Ischemic stroke is initiated by the interruption of cerebral blood flow and is followed by reperfusion, which further induces neuronal death, leading to brain tissue injury. Apoptosis is the primary mechanism of cell death and is regulated by several signaling pathways[19]. Investigation of novel treatment strategies to target cell death is still needed. Related studies have found that histone methylase inhibitors and histone demethylase inhibitors have effects in the treatment of ischemic stroke[20]. However, they have rarely been reported to be related to the apoptotic pathway of ischemic stroke. In this study, we demonstrated that the EZH2 inhibitor GSK-126 alleviated neuronal apoptosis induced by ischemic stroke. Moreover, MAPK/ERK pathway activation was required for this process (Fig. 6).
Methylation levels of H3K27 are mediated by the histone methylase EZH2 and the histone demethylase JMJD3. H3K27me3 is involved in a variety of biological processes by inhibiting gene expression. However, the role of H3K27me3 in neuronal death has been less well investigated. A previous report revealed that JMJD3 rescues apoptosis in hippocampal neurons by upregulating the expression of BDNF, and JMJD3 plays an important role in the expression of Bax and Caspase-3 during oxygen–glucose deprivation (OGD) injury [21, 22]. In addition, we previously found that elevation of H3K27me3 downregulated the expression of Na+/Ca2+ exchanger 3 (NCX3) in hippocampal neurons, while GSK-126 restored expression levels of NCX3[11]. NCX3 is a gene that reduces neuronal death after ischemic brain damage [23]. Combined with our observation in I/R rats, these results indicate that the increasing H3K27me3 levels in the brains of I/R rats were independent of the expression of EZH2 (Fig. 1). It would be interesting to examine whether an EZH2 inhibitor such as GSK-126 protect neurons from ischemic brain injury-induced neuron death.
GSK-126 was originally developed as a more effective inhibitor of EZH2[24]. It exhibits higher selective inhibition of EZH2 than other EZH2 inhibitors, such as DZNep and EPZ005687[25]. Moreover, GSK-126 has entered phase I clinical trials[26]. However, low oral bioavailability and blood-brain barrier permeability issues with GSK-126 were observed in animal models, which may limit its use in the central nervous system[27]. In contrast, reports have suggested that large doses of intraperitoneally administered GSK-126 reduce levels of H3K27me3 in the brain[11]. Therefore, to investigate the role of H3K27me3 in the ischemic brain, GSK-126 was administered to rats by intracerebroventricular injection in the present study. A time course experiment was performed and revealed that levels of H3K27me3 in the hippocampus were reduced in response to 7 consecutive days of intracerebroventricular injection of GSK-126 (Fig. 2A-C). Rats preadministered GSK-126 exhibited attenuated H3K27me3 levels (Fig. 2D-F) and more surviving neurons in response to I/R (Fig. 3).
The mechanism by which GSK-126 affects neuronal death is still unknown. According to the ChIP-seq analysis, the essential genes bound and differentially enriched by H3K27me3 span a multitude of various genes (Fig. 4A). Because GSK-126 inhibited levels of H3K27me3, we primarily analyzed the difference in H3K27me3 enrichment between I/R rats preadministered GSK-126 and I/R rats, along with control animals. Interestingly, we found that the three groups shared 9112 identical genes (Fig. 4B). These identical genes indicate that most genes affected by increasing levels of H3K27me3 after ischemia were reversed by GSK-126. Therefore, GO analyses were performed to detect the potential relationship between these genes and the pathology of ischemic stroke. The results revealed that negative regulation of the execution phase of apoptosis was the most significantly altered GO, and compared to ischemic rats, ischemic rats preadministered GSK-126 negatively regulated antiapoptotic genes (DFFA, NMAT1, PAM16, BCL2L1, FZD3, and CXCR3) (Fig. 4C, Table S1). Among these genes, NMAT1 and BCL2L1 are responsible for reducing cell death and apoptosis after ischemic stroke[28, 29]. CXCR3, the receptor of CXCL10, is known to reduce brain infarction and attenuate BBB disruption in stroke[30]. In addition, the known causes of neuronal apoptosis after stroke include reactive oxygen species (ROS), calcium overload, excitatory toxicity and inflammatory reactions. ROS are one of the early and most important components of cerebral brain ischemia, and excessive production of ROS leads to oxidative injury, including DNA damage, protein oxidation, and lipid peroxidation, ultimately leading to apoptosis[31]. After ischemic stroke, ATP deficiency leads to Ca2+ overload, which can induce neuronal apoptosis[32, 33]. Moreover, cerebral ischemia results in release of large amounts of glutamate, which stimulates NMDARs and induces calcium influx through these ionotropic receptors. The calcium-dependent activation of death-signaling proteins that are immediately downstream of the receptors triggers excessive signaling cascades that work cooperatively to induce neuronal death[34]. Therefore, excitotoxicity is considered one of the most important mechanisms of neuronal apoptosis after ischemia[35]. In the present study, the GO results were consistent with the above research. Altered G protein-coupled glutamate receptor signaling pathways were also identified between sham rats and I/R rats pretreated with GSK-126 (Figure S1, Table S3).
The KEGG pathway analysis suggested that the MAPK signaling pathway was one of the most significantly altered pathways. The MAPK pathway consists of three subpathways: ERK-MAPK, JNK-MAPK and p38-MAPK. ERK1/2, JNK, and p38 can induce both cell survival and cell death in response to ischemic stroke[19]. Among them, ERK1/2 was reported to play an antiapoptotic role in the process of ischemic brain injury, while JNK and p38 promote apoptosis after ischemia[36, 37]. Here, we found that 40 genes in the MAPK pathway were affected by H3K27me3 enrichment (Fig. 4D-E, Table S2). In addition, several key genes (BDNF, MEKK2, MYC, etc.) in the ERK1/2 pathway were included in these 40 genes. ERK can be activated by BDNF, MEK is an upstream protein of ERK, and MYC is activated by ERK[19, 38, 39]. This evidence suggests that GSK-126 may activate the entire ERK1/2 signaling pathway in the brains of I/R rats.
To test the above hypothesis, rats were administered the MEK inhibitor U0126 after 7 days of GSK-126 injection to block the ERK1/2 pathway before ischemia modeling. As expected, the role of GSK-126 in protecting CA1 neurons from ischemia-induced apoptosis was antagonized by administration of U0126 (Fig. 5). Levels of cleaved caspase 3 further confirmed this conclusion. Our results indicate that activation of the ERK1/2 pathway may contribute to the reduction of neuronal apoptosis after ischemia. Nevertheless, whether ERK activation promotes or inhibits neuronal apoptosis is still uncertain. Some reports indicate that inhibition of the ERK signaling pathway by U0126 inhibits apoptosis after ischemia[40, 41]. These conflicting results may be due to the complexity of the MAPK signaling pathway and the inconsistency among experimental conditions. In our research, it would be helpful to further study the impact of H3K27me3 on the JNK-MAPK and p38-MAPK pathways.
In addition, the GABAergic synapse pathway was among the top 10 significantly altered pathways in I/R rats preadministered GSK-126 compared to I/R rats (Fig. 4D-E), which is in line with the findings that GABA has a protective effect in ischemic stroke[42, 43]. During an ischemic episode, the extracellular cerebral GABA concentration increases. Related studies have revealed that GABAergic signaling is enhanced through pharmacological intervention just hours after stroke; neuroprotection can be achieved by reducing the excitotoxic index and lowering the release of glutamate [44, 45]. These reports further validate the accuracy of our ChIP-seq data. Moreover, they also suggest that studying the epigenetic regulation of the GABAergic signaling pathway may contribute to in-depth understanding of these excitotoxicity effects.
In conclusion, we demonstrated that the EZH2 inhibitor GSK-126 alleviates neuronal apoptosis in ischemic stroke by inhibiting H3K27me3. Furthermore, activation of the ERK-MAPK pathway is required for the neuroprotective function of GSK-126 in the ischemic brain.