Edible flowers are used to add special colors and flavors to dishes. The color of edible flowers is an important factor that influences consumer purchasing  . Many countries have dietary cultures that utilize edible flowers, and we can see them in side dishes, cakes, drinks, and so on. The use of edible flowers is very popular in European countries, such as England, France, Spain, Italy, etc. In Japan, the yellow-colored Chrysanthemum morifolium is a very popular edible flower that is used as an ornament or is sometimes cooked by boiling and seasoned with soy source.
Studies on edible flowers have mainly analyzed the chemical compositions   , identified the chemical structures that result in their distinct colors   , and evaluated their physiological functions     . For example, Rop et al. researched the chemical compositions of 12 edible flowers and found that Chrysanthemum, Dianthus, and Viola have the highest mineral content, of such minerals as potassium or sodium  . Pires et al. prepared hydromethanolic extracts from four flowers of different colors: Dahlia mignon, Rosa damascene, Calendula officinalis L., and Centaurea cyanus L., and identified the main compounds as naringenin-3-O-glucoside, quercetin-3-O-glucoside, isorhamnetin-3-O-rutinoside, and apigenin-glucuronide, respectively  . In the same reports, they also found that their bioactive potentials, referring to the anti-oxidant, anti-proliferative, and anti-bacterial capacities, were different based on the polyphenolic compositions.
Concerning the physiological functions of edible flowers, anti-oxidative activity has been well documented, while other activities have been less reported    . In the few reports regarding―physiological functions other than the anti-oxidative activity―anti-bacterial   , anti-adipogenic  , anti-allergic  , and anti-tumor activities  of edible flowers have been reported. Especially, as the interest in human health increases every year, it will be important to research the effects of edible flowers on lifestyle-related diseases, such as diabetes, hypertension, and adiposity.
As an abrupt increase in the blood glucose level is one of the risks of becoming diabetic, inhibition of α-amylase or α-glucosidase activities, which produce glucose from polysaccharides in the small intestine, is one of the effective strategies to protect against this disease. Hypertension is partially induced by the renin-angiotensin-aldosterone system. Angiotensin I converting enzyme (ACE) is a rate-controlling enzyme. It hydrolyzes angiotensin I to produce angiotensin II, which promotes the secretion of aldosterone in the adrenal gland. Aldosterone increases sodium reabsorption in the kidney, which results in hypertension. Lipase hydrolyzes adipose (triacylglycerol) to release monoaclyglycerol and free fatty acids. As the absorption of adipose starts with the hydrolysis of triacylglycerolin the small intestine, the inhibition of lipase activity leads to suppression of the absorption of adipose.
In this study, 70% ethanol extracts were prepared from 6 edible flowers of different color: Chrysanthemum morifolium (Chry. yellow), Rosa miniature (Rosa, red), Dendrobium phalaenopsis (Dend., purple), Viola x wittrockiana cv. fluna (Fluna, white), Viola x wittrockiana cv. pansy (Pansy, orange), and Primula x juliana (Prim., blue). These flowers are available in commercial. In addition, the physiological functions of the extracts for their anti-diabetics, anti-hypertensive, and anti-adiposity, as well as the anti-oxidative activities were evaluated and compared in vitro.
2. Experimental Procedures
2.1.1. Edible Flowers
Chry. was purchased from a local supermarket in Shobara, Hiroshima, Japan. Rosa and Dend. were obtained by mail-order from Marche Co., Ltd. (Shizuoka, Japan). Fluna, Pansy, and Prim. were obtained by mail-order from Maro Co., Ltd (Osaka, Japan).
α-Amylase from Bacillus subtilis (200 U/mg) and α-glucosidase from yest (75 U/mg) were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). ACE from rabbit lung (2.0 U/mg) and lipase from porcine pancreas (100 - 500 U/mg) were purchased from Sigma Chemical Co. (St. Louis, MO, USA).
Phenol reagent (Folin-Ciocalteu reagent) was purchased from the Kishida Chemical Co., Ltd. (Osaka, Japan). 1,1-diphenyl-2-picrylhydrazyl (DPPH), α-tocopherol, starch, hippuryl-histidyl-leucine, p-nitrophenyl laurate, chlorogenic acid, l-carnosineand captopril were purchased from Wako Pure Chemical Industries, Ltd. Ethyl gallate, 3,5-dinitrosalicylic acid, 4-nitrophenyl α-d-glucopyranoside, and o-phthalaldehyde were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Triton®X100 was purchased from nacalai tesque (Kyoto, Japan). Tannic acid was purchased from Sigma Chemical Co. Other reagents and solvents used in this study were of analysis grade.
2.2.1. Preparation of 70% Ethanol Extracts
Each flower petal (5 g) was cut smaller than 1 cm2 in area, and 100 mL of 70% ethanol was added. The mixture was stirred with a magnetic stirrer for 3 h in the dark. The solution was filtered with filter paper (Advantec® No.1), and the filtrate was evaporated to obtain the extract. The extract was dissolved in 70% ethanol and adjusted to an adequate concentration.
2.2.2. Polyphenol Content
The polyphenol content of the extracts was determined spectrophotometrically using Folin-Ciocalteu’s reagent, using a modified method of Folin and Ciocalteu  .
About 0.5 mL of each extract was mixed with 0.2 mL of the Folin-Ciocaltu reagent (1:1 v/v in distilled water) and 0.5 mL of 20% of sodium carbonate. Next, 4.3 mL of distilled water was added to each sample. After 30 min, the absorbance was measured at 765 nm with a spectrophotometer (U-2001, Hitachi High-Technologies Co., Tokyo, Japan). Ethyl gallate was used as a standard for the calibration curve (y = 0.073x + 0.0475, R2 = 0.9967), showing the relation of polyphenol content of sample solution and the absorbance of sample solution at 765 nm.
2.2.3. Polyphenol Composition
The polyphenol composition of the extracts was analyzed by high performanceliquid chromatography (HPLC)  . The HPLC system consisted of a Waters 2695 Separations module and a Waters 2487 Dual λ Absorbance Detector. An Inert Sustain C18 (4.6 × 250 mm, 5 μm, GL Science) column was used. The extracts were eluted with a gradient elution of mobile phase A (5% acetonitrile in 0.035% trifluoroacetic acid) and B (80% acetonitrile in 0.025% trifluoroacetic acid), where B increased from 10% in 10 min to 50% by 30 min, maintained at 50% for 5 min, returned to the initial condition (10%) in 5 min, and remained there for 5 min before injection. The flow rate was 0.7 mL/min, and the column temperature was 30˚C. Elutes were detected at 280 nm.
2.2.4. Anti-Oxidative Activity
The anti-oxidative activity was evaluated as the ability to trap a DPPH radical. In this assay  , 2 mL of extract was mixed with 2 mL of 0.1 M acetate buffer (pH 5.5) and 1 mL of 0.5 mM DPPH ethanol solution. Nest, the solutions were allowed to stand in a dark place at room temperature for 30 min. Absorbance was measured at 517 nm with a spectrophotometer (U-2001). The relative DPPH radical scavenging activity of the extract was evaluated and converted into the α-tocopherol equivalent (α-tocopherol = 1) with a calibration curve (y= −0.0067x + 0.8109, R2 = 0.9969).
2.2.5. Anti-Diabetic Activity
The anti-diabetic activity was evaluated as the ability to inhibit α-amylase  and α-glucosidase activities  . For the α-amylase inhibition system, 50 μL of extract was mixed with 50 μL of α-amylase solution (6 U/mL). After 10 min of incubation at 30˚C, 800 μL of 0.5% starch/1.0% NaCl in 0.1 M phosphate buffer (pH 6.8) was added. After 20 min of incubation at 30˚C, 125 μL of 2 M NaOH and 125 μL of 3,5-dinitrosalicylic acid was added. After 10 min of incubation at 99˚C, the absorbance of the solution at 540 nm was measured using a micro-plate reader (Varioskan Flash, Thermo Fisher Scientific, Waltham, MA, USA). The inhibition was calculated from Equation (1), where S, B, Cs, and Cb means the absorbance of the sample solution, the solution without enzyme, the solution without sample, and the solution without enzyme and sample, respectively. The inhibition activity was expressed as the IC50 from the inhibition curve (Figure 1).
Figure 1. α-Glucosidase-inhibition curve obtained from the extract from Chry. IC50 calculated from the equation inserted in this figure was 0.41 (μg/mL).
For the α-glucosidase inhibition system, 10 μL of extract was mixed with 40 μL of α-glucosidase solution (32 mU). After 5 min incubation at 37˚C, 950 μL of 0.7 mM 4-nitrophenyl α-d-glucopyranoside in 0.1 M phosphate buffer (pH 6.8) containing 100 mM NaCl was added. After 15 min of incubation at 37˚C, 500 μL of 1 M Tris was added, and the -absorbance of the solution at 400 nm was measured using a micro-plate reader (Varioskan Flash). Calculation of the IC50 was the same as the α-amylase inhibition system.
2.2.6. Anti-Hypertensive Activity
The anti-hypertensive activity was evaluated as ability to inhibit ACE  . Fifty microliters of the extract was mixed with 100 μL of ACE solution (10 mU/mL). After 10 min of incubation at 37˚C, 25 μL of 25 mM hippuryl-histidyl-leucine was added. After 40 min of incubation at 37˚C, 50 μL of 1 M NaOH was added, followed by 10 μL of 0.2% o-phthalaldehyde. After 15 min of incubation at room temperature, 15 μL of 3.6 M phosphate was added, and the fluorescence intensity (Ex: 360 nm, Em: 460 nm) was measured using a micro-plate reader (Varioskan Flash). Calculation of the IC50 was the same as the α-amylase inhibition system.
2.2.7. Anti-Adiposity Activity
The anti-adiposity activity was evaluated as the ability to inhibit lipase  . Fifty microliters of the extract was mixed with 50 μL of lipase solution (2 mg/mL) in 50 mL Tris-HCl buffer (pH 8.5). After 5 min of incubation at room temperature, 50 μL of 5 mM p-nitrophenyl laurate in dimethyl sulfoxide was added. After 5 min of incubation at 37˚C, 50 μL of 10% Triton®X100 was added, and the absorbance of the solution at 410 nm was measured using a micro-plate reader (Varioskan Flash). Calculation of the IC50 was the same as the α-amylase inhibition system.
2.2.8. Statistical Analysis
All experiments were conducted in triplicate. Data are shown as the means ± standard deviation. A significant difference was determined by Scheffe’s test (P < 0.01).
3. Results and Discussion
3.1. Yield, Colors, and Polyphenol Content of the Extracts
The yield, colors, and polyphenol content of the extracts are summarized in Table 1. The yields varied from 3.0% - 4.8%, based on their origin. As this study used fresh flowers, the yields were much lower those that have used dried flowers. For example, an approximate ~17% yield was obtained from a water or ethanol extraction of the dried daylily flower  . In addition, a 22% yield was obtained by methanol extraction of dried Rosa damascena  . However, Loizzo et al. reported 4.4% - 10.5% yield that was obtained by the ethanol extraction of several fresh edible flowers  . These values were slightly higher than those in this study.
The colors of the extracts were in accordance with their petal color before extraction, except for that of Fluna. As Fluna is primarily white color, but has a yellow color in the center of the petal, the yellow appeared in the extracts.
The polyphenol content of the extracts was highest for Fluna (18.6%) and lowest for Rosa (10.8%). Rocio et al.  reported that a polyphenol content of 15.2% was obtained for the Viola x wittrockiana cv. pansy, which is the same as the results of this study (14.4%). On the other hand, Özkan et al. reported a polyphenol content of 27.6% for Rosa damascena extracts  . It has been suggested that the polyphenol content differs even for the same genus.
3.2. Anti-Oxidative Activity of the Extracts
All of the extracts exerted more than 2-fold higher α-anti-oxidative activity, compared to α-tocopherol (Figure 2). Especially, the extracts from Fluna, which had highest polyphenol content (Table 1), exerted the highest activity. However, among the other extracts, there was no significant anti-oxidative activity. As the results showed that the polyphenol content of these extracts were varied in significance (Table 1), it was suggested that the composition of these polyphenols was varied. For example, Rosa, which had lowest polyphenol content, had the kind of polyphenols that have higher radical scavenging activity than the others.
3.3. Anti-Diabetics, Anti-Hypertensive, and Anti-Adiposity Activities of the Extracts
The anti-diabetic, anti-hypertensive, and anti-adiposity activities of the extracts were evaluated as the activity of inhibition of the corresponding enzyme. For anti-diabetes, the inhibition of α-amylase and α-glucosidase was evaluated. For anti-hypertension and anti-adiposity, the inhibition of ACE and lipase was evaluated, respectively. Table 2 summarizes corresponding enzyme-inhibitory activities of the extracts as IC50 values.
Table 1. Yields, colors, and polyphenol content of the extracts from 6 edible flowers.
adDifferent letters are significantly different in each other (Scheffe’s test, p < 0.01).
Figure 2. Anti-oxidative activity of the extracts from 6 edible flowers. a, b: different letters are significantly different in each other (Scheffe’s test, p < 0.01).
Table 2. Enzyme-inhibitory activities (IC50) of the extracts from 6 edible flowers.
For anti-diabetes, the extracts from Rosa, Dend., Fluna, and Pansy showed dose-dependent inhibition α-amylase in which, Dend. showed the highest inhibition with an IC50 value of 161.9 μg/mL, although it was only 1/8 of the activity of tannic acid, a positive control. On the other hand, Chry., Fluna, Pansy, and Prim. showed a dose-dependent inhibition of α-glucosidase, in which Chry. showed the highest inhibition with an IC50 value of 0.41 μg/mL that was 10-fold higher than chlorogenic acid, a positive control. Loizzo et al. evaluated the α-amylase/α-glucosidase inhibitory activities of 8 edible flower extracts and reported that 7 of them excluding Sambucus nigra, showed inhibition of both of the two enzymes  . In this study, it was unique that Rosa and Dend. showed selective α-amylase inhibitory activity, and Chry. and Prim. showed selective α-glucosidase inhibitory activity. The inhibition mode of methanol/ethanol extracts of Rosa damascena Mill. and banana flower on α-glucosidase were non-competitive mode   . Further studies are required to determine the mechanism of inhibition of the extracts used in this study.
For anti-hypertension, the extracts, except for the extract from Dend., showed dose-dependent inhibition of ACE, in which Chry. showed the highest inhibition with an IC50 value of 69.1 ng/mL that was 1000-fold higher than carnosine, a positive control. As carnosine is a food derivative peptide, we evaluated the activity of captopril, an ACE-inhibitory drug and revealed that its IC50 was 1.9 ng/mL. The inhibitory activity of peptides for ACE is well known, and a lot of corresponding peptide sequences have also been identified  . However, the inhibition level of these peptides is in the order of μg/mL. To our knowledge, few papers describing the inhibitory activity of Chrysanthemum toward ACE have been published. Kim et al.  reported the ACE inhibitory activity of a water extract from Chrysanthemum boreale Makino with an IC50 value of 440 ng/mL. They identified its structure as guanosine. Further studies are required to identify the chemical structures and inhibition mechanism of the compounds evaluated in this study.
For anti-adiposity, none of the extracts showed inhibitory activity for lipase in this assay.
3.4. Polyphenol Compositions of the Extracts
The HPLC chromatograms obtained from detection at 280 nm, showed the polyphenol composition of the extracts were different for each original flower (Figure 3). However, Fluna and Pansy, which are of the same genus, were alike in the chromatogram pattern. As these two flowers have different colors, but showed a similar pattern of physiological function (Table 2), the peak at a retention time of 20 min may be a compound that exerts the physiological function. Of course, it is difficult to identify the compounds present in each chromatogram by only this HPLC analysis. It is necessary to identify the corresponding compound in a future study.
Figure 3. HPLC chlomatograms obtained from each sample. Concentration of the samples applied into HPLC were 4.6 mg/mL (Chry.), 0.46 mg/mL (Rosa), 3.2 mg/mL (Dend.), 0.6 mg/mL (Fluna), 0.24 mg/mL (Pansy), and 0.38 mg/mL (Prim.), respectively.
As the color of edible flowers is an important factor that influences consumer purchasing, this study selected 6 edible flowers of different colors (yellow, red, orange, white, purple, and blue) and compared the physiological functions of their 70% ethanol extracts in vitro. Their anti-oxidative activities were 2 - 3-fold higher than α-tocopherol, which was not related to their color. All of the extracts from the edible flowers used in this study showed α-amylase/α-glucosidase inhibitory activity. Except for the extract from Dend., the extracts showed ACE inhibitory activity. Especially, Chry. showed strong inhibitory activity of ACE. Finally, none of them showed lipase inhibitory activity. These results suggested that daily consumption of the edible flowers used in this study will help prevent from diabetes and hypertensive diseases.
We would like to thank Editage (https://www.editage.jp/) for English language editing.
 Rop, O., Mlcek, J., Jurikova, T., Neugebauerova, J. and Vabkova, J. (2012) Edible Flowers—A New Promising Source of Mineral Elements in Human Nutrition. Molecules, 17, 6672-6683.
 Rocio, G.B., María, J.P, Cristina, L.R., Garcia-Alonso, F.J. and Inmaculada, N.G. (2018) Chemical Composition of the Edible Flowers, Pansy (Viola wittrockiana) and Snapdragon (Antirrhinum majus) as New Sources of Bioactive Compounds. Food Chemistry, 252, 373-380.
 Pires, T.G.S.P., Dias, M.I., Barros, L., Galhelha, R.G., Alves, M.J., Oliveira, M.B.P.P., Santos-Buelga, G. and Ferreira, I.G.F.R. (2018) Edible Flowers as Sources of Phenolic Compounds with Bioactive Potential. Food Research International, 105, 580-588.
 Maruyama, T., Yoda, H., Kobori, M., Shinmoto, H. and Tsushida, T. (2002) Evaluation of Three Antioxidants and Their Identification and Radical Scavenging Activities in Edible Chrysanthemums. Journal of the Japanese Society for Horitulcultural Science, 71, 236-242.
 Özkan, G., Sağdiç, O., Baydar, N.G. and Baydar, H. (2003) Antioxidant and Antibacterial Activities of Rosa Damascene Flower Extracts. Food Science and Technology International, 10, 277-281.
 Kim, G.C., Kim, J.S., Kim, G.M. and Choi, S.Y. (2017) Anti-Adipogenic Effects of Tropaeolum majus (Nasturtium) Ethanol Extract on 3T3-L1 Cells. Food & Nutrition Research, 61, 1339555.
 Toyoda, M., Tanaka, K., Hoshino, K., Akiyama, H., Tanimura, A. and Saito, Y. (1997) Profiles of Potentially Antiallergic Flavonoids in 27 Kinds of Health Tea and Green Tea Infusions. Journal of Agricultural and Food Chemistry, 45, 2561-2564.
 Yasukawa, K., Akihisa, T., Oinuma, H., Kaminaga, T., Konno, H., Kasahara, Y., Tamura, T., Kumaki, K., Yamanouch, S. and Takido, M. (1996) Inhibitory Effect of Taraxastane-Type Triterpenes on Tumor Promotion by 12-O-tetradecanoylphorbol-13-acetate in Two-Stage Carcinogenesis in Mouse Skin. Oncology, 53, 341-344.
 Mao, L.C., Pan, X., Que, F. and Fang, X.H. (2006) Antioxidant Properties of Water and Ethanol Extracts from Hot Air-Dried and Freeze-Dried Daylily Flowers. European Food Research and Technology, 222, 236-241.
 Tai, Z., Cai, L., Dai, L., Dong, L., Wang, M., Yang, Y., Cao, Q. and Ding, Z. (2011) Antioxidant Activity and Chemical Constituents of Edible Flower of Sophara Viciifolia. Food Chemistry, 126, 1648-1654.
 Xiong, L., Yang, J., Jiang, Y., Lu, B., Hu, Y., Zhou, F., Mao, S. and Shen, C. (2014) Phenolic Compounds and Antioxidant Capacities of 10 Common Edible Flowers from China. Journal of Food Science, 79, 517-525.
 Yu, J., Ahmedna, M. and Goktepe, I. (2005) Effects of Processing Methods and Extraction Solvents on Concentration and Antioxidant Activity of Peanut Skin Phenolics. Food Chemistry, 90, 199-206.
 Yamamoto, Y., Hiyama, S., Takase, Y., Kadowaki, A. and Hara, S. (2014) Effects of Antioxidants and Additional Emulsifiers on the Stability of Emulsified Milk Fat in the Photo/Radical Oxidation System. Journal of Oleo Science, 63, 893-901.
 Matsui, T., Ueda, T., Oki, T., Sugita, K., Terahara, N. and Matsumoto, K. (2001) α-Glucosidase Inhibitory Action of Natural Acylated Anthocyanins. 1. Survey of Natural Pigments with Potent Inhibitory Activity. Journal of Agricultural and Food Chemistry, 49, 1948-1951.
 Ishiguro, K., Sameshima, Y., Kume, T., Ikeda, K.I., Matsumoto, J. and Yashimoto, M. (2012) Hypotensive Effect of a Seetpotato Protein Digest in Spontaneously Hypertensive Rats and Purification of Angiotensin I-Converting Enzyme Inhibitory Peptides. Food Chemistry, 131, 774-779.
 Gupta, N., Rathi, P. and Gupta, R. (2002) Simplified para-Nitropheny Palmitate Assay for Lipase and Esterases. Journal of Analytical Biochemistry, 311, 98-99.
 Loizzo, M.R., Pugliese, A., Bonesi, M., Tenuta, M.C., Manichini, F., Xiao, J. and Tundis, R. (2015) Edible Flowers: A Rich Source of Phytochemicals with Antioxidant and Hypoglycemic Properties. Journal of Agricultural and Food Chemistry, 64, 2467-2474.
 Gholamhoseinian, A., Fallah, H. and Sharifi far, F. (2009) Inhibitory Effect of Methanol Extract of Rosa damascena Mill. Flowers on α-Glucosidase Activity and Postprandial Hyperglycemia in Normal and Diabetic Rats. Phytomedicine, 16, 935-941.
 Ramu, R., Shirahatti, P.S., Zameer, F., Ranganatha, L.V. and Prasad, M.N.N. (2014). Inhibitory Effect of Banana (Musa sp. var. Nanjangud rasa bale) Flower Extract and Its Constituents Umbelliferone and Lupeol on α-Glucosidase, Aldose Reductase and Glycation at Multiple Stages. South African Journal of Botany, 95, 54-63.
 Kim, J., Lee, S.H., Sun, N., Choung, D.H., Kim, W.K., Lee, S. and Song, K.B. (2006) Isolation of an Angiotensin Converting Enzyme Inhibitory Substance from Chrysanthemum boreale Makino. Food Chemistry and Toxicology, 68, 816-819.