Chrysanthemum morifolium extract protects HT22 cells from oxidative stress and enhances sleep via GABAergic modulation: contribution of luteolin‐7‐glucoside

Abstract

BACKGROUND

Chrysanthemum morifolium has long been utilized in traditional medicine for its antioxidant and sedative properties, yet its potential in sleep regulation and neuroprotection against oxidative stress remains underexplored. This study investigates the antioxidant and sleep‐enhancing effects of C. morifolium extract (CE) in HT22 hippocampal neuronal cells and animal models, focusing on its bioactive component, luteolin‐7‐glucoside (L‐7‐G) and the underlying molecular mechanisms.

RESULTS

CE exhibited potent antioxidant activity, significantly reducing reactive oxygen species and malondialdehyde levels while improving cell survival under oxidative stress. It upregulated antioxidant enzyme expression and activated the Nrf2/HO‐1 pathway, effects linked to L‐7‐G. In pentobarbital‐induced sleep models, CE increased sleep duration by up to 1.41‐fold compared to controls, with L‐7‐G showing similar efficacy. Electroencephalography revealed enhanced δ‐wave activity and prolonged non‐rapid eye movement (NREM) sleep with CE treatment. In a caffeine‐induced insomnia model, CE reduced wake time, restored NREM sleep, and counteracted sleep disturbances by enhancing δ‐wave activity. Mechanistic analysis showed CE upregulated GABA_A and 5‐HT_1A receptor expression, with its sleep‐promoting effects partially mediated through GABA_A receptor interactions.

CONCLUSION

CE, particularly through L‐7‐G, exhibits significant potential as a natural agent for reducing oxidative stress and enhancing sleep quality, offering a promising avenue for managing sleep disorders. © 2025 The Author(s). Journal of the Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

INTRODUCTION

Sleep is vital for regulating numerous physiological functions, including metabolic balance, immune system efficiency, and cognitive processes. ¹ However, the pressures of a modern 24‐hour society have resulted in widespread sleep deprivation, raising significant public health concerns. During wakefulness, the central nervous system is particularly vulnerable to oxidative stress due to the brain's relatively low levels of endogenous antioxidants and high concentrations of transition metals such as iron. ² These factors, combined with redox reactions occurring during metabolic processes like phosphorylation and prostaglandin synthesis, contribute to the production of reactive oxygen species (ROS). ³ Sleep plays a critical role in counteracting these effects by alleviating oxidative stress, facilitating cellular recovery, and enabling the removal of metabolic waste in both the brain and peripheral tissues. ⁴

Existing sleep aids, such as benzodiazepines and zolpidem, provide short‐term relief but carry risks of dependence and cognitive impairment. ⁵ Cognitive behavioral therapy for insomnia (CBT‐I) is effective but requires long‐term commitment. ⁶ Advanced sleep technologies, including smart mattresses and wearable devices, are being developed, yet high costs and limited validation hinder widespread use. ⁷ These limitations highlight the need for alternative approaches, such as natural products that target oxidative stress and neurotransmitter pathways to improve sleep.

Chrysanthemums, belonging to the Asteraceae family, consist of approximately 300 species, with 40 species widely distributed across East Asia. ⁸ In traditional Chinese medicine (TCM) and Korean medicine, Chrysanthemum morifolium has been used in the form of tea and decoctions to treat headaches, dizziness, and insomnia due to its calming and sleep‐promoting properties. ⁹ , ¹⁰ , ¹¹ Historically, texts such as the Shennong Bencao Jing (Shen Nong's Herbal Classic), a foundational TCM reference from the Eastern Han Dynasty (AD 25–220), document its use for calming the mind and promoting restful sleep. ¹² In Korean medicine, it is similarly recorded in the Donguibogam, a comprehensive medical text, as a remedy for insomnia and nervous tension (Heo, 1613). Studies suggest that C. morifolium plays a role in modulating oxidative stress and supporting cardiovascular health. ¹³ , ¹⁴ Additionally, studies have indicated that C. morifolium possesses antioxidant properties that may contribute to fatigue resistance, cardiovascular function enhancement, and serum lipid regulation. ¹⁵ These beneficial effects are attributed to the bioactive compounds found in chrysanthemums, including flavonoids and phenolic acids, which exhibit strong antioxidant properties. ¹⁶ Chrysanthemum morifolium extract (CE) contains key phytochemicals such as chlorogenic acid, quercetin, luteolin, apigenin, quercetin‐3‐O‐β‐d‐glucopyranoside (Q3G), luteolin‐7‐O‐β‐d‐glucopyranoside, and apigenin‐7‐O‐β‐d‐neohesperidoside. ¹⁷ These bioactive compounds are known for their antioxidant, antitumor, and anti‐inflammatory properties. ¹⁸ Among them, apigenin and Q3G have been reported to exhibit sedative–hypnotic effects through the benzodiazepine receptor‐binding mechanism. ¹⁹ Given their structural similarity, the potential hypnotic effects of luteolin and its glycosylated form remain largely unexplored. In particular, luteolin‐7‐glucoside (L‐7‐G) has been recognized for its antioxidant and neuroprotective properties, but its role in sleep regulation has yet to be fully elucidated.

CE was shown to mitigate oxidative damage by activating cytoprotective and antioxidant genes. ²⁰ Recent studies have also highlighted its potential role in improving sleep, as evidenced by its ability to prolong pentobarbital‐induced sleep duration through chloride channel activation. ²¹ However, despite these promising findings, scientific evidence regarding the direct effects of chrysanthemum on sleep regulation remains limited.

To address this gap, we systematically investigated the sleep‐enhancing effects of CE using both cellular and animal models. Specifically, we explored its potential to mitigate oxidative stress in hippocampal cells induced by hydrogen peroxide (H₂O₂) and analyzed its effects on sleep architecture through electroencephalogram (EEG) recordings. Furthermore, we examined the interaction of CE components with GABA_A receptors to elucidate possible mechanisms of action. By doing so, we seek to provide novel insights into the bioactive components of CE and their role in sleep regulation.

MATERIALS AND METHODS

Chemicals and reagent

CE is a 500 mL L⁻¹ ethanol extract, containing 3.5 g kg⁻¹ L‐7‐G (Lot. CM‐01‐210 921), provided by NOVAREX Co., Ltd (Cheongju, Korea). The following chemicals and reagents were used in this study: 2,4,6‐tripyridyl‐S‐triazine (TPTZ), 2,7‐dichlorofluorescin diacetate (DCF‐DA), p‐dimethylaminocinnamaldehyde, γ‐aminobutyric acid (GABA), and three types of antagonists, all of which were purchased from Sigma‐Aldrich (St Louis, MO, USA). The heptahydrate form of iron(II) sulfate and the hexahydrate form of iron(III) chloride were obtained from Shinyo Co., Ltd (Tokyo, Japan). Dulbecco's Modified Eagle's Medium (DMEM), fetal bovine serum (FBS), penicillin–streptomycin (PS), and l‐glutamine were sourced from Welgene (Gyeongsan, Korea).

Analysis of radical scavenging activities

The radical scavenging activities of CE against 2,2′‐azino‐bis(3‐ethylbenzthiazoline‐6‐sulfonic acid (ABTS) and 2,2‐diphenyl‐1‐picrylhydrazyl (DPPH) were evaluated using modified methods described by Blois ²² and Re et al., ²³ respectively. Absorbance was measured at 414 and 520 nm for ABTS and DPPH, respectively, and IC₅₀ values were calculated based on the obtained data.

Assessment of antioxidant capacity

The ferric reducing antioxidant power (FRAP) assay reagent was formulated by combining acetate buffer (300 mmol L⁻¹, pH 3.6), TPTZ (10 mmol L⁻¹ in 40 mmol L⁻¹ HCl), and iron(III) chloride hexahydrate (20 mmol L⁻¹) in a 10:1:1 ratio. For the assay, the sample (6 μL) was mixed with the FRAP reagent (180 μL) and allowed to incubate for 5 min. The absorbance was measured at 593 nm. Iron(II) sulfate heptahydrate served as the standard for determining the FRAP value. ²⁴

Polyphenol and flavonoid content analysis

The polyphenol and flavonoid contents of CE were analyzed using spectrophotometric methods. For total polyphenol content, the CE was reacted with 200 g L⁻¹ sodium carbonate and 0.9 N Folin–Ciocâlteu reagent at for 2 h. The absorbance was then recorded at 750 nm, with gallic acid serving as the standard for quantification. ²⁵ For flavonoid content, the p‐dimethylaminocinnamaldehyde method was employed. The absorbance was measured at 640 nm, with catechin used as the reference compound for the analysis. ²⁶

Cell culture conditions

HT22 cells – mouse hippocampus‐derived neuronal cells – were cultured in DMEM supplemented with 100 mL L⁻¹ FBS, 10 mL L⁻¹ PS, and 10 mL L⁻¹  l‐glutamine. The incubator used for cell culture was at 37 °C with 5% CO₂.

Cell viability assessment

To evaluate the impact of CE on HT22 cell viability under both normal and oxidative stress conditions, cells were seeded at 2 × 10⁴ cells per well in 96‐well plates and grown to 70–80% confluency. Under normal conditions, cells were treated with CE at concentrations of 0, 50, 100, 200, and 400 μg mL⁻¹ in DMEM containing 20 mL L⁻¹ FBS for 24 h before viability was assessed using the WST‐8 assay kit (Biomax, Gyeonggi‐do, Korea). For oxidative stress conditions, cells were pretreated with the same concentrations of CE for 2 h, followed by co‐treatment with 400 μmol L⁻¹ H₂O₂ (added without removing CE) for an additional 24 h, after which cell viability was measured.

Measurement of intracellular ROS and lipid peroxidation

To assess intracellular ROS levels, 2 × 10⁴ cells per well were plated in 96‐well black plates and cultured for 24 h. Following treatment with the sample solution for 2 h, the cells were exposed to 0.4 mmol L⁻¹ H₂O₂ for an additional 24 h. Subsequently, 10 μmol L⁻¹ DCF‐DA was added under light‐shielded conditions, and after 30 min fluorescence intensity was quantified with a multifunctional microplate reader (Victorx3, PerkinElmer, CT, USA) set to an excitation and emission wavelength of 485 and 535 nm. ²⁷ To determine the changes in malondialdehyde (MDA) levels in HT22 cells following CE pretreatment and H₂O₂ exposure, a thiobarbituric acid reactive substances (TBARS) assay kit (Biomax) was performed, following the manufacturer's protocol.

RNA isolation and quantitative polymerase chain reaction analysis

For RNA isolation, total RNA was extracted from HT22 cells and brain tissue using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). After 24 h of CE treatment, HT22 cells were rinsed with DPBS, lysed with 1 mL TRIzol, and the lysates were processed for RNA extraction. For brain tissue, 50 mg of tissue was homogenized in 1 mL TRIzol using a tissue homogenizer, and the supernatant was collected for RNA extraction. RNA purification, genomic DNA removal, and cDNA synthesis were carried out following the manufacturer's instructions. Quantitative reverse transcription polymerase chain reaction (PCR) was conducted on a real‐time PCR system utilizing the SYBR Green system (Applied Biosystems, CA, USA). Gene expression was analyzed using the ΔΔCt method as described by Livak and Schmittgen ²⁸ , with primer sequences listed in Supporting Information, Table S1. The genes analyzed by PCR included SOD1 (superoxide dismutase 1), GSH‐Px (glutathione peroxidase), CAT (catalase), Gabrg1 (γ‐aminobutyric acid type A receptor subunit γ1), Gabbr1 (γ‐aminobutyric acid type B receptor subunit 1), and Gabbr2 (γ‐aminobutyric acid type B receptor subunit 2).

Western blot analysis

Proteins extraction from HT22 cells and brain tissue was carried out using a lysis buffer supplemented with protease and phosphatase inhibitors. The lysates were then centrifuged (10 000 × g, 10 min, 4°C), and the supernatants were collected for analysis. Proteins (30 μg) were resolved via sodium dodecyl sulfate–polyacrylamide gel electrophoresis and subsequently transferred onto PVDF membranes. The membranes were blocked with 50 g L⁻¹ bovine serum albumin for 1 h, followed by incubation with primary antibodies (1:1000 dilution) at 4 °C overnight. After washing the membrane three times with TBST, they were treated with secondary antibodies (1:2000 dilution) for 2 h at room temperature. Details of the antibodies used are provided in Supporting Information, Table S2. Protein expression was detected using an enhanced chemiluminescence kit and visualized using an imaging system (EZ‐Capture ST, Tokyo, Japan). The analyzed proteins included Nrf2 (nuclear factor erythroid 2‐related factor 2), HO‐1 (heme oxygenase‐1), Keap1 (Kelch‐like ECH‐associated protein 1), GABA_A receptor, GABA_B receptor, and 5‐HT_1A receptor.

Experimental animals

ICR male mice (30 ± 2 g) were obtained from Orent Bio Inc. (Seongnam, Korea), and Sprague–Dawley (SD) rats (300 ± 20 g) were supplied by DBL (Eumseong, Korea). All procedures were approved by the Korea University Institutional Animal Care and Use Committee (KUIACUC 2024–0048). Before the experiments, the animals were housed for a minimum of 1 week to acclimate to the experimental environment, which included a temperature of 20 ± 2 °C, a relative humidity level of 50 ± 5%, and a 12 h light/dark cycle. This study was conducted in compliance with the Animal Research: Reporting In Vivo Experiments (ARRIVE) guidelines and other relevant reporting standards.

Pentobarbital‐induced sleep assessment

To evaluate the impact of CE on sleep parameters (latency and duration), a pentobarbital‐induced sleep test was performed. Male ICR mice were deprived of food for more than 18 h before the experiment. Following oral administration of CE, pentobarbital (42 mg kg⁻¹) was injected intraperitoneally 40 min later to induce sleep. Mice that did not fall asleep within 15 min after the pentobarbital injection were excluded from the study. The experimental groups (n = 7 per group) comprised: normal (NOR; water); positive control (PC; GABA 100 mg kg⁻¹); and low and high‐dose CE (CL 100 mg kg⁻¹ and CH 200 mg kg⁻¹). Sleep duration was recorded based on recovery of the righting reflex.

GABA_A receptor binding test using antagonist

A GABA_A receptor binding test was conducted to evaluate the effect of CE on sleep using specific antagonists. The doses of GABA_A receptor antagonists were selected based on previous studies, ²⁹ and optimized to ensure efficacy without adverse effects. Picrotoxin (PIX, 4 mg kg⁻¹), bicuculline (BIC, 6 mg kg⁻¹), and flumazenil (FMZ, 10 mg kg⁻¹) were each administered via intraperitoneal injection 20 min prior to CE administration. The antagonists were dissolved in edible oil solution containing 50 mL L⁻¹ dimethyl sulfoxide, and the NOR group received an intraperitoneal injection of the vehicle alone. CE was dissolved in distilled water and administered orally at 200 mg kg⁻¹. The experimental groups included: normal (NOR; vehicle), PIX, BIC, FMZ, and groups treated with both the respective antagonist and CE (200 mg kg⁻¹), namely PIX + CE, BIC + CE, and FMZ + CE. Sleep latency and sleep duration were evaluated following the same procedure as the pentobarbital‐induced sleep test.

EEG analysis

EEG electrodes were implanted following established surgical procedures. ³⁰ Briefly, under isoflurane anesthesia (Piramal Critical Care, Inc., Bethlehem, PA, USA) with oxygen and nitrous oxide mixture, rats were fixed in a stereotaxic frame (Stoeling Inc., Wood Dale, IL, USA). After shaving and disinfecting the scalp with 700 mL L⁻¹ ethanol, a midline incision was made to expose the skull. Four holes were drilled at specific coordinates relative to the bregma: +1 mm lateral/+1 mm anterior for the frontal cortex, and −1 mm lateral/−3 mm posterior for the hippocampus and striatum. EEG electrodes with mounting screws and sockets were fixed to the skull surface using dental cement (AgnThós AB, Lidingö, Sweden). After a 7‐day recovery period, EEG transmitters were attached, and recordings were performed for 7 h (10:00 to 17:00) on the day of oral CE administration. Each experimental group consisted of six rats. EEG signals were captured at a sampling speed of 15 mm s⁻¹. Sleep architecture was evaluated using the fast Fourier transform algorithm implemented in ecgAUTO3 software (version 3.3, emka Technologies, Paris, France). This analysis evaluated parameters such as sleep, wakefulness, rapid eye movement (REM) sleep, and non‐REM (NREM) sleep, including theta and delta waves.

To control for potential confounders, EEG analysis was conducted in soundproof chambers, and a consistent 12 h light/dark cycle was maintained throughout the experiment. This ensured uniform experimental conditions and minimized environmental influences on the EEG data.

In the basic EEG experiment, the following groups received their respective oral treatments: NOR (water), PC (GABA 100 mg kg⁻¹), CL (CE 100 mg kg⁻¹), and CH (CE 200 mg kg⁻¹). In the caffeine‐induced insomnia model, caffeine (CAF; 40 mg kg⁻¹) was administered to all groups except the NOR group, followed by the respective oral treatments: NOR (water), CAF (water), PC (GABA), and CH (CE 200 mg kg⁻¹). EEG data were subsequently analyzed to assess sleep parameters.

Analysis of gene and protein expression levels of receptors after 4 weeks of oral CE administration

After 4 weeks of oral CE administration, ICR mice were sacrificed, and brain tissue was extracted. RNA and protein were isolated following the previously described method. qPCR and western blotting were performed to measure the gene expression levels and protein expression levels of receptors.

Statistical analysis

The sample sizes were determined through a power analysis to ensure sufficient statistical power to detect significant differences. A total of seven animals were included per group, across four experimental groups (n = 28 animals in total). Data were analyzed using one‐way ANOVA with the Statistical Package for Social Sciences (SPSS, version 27.0, IBM, Armonk, NY, USA). Tukey's multiple comparison test was used for post hoc group comparisons.

RESULTS

Polyphenols and flavonoids content and radical scavenging activity of CE

The CE obtained using 500 mL L⁻¹ ethanol exhibited a total sugar content of 359.21 g kg⁻¹, along with polyphenol and flavonoid contents of 132.75 and 0.036 g kg⁻¹, respectively. The extract demonstrated significant radical scavenging activity, as reflected by its IC₅₀ values for ABTS and DPPH radicals, which were 1.02 and 0.30 mg mL⁻¹, respectively. Additionally, the FRAP assay, used to measure the reducing capacity of CE, indicated a reducing power of 978.03 mol kg⁻¹ (Table 1).

Table 1.

Quantification of bioactive compounds and antioxidant properties of Chrysanthemum morifolium extract (CE)

	IC50 value		Reducing power FRAP (mol kg−1 of CE)
	ABTS radical	DPPH radical
CE (mg mL−1)	1.02 ± 0.05	0.30 ± 0.0001	978.03 ± 6.08
Ascorbic acid (μg mL−1)	69.75 ± 0.26	22.00 ± 0.71	—
	Polyphenols	Flavonoids	Total sugar
CE (g kg−1)	132.75 ± 3.08	0.036 ± 0.002	359.21 ± 9.76

Note: Values indicate means ± SD for each group.

CE protects HT22 cells from oxidative stress by enhancing cell viability

This experiment evaluated the protective effects of CE against oxidative stress‐induced cell damage in HT22 cells. Initially, CE treatment at concentrations ranging from 0 to 400 μg mL⁻¹ showed no cytotoxicity under normal conditions, confirming its safety (Fig. 1(A)). Under oxidative stress conditions, however, CE treatment significantly improved cell survival compared to the control group (CON) with 200 μg mL⁻¹, increasing the cell survival rate by approximately 1.87‐fold compared to H₂O₂ treatment alone (P < 0.001, Fig. 1(B)).

Figure 1.

Protective effects of Chrysanthemum morifolium extract on HT22 cell viability under normal and oxidative stress conditions. (A) Cell viability under normal conditions after treatment with varying concentrations of C. morifolium extract (0, 25, 50, 100, 200, and 400 μg mL⁻¹). (B) Cell viability under oxidative stress induced by 400 μmol L⁻¹ H₂O₂, with C. morifolium extract treatment (50, 100, 200, and 400 μg mL⁻¹). Values are reported as mean ± SD with three replicates. ^### P < 0.001 versus NOR (normal control); ***P < 0.001, versus CON (oxidative stress control) (ANOVA, Tukey's test).

CE reduces oxidative stress by lowering ROS and MDA levels in HT22 cells

To assess the role of CE in mitigating oxidative stress, intracellular ROS levels and MDA content were measured. In the CON group treated with H₂O₂, ROS and MDA levels were significantly elevated relative to the normal group (NOR) (P < 0.001 and P < 0.01). However, CE treatment at 100 and 200 μg mL⁻¹ effectively reduced ROS production by 25.89% and 37.87%, respectively (P < 0.001, P < 0.001). Similarly, MDA levels were significantly reduced by 37.87% following CE treatment at 200 μg mL⁻¹ (P < 0.01). In addition, when L‐7‐G, a major bioactive compound in CE, was treated at concentrations equivalent to those in 100 μg mL⁻¹ of CE, ROS production was reduced by 26.81% (P < 0.001), which was consistent with the antioxidant effects of CE. These results suggest that CE exerts a potent protective effect against oxidative stress in HT22 cells, potentially due to the L‐7‐G contained in CE.

CE activates antioxidant enzyme expression and Nrf2 signaling pathway in HT22 cells

The effect of CE on the antioxidant defense mechanism in HT22 cells exposed to H₂O₂‐induced oxidative stress was evaluated by analyzing mRNA and protein expression levels of key antioxidant enzymes. The antioxidant enzymes SOD1, GSH‐Px, and CAT form a coordinated enzymatic defense system that neutralizes ROS and prevents cellular oxidative damage. In the H₂O₂‐treated control group (CON), the mRNA expression levels of these antioxidant enzymes were significantly reduced relative to the normal group (NOR) (P < 0.001). However, CE treatment, particularly at the high concentration (CH), markedly restored the mRNA expression levels of SOD1, GSH‐Px, and CAT (all P < 0.001, Fig. 3(A)–(C)).

Figure 3.

Effects of Chrysanthemum morifolium extract (CE) on antioxidant enzyme expression, Nrf2 signaling pathway activation, and associated proteins in HT22 cells exposed to H₂O₂‐induced oxidative stress. (A–C) mRNA expression levels of SOD1 (superoxide dismutase 1), GSH‐Px (glutathione peroxidase), and CAT (catalase) were quantified in the NOR (normal), CON (control, H₂O₂‐treated), CL (CE low dose), and CH (CE high dose) groups. (D–F) Western blot analysis showing relative protein expression levels of Nrf2, Keap1, and HO‐1 normalized to β‐actin. (G) Representative protein bands. Values are reported as mean ± SD with three replicates. ^# P < 0.05, ^## P < 0.01, ^### P < 0.001, compared to NOR; **P < 0.01, ***P < 0.001, compared to CON (ANOVA, Tukey's test).

Treatment with CE at both low (CL) and high (CH) concentrations led to a significant upregulation of Nrf2 and HO‐1 protein expression compared to the H₂O₂‐treated CON group (Fig. 3(D),(F)). In contrast, protein levels of Keap1, a negative regulator of Nrf2, were significantly decreased following CE treatment (P < 0.001). Specifically, Nrf2 expression, which was markedly reduced in the CON group under oxidative stress, was significantly restored upon CE treatment, with the most pronounced effect observed at the high dose (CH) (P < 0.001). Conversely, Keap1 protein levels, which were elevated in the CON group, were notably reduced by CE treatment, indicating activation of the Nrf2 signaling pathway. Additionally, HO‐1 protein expression was significantly upregulated in the CE‐treated groups, particularly at the high concentration (CH) (P < 0.001), further supporting activation of the Nrf2/HO‐1 pathway.

These results collectively suggest that CE, which contains L‐7‐G as the active ingredient, effectively alleviates oxidative stress by restoring the expression of key antioxidant enzymes (SOD1, GSH‐Px, and CAT) and activating the Nrf2/HO‐1 signaling pathway, thereby enhancing the cellular antioxidant defense mechanism.

Effect of CE on sleep duration and sleep quality

Oral administration of CE at doses of 100 or 200 mg kg⁻¹ did not result in significant changes in sleep latency (Fig. 4(A)). However, sleep duration increased significantly, showing a 1.18‐fold (P < 0.01) and 1.41‐fold (P < 0.001) enhancement compared to the NOR group (Fig. 4(B)). L‐7‐G, a key component of CE, administered alone, also showed a similar increase in sleep duration to the CE100 group (CE 100 mg kg⁻¹ administration group). This suggests that L‐7‐G contributes to the sleep‐enhancing effects of CE (P < 0.001).

Figure 4.

Effect of Chrysanthemum morifolium extract (CE) and luteolin‐7‐glucoside (L‐7‐G) on sleep parameters in mice with pentobarbital‐induced sleep. Mice were administered pentobarbital (42 mg kg⁻¹, i.p.) to induce sleep. NOR represents the water‐treated group, and PC represents the γ‐aminobutyric acid‐treated (100 mg kg⁻¹) group. (A) Sleep latency and (B) sleep duration were evaluated following CE administration (100 mg kg⁻¹). (C) Sleep latency and (D) duration were assessed after treatment with L‐7‐G (0.356 mg kg⁻¹). Results are presented as mean ± SEM (n = 7). **P < 0.01, ***P < 0.001, compared to NOR; ns: not significant (ANOVA, Tukey's test).

To assess sleep quality, NREM sleep duration was analyzed through EEG (Fig. 5). GABA, the positive control, reduced wakefulness, increased sleep time, and significantly enhanced δ‐wave activity (P < 0.05). CE administration dose‐dependently decreased wakefulness (P < 0.05 and P < 0.001, respectively) and increased sleep time (P < 0.05 and P < 0.001, respectively). The increase in sleep time by CE was attributed to the increase in δ‐wave activity (all P < 0.001) involved in deep sleep. The representative sleep–wake hypnogram for each group is provided in Supporting Information, Fig. S1.

Figure 5.

Effects of Chrysanthemum morifolium extract (CE) on sleep parameters analyzed via electroencephalogram in SD rats. Sleep‐related parameters were evaluated, including (A) awake time, (B) total sleep time, (C) REM sleep duration, (D) NREM sleep duration, (E) θ‐wave activity, and (F) δ‐wave activity. CE was orally administered at doses of 100 and 200 mg kg⁻¹, with NOR (water‐treated) and PC (100 mg kg⁻¹ γ‐aminobutyric acid‐treated) serving as controls. Results are shown as mean ± SEM (n = 6). *P < 0.05, **P < 0.01, ***P < 0.001, versus NOR (ANOVA with Tukey's test).

CE in the caffeine‐induced insomnia model

In the caffeine‐induced insomnia model (Fig. 6), caffeine administration led to increased wake time (P < 0.001) and decreased total sleep time (P < 0.001), negatively affecting deep sleep, as evidenced by reduced NREM sleep (P < 0.01) and lower theta‐ and delta‐wave activities (P < 0.05 and P < 0.01, respectively). However, both the positive control (PC) and CE reversed these effects, with CE showing the most pronounced results. CE treatment significantly reduced wake time (P < 0.001) and increased total sleep time (P < 0.001), with the highest levels of NREM sleep (P < 0.001) and delta‐wave activity (P < 0.001), highlighting its potential to restore deep sleep and counteract caffeine‐induced sleep disturbances. Furthermore, the sleep architecture patterns visualized in the caffeine‐induced insomnia model are presented in Supporting Information, Fig. S2.

Figure 6.

Effects of Chrysanthemum morifolium extract (CE) on EEG parameters in a caffeine‐induced insomnia model. Sleep‐related parameters were evaluated, including (A) awake time, (B) total sleep time, (C) REM sleep duration, (D) NREM sleep duration, (E) θ‐wave activity, and (F) δ‐wave activity. NOR (water‐treated), CAF (caffeine 40 mg kg⁻¹), PC (γ‐aminobutyric acid 100 mg kg⁻¹ with caffeine), and CE (CE 100 mg kg⁻¹ with caffeine). Results are expressed as mean ± SEM. ^# P < 0.05, ^## P < 0.01, ^### P < 0.001, compared to the NOR group; *P < 0.05, **P < 0.01, ***P < 0.001, compared to CON (ANOVA, Tukey's test).

Effects of CE on gene and protein expression related to sleep regulation

Figure 7 shows the expression levels of genes and proteins associated with sleep regulation, including GABAergic and serotonergic receptors, following 4 weeks of oral CE administration. The expression of Gabrg1 was significantly increased in the PC, CL, and CH groups compared to the NOR group (P < 0.05, P < 0.001, and P < 0.001, respectively), with the highest expression observed in the CH group. Additionally, the mRNA expression levels of Gabrr1 and Gabrr2, subunits of the GABA_B receptor, were elevated in the CE‐treated groups compared to the NOR group. The expression of Htr1a was also increased in the GABA‐ and CE‐treated groups relative to the NOR group. At the protein level, CE significantly increased the protein levels of GABA_A (P < 0.05) and 5‐HT_1A (P < 0.05) receptors, particularly in the CH group, while no significant changes were observed in GABA_B receptor protein level.

Figure 7.

Effects of Chrysanthemum morifolium extract (CE) on GABAergic and serotonergic receptor expression in mice. (A–D) Relative mRNA expression levels of Gabrg1 (γ‐aminobutyric acid type A receptor subunit γ1), Gabbr1 (γ‐aminobutyric acid type B receptor subunit 1), Gabbr2 (γ‐aminobutyric acid type B receptor subunit 2), and Htr1a (5‐hydroxytryptamine receptor 1A) were assessed in the NOR (normal), PC (γ‐aminobutyric acid 100 mg kg⁻¹), CL, and CH (CE 100 and 200 mg kg⁻¹, respectively). (E–G) Protein expression levels of GABA_A, GABA_B, and 5‐HT_1A receptors were quantified and normalized to GAPDH. (H) Representative protein bands. Data are presented as mean ± SEM (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001, compared to NOR (ANOVA, Tukey's test).

Interaction between CE and GABA_A receptors

The increased expression of the GABA_A receptor was confirmed at the protein level. To further investigate the impact of CE on the GABA_A receptor, a binding assay was performed using GABA_A receptor antagonists (Fig. 8). Treatment with CE alone, or CE combined with any of the antagonists (PIX, BIC, or FMZ), did not significantly alter sleep latency compared to the NOR or antagonist‐only groups (Fig. 8(A)). CE treatment significantly increased sleep duration compared to the NOR group (P < 0.001). Treatment with PIX, BIC, or FMZ alone did not significantly affect sleep duration compared to NOR. Co‐treatment with BIC and CE (BIC + CE) also did not significantly reduce the sleep‐enhancing effect of CE. In contrast, co‐treatment with PIX or FMZ alongside CE (PIX + CE and FMZ + CE) significantly reduced the sleep‐enhancing effect of CE, showing a statistically significant difference compared to CE alone (P < 0.001 and P < 0.001, respectively). Based on the results showing a reduction in the efficacy of CE in the presence of GABA_A receptor antagonists, it appears that the sleep‐enhancing effect of CE is mediated through interactions with GABA_A receptor subtypes.

Figure 8.

Sleep parameters following treatment with Chrysanthemum morifolium extract (CE) and antagonist in pentobarbital‐induced sleep model. (A) Sleep latency and (B) sleep duration in each group were measured after pentobarbital administration (42 mg kg⁻¹, i.p.). NOR (water‐treated); CE; PIX (picrotoxin); PIX + CE: combined treatment with PIX and CE; BIC (bicuculline); BIC + CE: combined treatment with BIC and CE; FMZ (flumazenil); FMZ + CE: combined treatment with FMZ and CE. Data are presented as mean ± SEM (n = 7). **P < 0.01, ***P < 0.001, compared to NOR; ^## P < 0.01, ^### P < 0.001, compared to CE. Different letters (a–c) indicate significant differences among groups (P < 0.05); ‘ns’ indicates no significant difference (ANOVA, Tukey's test).

DISCUSSION

Sleep is essential for maintaining overall health, particularly in the brain, where it facilitates the removal of metabolic waste products, including ROS. ³¹ Excessive accumulation of ROS can impair brain function and is closely associated with a decline in sleep quality. The hypothalamus and brainstem, regions critical for regulating the sleep–wake cycle, are particularly vulnerable to oxidative stress caused by ROS. Studies have shown that oxidative stress disrupts the neural circuits involved in sleep regulation, leading to disturbances in sleep duration and quality. Conversely, sleep plays a crucial role in reducing oxidative stress by enhancing antioxidant defense systems and promoting the clearance of ROS via mechanisms such as the glymphatic system, which is active during deep sleep phases, particularly during NREM sleep. ⁴ , ³¹

CE, rich in antioxidants such as L‐7‐G, was demonstrated to effectively scavenge ROS and alleviate oxidative stress in neuronal cells. ³² In this study, CE treatment significantly mitigated oxidative stress in HT22 cells exposed to H₂O₂‐induced damage by reducing intracellular ROS levels and lipid peroxidation markers such as MDA (Figs 1 and 2). Moreover, CE restored the expression of key antioxidant enzymes, including SOD1, GSH‐Px, and CAT, and activated the Nrf2/HO‐1 signaling pathway, underscoring its role in cellular protection against oxidative stress (Fig. 3). These findings align with previous reports suggesting that natural extracts with antioxidant properties can restore redox homeostasis and protect neuronal integrity under conditions of oxidative stress. ³³ , ³⁴

Figure 2.

Effects of Chrysanthemum morifolium extract (CE) on ROS production and lipid peroxidation in HT22 cells under oxidative stress conditions. (A) ROS production in cells treated with 400 μmol L⁻¹ H₂O₂ (CON) and CE at low (CL) and high (CH) concentrations. (B) Malondialdehyde (MDA) levels, an indicator of lipid peroxidation, in cells exposed to 400 μmol L⁻¹ H₂O₂ with and without CE treatment. (C) ROS production in cells treated with luteolin‐7‐glucoside (3.56 and 7.02 μg mL⁻¹) under 400 μmol L⁻¹ H₂O₂‐induced stress. Values are reported as mean ± SD with three replicates. ^## P < 0.01, ^### P < 0.001 versus NOR (normal control); **P < 0.01, ***P < 0.001, versus CON (oxidative stress control) (ANOVA, Tukey's test).

Oxidative stress is also a known contributor to sleep disturbances, including insomnia, through its impact on neurotransmitter systems such as GABAergic and serotonergic pathways. ³⁵ The present study further demonstrated that CE administration increased the expression of GABA_A and 5‐HT_1A receptors, which are known to modulate sleep duration and quality (Fig. 7). Deep sleep, characterized by delta‐wave activity, is particularly sensitive to oxidative stress, as ROS accumulation disrupts the cellular processes required for this restorative phase of sleep. ³⁶ CE's antioxidant activity likely contributes to its observed sleep‐enhancing effects by reducing oxidative damage and supporting the neural circuits involved in sleep regulation.

Recent studies have also suggested that luteolin – one of the bioactive flavonoids present in CE – may exert sleep‐promoting effects through its interaction with adenosine 1 receptor (A1R) and adenosine 2A receptor (A2AR), which are known to regulate sleep–wake homeostasis. ³⁷ Adenosine plays a crucial role in sleep induction, and activation of A1R and A2AR has been linked to increased NREM sleep. ³⁸ Given that CE contains L‐7‐G, a glycosylated form of luteolin, it is possible that a portion of its sleep‐promoting effects may be mediated through adenosine receptor interactions in addition to its modulation of GABA_A receptors. However, further studies are required to determine whether L‐7‐G retains similar adenosine receptor‐binding activity or if it must undergo metabolic conversion to aglycone luteolin to exert these effects.

The interaction between CE and neurotransmitter receptors was further confirmed through the use of GABA_A receptor antagonists, which partially attenuated the sleep‐enhancing effects of CE (Fig. 8). This suggests that CE not only alleviates oxidative stress but also directly modulates sleep‐related neural pathways, possibly through its bioactive components like L‐7‐G (Fig. 4). Similar findings have been reported for other natural extracts, highlighting the potential of antioxidants in addressing sleep disorders associated with oxidative stress. ³⁹ , ⁴⁰

CONCLUSION

In conclusion, CE's dual action of ROS scavenging and modulation of neurotransmitter systems underscores its potential as a therapeutic agent for sleep disorders linked to oxidative stress. By enhancing antioxidant defenses and promoting the recovery of neural circuits disrupted by ROS, CE offers a promising approach to improving sleep quality and duration. However, the variability in the composition of CE extracts remains a limitation, which may influence its therapeutic consistency. Future studies should explore the long‐term effects of CE on sleep regulation and its efficacy in clinical populations with oxidative stress‐related sleep disturbances. Additionally, identifying other biologically active compounds in CE, beyond L‐7‐G, and elucidating their mechanisms could further enhance its potential as a treatment for sleep disorders.

Despite these promising findings, several limitations warrant consideration. First, although GABA_A receptor involvement in CE's sleep‐enhancing effects was supported by partial attenuation with specific antagonists, the incomplete blockade suggests that other mechanisms may contribute. In particular, adenosine receptors (A1R and A2AR) – which have established roles in sleep regulation and are targets of luteolin ⁴¹ – were only briefly mentioned and not directly investigated in this study. Future research should explore CE's potential modulation of adenosine receptor pathways, as this may underlie additional sleep‐promoting effects. Second, the caffeine‐induced insomnia model employed here, while relevant, involves complex pharmacological actions beyond adenosine receptor antagonism, including effects on dopaminergic and other neurotransmitter systems. ⁴² These off‐target actions may confound interpretation of CE's effects in this model. Complementary studies using additional sleep disturbance models and broader neurochemical analyses would strengthen understanding of CE's mechanisms. Finally, the variability in CE composition, particularly the content of luteolin‐7‐glucoside across different extract batches, may affect reproducibility and translational potential. ⁴³ Future studies should include thorough compositional analyses and batch standardization to ensure consistency and facilitate clinical application. Addressing these limitations will provide a clearer framework for future investigations and help optimize CE's development as a potential therapeutic for sleep disorders associated with oxidative stress.

FUNDING INFORMATION

This work was supported by the Technology Innovation Program (20008980), funded by the Ministry of Trade, Industry and Energy (MOTIE, Korea).

AUTHOR CONTRIBUTIONS

Hyeongyeong Kim: writing – original draft, data curation. Seok Hyun Chung: data curation, methodology. Singeun Kim: methodology, investigation. Sooil Ko: software, investigation. Jae‐Chul Jung: methodology, conceptualization. Hyun Jung Lee: writing – review and editing, conceptualization. Hyowon Lee: writing – original draft, writing – review and editing, supervision.

CONFLICT OF INTEREST

The authors have declared that there are no conflicts of interest related to this study.

Supporting information

Table S1. PCR primer sequences.

Table S2. Antibody information for western blot.

Figure S1. Representative hypnograms showing the sleep–wake architecture over 7 h in rats under normal sleep conditions. Sleep stages were classified into wakefulness, rapid eye movement (REM) sleep, and non‐rapid eye movement (NREM). The experimental groups included NOR (water‐treated), PC (positive control; GABA 100 mg kg⁻¹), CE 100 mg kg⁻¹, and CE 200 mg kg⁻¹.

Figure S2. Representative hypnograms illustrating the sleep–wake architecture in a caffeine‐induced sleep disturbance model. Sleep stages were classified into wakefulness, REM, and NREM sleep. The experimental groups included NOR (water), CAF (caffeine 40 mg kg⁻¹), PC (GABA 100 mg kg⁻¹), and CE (CE 200 mg kg⁻¹).

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.