Genetic engineering of novel yellow color african violet (Saintpaulia ionantha) produced by accumulation of Aureusidin 6-O-glucoside

Background Flower color is one of the main characteristics of ornamental plants. Aurones are light yellow flavonoids produced in the petals of a limited number of plant species including snapdragon (Antirrhinum majus). As a commercially-recognized species, African violet can be found in various colors except yellow. This research, aiming at changing the petals’ color of African violet from white to yellow, was conducted using the simultaneous expressions of chalcone 4’-O-glucosyltransferase (4’CGT) and aureusidin synthase (AS1) genes without the need for silencing anthocyanin biosynthesis pathway genes via both transient and stable transfer methods. Results The transient gene transfer among transgenic plants led to a clear change of petals’ color from white to light yellow. This occurs while no change was observed in non-transgenic (Wild type) petals. In total, 15 positive transgenic plants, produced via stable gene transfer, were detected. Moreover, since their flower color was yellow, both genes were present. Meanwhile, the corresponding transformation yield was determined 20-30%. The transformation, expression and integration of genes among T0 transgenic plants were verified using the PCR, qRT-PCR and Southern blotting techniques, respectively. Furthermore, the probable color change of petals’ cross-section and existence of Aureusidin 6-O-glucoside (AOG) compound were determined using a light microscope and HPLC-DAD-MSn analysis, correspondingly. Conclusions Generally, the creation of aurones biosynthesis pathway is only viable through the simultaneous expression of genes which leads to color change of African violet’s petal from white to yellow. This conclusion can lead to an effective strategy to produce yellow color in ornamental plant species. Supplementary Information The online version contains supplementary material available at 10.1186/s12575-022-00164-0.


Background
Production of new ornamental cultivars with different colors of flowers is one of the main goals in floriculture [1]. Variability in flower color can increase the commercial value and marketing of ornamental plants [2]. The global use of ornamental plants is more than $300 billion and the area under cultivation in 2018 is estimated 2,600,000 and 4200 ha worldwide and Iran respectively [2,3]. Pigments in higher plants are generally grouped in three major classes: flavonoids, carotenoids, and betalains [4]. Anthocyanins are a class of flavonoids responsible for a range of colors: from orange to red, violet and blue. Similarly, carotenoids and betalains mainly yield yellow or red colors. Flavonoids such as chalcone and flavone are pale-yellow and often invisible with human eye [4].
Aurone is a class of rare flavonoids with brighter yellow present flowers in a limited number of species, such as A. majus, Cosmos bipinnatus (garden cosmos), and Rajabi et al. Biological Procedures Online (2022) 24:3 Limonium (sea-lavender) [4][5][6]. Figure 1 shows flavonoids and a simple diagram of their biosynthetic pathways in A. majus. Biosynthesis of Aureusidin 6-O-glucoside (yellow pigment) and associated enzymes (AS1 and 4'CGT ) have been well characterized in this plant [5]. 2' ,4' ,6' ,4 tetrahydroxychalcone (THC) is the first compound in this pathway, which is the catalytic product of the key enzyme chalcone synthase (CHS). THC is glucosylated at its 4'-hydroxyl group in the cytoplasm, before transfer to the vacuole, and later converted to aureusidin 6-O-glucoside via the catalysis of AS1 [5]. THC can later be converted to aurones, flavanones, and other classes of flavonoids, including anthocyanins.
African violet (Saintpaulia ionantha H. Wendl.; Gesneriaceae) is a commercial ornamental plant that can be easily propagated [7]. African violet has more than 22 species and 2000 cultivars with diverse petal color, floral shape, and color range of white, red, purple, and pink [8]. The Saintpaulia 'Jolly Diamond' cultivar has white petals; the number of petals is much more than its flowers and the petals hang slightly as they grow inwards [9]. Despite the variety of colors available among African violets, however, yellow has not yet been observed in this species [5,7].
Transformation technology has potential to produce novel flower phenotypes that are not available in nature [1]. Genetic engineering can be used to create novel flower colors in a variety of ways: (a) by introducing new genes that encode enzymes to create new biosynthetic pathways that do not exist, (b) by directing biosynthetic pathways through up-regulation of existing genes, or (c) by suppressing biosynthetic pathways through down-regulation of genes [10]. However, genetic transformation is a valuable tool that can be exploited in plant functional genomics, gene discovery, and gene characterization [11].
Transient gene transformation (agroinfiltration system) using Agrobacterium-mediated Transformation (ATMT) in different plant tissues allows for rapid and scalable development of functional genomics assays [12][13][14][15]. This method is an efficient tool aimed at studying the function of the gene construct and tests it before stable transfer [16]. Stable transfer, on the other hand, is a lengthy process that often requires tissue culture techniques for full plant growth from transformed cells or tissues. In this method, T-DNA is artificially integrated into the genome of host cell using ATMT; therefore, it subsequently passed on to the next generation [17,18]. Transgenic yellow flowers of Torenia hybrida are produced by co-expression of the 4'CGT and AS1 genes along with RNAi-mediated silencing of flavanone 3-hydroxylase (F3H) or dihydroflavonol 4-reductase (DFR) genes [5]. In transformed Ipomoea nil petals, the co-expression of the 4'CGT and AS1 genes induced expression of AOG and intensified the pale yellow color of the primary petals to visible yellow [19]. Genetically engineered production of yellow petals in important ornamental plants such as Geranium, Lathyrus odoratus, Cyclamen and Saintpaulia has not been reported [7,19].
In spite of the presence of various colors among African violet ornamental plant cultivars, it lacks the yellow cultivar. Therefore, due to the popularity of the African violet ornamental plant, the production of this plant with a new yellow color is very valuable commercially and is of considerable importance in the market. Thus, the present study aimed to change the color of African violet flowers from white to new yellow color through genetic manipulation (using both transient and stable transformation) and the simultaneous expression of 4'CGT and AS1 genes involved in aurone biosynthetic pathway without any necessity for silencing of anthocyanin biosynthesis genes (CHI, F3H and DFR).

Transient expression of 4'CGT and AS1 genes in African violet petals
The function of 4'CGT and AS1 in the aurone biosynthesis was verified using the agroinfiltration method (Fig. 2b, c). Three days after infiltration, the color of the petals has changed from white to yellow (pale yellow), while no change was observed in the control group and the petals were injected with agroinfiltration solution without gene structure. Successful infiltration was later confirmed by light microscope. The results clearly showed the color change to yellow in surrounding parts injected with needle (Fig. 2b). Additionally, the presence both of 4'CGT +AS1 genes in infiltrated petals was confirmed by PCR analysis and these two genes were not detected in control plants (non-injected plant and injected plant without gene construct) (Fig. 2c). This experiment was repeated twice and similar results were obtained (data not shown).

Transgenic Plants of African violet by simultaneous expression 4'CGT and AS1
To produce AOG, we inserted 4'CGT and AS1 genes into leaves through ATMT. Two binary vectors: pBI121 to express the 4'CGT gene and pCAMBIA1304 to express the AS1 gene were constructed. A few weeks after the transfermation of gene constructs, small white calli with green buds began to grow on some of the inoculated petiole explants. Totally, 15 putative independent transgenic plants were obtained from a total of 60 transformed explants (Table 1). 20-30% regeneration efficiency was achieved which enabled us to obtain 53 plantlets from 15 regenerated samples (Table 1; Fig. 3D). Regenerated plantlets were screened on Hygromycin and Kanamycin selection (Fig. 3E). Availability of Hygromycin and Kanamycin-resistant gene have successfully confirmed using PCR in putative transgenic plants. All putative transgenic plants acclimated and transferred to greenhouse until maturity (about 16 weeks). In transformants with white-yellow flowers, the area of yellow sectors increased in the white background of the petals during the late phase of flowering. With time, the color of most of the petals turned to yellow in all the late-born flowers of transgenic plants. In contrast, no change was observed in the petals of non-transgenic plants. No morphological difference other than the color change was observed between non-transgenic and transgenic plants ( Fig. 4a-f ).

Expression of 4'CGT and AS1 in transgenic African violet Flowers
We confirmed the expression of the transgenes with PCR. The size of PCR products was 1374 and 1689 bp for 4'CGT and AS1 genes, respectively (simultaneous expression 4'CGT and AS1 in T0 transgenic plants), identical with the positive control. These genes were not detected in non-transgenic plants (Fig. 5).
Two samples from PCR positive transgenic plants ( Fig. 5) were selected for further confirmation by Southern blot assay. For blotting, the researchers used specific 1374 and 1689 bp probes labeled with digoxin molecules. A Single signal was detected for each gene in the blotting of two transgenic African violet genomes, which were positive in PCR analysis, whereas no hybridization signal was observed in the non-transformed plant samples (African violet negative control). The lengths of hybridized fragments were 4000 and 6000 bp for 4'CGT and AS1 genes, respectively (Fig. 6).

Evaluation of the expression pattern of 4'CGT and AS1 genes in transgenic petals
The relative expression levels of 4'CGT and AS1 genes in transgenic plants with high expression levels were further confirmed by qRT-PCR. The amplification efficiency of genes ranged between 98.71% and 99.83% and correlation coefficients varied from 0.977 (4'CGT ) to 0.996 (AS1) (Supplementary Table S1). These results indicated the expression of both genes studied in transgenic petals as compared to Non-transgenic petals. The results showed that the expression level of the studied genes was not naturally present in African violet white (Non-transgenic plant), while in transgenic African violets (yellow), it expressed significantly and was to some extent equal to A. majus (yellow) (Fig. 7).

Microscopic analysis of transgenic petals
To further study the expression of 4'CGT +AS1 genes in petals of transgenic plants, transverse sections of petals were examined by microscopic analysis (Fig. 8). In wild type, pale-yellow pigments (Chalcones) were localized in the adaxial epidermis (L1 layer) and the underlying mesophyll (L2 layer) appeared to be white. In contrast, both L1 and L2 layers were yellow in petals of transgenic plant (Fig. 8). It generally represents the merger of both genes together in yellow transgenic African violet.

Aureusidin 6-O glucoside accumulation in the 4'CGT and AS1 expressing transgenic petals
Flavonoids were extracted from petals of not-transformed African violet, transformed African violet, and A. majus. These metabolites were then analyzed by HPLC. The HPLC chromatogram demonstrated that the yellow flowers of A. majus contained a compound that was not present in non-transgenic African violet ( Fig. 9e) but was observed in transgenic African violet petals (Fig. 9f ). Retention times (RT) in transgenic African violet and A. majus flowers were identical, thereby confirming the presence of AOG ( Fig. 9a; Table 2). The amount of AOG component (Peak 1ʹ, RT 3.9 min) in flowers of the transgenic African violet and A. majus was 2.95 and 3.45 mg/g, respectively ( Table 2). HPLC chromatogram showed that the components were properly separated. MS⁄ MS analysis of peak 1ʹ, which exhibited an [M + H]-ion at m⁄ z 449, yielded MS2 fragmentation at m⁄ z 287 due to the loss of 162 atomic mass units (amu), corresponding to one . Peak 1 is unlikely due to the saturation of peak 1ʹ as the two peaks are well separated and not tailing. In A. majus, a trace amount of molecular ion m⁄ z 465 was revealed to be co-eluted with broad peak 1ʹ, which was fragmented at m⁄ z 287 (data not shown) and tentatively identified as bracteatin-6-O-glucoside. Additionally, the mass spectra and UV⁄ V are features of the peaks 1 and 1ʹ, a compound identified in A. majus and transgenic African violet (Fig. 9d, f ) but not in the wild type (Fig. 9e). These features possibly correspond to a structural isomer of AOG.

Discussion
Genetic engineering supports the idea of changing as well as creating new colors in ornamental plants. This research has been conducted with the purpose of changing flower colors and creating new ones in African violet species. Flower and ornamental plants industry has been trying to develop new cultivars with specific characteristics such as new colors [20]. The development of ornamental cultivars with new flower colors, considering the major purpose of flower and ornamental plants industry, will result in increased economic value [21]. Ornamental cultivars of Petunia, chrysanthemums, Rosa hybrid, Ipomoea nil, Rosa rugose and Nierembergia species with respective colors of light pink, blue-violet, purple, light yellow, yellow and light violet have been produced using the simultaneous expression or overexpression of the corresponding genes [19,[22][23][24][25]. The anthocyanins were the principal pigments leading to the development of new colors including red, pink and blue. This is while aurone compound has been used to develop yellowcolored flowers in recent studies [19,26]. Although African violet can be found in various colors, no color change has been reported through genetic engineering of this precious species. In this investigation, the white-colored petals of S. 'Jolly Diamond' cultivar have been used. Therefore, no CHI is available to turn chalcone into AOG. Chalcone, considered as a key enzyme in flavonoid biosynthesis of flowering plants including aurones, is turned into colorless naringenin and plays an important role in color of ornamental flowers [2,21]. In most plant species, chalcone is not the final product. Combined with other enzymes, it turns into other classes of flavonoids such as aurones, flavonones, dehydroflavonols and finally anthocyanins as a major class of colors [27]. In a molecular analysis investigation, two separate genes of chalcone synthesis (i.e., SaCHSA and SaCHSD) were detected in African violet [28]. It should be noted that the identification of chalcone in this ornamental plant species is highly significant to conduct this research.
For the first time, the simultaneous transfer of AS1 and 4'CGT genes was successfully conducted in African violet using Agrobacterium tumefaciens strain LBA4404 with respective plasmids of pCAMBIA1304 and pBI121. In an investigation carried out on African violet by [29], two strains including A218 and EHA105 were employed. The results indicated that only strain A281 led to more efficient and stable transfer of traits in leafstalk microsamples while strain EHA105 caused no regeneration. In another research by [30], Agrobacterium tumefaciens strains LBA4404 and EHA101 were used in Pink Veil cultivar of the African violet. It was found that strain LBA4404 (pTOK233) had better performance. According to this research and the investigation conducted by Kushikawa et al. [30], LBA4404 can be considered an appropriate strain in transfer of genes to African violet.
The transient gene transfer enables investigating the genetic construct. Besides, the flower color change can be evaluated without the need for tissue culture. The transgenic plants can also be selected at lower cost and time if this technique is applied [12,31,32]. This method, showing high transfer yield in a wide range of plant species, has been employed to study the African violet petals. To this end, agrobacterium containing gene construct has been injected into the base of petals. Three days later, the phenotypic evaluation of color change of petals as well as the existence of 4'CGT and AS1 genes in the transgenic petals were verified using light microscope and PCR test. According to the results, the S. 'Jolly Diamond' cultivar, having white petals, is suitable for genetic modification and transfer of genes involved in the aurones biosynthesis pathway with the purpose of changing the flower color to yellow. Other researchers have also used the transient gene transfer method. Nazari et al [33] used this method to transfer monogenic (Viola-F3ʹ5ʹH gene) and constructs (Viola-F3ʹ5ʹH gene and Iris-DFR gene). After injection of gene constructs, they conducted to produce delphinine anthocyanin in petals of Gerbera flower (up to 44 and 75%, correspondingly), resulted in color change of petals into blue. In a similar research, the transfer of gene construct containing F3ʹ5ʹH and DFR gene from respective Petunia hybrida and Iris to the petals of Gerbera jamesonii caused a change in the level of anthocyanins among the injected petals and turned the stigma and pollen into blue [34]. Shang et al [35] [36]. The production of transgenic plants is an expensive and time-consuming process. According to the literature, transient transfer is a fast, simple and economic method to test and activate constructs prior to their mass production as gene constructs are most likely transferred following their verification.
In the stable gene transfer method, a certain part of DNA containing a new gene or a combination of multiple genes is artificially inserted into an organism's genome using laboratory methods [37]. Due to the complex nature of flavonoid biosynthesis pathway, the mere transfer of a gene to the plants might not have any specific effects on the biosynthesis pathways of flower's pigments. Therefore, the genetic engineering of flower color has been investigated via transferring of multiple genes [38]. In such research, the white-colored petals of African violet were turned into yellow using two vectors (pBI121 and pCAMBIA1304 vectors for 4'CGT and AS1 gene expressions, respectively) without any need for silencing anthocyanin biosynthesis pathway genes (CHI, DFR and F3H). The results of this research are similar to those of To and Wang [26]. In that research, the CHI and DFR genes corresponding to Petunia' cDNA pattern were cloned in pBI121 and pCAMBIA1304 carriers, respectively. By transferring them to tobacco via agrobacterium, different color patterns were developed compared to non-transgenic tobacco.
For the first time, the relative expression of both genes transferred to the petals was observed and verified in this research. While no analyses (qRT-PCR, Southern blotting and light microscope) were carried out to verify the results observed in previous investigations, the above-mentioned analyses have been conducted to demonstrate the transfer and integration of genes  According to the qRT-PCR analysis, 4'CGT and AS1 genes were successfully expressed in the African violet petals. This is while nontransgenic African violet plants lack these two genes. To verify the integration process of the transferred genes to the genome of genetically-modified plants as well as determining the number of gene transcripts, the Southern blotting test was applied. Findings showed that the genetically-modified plants possessed one or two transcripts of the transferred genes. By investigating the genetically-modified plants having 4'CGT and AS1 genes sequencing, the band lengths were determined 4500 and 6000 bp, respectively. In a recent research, HPLC analysis was used to verify and quantify these genes in the transgenic Ipomoea nil ornamental plant [19].
In another investigation, the 4'CGT and AS1 genes were separately transferred from A. majus to the bluecolored petals of Petunia. According to the results, 4 out of 9 transgenic SRY4'CGT plants developed bluewhite flowers. Moreover, the proportion of whitecolored sectors gradually increased until the petal color turned completely white. In contrast, 3 out of 13 transgenic SRYAS1 developed blue-white sectors. However, the amount of transgenic flowers did not increase and no complete white flowers were observed. Besides, the chalcone synthase significantly increased in the bluecolored sector while a decrease was observed in the white-colored Sect. [39]. In the blue-colored petals of Torenia hybrida ornamental plant, expression of these genes along with silencing DFR or F3H gene led to the development of yellow color. In an investigation by the RNAi technique, no change was reported in the appearance of flower [5]. Recently, the transfer of these genes to the petals of Ipomoea nil ornamental plant changed the color of flower from light yellow to yellow. However, most transgenic plants contained light brown-colored unbloomed flowers which might have consisted of necrotic cells [19]. In contrast, this research led to the production of 15 yellow-colored transgenic plants out of 60 explants transferred by ATMT. Meanwhile, there was no need for gene silencing (RNAi) and no other morphological difference was observed between transgenic and non-transgenic plants.
The flowers of yellow-colored A. majus cultivar produce a considerable amount of AOG and bracteatin 6-O-glucoside along with smaller quantities of the 4'-glucosides of THC and PHC [40,41]. According to the HPLC analysis conducted on transgenic Torenia hybrida ornamental plant, the AOG compound was produced in the transgenic plants (0.422 mg/g) while there was no trace of such compound in non-transgenic [5]. The enzymatic formation of aurones observed in the yellow-colored A. majus flowers extracts indicated that the main pigments pathway to produce AOG was open followed by reduction in PHC-glucoside, bractatin-6 glucoside, THC-4 glucoside, respectively. In this research, the HPLC-DAD-MS chromatogram revealed the existence of AOG compound in the transgenic African violet plants containing yellow-colored petals. This is while no such compound is produced in the white-colored petals of non-transgenic African violet plants naturally.

Conclusions
For the first time, the genetic engineering of aurone pigment biosynthesis pathway led to the production of yellow color in the white-colored African violet petals. This method was employed by simultaneous expression of 4'CGT and AS1 genes. Furthermore, agroinfiltration system was highly effective to evaluate the performance of genes and color of flowers. In this research, the S. 'Jolly Diamond' cultivar containing white-colored petals was used. The simultaneous expression of 4'CGT and AS1 genes led to the formation of new yellow color. This is while the existence such phenotype was not reported in the African violet. The results of this investigation indicated that the simultaneous expression of these genes in the white-colored petals, containing chalcone, contributed to the accumulation of AOG as the final compound of aurones. As the African violet petals produce chalcone, the existence of malonyltransferase caused accumulation of aurones. The phenotypic and molecular evaluation of integration indicated that these genes, detected in the genome of transgenic petals, were not in a single locus. This observation also demonstrated the production of transgenics by clonal reproduction. Furthermore, yellow pigments and accumulation of AOG were observed in the petals of 4'CGT/AS1 transgenic plants while no change was detected in the petals of wild type. Therefore, transferring these genes to other ornamental plants can lead to the production of yellow-colored petals provided that whitecolored petals contain chalcone to produce AOG inside the vacuoles. Moreover, the simultaneous expression of 4'CGT and AS1 was conducted without the need for silencing genes to reduce the anthocyanin biosynthesis adjustment. Thus, the mere application of these genes is applicable to change the petals' color from white to yellow. Besides, this method can be considered an appropriate strategy to change the color of white petals as well as producing new color in the flowers of ornamental plant species.

Vectors Construction
To clone the full-length cDNA of 4'CGT and AS1 genes, RNA from yellow petals of A. majus was isolated by TRIzol protocol as previously described [42]. Specific primer sets P1 and P2 (Supplementary Table S2) were used to amplify full-length reading frames of 4'CGT (1374 bp) and AS1 (1689 bp), respectively. PCR products were then analyzed by 1% agarose gel electrophoresis. After purification, PCR products were cloned into Pet28a Easy vector (Promega) and then sequenced by an automatic DNA sequencer (Sanger Sequencing method, ABI 3730 XL). The full-length cDNA of the 4'CGT gene was cloned into the pBI121 binary vector using SacI and BamHI restriction sites. AS1 gene was cloned into pCAMBIA1304 vector by NcoI and BstEII restriction sites, respectively (Supplementary Table S2; Fig. 2a). All these vectors were verified by restriction enzyme cutting and sequencing.

Agrobacterium-Mediated Transformation
Agrobacterium tumefaciens (strain LBA4404) provided from of the Institute ABRII (Agricultural Biotechnology Research Institute of Iran) was used to carry recombinant binary vectors and Agrobacterium-mediated plant transformation assay. For transformation, the researchers used the electroporation (BTX ™ ECM ™ 630, USA) at 25 µF, 400 Ω, and 1.25 kV. 1 ml of LB medium was then added immediately into the mixture, incubated at 28˚C for 3 h. 100 µl of the following bacterial mixture was transferred onto selective LB medium (containing 50 mg. L −1 Kanamycin and 100 mg.L −1 Rifampin) as described in the manufacture's protocol.

Plant materials and growth conditions
African violet (S. Jolly Diamond), was used in the present study. All Saintpaulia genotypes were provided from the genotype collection of the Institute GABIT (Genetics and Agricultural Biotechnology Institute of Tabarestan) and the plant material was maintained in greenhouses under controlled conditions. For plant transformation, the leaves were sterilized with 1% sodium hypochlorite for 20 min, washed thoroughly with sterile water, and cultured to germinate on MS basal medium (pH 5.7) containing MS salts [43] 2% sucrose, 0.8% bacto-agar, 1 mg. L −1 6-benzyl amino purine (BA), and 1 mg.L −1 indole-3-butyric acid (IBA). The cultures were then incubated in a growth chamber under the condition of 25±1ºC, a 16-h florescent light (100 µmol/m2/sec), and a 8-h dark cycle. 8 weeks later, young leaves were collected and used for the transformation test.

Transient expression via infiltration of petal
African violet flowers were infiltrated with Agrobacterium tumefactions strain LBA4404 harboring both pBI121 and pCAMBIA1304. For this, the LBA4404 cells were cultured in 10 ml LB broth with antibiotics overnight, pelleted and re-suspended in a medium #1003 (AB media salts + NaH2PO4 240 mg.L −1 + glucose 10 g/l + MES 14.693 g/l) supplemented with 100 µM acetosyringone and cultured for 4 h [35]. The cells were then pelleted and re-suspended to a concentration of A600 = 0.5 in 1% (w/v) glucose solution (PH 5.3) supplemented with 100 µM acetosyringone [35]. Flower buds or opened flowers were pierced with a needle and infiltrated with the Agrobacterium culture using a syringe. Three days after injection and subsequent change in specimen's conditions, the injected petals (both with and without gene construct) were separated from non-injected ones. After DNA extraction, the existence of two genetic parts was examined using PCR test. To further investigate the color change of petals, the cross section of the injected petals (with and without gene construct) was evaluated using a light microscope.

Stable transformation
A single colony of plasmid-carrying Agrobacterium was picked up and incubated in 10 ml LB medium containing 50 mg.L −1 Kanamycin (Km) and 100 mg.L −1 Rifampin (Rif ) at 28˚C overnight. Bacteria cells (OD 600 -0.5 -0.6) were then washed and resuspended in AB medium (

CAPS analysis DNA extraction and PCR
Total DNA was extracted from fresh petal of putative transgenic plants with CTAB method as previously described [44]. For PCR analysis, p1 and p2 specific primers were used for 4'CGT and AS1 genes (Supplementary

Quantitative realtime PCR (qRTPCR) for the expression analysis of 4'CGT and AS1 genes
For real-time quantitative PCR, total RNA was extracted from African violet white (non-transformation), transgenic A. violet, and A. majus petals all treated with DNase I, as previously described by wang et al. [45]. Next, cDNA was synthesized from total RNA (100 ng) using the Superscript III First-Strand Synthesis System (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) and oligo (dT) 20. The transcript levels of 4'CGT and AS1 were analyzed via RT-qPCR (Applied Biosystems 7900HT Fast Real-Time PCR System) using Power SYBRTM Green PCR Master Mix (Applied Biosystems, Thermo Fisher Scientific) according to the manufacturer's instructions. Transcript levels were calculated based on ∆∆Cq (formerly ∆∆Ct) method [46] using Actin (accession number: AB596843.1) gene as references. Statistical significance of the differential expression levels was assessed as independent experiments (with mean centering and autoscaling) [47]. The Tukey-Kramer test at the 1% level was used for the analysis of aurone biosynthesis-related genes. The results are presented as the standardized mean of SE.

Light Microscopy analysis
The morphology of regenerates transgenic and nontransgenic petals were examined under a VH-Z75 light microscope. Petals were cut into 5 × 5 mm pieces and embedded in 4% agarose. Thin sections (thickness, 100 mm) were cut with a Micro Slicer DTK-1000 (D.S.K.). Sections were examined using a VH-Z75 light microscope (Keyence).

Aureusidin 6-O glucoside identification by HPLC-DAD-MS n
The African violet white, transgenic African violet, and A. majus petals (12 mm) were evaluated separately. Aurone analyses were performed using an Agilent HPLC series 1200 equipped with ChemStation software, a degasser, quaternary pumps, autosampler with chiller, column oven, and diode array detector. The guard column operated at a temperature of 35 C. The mobile phase consisted of 0.1% TFA⁄water (eluent A) and 90% acetonitrile in 0.1% TFA⁄water (eluent B) at a flow of 0.8 mL⁄min using the following gradient program: 20% B (0-3 min); 20-60% B (3-20 min); 60% B isocratic (20-27 min); 60-90% B washing step (27-30 min); and equilibration for 10 min. The total run time was 40 min. The injection volume for all samples was 10 L. Specific wavelengths were monitored separately at 400 nm for aurone and 360 nm for flavones. Additionally, UV⁄Vis spectra were recorded at 520 nm for anthocyanins. The HPLC system was coupled online to a Bruker (Bremen, Germany) ion trap mass spectrometer fitted with an ESI source. Data acquisition and processing were performed using Bruker software. The mass spectrometer was operated in positive ion mode and auto MSn with a scan range from m⁄z 100 to 1000. HPLC-grade acetonitrile, water, trifluoroacetic acid (TFA), naringenin, and chalcone standards were purchased from Sigma (St. Louis, MO). All standards were prepared as stock solutions at 10 mg⁄mL in methanol and diluted in water except for chalcone, which was prepared in 50% methanol. UV external standard calibration was also used to obtain calibration curves of cyanidin-3-O-glucose, naringenin-7-O-rutinoside, and chalcone, which were used to quantify anthocyanins, flavones, and chalcones, respectively. UV external calibration of maritime was employed for the quantitation of AOG.