Genes Involved by Dexamethasone in Prevention of Long-Term Memory Impairment Caused by Lipopolysaccharide-Induced Neuroinflammation

Abstract

Inflammatory activation within the brain is linked to a decrease in cognitive abilities; however, the molecular mechanisms implicated in the development of inflammatory-related cognitive dysfunction and its prevention are poorly understood. This study compared the responses of hippocampal transcriptomes 3 months after the striatal infusion of lipopolysaccharide (LPS; 30 µg), resulting in memory loss, or with dexamethasone (DEX; 5 mg/kg intraperitoneal) pretreatment, which abolished the long-term LPS-induced memory impairment. After LPS treatment, a significant elevation in the expression of immunity/inflammatory-linked genes, including chemokines (Cxcl13), cytokines (Il1b and Tnfsf13b), and major histocompatibility complex (MHC) class II members (Cd74, RT1-Ba, RT1-Bb, RT1-Da, and RT1-Db1) was observed. DEX pretreatment did not change the expression of these genes, but significantly affected the expression of genes encoding ion channels, primarily calcium and potassium channels, regulators of glutamate (Slc1a2, Grm5, Grin2a), and GABA (Gabrr2, Gabrb2) neurotransmission, which enriched in such GO biological processes as "Regulation of transmembrane transport", "Cognition", "Learning", "Neurogenesis", and "Nervous system development". Taken together, these data suggest that (1) pretreatment with DEX did not markedly affect LPS-induced prolonged inflammatory response; (2) DEX pretreatment can affect processes associated with glutamatergic signaling and nervous system development, possibly involved in the recovery of memory impairment induced by LPS.

Keywords: RNA-sequencing; dexamethasone; glutamate; hippocampus; inflammation; lipopolysaccharide; memory.

Figure 1

Pro-inflammatory and pro-apoptotic effects of LPS in the striatum. In the right striatum, which was the site of endotoxin infusion, LPS significantly (* p < 0.05) increased gene expression of pro-inflammatory cytokine interleukin-1beta (a) and protein of active caspase-3 (b) compared with SAL group at 24 h. In the right striatum, immunohistochemical analysis, using Iba-1 as a specific marker for microglia, did not reveal any differences between the LPS and SAL groups in the number of cells that were only Iba1 immuno-positive, whereas the number of cells immunoreactivities for both Iba-1 and caspase-3 was significantly increased (* p < 0.05) (c). (d) Iba-1 (green) and caspase-3 (red) immunofluorescence of representative images from the right striatum after SAL (up) and LPS (down).

Figure 2

Expression levels of interleukin-1beta mRNA (a) and protein of Iba-1 (b) in ipsilateral and contralateral hippocampi 24 h following administration of LPS into right striatum. * p < 0.05 compared with an appropriate SAL group.

Figure 3

Short-term and delayed behavioral effects of LPS administered alone or with DEX. (a) The Garcia score at 24 h. Neurological scale of the LPS-treated rats significantly decreased compared to that of the SAL-treated rats. When animals were pretreated with DEX, the neurological scale of the DEX + LPS group was higher than after LPS alone. (b) The recognition index was assessed 3 months after the drug administration. LPS caused memory impairment, and this effect was prevented by DEX pretreatment. Each group of animals that were the same on (a,b) consisted of 10 rats. * p < 0.05 compared with an appropriate control group (SAL or DEX + SAL); # p < 0.05 compared with LPS group.

Figure 4

Volcano plots show the differentially expressed genes (padj < 0.05 for all dots). Red represents upregulated genes, and blue represents downregulated genes with │log2 (fold change)│ ≥ 1.

Figure 5

(a) Common genes among DEGs in LPS vs. SAL and DEX + LPS vs. DEX + SAL (22 genes), as well as DEX + LPS vs. DEX + SAL and DEX + LPS vs. LPS (84 genes). (b) Expression profiles of DEGs for LPS vs. SAL. Data were expressed as log2 transformation of fold changes. Asterisks indicate 22 DEGs common for LPS vs. SAL and DEX + LPS vs. DEX + SAL.

Figure 6

Responses of representative genes to LPS alone and with DEX pretreatment (relative to SAL group): (a) MHCII-related genes (RT1-Db1, RT1-Da, RT1-Ba, and Cd74), Ciita; (b) Il1b and Tnfsf13b; (c) verification of the RNA-sec results for IL1b by quantitative PCR. * p < 0.05 compared with an appropriate control group (SAL or DEX + SAL).

Figure 7

Protein–protein interaction (PPI) networks: (a) two k-means clusters of 24 DEGs between LPS and SAL; (b) network of the 16 DEGs shared between LPS vs. SAL and DEX + LPS vs. DEX + SAL. Colors indicate the association of DEGs with several GO biological processes: red—GO:0006955 immune response, 14 DEGs, FDR = 1.74 × 10⁻¹¹; blue—GO:0006952 defense response, 11 DEGs, FDR = 1.63 × 10⁻⁷; green—GO:0006954 inflammatory response, 7 DEGs, FDR = 2.53 × 10⁻⁵; yellow—GO:0051716 cellular response to stimulus, 14 DEGs, FDR = 0.0024; violet—GO:0051384 response to glucocorticoid, 4 DEGs, FDR = 0.0048. FDR: false discovery rate.

Figure 8

Responses of representative genes to LPS or DEX alone, as well as to LPS with DEX pretreatment (relative to SAL group). (a) Data from RNA-sec; (b) verification of the RNA-sec results for Slc1a2 using quantitative PCR (correlation between sec and PCR expression values was 0.63, p < 0.05). * p < 0.05 compared with an appropriate control group (SAL or DEX + SAL).

Figure 9

Protein–protein interaction (PPI) network of the DEGs specific for DEX + LPS vs. DEX + SAL (k-means cluster 1, 21 DEGs). Colors indicate the association of DEGs with several GO biological processes: red—GO:0034762 regulation of transmembrane transport, 10 DEGs (Cacna1e, Cacna2d1, Erbb4, Grin2a, Grm5, Kcnh5, Kcnh7, Kcnq3, Slc1a2, and Wnk3), FDR = 1.11 × 10⁻⁶; blue—GO:0050890 cognition, 5 DEGs (Cacna1e, Chl1, Creb1, Grin2a, and Grm5), FDR = 0.0165; violet—GO:0007612 learning, 4 DEGs (Cacna1e, Creb1, Grin2a, and Grm5), FDR = 0.0198; green—GO:0022008 neurogenesis, 9 DEGs (Cdkl5, Chl1, Creb1, Erbb4, Gabrb2, Grin2a, Grm5, Kcnq3, and Unc5d), FDR = 0.0223; yellow—GO:0007399 nervous system development, 10 DEGs (Chl1, Cdkl5, Creb1, Erbb4, Gabrb2, Grin2a, Grm5, Kcnq3, Slc1a2, and Unc5d), FDR = 0.0340. FDR: false discovery rate.