Elevated microRNA-129-5p level ameliorates neuroinflammation and blood-spinal cord barrier damage after ischemia-reperfusion by inhibiting HMGB1 and the TLR3-cytokine pathway

Abstract

Background: Ischemia-reperfusion (IR) affects microRNA (miR) expression and causes substantial inflammation. Multiple roles of the tumor suppressor miR-129-5p in cerebral IR have recently been reported, but its functions in the spinal cord are unclear. Here, we investigated the role of miR-129-5p after spinal cord IR, particularly in regulating high-mobility group box-1 (HMGB1) and the Toll-like receptor (TLR)-3 pathway.

Methods: Ischemia was induced via 5-min occlusion of the aortic arch. The relationship between miR-129-5p and HMGB1 was elucidated via RT-PCR, western blotting, and luciferase assays. The cellular distribution of HMGB1 was determined via double immunofluorescence. The effect of miR-129-5p on the expression of HMGB1, TLR3, and downstream cytokines was evaluated using synthetic miRs, rHMGB1, and the TLR3 agonist Poly(I:C). Blood-spinal cord barrier (BSCB) permeability was examined by measuring Evans blue (EB) dye extravasation and the water content.

Results: The temporal miR-129-5p and HMGB1 expression profiles and luciferase assay results indicated that miR-129-5p targeted HMGB1. Compared with the Sham group, the IR group had higher HMGB1 immunoreactivity, which was primarily distributed in neurons and microglia. Intrathecal injection of the miR-129-5p mimic significantly decreased the HMGB1, TLR3, interleukin (IL)-1β and tumor necrosis factor (TNF)-α levels and the double-labeled cell count 48 h post-surgery, whereas rHMGB1 and Poly(I:C) reversed these effects. Injection of miR-129-5p mimic preserved motor function and prevented BSCB leakage based on increased Basso Mouse Scale scores and decreased EB extravasation and water content, whereas injection rHMGB1 and Poly(I:C) aggravated these injuries.

Conclusions: Increasing miR-129-5p levels protect against IR by ameliorating inflammation-induced neuronal and BCSB damage by inhibiting HMGB1 and TLR3-associated cytokines.

Keywords: Blood-spinal cord barrier; High-mobility group box-1; Ischemia-reperfusion injury; MicroRNAs; Toll-like receptor 3.

Fig. 1

Experimental groups and protocol. Schematic representation of the six groups of mice exposed to the different treatments

Fig. 2

Time course of IR-induced alterations in the miR-129-5p and HMGB1 levels in a mouse model of spinal cord IR injury. a Quantification of miR-129-5p expression. b Representative western blots to detect HMGB1 expression. c Quantification of the integrated intensity of the HMGB1 bands. Data are expressed as the mean ± SD. n = 6 per group. *P < 0.05 versus the Sham group

Fig. 3

Confirmation of a direct interaction between miR-129-5p and HMGB1. a Schematic representation of the predicted interaction between miR-129-5p and HMGB1 based on TargetScan. b Binding site of miR-129-5p in the WT and MT 3′UTR of HMGB1. c Dual-luciferase reporter gene assay. Luciferase activity was measured after 48 h of co-transfection of HEK-293 cells with the miR-129-5p mimic and the WT or MT reporter vector. Compared to the control, miR-129-5p significantly decreased the luciferase activity of the WT vector but not that of the MT vector. Data are expressed as the mean ± SD. *P < 0.05 versus the control miR

Fig. 4

Double immunofluorescence of HMGB1 with markers of major cell types in the spinal cord after IR. a Representative fluorescence images of the distribution of HMGB1 (red) in neurons (NeuN; green), microglia (Iba1; green) and astrocytes (GFAP; green) in the spinal cords at 12 and 48 h after IR. Arrows indicate co-localization. Scale bars = 50 μm. b Quantification of HMGB1 signals was performed based on the average of three independent images. c Quantification of HMGB1-positive neurons and microglia in the spinal cords at 12 h and 48 h after IR. Data are expressed as the mean ± SD. *P < 0.05 versus the Sham group

Fig. 5

Effect of the miR-129-5p mimic and mimic control on the expression of HMGB1 and TLR3-mediated cytokines at 48 h after IR. a Representative fluorescence images of the distribution of TLR3(red), HMGB1(green) and nuclei (blue) in the spinal cords at 48 h after IR. b Representative western blots of HMGB1 and TLR3 expression. GAPDH was used as a loading control. c Quantification of the relative protein expression of HMGB1 and TLR3 at 48 h after IR. d Quantification of the relative mRNA expression of HMGB1 and TLR3 at 48 h after IR. e Quantification of spinal IL-1β and TNF-α production at 48 h after IR. All data were obtained from three independent experiments and expressed as the mean ± SD. *P < 0.05 versus the Sham group. ^# P < 0.05 versus the IR group

Fig. 6

Effects of the miR-129-5p mimic and mimic control on HMGB1 expression in specific cell types of the spinal cord after IR. a Representative photomicrographs showing the localization of the fluorescence signals for HMGB1 in neurons and microglia at 48 h after IR. Arrows indicate co-localization. Scale bars = 50 μm. b Quantification of HMGB1 signals was performed based on the average of three independent images. c Quantification of HMGB1-positive neurons and microglia in the spinal cords at 48 h after IR. Data are expressed as the mean ± SD. *P < 0.05 versus the Sham group. ^# P < 0.05 versus the IR group

Fig. 7

Effects of the miR-129-5p mimic and mimic control on hind-limb motor function and blood-spinal cord barrier (BSCB) integrity after IR. a Hind-limb motor function was assessed at 12-h intervals for 48 h after reperfusion using the Basso Mouse Scale for locomotion with a range from 0 (paraplegia) to 9 (normal). b Assessment of each mouse at 48 h post-surgery in the six groups. Each symbol represents one mouse. The bar represents the median. c Changes in BSCB permeability were measured based on EB extravasation visualized as red fluorescence by microscopy. d Quantification of the EB content of the spinal cord (μg/g tissue). e Quantification of the EB fluorescence intensity (INT/mm²). f Quantification of the spinal water content (%). Data are expressed as the mean ± SD. *P < 0.05 versus the Sham group. ^# P < 0.05 versus the IR group