Molecular hydrogen suppresses reactive astrogliosis related to oxidative injury during spinal cord injury in rats

Abstract

Aims: Spinal cord injury (SCI) can induce excessive astrocyte activation. Hydrogen has been deemed as a novel antioxidant. We investigated whether molecular hydrogen could act as an antiastrogliosis agent during SCI and oxidative injury in experimental rats and cultured astrocytes.

Methods: Hydrogen-rich saline (HS, 8 mL/kg, i.p.) was injected every 12 h after SCI in rats. The expression of STAT3, p-STAT3, and glial fibrillary acidic protein (GFAP); the release of IL-1β, IL-6, and TNF-α; and astrogliosis, along with the BBB score, were evaluated. Culturing astrocytes with hydrogen-rich medium, the intracellular reactive oxygen species (ROS), astrogliosis, and the release of proinflammatory cytokines were assessed after H2O2-induced injury.

Results: In the HS group, the expression of STAT3, p-STAT3, and GFAP and the proinflammatory cytokines were decreased in local spinal cord on postoperation day (POD) 3; on PODs 7 and 14, reactive astrogliosis was suppressed, and the locomotor function was also improved. Furthermore, hydrogen-rich medium attenuated the intracellular production of ROS (especially HO•), astrogliosis, and the secretion of proinflammatory cytokines in astrocytes 12 h after H2O2-induced injury.

Conclusions: Molecular hydrogen could suppress reactive astrogliosis after contusive SCI and reduce the release of proinflammatory cytokines produced by active astrocytes related to oxidative injury. Thus, molecular hydrogen is potential to be a neuroprotective agent.

Keywords: Astrogliosis; Glial scar; Hydrogen; Oxidative injury; Spinal cord injury.

Conflict of interest statement

We have not published or submitted the manuscript elsewhere simultaneously. The authors taking part in this study declared that they do not have any conflict of interests in this manuscript.

© 2014 John Wiley & Sons Ltd.

Figures

Figure 1

On postoperation day (POD) 3, local contusive spinal cord segments were dissected for assay. The content of IL‐1β, IL‐6, and TNF‐α was measured by ELISA, and the production of STAT3, p‐STAT3, and glial fibrillary acidic protein (GFAP) expressed by astrocytes was assayed by Western blot. (A) Content of IL‐1β, IL‐6, and TNF‐α (pg/mL) in each groups, respectively. (B) The expression of STAT3, p‐STAT3, and GFAP. (C) Western blot data were expressed as gray values normalized to β‐actin. The concentration of these proinflammatory cytokines was remarkably increased in the Allen + normal saline (NS) group and was notably attenuated by hydrogen‐rich saline administration. The expression of these three proteins was increased in the NS group (vs. sham, P < 0.01) and was significantly reduced by hydrogen‐rich saline administrated intraperitoneally on POD 3 (vs. NS, P < 0.01, P < 0.05, and P < 0.01, respectively). Results were means ± SEM of three independent experiments. **P < 0.01, Allen + NS compared with sham; #P < 0.05 and ##P < 0.01, Allen + hydrogen‐rich saline compared with Allen + NS.

Figure 2

(A, B) Immunohistochemistry of glial fibrillary acidic protein (GFAP) on postoperation day (POD) 7, and data were expressed with fluorescence intensity. (C) and (D) Immunohistochemistry of GFAP on POD 14, and its fluorescence intensity data. Morphologically, the characteristics of strongly activated astrocytes induced by contusive spinal cord injury were hypertrophy and hyperplasia, and the expression of GFAP (measured by the fluorescence intensity) was dramatically augmented on POD 7 and POD 14 (normal saline (NS) vs. sham, P < 0.01). Hydrogen‐rich saline administration significantly suppressed the reactive astrogliosis and alleviated the excessive expression of GFAP (vs. NS, P < 0.05 and P < 0.01, respectively). Results were means ± SEM of three independent experiments. **P < 0.01, Allen + NS vs. sham; #P < 0.05 and ##P < 0.01, Allen + hydrogen‐rich saline vs. Allen + NS. Scale bar = 100 μm.

Figure 3

Locomotor function was evaluated before surgery (D 0) and on postoperation days (PODs) 1, 7, and 14 using the Basso, Beattie, and Bresnahan locomotor scale (BBB scale). Spinal cord injury severely affected the locomotor function of animals (normal saline (NS) vs. sham, P < 0.01). Rats in the hydrogen‐rich saline (HS) group exhibited higher BBB scores on PODs 7 and 14 (vs. NS, P < 0.01) and revealed a treatment effect of HS over time (P < 0.01). Results were means ± SEM of three independent experiments. **P < 0.01 and ##P < 0.01, Allen + HS vs. Allen + NS.

Figure 4

Twelve hours after adding H2O2 (100 μM) and FeCl2 (15 μM) to the medium, with or without super‐saturated hydrogen, intracellular reactive oxygen species (ROS) levels of each cultured astrocyte group were assayed. (A) and (B) Intracellular ROS levels were measured by the intensity of DCF. (C) and (D) Intracellular HO• levels were measured by the intensity of hydroxyphenyl fluorescein. There was no difference between control and single hydrogen‐rich medium groups, but the total amounts of ROS and HO• were distinctly increased by H2O2 (vs. control, P < 0.01) and were notably decreased by hydrogen‐rich medium (HM) treatment (vs. normal medium, NM + H2O2, P < 0.01). Results were means ± SEM of three independent experiments. **P < 0.01, NM + H2O2 vs. control; ##P < 0.01, H2O2 + HM vs. H2O2 + NM. Scale bar = 50 μm.

Figure 5

Twelve hours after adding H2O2 (100 μM) and FeCl2 (15 μM) to the medium, with or without super‐saturated hydrogen, the hypertrophy and proliferation of cultured astrocytes were observed. (A) and (B) Immunocytochemistry of glial fibrillary acidic protein (GFAP) demonstrating the morphological characteristics of hypertrophy and excessive expression of GFAP, and the fluorescence intensity data. (C) and (D) Images demonstrating the DNA synthesis of BrdU, which reflects the proliferation of astrocytes, and the percentage of positive cells. Hypertrophy, proliferation, and excessive expression of GFAP were obviously induced by H2O2 in the normal medium (NM) + H2O2 group (vs. control, P < 0.01) and were significantly inhibited by treatment with hydrogen‐rich medium (vs. NM + H2O2, P < 0.01). Results were means ± SEM of three independent experiments. **P < 0.01, NM + H2O2 vs. control; ##P < 0.01, H2O2 + hydrogen‐rich saline vs. H2O2 + normal saline. Scale bar for A = 20 μm and for C = 50 μm.

Figure 6

Twelve hours after adding H2O2 (100 μM) and FeCl2 (15 μM) to the medium, with or without super‐saturated hydrogen, the supernatant of each group was collected for ELISA to assay the secretion of proinflammatory cytokines by active, cultured astrocytes. (A) Content of IL‐1β (pg/mL). (B) Content of IL‐6 (pg/mL). (C) Content of TNF‐α (pg/mL). H2O2 led to the abnormally increased secretion of IL‐1β, IL‐6, and TNF‐α (vs. control or single hydrogen‐rich medium (HM), P < 0.01). Secretion of these cytokines could be reduced by hydrogen‐rich medium (vs. NM + H2O2, P < 0.01). Results were means ± SEM of three independent experiments. **P < 0.01, NM + H2O2 vs. control; ##P < 0.01, H2O2 + hydrogen‐rich saline vs. H2O2 + normal saline.