Introduction

At present, natural products have been receiving substantial attention. Many active ingredients used in cosmetics and topical applications are derived from natural products1. This increase in attention is because consumers believe natural active ingredients are more effective and safer than synthetic ones2. One interesting plant-based ingredient is Chrysanthemum × morifolium (Ramat.) Hemsl. (C. morifolium) flowers3,4,5,6. C. morifolium is a species of perennial plant from the family Asteraceae. C. morifolium flowers have been used as tea or folk medicine in China as a traditional medicine to reduce fever and improve eyesight. In addition, these flowers have been dispensed to treat cold and eliminate heat3,4,7,8.

Previous studies have found and reported the presence of several phytochemicals from C. morifolium that are related to its activities, including phenolic compounds such as hydroxycinnamoylquinic acids, e.g., chlorogenic acid, caffeic acid, and dicaffeoylquinic acids (i.e., 3,5-,1,5- and 4,5-di-caffeoylquinic acids) and flavonoids, e.g., luteolin derivatives (i.e., luteolin, luteolin 7-O-glucoside, and luteolin-7-O-glucuronide)3,5,6,7,9,10. It has also been reported that water or methanolic extracts of C. morifolium flower contain chlorogenic acid, caffeic acid, and luteolin. In addition, Tao et al. reported that other substances, such as flavonoids, e.g., quercetin, acacetin, and apigenin, were found in the water extract of C. morifolium. The different amounts of these flavonoids depend on the genotypes and harvesting locations of C. morifolium flowers7.

Concerning the bioactivities of C. morifolium flower extracts, previous studies have revealed that C. morifolium flower extracts demonstrate cardiovascular protection, neuroprotection, anticancer, hepatoprotective, antioxidant, anti-glycation, anti-inflammatory, antibacterial, and antifungal properties3,5,6,10,11,12. Moreover, Siqun et al. reported the antitumor activity of a C. morifolium flower extract using an MTT assay. It was found that proanthocyanidins extracted from C. morifolium flowers could inhibit the proliferation of esophageal cancer (Eca-109 cells), cervical cancer (HeLa cells), mouse ascites hepatomas (H22 cells), and particularly H22 cells11.

Recent research has highlighted the remarkable potential of C. morifolium extracts in the cosmetic field due to their significant bioactive compounds, particularly phenolics and flavonoids, which exhibit potent antioxidant properties and tyrosinase inhibitory activities13. Researchers examined different extraction solvents, including ethanol, ethyl acetate, n-hexane, and water. It was found that ethanol-based extracts demonstrated exceptional DPPH radical scavenging activity at 96.2%, while water-based extracts exhibited notable tyrosinase inhibition at 39.34% when tested at 1.0 mg/mL13. Another study by Gao et al. examined different extraction techniques for C. morifolium cv. Fubaiju revealed that hot water extraction provided the highest antioxidant benefits, whereas enzyme-assisted methods proved most effective for inhibiting glycation products and AGE formation14. These findings indicate that the method and solvent used for extraction significantly affect the bioactivities of the obtained extract. Based on the above findings, C. morifolium extracts demonstrate potential for skin applications, particularly in anti-aging and skin-brightening products, due to their proven antioxidant, anti-tyrosinase, and antiglycation capabilities. To our knowledge, hydroglycolic extracts of purple, yellow, and white C. morifolium flowers have not been reported or compared. Therefore, we selected a hydroglycolic solvent system containing propylene glycol (PG) and water over traditional hydroalcoholic solvents due to its advantageous physicochemical properties. The dielectric constant of PG (32.1) is comparable to that of methanol (32.7) and higher than that of ethanol (24.5). indicating similar solvent characteristics and extraction capabilities. PG serves multiple purposes, acting as both a solvent and a humectant, and is commonly used in cosmetics. This is in contrast to methanol and ethanol, which can harm the skin’s barrier function and lead to dryness. Furthermore, plant extracts obtained using a hydroglycolic solvent system can be directly used in cosmetic formulations, eliminating the need for additional processing steps to remove the solvent, as methanol is limited in cosmetic products. Therefore, this study aimed to evaluate the phytochemicals, bioactivities, and safety regarding skin irritation and sensitization of the hydroglycolic extracts of purple, yellow, and white C. morifolium flowers for cosmetic purposes.

Materials and method

Materials

Fresh C. morifolium flowers of various colors (purple, yellow, and white) were obtained from Saraburi province, Thailand, and identified by a taxonomist to confirm the genus and species. Herbarium specimens (PBM05512, PBM05513, and PBM05514) were collected at the Faculty of Pharmacy, Mahidol University. Aluminum chloride, sodium nitrite, catechin, Folin-Ciocalteu’s reagent, gallic acid, D-glucose, monobasic sodium phosphate dihydrate, phosphate buffer, potassium persulfate, and 2,2'-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid (ABTS) were purchased from Merck (Germany). Sodium hydroxide and sodium carbonate were obtained from Ajax Finechem (Australia). Luteolin, caffeic acid, chlorogenic acid, and 6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid (Trolox) were acquired from Sigma-Aldrich Inc. (Germany). 2,2-Diphenyl-1-picrylhydrazyl (DPPH), aminoguanidine, and ascorbic acid were obtained from Sigma-Aldrich Inc (USA). The EnzChek gelatinase/collagenase assay kit (E-12055) was purchased from Thermo Fisher Scientific (USA). The reconstructed human epidermis (RhE) was purchased from Episkin (France). ADRA kit was purchased from Fujifilm Wako Pure Chemical Corporation (Japan). Dibasic sodium phosphate dihydrate was obtained from Scharlab S.L. (Spain). Acetonitrile and methanol were of analytical grade.

Plant extraction

Before extraction, C. morifolium flowers were washed with water for 10–15 min to remove impurities and subsequently dried at room temperature (30 ± 2 °C) for two hours. The petals were left at this temperature to allow the moisture on the petals to evaporate completely. Following this, the petals were carefully separated from the flowers and extracted using the maceration. A hydroglycolic solvent containing PG and distilled water (1:1 v/v) or 50% PG was used for extraction. Seventy grams of fresh petals were macerated in 1000 mL of hydroglycolic solvent for seven days at room temperature (30 ± 2 °C). The resulting macerated mixture was filtered through Whatman’s No. 1 filter paper using vacuum filtration to remove insoluble residues. All flower extracts were stored under refrigeration (2–8 °C) and protected from light before use.

Characterization of the C. morifolium flower extract

Nonvolatile matter (NVM) and density of the C. morifolium flower extracts

The percent NVM of each C. morifolium flower extract was evaluated by placing 1 g of the extract on a watch glass and drying at 105 °C for two hours in a hot air oven until no change of remaining weight was observed. The percent NVM of each extract on the watch glass was weighed and calculated using Eq. (1).

$${\text{NVM}}\;\left( \% \right) = \frac{{{\text{Weight}}\;{\text{of}}\;{\text{extract}}\;{\text{after}}\;{\text{drying}}}}{{{\text{Weight}}\;{\text{of}}\;{\text{extract}}\;{\text{before}}\;{\text{drying}}}} \times 100$$
(1)

A hydrometer was used to determine the densities of the C. morifolium flower extracts. The procedure involved filling a graduated cylinder with 100 mL of the extracts. Then, the hydrometer was gently inserted into the extracts. The hydrometer was allowed to reach static equilibrium. Once stabilized, the density of the extracts was read and recorded.

Total phenolic contents (TPC) of the C. morifolium flower extracts

The TPC of C. morifolium flower extracts was determined by the Folin-Ciocalteu method with some modifications15,16. A calibration curve was constructed using gallic acid as the standard, with concentrations ranging from 1.6 to 40 µg/mL. Each extract (20 µL) and 10% v/v Folin-Ciocalteu’s reagent (50 µL) were added to a transparent 96-well plate and allowed to react for 3 min. The reaction mixtures were neutralized by adding 7.5% w/v sodium carbonate (Na2CO3) (80 µL). After incubation for 30 min at room temperature, the absorbance of the resultant blue complex was measured spectrophotometrically at 765 nm using an infinite M200 microplate reader (Tecan, Männedorf, Switzerland). The TPC of the samples was calculated and expressed as mg of gallic acid equivalents (GAE)/g dry extract. All experiments were performed in triplicate.

Total flavonoid contents (TFC) of the C. morifolium flower extracts

The aluminum chloride colorimetric method was performed to evaluate the TFC of C. morifolium flower extracts17. The samples or standard reference (catechin) (20 µL), DI water (80 µL), and 5% sodium nitrite (6 µL) were added to a transparent 96-well plate and mixed well. After shaking for 5 min, a 10% aluminum chloride solution (6 µL) was added. Then, 1 M sodium hydroxide (40 µL) was added, followed by shaking for 10 min at 25 °C. The absorbance at 510 nm was measured with the infinite M200 microplate reader. The TFC of the samples is expressed as mg of catechin equivalents (CE)/g dry extract.

Qualitative and quantitative analysis of the chrysanthemum extracts

Based on previous studies, chlorogenic acid, caffeic acid, and luteolin have been found and reported to be present in C. morifolium flowers3,7,10,18,19. Therefore, the contents of these compounds in the extracts were qualitatively and quantitatively evaluated using HPLC. All samples were analyzed on a Shimadzu 10D instrument equipped with a UV detector. Analysis was performed with a reversed-phase C18 Platisil 5 µm ODS column (5 μm, 250 mm × 5 mm) and run with a mobile phase composed of 0.1% phosphoric acid and acetonitrile (72:28% v/v) at a flow rate of 1.0 mL/min. The UV detector was set at 350 nm. The injection volume was 25 μL. The run time was 25 min. Stock solutions of each reference standard (chlorogenic acid, caffeic acid, and luteolin) were prepared in methanol. The analytical method was validated regarding linearity, recovery, and precision. The limit of quantification (LOQ) and limit of detection (LOD) were also determined.

Antioxidant activity

Antioxidant activity by the DPPH radical scavenging activity assay

The antioxidant activity of the flower extracts was evaluated using a previously reported DPPH assay with slight modifications20,21. A DPPH solution (0.2 mmol/L) was prepared by dissolving in methanol. Ascorbic acid was used as a positive control for evaluating DPPH radical scavenging activity. Fitty microliters of the ascorbic acid solution (0.001–0.007 μg/mL) or different concentrations of C. morifolium flower extracts (20–200 μg/mL) were placed in the transparent 96-well plates, and 50 µL of DPPH solution or methanol (for a negative control) was added. Consequently, the samples were incubated and shaken gently for 30 min at 25 °C in the dark. The absorbance was measured at 515 nm with an infinite M200 microplate reader. The antioxidant activity is expressed as the half-maximal inhibitory concentration (IC50). Each sample was evaluated in triplicate. The percent DPPH radical scavenging activity was calculated according to Eq. (2).

$${\text{DPPH}}\;{\text{scavenging}}\;{\text{activity}}\;\left( \% \right) = \frac{{{\text{A}}_{0} { } - {\text{ A}}_{{\text{s}}} }}{{{\text{A}}_{0} }} \times 100$$
(2)

where A0 and As represent the absorbance of the control solution and samples, respectively.

Antioxidant activity by the ABTS cation radical assay

The ABTS radical scavenging activity of the extracts was investigated according to previously reported procedures for the decolorization of the ABTS cation radicals (ABTS+) 17,22,23 with some modifications. ABTS+ solution was generated by the reaction between 2.45 mM potassium persulfate and 7 mM ABTS in equal volumes and stored in the dark at room temperature for 12–16 h. Post-incubation, the freshly prepared ABTS+ solution was diluted with ethanol to achieve an absorbance of 0.70 ± 0.20 at 734 nm prior to use. The extract (40 µL) at different concentrations (20–200 μg/mL) was mixed with the ABTS+ solution (140 µL), incubated and shaken thoroughly at 25 °C for 10 min. Then, the absorbance was measured at 734 nm with an infinite M200 microplate reader. Trolox solution, a water-soluble analog of vitamin E, at different concentrations (0.5–10 μg/mL) was used as a positive control. The ABTS scavenging activity is expressed as IC50 values. The percent ABTS+ scavenging activity was calculated using Eq. (3).

$${\text{ABTS}}^{ + } \;{\text{scavenging}}\;{\text{activity}}\;\left( \% \right) = \frac{{{\text{A}}_{0} { } - {\text{ A}}_{{\text{s}}} }}{{{\text{A}}_{0} }} \times 100$$
(3)

where \({\text{A}}_{0}\) represents the absorbance of the control solution and \({\text{A}}_{\text{s}}\) represents the absorbance of the extracts or standard Trolox solutions.

In vitro determination of anti-glycation activity in the BSA-glucose model

In this study, the anti-glycation activity of extracts was evaluated using the BSA-glucose model24,25. A reaction mixture (2 mL), which consisted of 800 μg/mL bovine serum albumin (BSA), 200 mM D-glucose in phosphate buffer (0.05 M, pH 7.4) containing 0.2 g/L sodium azide (NaN3) and aminoguanidine (positive control) (0.03–4 mg/mL) or samples (0.1–0.5 mg/mL) at various concentrations, was incubated at 60 °C for 40 h in the dark. A microplate reader determined the presence of AGEs at an excitation wavelength of 370 nm and an emission wavelength of 440 nm. All data are expressed as the IC50 and a percentage of the maximal anti-glycation of the extracts. The percent inhibition was calculated using Eq. (4).

$${\text{Inhibition}}\;\left( \% \right) = \frac{{{\text{A}}_{0} - {\text{A}}_{{\text{s}}} }}{{{\text{A}}_{0} }} \times {1}00$$
(4)

where A0 and As represent the fluorescence intensities of the control solution and sample, respectively.

In vitro anti-collagenase assay

The in vitro anti-collagenase activity was assessed using an EnzChek® gelatinase/collagenase assay kit. The kit comprised collagenase (type IV from Clostridium histolyticum), DQ™ gelatin substrate (fluorescein conjugate derived from pigskin), and 1,10-phenanthroline monohydrate as a positive control. To evaluate the potential of the tested samples to inhibit collagenase activity, samples (1–6 mg/mL) or standard inhibitor solutions (0.1–0.2 μg/mL) with various concentrations (80 µL) were prepared in a 96-fluorescence plate. Then, a 0.4 U/mL of collagenase (100 µL) was added to each well, mixed, and preincubated for 5 min at 37 °C. Afterward, a 50 µg/mL DQ™ gelatin substrate (20 µL) was added to the above reaction mixture and further incubated at 37 °C protected from light for 90 min. After incubation, the fluorescence intensity was measured using a microplate reader with excitation and emission wavelengths of 485 nm and 520 nm, respectively. Finally, all data are expressed as IC50, and the percent collagenase inhibition was calculated according to Eq. (5):

$$\text{Collagenase inhibition }\left(\text{\%}\right)=\left(1-\frac{\text{B}}{\text{A}}\right)\times 100$$
(5)

where A is the fluorescence intensity of the control (absence of sample and collagenase inhibitor) and B is the fluorescence intensity in the presence of sample or collagenase inhibitor.

In vitro skin irritation

The skin irritation test was conducted following OECD 43926. The experiment was performed based on the in vitro test system of the reconstructed human epidermis (RhE) whose biochemical and physiological properties mimic the epidermis of human skin. This study used the EpiSkin™ small model as a skin model. Briefly, after receiving the EpiSkin™ model, they were taken from agar medium used for maintenance of the RhE during transport to a 12-well plate containing maintenance medium and then pre-incubated at 37 °C, 5% CO2, and 95% humidity for 18–24 h to activate cellular metabolism. Then, 10 µL of test samples were applied onto the surface of the RhE tissue. After 15 min of exposure, the samples were removed from the surface of the RhE tissue and rinsed with phosphate buffer saline (PBS). The RhE was further post-incubated for 42 h, and an MTT solution (0.3 mg MTT/mL in assay medium) was added to each well. It was further incubated for 3 h and protected from light. Then, the RhE tissues were transferred and removed from the transwell to a 24-well plate. The formation of purple insoluble formazan crystals by mitochondrial dehydrogenase activity was extracted by adding 500 μL of acidic isopropanol (0.04 N HCl in isopropanol) into the RhE tissues and incubated for 3 h at room temperature with gentle agitation. The purple solution was transferred into a 96-well plate, and the absorbance at 570 nm was measured using a microplate reader.

In vitro skin sensitization using in chemico method

To assess the skin sensitization potential of C. morifolium flower extracts, in chemico skin sensitization based on Amino acid Derivative Reactivity Assay (ADRA) as described in OECD 422c was performed27. This method categorizes test chemicals based on their minimal, low, medium, or high reactivity. Skin sensitizers can specifically form covalent bonds with skin proteins, leading to the depletion of these peptides. This can be evaluated using HPLC or UPLC connected with a UV detector. In this experiment, the flower extracts of C. morifolium contained multiple constituent substances of known and unknown compositions. Therefore, the test concentration of the extracts was prepared at 0.5 mg/mL. Briefly, NAC and NAL at a concentration of 6.667 µM in 100 mM of pH 8.0 phosphate buffer and 2 mM in 100 mM of pH 10.2 phosphate buffer, respectively, were mixed with the extracts or the positive control (phenylacetaldehyde) to achieve a final concentration of NAC and NAL at 5 µM. The mixture was then incubated for 24 h at 25 °C. Following incubation, NAC and NAL depletion were monitored using UPLC coupled with a UV detector and an Acquity UPLC BEH C18 column (2.1 × 100 mm, particle size 1.7 µm, Waters) as the stationary phase. The mobile phase consisted of 0.1% (v/v) trifluoroacetic acid in water (pump A) and 0.1% (v/v) trifluoroacetic acid in acetonitrile (pump B), using a gradient mode at a flow rate of 0.2 mL/min. The gradient program was as follows: from 0 to 4.56 min, 70% to 45% pump A; from 4.65 to 4.9 min, 45% to 0% pump A; from 4.9 to 6.4 min, 100% pump A; from 6.4 to 6.6 min, 70% pump A; and from 6.6 to 10.0 min, 70% pump A. The UPLC column was maintained at 40 °C. The standard curve for NAC and NAL was established in the concentration range of 0.156 µM to 5.0 µM.

Statistical analysis

All data are expressed as the mean and standard deviation. Significant differences among obtained values were analyzed using SPSS program version 18 using one-way ANOVA with Tukey’s multiple comparison tests. The graph was plotted using GraphPad Prism version 10.3.1 for Windows (GraphPad Software, La Jolla, CA). A p-value less than 0.05 was considered significant.

Results and discussion

Characterization of the C. morifolium flower extracts

Plant extracts and % NVM

The % NVM and densities of the C. morifolium flower extracts are shown in Table 1. The extracts were light yellowish orange to light yellow, with densities ranging from 1.03–1.05. The % NVM of the purple C. morifolium flower extract (2.48 ± 0.05% w/w) was higher (p < 0.05) than those of the yellow (1.07 ± 0.03% w/w) and white (1.64 ± 0.06% w/w) flower extracts.

Table 1 % NVM, density, TPC, TFC, and percentages of chlorogenic, caffeic acid and lutein in hydroglycolic extracts of C. morifolium flower extracts.
Full size table

TPC and TFC

Phenolic compounds and flavonoids are secondary plant metabolites found naturally in all plants, including C. morifolium flowers6,28. Therefore, the TPC and TFC in the purple, yellow, and white C. morifolium flower extracts were assessed and compared. The TPC and TFC were in the range of 4.0–6.5 mg GAE/g dry extract and 0.5–1.4 mg CE/g dry extract, respectively. Purple C. morifolium flower extracts showed the highest TPC (p < 0.05), followed by the yellow and white extracts, as represented in Table 1 and Fig. 1. The TFC values of the purple and yellow C. morifolium flower extracts were not statistically different (p > 0.05). The TFC of the white flower extract was the lowest (p < 0.05). Based on the above results, our research reveals differences in TPC and TFC across various flower color variants of C. morifolium. Notable differences in TPC and TFC have been noted among different cultivars of C. morifolium 29, indicating that specific cultivars accumulate these secondary metabolites in distinct ways. Comparable patterns of variation have been identified in other members of the Asteraceae family, such as Tagetes species, where TPC ranged from 7.35 to 12.02 g GAE/100g dry petals and TFC fluctuated between 3.00 and 5.20 g CE/100g dry petals among various flower color variants30.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Comparison of bioactive properties among purple, yellow, and white C. morifolium flower extracts. The analyses include TPC, TFC, DPPH radical scavenging activity (IC50, µg/mL), ABTS radical scavenging activity (IC50, µg/mL), anti-glycation activity (IC50, mg/mL), and anti-collagenase activity (IC50, mg/mL). Data are presented as mean ± SD (n = 3). Different letters above bars indicate significant differences between extracts (p < 0.05).

Qualitative and quantitative analysis of the C. morifolium extracts

Qualitative analysis of the hydroglycolic extracts of C. morifolium flowers was investigated by HPLC using a UV detector. Standard chlorogenic acid, caffeic acid, and luteolin solutions were prepared at 0.05–12.00 µg/mL. Based on the validation results, the correlation coefficients (r) for standard curves of chlorogenic acid, caffeic acid, and luteolin were 0.9996, 0.9995, and 0.9995, respectively. The repeatability showed %RSD values lower than 2 for chlorogenic acid, caffeic acid, and luteolin. The percent recovery was in the range of 90–110% for all standard markers. The LOQ and LOD of chlorogenic acid, caffeic acid, and luteolin were 0.0712 µg/mL and 0.0235 µg/mL, 0.1361 µg/mL and 0.0449 µg/mL, and 0.0480 µg/mL and 0.0158 µg/mL, respectively. HPLC chromatograms of the C. morifolium flower extracts and a mixture of standard chlorogenic acid, caffeic acid, and luteolin are shown in Fig. 2. The retention times of chlorogenic acid, caffeic acid, and luteolin were 3.9, 5.0, and 21.0 min, respectively. Figures 2A–C show the HPLC chromatograms of purple, yellow, and white C. morifolium flower extracts, respectively.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

HPLC chromatograms of C. morifolium flower extract from (A) purple flower, (B) yellow flower, (C) white flower, (D) mixed standards of (1) chlorogenic acid, (2) caffeic acid, and (3) luteolin.

As shown in Table 1, the purple C. morifolium flower extract showed the highest percentage of luteolin (0.2403 ± 0.0002% w/w) (p < 0.05), while chlorogenic acid (0.4320 ± 0.0022% w/w) and caffeic acid (0.0289 ± 0.0002% w/w) were more pronounced in the yellow C. morifolium flower extract (p < 0.05).

According to the literature, several compounds, including phenolic compounds and flavonoids, have been identified in C. morifolium. Wang et al. determined the presence of four flavonoids, including quercetin, luteolin, apigenin, and acacetin, in C. morifolium at different flowering stages and from various cultivation areas7. The flavonoid content varied depending on the cultivation area and changed throughout the flowering period. Specifically, luteolin content was found to range from 5.4 mg/g to 26.5 mg/g. Recently, Gong et al. investigated the phenolic compounds in a 70% methanolic extract of C. morifolium18. Their study showed that different harvest times affected the phenolic compositions, resulting in variations in antioxidant properties. Quantitative analysis revealed that the concentrations of chlorogenic acid, caffeic acid, and luteolin in C. morifolium flowers varied considerably, with ranges of 1,370–3,193 µg/g, 3.92–23.02 µg/g, and 24.08–247.09 µg/g dry weight, respectively. Our findings further indicate that the levels of these phenolic compounds differ significantly among C. morifolium flower color variants.

Antioxidant activity

All C. morifolium flower extracts were tested for antioxidant activity using DPPH and ABTS assays. The results of the antioxidative activities were expressed as IC50. Ascorbic acid and Trolox were positive controls for the DPPH and ABTS assays. The IC50 values of the test samples were 70–165 µg/mL, as shown in Table 2 and Fig. 1. The IC50 of ascorbic acid was 0.003 ± 0.000 µg/mL, significantly lower than all C. morifolium flower extracts (p < 0.05). The IC50 values for DPPH radical scavenging activity of the purple (71.112 ± 0.298 μg/mL) and yellow (71.519 ± 1.197 μg/mL) C. morifolium flower extracts were statistically comparable (p > 0.05) but significantly lower than that of the white (163.692 ± 1.420 μg/mL) C. morifolium flower extract (p < 0.05). These results suggest that the purple and yellow extracts exhibited higher antioxidant capacity in the DPPH assay than the white extract. In the ABTS radical scavenging assay, Trolox demonstrated significantly higher antioxidant capacity (IC50 = 5.924 ± 0.079 μg/mL) compared to all C. morifolium flower extracts (p < 0.05), as shown in Table 2 and Fig. 1. Among the flower extracts, the purple C. morifolium flower extract exhibited the lowest IC50 (77.502 ± 1.501 μg/mL), followed by the yellow (97.143 ± 0.439 μg/mL) and white (150.925 ± 1.900 μg/mL) extracts with statistically significantly differences between all groups (p < 0.05). Previous studies have established a correlation between antioxidant capacity and the presence of phenolic compounds and flavonoids in plant extracts24,31. Consistent with these findings, our results showed higher TPC and TFC in the purple and yellow C. morifolium flower extracts compared to the white extract. This suggests that the high antioxidant activities of the purple and yellow C. morifolium flower extracts compared to the white extract may be attributed to their higher TPC and TFC. This observation aligns with a previous report demonstrating a strong correlation between antioxidant activity and TPC in aqueous extracts of C. morifolium flowers32. Recently, Gong et al. studied the antioxidant activities of a 70% methanolic extract of C. morifolium flowers by DPPH and ABTS assays. Their analysis revealed a high correlation index (R2) between the phenolic contents and antioxidant activities, further supporting the relationship between these parameters. According to the literature, a compound with an IC50 lower than 50 ppm is classified as a very strong antioxidant, 50–100 ppm as a strong antioxidant, 101–150 ppm as a moderate antioxidant, and higher than 150 ppm as a weak antioxidant. Based on these criteria, the purple and yellow C. morifolium flower extracts were classified as strong antioxidants, while the white C. morifolium flower extract was classified as a weak antioxidant33,34.

Table 2 The IC50 of DPPH radical scavenging activity, ABTS radical scavenging activity, anti-glycation, and anti-collagenase of C. morifolium flower extracts.
Full size table

Anti-glycation activity in the BSA-glucose model

Glycation is a nonenzymatic reaction resulting from the covalent bonding between reducing sugars, e.g., glucose or fructose, and protein or lipid molecules24. Such a reaction causes the formation of advanced glycation end products (AGEs), which are the beginning step of glycation reactions in the human body. The formation of AGEs may lead to the cross-linking or breakdown of collagen or elastin in the skin, resulting in the loss of skin elasticity and skin aging35. In this study, the anti-glycation activity of C. morifolium flower extracts was expressed as IC50 values, as shown in Table 2. The IC50 of the positive control (aminoguanidine) was 0.422 mg/mL. From Table 2 and Fig. 1, it can be observed that the purple C. morifolium flower extract showed the lowest IC50 value (0.279 mg/mL) (p < 0.05), followed by the yellow flower extract (0.321 mg/mL). Conversely, the IC50 of the white C. morifolium flower extract could not be detected owing to the low activity in inhibiting the formation of AGEs. This indicated that the purple C. morifolium flower extract had the highest anti-glycation activity compared to other extracts.

Based on our findings, the superior anti-glycation activity observed in purple C. morifolium flower extract appears to be associated with its higher TPC and antioxidant activity as measured by the ABTS assay. A previous study evaluated and compared the inhibitory effects of the aqueous extracts of Chrysanthemum species, especially C. indium and C. morifolium, on the formation of AGEs36. It was found that both Chrysanthemum species could strongly inhibit the formation of non-fluorescent AGEs in BSA/glycose and BSA/fructose systems. However, the C. indium extract showed a lower ability to inhibit the formation of fluorescent AGEs, highlighting the broad-spectrum activity of C. morifolium as an AGE inhibitor. The difference in anti-glycation activities between both Chrysanthemum species correlated with the polyphenol and flavonoid contents. High correlations between the anti-glycation and phenolic and flavonoid contents have also been reported in other edible and medicinal plants, such as Punica granatum, Dimocarpus longan, and Mangifera indica24. The higher the phenolic and flavonoid contents, the higher the anti-glycation activity. Additionally, it has been reported that plant extracts with antioxidant activity can suppress the formation of AGEs during the glycation process37.

In vitro anti-collagenase assay

The main structural fibers in the dermis are elastin and collagen. Collagen is the most abundant protein in the extracellular matrix (ECM) and is responsible for the tensile strength of skin by adhering to connective tissues38. Consequently, the breakdown or depletion of collagen results in a lack of skin integrity and elasticity, leading to wrinkles and signs of aging39.

Table 2 and Fig. 1 display the anti-collagenase activity of the purple, yellow, and white C. morifolium flower extracts. The anti-collagenase activity of the yellow C. morifolium (IC50 = 2.263 ± 0.336 mg/mL) and purple (IC50 = 2.999 ± 1.429 mg/mL) flower extracts were significantly higher than that of the white extract (IC50 = 5.223 ± 0.402 mg/mL) as indicated by the lower IC50 value. All C. morifolium flower extracts showed anti-collagenase activity according to the above results. The IC50 of 1,10-phenanthroline monohydrate (0.124 ± 0.003 µg/mL) was significantly lower than that of all C. morifolium flower extracts (p < 0.05). From the literature, several plant-rich polyphenols have been reported to inhibit collagenase activity39,40,41. The inhibitory effects of chlorogenic acid, caffeic acid, and luteolin have also been documented42,43,44. The effects are caused by the interaction between the hydroxyl groups of polyphenol compounds and the backbone or side chain functional groups of the collagenase enzyme. Additionally, hydrophobic interactions occur between the benzene ring of the polyphenol and collagenase. These interactions lead to conformational changes, ultimately rendering the collagenase enzyme nonfunctional45,46.

In vitro skin irritation on the reconstructed human epidermis (RhE)

The potential for skin irritation from various colors of C. morifolium flower extract was studied using RhE. Figure 3 illustrates the percentage of tissue viability for 5% sodium dodecyl sulfate (SDS) (positive control), water (negative control), and C. morifolium flower extracts. The optical density (OD) of the negative control (water) was 1.31, within the range of 0.6 to 1.5. The cell viability percentage for 5% SDS was found to be 36.7%, which fell within the acceptance criteria (< 40%). For 5% SDS, the cell viability was less than 50%, confirming its status as a skin irritant. In contrast, all C. morifolium flower extracts demonstrated cell viability percentages ranging from 80 to 97%, significantly higher than 5% SDS (p < 0.05) With tissue viability exceeding 50%, all C. morifolium flower extracts were classified as non-irritants to the skin. While the RhE model offers valuable preliminary safety data, it is essential to recognize that this in vitro system has limitations in accurately mimicking the complex biological responses of human skin. Consequently, these findings should be viewed as initial safety indicators.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Percent cell viability of water, 5% SDS, and different colors of C. morifolium flower extracts. Each value represents a mean ± SD (n = 3). *Statistically significant difference compared to 5% SDS.

In vitro skin sensitization using in chemico method

In this experiment, the potential for skin sensitization from various colors of C. morifolium flower extract was assessed by measuring the depletion of peptides, i.e., NAC and NAL after exposure to the various colors of C. morifolium flower extract. The obtained results are shown in Table 3. The positive control, phenylacetaldehyde, exhibited mean depletion values of 58.19 ± 0.71% for NAC and 87.72 ± 0.01% for NAL. These results meet the established acceptance criteria for phenylacetaldehyde as a sensitizer, which specify that the percent depletion should fall within the ranges of 30–80% for NAC and 70–100% for NAL. The measured values are within these ranges, confirming the validity of the test results and their alignment with the anticipated outcomes for phenylacetaldehyde as a positive control.

Table 3 % NAC and NAL depletion of phenylacetaldehyde and different colors of C. morifolium flower extracts. Data are the mean ± standard deviation (n = 3).
Full size table

For the test samples, the percentage depletion of NAC and NAL ranged from 0.92% to 2.07%. Since the depletion of both NAC and NAL was less than 4.9%, the test samples were classified as negative for skin sensitization. However, it is important to recognize that the in chemico results represent only the initial protein binding events in the skin sensitization process. Although these findings indicate a low potential for sensitization, ADRA has certain inherent limitations. For instance, it cannot identify pro-haptens that require metabolic activation, and there may be issues with the solubility of certain compounds. Additionally, this method does not fully replicate the complex immunological cascade and biological transformations that occur in human skin27,47. Nevertheless, this study serves as a valuable tool for initial safety screening, providing a rapid and time-efficient evaluation of the extracts. Consequently, it is essential to interpret these results in conjunction with complementary testing methods to establish a thorough safety profile.

Conclusion

The preliminary results of this study emphasize the potential of hydroglycolic extracts from C. morifolium flowers, particularly those with yellow and purple colors, as plant-based active ingredients with antioxidant, anti-glycation, and anti-collagenase properties in vitro. These characteristics indicate promising applications for antioxidant and anti-aging effects. However, it is important to note that these findings are derived from in vitro models, which may not completely reflect the complexity of biological responses in living organisms. An initial safety assessment for skin application, concerning irritation and sensitization, concerning irritation and sensitization has also been confirmed through in vitro testing. Nonetheless, given the limitations of in vitro models in predicting in vivo outcomes, comprehensive future research, including human clinical trials, is essential to thoroughly support both the efficacy and safety of C. morifolium flower extracts for human use.