Lentivirus-mediated overexpression of OTULIN ameliorates microglia activation and neuroinflammation by depressing the activation of the NF-κB signaling pathway in cerebral ischemia/reperfusion rats

Adult male Sprague-Dawley (SD) rats (250–300 g) and neonatal (1 day old) SD rats were purchased from the Experimental Animal Center of the Chongqing Medical University. The adult animals were housed in a 12-h light/dark cycle condition with a temperature of 22 ± 2 °C and humidity of 65 ± 5%. The experimental protocol applied in the present study was approved by the Ethics Committee for Animal Experimentation of the First Affiliated Hospital of Chongqing Medical University. All procedures conducted were in accordance with the guidelines of the National Institutes for Animal Research.

A focal cerebral ischemia/reperfusion model was established by transient middle cerebral artery occlusion (tMCAO) as previously described [43, 44]. Briefly, adult SD rats were anesthetized with 3.5% chloral hydrate (1 ml/100 g) by intraperitoneal (IP) injection. During surgery, the core body temperature was continuously detected by a rectal probe and maintained at 38 ± 0.5 °C by a thermostatically controlled infrared lamp (FHC, Bowdoinham, ME, USA). A ventral midline neck incision was made to expose the right common carotid artery (CCA), the right external carotid artery (ECA), and the right internal carotid artery (ICA). Branches of the ECA were then ligated and cut off at 2.0 mm from the bifurcation of the CCA, and a heparin-dampened monofilament nylon suture (Ethicon Nylon Suture; Ethicon Inc., Osaka, Japan) with a rounded tip was gently inserted 18–20 mm from an incision in the ECA to occlude the right middle cerebral artery (MCA), confirmed with mild resistance. After 2 h of occlusion, the filament was withdrawn to restore blood flow. Successful occlusion was confirmed by a decrease in the regional cerebral blood flow to 20% and recovered to more than 80% of the baseline as detected by a laser-Doppler flowmeter (PeriFlux 5000, Perimed AB, Sweden). The animals that did not meet this requirement were excluded from the experiment. The Sham-operated rats underwent all the same surgical procedures except that the monofilament was not inserted to the MCA origin. All animals were closely monitored until they recovered from anesthesia, after which they were returned to their cages.

To induce a pure brain inflammation, lipopolysaccharide (LPS) (Sigma-Aldrich, St Louis, MO, USA, 0111: B4) was diluted with sterile PBS and administered at a dose of 500 μg/kg (IP). This dose of LPS is sufficient to cause brain inflammation without causing neuron death and has no effect on motor activity [45‐47]. The control group received the same volume of sterile saline. At 24 h after injection, the rats were sacrificed for later experiments. LPS-induced neuroinflammation models were successful, as shown by sharp increases in TNF-α, IL-1β, and IL-6 mRNA levels detected by real-time quantitative reverse transcription polymerase chain reaction (qRT-PCR; data not shown).

Primary cortical neurons were cultured with a modified method reported previously [48]. The meninges and blood vessels were isolated from the cerebral cortices of 1-day-old SD rats. The tissues were chopped mechanically, digested with 0.25% trypsin at 37 °C for 30 min, neutralized with 10% fetal bovine serum (FBS; Gibco Co., USA), and triturated with a Pasteur pipette. Then, the cells were seeded on poly-d-lysine pre-coated six-well plates (1.5 × 10⁶ cells per well) and cultured in Dulbecco’s modified Eagle’s medium/F12 medium (DMEM/F12; Gibco Co., USA) containing 10% FBS in a humidified chamber with 5% CO₂ at 37 °C for 5 h. The medium was replaced with Neurobasal medium (Gibco Co., USA) supplemented with 2% B27 and 0.5 mM glutamine. Half of the culture medium was replaced every 3 days. Arabinosylcytosine (sc-201628, Santa Cruz Co., USA) was added (5 μg/ml) at 3 days after incubation to inhibit the growth of non-neuronal cells. Cell immunofluorescence results showed that the cultured cells stained with NeuN (neuronal nuclei) exceeded 90% (data not shown). Neurons were identified by staining with anti-NeuN (MAB377, Millipore Co., Germany, 1:200).

At 7 days after culture, neurons were subjected to oxygen-glucose deprivation (OGD) according to a previously described method with some modifications [49]. Briefly, the neuronal culture medium was replaced with glucose-free D-Hanks solution (HyClone, USA). Next, the neurons were maintained in an anaerobic chamber filled with 94% N₂, 1% O₂, and 5% CO₂ for 3 h. For re-oxygenation, the cells were removed from the anaerobic chamber (Thermo 3111; Thermo Fisher Scientific Inc., USA) and returned to normal oxygen, and the D-Hanks solution was subsequently replaced with Neurobasal medium. After 24 h of re-oxygenation, the OGD-treated conditioned medium was collected for later use. The supernatant without OGD treatment served as a negative control.

A primary microglial culture was prepared following a method previously described with some modifications [50]. Briefly, the meninges and blood vessels in the rat cortex were removed from SD rats on postnatal day 1. Cortical tissue was dissociated with 0.25% trypsin at 37 °C for 30 min, and the cell suspension was filtered via a 70-μm filter. After centrifugation at 1000 rpm for 10 min, the cells were re-suspended by Pasteur pipette and then grown in flasks in a humidified atmosphere containing 5% CO_2. The culture medium, DMEM/F12 supplemented with 10% FBS, was refreshed every 3 days. After 2 weeks, microglial cells were isolated from the mixed glial cells by shaking the flasks on a rotary table concentrator at 200 rpm for 4 h at 37 °C. The purity of the cultured microglia was more than 95%, as detected by ionized calcium-binding adaptor molecule 1 (Iba-1; NB100-1028, Novus Co., USA, 1:200) immunocytochemical staining (data not shown). Iba-1 is a general marker for microglial cells.

N9 microglial cells (kindly provided by the Department of Anesthesia, the Affiliated Children’s Hospital of Chongqing Medical University) were cultured in DMEM-F12 supplemented with 10% FBS, 100 U/ml penicillin, and 100 U/ml streptomycin at 37 °C in a humidified environment containing 5% CO₂.

The lentivirus for overexpressing OTULIN (LV-OTULIN) and the control lentivirus (LV-Scramble) were obtained commercially from GenePharma Corporation (Shanghai, China).

Seven days before tMCAO surgery or LPS treatment, rats received an intracerebroventricular (ICV) injection of LV-OTULIN or LV-Scramble. The ICV injection was performed as previously described [51]. Briefly, rats were anesthetized with 3.5% chloral hydrate (1 ml/100 g) and placed in a stereotaxic apparatus (Stoelting, USA). A cranial hole on the right hemisphere was drilled 1.3 mm lateral and 1.5 mm posterior to bregma. A 10-μl Hamilton syringe (Hamilton Co., Reno, NV, USA) was stereotaxically inserted into the hole prepared for the ICV injection (3.8 mm under the dural surface). Five microliters of LV-OTULIN (1 × 10⁹ transduction units (TU)/ml) or LV-Scramble (1 × 10⁹ TU/ml) was injected into the right hemisphere at a rate of 0.5 μl/min. The needle was kept in place for an additional 5 min to prevent reflux and was then removed slowly. The expression of TREM2 mRNA and protein was significantly upregulated by OTULIN overexpression before the induction of tMCAO (Additional file 1: Figure S1).

Primary microglia and N9 microglial cells were seeded in six-well plates and transfected continuously with LV-OTULIN or LV-Scramble for 72 h. Subsequently, the medium was replaced with fresh medium. The effects of gene interference on OTULIN expression both in vivo and in vitro were verified using qRT-PCR and Western blotting.

To mimic the in vivo ischemia model, DMEM/F12 medium from primary microglia and N9 cells was replaced with OGD-treated conditioned medium (OGD[+]CM), and OGD-untreated conditioned medium (OGD[−]CM)-treated groups served as control groups.

Neurobehavioral assessment was performed at 72 h after reperfusion, based on the Longa score [43], modified Bederson score [52], and modified Garcia score [53] determined by an examiner blinded to the experimental conditions of the animals.

The Longa score was used to detect motor functions. It was graded as follows: 0, no observable deficits; 1, failure to fully extend the left forepaw; 2, difficulty in extending the left forelimb and circling to the left; 3, failing to the left side; and 4, no spontaneous walking and decreased level of consciousness. The higher the score is, the more severe the damage. Animals that scored 2 and 3 were included in this study, and the tMCAO rats that scored 0, 1, or 4 were excluded.

The Bederson score was applied to evaluate global neurological function. It was graded on a 5-point scale: 0, no observable deficits; 1, lost forelimb flexion; 2, lost forelimb flexion with lower resistance to lateral push; 3, unidirectional circling; 4, longitudinal spinning or seizure activity; and 5, no movement. The higher the score is, the more severe the damage.

The Garcia score was used to evaluate the sensorimotor function as follows: symmetry of limbs (0–3 points), spontaneous activity (0–3 points), forepaw outstretching (0–3 points), climbing (1–3 points), body proprioception (1–3 points), and response to vibrissal touch (1–3 points). The lower the score is, the more severe the damage.

The infarct volume was determined by 2,3,5-triphenyltetrazolium chloride (TTC; Sigma-Aldrich, USA) staining according to a previous method [54]. Briefly, rats were euthanized and decapitated immediately at 72 h after reperfusion. The brains were rapidly removed and frozen for 20 min at − 20 °C. Then, the brains were coronally sectioned into five consecutive slices (2 mm thick), stained with 2% TTC at 37 °C for 10 min, and then fixed with 10% formaldehyde. The slices were photographed with a digital camera (Canon IXUS, Canon Co., Japan), and image analysis software (Image-Pro Plus 6.0, Media Cybernetics Co. USA) was used to calculate the infarct volume via the following formula: percentage hemisphere lesion volume (% HLV = {[total infarct volume−(right hemisphere volume−left hemisphere volume)]/left hemisphere volume} × 100%). The measurements were performed in a blind manner.

The brain tissues in the ischemic penumbra (the location is illustrated in Fig. 1a) were removed at indicated times. The cultured cells were digested with 0.25% trypsin and then centrifuged at 1500 rpm for 15 min. The samples were homogenized in radioimmunoprecipitation assay (RIPA) lysis buffer (no. P0013B, Beyotime, Shanghai, China) supplemented with phenylmethane sulfonyl fluoride (PMSF; Beyotime, Shanghai, China) and additional phosphatase inhibitors to detect phosphorylated proteins. Total protein was extracted from the supernatants after centrifugation at 12,000 rpm for 15 min. Nuclear and cytoplasmic protein fractions were extracted by using a Nuclear and Cytoplasmic Protein Extraction Kit (no. AR0106, Boster, Beijing, China) according to the manufacturer’s instructions. Concentrations of the protein extracts were detected by using a BCA Protein Assay Reagent Kit (Beyotime, Shanghai, China). Total protein (50 μg) was separated by SDS-PAGE and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore Co., USA). The membranes were blocked with 5% non-fat milk for 1 h at room temperature and incubated at 4 °C overnight with the following primary antibodies: anti-OTULIN polyclonal rabbit antibody (no. 14127, Cell Signaling Technology, USA, 1:1000), anti-NF-κB/p65 rat monoclonal antibody (no. 8242, Cell Signaling Technology, USA, 1:1000), anti-IκBα rabbit monoclonal antibody (no. 4812, Cell Signaling Technology, USA, 1:1000), anti-phospho-IκBα (Ser32) rabbit monoclonal antibody (no. 2859, Cell Signaling Technology, USA, 1:500), and anti-β-actin rabbit monoclonal antibody (no. 4970, Cell Signaling Technology, USA, 1:1000). After washing three times with TBST, the membranes were incubated with a specific 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 used to scan the immunoblots and analyze the relative density of each band. The final results are presented as the ratios of the optical densities of targeted proteins to those of β-actin.

Total RNA of brain tissue or cultured cells was extracted by using Trizol (Takara Biotechnology, Japan) and converted to cDNA by using a Prime Script TM RT Reagent kit (Takara Biotechnology, Japan). The real-time quantitative reverse transcription polymerase chain reaction (qRT-PCR) was performed on a Q5 Gradient Real-Time PCR Detection System (Bio-Rad Co., USA) with the SYBR Green System (SYBR Premix Ex TaqTMII, Takara Co., Japan). Cycle conditions included heating for 30 s at 95 °C, followed by 40 cycles of 5 s at 95 °C, 30 s at 60 °C, and 60 s at 72 °C. The melting curve of each gene was performed to ensure specific amplification. All gene-specific expression values were normalized against β-actin gene expression in the same samples. All primer sequences are shown in Table 1.

The levels of TNF-α, IL-1β, and IL-6 in brain homogenate and cell supernatants were measured by using enzyme-linked immunosorbent assay (ELISA) kits (no. EK0526 96T, EK0393, and EK0412 96T, BOSTER Co., China) following the manufacturer’s instructions.

The slides of brain sections were fixed with 4% formaldehyde solution for 30 min at room temperature, incubated with 1% Triton X-100 for 30 min, and blocked with 5% goat or donkey serum for 1 h at 37 °C. Subsequently, sections were incubated overnight at 4 °C with the following primary antibodies: OTULIN (bs-14689R, Bioss Co., Beijing, China, 1:50), NeuN (MAB377, Millipore Co., Germany, 1:200), MAP2 (4542, Cell Signaling Technology, USA, 1:100) or Iba-1 (NB100-1028, Novus Co., USA, 1:200). After washing three times with PBS, sections were reacted with the following fluorescent secondary antibodies at 37 °C for 1 h: Alexa Fluor 594-conjugated goat anti-rabbit IgG (H+L; SA00006-4, Proteintech, 1:200), Alexa Fluor 488-conjugated goat anti-mouse IgG (H+L; SA00006-1, Proteintech, 1:200), FITC-conjugated AffiniPure donkey anti-goat IgG (H+L; SA00003-3, Proteintech, 1:200), and 594-conjugated AffiniPure donkey anti-rabbit IgG (H+L; SA00006-8, Proteintech, 1:200). DAPI (Sigma, USA, 1:200) was used to stain cellular nuclei at 37 °C for 10 min. All images were observed and acquired using an A1+R laser confocal microscope (Nikon, Tokyo, Japan).

The Longa score and Bederson score data were analyzed using Kruskal-Wallis tests followed by post hoc Dunn’s multiple comparison tests. Unpaired Student’s t tests were used when two groups were compared. All other quantitative data were analyzed using one-way ANOVA followed by Tukey’s post hoc test for multiple comparisons. Statistical analyses were performed using SPSS 19.0. All values are expressed as the means ± SEMs, and values of P < 0.05 were considered statistically significant.

To analyze the time course of OTULIN expression following cerebral ischemia/reperfusion, we detected the levels of OTULIN mRNA and protein in the ischemic penumbra of the cerebral cortex (Fig. 1a) within 72 h after reperfusion by Western blot and qRT-PCR. The rats were randomly divided into the Sham group and the tMCAO group. OTULIN mRNA expression in the tMCAO group increased gradually with the prolongation of reperfusion time and peaked at 48 h, followed by a decrease at 72 h. Moreover, the levels of OTULIN mRNA at indicated times in the tMCAO group remained significantly higher than those in the Sham group except at 6 h (Fig. 1b, n = 5 per group for RT-qPCR).

Consistent with the qRT-PCR results, the Western blot data indicated that OTULIN protein levels in the tMCAO group were markedly increased compared with those in the Sham group at each time point except at 6 h (Fig. 1c, d, n = 3 per group for Western blot). Together, these results suggest that cerebral ischemia induced an endogenous increase in OTULIN expression in the ischemic penumbra of the cerebral cortex.

To investigate the effect of OTULIN on stroke outcomes, cerebral infarct volume, neurobehavioral assessments, and neuronal loss in each group were detected at 72 h after reperfusion. OTULIN expression was enhanced by ICV injection of LV-OTULIN, and an empty vector (LV-Scramble) was injected as a control (Fig. 2a). The tMCAO model was established at 7 days after ICV injection, and 72 h later, animals were sacrificed for subsequent experiments (Fig. 2a). The rats were divided into four groups: Sham, tMCAO, tMCAO+LV-Scramble, and tMCAO+LV-OTULIN. As expected, the LV-OTULIN lentivirus effectively promoted OTULIN mRNA and protein expression as determined by qRT-PCR (Fig. 2b, n = 5 per group) and Western blot (Fig. 2c, d, n = 3 per group) compared to those in the tMCAO group and the tMCAO+LV-Scramble group.

The Longa score, Bederson score, and Garcia score consistently showed that the tMCAO group exhibited obvious neurological dysfunction compared with the Sham group. The neurological function deficits in the tMCAO+LV-OTULIN group were obviously improved compared to those in the tMCAO group and the tMCAO+LV-Scramble group in terms of the Longa score (Fig. 2e, n = 6 per group), Bederson score (Fig. 2f, n = 6 per group), and Garcia score (Fig. 2g, n = 6 per group). Moreover, rats in the tMCAO+LV-OTULIN group displayed significantly smaller infarct volume in the cortex and striatum compared to those in the tMCAO group or tMCAO+LV-Scramble group (Fig. 2h, i, n = 5 per group). The infarct volume in the tMCAO group was similar to that in the tMCAO+LV-Scramble group (P > 0.05; Fig. 2h, i). We further examined NeuN- and MAP2-stained neurons to assess the brain infarct damage. The qualitative analysis revealed that OTULIN overexpression markedly increased both NeuN-positive (Fig. 2j, k, n = 6 per group) and MAP2-positive (Fig. 2j, k, n = 6 per group) cells in the ischemic penumbra compared with that in the tMCAO group or the tMCAO+LV-Scramble group at 72 h following reperfusion. Collectively, these data indicate that the outcome of ischemic stroke could be improved by OTULIN overexpression.

To investigate the effect of OTULIN overexpression on microglia activation, we analyzed the immunoreactivity of Iba-1⁺ cells in the cortical ischemic penumbra. The rats were divided into four groups including: Sham, tMCAO, tMCAO+LV-Scramble, and tMCAO+LV-OTULIN (n = 6 per group). Morphological change is one of the classic features of microglia activation after ischemic stroke [55]. Iba-1⁺ cells were mainly classified into three types according to a common morphological classification: resting cells, activated-ramified cells, and amoeboid cells. Resting Iba-1⁺ cells featured small cell bodies and thin-branched processes (Fig. 3A(a)). Activated-ramified Iba-1⁺ cells (Fig. 3A(b), (c)) also had processes, but their cell bodies and processes were thick. Amoeboid Iba-1⁺ cells (Fig. 3A(d)) were characterized by enlarged cell bodies without processes. In the Sham group, almost all microglia remained in the resting state at 24 h, 72 h, and 7 days. At 24 h after stroke, Iba-1⁺ microglia mainly became activated-ramified cells, and a few amoeboid Iba-1⁺ cells in the penumbra were observed in all groups. The mean Iba-1 immunofluorescence intensity in the tMCAO+LV-OTULIN group was obviously more decreased than that in the tMCAO group or tMCAO+LV-Scramble group (Fig. 3B(d, g, j), C). At 72 h after reperfusion, amoeboid Iba-1⁺ cells obviously appeared and accumulated in the penumbra. The mean Iba-1 immunofluorescence intensity in the tMCAO+LV-OTULIN group (Fig. 3B(k), C) was significantly lower than that in the tMCAO group (Fig. 3B(e)) or the tMCAO+LV-Scramble group (Fig. 3B(h)). Almost all Iba-1⁺ activated cells become amoeboid in shape, and more prominent microglia accumulated in the penumbra at 7 days after reperfusion; moreover, the mean Iba-1 immunofluorescence intensity was more obviously decreased in the tMCAO+LV-OTULIN group (Fig. 3B(l), C) than in the tMCAO group (Fig. 3B(f)) or the tMCAO+LV-Scramble group (Fig. 3B(i)). In the contralateral cortex of all ischemic groups, Iba-1⁺ microglial cells existed diffusely in a resting shape, their Iba-1 immunoreactivity was quite weak, and activated-ramified or amoeboid Iba-1⁺ cells were rarely found at each time point after reperfusion (Fig. 3B(m–r)), the Iba-1 immunoreactivity in contralateral cortex of the Sham and tMCAO+LV-Scramble groups is not shown. Taken together, these results indicate that OTULIN overexpression limits the accumulation of microglia in the ischemic penumbra after cerebral ischemia, but this had no effect on the microglial cells in the contralateral cortex.

The extent of inflammation is often proportional to the infarct size [54, 56]. To exclude the possibility that the attenuation of microglia activation in OTULIN-overexpressing rats might be ascribed to the secondary effect of a reduction in infarct volume, we established a pure brain neuroinflammation rat model by IP injection of LPS, a direct promoter of microglial activation without a corresponding neuronal loss. The rats that received an injection of sterile saline served as the control group (LPS[−] group), and those that received the LPS injection served as the LPS(+) group. The rats were divided into four groups: the LPS(−) group, the LPS(+) group, the LPS(+)+LV-Scramble group, and the LPS(+)+LV-OTULIN group. The confocal immunofluorescence results showed that Iba-1⁺ microglial cells located in the cortex mainly existed in an activated-ramified state at 24 h after LPS treatment (Fig. 4A(b)), and the mean Iba-1 immunofluorescence intensity increased sharply compared with that in the LPS(−) group (Fig. 4A(a)). The quantitative analysis results (Fig. 4(b)) showed that overexpression of OTULIN (Fig. 4A(d)) resulted in an obvious decrease in mean Iba-1 immunofluorescence intensity compared to that in the LPS(+) (Fig. 4A(b)) and LPS(+)+LV-Scramble groups (Fig. 4A(c)).

To observe the expression of OTULIN in microglia after in vivo cerebral ischemia, a confocal immunofluorescence technique was performed to detect the double staining of OTULIN and Iba-1 in the tMCAO and Sham groups (n = 6 per group). Immunofluorescence results showed that Iba-1⁺ microglial cells were found in both hemispheres with a pronounced expression in the ischemic area at 72 h after reperfusion. The tMCAO group showed a greater increase in OTULIN immunoreactivity than did the Sham group. Interestingly, the expression of OTULIN was enriched in the activated microglial cells that possessed an amoeboid morphology, whereas little OTULIN was found in Iba-1⁺ resting microglia with a ramified shape in the Sham group (Fig. 5a).

We next quantitatively analyzed OTULIN expression levels in activated microglia in vitro. To mimic the in vivo stroke condition, we induced microglial cells with conditioned medium (CM) collected from OGD-treated neuronal cultures, which is a widely accepted cellular ischemic model [57, 58]. The groups were established as follows: the normal control (NC) group treated with regular microglial medium, the OGD(−)CM group cultured with OGD-untreated CM, and the OGD(+)CM group treated with CM that was not subjected to OGD. After 24 h of culture, the levels of OTULIN mRNA (Fig. 5b, n = 6 per group) and protein expression (Fig. 5c, d, n = 3 per group) were markedly increased in the PM and N9 cells in the OGD(+)CM group compared with the NC group or the OGD(−)CM group (Fig. 5b–d). Combined with the in vivo results, we could conclude that OTULIN levels were increased in ischemia-induced activated microglia.

To explore the effect of OTULIN on the production of inflammatory cytokines, ELISA was applied to evaluate the protein levels of TNF-α, IL-1β, and IL-6 at 24 h after cerebral ischemia. The rats were randomly divided into four groups: Sham group, tMCAO group, tMCAO+LV-Scramble group, and tMCAO+LV-OTULIN group (n = 6 per group for ELISA and n = 3 per group for Western blot). In line with the previous studies [7, 59, 60], ischemic stroke induced massive production of TNF-α, IL-1β, and IL-6. The treatment with LV-OTULIN in the tMCAO+LV-OTULIN group significantly decreased the contents of TNF-α (Fig. 6a), IL-1β (Fig. 6b), and IL-6 (Fig. 6c) compared to the levels of those proteins in the tMCAO group and the tMCAO+LV-Scramble group.

Next, we explored whether OTULIN overexpression decreased the expression of pro-inflammatory cytokines through the NF-κB pathway in tMCAO rats. Western blot was used to measure p-IκBα, IκBα, cytoplasmic p65, and nuclear p65 proteins in the ischemic area at 24 h after reperfusion. In the study, we detected enhanced degradation and phosphorylation of IκBα and nuclear translocation of NF-κB p65 in the tMCAO group compared to that in the Sham group (Fig. 6d, e). Moreover, OTULIN overexpression significantly attenuated NF-κB activity in the ischemic cortex, as determined by the inhibition of IκBα degradation and phosphorylation along with reduced nuclear translocation of p65, which manifested as less p65 in the nucleus and more p65 in the cytoplasm (Fig. 6d, e).

To further explore whether the post-ischemia induction of OTULIN acted directly via a microglial mechanism to modulate neuroinflammation, we activated microglial cells by using OGD(+)CM in vitro, which is a commonly used in vitro model for studying microglia-mediated neuroinflammation. PM or N9 cells transfected with LV-OTULIN or LV-Scramble were divided into NC, OGD(−)CM, and OGD(+)CM groups. ELISA results showed that only the OGD(+)CM markedly activated the PM and N9 cells, as evidenced by the robust production of TNF-α (Fig. 7a, n = 6 per group), IL-1β (Fig. 7b, n = 6 per group), and IL-6 (Fig. 7c, n = 6 per group), while the NC group and the OGD(−)CM group treated with LV-OTULIN or LV-Scramble displayed no changes in cytokine expression. The expression levels of these pro-inflammatory cytokines were significantly attenuated by LV-OTULIN but not by LV-Scramble (Fig. 7a–c).

Next, we analyzed whether OTULIN overexpression weakened the secretion of pro-inflammatory cytokines in PM and N9 cells via regulating the NF-κB pathway. Consistent with our in vivo results, we found that NF-κB activities induced by OGD(+)CM in the PM group (Fig. 7d, e, n = 3 per group), as well as in the N9 group (Fig. 7d, f, n = 3 per group), were significantly attenuated by LV-OTULIN, as shown by decreased IκBα degradation and phosphorylation and subsequently reduced nuclear translocation of p65. Collectively, these results reveal that OTULIN expression suppressed microglia-mediated neuroinflammation through the NF-κB pathway in the in vitro ischemic condition.