A Marigold (Tagetes erecta) MADS-Box Transcription Factor, TeSEP4, Regulates Petal Color by Modulating Chlorophyll and Carotenoid Biosynthesis

Abstract

Chlorophyll and carotenoid biosynthesis are essential metabolic processes in horticultural plants, critically influencing the coloration and economic value of fruits and petals. However, the regulatory mechanisms that coordinate the accumulation of both chlorophyll and carotenoids are still poorly understood. In this study, we demonstrated that the transcription factor TeSEP4 modulated petal color in marigolds by simultaneously modulating chlorophyll and carotenoid metabolic pathways. Overexpression of TeSEP4 produced dark-yellow petals, which were associated with enhanced carotenoid biosynthesis and suppression of chlorophyll biosynthesis. In contrast, silencing TeSEP4 triggered a shift toward yellow-green petals by reducing carotenoid accumulation while concurrently increasing chlorophyll content. Transcriptome and qRT-PCR analyses further revealed that TeSEP4 overexpression upregulated key carotenogenic genes such as TePSY1 and TePSY3, while downregulating chlorophyll-related genes, including TeCHLH, TeCHL27-1, and TePORA1. Silencing TeSEP4 expression caused the opposite changes in these genes. These findings reveal a transcriptional factor that coordinates chlorophyll and carotenoid biosynthesis, offering a strategy to simultaneously improve carotenoid content and modify petal color in marigolds.

1. Introduction

The coloration of plant tissues in nature primarily results from the accumulation of chlorophyll, carotenoids, and flavonoids [1,2]. Among these, chlorophyll and carotenoids, known as plastid pigments, are key participants in the process of photosynthesis [3]. They not only confer diverse coloration to plants but also fulfill vital physiological functions [4,5]. Chlorophyll, the primary photosynthetic pigment in plants, is responsible for their green coloration [6]. It is embedded in the thylakoid membranes as pigment-protein complexes, where its role is to capture and transfer light energy during photosynthesis [7,8]. Furthermore, chlorophyll exhibits physiological activities such as antioxidant and antimutagenic properties [8,9]. Carotenoids serve as essential and photoprotective pigments. They enhance photosynthetic efficiency by participating in light-energy transfer and protect photosynthetic tissues by quenching reactive oxygen species (ROS), thereby mitigating photo-oxidative damage [3]. The yellow, orange, and red colors conferred by carotenoids on plant tissues or organs provide vibrant coloration, which aids in attracting pollinators and seed dispersers, thereby facilitating plant reproduction [10]. Furthermore, carotenoids serve as essential nutrients for both humans and animals [11]. For instance, β-carotene can be converted into vitamin A upon ingestion, playing a critical role in maintaining vision, supporting growth and development, and enhancing immune function [12]. Concurrently, carotenoids exhibit antioxidant capacity, thereby lowering the propensity for chronic diseases and contributing positively to human and animal health [11].

The carotenoid biosynthesis pathway is well elucidated in plants [13,14]. The process begins with phytoene synthase (PSY) catalyzing the condensation of two geranylgeranyl pyrophosphate (GGPP) molecules to produce colorless phytoene. Phytoene then undergoes a series of sequential modifications facilitated by phytoene desaturase (PDS), ζ-carotene isomerase (ZISO), ζ-carotene desaturase (ZDS), and carotenoid isomerase (CRTISO), ultimately yielding red lycopene. Lycopene acts as the key branching point, where it is catalyzed by lycopene ε-cyclase (LCYE) and lycopene β-cyclase (LCYB) to form α-carotene and β-carotene, respectively, thereby establishing the distinct α- and β-branch pathways. Subsequently, α-carotene and β-carotene are modified by various enzymes, including hydroxylation catalyzed by HYDB and HYDE, and epoxidation by zeaxanthin epoxidase (ZEP), leading to the production of diverse xanthophylls. Finally, these xanthophyll degradations are cleaved by carotenoid cleavage dioxygenases (CCDs) and 9-cis-epoxycarotenoid dioxygenase (NCED) [15].

The pathway of chlorophyll biosynthesis has been well deciphered, and many enzymes are involved in this process [14,16]. For example, Mg-chelatase (CHLH/CHLM) functions in committing porphyrins to chlorophyll biosynthesis by catalyzing the insertion of a Mg²⁺ into protoporphyrin IX to yield Mg-protoporphyrin IX. Then, Mg-protoporphyrin IX methyltransferase (CHLM) methylates Mg-protoporphyrin IX to produce Mg-protoporphyrin IX monomethyl ester (MPE). CHL27 and PORA catalyze the production of chlorophyllide a from Mg-protoporphyrin IX monomethyl ester (MPE). Chlorophyllide a is then converted to chlorophyll a by the enzyme chlorophyll synthase (CS). Subsequently, chlorophyll a is transformed into chlorophyll b through two oxidation reactions catalyzed by chlorophyll a oxygenase (CAO). Chlorophyll a and chlorophyll b are degraded through a series of reactions catalyzed by key enzymes, including Chlorophyll b reductase (NOL/NYC), STAY-GREEN (SGR), pheophytin pheophorbide hydrolase (PPH), and pheophorbide a monooxygenase (PAO), leading to the production of non-fluorescent chlorophyll catabolites (NCCs) [17,18].

Transcription factors (TFs) regulate developmental and physiological processes in plants by binding to the promoters of target genes and controlling their expression. In horticultural plants, many TFs are known to regulate chlorophyll and carotenoid metabolic pathways, thereby modulating pigment accumulation. For example, in tomato (Solanum lycopersicum), citrus (Citrus reticulata) [19], and kiwifruit (Actinidia deliciosa) [20], MYB TFs have been shown to modulate chlorophyll and carotenoid accumulation by regulating key metabolic genes. It was further demonstrated that tomato SlZHD17 acts as a direct regulator of genes governing both chlorophyll metabolism and carotenoid biosynthesis in fruit [21]. More recently, pepper (Capsicum annuum L.) CaBBX10 and citrus CrWRKY42 have also been reported to coregulate chlorophyll degradation and carotenoid production by directly modulating the expression of respective biosynthetic genes [22,23]. However, the transcriptional regulation of these two biosynthetic pathways in ornamental plants remains poorly understood.

MADS-box TFs, representing a major TF family in plants, regulate diverse processes in plant development and hormone signaling [24]. They feature a highly conserved N-terminal MADS-box domain and are grouped into two main types: type I and type II (MIKC-type), based on their structure [25]. The MIKC-type genes are further subdivided into two lineages: MIKC* and MIKC^C [26]. A series of functionally related MIKC^C-type MADS-box genes, including APETALA1/FRUITFULL (AP1/FUL), AP3/PISTILLATA (AP3/PI), AGAMOUS (AG), and SEPALLATA (SEP) genes, are found to be involved in flowering, floral organ, and plastid development [27,28]. For example, the overexpression of petunia (Petunia hybrida) SOC1 and SOC1-like genes under heat stress promotes chloroplast biogenesis and chlorophyll accumulation in petals [28]. Similarly, ectopic expression of the birch (Betula platyphylla) gene BpMADS in tobacco (Nicotiana tabacum) promotes chloroplast growth and division and enhances photosynthetic activity [29]. Among these MADS genes, SEP genes function as key regulators of diverse processes, including floral development, flowering time, fruit ripening, and pigment biosynthesis [30,31]. Characterization of SEP genes in Arabidopsis, petunia, and rice (Oryza sativa) has shown a high degree of functional conservation in controlling the floral meristem and floral organ development [32,33]. However, SEP genes in many horticultural plant species are also involved in regulating the formation of pigments. In Phalaenopsis orchid (Phalaenopsis equestris), silencing the SEP3-like gene PeSEP3 results in tepal turning into a leaf-like organ and suppression of anthocyanin biosynthesis [34]. Similarly, in petunia, mutation of the SEP3 ortholog fbp2 results in greenish petals and ectopic inflorescences [35]. In gerbera (Gerbera hybrida), silencing of SEP-like genes GRCD4 and GRCD5 leads to the development of a greenish shade in the ligules of ray florets [36]. Additionally, suppression of SEP-like gene PaMADS7 in sweet cherry (Prunus avium) inhibits fruit ripening and delays softening, and slows ABA, sugar, and anthocyanin accumulation during ripening [37]. Independent overexpression of sweet cherry SEP1/2-like gene FaMADS1a also delays fruit ripening and reduces anthocyanin content [38]. Moreover, SEP-like gene SlMADS1 is a negative regulator of fruit ripening, because the ripening of SlMADS1-RNAi fruits is induced, and the carotenoid content is increasing [39]. Conversely, mutating another tomato SEP-like gene, SlMADS-SIN, leads to a reduction in carotenoid synthesis [40]. Recently, a SEP3-like gene, SnMADS37, in black nightshade berries (Solanum nigrum) was found to regulate chlorophyll biosynthesis [41]. These various studies clearly implicate SEP genes in regulating pigment biosynthesis.

Marigold (Tagetes erecta), a member of the Asteraceae family, exhibits a rich diversity of petal colors from white to deep orange-red and is highly valued as an ornamental plant. In addition, marigold petals are rich in carotenoids, especially lutein, making them a primary commercial source for these pigments in multiple industries [42]. Carotenoids constitute vital dietary micronutrients necessary for maintaining human health. Notably, approximately 10% of dietary carotenoids serve as precursors to vitamin A. This vitamin plays a crucial role in vision protection, immune enhancement, the prevention of cancer and cardiovascular diseases, and cognitive enhancement, making it an indispensable component in both human and animal nutrition [43,44]. Furthermore, in floriculture, enriched color intensity and novel hues can increase the market value of ornamental varieties. Therefore, elucidating the mechanisms of carotenoid biosynthesis in marigold petals is of considerable importance. However, research on carotenoid biosynthesis in marigold petals has primarily focused on the biosynthetic structural genes [45,46,47,48], with very limited knowledge about their transcriptional regulation. Our previous study identified a MADS-box gene, TeSEP4, which was primarily expressed in the floral organs [49]. Heterologous overexpression of TeSEP4 in Arabidopsis does not affect flower development or coloration [49]. However, mutation of its homologous gene, MADS-RIN, in tomato impairs lycopene biosynthesis in the fruit, resulting in a yellowish-red phenotype [40,50]. Furthermore, in white-egret flower (Habenaria radiata), a mutation in the SEP-like gene HrSEP-1 results in green flowers [51]. In this study, TeSEP4 expression was higher in the orange, carotenoid-rich petals of the marigold cultivar ‘Lady’ than in the cream-white, low-lutein petals of the ‘Vanilla’ cultivar. To further elucidate the function of TeSEP4 in marigold, we constructed both overexpression and RNAi recombinant vectors and introduced them into marigold plants. We found that TeSEP4 positively regulated carotenoid accumulation and controlled petal color. These findings significantly advance our understanding of the complex transcriptional regulation of carotenoid metabolism in plants.

2. Materials and Methods

2.1. Plant Materials

The marigold cultivars used were ‘Lady’ (deep-orange petals, high lutein) and ‘Va-nilla’ (cream-white petals, low lutein). The ‘Lady’ seeds were provided by Chifeng Seno Horticulture Co., Ltd. (Chifeng, China), and the ‘Vanilla’ seeds by the Thailand AmeriSeed Company (Chiang Mai, Thailand). Plants of these cultivars were grown in the experimental field of Nanjing Forestry University (Nanjing, China; Longitude: 118°48′, Latitude: 32°). When the plants reached the full bloom stage at 90 d post-growth, they were transplanted into a controlled growth chamber with a 16 h light/8 h dark photoperiod and temperatures maintained at 25 °C (day) and 22 °C (night). Following one week of acclimation, the outermost ray florets of inflorescences at different developmental stages were collected from the two marigold cultivars. Samples for each stage were collected from three healthy and vigorous plants per cultivar, with four inflorescences harvested from each plant. For each developmental stage, the outermost ray florets from the same cultivar were pooled, ground, and homogenized for RNA extraction. Based on previous descriptions, inflorescence development in both cultivars was divided into four key stages: S1 (Pre-flowering), with florets tight, packed, and green; S2 (Unopened), as outer florets elongated and started pigment synthesis; S3 (Semi-opened), showing sequential floret elongation and pigmentation outward from the center; and S4 (Full bloom), where complete floret expansion and pigmentation were observed [45].

Marigold cultivar ‘Milestone’ with yellow flowers from Chifeng Seno Horticulture Co., Ltd. (Chifeng, China) was used for genetic transformation experiments. When the transgenic plants had developed 5–6 leaves, both transgenic and wild-type plants were transplanted into a light-controlled growth chamber set at 25 °C and 60% relative humidity. The photoperiod was 16 h of light and 8 h of darkness, with light supplied at an intensity of 12,000 lux.

2.2. RNA Extraction and Gene Expression Quantification

The outermost ray florets were collected from the ‘Lady’ and ‘Vanilla’ cultivars at each of the four defined inflorescence developmental stages. Total RNA extraction from the collected samples was performed with the AG RNAex Pro RNA kit (Accurate Biology, Changsha, China). The first-strand cDNA was synthesized using the Evo M-MLV kit with gDNA Eraser (Accurate Biology, Changsha, China). The quantitative real-time PCR (qRT-PCR) was carried out using PC41-TRUEscript One Step qRT-PCR Kit (Aidlab, Beijing, China). The qRT-PCR was performed with three biological and three technical replicates per sample, using the beta-actin gene as the housekeeping control. The 2^−ΔΔCt method was used to determine the relative expression levels of the target genes [52]. Gene-specific primers used for qRT-PCR are listed in Table S1.

2.3. Vectors Construction and Plant Transformation

The TeSEP4-overexpression vector (35S:TeSEP4) was constructed in previous study [49]. The coding sequence of TeSEP4 was inserted into the V097 vector and subsequently subcloned into the RNAi vector pK7GWIWG2D(II) via the gateway system [53]. For marigold transformation, the recombinant vectors were harbored in Agrobacterium tumefaciens strain EHA105 [54]. All transgenic plants were initially selected on kanamycin-containing MS medium, with genomic integration confirmed by PCR using both gene-specific and universal primers (Table S1). The outermost ray florets were collected from both transgenic lines and wild-type plants at full bloom and were then immediately frozen in liquid nitrogen and maintained at −80 °C until RNA extraction. The outermost ray florets from four fully bloomed inflorescences per transgenic line and wild-type plant were sampled. The expression levels of TeSEP4 in transgenic and wild-type marigold plants were determined by qRT-PCR. Each reaction was carried out in three independent biological replicates. The phenotypic changes in transgenic lines were observed and analyzed. Based on the expression changes in TeSEP4 in the transgenic lines, plants from lines OE#2, OE#8, OE#9, RNAi#1, RNAi#3, and RNAi#5 were selected for subsequent phenotypic and biochemical analyses. To accurately and objectively quantify the petal color phenotype of transgenic plants, the color parameters (L*, a*, and b*) of the ventral petal surface from outermost ray florets at full bloom were quantified. Measurements were performed using a CM-5 spectrophotometer (Konica Minolta, Japan) on florets from wild-type plants and six transgenic lines (OE#2, OE#8, OE#9, RNAi#1, RNAi#3, and RNAi#5) [55]. For each transgenic line and wild-type plant, three inflorescences were used as biological replicates, with measurements taken from three or four petals of the outermost ray florets on each (Tables S2–S7).

2.4. Chlorophyll and Carotenoid Extraction and Analysis

The florets in the full-blooming stage of transgenic lines (OE#1, OE#2, OE#3, RNAi#1, RNAi#2, RNAi#3, and wild-type plants) were collected. For each plant, 10 fully bloomed inflorescences were collected, pooled, ground, and homogenized for chlorophyll and carotenoid extraction. For analysis of chlorophyll content in the transgenic lines, the floret samples were ground into powder and extracted using acetone as previously described [56,57]. Briefly, approximately 0.1 g of fresh outermost floret tissue was ground in 10 mL of 80% (v/v) acetone. After incubation in darkness at 4 °C for 24 h, the homogenate was subjected to centrifugation at 8000× g for 10 min at the same temperature. The supernatant was collected for the spectrophotometric determination of chlorophyll a and b concentrations using the following formulae: Chl a = (12.25 × A663.2 − 2.79 × A646.8); Chl b = (21.50 × A646.8 − 5.10 × A663.2); Total Chl = Chl a + Chl b. The chlorophyll was quantified using a microplate reader (EPOCH-SN, BioTek, Winooski, VT, USA), and data were recorded at 645 nm and 663 nm. Each sample was measured three times.

The carotenoid content of florets from transgenic lines (OE#2, OE#8, OE#9, RNAi#1, RNAi#3, and RNAi#5) and wild-type plants was extracted and measured by HPLC as previously described [45]. Briefly, the samples underwent freeze-drying, followed by grinding into a powder. The powder was subjected to extraction with 1.5 mL of n-hexane/acetone/ethanol (containing 0.01% BHT) by vortexing and centrifugation. The supernatant was collected after each of two consecutive extractions. The pooled supernatants were filtered through a 0.22 μm filter and then analyzed by a UPLC-APCI-MS/MS system. The LC analysis was performed under the following conditions: a YMC C30 carotenoid-specific column (4.6 mm × 150 mm, 5 μm; Wilmington, NC, USA) was used with the column temperature maintained at 28 °C. The mobile phase consisted of solvent A, acetonitrile: methanol (3:1, v/v) containing 0.01% Butylated Hydroxytoluene and 0.1% formic acid, and solvent B, methyl tertbutyl ether with 0.01% Butylated Hydroxytoluene. A gradient elution was applied at a flow rate of 0.8 mL/min: 0% B at 0–3 min, increased to 70% B by 5 min, further increased to 95% B by 9 min, and then ramped back to 0% B at 10–11 min. The injection volume was 2 μL. Carotenoids were identified according to their characteristic absorption spectra and by comparing with authentic standards. Each sample was measured three times. Peak areas for carotene (lycopene, α-carotene, and β-carotene) and xanthophylls (lutein, violaxanthin, zeaxanthin, neoxanthin, and antheraxanthin) were quantified, respectively. A triplicate design was employed for all experiments.

2.5. Transcriptome Sequencing and Bioinformatic Analysis

TeSEP4-overexpression lines exhibited initial phenotypic changes at the unopened inflorescence bud stage relative to wild-type plants. Furthermore, TeSEP4 demonstrated significantly reduced expression in the lutein-rich ‘Lady’ cultivar compared to the low-lutein ‘Vanilla’ cultivar from stages S2 to S4. Since both phenotypic divergence and key gene expression differences were most pronounced throughout these stages, we concluded that S2 to S4 represent the critical period for petal color determination in marigold. The outermost florets from unopened (S2) and fully bloomed (S4) inflorescences of the transgenic line OE#2, which was characterized by the highest TeSEP4 expression and carotenoid content, and of wild-type plants were collected to elucidate the molecular mechanisms underlying petal color variation. For biological replication, the outermost ray florets were harvested from three inflorescences at the S2 and S4 stages from both the transgenic line OE#2 and wild-type plants. The pooled samples were homogenized for total RNA extraction and subsequent transcriptome sequencing (RNA-seq).

RNA-seq was performed on three biological replicates from each plant material. The total RNA extraction and cDNA library construction were performed by Majorbio Bio-pharm Technology Co., Ltd. (Shanghai, China). RNA concentration and purity were determined with a Nanodrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), while integrity was evaluated by agarose gel electrophoresis and quantified as the RNA Integrity Number (RIN) using an Agilent 5300 system (Agilent Technologies, Santa Clara, CA, USA). For library construction, each sample required a total RNA amount of 1 μg, with a concentration ≥ 30 ng/μL, a RIN > 6.5, and an OD260/280 ratio between 1.8 and 2.2. Following mRNA enrichment via Oligo (dT) magnetic beads, the purified mRNA was subjected to a fragmentation buffer to generate short fragments. Reverse transcription was performed using random primers, followed by the generation of double-stranded DNA. The sticky-ended double-stranded DNA fragments were blunt-ended through end-repair, and a 3′ A-overhang was added for ligation to Y-shaped adapters. After ligation, the products underwent purification and size selection. The size-selected fragments were then PCR-amplified and purified to yield the final sequencing library. Subsequently, sequencing was performed on an Illumina HiSeqTM 2500 platform by Majorbio Bio-pharm Technology Co., Ltd. (Shanghai, China) for all qualified libraries. In this study, all RNA-seq libraries were prepared and sequenced in a single batch to avoid technical batch effects.

The in-house filtering software SOAPnuke v2.0.2 was employed to filter the raw reads, producing clean reads. The clean reads were then aligned to the reference genome (GCA_026213115.1) employing the HISAT2/Bowtie2 software package (http://daehwankimlab.github.io/hisat2/, accessed on 12 October 2024). On the Majorbio Cloud Platform (www.majorbio.com, accessed on 19 October 2024), RNA-seq quality was assessed, differentially expressed genes (DEGs) were identified, expression patterns were examined, and gene functional enrichment was analyzed via the Kyoto Encyclopedia of Genes and Genomes (KEGG). The Fragments Per Kilobase of transcript per Million mapped reads (FPKM) metric was employed to calculate gene expression levels. Expression levels were compared between transgenic lines and wild-type plants. DEGs were screened using a cutoff of an adjusted p-value < 0.05 and |log2-fold change| ≥ 1. KEGG pathway enrichment was analyzed via the KOBAS (http://bioinfo.org/kobas, accessed on 26 May 2025) with Fisher’s exact test as the statistical method. To control the false positive rate, the Benjamini–Hochberg (FDR) method was applied for multiple testing correction. KEGG pathways exhibiting an FDR-adjusted p-value of less than 0.05 were identified as significantly enriched. The raw transcriptome reads were deposited into the NCBI Sequence Read Archive (SRA) under the accession number PRJNA1330480.

2.6. cDNA Synthesis and qRT-PCR Validation

To validate the RNA-seq results, we identified and selected DEGs associated with chlorophyll and carotenoid metabolism based on the thresholds |log2FC| ≥ 1, adjusted p-value < 0.05, and FPKM > 20. These selected DEGs were then further verified by qRT-PCR. The outermost florets were collected from unopened and expanded and fully open inflorescences of wild-type and the transgenic line OE#2. The procedures included total RNA extraction, cDNA synthesis, and qRT-PCR, as described above. Gene-specific primers used for qRT-PCR are listed in Table S1.

Selection of key DEGs was performed by requiring a high consistency (R² > 0.8, p-value < 0.05) between their expression levels in the RNA-seq and qRT-PCR datasets. To further identify these key DEGs regulated by TeSEP4, the expressions of these selected DEGs were validated in the RNAi#3 transgenic line (severe phenotype) and wild-type plants. The outermost florets of unopened and fully open inflorescences were collected from both wild-type plants and the transgenic lines RNAi#3. The procedures included total RNA extraction, cDNA synthesis, and qRT-PCR, as described above.

2.7. Statistical Analysis

All statistical analyses were performed using IBM SPSS Statistics 26. With a sample size of at least three (n ≥ 3), all data are shown as the means ± SD, and differences were calculated for significance using one-way ANOVA. Statistical significance (p < 0.05) was assigned using an asterisk for the independent samples t-test. Statistical significance (p < 0.05 or p < 0.01) was assigned using different letters for Duncan’s test. To evaluate the consistency between the RNA-seq and qRT-PCR datasets, the Pearson correlation coefficients and corresponding p-values were calculated by Microsoft Excel 2019 for Windows V10.

3. Results

3.1. TeSEP4 Is Highly Expressed in the Orange Petals

qRT-PCR analysis showed an overall increase in TeSEP4 expression during inflorescence development in both marigold cultivars ‘Lady’ and ‘Vanilla’ (Figure 1a–c). Notably, TeSEP4 expression was consistently higher in ‘Lady’ than in ‘Vanilla’, with significant upregulation peaks concurrent with the S2 and S4 stages (Figure 1c). This expression profile suggests that TeSEP4 might play a role in regulating petal coloration and carotenoid biosynthesis.

3.2. TeSEP4 Regulates the Marigold Petal Color

Previous study indicated that ectopic expression of TeSEP4 in Arabidopsis does not affect flower development and petal color. To investigate the function of TeSEP4, we overexpressed TeSEP4 in marigolds in this study. A total of six transgenic lines overexpressing TeSEP4 were successfully established, as confirmed by PCR and RT-qPCR. Three transgenic lines (OE#2, OE#8, and OE#9) exhibiting significantly higher TeSEP4 expression levels compared to the wild-type plants were selected for further study (Figure 2b). Phenotypic observation revealed that overexpression of the TeSEP4 resulted in dark-yellow petals in marigold, with no effect on floral development (Figure 2a). Determination of petal color parameters demonstrated a significant increase in both redness (a*) and yellowness (b*), with no discernible alteration in lightness (L*) (Figure 2c, Tables S2–S4), suggesting TeSEP4 is involved in petal color.

To further investigate the function of TeSEP4, RNAi vectors were constructed and transformed into marigolds. Four RNAi-TeSEP4 transgenic lines were obtained, three of which exhibited successful silencing of TeSEP4 expression (Figure 3b). These three silenced lines (RNAi#1, #3, and #5) generated yellow-green florets compared to the control plants (Figure 3a). The RNAi-TeSEP4 transgenic lines exhibited a significant reduction in both petal redness-greenness (a*) and yellowness-blueness (b*) values compared to wild-type plants. Consistent with the observations in the TeSEP4-overexpressing lines, the lightness (L*) value was unaffected in the RNAi-TeSEP4 lines (Figure 3c, Tables S5–S7).

3.3. TeSEP4 Regulates the Chlorophyll and Carotenoid Accumulation in Marigold Petals

To elucidate the metabolic mechanisms underlying the petal color changes in transgenic plants, the chlorophyll and carotenoid profiles of wild-type plants and transgenic lines (OE#2, OE#5, OE#9, RNAi#1, RNAi#3, and RNAi#5) were compared (Figure 4 and Figure 5).

TeSEP4-overexpression lines showed a significant reduction in total chlorophyll concentration in florets compared to wild-type plants (Figure 4a). Similarly, the concentrations of chlorophyll a and chlorophyll b were also remarkably lower in TeSEP4-overexpression transgenic lines than in the wild-type plants (Figure 4b). Conversely, the contents of total chlorophyll and chlorophyll a in florets of RNAi-TeSEP4 marigold lines were remarkably higher than in the wild-type plants (Figure 4c,d). Additionally, chlorophyll b content was elevated in the RNAi-TeSEP4 marigold lines, with significant increases observed in the RNAi-TeSEP4#3 and RNAi-TeSEP4#5 lines (Figure 4d).

The HPLC analysis revealed a significant increase in total carotenoid content in TeSEP4-overexpressing lines compared to wild-type plants (Figure 5a). Specifically, the levels of lycopene, α-carotene, β-carotene, antheraxanthin, zeaxanthin, and lutein were all markedly increased (Figure 5b). Among these, lutein content rose by 26.8%, 22.5%, and 18.9% across the three transgenic lines. In contrast, the concentrations of neoxanthin were significantly reduced in OE#2 (Figure 5b). Violaxanthin content was somewhat elevated in the TeSEP4-overexpressing lines (Figure 5b). In contrast to the results observed in TeSEP4-overexpressing lines, the total carotenoid levels were dramatically decreased in the RNAi-TeSEP4 marigold lines compared to the control plants (Figure 5c). Additionally, the contents of lycopene, antheraxanthin, zeaxanthin, and lutein were markedly reduced in RNAi-TeSEP4 lines relative to control plants (Figure 5d). Notably, the violaxanthin content exhibited an overall decrease in the transgenic lines, with line #3 showing a significant reduction compared to the wild-type plants (Figure 5d). Similarly, the contents of α-carotene and β-carotene were somewhat decreased in the RNAi-TeSEP4 marigold lines. However, the neoxanthin level in RNAi-TeSEP4 transgenic lines was significantly increased compared to the wild-type plants (Figure 5d).

3.4. Transcriptome Sequencing and Functional Enrichment Analysis of DEGs

The phenotype changes in transgenic marigolds may be due to gene expression changes caused by the TeSEP4; the RNA-seq analysis was conducted on the outermost ray florets from unopened inflorescence and full-blooming inflorescence from the TeSEP4-overexpression transgenic line (OE#2) and wild-type plants. A total of 12 RNA libraries were constructed, each with three biological replicates, for transcriptome analysis. A total of 85.27 Gb of Clean Data was generated, with each sample producing at least 6.22 Gb and GC content ranging from 42.34% to 42.74% (Table S8). The total mapping ratio of clean reads mapped to the reference genome of each library ranged from 91.246% to 92.43% (Table S9). The normalized cluster and principal component analysis (PCA) revealed consistent clustering of biological replicates within each group, indicating high reproducibility (Figure S3a,b), suggesting the transcriptional sequencing data obtained in this study met the requirements for further analyses.

There were 6492 DEGs in OE_S2_vs_WT_S2 and 6643 DEGs in OE_S4_vs_WT_S4 comparisons (Figure 6a); 3505 DEGs were upregulated in OE_S2_vs_WT_S2, and 2930 DEGs in OE_S4_vs_WT_S4 comparisons, respectively; 2987 genes were downregulated in OE_S2_vs_WT_S2 comparison and 3713 DEGs in OE_S4_vs_WT_S4 comparison, respectively (Figure 6a). Furthermore, Venn diagram analysis revealed an overlap of 3995 DEGs between the OE_S2_vs_WT_S2 and OE_S4_vs_WT_S4 comparisons (Figure 6b). The KEGG enrichment analysis indicated most of these DEGs were found in OE_S2_vs_WT_S2 and OE_S4_vs_WT_S4 groups, mainly enriched pathways associated with the “metabolic pathways” (Figure 6c,d). Notably, photosynthetic antenna protein (map00196) and carotenoid biosynthesis (map00906) pathways were detected in this enriched pathway (Figure 6c,d), suggesting the chlorophyll and carotenoid metabolites were regulated by DEGs from both in OE_S2_vs_WT_S2 and OE_S4_vs_WT_S4 comparisons.

3.5. DEGs Involved in the Carotenoid and Chlorophyll Biosynthesis Pathway and Their Expression Patterns

To elucidate the genetic mechanisms underlying petal color changes in TeSEP4-overexpressing marigolds, we identified 16 chlorophyll biosynthesis-related and 14 carotenoid biosynthesis-related genes from the transcriptome (Tables S10 and S11). Among these, five chlorophyll-related genes (TePORA1, TeCHL27-1, TeCHLH, TeSGR1, and TeSGR2) and three carotenoid-related genes (TePSY1, TePSY3, and TeLCYB) were differentially expressed between the TeSEP4-overexpressing transgenic lines and the wild-type plants (Tables S12 and S13).

The biosynthetic pathways and expression profiles of these chlorophyll-related DEGs were listed in Table S12. Among these five DEGs, three were involved in chlorophyll synthesis (TePORA1, TeCHL27-1, and TeCHLH), while the other two genes (TeSGR1 and TeSGR2) were associated with chlorophyll degradation (Table S12). All three chlorophyll synthesis genes exhibited a similar expression pattern and showed significantly lower expression levels in OE_S4 than in WT_S4 (Figure 7a,b, Table S12). Notably, the TePORA1 transcript was markedly induced in OE_S2 compared to WT_S2 (Figure 7b, Table S12). Conversely, the expressions of TeSGR1 and TeSGR2 were significantly downregulated in OE_S2 and OE_S4, respectively, compared to WT_S2 and WT_S4 (Figure 7b, Table S12). Correlation analysis indicated that the qRT-PCR analysis of TePORA1, TeCHLH, TeCHL27-1, and TeSGR2 confirmed the transcriptomic data, demonstrating a significant correlation (R² > 0.90, p < 0.05) (Figure 7c and Figure S4). Conversely, TeSGR1 showed weaker correlation (R² < 0.50, p > 0.05) (Figure 7b), which might be related to the redundancy of the SGR gene family. The expressions of TePORA1, TeCHLH, and TeCHL27-1 were further analyzed in petals from unopened (RNAi_S2) and fully bloomed (RNAi_S4) inflorescences of TeSEP4-RNAi transgenic lines (Figure 7c). QRT-PCR analysis demonstrated that TePORA1, TeCHLH, and TeCHL27-1 expressions were significantly elevated in RNAi-TeSEP4_S4 compared to WT_S4 (Figure 7c). Similarly, TeCHLH and TeCHL27-1 expressions were also significantly higher in RNAi-TeSEP4_S2 than in WT_S2 (Figure 7c). However, the expression of TeSGR2 did not change significantly in the RNAi-TeSEP4 transgenic lines compared with wild-type plants (Figure 7c). Our results suggest that TeSEP4 may regulate chlorophyll biosynthesis by modulating the expression of chlorophyll-related genes.

The biosynthetic pathways and expression profiles of three carotenoid-related DEGs (TePSY1, TePSY3, and TeLCYB) are shown in Table S13, respectively. Two PSY genes exhibited distinct expression patterns between the OE_S2_vs_WT_S2 and OE_S4_vs_WT_S4 (Figure 8b, Table S13). Specifically, TePSY1 transcript levels were significantly elevated in both OE_S2 and OE_S4 petals relative to the wild-type plants (Figure 8b, Table S13). In contrast, TePSY3 displayed a stage-specific pattern, with significant upregulation in OE_S2 but downregulation in OE_S4 (Figure 8b, Table S13). TeLCYB was only upregulated in OE_S4 (Figure 8b, Table S13). Correlation analysis indicated that the qRT-PCR results of TePSY1 and TePSY3 were strongly correlated with the RNA-seq data (Figure 8b, Table S13). However, TeLCYB showed weaker correlation (R² < 0.50, p > 0.05) (Figure 8b), which might be related to the primers. Furthermore, qRT-PCR analysis revealed significantly reduced TePSY1 expression in RNAi-TeSEP4#3 lines across both the S2 and S4 stages relative to wild-type plants (Figure 8c). The transcript level of TePSY3 was significantly lower in the S2 stage of RNAi-TeSEP4#3 lines than in that of wild-type plants, whereas no significant change was observed at the S4 stage (Figure 8c). Collectively, these results revealed that TeSEP4 might regulate the transcripts of TePSY1 and TePSY3.

4. Discussion

Chlorophylls and carotenoids are vital plastid pigments responsible for the coloration of fruits, vegetables, and flowers [1,58]. In horticultural plants, their content directly determines ornamental and industrial value [59]. However, the molecular mechanisms coordinating the biosynthesis of these two pigments remain poorly understood. Here, we report that TeSEP4 regulates marigold petal coloration by simultaneously modulating the biosynthesis of both chlorophylls and carotenoids.

4.1. TeSEP4 Simultaneously Regulates Chlorophyll and Carotenoid Biosynthesis in Marigold Petals

Numerous studies have revealed that MADS TFs in horticultural plants are involved in fruit ripening and carotenoid accumulation [40,60,61,62]. For example, overexpression of CsMADS5 in citrus fruits and tomato promotes the carotenoid contents [63]. Among MADS genes, many SEP genes play crucial roles in floral organ development, fruit ripening, and flower coloration [31,41,64]. In this study, overexpression of TeSEP4 in marigold promoted the accumulation of carotenoids, yielding dark-yellow petals (Figure 2 and Figure 5a). Conversely, silencing TeSEP4 led to a decrease in the accumulation of carotenoids, resulting in yellow-green petals (Figure 3, Figure 4 and Figure 5). The established role of TeSEP4 as a key regulator of carotenoid biosynthesis offers substantial potential for both floriculture and lutein production. Modulating its expression enables the creation of novel marigold varieties with a range of petal colors, enhancing ornamental value. Moreover, manipulating TeSEP4 can boost lutein yield through metabolic engineering. Previous studies have revealed that the color of marigold petals is related to the content of carotenoids. Carotenoid concentrations in dark petals are higher than those in light petals [65]. Given the conserved amino acid sequence structure of SEP proteins within the Asteraceae family, it is plausible that TeSEP4 or its orthologs may similarly regulate carotenoid biosynthesis in other marigold cultivars and closely related species. Future comparative studies across diverse Asteraceae taxa could help validate the broader applicability of these findings and reveal potential variations in their regulatory mechanisms. Silencing the CsMADS3 in citrus fruit peel led to a significant accumulation of chlorophyll and impaired the color transition from green to yellow [66] (Zhu et al., 2023). Similarly, suppression of the MADS gene CaRIN in pepper increased total chlorophyll and chlorophyll a content and yielded uneven fruit coloration [61]. In this study, inhibiting TeSEP4 in marigolds led to an increase in total chlorophyll, chlorophyll a, and chlorophyll b contents, which in turn caused the color of the petals to change from yellow to yellow-green (Figure 3, Figure 4 and Figure 5). Furthermore, the chlorophyll contents were significantly down-regulated in petals of TeSEP4-overexpression lines (Figure 4a). Collectively, TeSEP4 functions as a transcriptional dual regulator, repressing chlorophyll biosynthesis while stimulating carotenoid accumulation in marigold petals. However, it is important to acknowledge that carotenoid accumulation is regulated not only through biosynthetic pathways but also by processes such as chromoplast differentiation, pigment sequestration capacity, and carotenoid cleavage activity. While the present study focused on the transcriptional role of TeSEP4 in coordinating chlorophyll and carotenoid biosynthesis, these additional regulatory layers may likely contribute to the overall pigment phenotype and represent important directions for future research.

4.2. TeSEP4 Might Be a Key TF Coordinating Chlorophyll and Carotenoid Biosynthesis

The biosynthesis of chlorophyll and carotenoids in plants is governed by a complex, coordinated regulatory network. Current research has revealed that various TFs modulate the accumulation of these pigments by regulating the enzyme-encoding genes within their respective metabolic pathways [22,23]. Here, KEGG enrichment analysis revealed that “Metabolism” was identified as the most significantly enriched pathway category among the DEGs in TeSEP4-overexpressing transgenic marigold, with many DEGs identified as belonging to the enrichment category of ‘photosynthetic antenna protein (map00196) and arotenoid biosynthesis (map00906) pathways (Figure 6c,d), suggesting TeSEP4 might affect the transcription of genes encoding enzymes in the chlorophyll and carotenoid metabolic pathways. However, it should be explicitly noted that these transcriptomic results indicated an association rather than direct regulatory evidence, and whether TeSEP4 exerts direct transcriptional control requires further experimental validation.

The alteration of chlorophyll synthesis is frequently due to changes in the expression of structural genes, which in turn lead to color changes in plant tissues [67,68]. The enzymes of CHLH, PORA, and CHL27 play key roles in this process. In tobacco (Nicotiana benthamiana), suppression of CHLH decreases the chlorophyll content in leaves and induces a chimeric phenotype with green and yellow/white leaves [69]. The mutation SlCHLH in tomato delayed the progression of fruit ripening, reduced carotenoid content, and delayed plastid conversion from chloroplast to chromoplast [70]. Silencing of LcPORA in Chinese tulip tree (Liriodendron chinense) petal inhibits chlorophyll biosynthesis and causes the petal color to change from dark green to light green [71]. The CHL27 is required for the synthesis of protochlorophyllide; the mutation of CHL27 in Arabidopsis generates pale-green leaves and disrupts the Chl biosynthesis [72,73]. In this study, overexpression of TeSEP4 produced dark-yellow petals with lower chlorophyll content (Figure 2 and Figure 4), while suppression of TeSEP4 generated yellow-green petals with higher chlorophyll content, compared to the wild-type plants (Figure 3 and Figure 4). Additionally, the changes in chlorophyll content in transgenic plants were along with the changes in the expression levels of TeCHLH, TeCHL27-1, and TePORA1, suggesting TeSEP4 might regulate these three genes’ transcripts. Current studies have revealed that transcriptional regulation of chlorophyll biosynthesis is achieved by the regulated expression of chlorophyll synthesis-related genes [23,74]. For example, rice OsGATA16 directly promotes the expression of OsHEMA, OsCHLH, and OsPOR and increases chlorophyll accumulation in leaves [75]. Similarly, chrysanthemum (Chrysanthemum morifolium) CmNAC73 promotes chlorophyll biosynthesis by directly binding to the promoter of chlorophyll synthesis-related genes CmHEMA1 and CmCRD1 [67]. In this study, TeSEP4 overexpression in marigold down-regulated the transcripts of TeCHLH, TeCHL27-1, and TePORA1, while suppression of TeSEP4 in marigold up-regulated these gene expressions (Figure 7), suggesting TeSEP4 potentially acts as a negative regulator of these genes, necessitating further experimental validation.

Carotenoid metabolism is important for fruit and flower coloration alternation [76,77,78]. In this process, PSY, as the foremost key-limiting enzyme, is the core determinant of the overall metabolism of carotenoids. In apple (Malus × domestica) and rose (Rosa rugosa), PSY1 promotes carotenoid accumulation [79,80]. In this study, the expressions of TePSY1 and TePSY3 transgenic lines were significantly increased in the S1 stage of TeSEP4-overexpression lines, while the expressions of TePSY1 and TePSY3 were significantly decreased in the S1 stage of RNAi-TeSEP4 transgenic lines (Figure 8), which might explain the alteration of carotenoid content in transgenic lines. Notably, contents of lycopene and lutein were remarkably changed in both TeSEP4-overexpression and RNAi-TeSEP4 transgenic marigolds (Figure 5). However, overexpression of TePSY3 in tomato increases the lutein content, but does not affect the lycopene biosynthesis [47]. We speculate that the change in lycopene content in transgenic lines may be related to the change in TePSY1 expression. In many plants, two or more PSY genes are identified, which exhibit tissue-specific expression patterns and present overlapping and specific functions in carotenogenesis during evolution [81,82,83]. For example, in wheat (Triticum aestivum), the three PSY genes exhibit functional specialization, with PSY1 driving carotenoid biosynthesis in endosperm, while PSY2 and PSY3 function in green tissues and roots, respectively [84]. However, in tobacco (Nicotiana tabacum), PSY1 and PSY2 are specifically expressed in leaves, while gene PSY3 is not expressed in leaves [82]. Additionally, virus-induced silencing of both PSY1 and PSY2 in tobacco leads to a marked decrease in carotenoid and chlorophyll content [82]. There are three PSY genes in marigolds, and TePSY1 and TePSY3 share similar expression pattern in the yellow-orange flower from the variety ‘Quanwang’ [47]. Collectively, these findings point to the coordinated functions of TePSY1 and TePSY3 as the underlying cause for the altered carotenoid profiles in the transgenic marigolds. Similarly to our results, citrus CsMADS5 and CsMADS6 promote carotenoid content by inducing the CsPSY transcript. Silencing CaNAC81 in chili peppers inhibits the PSY expression and decreases the carotenoid content [85]. These findings suggested that TeSPE4 likely acts as a positive regulator of TePSY1 and TePSY3.

Previous studies have revealed that SEP proteins typically function as heterodimers or higher-order complexes with other MADS-box transcription factors [64]. In this study, many DEGs belonging to the MADS family were identified in transcriptome data, suggesting that TeSEP4 may also operate within such protein complexes to coordinate petal development and pigment metabolism in marigold. However, the specific interaction partners of TeSEP4 remain to be elucidated, necessitating further experimental studies.

5. Conclusions

In this study, we demonstrate that TeSEP4 alters marigold petal color by coordinately regulating the biosynthesis of chlorophyll and carotenoids. Specifically, TeSEP4 likely promotes carotenoid accumulation by activating TePSY1 and TePSY3, while simultaneously inhibiting chlorophyll accumulation by repressing the synthesis-related genes TeCHLH, TeCHL27-1, and TePORA1. Therefore, we propose a potential working model for TeSEP4 (Figure 9), which illustrates its role in the coordinated regulation of these two pigment pathways. Our work elucidates a key regulatory mechanism governing pigment biosynthesis in plants. Furthermore, manipulating TeSEP4 or its downstream targets unlocks the potential to create novel marigold varieties with unique petal color patterns, enhancing their ornamental value. Beyond floriculture, this regulatory mechanism offers genetic insights for boosting carotenoid levels in other crops.