Introduction

Flavonoids, a class of polyphenolic compounds produced via the phenylpropanoid biosynthetic pathway in higher plants, consist of more than 9,000 structures (Williams and Grayer 2004). Flavonoids have many functions in plants including protection from UV light, defense against pathogenic microorganisms, or coloration of flowers (Dixon and Paiva 1995; Dixon 2004; Tanaka et al. 2008). Over the past decade flavonoids have been attracting the attention of many researchers for their pharmacological and nutritional properties. Among the flavonoids, the anticancer and antioxidant activities of kaempferol and quercetin make them good candidates for inclusion in the human diet (Pietta 2000; Ren et al. 2003; Williams et al. 2004).

Lignins, polymers of 4-hydroxyphenylpropanoids, are also derived from the phenylpropanoid pathway (Boerjan et al. 2003). In plants, lignins play important roles in the development and resistance to biotic and abiotic stresses because the polymers confer strength and rigidity to the cell wall. On the other hand, lignins hinder the use of plant biomass in chemical pulping and processing to biofuels, besides decreasing the digestibility of forage.

The initial step in the phenylpropanoid biosynthetic pathway in higher plants is catalyzed by phenylalanine ammonia-lyase (PAL), which converts phenylalanine to trans-cinnamic acid by the elimination of ammonia. Cinnamate 4-hydroxylase (C4H), a membrane-anchored cytochrome P450 monooxygenase enzyme, subsequently hydroxylates trans-cinnamic acid to p-coumaric acid. p-Coumaric acid is a precursor of 4-coumaroyl-CoA, which serves as a substrate to form the basic skeleton of all flavonoid derivatives (Ehlting et al. 1999).

PAL is a metabolically important enzyme because its reaction links primary metabolism to secondary metabolism. Furthermore, PAL has been known as a rate-limiting enzyme in the phenylpropanoid pathway (Hahlbrock et al. 1976). The genomic DNA of the model plant Arabidopsis encodes four PAL isoforms, and Cochrane et al. (2004) reported that all four recombinant PAL proteins recognize phenylalanine, but not tyrosine as a substrate. It has been reported that multiple mechanisms, such as feedback inhibition and transcriptional regulation, regulate PAL activity in plants (Bolwell et al. 1986, 1988; Jorrin and Dixon 1990; Mavandad et al. 1990; Appert et al. 1994). In fact, increased expression of PAL enhances the accumulation of phenylpropanoids such as chlorogenic acid while major flavonoids such as rutin were not altered in the tobacco plants overexpressing PAL (Bate et al. 1994; Howles et al. 1996; Shadle et al. 2003). It has also been postulated that the initial metabolic flux into the phenylpropanoid biosynthetic pathway is regulated by an enzyme complex consisting of PAL and C4H, which are co-localized on the ER membrane (Blount et al. 2000; Achnine et al. 2004). Thus, the regulatory mechanisms governing the initial steps of phenylalanine metabolism are complicated.

On the other hand, it has been reported that some photosynthetic bacteria possess a unique soluble enzyme, tyrosine ammonia-lyase (TAL), that catalyzes the direct conversion of tyrosine to p-coumaric acid in order to produce a chromophore of a photosensory protein (Kyndt et al. 2002; Xue et al. 2007). We therefore expected that the introduction of the TAL gene into plants would open an extra p-coumaric acid pathway and increase the initial flux into phenylpropanoid biosynthesis pathway (Fig. 1). In the present study, we describe the characterization of Arabidopsis transformants expressing TAL gene that originates from Rhodobacter sphaeroides. We analyzed metabolites in the transformed plants, and found that the expression of TAL efficiently enhanced the metabolic flux into the phenylpropanoid pathway, resulting in the accumulation of phenylpropanoid derivatives such as quercetin glycosides.

Fig. 1
figure 1

Summary of phenylpropanoid and flavonoid biosynthetic pathways in plants. TAL tyrosine ammonia-lyase, PAL phenylalanine ammonia-lyase, C4H cinnamate 4-hydroxylase, 4CL 4-coumarate:CoA ligase, CHS chalcone synthase, CHI chalcone isomerase, F3H flavanone 3-hydroxylase, F3′H flavonoid 3′-hydroxylase, FLS flavonol synthase, DFR dihydroflavonol 4-reductase, LDOX leucoanthocyanidin dioxygenase. The boxed part indicates the introduced conversion step catalyzed by TAL in this study

Full size image

Materials and methods

Plant material and growth conditions

Arabidopsis thaliana ecotype Columbia (RIKEN Plant Science Center, Yokohama, Japan) was used in this study. Seeds were sown on soil medium and stratified at 4°C for 3 days before transfer to a growth chamber maintained under long-day (16-h light, 8-h dark) conditions. Temperature and humidity were maintained at 23°C and 50%, respectively.

For non-targeted metabolic profiling, transformed and wild-type Arabidopsis were cultured on solid MS medium [Murashige and Skoog plant salt mixture, Gamborg’s B5 medium vitamin mixture, 1% sucrose, 0.05% Mes buffer (pH 5.7), 0.8% agar] for 3 weeks under long-day (16-h light, 8-h dark) condition. Whole aerial part of seedlings were collected and frozen in liquid nitrogen and stored at −85°C before use.

For flavonoid-targeted analysis, plants were cultured on solid 1/2 MS medium with 1% sucrose under long-day (16-h light, 8-h dark) condition. Plates were set up in vertical orientation. Five-day old seedlings were collected, frozen in liquid nitrogen and stored at −85°C until analysis.

Cloning of TAL from Rhodobacter

Rhodobacter sphaeroides (NBRC no. 12203) was obtained from NBRC [National Institute of Technology and Evaluation (NITE) Biological Resource Center, Kisarazu, Japan]. Genomic DNA of R. sphaeroides was extracted and used as a template to amplify a 1.6-kb DNA fragment containing the open reading frame (ORF) of the RsTAL gene. Primers used are as follows: RsTAL1 (5′-AAAACTAGTATGCTCGCCATGAGC-3′, SpeI site underlined) and RsTAL2 (5′-AAAGTCGACTCAGACGGGAGATTGC-3′, SalI site underlined). The amplified fragment was cloned into the SpeI and SalI sites of pEU3S (Kanno et al. 2005) to yield pEU3S-RsTAL. The DNA sequence of the inserted fragment was identical to that of the R. sphaeroides TAL gene as reported previously (GenBank accession no. YP_355075).

Construction of expression vector and plant transformation

The Sse8387I fragment containing RsTAL was excised from the pEU3S-RsTAL and cloned into the Sse8387I site of the binary vector pYT105 (Yun et al. 2008) to obtain pYT105-RsTAL. The binary vector pYT105-RsTAL and empty vector pYT105 were introduced into Agrobacterium tumefaciens C58C1 by the freeze–thaw method (Höfgen and Willmitzer 1988), respectively. Transformation of A. thaliana (Columbia) was performed by in planta infiltration (Clough and Bent 1998). Transformed seedlings (T1) were identified by selection on solid MS medium containing 50 mg/l kanamycin and 200 mg/l claforan, then transferred to soil. Homozygous T3 plants: RsTAL-expressing plants, TA3, TB7, TE3, TM1 and TL6; and plants transformed with empty vector, YT1, YT4 and YT5, were obtained and used in subsequent analyses.

Detection of RsTAL transcript

For RNA analysis, total RNA was extracted from whole seedlings of wild-type and transgenic Arabidopsis plants with the use of RNeasy Plant Mini Kit (Qiagen, Hilden, Germany), and 5 μg of the RNA were subjected to RNA gel-blot hybridization. A digoxigenin-labeled antisense RNA probe for the detection of RsTAL transcripts was prepared with a DIG RNA labeling kit (Roche, Mannheim, Germany). A 375-bp fragment of the RsTAL ORF (nucleotides 91–485) was amplified by PCR with the primers RsTAL3 (5′-AAAGAATTCCGGATTGTCCTTG-3′) and RsTAL4 (5′-CCCAAGCTTCAGAGCACCATAT-3′), then cloned into pSPT19 (Roche). The antisense RNA probe was synthesized from the obtained vector by in vitro run-off transcription with T7 RNA polymerase in the presence of digoxigenin-labeled UTP. RNA gel-blot hybridization was performed as described previously (Tozawa et al. 1998).

Non-targeted metabolic profiling analysis using LC-PDA-MS

The frozen tissues were mixed with five volumes of 80% aqueous methanol containing 0.5 mg/l of lidocaine and d-camphorsufonic acid and homogenized in a mixer mill (MM 300, Retsch, Haan, Germany) with a zirconia bead for 10 min at 20 Hz. Following centrifugation at 15,000g for 10 min, the supernatants were passed through a filter (Ultrafree-MC, 0.2 μm, Millipore, Bedford, MA, USA) to remove the insoluble materials. The sample extracts (3 μl) were analyzed using an LC-MS system equipped with an electrospray ionization interface (HPLC: Waters Acquity UPLC system; MS: Waters Q-TOF Premier, Waters, Milford, MA, USA). Data acquisition and processing were performed as previously described (Matsuda et al. 2009). Flavonoid-targeted analyses using PDA chromatograms were performed by use of reported profiles of flavonoids accumulated in A. thaliana (Tohge et al. 2005).

Flavonoid-targeted analysis of 5-day-old seedlings

Flavonoid-targeted analyses were performed in triplicate. Frozen seedlings were homogenized in 5 μl of extraction solvent (methanol:CH3COOH:H2O = 9:1:10, 0.02 mM naringenin-7-O-glucoside) per milligram of fresh weight of tissue in the mixer mill (MM 300, Retsch) for 5 min at 30 Hz. After centrifugation at 12,000g, the supernatants were immediately used for flavonoid analysis. The UPLC-MS system mentioned above was used. Data acquisition and processing were performed as previously described (Yonekura-Sakakibara et al. 2009).

Results

Transgenic Arabidopsis expressing RsTAL

To enhance the carbon flow into the phenylpropanoid pathway in plants, we cloned the gene encoding TAL from R. sphaeroides, which has been reported to manifest specific and efficient catalytic activity in the conversion of tyrosine to p-coumaric acid when expressed in Escherichia coli (Watts et al. 2004). To express RsTAL protein, the RsTAL gene under the control of CaMV 35S promoter was introduced into Arabidopsis (Fig. 2a). Kanamycin-resistant T1 plants were selected and, eventually, five independent lines (TA3, TB7, TE3, TM1 and TL6) were obtained as homozygous T3 plants. The accumulation of RsTAL transcripts in seedlings of these lines was confirmed by RNA gel-blot hybridization (Fig. 2b).

Fig. 2
figure 2

Expression of RsTAL in transgenic Arabidopsis. a Structure of the T-DNA region of the binary vector pYT105-RsTAL. RB and LB right and left borders, respectively, P 35S 35S promoter of cauliflower mosaic virus, Pnos and Ter promoter and terminator of the nopaline synthase gene, respectively, NPTII neomycin phosphotransferase gene. b Detection of RsTAL mRNA in transgenic Arabidopsis lines. Total RNA (5 μg) from wild-type and five independent lines of T3 transgenic seedlings (TA3, TB7, TE3, TM1, TL6) were subjected to the analysis. The 2.1-kb RsTAL transcript is indicated by an arrowhead. The membrane was also stained with methylene blue for visualization of 25S and 17S rRNA

Full size image

We detected no significant differences in morphology or growth rate among the five RsTAL homozygous lines and wild-type plants (data not shown), indicating that RsTAL expression, even at a high level, does not affect normal development of Arabidopsis. Four independent lines (TA3, TB7, TM1 and TL6) exhibiting comparatively higher accumulation of transcripts were selected for further investigation.

LC-PDA-MS analysis of phenylpropanoid compounds in transgenic plants

To analyze the effect of TAL expression on phenylpropanoid metabolism, we conducted LC-PDA-MS analysis of 21-day-old seedlings expressing RsTAL.

The PDA chromatogram showed that RsTAL-expressing plants accumulated more quercetin glycosides and sinapoylmalate than control plants. Sinapoylmalate and total quercetin glycosides in the RsTAL-expressing plants are 1.7–2.5 and 5–10 times higher than those in wild type, respectively (Fig. 3a, b). The content of quercetin 3-O-(2″-O-rhamnosyl)-glucoside 7-O-rhamnoside, which is a more modified version of a quercetin-derivative than quercetin 3-O-glucoside-7-O-rhamnoside, is constantly higher in plants expressing RsTAL than in the wild type and vector control. On the other hand, kaempferol glycosides, such as kaempferol-3-O-rhamnoside 7-O-rhamnoside and kaempferol-3-O-(2″-O-rhamnosyl)-glucoside 7-O-rhamnoside, were not increased in three RsTAL-expressing lines, TA3, TL6 and TM1 (Fig. 3c). Anthocyanins were not detected in the 21-day-old seedlings.

Fig. 3
figure 3

Phenylpropanoid compounds accumulated in the transgenic lines expressing RsTAL. Extracts from 21-day-old seedlings were analyzed by LC-PDA-MS. Flavonol glycosides and sinapoylmalate were identified in the PDA chromatogram and their amounts were compared. Analyses were performed in quintuplicate for RsTAL-expressing lines and wild type, and in triplicate for empty vector control (YT1). Mean values with SD are shown. a Relative quantification of quercetin glycosides. b Relative quantification of sinapoylmalate. c Relative quantification of kaempferol glycosides

Full size image

Non-targeted metabolite profiling was also conducted on the same samples. From the 2,250 peaks detected in positive-ion mode, peaks that displayed more than threefold higher or lower average ion intensity in RsTAL-expressing plants when compared with those in wild-type controls were selected. There were 16 peaks with increased ion intensity (UK1–16) while no peak with decreased intensity was detected (Table 1). UK14 was identified as sinapoylmalate by comparison of the chromatographic behavior with that of authentic compound. We could not detect ions derived from quercetin glycosides in MS analysis because quercetin glycosides co-migrated with other unknown compounds during LC separation.

Table 1 Compounds increased in the RsTAL expressing Arabidopsis detected by LC-MS analyses
Full size table

Among 16 peaks with increased ion intensity, there were four pairs of peaks with the same m/z value (UK1, UK2 (m/z 277); UK3, UK5 (m/z 369); UK4, UK6 (m/z 353); UK10, UK12 (m/z 354)). Each pair was assumed to be a set of isomers because two peaks showed similar MS/MS spectrum. MS/MS analysis of peaks UK4, UK6 (m/z 353), UK10, UK12 (m/z 354), UK3 and UK5 (m/z 369) gave fragments with m/z 207, 175, 147, 119, 91. The fragment pattern was considered to be derived from a sinapoyl moiety. Based on an accurate estimation of their molecular masses, UK4 and UK6 (m/z 353.1318) and peaks UK10 and UK12 (m/z 354.1119) were deduced to be sinapoylglutamine (C16H20N2O7, theoretical molecular weight of protonated molecule 353.1349) and sinapoylglutamate (C16H19NO8, theoretical molecular weight of protonated molecule 354.1189), respectively. Each pair was supposed to consist of cis/trans isomers. However, the pertinent compounds were not found in CAS database. MS/MS analysis of peaks UK1 and UK2 (m/z 277.1623) gave fragments derived from cleavage of p-coumaroyl residue, i.e. fragments with m/z 147, 119 and 91. We annotated these peaks as cis/trans isomers of p-coumaroylagmatine (C14H21N4O2, theoretical molecular weight of protonated molecule 277.1665) based on a search of the CAS database. We could not deduce structures of peaks UK3 and UK5.

Increase of the compound corresponding to UK16 was also confirmed in the comparison of PDA chromatograms. MS/MS analysis of the peak gave fragments with m/z 207, 175, 147 and 119 though the intensities of fragment ions were relatively low. The compound was deduced to be a sinapoyl derivative whose molecular formula is C17H25NO6 (theoretical molecular weight of protonated molecule 338.1603). Mass spectrometry data and deduced structures of UK1, UK2, UK4, UK6, UK10 and UK12 are shown in supplementary material (Suppl. data 1).

In brief, we identified one peak and putatively annotated 6 peaks out of 16 increased peaks based on their UV–Visible absorption spectra and comprehensive analysis of mass fragmentation patterns obtained by tandem MS spectroscopy (Table 1). Five peaks were deduced to have sinapoyl residues while two peaks were apparently p-coumaroyl derivatives among annotated peaks. Three unannotated peaks were also deduced to be sinapoyl derivatives.

Taken together, LC-PDA-MS analysis showed that plants expressing RsTAL had higher levels of quercetin glycosides and phenylpropanoid metabolites, which have sinapoyl or p-coumaroyl residues, as compared to wild-type plants.

Flavonoid-targeted analysis of 5-day-old seedlings

In order to analyze phenylpropanoid metabolism in other growth stages, we carried out metabolic profiling of flavonoids in 5-day-old seedlings of Arabidopsis expressing RsTAL. Flavonoid-targeted analysis of transgenic plants revealed that the expression of RsTAL resulted in slightly higher accumulation of quercetin 3-O-rhamnoside 7-O-rhamnoside in two lines, TA3 and TM1, when compared with wild type (1.6 and 1.9 times, respectively) (Fig. 4a). There was no apparent change in other quercetin glycosides. Increase of sinapoylmalate was less obvious than that of quercetin-derivatives, and we only observed a slight increase in TM1 line (Fig. 4b). In TM1, kaempferol glycosides, kaempferol 3-O-rhamnoside 7-O-rhamnoside and kaempferol 3-O-glucoside 7-O-rhamnoside, also increased slightly compared to wild type (1.4 and 1.5 times, respectively) (Fig. 4c). Changes in amount of these compounds in 5-day-old seedlings were not prominent as those in 21-day-old seedlings.

Fig. 4
figure 4

Analyses of flavonoid accumulation in 5-day-old seedlings of transgenic plants expressing RsTAL. Extracts from 5-day-old seedlings were analyzed by LC-PDA-MS. Relative MS peak intensities of ions derived from flavonol glycosides, sinapoylmalate and anthocyanins were compared with that of wild type. YT1, YT4 and YT5 are control plants transformed with an empty vector. Mean values and SD of three biologically independent experiments are shown. a Relative quantification of quercetin glycosides. b Relative quantification of sinapoylmalate. c Relative quantification of kaempferol glycosides. d Relative quantification of anthocyanin (cyanidin 3-O-[2′′-O-(2′′′-O-(sinapoyl)-xylosyl)-6′′-O-(p-O-(glucosyl)-p-coumaroyl)-glucoside] 5-O-(6′′′′-O-malonyl) glucoside)

Full size image

In contrast, RsTAL expression led to significantly higher accumulation of major anthocyanins when compared with the wild-type controls (Fig. 4d). The RsTAL transgenic lines accumulated 2.2–4.0 times higher amount of cyanidin 3-O-[2′′-O-(2′′′-O-(sinapoyl)-xylosyl)-6′′-O-(p-O-(glucosyl)-p-coumaroyl)-glucoside] 5-O-(6′′′′-O-malonyl) glucoside than the wild-type controls (Fig. 4d). Increase of another anthocyanin, cyanidin 3-O-[2′′-O-(2′′′-O-(sinapoyl)-xylosyl)-6′′-O-(p-coumaroyl)-glucoside] 5-O-(6′′′′-O-malonyl) glucoside, was also dependent on the expression of RsTAL (data not shown).

Discussion

To date, there have been a number of reports of metabolic engineering approaches aimed at increased accumulation of flavonoids. Some reports have described modification of enzyme activity that limited the carbon flow into the phenylpropanoid pathway. Transgenic tobacco plants with higher PAL activity were selected from transgenic plants expressing a heterologous PAL gene (Elkind et al. 1990; Howles et al. 1996). Chlorogenic acid was increased in plants with higher PAL activity while the flavonoid rutin was not increased. Over-expression of C4H in tomato gave rise to some plants with increased rutin and naringenin, though increased level of those compounds was not directly correlated to increased level of C4H (Millar et al. 2007). Some reports revealed an increase in flavonoids by activating transcription factors. For example, anthocyanins and quercetins were specifically increased in PAP1-overexpressing Arabidopsis (Borevitz et al. 2000; Tohge et al. 2005) while quercetins were specifically increased in MYB12-overexpressing Arabidopsis (Mehrtens et al. 2005). These results indicate that over-expression of transcription factor is a useful tool in metabolic engineering for increased production of specific metabolites. Over-production of proanthocyanidins was reported by over-expression of PAP1 in combination with MYB transcription factor TT2 only in a subset of cells (Sharma and Dixon 2005). Co-expression of PAP1 and anthocyanidin reductase resulted in ectopic production of proanthocyanidins in transgenic plants (Xie et al. 2006). These results suggest that co-expression of multiple genes may help to produce a specific class of metabolites.

In this report, we increased the metabolic flux to the phenylpropanoid biosynthetic pathway by expressing TAL, which catalyzes the reaction from tyrosine to p-coumaric acid. TAL has been found to be involved in the development of the photosensory system in photosynthetic bacteria, R. sphaeroides, by producing p-coumaric acid as the chromophore of the photosensory yellow protein (Kort et al. 1998; Kyndt et al. 2002). As trans-cinnamic acid is not detected in the bacteria, tyrosine has been considered to be the natural substrate of RsTAL in vivo. In fact, the K m values of RsTAL are 60 and 560 μM for tyrosine and phenylalanine, respectively, while k cat values for both substrates are similar (0.02 and 0.01 s−1 for tyrosine and phenylalanine, respectively) (Xue et al. 2007). These parameters indicate that RsTAL can catalyze the deamination of tyrosine 19 times more efficiently than that of phenylalanine.

Maize PAL was confirmed to catalyze tyrosine deamination besides phenylalanine deamination using recombinant protein (Rösler et al. 1997). k cat/K m values of recombinant ZmPAL1, measured at pH 7.7, are 0.018 and 0.023 μM−1 s−1 for phenylalanine and tyrosine, respectively. ZmPAL1 catalyzes the deamination of phenylalanine and tyrosine at similar efficiencies. On the other hand, all four PAL isoforms were supposed to accept only phenylalanine in Arabidopsis (Cochrane et al. 2004). For AtPAL1, 2 and 4, enzyme efficiencies for the deamination of tyrosine are 340-, 1,200- and 970-times lower, respectively, than those for phenylalanine. AtPAL3 is essentially inactive with tyrosine. Based on these catalytic properties of AtPALs and RsTAL, we expected that the expression of RsTAL in Arabidopsis will open a novel phenylpropanoid pathway by utilizing tyrosine as an extra primary substrate.

Metabolite profiling using LC-PDA-MS revealed that RsTAL expression in Arabidopsis leads to higher accumulation (5–10-fold) of quercetin glycosides. On the other hand, we did not observe increased accumulation of kaempferol glycosides in the transgenic lines. It seems that the downstream effect of the enhanced influx into flavonoid pathway is limited to an increase of quercetin. Similarly, a previous report has shown that PAP1-overexpressing plants only increased the level of quercetin glycosides while kaempferol glycosides decreased or remained unchanged (Tohge et al. 2005).

Besides quercetin glycosides, we expected that phenylpropanoids with sinapoyl or coumaroyl residues will be highly accumulated in seedlings expressing RsTAL. MS/MS analyses revealed that eight peaks with sinapoyl residue and two peaks with coumaroyl residue were increased in transgenic plants (Table 1). In addition, flavonoid-targeted analysis of 5-day-old seedlings revealed that anthocyanin is increased in RsTAL-expressing seedlings (Fig. 4). Taken together, metabolite analyses revealed that 21- and 5-day-old seedlings have different metabolic profiles, and confirmed that RsTAL expression results in the accumulation of various end products instead of just a specific class of metabolites, most likely in a growth stage-specific manner. We also conducted semi-quantitative RT-PCR analysis. However, no obvious change was observed in the expression of genes related to flavonoid biosynthesis, such as PAL, C4H, 4CL, CHS, CHI, F3H, F3′H, FLS, MYB12, and MYB111, in 21-day-old seedlings of transgenic lines (Suppl. data 2). Similarly, the expression of genes related to anthocyanin biosynthesis, such as DFR, LDOX, and PAP1, was not changed in 5-day-old seedlings (Suppl. data 2). The expression of RsTAL therefore does not seem to affect gene regulation of related enzymes in the flavonoid biosynthetic pathway.

Increased amount of various end products derived from phenylpropanoid pathway suggested that TAL expression enhanced the metabolic flux into the phenylpropanoid pathway. By expressing TAL protein, plants can utilize tyrosine as an alternative initial substrate of the phenylpropanoid pathway. Increased supply of p-coumaric acid was assumed to enhance the downstream metabolic flux in the phenylpropanoid pathway, though we could not detect p-coumaric acid in either RsTAL expressing nor wild-type seedlings. It is possible that the conversion of p-coumaric acid into 4-coumaroyl-CoA by 4-coumarate:CoA ligase may be efficient enough to prevent accumulation of p-coumaric acid and its consequent detection by LC-PDA-MS.

Plants with high levels of RsTAL expression had normal wild-type morphotypes. This suggests that overexpression of RsTAL has no visibly adverse effect on plants. In some cases, altered accumulation levels of secondary metabolites cause adverse effects on growth, morphology, and fertility (Ylstra et al. 1994; Fischer et al. 1997; Sharma and Dixon 2005; Millar et al. 2007). Normal growth and fertility are advantageous characteristics in metabolic engineering. TAL is expected to be co-expressed with other gene(s) for accumulation of specific class of metabolites. We reported functional expression of flavone synthase type I in Arabidopsis (Yun et al. 2008). RsTAL expression can increase the supply of the substrate naringenin to flavone synthase.

In summary, the phenylpropanoid pathway in Arabidopsis was activated by expressing a TAL gene cloned from a photosynthetic bacterium. This approach will facilitate metabolic engineering of the phenylpropanoid pathway in many kinds of plants.