Hypoxic preconditioned mesenchymal stem cells ameliorate rat brain injury after cardiopulmonary resuscitation by suppressing neuronal pyroptosis

Abstract Cardiac arrest (CA) can result in cerebral ischaemia–reperfusion injury and poor neurological outcomes. While bone marrow‐derived mesenchymal stem cells (BMSCs) have been shown to have protective effects in brain ischaemic disease, their efficacy can be reduced by the poor oxygen environment. In this study, we investigated the neuroprotective effects of hypoxic preconditioned BMSCs (HP‐BMSCs) and normoxic BMSCs (N‐BMSCs) in a cardiac arrest rat model by examining their ability to ameliorate cell pyroptosis. The mechanism underlying the process was also explored. Cardiac arrest was induced in rats for 8 min and surviving rats received 1 × 106 normoxic/hypoxic BMSCs or PBS via intracerebroventricular (ICV) transplantation. Neurological function of rats was evaluated using neurological deficit scores (NDSs) and examined for brain pathology. Serum S100B and neuron‐specific enolase (NSE) levels and cortical proinflammatory cytokines were measured to evaluate brain injury. Pyroptosis‐related proteins in the cortex after cardiopulmonary resuscitation (CPR) were measured using western blotting and immunofluorescent staining. Transplanted BMSCs were tracked using bioluminescence imaging. Results showed significantly better neurological function and neuropathological damage after transplantation with HP‐BMSCs. In addition, HP‐BMSCs reduced levels of pyroptosis‐related proteins in the rat cortex after CPR and significantly reduced levels of biomarkers for brain injury. Mechanistically, HP‐BMSCs alleviated brain injury by reducing the expressions of HMGB1, TLR4, NF‐κB p65, p38 MAPK and JNK in the cortex. Our study demonstrated that hypoxic preconditioning could enhance the efficacy of BMSCs in alleviating post‐resuscitation cortical pyroptosis. This effect may be related to the regulation of the HMGB1/TLR4/NF‐κB, MAPK signalling pathways.


| INTRODUC TI ON
Cardiac arrest (CA) is a serious public health emergency with high morbidity and mortality rates. 1,2 Although the treatment of CA has improved, only 20%-40% of victims regain spontaneous circulation (ROSC). 3 Brain damage is responsible for most deaths after ROSC. 1 ROSC, followed by post-CA syndrome, is characterized by global cerebral ischemia/reperfusion (I/R)-induced injury, which contributes to a poor prognosis. 4 Since the brain is highly susceptible to I/R injury, brain damage after cardiopulmonary resuscitation CPR accounts for two-thirds of deaths in patients resuscitated out-of-hospital CA and nearly a quarter of patients who survive in-hospital CA. 5 Approximately 30% of survivors suffer from permanent injury brain, and half of the resuscitated patients are discharged from the hospital with varying degrees of neurological deficits. 6,7 Therefore, it is essential to explore the mechanisms underlying post-resuscitation brain damage and identify neuroprotective drugs for resuscitated patients.
The mechanism underlying the increased brain damage after CPR resulting from global cerebral I/R remains unclear. However, inflammation is thought to play a significant role in secondary damage in brain ischaemic diseases. 8 A growing body of evidence suggests that neuroinflammation triggered by brain I/R activates programmed cell death (PCD) signalling pathways within hours to days. 9,10 Numerous studies have indicated that regulating the apoptosis, necrosis, autophagy and pyroptosis pathways can achieve neuroprotection. [11][12][13] Pyroptosis, a regulated cell death, has been observed in CPR models and is characterized by cell swelling and plasma-membrane breakage. 13 In addition, two recent studies have suggested that cell pyroptosis is extensively involved after CPR and could become a novel target for treating CPR. 13,14 Brain I/R damage generally activates the nod-like receptor family protein 3 (NLRP3) inflammasome and secretion of pro-inflammatory cytokines, including Interleukin-1β (IL-1β) and Interleukin-18 (IL-18). 15 The NLRP3 inflammasome, including NLRP3, apoptosis-associated speck-like protein containing a CARD (ASC), and precursor of caspase-1 (pro-caspase-1), plays an essential role in the initial phase of pyroptosis. 16,17 Subsequently, the activated caspase-1 is cleaved, then processed by the inactive pro-IL-1βand pro-IL-18 and a protein named gadermin D (GSDMD), resulting in the formation of membrane pores and the release of abundant IL-1β and IL-18, causing cell pyroptosis. 18,19 Studies have confirmed that NLRP3 inflammasomedependent cell pyroptosis is involved in the pathogenic mechanism of whole brain I/R injury after CA/CPR. 13,20 Meanwhile, the high-mobility family box-1 (HMGB1) is also secreted extracellularly. HMGB1 binds to its receptors, toll-like receptor 4 (TLR4) and RAGE, and regulates the expression of nuclear factor-kB (NF-κB), which expands the inflammatory response. 21 The mitogenactivated protein kinase (MAPK) is also a critical signalling pathway in the expression of NLRP3. 22 Recent studies have demonstrated that suppressing pyroptosis can alleviate neuronal damage and improve neuropathy outcomes after resuscitation. 23 However, research on neural function and pyroptosis after CA is still insufficient.
Mesenchymal stem cells (MSCs) have great potential in treating inflammation-related diseases. Recent research has shown that MSCs harvested from various tissues, such as bone marrow, adipose tissue, umbilical cord blood and dental pulp, possess the ability to home inflamed tissue. [24][25][26] Among these, bone marrow-derived MSCs (BMSCs) are easy to obtain, have low antigenicity and are nearly non-cytotoxic or oncogenic, making them a promising therapy against neurological disorders. 27,28 BMSCs release abundant neurotrophic factors and cytokines and activate signalling pathways to repair the I/R-induced cell death, which potentially reduces brain damage and promotes nervous system recovery. [29][30][31] Previous studies have found that transplantation with BMSCs relieves brain pathology and neurofunctional disturbance in CA/CPR rat model. 32,33 Our recent study demonstrated that transplantation with BMSCs attenuates brain damage in a rat CA model by decreasing levels of inflammatory factors and increasing levels of anti-inflammatory cytokines. 34 Although few studies have examined the effect of BMSCbased therapies on reducing neural pyroptosis, there is evidence that BMSCs transplantation can improve neurological function. [35][36][37] However, recent studies have shown that transplanted cells can be limited by the global and regional tissue microenvironment, decreasing the efficacy of BMSCs. [38][39][40] Therefore, improving the viability of transplanted cells in the injured brain is crucial. Hypoxic preconditioning of BMSCs effectively increases cell survival during cerebral ischaemia, 41,42 as it can improve the cell's capacity to tolerate ischaemic areas. 43 In addition, BMSCs can maintain their undifferentiated state and show migratory behaviour towards brain lesions under hypoxic culture conditions. 44,45 These findings suggest that HP may be an effective preconditioning method before transplantation. However, previous studies have only focused on localized brain ischaemia, and there have been no studies on the protective effect of hypoxic preconditioned BMSCs (HP-BMSCs) against CA and post-resuscitation. Notably, in vivo bioluminescence imaging (BLI) can accurately track the survival and migration of transplanted normoxic/hypoxic preconditioned BMSCs in the ventricles of the rat brain in a spatiotemporal manner. Therefore, our study investigated the effect of HP on the neuroprotective potential of BMSCs in treating whole cerebral I/R injury.
CA/CPR was induced by asphyxia, and HP-BMSCs were cultured with oxygen-glucose deprivation (OGD)-injured neurons. We hypothesized that HP improves the efficacy of BMSCs, alleviating neurological dysfunction and relieving post-resuscitation brain injury by inhibiting NLRP3 inflammasome-mediated pyroptosis. Through this study, we furthered the understanding of the role of neural cell pyroptosis and the therapeutic effects of BMSCs against CA. How HP improves the therapeutic effects of transplanted BMSCs was also investigated. The results of our study are expected to provide valuable insights into the potential of HP-BMSCs therapy as a treatment for post-resuscitation brain injury.
All The experimental procedures for BMSC therapy are shown in Figure 1A. The protocol for this study was approved by the Animal Welfare and Ethics Committee of Fujian Medical University (license no. IACUC FJMU 2022-0577).
The rats were randomly allocated into the following five groups: sham group (n = 15): rats underwent similar surgery without CA/ CPR; CPR group (n = 18); rats underwent 8 min of asphyxial cardiac arrest and received CPR without treatment; CPR + PBS group (n = 18), in which rats received 15 μL PBS via lateral ventricle 2 h after ROSC; CPR + normal-cultured BMSCs (N-BMSCs) group (n = 18), in which BMSCs were injected into lateral ventricle 2 h after ROSC and CPR + HP-BMSCs group (n = 18), in which the lateral ventricle was injected with HP-BMSCs 2 h after resuscitation. The five groups were F I G U R E 1 Experimental workflow and physiological surveillance data. (A) The schematic illustration represented the experimental plan. (B) Representative waveforms of femoral arterial pressure (red line) and electrocardiogram (green line) during the process of model establishment. CPR, cardiopulmonary resuscitation; ROSC, the return of spontaneous circulation (horizontal scaling = 1000:1). divided into three subgroups: 6, 12 and 24 h, with six test rats in each subgroup, and each sham subgroup had five rats. All efforts were made to minimize the consumption and pain of animals in the study.

| Isolation and characterization of BMSCs
Isolation and culture of BMSCs were performed according to a previously published protocol but with minor modifications. 46 Briefly, young rats were rapidly anaesthetized with sevoflurane and sacrificed by dislocating cervical vertebrae. BMSCs were collected from the femur and tibia by flushing the medullary canal with Dulbecco's Modified Eagle Medium (DMEM)/F12 (Gibco). The medium was centrifuged at 1500 rpm for 5 min to isolate the BMSCs. The cells were resuspended in 4 mL DMEM/F12 containing 10% foetal bovine serum (Clark) and cultured in 25 cm 2 culture flasks in an incubator (Sanyo) at 37°C and 5% CO 2 . The cells were detached with 0.25% trypsin-ethylenediaminetetraacetic acid (EDTA) (Gibco) when they reached 90% confluence. Passage 3 (P3) cells were identified using flow cytometry (BD Bioscience) as described previously, 34 using the following antibodies: anti-CD29-FITC, CD45-FITC, anti-CD90-PE and anti-CD11b-PE (all from BioLegend). Wellgrowing cells at passages 3-8 were used in subsequent experiments.

| Transfection of BMSCs with luciferase lentiviral vectors
Before transplantation or hypoxic preconditioning, half of the BMSCs in each normoxic/hypoxic treated subgroup (6, 12 and 24 h, n = 6) were suspended in 3.6 mL medium and infected with 400 μL of 1 × 10 8 transduction units (TU)/mL lentiviral vectors carrying luciferase (Genepharma). The cultures were maintained in 25 cm 2 flasks at 37°C with 5% CO 2. The other three cell samples were not labelled with lentivirus.

| Determination of cell viability
Cell viability was determined using the Cell Counting Kit-8 (CCK-8, Meilun Biotech). Briefly, P3 cells were inoculated in 96-well plates at a density of 2 × 10 3 cells per well and incubated overnight. Cells were then maintained under normoxic conditions or exposed to hypoxia, as described above, for 48 h. Four hours before hypoxia exposure was terminated, CCK-8 solution (10% in the growth medium, 100 μL/well) was added, and cells were incubated for an additional 4 h. Absorbance was measured at 450 nm using a microplate reader (BioRad).

| Determination of cell apoptosis
The apoptosis of BMSCs was evaluated using an Annexin V-FITC/ propidium iodide (PI) apoptosis detection kit (Meilun Biotech).
Briefly, cells were treated with or without HP and then dissociated and washed twice with PBS. The cell suspensions were incubated with 5 μL Annexin V-FITC and 5 μL PI at room temperature for 15 min in the dark. The cells were analysed using a flow cytometer (BD), and the apoptosis was quantified using Accuri C6 Plus software (BD). All experiments were performed five times.

| Transwell migration assay
Cell migration was evaluated using a 24-well Transwell chamber with 8 μm filter inserts (Corning). Briefly, normoxic/hypoxic BMSCs (1 × 10 5 cells) were plated in the upper layer with 200 μL serum-free medium, and 1 mL of complete medium supplemented with 10% FBS was added to the lower compartment. Cells were cultured at 37°C, 5% CO 2 for 24 h. Subsequently, a wet cotton swab was used to wipe off non-migrated cells from the upper layer, and migrated cells on the lower surface were fixed in 4% paraformaldehyde for 15 min. The cells were then stained with 0.1% crystal violet (Solaibao) and observed under an inverted microscope (Leica). The migrated cells were counted using the ImageJ software (NIH). Three random fields were selected from each sample, and the average was calculated.

| Establishment of CA/CPR model
The rats were starved overnight but with access to water. The rats were initially anaesthetized with sevoflurane (Abbvie) and then deeply anaesthetized with pentobarbital (45 mg/kg) by intraperitoneal injection. Anaesthesia was maintained by giving additional pentobarbital (10 mg/kg) as needed.
A 14-gauge tracheal catheter (BD) was inserted through the orotracheal tube, and the cannula was connected to a small animal ventilator (Acott Biotech) with the following settings: ventilation frequency at 100 breaths/min; tidal volume of 6.5 mL/kg, and inhaled oxygen concentration of 21%. A polyethylene-30 (PE-30) tube (SDR Scientific) was inserted into the right femoral artery and connected to a pressure transducer for continuous monitoring of mean arterial pressure (MAP). The catheter was flushed with heparinized saline (2.5 IU/mL) when necessary. Heart rhythm was monitored by a standard lead II electrocardiogram. The MAP and electrocardiogram signals were recorded and digitized via the Power lab data acquisition system (AD Instrument) and continuously collected for later analysis.
According to our previous study, 34 CA is caused by 6-min asphyxia of less pronounced brain damage symptoms. Therefore, in the present study, CA time was extended to 8 min. The rat's temperature was measured through a rectal probe and maintained between 36.5°C and 37°C by a thermostat (RWD Life Science).
After establishing a stable baseline for 5 min, vecuronium (Xianju Pharmaceutical, 1 mg/kg) was administered through the right femoral artery, and asphyxia CA was induced by switching off the ventilator. CA was defined as MAP less than 25 mmHg in combination with pulseless electrical activity, ventricular fibrillation or cardiac arrest. The circulatory arrest induction was performed for 3-4 min ( Figure 1B). Ventilation assistance was provided after 7 min 45 s of cardiac arrest with 100% O 2 , and the remaining parameters were unchanged. Chest compressions were initiated when CA lasted for 8 min, and manual chest compressions were maintained at a rate of 200 times per min. The compression depth was adjusted to maintain a MAP ≥20 mmHg. Epinephrine (0.05 mg/kg) and heparinized saline (0.5 IU in 0.1 mL) were administered via the femoral artery catheter 1 min after the start of CPR, ROSC was defined as MAP ≥60 mmHg, lasting for more than 5 min. If ROSC did not occur within 2 min, the CPR was stopped. The respiratory machine was kept on to provide 100% O 2 for 1 h after ROSC. The inhaled oxygen concentration was reduced to 21% for 1 h. The endotracheal tube and catheter were removed, and the skin was sutured. Each rat received an intraperitoneal injection of 0.1 million units of penicillin to prevent infection after surgery. The operation was performed by the same operator to reduce experimental variations. All animals were operated under a sterile environment, and no infected wounds were observed. The rats were returned to their individual cages, and corncob was used for bedding. Rats meeting any of the following criteria were excluded from the study: (1) Rats that failed to achieve ROSC within 2 min of CPR initiation; (2) Rats that were difficult to wean from the ventilator 2 h after ROSC and (3) Rats that died prior to sampling.

| Treatment processes
Two hours after resuscitation, the rats in the CPR + PBS group re-

| In vivo imaging of luciferaseexpressing BMSCs
The survival and distribution of the transplanted cells were analysed using an in vivo imaging system (IVIS, PerkinElmer). Rats were anaesthetised with pentobarbital and received D-luciferin potassium salt (PerkinElmer, 150 mg/kg) intraperitoneally at 6, 12 and 24 h after ROSC. Images were captured 20 min after luciferin administration, and cell survival was observed in the brain ventricle. The luminescence of BMSCs was measured using Living Image software (PerkinElmer).

| Specimen collection
Neurological outcomes were accessed at 6, 12 and 24 h after ROSC, while rats were under sevoflurane anaesthesia. The animals were sacrificed by cervical dislocation immediately after blood samples were obtained through heart puncture. The rats were then transcardially perfused with ice-cold physiological saline to remove residual blood until no red perfusate was effused. The brains were quickly removed and placed on ice for dissection. The right frontal cortex of each rat was used for electron microscopy. The right parietal cortex was carefully collected for histopathological and immunofluorescence staining analysis. Meanwhile, the left frontal cortical tissues were homogenized, and the supernatants were used for enzymelinked immunosorbent assay (ELISA). The left parietal cortex was saved to detect the expressions of target proteins via western blotting. Operators involved in detection and data collection were blinded to the experimental treatments.

| Neurological function assessment
Neurological function was evaluated at each time point according to a previously described protocol. 47 The neurological deficit scores (NDS) were based on assessments of overall behaviour, brainstem function, motor function, sensory function, movement behaviour and signs of seizures. Scores ranged from 0 to 80, with a score of 80 indicating normal neurological function and a score of 0 indicating brain death. Lower scores indicated severe neurological deficits.

| Histological evaluation of brain cerebral tissues
Brain tissues were fixed in 4% paraformaldehyde overnight, embedded in paraffin after dehydration with gradient alcohol, and cut into 5μm-thick coronal sections. Sectioned slides were stained by haematoxylin-eosin according to the protocol and examined by light microscopy.

| Immunofluorescence assay
The brain sections were first dewaxed and then underwent antigen

| ELISA
The levels of IL-1β and IL-18 in the left frontal cortex samples were measured using rat IL-1β and IL-18 ELISA kits (Cloud-Clone) according to the manufacturer's instructions. In addition, the concentrations of serum S100B and neuron-specific enolase (NSE) were measured using the rat S100B test kit (Cloud-Clone) and the rat NSE reagent kit (Cloud-Clone), respectively. Optical densities at 450 nm were measured using a microplate reader C (Rayto Life Science).

| Western blotting
First, cortical tissues were washed three times with PBS to remove the blood and cut into small pieces. Next, tissue samples were lysed and centrifuged, and the resulting supernatant was collected as the total protein solution. Total protein concentration was determined using a bicinchoninic acid (BCA) protein assay kit (Beyotime).

| Electron microscopy examination
The right frontal cortical tissues were fixed in 2.5% glutaraldehyde

| Establishment of the OGD/R model
The primary cultured neurons were washed twice with PBS and cultured in a glucose-free DMEM medium (Gibco). Subsequently, the neurons were exposed to oxygen-glucose deprivation (OGD) in an anaerobic chamber (Biospherix) for 2 h, with a gas mixture of 95% N 2 and 5% CO 2 . After the OGD period, the neurons were returned to their original medium and co-cultured with normoxic/hypoxic BMSCs.

| Co-culture of injured neurons with BMSCs under different oxygen levels
For the co-culture experiment, normoxic or hypoxic BMSCs were seeded on the upper layer of the Transwell chamber (Corning) at a density of 5 × 10 4 cells/well and cultured for 4 days in a hypoxic chamber (3% O 2 , 5% CO 2 and 92% N 2 ). Next, the DMEM/F12 medium was replaced with a Neurobasal medium. Neurons that had undergone OGD were cultured in the lower layer of the Transwell with a Neurobasal medium. The cells were co-cultured for 24 h at 37°C with 5% CO 2 . This part of the experiment involved seven groups Specific pathway inhibitors were added to the lower Transwell chamber simultaneously with co-culture.

| Statistical analysis
The Shapiro-Wilk normality test was used to test for normality. All experiments were replicated at least three times. Data that satisfied the normality conditions were presented as means ± standard deviations (SD). Statistical analysis and plotting were performed using SPSS 19.0 (IBM) and GraphPad Prism 9.0 software (GraphPad Prism Software). For normally distributed data, comparisons between two groups were analysed by Student's t-test, while one-way anova was used to compare data from more than two groups. In cases where the Brown-Forsythe test showed data variances were not equal, the Brown-Forsythe anova test was performed, followed by Dunnett's T3 multiple comparisons test. p < 0.05 were considered statistically significant.

| The effect of hypoxic conditions on cell migration in BMSCs
Observation using an inverted microscope showed that BMSCs at  Figure 2C). These results confirmed that the purity of BMSCs was higher than 99%, appropriate for the subsequent experiments. Cells were then exposed to normoxia (21% O 2 ) or hypoxia (3% O 2 ) for 48 h, and the CCK-8 assay showed that cells cultured in hypoxia for 48 h did not exhibit higher proliferation ( Figure 2D).
Subsequently, the apoptosis rate was measured by flow cytometry assay using Annexin V/PI labelling, and the results indicated that the apoptotic rate in BMSCs cultured under hypoxic conditions did not increase, and there was no significant statistical difference between N-BMSCs and HP-BMSCs ( Figure 2E,F). Finally, the Transwell assay was conducted to evaluate the cell migratory capacity under hypoxic conditions. The results revealed that HP significantly increased the migration of BMSCs ( Figure 2G,H; p < 0.05). These findings suggested that non-fatal hypoxic preconditioning for 48 h enhanced the migratory ability of BMSCs, which would make HP-BMSCs more favourable for engraftment than N-BMSCs.

| Physiological parameters at baseline and during the CPR modelling
In the current study, there were 72 male rats were included who achieved ROSC. These rats were randomly selected and assigned to four groups, each consisting of 18 rats. Each group was further divided into three subgroups based on the sample collection time points (6, 12 and 24 h after ROSC, n = 6 per subgroup). Additionally, 15 rats were selected as the sham group, with five rats in each subgroup.
There were no statistically significant differences in the physiological parameters among the five groups. Moreover, the time from asphyxia to cardiac arrest, adrenaline dosage, time from CPR to achieve ROSC and duration of hypoxia was similar in all groups, with no statistically significant differences ( Table 1).  In addition, S100B and NSE were considered biomarkers of brain injury following CPR. 48 Serum S100B and NSE levels were measured to evaluate brain damage following global cerebral ischaemiareperfusion injury. At 12 and 24 h post-resuscitation, ELISA reported that the blood S100B and NSE levels were lower in the CPR + HP-BMSCs group compared to the CPR + PBS group ( Figure 4C,D; p < 0.05). Similarly, lower serum S100B and NSE expressions were detected in CPR + HP-BMSCs group than in CPR + N-BMSCs at 12 h and 24 h after ROSC (p < 0.05). Also, NSE levels significantly reduced in CPR + HP-BMSCs group at an earlier time point (p < 0.05). These findings indicated that HP-BMSCs treatment might prevent the brain from ischaemia-reperfusion injury following resuscitation and is better than N-BMSCs transplantation.

| Hypoxic preconditioned BMSCs transplantation alleviated neuronal pyroptosis in the cerebral cortex after CPR
The expressions of NLRP3 and cleaved-caspase-1 were detected using double-label immunofluorescence staining ( Figure 5A Membrane pore development is a common aspect of pyroptosis. We used TEM to examine the ultrastructural changes in cortical neurons at 6, 12 and 24 h after resuscitation ( Figure 7A).
After whole-brain I/R injury and CA/CPR, pores emerge in the neurons' plasma membrane, and a membrane integrity breakdown ( Figure 7B). Damaged neuronal boundaries were blurred and irregular, and mitochondrial disruption ensued. Nevertheless, according to previous findings, intracerebroventricular HP-BMSCs injection significantly reduced this trend 12 h after resuscitation compared to N-BMSCs injection. Our results also showed that HP-BMSCs reduced NLRP3 inflammasome-mediated cytokine release and thus protected the cerebral cortex from CPR-induced whole-brain injury.  Note: Data are expressed as means ± SD. Abbreviations: CPR, cardiopulmonary resuscitation; ROSC, the return of spontaneous circulation.

| Hypoxic preconditioned BMSCs treatment prevented neurological impairment by repressing activation of NF-κB and MAPK signalling pathways
To better understand the impact of HP-BMSCs on neurological dysfunction after ROSC, we investigated the protein expression levels of HMGB1, TLR4 and NF-κB in the cortex ( Figure 8A). Western blotting results confirmed that the proteins above significantly increased following CA compared to those in the control group (p < 0.05).
Compared to the CPR or CPR + PBS groups, treatment with HP-BMSCs reversed the CA-induced alternation in HMGB1, TLR4 and NF-B 24 h after resuscitation (p < 0.05). Moreover, compared to the CPR + N-BMSCs group, the expression levels of these proteins began to decrease significantly after only 12 h following HP-BMSCs implantation ( Figure 8B,C,F, p < 0.05).
Subsequently, we evaluated the expressions of p38 MAPK and JNK in the cortex to better understand the impact of HP-MSCs transplanted into the lateral ventricle on the MAPK signalling pathway of CA/CPR rats ( Figure 8A). Our findings showed that p38 and JNK expression levels were considerably higher in CA rats compared to the sham group (p < 0.05). In contrast, intracerebroventricular injection of BMSCs resulted in a slight decrease in p38 and JNK expression, although this was not statistically significant (p > 0.05).
Concurrently, this decline was significant for HP-BMSCs treatment against the CPR + PBS group (p 0.05), and these proteins declined significantly at earlier time points (Figure 8D,E). It is important to note that the signalling pathway-associated proteins followed the same pattern as the pyroptosis-related proteins in response to HP-BMSCs therapy (Replicate WBs are shown in Figure S1).

SB203580 (p38 MAPK inhibitor), SP600125 (JNK inhibitor) and
Bay11-7082 (NF-B inhibitor) were added to the co-culture system further to examine the pathway and mechanism of HP-BMSCs on neurons. Cultured cells were stained positively for MAP2 (a neuronal marker, Figure 9A) in immunofluorescence labelling, showing that most cultured cells were neurons. As shown in Figure 9B, The serum S100B and neuron-specific enolase (NSE) levels were reduced with intracerebroventricular HP-BMSCs injection (n = 5-6). All data are presented as means ± SD. *p < 0.05 vs. the sham group. # p < 0.05 vs. the CPR + PBS group. & p < 0.05 vs. the CPR + normal-cultured BMSCs (N-BMSCs) group.
neurological function recovery post-transplantation in the CA/CPR model. This mechanism most probably involves HMGB1/TLR4/NF-B and MAPK signalling pathways simultaneously in time ( Figure 10).
Brain dysfunction caused by CA is the primary reason contributing to substantial morbidity and mortality following an initially successful CPR due to whole-body I/R injury. According to statistics, the survival rate to discharge for those receiving out-of-hospital CPR is only 10.4%, and the survival rate with a favourable neurological prognosis is 8.2%. 55 Due to the elevated impairment and mortality rates from CA in this population, further in-depth study into post-CA syndrome's physiological and pathophysiological processes is essential. 56 Programmed cell death is one crucial pathogenic factor for  F I G U R E 1 0 Schematic diagram illustrating the proposed mechanism by which hypoxic preconditioned bone marrow-derived mesenchymal stem cells (HP-BMSCs) alleviate neuronal pyroptosis via the NLRP3 inflammasome, MAPK and NF-κB signalling pathways following cardiopulmonary resuscitation (CPR). The expression level of HMGB1 and TLR4 was upregulated following resuscitation, including MAPK pathway activation and pyroptosis initiation. This triggered the nuclear translocation of NF-κB and increased HMGB1 leading to NLRP3 inflammasome formation. Subsequently, NLRP3 inflammasome activation promoted the cleavage and activation of pro-caspase-1, which contributed to elevation in IL-1β, IL-18 and GSDMD-N levels, and induced pore formation on the membrane, causing pyroptotic cell death. Hypoxic preconditioning (HP) enhanced the tolerance capacity of BMSCs to local adverse microenvironment. HP-BMSCs inhibited HMGB1 and TLR4 activation, which decreased MAPK and NF-κB signals, as well as preventing the formation of NLRP3 inflammasome, which in turn ameliorated neuronal pyroptosis.
Our non-lethal hypoxic culture increased cell migration, according to an in vitro Transwell experiment. Cell migration after injection is a crucial factor in the effectiveness of MSCs transplantationbased therapy. 84 Here, we quantitatively measured luciferase signals expressed in BMSCs to evaluate cell survival and migration in vivo. To the best of our knowledge, we reported, for the first time, the application of living imaging for BMSCs-based therapy in the CPR model. Bioluminescent monitoring was characterized by a high signal-to-noise ratio and playing function only at the mammalian cell temperature of nearly 37°C. 85 The generation of bioluminescence was significantly dependent on the host and grafted cells, and only live elicited a bioluminescent reaction. 86 As a result, the biolumines-  102 Also, NF-κB, an essential transcription regulatory factor, is crucial in the pathogenesis of neuroinflammation. 103 Research has demonstrated that MSCs inhibit cerebral cell apoptosis by downregulating the NF-κB signalling pathway after brain damage. 104,105 In an I/R stroke model, TREM-1 can activate CARD9/ NF-κB and NLRP3/caspase-1, initiating microglia pyroptosis and releasing inflammatory factors. Another group has demonstrated that stroke-induced white matter damage and microglial pyroptosis can be partially attenuated by suppressing the NF-κB/NLRP3 signalling pathway. 106 The latest study has indicated that the NF-κB signalling pathway regulates pyroptosis in microglia post-CA. 107 Hence, the NF-κB pathway was strongly connected to cell pyroptosis in nervous system diseases. Consist with studies mentioned above. According Although our findings validated the neuroprotective benefits of HP-BMSCs, additional study is needed to determine the best culture conditions. Second, our research did not address the molecular mechanism of HP leading to enhanced BMSC migration, which needs to be studied more in the future. Third, since many cell injections likely predispose to vascular embolization or local hematoma, we performed only single injections and only evaluated short-term neurological function after ROSC.
Consequently, the effects of multiple HP-BMSCs injections on long-term neurological outcomes still require further study. Fourth, we did not detail the reasons for failing to achieve ROSC. Based on our observation, we found that the disappearance of cardiac electrical activity probably caused this during the no-flow time. Finally, we did not perform the rescue experiment on HP-BMSCs and signalling pathway activators. Despite being readily available, MAPK and NF-κB activators have an unfavourable effect on the ROSC rate in our model, causing high animal attrition. Three-time points were selected to observe the same trend for pyroptotic and signalling pathway-related proteins, reducing experimental variations.
In summary, the vast majority of current research studies are still at the experimental stage, indicating that it will take a significant amount of time before the translation of HP-BMSCs into clinical applications can be achieved. Nevertheless, the use of HP-BMSCs represents a promising novel approach for cerebral protection post-resuscitation.

| CON CLUS ION
Our results showed that transplanted HP-BMSCs alleviated CPRinduced brain damage associated with the inhibition of NLRP3 Conceptualization (lead).

ACK N O WLE D G E M ENTS
We appreciate the assistance of experimenter Jiuyun Zhang from Fujian Provincial Key Laboratory of Emergency Medicine. We also thank the Home for Researchers editorial team for the language editing service.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors have no financial conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data and models in this study are included in the submitted article.