Physiological fluid shear stress synergistically enhances the protective effects of Salvia miltiorrhiza extract on endothelial cells

Abstract

Natural medicines are gaining increasing attention in drug development, yet traditional static cell models fail to mimic the dynamic mechanical microenvironment in vivo, potentially leading to misleading evaluations. This study investigated how physiological fluid shear stress influences the protective effects of Salvia miltiorrhiza extract (SME) on endothelial cells (ECs). Using a parallel plate flow chamber, ECs were exposed to 15 dyn/cm² shear stress or static conditions, followed by H₂O₂-induced oxidative damage. Cell viability was assessed via MTT and live/dead staining; morphology and cytoskeleton were examined through phalloidin/DAPI staining; oxidative markers (ROS, SOD, MDA) were quantified; and SME bioactive component uptake was analyzed using HPLC. Under shear stress, SME’s protection was significantly enhanced: metabolic viability reached 86.83 ± 6.4% (vs. 72.38 ± 6.8% static), protection rate nearly doubled (68.00% vs. 34.14%), morphology improved (shape index: 0.63 ± 0.11 vs. 0.73 ± 0.11), and oxidative damage was reduced (ROS decreased to 13.73 ± 1.47; SOD increased to 41.24 U/mg; MDA decreased to 0.74 nmol/mg). HPLC revealed enhanced absorption of all compounds and uptake of two additional bioactive constituents. These findings demonstrate that physiological shear stress potentiates SME’s efficacy through improved compound delivery and amplified cellular responses, advocating for shear stress-responsive screening as a physiologically relevant strategy for natural medicine evaluation.

Introduction

Natural medicines have long been recognized as an important resource for the development of new drugs, with a rich history that dates back to ancient civilizations. The utilization of herbal remedies has been documented throughout history, and modern scientific research has delved into the active components of these natural medicines. In recent years, there has been a renewed interest in natural medicines as the development of synthetic drugs has faced challenges. This renewed focus is due to the unique chemical diversity and biological activity of natural medicines, making them an attractive area of research. The complex composition of natural medicines requires a thorough approach to screening their active components and conducting pharmacological research. This process typically involves a multi-level approach, encompassing techniques such as network pharmacology, virtual screening, in vitro molecular experiments, cell studies, tissue studies, organ studies, and animal studies. By utilizing these various methods, researchers can gain a comprehensive understanding of the potential benefits and mechanisms of action of natural medicines, further solidifying their place in the medical domain^1–3. Network pharmacology, through the utilization of database resources, builds a comprehensive “drug-target-disease” network to forecast the potential targets of the bioactive compounds present in natural medicines, as well as elucidate the interplay between these targets and the pathways implicated in disease treatment⁴. Virtual screening utilizes computer simulation technology to efficiently screen natural products that may possess affinity for particular targets from vast compound databases⁵. However, the findings obtained from both network pharmacology and virtual screening methods must be validated through rigorous in vitro and in vivo experiments to establish the credibility of the research. Investigating the effects of active compounds in natural medicines and understanding their pharmacological mechanisms is crucial at the cellular level. Cells, as the fundamental functional units of living organisms, play a crucial role in bridging virtual experiments and in vitro molecular experiments with in vivo research in the study and screening of active substances in natural medicines. Endothelial cells (ECs), as a monolayer of flat cells lining the inner surface of blood vessels, are crucial regulators of vascular homeostasis and play vital roles in various physiological and pathological processes^6,7. The targeting of ECs for screening active substances in natural medicines and pharmacological research is of immense theoretical importance and clinical relevance. The predominant approaches utilized for investigating natural medicines through ECs involve both two-dimensional monolayer cultured cell models and three-dimensional cell models. Two-dimensional cell models serve as a useful tool for the initial assessment of the effects of bioactive compounds present in natural medicines by examining changes in various biological behaviors of ECs, including morphology, proliferation, migration, and apoptosis⁸. On the other hand, three-dimensional cell models are instrumental in creating a cellular microenvironment that closely mimics in vivo conditions, allowing for a more accurate representation of cell-cell interactions, cell-matrix adhesion, and substance exchange, thus better reflecting physiological responses⁹. The effects of bioactive compounds present in natural medicines on the angiogenic process can be more precisely and meaningfully elucidated. This holds paramount significance for research on anti-angiogenic properties of natural medicines and aids in identifying bioactive compounds that possess the ability to efficiently inhibit tumor angiogenesis in living organisms¹⁰. However, the identification of active compounds in natural medicines and pharmacological studies at the cellular level predominantly utilize static models. These models are limited in their ability to accurately replicate the dynamic microenvironment in which cells exist in vivo, potentially leading to discrepancies between the observed pharmacological effects of natural medicines in research settings and their real-world applications. As a result, there is a risk of suboptimal efficacy or unexpected side effects when these identified compounds are administered in vivo.

In the human body, cells are located in a complex and diverse microenvironment. One of the crucial mechanical factors in this microenvironment is fluid shear stress, which plays a significant role in shaping the behavior and functions of cells. ECs form a monolayer on the inner wall of blood vessels, placing them in direct contact with blood flow¹¹. These cells are subject to constant shear stress due to the frictional force generated by the flowing blood¹². Shear stress is a critical factor in preserving the equilibrium of the vascular system, regulating diverse physiological functions, and initiating pathological changes. In the absence of any abnormalities, the stimulation of shear stress can direct ECs to orient themselves in a coordinated manner, promoting the stability of vascular structure and facilitating uninterrupted blood flow¹³. Recent studies have demonstrated that physiological levels of laminar shear stress have the ability to activate the Akt signaling pathway in ECs. This activation promotes cellular survival and proliferation while also contributing to the maintenance of the vascular endothelium integrity¹⁴. Additionally, research suggests that consistent shear stress can stimulate ECs to produce a substantial quantity of nitric oxide, leading to vasodilation. This process effectively reduces the tension of vascular smooth muscle and aids in stabilizing blood pressure¹⁵. Furthermore, the stimulation from shear stress prompts ECs to release antithrombotic factors, such as tissue plasminogen activator (t-PA), which contributes to the prevention of blood clot formation¹⁶. Moreover, shear stress acts as a crucial regulator in the transendothelial transport of ECs, significantly impacting the exchange of substances between the bloodstream and tissues. Shear stress plays a role in modifying the activity and distribution of transport proteins within ECs, thereby governing substance transport¹⁷. For instance, at normal physiological shear stress levels ranging from 10 to 20 dyn/cm², the expression levels of glucose transporters 1 and 4, responsible for facilitating glucose transport, remain stable and evenly distributed on the EC surface. This ensures that glucose is transported into tissues from the bloodstream at a suitable rate to meet cellular metabolic demands¹⁸. Additionally, shear stress influences the integrity of intercellular junctions in ECs. Physiological shear stress levels promote the upregulation of tight junction proteins like occludin and claudin between ECs, tightening the intercellular spaces and limiting the passage of large molecules in and out of blood vessels¹⁹. Considering the impact of fluid shear stress on EC physiological functions and substance exchange processes, drug absorption by ECs is also influenced by shear stress. For instance, elevated shear stress levels lead to a significant increase in the gene transcription level and protein expression of P-glycoprotein, a crucial drug efflux transporter, enhancing its transport activity and resulting in increased drug efflux from cells and decreased drug uptake efficiency²⁰. Conversely, appropriate shear stress stimulation leads to a moderate increase in the expression and activity of organic cation transporters, a group of transport proteins that promote drug uptake. This enhances the absorption of corresponding drugs by cells and optimizes the intracellular drug concentration, ultimately improving drug efficacy²¹. Research has demonstrated the significant impact of shear stress on cellular uptake of various drugs. In the realm of cardiovascular medications, it has been found that physiological shear stress can facilitate the absorption of antiplatelet drug aspirin by ECs, consequently enhancing its ability to inhibit platelet aggregation. Conversely, abnormal shear stress resulting from conditions like hypertension has been shown to diminish drug uptake by cells, thereby diminishing therapeutic efficacy²². Additionally, in the context of antitumor medications such as paclitaxel, consistent laminar shear stress has been observed to upregulate specific transport proteins, leading to increased uptake of paclitaxel by ECs within tumor blood vessels²³. This serves to elevate drug concentration at tumor sites, ultimately enhancing the effectiveness of chemotherapy. On the other hand, irregular shear stress has been linked to the development of drug resistance in cells, impeding drug absorption. While existing research has established a foundation regarding the impact of shear stress on drug effects on ECs, the majority of studies have focused on individual chemical drugs, with limited exploration of complex natural medicines.

This research utilized Salvia miltiorrhiza extract (SME) as a model drug to investigate the impact of fluid shear stress on EC repair using natural medicines. SM has effectiveness in treating cardiovascular diseases, supported by modern pharmacological research confirming its vascular protective properties²⁴. The bioactive components of SM, such as salvianolic acids and tanshinones, have been proven to have mechanisms such as antioxidant activity (e.g., through activation of the Nrf2/HO-1 pathway) and anti-inflammatory effects (e.g., by inhibiting NF-κB), which are crucial for EC function and the management of vascular disorders²⁵. Furthermore, the diverse composition of SM, which encompasses both hydrophilic and lipophilic compounds, makes it an excellent candidate for investigating how dynamic mechanical forces influence the pharmacological actions of intricate natural medicines, thereby offering a more realistic representation of their clinical utility.

A parallel plate flow chamber was employed to mimic the physiological shear stress environment experienced by ECs, which were then cultured and exposed to oxidative damage. The morphological changes in ECs, cell viability, and levels of oxidative stress markers were assessed to evaluate the reparative effects of SME under shear stress. Additionally, the uptake of active compounds from SME by ECs was compared between static and dynamic cell models. The findings of the study indicated that under appropriate shear stress conditions, SME significantly enhanced the repair and regeneration of ECs, resulting in improved cell morphology, increased cell viability, and reduced oxidative stress. Furthermore, the absorption of bioactive compounds was significantly enhanced when subjected to shear stress. Overall, these results underscore the influence of shear stress on the efficacy of natural medicines in protecting vascular ECs and facilitating the uptake of active components, offering novel insights and a foundation for the development of new natural medicines.

Results

Estimation of shear stress

The in vitro cell culture flow device enables precise control over the magnitude of shear stress experienced by cells through regulation of the flow rate of the culture medium. In the context of a parallel plate flow chamber, the relationship between shear stress (τ), flow rate (Q), and the geometric dimensions (L, W and H) of the flow chamber can be described by the formula: τ = 6ηQ/WH², where τ represents the shear stress at the bottom surface of the flow chamber in dyn/cm², η denotes the viscosity of the perfusion fluid in dyn·s/cm², Q is the flow rate of the perfusion fluid in mL/s, W is the width of the flow chamber in cm, and H is the height of the flow chamber in cm. The viscosity of the culture medium at 37 °C was determined to be 6.684 × 10⁻³ dyn·s/cm² using an Ubbelohde viscometer, while the width (W) and height (H) of the flow chamber were measured to be 2.5 cm and 0.2 mm, respectively, utilizing a vernier caliper. Additionally, the relationship between the rotational speed of the peristaltic pump and the flow rate of the flow device was established through experimentation (Table 1), and the connection between shear stress and the rotational speed of the peristaltic pump was elucidated (Fig. 1). Notably, the regression curve displayed a remarkably high linear correlation (R² = 0.999). Based on the linear regression equation derived from the data, it can be predicted that at a rotational speed of 1.06 r/s, the shear stress (τ) would approximate 15 dyn/cm², aligning closely with the typical physiological shear stress experienced by blood in vivo.

Table 1.

The relationship between rotary speed of peristaltic pump and flow rate.

Rotary speed (rpm)	t (s)/100mL			Q (mL·s− 1)			(mL·s− 1)	RSD (%)
	t1	t2	t3	Q1	Q2	Q3
0.2	2204.6	2204.7	2204.8	0.0454	0.0454	0.0454	0.0454	0
0.3	1198.0	1198.1	1198.0	0.0835	0.0835	0.0835	0.0835	0
0.4	829.9	829.9	829.9	0.1205	0.1205	0.1205	0.1205	0
0.5	632.6	632.6	632.6	0.1581	0.1581	0.1581	0.1581	0
0.6	511.1	511.2	511.2	0.1957	0.1956	0.1956	0.1956	0.01
0.7	429.0	429.0	429.0	0.2331	0.2331	0.2331	0.2331	0
0.8	369.3	369.3	369.2	0.2708	0.2708	0.2709	0.2708	0.02
0.9	327.3	327.2	327.4	0.3055	0.3056	0.3054	0.3055	0.03
1.0	290.7	290.8	290.6	0.3440	0.3439	0.3441	0.3440	0.03
1.1	259.9	259.9	259.9	0.3848	0.3848	0.3848	0.3848	0
1.2	234.3	234.3	234.4	0.4268	0.4268	0.4266	0.4267	0.02
1.3	215.0	215.1	215.0	0.4651	0.4649	0.4651	0.4650	0.03
1.4	200.0	200.0	200.0	0.5000	0.5000	0.5000	0.5000	0

Fig. 1.

The relationship between rotary speed of peristaltic pump and shear stress.

The effective concentrations of H₂O₂ and SME

Various concentrations of H₂O₂ were found to impair EC metabolic viability (Fig. 2A). Within the range of 100 µmol/L to 1600 µmol/L, the reduction in EC metabolic viability increased with the concentration of H₂O₂. Significant decreases were observed at H₂O₂ concentration of 100 µmol/L and above (P < 0.01 for all), indicating substantial cellular damage. At a concentration of 200 µmol/L, the cell metabolic viability was reduced by 43.65%±3.4% relative to the normal control group, corresponding to a residual metabolic activity of 56.35%±3.4%, a level close to the 50% threshold recommended in literature for optimal assessment of drug protective effects²⁶. This concentration aligns with pathophysiological H₂O₂ levels (100–500 µmol/L) observed during early oxidative stress in vivo, ensuring the injury model’s biological relevance²⁷. Hence, the concentration of 200 µmol/L H₂O₂ was chosen for inducing oxidative damage in ECs.

Fig. 2.

The effect of H₂O₂ and SME on the metabolic viability of ECs. (A) Cell metabolic viability of ECs treated with different concentrations of H₂O₂ (100–1600 µmol/L). (B) Effect of different concentrations of SME (10–80 µg/mL) on the metabolic viability of ECs injured by 200 µmol/L H₂O₂. Data are presented as mean ± SD (n = 5 independent experiments). Statistical significance was determined using Welch’s ANOVA with Games-Howell post hoc test; compared to the normal control group in (A): ^**P < 0.01; compared to the H₂O₂ damage group in (B): ^*P < 0.05, ^**P < 0.01.

Different concentrations of SME displayed protective effects on oxidatively damaged ECs (Fig. 2B). However, the protective effect diminished with increasing concentrations of SME. Concentrations ranging from 10 µg/mL to 20 µg/mL SME demonstrated significant protection of ECs, with notable differences compared to the H₂O₂ damage group (P < 0.05 for all). Among these, 10 µg/mL exhibited the most potent protective effect, increasing cell metabolic viability by 28.01%±2.16% compared to the H₂O₂ damage group. Therefore, a concentration of 10 µg/mL was selected as the optimal dose of SME for protecting ECs.

Protective effects of SME on oxidative damage of ECs under fluid shear stress

Study on the viability of ECs

Under both static conditions and shear stress, SME demonstrates a protective effect on the metabolic viability of ECs. However, this protective ability is significantly enhanced under shear stress conditions (Fig. 3). Under static conditions, SME treatment restored cell metabolic viability to 72.38 ± 6.8% (compared to the H₂O₂ damage group, 48.75 ± 6.7%, P = 0.023). In contrast, under shear stress, SME enhanced cell metabolic viability to 86.83 ± 6.4% (compared to H₂O₂ damage group, 52.67 ± 6.1%, P = 0.004), yielding a significantly higher protection efficacy compared to the static condition (two-way ANOVA, P = 0.049 for interaction). This represents a ~ 1.3-fold increase in the protective efficacy of SME under dynamic flow.

Fig. 3.

Cell metabolic viability of ECs under static and dynamic conditions. Data are presented as mean ± SD (n = 5 independent experiments). Statistical significance was determined using two-way ANOVA with an interaction model, followed by LSD post hoc test and simple-effect analysis. In the same condition, compared to the H₂O₂ damage group: ^*P < 0.05, ^**P < 0.01; for comparisons within the different condition: ^#P < 0.05, ^##P < 0.01.

This enhanced protective role of SME under shear stress was further corroborated by a live/dead cell staining assay (Fig. 4), which directly assesses cell membrane integrity and viability. Representative fluorescent images (Fig. 4A) revealed a marked reduction in propidium iodide (PI)-positive (dead) cells and a concomitant increase in calcein-AM-positive (live) cells with healthy, spread morphologies in the SME-treated groups, particularly under shear stress. Quantification of live cell numbers (Fig. 4B) demonstrated that SME significantly increased viability against H₂O₂-induced damage under both static and dynamic conditions (P < 0.01 for all). Crucially, the protective effect of SME was markedly potentiated by shear stress, as evidenced by a significantly greater number of live cells in the dynamic SME protection group (120 ± 11) compared to its static counterpart (99 ± 6) (P < 0.001), a phenomenon not observed in the normal control or H₂O₂ damage groups. This significant interaction between flow and treatment was confirmed by two-way ANOVA (P < 0.001). Consequently, the protection rate of SME was calculated to be nearly twofold higher under shear stress (68.00%±7.48%) than under static conditions (34.14%±6.33%) (P < 0.001, Fig. 4C).

Fig. 4.

SME protects ECs from H₂O₂-induced oxidative damage as demonstrated by live/dead cell staining. (A) Representative fluorescent images of ECs stained with calcein-AM (green, live cells) and PI (red, dead cells) under static and dynamic conditions (scale bar = 100 μm). (a) Static normal control; (b) Static H₂O₂ damage group; (c) Static SME protection group; (d) Dynamic control group; (e) Dynamic H₂O₂ damage group; (f) Dynamic SME protection group. Arrows highlight representative morphological features: blue arrows, healthy cells; yellow arrows, viable cells with compromised morphology; red arrows, dead cells. (B) Quantitative analysis of live cell numbers. Data are presented as mean ± SD (n = 5 independent experiments). Statistical significance was determined using two-way ANOVA with an interaction model, followed by LSD post hoc test and simple-effect analysis. In the same condition, compared to the H₂O₂ damage group: ^**P < 0.01; for comparisons between different conditions: ^#P < 0.05, ^##P < 0.01. (C) Protection rate of SME. Data are presented as mean ± SD (n = 5 independent experiments). Statistical significance was determined using independent samples t-test: ^**P < 0.01.

Together, these data from two distinct methodologies consistently demonstrate that physiological shear stress dramatically augments the ability of SME to protect endothelial cells from oxidative damage.

Study on cell morphology

Under both static and dynamic conditions, normal ECs exhibited a characteristic cobblestone morphology, appearing as well-spread polygons (e.g., quadrilaterals or hexagons) with serrated edges and tight intercellular connections (Fig. 5A a/d). H₂O₂-induced oxidative damage severely disrupted this architecture, resulting in irregular cell shapes, including rounded and mulberry-like forms, increased intercellular gaps, and cell detachment (Fig. 5A b/e, red arrows and yellow circles). SME treatment markedly attenuated these morphological alterations. Under static conditions, SME improved cell density and connections, though many cells remained rounded (Fig. 5A c). Notably, under shear stress, SME’s protective effect was significantly enhanced, with most cells recovering a typical polygonal morphology and re-establishing tight junctions, closely resembling the normal control group (Fig. 5A f).

Fig. 5.

Protective effects of SME on EC morphology under static and dynamic conditions. (A) Phase-contrast micrographs of ECs (scale bar = 200 μm). (a) Static normal control; (b) Static H₂O₂ damage group; (c) Static SME protection group; (d) Dynamic control group; (e) Dynamic H₂O₂ damage group; (f) Dynamic SME protection group. Arrows highlight representative morphological features: blue arrows, healthy cells; yellow arrows, viable cells with compromised morphology; red arrows, dead cells. Yellow dashed circles highlight areas of extensive intercellular gaps. (B) Quantification of CSI under static and dynamic conditions. Data are presented as mean ± SD (n = 5 independent experiments, with 100 cells analyzed per group per experiment). Statistical significance was determined using two-way ANOVA with an interaction model, followed by LSD post hoc test and simple-effect analysis. In the same condition, compared to the H₂O₂ damage group: ^*P < 0.05, ^**P < 0.01; for comparisons within the different condition: ^#P < 0.05.

Under the static condition, the CSI values were 0.62 ± 0.10 for the normal control group, 0.86 ± 0.12 for the H₂O₂ damage group, and 0.73 ± 0.11 for the SME protection group. Under dynamic condition, the CSI values were 0.64 ± 0.11 for the normal control group, 0.85 ± 0.12 for the H₂O₂ damage group, and 0.63 ± 0.11 for the SME protection group (Fig. 5B). The analysis of CSI revealed a 40.23%±5.41% and 33.48%±3.42% increase in CSI of ECs under oxidative damage in static and dynamic conditions, respectively. In contrast, treatment with SME resulted in a significant decrease in CSI compared to the H₂O₂ damage group in both static (P = 0.025) and dynamic conditions (P < 0.001). The protective effect of SME on EC morphology was significantly enhanced under shear stress (two-way ANOVA, P < 0.001 for interaction), corroborating the findings of cell morphology observations, indicating a more pronounced morphological recovery under shear stress.

To gain deeper insight into the protective effects of SME at the subcellular level, we performed phalloidin/DAPI staining to assess cytoskeletal integrity and nuclear morphology (Fig. 6A-D). Representative images revealed that normal ECs under both conditions exhibited well-organized stress fibers (Fig. 6A a/d). In contrast, H₂O₂ exposure resulted in severe cytoskeletal disintegration, characterized by a nearly complete loss of actin filaments and cell contraction (Fig. 6A b/e). SME treatment attenuated this damage, with the more remarkable preservation of the actin cytoskeleton observed under shear stress (Fig. 6A c/f).

Fig. 6.

SME attenuates oxidative stress-induced cytoskeletal disruption and nuclear morphological changes in ECs. (A) Representative fluorescent images of ECs stained for F-actin (phalloidin, red) and nuclei (DAPI, blue) (scale bar = 50 μm). (a) Static normal control; (b) Static H₂O₂ damage group; (c) Static SME protection group; (d) Dynamic control group; (e) Dynamic H₂O₂ damage group; (f) Dynamic SME protection group. Arrows highlight representative morphological features: blue arrows, well-organized stress fibers; red arrows, fragmented actin or cells with only cortical actin. Yellow dashed circles highlight areas of improved cytoskeletal recovery. (B) Quantitative analysis of F-actin content IOD and cell area. Data are presented as mean ± SD (n = 5 independent experiments). (C) Quantitative analysis of nuclear circularity. Data are presented as mean ± SD (n = 5 independent experiments). Statistical significance was determined using two-way ANOVA with an interaction model, followed by LSD post hoc test and simple-effect analysis. In the same condition, compared to the H₂O₂ damage group: ^**P < 0.01; for comparisons within the different condition: ^##P < 0.01. (D) Frequency distribution of actin filament lengths from skeletonized images. Note the leftward shift (toward shorter filaments) and narrowing of the curve in damage groups, which is ameliorated by SME treatment, particularly under shear stress.

This visual assessment was confirmed quantitatively. Analysis of the integrated optical density (IOD) of phalloidin staining, which reflects F-actin content, showed that SME significantly restored the loss of F-actin induced by H₂O₂ under both static and dynamic conditions (P < 0.01, Fig. 6B). Crucially, the restorative effect of SME on F-actin content was significantly potentiated by shear stress (IOD of SME protection group dynamic 29393 ± 2549 vs. static 26929 ± 2685, P = 0.004; two-way ANOVA, P = 0.031 for interaction). Concurrent measurement of cell area (Fig. 6B) demonstrated a parallel trend, confirming that SME mitigated H₂O₂-induced cell contraction and that this protection was enhanced by flow (area of SME protection group dynamic 1680 ± 172 μm² vs. static 1265 ± 148 μm², P < 0.001; two-way ANOVA, P = 0.042 for interaction).

Furthermore, oxidative stress induced nuclear condensation, as evidenced by a significant increase in nuclear circularity (Fig. 6C). SME treatment effectively reduced circularity under both conditions. Again, the effect was significantly stronger under shear stress, with the nuclear circularity of the dynamic SME group (0.89 ± 0.04) being fully restored to the level of the normal control and significantly lower than that of the static SME group (0.92 ± 0.05, P < 0.01).

Finally, skeletonization analysis of the actin network provided additional evidence at the single-fiber level (Fig. 6D). The frequency distribution of actin filament lengths showed that H₂O₂ injury shifted the distribution leftward (toward shorter fibers) and narrowed the curve. While static SME offered a modest improvement, dynamic SME treatment resulted in a rightward shift and a broader distribution of fiber lengths. This indicated that shear stress not only enhanced the quantity of F-actin preserved by SME but also improved its architectural organization, promoting the formation of longer, more mature stress fibers.

Oxidative stress factor test

The study conducted a comparison of oxidative stress factors levels among the normal control group, H₂O₂ damage group, and SME protection group under both static and dynamic conditions (Table 2).

Under static conditions, H₂O₂ injury induced a severe oxidative stress response, with ROS levels surging to 38.64 ± 3.28 (P < 0.001 vs. normal control). SME treatment significantly ameliorated this damage, reducing ROS levels to 17.90 ± 1.35 (P < 0.001 vs. H₂O₂ group). Strikingly, the protective efficacy of SME was markedly enhanced under dynamic shear stress (two-way ANOVA, p < 0.001 for interaction). While H₂O₂ stimulation elevated ROS to 47.16 ± 5.29, SME administration potently suppressed this increase, bringing ROS levels down to 13.73 ± 1.47 (P < 0.001 vs. H₂O₂ group). This marked enhancement demonstrates a clear synergistic effect between the mechanical stimulus and the pharmacological action of SME in alleviating oxidative damage.

With respect to SOD content, in the static state, the normal control group exhibited a level of 51.26 ± 2.31 U/mg protein, the H₂O₂ damage group displayed a decrease to 17.26 ± 1.70 U/mg protein, and the SME protection group showed a level of 36.25 ± 2.62 U/mg protein. The SOD content in the SME protection group was significantly higher than that in the H₂O₂ damage group (P < 0.001). In the dynamic state, the normal control group had a SOD content of 49.89 ± 3.16 U/mg protein, the H₂O₂ damage group decreased to 13.31 ± 1.03 U/mg protein, and the SME protection group exhibited a level of 41.24 ± 2.23 U/mg protein. This indicates a significant increase in the SOD content in the SME protection group when compared to the H₂O₂ damage group (P < 0.001). Moreover, the increase in SOD content in the SME protection group was more pronounced under dynamic conditions (two-way ANOVA, P < 0.001 for interaction).

Furthermore, with regards to MDA content, in the static state, the normal control group showed a measurement of 0.63 ± 0.03 nmol/mg protein, which increased to 1.53 ± 0.18 nmol/mg protein in the H₂O₂ damage group, and was recorded at 0.88 ± 0.15 nmol/mg protein in the SME protection group. In the dynamic state, the normal control group exhibited a measurement of 0.71 ± 0.12 nmol/mg protein, which increased to 1.65 ± 0.11 nmol/mg protein in the H₂O₂ damage group, and to 0.74 ± 0.08 nmol/mg protein in the SME protection group decreased. The MDA content in the SME protection group was significantly lower than that in the H₂O₂ damage group (P < 0.001). Similarly, the reduction effect of MDA content in the SME protection group was more prominent under dynamic conditions (two-way ANOVA, P = 0.035 for interaction).

In conclusion, it can be observed that under dynamic (shear stress) conditions, SME demonstrated superior performance in reducing ROS levels, increasing SOD content, and decreasing MDA production, thereby providing enhanced protection for ECs.

Table 2.

The influence of SME on the production of oxidative factors by ECs.

Group		Normalized ROS level (fluorescence intensity/mg protein)	SOD content (U/mg protein)	MDA content (nmol/mg protein)
Static condition	Normal Control	6.40 ± 1.11**	51.26 ± 2.31**	0.63 ± 0.03**
	H2O2 Damage Group	38.64 ± 3.28	17.26 ± 1.70	1.53 ± 0.18
	SME Protection Group	17.90 ± 1.35**	36.25 ± 2.62**	0.88 ± 0.15**
Dynamic condition	Normal Control	6.87 ± 1.03**	49.89 ± 3.16**	0.71 ± 0.12**
	H2O2 Damage Group	47.16 ± 5.29	13.31 ± 1.03	1.65 ± 0.11
	SME Protection Group	13.73 ± 1.47** ##	41.24 ± 2.23** #	0.74 ± 0.08**

Data are presented as mean ± SD (n = 5 independent experiments). Statistical significance was determined using two-way ANOVA with an interaction model, followed by LSD post hoc test and simple-effect analysis. In the same condition, compared to the H₂O₂ damage group: ^**P < 0.01; for comparisons within the different condition: ^#P < 0.05, ^##P < 0.05.

Absorption of SME bioactive components by ECs under shear stress

In this study, the impact of shear stress on the absorption of bioactive components from SME by ECs was investigated utilizing high-performance liquid chromatography (HPLC). The primary active compounds analyzed included danshensu, protocatechuic acid, protocatechuic aldehyde, salvianolic acid A, salvianolic acid B, dihydrotanshinone, tanshinone I, cryptotanshinone, and tanshinone IIA (Table 3).

Table 3.

Identification of bioactive compounds from SME extract.

Peak	Retention Time (min)	Identified compound	Absolute concentration (µg/mL)		Normalized concentration (µg/mg protein)
			Static Condition	Dynamic Condition	Static Condition	Dynamic Condition
1	3.055	danshensu	0.53 ± 0.03	1.88 ± 0.03**	0.19 ± 0.02	0.36 ± 0.02**
2	8.718	protocatechuic acid	1.08 ± 0.05	1.80 ± 0.05**	0.39 ± 0.05	0.37 ± 0.05**
3	13.770	protocatechuic aldehyde	0.51 ± 0.03	0.95 ± 0.03**	0.18 ± 0.02	0.19 ± 0.01
4	44.416	salvianolic acid B	1.63 ± 0.14	2.22 ± 0.05**	0.58 ± 0.08	0.54 ± 0.02
5	46.112	salvianolic acid A	Not Detected	0.51 ± 0.14	Not Detected	0.08 ± 0.02
6	56.266	dihydrotanshinone	0.24 ± 0.03	0.43 ± 0.04**	0.09 + 0.01	0.10 ± 0.01
7	59.900	tanshinone I	Not Detected	Not Detected	Not Detected	Not Detected
8	60.801	cryptotanshinone	Not Detected	0.60 ± 0.10	Not Detected	0.06 ± 0.01
9	67.011	tanshinone IIA	0.39 ± 0.04	0.71 ± 0.04**	0.14 ± 0.02	0.15 ± 0.02

The HPLC chromatogram of SME’s standard reference materials clearly resolved 9 characteristic peaks for active components (Fig. 7a). Under static conditions, ECs predominantly absorbed six active compounds comprising danshensu, protocatechuic acid, protocatechuic aldehyde, salvianolic acid B, dihydrotanshinone, tanshinone I and tanshinone IIA (Fig. 7b). However, when exposed to shear stress, ECs not only absorbed these six active compounds but also salvianolic acid A and cryptotanshinone (Fig. 7c).

Fig. 7.

Absorption of SME active components by ECs under different conditions. (a) denotes SME active component standard substance; (b) denotes absorption of SME active components by ECs under static condition; (c) denotes absorption of SME active components by ECs under dynamic condition.

Quantitative analysis, performed via calibration curves with excellent linearity (R²>0.999), revealed that the absolute concentrations of all detectable compounds were significantly higher under shear stress compared to static conditions (P < 0.01 for all, Table 3), a finding likely correlated with the observed increase in cell viability under dynamic culture.

To precisely evaluate cellular uptake efficiency, the absolute concentrations were normalized to the total cellular protein content. This normalized concentration (µg/mg protein) data provided a more accurate comparison (Table 3). Notably, following normalization, the statistically significant enhancement by shear stress was retained for danshensu and protocatechuic acid (P < 0.01 for all), while the differences for other compounds were no longer significant.

All data are presented as mean ± SD (n = 5 independent experiments). Statistical significance was determined by independent samples t-test to investigate the difference in the uptake of each SME compound between culture conditions (static vs. dynamic). Compared to the static condition: ^**P < 0.01. For compounds that were not detected, no statistical comparison was made.

Discussion

This study establishes that physiological laminar shear stress significantly amplifies the protective effects of SME on ECs through three concerted mechanisms: by preconditioning ECs to enhance their intrinsic resilience, by improving the convective delivery and cellular uptake of bioactive compounds, and by inducing a critical bio-synergy that potentiates SME’s pharmacological efficacy. These findings demonstrate that the mechanical microenvironment is a decisive determinant of natural medicine effectiveness, a factor largely overlooked in conventional static screening models.

Firstly, previous studies have demonstrated that shear stress acts as a critical regulator of EC homeostasis by promoting cytoskeletal reorganization, strengthening intercellular junctions, and activating pro-survival signaling pathways²⁸. This mechanical preconditioning renders ECs more resilient to oxidative insult. Indeed, the baseline protective effect of shear stress is corroborated by the cell viability data, which revealed a marked attenuation of H₂O₂-induced disruption under dynamic conditions relative to static cultures. Laminar shear stress fosters a healthier, more intact endothelium, equipping it with a greater inherent capacity to counteract oxidative stress.

Secondly, the enhanced protection under flow can be attributed to a fundamental improvement in drug delivery. In a static culture environment, the transport of SME’s constituents to the cell surface relies solely on passive diffusion, a slow and inefficient process that often creates a nutrient and drug-depleted zone around the cells (known as the diffusion boundary layer)²⁹, which constitutes a major physical barrier to effective compound exposure. In contrast, the application of laminar shear stress actively perfuses the culture medium, effectively eliminating this stagnant boundary layer and ensuring a much more consistent and efficient convective delivery of these compounds directly to the cell surface³⁰. This critical improvement in physical delivery is the primary mechanism robustly corroborated by the HPLC data, which showed the absolute concentrations of every detectable compound were significantly higher in the dynamic group. Moreover, the flow-enabled detection of specific components that were undetectable under static conditions, namely salvianolic acid A and cryptotanshinone, offers strong support for this elimination of diffusion barriers and enhanced bioavailability. However, when these absolute concentrations were normalized to cellular protein content to isolate and assess uptake efficiency per cell, a more nuanced picture emerged: only the normalized concentrations of danshensu and protocatechuic acid were significantly increased under dynamic conditions compared to the static condition. This suggests that their uptake may involve shear-sensitive transporters or pathways, a hypothesis that requires future investigation.

Thirdly, the data on morphology, cytoskeleton, and oxidative stress reveal a potent synergistic interplay between shear stress and SME’s pharmacological mechanisms. While the improved compound delivery (as shown by HPLC) is a key factor, it may not fully account for the magnitude of the functional improvement. Based on established literature, we hypothesize that this synergy may arise from a convergence of signaling pathways. For instance, it is well-documented that shear stress activates the Nrf2-Keap1 antioxidant pathway³¹, and salvianolic acids in SME are also recognized Nrf2 agonists³². It is therefore plausible that the concomitant application of both could elicit a more robust antioxidant response (e.g., greater SOD upregulation and ROS suppression) than either alone, though this specific mechanistic interplay remains to be directly confirmed. Similarly, the convergence of shear stress and SME on the well-documented PI3K/Akt pathway^14,33 presents a compelling mechanistic candidate to explain the superior restoration of metabolic activity and cytoskeletal architecture observed under flow. Thus, beyond enhancing delivery, our functional data suggest that shear stress may critically modulate the pharmacodynamic response to SME. Future work specifically measuring the activation status of these pathways under combined stimulation is needed to validate these hypotheses. This study provides the foundational evidence that such a synergy exists, positioning the mechanical microenvironment as a decisive factor in natural medicine efficacy.

In summary, these findings demonstrate that physiological laminar shear stress profoundly enhances the efficacy of SME through three integrated mechanisms: it preconditions ECs to bolster their intrinsic resilience against oxidative injury; it fundamentally enhances drug delivery by overcoming diffusional barriers, thereby increasing the cellular availability of bioactive compounds; and it induces a critical bio-synergy that potentiates SME’s pharmacodynamic actions. Thus, by accounting for the mechanical microenvironment, this study provides a more physiologically relevant framework for evaluating natural medicines, revealing that shear stress can fundamentally reshape the functional therapeutic profile of complex botanical extracts.

This work challenges the conventional static screening paradigm and firmly integrates the evaluation of natural medicines into the emerging field of “mechanopharmacology.” By providing the first systematic evidence that shear stress simultaneously reshapes both the pharmacokinetic and pharmacodynamic profiles of a natural extract, this study establishes a more physiologically relevant framework for future drug discovery and efficacy assessment.

While the parallel-plate flow chamber system provides a controlled method for applying defined shear stress, it is a simplification of the complex, pulsatile, and multi-directional flow found in human vessels. Furthermore, this study focused on a single shear stress magnitude, representative of healthy arterial flow. Future experiments will explore a range of shear stresses (including disturbed flow patterns) to establish a more comprehensive dose-response relationship between mechanical force and drug effect.

Beyond these technical considerations, the physiological relevance of these findings is supported by the use of the EA.hy926 cell line, a well-established model for endothelial mechanobiology³⁴. This model demonstrates robust responsiveness to fluid shear stress and expresses a repertoire of functional transporters, as directly evidenced by the observed enhancement in compound uptake^35,36. Nevertheless, it is acknowledged that, as an immortalized line, EA.hy926 may not fully replicate all aspects of primary endothelial cells. Thus, future studies employing primary endothelial cells represent a valuable next step to further confirm the translational potential of the synergism between fluid shear stress and natural medicine efficacy uncovered in this work.

Conclusions

This study demonstrates that physiological laminar shear stress significantly amplifies the protective effects of SME on ECs against H₂O₂-induced oxidative damage. The results establish the mechanical microenvironment as a critical determinant of natural medicine efficacy in vitro. Consequently, incorporating physiological shear stress provides a more physiologically relevant model for evaluating the pharmacological activity of complex natural products like SME, primarily by improving compound delivery and modulating cellular physiology. Although this study highlights the importance of mechanical forces, other factors such as pharmacokinetic parameters and compound metabolism also significantly impact the translational potential of natural medicine research and warrant further investigation.

Methods

Cells and culture conditions

ECs were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum, 100 U/mL penicillin, and 100 µg/mL streptomycin. All reagents were obtained from Thermo Fisher Scientific China (Shanghai, China). The cells were incubated in a humidified atmosphere containing 5% CO₂ at 37℃. The EC line was a human umbilical vein endothelial cell line (EA.hy926) commercially purchased from Procell Co., Ltd. (Wuhan, China). Cells between passages 4–6 were used for all experiments to maintain phenotypic stability.

Fabrication of the in vitro cell culture flow device

The in vitro cell culture flow device consists of a parallel plate flow chamber, a peristaltic pump, medical silicone tubes, and a reservoir (Fig. 8a). The perfusion fluid is delivered by the peristaltic pump, flowing through the tubes into the flow chamber, where it exerts shear stress on the cells. The fluid then returns to the reservoir through the tubes, creating a closed circulation system. The dimensions of the flow chamber are designed to ensure laminar, two-dimensional, and fully developed flow, with the size of L > > h, W > > h, L/h > > 10 (Fig. 8b). Cells cultured in this flow chamber experience continuous and stable stimulation from shear stress, mimicking the physiological conditions of blood flow in vessels.

Fig. 8.

The schematic diagram of cell culture flow device. (a) denotes cell culture flow device (1. reservoir, 2. peristaltic pump, 3. parallel plate flow chamber); (b) denotes parallel plate flow chamber: A: flow chamber (L = 7.5 cm, W = 2.5 cm, H = 0.2 mm), B: silicone gasket (h = 1.2 mm), C1, C2, C3: organic glass, D: perfusion fluid inlet, E: perfusion fluid buffer dispersion zone, F: downward dispersion zone of perfusion fluid, G: perfusion fluid outlet.

Screening of effective concentrations of H₂O₂ and SME

The MTT colorimetric method was used to screen the effective concentration of H₂O₂ that causes damage to ECs and the protective concentration of SME for ECs. The experimental design comprised four distinct groups. The normal control group consisted of ECs cultured in growth medium without any treatment. The H₂O₂ damage group was established by treating ECs with a specific concentration of H₂O₂ to induce oxidative injury. The SME protection group involved ECs pre-treated with various concentrations of SME prior to the H₂O₂ challenge. Finally, a blank control group, containing medium without cells, was included to account for background absorbance in the assay. All concentration-screening experiments (H₂O₂ and SME) were conducted under static conditions to determine baseline effective concentrations, which were then used in subsequent dynamic (shear stress) experiments.

ECs in the logarithmic growth phase were seeded in 96-well plates at a density of 2 × 10⁵/mL and cultured until reaching confluence. The culture medium was then replaced with serum-free RPMI 1640 medium, and the cells were starved for 12 h. After 4 h of H₂O₂ (with concentrations ranging from 100 µmol/L to 1600 µmol/L) treatment, cell metabolic viability was assessed using the MTT colorimetric method (details provided in Sect. 5.4) to determine the effective damage concentration of H₂O₂ for ECs.

In the screening of the protective concentration of SME for ECs. The procedure for the blank control group, normal control group, and H₂O₂ damage group was consistent with the above description. In the SME protection group, after 12 h of cell starvation, varying concentrations of SME (ranging from 10 µg/mL to 80 µg/mL) were added to the wells, and the cells were cultured for 24 h before being exposed to the previously determined damaging concentration of H₂O₂. Cell metabolic viability was then assessed to determine the effective protective concentration of SME. SM was procured from CI YUAN BIOCHEMICAL TECHNOLOGY Co., Ltd. (Shanxi, China), and SM was subjected to reflux extraction with 95% ethanol, retaining the medicinal residue. The residue was then extracted with 8–10 times the volume of water (refluxed at 85℃ for 1.5 h, repeated twice), filtered, concentrated, and combined with the ethanol and water extracts. The extracted compounds were subjected to a single-step reflux extraction using a mixture of ethanol and water (70:30). The mixture was then concentrated under vacuum at temperatures ranging from 50 to 75 °C to remove the ethanol. Subsequently, the concentrated solution was freeze-dried at temperatures between − 20 and − 40 °C to obtain a powder form. For experimental purposes, the powder was dissolved directly in serum-free RPMI 1640 medium.

Cell metabolic viability experiment

After coating the glass slide with a gelatin solution for 3 h, a prepared suspension of 2 × 10⁵/mL ECs was seeded onto the slide. Once the cells adhered to the surface, the slide was placed in a parallel plate flow chamber. The in vitro cell culture flow device was installed, and the reservoir was filled with RPMI 1640 medium containing 10% fetal bovine serum. The rotational speed of the peristaltic pump was adjusted, and the flow device was initiated before being placed in a 37 °C, 5% CO₂ incubator for 8 h. The medium was then changed to a serum-free medium, and the culture continued for an additional 12 h. The concentration of the screened SME was then added for another 24 h of culture, followed by the addition of a certain concentration of H₂O₂ to damage the ECs for 4 h. As a control, ECs were grown without exposure to shear stress. Normal control groups and H₂O₂ groups were set up simultaneously. The EC suspensions were collected from the slides and centrifuged at 1000 rpm for 5 min at 25 °C. The cells were then resuspended in medium and added to a 96-well plate. An MTT cell proliferation and cytotoxicity assay kit (COIBO BIO, China) was used to assess cell metabolic viability. In brief, the MTT reagent was added to the cell suspensions in the 96-well plates at a volume of 10 µL per well, followed by the addition of 180 µL of serum-free RPMI1640 medium. The cells were then kept in a humidified cell culture chamber for 4 h. After carefully removing the culture medium from each well, 150 µL of dimethyl sulfoxide (DMSO) was added to solubilize the formazan crystals. The samples were gently shaken for 10 min, and the optical density (OD) was measured at 570 nm using a universal microplate spectrophotometer (Bio-Rad). Cell metabolic viability was determined using the formula: Cell Metabolic Viability (%) = (OD_T-OD_B)/(OD_C-OD_B)×100%. Here OD_T represents the OD value of the test group, OD_B denotes the OD value of the blank control group and ODc stands for the OD value of the normal control group.

Examination of cell morphology

After treatment, the morphology of ECs was examined using an IX71 inverted microscope equipped with phase contrast capability (Olympus, Shinjuku, Tokyo, Japan). Images were captured using a charged-coupled device (CCD) Digital Camera 320 (Tykor, Guangzhou, China) and analyzed with Image-Pro Plus software (version6.0.0.260, Media Cybernetics, USA, https://my.mediacy.com/).

The outlines of 100 individual cells were manually traced, and their respective areas and perimeters were measured using the software. The cell shape index (CSI) was then calculated for each cell using the formula CSI = 4π × (cell area)/(cell perimeter)², with the mean value being determined. The CSI reflects the degree of cellular morphological deviation from a circular shape: a CSI of 1 signifies a perfectly circular cell, while a value approaching 0 indicates a more elongated or polygonal shape, which is closely associated with the functional state of ECs. Under normal physiological conditions, functional ECs typically exhibit a lower CSI due to their elongated morphology, whereas abnormal or damaged ECs often display a higher CSI with a more rounded, “cobblestone-like” appearance³⁷.

Cell staining assays

To assess cell viability and cytoskeletal morphology, the following two staining methods were employed.

Calcein-AM and PI staining for cell viability

After treatment, ECs were stained directly in culture medium with Calcein-AM (1 µL/mL) and PI (10 µL/mL) (Beyotime, China) and incubated for 20 min. The samples were then observed and imaged under a fluorescence microscope (ECLIPSE Ti-s, Nikon, Japan). Viable cells exhibited green fluorescence, while dead cells showed red fluorescence. The numbers of live and dead cells were counted to calculate the dead cell rate (DCR) for each group. The protection rate of SME was calculated as follows: Protection rate (%) = (DCR_D − DCR_P)/(DCR_D − DCR_C) × 100%, where C denotes the normal control group, D denotes the H₂O₂ damage group, and P denotes the SME protection group.

Phalloidin and 4’,6-diamidino-2-phenylindole (DAPI) staining for cytoskeleton and nuclear morphology

After treatment, the culture medium was aspirated and cells were washed twice with pre-warmed 1×PBS (pH 7.4). Cells were fixed with 4% paraformaldehyde in PBS for 10 min at room temperature, followed by washing with PBS 2–3 times (10 min each). Permeabilization was performed using acetone (≤−20 °C) or 0.5% Triton X-100 for 5 min, followed by PBS washing. Cells were incubated with 200 µL of TRITC-conjugated phalloidin working solution (Solarbio, China) for 30 min at room temperature in the dark. After washing with PBS, nuclei were counterstained with DAPI (100 nM; Beyotime, China) for approximately 30 s. Coverslips were mounted with Fluoromount-G™ (Beyotime, China) and sealed with nail polish. Images were acquired using a confocal laser scanning microscope (Leica SP8, Germany) with excitation/emission filters set at 540/570 nm for TRITC and 364/454 nm for DAPI.

Quantitative analysis of the acquired images was performed using Image-Pro Plus software (version6.0.0.260, Media Cybernetics, USA, https://my.mediacy.com/). Following OD calibration, the fluorescence intensity of F-actin (TRITC channel) and nuclear morphology (DAPI channel) were assessed. For each cell, the IOD and area were measured. The circularity of nuclei was calculated using the formula: Circularity = 4π × Area/Perimeter². Higher nuclear circularity values (closer to 1, representing a perfect circle) are characteristic of apoptotic and necrotic cells, which exhibit a rounded morphology. Furthermore, the phalloidin-stained images were subjected to skeletonization analysis to determine the length distribution of actin filaments. Data collected from at least 20 cells per group across five independent experiments were averaged and subjected to statistical analysis.

Detection experiments of oxidative stress factors (ROS, SOD and MDA)

After being treated with SME and H₂O₂ under both dynamic or static conditions, the cell suspension was prepared and subjected to centrifugation for cell collection. The collected cells were then incubated with HBSS buffer containing 2’,7’-dichlorofluorescein-diacetate (DCF-DA) in a 37 °C cell culture incubator in the dark for 30 min. Following incubation, the cells were washed 1–2 times with serum-free culture medium to ensure complete removal of uninternalized DCF-DA. The intracellular ROS levels were assessed using the fluorescent probe DCF-DA. The DCF fluorescence was measured at an excitation wavelength of 488 nm and an emission wavelength of 525 nm using a fluorescence microplate reader (YOUYUNPU, China). To eliminate potential interference from differences in cell viability or number, the ROS fluorescence intensity was normalized to the total protein content of each sample. After detecting ROS fluorescence, the same cell pellets were lysed, and protein concentration was determined using a BCA kit (Beyotime, China). The ROS levels were ultimately expressed as “fluorescence intensity/mg protein”.

For the detection of SOD and MDA, ECs were harvested through trypsinization, washed twice with PBS at 4 °C, and centrifuged at 1000 rpm for 5 min at 25 °C. The supernatant was then discarded, and the cell pellet was lysed on ice with dedicated SOD sample preparation solution (for SOD detection) or Western and IP cell lysis buffer (for MDA detection). Following centrifugation at 12,000 rpm for 10 min at 4 °C, the supernatant was collected as the sample for SOD and MDA content determination. The total SOD activity and MDA content were measured using the total SOD activity assay kit (WST-8 method, Beyotime, China) and lipid peroxidation (MDA) assay kit (Beyotime, China) respectively as per manufacturers’ instructions. The absorbance were measured using a microplate reader (xMark, Bio-Rad, USA) with SOD activity assessed at 450 nm and MDA detected at 530 nm. The calculation formula for SOD enzyme activity was: SOD enzyme activity unit (U) in the sample = inhibition percentage/(1-inhibition percentage). The SOD activity unit was converted to U/mg protein based on the protein concentration and dilution factor of the sample. The MDA molar concentration of the sample was determined using a standard curve. The MDA content in the original sample was expressed based on the protein content per unit weight. Protein concentration in cell lysates was quantified using the BCA kit (Beyotime, China) following the manufacturer’s instructions, with absorbance measured using the same xMark microplate reader (Bio-Rad, USA).

HPLC analysis

ECs were subjected to treatment with SME and H₂O₂, followed by the addition of 2 mL 75% ethanol to lyse the cells and extract their intracellular components. The samples were then subjected to ultrasonic homogenization using a Scientz instrument (China) and subsequently centrifuged. The resulting supernatant was lyophilized using a vacuum freeze dryer (Thermo Fisher Scientific, Shanghai, China), dissolved in 70% methanol, and filtered through a 0.45 μm nylon membrane (Xingya, Shanghai, China) prior to HPLC analysis. Mixed standard samples, as well as EC samples treated with SME under static or dynamic conditions, were prepared for analysis by HPLC. The specific chromatographic analysis conditions used in this study have been described in previous publications³⁸. Peaks in the HPLC chromatogram were identified based on a threshold criterion, with the peak area being required to be at least 3% of that of protocatechuic aldehyde.

The absolute quantification of each target compound was achieved using a calibration curve established by analyzing a series of standard solutions with known concentrations (ranging from 1 to 10 µg/mL). The concentrations in the cell lysates were calculated by interpolating the measured peak areas onto their respective calibration curves. To account for variations in cell number and lysate volume, these absolute concentrations were subsequently normalized to the total cellular protein, which was determined using a BCA kit (Beyotime, China) according to the manufacturer’s instructions. The final uptake data are expressed as micrograms of compound per milligram of total cellular protein (µg/mg protein).

Statistical analysis

All data were expressed as mean ± standard deviation (SD) and analyzed using SPSS, with significance defined as P < 0.05 (significant) and P < 0.01 (highly significant). All datasets were first assessed for normality (Shapiro–Wilk test) and homogeneity of variance (Brown–Forsythe test) (α = 0.05). For control-normalized data, based on the principle of mathematical equivalence, statistical comparisons were performed using blank-subtracted original OD values rather than percentages, which is both rigorous and avoids issues with variance homogeneity. Depending on whether the assumptions were met, one-way ANOVA (followed by Dunnett’s test) or Welch’s ANOVA (followed by Games-Howell test) was used to compare H₂O₂ or SME treatment groups against the normal or damage control groups, respectively. For comparing static versus dynamic conditions, two-way ANOVA with an interaction model or independent samples t-test was applied to evaluate the significance.

Author contributions

Y.X. designed the experimental protocols and oversaw the project implementation. T.Z. performed the cell culture experiments and collected the morphological data. L.W. conducted the biochemical assays for ROS and MDA detection, and analyzed the HPLC results. M.W. contributed to the statistical analysis and prepared the initial figures. Y.L. drafted the manuscript, with critical revisions from GH on the theoretical interpretation. Y.W. supervised the study. All authors read and approved the final manuscript.

Data availability

All data generated or analysed during this study are included in this published article.

Declarations

Competing interests

The authors declare no competing interests.