Preventive effects of intrathecal methylprednisolone administration on ...

Neuroscience 147 (2007) 294 –303

PREVENTIVE EFFECTS OF INTRATHECAL METHYLPREDNISOLONE ADMINISTRATION ON SPINAL CORD ISCHEMIA IN RATS: THE ROLE OF EXCITATORY AMINO ACID METABOLIZING SYSTEMS G.-J. WU,a,b,c1 W.-F. CHEN,d1 C.-S. SUNG,e Y.-H. JEAN,f C.-M. SHIH,g C.-Y. SHYUh AND Z.-H. WENi*

ventral portion of the lumbar spinal cord was partly inhibited by pretreatment with i.t. MP. However, MP did not affect the down-regulation of EAAC1 in the dorsal portion of the lumbar spinal cord after spinal cord ischemia. The i.t. injection of MP alone did not change the neurological functions and the expression of proteins of the glutamate metabolizing system in the spinal cord. Our results indicate that spinal cord ischemia-induced neurological deficits accompany the decrease in the expression of proteins of the glutamate metabolizing system in the lumbar portion of the spinal cord. The i.t. MP pretreatment significantly prevented these symptoms. These results support the observation that MP delivery through an i.t. injection, is beneficial for the treatment of spinal cord ischemic injuries. © 2007 IBRO. Published by Elsevier Ltd. All rights reserved.

a

Department of Anesthesiology, Shin Kong Wu Ho-Su Memorial Hospital, Taipei, Taiwan

b Department of Medicine, School of Medicine, Fu-Jen Catholic University, Taipei, Taiwan c

Graduate Institute of Medical Sciences, College of Medicine, Taipei Medical University, Taipei, Taiwan

d Department of Neurosurgery, Chang Gung Memorial Hospital, Chang Gung University College of Medicine, Kaohsiung, Taiwan e Department of Anesthesiology, Veterans General Hospital, Taipei, Taiwan f

Section of Orthopedic Surgery, Pingtung Christian Hospital, #60 Da-Lan Road, Pingtung 900, Taiwan

g Department of Biochemistry, School of Medicine, Taipei Medical University, Taipei, Taiwan

Key words: spinal cord ischemia, glutamate metabolizing systems, methylprednisolone, intrathecal.

h Cook College, Rutgers, The State University of New Jersey, New Brunswick, NJ, USA

Spinal cord ischemia, which often occurs after the repair of thoracoabdominal aortic aneurysms or dissection, is an uncommon but devastating entity in clinical practice. The subsequent paraplegia that occurs can be a serious complication following aortic surgery (Kouchoukos, 1991; Svensson et al., 1993). Researchers have yet to find an effective cure to prevent it. The results of studies on the mechanisms and prevention of spinal cord ischemia are still unclear and require further investigation. The excitatory amino acid (EAA) glutamate, which is a major excitatory neurotransmitter in the CNS, also plays an important role in the pathogenesis of neuronal injury. Pharmacological studies in rodents and recent clinical studies in humans have shown that excessively high extracellular concentrations of glutamate caused by ischemia can be toxic to neurons (Nishizawa, 2001). Marsala et al. (1994) and Rokkas et al. (1995) had demonstrated a strong positive relationship between the increase in glutamate concentrations and aortic cross-clamping induced by spinal cord ischemia. The presence of excess extracellular glutamate activates neuronal glutamate receptors; these then induce massive lethal Ca2⫹ influx and subsequently result in neurotoxicity (Arundine and Tymianski, 2003). To ensure a high signal-to-noise ratio during synaptic signaling and to protect the neurons, the extracellular concentration of glutamate in the synapse needs to be maintained at an appropriate level (⬍1 ␮M) (Danbolt, 2001). There is no evidence to indicate that extracellular metabolism of glutamate occurs since the concentration is basically maintained by plasma membrane transporters. Glutamate transporters (GTs) are membrane proteins that reuptake

i

Department of Marine Biotechnology and Resources, National Sun Yat-Sen University, #70 Lien-Hai Rd, Kaohsiung 804, Taiwan

Abstract—Spinal cord ischemic injury usually results in paraplegia, which is a major cause of morbidity after thoracic aorta operations. Ample evidence indicates that massive release of excitatory amino acids (EAAs; glutamate) plays an important role in the development of neuronal ischemic injuries. However, there is a lack of direct evidence to indicate the involvement of EAAs in the glutamate metabolizing system (including the glutamate transporter isoforms, i.e. the GluAsp transporter (GLAST), Glu transporter-1 (GLT-1), and excitatory amino acid carrier one (EAAC1); glutamine synthetase (GS); and glutamate dehydrogenase (GDH)) in spinal cord ischemia. In the present results, we found that methylprednisolone (MP; intrathecal (i.t.) injection, 200 ␮g twice daily administered for 3 days before ischemia), a synthetic glucocorticoid, is the therapeutic agent for the treatment of spinal injuries in humans, can significantly reduce the ischemia-induced motor function defect and down-regulate the glutamate metabolizing system (including GLAST, GLT-1, GS, and GDH) in male Wistar rats. The spinal cord ischemiainduced down-regulation of EAAC1 protein expression in the 1

These authors contributed equally to this work. *Corresponding author. Tel: ⫹886-7-5252021; fax: ⫹886-7-5255020. E-mail address: [email protected] (Z.-H. Wen). Abbreviations: ANOVA, analysis of variance; CSF, cerebrospinal fluid; EAA, excitatory amino acid; EAAC1, excitatory amino acid carrier 1; EAAT, excitatory amino acid transporter; GDH, glutamate dehydrogenase; GFAP, glial fibrillary acid protein; GLAST, Glu-Asp transporter; GLT-1, Glu transporter-1; GS, glutamine synthetase; GT, glutamate transporter; H&E, hematoxylin and eosin; i.t., intrathecal; MDI, motor deficit index; MP, methylprednisolone; SOD, superoxide dismutase; TTBS, 5% non-fat dry milk in 0.1% Tween 20 in 20 mM Tris–HCl, 137 mM NaCl, pH 7.4..

0306-4522/07$30.00⫹0.00 © 2007 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2007.04.040

294

G.-J. Wu et al. / Neuroscience 147 (2007) 294 –303

glutamate into the nerve terminals or glial cells in order to remove it from the extracellular space around the neurons. Five related but distinct eukaryotic high-affinity GTs have been cloned: the neuronal transporter excitatory amino acid carrier 1 (EAAC1); excitatory amino acid transporter 4 (EAAT4); the glial transporter, i.e. Glu-Asp transporter (GLAST); Glu transporter-1 (GLT-1); and a recently identified retinal transporter EAAT5 (Danbolt, 2001). Homologs of GLAST (EAAT1), GLT-1 (EAAT2), and EAAC1 (EAAT3) have been identified in the mammalian CNS (Rothstein et al., 1994; Lehre et al., 1995). Glutamate removal is essential for maintaining functional communication between neurons and preventing the concentration of glutamate from reaching toxic levels. The purpose of the present study is to evaluate how spinal cord ischemia induces changes in the expression of GT subtypes in various rat spinal cord regions. Glutamate is taken up by the glial cells and subsequently metabolized by glutamine synthetase (GS) and glutamate dehydrogenase (GDH) into neutral metabolites, such as glutamine, thus preventing the overexcitation of neurons and excitotoxicity. Martin and Waniewski (1996) reported that GS is important for converting glutamate into nontoxic glutamine in astrocytes; this glutamine is then used in the neuronal TCA cycle. GDH converts glutamate into 2-oxoglutarate and other related metabolites (malate, pyruvate, and lactate), which provide energy to the neuron (Nicklas, 1984; Kaneko et al., 1988). Thus, GS and GDH might play an important role in modulating extracellular glutamate levels. In this study, we also examine the expression of GS and GDH in the spinal cord during spinal cord ischemia. Methylprednisolone (MP), a synthetic glucocorticoid, has long been approved by the FDA, on the U.S. and worldwide markets and available for a variety of uses including acute SCI. The most significant characteristic of intrathecal (i.t.) injections is that the dosage of medication required and the side-effects caused by it are reduced. Therefore, i.t. MP administration is an effective mode of treatment for postherpetic neuralgia (Kotani et al., 2000). However, its effect is very limited since a technique for direct glucocorticoid delivery to the injury site in spinal cord ischemia is yet to be developed. The mechanisms responsible for the effect of MP on neuronal injuries remain to be investigated. Several studies have shown that the vascular anatomy of rats and humans is similar (Koyanagi et al., 1993; Scremin, 1995). These studies also provided a suitable animal model for studying spinal cord ischemia-induced neurobehavioral deficits and evaluating the neuroprotective effect of MP. Therefore, in this study, we have elucidated the role of glutamate metabolizing systems in the development of spinal ischemia. Moreover, we have investigated the effect of i.t. MP on neurobehavioral deficits as well as on the EAA metabolizing system in the spinal cord of ischemic rats. During the study, we have noted a correlation between the MP-induced protective effect and the changes in the expression of proteins of the glutamate metabolizing system in the spinal cord of ischemic rats. The present

295

results suggest that down-regulation of the expression of GTs (GLSAT, GLT-1, and EAAC1), GDH, and GS might contribute to an increase in EAA-induced neurotoxicity in the spinal cord of ischemic rats. Both spinal cord ischemiainduced neurological dysfunction and the down-regulation of the expression of proteins of the glutamate metabolizing systems were significantly inhibited by i.t. MP injections.

EXPERIMENTAL PROCEDURES Implantation of i.t. catheters Male Wistar rats (400 – 450 g) were used in the experiments. As shown in our previous study (Wen et al., 2005), i.t. catheters (PE5 tubes 9 cm, 0.008 inch inner diameter, 0.014 inch outer diameter) were inserted via the atlantooccipital membrane into the i.t. space at the level of the lumbar enlargement of the spinal cord and externalized and fixed to the cranial aspect of the head. The rats were then returned to their home cages for a 4-day recovery period. Each rat was housed individually and lived on a 12-h light/dark daily cycle with food and water freely available. Rats were excluded from the study if they showed evidence of gross neurological injury or the presence of fresh blood in the cerebrospinal fluid (CSF). The use of animals conformed to the Guiding Principles in the Care and Use of Animals of the American Physiology Society and was approved by the National Sun-Sen University Animal Care and Use Committee. Every effort was made to minimize the number of animals used and their suffering.

The induction of spinal cord ischemia in rats Male Wistar rats (400 – 420 g) implanted with one i.t. catheter were used. This rat spinal cord ischemia model was a modification of the one that was previously described by Taira and Marsala (1996). Briefly, animals were anesthetized in a plastic box with 4% isoflurane in room air. After induction, 2.5% isoflurane in an air/O2 mixture was delivered to the rats through a mask. The tail artery was cannulated with a 22-gauge polytetrafluoroethylene catheter in order to monitor distal arterial pressure and to administer heparin intra-arterially. The left carotid was cannulated with a 20gauge polytetrafluoroethylene catheter in order to collect the blood sample. In order to induce spinal cord ischemia, the left femoral artery was exposed. Aortic occlusion was induced by the inflation of a 2F Fogarty catheter placed into the thoracic aorta for 12 min, and the left carotid artery was also cannulated with a 20-gauge polytetrafluoroethylene catheter to collect blood and pump it into the peripheral stream during aortic occlusion. Immediately after the completion of arterial cannulations, all rats received 100 U of heparin (0.1 ml) through the tail artery. After the completion of all surgical procedures, 0.4 ml of protamine sulfate (4 mg) was administered intraperitoneally. Animals were then returned to their cages for recovery of motor function and finally killed for spinal sample collection.

Study groups and MP treatment After i.t. catheter insertion, the rats were randomly divided into the following four groups: (1) the control group (C, n⫽8), in which an i.t. injection of saline was administered and a balloon catheter was placed in the thoracic aorta without inflation; (2) the ischemiaoperated group (I, n⫽10), in which an i.t. injection of saline was given and a balloon catheter was placed in the thoracic aorta with inflation; (3) the ischemia plus MP group (I⫹M, n⫽6), in which an i.t. injection of MP (200 ␮g twice daily) was administered from 3 days prior to the induction of ischemia; and (4) the only MPadministered group (M, n⫽6), in which ischemia was not induced. Rats from each group were killed 48 h after ischemia had been induced and the behavioral test had been performed.

296


Evaluation of the neurobehavioral outcome

Histopathology

Motor function deficits in the hind limbs were evaluated according to the following system, which was modified from Marsala and Yaksh’s (1994) report. A motor deficit index (MDI) was calculated for each rat at each assessment based on the following criteria: 0⫽normal gait; 1⫽mildly impaired gait, i.e. toes flat under body, but weakness or spasticity apparent; 2⫽knuckle walking; 3⫽dragging legs with movement in the knees; 4⫽complete paraplegia with or without spasticity. Spasticity was defined as the continuous or intermittent tonic positioning of the hind limbs in extension, particularly with the feet in plantar flexion. We also assessed the placing/stepping reflex by dragging the dorsum of the hind paw over the edge of a surface. This normally evokes a coordinating lifting and placing response that was graded as follows: 0, normal; 1, weak; and 2, no stepping. The MDI was calculated for each rat as the sum of both scores, with the maximum score being 6.

Under deep anesthesia with sodium pentobarbital (100 mg/kg), rats were first perfused intracardially with 500 ml of cold phosphate-buffered saline (PBS) containing 1% sodium nitrite and heparin (0.2 U/ml) and then with 4% paraformaldehyde in 500 ml of 0.1 M phosphate buffer (PB; pH 7.4). The spinal cord samples were then harvested and post-fixed in the same fixative for 3 days before they were embedded in paraffin. The spinal cord segments were cut on a cryostat into sections of thickness 3 ␮m. Transverse sections were obtained through the middle of the 4th lumbar, 10th thoracic, 3rd thoracic, and 5th cervical spinal segments. Three individual spinal cords were used at each group. The sections were stained with hematoxylin and eosin (H&E) for histopathological observations. Spinal cords of four animals taken from each experimental group were used for the analysis. The spinal cord sections were photographed in a series of five frames, and were assessed by two examiners in a double blind assessment. The measurement of lymphocyte infiltration was present in number/mm2 by the average of five frame of each rat. The total number of apparently viable ventral horn motor neurons was determined in each section. Ventral horn motor neurons were assessed by morphologic observation of motoneurons as reported previously (Lips et al., 2000) and defined according to the following criteria: fine granular cytoplasm with basophilic stippling, prominent nucleoli, and a soma diameter of 30 – 60 ␮m. Results were expressed the percentage of eosinophilic motor neurons.

Preparation of spinal cord tissue homogenates and Western blotting Rats anesthetized with isoflurane were rapidly decapitated, and the dorsal and ventral portions of the cervical (C2–C5), thoracic (T3–T10), and lumbar (L2–L5) spinal cords were then removed. Six individual spinal cord homogenates were used at each group. The dorsal portions of the spinal cord were homogenized in an ice-cold lysis buffer (50 mM Tris, pH⫽7.5, 150 mM NaCl, 2% Triton X-100, 100 ␮g/ml phenylmethylsulfonyl fluoride, 1 ␮g/ml aprotinin), and then centrifuged at 68,000 r.p.m. (TXL-100, Beckman, Fullerton, CA, USA) for 30 min at 4 °C. The supernatant was decanted from the pellet and retained for Western blot analysis. Protein concentrations were determined by the DC protein Assay kit (Bio-Rad, Hercules, CA, USA) modified form by the method of Lowry et al. (1951). Western blotting was performed as our previous study (Wen et al., 2005). In brief, an equal volume of sample buffer (2% SDS, 10% glycerol, 0.1% Bromophenol Blue, 2% 2-mercaptoethanol, and 50 mM Tris–HCl, pH 7.2) was added to the sample, which was then loaded onto a 10% SDS–polyacrylamide gel and electrophoresis performed at 150 V for 60 min. For GTs analysis as well as GS and GDH analysis, the spinal homogenates sample was taken 100 ␮g and 10 ␮g respectively. The proteins were transferred to a polyvinylidene difluoride membrane (Immobilon-P, Millipore, Billerica, MA, USA; 0.45 ␮M pore size) at 125 mA overnight at 4 °C in transfer buffer (50 mM Tris–HCl, 380 mM glycine, 1% SDS, and 20% methanol), then the membrane was blocked for 50 min at room temperature with 5% non-fat dry milk in 0.1% Tween 20 in 20 mM Tris–HCl, 137 mM NaCl, pH 7.4 (TTBS), and incubated for 180 min at room temperature with antibody against GLAST (EAT1, 1:1000 dilution; Chemicon, Temecula, CA, USA), EAAC1 (EAT3, 1:1000 dilution; Chemicon), GLT-1 (EAT2, 1:1000 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA), GDH (1:1000 dilution; Biogenesis, Poole, UK; cat no: 4670 –5488), GS (1:1000 dilution; Biogenesis, cat no:4673–5007) or glial fibrillary acid protein (GFAP; 1:2000 dilution; Chemicon). It was then washed three times in TTBS for 10 min, blocked with 5% non-fat dry milk/TTBS, then incubated for 1 h at room temperature with horseradish peroxidase-conjugated secondary antibodies (dilution 1:2000). The GLAST antibody recognized at a band at ⬃65 kDa, GLT-1 antibody recognized at a band at ⬃70 kDa, EAAC1 antibody recognized at a band at ⬃70 kDa, GDH antibody recognized at a ⬃55 kDa, GS antibody recognized at a band at ⬃55 kDa and GFAP antibody recognized at a band at ⬃ 52 kDa. The blots were then visualized in ECL solution (NEN) for 30 s and finally exposed to X-ray film (Koda X-OMAT LS, Kodak, Rochester, NY, USA) and quantified by densitometry. The membranes were re-probed with a monoclonal mouse anti-␤-actin antibody (1:2500, Sigma, St. Louis, MO, USA) as the loading control. Each membrane was re-probed on one occasion.

Statistical analysis All data are presented as the mean⫾S.E.M. The physiological parameters were analyzed by one-way analysis of variance (ANOVA) followed by a post hoc Tukey’s test. In the immunoreactivity data, the intensity of each test band was expressed as the relative optical density (ROD) calculated with respect to the average control optical density value for all of the control X-ray films. Wherever applicable, data from testing were analyzed using ANOVA followed by Dunnett’s test. A P value less than 0.05 was considered to be statistically significant.

RESULTS The effect of i.t. MP administration on spinal cord ischemia-induced neurological dysfunction The dose of i.t. MP was calculated and modified according to the method of Xu et al. (2001). In general, the i.t. doses of drugs are approximately 1%–2% of their systemic dose. However, a single i.t. injection of 100 or 200 ␮g MP did not attenuate the spinal cord ischemia-induced neurological dysfunction at 15 min after the surgery for spinal cord ischemia; this was our preliminary finding. The i.t. pretreatment with i.t. 200 ␮g MP twice daily for 3 days prior to the surgery for spinal cord ischemia produced a slight but not significant improvement in the neurological deficit. All the rats in the control group and those treated with i.t. MP alone exhibited a normal neurological outcome (MDI⫽0). In the ischemia group, all the rats exhibited flaccid paraplegia after recovery from anesthesia and then developed spasms over the next 48 h (MDI⫽6). The MDI values recorded at 4 h, 12 h, 24 h, and 48 h after ischemia in the spinal cord ischemia plus MP group were 3.88⫾0.26, 3.44⫾0.29, 2.66⫾0.33, and 1.77⫾0.44, respectively. The i.t. MP pretreatment (200 ␮g twice daily for 3 days prior to the surgery for ischemia) significantly


297

rological function was not affected by the i.t. administration of MP alone (Fig. 1). The effect of i.t. MP administration on spinal cord ischemia-induced changes in the expression of proteins of the glutamate metabolizing systems

Fig. 1. Time course of the effect of i.t. MP pretreatment on the MDI following spinal cord ischemia. The neurological status of the ischemia plus MP group was significantly better than that of the control group at 4, 12, 24, and 48 h after induction of spinal cord ischemia. (a) P⬍0.05 compared with the control group; (b) P⬍0.05 compared with the ischemia group; (c) P⬍0.05 compared with the ischemia group at 4 h.

inhibited the spinal cord ischemia-induced neurological dysfunction during 4 – 48 h after the surgery. The neu-

From Fig. 2, it can be inferred that the expressions of GLAST, GLT-1, and GFAP proteins were significantly increased in the ventral and dorsal parts of the cervical spinal cord at 48 h after the spinal cord ischemia. The ischemia-induced upregulation of GLAST, GLT-1, and GFAP protein expression was significantly prevented by the i.t. MP. Both ischemia and MP alone administration did not interfere with the expressions of EAAC1, GS, and GDH in the cervical spinal cord. The expressions of the three proteins—GTs-GLAST, GLT-1, and EAAC1—were significantly reduced in the ventral and dorsal parts of the thoracic and lumbar spinal cord after the spinal cord ischemia (Fig. 3 and 4). This reduction in the expressions of the GLAST, GLT-1, and EAAC1 proteins was significantly inhibited by the i.t. MP pretreatment. However, MP could not completely inhibit the reduction of the GLAST expression in the ventral part of the thoracic cord (Fig. 3). The expressions of the GS and GDH proteins in the thoracic spinal cord were not affected by the spinal cord ischemia injury.

Fig. 2. Protein/␤-actin density ratios upon Western blotting of the expression of the GLAST, GLT-1 and GFAP proteins following various treatments of the dorsal and ventral regions of the cervical (segments 2–5) spinal cord. Control values are defined as 100%. D, dorsal region; V, ventral region; C, cervical 2–5; C, control; I, ischemia after 48 h; I⫹MP, pretreatment with ischemia plus MP (i.t. injection 200 ␮g twice daily for 3 days before surgery); MP, i.t. MP alone (200 ␮g twice daily for 3 days). An asterisk indicates statistical significance (P⬍0.05) compared with the control.

298


Fig. 3. Protein/␤-actin density ratios upon Western blotting of the expression of the GLAST, GLT-1 and EAAC1 proteins following various treatments of the dorsal and ventral regions of the thoracic (segments 3–10) spinal cord. Control values are defined as 100%. D, dorsal region; V, ventral region; C, control; I, ischemia after 48 h; I⫹MP, pretreatment with ischemia plus MP (i.t. injection 200 ␮g twice daily for 3 days before surgery); MP, i.t. MP alone (200 ␮g twice daily for 3 days). An asterisk indicates statistical significance (P⬍0.05) compared with the control. # P⬍0.05 compared with the ischemia group.

Spinal cord ischemia induced the down-regulation of GLAST, GLT-1, GS, GDH, and EAAC1 in the lumbar spinal cord, and this down-regulation was completely inhibited by the i.t. MP pretreatment (Figs. 4 and 5). This MP pretreatment could significantly inhibit the ischemia-induced decrease in the EAAC1 expression in the ventral but not in the dorsal part of the lumbar spinal cord (Fig. 4). The i.t. administration of MP alone did not influence the protein expression in any of the glutamate metabolizing systems at any spinal cord segment. Histopathological observations The histopathological findings of the spinal cord in the four groups at 48 h after surgery are shown in Fig. 6. In the present study, we found that spinal cord ischemia induced a caudal–rostral gradient of spinal damage; the most severe damage was observed in the lumbar spinal cord (Fig. 6); less damage, in the lower thoracic cord; and no damage, in the cervical cord. The lumbar spinal section from rats with an MDI of 5 or 6 (ischemia group) showed the presence of a large number of eosinophilic neurons distributed in the gray matter (Figs. 6 and Fig. 7). Neurons exhibiting the morphological features of necrosis were frequently observed. Moreover, we also observed pale areas representing infarcts in the vacuolated gray matter in the

lumbar cord of the ischemic rats. The infarcted areas were characterized by the destruction of normal tissue as well as by the conspicuous presence of large numbers of infiltrating neutrophils and mononuclear phagocytes. The gray matter infarcts did not extend into the white matter after spinal cord ischemia even in the cases of severe injury. The i.t. MP pretreatment could significantly inhibit the severe spinal cord pathology that was induced by the spinal cord ischemia. Eosinophilic neurons, tissue destruction, and lymphocyte infiltration were less severe in the ischemia plus MP group than in the ischemia groups (Fig. 7). However, the red neuron and ghost neurons were also observed in the ischemia plus MP rats. Sections from the lumbar spinal cord, which were obtained from both control and MP alone groups, showed either no abnormalities or the occasional presence of a few isolated eosinophilic neurons in the gray matter.

DISCUSSION This study has demonstrated spinal cord ischemia-induced neurological dysfunctions. Forty-eight hours after spinal cord ischemia, significant down-regulation of GLAST, GLT-1, and EAAC1 was observed in both the ventral and dorsal parts of the lumbar and thoracic regions of the


299

Fig. 4. Protein/␤-actin density ratios after Western blotting of the expression of the GLAST, GLT-1 and EAAC1 proteins following various treatments of the dorsal and ventral regions of the lumbar (segments 2–5) spinal cord. Control values are defined as 100%. D, dorsal region; V, ventral region; L, lumbar 2–5; C, control; I, ischemia after 48 h; I⫹MP, pretreatment with ischemia plus MP (i.t. injection 200 ␮g twice daily for 3 days before surgery); MP, i.t. MP alone (200 ␮g twice daily for 3 days). An asterisk indicates statistical significance (P⬍0.05) compared with the control. # P⬍0.05 compared with the ischemia group.

spinal cord. When the i.t. MP injection was administered for 3 days prior to the induction of ischemia, the spinal cord ischemia-induced neurological defects were significantly attenuated, and the down-regulation of GT expression in the lumbar and thoracic spinal cord was significantly blocked. Interestingly, 48 h after spinal cord ischemia, there was a significant increase in the expression of GLAST and GLT-1 proteins in the dorsal and ventral cervical spinal cords. The i.t. MP injection alone had no effect on the neurological functions and the expression of proteins of the glutamate metabolizing system in the spinal cord of rats. The role of GTs in spinal cord ischemia In rats, it is known that after spinal cord ischemia injury, there is a significant increase in the levels of neurotoxic glutamate in the CSF, and this is accompanied by a neurological deficit (Marsala et al., 1994). During CNS ischemia, the extracellular glutamate concentration increases, reaching levels that activate the N-methyl-D-aspartate (NMDA) type of glutamate receptor, thereby causing neuronal death (Sattler and Tymianki, 2001). GTs are membrane proteins that reuptake glutamate into the nerve terminals or glial cells in order to remove it from the extracellular space around neurons. Glutamate removal is essential for maintaining functional communication be-

tween neurons and extremely important for preventing the concentration of glutamate from reaching neurotoxic levels (Danbolt, 2001). It has been know that excessive glutamate accumulation in the extracellular space occurs in various neurological degeneration diseases. The positive glutamate transport regulator riluzole has demonstrated efficacy in ischemia spinal cord injury in rats (Lang-Lazdunski et al., 2000). In the present study, the down-regulation of the expression of the GT proteins in the lumbar spinal cord correlated well with the severity of hindlimb paraparesis. The i.t. MP pretreatment significantly inhibits spinal cord ischemia-induced GLAST and GLT-1 downregulation in the lumbar spinal cord. However, MP did not completely prevent the ischemia-induced reduction in EAAC1 expression in the lumbar cord. The results also correlated with the histopathological observations of motor neuron damage and the presence of some infiltrating cells in the lumbar spinal region in the ischemia plus MP group. Moreover, the i.t. MP injection could significantly protect the spinal cord ischemia-induced neurological defects and spinal pathophysiology. In the present study, using the spinal cord ischemia animal model, we only observed the severity of hindlimb paraparesis and changes in the pathology and glutamate metabolizing systems of the thoracic and lumbar spinal cords in rats. Based on past studies and our present observations, the spinal cord ischemia

300


Fig. 5. Protein/␤-actin density ratios after Western blotting of the expression of the GS and GDH following various treatments of the dorsal and ventral regions of the lumbar (segments 2–5) spinal cord. Control values are defined as 100%. D, dorsal region; V, ventral region; L, lumbar 2–5; C, control; I, ischemia after 48 h; I⫹MP, pretreatment with ischemia plus MP (i.t. injection 200 ␮g twice daily for 3 days before surgery); MP, i.t. MP alone (200 ␮g twice daily for 3 days). An asterisk indicates statistical significance (P⬍0.05) compared with the control.

animal model mainly affects the motor function of the lower half of the body (Taira and Marsala, 1996). On the other hand, the upper body appears to be unaffected by the disease. This observation is in agreement with anatomical knowledge and the down-regulation of neuroprotective proteins (GTs and intracellular glutamate metabolizing enzymes) at the lumbar spinal cord level. Surprisingly, in the present study, spinal cord ischemia significantly resulted in the upregulation of GLAST, GLT-1 and GFAP protein expression in the cervical spinal region. As GFAP is a commonly accepted astrocytic marker, changes in GFAP protein expression are normally considered to represent the astroglial response in spinal cord ischemia model. It is possible that the increase in GT protein expression in the cervical spinal cord tissue is a compensatory mechanism of the body that is a response to the ischemic insult. It is proposed that cervical spinal neurons are protected by an increase in the reuptake of extracellular neurotoxic glutamate via up-regulation of glial GT GLAST and GLT-1.

injury, it has been identified that the oxidative damage also is an important contributing factor (Chan, 1996). Increasing evidence indicates extracellular glutamate uptake inhibition by oxidative stress (Blanc et al., 1998; Muller et al., 1998; Chen et al., 2000). A significant protection against oxidative stress can be provided by superoxide dismutase (SOD) and SOD is down-regulated following spinal cord ischemia injury (Erten et al., 2003). Furthermore, it has been established that significant down-regulation of GLT-1 protein of spinal cord in mice with SOD causes mutation (Bendotti et al., 2001). The oxidative stress might play an important role in down-regulation of GT protein expression in spinal cord ischemic rats. Schmidt et al. (2002) have also reported that MP can produced anti-oxidative functions in neuronal and glial cell. On the basis of the above evidence and data, we suggest that the MP-attenuated cord ischemia-induced down-regulation of GTs might be via inhibition of oxidative stress. The effect of MP on GS and GDH

The effect of MP on GTs GTs are down-regulated in the lumbar and thoracic spinal cord of ischemic rats by an as yet unknown mechanism, and glutamate-induced toxicity is well known for oxidative stress after neuronal injury (Dykens et al., 1987; LafonCazal et al., 1993). For cell death following central neural

GTs and the glutamate-metabolizing enzymes—GS and GDH—are found to be widely expressed in glial cells, particularly in the astrocytes that surround the glutamatergic synapse, and play an important role in maintaining normal EAA concentrations in the neuronal synapse (Miller et al., 2002). Tight control of the extracellular glutamate


301

Fig. 6. H&E staining of the lumbar spinal cord after a sham operation (A, B, C), ischemia after 48 h (D, E, F), and pretreatment with ischemia plus MP (G, H, I). Two days after ischemia, spastic paraplegia was observed; the gray matter appeared to be spongy and vacuolated (D, E arrow), and motor neurons (F, arrow) showed ischemic cell changes. Many infiltrating cells can be observed in the gray matter of the lumbar spinal cord after ischemia (D, E). The i.t. MP pretreatment significantly decreases the number of cells infiltrating into the gray matter of the lumbar spinal cord (G). Some motor neurons were observed in which nuclear hematoxylin staining was absent, i.e. they appeared like ghost neurons (H, arrow). I, the arrow points to surviving motor neurons, and some necrotic neurons (arrowhead) showed eosinophilia upon H&E staining (red neurons).

concentration in the synapse is crucial not only for nociception transmission but also for preventing glutamateinduced neurotoxicity. The uptake of glutamate by glial cells prevents overexcitation of the neurons and excitotoxicity. Glutamate is then metabolized by GS and GDH into the neutral metabolites glutamine and 2-oxoglutarate, respectively. Moreover, Gemba et al. (1994) have also suggested that ischemia-induced CNS acidosis (including an increase in glutamate) can lead to compromised GTs in astrocytes and therefore enhanced potential for excitotoxicity. Thus, GS and GDH also play important roles in modulating extracellular glutamate levels. The ventral parts of the spinal cord contain motor neurons that directly control motor function. In the present study, following spinal cord ischemia for 48 h, GS and GDH were also significantly down-regulated in the ventral part of the lumbar spinal cord. This might result in glutamate-induced neurotoxicity of the ventral horn of the spinal cord, which controls motor functions. Similar to the effects of GTs, the down-

regulation of GS and GDH expression might also contribute to a decrease in glutamate metabolism and enhance glutamate-induced toxicity. Taken together, these results suggest that in addition to the down-regulation of GT expression the reduction in GDH and GS expression following spinal cord ischemia can lead to an enhanced excitotoxic potential and is a major contributor to the resulting histological and motor function damage. The possible mechanism of MP in neuroprotection Glucocorticoids are secreted from adrenal glands and bind to intracellular corticoid receptors; they subsequently activate transcription factors and influence the expression of different genes in the CNS. The corticosteroids can suppress local inflammatory responses that produce edema and swelling and worsen neuronal injuries. They may, via the inhibition of pathological stress, induce the release of proinflammatory mediators (prostaglandins, interleukin-1␤,

302


Fig. 7. Quantitative assessment of lymphocyte infiltration (n/mm2) (A), eosinophilic motor neurons (B) in gray matter of the ventral lumbar spinal cord of animals killed after 48 h of ischemia surgery. An asterisk indicates statistical significance (P⬍0.05) compared with the control. # P⬍0.05 compared with the ischemia group.

tumor necrosis factor-␣, and nitric oxide) (Chikawa et al., 2001; Bartholdi and Schwab, 1995; Rhodes, 2003). Many reports have indicated that glutamate induces the upregulation of proinflammatory mediators in the nervous system (Minami et al., 1991; Yabuuchi et al., 1993). Zou and Crews (2005) demonstrated that inhibition of glutamate uptake by the proinflammatory mediator TNF-␣ could lead to potential glutamate-induced neurotoxicity. Korn et al. (2005) had demonstrated that the activation of the proinflammatory mediator tumor necrosis factor-␣ pathway decreased GALST protein expression in astrocytes. MP may regulate the protein expression of GTs via an anti-inflammatory pathway. The GS and GDH genes contain a glucocorticoid response element (GRE), and glucocorticoids promote GS and GDH expression (Vardimon et al., 1988; Hardin-Pouzet et al., 1996). It is surprising that MP alone did not increase GS and GDH expression in nonischemic rats in the study; this observation contradicts several previous reports that glucocorticoids regulate GS and GDH expression (Moscona, 1975; Gorovits et al., 1996). We hypothesized that glucocorticoids plus ischemic insult could modulate GS and GDH expression and that the anti-inflammatory activity of MP would indirectly inhibit the down-regulation of GTs during neuronal ischemia.

CONCLUSION In conclusion, after spinal cord ischemia, the expression of the GT proteins and intracellular glutamate metabolizing enzymes was decreased. This might contribute to the accumulation of extracellular glutamate, resulting in neuronal damage, which eventually leads to neurological dysfunctions. The increased expression of GLAST and GLT-1 in the cervical spinal cord observed in this study provides an insight into the events that follow spinal cord ischemia. The data suggest that glial GTs are important in regulating glutamate homeostasis. They prevent the accumulation of

excess glutamate and therefore further reduce the level of toxins in the cervical region after spinal cord injury. Both spinal cord ischemia-induced neurological defects and changes in the expression of proteins of the glutamate metabolizing system were significantly and directly inhibited by i.t. MP administration. We anticipate that i.t. MP delivery may provide a significant advantage in cases where surgery is performed in the descending thoracic or thoracoabdominal aorta as a result of neurological dysfunctions. Acknowledgments—This study is supported by the “Aim for the Top University Plan” of the National Sun Yat-Sen University and Ministry of Education, Taiwan. It is also partly supported by research grants SKH-TMU-94-19 and CMRPG850201 from the Shin Kong Wu Ho-Su Memorial Hospital and Chang Gung Memorial Hospital, respectively.

REFERENCES Arundine M, Tymianski M (2003) Molecular mechanisms of calciumdependent neurodegeneration in excitotoxicity. Cell Calcium 34:325–337. Bartholdi D, Schwab ME (1995) Methylprednisolone inhibits early inflammatory processes but not ischemic cell death after experimental spinal cord lesion in the rats. Brain Res 672:177–186. Bendotti C, Tortarolo M, Suchak SK, Calvaresi N, Carvelli L, Bastone A, Rizzi M, Rattray M, Mennini T (2001) Transgenic SOD1 G93A mice develop reduced GLT-1 in spinal cord without alterations in cerebrospinal fluid glutamate levels. J Neurochem 79:737–746. Blanc EM, Keller JN, Fernandez S, Mattson MP (1998) 4-Hydroxynonenal, a lipid peroxidation product, impairs glutamate transport in cortical astrocytes. Glia 22:149 –160. Chan PH (1996) Role of oxidants in ischemic brain damage. Stroke 27:1124 –1129. Chen Y, Ying W, Simma V, Chen Y, Copin JC, Chan PH, Swansom RA (2000) Overexpression of Cu, Zn superoxide dismutase attenuates oxidative inhibition of astrocyte glutamate uptake. J Neurochem 75:939 –945.

G.-J. Wu et al. / Neuroscience 147 (2007) 294 –303 Chikawa T, Ikata T, Katoh S, Hamada Y, Kogure K, Fukuzawa K (2001) Preventive effects of lecithinized superoxide dismutase and methylprednisolone on spinal cord injury in rats: transcriptional regulation of inflammatory and neurotrophic genes. J Neurotrauma 18:93–103. Danbolt NC (2001) Glutamate uptake. Prog Neurobiol 65:1–105. Dykens JA, Stern A, Trekner E (1987) Mechanism of kainate toxicity to cerebellar neurons in vitro is analogous to reperfusion tissue injury. J Neurochem 49:1222–1228. Erten SF, Kocak A, Ozdemir I, Aydemir S, Colak A, Reeder BS (2003) Protective effect of melatonin on experimental spinal cord ischemia. Spinal Cord 41:533–538. Gemba T, Oshima T, Ninomiya M (1994) Glutamate efflux via the reversal of the sodium-dependent glutamate transporter caused by glycolytic inhibition in rat cultured astrocytes. Neuroscience 63: 789 –795. Gorovits R, Yakir A, Fox LE, Vardimon L (1996) Hormonal and nonhormonal regulation of glutamine synthetase in the developing neural retina. Mol Brain Res 43:321–329. Hardin-Pouzet H, Giraudon P, Belin MF, Didier-Bazes M (1996) Glucocorticoid upregulation of glutamate dehydrogenase gene expression in vitro in astrocytes. Mol Brain Res 37:324 –328. Kaneko T, Shigemoto R, Mizuno N (1988) Metabolism of glutamate and ammonia in astrocyte: an immunocytochemical study. Brain Res 457:160 –164. Korn T, Magnus T, Jung S (2005) Autoantigen specific T cells inhibit glutamate uptake in astrocytes by decreasing expression of astrocytic glutamate transporter GLAST: a mechanism mediated by tumor necrosis factor-alpha. FASEB J 19:1878 –1880. Kotani N, Kushikata T, Hashimoto H, Kimura F, Muraoka M, Yodono M, Asai M, Matssuki A (2000) Intrathecal methylprednisolone for intractable postherpetic neuralgia. N Engl J Med 343:1514 –1519. Kouchoukos NT (1991) Spinal cord ischemia injury: is it preventable? Semin Thorac Cardiovasc Surg 3:323–328. Koyanagi I, Tator CH, Lea PJ (1993) Three-dimensional analysis of the vascular system in the rat spinal cord with scanning electron microscopy of vascular corrosion casts, part 1: Normal spinal cord. Neurosurgery 33:277–284. Lafon-Cazal M, Pietri S, Culcasi M, Bockaert J (1993) NMDA-dependent superoxide production and neurotoxicity. Nature 364:535–537. Lang-Lazdunski L, Heurteaux C, Mignon A, Mantz J, Widmann C, Desmonts J, Lazdunski M (2000) Ischemic spinal cord injury induced by aortic cross-clamping: prevention by riluzole. Eur J Cardiothorac Surg 18:174 –181. Lehre KP, Levy LM, Ottersen OP, Storm-Mathisen J, Danbolt NC (1995) Differential expression of two glial glutamate transporters in the rat brain: quantitative and immunocytochemical observations. J Neurosci 15:1835–1853. Lips J, Haan P, Bodewits P, Vanicky I, Dzoljic M, Jacobs MJ, Kaklman CJ (2000) Neuroprotective effects of riluzole and ketamine during transient spinal cord ischemia in the rabbit. Anesthesiology 93:1303–1311. Lowry DH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the folin phenol reagent. J Biol Chem 193:265–275. Marsala M, Sorkin LS, Yaksh TL (1994) Transient spinal ischemia in rat: characterization of spinal cord blood flow, extracellular amino acid release, and concurrent histopathological damage. J Cereb Blood Flow Metab 14:604 – 614. Martin DL, Waniewski RA (1996) Precursor synthesis and neurotransmitter uptake by astrocyte as target of neurotoxicants. In: The role of glial in neurotoxicity (Aschner M, Kimelberg HK, eds), pp 335– 357. Boca Raton: CRC Press.

303

Miller KE, Richards BA, Kriebel RM (2002) Glutamine-, glutamine synthetase-, glutamate dehydrogenase- and pyruvate carboxylase-immunoreactivities in the rat dorsal root ganglion and peripheral nerve. Brain Res 945:202–211. Minami M, Kuraishi Y, Satoh M (1991) Effects of kainic acid on mRNA levels of IL-1 beta, IL-6, TNF alpha and LIF in the rat brain. Biochem Biophys Res Commun 176:593–598. Moscona AA (1975) Hydrocortisone-mediated regulation of gene expression in embryonic neural retina: induction of glutamine synthetase. J Steroid Biochem 6:633– 638. Muller A, Maurin L, Bonne C (1998) Free radicals and glutamate uptake in the retina. Gen Pharmacol 30:315–318. Nicklas WJ (1984) Amino acid metabolism in the central nervous system: role of glutamate dehydrogenase. In: The olivopontocerebellar atrophies (Duvoisin RC, Platitakis A, eds), pp 245–253. New York: Raven. Nishizawa Y (2001) Glutamate release and neuronal damage in ischemia. Life Sci 69:369 –381. Rhodes JK (2003) Actions of glucocorticoids and related molecules after traumatic brain injury. Curr Opin Crit Care 9:86 –91. Rokkas CK, Cronin CS, Nitta T, Helfrich LR Jr, Lobner DC, Choi DW, Kouchoukos NT (1995) Profound systemic hypothermia inhibits the release of neurotransmitter amino acids in spinal cord ischemia. J Thorac Cardiovasc Surg 110:27–35. Rothstein JD, Martin L, Levey AI, Dykes-Hoberg M, Jin L, Wu D, Nash N, Kuncl RW (1994) Localization of neuronal and glial glutamate transporter. Neuron 13:713–725. Sattler R, Tymianki M (2001) Molecular mechanisms of glutamate receptor-mediated excitotoxic neuronal cell death. Mol Neurobiol 24:107–129. Schmidt AJ, Krieg JC, Vedder H (2002) Differential effects of glucocorticoids and gonadal steroid on glutathione levels in neuronal and glial cell systems. J Neurosci Res 67:544 –550. Scremin OU (1995) Cerebral vascular system. In: The rat nervous system, 2nd edition (Paxinos GT, ed), pp 3–35. San Diego, CA: Academic Press. Svensson LG Crawford ES, Hess KR, Coselli JS, Safi HJ (1993) Experience with 1509 patient undergoing thoracoabdominal aortic operations. J Vasc Surg 17:357–370. Taira Y, Marsala M (1996) Effect of proximal arterial perfusion pressure on function, spinal cord blood flow, and histopathologic changes after increasing intervals of aortic occlusion in the rat. Stroke 27:1850 –1858. Vardimon L, Fox LL, Degenstein L, Moscona AA (1988) Cell contacts are required for induction by cortisol of glutamine-synthetase genetranscription in the retina. Proc Natl Acad Sci U S A 856: 5981–5985. Wen ZH, Wu GJ, Chang YC, Wang JJ, Wong CS (2005) Dexamethasone modulates the development of morphine tolerance and expression of glutamate transporters in rats. Neuroscience 133: 807– 817. Xu J, Kim GM, Ahmed SH, Xu J, Yan P, Xu XM, Hsu CY (2001) Glucocorticoid receptor-mediated suppression of activator protein-1 activation and matrix metalloproteinase expression after spinal cord injury. J Neurosci 21:92–97. Yabuuchi K, Minami M, Katsumata S, Satoh M (1993) In situ hybridization study of interleukin-1 beta mRNA induced by kainic acid in the rat brain. Mol Brain Res 20:153–161. Zou JY, Crews FT (2005) TNF␣ potentiates glutamate neurotoxicity by inhibiting glutamate uptake in organotypic brain slice cultures: neuroprotection by NF␬B inhibition. Brain Res 1034:11–24.

(Accepted 10 April 2007) (Available online 1 June 2007)