Ursolic acid, a key anticancer compound derived from Prunella vulgaris L., induces apoptosis in HepG2 cells by directly targeting the STAT3 signaling pathway

2.1 Chemicals and reagents

UA was sourced from MedChemExpress (Monmouth Junction, NJ, USA), the AnnexinV-FITC/PI Apoptosis Kit was obtained from BD (San Diego, CA, USA), the GreenNuc™ Caspase-3 Assay Kit was acquired from Beyotime Biotechnology (Haimen, China). Additionally, caspase 3 (14220), β-actin (4970s), Bcl2 (3498t), Bax (2772t), cleaved caspase 9 (20750), STAT3 (4904t), phospho-STAT3 (Tyr705) (9145t), cleaved-caspase 3 (9664), and Lamin A/C (2032) were purchased from CST (Danvers, MA, USA); Caspase 9 was obtained from Proteintech (66169-1-ig, Wuhan, Hubei, China). The ribo FECTtm CP Transfection kit was purchased from Ribobio (Guangzhou, China). HepG2 cells were procured from Haixing Biosciences (Suzhou, China), and Prunella vulgaris was acquired from the Pharmacy of TCM of the out-patient Department of Zhejiang Chinese Medical University (ZCMU). A voucher specimen (PV20230510) was stored in the Academy of Chinese Medical Sciences of ZCMU.

2.2 Preparation and HPLC characterization of the EEPV and the UA knockout fraction

Following pulverization, Prunella vulgaris powder was sieved (40-mesh), precisely weighed (20.0 g), then reflux-extracted with 400 mL 95% ethanol for 60 min. The 95% EEPV was obtained through filtration and concentration under reduced pressure. An appropriate amount of EEPV was taken, dissolved in 95% acetonitrile to prepare a 20 mg/mL solution. The Dionex semi-prepared high-performance liquid chromatography (HPLC) system (U3000, Sunnyvale, CA, USA) was used for UA knockout. UA and other remaining components were collected separately (Fig. S1) and concentrated to obtain UA and the EEPV knockout UA fraction (KO-UA).

The extracts were dissolved in 95% acetonitrile, and subsequently injected into a Dionex analysis HPLC system (Sunnyvale, CA, USA) that was equipped with a PDA detector. The determination was performed on a Thermo Scientific Acclaim^TM 120 C18 column (4.6 × 150 mm, 120 Å, 5 μm) at 30°C. The elution process began with an initial composition of 5% ACN/95% aqueous phase (0.1% HCOOH) for the first 5 min, which then gradually increased to reach full acetonitrile concentration over the next 15 min, and 100% acetonitrile elution was maintained for 5 min. Then, acetonitrile was decreased from 100% to 5% within 1 min, and finally balanced for 10 min. The flow rate was maintained at 1 mL/min, with UV detection performed at 210 nm.

2.3 Cell proliferation inhibition assay

HepG2 cells were inoculated into 96-well plates at a density of 1 × 10⁴ cells per well. Following an overnight incubation, extracts of varying concentrations were introduced, and the culture was continued for an additional 24 h using 0.1% DMSO as a control. After discarding the supernatant, each well received 100 μL of diluted CCK-8 solution and was incubated at 37°C for 2 h. Optical density (OD_{450 nm}) was quantified using a multifunctional enzyme labeler (PEEnspire, USA), from which cell survival rate and IC₅₀ were calculated.

For the EdU assay, the UA-treated cells were stained following the methodology we previously reported (Pan et al. 2021). Ultimately, images of the cells were captured using a high-content confocal cell imaging system (Molecular Devices, Sunnyvale, USA). Quantitative analysis of EdU-positive cells was performed through systematic evaluation of three randomly selected microscopic fields per experimental group, using IPP9.0 (Media Cybernetics, Rockville, MD, USA). And the proportion of EdU-positive cells in each group was calculated as (the number of EdU-positive cells/the total number of cells) × 100%.

2.4 Cell apoptosis assay

HepG2 cells in 24-well plates were subjected to varying concentrations of UA for 24 h, with 0.1% DMSO serving as the control. Subsequently, cell apoptosis was analyzed by flow cytometry according to the method we previously reported (Pan et al. 2021).

Additionally, HepG2 cells in 96-well plates were exposed to different concentrations of UA for 24 h. Then, the caspase-3 probe was added in accordance with the instructions provided by the GreenNuc™ Caspase-3 Assay Kit and incubated at 37°C for 30 min. Images were captured using a high-content confocal cell imaging system. Three images per group were randomly selected to analyze the apoptotic cell ratio using IPP9.0. The proportion of caspase 3 active cells in each group was calculated using the formula: (number of caspase 3 active cells/total cell count) × 100%.

2.5 Network pharmacological analysis and molecular docking assay

UA’s canonical SMILES was acquired from PubChem (CID: 64945) for Swiss Target Prediction analysis. Traditional Chinese Medicine Systems pharmacology database and analysis platform (TCMSP) provided supplementary targets. HCC-associated targets were curated using Genecards (v4.14) and DisGeNET (v7.0) with “hepatocellular carcinoma” as the key query. Venny 2.1 processed intersections as potential anti-HCC targets. The UA-target-HCC network was visualized in CytoScape 3.10.1. UA-HCC intersecting targets were subjected to PPI network construction on STRING (v11.5, confidence score >0.7). Network topology analysis via CentiScaPe (v2.2) pinpointed disease-modulating hubs based on betweenness centrality >500.

UA’s 2D structure was acquired via PubChem (CID: 64945), while STAT3’s 3D conformation (PDB: 6NJS) was obtained from the Protein Data Bank. Both protein and compound structures were uploaded onto CB-Dock2, an online docking server developed by Yang C’s laboratory. Subsequently, the structure-based blind docking option was chosen for the molecular docking simulation (Liu et al. 2024).

2.6 Quantitative fluorescence PCR (qPCR) analysis

Drug-treated HepG2 cells underwent Trizol-based RNA extraction. cDNA was synthesized with a reverse transcription kit from TaKaRa. Quantitative PCR analysis was conducted on an Applied Biosystems 7500 Real-Time PCR System (Foster City, CA, USA), using specific primers (see Table 1). Transcript quantification utilized the 2^-⊿⊿Ct method (Pan et al. 2021).

2.7 Immunofluorescence analysis

HepG2 cells were subjected to drug treatment for a duration of 24 h. Subsequently, the cells were fixed using 4% paraformaldehyde, then permeabilized with 0.1% Triton X-100. Thereafter, the cells were blocked at room temperature for 1 h with 10% FBS and incubated overnight at 4°C with an anti-STAT3 antibody (1:400, cst4904t). After washing, cells were incubated at room temperature for 2 h with a secondary antibody (ab60314, Abcam) and were stained for 10 min using Hoechst dye at a concentration of 10 μM, and images were captured by a high-content confocal cell imaging system to analyze the translocation of STAT3 to the nucleus.

2.8 Western blot analysis

Post-treatment cellular proteins were isolated with the M-PER Mammalian Protein Extraction Reagent kit (78503, Waltham, MA, USA). Protein quantification utilized the BCA method, with concentrations normalized by RIPA buffer solution (Beyotime Biotechnology, Haimen, China). Target protein levels were detected on a ProteinSimple simple western system (San Jose, CA, USA) using β-actin for normalization. Results were analyzed by Compass software 6.0.0.

2.9 Cellular thermal shift assay

HepG2 lysates were prepared by centrifugation (12000 rpm, 15 min). Aliquots received either: (a) 0.1% DMSO (vehicle control), or (b) 100 μM UA with 1-h incubation at 25°C. Then, they were transferred to PCR tubes and placed in an Applied Biosystems PCR instrument (Foster City, CA, USA), heated at different temperatures for 5 min. After centrifugation (12000 rpm, 10 min), supernatants were immunoprobed via Jess^TM automated Simple western system (Yoon et al. 2019).

2.10 RNA interference

The siRNA sequences of STAT3 were designed in accordance with the publication of Choi, H.J. (Choi et al. 2009). The sense strand: 5’ - AACUUCAGACCCGUCAACAAA dTdT- 3’, and the antisense strand: 5’ - UUUGUUGACGGGUCUGAAGUUdTdT - 3’. The siRNA was synthesized by Ribobio and transfected into HepG2 cells using the Ribo FECTtm CP Transfection kit. The cells were then cultured for 48 h before proceeding with the subsequent experiments.

2.11 Statistical analysis

The data are expressed as mean ± standard deviation and were analyzed and plotted using GraphPad Prism software (Version 6.0, Graphpad Software Inc., San Diego, CA, USA). Multi-group comparisons utilized one-way ANOVA with post-hoc Tukey test, whereas pairwise comparisons employed unpaired Student’s t-test. The statistical significance threshold was set at p < 0.05.

3.1 UA serves as a crucial active constituent of the ethanol extract from Prunella vulgaris that inhibits the growth of HepG2 cells.

Based on the determination via HPLC, the content of UA in Prunella vulgaris was as high as 5.872 mg/g, aligning with prior reports (Xiang et al. 2019). The 95% EEPV concentrated UA to 45.333 mg/g, establishing it as a principal constituent (Fig. 1a). To determine the contribution of UA to the inhibition of HepG2 cell proliferation in EEPV, we utilized semi-preparative liquid chromatography (Fig. S1) to perform component knockout of UA. The HPLC chromatogram of the KO-UA has been presented in Fig. 1(b), and the collected component of UA has been shown in Fig. 1(c), namely, the UA content in KO-UA significantly decreased, and the purity of the collected UA was also high. CCK-8 assays demonstrated concentration-dependent growth inhibition by EEPV, KO-UA, and UA (Figs. 1d-f). Notably, UA knockout markedly elevated EEPV’s IC₅₀ value from 110.933 ± 5.543 μg/mL to 366.167 ± 22.945 μg/mL (Fig. 1g), indicating diminished efficacy. While the purified UA exhibited potent activity (IC₅₀=9.315 ± 0.525 μg/mL; 20.396 μM) (Fig. 1g), EdU assays corroborated UA’s anti-proliferative effect: increasing UA concentrations progressively reduced EdU positive cell populations (Fig. 1h). Collectively, UA is identified as the pivotal bioactive constituent mediating Prunella vulgaris ‘s suppression of HepG2 proliferation.

Close

Fig. 1.

UA is a key active component in the 95% alcohol extract of Prunella vulgaris for inhibiting HepG2 cell proliferation. HPLC chromatograms of (a) 95% ethanol extract (EEPV) , (b) UA knockout fraction (KO-UA), and (c) UA are shown. HepG2 cells were treated with (d) EEPV, (e) KO-UA, and (f) UA at various concentrations for 24 hours, and cell viability was determined by CCK8 assay; (g) corresponding IC50 values were calculated. Furthermore, after HepG2 cells were treated with various concentrations of UA for 24 h, EdU staining was carried out. Subsequently, (h) the proportion of EdU-positive cells in each group was calculated as (the number of EdU-positive cells / the total number of cells) × 100% . Compared with the control group, *p < 0.05, **p < 0.01; significant differences observed among groups labeled with different letters (a, b, and c), p < 0.01.

Export to PPT

3.2 UA can induce apoptosis of HepG2 cells in a dose-dependent manner

To further explore the anti-cancer effects and mechanisms of UA in Prunella vulgaris, apoptosis induced by UA was analyzed using flow cytometry based on morphological changes observed in drug-treated cells, such as cell shrinkage. As depicted in Fig. 2(a), the rate of apoptosis in HepG2 cells significantly increased with escalating concentrations of UA treatment, reaching a total apoptosis rate of (32.753±3.742) % at a 30 μM concentration. To validate the pro-apoptotic properties of UA, staining analysis with GreenNuc™ Caspase-3 active probe was conducted. As depicted in Fig. 2(b), caspase-3 active cell frequency escalated concentration-dependently with UA treatment, attaining 47.039 ± 3.965% positivity at 30 μM. This concordance between assays confirms UA partially suppresses HepG2 proliferation by triggering programmed cell death. Notably, EEPV similarly elicited concentration-responsive apoptotic effects in HepG2 cells at 50-200 μg/mL (Fig. S2).

Close

Fig. 2.

UA triggers apoptosis in HepG2 cells via the mitochondrial pathway. HepG2 cells were treated with different concentrations of UA for 24 h, then the relative mRNA levels of (a) Bcl-2, (b) Bax, (c) caspase 9 and (d) caspase 3 were detected by qPCR. The relative protein expression levels of (e) Bcl-2, (f) Bax, (g) caspase-9, (h) cleaved-caspase-9, (i) caspase-3, and (j) cleaved-caspase-3, normalized to β-actin, (k) were analyzed using the Simple Western system. (l) Mitochondrial membrane potential was assessed by JC-1 staining. * p < 0.05, ** p < 0.01.

Export to PPT

3.3 UA triggers apoptosis in HepG2 cells by disrupting the intrinsic mitochondrial pathway

To decipher molecular mechanisms of UA-driven apoptosis in HepG2 cells, apoptosis-related genes were profiled. As Fig. 3 illustrates, UA concentration-dependently suppressed BCL2 transcription (Fig. 3a) while elevating BAX mRNA (Fig. 3b), without affecting caspase-9/3 expression. Western blotting confirmed dose-responsive Bcl-2 protein downregulation and Bax upregulation (Figs. 3e-k), indicating UA triggers mitochondrial apoptosis via modulating membrane permeabilization. Concomitantly, pro-caspase-9/3 diminished with parallel increases in cleaved forms (Figs. 3e-k), validating apoptotic cascade activation. Additionally, JC1 staining results (Fig. 3l) demonstrated that with increasing concentrations of UA, red fluorescence was notably diminished while green fluorescence was markedly enhanced, suggesting a reduction in mitochondrial membrane potential and damage to mitochondria. These findings suggest that UA treatment triggers endogenous mitochondrial pathways leading to apoptosis in HepG2 cells.

Close

Fig. 3.

UA triggers apoptosis in HepG2 cells via the mitochondrial pathway. HepG2 cells were treated with different concentrations of UA for 24 h, then the relative mRNA levels of (a) Bcl-2, (b) Bax, (c) caspase 9 and (d) caspase 3 were detected by qPCR. The relative protein expression levels of (e) Bcl-2, (f) Bax, (g) caspase-9, (h) cleaved-caspase-9, (i) caspase-3, and (j) cleaved-caspase-3, normalized to β-actin, were analyzed using the (k) Simple Western system. (l) Mitochondrial membrane potential was assessed by JC-1 staining. ^* p < 0.05, ^** p < 0.01.

Export to PPT

3.4 Network pharmacological screening of UA against core targets in HCC

The TCMSP and Swiss Target Prediction databases were used to obtain 126 targets of UA, while the Disgenet and Genecards databases provided 1991 targets related to HCC. By intersecting these datasets, 76 common targets for both UA and HCC were identified. Venny2.1 was utilized to create a Venn diagram (Fig. 4a) illustrating the potential anti-HCC target of UA. Furthermore, a visualization network illustrating the gene target information associated with UA and HCC was created using Cytoscape software, resulting in the establishment of the “UA-target-HCC” network (Fig. 4b). The shared target proteins were then imported into the STRING12.0 database to elucidate their interaction relationships, yielding a network comprising 20 nodes and 186 edges (Fig. 4c), where each node represents a protein and each edge signifies their relationship. Subsequently, using Cytoscape 3.10.1 for network topology analysis, a total of 20 core targets were selected, including peroxisome proliferator-activated receptor-alpha (PPARA), AR, CDK4, STAT3, Caspase-3, Bcl-2, etc. Gene ontology (GO) analysis performed on the set of common targets revealed significant enrichment in biological processes such as those related to Bcl-2 family protein complex and regulation of gene transcription expression; molecular functions including nuclear receptor activity, as well as cell components like transcription regulator complex and DNA-binding transcription activator activity (Fig. 4d). Moreover, KEGG analysis indicated that these target proteins are primarily enriched in cancer and apoptosis pathways (Fig. 4e).

Close

Fig. 4.

Network pharmacological analysis results. (a) Venn Diagram illustrating the overlap between UA and hepatocellular carcinoma targets, (b) UA-target-hepatocellular carcinoma network, (c) Protein-Protein Interaction (PPI) network of intersection target proteins for UA and hepatocellular carcinoma, (d) Gene Ontology (GO) analysis of intersection target proteins for UA and hepatocellular carcinoma, including biological process (BP), molecular function (MF), and cell component (CC) annotations; as well as (e) the top 20 pathways identified in KEGG analysis of intersection target proteins for UA and hepatocellular carcinoma.

Export to PPT

Close

Fig. 5.

UA inhibits the phosphorylation and nuclear translocation of STAT3 in a dose-dependent manner by binding to STAT3. (a) Molecular docking of UA with STAT3, (b) pre-dicted conformation of UA in the pocket of STAT and (c) the results of cellular thermal shift as-say. HepG2 cells were treated with various concentrations of UA for 24 hours. Subsequently, (d) total RNA was isolated and utilized to quantify STAT3 expression via qPCR. (e) Total protein extraction from the cells facilitated the determination of STAT3 and pSTAT3 levels using a sim-ple western system. Following cell fixation, immunofluorescence was employed to examine the subcellular localization of STAT3 within the nucleus (f). **, p < 0.01.

Export to PPT

The enriched target proteins, particularly those involved in apoptosis, have been validated through qPCR and western blot analyses. Specifically, UA treatment was found to reduce Bcl-2 levels while enhancing Bax expression. Simultaneously, cleaved caspase 9 and cleaved caspase 3 were markedly elevated, suggesting that the apoptosis of HepG2 cells induced by UA was associated with the interference in the mitochondrial pathway (Fig. 3). Notably, STAT3, which belongs to the STAT family of cytoplasmic transcription factors, plays a crucial role in various cellular functions that facilitate the transmission of signals from the plasma membrane to the nucleus. It has been widely recognized as a carcinogenic factor in various human cancers and plays a crucial role in HCC occurrence, development, metastasis, drug resistance, and immune evasion. Overactivation of STAT3 is observed in more than 60% of HCC tissues (He et al. 2010), making it one of the prominent targets for HCC drug development (Lee et al. 2019). Upon phosphorylation and activation leading to homodimer formation, STAT3 translocates into the nucleus and binds to the promoter sequences of anti-apoptotic genes such as BCL2, BCLxL, and survivin to up-regulate gene expression and inhibit apoptosis (Carpenter et al. 2014). Based on the aforementioned pharmacological analysis results presented in Fig. 4(c), STAT3 emerges as one of the core target proteins involved in interactions between UA and HCC. Furthermore, given that STAT3 regulates BCL2 expression and considering the dose-dependent inhibitory effect exhibited by UA on Bcl-2, it can be inferred that UA might promote apoptosis in HepG2 cells through the inhibition of STAT3 activation.

3.5 UA targets STAT3 and inhibits its activation into the nucleus

To explore the possible interaction between UA and STAT3, molecular docking methods were carried out to assess their binding activity. The findings indicated a robust binding affinity, with a calculated binding free energy of -8.0 kcal/mol for STAT3 and UA (Figs. 5a and b). Subsequent cellular thermal shift assay revealed that incubation of 100 μM UA with HepG2 cell lysate significantly enhanced the thermal stability of STAT3 protein (Fig. 5c), providing evidence that UA directly targets and binds to STAT3. Furthermore, qPCR and western blot analyses demonstrated that UA had no significant effect on STAT3 mRNA expression (Fig. 5d) or total STAT3 protein levels (Fig. 5e), but it dose-dependently reduced phosphorylated STAT3 levels (Fig. 5e). Immunofluorescence results further supported these findings by showing a significant decrease in nuclear STAT3 levels alongside increased cytoplasmic fluorescence intensity following treatment with increasing concentrations of UA (Fig. 5f). Collectively, these findings indicate that UA can inhibit the phosphorylation of STAT3 by directly binding to it, thereby impeding its translocation into the nucleus.

3.6 UA has the potential to trigger apoptosis in HepG2 cells by suppressing STAT3 activation and decreasing Bcl-2 expression

Numerous studies have demonstrated that STAT3 functions as a transcription factor, with Bcl-2 identified as its downstream regulatory target gene (Bhattacharya et al. 2005; Choi et al. 2009; Liu et al. 2017) Consequently, UA may suppress Bcl-2 expression by interfering with STAT3, which in turn leads to apoptosis in HepG2 cells. Following treatment with STAT3 siRNA, a marked reduction in the expression levels of STAT3 in HepG2 cells was observed (Fig. 6a), accompanied by notable reductions in pSTAT3 levels and BCL2 expression (Fig. 6a), leading to a substantial increase in apoptotic cell population (Fig. 6b). These observations confirm the role of the transcription factor STAT3 in regulating apoptosis through modulation of BCL2 expression within HepG2 cells. Furthermore, treatment with UA resulted in a dose-dependent reduction in the protein levels of phosphorylated STAT3 and Bcl-2, effectively promoting apoptosis of HepG2 cells through initiation of the mitochondrial apoptosis pathway (Fig. 5). These results indicate that UA has the potential to induce apoptosis in HepG2 cells by suppressing STAT3 activation and reducing Bcl-2 levels, thereby triggering intrinsic mitochondrial apoptotic pathways.

Close

Fig. 6.

UA treatment induced apoptosis in HepG2 cells, partially through the inhibition of the STAT3/Bcl-2 pathway. Following siRNA treatment for 24 h, HepG2 cells were collected. (a) Total protein was extracted for simple western analysis to assess changes in the expression of STAT3, pSTAT3, and Bcl-2. (b) Cell apoptosis was evaluated using flow cytometry with Annexin V-FITC/PI staining. Compared to the control group, ^**, p < 0.01

Export to PPT

3.7 The knockout of UA diminished the ability of EEPV to induce apoptosis via the STAT3/Bcl-2 pathway in HepG2 cells

To further explore the involvement of UA in the antitumor effects of EEPV, HepG2 cells were exposed to the IC₅₀ dose of EEPV, along with equivalent levels of UA and KO-UA. The inhibitory effects on STAT3 phosphorylation, Bcl-2 expression, and apoptosis were observed. The results demonstrated that EEPV, UA, and KO-UA+UA significantly reduced pSTAT3 and Bcl-2 levels (Fig. 7a). However, KO-UA did not effectively reduce STAT3 phosphorylation or Bcl-2 levels. Apoptosis detection results indicated a significant decrease in apoptosis rate after UA knockout, which was essentially restored after UA supplementation (Fig. 7b). These findings further support the important role of UA in EEPV’s interference with the STAT3/Bcl-2 pathway to induce apoptosis in HepG2 cells.

Close

Fig. 7.

UA plays a crucial role in EEPV by inhibiting STAT3 phosphorylation, reducing Bcl-2 expression, and inducing apoptosis of HepG2 cells. After 24 hours of treatment with various drugs, (a) total protein was extracted from HepG2 cells for simple western analysis. The relative ex-pression levels of STAT3, pSTAT3, and Bcl-2 were calculated using β-actin as the internal reference. (b) Apoptosis in HepG2 cells was assessed by flow cytometry with Annexin V-FITC/PI staining. Compared to the control group, *, p < 0.05, **, p < 0.01; Comparison between the two groups, &, p < 0.05, &&, p < 0.01.

Export to PPT