Enhanced neuronal damage by co-administration of quinolinic acid and free radicals, and protection by adenosine A_2A receptor antagonists

Intrahippocampal injections

Male Wistar rats weighing between 200 and 250 g were used. All animals were housed singly and provided with free access to food and water. Animals were anaesthetised with chloral hydrate (400 mg kg⁻¹, intraperitoneally) and placed in a stereotaxic frame. The scalp was incised and a burr hole made through the skull to permit access of the injection needle into the hippocampus at the desired co-ordinates (anteroposterior: 3.0 mm behind the bregma suture, dorsoventral: 2.8 mm below the cortical surface and lateral: 3.0 mm from the midline suture; Paxinos & Watson, 1986). The needle was then lowered to these co-ordinates and left in place for 2 min before the injection of test agents. The compounds used in this study were all introduced through a 29-gauge needle, injections being made in a volume of between 1 and 2 μl at a constant rate of 0.3 ml min⁻¹ using a Sage infusion pump (Jones et al., 1998a, 1998b). The injection needle was allowed to remain in place for 2 min after ending the injection so as to prevent leakage of drug along the needle track. The scalp was then sutured and the animals left to recover for 7 days.

Quinolinic acid was injected at doses of 40, 80 and 120 nmols. A mixture of xanthine (100 μM) and xanthine oxidase (0.1 U ml⁻¹) was co-administered by mixing solutions of the individual agents with the stock quinolinic acid solution in the appropriate proportions. The dose of xanthine and xanthine oxidase was limited by the solubility of the compounds, so that higher dosage was not practicable. Quinolinic acid was dissolved in 1 N NaOH and then diluted with 165 mM NaCl solution. The pH of the solution was then adjusted using 1 N HCl to between 7 and 7.6 before making up to volume by the addition of further saline. S-nitroso-N-acetylpenicillamine (SNAP) was injected at doses of 100 and 500 nmols.

Tissue fixing and slicing

Seven days after recovery from the intrahippocampal injections, rats were killed by an overdose of sodium pentobarbitone. The chest was opened to expose the heart and 10 ml of 0.9% sodium chloride solution was injected via a 26-gauge needle inserted into the left cardiac ventricle to wash blood from the cerebral vessels. This was immediately followed by 4% formaldehyde in phosphate buffered saline. The brain was then removed and stored in fixative for up to 1 week. A coronal slice of brain, approximately 3 mm thick, was prepared to include the location of the injection track, which was normally apparent from the residual dimpling of the cortical surface produced by the needle penetration. The block of brain was dehydrated and impregnated with paraffin wax throughout before embedding in wax. Sections were cut 6 μm thick, mounted on slides and stained with cresyl fast violet.

Sections were subsequently examined under a light microscope and areas CA1 and CA3 examined for damage. The damage was quantified in the CA1 region by selecting three sections approximately 2000 – 2500 μm from the site of the needle track and taking the average number of intact, surviving neurones at a magnification of 100×. A comparable count was made of neurons in the contralateral, unaffected side of the hippocampus, and the number of intact cells on the damaged side (a mean of the three sections counted) was then expressed as a percentage of the control side. As an indication of the number of cells per field counted for analysis, the number counted in a series of 10 control brains was 282×14. In all cases, the damaged and control sides were examined in the same coronal sections. Four animals were used for each data point, except for the preliminary data with quinolinic acid, where n=3. In a separate preliminary series of five brains, it was confirmed that there was no significant difference between the number of neurons in a CA1 field of view from an animal which had received no treatment or surgery, and a field from the contralateral side of a test animal.

Data analysis

The results are presented as mean±s.e.mean. Analysis of variance (ANOVA) followed by the Bonferroni post-test for multiple comparisons was used to determine any statistical significance. Significance refers to results where P<0.05 was obtained.

Drugs

All compounds were obtained from Sigma Chemicals, with the exception of 7-(2-phenylethyl)-5-amino-2-(2-furyl)-pyrazolo-[4,3e]-1,2,4-triazolo[1,5c]pyrimidine (SCH58261) and 4-(2-[7-amino-2-{2-furyl}{1,2,4}triazolo{2,3-a}(1,3,5}triazin-5-yl-amino]ethyl)phenol (ZM241385), which were partly obtained as gifts from Dr E. Ongini, Schering-Plough Research Institute, and Dr S. Poucher, Zeneca Pharmaceuticals respectively. The latter compound was also purchased from Tocris Chemicals. 8-(3-chlorostyryl)caffeine (CSC) was obtained from Research Biochemicals, Inc. All compounds were dissolved initially in dimethyl-sulphoxide and subsequently diluted to the required concentration for injection. The final concentration of dimethyl-sulphoxide was never more than 0.1%, which in earlier pilot experiments did not influence neuronal viability.

Effects of quinolinic acid

Normal intact pyramidal neurones had a clearly rounded appearance with a clear nucleus and nucleolus (Figure 1). Quinolinic acid produced a dose-dependent effect in the CA1 and CA3 regions of the hippocampus with no obvious or significant damage being seen at the lower doses of 40 nmols (4.3% cell loss±2.4, n=3) and 80 nmols (14.1% loss±3.3, n=3) but a substantial and usually almost total destruction at the highest dose (91.1% loss±6.5, n=3). This relationship was consistent with our previous studies (Behan et al., 1999; Behan & Stone, 2000), in which a dose of 120 nmols was found to produce a substantial degree of damage against which neuroprotective agents could be assessed. The border between damaged areas of CA1 and adjacent undamaged areas was sharply defined, as noted by Foster & Schwarcz (1989). There was also an infiltration of the pyramidal cell layers and surrounding tissue by microglial cells after the higher dose. Of the damaged areas, the CA1 area was selected for the quantification of the damage and protection as described in Methods, and the intermediate 80 nmols dose of quinolinic acid was selected for some of the later combination studies with potentially neuroprotective agents.

Effects of free radical generators

At the dose employed in this study, which was largely determined by the limited solubility of the compounds used, the mixture of xanthine and xanthine oxidase produced no obvious neuronal damage when injected alone (Figure 2). Similarly, the nitric oxide donor S-nitroso-N-acetylpenicillamine (SNAP) produced no significant neuronal loss when administered alone into the hippocampus at either of the doses tested (100 and 500 nmols). The higher dose was therefore selected for further experiments (Figure 3).

Combined administration

The two lower doses of quinolinate were tested in combination with the xanthine/xanthine oxidase mixture. With the lowest dose damage was apparent as a loss of healthy cells in the CA1 area when the two agents were administered simultaneously, with a proportion of cells showing general shrinkage, with enlarged and disrupted but darkly-staining nuclei or nuclear remnants (Figure 1). The damage induced by the combinations of xanthine/xanthine oxidase with either dose of quinolinic acid was very significantly greater than that produced by either agent alone (Figure 2), the combination based on quinolinic acid at 40 nmols producing a combined loss of 35% of neurons, and the combination involving 80 nmols of quinolinic acid yielding a loss of 81% of CA1 neurones.

The combined administration of quinolinate and SNAP (500 nmols) also produced significant damage with either the 40 or 80 nmol doses of the excitotoxin (Figure 3), the combination with the higher dose producing a loss of 66% of neurons in the CA1 region.

Effect of 5,7-dichlorokynurenic acid

The substantial degree of neuronal loss resulting from the administration of quinolinic acid at the high dose of 120 nmols is indicated in Figure 4. In this series of animals, only 3.5% of neurons remained intact after 7 days. 5,7-dichlorokynurenic acid was administered at a dose of 1 nmol, which had been found in preliminary experiments to block the damage produced by 120 nmols of quinolinic acid. This activity was confirmed in the present series, in which 5,7-dichlorokynurenate produced a highly significant protection when combined with 120 nmols of quinolinic acid compared with the administration of quinolinic acid alone (Figure 4).

A combination of 80 nmols of quinolinic acid with the xanthine/xanthine oxidase free-radical generating mixture generated a loss of 69% of the CA1 neurons, a value not significantly different from that seen in the experimental group illustrated in Figure 2. Against this combination, however, 5,7-dichlorokynurenate was not able to afford a significant degree of protection, yielding only a tendency towards protection and a neurone count of 52% of that in control sections (Figure 4).

Protection by purines

Using the combination of 80 nmols quinolinic acid with xanthine/xanthine oxidase, the potentially protective effects of antagonists at adenosine receptors were also examined. Administration of the A_2A receptor antagonist ZM241385 at a dose of 2.5 pmols, which we have previously reported to protect against damage induced by kainic acid (Jones et al., 1998a, 1998b) produced a small change in the proportion of surviving neurons from 27%±5.4 to 40%±7.7 (n=3) which was not statistically significant. At a higher dose of 25 pmols, significant protection was seen compared with the control quinolinic acid and xanthine/xanthine oxidase mixture (Figure 5). Similarly, another non-xanthine compound SCH58261 had little effect at a dose of 5 pmols, increasing the proportion of surviving neurons from 22%±8.1 to 38%±6.4, whereas a higher dose of 50 pmols did produce a significant protection as illustrated in Figure 5. The xanthine-derived A_2A receptor antagonist chlorostyrylcaffeine (CSC) also protected against the damage produced by the toxic combination, at a dose of 100 pmols (Figure 5).

Exactly comparable results were obtained when the adenosine receptor ligands were examined against the combination of quinolinic acid and SNAP (Figure 6). Chlorostyrylcaffeine and the higher doses of ZM241385 and SCH58261 were able to protect against cell loss in the CA1 region induced by the administration of the 500 nmols dose of SNAP (Figure 6).

Neuroprotection by adenosine antagonists

Adenosine or its analogues can reduce excitotoxicity and ischaemic damage via A₁ receptors (von lubitz et al., 1988; Macgregor & Stone, 1993; Heron et al., 1994) or A_2A receptors (Sheardown & Knutsen, 1996; Jones et al., 1998a, 1998b). Evidence for neuroprotection by an A_2A receptor antagonist was first suggested by Gao & Phillis (1994). This was later confirmed and extended to other ischaemic models and excitotoxic damage, using a range of A_2A receptor antagonist ligands (Phillis, 1995; von lubitz et al., 1995; Ongini et al., 1997; Monopoli et al., 1998; Jones et al., 1998a, 1998b). The selective adenosine antagonist ZM241385, which is 80-fold selective for A_2A versus A_2B receptors, and 500 – 1000-fold selective for A_2A versus A₁ receptors (Palmer et al., 1995), was reported to protect the hippocampus against kainic acid to a similar degree as the agonist CGS21680 (Jones et al., 1998a, 1998b). In support, work using A_2A receptor knockout mice has revealed that ischaemic brain injury is decreased compared with normal mice (Bona et al., 1997; Chen et al., 1999).

In the present study three chemically distinct antagonists at A_2A receptors, including the xanthine CSC and non-xanthines ZM241385 and SCH58261, protected against quinolinic acid. The doses used were chosen to target A_2A receptors with some selectivity at the sections used for cell counting. The dilution factor of an agent injected in a volume of 1 μl but subsequently diluting into a sphere of radius 2500 μm (the distance selected for assessment, see Methods) is approximately 30. This would yield a concentration of around 4 μM for CSC, the K_i for which at A_2A receptors is 54 nM, while the K_i at rat A1 receptors is 28 μM (Jarvis & Williams, 1989; Jacobson et al., 1993). Similarly, ZM241385 is likely to reach a concentration of about 1 μM, compared with its K_i at A_2A receptors of approximately 1 nM, a K_i of 3 μM at A1 receptors and 100 μM at A3 receptors (Poucher et al., 1995). Both these compounds should, therefore, completely saturate A_2A receptors, while leaving A1 receptors relatively unaffected. SCH58261 was used at a relatively high concentration which is expected to reach around 2 μM at the site or analysis, compared with a K_i of 1 nM at striatal A_2A receptors and over 100 nM at A1 receptors (Cunha et al., 1996; Ongini et al., 1997). The high dose was chosen because of evidence that the hippocampal A_2A receptor is less sensitive to SCH58261 than the striatal receptors (Johansson et al., 1993; Johansson & Fredholm, 1995; Lindstrom et al., 1996). In all cases, however, blockade of A1 receptors should increase cell damage, rather than protect, and would tend to oppose, rather than explain, the present data.

The most likely mechanism by which blockade of A_2A receptors can produce protection is the suppression of glutamate release. Activation of A_2A receptors has an excitatory action on neurones, partly via an increased release of glutamate (Simpson et al., 1992; Sebastião & Ribeiro, 1992; Popoli et al., 1995). The blockade of A_2A receptors, therefore, may reduce the extracellular concentrations of glutamate below a threshold necessary for cell damage. In addition, since A_2A receptors suppress responses mediated by A₁ sites (Lopes et al., 1999; O'kane & Stone, 1998), A_2A receptor antagonists will ‘unmask' the A₁ receptors and thus permit a degree of protection by A₁ receptors.

There may also be effects of A_2A receptor antagonists on glial cell function. Activation of A_2A receptors suppresses phagocytic activity (Zalavary et al., 1994); antagonists should therefore facilitate the removal of dying cells and toxic materials from a region of injury. In addition, although A_2A receptors inhibit the production of several pro-inflammatory cytokines from cells (Dianzani et al., 1994), they can also potentiate the pro-inflammatory effects of those compounds (Scholz-Pedretti et al., 2001). The protective effect of A_2A receptor antagonists may therefore reflect a net reduction of pro-inflammatory activity in the damaged region. Activation of A_2A receptors can promote glial proliferation after brain injury (Rathbone et al., 1999; Hindley et al., 1994). Since these cells produce anti-inflammatory cytokines as well as pro-inflammatory compounds and kynurenines, the net balance of these may influence neuronal survival.

Whatever the mechanism, the ability of A_2A receptor antagonists to protect not only aginst excitotoxins, but also against combined excitotoxic/free radical injury suggests that this class of compounds could be of value in preventing neuronal damage resulting from a variety of insults.