Lipidomics Analysis Reveals a Protective Effect of Myriocin on Cerebral Ischemia/Reperfusion Model Rats

Ceramide accumulation has been associated with ischemic stroke. Myriocin is an effective serine palmitoyltransferase (SPT) inhibitor that reduces ceramide levels by inhibiting the de novo synthesis pathway. However, the role of myriocin in cerebral ischemia/reperfusion (I/R) injury and its underlying mechanism remain unknown. The present study established an experimental rat model of middle cerebral artery occlusion (MCAO). We employed ultra-performance liquid chromatograph quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF/MS)–based lipidomic analysis to identify the disordered lipid metabolites and the effects of myriocin in cerebral cortical tissues of rats. In this study, we found 15 characterized lipid metabolites involved in sphingolipid and glycerophospholipid metabolism in cerebral I/R-injured rats, and these alterations were significantly alleviated by myriocin. Specifically, the mRNA expression of metabolism-related enzyme genes was detected by real-time quantitative polymerase chain reaction (RT-qPCR). We demonstrated that myriocin could regulate the mRNA expression of ASMase, NSMase, SGMS1, SGMS2, ASAH1, ACER2, and ACER3, which are involved in sphingolipid metabolism and PLA2, which is involved in glycerophospholipid metabolism. Moreover, TUNEL and Western blot assays showed that myriocin plays a key role in regulating neuronal cell apoptosis. In summary, the present work provides a new perspective for the systematic study of metabolic changes in ischemic stroke and the therapeutic applications of myriocin.


Introduction
Cerebral ischemia/reperfusion (I/R) injury is caused by blood supply restriction and followed by the restoration of blood flow and re-oxygenation (Peralta et al. 2013), and it typically occurs during the treatment stage of ischemic disease, aggravating irreversible damage to brain tissue (Obadia et al. 2017;Yu et al. 2018). Numerous studies have confirmed that ischemic/reperfusion of the brain leads to a cascade of pathological mechanisms, mainly involving neuronal excitotoxicity, oxidative stress, ionic imbalance, and so on (Khandelwal et al. 2016;Lai et al. 2014;Schaller et al. 2003). In clinical practice, reperfusion treatments based on thrombolysis and mechanical thrombectomy are major therapeutic strategies for ischemic stroke patients. The risk of cerebral I/R injury following reperfusion therapy is a critical challenge for patient outcomes (Cas et al. 2020). Moreover, in view of the limited clinical therapeutic options for ischemic stroke, the development of efficient therapeutic strategies and potential drugs to treat ischemic stroke is urgently required.
Lipids are essential components of cellular membranes and play a vital role in cell development and death (Wu et al. 2018). Compared with other mammalian organs, the brain not only contains extremely abundant lipids, but a large diversity of lipid species is identified (Fitzner et al. 2020). Notably, growing evidence have indicated that lipid metabolism disorders are tightly associated with the occurrence of disease (Doria et al. 2011;Naudí et al. 2015;Orei et al. 2012). Ceramides, which are bioactive sphingolipids, comprise a critical class of signalling molecules that act as second messengers in sphingolipid signal transduction (Fyrst and Saba 2010). Previous studies have shown that ceramide accumulation can be observed after brain injury, which supportd a distinct link between ceramides and ischemic stroke (Bhuiyan et al. 2010;Novgorodov et al. 2015;Yu et al. 2007a, b).
Myriocin is a powerful inhibitor of the serine palmitoyltransferase (SPT), which catalyzes the first step of ceramide de novo synthesis (Wadsworth et al. 2013). Furthermore, myriocin (Fig. 1A) is a new small-molecule immunosuppressant extracted from Mycelia sterilia, Isaria sinclairii, and Cordyceps cicadae (Wadsworth et al. 2013). In particular, myriocin has been extensively utilized for studying certain diseases related to sphingolipid metabolism (Cheng and Lee 2016;Lee et al. 2011). Emerging studies have demonstrated myriocin treatment may serve as an effective therapeutic strategy for myocardial I/R injury (Bonezzi et al. 2019;Reforgiato et al. 2016). However, the underlying protective mechanism exerted by myriocin against ischemic stroke has not been reported.
Lipidomics presents a comprehensive analysis of lipid metabolic profiles in various biological samples, which provides a novel approach to identify key metabolic biomarkers and pathways (Sun et al. 2019). Importantly, due to the homeostatic alterations in the body, a multitude of abnormal Fig. 1 Myriocin attenuated cerebral I/R injury in rats. A The chemical structure of myriocin. B Neurological deficit scores were assessed after 24 h of rat cerebral I/R (n = 6). C Percentage of infarct volume in whole cerebral tissue (n = 3). D TTC staining of the brain tissue. E Observation of pathological changes by Nissl's staining in each group (magnification × 200, n = 3). Data are expressed as the mean ± SD. Compared with the sham group, # P < 0.05; Compared with the I/R group, *P < 0.05 lipid metabolic activities have been observed during the progression of ischemic stroke (Au 2018). In our study, a systemic lipidomics study based on ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry (UPLC-Q-TOF/MS) has certain significance for studying the mechanism by which myriocin intervenes in the pathophysiology of cerebral I/R injury. We also determined the expressions levels of relevant lipid metabolism-related genes by real-time quantitative polymerase chain reaction (RT-qPCR) analysis. Additionally, the underlying apoptosis mechanism of myriocin concerning lipid metabolites was further explored.

Experimental Animals and Drug Administration
Male Sprague-Dawley rats weighing 250-300 g (SPF grade, license No. SCXK (Yu) 2017-0001) were obtained from the Experimental Animal Center of Zhengzhou University (Zhengzhou, China). The Ethics Committee approved animal experiments of Zhengzhou University (approval number: ZZU-LAC20200807). Animals were housed with free water and diet in colony cages on a 12 h light/dark cycle. Rats were randomly assigned into three groups with 20 rats for each group: the sham group, the I/R group, and the myriocin-treated group.
Cerebral ischemia/reperfusion rat model was performed by middle cerebral artery occlusion (MCAO) as previously described (Mavroudakis et al. 2020). In brief, the left middle cerebral artery (MCA) of rats was occluded by inserting a 5-0 monofilament long nylon suture into the ICA for 2 h and then to reperfusion for 22 h. Rats in I/R and myriocin group were performed with ischemia/reperfusion surgery, while the sham group received the same surgical procedure except there was no thread insertion.
Myriocin was administered intraperitoneally for 7 days with 0.3 mg/kg dissolved in DMSO before MCAO in rats. The sham and model groups were given an equal volume of dimethylsulfoxide (DMSO) for injection in the same way.

Sample Preparation and Lipid Extraction
After the behavioral assessment, rats were sacrificed and the ischemic side of cerebral cortical tissue samples were quickly extracted. Total lipids are extracted from brain homogenate using a modified Folch method (Orei et al. 2012). Briefly, the brain tissues (n = 6 per group) were weighed precisely (50.0 ± 0.5 mg) and then 1 mL of chloroform/methanol (2:1, v/v) was added. The mixture was added ice-cold water to induce phase separation, and then centrifuged at 13,000 rpm for 10 min. Next, the layer containing chloroform was collected and dried under a stream of N2. Finally, the extracts were reconstituted in acetonitrile for further lipid analyses. Meanwhile, quality control (QC) samples were prepared by mixing equivalent lipid extracts of each cerebral cortical tissue and injected every six samples during the analysis to monitor the stability of the system and the method.
Mass spectrometry (MS) analysis was performed with an electrospray ionization (ESI) source in positive and negative ion modes. ESI source conditions were set as follows: the nebulizer gas pressure, curtain gas, and temperature were set at 55 psi, 35 psi, and 500 °C, respectively. The ion spray voltage and declustering potential were set at ± 4500 V and ± 100 V, respectively. The first-order spectrum was obtained by full scanning, and the secondary spectra were obtained by information-dependent acquisition (IDA) mode. The scanning range of the first mass spectrometry and the secondary mass spectrometry were m/z 100-2000 and 50-2000, and the collision voltage was ± 45 eV in positive mode and negative mode.

Evaluation of Neurological Deficit Scores and Cerebral Infarct Volume
Neurological deficits were assessed at 24 h after reperfusion with Zea Longa 5-point scale (Longa et al. 1989) to evaluate the MCAO model. Briefly, 0 points, no neurological deficit; 1 point, forelimb flexion; 2 points, circling to the contralateral side; 3 points, falling towards the contralateral side when walking; and 4 points, difficult to walk. The higher score represents the more severe behavior disorder in rats. The model was considered successful when the score was 1-3. Brains (n = 3 per group) were quickly placed on ice and cut into five transverse slices. After that the brain slices were stained with 2% TTC (2,3,5-triphenyltetrazolium chloride, Sigma) at 37 °C for 30 min, and then fixed in 4% paraformaldehyde for overnight. All stained slices were photographed and the infarct volume was quantified by Image-pro plus 6.0 software.

Nissl Staining and TUNEL Staining
Nissl staining was performed to evaluate the morphological changes in the brain. According to previously described methods (Zhang et al. 2002), the rat brain tissues were fixed in 4% paraformaldehyde fixed overnight, embedded in paraffin and cut into coronal sections. The brain sections were stained with Nissl staining solution (Beyotime, China) and observed under a fluorescence microscope (magnification × 200). The TUNEL Apoptosis Assay Kit (Beyotime, China) was used to examine cell apoptosis in rat brain tissue for TUNEL staining. Briefly, nuclei were stained by DAPI dye after TUNEL staining. Then, the brain sections were observed under a fluorescence microscope (magnification × 200), and the cells with green fluorescence were considered as apoptotic cells. Quantification of positive cells was performed by Image J software. TUNEL-positive cells (%) = (apoptotic cells/total cells) × 100%.

Reverse Transcription Quantitative PCR (RT-qPCR) Analysis
Total RNA was extracted from the cerebral cortex tissue with Trizol reagent (Invitrogen, USA) and then reversely transcribed to cDNA using PrimeScript RT reagent (Thermo Fisher Scientific, USA), and RT-qPCR reaction was performed in triplicate using SYBR Green PCR Master Mix (Applied Biosystems). The amplification procedures were as follows: 95 °C for 10-min denaturation, followed by 40 cycles of 95 °C for 15 s, and 60 °C for 60 s. Data were measured using the 2 −ΔΔCt method using GAPDH as the reference gene. All gene analyses were done at least three times. The primer sequences used for RT-qPCR are listed in Table S1.
After being washed three times with TBST, the membranes were next incubated with appropriate secondary antibodies for 1 h at room temperature. GAPDH served as an internal control. The membranes were visualized with Bio-Rad ChemiDoc system and protein amount was quantified by Image J software.

Statistical Analysis and Multivariate Data Analysis
GraphPad Prism 8.3 (GraphPad Prism Software, San Diego, CA, USA) software was used for statistical analysis. The data were analyzed using one-way ANOVA for multiple group comparisons. All data were expressed as the mean ± SE (standard error). Values with P < 0.05 were considered significant.
ProteoWizard software converted the original data into mzXML format, and then off-line XCMS software was used for peak alignment, retention time correction, and peak area extraction. The data matrices after 80% principle screening and normalization were input into SIMCA-P 14.1 software (Umetrics, Sweden) for multivariate analysis, including principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA). In addition, the OPLS-DA models were validated by the response values of the 200 permutation tests, which were applied to prevent overfitting. Variable importance for the projection (VIP) values was obtained from the OPLS-DA. The selected metabolites data were performed with Metaboanalyst 5.0 (http:// www. metab oanal yst. ca/) used for potential metabolic pathway analysis and clustering heatmap analysis.

Myriocin Attenuated Cerebral I/R Injury in Rats
To evaluate the protective effect of myriocin on cerebral I/R model, neurological scores and TTC staining were performed to assess neurological function and infarct volume, respectively. As shown in Fig. 1B, the neurological scores of the I/R group were markedly increased compared to sham group (P < 0.05). Compared with the I/R group, myriocin treatment decreased the neurological defect scores (P < 0.05). After staining with 2% TTC solution, the normal brain tissue of each group was stained red while the infarcted area was colored pale (Fig. 1C). The quantitative statistical results of infarct volumes showed that the cerebral infarct volume in I/R group was increased significantly compared to sham group (Fig. 1D, P < 0.05). Compared with the I/R group, myriocin treatment attenuated the infarct volume (P < 0.05).
Nissl staining showed the pathological changes of the ischemic area in cerebral I/R model rats. As shown in Fig. 2E, compared with the sham group, the I/R group showed neural cells were poorly arranged and became slightly swollen. Additionally, Nissl bodies decreased markedly; some of the nucleoli were shrunken and deep stained. After intervention with myriocin, the neuron damage was significantly reduced (P < 0.05).

Myriocin Inhibited Ceramide Accumulation of the De Novo Synthesis Pathway in the Cerebral Cortex of Rats with Cerebral I/R Injury
We then assessed SPT enzyme expression, which is the target of myriocin. The qRT-PCR results showed that I/R significantly increased the mRNA levels of SPTLC2, which was successfully decreased by myriocin treatment (Fig. 2A). However, there were no significant differences in SPTLC1 mRNA expression among groups. Similarly, we found consistent results on SPTLC1 and SPTLC2 protein levels via Western blotting (Fig. 2B). Overall, the above finding suggested that ceramide accumulation induced through a transcriptional activation of the de novo synthesis pathway in cerebral I/R-injured rats, and this effect could be inhibited by myriocin.

Multivariate Statistical Analysis
UPLC-Q-TOF/MS acquired the lipid profiles of cerebral cortex samples. Figure S1 shows a representative total ion chromatogram (TIC) of QC group in positive and negative ion modes. Multivariate statistical analysis of PCA and OPLS-DA were applied to reveal the lipid metabolic differences among groups. From the processed data, PCA analysis was used to distinguish the differential regulation of lipid species between the compared groups according to the similarity of data. In our study, a clear separation between each group was observed in PCA score plots, and the QC samples displayed significant aggregation (Fig. 3). The results showed that myriocin group was separated from the model group with approaching the sham group. It indicates that cerebral I/R injury could be effectively reversed by myriocin intervention. Additionally, the OPLS-DA was used to analyze the data in order to further separate the differences between groups and facilitate subsequent search for potential biomarkers. In our study, as illustrated the establishment of the scatter plots of OPLS-DA models established, the separation of Sham group and I/R group was obvious. Besides, the Myriocin group and the I/R group were clearly separated (Supplementary, Fig. S2A, C). The 200 permutation tests were used to validate the performance of the OPLS-DA models, and there was no overfitting phenomenon (Supplementary, Fig. S2B, D).

Potential Biomarkers and Metabolic Pathway Analysis
The structure of the metabolites were determined with MS/MS information including m/z, retention time, and characteristic fragments. Metabolites identified by comparing with authentic standard and database such as HMDB (http:// www. hmdb. ca/), MassBank (http:// www. massb ank. jp/), and METLIN (http:// metlin. scrip ps. edu/) and the database established by us. When VIP > 1and P < 0.05, the metabolites were exhibited statistically significant, which were identified as potential biomarkers. A total of 15 potential biomarkers were finally selected, as listed in Table 1. Compared with the sham group, there were six different lipid metabolites that were significantly upregulated in the I/R group (P < 0.05), including Cer Fig. 2 Myriocin inhibited ceramide accumulation of the de novo synthesis pathway in the cerebral cortex of rats with cerebral I/R injury. A Relative mRNA levels of SPTLC1 and SPTLC2 of each group were determined by quantitative real-time. B Protein lev-els of SPTLC1 and SPTLC2 of each group detected by Western blot. GAPDH was used as an internal control. Data are presented as mean ± SD (n = 3). Compared with the Sham group, ## P < 0.01; Compared with the I/R group *P < 0.05, **P < 0.01 (d18:1/16:0), Cer (d18:1/18:0), Cer (d18:0/18:0), C18 Sphingosine, LPE (18:0), and LPC (20:4). The other nine different lipid metabolites were significantly downregulated in the I/R group (P < 0.05). After myriocin treatment, all the above 15 lipid metabolic disturbances were effectively ameliorated.
To reveal the metabolic processes of the lipid metabolites, we performed pathway enrichment analysis using MetaboAnalyst 5.0 ( detailed information of the metabolic pathways is shown in Table S2. A clustering heatmap was conducted to more visually exhibit the trend of the variation of 15 differential metabolites among the three groups. The obtained clustering results in the form of heatmap are shown in Fig. 4A. Each little square represents a sample, red color represents the increase, and blue color represents the decrease in intensity, respectively. As demonstrated in Fig. 4B, the most relevant metabolic pathways regulated by myriocin against cerebral I/R injury is sphingolipid metabolism and glycerophospholipid metabolism with the impact value > 0.1.

Myriocin Influenced the Expression Levels of Enzymes Genes Involved in the Pathways of Altered Metabolites
To investigate the role of myriocin in sphingolipid and glycerophospholipid metabolism, we determined the expression of relevant enzymes that were involved in cerebral I/R injury in rats, including ASMase, NSMase, SGMS1, SGMS2, ASAH1, ASAH2, ACER2, ACER3, and PLA2 by RT-qPCR analysis. As a result, eight out of nine genes showed a differentially changed mRNA expression. As seen in Fig. 5, the mRNA expression levels of ASMase, NSMase, SGMS2, ASAH1, ACER2, ACER3, and PLA2 were upregulated obviously, while SGMS1 was downregulated drastically in the I/R group when compared with the sham group. Upon the myriocin treatment, the mRNA levels of ASMase, NSMase, SGMS2, ASAH1, ACER2, ACER3, and PLA2 were all significantly decreased whereas SGMS1 was increased when compared with the I/R group (P < 0.05). However, the level of ASAH2 showed no significant difference among groups.

Myriocin Inhibited Relative Apoptotic Molecular Expression Levels
TUNEL staining was adopted to quantify the number of TUNEL-positive cells from the rat cerebral cortex. The results showed that compared with the sham-operated rats, apoptotic cells in the I/R group were significantly increased and remarkably decreased after myriocin treatment (Fig. 6A,  B). Moreover, to further explore the potential anti-apoptotic mechanism induced by myriocin, we next estimated the protein expression of the apoptosis-related proteins by western blot analysis (Fig. 6C, D). The results showed that myriocin decreased the protein expressions of Bax, caspase-3, and cleaved-caspase-3, whereas increased the protein expressions  Fig. 4 Potential biomarkers and metabolic pathway analysis. A Heatmap visualization of key metabolite expression of the brain regulated by the myriocin. B Metabolic pathway enrichment analysis of the brain regulated by the myriocin of Bcl-2 compared with the I/R group (P < 0.05). These results suggested that myriocin inhibited apoptosis in the cerebral cortex of rats with cerebral I/R injury.

Discussion
Collectively, lipidomics and gene expression analysis indicated that myriocin treatment induced complex responses from multiple interconnected metabolic pathways (Fig. 7). The potential causal mechanisms of cerebral I/R injury in rats can be better understood by analyzing the metabolic network.
In this study, we verified that myriocin could significantly decrease cerebral infarction, neurological deficits, and pathological changes in the cerebral cortex of rats, which clearly support a neuroprotective role for myriocin against cerebral I/R injury in MCAO rats. It is interesting to note that myriocin was found to significantly reduce SPTLC2 but not SPTLC1 gene and protein expression following injury, indicating that SPTLC2 may be the target mediating the effect of myriocin. Besides, our lipidomics analysis revealed that a total of 15 potential lipid biomarkers between groups were identified, but more importantly, these lipids play essential roles in glycerophospholipid and sphingolipid metabolism.
Sphingolipids play a critical physiological role in maintaining the structure of cell membranes and are extensively involved in regulating various biological signal transduction processes, such as cell growth, apoptosis, and signal transduction (Grösch et al. 2018;Sun et al. 2016). Meanwhile, sphingolipids are mainly composed of ceramide (Cer), sphingomyelin (SM), and sphingosine (Hannun et al. 2001). Among these lipids, ceramide is the central component of sphingolipid metabolism and has emerged as a second messenger involved in regulating the physiological activities and metabolism of cells (Hannun and Obeid 2008). The involvement of ceramide accumulation in the pathogenetic mechanism of ischemic stroke have been observed in previous studies, which is consistent with our observation (Liu et al. 2000;Takahashi et al. 2004;Yu Z et al. 2007a, b). In our study, we found that the long-chain ceramides, Cer (d18:1/16:0) and Cer (d18:1/18:0), were significantly increased in the I/R group when compared with the sham group. However, notably, we also found a very-long-chain ceramide, Cer (d18:1/24:1) was markedly decreased. Emerging data have indicated that ceramides with distinct chain lengths showed different responses to disease and regulated different physiological processes (Cha et al. 2016;Chan and Goldkorn 2000;Chao et al. 2019).Based on the above research results, we speculated that long-chain and verylong-chain ceramides may play different biological roles in cerebral I/R injury.
Ceramide levels are regulated by complex metabolic pathways; in the present study, we aimed to clarify the effects of inhibition of de novo ceramide synthesis by myriocin on other metabolic mechanisms of ceramide in rats with cerebral I/R injury. Ceramides are produced via hydrolysis of SM by sphingomyelinases (SMases) and are converted to SM by sphingomyelin synthase (SGMS) (Tian et al. 2009). In our results, a decreased trend of SM (d18:0/18:0) and SM(d18:0/20:0) was observed after cerebral I/R injury in rats. We detected the mRNA expression levels of key sphingomyelin cycle-related enzymes (ASMase, NSMase) tended to increase and SGMS1 mRNA expression levels tended to decrease after cerebral I/R injury in rats, which might explain the reduction in the ceramide/sphingomyelin ratio. Additionally, there was an increase in C18 sphinganine level in the I/R group, which is one of the intermediate molecules in ceramide metabolism. Sphinganine is a well-known mediator of cell-growth arrest and apoptosis (Cuvillier 2002). In brain tissue, ceramides are hydrolyzed into sphingosine under the action of ceramidase, which is encoded by four genes, including acid ceramidase (ASAH1), neutral ceramide (ASAH2), alkaline ceramidase 2 (ACER2), and alkaline ceramidase 3 (ACER3). In the present study, the increased mRNA expressions of ceramide conversionrelated enzymes (ASAH1, ACER2, ACER3) after cerebral I/R injury in rats might account for the accumulation of C18 sphingosine. Remarkably, myriocin could ameliorate the dysregulation of sphingolipid metabolism at the level of the above metabolic enzyme genes, which was first investigated in cerebral I/R injury in this study. Meanwhile, a significant increase of DAG (16:0/18:1) level was also observed in the I/R group. SMS enzymes have been shown to convert ceramide and phosphatidylcholine (PC) into sphingomyelin and diacylglycerol (DAG), respectively (Jatooratthawichot et al. 2020). According to our results, a higher level of DAG (16:0/18:1) might be associated with increased mRNA expression of SGMS2, which needs more research.
Glycerophospholipids are the principal lipid components of biological membranes, which are involved in diverse signal transduction processes such as apoptosis and membrane fusion (Farooqui et al. 2000). In animal tissues, PC and phosphatidylethanolamine (PE) are the most abundant and important intermediates in glycerophospholipid metabolism (Vance 2015). In our study, PCs (PC (16:0/18:1), PC (18:1/18:1), PC (18:0/20:4)) and PEs (PE(P-18:0/20:1), PE (16:0/22:6)) were significantly decreased, whereas LPC (20:4) and LPE (18:0) were significantly increased in the Fig. 5 Myriocin influenced the expression levels of enzymes genes involved in the pathways of altered metabolites. RT-qPCR analysis of the enzymes gene involved in sphingolipid and glycerophosphospholipid metabolism of the cerebral cortical tissue of rats cortex. Data are presented as mean ± SD (n = 3). Compared with the Sham group, ## P < 0.01; Compared with the I/R group *P < 0 .05, **P < 0.01 ◂ I/R group. After myriocin treatment, the levels of these lipid metabolites showed a trend toward the normal status. A previous study found that the plasma levels of PE and PC in patients with ischemic stroke were significantly lower than those in the control group, which was consistent with our findings (Ma et al. 1998). Phospholipids and lysophospholipids can interconvert through the "Lands cycle" to maintain lipid homeostasis (Wu et al. 2016).
As is known to all, PC and PE in organisms can be hydrolyzed by phospholipase A2 (PLA2) to produce the corresponding single-stranded lysophospholipids, lysophosphatidylcholine (LPC) and lysophosphatidylethanolamine (LPE), respectively. Studies have also demonstrated that activation of PLA2 plays a pivotal role in acute cerebral ischemia and is linked to neuronal cell death, which is in line with the results of our study (Sun et al. 2009). Notably, our study showed that myriocin could ameliorate glycerophospholipid metabolism dysregulation as a result of PLA2, which is related to the balance of Lands cycle in the cerebral I/R model rats.
Importantly, apoptosis is the key factor for neuronal cell death in cerebral I/R injury (Radak et al. 2017). In our study, we observed the morphological changes in neuronal cells in the cortical brain regions through TUNEL staining. Then, Fig. 6 Myriocin inhibited relative apoptotic molecular expression levels. A Representative images of TUNEL staining of rat brain tissue. B The quantification of TUNEL positive cell in each group (magnification × 200, n = 3). C The protein levels of Bax, Bcl-2, caspase-3, and cleaved caspase-3 were analyzed by Western blot. D The quantification of relative protein in each group. Data are presented as mean ± SD (n = 3). Compared with the sham group, ## P < 0.01; Compared with the I/R group *P < 0.05, **P < 0.01 we further explored the underlying mechanisms of myriocin on inhibiting cortical neuronal apoptosis. Our results clearly confirmed that myriocin downregulated the protein levels of Bax, caspase-3, and cleaved caspase-3 while upregulating Bcl-2, thus promoting the survival of neural cells after cerebral I/R injury. It is recognized that ceramides are known to be involved in apoptosis (Hannun 1996). On the basis of our present results, we speculated that myriocin might regulate ceramide and eventually exerted anti-apoptotic effect in ischemic brain, but it needs more experiments to validate our hypothesis.

Conclusions
Taken together, the present work provided a new perspective for a systematic study of the therapeutic effects of myriocin in ischemic stroke. A UPLC-Q-TOF/MS-based lipidomics approach was firstly applied in the present study to elucidate the neuroprotective effect of myriocin against cerebral I/R injury. Combined lipidomics and RT-qPCR analysis results, this study revealed that myriocin could improve the disruption of 15 lipid metabolites by regulating the expression of relevant enzyme genes in cerebral cortical tissue of rats. Studies also shown that the protective effect of myriocin might be mediated by inhibiting the apoptosis response induced by cerebral I/R injury. Therefore, the findings of this study suggest that myriocin may serve as a promising therapeutic drug to alleviate cerebral I/R injury. Fig. 7 Disturbed lipid metabolic regulatory network in cerebral I/R model rats and the interventional effects of myriocin. The red and blue fonts represent the related enzymes were increased or decreased, respectively, in I/R group compared with sham group. The related enzymes reversed by myriocin are marked with up and down arrows