Anatomical and Biochemical Traits Related to Blue Leaf Coloration of Selaginella uncinata

Selaginella uncinata shows particularly rare blue leaves. Previous research has shown that structural interference by the cell wall of adaxial epidermal cells imparts blue coloration in leaves of S. uncinata; the objective of this study was to see whether anthocyanins might additionally contribute to this color, as changes in pH, and conjugation with metals and other flavonoids is also known to result in blue coloration in plants. We compared anatomical and biochemical traits of shade-grown (blue) S. uncinata leaves to high light (red) leaves of the same species and also to a non-blue (green) leaves of a congeneric S. kraussiana. By examining the anatomical structure, we found that the shape of adaxial epidermis of S. uncinata leaves was convex or lens-shaped on the lateral view and irregular circles with smooth embossment on the top view. These features were different from those of the abaxial and adaxial epidermis of S. kraussiana. We suspect that these structures increase the proportion of incident light entering the cell, deepening the leaf color, and therefore may be related to blue leaf color in S. uncinata. By examining biochemical traits, we found little difference in leaf pH value among the leaf types; all leaves contained several metal ions such as Mg, Fe, Mn, and copigments such as flavones. However, because there was no anthocyanin in blue S. uncinata leaves, we concluded that blue coloration in S. uncinata leaves is not caused by the three hypotheses of blue coloration: alkalization of the vacuole pH, metal chelation, or copigmentation with anthocyanins, but it may be related to the shape of the leaf adaxial epidermis.


Introduction
Colorful leaves are attractive features that characterize ornamental plants. Red, purple, yellow, and variegated leaves are common, whereas blue leaves are particularly rare. Selaginella uncinata, a fern species that is adapted to the shaded conditions, is such a blue leaf plant. It has a blue upper and green lower surface. ere have been some reports related to S. uncinata focusing on structural anatomy [1], developmental anatomy [2], cell genetics [3], chloroplast genome [4][5][6], and chemicals and medicine [7][8][9][10][11][12][13]. Our previous research showed that the leaf color of S. uncinata appears normally blue in the shade, while it changes to red under full light exposure [14], and this color change under high light corresponded with a reduction in chlorophyll and anthocyanin and an increase in carotenoids, resulting in a dominant orange color. Based on the transcriptome sequencing and quantitative real-time polymerase chain reaction (qRT-PCR) analysis, we concluded that the primary pathway of pigment metabolism in S. uncinata may be the chlorophyll metabolism pathway rather than the anthocyanin biosynthesis pathway [15]. According to the research of leaf coloration mechanism, leaf coloration in plants can be due to either pigments or structural coloration. ese two groups differ in their appearance-pigmented colors look the same from all angles, while structural colors appear with different hues when viewed from different angles, a unique attribute of structural color called iridescence [16].
ere have been very few studies on the production of blue leaves. Hébant and Lee [1] found that the iridescent blue color of S. uncinata and S. willdenowii is caused by thin-film interference (a physical effect). In other blue iridescent plants, the iridescent ultrastructural basis is relevant to their adaxial epidermis, but they are different in detail. In Diplazium tomentosum, Lindsaea lucida, and Danaea nodosa, the iridescent ultrastructure is that in the uppermost cell walls of the adaxial epidermis, the arrangement of multiple layers of cellulose microfibrils is helicoidal [17,18]. In Begonia pavonina, Phyllagathis rotundifolia, and Elaeocarpus angustifolius, blue coloration is due to the parallel lamellae in specialized plastids (iridoplasts) adjacent to the abaxial wall of the adaxial epidermis [18,19], while in Trichomanes elegans, it results from the remarkably uniform thickness and arrangement of grana in specialized chloroplasts adjacent to the adaxial wall of the adaxial epidermis [17]. In these studies, it was not possible to extract blue pigment from the study material, such that in all cases, blue iridescence was considered to be a structural color.
However, according to studies of blue flowers, pH in the vacuole [20][21][22], metal chelation [23], and copigmentation [24,25] may also be related to blue coloration. It seems clear that blue leaves in S. uncinata have a structural mechanism, but whether it is also affected by any or all of these three factors remains unknown. In our previous study, we detected low content of anthocyanins in S. uncinata [14]. e objective of this study was to further investigate the possibility that anthocyanins may contribute to blue coloration in S. uncinata, by examining leaf pH, metal ions, and pigment composition, in addition to anatomical structure.

Plant Material.
We conducted tests on three leaf types: blue S. uncinata leaves grown under a sunshade net (light intensity: 65-105 umol m −2 s −1 ), green S. kraussiana leaves grown under the same conditions, and red S. uncinata leaves grown in full exposure (light intensity: 500-520 umol m −2 s −1 ) [14]. ere were 6 POTS for each leaf type and 3 replicates for a total of 54 POTS. All plants were 6 months old, given normal water and fertilizer management, and cultivated in the nursery of the Forestry College, Guangxi University, Nanning, China. Mature normal leaves were selected randomly in different directions from various individuals when sampling.

Methods.
We compared observations within species (blue and red S. uncinata leaves) and between species (blue S. uncinata and green S. kraussiana leaves). e traits examined included morphology, color parameters, leaf paraffin transverse sections, freehand sections, and scanning electron microscopy (SEM) photomicrographs to determine the structural mechanism. We also compared reported leaf pH values, and metal ion, anthocyanin, and flavonoid content to examine the physiological and biochemical mechanisms related to blue coloration.

Morphological Traits and Leaf Color Parameters.
Morphological observations and measurements included leaf type, leaf texture, leaf color on the adaxial and abaxial sides, leaf size, and leaf thickness. e procedure is repeated 6 times, and the results are averaged.
Fresh leaves are taken, and leaf color in the middle of the upper epidermis was measured by Royal Horticultural Society Colour Chart (RHSCC) and a General Colorimeter (NR10QC, 3nh, Shenzhen). In a daylight condition, lightness (L * ) and two chromatic components a * and b * of the CIEL * a * b * color coordinate were measured. Based on the equations: C * � (a * 2 + b * 2 ) 1/2 and h � arctan (b * /a * ), chroma (C * ) and hue angle (h) were calculated. e procedure is repeated 5 times, and the results are averaged.

Microscopic Observation of Leaf Transverse Sections.
e transverse sections of leaves were prepared according to Li [26]. Leaves were collected and fixed in a formalin-acetic acid-alcohol (FAA; absolute ethyl alcohol: glacial acetic acid, 3 : 1) solution for 30 min; fixed samples were washed three times in 50% ethanol and dehydrated through a series of ethanol concentrations: 60% (30 min), 70% (30 min), 85% (30 min), 95% (5 min), and 100% (5 min, twice). Ethanol in dehydrated samples was then replaced with xylol and paraffin, and samples were embedded and cut into sections of 8-14 μm thickness using a fully motorized rotary microtome (Leica RM2245, Germany). e sections were stained with Safranin Fast Green, washed with 50% ethanol, and then observed with a digital microscope (×10) (Nikon Eclipse E100, Japan).

Microscopic Observation of Leaf Epidermal Cells.
Freehand sections were prepared for leaf shape observations. We rinsed 1.5 × 2 cm leaf samples with distilled water, put them into a 1 : 1 solution of glacial acetic acid and 30% peroxide water, and then placed them in a 60°C incubator for 2-3 h. e samples were rinsed with distilled water, and peels (at least 2 mm long) of the upper and lower surfaces were made with fine-tipped tweezers from the central area of a single leaf and mounted in water, stained with Safranin for 30-60 s, washed, and observed with a digital microscope (Nikon Eclipse E100) [27].

Scanning Electron Microscopy.
Leaf epidermal threedimensional structure was observed by SEM (Hitachi, S-3400N, Japan). 1.5 × 2 cm samples from each of the three samples were cut, respectively, from the middle of each leaf and fixed with 2.5% glutaraldehyde solution for 2 h at room temperature, rinsed with 0.1 mol L −1 phosphate saline buffer, and dehydrated through increasing alcohol series, and then, the alcohol was replaced with isoamyl acetate. e samples were dried naturally, cut into appropriate sizes, and coated using a sputter coater. ey were subsequently observed and photographed using SEM [28].

Measurement of Leaf pH.
e three samples were collected and rinsed, and 5 g of leaves was weighed and cut into pieces. We then added 50 ml distilled water, vibrated the samples for 10 min after soaking for 12 h, and measured the pH of the solution at 30°C using a pH meter (PHS-25, Hongyi, Shanghai) [29]. e procedure is repeated 3 times, and the results are averaged.

Leaf Metal Ion
Measurements. Dried leaves were ground into a fine powder, and a 0.6 g sample of dried material was digested in 5 ml of concentrated HNO 3 and 1 ml of H 2 O 2 , followed by the treatment in a high-performance microwave digestion unit (CEM, Mars, Matthews, NC, USA). Settings used were as follows: time (minutes)/ power (watts)/temperature (°C): 5/1,200/120, 10/1,200/160, and 20/1,200/180. After complete digestion and acid removal, the samples were diluted with ultrapure water for measurement. e procedure is repeated 3 times, and the results are averaged. Sample solutions were analyzed for elements by ICP-MS (NexION 350X, PerkinElmer, Waltham, MA, USA). e parameters for analysis were as follows: plasma power: 1,400 W, plasma flow: 18 l/min, auxiliary flow: 1.8 l/min, and sampling depth: 7.5 mm [30].

Anthocyanin Analysis.
Only blue leaf samples were analyzed for anthocyanin components using ultra-performance liquid chromatography (UPLC).
e methods of extraction were as previously described [31] with some modifications. Anthocyanins were extracted from 0.5 g freeze-dried leaf powder from blue leaves in 25 ml of 2% formic acid/methyl alcohol for 24 h at 4°C. e supernatant was removed and stored under the same conditions. e extraction was repeated once. ese two extractions were merged and subjected to rotary evaporation at 30°C until the anthocyanins were dry. We then added a moderate amount of 2% formic acid solution to dissolve the residue and ethyl acetate to extract the anthocyanins in the aqueous phase.

Flavonoid Analysis.
e extraction methods as previously described by Zhu et al. [27] were referenced with some modifications. Flavonoids were extracted from 1.0 g freeze-dried leaf powder from three 2% formic acid/methyl alcohol samples, after oscillation in ultrasonic cleaners for 20 min at 20°C, and clarified by centrifugation at 12,235 × g for 10 min. e supernatant was then collected. e extraction was repeated twice, and the total extraction volume was 25 ml. After the combined extraction was filtered with a 0.22 μm nylon microporous filter, the solution was tested.
MS was performed on an UPLC-Triple-TOF 5600 Plus System (AB Sciex, Framingham, MA, USA) equipped with an electrospray ionization source (ESI) system. e optimal MS conditions were as follows: scan range m/z of 100-1500. e experiment was conducted in negative ion mode, with a source voltage of −4.5 kV and source temperature of 550°C. e pressure of both gas 1 (air) and gas 2 (air) was set to 50 psi. e pressure of curtain gas (N2) was set to 35 psi. e maximum allowed error was set to ±2 ppm. e collision energy was 40 V, with a collision energy spread of 20 V. Exact mass calibration was performed automatically before each analysis, employing the automated calibration delivery system.

Morphological Traits and Leaf Color Parameters.
Blue S. uncinata leaves were soft and thin; newly formed leaves were grass green. e adaxial side of mature leaves was glaucous, showing iridescence ( Figure 1(a)), whereas the abaxial side was green (Figure 1(b)). Red S. uncinata leaves were hard and crisp; newly formed leaves were also grass green, and the adaxial and abaxial sides of the mature leaf were red (Figures 1(c) and 1(d)). S. kraussiana leaves were thick and soft; newly formed leaves and the adaxial and abaxial sides of mature leaves were all green (Figures 1(e) and 1(f )). e measurement data for the three samples are shown in Table 1. e results indicated that the thickness of red leaves was significantly greater than that of blue leaves, whereas the area of blue leaves was significantly greater than that of red leaves. e color parameters for the three samples are shown in Table 2. e results indicated that their hue angles (h) are all near 0°and belong to the red-purple area. e chroma (C * ) of blue leaves was significantly greater than that of red leaves.
Note: data analysis used Duncan's method, and the data of L * a * b * � mean value ± standard deviation (n � 5); A and B show the different significant differences at P � 0.05 level in SNK test.

Anatomical Structure of Leaf Transverse Section.
e anatomical structure was observed in leaf transverse sections by the paraffin section. e three leaf types shared some similar features: the epidermis is covered by the cuticle; between the upper epidermis and the lower epidermis are irregularly shaped mesophyll cells; and there is no obvious differentiation of palisade and spongy tissue in the mesophyll. ere are large chloroplasts in the mesophyll cells; these are long, narrow, and moniliform distribution. Cell arrangement is loose, with large gaps between the cells,  Note: data analysis used Duncan's method, the data of leaf thickness, and length and width � mean value ± standard deviation (n � 6); A, B, and C show the different significant differences at P � 0.05 level in SNK test.
forming well-developed aeration tissue in the mesophyll cells. e shapes of leaf adaxial and abaxial epidermal cells in blue and red S. uncinata were different from S. kraussiana. e shape of leaf adaxial and abaxial epidermal cells was different in blue leaves: the adaxial epidermis consisted of convex or lens-shaped cells, whereas the abaxial epidermis consisted of long cylindrical cells on the lateral view. In blue leaves, chloroplasts were distributed in the upper and lower epidermal cells, and in mesophyll cells, they were mainly at the bottom of the upper epidermal cells (Figure 2(a)). e shape of leaf adaxial and abaxial epidermal cells in red leaves was similar to that of blue leaves; however, the cells were much more closely aligned, and chloroplasts were significantly reduced (Figure 2(b)). e leaf adaxial and abaxial epidermal cells in S. kraussiana were long and cylindrical on the lateral view. Chloroplasts were distributed in the upper and lower epidermal and mesophyll cells, but primarily in the mesophyll cells (Figure 2(c)).

e Shape of Epidermal Cell.
We used the freehand section to observe epidermal cells from the upper face and SEM photomicrographs to observe the three-dimensional shape of the epidermal cells. We found that the shapes of leaf adaxial and abaxial epidermal cells in blue S. uncinata were different: the adaxial epidermis was irregular circles, with smooth embossment, and by contrast, the abaxial epidermis was a long, wavy, irregular strip, with elongated embossment on the top view (Figure 3(e)). e shape of leaf adaxial and abaxial epidermal cells in red leaves was similar to that of blue leaves (Figure 3(f )). e leaf adaxial and abaxial epidermal cells in S. kraussiana were both shaped as irregular long strips, with elongated embossment on the top view ( Figure 3(g)).

Leaf pH.
Leaf pH values were in the range of 4.5-5.0 (Table 3), in the order blue S. uncinata leaves > red S. uncinata leaves > S. kraussiana leaves. e pH value of blue S. uncinata leaves was significantly greater than that of other two.

Leaf Metal Ion Content.
Mg ions were very abundant in all three leaf types (>3000 mg/kg). e Ca, Mn, Fe, Zn, and Al ion contents were also relatively high (55-1200 mg/kg).
ere was little Cu ion content (<9 mg/kg). Cd ion content was the lowest, ranging from 0.03 to 0.04 mg/kg (Table 4).

Anthocyanin Analysis.
e extract solution in blue S. uncinata leaves contained no anthocyanins ( Figure 4). No anthocyanin ion peaks were observed through mass spectrometry. e result is in line with our conclusion that was previously published: the primary pathway of pigment metabolism in S. uncinata might not be the anthocyanin biosynthesis pathway, but rather the chlorophyll metabolism pathway [15].

Flavonoid Analysis.
ere was good flavonoid separation among the three leaf types (Figures 5-7). We compared the peaks on the chromatographic map and the results of the mass spectrogram, total ion flow diagram, and debris ion mass spectrum analysis to the SciFinder and Reaxys databases and finally inferred 15, 20, and 9 types of flavonoids in blue S. uncinata leaves, red S. uncinata leaves, and S. kraussiana leaves, respectively ( Figure 8 and Table 5).
We find that the vast majority of the flavonoids present (14 compounds) in blue and red leaves are similar; the component specific to blue leaves was genistein, which has an antioxidant effect. Components specific to red leaves were 6,8-C-diglucosylapigenin, apigenin 6,8-di-C-α-L-arabinopyranoside, apigenin 6,8-di-C-α-L-arabinopyranoside isomeride, 3-O-methylamentoflavone, dihydrobilobetin, and dihydrorobustaflavone 4'-methyl ether. ese components are likely to play an adaptive role in high light intensity S. uncinata environments.

e Adaptability of Leaf Morphology and Anatomical
Structure to Environmental Conditions. Leaf thickness is an Journal of Healthcare Engineering important indicator of plant shade tolerance. inner leaves make fern chloroplasts more fully capable of absorbing light energy; improving the photosynthetic efficiency of ferns; and making them better adapted to shaded environments [35]. Morphology traits and measurement data show that blue S. uncinata leaves, including increased leaf area and decreased leaf thickness, are adaptations to low-light environments and weak-light intensity.
Many researchers consider that shade-tolerant trees have a greater ability to change their leaf anatomical structure [36]. Leaf anatomical structure traits such as wax coating, shape of epidermal cells, epidermal thickness, and epidermal  hair play an important role in light absorption, even determining the light use efficiency of the plant [37]. Deepshade plants change their morphology and physiology traits of cells and chloroplasts to fit the low-light conditions [38]. In the paraffin sections of the three leaf types, we observed no clear differentiation of the palisade and spongy tissues in the mesophyll, and mesophyll cells were irregularly distributed. is trait contributes to reducing projection loss of quantum light, allowing the plant to fully utilize limited light to carry out photosynthesis and accumulate organic matter, thus adapting to the shaded environment [35].
Previous research has shown that fern mesophyll cells possess more intercellular space, forming well-developed aeration tissue that can be used to store gases for photosynthesis and respiration, to make up for deficiencies in gas absorption, which is also an adaptation to a shaded environment [35]. e chloroplast is the main site of photosynthesis; in ferns, this structure appears long and narrow and is distributed as a moniliform, reducing the amount of photons penetrating the leaf and raising the utilization rate of quantum light in weak-light conditions, to improve the efficiency of leaf photosynthesis [35]. Plant chloroplasts typically exist in mesophyll cells, but through paraffin section observation, we found that chloroplasts in blue S. uncinata leaves exist mainly in epidermal cells and are larger.
is result is consistent with that of Hébant and Lee [1], who examined transverse leaf sections of S. uncinata by light microscopy, but is different from those of Sheue et al., and they found several chloroplasts in the mesophyll cells and the ventral epidermal cells, while only one single giant chloroplast (bizonoplast, BP) per dorsal epidermal cell [39]. e preferential localization of chloroplasts in the lower part of the epidermal cells in S. uncinata would allow more light to penetrate and reach mesophyll cells [39], which is an adaptation of plants to weak-light environments.

Effects of the Shape of Leaf Epidermal Cells on Blue Coloration in S. uncinata.
e shape of petal epidermal cells has a great influence on the formation of flower color. Noda et al. [40] found that a conical shape in the epidermal cells of petal was believed to enhance light absorption and thus intensified its color, while the flat shape could reflect more incident light and thus lightened its color. Quintana et al. [41] found that in Anagallis, the epidermis contains anthocyanins, and most epidermal cells are flat, with domeshaped and conical cells in the outer layer. Mudalige et al. inspected the perianths of 34 Dendrobium Sw. species and hybrids to clarify the relevance of pigment distribution, the shape of upper epidermal cells, color intensity, perception, and visual texture [42]. Four types of epidermal cell shapes were identified in these Dendrobium flowers: flat, domeshaped, elongated dome-shaped, and papillate [42]. Yue [43] observed using SEM measurements that the epidermal cell shapes of 17 monocotyledon flowers could be grouped into five classes: conical, flat, oval, strip-shaped, and irregular mosaic. at study suggested that convex epidermal cells increased the refraction of light, making petal color appear deeper, and that bulging cells appeared to be more       Journal of Healthcare Engineering conducive to pigment, whereas flat epidermal cells decreased the effect, making their color appear lighter. By comparing leaf paraffin transverse sections, freehand sections, and SEM photomicrographs, we found that the shape of the adaxial epidermis of S. uncinata leaves was not only different from the abaxial epidermis, but also different from the adaxial epidermis of S. kraussiana. e shape appeared convex or lens-shaped on the lateral view and irregular circles with smooth embossment on the top view.
is result corresponds with those of Hébant and Lee [1], who examined the convexly curved upper epidermal cells of S. willdenowii by SEM. e structure increases the proportion of incident light entering the cell, deepens the leaf color [40,43], and therefore may be related to blue leaf coloration.
According to Hébant and Lee [1], the blue color of S. uncinata results from thin-film interference. Which contributes more to blue coloration? Is it thin-film interference or convex or lens-shaped epidermal cells? Do they work complementary? It needs more intensive research.

Effects of the pH on Blue Coloration in S. uncinata.
e color of plant leaves is affected to some extent by the pH within vacuoles, which has a great influence on the coloration of anthocyanins, with varying performance among 1.196 IDA Survey from hong-neg.wiff (sample 1) -Sample002     Journal of Healthcare Engineering different plant species. Tang et al. [44] found that the pH affected anthocyanin synthesis and stability. e degradation rate of anthocyanins has been accelerated by increasing the pH during the process of red turning green in Loropetalum chinense var. rubrum. Studies on the relationship between leaf pigment content and leaf color change in Liquidambar formosana have shown that a reduction in pH was one reason their leaves turned red [45]. Research by Shi [46] indicated that Prunus cerasifera leaf color appeared red in a medium with pH <5, and the stronger the acidity, the more red the pigment. Red color was stable when the pH ranged from 4 to 5, and the solution turned green at pH >5. e stronger the alkalinity, the more green the pigment. Most research results have indicated that anthocyanins present stable red when the pH of the vacuole is lower, and unstable blue occurs as the pH increases. e vacuole is the largest organelle in the mature leaf cell. e pH of leaf juice is often used to approximate the pH of the vacuole [47]. We used this method in our experiment, with results indicating that the pH of blue leaves was greater than that of red leaves. is result is consistent with previous results that the pH of blue flowers was greater than that of red flowers, in Hydrangea macrophylla [48] and Pharbitis nil (Linn.) Choisy [22]. However, for the specific value, the difference between the two pH values was only 0.08, and the difference between those of blue S. uncinata leaves and S. kraussiana leaves is only 0.11, whereas that between blue and red cultivars of H. macrophylla was 0.8 [48] and that between blue in the full-bloom stage and red in the burgeoning stage of P. nil (Linn.) Choisy was 1.1 [22]. erefore, we conclude that blue leaf coloration in S. uncinata has no concern with alkalization of the pH in the vacuole.

Effects of Metal Ion Content on Blue Coloration in
S. uncinata

Effects of Metal Ion Content on Anthocyanin
Coloration. Anthocyanins can be combined with metal ions and flavonoids in a stoichiometric ratio or not, to be assembled into metal pigment complexes [49], and these complexes can affect the coloration of plant leaves. ere have been several studies of metal anthocyanins making color tend toward bluish, and these concentrate primarily on Mg, Al, Fe [23-25, 48, 50], Ca [30], and Mn [30]. Metal ions have a stable and protective effect on anthocyanins, and the pigments are often chelated if the cell sap contains metal ions such as Al, Fe, Mg, or Mo. In particular, anthocyanins, which change their color to some degree, often tend toward purple after chelation [51]. In our experiment, Mg, Ca, Mn, Fe, Zn, and Al ion contents were all relatively high in the three leaf types; however, there was no anthocyanin in blue S. uncinata leaves, so we concluded that blue leaf coloration in S. uncinata was unrelated to metal chelation with anthocyanin. Mg is a part of the molecular composition of chlorophyll; chlorophyll formation will be affected if it is lacking. e concentration of Mg 2+ influences the activity of CHLH [53]. Fe is necessary for protochlorophyllide formation; Mgprotoporphyrin IX and Mg-protoporphyrin IX methyl ester accumulate when short of Fe, and protochlorophyllide cannot form chlorophyll [53,54]. Chlorophyll synthesis is also affected by the content of Cu, Mn, and other ions.

Effects of Metal Ion
In our experiment, the Cu content was not high, but those of Mg, Fe, and Mn were very high, particularly that of Mg, with content reaching 4311.7 mg/kg, which was 1.24fold higher in blue leaves than in red leaves. We speculate that such high Mg levels may be associated with chlorophyll synthesis in S. uncinata.

Effects of Anthocyanin and Copigment on Blue Coloration in S. uncinata.
Copigments, often flavone and flavonols, are the two branches of the flavonoid metabolic pathway [55]. Combined with anthocyanins, they can stabilize pigments, and the compounds they form will influence the coloration of anthocyanins to some degree [56]. Research by Li [57] found that the effect of copigments turned purple or pink delphinidin flowers blue. Malvidin-3-glucoside is the basic anthocyanin of Primula sinensis; the flower appears purple when it is combined with flavonol, but appears garnet when not combined [58]. Under certain conditions, the larger the molar ratio of flavonols and anthocyanins, the more significant the copigmentation effect.
In our research, we detected anthocyanins in a preliminary experiment by measuring absorbance values using an enzyme-linked immunosorbent assay (ELISA). Although the overall content was low, and the highest content was only 1.2 pigment units [14], we still initially speculate that blue leaf coloration in S. uncinata may be related to delphinidin in anthocyanins, or a copigment with anthocyanins. We did not detect anthocyanins in blue leaves using liquid chromatography-MS (this is accordant with the conclusion that compared with the anthocyanin biosynthesis pathway, the chlorophyll metabolism pathway may be the primary pigment metabolism pathway of S. uncinata [15], although   there were copigments such as flavone). If anthocyanins are not present, the copigmentation of flavone cannot occur. erefore, we infer that blue leaf coloration in S. uncinata was not caused by copigmentation of anthocyanins.

Conclusion
rough comparison of leaf paraffin transverse sections, freehand sections, and SEM photomicrographs, we found that the shape of the adaxial epidermis of S. uncinata leaves was convex or lens-shaped on the lateral view and irregular circles with smooth embossment on the top view. ese shapes were different from those on the abaxial epidermis and the adaxial epidermis of S. kraussiana leaves. We speculated these structures increase the proportion of incident light entering the cell, deepening the leaf color, and therefore may be related to blue leaf coloration.
rough comparison of previously published values of leaf pH and metal ion content, anthocyanins, and flavonoids with those of the three leaf types in our study, we found that leaf pH was similar among the leaf types and that the leaves all contained high levels of metal ions such as Mg, Fe, Mn, and copigments such as flavones. However, because there was no anthocyanin present in blue S. uncinata leaves, we conclude that blue leaf coloration in S. uncinata was not related to the three hypotheses of blue coloration: alkalization of vacuole pH, metal chelation, and copigmentation with anthocyanins.
Data Availability e raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Conflicts of Interest
e authors declared that they have no conflicts of interest regarding this work.