Endothelial TREM‐1 receptor regulates the blood–brain barrier integrity after intracerebral hemorrhage in mice via SYK/β‐catenin signaling

Abstract Background Intracerebral hemorrhage (ICH) is a high mortality and disability stroke subtype. Destruction of the blood–brain barrier (BBB) is a crucial contributor to brain edema and neurological deficit after ICH. Triggering receptor expressed on myeloid cells 1 (TREM‐1) has been reported to be expressed in endothelial cells, but its role in ICH remains unclear. This study aims to evaluate the role of TREM‐1 on BBB permeability after ICH in mice. Methods Two hundred and forty‐two CD1 mice were used in this study. The ICH model was established by collagenase injection. LP17 was administered intranasally at 2 or 8 h after ICH to inhibit TREM‐1. To explore the underlying mechanism, SYK activation CRISPR was administered intracerebroventricularly with LP17, and Anti‐mouse TREM‐1 rat IgG2a (a specific TREM‐1 agonist) was injected intracerebroventricularly with R406 (a specific SYK inhibitor) intraperitoneally. Neurobehavioral outcome, brain water content, BBB permeability, and protein expression were evaluated. Results The expression level of the TREM‐1 receptor increased rapidly as early as 6 h after ICH, and it was mainly expressed on the endotheliocytes in the neurovascular unit. Early and delayed administration of LP17 significantly decreased brain edema and improved neurobehavioral outcomes at 24 h after ICH. LP17 reduced the BBB permeability by increasing β‐catenin, claudin‐5 and ZO‐1 expression. Furthermore, SYK activation CRISPR abolished the beneficial effect of LP17 on the expression of the above junction molecules. Meanwhile, R406 reversed the impact of the TREM‐1 activator on the downregulation of β‐catenin, claudin‐5 and ZO‐1 expression. Conclusions This study demonstrated that TREM‐1 deteriorated BBB permeability via modulating the expression of interendothelial junction molecules after ICH, and this regulation is partly mediated by the SYK/β‐catenin signaling pathway.

Intracerebral hemorrhage (ICH), a severe stroke subtype with high morbidity and mortality, accounts for 15%-20% of all strokes. 1 As one of the critical pathological processes in second brain injury after ICH, disruption of the blood-brain barrier (BBB) integrity results in perihematomal cerebral edema formation and peripheral inflammatory factors infiltration and contributes to high mortality and poor outcome of ICH. 2 Therefore, inhibition of BBB disruption can be a therapeutic target for ICH patients.
Triggering receptor expressed on myeloid cells 1 (TREM-1) is an immune-receptor expressed on the surface of myeloid cells. 3though the pathophysiological role of TREM-1 was first identified during infectious diseases, increasing studies suggested it participated in no-infectious disorders. 4Furthermore, TREM-1 was also found expressed on the surface of endothelial cells and regulated the vascular function recently. 5However, whether TREM-1 medicated the disruption of BBB integrity remained unclear.
Activated TREM-1 will provide a docking site for Spleen Tyrosine Kinase (SYK) via DAP12, then SYK will recruit and phosphorylate various signals. 4,6β-catenin, regarded as a critical signal transducer that regulates gene expression of endothelial junctions, 7 was also reported phosphorylated by SYK in recent studies. 8Inhibition or phosphorylation of β-catenin decreases junction molecules' expression level and amplifies BBB leakage. 9is study hypothesized that the endothelial TREM-1 receptor could modulate blood-brain barrier integrity by regulating the expression of junction molecules after intracerebral hemorrhage in mice.This effect was mediated through the SYK/β-catenin signaling pathway.

| Animals
Total 242 adults male CD1 mice were used as listed in Appendix S1.
All mice were housed in a 12-h light/dark cycle room with temperature and humidity control.Animals had free access to food and water.
All the experimental procedures were approved by the Institutional Animal Care and Use Committee at Sir Run Run Shaw Hospital.

| Study design
As shown in Appendix S1, five parts of experiments were designed and all mice were randomly assigned.

| Experiment 1
To explore the time-course expression of TREM-1 in collagenaseinduced ICH mice.Six groups were set randomly (six mice per group): sham, 6 h, 12 h, 24 h, 72 h and 7 days after ICH, and western blot was used to detect the expression level of TREM-1.Furthermore, two additional sham mice and two ICH mice at 24 h were used to assess neurovascular unit localization of TREM-1 by double immunohistochemistry staining.

| Experiment 2
LP17 with a dose of 1.0 μg/g was applied to evaluate the role of TREM-1 on the outcome of ICH mice at two different time points.

| Experiment 3
To explore the role of TREM-1 on BBB permeability, 30 mice were divided into three groups: sham, ICH + control peptide, and ICH + LP17 (1.0 μg/g).Evans blue (EB) test and IgG staining were used to evaluate BBB permeability at 24 h after ICH.

| Experiment 4
To further study the long-term outcome of LP17 treatment mice, 24 (n = 8 per group) mice were randomly divided into sham, ICH + control peptide, and ICH + LP17 (1.0 μg/g).Long-term neurobehavioral experiments and Nissl staining were analyzed.
The information of the above groups was blinded to the researchers who performed surgeries, drug administration, neurobehavioral tests, immunofluorescence staining western blot, Nissl staining, and data analysis.

| ICH model
The ICH model was established by injection of bacterial collagenase as previously described. 10Mice were anesthetized by intraperitoneal injection with a mixture of ketamine (100 mg/kg) and xylazine (10 mg/kg) (2:1).A 1-mm cranial burr hole was drilled under a stereotaxic head frame (Kopf Instruments) with the coordinates 0.2 mm posterior and 2.2 mm right lateral to the bregma.Then 0.5 mL sterile phosphate-buffered saline (PBS) with 0.075 U bacterial collagenase type VII was injected into the right basal ganglia at a rate of 0.167 mL/ min under a microinjection pump (Stoelting, MA).The needle was left in place for an additional 5 min, then withdrawn 1 mm per minute.The burr hole was covered with bone wax before the scalp was sutured.Finally, 0.4 mL of normal saline was injected subcutaneously to avoid postsurgical dehydration.Meanwhile, the sham operation was performed following the same protocol without injection.

| Drug administration
According to previous studies, a selective inhibitor of TREM-1 called LP17 (GenScript) with a dose of 1.0 μg/g and a control peptide was synthesized and administered intranasally at 2 h after ICH. 11ditionally, LP17 was administered at 8 h after ICH to evaluate the effect of delayed treatment.Anti-mouse TREM-1 rat IgG2a (0.25 μg/g; Thermo Fisher Scientific), as a specific TREM-1 agonist, or control rat IgG2a (Thermo Fisher Scientific), was administered intracerebroventricularly. R406 (5 μg/g; Selleckchem) as an inhibitor of SYK was dissolved in dimethyl sulfoxide (DMSO) and administered intraperitoneally.SYK activate (ACT) CRISPR (Santa Cruz Biotechnology) was used to activate syk gene expression. 1 μg ACT CRISPR was given intracerebroventricularly 48 h before the animal model set at a rate of 1 μL/min with a micro-injection pump.The scrambled CRISPRs (Santa Cruz Biotechnology) were injected with the same protocol.

| Intracerebroventricular injection
Intracerebroventricular administration was performed as previously described. 12A 10 μL Hamilton syringe was inserted into the left lateral ventricle stereotactically, and the coordinates were given as follows: 0.3 mm posterior, 1.0 mm lateral, and 2.3 mm below the dura.
The injection rate was 1 μL/min, and the needle was left in place for an additional 5 min before being slowly withdrawn for about 3 min.

| Neurobehavior tests
For short-term assessments, neurobehavioral outcomes were evaluated at 24 and 72 h after ICH by an independent researcher blinded to the procedure including three tests: modified Garcia test, forelimb placement test and corner turn test.For long-term evaluations, the rotarod test was performed every week after ICH, and the water maze test was assessed on days 21-25 post-ICH as previously described. 13The details of the tests were also listed in the Appendix S1.

| Brain water content
Brain water content was measured by the wet/dry method. 14ole-brain was collected at 24 h after ICH and separated into the ipsilateral cortex, ipsilateral basal ganglia, contralateral cortex, contralateral basal ganglia and cerebellum.The wet weight of brain tissues was obtained by weighing immediately, and then the tissues were dried at 100°C for 48 h before determining the dry weight.

| BBB permeability assessment
As reported, BBB permeability was evaluated by EB dye extravasation assay. 15Briefly, 4% EB dye (4 mL/kg, Sigma-Aldrich) was injected intraperitoneally at 3 h before sacrifice evaluation, and mice were sacrificed 24 h after ICH.The right hemisphere was homogenized and sonicated within 1100 μL PBS and then centrifuged for 30 min at 15,000 g. 500 μL supernatant from each sample was mixed with 500 μL 50% trichloroacetic acid overnight at 4°C.After centrifugation, the supernatant was collected, and EB dye extravasation was evaluated at 610 nm and quantified according to a standard curve.

| Western blot
Western blot was performed as previously described. 16The right brain hemispheres were homogenized in RIPA (Santa Cruz Biotechnology) and centrifuged at 14,000 g at 4°C for 30 min.

| Nissl staining
In long-term assessment, mice were sacrificed at 28 days after ICH.
Brains were fixed in formalin, dehydrated at 10%-30% sucrose, and sectioned at 20 μm.Sections were processed for Nissl staining to evaluate neuronal degeneration as previously described. 18The number of surviving neurons was counted in the hippocampal subfield (including CA1, CA3 and DG) under a light microscope.

| Statistical analysis
Data were shown as the mean and standard deviation (mean ± SD).
Statistical analysis was performed with GraphPad Prism 6.0.The Kolmogorov-Smirnov test was applied to check the normal distribution of the data, while F-test was used to check the homogeneity of variances.Multiple comparisons were statistically analyzed using one-way analysis of variance (ANOVA) followed by Tukey's HSD post hoc test.Long-term neurobehavioral results were analyzed by two-way ANOVA.p < 0.05 was considered statistically significant.

| ICH increased the expression of TREM-1 in the neurovascular unit
The temporal expression of the TREM-1 receptor after ICH was evaluated by western blot.The result showed that the expression of TREM-1 significantly upregulated surrounding hematoma, which started at 6 h and extended to 7 days after ICH (p < 0.05, Figure 1A,B).Furthermore, double immunofluorescence staining was performed to determine the cellular localization of TREM-1 in the neurovascular unit after ICH.As shown in Figure 1, TREM-1 is mainly expressed on the endotheliocytes and almost not on astrocytes.
Meanwhile, LP17 significantly decreased the brain water content in basal ganglia of the ipsilateral hemisphere at 24 and 72 h after ICH (p < 0.05, Figure 2D,H).

| LP17 preserved the blood-brain barrier permeability after ICH
Evans blue extravasation assay was measured at 72 h after ICH to assess the effect of TREM-1 on the BBB permeability.EB extravasation in the right hemisphere increased significantly after ICH, while LP17 treatment obviously reduced EB leakage compared with the control peptide group (p < 0.05, Figure 3A,B).To further evaluate the BBB permeability after LP17 treatment, IgG protein staining with immunohistochemistry was used.As shown in Figure 3C and Appendix S1, the LP17 treatment group had a lower perivascular IgG protein leakage when compared with ICH plus control peptide group.

| LP17 improved the long-term neurobehavioral outcome after ICH
Rotarod test and Morris water maze were applied to assess the effect of LP17 on long-term neurobehavior.The Rotarod test showed a shortened falling latency in the control group at 5 and 10 rpm accelerating velocity tests compared with the sham group.And LP17 significantly increased falling latency at 5 rpm in 1-3 weeks after ICH (p < 0.05, Figure 4A,B).In the Morris water maze, mice in the control group had a significantly longer swim distance and time of escape latency to find the platform than sham mice.However, LP17 improved the performance on days 1 and 3 (Figure 4C,D), and LP17 treatment also increased the time spent in the probe quadrant trials in ICH mice (Figure 4F).and the result suggested that LP17 spatial memory and learning ability of ICH mice on long-term.Furthermore, Nissl staining was used to evaluate the survival of neurons in the hippocampus at 28 days after ICH.As shown in Figure 4, the survival of neurons in the CA1 region of the right hippocampus from the control group significantly decreased compared with the sham.And there was less neuron loss and shrinkage neuron in the LP17 treatment group.Meanwhile, similar results were found in the CA3 and DG regions of the hippocampus in three groups (Figure 4G-J).

| TREM-1 regulated the expression of junction molecules via the SYK/β -catenin signaling pathway
To further explore the intracellular signal of TREM-1 that regulated the expression of junction molecules, an SYK activation (ACT) CRISPR was applied to activate the predicted pathway before LP17 administration.As expected, the level of SYK was increased after ICH, accompanied with a downregulation of β-catenin and upregulation of pβ-catenin, which indicated that SYK may participate in the phosphorylation of β-catenin and influence the expression of junction molecules.LP17 treatment effectively decreased the expression of SYK and further attenuated β-catenin phosphorylation.
However, SYK ACT CRISPR significantly reversed the effect of LP17 on the phosphorylation of β-catenin and the expression of junction molecules, which led to a downregulation of ZO-1 and Claudin-5 Furthermore, we also evaluate the effect of the TREM-1 activator on the expression of junction molecules using Anti-TREM-1 mAb before ICH.We found that activation of the TREM-1 receptor decreased the level of junction molecules via the SYK/β-catenin signaling pathway.Meanwhile, a selective inhibitor of SYK, called R406, was injected after ICH to reverse the signaling pathway.As expected, the phosphorylation of β-catenin and the expression of junction molecules were reversed when compared to the Anti-TREM-1 mAb group (Figure 6).

| DISCUSS ION
In the present study, the following novel findings were made:  The blood-brain barrier is mainly composed of cerebral endothelial cells and interendothelial tight junctions.Under physiological conditions, the tightness of junction molecules contributes to a very low permeability of BBB unless some lipophilic or transported compounds, which provide the central nervous system (CNS) a highly controlled and healthy microenvironment. 19Numbers of studies verify that disrupted BBB integrity and increased permeability subsequently are critical aspects inducing secondary brain injury after ICH. 20BBB injury will result in the formation of brain edema, the neuroinflammation caused by leukocyte facilitation, and the entry of some harmful molecules. 21Thus, maintaining BBB integrity and permeability is a potential target for ICH.
Triggering receptor expressed on myeloid cells 1 usually expresses on the majority of innate immune cells, and it will amplify the inflammatory response when activated.Our previous study suggested that the TREM-1 receptor activated rapidly on the surface of microglia after ICH and participated in the polarization of microglia.Furthermore, inhibition of TREM-1 with LP17 could improve neurobehaviour outcomes and brain edema.Brain edema is related to the broken blood-brain barrier directly.To further investigate the role of TREM-1 in the blood-brain barrier integrity, we conducted this study.As a result, we found TREM-1 expressed on the endothelial cells in a collagenase-induced ICH mouse and modulated the BBB permeability via regulating the expression of interendothelial adhesion molecules.Consistent with our result, the level of TREM-1 expression was increased in microglia and endothelial cells in subarachnoid hemorrhage rats.It may aggravate early brain injury by interacting with the TLR4 pathway. 22wever, the relation between the TREM-1 receptor and BBB permeability had not been dissected.In our study, EB dye extravasation assay and IgG staining showed an increasing BBB permeability surrounding hematoma in ICH mice as reported before. 23
ICH. Compared with ICH mice, 1.0 μg/g LP17 treated mice had higher scores in the modified Garcia test (12.7 ± 0.8 vs. 10.8 ± 1.3, p < 0.05) but no improvement in the other two tests (Figure2A-C).And similar results of neurobehavioral tests were found at 72 h after ICH (14.8 ± 0.8 vs. 12.8 ± 0.8, p < 0.05, Figure 2E-G).Furthermore, delayed administration of LP17 also improved neurobehavioral outcomes in modified Garcia test at 24 and 72 h after ICH (12.3 ± 0.5 vs. F I G U R E 2 LP17 improved neurobehavioral deficits and decreased brain edema at 24 and 72 h after ICH.(A) modified Garcia test, (B) limb placement test and (C) corner turn test 24 h after ICH.(D) Analysis of brain water content in different brain regions at 24 h after ICH.(E-H) Result of neurobehavioral test and brain water content at 72 h after ICH.The error bars represent the mean ± SD. *p < 0.05 versus sham, # p < 0.05 versus ICH.N = 6 per group.BG, basal ganglia; Cont, contralateral; CX, cortex; Ispi, ipsilateral.

F I G U R E 3
Inhibition of TREM-1 reduced BBB leakage 72 h after ICH.(A) Pictures of brain slices 2 mm thickness with extravasated EB dyes in different groups.(B) Statistical analysis of EB extravasation.The error bars represent the mean ± SD, *p < 0.05 versus sham, # p < 0.05 versus ICH + Control peptide.N = 6 per group.(C) Immunohistochemistry for perivascular IgG staining surrounding hematoma.

( 1 )
TREM-1 receptor activated rapidly in collagenase-induced ICH, F I G U R E 4 The effect of LP17 on the long-term neurobehavioral function and neuron survival after ICH.(A, B) Rotarod test before operation and every week after ICH.(C, D) Swimming distance and escape latency of Morris water maze test on days 23-27 after ICH.(E) Representative heat map in probe trial at day 28.The white circles indicate the positions of the probe platform.(F) Probe quadrant duration.(G) Representative Nissl staining figures in difference regions of hippocampus including CA1, CA3 and DG.(H-J) Quantifications of survival neuron in CA1 (H), CA3 (I) and DG (J).The error bars represent the mean ± SD, *p < 0.05 versus sham, # p < 0.05 versus ICH, N = 8 per group.

| 3235 XIE
et al.which is similar to our previous study11 and is mainly expressed on the endotheliocytes in the vascular endothelial unit.(2) A selective inhibitor of TREM-1, called LP17, decreased brain edema and improved neurobehavioral outcomes at 24 and 72 h after ICH, with a similar result in delayed LP17 treatment.(3) TREM-1 receptor modulated the blood-brain barrier integrity after ICH via SYK/ βcatenin signaling.
LP17 could reduce the leakage of dye and IgG from microvascular in the brain via inhibiting TREM-1 activation.Our study indicated that TREM-1 activated in the endothelial cells and deteriorated BBB F I G U R E 5 SYK activation CRISPR abolished the effect of LP17 at 24 h after ICH.(A) Representative Western blot bands.(B-G) Quantitative analyses of TREM-1, SYK, β-Catenin, pβ-Catenin, ZO-1 and Claudin 5 in the ipsilateral hemisphere at 24 h after ICH.The error bars represent the mean ± SD. *p < 0.05 versus sham, # p < 0.05 versus ICH, @ p < 0.05 versus LP17, One-way ANOVA, Tukey's test.N = 6 per group.