Cloning and molecular characterization of R2R3-MYB and bHLH-MYC ...

Tree Genetics & Genomes (2010) 6:101–112 DOI 10.1007/s11295-009-0232-y

ORIGINAL PAPER

Cloning and molecular characterization of R2R3-MYB and bHLH-MYC transcription factors from Citrus sinensis Antonietta Cultrone & Paola S. Cotroneo & Giuseppe Reforgiato Recupero

Received: 13 June 2008 / Revised: 19 August 2009 / Accepted: 26 August 2009 / Published online: 17 September 2009 # Springer-Verlag 2009

Abstract Members of the MYB and MYC family regulate the biosynthesis of phenylpropanoids in several plant species. Two sequences, called CsMYB8 and CsMYC2, were identified from Citrus sinensis, and both the cDNA and the genomic clones were isolated and characterized from the flesh of common and blood oranges. Analysis by real-time polymerase chain reaction showed that the expression pattern of CsMYC2 is generally higher in rind than in flesh and in blood oranges than in common ones. In contrast, no significant difference in expression was observed for CsMYB8. The expression pattern of the structural genes chalcone synthase, anthocyanidin synthase, and UDP-glucose–flavonoid 3-O-glucosyltransferase, which code for three enzymes involved in the anthocyanin biosynthetic pathway, was also analyzed and correlated with CsMYC2, in flesh, rind, and leaf of the common and blood oranges, and in leaf of Citrus limon cultivars (characterized by anthocyanin absence or variable content). Surprisingly, CsMYC2 is highly expressed in the leaf and expression is correlated with UFGT expression in this Communicated by F. GMITTER A. Cultrone : P. S. Cotroneo : G. Reforgiato Recupero C.R.A-ACM Centro di Ricerca per l’Agrumicoltura e le Colture Mediterranee, Corso Savoia 190, 95024 Acireale, Italy A. Cultrone Parco Scientifico e Tecnologico della Sicilia, Zona Industriale, Blocco Palma 1, Stradale G. Agnelli angolo Stradale V Lancia, 95100 Catania, Italy Present Address: A. Cultrone (*) INRA UEPSD, bat 405—Domaine de Vilvert, 78352 Jouy-en-Josas cedex, France e-mail: [email protected]

organ. These results suggest that CsMYC2 is involved in the regulation of the flavonoid biosynthetic pathway in Citrus. Keywords Citrus . Flavonoid . Regulation Abbreviations CHS Chalcone synthase ANS Anthocyanidin synthase UFGT UDP-glucose–flavonoid 3-O-glucosyltransferase bHLH Basic helix-loop-helix TF Transcription factor

Introduction Flavonoids are polyphenolic, secondary plant metabolites highly represented in commonly consumed fruits, vegetables, and beverages. The flavonoid family is divided into a number of subgroupings; the six main classes are flavonols, flavones, flavan-3-ols, isoflavones, flavanones, and anthocyanidins. In recent decades, interest in flavonoids has been constantly growing, due to the increased awareness of their antioxidant activity associated with the prevention of coronary heart disease. Anthocyanins, particularly anthocyanidins glycosides, have been the subject of many recent studies (Bagchi et al. 2000, 2004; Lila 2004). The aromatic rings of the flavonoid molecule allow the donation and acceptance of electrons from free radical species (Kanner et al. 1994). In addition to quenching free radicals, flavonoids are able to regenerate the traditional antioxidant vitamins, vitamin C and vitamin E (Vinson et al. 1995). Anthocyanins are responsible for purple and blue colors and for some of the red tonalities of many fruits, vegetables, cereal grains, and flowers (Mol et al. 1998). They have long been the subject of investigation by botanists

102

and plant physiologists, due to their role as pollination attractants and phytoprotective agents (Gould 2004; Grotewold 2006). In Citrus, anthocyanins are expressed in young shoots and fruits, in some floral tissues of lemon [Citrus limon (L.) Burm. f.], citron (Citrus medica L.), and in some species of Papeda subgenus. In mature fruits, they are present exclusively in blood oranges and their hybrids. Flavonols, like quercetin, may also function as copigments for anthocyanins in fruits and flowers (Koes et al. 1994). Structural genes encoding the enzymes involved in the anthocyanin biosynthetic pathway are conserved among different species (Holton and Cornish 1995). This suggests that the variability of pigmentation patterns depends on differences in the regulation of these structural genes. The mechanisms regulating anthocyanin biosynthesis have been studied in various plant species (Dooner et al. 1991). Regulatory genes controlling the tissue specificity of the structural genes have been identified by mutant analysis in maize (Paz-Ares et al. 1986, 1987), snapdragon (Martin et al. 1991), petunia (Beld et al. 1989; Quattrocchio et al. 1993), and grapes (Azuma et al. 2008; Kobayashy et al. 2002). These genes have been found to belong to MYB and MYC transcription factors families and to the WD40 repeats family. In plants, MYB proteins are characterized by the R2R3-type MYB domain, comprising a conserved DNA-binding domain and an activation/repression domain (Martin and Paz-Ares 1997; Broun 2004; Koes et al. 2005). MYC proteins are characterized by the basic helix-loophelix (bHLH) domain that is required for the dimerization of two bHLH proteins (Ferré-D’Amaré et al. 1993, 1994). Moreover, it has been recently reported that the regulation of their transactivation strength is dependent on the conserved lysine and alanine in the interaction domain (Pattanaik et al. 2008). The WD40 repeats contain a protein binding domain characterized by the presence of seven regions of 40 amino acids each; each region is rich in tryptophan (W) and aspartic acid (D). The WD40 repeats can form complexes with MYB and MYC proteins. It has been proposed that these regulators interact with each other and constitute transcription complexes that bind to the promoters of the structural genes. Moreover, the regulation is specific for groups of genes acting at different stages in the biosynthetic pathway (Davies and Schwinn 2003). VvMYBA1 and VvMYBA2, for example, have been found to control berry skin color in Vitis vinifera, acting on the promoter of the last gene of the pathway (Kobayashy et al. 2002; Walker et al. 2007). Finally, transcriptional regulation of structural gene expression can also be mediated by the action of repressors; in some cases, a transcription factor can act as either an activator or a repressor (Park et al. 2008; Dubos et al. 2008; Coffman et al. 1997). In blood oranges [Citrus sinensis (L.) Osbeck], anthocyanin content varies considerably between the flesh and

Tree Genetics & Genomes (2010) 6:101–112

the rind, depending on genotype, rootstock, environmental conditions, and storage temperature (Christie et al. 1994; Shvarts et al. 1997; Kim et al. 2003; Lo Piero et al. 2005; Takos et al. 2006). In Cotroneo et al. (2006), the expression levels of some of the structural genes of the anthocyanin biosynthetic pathway, chalcone synthase (CHS), anthocyanidin synthase (ANS), and UDP-glucose–flavonoid 3-Oglucosyltransferase (UFGT), were analyzed in the flesh of blood and common oranges of 11 different genotypes and found to be strongly correlated with the pigment content. To date, no reports have been published on any regulatory genes associated with this pathway; thus, the identification of the factors controlling anthocyanin synthesis in citrus would be valuable for understanding the mechanism leading to its accumulation and distribution in different tissues. Since classical genetic studies are difficult in Citrus, due to their long juvenile period and high heterozygosity level, molecular characterization and gene expression analysis represent worthy instruments to improve upon the knowledge in this field and support the genetic improvement of fruit quality. In the present study, we report on the isolation and the molecular characterization of two putative regulatory genes from C. sinensis, CsMYB8 and CsMYC2. Correlations of CsMYC, with the structural genes CHS, ANS, and UFGT in different C. sinensis organs, as well as in the leaf of some cultivars of C. limon, are also presented. This is the first reported data on a flavonoid biosynthesis regulator gene in Citrus.

Materials and methods Plant material Fruits and leaves of “58-8D-1 Moro” nucellar selection (a blood orange) and “Valencia” cultivar (a common orange; Fig. 1) were sampled at the experimental orchard of the CRA-ACM Centro di Ricerca per l’Agrumicoltura e le Colture Mediterranee, located in Palazzelli (Siracusa), Italy, at eight progressive maturation stages and immediately processed: I = 25 Oct. 2005, II = 15 Nov. 2005, III = 1 Dec. 2005, IV = 29 Dec. 2005, V = 24 Jan. 2006, VI = 8 Feb. 2006, VII = 7 Mar. 2006, and VIII = 28 Mar. 2006. Lemon leaves were sampled from the following varieties: Femminello Siracusano, a common variety owning purple tinted young leaves and shoots; Femminello Zagara Bianca, carrying a mutation characterized by green young leaves and shoots; and Poros acidless, a chimeric variety producing purple, chimeric (characterized by purple tinted mid and major lateral veins) and green leaves variant branches, respectively (Fig. 2). Poros acidless lemon purple-tinted leaves represent the lemon tissue with the highest pigmentation in our experiment.


103

Fig. 1 Blood (“58-8D-1 Moro” nucellar selection) and common (“Valencia”) orange fruit flesh sampled at the experimental orchard of the CRA-ACM Centro di Ricerca per l’Agrumicoltura e le Colture Mediterranee, located in Palazzelli (Siracusa), Italy, at eight progressive maturation stages: I = 25 Oct. 2005, II = 15 Nov. 2005, III = 1 Dec. 2005, IV = 29 Dec. 2005, V = 24 Jan. 2006, VI = 8 Feb. 2006, VII = 7 Mar. 2006, VIII = 28 Mar. 2006

Isolation of nucleic acids and synthesis of cDNA Plasmid isolation from Escherichia coli strains and DNA manipulation were performed as described by Sambrook and Russell (2001). The preparations of total DNA and RNA were performed as described by Specht et al. (1982) and Lockington et al. (1985). The quality of RNA was verified by the demonstration of intact ribosomal bands following agarose gel electrophoresis and by the absorbance ratios (A260/280) of 1.8 to 2.0. For cDNA synthesis, 0.5 to 1 μg of total RNA, previously treated with DNaseI (Amersham Biosciences, Piscataway, NJ, USA), was reverse transcribed using the High Capacity cDNA Archive Kit (Applied Biosystems, Foster City, CA, USA), following the protocol of the supplier. Absence of genomic DNA contamination in the RNA preparation was verified by polymerase chain reaction (PCR) amplification of a CsMYC2 intron-containing region, using the cDNA as templates.

like amplicons were isolated from C. sinensis (nucellar cv. 58-8D-1) flesh cDNA, using primers designed from a highly conserved region of plant MYBs, as described by Kobayashy et al. (2002). The MYC-like DNA partial sequence (800 bp) was amplified from genomic DNA purified from the leaf of the nucellar blood cultivar “58-8D1 Moro”. Degenerate primers were designed from published sequences of the anthocyanin MYC-like regulatory genes, as described by Fan et al. (2004). Fragments were isolated from agarose gel, cloned into the TOPO TA vector (Invitrogen Corp., Carlsbad, CA, USA), and sequenced. The cloned PCR products were subjected to sequence analysis by BLAST. The 5′ and 3′ ends of the two sequences were determined by SMART RACE PCR kit (BD Biosciences, Franklin Lakes, NJ, USA), using total RNA purified from fully colored blood orange flesh as template. All primer pairs used are listed in Tables 1 and 2. Phylogenetic analysis

Isolation of CsMYB8 and CsMYC2 MYB and MYC-like partial sequences were obtained by means of a degenerate primer PCR-based strategy. MYB-

Protein sequences were aligned using the CLUSTAL W program version 1.83 (Thompson et al. 1994). Phylogenetic analysis was performed by PHYLIP package of

Fig. 2 Variant leaves of C. limon chimeric acidless cultivar “Poros”. Purple leaf (a), lemon organ with the highest anthocyanin content in our experiment; chimeric leaves (b), characterized by purple tinted mid and major lateral veins; green leaves (c)

104 Table 1 Primers used for the isolation of the complete sequences

Tree Genetics & Genomes (2010) 6:101–112 Name

Sequence

Function

Myb8-fw Myb10-fw Myb11-rw Myb12-rw M8st-fw M8end-rw 3Myc2 3nesMyc2 5Myc2 5nesMyc2 Race3myc2-fw Race4myc2-fw Race6myc2-rw Race7myc2-rw Myc+1-fw

5-GAGATGATCATACGCGCCCACGCTCG-3 5-TGAATGTGGTGCCCGAGCCGGAAACGAC-3 5-ACCGAGCGTGGGCGCGTATGATCATCTC-3 5-CGATGACAAACCCGGCACGCTTGAATCG-3 5-ATGGCGTTTAATAGAAGAGAAGTGG-3 5-CTAACCAGTCTTTTGCATGG-3 5-GTGATGTTGCCAGAGATGGGAATCAAC-3 5-GACAATGCAAGCAATGGAACTTACACCTG-3 5-CAGGTGTAAGTTCCATTGCTTGCATTGTC-3 5-GTTGATTCCCATCTCTGGCAACATCAC-3 5-CTGGTCCCAGAGGATCCTAGTCTCCTTC-3 5-GATGACAGTGACCCTCTGTGTGCCAAGG-3 5-GAAGGAGACTAGGATCCTCTGGGACCAG-3 5-GTGCTCTTCCTGGCAAACCTTGGCCAG-3 5-ATGGCTAGTGCTGCTCAAAACCAGG-3

3end myb 8 RACE 3end myb 8 RACE nest 5end myb 8 RACE 5end myb 8 RACE nest myb8 transcription start myb8 transcription end 3end myc 2 RACE 3end myc 2 RACE nest 5end myc 2 RACE 5end myc 2 RACE nest 3end myc 2 RACE 3end myc 2 RACE nest 5end myc 2 RACE 5end myc 2 RACE nest myc2 transcription start

Myc-endbig-rw Myc+167-fw Myc+644-fw Myc1505-rw Myc1218-rw Myc1085-rw Myc+1086-fw Myc+1520-fw Myc1786-rw

5-TTAACACTTACCAGCGATCTTCC-3 5-GACTAGGAAGACAATGCAAG-3 5-CGATGAAGATGATGACAGTG-3 5-CCAACTCCTCTACTCTTGCCTCAAGC-3 5-CCTGGCCAATGCCTCTTAACCATTCCTCC-3 5-GTGTGCACCATCATCTATCCCAAGATCC-3 5-GGATCTTGGGATAGATGATGG-3 5-CAGAGATGGTGGAGCAGACA-3 5-CTGAATAGGCATCCAAGTGC-3

myc2 transcription end Primer used to sequence Primer used to sequence Primer used to sequence Primer used to sequence Primer used to sequence Primer used to sequence Primer used to sequence Primer used to sequence

programs (version 3.66; Felsenstein 1989), using amino acid sequence from the bHLH DNA binding domain, plus other regions that were clearly homologous across all proteins in the alignment. A distance matrix method, emTable 2 Primers and probes used in quantitative real-time PCR experiments

Gene detected

SF

CHS

ANS

UFGT

CsMYC2

ploying the Jones–Taylor–Thornton model (Jones et al. 1992), was used to compare the sequences, and a tree was derived using the neighbor-joining clustering method (Saitou and Nei 1987).

Primer/TaqMan probe (FAM/Tamra) Name

Sequence

EF-161-fw EF-88-rw

5-CTGCTGGACGCTCTTGACAA-3 5-TCCTGGAGTGGCAGACGAA-3

EF-161-88 (probe) CHS-261-fw CHS-333-rv CHS-261-333 (probe) ANS-702-fw ANS-771-rw ANS-702-771(probe) UFGT-191-fw UFGT-262-rw UFGT-191-262(probe) MYC2-9-fw MYC2-10-rw MYC2(probe)

5-ATCAATGAGCCGAAGAGGCCCTCAGA-3 5-TGACACCCATCTCGATAGTCTTGT-3 5-GGCGCCGATGATAATAGCA-3 5-CCTTGTTTGGCGATGGTGCGG-3 5-TCACTTTCATTCTTCACAACATGGT-3 5-TTAGCAGTCACCCATTTGTCTTTG-3 5-CCGGGCCTGCAGCTCTTCT-3 5-TCACTTCATTCCAGGAATGAATAAGA-3 5-TCCAAGTCTCCGGAAACAACTC-3 5-ACGCGTCGCCGACTTGCCTG-3 5-CGTGAATCAAGCAGAGGGTTTT-3 5-ATGATTGCCTTCTTGAAGTTCTTTC-3 5-CTTCCAAGGATGAAAACATGAGCCACATTC-3


Real-time PCR expression analysis Real-time PCR analyses were carried out using a 7300 realtime PCR System (Applied Biosystems). Amplification was conducted in 50 μl, following the TaqMan universal PCR Master Mix protocol (Applied Biosystems) and using 50 ng of cDNA as template. The thermal cycling conditions were as follows: 50°C for 2 min (one cycle), 95°C for 10 min (one cycle), 40 cycles at 95°C for 15 s, and 60°C for 1 min. For the specific detection of mRNA, primers and TaqMan probes (Tables 1 and 2) were designed using the ABI PRISM Primer Express version 3.0 (Applied Biosystems). All samples were measured in triplicate, and positive and negative controls were performed. The analytical method used was the relative quantification standard curve method. Data collected from the sequence detection system were expressed as Ct (threshold cycle for target amplification) values and treated as follows: C. sinensis elongation factor 1-alpha (accession number AY498567) was used for the normalization of the target genes CsMYC2, CitCHS (AB09351), ANS (AY500593), and UFGT (CF972319) expression, as it has been reported to be constitutively expressed (Mahe et al. 1992). A standard curve was generated for each gene using a cDNA serial dilution. The log input amount for the unknown samples was calculated using the formula: [(Ct value)−y]/m, where y=standard curve intercept value and m=standard curve slope value; the final input amount was calculated using the formula: 10^ (log input amount). The units of the calculated amount are the same as the units used to construct the standard curve, which are nanograms of cDNA. The amount of the gene of interest was then divided by the amount of the endogenous control gene to determine the normalized amount of target. The sample showing the lowest expression level of each target gene was then designated arbitrarily, as a calibrator for all the samples analyzed for that gene. Final expression values refer to fold increase relative to the endogenous control gene. Pearson correlation coefficient statistical analysis was performed by means of STATISTICA software (version 6; StatSoft, Tulsa, OK, USA). Correlations were marked significant for P=0. Evidence of coregulation was found for values approximated to r=+1 (perfect positive correlation).

Results Cloning and sequence characterization of a C. sinensis MYB domain transcription factor Three fragments of approximately 180 bp each were amplified from C. sinensis cDNA. BLAST analysis revealed that only one of the three fragments shared high homology

105

with MYB transcription factors. By rapid amplification of cDNA ends (RACE) PCR, both the 5′ and 3′ ends were identified, resulting in a full-length cDNA sequence called. CsMYB8 has a 954-bp open reading frame and encodes a protein of 317 amino acids. No introns were found to interrupt the genomic clone. We identified two different CsMYB8 cDNA clones (GenBank accession numbers EF537874 and EF537875) differing from each other at positions +271 and +298 in the protein sequence. CsMYB8 was also amplified and sequenced from “Valencia” common orange cultivar. Two cDNA clones (accession numbers EF537876 and EF537877), differing at positions +135, +172, +221, +241, and +253 in the protein sequence, were obtained. These are presumably products of different genes, in accordance with the observation that MYB and MYC transcription factors represent multigene families. All of the observed differences were located outside the R2R3 domain. The deduced CsMYB8 amino acid sequence showed the highest homology with MYB6 (Malus x domestica, accession number AAZ20429) at 62% identity and with VvSREBP, a berry protein of V. vinifera (accession number AY953543) at 59% identity, which has been shown to be a repressor of sucrose transporters. At the level of the nucleotide sequence, the highest similarity was found with many MYB-like genes from Arabidopsis thaliana (80% identity), involved in the response to different stimuli; these stimuli include abscisic acid, auxins, ethylene, gibberellins, and jasmonic acid.. No similarity was found with MYB-like proteins involved in the regulation of the anthocyanin biosynthetic pathway. The expression level of CsMYB8 transcripts was analyzed by reverse transcriptase (RT)-PCR in the flesh and rind of the common and the blood cultivars, sampled at different maturation stages. No difference in CsMYB8 transcript level was observed between the two cultivars or among the different maturation stages (data not shown). Cloning and sequence characterization of a C. sinensis MYC domain transcription factor A 800-bp DNA fragment was amplified from genomic DNA of the blood cultivar “58-8D-1 Moro” by PCR using degenerate primers designed from published sequences of the MYC-like anthocyanin regulatory genes as described by Fan et al. (2004). BLAST analysis of the cloned PCR product revealed a high similarity between the N-terminal 183 nucleotides of the sequence and various anthocyanin biosynthesis regulatory genes. In order to obtain the fulllength cDNA of this putative gene, specific primers on the 183-bp sequence were designed for RT-PCR. Total cDNA purified from “Moro” orange flesh was used as template. The complete cDNA sequence was found to be 1,971 bp and to encode a protein of 656 amino acids. The start of transcription was predicted at −15 from the first ATG

106

codon. The 3′-untranslated region contains the polyadenylation consensus element, AAUAAA. Northern blot analysis showed a transcript of about 2.1 kb (data not shown). The size of the mRNA indicates that the end point of transcription is within 200 nucleotides from the stop codon. The gene was named CsMYC2. We cloned the full-length CsMYC2 cDNA from the flesh of both the pigmented nucellar “58-8D-1 Moro” (accession number EF645810) and the common “Valencia” (accession number EF645811) cultivars. The two sequences differ from each other at positions +144, +169, +261, +311, +322, and +606 of the protein sequence. These differences are located outside the bHLH motif. We have also characterized CsMYC2 cloned from total DNA. The genomic sequence is 3,360 bp and consists of eight exons and seven introns (Fig. 3). The intron–exon structure of CsMYC2 is conserved among different anthocyanin biosynthesis regulators, showing highly variable exon size, ranging from only 15 bp for exon IV to 907 bp for exon VI. Phylogeny of the C. sinensis CsMYC2 bHLH-type factors Database searches have shown that the predicted CsMYC2 protein has structural similarities to several plant bHLH proteins. In Fig. 4, we show that CsMYC2 shares a highly conserved, 200 amino acid N-terminal domain with other bHLH proteins. This bHLH domain is located between position +462 and +503 of the protein sequence. The Cterminal region downstream the bHLH domain is generally poorly conserved, while the central domain is the least conserved. In order to visualize the level of similarity among bHLH heterologous proteins involved in anthocyanin biosynthesis, we constructed a phylogenetic tree. Twenty bHLH protein sequences from both monocotyledonous and dicotyledon-

Fig. 3 Map showing the overall structure of the CsMYC2 gene as deduced from the cDNA and genomic sequence. The entire CsMYC2 genomic sequence is 3,360 bp long and consists of eight exons and seven introns. Blocks represent exons and lines represent introns. The intron–exon structure of CsMYC2 is conserved among different anthocyanin biosynthesis regulators. The size of the exons is similar to those observed in some other anthocyanin regulators, and it is highly variable ranging from only 15 bp for exon IV to 907 bp for exon VI. The complete cDNA sequence is 1,971 bp long and encodes a 656-amino acid protein. The position of the ATG translation start and the TAG translation stop codon are indicated


ous species were aligned and subjected to phylogenetic analysis. The results, including bootstrap values, are shown in Fig. 5. In this phylogenetic tree, where monocot and dicot MYC-like proteins are clearly separated, one group represents the R-family proteins B, Sn, and LC from Zea mays. They present 55% identity with CsMYC2, and all of them are specific for the regulation of anthocyanin biosynthesis and are more distantly related to CsMYC2. On the contrary, CsMYC2 is closely related to AtMYC1 (AtbHLH12 accession number AF251697) from A. thaliana, bHLH33 (accession number DQ266451) from M. x domestica, and VvMYCA1 (ABM92332) from V. vinifera. AtMYC1 is a transcription factor placed in group III, subgroup f, of the A. thaliana bHLH TF classification (Heim et al. 2003). Subgroup IIIf includes the transcription factor involved in the regulation of anthocyanin biosynthesis (Heim et al. 2003). bHLH33 is a transcription factor from the apple, and it is a putative homolog of DELILA from Antirrhinum majus (accession number AAA32663), a known regulator of anthocyanin synthesis (Martin et al. 1991). VvMYCA1 also seems to be involved in the regulation of anthocyanin biosynthesis in the grape (P. Arce-Johnson and J.-T. Matus, unpublished). Moreover, CsMYC2 is also strongly related to Cornus bHLH transcription factors, to DELILA from A. majus, and to JAF13 (accession number AAC39455) of Petunia x hybrida, all of which have been characterized as anthocyanin biosynthesis regulators. Expression pattern of CsMYC2 supports its involvement in the regulation of flavonoids The expression of CsMYC2 was investigated during eight progressive maturation periods in the flesh (Fig. 6a) and the rind (Fig. 6b) of blood and common oranges, during the season 2005–2006, first by RT-PCR (data not shown) and then by real-time PCR). The probe used in this analysis was designed in a nonconserved region (central domain) of the corresponding protein (Fig. 4). Transcript levels of CsMYC2 were generally weak, but higher in the rind than in the flesh. This was true for both the blood and the common oranges (Fig. 6a, b); however, expression was higher in the blood cultivar. CsMYC2 expression was also analyzed in the leaves of blood and common oranges. Weak or null expression was expected, as C. sinensis do not synthesize anthocyanin in this organ. Surprisingly, CsMYC2 is highly expressed in the leaf, but no difference was found between the two cultivars (Fig. 7). On the contrary, differences in CsMYC2 expression were observed in leaf of lemon cultivars, characterized by the absence or variable content of anthocyanin (“Femminello Siracusano”, “Femminello Zagara Bianca”, and three variant leaves of “Poros” acidless; Figs. 2a–c). The maximum level was observed in the case of “Poros”


Fig. 4 Alignment of amino acid sequences from CsMYC2 and related plant bHLH proteins. C. sinensis CsMYC2 (EF645810), A. majus DEL (AAA32663), Cornus canadensis MYC-like anthocyanin regulatory protein (AY465415), M. x domestica bHLH33 (DQ266451), Z. mays Lc (P13526), P. x hybrida AN1(AAG25928). CsMYC2 shares a highly conserved N-terminal domain of about 200 amino acids with other bHLH proteins. The bHLH domain is located between positions

107

+462 and +503 of the protein sequence. Among these proteins, the Cterminal region downstream the bHLH domain is in general poorly conserved, while the middle domain is always the less conserved one. Identical amino acids are in black; similar amino acids are in gray. The bHLH domain and the region recognized by the CsMYC2 probe are indicated

108

Fig. 5 Phylogenetic tree of the CsMYC2 homologs constructed aligning 20 bHLH protein sequences from both monocotyledonous and dicotyledonous, detected by BLAST program. The sequences were aligned using CLUSTALW (Thompson et al. 1994), and the phylogenetic tree was created using the PHYLIP package (Saitou and Nei 1987). Bootstrap values greater than 700 are shown. Monocot and dicot MYC-like proteins are clearly separated, one group representing the Rfamily proteins B, Sn, and LC from Z. mays. They present 55% identity with CsMYC2, and all of them are specific for the regulation of anthocyanin biosynthesis and are more distantly related to CsMYC2. On the contrary, CsMYC2 is closely related to AtMYC1 (AtbHLH12 accession number AF251697) from A. thaliana, bHLH33 (accession number DQ266451) from M. x domestica, and VvMYCA1 (ABM92332) from V. vinifera. AtMYC1 is a transcription factor placed in group III, subgroup f, of the A. thaliana bHLH TF classification, involved in the regulation of anthocyanin biosynthesis. bHLH33, a TF from apple, is a putative homolog of DELILA from A. majus (accession number AAA32663), regulator of anthocyanin synthesis. VvMYCA involved in the regulation of anthocyanin biosynthesis in grape. CsMYC2 is also strongly related to Cornus bHLH TFs, to DELILA from A. majus, and to JAF13 (accession number AAC39455) of P. x hybrida, anthocyanin biosynthesis regulators

acidless lemon purple-tinted leaves, which represents the lemon organ with the highest pigmentation in our experiment (Fig. 2a). Particularly, CsMYC2 transcript levels in the “Poros” purple leaf resulted in twice as much CsMYC2 than observed in the anthocyaninless cultivar “Femminello Zagara Bianca”. Statistical analysis supports positive correlation between CsMYC2 and UFGT in leaf In order to investigate the possible role of CsMYC2 in the anthocyanin biosynthetic pathway, its transcription level, measured by real-time PCR, was correlated with the expression profiles of the structural genes chalcone synthase, anthocyanidin synthase, and UDP-glucose–flavonoid 3-O-glucosyltransferase, all of which have been previously


proven to be involved in this pathway in orange flesh (Cotroneo et al. 2006). However, no correlation was found between CsMYC2 and CHS, ANS, and UFGT in the flesh and the rind (data not shown). Alternatively, different results were obtained with the leaf (Figs. 7 and 8b–d). A basal CHS transcript level was observed in all C. sinensis and C. limon nonpigmented leaves, increasing in pigmented samples and showing the highest level in “Poros” lemon purple leaves. ANS was strongly expressed in the purple leaf of the “Poros” lemon, with a transcription level five times higher than observed for the “Femminello Siracusano” purple leaf and 33 times higher than the “Poros” chimeric leaf. No significant expression was detected in the green leaves analyzed. On the other hand, a significant level of UFGT transcript was detected in the leaf of both the anthocyanin and the anthocyaninless genotypes. Data were subjected to Pearson statistical analysis, and a P value of