Melatonin promotes recovery from ischemic stroke by modulating microglia polarization and inhibiting oligodendrocyte pyroptosis via the RORα/AMPKα/STAT1 pathway

Abstract

Background

This study investigates how melatonin affects white matter damage and oligodendrocyte pyroptosis after ischemic stroke by modulating microglia polarization through the RORα/AMPKα/STAT1 pathway.

Methods

A mouse model of middle cerebral artery occlusion (MCAO) was used, with melatonin (20 mg/kg) administered post-reperfusion and daily for 14 days. Neurological function and white matter damage were assessed, along with gasdermin D (GSDMD) and caspase-1 expression to evaluate pyroptosis. In vitro, a Transwell co-culture system of microglia and oligodendrocytes exposed to oxygen-glucose deprivation (OGD) was employed. The role of RORα in modulating microglia polarization and oligodendrocyte pyroptosis was explored using siRNA-mediated knockdown and AAV-based RORα shRNA microinjection.

Results

Melatonin reduced white matter injury, inhibited oligodendrocyte pyroptosis, and improved sensorimotor function. These effects were linked to reduced APC/caspase-1 and APC/GSDMD double-positive cells and were dependent on RORα signaling. Melatonin also shifted microglia from a pro-inflammatory to an anti-inflammatory phenotype, an effect reversed by RORα knockdown. In vitro, melatonin inhibited OGD-induced pyroptosis via the RORα/AMPKα/STAT1 pathway.

Conclusions

These findings suggest that melatonin promotes long-term recovery by reducing neuroinflammation and protecting white matter integrity through RORα signaling, highlighting its potential as a therapeutic agent for stroke recovery.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12967-026-08596-y.

Introduction

Stroke, a prominent cause of global morbidity and mortality with escalating disability rates in developing countries, poses significant challenges for healthcare and society due to its long-term adverse outcomes [1]. Effective rehabilitation, aimed at symptom improvement and restoration of social and vocational functions, is crucial to mitigate the associated socio-economic burdens [2, 3]. A pivotal aspect of neurological recovery involves repairing intracranial white matter injuries, essential for maintaining white matter integrity, a process facilitated by rehabilitation-induced white matter remodeling during the subacute stroke phase [4, 5]. Recent research underscores the importance of oligodendrocytes in white matter protection, with a focus on inhibiting oligodendrocyte pyroptosis as a potential therapeutic strategy [6–8]. White matter, accounting for over half of the human brain, is vital for neural signal transmission and network formation, comprising axons, glial cells, myelin-producing oligodendrocytes, astrocytes, and microglia. Post-stroke white matter damage results from demyelination, axonal degeneration, and neuroinflammation, highlighting the vulnerability of myelin to inflammation [9–11]. Thus, therapies enhancing anti-inflammatory factors while suppressing pro-inflammatory factors may provide effective protection against white matter injuries [12].

Microglia are the resident immune cells in the brain, constituting approximately 10–15% of the total cell population [13]. They play a crucial role in maintaining brain immune homeostasis by continuously monitoring their surrounding microenvironment [14, 15]. When a stroke occurs, the infarcted areas become infiltrated with deleterious inflammatory cells, exacerbating cerebral injury. Emerging evidence suggests that microglia can exhibit either damaging or beneficial phenotypes, depending on the context [16]. In the acute phase, typically within the first week after a stroke, microglia tend to adopt a pro-inflammatory phenotype, which can contribute to further brain damage. Conversely, during the chronic phase, microglia are more likely to shift toward an anti-inflammatory phenotype, which supports tissue repair and recovery. Therefore, guiding microglia polarization toward the anti-inflammatory phenotype is critical for protecting the ischemic brain [17, 18].

Melatonin (N-acetyl-5-methoxy-tryptamine), initially identified as a hormone secreted by the pineal gland, is best known for its role in regulating the sleep/wake cycle and circadian rhythms. However, melatonin’s functions extend beyond its endocrine role; it is also recognized as an immunomodulator and neuroprotective agent [19]. In our previous studies, we observed that melatonin could promote the polarization of microglia and macrophages toward an anti-inflammatory phenotype, thereby offering protection against white matter injuries [20]. This neuroprotective effect of melatonin is particularly relevant given the emerging evidence that oligodendrocyte pyroptosis contributes to inflammatory damage in white matter [21]. The selection of melatonin for this study stems from its diverse biological effects, apt for addressing stroke’s complex pathophysiology. While its anti-inflammatory microglial polarization role is recognized, the mechanisms underlying melatonin’s regulation of oligodendrocyte pyroptosis remain elusive. This investigation is pivotal, given the incomplete understanding of oligodendrocyte cell death pathways under ischemia [22]. By exploring how melatonin influences these processes, this study aims to shed light on potential therapeutic strategies for mitigating white matter damage after stroke.

The present study found that melatonin can reduce cerebral ischemic injury after stroke through various mechanisms. Our previous research confirmed that melatonin partially modulates microglial polarization by a STAT3-dependent pathway, contributing to neuroprotection [20]. Additionally, Ding et al. revealed that melatonin, acting through a RORα-dependent mechanism, regulates the AMPKα-STAT pathway, thereby directing macrophage polarization, enhancing plaque stability, and reducing plaque rupture [23]. Furthermore, another study demonstrated that melatonin inhibits the TLR4/NF-κB signaling pathway, promotes the proliferation of endogenous oligodendrocyte precursor cells (OPCs), and suppresses IL-1β protein expression, ultimately reducing white matter lesions in a rat model of focal cerebral ischemia [24]. These findings suggest that melatonin regulates the polarization and phenotype of microglia and macrophages through multiple signaling pathways, thereby offering protection against cerebral ischemic injury. However, whether melatonin can reduce oligodendrocyte pyroptosis during stroke rehabilitation remains to be elucidated. Therefore, this study aimed to explore the role of melatonin in the long-term post-stroke recovery. Specifically, we sought to investigate the potential involvement of microglial polarization and oligodendrocyte pyroptosis, mediated by the RORα/AMPKα/STAT1 signaling pathway, in the long-term neuroprotective effects of melatonin.

Result

Melatonin treatment enhanced long-term sensorimotor recovery by protecting against white matter injury after ischemia

To evaluate the effects of melatonin on white matter injury and long-term sensorimotor recovery after stroke, we induced MCAO in mice and administered melatonin (20 mg/kg, intraperitoneally) immediately after reperfusion. The treatment was continued daily for 14 days post-surgery (Fig. 1A). White matter pathology was assessed using double immunofluorescence staining with anti-MBP (myelin basic protein) and anti-SMI-32 (a marker of axonal damage) antibodies in the cortex (CTX), corpus callosum (CC), and striatum (STR) at 21 days post-MCAO (Fig. 1B, C). Compared to the sham group, vehicle-treated stroke mice exhibited increased SMI-32 immunostaining and a loss of MBP-positive myelin, resulting in a higher SMI-32/MBP ratio. However, melatonin treatment significantly reduced white matter injury, as indicated by a lower SMI-32/MBP ratio in melatonin-treated mice (Fig. 1D-F). The neuroprotective effect of melatonin was further supported by behavioral assessments performed pre-MCAO and at 3, 10, and 21 days post-MCAO to track sensorimotor recovery. Melatonin-treated mice exhibited significant improvements in sensorimotor function as early as 3 days post-stroke, with these benefits sustained throughout the 21-day recovery period (Fig. 1G, H). To explore the relationship between white matter integrity and behavioral outcomes, we analyzed the correlation between the SMI-32/MBP ratio and corner test performance. A positive correlation was found, with melatonin treatment leading to a marked reduction in poor behavioral outcomes (Fig. 1I-K). In contrast, a negative correlation was observed between the SMI-32/MBP ratio and rotarod performance, with melatonin-treated mice displaying longer retention times on the rotarod, indicating better motor coordination and recovery (Fig. 1L-N). These results demonstrate that melatonin treatment not only protects against white matter damage but also promotes long-term functional recovery following ischemic stroke, with significant improvements observed from the acute phase through to later recovery stages.

Fig. 1.

Melatonin reduced white matter damage and promoted long-term sensorimotor recovery in stroke mice. (A) Experimental flow chart. (B) Representative brain sections from the cortex (CTX), corpus callosum (CC), and striatum (STR) stained with SMI-32 and MBP. (C) Zonal brain map. (D-F) Relative quantification of the SMI-32/MBP ratio in the CTX, CC, and STR. ^***P < 0.001. Data are presented as the mean ± SD. n = 8 mice per group, with 1 section per animal and 3 FOVs per section. One-way ANOVA followed by Bonferroni post hoc test. (G, H) Performance in the corner and Rota-rod tests. ^*indicates Sham vs. MCAO + Vehicle; ^#indicates MCAO + Vehicle vs. MCAO + Melatonin. ^***P < 0.001, ^#P < 0.05, ^##P < 0.01, ^###P < 0.001. Data are presented as the mean ± SD. n = 8 mice per group. Two-way ANOVA followed by Bonferroni post hoc test. (I-K) Correlation between white matter injury (SMI-32/MBP) and corner test performance. (L-N) Correlation between white matter injury (SMI-32/MBP) and Rota-rod test performance. n = 8 mice per group. The Pearson product linear regression analysis was used to correlate white matter injury and behavioral tests

Melatonin treatment decreased oligodendrocyte pyroptosis after brain ischemia in mice

To assess whether melatonin can mitigate ischemia-induced pyroptosis, we performed double immunostaining for the oligodendrocyte marker APC and the pyroptotic markers Caspase-1 and GSDMD in the CTX, CC, and STR of mice at 21 days post-MCAO. MCAO induced a significant increase in Caspase-1 and GSDMD expression, indicating elevated pyroptosis in these regions (Fig. 2A-C). However, melatonin treatment markedly inhibited the expression of both Caspase-1 and GSDMD, suggesting suppression of oligodendrocyte pyroptosis. Quantification of the ratio of APC and Caspase-1 double-positive cells to total APC-positive cells revealed a significant reduction in the melatonin-treated group compared to the vehicle group across all peri-infarct regions, including the CTX, CC, and STR (Fig. 2D-F). Similarly, double immunostaining for APC and GSDMD showed a consistent trend, with melatonin treatment substantially reducing GSDMD-positive oligodendrocytes, further supporting the inhibition of pyroptosis (Fig. 2G-I). These results indicate that melatonin exerts a protective effect on white matter by inhibiting oligodendrocyte pyroptosis, likely contributing to the suppression of pro-inflammatory responses following ischemic stroke.

Fig. 2.

Melatonin reduced oligodendrocyte pyroptosis 21 days after brain ischemia. (A) Representative images of the CTX, CC, and STR immunostained with APC antibodies to label oligodendrocytes (OL) and Caspase-1 antibodies to detect pyroptosis. (B) Representative images of the CTX, CC, and STR immunostained with APC antibodies to label OL and GSDMD antibodies to detect pyroptosis. (C) Zonal brain map. (D-F) Quantification of the relative number of APC/Caspase-1 double-positive cells within the APC⁺ areas in brain sections from the sham, MCAO + Vehicle, and MCAO + Melatonin groups. (G-I) Quantification of the relative number of APC/GSDMD double-positive cells within the APC⁺ areas across the same groups. ^***P < 0.001. Data are presented as mean ± SD. n = 8 mice per group, with 1 section per animal and 3 FOVs per section. One-way ANOVA followed by Bonferroni post hoc test

Melatonin treatment shifted microglia polarization from pro-inflammatory to anti-inflammatory phenotype in the ischemic brain of mice

To assess the effects of melatonin on microglial polarization, we examined pro-inflammatory and anti-inflammatory microglial states at 21 days post-MCAO using immunofluorescent markers. Pro-inflammatory microglia were identified as CD16/32/Iba-1 double-positive cells in the CTX, CC, and STR (Fig. 3A, B). MCAO significantly increased the proportion of pro-inflammatory microglia in these regions. However, melatonin treatment markedly reduced the relative number of pro-inflammatory microglia compared to the vehicle-treated group across all peri-infarct areas (Fig. 3C-E). To explore the link between pro-inflammatory microglia and behavioral outcomes, we analyzed correlations between pro-inflammatory marker expression and sensorimotor function. A reduction in pro-inflammatory microglia following melatonin treatment was associated with improved performance in the corner test (Fig. 3F-H). Similarly, melatonin treatment correlated with longer retention times in the rotarod test, showing a negative relationship between pro-inflammatory microglia and motor coordination (Fig. 3I-K). We further examined the relationship between pro-inflammatory microglial markers and oligodendrocyte pyroptosis. CD16/32/Iba-1 expression was positively correlated with oligodendrocyte pyroptosis markers (APC/Caspase-1 and APC/GSDMD double staining) (Fig. 3L-Q). Melatonin treatment reduced both pro-inflammatory microglial activation and oligodendrocyte pyroptosis, suggesting that the anti-inflammatory effects of melatonin contribute to mitigating cell death pathways in the ischemic brain.

Fig. 3.

Melatonin shifted microglia polarization from pro-inflammatory to anti-inflammatory phenotype in ischemic mice 21 days post-injury. (A) Representative images of the CTX, CC, and STR immunostained with CD16/32 antibodies to label pro-inflammatory microglia and Iba-1 antibodies to identify activated microglia. (B) Zonal brain map. (C-E) Quantification of the relative number of CD16/32⁺Iba-1⁺ cells within the Iba-1⁺ areas in brain sections from the sham, MCAO + Vehicle, and MCAO + Melatonin groups. ^***P < 0.001. Data are presented as mean ± SD. n = 8 mice per group, with 1 section per animal and 3 FOVs per section. One-way ANOVA followed by Bonferroni post hoc test. (F-H) Correlation between microglia polarization (CD16/32⁺/Iba-1⁺) and Corner test performance. (I-K) Correlation between microglia polarization (CD16/32⁺/Iba-1⁺) and Rota-rod test performance. n = 8 mice per group. The Pearson product linear regression analysis was used to correlate microglia polarization and behavioral tests. (L-N) Correlation between microglia polarization (CD16/32⁺/Iba-1⁺) and oligodendrocyte pyroptosis (APC⁺/Caspase-1⁺). (O-Q) Correlation between microglia polarization (CD16/32⁺/Iba-1⁺) and oligodendrocyte pyroptosis (APC⁺/GSDMD⁺). n = 8 mice per group. The Pearson product linear regression analysis was used to correlate microglia polarization and oligodendrocyte pyroptosis

The anti-inflammatory phenotype of microglia was evaluated using Arg-1/Iba-1 double staining (Fig. S1A and B). While MCAO induced a modest increase in anti-inflammatory microglia, melatonin treatment significantly enhanced the presence of anti-inflammatory microglia across the CTX, CC, and STR (Fig. S1C-E). Enhanced anti-inflammatory microglia correlated with improved outcomes in both the corner test and the rotarod test (Fig. S1F-K). Furthermore, markers of oligodendrocyte pyroptosis (APC/Caspase-1 and APC/GSDMD) showed a strong negative correlation with the increased presence of anti-inflammatory microglia following melatonin treatment (Fig. S1L-Q). These findings demonstrate that melatonin promotes a shift in microglial polarization from the pro-inflammatory phenotype toward the anti-inflammatory phenotype, partly by inhibiting oligodendrocyte pyroptosis. This shift contributes to the reduction of inflammation and the promotion of functional recovery in the ischemic brain.

Melatonin regulated the STAT1 pathway in a RORα- and AMPKα-dependent manner in the ischemic brain and modulated microglia polarization via the RORα/AMPKα/STAT1 pathway in vitro

To investigate the protective effects and underlying mechanisms of Melatonin on white matter injury (WMI) following ischemic stroke, we conducted a series of bioinformatics analyses, as shown in the figure. First, we screened the melatonin-related dataset GSE179519 from the Gene Expression Omnibus (GEO) database and identified differentially expressed genes (DEGs) using the thresholds of log₂FC ≥ 1 or ≤ − 1 and p-value < 0.05 (Fig. 4A). We then retrieved WMI-related genes from the GeneCards database and intersected them with melatonin-associated DEGs to generate a Venn diagram, yielding 878 enriched genes (Fig. 4B). To elucidate the biological functions of these key genes, KEGG and Gene Ontology (GO) enrichment analyses were performed, which revealed significant enrichment in several critical signaling pathways, including the AMPK signaling pathway and the JAK-STAT signaling pathway (Fig. 4C–D). These findings suggest that melatonin may exert neuroprotective effects after ischemic stroke by modulating the AMPK and JAK-STAT signaling pathways to attenuate inflammation and promote myelin repair, thereby protecting white matter and improving neurological function. This provides novel theoretical insights into the neuroprotective mechanisms of melatonin. Microglia polarization is regulated by several transcription factors, including STATs and PPAR-γ, with STAT1 playing a key role in the RORα-mediated modulation of microglia function. To explore whether melatonin regulates the STAT1 signaling pathway, we analyzed the levels of RORα, phosphorylated STAT1 (phos-STAT1, Tyr701), STAT1, phosphorylated AMPKα (phos-AMPKα, Thr172), and AMPKα in the brains of mice 3 and 21 days post-MCAO. MCAO significantly increased the expression of RORα and phos-AMPKα (Thr172) 3 days post-MCAO (Fig. 4E-G). Melatonin treatment further enhanced the RORα and phos-AMPKα levels (Fig. 4E-G). MCAO significantly suppressed the expression of RORα and phos-AMPKα (Thr172) 21 days post-MCAO (Fig. 4K-M). However, melatonin treatment mitigated this suppression, restoring RORα and phos-AMPKα levels (Fig. 4K-M). Additionally, melatonin reduced phos-STAT1 expression in the brains of mice 3 and 21 days post-MCAO, suggesting inhibition of pro-inflammatory microglia polarization (Fig. 4E, I, K and O). No significant changes were observed in the total levels of AMPKα or STAT1 across the groups (Fig. 4E, H, J, K, N and P). These results indicate that melatonin dynamically modulates RORα/AMPKα signaling across different post-ischemic stages and consistently suppresses excessive STAT1 phosphorylation, thereby influencing microglial polarization after stroke.

Fig. 4.

Melatonin dynamically modulates the RORα/AMPKα/STAT1 pathway in ischemic brain tissues at 3 and 21 days after MCAO. (A) Volcano plot of differentially expressed genes (DEGs). Each dot represents a gene. Red dots indicate significantly upregulated genes (log₂FC ≥ 1, p-value < 0.05), blue dots indicate significantly downregulated genes (log₂FC ≤ − 1, p-value < 0.05), and gray dots represent non-significant genes. The dashed lines denote the threshold boundaries for differential expression. (B) Venn diagram showing the overlap between white matter injury (WMI)-related gene sets and Melatonin-regulated differentially expressed genes. The overlapping region represents the common potential target genes. (C) Bar chart of Gene Ontology (GO) enrichment analysis based on enrichment scores. The top 10 most significantly enriched terms in Biological Process, Cellular Component, and Molecular Function categories are shown. (D) Bubble plot of pathway enrichment analysis. Each bubble represents an enriched GO or KEGG pathway term. Bubble size indicates the number of enriched genes, and color reflects the significance level of enrichment (− log₁₀ adjusted p-value). The X-axis represents the gene ratio (Rich Factor). (E) Representative Western blot images showing the protein levels of RORα, phosphorylated AMPKα (Thr172), total AMPKα, phosphorylated STAT1 (Tyr701), and total STAT1 in brain tissues from Sham, MCAO + Vehicle, and MCAO + Melatonin mice at 3 days after MCAO. GAPDH was used as the loading control. (F–J) Quantitative analysis of RORα, phosphorylated AMPKα, total AMPKα, phosphorylated STAT1, and total STAT1 protein levels at 3 days after MCAO. Data are presented as mean ± SD; n = 5 mice per group. ***P < 0.001. One-way ANOVA followed by Bonferroni post hoc test. (K) Representative Western blot images showing the protein levels of RORα, phosphorylated AMPKα (Thr172), total AMPKα, phosphorylated STAT1 (Tyr701), and total STAT1 in brain tissues from Sham, MCAO + Vehicle, and MCAO + Melatonin mice at 21 days after MCAO. GAPDH was used as the loading control. (L–P) Quantitative analysis of RORα, phosphorylated AMPKα, total AMPKα, phosphorylated STAT1, and total STAT1 protein levels at 21 days after MCAO. Data are presented as mean ± SD; n = 8 mice per group. **P < 0.01, ***P < 0.001. One-way ANOVA followed by Bonferroni post hoc test

Molecular docking/interface analysis suggests a potential spatial association between melatonin and RORα

To further explore whether melatonin may potentially associate with RORα, we performed molecular docking/interface analysis using melatonin as the small-molecule ligand and RORα as the receptor. The predicted model showed that melatonin could be spatially accommodated within a pocket-like region of the RORα structure. Interface analysis identified several RORα residues surrounding melatonin, including Val364, Phe365, Ile366, Arg367, Met368, Cys369, Arg370, Ala371, and Phe372. The interface summary showed an accessible surface area of 434.07 Ų and a buried surface area of 396.09 Ų, with a calculated solvation energy gain of ΔiG = 0.00 kcal/mol (Fig. S6). These findings suggest a potential spatial association between melatonin and RORα. However, because the calculated ΔiG was 0.00 kcal/mol, this computational result should be interpreted cautiously and does not prove direct physical binding.

To further investigate the effects of melatonin on microglia polarization, we conducted in vitro experiments using primary microglia cultures subjected to OGD to simulate ischemic conditions. We performed siRNA-mediated knockdown of RORα to determine whether melatonin’s effects on STAT signaling were RORα-dependent (Fig. S2). RORα knockdown alone significantly increased STAT1 phosphorylation and diminished melatonin’s inhibitory effect on STAT1 phosphorylation during OGD exposure (Fig. 5A, G). We also examined AMPKα phosphorylation, an upstream regulator of STATs, to better understand the mechanism. Melatonin treatment increased RORα expression (Fig. 5A, C) and AMPKα phosphorylation (Fig. 5A, D), but these effects were largely abolished following RORα knockdown. The STAT1 pathway is a critical regulatory mechanism for microglia polarization, with STAT1 phosphorylation being essential for pro-inflammatory microglia activation. Consistent with in vivo results, melatonin significantly suppressed OGD-induced STAT1 phosphorylation (Fig. 5A, G, I) without altering total AMPKα and STAT1 levels (Fig. 5A, E, H). Additionally, melatonin downregulated the expression of pro-inflammatory microglial markers (CD11b/CD16; Fig. 5B, J, K) and upregulated anti-inflammatory markers (CD206/Arg1; Fig. 5B, L, M), indicating a shift in microglial phenotype from pro-inflammatory to anti-inflammatory. These effects were largely reversed following RORα knockdown.

Fig. 5.

Melatonin modulated OGD-induced microglia polarization via the RORα/AMPKα/STAT1 pathway. (A) Representative Western blots showing RORα signaling proteins in primary microglia cells. (B) Representative Western blots of microglia phenotype markers. (C-I) Relative expression levels of RORα, phos-AMPKα (Thr172), AMPKα, phos-STAT1 (Tyr701), and STAT1 in primary microglia cells. (J-M) Relative expression levels of CD11b, CD16, Arg1, and CD206 in primary microglia cells. Eight groups of primary microglia cells were prepared: (1) Scramble + Vehicle; (2) Scramble + Melatonin (400 nM); (3) Scramble + OGD + Vehicle; (4) Scramble + OGD + Melatonin (400 nM); (5) siRORα + Vehicle; (6) siRORα + Melatonin (400 nM); (7) siRORα + OGD + Vehicle; (8) siRORα + OGD + Melatonin (400 nM). (N) Representative Western blot images showing protein levels of phos-AMPKα, AMPKα, CD16 and CD206. (O-R) Quantification of phos-AMPKα, AMPKα, CD16 and CD206 protein expression in the primary microglia from the Vehicle, OGD + Vehicle, OGD + Melatonin, and OGD + Melatonin + Compound C groups. ^*P < 0.05, ^***P < 0.001. Data are presented as mean ± SD. Data are from three independent experiments per group. One-way ANOVA followed by Bonferroni post hoc test

Furthermore, we used the AMPKα inhibitor Compound C to assess whether AMPKα activation is required for melatonin-regulated microglial phenotype switching. Our results showed that melatonin markedly inhibited the expression of CD16 and increased CD206 levels in OGD-induced primary microglia (Fig. 5N, Q, R), whereas AMPKα inhibition with Compound C (5µM) attenuated melatonin-induced neuroprotective effects (Fig. 5N-R). These findings suggest that melatonin regulates microglia polarization through the AMPKα-STAT1 pathway in a RORα-dependent manner, highlighting the potential of melatonin to modulate inflammation and promote recovery in ischemic conditions.

In vitro transwell experiments demonstrated that melatonin regulated microglia polarization via the RORα signaling pathway, indirectly affecting oligodendrocyte pyroptosis and survival

To investigate whether melatonin indirectly modulates the interaction between oligodendrocyte pyroptosis and survival, we performed cell viability and survival assays using a microglia-oligodendrocyte Transwell co-culture system, both with and without OGD (Fig. 6A). OGD stimulation significantly reduced cell viability (Fig. 6B) and increased LDH release (Fig. 6C) compared to the vehicle group. However, melatonin treatment significantly ameliorated post-OGD oligodendrocyte cell death, evidenced by increased cell viability and decreased LDH release. To further explore the effects of melatonin on pyroptosis-related protein markers, we conducted Western blot analysis of primary oligodendrocytes co-cultured with microglia under both non-OGD and OGD conditions (Fig. 6D, E). In the absence of OGD (Fig. 6D), we observed baseline expression levels of pro-Caspase-1, cleaved Caspase-1, full-length GSDMD, GSDMD-N, and GAPDH (as a loading control) in oligodendrocytes, with scramble or siRORα oligonucleotides included to assess their influence on protein expression. Following OGD treatment (Fig. 6E), the expression profiles of pyroptosis-related markers changed significantly. While the levels of pro-Caspase-1 and full-length GSDMD remained unchanged under both normal and OGD conditions (Fig. 6F, H), OGD stimulation led to a marked increase in cleaved Caspase-1 and GSDMD-N levels (Fig. 6G, I). These increases were further amplified by siRORα treatment, indicating the involvement of RORα in regulating pyroptosis. Notably, melatonin treatment reversed the OGD-induced increases in cleaved Caspase-1 and GSDMD-N levels, demonstrating that melatonin inhibits OGD-induced oligodendrocyte pyroptosis via the RORα signaling pathway. These results indicate that melatonin not only mitigates the harmful effects of OGD on oligodendrocyte survival but also regulates pyroptosis through modulation of the RORα signaling pathway, further supporting its potential role in neuroprotection during ischemic conditions.

Fig. 6.

Melatonin modulated microglia polarization via the RORα signaling pathway, indirectly affecting oligodendrocyte pyroptosis and survival. (A) Schematic diagram of the Transwell co-culture system involving microglia and oligodendrocytes treated with OGD at different time points. (B) Oligodendrocyte cell viability measurement. (C) LDH release assay from oligodendrocytes. (D) Representative Western blots of pyroptosis protein markers in oligodendrocytes co-cultured with microglia under non-OGD conditions. (E) Representative Western blots of pyroptosis protein markers in oligodendrocytes co-cultured with microglia after OGD treatment. (F-I) Relative expression levels of pro-Caspase1, cleaved Caspase1, full-length GSDMD (GSDMD-FL), and GSDMD-N in oligodendrocytes. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001. Data are presented as mean ± SD. Data are from three independent experiments per group. One-way ANOVA followed by Bonferroni post hoc test

Melatonin directed microglia polarization from a pro-inflammatory to anti-inflammatory phenotype in the ischemic brain via RORα signaling

At 21 days post-stroke, we evaluated the expression of CD16/32/Iba-1 double-positive cells to identify the pro-inflammatory microglia, markers of pro-inflammatory activation, within ischemic brain regions (Fig. 7A, B). Our results showed that melatonin treatment significantly reduced the proportion of pro-inflammatory microglia (CD16/32⁺/Iba-1⁺ cells). However, when RORα was knocked down using AAV-based shRNA (Fig. S3), the effect of melatonin on reducing pro-inflammatory polarization was abolished. Across all peri-infarct regions, including the CTX, CC, and STR, the pro-inflammatory marker ratio was significantly elevated in the RORα knockdown group versus the melatonin-treated group, highlighting the crucial role of RORα in mediating this effect (Fig. 7C-E).

Fig. 7.

Melatonin reduced pro-inflammatory microglia via RORα signaling 21 days after MCAO in mice. (A) Representative images of the CTX, CC, and STR immunostained with CD16/32 antibodies to label pro-inflammatory microglia and Iba-1 antibodies to identify activated microglia. (B) Zonal brain map. (C-E) Quantification of the relative number of CD16/32⁺Iba-1⁺ cells within the Iba-1⁺ areas in brain sections from the MCAO + shRNA-Control + Vehicle, MCAO + shRNA-Control + Melatonin, MCAO + shRNA-RORα + Vehicle, and MCAO + shRNA-RORα + Melatonin groups. **P < 0.01, ***P < 0.001; n = 5 mice per group. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, ns denotes not significant. Data are presented as mean ± SD. n = 5 mice per group, with 1 section per animal and 3 FOVs per section. One-way ANOVA followed by Bonferroni post hoc test

Next, we analyzed the expression of Arg-1/Iba-1 double-positive cells as markers of the anti-inflammatory microglia phenotype under different melatonin treatment conditions (Fig. S4A, B). Melatonin treatment significantly increased the proportion of anti-inflammatory microglia (Arg-1⁺/Iba-1⁺ cells) in the ischemic brain. However, consistent with the pro-inflammatory results, RORα knockdown weakened melatonin’s effect on promoting anti-inflammatory polarization. Across all peri-infarct regions (CTX, CC, and STR), the anti-inflammatory marker ratio was significantly diminished in the RORα knockdown group compared to the melatonin-treated group, indicating that RORα is essential for facilitating the anti-inflammatory transition (Fig. S4C-E). These findings suggest that melatonin drives microglia polarization from a pro-inflammatory phenotype toward an anti-inflammatory phenotype through a mechanism highly dependent on RORα signaling. By promoting this shift, melatonin may contribute to the resolution of neuroinflammation and the enhancement of functional recovery following ischemic brain injury.

Melatonin inhibited oligodendrocyte pyroptosis via RORα signaling, promoting long-term sensorimotor function recovery in ischemic mice

At 21 days post-stroke, we measured oligodendrocyte pyroptosis markers, specifically APC/Caspase-1, in the ischemic brain regions (Fig. 8A and B). Our findings revealed that melatonin administration markedly decreased the proportion of these markers. Notably, the inhibitory effect of melatonin on oligodendrocyte pyroptosis was abrogated after RORα knockdown. Across all peri-infarct zones, encompassing the CTX, CC, and STR, RORα knockdown led to a significant increase in the ratio of oligodendrocyte pyroptosis markers within the melatonin-treated groups (Fig. 8C-E).

Fig. 8.

Melatonin inhibited oligodendrocyte pyroptosis via RORα signaling in ischemic mice. (A) Representative images of the CTX, CC, and STR immunostained with APC antibodies to label OL and Caspase-1 antibodies to identify OL pyroptosis. (B) Zonal brain map. (C-E) Quantification of the relative number of APC/Caspase-1 double-positivecells within the APC⁺ areas in brain sections from the MCAO + shRNA-Control + Vehicle, MCAO + shRNA-Control + Melatonin, MCAO + shRNA-RORα + Vehicle, and MCAO + shRNA-RORα + Melatonin groups. ^**P < 0.01, ^***P < 0.001, ns denotes not significant. Data are presented as mean ± SD. n = 5 mice per group, with 1 section per animal and 3 FOVs per section. One-way ANOVA followed by Bonferroni post hoc test

To corroborate these findings, we conducted additional animal tests using the oligodendrocyte pyroptosis markers APC/GSDMD, which yielded similar results (Fig. S5A and B). In all peri-infarction regions, the ratio of APC/GSDMD double-positive cells significantly increased following RORα knockdown in the melatonin-treated groups (Fig. S5C-E). At preoperative, as well as at days 3, 10, and 21 following MCAO, assessments were conducted using the Corner test (Fig. S5F) and Rota-rod test (Fig. S5G). The results demonstrated that melatonin facilitates the long-term recovery of sensorimotor function in mice with stroke. This effect, however, can be reversed by RORα gene knockdown. These results suggest that melatonin promotes the recovery of long-term sensorimotor function in ischemic mice by inhibiting oligodendrocyte pyroptosis via RORα signaling.

Discussion

Our findings reveal that melatonin enhances long-term recovery following ischemic stroke by regulating microglia polarization and inhibiting oligodendrocyte pyroptosis via the RORα/AMPKα/STAT1 pathway. Melatonin modulated microglia from a pro-inflammatory phenotype to an anti-inflammatory phenotype, reducing inflammation and promoting white matter integrity. This polarization shift is crucial for mitigating neuroinflammation and tissue damage, supporting the therapeutic potential of melatonin in post-stroke recovery. Maintaining white matter integrity is essential for neurofunctional recovery, as white matter tracts facilitate interhemispheric communication and sensorimotor coordination [25–27]. Our results align with earlier studies that highlight the importance of white matter preservation for stroke rehabilitation [17]. Melatonin treatment improved sensorimotor performance in behavioral tests such as the corner and rotarod tests, further reinforcing its role in promoting functional recovery. The observed improvement in white matter integrity was associated with a decrease in oligodendrocyte pyroptosis, evidenced by the reduction in APC/caspase-1 and APC/GSDMD-N double-positive oligodendrocytes.

Oligodendrocytes are highly vulnerable to inflammatory injury, making their preservation vital for white matter repair [28]. Pyroptosis, a form of inflammatory cell death, plays a key role in demyelination and white matter injury, particularly in diseases such as multiple sclerosis (MS) and experimental autoimmune encephalomyelitis (EAE) [7, 29, 30]. Our results indicate that melatonin effectively suppresses pyroptosis by inhibiting caspase-1 and GSDMD activation. This finding complements earlier studies showing that inflammasome components, including NLRPs and ASC, contribute to post-stroke pyroptosis [10, 30]. In addition to reducing pyroptosis, melatonin shifted microglial polarization toward the anti-inflammatory phenotype. This polarization switch helps protect oligodendrocytes from inflammatory damage and promotes tissue repair. Microglia exhibit dynamic roles during stroke recovery: they initially adopt an anti-inflammatory state but may transition to a pro-inflammatory phenotype during the later stages of recovery, exacerbating tissue injury [31–33]. Thus, modulating microglial activity is crucial for maintaining white matter integrity and improving outcomes after stroke.

Our study highlights RORα-associated signaling as a key mediator of the neuroprotective effects of melatonin after ischemic stroke. Importantly, our data do not support a simple inhibitory effect of melatonin on RORα. Instead, melatonin dynamically modulated RORα-associated signaling in a stage-dependent manner. At the acute stage after MCAO, RORα expression and AMPKα phosphorylation were increased, and melatonin further enhanced these responses. At the chronic stage, MCAO reduced RORα expression and AMPKα phosphorylation, whereas melatonin partially restored both RORα expression and AMPKα activation. In both acute and chronic phases, melatonin consistently suppressed excessive STAT1 phosphorylation. These findings indicate that melatonin restores or enhances RORα/AMPKα signaling depending on the post-ischemic stage, while maintaining an inhibitory effect on aberrant STAT1 activation.

Mechanistically, our findings suggest that RORα functions upstream of AMPKα/STAT1 signaling in the protective response to melatonin. However, AMPKα and STAT1 phosphorylation are post-translational signaling events and are unlikely to represent direct transcriptional outputs of RORα itself. Rather, RORα may indirectly coordinate AMPKα activation and STAT1 suppression by regulating intermediate metabolic, inflammatory, or cellular stress-response pathways. Consistent with this interpretation, RORα knockdown weakened the protective effects of melatonin on microglial polarization and oligodendrocyte pyroptosis, temporal profiling showed coordinated changes in RORα expression, AMPKα phosphorylation, and STAT1 phosphorylation after MCAO, and AMPKα inhibition by Compound C partially reversed the effects of melatonin on microglial polarization. Thus, the RORα/AMPKα/STAT1 axis should be interpreted as a functionally connected signaling cascade rather than a simple direct transcriptional pathway.

To further explore whether melatonin may potentially associate with RORα, we performed molecular docking/interface analysis. The predicted model suggested a possible spatial association between melatonin and RORα, with melatonin located within a pocket-like region of the RORα structure. Interface analysis identified several residues surrounding melatonin, including Val364, Phe365, Ile366, Arg367, Met368, Cys369, Arg370, Ala371, and Phe372. However, the calculated solvation energy gain was ΔiG = 0.00 kcal/mol, indicating that this computational result should be interpreted cautiously. Therefore, our data do not prove direct physical binding between melatonin and RORα. Instead, they support the functional involvement of RORα-associated signaling in melatonin-mediated neuroprotection [34, 35].

Our findings align with previous studies demonstrating that modulating microglial activity can mitigate brain injury and promote recovery [36, 37]. The STAT and TLR/NF-κB pathways are well-known regulators of microglial polarization, with STAT3 playing a central role in promoting anti-inflammatory responses in macrophages and microglia [38–40]. Additionally, other studies have shown that activation of the STAT3 pathway by agents such as resveratrol can inhibit LPS-induced pro-inflammatory responses in microglia [41]. In line with our study, Ding et al. reported that melatonin regulates macrophage polarization via the RORα/AMPKα/STAT pathway, further supporting the pathway’s relevance in stroke recovery [23]. Therefore, our findings do not exclude the involvement of other melatonin-sensitive inflammatory pathways. Rather, they identify RORα/AMPKα/STAT1 signaling as one important mechanism by which melatonin modulates microglial phenotype and protects oligodendrocytes in ischemic white matter injury. Given the pleiotropic nature of melatonin, crosstalk between RORα/AMPKα/STAT1 signaling and other inflammatory pathways, such as STAT3 and TLR4/NF-κB, may jointly shape the inflammatory microenvironment during post-stroke recovery.

Given melatonin’s ability to target multiple interconnected pathways, it presents a promising candidate for stroke rehabilitation [42]. However, further in vivo and in vitro studies are needed to elucidate the full range of its effects on oligodendrocytes, microglia, and other cell types involved in stroke pathology [43, 44]. Identifying the precise molecular targets and pathways engaged by melatonin will be essential for translating these findings into clinical interventions that can enhance recovery and improve outcomes for stroke patients.

This study provides valuable insights, but several limitations should be acknowledged. First, although our loss-of-function experiments using siRNA and AAV-shRORα indicate that RORα is required for melatonin-mediated protection, RORα gain-of-function experiments were not performed. Therefore, whether RORα overexpression alone is sufficient to mimic melatonin-like effects on microglial polarization and oligodendrocyte pyroptosis remains to be determined. Second, although molecular docking/interface analysis suggested a potential spatial association between melatonin and RORα, this computational result does not prove direct physical binding. Future studies using SPR, MST, ITC, DSF, and RORα response element-based reporter assays are needed to determine whether melatonin directly binds to RORα and regulates RORα-dependent transcriptional activity. Third, although we added representative acute- and chronic-phase analyses at 3 and 21 days after MCAO, a complete time-course analysis including multiple early OGD time points and additional in vivo stages such as 1, 7, and 14 days after MCAO was not performed. Such studies would help define the immediate and sustained molecular events triggered by melatonin.

Fourth, although Compound C partially reversed the effects of melatonin on AMPKα phosphorylation and microglial polarization, additional pharmacological studies using AMPK activators and STAT1 inhibitors are needed to further clarify the hierarchical organization of the RORα/AMPKα/STAT1 pathway. Moreover, because Compound C may have off-target effects, these findings should be interpreted as supportive rather than definitive evidence for AMPKα dependency. Fifth, although the Transwell co-culture system allowed us to investigate microglia-oligodendrocyte crosstalk under controlled conditions, it cannot fully reproduce the complex architecture of in vivo white matter. This model lacks axon-myelin interactions, vascular components, astrocyte involvement, peripheral immune cell infiltration, extracellular matrix organization, and the full cellular heterogeneity of the post-stroke brain. Therefore, the in vitro findings should be interpreted as mechanistic support, while the in vivo MCAO data remain essential for validating the relevance of this pathway in ischemic white matter injury.

Finally, while our findings demonstrate the role of melatonin in regulating microglial polarization and inhibiting oligodendrocyte pyroptosis, the direct effects of melatonin on OPC differentiation and oligodendrocyte apoptosis remain unaddressed. In addition, only male mice were used in the present study, and future studies should investigate potential sex differences in melatonin-mediated white matter protection. Although the selected melatonin doses were based on previous experimental studies and preliminary cytotoxicity assessment, further pharmacokinetic and dose-response studies are needed to evaluate their translational relevance in clinical stroke recovery. Future clinical studies will also be required to confirm the therapeutic efficacy of melatonin in human stroke populations.

In conclusion, our study demonstrates that melatonin promotes functional recovery following ischemic stroke by modulating microglial polarization and inhibiting oligodendrocyte pyroptosis through RORα-associated AMPKα/STAT1 signaling. The neuroprotective effects of melatonin are highly dependent on RORα-associated signaling, as RORα knockdown weakened the anti-inflammatory/reparative-like shift in microglia and attenuated the inhibitory effect of melatonin on oligodendrocyte pyroptosis. These findings highlight the therapeutic potential of melatonin for preserving white matter integrity and promoting post-stroke recovery.

Materials and methods

Experimental animals

All animal experiments comply with ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines. Adult male C57BL/6J mice (8–12 weeks old, 25–30 g) were obtained from Charles River Laboratory (Beijing, China). The mice were housed in temperature-controlled cages (22 ± 2 °C) with ad libitum access to food and water. All animal procedures were conducted in accordance with institutional guidelines and approved by the Animal Ethical Committee of Beijing Rehabilitation Hospital, Capital Medical University (Approval ID: 2019bkky-08-012, August 2, 2019) and complied with Institutional Animal Care and Use Committee (IACUC) standards. The design, conduct, and reporting of animal experiments in this study comply with the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines 2.0 to ensure transparency and reproducibility in animal research. Efforts were made to minimize animal suffering throughout the study. Mice were randomly divided into three groups: sham, MCAO + vehicle, and MCAO + melatonin (n = 8 per group). Each mouse was used for multiple outcome evaluations, including neurobehavioral tests, white matter injury assessments, and protein expression analyses. Behavioral assessments were conducted by independent investigators blinded to the treatment groups to ensure unbiased results.

MCAO model and stereotaxic microinjection

A transient middle cerebral artery occlusion/reperfusion (MCAO/R) model was used to induce stroke, as described previously [45]. Mice were anesthetized with 1.5% isoflurane in a 30% O₂/70% N₂ mixture. A commercial monofilament (Doccol, 602156PK5Re, USA) was introduced through the right common carotid artery (CCA) to block the right middle cerebral artery (MCA) for 60 min. Sham-operated mice underwent the same anesthesia and surgery without MCA occlusion. Adeno-associated virus (AAV) vectors encoding shRNA targeting RORα (AAV-shRORα) and a negative control sequence were designed and synthesized by GeneChem (Shanghai, China). The nucleotide sequences are listed in Table S1. Eight-week-old C57BL/6J mice received a single stereotaxic microinjection of either AAV-shRORα or the control vector (1 µL of 1 × 10¹⁰ viral genomes/µL) into the ipsilateral cortex of ischemic lesions. The injection coordinates were 0.3 mm anterior to the bregma, 3 mm lateral to the midline, and 2 mm deep (AP: 0.3, ML: 3, depth: -2) as well as 1.9 mm posterior to the bregma, 3 mm lateral, and 2 mm deep (AP: -1.9, ML: 3, depth: -2). A 10 µL microsyringe was used to inject the vector at a rate of 0.1 µL/min. Following the injection, the needle was kept in place for 10 min before being slowly withdrawn to prevent reflux. Two weeks after AAV microinjection, the mice underwent the MCAO procedure to assess the effects of shRORα on pyroptosis. They were divided into four groups (n = 5 per group): MCAO-AAV (Ctrl) + Vehicle, MCAO-AAV (Ctrl) + Melatonin, MCAO-AAV-shRORα + Vehicle, MCAO-AAV-shRORα + Melatonin.

Melatonin preparation and treatments

Melatonin (Sigma-Aldrich) was dissolved in dimethylsulfoxide (DMSO, Sigma-Aldrich) and further diluted in saline to the desired concentration. The same concentration of DMSO in saline was used as the vehicle control. Mice received intravenous injections of either melatonin (20 mg/kg) or an equivalent volume of DMSO-saline at 0 and 24 h after reperfusion of the CCA. Daily injections were administered for 14 consecutive days [20].

Behavior test

Motor function was assessed using the rotarod test, as previously described [46]. In brief, mice were placed on a rotating drum (IITC Life Science) with speeds starting at 4rpm, accelerating to 40rpm over 300 s. Each mouse underwent three consecutive trials, with a 15-minute interval between trials. The time at which the mouse fell off the drum was recorded as the latency to fall. The corner test, as previously described [47], was also conducted to assess sensorimotor function. Two boards (30 × 20 × 1 cm each) were positioned at a 30° angle inside the mouse’s home cage, with a small opening at the joint to encourage entry into the corner. The mouse was placed between the boards facing the corner and halfway toward it. Upon entering the corner, the mouse’s vibrissae were stimulated, prompting the mouse to rear and turn back. The direction of each turn was recorded. Turns not associated with rearing were excluded. Healthy mice typically show no preference for left or right turns, while injured mice tend to turn preferentially toward the non-impaired, ipsilateral side. Each mouse underwent 10 trials per testing day, and the data were expressed as the percentage of right turns out of the total 10 trials.

Immunofluorescent staining and quantification

Oligodendrocyte pyroptosis and microglial polarization were detected using double immunofluorescence staining. Brain sections were fixed in 4% paraformaldehyde (PFA), washed in phosphate-buffered saline (PBS), and permeabilized with 0.25% Triton X-100 in PBS. After blocking with 10% normal goat serum (ZSGB-BIO, China), sections were incubated with a primary antibody mixture (Table S2) containing rabbit anti-Iba-1 antibody (1:500, ab178846, Abcam) and either mouse anti-GSDMD antibody (1:50, sc-393581, Santa Cruz) or mouse anti-caspase-1 antibody (1:50, sc-392736, Santa Cruz) for 36 h at 4 °C. This was followed by a 4-hour incubation at room temperature with a secondary antibody cocktail consisting of mouse anti-rabbit IgG-CFL 594 (1:100, sc-516250, Santa Cruz) and mouse IgGκ BP-CFL 488 (1:50, sc-516176, Santa Cruz). Samples were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (Southern Biotech, USA) and imaged using a laser confocal microscope (Nikon, Japan). Quantification of positive cells was performed using ImageJ software (NIH, USA), with analysis blinded to the experimental groups. The number of positive cells in the cortex (CTX), corpus callosum (CC), and striatum (STR) was calculated per square millimeter from three random microscopic fields (200× magnification). n = 5 or 8 animals per group. From each animal, one coronal brain section was collected, and 3 random fields of view (FOVs) per section were imaged.

Western blot

Collected cells were lysed in RIPA buffer containing the protease and phosphatase inhibitors (Roche). The extracted proteins were quantified with bicinchoninic acid (BCA) kit (Thermo Fisher Scientific). Proteins were separated by 12% SDS-PAGE gels and further transferred onto a PVDF membrane (Bio-Rad). The membranes were blocked with 5% fat-free milk in TBST for 1 h at room temperature, and incubated with the primary antibodies (Table S3), including signal transducer and activator of transcription 1 (STAT1, 1:1000, Abcam) and phospho- STAT1 (p-STAT1, 1:1000, Abcam) in TBST at 4 °C overnight. Next, the membrane was incubated with the secondary antibody (1:5000, Abcam) for 1 h at room temperature. All protein bands were visualized by ECL detection systems (Bio-Rad). GAPDH (Abcam) was used as an internal control. The intensities of protein bands were analyzed densitometrically using Image J software.

Primary microglia and oligodendrocyte culture

Newborn C57BL/6 mice were euthanized and promptly placed on ice. Under sterile conditions, brains were extracted, with the brainstem, hippocampus, and meninges removed. The brain tissue was rinsed thrice with pre-cooled D-Hanks solution and diced into ~ 1 mm³ cubes. These were digested with 1.25 g/L trypsin at 37 °C for 10 min, followed by neutralization with DMEM/F12 complete medium containing 10% fetal bovine serum (FBS). Gentle pipetting dispersed the cell clumps, which were then filtered through a 200-mesh screen. The filtrate was centrifuged at 1000 × g for 10 min at 4 °C, supernatant discarded, and cells resuspended in DMEM/F12 with 10% FBS to a density of 1 × 10⁶ cells/mL. Cells were seeded into poly-L-lysine-coated 75 cm² flasks for primary culture, left undisturbed for 48 h, and the medium was changed on the third day. Subsequently, cells were cultured under nutrient-deficient conditions for 7 days without further medium changes. For microglia and oligodendrocyte isolation, the primary culture was subjected to oscillatory separation [45, 48].

The medium was replaced with fresh DMEM/F12, and flasks were shaken at 200 g for 2 h at 37 °C. The cell suspension was centrifuged, supernatant discarded, and cells resuspended in complete medium before seeding into petri dishes. Adherent cells were collected after 4 h, and fresh medium was added for continued culture. The original flasks were washed, refilled with fresh medium, and cultured for another 7 days without medium changes. Microglia were re-isolated using the same method. After 7–9 days, mixed glial cells were observed under an inverted microscope, displaying confluence and stratified growth with microglia on the upper layer. Predominantly amoeboid and round cells indicated readiness for oscillatory separation, which was performed to remove microglia. Post-microglia removal, flasks were washed gently with PBS and refilled with fresh medium, then equilibrated in a CO₂ incubator for 2 h. Subsequently, flasks were shaken at 220 g for 18–22 h at 37 °C to isolate oligodendrocytes. The cell suspension was filtered through a 400-mesh sieve, placed in a 10 cm² dish for 15 min to allow differential adhesion, and then collected and centrifuged. Cells were resuspended in a chemically defined medium for oligodendrocyte precursors, adjusted to appropriate density, and seeded onto poly-L-lysine-coated plates. Cells were cultured in mouse oligodendrocyte complete medium with medium changes every other day for 3–5 days.

OGD cell model

To induce OGD in mouse primary microglia, the culture medium was discarded, and the cells were washed twice with PBS before being replaced with DMEM/F12 glucose-free culture medium. The cells were then placed in a carbon dioxide incubator set to 1% O₂ and 5% CO₂ and cultured under anoxic conditions at 37 °C for 4 h. After the OGD period, the glucose-free medium was discarded and replaced with complete culture medium to continue the culture. For primary oligodendrocytes, the culture medium was similarly discarded, and the cells were washed twice with PBS before being replaced with DMEM/F12 glucose-free medium. The cells were then incubated under the same anoxic conditions (1% O₂, 5% CO₂) at 37 °C for 40 min. After the OGD period, the glucose-free medium was discarded and replaced with complete medium to resume culture and treated with the vehicle or 400 nM melatonin (MLT) for 12 h [20].

siRORα transfection

Cells were seeded in 24-well plates one day prior to transfection and cultured in 0.5 mL of complete medium without antibiotics. Transfection was initiated when the cell density reached approximately 50%. For each transfection, 20 pmol of siRNA was mixed with 150 µL of OPTI-MEM medium, while 1 µL of Lipofectamine 3000 (Lipo3000) was added to another 150 µL of OPTI-MEM medium. Both mixtures were incubated separately at room temperature for 5 min, then combined and allowed to stand for an additional 15 min. Meanwhile, the culture medium in the 24-well plates was removed, and the cells were rinsed three times with 1 mL of PBS. Subsequently, 400 µL of OPTI-MEM medium was added to each well, and the plates were placed in a 5% CO₂ incubator at 37 °C. After 15 min, the medium was removed, and the siRNA-transfection reagent mixture was gently added to the corresponding wells in the 24-well plate. The plate was then gently shaken to ensure even distribution and incubated at 37 °C for 6 h. After the incubation period, the medium was replaced with serum-free medium to continue the culture. The nucleotide sequences for siRNA targeting RORα (siRNA-RORα) and the control siRNA (siRNA-Con) are listed in Table S4.

Transwell co-culture system of microglia and oligodendrocytes

Following 4 h of OGD treatment, microglia cells were returned to normal culture medium, and treated with the vehicle or MLT (400 nM). After 12 h of incubation, the cells were washed three times with fresh medium to remove residual reagents. For co-culture, primary oligodendrocytes were reoxygenated 40 min after OGD and replaced with complete medium at a 1:1 ratio of microglia to oligodendrocytes. The previously cleaned microglia (including the nested cells) were then layered onto the oligodendrocytes. After 12 h of continuous culture in the incubator, microglia were removed, and oligodendrocytes were collected for further analysis. The microglia were divided into eight groups for experimentation: Microglia-scramble (control); Microglia-scramble + melatonin (MLT, 400 nM); Microglia-scramble + OGD; Microglia-scramble + OGD + MLT (400 nM); Microglia-siRORα; Microglia-siRORα + MLT (400 nM); Microglia-siRORα + OGD; Microglia-siRORα + OGD + MLT (400 nM). For each condition, primary microglia (either scramble control or siRORα-treated) were initially subjected to 4 h of OGD and subsequently cultured in standard medium. Following a 12-hour incubation, cells were rinsed thrice with fresh medium. Two sets of primary oligodendrocytes were prepared: one as untreated control (Normal oligodendrocytes) and the other exposed to 40 min of OGD followed by reoxygenation and replacement with normal medium (Oligodendrocytes post-OGD). After washing, microglia were overlaid onto the oligodendrocytes. After an additional 12 h of coculture, microglia were removed, and oligodendrocytes were harvested for further analysis, with each group replicated three times (n = 3).

MTT test and LDH test

Cell survival was evaluated using the MTT assay (3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H-tetrazolium bromide, Thermo Fisher Scientific), the lactate dehydrogenase (LDH) assay (Point Scientific), and the Live/Dead staining kit (Invitrogen). All assays were performed according to the manufacturers’ protocols to ensure accuracy and consistency.

Bioinformatic analysis of melatonin-related transcriptomic data

To investigate the potential signaling pathways regulated by Melatonin, a transcriptomic dataset (GSE179519) was retrieved from the Gene Expression Omnibus (GEO) database. The raw expression matrix was downloaded and normalized using the limma package in R (version 4.3.2). Differentially expressed genes (DEGs) between melatonin-treated and control samples were identified using the criteria of |log₂ fold change (log₂FC)| ≥ 1 and p < 0.05. Volcano plots and heatmaps were generated using the ggplot2 and pheatmap packages, respectively. Subsequently, functional enrichment analyses were performed on the identified DEGs using the clusterProfiler package. Gene Ontology (GO) and KEGG pathway analyses were conducted to explore the biological processes and signaling pathways enriched among the DEGs. Enrichment results were considered significant at an adjusted p-value < 0.05.

Sex as a biological variable

For this study, only male adult C57BL/6J mice were used. This decision was based on the fact that the MCAO model is well-established in male mice, and previous studies have demonstrated significant sex differences in the response to ischemic stroke. Since this study focuses on stroke and neuroinflammation, which are known to have pronounced effects in male mice in this model, male mice were chosen to ensure consistency in experimental conditions. The findings from this study are expected to be relevant for both sexes, but further studies will be required to investigate potential sex differences in these outcomes.

Statistics

Statistical analyses were performed using GraphPad Prism (version 9.0, GraphPad Software, USA) and R software (version 4.3.2). Sample sizes were determined based on power analysis using prior experimental data (α = 0.05, β = 0.20). Data are expressed as mean ± standard deviation (SD). For comparisons between two groups, the Student’s two-tailed t-test was applied. For multiple group comparisons, one- or two-way analysis of variance (ANOVA) was used, followed by Bonferroni post hoc tests to assess differences between specific groups. Statistical significance was set at P < 0.05.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Zongjian Liu, Guohua Wang, Shuyan Qie and Jianing Xi conceived, designed, and supervised the entire study. Yuanyuan Ran, Haiyue Zhang, Linhong Mo, Chenye Qiao and Dajiang Wang performed the experiments, acquired data, and analyzed data. Zitong Ding, Haiqing Song, Fangmin Liu and Lisha Ye participated in part of the experiments, acquired data, and analyzed data. Yuanyuan Ran, Guohua Wang and Zongjian Liu wrote and edited the manuscript. All authors contributed to reading and approved the submitted version.

Funding

This work was supported by the Beijing Municipal Natural Science Foundation (Grants 7252076), National Natural Science Foundation of China (Grants 82272616 and 82171190), Major National Science and Technology Special Project of China (Grant 2023ZD0505304), Beijing Municipal Natural Science Foundation (Grants 7222102).

Data availability

The datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request. Supporting analytic code and relevant data are available in the supplementary materials. Further information regarding data access can be provided by the corresponding author.

Declarations

Study approval

All animal procedures were approved by the Animal Ethical Committee of Beijing Rehabilitation Hospital, Capital Medical University (Approval ID: 2019bkky-08-012, August 2, 2019). All experiments were carried out in accordance with institutional guidelines and complied with the Institutional Animal Care and Use Committee (IACUC) standards. The design, conduct, and reporting of animal experiments followed the ARRIVE guidelines to ensure transparency and reproducibility. For human studies, written informed consent was obtained from all participants prior to participation.

Competing interests

The authors declare that they have no competing interests.