Differential Accumulation of Anthocyanins in Dendrobium officinale Stems with Red and Green Peels

Dendrobium officinale stems, including red and green stems, are widely used as a dietary supplement to develop nutraceutical beverages and food products. However, there is no detailed information on pigment composition of red and green stems. Here, we investigated the content and composition of pigments in red and green stems by Ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry and assessed the differential accumulation of anthocyanins at the molecular level. The color of peels in red stems was caused by the presence of anthocyanins in epidermal cells unlike the peels of green stems. The glucoside derivatives delphinidin and cyanidin are responsible for the red color. Within the D. officinale anthocyanidin biosynthetic pathway, DoANS and DoUFGT, coding for anthocyanidin synthase and UDP-glucose flavonoid-3-O-glucosyltransferase, respectively, are critical regulatory genes related to the differential accumulation of anthocyanidin. These findings provide a more complete profile of pigments, especially anthocyanin, in D. officinale stems, and lay a foundation for producing functional foods.


Introduction
Dendrobium officinale (Chinese name: Tie-pi-shi-hu) is an edible perennial orchid plant that has commonly been employed as a functional food in China and other Asian countries for thousands of years [1]. D. officinale is frequently infused in juice, tea and wine as a functional liquid, but can also be chewed directly or stewed in soup, porridge and other dishes as a high-quality agricultural vegetable in dietary supplementation [2], and has been approved by the China Food and Drug Administration (http: //eng.sfda.gov.cn/WS03/CL0755/). Phytochemical analyses of stems in D. officinale have validated the presence of antioxidants [3], which are associated with anti-angiogenesis [4], immuno-stimulating activity [5] and the attenuation of ulcerative colitis [6], biofunctional properties that are due to large amounts of polysaccharides, flavonoid, phenanthrenes, bibenzyls and polyphenolic compounds [7].
Currently, the quality of D. officinale is irregular throughout the course of artificial culture as a result of distinct phenotypic traits, i.e., non-clonal material [8]. Differences in the biological characters of D. officinale stems, such as color, size, hardness and thickness, may exhibit different health functions [9]. D. officinale stems are a widely used phytomedicinal organ [10]. Depending on the absence or presence of purple spots on their stems, D. officinale can be divided into stems with a green peel (hereafter green stems) and stems with a red peel (hereafter red stems), both of which exist widely on the current market. It is generally believed that these two phenotypes of D. officinale stems result from a distinct content and composition of pigments. However, information on pigment composition of the red and green D. officinale stems, as well as the differential regulation of anthocyanin pigmentation in D. officinale stems, is scanty.
The diversity of colors in higher plant organs, such as flowers, leaves, seeds and fruits, is primarily determined by the differential accumulation of anthocyanins [11], but these colors can be enhanced by co-pigmentation with colorless flavonoids. Anthocyanin is divided into three distinct lineages depending on its hydroxylation state, including pelargonidin (orange/red), cyanidin (pink/magenta) and delphinidin (purple/blue) [12]. Anthocyanin biosynthetic pathways, key enzymes as well transcription factors that affect anthocyanin content have been intensively elucidated in plants [13]. In the Orchidaceae, both chalcone synthase (CHS) and dihydroflavonol 4-reductase (DFR) have been cloned from Bromheadia finlaysoniana [14], CHS, DFR and a flavonoid 3 5 -hydroxylase (F3 5 H) from Dendrobium hybrids including 'Sonia Earsakul', 'Sirin classic', 'Suree white' and 'Jasmine white' [15] and CHS, DFR and F3 5 H from Phalaenopsis [16,17]. Anthocyanin synthase (ANS, EC 1.14.11.19) is responsible for converting colorless leucoanthocyanidins to colored anthocyanidins, in particular, via a conserved and well-characterized pathway in a wide range of plant species, including Antirrhinum majus, Arabidopsis thaliana, Malus sieversii, Petunia hybrida, Punica granatum, Vitis vinifera and Zea mays [18]. UDP-glucose flavonoid 3-O-glucosyl transferase (UFGT, EC 2.4.1.115), which is involved in the last dedicated step of anthocyanin biosynthesis, catalyzes the glycosylation of anthocyanidins and produces the first stable anthocyanins [19]. The transcript levels of ANS and UFGT are correlated with anthocyanin accumulation in the skin of Vitis vinifera grapes [20]. However, the genes coding for anthocyanin synthase (DoANS) and UFGT from D. officinale, i.e., DoUFGT, have not been detected yet, and no studies have attempted to quantify the expression levels among genes of the anthocyanin biosynthetic pathway in this medicinal orchid.
D. officinale with red and green stems are widely available and are considered in China as a traditional promising functional food [1]. Thus, the aim of the present study was to compare the pigment (flavonoids, anthocyanins, and carotenoids) profiles of red and green stems using ultra performance liquid chromatography coupled with photodiode array detection quadrupole time-of-flight mass spectrometry (UPLC-PDA-QTOF-MS) analysis. In addition, we aimed to unveil differential anthocyanin accumulation between red and green stems at the molecular level. These findings will advance our understanding of pigment content and composition in D. officinale stems, and will provide additional insight into the regulation of anthocyanin biosynthesis in orchids.

Red D. officinale Stems Exhibit High Antioxidant Activity
The antioxidant activity of D. officinale stems was determined by the DPPH (2, 2-diphenyl-1picryhydrazyl) free-radical scavenging assay. As shown in Figure 1, the DPPH free-radical scavenging activity in red stems (93.5%) was slightly higher than in green stems (90.1%) when used at 800 µg/mL. Similarly, red stems showed high antioxidant activity at all concentrations. Moreover, red stems had a much lower IC 50 value than green stems. The IC 50 value of red stems was statistically similar to that of ascorbic acid (AA) (IC 50 = 38.21 µg/mL) (Figure 1). Antioxidant capacity is strongly associated with anthocyanin content in pomegranate juice [21], indicating that anthocyanins contribute to antioxidant activity. These results correspond to a previous study in which D. officinale stems were an important source of antioxidants with 97.1% DPPH scavenging activity at 10.0 mg/mL [3], suggesting that antioxidant activity of D. officinale stems is likely to be related to pigment content, especially anthocyanin content. DPPH radical scavenging activity of AA, green and red stems extracts; (B) IC50 value of extract from red and green D. officinale stems. AA: ascorbic acid; DPPH: 2, 2-diphenyl-1-picryhydrazyl. Each data bar represents the mean ± standard deviation (SD; n = 3). Different letters above bars indicate significant differences among different treatments at p < 0.05 based on student's t-test. Both red and green stems exhibit a significantly lower IC50 value according to a student's t-test (p < 0.05) when compared with AA, and red stems demonstrate better antioxidant activities than green stems.

Red D. officinale Stems Possess High Anthocyanin Content
In general, plant pigments are composed of anthocyanins, carotenoid and flavonoids [12]. Total flavonoid content in green and red D. officinale stems was significantly different (Figure 2). Green stems exhibited significantly lower total flavonoid content (3.6 mg/g DW) than red stems (4.1 mg/g DW). The anthocyanin content in red stems (106.4 μg/g DW) was significantly (5.85-fold, p < 0.01) higher than in green stems (18.2 μg/g DW). However, the carotenoid content in green stems (122.1 μg/g DW) and red stems (129.9 μg/g DW) were not significantly different. These findings indicate that anthocyanins might be responsible for the red color in red D. officinale stems and higher anthocyanin content corresponds to a deeper red color, which also indicates that anthocyanins are found in small amounts in green stems. A detailed composition of anthocyanins in D. officinale stems still has to be clearly determined. The red color in peels of Pyrus pyrifolia fruit is highly correlated with the content and composition of anthocyanins [22].

Red D. officinale Stems Possess High Anthocyanin Content
In general, plant pigments are composed of anthocyanins, carotenoid and flavonoids [12]. Total flavonoid content in green and red D. officinale stems was significantly different (Figure 2). Green stems exhibited significantly lower total flavonoid content (3.6 mg/g DW) than red stems (4.1 mg/g DW). The anthocyanin content in red stems (106.4 µg/g DW) was significantly (5.85-fold, p < 0.01) higher than in green stems (18.2 µg/g DW). However, the carotenoid content in green stems (122.1 µg/g DW) and red stems (129.9 µg/g DW) were not significantly different. These findings indicate that anthocyanins might be responsible for the red color in red D. officinale stems and higher anthocyanin content corresponds to a deeper red color, which also indicates that anthocyanins are found in small amounts in green stems. A detailed composition of anthocyanins in D. officinale stems still has to be clearly determined. The red color in peels of Pyrus pyrifolia fruit is highly correlated with the content and composition of anthocyanins [22].

The Red Color of Stem Peels Is Mainly Localized in Epidermal Cells
To further understand the function of different pigments (including anthocyanins, flavonoids, and carotenoids), the distribution of pigment-containing cells in green and red D. officinale stems was investigated ( Figure 3). Red pigments (mainly anthocyanins) were only observed in the epidermal cells of red stems, but no red pigments were detected in green stems, corresponding to the visual phenotype of D. officinale stems ( Figure 2). This is also related to the relatively high anthocyanin content in the peels of red stems (222.3 mg/g DW). Carotenoids might play a greater role than flavonoids in producing yellow color in D. officinale stems, even though carotenoid content was much lower than flavonoid content ( Figure 2). The rationale is based on studies that have demonstrated that carotenoids play a major role in contributing a yellow color among a wider range of vegetables and fruits [26]. Overall, individual epidermal cells in the outer layer of green stems are arranged loosely, while epidermal cells in the outer layer of red stems are aligned tightly, and the intercellular space is small ( Figure 3G,H). Furthermore, red pigmented cells showed a dot-like pattern, as seen for the epidermis of red stems ( Figure 3J), but this phenomenon could not be observed in epidermal cells of green stems. After the epidermis was peeled off from green and red stems, it was found that anthocyanin content was significantly higher in the peels of red stems (222.3 mg/g DW) than in the peels of green stems (38.7 mg/g DW). The presence of pink/magenta and purple/blue in red D. officinale stems suggests that cyanidin-and delphinidin-based anthocyanins might accumulate [11]. The red D. officinale stems looked the same as red-fleshed grape berries [27], suggesting that the anthocyanin composition, including diglucosides of delphinidin, cyanidin, malvidin and peonidin, might be similar.

The Red Color of Stem Peels Is Mainly Localized in Epidermal Cells
To further understand the function of different pigments (including anthocyanins, flavonoids, and carotenoids), the distribution of pigment-containing cells in green and red D. officinale stems was investigated (Figure 3). Red pigments (mainly anthocyanins) were only observed in the epidermal cells of red stems, but no red pigments were detected in green stems, corresponding to the visual phenotype of D. officinale stems (Figure 2). This is also related to the relatively high anthocyanin content in the peels of red stems (222.3 mg/g DW). Carotenoids might play a greater role than flavonoids in producing yellow color in D. officinale stems, even though carotenoid content was much lower than flavonoid content ( Figure 2). The rationale is based on studies that have demonstrated that carotenoids play a major role in contributing a yellow color among a wider range of vegetables and fruits [26]. Overall, individual epidermal cells in the outer layer of green stems are arranged loosely, while epidermal cells in the outer layer of red stems are aligned tightly, and the intercellular space is small ( Figure 3G,H). Furthermore, red pigmented cells showed a dot-like pattern, as seen for the epidermis of red stems ( Figure 3J), but this phenomenon could not be observed in epidermal cells of green stems. After the epidermis was peeled off from green and red stems, it was found that anthocyanin content was significantly higher in the peels of red stems (222.3 mg/g DW) than in the peels of green stems (38.7 mg/g DW). The presence of pink/magenta and purple/blue in red D. officinale stems suggests that cyanidin-and delphinidin-based anthocyanins might accumulate [11]. The red D. officinale stems looked the same as red-fleshed grape berries [27], suggesting that the anthocyanin composition, including diglucosides of delphinidin, cyanidin, malvidin and peonidin, might be similar.

Red-Peel Color Was Determined by Delphinidin and Cyanidin Derivatives of Anthocyanins
The present study is the first to report the main composition of anthocyanins in D. officinale stems. As mentioned above, red and green peels of D. officinale stems were chiefly determined by the presence of anthocyanin-containing epidermal cells in their peels. Therefore, peels were used to analyze anthocyanin composition by UPLC-QTOF-MS. Five categories of anthocyanins, i.e., containing pelargonidin (m/z 271), cyanidin (m/z 287), peonidin (m/z 301), delphinidin (m/z 303) and malvidin (m/z 331), were identified in the peels of red and green D. officinale stems via precursor-ion analysis. As shown in Table 1 and Figure S1, seven anthocyanin compounds were determined. Next, we provide a summary of the differences in retention times, molecular ions, aglycone ions, main MS 2 fragments, and the ratio of each compound in peels of red and green stems. Red/green peak area ratio is equal to the ratio of (the peak area of red stems) to (the peak area of green stems). + indicates that the compound was only detected in red stems, but not in green stems. − indicates that the compound was only detected in green stems, but not in red stems.  (Table 1). In addition, delphinidin 3,5-O-diglucoside, cyanidin 3-O-galactoside and cyanidin 3-O-glucoside were only detected in red stems, but not in green stems. Although the composition of anthocyanins was identified in D. officinale stems, these compounds were similar to banana fruit [28] and grape berries [27]. These results indicate that diglucoside or glucoside derivatives of delphinidin and cyanidin are responsible for the purple-red color of red D. officinale stems.

DoANS and DoUFGT Are Vital Regulatory Genes during Accumulation of Anthocyanidins
To further clarify the correlation of anthocyanin content with transcript levels of genes involved in anthocyanin biosynthesis, 13 anthocyanin-related genes in red and green D. officinale stems were detected by quantitative real-time PCR (Figure 4). The transcript levels of most genes were much higher in red stems than in green stems. Anthocyanin biosynthesis in higher plants is controlled by structural and regulatory genes, i.e., transcription factors [11]. The expression levels were in line with the distribution of anthocyanins: in red stems, the level expression of F3H and F3 H was relatively higher than in green stems, 2.89-4.13-fold and 3.55-4.37-fold, respectively. In addition, relative to green stems, F3 5 H1 and F3 5 H2 exhibited a significantly lower expression (0.10-and 0.02-fold, respectively) in red stems, while F3 5 H3 was highly expressed (8.23-fold) in red stems, indicating that F3 5 H3 might exercise a different function. The hydroxylation state of flavonoids is dominated by the enzymatic activity of F3 H and F3 5 H, corresponding to 'red' and 'purple' genes, respectively [12], caused by the generation of cyanidin (pink/magenta) and delphinidin (purple/blue)-based anthocyanins [19]. The expression levels of genes related to phenylpropanoid and early anthocyanin biosynthesis were lower in red stems, while genes related to late anthocyanin biosynthesis were higher in red stems ( Figure 4). Interestingly, the abundance of anthocyanidin synthase gene (DoANS) transcript was 22.12-fold higher in red stems than in green stems ( Figure 4). ANS, an enzyme in the anthocyanin biosynthetic pathway, catalyzes the reaction from colorless leucoanthocyanidins to colored anthocyanidins in plants [18]. The UDP-glucose flavonoid 3-O-glucosyltransferase (UFGT) is the final enzyme in the anthocyanin pathway, transferring a glucosyl moiety from UDP-glucose to the 3-hydroxyl group, and this process is vital for forming stable anthocyanidins [29]. In addition to DoANS, the same phenomenon occurred in DoUFGT1 and DoUFGT2, with 8.73-fold and 5.24-fold higher levels, respectively in red stems than in green stems. Given that the high expression levels of DoANS, DoUFGT1 and DoUFGT2 in red stems are significantly positively correlated with anthocyanidin content (R 2 = 0.98, 0.96 and 0.90, respectively, p < 0.01) (Figure 4), it is possible that they might be critical regulatory genes in anthocyanidin accumulation in red D. officinale stems. To date, several structural genes as well as some regulatory genes in the anthocyanidin biosynthetic pathway of the Dendrobium genus have been isolated [7,30,31], but no reports exist on DoANS, DoUFGT1 and DoUFGT2 in this pathway of D. officinale.
A target sequence (Unigene0101519) coding for anthocyanidin synthase (EC: 1.14.11.19) was mined from the D. officinale transcriptome [32]. Nested PCR yielded the full-length cDNA of DoANS, and possessed a 1077-bp open reading frame (ORF) and a polypeptide of 358 amino acids, which was submitted to NCBI (GenBank accession number MH458949). Just as important, DoUFGT1 and DoUFGT2 were isolated and submitted to NCBI, with a 1359-bp ORF coding for 452 amino acids, and a 999-bp ORF coding for 332 amino acids, respectively, which were also submitted to NCBI (GenBank accession number MH663506 and MH663505, respectively). Phylogenetic analysis revealed that DoANS was evolutionarily closer to monocotyledonous plants than to dicotyledonous plants ( Figure S2). Multi-alignment by ClustalX suggested that DoANS was phylogenetically closer to Oncidium hybridum (77.99% identity) and Cymbidium dayanum (84.12% identity). Moreover, a highly conserved 2OG-Fe(II) oxygenase domain, which is a typical ANS binding site [33], as well as three critical binding sites (ferrous-iron coordination, substrate binding site and a binding site of 2-oxoglutarate) were strictly conserved in DoANS ( Figures S3 and S4), as well as in most well-characterized ANS proteins in plants such as Arabidopsis thaliana [33], Oncidium Gower Ramsey [30], Pyrus pyrifolia [22], Solanum melongena [34] and Vitis vinifera [20]. As for DoANS, a plant secondary product glycosyltransferase (PSPG) box (belonging to the UDP-glucuronosyltransferase family) was conserved in DoUFGT1 and DoUFGT2 ( Figure S5). Unstable anthocyanidins are stabilized by glycosylation at the 3-O-position or the 5-O-position by UF3GT and UF5GT, respectively, and then transported to acidic vacuoles, which is essential for coloration [13]. Phylogenetic analysis revealed that DoUFGT1 was clustered in the UF3GT group, while DoUFGT2 was clustered in the UF5GT group ( Figure S6). As shown in Figure 4, DoUFGT1 exhibited higher transcript levels in red stems than DoUFGT2, which corresponded well with the higher number of anthocyanidin compounds with a 3-O-position (Table 1). Thus, DoUFGT was thought to be the key enzyme for anthocyanin biosynthesis in D. officinale. SmFAS, which encodes anthocyanidin synthase, is a putative candidate gene responsible for the purple-color peels of Solanum melongena L. [34]. In addition, UFGT might strongly affect the coloration of grapevine peels [20]. Taken together, the expression of DoANS and DoUFGT was determined to be much higher in red stems than in green stems, which coincides with the results observed in peel color and anthocyanin content, indicating the role of DoANS and DoUFGT in regulating the accumulation of anthocyanins. Phylogenetic analysis revealed that DoUFGT1 was clustered in the UF3GT group, while DoUFGT2 was clustered in the UF5GT group ( Figure S6). As shown in Figure 4, DoUFGT1 exhibited higher transcript levels in red stems than DoUFGT2, which corresponded well with the higher number of anthocyanidin compounds with a 3-O-position (Table 1). Thus, DoUFGT was thought to be the key enzyme for anthocyanin biosynthesis in D. officinale. SmFAS, which encodes anthocyanidin synthase, is a putative candidate gene responsible for the purple-color peels of Solanum melongena L. [34]. In addition, UFGT might strongly affect the coloration of grapevine peels [20]. Taken together, the expression of DoANS and DoUFGT was determined to be much higher in red stems than in green stems, which coincides with the results observed in peel color and anthocyanin content, indicating the role of DoANS and DoUFGT in regulating the accumulation of anthocyanins.

Plant Materials
Green stem D. officinale 'T32-5 (mainly distributed in sheltered and moist places in forests) and red stem D. officinale 'T32-4 (epiphytes growing on rocks in sunlight) were sampled under natural conditions in a greenhouse of the South China Botanical Garden, Chinese Academy of Sciences, Guangzhou, China (23 • 10 N, 113 • 21 E). Ten stems were gathered from an individual potted plant, and 10 independent plotted plants were sampled for each stem type. The plants were two years old. A total of 200 stems (including 100 green stems and 100 red stems, with a uniform size approximately 20 cm in height and 0.5 cm in diameter) were collected in a ready-to-eat stage (12 months after seedlings were transplanted), which is the optimum harvest time according to the Chinese pharmacopoeia [10]. Half of the green and red stems were stored separately at −80 • C in a refrigerator (Haier Group Co., Qingdao, China) until the analyses of metabolites. To determine the content of total flavonoids, anthocyanins and carotenoids, the remaining stems were oven dried, ground into powder with a diameter less than 0.3 mm, and stored in a vacuum pack (VIP320, Beijing Torch SMT Inc. Co., Beijing, China) at 4 • C.

Assessment of Total Flavonoids, Anthocyanins and Carotenoids Content in D. officinale Stems
Total flavonoid content was detected by a colorimetric method [28] under natural light at room temperature with minor modifications. Briefly, powdered oven-dried stems (1.0 g) were extracted in an sxt-06 Soxhlet extractor (Hangzhou Chincan Trading Co., Hangzhou, China) with 20 mL of 80% aqueous ethanol for 1 h at 80 • C, and the residues were re-extracted by the same solvent. Both supernatants were pooled and insoluble material was removed by a 0.45 µm filter (jtsf0311, Tianjin Jinteng Experiment Equipment Co., Ltd., Tianjin, China) until a final volume of 50 mL was obtained. Extract (0.5 mL) was transferred to a 10-mL plastic tube containing 0.3 mL NaNO 2 (5%, w/v) for 5 min. Then, 0.3 mL Al(NO 3 ) 3 ·9H 2 O (10%, w/v) was added. After incubation for 6 min, the reaction was terminated by adding 2 mL of 1 m NaOH, and further diluted with 30% aqueous ethanol up to 10 mL. A standard curve was calibrated using diluted rutin with 30% aqueous ethanol (0-400 µg/mL). Absorbance at 510 nm was recorded with a UV-6000 spectrophotometer (Shanghai Metash Instruments Co., Shanghai, China). Total flavonoid content was quantified as mg of rutin equivalent per gram dry weight (DW).
Total anthocyanin content of green and red stems was measured by a pH differential protocol [35]. Detailed information is provided in the Supplementary Information.
Total carotenoid content was determined using the protocol of Lichtenthaler and Wellburn [36]. Detailed information is described in the Supplementary Information.

Compositional Analysis of Flavonoids and Anthocyanins by UPLC-QTOF-MS
Compositional analysis of flavonoids and anthocyanins in red and green stems of D. officinale was identical to a previously reported procedure [28]. The detailed protocol is described in the Supplementary Materials and methods (Supplementary Information).

Evaluation of Antioxidant Activity In Vitro
Antioxidant activity in vitro was determined using a DPPH free-radical scavenging test following Luo et al. [37]. Briefly, 0.1 mL of extract at five concentrations (50, 100, 200, 400 and 800 µg/mL) was separately mixed with 3.9 mL of DPPH solution (0.1 mmol/L in methanol). The reaction mixture was shaken vigorously and placed in the dark at 25 • C for 30 min. The decolorization of DPPH was monitored at 517 nm with a UV-6000 spectrophotometer (Shanghai Metash Instruments Co., Shanghai, China). DPPH radical scavenging activity (%) was determined using the following equation: DPPH radical scavenging activity (%) = 1 − (A sample − A control )/A blank × 100, where A blank is the absorbance of the control (DPPH solution without extract), A control is the absorbance of the extract without DPPH, and A sample is the absorbance of the extract with DPPH. The activity of the extract was expressed as 50% inhibition concentration (IC50, µg/mL), or half of radical scavenging activity, which was calculated using a forecast linear regression in Excel 2013 (Microsoft Co., Redmond, WA, USA). Ascorbic acid solution (identical concentration as extract) was used as the positive control.

Light Microscopy of D. officinale Stem
To determine pigment distribution in D. officinale stems, green and red stems were cut by hand using a razor blade (AS-D2, Shanghai Razor Blade Co., Ltd., Shanghai, China). Samples were mounted on glass slides (BM5116, SailBoat Lab Co., Ltd., Ningbo, China) and covered with a drop of distilled water. Epidermal micro-morphology was visualized using a Leica DM1000 LED microscope (Leica, Wetzlar, Germany) with bright-field illumination.

Gene Expression Analysis by Quantitative Real-Time PCR
RNA was extracted and cDNA was prepared from D. officinale stems according to Yu et al. [38]. The gene-specific primers for anthocyanin biosynthesis listed in Table S2 were designed by Primer Premier 5.0 (Premier Biosoft International, Palo Alto, CA, USA). Transcript abundance of the anthocyanin biosynthesis-related genes were analyzed on an ABI 7500 real-time PCR instrument (Applied Biosystems, Foster City, CA, USA), using iTaq TM Universal SYBR ® Green Supermix (Takara Bio Inc., Kyoto, Japan) in accordance with the manufacturer's protocol. D. officinale actin (GenBank accession: JX294908) was used as an internal standard to quantify gene targets. Relative RNA levels were calculated using the 2 −∆∆CT method [39].

Cloning of Genes and Phylogenetic Analysis
Based on our reported D. officinale transcriptome [32], a target sequence (Unigene0101519) coding anthocyanidin synthase, designated as DoANS, was amplified from cDNA of red and green D. officinale stems separately by using the Premix HS PrimeSTAR HS DNA Polymerase kit (Takara Bio Inc., Dalian, China) with specific primers (Table S2). Likewise, DoUFGT1 (Unigene0117481) and DoUFGT2 (Unigene0142333) were individually cloned from red and green stems. Thermal cycling was carried out as follows: 94 • C for 3 min, 30 cycles of 94 • C for 30 s, 55 • C for 30 s, and 72 • C for 1 min, and then 72 • C for 10 min. Purified PCR products were sub-cloned into pMD18-T vector (Takara Bio Inc., Dalian, China) and then sequenced at the Beijing Genomics Institute (BGI, Shenzhen, China).

Statistical Analysis
All analyses were conducted in triplicate. Data were analyzed using SPSS version 22.0 software (IBM Corp., Armonk, NY, USA). Statistical analysis of metabolites in green and red stems was performed by one-way analysis of variance (ANOVA) with mean separation by a student's t-test (p < 0.05). The level of statistical significance was set at p < 0.05. Correlation analysis was performed with Pearson's correlation coefficient (R 2 ) at p < 0.01. Graphs were generated with SigmaPlot 12.5 (Systat Software Inc., San Jose, CA, USA).

Conclusions
As shown in Figure 5, pigment content, composition, and distribution in green and red D. officinale stems have been investigated. A significantly higher content of anthocyanins (5.82-fold) was observed in red stems relative to green stems. The color of red stems is primarily determined by anthocyanin-containing epidermal cells in the stem peel. Among anthocyanins that were detected, delphinidin 3,5-O-diglucoside and cyanidin 3-O-glucoside are likely responsible for the red peel color. These two phenotypes based on anthocyanidin accumulation in green and red D. officinale stems may be controlled by DoANS and DoUFGT, which are well correlated with the amount of anthocyanidins, and reflect a highly homologous level with reported proteins involved in anthocyanidin biosynthesis. In addition, apigenin and genistin 7-O-gentiobioside are cardinal components in flavonoids, with a consistently high content of total flavonoids in red D. officinale stems, which provides them with high antioxidant activity. This information contributes to a broader understanding of the formation and pigmentation of D. officinale stems, and is thus closer to the ability to create overproduced anthocyanin-based functional foods.
performed by one-way analysis of variance (ANOVA) with mean separation by a student's t-test (p < 0.05). The level of statistical significance was set at p < 0.05. Correlation analysis was performed with Pearson's correlation coefficient (R 2 ) at p < 0.01. Graphs were generated with SigmaPlot 12.5 (Systat Software Inc., San Jose, CA, USA).

Conclusions
As shown in Figure 5, pigment content, composition, and distribution in green and red D. officinale stems have been investigated. A significantly higher content of anthocyanins (5.82-fold) was observed in red stems relative to green stems. The color of red stems is primarily determined by anthocyanin-containing epidermal cells in the stem peel. Among anthocyanins that were detected, delphinidin 3,5-O-diglucoside and cyanidin 3-O-glucoside are likely responsible for the red peel color. These two phenotypes based on anthocyanidin accumulation in green and red D. officinale stems may be controlled by DoANS and DoUFGT, which are well correlated with the amount of anthocyanidins, and reflect a highly homologous level with reported proteins involved in anthocyanidin biosynthesis. In addition, apigenin and genistin 7-O-gentiobioside are cardinal components in flavonoids, with a consistently high content of total flavonoids in red D. officinale stems, which provides them with high antioxidant activity. This information contributes to a broader understanding of the formation and pigmentation of D. officinale stems, and is thus closer to the ability to create overproduced anthocyanin-based functional foods.