Rebaudioside A Enhances LDL Cholesterol Uptake in HepG2 Cells via Suppression of HMGCR Expression

Background: Rebaudioside A is one of the major diterpene glycosides found in Stevia had been reported to possess anti-hyperlipidemic effects. In this study, we explore the potential cholesterol-regulating mechanisms of Rebaudioside A in the human hepatoma (HepG2) cell line in comparison with simvastatin. Methods: Cells were incubated with Rebaudioside A at several concentrations (0-10 μM) to determine the cytotoxicity by the MTT assay. Cells were treated with selected dosage (1 and 5 μM) in further experiments. Total cellular lipid was extracted by Bligh and Dyer method and subjected to quantitative colorimetric assay. To illustrate the effect of Rebaudioside A on cellular lipid droplets and low-density lipoprotein receptors, treated cells were subjected to immunofluorescence microscopy. Finally, we investigated the expression of experimental gene patterns of cells in response to treatment. Results: In this study, cytotoxicity of Rebaudioside A was determined at 27.72 μM. Treatment of cells with a higher concentration of Rebaudioside A promotes better hepatocellular cholesterol internalization and ameliorates cholesterol-regulating genes such as HMGCR, LDLR, and ACAT2. Conclusions: In conclusion, our data demonstrated that Rebaudioside A is capable to regulate cholesterol levels in HepG2 cells. Hence, we proposed that Rebaudioside A offers a potential alternative to statins for atherosclerosis therapy.


Introduction
An evidence-based phytotherapy pathway that relies on scientific and clinical data is a crucial approach before the establishment of plantbased derivatives in medical practice (1). Stevia rebaudiana Bertoni is an intense, non-nutritive natural sweetener that has been a popular sugar substitute used in various industries as well as phytotherapy (2). Following the first domestication in Japan, large-scale cultivation now exists in other countries to fulfil the economical demands (3). Phytochemical of Stevia leaves revealed the presence of steviol glycosides, and other important secondary metabolites each with various therapeutic potentials (3,4). To date, there are 11 major steviol glycosides detected in the plant (5).
Rebaudioside A accounted for up to 6.5% of the total dry weight of the leaf is of interest due to its potent sweetness estimated to be 400 times sweeter than sucrose (6). There exist quite a lot of discussion as to therapeutic applications of Rebaudioside A (7)(8)(9)(10). Recent investigations demonstrated the potential role of Rebaudioside A as a hepatoprotective agent (11,12). More importantly, it has been shown that an improvement in the lipid profile of experimentally induced diabetic rats can be achieved by oral administration of this diterpenoid glycoside (13) in which further support the notion that Rebaudioside A possesses potential lipid-regulating activity. Atherosclerosis is a chronic inflammatory disease affecting the intima of medium and large-sized arteries, likely to occur at arterial bifurcations (14,15). The disease has been recognized as the principal cause of clinically important cardiovascular diseases (CVDs) (16). Pioneering work on atherosclerosis has been carried out earlier by Alexander Iosifovich Ignatowski in 1907 in which he provided evidence of aortic atherosclerotic lesion in experimental rabbits fed with various types of animal proteins (17). Conceptually similar work has also been carried out by Anichkov and Semen Chalatov in 1912 that extends the idea of cholesterol in atherosclerosis development (17). In addition to non-pharmacological approaches, few medications have been developed to intervene with abnormal plasma cholesterol levels (18). A class of cholesterol-modifying medications known as statins is commonly prescribed to hyperlipidemic patients (18). Likewise, considering that reactive oxygen species (ROS) and oxidative stress initiate atherosclerosis which cause coronary heart disease, antioxidants are beneficial to prevent such events (19). Despite the efficiency of statins in lowering cholesterol, many side effects have been discussed (20)(21)(22) thus further motivates the need for an alternative to replacing statins. Therefore, in this study, we are interested in examining the unexplored cholesterol-lowering potential of Rebaudioside A in the HepG2 cell line.

Preparation of Rebaudioside A and Simvastatin
HPLC grade of Rebaudioside A (≥ 96%) and Simvastatin (≥ 97%) were from Sigma Aldrich (USA). Rebaudioside A (RA) was prepared as a stock solution in a sterile dimethyl sulfoxide with a final concentration of 1mM. Simvastatin was activated as mentioned by (23). Both solutions were stored at 4 °C until further analysis.

Cell viability assay
Cells (1x10 4 cells per well) were incubated in 96 well-plates for 24 hours. The following day, the culture medium was removed and cells were further incubated with RA (0-10 µM) overnight. Upon 24 hours, the cytotoxic effect was determined by the MTT assay (24). The halfmaximal inhibitory concentration ( cDNA analysis, the treatment medium was removed, cells were washed with PBS, scraped off, and collected into a 1.5ml tube. For immunofluorescence imaging, cells were stained, mounted on coverslips, and viewed.

Cholesterol quantification and protein determination
Before analysis, cells were homogenized for 2 minutes at 20% amplitude on ice using an ultrasonic dismembrator (Thermo Fisher Scientific, USA) followed by lipid extraction according to Bligh and Dyer method (25). Total cellular cholesterol (TCC) was quantified by Cholesterol Liquicolor Kit (Human Diagnostic, USA) according to the manufacturer's instruction at 540nm wavelength using a spectrophotometer. TCC was calculated as follow: Protein concentration was determined using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific, USA). Data acquired from protein determination was used to normalize data of cholesterol quantification analysis.

RNA isolation and optimization of primers
RNA was extracted using the RNeasy Mini Kit (Qiagen, Germany) as mentioned in the manual. The purity and quantity of total cellular RNA were assessed spectrophotometrically using a NanoQuant Plate™ (Tecan, Switzerland). Optimization of experimental primers (IDT, Singapore) in Table 1 with standardized total RNA (100 ng/µl) was carried out by conventional reverse transcriptase-polymerase chain reaction (RT-PCR) using the Access RT-PCR System master mix (Promega, USA) in the Thermoblock 96 thermocycler (Sensoquest, Germany). The resulting PCR products were separated by agarose gel electrophoresis (2% agarose, w/v; at 80V for 40 minutes) containing SYBR ® Safe DNA gel stain (Thermo Fisher Scientific, USA) and visualized using the Gel DOC XR+ System Image Analyser (Bio-Rad, USA).

Gene expression analysis by 2-step real-time quantitative PCR (2-step RTPCR)
The

Statistical analyses
Data were analyzed by One-Way ANOVA test using the GraphPad Prism version 7 for Windows, GraphPad Prism (USA). Results were presented as mean ± standard deviation and P-value< 0.05 was considered as statistically significant.

Cell viability assay
HepG2 cells were treated with RA (0-10 µM) for 24 hours and the results were depicted in Figure  1. The trend of the results illustrated that cell viability decreases dose-dependently with RA incubation for 24 hours. Notably, the IC50 of RA was found to be at 27.72 µM. In line with the findings from this assay, 1 µM and 5 µM of RA were used to treat cells in the following experiments.

Cholesterol quantification and protein determination
Data illustrated in Figure 2 indicated the extracted total cellular cholesterol (TCC) level in different groups of cells was considerably affected by treatments. According to the results, the highest value for TCC was recorded in HRA following PC> LRA> NC, while the lowest was observed in BC. Surprisingly, treatment of cells with LRA has little to no effects on TCC as the result is almost similar to NC.

Immunofluorescence microscopy
To determine the expression of LDLR and LDs, a qualitative assessment by fluorescent staining was employed. The microphotographs depicted in Figure 3 suggest that HRA promoted the upregulation of LDLR expression on cells' surface causing higher intracellular lipid mobilization and further reserved into LDs. The expression of LDLR was qualitatively prominent in HRA> LRA> PC but absent in both NC and BC. However, the presence or absence of LDLR does not affect LDs formation observed in all groups. It is important to highlight that no strong conclusions could be drawn based on the qualitative assessment of immunofluorescence staining although it reflects the results of the cholesterol quantification assay. Therefore, to ensure the molecular mechanism of RAinducing cellular uptake of cholesterol is properly understood, gene expression analysis was conducted.

Gene expression
The purity of extracted RNA samples was in a range of 2.0 to 2.1 at an absorbance ratio of 260-280 nm wavelengths. The annealing temperature of each experimental and internal control gene was determined by conventional RT-PCR followed and qualitative assessment of bands via agarose gel electrophoresis (GAPDH: 490 bp, HMGCR: 747 bp, LDLR: 258 bp, and ACAT2: 117 bp). The optimized annealing temperatures were adapted to cycling conditions for gene expression analysis by 2-step RT-PCR. The expression of experimental genes was illustrated as expression-fold in Figure 4. It was found that HRA significantly down-regulated HMGCR expression and at the same time up-regulating LDLR and ACAT2 genes as compared to other treatment groups.

Discussion
To our knowledge, this work is novel in describing the anti-hypercholesterolemia activity of RA and the potential molecular regulation of cholesterol levels in HepG2 cells. The cell line was used as an in-vitro model as it expresses multi-differentiated functions of the liver, especially in cholesterol metabolism (26,27). Available data from this in-vitro study could provide scientific evidence on the application of RA therapeutical outcomes.
We first validated the cytotoxic effects of RA in HepG2 cell lines via MTT assay. The present experiment has illustrated three interesting findings. First, the incubation of cells with 1 µM and 2 µM of RA did not show any significant difference in cell viability compared to untreated cells (0 µM). Second, a decrease in cellular viability with a significant difference was observed as the concentration of RA given to cells increased (2.5-10 µM). Third, cellular proliferation was absent in all treatments which may suggest the antiproliferative effects of RA. Following the current findings, the toxicity effect of RA is observed to be cell dependant. The higher concentration of RA significantly decreased the cellular viability and proliferation of these cells due to membrane ruptures induced by high osmotic pressure (28).
Next, we examined the effects of RA in regulating intracellular cholesterol mobilization. Although cholesterol is important in maintaining cellular architecture, dynamics, and functions (27,29) high circulating cholesterol levels are a predisposing factor of atherosclerotic cardiac disease (27,30,31). Here we observed the TCC levels in the HRA group are the highest. The analysis presumed that HRA in this study is a potential candidate of cholesterol-lowering substitute but could not be definitively proven by comparing the TCC levels and fluorescence staining entirely. Therefore, to ensure the molecular mechanism is properly understood, gene expression analysis was done.
In this analysis, cells treated with HRA show promising results by significantly altered cholesterol-regulating genes (HMGCR, LDLR, and ACAT2) in comparison to other treatment groups. The first experimental gene observed, HMGCR was down-regulated in HRA further suggest that RA is a potential cholesterollowering alternative to simvastatin. This gene encodes the rate-limiting enzyme in endogenous cholesterol synthesis, the 3hydroxy-3-methylglutaryl coenzyme A reductase (HMGCR) (32). Statins possess a similar structure as HMGCR which competitively inhibits the enzyme results in decrease cholesterol levels in cells (33). We postulated that RA exerts similar biochemical and physiological effects as statins. Treatment of cells with HRA also caused an up-regulation of LDLR and ACAT2 gene expression. The mechanisms could be further elaborated by a brief understanding of the negative feedback mechanism evoked by strings of proteins interaction initiated by sterol regulatory element-binding protein (SREBP).
SREBP is a transcriptional factor protein that controls lipids synthesis in cells (34,35). This protein is encoded by two genes (SREBF1: SREBP-1 and SREBF2: SREBP-2) that regulate cholesterol synthesis genes (27,36). Among the existing isoforms, SREBP-2 is expressed ubiquitously and tightly regulated by cellular cholesterol levels (34,35,37). SREBP-2 is present in the endoplasmic reticulum (ER) membrane as an inactive precursor bound to SREBP cleavage-activating protein (SCAP) (38). SCAP possesses a cholesterol-sensing domain that responds to intracellular cholesterol depletion causing translocation of SREBP-2 to the Golgi in a vesicle. Here, the Site-1 and -2 proteases are cleaved which release the NH2-terminal transcription-activation domain of the SREBP-2. The NH2-terminal enter the nucleus binds to a sterol response element and ameliorates the expression of SREBP-2 targeted genes (34,(37)(38)(39).
We also demonstrated that RA decreased the expression of HMGCR and its transcription product, the HMGCR enzyme. Inhibition of the enzyme results in alleviated intracellular cholesterol levels which initiate the proteolytic activation of SREBP-2 (33). The activation of SREBP-2 initiates the transcription of the LDLR gene results in increased expression of LDLR on the cell surface. LDLR gene encodes the genetic information required for LDLR synthesis involved in cellular uptake of circulating LDLC (33,40). The regulation of LDLC is important in providing balance cholesterol homeostasis necessary to prevent certain cardiovascular diseases development (41). Circulating LDLC binds to LDLR and endocytosed through clathrin-mediated endocytosis. Through chains of enzymatic reactions, LDLR is recycled to the cell surface via an endocytic recycling compartment, whereas non-esterified cholesterol is hydrolyzed from LDLC by lysosomal acid lipase (LAL).
Next, we observed that the expression of the ACAT2 gene in HRA is significantly up-regulated in parallel to the expression of LDLR. ACAT2 encodes the acyl-CoA:cholesterol acyltransferase (ACAT2), an enzyme that catalyzes the esterification of cholesterol (42). ACAT2 is one of the membrane-bound Oacyltransferase (MBOAT) families that plays important role in neutral lipid synthesis for lipid droplets biogenesis (43). This enzyme is primarily localized to the ER that utilizes nonesterified lipids (e.g. fatty acyl CoA and cholesterol) to form cholesterol esters which are reserved in between the leaflets of the ER bilayer and give rise to oil lens as the concentration increases (43,44). Subsequently, LDs will bud off into the cytosol (43). LDs are highly dynamic intracellular organelles that prevent lipotoxicity and oxidative stress in cells by buffering potentially toxic lipids and proteins accumulation (45). Figure 5 depicted the proposed molecular pathway of RA in regulating the intracellular intake of LDLC.  5. A proposed molecular pathway of Rebaudioside A in ameliorating the expression of cholesterol-synthesis genes (HMGCR, LDLR and ACAT2). Rebaudioside A competitively inhibits the enzymatic reaction of HMGCR causing intracellular cholesterol depletion. Proteolytic activation of SREBP-2 was stimulated as the SCAP/SREBP-2 complex translocated to the Golgi followed by releasing of NH2-terminal into the nucleus. The consequences of this event lead to LDLR gene transcription and more LDLR proteins exist on the cell surface. Circulating LDLC binds to LDLR and internalized via the clathrin-mediated endocytosis. In cells, LDLC segregates from its receptor (recycle to the surface) due to acidic pH and the former was hydrolysed by LAL. The non-esterified cholesterol molecules (free cholesterol) are toxic to cells, thereby esterified by ACAT2 protein into neutral lipids (cholesterol esters) and further stored in lipid droplets (This figure was created with BioRender.com). In summary, this novel in-vitro study shows that Rebaudioside A is a potential cholesterollowering candidate for statins. Considering its natural and non-cytotoxic characteristics, Rebaudioside A possesses few other therapeutical advantages to statins drawback. Despite the success demonstrated, this work is limited by its absence of other cholesterol synthesis genes namely SREPB-2, HMGCS, and MVK. Hence, future research should therefore seek to address this issue by implying these genes as well as other possible approaches such as large-scale multi-omics.
Regardless, the findings in this study provide a very useful insight to understand the mechanism of Rebaudioside A in lowering cholesterol levels.