Isolation and Characterization of cDNAs Encoding Leucoanthocyanidin Reductase and Anthocyanidin Reductase from Populus trichocarpa

Proanthocyanidins (PAs) contribute to poplar defense mechanisms against biotic and abiotic stresses. Transcripts of PA biosynthetic genes accumulated rapidly in response to infection by the fungus Marssonina brunnea f.sp. multigermtubi, treatments of salicylic acid (SA) and wounding, resulting in PA accumulation in poplar leaves. Anthocyanidin reductase (ANR) and leucoanthocyanidin reductase (LAR) are two key enzymes of the PA biosynthesis that produce the main subunits: (+)-catechin and (−)-epicatechin required for formation of PA polymers. In Populus, ANR and LAR are encoded by at least two and three highly related genes, respectively. In this study, we isolated and functionally characterized genes PtrANR1 and PtrLAR1 from P. trichocarpa. Phylogenetic analysis shows that Populus ANR1 and LAR1 occurr in two distinct phylogenetic lineages, but both genes have little difference in their tissue distribution, preferentially expressed in roots. Overexpression of PtrANR1 in poplar resulted in a significant increase in PA levels but no impact on catechin levels. Antisense down-regulation of PtrANR1 showed reduced PA accumulation in transgenic lines, but increased levels of anthocyanin content. Ectopic expression of PtrLAR1 in poplar positively regulated the biosynthesis of PAs, whereas the accumulation of anthocyanin and flavonol was significantly reduced (P<0.05) in all transgenic plants compared to the control plants. These results suggest that both PtrANR1 and PtrLAR1 contribute to PA biosynthesis in Populus.


Introduction
Flavonoids are a large group of plant secondary metabolites that comprise several classes of compounds (e.g. anthocyanins, flavonols, isoflavones and flavan-3-ols) and accumulate in a wide variety of plant tissues [1]. Proanthocyanidins (PAs), also called as condensed tannins, are oligomers or polymers of flavan-3-ols and are among the major flavonoid compounds found in higher plants [2,3]. As a final product of the flavonoid pathway, PAs play an important role in the protection of plants against herbivores and pathogens [1,4,5]. PAs in forage crops can protect ruminants against pasture bloat [6,7,8]. PAs also act as antioxidants with beneficial effects for human health by protection against free radical-mediated injury and cardiovascular disease [9,10]. Additionally, PAs also contribute to the astringency and taste of many fruits and the quality of other plant products, such as wine, tea (Camellia sinensis), and cocoa [11]. Therefore, an understanding of the mechanisms leading to the formation of PA polymers and its regulation is important for regulation of PA biosynthesis in plants.
Flavonoids are synthesized through the phenylpropanoid pathway. PA biosynthesis shares the common avonoid biosynthetic pathway with anthocyanins and avonols. The common pathway has been characterized genetically and biochemically in a number of plant species including Arabidopsis thaliana and Medicago truncatula [12]. At the beginning of the flavonoid biosynthetic pathway, the enzyme chalcone synthase (CHS) carries out the condensation of one molecule of 4-coumaroyl-CoA and three molecules of malonyl-CoA to yield naringenin chalcone. Chalcone is isomerised to a flavanone via the enzyme chalcone flavanone isomerase (CHI). Sharing these central intermediates, the pathway divided into several side branches, each producing a different class of flavonoids. The stereospecific 3ß-hydroxylation of (2S)-flavanones are catalyzed to dihydroflavonols by flavanone 3-hydroxylase (F39H). For the biosynthesis of anthocyanins, dihydroflavonols are catalyzed to flavan-3,4-diols (leucoanthocyanins) by dihydroflavonol reductase (DFR), then leucoanthocyanins are converted to anthocyanidins by anthocyanidin synthase (ANS). The formation of glucosides is catalyzed by UDP glucose-flavonoid 3-o-glucosyl transferase (UFGT), which stabilizes the anthocyanidins by 3-Oglucosylation [13,14].
The biosynthesis of PAs and anthocyanins begins with the same upstream avonoid pathway, leading to the synthesis of avan-3-ol units such as catechin and epicatechin [15,16]. Catechin is traditionally derived from leucocyanidin by the catalyzation of leucoanthocyanidin reductase (LAR) [15], while epicatechins are synthesized from cyanidin by anthocyanidin reductase (ANR), which is encoded by the BANYULS (BAN) gene and has been initially characterized in Arabidopsis [17]. Ectopic expression of BAN in tobacco flower petals and Arabidopsis leaves results in loss of anthocyanins and accumulation of PAs, suggesting that there is an interaction between anthocyanidin and PA pathways [17]. VvANR, encoded by a single gene, has been also isolated and characterized in grapevines (Vitis vinifera) [18]. More recently, three ANR genes were reported in apple (Malus6domestica Borkh.), and introduction of these genes into tobacco inhibits expression of both CHI and DFR genes in flowers, leading to loss of anthocyanin [19]. This finding suggests that the ANR gene may be capable of generating catechin via an alternative route.
LAR activity has been found in several plants and its activity correlated with PA accumulation [2,15,20,21]. The functionality of LAR has been determined in Desmodium uncinatum, in which the recombinant LAR protein catalyzed the conversion of leucocyanidin, leucodelphinidin, or leucopelargonidin to the corresponding 2,3-trans-avan-3-ol (catechin) [5]. This finding clearly established the role of DuLARs in PA biosynthesis. Two grapevine LAR orthologs were also isolated and had different patterns of expression in skin and seeds [18]. In M. truncatula, a single LAR gene has been cloned and characterized and transgenic tobacco plants constitutively overexpressing MtLAR showed reduced anthocyanin content, but no catechin or increased levels of PAs were detected either in leaves or in flowers, suggesting the poor correlation between LAR expression and PA biosynthesis in Medicago [22]. In addition, two LAR genes were reported in Lotus corniculatus, but only LcLAR1 produced active proteins following heterologous expression in Escherichia coli [21]. Therefore, these studies mentioned above suggest a complex genetic control for PA biosynthesis in the same plant species.
Poplar (Populus spp.) is an important forest tree with significant economic and ecological application, but many native poplar species are susceptible to diseases caused by Melampsora medusae, Marssonina brunnea and Septoria musiva. The flavonoid-derived PAs are one class of the major defense phenolics in poplar. Therefore, it is important to develop a model system for studies on the biochemical and genetic mechanisms of PA biosynthesis in tree and woody perennial plants [23]. The transcriptional response of hybrid poplar (P. trichocarpa 6 P. deltoides) to poplar leaf rust (M. medusae) infection showed that these genes for enzymes of PA biosynthesis were strongly induced in the infection process, linking this pathway for the first time to the pathogen defense response in poplar [24]. Similarly, these M. medusae-induced gene family members of PA biosynthesis pathway correspond to those genes most highly induced by wounding in leaves of P. fremontii 6P. angustifolia [25]. With the completion of the P. trichocarpa genome sequence, bioinformatic analysis showed that three putative LAR genes (PtrLAR1, PtrLAR2 and PtrLAR3) were identified in Populus genome [25]. When displayed in a phylogenetic tree, PtrLAR1/ PtrLAR2 and PtrLAR3 proteins occurred in two distinct phylogenetic lineages [25]. We have demonstrated previously that a Populus LAR, PtrLAR3, was expressed in various tissues and the highest level expression in roots. Overexpression of PtrLAR3 in Chinese white poplar (P. tomentosa Carr.) led to a significant increase in PA levels and improved resistance to fungal pathogens in transgenic plants [26]. However, it is not yet clear whether PtrLAR1 is involved in PA biosynthesis in Populus.
In this report, transcription profiles of flavonoid genes of poplar leaves in a compatible interaction with wounding, salicylic acid (SA) and fungal pathogens were analyzed by quantitative RT-PCR.
Furthermore, these genes encoding ANR and LAR isoforms from P. trichocarpa were cloned and characterized. Expression analysis of these genes in the empty-vector control and transgenic lines showing different accumulation of PAs has developed our understanding of the genetic control of PA biosynthesis in poplar. Our findings established the roles of PtrANR and PtrLAR in the PA biosynthesis pathway in poplar and may be of importance in understanding the genetic mechanisms controlling PA accumulation in poplar as well as other plants.

Plant Growth Conditions and Stress Treatments
Populus tomentosa Carr. (clone 73) is native in China. In this study, we collected poplar leaves from the vicinity of Yangling. No specific permission was required for any locations/activities. It was sure that the field studies did not involve endangered or protected species. Poplar plants were grown in the greenhouse at 25uC under a 14-/10-h light/dark cycle with supplemental light (4500 lux). Three-month-old poplar seedlings grown in the greenhouse were subjected to stress experiments analysis of transgenics. Upon harvest of leaves, midveins and necrotic tissue were removed and tissues were frozen in liquid nitrogen and stored at 280uC until analyzed. Leaves within the range of Leaf Plastochron Index (LPI)-8 to 212 [27], which correspond to the youngest fully expand leaves, were used for all stress experiments.
For pathogen inoculations, leaf discs were inoculated with Melampsora brunnea f.sp. multigermtubi (Mb) as previously described [26]. Briefly, leaf discs were inoculated with Mbm urediniospores suspended in 0.01% Tween 20 at a density of 800-1000 spores cm 22 . Control samples were mock-inoculated with 0.01% Tween 20. After inoculation, leaves were kept on wet paper in Petri dishes and incubated in a growth chamber at 18uC with a 16-h photoperiod. Infected and control leaves were harvested at 3 days post inoculation. For salicylic acid (SA) treatment, solution with 5 mM was generously sprayed on leaves until saturation. The Leaves were harvested 24 h after treatments. The wounding treatment consisted of making small holes on leaves with disposable syringe. Control plants were untouched for the wound treatment or sprayed with H 2 O for elicitor treatment. All tested tissues were harvested and frozen in liquid nitrogen until further processing. Biological materials were collected from three replicate trees and processed separately for each treatment.
Plant height from the base of the stem to leaf plastochron index (LPI)-0, of the youngest unfurled leaf, 3 cm in length [27], was recorded, and the petiole of LPI-0 was marked. LPI-1 to LPI-5 of three plants for each genotype were monitored every third day over a 9-day period for the analysis.

Cloning of PtrLAR1 and PtrANR1
Total RNA was isolated from frozen tissues of poplar plants using a RNA RNeasy Plant Mini Kit (Qiagen, Germany) following the manufacturer's instructions. Leaves and petioles were excised from stems, including the fourth (young) and fifth (mature) internodes from the top of the stems. First-strand cDNA was synthesized from 2 mg DNase-treated RNA with RT-AMV transcriptase (TaKaRa, Dalian, China) in a total volume of 20 ml using oligo d(T) at 42uC for 30 min. The full open reading frame of PtrLAR1 was amplified with gene-specific primers (LAR1-F: 59-GCATGACTGTTTCAGCTTCT-39; LAR1-R: 59-TCA-CACCAACAAACCACA GG-39) Joint Genome Institute, http:// genome.jgi-psf.org/poplar/poplar.info.html) by RT-PCR with cDNA from roots. The PCR reaction was carried out with Pfu DNA polymerase (TaKaRa) in a total volume of 50 ml with an initial denaturing step at 94uC for 3 min, 32 cycles of 94uC for 45 s, 56uC for 45 s, and 72uC for 90 s and a final extension step at 72uC for 10 min. The amplification products were cloned into the plant binary vector pCXSN, which is a zero-background TA cloning system that provides simple and high-efficiency direct cloning of PCR-amplified DNA fragments [28]. A 1155-bp PCRfragment of PtrANR1 was amplified from poplar cDNA with gene-specific primers (ANR1-F:59-GCTAGAAAGCTTGCA-TAGCATCCCAGTTGACC-39; ANR1-R: 59-GTAGCTCTA-GATGACGCTAGATTGCTTCAGC-39). Restriction enzyme sites HindIII (in the ANR1-F primer) and XbaI (in the ANR1-R primer) were marked in bold, which were used for subcloning. After verification by DNA sequence analysis in both directions, the fragment digested with HindIII/XbaI was inserted into the corresponding sites of pBI121 (Clontech). The resulting vectors 35S:PtrLAR1 and 35S:PtrANR1, contain the open reading frame down-stream of the cauliflower mosaic virus 35S promoter and the hygromycin phosphotransferase gene (Hpt) as well as neomycin phosphotransferase II (NPTII) as a plant-selectable marker conferring hygromycin and kanamycin resistance, respectively. These plant transformation vectors were transferred into A. tumefaciens EHA105 by the freeze-thaw method [29].

Transformation of P. tomentosa Carr. Plants
Transgenic Chinese white poplar (P. tomentosa Carr.) plants were generated by Agrobacterium-mediated transformation as described previously [30]. Recombinant Agrobacterium was used to infect poplar leaf discs and putative transgenic plants were selected on woody plant medium (WPM) [31] supplemented with 10 mg l -1 hygromycin or 25 mg l 21 kanamycin. Rooted plantlets were acclimatized in pots at 25uC in a 14/10 light/dark cycle and then transferred to the greenhouse for further studies.

Semi-quantitative RT-PCR and Quantitative Real-time PCR Analysis
Total RNA was extracted from leaves, roots, stems, and petioles of poplar plants and treated with DNase I (TaKaRa) according to the manufacturer's instructions. All RNA was purified and firststrand cDNA was synthesized as described above. The reversetranscribed cDNA samples were used for quantitative real-time PCR, which was performed on a TaKaRa real-time-PCR detection system. 18S rRNA was used as an internal control. The RT-PCR conditions were an initial denaturation step at 94uC for 3 min, 28 cycles of 94uC for 30 s, 58uC (PtrLAR1), 55uC (PtrANR1) for 30 s, and 72uC for 1 min, and an extension step at 72uC for 10 min. The amplification products were resolved by 1% (w/v) agarose gel electrophoresis and visualized with ethidium bromide under UV light. Quantitative real-time PCR analysis was performed as described Tsai et al. [25] in a 20 ml reaction volume containing 10 ml of SYBR Green master mix reagent (TaKaRa). The primers were designed using Primer 5.0 software: forward and reverse primers for PtrLAR1 and PtrANR1 amplifications were PtrLAR1-F (59-ACCTACTTGCTATTGCTGCAG-39), PtrLAR1-R(59-CCAAGGTA CGAAAAGCTTCAT-39), PtrANR-F(59-ACAGGGTTTGTGGCATC-39) and PtrANR-R (59-TCTGGGAGCATTGAAGC-39). Each reaction was performed in duplicate and with three biological replicates along with no-template controls. The gene-quantification method was based on the relative expression of the target gene versus the reference gene (18S) [25].

Phylogenetic Analysis
The deduced amino acid sequences were analysed using the program DNAMAN and the software MAGE version 4.0 (Lynnonon Biosoft, Quebec, Canada). Alignment of the deduced amino acid sequences was performed using DNAMAN. The phylogenetic relationships of LARs were analyzed with the neighbour-joining method using MAGE version 4.0.

Extraction and Quantification of Proanthocyanidins, Anthocyanins and Flavonols
For extraction of PAs, tissues were ground in liquid nitrogen and 500 mg of ground tissue was used for extraction in 5 ml extraction buffer (70% [V/V] acetone containing 0.1% [W/V] ascorbic acid) for 24 h at room temperature on a rotating shaker in darkness. The water phase was separated from the acetone phase by adding sodium chlorid etosaturation. After removal of the acetone phase, the water phase was extracted with additional sodium chloride-saturated 100% acetone, and the resulting acetone phase was combined with the first acetone phase. The samples were dried under a stream of nitrogen, the pellet redissolved in 2 ml of methanol acidified with 0.1% ascorbic acid, centrifuged at 16,100 g for 10 min, and the final supernatant kept in darkness and under refrigeration until analysis.
The insoluble PA content was tested using the butanol/HCl method [33]. The residues from the above tissue extractions were dried in air for 2 days, and then 1 ml butanol/HCl reagent was added and the mixture was sonicated at room temperature for 60 min and centrifuged at 2,500 g for 10 min. Supernatants were transferred to cuvettes for determination of absorption at 550 nm and were then boiled for 1 h. After cooling to room temperature, the A 550 was recorded again and the first value subtracted from the second. Absorbance values were converted into PA equivalents using a standard curve (2.5, 5, 10, 20, and 40 mg) of procyanidin B1 (Indofine). Three independent experiments were performed for each sample.
For analysis of anthocyanin levels, 5 ml of methanol:0.1% HCl was added to 0.5 g of ground tissue and sonicated for 1 h, followed by shaking overnight at 120 rpm. After centrifugation at 2,500 g for 10 min, 1 ml of water was added to 1 ml of extract, followed by addition of 1 ml of chloroform to remove chlorophyll. Absorption of the clear supernatant was then measured at 530 nm. Total anthocyanin concentration was calculated using the molar absorbance of cyanidin-3-O-glucoside.
For analysis of flavonols, 100 mg of plant tissue was extracted with 3 ml 80% methanol, sonicated for 1 h, and allowed to stand overnight at 4uC. The extract was centrifuged to remove debris and the supernatant dried under nitrogen. Dried samples were incubated with 3 ml of 1 N HCl at 90uC for 2 h and extracted twice with 3 ml of ethyl acetate. Ethyl acetate extracts were pooled, dried under nitrogen, and resuspended in 200 ml of methanol.
HPLC analysis and quantification of total anthocyanins and PA-derived compounds harvested from transgenic and control plants were performed using a LC-20AD HPLC system (Shimadzu Co., Kyoto, Japan) as described in Deluc et al. [34]. Separation was performed with a linear elution gradient from A (0.2% phosphoric acid in H2O) and B (acetonitrile) over 30 min at a flow rate of 0.5 ml min 21 . For favonols separation, 60% methanol solution was used for isocratic eluting with volume of injecting sample of 10 ml and at a flow-rate of 1 ml min -1 . In each case, the column was maintained at 40uC and the diode array detector was used to record absorption at 280 nm and 368 nm. Catechin, epicatechin, kaempferol and quercetin were used to create standard absorption curves.
Histochemical Staining with DMACA DMACA reacts quite specifically with flavan 3-ol monomers and PAs to form a blue chromophore [35]. Histochemical analysis of PA accumulation in various tissues was detected as described by Li et al. [36]. In brief, plant tissues were decolourized in 5 ml of 30% acetic acid in ethanol for 12 h and washed with 75% ethanol. PAs were detected by staining tissues for 3 h with 1% (w/v) DMACA in ethanol: 6 N HCl (1:1, v/v). Images of stem and petiole sections were recorded using a Nikon microscope. Figures were formatted and assembled with Adobe Photoshop 7.0.

Stress-induced Expression of PA Biosynthetic Genes in Poplar
PA biosynthesis is often induced by various stresses such as mechanical wounding, pathogen infection and insect herbivory in poplar [4,24,37]. In our previous study, transcriptional profiling of PA biosynthetic genes was detected following infection by the fungus Marssonina brunnea f.sp. multigermtubi using digital gene expression (DGE) analysis. The result demonstrated that at least one gene in each enzymic step was upregulated, with the exception of cinnamate 4-hydroxylase (C4H), resulting in PA accumulation in leaves [26]. To investigate association of the Populus flavonoid genes with PA biosynthesis, we carried out gene-specific quantitative real-time polymerase chain reaction (qPCR) analysis using a set of induced flavonoid pathway genes and noninduced genes for comparison. The result showed that transcript abundance of these flavonoid genes essential for PA biosynthesis was induced by fungal infection, treatments of SA and wounding (Fig. 1). The qPCR expression profiles confirmed the trends seen on the DGE analysis for most of PA biosynthesis-related genes.
Isolation and Characterization of PtrLAR1 and PtrANR1 from P. trichocarpa With the completion of the P. trichocarpa genome sequence, all enzymatic steps required for PA biosynthesis annotated in the poplar genome [25,38]. LAR catalyzed early steps in PA biosynthesis are NAPDH-dependent reductases, encoded by three genes in Populus [25]. All three PtrLAR genes are composed of five exons and four introns (Fig. S1A). PtrLAR1/PtrLAR2 and PtrLAR3 proteins occurred in two distinct lineages in a phylogenetic tree [25]. The PtrLAR3 gene was previously isolated and overexpression of PtrLAR3 in Chinese white poplar (P. tomentosa Carr.) led to a significant plant-wide increase in PA levels [26]. In the present study, a full-length cDNA sequence of PtrLAR1 was cloned by reverse transcription (RT)-PCR and contained an open reading frame (ORF) of 1,056 bp encoding a putative protein of 352 amino acids. Sequence comparison of PtrLAR1 showed the highest identity with LAR proteins from Malus domestica (69.83%), Diospyros kaki (66.85%), Lotus corniculatus (62.61%), L. uliginosus (62.61%), Medicago truncatula (57.91%) and Desmodium uncinatum (57.70%) (Fig. S1B). The highly conserved putative motifs (RFLP, ICCN and THD) found in other LAR genes [5,18] was also the deduced PtrLAR1 protein.
Two possible ANR sequences were identified based on the sequences deposited in the Populus genome database (Phytozome version 2.0). The duplicated ANR genes, named PtrANR1 and PtrANR2, have similar exon/intron structures (Fig. S2A) and exhibited similar expression patterns, both with a high transcript level in roots [25]. The cDNA fragment of PtrANR1 with homology to the Arabidopsis ANR (AtANR) was cloned from RNA isolated from P. trichocarpa leaves and contained an ORF of 1,005 bp encoding a protein of 335 amino acids. PtrANR1 has 62.06% identity to ANR from Arabidopsis, 76.18% identity to ANR  Characterization of ANR1 and LAR1 from Populus PLOS ONE | www.plosone.org from D. kaki, 68.14% identity to ANR from L. corniculatus, 77.29% identity to ANR from L. uliginosus, 81.42% identity to ANR, 72.86% identity to ANR from M. truncatula and 80.77% identity to ANR from Vitis vinifera in the coding region (Fig. S2B). Analysis of PtrANR1 indicated that the highly conserved putative NAPDHbinding domain reported in the other ANR genes [21] was also present (Fig. S2B).
A phylogenetic tree was constructed using the predicted amino acid sequences of the putative LAR and ANR proteins from poplar as well as other species. LAR and ANR are related members of the reductase epimerase-dehydrogenase (RED) protein superfamily. As shown in Fig. 2B, PtrLAR1 and PtrLAR2 proteins belonged to one group, while PtrLAR3 was clustered into a distinct branch with TcLAR from Theobroma cacao. This finding is consistent with a recently published phylogenetic analysis [26]. The deduced amino acid sequences of PtrANR1 and ANRs from several plant species were also subjected to phylogenetic analysis. The results demonstrated that PtrANR1 was more closely related to ANR proteins than other RED superfamily proteins (Fig. 2), suggesting that PtrANR1 is more likely to have function of anthocyanidin reductase rather than other RED protein functions.

Expression Profiles of PtrANR1 and PtrLAR1, and PA Accumulation in Poplar Leaves
The expression profiles of PtrANR1 and PtrLAR1 were investigated by semi-quantitative RT-PCR with total RNA from various tissues of poplar. Transcript accumulation was detected in all analyzed tissues, including roots, stems, leaves and petioles but the highest mRNA level was found in roots (Fig. 3A), consistent with PA accumulation in different tissues of P. trichocarpa [26]. Expression patterns of the PtrANR1 and PtrLAR1 genes were further confirmed by quantitative real-time PCR. As shown in Fig. 3B, PtrANR1 transcripts in the roots were approximately 3fold higher than in the stems and petioles, and were more than 15 times as abundant as in the mature leaves. Expression of PtrLAR1 showed a similar pattern to PtrLAR3 as reported previously by Tsai et al. [25], with the highest expression level in roots, with lower expression in other organs (Fig. 3C).
In a previous study, the relatively high levels of PAs were detected in young leaves of poplar compared with mature leaves [26]. Here, we extracted PAs from developing leaves at various stages. The concentration of PAs was the highest in Stage 4 leaves (around 0.17 mg catechin equivalents/g FW) and was lower in Stage 1 and 2 leaves, indicating that PA synthesis increased in the later stages of leaf development (Fig. 4A). Further, expression patterns for the Populus genes encoding LAR and ANR were determined in poplar leaves of different developmental stages. Consistent with their role in PA biosynthesis, PtrLAR1, PtrLAR3 and PtrANR1 were expressed highly in young leaves (LPI-1, -2 and -3) but there was a sharp decrease in their expression in mature leaves (LPI-4 and -5) (Fig. 4B-C).

Functional Analysis of PtrANR1 Gene in Transgenic Poplar Plants
To investigate the function of PtrANR1, the ORF in sense or antisense orientation was introduced into P. tomentosa Carr. plants for ectopic expression under the control of the cauliflower mosaic virus 35S promoter, respectively. More than twenty kanamycinresistant lines for each construct were used to produce plants grown to an average height 1.00 m in the greenhouse. PCR analysis using gene-specific primers showed that an expected amplification product specific for NPTII was obtained from all transgenic lines tested, whereas no signal was detected from wildtype plants (Fig. S3A), confirming the integration of the transgene into the poplar genome. No phenotypic changes were observed in all of the transgenic lines compared with the controls (Data not shown).
To estimate PA localization in PtrANR1-overexpressing and PtrANR1-antisense poplar plants, petiole and stem sections were stained with dimethylaminocinnamldehyde (DMACA), which reacts specifically with PAs and flavan-3-ols to form a blue chromophore [39]. In stem sections, much higher concentration of PAs was observed in the epidermis in the PtrANR1 overexpressor compared with the control and PtrANR1-antisense plants (Fig. 5A-C). In petioles of control plants, DMACA staining showed that PAs were present in the epidermal phloem and xylem cells (Fig. 5D). In PtrANR1-antisense petioles, staining was only detected in phloem and xylem layers, while strong staining was observed in these cells of PtrANR1 overexpressors (Fig. 5E and F).
Spectrophotometric analysis revealed that the insoluble PA levels found in PtrANR1-overexpressors were similar to the levels found in the control plants but relatively higher than that in the PtrANR1-antisense lines (Fig. 5I). All PtrANR1-overexpressing lines except for line 2 showed remarkably increased contents of soluble PAs in their leaves compared with those in the transgenic lines harboring the PtrANR1-antisense construct (Fig. 5H). Moreover, all transgenic 35S:PtrANR1 lines accumulated higher levels of epicatechin than did the control plants, and no difference (Fig. S4). Quantification of anthocyanin content was achieved by spectrophotometry (530 nm) in transgenic poplar. The PtrANR1-overexpressors accumulated much lower levels of anthocyanin when compared with transgenic PtrANR1-antisense lines and the emptyvector control (Fig. 5J). Quantitative RT-PCR analysis confirmed reduced PtrANR1 transcript levels in the antisense lines. These results indicate that constitutive expression of PtrANR1 in poplar

PtrLAR1 Overexpression in Transgenic Poplar Plants
The potential function of PtrLAR1 was investigated in transgenic poplar by expression of a 35S-promoter-driven sense construct. PA accumulation in various tissues of the control and 35S:PtrLAR1 plants were determined. There was a significant increase in the levels of PAs in all transgenic lines compared with the emptyvector control lines (Fig. 6A), consistent with involvement of LAR function in PA biosynthesis. HPLC analysis revealed that transgenic 35S:PtrLAR1 lines accumulated higher levels of catechin and epicatechin, than did the controls (Fig. S5). However, all transgenic lines exhibited a reduction in anthocyanin levels, but flavonols (kaempferol and quercetin) were not significantly changed in the transgenic lines compared with the empty-vector plants (Table 1). Real-time qPCR analysis revealed that overexpression of the PtrLAR1 transgene was confirmed in the transformed lines (Fig. 6B). These findings indicated that PtrLAR1 might play important roles in the PA bisosynthesis in poplar.

Discussion
Numerous factors like insects and diseases affect the health of perennial poplar plants, resulting in reduced growth. Poplar leaves constitutively accumulate PAs, but their biosynthesis is often upregulated by biotic and abiotic stresses such as insect herbivory, mechanical wounding as well as pathogen infection [4,24,26,37]. PA accumulation response to wounding and herbivory occurs both locally at the site of damage and systemically in distal leaves [4], indicating that these compounds function in defence against herbivores and pathogens. In the present study, the strong induction of genes encoding the flavonoid pathway enzymes, including PtrPAL5, PtrCHS3, PtrCHI, PtrDFR1, PtrANR1 and PtrLAR1, was detected in poplar leaves after treatments with wounding, SA and pathogen infection (Fig. 1). The developmentally regulated PA accumulation was found in poplar leaves. Accumulation of PAs began in the youngest leaves (LPI-1) and peaked in mature leaves (LPI-4) (Fig. 4A). This is not consistent with the temporal expression profiles of PtrANR1, PtrLAR1 and PtrLAR3 during leaf development (Fig. 4B-D), indicating that PA content constantly increases during leaf development, but the rate of synthesis decreases in old leaves [18].
PAs are important polyphenolic compounds for plant adaptation to the environment, and there lies a considerable interest in their biosynthesis. The PA biosynthetic pathways share common intermediates until leucocyanidin, which may be used by anthocyanidin reductase (ANR) and leucoanthocyanidin reductase (LAR) to produce epicatechin and catechin (precursors of PAs), respectively. ANR has been initially identified in Arabidopsis, and it is encoded by the BANYULS (BAN) gene [17]. ANR utilizes cyanidin as a substrate, rather than leucocyanidin, which is consistent with the fact that leucoanthocyanidin dioxygenase (LDOX) is essential for PA synthesis in Arabidopsis [40]. Ectopic  expression of BAN in tobacco flower petals and Arabidopsis leaves results in loss of anthocyanins and accumulation of condensed tannins, suggesting that there is an interaction between anthocyanidin and PA pathways [17]. Recently, ectopic expression of apple MdANR genes in tobacco positively and negatively regulates the biosynthesis of proanthocyanidins (PAs) and anthocyanin, respectively, resulting in white, pale pink-coloured, and white/red variegated flowers [19]. In Our study, PtrANR1 transcripts are most highly expressed in roots, which are the sites of maximal accumulation of PAs (Fig. 3). Overexpression of PtrANR1 in poplar resulted in a significant plant-wide increase in PA levels, but it also influences the biosynthesis of anthocyanin by competing with UFGT (UDP-glucose:flavonoid 3-O-glucosyltransferase) activity by which it converts anthocyanidin to anthocyanin (Fig. 5).
Antisense down-regulation of PtrANR1 in poplar reduced the levels of soluble PAs but improved the accumulation of anthocyanin in leaves (Fig. 5). Thus, PtrANR1 plays an important role in both anthocyanin and PA synthesis in poplar. Our results will provide insights into the interaction between PtrANR genes and other genes involved in flavonoid biosynthesis. To date, LAR genes have been isolated from many plant species including tea [41], grape [18,42,43]. The LAR activity of correlated with PA accumulation has been clearly characterized [2,15,20,21,22]. We previously showed that overexpression of PtrLAR3 in Chinese white poplar (P. tomentosa Carr.) resulted in a significant plant-wide increase in PA levels and enhanced fungal resistance in transgenic plants [26]. Here, we have identified PtrLAR1 and its expression is high in roots (Fig. 3C). Ectopic expression of PtrLAR1 in poplar induced qualitative and quantitative changes of the proanthocyanidin profiles (Fig. 6), indicating a good correlation of PtrLAR1 transcript levels with PA accumulation. Based on our studis, both of PtrLAR1 and PtrLAR3 contribute to PA accumulation in Populus. But obviously, expression level of PtrLAR3 in Populus tissues was higher than that of PtrLAR1 [26]. Furthermore, overexpression of PtrLAR3 resulted in more accumulation of PAs in transgenic plants as showed in previous study [26] and Fig. 6. Therefore, these results suggested that PtrLAR1 and PtrLAR3 as LAR homologs in poplar, are required for PAs accumulation, but PtrLAR3 play a more significant role than PtrLAR1 in PA synthesis.
However, PA accumulation was found restricted to some tissues of transgenic plants despite the use of a 35S constitutive promoter. Transgenic tobacco plants overexpressing MtLAR showed reduced anthocyanin content, but no catechin or increased levels of PAs were detected either in leaves or in flowers [22], indicating the poor correlation between MtLAR expression and PA accumulation. In Arabidopsis, PA biosynthetic pathway has been best characterized, but this species accumulates only epicatechin-based starter units and lacks any obvious LAR ortholog [40]. In the present work, overexpression of PtrLAR1 in Arabidopsis did not result in an increase in PA levels, although the transgene transcript accumulated at high levels (Data not shown). Thus, Arabidopsis does not offer a model system for the genetic study of PA biosynthesis in important crops where parallel catechin-and epicatechin-based pathways contribute to PA biosynthesis [5,40,44]. This demonstrates that Populus LARs catalyzed the conversion of leucocyanidin to catechin clearly established its role in PA biosynthesis.
In the current model of PA biosynthesis, LAR and ANR provide two separate pathways for the synthesis of the terminal units for PA polymers [3]. It is the absence of evidence whether ANR and LAR provide alternate pathways to PA biosynthesis in different plant tissues or if both branches can be active in some tissues that accumulate high levels of PA. In a previous study, ectopic expression of apple MdANR genes in tobacco up-regulated the biosynthesis of proanthocyanidins (PAs) but down-regulated the biosynthesis of anthocyanin. Anthocyanin accumulation was significantly reduced in all tobacco transgenic flowers, while contents of both catechin and epicatechin in transgenic flowers were significantly higher than those in flowers of nontransgenic plants [19]. This finding suggests that the ANR gene may be capable of generating catechin via an unknown route. Here, we found that leaves of all transgenic lines overexpressing PtrLAR1 accumulated lower anthocyanin than the control plants (Table 1). Interestingly, following HPLC analysis, it was revealed that transgenic lines accumulated higher levels of catechin and epicatechin when compared with the empty-vector control. Moreover, leaves of all transgenic lines produced lower levels of kaempferol than did the control plants (Table 1). It is necessary to understand how the ANR and LAR pathways co-occur and function to contribute to the formation of PAs in the future.