exons 4 and 5 (Figure 6a). Homozygous tydc plants were obtained, as confirmed by genomic PCR analysis with gene-specific primers flanking the insertion sites that failed to amplify the respective gene region (Figure 6b). No AtTYDC transcripts were identified in flowers of either tydc mutant by RT-PCR (Figure 6c). Tyramine levels were analyzed in mutant and wild type flowers where AtTYDC is highly expressed (Figure 1c). Under normal growth conditions tyramine was undetectable in wild type flowers, but its levels increased drastically after treatment with methyl jasmonate due to an increase in the level of the substrate Tyr (Figure 6d and e). These results are supported by the fact that methyl jasmonate strongly upregulates the expression of shikimate pathway genes (Pauwels et al., 2008), but has no effect on AtTYDC expression (Figure S3a; Winter et al., 2007; Goda et al., 2008). No tyramine was detected in flowers of either tydc knock-out mutant after treatment with methyl jasmonate, providing genetic evidence that AtTYDC is essential for tyramine formation in planta. The absence of tyramine in mutant flowers led to an increase in the internal pools of Tyr, the AtTYDC substrate, by 1.5- to 1.7-fold relative to the wild type (Figure 6e).
A transcriptome analysis of the haploid male gametophyte development revealed that AtTYDC (At4g28680) is highly expressed in developing pollen, with a maximum at the tricellular stage (Honys and Twell, 2004). In addition, co-expression analysis using BAR-ExpressionAngler (Toufighi et al., 2005) indicated that AtTYDC is co-expressed with callose synthase 5 (CalS5), which is involved in pollen primary cell wall formation (Dong et al., 2005; Nishikawa et al., 2005). To test the potential role of AtTYDC in pollen development, scanning electron microscopy of mature pollen grains from wild type and tydc plants was performed. Only slight changes in the dimensions of tydc pollen were observed relative to the wild type (Figure S3b and c), and the pollen exine wall was properly formed, in contrast to cals5 mutants (Dong et al., 2005; Nishikawa et al., 2005).
As TYDCs are known to be induced by plant pathogens in some plant species (Schmelzer et al., 1989; Kawalleck et al., 1993; Keller et al., 1996; Schmidt et al., 1998), disease symptoms and bacterial growth were analyzed upon infiltration of A. thaliana leaves with the bacterial pathogen Pseudomonas syringae (PstDC3000). However, no differences were observed between the wild type and tydc mutants (Figure S3d), suggesting that AtTYDC is not required for responses to pathogens.
To date, bona fide TYDCs and TDCs have been functionally characterized from a variety of plant species (Facchini et al., 2000), whereas PAASs have been identified only in scent-producing P. hybrida and Rosa hybrida (Kaminaga et al., 2006). Plant TYDCs, TDCs and PAASs belong to group II of amino acid decarboxylases, and their extensive amino acid identity makes it difficult to predict the actual function of the protein based on amino acid sequences. A search for PAAS homologs in A. thaliana resulted in the identification of two genes encoding proteins with approximately 75% homology to P. hybrida PAAS (Kaminaga et al., 2006), and which form a separate subgroup within the plant AADCs (Figure 1a). One gene was recently shown to encode a TYDC (Lehmann and Pollmann, 2009), whereas the biochemical and genetic analyses of the other revealed that it is a functional AAS (Figures 2, 3d–f and S1). Similar to P. hybrida PAAS, AtAAS is a bifunctional enzyme that catalyzes Phe decarboxylation/deamination with PHA formation (Figures 2 and 7), but its apparent Km for Phe is 3.5-fold higher, and catalytic efficiency (kcat/Km ratio) is 56-fold lower, than those of P. hybrida PAAS (Table 1; Kaminaga et al., 2006). In contrast to the strict substrate specificity of PAAS, AtAAS can also use l-Dopa as a substrate (Figure S1), with a 16-fold greater catalytic efficiency than that with Phe (Table 1). Although l-Dopa is a preferred substrate for AtAAS in vitro, to date very little is known about l-Dopa levels in A. thaliana, and we were unable to detect it in planta. In addition, benzylisoquinoline alkaloids, for which l-Dopa is a direct precursor, have not been found in A. thaliana (Liscombe et al., 2005; Ziegler and Facchini, 2008). Although we do not exclude that AtAAS might use l-Dopa as a substrate in vivo, our results show that AtAAS is responsible for PHA formation in planta (Figure 7). A reduction of AtAAS transcript levels by both RNAi suppression and by T-DNA insertion (Figure 3d) led to a corresponding decrease in PHA emission following mechanical damage to leaves (Figure 3e).
It is well documented that plants emit blends of volatiles from their tissues in response to herbivory damage. As aromatic compounds often represent only minor constituents, very little is known about their contribution to plant defense (Qualley and Dudareva, 2008). We found that A. thaliana (ecotype Col-0) leaves emit elevated PHA levels in response to herbivory and mechanical wounding (Figure 4b), implying its potential involvement in plant defense. However, leaf tissue consumption by P. rapae caterpillars on controls and the aas mutants did not differ, suggesting that PHA, the product of AtAAS, might be involved in indirect rather than direct plant defense by attracting specialist parasitoids of P. rapae, such as the parasitic wasp Cotesia rubecula (Agelopoulos and Keller, 1994; Geervliet et al., 1994). Indeed, recently it has been shown that PHA can function as an attractant of the green lacewing Chrysoperla carnea s.l. (Tóth et al., 2006, 2009), an important predator of aphids, caterpillars and other pests of many crops (McEwen et al., 2001).
Phenylacetaldehyde, a common constituent of flower scent blends emitted from many plant species (Knudsen et al., 1993), is an important floral attractant (Raguso et al., 1996; Schiestl and Marion-Poll, 2002; Theis, 2006). Although A. thaliana flowers (ecotype Col-0) do not produce PHA under normal growth conditions, flowers of the Sei-0 and Di-G ecotypes, similar to Arabidopsis lyrata flowers (Peer and Murphy, 2003; Abel et al., 2009), were found to emit PHA as a major floral volatile (Figure 5a). The observed differences in PHA emission between flowers from different ecotypes can be explained by the high levels of AtAAS expression, which are approximately 15–20-fold higher in Sei-0 and Di-G flowers than in the Col-0 ecotype (Figure 5b). In contrast to the Col-0 ecotype, where PHA is emitted upon wounding and insect feeding, mechanical wounding of Sei-0 leaves does not affect AtAAS expression or PHA emission (Figure 5e,f), suggesting that PHA is likely to be involved in pollinator attraction in this ecotype, thus contributing to the outcrossing observed to some extent in natural A. thaliana populations (Snape and Lawrence, 1971; Abbott and Gomes, 1989; Hoffmann et al., 2003). Overall, our results show that AtAAS function is not conserved among different A. thaliana ecotypes. A similar divergence has been observed among Arabidopsis species for the formation of the sesquiterpene (E)-β-caryophyllene, which is synthesized as a predominant floral volatile in A. thaliana, but is produced in leaves of A. lyrata upon herbivore damage (Abel et al., 2009).
The closest AtAAS homolog in A. thaliana, AtTYDC was recently shown to exclusively convert l-tyrosine to tyramine in vitro (Lehmann and Pollmann, 2009). Our results using tydc knock-out lines provide genetic evidence that AtTYDC is indeed responsible for tyramine formation in planta (Figures 6 and 7). Tyramine can be incorporated either directly, or following conjugation with 4-coumaryl or feruloyl residues into the plant cell wall (Negrel and Jeandet, 1987), thus contributing to plant defense by creating a barrier against pathogens (Facchini et al., 2002). However, there were no significant differences in disease symptoms and bacterial growth upon infiltration of wild type and tydc mutant leaves with the bacterial pathogen P. syringae (Figure S3d). Moreover, no AtTYDC upregulation was observed after infection of A. thaliana leaves by the necrotrophic fungus Botrytis cinerea (AbuQamar et al., 2006) or P. syringae (Wang et al., 2008), in contrast to pathogen-responsive plant TYDCs from other plant species (Facchini et al., 2000). In addition, no tyramine-derived hydroxycinnamic acid amides were detected in A. thaliana leaves infected with the necrotrophic fungus Alternaria brassicicola (Muroi et al., 2009), suggesting that AtTYDC is probably not involved in the Arabidopsis response to pathogens. High AtTYDC expression and tyramine accumulation in A. thaliana flowers (Figures 1 and 6d), also observed by Matsuda et al. (2009), is not unique in the plant kingdom, and was found in inflorescences of other plant species including Nicotiana tabacum (Perdrizet and Prevost, 1981) and P. hybrida (Miersch et al., 1998). Further investigation is required to determine the exact function of tyramine and AtTYDC in flowers.
Plant material and growth conditions
Arabidopsis thaliana T-DNA insertion lines (N672304/aas, N590725/tydc1, N671930/tydc2; Alonso et al., 2003) were obtained from the Arabidopsis Stock Center at Ohio State University. Arabidopsis thaliana (ecotype Col-0) wild-type, transgenic and mutant plants, as well as A. thaliana ecotypes Sei-0 and Di-G were grown under a 16-h photoperiod in standard glasshouse conditions, and in growth chambers under a 21°C 12-h day/18°C 12-h night cycle during metabolic and volatile analyses.
Chemicals and substrates
All reagents were purchased from Sigma-Aldrich (http://www.sigmaaldrich.com), unless otherwise noted. l-[U-14C]Phe (425 mCimmol−1) and l-3,4-dihydroxyphenyl [3-14C]alanine (54 mCi mmol−1) were purchased from American Radiolabeled Chemicals Inc. (http://www.arc-inc.com) and Amersham (now GE Healthcare, http://www.gelifesciences.com), respectively. Dopaldehyde standard was kindly provided by Dr Kenneth L. Kirk (National Institutes of Health, Bethesda, MD, USA).
Homozygous aas, tydc1 and tydc2 mutant plants were identified by PCR on genomic DNA using gene-specific primers flanking the T-DNA insertions as well as a T-DNA border-specific primer (LBb1): aas forward 5′-TGGTTGTTGCATACCATCATC-3′ and reverse 5′-ATAACCGAAACCGTGAACTCC-3′; tydc1 forward 5′-CAACAGCTGCCACTACCTTTC-3′ and reverse 5′-GGTCACAAAAACGTATGGTCG-3′; tydc2 forward 5′-CCTGGTTATCTCCGTGACATG-3′ and reverse 5′-CAAGTGATTCTGGAGGCATTC-3′; LBb1, 5′-GCGTGGACCGCTTGCTGCAACT-3′. The PCR products obtained with LBb1/forward gene-specific primers and LBb1/reverse gene-specific primers were sequenced.
Generation of AtAAS-RNAi Arabidopsis plants
A 238-bp fragment of the 3′ UTR of the AtAAS cDNA (nucleotides 1568–1805), obtained by PCR, was cloned into the binary vector pGSA1131 under the control of the cauliflower mosaic virus 35S promoter in sense and antisense orientations, separated by the GUS intron spacer region. Transgenic A. thaliana ecotypes Col-0 and Sei-0 were obtained via Agrobacterium tumefaciens (strain LBA4404 carrying the AtAAS-RNAi construct)-mediated transformation using the floral-dip method (Clough and Bent, 1998). Seeds of infiltrated plants were harvested and germinated. Transgenic seedlings were selected by spraying with 50 μl L−1 BASTA.
RNA isolation and qRT-PCR
Total RNA was isolated from different tissues of wild-type, transgenic and mutant A. thaliana plants, as described by Eggermont et al. (1996). For qRT-PCR analysis, total RNA was pre-treated with RNase-free DNase (Promega, http://www.promega.com), and cDNA was synthesized using Reverse Transcriptase (Superscript II; Invitrogen, http://www.invitrogen.com). Gene-specific primers AtAAS forward 5′-CCGAAACCGTGAACTCCTAGAC-3′ and reverse 5′-CGCCTCCTTAACGTGCTTCT-3′, AtTYDC forward 5′-GATCCAAGTTTTGAGGTTGTCACTA-3′ and reverse 5′-ACGGTTACGTTCGTTACATTGG-3′, and Ubc (ubiquitin conjugating enzyme) forward 5′-CTGCGACTCAGGGAATCTTCTAA-3′ and reverse 5′-TTGTGCCATTGAATTGAACCC-3′, were designed using primerexpress software (Applied Biosystems, http://www.appliedbiosystems.com). All primers showed more than 90% efficiency at a final concentration of 300 nm. qRT-PCR reactions were performed as described previously (Orlova et al., 2009) using the StepOne Real-Time PCR system (Applied Biosystems). Absolute and relative AtAAS and AtTYDC transcript levels were identified as described by Maeda et al. (2010). Each data point represents an average of at least three or four independent biological samples, with three technical replicates for each sample.
Expression of AtAAS in Escherichia coli, and purification of recombinant enzyme
The coding region of AtAAS was PCR-amplified using forward, 5′-GCTAGCATGGAAAATGGAAGCGGGAAGGTG-3′ and reverse, 5′-GGATCCTTACTTGTGAAGCAAGTAAG-3′ primers and subcloned into the NheI/BamHI restriction sites of the pET28a expression vector containing an N-terminal hexahistidine tag (Novagen, now part of Merck, http://www.merck-chemicals.com). Sequencing confirmed that no errors had been introduced during PCR amplification.
Expression in E. coli BL21 Rosetta cells, induction, harvesting and crude extract preparation were performed as previously described (Kaminaga et al., 2006). Protein purification was performed by affinity chromatography on a HisTrap FF column (1 ml; GE Healthcare) using an FPLC system (AKTA 900 series; GE Healthcare). After washing the column with 20 and 40 mm imidazole in a buffer containing 50 mm potassium phosphate, pH 8.0, and 200 mm PLP, the (His)6-tagged enzyme was eluted with 500 mm imidazole in the same buffer. The fractions with the highest AAS activity were desalted on Econo-Pac 10 DG columns (Bio-Rad, http://www.bio-rad.com) into a buffer containing 20 mm potassium phosphate, pH 8.0, and 4% glycerol. The purity of the protein, determined by densitometry of SDS-PAGE gels after Coomassie Brilliant Blue staining, was 74%, which was taken into account for the determination of kcat values. Protein concentrations were determined using the Bradford method (Bradford, 1976).
Aromatic aldehyde synthase enzyme assays and product identification
Aromatic aldehyde synthase (AAS) activity was analyzed as described previously (Kaminaga et al., 2006) using 2 mm l-[U-14C]Phe (45 nCi) or 2 mm l-[14C]Dopa (25 nCi) and 83.4 or 16.7 mg purified AtAAS protein, respectively. The pH of reactions with l-Dopa was adjusted to 7.5 to avoid spontaneous substrate degradation (Sherald et al., 1973). After terminating the reaction with 5 μl of 50% trichloroacetic acid, the product was extracted with 250 μl of ethyl acetate and 200 μl of the organic phase was counted in a scintillation counter.
For product verification, reactions were performed with unlabeled substrates and product formation was analyzed either spectrophotometrically at 280 nm on the Agilent 1200 HPLC system using an Agilent ZORBAX SB-C18 column (4.6 × 150 mm × 3.5 mm) or by GC-MS (5975 inert XL EI/CI mass spectrometer detector combined with 6890N GC; Agilent Technologies, http://www.agilent.com) using an Agilent 19091S-433 HP-5MS capillary column (30 m × 0.25 mm; film thickness 0.25 μm). For HPLC separation, mobile phases were (A) 1% methanol in 88 mm KH2PO4, pH 3.9, 1 mm EDTA and (B) 100% methanol. A linear gradient from 0 to 30%B was applied over 20 min with holding 30%B for 20 min followed by 5 min equilibration back to 0%B at a 0.2 ml min−1 flow rate and 40°C column temperature. Analysis of NH4+ and H2O2 formation was performed as described by Kaminaga et al. (2006).
For kinetic analysis an appropriate enzyme concentration was chosen so that the reaction velocity was proportional to the enzyme concentration, and remained linear during the incubation time period. Kinetic data were evaluated by hyperbolic regression analysis (hyper.exe v1.00, 1992). Triplicate assays were performed for all data points.
Analysis of PHA, tyramine and aromatic amino acid internal pools
Shoots of A. thaliana plants were incubated in 100 mm methyl jasmonate for 12 h at room temperature (23°C), and flower tissues were collected, frozen in liquid nitrogen and stored at −80°C. The analysis of PHA, tyramine and aromatic amino acid internal pools was performed as described previously (Orlova et al., 2006; Maeda et al., 2010).
Wounding, insect feeding and pathogen treatments of Arabidopsis plants
Rosette leaves of 4-week-old A. thaliana plants were wounded using an array of 50 household needles, and volatiles were collected for 24 h. Herbivory experiments were performed with larvae of the white brassica butterfly P. rapae. Eggs (Carolina Biological Supply, http://www.carolina.com) were hatched and larvae were reared on cabbage leaves. For feeding experiments, four or five larvae were placed on 4-week-old A. thaliana plants for 24 h. After removal of the larvae, plants were used for volatile collection.
For the analysis of pathogen response, 4-week-old A. thaliana plants were inoculated with P. syringae pv. tomato DC3000 (PstDC3000) as described by Veronese et al. (2006). Bacterial disease assays were performed as described by Mengiste et al. (2003) and Veronese et al. (2006). To determine bacterial growth, leaf discs of the same size were collected from infected leaves at 0 and 3 days after inoculation, and bacterial titer per leaf area was determined. Experiments were performed in triplicate, with each experiment containing 10 leaves.
Volatile collection and analysis
Leaf and floral volatiles were collected from A. thaliana plants using a closed-loop stripping method under growth chamber conditions (21°C, 50% relative humidity, 150 μmol m−2 s−1 light intensity and a 12-h photoperiod; Donath and Boland, 1995; Dudareva et al., 2005). Root balls of individual 4-week-old plants were wrapped in aluminum foil and leaf volatile collections were performed for 24 h using Porapak Q traps (80/100 mesh size; Alltech Associates, Deerfield, IL, USA) and analyzed as described previously (Dudareva et al. (2005). To increase the sensitivity and reproducibility of PHA detection and quantification, volatile samples were analyzed in selective ion monitoring (SIM) mode. For floral volatiles, 70 inflorescences were transfered to small glass beakers filled with water, and collections were performed for 8 h as described by Tholl et al. (2005).
Scanning electron microscopy
Wild type and tydc mutant pollen was attached to a sample holder, coated with gold-palladium using an ion sputter coater (Hummer™ Sputter coater; Anatech USA, http://www.anatechusa.com) and analyzed under a scanning electron microscope (Nova™ NanoSEM; FEI, http://www.fei.com). Size and shape measurements were obtained with imagej (http://rsbweb.nih.gov/ij).
The sequence reported in this paper has been deposited in the GenBank database (accession no. HQ843094).
This work was supported by the Agricultural and Food Research Initiative Competitive Grants no. 2010-65115-20385 and no. 2010-65115-20383 from USDA National Institute of Food and Agriculture to ND and FNZ, respectively, by grant MCB-0615700 from the National Science Foundation to ND, and by a National Science Foundation Advance VT research seed grant and the Thomas and Kate Jeffress Memorial Trust (J-850) to DT.