Exercise training and Silymarin consumption can ameliorate mitophagy signaling flux in hepatocytes of rats with dexamethasone-induced non-alcoholic fatty liver disease

Abstract

This study aimed to investigate the effects of exercise training combined with silymarin supplementation on mitophagy markers in hepatocytes of rats with dexamethasone-induced non-alcoholic fatty liver disease (NAFLD). Forty-two male Wistar rats (6 weeks old) were divided into 7 groups (n = 6 per group): 1- Control (CON), 2- Dexamethasone (DEX), 3- DEX + moderate-intensity training (DEX-MIT), 4- DEX + high intensity training (DEX-HIT), 5- DEX + silymarin (DEX-S), 6- DEX + moderate intensity training + silymarin (DEX-MIT-S), 7- DEX-high intensity training + silymarin (DEX-HIT-S). NAFLD was induced by subcutaneous administration of dexamethasone for 7 days. Exercise groups underwent 8 weeks of treadmill running (5 sessions/week) at matched distances for MIT and HIT protocols. Silymarin was administered via oral gavage at a dose of 300 mg/kg body weight/day. Gene expression levels of mTORC1, AMPKα2, Bcl-2, Parkin, and LC3 were measured using real-time PCR. Protein levels of PINK1, Beclin-1, and P62 were assessed by western blotting. Moderate and high intensity training significantly reduced Bcl-2 and LC3 gene expression and increased P62 protein levels compared to the DEX group (P < 0.05). Silymarin supplementation significantly decreased expression of Parkin, Bcl-2, LC3, and PINK1 compared to the DEX (P < 0.05). Bcl-2 and LC3 gene expressions were lower in DEX-MIT-S and DEX-HIT-S compared to DEX (P < 0.05). PINK1 levels were reduced in the DEX-MIT-S relative to DEX (P < 0.05). LC3 gene expression was higher in DEX-HIT-S compared to DEX-MIT-S (P < 0.05). The findings suggest that both exercise training and silymarin supplementation can attenuate excessive mitophagy signaling in hepatocytes of rats with dexamethasone-induced NAFLD, potentially providing hepatoprotective effects against further damage.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-20512-w.

Introduction

Autophagy is an essential cellular process in eukaryotic cells that involves the recycling of surplus or abnormal cellular components and damaged organelles-including mitochondria-via lysosomes. This process facilitates the generation of vital nutrients such as amino acids, glucose, and fatty acids, thereby maintaining cellular homeostasis and promoting cell survival¹. Mitophagy, a specialized form of autophagy, is responsible for the selective elimination of damaged mitochondria and plays a crucial role in preserving liver homeostasis. It is vital for sustaining mitochondrial and cellular balance during liver stress. Reduced mitophagy has been associated with mitochondrial dysfunction and liver failure, underscoring its importance in maintaining liver health^2,3. In response to cellular stressors such as high nutrient intake or elevated levels of reactive oxygen species, mitophagy is triggered when mitochondrial membrane potential is compromised. Under these conditions, PTEN-induced kinase 1 (PINK1) accumulates on the outer mitochondrial membrane and phosphorylates mitofusin 2 (Mfn2). This phosphorylation event enables the recruitment of Parkin, an E3 ubiquitin ligase, to the damaged mitochondria, initiating subsequent steps in the mitophagy process⁴. Following Parkin’s recruitment, parkin ubiquitinates various mitochondrial outer membrane proteins, effectively tagging the mitochondria for selective degradation. These ubiquitinated organelles are then engulfed by autophagosomes and delivered to lysosomes for breakdown—a process collectively referred to as mitophagy⁵. The intricate regulation of this system enables cells to eliminate dysfunctional mitochondria effectively, thereby limiting the spread of mitochondrial damage. Preserving mitochondrial integrity is critical for maintaining energy production and ensuring cell survival under stress. Thus, proper mitophagy function is essential for overall cellular health and organismal well-being.

Non-alcoholic fatty liver disease (NAFLD) is recognized as one of the most common causes of chronic liver disease and is ranked as the eighth leading cause of mortality worldwide. NAFLD encompasses a spectrum of conditions, including hepatic steatosis, non-alcoholic steatohepatitis (NASH), and cirrhosis^2,3. Mitochondrial dysfunction in NAFLD involves several interconnected factors such as impaired mitophagy, oxidative stress, and alterations in mitochondrial biogenesis. Studies have demonstrated that defective mitophagy is a hallmark of NAFLD, leading to increased oxidative stress and inflammation in hepatic tissue⁶. This dysfunction not only exacerbates lipid accumulation⁷ but also promotes hepatic stellate cell activation, contributing to the development of liver fibrosis^6,8. Therefore, understanding and targeting these underlying mechanisms may offer novel therapeutic strategies for managing NAFLD and halting its progression.

Exercise training is considered a first-line preventive and therapeutic non-pharmacological approach for various metabolic diseases, including disease NAFLD⁹. Both clinical studies and experimental models—whether genetic or diet-induced—have demonstrated the positive effects of exercise training on NAFLD, primarily through the activation of autophagy. Studies suggest that physical exercise exerts its protective role in pathological conditions by improving mitochondrial function, morphology, and bioenergetics, particularly in disorders associated with mitochondrial dysfunction^10–12. In a zebrafish model of NAFLD, regular exercise has been shown to stimulate mitophagosome formation and enhance mitochondrial quality. The underlying mechanism involves the restoration of the PINK1/Parkin signaling pathway, which plays a key role in promoting mitophagy¹³. Similarly, a study by Rosa-Caldwell et al. (2022) demonstrated that combining exercise with weight loss interventions enhances hepatic PINK1/Parkin-mediated mitophagy, facilitating the clearance of dysfunctional mitochondria and potentially reversing metabolic disturbances associated with NAFLD¹⁴. Moderate physical activity appears to promote basal autophagy in the liver, offering protection against hepatic fat accumulation¹¹. High-intensity interval training (HIIT) has emerged as an effective exercise strategy for managing NAFLD, particularly in individuals with limited time. Although its underlying mechanisms remain unclear, animal studies suggest that the benefits of HIIT may involve AMPK-mediated improvements in mitochondrial function through enhanced autophagy¹⁵.

Silymarin, a compound derived from milk thistle- and its major flavonolignan component, silybin, are widely used in phytotherapy for liver disorders. Silymarin exhibits multiple pharmacological properties, including antioxidant, anti-inflammatory, antidiabetic, anticancer, and anti-fibrotic effects—making it a promising agent for liver health and disease management¹⁶. In a randomized, double-blind, placebo-controlled trial, 48 weeks of silymarin administration in non-cirrhotic patients with non-alcoholic steatohepatitis (NASH) did not significantly improve the NAFLD Activity Score (NAS) compared to placebo. However, a high completion rate of 80% indicated good tolerability. Moreover, secondary outcomes—such as improvements in fibrosis and normalization of ALT and AST levels—offered preliminary evidence supporting the safety and potential efficacy of silymarin¹⁷.

Exercise and silymarin have both shown therapeutic potential for managing NAFLD. While exercise improves metabolic health and reduces hepatic fat, silymarin exhibits hepatoprotective properties through its antioxidant and anti-inflammatory effects. However, the combined effect of exercise training and silymarin supplementation on dexamethasone-induced NAFLD, particularly in relation to mitophagy pathways, remains poorly understood. Therefore, the present study aimed to investigate the impact of this combined intervention on mitophagy-related markers in the livers of rats subjected to dexamethasone-induced hepatic steatosis.

Materials and methods

Animal care

The study protocol was approved by the specialized Ethics Committee in Biomedical Research at the University of Mazandaran (approval number IR.MUI.MED.REC.1400.677). Furthermore, all animal protocols were conducted in accordance with the ARRIVE guidelines (https://arriveguidelines.org).

Forty-two 6-week-old male Wistar rats were purchased from the Pasteur Research Institute of Iran. The animals had free access to water and standard chow and were housed in cages (four rats per cage) under a 12-h light/dark cycle at a room temperature of 21–23 °C. After one week of adaptation, the animals were divided into seven weight-matched groups (Table 1): 1. Control (CON; n = 6), 2. Dexamethasone (DEX; n = 6), 3. DEX + Moderate Intensity Training (DEX-MIT; n = 6), 4. DEX + High Intensity Training (DEX-HIT; n = 6), 5. DEX + Silymarin (DEX-S; n = 6), 6. DEX + Moderate Intensity Training + Silymarin (DEX-MIT-S; n = 6), 7. DEX + High Intensity training + Silymarin (DEX-HIT-S; n = 6).

Table 1.

Animals’ weight.

Groups	Initial Weight (gr)	Final Weight (gr)
CON	261.4 ± 4.97	316.0 ± 5.94
DEX	236.66 ± 8.33	325.83 ± 12.89
DEX-MIT	227.37 ± 7.27	325.0 ± 9.56
DEX-HIT	240.5 ± 6.98	319.75 ± 7.58
DEX-S	215.75 ± 12.96	331.25 ± 18.76
DEX-MIT-S	232.85 ± 2.65	316.0 ± 5.79
DEX-HIT-S	234.28 ± 8.77	323.85 ± 13.84

Values are mean ± standard error of means; n = 6.

Dexamethasone administration

Dexamethasone, a synthetic glucocorticoid (obtained from Darupakhsh Pharmaceutical Chemistry Company, Iran), was administered subcutaneously in incrementally increasing doses, starting at 2.5 mg/kg on the first day and reaching 10 mg/kg body weight on the final day, over a period of seven consecutive days, to induce dyslipidemia and fatty liver in rats¹⁸. A preliminary induction was carried out on three rats through dexamethasone administration to histologically assess the presence of hepatic fat droplets. As shown in Fig. 1, the conspicuous accumulation of lipid vacuoles confirmed the successful establishment of fatty liver, following which the full experimental protocol was initiated.

Fig. 1.

A representative H&E-stained liver section from a dexamethasone-treated rat in the pilot study, demonstrating distinct fat vacuolation within hepatocytes (bar = 60 μm for A and bar = 20 μm for B).

Silymarin supplementation

Silymarin powder (Barij Essential Pharmaceutical Company, Iran) was dissolved in distilled water. Rats in the silymarin-treated groups received an oral dose of 300 mg/kg body weight, administered 30–40 min prior to their respective training sessions¹⁹.

Training protocol

After the adaptation phase, the animals underwent a maximum exercise capacity (MEC) test. They began running on a treadmill at 6 m/min with a 25% incline for 5 min. The speed was progressively increased by 0.5 m/min every 2 min until the animals could no longer continue running. Exhaustion was defined as the point at which the animals were unable to maintain the treadmill pace, at which time the test was terminated.

Rats in the exercise training groups underwent 8 weeks of distance-matched MIT or HIT on a treadmill, five times per week. Each session included a 10-min warm-up and cool-down period at a speed of 10 m/min and 0% incline. The HIT protocol consisted of six 3-min bouts at 40 m/min (85–95% of MEC), interspersed with 3-min active recovery at 20 m/min (40 –50% of MEC), all performed on a treadmill with a 15° incline Over four weeks, the duration, speed, and incline were gradually increased to 36 min per session, 40 m/min, and 15°, respectively.

The MIT groups exercised at the same speed as the active recovery phase of the HIT group (20 m/min on a treadmill with a 15° incline), gradually reaching 54 min of running by the eighth week (Table 2). The MEC was reassessed every two weeks using an incremental test to adjust the training workload accordingly²⁰.

Table 2.

Summary of training protocols.

Weeks	Type of training	Number of bouts	Speed (m/min)	Incline	Total duration (min)
Week 1	HIT	2	I:15–20 R:12–14	5°	12
	MIT	1	12–14		12–15.5
Week 2	HIT	3	I:20–25 R: 14–16	10°	18
	MIT	1	14–16		15–20
Week 3	HIT	4	I:25–30 R:16–18	15°	24
	MIT	1	16–18		20–28
Week 4	HIT	5	I:30–35 R:18–20	15°	30
	MIT	1	18–20		28–39
Week 5	HIT	6	I:35–40 R:20	15°	36
	MIT	1	20		40–54
Week 6	HIT	6	I:40 R:20	15°	36
	MIT	1	20		54
Week 7–8	HIT	6	I:40 R:20	15°	36
	MIT	1	20		54

HIT: High-intensity interval training, MIT: Moderate-intensity training, I: Intensive phase, R: Recovery phase.

Animal sacrifice and liver extraction

Following the 8-week intervention period and 48 h after the final exercise training session, the animals were fasted overnight (10 h) and then anesthetized with Ketamine (90 mg/kg) and Xylazine (10 mg/kg). Blood samples (10 mL) were collected via cardiac puncture and centrifuged at 3,000 rpm for 15 min. The resulting serum was separated and stored at − 20 °C for subsequent analysis. Following blood collection, the animals were sacrificed by cervical dislocation. The livers were then rapidly excised, immediately snap-frozen in liquid nitrogen, and stored at − 80 °C until further analysis.

Biochemical analysis

Serum concentrations of high-density lipoprotein (HDL), low-density lipoprotein (LDL), total cholesterol (TC), and triglycerides (TG) were analyzed using a colorimetric method, following the manufacturer’s instructions (Bionik, Iran; sensitivities: 2.5 mg/dL, 7 mg/dL, 0.113 mg/dL, and 0.53 mg/dL, respectively).

Real-time quantitative PCR

Total RNA was extracted from hepatocytes using an RNA isolation kit (Denazist, Iran).

Complementary DNA (cDNA) was synthesized using a cDNA synthesis kit (Yekta Tajhiz Azma, Iran), according to the manufacturer’s protocol. Quantitative real-time PCR was performed using the Rotor-Gene 6000 system (Corbett Research, Australia) and SYBR Green PCR Master Mix (Amplicon, Denmark) under the following conditions: Initial denaturation was performed for 15 min at 95 °C, followed by 40 cycles of denaturation for 20 s at 94 °C, annealing for 30 s at 58 °C, and extension for 30 s at 72 °C. The reaction mixture consisted of 2 µL cDNA, 8 µL Real Q Plus 2 × Master Mix Green (Amplicon, Denmark), 0.5 µL of each primer (10 pmol), and 9 µL of distilled water. Specific primers were designed using Primer Premier 5 software and then synthesized by Bioneer (South Korea). Primer sequences are listed in Table 3. The expression levels of mTORC1, AMPKα2, Bcl-2, Parkin, and LC3 were normalized to the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Relative gene expression levels were calculated using the Livak method (2^ − ΔΔCt).

Table 3.

Forward and reverse primer sequences used in this study.

Gene	Primer sequence (5′ → 3′)	Accession number	Product size (bp)
AMPKα2	F: GTTCTACCTCGCCTCCAGTCC R: GCTTAGTTGTGTTGAGTGCATCCA	NM_023991.1	158
Parkin	F: GTCTTCCAATGTAACCACCGC R: GGAGTAGCCAAGTTGAGCGTC	NM_020093.1	111
LC3	F: CTAACCAAGCCTTCTTCCTCC R: GCCGTCTTCATCTCTCTCGC	NM_022867.2	98
mTOR	F: ATTTGCCAACTACCTTCGGAAC R: ACTTCAAACTCCACATACTCAGCAG	NM_019906.2	129
Bcl-2	F: GCTACGAGTGGGATACTGGAGATGA R: ACAGCGGGCGTTCGGTTG	NM_016993.2	103
GAPDH	F: GGCAAGTTCAACGGCACAG R: CGCCAGTAGACTCCACGACATA	NM_017008.4	142

Western blot analysis of Beclin-1, P62 and PINK1

Dissected liver tissues were weighed and homogenized in lysis buffer containing 10 mM Tris–HCL (500 µl, pH = 8.0), 0.08 g NaCl, 1 mM EDTA, 1% Triton X-100 (10 µL), 0.025 g sodium deoxycholate, 0.01 g SDS and one tablet protease inhibitor cocktail. After centrifugation, the supernatants were collected and heat-denatured at 90 °C for 10 min. Equal amounts of protein (10 µg) were separated by SDS-PAGE using 10% polyacrylamide gels and transferred to polyvinylidene fluoride (PVDF) membranes. The membranes were blocked with 1% non-fat dry milk in TBS-T for 1 h at room temperature. Membranes were then incubated overnight at 4 °C with mouse monoclonal primary antibodies: β-Actin (C4: sc-47778), BECN1 (E-8: sc-48341), PINK1 (38CT20.8.5: sc-517353), and SQSTM1 (P-15: sc-10117), all purchased from Santa Cruz Biotechnology (USA), at a dilution of 1:300. After washing with TBS-T, membranes were incubated with HRP-conjugated secondary antibody (1:1000 in 1% [w/v] milk in TBS-T) for 1 h at room temperature. Protein bands were visualized using a chemiluminescence reagent and captured with an imaging system. Band intensities were quantified using ImageJ software (NIH, USA).

Statistical analysis

Data were statistically analyzed using SPSS software (version 24, IBM Corp., Armonk, NY, USA). The Shapiro–Wilk test was used to assess the normality of data and Levene’s test was applied to evaluate the homogeneity of variances. One-way analysis of variance (ANOVA) followed by the least significant difference (LSD) post hoc test was used to compare differences among the experimental groups. Effect sizes were calculated using partial eta squared (η²p) for ANOVA models, and Cohen’s d was used to quantify the magnitude of pairwise post hoc comparisons. Effect sizes were interpreted according to the following thresholds: Partial eta squared (η2p): small (≈0.01), medium (≈0.06), large (≥ 0.14). Cohen’s d: trivial (< 0.2), small (0.2–0.6), moderate (0.6–1.2), large (1.2–2.0), very large (2.0–4.0). All data are presented as mean ± standard error of the means (SEM), and statistical significance was accepted at p ≤ 0.05.

Results

Serum biochemical parameters

A one-way ANOVA revealed a statistically significant difference in serum glucose levels among the groups, F (6, 35) = 8.94, p < 0.0001, with a large effect size (η² = 0.585), indicating a strong impact of the interventions on glucose metabolism and homeostasis. Post hoc analysis demonstrated that the DEX group had significantly higher serum glucose levels compared to the control group (d = 2.04, p < 0.001). All intervention groups exhibited significantly lower glucose levels relative to the DEX group (p < 0.0001), with very large effect sizes observed for the following comparisons: DEX + S (d = 2.45), DEX—HIT (d = 2.37), DEX -MIT (d = 2.25), DEX—MIT—S (d = 2.28), and DEX—HIT—S (d = 2.30) (Table 4).

Table 4.

Effects of exercise training and silymarin treatment on serum biochemical parameters in rats.

Group	CON	DEX	DEX-MIT	DEX-HIT	DEX-S	DEX-MIT-S	DEX-HIT-S
Glucose (mg/dL)	116.6 ± 4.91	169.6 ± 13.63a	109 ± 6.94b	105 ± 7.62b	96.75 ± 4.13b	107.00 ± 7.52b	114 ± 1.29b
TG (mg/dL)	40 ± 6.42	61.33 ± 6.79a	47.87 ± 5.99	27 ± 1.70bc	25.5 ± 4.73b	43.42 ± 6.62b	49.0 ± 6.18e
TC (mg/dL)	85.8 ± 2.69	59.06 ± 3.34a	84.37 ± 2.72	81.25 ± 2.03b	73.75 ± 5.63b	79.28 ± 4.62b	82.42 ± 3.30b
HDL (mg/dL)	45 ± 2.50	32.83 ± 1.68a	44.87 ± 2.19b	44.14 ± 1.38b	35.75 ± 4.09a	40.28 ± 3.11b	45.14 ± 2.53b
LDL (mg/dL)	25.4 ± 0.67	11.33 ± 0.49a	22.37 ± 1.06b	26.37 ± 1.61bc	24.25 ± 1.43b	23.14 ± 1.65b	20.71 ± 1.84b

CON: Control; DEX: Dexamethasone; MIT: Moderate Intensity Training; HIT: High Intensity Training; S: Silymarin; Data were expressed as Means ± SEM; n = 6 per group; ^aversus the CON group; ^b versus the DEX group; ^c versus the DEX-MIT; ^d versus the DEX-HIT group; ^e versus the DEX-S group; P ≤ 0.05.

A statistically significant difference in serum triglyceride (TG) levels among the groups was identified via one-way ANOVA, F (6, 35) = 4.51, p = 0.002, indicating a large effect size (η² = 0.415). These findings suggest that the interventions had a meaningful impact on serum triglyceride concentrations. Post hoc comparisons revealed that the DEX group exhibited significantly higher TG levels compared to the control group (d = 1.26, p = 0.021). Significant reductions in TG levels relative to the DEX group were observed in the following intervention groups: DEX-HIT (d = 1.70, p < 0.0001), DEX-S (d = 1.77, p = 0.001), and DEX-MIT-S (d = 1.04, p = 0.034). Additionally, TG levels in the DEX-HIT group were significantly lower than in the DEX-MIT group (d = 1.22, p = 0.007). Conversely, the DEX-HIT-S group demonstrated significantly higher TG levels compared to both the DEX-S (d = 1.30, p = 0.014) and DEX-HIT (d = 1.45, p = 0.006) groups.

Total cholesterol (TC) levels showed significant variation among the groups, as determined by one-way ANOVA, F (6, 35) = 6.79, p < 0.001, with a large effect size (η2 = 0.517). Post hoc analysis revealed that the DEX group had significantly lower TC levels compared to the control group (d = 2.33, p < 0.0001). Importantly all experimental groups demonstrated significantly increased TC levels relative to the DEX group: indicating a return toward normal levels: DEX-MIT vs DEX (d = 2.70, p < 0.001), DEX-HIT vs DEX (d = 2.33, p < 0.001), DEX-S vs DEX (d = 1.47, p = 0.013), DEX-MIT-S vs DEX (d = 2.16, p < 0.001), and DEX-HIT-S vs DEX (d = 2.45, p < 0.001).

HDL levels differed significantly among the groups, as indicated by one-way ANOVA, F (6, 35) = 3.80, p = 0.005, with a large effect size (η2 = 0.381). AS shown in Table 4, the DEX group exhibited a significant reduction in serum HDL levels compared to the control group (d = 1.54, p = 0.003). in contrast HDL level were significantly elevated in the following groups compared to DEX group: DEX-MIT (d = 1.62, p = 0.001), DEX-HIT (d = 1.56, p = 0.002), DEX-MIT-S (d = 1.15, p = 0.038), and DEX-HIT-S (d = 1.70, p = 0.001). Moreover, the DEX-HIT-S group showed significantly higher HDL levels compared to the DEX–S group (d = 2.32, p = 0.021).

A highly significant difference in LDL levels was observed among groups, F (6, 35) = 11.64, p < 0.001, with a large effect size was (η2 = 0.648), indicating a strong impact of the interventions on serum LDL concentrations. As shown in Table 4, all experimental groups demonstrated significantly higher LDL levels compared to the DEX group, indicating a reversal toward normal levels. DEX-MIT vs DEX (d = 3.66, p < 0.001), DEX-HIT vs DEX (d = 4.20, p < 0.001), DEX-S vs DEX (d = 6.06, p < 0.001), DEX-MIT-S vs DEX (d = 3.23, p < 0.001), and DEX-HIT-S vs DEX (d = 2.54, p < 0.001). Additionally, LDL levels in the DEX-HIT group were significantly higher than the DEX-MIT group (d = 1.09, p = 0.035).

Mitophagy-related genes and proteins

A one-way ANOVA revealed no statistically significant differences in AMPKα2 expression among the groups, F (6, 28) = 1.84, p = 0.12. However, the observed moderate effect size (η² = 0.29) suggests that the interventions may still have exerted a biologically meaningful impact.

Similarly, the analysis of mTOR expression showed no statistically significant differences among the groups, F (6, 28) = 2.26, p = 0.069. Nevertheless, a moderate effect size (η2 = 0.34) was observed, indicating potentially meaningful biological trends. Notably, mTOR mRNA expression was reduced in all experimental groups compared to the control group, with statistically significant reductions observed in the DEX (d = 1.444, p = 0.014) and DEX-MIT (d = 2.139, p = 0.001) groups. Interestingly, mTOR expression was significantly higher in the DEX-MIT-S group compared to the DEX-MIT group (d = 1.865, p = 0.043), suggesting a possible modulatory effect of silymarin when combined with moderate-intensity training (Fig. 2).

Fig. 2.

Effects of exercise training and silymarin treatment on the expression of autophagy genes. CON: Control; DEX: Dexamethasone; MIT: Moderate Intensity Training; HIT: High Intensity Training; S: Silymarin; Data were expressed as Means ± S.E.M; n = 5 per group; ^a versus the CON group; ^b versus the DEX group; ^c versus the DEX-MIT; ^d versus the DEX-HIT group; ^e versus the DEX-S group; ^f versus the DEX-MIT-S group; P ≤ 0.05.

Significant differences were observed in Bcl2 expression across the groups, F (6, 28) = 5.60, p = 0.001, indicating a large effect size (η2 = 0.545). These findings suggest that the interventions had a strong impact on Bcl2 expression, consistent with its role in regulating cell survival. Notably, DEX treatment led to a significant increase in Bcl2 mRNA levels compared to the control group (p = 0.004, d = 1.25). However, all intervention groups showed a reversal of this effect, exhibiting significantly lower Bcl2 expression compared to the DEX group: DEX-MIT vs. DEX (d = 2.08, p < 0.001), DEX-HIT vs. DEX (d = 1.77, p < 0.001), DEX-S vs. DEX (d = 1.47, p = 0.001), DEX-MIT-S vs. DEX (d = 1.93, p < 0.001), and DEX-HIT-S vs. DEX (d = 1.55, p = 0.001) (see Fig. 2).

Beclin-1 expression showed a statistically significant difference among the groups, F (6, 28) = 3.27, p = 0.014, with a large effect size (η2 = 0.412). As illustrated in Fig. 3, Beclin-1 protein levels were significantly elevated in the DEX (d = 2.772, p = 0.026), DEX-MIT (d = 2.44, p = 0.002), DEX-HIT (d = 3.753, p = 0.001), and DEX-HIT-S (d = 2.01, p = 0.003) groups compared to the control (CON) group.

Fig. 3.

Effects of exercise training and silymarin treatment on the expression of autophagy proteins. CON: Control; DEX: Dexamethasone; MIT: Moderate Intensity Training; HIT: High Intensity Training; S: Silymarin; Data were expressed as Means ± S.E.M; n = 5 per group; ^a versus the CON group; ^b versus the DEX group; P ≤ 0.05. The bands shown were cropped from different regions of the same gel and arranged side-by-side for clarity.

PINK1 expression varied significantly among the groups, F (6, 28) = 11.72, p < 0.001, with a very large effect size (η² = 0.715). Compared to the control group, PINK1 expression was significantly elevated in all dexamethasone-treated groups: DEX (d = 5.565, p < 0.0001), DEX-MIT (d = 6.68, p < 0.0001), DEX-HIT (d = 7.98, p < 0.0001), DEX-S (d = 2.018, p = 0.001), DEX-MIT-S (d = 3.28, p = 0.001), and DEX-HIT-S (d = 3.55, p = 0.0001). Furthermore, PINK1 expression was significantly lower in the DEX-S (d = 1.119, p = 0.03) and DEX-MIT-S (d = 1.599, p = 0.03) groups compared to the DEX group, suggesting a potential modulatory effect of silymarin and exercise co-interventions on mitochondrial quality control (Fig. 3).

No statistically significant difference was observed in Parkin expression among the groups, F (6, 28) = 2.39, p = 0.055; however, a moderate effect size was detected (η² = 0.347). As shown in Fig. 2, DEX administration increased Parkin mRNA levels compared to CON group, although this difference did not reach statistical significance. Parkin mRNA level was higher in DEX-HIT group when compared to the CON group (d = 1.784, P = 0.011). Additionally, Silymarin treatment resulted in a significant reduction in Parkin mRNA levels compared to DEX group (d = 1.445, P = 0.024).

A one-way ANOVA revealed significant differences in LC3 gene expression among the groups, F (6, 28) = 12.44, p < 0.001, with a very large effect size (η2 = 0.727), indicating robust group effects. Compared to the control group, LC3 mRNA expression was significantly elevated in the DEX (d = 2.339, p < 0.0001) and DEX-HIT-S (d = 1.869, p = 0.031) groups.

Notably, all intervention groups showed a significant reduction in LC3 expression relative to the DEX group, including: DEX-MIT (d = 3.045, p < 0.001), DEX-HIT (d = 2.120, p < 0.001), DEX-S (d = 2.677, p < 0.001), DEX-MIT-S (d = 2.928, p < 0.001), and DEX-HIT-S (d = 1.334, p = 0.001). Moreover, LC3 expression in DEX-HIT-S was significantly higher than in both DEX-S (d = 2.434, p = 0.003) and DEX-MIT-S (d = 3.105, p = 0.001), suggesting a distinct modulatory effect of high-intensity interval training combined with silymarin on mitophagy- related gene regulation (see Fig. 2).

A significant group effect was observed for p62 expression, F (6, 28) = 2.68, p = 0.037, accompanied by a moderate-to-large effect size (η² = 0.382). Compared to the control group, p62 expression was significantly increased in the DEX-MIT (d = 3.079, p = 0.005) and DEX-HIT (d = 2.28, p = 0.003) groups. Furthermore, both the DEX-MIT (d = 1.96, p = 0.038) and DEX-HIT (d = 1.574, p = 0.024) groups exhibited significantly higher p62 levels than the DEX group (see Fig. 3).

Discussion

To the best of our knowledge, this is the first study to investigate the combined effects of exercise training and silymarin supplementation on the expression of mitophagy-related genes (mTORC1, AMPKα2, Bcl-2, Parkin, and LC3,) and proteins (Beclin-1, PINK1, and p62) involved in a rat model of dexamethasone-induced NAFLD.

In this study, as illustrated in Fig. 1, both microvesicular and macrovesicular fat vacuoles are prominently visible within the hepatocyte cytoplasm, confirming the successful induction of hepatic steatosis. Dexamethasone has been extensively investigated for its effects on liver function, particularly in relation NAFLD in animal models. Evidence suggests that dexamethasone disrupts hepatic lipid metabolism and may exacerbate liver injury under certain conditions²¹. Animal studies have highlighted the upregulation of hepatic CD36 mediated by glucocorticoid receptor activation as a key factor contributing to lipid metabolic disturbances induced by dexamethasone²². Additionally, chronic administration of dextran-conjugated dexamethasone has been shown to activate pro-inflammatory macrophage and promote adipose tissue inflammation, thereby linking glucocorticoid exposure to inflammatory pathways associated with obesity and NASH²³. Taken together, these findings underscore the complex and multifaceted role of dexamethasone in modulating liver function and its potential implications in NAFLD progression.

In this study, injection of DEX and the subsequent development of fatty liver caused a significant change in the variables related to mitophagy. The expression level of Beclin-1, PINK1 proteins, and of LC3 gene were elevated, suggesting increased stimulation of the clearance and removal of damaged mitochondria. It has been shown that DEX induces activation of the autophagy process through the AMPK signaling and upregulation of several autophagy-related genes²⁴. However, In the present study, one of the central regulators of cellular homeostasis, AMPKα2, showed no significant change between the CON and DEX groups, which may reflect inhibitory effects of dexamethasone on cellular energy- sensing mechanisms. Dexamethasone activates glucocorticoid receptors, promoting gluconeogenesis, which elevates blood glucose levels (see Table 4) and decreases AMP/ATP ratio—factors that collectively contribute to AMPK inhibition²⁵. In our study, although AMPKα2 expression did not change, mTOR expression was significantly reduced. It is now widely recognized that mTOR plays a key role in suppressing autophagy. In mammals, the first step of mitophagy requires the induction of Atg-dependent autophagy, which can be triggered either by AMPK activation (via an increased AMP/ATP ratio) or by mTOR inhibition induced by mitochondrial produced reactive oxygen species (ROS)²⁶. Although we did not measure the levels of free radical levels in this study, it has been shown that oxidative stress is a major contributor to dexamethasone-induced liver injury, primarily due to excessive free radical production²⁷.

The present study revealed that the expression levels of both Bcl-2 and Beclin-1 were elevated in hepatocytes of the dexamethasone-treated group compared to the control group. Upregulation of Bcl-2 protein in the liver tissue of patients with non-alcoholic and alcoholic fatty liver disease has been reported²⁸. Bcl-2, known as an anti-apoptotic protein, also functions as an anti-autophagic protein by inhibiting Beclin-1 activity²⁹. Bcl-2 influences autophagy by binding to Beclin-1 and participating in mitochondrial quality control mechanisms. This interaction suppresses autophagy primarily by preventing Beclin-1 from initiating autophagosome formation under normal physiological conditions. However, under condition of metabolic or oxidative stress, this inhibitory interaction can be disrupted, thereby enabling Beclin-1 to activate autophagy³⁰. It has also been reported that the anti-autophagic function of Bcl-2 depends on its subcellular localization. Notably, Bcl-2 localized at the endoplasmic reticulum can suppress autophagy, whereas mitochondrial Bcl-2 appears not to exert this repressive effect³¹.

In the present study, the levels of PINK1, Parkin, and LC3 were elevated in the dexamethasone-treated group compared to controls; However, the increase in Parkin did not reach statistical significant. The PINK1/Parkin pathway serves as a crucial mitochondrial quality control mechanism by identifying and eliminating damaged mitochondria, thereby preserving cellular homeostasis and mitochondrial integrity. In healthy mitochondria, PINK1 is synthesized in the cytoplasm and imported into the organelle via molecular channels in the mitochondrial membranes. Once inside, PINK1 is rapidly degraded by intramitochondrial proteases, preventing its accumulation and thereby suppressing unnecessary mitophagy signaling⁵. Mitochondrial damage disrupts the inner membrane potential, impeding PINK1 import and leading to its accumulation on the outer mitochondrial membrane. Accumulated PINK1 phosphorylates and activates Parkin on Ser65 within its ubiquitin-like domain³². Activated Parkin is recruited to the outer mitochondrial membrane, where it ubiquitinates several surface proteins such as Mitofusin 1/2, Miro1, and voltage-dependent anion channel⁵.This ubiquitination acts as a signal for adaptor proteins such as p62, which plays a key role by binding ubiquitinated proteins via its UBA domain and interacting with the autophagy marker LC3 through its LC3-interacting region. Through this dual interaction, p62 facilitates the targeting of damaged mitochondria to autophagosomes³². Eventually, these autophagosomes fuse with lysosomes, allowing for the degradation and recycling of dysfunctional mitochondria.

Several studies have demonstrated that exercise training enhances autophagy, thereby contributing to the prevention of NAFLD^33–35. In the present study, the expression levels of Beclin-1, PINK1, and Parkin tended to increase in the DEX-MIT group compared to the control group; however, their levels remained comparable to those observed in the DEX group. Both MIT and HIT training protocols significantly reduced the expression of Bcl-2 and LC3 relative to the DEX group. Furthermore, previously our published data from the same experimental samples revealed a significant increase in apoptotic markers in the DEX group³⁶. Given that apoptosis can be induced by excessive mitophagy, it appears that dexamethasone administration may have triggered extensive mitophagic activity. These findings suggest that both MIT and HIT exercise protocols may contribute to the restoration of mitochondrial quality and improvement cellular metabolism, thereby partially mitigating the detrimental mitophagy-related alterations induced by dexamethasone. It has been shown that chronic exercise training for 12 weeks induced beneficial alterations on liver mitochondrial morphology and increased mitochondrial biogenesis markers (PGC-1α and mtTFA), mitochondrial dynamics proteins (OPA1, DRP1), and autophagy-related proteins (Beclin-1, LC3-II, LC3II/LC3I). Additionally, Mitochondrial density and circularity were increased in both exercised groups³⁷. Exercise training also appears to modify mitochondrial phenotype, rendering it more resistant to deleterious stimuli, such as drug-induced damage, and promotes hepatocyte renewal¹².

In the present study, we observed that p62 protein levels remained significantly elevated in both exercise groups compared to the DEX group. Although p62 accumulation is often interpreted as an indicator of impaired autophagic flux—a dysfunction linked to the progression of NAFLD, its biological role appears to be more complex. Previous research by Komatsu et al. demonstrated that p62 can activate the transcription factor Nrf2 by binding to and sequestering its inhibitor, Keap1. This interaction stabilizes Nrf2, allowing it to translocate to the nucleus and stimulate the expression of a broad range of antioxidant and cytoprotective genes, including those involved in glutathione synthesis³⁸. Through this mechanism, the p62–Nrf2 signaling axis may contribute to the reduction of oxidative stress and preservation of mitochondrial function³⁹.

In this study, silymarin supplementation significantly reduced the expression of Bcl-2, PINK1, Parkin, and LC3 compared to the DEX group. Furthermore, serum glucose and triglyceride levels—which were elevated in DEX-induced NAFLD rats —were effectively restored to near-control levels following treatment. These findings are consistent with the well-documented antioxidant, anti-inflammatory, and antifibrotic properties of silymarin, which has been shown to prevent mitochondrial dysfunction in rodent models of NASH and to ameliorate oxidative stress, dyslipidemia, and hepatic steatosis in NAFLD.

Interestingly, the observed downregulation of mitophagy-related genes by silymarin may reflect a secondary effect of reduced mitochondrial injury and reactive oxygen species (ROS) burden, thereby diminishing the need for compensatory mitophagy. Although the precise molecular mechanisms through which silymarin modulates autophagy remain incompletely elucidated, recent findings by Liu et al. (2025) demonstrated that silibinin—the major bioactive component of silymarin—attenuated hepatic lipid accumulation and preserved mitochondrial integrity via modulation of OPA1 (Optic Atrophy 1), a key regulator of mitochondrial fusion⁴⁰.

The present study revealed that the combined intervention of exercise training and silymarin supplementation did not yield a synergistic effect on the expression of key mitophagy-related genes (Bcl-2, Parkin, LC3) and proteins (Beclin-1, PINK1, and p62) when compared to either treatment alone. This lack of additive effect suggests that both interventions may exert their effects through overlapping intracellular signaling pathways. Recent studies suggest that both silymarin and exercise may exert their effects on mitophagy through overlapping intracellular signaling pathways—particularly those involving AMPK activation, SIRT1 upregulation, and NF-κB inhibition. Exercise training enhances AMPK phosphorylation, facilitating mitochondrial turnover and promoting cellular energy homeostasis⁴¹. Likewise, silymarin activates AMPK and appears to support similar outcomes under metabolic stress conditions⁴². Furthermore, both interventions upregulate SIRT1, a deacetylase involved in mitophagy induction and mitochondrial biogenesis⁴³. These shared molecular targets also extend to the suppression of NF-κB, a key driver of inflammatory mitochondrial damage. Given these mechanistic overlaps, it is reasonable to hypothesize that the co-application of silymarin and exercise training may not lead to additive effects under certain conditions, potentially due to target saturation or pathway redundancy. This possibility could account for the plateaued response observed when both interventions were applied simultaneously in our study. However, further mechanistic work is needed to delineate whether redundant signaling explains the lack of synergy, or whether other compensatory pathways are involved.

Conclusion

These results suggest that under pathological conditions of excessive autophagy, both exercise training and silymarin consumption can ameliorate mitophagy signaling activation, likely through reducing oxidative stress levels, improving mitochondrial dynamics and integrity, and protecting the liver from further damage. However, the combination of these two interventions does not provide additional benefits in improving mitophagy flux in rats with dexamethasone-induced non-alcoholic fatty liver disease.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

NAFLD

Non-alcoholic fatty liver disease

NASH

Non-alcoholic steatohepatitis

DEX

Dexamethasone

HIT

High-intensity interval training

MIT

Moderate-intensity training

S

Silymarin

AMPKα2

Adenosine monophosphate-activated protein kinase α2

mTORC1

Mammalian target of rapamycin complex 1

Bcl-2

B-cell lymphoma 2

LC3

Microtubule-associated protein 1 light chain 3

Beclin-1

Bcl-2-interacting myosin-like coiled-coil protein 1

PINK1

PTEN-induced putative kinase 1

HDL

High-density lipoprotein

LDL

Low-density lipoprotein

TC

Total cholesterol

TG

Triglyceride

Author contributions

Fatemeh Mokhtari-Andani: Data curation, Methodology, Formal analysis, Writing manuscript Elahe Talebi—Garakani: Conceptualization, Formal analysis, Writing—review & editing Khadijeh Nasiri: Methodology (real time PCR data evaluation) Abolfazle Akbari: Methodology (biochemical evaluation). Elahe Talebi—Garakani: Conceptualization, Formal analysis, Writing—review & editing. Khadijeh Nasiri: Methodology (real time PCR data evaluation). Abolfazle Akbari: Methodology (biochemical evaluation).

Funding

This research was supported by the Iran National Science Foundation (INSF) (grant number 99014697) who helped of the financial costs of this project.

Data availability

All data are available from the corresponding author on reasonable request.

Declarations

Ethics approval

All care and ethical principles were observed as per the guidelines for the use and care of laboratory animals approved by the ethics committee of Mazandaran University of medical science with the Code of Ethics IR.MUI.MED.REC.1400.677.

Competing interests

The authors declare no competing interests.