OTULIN is a New Target for EA Treatment in Alleviating Brain Injury and The Activation of Glial Cells By Depressing NF-κB Signalling Pathway in Acute Ischaemic Stroke Rats

Objective: Ovarian tumour domain deubiquitinase with linear linkage specicity (OTULIN), a potent negative regulator of nuclear factor-κB (NF-κB) signalling pathway, exerted strong neuroprotective role following acute ischemic stroke. Electroacupuncture (EA) is an effective adjuvant treatment for reducing brain injury and neuroinammtion through inhibiting the nuclear translocation of NF-κB p65, while the underlying mechanism remains unclear. This study investigated whether OTULIN was necessary for EA to mitigate brain injury and the activation of glial cells, following by a transient middle cerebral artery occlusion (tMCAO) model in rats. Methods: Acute ischaemic stroke model was established by performing tMCAO surgery in Sprague-Dawley (SD) rats. EA was performed once daily at “Baihui (GV 20)”, “Hegu (LI 4)”, and “Taichong (LR 3)” acupoints. EA’s effect on the spatiotemporal expression of OTULIN in the ischaemic penumbra of cerebral cortex was detected within 7 d after reperfusion. Following gene interference to silence OTULIN expression, the effects of OTULIN on EA neurological decits, cerebral infarct volume, neuronal damage, the activation of microglia and astrocytes, the contents of tumor necrosis factor alpha (TNF-α), interleukin-1 beta (IL-1β) and interleukin-6 (IL-6), and the expression of p-IκBa, IκBa and nucleus/cytoplasm NF-κB p65 protein were assessed. Results: p65 nuclear translocation. Furthermore, EA partly inhibited the transformation of microglia and astrocytes to from resting states to activated states, and reduced the secretion of TNF-α, IL-1β and IL-6, but this preventive effects were reversed after silencing OTULIN expression. Conclusions: OTULIN provides a new potential therapeutic target for EA to alleviate acute ischaemic stroke induced brain injury and the activation of glial cells, which is related to depression of NF-κB signalling pathway.


Introduction
Ischaemic stroke induced cell injury initiates the release of danger signals that activate the immune system in central nervous system (CNS) (Corps, Roth et al., 2015). The subsequent sterile in ammatory response mainly involves the innate immune system with activation of resident immune cells of the central nervous system (CNS) and a rapid in ltration of peripheral immune cells (Lin, Wang et  . Nuclear factor-kappa B (NF-κB) signalling pathway has been considered a "master regulator", as it dominates the initiation and development of brain injury induced by cerebral ischaemia(Kaushal and Schlichter, 2008; Liu, . Physically, the p65/p50 subunit dimer, the main form of NF-κB, is mainly located in the cytoplasm and combines with NF-κB inhibitor protein α (IκBα) in a resting form. Following activation by upstream kinases after ischaemia onset, IκBα is degraded and phosphorylated, subsequently leading to the nuclear translocation of the p65 to bind NF-κB-responsive genes (Mattson and Camandola, 2001).
Inhibiting NF-κB activation is an effective strategy for reducing acute ischaemic stroke induced brain injury and in ammation (Li,  Currently, recombinant tissue plasminogen activator (r-tPA) remains the only FDA-approved pharmacological treatment for stroke patients. However, few patients could bene t from it due to the rigid therapeutic time window and severe side effects. Thus, complementary and/or alternative medicine is urgently needed to treat ischaemic stroke worldwide. Electroacupuncture (EA) has been identi ed as a general complementary therapy for post-stroke patients in China, as it is more readily controlled, standardized, and objectively measurable (Cai, Zhang et  . OTULIN is expressed in immune cells, such as T cells, B cells, and natural killer cells, and prominently in dendritic cells and macrophages; moreover, it has been proven to be critical in the NF-κΒ-dependent in ammatory signalling pathway (Damgaard, Walker et al., 2016). OTULIN was also reported to protect liver against cell death, in ammation, brosis, and cancer (Damgaard, Jolin et al., 2020;Verboom, Martens et al., 2020) and is involved in regulating autophagy initiation and autophagosomes (Chu, Kang et al., 2020). However, the role of OTULIN in CNS is poorly understood, and its role in cerebral ischaemia is limited to a paper we published previously . OTULIN was enriched in activated microglia, and OTULIN overexpression ameliorated microglia-mediated neuroin ammation by regulating NF-κB activation in focal cerebral ischaemia/reperfusion rats .
Our previous studies demonstrated that neuronal zinc nger protein A20 and cylindromatosis (CYLD), which are important deubiquitinases that negatively regulate the NF-κB pathway, were required for EA to alleviate excessive in ammatory reaction in acute ischaemic rats (Zhan, Qin et al., 2016;Jiang, Luo et al., 2017). As the deubiquitination activity of OTULIN is more critical than A20 and CYLD in terms of regulating NF-κB pathway (Keusekotten, Elliott et al., 2013;Aksentijevich and Zhou, 2017), we speculated that EA may also alleviate brain injury and glial activation neuroin ammation following cerebral ischaemia by regulating OTULIN. Therefore, this study aimed to characterize the spatiotemporal expression of OTULIN, the effect of EA on OTULIN expression, and the possible mechanism by which EA regulates OTULIN-mediated neuroprotection.

Animals
A total of 313 Sprague-Dawley (SD) rats (250-300 g) were purchased from the Experimental Animal Center of Chongqing Medical University. All rats were housed under a 12/12 h light and dark cycle condition and allowed free access to food and water.
Establishment of the transient middle cerebral artery occlusion (tMCAO) model The model was performed according to a method described previously (Longa, Weinstein et al., 1989;Qin, Luo et al., 2013). Brie y, rats were initially anaesthetized with 3.5% chloral hydrate (1 ml/100 g) intraperitoneally. After a midline neck incision, the bifurcation of the right common carotid artery (CCA) was exposed, and a heparin-dampened nylon mono lament with a rounded tip was advanced to block the origin of the right middle cerebral artery (MCA). Two hours after MCA occlusion, rats were reanaesthetized, and the mono lament was gently withdrawn to restore blood ow. All the same surgical procedures except the insertion of a mono lament to the origin of the MCA were performed in the Sham group. Successful tMCAO and reperfusion was determined by a decrease in the regional cerebral blood ow to 20% and recovery to > 80% of the baseline monitored by a laser-Doppler owmeter (PeriFlux 5000, Perimed AB, Sweden). Throughout the operation, the body temperature was continuously maintained at 38±0.5°C by a thermostatically controlled infrared lamp (FHC, Bowdoinham, ME, USA). After rats recovered from anaesthesia, the Longa score (Longa, Weinstein et al., 1989) was used to assess neurological de cits. Rats scoring 2 and 3 were included as previously described . Electroacupuncture (EA) treatment EA treatment was performed as before (Qin, Luo et al., 2013;Xie, Gao et al., 2016). The acupoints "Baihui (GV 20)", "Hegu (LI 4)", and "Taichong (LR 3)" on the left side of the rats were chosen (as shown in Fig.   1A). "Baihui (GV 20)" is located at the intersection of the sagittal midline and the line connecting the two ears, "Hegu (LI 4)" is at the second metacarpal midpoint of the radial side, and "Taichong (LR 3)" is at the second toe tibial collateral at the rear of the phalanx. The needle was inserted into each acupoint at an angle of 15°-45° with a 0.5 inch depth and then connected with the EA instrument (Model no. 227033; Beijing Jinggong Ltd., China). The acupoints were stimulated with an intensity of 1 mA and a frequency of 2/20 Hz for 30 min as the stimulation parameters can induce visible muscle contraction. The initial EA treatment was conducted once the reperfusion was completed and performed once daily thereafter.
Seven days before tMCAO, SD rats received intracerebroventricular (i.c.v) injection of LV-shOTULIN or LV-Scramble as before . Brie y, rats were anaesthetized and xed in a stereotaxic apparatus (Stoelting, USA). LV-shOTULIN or LV-Scramble was injected into the right i.c.v. (bregma coordinates: 1.3 mm lateral, 1.5 mm posterior, and 3.8 mm under the dural surface) by a 10-µl Hamilton syringe (Hamilton Co., Reno, NV, USA) at a rate of 0.5 µl/min, and the needle was left in place for 5 min to prevent back ow. The animals were carefully monitored until recovery from anaesthesia. To verify the effect of gene silencing on OTULIN expression, OTULIN mRNA and protein were detected by RT-qPCR and Western blotting.

Neurobehavioural assessment
Assessments of neurobehavioural de cits were performed at 72 h with the modi ed Neurological Severity Score (mNSS) (Chen, Li et al., 2001), rotarod test (Linden, Fassotte et al., 2014) and inclined board test (Zhang, Wei et al., 2006) with some modi cations by an independent investigator blinded to the experiment.
mNSS is a composite test including motor (muscle status and abnormal movement), sensory (visual, tactile and proprioceptive), re ex (pinna, corneal and startle), and balance assessments. It is graded on a scale of 0-18, and the higher the score is, the more severe the injury.
The rotarod test was used to evaluate motor coordination by an accelerating rotarod apparatus (YLS-4C, Shanghai Jinggong Industry Co., Ltd, China). Brie y, each rat was placed on the rotating rod, which accelerates from 4 to 40 rpm per min and went through three trials with 10 min breaks. A trial ended if the rat fell from the rotating rod or clung to the rod without walking for two consecutive rotations, and the latency to end was recorded. Before tMCAO, all rats were trained to stay on the rotating rod at a speed of 4 rpm three times per day for three days. The pre-tMCAO data were recorded as the internal control. The rotarod test results are presented as the percentage of the mean latency time post-tMCAO compared with that pre-tMCAO. The lower the score was, the more severe the injury.
An inclined board test was used to assess balance and coordination. Animals were placed on a board (50 ×30 cm) covered by copper wire mesh (0.2 mm) and stabilized. The board was gradually inclined from the horizontal to the vertical plane. The holding angle at which the animal fell from the board was recorded. The test was repeated and measured three times with two min intervals. The scores were recorded according to the holding angle as follows: 0, >70°; 1, 65-69°; 2, 60-64.9°; 3, 55-59.9°; 4, <55°.
The higher the score was, the more severe the injury.

TTC staining
The cerebral infarct volume was detected by 2,3,5-triphenyltetrazolium chloride (TTC, Sigma-Aldrich, USA) staining as described previously . Rats were euthanized, and brains were quickly removed and frozen for 20 min at -20°C. Then, the brains were coronally sliced into ve 2-mm thick sections, stained with 2% TTC at 37°C for 10 min, and then xed with 10% formaldehyde. The stained sections were photographed by camera (Canon IXUS, Canon Co., Japan) and analysed by Image-Pro Plus (version 6.0, Media Cybernetics Co., USA). The infarct volume was calculated using the following formula: percentage hemisphere lesion volume=[total infarct volume-(right hemisphere volume-left hemisphere volume)]/left hemisphere volume×100%.

Cresyl violet staining
Frozen coronal brain sections were used for cresyl violet staining to label the ischaemic penumbra of the cerebral cortex in rats as previously described (Marchese, Occhieppo et al., 2020). The brain sections were dehydrated through descending grades of ethanol (95%, 90%, 80%, 70%, and 50% for 5 min) and rinsed with water. The sections were then immersed in 0.1% cresyl violet solution for 10 min, rinsed quickly in water, dehydrated though ascending grades of ethanol, immersed 3 times for 2 min each in xylene, and then cover-slipped.

Nissl staining
The frozen coronal brain sections were dehydrated through descending grades of ethanol (95%, 90%, 80%, 70%, and 50% for 5 min), immersed in 0.1% cresyl violet solution for 10 min, rinsed quickly in water, rehydrated though ascending grades of ethanol, immersed 3 times for 2 min each in xylene, and then cover-slipped. The sections were photographed by a digital camera.

Western blot
Rats were sacri ced, and the brain tissues in the ischaemic penumbra were rapidly dissected. The tissues were homogenized in RIPA lysis buffer (no. P0013B, Beyotime, Shanghai, China) supplemented with PMSF (Beyotime, Shanghai, China) and phosphatase inhibitors for the detection of phosphorylated proteins. After centrifugation, total protein was extracted from the supernatant. The nuclear and cytoplasmic proteins were extracted by a Nuclear and Cytoplasmic Protein Extraction Kit (no. AR0106, Boster, Beijing, China), and the protein concentrations were detected by a BAC kit (Beyotime, Shanghai, China). SDS-PAGE was used to separate proteins, and PVDF membranes (Millipore Co., USA) were used to transfer proteins. After blocking in 5% non-fat milk for 2 h, the PVDF membranes were incubated with the following primary antibodies at 4°C: anti-OTULIN polyclonal rabbit antibody (no. 14127, Cell Signaling Technology, USA, 1:1000), anti-NF-κB p65 rabbit antibody (no. 8242, Cell Signaling Technology, USA, 1:1000), anti-IκBα rabbit antibody (no. 4812, Cell Signaling Technology, USA, 1:1000), anti-phospho-IκBα (Ser32) rabbit antibody (no. 2859, Cell Signaling Technology, USA, 1:500), and anti-β-actin rabbit antibody (no. 4970, Cell Signaling Technology, USA, 1:1000). The next day, the membranes were incubated with a speci c horseradish peroxidase-conjugated secondary antibody for 1 h at 37°C. A gel imaging instrument (Vilber Lourmat fusion FX 7 Spectra, France) and analysis software (FUSION-CAPT, France) were applied to scan and analyse the immunoblots.

Real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) analysis
Total tissue RNA in the ischaemic penumbra of the cerebral cortex was isolated by TRIzol (Takara Biotechnology, Japan). The mRNA was used as a template to synthesize cDNA by using the PrimeScript™ RT reagent kit with gDNA Eraser (TaKaRa) at 42°C for 2 min. The synthesized cDNA was used for RT-qPCR on an iQ5 Gradient Real-Time PCR detection system (Bio-Rad Co., USA) with SYBR Green (SYBR Premix Ex Taq™ II, TaKaRa). The cycling conditions were as follows: 10 min at 95°C followed by 40 cycles of 5 s at 95°C and 30 s at 60°C. The melting curve was used to analyse the gene speci city of OTULIN and housekeeping gene β-Actin. Relative quanti cation was carried out with the 2 −∆∆Ct method. The expression of each targeted gene was normalized to the expression of β-actin in the same sample. All primer sequences used in the study were as follows: for OTULIN, forward primer: TGTGGCTCCTGAAATGGATATTATG, reverse primer: CTCTGACAGGGATGTTATAGTGCCG; for β-Actin, forward primer: TGTCACCAACTGGGACGATA, reverse primer: GGGGTGTTGAAGGTCTCAAA.
Iba-1 is a commonly used microglial marker. In this study, the morphology of reactive microglia in the ischaemic penumbra of the cerebral cortex was determined by a classic method (Ito, Tanaka et al., 2001;Sawano, Watanabe et al., 2015). Brie y, according to the length of branches, thickness of branches and cell body size, Iba-1 + cells were categorized into the following three forms: 1) rami ed microglia, with a small soma, and very ne, long processes; 2) internal microglia, with a larger cell body and thicker processes than rami ed microglia; and 3) ameboid microglia, characterized by a large cell body without processes.
The morphological analysis of GFAP + astrocytes was mainly completed through a method described previously (Ramirez-Sanchez, Pires et al., 2018). Brie y, GFAP + cells with clearly visible nuclei and soma were selected, and the length of the longest cellular processes (the distance of the nucleus to the tip of the extension process) was measured as an indicator to assess the degree of hypertrophy of astrocytes according to a method described previously.

Enzyme-linked immunosorbent assay (ELISA)
The levels of tumor necrosis factor alpha (TNF-α), interleukin-1 beta (IL-1β) and interleukin-6 (IL-6) in brain tissues from the ischaemic penumbra of the cerebral cortex were detected using ELISA kits (no. EK0526 96T, EK0393 and EK0412 96T, BOSTER Co.) according to the manufacturer's instructions. The whole experimental owchart is shown in

Statistical analysis
The neurological scores were analysed using Kruskal-Wallis tests followed by post hoc Dunn's multiple comparison tests, and all other data were expressed as the mean±SEM. The expression of OTULIN mRNA and protein among the Sham, tMCAO and EA groups was compared using two-way ANOVA with Bonferroni post hoc tests. All other quantitative data were analysed using one-way ANOVA followed by Tukey's post hoc test for multiple comparisons. Statistical analysis was performed using SPSS 19.0. P<0.05 was considered statistically signi cant.

EA enhanced OTULIN expression following focal cerebral ischaemia/reperfusion injury
To detect the effect of EA on the time course of OTULIN expression, OTULIN mRNA and protein in the ischaemic penumbra of the cerebral cortex ( Fig.2A) from 2 h to 7 d after reperfusion were measured by RT-qPCR and Western blot, respectively. Rats were randomly divided into the Sham, tMCAO and tMCAO+EA groups. Compared with those in the Sham group, OTULIN mRNA and protein in the tMCAO group decreased obviously as early as 2 h and then increased gradually with a peak at 48 h. The concentrations of OTULIN mRNA and protein in the tMCAO+EA group were increased signi cantly compared with those in the tMCAO group at the indicated time points except at 2 h and 6 h (Fig. 2B-D).
Next, the immune response of OTULIN protein in the ischaemic penumbra of the cerebral cortex was determined by immuno uorescence at 48 h. The number of OTULIN + cells in the tMCAO group increased obviously compared with that in the Sham group (P<0.01), and the count in the tMCAO+EA group was further increased compared with that in the tMCAO group (P<0.01). In addition, the OTULIN protein was mainly located in the cytoplasm, and little OTULIN was expressed in the nucleus in all groups. (Fig. 2E, F).
Enhanced OTULIN was mainly expressed in microglia and neurons in focal cerebral ischaemia/reperfusion rats Next, we explored the main cellular localization of OTULIN protein in the ischaemic penumbra of the cerebral cortex. Rats were randomly divided into the Sham, tMCAO and tMCAO+EA groups. Forty-eight hours, the peak time for OTULIN expression, was selected as the observation time point. OTULIN was costained with NeuN, Iba-1 and GFAP by immuno uorescence staining.
As shown in Fig. 3, in the Sham group, OTULIN was mainly located in NeuN + cells, and little OTULIN was detected in Iba-1 + cells. OTULIN was predominantly expressed in Iba-1 + cells, and little OTULIN was observed in reduced NeuN + cells in the tMCAO group. Moreover, OTULIN protein in the tMACO+EA group was upregulated in both NeuN + and Iba-1 + cells compared with that in the tMCAO group. No OTULIN expression was observed in GFAP + cells in any group.
OTULIN knockdown reversed EA's neuroprotective effect against ischaemic brain injury To explore whether enhanced OTULIN expression was required for EA to exert a neuroprotective effect, gene interference by LV-shOTULIN was applied to silence OTULIN expression. Rats were randomly divided into ve groups: Sham, tMCAO, tMCAO+EA, tMCAO+EA+LV-Scramble, and tMCAO+EA+LV-shOTULIN. Seventy-two hours after reperfusion, the levels of OTULIN mRNA and protein in the tMCAO+EA+LV-shOTULIN group were signi cantly decreased compared with those in the tMCAO+EA and tMCAO+EA+LVscramble groups (Fig. 4A-C).
The neurobehavioural assessment with the mNSS, inclined board test and rotarod test, and cerebral infarct volume detection were performed at 72 h. As shown in Fig. 4D-H, rats in the tMCAO group showed more severe neurobehavioural dysfunction than rats in the Sham group. EA obviously improved neurological function, as demonstrated by a lower mNSS score, higher holding angle and relative latency time than those in the tMCAO group. However, this improvement was weakened by LV-shOTULIN, as evidenced by a poor neurobehavioural improvement in the tMCAO+EA+LV-shOTULIN group. Moreover, as shown in Fig. 4G and H, no detectable cerebral infarct volume was found in the Sham group (Fig. 4G, H).
The infarct volume in the tMCAO+EA group was decreased compared with that in the tMCAO group. After silencing OTULIN, EA failed to reduce the infarct volume.
Enhanced OTULIN expression was necessary for EA to attenuate neuronal injury in focal cerebral ischaemia/reperfusion rats Since OTULIN was also detected in neurons, we sought to investigate the effects of OTULIN on EA's role in attenuating neuronal injury in the ischaemic penumbra of the cerebral cortex at 24 h. Rats were randomly divided into the following ve groups: Sham, tMCAO, tMCAO+EA, tMCAO+EA+LV-Scramble and tMCAO+EA+LV-shOTULIN groups. First, viable neurons were detected by Nissl staining. As shown in Fig.  5A and B, neurons were arranged regularly, and most cells exhibited a normal shape in the Sham group. In the tMCAO group, most neurons were shrunken, deeply stained and sparsely distributed with a cytosolic concentration, karyopyknosis and karyorrhexis. In the tMCAO+EA group, the morphology of neurons was improved, with many normal-shaped cells, and the number of viable neurons was increased compared with that in the tMCAO group (P<0.01). However, this improvement was obviously blocked by OTULIN silencing (P<0.01). In Fig. 5C-E, a large number of TUNEL + and TUNEL + NeuN + cells were observed in the tMCAO group, whereas few apoptotic cells were detected in the Sham group. EA treatment signi cantly decreased the proportion of TUNEL + and TUNEL + NeuN + cells compared with tMCAO (P<0.001), but the effect was inhibited by OTULIN silencing.
Next, we explored whether OTULIN participated in EA regulation of neuronal NF-κB p65 nuclear translocation. As shown in Fig. 5F, neuronal NF-κB p65 was mainly expressed in the nucleus in the tMCAO group. An increased concentration of NF-κB p65 protein was translocated to neuronal cytoplasm in the tMCAO+EA group compared with the tMCAO group, but this effect was reversed by OTULIN silencing.
Increased OTULIN expression was required for EA to inhibit the NF-κB signalling pathway in focal cerebral ischaemia/reperfusion rats To further investigate whether enhanced OTULIN was required for EA to inhibit the NF-κB signalling pathway, OTULIN, p-IκBα, IκBα, cytoplasm-p65 and nucleus-p65 proteins from brain tissues in the ischaemic penumbra of the cerebral cortex were detected 24 h after reperfusion. p-IκBα/IκBα represents the phosphorylation ratio of IκBα, and nuclear/cytoplasmic NF-κB p65 represents the nuclear translocation of NF-κB p65 protein. Rats were randomly divided into the tMCAO, tMCAO+EA, tMCAO+EA+LV-Scramble and tMCAO+EA+LV-shOTULIN groups. As shown in Fig. 6A and B, increased OTULIN protein content, a lower phosphorylation ratio of IκBα, and attenuated nuclear translocation of p65 protein were detected in the tMCAO+EA group compared with the tMCAO group. In the MCAO+EA+LV-shOTULIN group, little OTULIN was detected, and an elevated phosphorylation ratio of IκBα and increased nuclear translocation of NF-κB p65 were detected compared with those in the tMCAO+EA and tMCAO+EA+LV-Scramble groups.
In addition, immunohistochemical analysis was also applied to detect IκBα and NF-κB p65 proteins 24 h after reperfusion. As shown in Fig. 6C, EA obviously enhanced IκBα staining, and IκBα staining was obviously attenuated in the tMCAO+EA+LV-shOTULIN group compared with the tMCAO+EA group. NF-κB p65 protein was predominantly located in the nucleus in the tMCAO group but partly transferred to the cytoplasm in the tMCAO+EA group. In the tMCAO+EA+LV-shOTULIN group, NF-κB p65 staining increased in the nucleus and decreased obviously in the cytoplasm compared with the tMCAO+EA and tMCAO+EA+LV-Scramble groups.
OTULIN silencing reversed the depression of microglial activation by EA after focal cerebral ischaemia/reperfusion injury The transformation of microglial shape is a manifestation of microglia activation. Microglia located in the ischaemic penumbra of the cerebral cortex were mainly in the following three forms: rami ed cells (Fig. 7Aa), internal cells (Fig. 7Ab) and ameboid cells (Fig. 7Ac). To investigate the effect of OTULIN on EA's role in attenuating microglial activation, mean Iba-1 immuno uorescence intensity and the proportions of microglia with different forms were analysed quantitatively. All rats were randomly divided into the following ve groups: Sham, tMCAO, tMCAO+EA, tMCAO+EA+LV-Scramble and tMCAO+EA+LV-shOTULIN groups.
First, the Iba-1 mean immuno uorescence intensity in each group was analysed. Twenty-four hours and 72 h after reperfusion, as shown in Fig. 7, the mean Iba-1 immuno uorescence intensity was very low in the Sham group and increased sharply in the tMCAO group, but it was obviously reduced in the tMCAO+EA group. After silencing OTULIN, the Iba-1 mean immuno uorescence intensity in the tMCAO+EA+LV-shOTULIN group was increased signi cantly compared with that in the tMCAO+EA and tMCAO+EA+LV-Scramble groups.
Next, the effect of EA and OTULIN silencing on the cell shape of microglia in the ischaemic penumbra was analysed. As shown in Fig. 7B, D, the main type of microglia was internal shape in the tMCAO group at 24 h. In the tMCAO+EA group, the proportion of internal microglia was decreased, and the proportion of rami ed microglia was obviously increased compared with the tMCAO group. After silencing OTULIN, the proportion of rami ed microglia was signi cantly lower, while the proportion of internal microglia was signi cantly higher in the tMCAO+EA+LV-shOTULIN group than those in the tMCAO+EA and tMCAO+EA+LV-Scramble groups. At 72 h, activated microglia were mainly in ameboid form (Fig. 7E, G). In the tMCAO+EA group, the proportion of internal microglia was increased, and the proportion of ameboid microglia was obviously decreased compared with those in the tMCAO group. In the tMCAO+EA+LV-shOTULIN group, the proportion of internal microglia was obviously decreased, and the proportion of rami ed microglia was higher than those in the tMCAO+EA and tMCAO+EA+LV-Scramble groups.
To investigate the effect of OTULIN silencing on EA's role in depressing the secretion of pro-in ammatory cytokines, the contents of TNF-α, IL-1β and IL-6 were detected at 24 h and 72 h by ELISA. As shown in Fig.  7H-J, the levels of TNF-α, IL-1β, and IL-6 in the tMCAO group were signi cantly higher than those in the Sham group, and they decreased after EA treatment. After silencing OTULIN expression, the contents of TNF-α, IL-1β and IL-6 in the tMCAO+EA+LV-shOTULIN group were signi cantly higher than those in the tMCAO+EA and tMCAO+EA+LV-Scramble groups.
OTULIN reversed the depression of astrogliosis by EA in focal cerebral ischaemia/reperfusion rats Although OTULIN protein was not detected in astrocytes, we tried to explore whether the attenuated neuroin ammation could adversely affect the activation of astrocytes. Rats were randomly divided into the Sham, tMCAO, tMCAO+EA, tMCAO+EA+LV-Scramble and tMCAO+EA+LV-shOTULIN groups.
Immuno uorescence was used to observe GFAP + astrocytes in the ischaemic penumbra of the cerebral cortex at 24 h and 72 h. The mean immuno uorescence intensity of GFAP and the mean longest process of GFAP + cells were quantitatively analysed.
In the tMCAO group, the mean immuno uorescence intensity of GFAP and the mean longest process of GFAP + cells were increased signi cantly compared with the Sham and EA group at both 24 h and 72 h.
After silencing OTULIN, the mean immuno uorescence intensity of GFAP in the tMCAO+EA+LV-shOTULIN group was signi cantly higher than that in the tMCAO+EA group at 72 h, but not at 24 h, while the mean longest process of GFAP + cells was increased signi cantly compared with that in the tMCAO+EA group  (Damgaard, Walker et al., 2016). Mice that lacked OTULIN in myeloid cells exhibited enlarged lymphoid organs and liver, increased immune cell in ltration in liver, and increased cytokine levels in serum (Damgaard, Walker et al., 2016). We previously demonstrated that OTULIN overexpression ameliorated microglial activation and neuroin ammation by depressing the NF-κB pathway in cerebral ischaemia/reperfusion rats . In this study, rapidly decreased OTULIN was detected as early as 2 h following perfusion. It is speculated that hyperacute brain injury inhibits brain cells from synthesizing OTULIN. Subsequently, endogenous OTULIN expression increased gradually with a peak at 48 h, which is considered an endogenous protection, as we previously demonstrated that enhanced OTULIN exerted a neuroprotective role in ischaemic stroke.
In this study, for the rst time, OTULIN was detected to be located mainly in neurons and microglia in both rats in a normal physiological state and those with ischaemic stroke, but no OTULIN protein was detected in astrocytes. EA treatment could promote OTULIN gene transcription and protein synthesis in the brain, and enhanced OTULIN was mainly located expressed in neurons. Importantly, the neuroprotective role of EA was obviously reversed by silencing OTULIN, which indicated that OTULIN was involved in the antipreventive role of EA following acute ischemic stroke.
NF-κB signalling pathway is widely activated in brain cells after cerebral ischaemia, while its effect on neuronal injury is controversial. Some studies reported that neuronal conditional NF-κB knockout mice exhibited impaired synaptic transmission, spatial memory formation, and plasticity(O'Mahony, Raber et al., 2006), while other studies showed that NF-κB activation in neurons contributed to neuronal cell death in cerebral ischaemia (Zhang, Potrovita et al., 2005; Ridder and Schwaninger, 2009; Ingrassia, Lanzillotta et al., 2012). Sarnico et al (Sarnico, Lanzillotta et al., 2009) found that overexpression of p65 reduced neuron susceptibility to anoxia, and p65 silencing signi cantly reduced neuronal cell death in vitro.
Although the functional consequences of neuronal NF-κB and neuronal survival are still disputed, EA has been con rmed to decrease neuronal apoptosis and improve neurological function by inhibiting NF-κB activation in the whole brain. In this study, it was shown that alleviated neuronal injury and apoptosis by EA was accompanied by enhanced OTULIN expression and depressed activation of the neuronal NF-κB signalling pathway, while the above ndings were obviously reversed when OTULIN expression was inhibited. However, whether enhanced neuronal OTULIN is directly required for attenuating neuronal injury by EA is unclear and needs further exploration, as this effect may partly be attributed to alleviated neuroin ammation. ). However, little was known about EA's effect on microglial morphology following cerebral ischaemia until now. By quantitative analysis of microglia with different shapes using immuno uorescence, this study revealed that EA inhibited the transformation of microglia from the rami ed to internal state at 24 h and transformation of microglia from the internal to ameboid shape at 72 h. Importantly, these transformations were partly inhibited by OTULIN silencing, which suggests that OTULIN is involved in EA's role in inhibiting the transformation of microglia from resting state to active state. In line with our hypothesis, OTULIN silencing reversed the inhibitory effect of EA on the secretion of pro-in ammatory factors in the ischaemic cerebral cortex. EA effectively inhibits the activation of microglia and the secretion of in ammatory mediators by regulating OTULIN expression.
Astrocytes, the most abundant cells in the CNS, change to a reactive phenotype (so-called reactive astrogliosis) characterized by enhanced glial brillary acidic protein (GFAP) expression and cellular hypertrophy (Sofroniew, 2009;Pekny and Pekna, 2014;Rossi, 2015) in response to ischaemic stroke.
Numerous studies have proven that EA can inhibit astrocyte activation induced by acute cerebral ischaemia/reperfusion injury ( (Tang, Hsu et al., 2016), neuropathic pain (Liang, Qiu et al., 2016), and in ammatory pain (Liao, Hsieh et al., 2017). Although astrogliosis plays a dual role in ischaemic stroke, inhibition of excessive astrogliosis was considered bene cial , as reactive astrogliosis and glial scar formation are the main reasons for functional recovery di culty after cerebral ischaemia (Cregg, DePaul et al., 2014;Cheon, Cho et al., 2016;Liu and Chopp, 2016). It is well demonstrated that EA could inhibit astrogliosis (Tian, Tao et al., 2016;Zhan, Qin et al., 2016), and this is bene cial to the antiin ammatory effect of EA (Zhan, Qin et al., 2016). Microglia and astrocytes are the primary immune cell types that quickly respond to cerebral ischaemia, but their functions are increasingly recognized as being complex and interactive (sometimes even synergistic) (Liddelow, Guttenplan et al., 2017;Zhang, Zhao et al., 2018). Actually, the complex cell-cell interactions have hampered the mechanistic understanding of astrocyte reactivity. As OTULIN was required for EA to alleviate microglial activation and neuroin ammation, we wondered whether astrogliosis could be affected after silencing OTULIN.
Although no OTULIN was detected to be expressed in astrocytes, EA failed to depress astrocyte activation after silencing OTULIN. It was speculated to be related to attenuated microglia activation, as it was reported that reactive microglia were required to induce reactive astrocytes by secreting various in ammatory factors, such as TNF-α, IL-1α, and complement component 1q (C1q) (Liddelow, Guttenplan et al., 2017). However, the speci c mechanism needs deeper study.
However, there are some limitations in this study. OTULIN is a new type of deubiquitinating enzyme. After EA up-regulated the expression of OTULIN, how does OTULIN inhibited the activation of NF-κB signaling pathway? IKKγ is an important modi cation target of OTULIN in the NF-κB signaling pathway, and how does its ubiquitination modi cation level change? Which will be demonstrated in later studies.

Conclusion
Our data highlight the temporal and spatial distribution of OTULIN in the cerebral cortex following ischaemic stroke and revealed that enhanced OTULIN is required for EA to alleviate neuronal injury and the activation of glial cells, which was associated with NF-κB signalling pathway. Although there are still many problems that need to be solved, this study proposes OTULIN as a new target that exerts a neuroprotective role in ischaemic stroke, and this may be a complex process in which a variety of cells participate and coordinate with each other.