Comparative Analysis of the Effects of Internal Factors on the Floral Color of Four Chrysanthemum Cultivars of Different Colors

Abstract

Flower color, a critical phenotypic trait of ornamental plants, is an essential indicator for flower variety classification. Many physical and internal factors that affect flower color have been widely investigated; however, the effects of internal factors during the flowering period remain unknown. In this study, we evaluated the effects of internal factors on floral coloration during the flowering period of four chrysanthemum cultivars of different colors. colorimetric measurements showed that L*, a*, and b* were in correlation with the lightness and color development in the four chrysanthemum cultivars. The distinctive shape of upper epidermal cells was observed in each flowering stage of different colored chrysanthemums. With progression of the flowering process, the content of anthocyanins and carotenoids increased during early stages, decreased at the senescence stage, and was the highest at the full-bloom stage. The vacuolar pH of flowers gradually decreased as the flower bloomed. Metal contents in flowers varied across different chrysanthemum varieties. Anthocyanins biosynthesis genes, such as CmCHS and CmCHI, were expressed and responsible for pigment changes in red chrysanthemums. Moreover, the expression pattern of cytosol pH-related genes, such as CmVHA-a1, CmVHA-C, and CmVHA-C″1, was in accordance with a decrease in pH during flowering stages. Our results revealed the effects of main internal factors on floral color during the flowering period in four Chrysanthemum varieties, providing insights into the introcellular and molecular regulatory mechanisms of flower coloration and laying key foundations for the improvement of color breeding in chrysanthemums.

1. Introduction

Flower color is one of the most important characteristics of ornamental plants and is essential in economy and biology [1]. The bright-colored corolla is not only the key appreciation merit but also a strong signal to attract insect pollinators, which is conducive to plant reproduction [2]. Over the years, several studies have reported on the mechanism of floral color regulation in plants. Many environmental factors affect the color of flowers [3]. For example, temperature exerts obvious effects on petal color by causing changes in the anthocyanins content [4,5]. Light quality, intensity, and duration are also important regulators of the floral color of many ornamental plants [6,7,8]. Cultivation management, such as watering, can also affect the floral color to some extent [9].

On the other hand, floral color is affected by some internal factors, especially pigments, epidermal cell shapes, vacuolar pH, metal ions, and gene expression. Pigment type, content, and distribution have a direct relationship with petal color in various ornamental plants [10]. Anthocyanins are widely distributed in different chemical forms in many plants, which results in variations in flower color across different plants [11,12]. For instance, delphinidin aglycone tends to confer brilliant blue or purple color to flowers such as cineraria [13]. Carotenoids confer bright red, orange, and yellow colors to flowers such as osmanthus and yellow oncidium [14,15]. Petals of yellow-flowered marigold “Lady” contain only lutein [16]. In addition, the shape of epidermal cells affects pigments’ optical properties and, thus, the perceived color of petals. For example, diffraction caused by regularly arrayed petal epidermal cells could result in an iridescent effect in Hibiscus trionum [17].

Vacuolar pH regulates petal color by altering anthocyanins conformation and absorption spectrum [18]. When the flowers of Morning glory bloom, the color of their petals changes from purplish-red to blue, which involves alternations of anthocyanins conformation [19]. However, in petunia, seven loci (pH1–pH7) have been identified; mutation of these loci results in bright blue color of the flowers, which is due to an increase in the vacuolar pH of petal extracts but not due to alternation of anthocyanins composition [20]. Metal ions co-existing with pigments play an essential role in floral color development by altering the activity of enzymes involved in pigment biosynthesis, transition, and destruction, as well as by generating metal–pigment complexes [21,22]. The bright blue color of cornflowers is a result of the stacking of apigenin, cyanidin, Fe³⁺, Mg²⁺, and Ca²⁺-complexes [23].

Genes are essential in the transcriptional regulation of the pigmentation pattern in flowers. Anthocyanins biosynthesis genes, such as chalcone synthase (CHS), chalcone flavanone (CHI), flavonoid 3-hydroxylase (F3H), and dihydroflavonol 4-reductase (DFR), have been well identified, and the expression of these genes is closely related to the anthocyanins content, thereby affecting petal color in plants [24]. F3′5′H encoding flavonoid 3′5′-hydroxylase is a key enzyme of delphinidin biosynthesis in most blue flowers. Genetic modifications of F3′5′H are an effective method for obtaining blue-flower cultivars in some ornamental plants [25,26]. Moreover, members of the WD40-bHLH-MYB regulatory complex and genes regulating the vacuolar pH could regulate anthocyanins accumulation and composition, which in turn affects floral color development [27,28,29].

Chrysanthemum (C. morifolium Ramat.) is one of the most important ornamental plants in the world that exhibits great flower color diversity. Its color trait is a highly dominant trait compared with that of other floricultural crops. The chrysanthemum inflorescence consists of central disc florets and outer ray florets. The color of disc florets is generally pale yellow or light green, whereas that of ray florets (petals) is highly diverse and is the most attractive part, which is specific to each cultivar. In this study, we evaluated the effect of pigment content and composition, upper epidermal cell shape, vacuolar pH, metal ion content, and genetics of ray florets’ color formation during the flowering period on floral coloration of four colored (white, pink, red, and yellow) chrysanthemums. Understanding the regulatory mechanisms of these factors in the coloration of ray florets will not only provide professional ornamental knowledge, but also have great implications for the breeding manipulation of flower color in chrysanthemum.

2. Materials and Methods

2.1. Materials and Maintenance Management

Four colored chrysanthemum cultivars used in this study were grown in the Shandong Agricultural University Nursery (China). To reproduce them, healthy and disease-free stems were selected from the parent chrysanthemums in December 2020. The stems were cut into 6-cm pieces containing only one leaf and then rooted in a 1:1 mixture of vermiculite and perlite. The cuttings with new roots were transplanted in flowerpots in January 2021. When these new plants reached a height of approximately 20 cm, apical meristems were truncated for the first time. The second truncation was performed when new branches grew to approximately 10–12-cm long. After approximately 2 months of cultivation, or when the height of plants reached approximately 40 cm, flowering treatment was performed under a short-day photoperiod (light/dark, 9 h/15 h, 23 °C/18 °C). The ray floret samples of the four chrysanthemum cultivars were collected in June 2021 for further study.

2.2. Colorimetric Measurement

The color intensities of ray florets in different flowering stages of the four colored chrysanthemums were measured using a colorimeter (NF333, Nippon Denshoku Industries Co., Ltd., Tokyo, Japan) and a C/2° light source. Six ray floret samples were collected from three different plants, and the middle part of each ray floret was used for colorimetric measurement [30]. The values of L*, a*, and b* were measured.

2.3. Extraction and Measurement of Total Flavonoids

The ray floret samples were collected and ground with mortar and stored in liquid N₂. Subsequently, 0.2 g of fine powder was added to 3 mL of methanol for extraction under the dark condition for 24 h. The mixture was filtrated through a 0.22-µm organic membrane filter. Thereafter, 0.3 mL of the supernatant and 4.7 mL of 1% AlCl₃·6H₂O methanol solution were mixed for 10 min at room temperature. Thereafter, OD_405nm was measured using a spectrophotometer (UV-2450, Shimadzu, Tokyo, Japan) [31].

2.4. Extraction and Determination of the Anthocyanins Content

Total anthocyanins were extracted using the ethanol-HCl method and detected as described by Hu et al. [32]. Approximately 0.1 g of fresh ray floret sample was ground and mixed with 10 mL of anthocyanins extract (absolute ethanol alcohol:sterile water:hydrochloric acid = 43:9:1, v/v/v). Thereafter, OD_340nm was measured using a spectrophotometer. Each treatment was repeated three times.

2.5. Extraction and Determination of Carotenoid Content

Total carotenoids of ray florets were extracted according to a previously reported protocol [33]. Briefly, approximately 0.2 g of fresh sample was used for carotenoid release by using acetone:petroleum ether (1:4, v/v) as the extraction agent. After 48 h-extraction at 4 °C in the dark, the mixture was centrifuged at 4000 r * min⁻¹ for 10 min. The OD_450nm value of the supernatant was measured using a spectrophotometer. Each treatment was repeated three times.

2.6. Observation of Upper Epidermal Cells of Ray Florets at Different Stages

At each flowering stage, the middle parts of ray florets were subjected to upper epidermal cell observation [34]. The images of upper epidermal cells were obtained using an optical microscope (Olympus CX31, Tokyo, Japan).

2.7. pH Measurement of ray Floret Homogenates

The complete inflorescence at each flowering stage was ground with quartz sand in 6 mL of distilled water and immediately centrifuged at 4 °C and 1200 rpm for 1 min. The supernatant was used for pH measurement using a pH electrode (Bell Instruments, pocket-sized pH meter) [35].

2.8. Determination of Metal Ions

The ray florets at each flowering stage were collected and washed with distilled water three times. Thereafter, the samples were kept at 95 °C for 15 min, and dried to a constant weight at 110 °C for several hours. The dried samples were ground into powder and sieved through 60-mesh nylon. The powder was used for the determination of metal ions. Na⁺, K⁺, Fe³⁺, Ca²⁺, Mg²⁺, Cu²⁺, and Zn²⁺ in the powder of ray florets were measured through flame atomic absorption spectrometry according to a previously reported method [36].

2.9. Quantitative Reverse Transcription-Polymerase Chain Reaction

Total RNA was extracted from ray florets of the four chrysanthemum varieties by using TRIzol Reagent (Vazyme, Nanjing, China). Reverse transcription was performed using a PrimeScript first-strand cDNA synthesis kit (TaKaRa, Dalian, China). The relative expression of anthocyanins- and pH-related genes was calculated using the 2^−ΔΔCt method [37]. The primers used for quantitative reverse transcription-polymerase chain reaction (qRT-PCR) are listed in Supplementary Tables S10 and S11. All reactions were performed in triplicate.

2.10. Statistical Analysis

All assays were performed in triplicate, and data are expressed as the mean ± standard deviation. Statistical significance was determined using the Duncan’s multiple range test (p < 0.05).

3. Results

3.1. Measurement of Colorimetric Values of Four Chrysanthemum Varieties during the Flowering Period

To better interpret the effects of internal factors on flower color development during the flowering period, we chose four colored chrysanthemum varieties as plant materials (Figure 1A–D). As shown in Figure 1E–H, the flowering period of chrysanthemums comprised six stages from S1 to S6. The description of six flowering stages is given in Supplementary Table S1, wherein S5 represents a completely unfolded inflorescence of chrysanthemums with the optimum performance. The CIELAB color system is extensively used for the identification of floral coloration in plants [38]. The colorimetric measurement showed that color space values, such as L*, a*, and b*, were in connection with the lightness and color development of ray florets. The L* value showed a negative correlation with flower color development. The deeper the color, the lower the L* value. The L* value of red-ray florets was lower than that of white- and yellow-ray florets. In addition, the L* value increased with progression of the flowering process and reached the highest in S6 of all chrysanthemum varieties (Figure 1I). The values of a* and b* represent the degree of redness and yellowing of flower color, respectively. An increase in a* value represents a lesser greenish color but a redder color, whereas an increase in b* value represents a lesser bluish color but a more yellowish color. In our study, the a* value of the four varieties was relatively variable, ranging from −4.35 in S5 of yellow chrysanthemums to 9.45 in S5 of red chrysanthemums (Figure 1J), which indicated a rare change of red color during the flowering period, except for red chrysanthemums. A nonsignificant correlation was observed between L* and a*, although the result is not shown. The b* value was rarely changed from S1 to S4 of these four chrysanthemums. However, a rapid increase in the b* value in S5 and S6 of red and yellow chrysanthemums was observed (Figure 1K), which suggested an increase in yellow-colored pigments in the ray florets of two chrysanthemum varieties. Red chrysanthemums used in this study showed a dark orange color at the end of flowering stages, which was in accordance with the rapid increase in the b* value.

3.2. The Effect of Upper Epidermal Cells of Ray Florets on Floral Color during the Flowering Period

The shapes of petal epidermal cells exhibit some effects on floral color via the regulation of light absorption and reflection by pigments in petal cells [39]. Therefore, we observed the upper epidermal cells of ray florets of the four chrysanthemums during flower opening stages. Our results showed that the upper epidermal cells expanded during the flowering process, with a continuous increase of cell area from S1 to S6. However, some differences were observed in cell size and shape among these four varieties. In terms of cell size, the pink chrysanthemums had a larger upper epidermal cell at each flowering stage than the other three chrysanthemums, whereas the cell size of yellow chrysanthemums was the smallest (Figure 2E). Regarding cell shape, which plays a greater role in flower color than cell size, a different pattern was observed. Different colored chrysanthemums had distinctive upper epidermal cell shape at each flowering stage. The upper epidermal cells were present as crowded cuboids in W1, which expanded into irregular hexagons from W4 to W6 of white chrysanthemums (Figure 2A). More irregularly shaped epidermal cells were observed in pink chrysanthemums, with sinuous undulations from P1 to P5 but few undulations in P6 (Figure 2B). In red chrysanthemums, the conical epidermal cells of ray florets in R2 gradually expanded and became flat in the later stages and round-shaped in R6 (Figure 2C). Moreover, the square or striped shape of upper epidermal cells with anticlinal walls was nonuniform in Y1. Then, the epidermal cells became longer and undulated in Y3 to Y5 and grew into irregularly flat hexagons in Y6 of yellow chrysanthemums (Figure 2D). To conclude, the shape and size of upper epidermal cells did not affect the flower color of chrysanthemums.

3.3. The effect of Pigments and pH on the Floral Color of Four Chrysanthemum Varieties

Previous studies have shown that anthocyanins, carotenoids, and flavonoids are major floral pigments that contribute to the range and type of petal color in plants [12]. In this study, the detection and analysis of pigments in ray florets showed that the anthocyanins content in pink chrysanthemums gradually increased and then decreased during the flowering process. On the other hand, anthocyanins content was sharply decreased in S6 of red chrysanthemums and gradually increased in yellow chrysanthemums (Figure 3A), which is consistent with the factual color change of ray florets. The carotenoid content in red and yellow chrysanthemums increased gradually with progression of the flowering process (from S1 to S5) and showed an obvious decrease in the late senescence stage (S6). Additionally, the carotenoid content in red and yellow flowers was much higher than that in white and pink flowers (Figure 3B), which is in accordance with the results of previous studies [40,41]. Flavonoids, a secondary metabolite, are one of the most important pigments; however, their compositions vary across different colored chrysanthemums [10]. Our results showed that the total flavonoid content of white and red chrysanthemums decreased gradually during the flowering process, whereas that of pink chrysanthemums showed a relatively slower decreasing trend during the flowering period. By contrast, an increase in the flavonoid content was observed in S6 of yellow chrysanthemums (Figure 3C).

Cytosol pH or vacuolar pH is important to determine the floral color because it affects the coloration of anthocyanins in cells [42,43]. In this study, we measured the cytosol pH of ray floret homogenate of the four chrysanthemum varieties. Cytosol pH of ray florets of white and red chrysanthemums decreased slightly from S1 to S5 but increased suddenly in S6. A slight reduction in cytosol pH was observed in pink chrysanthemums during the flower blooming stage. A decrease in cytosol pH was significant during the flowering process in yellow chrysanthemums (Figure 3D). The values of cytosol pH during the flowering period were in accordance with the color development of red and pink chrysanthemums to some extent.

3.4. Quantification of Metal Contents in Ray Florets of Four Chrysanthemum Varieties

Metals as the cofactor of pigments are another internal factor that affect floral coloration in plants [21]. In this study, we quantified seven types of metal ions in ray florets during the flowering process. The result showed a variation of metal ions during the flowering period in different colored chrysanthemums. The content of potassium ion (K⁺) showed an increasing and decreasing trend in six flowering stages, with the highest content in S1 or S6 of white, pink, and red chrysanthemums. In yellow chrysanthemums, the content of K⁺ was highest in S1 to S3, and decreased from S4 to S6 (Figure 4A). The content of calcium ions (Ca²⁺) exhibited large fluctuations during the flowering process. Ca²⁺ content was higher in white chrysanthemums than in the other three chrysanthemums in each stage and showed a gradual decrease from S1 to S3 but a gradual increase from S4 to S6, reaching the highest in S6. The Ca²⁺ content in pink chrysanthemums showed an irregular wave during the flowering process, with the highest content observed in S2 and lowest content observed in S5. Among all chrysanthemums, the red chrysanthemums exhibited the lowest Ca²⁺ content in each flowering stage. The ray florets in the first three stages had a higher Ca²⁺ content than those in the latter three stages, with two peaks (S2 and S5) during flower blooming. In yellow chrysanthemums, Ca²⁺ content remained unchanged from S1 to S4 but increased in S5 and S6 (Figure 4B). The sodium (Na⁺) content in the four colored chrysanthemums showed distinct changes. There were two peak variations of Na⁺ content, S1 to S3 and S4 to S6, with the Na⁺ content being the highest in S6 and lowest in S2 in white chrysanthemums. The Na⁺ content in ray florets of pink chrysanthemums decreased in S2 and increased in S3 but did not change much in the later three stages. The ray florets of red chrysanthemums showed lower Na⁺ content than those of the other three colored chrysanthemums; however, its Na⁺ content increased in the late flowering period. Moreover, the Na⁺ content in yellow chrysanthemums was constantly high from S1 to S3 but showed a gradual decrease from S4 to S6, with the lowest content in S6 (Figure 4C). The zinc ion (Zn²⁺) content showed an overall decreasing trend in ray florets of white chrysanthemums, a slight increase and decrease in ray florets of pink chrysanthemums, a fairly smooth change in ray florets of red chrysanthemums, and a gradual decrease in ray florets of yellow chrysanthemums (Figure 4D). The iron ion (Fe²⁺) content was distinctively changed during flowering stages in ray florets of four colored chrysanthemums. The Fe³⁺ content in white flowers was the lowest in each stage compared with that in the other three chrysanthemums but increased as the flowering process proceeded. In pink flowers, the Fe³⁺ content was highest in S3, decreased sharply in S4, and increased in S5 and S6. In red chrysanthemums, the Fe³⁺ content was higher from S1 to S3 than that from S4 to S6, with the Fe³⁺ content being lowest in S6. Conversely, the Fe³⁺ content in yellow flowers was higher from S4 to S6 than from S1 to S3, which remained unchangeable in the early stages of blooming (Figure 4E). Finally, the contents of copper ions (Cu²⁺) and magnesium ions (Mg²⁺) were unstable during the flowering process in the four colored chrysanthemums (Figure 4F,G).

To conclude, the metal ions played different roles in affecting the floral color of the four Chrysanthemum varieties. Changes in the Na⁺ content were most remarkable across these plants during the flowering period. Fe³⁺ and Mg²⁺ were relevant to ray floret coloration of red and pink chrysanthemums in the early stages of flowering, whereas Fe³⁺ may be related to yellow color formation in the later stages of flowering. In other words, floral color may not be affected by a single determining factor.

3.5. The Expression of Genes Associated with Anthocyanins Synthesis and Cytosol pH in Ray Florets of Chrysanthemums

Anthocyanins and its essential cofactor cytosol pH are strongly associated with flower coloration, therefore, we analyzed the relative expression of genes associated with anthocyanins biosynthesis and cytosol pH. The anthocyanins biosynthesis genes have been identified [24,44], and cytosol pH is generally regulated by vacuolar ATPase, pyrophosphatase, and ion exchangers on cell membranes [28,29]. In our study, the key genes of anthocyanins biosynthesis, such as CmCHI, CmCHS, CmANS, CmDFR, CmF3H, and CmF3′H, were detected through qRT-PCR during the flowering period. The assay showed that the expression changes of some genes were strongly associated with pigment coloration in the four colored chrysanthemums during the flowering period. For instance, CmCHS and CmCHI showed the highest transcript levels in S2 and S5 of red chrysanthemums, but in S1 of the other three colored chrysanthemums (Figure 5A,B). CmF3′H and CmANS were mainly expressed in red and pink chrysanthemums, with high expression from S4 to S6 (Figure 5D,F). However, CmDFR and CmF3H showed distinctive expression patterns. The expression of CmF3H varied during the flowering period of red chrysanthemums; the expression was high in S1 and S2, but decreased markedly in S3 and S4 and then increased suddenly in the later stages. A few transcript abundances of CmF3H were observed in the other three chrysanthemums (Figure 5C). CmDFR was expressed mainly in red, pink, and yellow chrysanthemums to varying degrees, with the highest transcripts in S6. The CmDFR transcript abundance exhibited a variable trend, with a gradual increase in pink chrysanthemums, two transcript peaks in S2 and S5 of yellow chrysanthemums, and marked decrease in S3 and S4 of red chrysanthemums (Figure 5E).

We confirmed the expression pattern of cytosol pH-related genes during different flowering stages. CmVHA-C″1, CmVHA-a1, and CmVHA-C showed the maximum expression in S4 of white, pink, and red chrysanthemums (Figure 6A,C,E). CmVHA-A, CmVHA-C1, CmNHX1, and CmpH1 showed the maximum expression in S2 of yellow chrysanthemums. CmVHA-B2 showed the highest transcript abundance in S3 of red and yellow chrysanthemums, as well as in S5 of white and pink chrysanthemums (Figure 6B). In general, the expression pattern of cytosol pH-related genes was in accordance with a decrease in pH during flowering stages. CmVHA-a1, CmVHA-C, and CmVHA-C″1 were expressed more in the early flowering stages of red and pink chrysanthemums, which together with anthocyanins coloration rendered the color of red and pink chrysanthemums.

3.6. The Comprehensive Analysis of the Measurement Indexs of Ray Florets of Four Colored Chrysanthemums

According to the principal component analysis (PCA), the variance contribution rate of white chrysanthemums at different stages was 41.41%, 23.79%, 15.24%, and 12.78%, respectively, and the cumulative variance contribution rate was 93.21% (Supplementary Table S2). The correlation analysis showed that the relative expression of CmVHA-C, CmF3H, CmF3′H, cytosol pH, and Na⁺ had a strong correlation with color development of white chrysanthemums during different flowering stages (Supplementary Table S3). The PCA at different stages of pink flowers showed that the eigenvalues of these four principal components were >3, and their variance contribution rates were 44.28%, 20.83%, 13.91%, and 13.19%, respectively, and the cumulative variance contribution rate was 92.21% (Supplementary Table S4). Meanwhile, the relative expression of CmCHS and CmF3H, upper epidermal cell shape, and Mg²⁺ content had a strong correlation with pink color development at different stages (Supplementary Table S5). In red chrysanthemums, PCA showed that the eigenvalues of these four principal components were >2; their variance contribution rates were 43.31%, 24.41%, 19.81%, and 7.36% (Supplementary Table S6), respectively, and the cumulative variance contribution rate was 94.9%. The relative expression of CmVHA-C1, color difference values a* and b*, upper epidermal cell shape, cytosol pH, and K⁺ and Cu²⁺ contents had a strong correlation with color development of red chrysanthemums during different stages (Supplementary Table S7). Finally, the PCA of yellow chrysanthemums showed that the eigenvalues of four principal components were >2; their variance contribution rates were 51.54%, 23.3%,12.42%, and 8.17%, respectively, and the cumulative variance contribution rate was 95.42% (Supplementary Table S8). The correlation analysis showed that the relative expression of CmVHA-a1, CmVHA-C, CmVHA-C″1, and CmCHI, color difference value b*, upper epidermal cell shape, carotenoids, and Na⁺ and Ca²⁺ contents exhibited a strong correlation with the color development of yellow chrysanthemums at different stages (Supplementary Table S9). Taken together, our analysis suggested that some of these factors may represent the dominant regulators in the certain flowering stage of chrysanthemums. Moreover, the dominant regulators varied across different color chrysanthemums.

4. Discussion

To date, many studies have been conducted on flower color formation and classification, pigment types and content, and pigment genetic regulation in chrysanthemums [38,45]; however, the regulatory mechanisms of these factors in floral color development during the flowering period are unclear. This study evaluated the effects of pigment composition, cytosol pH, metals, and genetics on ray petal color development during the flowering period in four classic, colored chrysanthemums. Based on the comprehensive analysis of all these measurement indexes of ray florets during flowering stages of four colored chrysanthemums, we draw a conclusion that different factors served as the decisive regulator of flower color between different colored chrysanthemums, and some of which may be specific to one certain colored chrysanthemum. For example, Mg²⁺ was related to pink color formation in early flowering stages of pink chrysanthemums (Figure 4G, Supplementary Table S5). K⁺ and Cu²⁺ had a strong correlation with color development of red chrysanthemums (Figure 4A,F, Supplementary Table S7). Carotenoids had a strong correlation with the color development of red and yellow chrysanthemums (Figure 3B; Supplementary Table S9).

Although we obtained some new knowledge about the effects of internal factors on floral color development during the flowering period of four colored chrysanthemums, some unexpected results still remained to be resolved in the near future. For example, introcellular co-pigmentation of flower color involves the interaction of anthocyanins with metal ions such as Fe³⁺ and Mg²⁺ [46,47]. In red chrysanthemums, the change of Fe³⁺ and Mg²⁺ contents was in accordance with the change of the anthocyanins content and, thus, with the ray floret coloration during the flowering period (Figure 2A and Figure 4E). However, the Mg²⁺ content showed a continuous decrease during flowering stages, which was opposite to the changes in the anthocyanins content in yellow chrysanthemums (Figure 2A and Figure 4G). Therefore, the regulatory mechanism of Mg²⁺ in red chrysanthemums would not apply to that in yellow chrysanthemums. Thus, the same metal ion may have distinct roles in floral color development in different plant varieties. Additionally, metal ions may exert their effects on floral color through various mechanisms, except for metal–pigment chelation. Finally, petal color development is determined by several factors, which work together conferring various color of flowers.

5. Conclusions

In this study, we measured and analyzed the internal factors that affect the color of ray florets of chrysanthemums during the flowering period. The shape of upper epidermal cells varied across four colored chrysanthemum varieties during different flowering stages. Anthocyanins accounted for coloration of ray florets in red and pink chrysanthemums, whereas carotenoids accounted for coloration of ray florets in red and yellow chrysanthemums. Cytosol pH was correlated with the color development of red and pink chrysanthemums. Although the Na⁺ and Fe³⁺ content exhibited dramatic changes in all four colored chrysanthemums during the flowering period, it was more correlated with the coloration of red and yellow chrysanthemums. The effects of metal ions on floral color may not be determined by a single determining factor. Moreover, anthocyanins biosynthesis genes such as CmCHS and CmCHI were expressed and responsible for pigment coloration in red chrysanthemums. Cytosol pH-related genes such as CmVHA-a1, CmVHA-C, and CmVHA-C″1 were associated with a decrease in pH during different flowering stages. Our results indicated the effects of main internal factors on ray florets’ color formation, laying a foundation for the improvement of color breeding in chrysanthemum.