Ginsenoside Rb1 ameliorates endothelial dysfunction in sepsis: Targeting the RhoA/ROCK pathway to modulate mitochondrial homeostasis and inflammation

Abstract

Background

Sepsis-induced endothelial dysfunction drives organ injury through microcirculatory failure, characterized by barrier disruption and heterogeneous blood flow. Ginsenoside Rb1 (G-Rb1), a bioactive ginseng constituent, exhibits potent anti-inflammatory and antioxidant properties. This study investigates whether G-Rb1 ameliorates sepsis by preserving endothelial barrier integrity via modulating mitochondrial homeostasis and inhibiting the RhoA/ROCK pathway.

Methods

A murine cecal ligation and puncture (CLP) sepsis model and lipopolysaccharide (LPS)-stimulated human umbilical vein endothelial cells (HUVECs) were employed. Endothelial barrier function, mitochondrial dynamics, and inflammatory responses were assessed via histology, flow cytometry, Western blot, and Seahorse XF analysis. RhoA/ROCK inhibitor (Fasudil) and Drp1 inhibitor (Mdivi-1) served as mechanistic controls.

Results

G-Rb1 significantly improved survival in CLP mice and attenuated lung injury. It suppressed pro-inflammatory cytokines (TNF-α, IL-6, IL-1β) in bronchoalveolar lavage fluid and rebalanced T cell percentage. In HUVECs, G-Rb1 restored endothelial junction proteins (VE-cadherin, ZO-1, Claudin-5), reduced ROS/MDA, and enhanced SOD/GSH/ATP levels. Mechanistically, G-Rb1 inhibited RhoA activation and ROCK-mediated Drp1 phosphorylation (Ser616), normalizing mitochondrial fission, membrane potential, and oxidative phosphorylation.

Conclusion

G-Rb1 alleviates sepsis-induced endothelial dysfunction by suppressing the RhoA/ROCK/Drp1 axis, thereby restoring mitochondrial homeostasis and barrier integrity. These findings highlight G-Rb1 as a promising therapeutic candidate for sepsis management via dual modulation of cytoskeletal dynamics and mitochondrial quality control.

Graphical abstract

Highlights

• •Ginsenoside Rb1 regulates RhoA/ROCK to inhibit excessive immune responses in sepsis.
• •Ginsenoside Rb1 protects endothelial barrier to attenuate tissue damage in sepsis.
• •Ginsenoside Rb1 inhibits RhoA/ROCK maintain mitochondrial homeostasis in endothelium.

1. Introduction

The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) define sepsis as “life-threatening organ dysfunction caused by a disordered host response to infection” [1]. Sepsis is an infectious syndrome with high morbidity and mortality, characterized by a severe dysregulated inflammatory response that can rapidly progress to multiple organ dysfunction, with the respiratory and cardiovascular systems being particularly susceptible [2]. Severe endothelial injury and the resulting microcirculatory dysfunction, including impaired vascular reactivity and endothelial barrier disruption, are critical in sepsis pathogenesis [3,4]. This is especially evident in the lungs, where early endothelial injury causes fluid and protein leakage into alveolar walls, leading to interstitial thickening, impaired oxygenation, and potentially Acute Respiratory Distress Syndrome (ARDS) [5,6]. Therefore, reducing endothelial permeability, restoring normal signaling, and re-establishing homogeneous blood flow are key to preventing or reversing sepsis-induced lung injury.

Among the various intracellular mechanisms driving this endothelial barrier disruption and hyperpermeability, the Ras homologous gene family member A (RhoA) signaling pathway has emerged as a pivotal regulator. RhoA, a small GTPase, is a key regulator of fundamental cellular processes such as cell motility, morphology, and contraction [7]. In sepsis, vasoactive substances like endotoxin and thrombin activate RhoA, which then regulates pathological pathways such as NF-κB phosphorylation, actin cytoskeleton dynamics, and reduced expression of tight and adherens junction proteins. This induces a contractile, high-permeability endothelial phenotype, establishing RhoA as a key regulator of endothelial barrier function and vascular reactivity [8,9].

Imbalance in mitochondrial homeostasis, causing abnormal energy metabolism and redox stress, plays a pivotal role in endothelial inflammatory activation and dysfunction during sepsis, directly impacting organ failure and mortality [7,10]. Consequently, restoring mitochondrial homeostasis is a viable therapeutic target. Mitochondria maintain function through a coordinated fission-fusion cycle, which is disrupted in sepsis, leading to excessive mitochondrial fission, a process dependent on the actin cytoskeleton [11]. Excessive inflammatory cytokines and reactive oxygen species induce RhoA nitration and activation. Activated RhoA subsequently stimulates ROCK, leading to remodeling of the actin cytoskeleton. In addition, ROCK promotes Drp1 phosphorylation at serine 616. This phosphorylation facilitates Drp1 oligomerization on the outer mitochondrial membrane (OMM). Through GTP hydrolysis, this complex mediates OMM division, while calcium influx promotes inner mitochondrial membrane (IMM) constriction [12,13]. Therefore, inhibiting RhoA activity to restore mitochondrial homeostasis is a potential therapeutic strategy to alleviate microcirculatory dysfunction in sepsis.

Various ginsenosides, the primary active constituents of Panax ginseng C. A. Meyer, have demonstrated protective effects against sepsis through multiple mechanisms [14]. Among these, G-Rb1 has reported anti-inflammatory, antioxidant, and anti-apoptotic properties and protects against inflammation-induced acute lung injury [[15], [16], [17], [18], [19]]. This study investigates whether G-Rb1 ameliorates sepsis by modulating the RhoA signaling pathway and Drp1-mediated mitochondrial homeostasis, thereby inhibiting aberrant actin remodeling and preserving endothelial function. The RhoA/ROCK inhibitor Fasudil and the Drp1 inhibitor Mdivi-1 were used as mechanistic controls to evaluate G-Rb1's regulatory effects on pulmonary vascular injury.

2. Materials and methods

2.1. Animals

Male C57BL/6JNifdc mice (8–10 weeks, 20–25 g) were obtained from Beijing Vital River Laboratory Animal Technology (Beijing, China). All procedures followed the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Animal Ethics Committee of the Experimental Research Center, China Academy of Chinese Medical Sciences (Approval Number: ERCCACMS21-2502-12). Mice (n = 15/group) were randomized into Sham, CLP, G-Rb1 (20 mg/kg, i.p.), and Fasudil (30 mg/kg, i.p.) groups. The selected dose of G-Rb1 (20 mg/kg) was based on previous literature demonstrating its efficacy in improving sepsis outcomes in murine models [20]. As a prophylactic intervention, G-Rb1 and Fasudil were administered in saline once daily for 5 days prior to modeling. The final dose was given 2 h prior to modeling on the day of surgery. Sepsis was induced via cecal ligation and puncture (CLP) under general anesthesia induced with 5% isoflurane and maintained with 2% isoflurane; the cecum was ligated at its midpoint, punctured once with a 21-gauge needle. Fecal extrusion was gently performed to ensure steady fecal flow into the peritoneal cavity, excess feces were wiped away, and the cecum was then returned to the abdominal cavity. Sham mice underwent laparotomy without ligation or puncture. Post-operatively, all mice received 1 mL subcutaneous saline. Following surgery, the mice were continuously monitored for humane endpoints. Euthanasia was promptly performed via intraperitoneal injection of an overdose of pentobarbital (100 mg/kg) if the mice exhibited any of the following conditions after CLP: (1) loose dorsal skin, severely sunken eyes, dry and cyanotic mucous membranes, or severe dehydration; (2) open-mouth breathing, apnea, or obvious respiratory distress (abdominal breathing); (3) convulsions, seizures, opisthotonos, or coma; (4) oliguria or anuria; or (5) extreme abdominal distension accompanied by dehiscence of the surgical incision and the discharge of a large amount of turbid, foul-smelling fluid. Separate mouse cohorts were used for survival analysis and for tissue/biochemical analysis, respectively. For survival monitoring: 15 mice per group were used, and mortality was recorded within 96 h after CLP. For sample collection: Independent batches of mice were deeply anesthetized with pentobarbital (50 mg/kg, i.p.) at 24 h post-surgery. Blood samples were collected via retro-orbital bleeding, after which the mice were euthanized by cervical dislocation for tissue harvesting and subsequent measurements. The exact number of biological replicates (n) is specified in each corresponding figure legend.

2.2. Histopathological observation and evaluation

Lung samples were fixed, paraffin-embedded, and sectioned (∼5 μm) for H&E and immunohistochemistry (IHC). H&E staining assessed pathological changes according to established criteria [21]. For IHC, sections were deparaffinized and blocked for endogenous peroxidase with 3% H₂O₂ for 5 min. Subsequently, sections were blocked with 5% bovine serum albumin (BSA) for 30 min at room temperature to prevent non-specific binding. After blocking, sections were incubated with primary antibodies (VE-cadherin, ZO-1, Claudin-5; 1:500) overnight at 4°C, followed by a biotinylated goat anti-rabbit secondary antibody (1:50) for 30 min. Visualization was performed using a DAB kit. Images were acquired with a VS120-S6-W system (Olympus, Japan) and analyzed using OlyVIA software (v4.1.1). For transmission electron microscopy, mesenteric vessels were selected because they represent a typical and representative site of systemic microcirculation, which is highly susceptible to sepsis-induced endothelial injury and barrier dysfunction [22]. Mesenteric vessels were fixed in 2.5% glutaraldehyde (4°C, 12 h) and post-fixed in 1% osmium tetroxide (4°C, 1 h). After dehydration, samples were embedded in Epon 812, and ultrathin sections (50 nm) were prepared with an EM UC6 ultramicrotome (Leica, Germany). Stained sections were examined under an H-7650 microscope (Hitachi, Japan) for mitochondrial morphology.

2.3. Mesenteric blood flow assessment by laser speckle contrast imaging (LSCI)

Mesenteric blood flow was evaluated using the Moor FLPI-2 full-field laser perfusion imager (Moor Instruments Ltd, UK). Mice were deeply anesthetized with pentobarbital (50 mg/kg, I, p.), and a midline abdominal incision (2–3 cm in length) was made to fully expose the mesentery. The mesenteric vasculature was then gently spread horizontally and immersed in 37°C saline to maintain physiological conditions. Images were acquired at a rate of 30 frames per second over a 4-s period to generate representative color-coded perfusion maps. Blood flow velocity in the mesentery and the corresponding intestinal regions was quantified using the moor FLPI-2 V3.0 software, with results expressed as mean blur rate (MBR).

2.4. Cell culture and cytotoxicity detection

Human umbilical vein endothelial cells (HUVECs) used in this study were provided by the Functional Laboratory of the Experimental Research Center, China Academy of Chinese Medical Sciences. Cells were cultured in endothelial cell medium (ECM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin, and maintained at 37°C in a humidified atmosphere containing 5% CO₂. Cells between passages 3 and 7 were employed for all experiments, and assays were conducted within 2–3 days after the cells reached 70-80% confluence.

HUVECs were seeded in 96-well plates at a density of 10,000 cells per well. When cell confluence reached approximately 70%–80%, the cells were divided into LPS-treated groups with a gradient of LPS concentrations (0, 0.1, 0.3, 1.0, 3.0, and 10.0 μg/ml) for 24 h. Absorbance was then measured at 490 nm to determine LDH content released by the cells into the medium, thereby identifying the optimal LPS concentration for establishing a cell-based sepsis model.

In a separate set of experiments, cells were divided into G-Rb1 treatment groups co-treated with the optimal LPS concentration and a gradient of G-Rb1 concentrations (3, 10, 30, 50, and 100 μmol/L). After 24 h of treatment, LDH levels were measured to select an appropriate G-Rb1 treatment concentration.

For subsequent experiments, cells were divided into the following groups: control, LPS, G-Rb1 treatment, Fasudil positive control, and Mdivi-1 positive control. Cells in the LPS model and treatment groups were exposed to the optimal LPS concentration for 24 h. Fasudil and Mdivi-1 were applied at concentrations of 25 μmol/L and 15 μmol/L, respectively, concurrently with LPS exposure.

2.5. Assessment of mitochondrial morphology

HUVECs were seeded at a density of 50,000 cells per well in 35 mm confocal dishes (coverglass-bottom dishes) pre-coated with Matrigel Matrix. After 12 h of attachment, cells were subjected to the corresponding treatments. Subsequently, mitochondria were stained using MitoTracker™ Green diluted to 2 μmol/L in serum-free ECM and incubated at 37°C for 30 min. Following PBS washes to remove background fluorescence, images of mitochondrial morphology were acquired using an FV4000 confocal microscope (Olympus, Japan). Using ImageJ software, the mitochondrial length and width were measured, and the length/width ratio was calculated to evaluate the degree of mitochondrial fragmentation.

2.6. Statistical analysis

Statistical analyses were performed using GraphPad Prism version 10.0.0 (GraphPad Software, Inc., USA). Data are presented as mean ± standard error of the mean (SEM). Comparisons between two groups were analyzed using unpaired Student's t-test, while one-way analysis of variance (ANOVA) followed by the Newman-Keuls post-hoc test was applied for comparisons among three or more groups. Survival curves were generated using the Kaplan–Meier method and compared by the log-rank test. A p-value < 0.05 was considered statistically significant for all analyses. Unless otherwise stated, all experiments were performed with at least three independent replicates.

More details on the experimental methodology are included in the supplementary file 1- Materials and methods.

3. Result

3.1. G-Rb1 protects against sepsis-induced inflammation and lung injury in mice

To investigate the role of G-Rb1 in sepsis, we established a sepsis model in male C57BL/6JNifdc mice using CLP (Fig. 1 A). Survival analysis revealed that the mortality rate in the CLP group reached 80% within 96 h post-surgery, whereas G-Rb1 treatment significantly reduced mortality to 33.3%. Notably, at the 24-h endpoint, the mortality rate in the CLP group was twice that of the G-Rb1-treated group (Fig. 1 B). H&E staining demonstrated that CLP induced pathological alterations in lung tissue resembling acute respiratory distress syndrome (ARDS), characterized by fluid accumulation in the lungs, inflammation, edema, hemorrhage (Fig. 1 C). The lung injury score was significantly increased in the CLP group compared with the sham group (Fig. 1 D). These pathological changes were markedly attenuated by treatment with either G-Rb1 or the ROCK inhibitor Fasudil. It is reported that Fasudil effectively protects the vascular endothelium, thereby alleviating acute lung injury.

Fig. 1.

G-Rb1 alleviates lung injury and inflammation in sepsis. (A) Schematic overview of the experimental timeline showing CLP-induced sepsis model and drug administration in C57BL/6JNifdc mice. (B) Survival analysis of mice 24 h after CLP surgery (n = 15 per group). Log-rank (Mantel–Cox) test was used to compare survival curves between sham and CLP groups, and between CLP and G-Rb1 treatment groups.(C) Representative hematoxylin and eosin (H&E)-stained lung tissue sections and (D) corresponding lung injury scores at 24 h post-CLP (n = 6).(E) Immunohistochemical (IHC) staining of iNOS in lung tissues and (F) quantitative analysis of the percentage of iNOS-positive cells (n = 6). Error bars represent mean ± s.e.m. Scale bars: 100 μm and 50 μm (C, E).

Immunohistochemical (IHC) analysis revealed that the expression level of inducible nitric oxide synthase (iNOS) in the lung tissue of the CLP group was significantly higher than that of the Sham group (Fig. 1 E). In contrast, iNOS expression was significantly suppressed in both the G-Rb1 and Fasudil groups (Fig. 1 F). During sepsis, iNOS is extensively induced, subsequently catalyzing the production of peroxynitrite. The latter is a key mediator that promotes inflammation, oxidative stress, and hemodynamic instability, leading to impairment of both endothelial barrier and mitochondrial function.

Flow cytometric analysis revealed that at 24 h post-CLP, the percentages of CD45⁺CD3⁺ (Fig. S1 A, B) and CD3⁺CD4⁺ cells (Fig. 2A and B), along with the absolute cell numbers of CD3⁺CD4⁺ cells (Fig. S1 E) in the peripheral blood of septic mice, were significantly decreased compared to the Sham group. Conversely, the percentage of CD3⁺CD8⁺ cells was significantly increased (Fig. 2A–Fig. S1 C), resulting in a markedly reduced CD4⁺/CD8⁺ ratio (Fig. S1 D). Both G-Rb1 and Fasudil effectively reversed these alterations in cell proportions, thereby ameliorating sepsis-induced lymphocyte suppression. In addition, cytokine detection in BALF indicated that, 24 h after CLP, the levels of pro-inflammatory factors TNF-α (Fig. 2 M), IL-6 (Fig. 2 N), IL-1β (Fig. 2 O) and chemokine MCP-1 (Fig. 2 P) in the BALF of septic mice were markedly higher than those in sham mice. This finding aligns with the “cytokine storm” feature seen in patients with clinical sepsis. Treatment with G-Rb1 and Fasudil was able to significantly reverse the abnormal elevation of these cytokines.

Fig. 2.

G-Rb1 regulates immune cells and cytokine profiles in sepsis. (A) Representative gating strategy for flow cytometric analysis of CD3⁺CD4⁺ cells and CD3⁺CD8⁺ in peripheral blood at 24 h post-CLP. (B) Quantitative analysis of the percentage of CD4⁺ T cells among CD3⁺ T cells. (C, E, G, I) Representative flow cytometry plots showing Th1, Th2, Th17, and Treg cell populations gated from CD3⁺ T cells in peripheral blood. (D, F, H, J). Quantitative analysis of the percentages of Th1, Th2, Th17, and Treg cells among CD3⁺ T cells (n = 6). (K) Ratio of Th1 to Th2 cells and (L) ratio of Th17 to Treg cells among CD3⁺ T cells (n = 6). (M − P) Levels of TNF-α, IL-6, IL-1β, and MCP-1 in BALF measured by cytometric bead array at 24 h post-CLP (n = 6). Error bars represent mean ± s.e.m.

Further flow cytometric analysis revealed that the percentages of Th1 (Fig. 2C and D), Th2 (Fig. 2E and F), Th17 (Fig. 2G and H), and Treg cells (Fig. 2I and J) were significantly elevated in the peripheral blood of CLP mice relative to the Sham group. Treatment with G-Rb1 and Fasudil led to a significant reduction in the percentages of Th1, Th2, and Th17 cells, accompanied by an increase in the Treg cell percentage. While the Th1/Th2 ratio showed no significant change (Fig. 2 K), the Th17/Treg ratio was notably lower in the G-Rb1 group and Fasudil group (Fig. 2 L). Taken together, these results demonstrate that G-Rb1 effectively mitigates lung tissue injury and systemic inflammation caused by sepsis.

3.2. G-Rb1 ameliorates microcirculatory dysfunction in septic mice

At 24 h post-CLP surgery, we observed that mice in the CLP group exhibited significant edema and hemorrhage in the lung tissue (Fig. 3 A), along with a significantly increased lung wet/dry weight ratio (Fig. 3 B). In comparison, treatment with G-Rb1 and Fasudil alleviated the extent of lung tissue damage and significantly reduced pulmonary edema and the increase in lung weight. The Evans Blue leakage assay revealed a significant increase in pulmonary vascular permeability 24 h post-CLP, while both G-Rb1 and Fasudil markedly reduced this leakage (Fig. 3F and G). In addition to the lungs, the liver (Fig. 3 C), renal (Fig. 3 D), and spleen (Fig. 3 E) tissues of CLP mice also showed various degrees of edema; intervention with G-Rb1 and Fasudil appeared to alleviate the degree of edema in these organs.

Fig. 3.

G-Rb1 ameliorates microcirculatory dysfunction in sepsis. (A) Representative lung tissue morphology at 24 h post-CLP. (B–E) Wet-to-dry weight ratios of lung, liver, kidney, and spleen tissues at 24 h post-CLP (n = 8). (F) Representative images of lung tissues after Evans blue dye injection at 24 h post-CLP. (G) Quantification of Evans blue content per milligram of dry lung tissue weight (n = 8). (H, J, L) Immunohistochemical staining of Claudin-5, ZO-1, and VE-cadherin in lung tissues. (I, K, M) Quantitative analysis of the percentage of Claudin-5-, ZO-1-, and VE-cadherin-positive cells (n = 6). (N) Laser speckle contrast imaging of mesenteric blood flow at 24 h post-CLP. (O) Quantitative analysis of mesenteric blood flow perfusion rate (n = 6). Error bars represent mean ± s.e.m. Scale bars: 100 μm (A, F); 40 μm and 20 μm (H, J, L).

The disruption of vascular barrier function is primarily associated with the degradation of proteins in adherens junctions (e.g., VE-cadherin) and tight junctions (e.g., Claudin-5, ZO-1). Immunohistochemistry (IHC) results showed that in the lung tissue of the CLP group, the continuity of Claudin-5 (Fig. 3H and I), ZO-1 (Fig. 3J and K) and VE-cadherin (Fig. 3L and M) was disrupted, and their expression levels were significantly decreased. In contrast, treatment with G-Rb1 and Fasudil effectively maintained the expression levels of these three junctional proteins.

To further evaluate the effect of G-Rb1 on microcirculatory perfusion, we used a laser speckle flow imaging system to measure mesenteric microcirculatory blood flow. The results showed that the microcirculatory blood flow in the CLP group was significantly decreased compared to the sham group (Fig. 3N and O). Although intervention with G-Rb1 and Fasudil did not fully restore it to normal levels, they significantly improved blood flow distribution and maintained effective microcirculatory perfusion.

Above results indicate that G-Rb1 can alleviate CLP-induced vascular endothelial barrier disruption and mesenteric blood flow disorders, reduce vascular hyperpermeability, and thereby mitigate the sepsis phenotype. These findings support a beneficial effect of G-Rb1 on regional microcirculation.

3.3. G-Rb1 attenuates LPS-induced endothelial inflammation and hyperpermeability In vitro

To determine whether G-Rb1 directly modulates endothelial inflammatory responses and permeability, we conducted in vitro experiments using HUVECs treated with LPS. By measuring the lactate dehydrogenase (LDH) release rate, we found that stimulation with 10 μg/ml LPS for 24 h led to a significant increase in LDH release (Fig. 4 A). HUVECs were then co-treated with 10 μg/mL LPS and various concentrations of G-Rb1 for 24 h. The results showed that 50 μmol/L G-Rb1 effectively inhibited LPS-induced cytotoxicity (Fig. 4 B).

Fig. 4.

G-Rb1 attenuates LPS-induced inflammatory injury in endothelial cells. (A) LDH release levels in HUVECs treated with different concentrations of LPS for 24 h (n = 3). (B) LDH release levels in HUVECs co-treated with 10 μg/mL LPS and different concentrations of G-Rb1 for 24 h (n = 3). (C-G) Levels of (C) iNOS, (D) MDA, (E) SOD, (F) GSH, and (G) ATP in HUVECs after corresponding treatments for 24 h (n = 3 for C-F; n = 6 for G). (H) Representative fluorescence images of intracellular ROS in differently treated HUVECs. (I) Quantitative analysis of ROS fluorescence intensity (n = 120-150 cells). Error bars represent mean ± s.e.m. Scale bar: 100 μm (H). Representative immunofluorescence images of (J) Claudin-5, (L) ZO-1, and (N) VE-cadherin in HUVECs after respective treatments for 24 h. Quantitative analysis of (K) Claudin-5, (L) ZO-1, and (N) VE-cadherin expression levels based on fluorescence intensity (n = 8). Error bars represent mean ± s.e.m. Scale bars: 50 μm (J, L, N).

Analysis revealed that following LPS stimulation, the levels of iNOS (Fig. 4 C) and malondialdehyde (MDA) (Fig. 4 D) in HUVECs were significantly increased. Immunofluorescence staining and quantitative analysis demonstrated that LPS exposure markedly elevated intracellular ROS levels in HUVECs (Fig. 4H and I). While the levels of SOD (Fig. 4 E) and GSH (Fig. 4 F) were decreased. These results indicate that the cells were in a significant state of oxidative stress. The ATP level in HUVECs was also markedly decreased after LPS stimulation (Fig. 4 G); however, G-Rb1 treatment significantly reversed these changes.

Additionally, immunofluorescence staining showed that LPS caused a marked reduction in the fluorescence intensity of Claudin-5 (Fig. 4J and K), ZO-1 (Fig. 4L and M), and VE-cadherin (Fig. 4N and O) whereas G-Rb1, Fasudil, and Mdivi-1 (a Drp1 inhibitor) all significantly enhanced the expression of these three junctional proteins.

In summary, G-Rb1 can directly alleviate LPS-induced inflammation and oxidative damage in HUVECs at the cellular level and increase cellular ATP levels, suggesting a partial modulation of cellular energy metabolism and redox balance. Additionally, the expression levels of specific intercellular junction proteins (e.g., ZO-1) were modulated by G-Rb1 intervention in LPS-treated HUVECs. The effects of Fasudil and Mdivi-1 further confirm that inhibiting the Rho/ROCK and Drp1-related signaling pathways is an effective approach for protecting endothelial barrier function.

3.4. G-Rb1 exerts therapeutic effects in sepsis by inhibiting the RhoA signaling pathway

To elucidate the molecular mechanisms by which G-Rb1 alleviates inflammatory damage and maintains the endothelial barrier, we performed transcriptomic analysis on HUVECs exposed to LPS alone or with concurrent G-Rb1 treatment. Heatmap (Fig. 5 A) and volcano plot analyses (Fig. 5 B) showed significant differences in gene expression between the LPS group and the G-Rb1-treated group, with a total of 519 genes significantly upregulated and 549 significantly downregulated, amounting to 1068 differentially expressed genes (DEGs). Reactome pathway analysis revealed significant enrichment of pathways such as “Signaling by Rho GTPases, Miro GTPases and RHOBTB3,” “Signaling by Rho GTPases,” and “RHO GTPases cycle” (Fig. 5 C). Gene Ontology (GO) analysis showed significant enrichment of functional terms including “regulation of small GTPase mediated signal transduction,” “Rho protein signal transduction,” “small GTPase binding,” and “GTP regulator activity” (Fig. 5 D). The differentially expressed genes between the two groups were significantly enriched in pathways related to Rho GTPase signaling. These results indicate that G-Rb1 may potentially modulate Rho GTPase signal transduction. RhoA, a member of the small Rho family, plays a key role in small GTPase-mediated signal transduction, and studies have confirmed its involvement in the regulation of inflammation and immunity in sepsis, as well as in the process of vascular endothelial permeability.

Fig. 5.

G-Rb1 regulates Rho-GTPase activity in LPS-stimulated HUVECs. (A) Heat map displaying an overview of the transcriptome profiles of HUVECs treated with LPS (10 μg/ml) or LPS (10 μg/ml) + G-Rb1 (50 μmol/L) for 24 h (n = 3). (B) Volcano plot showing differentially expressed genes (DEGs) between the LPS group and LPS + G-Rb1 group, filtered by |Fold Change| ≥ 1.5 and p-value < 0.05. (C) Reactome pathway enrichment analysis of DEGs between LPS and LPS + G-Rb1 groups (adjusted p-value < 0.05). The 10 significantly enriched pathways/gene sets are displayed. (D) Gene Ontology (GO) enrichment analysis of DEGs (adjusted p-value < 0.05). (E) Western blot analysis of RhoA expression in HUVECs after respective treatments for 24 h. (F) Quantitative analysis of RhoA protein levels normalized to GAPDH (n = 5). (G) Western blot analysis of RhoA expression in lung tissues of mice at 24 h post-CLP. (H) Quantitative analysis of RhoA protein levels in lung tissues normalized to GAPDH (n = 5). (I) Quantitative analysis of RhoA reactivity in HUVECs measured by G-LISA assay after respective treatments for 24 h (n = 6). Error bars represent mean ± s.e.m.

To verify whether G-Rb1 exerts its therapeutic effects by regulating endothelial RhoA activity, we performed Western Blot (WB) and G-LISA assays. The results showed that LPS stimulation significantly increased RhoA protein expression in HUVECs compared to the Control group, while G-Rb1 treatment markedly suppressed this LPS-induced increase (Fig. 5E and F). Similarly, in animal tissues, RhoA expression in the lung tissue of the CLP group was significantly elevated compared to the sham group, and G-Rb1 treatment significantly inhibited this effect (Fig. 5G and H). The G-LISA results showed that LPS significantly enhanced RhoA activity in HUVECs, whereas co-treatment with G-Rb1 inhibited this increase in activity (Fig. 5 I).

These results indicate that in both in vivo and in vitro experiments, G-Rb1 exerts anti-inflammatory, anti-oxidative stress, and vascular barrier-protective effects by inhibiting the expression and signal transduction of RhoA.

3.5. G-Rb1 protects against sepsis-induced endothelial injury by preserving mitochondrial quality

To further investigate whether G-Rb1 exerts its therapeutic effects by protecting mitochondrial function through the regulation of the RhoA-related signaling pathway, we conducted a systematic evaluation of mitochondrial function and quality in human umbilical vein endothelial cells (HUVECs) and detected changes in the expression of Drp1, a protein related to mitochondrial homeostasis. The results showed that LPS stimulation caused a significant increase in mitochondrial ROS production in HUVECs (Fig. 6A and B). Dysfunctional mitochondria have been reported to be a major source of ROS, which in turn triggers oxidative stress and forms a positive feedback loop with inflammatory factors, exacerbating cellular damage. In contrast, G-Rb1 treatment significantly suppressed the LPS-induced increase in mitochondrial ROS, an effect comparable to that of Fasudil and the Drp1 inhibitor Mdivi-1.

Fig. 6.

G-Rb1 ameliorates sepsis-induced mitochondrial homeostasis disruption. (A) Representative fluorescence images showing mitochondrial ROS levels (red signal) in HUVECs after respective treatments for 24 h. Nuclei were stained with DAPI (blue signal), and mitochondrial locations were labeled with MitoTracker (green signal). (B) Quantitative analysis of mitochondrial ROS fluorescence intensity (n = 120-150 cells). (C) Mitochondrial membrane potential assessed by JC-1 staining after 24-h treatments. (D) Quantitative analysis of the JC-1 red/green fluorescence intensity ratio (n = 120-150 cells). (E-I) Mitochondrial oxidative phosphorylation (OXPHOS) parameters in HUVECs measured using the Seahorse XF Analyzer after respective treatments. (J) Western blot analysis of Drp1 and p-Ser616 Drp1 expression in HUVECs after 24-h treatments. (L, M) Quantitative analysis of Drp1 and p-Ser616 Drp1 protein levels normalized to GAPDH (n = 5). (K) Western blot analysis of Drp1 and p-Ser616 Drp1 expression in lung tissues of mice at 24 h post-CLP. (N, O) Quantitative analysis of Drp1 and p-Ser616 Drp1 protein levels in lung tissues normalized to GAPDH (n = 5). Error bars represent mean ± s.e.m. Scale bars: 50 μm (A); 100 μm (C). (P)Representative images of mitochondrial morphology stained with MitoTracker in HUVECs after respective treatments for 24 h. (Q) Quantitative analysis of mitochondrial aspect ratio to evaluate mitochondrial fission (n = 8). (R) Mitochondrial calcium levels detected by Rhod-2 fluorescence probe in treated HUVECs. (S) Quantitative analysis of Rhod-2 fluorescence intensity (n = 120-150 cells). (T) Fluorescence visualization of F-actin using phalloidin staining in differently treated HUVECs. (U) Transmission electron micrographs of mesenteric vessels from CLP mice at 24 h post-surgery, showing endothelial cell tight junctions and mitochondrial morphology (E, endothelial cell; L, vascular lumen; white triangles, tight junctions; white arrows, mitochondria). Error bars represent mean ± s.e.m. Scale bars: 50 μm and 12.5 μm (P); 50 μm (R, T); 1 μm and 500 nm (U).

Additionally, LPS stimulation led to a decrease in the mitochondrial membrane potential of HUVECs, which was manifested as a reduction in J-aggregates and an increase in JC-1 monomers from the JC-1 probe (Fig. 6C and D). G-Rb1 treatment, however, reversed this change, leading to a significant restoration of the membrane potential, an effect comparable to that of Fasudil and Mdivi-1. Measurement of bioenergetic parameters related to oxidative phosphorylation (Fig. 6 E) revealed that LPS stimulation significantly reduced the cells’ basal respiration (Fig. 6 A), spare respiratory capacity (Fig. 6 G), ATP production (Fig. 6 H), and maximal respiration (Fig. 6 I), indicating impaired mitochondrial function under stress conditions. G-Rb1 treatment effectively maintained these bioenergetic parameters, keeping them significantly higher than in the LPS group.

Western blot results showed that while total Drp1 protein expression did not differ significantly among the groups (Fig. 6J–L), LPS stimulation significantly enhanced the phosphorylation of Drp1 at the Ser616 site (p-Ser616 Drp1) in HUVECs (Fig. 6J–M), an effect effectively inhibited by G-Rb1. Similarly, we observed in the lung tissues of mice that although Drp1 expression was unchanged (Fig. 6K–N), G-Rb1 significantly suppressed the CLP-induced increase in p-Ser616 Drp1 (Fig. 6K–O).

Furthermore, MitoTracker Green staining was utilized to evaluate mitochondrial mass and morphology. By calculating the mitochondrial length-to-width ratio, we quantified the level of mitochondrial fragmentation. The results revealed that LPS-induced mitochondrial fragmentation was significantly alleviated in both the G-Rb1- and Mdivi-1-treated groups (Fig. 6P and Q). Rhod-2 staining showed that LPS induced significant mitochondrial calcium accumulation, which was markedly reduced by both G-Rb1 and Mdivi-1 (Fig. 6R and S). It is known that mitochondrial fission in endothelial cells depends on the actin cytoskeleton, whose aggregation is mediated by RhoA/ROCK activation. We found that F-actin aggregation increased in HUVECs after LPS treatment, a process reversed by both G-Rb1 and Fasudil (Fig. 6 T). As previous experiments demonstrated that G-Rb1 significantly inhibits LPS-induced enhancement of RhoA activity in HUVECs (Fig. 5 I), these findings suggest that G-Rb1 may suppress LPS-induced excessive mitochondrial fission through the RhoA/ROCK pathway.

Observation via transmission electron microscopy further revealed that in CLP mice, the gaps between mesenteric vascular endothelial cells at their tight junctions were enlarged and separated. Concurrently, there was a marked mitochondrial homeostatic imbalance within these cells, manifested as extensive vacuolization and destruction of the cristae structure (Fig. 6 U). Intervention with G-Rb1 significantly maintained the normal morphology of both the endothelial cells themselves and their intercellular junctions, improved mitochondrial cristae integrity, and reduced the extent of vacuolization, confirming that G-Rb1 improves mitochondrial quality and the endothelial barrier at the tissue and organ level.

4. Discussion

Sepsis, a critical condition with high morbidity and mortality, often leads to multi-organ dysfunction, with acute lung injury (ALI) as a common complication [5]. Its pathophysiology involves complex processes where microcirculatory dysfunction, characterized by vascular barrier injury, is central to sepsis-associated ALI. In sepsis, activated endothelial cells amplify inflammation and disrupt the alveolar-capillary barrier, causing alveolar leakage and driving ALI [[23], [24], [25]].

The role of RhoA signaling in the inflammatory progression, immune dysregulation, and vascular barrier injury in sepsis has been widely confirmed [26,27]. In the CLP model, inhibiting RhoA activation effectively alleviates the inflammatory response and protects vascular barrier function. Inhibition of the RhoA/ROCK pathway reduces MPO activity and the release of inflammatory mediators like IL-1β, IL-6, tissue factor, and PAI-1 [[28], [29], [30]], and antagonizes the systemic inflammatory response by inhibiting NF-κB [31]. Concurrently, this inhibition modulates STAT phosphorylation, suppressing Th1 over-activation and restoring the Treg/Th17 balance, thereby alleviating MODS and improving immune homeostasis [32]. Rho kinase activation is also closely associated with enhanced oxidative stress, marked by decreased GSH and elevated lipid peroxidase, nitrite, and ROS levels [33]. Furthermore, RhoA/ROCK signaling disrupts endothelial junction structure by inhibiting VE-cadherin, Claudin-5, and ZO-1 expression and localization [34], and promotes actin cytoskeletal reorganization and membrane contraction, increasing intercellular gaps and exacerbating junctional dissociation [35].

G-Rb1 has demonstrated ameliorating effects on sepsis at multiple levels [14]. Notably, Rg3-enriched red ginseng extract, in which G-Rb1 and Rg3 are major active components, has also been shown to mitigate LPS-induced sepsis by inhibiting platelet-leukocyte aggregates and regulating thrombo-inflammatory pathways, further supporting the broad therapeutic potential of ginseng-derived compounds in this condition [36]. Previous studies indicate G-Rb1 can inhibit sepsis-induced ferroptosis [20], suppress inflammatory pathways by downregulating TLR4 and TNF-α [37], and improve microvascular hyperpermeability by inhibiting caveolin-1 (Cav-1) in endothelial cells [16].

This study further confirms that G-Rb1 effectively reduces pulmonary inflammatory infiltration, iNOS expression, and pro-inflammatory factors in BALF of septic mice. G-Rb1 treatment not only alleviated sepsis-induced immunosuppression by restoring the proportions of CD3⁺ and CD4⁺ cells, but also decreased pro-inflammatory T cell subsets (Th1, Th17) and rebalanced the Th1/Th2 and Th17/Treg ratios. In vivo, G-Rb1 maintained the expression of VE-cadherin, Claudin-5, and ZO-1, reducing pulmonary edema and capillary permeability. Laser speckle imaging showed G-Rb1 restored mesenteric microcirculatory blood flow and tissue perfusion. These protective effects, comparable to the RhoA/ROCK inhibitor Fasudil, are linked to the regulation of Rho/ROCK signaling.

To investigate the cellular mechanism, we used an LPS-induced HUVEC model. G-Rb1 reduced cytotoxicity, maintained VE-cadherin, Claudin-5, and ZO-1 levels, and reversed LPS-induced oxidative stress, an effect analogous to Fasudil. Transcriptomic analysis identified Rho GTPases as a potential pathway for G-Rb1's intervention. Based on evidence that G-Rb1 can directly bind RhoA [38], we hypothesized RhoA is the core target. We consistently demonstrated that G-Rb1 significantly reduced RhoA expression and activation in both lung tissues of CLP mice and LPS-induced HUVECs.

Sepsis-induced endothelial dysfunction is closely associated with mitochondrial dysfunction, particularly excessive fission. We hypothesized G-Rb1's effects were achieved by inhibiting this process via the RhoA pathway. Experiments confirmed G-Rb1 reduced mitochondrial ROS, restored membrane potential, and mitigated fragmentation in LPS-stimulated HUVECs, with effects comparable to Fasudil and the Drp1 inhibitor Mdivi-1. Drp1 is a key regulator of fission; activated RhoA promotes Drp1 phosphorylation at serine 616 (Ser616) via ROCK [39]. p-Ser616 Drp1 translocates to the outer mitochondrial membrane to mediate division [40]. This process is dependent on actin cytoskeleton reorganization, which is also regulated by RhoA/ROCK through myosin light chain (MLC-2) phosphorylation [[41], [42], [43]]. F-actin polymerization is essential for Drp1 recruitment, while cytoskeletal reorganization simultaneously disrupts endothelial junctions [44,45]. Our study showed that while total Drp1 was unchanged, its phosphorylation at Ser616 was markedly increased in sepsis, and G-Rb1 treatment effectively inhibited this at both tissue and cellular levels, improving mitochondrial morphology and function. Electron microscopy confirmed G-Rb1's protective effect on mitochondrial ultrastructure and endothelial junctions.

Taken together, our findings suggest that G-Rb1 may exert its protective effects against sepsis at least partly through the modulation of the RhoA/ROCK pathway, subsequently attenuating Drp1-mediated excessive mitochondrial fission. This proposed mechanistic link is supported by the parallel improvements we observed in inflammation, redox balance, and vascular barrier integrity, which collectively contributed to reduced ALI and improved survival in septic mice. While further mechanistic validation is warranted, these results highlight the potential of G-Rb1 as a promising therapeutic candidate for sepsis-induced microcirculatory dysfunction.

Although this study offers novel mechanistic insights, limitations remain. The CLP and LPS models cannot fully recapitulate human sepsis, and the focus on pulmonary microcirculation means organ-specific effects of G-Rb1 require further investigation. The precise interaction between G-Rb1 and RhoA, and its influence on other mitochondrial dynamics proteins beyond Drp1, also need clarification [46,47]. Furthermore, it is important to note that the current dosing regimen-administered G-Rb1 and Fasudil for five consecutive days prior to CLP surgery-represents a prophylactic rather than a therapeutic intervention. While this preventive design effectively elucidates the pharmacological potential of G-Rb1 in averting subsequent endothelial and mitochondrial damage, it differs from standard clinical practice, where treatment is initiated after the onset of sepsis. Therefore, evaluating the efficacy of G-Rb1 when administered post-CLP (as a true therapeutic rescue) will be a crucial next step to further validate its potential for clinical application. Despite this, our study systematically reveals the molecular mechanism by which G-Rb1 protects mitochondrial function and the endothelial barrier by modulating the RhoA/ROCK/Drp1 pathway. This provides a new theoretical basis for G-Rb1 as a potential therapeutic agent for sepsis-induced microcirculatory dysfunction.

5. Conclusion

Microcirculatory dysfunction is a key factor in organ damage and subsequent organ failure during sepsis. This study provides new insights into the application of G-Rb1 in sepsis treatment. The rational use of G-Rb1 during sepsis can alleviate inflammatory damage via RhoA, preserve vascular barrier integrity, and improve mitochondrial quality in vascular endothelial cells, ultimately reducing vascular leakage and disordered blood flow distribution. This finding highlights the potential of G-Rb1, a monomer from traditional Chinese medicine, as a prophylactic and therapeutic agent for sepsis.

CRediT authorship contribution statement

Shixi Shang∗: Conceptualization, Methodology, Formal analysis, Writing - original draft, Writing - review & editing. Jingju Wang∗: Data curation, Investigation, Methodology. Ying Huang: Project administration, Software. Lihua Sun: Investigation, Methodology. Junsong Yin: Formal analysis, Validation. Jiayao Shi: Methodology, Visualization. Guangzhao Cao: Project administration, Formal analysis. Hongjun Yang#: Writing - review & editing, Funding acquisition, Supervision, Conceptualization. Xiaolu Shi#: Writing - review & editing, Funding acquisition, Supervision, Conceptualization.

Funding

Scientific and Technological Innovation Project of China Academy of Chinese Medical Sciences (NLTS2025009 CI2021A05014); National Natural Science Foundation of China, 82004244.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: