Takeda G Protein-Coupled Receptor 5 and Peroxisome Proliferator-Activated Receptor-Gamma Activation by Pinocembrin and Pinostrobin Isolated from Lindera sericea

Abstract

Lindera sericea var. sericea (Japanese common name: “Kekuromoji”) is a deciduous shrub belonging to the Lauraceae family. Mainly distributed in Japan and Korea, L. sericea has been traditionally used as a source of essential oils and has not been characterized as a medicinal plant. In this study, we aimed to isolate and identify compounds that activate Takeda G protein-coupled receptor 5 (TGR5) and peroxisome proliferator-activated receptor-γ (PPARγ). Bioactivities were evaluated using a dual-color real-time bioluminescence monitoring system. The methanolic extract of L. sericea showed significant dose-dependent TGR5 activation and modest PPARγ activation. Spectroscopic analysis identified rac-pinocembrin (rac-1) and rac-pinostrobin (rac-2) as the major bioactive compounds in the methanolic extract. Reporter assays revealed that rac-2 is a TGR5 activator, whereas both rac-1 and rac-2 are modest PPARγ activators. We then separated the enantiomers of pinocembrin (1) and pinostrobin (2) using chiral column chromatography. Both enantiomers of 2 contributed comparably to TGR5 activation, whereas each pair of enantiomers (1 and 2) exhibited similar activity within the same compound toward PPARγ. These findings suggest that L. sericea is a promising source of bioactive compounds with potential metabolic regulatory activity.

1. Introduction

Lindera sericea var. sericea (Japanese common name: “Kekuromoji”) is a deciduous shrub belonging to the Lauraceae family and is mainly distributed in Japan and Korea [1]. L. sericea is a source of essential oils and a cosmetic ingredient. However, it has never been studied as a medicinal plant. Therefore, it is important to identify its bioactive compounds and potential industrial applications.

Takeda G protein-coupled receptor 5 (TGR5), also known as G protein-coupled bile acid receptor 1 (GPBAR1), is a member of the G protein-coupled receptor (GPCR) superfamily specific for bile acids [2]. TGR5 is widely distributed across various organs and tissues, including the liver, gallbladder, intestine, kidney, spleen, brain, skeletal muscle, and brown adipose tissue [3,4]. Bile acids, such as lithocholic acid and deoxycholic acid, activate TGR5 in vivo, stimulating energy consumption in brown adipose tissue and skeletal muscle tissue, as well as mitochondrial oxidative phosphorylation [5]. In addition, the activation of TGR5 expressed in intestinal endocrine L cells promotes glucagon-like peptide 1 secretion, inhibiting liver fat production and maintaining glucose homeostasis [6]. Therefore, TGR5 activation helps prevent obesity and type II diabetes and is considered an essential biological target [7]. Conversely, synthetic TGR5 agonists such as INT-777 have been reported to exhibit undesirable side effects, including gallbladder filling, pruritus, and off-target metabolic responses [8,9]. In this regard, naturally occurring agonists are considered attractive alternatives to synthetic agonists owing to their structural diversity, multi-target potential, and favorable safety profiles [10,11,12]. A variety of natural compounds have been reported as TGR5 agonists. These compounds improve metabolic parameters in preclinical obesity models [13]. Some naturally occurring TGR5 agonists exhibit gut-restricted activity, potentially avoiding systemic side effects while retaining efficacy through local TGR5 activation in intestinal tissues [14]. Hence, there is a need to further explore naturally occurring TGR5 agonists to develop safer, more physiological approaches [13].

Peroxisome proliferator-activated receptor-gamma (PPARγ) is also a valuable biological target. A nuclear receptor predominantly expressed in adipose tissue, PPARγ plays a vital role as a master regulator of adipocyte differentiation [15]. PPARγ activation has been reported to increase the proportion of small adipocytes with high insulin sensitivity [16]. For these reasons, PPARγ agonists, such as pioglitazone and rosiglitazone, are used as insulin-sensitizing drugs for the treatment of insulin resistance in type II diabetes [17].

In this study, we investigated the major bioactive compounds in L. sericea to demonstrate its potential as a valuable plant resource. To evaluate TGR5 activation by L. sericea, we generated a reporter cell line using an artificial chromosome vector that stably expresses dual-color beetle luciferases. In addition, PPARγ activation was evaluated using a previously generated reporter cell line expressing dual-color beetle luciferases [18]. Dual-color real-time bioluminescence monitoring revealed that L. sericea methanolic extract activates both TGR5 and PPARγ. Spectroscopic analysis identified rac-pinocembrin (rac-1) and rac-pinostrobin (rac-2) as the major bioactive compounds in L. sericea. rac-2 exhibited clear TGR5 activation, whereas both rac-1 and rac-2 showed modest PPARγ activation.

2. Results and Discussion

2.1. Generation of Reporter Cell Line Expressing Dual-Color Beetle Luciferases for Monitoring TGR5 Activation

To evaluate TGR5 activation by L. sericea, we generated a reporter cell line from human hepatoma HepG2 cells using a multi-integrase mouse artificial chromosome (MI-MAC) vector, in which TGR5 is stably expressed under the control of the CMV promoter. TGR5 activation was monitored by CRE-dependent transactivation using SLR3, a red-emitting beetle luciferase [19,20]. ELuc, a green-emitting beetle luciferase [21,22] expressed under the control of the TK promoter, served as the internal control reporter (Figure 1a). To account for other endogenously expressed GPCRs, we also generated a negative reporter cell line from HepG2 cells that stably express the CRE-driven SLR3 reporter without exogeneous TGR5 (herein referred to as the TGR5(−) negative reporter cell line). In both cell lines, internal control reporter ELuc was employed to accurately correct the baseline of CRE-dependent transcription monitored by SLR3, ensuring precise assessment of TGR5-dependent and -independent activation dynamics.

First, to examine the responsiveness of the generated TGR5-expressing reporter cell line, we created concentration–response curves for INT-777 and ursolic acid, synthetic and natural TGR5 agonists, respectively [23], by simultaneously recording the luminescence intensities of SLR3 and ELuc in real time. TGR5 activation was expressed as fold change, with SLR3 luminescence intensity normalized to ELuc luminescence intensity. INT-777 (Figure 1b and Figure S1a) and ursolic acid (Figure 1c and Figure S2a) showed rapid and dose-dependent activation kinetics, with a peak appearing after approximately 5–7 h. The concentration–response curves were obtained by plotting the area under the curve (AUC) values of the dose-dependent activation kinetics at each concentration from 0 to 20 h. From the concentration–response curves, the EC₅₀ values of INT-777 and ursolic acid were estimated to be 5.0 and 5.8 µM, respectively (Figure 1b,c). Compared with other luciferase-based cell assay systems that reported EC₅₀ values of INT-777 (0.8 µM) [24,25] and ursolic acid (1.1 and 1.4 µM) [26,27], the real-time reporter system using the generated reporter cell line exhibited slightly higher EC₅₀ values; however, it consistently demonstrated sufficient responsiveness for evaluating TGR5 activation.

We also examined the responsiveness of the TGR5(−) negative reporter cell line using the adenylate cyclase activator forskolin (Figure S3). The EC₅₀ value of forskolin was estimated at 2.4 µM, indicating that this cell line also exhibits sufficient responsiveness for monitoring CRE-dependent transactivation via the CREB signaling pathway. As expected, compared with forskolin, INT-777 and ursolic acid showed no significant activation in the TGR5(−) negative reporter cell line (Figures S1b and S2b). These results further demonstrate the suitability of the stable TGR5-expressing reporter cell line for the specific assessment of TGR5 activation.

2.2. Evaluation of TGR5 and PPARγ Activation by L. sericea Methanolic Extract

Using the generated TGR5-expressing reporter cell line, we evaluated TGR5 activation by L. sericea methanolic extract. The methanolic extract exhibited TGR5 activation in a dose-dependent manner (Figure 2a and Figure S4a), although an irregular peak at around 30 h was detected at 100 µg/mL, which may be due to non-specific activation of TGR5. In contrast, no detectable activation was observed in the TGR5(−) negative reporter cell line (Figure S4b). We also evaluated PPARγ activation by L. sericea methanolic extract using a previously established GAL4-based reporter cell line, in which SLR3 was used as the reporter for PPARγ activation, and the green-emitting beetle luciferase SLG, constitutively expressed under the control of TK promoter, was employed as the internal control (Figure S5) [18]. The methanolic extract showed modest PPARγ activation at concentrations of 25 and 50 µg/mL (Figure 2b and Figure S6). These results suggest that L. sericea contains bioactive compounds capable of activating both TGR5 and PPARγ. Subsequently, we determined the chemical structures of the major bioactive compounds in L. sericea methanolic extract by spectroscopy.

2.3. Identification of Major Bioactive Compounds in L. sericea

Two major peaks were observed in the high-performance liquid chromatography (HPLC) chromatogram of L. sericea methanolic extract (Figure 3). We isolated these compounds using Reversed-Phase HPLC (RP-HPLC) to determine their chemical structures.

Compound 1 was isolated as a white powder. Its molecular formula, C₁₅H₁₂O₄, was determined from the HRESIMS ion peak at m/z 255.0655 ([M−H]⁻, calcd 256.0663). The ¹H nuclear magnetic resonance (NMR) spectrum of 1 showed several typical aromatic ring signals at δ_H 6.0–7.5 (H-6/H-8 and 2-phenyl). In addition, typical non-equivalent methylene signals of flavanone were observed at 2.83 and 3.09 (H-3). From 1D and 2D NMR spectra, MS data, and comparisons with data from the literature [28], the chemical structure of 1 was identified as pinocembrin (Figure 4).

Compound 2 was isolated as a white powder. Its molecular formula, C₁₆H₁₄O₄, was determined from the HRESIMS ion peak at m/z 269.0811 ([M−H]⁻, calcd 269.0819). The ¹H and ¹³C NMR spectra of 2 resembled those of 1. Compared with the molecular formula of 1, that of 2 has an additional CH₃ group. In addition, the presence of a methoxy group was indicated by a proton signal at δ_H 3.82 and a carbon signal at δ_C 55.7 (7-OCH₃). The heteronuclear multiple bond coherence correlation of the 7-OCH₃ proton to C-7 establishes that the methoxy group is attached to C-7. From 1D and 2D NMR spectra, MS data, and comparisons with data from the literature [29], the chemical structure of 2 was identified as pinostrobin (Figure 4).

To determine the absolute configurations of 1 and 2 at the C-2 position, we measured their specific rotations. Compounds 1 and 2 exhibited low specific rotations (${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{28}$ –3.7 and −1.6, respectively), suggesting that they exist in nearly racemic forms. In addition, both compounds were analyzed using a chiral column, and their (R)- and (S)-forms were present in a 1:1 ratio (Figures S13 and S14). From these analyses, we established that 1 and 2 were isolated as racemic (R/S) mixtures (Figure 4). We also analyzed the L. sericea methanolic extract using a chiral column to confirm whether 1 and 2 exist as racemic (R/S) mixtures in the methanolic extract. The peak areas of both compounds in the chiral HPLC chromatograms showed a 1:1 ratio (Figure S23). Therefore, both compounds exist as racemic (R/S) mixtures in the methanolic extract. We also performed quantitative analysis of rac-1 and rac-2 in the L. sericea methanolic extract and found that the amounts of rac-1 and rac-2 in the extract were 12.6 ± 0.1 and 32.1 ± 0.0 µg/mg, respectively. The amounts of rac-1 and rac-2 in the dried L. sericea sample were calculated to be 2.6 and 6.6 mg/g dry weight, respectively.

Flavanones, such as pinocembrin (1) and pinostrobin (2), are biosynthesized through the cyclization of chalcones mediated by chalcone isomerase [30,31]. Chalcone isomerase exhibits remarkable enantiospecificity for the cyclization of chalcones, producing only (S)-form flavanone [32,33]. For these reasons, flavanones exist as an (S)-form in various plants. Interestingly, in this study, we found that pinocembrin (1) and pinostrobin (2) exist as racemic (R/S) mixtures. Although we could not elucidate the mechanism, it is possible that in L. sericea plants, (S)-pinocembrin and (S)-pinostrobin undergo racemization under unknown plant conditions, including a novel biosynthetic pathway. It has also been reported that the cyclization of chalcone to flavanone can proceed spontaneously. However, the chalcone isomerase-catalyzed cyclization is 10⁷ times faster than the spontaneous one. In addition, the spontaneous cyclization of chalcone to flavanone produces a racemic (R/S) mixture. Therefore, for another reason, it is reasonable to assume that the relatively weak chalcone isomerase activity in L. sericea could lead to the biosynthesis of racemic (R/S) flavanones mainly via spontaneous cyclization reactions.

To evaluate the difference in TGR5 and PPARγ activation between the (R)- and (S)-forms, we performed the optical resolution of rac-1 and rac-2 using a preparative chiral column (Figure 4). The (R)- and (S)-forms of both compounds were quantitatively separated in almost equal amounts, yielding approximately 50 mg of each enantiomer from 100 mg of the racemic mixture, as described in the Materials and Methods Section.

2.4. Evaluation of TGR5 and PPARγ Activation by Rac-Pinocembrin and Rac-Pinostrobin

Next, we evaluated TGR5 and PPARγ activation by pinocembrin (1) and pinostrobin (2). First, we evaluated TGR5 activation by racemic (R/S) mixtures. rac-1 showed no TGR5 activation (Figure 5a and Figure S24). In contrast, rac-2 showed rapid and dose-dependent activation kinetics with a peak around 5 h (Figure 5b and Figure S25a), similar to those of INT-777 and ursolic acid (Figure 1b,c). This activation was not observed in the TGR5(−) negative reporter cell line (Figure S25b), indicating that rac-2 indeed activates TGR5 but not through other endogenously expressed GPCRs.

Regarding PPARγ activation, rac-1 showed clear activation kinetics in the concentration range of 13 to 30 µM, exhibiting a peak around 7–8 h (Figure 6a). However, at higher concentrations, a decrease in GAL4–UAS-dependent transcription monitored by SLR3 and an increase in the internal control reporter SLG intensity were observed (Figure S26a), resulting in an apparent reduction in the calculated fold change. Similarly, rac-2 exhibited very weak but statistically significant activation kinetics in the concentration range of 13 to 44 μM (Figure 6b); at higher concentrations, a decrease in SLR3 intensity and a concomitant increase in SLG intensity were also observed (Figure S26b).

In this study, we evaluated PPARγ activation using SLR3 and SLG luminescence kinetics from 0 to 20 h (see the Materials and Methods Section), because representative PPARγ activators, such as pioglitazone, exhibit a peak at 7–8 h, as reported previously [18]. Based on this defined method, we concluded that rac-1 and rac-2 have no PPARγ activity at high concentrations, since no statistically significant differences from the vehicle control were detected.

In the dual-color real-time monitoring system, an internal control reporter driven by a constitutive promoter reflects cell viability. Therefore, in this study, cytotoxic effects were not evaluated by a conventional assay but were inferred from changes in the intensity of the internal control reporter. As shown in Figure S26, no decrease in SLG intensity compared with vehicle control was observed even at high concentrations, suggesting that the lack of activation at high concentrations was not due to cytotoxic effects of the compounds. On the other hand, the decrease in SLR3 intensity at higher concentrations in both compounds could be considered a nonspecific side effect, potentially including reporter interference or off-target transcriptional effects impacting the reporter system, although the detailed mechanism remains unclear.

A series of experiments revealed that rac-2, identified as rac-pinostrobin, in L. sericea methanolic extract activates TGR5. This is the first report demonstrating TGR5 activation by pinostrobin. 5,7-Dihydroxy-6,4′-dimethoxyflavanone, a methoxyflavanone, was reported to exhibit high affinity in a docking study of the human adenosine A2a receptor, a model for TGR5 [34], thereby supporting the possibility that rac-pinostrobin (rac-2) functions as a TGR5 activator. Pinostrobin (2) has been reported to exhibit various biological activities, including antioxidant, anti-inflammatory, and anti-adipogenic effects [35]. Although the potency of TGR5 activation of rac-2 is moderate, the combination of these previously reported biological activities with the TGR5 activation identified in this study suggests that rac-pinostrobin (rac-2) may be a valuable nutraceutical candidate for the improvement of metabolic disorders. Furthermore, rac-1 and rac-2 were also identified as modest PPARγ activators, although their activation was visible only within a limited concentration range. Flavanols naringin and alpinetin have been reported to function as PPARγ activators [36,37,38]. Because of the structural similarities to these flavanols, it is possible that rac-pinocembrin (rac-1) and rac-pinostrobin (rac-2) showed PPARγ activation.

2.5. Evaluation of TGR5 and PPARγ Activation by Enantiomers of Pinocembrin and Pinostrobin

Finally, we evaluated TGR5 and PPARγ activation by the enantiomers of pinocembrin (1) and pinostrobin (2) to identify the enantiomers responsible for the observed activity. Regarding TGR5 activation, both (R)-2 and (S)-2 showed comparable and significant activity (Figure 7, Figures S27a and S28a), suggesting that both enantiomers contributed to TGR5 activation in the racemic mixture. Furthermore, neither enantiomer showed activity in the TGR5(−) negative reporter cell line (Figures S27b and S28b), confirming their specificity for TGR5 activation.

Regarding PPARγ activation, both enantiomers of pinocembrin (1) and pinostrobin (2) exhibited similar dose-dependent activation kinetics (Figure 8, Figures S29 and S30), indicating that, as observed for TGR5 activation, there was no apparent difference in PPARγ activation between the enantiomers. Taken together, the enantiomers of both compounds exhibit comparable activities toward TGR5 and PPARγ, suggesting that both enantiomers are biologically effective constituents in L. sericea.

The pre-clinical pharmacokinetics of chiral pinocembrin (1) and pinostrobin (2) have been investigated in male rats [39]. In that study, cardiomyocyte size was measured, and several biological activities, including α-glucosidase and α-amylase inhibitory activities, were also compared between the enantiomers of 1 and 2. (R)-1 prevented cardiac hypertrophy, whereas (S)-1 did not do so. (R)-2 exhibited stronger α-amylase inhibitory activity than (S)-2. Thus, this study showed differences in several biological activities between the enantiomers.

However, in this study, the activation of TGR5 and PPARγ by the chiral pinocembrin (1) and pinostrobin (2) exhibited similar response levels. This suggests that the binding sites of these compounds on TGR5 and PPARγ may be positioned in regions not strongly influenced by the chiral carbon of the flavanone scaffold. Additionally, previous studies have reported that PPARγ possesses a large ligand-binding pocket capable of accommodating a broad range of ligands [40], whereas TGR5 contains two distinct ligand-binding sites [2,13]. These structural features may account for the absence of apparent stereoselective differences in TGR5 and PPARγ activation between the enantiomers of pinocembrin (1) and pinostrobin (2). Overall, our findings provide new insight into the biological activities of chiral pinocembrin (1) and pinostrobin (2).

In this study, we revealed that pinocembrin (1) and pinostrobin (2) showed TGR5 or PPARγ activation using dual-color real-time bioluminescence monitoring. Since the assay system we developed enables monitoring of receptor activation kinetics triggered by these compounds, it provides a more accurate evaluation than conventional end-point assays that require cell lysis. Nevertheless, the biological activities identified in this study will still require further functional verification through cellular physiology studies and in vivo experiments.

3. Materials and Methods

3.1. Cell Culture

Human hepatoma HepG2 cells harboring a multi-integrase mouse artificial chromosome (MI-MAC) vector (a gift from Dr. M. Oshimura and Dr. Y. Kazuki, Tottori University) [41,42] for evaluating TGR5 activation were grown in Dulbecco’s Modified Eagle Medium (DMEM; FUJIFILM Wako, Osaka, Japan) supplemented with 10% fetal bovine serum (FBS; Merck, Darmstadt, Germany), 1% Minimum Essential Medium Non-Essential Amino Acids (MEM NEAA; Thermo Fisher Scientific, Waltham, MA, USA), and 1 mM sodium pyruvate (Thermo Fisher Scientific) at 37 °C in a 5% CO₂ atmosphere.

Mouse fibroblast A9 cells harboring an MI-MAC vector for evaluating PPARγ activation generated in a previous study [18] were grown in DMEM (FUJIFILM Wako) supplemented with 10% FBS (Merck) at 37 °C in a 5% CO₂ atmosphere.

3.2. Plasmid Construction

To construct a reporter vector for evaluating TGR5, a reporter cassette containing three tandem repeats of cAMP response element (CRE; 5′-TGACGTCA-3′), herpes simplex thymidine kinase (TK) promoter, and red-emitting beetle luciferase Stable Luciferase Red 3 (SLR3; Toyobo Co., Osaka, Japan) [19] from Phrixotrix hirtus [20] was synthesized by GenScript (Tokyo, Japan). The cassette was ligated into the HindIII/StuI site of phiNeo_imaimai vector (a gift from Dr. S. Suzuki, Kagawa University) [43], which carries the φC31 attB site, chicken HS4 insulators, and neomycin resistance gene, resulting in pCRE-SLR3. Codon-optimized human TGR5 (NP_001070659.1) was synthesized by GenScript and ligated into the HindIII/XhoI site of pcDNA3 (+) (Thermo Fisher Scientific). An expression cassette containing CMV promoter, TGR5, and late polyA signal was amplified by PCR using 5′-ACTAATGAGAGGAAGGTACCGATGTACGGGCCAGATATAC-3′ as the forward primer and 5′-CCTCCCTCTTAGATCTCCCCAGCATGCCTGCTATTG-3′ as the reverse primer. The amplified fragment was cloned into the KpnI/BglII site of pCRE-SLR3 using an In-Fusion cloning kit (Takara Bio Inc., Shiga, Japan) in accordance with the manufacturer’s instructions, resulting in pCRE-SLR3::CMV-TGR5.

3.3. Generation of Stable Cell Line for Evaluating TGR5 Activation

To generate a stable cell line for evaluating TGR5 activation, internal control reporter plasmid pTK-ELuc-PEST-R4-Bsd [44] carrying TK promoter and PEST-fused green-emitting Emerald Luc [21] (ELuc; Toyobo) from Pyrearinus termitilluminans [22] was integrated into HepG2 cells harboring the MI-MAC vector [41,42] by co-transfection with R4 integrase expression plasmid pCMV-R4 [45] and subcultured for selection with 2 μg/mL blasticidin (Thermo Fisher Scientific), resulting in TK-dELuc/HepG2 cells. pCRE-SLR3::CMV-TGR5 was further integrated into the TK-dELuc/HepG2 cells by co-transfection with φC31 integrase expression plasmid pCMV-φC31 [45] and subcultured for selection with 1 mg/mL G418 (Nacalai Tesque, Kyoto, Japan), resulting in CRE-SLR3::CMV-TGR5::TK-dELuc/HepG2 cells. To generate negative control cells that do not express TGR5 (TGR5(−) negative reporter cell line), pCRE-SLR3 was integrated into the TK-dELuc/HepG2 cells by co-transfection with pCMV-φC31 and subcultured for selection with 1 mg/mL G418 (Nacalai Tesque), resulting in CRE-SLR3::TK-dELuc/HepG2 cells.

3.4. Dual-Color Real-Time Bioluminescence Monitoring

The generated stable HepG2 reporter cell line was used to evaluate TGR5 activation. For PPARγ activation, A9 cells harboring the MI-MAC vector generated in a previous study were used (Figure S5) [18]. Both reporter cell lines were seeded onto a 96-well white clear-bottom plate (Thermo Fisher Scientific) at 3 × 10⁴ cells/well. After incubation for one day, the medium was replaced with DMEM without phenol red (Thermo Fisher Scientific) containing 10% charcoal-stripped FBS (Serana Europe GmbH, Brandenburg, Germany), 25 mM HEPES (pH 7.0; Nacalai Tesque), and 300 µM D-luciferin potassium salt (RESEM BV, Linden, Netherlands), containing 1% Glutamax (Thermo Fisher Scientific) for HepG2 cells or 4 mM L-glutamine (Nacalai Tesque) for A9 cells. INT-777 (BLD Pharmatech Ltd., Shanghai, China) and ursolic acid (Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) were used as positive controls for TGR5 activation. Forskolin (Enzo Life Sciences, Inc., Farmingdale, NY, USA) and pioglitazone (FUJIFILM Wako) were used as positive controls for CREB pathway activation and PPARγ activation, respectively. Bioluminescence was recorded in real time for 5 s at 30 min intervals in the absence or presence of an R62 long-pass filter (HOYA, Tokyo, Japan) at 37 °C in a 5% CO₂ atmosphere under saturated humidity using a microplate-type luminometer (WSL-1563 Kronos HT, ATTO, Tokyo, Japan). After reducing noise in raw data using Fourier transform [46], ELuc or SLG, and SLR3 luminescence intensities were calculated as previously reported [47]. Fold-change values were calculated using the following equation: Fold change = [(SLR3 intensity of treated sample/ELuc or SLG intensity of treated sample)/(SLR3 intensity of untreated sample/ELuc or SLG intensity of untreated sample)]. Dose-dependent activation kinetics were assessed by plotting the fold change in luminescence over time. The concentration–response curves for TGR5 and PPARγ activation were constructed by plotting the AUC values of the activation kinetics (integrated from 0 to 20 h) against the corresponding sample concentration. The concentration–response curve for internal control, including cytotoxicity evaluation, was generated by plotting the AUC of ELuc or SLG luminescence kinetics (integrated from 0 to 20 h) against concentration. Luminometer performance, including the dynamic range and the intermediate precision, was verified in accordance with ISO 24421:2023 [48].

3.5. Plant Material

L. sericea was collected in Kochi, Japan, from June to October 2023. The voucher numbers of the specimens are 20230904A–20231113A.

3.6. General Experimental Procedures for Spectral Analysis

Optical rotations were measured using a DIP-1000 digital polarimeter (Jasco, Tokyo, Japan). 1D and 2D NMR spectra were acquired on a Bruker AVANCE NEO (400 MHz) spectrometer (Billerica, MA, USA), with chemical shifts expressed in ppm. The NMR spectra were referenced to residual solvent peaks (CDCl₃: ¹H NMR 7.26 ppm, ¹³C NMR 77.0 ppm). High-resolution electrospray ionization mass spectrometry (HRESIMS) spectra were acquired on a Thermo Fisher Scientific Q-Exactive HR-ESI-Orbitrap-MS. RP-HPLC separations were performed using a PU-4086 Semi-preparative pump (Jasco), a UV-4570 UV/Vis detector (Jasco), a Capcell Pak UG 120 C18 column (5 μm, 20 × 250 mm, Osaka Soda, Osaka, Japan), a Capcell Pak UG 120 C18 column (5 μm, 10 × 250 mm, Osaka Soda), and HPLC-grade solvents. For analytical HPLC, a DGU-20A₃ Prominence Degasser (Shimadzu, Kyoto, Japan), an LC-20AB Liquid Chromatograph (Shimadzu), an SIL-20AC Prominence Autosampler (Shimadzu), a CTO-20AC Prominence Column Oven (Shimadzu), and an SPD-M40 Photodiode Array Detector (Shimadzu) were used. Data were analyzed using LabSolutions software (version 5.111, Shimadzu).

3.7. Extraction and Fractionation

The dried leaves and stems of L. sericea (leaf:stem = 3:7, 240 g) were extracted with 80% methanol (2.4 L) overnight. The supernatant was collected, and the residue was re-extracted with 80% methanol (2.4 L) under stirring at 70 °C for 3 h. After collecting the supernatant, the residue was further re-extracted with 80% methanol (2.4 L) with sonication at 60 °C for 1 h. The collected supernatants were combined and filtered, and the filtrate was concentrated under reduced pressure to give the concentrated methanolic extract (49.1 g). A portion of the methanolic extract (43.8 g) was suspended in H₂O (500 mL) and successively partitioned with CHCl₃ (2 × 500 mL) to furnish the CHCl₃-soluble extract (11.8 g). The CHCl₃-soluble extract (11.8 g) was subjected to middle-pressure liquid chromatography using a Biotage^® Selekt system equipped with a Sfär C18 D-Duo 100 Å 30 μm 400 g column (Biotage, Uppsala, Sweden) with H₂O/acetonitrile (50:50 (0.5 column volume (CV))–50:50–0:100 (3 CV)–0:100 (1 CV), 0.1% trifluoroacetic acid (TFA))—2-propanol (1.5 CV)—methanol (1.5 CV) to yield nine fractions (fr. 1, 2.4 g; fr. 2, 1.0 g; fr. 3, 540 mg; fr. 4, 1.5 g; fr. 5, 830 mg; fr. 6, 520 mg; fr. 7, 750 mg; fr. 8, 210 mg; fr. 9, 2.3 g). Fraction 2 was subjected to preparative RP-HPLC with H₂O/acetonitrile (45:55, 0.1% TFA) as the eluent to give rac-pinocembrin (rac-1; 510 mg). Fraction 4 was subjected to preparative RP-HPLC with H₂O/acetonitrile (35:65, 0.1% TFA) as the eluent to give rac-pinostrobin (rac-2; 800 mg).

rac-Pinocembrin (rac-1): white powder; ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{28}$ –3.7 (c 0.22, acetone); ¹H NMR (CDCl₃, 400 MHz), δ_H = 12.04 (s, 1H, 5-OH), 7.48–7.36 (m, 5H, 2-phenyl), 6.01 (brs, 2H, H-6/H-8), 5.43 (dd, J = 13.0, 3.0 Hz, 1H, H-2), 3.09 (dd, J = 17.2, 13.0 Hz, 1H, H-3), 2.83 (dd, J = 17.2, 3.0 Hz, 1H, H-3); ¹³C NMR (CDCl₃, 100 MHz), δ_C = 195.8 (C-4), 164.5 (C-7), 164.4 (C-5), 163.2 (C-9), 138.3 (C-1′), 128.91 (C-4′), 128.88 (C-3′/C-5′), 126.1 (C-2′/C-6′), 103.2 (C-10), 96.8 (C-6), 95.5 (C-8), 79.2 (C-2), 43.3 (C-3); HRESIMS m/z 255.0655 [M−H]⁻ (calcd for C₁₅H₁₁O₄, 255.0663).

rac-Pinostrobin (rac-2): white powder; ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{28}$ –1.6 (c 0.27, acetone); ¹H NMR (CDCl₃, 400 MHz), δ_H = 12.02 (s, 1H, 5-OH), 7.49–7.37 (m, 5H, 2-phenyl), 6.08 (d, J = 2.3 Hz, 1H, H-6), 6.07 (d, J = 2.3 Hz, 1H, H-8), 5.43 (dd, J = 13.0, 3.0 Hz, 1H, H-2), 3.82 (s, 3H, 7-OCH₃), 3.09 (dd, J = 17.2, 13.0 Hz, 1H, H-3), 2.83 (dd, J = 17.2, 3.0 Hz, 1H, H-3); ¹³C NMR (CDCl₃, 100 MHz), δ_C = 195.8 (C-4), 168.0 (C-7), 164.2 (C-5), 162.8 (C-9), 138.4 (C-1′), 128.9 (C-3′/C-4′/C-5′), 126.1 (C-2′/C-6′), 103.2 (C-10), 95.2 (C-6), 94.3 (C-8), 79.2 (C-2), 55.7 (7-OCH₃), 43.4 (C-3); HRESIMS m/z 269.0811 [M−H]⁻ (calcd for C₁₆H₁₃O₄, 269.0819).

3.8. Quantitative Analysis Using HPLC

The methanolic extract of L. sericea was analyzed under the following conditions: column, Capcell Pak UG 120 C18 column (5 μm, 4.6 × 250 mm, Osaka Soda); flow rate, 1.0 mL/min; mobile phase, H₂O/acetonitrile with 0.1% TFA: gradient, 50:50 (0 min)–0:100 (20 min)–0:100 (30 min); and detection wavelength, 280 nm. Isolated compounds were used as standards to obtain calibration curves. The limit of detection and the limit of quantification of rac-1 and rac-2 were 3 and 10 ng/mL, respectively (defined as signal-to-noise ratios of 3 and 10, respectively). The extracts were analyzed independently three times, and the standard deviation (SD) was calculated. To confirm compound loss during sample preparation, a recovery test was conducted for rac-2. Because the recovery rate was 103.1 ± 0.1%, the quantitative values were not corrected.

3.9. Optical Resolution of Rac-Pinocembrin and Rac-Pinostrobin

rac-Pinocembrin (rac-1) and rac-pinostrobin (rac-2) were separated optically using a CHIRALPAK IG chiral column (5 μm, 10 × 250 mm, Daicel, Osaka, Japan). rac-1 (100 mg) was subjected to preparative HPLC with H₂O/acetonitrile (45:55, 0.1% TFA) as the eluent to give (R)-1 (44 mg) and (S)-1 (48 mg). rac-2 (100 mg) was subjected to preparative HPLC with H₂O/acetonitrile (15:85, 0.1% TFA) as the eluent to give (R)-2 (51 mg) and (S)-2 (49 mg).

(R)-Pinocembrin ((R)-1): ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{28}$ +43.0 (c 0.18, acetone).

(S)-Pinocembrin ((S)-1): ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{28}$ –38.5 (c 0.19, acetone).

(R)-Pinostrobin ((R)-2): ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{28}$ +41.7 (c 0.21, acetone).

(S)-Pinostrobin ((S)-2): ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{28}$ –36.6 (c 0.21, acetone).

3.10. Statistical Analysis

All assays were conducted in triplicate. Data are expressed as means ± SD. Comparisons were performed by one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test. GraphPad Prism 10.1.2 (GraphPad Software Inc., San Diego, CA, USA) was used for statistical analysis, and p-values less than 0.05 were considered statistically significant.

4. Conclusions

In this study, we demonstrated that L. sericea contains biologically active flavanones capable of modulating TGR5 and PPARγ activation. By establishing dual-color bioluminescence reporter cell lines to evaluate TGR5 activation, we characterized receptor-specific activation kinetics in real time. We identified rac-pinostrobin (rac-2) as a previously unrecognized natural TGR5 activator. In addition, we showed that rac-pinocembrin (rac-1) and pinostrobin (rac-2) exhibit modest PPARγ activation. The presence of racemic mixtures of 1 and 2 suggests unique biosynthetic features in L. sericea, potentially involving relatively weak chalcone isomerase activity. Furthermore, we examined TGR5 and PPARγ activation by the separated enantiomers of 1 and 2 and found that all separated enantiomers exhibited activation profiles comparable to those of the corresponding racemic mixtures. Taken together, our findings suggest that L. sericea is a promising functional plant resource with potential medicinal applications and provide new insights into the biological potential of its flavanone constituents.