Protective Effects of Duchesnea chrysantha Ethanol Extract Against UVB‐Induced Skin Damage via Modulation of TAK1‐Associated Signaling Pathways

ABSTRACT

Duchesnea chrysantha (Zoll. & Moritzi) Miq. (Rosaceae) has traditionally been used in East Asian medicine for inflammatory and skin‐related disorders. This study investigated the phytochemical composition and protective effects of D. chrysantha ethanol extract (Dc‐EE) against ultraviolet B (UVB)‐induced skin damage. Untargeted LC–MS/MS analysis tentatively annotated various phytochemical constituents, including flavonoids and phenolic acids. Dc‐EE reduced UVB‐induced oxidative stress, reactive oxygen species production, and apoptosis in human keratinocytes (HaCaT) and dermal fibroblasts (HDF). It also suppressed cyclooxygenase‐2 and matrix metalloproteinases (MMP‐1, MMP‐2, and MMP‐9) in HaCaT cells while restoring collagen type I alpha 1 expression in HDF cells. In addition, Dc‐EE enhanced the expression of skin barrier‐ and hydration‐related factors, including filaggrin, transglutaminase‐1, and hyaluronan synthase‐2. Dc‐EE attenuated UVB‐induced MAPK signaling by reducing JNK, ERK, and p38 phosphorylation, together with AP‐1 and NF‐κB transcriptional activities. Reduced activation of transforming growth factor‐β‐activated kinase 1, an upstream stress‐responsive regulator, was also observed. These findings indicate that Dc‐EE attenuates UVB‐induced cellular damage through antioxidant, anti‐inflammatory, and barrier‐related effects, supporting its potential for skin protection.

An ethanol extract of Duchesnea chrysantha attenuates UVB‐induced oxidative stress, cell death, and inflammatory responses in keratinocytes while modulating extracellular‐matrix‐related markers in keratinocytes and dermal fibroblasts. These protective effects are associated with reduced TAK1 activation and suppression of NF‐κB, AP‐1, and MAPK signaling, highlighting the extract's potential against UVB‐induced skin damage.

1. Introduction

The skin is the largest organ of the human body and serves as a primary protective barrier against environmental stressors, including ultraviolet (UV) radiation, chemical agents, and pathogenic microorganisms [1, 2]. By maintaining epidermal integrity and regulating inflammatory and oxidative responses, the skin plays a crucial role in preserving tissue homeostasis. Structurally, the skin is composed of the epidermis, dermis, and subcutaneous tissue, among which the epidermis—dominated by keratinocytes—is the first line of defense against external insults [3].

Solar UV radiation, particularly ultraviolet B (UVB; 280–320 nm), is a major environmental factor responsible for skin damage and photoaging [4]. UVB predominantly affects epidermal keratinocytes and induces excessive production of reactive oxygen species (ROS), which leads to oxidative stress, apoptosis, and inflammatory responses. Elevated intracellular ROS levels activate multiple signaling cascades, including mitogen‐activated protein kinases (MAPKs) and nuclear factor‐κB (NF‐κB), resulting in increased expression of pro‐inflammatory mediators such as cyclooxygenase‐2 (COX‐2) and matrix metalloproteinases (MMPs) [5]. MMPs, particularly MMP‐1, MMP‐2, and MMP‐9, play critical roles in collagen degradation and extracellular matrix (ECM) remodeling, thereby contributing to wrinkle formation and loss of skin elasticity during photoaging [6].

In addition to ECM degradation, UVB‐induced oxidative stress disrupts epidermal barrier function and skin hydration. Key molecules involved in maintaining skin barrier integrity and moisture include filaggrin (FLG), transglutaminase‐1 (TGM‐1), and hyaluronan synthase‐2 (HAS‐2) [7, 8]. Downregulation of these factors compromises epidermal differentiation and hyaluronic acid synthesis, leading to impaired barrier function and dehydration. Therefore, modulation of ROS‐dependent signaling pathways and restoration of collagen synthesis and barrier‐related factors represent important strategies for protecting skin from UVB‐induced damage.

Current therapeutic approaches for photoaging, such as topical retinoids, exert their effects primarily by suppressing AP‐1–mediated MMP expression and regulating melanogenesis. However, their clinical use is frequently limited by adverse effects, including erythema, dryness, peeling, and irritation. In addition, many synthetic agents target a limited number of molecular pathways and may raise safety concerns during long‐term use. Consequently, there is growing interest in naturally derived agents that can modulate oxidative stress and inflammatory signaling with improved tolerability [9, 10].

Plants are continuously exposed to UV radiation and have evolved diverse secondary metabolites to protect against oxidative damage. Numerous plant‐derived polyphenols and flavonoids have been reported to exhibit antioxidant and anti‐inflammatory activities by scavenging ROS and regulating redox‐sensitive signaling pathways [11, 12]. These observations provide a pharmacological basis for investigating traditional medicinal plants used for inflammatory and skin‐related disorders as potential sources of photoprotective agents.

Duchesnea chrysantha, commonly known as San‐baem‐ttalgi in Korea, is a perennial herb traditionally used in East Asian countries, including Korea, China, and Japan [13]. In traditional medicine, D. chrysantha has been prescribed for inflammatory disorders and has also been applied topically for skin conditions such as eczema, acne, and abscesses [14, 15], as described in classical East Asian medical texts such as Donguibogam [16]. Previous pharmacological studies have reported that extracts of D. chrysantha possess antioxidant, anti‐inflammatory, and immunomodulatory activities. Phytochemical analyses have further suggested that the extract contains bioactive compounds with strong redox‐modulating potential. Nevertheless, the effects of D. chrysantha on UVB‐induced skin damage and the associated molecular mechanisms remain insufficiently characterized [15, 17].

Transforming growth factor‐β–activated kinase 1 (TAK1) is a central upstream regulator of MAPK and NF‐κB signaling pathways and plays a pivotal role in mediating cellular responses to oxidative stress and inflammatory stimuli [9, 18, 19, 20]. Activation of TAK1 leads to phosphorylation of downstream kinases, including ERK, JNK, and p38, as well as transcription factors such as AP‐1 and NF‐κB, which collectively regulate the expression of inflammatory mediators, MMPs, and apoptosis‐related proteins [21, 22]. Dysregulation of TAK1‐dependent signaling has been implicated in UVB‐induced skin inflammation and matrix degradation [23, 24]. However, whether D. chrysantha directly modulates TAK1‐associated signaling in the context of UVB‐induced skin damage has not yet been elucidated.

Therefore, this study aimed to investigate whether Dc‐EE can mitigate UVB‐induced cellular damage and to examine changes in TAK1‐associated stress signaling in relation to its protective effects. Furthermore, we investigated the MAPK/AP‐1 and NF‐κB signaling pathways associated with these responses.

2. Results and Discussion

2.1. Phytochemical Characterization, Antioxidant Activity, and Target Prediction of Dc‐EE

Untargeted LC–MS/MS analysis tentatively annotated a chemically diverse range of phytochemical features in Dc‐EE, including flavonoid glycosides, phenolic acids, and related phytochemical candidates (Figure 1A and Table 1). Several tentatively annotated features corresponded to quercetin derivatives, luteolin glycosides, chlorogenic acid, and ellagic acid, compounds previously associated with antioxidant and anti‐inflammatory activities. Although the individual constituents were not confirmed or quantified, this phytochemical profile provides a plausible chemical context for the observed biological activities of Dc‐EE.

FIGURE 1.

Chemical profiling and skin‐associated biological activities of Dc‐EE. (A) Representative LC–MS/MS chromatograms of Duchesnea chrysantha ethanol extract (Dc‐EE) acquired in positive ion mode (ES⁺), negative ion mode (ES⁻), and diode array detection (DAD). Major phytochemical features were putatively or tentatively annotated based on accurate precursor mass, retention behavior, and product‐ion/library matching, as summarized in Table 1. (B–E) Antioxidant activities of Dc‐EE assessed using complementary in vitro assays, including DPPH radical scavenging (B), ABTS radical scavenging (C), cupric ion reducing antioxidant capacity (CUPRAC) (D), and ferric reducing antioxidant power (FRAP) (E). Ascorbic acid or Trolox was used as a reference antioxidant. (F,G) Total flavonoid content (TFC) (F) and total phenolic content (TPC) (G) of Dc‐EE determined at increasing concentrations. (H,I) Effects of Dc‐EE on epidermal hydration–related parameters, evaluated by measuring hyaluronic acid secretion (H) and intracellular hyaluronic acid levels (I) in HaCaT keratinocytes after 24 h of treatment. (J–M) Cytotoxicity evaluation of Dc‐EE in HaCaT keratinocytes (J), B16F10 melanoma cells (K), HEK293T cells (L), and human dermal fibroblasts (HDFs) (M) using cell viability assays. Data are presented as the mean ± SD of at least three independent replicates (n ≥ 3). Statistical significance was assessed using Student's t‐test. *p < 0.05 and **p < 0.01 versus the control group; #p < 0.05 and ##p < 0.01 versus the normal group.

TABLE 1.

Phytochemical constituents tentatively annotated in Duchesnea chrysantha ethanol extract by untargeted LC–MS/MS analysis.

Detection mode	RT (min)	Tentative annotation	Molecular formula	Calculated precursor m/z	Observed precursor m/z	Ion form	Mass error (ppm)	Matched product ions (n)	Annotation confidence
ES+	3.79	Luteolin‐7‐O‐glucuronide	C21H18O12	485.0690	485.0673	[M+Na]+	−3.6	15	Level 2 (putative annotation)
ES+	8.84	δ‐Tocotrienol	C27H40O2	397.3101	397.3106	[M+H]+	1.2	9	Level 2 (putative annotation)
ES+	11.32	Trigoneoside IIIa	C45H76O18	905.5104	905.5086	[M+H]+	−2.0	26	Level 2 (putative annotation)
ES+	11.55	Sophoranodichromane B	C25H28O5	431.1829	431.1833	[M+Na]+	0.9	2	Level 2 (putative annotation)
ES+	11.84	Mulberrofuran A	C25H28O4	415.1880	415.1872	[M+Na]+	−1.9	0	Level 3 (tentative candidate)
ES+	11.97	Leonuridine	C15H24O9	371.1313	371.1309	[M+Na]+	−1.0	5	Level 2 (putative annotation)
ES+	12.69	Ginsenoside Rh4	C36H60O8	621.4361	621.4378	[M+H]+	2.8	1	Level 2 (putative annotation)
ES+	12.84	Ginsenoside Rh4	C36H60O8	621.4361	621.4383	[M+H]+	3.5	2	Level 2 (putative annotation)
ES+	13.15	20(S)‐Ginsenoside Rh1	C36H62O9	639.4467	639.4497	[M+H]+	4.7	16	Level 2 (putative annotation)
ES+	13.97	Astragaloside VIII	C47H76O17	913.5155	913.5188	[M+H]+	3.6	49	Level 2 (putative annotation)
ES−	1.81	Casuariin	C34H24O22	783.0686	783.0688	[M−H]−	0.2	20	Level 2 (putative annotation)
ES−	3.43	Ellagic acid	C14H6O8	300.9990	300.9994	[M−H]−	1.4	4	Level 2 (putative annotation)
ES−	5.55	Arjunetin	C36H58O10	695.4012	695.4007	[M+HCOO]−	−0.8	24	Level 2 (putative annotation)
ES−	8.29	Stearidonic acid	C18H28O2	275.2017	275.2013	[M−H]−	−1.3	12	Level 2 (putative annotation)
ES−	8.87	Raddeanoside R0	C35H56O7	587.3953	587.3969	[M−H]−	2.6	18	Level 2 (putative annotation)
ES−	9.17	Ginsenoside Rh4	C36H60O8	619.4215	619.4208	[M−H]−	−1.3	17	Level 2 (putative annotation)
ES−	12.89	γ‐Tocotrienol	C28H42O2	409.3112	409.3128	[M−H]−	4.0	0	Level 3 (tentative candidate)
ES−	13.95	Hydroxyvalerenic acid	C16H24O3	309.1707	309.1713	[M+HCOO]−	1.8	0	Level 3 (tentative candidate)
DAD‐associated peak	2.14	Geraniin	C41H28O27	951.0745	951.0756	[M−H]−	1.2	5	Level 2 (putative annotation)
DAD‐associated peak	2.51	Chlorogenic acid	C16H18O9	353.0878	353.0873	[M−H]−	−1.5	18	Level 2 (putative annotation)
DAD‐associated peak	3.14	Quercetin‐3‐O‐β‐D‐glucuronide	C21H18O13	479.0820	479.0825	[M+H]+	0.9	17	Level 2 (putative annotation)
DAD‐associated peak	3.38	Patuletin‐7‐O‐glucoside	C22H22O13	493.0988	493.0990	[M−H]−	0.5	22	Level 2 (putative annotation)

Calculated precursor m/z values were derived from the corresponding molecular formulas and indicated ion forms using monoisotopic masses. Mass error was calculated as [(observed m/z − calculated m/z)/calculated m/z] × 10⁶. The matched product‐ion values represent the numbers of product ions matched during spectral‐library analysis rather than individual diagnostic fragment m/z values. Level 2 indicates a putative annotation supported by accurate precursor mass and product‐ion/library matching, whereas Level 3 indicates a tentative candidate based primarily on accurate precursor mass and retention evidence without matched product‐ion support. Because authentic standards were not analyzed, none of the assignments represents a confirmed Level 1 identification. Mulberrofuran A was retained as a Level 3 tentative candidate; accordingly, its docking result is interpreted solely as an exploratory computational hypothesis and does not confirm either its presence in Dc‐EE or its direct interaction with TAK1.

To evaluate the antioxidant capacity of Dc‐EE, multiple complementary assays were performed. Dc‐EE exhibited concentration‐dependent radical scavenging activity in both DPPH and ABTS assays, with strong activity observed at higher concentrations (Figure 1B,C). Consistently, Dc‐EE also demonstrated pronounced reducing capacity in CUPRAC and FRAP assays, indicating effective electron‐donating ability and redox modulation (Figure 1D,E). In parallel, the total flavonoid and phenolic contents increased with concentration, supporting the contribution of phenolic constituents to the antioxidant activity (Figure 1F,G).

Given the importance of epidermal hydration in skin homeostasis, the effect of Dc‐EE on hyaluronan (HA) production was further assessed. Dc‐EE significantly increased both extracellular HA secretion and intracellular HA levels in HaCaT keratinocytes in a dose‐dependent manner (Figure 1H,I). In addition, no significant cytotoxic effects were observed in HaCaT, HEK293T, HDF, and B16F10 cells at the tested concentrations, indicating good cellular tolerance (Figure 1J–M).

Collectively, these findings indicate that Dc‐EE possesses strong antioxidant activity, enhances skin‐related functional markers, and exhibits low cytotoxicity, supporting its potential role in protecting skin against oxidative stress–induced damage.

To gain insight into the molecular basis underlying the antioxidant and anti‐inflammatory activities of Dc‐EE, putative targets of Dc‐EE–derived phytochemicals were first collected and intersected with genes associated with antioxidative and anti‐inflammatory functions. As illustrated in the Venn diagram (Figure 2B), 307 common targets were identified, indicating a substantial overlap between Dc‐EE–associated targets and genes involved in oxidative stress and inflammatory regulation. To further characterize the biological relevance of these overlapping targets, Gene Ontology (GO) enrichment analysis was performed. As shown in Figure 2A, enriched Biological Process (BP) terms were mainly related to responses to oxidative stress, regulation of inflammatory responses, apoptotic signaling, and cellular stress responses. In the Cellular Component (CC) category, the targets were predominantly associated with cytosolic, membrane‐associated, and protein complex–related components. Molecular Function (MF) analysis further revealed enrichment in functions related to protein binding, kinase activity, and enzyme regulation, suggesting extensive involvement of signaling and regulatory proteins. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis provided additional insight into the signaling networks potentially modulated by Dc‐EE. As presented in Figure 2C, the overlapping targets were significantly enriched in several pathways closely associated with oxidative stress and inflammation, including the MAPK signaling pathway, apoptosis, HIF‐1 signaling pathway, AGE–RAGE signaling pathway, and ROS–related pathways. Among these, the MAPK signaling pathway emerged as a prominent node, highlighting its potential involvement in mediating the antioxidant and anti‐inflammatory effects of Dc‐EE.

FIGURE 2.

Identification of Dc‐EE–associated targets and functional enrichment analysis. (A) GO enrichment analysis of the overlapping targets between Dc‐EE–associated genes and antioxidative/anti‐inflammatory–related genes. Enriched terms are classified into BP, CC, and MF categories. (B) Venn diagram illustrating the intersection between Dc‐EE–associated targets and genes related to antioxidative and anti‐inflammatory functions. A total of 307 common targets were identified. (C) KEGG pathway enrichment analysis of the overlapping targets. Significantly enriched pathways are presented based on −log10(Q value), highlighting pathways associated with oxidative stress, inflammation, and cell survival signaling.

Taken together, these bioinformatics analyses suggest that Dc‐EE may exert its protective effects through coordinated regulation of oxidative stress– and inflammation‐related signaling pathways. These findings provided a rationale for subsequent experimental validation focusing on MAPK‐associated inflammatory signaling and UVB‐induced skin damage.

The phytochemical complexity of Dc‐EE may represent a major advantage in the context of photoprotection. UVB‐induced skin damage is intrinsically multifactorial, involving oxidative stress, inflammation, apoptosis, ECM degradation, and barrier dysfunction. Therefore, a multi‐component extract containing phenolic acids, flavonoids, and related phytochemicals may provide broader protection than single‐target approaches through simultaneous modulation of multiple stress‐responsive pathways [25, 26, 27]. The enrichment of MAPK‐related pathways identified by network pharmacology analysis further supports the possibility that these phytochemicals collectively regulate oxidative stress– and inflammation‐associated signaling networks.

2.2. Protective Effects of Dc‐EE Against UVB‐induced Oxidative Stress and Photoaging

To assess whether Dc‐EE confers cytoprotection under oxidative stress conditions, HaCaT keratinocytes were challenged with sodium nitroprusside (SNP, 1.5 mM), a nitric oxide (NO) donor. SNP exposure markedly reduced cell viability to approximately 30%–40% of control levels, whereas pretreatment with Dc‐EE significantly restored cell survival in a concentration‐dependent manner. In particular, Dc‐EE at 25, 50, and 100 µg/mL progressively increased cell viability, reaching nearly 70% of control values at 100 µg/mL (Figure 3A). In parallel, SNP‐induced NO production was robustly elevated in HaCaT cells, while Dc‐EE pretreatment significantly suppressed NO levels at concentrations of 25–100 µg/mL (Figure 3B). The inhibitory effect on NO production was most pronounced at 100 µg/mL, indicating that Dc‐EE effectively attenuates nitrosative stress under oxidative challenge. The protective activity of Dc‐EE was further examined using a UVB‐induced keratinocyte damage model. Exposure to UVB irradiation (30 mJ/cm²) reduced HaCaT cell viability to approximately 60% of control levels. Pretreatment with Dc‐EE significantly and dose‐dependently restored cell viability, with concentrations of 25–150 µg/mL progressively improving survival and nearly complete recovery observed at 150 µg/mL (Figure 3C). Consistent with the quantitative viability data, live/dead fluorescence staining revealed that UVB irradiation caused extensive cell death accompanied by cell rounding, loss of adhesion, and disrupted monolayer structure. In contrast, Dc‐EE pretreatment at 25–100 µg/mL markedly reduced the proportion of dead cells and preserved normal keratinocyte morphology, including intact cell–cell contacts and spread cellular architecture, compared with the UVB‐only group (Figure 3D). Given the central role of oxidative stress in UVB‐mediated skin damage, intracellular ROS levels were next assessed by flow cytometry. UVB irradiation induced a pronounced increase in intracellular ROS, with ROS‐positive cell populations rising to approximately 40%. Dc‐EE pretreatment significantly reduced ROS accumulation in a concentration‐dependent manner, decreasing ROS‐positive cells to ∼10%, ∼6%, and <1% at 25, 100, and 200 µg/mL, respectively (Figure 3E), demonstrating a strong antioxidant effect at higher concentrations. To further determine whether Dc‐EE influences apoptotic signaling pathways, the expression of apoptosis‐related proteins was analyzed. UVB exposure markedly increased the levels of cleaved caspase‐3 and the pro‐apoptotic protein Bax. Dc‐EE treatment significantly suppressed UVB‐induced caspase‐3 activation at 25–100 µg/mL, with near‐complete inhibition observed at 100 µg/mL. Similarly, Bax expression was dose‐dependently reduced by Dc‐EE pretreatment, indicating effective suppression of UVB‐triggered apoptotic signaling (Figure 3F, left panel). Band intensities were quantified, with cleaved caspase‐3 normalized to total caspase‐3 and Bax normalized to β‐actin (Figure 3F, middle and right panels).

FIGURE 3.

Dc‐EE protects HaCaT keratinocytes from oxidative stress– and UVB‐induced cellular injury. (A) Cell viability of HaCaT keratinocytes following exposure to sodium nitroprusside (SNP, 1.5 mM) in the presence or absence of Dc‐EE. (B) Nitric oxide (NO) levels measured in SNP‐treated HaCaT cells with Dc‐EE pretreatment. (C) Viability of HaCaT keratinocytes after UVB irradiation (30 mJ/cm²) with or without Dc‐EE pretreatment. (D) Representative live/dead fluorescence images illustrating the protective effects of Dc‐EE against UVB‐induced cell damage, including improved cell viability and preservation of normal cell morphology. (E) Flow cytometric profiles showing intracellular ROS levels in UVB‐irradiated HaCaT cells treated with Dc‐EE. (F) Immunoblot analysis of apoptosis‐related proteins, cleaved caspase‐3 and Bax, in UVB‐exposed HaCaT cells following Dc‐EE treatment, with corresponding quantitative analyses of cleaved caspase‐3 normalized to total caspase‐3 and Bax normalized to β‐actin. Quantitative data are presented as the mean ± SD of at least three independent replicates (n ≥ 3). Statistical significance was assessed using Student's t‐test. *p < 0.05 and **p < 0.01 versus the corresponding SNP‐ or UVB‐treated control group; #p < 0.05 and ##p < 0.01 versus the untreated normal group.

Taken together, these findings demonstrate that Dc‐EE effectively protects HaCaT keratinocytes from oxidative stress– and UVB‐induced injury in a concentration‐dependent manner. The cytoprotective effects of Dc‐EE are mediated through attenuation of nitrosative stress, suppression of intracellular ROS accumulation, preservation of normal cell morphology, and inhibition of apoptosis‐related signaling pathways.

To further characterize the protective effects of Dc‐EE against UVB‐induced skin damage, the expression of ECM–degrading enzymes was first examined in HaCaT keratinocytes. UVB irradiation (30 mJ/cm²) robustly induced the mRNA expression of MMP‐1, MMP‐2, and MMP‐9, reflecting enhanced matrix‐degrading activity. Pretreatment with Dc‐EE significantly and dose‐dependently suppressed the UVB‐induced upregulation of all three MMPs. Notably, Dc‐EE at 25 µg/mL partially reduced MMP expression, while stronger inhibition was observed at 100 µg/mL, and near‐complete suppression occurred at 200 µg/mL (Figure 4A). In parallel, the effect of Dc‐EE on UVB‐induced inflammatory signaling was evaluated by assessing COX‐2 expression in HaCaT cells. UVB exposure markedly increased COX‐2 mRNA levels, whereas Dc‐EE treatment significantly attenuated this induction in a concentration‐dependent manner. A modest reduction was observed at 25 µg/mL, with more pronounced suppression at 100 and 200 µg/mL, indicating effective inhibition of UVB‐triggered inflammatory responses at higher concentrations (Figure 4B). Given the critical role of dermal fibroblasts in collagen production and dermal integrity, the effects of Dc‐EE on collagen synthesis were next examined in human dermal fibroblast (HDF) cells. UVB irradiation substantially reduced the expression of collagen type I alpha 1 (COL1A1). Dc‐EE treatment significantly restored COL1A1 mRNA expression in a dose‐dependent manner, with 25 µg/mL partially reversing the UVB‐induced suppression and higher concentrations (100–200 µg/mL) leading to a marked recovery toward basal levels (Figure 4C). Finally, the impact of Dc‐EE on epidermal barrier– and hydration‐related gene expression was assessed in HaCaT keratinocytes. UVB irradiation markedly downregulated the expression of TGM‐1, FLG, and HAS‐2, genes essential for epidermal differentiation, barrier integrity, and skin hydration. Dc‐EE treatment dose‐dependently reversed the UVB‐induced suppression of all three genes. While partial recovery was evident at 25 µg/mL, Dc‐EE at 50 and 100 µg/mL significantly enhanced TGM‐1, FLG, and HAS‐2 expression, with HAS‐2 showing the most prominent induction at 100 µg/mL (Figure 4D). Band intensities of MMP‐1, MMP‐2, MMP‐9, COX‐2, COL1A1, TGM‐1, FLG, and HAS‐2 were quantified and normalized to GAPDH (Figure 4A–D, right panels).

FIGURE 4.

Dc‐EE attenuates UVB‐induced extracellular matrix degradation and inflammatory responses while restoring skin barrier–related gene expression. (A) mRNA expression of matrix metalloproteinases MMP‐1, MMP‐2, and MMP‐9 in HaCaT keratinocytes following UVB irradiation (30 mJ/cm²) and Dc‐EE treatment for 6 h (left panel), with relative band intensities quantified and normalized to GAPDH (right panel). (B) COX‐2 mRNA expression in UVB‐irradiated HaCaT cells treated with Dc‐EE (left panel), and corresponding densitometric analysis normalized to GAPDH (right panel). (C) Effects of Dc‐EE on collagen type I alpha 1 (COL1A1) mRNA expression in UVB‐irradiated human dermal fibroblasts (HDFs) (left panel), with relative band intensities normalized to GAPDH (right panel). (D) mRNA expression of epidermal barrier– and hydration‐related genes, including transglutaminase‐1 (TGM‐1), filaggrin (FLG), and hyaluronan synthase‐2 (HAS‐2), in HaCaT keratinocytes exposed to UVB irradiation and treated with Dc‐EE (left panel), with corresponding quantitative analysis normalized to GAPDH (right panel). Quantitative data are presented as the mean ± SD of at least three independent replicates (n ≥ 3). Statistical significance was assessed using Student's t‐test. *p < 0.05 and **p < 0.01 versus the UVB‐treated control group; #p < 0.05 and ##p < 0.01 versus the untreated normal group.

Collectively, these results demonstrate that Dc‐EE effectively counteracts UVB‐induced skin damage by suppressing ECM degradation and inflammatory gene expression while restoring collagen synthesis and key epidermal barrier– and hydration‐related genes in a concentration‐dependent manner.

Collectively, these findings demonstrate that Dc‐EE effectively protects skin cells from UVB‐induced oxidative stress, apoptosis, inflammation, ECM degradation, and barrier dysfunction. UVB‐induced skin injury is generally characterized by a tightly interconnected ROS–inflammation–matrix degradation cascade, in which excessive ROS production initiates inflammatory signaling and promotes ECM breakdown, ultimately leading to photoaging‐associated skin deterioration [28]. Consistent with this concept, Dc‐EE significantly reduced intracellular ROS accumulation, suppressed apoptosis‐related signaling, and restored cellular viability under both SNP‐ and UVB‐induced oxidative stress conditions. The preservation of keratinocyte morphology and reduction of ROS‐positive populations further suggest that Dc‐EE maintains epidermal homeostasis by limiting oxidative injury and preventing the cellular damage that precedes apoptosis [29].

Phytochemicals can protect stressed cells not only through direct radical‐scavenging activity but also by modulating endogenous redox‐responsive and cellular quality‐control pathways. In particular, phytochemical activation of the Nrf2‐mediated antioxidant response has been proposed to enhance cellular defense against oxidative stress and subsequent inflammatory and apoptotic injury [30]. In addition, recent evidence indicates that natural compounds can alleviate mitochondrial dysfunction and inflammatory cell injury by promoting mitophagy and suppressing NLRP3 inflammasome activation [31]. Although Nrf2 signaling, mitophagy, and inflammasome activation were not directly examined in the present study, these mechanisms provide a broader context for the antioxidant and cytoprotective activities of Dc‐EE and warrant investigation in future studies.

Beyond cytoprotection, Dc‐EE also modulated molecular markers directly associated with photoaging. UVB exposure is known to induce MMPs, thereby accelerating collagen degradation and wrinkle formation [6, 32, 33]. In the present study, Dc‐EE suppressed UVB‐induced MMP‐1, MMP‐2, and MMP‐9 expression in HaCaT keratinocytes while restoring COL1A1 expression in human dermal fibroblasts, suggesting coordinated regulation of epidermal stress responses and dermal collagen homeostasis. Furthermore, Dc‐EE reversed the UVB‐mediated downregulation of TGM‐1, FLG, and HAS‐2, genes closely associated with epidermal differentiation, barrier integrity, and skin hydration [34, 35, 36]. These observations indicate that Dc‐EE exerts multifunctional protective effects against UVB‐induced skin damage by targeting both dermal and epidermal components involved in photoaging.

2.3. Regulation of TAK1‐Associated Signaling Pathways

To further elucidate the molecular mechanisms underlying the protective effects of Dc‐EE against UVB‐induced skin damage, the activation of NF‐κB– and AP‐1–dependent signaling pathways was examined. Using luciferase reporter assays in HEK293T cells, Dc‐EE significantly and dose‐dependently suppressed both NF‐κB– and AP‐1–mediated transcriptional activities (Figure 5A,B), indicating its ability to modulate stress‐responsive transcriptional programs associated with oxidative stress and inflammatory signaling. Notably, Dc‐EE at 25, 50, and 100 µg/mL reduced NF‐κB activity to approximately 80%, 55%, and 35% of the activated control level, respectively (Figure 5A). A similar inhibitory trend was observed for AP‐1–dependent transcriptional activity, with Dc‐EE at 100 µg/mL decreasing AP‐1 activity by more than 60% compared with the activated control (Figure 5B). Consistent with these transcriptional findings, UVB irradiation markedly increased the phosphorylation of NF‐κB subunits p50 and p65 in HaCaT keratinocytes. Pretreatment with Dc‐EE significantly attenuated UVB‐induced phosphorylation of both p50 and p65 in a concentration‐dependent manner, with pronounced suppression observed at 50 and 100 µg/mL (Figure 5C, left panel). At the upstream regulatory level, UVB‐induced phosphorylation of IKKα/β and degradation of IκBα were also markedly reduced by Dc‐EE treatment (Figure 5D, left panel), indicating effective inhibition of NF‐κB activation along the signaling cascade. The effects of Dc‐EE on AP‐1 signaling were further evaluated by assessing the phosphorylation status of c‐Fos and c‐Jun. UVB exposure strongly enhanced phosphorylation of both AP‐1 components, whereas Dc‐EE treatment dose‐dependently suppressed their activation without significantly affecting total protein levels. Among the tested concentrations, Dc‐EE at 50 and 100 µg/mL exerted the most pronounced inhibitory effects (Figure 5F, left panel). Given that NF‐κB and AP‐1 signaling pathways are commonly regulated by MAPKs, the activation of ERK, p38, and JNK was subsequently assessed. UVB irradiation markedly induced phosphorylation of all three MAPKs, while Dc‐EE treatment significantly attenuated their activation in a concentration‐dependent manner. In particular, phosphorylation of p38 and JNK was strongly suppressed at 25 µg/mL and was nearly abolished at 100 µg/mL (Figure 5G, left panel). To further delineate upstream regulatory events, the phosphorylation of MAPK kinases MKK4 and MKK7 was examined. Dc‐EE significantly reduced UVB‐induced phosphorylation of both kinases, with stronger suppression observed at higher concentrations (Figure 5E, left panel). Importantly, phosphorylation of transforming growth factor‐β–activated kinase 1 (TAK1), a critical upstream regulator linking UVB‐induced stress to both NF‐κB and MAPK pathways, was markedly increased by UVB irradiation but dose‐dependently attenuated by Dc‐EE treatment, with substantial inhibition observed at 50 and 100 µg/mL (Figure 5H, left panel). Band intensities of phosphorylated forms of p50, p65, IKKα/β, IκBα, c‐Fos, c‐Jun, ERK, p38, JNK, MKK4, MKK7, and TAK1 were quantified, as indicated (Figure 5C–H, right panels).

FIGURE 5.

Dc‐EE attenuates UVB‐induced stress‐responsive signaling associated with the TAK1, NF‐κB, and MAPK pathways. (A, B) NF‐κB–dependent (A) and AP‐1–dependent (B) luciferase activities assessed in HEK293T cells after treatment with Dc‐EE (25–100 µg/mL) for 24 h. (C) Phosphorylation of NF‐κB subunits p50 and p65 in UVB‐irradiated HaCaT cells treated with Dc‐EE, with p‐p50 and p‐p65 levels normalized to their corresponding total p50 and p65 levels, respectively. (D) Effects of Dc‐EE on UVB‐induced phosphorylation of IKKα/β and IκBα in HaCaT cells, with p‐IKKα/β normalized to total IKKα and p‐IκBα normalized to total IκBα. (E) Phosphorylation of MKK4 and MKK7 in UVB‐irradiated HaCaT cells, with p‐MKK4 and p‐MKK7 normalized to their corresponding total MKK4 and MKK7 levels, respectively. (F) Phosphorylation of AP‐1 subunits c‐Fos and c‐Jun in UVB‐irradiated HaCaT cells, with p‐c‐Fos and p‐c‐Jun normalized to their corresponding total c‐Fos and c‐Jun levels, respectively. (G) Effects of Dc‐EE on UVB‐induced phosphorylation of ERK, p38, and JNK, with the phosphorylated proteins normalized to their corresponding total protein levels. (H) Phosphorylation of TAK1 in UVB‐irradiated HaCaT cells, with p‐TAK1 normalized to total TAK1. β‐Actin was used as a loading control for all immunoblot analyses in panels C–H. Relative band intensities were quantified and are shown in the corresponding graphs. Quantitative data are presented as the mean ± SD of at least three independent replicates (n ≥ 3). Statistical significance was assessed using Student's t‐test. *p < 0.05 and **p < 0.01 versus the corresponding activated or UVB‐treated control group; #p < 0.05 and ##p < 0.01 versus the untreated normal group.

Collectively, these results indicate that Dc‐EE attenuates UVB‐induced stress‐responsive and inflammatory signaling in association with reduced TAK1 activation and suppression of the NF‐κB, AP‐1, and MAPK signaling pathways in keratinocytes.

To further examine the association between Dc‐EE treatment and TAK1‐related signaling, a cellular thermal shift assay (CETSA) was performed to evaluate the thermal stability of TAK1 protein in cells. In HEK293T cells, Dc‐EE treatment markedly increased the thermal stability of TAK1 compared with the control group, as evidenced by enhanced resistance of TAK1 to heat‐induced denaturation across a broad temperature range (Figure 6A). These results indicate that Dc‐EE stabilizes TAK1 protein in cells, supporting an association between Dc‐EE treatment and TAK1‐related signaling. To assess the functional relevance of TAK1 modulation, HEK293T cells were transfected with HA‐tagged TAK1 to induce pathway activation. Overexpression of TAK1 markedly increased the expression of downstream stress‐ and inflammation‐associated mediators, including MMP‐1, MMP‐2, and COX‐2. Dc‐EE treatment significantly and dose‐dependently suppressed TAK1‐induced upregulation of MMP‐1, MMP‐2, and COX‐2 (Figure 6B,C), indicating effective attenuation of TAK1‐driven transcriptional responses. Consistent with these observations, TAK1 overexpression significantly enhanced AP‐1– and NF‐κB–dependent luciferase activities in HEK293T cells. Dc‐EE treatment dose‐dependently reduced TAK1‐induced activation of both AP‐1 and NF‐κB transcriptional reporters, with pronounced inhibition observed at higher concentrations (Figure 6D,E). To further support the involvement of TAK1 in Dc‐EE–mediated signaling regulation, quercetin—a polyphenolic compound reported to modulate stress‐responsive and inflammatory signaling—was employed as a comparative control. Similar to Dc‐EE, quercetin treatment markedly enhanced the thermal stability of TAK1 protein in CETSA experiments (Figure 6F) and significantly suppressed TAK1‐induced AP‐1– and NF‐κB–dependent luciferase activities (Figure 6G,H). In addition, molecular docking analysis was performed to explore potential interactions between TAK1 and selected candidate compounds associated with the Dc‐EE phytochemical profile. Docking results indicated that quercetin, patuletin‐7‐O‐glucoside, and mulberrofuran A could be accommodated within the TAK1 binding pocket and formed multiple hydrogen bonding and hydrophobic interactions with key residues of TAK1 (Figure 6I–K and Table 2). These exploratory computational findings provide hypotheses regarding possible ligand–TAK1 interactions but do not demonstrate direct binding or target engagement. Relative soluble TAK1 band intensities were quantified for the CETSA analyses (Figure 6A,F), whereas the RT‐PCR band intensities of MMP‐1, MMP‐2, and COX‐2 were normalized to GAPDH (Figure 6B,C).

FIGURE 6.

Dc‐EE attenuates TAK1‐associated stress‐responsive signaling and increases the thermal stability of TAK1. (A) Cellular thermal shift assay (CETSA) showing the thermal stability of TAK1 in HEK293T cells treated with Dc‐EE (100 µg/mL) for 24 h. Relative soluble TAK1 band intensities at each temperature were quantified, with the signal at 40°C in each treatment group set to 100%. (B) Effects of Dc‐EE on TAK1‐induced MMP‐1 and MMP‐2 mRNA expression in HA‐TAK1‐transfected HEK293T cells, with RT‐PCR band intensities normalized to GAPDH. (C) Effects of Dc‐EE on TAK1‐induced COX‐2 mRNA expression in HA‐TAK1‐transfected HEK293T cells, with RT‐PCR band intensities normalized to GAPDH. (D,E) AP‐1–dependent (D) and NF‐κB–dependent (E) luciferase activities in HA‐TAK1‐transfected HEK293T cells following Dc‐EE treatment for 24 h. (F) CETSA analysis of TAK1 thermal stability in HEK293T cells treated with quercetin (10 µM) for 24 h. Relative soluble TAK1 band intensities at each temperature were quantified, with the signal at 40°C in each treatment group set to 100%. (G, H) AP‐1–dependent (G) and NF‐κB–dependent (H) luciferase activities in HA‐TAK1‐transfected HEK293T cells following quercetin treatment. (I–K) Predicted binding poses of quercetin (I), patuletin‐7‐O‐glucoside (J), and mulberrofuran A (K) within the putative TAK1 binding pocket. Predicted hydrogen‐bonding and hydrophobic interactions between the ligands and TAK1 residues are shown. Quantitative data are presented as the mean ± SD of at least three independent replicates (n ≥ 3). Statistical significance was assessed using Student's t‐test. *p < 0.05 and **p < 0.01 versus the corresponding DMSO‐treated or TAK1‐overexpressing control group; #p < 0.05 and ##p < 0.01 versus the untreated or empty‐vector normal group.

TABLE 2.

Predicted docking scores and pKi values of selected compounds against TAK1.

Compound	Docking scores (kcal/mol) ± SD	Predicted pKi
Quercetin	−7.50 ± 0.51	5.52
Mulberrofuran A	−9.40 ± 0.20	6.93
Patuletin‐7‐O‐glucoside	−7.90 ± 0.18	5.83

The docking scores are presented as the mean ± SD of three independent docking runs (n = 3).

Collectively, these results demonstrate that Dc‐EE increases the thermal stability of TAK1 protein, suggesting a functional association with TAK1‐related signaling and attenuates stress‐responsive and inflammatory signaling associated with TAK1 activation, thereby suppressing downstream activation of AP‐1 and NF‐κB pathways.

Mechanistically, the present findings indicate that the protective effects of Dc‐EE were accompanied by changes in TAK1‐associated signaling. UVB irradiation markedly increased phosphorylation of TAK1 and activated downstream NF‐κB and MAPK signaling cascades, including phosphorylation of IKKα/β, p50, p65, ERK, p38, JNK, c‐Fos, and c‐Jun. Dc‐EE attenuated these signaling events in a concentration‐dependent manner, which corresponded closely with reduced expression of COX‐2 and MMPs, decreased apoptosis‐related signaling, and improved cellular survival. Such consistency between upstream signaling changes and downstream biological responses demonstrates a consistent association between attenuation of TAK1‐related signaling and the observed protective effects.

Additional experiments further strengthened the involvement of TAK1 in Dc‐EE‐mediated signaling regulation. CETSA analysis showed increased thermal stability of TAK1 following Dc‐EE treatment, suggesting a possible association with TAK1‐related signaling. Because CETSA was performed using the whole extract or quercetin in TAK1‐overexpressing cells, the observed thermal stabilization does not by itself establish direct binding of individual Dc‐EE constituents to TAK1. Moreover, Dc‐EE suppressed TAK1‐induced expression of MMP‐1, MMP‐2, and COX‐2 and attenuated TAK1‐induced NF‐κB and AP‐1 transcriptional activities. Similar effects were observed with quercetin, which was used as a comparative bioactive flavonoid in this study [37, 38, 39]. Exploratory molecular docking predicted energetically favorable poses of quercetin, patuletin‐7‐O‐glucoside, and mulberrofuran A within a putative TAK1 binding pocket. However, these computational predictions do not demonstrate direct biochemical binding or target engagement. In particular, mulberrofuran A was retained only as a Level 3 tentative candidate because no matched product‐ion evidence was available. Therefore, its predicted docking pose should not be interpreted as confirmation of its presence in Dc‐EE, direct TAK1 binding, or contribution to the biological effects of the extract.

Although docking analysis alone does not establish direct target engagement, the convergence of CETSA‐based stabilization, suppression of TAK1‐driven transcriptional responses, attenuation of downstream phosphorylation events, and structural predictions collectively support the hypothesis that TAK1 functions as a central signaling node influenced by Dc‐EE. Notably, TAK1 has been recognized as an important upstream regulator linking oxidative stress and inflammatory signaling pathways in skin pathology [18]. Therefore, modulation of TAK1‐associated signaling may provide a plausible molecular explanation for both the anti‐inflammatory activity observed in this study and the traditional use of D. chrysantha in skin‐related disorders.

A limitation of the present study is that the causal requirement for TAK1 in the protective effects of Dc‐EE was not directly established using TAK1 knockdown, knockout, or a selective pharmacological inhibitor. Therefore, the current findings support an association between Dc‐EE treatment and modulation of TAK1‐related signaling rather than demonstrating that TAK1 directly mediates the protective effects of Dc‐EE. Future studies employing genetic or pharmacological inhibition of TAK1 will be required to determine whether TAK1 is necessary for the effects of Dc‐EE on ROS accumulation, apoptosis, COX‐2 expression, and MMP regulation.

In addition, the gene‐expression findings obtained by conventional endpoint RT‐PCR were not independently validated using quantitative RT‐qPCR or protein‐level analyses. Future studies should confirm these major endpoints using validated quantitative and protein‐based approaches.

2.4. Proposed Mechanism of Action

Based on the collective experimental findings of this study, a schematic model was proposed to summarize the molecular mechanisms by which Dc‐EE protects skin cells from UVB‐induced damage (Figure 7). UVB irradiation triggers excessive production of ROS, which in turn activates stress‐responsive signaling pathways, leading to inflammatory responses, apoptotic cell death, and degradation of ECM components in skin cells. Dc‐EE effectively attenuates UVB‐induced ROS accumulation and suppresses NF‐κB and AP‐1 signaling in association with reduced TAK1 activation in keratinocytes. This suppression results in reduced expression of inflammatory mediators, such as COX‐2, and inhibition of apoptosis‐related signaling, thereby protecting HaCaT keratinocytes from UVB‐induced cell death. In parallel, Dc‐EE inhibits UVB‐induced upregulation of MMPs in HaCaT keratinocytes and restores collagen type I synthesis in human dermal fibroblasts, as reflected by increased COL1A1 expression. These coordinated effects contribute to the preservation of dermal ECM integrity and attenuation of UVB‐associated skin aging processes.

FIGURE 7.

Proposed schematic model illustrating the protective mechanisms of Dc‐EE against UVB‐induced skin damage. UVB irradiation induces excessive generation of reactive oxygen species (ROS), which activates stress‐responsive signaling pathways and promotes inflammation, apoptosis, and extracellular matrix degradation in skin cells. Dc‐EE attenuates ROS accumulation and suppresses NF‐κB and AP‐1 signaling in association with reduced TAK1 activation in keratinocytes, resulting in reduced COX‐2 expression and inhibition of apoptosis, thereby protecting epidermal cells from UVB‐induced damage. Dc‐EE suppresses matrix metalloproteinase (MMP) expression in HaCaT keratinocytes and enhances COL1A1 expression in human dermal fibroblasts, thereby potentially contributing to the maintenance of extracellular matrix integrity and attenuation of skin aging. This schematic summarizes the proposed molecular mechanisms of Dc‐EE based on the experimental results obtained in this study.

Taken together, this proposed model highlights Dc‐EE as a multifunctional protective agent that mitigates UVB‐induced oxidative stress and downstream stress‐responsive signaling, ultimately preserving epidermal cell viability and dermal structural integrity through coordinated regulation of TAK1‐associated pathways.

The phytochemical constituents tentatively annotated in Dc‐EE may collectively contribute to the observed biological effects through coordinated regulation of oxidative stress and inflammatory signaling pathways [25, 26, 27, 40]. Such multi‐target activity may be advantageous for preventing UVB‐induced skin damage, which involves simultaneous activation of multiple pathological processes. Overall, the proposed mechanism highlights Dc‐EE as a multifunctional phytochemical mixture capable of mitigating oxidative stress–associated skin damage through regulation of TAK1‐associated signaling networks.

3. Conclusion

In conclusion, D. chrysantha ethanol extract (Dc‐EE) exhibited significant antioxidant, anti‐inflammatory, and skin‐protective activities against UVB‐induced cellular damage. Phytochemical profiling by LC–MS/MS revealed the presence of diverse bioactive constituents, including flavonoids, phenolic compounds, and related secondary metabolites, which may collectively contribute to the observed biological effects. Dc‐EE effectively attenuated UVB‐induced oxidative stress, apoptosis, inflammatory responses, and ECM degradation while restoring collagen synthesis and epidermal barrier–related factors associated with skin hydration and homeostasis.

Mechanistic investigations indicated that these protective effects are associated with modulation of TAK1‐associated signaling and subsequent suppression of NF‐κB, AP‐1, and MAPK activation. The coordinated regulation of oxidative stress– and inflammation‐related pathways highlights the multifunctional nature of Dc‐EE and provides a molecular basis for its traditional use in skin‐related disorders.

Taken together, the present findings identify D. chrysantha as a promising source of bioactive phytochemicals with potential applications in skin protection and the prevention of UVB‐induced photoaging. Further studies aimed at the characterization of active constituents and validation in advanced skin models and in vivo systems will facilitate the development of Dc‐EE–based dermatological and cosmetic applications.

4. Experimental

4.1. Materials

Fetal bovine serum (FBS), penicillin‐streptomycin, dulbecco's modified Eagle's medium (DMEM), ascorbic acid, phosphate‐buffered saline (PBS), 1‐diphenyl‐2‐picryl‐hydrazyl (DPPH), potassium sulfate, 2,2'‐azino‐bis (3‐ethylbenzothiazoline‐6‐sulphonic acid) diammonium salt (ABTS), and trypsin were purchased from Hyclone (Grand Island, NY, USA). 3‐(4,5‐Dimethylthiazol‐2‐yl)‐2,5‐diphenyl tetrazolium bromide (MTT), sodium nitroprusside (SNP), 2′,7′‐dichlorodihydrofluorescein diacetate (H₂DCFDA), 4′,6‐diamidino‐2‐phenylindole (DAPI), TRIzol, polyethylenimine (PEI) and FeCl₃·6H₂O, 2,4,6‐tri(2‐pyridyl)‐s‐triazine (TPTZ), dimethyl sulfoxide (DMSO), CuCl₂·2H₂O, Trolox, NH₄Ac, neocuproine were purchased from Sigma Aldrich Chemical Co. (St. Louis, MO, USA). Forward and reverse primers which used as RT‐PCR were combined by Macrogen (Seoul, Korea), and PCR premix was received from Bio‐D Inc. (Seoul, Korea). Polyvinylidene difluoride (PVDF) membranes were purchased from Millipore (Billerica, MA, USA). Primary antibodies against total and phosphorylated forms of p50, p65, IKKα/β, IκBα, MKK4, MKK7, c‐Fos, c‐Jun, ERK, p38, JNK, and β‐actin were purchased from Cell Signaling Technology (CST, USA) or Santa Cruz Biotechnology (Santa Cruz, CA, USA), and enhanced chemiluminescence (ECL) were acquired from Bio‐Rad (Hercules, CA, USA).

4.2. Extract Processing for Dc‐EE Preparation and LC–MS (TOF‐MS) Analysis

The plant material of D. chrysantha (Zoll. & Moritzi) Miq. was collected in Daegang‐myeon, Danyang‐gun, Chungcheongbuk‐do, Republic of Korea, on June 20, 2018, by Jin Yeonghoon and Kang Kyoungsuk of the Northeast Asia Biodiversity Institute. The material was provided and verified by the Wildlife Natural Products Bank at the National Institute of Biological Resources (NIBR), Incheon, Republic of Korea, under accession number NIBRGR0000611257. A voucher specimen was deposited at NIBR under voucher number NIBRVP0000725025. The dried leaves and stems (1 kg) were cut into small pieces and extracted three times with 70% (v/v) ethanol for 3 h each using ultrasonic‐assisted extraction. The combined extracts were filtered, concentrated by rotary evaporation, and subsequently freeze‐dried at −80°C for 72 h, yielding 141.4 g of dried Dc‐EE from 1 kg of dried plant material, corresponding to an extraction yield of 14.14%. The primary stock solution of Dc‐EE was prepared in DMSO at a concentration of 100 mg/mL and diluted with culture medium for each experiment.

Untargeted metabolomics analysis was performed using liquid chromatography–tandem mass spectrometry (LC‐MS). Specifically, an ultra‐performance liquid chromatography system (UPLC, Waters, Milford, USA) coupled with a high‐resolution quadrupole time‐of‐flight mass spectrometer (SYNAPT XS QTOF, Waters) was used. Chromatographic separation was conducted on an ACQUITY UPLC HSS T3 column (100 × 2.1 mm, 1.8 µm, Waters) maintained at 40°C, with a flow rate of 0.5 mL/min. The mobile phase consisted of solvent A (water with 0.1% formic acid) and solvent B (acetonitrile with 0.1% formic acid). The gradient elution program was as follows: 0–5 min, 97% A; 5–16 min, linear increase from 3% to 100% B; 16–17 min, 100% B; 17–19 min, linear return from 100% to 3% B; and 19–25 min, re‐equilibration with 97% A. The injection volume for each sample was 5 µL. The mass spectrometer was operated in both positive and negative electrospray ionization modes. For positive ion mode, the capillary voltage and cone voltage were set at 2 kV and 40 V, respectively; for negative mode, they were set at 1 kV and 40 V. Data were acquired in centroid MSE mode. The scan range for both MS and MS/MS was m/z 50–1200, with a scan time of 0.2 s. Precursor ions were fragmented using collision energies set between 20 and 40 eV. To ensure mass accuracy, leucine enkephalin ([M + H]⁺ = 556.2771; [M − H]⁻ = 554.2615) was infused continuously at 10 µL/min via a lock spray interface and sampled every 3 s. Data acquisition and processing were performed using UNIFI software (version 1.71, Waters).

4.3. Antioxidant Assays: DPPH, ABTS, FRAP, and CUPRAC

The antioxidant activity of Dc‐EE was evaluated using four well‐established assays (DPPH, ABTS, FRAP, and CUPRAC), as described in previous studies with slight modifications [41].

4.4. Hyaluronan (HA) Quantification

HA levels were determined using a carbazole‐based colorimetric assay, as previously described [42]. Briefly, culture supernatants and cell pellets were prepared and reacted with borax/sulfuric acid and carbazole reagents. After heating for color development, absorbance was measured at 550 nm using a microplate reader.

4.5. Cell Culture

Human keratinocytes (HaCaT; RRID: CVCL_0038), human embryonic kidney cells (HEK293T; RRID: CVCL_0063), and mouse melanoma cells (B16F10; RRID: CVCL_0159) were purchased from the American Type Culture Collection (ATCC; Rockville, MD, USA). Neonatal primary human dermal fibroblasts (HDFs; SKU: FC‐0001) were obtained from Lifeline Cell Technology (Oceanside, CA, USA). All cells were cultured in DMEM supplemented with 10% FBS and 1% penicillin–streptomycin and maintained at 37°C in a humidified incubator containing 5% CO₂.

4.6. Cell Viability

To evaluate the cytotoxicity of Dc‐EE, a colorimetric MTT assay was performed according to a previously described protocol [43].

4.7. Nitric Oxide (NO) Production Assay

SNP‐derived NO production was determined using a commercial Griess reagent kit, following a previously described protocol [44]. HaCaT cells were seeded into 96‐well plates at a density of 5 × 10⁵ cells/mL. After incubation for 18 h, the cells were treated with Dc‐EE for 30 min, followed by stimulation with SNP. After a further 24 h incubation, the NO levels in the culture supernatants were quantified by measuring absorbance at 540 nm.

4.8. UVB Irradiation

HaCaT and HDF cells were exposed to UVB irradiation as previously described [43]. Briefly, cells were pretreated with Dc‐EE, washed with PBS, and exposed to UVB irradiation at a dose of 30 mJ/cm². Following irradiation, cells were incubated with Dc‐EE for the indicated time periods prior to subsequent analyses.

4.9. Cell Morphology Observation

HaCaT cells were seeded at a density of 1 × 10⁵ cells/mL and pretreated with various concentrations of Dc‐EE (25–100 µg/mL) for 30 min. After pretreatment, the cells were rinsed with PBS and exposed to UVB irradiation. Subsequently, fresh culture medium supplemented with Dc‐EE at the corresponding concentrations was added, and the cells were incubated for an additional 24 h. Cellular morphological changes were then examined and recorded using an epifluorescence microscope (Olympus, Tokyo, Japan).

4.10. Intracellular ROS Levels

Intracellular ROS levels were measured using the fluorescent probe H₂DCFDA. HaCaT cells were seeded in 6‐well plates and allowed to attach overnight, then pretreated with Dc‐EE prior to UVB irradiation at a dose of 30 mJ/cm². Following UVB exposure, cells were further incubated with Dc‐EE for 24 h. Intracellular ROS levels were analyzed by flow cytometry and fluorescence microscopy (excitation/emission: 492–495/517–527 nm).

4.11. Western Blot Analysis

HaCaT cells were lysed in extraction buffer containing 50 mM Tris‐HCl (pH 7.5), 120 mM NaCl, 20 mM NaF, 25 mM β‐glycerophosphate (pH 7.5), 2% NP‐40, and a protease/phosphatase inhibitor cocktail. The lysates were clarified by centrifugation at 13 000 rpm for 15 min at 4°C to remove cellular debris. Protein concentrations were determined using the Bradford assay. Equal amounts of protein (20 µg per lane) were separated by SDS‐PAGE using Tris‐glycine gels and transferred onto PVDF membranes. Membranes were blocked with 3% BSA at room temperature for 1 h and then washed three times (10 min each) with TBST (TBS containing 0.1% Tween‐20). Primary antibodies were incubated overnight at 4°C, followed by incubation with secondary antibodies for 2 h at room temperature. Protein bands were visualized using EzWestLumi Plus chemiluminescent substrate (ATTO Corporation, Tokyo, Japan). Band intensities were quantified using ImageJ. Phosphorylated proteins were normalized to their corresponding total protein levels. Cleaved caspase‐3 was normalized to total caspase‐3, whereas Bax was normalized to β‐actin. β‐Actin was used as the loading control for the immunoblot analyses.

4.12. MRNA Expression Analysis

Reverse transcription PCR (RT‐PCR) was performed to analyze mRNA expression levels. HaCaT and HDF cells were seeded in 6‐well plates at a density of 3 × 10⁵ cells per well. Total RNA was extracted using TRIzol reagent, and RNA concentrations were measured spectrophotometrically, as previously described [44]. PCR amplification was conducted under the following thermal cycling conditions: initial denaturation at 98°C for 5 min, followed by 35 cycles of denaturation at 98°C for 15 s, annealing at 56°C–61°C for 15 s, and extension at 72°C for 1 min, with a final extension at 72°C for 5 min. The primer sequences used in this study are listed in Table 3. RT‐PCR band intensities were quantified using ImageJ and normalized to GAPDH. Because conventional endpoint RT‐PCR followed by gel densitometry was used, the results were interpreted as semiquantitative estimates of relative gene expression rather than precise quantitative measurements.

TABLE 3.

Primer sequences used in this study for RT‐PCR.

Gene name	Direction	Sequences (5′ to 3′)
MMP‐1 (Homo sapiens)	Forward	TCTGACGTTGATCCCAGAGAGCAG
	Reverse	CAGGGTGACACCAGTGACTGCAC
MMP‐2 (H. sapiens)	Forward	ACGACCGCGACAAGAAGTAT
	Reverse	CTGCAAAGAACACAGCCTTCTC
MMP‐9 (H. sapiens)	Forward	GCCACTTGTCGGCGATAAGG
	Reverse	CACTGTCCACCCCTCAGAGC
COX‐2 (H. sapiens)	Forward	CAAAAGCTGGGAAGCCTTCT
	Reverse	CCATCCTTCAAAAGGCGCAG
COL1A1 (H. sapiens)	Forward	CAGGTACCATGACCGAGACG
	Reverse	AGCACCATCATTTCCACGAG
TGM‐1 (H. sapiens)	Forward	GAAATGCGGCAGATGACGAC
	Reverse	AACTCCCCAGCGTCTGATTG
FLG (H. sapiens)	Forward	AGGGAAGATCCAAGAGCCCA
	Reverse	ACTCTGGATCCCCTACGCTT
HAS‐2 (H. sapiens)	Forward	CCACCCAGTACAGCGTCAAC
	Reverse	CATGGTGCTTCTGTCGCTCT
GAPDH (H. sapiens)	Forward	GCACCGTCAAGGCTGAGAAC
	Reverse	ATGGTGGTGAAGACGCCAGT

4.13. Luciferase Reporter Assay

HEK293T cells were seeded in 24‐well plates at a density of 1 × 10⁵ cells per well and cultured overnight. The next day, the medium was replaced with 400 µL of fresh DMEM. Cells were then co‐transfected with 0.8 µg/mL of either AP‐1‐luc or NF‐κB‐luc plasmid, along with β‐galactosidase, using polyethylenimine (PEI) as the transfection reagent. After 24 h of transfection, the medium was replaced with DMEM containing Dc‐EE (25–100 µg/mL), and cells were incubated for an additional 24 h. Luciferase activity was quantified using a luminometer (BioTek Instruments Inc., Winooski, VT, USA).

4.14. Overexpression

HEK293T cells (1 × 10⁶ cells/mL) were seeded into 6‐well plates and incubated overnight. For overexpression, 4 µg of either HA‐Tag or HA‐TAK1 plasmid was transfected using PEI. After 24 h of transfection, cells were treated with Dc‐EE or DMSO and incubated for an additional 6 or 24 h. Following treatment, cells were harvested for subsequent experiments, including the CETSA and RT‐PCR.

4.15. Cellular Thermal Shift Assay (CETSA)

CETSA was performed as previously described [45] to assess the thermal stability of TAK1. HEK293T cells overexpressing HA‐TAK1 were treated with DMSO, Dc‐EE (100 µg/ml), or quercetin (10 µM) for 24 h. Cells were harvested, washed with cold PBS, and resuspended in PBS containing protease inhibitors. Cell suspensions were aliquoted and heated at the indicated temperatures (40°C–70°C) for 3 min, followed by cooling on ice. After heating, cells were lysed by three freeze–thaw cycles and centrifuged to remove aggregated proteins. The soluble fractions were subjected to SDS–PAGE and immunoblotting using an anti‐TAK1 antibody. Soluble TAK1 band intensities at each temperature were quantified using ImageJ and expressed relative to the corresponding signal at 40°C, which was set to 100% for each treatment group.

4.16. Network Pharmacology and In Silico Analysis

Publicly available online databases and software tools were used for network pharmacology analysis. The details of databases and software are summarized in Tables 4 and 5.

TABLE 4.

Database URLs used in this paper.

Database name	URL
PubChem	https://pubchem.ncbi.nlm.nih.gov/
Swiss Target Prediction	http://www.swisstargetprediction.ch/
GeneCards v5.26	https://www.genecards.org/
HGNC	https://www.genenames.org/
Omicshare	https://www.omicshare.com/tools/
Protein Data Bank (PDB)	http://www.rcsb.org/pdb/
CB‐Dock	http://cao.labshare.cn/cb‐dock/

TABLE 5.

Programs used in this paper.

Program name
SigmaPlot 10.0
CSAnalyzer4
ImageJ
Discovery Studio Visualizer

4.17. Functional Characterization via GO and KEGG Enrichment

To functionally characterize the predicted targets associated with Dc‐EE, GO, and KEGG enrichment analyses were performed. GO enrichment was conducted across the BP, CC, and MF categories, while KEGG analysis was used to identify signaling pathways potentially involved in the protective effects of Dc‐EE.

4.18. Target Identification Criteria in SwissTargetPrediction

Compound‐related targets were identified using SwissTargetPrediction. The analysis was restricted to Homo sapiens, and only targets with a nonzero probability score were retrieved. Following commonly used criteria in network pharmacology, targets with a probability ≥0.1 or ranked within the top 15 predicted targets were selected for downstream analyses.

4.19. Molecular Docking Simulation

Quercetin was included as a comparative bioactive flavonoid based on its reported involvement in stress‐responsive and inflammatory signaling and its experimentally observed effects on TAK1 in this study. Patuletin‐7‐O‐glucoside, assigned as a Level 2 putative annotation, and mulberrofuran A, retained as a Level 3 tentative candidate, were selected for exploratory molecular docking with TAK1. Because the identity of mulberrofuran A was not supported by matched product‐ion evidence or an authentic standard, its docking result was interpreted solely as a computational hypothesis. The crystal structure of human TAK1 (PDB ID: 5JGD, chain A; resolution, 3.10 Å) was obtained from the Protein Data Bank. The three‐dimensional structures of quercetin, mulberrofuran A, and patuletin‐7‐O‐glucoside were obtained from PubChem in SDF format. Molecular docking was performed using CB‐Dock, which automatically detects potential protein cavities and performs blind docking using AutoDock Vina. For each ligand, the binding region was defined as the top‐ranked cavity automatically identified by CB‐Dock, and the pose with the most favorable predicted binding energy within that cavity was selected for subsequent interaction analysis. Docking was performed independently three times for each ligand, and the resulting docking scores are presented as the mean ± standard deviation. All docking calculations were performed using the default CB‐Dock settings. Predicted pKi values were calculated at 298 K using the equation pKi = −ΔG/(2.303RT), where ΔG is the docking‐derived binding energy expressed in kcal/mol and R is the gas constant (0.001987 kcal·mol⁻¹·K⁻¹). Predicted pKi values are dimensionless. The selected ligand–protein complexes were visualized using Discovery Studio Visualizer to examine the predicted binding orientations and potential hydrogen‐bonding and hydrophobic interactions between each ligand and TAK1 residues. Because this exploratory docking protocol was not validated by redocking a co‐crystallized ligand and the predicted interactions were not confirmed using orthogonal biochemical binding assays, the docking poses, binding energies, and predicted pKi values were interpreted as computational hypotheses rather than evidence of direct binding or target engagement with TAK1.

4.20. Statistical Analysis

Data are expressed as the mean ± standard deviation (SD) of at least three independent replicates (n ≥ 3). Graphical representations were generated using SigmaPlot software (Systat Software Inc., San Jose, CA, USA). Statistical significance between groups was assessed using Student's t‐test. A p‐value < 0.05 was considered statistically significant.

Author Contributions

Yerkyesh Khamit: writing – original draft, visualization, investigation, formal analysis, data curation, conceptualization. Lei Huang: writing – original draft, visualization, investigation, formal analysis, data curation, conceptualization. Yena Oh: investigation. Jinghan Su: investigation. Dilda Kaskadamova: investigation. Segon Jang: investigation. Weizhaole Wei: investigation. Sarah Lee: investigation. Eun Sil Kim: investigation. Youn Kyoung Son: investigation. Byong Chul Yoo: writing – review and editing, supervision, conceptualization. Jongsung Lee: writing – review and editing, supervision, project administration, conceptualization. Jae Youl Cho: writing – review and editing, supervision, project administration, funding acquisition, formal analysis, data curation, conceptualization.

Funding

This work was supported by a grant from the National Institute of Biological Resources (NIBR), funded by the Ministry of Climate, Energy and Environment (MCEE) of the Republic of Korea (NIBR202419202, NIBR202619101).

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data are available from the corresponding author upon reasonable request.