The combination of EGCG with warfarin reduces deep vein thrombosis in rabbits through modulating HIF-1α and VEGF via the PI3K/AKT and ERK1/2 signaling pathways

Deep vein thrombosis (DVT) is a common occurrence that can lead to debilitating conditions such as pulmonary embolism and is the cause of complications in hospitalized patients that can result in death ^[1-3]. Endothelial cells play a role in the development of DVT by participating in thrombus formation in response to injury ^{[4, 5]}. Endothelial cells normally prevent platelet and thrombin activation by expressing and producing mediators such as heparan sulfate, thrombomodulin, and nitric oxide (NO), which inhibit platelet adhesion and thrombin-induced coagulation ^[6-8]. Endothelial cells help reduced blood loss and clot formation by producing endothelin-1 (ET-1), angiotensin II (Ang II), and adhesion glycoproteins, which enhance vasoconstriction and platelet and collagen fibrils aggreatioon in response to injury ^{[9, 10]}. Ang II and ET-1 induce the production of 20-HETE, which activates the ERK pathway, causing endothelial activation ^[11].

Coagulation is cuased by the activation of pathways, such as ERK1/2, which are activated by damaged blood vessels or exposure to collagen ^[12]. These pathways trigger the generation of thrombin, which converts fibrinogen to fibrin and forms the thrombus scaffold ^[13]. Endothelial cells are also stimulated by thrombin to release cell surface adhesion molecules including P-selectin and von Willebrand factor (vWF), which attract platelets and leukocytes to the injury site ^[14]. In the absence of injury, endothelial cell dysfunction can cause fibrin production and platelet aggregation by causing inflammation and contributing to the endothelial cell damage associated with DVT ^[15-17]. Coagulation factor XII (FXII) was discovered to increase DVT in a rat model by activating a PI3K/AKT signaling-induced inflammatory response ^[18].

Polyphenols from tea (Camellia sinensis) are plant-based antioxidants with a wide range of health benefits ^{[19, 20]}. Tea polyphenols (TPs), also known as catechins, make up 30%−42% of the dry weight of the tea leaves ^[19]. Their antioxidant effects are enhanced by their vicinal dihydroxy or trihydroxy structure, which allows for electron delocalization and chelation of metal ions, resulting in free radicals quenching ^[21]. TPs have been found to attenuate oxidative stress by regulating the Keap1/Nrf2/ARE pathway ^[22]. The reagent utilized in this investigation was epigallocatechin-3-gallate (EGCG), which is one of the key polyphenols known to have anti-inflammatory and antioxidant properties in C. sinensis.

In compared to venous blood, newly formed thrombus are hypoxic, which causes the expression of hypoxia-inducible factor 1 (HIF-1) ^[23]. HIF-1 is one of two subunits that make up HIF1 and is increased during hypoxia. HIF-1 translocates to the nucleus under hypoxia, where it combines with the other HIF-1 subunit to produce activated HIF-1 ^[24]. HIF-1 may enhance blood clots dissolve faster by interacting with hypoxia-responsive elements, which upregulate the transcription of angiogenesis-related genes such as vascular endothelial growth factor (VEGF) ^{[25, 26]}. VEGF is elevated during DVT and has been found in endothelial cells in thrombus resolution and newly formed vascular tissue ^[27].

By ligating the inferior vena cava (IVC) and left common iliac vein, we were able to test whether the combination of EGCG and warfarin could reduce blood clot formation. In human umbilical vein endothelial cells (HUVECs) treated with hydrogen peroxide (H₂O₂), we measured and observed hemorheological parameters as well as the expression of HIF-1α, VEGF, AGTR-1, and components of the PI3K/AKT and ERK1/2 signaling pathways. Our results demonstrate that combining EGCG with warfarin could be a beneficial complementary medication in the treatment of DVT.

ECGC was purchased from Hangzhou Yibeijia Tea Technology Co., Ltd. (S18152, Zhejiang, China). Warfarin was purchased from Qilu Pharmaceutical Co., Ltd. (H37021314, Jinan, China). HIF-1α (NB100-105, 1 : 1000, Novus, USA) and AGTR-1 (NBP1-77078, 1 : 1000, Novus, USA) were purchased from Novus (Colorado, USA). p-ERK (bs-3016R, 1 : 1500), ERK (bs-0022R, 1 : 1000), VEGF (bs-1313R, 1 : 1500), VEGFR-1 (bs-20692R, 1 : 1500), Akt (bs-0115M, 1 : 1000), p-PI3K (bs-3332R, 1 : 1500), p-P70S6K (bs-1656R, 1 : 1000), P70S6K (bs-3617R, 1 : 1000), p-Akt (bs-0876R, 1 : 1000) and PI3K (bs-2067R, 1 : 2000) were purchased from Bioss antibodies (Beijing, China). GAPDH (ab9485, 1 : 1000) antibodies were purchased from Abcam (Cambridge, MA, USA). Primary HUVECs were purchased from ThermoFisher Scientific (C0035C, Waltham, MA, USA). Dulbecco’s modified Eagle’s medium (DMEM) was purchased from GE Healthcare Life Sciences (SH30022, Logan, UT, USA).

Male New Zealand white rabbits weighting an average of 2.5 kg were obtained from the Medical Animal Experimental Center at Nanjing Medical University . The rabbits were caged at a controlled temperature of 24 °C with free access to food and water. The animals were maintained following the ethical guidelines recommended by our institute (Ethical code: IACUC-20210207, time: 2021-07-02) and all experiments were conducted in accordance with the internationally accepted principles for laboratory animal use and care as found in the US guidelines (NIH publication #85-23, revised in 1985).

The rabbits were divided randomly into five experimental groups: control (n = 10), model group (n = 10), model with warfarin (2.00 mg·kg⁻¹·d⁻¹, n = 10), and model with warfarin and EGCG at a low-dose (200 mg·kg⁻¹·d⁻¹, n = 10), middle-dose (600 mg·kg⁻¹·d⁻¹, n = 10), and high-dose (1000 mg·kg⁻¹·d⁻¹, n = 10). To create the DVT model, animals were first anesthetized with 20% urethane (5 mL·kg⁻¹). DVT was induced by the ligation of the IVC and left common iliac vein with 6-0 silk thread via access through a midline laparotomy. Postoperatively, animals in each group were given the appropriate dose of EGCG and/or warfarin (2.00 mg·kg⁻¹·d⁻¹) or vehicle orally after surgery for 7 days.

Whole blood (5 mL) was obtained from the ear artery of rabbits 1, 3, and 7 days after the DVT procedure. Plasma viscosity, whole blood viscosity at high and low shear, and red cell assembly, and other hemorheology indexes were measured using a hemorheology detector (Beijing Succeeder Science and Technology Development, Beijing, China). Coagulation function indexes, such as fibrinogen content, prothrombin time (PT), thrombin time (TT), and activated partial thromboplastin time (APTT), were determined from blood samples using a coagulation convention detector (Nanjing Perlong Image Documentation Equipment, Nanjing, China).

HUVECs were cultured in DMEM with 10% fetal bovine serum and 100 U·mL⁻¹ penicillin/streptomycin (15140122, Gibco, Gaithersburg, MD, USA) in a humidified atmosphere at 37 °C with 5% CO₂. After 24 h, cells (1 × 10⁴ cells/mL) were resuspended in DMEM with 0, 5, 10, and 20 μmol·L⁻¹ EGCG. HUVECs were incubated for 2 h and then subjected to 200 µmol·L⁻¹ H₂O₂ for 24 h. Cells were harvested at 150 × g for 5 min for analysis.

To assess the extent of injury, rabbit femoral veins were excised and stained with H&E. The tissue was first fixed with 4% paraformaldehyde befroe being dehydrated in a series of ethanol and finally embedded in paraffin. Fixed tissue was cut into sections and dewaxed and rehydrated. Dewaxed tissue was stained with H&E and images were captured at × 400 magnification using a light microscope (IX-81, Olympus, Tokyo, Japan).

Sections (4 µmol·L⁻¹) of fixed femoral vein tissue were first dewaxed and rehydrated. A TUNEL Apoptosis Detection kit (40306ES20, Yeasen Biotechnology, Shanghai, China) was used according to the manufacturer’s instructions. The tissue was counterstained with DAPI (C1002, Beyotime, Nanjing, China).

Sections of femoral vein tissue were incubated with antibodies toward CD34 (1 : 100, bs-0646R, Bioss, Beijing, China). They were then incubated at room temperature for 1 h in HRP-conjugated with secondary antibodies (1 : 50). After washing, sections were counterstained with DAPI and visualized with a Dako REAL EnVision Detection System (K5007, Agilent, Santa Clara, CA, USA).

The viability of HUVECs was measured with an MTT assay. After treatment, the cells were washed and incubated with MTT solution (5 mg·mL⁻¹, M5655, Sigma-Aldrich, St Louis, MO, USA) at 37 °C for 4 h in a microplate. Dimethyl sulfoxide (67-68-5, 100 µL, Carl Roth GmbH, Karlsruhe, Germany) was added to each well and left to stand for 4 h at 37 °C. Optical density was determined at 570 nm in a microplate reader (Model 680, Bio-Rad, Hercules, CA, USA).

A CCK-8 assay was used to determine the number of viable HUVECs. HUVECs (2 × 10³ cells/well) were seeded into 96-well plates with 10 μL CCK-8 solution (M4839, ApexBio, Shanghai, China) and incubated for 1 h. A CytoFLEX flow cytometer (Beckman Coulter, Miami, FL, USA) was used to measure cell apoptosis. ROS content was assessed by staining cells with 2′, 7′-dichlorofluorescein (DCF) and then the percentage of stained cells was determined.

Slides were recovered with EDTA solution (pH 9.0), immersed in 3% H₂O₂, blocked with 10% goat serum, and incubated with antibody to VEGF and HIF-1α (1 : 200) at 4 °C overnight. After that, the sections were incubated for 1 h at room temperature with an HRP-conjugated secondary antibody (KGAA35, 1: 125, KeyGEN BioTECH, Nanjing, China) before being colored with diaminobenzidine for 40 s. As a negative control, normal goat IgG was employed. Under an optical microscope (BX53; Olympus, Tokyo, Japan), pathological alterations were seen.

In Table 1, the sequences of HIF-1 siRNAs are presented. We utilized a random siRNA sequence (5'-GCGCGCUUUGUAGGAUUCG dTdT-3') as a control for each siRNA. The Lipofectamine^TM 3000 reagent (Invitrogen, Grand Island, NY, USA) was used to transfect cells with each siRNA according to the manufacturer’s instructions.

RNA was extracted with Trizol reagent (15596018, Invitrogen, Carlsbad, CA, USA). An RT reagent kit (RR037, Takara, Shiga, Japan) was used to synthesize cDNA. Genes were amplified with SYBR Green Supermix (1725121, Bio-Rad, Hercules, USA). The PCR conditions were 95 °C for 5 min then 95 °C for 3 s and 60 °C for 20 s for 40 cycles using a LightCycler480 II (Roche, Basel, Switzerland). Relative gene expression was determined against GAPDH using the 2^−ΔΔCt method. The primer sequences are listed in Table 1.

To perform western blotting cells were first lysed with RIPA buffer, and protein levels were measured using a protein quantification kit (P0012S, Beyotime, Shanghai, China). SDS-PAGE was used to separate the proteins, which were then transferred to PVDF membranes (1620177, Bio-Rad, Hercules, CA). PVDF membranes were blocked in non-fat milk and then incubated with antibodies against HIF-1α (1 : 1000), VEGF (1 : 1500), VRGFR-1 (1 : 500), AGTR-1 (1 : 1000), p-PI3K (1 : 1500), PI3K (1 : 2000), p-Akt (1 : 1000), Akt (1 : 1000), p-ERK (1 : 1500), ERK (1 : 1000), p-P70S6K (1 : 1000), P70S6K (1 : 1000) and GAPDH (1 : 1000) overnight at 4 °C. The PVDF membranes were probed with secondary antibody for 1 h at room temperature and the immunoreactive bands were detected with an ECL luminescence kit (PE0010, Solarbio, Beijing, China). Images were processed with ImageJ2x software (v2.1.5.0, Rawak Software, Stuttgart, Germany).

Statistical analysis

Data are presented as the mean ± standard error of the mean (SEM). Differences between groups were compared with one-way ANOVA and the Tukey post-hoc test. Student’s t-test was used to compare the differences between two groups. P < 0.05 was considered significant. All statistical analyses were conducted using SPSS 20.0 (SPSS, Chicago, IL USA) and GraphPad Prism v8.0 (GraphPad Software, San Diego, CA, USA) software.

To determine whether combination of EGCG with warfarin could ameliorate the formation of blood clots, we induced a rabbit model of DVT by ligation of the IVC and left common iliac vein. The numerical data of hemorheological parameters and coagulation function measured 7 days after the ligation model is shown in Table 3. According to the data, there is a significant difference between the model group and the control group in terms of PT, TT, APTT, and fibrinogen levels, indicating a rise in the blood’s clotting properties. PT, TT, and APTT values are significantly higher and improve as the dose of EGCG is increased in animals that had received EGCG for 7 days with warfarin, and fibrinogen levels are lower than in control rabbits or those treated with an warfarin alone. Similarly, when EGCG is taken in combination with warfarin, thrombus formation, measured by length and weight, (Figs. 1A−1C) and vein wall thickness (Fig. 1D) are ameliorated in the DVT model and the degree of amelioration is dose-dependent. Overall, the results we have obtained using the DVT model, support the use of EGCG as a useful adjunct to warfarin.

Figure 1.  EGCG combined with warfarin improves thrombosis in vivo. A, Images of thrombus from different groups. B and C, Thrombus length and weight in each group. D, H&E-stained sections of vein wall thickness indicated by arrows, Scale bar, 20 μm. E, Vein wall thickness in different treatment groups. Data are expressed as mean ± SEM (n = 10). ^aP < 0.05 model groups vs control group; ^bP < 0.05 model + warfarin group vs model group; ^cP < 0.05 model + warfarin + EGCG group vs model + warfarin group

The beneficial effects of the EGCG-warfarin combination were confirmed by measuring blood parameters related to coagulation and thrombus formation over 7-day period. D-dimer, a fibrin degradation product, was found to be lowest in the control animals and highest in the DVT model without treatment (Fig. 2A). Levels are significantly reduced when warfarin is administered, but they are further reduced when EGCG is added, and the improvement is dose-dependent. P-selectin, a cell adhesion molecule found on the surface of endothelial cells in blood vessels and activated platelets, exhibits a similar pattern (Fig. 2B). In contrast, NO levels in the control are significantly higher than in the DVT model (Fig. 2C). NO levels rise in animals given warfarin, and they rise even more in those given warfarin and EGCG. NO regulates the levels of 20-HETE, which accelerates thrombosis, therefore, 20-HETE levels in the control are lower than in the DVT model (Fig. 2D). The combination of warfarin and EGCG significantly reduces 20-HETE levels in a dose-dependent manner. Similarly, in the DVT model, levels of vWF, which stimulates the adhesion of platelets to endothelial cells, increases in the DVT model thereby exacerbating the formation of thrombi (Fig. 2E). The combination of warfarin with the highest dose of EGCG had the largest influence on vWF levels, with the greatest effect observed on the first day of adm inistration.

Figure 2.  In vivo effects of EGCG and warfarin combinations on DVT-associated blood parameters. D-dimer (A), P-selectin (B), NO (C), 20-HETE (D), and vWF (E) Data are expressed as mean ± SEM (n = 8). ^aP < 0.05 model group vs control group, ^bP < 0.05 model + warfarin group vs model group, ^cP < 0.05 model + warfarin group + EGCG vs model + warfarin group

Hypoxia is associated with the thrombus environment, therefore, we determined levels of HIF-1α, VEGF, and components of the PI3K/AKT and ERK1/2 signaling pathways in the DVT model. Immunofluorescence analysis indicated that levels of CD34, a cell surface glycoprotein that functions as a cell-cell adhesion factor in the endothelial cells of blood vessels, were higher in the control than in the DVT model. CD34 expression was considerably reduced in the model group, but dramatically increased after treatment with warfarin alone compared to the model group. The use of different doses of EGCG and warfarin in combination increased CD34 expression, with the degree of elevation corresponding to EGCG concentration (Fig. 3A). In the DVT model, immunohistochemistry revealed a higher level of HIF-1 and VEGF in the periphery of blood arteries and thrombus than in the tissue of animals treated with EGCG (Figs. 3B−3F). The most intense staining was seen in the DVT model, however as the amount of EGCG was increased, the intensity of the staining decreased. HIF-1α, VEGF, VEGFR1 and ATGR-1 protein levels were confirmed by western blotting, and mRNA expression matched these values (Figs. 3D−3F). The DVT model had the highest levels of HIF-1, ATGR-1, VEGF, and VEGFR-1, while the control and animals that received warfarin and the highest dose of EGCG had the lowest. We also measured PI3K, AKT, and ERK phosphorylation levels in relation to EGCG dose (Fig. 3D). The phosphorylation of PI3K, AKT, and ERK is elevated in the DVT model and is reduced dose-dependently by the addition of EGCG. These results suggest that combining EGCG with warfarin reduces the expression of HIF-1α and VEGF through a mechanism that could involve the PI3K/AKT and ERK1/2 signaling pathways.

Figure 3.  Combined use of warfarin and EGCG regulate HIF-1 and VEGF protein levels in vivo. A, Immunofluorescence analysis of CD34 in different groups, bar, 20 μm. B and C, immunohistochemical staining analysis: representative immunohistochemical staining of (B) HIF-1α and (C) VEGF in different groups, bar, 20 μm. D, Expressions of HIF-1α, VEGF, VEGFR-1, AGTR-1, PI3K/AKT, and ERK1/2 proteins determined by western blotting, GAPDH was used as the loading control. E and F, mRNA expression of AGTR-1 and VEGF in different groups was investigated by reverse transcription-quantitative polymerase chain reaction. Data are presented as mean ± SEM (n = 5); ^aP < 0.05 model group vs control group; ^bP < 0.05 model + warfarin group vs model group; ^cP < 0.05 model + warfarin group + EGCG vs model + warfarin group

To verify the protective effect of EGCG combined with warfarin on endothelial cell injury we measured the influence of combination of EGCG with warfarin on cell proliferation in HUVECs treated with 0, 5, 10, and 20 µmol·L⁻¹ EGCG for 2 h followed by 200 µmol·L⁻¹ H₂O₂ for 24 h. Cell proliferation and apoptosis in HUVECs are shown in Figs. 4A−4D, using the CCK8 assay and flow cytometry, respectively. Cell proliferation was reduced the most in cells exposed to H₂O₂ in the absence of warfarin (Fig. 4A). The addition of warfarin improved cell proliferation, which was enhanced by EGCG treatment in a dose-dependently manner. Similar results were obtained in terms of apoptosis (Figs. 4B and4C). Lower numbers of cells survived H₂O₂ treatment without warfarin or EGCG. Flow cytometry and DCF staining confirmed that EGCG in combination with warfarin protects endothelial cells from oxidative stress injury in vitro.

Figure 4.  Endothelial cell damage is prevented in vitro by combining EGCG with warfarin. A, Cell proliferation ability of HUVECs was detected by CCK8 assay. B and C, Effects of EGCG on cell apoptosis by flow cytometry in three independent experiments. D, Determination of apoptosis in HUVECs stained with DCF. E, Statistical results of flow cytometry. Data are presented as mean ± SEM (n = 3), ^aP < 0.05 model groups vs control group, ^bP < 0.05 model + warfarin group vs model group, ^cP < 0.05 model + warfarin group + EGCG vs model + warfarin group

To verify that combination of EGCG with warfarin could influence the expression of HIF-1α and VEGF in vitro, we assessed levels of HIF-1α in HUVECs after treatment using immunofluorescence (Fig. 5A). To substantiate the results we obtained in vivo, we found that the expression of HIF-1α was increased in response to oxidative stress in vitro and EGCG could lower expression when applied in combination with warfarin. Protein levels and phosphorylation were also similar to the results in vivo (Fig. 5B). HIF-1α, VEGF, VEGFR-1 and ATGR-1 were all upregulated in response to oxidative stress. However, the addition warfarin with EGCG could protect cells from the damage caused by oxidative stress. In addition, the phosphorylation of PI3K, AKT, P70S6K, and ERK1/2 were all increased under oxidative stress but lower levels were observed in cells pretreated with warfarin and EGCG. Relative HIF-1α and ATGR-1 mRNA expression was also elevated under oxidative stress but not to the same level in cells pretreated with warfarin and EGCG. Overall, our results suggest that EGCG in combination with warfarin could protect cells from the consequences of oxidative stress and could lower the expression of HIF-1α and VEGF through a mechanism that may involve PI3K/AKT and ERK1/2 signaling pathways.

Figure 5.  EGCG combined with warfarin reduces the expression of HIF-1α and VEGF in vitro. A, Immunofluorescence analysis of HIF-1α in different groups, bar, 20 μm. B, Expressions of HIF-1α, VEGF, VEGFR-1, AGTR-1, PI3K/AKT, and ERK1/2 protein levels in vitro determined by western blotting, GAPDH was used as the loading control. C, mRNA expression of AGTR-1 and VEGF in the different groups was investigated by reverse transcription-quantitative polymerase chain reaction. Data are presented as mean ± SEM (n = 5); ^aP < 0.05 model groups vs control group; ^bP < 0.05 model + warfarin group vs model group; ^cP < 0.05 model + warfarin group + EGCG vs model + warfarin group

Next, we determined whether the inhibition of HIF-1α would influence the effects of EGCG combined with warfarin on HUVECs. HIF-1α was inhibited in HUVECs by the use of small interfering RNA (Fig. 6A). The levels of proliferation and apoptosis were then measured in HUVECs (Figs. 6B−6D). We found that the inhibition of HIF-1α in combination with warfarin and EGCG (20 μmol·L−1) slightly increased proliferation and reduced apoptosis in response to oxidative stress compared to the model but was significantly less effective than the EGCG-treated control with siRNA (Figs. 6B−6D). Cell migration and invasion were also improved when cells were treated with both warfarin and EGCG but the inhibition of HIF-1α impeded this improvement (Figs. 6E−6G). Similar results were obtained for cell viability, which was measured by flow cytometry (Fig. 6H). DCF staining indicated that the levels of ROS were higher in the model (Fig. 6I). The oxidative properties of EGCG and warfarin lowered the levels of ROS. Silencing HIF-1α appears to increase the levels of ROS and illustrates that HIF-1α may be overexpressed in the hypoxic thrombus environment. These results indicate that EGCG may be involved in the modulation of HIF-1α to protect HUVECs from oxidative stress.

Figure 6.  Inhibition of HIF-1α impedes the beneficial effect of EGCG on HUVECs. A, Expression of HIF-1α was determined by RT-PCR. B, Cell proliferation ability of HUVEC was detected by CCK8 assay. C and D, Cell apoptosis was assessed by flow cytometry. E and F, the migration and invasion of endothelial progenitor cells was confirmed by Transwell assay (200 ×). G, Histogram of cell migration and invasion ability. H, Statistical results of flow cytometry. I, Percentage of HUVECs stained with DCF. Data were presented as mean ± SEM (n = 3); ^aP < 0.05 model group vs control group; ^bP < 0.05 model + warfarin group vs model group; ^cP < 0.05 model + warfarin group + EGCG 20 μmol·L⁻¹ vs model + warfarin group; ^dP < 0.05 model + warfarin + EGCG 20 mmol·L⁻¹ + si-HIF-1α group vs model + warfarin + EGCG 20 μmol·L⁻¹ + si-NC group

To further explore the role of EGCG and warfarin in protecting cells from the adverse effects of H₂O_2, we analyzed the protein levels and gene expression of HIF-1α under the experimental conditions. As anticipated, immunofluorescence shows that HIF-1α is upregulated in cells subjected to oxidative stress (Fig. 7A). The addition of warfarin and EGCG (20 μmol·L⁻¹) greatly reduced the levels of HIF-1α. Likewise, the addition of EGCG and warfarin reduced levels of the proteins associated with HIF-1α (VEGF, VEGFR-1 and ATGR-1) and the phosphorylation of proteins in associated pathways (P13K, Akt, P706SK, and Erk1/2) in cells subjected to oxidative stress (Fig. 7B). Quantification of VEGF and ATGR-1 mRNA indicated that both genes are expressed at significantly higher levels in the model and in the model treated with warfarin. The expression level was significantly reduced in the model after combined treatment with EGCG and warfarin, but began to increase when HIF-1α was inhibited (Fig. 7C). These results indicate that the antioxidative effects of EGCG and warfarin combination on HUVECs are associated with the expression of HIF-1α and the activity of the PI3K/AKT and ERK1/2 signaling pathways.

Figure 7.  EGCG and warfarin combination regulates HIF-1α and VEGF expression through PI3K/AKT and ERK1/2 signaling pathways and attenuates cellular injury. A, Immunofluorescence analysis of HIF-1α in different groups bar, 20 μm. B, Expressions of HIF-1α, VEGF, VEGFR-1, AGTR-1, PI3K/AKT, and ERK1/2 proteins in vitro was determined by western blotting, GAPDH was used as the loading control. C, mRNA expression of AGTR-1 and VEGF in the different groups was investigated by reverse transcription-quantitative polymerase chain reaction. Data are presented as mean ± SEM (n = 5); ^aP < 0.05 model groups vs control group; ^bP < 0.05 model + warfarin group vs model group; ^cP < 0.05 model + warfarin group + EGCG 20 μmol·L⁻¹ vs model + warfarin group; ^dP < 0.05 model + warfarin + EGCG 20 μmol·L⁻¹ + si-HIF-1α group vs model + warfarin + EGCG 20 μmol·L⁻¹ + si-NC group

DVT is a major cause of mortality and morbidity worldwide and causes complications in hospitalized patients ^{[28, 29]}. Anticoagulants like heparin and warfarin are frequently used to treat DVT but they can cause spontaneous bleeding ^[30]. Therefore, accessible prophylaxis and non-toxic complimentary medication provide appealing solutions. The study by Ikeda et al. demonstrated that EGCG interacts with Trp-213 at the drug binding site I via π-π stacking ^[31]. Besides, warfarin has been found to bind to drug binding sites I, II, and III in several investigations ^{[32, 33]}. In this study, we investgiated the ability of EGCG as a complementary medication to improve anticoagulant properties in a rabbit model of DVT and in HUVECs exposed to H₂O₂.

Several studies have reported ethnopharmacological methods that reduce the peevalence of DVT ^[34-37]. For instance, the aqueous extract of Whitmania pigra was found to reduce the burden of DVT by alleviating inflammatory responses through the SIRT1/nuclear factor-kappa B (NF-κB) signaling pathway ^[35] and antioxidation ^[34]. Similarly, Spatholobi Caulis, a traditional Chinese medicine used to treat blood-related illnesses, was found to ameliorate DVT by suppressing platelet aggregation and reducting inflammation ^[38]. Earlier, Zhang discovered that Danhong huayu koufuye (DHK) could also prevent DVT in rats through the inhibition of inflammation ^[37]. DHK administration to a rat model of DVT significantly reduced levels of neutrophils, lymphocytes, and matrix metalloproteinases-9 and resulted in a significant reduction in the size of thrombi and vein wall thickness. EGCG has been found to have antioxidant and anti-inflammatory properties. EGCG was discovered to regulate the expression of HIF-1 and VEGF in age-related macular degeneration, inhibiting the progression of choroidal neovascularization ^[39]. EGCG has also been used successfully as complementary medicine because of its antioxidant and apoptotic effects in cancer ^[40]. In the current study, we discovered that using EGCG in combination with warfarin improved hemorheological parameters and coagulation function in a rabbit model of DVT. In a dose-dependent manner, EGCG in combination with warfarin could reduce thrombus volume and vein wall thickness. The determination of parameters associated with coagulation, such as D-dimer, P-selectin, NO levels, 20-HETE, and vWF supports the beneficial effects of EGCG as a complementary medication to prevent and resolve thrombus formation.

The antioxidant properties of EGCG and warfarin combination during DVT were also studied in this study by evaluating the expression of HIF-1, VEGF, VEGFR1, and ATGR-1 in the rabbit model and HUVECs. The DCF reagent was also used to measure ROS production. There was a dose-dependent decrease in expression of HIF-1, VEGF, VEGFR1, and ATGR-1 in the DVT model in response to H2O2. Although HIF-1 and VEGF are known to be upregulated in DVT, they are thought to function independently in the IVC. According to reference 9, VEGF is an angiogenic growth factor that promotes epithelial cells growth and infiltrates the thrombus during resolution ^[25]. However, they found that levels of HIF-1α in the thrombus were lower than in the IVC and speculated that HIF-1α may induce the expression of VEGF via interactions with VEGFR-1. Another study found that suppressing HIF-1 by Rhaponticin inhibited the expression of VEGF and reduced the levels of metastatic and angiogenic activities in cancer ^[41]. Similar results were obtained in our study, where HIF-1 suppression hampered HUVEC migration and invasion while increasing ROS levels. When HIF-1 is inhibited, the mRNA levels of VEGF and AGTR-1 in HUVECs rise.

Pathways including PI3K/AKT and ERK1/2 have been linked to DVT ^{[18, 42]}. Recent research suggests that thrombin activates the ERK1/2 pathway to initiate a VEGF-induced proangiogenic response ^[43]. Rumex acetosa’s antiplatelet action was reported to reduce thrombus formation in rats through modulating PI3K/AKT and ERK1/2 ^[44]. Elevated levels of ROS are known to activate platelets by disrupting the PI3K/AKT signaling pathway ^[45]. Our results demonstrate that the combination of EGCG and warfarin could reduce levels of HIF-1α and the phosphorylation of proteins in associated pathways, such as P13K, Akt, P706SK, and Erk1/2, in HUVECs under oxidative stress.

In this study, we examined the combined use of EGCG and warfarin in a DVT animal model induced by ligation of the IVC and in vitro studies in HUVECs subjected to oxidative stress. In combination with warfarin, EGCG significantly reduced thrombus formation and protected HUVECs from oxidative stress and apoptosis. However, the silencing of HIF-1α reduced the advantages of combining EGCG and warfarin and upregulates VEGF expression as well as activation of the PI3K/AKT and ERK1/2 signaling pathways. Overall, EGCG combined with warfarin allievate DVT by modulating the expression of HIF-1α and VEGF through the PI3K/AKT and ERK1/2 signaling pathways.