Female Sprague-Dawley rats (200–220 g) were used for the induction of SCI. All animal experiments were approved by the Institutional Animal Care and Use Committee of Nanjing Medical University for the use of laboratory animals and performed in accordance with the Guide for the Care and Use of Laboratory Animals (National Research Council, 1996, USA). The rats were randomly assigned to control and experimental groups. Contusive SCI was performed as described previously (6,10,11). Briefly, the rats were anesthetized by pentobarbital sodium (50 mg/kg; intraperitoneal injection) and the spinal cords were exposed by laminectomy at the thoracic 9 (T9) level. Body temperature was maintained at 37°C during anesthesia. Following the exposure of the dorsal surface of the spinal cord, with the dura remaining intact, a weight-drop injury was performed by dropping a 10-g rod (diameter, 2.5 mm) from a height of 25 mm using the New York University impactor (purchased from New York University Neurosurgery Laboratory, New York, NY, USA). Sham-operated rats were laminectomized but received no contusion. The rats were perfused transcardially with 100 ml 0.9% saline followed by 200 ml 4% paraformaldehyde at day 1 or week 1 post-injury. Following perfusion, the entire spinal cord was removed. The spinal cord tissue (cervical to sacral) was separated into 5-mm segments. Frozen transverse sections (20-μm thick) were obtained using a cryostat (Leica CM1900; Leica Microsystems, Inc., Nussloch, Germany) and thaw-mounted on poly-L-lysine-coated slides (Sigma, St. Louis, MO, USA). The results are representative of sections from 3 male and 3 female rats.

The immunofluorescence double-labeling method was performed as described previously (12). Briefly, the sections were premobilized and blocked with 0.3% Triton X-100 or 3% normal goat serum in 0.01 M PBS for 30 min at room temperature (RT). A mixture of GS rabbit polyclonal (1:2,000; Abcam, Cambridge, UK) and mouse monoclonal antibodies were applied to the sections overnight at 4°C. The mouse monoclonal antibodies used in this experiment included anti-β-III-tubulin (1:800; Sigma) to identify neurons, anti-glial fibrillary acidic protein (GFAP; 1:800; Sigma) to identify astrocytes, anti-RIP (1:25, a gift from Dr Scott R. Whittemore, University of Louisville, KY, USA) to identify oligodendrocytes, anti-OX42 (CD11b; 1:200; Chemicon, Temecula, CA, USA) to recognize resting microglial cells and anti-CD68 (1:200; R&D systems, Minneapolis, MN, USA) to recognize activated microglia and macrophages. On the following day, the sections were incubated with FITC-conjugated goat anti-rabbit and TRITC-conjugated goat anti-mouse (1:100; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) antibodies. The sections were washed, mounted and examined using the Olympus BX60 light (Olympus, Center Valley, PA, USA) or Zeiss LSM 510 confocal microscopes (Carl Zeiss AG, Oberkochen, Germany). Primary antiserum omission and normal mouse and goat serum controls were used to confirm the specificity of the immunofluorescent labeling.

Immunohistochemistry was performed using the avidin-biotinylated peroxidase complex (ABC) immunoperoxidase-diaminobenzidine tetrahydrochloride (DAB) method according to the manufacturer’s instructions (Zymed; Invitrogen Life Technologies, Carlsbad, CA, USA). Briefly, the sections were incubated with mouse anti-GS antibodies (1:1,000; BD Biosciences, Franklin Lakes, NJ, USA) and then reacted with biotinylated goat anti-mouse IgG for 0.5 h at RT and subsequently with ABC for 0.5 h at RT. The reaction product was revealed by incubating the slices with DAB and H₂O₂ in substrate buffer for 15 min. Next, the sections were dehydrated, cleared and coverslipped. The slides were examined using an Olympus BX60 light microscope. The primary anti-serum omission and normal goat controls were used to confirm the specificity of the immunohistochemical labeling.

Western blot analysis was performed as described previously (6). Briefly, a 10-mm cord segment containing the injury epicenter was removed and homogenized by sonication in a lysis buffer. Protein (20 μg) from the sample supernatant was measured using the BCA method and loaded onto a 12% polyacrylamide gel, separated by SDS-PAGE and transferred to PVDF membranes by electrophoresis. Following membrane blocking, the mouse anti-GS (1:5,000; BD Biosciences) or anti-GFAP antibodies (1:1,000; Sigma-Aldrich) were added to the membrane and incubated at 4°C overnight. The membrane was washed and incubated with secondary antibody goat anti-mouse IgG conjugated with horseradish peroxidase (1:2,000; Amersham Pharmacia Biotech, Arlington Heights, IL, USA) at RT for 2 h. Subsequent to being washed, the membrane was analyzed by enhanced chemiluminescence (Amersham Pharmacia Biotech).

GS activity in the tissue was measured using the colorimetric assay as described previously (3) and then expressed as c-glutamylhydroxamate (μM)/h/mg cell protein. The spinal cord (10-mm section containing the injury epicenter) was obtained at varying post-operative survival times. The tissue protein concentrations were determined using the BCA protein estimation kit (Pierce Biotechnology, Inc., Rockland, IL, USA).

The glutamate content in the spinal cord samples was quantified by measuring the absorbance at 492 nm using the Glutamate Colorimetric Assay kit (Genmed Scientific, Inc., Arlington, MA, USA) and expressed as glutamate (μM)/mg tissue protein. A standard curve was constructed in each assay using known concentrations of glutamate. The concentration of tissue glutamate in the spinal cord (10-mm section containing the injury epicenter) was estimated from the standard curve.

All data are expressed as mean ± SD. A one-way ANOVA with Tukey’s HSD post hoc t-test was used to determine the level of statistical significance. P<0.05 was considered to indicate a statistically significant difference.

Immunofluorescence double staining was employed to confirm the cellular localization of GS in the rat spinal cords. Sections were obtained from an area 5 mm caudal to the epicenter of the sham-operated rat spinal cord. GS-expressing glial cells, including GFAP-positive astrocytes (Fig. 1A-C), RIP-positive oligodendrocytes (Fig. 1D-F) and OX42-positive microglia (Fig. 1G-I) were identified. However, GS expression was not observed in β-III-tubulin-positive neurons (Fig. 1J-L).

GS is the key glial-specificity enzyme that converts glutamate to glutamine to affect the clearance of synaptic glutamate (13). In the present study, the correlation between GS activity and glutamate concentration following SCI was analyzed. As shown in Fig. 2A, the concentration of glutamate increased rapidly and reached a maximum of 1.26±0.08 fold at 1 h compared with the sham-operated group (P=0.0048) post-injury. The concentration returned to baseline rapidly at 4 h post-injury. Fig. 2B demonstrates that a significant decrease in GS activity occurred at 4 h post-injury (P=0.018), then the GS activity increased, prior to returning to the basal level at day 1 and then increasing to peak levels at day 7 post-injury (P=0.0037). The GS activity then remained at a high level. SCI also caused a significant decrease in the GS expression at 4 h (P=0.018) and day 1 (P=0.0079) post-injury. The GS protein levels also revealed a similar time course change following SCI (Fig. 2C and E). To test whether the change in GS levels following SCI was associated with the activity of glial cells, the correlation between the changes in GS and GFAP, CD68 and PLP, the specific markers for astrocytes, active microglia/macrophages and oligodendrocytes, respectively, was determined (Fig. 2C-H). The western blot analysis revealed that GFAP levels decreased 1 h post-injury, decreased further at 4 h (p=0.018) and day 1 (P=0.0022) post-injury, then increased rapidly from day 3 (P=0.0099) and peaked at day 7 post-injury (P=0.000016). The GFAP protein concentration remained at a high level from then onwards (Fig. 2C and D). Fig. 2E and G revealed that CD68 was almost undetectable in the sham-operated spinal cords and increased significantly from 4 h following SCI (P=0.0011; Fig. 2F and G). However, the PLP expression decreased gradually and was significantly decreased 1 day after SCI (P=0.0033; Fig. 2F and H).

To determine the spatial distribution of GS following SCI, immunochemistry staining for GS was performed at days 1 and 7 post-injury, when the GS protein was minimally and maximally expressed, respectively. In the sham-operated rats (Fig. 3, left column), the GS immunoreactivities revealed comparable basal levels in the glial cells in the white and gray matter of the spinal cord. At day 1 post-injury, the GS immunoreactivities were markedly decreased in the injured epicenter dorsal column and ventral horn gray and white matter (Fig. 3B and E). However, no significant difference was observed in the GS levels in the region 5-mm caudal to the injury site (Fig. 3H and K). By contrast, the GS immunoreactivities were markedly increased in the glial cells in the white and gray matter 7 days after SCI (Fig. 3C and F). The strongest labeling was observed at the lesion cavity and surrounding area. Specific GS immunoreactivities were morphologically characteristic of activated astrocytes. It appeared that certain cells with spherical morphologies were derived from cells that infiltrated into the lesion site. Furthermore, the increased GS immunoreactivities were also observed throughout the entire length of the specimen (10 mm; Fig. 3I and L).

The astrocyte response to injury was confined to the immediate penumbra surrounding the lesion center. Astrocyte reactivity was characterized by a cellular hypertrophy with a process extension and increased production of intermediate filaments, including GFAP (Fig. 4A). The upregulation of GS in the tissue surrounding the lesion center was produced by GFAP-positive astrocytes (Fig. 4A-D). Since macrophages and microglial cells express GS (14,15), CD68 was used to identify the GS-positive infiltrative cells. As revealed in Fig. 4E-H, colocalization of GS and CD68 was observed in certain spherical cells (indicated by arrows). Thus, these results indicate that the infiltrative macrophages/activated microglia in areas close to the injury site express GS.