l -Arginine Decreases Fluid-Percussion Injury-Induced Neuronal Nitrotyrosine Immunoreactivity in Rats

Abstract

Peroxynitrite is a powerful oxidant capable of nitrating phenolic moieties, such as tyrosine or tyrosine residues in proteins and increases after traumatic brain injury (TBI). First, we tested the hypothesis that TBI increases nitrotyrosine (NT) immunoreactivity in the brain by measuring the number of NT-immunoreactive neurons in the cerebral cortex and hippocampus of rats subjected to parasagittal fluid-percussion TBI. Second, we tested the hypothesis that treatment with l-arginine, a substrate for nitric oxide synthase, further increases NT immunoreactivity over TBI alone. Rats were anesthetized with isoflurane and subjected to TBI, sham TBI, or TBI followed by treatment with l-arginine (100 mg/kg). Twelve, 24, or 72 h after TBI, brains were harvested. Coronal sections (10 μm) were incubated overnight with rabbit polyclonal anti-NT antibody, rinsed, and incubated with a biotinylated secondary antibody. The antigen—antibody complex was visualized using a peroxidase-conjugated system with diaminobenzidine as the chromagen. The number of NT-positive cortical and hippocampal neurons increased significantly in both ipsilateral and contralateral hemispheres up to 72 h after TBI compared with the sham-injured group. Remarkably, treatment with l-arginine reduced the number of NT-positive neurons after TBI in both cortex and hippocampus. Our results indicate that l-arginine actually prevents TBI-induced increases in NT immunoreactivity.

Introduction

Traumatic brain injury (TBI) decreases cerebral blood flow (CBF) (Bouma et al, 1991; Marion et al, 1991). Although the exact mechanism of action is unknown, posttraumatic hypoperfusion is likely caused by a combination of trauma-induced release of vasoconstrictors and inhibition of cerebral vasodilators (DeWitt and Prough, 2003; Golding et al, 1999). Recent evidence suggests that oxygen free-radical-mediated inactivation of nitric oxide (NO), a powerful cerebral vasodilator (Furchgott and Zawadzki, 1980), plays an important role in this process (DeWitt et al, 1997; Cherian and Robertson, 2003). After head injury in humans and experimental animals (Cherian et al, 2000; Hlatky et al, 2002), brain tissue NO concentrations increase transiently, then significantly decrease. Traumatic brain injury produces superoxide anion radicals (Fabian et al, 1995) that can inactivate NO by converting it to peroxynitrite (ONOO⁻). Peroxynitrite is a powerful oxidant that, when protonated, can isomerize to trans-peroxynitrous acid, which in turn forms an excited state with reactive properties of the powerful oxidants nitrogen dioxide and the hydroxyl radical (Beckman et al, 1990).

l-Arginine, a substrate for nitric oxide synthase (NOS), contributes to the maintenance of normal CBF, but, under conditions of oxidative stress, may contribute to the production of ONOO⁻ by enhancing NO levels. We previously found that l-arginine (100 mg/kg) prevented hypoperfusion after fluid-percussion TBI (DeWitt et al, 1997), probably by increasing NO production, and l-arginine restored CBF and reduced neuronal damage after TBI (DeWitt et al, 1997; Cherian and Robertson, 2003). l-Arginine treatment actually protected against increases in intracranial pressure up to 2 h after TBI (Prough et al, 2006) as well as contributed to improved neuronal and behavioral outcomes up to 15 days after TBI (Sell et al, 2008). Although l-arginine restored CBF after TBI leading to improved neuronal and behavioral outcomes, l-arginine-mediated increases in NO are likely to lead to increased ONOO⁻ formation.

Peroxynitrite exposure can cause systemic and cerebral vascular dysfunction. In rat spinal cord, ONOO⁻ induced oxidation, tyrosine nitration (Bao et al, 2003), and lipid peroxidation (Liu et al, 2005). Thus, NO produced as a result of treatment with l-arginine may increase CBF after TBI but may also cause oxidation, nitration, and peroxidation in proteins and lipids, and may impair cerebral vascular reactivity by increasing the production of ONOO⁻.

Investigations of ONOO⁻ are complicated as the ion is short-lived. ONOO⁻ is capable of hydroxylation, nitration, and nitrosation reactions. Phenols react readily with ONOO⁻ and the most common phenol in biologic systems is tyrosine. When ONOO⁻ reacts with tyrosine residues in proteins, nitrotyrosine (NT) is produced; NT is therefore considered a biomarker of ONOO⁻ production. Nitrotyrosine can be stained and localized using specific antibodies (Viera et al, 1999). Immunohistochemical staining revealed the presence of increased numbers of NT-positive neurons after spinal cord injury in rats (Liu et al, 2000), controlled cortical impact injury in mice (Whalen et al, 1999), fluid-percussion injury in rats (Besson et al, 2003), and head injury in humans (Forster et al, 1999). Although these reports indicate that central nervous system injury is associated with increases in NT immunoreactivity, the time course of protein nitration has not been established for TBI.

To measure the temporal course of protein nitration and determine whether l-arginine treatment increases protein nitration after TBI, we compared the number of NT-immunoreactive neurons in the cerebral cortex and hippocampus of rats subjected to TBI, sham TBI, or TBI followed by treatment with l-arginine at 12, 24, and 72 h after injury.

Materials and methods

Animals

Sixty-five adult (3 to 4 months) male Sprague—Dawley rats (350 to 400 g) were obtained from Harlan (Houston, TX, USA), housed 2 per cage with food and water ad libitum, and maintained at a constant temperature (21°C to 23°C) and humidity (45% to 50%) with lights on from 0700 to 1900 hours. After surgical manipulations, animals were housed singly. All experiments were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (National Research Council, 1996) and approved by the Institutional Animal Care and Use Committee of the University of Texas Medical Branch.

Surgical Procedures

Rats were anesthetized with 4% isoflurane, intubated and mechanically ventilated (NEMI Scientific; New England Medical Instruments, Medway, MA, USA) with 1.5% to 2.0% isoflurane in an oxygen-to-room-air ratio of 70:30, and prepared for parasagittal fluid-percussion TBI, as described previously (DeWitt et al, 1997). Briefly, the skin over the dorsal surface of the skull was incised in the midline and reflected, and a needle temperature probe (Yellow Springs Instruments, Yellow Springs, OH, USA) was placed under the reflected temporalis muscle. A 3 mm craniotomy was trephined 2.5 mm to the right of the midline, midway between lambda and bregma, and the dura was left intact. A modified plastic Luer-lok needle hub was cemented into the craniotomy using cyanoacrylic and dental cement. After surgery, rats were randomly assigned to receive sham TBI, moderate (2.0 atm) TBI, or moderate TBI + l-arginine (100 mg/kg intravenously, diluted in 3 ml of 0.9% saline, 5 mins after trauma). Rectal temperature was monitored using a Physitemp Thermalert Model TH-8 (Clifton, NJ, USA) and maintained at 37°C using a thermostatically controlled water blanket (Gaymar, Orchard Park, NY, USA). The right common jugular vein was cannulated and a 3 ml bolus of 0.9% saline was administered to all rats (in the l-arginine group this volume was the vehicle for the drug). All animals were administered 120 mg/kg of acetaminophen for analgesia. Rats were randomly assigned to one of the following three treatment groups: vehicle-treated sham TBI surgery (sham; n = 24), vehicle-treated TBI surgery (TBI; n = 20), and l-arginine-treated TBI surgery (TBI + l-arginine; n = 21). Each treatment group was subdivided into three separate groups for tissue harvesting at different time points (12, 24, and 72 h after surgery; n = 6 to 10 per group). There were no deaths or unexpected complications because of surgical manipulations.

Immunohistochemistry

Rats were reanesthetized with 4% isoflurane and ketamine (60 to 70 mg/kg intraperitoneally) plus xylazine (5 to 10 mg/kg intraperitoneally) either 12, 24, or 72 h after TBI or sham injury, and transcardially perfused with saline followed by 4% paraformaldehyde. The brains were removed, dehydrated, embedded in paraffin, and sectioned. Coronal sections (10 μm) were collected, deparaffinized with xylene and graded alcohols, placed in an antigen retrieval solution, and then immersed in 3% H₂O₂ in methanol (to block endogenous peroxidase activity) and goat-blocking serum (to block nonspecific binding).

The sections were then incubated with rabbit polyclonal anti-NT antibody (1:100 dilution; Upstate Biotechnology, Lake Placid, NY, USA) for 45 mins at 27°C and then overnight at 4°C, rinsed, and incubated with a biotinylated goat anti-rabbit IgG (1:200 dilution; Vectastain Elite ABC kit (Universal); Vector Laboratories, Burlingame, CA, USA) for 30 mins. The antigen—antibody complex was detected with a peroxidase-conjugated system (Vectastain Elite ABC kit (Universal); Vector Laboratories), developed with diaminobenzidine, and counterstained with hematoxylin. Additional sections were treated identically but were not incubated with the primary antibody (negative control).

Nitrotyrosine-immunoreactive neurons in sections on coded slides were counted in preselected regions of the hemispheres and hippocampal cell layers CA1 and CA3 and the polymorphic layer of the dentate gyrus (Figure 1) bilaterally by an investigator unaware of the treatment groups, and expressed as a percentage of total neurons in high-power microscopic fields. Neurons were identified by their large size and large, prominent euchromatic nuclei. Two slices were counted per animal, one each at two different levels of bregma (−3.3 and −5.2 mm). In the cortex, five zones were counted in each section. In the hippocampus, three zones were counted at bregma −3.3 (CA1, CA3, and dentate gyrus; see Figure 1).

Figure 1

Coronal sections of rat brain are shown at −3.3 mm from bregma (A) and −5.2 mm from bregma (B). Counts of NT-positive neurons were performed on designated areas of the ipsilateral (to trauma site) hemisphere and the equivalent areas on the contralateral side. Original sections adapted from Paxinos G and Watson C (1986), The brain in stereotaxic coordinates. 2nd ed. CA, USA: Academic Press; with permission.

Statistical Analysis

Group descriptive statistics are expressed as mean ± s.e. of the mean and differences between groups were evaluated using analysis of variance (Table 1). The neuronal counting data were not normally distributed and are therefore expressed as quartiles. The effects of treatment group and time after injury were evaluated using a nonparametric two-way analysis of variance followed by Tukey's procedure using a two-sided alpha level of significance (P < 0.05). Comparisons between ipsilateral and contralateral hemispheres were evaluated using the Wilcoxon Signed Rank Test, with P < 0.05 considered significant. Neuronal counting statistical analyses were conducted using SAS (SAS Institute Inc (2004), SAS/STAT 9.1 User's Guide, Cary, NC, USA).

Table 1

Group descriptive statistics

	Sham	TBI	TBI+l-arginine
Weight (g)	480 ± 33	499 ± 34	489 ± 41
Trauma level (atm)		2.0 ± 0.4	2.0 ± 0.3

TBI, traumatic brain injury.

There were no significant differences in body weight among groups (sham versus TBI, P = 0.12; sham versus TBI + l-arginine, P = 0.41; TBI versus TBI + l-arginine, P = 0.43) and no significant differences in TBI level between the two injured groups (TBI versus TBI + l-arginine, P = 0.29).

Results

There were no significant differences in body weight among the three groups and no significant differences in TBI level between the two injured groups. Also, there were no deaths or unexpected complications because of surgery (Table 1). Nitrotyrosine staining was not observed in the negative control slides that were not incubated with the primary antibody. Nitrotyrosine immunoreactivity was observed in the cytoplasm (perikarya and processes) of neurons and occasionally in neuroglia. Only cells identified as neurons were counted.

Cerebral Cortex

Although there appeared to be fewer NT-positive neurons in the contralateral (relative to the injury site) compared with the ipsilateral cortex, there were no significant differences in the number of NT-positive cells between the ipsilateral and contralateral cortical hemispheres in the sham (P = 0.930), TBI (P = 0.221), or l-arginine-treated TBI groups (P = 0.650).

In the ipsilateral cortex, a main effect of treatment (F_2,63 = 12.06; P < 0.0001) was observed, whereas a main effect of time was not quite significant (F_2,63 = 2.91; P = 0.062). Post hoc analysis revealed that the TBI group showed increased NT immuno-reactivity compared with the sham-operated group. Furthermore, animals that received l-arginine treatment after TBI showed significantly lower levels of NT immunoreactivity compared with vehicle-treated animals that received TBI (Figures 2A and 3).

Figure 2

Percentages of NT-immunoreactive neurons in the ipsilateral (A) or contralateral (B) cerebral cortex of rats subjected to sham TBI (sham; n = 7, 7, and 10 at 12, 24, and 72 h after surgery, respectively), fluid-percussion TBI (TBI; n = 6, 7, and 7 at 12, 24, and 72 h after injury, respectively), or TBI followed by treatment with 100 mg/kg l-arginine (TBI + l-arg; n = 7, 6, and 8 at 12, 24, and 72 h after injury, respectively). Data are presented as median (solid line), first and third quartiles (boundaries of box), and range (error bars). The mean is also shown (dashed line). *P < 0.05 TBI versus sham, ^P < 0.05 TBI + l-arg versus TBI (comparisons are between groups collapsed across time).

Figure 3

Examples of NT immunoreactivity taken from the ipsilateral cortex (relative to trauma site) are shown: sham TBI surgery (left column: A, D, and G), moderate (2.0 atm) fluid-percussion TBI (middle column: B, E, and H), and TBI followed by treatment with 100 mg/kg l-arginine (right column: C, F, and I) at 12 h (top row: A—C), 24 h (middle row: D—F), and 72 h (bottom row: G—I).

In the contralateral cortex, there was a main effect of treatment group (F_2,63 = 6.51; P = 0.003) but not time (F_2,63 = 2.34; P = 0.106). Post hoc analysis revealed effects similar to the ipsilateral cortex. The TBI group showed significantly higher percentages of NT-immunoreactive neurons compared with the sham group, and the l-arginine-treated group showed significantly lower percentages of NT-immunoreactive neurons compared with the TBI group (Figure 2B).

Hippocampus

Comparison between ipsilateral and contralateral regions showed no differences between the number of NT-positive cells in the sham (P = 0.051), vehicle-treated TBI (P = 0.332), or l-arginine-treated TBI groups (P = 0.463).

In the ipsilateral hippocampus, there was no main effect of treatment (F_2,64 = 1.85; P = 0.166) or time (F_2,64 = 0.83; P = 0.441). However, in the contralateral hippocampus, a significant main effect of treatment (F_2,64 = 3.82; P = 0.028) but not time (F_2,64 = 1.16; P = 0.321) was observed. Post hoc analysis revealed that animals treated with l-arginine after TBI showed reduced NT immunoreactivity compared with animals treated with vehicle after TBI (Figure 4).

Figure 4

Percentages of NT-immunoreactive neurons in the ipsilateral (A) or contralateral (B) hippocampus of rats subjected to sham TBI (sham; n = 7, 7, and 10 at 12, 24, and 72 h after surgery, respectively), fluid-percussion TBI (TBI; n = 6, 7, and 7 at 12, 24, and 72 h after injury, respectively), or TBI followed by treatment with 100 mg/kg l-arginine (TBI + l-arg; n = 7, 6, and 8 at 12, 24, and 72 h after injury). Data are presented as median (solid line), first and third quartiles (boundaries of box), and range (error bars). The mean is also shown (dashed line). ^P < 0.05 TBI + l-arg versus TBI (comparisons are between groups collapsed across time).

Discussion

This is the first quantitative report of the time course of changes in NT immunoreactivity after fluid-percussion TBI in rats, and the first demonstration that treatment with l-arginine actually prevents TBI-induced increase in the number of NT-immunoreactive neurons in the rat brain after injury. Moderate parasagittal fluid-percussion TBI resulted in increases in NT immunoreactivity in both ipsilateral and contralateral cerebral cortices and hippocampi. Nitrotyrosine immunoreactivity was increased earlier in the cortex (12 h) than in the hippocampus (24 h), and remained elevated up to 72 h after injury. In sham-operated rats, the percentage of NT-immunoreactive neurons was generally lower than in TBI and was not time-dependent. Treatment with l-arginine 5 mins after TBI kept NT immunoreactivity at approximately the same level as that observed in the sham group in the cortex and hippocampus. As NT is generally considered as an indicator of ONOO⁻ production, our results show that l-arginine treatment actually reduced cortical and hippocampal production of ONOO⁻ after TBI.

Our results are consistent with previous reports in regard to NT formation after traumatic injury in the central nervous system. In the brain, fluid-percussion injury in rats produced qualitative increases in NT immunoreactivity in the corpus callosum 30 mins and 2 h after injury, as well as in the temporoparietal cortex 4 h to 72 h after injury (Besson et al, 2003). Controlled cortical impact TBI followed by hyperoxic ventilation in rats also resulted in significant increases in the number of NT-positive neurons in hippocampal layers CA1 and CA3, 24 h after injury (Ahn et al, 2005). In mice, time-dependent increases in NT staining (Hall et al, 2004; Mesenge et al, 1998) or expression (Bayir et al, 2005) have also been observed after TBI. Furthermore, in human brains, autopsy revealed the presence of NT-immunoreactive neurons after TBI (Forster et al, 1999). Likewise, in the spinal cord, traumatic injury increased NT protein expression (Xu et al, 2001), the number of NT-immunoreactive neurons, and extracellular NT levels (Liu et al, 2000). Clearly, NT expression increases after different types of neuronal injury in rodents as well as humans.

After fluid-percussion TBI in rats, constitutive NOS activity increased significantly at the lesion site (Wada et al, 1998) and direct measurements of changes in brain tissue NO levels revealed significant, transient NO increases after controlled cortical impact (Cherian et al, 2000) or parasagittal fluid-percussion TBI in rats (Ahn et al, 2004). In a closed head injury model, the nonspecific NOS inhibitor l-nitroarginine methyl ester reduced NT immunoreactivity 4 and 24 h after TBI (Mesenge et al, 1998), results that are consistent with trauma-induced increases in NO contributing to NT formation. Nitric oxide reacts with superoxide to form ONOO⁻(Beckman et al, 1990). Normally, superoxide is scavenged by superoxide dismutase and higher concentrations of superoxide dismutase and superoxide, compared with NO, favor the reaction between them. However, when NO levels increase, the faster rate of the reaction between NO and superoxide drives the formation of ONOO⁻ (Beckman et al, 1990). After fluid-percussion TBI, levels of superoxide and other oxygen free radicals increase significantly (Fabian et al, 1995; Wei et al, 1996), possibly as a byproduct of trauma-induced increases in prostaglandin synthesis (Ellis et al, 1981). Catalysis of the metabolism of arachidonic acids by cyclooxygenase-2, which also increases after TBI (Dash et al, 2000), likely contributes to increased formation of the superoxide anion radical (Pickard, 1981).

l-Arginine, a substrate for NOS enzymes, contributes to the maintenance of normal CBF, but under conditions of oxidative stress, may also increase the production of ONOO⁻ by enhancing NO levels. Thus, an excess of l-arginine in an environment where there is already an overproduction of superoxide anion radicals (such as after traumatic injury) could potentially increase NO concentrations and result in increased ONOO⁻levels. However, our results indicate that l-arginine treatment reduces rather than increases cortical and hippocampal NT staining after TBI. Although cerebral ischemia and reperfusion were associated with ONOO⁻ formation (Beckman, 1991; Kumura et al, 1996) and l-arginine improved CBF after TBI (DeWitt et al, 1997), it is unlikely that l-arginine reduced NT staining through its effects on CBF. This is because after moderate fluid-percussion TBI in rats, CBF reductions generally do not reach ischemic levels (Yamakami and Mcintosh, 1991). However, CBF may be moderately reduced, so as not to result in ischemia, but to a susceptible level where an increase in flow could produce long-lasting effects.

Alternatively, the reduction in the number of NT-positive neurons in the cortex and hippocampus with l-arginine treatment after TBI may be because of increased production of NO leading to increased antioxidant activity of this radical (Kagan and Laskin, 2001; O'Donnell et al, 1997). For instance, Bayir et al (2005) observed increased consumption of the antioxidant ascorbate in the brain of inducible NOS (iNOS)-knockout mice compared with wild type after closed cortical impact TBI, suggesting that iNOS-derived NO was acting as an antioxidant after TBI in wild-type mice. Although the antioxidant attribute of NO has been primarily to scavenge lipid peroxyl (O'Donnell et al, 1997; Kagan and Laskin, 2001) and hydroxyl (Rauhala et al, 1996) rather than the superoxide radicals that contribute to ONOO⁻ formation, the preservation of other endogenous antioxidants, such as ascorbate, by increased antioxidant activity of NO because of l-arginine treatment could lead to a reduction in superoxide levels, resulting in decreased formation of ONOO⁻.

Other explanations for l-arginine preventing increases in NT after TBI are based on the idea that l-arginine depletion can initiate superoxide generation by iNOS, which can then be inhibited by l-arginine (Xia et al, 1998). For example, neuronal NOS and iNOS can generate superoxide and NO in the absence of l-arginine (Xia et al, 1998). Although l-arginine levels in rodent brain tissue and micro-dialysate were not reduced by TBI (Cherian et al, 2000), intracellular compartmentalization may result in low l-arginine levels in certain subcellular compartments (Morikawa et al, 1994). Other potentially favorable effects of l-arginine treatment include an osmolar effect that decreased intracranial pressure (Cherian and Robertson, 2003) or a hydroxyl radical-scavenging effect (a property shared with l- and d-nitroarginine methyl ester; Rehman et al, 1997), which reduced NT staining after treatment with l-arginine after TBI.

Although it may seem unlikely that a single treatment with l-arginine 5 mins after TBI reduces NT immunoreactivity for up to 72 h after injury, a variety of potential explanations exist. Considering that the half-life of l-arginine after intravenous administration in rats is relatively short (0.76 h, peak 20 to 30 mins, return to baseline 4 to 5 h; Wu et al, 2007), clearly the effect of this single treatment outlasts its presence in the circulation. Long-lasting effects after single treatments may be the result of changes in gene expression, activation of specific enzymes or metabolic pathways, protective effects of antioxidants, or interactions between the aforementioned. Nutritional supplementation studies suggest that effective physiologic responses can be maintained in the absence of sustained increases in circulating arginine concentrations (Mateo et al, 2007). One potential mechanism is through changes in gene expression, such as arginine regulation of iNOS protein translation (Lee et al, 2003). In addition, TBI studies suggest that alternative metabolic pathways are activated after injury. For example, Vespa et al (2007) reported elevated lactate/pyruvate ratios in the absence of ischemia in TBI patients, and Bartnik et al (2005, 2007) observed that fluid-percussion TBI was associated with significant increases in glucose metabolism through the pentose phosphate pathway, which would result in increased NADPH synthesis.

This shift in cerebral metabolic activity toward energy production through alternative pathways after TBI may be associated with changes in gene expression and/or influenced by antioxidant activity. For example, reduced levels of NAD⁺ and ATP after TBI may be because of the activation of the gene for poly(ADP-ribose) polymerase-1 (PARP-1) an enzyme that repairs DNA strand breaks. It has been shown that TBI results in the activation of PARP-1 (LaPlaca et al, 1999; Ha and Snyder, 2000). However, activation of PARP-1 may deplete NAD⁺, decrease the rate of electron transport and ATP formation, and eventually result in energy failure and cell death (Endres et al, 1997). Poly(ADP-ribose) polymerase-1 can be activated by superoxide anion radicals and ONOO⁻ (Ha and Snyder, 2000), both of which are produced immediately after TBI (Fabian et al, 1995; Kontos and Wei, 1986). It is possible that NO produced by l-arginine treatment immediately after TBI acted as an antioxidant and reduced oxidant-induced DNA strand breaks such that PARP-1 activation was avoided. Besson et al (2003) reported that PARP-1 activation was associated with increased NT immunoreactivity for up to 72 h after TBI and that the inhibition of PARP-1 reduced NT staining. Thus, a single treatment with l-arginine may have been sufficient to prevent PARP-1 activation and thereby avoid persistent generation of peroxynitrite.

The detection of NT-immunoreactive neurons in both ipsilateral and contralateral hemispheres (relative to the injury site) is because of the bilateral nature of parasagittal fluid-percussion TBI. Using parasagittal (injury site 2 to 3 mm from midline) fluid-percussion injury, others and we have shown bilateral consequences of injury. Neuronal counting at 15 days after injury showed bilateral decreases in neuronal survival, whereas treatment with hypertonic l-arginine (100 mg/kg) produced bilateral increases in neuronal survival (Sell et al, 2008). Others have observed bilateral reductions in CBF (Yamakami and Mcintosh, 1991) and bilateral increases in blood—brain barrier permeability in the hippocampus and septal nuclei (Schmidt and Grady, 1993). Furthermore, Smith et al (1997) reported bilateral ventriculomegaly and a significant loss of dentate neurons in the contralateral hippocampus in rats 2 weeks to 1 year after parasagittal fluid-percussion TBI (Smith et al, 1997). Thus, our results are consistent with other evidence of bilateral cerebral injury after parasagittal fluid-percussion TBI.

The accuracy of our findings may be influenced by two general categories of limitations. First, limitations related to the accuracy of neuronal counting and second, limitations related to the specificity and accuracy of NT as a marker for ONOO⁻. Neuronal counting may be influenced by morphologic damage at the injury site, making the identification of neurons by morphologic criteria difficult and possibly resulting in an underestimation of the number of NT-positive neurons. Definitive differentiation of NT-positive neuronal from nonneuronal cells requires the use of double-labeling with specific neuronal antibodies (e.g., Neu-N), which was not performed in this study. Secondly, NT immunoreactivity may not be highly specific to the effects of ONOO⁻. Rather, NT immunoreactivity is likely a marker of ONOO⁻-related tyrosine nitration in vivo (Beckman et al, 1990). The formation of NT detected with myeloperoxidase (Eiserich et al, 1998) and eosinophil peroxidase (Duguet et al, 2001) suggests that tyrosine nitration is an indicator of reactive nitrogen species in general rather than ONOO⁻ in particular.

The enzyme myeloperoxidase is secreted by activated neutrophils, and uses peroxide and chloride ions to produce hypochlorous acid (HOCl), an antibacterial agent. Under conditions when there is concomitant ONOO⁻ and HOCl formation (e.g., sites of inflammation), HOCl may remove NT (Whiteman and Halliwell, 1999). Traumatic brain injury produces such sites of inflammation. In humans, TBI has been shown to induce a strong, locally restricted inflammatory response in which activated neutrophils infiltrate the site of tissue destruction (Bank et al, 1999). Therefore, the production of HOCl is possible in this scenario, making this mechanism another potential source of inaccuracy in our study. However, although multiple reactive pathways have been shown to effectively increase NT residues in vitro, only few have been replicated in vivo (Ischiropoulos, 1998). Interestingly, nitryl chloride, a potential nitration agent formed from the reaction between nitrite and neutrophil myeloperoxidase-derived HOCl, contributed minimally to the formation of intracellular NT in vivo (Whiteman et al, 2003), confirming that the NT immunoreactivity we detected originated from ONOO⁻.

In summary, we showed that parasagittal fluid-percussion TBI in rats resulted in time-dependent, significant increases in the number of NT-immunoreactive neurons bilaterally in the cerebral cortex and hippocampus. Post-injury treatment with l-arginine prevented the TBI-induced increase in the number of NT-positive neurons at all three time points and in both brain regions. Future studies will target the potential mechanisms by which l-arginine, a substrate for NOS, protects against tyrosine nitration after traumatic neuronal injury.