Introduction

Top Abstract Introduction Materials and Methods Results Discussion Conclusions References

The importance of endogenous norepinephrine (NE) as an anticonvulsant neurotransmitter is well established. For example, specific lesions of the central noradrenergic system using 6-hydroxydopamine (Corcoran et al., 1974) or N-(2-chloroethyl)-N-2-bromobenzylamine (Carre and Harley, 1986) result in increased seizure sensitivity. Conversely, stimulation of the locus coeruleus, the primary central noradrenergic nucleus, significantly inhibits seizures (Weiss et al., 1990). More recently, antiepileptic therapies used clinically have been shown to either increase central NE content or require an intact noradrenergic system for their efficacy (Waller and Buterbaugh, 1985; Baf et al., 1994; Krahl et al., 1998; P. Szot, D. Weinshenker, J. M. Rho, T. W. Storey, and P. A. Schwartzkroin, unpublished data). Despite the long-standing interest and experimentation along these lines, it is still unclear exactly how NE inhibits seizure activity. There are three primary reasons for this confusion.

First, the noradrenergic signaling system is extremely complex; three distinct classes of receptor with multiple subtypes have been cloned, and each class activates different G proteins. The ₁ class of adrenoreceptors (ARs) is linked to G_o/G_q and activates phospholipase C and intracellular Ca²⁺ release. ₂ARs are linked to G_i and inhibit adenylate cyclase. ARs (₁AR and ₂AR) are linked to G_s and activate adenylate cyclase (for review, see Cooper et al., 1996). All of these receptor classes are widely distributed throughout the brain and are found in regions implicated in regulating seizures such as the hippocampus, cortex, and amygdala. Differences in receptor distribution between animal species, strain, and the brain region(s) affected by seizure-inducing paradigms probably contribute to the variety of ARs shown to possess anticonvulsant activity (Papanicolaou et al., 1982; Lints and Nyquist-Battie, 1985; Neuman, 1986; Löscher and Czuczwar, 1987; Ferraro et al., 1994; Yan et al., 1998). Second, despite this complexity, many seizure susceptibility studies have not been pharmacologically comprehensive for any given seizure paradigm. For example, data have been presented exclusively on nonselective AR antagonists (Lints and Nyquist-Battie, 1985), nonselective AR agonists and antagonists (Ferraro et al., 1994), and AR agonists and antagonists (Löscher and Czuczwar, 1987; Tsuda et al., 1990). In only a few studies have compounds that target both ARs and ARs been tested (Neuman, 1986; Gellman et al., 1987; Micheletti et al., 1987), and even these were somewhat limited in their scope.

Third, although the ₂AR receptor is probably the most promiscuous anticonvulsant receptor in terms of efficacy across species, strain, and seizure paradigm (Papanicolaou et al., 1982; Baran et al., 1985; Scotti de Carolis et al., 1986; Löscher and Czuczwar, 1987; Jackson et al., 1991), proconvulsant effects have also been reported (Oishi et al., 1979; Löscher and Czuczwar, 1987; Wu et al., 1987). In addition, there are inherent difficulties interpreting any of these results. Three different subtypes of ₂AR exist (₂A, ₂B, and ₂C) and they are localized both pre- and postsynaptically. Activation of presynaptic ₂AR autoreceptors decreases noradrenergic firing and NE release (L'Heureux et al., 1986; Jorm and Stamford, 1993), while activation of postsynaptic ₂ARs mimics the effect of released NE on target cells expressing these receptors and inhibits their firing (Gobert et al., 1998). Because available ₂AR agonists cannot distinguish between ₂AR subtypes or pre- versus postsynaptic ₂ARs, it is nearly impossible to interpret the effect of ₂AR agonists. In addition, there is some evidence that activation of presynaptic ARs facilitate central NE release (Murugaiah and O'Donnell (1995).

To address these issues, we have tested the effects of subtype-selective agonists and antagonists for most known ARs on susceptibility to pentylenetetrazole (PTZ)-induced seizures in dopamine -hydroxylase knockout (Dbh /) mice that are deficient in NE synthesis and lack NE, thereby isolating the postsynaptic effects of these compounds and eliminating any influence on NE release. Dbh / mice have increased susceptibility to PTZ-induced seizures, demonstrating that endogenous NE is anticonvulsant in this model of epilepsy (Szot et al., 1999). Both anticonvulsant (Löscher and Czuczwar, 1987; Amabeoku et al., 1994) and proconvulsant (Oishi et al., 1979; Fletcher and Forster, 1988) effects of clonidine on PTZ-induced seizures in mice have been reported. We predicted that the effects of clonidine would be abolished in Dbh / mice if it acts via presynaptic ₂ARs, but Dbh / mice would still be affected if the postsynaptic ₂ARs are involved.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Conclusions References

Animals. Dbh knockout (Dbh /) mice, maintained on a 129/SvEv and C57BL/6J hybrid background, were developed and generated as described (Thomas et al., 1998). Mice were reared in a specific pathogen-free facility with a 12-h light/dark cycle at the University of Washington (Seattle, WA), but moved to a conventional facility for seizure experiments. Mice between 3 and 6 months of age were used for all experiments, and food and water were available ad libitum.

/

/

+/

+/

Seizure Induction. Seizures were induced with an i.p. injection of PTZ dissolved in water. Because Dbh / mice are more sensitive to PTZ-induced seizures (Fig. 1; Szot et al., 1999), doses of PTZ were adjusted based on genotype in many experiments to elicit seizures of similar severity in Dbh +/ and Dbh / mice. PTZ was administered at a dose of 25, 30, or 40 mg/kg (6.25, 7.5, or 10 mg/ml) for Dbh / mice and 40 or 50 mg/kg (10 or 12.5 mg/ml) for Dbh +/ mice in a volume of 4 ml/kg. The lower doses of PTZ for each genotype, which produced mild seizures, were used when testing a drug that had a proconvulsant effect based on pilot experiments, while the higher doses produced severe seizures and were used when testing anticonvulsant compounds. Mice were placed in a clear Plexiglas chamber and closely observed for 10 min. This observation time was chosen because mice that displayed seizure activity did so within the first few minutes after PTZ administration and were typically postictal by 10 min. Latency to first myoclonic jerk (MJ) and clonic/tonic (C/T) seizure was recorded. Animals that did not experience one of these behavioral seizure landmarks were assigned a latency of 600 s for that measurement.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Susceptibility of Dbh +/ and Dbh / mice to PTZ-induced seizures. Shown are latencies to first MJ and C/T seizure. These data are a compilation of results with water and vehicle-injected mice followed by administration of 40 mg/kg PTZ from Figs. 2 through 4 (Dbh +/, n = 49; Dbh /, n = 67). ****P < 0.0001 compared with Dbh +/.

	Results

Top Abstract Introduction Materials and Methods Results Discussion Conclusions References

Dbh / Mice Are Hypersensitive to PTZ-Induced Seizures. We previously reported that Dbh / mice have shorter latencies to first MJ and C/T seizure than Dbh +/ mice following PTZ administration (Szot et al., 1999). We repeated this experiment with a larger number of mice and obtained the same result (Fig. 1), demonstrating that a lack of endogenous NE results in increased sensitivity to PTZ-induced seizures.

₂AR Agonists Act via Autoreceptors to Exacerbate PTZ-Induced Seizures. To determine whether ₂AR agonists increased or decreased sensitivity to PTZ-induced seizures, we treated Dbh +/ mice with the ₂AR agonist clonidine prior to PTZ administration. Clonidine (0.1 mg/kg) had a proconvulsant effect that was most pronounced for latency to MJ (Fig. 2A). To determine whether this proconvulsant effect was mediated by an inhibition of NE release via presynaptic autoreceptors or via actions on postsynaptic heteroreceptors, we tested the effect of clonidine on PTZ-induced seizures in Dbh / mice. Clonidine did not alter seizure latency in Dbh / mice in response to 30 mg/kg PTZ (Fig. 2A) or 25 mg/kg PTZ (latency to MJ: vehicle = 317 ± 73 s, 0.1 mg/kg clonidine = 265 ± 29 s, P = 0.85; no C/T seizures observed), suggesting that the proconvulsant effect seen in control mice resulted from decreased NE release mediated by activation of presynaptic ₂AR autoreceptors. Clonidine treatment caused piloerection in both genotypes and rescued the bilateral ptosis phenotype of Dbh / mice, demonstrating that the drug possessed biological activity at this dose in the knockouts (data not shown). Blockade of ₂ARs with yohimbine (10 mg/kg) slightly increased seizure latency in both genotypes, but the effect was not significant (Fig. 2B). The weak anticonvulsant effect of yohimbine may have been nonspecific and caused by the significant sedation observed with this treatment. The ₂AR antagonist idazoxan (3 mg/kg), which lacks the sedative effects of yohimbine, also did not have a significant effect (data not shown). These results suggest that while exogenous activation of ₂ARs can exacerbate PTZ-induced seizures, activation of ₂ARs by endogenous NE is not involved under these conditions.

View larger version (44K): [in this window] [in a new window]   Fig. 2.   Effects of 2AR-acting drugs on PTZ-induced seizure susceptibility. Shown are latencies to first MJ and C/T seizure. A, water (Dbh +/, n = 13; Dbh /, n = 14) or clonidine (0.1 mg/kg: Dbh +/, n = 14; Dbh /, n = 14) was administered 30 min prior to PTZ (40 mg/kg for Dbh +/, 30 mg/kg for Dbh /). *P < 0.05 compared with water control. B, water (Dbh +/, n = 12; Dbh /, n = 8) or yohimbine (10 mg/kg: Dbh +/, n = 12; Dbh /, n = 8) was administered 30 min prior to PTZ (50 mg/kg for Dbh +/, 40 mg/kg for Dbh /).

Activation of Central ₁ARs Inhibits PTZ-Induced Seizures in Dbh / Mice. While clonidine is a fairly selective ₂AR agonist, it has some activity at ₁ARs at high doses (Wu et al., 1987). We considered the possibility that our results with clonidine were due to concurrent activation of ₁ARs, but rejected this hypothesis based on the results of two experiments. First, decreasing the dose of clonidine to 1 µg/kg failed to reveal an anticonvulsant effect (data not shown). This dose was chosen because it is maximally effective at inhibiting PTZ-induced seizures in normal rats and electrically induced seizures in rats with 6-hydroxydopamine lesions of their noradrenergic systems (Dalton et al., 1985). Second, cirazoline (0.2 mg/kg), a selective ₁AR agonist, significantly increased latencies to MJ and C/T seizures in Dbh / mice (Fig. 3A). While eight of nine vehicle-treated Dbh / mice progressed to a C/T seizure at this dose of PTZ, only 2 of 10 cirazoline-treated Dbh / mice had seizures of this severity. Similar effects were obtained with 0.1 and 0.5 mg/kg cirazoline (data not shown). This anticonvulsant effect was blocked by pretreatment with prazosin (1 mg/kg), a selective ₁AR antagonist, further verifying the importance of ₁ARs (Fig. 3B). While cirazoline did not have a significant anticonvulsant effect on Dbh +/ controls (Fig. 3A), prazosin alone (1 mg/kg) had a proconvulsant effect (Fig. 3C), reducing latencies to MJ and C/T seizures. These data suggest that the inhibition of seizures by endogenous NE in normal animals involves ₁ARs.

View larger version (31K): [in this window] [in a new window]   Fig. 3.   Effects of 1AR-acting drugs on PTZ-induced seizure susceptibility. Shown are latencies to first MJ and C/T seizure. A, water (n = 8 for each genotype) or cirazoline (0.2 mg/kg: Dbh +/, n = 8; Dbh /, n = 9) was administered 30 min prior to PTZ (50 mg/kg for Dbh +/, 40 mg/kg for Dbh /). *P < 0.05 compared with water control; **P < 0.01 compared with water control. B, vehicle (1.5% DMSO, 1.5% Cremaphor EL in water) or prazosin (1 mg/kg) was administered 30 min prior to water or cirazoline (0.2 mg/kg; n = 10 for each drug treatment group), which was administered to Dbh / mice 30 min prior to PTZ (40 mg/kg). *P < 0.05 compared with water control. ***P < 0.001 compared with water control. P < 0.01 compared with cirazoline alone. C, vehicle (1.5% DMSO, 1.5% Cremaphor EL in water) or prazosin (1 mg/kg; n = 12 for each drug treatment group) was administered to Dbh +/ mice 30 min prior to PTZ (40 mg/kg). *P < 0.05 compared with vehicle control; **P < 0.01 compared with vehicle control. D, water (n = 6) or phenylephrine (3 mg/kg; n = 7) was administered to Dbh / mice 30 min prior to PTZ (40 mg/kg).

/

/

₂AR Signaling Inhibits PTZ-Induced Seizures. To investigate the contribution of AR signaling to the anticonvulsant effect of NE, we treated mice with isoproterenol, a nonselective AR agonist. We found that isoproterenol (10 mg/kg) increased latencies to PTZ-induced seizures in both Dbh / and Dbh +/ mice (Fig. 4A). To determine which AR subtype was responsible for the anticonvulsant effect of isoproterenol, mice were treated with a preferential ₁AR (dobutamine) or a preferential ₂AR (albuterol) agonist prior to PTZ administration. While dobutamine (10 mg/kg) did not significantly affect seizure susceptibility in Dbh / mice, albuterol (10 mg/kg) prolonged the latency to both MJ and C/T seizures (Fig. 4B). Nearly all water-treated Dbh / mice (22 of 23) progressed to C/T seizures, while only 6 of 22 albuterol-treated Dbh / mice had seizures of that severity (data from Fig. 4, B and C, combined). Albuterol also significantly protected Dbh +/ control mice (Fig. 4B). Propanolol (10 mg/kg), a nonselective AR antagonist, blocked the anticonvulsant effect of albuterol in terms of latency to MJ but not C/T seizure in Dbh / mice (Fig. 4C). To determine whether the inability of propanolol to completely block the anticonvulsant effect of albuterol was due to blockade of ₁ARs as well as ₂ARs, we tested the ₂AR-selective antagonist ICI-118,551. ICI-118,551 (10 mg/kg) abolished the anticonvulsant effect of albuterol on MJ, but only partially blocked its effect on C/T seizures (Fig. 4C). Nadolol (10 mg/kg), a nonselective AR antagonist that cannot cross the blood-brain barrier, had effects similar to ICI-118,551 (Fig. 4C). Last, neither propanolol (10, 20, or 40 mg/kg) nor ICI-118,551 affected susceptibility to PTZ-induced seizures in Dbh +/ control mice (Fig. 4D; data not shown). These results suggest that ₂ARs do not mediate the anticonvulsant effect of endogenous NE, but that exogenous activation of ₂ARs can inhibit PTZ-induced seizures in mice.

View larger version (40K): [in this window] [in a new window]   Fig. 4.   Effects of AR-acting drugs on PTZ-induced seizure susceptibility. Shown are latencies to first MJ and C/T seizure. A, water (Dbh +/, n = 10; Dbh /, n = 9) or isoproterenol (10 mg/kg: Dbh +/, n = 10; Dbh /, n = 9) was administered 30 min prior to PTZ (50 mg/kg for Dbh +/, 40 mg/kg for Dbh /). *P < 0.05 compared with water control. **P < 0.01 compared with water control. ****P < 0.0001 compared with water control. B, water (Dbh +/, n = 9; Dbh /, n = 13), albuterol (10 mg/kg: Dbh +/, n = 9; Dbh /, n = 5), or dobutamine (10 mg/kg: Dbh /, n = 8) was administered 30 min prior to PTZ (50 mg/kg for Dbh +/, 40 mg/kg for Dbh /). **P < 0.01 compared with water control. ***P < 0.001 compared with water control. C, water, propanolol (10 mg/kg), ICI-118,551 (10 mg/kg), or nadolol (10 mg/kg) was administered 30 min prior to water or albuterol (10 mg/kg), which was administered to Dbh / mice (water + water, n = 18; water + albuterol, n = 17; propanolol + albuterol, n = 9; ICI-118,551 + albuterol, n = 12; nadolol + albuterol, n = 16) 30 min prior to PTZ (40 mg/kg). *P < 0.05 compared with water control. **P < 0.01 compared with water control. ***P < 0.001 compared with water control. P < 0.05 compared with albuterol alone. P < 0.01 compared with albuterol alone. D, water, propanolol (10 mg/kg), or ICI-118,551 (10 mg/kg) was administered to Dbh +/ mice (n = 10 for each drug treatment group) 30 min prior to PTZ (40 mg/kg).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Conclusions References

The results of this study are summarized in Table 1 and reveal three important points. First, stimulation of ₂ARs has a proconvulsant effect on PTZ-induced seizures. This effect is probably mediated by the inhibition of NE release via presynaptic ₂ autoreceptors, because the proconvulsant effect is abolished in Dbh / mice. Second, stimulation of either ₁ARs or ₂ ARs restores normal PTZ-induced seizure susceptibility upon Dbh / mice. Third, ₁AR but not ₂AR antagonists confer increased seizure susceptibility upon control mice, suggesting that endogenous NE inhibits seizures by activating ₁ARs in animals with normal NE content.

View this table: [in this window] [in a new window]   TABLE 1 Summary of AR agonist and antagonist effects on PTZ-induced seizure susceptibility Dbh / mice are more susceptible to PTZ-induced seizures than Dbh +/ controls, and some AR agonists and antagonists affect susceptibility differently in the two genotypes.  represents an increase in latency,  represents a decrease in latency, and  represents no change in latency to PTZ-induced seizure compared with the no drug control for each genotype.

Role of ₂ARs in PTZ Seizures. The literature indicates that ₂ARs are the most likely receptor type to mediate the anticonvulsant effect of endogenous NE to PTZ-induced seizures in rats. Multiple independent investigators have found that ₂AR agonists inhibit PTZ-induced seizures (Papanicolaou et al., 1982; Scotti de Carolis et al., 1986; Löscher and Czuczwar, 1987), and that ₂AR antagonists block the anticonvulsant effect of ₂AR agonists and potentiate PTZ-induced seizures on their own (Lazarova and Samanin, 1983; Scotti de Carolis et al., 1986). Because endogenous NE is anticonvulsant, the interpretation of these results was that the anticonvulsant effect of ₂AR agonists was mediated by postsynaptic receptors. However, one of the earliest publications on this theme reported a proconvulsant effect of clonidine on PTZ-induced seizures in mice, and proposed that activation of presynaptic ₂ARs decreased NE release and exacerbated seizures (Oishi et al., 1979). Since then, both proconvulsant (Fletcher and Forster, 1988) and anticonvulsant (Löscher and Czuczwar, 1987; Amabeoku et al., 1994) effects of clonidine in mice have been reported. We undertook this study using Dbh / mice that lack endogenous NE to determine whether clonidine is pro- or anticonvulsant and whether these effects are mediated by pre- or postsynaptic ₂ARs. We found that clonidine has a proconvulsant effect on control mice but no effect on Dbh / mice, suggesting that in our paradigm clonidine decreases NE release via presynaptic ₂ARs in normal animals.

Role of ₁ARs in PTZ Seizures. There are three results that support our conclusion that ₁AR activation mediates the anticonvulsant effects of endogenous NE in PTZ-induced seizures. First, Dbh / mice that lack endogenous NE are hypersensitive to PTZ-induced seizures, and administration of an ₁AR-selective agonist can restore normal sensitivity. Second, inhibition of PTZ-induced seizures by cirazoline can be blocked by prazosin, a selective ₁AR antagonist. Third, prazosin administration in control mice results in increased PTZ sensitivity. The inability of cirazoline to significantly protect Dbh +/ mice may mean that ₁ARs are already maximally activated in normal animals during seizures. Although ₁ARs have not previously been implicated in PTZ-induced seizures (Löscher and Czuczwar, 1987), ₁ agonists have anticonvulsant activity in many other seizure models, including amygdala kindling (Löscher and Czuczwar, 1987), air blast stimulation (Löscher and Czuczwar, 1987), focal cortical penicillin (Neuman, 1986), audiogenic seizures (Yan et al., 1998), quinolinic acid (Wu et al., 1987), and spontaneous petit mal seizures (Micheletti et al., 1987). Further evidence that ₁ARs mediate the anticonvulsant effect of endogenous NE in some situations comes from studies showing changes in ₁AR density in seizure-sensitive rats (Nicoletti et al., 1986; Gundlach et al., 1995), mice (Jazrawi and Horton, 1986), and humans with epilepsy (Brière et al., 1986).

Role of ₂ARs in PTZ Seizures. Our results indicate that activation of ₂ARs, but not ₁ARs, inhibits PTZ-induced seizures in both norepinephrine-deficient Dbh / mice and Dbh +/ mice, which have normal NE content. Because mice lacking NE were affected, we conclude that albuterol exerts its anticonvulsant action via postsynaptic ₂ARs and does not involve an increase in NE release via presynaptic ₂ARs. The lack of a proconvulsant effect of the AR antagonists propanolol or ICI-118,551 on Dbh +/ mice suggests that ₂ARs do not mediate the anticonvulsant effect of endogenous NE, even though ₂AR agonists can have an anticonvulsant affect on mice with normal NE content. In fact, it is unclear whether endogenously released NE typically activates ₂ARs in the brain. For example, ₂ARs have a much lower affinity than ₁ARs for NE (for review, see Molinoff, 1984). ₁AR but not ₂AR antagonists affect behavior (Pandey et al., 1995) and only ₁ARs undergo changes in response to treatments that chronically alter synaptic NE (Minneman et al., 1982). However, exogenously supplied ₂AR agonists affect behavior, ₂AR density (O'Donnell, 1990), and NE release (Murugaiah and O'Donnell, 1995), suggesting that ₂ARs are functional when artificially stimulated, as they appear to be in our study.

	Conclusions

Top Abstract Introduction Materials and Methods Results Discussion Conclusions References

This study comprehensively examines the anticonvulsant pharmacology of NE for a single convulsant agent, PTZ. The use of Dbh / mice, which lack NE, circumvents many of the caveats in the interpretation of results obtained in the presence of endogenous NE, and provides an example of how a combination of genetics and pharmacology can be a powerful approach in the study of neurotransmitter function.

It is important to note that each class of convulsant agent probably elicits seizures by distinct mechanisms that most likely involve different regions of the brain. Therefore, it is not surprising that there is not one "universal" noradrenergic receptor that mediates the anticonvulsant effect of NE for all seizure models. Because Dbh / mice are also more susceptible to seizures induced by kainic acid, flurothyl, and sound (Szot et al., 1999), it will be possible to apply the same approach to the identification of the anticonvulsant NE receptor(s) involved in other seizure models.