Blocking PSD-93-CX3CL1 Interaction Promotes the Phenotypic Transformation of Microglia During Acute Ischemic Stroke

Background Postsynaptic density 93 (PSD-93) plays an important role in ischemic brain injury by mediating neurotoxicity and neuroinammation. Different photypes of microglia perform an important role in ischemic crerbral injury and repair. Blocking the combination of PSD-93 and CX3C chemokine ligand 1 (CX3CL1) is benecial in acute ischemic stroke, but the underlying mechanism remains unclear. Methods Middle cerebral artery occlusion (MCAO) model was established in male C57BL/6 mice. The peptide Tat-CX3CL1 (357-395aa) which disturbing the interaction of PSD-93 and CX3CL1 was used in this study to explore the mechanism of its neuroprotective effect. The production and secretion of cytokines associated with M1 and M2 type of microglia was detected by PCR and ELISA, respectively. Neurologic damage was evaluated by behavior, triphenyl tetrazolium chloride staining, and brain water content. MBP and SMI32 double immunostaining were used to detect white matter injury and double staining for Iba1 and CD68 to assess M1 type microglia polarization. artery necrosis factor-alpha; cerebral artery occlusion model; neurological function


Background
Ischemic stroke induces neurotoxicity and neuroin ammation, is a leading cause of death and disability [1][2][3][4]. In ammatory responses induced by microglia polarization exacerbate cerebral infarction [5]. Microglial heterogeneity is of paramount importance for ischemic brain injury, and the polarization of microglia toward M2 phenotype was protective for ischemic stroke [6]. However, the polarization of microglia is highly dependent on environmental signals during cerebral ischemiareperfusion [7,8]. Therefore, intervention to promote the microglia to maintain M2 phenotype is a good strategy to achieve functional recovery after stroke.
Our previous study showed that PSD-93 binds directly to 670-685 amino acid sequence of GTPaseactivating protein for Ras (SynGAP) and promotes SynGAP ubiquitination in ischemic brain injury [15]. Knockout of PSD-93 improves neurological de cit by promoting the expression of anti-in ammatory cytokines and inhibiting pro-in ammatory cytokines [16]. Furthermore, we found that PSD-93 interacted with CX3CL1 to mediate neuron-microglia crosstalk and induce neuroin ammation [17]. Using yeast two hybrid and co-immunoprecipitation assay, we identi ed the binding sites and constructed a small peptide Tat-CX3CL1 (357-395aa) to disturb the combination of PSD-93 and CX3CL1 and attenuate cerebral infarct volume [17], but the underlying mechanism remains elusive.
In this study, we showed that Tat-CX3CL1 (357-395aa) peptide improved functional recovery after ischemic stroke by promoting M2 type microglia polarization. Delivery of Tat-CX3CL1 (357-395aa) inhibited the production of M1 type proin ammatory mediators, facilitated M2 type anti-in ammatory cytokines production and improved the integrity of blood-brain barrier (BBB) after stroke. Furthermore, Tat-CX3CL1 (357-395aa) reduced cerebral infarct volume and improved long-term cognitive function after stroke.

Antibodies and Reagents
The sequence of peptide Tat Animals C57BL/6 mice (male, 22-26 g weight) were purchased from Pengyue Company (Jinan, Shandong, China), and kept in a 12 h/12 h light-dark cycle with ad libitum food and water. The protocols were approved by Animal Care Committee of Xuzhou Medical University (No. 201702w012), and animal experiments were carried out in accordance with ARRIVE guidelines and the National Institutes of Health guide for the care and use of laboratory animals. Only male mice were used to exclude the effects of estrogen on ischemic injury and stress. Middle cerebral artery occlusion (MCAO) model was established in experimental groups following previous protocol [18]. In the sham operation group, mice received the same surgical procedures without occluding the carotid arteries.
Mice were euthanized by cervical dislocation and the mortality rate after surgery was 8.05%. Among 261 mice used in this study, 26 mice were excluded because of the failure of ischemia induction (5 mice), death (5 mice), cerebral hemorrhage (8 mice), and consciousness disturbance (8 mice).
Tat-CX3CL1 (357-395aa) was diluted with DMSO and administrated into the mice by intracerebroventricular injection. Mice received peptide or DMSO as control before MCAO and at 1d, 2d, and 3d postsurgery (see in supplemental 1). The right cerebral ventricle was selected to inject the peptide or DMSO (from the bregma: anteroposterior-1 mm; lateral 1 mm; depth 2 mm) [17].
Triphenyl tetrazolium chloride (TTC) staining TTC staining was performed as described previously [15] . Brie y, on the 7th day after surgery, the mice were anesthetized and decapitated. Brain slices were stained with 2% TTC (Catalog No. BCBW4269, Sigma, USA) for 15 min at 37°C in the dark. The normal tissue was stained red and the infarcted tissue was stained white. Infarct volume ratio (%)= V1/V2×100%, V1= ∑S1×d,V2= ∑S2×D (S1: infarct area of each section; S2: total area of each section; d: thickness of each piece was 1 mm).

Quantitative PCR
Total RNA was extracted using Trizol (Invitrogen, Carlsbad, CA, USA) and reverse transcribed to cDNA using cDNA synthesis kit (TAKARA, Japan). The reaction conditions were 95 °C for 10 min, followed by 40 cycles at 95 °C for 15 s and 60 °C for 60 s. The primers were purchased from SANGON Biotech (Shanghai, China) ( Table 1). The expression of target genes was calculated using 2 −ΔΔCT method, with glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as internal control.

Immuno uorescent assay
The brains were removed from mice and xed in 4% paraformaldehyde at 4℃ for 24 h. Brain tissues were treated with 30% sucrose solution for 72 h for dehydration and cut into slices (20 μm thick).
Sections were blocked with 5% BSA solution with 0.3% Triton X-100 in PBS (PBST) (Solarbio, Beijing, China) at room temperature for 1 h and then incubated with primary antibodies at 4℃ overnight. After washing with PBST, the sections were incubated with speci c uorescent secondary antibodies for 1 h at room temperature. The sections were observed under confocal microscope.
Evaluation of brain water content Brain water content was measured with wet-dry weight method [20]. Mice were euthanized and the brains were removed carefully and quickly. The wet weight (WW) was measured after removing olfactory bulbs, cerebellum and pons. Subsequently, dry weight (DW) was measured after the brains were put in an oven at 110℃ for 6 h. Brain water content was evaluated by formula: WC = (WW−DW)/WW×100%.

Modi ed Neurological function score (mNSS)
Based on motor, sensory, re ex, and balance tests, neurological function was evaluated at 1 d, 3 d, 5 d and 1 w after stroke with modi ed Neurological Severity Scores (mNSS) as previously described [15]. Total mNSS score were 18 points, and the higher the score, the more severe neurological impairment.

Morris water maze test
Morris water maze test was performed to assess cognition function including learning and memory ability as previously described [18]. Brie y, mice were trained on three trials one day for three consecutive days before MCAO, and the experiments were carried out 31-34 d after MCAO surgery. Each mouse was trained 4 times a day with an interval of 10-15 min for 3 consecutive days. The time spent to reach the platform was recorded. The memory test was performed on 34 day to record the time of mice in the target quadrant (the quadrant where the platform was originally placed), the number of times of entering quadrant platform and total distance of swimming in 1 min.

Open eld test.
Open eld test was performed to evaluate the state of depression 28 days after stroke as previously described [21]. Brie y, the equipment was mainly composed of a square open box of 50cm×50cm× 50cm and was divided into peripheral area and central area. Elevated plus maze Elevated plus maze test was performed 29 days after stroke to evaluate anti-anxiety behavior by using the contradictory tendency of mice to explore new environment and the fear of the open arm hanging high as previously described [22]. Brie y, the elevated cross maze consisted of two open arms with 50cm long and 10cm wide and closed arms. The four arms were connected by a central platform with a 10cm×10cm open part. The mice were placed on one side of the maze with open arms at the beginning of the experiment, and each animal was placed in the same position thereafter. The number of entries and the time spent in each arm were recorded for 5 min.
ELISA Soluble CX3CL1 was detected by using ELISA kit (R&D Systems, Stillwater, MN, USA) following the manufacturer's instructions.

Statistical analysis
Data were shown as mean ± standard error of the mean (SEM) and analyzed by using GraphPad Prism software 8.1.0 (La Jolla, CA, USA). Data were tested for normal distribution using Kolmogorov-Smirnov test. Continuous variables with normal distributions were analyzed with Student's t test, non-normal distributions data were analyzed with Mann-Whitney test. One-way ANOVA was applied to compare the differences among multiple groups, followed by post hoc Bonferroni test for pairwise comparison. Table 1
To explore the potential function of Tat-CX3CL1 (357-395aa) in ischemic stroke, we rst detected the expression of M1/M2 type in ammatory factors at different time points after ischemia/reperfusion. The results showed that TNF-α (microglia M1 type in ammatory factors) was increased at 6 h after stroke and peaked at 24 h (Supplemental Fig.2B), while the levels of other cytokines including iNOS and IL-1β also peaked at 24 h after stroke (Supplemental Fig.2A, C). However, the levels of M2 type in ammatory factors (CD-163, IL-10, and VEGF) were decreased at 3 h after stroke and kept stable to 24 h (Supplemental Fig. 2D-F). CD163 and IL-10 levels were increased from 48 h after stroke (Supplemental Fig.2D, E). Therefore, we selected 24 h time point to investigate microglia polarization.
PCR analysis of M1 and M2 type cytokines production showed that Tat-CX3CL1 (357-395aa) inhibited the production of proin ammatory factors (iNOS, TNF-α, and IL-1β) related to M1 type ( Fig. 2A-C), and increased the level of anti-in ammatory factors (IL-10, CD163, and VEGF) related to M2 type at 24 h after stroke ( Fig. 2D-F). Immuno uorescence assay of M1-type microglia (Iba1 and CD68 double staining) in the cortex and striatum around the infarct area after 24 h reperfusion showed that the number of M1 microglia decreased signi cantly in Tat-CX3CL1 (357-395aa) group not only in cortex region but also in striatum region (Fig. 3A-D). These data suggest the Tat-CX3CL1 (357-395aa) may affect the microglia M1/M2 phenotype switch during ischemic stroke.
Tat-CX3CL1 (357-395aa) promoted functional recovery of cognitive dysfunction after stroke by improving the integrity of myelinated bers.
White matter injury after stroke is associated with cognitive de cits, neuroin ammation, demyelination, and the degeneration of axons [32][33][34]. To explore whether Tat-CX3CL1 (357-395aa) could improve cognitive dysfunction, we performed Morris water maze test and found that Tat-CX3CL1 (357-395aa) reduced the escape latency and extended the time spent in the target quadrant, but did not affect swimming speed (Fig. 6A-D). Next, we evaluated whether Tat-CX3CL1 (357-395aa) could improve white matter integrity in stroke. Dual staining for SMI32 (a marker of demyelinated axons) and myelin basic protein (MBP, a major myelin protein) showed the lesion in white matter (Fig. 5A, B) was alleviated. The immuno uorescence intensity of MBP staining was decreased in corpus callosum (CC) and cortex regions at 35 d after MCAO (Fig 5C, 5D). However, Tat-CX3CL1 (357-395aa) increased the loss of myelin protein signi cantly. SMI32/MBP ratio was used to analyze the repair of myelin sheath and white matter.
The results showed that the SMI32/MBP ratio in corpus callosum (CC) and cortex regions was increased signi cantly in I/R and DMSO groups compared to sham group, but was decreased in Tat-CX3CL1 (357-395aa) group compared to I/R and MDSO group (Fig. 5E, F).
Post-stroke depression and anxiety exacerbate cognitive dysfunction, we suspect the underlying mechanism might be related to microglia polarization. We observed the anxiety-like behavior and

Discussion
In present study, we demonstrated protective effects of Tat-CX3CL1 (357-395aa) on stroke by blocking the interaction of PSD-93 and CX3CL1 and reducing the production of soluble CX3CL1, resulting in inhibiting the communication between neuron and microglia. We found that the peaks for the expression of pro-in ammatory cytokines (M1 like) and anti-in ammatory cytokines (M2 like) were different. Thus we proposed that the peptide Tat-CX3CL1 (357-395aa) reduced pro-in ammatory cytokines secretion while promoted anti-in ammatory cytokines expression in acute ischemia-reperfusion due to M1 and M2 phenotypic polarization shift (Fig. 8). Furthermore, the peptide Tat-CX3CL1 (357-395aa) diminished neurological impairment and improved long-term cognitive dysfunction after stroke. Collectively, these data support the bene cial effects of Tat-CX3CL1 (357-395aa) and the peptide may be a therapeutic agent for ischemic stroke.
Accumulating evidences showed that microglia are polarized into different states within hours following the onset of stroke. Differential polarization of microglia including classic pro-in ammatory type (M1like) and alternative protective type (M2-like) is activated at different stage, and exerts detrimental or bene cial potential role [10,36,37]. However, molecular mechanism underlying microglia polarization during stroke remains to be elucidated.
In this study, we reported M1/M2-like microglia activation at different stage after stroke. M1-like microglia was activated at 6 h after stoke and peaked at 24 h and then persisted several days, consistent with the peak expression of sCX3CL1 [17], which indicates that splicing into sCX3CL1 promotes M1-like microglia activation. Conversely, M2-like microglia was suppressed at 24 h following stroke. Meanwhile, we found that Tat-CX3CL1 (357-395aa) facilitated microglia polarization from M1 to M2 phenotype by inhibiting M1-phenotype cytokines (iNOS, TNF-α, and IL-1β) and promoting M2-phenotype cytokines (CD-163, IL-10, and VEGF) at 24 h after stroke.
Previous study revealed that CX3CL1-CX3CR1 signaling regulated synaptic plasticity and cognitive function [39][40][41][42]. Ischemic stroke impacts not only gray matter but also white matter and induces cognitive de cits in memory and learning. Furthermore, the incidence of post-stroke anxiety and depression is about 36.7% within 2 weeks after stroke [43,44]. In this study, we showed that Tat-CX3CL1 (357-395aa) promoted white matter repair and cognitive improvement. Additionally, Tat-CX3CL1 (357-395aa) showed bene cial potential role on post-stroke anxiety and depression. These results suggest that the inhibition of microglial polarization in acute stage of ischemic infarction will be bene cial for the rehabilitation of nerve function in the later stage, and Tat-CX3CL1 (357-395aa) is an attractive agent for stroke therapy.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.