Thrombin disrupts vascular endothelial‐cadherin and leads to hydrocephalus via protease‐activated receptors‐1 pathway

Previous studies indicated that intraventricular injection of thrombin would induce hydrocephalus. But how thrombin works in this process remains unclear. Since cadherin plays a critical role in hydrocephalus, we aimed to explore the mechanisms of how thrombin acted on choroid plexus vascular endothelium and how thrombin interacted with vascular endothelial‐cadherin (VE‐cadherin) during hydrocephalus.

Cerebrospinal fluid (CSF) is mainly produced in the choroid plexuses, which consist of plenty of capillaries constructing the blood-CSF barrier. 11 Choroid plexuses dysfunction would be expected to cause an increase in the production of CSF. 12,13 The transmembrane protein vascular endothelial-cadherin (VE-cadherin) exists in the central nervous system endothelium. 14 VE-cadherin belongs to the cadherin superfamily which is involved in cell-cell adhesion and signaling. 15 Moreover, it has been recognized that cadherin plays a critical role in the formation of hydrocephalus. [16][17][18] It is also found that thrombin could lead to the degradation of VE-cadherin as well as the breakdown of the blood-brain barrier (BBB) integrity and thus induce the brain edema after ICH. 19 Furthermore, PAR1 is generally acknowledged as the main thrombin receptor on the endothelial cells. 20,21 Previous studies suggested that thrombin increased the microvascular permeability by activating PAR1 pathway, which increased the activity of downstream protein kinases, such as p21-activated kinase-1 (PAK1). 19 Several lines of evidence also indicated that the phosphorylation of Src or PAK1 was involved in the degradation and loss of VE-cadherin. [22][23][24][25] In this study, we hypothesized that thrombin-induced hydrocephalus was related to the PAR1 signaling pathway which resulted in the decline of VE-cadherin in choroid plexuses. And this process may shed light on the formation of hydrocephalus after IVH. (3 months old, Zhejiang University Laboratories), at the weight of 280-320 g, were used in this study. Animals were anesthetized with pentobarbital (50 mg/kg intraperitoneally (IP)) and were positioned in a stereotaxic frame. A cranial burr hole (1 mm) was drilled, and a 26-gauge needle was inserted stereotaxically into the right lateral ventricle (coordinates: 0.6 mm posterior, 4.5 mm ventral, and 1.6 mm lateral to the bregma). Saline or thrombin (3U in 50 μL) was infused using a microinfusion pump (World Precision Instruments). The needle was removed after injection, the burr hole was filled with bone wax, and the skin incision was sutured.

| Experimental groups
In brief, the rats were randomly divided into two parts. First, rats received an intracerebroventricular (ICV) injection of three units rat thrombin (Sigma-Aldrich, St Louis, MO, USA) in saline or saline alone (50 μL) in 7 minutes and were anesthetized at 24 hours for magnetic resonance imaging (MRI) scanning. And then, rats were euthanized for intracardiac perfusion and the brains were used for Western blot (n = 4 for each group), brain histology (n = 4 for each group), and Evans blue assay (n = 4 for each group). Second, rats were treated with vehicle (1% dimethyl sulfoxide in saline) or PAR1 inhibitor (SCH79797, Abcam, USA, 25 μg/kg, IP), Src inhibitor (PP2, Abcam, USA, 1 mg/kg, IP), PAK1 inhibitor (IPA3, Abcam, USA, 1 mg/kg, IP) immediately while thrombin infusion. Serial MRI scanning was performed 24 hours after injection, and the brains were used for Western blot (n = 4 for each group, Figure 1).

| MRI scanning and ventricle volume measurement
Under chloral hydrate anesthesia, the rats were placed in a rodent coil. Magnetic resonance imaging was performed in a 3.0-T Varian F I G U R E 1 Experimental designs. (A) Experiment 1 was designed to investigate the effect of thrombin on dilation of ventricles and the VE-cadherin expression in choroid plexus. (B) Experiment 2 was designed to evaluate the effect of PAR1/p-Src/p-PAK1 pathway on thrombin-induced hydrocephalus MR scanner (GE Signa HDxt) with a T2 sequence using a view field of 50 mm × 50 mm, a matrix of 128 × 128, and 0.8 mm slice thickness. Ventricular volumes were measured and calculated as described previously. 8 Bilateral ventricles from frontal horns to the fourth ventricle were outlined and measured. Ventricular volume was calculated by summing the ventricle areas over all slices and multiplying by section thickness. All image analysis was performed using ImageJ program by a masked observer.

| Immunohistochemistry and immunofluorescence
Rats were anesthetized with pentobarbital (100 mg/kg, IP) and perfused with 4% paraformaldehyde in 0.1 mol/L phosphate-buffered saline (pH 7.4). The brains were removed and kept in 4% paraformaldehyde for 24 hours and then immersed in 30% sucrose for 2-3 days at 4°C. Brains were embedded in optimal cutting temperature compound and 18-µm thick slices cut using a cryostat (LEICA, CM3050S).

| Western blot analysis
Briefly, brain tissue was immersed in western sample buffer and lysed. The lysate was centrifuged at 14 000 g for 30 minutes at 4°C, and the supernatant was collected. Protein concentration was determined by the BCA protein assay kit (Pierce, Thermo Scientific, USA). Protein samples were separated using SDS-polyacrylamide gel electrophoresis and transferred to the nitrocellulose membrane.
After blocking with 5% nonfat milk, the membranes were incubated

| Evans blue extravasation assay
Evans blue dye (2%; 5 mL/kg) was injected intraperitoneally and allowed to circulate for 3 hours. Rats were then euthanized by intracardiac perfusion with 0.01 mol/L ice-cold phosphate-buffered saline, and brains were divided into six regions. Tissues were homogenized in 50% trichloroacetic. After centrifugation (4°C, 20 000 g for 20 minutes), the amount of extravascular Evans blue dye was measured by spectrofluorophotometer. Evans blue extravasation was calculated as thrombin/saline ratios according to different regions.

| Statistical analysis
Values are given as means ± SD. T tests were used to compare the data between saline and thrombin group. One-way ANOVA with Tukey's post hoc test was used to analyze the data between multiple groups. Differences were considered significant at P < 0.05.

| Intraventricular injection of thrombin caused hydrocephalus
This current study confirmed that injection of thrombin resulted in obvious ventricles dilation on MRI (Figure 2A). Lateral ventricular volumes in thrombin-injected rats were significantly larger than in rats receiving the saline injection at 24 hours (21.82 ± 1.54 vs 8.73 ± 0.56 mm 3 in saline control, P < 0.01, Figure 2B).

| Intraventricular injection of thrombin-induced impairment of blood-brain barrier
Compared with the saline injection group, we observed more albumin deposit and Evans blue leakage around the brain ventricles after thrombin injection ( Figure 3A

| Intraventricular injection of thrombin resulted in downregulation of VE-cadherin in choroid plexus
We observed the expression of VE-cadherin in choroid plexus during thrombin-induced hydrocephalus. At 24 hours after thrombin injection, immunofluorescence staining of choroid plexus was performed. In the saline injection group, VE-cadherin abundantly distributed in the borders of choroid plexus endothelial cells.
After thrombin injection, the expression of VE-cadherin was remarkably decreased ( Figure 4A) in the choroid plexus compared to the saline-injected group, which was assessed by Western blot (0.54 ± 0.06 vs 1.05 ± 0.07 in the saline group, P < 0.01, Figure 4B, C).

| Thrombin downregulated VE-cadherin in choroid plexus via activating PAR1/p-Src/p-PAK1 pathway
We used the PAR1 inhibitor (SCH79797), Src inhibitor (PP2), and PAK1 inhibitor (IPA3) to explore the downstream mechanism in the thrombin-induced hydrocephalus. We found that PAR1, p-Src, and p-PAK1 were significantly increased in the thrombin group (P < 0.05; Figure 6A-C). The increase of PAR1 was reversed by SCH79797, but not PP2 or IPA3 treatment (P < 0.05; Figure 6A), it's reasonable to conclude that PAR1 is upstream of p-Src and p-PAK1. Meanwhile, the increase of p-Src was reversed by SCH79797 and PP2 treatment, but not IPA3 (P < 0.05; Figure 6B), this indicated p-Src is an upstream  Figure 6C). In addition, administration of SCH79797, PP2, or IPA3 led to improved outcome in the level of VE-cadherin (P < 0.05; Figure 6D). Overall, it indicated that PAR1/p-Src/p-PAK1 pathway played an important role in thrombin-induced downregulation of VE-cadherin.

| DISCUSS IONS
The major findings of this study are as follows: (a) thrombin contrib- The mechanisms of brain damage caused by IVH have been poorly illustrated. Dilatation of the ventricles is a specific character of severe IVH or subarachnoid hemorrhage (SAH) and has been reported as an unfavorable predictor of the patients' survival and neurological outcome. [26][27][28] Previous theory supported that the mass effect of blood clots may be one of the major factors accounting for hydrocephalus following IVH, but these mechanical factors could not explain all aspects of clinical situations. 29,30 IVH secondary to ICH is a medical emergency, and the deterioration of neurological functions occurs within the first 24-48 hours. 1,31,32 Thrombin is an essential component of the coagulation cascade and can be produced immediately in the brain after IVH. The increased plasma thrombin after IVH can result from the disruption of arteries and the endogenous thrombin production. 19 Thrombin production is related to early brain injury (within the first 2 days) after hemorrhage stroke. 6,33 Our current study focused on the time point of 24 hours after thrombin injection to explore the role of thrombin in IVH. In the previous study, we found that heparin could alleviate IVH-induced hydrocephalus and intraventricular injection of thrombin alone could induce hydrocephalus 8 ; all these studies suggested that thrombin should be responsible for the IVH-induced hydrocephalus. Our present study aimed to identify how thrombin causes hydrocephalus.  19,38 In the present study, we actually found the hyperpermeability in choroid plexus and/or around the brain ventricles. The downregulation of VE-cadherin in choroid plexuses was also observed, which impaired cellular adhesion and resulted in an irreversible blood-CSF barrier disruption.
Finally, this change contributed to an increased vessel permeability and the extra-leaking of CSF, which was shown as hydrocephalus.
Moreover, we explored the mechanisms of how the thrombin influences the function of VE-cadherin during hydrocephalus. PARs are G-protein-coupled receptors specific for thrombin signaling. 21 Activating PAR1 could generate increased microvascular permeability, which is confirmed in the SAH model. 19 Our study also observed a significant increase of PAR1 in the hydrocephalus model, and we hypothesized that thrombin increased the choroid plexus permeability by activating PAR1 and degrading VE-cadherin. Activation of the PARs accelerated the activation of Src kinases and PAK1. 19 Src contributed to increased vascular permeability, brain edema, and mortality after intracerebral hemorrhage. 39,40 And PAK1 is a downstream Rac effector and a major cyclin-dependent kinase 5 (Cdk5) substrate which is related to neurons apoptosis following human stroke. 41 In this study, the level of phosphorylated Src and PAK1 was increased after thrombin injection, and SCH79797 treatment markedly suppressed the phosphorylation of Src and PAK1. This is consistent with the mechanism that thrombin leads to brain edema in the SAH model. 19 What is more important, inhibition of PAR1, Src, or PAK1 can all attenuate the thrombin-induced hydrocephalus with the reversion of the VE-cadherin expression. This is in line with the hypothesis that activation of PAR1/p-Src/p-PAK1 would accelerate the degradation of VE-cadherin and ultimately give rise to hydrocephalus.
However, there is also considerable evidence on other roles of thrombin in brain damage. For example, thrombin triggered platelet activation which could release inflammatory cytokines 42 and disrupt the integrity of endothelial cell monolayers. 43 Claudin-5, ZO-1, and other tight junction proteins also contributed to the thrombininduced increase in vascular permeability. 38,44,45 But their roles in hydrocephalus still remain to be understood. Future studies will examine more detailed scenarios of hydrocephalus after IVH.

| CON CLUS ION
In summary, the present study demonstrated that thrombin-induced hydrocephalus can be partly attributed to the increased permeability of blood-CSF barrier, which may be due to the degradation of VE-cadherin via the activation of PAR1/ p-Src/ p-PAK1 pathway.

CO N FLI C T S O F I NTE R E S T
The authors declare no conflict of interest.