Screening of Genes Important for Artificial Accumulation of 6′-Deoxychalcone in Nicotiana benthamiana

Abstract

6′-Deoxychalcone (isoliquiritigenin and butein) is a yellow flavonoid pigment that is promising target for molecular breeding of yellow flowers. In our previous study, it was suggested that co-overexpression of CaMYBA, a positive anthocyanin MYB transcription factor in pepper (Capsicum annuum) with a dahlia (Dahlia variabilis) aldo-keto reductase (AKR) DvAKR1 and snapdragon (Antirrhinum majus) chalcone 4′-O-glucosyltransferase (4′CGT) Am4′CGT was sufficient to accumulate isoliquiritigenin. However, CaMYBA expression induced not only accumulation of 6′-deoxychalcone, but also that of delphinidin-based anthocyanin. Since this anthocyanin accumulation is unnecessary for molecular breeding of yellow flowers, it is necessary to identify genes important for the accumulation of only 6′-deoxychalcone. In this study, we conducted comparative RNA-seq analysis between co-overexpressing CaMYBA, DvAKR1, and Am4′CGT (referred as the CaMYBA combination) and co-overexpressing NtAN2, DvAKR1, and Am4′CGT (referred as the NtAN2 combination) Nicotiana benthamiana leaves. The CaMYBA combination successfully induced isoliquiritigenin, while the NtAN2 combination failed to do so. Because the NbCHS2 gene was detected as a differentially expressed gene, CHS genes from N. benthamiana, pepper, and dahlia were co-overexpressed with DvAKR1 and Am4′CGT. Our results reconfirm that CHS plays an important role in the accumulation of isoliquiritigenin, but the detected peak of isoliquiritigenin was small, also indicating that other genes are required for the abundant accumulation of isoliquiritigenin. We also analyzed NbPAL4, which was more highly expressed in the CaMYBA combination than in the NtAN2 combination, but our results indicated that NbPAL4 is not essential for isoliquiritigenin accumulation. Finally, we investigated why the NtAN2 combination did not work for isoliquiritigenin and anthocyanin accumulation. RNA-seq analysis indicated one bHLH transcription factor was down-regulated in the NtAN2 combination, so we co-overexpressed a dahlia bHLH transcription factor DvIVS with the NtAN2 combination. Isoliquiritigenin accumulation was successfully detected suggesting that the failure of isoliquiritigenin accumulation in the NtAN2 combination is due to the weak ability of NtAN2 to activate bHLH transcription factors.

Introduction

Flower color is one of the most important characteristics for ornamental plants. Yellow flower color is conferred by carotenoids, betaxanthin, and flavonoids in many species. Among these, carotenoids are major yellow pigments in most species, so plant species that cannot accumulate carotenoids rarely produce yellow flowers. On the other hand, flavonoids are secondary metabolites that can be synthesized in most plant species. Because yellow flavonoids induce yellow flowers in some limited species, these yellow flavonoids are promising targets for molecular breeding of yellow flowers.

6′-Deoxychalcone is a yellow flavonoid pigment important for yellow coloration of dahlia (Dahlia variabilis) flowers, and dahlias accumulate isoliquiritigenin and butein derivatives such as butein 4′-malonylglucoside, butein 4′-sophoroside, and isoliquiritigenin 4′-malonylglucoside in their flowers (Harborne et al., 1990; Ohno et al., 2021; Price, 1939). For the biosynthesis of butein derivatives, isoliquiritigenin is first synthesized from the common flavonoid precursors 3-malonyl-CoA and 4-coumaroyl-CoA by chalcone synthase (CHS) and aldo-keto reductase (AKR) (Ohno et al., 2022), and then isoliquiritigenin is converted to butein by chalcone 3-hydroxylase (CH3H) (Schlangen et al., 2010) (Fig. 1). Both isoliquiritigenin and butein receive glycosylation by butein 4′-O-glucosyltransferase (4′BGT) or chalcone 4′-O-glucosyltransferase (4′CGT) (Maruyama et al., 2024).

Fig. 1

6ʹ-Deoxychalcone and anthocyanin biosynthesis pathway. Abbreviations: PAL, phenylalanine ammonialyase; C4H, cinnamate 4-hydroxylase; 4CL, 4-coumaroyl-CoA ligase; CHS, chalcone synthase; EFP, enhancer of flavonoid production; CHI, chalcone isomerase; F3H, flavanone 3-hydroxylase; F3ʹ5ʹH, flavonoid 3′,5′-hydroxylase; DFR, dihydroflavonol 4-reductase; ANS, anthocyanidin synthase; GST, glutathione S-transferase; GT, glycosyltransferase; AKR, aldo-keto reductase; CH3H, chalcone 3-hydroxylase; 4ʹCGT, chalcone 4ʹ-glucosyltransferase.

In our previous study, agroinfiltration of DvAKR1 and snapdragon (Antirrhinum majus) Am4′CGT was not sufficient to induce artificial isoliquiritigenin accumulation in Nicotiana benthamiana, but agroinfiltration of CaMYBA, a positive anthocyanin MYB transcription factor of pepper (Capsicum annuum) (Borovsky et al., 2004), with DvAKR1 and Am4′CGT was sufficient (Ohno et al., 2022). From the perspective of the biosynthetic pathway, ectopic expression of AKR and 4′CGT may be sufficient since endogenous CHS is expressed in N. benthamiana leaves. However, the fact that CaMYBA was required for the accumulation of 6′-deoxychalcone indicates that other genes are required for the artificial accumulation of 6′-deoxychalcone. In addition, ectopic CaMYBA expression induced not only accumulation of 6′-deoxychalcone, but also that of delphinidin-based anthocyanin (Ohno et al., 2022). Since this anthocyanin accumulation is unnecessary for molecular breeding of yellow flowers, it is necessary to identify genes important for the accumulation of only 6′-deoxychalcone.

In our preliminary experiments, when NtAN2 from tobacco (Nicotiana tabacum), a gene orthologous to CaMYBA, was used instead of CaMYBA, neither 6′-deoxychalcone nor delphinidin was detected. Both CaMYBA and NtAN2 are positive anthocyanin regulators and belong to the same AN2 sub-group (Borovsky et al., 2004; Pattanaik et al., 2010). However, tobacco plants accumulate cyanidin-based anthocyanins and primarily lack delphinidin, suggesting that CaMYBA and NtAN2 regulate different sets of structural genes. We hypothesized that genes regulated by CaMYBA, but not by NtAN2, are involved in accumulation of 6′-deoxychalcone, so we conducted comparative RNA-seq analysis to screen the genes important for artificial accumulation of 6′-deoxychalcone in N. benthamiana.

Materials and Methods

Vector construction and agro-infiltration

Total RNA was extracted using Sepasol RNA I Super G (Nacalai Tesque, Inc., Kyoto, Japan) and purified with High-Salt solution (Takara Bio Inc., Kusatsu, Japan). cDNA was synthesized using ReverTra Ace (TOYOBO Co., Ltd., Osaka, Japan) and an oligo dT primer. NtAN2 cDNA (FJ472650) was amplified from an N. tabacum flower. NbCHS1 (LC875493), NbCHS2 (Nbe.v1.s00050g24620), and NbPAL4 (LC875494) cDNAs were amplified from agro-infiltrated leaves of N. benthamiana. DvAKR1 (LC671883) cDNA was isolated from dahlia ‘Shukuhai’ ray florets (Ohno et al., 2022), DvCHS1 (AB576660) and DvCHS2 (AB591825) cDNAs were isolated from dahlia ‘Yuino’ ray florets (Ohno et al., 2011b), and DvIVS (AB601005) cDNA was isolated from dahlia ‘Michael J’ ray florets (Ohno et al., 2011a). CaMYBA (LC473089), CaCHS1B (XM_016710598) and CaCHS2 (NM_001325005) cDNAs were isolated from pepper ‘Peruvian Purple’ flowers (Ohno et al., 2020). Am4′CGT (AB198665) cDNA was isolated from a yellow snapdragon flower (Ono et al., 2006). The cDNAs of CaMYBA, NtAN2, DvAKR1, Am4′CGT, NbCHS1, NbCHS2, NbPAL4, CaCHS1B, CaCHS2, DvCHS1, DvCHS2, and DvIVS, were first subcloned into a pDONR221 vector (Invitrogen, Carlsbad, CA, USA) and then introduced into a pGWB2 Gateway binary vector (Nakagawa et al., 2007). All constructs were transformed into the Agrobacterium tumefaciens strain EHA105 by an electroporation method. Primers used for vector construction are shown in Table S1.

The agro-infiltration procedure was according to Ohno et al. (2022). In summary, transgenic A. tumefaciens were co-infiltrated to 3–4 week-old seedling leaves of N. benthamiana plants. For co-infiltration, each Agrobacterium suspension with 0.2–1.0 cell density at OD₆₀₀ was mixed in equal ratios before infiltration. The tomato bushy stunt virus (TBSV) p19 silencing suppressor expressed in the pDGB3alpha2_35S:P19:Tnos (GB1203) vector (addgene) was also mixed before infiltration. N. benthamiana plants were grown in an incubator (MIR-554; PHC, Tokyo, Japan) with LED lighting under 25°C. Leaves used for pigment extraction and RT-PCR were sampled between the fifth and the seventh day post infiltration. Samples were obtained from three biological replicates consisting of at least three different leaves for each experiment. Primers used for confirmation RT-PCR are shown in Table S2.

High performance liquid chromatography (HPLC) analysis

HPLC analysis was performed according to Ohno et al. (2022). Three leaves of Agro-infiltrated N. benthamiana plants were ground in liquid nitrogen. One mL solution consisted of 5% acetic acid and 50% methanol was added. Then, the mixture was centrifuged at 4°C for 15 min at 15,000 rpm, and the supernatant was collected. Extracted solutions were evaporated and redissolved in hydrochloric acid, water, and methanol in a ratio of 5%:45%:50%. Each sample was diluted five times with the same solvent then boiled at 100°C for 2 hours to hydrolyze it. The pigment contents were quantified using HPLC. The analysis was performed using an HPLC Shimazu series, SCL-10AVP, SPD-M10AVP, CTO-10AVP. SIL-10ADVP, LC-10ADVP, FCV-10ALVP, and DGU-14A (LCsolutions software; Shimazu Corp., Kyoto, Japan) with a C18 column (Nihon Waters K.K., Tokyo, Japan) maintained at 40°C. The detection wavelength was 530 nm for anthocyanin and 380 nm for isoliquiritigenin. Isoliquiritigenin (Tokyo Chemical Industry Co., Tokyo, Japan) and delphinidin chloride (Nagara Science, Gifu, Japan) were used as standard chemicals.

RNA-sequencing analysis and de novo assembly

Total RNAs were extracted from five leaves each of co-overexpressing CaMYBA, DvAKR1, Am4′CGT, and p19 (referred as the CaMYBA combination) plants and co-overexpressing NtAN2, DvAKR1, Am4′CGT, and p19 (referred as the NtAN2 combination) plants using Sepasol RNA I Super G (Nacalai Tesque, Inc.) and purified with High-Salt solution (Takara Bio Inc.). Five RNA samples for each combination were mixed equally, and the mixed RNA samples were sequenced using the Illumina NovaSeq platform with a 101-bp paired-end. A total of 54,000,358 and 44,112,342 raw reads were obtained from the CaMYBA combination and the NtAN2 combination, respectively. Data from both libraries were trimmed with Trimmomatic, and the trimmed reads from the CaMYBA combination were de novo assembled by Trinity (Grabherr et al., 2011). Each set of RNA-seq data was mapped to the de novo assembled CaMYBA combination transcriptome data as a reference sequence using bowtie2 software, and quantification of the transcript per million (TPM) score was calculated using Salmon software v.1.2.1 (Patro et al., 2017).

cDNA sequences of flavonoid biosynthesis-related genes were identified from de novo assembled transcriptome data by local blast analysis. Query sequences were Am4′CGT (AB198665), AtABCC2 (NM_129020), AtTT12 (AJ294464), CaMYBA (AJ608992), DvAKR1 (LC671884), Nt4CL (D43773), NtANS2 (NM_001325254), NtC4H (MW260510), Pa3GT (Peaxi162Scf00050g00423.1), Pa5GT (Peaxi162Scf00016g02128.1), PhAN1 (AF260919), PhAN11 (U94748), PhCHIA (X14589), PhCHSA (S80857), PhDFR (AF233639), PhEFP (AB543054), PhF3H (X60512), PhF3′H (AF155332), PhF3′5′H (Z22545), PhFLS (Z22543), PhGST (Y07721), PhJAF13 (AF020545), and PhPH5 (DQ334807).

Data availability statement

The data that support the findings of this study are openly available in the Genbank and Sequence Read Archive under the accession number PRJDB20472 (https://www.ncbi.nlm.nih.gov/bioproject/PRJDB20472). Accession numbers: RNA-seq: CaMYBA combination (DRR656879); NtAN2 combination (DRR656880).

Results

The CaMYBA combination could induce isoliquiritigenin accumulation, but the NtAN2 combination could not do so

We initially attributed the failure of isoliquiritigenin accumulation by DvAKR1 and Am4′CGT to be due to the incredibly low overall flavonoid biosynthetic activity in agro-infiltrated N. benthamiana leaves. Therefore, we used anthocyanin MYB transcription factors, CaMYBA or NtAN2, in addition to DvAKR1 and Am4′CGT, to enhance flavonoid biosynthetic activity. Unexpectedly, isoliquiritigenin was detected in the CaMYBA combination, but not in the NtAN2 combination (Fig. 2). We also tested for expression of Petunia hybrida Enhancer of Flavonoid Production (PhEFP), an enhancer of flavonoid production by binding to CHS to enhance 2′,4,4′,6′-tetrahydroxychalcone production (Morita et al., 2014; Waki et al., 2020), with DvAKR1 and Am4′CGT, but again no isoliquiritigenin was detected (Fig. 2). Focusing on the contrasting results between the CaMYBA and NtAN2 combinations, we performed comparative RNA-seq analysis to screen genes important for artificial accumulation of 6′-deoxychalcone.

Fig. 2

Effect on isoliquiritigenin accumulation of transient overexpression of CaMYBA, NtAN2 or PhEFP in N. benthamiana leaves. (a) RT-PCR validation of the introduced genes. 1: CaMYBA × DvAKR1 × Am4′CGT × p19, 2: NtAN2 × DvAKR1 × Am4′CGT × p19, 3: PhEFP × DvAKR1 × Am4′CGT × p19. (b) HPLC chromatograms of over-expressed leaves and the isoliquiritigenin standard analyzed at 380 nm.

Differentially expressed genes between the CaMYBA combination and the NtAN2 combination

RNA-seq data of the CaMYBA combination was de novo assembled and used for comparison reference sequences. Fifty-two contigs showed more than 50-fold higher expression in the CaMYBA combination than in the NtAN2 combination, and 19 of them were flavonoid-related genes (Table S3). This result indicated that the flavonoid biosynthetic pathway was not activated in the NtAN2 combination. Therefore, we focused on flavonoid biosynthesis-related genes. Most flavonoid biosynthetic genes were more highly expressed in the CaMYBA combination than NtAN2 combination (Table 1). For example, NbCHS2, NbF3H, NbF3′5′H, NbGST, and Nb3GT were expressed more than 100 times higher in the CaMYBA combination than in the NtAN2 combination. All flavonoid biosynthetic pathway genes were upregulated in the CaMYBA combination, indicating CaMYBA functioned as a very strong anthocyanin regulator, while NtAN2 did not. We also analyzed endogenous genes which share high homology with DvAKR1 and Am4′CGT. For AKR, two contigs, TRINITY_DN15473_c0_g1_i3 (Niben101Scf00270g14002.1) and TRINITY_DN55_c0_g1_i9 (Niben101Scf06283g00001.1) were detected. The expression level of TRINITY_DN15473_c0_g1_i3 was similar between the CaMYBA combination and the NtAN2 combination (TPM value was 6.0 and 4.6, respectively), while TRINITY_DN55_c0_g1_i9 was expressed at a 71 times higher level in the CaMYBA combination than in the NtAN2 combination (Table S3). In a putative amino acid sequence, the identities of TRINITY_DN15473_c0_g1_i3 and TRINITY_DN55_c0_g1_i9 to DvAKR1 were 71% and 69%, respectively. Based on phylogenetic analysis, TRINITY_DN15473_c0_g1_i3 formed a clade with Perilla setoyensis Alcohol Dehydrogenase (AFV99150), while TRINITY_DN55_c0_g1_i9 formed a clade with Arabidopsis thaliana AT1G60710 (OAP19317) and Zea mays AKR2 (PWZ18047), and both located to a different clade from DvAKR1 (data not shown), suggesting these two N. benthamiana AKR homologues are not involved in isoliquiritigenin biosynthesis. For 4′CGT, no contig shared high homology with Am4′CGT. Therefore, considering the biosynthetic pathway of 6′-deoxychalcone (Fig. 1), only CHS is included in this pathway. Two different CHS genes were detected from the transcriptome data, NbCHS1 and NbCHS2, and NbCHS1 expression was not different between the CaMYBA combination and the NtAN2 combination, whereas NbCHS2 expression was drastically different and 7,500-times higher in the CaMYBA combination than in the NtAN2 combination (Table 1). CHS consists of a multigene family, and multiple CHS genes were identified in many species. We hypothesized that some of the multiple CHSs could be involved in isoliquiritigenin accumulation, whereas others could not. To test this hypothesis, we constructed 35S driven CHS vectors of NbCHS1 and NbCHS2 from N. benthamiana, CaCHS1B and CaCHS2 from pepper, and DvCHS1 and DvCHS2 from dahlia and expressed them with DvAKR1 and Am4′CGT solely or in combination. No isoliquiritigenin peak was detected from DvAKR1 × Am4′CGT, while a small isoliquiritigenin peak was detected from NbCHS1 × DvAKR1 × Am4′CGT, NbCHS2 × DvAKR1 × Am4′CGT and NbCHS1 × NbCHS2 × DvAKR1 × Am4′CGT (Fig. 3). Similar results were obtained when CaCHS1B and CaCHS2 or DvCHS1 and DvCHS2 were used instead of NbCHS1 and NbCHS2. These results reconfirm that CHS plays an essential role in the accumulation of isoliquiritigenin. However, compared with the CaMYBA combination, the detected peak was very low, also indicating that another gene is required for the abundant accumulation of isoliquiritigenin.

Table 1

TPM values of genes related to flavonoid biosynthesis.

Fig. 3

Effect on isoliquiritigenin accumulation through transient overexpression of CHS genes in N. benthamiana leaves. (a) RT-PCR validation of the introduced genes. 1: NbCHS1 × DvAKR1 × Am4′CGT × p19, 2: NbCHS2 × DvAKR1 × Am4′CGT × p19, 3: NbCHS1 × NbCHS2 × DvAKR1 × Am4′CGT × p19, 4: DvAKR1 × Am4′CGT × p19, 5: CaCHS1B × DvAKR1 × Am4′CGT × p19, 6: CaCHS2 × DvAKR1 × Am4′CGT × p19, 7: CaCHS1B × CaCHS2 × DvAKR1 × Am4′CGT × p19, 8: DvCHS1 × DvAKR1 × Am4′CGT × p19, 9: DvCHS2 × DvAKR1 × Am4′CGT × p19, 10: DvCHS1 × DvCHS2 × DvAKR1 × Am4′CGT × p19. (b) HPLC chromatograms of overexpressed leaves and the isoliquiritigenin standard analyzed at 380 nm.

Because CHS is a necessary, but not the only factor needed to generate abundant amounts of isoliquiritigenin, we next focused upstream of flavonoid biosynthesis. It is well known that a precursor of flavonoid pigments, 4-coumaloyl-CoA, is synthesized through the phenylpropanoid pathway. It begins with phenylalanine and is catalyzed by the actions of phenylalanine ammonialyase (PAL), cinnamate 4-hydroxylase (C4H) and 4-coumaroyl-CoA ligase (4CL) to form 4-coumaloyl-CoA (Fig. 1). Among them, one PAL gene, NbPAL4, was more highly expressed (87.5 times) in the CaMYBA combination compared to the NtAN2 combination (Table 1; Fig. 4). Therefore, we generated an 35S driven NbPAL4 construct and co-overexpressed it with DvAKR1 × Am4′CGT. However, no isoliquiritigenin peak was detected from NbPAL4 × DvAKR1 × Am4′CGT (Fig. 4c). We also co-expressed NbPAL4 with NbCHS1 and/or NbCHS2, DvAKR1, and Am4′CGT. However, even though a small peak of isoliquiritigenin was detected from NbPAL4 × NbCHS1 × DvAKR1 × Am4′CGT, NbPAL4 × NbCHS2 × DvAKR1 × Am4′CGT and NbCHS1 × NbCHS2 × DvAKR1 × Am4′CGT, the peaks were similar to those observed when CHS was introduced, but were not enhanced by the introduction of NbPAL4 (Fig. 4). These results indicate that NbPAL4 is not essential for isoliquiritigenin accumulation.

Fig. 4

Effect on isoliquiritigenin accumulation of transient overexpression of NbPAL4 and CHS genes in N. benthamiana leaves. (a) RT-PCR of PAL, C4H and 4CL genes in the CaMYBA or NtAN2 combination overexpressed N. benthamiana leaves. (b) RT-PCR validation of introduced genes. 1: NbPAL4 × DvAKR1 × Am4′CGT × p19, 2: NbPAL4 × NbCHS1 × DvAKR1 × Am4′CGT × p19, 3: NbPAL4 × NbCHS2 × DvAKR1 × Am4′CGT × p19, 4: NbPAL4 × NbCHS1 × NbCHS2 × DvAKR1 × Am4′CGT × p19. (c) HPLC chromatograms of overexpressed leaves or isoliquiritigenin standard analyzed at 380 nm. Due to a malfunction of the HPLC equipment, NbPAL4 × NbCHS1 × DvAKR1 × Am4′CGT × p19 had a slightly early retention time, so we confirmed this using a photodiode array.

Multidrug and toxin extrusion protein (MATE) and ATP-binding cassette (ABC) protein are known to be involved in the transport of flavonoid pigments (Klein et al., 2000; Marinova et al., 2007). Although one MATE gene and one ABC gene were detected from the RNA-seq data, no drastic differences were found between the CaMYBA and NtAN2 combinations (Table 1). These results indicated the MATE and ABC genes were not the determinant genes for isoliquiritigenin accumulation between the CaMYBA combination and the NtAN2 combination.

Agroinfiltration of the NtAN2 combination with DvIVS

To investigate why the NtAN2 combination did not cause isoliquiritigenin and anthocyanin accumulation, we analyzed transcription factors. It is well known that MYB, bHLH, and WDR transcription factors form an MBW complex to regulate enzymatic genes. We detected two bHLH transcription factors and two WDR transcription factors from the transcriptome data of the CaMYBA combination. Among them, one bHLH transcription factor (TRINITY_DN1420_c0_g1_i2) that is orthologous to petunia ANTHOCYANIN 1, arabidopsis (Arabidopsis thaliana) Transparent Testa 8, and dahlia DvIVS, was down-regulated (Table 1). We made a 35S driven construct for DvIVS and expressed it with DvAKR1 × Am4′CGT or co-expressed it with the NtAN2 combination. As a result, while isoliquiritigenin was not detected from DvIVS × DvAKR1 × Am4′CGT, abundant amounts of isoliquiritigenin and delphinidin were detected from DvIVS × NtAN2 × DvAKR1 × Am4′CGT (Fig. 5). These results indicated that the low activity of NtAN2 to induce anthocyanin and isoliquiritigenin accumulation was due to low activation of a bHLH transcription factor.

Fig. 5

Effect on isoliquiritigenin accumulation of transient overexpression of the NtAN2 combination with DvIVS in N. benthamiana leaves. (a) RT-PCR validation of the introduced genes. 1: CaMYBA × DvAKR1 × Am4′CGT × p19, 2: NtAN2 × DvAKR1 × Am4′CGT × p19, 3: NtAN2 × DvIVS × DvAKR1 × Am4′CGT × p19, 4: DvIVS × Am4′CGT × p19. (b) HPLC chromatograms of overexpressed leaves analyzed at 380 nm. (c) HPLC chromatograms of overexpressed leaves analyzed at 530 nm.

Discussion

6′-Deoxychalcones are promising compounds for molecular breeding of yellow flowers, but there are challenges in terms of their artificial accumulation. Regarding the artificial accumulation of isoliquiritigenin or isoflavones synthesized via isoliquiritigenin, at least two attempts have been successful. One attempt was by Davies et al. (1998) in which Medicago CHR7 was introduced to the petunia ‘Mitchell’. Although de novo accumulation of butein 3-glucoside and a decrease in flavonol were detected, the color of the limbs and tubes of the flowers was still pale yellow, indicating that only a limited amount of 6′-deoxychalcones had accumulated. Another trial was that of Yu et al. (2000) that investigated production of isoflavones in non-legume species by introducing soybean isoflavone synthase (IFS) 1 to maize cells. In this study, isoflavone was not detected by sole expression of IFS1, but was detected when IFS1 was co-expressed with CRC, a chimeric transcription factor containing maize C1 (MYB) and R (bHLH) coding regions. These papers suggested that the introduction of AKR or IFS genes alone is not sufficient for the accumulation of isoliquiritigenin or isoflavones, and that other factors such as transcription factors are required. In fact, introduction of DvAKR1 and Am4′CGT was not sufficient to induce isoliquiritigenin accumulation in N. benthamiana (Ohno et al., 2022; Fig. 2).

Isoliquiritigenin was successfully detected by introducing CaMYBA with DvAKR1 and Am4′CGT in N. benthamiana, indicating one or more genes under regulation of CaMYBA is involved in isoliquiritigenin accumulation (Ohno et al., 2022). Comparing the CaMYBA combination and the NtAN2 combination, several flavonoid-related genes were up-regulated in the CaMYBA combination; however, in the 6′-deoxychalcone biosynthesis pathway, only CHS could be the candidate. CHS is known as a multigene family gene and many species have several copies of CHS. In N. benthamiana, two different CHS, NbCHS1 and NbCHS2 genes were expressed in leaves, and only NbCHS2 was differentially expressed between the CaMYBA combination and the NtAN2 combination. Therefore, we hypothesized that subfunctionalization of CHS could affect isoliquiritigenin biosynthesis, namely NbCHS1 and NbCHS2 have different functions and only NbCHS2 can be involved in isoliquiritigenin biosynthesis. However, all the tested CHSs had similar results with a small isoliquiritigenin peak detected (Fig. 3), indicating whether under the regulation of CaMYBA or not, all the tested CHSs have a similar essential role in isoliquiritigenin accumulation and that another gene is required for abundant accumulation of isoliquiritigenin. Although endogenous NbCHS1 expression was detected in the NtAN2 combination or with DvAKR1 × Am4′CGT, failure of isoliquiritigenin accumulation may be due to insufficient expression of endogenous NbCHS1.

As well as NbCHS2, NbPAL4 was differentially expressed between the CaMYBA combination and the NtAN2 combination. PAL is known to act in the deamination of L-phenylalanine to form trans-cinnamic acid in the phenylpropanoid pathway. Because PAL acts upstream in the flavonoid biosynthesis pathway, it was hypothesized that overexpression of PAL may lead to enhanced substrate flow of flavonoids. However, agro-infiltration of NbPAL4 × DvAKR1 × Am4′CGT could not induce isoliquiritigenin, and agro-infiltration of NbPAL4 × NbCHS1 × DvAKR1 × Am4′CGT, NbPAL4 × NbCHS2 × DvAKR1 × Am4′CGT and NbPAL4 × NbCHS1 × NbCHS2 × DvAKR1 × Am4′CGT could not increase the size of the isoliquiritigenin peaks (Fig. 4), indicating the presence of NbPAL4 does not affect isoliquiritigenin accumulation. Howles et al. (1996) reported that PAL-overexpressing tobacco plants with increased PAL activity showed increased levels of chlorogenic acid, but not flavonol (rutin). Our results also indicated that overexpression of PAL does not affect flux into the flavonoid pathway. We also analyzed gene expressions of flavonoid transporters (MATE and ABC); however, both genes expressed similar levels between the CaMYBA combination and the NtAN2 combination (Table 1). Therefore, since the high accumulation of isoliquiritigenin only in the CaMYBA combination cannot be explained by the existing biosynthetic pathways, indicating unknown factors such as genes required for metabolon formation may be indirectly involved. At least, a gene(s) must be under regulation by an MYB transcription factor, and this gene may be important for molecular breeding of yellow flowers by artificial accumulation of 6′-deoxychalcone.

Compared with the CaMYBA combination, the NtAN2 combination failed to induce isoliquiritigenin (Fig. 5). NtAN2 was isolated as a key gene controlling anthocyanin production in tobacco reproductive tissues (Pattanaik et al., 2010). Comparing the transcriptome data of the CaMYBA and the NtAN2 combinations, not only anthocyanin biosynthetic genes, but also a bHLH transcription factor that is orthologous to petunia ANTHOCYANIN 1, arabidopsis Transparent Testa 8 and dahlia DvIVS were less expressed in the NtAN2 combination (Table 1). We then introduced DvIVS (Ohno et al., 2011a) in addition to the NtAN2 combination genes, and succeeded in inducing isoliquiritigenin and delphinidin (Fig. 5). These results indicate that NtAN2 partially retains transcription factor activity to promote structural genes, but does not function in bHLH activation. Overexpression of the potato (Solanum tuberosum) MYB transcription factor StAN1 induced anthocyanin accumulation in Nicotiana leaves, whereas anthocyanin accumulation was enhanced by co-overexpression of StAN1 and StbHLH1 (D’Amelia et al., 2014). In addition, a hierarchical mechanism has been proposed in Solanaceae plants in which MYB transcription factors first activate bHLH-AN1 transcription factors and then establish a MYB-bHLH-WDR transcription complex to activate flavonoid biosynthetic enzyme genes (Montefiori et al., 2015). Our results suggested that the inability of the NtAN2 combination to induce isoliquiritigenin was due to the weak ability of NtAN2 to activate bHLH transcription factors, and that genes important for isoliquiritigenin accumulation may be regulated by the MBW complex. Further study will elucidate the mechanism for the artificial accumulation of 6′-deoxychalcones.

Acknowledgements

Computations were partially performed on the NIG supercomputer at the ROIS National Institute of Genetics.

Literature Cited

•  Borovsky,  Y.,  M.  Oren-Shamir,  R.  Ovadia,  W. D.  Jong and  I.  Paran. 2004. The A locus that controls anthocyanin accumulation in pepper encodes a MYB transcription factor homologous to Anthocyanin2 of Petunia. Theor. Appl. Genet. 109: 23–29.
•  D’Amelia,  V.,  R.  Aversano,  G.  Batelli,  I.  Caruso,  M. C.  Moreno,  A. B.  Castro-Sanz,  P.  Chiaiese,  C.  Fasano,  F.  Palomba and  D.  Carputo. 2014. High AN1 variability and interaction with basic helix-loop-helix co-factors related to anthocyanin biosynthesis in potato leaves. Plant J. 80: 527–540.
•  Davies,  K. M.,  S. J.  Bloor,  G. B.  Spiller and  S. C.  Deroles. 1998. Production of yellow colour in flowers: redirection of flavonoid biosynthesis in Petunia. Plant J. 13: 259–266.
•  Grabherr,  M. G.,  B. J.  Haas,  M.  Yassour,  J. Z.  Levin,  D. A.  Thompson,  I.  Amit,  X.  Adiconis,  L.  Fan,  R.  Raychowdhury,  Q.  Zeng,  Z.  Chen,  E.  Mauceli,  N.  Hacohen,  A.  Gnirke,  N.  Rhind,  F.  Di Palma,  B. W.  Birren,  C.  Nusbaum,  K.  Lindblad-Toh,  N.  Friedman and  A.  Regev. 2011. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat. Biotechnol. 29: 644–652.
•  Harborne,  J. B.,  J.  Greenham and  J.  Eagles. 1990. Malonylated chalcone glycosides in Dahlia. Phytochemistry 29: 2899–2900.
•  Howles,  P. A.,  V. J. H.  Sewalt,  N. L.  Paiva,  Y.  Elkind,  N. J.  Bate,  C.  Lamb and  R. A.  Dixon. 1996. Overexpression of L-phenylalanine ammonia-lyase in transgenic tobacco plants reveals control points for flux into phenylpropanoid biosynthesis. Plant Physiol. 112: 1617–1624.
•  Klein,  M.,  E.  Martinoia,  G.  Hoffmann-Thoma and  G.  Weissenböck. 2000. A membrane-potential dependent ABC-like transporter mediates the vacuolar uptake of rye flavone glucuronides: regulation of glucuronide uptake by glutathione and its conjugates. Plant J. 21: 289–304.
•  Marinova,  K.,  L.  Pourcel,  B.  Weder,  M.  Schwarz,  D.  Barron,  J. M.  Routaboul,  I.  Debeaujon and  M.  Klein. 2007. The Arabidopsis MATE transporter TT12 acts as a vacuolar flavonoid/H⁺-antiporter active in proanthocyanidin-accumulating cells of the seed coat. Plant Cell 19: 2023–2038.
•  Maruyama,  K.,  H.  Yamada,  M.  Doi and  S.  Ohno. 2024. Identification of two 6′-deoxychalcone 4′-glycosyltransferase genes in dahlia (Dahlia variabilis). Planta 259: 114. DOI: 10.1007/s00425-024-04395-1.
•  Montefiori,  M.,  C.  Brendolise,  A. P.  Dare,  K.  Lin-Wang,  K. M.  Davies,  R. P.  Hellens and  A. C.  Allan. 2015. In the Solanaceae, a hierarchy of bHLHs confer distinct target specificity to the anthocyanin regulatory complex. J. Exp. Bot. 66: 1427–1436.
•  Morita,  Y.,  K.  Takagi,  M.  Fukuchi-Mizutani,  K.  Ishiguro,  Y.  Tanaka,  E.  Nitasaka,  M.  Nakayama,  N.  Saito,  T.  Kagami,  A.  Hoshino and  S.  Iida. 2014. A chalcone isomerase-like protein enhances flavonoid production and flower pigmentation. Plant J. 78: 294–304.
•  Nakagawa,  T.,  T.  Kurose,  T.  Hino,  K.  Tanaka,  M.  Kawamurai,  Y.  Niwa,  K.  Toyoda,  K.  Matsuoka,  T.  Jimbo and  T.  Kumura. 2007. Development of series of gateway binary vectors, pGWBs, for realizing efficient construction of fusion genes for plant transformation. J. Biosci. Bioeng. 103: 34–41.
•  Ohno,  S.,  M.  Hosokawa,  A.  Hoshino,  Y.  Kitamura,  Y.  Morita,  K. I.  Park,  A.  Nakashima,  A.  Deguchi,  F.  Tatsuzawa,  M.  Doi,  S.  Iida and  S.  Yazawa. 2011a. A bHLH transcription factor, DvIVS, is involved in regulation of anthocyanin synthesis in dahlia (Dahlia variabilis). J. Exp. Bot. 62: 5105–5116.
•  Ohno,  S.,  M.  Hosokawa,  M.  Kojima,  Y.  Kitamura,  A.  Hoshino,  F.  Tatsuzawa,  M.  Doi and  S.  Yazawa. 2011b. Simultaneous post-transcriptional gene silencing of two different chalcone synthase genes resulting in pure white flowers in the octoploid dahlia. Planta 234: 945–958.
•  Ohno,  S.,  M.  Ueno and  M.  Doi. 2020. Difference in the CaMYBA genome among anthocyanin pigmented cultivars and non-pigmented cultivars in pepper (Capsicum annuum). Hort. J. 89: 30–36.
•  Ohno,  S.,  H.  Yamada,  K.  Maruyama,  A.  Deguchi,  Y.  Kato,  M.  Yokota,  F.  Tatsuzawa,  M.  Hosokawa and  M.  Doi. 2022. A novel aldo-keto reductase gene is involved in 6′-deoxychalcone biosynthesis in dahlia (Dahlia variabilis). Planta 256: 47. DOI: 10.1007/s00425-022-03958-4.
•  Ohno,  S.,  M.  Yokota,  H.  Yamada,  F.  Tatsuzawa and  M.  Doi. 2021. Identification of chalcones and their contribution to yellow coloration in dahlia (Dahlia variabilis) ray florets. Hort. J. 90: 450–459.
•  Ono,  E.,  M.  Fukuchi-Mizutani,  N.  Nakamura,  Y.  Fukui,  K.  Yonekura-Sakakibara,  M.  Yamaguchi,  T.  Nakayama,  T.  Tanaka,  T.  Kusumi and  Y.  Tanaka. 2006. Yellow flowers generated by expression of the aurone biosynthetic pathway. Proc. Natl. Acad. Sci. USA 103: 11075–11080.
•  Patro,  R.,  G.  Duggal,  M. I.  Love,  R. A.  Irizarry and  C.  Kingsford. 2017. Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14: 417–419.
•  Pattanaik,  S.,  Q.  Kong,  D.  Zaitlin,  J. R.  Werkman,  C. H.  Xie,  B.  Patra and  L.  Yuan. 2010. Isolation and functional characterization of a floral tissue-specific R2R3 MYB regulator from tobacco. Planta 231: 1061–1076.
• Price, J. R. 1939. The yellow colouring matter of Dahlia variabilis. J. Chem. Soc. 1017–1018.
•  Schlangen,  K.,  S.  Miosic,  J.  Thill and  H.  Halbwirth. 2010. Cloning, functional expression, and characterization of a chalcone 3-hydroxylase from Cosmos sulphureus. J. Exp. Bot. 61: 3451–3459.
•  Waki,  T.,  R.  Mameda,  T.  Nakano,  S.  Yamada,  M.  Terashita,  K.  Ito,  N.  Tenma,  Y.  Li,  N.  Fujino,  K.  Uno,  S.  Yamashita,  Y.  Aoki,  K.  Denessiouk,  Y.  Kawai,  S.  Sugawara,  K.  Saito,  K.  Yonekura-Sakakibara,  Y.  Morita,  A.  Hoshino,  S.  Takahashi and  T.  Nakayama. 2020. A conserved strategy of chalcone isomerase-like protein to rectify promiscuous chalcone synthase specificity. Nat. Commun. 11: 870. DOI: 10.1038/s41467-020-14558-9.
•  Yu,  O.,  W.  Jung,  J.  Shi,  R. A.  Croes,  G. M.  Fader,  B.  McGonigle and  J. T.  Odell. 2000. Production of the isoflavones genistein and daidzein in non-legume dicot and monocot tissues. Plant Physiol. 124: 781–794.