Arogenate Dehydratase Isoforms Differentially Regulate ... - Cell Press

Molecular Plant Research Article

Arogenate Dehydratase Isoforms Differentially Regulate Anthocyanin Biosynthesis in Arabidopsis thaliana Qingbo Chen1,2, Cong Man1,2, Danning Li1, Huijuan Tan1, Ye Xie1 and Jirong Huang1,* 1

National Key Laboratory of Plant Molecular Genetics, CAS Center for Excellence in Molecular Plant Sciences, Shanghai Institute of Plant Physiology and Ecology, Chinese Academy of Sciences, 300 Fenglin Road, Shanghai 200032, China

2These

authors contributed equally to this article.

*Correspondence: Jirong Huang ([email protected]) http://dx.doi.org/10.1016/j.molp.2016.09.010

ABSTRACT Anthocyanins, a group of L-phenylalanine (Phe)-derived flavonoids, have been demonstrated to play important roles in plant stress resistance and interactions between plants and insects. Although the anthocyanin biosynthetic pathway and its regulatory mechanisms have been extensively studied, it remains unclear whether the level of Phe supply affects anthocyanin biosynthesis. Here, we investigated the roles of arogenate dehydratases (ADTs), the key enzymes that catalyze the conversion of arogenate into Phe, in sucrose-induced anthocyanin biosynthesis in Arabidopsis. Genetic analysis showed that all six ADT isoforms function redundantly in anthocyanin biosynthesis but have differential contributions. ADT2 contributes the most to anthocyanin accumulation, followed by ADT1 and ADT3, and ADT4–ADT6. We found that anthocyanin content is positively correlated with the levels of Phe and sucrose-induced ADT transcripts in seedlings. Consistently, addition of Phe to the medium could dramatically increase anthocyanin content in the wild-type plants and rescue the phenotype of the adt1 adt3 double mutant regarding the anthocyanin accumulation. Moreover, transgenic plants overexpressing ADT4, which appears to be less sensitive to Phe than overexpression of ADT2, hyperaccumulate Phe and produce elevated level of anthocyanins. Taken together, our results suggest that the level of Phe is an important regulatory factor for sustaining anthocyanin biosynthesis. Key words: phenylalanine, metabolic flux, ADT, anthocyanin, Arabidopsis Chen Q., Man C., Li D., Tan H., Xie Y., and Huang J. (2016). Arogenate Dehydratase Isoforms Differentially Regulate Anthocyanin Biosynthesis in Arabidopsis thaliana. Mol. Plant. 9, 1609–1619.

INTRODUCTION Phenylalanine (Phe) is a primary building block not only utilized for biosynthesis of proteins but also for numerous phenylpropanoid compounds (Vogt, 2010; Fraser and Chapple, 2011). In plants, Phe biosynthesis initiates from the conversion of the final product (chorismate) of the shikimate pathway to prephenate by chorismate mutase (Maeda et al., 2011; Yoo et al., 2013). Two branched pathways have been proposed from prephenate to Phe: one is via phenylpyruvate and the other is via arogenate intermediates. In the arogenate pathway, prephenate is converted into arogenate by prephenate aminotransferase (PPA-AT) and then subsequently to Phe by arogenate dehydrotase (ADT). In contrast, prephenate first undergoes decarboxylation/ dehydration by prephenate dehydratase (PDT), followed by transamination to Phe by phenylpyruvate amino transferase. It has been demonstrated that the arogenate pathway is

predominant in Phe biosynthesis in higher plants (Cho et al., 2007; Yamada et al., 2008; Huang et al., 2010; Maeda et al., 2010, 2011). There are six ADT genes in the Arabidopsis genome, designated as ADT1–ADT6. They are ubiquitously expressed in various tissues or organs (Cho et al., 2007). However, different ADTs were reported to make differential contributions to Phe accumulation depending on the tissue. ADT2 was suggested to be a housekeeping gene in leaves and seeds, whereas ADT4 and ADT5 were dominant in stems and roots (Cho et al., 2007; Rippert et al., 2009). All ADTs were reported to be localized in the plastid (Cho et al., 2007; Rippert et al., 2009). Besides ADT activity in vitro, ADT1, ADT2, and ADT6 have weak PDT activity,

Published by the Molecular Plant Shanghai Editorial Office in association with Cell Press, an imprint of Elsevier Inc., on behalf of CSPB and IPPE, SIBS, CAS.

Molecular Plant 9, 1609–1619, December 2016 ª The Author 2016. 1609

Molecular Plant

Phe Levels Influence Anthocyanin Biosynthesis

Figure 1. ADT Isoforms Regulate Anthocyanin Biosynthesis in a Redundant and Differential Manner. (A) Anthocyanin accumulation in 4-day-old seedlings of the representative mutants. Seeds were germinated in 1/2 MS liquid medium without sucrose for 2 days and then transferred to the anthocyanin induction medium containing 3% sucrose for an additional 2 days. Bars, 1 mm. (B–F) Anthocyanin content of WT and various adt mutant seedlings grown under the same condition as in (A). Error bars indicate SD values for the means of three independent replicates. The means that do not share a letter are significantly different (P < 0.05).

but only ADT1 and ADT2 can complement the pha2 yeast mutant (Cho et al., 2007; Bross et al., 2011). Consistent with their expression patterns, genetic and biochemical analyses showed that ADT isoforms are redundant in Phe biosynthesis, whereas ADT4 and ADT5 play a dominant role in lignin biosynthesis (Corea et al., 2012a, 2012b). In addition, ADT isoforms can differentially regulate lignin content and composition (Corea et al., 2012b).

Anthocyanins are a class of water-soluble flavonoid pigments synthesized from Phe in higher plants. They have important biological functions, including defense against UV-B radiation, attracting pollinators and scavenging reactive oxygen species (Ferreyra et al., 2012). The whole anthocyanin biosynthetic pathway can be divided into three parts. The upstream reactions, which are catalyzed by phenylalanine ammonia lyase, cinnamate 4-hydroxylase (C4H), and 4-coumaroyl-CoA

1610 Molecular Plant 9, 1609–1619, December 2016 ª The Author 2016.

Molecular Plant


stress-induced synthesis but not to metabolic redirection (Bonawitz et al., 2014). Thus, additional evidence is required to address whether metabolic flux affects anthocyanin biosynthesis. In this study, we investigated the role of ADTs in anthocyanin biosynthesis using the sucrose-induced anthocyanin biosynthesis system. Our results revealed that the level of Phe is an important regulatory factor for anthocyanin biosynthesis.

RESULTS Identification of adt Mutants

Figure 2. ADT Mutations Affect Cold-Induced Anthocyanin Accumulation. Anthocyanin levels were measured from WT, adt1/3, and adt4/5 seedlings that were grown on 1/2 MS sucrose-free plates for 3 days at 22 C, and then transferred to 1/2 MS plates with 1% sucrose for an additional 3 days at 22 C or 7 C. Data are means ± SD from three independent replicates. The means that do not share a letter are significantly different (P < 0.05).

ligase (4CL), convert Phe into p-coumaroyl CoA. p-Coumaroyl CoA is located at the first crossroad, which can be turned into flavonoids or lignins. The middle stream of the pathway needs three enzymes, including chalcone synthase (CHS), chalcone isomerase (CHI), and flavonone 3-hydroxylase (F3H), to convert p-coumaroyl CoA into dihydroflavonol, which is the common substrate for flavonol, anthocyanin, and proanthocyanidin biosynthesis. The downstream steps are anthocyanin specific and are catalyzed by dihydroflavonol 4-reductase (DFR), anthocyanidin synthase (ANS), and UDP-glucose:flavonoid 3-O-glucosyl-transferase (UFGT). It is known that expression of the anthocyanin-specific genes is regulated by a ternary complex, called MBW (MYB-bHLH-WD40), which is composed of a basic helix-loop-helix (bHLH), R2R3-MYB, and WD40 transcription factors (Gonzalez et al., 2008; Shan et al., 2009). In many plant species, overexpressing an MYB transcription factor, PAP1, activates the expression of the late anthocyanin biosynthetic genes and subsequently promotes anthocyanin accumulation (Borevitz et al., 2000; Mathews, 2003; Li et al., 2010b; Shi and Xie, 2011; Ben Zvi et al., 2012). The activity of the MBW complex is also post-translationally regulated by many cotranscriptional factors, including MYBL2, DELLA, and JAZ, which coordinate plant growth and development with environmental changes (Matsui et al., 2008; Qi et al., 2011; Xie et al., 2016). In addition, anthocyanin biosynthesis was demonstrated to be regulated by metabolic flux (Albert et al., 1997; Devic et al., 1999; Xie et al., 2003; Sun et al., 2012). For example, Zhang et al. (2015) demonstrated that the fruit-specific expression of AtMYB12 in tomato increases the supplies of aromatic amino acids for secondary metabolism via positively regulating the shikimate and phenylalanine biosynthetic pathways; a reduction in lignin biosynthesis leads to anthocyanin accumulation (Li et al., 2010a). Recently, however, hyperaccumulation of flavonoids in lignin biosynthesis-deficient mutants was shown to be attributed to

To elucidate the role of different ADT isoforms in anthocyanin biosynthesis, we isolated five adt single mutants created by T-DNA insertion mutagenesis from the Arabidopsis Biological Resource Center (ABRC) except for adt2 (Supplemental Figure 1). We then generated all double, triple, and quadruple mutants using these single mutants. As previously reported (Corea et al., 2012b), single mutants had no obvious defects in plant growth compared with the wild-type (WT), whereas the adt4/5 double mutant and adt4/5 combined with other mutants including adt1, adt3, adt1/3, adt3/6, and adt1/3/6 displayed a dwarf phenotype (Supplemental Figure 2). These results indicate that ADT isoforms are functionally redundant, and ADT4 and ADT5 play a dominant role in plant growth. In this study, we investigated the effect of ADT isoforms on anthocyanin biosynthesis.

ADT Isoforms Function Redundantly but Differentially to Affect Anthocyanin Accumulation To evaluate the effect of ADTs on anthocyanin synthesis, we established a sucrose-induced anthocyanin synthesis system in which seeds were first germinated in sucrose-free liquid media for 2 days and then treated with 3% sucrose for an additional 2 days as previously described (Sun et al., 2012). Anthocyanins usually accumulate at the abaxial side, edge and petiole of cotyledons, and the top part of hypocotyls (Figure 1A). Quantification of anthocyanin content showed that five adt single mutants accumulated slightly lower levels of anthocyanin than WT but without significant difference (Figure 1B). Double mutants knocking out adt1 or adt3 with another single mutant, except for adt3/6, accumulated a significantly lower level of anthocyanin than WT (Figure 1C, 1D, and 1F). Among other double mutants, adt4/5 and adt4/6 had the same content of anthocyanin with WT, whereas adt5/6 was significantly lower in anthocyanin content than WT (Figure 1D). These results indicate that ADT4–ADT6 act synergistically with ADT1 and ADT3 in anthocyanin biosynthesis. Consistently, in all triple mutants, the mutants with adt1/3 produced the lowest levels of anthocyanin, about 40%–55% of WT, compared with other triple mutants such as adt1/4/5 (Figure 1D and 1E). A similar contribution of ADTs to anthocyanin biosynthesis was observed in quadruple mutants (Figure 1F). Although the quintuple adt1/3/4/5/6 mutant produced the lowest level (about 35% of the WT level) of anthocyanin, no significant difference was detected between the quintuple mutant and quadruple mutants (adt1/3/4/5, adt1/3/4/6, and adt1/3/5/6) (Figure 1F). These results suggest that ADTs function redundantly but differentially to influence sucrose-induced anthocyanin biosynthesis.


Molecular Plant

Phe Levels Influence Anthocyanin Biosynthesis Figure 3. Anthocyanin Profile Is Not Altered in adt Mutants. (A) HPLC analysis of anthocyanin components of WT, adt1/3, and adt4/5 seedlings grown in the sucrose-induced anthocyanin system. The molecules of each labeled component are: a1, cyanidin3-O-[2-O-(xylosyl)-6-O-(4-O-(glucosyl)-p-coumaroyl) glucoside]-5-O-[6-O–(malonyl) glucoside]; a2, cyanidin3-O-[2-O-(2-O-(sinapoyl) xylosyl)-6-O-(4-O-(glucosyl)-p-coumaroyl)glucoside]-5-O-glucoside; a3, cyanidin3-O-[2-O–(2-O– (sinapoyl)xylosyl)-6-O–(4-O–(glucosyl)-p-coumaroyl)glucoside]-5-O-[6-O-(malonyl)glucoside]; a4, cyanidin3-O-[2-O–(2-O–(sinapoyl)xylosyl)-6O–(p-coumaroyl) glucoside]-5-O-glucoside. HPLC chromatogram profiles of anthocyanins were detected at 520 nm. mAU, milliabsorbance unit. (B) Quantification of the four main anthocyanin components based on peak areas. Cyanindin 3-O-glucoside was used as an internal standard. Data are means ± SD from three independent replicates. The means that do not share a letter are significantly different (P < 0.01).

induced anthocyanin biosynthesis. Taken together, our data indicate that all ADTs are functionally redundant, but ADT1 and ADT3 play more important roles in sucroseand cold-induced anthocyanin synthesis.

Anthocyanin Profile Is Not Affected by ADT Mutations

Recently, anthocyanin biosynthesis has been reported to be directly related to plant resistance to cold stress (Wang et al., 2013). We tested whether ADT mutations led to a decrease in anthocyanin accumulation under cold conditions. Our results showed that WT, adt4/5, and adt1/3 plants accumulated similar levels of anthocyanin at 22 C, whereas the anthocyanin level is much lower in adt1/3 than that in WT and adt4/5 at 7 C, although the cold treatment significantly promoted anthocyanin biosynthesis in all genotypes (Figure 2). These results indicate that ADT1 and ADT3 are also involved in cold-

Besides anthocyanin content, we investigated the effect of ADT mutations on anthocyanin components using high-performance liquid chromatography (HPLC). HPLC analysis showed that all the anthocyanin peaks detected in adt1/3 and adt4/5 double mutants were detected in WT (Figure 3A). The chemical identity of the main a1, a2, a3, and a4 peaks has been reported in our previous paper (Sun et al., 2012). Quantification of each peak showed that the levels of all components were dramatically reduced in adt1/3 compared with WT, whereas no difference was detected between WT and adt4/5 (Figure 3B). In adt1/3, the level of peak a3 was decreased most significantly among all components (Figure 3B). These results indicate that mutations in ADT genes have no effect on the anthocyanin profile, but reduce their levels.

ADT2 Plays the Most Important Role in Anthocyanin Biosynthesis To investigate the role of ADT2 in anthocyanin biosynthesis, we generated transgenic plants, designated adt2-amiR, in which expression of ADT2 was significantly downregulated by artificial microRNA interference. Quantitative PCR analysis showed that,


Molecular Plant


Figure 4. ADT2 Plays a Dominant Role in Anthocyanin Biosynthesis. (A) qPCR analysis of transcriptional levels of ADT1–ADT6 in WT and two adt2-amiR transgenic lines. Transcript levels were normalized to ACTIN2, and then expression of each gene in WT was set to 1. Data are means ± SD from three independent replicates. *P < 0.05 and **P < 0.01 (Student t-test). (B) Anthocyanin content of WT and two adt2-amiR transgenic plants. Seedlings were grown in the sucrose-induced anthocyanin system. Data are means ± SD from three independent replicates. The means that do not share a letter are significantly different (P < 0.01).

in the two adt2-amiR lines, the level of ADT2 mRNA was about 10% of the WT level (Figure 4A), whereas expression of other ADTs was not dramatically altered despite statistically significant changes detected in some ADTs. Similar to other single mutants, adt2-amiR plants had no obvious phenotypes in growth compared with WT. However, the two adt2-amiR lines produced about 35% of WT anthocyanin content in the sucroseinduced system (Figure 4B). This result may explain why the adt1/ 3/4/5/6 quintuple mutant still produces about 30% of WT anthocyanin content (Figure 1E). Thus, our data suggest that ADT2 contributes most to anthocyanin synthesis among all ADT isoforms.

Expression of ADT1–ADT3 Is Significantly Induced by Sucrose To understand the mechanism by which ADT isoforms had a differential and redundant effect on sucrose-induced anthocyanin biosynthesis, we investigated tissue-specific expression patterns of all ADT genes. As previously reported (Cho et al., 2007; Rippert et al., 2009), we observed that all six ADTs were expressed ubiquitously in various tissues with differential intensity (Supplemental Figure 3), supporting that ADTs have redundant and differential roles in plant growth and development. We also found that ADT4 and ADT5 transcripts were dominant in roots, seedlings, and stems, whereas all ADTs were expressed in a relatively equal manner in leaves. These tissue-specific profiles of ADT expression were generally consistent with the results of our histochemical analysis of b-glucuronidase (GUS) activity in transgenic plants harboring the reporter GUS gene driven by the promoter of ADTs (Supplemental Figure 4). Thus, expression patterns of ADTs are not in agreement with their individual contributions to sucrose-induced anthocyanin accumulation. This prompted us to test whether expression of ADTs is differentially induced by sucrose. qPCR analysis showed that sucrose suppressed the transcriptional level of ADT4 but induced expression of other ADTs to different degrees (Figure 5). Expression of ADT1– ADT3 was induced more rapidly and more highly than that of ADT5 and ADT6, and the highest induction of ADT expression occurred 10 h and/or 16 h after sucrose treatment. In agreement with the role of ADTs in anthocyanin biosynthesis, ADT2 had the highest expression level among all ADTs, followed by ADT1 and ADT3 (Figure 5). These results indicate that anthocyanin content is positively correlated with the sucrose-induced expression levels of ADTs.

Reduced Anthocyanin Accumulation in adt1/3 Is Irrelevant to the Expression Levels of Anthocyanin Biosynthetic Genes It was also possible that the observed difference in anthocyanin content between adt1/3 and WT was associated with the expression level of anthocyanin biosynthetic and regulatory genes. To verify this hypothesis, we examined transcriptional levels of several key genes involved in anthocyanin biosynthesis in WT, adt1/3, and adt4/5 using qPCR. These key genes include biosynthetic genes, such as CHALCONE SYNTHASE (CHS), DIHYDROFLAVONOL REDUCTASE (DFR), and UDP-GLUCOSE: FLAVONOID 3-O-GLUCOSYLTRANSFERASE (UF3GT), and regulatory genes, such as PRODUCTION OF ANTHOCYANIN PIGMENTATION 1 (PAP1), TRANSPARENT TESTA 8 (TT8), and TRANSPARENT TESTA GLABRA 1 (TTG1). Our results showed that mRNA levels of all genes examined were coordinately upregulated in a time course by sucrose treatment in WT, adt1/3, and adt4/5 seedlings (Figure 6). Gene expression reached a peak 16 h after sucrose induction and then maintained a relatively higher level to 24 h. In general, we did not observe a significant difference in expression of the genes examined between WT and adt1/3. Interestingly, levels of PAP1, TT8, and DFR transcripts were much higher in adt4/5 than in adt1/3 and WT after 10 h of sucrose treatment, whereas no dramatic changes in CHS, UF3GT, and TTG1 expression were detected. These results imply that the reduced anthocyanin level in adt1/3 is unlikely caused by the altered expression of genes involved in the anthocyanin biosynthetic pathway.

Phe Feeding Rescues the Reduced Anthocyanin Accumulation in adt1/3 Based on the above data, we proposed that the reduced level of Phe in the adt mutants affects anthocyanin biosynthesis. To test his hypothesis, we first measured the level of Phe in WT, adt1/3, and adt4/5 seedlings during sucrose-induced anthocyanin synthesis. As shown in Figure 7A, Phe content was significantly lower in adt1/3 than in WT, but similar between WT and adt4/5. Then, we tested whether Phe feeding was able to rescue the phenotype of adt1/3 with low anthocyanin content. Our results showed that addition of 0.4 mM Phe to the culture media markedly promoted anthocyanin accumulation in WT, adt4/5, and adt1/3 seedlings (Figure 7B), indicating that Phe content is a critical factor for anthocyanin biosynthesis in vivo. Importantly, adt1/3 produced almost the same level of anthocyanin as WT and adt4/5 in the presence of 0.4 mM Phe (Figure 7B). Taken


Molecular Plant

Phe Levels Influence Anthocyanin Biosynthesis Figure 5. Expression of ADT1–ADT6 Genes Are Differentially Induced by Sucrose. Gene expression was analyzed by qPCR using total RNA extracted from WT seedlings grown in sucrose-free 1/2 MS solution for 2 days and then treated with 3% sucrose for indicated times. Transcript levels were normalized to ACTIN2, and then the value for the transcript of each gene at 0 h was set to 1. Data are means ± SD from three independent replicates.

together, these results suggest that the level of Phe in adt mutants is positively correlated with anthocyanin accumulation.

Ectopic Expression ADT4 or ADT5 Lead to Overaccumulation of Anthocyanin To assess whether overexpressing ADTs could promote sucroseinduced anthocyanin biosynthesis, we made transgenic plants overexpressing ADTs. Interestingly, we observed that overexpressing ADT4 or ADT5 but not other ADTs resulted in severe phenotypes. Figure 8A shows the representative phenotypes of transgenic plants overexpressing ADT2, ADT4, and ADT5. The leaves of ADT4/ADT5 overexpressing plants were yellow/white, narrow, small, and upwardly curled (Figure 8A). Some ADT4/ ADT5 overexpression lines were dwarf and sterile. These phenotypes are the same as those observed in the adt2-1D mutant (Huang et al., 2010). Thus, we suggest that ADT4 and ADT5 may not be allosterically regulated by the product Phe. qRT–PCR analysis showed that mRNA levels of ADTs were dramatically increased in transgenic plants (Figure 8B). ADT4 or ADT5 overexpression lines produced 60%–120% higher levels of anthocyanin than WT, whereas ADT1, ADT2, ADT3, or ADT6 overexpression lines had no significant effect on anthocyanin accumulation (Figure 8C). These results indicate that Phe generated by ADT is a rate-limiting step in anthocyanin biosynthesis. To test our hypothesis that ADT4/5 but not other ADTs are insensitive to Phe, we purified recombinant ADT2 and ADT4 from Echerichia coli and analyzed their catalytic activity in vitro in the presence or absence of 100 mM Phe. Our results showed that Phe did not inhibit ADT4 activity but dramatically inhibited ADT2 activity (Figure 8D and 8E). These data indeed revealed that ADT4, different from ADT2, is not feedback regulated by Phe.

DISCUSSION Expression Patterns of ADT Genes Are Associated with Their Biological Functions It has been reported that ADT isoforms are expressed ubiquitously but also in a tissue-specific manner, which defines a role of ADT isoforms in plant growth and development

(Corea et al., 2012a, 2012b). For example, ADT4 and ADT5 play a dominant role in lignin biosynthesis and components that affect plant height and prostrate phenotypes. Here, we found that ADT1–ADT3 play a more important role in anthocyanin biosynthesis than other ADTs in sucroseand cold-induced anthocyanin biosynthesis, although all ADT isoforms are functionally redundant. Besides their tissue-specific expression patterns, ADT genes respond differentially to various stimuli. We discovered that sucrose dramatically induced ADT1–ADT3 transcripts and slightly induced ADT5 and ADT6 expression, but significantly suppressed ADT4 expression. This character of ADT gene expression is in accord with the role of individual ADT isoforms in sucroseinduced anthocyanin biosynthesis during the seedling growth stage. However, it is difficult to finely evaluate the contribution of each ADT isoform to the Phe pool because expression of certain ADT genes is significantly altered in some adt mutants (Corea et al., 2012a). In the future, it will be important to investigate the molecular mechanisms by which expression of ADT genes is differentially regulated by various stimuli. Generally, the level of Phe in plants, similar to microorganisms, is primarily controlled by an allosteric feedback mechanism through the branch point enzyme ADT (Tzin and Galili, 2010). It was reported that ADT activity was positively regulated by Tyr and negatively by Phe in tobacco, spinach, and sorghum (Jung et al., 1986; Siehl and Conn, 1988). Genetic screening for resistant mutants to analogs of Trp (5-methyltryptophan, 5MT) in rice and to phenylalanine (m-tyrosine) in Arabidopsis identified the same mutation that leads to the overaccumulation of Phe due to a reduced feedback sensitivity of the ADT protein to Phe (Yamada et al., 2008; Huang et al., 2010). This point mutation is located in the conserved ESRP motif of the ACT regulatory domain. Overexpressing Phe-insensitive ADT2 (adt21D) but not the WT ADT2 led to various severe phenotypes including changed rosette leaf morphology and dwarf plants without setting seeds (Huang et al., 2010). In this study, we found that ADT4 or ADT5 overexpression lines displayed the same morphological phenotype as observed in the adt2-1D mutant, which accumulates up to 160-fold higher free Phe in rosette leaves (Huang et al., 2010). In addition, ectopic expression of ADT4 or ADT5 led to a significant increase in anthocyanin content, compared with WT. We therefore infer that hyperaccumulation of anthocyanin would be observed in the adt2-1D mutant. However, other ADT overexpression lines had the same phenotype as WT plants. These data indicate


Molecular Plant


Figure 6. Reduced Anthocyanin Accumulation in adt1/3 Is Not due to Low Expression Levels of Anthocyanin Biosynthetic Genes. Gene expression was analyzed by qPCR using total RNAs extracted from WT, adt1/3, and adt4/5 seedlings that were grown in sucrose-free 1/2 MS solution for 2 days and subsequently treated with 3% sucrose for indicated times. Transcript levels were normalized to ACTIN2, and then the value of the transcript of each gene at 0 h was set to 1. Data are means ± SD from three independent replicates.

that activity of ADT4 and ADT5 is not feedback inhibited by Phe. Indeed, our in vitro enzymatic activity assays revealed that ADT4 but not ADT2 was insensitive to Phe. Since the ESRP motif is conserved in all ADTs, we suggest that the ESRP motif is not

sufficient for the allosteric inhibition of Phe biosynthesis. We analyzed the ACT domain between ADT4 or ADT5 and other ADT proteins, and found that some specific amino acids are present only in ADT4 and ADT5. It is possible that, besides the


Molecular Plant

Figure 7. Anthocyanin Accumulation Is Regulated by Phe Levels In Vivo. (A) UPLC analysis of Phe levels in WT, adt1/3, and adt4/5 seedlings grown in the sucrose-induced anthocyanin system. The seedlings were sampled at 24 h after addition of 3% sucrose. DW, dry weight. The means labeled with an asterisk are significantly different between WT and mutant at P < 0.05. (B) Anthocyanin content of WT, adt1/3, and adt4/5 seedlings. Seeds were germinated in 1/2 MS liquid medium without sucrose for 2 days and then grown in anthocyanin induction medium with or without 0.4 mM Phe for an additional 2 days. Data are means ± SD from three independent replicates. The means labeled with double asterisks are significantly different between WT and mutants at P < 0.01.

ESRP motif, other motifs in the ACT domain are also required for feedback regulation. In the future, it will be interesting to investigate the structural basis determining the sensitivity of ADTs to Phe.

Metabolic Flux Regulates Anthocyanin Biosynthesis To date, the effect of metabolic flux on anthocyanin biosynthesis has been concentrated on the metabolic interaction between anthocyanin and other phenylpropanoid compounds. For example, anthocyanin and flavonol biosynthesis was regulated in an opposite way by miR156 via its targeted SPL9 gene, whose product directly binds to MYB transcription factors and inhibits the formation of the MYB-bHLH-WD40 transcriptional activation complex in Arabidopsis (Gou et al., 2011). In strawberry fruits, profiling analysis combined transcriptomes with metabolites and showed that downregulation of CHS expression and concomitant induction of a peroxidase FaPRX27 gene led to phenylpropanoid flux from anthocyanin to lignin (Ring et al., 2013). These data highlight that the branched pathways of phenylpropanoid biosynthesis compete for their common precursors. Direct evidence supporting that anthocyanin biosynthesis is influenced by metabolic flux is derived from Phe-feeding experiments. Addition of Phe to the media significantly promoted

Phe Levels Influence Anthocyanin Biosynthesis sucrose-induced anthocyanin biosynthesis of WT and adt4/5 seedlings and completely rescued anthocyanin biosynthesis of the adt1/3 double mutant. We also provided indirect evidence that transgenic plants overexpressing ADT4 or ADT5 produced higher levels of Phe and anthocyanin than WT (Figure 8). These data indicate that Phe content is a rate-limiting factor to maximize anthocyanin biosynthesis in sucrose-induced anthocyanin biosynthesis. Consistently, a decrease in the Phe level was reported to be correlated with reduction in the lignin content (Corea et al., 2012a, 2012b). The same result was observed in petunia petals where ADT activity is a rate-limiting factor for Phe biosynthesis via the arogenate pathway (Maeda et al., 2010, 2011). Furthermore, we did not observe that expression levels of anthocyanin biosynthetic and regulatory genes were positively correlated with anthocyanin accumulation among WT, adt4/5, and adt1/3 in our system, indicating that the capacity to convert Phe into anthocyanin is not limited. It is an interesting phenomenon that expression levels of DFR, TT8, and PAP1 genes were higher in adt4/5 than in WT (Figure 6), but no difference in anthocyanin accumulation was found between WT and adt4/5 in the Phefeeding experiment (Figure 7). We therefore suggest that the effect of adt4/5 mutations on anthocyanin accumulation is not due to compensatory upregulation of anthocyanin pathway genes. Taken together, our data reveal that anthocyanin biosynthesis is also regulated by the level of Phe. This may provide a new target for metabolic engineering to modulate anthocyanin content in plants.

METHODS Plant Materials and Growth Conditions The Arabidopsis thaliana ecotype Columbia-0 was used as WT. All T-DNA insertion mutants except for adt2 were identified from the ABRC seed stocks using gene-specific primers (Supplemental Table 1). The adt mutants were named as previously reported (Corea et al., 2012b). adt12, adt3-1, adt4-1, adt5-2, and adt6-2 were used in the experiments and to make double, triple, quadruple, and quintuple mutants by genetic crosses. Seeds were sterilized and treated at 4 C for 2 days, then germinated in half-strength Murashige and Skoog medium (MS salts [Sigma], 1% sucrose, and 0.7% Phyto agar [Duchefa Biochenie]) under light conditions of 90–100 mmol m 2 s 1 with 16 h light/8 h dark cycles at 22 C. Five days after germination, seedlings were transplanted into soil in a growth chamber with 16 h light/8 h dark cycles at 22 C. To study the effect of ADT mutations on anthocyanin biosynthesis and Phe content, seeds were germinated in 1/2 MS liquid medium without sucrose for 2 days, and then transferred to the anthocyanin induction liquid medium, which contained 1/2 MS liquid and 3% sucrose, for another 2 days under constant light conditions. For cold treatment, 3-day-old seedlings grown in 1/2 MS medium without sucrose were transferred to the medium containing 1% sucrose with or without cold (7 C) treatment for an additional 3 days under constant light conditions.

Plasmid Construction and Transformation Full-length cDNAs of ADT1 (U16103), ADT3 (U82642), ADT4 (U16705), and ADT6 (U11191) were obtained from the Salk Institute Genomic Analysis Laboratory (Yamada et al., 2003). Full-length cDNAs of ADT2 and ADT5 were cloned from a cDNA library. The various plasmids were constructed using the Gateway cloning system (Invitrogen). ADT2 artificial microRNAs were constructed using the plasmid pRS300 as reported by Schwab et al. (2010). After insertion into pENTR/SD/D-TOPO (Invitrogen), the construct was then recombined into pGWB destination binary vectors for making transgenic plants (Research Institute of Molecular Genetics,


Molecular Plant


Figure 8. Effect of ADT Overexpression on Anthocyanin Biosynthesis and Phe Feedback Regulation of ADT2 and ADT4 Activity. (A) Phenotypes of 35-day-old WT and transgenic plants overexpressing the ADT2, ADT4, or ADT5 gene. Bars, 1 cm. (B) qPCR analysis of the level of ADTs expression in WT and transgenic lines. Total RNAs were extracted from the seedlings grown in the sucrose-induced anthocyanin synthesis system. Expression levels were normalized to ACTIN2, and then the value for the transcript of each gene in WT was set to 1. Data are means ± SD from three independent replicates. (C) Anthocyanin content of WT and overexpression lines. Anthocyanin was extracted from the same seedlings as in (B). Data are means ± SD from three independent replicates. The means that do not share a letter are significantly different (P < 0.05). (D and E) In vitro assays of ADT2 and ADT4 activity in the absence or presence of 100 mM Phe. Data are means ± SD from three independent replicates. Shimane University, Japan). For analysis of ADT promoter activity, GUS was cleaved from pBI101 and ligated into pCAMBIA1300, and ADT promoter fragments up to 1000 bp from the start codon were cloned into

pCAMBIA1300. Primers used for plasmid construction are listed in Supplemental Table 2. Transgenic plants were made using the floral dip method (Clough and Bent, 1998).


Molecular Plant


Isolation of RNA and Transcription Analysis

FUNDING

Total RNA was extracted using RNAgents Total RNA Isolation System (Promega) and treated with DNase using a DNA-free Kit (Applied Biosystems). Then, 1 mg of total DNase-treated RNA was reverse transcribed with AMV reverse transcriptase (Promega). Gene expression was determined using gene-specific primers by RT–PCR or quantitative real-time PCR. The amplified DNA fragments were separated by 1.2% agarose gel and analyzed using GIS-2500 (Tanon). Primers used for RT–PCR and qRT–PCR are listed in Supplemental Table 2.

This work was supported by grants from the 973 Program (2013CB127000), the National Science Fund (31370326), and Special Fund for strategic pilot technology Chinese Academy of Sciences (XDA08020109).

AUTHOR CONTRIBUTIONS Q.-B.C., C.M., and J.-R.H. designed the experiments; Q.-B.C., C.M., D.-N.L., H.-J.T., and Y.X. performed the experiments and data analysis; and Q.-B.C. and J.-R.H. wrote the manuscript.

GUS Staining and Confocal Microscopy GUS staining was performed as described by Sun et al. (2012). All fluorescent images were recorded with a confocal laser-scanning microscope (Olympus, FV1000).

Analyses of Anthocyanin and Phe Content Anthocyanin was extracted by gently shaking samples in 600 mL of a solution with 1% HCl in methanol (v/v) overnight in the dark at 4 C. After extraction, 400 mL of water and 400 mL of chloroform were added and mixed. After centrifugation at 12 000 rpm for 2 min, the absorbance (A) of the supernatant was measured at 530 and 657 nm, and the concentration of anthocyanin was calculated using (A530–0.25A657)/W (Rabino and Mancinelli, 1986). For HPLC analysis of anthocyanin, anthocyanin was extracted using the solvent MeOH:H2O:CH3COOM (9:10:1), which contains 0.0025 mg/mL cyanidin 3-O-glucoside (C3G) as a standard for anthocyanins at 520 nm, and the level was determined as described (Sun et al., 2012). The samples used for extraction of total amino acids were the same as those for the anthocyanin analysis. Amino acids were extracted and separated as described by Salazar et al. (2012). Five milligrams of the dried Arabidopsis leaves were ground with 125 mL of 50% (v/v) methanol:water solution for 60 s, incubated on dry ice for 5 min, and sonicated in a water bath for 1 min. The liquid was collected after centrifugation at 13 000 rpm, 4 C for 8 min. The precipitate was used for amino acid extraction again with the same procedure, and the extract was mixed with the previous one. Amino acid derivatization was carried out using the AccQ,Tag Ultra derivatization kit (Waters Corp.) according to the manufacturer’s instructions. Derivatized amino acids were separated on a Waters AccQ,Tag Ultra column and detected by the ACQUITY UPLC Fluorescence Detector. The area of the Phe peak was used for quantification via the standard curve made from reagent grade Phe (Sigma-Aldrich).

Recombinant Protein Purification and Enzyme Assays ADT2 and ADT4 were cloned to the vector PET28b, transformed to Escherichia coli BL21, and purified by Ni-NTA agarose (QIAGEN). Arogenate, the substrate of ADTs, was isolated from a Neurosporacrassa mutant (ATCC 36373) as described as Yamada et al. (2008) without HPLC purification (Huang et al., 2010). The level of arogenate in the purified product was determined by the conversion of arogenate to Phe in the presence of 1 N HCl. The activity of ADT2 and ADT4 was measured by adding 150 mL of purified arogenate (100 mM) and 15 mg of protein to 200 mL reaction buffer (100 mM KH2PO4/K2HPO4 [pH 7.0]). For allosteric inhibition analysis, 100 mM Phe was incubated with ADT2 and ADT4 protein for 10 min at 37 C before adding arogenate. The reaction was stopped by adding the same volume of methanol, and proteins were removed by centrifugation at 12 000 rpm for 5 min. Twenty microliters of the product was injected into a Luna C18 column (250 3 4.6 mm, Phenomenex). The elution speed was set at 0.6 mL/min with 1 mM potassium phosphate (pH 7.2):methanol (9:1, v/v), and Phe was detected at 210 nm.

SUPPLEMENTAL INFORMATION Supplemental Information is available at Molecular Plant Online.

ACKNOWLEDGMENT We thank Dr. Wenli Hu at the Shanghai Institute of Plant Physiology and Ecology for skilled technical assistance. No conflict of interest declared. Received: June 15, 2016 Revised: September 24, 2016 Accepted: September 26, 2016 Published: October 5, 2016

REFERENCES Albert, S., Delseny, M., and Devic, M. (1997). BANYULS, a novel negative regulator of flavonoid biosynthesis in the Arabidopsis seed coat. Plant J. 11:289–299. Ben Zvi, M.M., Shklarman, E., Masci, T., Kalev, H., Debener, T., Shafir, S., Ovadis, M., and Vainstein, A. (2012). PAP1 transcription factor enhances production of phenylpropanoid and terpenoid scent compounds in rose flowers. New Phytol. 195:335–345. Bonawitz, N.D., Kim, J.I., Tobimatsu, Y., Ciesielski, P.N., Anderson, N.A., Ximenes, E., Maeda, J., Ralph, J., Donohoe, B.S., Ladisch, M., et al. (2014). Disruption of Mediator rescues the stunted growth of a lignin-deficient Arabidopsis mutant. Nature 509:376–380. Borevitz, J.O., Xia, Y., Blount, J., Dixon, R.A., and Lamb, C. (2000). Activation tagging identifies a conserved MYB regulator of phenylpropanoid biosynthesis. Plant Cell 12:2383–2393. Bross, C.D., Corea, O.R., Kaldis, A., Menassa, R., Bernards, M.A., and Kohalmi, S.E. (2011). Complementation of the pha2 yeast mutant suggests functional differences for arogenate dehydratases from Arabidopsis thaliana. Plant Physiol. Biochem. 49:882–890. Cho, M.H., Corea, O.R., Yang, H., Bedgar, D.L., Laskar, D.D., Anterola, A.M., Moog-Anterola, F.A., Hood, R.L., Kohalmi, S.E., Bernards, M.A., et al. (2007). Phenylalanine biosynthesis in Arabidopsis thaliana. Identification and characterization of arogenate dehydratases. J. Biol. Chem. 282:30827–30835. Clough, S.J., and Bent, A.F. (1998). Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J. 16:735–743. Corea, O.R., Bedgar, D.L., Davin, L.B., and Lewis, N.G. (2012a). The arogenate dehydratase gene family: towards understanding differential regulation of carbon flux through phenylalanine into primary versus secondary metabolic pathways. Phytochemistry 82:22–37. Corea, O.R., Ki, C., Cardenas, C.L., Kim, S.J., Brewer, S.E., Patten, A.M., Davin, L.B., and Lewis, N.G. (2012b). Arogenate dehydratase isoenzymes profoundly and differentially modulate carbon flux into lignins. J. Biol. Chem. 287:11446–11459. Devic, M., Guilleminot, J., Debeaujon, I., Bechtold, N., Bensaude, E., Koornneef, M., Pelletier, G., and Delseny, M. (1999). The BANYULS gene encodes a DFR-like protein and is a marker of early seed coat development. Plant J. 19:387–398. Ferreyra, M.L., Rius, S.P., and Casati, P. (2012). Flavonoids: biosynthesis, biological functions, and biotechnological applications. Front. Plant Sci. 3:1–15.


Molecular Plant

Phe Levels Influence Anthocyanin Biosynthesis Fraser, C.M., and Chapple, C. (2011). The phenylpropanoid pathway in Arabidopsis. Arabidopsis Book 9:e0152.

derivatization for targeted amino acid analysis: application to screening of Arabidopsis thaliana mutants. Metabolites 2:398–428.

Gonzalez, A., Zhao, M., Leavitt, J.M., and Lloyd, A.M. (2008). Regulation of the anthocyanin biosynthetic pathway by the TTG1/ bHLH/Myb transcriptional complex in Arabidopsis seedlings. Plant J. 53:814–827.

Schwab, R., Ossowski, S., Warthmann, N., and Weigel, D. (2010). Directed gene silencing with artificial microRNAs. Methods Mol. Biol. 592:71–88.

Gou, J.Y., Felippes, F.F., Liu, C.J., Weigel, D., and Wang, J.W. (2011). Negative regulation of anthocyanin biosynthesis in Arabidopsis by a miR156-targeted SPL transcription factor. Plant Cell 23:1512–1522. Huang, T., Tohge, T., Lytovchenko, A., Fernie, A.R., and Jander, G. (2010). Pleiotropic physiological consequences of feedbackinsensitive phenylalanine biosynthesis in Arabidopsis thaliana. Plant J. 63:823–835. Jung, E., Zamir, L.O., and Jensen, R.A. (1986). Chloroplasts of higher plants synthesize L-phenylalanine via L-arogenate. Proc. Natl. Acad. Sci. USA 83:7231–7235. Li, X., Bonawitz, N.D., Weng, J.K., and Chapple, C. (2010a). The growth reduction associated with repressed lignin biosynthesis in Arabidopsis thaliana is independent of flavonoids. Plant Cell 22:1620–1632. Li, X., Gao, M.J., Pan, H.Y., Cui, D.J., and Gruber, M.Y. (2010b). Purple canola: Arabidopsis PAP1 increases antioxidants and phenolics in Brassica napus leaves. J. Agric. Food Chem. 58:1639–1645. Maeda, H., Shasany, A.K., Schnepp, J., Orlova, I., Taguchi, G., Cooper, B.R., Rhodes, D., Pichersky, E., and Dudareva, N. (2010). RNAi suppression of Arogenate Dehydratase1 reveals that phenylalanine is synthesized predominantly via the arogenate pathway in petunia petals. Plant Cell 22:832–849. Maeda, H., Yoo, H., and Dudareva, N. (2011). Prephenate aminotransferase directs plant phenylalanine biosynthesis via arogenate. Nat. Chem. Biol. 7:19–21. Mathews, H. (2003). Activation tagging in tomato identifies a transcriptional regulator of anthocyanin biosynthesis, modification, and transport. Plant Cell 15:1689–1703. Matsui, K., Umemura, Y., and Ohme-Takagi, M. (2008). AtMYBL2, a protein with a single MYB domain, acts as a negative regulator of anthocyanin biosynthesis in Arabidopsis. Plant J. 55:954–967. Qi, T., Song, S., Ren, Q., Wu, D., Huang, H., Chen, Y., Fan, M., Peng, W., Ren, C., and Xie, D. (2011). The Jasmonate-ZIM-domain proteins interact with the WD-Repeat/bHLH/MYB complexes to regulate jasmonate-mediated anthocyanin accumulation and trichome initiation in Arabidopsis thaliana. Plant Cell 23:1795–1814. Rabino, I., and Mancinelli, A.L. (1986). Light, temperature, and anthocyanin production. Plant Physiol. 81:922–924. € cherig, S., Hoffmann, T., Blanco-Portales, R., Ring, L., Yeh, S.Y., Hu Fouche, M., Villatoro, C., Denoyes, B., Monfort, A., Caballero, J.L., et al. (2013). Metabolic interaction between anthocyanin and lignin biosynthesis is associated with peroxidase FaPRX27 in strawberry fruit. Plant Physiol. 163:43–60. Rippert, P., Puyaubert, J., Grisollet, D., Derrier, L., and Matringe, M. (2009). Tyrosine and phenylalanine are synthesized within the plastids in Arabidopsis. Plant Physiol. 149:1251–1260. Salazar, C., Armenta, J.M., and Shulaev, V. (2012). An UPLC-ESI-MS/ MS assay using 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate

Shan, X., Zhang, Y., Peng, W., Wang, Z., and Xie, D. (2009). Molecular mechanism for jasmonate-induction of anthocyanin accumulation in Arabidopsis. J. Exp. Bot. 60:3849–3860. Shi, M.Z., and Xie, D.Y. (2011). Engineering of red cells of Arabidopsis thaliana and comparative genome-wide gene expression analysis of red cells versus wild-type cells. Planta 233:787–805. Siehl, D.L., and Conn, E.E. (1988). Kinetic and regulatory properties of arogenate dehydratase in seedlings of Sorghum bicolor (L.) Moench. Arch. Biochem. Biophys. 260:822–829. Sun, Y., Li, H., and Huang, J.R. (2012). Arabidopsis TT19 functions as a carrier to transport anthocyanin from the cytosol to tonoplasts. Mol. Plant 5:387–400. Tzin, V., and Galili, G. (2010). The biosynthetic pathways for shikimate and aromatic amino acids in Arabidopsis thaliana. Arabidopsis Book 8:e0132. Vogt, T. (2010). Phenylpropanoid biosynthesis. Mol. Plant 3:2–20. Wang, H., Fan, W., Li, H., Yang, J., Huang, J., and Zhang, P. (2013). Functional characterization of dihydroflavonol-4-reductase in anthocyanin biosynthesis of purple sweet potato underlies the direct evidence of anthocyanins function against abiotic stresses. PLoS One 8:e78484. Xie, D.Y., Sharma, S.B., Paiva, N.L., Ferreira, D., and Dixon, R.A. (2003). Role of anthocyanidin reductase, encoded by BANYULS in plant flavonoid biosynthesis. Science 299:396–399. Xie, Y., Tan, H., Ma, Z., and Huang, J. (2016). DELLA proteins promote anthocyanin biosynthesis via sequestering MYBL2 and JAZ suppressors of the MYB/bHLH/WD40 complex in Arabidopsis thaliana. Mol. Plant 9:711–721. Yamada, K., Lim, J., Dale, J.M., Chen, H., Shinn, P., Palm, C.J., Southwick, A.M., Wu, H.C., Kim, C., Nguyen, M., et al. (2003). Empirical analysis of transcriptional activity in the Arabidopsis genome. Science 302:842–846. Yamada, T., Matsuda, F., Kasai, K., Fukuoka, S., Kitamura, K., Tozawa, Y., Miyagawa, H., and Wakasa, K. (2008). Mutation of a rice gene encoding a phenylalanine biosynthetic enzyme results in accumulation of phenylalanine and tryptophan. Plant Cell 20:1316– 1329. Yoo, H., Widhalm, J.R., Qian, Y., Maeda, H., Cooper, B.R., Jannasch, A.S., Gonda, I., Lewinsohn, E., Rhodes, D., and Dudareva, N. (2013). An alternative pathway contributes to phenylalanine biosynthesis in plants via a cytosolic tyrosine:phenylpyruvate aminotransferase. Nat. Commun. 4:2833. Zhang, Y., Butelli, E., Alseekh, S., Tohge, T., Rallapalli, G., Luo, J., Kawar, P.G., Hill, L., Santino, A., Fernie, A.R., et al. (2015). Multilevel engineering facilitates the production of phenylpropanoid compounds in tomato. Nat. Commun. 6:8635.
