Regulation of interstitial excitatory amino acid concentrations after cortical contusion injury

Increases in brain interstitial excitatory amino acid (EAA(I)) concentrations after ischemia are ameliorated by use-dependent Na+ channel antagonists and by supplementing interstitial glucose, but the regulation of EAA(I) after traumatic brain injury (TBI) is unknown. We studied the regulation of EAA(I) after TBI using the controlled cortical impact model in rats. To monitor changes in EAA(I), microdialysis probes were placed in the cortex adjacent to the contusion and in the ipsilateral hippocampus. Significant increases in dialysate EAA(I) after TBI were found compared to levels measured in sham controls. Treatment with the use-dependent Na+ channel antagonist 619C89 (30 mg/kg i.v.) did not significantly decrease dialysate glutamate compared to vehicle controls in hippocampus (10.4+/-2.4 vs. 11.9+/-1.6 microM), but there was significant decrease in dialysate glutamate in cortex after 619C89 treatment (19.3+/-3 vs. 12.6+/-1.1 microM, P<0.05). Addition of 30 mM glucose to the dialysate, a treatment that decreases EAA(I) after ischemia, had no significant effect upon dialysate glutamate after TBI in cortex (20.0+/-4.9 vs. 11.7+/-3.4 microM) or in hippocampus (10.9+/-2.0 vs. 8.9+/-2.4 microM). These results suggest that neither increased release of EAAs due to Na+ channel-mediated depolarization nor failure of glutamate reuptake due to glucose deprivation can explain the majority of the increase in EAA(I) following TBI.


Abstract
Increases in brain interstitial excitatory amino acid (EAA ) concentrations after ischemia are ameliorated by use-dependent Na1 I channel antagonists and by supplementing interstitial glucose, but the regulation of EAA after traumatic brain injury (TBI) is unknown. I We studied the regulation of EAA after TBI using the controlled cortical impact model in rats. To monitor changes in EAA , I I microdialysis probes were placed in the cortex adjacent to the contusion and in the ipsilateral hippocampus. Significant increases in dialysate EAA after TBI were found compared to levels measured in sham controls.  : depolarization of neurons mediated by Na1  ing, Monee, IL, USA), mounted on an adjustable crossbar, channels, and decreased supply of glucose. Use-dependent was positioned to provide vertical impact with a 5-mm inhibitors of brain (Type II) Na1 channels prevent pro-rounded impactor tip. Velocity and duration of impact were longed, pathological opening of neuronal sodium channels determined using an integrated accelerometer and comand thus ameliorate depolarization of neurons [5,11,13]  Isoflurane was lowered to 2% for surgery (1% for maintenance, and rats ventilated (Harvard Small Animal Ven-2.4. Sham versus CCI groups tilator, Harvard Apparatus, South Natick, MA, USA) with a tidal volume of 1 cc / 100 g, adjusting the rate to yield Rats were divided into two groups (n54-8 per group), PaO .90 mmHg and PaCO 35-45 mmHg. The right 2 2 CCI and sham, with all surgeries and physiologic paramefemoral artery was cannulated using PE50 tubing for ters identical except for the omission of CCI. Microdialysis continuous blood pressure monitoring, blood gases, hemaprobes were prepared and loaded as described above, tocrit and glucose.
collecting dialysate 1 h prior to trauma. CCI animals Rats were placed in a stereotactic device (Kopf Instruunderwent trauma while sham animals had probes removed ments, Tujunga, CA, USA) and prepared for trauma by for 2 min with no CCI. Probes for both groups were exposure of the left parietal cortex via craniotomy. A brain reinserted and dialysate collected every 10 min for 1 h temperature probe (Physitemp IT23, Physitemp Instruafter trauma. ments, Clifton, NJ, USA) was placed in the ipsilateral frontal cortex via a burr hole, and both brain and rectal temperatures were maintained at 3760.5 8C by use of a 2.5. Glucose microdialysis treatment group heating pad and lamp.
Artificial CSF was prepared as above with either 0 or 30 2.2. Controlled cortical impact mM glucose. Animals were divided into two groups (n5 10-15 per group): CCI with 0 mM glucose CSF, and CCI Rats underwent lateral cortical contusion injury (CCI) with 30 mM glucose CSF. Probes were inserted and using a vertical impactor as previously described [10]. A trauma performed exactly as described above.
2.6. 619C89 treatment group 3. Results Rats underwent the same microdialysis procedure as Figs. 1 and 2 and illustrate the effects of various described above. They were divided into two groups (n58 treatments upon dialysate glutamate concentrations in per group): vehicle control (double-distilled water), and hippocampus and cortex after CCI. Since the probe was 619C89 (30 mg / kg). This dose was the maximally effec-removed from the brain just prior to and reinserted just tive dose in previous temporary focal ischemia experi-after trauma, changes in dialysate glutamate were first ments. Furthermore, doses higher than 30 mg / kg were studied in traumatized rats and rats that were subjected to found to produce hypotension, which could confound the interpretation of these experiments [7]. The 30 mg / kg dose was found to afford behavioral protection in the rat fluid percussion model [22]. Vehicle or drug was administered i.v. over 1 min after the equilibration period, with drug concentration adjusted to yield a final injection volume of 1 ml. Five minutes after injection, CCI was performed and samples collected for 1 h post trauma.

HPLC
The concentration of glutamate was determined by HPLC with scanning fluorescence detection (Waters 470 scanning fluorescence detector, excitation wavelength 334 nm, emission wavelength 424 nm). 25 ml samples were diluted 1:1 with 0.01 M HCl, mixed with 80 ml ophthaldialdehyde-mercaptoethanol (Sigma) and injected into the column after 1 min (Waters 715 Ultra Wisp). A reversed-phase column and guard column (10034.6 mm I.D. and 1534.6 mm I.D., respectively; C Microsorb, 18 Varian, Sugarland, TX, USA), packed with 3 mm octadecylsilane particles, were contained in an axial compression unit (Dynamax, Rainin Instrument) at 30 8C. Mobile phase was comprised of 50 mM o-phosphoric acid, 50 mM EDTA, pH 5.7 with NaOH (all chemicals: Sigma). This mobile phase was mixed with methanol (Optima, Fisher Scientific, Pittsburgh, PA, USA) in a ratio of (90:10, v / v) for solvent A and (10:90, v / v) for solvent B. Samples were injected using a curvilinear gradient, changing the proportion of solvent A from 80% to 10% over 16 min, returning to 80% over 5 min. Peaks were identified by comparing retention times with those of standards, and quantified using Maxima 820 software (Dynamic Solutions, Ventura, CA, USA) after calibration to standard curves. Amino acid concentrations are expressed as the actual dialysate concentrations and were not corrected for recovery.

Statistical analysis
The a priori hypothesis was that the mean peak dialysate concentrations in samples obtained immediately after trauma would differ for treatment groups versus controls. Accordingly, this comparison was made with a t-test. The microdialysate glutamate concentrations in the sample immediately after probe reinsertion. There were also increases in the excitatory amino acids aspartate and glycine in the CCI group (Table 1).
Treatment with the use-dependent Na1 channel antagonist 619C89 produced a significant effect on microdialysis glutamate in cortex, but not in hippocampus (Figs. 1B and  2B). There was also a significant effect on dialysate aspartate in cortex. Paradoxically, there was an increase in dialysate glycine in the 619C89 treatment group ( Table 1). The magnitudes of these changes are less compared to the near total suppression of ischemia-induced changes in dialysate glutamate and other EAAs by 619C89 treatment [7].
There was no significant effect of altering the glucose concentration in the dialysate upon microdialysate glutamate in either cortex (Fig. 1C) or hippocampus (Fig. 2C). Paradoxically, there was an increase in dialysate aspartate in cortical samples obtained in the 30 mM glucose dialysate group, although there was no significant effect of altering dialysate glucose upon any dialysate amino acid measured in hippocampus (Table 1). Thus, in contrast to cerebral ischemia, dialysate glucose does not have a significant effect on interstitial EAA concentrations after CCI.

Discussion
The major findings of this study are: (1) increased interstitial concentrations of the EAA glutamate occur in cortex after CCI. Increased concentrations also occur in hippocampus, a site remote to the contusion but synaptically connected to cortex. (2) Treatment with the use-dependent glutamate inhibitor 619C89 decreased concentrations of dialysate glutamate in cortex, but not in hippocampus.
(3) Altering the concentration of microdialysate glucose produced no effect on [EAA ] in either region. I The CCI model of traumatic brain injury produces a circumscribed contusion in rat cortex; however, this injury produces additional effects at sites distant from the contusion. The hippocampus is one area where CCI results in death of rare pyramidal neurons in CA3 and additional loss synaptic activity in hippocampus. In this study, microdialysate glutamate and other neurotransmitter amino the sham operation, including probe removal and reinser-acids are increased in hippocampus, demonstrating that tion. There was a slight trend toward increased mi-CCI has effects at this distant site. crodialysis glutamate in the sample immediately after Ischemia-induced increases in [EAA ] can be amelior-I probe reinsertion in the sham-operated group, but these ated by supplying additional glucose to the probe surchanges were not significantly different from baseline rounds via 30 mM glucose in the dialysate. This treatment concentrations. However, in keeping with previous studies, almost completely blocked increases in [EAA ] after I the rats subjected to CCI showed a significant increase in complete ischemia induced by cardiac arrest [24]. Further- There is also evidence that glutamate released from continues at the normal rate even when mitochondrial synapses plays a role in hippocampal pathology after CCI. respiration is blocked by fluoroacetate and other such Lesioning CA3 with kainate prior to CCI prevented the agents [23,25]. Thus, the failure of glial reuptake due to trauma-induced increases in glucose utilization [26]. Treatdeprivation of glucose supplies may be a major factor ments with drugs that prevent the increase of [EAA ] by I accounting for glial reuptake in ischemia. inhibiting Na1 channels also have neuroprotective effects In the CCI model of TBI, however, there was no in TBI models. Treatment with 619C89 has also been significant effect of 30 mM glucose in the dialysate upon shown to improve behavioral outcome and reduce neuronal the EAA glutamate in either cortex or hippocampus. loss in hippocampus after fluid percussion injury in rats Although there are decreases in cerebral blood flow after [22]. In the current study, treatment with the use-dependent trauma, these decreases do not reach the threshold for Na1 channel inhibitor 619C89 did significantly reduce significant ischemia, except in the center of the contusion peak glutamate microdialysate concentrations in cortex, itself [10]. Therefore, one would not expect a lack of but not in hippocampus. These results suggest that depolarglucose to be the primary factor in this TBI model. While ization of neurons after TBI is not the major determinant the degree of blood flow deprivation is not as severe as in of increased [EAA ]  then reinserted was used to control for these potential able by microdialysis during status epilepticus [21]. Thus,  mimics some aspects of human head injury but is a less