Arctigenin Attenuates Hepatic Stellate Cell Activation via Endoplasmic Reticulum-Associated Degradation (ERAD)-Mediated Restoration of Lipid Homeostasis

Abstract

Arctigenin, a natural lignan from Arctium lappa L., exhibits potent antifibrotic activity, yet its molecular mechanisms remain unclear. Endoplasmic reticulum (ER) stress is known to promote hepatic stellate cell (HSC) activation and liver fibrosis. This study investigates the therapeutic potential of arctigenin in HSC activation through ER stress modulation. Primary rat HSCs were activated (3–7 days) and treated with tunicamycin (ER stress inducer) or 4-PBA (ER stress inhibitor). Arctigenin attenuated ER stress markers (e.g., GRP78) and suppressed the expression of fibrotic marker α-SMA in ER stress-challenged activating (day 3) and activated (day 7) HSCs. Arctigenin restored lipid homeostasis by modulation of both lipogenesis (via Dgat2 and Ppar-γ upregulation) and lipolysis (suppression via ATGL inhibition). ER stress activated ER-associated degradation (ERAD), triggering the formation of small lipid droplets (LD). Arctigenin normalized the ERAD activity, thereby rescuing LD integrity and suppressing HSC activation. Our findings demonstrate that arctigenin mitigates HSC activation by suppressing ER stress and restoring lipid homeostasis via modulating ERAD-mediated lipid dysregulation. As a dietary and medicinal compound, arctigenin emerges as a promising therapeutic candidate for liver fibrosis.

1. Introduction

Liver fibrosis is a dynamic pathological process resulting from chronic liver disease, including metabolic dysfunction-associated steatohepatitis, alcohol-related liver disease, and hepatitis C or B viral infections. Advanced liver fibrosis progresses to cirrhosis, hepatocellular carcinoma, and ultimately liver failure, the leading cause of liver transplantation. The onset and development of liver fibrosis is characterized by the accumulation of excessive extracellular matrix proteins (ECM) in the liver, particularly the deposition of collagen types I and III. Activated hepatic stellate cells (aHSCs) synthesize excessive amounts of the extracellular matrix and contribute to fibrotic scar formation in response to chemokine and cytokine stimulation from inflammatory or injured cells. Therefore, aHSCs are recognized as the primary drivers of liver fibrosis, and inhibition and reversal of HSC activation have been proposed as novel therapeutic strategies for liver fibrosis. ,

The endoplasmic reticulum (ER) is an important compartment for protein biosynthesis, folding, and trafficking in cells, and many proteins are involved in regulating proteostasis and ER homeostasis. When the protein-folding demand exceeds ER capacity or when ER function is compromised, the accumulation of misfolded proteins triggers ER stress. , The Unfolded Protein Response (UPR) is the initial response to ER stress and serves to restore ER homeostasis by increasing the production of proteins related to the protein-folding capacity of the ER. When the UPR is unable to restore protein-folding capacity, persistent ER stress triggers cellular damage and apoptotic cell death in various cell types, initiating pathogenic cascades implicated in multiple human diseases. , Accumulating evidence has demonstrated that ER stress as an inducer of HSCs activation drives hepatic fibrogenesis. ,, Chronic ER stress induces paradoxical cellular responses in HSCs: ER stress induces apoptosis but also promotes HSC transdifferentiation into fibrogenic myofibroblast-like cells via the IRE1α-dependent p38/MAPK and PERK-associated UPR signaling pathways. ,, This ER stress-mediated phenotype facilitates excessive ECM deposition and creates a pro-fibrotic microenvironment that perpetuates hepatic scarring. Clinical analysis reveals coordinated upregulation of UPR markers, e.g., Grp78 and Chop, and HSC activation markers, e.g., α-SMA, in fibrotic human livers, suggesting the spatial correlation between ER stress hotspots and fibrotic regions. Importantly, pharmacological induction of ER stress accelerates HSC activation in both in vitro models and in vivo murine models. , These studies indicate that the modulation of ER stress is a valid therapeutic strategy against liver fibrosis.

ER-associated degradation (ERAD) is a conserved pathway that eliminates misfolded ER proteins via the ubiquitin-proteasome system, preserving ER homeostasis. ERAD operates by recognizing misfolded substrates, translocating them across the ER lipid bilayer into the cytoplasm, and facilitating their degradation via membrane-associated E2/E3 ubiquitin ligases such as SYVN1 and GP78. Emerging evidence implicates ERAD in the pathogenesis of liver fibrosis, where the E3 ligase SYVN1 has been identified as essential for HSC activation. SYVN1 suppression attenuates the progression of fibrosis, indicating that ERAD is a pivotal regulator of this process. ERAD also plays a central role in lipid droplet (LD) homeostasis, since the ER mediates LD biogenesis. LDs are ubiquitous organelles that store lipids for energy production or membrane synthesis. It also acts as a hub for various metabolic processes. , ERAD functions as a rate-limiting factor in LD formation, since it regulates conversion of diacylglycerol (DAG) to triacylglycerol (TAG) through the ERAD sensor UbxD8 and its membrane partner Ubx2, thereby fine-tuning neutral LD biosynthesis. , Inhibition of ERAD ubiquitin ligases increases TAG synthesis enzymes GPAT3, MOGAT2, and DGAT2, resulting in increased fatty acid re-esterification and tissue triacylglycerol synthesis, leading to increased LD formation and accumulation. In addition, the ERAD adaptor facilitates lipase maturation and secretion, promoting LD mobilization. As a major site of lipid storage in the liver, quiescent (non-activated) HSCs store mainly retinoids and neutral LDs. Upon activation, the quiescent HSCs lose their LDs as a marker of ER stress increase, , suggesting that the interplay among ER stress, ERAD, and LD formation are significant in HSC activation. Consequently, modulation of this interplay may restore LD content in HSCs and reverse their pro-fibrotic activation.

Arctigenin is a bioactive lignan extracted from Fructus arctii, the dried, ripe fruit of Arctium lappa L., which is a medicine and food homologue (MFH) present in functional foods due to its dual dietary and therapeutic potential. Emerging evidence highlights the multifaceted pharmacological activities of arctigenin, particularly in ameliorating ER stress through AMPK activation and suppression of the unfolded protein response (UPR). − Despite these findings, the therapeutic properties of arctigenin on ER stress-driven liver fibrogenesis remain unexplored, with systematic investigations lacking in both in vitro and in vivo models. In the present study, we used an in vitro HSC activation model to investigate the antifibrotic mechanisms of arctigenin with a focus on ER stress modulation and its interplay with the ERAD system. Our findings reveal that arctigenin suppresses HSC activation by restoring ERAD-mediated lipid homeostasis. This discovery not only elucidates a novel molecular mechanism underlying the antifibrotic activity of arctigenin but also establishes a theoretical foundation for its dual applicability as both a functional food ingredient and a therapeutic agent for liver fibrosis.

2. Materials and Methods

2.1. Primary Rat Hepatic Stellate Cell Isolation and Culture

Adult male Wistar rats (250–300 g) were purchased from Charles River Laboratories Inc. (Wilmington, MA) and housed in the central animal facility of the University Medical Center Groningen. All animals had ad libitum access to food and water. The animal procedures in this study were approved by the Animal Welfare Body of the University of Groningen under an ethical license No. 2115139-01-001. Primary hepatic stellate cells (HSCs) were isolated from rats as previously described. Briefly, livers were perfused in situ with Pronase (Merck, Amsterdam, The Netherlands) followed by collagenase P (Roche, Almere, The Netherlands). Dispersed cell suspensions were layered on a 13% Nycodenz density gradient (Axis-Shield POC, Oslo, Norway), and HSCs were subsequently separated by density gradient centrifugation. Freshly isolated HSCs were cultured in Iscove’s Modified Dulbecco’s Medium (IMDM; Thermo Fisher Scientific, Breda, The Netherlands) supplemented with 20% fetal bovine serum, 1% MEM nonessential amino acids, 1% sodium pyruvate (all from Thermo Fisher Scientific), and antibiotics: 50 μg/mL gentamycin (Thermo Fisher Scientific), 100 units/mL streptomycin (Lonza, Vervier, Belgium), 100 units/mL penicillin (Lonza), and 250 ng/Ml Fungizone (Lonza). HSCs were maintained at 37 °C in a humidified atmosphere containing 5% CO₂. HSCs isolated on day 1 exhibit a quiescent phenotype characterized by abundant retinoid-containing lipid droplets (LDs; Figure S4a). Upon culture on plastic surfaces, spontaneous progressive activation occurs: by day 3, HSCs transitioned to an activating stage, i.e., intermediate activation state, retaining partial LDs but displaying reduced retinoid storage (Figure S4a). By day 7, cells reached full activation, marked by complete LD depletion and concurrent declines in neutral lipids (e.g., triglycerides). This activation cascade correlated with elevated expression of α-smooth muscle actin (α-SMA, indicative of myofibroblastic transdifferentiation) and collagen type I α1 (Col1α1), a hallmark of extracellular matrix synthesis. Additionally, progressive increased proliferation is observed during this activation process.

2.2. Chemicals

Arctigenin (purity >98%) was obtained from Dilger Medicine (Nanjing, China), Tunicamycin from Streptomyces sp. and sodium 4-phenylbutyrate (4-PBA) were from Sigma-Aldrich (Zwijndrecht, The Netherlands), and Eeyarestatin-I (Eer I) was purchased from Santa Cruz (Heidelberg, Germany).

2.3. Experimental Design

This study investigated the antifibrotic effects of arctigenin on HSC activation, focusing on its protective effect against ER stress-driven lipid dysregulation. Freshly isolated primary HSCs were seeded at 1 × 10⁶ cells/mL and cultured for 3 days (activating phase) or 7 days (fully activated state), as outlined in Figure S1. In phase 1, activating and activated HSCs were treated with ATG to evaluate the dose-dependent suppression of proliferation, activation markers (α-SMA, Col1a1), and ER stress mediators (Grp78). Phase 2 employed ER stress modulation, activating HSCs (day 3) were exposed to tunicamycin (2 μg/mL, 12 h) to induce ER stress, while activated HSCs (day 7) received 4-PBA (3 mM, 12 h) to alleviate ER stress. The ability of arctigenin to ameliorate ER stress-driven HSC activation was validated by assessing the lipid droplet dynamics and the analysis of lipid metabolism pathways, confirming its role in restoring lipid homeostasis. Phase 3 mechanistically linked ERAD to lipid regulation by inhibiting ERAD with Eeyarestatin-I (Eer I), demonstrating that arctigenin attenuates HSC activation through the modulation of ERAD activity.

2.4. Cell Toxicity Assay

Necrotic cells were determined using SYTOX Green nucleic acid staining (Thermo Fisher Scientific) at a dilution of 1:40,000 in Hanks’ Balanced Salt Solution for 15 min at 37 °C under CO₂-free conditions. Hydrogen peroxide (1 mmol/L) was used as a positive control for necrosis. Fluorescent nuclei were visualized using a Leica microscope (Amsterdam, The Netherlands) equipped with an excitation filter 450–490 nm bandpass (Leica #11513867, Chroma ET470/40x equivalent).

Quantification of cytotoxicity was evaluated by a WST-1 assay (Roche). About 30,000 HSCs were seeded per well in 96-well plates and treated with arctigenin for 48 h. 10 μL of WST-1 solution per 100 μL of culture medium was added to each well and incubated for 2 h at 37 °C. The absorbance was quantified using a microreader (Bio-Tek, Winooski, VT).

2.5. Cell Proliferation Measurement

HSC proliferation was measured using a Real-Time xCELLigence system (RTCA DP; ACEA Biosciences Inc., Santa Clara, CA). Around 3000–5000 HSCs were seeded per well in a 16-well E-plate (Agilent, Santa Clara, CA) and treated with arctigenin. Normalized cell index was determined by measuring real-time cellular impedance.

2.6. Detection of Intracellular Neutral Lipid Droplets

HSCs were seeded on coverslips in 12-well plates and treated as described. Neutral lipid droplets (LDs) inside HSCs were visualized using BODIPY-LD staining as previously reported. Briefly, cells were fixed with 3.7% paraformaldehyde solution (Merck Darmstadt, Germany) for 15 min, followed by washes with phosphate-buffered saline (PBS, Thermo Fisher Scientific). LDs were fluorescently labeled with BODIPY 493/503 dye (1:2500 dilution in PBS, Thermo Fisher Scientific) for 15 min. Nuclei were subsequently counterstained with 4′,6-diamidino-2-phenylindole (DAPI, Sigma-Aldrich) for 10 min. Cell imaging was performed using a fluorescent microscope (Leica, Amsterdam, The Netherlands).

2.7. Detection of Vitamin A (Retinoids) Autofluorescence

HSCs were cultured in 12-well plates, and intracellular vitamin A autofluorescence was determined using a fluorescence microscope (Leica) with an excitation filter of 365 nm.

2.8. Live-Cell Imaging of Endoplasmic Reticulum (ER)

HSCs were cultured on coverslips and treated with the ER stress inducer tunicamycin and the ER stress-relieving compounds 4-PBA and arctigenin. After the treatment, HSCs were washed 3 times with PBS (Thermo Fisher Scientific) to remove cell debris and then incubated with 1 μmol/L ER-tracker (Thermo Fisher Scientific) for 30 min in the dark. After incubation, HSCs were stained with Hoechst (Thermo Fisher Scientific) for 10 min. Fluorescent ER was detected using a microscope (Leica) at an excitation wavelength of 587 nm.

2.9. Quantitative Real-Time Reverse Transcription Polymerase Chain Reaction

HSC RNA was extracted using TriReagent (Sigma-Aldrich) according to the manufacturer’s protocol. RNA quality and quantity were measured by using a Nano-Drop 2000c spectrophotometer (Thermo Fisher Scientific). 1.5 μg portion of RNA was used for reverse transcription using M-MLV polymerase (Thermo Fisher Scientific). Quantitative real-time polymerase chain reaction was conducted using the QuantStudio 3 system (Thermo Fisher Scientific). The mRNA levels of Plin2, Pnpla2, Cyp26a1, Syvn1, Dnajc10, Dnajb9, and Herpud1 were quantified using commercial tests (Thermo Fisher Scientific), and other genes were quantified using the Taqman primers and probes listed in Table S1. All samples were analyzed in duplicate. Relative mRNA expression was calculated using the 2^–ΔΔCt method, with 36b4 as the normalizing gene.

2.10. Western Blot Analysis

After treatment, HSCs were collected using lysis buffer (HEPES 25 mmol/L, KAc 150 mmol/L, EDTA 2 mmol/L, NP-40 0.1%, NaF 10 nmol/L, PMSF 50 mmol/L, aprotinin 1 μg/μL, pepstatin 1 μg/μL, leupeptin 1 μg/μL, and DTT 1 mmol/L). Total protein in HSCs was quantified using the Bio-Rad protein assay (Bio-Rad, Hercules, CA). About 20–30 μg of protein was separated by SDS-PAGE gels and transferred to nitrocellulose membranes (Amersham, Piscataway, NJ) using a Trans-Blot Turbo Blotting system (Bio-Rad). Protein was detected using the following primary antibodies: anti-Actin α-Smooth Muscle (A5228, Sigma-Aldrich, 1:2000), anti-COL1A1 (1310-01, Southern Biotech, 1:1000, Birmingham, AL), anti-BiP (also called anti-Grp78, 3183, Cell Signaling Technology, 1:1000, Leiden, The Netherlands), anti-P21 (ab109199, Abcam, 1:1000), anti-phospho-HSL (4126, Cell Signaling Technology, 1:1000), anti-HSL (4107, Cell Signaling Technology, 1:1000), anti-ATGL (2138, Cell Signaling Technology, 1:1000), anti-phospho-eIF2α (9721, Cell Signaling Technology, 1:1000), anti-eIF2α (9722, Cell Signaling Technology, 1:1000), and anti-Tubulin (T9026, Sigma-Aldrich, 1:1000). Protein signals were detected by using a ChemiDoc MP Imaging system (Bio-Rad).

2.11. Statistical Analysis

All data are presented as means ± standard deviation (means ± SD). Every experiment was repeated at least three times using HSCs from different isolations. Statistical analysis was performed with GraphPad Prism (V..8.0.1, GraphPad Software, San Diego, CA). One-way analysis of variance (ANOVA) test followed by Tukey’s multiple comparison tests was used to evaluate group differences. P values < 0.05 were considered statistically significant.

3. Results and Discussion

3.1. Arctigenin Inhibits Hepatic Stellate Cell Activation In Vitro

Primary rat hepatic stellate cells (rHSCs) were activated via a standard plastic adherence culture, a validated experimental model for recapitulating spontaneous in vitro activation. To assess the toxicity of arctigenin, cells at distinct activation stages, early activating (day 3) and fully activated myofibroblasts (day 7, denoted as aHSCs in figures), were exposed to arctigenin (1–50 μM) for 48 h (experimental design detailed in Figure S1a). This concentration range was selected based on prior evidence of its bioactivity in HSCs, spanning concentrations with demonstrated biological activity.

Membrane integrity, a definitive criterion for distinguishing viable and nonviable cells in vitro, was evaluated using SYTOX Green. This fluorescent dye selectively binds nucleic acids upon loss of membrane integrity. Hydrogen peroxide (H₂O₂) was used as a positive control for the loss of membrane permeability (Figure S2a,b). Arctigenin treatment (1–50 μM) induced no detectable SYTOX Green uptake in HSCs (Figure S2a,b). Cellular viability, reflecting metabolic competence, was also quantified via the WST-1 assay. As shown in Figure S2c,d, the viability of activating and activated HSCs treated with arctigenin (1–50 μM) remained comparable to that of untreated controls (∼100%), confirming the absence of cytotoxicity across all tested concentrations. These data collectively demonstrate that arctigenin exhibits no cytotoxic effects on HSCs, irrespective of their activation status (activating or fully activated).

Under normal conditions, HSCs remain in a quiescent, non-proliferative state; however, upon activation, they become highly proliferative. This expansion of the activated HSC population drives ECM deposition, thereby accelerating fibrosis progression. Consequently, the pharmacological inhibition of HSC proliferation emerges as a promising therapeutic strategy for hepatic fibrosis management. In the present study, we employed the xCELLigence real-time cell analysis system to quantitatively assess proliferative dynamics through continuous monitoring of impedance variations in specialized microelectronic plates (experimental design illustrated in Figure S1b). This label-free methodology enables precise measurement of cellular spreading and the number of HSCs, offering distinct advantages over conventional end point quantification approaches. Our experimental design captured proliferation profiles during critical transitional phases: early activating and fully activated states (Figure S1b), with data acquisition completed within a 48 h observation window. Notably, administration of arctigenin significantly attenuated the real-time cell index in early activating HSCs compared to the control group (Figure a). This inhibitory effect persisted in fully activated HSCs, where increasing arctigenin concentrations produced a progressive reduction in proliferative indices (Figure b). These findings collectively demonstrate the efficacy of arctigenin in inhibiting HSC proliferation regardless of their activation state, suggesting its potential for therapeutic intervention at multiple stages of fibrotic progression.

1.

Arctigenin inhibits hepatic stellate cell proliferation in vitro. Activating (Day 3) hepatic stellate cells (HSCs) and activated (Day 7) HSCs (aHSCs) were treated with arctigenin (10, 20, and 50 μM) for 48 h. (a, b) Proliferation of activating and activated HSCs was quantified using the xCELLigence real-time cell index system. (c) Quantitative real-time PCR (qPCR) analysis of mRNA expression of the cell cycle inhibitor p21^Cip1 in activating HSCs following 48 h arctigenin treatment. (d) Western blot analysis of p21^Cip1 protein expression in activating HSCs treated with arctigenin for 48 h. Protein band intensities were quantified using ImageJ software. (e) qPCR analysis of p21^Cip1 mRNA expression in activated HSCs after 48 h of arctigenin treatment. (f) Western blot analysis of p21^Cip1 protein expression in activated HSCs treated with arctigenin for 48 h. Protein band intensities were quantified using ImageJ software. (g) qPCR analysis of p27^Kip1 mRNA expression in activating and activated HSCs following 48 h arctigenin treatment. HSC, Hepatic stellate cell; D3 HSC, activating HSCs; aHSC, D7 (fully) activated HSCs; ATG, arctigenin. Data are presented as mean ± standard deviation (mean ± SD, n ≥ 3 per group). Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. Significance: ns: not significant, P > 0.05; *: P < 0.05; **: P < 0.01.

Pharmacological suppression of cellular proliferation may result from compromised cellular functionality. Notably, the absence of cytotoxicity observed in arctigenin-treated HSCs demonstrates that its antiproliferative effects are not mediated through cellular dysfunction or cytotoxic damage (Figure S2). A critical regulatory mechanism governing cell cycle progression involves the activation of cyclin-dependent kinase inhibitors (CKIs), which modulate cyclin protein-CDK complexes, thereby leading to cell cycle arrest. Contrary to LX2 cells, where arctigenin induced P27 ^Kip1-mediated cell cycle arrest, our study revealed that 50 μM arctigenin upregulated both mRNA and protein levels of P21^Cip1 in activating and fully activated HSC populations, whereas lower concentrations failed to elicit this response (Figure c–f). Interestingly, P27^Kip1 expression remained unaltered by 50 μM arctigenin in activating HSCs but exhibited a significant elevation in activated HSCs (Figure g). The divergence between our results and prior reports could be attributed to phenotypic heterogeneity in HSC activation states and/or interspecies variations between rodent and human cell models, and the fact that LX-2 is an immortalized cell line. This study further underscores the dynamic regulation of cell cycle progression during HSC activation, wherein distinct CDK inhibitors may govern the proliferative capacity at different differentiation stages. Such stage-specific regulatory mechanisms warrant consideration when developing targeted antifibrotic therapies.

Quiescent HSCs undergo a dramatic phenotypic transition into myofibroblasts in response to hepatic injury. This phenotypic alteration is identified by an elevated expression of α-smooth muscle actin (α-SMA), a key biomarker associated with HSC activation, fibrogenic progression, and fibrosis severity in MASLD patients. Our findings demonstrate that 50 μM arctigenin significantly attenuated both mRNA and protein expression of α-SMA in activating and activated HSCs compared to their corresponding controls (Figure a–d), whereas lower concentrations (10–20 μM) exhibited no significant effects (Figure a–d). The antifibrotic efficacy of arctigenin was corroborated by parallel reductions in collagen type I α 1 (Col1a1) expression, a predominant ECM component overproduced by activated HSCs during fibrogenesis. In activating HSCs, arctigenin induced dose-dependent suppression of Col1a1 mRNA and protein levels across the 10–50 μM concentration range (Figure a,b). In fully activated HSCs, significant Col1a1 inhibition required higher therapeutic thresholds, with both 20 and 50 μM concentrations achieving inhibitory efficacy (Figure c,d). The antifibrotic potential of arctigenin was further substantiated through immunofluorescence analyses, revealing marked attenuation of α-SMA abundance specifically at 50 μM arctigenin in activating and activated HSCs and progressive Col1a1 reduction with increasing arctigenin concentrations (Figure e,f). Collectively, these observations demonstrate that arctigenin treatment displays antifibrotic effects by blocking HSC proliferation and suppressing HSC activation markers (α-SMA, Col1a1). The antiproliferative effects of arctigenin during HSC activation may operate through mechanisms dependent on canonical Cip/Kip-mediated cell cycle arrest, given the upregulation of specific CKIs.

2.

Arctigenin inhibits hepatic stellate cell activation in vitro. Activating (Day 3) HSCs and activated (Day 7) HSCs (aHSCs) were treated with arctigenin (10, 20, and 50 μM) for 48 h. (a) Quantitative real-time PCR (qPCR) analysis of mRNA expression for the HSC activation marker α-smooth muscle actin (α-SMA) and extracellular matrix (ECM) molecule collagen type I α 1 (Col1α1) in activating HSCs treated with arctigenin for 48 h. (b) Western blot analysis of α-SMA and Col1α1 protein expression in activating HSCs treated with arctigenin for 48 h. Protein band intensities were quantified using ImageJ software. (c) qPCR analysis of α-SMA and Col1α1 mRNA expression in avtivated HSCs treated with arctigenin for 48 h. (d) Western blot analysis of α-SMA and Col1α1 protein expression in activated HSCs treated with arctigenin for 48 h. Protein band intensities were quantified using ImageJ software. (e) Immunofluorescence analysis of α-SMA and Col1α1 expression in activating HSCs treated with arctigenin for 48 h. Scale bar, 100 μm. (f) Immunofluorescence analysis of α-SMA and Col1α1 expression in activated HSCs treated with arctigenin for 48 h. Scale bar: 100 μm. HSC, Hepatic stellate cell; D3 HSC, activating HSCs; aHSC, D7 (fully) activated HSCs; ATG, arctigenin. Data are presented as mean ± standard deviation (mean ± SD, n = 3 per group). Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. Significance: ns: not significant, P > 0.05; *: P < 0.05; **: P < 0.01.

3.2. Arctigenin Suppresses ER Stress-Induced Hepatic Stellate Cell Activation

ER stress is known to promote HSC activation and increase the expression of UPR markers in fibrotic liver. , Therefore, alleviation of ER stress using functional food-derived bioactive molecules, particularly those modulating the UPR, may constitute an innovative therapeutic strategy for hepatic fibrosis intervention. To investigate the therapeutic efficacy of arctigenin against ER stress during HSC activation, we first quantified Grp78 levels, which are canonical UPR biomarkers across progressive HSC activation phases. As illustrated in Figure a,b, arctigenin treatment at 20 and 50 μM concentrations significantly reduced Grp78 mRNA and protein expression in activating HSCs, while 50 μM arctigenin effectively suppressed Grp78 in fully activated HSCs. Live-cell imaging using ER-Tracker dye revealed dose-dependent attenuation of ER fluorescence intensity in both activation stages following arctigenin intervention (Figure c,d), suggesting that arctigenin mediates the suppression of ER expansion. We further verified this effect in pharmacological ER stress-exposed HSC models: Day 3 HSCs (activating phase) were challenged with tunicamycin to exacerbate ER stress. Day 7 HSCs (completely activated) received 4-PBA to ameliorate ER stress (Figure S1c,d). The UPR involves three ER membrane sensors: IRE1 (inositol-requiring transmembrane kinase/endoribonuclease), PERK (double-stranded RNA-dependent protein kinase-like eukaryotic initiation factor 2a [eLF2a] kinase), and ATF6 (activating transcription factor 6). Under physiological conditions, these sensors remain inactive through binding to the ER chaperone Grp78. ER stress triggers their activation through Grp78 dissociation, initiating the UPR cascade signaling. IRE1 leads to the activation of downstream Xbp1s and Chop, PERK facilitates the expression of Atf4 and phosphorylation of eIF2α, whereas dissociated ATF6 translocates from the ER to the Golgi complex to be activated and then acts as a transcription factor after splicing , As shown in Figure S3a,b, tunicamycin significantly upregulated Grp78 expression (transcriptional and translational) and key UPR mediators (Xbp11s, Atf4, Chop) in activating HSCs. The ER stress activation was also demonstrated by enhanced phosphorylation of the UPR-related regulator eIF2α (p-eiF2α) and ER expansion in activating HSCs (Figure S3b). Arctigenin treatment normalized the increased level of Grp78, and significantly attenuated UPR signaling through downregulation of Xbp1s, Atf4, and Chop expression (Figure S3a,b). This protective effect was further evidenced by the inhibition of p-eIF2α and mitigation of tunicamycin-induced ER expansion (Figure S3b). In activated HSCs, the ER stress reliever 4-PBA mimicked and increased the effect of arctigenin. 4-PBA alone mimicked arctigenin-mediated suppression of Grp78, Chop, and eIF2α phosphorylation, whereas their combined administration was even more effective in attenuating ER stress (Figure S3c,d). Both 4-PBA alone and its combination with arctigenin effectively restored the ER structure (Figure S3d). These findings collectively demonstrated the capacity of arctigenin to restore ER homeostasis by attenuating the UPR and restoring ER structure.

3.

Arctigenin decreases ER stress initiation in hepatic stellate cells. Activating (Day 3) HSCs and activated (Day 7) HSCs (aHSCs) were treated with arctigenin (10, 20, and 50 μM) for 48 h. (a) Quantitative real-time PCR (qPCR) and Western blot analysis of mRNA and protein expression of the ER stress chaperone Grp78 in activating HSCs treated with arctigenin for 48 h. Protein intensity was analyzed by ImageJ software. (b) qPCR and Western blot analysis of Grp78 mRNA and protein expression in activated HSCs treated with arctigenin for 48 h, respectively. Protein intensity was analyzed by ImageJ software. (c) Live-cell ER-Tracker staining to assess ER abundance in activating HSCs treated with arctigenin for 48 h. Scale bar: 200 μm. (d) Live-cell ER-Tracker staining to evaluate ER abundance in activated HSCs treated with arctigenin for 48 h. Scale bar: 200 μm. HSC, Hepatic stellate cell; D3 HSC, activating HSCs; aHSC, D7 (fully) activated HSCs; ATG, arctigenin. Data are presented as mean ± standard deviation (mean ± SD, n = 3 per group). Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. Significance: ns: not significant, P > 0.05; *: P < 0.05; **: P < 0.01.

Further analysis uncovered the therapeutic potential of arctigenin in alleviating HSC activation via ER stress modulation: tunicamycin significantly increased the expression of HSC activation marker α-SMA in activating HSCs (Figure a–c), demonstrating that ER stress is a driver of HSC transdifferentiation and activation. Despite HSC activation, Col1a1 expression was significantly downregulated at both transcriptional and translational levels rather than upregulated following tunicamycin challenge (Figure a,b), which is consistent with a previous study demonstrating ER stress-mediated suppression of collagen synthesis. This phenomenon appears independent of transcriptional regulation of collagen biosynthetic enzymes (e.g., P4ha1 and P4ha2) but may be dependent on ER dysfunction-induced defects in post-translational processing of Col1α1, thereby resulting in the accumulation of misfolded Col1α1 in the ER lumen and depletion of secreted mature Col1α1. Both 4-PBA and arctigenin and their combination reduced both α-SMA and Col1a1 expression in activated HSCs (Figure d–f). Collectively, these data demonstrate that arctigenin suppresses HSC activation via alleviation of ER stress and restoration of ER homeostasis. Moreover, arctigenin significantly inhibits collagen biosynthesis, suggesting its potential to attenuate ECM remodeling in hepatic fibrosis. However, whether arctigenin directly targets collagen synthesis/secretion or acts indirectly by modulating ER expansion remains unclear. Further studies are required to clarify these mechanisms and the molecular targets involved.

4.

Arctigenin suppresses ER stress-induced hepatic stellate cell activation. Activating and activated HSCs were pretreated with arctigenin (50 μM, 12 h), followed by treatment with 2 μg/mL tunicamycin or 3 mM sodium 4-phenylbutyrate (4-PBA) for 12 h, respectively. (a) Quantitative real-time PCR (qPCR) analysis of mRNA expression of HSC activation marker α-smooth muscle actin (α-SMA) and extracellular matrix (ECM) molecule collagen type I α 1 (Col1α1) in activating HSCs. (b) Western blot analysis of α-SMA and Col1α1 protein expression in activating HSCs. Protein intensity was analyzed by ImageJ software. (c) Immunofluorescence analysis of α-SMA in activating HSCs. Fluorescence intensity was analyzed by ImageJ software. Scale bar: 200 μm. (d) qPCR analysis of α-SMA and Col1α1 mRNA expression in activated HSCs. (e) Western blot analysis of α-SMA and Col1α1 protein expression in activated HSCs. Protein intensity was analyzed by ImageJ software. (f) Immunofluorescence analysis of α-SMA in activated HSCs. Fluorescence Intensity was analyzed by ImageJ software. Scale bar: 200 μm. HSC, Hepatic stellate cell; D3 HSC, activating HSCs; aHSC, D7 (fully) activated HSCs; ATG, arctigenin; Tun, tunicamycin; 4-PBA, sodium 4-phenylbutyrate. Data are presented as mean ± standard deviation (mean ± SD, n ≥ 3 per group). Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. Significance: ns: not significant, P > 0.05; *: P < 0.05; **: P < 0.01; ***: P < 0.001.

3.3. Arctigenin Increases Neutral Rather Than Retinoid-Containing Lipid Droplets in ER Stress-Exposed Hepatic Stellate Cells

Lipid droplets (LDs), a characteristic feature of quiescent HSCs, serve as dynamic pools for retinoids and non-retinoid lipids. The loss of LDs and their lipid cargo is not only a hallmark of early HSC activation but is also hypothesized to supply bioenergetic substrates necessary for HSC activation. , Consequently, maintaining LD integrity may represent a valid strategy to maintain HSC quiescence. ER dynamics, acting as a master regulator of LD homeostasis, orchestrates LD biogenesis: LD formation initiates within the ER bilayer and is dynamically controlled by lipogenic enzymes embedded in the ER membrane. ,, Moreover, ER stress pathways critically regulate LD turnover, as indicated by studies demonstrating that UPR-driven autophagy promotes LD degradation. , These findings position LDs as a central mechanism linking ER stress to HSC activation, although the precise regulatory crosstalk remains unresolved.

Since arctigenin has been shown to mitigate ER stress in HSCs (Figure S3), we hypothesized that arctigenin could reverse ER stress-mediated LD loss during HSC activation (experimental design shown in Figure S1c,d). To visualize retinoid-enriched LDs, we monitored alterations in vitamin A, the primary retinoid derivative, using its intrinsic fluorescence as a proxy for retinoid storage. Consistent with LD depletion during activation, in vitro HSC cultures exhibited a progressive decline in vitamin A autofluorescence (Figure S4a), reflecting retinoid loss during in vitro HSC activation. Arctigenin treatment significantly attenuated this reduction, preserving the LD-associated retinoid content. Retinoid homeostasis in HSCs is governed by a balance between storage (via re-esterification of retinol) and mobilization (via hydrolysis of retinyl esters). Lrat, the key enzyme catalyzing retinol esterification, was upregulated by arctigenin during HSC activation (Figure S4b), suggesting that arctigenin enhances re-esterification to preserve retinoid pools. We next investigated whether arctigenin restores retinoid-filled LDs by alleviating the ER stress. As shown in Figure a, tunicamycin triggered rapid retinoid loss in activating HSCs. However, arctigenin failed to rescue this depletion (Figure a). Transcriptional profiling revealed that tunicamycin significantly upregulated Pnpla3 and Cyp26a1, which drive retinol hydrolysis and metabolism, , while suppressing Lrat expression (Figure b). Although arctigenin reduced Pnpla3 expression, it did not affect Lrat or Cyp26a1 levels in Tunicamycin-treated HSCs (Figure b). Collectively, these findings suggest that ER stress induces retinoid loss in HSCs via suppressed re-esterification (Lrat downregulation) and enhanced catabolism (Pnpla3/Cyp26a1 upregulation). Importantly, arctigenin attenuates ER stress-driven HSC activation through mechanisms independent of retinoid storage.

5.

Arctigenin increases neutral but not retinoid lipid droplets in ER stress-exposed hepatic stellate cells. (a) Activating HSCs were pretreated with arctigenin (50 μM, 12 h), followed by treatment with 2 μg/mL tunicamycin for varying time intervals. Retinoid content within HSCs was visualized via intrinsic autofluorescence. Scale bar: 200 μm. (b–d) Activating HSCs were pretreated with arctigenin (50 μM, 12 h), followed by treatment with 2 μg/mL tunicamycin for 12 h. (b) Quantitative real-time PCR (qPCR) analysis of mRNA expression for genes involved in retinoid metabolism in activating HSCs. (c) BODIPY-LD staining of neutral lipid droplets (LDs) in activating HSCs. Scale bar: 25 μm. (d) Quantification of LD intensity and size in activating HSCs using ImageJ software. (e) and (f) Activated HSCs were pretreated with arctigenin (50 μM, 12 h), followed by treatment with 3 mM 4-phenylbutyric acid (4-PBA) for 12 h. (e) BODIPY-LD staining of neutral LDs in activated HSCs. Scale bar: 25 μm. (f) Quantification of LD intensity and size in activated HSCs using ImageJ software. HSC, Hepatic stellate cell; D3 HSC, activating HSCs; aHSC, D7 (fully) activated HSCs; ATG, arctigenin; Tun, tunicamycin; 4-PBA, sodium 4-phenylbutyrate. Data are presented as mean ± standard deviation (mean ± SD, n ≥ 3 per group). Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. Significance: ns: not significant, P > 0.05; *: P < 0.05; **: P < 0.01; ***: P < 0.001.

Neutral lipids, including triglycerides, cholesterol esters, non-esterified fatty acids, and diverse phospholipid species, constitute the predominant non-retinoid lipid components of LDs in HSCs. A key feature of HSC activation is the selective release of neutral lipids, particularly TGs, from LDs. Notably, pharmacological or genetic inhibition of lysosomal acid lipase LAL, a key enzyme driving triglyceride catabolism, attenuates HSC activation, , underscoring the critical role of neutral lipid mobilization in driving fibrogenic signaling. To determine whether arctigenin modulates neutral lipid dynamics during ER stress, we visualized intracellular neutral LDs by using BODIPY staining. Interestingly, arctigenin supplementation not only restored neutral LD size but also increased fluorescence intensity in tunicamycin-treated activating HSCs (Figure c,d). Furthermore, attenuation of ER stress using 4-PBA, either alone or in combination with arctigenin, markedly increased both the intensity and the size of neutral LDs in activated HSCs (Figure e,f). These findings indicate that ER stress serves as a critical driver of neutral LD mobilization during HSC activation. Furthermore, arctigenin preserves neutral LD integrity by mitigating ER stress, which may suppress lipid-driven pro-fibrotic signaling.

It is critical to emphasize that our findings on LD dynamics under ER stress in HSCs differ from reports describing ER stress-induced lipid accumulation in hepatocytes. In hepatocytes, ER stress, even in the absence of lipid overload, promotes intracellular lipid retention through suppression of apolipoprotein ApoB100 expression and impaired lipoprotein secretion, resulting in hepatocyte steatosis. Conversely, HSCs, which inherently function as lipid reservoirs under physiological conditions, exhibit an ER stress-driven LD loss. Mechanistically, HSC activation downregulates apolipoprotein ApoE, , a critical mediator of lipid secretion, while ER stress exacerbates lipid depletion by suppressing compensatory lipid deposition pathways in the absence of ApoE. These observations are consistent with prior evidence demonstrating that activated human HSCs treated with palmitate exhibit reduced lipid accumulation under high ER stress conditions, highlighting the unique capacity of HSCs to mobilize lipids in response to cellular stress, which is distinct from other hepatic cell types.

3.4. Arctigenin Modulates Lipid Droplet Size through Adaptive Lipogenesis and Lipolysis in ER Stress-Exposed Hepatic Stellate Cells

De novo lipogenesis (DNL) governs LD size by driving lipid biosynthesis and facilitating LD expansion through lipogenic enzymes, which catalyze triglyceride synthesis and mediate their deposition into the LD core. ,, Under ER stress conditions, activation of the UPR suppresses lipogenesis by downregulating key enzymes, disrupting LD homeostasis. , In tunicamycin-induced activating HSCs, we observed marked downregulation of Srebp1c (the transcriptional master regulator of fatty acid and triglyceride synthesis), and its downstream targets Acsl3 (essential for long-chain fatty acid activation) and Dgat2 (catalyzes the final step of triglyceride synthesis); , however, enzymes involved in fatty acid elongation (Elovl5) or alternative lipid storage molecules (Dgat1, Ppar-γ) remained unaffected (Figure a). − These findings indicate that ER stress selectively impairs the Srebp1c-driven DNL axis rather than globally disrupts lipid metabolism. Arctigenin treatment failed to restore Srebp1c, Acsl3, or Dgat2 expression but elevated Ppar-γ level (Figure a), suggesting that Srebp1c suppression under ER stress may involve irreversible transcriptional or epigenetic modifications, while Ppar-γ induction occurs via a distinct stress-responsive pathway. Furthermore, in activated HSCs, the ER stress alleviator 4-PBA did not alter the expression of Srebp1c, Dgat1, Acsl3, and Elovl5 but increased Dgat2 and Ppar-γ expression in the presence of arctigenin (Figure b). The partial restoration of Dgat2 implies that alleviating ER stress may ameliorate defects in lipid packaging, thereby promoting LD expansion through a Dgat2-dependent mechanism. Taken together, arctigenin may indirectly regulate lipid storage through enhancing the expression of compensatory molecules, i.e., Dgat2 and Ppar-γ, rather than restoring DNL in response to ER stress.

6.

Arctigenin modulates lipogenesis and lipolysis in ER stress-exposed hepatic stellate cells. Activating and activated HSCs were first treated with arctigenin at 50 μM for 12 h followed by treatment with 2 μg/mL Tunicamycin or 3 mM 4-PBA for 12 h, respectively. (a) Quantitative real-time PCR (qPCR) analysis of mRNA expression for lipogenesis-related genes in activating HSCs. (b) qPCR analysis of lipogenesis-related gene mRNA expression in activated HSCs. (c) qPCR analysis of Pnpla2 (involved in lipolysis) mRNA expression and Western blot analysis of ATGL and HSL protein expression in activating HSCs. Protein intensity was analyzed by ImageJ software. (d) qPCR analysis of Pnpla2 mRNA expression and Western blot analysis of ATGL and HSL protein expression in activated HSCs. Protein intensity was analyzed by ImageJ software. (e) qPCR analysis of Plin2 and Foxo1 mRNA expression and Western blot analysis of PLIN2 protein expression in activating HSCs. Protein intensity was analyzed by ImageJ software. (f) qPCR analysis of Plin2 and Foxo1 mRNA expression in activated HSCs. HSC, Hepatic stellate cell; D3 HSC, activating HSCs; aHSC, D7 (fully) activated HSCs; ATG, arctigenin; Tun, tunicamycin; 4-PBA, sodium 4-phenylbutyrate. Data are presented as mean ± standard deviation (mean ± SD, n ≥ 3 per group). Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. Significance: ns: not significant, P > 0.05; *: P < 0.05; **: P < 0.01; ***: P < 0.001.

Lipolysis, the enzymatic hydrolysis of lipids within LDs, is another critical mechanism that regulates LD size apart from DNL. In contrast to DNL, ER stress actively enhances lipolysis by stimulating lipase-dependent pathways, − thereby mobilizing stored lipids via lipases such as ATGL and HSL. In activating HSCs, tunicamycin-induced ER stress significantly upregulated both Pnpla2 (ATGL mRNA) and ATGL protein levels and increased phosphorylation of HSL (p-HSL, the active form) without altering total HSL expression (Figure c). This suggests that ER stress enhances lipolytic capacity through transcriptional (Pnpla2), translational (ATGL), and post-translational (HSL phosphorylation) mechanisms, likely as an adaptive response to mobilize fatty acids for energy production to facilitate HSC activation. Arctigenin treatment markedly reduced ATGL expression but did not affect p-HSL or total HSL (Figure c). This selective inhibition implies that arctigenin may interfere with transcriptional regulation of Pnpla2 or destabilize ATGL protein, while HSL phosphorylation is regulated independently of ATGL. Conversely, in activated HSCs, neither 4-PBA alone nor combined arctigenin/4-PBA altered ATGL or HSL expression (Figure d). This may be because expression of ATGL is already very low in activated HSCs and cannot be further reduced to maintain a minimal lipolytic activity necessary for cellular homeostasis. Hence, arctigenin may directly regulate ATGL-mediated lipolysis by suppressing its expression, thereby limiting LD degradation under the ER stress conditions.

Plin2 and Foxo1 function as regulators of ATGL-mediated lipolysis. Plin2, an LD-associated protein, colocalizes with Pnpla2 (the gene encoding ATGL) on LDs. Overexpression of Plin2 inhibits ATGL-dependent lipid hydrolysis, thus stabilizing LD content. Conversely, Foxo1, a transcription factor, directly activates Pnpla2 by binding to its promoter region, thereby enhancing triglyceride degradation. These opposing roles suggest that ER stress-induced changes in ATGL activity in HSCs may be mediated by Plin2 and Foxo1. As demonstrated in Figure e, tunicamycin-induced ER stress in activating HSCs did not alter Plin2 expression at either the mRNA or protein level, indicating that Plin2 is not directly involved in ER stress-driven ATGL activation. However, treatment with arctigenin significantly upregulated Plin2 levels, likely as a compensatory response to ATGL inhibition (Figure e). Conversely, tunicamycin drastically increased Foxo1 expression, which was abolished by arctigenin treatment (Figure e). This indicates that Foxo1 is a critical ER stress-sensitive transcriptional activator of ATGL, linking proteotoxic stress to enhanced lipolytic activity. In activated HSCs, neither the ER stress inhibitor 4-PBA nor arctigenin modulated Plin2 or Foxo1 levels (Figure f). This aligns with their lack of effect on ATGL in activated HSCs, suggesting that Plin2 and Foxo1 respond to ER stress in an activation-stage-dependent manner and exert their regulatory effects primarily during the early stress phase (activating HSCs) rather than the chronic phase (fully activated HSCs). Collectively, these findings indicate that Foxo1 is the primary ER stress-responsive regulator of ATGL, and its activity is regulated by the antilipolytic effect of arctigenin. Further mechanistic analyses are required to fully elucidate the precise role of Foxo1 in this pathway as well as its potential mechanism in the regulation of ATGL.

3.5. ERAD Is Involved in Arctigenin-Mediated Lipid Homeostasis in ER Stress-Exposed Hepatic Stellate Cells

The ERAD system serves as a critical protein quality control mechanism within the ER, functioning to preserve ER homeostasis by mediating the ubiquitin-dependent degradation of misfolded ER-resident proteins. Recent studies have further elucidated the pivotal role of ERAD in the regulation of lipid metabolism, where it governs the turnover of lipid biosynthetic enzymes via ubiquitination and subsequent proteasomal degradation, and enhances secretion of lipases. ,, These findings suggest a potential relationship between the ERAD machinery and arctigenin-mediated lipid homeostasis, particularly under ER stress conditions. To investigate this hypothesis, we first assessed ERAD activity by analyzing the Syvn1 (Hrd1) ubiquitin ligase complex, a central regulatory branch of the mammalian ERAD system. ERAD substrates are recruited to the Syvn1 complex for degradation, facilitated by ER-resident cochaperones such as Dnajb9 or Dnajc10, which mediate substrate recognition. Following ubiquitination, substrates are transported to the cytosolic proteasome via the cytoplasmic adaptor Herpud1. , As shown in Figure a, tunicamycin-induced ER stress significantly upregulated the expression of Dnajb9, Herpud1, and Syvn1 in activating HSCs. However, this upregulation was absent for Dnajc10 (Figure a), indicating that ERAD substrates are selectively recognized and interact with Dnajb9, but not Dnajc10, under these conditions, with Herpud1 mediating the delivery of ubiquitinated substrates to the proteasome, completing the ERAD cascade. Supplementation of arctigenin effectively downregulated the expression of Dnjb9, Herpud1, and Syvn1 in activating HSCs (Figure a), suggesting that arctigenin attenuates Syvn1-mediated ERAD activity during ER stress. Consistent with this, neither arctigenin nor the chemical chaperone 4-PBA altered Dnajc10 expression in activated HSCs (Figure b). Both arctigenin and 4-PBA reduced Dnajb9 and Herpud1 levels compared with untreated controls (Figure b). Although arctigenin alone or in combination with 4-PBA suppressed Syvn1 expression, 4-PBA alone had no effect on Syvn1 levels (Figure b). These results confirm that arctigenin inhibits Syvn1-ERAD activity at various stages of HSC activation by alleviating ER stress. Furthermore, the inability of 4-PBA to modulate Syvn1 implies that this ER stress mitigator may target alternative ERAD ubiquitin ligases, such as gp78, rather than Syvn1. Collectively, these findings demonstrate that the ERAD system is dynamically engaged in ER stress responses during HSC activation and modulated by arctigenin. This regulatory crosstalk indicates the potential role of ERAD in lipid homeostasis within the ER.

7.

Arctigenin regulates ERAD homeostasis in ER stress-exposed hepatic stellate cells. Activating and activated HSCs were first treated with arctigenin at 50 μM for 12 h followed by treatment with 2 μg/mL Tunicamycin or 3 mM 4-PBA for 12 h, respectively. (a) Quantitative real-time PCR (qPCR) analysis of mRNA expression for genes associated with the ERAD system in activating HSCs. (b) qPCR analysis of ERAD-related gene mRNA expression in activated HSCs. HSC, Hepatic stellate cell; D3 HSC, activating HSCs; aHSC, D7 (fully) activated HSCs; ATG, arctigenin; Tun, tunicamycin; 4-PBA, sodium 4-phenylbutyrate. Data are presented as mean ± standard deviation (mean ± SD, n ≥ 3 per group). Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. Significance: ns: not significant, P > 0.05; *: P < 0.05; **: P < 0.01; ***: P < 0.001.

To further investigate the role of ERAD in the regulation of lipid homeostasis under ER stress, we treated ER stress-exposed HSCs with Eeyarestatin-I (Eer I), a potent ERAD inhibitor that selectively blocks deubiquitinating enzymes (experimental design shown in Figure S1e,f). As shown in Figure a,b, Eer I alone significantly reduced both mRNA and protein levels of the lipase ATGL (gene symbol Pnpla2) compared to tunicamycin-treated activating HSCs, even when co-administered with arctigenin. Eer I also downregulated the mRNA expression of Foxo1, an ER stress-responsive regulator of ATGL (Figure ), in ER stress-induced activating HSCs. Notably, Eer I selectively increased the lipogenic gene Ppar-γ but not Dgat2 expression (Figure a). In activated HSCs, Eer I alone or combined with arctigenin markedly reduced ATGL protein (but not mRNA) levels compared with 4-PBA-treated activated HSCs (Figure c,d). Interestingly, Foxo1 mRNA levels remained unchanged by Eer I, even in combination with arctigenin (Figure c). Although Eer I alone or with arctigenin treatment did not alter the expression of lipogenic markers Dgat2 and Ppar-γ compared to 4-PBA-treated activated HSCs (Figure c), the unchanged Dgat2 levels may reflect a limitation to further upregulation following prior ER stress alleviation. These findings suggest that ERAD contributes to arctigenin-mediated lipid homeostasis through two mechanisms: suppression of adaptive lipogenesis and enhancement of ATGL-mediated lipolysis. ERAD inhibition upregulated the transcription level of adaptive lipogenic factor Ppar-γ and Dgat2, possibly due to impaired ubiquitin-proteasomal degradation of their upstream regulating (transcription) factors, , implying that they may be indirectly regulated by ERAD modification. ERAD is essential for ATGL synthesis, likely via transcriptional (Foxo1-dependent) and post-translational mechanisms. In tunicamycin-treated HSCs (high ER stress), ERAD inhibition suppresses Foxo1 activity, thereby reducing the level of ATGL transcription. In 4-PBA-treated HSCs (low ER stress), ERAD directly stabilizes the ATGL protein. ERAD inhibition may induce immature ATGL retention within the ER lumen, triggering autophagic degradation or impairing its translocation to ER. , Further studies are needed to identify ERAD-targeted lipogenic factors as well as the mechanism by which ERAD stabilizes ATGL, to fully elucidate ERAD as a key mediator of arctigenin-mediated LD homeostasis.

8.

ERAD is involved in arctigenin-mediated lipid homeostasis in ER stress-exposed hepatic stellate cells. (a) Activating HSCs were first treated with arctigenin at 50 μM for 12 h followed by treatment with 2 μg/mL tunicamycin alone or in combination with 2 μM Eer I for 8 h. Real-time PCR analysis of mRNA expression of genes involved in lipolysis and lipogenesis in activating HSCs. (b) Activating HSCs were first treated with arctigenin at 50 μM for 12 h followed by treatment with 2 μg/mL tunicamycin alone or in combination with 2 μM Eer I for 10 h. Western blot analysis of protein expression of Pnpla2 in activating HSCs. Protein intensity was analyzed by ImageJ software. (c) Activated HSCs were first treated with arctigenin at 50 μM for 12 h followed by treatment with 3 mM 4-PBA alone or in combination with 2 μM Eer I for 8 h. Real-time PCR analysis of mRNA expression of genes involved in lipolysis and lipogenesis in activated HSCs. (d) Activated HSCs were first treated with arctigenin at 50 μM for 12 h followed by treatment with 3 mM 4-PBA alone or in combination with 2 μM Eer I for 10 h. Western blot analysis of protein expression of Pnpla2 in activated HSCs. Protein intensity was analyzed by ImageJ software. HSC, Hepatic stellate cell; D3 HSC, activating HSCs; aHSC, D7 (fully) activated HSCs; ATG, arctigenin; Tun, tunicamycin; 4-PBA, sodium 4-phenylbutyrate; Eer I, Eeyarestatin-I. Data are presented as mean ± standard deviation (mean ± SD, n ≥ 3 per group). Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. Significance: ns: not significant, P > 0.05; *: P < 0.05; **: P < 0.01; ***: P < 0.001.

In summary, our study elucidates a previously unrecognized role and mechanistic pathway of arctigenin, a bioactive lignan derived from Arctium lappa L., in mitigating ER stress-driven HSC activation. We demonstrate that arctigenin suppresses HSC activation by restoring lipid homeostasis through ERAD-dependent mechanisms under ER stress. ER stress induces lipid dysregulation, characterized by the loss of LDs, which is mechanistically dependent on ERAD activation through modulating the imbalance of adaptive lipogenesis and lipolysis. Arctigenin alleviates these effects by normalizing ERAD activity, thereby preserving LD abundance and size through dual modulation, upregulating lipogenic factors (Ppar-γ, Dgat2) and suppressing lipolytic drivers (ATGL). A critical limitation of this work is its exclusive reliance on in vitro models, which constrains translational applicability. To advance arctigenin toward clinical practice or food supplements, future studies should validate these findings in vivo using animal models of liver fibrosis to assess the efficacy, pharmacokinetics, and safety of arctigenin. Furthermore, the molecular targets of arctigenin within the ERAD cascade remain unclear, warranting proteomic or structural studies. Investigating the synergistic effects of arctigenin with existing antifibrotics and optimizing its bioavailability as a nutraceutical agent are crucial for clinical translation. Addressing these gaps will enhance understanding of the therapeutic potential of arctigenin and ER stress-lipid interplay in fibrotic diseases.

Glossary

Abbreviations

Acsl3

acyl-CoA synthetase long-chain family member 3

Acta2

actin α 2, smooth muscle

α-SMA

α-smooth muscle actin

Atf4

activating transcription factor 4

ATGL

adipose triglyceride lipase

Chop

C/EBP homologous protein

Col1a1

collagen type I α 1 chain

Cyp26a1

cytochrome P450 family 26 subfamily A member 1

Dgat1

diacylglycerol O-acyltransferase 1

Dgat2

diacylglycerol O-acyltransferase 2

Danjb9

DnaJ heat shock protein family (Hsp40) member B9

Dnajc10

DnaJ heat shock protein family (Hsp40) member C10

Eer I

eeyarestatin-I

eIF2α

eukaryotic initiation factor 2

Elovl5

ELOVL fatty acid elongase 5

ER

endoplasmic reticulum

ERAD

endoplasmic reticulum-associated degradation

Foxo1

Forkhead box protein O1

Grp78

glucose-regulated protein 78

Herpud1

homocysteine inducible ER protein with ubiquitin like domain 1

HSC

hepatic stellate cell

LD

lipid droplet

Lrat

lecithin retinol acyltransferase

4-PBA

sodium 4-phenylbutyrate

Plin2

perilipin 2

Pnpla2

patatin like phospholipase domain containing 2

Pnpla3

patatin like phospholipase domain containing 3

Ppar-γ

peroxisome proliferator-activated receptor γ

Srebp1c

sterol regulatory element binding protein 1c

Syvn1

synoviolin 1

UPR

unfolded protein response

Xbp1s

spliced form of X-box binding protein 1

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jafc.5c01366.

• Schematic flowchart illustrating experimental design (Figure S1); determination of optimal concentrations of arctigenin (Figure S2); arctigenin inhibits the UPR in ER stress-exposed hepatic stellate cells (Figure S3); and arctigenin increases retinoid content in activating hepatic stellate cells (Figure S4) (PDF)

M.X.: conceptualization, methodology, investigation, formal analysis, writingoriginal draft, and writingreview and editing. J.L.: methodology, investigation, formal analysis, and data curation. L.M.M.A.: Methodology, investigation, and data curation. J.W.: methodology, formal analysis, software, and data curation. M.C.T.-A.: methodology and formal analysis. Y.L.: methodology, software. M.B.-H.: methodology and writingreview and editing. H.M.: conceptualization, supervision, funding acquisition, and writingreview and editing.

The authors declare no competing financial interest.