Introduction
Venous thromboembolism consists of deep vein thrombosis and pulmonary embolism, and is the third most common cardiovascular disease (CVD) worldwide (1,2). Deep venous thrombus is a CVD and a serious clinical issue that has shown a significantly increasing incidence over the last 20 years leading to pulmonary thromboembolism, and even the death of patients with acute deep venous thrombosis (3). Deep vein thrombosis is a pathological CVD and is induced by a large number of risk factors including various genetic factors, dietary habits, obesity, pregnancy, aging, drugs, trauma and cancer (4,5). Deep vein thrombosis frequently leads to metabolic syndrome and other diseases, resulting in a higher risk of deep vein thrombosis-caused death (1,6). Clinical investigation has revealed that the incidence is associated with gender and has also shown that thrombus embolization into inferior vena cava filters is the central nodes during factor-induced thrombolysis for proximal deep venous thrombosis (7–9).
The imbalance between the coagulation and fibrinolytic system plays an important role in the progression and pathogenesis of arterial thrombosis (10). Previous studies have suggested that the activities of thrombin-activatable fibrinolysis inhibitor (TAFI) and plasminogen activator inhibitor-1 (PAI-1) play crucial roles in the initiation and development of deep venous thrombus (11–13). TAFI is an anti-fibrinolytic factor, lower levels of which have been associated with certain comorbidities, such as CVD (14,15). However, previous studies have demonstrated that the influence of genetic variations in the TAFI gene on the risk of CVD is inconclusive. Plug and Meijers have put forward new clues regarding the mysterious mechanism of activated TAFI self-destruction (16). In addition, previous research has also indicated the PAI-1 regulates the balance of the plasma fibrinolytic and blood coagulation system and further initiates or promotes the progression of CVD (17,18). Heineking et al indicated the relationships between sinus venous thrombosis and homozygosity for the PAI-1 4G/4G polymorphism in intraventricular hemorrhage (19). Lichy et al suggested that the evidence of the PAI-1 genotype as a risk factor for cerebral venous thrombosis is controversial (20). Furthermore, Ringelstein et al analyzed the promotor polymorphisms of PAI-1 and other thrombophilic genotypes in cerebral venous thrombosis in a clinical study (21). Moreover, the plasma concentrations of carboxypeptidase, CPU and TAFIa were found to inhibit clot lysis as tissue plasminogen activator (tPA) leading to thrombus stability in vivo (22). These studies suggest that TAFI and PAI-1 may be associated with the process and progression of venous thrombus. We found that the anticoagulant therapy of rivaroxaban can inhibit the expression and activities of TAFI and PAI-1 through matrix metalloproteinase-9 (MMP-9)-mediated NF-κB signaling pathway in venous endothelial cells and in a rat model of deep venous thrombus.
The family of NF-κB transcription factors has five cellular members and it has been reported that NF-κB is involved in the process of venous thrombus (23). NF-κB transcription factors can regulate the expression of tissue factor, which plays a crucial role as a principal initiator of the coagulation cascade by regulating p50/p65 heterodimer (24). Hashikata et al suggested that rivaroxaban inhibits angiotensin II-induced activation in cultured mouse cardiac fibroblasts through modulation of the NF-κB signaling pathway (25). Therefore, we assumed that rivaroxaban may attenuate deep venous thrombus through the MMP-9-mediated NF-κB signaling pathway. Our data showed that rivaroxaban not only inhibited TAFI, PAI-1, ADP, PAIs, von Willebrand factor (vWF) and thromboxane expression levels, but also improved neutrophils, tissue factor, neutrophil extracellular traps (NETs), myeloperoxidase and macrophages in microvascular endothelial cells and in a rat model of deep venous thrombus. We also investigated whether rivaroxaban can improve fibrinolysis and impact deep venous thrombus through the MMP-9-mediated NF-κB signaling pathway in a rat model.
In this study, we investigated the efficacy and related molecular mechanism of rivaroxaban-mediated differentiation changes in TAFI, PAI-1, inflammatory factors, thrombosis factors and pathological characteristics in vein endothelial cells and in rats with venous thrombosis. Our data suggest that rivaroxaban presents anti-inflammatory and pro-fibrinolytic properties determined by both in vitro and in vivo analysis through MMP-9-mediated NF-κB signaling. These findings suggest that rivaroxaban may be a potential anti-thrombotic drug for the treatment of deep venous thrombosis.
Materials and methods
Ethical approval and participant consent
This study was approved by the Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University. All surgery and euthanasia of experimental rats were performed under sodium pentobarbital anesthesia to minimize suffering.
Cells and reagents
Vein endothelial cells were isolated from SD rats and cultured in MEM medium (Sigma-Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, Carlsbad, CA, USA). Vein endothelial cells were treated with rivaroxaban or phosphate-buffered saline (PBS) as the control for 72 h. Cells were cultured in a 37°C humidified atmosphere of 5% CO2.
Western blotting
Vein endothelial cells were homogenized in lysis buffer containing protease inhibitor and centrifuged at 8,000 rpm at 4°C for 10 min. The supernatant of the mixture was used for analysis of the relevant protein using sodium dodecyl sulfate (SDS) assay according to the manufacturer's instructions (26). The primary goat anti-rat antibodies [anti-TAFI (1:1,000; ab181990), anti-PAI-1 (1:1,000; ab125687), anti-vWF (1:1,000; ab6994), anti-ADP (1:1,000; ab22554), anti-MMP9 (1:1,000; ab73734), anti-NF-κB (1:1,000; ab32360) (all from Abcam, Cambridge, UK)] were added after blocking (5% skimmed milk) for 60 min at 37°C and then washing with PBS three times. Subsequently, incubation with the secondary rabbit anti-goat antibody (1:1,000; ab6741; Abcam, UK) was carried out for 24 h at 4°C. The results were visualized using a chemiluminescence detection system.
Fluorescent quantitative RT-PCR
Total RNA was extracted from vein endothelial cells and the identified RNA was applied to the cDNA synthesis by reverse transcription PCR. One-tenth of the cDNA was used for fluorescent quantitative RT-PCR by using the iQ SYBR-Green system. Relative multiples of change in mRNA expression was calculated by 2−ΔΔCt. The results are expressed as the n-fold difference relative to normal β-actin control.
Animal study
SD rats were purchased from Vital River Laboratory Animal Technology Co., Ltd. (Shanghai, China). Rats were used to establish the model of deep venous thrombosis by using heparin according to a previous study (27). Heparin-induced rats with deep venous thrombosis were divided into two groups and received treatment with rivaroxaban or PBS as a control for 60 days. Rats were treated with an intravenous injection of rivaroxaban (10 mg/kg body weight) or PBS once a day and the total treatment continued for 60 days. All rats were housed at a suitable temperature with a 12 h light/dark cycle and free access to food and water. The rats were sacrificed for further analysis.
Histologic and immunohistochemical analyses
Vein endothelial tissues were isolated from experimental mice after the 60-day treatment with rivaroxaban (10 mg/kg body weight) or PBS. The brains were frozen and coronal sections were cut in a cryostat after perfusion, fixation and cryoprotection. Free-floating sections were rinsed and placed in the solution with the primary antibody of goat anti-mouse Aβ-42 (1:1,000; K10054; Baiaolaibo Science and Technology. Ltd., Beijing, China), Aβ-40 (1:1,000; K10098; Baiaolaibo Science and Technology. Ltd.) and APP (1:1,000; ab180140; Abcam). After incubation for 60 min, the sections were washed and incubated with the secondary rabbit anti-goat antibodies (1:500; Chemicon International, Temecula, CA, USA) for MMP-9, NF-κB, apolipoprotein and thrombomodulin staining, respectively. The sections were washed and observed by fluorescence video microscopy (BZ-9000; Keyence Co., Osaka, Japan). Immunohistochemical staining was used to examine the content of neuroprotection-related proteins in the hippocampus. Immunohistochemical procedures were previously reported in detail (28).
Preparation of platelets and leukocytes for intravital microscopy
Rat platelets in the location of the deep venous thrombus were analyzed and labeled with 5-carboxy-flourescein diacetate succinimidyl ester (DCF) as previously reported [Massberg et al (29)]. The neutrophils and monocytes in the lesions of deep venous thrombus were analyzed as detailed in a previous study (30).
Assessment of thrombus formation in vivo
Thrombus formation in vivo was measured in HCV using an Alexa Flour 488 (Invitrogen, Carlsbad, CA, USA) ex vivo labeled anti-fibrin antibody. Fluorescence intensity was quantified by intravital video microscopy (BX51WI; Olympus, Tokyo, Japan).
Activity analysis
The activities of MMP-9, NF-κB, tissue factor (TF), TAFI and PAI-1 in vein endothelial cells and deep venous thrombus were analyzed by commercialized kits [MMP (ab100732; Abcam), NF-κB (JK-(a)-6261; Jiangkang Bioscience, Shanghai, China), TF (ab214091; Abcam), TAFI (MA143023; Beinuo Bioscience, Shanghai, China), PAI-I (ab197752; Abcam)] and performed according to the manufacturer's instructions.
Statistical analysis
All data are presented as mean and SEM. Statistical significance was determined utilizing two-tailed Student's t-test determined by SPSS 19.0. Two-way ANOVA, Kaplan-Meier, one-way analysis of variance (ANOVA), followed by Student's two-tailed t-test and log-rank statistical analyses were performed utilizing GraphPad software. P<0.05 was considered to indicate a significant difference between rivaroxaban and control group.
Results
Rivaroxaban inhibits expression and activity of TAFI and PAI-1 in vein endothelial cells and a rat model of deep venous thrombosis
A model of thromboplastin-induced thromboembolism was established for evaluation of the profibrinolytic agent, rivaroxaban. We first analyzed baseline levels of PAI-1 and PAI-1 in vein endothelial cells and in plasma concentrations in the rat. As illustrated in Fig. 1A and B, TAFI and PAI-1 protein expression levels were upregulated in thromboplastin-treated vein endothelial cells (ThrVEC). However, rivaroxaban treatment downregulated TAFI and PAI-1 protein expression levels in the vein endothelial cells. We also found that rivaroxaban treatment had inhibitory effects on TAFI and PAI-1 activity (Fig. 1C and D). In addition, we analyzed changes in TAFI and PAI-1 in the rats with deep venous thrombosis. As shown in Fig. 1E and F, plasma concentration levels of TAFI and PAI-1 were decreased in serum samples of the rivaroxaban-treated rats. Furthermore, activity of TAFI and PAI-1 was also decreased in the rivaroxaban-treated rats (Fig. 1G and H). Moreover, plasminogen activators and thrombin-activatable fibrinolysis were upregulated both in vitro and in vivo after treatment with rivaroxaban (Fig. 1I and J). Taken together, these results suggest that rivaroxaban can inhibit expression and activity of TAFI and PAI-1 both in thromboplastin-treated vein endothelial cells and heparin-induced deep venous thrombosis.
Rivaroxaban inhibits expression levels of inflammatory factors in endothelial cells and the rat model of deep venous thrombosis
In order to analyze the efficacy of rivaroxaban on inflammatory factors in endothelial cells and experimental rats with deep venous thrombosis, we examined the inflammatory signals, inflammatory factors and monocytes, neutrophils and platelets. As shown in Fig. 2A and B, inflammatory signals were increased along the thrombus edges in vivo, while rivaroxaban decreased inflammatory signals along the deep venous thrombosis edges. We also found that expression levels of inflammatory factors including IL-1, IL-6, IL-17 and TNFα were decreased in endothelial cells and the experimental rats with deep venous thrombosis after treatment with rivaroxaban (Fig. 2C and D). We showed venous inflammatory leukocytes were distributed in clusters or layers adjacent to the intact endothelium and were attenuated by rivaroxaban treatment (Fig. 2E and F). We then determined whether accumulation of innate immune cells is a cause of deep venous thrombosis formation. The results in Fig. 2G and H revealed that leukocytes adhered directly to the venous endothelium from the experimental rats, whereas endothelial disruption was attenuated in the rivaroxaban-treated rats. Furthermore, neutrophils and monocytes were analyzed in the main leukocyte subsets accumulating during the initiation of deep venous thrombosis. We found that rivaroxaban decreased the numbers of neutrophils and monocytes in vivo as determined by intravital two-photon microscopy (Fig. 2I and J). Taken together, these results suggest that rivaroxaban inhibits inflammatory signals and expression levels of inflammatory factors in endothelial cells and in the rat model of deep venous thrombosis in vivo.
Rivaroxaban regulates the balance between clotting and anti-clotting factors in vein endothelial cells
Previous research has reported that thrombogenic factors are indicators and play a crucial role in patients with deep venous thrombosis (31). Therefore, we analyzed changes in the balance between clotting and anti-clotting factors in endothelial cells and experimental rat. We observed that expression levels of ADP, vWF and thromboxane were increased in the endothelial cells and experimental rats, whereas rivaroxaban significantly downregulated these in the endothelial cells and experimental rats (Fig. 3A–C). In addition, transsulfuration enzymes (ETS), CBS and CGL activities in vein endothelial cells were upregulated in the endothelial cells and experimental rats after rivaroxaban treatment (Fig. 3D–F). Furthermore, we found that the plasma concentration levels of monounsaturated and saturated fatty acids were increased and the ratio of oleic to palmitic acid (MUFA:SFA) was imbalanced in the rats with deep venous thrombosis (Fig. 3G and H). Moreover, we observed that concentrations of triacylglycerols, TF, fibrinogen, tissue-type plasminogen activator (t-PA) were decreased by rivaroxaban treatment (Fig. 3I). Taken together, these results suggest that rivaroxaban can regulate the balance between clotting and anti-clotting factors in thromboplastin-treated vein endothelial cells and heparin-induced deep venous thrombosis.
MMP-9 contributes to inflammatory cell recruitment and collagen metabolism in thrombus resolution
We investigated the expression level, activity and function of MMP-9 in vein endothelial cells in rat thrombus resolution. As shown in Fig. 4A and B, the expression level and activity of MMP-9 were increased in the rats with heparin-induced deep venous thrombosis compared to the controls. However, rivaroxaban treatment significantly inhibited expression levels and activity of MMP-9 in vein endothelial cells. We also examined the effects of MMP-9 on collagen metabolism and expression of inflammatory fibrotic mediators. We found that there were significant decreases in the number of macrophages in the rivaroxaban-treated rats compared to the control (Fig. 4C). In addition, vWF staining for endothelial cells presented significant differences compared to the control (Fig. 4D). In addition, MMP-9 also decreased collagen metabolism in vein endothelial cells and rivaroxaban treatment increased collagen metabolism in the vein endothelial cells (Fig. 4E). Deep venous thrombosis also influenced vein endothelial cell viability and rivaroxaban increased the viability of the vein endothelial cells after venous thrombosis (Fig. 4F). Furthermore, rivaroxaban also improved elastin fibers and the stiffness of collagen (K1 parameter) after thrombus resolution compared to the control, whereas MMP-9 decreased these effects of rivaroxaban (Fig. 4G and H). Taken together, these results indicate that MMP-9 can attribute to inflammatory cell recruitment and collagen metabolism in thrombus resolution.
Rivaroxaban regulates expression and activities of TAFI and PAI-1 through the MMP-9-induced NF-κB signaling pathway
In order to analyze the molecular mechanism of rivaroxaban-mediated activities of TAFI and PAI-1, we examined the NF-κB signaling pathway in vein endothelial cells and the rat model of deep venous thrombosis. We first analyzed the promoter activity of NF-κB target genes in vein endothelial cells from the experimental rats. As shown in Fig. 5A, 7p105/p50 (NF-κB1), p100/p52 (NF-κB2), p65 (RelA), RelB and c-Rel expression levels were increased in vein endothelial cells in the rats with deep venous thrombosis. However, rivaroxaban increased NF-κB transcription factors. In addition, IκBα, IκBβ and IκBε expression levels were decreased in the vein endothelial cells in the rivaroxaban-treated rats (Fig. 5B). In addition, we observed that promotion of MMP-9 (MMP-9P) activity canceled rivaroxaban-inhibited activities of TAFI and PAI-1 in the vein endothelial cells (Fig. 5C and D). In addition, we found that rivaroxaban decreased and restoration of MMP-9 activity decreased NF-κB activity in the vein endothelial cells (Fig. 5E). The expression levels of E-selectin and VCAM-1 were downregulated by rivaroxaban and upregulated by the MMP-9 promotor in vein endothelial cells (Fig. 5F and G). Furthermore, we found that the MMP-9 promoter promoted TF and ETS, CBS and CGL activities in the vein endothelial cells (Fig. 5H). Taken together, these findings suggest that rivaroxaban can regulate expression and activities of TAFI and PAI-1 through the MMP-9-induced NF-κB signaling pathway.
Rivaroxaban exhibits benefits for rats with heparin-induced deep venous thrombus
After analysis of the molecular mechanism of rivaroxaban in vein endothelial cells, we further examined the in vivo effects of rivaroxaban on rats with heparin-induced deep venous thrombus. As shown in Fig. 6A, our data demonstrated that representative thrombus apparent diffusion coefficient maps at thrombus organization prior and post treatment of rivaroxaban. The average apparent diffusion coefficient values prior and post treatment with rivaroxaban or PBS of thrombus organization are shown in Fig. 6B. Axial histological sections of the thrombosed femoral vein in rats were further analyzed for thrombus burden (Fig. 6C). Rivaroxaban treatment resulted in a mean 34.32% reduction in luminal thrombus burden compared to PBS-treated group (Fig. 6D). In addition, fibrin and collagen plasma levels were decreased in the rivaroxaban-treated rats compared to PBS (Fig. 6E). We also identified that rivaroxaban decreased MMP-9 and NF-κB staining in the presence of endothelium lined channels within the thrombi (Fig. 6F). Furthermore, histopathological analyses of deep venous thrombi obtained from experimental rats showed that rivaroxaban treatment markedly improved thrombus samples stained with H&E or Masson trichrome solution (Fig. 6G). Moreover, we found that expression levels of apolipoprotein and thrombomodulin were decreased in vein endothelial cells after rivaroxaban treatment compared to controls (Fig. 6H). Taken together, these results revealed that rivaroxaban is an efficient anti-thrombotic drug for the treatment of heparin-induced deep venous thrombosis.
Discussion
Rivaroxaban has been reported as a novel anticoagulation agent for the treatment of venous thrombosis (32,33). Clinical research indicates that rivaroxaban can successfully treat heparin-induced thrombocytopenia presenting with deep venous thrombosis and pulmonary embolism (27). Although rivaroxaban is commonly considered as a factor Xa inhibitor that can directly suppress factor Xa for the treatment of venous thrombosis (34), the association between rivaroxaban and the NF-κB signaling pathway in the progression of heparin-induced deep venous thrombosis has not been investigated. In this study, we investigated the molecular mechanism of the rivaroxaban-mediated NF-κB pathway in cultured endothelial cells and a rat model of heparin-induced deep venous thrombus. We first found that deep venous thrombus promoted expression and activities of TAFI and PAI-1 in vein endothelial cells in a rat model of deep venous thrombosis. In addition, deep venous thrombus enhanced expression levels of inflammatory factors in endothelial cells and the rat model of deep venous thrombosis. Furthermore, deep venous thrombus disturbed the balance between clotting and anti-clotting factors in vein endothelial cells from the rat model of deep venous thrombosis. Notably, the most significant finding in this study is that deep venous thrombus induces transcription and activity of MMP-9 resulting in stimulation of the NF-κB signaling pathway in vein endothelial cells. Interestingly, rivaroxaban treatment not only regulated activities of TAFI and PAI-1 and the inflammatory response, but also inhibited MMP-9-induced inflammatory cell recruitment, collagen metabolism and NF-κB signaling pathway in thrombus resolution.
Venous thromboembolism is a severe life-threatening disease that comprises deep vein thrombosis and pulmonary embolism that significantly affects the life quality of patients with venous thrombosis (27,35). Previous studies have indicated that recruitment of inflammatory factors contributes to thrombogenesis and vascular inflammation, and immunomodulation (36,37). Terry et al showed that rivaroxaban improved patency and decreased inflammation in a mouse model of catheter thrombosis (38). Our results demonstrated that rivaroxaban decreased monocytes, neutrophils and inflammatory signals along the deep venous thrombosis edge, which contributed to the thrombolytic effects in the rat in vivo. The pleiotropic effects of rivaroxaban on endothelial function were systematacially investigated both in vitro and in vivo.
Currently, thrombogenic risk factors for atherothrombosis play an essential role in the initiation of deep venous thrombosis that is the proximate event that triggers most acute ischemic syndromes and episodes of sudden cardiac death (39). In this study, we reported that the balance of thrombogenic risk factors were disturbed in vein endothelial cells. However, these thrombogenic risk factors can be regulated by rivaroxaban in the rat with heparin-induced deep venous thrombosis. Evidence suggests that ETS, CBS and CGL activities are downregulated in serum during deep venous thrombosis (40). Our results showed that ETS, CBS and CGL activities were upregulated in endothelial cells and experimental rats after rivaroxaban treatment. In addition, Pacheco et al found that the ratio of oleic to palmitic acid is imbalanced between thrombogenic and fibrinolytic factors in patients with thrombosis (41). The plasma concentration levels of monounsaturated and saturated fatty acids and imbalance in the ratio of oleic to palmitic acid (MUFA:SFA) were improved by rivaroxaban treatment in the rats with heparin-induced deep venous thrombosis. Furthermore, rivaroxaban treatment also regulated plasma concentrations of triacylglycerols, TF, fibrinogen and t-PA, which is in accordance with previous studies (42,43).
Deep venous thrombosis resolution involves the plasmin and the matrix metalloproteinase (MMP) system. We also identified the functions of MMP-9 in the progression of thrombus resolution. Dewyer et al showed that inhibition of plasmin increases activity of MMP-9 and decreases vein wall stiffness during venous thrombosis resolution (44). Our data indicated that MMP-9 attributes to inflammatory cell recruitment and collagen metabolism in thrombus resolution. Inhibition of MMP-9 activity increased the viability of the vein endothelial cells and rivaroxaban improved the stiffness of collagen and elastin fibers (K1 parameter) induced by MMP-9 after thrombus resolution compared to the control. Nosaka et al found that immunohistochemical detection of MMP-9 in a stasis-induced deep vein thrombosis model can estimate thrombus age (45). Furthermore, MMP-9 polymorphisms are associated with increased risk of deep vein thrombosis in cancer patients (46). These studies indicate that MMP-9 may be a predicator of deep vein thrombosis and our findings show that MMP-9 is active in the progression of heparin-induced deep vein thrombosis.
In previous studies, the underlying molecular mechanism of NF-κB and prevention of NF-κB activation in inflammation and neointimal hyperplasia has been investigated (47,48). Our data also showed that rivaroxaban improved inflammation of deep vein thrombosis via the NF-κB pathway. In addition, NF-κB transcription factor p50 can critically regulate expression of tissue factor in deep vein thrombosis (24). Furthermore, inhibition of tissue factor expression by drugs for venous thrombosis via the Akt/GSK3β-NF-κB signaling pathway in the endothelium also has been reported in a previous study (23). In this analysis, tissue factor was also inhibited by rivaroxaban-mediated NF-κB signaling pathway. Notably, the p65/c-Rel heterodimer of NF-κB transcription factors has been previously shown to critically regulate TF expression (49). NF-κB transcription factors presented a decreasing trend in the vein endothelial cells after treatment with rivaroxaban.
In conclusion, in the present study, we investigated the efficacies and molecular mechanism of the rivaroxaban-mediated NF-κB pathway in vein endothelial cells and in rats with deep venous thrombosis. Our study design showed that deep venous thrombosis in association with perturbed blood flow is driven by MMP-9-induced NF-κB activity. Expression and activity of MMP-9 present a concerted interaction of monocytes, netting neutrophils, thrombotic risk factors, platelets and collagen metabolism, which reveals the mechanisms linking inflammation and deep venous thrombosis. Our data demonstrated that rivaroxaban can significantly improve pathological characteristics of deep venous thrombosis by suppressing these adverse factors by decreasing MMP-9 expression and activity in the NF-κB signaling pathway in venous endothelial cells both in vitro and in vivo. These findings may improve the benefit-to-risk profile of anticoagulant therapy and indicate that rivaroxaban may be a potential anti-thrombotic drug for the treatment of deep venous thrombosis.