Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Handb Exp Pharmacol. Author manuscript; available in PMC 2012 January 1.
Published in final edited form as:
PMCID: PMC2945384

10 Methylxanthines, seizures and excitotoxicity


Clinical evidence, in particular the wide use of theophylline as bronchodilator, suggests that methylxanthines can cause seizures in patients without known underlying epilepsy. Theophylline is also known to be an added risk factor for seizure exacerbation in patients with epilepsy. The proconvulsant activity of methylxanthines can best be explained by antagonizing the brain’s own anticonvulsant adenosine. Recent evidence suggests that adenosine dysfunction is a pathological hallmark of epilepsy contributing to seizure generation and seizure spread. Conversely, adenosine augmentation therapies are effective in seizure suppression and prevention, whereas adenosine receptor antagonists such as methylxanthines generally exacerbate seizures. The impact of the methylxanthines caffeine and theophylline on seizures and excitotoxicity depends on timing, dose, and acute versus chronic use. New findings suggest a role of free radicals in theophylline-induced seizures and adenosine-independent mechanisms for seizure generation have been proposed.

A. Introduction

Seizures, ranging from altered states of consciousness to clonic and/or tonic convulsions, are commonly encountered in patients who do not have epilepsy (Delanty et al. 1998). Among other potential triggers, such non-epileptic seizures can be provoked by medication or medication withdrawal. Within this context, seizures are potentially severe or fatal complications of theophylline therapy. Theophylline can trigger seizures in patients without known underlying epilepsy and is an added risk factor for seizure exacerbation in patients with epilepsy. Most of these seizures result from toxic theophylline serum concentrations and are difficult to control. Nevertheless, clinical diagnosis and management of theophylline-induced seizures are underappreciated compared to other drug toxicities. Despite a long clinical history of theophylline-induced seizures, relatively little is known about the underlying molecular mechanisms that contribute to methylxanthine-induced seizure generation. Knowledge gained from patient data, but most notably from animal or in vitro studies aimed at elucidating the role of endogenous adenosine in seizure control contributes to our current understanding how methylxanthines influence the excitability of the brain.

B. Clinical findings

Anecdotally, caffeinated beverages are “known” to lower seizure thresholds in patients with epilepsy and the avoidance of excessive caffeine has been recommended in patients with epilepsy (Kaufman and Sachdeo 2003). However, due to the lack of well-designed, randomized, and placebo-controlled clinical trials, this concept has been challenged (Asadi-Pooya et al. 2008). Clinical findings in support of a proconvulsive role of methylxanthines are largely based on theophylline (or aminophylline, a mixture of theophylline with ethylenediamine that is 20 times more soluble than theophylline alone) which, clinically, is widely used to manage bronchospasms in reversible airway obstruction associated with stable asthma and chronic bronchitis (Barnes 2005; Van Dellen 1979). In addition, aminophylline is indicated in asystolic cardiac arrest and periarrest bradycardia refractory to atropine, whereas caffeine is used to treat diabetic cardiac autonomic neuropathy (Duby et al. 2004). Theophylline has a narrow therapeutic window with an optimal plasma concentration of 10–20 mg l−1 (55–110mmol l−1). Above this concentration, side effects such as arrhythmias and convulsions may occur, especially when given rapidly by intravenous injection (Nolan et al. 2005). Theophylline-associated seizures (TAS) or status epilepticus are considered a neurological emergency with potentially fatal outcome (Nakada et al. 1983). These seizures – largely focal onset generalized motor seizures – tend to be the only sign of theophylline toxicity, and can occur in neurologically intact patients (Aminoff and Simon 1980; Nakada et al. 1983). Remarkably, anticonvulsant therapy is ineffective in controlling these seizures, which often progress to status epilepticus and become intractable (Nakada et al. 1983; Yoshikawa 2007). Thus, in a recent clinical study the usual first-line treatment of diazepam was found to be more likely ineffective in TAS cases compared to non-TAS cases (Yoshikawa 2007); the failure of diazepam to stop those seizures might be based on interactions of theophylline with benzodiazepines (see below) (Yoshikawa 2007).

Interestingly, TAS is most common in pediatric patients under 5 years of age (Korematsu et al. 2008; Yoshikawa 2007), which can be considered to be naïve to theophylline or caffeine. In a recent study of eight pediatric TAS cases without underlying epilepsy, all had fever at the onset of TAS (> 38 °C), and six out of eight had a family history of febrile seizures and/or idiopathic epilepsy (Korematsu et al. 2008). The authors of this study concluded that in infants with an idiopathic reduced seizure threshold and fever, theophylline administration might possibly be sufficient to trigger a seizure. Apart from TAS discussed here, methylxanthine-induced seizures have also been described after the consumption of caffeinated energy drinks (Iyadurai and Chung 2007), and theophylline, caffeine, and aminophylline are used clinically to prolong seizure durations in electroconvulsive therapy for major depression (Stern et al. 1999). The potential risks associated with theophylline therapy are now well recognized. Due to concerns of CNS stimulant effects, theophylline use in patients with insomnia has been included in the 2002 Criteria for Potentially Inappropriate Medication Use in Older Adults (Fick et al. 2003).

Pharmacokinetic drug interactions of methylxanthines need also to be considered. Theophylline is largely metabolized by the hepatic enzyme CYP1A2, which is induced not only by a variety of antibiotics (Gillum et al. 1993) but also by the commonly used enzyme-inducing antiepileptic drugs, phenobarbital, phenytoin, carbamazepine, and primidone, and might require an increase in the therapeutic dose of theophylline (Patsalos et al. 2002; Spina et al. 1996). In view of the potential seizure-inducing effects of theophylline, the use of theophylline in patients with epilepsy is now limited despite the fact that second generation antiepileptic drugs do not interfere with the pharmacokinetics of theophylline (Patsalos et al. 2002). Of note, caffeine comedication in combination with phenobarbital during the first trimester of pregnancy leads to a significant increase in congenital malformations in offspring (Samren et al. 1999).

C. Experimental findings

The preconvulsive potential of methylxanthines has been corroborated in countless animal studies that reach back to more than 35 years (Roussinov et al. 1974). Early studies suggest slight differences in the convulsive role of methylxanthines: intraperitoneal administration of caffeine produced immediate excitation and seizures followed by an encephalopathy, whereas progression from encephalopathy to seizures was observed following aminophylline administration (Chu 1981). The proconvulsant and convulsant effects of methylxanthines are generally depend on dose and mode of application: Aminophylline at 100 mg/kg is known to increase the susceptibility of rats to pilocarpine or pentylenetetrazol induced seizures (Chakrabarti et al. 1997; Turski et al. 1989), whereas higher doses of aminophylline (250 mg/kg) lead to seizures and death in rats (Chakrabarti et al. 1997). These detrimental effects of high doses of aminophylline could be avoided by using equivalent doses of theophylline in preparations of acepifylline (theophylline ethanoate of piperazine) (Chakrabarti et al. 1997). Aminophylline-induced seizures directly depended on cerebrospinal fluid concentrations of theophylline and were not influenced by metabolites of theophylline (Ramzan and Levy 1986). In several experimental combinations it was shown that methylxanthines reduce or abolish the anticonvulsant activity of several antiepileptic drugs (Kulkarni et al. 1991). In contrast, the anticonvulsant effectiveness of felbamate was only affected at higher doses of aminophylline and caffeine (Gasior et al. 1998) and aminophylline did not alter the ability of gabapentin to protect mice against seizures induced by electroconvulsive shock (Luszczki et al. 2007). The concept that methylxanthines can exacerbate seizures in epilepsy has recently been challenged by Loscher, arguing that CNS stimulants exert (pro)convulsant activity only at supratherapeutic doses (Loscher 2009).

Whereas methylxanthine induced seizures are refractory to diazepam in patients, it is important to point out that levetiracetam and several other antiepileptic drugs that do not act via activation of GABAA receptors are highly effective in suppressing caffeine-induced seizures in mice (Klitgaard et al. 1998). Astemizole, a novel histamine H1 receptor antagonist, at a dose of 2 mg/kg increased the threshold for aminophylline-induced seizures (Swiader et al. 2005), an interesting observation since these drugs are usually combined during the treatment of asthma.

Pharmacokinetic and pharmacodynamic drug interactions have also been studied in animal models. Of note are interactions of the fluoroquinolone class of antibacterials with theophylline. In one study, chronic pretreatment of rats with the fluoroquinolone pefloxacine was shown to exacerbate aminophylline-induced seizures without altering brain concentrations of theophylline (Imperatore et al. 1997). Likewise certain environmental toxins, such as toluene, were shown to reduce thresholds for methylxanthine induced seizures (Chan and Chen 2003).

D. Adenosine, seizures, and excitotoxicity

Several potential mechanisms have been discussed that could explain the proconvulsive role of acute theophylline (Yoshikawa 2007): (i) general decrease of seizure thresholds; (ii) inhibition of adenosine A1 receptors (A1Rs) that normally suppress seizures by blocking the release of excitatory amino acids; (iii) inhibition of cerebral blood flow via adenosine antagonism (Puiroud et al. 1988); (iv) inhibition of 5’-nucleotidase and decrease in endogenous adenosine production; (v) inhibition of pyridoxal kinase, an enzyme needed for the synthesis of GABA; (vi) increase in cyclic GMP that is involved in maintaining the epileptic discharge; and (vii) a presumed direct inhibition of the GABAA receptor (Sugimoto et al. 2001), although interactions between GABAA receptors and the adenosine system might also be involved (Bonfiglio and Dasta 1991; Phillis 1979). Overall, it appears that theophylline does not trigger seizures as such, but rather potentiates pre-existing brain hyperexcitability, a mechanism consistent with the role of A1Rs in preventing seizure spread and in mediating seizure arrest (Fedele et al. 2006; Lado and Moshe 2008; Young and Dragunow 1994). Given the dominant role of the adenosine system in seizure control within the context of theophylline toxicity, the following sections focus on the role of adenosine in epilepsy.

1. Adenosine deficiency and seizure generation

The role of adenosine as an endogenous regulator of hippocampal excitability was first recognized by Dunwiddie almost 30 years ago (Dunwiddie 1980). In a subsequent study it was shown that theophylline and other alkylxanthines antagonized electrophysiological responses to adenosine and adenosine-stimulated cyclic AMP formation, indicating that alkylxanthines increase hippocampal excitability by antagonizing the actions of adenosine (Dunwiddie et al. 1981). Thus, several adenosine receptor agonists that activate the A1R were shown to suppress seizures in a variety of models, albeit accompanied by sedative and hypothermic side effects (Dunwiddie and Worth 1982). Endogenous adenosine is a potent regulator of hippocampal activity and was recently shown to control hippocampal sharp waves in CA3 via activation of A1Rs (Wu et al. 2009). It is now well recognized that adenosine is an endogenous anticonvulsant and regulator of brain activity (Boison 2005; Dunwiddie and Masino 2001; Fredholm et al. 2005a; Fredholm et al. 2005b; Ribeiro et al. 2002). The anticonvulsant activity of adenosine is largely mediated by activation of A1Rs, since A1R knockout mice experience spontaneous seizures (Li et al. 2007a) and are highly susceptible to seizure spread (Fedele et al. 2006). Conversely, A1R agonists are highly effective in the suppression of seizures (Benarroch 2008; Fredholm 2003; Jacobson and Gao 2006), and have been demonstrated to suppress seizures that are resistant to conventional antiepileptic drugs (Gouder et al. 2003). Decreased extracellular adenosine levels and reduced A1R activation as a consequence of kindling or caused by hypercapnia in a hippocampal slice preparation provide a plausible mechanisms for seizure generation (Dulla et al. 2005; Rebola et al. 2003).

In adult brain synaptic levels of adenosine are largely regulated by an astrocyte-based adenosine cycle (Boison 2008). Under physiological conditions synaptic adenosine is largely derived from vesicular release of ATP from astrocytes followed by extracellular cleavage into adenosine (Pascual et al. 2005), although astrocytic release of ATP via hemichannels has been demonstrated (Iglesias et al. 2009; Kang et al. 2008). In adult brain adenosine is rapidly phosphorylated into 5′-adenosine monophosphate by the astrocyte-based enzyme adenosine kinase (ADK; EC (Boison 2006, 2008). In contrast to conventional neurotransmitters, such as glutamate or glycine, there is no transporter-based regulatory mechanism to terminate the synaptic activity of adenosine. Due to the presence of two types of equilibrative nucleoside transporters in the astrocyte membrane (Baldwin et al. 2004), intracellular ADK is able to fulfill the role of a metabolic reuptake system for adenosine (Boison 2008). Based on its low KM for adenosine, ADK is the key regulator for ambient concentrations of adenosine (Boison 2006; Etherington et al. 2009; Lloyd and Fredholm 1995).

ADK has recently been identified as a molecular link between astrogliosis and neuronal hyperexcitability in epilepsy (Li et al. 2008). Astrogliosis – a pathological hallmark of the epileptic brain – is associated with upregulation of the adenosine removing enzyme ADK (Gouder et al. 2004; Li et al. 2008). Remarkably, the development of spontaneous electrographic seizures coincides both spatially (Li et al. 2008), as well as temporally (Li et al. 2007a) with astrogliosis and upregulated ADK. Uncoupling of astrogliosis from epileptogenesis in ADK-transgenic mice (Adk-tg) (Li et al. 2009) has demonstrated that overexpression of ADK, rather than astrogliosis per se, can be the cause for seizures. In line with these findings, Adk-tg mice express spontaneous recurrent electrographic seizures (Li et al. 2007a). Conversely, therapeutic augmentation of the adenosine system is very effective in suppressing seizures (Boison 2009). Together, these findings demonstrate that adenosine-deficiency and therefore deficient activation of A1Rs can be a direct cause for seizures. This conclusion supports the notion that methylxanthines have proconvulsant activity due to antagonizing the function of the endogenous anticonvulsant adenosine.

2. Adenosine deficiency and excitotoxicity

Adenosine, acting via A1Rs, is not only an endogenous anticonvulsant of the brain, but also a powerful neuroprotectant (Cunha 2005; Fredholm 1997). Thus, in addition to a proconvulsant role of A1R deficiency or increased adenosine clearance (overexpression of ADK), these conditions lead to increased vulnerability to excitotoxic injury. Consequently, A1R knockout mice are highly susceptible to seizure-induced (Fedele et al. 2006) or traumatic (Kochanek et al. 2006) brain injury and A1R knockout mice experience highly aggravated neuronal cell loss after status epilepticus (Li et al. 2007a).

Pharmacological studies in a model of oxygen glucose deprivation (OGD) suggest that whereas A1Rs desensitize after prolonged agonist exposure, A2AR mediated facilitation of glutamate release by endogenous adenosine remains fully operational under long-term OGD (Sperlagh et al. 2007). Thus, the inhibition of A2ARs might be a more effective approach to attenuate glutamatergic excitotoxicity than the stimulation of A1Rs (Cunha 2005). Consequently, A2AR antagonists are actively investigated clinically for their neuroprotective potential (Chase et al. 2003; Hauser et al. 2003).

3. Adenosine-based therapeutic approaches

Given the prominent role of adenosine as endogenous anticonvulsant and neuroprotectant, adenosine augmentation therapies (AATs) are highly effective in preventing seizures (Boison 2009). Pharmacologically, seizures can be suppressed by A1R agonists (Benarroch 2008) or by ADK inhibitors (McGaraughty et al. 2005), however systemic augmentation of the adenosine system is associated with significant side effects, including the suppression of cardiac function and depression of blood pressure, and therefore not a therapeutic option (Dunwiddie and Masino 2001). Alternatives are focal AATs to avoid systemic side effects and to restore adenosinergic signalling within a localized area of adenosine dysfunction, which can be equated with an epileptogenic focus (Li et al. 2008). Strategies that have been explored include the implantation of adenosine-releasing silk-based polymers into the infrahippocampal fissure in kindled rats. Rats treated with these polymers were protected both from established seizures, as well as from developing epilepsy (Szybala et al. 2009). Likewise, rats with focal implants of adenosine-releasing encapsulated fibroblasts or ADK-deficient stem cells were protected from kindled seizures or kindling development, respectively (Huber et al. 2001; Li et al. 2007b). Stem-cell derived adenosine-releasing implants that were placed into the infrahippocampal fissure in mice were shown to suppress acute chemoconvulsant induced seizures with associated injury (Ren et al. 2007), and to suppress epilepsy-development and spontaneous seizure expression in a model of CA3-selective focal epileptogenesis (Li et al. 2008). Together, these data demonstrate that focal re-constitution of adenosine signalling within an area of acquired adenosine dysfunction (i.e. within an epileptogenic focus) constitutes a powerful approach to suppress seizures.

E. Methylxanthines, seizures, and excitotoxicity

The above paragraphs suggest that methylxanthines – via antagonizing adenosine’s anticonvulsant and neuroprotective actions (Fredholm et al. 1999; Nehlig et al. 1992) – are proconvulsants that aggravate excitotoxicity. There are, however, additional interactions that need to be considered: the influence of methylxanthines on seizures and excitotoxicity is context- and receptor-dependent, and appears to be influenced by pathways not related to adenosine.

1. Acute versus chronic caffeine

Whereas the proconvulsive activity of acute methylxanthines has long been recognized (see above), the chronic dosing of caffeine has different effects. Thus, caffeine administered at a dose of 60–70 mg/kg per day in mice over a period of two weeks (resulting in plasma levels of caffeine in the range of 6 to 14 μM, corresponding to chronic caffeine use in humans) reduced N-methyl-D-aspartate (NMDA)-, bicuculline-, and pentylenetetrazol- induced seizures in mice in the absence of changes in A1R, A2ARs, or GABAARs (Georgiev et al. 1993; Johansson et al. 1996). The effect was due to the combined effects of theophylline, to which caffeine is metabolized in brain, and caffeine itself but could not be ascribed to changes in A1 and A2A adenosine or GABAA receptors (Johansson et al. 1996). In contrast, higher plasma concentrations of caffeine (100 μM) after chronic dosage for 12 days resulted in increased A1R densities, whereas mRNA levels or A2ARs were not affected (Johannson et al. 1993). Remarkably, chronic caffeine administration in rats (40 mg/kg, twice daily for seven days) increased the thresholds for subsequent theophylline-induced seizures (Zhi and Levy 1990). This phenomenon of effect inversion might be an explanation why children (which are considered to be caffeine-naïve) appear to be more sensitive to TAS-induced seizures (see above). Effect inversion of chronic adenosine receptor antagonists has also been described within the context of ischemic excitotoxicity (de Mendonca et al. 2000). Whereas acute methylxanthines generally aggravate ischemic injury, the chronic use of caffeine or of the A1R-selective antagonist DPCPX protects the brain from ischemic injury (de Mendonca et al. 2000). The phenomenon of effect inversion of acute versus chronic caffeine has intensively been studied and has been explained by antagonism of an endogenous agonist that downregulates A1Rs without affecting gene transcription (Jacobson et al. 1996). Evidence for effect inversion by caffeine or adenosine receptor ligands has been obtained through changes in physiological outcome parameters such as susceptibility to seizures or to seizure- and ischemia- induced neuronal cell death (Jacobson et al. 1996). Despite these clear physiological changes the molecular mechanisms behind this phenomenon appear to be more complex since upregulation of A1Rs as a consequence of chronic caffeine was not always observed (Georgiev et al. 1993; Johansson et al. 1996). Later studies have ruled out upregulation of A1Rs as a consequence of the long-term use of caffeine or theophylline in reasonably normal doses, indicating that upregulation of A1Rs is triggered only by excessively high or toxic doses of methylxanthines (Svenningsson et al. 1999). Thus, selected doses and durations of exposure and withdrawal, as well as A2AR mediated effects (see below) might play an important role.

In a recent study a single dose of acute caffeine (40 mg/kg i.p.) given after the onset of seizures in a new mouse model of sudden unexplained death in epilepsy (SUDEP) significantly increased the survival time from 24 to 55 minutes (Shen et al. 2009). This protective effect of acute caffeine can best be explained by antagonizing a seizure-induced surge of adenosine, which had experimentally been exacerbated by pharmacological disruption of adenosine clearance. In this model of SUDEP excessive seizure-induced concentrations of adenosine are thought to induce cardiac and respiratory failure by overstimulation of brainstem adenosine receptors, an effect that can be ameliorated by caffeine-induced blockade of these receptors (Shen et al. 2009).

2. Caffeine: A1 and A2A receptor-mediated actions

Whereas the anticonvulsant role of A1Rs is well established, newer findings suggest that A2ARs play an important role in modulating the susceptibility to seizures. Thus, A2AR knockout mice are partially resistant to limbic seizures induced by chomoconvulsants or to seizures induced by ethanol withdrawal (El Yacoubi et al. 2009; El Yacoubi et al. 2001). Interestingly, the attenuation of clonic pentylenetetrazole-induced seizures in A2AR knockout mice could be mimicked in wild-type mice exposed to chronic caffeine (0.3 g/l caffeine in drinking water) during a period of 14 days prior to the seizure tests (El Yacoubi et al. 2008). However, A2AR knockout mice under chronic caffeine were less protected from clonic seizures than water treated A2AR KO mice, a conflicting result that was not further addressed (El Yacoubi et al. 2008). Together, these findings indicate that the protective effects of chronic caffeine might best be explained by antagonizing the A2AR and thus causing a state of decreased neuronal excitability; however, these studies also indicate a proconvulsive role of chronic caffeine under conditions, during which A2AR-dependent signalling is abolished.

3. GABAA receptor and phosphodiesterase (PDE) inhibition

In contrast to adenosine receptors, which are affected by caffeine plasma concentrations attainable by normal human caffeine consumption, 10 to 100 times higher concentrations are needed to inhibit GABAARs or PDE (Fredholm et al. 1999). Therefore, a direct preconvulsive role of “physiological” doses of methylxanthines via GABAARs or PDE appears to be unlikely. However, caffeine can inhibit the binding of benzodiazepines to the GABAAR (Marangos et al. 1979), which might contribute to a convulsant role of high or toxic doses of methylxanthines. Inhibition of benzodiazepine binding to GABAARs might be an explanation for the clinical findings that TAS-induced seizures are usually refractory to treatment with diazepam or other drugs that act via the GABAAR (Yoshikawa 2007).

4. Ryanodine receptor activated calcium induced calcium release

Changes in Ca2+ homeostasis and persistent increases in intracellular Ca2+ contribute to the initiation and maintenance of acquired epilepsy (DeLorenzo et al. 2005). Ryanodine receptor (RyR) mediated calcium-induced calcium release (CICR) plays a key role in regulating intracellular calcium concentrations in epileptic conditions (Pal et al. 2001). The brain RyR is a caffeine-sensitive calcium release channel and mediates the caffeine induced mobilization of Ca2+ from internal stores (McPherson et al. 1991; Usachev et al. 1993). The caffeine-induced release of Ca2+ from ryanodine-sensitive calcium stores in the neuronal endoplasmic reticulum and pathological mechanisms that potentiate this response may render neurons more vulnerable to excitotoxicity and to the expression of seizures (Chan et al. 2000; Verkhratsky 2005). Interestingly, in cultered hippocampal neurons the newer antiepileptic drug levetiracetam led to a 61% decrease in caffeine-induced peak height of intracellular Ca2+ (Nagarkatti et al. 2008), indicating that levetiracetam might interact with adenosine-related signsalling.

5. Free radicals in theophylline-induced seizures

A possible role of free radicals in theophylline-induced seizures was recently suggested (Gulati et al. 2005; Gulati et al. 2007; Ray et al. 2005). In the underlying studies aminophylline (50 to 250 mg/kg) dose dependently induced convulsions and mortality in rats. Seizures and mortality were attenuated by anti-oxidants (melatonin, N-acetylcysteine) and by nitric oxide (NO) synthase inhibitors (L-NAME, 7-nitroindazole). Combination of anti-oxidant and NO-reducing treatments augmented the anticonvulsant effects of single treatments. Further, the authors found increased concentrations of malondialdehyde and NO metabolites in brain homogenates of mice with aminophylline-induced seizures; accumulation of these metabolites could be attenuated by melatonin or L-NAME pretreatment. These studies suggest the contribution of free radicals in the mechanism of theophylline induced ictogenesis.

6. Inhibition of TREK-1 channels by methylxanthines

TREK-1, a member of the two-pore-domain K(+) (K(2P)) channel superfamily, plays a major role in regulating the resting membrane potential of neurons, and thus contributes to controlling neuronal excitability (Honore 2007). Using whole-cell patch-clamp recordings on human TREK-1 channel expressing CHO cells, Harinath and Sikdar demonstrated reversible inhibition of the channels, and depolarization of the membrane potential, by caffeine and theophylline in a concentration-dependent manner (Harinath and Sikdar 2005). Inhibition by caffeine and theophylline was attenuated in channels with a mutation of a protein kinase A (PKA) consensus sequence indicating involvement of the cAMP/PKA pathway. Thus, inhibition of TREK-1 dependent membrane-depolarization may contribute to seizure generation by toxic doses of caffeine or theophylline.

F. Conclusions and outlook

Although adenosine-independent mechanisms have been proposed, the majority of evidence indicates that the proconvulsant roles of methylxanthines are based on antagonism of the brain’s endogenous adenosine-based seizure control system. Whereas inhibition of A1Rs by methylxanthines can directly contribute to ictogenesis and seizure spread, under certain conditions methylxanthines can also contribute to seizure suppression. First, this can be the case after chronic drug exposure leading effect inversion and alterations in gene expression (Svenningsson et al. 1999). Second, antagonism of A2ARs by methylxanthines may have direct anticonvulsant and neuroprotective consequences.

A detailed understanding of the convulsant role of methylxanthines is of importance since many new drugs are in clinical trials that act on adenosine receptors. For example, in recent clinical trials conducted with the A1R antagonist rolofylline, which facilitates diuresis and preserves renal function in patients with acute heart failure (AHF) with renal impairment, the occurrence of seizures was described in some patients that were treated with higher doses of the drug (Cotter et al. 2008). This example demonstrates that caution is needed when evaluating the clinical use of new adenosine-related therapeutic agents; however, understanding the mechanisms involved in the adenosine-related control of seizure mechanisms will allow the safe use of novel drugs that act on new therapeutic principles. New approaches using gene-array based strategies might unravel novel pathways and interactions that might help explain the complex role of methylxanthines in determining the brain’s susceptibility to seizures and excitotoxicity (Yu et al. 2009).


The work of the author is supported by grants R01NS058780, R01NS061844, R01MH083973, R21NS057475-01, and R21NS057538-01 from the National Institutes of Health (NIH), from Citizens United in Research against Epilepsy (CURE) in collaboration with the Department of Defense (DoD), and from the Legacy Hospital Foundation.

G. References

  • Aminoff MJ, Simon RP. Status epilepticus. Causes, clinical features and consequences in 98 patients. Am J Med. 1980;69:657–66. [PubMed]
  • Asadi-Pooya AA, Mintzer S, Sperling MR. Nutritional supplements, foods, and epilepsy: is there a relationship? Epilepsia. 2008;49:1819–27. [PubMed]
  • Baldwin SA, Beal PR, Yao SY, King AE, Cass CE, Young JD. The equilibrative nucleoside transporter family, SLC29. Pflugers Arch. 2004;447:735–43. [PubMed]
  • Barnes PJ. Targeting histone deacetylase 2 in chronic obstructive pulmonary disease treatment. Expert Opin Ther Targets. 2005;9:1111–21. [PubMed]
  • Benarroch EE. Adenosine and its receptors: multiple modulatory functions and potential therapeutic targets for neurologic disease. Neurology. 2008;70:231–6. [PubMed]
  • Boison D. Adenosine and epilepsy: from therapeutic rationale to new therapeutic strategies. Neuroscientist. 2005;11:25–36. [PubMed]
  • Boison D. Adenosine kinase, epilepsy and stroke: mechanisms and therapies. Trends Pharmacol Sci. 2006;27:652–658. [PubMed]
  • Boison D. The adenosine kinase hypothesis of epileptogenesis. Progress in Neurobiology. 2008;84:249–262. [PMC free article] [PubMed]
  • Boison D. Adenosine augmentation therapies (AATs) for epilepsy: prospect of cell and gene therapies. Epilepsy Res. 2009;85:131–141. [PMC free article] [PubMed]
  • Bonfiglio MF, Dasta JF. Clinical significance of the benzodiazepine-theophylline interaction. Pharmacotherapy. 1991;11:85–7. [PubMed]
  • Chakrabarti A, Saini HK, Garg SK. A comparative study of aminophylline- and acepifylline-induced seizures and death in the chemoconvulsion model in rats. J Pharm Pharmacol. 1997;49:812–5. [PubMed]
  • Chan MH, Chen HH. Toluene exposure increases aminophylline-induced seizure susceptibility in mice. Toxicol Appl Pharmacol. 2003;193:303–8. [PubMed]
  • Chan SL, Mayne M, Holden CP, Geiger JD, Mattson MP. Presenilin-1 mutations increase levels of ryanodine receptors and calcium release in PC 12 cells and cortical neurons. J Biol Chem. 2000;275:18195–200. [PubMed]
  • Chase TN, Bibbiani F, Bara-Jimenez W, Dimitrova T, Oh-Lee JD. Translating A2A antagonist KW6002 from animal models to parkinsonian patients. Neurology. 2003;61:S107–11. [PubMed]
  • Chu NS. Caffeine- and aminophylline-induced seizures. Epilepsia. 1981;22:85–94. [PubMed]
  • Cotter G, Dittrich HC, Weatherley BD, Bloomfield DM, O’Connor CM, Metra M, Massie BM. The PROTECT Pilot Study: A Randomized, Placebo-Controlled, Dose-Finding Study of the Adenosine A(1) Receptor Antagonist Rolofylline in Patients With Acute Heart Failure and Renal Impairment. Journal of Cardiac Failure. 2008;14:631–640. [PubMed]
  • Cunha RA. Neuroprotection by adenosine in the brain: from A1 receptor activation to A2A receptor blockade. Purinergic Signaling. 2005;1:111–134. [PMC free article] [PubMed]
  • de Mendonca A, Sebastiao AM, Ribeiro JA. Adenosine: does it have a neuroprotective role after all? Brain Res Brain Res Rev. 2000;33:258–274. [PubMed]
  • Delanty N, Vaughan CJ, French JA. Medical causes of seizures. Lancet. 1998;352:383–90. [PubMed]
  • DeLorenzo RJ, Sun DA, Deshpande LS. Cellular mechanisms underlying acquired epilepsy: The calcium hypothesis of the induction and maintainance of epilepsy. Pharmacology & Therapeutics. 2005;105:229–266. [PMC free article] [PubMed]
  • Duby JJ, Campbell RK, Setter SM, White JR, Rasmussen KA. Diabetic neuropathy: an intensive review. Am J Health Syst Pharm. 2004;61:160–73. quiz 175–6. [PubMed]
  • Dulla CG, Dobelis P, Pearson T, Frenguelli BG, Staley KJ, Masino SA. Adenosine and ATP link P-CO2 to cortical excitability via pH. Neuron. 2005;48:1011–1023. [PMC free article] [PubMed]
  • Dunwiddie TV. Endogenously released adenosine regulates excitability in the in vitro hippocampus. Epilepsia. 1980;21:541–8. [PubMed]
  • Dunwiddie TV, Hoffer BJ, Fredholm BB. Alkylxanthines elevate hippocampal excitability. Evidence for a role of endogenous adenosine. Naunyn Schmiedebergs Arch Pharmacol. 1981;316:326–30. [PubMed]
  • Dunwiddie TV, Masino SA. The role and regulation of adenosine in the central nervous system. Annu Rev Neurosci. 2001;24:31–55. [PubMed]
  • Dunwiddie TV, Worth T. Sedative and anticonvulsant effects of adenosine analogs in mouse and rat. J Pharmacol Exp Ther. 1982;220:70–76. [PubMed]
  • El Yacoubi M, Ledent C, Parmentier M, Costentin J, Vaugeois JM. Evidence for the involvement of the adenosine A(2A) receptor in the lowered susceptibility to pentylenetetrazol-induced seizures produced in mice by long-term treatment with caffeine. Neuropharmacology. 2008;55:35–40. [PubMed]
  • El Yacoubi M, Ledent C, Parmentier M, Costentin J, Vaugeois JM. Adenosine A2A receptor deficient mice are partially resistant to limbic seizures. Naunyn Schmiedebergs Arch Pharmacol. 2009;380:223–32. [PubMed]
  • El Yacoubi M, Ledent C, Parmentier M, Daoust M, Costentin J, Vaugeois J. Absence of the adenosine A(2A) receptor or its chronic blockade decrease ethanol withdrawal-induced seizures in mice. Neuropharmacology. 2001;40:424–32. [PubMed]
  • Etherington LA, Patterson GE, Meechan L, Boison D, Irving AJ, Dale N, Frenguelli B. Astrocytic adenosine kinase regulates basal synaptic adenosine levels and seizure activity but not activity-dependent adenosine release in the hippocampus. Neuropharmacology. 2009;56:429–437. [PubMed]
  • Fedele DE, Li T, Lan JQ, Fredholm BB, Boison D. Adenosine A1 receptors are crucial in keeping an epileptic focus localized. Exp Neurol. 2006;200:184–190. [PubMed]
  • Fick DM, Cooper JW, Wade WE, Waller JL, Maclean JR, Beers MH. Updating the Beers criteria for potentially inappropriate medication use in older adults: results of a US consensus panel of experts. Arch Intern Med. 2003;163:2716–24. [PubMed]
  • Fredholm BB. Adenosine and neuroprotection. Int Rev Neurobiol. 1997;40:259–80. [PubMed]
  • Fredholm BB. Adenosine receptors as targets for drug development. Drug News Perspect. 2003;16:283–9. [PubMed]
  • Fredholm BB, Battig K, Holmen J, Nehlig A, Zvartau EE. Actions of caffeine in the brain with special reference to factors that contribute to its widespread use. Pharmacol Rev. 1999;51:83–133. [PubMed]
  • Fredholm BB, Chen JF, Cunha RA, Svenningsson P, Vaugeois JM. Adenosine and brain function. Int Rev Neurobiol. 2005a;63:191–270. [PubMed]
  • Fredholm BB, Chen JF, Masino SA, Vaugeois JM. Actions of adenosine at its receptors in the CNS: Insights from knockouts and drugs. Annu Rev Pharmacol Toxicol. 2005b;45:385–412. [PubMed]
  • Gasior M, Swiader M, Przybylko M, Borowicz K, Turski WA, Kleinrok Z, Czuczwar SJ. Felbamate demonstrates low propensity for interaction with methylxanthines and Ca2+ channel modulators against experimental seizures in mice. Eur J Pharmacol. 1998;352:207–14. [PubMed]
  • Georgiev V, Johansson B, Fredholm BB. Long-term caffeine treatment leads to a decreased susceptibility to NMDA-induced clonic seizures in mice without changes in adenosine Al receptor number. Brain Res. 1993;612:271–7. [PubMed]
  • Gillum JG, Israel DS, Polk RE. Pharmacokinetic drug interactions with antimicrobial agents. Clin Pharmacokinet. 1993;25:450–82. [PubMed]
  • Gouder N, Fritschy JM, Boison D. Seizure suppression by adenosine A1 receptor activation in a mouse model of pharmacoresistant epilepsy. Epilepsia. 2003;44:877–885. [PubMed]
  • Gouder N, Scheurer L, Fritschy J-M, Boison D. Overexpression of adenosine kinase in epileptic hippocampus contributes to epileptogenesis. J Neurosci. 2004;24:692–701. [PubMed]
  • Gulati K, Ray A, Pal G, Vijayan VK. Possible role of free radicals in theophylline-induced seizures in mice. Pharmacol Biochem Behav. 2005;82:241–5. [PubMed]
  • Gulati K, Ray A, Vijayan VK. Free radicals and theophylline neurotoxicity: an experimental study. Cell Mol Biol (Noisy-le-grand) 2007;53:42–52. [PubMed]
  • Harinath S, Sikdar SK. Inhibition of human TREK-1 channels by caffeine and theophylline. Epilepsy Res. 2005;64:127–35. [PubMed]
  • Hauser RA, Hubble JP, Truong DD. Randomized trial of the adenosine A(2A) receptor antagonist istradefylline in advanced PD. Neurology. 2003;61:297–303. [PubMed]
  • Honore E. The neuronal background K2P channels: focus on TREK1. Nat Rev Neurosci. 2007;8:251–61. [PubMed]
  • Huber A, Padrun V, Deglon N, Aebischer P, Mohler H, Boison D. Grafts of adenosine-releasing cells suppress seizures in kindling epilepsy. Proc Natl Acad Sci USA. 2001;98:7611–6. [PubMed]
  • Iglesias R, Dahl G, Qiu F, Spray DC, Scemes E. Pannexin 1: the molecular substrate of astrocyte “hemichannels” J Neurosci. 2009;29:7092–7. [PMC free article] [PubMed]
  • Imperatore C, Trimarchi GR, De Sarro A. Interaction between pefloxacin and aminophylline in genetically epilepsy-prone rats. J Pharm Pharmacol. 1997;49:1025–9. [PubMed]
  • Iyadurai SJ, Chung SS. New-onset seizures in adults: possible association with consumption of popular energy drinks. Epilepsy Behav. 2007;10:504–8. [PubMed]
  • Jacobson KA, Gao ZG. Adenosine receptors as therapeutic targets. Nat Rev Drug Discov. 2006;5:247–64. [PMC free article] [PubMed]
  • Jacobson KA, von Lubitz DKJE, Daly JW, Fredholm BB. Adenosine receptor ligands: differences with acute versus chronic treatment. TiPS. 1996;17:108–113. [PMC free article] [PubMed]
  • Johannson B, Ahlberg S, van der Ploeg I, Brené S, Lindefors N, Persson H, Fredholm BB. Effect of long term caffeine treatment on A1 and A2 adenosine receptor binding and on mRNA levels in rat brain. Naunyn-Schmiedeberg’s Arch Pharmacol. 1993;347:407–414. [PubMed]
  • Johansson B, Georgiev V, Kuosmanen T, Fredholm BB. Long-term treatment with some methylxanthines decreases the susceptibility to bicuculline- and pentylenetetrazol-induced seizures in mice. Relationship to c-fos expression and receptor binding. Eur J Neurosci. 1996;8:2447–58. [PubMed]
  • Kang J, Kang N, Lovatt D, Torres A, Zhao Z, Lin J, Nedergaard M. Connexin 43 hemichannels are permeable to ATP. J Neurosci. 2008;28:4702–11. [PubMed]
  • Kaufman KR, Sachdeo RC. Caffeinated beverages and decreased seizure control. Seizure. 2003;12:519–21. [PubMed]
  • Klitgaard H, Matagne A, Gobert J, Wulfert E. Evidence for a unique profile of levetiracetam in rodent models of seizures and epilepsy. Eur J Pharmacol. 1998;353:191–206. [PubMed]
  • Kochanek PM, Vagni VA, Janesko KL, Washington CB, Cramrine PK, Garman RH, Jenkins LW, Clark RS, Homanics GE, Dixon CE, Schnermann J, Jackson EK. Adenosine A1 receptor knockout mice develop lethal status epilepticus after experimental traumatic brain injury. J Cereb Blood Flow Metab. 2006;26:565–575. [PubMed]
  • Korematsu S, Miyahara H, Nagakura T, Suenobu S, Izumi T. Theophylline-associated seizures and their clinical characterizations. Pediatr Int. 2008;50:95–8. [PubMed]
  • Kulkarni C, Joseph T, David J. Influence of adenosine receptor antagonists, aminophylline and caffeine, on seizure protective ability of antiepileptic drugs in rats. Indian J Exp Biol. 1991;29:751–4. [PubMed]
  • Lado FA, Moshe SL. How do seizures stop? Epilepsia. 2008;49:1651–1664. [PMC free article] [PubMed]
  • Li T, Lan JQ, Boison D. Uncoupling of astrogliosis from epileptogenesis in adenosine kinase (ADK) transgenic mice. Neuron Glia Biology. 2009 in press. [PMC free article] [PubMed]
  • Li T, Lan JQ, Fredholm BB, Simon RP, Boison D. Adenosine dysfunction in astrogliosis: cause for seizure generation? Neuron Glia Biology. 2007a;3:353–366. [PMC free article] [PubMed]
  • Li T, Ren G, Lusardi T, Wilz A, Lan JQ, Iwasato T, Itohara S, Simon RP, Boison D. Adenosine kinase is a target for the prediction and prevention of epileptogenesis in mice. J Clin Inv. 2008;118:571–582. [PMC free article] [PubMed]
  • Li T, Steinbeck JA, Lusardi T, Koch P, Lan JQ, Wilz A, Segschneider M, Simon RP, Brustle O, Boison D. Suppression of kindling epileptogenesis by adenosine releasing stem cell-derived brain implants. Brain. 2007b;130:1276–88. [PubMed]
  • Lloyd HG, Fredholm BB. Involvement of adenosine deaminase and adenosine kinase in regulating extracellular adenosine concentration in rat hippocampal slices. Neurochem Int. 1995;26:387–95. [PubMed]
  • Loscher W. Preclinical assessment of proconvulsant drug activity and its relevance for predicting adverse events in humans. Eur J Pharmacol. 2009;610:1–11. [PubMed]
  • Luszczki JJ, Jankiewicz K, Jankiewicz M, Czuczwar SJ. Influence of aminophylline on the anticonvulsive action of gabapentin in the mouse maximal electroshock seizure threshold model. J Neural Transm. 2007;114:1539–45. [PubMed]
  • Marangos PJ, Paul SM, Parma AM, Goodwin FK, Syapin P, Skolnick P. Purinergic inhibition of diazepam binding to rat brain (in vitro) Life Sci. 1979;24:851–7. [PubMed]
  • McGaraughty S, Cowart M, Jarvis MF, Berman RF. Anticonvulsant and antinociceptive actions of novel adenosine kinase inhibitors. Curr Top Med Chem. 2005;5:43–58. [PubMed]
  • McPherson PS, Kim YK, Valdivia H, Knudson CM, Takekura H, Franzini-Armstrong C, Coronado R, Campbell KP. The brain ryanodine receptor: a caffeine-sensitive calcium release channel. Neuron. 1991;7:17–25. [PubMed]
  • Nagarkatti N, Deshpande LS, DeLorenzo RJ. Levetiracetam inhibits both ryanodine and IP3 receptor activated calcium induced calcium release in hippocampal neurons in culture. Neurosci Lett. 2008;436:289–93. [PMC free article] [PubMed]
  • Nakada T, Kwee IL, Lerner AM, Remler MP. Theophylline-induced seizures: clinical and pathophysiologic aspects. West J Med. 1983;138:371–4. [PMC free article] [PubMed]
  • Nehlig A, Daval JL, Debry G. Caffeine and the central nervous system: mechanisms of action, biochemical, metabolic and psychostimulant effects. Brain Res Brain Res Rev. 1992;17:139–70. [PubMed]
  • Nolan JP, Deakin CD, Soar J, Bottiger BW, Smith G. European Resuscitation Council guidelines for resuscitation 2005. Section 4. Adult advanced life support. Resuscitation. 2005;67(Suppl 1):S39–86. [PubMed]
  • Pal S, Sun D, Limbrick D, Rafiq A, DeLorenzo RJ. Epileptogenesis induces long-term alterations in intracellular calcium release and sequestration mechanisms in the hippocampal neuronal culture model of epilepsy. Cell Calcium. 2001;30:285–96. [PubMed]
  • Pascual O, Casper KB, Kubera C, Zhang J, Revilla-Sanchez R, Sul JY, Takano H, Moss SJ, McCarthy K, Haydon PG. Astrocytic purinergic signaling coordinates synaptic networks. Science. 2005;310:113–6. [PubMed]
  • Patsalos PN, Froscher W, Pisani F, van Rijn CM. The importance of drug interactions in epilepsy therapy. Epilepsia. 2002;43:365–85. [PubMed]
  • Phillis JW. Diazepam potentiation of purinergic depression of central neurons. Can J Physiol Pharmacol. 1979;57:432–5. [PubMed]
  • Puiroud S, Pinard E, Seylaz J. Dynamic cerebral and systemic circulatory effects of adenosine, theophylline and dipyridamole. Brain Res. 1988;453:287–98. [PubMed]
  • Ramzan IM, Levy G. Kinetics of drug action in disease states. XVI. Pharmacodynamics of theophylline-induced seizures in rats. J Pharmacol Exp Ther. 1986;236:708–13. [PubMed]
  • Ray A, Gulati K, Anand S, Vijayan VK. Pharmacological studies on mechanisms of aminophylline-induced seizures in rats. Indian J Exp Biol. 2005;43:849–53. [PubMed]
  • Rebola N, Coelho JE, Costenla AR, Lopes LV, Parada A, Oliveira CR, Soares-da-Silva P, de Mendonca A, Cunha RA. Decrease of adenosine A1 receptor density and of adenosine neuromodulation in the hippocampus of kindled rats. Eur J Neurosci. 2003;18:820–8. [PubMed]
  • Ren G, Li T, Lan JQ, Wilz A, Simon RP, Boison D. Lentiviral RNAi-induced downregulation of adenosine kinase in human mesenchymal stem cell grafts: a novel perspective for seizure control. Exp Neurol. 2007;208:26–37. [PMC free article] [PubMed]
  • Ribeiro JA, Sebastiao AM, de Mendonca A. Adenosine receptors in the nervous system: pathophysiological implications. Prog Neurobiol. 2002;68:377–92. [PubMed]
  • Roussinov KS, Lazarova MB, Atanassova-Shopova S. Experimental study of the effect of lithium, haloperidol, caffeine and theophylline on convulsive seizure reactions. Acta Physiol Pharmacol Bulg. 1974;2:67–75. [PubMed]
  • Samren EB, van Duijn CM, Christiaens GC, Hofman A, Lindhout D. Antiepileptic drug regimens and major congenital abnormalities in the offspring. Ann Neurol. 1999;46:739–46. [PubMed]
  • Shen H-Y, Li T, Boison D. A novel mouse model for sudden unexpected death in epilepsy (SUDEP): role of impaired adenosine clearance. Epilepsia Published Online. 2009 Aug 8; doi: 10.1111/j.1528-1167.2009.02248.x. 8:31PM. [PMC free article] [PubMed] [Cross Ref]
  • Sperlagh B, Zsilla G, Baranyi M, Illes P, Vizi ES. Purinergic modulation of glutamate release under ischemic-like conditions in the hippocampus. Neuroscience. 2007;149:99–111. [PubMed]
  • Spina E, Pisani F, Perucca E. Clinically significant pharmacokinetic drug interactions with carbamazepine. An update. Clin Pharmacokinet. 1996;31:198–214. [PubMed]
  • Stern L, Dannon PN, Hirschmann S, Schriber S, Amytal D, Dolberg OT, Grunhaus L. Aminophylline increases seizure length during electroconvulsive therapy. J Ect. 1999;15:252–7. [PubMed]
  • Sugimoto T, Sugimoto M, Uchida I, Mashimo T, Okada S. Inhibitory effect of theophylline on recombinant GABA(A) receptor. Neuroreport. 2001;12:489–93. [PubMed]
  • Svenningsson P, Nomikos GG, Fredholm BB. The stimulatory action and the development of tolerance to caffeine is associated with alterations in gene expression in specific brain regions. J Neurosci. 1999;19:4011–22. [PubMed]
  • Swiader MJ, Luszczki JJ, Wielosz M, Czuczwar SJ. Effect of histamine receptor antagonists on aminophylline-induced seizures and lethality in mice. Pharmacol Rep. 2005;57:531–5. [PubMed]
  • Szybala C, Pritchard EM, Wilz A, Kaplan DL, Boison D. Antiepileptic effects of silk-polymer based adenosine release in kindled rats. Exp Neurol. 2009 in press. [PMC free article] [PubMed]
  • Turski L, Ikonomidou C, Turski WA, Bortolotto ZA, Cavalheiro EA. Review: cholinergic mechanisms and epileptogenesis. The seizures induced by pilocarpine: a novel experimental model of intractable epilepsy. Synapse. 1989;3:154–71. [PubMed]
  • Usachev Y, Shmigol A, Pronchuk N, Kostyuk P, Verkhratsky A. Caffeine-induced calcium release from internal stores in cultured rat sensory neurons. Neuroscience. 1993;57:845–59. [PubMed]
  • Van Dellen RG. Clinical pharmacology. Series on pharmacology in practice. 4. Theophylline. Practical application of new knowledge. Mayo Clin Proc. 1979;54:733–45. [PubMed]
  • Verkhratsky A. Physiology and pathophysiology of the calcium store in the endoplasmic reticulum of neurons. Physiol Rev. 2005;85:201–79. [PubMed]
  • Wu C, Wong T, Wu X, Sheppy E, Zhang L. Adenosine as an endogenous regulating factor of hippocampal sharp waves. Hippocampus. 2009;19:205–20. [PubMed]
  • Yoshikawa H. First-line therapy for theophylline-associated seizures. Acta Neurol Scand. 2007;115:57–61. [PubMed]
  • Young D, Dragunow M. Status epilepticus may be caused by loss of adenosine anticonvulsant mechanisms. Neurosci. 1994;58:245–261. [PubMed]
  • Yu L, Coelho JE, Zhang X, Fu Y, Tillman A, Karaoz U, Fredholm BB, Weng Z, Chen JF. Uncovering multiple molecular targets for caffeine using a drug target validation strategy combining A 2A receptor knockout mice with microarray profiling. Physiol Genomics. 2009;37:199–210. [PubMed]
  • Zhi JG, Levy G. Effect of chronic caffeine administration on theophylline concentrations required to produce seizures in rats. Proc Soc Exp Biol Med. 1990;193:210–3. [PubMed]