Objective To explore the activation of astrocytes induced by the perivascular inflammatory microenvironment during the pathogenesis of the brain during the pathogenesis process of cerebral malaria and its mechanism, as well as its impact on neuronal damage. Methods C57BL/6 male mice aged 4-5 weeks were intraperitoneally inoculated with 5 × 10⁶ red blood cells infected with the ANKA strain of Plasmodium berghei, which would be died 7-10 d later and served as an experimental cerebral malaria model. Newborn suckling mice aged 3-5 d were euthanized to isolate primary astrocytes from the cerebral cortex; while the other suckling mince were euthanized within 24 hours for isolation of primary cortical neurons. The cell culture containing 20 ng/ml γ interferon (IFN-γ), 1 ng/ml tumor necrosis factor α (TNF-α) and mice red blood cells (pRBC) infected with malaria parasites were used to induce activation of astrocytes, and assigned as the activation group, while untreated resting-state cells were set as the control group. After cultivation for 24 h, total RNA was extracted by cell lysis and sequenced. Cluster analysis was performed to generate heatmaps, and representative genes were selected to draw violin diagram. ELISA was used to determine the expression level of CXCL10 in cell culture supernatant. The spleen CD8⁺ T cells of mice with cerebral malaria were separated and co-cultured with activated astrocytes to observe immune synapses and CXCL10 levels after co-incubation with CXCL10 antibodies by fluorescence microscopy, and CD80 and other molecular expression levels were detected using flow cytometry. Primary neurons were added into two sets of astrocyte culture supernatants and incubated in MAP2 antibody. Blank culture medium was set as the blank group, and lactate dehydrogenase (LDH) and CCK-8 assay kit were used to detect the degree of neuronal damage. Western blotting was used to detect the levels of STAT1, STAT3, and their phosphorylated molecules. The activation group was treated with STAT1 pathway inhibitor fludarabine, and the morphology of neurons was observed by immunofluorescence staining. Statistical analysis was conducted using PRISM Graph Pad 8.0 software, and pairwise comparisons were conducted using t-tests. Results The morphology of activated astrocytes changed from flat star-shaped to long spindle-shaped, and the relative content of glial fibrillary acidic protein (GFAP) was (1.36 ± 0.03), higher than that of the control group (1.00 ± 0.00) (t = 13.33, P < 0.01). The CXCL10 content in the supernatant of the activation group cell culture was (7.07 ± 0.81) ng/ml, which was higher than that of the control group (2.57 ± 0.28) ng/ml (t = 9.05, P < 0.01). Transcriptome analysis showed that antigen processing, presentation, and transcription levels of T cell chemokines were upregulated in the activation group (all P < 0.01), and the expression levels of CD80, CD86, and MHC Ⅰ were upregulated. The activation group co-incubated with CD8⁺ T cells to form "immune synapses", and the CD69 expression level of CD8⁺ T cells increased. The LDH results showed that the relative mortality rate of neurons in the activation group after supernatant stimulation was (50.2 ± 2.4)%, which was higher than that in the blank group (0% ) and the control group (0%) (t = 20.62, 20.62, both P < 0.01). The CCK-8 detection results showed that the relative activity of neurons in the activation group after supernatant stimulation was 0.52 ± 0.03, lower than that in the blank group (1.00 ± 0.00) and the control group (1.42 ± 0.06) (t = 18.92, 16.65, both P < 0.01). Western blotting results showed that relative expression level of neurons stimulated with activated astrocytes supernatant β-amyloid precursor protein (β-APP) was 0.44 ± 0.02, which was lower than that of the blank group (1.00 ± 0.00) and the control group (0.55 ± 0.02) (t = 37.28, 4.93, both P < 0.01). The ratio of pro apoptotic protein Bax to anti apoptotic protein Bcl-2 was 1.01 ± 0.07, and there was no statistically significant difference compared with the blank group (1.00 ± 0.00) and the control group (1.00 ± 0.06) (t = 0.31, 0.13, both P > 0.05). Western blotting results showed that the relative expression level of STAT1 and p-STAT1 protein in the cytoplasm of activation group were 3.40 ± 1.08 and 4.00 ± 0.82, both of which were higher than the control group (1) (t = 3.13, 5.13, both P < 0.05). The relative expression level of STAT3 and p-STAT3 protein in the cytoplasm of activation group were 1.00 ± 0.03 and 1.01 ± 0.05. There was no statistically significant difference compared to the control group (1) (t = 0.27, 0.52, both P > 0.05). The relative expression level of p-STAT1 protein in the nucleus of the activation group was 1.78 ± 0.21, higher than that of the control group (1) (t = 5.081, P < 0.01), and the relative expression level of p-STAT3 protein was 1.02 ± 0.02, with no statistically significant difference compared to the control group (1) (t = 1.38, all P > 0.05). The results of MAP2 immunofluorescence staining showed that the activation group supernatant could cause significant neuronal damage, which was alleviated after treatment with STAT1 inhibitors. Conclusion During the pathogenic process of cerebral malaria, the brain perivascular inflammatory microenvironment may induce production of neurotoxic astrocytes via the STAT1, and may cause death of neurons. This activated astrocytes could interact with activated CD8⁺ T cells, and further aggravate the damage of central nervous system.