Introduction

Hypertension (HTN) is a chronic condition characterized by sustained elevation in systemic arterial blood pressure and is the predominant cause of mortality and morbidity global scale.^1–3 The pathogenesis of HTN is complex, and includes dietary, environmental and genetics.⁴ The predominant pathophysiological features of HTN are microvascular rarefaction and endothelial dysfunction.⁵ Clinical studies have shown that endothelial fibrinolysis induced by endothelial cell injury decreases in hypertensive individuals.⁶

Vascular inflammation is an early manifestation and the pathophysiological basis of HTN.⁷ Endothelial cell-cell interactions shield tissues and organs from edema and unchecked infiltration of inflammatory cells, which are vital for vascular integrity and function.⁸ Angiotensin II (Ang II) is the main bioactive peptide of the renin/angiotensin system, which causes oxidative stress, vascular inflammation, and cell apoptosis, leading to the dysfunction of vascular endothelial cells and is closely related to the occurrence and development of HTN.^9,10 Specifically, AngII could stimulate the production of thromboxane A2 (TXA2), endothelin 1 (ET-1), inducible nitric oxide synthase (iNOS), which increased platelet aggregation and vasoconstriction, elevated vascular permeability and promoted smooth muscle cell proliferation, and contributed to the formation of peroxynitrite, respectively, exacerbating endothelial damage and promoting inflammatory responses.¹¹ Furthermore, an increase in inflammation and apoptotic cells further aggravates metabolic disorder.¹²

Sphingolipids are essential components of cell membranes and organelles that contribute to maintaining cell homeostasis and regulating cell development, differentiation, aging, and programmed cell death.¹³ Furthermore, sphingolipids can retain endothelial cells in a normal state by providing conditions to maintain normal vascular tension, permeability, and oxidative stress responses.¹⁴ When sphingolipid metabolism is disrupted, endothelial cells are inevitably injured, which leads to various cardiovascular disorders.¹⁵ Therefore, the regulation of sphingolipid metabolism to inhibit inflammation and apoptosis of endothelial cells is expected to provide a new avenue for the treatment of HTN.

Gastrodiae Rhizoma is the dried tuber of Gastrodia elata Bl., of the family Orchidaceae called tianma (TM) in Chinese, a popular traditional medicine with a wide range of biological activities, including circulatory regulation, anti-inflammatory, anti-oxidant, anti-ageing, anti-viral, and anti-tumour effects.^16,17 Studies have shown that Gastrodiae Rhizoma can attenuate HTN and ameliorate hypertension-induced inflammation.¹⁸ Lycii Fructus is the dried mature fruit of Lycium barbarum L. of the family Solanaceae, which is known in China as gouqizi (GQZ). Modern pharmacological studies have shown that GQZ has a variety of pharmacological effects, such as strengthening immunity, enhancing hypotensive action, improving memory, and suppressing tumor growth.¹⁹ Studies have indicated that the polyphenolic compounds in GQZ, particularly GA, can not only alleviate hypertensive disease by dilating thoracic aortic vessels in both endothelium-dependent and non-endothelium-dependent manners,²⁰ but can also significantly reduce blood pressure in spontaneously hypertensive rats by inhibiting angiotensin-converting enzyme, and have comparable potency to the pharmacological effects of captopril.²¹ However, whether regulating sphingolipid metabolism could inhibit the inflammation and apoptosis of endothelial cells and the underlying mechanisms remain unclear.

Human Umbilical Vein Endothelial Cells (HUVECs) are a widely accepted and well-established model for studying endothelial cell function and injury. They closely resemble the endothelial cells found in human blood vessels and are particularly useful for investigating mechanisms underlying vascular injury and inflammation.^22,23 HUVECs are known to exhibit characteristic responses to Ang II, such as increased oxidative stress, inflammation, and apoptosis.^11,24,25 This makes them a suitable model for assessing the protective effects of pharmacological agents against Ang II–induced endothelial damage.

In the current study, the integration of network pharmacology and molecular docking was initially used to explore the bioactive components, potential targets, and pathways of TM and GQZ in the treatment of HTN. The synergistic protective effects of gastrodin and gallic acid were evaluated in Ang II–induced Human umbilical vein endothelial cells (HUVECs) injury. Moreover, the Chou-Talalay model was used to assess synergistic effects, and the vital cell transduction pathway related to sphingolipids was experimentally verified by Western blotting and molecular dynamics simulation. This study provides evidence for the elucidation of the synergistic effects of a combination of gastrodin and gallic acid from the perspective of sphingolipid metabolism in endothelial cells.

Materials and Methods

Materials

Human umbilical vein endothelial cells (HUVECs), DMEM medium, F12-K medium, penicillin/streptomycin solution, fetal bovine serum (FBS), the kits for measuring Nitric Oxide Content (NO), lactate dehydrogenase (LDH), interleukin-6 (IL-6), interleukin-1β (IL-1β), Tumor Necrosis Factor Alpha (TNF-α), Endothelin-1 (ET-1), Thromboxane B2 (TX-B2), inducible nitric oxide synthase (iNOS) and the secondary antibody for fluorescence were purchased from Sangon Biotech (Shanghai, China). The Annexin V-FITC apoptosis detection kit was purchased from Solarbio (Beijing, China). Gastrodin (GAS), 4-Hydroxybenzyl alcohol (PHE), Parishin A (PA), Lycium barbarum polysaccharides (LBP), luteolin (LUT), and gallic acid (GA) (purity ≥ 98%) were purchased from Jingzhu Biotech (Nanjing, China). The anti-Caspase-3 antibody and anti-cleaved caspase-3 antibodies were purchased from Cell Signaling Technology (USA). The anti-EDG1, anti-ROCK2, anti-phospho-NF-κB p65 (Ser536), and anti-β-actin antibodies were purchased from Bioss Biotech (Beijing, China). DNase/RNase-free Water, TransZol Up Plus RNA kit, all-in one first-strand cDNA synthesis SuperMix kit, and SYBR PCR Master Mix were purchased from TransGen (Beijing, China).

Compounds Screen and Targets Prediction

The bioactive compounds were identified using TCMSP, TCMIP, and SwissTarget Prediction database with the keywords “Gastrodiae Rhizoma” and “Lycii Fructus”, and were filtered based on oral bioavailability (OB) ≥30% and drug-likeness (DL) ≥0.18. The bioactive compounds reported in the literature^26–29 were also collected. The bioactivity and relevance of these compounds based on their reported effects and interactions as documented in the literature. Furthermore, compounds, such as gastrodin and Lycium barbarum polysaccharide with clear pharmacological effects were chosen for additional investigation. SEA (https://sea.bkslab.org) and STITCH (http://stitch.embl.de) databases were used to screen the targets of these bioactive compounds. Genecards (https://www.genecards.org/), DisGeNET (https://www.disgenet.org/home/), OMIM (https://www.omim.org/), Malacard (https://previous.malacards.org/), and TTD (https://db.idrblab.net/ttd/) databases were used to search for hypertension-related targets. We employed search terms related to HTN, such as “hypertension”, “high blood pressure”, and assessed the relevance of each target by evaluating the strength of the evidence supporting its involvement in HTN. This included reviewing published literature, experimental validation, and known interactions with antihypertensive drugs.

Construction of Protein-Protein Interaction (PPI) Network

To better analyze the functional interaction between targets, the common targets of TM-GQZ and HTN were input into the STRING online platform (version 12.0) (https://string-db.org/), selected “Multiple proteins”, and limited species to “Homo sapiens”. We used a minimum interaction score of >0.400 to filter the results based on the strength of interaction evidence, as this threshold was chosen to balance between sensitivity and specificity in identifying meaningful interactions. The intersection targets were input into Cytoscape 3.7.2 software to construct the PPI, and Cytohubba was used to identify the densely connected network to screen the target correlation.

Enrichment Analysis

Common drug-disease targets were entered into the DAVID database (https://david.ncifcrf.gov/) for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis. “Homo sapiens” was selected as the species. Biological processes (BP), cellular components (CC), and molecular functions (MF) were screened at P<0.05, and the obtained items were sorted according to the P value. The top 10 items were visualized using R software 3.4.1.

Molecular Docking

Molecular docking can effectively predict the possibility of a compound binding to a target. AutoDock-Vina software was used for semi-flexible docking. Semi-flexible docking offers a balance of efficiency and accuracy by allowing ligand flexibility while keeping the protein rigid, which is computationally less demanding than fully flexible docking. The top five target proteins in the above analysis were verified by molecular docking with the core active components. The 3D structure of the target protein and chemical structures of GAS, PHE, PA, LBP, LUT, and GA were obtained from the RCSB PDB protein database (https://www.rcsb.org/) and PubChem database (https://pubchem.ncbi.nlm.nih.gov/), respectively, the details of structure IDs which are available in Supplementary Tables 1 and 2. AutoDock Tools software was used to prepare input files, set docking parameters, and analyze results for our protein-ligand simulations, and the Autogrid plug-in was used to obtain the docking active sites for molecular docking to determine the affinity. Here, we used PyMOL v.0.99 software to visualize the 3D structures of docking results.

Cell Culture and Drug Treatment

The HUVECs were seeded with 1.2×10⁵ cells/mL and cultured in F-12K Medium containing 10% FBS, 0.1 mg/mL heparin, and 30 µg/mL ECGS at 37 °C with 5% CO₂.³⁰ GAS, PHA, PA, LBP, LUT, and GA were dissolved in DMSO at concentrations of 100 mg/mL. The cells were treated with 1 μM Ang II for 12 h and then treated with the above compounds.

Cell Viability Assay

HUVECs were grown in 96-well plates at a density of 1.2×10⁴ per/well. After the cells were treated as described above, the supernatant was discarded and the cells were treated with MTT for 4 h. Subsequently, DMSO was added, and the absorbance was recorded at 570 nm using a microplate reader (Tecan, CH).³¹

Cell Morphology Observation

HUVECs were seeded in 6-well plates at a density of 1.8×10⁵ cells per well. After treatment as described above, morphological changes in the cells were visualized using a microscope (Nikon, Japan).

Evaluation of the Synergistic Effect

The hypotensive effects of GAS and GA were tested at concentrations of 0.8, 4, 20, and 100 µM and 0.16, 0.8, 4, and 20 µM. The combination doses were determined based on the optimal hypotensive effects of the treatment groups. Sixteen combinations of GAS and GA were used to determine the optimal synergistic dose and dose range. The combination treatment cell viability results were further analyzed using the Chou-Talalay model to determine the synergistic effect. Chou-Talalay model analysis was performed using CompuSyn software.³² The CompuSyn program was used to compute the combination index (CI) for the studied drug combinations. The resulting CI offers a quantitative definition of the additive effects (CI = 1), synergism (CI < 1), and antagonism (CI > 1) in drug combinations.³³ The optimal synergistic combination dose and single dose were used in subsequent experiments to identify possible synergistic hypotensive effect mechanisms.

NO, LDH, IL-6, IL-1β, TNF-α, ET-1, TX-B2, iNOS Measurement

HUVECs were treated as described above. After drug treatment, the supernatants were collected. The levels of NO and LDH were measured using commercial kits according to the manufacturer’s instructions. The secretion of IL-6, IL-1β, and TNF-α was detected using qPCR and ELISA kits following the manufacturer’s instructions. The expression level of ET-1, TX-B2, iNOS was measured using ELISA kits following the manufacturer’s instructions. The OD values were recorded at 450 nm, and the levels of IL-6, IL-1β, and TNF-α, Thromboxane A2 and iNOS were found to be directly proportional to the OD values. The concentrations of IL-6, IL-1β, and TNF-α, ET-1, TX-B2, iNOS were determined using a standard curve.³⁴

Annexin V-FITC/PI Staining

Apoptotic cells were analyzed using Annexin V-FITC/PI staining.³⁵ The cells were harvested and incubated with Annexin V-FITC/PI (1×) working solution for 15 min. Apoptotic cells were quantified using a FACSCalibur II flow cytometer (BD Biosciences).

qRT-PCR Analysis

The primers were synthesized by Sangon Biotech (Shanghai, China), and the sequences are shown in Supplementary Table 3. Total RNA from T98G cells was extracted and reverse-transcribed to cDNA using PrimeScript^TMRT Master Mix and TB Green^® Premix Ex Taq^TM II. Relative mRNA expression was analyzed using the 2 ^−ΔΔCt method as described previously.³⁶

Western Blot Analysis

After drug treatment and cell sampling, Western blotting was performed to detect the expression of BAX, BCL-2, S1PR1, ROCK2, and p-NF-κB. The cells were lysed to obtain protein lysates. Equal amounts of protein were separated by 10% electrophoresis and transferred onto PVDF membranes. After sealing with 5% skim milk, primary and secondary antibodies were used for the analysis.³⁷ Finally, the proteins were scanned using a ChemiDocTM XRS+ system (Bio-Rad, USA) and analyzed using the ImageJ software.

Molecular Dynamics Simulation

Molecular dynamics simulations (MD) were conducted using the GROMACS 2022 ³⁸ program with constant temperature, pressure, and periodic boundary conditions. The proteins obtained from the docking results were separated from the best small molecule ligands, the details of structure IDs which are available in Supplementary Tables 1 and 2. Small-molecule force field files were generated using the antechamber tool in Ambertools software and then converted to gromacs force field files using the Acpype software tool. The simulation system for the complex was created by combining the protein and small-molecule ligand files. The GAFF force field was used for small molecules, whereas the AMBER14SB force field and TIP3P water model were used for proteins. The LINCS algorithm was employed to constrain all hydrogen bonds, with an integration step of 2 fs. Electrostatic interactions were calculated using the Particle-mesh Ewald (PME) method, with a cutoff value of 1.2 nm. The cutoff for non-key interactions was set to 10 Å and updated every 10 steps. The temperature was maintained at 298 K using the V-rescale temperature-coupling method, whereas the pressure was maintained at 1 bar using the Berendsen method. Equilibrium simulations were conducted at 298 K for 100 ps using the NVT and NPT. MD simulations were then performed for 100 ns on the complex system, with conformations saved every 10 ps. Finally, the simulated trajectories were analyzed using VMD and PyMOL.

Statistical Analysis

Statistical analysis was performed using one-way analysis of variance (ANOVA) followed by Dunnett’s test and is presented as mean  ± SEM (GraphPad Prism 8.0). * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 were defined as statistically significant.

Results

Screening of Major Bioactive Components and Construction of the Anti-Hypertension Targets Network of TM-GQZ

To explore the therapeutic mechanism of TM-GQZ in the treatment of HTN, 3302 disease-related targets were initially identified (Figure 1A) he details of which are available in Supplementary Table 4. Furthermore, the Venn diagram indicated 133 and 202 targets for TM and GQZ, respectively, in the treatment of HTN, the details of which are available in Supplementary Table 5. The combination of TM and GQZ resulted in 39 shared targets for the treatment of HTN (Figure 1B), the details of which are available in Supplementary Table 6. 18 active compounds of TM and 29 active compounds of GQZ were selected using OB and DL filtering criteria. Comprehensive details about these active ingredients can be found in Supplementary Table 7. The drug-compound-target network was constructed using Cytoscape (Figure 1C). Through degree ranking, GAS, PHE, PA, LBP, LUT, and GA were selected as important bioactive compounds for further verification. The antihypertensive targets of TM-GQZ were imported into the STRING database to obtain a protein-protein interaction network (Figure 1D). The interaction network was analyzed using the MCODE plug-in of Cytoscape software, and the subnetwork with the highest MCODE score was obtained (Figure 1E). In the subnetwork, the top five targets with larger weights, including NF-κB1, JAK2, BAX, PPARG, and TNF, may be the key targets for TM-GQZ to exert antihypertensive effects. (Figure 1F).

Figure 1 The network construction and topological analysis of anti-hypertension targets of TM-GQZ. (A and B) Venn analysis of hypertension-related targets (A) and common targets of TM-GQZ in treating hypertension (B). (C) Compound-target network of TM-GQZ. (D) The PPI network construction of the common targets. (E) The construction of sub-network. (F) The construction of interaction network of key targets.

Biological Processes and KEGG Pathways Affected by TM-GQZ in Treating HTN

To explore the functional annotation of the targets regulated by TM-GQZ in the treatment of HTN, GO and KEGG enrichment analysis was performed with the “clusterProfiler” package (v. 4.4.4) with a p-value threshold set to 0.05. The results revealed that these targets were mainly enriched in the inflammatory response and apoptotic process (Figure 2A), indicating that inflammation and apoptosis are both important biological events that are altered during HTN. Furthermore, KEGG pathway analysis showed that multiple inflammation-related, apoptosis-related, and sphingolipid signaling pathways were highly enriched (Figure 2B).

Figure 2 The enrichment analysis and molecular docking. (A and B) Biological process analyses (A) and KEGG pathway analyses (B) of common targets of TM-GQZ in treating hypertension (p-value < 0.05). (C) The docking scores of key active components of TM-GQZ with the five hub targets. (D) Three-dimensional visual representation and two-dimensional visual representation of the molecular docking interactions between active ingredients and hub targets.

Molecular Docking Results

To further explore the possible binding modes between the active components (GAS, PHE, PA, LBP, LUT, and GA) and hub genes (PPARG, NF-κB1, TNF, JAK2, and BAX), binding affinities were predicted using AutoDock Vina. Binding stability between receptors and ligands is governed by their binding energy. A lower binding energy signifies greater stability in the receptor-ligand conformation.^39,40 Consequently, we identified the most effective ligand (having the lowest binding energy) for receptor binding, using binding energy as the primary criterion. As shown in Figure 2C, the results showed that the binding energies of hub genes with key active components. Figure 2D, Figure 7A and B illustrated the interactions between the receptors and ligands. Among them, GA bound to BAX through 6 hydrogen bonds formed by ALA-42, PRO-130, ARG-134, and LEU-47, GA bound to JAK2 through 3 hydrogen bonds formed by LYS-263, LYS-265 and GLU-343, GA bound to PPAARG through 4 hydrogen bonds formed by LYS-518, GLN-626, GA bound to TNF through 7 hydrogen bonds formed by GLY-148, LEU-93, ASN-34, and ALA-33, ARG-32, GA bound to NF-κB1 through hydrogen bond formed by ARG-156, GAS bound to BAX through 7 hydrogen bonds formed by LYS-607, ASP-789, HIS-770 and HIS-662, GAS bound to JAK2 through 3 hydrogen bonds formed by LYS-263, LYS-265, GLU-343, GAS bound to PPAARG through hydrogen bond formed by ASP-220, GAS bound to TNF through 6 hydrogen bonds formed by ASN-34, ARG-82, PRO-8, and LEU-37, GAS bound to NF-κB1 through 5 hydrogen bonds formed by ASN-138, ALA-137 and LEU-69. The above results indicated that these five components can bind well with these two proteins.

AngII Induces Cytotoxicity and Injury of HUVECs

The physiological function of endothelial cells play an important role in maintaining normal cardiovascular functions. Ang II–induced endothelial dysfunction is a pathological mechanism underlying the occurrence and development of cardiovascular disease. MTT results showed that 0.1 μM and 1 μM AngII markedly reduced cell viabilities as compared with control group, and 1 μM AngII decreased the cell viability by 39.83% (Figure 3A). Furthermore, the results revealed that 1 μM Ang II significantly decreased NO production and promoted LDH release (Figure 3B and C). Furthermore, inverted microscopy revealed that the cells in the control group were well-grown, numerous, with tight intercellular connections, polygonal or pike-shaped, uniform in size, and with clear boundaries. Compared with the control group, the number of cells in the model group after Ang II intervention was significantly reduced, the cells were less connected to each other, disorderly arranged, unevenly distributed in size, and the phenomenon of detachment was obvious. This was particularly evident in the HUVECs after 1 μM and 10 μM Ang II treatment (Figure 3D), suggesting that 1 μM Ang II induced injury in HUVECs, the concentration of which was chosen for the follow-up experiments.

Figure 3 Ang II induces cytotoxicity in HUVECs. (A) Viability of HUVECs incubated with 0, 0.1, 1, and 10 μM Ang II for 12 h (n=6). (B) NO release in HUVECs stimulated with 0, 0.1, 1, and 10 μM Ang II for 12 h (n=6). (C) LDH levels of HUVECs treated with 0, 0.1, 1, or 10 μM Ang II for 12 h (n=6). (D) Morphological changes in HUVECs treated with 0, 0.1, 1, and 10 μM Ang II for 12 h (n=5). Scale bar, 100 μm. *P < 0.01, ****P < 0.0001 vs control group. Data are expressed as the mean ± SEM. Statistical significance was determined using one-way ANOVA for multiple comparisons. All experiments were repeated at least three times.

The Protective Effects of Bioactive Components in AngII-Induced HUVECs

To investigate the effects of GAS, PHE, PA, LBP, LUT, and GA on HUVECs and Ang II-treated HUVECs, a cell viability assay was performed. The results demonstrated that GAS, PA, LBP, LUT, and GA were not cytotoxic to the HUVECs (Figure 4A–F). However, 20 and 100 μM PHE had slight toxic effects on HUVECs (Figure 4B). Moreover, MTT assay showed that GAS (0.8, 4, 20, 100 μM), PHE (100 μM), PA (100 μM), LBP (40, 200 μg/mL), LUT (10, 50 μM) and GA (0.8, 4, 20 μM) significantly increased cell viabilities of AngII-induced HUVECs (Figure 4G–L), and the effect of GAS and GA are significantly higher than other four components.

Figure 4 Protective effects of GAS, PHE, PA, LBP, LUT and GA on AngII-induced HUVECs. (A-F) Effects of GAS, PHE, PA, LBP, LUT and GA on cell viability of HUVECs (n=6). (G-L) Effects of GAS, PHE, PA, LBP, LUT and GA on cell viability of AngII-induced HUVECs (n=6). ***P < 0.001, ****P < 0.0001 vs control group. #P < 0.05, ###P < 0.001, ####P < 0.001 vs AngII group. Data were expressed as mean ± SEM. Statistical significance was determined by one-way ANOVA for multiple comparisons. Six replicates of each sample were subjected to an independent experiment, and the experiment was repeated six times.

The Synergistic Effect of GAS and GA Based on Chou-Talalay Algorithm

Thus, a combination of GAS and GA may have synergistic effects. For this purpose, Ang II-treated HUVECs were treated with the indicated concentrations of GAS and GA, alone or in combination, and cell viability was measured using an MTT assay. The result revealed that The synergistic dose range was 4–100 µM GAS and 0.8–20 µM (Figure 5A). CI values were measured using the CompuSyn software. CI < 1.0 for virtually all conditions (except for one condition for the GAS and GA combination) indicated synergism between GAS and GA (Figure 5B), and the optimal synergistic dose was 20 µM of GAS and 4 µM of GA (Figure 5C). These results revealed that the combination of GAS and GA exerted synergistic effects. Thus, the concentrations of 20 µM GAS and 4 µM GA were chosen for subsequent studies.

Figure 5 The synergistic neuroprotection effect of GAS and GA. (A) Effects of GAS and GA on cell viability of AngII-induced HUVECs (n=6). (B and C) Synergy analysis based on Chou-Talalay model (CI < 1.0, = 1.0, and > 1.0, indicating synergistic, additive, and antagonistic effect, respectively). ****P < 0.0001 vs control group. ##P < 0.01, ####P < 0.001 vs AngII group. ns, not significant. All data were shown as mean ± SEM. Statistical significance was determined by one-way ANOVA for multiple comparisons. All experiments were repeated independently for six times.

The Synergistic Effect of GAS and GA on Inhibiting Inflammation and Apoptosis in AngII Induced HUVECs

To validate the synergistic anti-inflammatory and anti-apoptotic effects of GAS and GA, the production of NO, LDH leakage, the level of ET-1, iNOS and TX-B2, inflammatory cytokine secretion, apoptotic cell rate, and BAX and BCL-2 expression levels were measured. When cells were stimulated with 1 µM Ang II, the expression level of NO was significantly reduced by 64.80%. LDH release was markedly increased by 132.30% (Figure 6B), the ATP level was notably reduced by 56.95% (Figure 6C). The expression level of ET-1, iNOS, TX-B2 were notably increased by 133.90%, 109.61%, 255.11%, respectively (Figure 6D–F), the secretion of inflammatory cytokines (IL-6, IL-1β, TNF-α) was significantly increased (Figure 6G–I), the apoptotic cell rate was notably increased (Figure 6M–N), the expression of pro-apoptotic marker BAX was upregulated by 103.54%, and the protein level of anti-apoptotic marker BCL-2 was downregulated by 66.30% (Figure 6O–Q).

Figure 6 The effects of GAS and GA combination on inflammation and apoptosis. (A) The NO production (n=6). (B) The LDH leakage (n=4). (C) The ATP level (n=6). (D) The ET-1 level (n=5). (E) The TX-B2 level (n=5). (F) The iNOS level (n=5). (G–I) The mRNA expression of IL-6, IL-1β and TNF-α detected by qPCR (n=3). (J–L) The secretion of IL-6, IL-1β and TNF-α detected by ELISA assay (n=4). (M and N) The effects of GAS and GA combination on apoptotic cells rate (n=3). (O–Q) The expression of BAX and BCL-2 (n=3). **P < 0.01, ***P < 0.001, ****P < 0.0001 vs control group. #P < 0.05, ##P < 0.01, ###P < 0.001, ####P < 0.0001 vs AngII group. All data were shown as mean ± SEM. Statistical significance was determined by one-way ANOVA for multiple comparisons. All experiments were repeated independently at least three times.

However, treatment with GAS (20 μM) and GA (4 μM) markedly increased NO levels by 90.21% and reduced LDH leakage by 53.33%. Meanwhile, treatment with GAS and GA elevated the ATP levels by 106.28%. Furthermore, combination of GAS and GA significantly decreased the expression level of ET-1, iNOS, TX-B2 by 44.12%, 38.68%, and 49.61%, respectively. qPCR and ELISA results showed that the combination of GAS (20 μM) and GA (4 μM) significantly decreased the levels of inflammatory cytokines (IL-6, IL-1β, and TNF-α). Moreover, the combination of GAS and GA markedly decreased the rate of apoptosis, downregulated BAX expression by 46.55%, and up-regulate the expression of BCL-2 by 186.33%. These results demonstrated that the combination of GAS and GA significantly alleviated Ang II–induced inflammation and apoptosis in HUVECs, the details of which are available in Supplementary Table 8.

The Anti-Inflammatory and Anti-Apoptotic Effects of GAS and GA Combination Regulated by Sphingolipid Metabolism

To explore the potential mechanisms of the GAS and GA combination on sphingolipid suppression, the binding of GAS or GA with NF-κB1 was predicted by molecular dynamics simulations, and the protein levels of S1PR1, ROCK2, and p-NF-κB were detected by Western blotting in Ang II-treated HUVECs. Molecular docking results showed that amino acids ASN-138 and LYS-79 in NF-κB1 form hydrogen bonds with GAS, PRO-70, and LYS-79 from Pi-Alkyl and Pi-Cation hydrophobic interactions with GAS, and amino acids LYS-76 and TYR- 81 form van der Waals force interactions with GAS. Amino acids ILE-122 and ALA-160 in NF-κB1 formed Pi-alkyl hydrophobic interactions with GA, ARG-156 formed a salt bridge with GA, and amino acids VAL-97 and TYR-165 formed van der Waals interactions with GA (Figure 7A-B). The VDW for GAS, GA, and NF-κB1 binding was more stable relative to the ELE, and the binding between small molecules and proteins remained stable as the simulation progressed for GAS, GA, and NF-κB1 binding (Figure 7C and D). The number of hydrogen bonds between GAS and NF-κB1 was more stable, fluctuating mainly between one and five. The number of hydrogen bonds between GA and NF-κB1 was mainly below two after 30 ns, and the number of hydrogen bonds between small molecules and proteins was lower (Figure 7E and F). The RMSD of NF-κB1, the RMSD of the GA and NF-κB1 binding complex, and the RMSD of GAS and NF-κB1 binding remained stable after 50 ns. (Figure 7G and H). The Rg values of the complexes bound by GAS, GA, and NF-κB1 stabilized progressively during the simulation (Figure 7I and J). The distance between GAS and the initial binding site appeared to change before 50 ns (Figure 7K); however, analysis of its conformation (Figure 7M) revealed that the GAS small molecule moved near the initial binding site and stably bound to the initial binding site of NF-κB1 after 50 ns. The distance between GA and the initial binding site, the docking site, changed in a stepwise manner (Figure 7L), and the trajectory of the small molecule was analysed (Figure 7N), which showed that GA moved on the surface of NF-κB1 from the initial binding site at 0 ns to the site at 50 ns, and then to the 100 ns site, which were all close to the initial binding site, and the distance between GA and the centre of the NF-κB1 protein did not change by more than 2 nm.The Buried SASA for GAS binding to NF-κB1 remained stable after 50 ns (Figure 7O). The Buried SASA for GA binding to NF-κB1 fluctuated in accordance with the distance between GA and NF-κB1, with the highest Buried SASA at the initial binding site (Figure 7P). The superimposed conformations of GAS, GA, and NF-κB1-bound GAS were located near the initial binding site (Figure 7Q and R). The major conformations of GA and NF-κB1-bound GA were distributed in three positions, corresponding to the three sites previously analyzed for the binding of GA small molecules to NF-κB1. The above results indicated that when GAS and NF-κB1 are combined, van der Waals interactions play a major role, electrostatic interactions play a minor role, and hydrophobic interactions play a complementary role.When both GA and NF-κB1 are combined, electrostatic interactions play a major role, van der Waals interactions play a minor role, and hydrophobic interactions play a complementary role. GAS or GA with NF-κB1 combined with Good stability, high binding energy and affinity.

Figure 7 Sphingolipid metabolism is involved in the anti-inflammatory and anti-apoptotic effects of GAS and GA combination. (A and B) The simulated binding conformations of GAS, GA binding with NF-κB separately. (C and D) The van der Waals and electrostatic interactions between GAS, GA and NF-KB. (E and F) The Hbond number of GAS and GA to NF-κB. (G–J) The RMSD, Rg results of GAS, GA binding with NF-κB separately. (K and L) The Dock Site-ligand for the binding site of GAS, GA binding with NF-κB separately. (M and N) The binding trajectories of GAS, GA binding with NF-κB separately. (O and P) The Buried SASA of GAS, GA binding with NF-κB separately. (Q and R) The Simulated conformational superposition of GAS and GA to NF-κB. (S–V) The protein level of S1PR1, ROCK2 and p-NF-κB (n=3). **P < 0.01, ****P < 0.0001 vs control group. ####P < 0.001, ####P < 0.001 vs AngII group. ns, not significant. All data were shown as mean ± SEM. Statistical significance was determined by one-way ANOVA for multiple comparisons. All experiments were repeated at least three times.

Western blotting showed that 1μM AngII II decreased S1PR1 expression by 29.83% and notably increased the protein levels of ROCK2 and p-NF-κB by 409.87% and 402.31%, respectively. However, treatment with GAS (20 μM) and GA (4 μM) markedly elevated S1PR1 levels by 63.91% (Figure 7S and T), and suppressed the expression of ROCK2 and p-NF-κB by 77.08% and 71.53% (Figure 7S, U and V). The above results revealed that the combination of GAS and GA could regulate the protein levels of S1PR1, ROCK2, and p-NF-κB in Ang II-treated HUVECs, which are the key targets participating in sphingolipid metabolism.

Discussion

HTN continues to pose a major hazard to human health, and has emerged as a high-incidence disease. Current treatment for HTN includes medications such as diuretics and ACE inhibitors, as well as lifestyle changes like diet, exercise, and weight management. However, there is no cure for HTN at present.⁴¹ Numerous studies have demonstrated the effectiveness of Chinese medicine may be in treating HTN.⁴² Chinese herbal formulas are frequently used in the clinical practice. Compounding enables drug interactions, which can result in a series of biochemical reactions that enhance efficacy and decrease toxicity.^43,44 The compatibility concept of traditional Chinese medicine is based on experience, but its potential interactions and mechanisms are yet to be elucidated.

Studies have demonstrated the beneficial antihypertensive effects of TM and GQZ; however, the precise active chemical components and associated mechanisms remain unknown.^27,28,45,46 Specifically, TM and GQZ contain substances, such as polyphenols and polysaccharides, which are the main biologically active compounds with antihypertensive properties.^28,47,48 The polyphenolic components in the GAS and gallic GA formulations have been the most widely studied for their antihypertensive activity.^26,27,49,50 Therefore, GAS and GA are the primary active components of TM and GQZ, respectively, both of which have antihypertensive effects. Nonetheless, the antihypertensive mechanisms of TM and GQZ remain unknown; thus, network pharmacology could help identify key targets. The topology-based network analysis revealed that GAS, PHE, PA, LBP, LUT, and GA were the main active components. Moreover, the top five targets with the greatest weight, including NF-κB1, JAK2, BAX, PPARG, and TNF, were considered potential targets for the compatibility of TM and GQZ.

Recent studies have shown that GAS reduces inflammation by inhibiting the NF-κB pathway and enhancing antioxidant defenses, while GA modulates inflammation by suppressing pro-inflammatory cytokines and MAPK pathways.⁵¹ Both substances lower blood pressure by improving endothelial function and inhibiting angiotensin-converting enzyme activity.^52–54 Additionally, GAS and GA counteract atherosclerosis by improving lipid profiles, reducing oxidative stress, and enhancing endothelial health.^55,56 These findings align with recent literature and highlight the synergistic potential of these substances in managing inflammation, HTN, and atherosclerosis. The results of our current study revealed that GAS and GA synergistically protected HUVECs from Ang II–induced injury. Ang II plays a critical role in the inflammatory cascade by inducing vasoconstriction, promoting atherosclerosis, and contributing to cardiovascular events.⁵⁷ It elevates blood pressure through direct vasoconstriction and stimulating aldosterone release, leading to endothelial damage.⁵⁸ This damage exacerbates inflammation, enhances oxidative stress, and accelerates plaque formation in arteries. Chronic Ang II exposure further impairs endothelial function and increases the risk of cardiovascular events such as heart attacks and strokes, ultimately contributing to increased mortality.⁵⁹ Consistent with previous studies, Ang II initiated HUVECs damage, inhibited NO release, increased ET-1, iNOS and TX-B2 level, and augmented LDH release, further aggravating the course of HTN.^60–62 Furthermore, our cell viability experiment indicated considerable synergy between GAS and GA. Numerous studies have shown that GAS and GA improve the specific mode of action of endothelial cell injury, further supporting our findings.^21,63–66 In our study, synergistic effects are defined as the enhanced efficacy of combining GAS and GA compared to their individual use in attenuating AngII-induced injury. We evaluated this by comparing the outcomes in the combination treatment group with those in the individual treatment groups and the control group. For example, our result show that the combination treatment resulted in a significantly greater increase in cell viability compared to either Gastrodin or Gallic Acid alone.

Pathway enrichment analysis indicated that GAS and GA differentially affected the inflammatory response- and apoptosis-related pathways. The role of the inflammatory response in HTN is significant because it regulates blood pressure by affecting vascular function and endothelial cell activity.⁶⁷ Inflammatory factors act on vascular endothelial and smooth muscle cells, leading to vascular wall inflammation. This, in turn, affects the regulation of vascular tone and blood pressure. For example, TNF-α stimulates the release of NO by vascular endothelial cells, resulting in vasodilation. In contrast, IL-6 increases the sensitivity of vascular endothelial cells to vasoconstrictor factors, leading to vasoconstriction and affecting both the vasodilatory and constrictor functions. Furthermore, apoptosis plays a crucial role in HTN by regulating the structure and function of the vascular wall through its effects on the number and function of vascular endothelial and smooth muscle cells, and by influencing the regulation and maintenance of blood pressure.⁶⁸ Abnormal regulation of apoptosis in hypertensive states results in an increased rate of apoptosis in the vascular endothelial and smooth muscle cells. Endothelial cells are a significant component of the vascular lining and their abnormal function is closely linked to vascular wall damage and the development of HTN. Emerging studies suggest that inflammation and apoptosis play crucial roles in the pathogenesis of metabolic dysregulation. Inflammation and apoptosis are intricately linked processes that contribute to metabolic dysregulation through various mechanisms including insulin resistance, adipose tissue dysfunction, and lipid accumulation-induced toxicity.^28,69 Therefore, targeting inflammation and apoptosis pathways holds promise for the development of novel therapeutic strategies to combat metabolic disorders. The current study confirmed that treatment with a combination of GAS (20 μM) and GA (4 μM) markedly increased NO levels, reduced LDH leakage, suppressed the secretion of inflammatory cytokines, decreased the apoptotic cell rate, and alleviated apoptosis-related marker protein levels. Therefore, our results demonstrated that GAS and GA could alleviate Ang II–induced HUVECs injury by inhibiting inflammation and apoptosis. While our study is primarily preclinical, the results suggest that the combination of Gastrodin and Gallic Acid may offer superior therapeutic benefits over their individual use in clinical settings, particularly in the management of chronic inflammation and apoptosis-related diseases. These findings provide a theoretical basis for considering this combination therapy in clinical trials.

Disorders in sphingolipid metabolism affect the occurrence and development of HTN.¹⁵ The development and homeostasis of endothelial cells depend heavily on the proper functioning of sphingolipid metabolism, which is intimately linked to endothelial cell function. Sphingosine, ceramide (Cer), sphingosine 1-phosphate (S1P), serine, sphingosine kinase (SPHK), ceramide kinase (Cerk), and sphingosine 1-phosphate lyase (S1PL) are key sphingolipid metabolites and important enzymes in pathological situations.¹⁴ Numerous inflammatory factors are generated when the downstream inflammation-related signaling pathway is activated, ultimately leading to the destruction of endothelial cell function.⁷⁰ In our study, Ang II induced perturbation of sphingolipid metabolism by suppressing the level of S1PR1 and increasing the downstream signaling molecules of ROCK2 and p-NF-κB. In particular, the combination of GAS (20 μM) and GA (4 μM) reversed this phenomenon by elevating the expression of S1PR1 and reducing ROCK2 and p-NF-κB levels. Our study suggests that sphingolipid metabolism is involved in the inhibition of inflammation and apoptosis by GAS and GA in Ang II–induced HUVECs.

Conclusion

In summary, our findings reveals that the combination of GAS and GA significantly mitigates AngII-induced injury in HUVECs by effectively suppressing both inflammation and apoptosis. These effects are closely linked to the modulation of sphingolipid metabolism. Additionally, our findings suggest that targeting sphingolipid metabolism not only provides insights into the progression and severity of HTN but also highlights its potential as a novel therapeutic target for future interventions. This study underscores the promising therapeutic potential of GAS and GA in managing endothelial dysfunction and HTN.