Hepatic Cholesterol Regulation Through Multi-Botanical Extract Targeting of the PCSK9–LDLr–SREBP-2 Axis in HepG2 Cells

Abstract

Background/Objectives: Botanical and nutraceutical approaches have increasingly been considered as alternatives or complements to conventional lipid-lowering therapies, particularly in individuals with mild-to-moderate dyslipidemia or statin intolerance. This study aimed to evaluate a multi-botanical formulation, combining black garlic, sesame, Gastrodia elata, and Primula veris extracts, for its effects on hepatic cholesterol regulation and the PCSK9–LDLr–SREBP-2 axis in vitro. Methods: Each extract was chemically characterised for its polysaccharide, polyphenol, flavonoid, and sesamin content. HepG2 cells were exposed to normal (5 mM) or high-glucose (30 mM) conditions to mimic metabolic stress. Dose–response studies identified optimal concentrations for cell viability. Hepatic safety was assessed via MTT and ROS assays, while cholesterol metabolism was evaluated by measuring HMG-CoA reductase levels, total cholesterol, LDL levels, bile acid production, free cholesterol levels, and the expression of PCSK9, LDLr, and SREBP-2 using ELISA and Western blot. Results: All individual extracts improved cell viability, reduced oxidative stress, and moderately modulated cholesterol metabolism. The multi-botanical combination exhibited synergistic effects, enhancing cell viability (+47.5% vs. untreated), suppressing ROS, reducing HMGR levels, and lowering total intracellular cholesterol more effectively than single extracts or the statin-like reference RYRF. Importantly, the combination strongly downregulated PCSK9 and inhibited SREBP-2 proteolytic activation while upregulating LDLr, indicating coordinated transcriptional and post-translational regulation. Bile acid production and free cholesterol excretion were also significantly increased, supporting improved cholesterol clearance. Conclusions: This four-botanical formulation effectively modulates hepatic cholesterol homeostasis via a multifactorial, synergistic mechanism distinct from statin-like agents. The results suggest its potential as a safe, non-statin strategy to support cardiometabolic health. Future studies are warranted to confirm long-term efficacy and clinical relevance.

1. Introduction

One significant modifiable risk factor for cardiovascular disease is hypercholesterolemia, which also contributes to atherosclerosis, hepatic disease, and associated clinical consequences [1]. Despite statins’ proven effectiveness as first-line therapy, there is growing interest in complementary strategies for patients with intolerance, poor adherence, or mild to moderate risk levels requiring non-pharmacological preventive approaches. Elevated low-density lipoprotein (LDL) cholesterol (LDL-C) is linked to an increased risk of major cardiovascular events [2]. At the molecular level, cholesterol homeostasis is controlled by important regulators such as 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGR), LDL receptor (LDLr), proprotein convertase subtilisin/kexin type 9 (PCSK9), and transcription factors such as sterol regulatory element-binding protein (SREBP), as well as hepatic production, catabolism, and bile acid-mediated excretion [3,4]. Modulating these pathways, increasing LDLr activity, decreasing PCSK9, and encouraging cholesterol conversion to bile acids can enhance LDL clearance and lipid homeostasis [5,6]. In vitro models, particularly HepG2 cells, are commonly used to study these pathways and evaluate the effects of nutraceuticals on lipid metabolism, oxidative stress, and inflammation [7,8]. Compared with conventional medication therapy, nutraceuticals offer a therapeutic option that can act through a variety of pathways and have superior safety profiles [9,10].

The discovery of monacolin K, a molecule chemically equivalent to lovastatin and capable of inhibiting HMGR, has made fermented red rice (RYRF) one of the most researched nutraceuticals for cholesterol management in recent years [11]. Due to its statin-like activity, potential hepatic and muscular effects have been reported at doses of approximately 3 mg/day, and EU authorities are discussing appropriate use conditions and intake limits for monacolins in supplements [12,13]. The evaluation of the benefit/risk profile of RYRF-based products is further complicated by the wide variability in monacolin levels reported in commercial supplements, sometimes differing by orders of magnitude between products and by quality issues, including the presence of impurities such as citrinin, a nephrotoxic mycotoxin that has been identified in certain red yeast rice formulations [14,15,16]. Research into nutraceutical substitutes that lack direct statin-like action but can modulate lipid metabolism through complementary mechanisms has been spurred by these safety and regulatory constraints [17].

Among them, plant extracts with various phytochemical profiles that might affect hepatic redox and inflammatory homeostasis, bile acid synthesis, intestinal lipid absorption, PCSK9, and LDLr modulation are particularly interesting [6,18]. Black garlic, obtained through controlled fermentation of fresh garlic, has emerged as a promising nutraceutical due to its enriched content of bioavailable organosulfur compounds, particularly S-allyl-cysteine (SAC) [19]. Preclinical and clinical evidence indicate that black garlic favourably modulates plasma lipid profiles, hepatic redox balance, and inflammatory pathways implicated in atherosclerosis [20,21]. Beyond these pleiotropic effects, in vitro studies in HepG2 cells suggest that garlic-derived organosulfur compounds can influence key regulators of cholesterol homeostasis, including downregulation of PCSK9 expression and upregulation of LDLr activity, thereby enhancing hepatic LDL clearance [20,22].

Other plant extracts that have historically been employed in herbal medicine are garnering attention for their potential role in lipid homeostasis, within the broader context of nutraceuticals that operate through pleiotropic, non-statin-like mechanisms. East Asian traditional medicine extensively uses Gastrodia elata, which contains phenolic compounds such as gastrodin, a hydroxybenzyl alcohol β-D-glucosid, with metabolic, anti-inflammatory, and antioxidant properties. Experimental studies indicate that gastrodin may indirectly modulate hepatic cholesterol metabolism by reducing oxidative stress and by regulating key mechanisms involved in LDL receptor expression and PCSK9 signalling. However, information on lipid metabolism remains mostly preclinical [23,24].

Similarly, sesame (Sesamum indicum) seeds, which contain lignans such as sesamin and sesamolin, characterised by a dibenzylbutane skeleton with multiple methylenedioxy-substituted phenolic rings, have been shown to directly influence hepatic lipid metabolism in in vitro HepG2 models [25]. Sesamin has been demonstrated to decrease intracellular lipid accumulation and modify the expression of genes involved in fatty acid β-oxidation, lipid uptake, and cholesterol homeostasis, including PCSK9/LDLr and SREBP-1/AMP-activated protein kinase (AMPK)-related pathways [26].

In addition, cowslip (Primula veris), traditionally recognised for its anti-inflammatory and antioxidant properties, contains flavonoids with a polyphenolic C6–C3–C6 backbone (e.g., apigenin, quercetin, kaempferol, and rutin) and triterpene saponins, composed of a pentacyclic triterpene aglycone linked to one or more sugar moieties [27,28] that may contribute to lipid metabolism regulation through modulation of redox balance, inflammatory tone, as well as metabolic mechanisms relevant to the control of metabolic syndrome [29]. While direct evidence from HepG2 lipid models remains limited, its phytochemical profile supports a complementary role in a multifactorial nutraceutical strategy aimed at improving cardiometabolic health [30,31].

Taken together, despite the growing body of evidence supporting the lipid-modulating effects of individual botanicals, direct comparative studies evaluating their effects on key cholesterol homeostasis regulators, particularly PCSK9 and LDLr, within a unified in vitro model are still lacking. Moreover, the potential synergistic or complementary effects of combining non-statin-like nutraceuticals remain largely unexplored.

The aim of this in vitro study was to compare the hepatoprotective and cholesterol-lowering effects of black garlic, Gastrodia elata, sesame, and Primula veris, administered alone or in combination, compared to a RYRF extract, in HepG2 cells under hypercholesterolemic conditions. Specifically, their impact on key regulators of cholesterol homeostasis, including LDLr, PCSK9, SREBP-1, and HMGR, as well as their effects on intracellular lipid accumulation, were investigated. In parallel, the antioxidant properties of the extracts were evaluated to elucidate their contribution to hepatocellular protection. Overall, this study aims to clarify the molecular mechanisms underlying the nutraceutical potential of these botanicals and to highlight their complementary, multifactorial role in modulating lipid metabolism in an in vitro model. The formulation was selected based on a mechanism-driven rationale, combining extracts with complementary actions on cholesterol synthesis, elimination, and oxidative stress pathways. This approach aims to enhance efficacy while mitigating compensatory responses observed with single-agent interventions.

2. Materials and Methods

2.1. Characterisation of Botanical Extracts

2.1.1. Determination of Total Polyphenol Content

Total polyphenol content (TPC) was determined using the Folin–Ciocalteu colourimetric method, following established literature procedures with minor modifications [32]. Briefly, an aliquot of the sample solution (botanical extracts of Gastrodia elata, black garlic, and Primula veris, supplied by Nutra Futura Srl (Legnano, Italy) (see the Section 2.2 for details) was mixed with the Folin–Ciocalteu reagent and allowed to react for 3–5 min at room temperature. Subsequently, an aqueous sodium carbonate solution was added to alkalinise the reaction medium. The mixture was incubated at room temperature for 30–60 min in the dark to allow colour development.

Absorbance was measured at 760 nm using a UV–Vis spectrophotometer (Infinite 200 Pro MPlex plate reader, Tecan, Männedorf, Switzerland). Gallic acid was used as the reference standard, and a calibration curve was constructed using appropriate concentrations. The experiments were performed at least in triplicate, and the results were expressed as mean (%) ± standard deviation (SD).

2.1.2. Determination of Total Polysaccharide with Phenol–Sulfuric Acid

Total polysaccharide content was determined using the phenol–sulfuric acid colourimetric method [33]. Briefly, an aliquot of the Gastrodia elata extract was accurately dissolved in purified water and appropriately diluted to fall within the linear range of the assay. A volume of 1.0 mL of the sample solution was mixed with 1.0 mL of 5% (w/v) phenol solution, followed by the rapid addition of 5.0 mL of concentrated sulfuric acid along the wall of the test tube to ensure thorough mixing. The reaction mixture was allowed to stand at room temperature for 30 min to allow complete colour development. Absorbance was measured at 490 nm against a reagent blank using a UV–Vis spectrophotometer. Quantification was performed using a calibration curve constructed with D-glucose as the reference standard, and total polysaccharide content was expressed as mean (%) ± SD. All measurements were carried out in triplicate.

2.1.3. SAC Analyses Employing LC–MS/MS

SAC in black garlic extract samples was determined following the method reported in the literature [34]. SAC was analysed using a high-performance liquid chromatography coupled with tandem mass spectrometry (HPLC–MS/MS) system (TSQ Quantum Access Max, Thermo Scientific, Waltham, MA, USA) equipped with an auto-sampler and a nitrogen gas generator. MS operating conditions, including the electrospray ionisation source (positive mode), scan rate (0.5 spectra/s), collision energy (10 eV), and selected reaction monitoring (precursor m/z: 162; product m/z: 73, 145), were previously established. For chromatographic separation, 5 μL of each sample was injected onto a Zorbax^® C18 column (250 mm × 4.6 mm, 5 μm, Agilent Technologies Inc., Santa Clara, CA, USA). A linear gradient of mobile phase A (0.1% formic acid in water) and B (acetonitrile) was used: 0–7 min, 20% B; 7–14 min, 100% B; 14–17 min, 100% B; 17–25 min, 1% B. SAC quantification was performed using the Xcalibur 2.1 software (Thermo Scientific). The regression analysis demonstrated linear behaviour within the selected range (R² ≥ 0.99). The results were reported as mean (%) ± SD.

2.1.4. Determination of Total Flavonoid Content

A previously established colourimetric approach was used to determine the total flavonoid content (TFC) [35]. A 10 mL volumetric flask was filled with 1 mL of the Primula veris, sesame extract sample (containing 0.1 mg/mL dry material), and 4 mL of water. First, 3 millilitres of 5% NaNO₂ solution were added, followed by 0.3 millilitres of 10% AlCl₃ after 5 min and 2 millilitres of 1 M NaOH after 6 min. Distilled water was added to the flask until it reached the calibration mark. After mixing the solution, its absorbance at 510 nm was measured. The results were reported as mean (%) ± SD.

2.1.5. Determination of Sesamin in Sesame Extract

Sesamin in sesame extract was determined using HPLC–MS/MS, following a modified procedure based on literature methods [36]. An aliquot of the extract (5 μL) was injected onto a Zorbax^® C18 column (250 mm × 4.6 mm, 5 μm, Agilent Technologies Inc., Santa Clara, CA, USA). Chromatographic separation was achieved using a linear gradient of mobile phase A (0.1% formic acid in water) and B (acetonitrile): 0–5 min, 30% B; 5–15 min, 70% B; 15–20 min, 100% B; and 20–25 min, 30% B.

Detection was carried out with a triple quadrupole mass spectrometer (TSQ Quantum Access Max, Thermo Scientific, Waltham, MA, USA) using electrospray ionisation in positive mode. MS conditions were: scan rate 0.5 spectra/s, collision energy 15 eV, and selected reaction monitoring (SRM) transitions with precursor m/z 353 and product m/z 161, 175. Quantification was performed using the Xcalibur 2.1 software (Thermo Scientific), with calibration curves prepared from authentic sesamin standards, and the results were reported as mean (%) ± SD.

2.2. Agent’ Preparation

Specifically, black garlic extract (0.66% SAC), SesameQ extract (Q-antek seed extract; 10% sesamine, named Sesame), Primula veris extract (4:1 leaf extract), Gastrodia elata extract (rhizoma extract; 10.9% polysaccharides), and RYRF extract (Oryza sativa L. extract; 5.2% monacolines) were freshly prepared.

Before performing each stimulation, all the botanicals were dissolved in Dulbecco’s Modified Eagle’s Medium without phenol red (DMEM, Merck Life Science, Rome, Italy) supplemented with 0% foetal bovine serum (FBS), 50 IU/mL penicillin-streptomycin (Merck Life Science, Rome, Italy), and 2 mM L-glutamine solution (Merck Life Science, Rome, Italy).

A dose–response study was used to evaluate the various concentrations chosen from scientific literature based on similar previous studies [37,38,39,40,41,42,43] for the extract under investigation at the level of the hepatic cellular model, except for RYRF, whose concentration in vitro (1 µM = 8 µg/mL) was selected from a previous in vitro study on a similar extract [44].

The dose–response study examines the following in detail: black garlic extract, concentrations of 1 μg/mL, 5 μg/mL, 10 μg/mL, 25 μg/mL, and 50 μg/mL; sesame extract, 9 μg/mL, 18 μg/mL, 27 μg/mL, 36 μg/mL, and 45 μg/mL; Primula veris (indicated as Primula in the Section 3 and graphs), 25 μg/mL, 50 μg/mL, 100 μg/mL, 125 μg/mL, and 150 μg/mL; and Gastrodia elata, 25 μg/mL, 50 μg/mL, 100 μg/mL, 150 μg/mL, 200 μg/mL, and 250 μg/mL.

To ensure consistency across experiments, the proportion of each extract in the combination (called MIX) was determined by its selected concentration (μg/mL) identified in preliminary dose–response assays.

2.3. Cell Cultures

HepG2 cells (ATCC, Manassas, VA, USA), derived from human hepatocellular carcinoma, were maintained under standard culture conditions (37 °C, 5% CO₂) [45]. The cells were grown in DMEM supplemented with 10% FBS, 2 mM L-glutamine, and 1% penicillin-streptomycin (Merck Life Science). To ensure experimental consistency, cultures at 90–95% confluence were used [46]. For dose–response experiments, 1 × 10⁴ cells were seeded in 96-well plates; for other analyses, 3.5 × 10⁴ cells were seeded in 24-well plates equipped with 6.5 mm Transwell^® inserts, and the formation of a mature monolayer was confirmed by measuring trans-epithelial electrical resistance (TEER), which reached 486 Ω·cm² [47]. Throughout the experiments, the culture medium was replaced with fresh medium containing a high glucose concentration (30 mM; Merck Life Science, Milan, Italy). Control cells were maintained either in standard glucose conditions or in the same high-glucose medium (30 mM) to serve as untreated controls [44,45].

2.4. Experimental Protocol

The study was structured in three main phases. In the first phase, a preliminary chemical characterisation of each individual botanical extract was performed. Botanical extract was chemically characterised for its polysaccharide, polyphenol, SAC, flavonoid, and sesamin content.

After that, in the second phase, the preliminary dose–response analysis was performed to identify non-cytotoxic concentrations of the tested natural extracts (black garlic, sesame extract, Primula veris, and Gastrodia elata) in HepG2 cells. The cells were synchronised overnight in DMEM without phenol red and FBS, supplemented with penicillin/streptomycin, L-glutamine, and sodium pyruvate, and maintained in normal glucose medium (5 mM); they served as controls [44]. Cell viability was assessed through the MTT assay to select appropriate concentrations for subsequent experiments. These ratios were also taken as a reference for integrative chemical analyses (LC-MS/MS) of the MIX characterisation (Appendix A).

In the third phase, HepG2 cells were exposed to the high-glucose medium (30 mM) to mimic hyperglycaemic conditions [45]. Then, they were stimulated with the selected concentrations of each botanical extract, individually and in combination (MIX), for 24 h in comparison with RYRF 8 µg/mL. The proportion (µg/mL) of each extract in the MIX was defined according to their individually selected effective concentrations on phase two. This phase evaluated the samples’ biological effects on cholesterol metabolism, including total and free cholesterol levels, LDL determination, and bile acid production through specific assay kits while excluding any cytotoxic effects by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay and reactive oxygen species (ROS) production analysis). Finally, the main intracellular pathways regulating cholesterol homeostasis were investigated, with a focus on those involved in cholesterol synthesis, uptake, and regulation. Specifically, in HepG2 cells under the same experimental conditions as in the second phase, key proteins were quantified using the ELISA kits: HMGR, PCSK9, SREBP-2, and LDLr. In addition, Western blot analyses were conducted to confirm LDL receptor protein expression in HepG2 lysates.

2.5. Cell Viability (MTT Assay)

The MTT assay was conducted according to a standard procedure [48] to assess cytotoxicity. In brief, following stimulation, both cell types were incubated with 1% MTT solution (Merck Life Science, Milan, Italy) in DMEM for 2 h at 37 °C. The resulting purple formazan crystals were then dissolved using an equal volume of MTT solubilization solution. Cell viability was measured using an Infinite 200 Pro MPlex plate reader (Tecan, Männedorf, Switzerland) by recording absorbance at 570 nm with a 690 nm reference correction. Results are expressed as mean (%) ± SD, relative to untreated control cells (0% line), from five independent experiments, each performed in triplicate.

2.6. ROS Production

Superoxide anion production was assessed using a standard method [49] by measuring the reduction in cytochrome C in culture supernatants at 550 nm with an Infinite 200 Pro MPlex plate reader (Tecan, Männedorf, Switzerland). The rate of O₂⁻ release was expressed as the mean ± (SD) of nanomoles of reduced cytochrome C per microgram of protein as percentage relative to the untreated control (0% line), from five independent experiments, each performed in triplicate.

2.7. HMGR ELISA Kit

The HMGR levels were quantified using the HMGR ELISA Kit (LSBio, Seattle, DC, USA), following the manufacturer’s instructions [50]. After stimulation, the cells were collected, lysed by three freeze–thaw cycles, and centrifuged at 1500× g for 10 min at 2–8 °C. The samples (100 µL) were incubated with 100 µL of biotinylated detection antibody for 60 min at 37 °C, followed by 100 µL of HRP-streptavidin conjugate for 30 min at 37 °C. Then, 90 µL of TMB substrate was added for 15 min at 37 °C in the dark, and the reaction was stopped with 50 µL of Stop Solution. Absorbance was measured at 450 nm using an Infinite 200 Pro MPlex plate reader (Tecan, Männedorf, Switzerland), and concentrations were calculated from a standard curve (0.625–40 ng/mL). Results are expressed as mean ± SD reported in ng/mL from five independent experiments, each performed in triplicate.

2.8. Determination of Total and Free Cholesterol Levels

The cholesterol levels were measured using a Cholesterol Quantitation Kit (Merck Life Science, Milan, Italy) according to the manufacturer’s instructions [51]. After stimulation, the cells were extracted with chloroform:isopropanol:IGEPAL CA-630 (7:11:0.1) and centrifuged at 13,000× g for 10 min; the supernatants were dried at 50 °C. Lipids were resuspended in cholesterol assay buffer with reaction mix and incubated for 60 min at 37 °C in the dark. Absorbance at 570 nm was measured using an Infinite 200 Pro MPlex plate reader (Tecan, Männedorf, Switzerland). The results are expressed as mean (%) ± SD, normalised to protein content (µg/µL), relative to untreated control cells (0% line), from five independent experiments, each performed in triplicate.

2.9. LDL Determination Quantification

The LDL levels were determined using LDLC colourimetric assay kits (cholesterol oxidase/phenol aminophenazone method, Elabscience, Wuhan, China), following the manufacturer’s instructions [44,52]. Absorbance at 546 nm was measured using an Infinite 200 Pro MPlex plate reader (Tecan, Männedorf, Switzerland). The results are expressed as mean (%) ± SD, normalised to protein content (µg/µL), relative to untreated control cells (0% line), from five independent experiments, each performed in triplicate.

2.10. Bile Acid Production Assay

Bile acid production by HepG2 cells after stimulation was measured using a fluorometric total bile acid assay, following established protocols [53]. Briefly, the cells were washed with cold phosphate-buffered saline (PBS), lysed by sonication in cold 1× PBS (Merck Life Science, Milan, Italy), and centrifuged at 10,000× g for 10 min at 4 °C. Bile acids were quantified based on the fluorescence of resorufin, measured at an excitation wavelength of 560 nm and an emission wavelength of 590 nm using an Infinite 200 Pro MPlex plate reader (Tecan, Männedorf, Switzerland). The results are expressed as mean (%) ± SD, normalised to protein content (µg/µL), relative to untreated control cells (0% line), from five independent experiments, each performed in triplicate.

2.11. PCSK9 Levels (ELISA Kit)

The PCSK9 levels in HepG2 culture supernatants were quantified using an ELISA kit (BioVision, Milpitas, CA, USA; Elabscience Biotechnology, Wuhan, China) according to the manufacturer’s instructions [7]. Absorbance at 450 nm was measured using an Infinite 200 Pro MPlex plate reader (Tecan, Männedorf, Switzerland). PCSK9 concentrations were determined from a standard curve (0–10 pg/mL). Results are expressed as mean (%) ± SD, relative to untreated control cells (0% line), from five independent experiments, each performed in triplicate.

2.12. SREBP-2 Levels (ELISA Kit)

Following the stimulation, the cells were lysed, and SREBP-2 levels were measured using an ELISA kit (LSBio, Seattle, DC, USA) according to the manufacturer’s instructions [50]. Absorbance at 450 nm was measured using an Infinite 200 Pro MPlex plate reader (Tecan, Männedorf, Switzerland), and concentrations were calculated from a standard curve (0.312–20 ng/mL) and expressed as ng/µL. The results are expressed as mean (%) ± SD, relative to untreated control cells (0% line), from five independent experiments, each performed in triplicate.

2.13. LDLr Levels (ELISA Kit)

The LDLr levels in HepG2 culture supernatants were quantified using a commercially available ELISA kit (Human LDLR Quantikine ELISA Kit, R&D Systems, Bio-Techne, Minneapolis, MN, USA) according to the manufacturer’s instructions [54]. Absorbance at 450 nm was measured using an Infinite 200 Pro MPlex plate reader (Tecan, Männedorf, Switzerland). LDLr concentrations were determined from a standard curve (range 62.5–4000 pg/mL) and expressed as pg/mL. Results are expressed as mean (%) ± SD, relative to untreated control cells (0% line), from five independent experiments, each performed in triplicate.

2.14. Western Blot

The HepG2 cells were lysed on ice using Complete Tablet Buffer (Roche, Basel, Switzerland) supplemented with 1 mM phenylmethanesulfonylfluoride (PMSF), 2 mM sodium orthovanadate (Na₃VO₄), phosphatase inhibitor cocktail (1:50), and protease inhibitor cocktail (1:200), following standard procedures [55]. Protein extracts (35 µg) were separated by 8% SDS-PAGE and transferred onto PVDF membranes (GE Healthcare Europe GmbH, Milan, Italy). Membranes were incubated overnight at 4 °C with primary antibodies (all 1:500, mouse monoclonal, Santa Cruz Biotechnology, Santa Cruz, CA, USA) against LDLr (Cat. No. sc-18823), HMGR (Cat. No sc-271595), and SREBP-2 precursor-full-length and cleaved forms (Cat. No sc-271615). Secondary antibodies used were HRP-conjugated anti-mouse IgG (1:5000; Cat. No A9044, Merck Life Science, Rome, Italy). Protein loading was verified using an anti-β-actin antibody (1:4000; Cat. No. A5441, Merck Life Science, Rome, Italy). Results are expressed as mean (%) ± SD, relative to untreated control cells (0% line). Raw and uncropped Western blot images are provided in Supplementary Materials (Figures S1–S5).

2.15. Statistical Analysis

All experiments were performed independently at least five times, with each experiment conducted in triplicate. All experimental data, including colorimetric/functional assays and Western blot analyses, were normalised to the untreated control, which was set to 0% and presented as mean ± SD (except for HMGR data reported as ng/mL). For Western blot experiments, densitometric band intensities were first normalised to the corresponding loading control (β-actin) before calculating percentage variation relative to the control.

$$\mathbf{\%} \mathrm{variation} =({\displaystyle \frac{\mathrm{normalized} \mathrm{OD} \mathrm{treated}}{ \mathrm{normalized} \mathrm{OD} \mathrm{control}}}-1)\times 100$$

The fold changes reported in the Section 3 were calculated as the ratio between the value measured in the treated condition and the reference value. References included the untreated control, the mean response of all botanical extracts, or a specific comparator (e.g., RYRF at 8 μg/mL). Fold change was defined as:

Fold change = Treated value/Reference value

Fold changes were expressed as percentage variation relative to the reference according to the following formula:

Percentage change = [(Treated − Reference)/Reference] × 100

Normality of all datasets was assessed using the Shapiro–Wilk test and visually confirmed with Q–Q plots. Although some variables exhibited relatively high dispersion, variances between groups were comparable, and one-way ANOVA is generally robust to moderate deviations from normality in balanced designs. Statistical comparisons were performed using one-way ANOVA followed by Bonferroni’s post hoc test, with p < 0.05 considered statistically significant. For variables with higher dispersion, complementary non-parametric analyses (Kruskal–Wallis followed by Dunn’s post hoc test) were also conducted, confirming the robustness of the results. Analyses were carried out using GraphPad Prism 10.2.3 (GraphPad Software, Inc., San Diego, CA, USA). In specific parameters, the synergistic effect was evaluated post hoc using the Bliss independence model. Expected additive effects were calculated from single agent responses and compared with the observed combined effect.

3. Results

3.1. Chemical Characterisation of the Extracts

Before proceeding with biological assays, all extracts were subjected to preliminary chemical characterisation to determine their main bioactive compounds (

Table 1. Total polyphenols, flavonoids, polysaccharides, S-allyl-L-cysteine (SAC), and sesamin were quantified in Gastrodia elata (rhizome extract), black garlic extract, Primula (leaf extract 4:1), and sesame (seed extract) using colourimetric methods and LC–MS/MS. Data are expressed as mean ± SD (n = 3) or as indicated from literature titrations. The table highlights the differences in bioactive compound content among the four extracts, providing a basis for their subsequent biological evaluations.

Natural Extract	Methods	Main Bioactive Compounds	Content (%)
Gastrodia elata	UV-Vis	Polysaccharides	10.9 ± 0.3
	UV-Vis	Polyphenols	3.2 ± 0.1
Black garlic	HPLC	SAC	0.66 ± 0.02
	UV-Vis	Polyphenols	5.1 ± 0.2
Primula veris	UV-Vis	Polyphenols	2.4 ± 0.1
	UV-Vis	Flavonoids	1.2 ± 0.05
Sesame	HPLC	Sesamin	10 ± 0.25
	UV-Vis	Flavonoids	1.5 ± 0.06

). Specifically, total polyphenol content was evaluated using the Folin–Ciocalteu colourimetric method, while the content of polysaccharides, flavonoids, SAC, and sesamin was determined using specific colourimetric methods or LC–MS/MS. In this context, Gastrodia elata (rhizome extract) showed a high polysaccharide content (10.9%) and a polyphenol content of 3.2%, consistent with literature values, confirming its relevance as a source of neuroprotective compounds. Similarly, black garlic extract exhibited a SAC content of 0.66% and a polyphenol content of 5.1%, in agreement with previously reported data, indicating both antioxidant potential and the presence of cardiovascular-functional compounds. Moreover, Primula (leaf extract 4:1) presented polyphenols at 2.4% and flavonoids at 1.2%, suggesting antioxidant potential. Finally, sesame contained 10% sesamin and 1.5% flavonoids, confirming the presence of lignans and bioactive compounds with antioxidant and lipid metabolism-modulating properties. Taken together, these results provide a comprehensive overview of the chemical composition of the four extracts, allowing the correlation of the presence of specific bioactive compounds with subsequent biological evaluations.

3.2. Dose–Response Study of Botanical Extracts in HepG2 Cell Line

The initial phase of the investigation was to identify the optimal concentration of each extract associated with the best biological effect, as measured by viability in HepG2. In particular, the cells were kept at a normal glucose concentration (D-glucose 5 mM) as specified in the experimental literature [56]. Before being stimulated with the samples, they were cultured in DMEM without red phenol. This was done to synchronise them to the same stage of the cell cycle. Each extract was investigated at five different concentrations. For each concentration under consideration, HepG2 cells were treated for 24 h, and cell viability was then determined using the MTT test (Figure 1).

As shown in the graphs, for each sample, a concentration was identified that yielded the best maintenance and an increase in viability compared to the other tests and the untreated control (indicated as control; line 0%). Regarding the black garlic extract investigated (Figure 1A), it exhibited a clear dose-dependent effect on cell viability. Black garlic 50 μg/mL yielded the most significant results compared to the other concentrations (p < 0.05), with an increase of 14% compared to the untreated control (p < 0.05). In contrast, the extracts of sesame (10%), Primula, and Gastrodia elata (Figure 1B–D) exhibited different biological behaviour, characterised by the identification of an optimal concentration associated with the highest statistically significant increase in cell viability compared with other tested doses (p < 0.05), followed by a decline at higher concentrations. Specifically, for sesame (10%), the concentration of 18 μg/mL produced the most favourable effect, resulting in a 21% increase in cell viability relative to the untreated control (p < 0.05). For Primula, the optimal concentration was 50 μg/mL, corresponding to a 20.3% increase compared with the untreated control (p < 0.05). Finally, Gastrodia elata showed the most significant enhancement of cell viability at 100 μg/mL, with an increase of 14.4% compared with the untreated control (p < 0.05).

Based on the results obtained, the concentrations of black garlic 50 μg/mL, Sesame (10%) 18 μg/mL, Primula 50 μg/mL, and Gastrodia elata 100 μg/mL were selected for subsequent experimental phases at the liver level in vitro. Following the selected concentrations, an integrative chemical analysis (LC-MS/MS) of the MIX was performed (Appendix A;

Table A1. LC–MS/MS analysis of the main bioactive compounds in the MIX prepared according to the selected experimental ratios of the four extracts after the dose–response screening step. Polyphenols and flavonoids are reported as cumulative contributions of their respective source extracts.

Bioactive Compound	Content in MIX (%)	Analytical Method	Source Extract(s)
SAC	0.15 ± 0.01	LC–MS/MS	Black garlic
Polysaccharides	5.0 ± 0.25	LC–MS/MS	Gastrodia elata
Polyphenols	3.2 ± 0.16	LC–MS/MS	Black garlic, Gastrodia elata, Primula veris
Flavonoids	0.4 ± 0.03	LC–MS/MS	Primula veris, Sesame
Sesamin	0.8 ± 0.05	LC–MS/MS	Sesame

). The analysis confirmed that the MIX contains the expected amounts of each bioactive compound, fully consistent with the weighted contribution of the individual extracts, with no evidence of significant quantitative changes upon mixing.

3.3. Hepatic Safety Profile of the Botanical Extract Combination

In this part, the study examined the biological effects of the combination of four botanical extracts (MIX) and compared these effects with RYRF at 8 μg/mL as a separate comparator. Initially, the safety profile (Figure 2) was thoroughly examined with respect to cell viability (MTT assay) and oxidative stress (cytochrome C colourimetric assay). HepG2 cells were exposed to high glucose concentrations (30 mM D-glucose) to promote an increase in intracellular cholesterol levels. At this point, all data were compared to untreated cells in 30 mM D-glucose conditions (high-glucose control; 0% line on graph).

As shown in Figure 2A, all individual extracts exhibited excellent biological effects, increasing cell viability compared to the untreated control (p < 0.05). At the same time, they outlined comparable (black garlic 50 μg/mL) or better results (Primula 50 μg/mL; Sesame (10%) 18 μg/mL and Gastrodia elata 100 μg/mL) compared to the reference RYRF 8 μg/mL extract, and in some cases, as for Primula 50 μg/mL; sesame (10%) 18 μg/mL statistically significant (p < 0.05 vs. RYRF 8 μg/mL). The MIX of the four extracts produced markedly stronger effects than both the individual extracts and RYRF at 8 μg/mL (p < 0.05), preliminarily suggesting a potential synergistic interaction exceeding the additive effects of the single components. Specifically, MIX increased cell viability by 47.5% compared with the untreated control, by 84% compared with black garlic 50 μg/mL, by 78% compared with Gastrodia elata 100 μg/mL, by 67% compared with Primula, by 66% compared with sesame (10%) 18 μg/mL, and by 84% compared with RYRF 8 μg/mL.

At the same time, in terms of oxidative stress (Figure 2B), all test samples helped maintain ROS production levels below those of the untreated control (p < 0.05, except for RYRF at 8 μg/mL). In detail, all individual botanical extracts showed better pro-antioxidant effects than RYRF at 0.1 μg/mL alone, with statistically significant results (p < 0.05, except for black garlic 50 μg/mL). MIX promoted an increase in biological data, as highlighted by individual agents, by amplifying the antioxidant action of the extracts (p < 0.05). MIX promoted an improvement in the reduction in ROS production by 14.9% compared with the untreated control, by 63% compared with black garlic 50 μg/mL, by 56% compared with Gastrodia elata 100 μg/mL, by 42% compared with Primula, by 57% compared with sesame (10%) 18 μg/mL, and by 87% compared with RYRF 8 μg/mL.

3.4. Biological Effects of the Botanical Extract Combination on Cholesterol Homeostasis

To investigate the variation in cholesterol homeostasis under hyperglycemic conditions, HepG2 cells were maintained under the same high-glucose conditions as previously described to induce metabolic changes in vitro. In this regard, the potential bioactivity of botanical extracts and their combinations was investigated with respect to key hepatic cholesterol metabolism parameters, including HMGR levels, total cholesterol, and LDL (Figure 3 and Appendix A).

Figure 3A shows that 24 h treatment with single botanical extracts and RYRF at 8 μg/mL significantly reduced HMGR protein levels (ng/mL) compared to the untreated control (p < 0.05). Each extract performed worse than RYRF 8 μg/mL alone. The most bioactive effects on HMGR were observed following treatment with Primula at 50 μg/mL and sesame (10%) 18 μg/mL. MIX significantly reduced HMGR activity compared to individual extracts and RYRF 8 μg/mL (p < 0.05), indicating effective suppression of the cholesterol synthesis pathway under hyperglycemic conditions and a potential synergistic effect. MIX promoted an improvement in the reduction in HMGR levels by 67.6% compared with the untreated control, by 84% compared with black garlic 50 μg/mL, by 86% compared with Gastrodia elata 100 μg/mL, by 79% compared with Primula 50 μg/mL and sesame (10%) 18 μg/mL, and by 64% compared with RYRF 8 μg/mL. A post hoc Bliss independence analysis, reported in Appendix A (

Table A2. Bliss independence analysis on HMGR levels reduction (ELISA kit).

Treatment	Expected Data	Observed Data	Interaction
MIX	−48.83%	−67.64%	Synergistic

), associated with the observed combined effects, exceeded the expected additive responses.

Figure 3B provides further information on the measurement of normalised HMGR protein levels following a 24 h treatment with RYRF, single botanical extracts, and MIX in high-glucose conditions. All treatments induced a significant reduction in the HMGR/β-actin ratio compared to the untreated high-glucose control (p < 0.05). RYRF reduced the ratio by approximately 12% with better reduction data than all individual extractts (reduction between 6 and 9% vs. control high glucose). The most substantial reduction was observed with the MIX treatment, which reduced the HMGR/β-actin ratio by 31.3%, indicating a possible synergistic effect of the combined extracts in downregulating HMGR protein expression. The post hoc Bliss independence analysis, reported in Appendix A (

Table A3. Bliss independence analysis on HMGR expression reduction (Western blot).

Treatment	Expected Data	Observed Data	Interaction
MIX	−23.32%	−31.30%	Synergistic

), showed that the observed combined effects exceeded the expected additive responses.

Consistent with previous findings, Figure 3C shows the effects of the treatments on the total liver cholesterol levels. While individual extracts and RYRF produced a moderate but significant decrease in total intracellular cholesterol compared to the high-glucose control (p < 0.05), MIX resulted in a substantially stronger reduction, highlighting its superior ability to counteract cholesterol accumulation in HepG2 cells exposed to high glucose. MIX reduced total cholesterol in HepG2 by 7.3-fold compared with the untreated control, by 4-fold compared with all botanical extracts, and by 1.6-fold compared with RYRF 8 μg/mL.

Additional supportive data on intracellular cholesterol content measured by enzymatic/colourimetric assay are provided in Appendix A.1 and Appendix A.2 (

Table A4. Bliss independence analysis on LDL content.

Treatment	Expected Data	Observed Data	Interaction
MIX	−16.84%	−36.77%	Synergistic

and Figure A1) and are not intended to reflect LDL particle uptake or receptor-mediated internalisation. This assay indicates total intracellular cholesterol accumulation rather than intact lipoprotein particle internalisation.

Within these limitations, individual botanical extracts showed a trend towards reducing intracellular cholesterol content compared to RYRF at 8 μg/mL, with Gastrodia elata displaying the most consistent effect relative to untreated controls. The combined formulation (MIX) demonstrated a more significant reduction, resulting in a decrease of 36.7% compared with the untreated control, 88% compared with the mean of the individual botanical extracts, and 96% compared with RYRF at 8 μg/mL. Detailed quantitative data and Bliss independence analysis are presented in Appendix A (Table A4).

Further analyses were also conducted on key parameters of cholesterol excretion, again in the context of cholesterol metabolism and homeostasis at the liver level. In detail, bile production in terms of bile acids was examined in addition to the determination of free cholesterol (Figure 4).

Regarding bile production (Figure 4A), all treatments significantly increased bile production compared to the untreated control (p < 0.05). In particular, the individual botanical and RYRF 8 μg/mL extracts have very overlapping effects with lower data from black garlic 50 μg/mL. MIX resulted in a significantly greater increase in bile production than all other treatments (p < 0.05). The mean increase in bile production by MIX was 43.4% compared to the untreated control and 88% compared to the constituent botanical extracts and RYRF 8 μg/mL. The post hoc Bliss independence analysis, reported in Appendix A (

Table A5. Bliss independence analysis on bile production.

Treatment	Expected Data	Observed Data	Interaction
MIX	19.54%	43.38%	Synergistic

), showed that the observed combined effects exceeded the expected additive responses.

Inline data are reported in Figure 4B in terms of determining the levels of excreted free cholesterol. The data were also compared with free cholesterol levels obtained for control under normal glucose conditions, indicated by the dotted line. Treatments with individual botanical extracts and RYRF at 8 μg/mL result in a significant increase in free cholesterol levels compared to the high-glucose control (p < 0.05). However, the values remain lower than those observed under normal glucose conditions. At the same time, MIX, on the other hand, shows a distinctive bioactive action, inducing a significant increase in free cholesterol compared to the high-glucose control (p < 0.05), exceeding the levels of the normal glucose control (p < 0.05) with significantly greater data for all individual agents (p < 0.05). In detail, the mean increase in free cholesterol levels with MIX was 1.12-fold compared to the untreated control high-glucose group, 20% compared to the untreated control normal-glucose group, 57% compared to the constituent botanical extracts, and 35% compared to RYRF 8 μg/mL.

3.5. Biological Effects of the Botanical Extract Combination on the PCSK9–LDLr–SREBP-2 Regulatory Axis

Under conditions of in vitro-induced hypercholesterolemia, further molecular-level analyses were performed to characterise the bioactive effects of the test samples on the PCSK9–LDLr–SREBP-2 regulatory axis after 24 h of treatment (Figure 5).

In the PCSK9 panel (Figure 5A), RYRF at 8 μg/mL shows a significant increase in expression compared to the untreated high-glucose control (p < 0.05). All botanical extracts individually induced in an almost overlapping manner a modest but significant reduction in PCSK9 levels compared to RYRF 8 μg/mL and, in some cases, also to the untreated high-glucose control (p < 0.05 except for black garlic and Gastrodia elata). As previously observed, MIX resulted in a marked and significant downregulation of PCSK9 levels, significantly higher than that observed with the individual components (p < 0.05). MIX decreased PCSK9 levels by 46.3% compared to the untreated control with high glucose, by 86% compared to the constituent botanical extracts, and by 62% compared to RYRF 8 μg/mL. The SREBP-2 panel (Figure 5B) shows a similar trend with botanical extracts, with RYRF at 8 μg/mL also inducing a small but significant reduction in SREBP-2 compared to the untreated control under high-glucose conditions (p < 0.05). In this case, the effect of the individual botanical extracts was less than that of RYRF 8 μg/mL alone. Again, MIX exerts the most pronounced effect, amplifying the effect of individual botanical extracts and even exceeding RYRF 8 μg/mL alone with a robust decrease in SREBP-2 levels. MIX decreased SREBP-2 levels by 63.8% compared to the untreated control under high glucose, by 88% compared to the constituent botanical extracts, and by 67% compared to RYRF 8 μg/mL. For PCSK9 and SREBP-2 levels (ELISA kit), post hoc Bliss independence analysis, presented in Appendix A (

Table A6. Bliss independence analysis on PCSK9 levels.

Treatment	Expected Data	Observed Data	Interaction
MIX	−23.21%	−46.26%	Synergistic

and

Table A7. Bliss independence analysis on SREBP-2 levels (ELISA kit).

Treatment	Expected Data	Observed Data	Interaction
MIX	−27.30%	−63.75%	Synergistic

), indicated that the observed combined effects were greater than the expected additive responses.

Figure 5C provides additional information on the measurement after Western blot of the SREBP-2 (precursor/active cleaved ratio) after 24 h treatment with RYRF, single botanical extracts and MIX under high-glucose conditions. RYRF and selected single extracts induced a significant increase in the ratio compared to the untreated control (p < 0.05), reflecting reduced proteolytic activation of the transcription factor. RYRF increased the ratio by approximately 11%, showing greater efficacy than individual extracts (ranging between 3% and 7% compared to the untreated control). The most substantial modulation was observed in the MIX treatment, which increased the ratio by approximately 34%, indicating a strong ability of the combined formulation to prevent SREBP-2 cleavage.

Considering this, the LDLr tests performed using ELISA kits and Western blots (Figure 5D,E) suggested that the botanical substances under investigation had excellent bioactive potential. In contrast to the high-glucose control, the RYRF is 8 μg/mL control on the screen devoted to determining LDLr displays low values. With quantitative differences between the various substances and statistically significant findings repeated at RYRF 8 μg/mL and untreated control (p < 0.05), single extracts cause a gradual rise in LDLr levels. Following 24 h of therapy with Primula 50 μg/mL and sesame (10%) 18 μg/mL, the best results were obtained. Among all treatments, MIX produced the greatest and statistically significant increase in LDLr values (p < 0.05). In detail, the MIX increased LDLr levels by 20.2% compared with the untreated control, by 68% compared with black garlic 50 μg/mL, by 74% compared with Gastrodia elata 100 μg/mL, and by 57% compared with Primula 50 μg/mL and sesame (10%) 18 μg/mL, and by 93% compared with RYRF 8 μg/mL. Figure 5D confirmed these results by showing the LDLr/β-actin ratio. Here, too, single extracts increase the ratio compared to the control, while MIX produces the maximum increase, amplifying the bioactive action of the individual constituent botanical extracts.

4. Discussion

Botanical and nutraceutical approaches have received greater attention in recent years as potential alternatives or supplements to traditional lipid-lowering treatments, especially among people with low-to-intermediate cardiovascular risk, statin intolerance, or mild-to-moderate dyslipidemia [57,58]. Botanical extracts are distinguished by pleiotropic and multitarget mechanisms that can simultaneously modulate lipid homeostasis, oxidative stress, and inflammatory pathways, frequently with a more favourable safety profile than statin-based interventions, which primarily act through direct inhibition of HMGR [59]. In this context, the current study provides a thorough in vitro characterisation of a multi-component botanical formulation and each of its constituents, emphasising their ability to regulate hepatic cholesterol homeostasis through mechanisms distinct from those of RYRF. These beneficial effects are likely related to the specific bioactive profiles of each extract, as observed in the chemical tests, since differences in polysaccharide, polyphenol, flavonoid, SAC, and sesamin content may contribute to their distinct antioxidant, neuroprotective, and lipid-modulating activities.

Employing HepG2 cells subjected to hyperglycemic and hypercholesterolemic circumstances, we found that all four botanical extracts tested, black garlic, Gastrodia elata, sesame, and Primula veris, had a positive effect on hepatocyte survival and redox balance. Importantly, the botanical combination demonstrated much greater cytoprotective and antioxidant activity than either of the isolated extracts or RYRF. Because hyperglycemia-induced oxidative stress is a major contributor to hepatic metabolic dysfunction and impaired cholesterol handling [60], the botanical combination’s significant reduction in ROS production most likely represents an upstream mechanism that facilitates the restoration of lipid regulatory pathways [61,62]. This antioxidant-driven hepatoprotection separates botanical methods from RYRF, which has predominantly enzymatic effects that are not directly related to redox regulation.

In terms of cholesterol production, all treatments significantly decreased HMGR activity. As predicted, RYRF achieved this impact through its monacolin K content, which mimics lovastatin [63]. Individual botanical extracts, despite lacking direct statin-like activity, reduced HMGR levels and expression after Western blot analysis, indicating indirect regulation via improved cellular redox status, modulation of transcriptional regulators, or activation of metabolic sensors such as AMPK, as previously reported for sesame lignans and garlic-derived organosulfur compounds [22,64]. Gastrodin, a key bioactive component from Gastrodia elata, in addition to exerting a possible hepatoprotective effect already observed in HepG2 in line [34] with the safety data obtained, suggests potential for modulation of lipid metabolism [65]. While Primula veris, rich in phenolic antioxidants, may contribute [28], direct evidence in HepG2 cells is currently limited. Notably, MIX suppressed HMGR more effectively than either single extract or RYRF, showing a hypothetical synergistic activity that does not rely on direct enzyme inhibition [63]. Nutraceuticals with diverse mechanisms of action can modulate cholesterol metabolism, including the inhibition of HMGR, enhancement of hepatic uptake, and altered lipid absorption or efflux. Nutraceutical combinations based on multiple bioactive compounds, such as plant sterols, policosanol, or berberine, have been shown in clinical trials to effectively manage dyslipidemia, particularly by reducing LDL and total cholesterol while maintaining favourable safety profile [57,66]. Consistent with HMGR regulation, all treatments significantly reduced intracellular cholesterol in HepG2 cells, with MIX showing the greatest effect. Black garlic reduces neutral lipid and cholesterol accumulation in vitro [67]. At the same time, for sesame, sesamin alleviates oleic acid-induced triglycerides and total cholesterol accumulation via autophagy and modulation of lipid-related gene [25]. Gastrodia elata improves lipid and glucose metabolism [68,69], and Primula veris, rich in saponins [70], may support lipid homeostasis through mild cholesterol-lowering and membrane-stabilising effects [71]. These findings suggest that multitarget botanical formulations can efficiently counteract cholesterol accumulation by coordinating pathway regulation rather than by single-enzyme inhibition. In parallel, bile acid production was significantly increased, particularly in cells treated with the combination, indicating enhanced conversion of cholesterol into bile acids, one of the principal routes of cholesterol elimination [72]. The associated increase in free cholesterol levels, even those observed under normal glucose conditions, further supports the notion of improved cholesterol mobilisation and export [73,74]. These effects collectively highlight a key difference from RYRF, which primarily reduces cholesterol synthesis but does not directly promote cholesterol catabolism or excretion to the same extent.

A central and novel aspect of this study is the detailed analysis of the PCSK9–LDLr–SREBP-2 regulatory axis. Under high-glucose conditions, RYRF significantly increased PCSK9 expression, consistent with clinical and experimental evidence showing that statin therapy can induce compensatory PCSK9 upregulation, thereby limiting LDL receptor availability and attenuating LDL clearance [75]. In contrast, all botanical extracts reduced PCSK9 levels, and MIX dramatically enhanced this effect. The strong downregulation of PCSK9 was paralleled by a marked reduction in SREBP-2, indicating coordinated transcriptional control rather than a compensatory response [7]. The Western blot analysis revealed that RYRF induced a significant increase in the SREBP-2 full-length/active cleaved ratio, indicating modulation of SREBP-2 proteolytic activation. Notably, RYRF increased PCSK9 expression despite reduced SREBP-2 activation, suggesting the involvement of alternative monacolin/statin-related regulatory mechanisms, such as hepatocyte nuclear factor-1α (HNF-1α) activation [76,77]. In contrast, botanical extracts inhibited both SREBP-2 proteolytic activation and PCSK9 levels. This distinction is mechanistically crucial. While RYRF acts primarily upstream by inhibiting cholesterol synthesis, thereby triggering SREBP-2 activation and secondary PCSK9 induction [78], the botanical formulation appears to act downstream and laterally, attenuating SREBP-2 signalling itself and preventing PCSK9-mediated LDL receptor degradation [78]. As a result, LDL receptor expression and stability were significantly increased, as confirmed by both ELISA and Western blot analyses, with MIX producing the highest LDLr levels and LDLr/β-actin ratio. Notably, the mechanism observed in the combination of botanical extracts on limited SREBP-2/PCSK9 activation appears consistent with previous evidence highlighting the involvement of this regulatory pathway in the suppression of HMGR expression [79,80]. This mechanism closely resembles that of PCSK9-targeting pharmacological agents but is achieved here through nutraceutical modulation without direct enzymatic inhibition or monoclonal antibody intervention.

The individual botanical extracts contributed complementary actions to this effect. Black garlic, which contains SAC, has been shown to modulate PCSK9 expression and improve LDL receptor activity through redox-sensitive mechanisms [81]. Sesame lignans such as sesamin influence lipid-regulatory genes and AMPK-related pathways, impacting PCSK9/LDLr signalling [25,26]. At the same time, Gastrodia elata and Primula veris contribute antioxidant, anti-inflammatory, and metabolic regulatory effects that likely support the observed synergy in combination [24,82]. Both in vivo and in vitro, Gastrodia elata, thanks to gastrodin, has been shown to reduce PCSK9 expression by inhibiting the Janus kinase 2/signal transducer and activator of transcription 3 (JAK2/STAT3) pathway, thereby reducing HNF-1α and activating forkhead box protein O3 (FoxO3a) and LDLr transcription [24]. Regarding Primula veris, the correlation with the PCSK9–LDLr–SREBP-2 mechanism is not yet fully understood. Still, it has been demonstrated that analogous bioactive plant compounds (such as quercetin, kaempferol, and other flavonoid derivatives) also present in garlic extract, for example [83,84], modulate the expression of PCSK9 and LDLr by influencing the main transcriptional regulators, including SREBP-2 and HNF-1α [85]. These results suggest that the antioxidant and regulatory phytochemicals present in Primula veris [27,28,86] and verified during chemical tests could support beneficial metabolic effects and could act synergistically with other botanical components in the combined extracts.

This study provides evidence that, for the first time, a multi-component botanical formulation combining four distinct extracts can modulate hepatic cholesterol homeostasis via a multifactorial, synergistic mechanism distinct from that of RYRF. Unlike RYRF, which primarily inhibits cholesterol synthesis and triggers compensatory PCSK9 upregulation, the botanical combination coordinates antioxidant, transcriptional, and post-translational regulation of the PCSK9–LDLr–SREBP-2 axis, enhancing cholesterol clearance and showing no cytotoxicity in HepG2 cells under the tested conditions. Importantly, this work also provides new mechanistic insights into some of the less studied extracts (Primula veris) within the formulation, expanding the understanding of their potential roles in lipid metabolism based on in vitro and mecahnistic evidence. These results may support the potential of botanical nutraceuticals as non-statin, pleiotropic strategies for lipid regulation in preclinical models. Notably, the study has some limitations: it relied mainly on in vitro and preclinical models, the contribution of individual components versus the whole combination was not fully dissected, the duration was short, and variability in botanical composition may affect reproducibility. It should be noted that the long-term safety and clinical efficacy of this multicomponent botanical formulation have not yet been established. The present results are limited to in vitro models and cannot be directly extrapolated to human physiology. Future studies should include in-depth systematic combination analyses to further evaluate component interactions, as well as preclinical and rigorously designed clinical investigations, to validate both the efficacy and safety of the formulation. Furthermore, future research employing specific LDL uptake assays using labelled LDL will be crucial for quantitatively assessing receptor-mediated internalisation, which will provide further insights into the formulation’s mechanism of action and its potential therapeutic applicability.

5. Conclusions

This study shows, for the first time, that a multi-component botanical formulation combining four extracts can modulate hepatic cholesterol homeostasis via a synergistic mechanism distinct from that of statin-like agents such as RYRF. The combination regulates the PCSK9–LDLr–SREBP-2 axis, enhances cholesterol clearance, and maintains hepatic safety, providing novel insights into less studied extracts. Limitations include the study’s preclinical nature, short treatment duration, and variability in extract composition. Further studies are needed to confirm the efficacy and long-term safety in humans.