Anti-Adipogenic Activity of Rhaponticum carthamoides and Its Secondary Metabolites

Besides their common use as an adaptogen, Rhaponticum carthamoides (Willd.) Iljin. rhizome and its root extract (RCE) are also reported to beneficially affect lipid metabolism. The main characteristic secondary metabolites of RCE are phytoecdysteroids. In order to determine an RCE’s phytoecdysteroid profile, a novel, sensitive, and robust high-performance thin-layer chromatography (HPTLC) method was developed and validated. Moreover, a comparative analysis was conducted to investigate the effects of RCE and its secondary metabolites on adipogenesis and adipolysis. The evaluation of the anti-adipogenic and lipolytic effects was performed using human Simpson–Golabi–Behmel syndrome cells, where lipid staining and measurement of released glycerol and free fatty acids were employed. The HPTLC method confirmed the presence of 20-hydroxyecdysone (20E), ponasterone A (PA), and turkesterone (TU) in RCE. The observed results revealed that RCE, 20E, and TU significantly reduced lipid accumulation in human adipocytes, demonstrating their anti-adipogenic activity. Moreover, RCE and 20E were found to effectively stimulate basal lipolysis. However, no significant effects were observed with PA and TU applications. Based on our findings, RCE and 20E affect both lipogenesis and lipolysis, while TU only restrains adipogenesis. These results are fundamental for further investigations.

Although many bioactive compounds of plant origin possess anti-obesity potential, the specific role of phytoecdysteroids (PDs) remains incompletely understood. Noratto et al. reported that the intake of quinoa (a phytoecdysteroid-rich plant) is associated with positive effects on obesity in mice [21]. However, further research is needed to reveal the potential of PDs and phytoecdysteroid-containing plants on obesity. Among the phytoecdysteroid-rich plants is the endemic perennial plant Rhaponticum carthamoides (Willd.) Iljin. from the Asteraceae family, commonly known as maral root or Russian leuzea [22,23]. In traditional medicine, it has been used to improve physical strength [22]. In 1969, leuzea was systemized as one of the plant adaptogens by Brekman and Dardimov [22]. In recent decades, extracts from its rhizomes and roots have been used for physical weakness, to promote muscle growth, to treat impotency, etc. [22]. Various medicinal preparations from R. carthamoides rhizomes and roots have been reported to possess not only adaptogenic effects but also a broad spectrum of biological effects, such as antioxidant, immunomodulatory, anticancerogenic, antimicrobial, antiparasitic, and repellent activities [22][23][24]. The main isolated chemical classes from R. carthamoides rhizomes and roots are not only PDs but also phenolics (flavonoids and phenolic acids) [22]. Previous phytochemical reports regarding ecdysteroids of R. carthamoides from the underground parts of the plants revealed the isolation of 20-hydroxyecdysone, also known as ecdysterone (20E), ponasterone A (PA), and turkesterone (TU) [22,25,26]. Ecdysteroids are a group of polyhydroxylated sterols, structurally similar to androgens ( Figure 1) [24,27,28]. PDs possess a wide range of pharmacological properties, including antidiabetic and hepatoprotective properties [24,29]. Previous studies reported that 20E possesses anabolic, neuroprotective, and antitumor effects, restores renal dysfunction, and decreases triglycerides [28][29][30][31][32][33][34]. It has the potential to prevent adiposity, dyslipidemia, and hyperglycemia [35,36], while PA possesses anabolic activity [29]. Turkesterone has been associated with anabolic, antidiabetic, and hypoazotemic effects, as well as the ability to decrease cholesterol levels and restore renal function, according to previous investigations [29].
The current study aimed to evaluate the phytochemical profile of R. carthamoides rhizomes and roots extract (RCE) via an innovative high-performance thin layer chromatog- PDs possess a wide range of pharmacological properties, including antidiabetic and hepatoprotective properties [24,29]. Previous studies reported that 20E possesses anabolic, neuroprotective, and antitumor effects, restores renal dysfunction, and decreases triglycerides [28][29][30][31][32][33][34]. It has the potential to prevent adiposity, dyslipidemia, and hyperglycemia [35,36], while PA possesses anabolic activity [29]. Turkesterone has been associated with anabolic, antidiabetic, and hypoazotemic effects, as well as the ability to decrease cholesterol levels and restore renal function, according to previous investigations [29].
The current study aimed to evaluate the phytochemical profile of R. carthamoides rhizomes and roots extract (RCE) via an innovative high-performance thin layer chromatography (HPTLC) method. Moreover, the PDs in the extract-20E, PA, and TU-were quantified. Following the chemical analysis, the effect of RCE as well as its secondary metabolites on adipogenesis and adipolysis were investigated in an in vitro obesity model of human adipocytes.

Plant Material and Extraction
The dried R. carthamoides rhizomes and roots were purchased from Russia. The plant material was characterized according to the Russian Pharmacopoeia by the Department of Pharmacognosy and pharmaceutical chemistry, Medical University of Plovdiv. The plant material was further frozen, freeze-dried, and ground before ultrasound-assisted extraction with 50% aqueous methanol at 20 • C for 30 min. The obtained RCE was filtered and concentrated via a rotary vacuum evaporator at 40 • C, further freeze-dried, and stored at −20 • C before use.

Sample Preparation
The stock solutions for HPTLC analysis of 20E, PA, T, and the extract were prepared in acetonitrile in concentration 1 mg·mL −1 . Ultrasound was used for better dissolution. The prepared stock solutions were stored before use in brown vials, protected from light at 4 • C.

Method Development
The analyses were carried out using silica gel 60 F254 glass TLC plates of 10 × 20 cm size and with 200 µm layer thickness (E. Merck KGaA, Darmstadt, Germany). The mobile phase comprised methanol: acetonitrile at a ratio of 10:90 (v/v). The volume of the mobile phase was 10 mL. The application type was a band, and the front was 70 mm. The time for development was 10 min, followed by drying for 5 min. Detection was performed at 254 nm using CAMAG TLC Visualizer 2.

Method Validation
The developed method was validated according to the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) with the following validation parameters: linearity, range, precision, accuracy, and limits of detection (LD) and quantification (LQ) [37].

Cell Culture and Treatment
Human Simpson-Golabi-Behmel syndrome (SGBS) preadipocytes were kindly provided by Professor Martin Wabitsch (University of Ulm, Germany). The cells were cultured under optimal conditions [38,39]. The differentiation of near-confluent preadipocytes was carried out with the presence of RCE (5-50 µg·mL −1 ), 20E (5-50 µM), PA (5-50 µM), TU (5-50 µM), or 0.02% DMSO as a vehicle. These concentrations were selected based on cell viability evaluation with the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. The experimental treatments were applied upon differentiation and on the fourth and eighth days with every culture media replacement process. Sample collection and subsequent analyses were performed 24 h after the last treatment. Each assay was performed at least in three independent experiments.

Cell Viability Assay
The preadipocytes were seeded in 96-well plates, grown to near confluence, and incubated for 48 h with increasing concentrations from 0.1 to 100 µM for the pure compounds and 0.1 to 100 µg·mL −1 for the extract or 0.02% DMSO as a vehicle. The results are presented as the percentage of cell viability compared to the vehicle as the mean ± SEM and are representative of three independent experiments.

Lipid Staining
The procedure was performed as previously described [40]. Briefly, on day 9 of differentiation, the SGBS adipocytes were fixed with formalin and subsequently stained with freshly prepared ORO solution. Then, the representative microphotographs were taken using an Oxion Inverso OX.2053-PLPH inverted microscope, equipped with a DC.10,000-Pro CMEX camera (Euromex, Arnhem, The Netherlands). For the quantification of accumulated lipids for each group, the absorbance of the extracted lipid dye at 495 nm was measured using an Anthos Zenyth 340 multiplate reader from Biochrom Ltd. (Cambridge, United Kingdom). The results were represented as the percentage of accumulated lipids compared to the vehicle-treated group.

Analysis of Basal and Stimulated Lipolysis
The effect of increasing concentrations of RCE, 20E, PA, and TU, was evaluated through quantification of released glycerol and free fatty acids (FFAs) in the culture media, as products of lipid hydrolysis. Along with the last treatment, on day 8 of adipogenic differentiation, lipolysis was stimulated with 10 µM isoproterenol for 24 h [41]; then, culture media samples were collected from the tested treatments subjected to both basal and isoproterenol-stimulated conditions. Glycerol and FFA concentrations were determined using a glycerol assay kit (#MAK117) and a free fatty acid assay kit (#MAK044) from Merck KGaA according to the manufacturer's instructions.

Statistical Analysis
The resulting data were expressed as the mean ± SEM and statistical significance between groups was determined by one-way ANOVA, followed by Tukey's post hoc test, using SigmaPlot v11.0 software from Systat Software GmbH (Erkrath, Germany). Values of p < 0.05 were considered significant.

Method Development
A rapid and sensitive HPTLC method was developed to quantify the PDs present in RCE. The method was effective for the estimation of 20E, PA, and TU.
The first step of the method development was to choose a suitable solvent system for the analyzed compounds. For the determination of a suitable phase, various proportions of acetonitrile and methanol were used. Proportions including acetonitrile/methanol (85:15, v/v), acetonitrile/methanol (80:20, v/v), acetonitrile/methanol (10:90, v/v), acetonitrile/methanol (95:5, v/v), and acetonitrile/methanol (50:50, v/v) were investigated as the solvent systems for the development of a suitable band for quantitation. From the results, it was observed that the acetonitrile/methanol (85:15, v/v), acetonitrile/methanol (80:20, v/v), acetonitrile/methanol (10:90, v/v), acetonitrile/methanol (95:5, v/v), and acetonitrile/methanol (50:50, v/v) solvent systems presented a poor chromatogram of the examined PDs with a poor asymmetry factor. Among the tested solvent systems, acetonitrile/methanol (90:10, v/v) provided well-separated and compact chromatographic peaks of TU, 20E, and PA at Rf 0.2, 0.3, and 0.6, respectively. Hence, the acetonitrile/methanol (90:10, v/v) proportion was considered as a proper solvent system for the determination of 20E, PA, and TU in the HPTLC method. Figures 2 and 3 represent the HPTLC chromatogram and profiles of different concentration levels of the PDs and RCE, respectively.
A rapid and sensitive HPTLC method was developed to quantify the PDs present in RCE. The method was effective for the estimation of 20E, PA, and TU.
The first step of the method development was to choose a suitable solvent system for the analyzed compounds. For the determination of a suitable phase, various proportions of acetonitrile and methanol were used.  (50:50, v/v) solvent systems presented a poor chromatogram of the examined PDs with a poor asymmetry factor. Among the tested solvent systems, acetonitrile/methanol (90:10, v/v) provided well-separated and compact chromatographic peaks of TU, 20E, and PA at Rf 0.2, 0.3, and 0.6, respectively. Hence, the acetonitrile/methanol (90:10, v/v) proportion was considered as a proper solvent system for the determination of 20E, PA, and TU in the HPTLC method. Figures 2 and 3 represent the HPTLC chromatogram and profiles of different concentration levels of the PDs and RCE, respectively.

Method Validation
The method was validated according to ICH guidelines [37].

Linearity
To establish linearity, an external standard curve was employed. The calibration curves were plotted by concentrations and the peak area of each PD. To prepare the standard solution, 20E (0.5 to 1.5 µg·band −1 ), PA (0.5 to 1.5 µg·band −1 ), and TU (0.5 to 1.5 µg·band −1 ) were dissolved in acetonitrile. The regression line was calculated with y = ax ± b, where x is the concentration and y is the peak area of each PD, b is the y-intercept, and a is the slope of the regression line. Moreover, the coefficient of determination (R 2 ) was established for the linearity.
The linearity of the method was determined in the 0.5-1.5 µg.band −1 range for the tested substances. The regression line equation and the R 2 for 20E were y = 0.0027x + 0.0002 and R 2 = 0.9988, for PA, they were y = 0.0042x + 0.0003 and R 2 = 0.9986, and for TU, they were y = 0.004x + 0.0008 and R 2 = 0.997, respectively. These results showed a significant correlation and demonstrated the reliability of the method for estimating these PDs. Table 1 presents the results of the regression analysis, LD, and LQ.

Method Validation
The method was validated according to ICH guidelines [37].

Linearity
To establish linearity, an external standard curve was employed. The calib curves were plotted by concentrations and the peak area of each PD. To prepa standard solution, 20E (0.5 to 1.5 µg·band −1 ), PA (0.5 to 1.5 µg·band −1 ), and TU (0.5 µg·band −1 ) were dissolved in acetonitrile. The regression line was calculated with y b, where x is the concentration and y is the peak area of each PD, b is the y-intercep a is the slope of the regression line. Moreover, the coefficient of determination (R established for the linearity.
The linearity of the method was determined in the 0.5-1.5 µg.band −1 range f tested substances. The regression line equation and the R 2 for 20E were y = 0.0027x + 0 and R 2 = 0.9988, for PA, they were y = 0.0042x + 0.0003 and R 2 = 0.9986, and for TU were y = 0.004x + 0.0008 and R 2 = 0.997, respectively. These results showed a signi correlation and demonstrated the reliability of the method for estimating these PDs. 1 presents the results of the regression analysis, LD, and LQ.

Accuracy
Accuracy was established across the specified range of the analytical procedure, which was determined to be from the 0.5 to 1.5 µg·band −1 for 20E, PA, and TU.
The accuracy of the suggested HPTLC method was evaluated using the percentage of recovery of three concentration levels (low, medium, and high) with six replicates of each concentration. For the accuracy test, from each examined substance, three different quality control (QC) levels were used: lower QC (LQC: 0.75 µg·band −1 ), middle QC (MQC: 1 µg·band −1 ), and high QC (HQC: 1.25 µg·band −1 ) with six replicates. Table 2 presents the results of the accuracy of the developed HPTLC method.

Precision
The precision of the proposed HPTLC method was evaluated for both intra-day and inter-day precision, with six replicates of the injection. Examining the intra-day variation for the examined substances involved quantifying fresh solutions at LQC, MQC, and HQC on the same day in six replicates (n = 6). Inter-day variability for the examined substances was examined using the quantitation of freshly generated solutions at LQC, MQC, and HQC on three consecutive days in six replicates (n =6). Table 3 presents the results for the precision of these PDs.  Detection Limit (DL) and Quantitation Limit (QL) The detection limit and quantification limit were expressed by the standard deviation of the slope (σ) and the slope of the calibration curve (S) using the following formulas: DL = 3.3 σ/S and QL = 10 σ/S, respectively.

Robustness
The robustness of the proposed method was assessed by deliberately introducing variations in the mobile phase compositions and total run length. The solvents ratio of acetonitrile/methanol (90:10, v/v) was modified within a range of ±1%, and the HPTLC response was recorded for each set of conditions. The total solvent distance was altered to 72 mm and 68 mm from the initial 70 mm, and the HPTLC response was recorded. The observed changes in Rf values were within the range ± 0.02, which indicated that the method was robust.
In order to assess the stability of the standard solutions, they were stored at 2-8 • C for a week, visual inspection confirmed the clarity of the solutions, and subsequently, the obtained chromatograms from the freshly prepared solutions were compared with those The close values of correlation factors to one, the high percentage of accuracy, and the low values of standard deviation suggested that the developed method is linear, accurate, precise, and reliable for the determination and quantification of 20E, PA, and TU. The developed and validated method was used for the quality determination of the three compounds in RCE. The amount of 20E was found to be 2.96 mg·g −1 , PA was found to be 1.75 mg·g −1 , and the amount of TU was found to be 1.65 mg·g −1 crude dry extract. The obtained HPTLC results were confirmed through HPLC/UV analysis using the previously validated method [42].

Effect of RCE, 20E, PA, and TU on Cell Viability
The performed MTT assay revealed that the cell viability of the near-confluent preadipocytes was not affected upon incubation for 48 h with RCE µg·mL −1 and 20E, PA, and TU in 0.1-100 µM, respectively ( Figure 4). Consequently, the selected treatment concentrations are safe for application in the following experiments.

Effect of RCE, 20E, PA, and TU on Adipogenesis in Human Adipocytes
The chemical profiling of RCE affirmed the presence of 20E, PA, and TU. To evaluate whether RCE and the identified PDs modulate adipogenesis, ORO lipid staining was performed. The observed tendency toward a reduction in accumulated lipids is represented through microscopic images of the treated groups ( Figure 5A).

Effect of RCE, 20E, PA, and TU on Adipogenesis in Human Adipocytes
The chemical profiling of RCE affirmed the presence of 20E, PA, and TU. To evaluate whether RCE and the identified PDs modulate adipogenesis, ORO lipid staining was performed. The observed tendency toward a reduction in accumulated lipids is represented through microscopic images of the treated groups ( Figure 5A).
Collectively, the screening, based on lipid staining, affirmed that RCE and the identified PDs-20E and TU-possess promising anti-adipogenic activity. Further experiments evaluated whether the modulation of adipolysis is involved in the observed decrease in total lipid content.

Effect of RCE, 20E, PA, and TU on Adipogenesis in Human Adipocytes
The chemical profiling of RCE affirmed the presence of 20E, PA, and TU. To evaluate whether RCE and the identified PDs modulate adipogenesis, ORO lipid staining was performed. The observed tendency toward a reduction in accumulated lipids is represented through microscopic images of the treated groups ( Figure 5A).

Effect of RCE, 20E, PA, and TU on Basal and Isoproterenol-Stimulated Lipolysis in Human Adipocytes
To determine the effect of RCE, 20E, PA, and TU on basal and isoproterenol-stimulated lipolysis, quantification of glycerol and FFAs released in the culture media was performed.
Incubation with RCE (50 µg·mL −1 ) significantly increased the concentration of both glycerol ( Figure 6A) and FFAs ( Figure 6C) in the culture media under basal conditions. In a similar manner, in unstimulated adipocytes, 20E (50 µM) application only significantly increased the FFA concentration ( Figure 6C). Treatment with PA and TU affected neither glycerol ( Figure 6A) nor FFA ( Figure 6C) concentrations. Therefore, we can suggest that both RCE and 20E significantly increased basal lipolysis, while such an effect was not observed upon PA or TU treatment ( Figure 6A,C). Isoproterenol stimulation elevated the released glycerol and FFAs in comparison to the basal level. However, no significant effect on both glycerol ( Figure 6B) and FFAs ( Figure 6D) was detected upon all the treatments applied compared to the isoproterenol-stimulated vehicle. Basal lipolysis was elevated only in the highest concentrations of RCE and 20E. However, none of the treatments affected isoproterenol-stimulated adipolysis.

Discussion
Considering the continuously increasing interest in plant products as food supplements due to the assumption of their safety [43], it is crucial to address the limited data regarding the control of their quality. Moreover, a lack of relevant scientific evidence confirming the biological activity and efficacy of some plant products [44] raises additional questions about the rationale of their use. Therefore, the development of fast, precise, and sensitive analytical methods is of great importance for the adequate control of plant products. RCE and PDs are widely used as adaptogens [22,45]. However, the available HPTLC methods for identifying and quantifying PDs, especially for PDs isolated from RCE, are currently quite limited, as evidenced by the data presented in Table 4 [46][47][48][49][50]. For that reason, a new HPTLC method that offers high sensitivity, efficiency, and reproducibility has been developed. The HPTLC technique has several advantages, including rapidity, cost-effectiveness, etc. [51,52], making it a highly suitable method for the future analysis of PDs.

Purpose
Chromatographic Conditions LD/LQ Ref. Basal lipolysis was elevated only in the highest concentrations of RCE and 20E. However, none of the treatments affected isoproterenol-stimulated adipolysis.

Discussion
Considering the continuously increasing interest in plant products as food supplements due to the assumption of their safety [43], it is crucial to address the limited data regarding the control of their quality. Moreover, a lack of relevant scientific evidence confirming the biological activity and efficacy of some plant products [44] raises additional questions about the rationale of their use. Therefore, the development of fast, precise, and sensitive analytical methods is of great importance for the adequate control of plant products. RCE and PDs are widely used as adaptogens [22,45]. However, the available HPTLC methods for identifying and quantifying PDs, especially for PDs isolated from RCE, are currently quite limited, as evidenced by the data presented in Table 4 [46][47][48][49][50]. For that reason, a new HPTLC method that offers high sensitivity, efficiency, and reproducibility has been developed. The HPTLC technique has several advantages, including rapidity, cost-effectiveness, etc. [51,52], making it a highly suitable method for the future analysis of PDs. - [50] Except for the traditional adaptogenic activity of RCE, diverse biological activities are reported either for the extract or for the PDs investigated in this study. The extract has been evaluated for anti-neoplastic activity [53,54], cardioprotective effects [55], and the stimulation of muscle protein synthesis [56]. Moreover, an in vivo study indicates the beneficial effects of RCE on fat tissue expansion and hepatic triglyceride accumulation [57]. The reported biological activity of 20E includes anti-neoplastic activity [31,58], the modulation of mitochondrial bioenergetics [59], immunomodulatory effects [60], an increase in the muscle mass amelioration of the radiation-induced damage of oral mucosa [61], and neuroprotective [30,62,63], anti-fibrotic [64], wound-healing [65], and anti-inflammatory [66,67] activities. Several reports have proposed the potential of 20E to benefit metabolic disturbances such as obesity [36,68]; in addition, it has been reported to exert anti-diabetic [36,65,69,70] and anti-osteoporotic [71][72][73] effects. Moreover, as a food supplement in humans, 20E was found to increase strength performance with no effect on steroid profile [32]. Both PA and TU have not been investigated as natural compounds with anti-obesogenic effects.
The balance between adipogenesis and adipolysis determines the size of fat cells [74]. Thus, a decrease in lipid accumulation along with the stimulation of triglyceride mobilization are among the anti-obesogenic mechanisms of plant extracts and their constituents [75].
In order to affirm the available data for the potential of RCE and 20E in obesity management, as well as to evaluate the anti-adipogenic activity of PA and TU, the current investigation assessed the effect of RCE, 20E, PA, and TU on adipogenesis and lipolysis in vitro in human adipocytes. The applied cell-based platform provides fast and reliable screening of anti-adipogenic potential and accelerates the identification and selection of drugs which leads to their effects being subsequently evaluated in vivo.
The observed anti-adipogenic activity of RCE is in accordance with the previously reported decrease in the weight of epididymal fat tissue in rats [57]. Moreover, the obtained results suggested that among the identified PDs from the extract, only 20E and TU significantly decreased lipid accumulation during adipocyte differentiation. The detected effect upon 20E treatment is consistent with the literature data for reduced adipocyte size in dietinduced obesity in a murine model [68]. Interestingly, the current investigation suggested the notable anti-adipogenic activity of TU, which has not been previously reported.
Lipolysis is the process of triglyceride hydrolysis which is assumed to decrease the size of adipocytes [75] and can also be accepted as an indicator of energy expenditure [76]. Principally, there are two types of adipolysis-basal and stimulated (upon β3-adrenergic receptor activation by isoproterenol or catecholamines) [75]. Our findings suggest that RCE and 20E elevated basal lipolysis, which apparently contributes to the observed decrease in total lipid content.
In the current study, in comparison to 20E and TU, PA had no effect on adipogenesis and adipolysis, which have not been reported to our knowledge. Despite the common ecdysteroid structure in the investigated natural compounds, we could suggest that the observed difference in biological response is attributed to the lack of hydroxyl group on position 25 in PA, compared to 20E and TU.
The results of the present study demonstrated that RCE and 20E exhibit anti-obesity potential by reducing adipogenesis and promoting lipolysis in human adipocytes, while turkesterone promotes only adipogenesis. Further investigation is needed to fully understand the mechanism and affirm its potential therapeutic applications. Nevertheless, the findings highlight the importance of exploring the diversity of plant metabolites for drug discovery and development and suggest that R. carthamoides could be a promising source of natural anti-obesity agents or a combination of the most abundant secondary metabolites.

Conclusions
In summary, the developed and validated HPTLC method was demonstrated to be reliable for the estimation and quantification of the PDs 20E, PA, and TU. Using the developed HPTLC method, identification and quantification of these PDs in RCE were performed. Additionally, evaluation of anti-adipogenic activity revealed that RCE, 20E, and TU considerably decreased lipid accumulation in human adipocytes. Further experiments indicated that RCE and 20E significantly stimulated basal lipolysis, while no effect was observed upon PA and TU application. The obtained results from RCE, 20E, and TU are worth further mechanistic evaluation, which would provide a scientific rationale for subsequent in vivo experiments.