Withaferin A was purchased from Xenon Biosciences (India), and the chow diet was procured from Adita Biosys Pvt Ltd. The high-fat diet was procured from VRK Nutritional Solutions, India. Glucose and fructose were procured from Sisco Research Laboratories Pvt. Ltd. Commonly used liver function test enzymes, like aspartate transaminases or aspartate aminotransferase (AST), alanine transaminase or alanine aminotransferase (ALT), and alkaline phosphatase (ALP), and lipid molecules like total cholesterol, triglycerides (TG), and high-density lipoproteins (HDL) kits were purchased from Agape Diagnostics Ltd. TRIzol reagent, sodium palmitate, oleate and Oil Red O stain solution were purchased from Sigma Aldrich, St. Louis, Missouri, United States. cDNA synthesis and SYBR green kits were purchased from Thermo Fisher Scientific. A TGF-β1 ELISA kit was purchased from Krishgen Biosystems. HepG2 and Huh7 cells were purchased from NCCS Pune, India. The cell culture media Minimum Essential Medium Eagle (MEM), Ham DMEM/F-12, 1:1 mixture, bovine serum albumin, and hematoxylin were purchased from HiMedia, India. Fetal bovine serum (FBS) and antibiotics like penicillin/streptomycin were purchased from Gibco. 25-Hydroxycholesterol was a gift provided by Dr. Perumal Madan Kumar, CSIR-CFTRI, Mysuru. Taurochenodeoxycholic acid and deoxycholic acid were a gift from Dr. Ramprasad Talahalli, CSIR-CFTRI, Mysuru.

Male Swiss albino mice (4–6 weeks old) were fed ad libitum sugar water (SW) containing glucose (18.9 g/L) and fructose (23.1 g/L) and a high-fat diet (Western Diet, WD) containing 42% kcal from fat and 0.1% cholesterol. A standard chow diet and normal water were given to the control mice.

All animal experiments were conducted following the ethical clearance and approval from the Jagadguru Sri Shivarathreeshwara Academy of Higher Education and the Research Institutional animal ethics committee (JSSAHER/CPT/IAEC/019/2020), JSS AHER, Mysore, Karnataka, India. In this study, male Swiss albino mice (weighting 15–20 g) were selected and separated equally into five groups of six animals as follows: Group 1: control, standard chow diet, and normal water (CDNW); Group 2: western diet (high-fat diet) and sugar water (WDSW); Group 3: WDSW with withaferin A up to 12 weeks (treatment of withaferin A from 8 to 12 weeks); Group 4: WDSW with withaferin A up to 16 weeks (treatment of withaferin A from 8 to 16 weeks); and Group 5: WDSW with withaferin A up to 20 weeks (treatment of withaferin A from 8 to 20 weeks). Every 3 days, the treatment groups received withaferin A (1.25 mg/kg body weight, DMSO 0.1%), while the control group received DMSO (0.1%) intraperitoneally before the dark cycle of each day. To compare the treatment groups, Group 1—CDNW and Group 2—WDSW mice also received diet specifications, as previously mentioned for a duration of 12, 16, and 20 weeks. Briefly, Group 3 mice [mice that received WDSW (until 12 weeks) + withaferin A treatment for 4 weeks (from the 8^th week until the 12^th week)] were compared with the mice group that received CDNW and WDSW for 12 weeks. Group 4 mice [mice that received WDSW (16 weeks) + Withaferin A treatment for 8 weeks (from the 8th week until the 16th week)] were compared with the mice group that received CDNW and WDSW for 16 weeks. Group 5 mice [mice that received WDSW (20 weeks) + Withaferin A treatment for 12 weeks (from the 8th week until the 20^th week)] were compared with the mice group that received CDNW and WDSW for 20 weeks. Furthermore, images of individual groups of mice that received the WDSW + Withaferin A treatment for 12, 16, and 20 weeks along with their corresponding controls and WDSW groups are depicted in Supplementary Figures S6A–C.

Hepatic function enzymes, such as ALT, AST, and ALP, and the lipid profile comprising triglycerides, cholesterol, and HDL were determined using commercially available kits and is based on the associated manuals (Agappe Diagnostics Ltd.). The traditional GOD-POD method was used to measure the level of serum glucose. LDL was calculated using a simple mathematical formula as follows:

LDL = Total cholesterol–HDL–Triglycerides/5

Mice were sacrificed at the following time intervals (12, 16, and 20 weeks) to collect the liver tissue. The tissue samples were immediately formalin-fixed and stored at room temperature for subsequent processing and embedded by a standard technique in paraffin blocks for future usage. H&E and Trichome Masson’s (TMS) staining were performed to assess inflammation and visualize fibrosis, respectively.

The liver tissue was preserved at −80 °C. The TRIzol reagent (Thermo Fisher Scientific) was used to extract total RNA from frozen livers. The verso cDNA synthesis kit was used to create cDNA from 1 μg of total RNA. qPCR was performed on a Rotor-Gene Q (Qiagen) PCR system using the SYBR green kit. These qPCR results were expressed as a fold change relative to the control group, and values were normalized to β-actin mRNAs. The primer sequences used in our experiments were as follows:

TNF-α forward: 5′-ATG​GCC​TCC​CTC​TCA​TCA​GT-3′

TNF-α reverse: 5′-TTT​GCT​ACG​ACG​TGG​GCT​AC-3′

IL-6 forward: 5′-GTC​CTT​CCT​ACC​CCA​ATT​TCC​A-3′

IL-6 reverse: 5′-TAA​CGC​ACT​AGG​TTT​GCC​GA-3′

IL-1β forward: 5′-TGC​CAC​CTT​TTG​ACA​GTG​ATG-3′

IL-1β reverse: 5′-AAG​GTC​CAC​GGG​AAA​GAC​AC-3′

MCP1 forward: 5′-AGG​TGT​CCC​AAA​GAA​GCT​GT-3′

MCP1 reverse: 5′-AAG​ACC​TTA​GGG​CAG​ATG​CAG-3′

COL1A1 forward: 5′-CGA​TGG​ATT​CCC​GTT​CGA​GT-3′

COL1A1 reverse: 5′-GCT​GTA​GGT​GAA​GCG​ACT​GT-3′

COL3A1 forward: 5′-GAG​GAA​TGG​GTG​GCT​ATC​CG-3′

COL3A1 reverse: 5′-TTG​CGT​CCA​TCA​AAG​CCT​CT-3′

α-SMA forward: 5′-GCC​GAG​ATC​TCA​CCG​ACT​AC-3′

α-SMA reverse: 5′-ATA​GGT​GGT​TTC​GTG​GAT​GC-3′

β-actin forward: 5′-TGG​ATC​AGC​AAG​CAG​GAG​TAT​G-3′

β-actin reverse: 5′-GCA​TTT​GCG​GTG​GAC​GAT-3′

An ELISA kit (Krishgen Biosystems) was used to quantify TGF-β1 in the plasma from the study animals following the manufacturer’s instructions. The results are presented as pg/mL after sample values were calculated using a standard curve (created by serially diluting known standards).

Human hepatoma cells, namely, HepG2 and Huh7 cell lines, were cultured in MEM medium and Ham DMEM/F-12, 1:1 mixture, respectively, complemented with 10% penicillin (100 U/mL)/streptomycin (100 mg/mL) antibiotics in a humid incubator with 5% CO₂ at 37 °C.

Stock solutions of sodium palmitate (SP) and oleate (OA) (Sigma-Aldrich, United States) were prepared, as previously described (Römer et al., 2021; Cao et al., 2012). Briefly, 100 μM of SP and OA were incubated for 30 min at 50 °C. Later, fatty acids were mixed with BSA in a culture medium (the fatty acid to BSA molar ratio was 4:1). To induce steatosis, HepG2 and Huh7 cells were exposed to SP and OA conjugated with fatty acid-free BSA. After incubation for 24 h, the cells were treated for 24 h with various concentrations of withaferin A (1, 2.5, and 5 μM). Cells used as controls were treated with fatty acid-free media containing ethanol as a vehicle.

Following treatment, the cells were fixed for 1 h in 10% formalin. The fixative solution was removed and rinsed with PBS once and then treated with 60% isopropanol for 15 s to facilitate the staining of neutral lipids. Cells were then incubated with a 6:4 diluted Oil Red O solution for 1 min and then washed with PBS to remove the excess stain before being counter-stained for 1 min with hematoxylin stain. Then, cells were washed with distilled water to remove excess stains. Pictures of the lipid droplets were taken using an inverted microscope.

Oil Red O staining was measured semi-quantitatively after staining with Hematoxylin, washing with dH₂O, and additionally washing with 60% isopropanol. The extracted Oil Red O stain was then treated with 100% isopropanol and gentle rocking. Then, the red color absorbance was measured at 492 nm.

The binding affinity of withaferin A with the LXR and FXR was assessed using AutoDock Vina (Eberhardt et al., 2021; Trott et al., 2010). The crystal structure of the LXR/FXR complexed with withaferin A was used as the target structure in the docking study. Water molecules were eliminated from the docking study and checked for prior attachment to the ligand (withaferin A) before being removed from the dimensional structure using version 2.4 of the PyMOL tool. Discovery Studio software, which offers near-able binding residues, was used to further visualize the favored possessions from the findings obtained from the docking technique. This verified the selected ligands’ ideal docking postures, and their binding affinities were recorded. Docking poses and score calculations were used to determine the binding affinity of withaferin A as a ligand with LXR and FXR.

The one-way ANOVA (Bonferroni post hoc test) test was used for the data average value calculation of the results and for statistical analysis, where p < 0.05 was considered significant. The significance of different groups was expressed using mean ± SEM values. *p < 0.05, **p < 0.01, and ***p < 0.001

Swiss Albino mice nourished with WDSW showed advancement in weight and adipose tissue mass contrary to mice nourished with a chow diet and regular water (Figures 1A–D). In addition to progression in diet-induced obesity, nourishment with WDSW showed an elevation in blood glucose levels. Another group of mice fed with WDSW was treated with withaferin A for 12–16 weeks, which showed a considerable reduction in the total body weight (including fat mass) compared to WDSW-fed mice. Withaferin A also reduced blood glucose levels in mice with a high-fat diet and sugar water (Figure 1E). Our data are strongly supported by an elegant study from Harvard Medical School, which clearly shows the anti-obesity and anti-diabetic properties of withaferin A through its leptin sensitizing action, and it may also impact the appetite of the animal (Lee et al., 2016).

Liver enzymes AST, ALT, and ALP increased in mice fed WDSW contrary to the control group (Figure 2A). The treatment of withaferin A reduced liver enzyme levels in different time courses. Withaferin A effectively lowered the lipid profile parameters in mice fed WDSW (Figure 2B). These data provide preliminary evidence showing the hepatoprotective effect of withaferin A.

Mice nourished with a chow diet and regular water showed normal liver anatomy and weight (Figures 3A, B). In contrast, mice on WDSW developed major hallmarks of steatohepatitis (grade 3 macrovesicular steatosis, immune cell infiltration, and hepatocellular ballooning along with some microvesicular steatosis), as confirmed by H&E staining (Figure 3C). The withaferin A treatment also inhibited inflammatory markers, such as TNF-α, IL-6, IL-β, and MCP1 expression, in diet-induced obese mice (Figure 3D) compared to the high-fat diet-fed mice in our qPCR data. In vitro steatosis was confirmed by routine and common Oil Red O staining using HepG2 and Huh7 cells (Figures 4A, C). The withaferin A treatment inhibited sodium palmitate- and oleate-induced lipid droplet accumulation in human Huh7 and HepG2 cells, and the graphs depict a quantitative decrease of lipid accumulation in the cells ( Figures 4B, D ). This experimental evidence strongly shows the anti-steatohepatitis effect of withaferin A.

Mice fed a regular low-calorie chow diet and normal water revealed normal liver architecture, whereas mice fed with high-calorie WDSW for 16–20 weeks revealed early fibrotic characteristics such as deposition of collagen and activation of TGF-β signaling. The withaferin A treatment inhibited diet-induced liver fibrosis in WDSW-fed mice (Figure 5A). The withaferin A treatment also inhibited TGF-β secretion and its target genes, Collagen 1 and Collagen 3, expression in WDSW-fed mice and was confirmed by ELISA and qPCR data (Figures 5B, C).

Our study revealed that withaferin A acts as a potent molecular ligand for LXR-α in HCC and inhibits NF-κB target genes (Shiragannavar et al., 2021; Shiragannavar et al., 2022). To determine the possible molecular mechanism of the hepatoprotective nature of withaferin A in the NAFLD model, we conducted a docking study and our results indicated a strong binding of withaferin A to both LXR-α and FXR, which is a known bile acid nuclear receptor (Figures 6A, B). Some of these common genes and signaling pathways are mutually regulated by ligand-dependent activation of both LXR-α and FXR (Ding et al., 2014), (Dong et al., 2019). Both inflammatory and fibrotic signaling is negatively regulated by LXR-α and FXR through their cognate ligands (Rudraiah et al., 2016; Wang et al., 2008). Based on our docking study results and the previously published evidence, we inferred that withaferin A acts as a dual LXR/FXR receptor activator and inhibits diet-induced steatosis, steatohepatitis, and fibrosis. To support our claim, we validated the expression of LXR-α and FXR and their target genes in both in vitro and in vivo NAFLD models treated with withaferin A. Along with withaferin A, we used the LXR-α specific agonist 25-hydroxycholesterol (25HC) and the FXR-specific agonist taurochenodeoxycholate (TCDC) and TGR5 (bile acid membrane receptor)-specific agonist deoxycholic acid (DCA). Our results showed that withaferin A activated both LXR-α and FXR and induced their canonical target genes (ABCA1, ApoE, ABCB11, and ApoCII) (Supplementary Figures S1–S3, S5, S7).