RhTrx-1 ameliorates miroglial neuroinflammation after cerebral ischemic stroke

Background Microglia are rapidly activated after ischemic stroke and participate in the occurrence of neuroinflammation, which exacerbates the injury of ischemic stroke. Receptor Interacting Serine Threonine Kinase 1 (RIPK1) is thought to be involved in the development of inflammatory responses, but its role in ischemic microglia remains unclear. Here, we applied recombinant human thioredoxin-1 (rhTrx-1), a potential neuroprotective agent, to explore the role of rhTrx-1 in inhibiting RIPK1-mediated neuroinflammatory responses in microglia. Method Middle cerebral artery occlusion (MCAO) and Oxygen and glucose deprivation (OGD) were conducted for in vivo and in vitro experimental stroke models. The expression of RIPK1 in microglia after ischemia was examined. The inflammatory response of microglia was analyzed after treatment with rhTrx-1 and Necrostatin-1 (Nec-1, inhibitors of RIPK1), and the mechanisms were explored. In addition, the effects of rhTrx-1 on neurobehavioral deficits and cerebral infarct volume were examined. Recombinant RIPK1: RIPK3: Receptor Interacting Serine Threonine Kinase 3; ROS: Reactive oxygen species; TEM: Transmission Electron Microscope; TGF-β: Transforming Growth Factor-β; TNF-α: Tumor Necrosis Factor-α; Trx: Thioredoxin; TTC: 2,3,5-triphenyltetrazolium chloride.

neuroinflammatory response in MCAO mice [18][19][20]. These findings provide a strong scientific basis for the clinical translational application of Trx-1. RhTrx-1 is a compound extracted and synthesized from Escherichia coli. Current studies have shown that using rhTrx-1 treatment could reduce cerebral infarction volume and improve neurological deficits in MCAO mice. Besides, rhTrx-1 treatment could alleviate ischemia-induced neuron injury [21][22][23]. However, the protective and regulatory effects of rhTrx-1 on microglia after cerebral ischemia remain unclear. On these bases, our study aims to investigate the therapeutic potential of rhTrx-1 in inhibiting RIPK1-driven neuroinflammation in microglia after cerebral ischemia.
group. All mice were anesthetized with pentobarbital sodium before inducing MCAO. During the operation, mice body temperature was kept at 37℃. For MCAO procedure, common carotid artery (CCA), external carotid artery (ECA) and internal carotid artery (ICA) were exposed and separated.
After clipping the common carotid artery, ligation was performed at the ECA and a small incision was made. Through the incision, thread embolism was slowly inserted into ICA. After 60 min occlusion, thread embolism was pulled out and the reperfusion was carried out. Mice in Sham group received the same procedure, but did not receive the embolus. The mice in MCAO + rhTrx-1 group received 10 mg/kg rhTrx-1 by tail vein injection after reperfusion. As control, mice in MCAO group were injected with equal volume of 0.9% sterile saline. All mice were sacrificed 24 h after reperfusion for further analysis.
Infarct Volume Assessment TTC staining was executed to evaluate the size of cerebral infarction. In brief, the brain was taken to make coronal brain sections after the mice were sacrificed. The slices were placed in 2% TTC staining solution and immersed for 10 min at 37℃ avoiding meeting up. After staining, the slices were fixed in 4% PFA overnight, and pictured the next day. Each infarct area was measured by using Image J software.

Neurobehavioral Testing
To assess the degree of sensory and motor damage in each group of mice, Bederson score and corner test were carried out based on previous studies [25,26]. In order to evaluate Bederson score, scores were recorded according to the physical signs of the mice in the tail suspension state as followed: 0 points, normal; 1 point, the contralateral forelimb could not be fully extended; 2 points, the resistance to thrust of the contralateral forelimb decreased; 3 points, turning to the contralateral side of the lesion. For evaluation corner test, mice were placed in the depth of 30-degree angle, and the direction of turning was observed. Normal mice turned randomly and with equal probability toward both sides, whereas mice turned toward the lesion side after cerebral ischemia.
OGD Procedure And Treatment BV-2 cell line were cultured in complete medium, and incubated in 37℃ incubator containing 5% CO 2 .
The complete medium was composed of DMEM containing 10% Fetal Bovine Serum (FBS) and 1% penicillin/streptomycin solution. For the OGD program, after replacing the cell supernatant with EBSS solution, the cells were transferred to three gas incubators (37℃, containing 95% N 2 and 5% CO 2 ) and cultured for 1, 2, or 4 h, respectively. Cells in the Nec-1 group were treated with Nec-1 (20 µM) after re-oxygenation and maintained for 6 h, 12 h, 24 h. Cells in the rhTrx-1 group were incubated with 5, 10 or 25 µg/ml rhTrx-1 (dissolved in sterile PBS) for 24 h after reoxygenation.

Flow Cytometry
To detect apoptotic rate and intracellular ROS expression level, Annexin V-PE/7AAD kit and DCFH-DA solution were used according to the manufacturer's instructions. Briefly, to analyze the extent of apoptosis, cells were harvested and double-stained with Annexin V-PE and 7AAD for 15 min at room temperature in the dark. To detect intracellular ROS levels, cells were harvested and stained with DCFH-DA solution for 20 min in the dark at 37℃. Immediately after the above staining, the cells were detected and analyzed using a Beckman CytoFLEX flow cytometer (CA, USA).

Western Blot Analysis
The protein levels were detected by western blot. In a brief, cells or brain tissues were lysed in precooled RIPA buffer and centrifuged to collect the supernatant for protein extraction. Samples were loaded onto SDS-PAGE and then transferred onto nitrocellulose membranes. Membranes were blocked with 5% BSA or 5% non-fat dry milk and finally incubated with primary antibodies (anti-RIPK1, anti-RIPK3, anti-MLKL, anti-pMLKL, anti-CCL2, anti-MMP-9, anti-NLRP3, anti-ASC, anti-caspase-1, anticaspase-3 and anti-β-actin) at 4℃ overnight. The second day membranes were incubated with Alexa Fluor 800-conjugated Goat-anti rabbit antibody at room temperature. Protein bands were imaged and analyzed with the Odyssey system (LI-COR Biosciences, Lincoln NE, USA).

Transmission Electron Microscope (TEM)
After collecting the cell precipitation, glutaraldehyde was added slowly for fixation. Ultrathin cell sections (100 nm thick) prepared using ultramicrotome were mounted on a copper grid, and then stained with uranyl acetate and lead citrate. Observed the sections at an accelerating voltage of 80 kV on a transmission electron microscope (Hitachi H-7100, Hitachinaka, Japan) by single blind.

Immunofluorescence Staining
Briefly, for in vitro test, cells were fixed with 4% paraformaldehyde (PFA) for 15 min and then washed with PBS for 3 times and then incubated with first antibodies. After incubated with the anti-CD206 or anti-CD16 primary antibody respectively at 4℃ overnight, cells were incubated with the secondary antibodies at room temperature for 1 h, and were added Hoechst finally. For in vivo test, frozen slices were only washed with PBS for 3 times and then blocked with 5% BSA at room temperature for 1 h.

ELISA
To detect the inflammatory factor levels, cell supernatant or brain tissues were collected and ELISA were performed according to the instructions. Briefly, 50 µL sample and 100 µL antibody being tested were added to the reaction wells respectively and then placed the wells-plate in a 37℃ incubator and incubated for 60 min in the dark. After incubation with termination solution, the OD value was measured at 450 to measure the expression levels of TNF-α, IL-1β and TGF-β.

Molecular Docking
The structures of rhTrx-1 and RIPK1 were processed and optimized using Accelrys Discovery Studio 2016 platform (San Diego, CA, USA), and protein docking was performed and calculated in ZDOCK module. Poses with the best scores were selected and then further energy optimization was performed using the RDock program. The interaction between rhTrx-1 and RIPK1 was analyzed using the Analyze Protein Interface module. Finally, Pymol (DeLano Scientific. Palo Alto, CA, USA) was used for mapping.

Statistical Analysis
GraphPad Prism (Version 6.0c) software was used for statistics. Student's T test and One-factor Analysis of Variance (One-Way ANOVA) were used to evaluate the statistical significance. The results were presented using mean standard deviation. P < 0.05 was considered statistically significant.

RIPK1 was induced in the microglia of MCAO following reperfusion
We applied the MCAO model and examined the expression of RIPK1 in the mouse brain 24 h after reperfusion. Firstly, we removed samples from the contralateral hemisphere and the ipsilateral hemisphere, and then examined samples by using western blot. As described in Fig. 1a and 1b, expression of RIPK1 was significantly increased on the ipsilateral hemisphere compared with the contralateral hemisphere. Next, to assess the localization of RIPK1 in microglia, the double staining of RIPK1 and Iba-1 was examined by the whole scan analysis. We found that RIPK1 exhibited in microglia of MCAO (Fig. 1c, d). These results indicated that RIPK1 was induced in the microglia of MCAO following reperfusion.

RIPK1-induced Necroptosis Protein Levels Of OGD Microglia In Vitro
To investigate the degree of microglial necroptosis at different OGD time, flow cytometry was performed. The results showed that the number of early apoptotic and late apoptotic microglia gradually increased with the prolongation of OGD time (Fig. 2a) and was the highest at 24 h of reoxygenation after 4 h of GOD (Fig. 2b). Based on this, further investigation of RIPK1 level at different reoxygenation time after 4 h of GOD presented that RIPK1 peaked at 6 h of reoxygenation ( Fig. 2c). Furthermore, therapeutic treatment of Nec-1 restrained RIPK1, RIPK3, and pMLKL/MLKL levels compared with untreated groups (Fig. 2d, e). The inhibitory effect was most significant at 24 h of reoxygenation. Taken together, these results indicated that necroptosis was augmented in microglia following ischemia reperfusion.

RhTrx-1 decreased RIPK1-induced necroptosis and apoptosis of OGD microglia in vitro
To characterize the role of rhTrx-1, we computerized the binding interaction of rhTrx-1 and RIPK1.
Docking analysis illustrated that total six Pi interactions, five Hydrogen Bonds and one Salt Bridges were exhibited between rhTrx-1 and RIPK1 (Table S1), which indicated rhTrx1 directly binds with RIPK1 ( Fig. 3a, b). Further detection showed 5 µg/mL, 10 µg/mL and 25 µg/mL dose of rhtTx-1 treatment inhibited the expression of RIPK1, RIPK3, and pMLKL/MLKL after 4 h of OGD, and the inhibitory effect of rhTrx-1 was more obvious in the dose of 25 µg/mL (Fig. 3c, d). According to these, follow-up experiments were carried out with 25 µg/mL concentration as the rhTrx-1 action concentration. Subsequent investigation revealed that rhTrx-1 and Nec-1 treatment reduced Cleaved-Caspase-3 level (Fig. 3e). Ultrastructural images of TEM showed cell structures of OGD-induced microglia in rhTrx-1 and Nec-1 treatment groups were more similar to the control group (Fig. 3f). In addition, compared to the untreated groups, fewer microglia underwent early and late apoptosis in rhTrx-1 and Nec-1 treatment group after OGD (Fig. 3g, h).
Therapeutic treatment with rhTrx-1 mitigated OGD-induced microglia mitochondrial injury and NLRP3 inflammasome activation To determine the impact of rhTrx-1 on the mitochondrial injury in ischemia microglia, we measured the mitochondrial potential by using JC-1 staining. The mitochondrial membrane potential was dwindled (aggregate was decreased and monomer increased) after OGD 4 h/reoxygenation 24 h, whereas rhTrx-1 and Nec-1 treatment reversed the reduction of potential (Fig. 4a, b). Further detection proved that rhTrx-1 and Nec-1 treatment eliminated the accumulation of ROS in OGD-induced microglia (Fig. 4c, d). Lastly, inflammasome proteins levels were examined to determine the activation of inflammasome. Compared to untreated groups, NLRP3, ASC and Cleaved-Caspase-1 levels in rhTrx-1 and Nec-1 treatment groups were significance decreased (Fig. 4e, f), and the release of IL-1β was cleared simultaneous as well (Fig. 4g).
Therapeutic treatment with rhTrx-1 regulated OGD-induced microglia polarization and inhibited the release of inflammation mediators To clarify the effect of rhTrx-1 on the microglia polarization, we observed the intensity of CD16 and CD206. As demonstrated in Fig. 5a, b, the fluorescence intensity of CD16, representing M1-type microglia, enhanced in OGD and significantly abated after treatment with rhTrx-1 and Nec-1. Contrast to this, the fluorescence intensity of M2 microglia-labeled CD206 was increased and treatment with rhTrx-1 and Nec-1 strengthened the augment (Fig. 5c, d). Moreover, the increased inflammation mediators CCL2, MMP-9, and TNF-α level were erased after rhTrx-1 and Nec-1 treatment, and the expression of TFG-β increased further compared to untreated group.

Administration Of rhTrx-1 Reduced Acute Cerebral Ischemic Stroke Injury
We examined the neurological function of mice after 24 h of MCAO reperfusion. Berderson score and corner test results showed that neurological deficits of rhTrx-1 treated MCAO mice were significantly reduced compared with untreated MCAO mice (Fig. 6a, b). The volume size of cerebral infarction was detected by TTC staining, and the size of cerebral infarction in the rhTrx-1 treatment group was significantly smaller than that in the untreated MCAO group (Fig. 6c). The results of TTC staining supported the results of neurobehavioral testing.

Discussion
The microglial neurotoxicity is directly related to the adverse outcome of ischemic brain injury [5, 6, 27, 28]. In this study, we performed MCAO and OGD procedure assessing RIPK1-mediated inflammatory response in microglia, and found that rhTrx-1 can be used as an inhibitor of RIPK1 to regulate the inflammatory activation of microglia after cerebral ischemia.
RIPK1 is considered to be a key regulator of innate immunity, which can regulate the occurrence of inflammatory response [9]. Study has shown that mice lacking RIPK1 gene die of systemic multiple organ inflammation at birth [29]. As a promoter of programmed cell death, RIPK1 can recruit RIPK3 to form necrosome and finally activate the phosphorylation of MLKL to activate the necroptosis-signaling pathway. And therefore causes the cells to undergo necrotic changes and release pro-inflammatory mediators such as MMP-9 and TNF-α to intensify the neuroinflammation [30, 31]. Huang and Fan et al proved that Nec-1 could inhibit the inflammatory response mediated by microglia in the retina and spinal cord by activating necroptosis pathway [32,33]. These suggest that treatment of inhibiting necroptosis pathway in microglia may be beneficial to provide anti-inflammatory effects for ischemic stroke. In our data, we showed that RIPK1 level increased after MCAO, and found that RIPK1 localized in microglia. To figure out its role in ischemic stroke-induced microglial neuroinflammation, we detected the expression level of RIPK1 in vitro. We found that the expression of RIPK1 increased with the prolong of reoxygenation time in microglia. In the context of Nec-1 or rhTrx-1 treatment, the induction of necroptosis by RIPK1 was significantly decreased in OGD-induced microglia indicating that rhTrx-1 has similar pharmacological effects to Nec-1. In accordance with these results, we further used a docking model and found that rhTrx-1 directly binds to RIPK1 with Pi interactions, salt bridge and hydrogen bonds. It illustrated that rhTrx-1 can be directly combined with RIPK1 and produced a marked effect in the inhibition of RIPK1.
In addition to activating necroptosis, RIPK1 can initiate apoptosis as a scaffold protein [34]. Although it is believed that apoptosis does not involve in the release of inflammatory factors, the apoptotic microglia may indirectly aggravate the damage of cerebral ischemic tissues and cells due to the loss of normal physiological functions. We found that rhTrx-1 and Nec-1 treatment reduced the activation of cleave-caspase-3 after OGD in microglia, suggesting rhTrx-1 has anti-apoptotic effect in ischemic microglia by inhibiting RIPK1 level. Besides, RIPK1 can activate the expression of pyruvate dehydrogenase (PDH) complex by recruiting RIPK3, which promotes the release of ROS accumulation in mitochondria. And that make further efforts to increase the expression of ROS in a feed forward manner [35,36]. In this way, that kind of loop promotes the accumulation of ROS, leading to mitochondrial damage and activation of mitochondrial-mediated apoptosis [37]. When we studied the expression of ROS in microglia and the change of mitochondrial membrane potential, we found that rhTrx-1 effectively inhibited the pathological accumulation of ROS and rescued mitochondrial injury.
Such protective effect might be due to the inhibition of the cascade reaction of RIPK1. Of course, rhTrx-1 has antioxidant effect on its own biological characteristics, which affects mitochondrial function and ROS release by affecting cell oxidative respiratory chain during treatment [38]. However, due to the complexity of oxidative stress injury mechanism, the study of oxidative stress therapy cannot be single. At least, our study provides evidence for the therapeutic effect of rhTrx-1 from another perspective.
It has been reported that the activation of RIPK3-MLKL can directly trigger the activation of NLRP3 inflammasome in renal tubular cells [39]. Furthermore, RIPK1 can trigger the activation of NLRP3 inflammasome by destroying mitochondria membrane and promoting the release of ROS, which further promotes the production and secretion of IL-1β to extracellular and initiate neuroinflammation. To investigate the role of rhTrx-1 in RIPK1-induced NLRP3 inflammasome activation, we examined the levels of NLRP3, ASC and cleaved-caspase-1 and indicated that rhTrx-1 ultimately reduced the release of IL-1β, depending on the inhibition of NLRP3 inflammasome. These results highlight the role of rhTrx-1 in inhibiting microglia associated-inflammation in cerebral ischemic stroke.
Compared with MCAO group, rhTrx-1 treatment alleviated the absence of neurological deficit after MCAO, and effectively inhibited the volume of cerebral infarction. Similar to our results, Hattori et al found that rhTrx-1 was able to permeate the blood-brain barrier and significantly ameliorated neurological function deficit and infarct volume size [21]. Ma et al investigated that rhTrx-1 exerts an antioxidative and antinitrative effect against neurological dysfunction and cerebral infarction [22]. On these basis, we further considered the activation of microglia in vitro and in vivo. The activated microglia after ischemia details a dual function of promoting inflammation and anti-inflammation due to its polarization to M1/M2 phenotype conversion [6, 40, 41]. We found that rhTrx-1 treatment increased M2 phenotype microglia and decreased M1 phenotype after OGD procedure. Consistently, we also observed that rhTrx-1 treatment predisposed the microglia phenotype to M2 type in the in vivo MCAO model. Moreover, we showed that TGF-β released by microglia was increased, while the release of TNF-α, CCL2 and MMP-9 were decreased compared with the untreated group. These data provides evidences for rhTrx-1 inhibiting inflammation mediated by RIPK1. The results also prove from the side that inhibiting the expression of RIPK1 can make microglia gradually show its original advantages in repairing brain tissue damage and promoting angiogenesis.

Conclusions
In conclusion, our study found that RIPK1 plays an important role in microglia-mediated neuroinflammatory response after cerebral ischemia. It can regulate the conversion of microglia        Mechanisms of rhTrx-1 treatment on cerebral ischemic stroke-induced microglia activation involved RIPK1-mediated necroptosis, apoptosis and NLRP3 inflammasome.

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