- Email: [email protected]
A unifying concept of seizure onset and termination
Medical Hypotheses (2004) 62, 740–745
http://intl.elsevierhealth.com/journals/mehy
A unifying concept of seizure onset and terminationq Glenn Doman, Ralph Pelligra* The Institutes for the Achievement of Human Potential, 8801 Stenton Avenue, Wyndmoor, PA 19038, USA Received 4 March 2003; accepted 22 October 2003
Summary Recent discoveries in molecular biology and human genetics have contributed greatly to an understanding of the nature of seizure (ictal) activity. However, two questions of fundamental clinical importance continue to resist scientific inquiry: when and why does a seizure begin; and when and why does a seizure end? This paper cites evidence from the medical literature in support of two counterintuitive concepts that address this issue. First, that despite the diversity of conditions that are associated with seizures, the ictal response results from disturbances of a mitochondrial metabolic pathway that is common to them all. Second, that the seizure is not inherently harmful but is, instead, associated with massive intracerebral circulatory changes that are intended to restore impaired mitochondrial function. We hypothesize that the protogenic pathophysiological condition leading to neuronal hyperexitability and seizures results from inadequate mitochondrial energy production due to hypoxia or a hypoxia-equivalent state. Failure to generate sufficient adenosine triphosphate compromises ionic pump function and the ability to maintain neuronal homeostasis and stability. The seizure cascade is a heroic effort to perfuse the brain when local mechanisms fail to restore energy production and ionic equilibrium. In summary, a seizure starts when the neuron’s aerobic machinery fails to maintain effective ionic pump function and terminates when increased cerebral perfusion, associated with the seizure response, restores adequate supplies of metabolic nutrients required for mitochondrial respiration. This unorthodox unifying concept that views ictogenesis as part of a restorative process rather than as a life threatening event may provide the basis for a much needed paradigm shift in the management of seizures. Current antiepileptic drugs are associated with many serious side effects, including death, and fail to control seizures in 20% of patients with primary generalized epilepsy and 35% of patients with partial epilepsy. We propose that efforts to prevent and control seizures should be directed away from pharma-chemical suppression towards removing the causes of disturbed neuronal energy production and developing methods and bioactive agents that promote an optimized physiological milieu within the brain. c 2004 Elsevier Ltd. All rights reserved.
Introduction q The Matsuzawa-Samoto Research Fund, the institutes for the Achievement of Human Potential. * Corresponding author. Present address: The National Aeronautics and Space Administration (NASA), Moffett Field, CA 94035-1000, USA. Tel.: +1-650-604-5163; fax: +1-650-604-1273. E-mail address: [email protected] (R. Pelligra).
Seizure behavior in humans has been observed and recorded since ancient times and remains a clinical enigma to this day. Modern medical science has had some limited success in controlling seizures by pharma-chemical suppression of the central
0306-9877/$ - see front matter c 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.mehy.2003.10.020
A unifying concept of seizure onset and termination nervous system (CNS) and by the surgical ablation of epileptigenic foci in the brain. Progress in cellular and molecular biology has identified various classes of ion channels and neurotransmitter receptors that are major determinants of neuronal excitability and synchronization, the presumed ‘‘causes’’ of seizures. However, despite these advances, there is still no satisfactory explanation as to when and why a seizure begins, and when and why a seizure terminates. A clear understanding of these fundamental processes has obvious therapeutic implications. The search for a unifying concept of seizure onset and termination is compelled by the observation that the convulsive symptom complex is ubiquitous and stereotypical in nature. Seizures can be induced in any human and throughout the animal world. Moreover, the full convulsive response is essentially the same whether elicited by electroshock, toxic stimulus, or clinical pathology and whether it occurs in a child or adult. It has been proposed that these features define the seizure as a primitive, innate reflex mechanism that is provoked by a lethal threat to the brain [1,2]. Evidence is cited from the current medical literature that supports both this view and the related notion that hypoxia or a hypoxia equivalent state is the protogenic neuronal condition that precedes virtually every convulsive response. A theoretical model is presented that views seizure activity as both a consequence of mitochondrial metabolic insufficiency and as an integral part of the protective process that attempts to restore mitochondrial and neuronal homeostasis.
Discussion Ion channels and ictogenesis Nerve cell excitability has long been known to be associated with the flow of ions through specialized protein aggregates within neuronal membranes. These ion channels, whose primary function is to generate transient electrical signals, are present in great abundance and variety in central neurons. The characteristic electrical activity of neurons – their ability to conduct, transmit, and receive electric signals – results from the opening and closing of specific ion channel proteins in the neuron plasma membrane [3]. Epileptic discharges are a pathological extreme of normal neuronal excitability and synchrony that are modulated by ion channel function. Impaired ion channel function associated with marked changes in extracellular potassium concen-
741 tration ([Kþ ]o), was initially shown to be correlated with the onset of seizures both in vivo [4] and in vitro [5]. Later studies demonstrated that increases in [Kþ ]o can induce seizure activity [6] and increase the frequency of spontaneous ictal discharges [7]. The effects of increased [Kþ ]o on neuronal excitability are complex but include the increased release of excitatory neurotransmitters [8]. The indispensable role of neuronal excitability for normal CNS function, and the abundant presence in the brain of excitatory ligand-gated channels such as fast inotropic glutamate (NMDA) receptors, raise a puzzling question: why is not every brain epileptic? Some investigators, in an effort to resolve this apparent paradox, posit the existence of specific ‘‘epileptic’’ ion channels that result from physical or functional alterations of the ion channels [9]. And, in fact, mutations have been identified in two novel and related voltage-gated Kþ channel genes (KCNQ2 and KCNQ3) that are associated with a form of epilepsy known as benign familial neonatal convulsions (BFNC) [10]. However, in BFNC, the seizures disappear spontaneously in the post-natal period even though the gene defect persists. Genetic factors are also implicated in the ‘‘idiopathic generalized epilepsies’’ (IGE), a heterogeneous seizure group that constitutes a large proportion of all epilepsy as classified by the International League Against Epilepsy (ILAE) [11]. Yet, despite the presence of genetic alterations at birth, seizure activity in IGE is often not manifest until adolescence. It would appear then that the genetic alterations of ion channel and neuronal receptor function are icto-predisposing rather than icto-causal factors. It is our belief that every brain is ‘‘epileptic, ’’ in that every brain is capable of repeated seizure activity. We take issue with the term ‘‘unprovoked’’ as a criterion for ‘‘epilepsy’’ since there must be a transitory event between the seizurefree and the seizure-active state. We propose that some brains, for genetic and other predisposing factors, are more vulnerable to external and internal perturbations of biological functions that ultimately interfere with neuronal energy production. Disturbances of mitochondrial respiration, adenosine triphosphate (ATP) synthesis, and ionic equilibrium are, in our view, the ultimate determinants of seizure occurrence and frequency.
Hypoxia, hypoxia equivalents and ATP synthesis Ion channels, although diverse in structure and function, share a common feature. They depend on energy from the binding or hydrolysis ATP to drive
742 metabolic pumps that actively transport ions against their concentration gradients. The energy expended by nerve cells to maintain the concentration gradients of Naþ , Kþ , Hþ and Ca2þ across the plasma and intracellular membranes is considerable, requiring approximately 25% of the total ATP produced by the cell [3]. Failure to provide sufficient energy to ensure effective ion channel function results in disturbed ionic homeostasis, which, as mentioned in the previous section, may lead to enhanced neuronal excitability and seizure activity. In support of this concept, Hagland and Schwartskroin [8], building on the work of Traynelis and Dingledine [6], have developed a convincing hypothetical model that illustrates how low Na–K pump activity can lead to increased [Kþ ]o and seizures. It is well known that ATP synthesis by mitochondrial respiration, and, therefore, Na–K pump function, is critically dependent on the presence of oxygen and oxidative substrate. Ischemia (diminished oxygen and glucose), hypoglycemia (diminished glucose) and hypoxia (resulting from a disturbance of oxygen availability, transport or utilization), all diminish the supply of ATP available to the cell. Hypoxia equivalents produce the identical ATP depleting effects as hypoxia by interfering with the oxidation of glucose to CO2 at various metabolic sites. This can occur in the glycolytic pathway that converts glucose to pyruvate, in the production of nicotinamide adenine dinucleotide (NADH) in the citric acid cycle, and anywhere in the electron transport chain where these products are re-oxidized by oxygen to form ATP. Some examples of hypoxia equivalents include, toxins and poisons such as 2,4-dinitrophenol (DNP) and cyanide that act as uncouplers allowing oxidation of NADH and the reduction of O2 to continue at high levels, but do not permit ATP synthesis [3]; metabolic inhibitors such as oligomycin and ouabain that interfere with mitochondrial ATPase; antimycin A and rotenone which are proximal inhibitors of the mitochondrial electron transport chain [12]; and an inherited disease, myoclonic epilepsy and ragged-red fibers (MERRF) that is associated with a genetic defect of mitochondrial function [13]. The end result of ischemia, hypoxia and hypoxia equivalents is the same; bioenergetically incompetent mitochondria that cannot support the functional needs of ion channels and their ability to harness and orchestrate normal neuronal excitability and synchronization. If local compensatory mechanisms fail to relieve the hypoxia/hypoxia equivalent state, a critical threshold of neuronal excitability is exceeded, and a seizure ensues.
Doman, Pelligra
The seizure as a response to, and protective reflex against, mitochondrial metabolic insufficiency There are two lines of reasoning and evidence that defend the counter-intuitive notion that a seizure is a physiological defense mechanism intended to restore impaired mitochondrial energy production. They are: (1) the consistent, often dramatic changes in cerebral blood flow prior to and during a seizure, and (2) a teleological justification of the purpose and value of the convulsive symptom complex. 1. Cerebral blood flow: Several potent vasodilator mechanisms have been identified in the cerebral circulation that are activated in response to hypoxia and seizures. These include nitric oxide (NO), trigeminal peptides and potassium channels [14]. A disturbance of potassium channel function can lead to hyperpolarizaton of cerebral vascular smooth muscle which appears to be a major mechanism for dilation of cerebral arteries. Calcitonin gene-related peptide (CGRP) and NO are also associated with the marked vasodi- latation during the phenomenon of cortical spreading depression [15], a condition analogous to seizures in humans [8]. Interictal hypoperfusion and cerebral blood flow alterations have been demonstrated in a wide range of seizure states including temporal lobe epilepsy [16], childhood absence seizures [17], infantile spasms [18], status epilepticus [19], West syndrome [20], and early onset benign childhood occipital seizure susceptibility syndrome (EBOSS) [21]. Weinand et al. [22] demonstrated that progressive hypoperfusion of the temporal lobe epileptic focus correlated with increased epileptogenicity and that the increased perfusion of the focus correlated with decreased epileptogenicity. Duncan [23], in an extensive review of imaging and epilepsy, states that, ‘‘the hallmark of an epileptic focus is an area of reduced glucose metabolism. . .that is commonly more extensive than the underlying anatomical abnormality.’’ It is reasonable to speculate, in view of these observations, that the profound vasodilatation and hyperperfusion associated with seizure activity is a protective response to pre-ictal hypoperfusion/ hypometabolism and reduced mitochondrial energy production. But why should these presumably protective, intracerebral vascular changes be associated with a violent, gross motor convulsive reaction?
A unifying concept of seizure onset and termination 2. A teleological basis for the concurrent intracerebral and convulsive components of a seizure: Fay [1,2], observes that the typical pattern of a convulsive seizure includes the turning of the head and eyes, the twisting of the trunk or extension of the back and the repetitive flexor–extensor movements of the prime muscles of the extremities. Most notably, it is devoid of the skilled rotary and prehensile coordinating movements of the motor cortex. He recognizes this pattern as characteristic of the amphibian level of motor development. Fay postulates that the convulsive symptom complex, retained in all humans, is a primitive, innate reflex inherited from our evolutionary forebears, the amphibians. It evolved as an emergency reflex response to ensure availability of life sustaining water, oxygen, salts and electrolytes as organisms struggled to transition from the sea to the land environment. Primitive life forms embarking on this journey, evolved simple motor patterns concerned with swimming, paddling and crawling about on the belly aided by crude legs responding to flexor–extensor impulses. A ‘‘defense’’ pattern in the CNS involving tail twisting, flapping, jerking and slapping motions ultimately evolved that could return the adventurous vertebrate to the life-nurturing sea if it wandered beyond the safe harbor of a water hole on land. A familiar example is the fisherman’s catch whose violent convulsive behavior will on occasion propel it from the bottom of the rowboat back into sea. The marked intracerebral vasodilation combined with the convulsive attempt to return the organism to a more biochemically hospitable environment are seen as a final, desperate, attempt to compensate for failure of local mechanisms to restore mitochondrial energy production and neuronal homeostasis.
A unifying model of seizure onset and termination The intracerebral (vascular) and extracerebral (convulsive) consequences of mitochondrial metabolic insufficiency described above, support the following theoretical schema of seizure onset and termination: Cerebral hypoxia/hypoxia equivalent state ! diminished mitochondrial ATP synthesis ! impaired Na–K pump function ! increased extracellular [Kþ ]o ! enhanced neuronal excitability ! seizure
743 response ! marked cerebral vasodilation ! increased cerebral perfusion (renewed access to oxygen) ! restored metabolic synthesis of ATP ! restored Na–K pump function ! restored ion channel homeostasis ! reduced neuronal excitability ! termination of seizure activity (cerebral normoxia state). It can be seen from this model that the seizure response is inherently self-limiting. However, if the seizure-associated hyperperfusion is unable to override the hypoxic stimulus, continuing inability of the Na–K pump to restore ionic homeostasis will lead to persistent ictal discharge, i.e. status epilepticus.
Brain trauma, electroshock and spreading depression Seizures that are associated with non-metabolic events such as brain trauma and electroshock may be related to the phenomenon of Leao’s spreading cortical depression (SD) [15]. SD is in concert with our own model and allows for seizures of nonmetabolic origin to be incorporated within it. SD is a reversible, protective response of brain tissue to a local blow, compression or other insult [15]. The most potent stimulus for SD is extracellular [Kþ ]o which may be due to cell membrane injury [24], hypoxia [25] or, inability of the Na–K pump to maintain ionic homeostasis. In electroshock, stimulation of non-synaptic mechanisms (gap junction) develop intense electrical fields [8] that generate SD. SD is not in itself pathological and may be repeated indefinitely without neural damage [26]. It is associated with marked vasodilatation [15] mediated by nitric oxide [14]. SD is manifest as a wave form that moves away from the insult site at a known rate [25] which coincides with the rate of propagation across motor cortex during the characteristic ‘‘Jacksonian March’’ of some forms of epilepsy [8]. Local increase in extracellular [Kþ ]o due to injury or any cause is immediately countered by a concurrent increase in Na–K pumping activated by an influx of Naþ into the cytosol. Consequently, a small region of Kþ release may increase or decrease depending on the local balance of forces. If [Kþ ]o diminishes, the region of cortex involved is stabilized. If Kþ cannot be contained by re-uptake, diffusion or transport away from the active site through the glial cell network, SD ensues [26]. Recovery from SD occurs primarily by re-uptake of Kþ . The rate of recovery is limited by the metabolic energy available and is reduced by hypoxia or hypoglycemia [27].
744
Concluding remarks We submit that an eventual solution to the perplexing clinical problem of seizure control will depend on a clearer understanding of the pathogenic role of mitochondrial metabolic insufficiency and a recognition that the seizure is both a consequence of mitochondrial dysfunction and an integral part of the protective process that attempts to restore mitochondrial energy production. The simplified model of seizure onset and termination presented here does not explore the many, complex, pathophysiological processes that comprise its component parts. However, its value may reside in providing insights that question the wisdom and validity of the current direction of research and therapeutic intervention. Antiepileptic drugs fail to control seizures in 20% of patients with primary generalized epilepsy and 35% of patients with partial epilepsy and are associated with many serious side effects, including death [28]. If, in fact, seizures are a protective reflex response to mitochondrial metabolic insufficiency, pharma-chemical suppression may not only be potentially harmful, but counterproductive as well. Efforts to prevent and control seizures should be directed towards removing the causes of disturbed neuronal energy production and developing methods and bioactive agents that create an optimum physiologic milieu within the CNS. At the gross level, for example, by maintaining fluid and electrolyte balance and correcting or facilitating cardiovascular, pulmonary and hematological functions that affect the delivery, uptake and utilization of oxygen by the brain; at the cellular level, by providing adequate substrate and biochemicals that affect function of the neuronal aerobic machinery. Nutritional pharmaceuticals such as ubiquinone (coQ10), alpha-lipoic acid, and acetyl-L -carnitine can help to promote efficient electron transport and oxidative phosphorylation and to reduce oxidative stress. Physiological ‘‘antihypoxic’’ methods and agents should be developed which protect against episodic tissue hypoxia and ischemia. This multilayered therapeutic approach may not be as convenient or appealing as a single pharmaceutical drug that can eliminate the visual offense of a seizure by inhibiting the CNS response. But, it is medically sound, and may ultimately offer the best hope for resolving this clinical dilemma that has plagued humanity for centuries. We recognize that a concept must be testable in order to qualify as a valid hypothesis. Evidence from the medical literature in support of the ideas
Doman, Pelligra presented in this paper, although compelling, is indirect. It may not be possible to construct a basic research model due to the inherent flaws in inducing seizures in animals and their random occurrence in humans. However, advanced imaging techniques such as functional magnetic resonance imaging and positron emission tomography offer the best opportunity for understanding metabolic and biochemical changes associated with pre- and post-ictal activity. Clinical validation studies that incorporate these tools and detailed outcome measures of seizure prevention and control are needed.
Acknowledgements The authors wish to thank Dr. Henry Leon, Dr. Leland Green, Dr. Ernesto Vasquez, Mr. Teruki Uemura, Mr. Phillip Phillips and Mr. David Bergner for their critical reviews of the manuscript and insightful suggestions.
References [1] Fay T. The other side of a fit. Amer J Psychiatry 1942; 39:196–200. [2] Doman G, Pelligra R. Ictogenesis: the origin of seizures in humans. A new look at an old theory. Med Hypoth 2003;60(1):129–32. [3] Lodish HF, Berk A, Zipursky S, Masudaira P, et al. editors. Nerve Cells. In: Molecular cell biology, 4th ed. New York: W.H. Freeman & Co; 1999. p. 647, 917 [Chapter 21]. [4] Futamachi KJ, Mutani R, Prince DA. Potassium activity in rabbit cortex. Brain Res 1974;75:5–25. [5] Hablitz JJ, Heinemann U. Extracellular Kþ and Ca2þ changes during epileptiform discharges in the immature cortex. Dev Brain Res 1987;36:299–303. [6] Traynelis SF, Dingledine R. Potassium induced spontaneous electrographic seizures in the rat hippocampal slice. J Neurophysiol 1988;59:259–76. [7] Rutecki PA, Lebeda F, Johnston D. Epileptiform activity induced by changes in extracellular potassium in hippocampus. J Neurophysiol 1985;54:1363–74. [8] Haglund NM, Schwartskroin PA. Role of Na–K pump potassium regulation and IPSPs in seizures and spreading depression in immature rabbit hippocampal slices. J Neurophysiol 1990;63(2):225–39. [9] Mody I. Ion channels in epilepsy. Int Rev Neurobiol 1998;42:199–225. [10] Schroeder BC, Kubisch C, Stein V, Jentsch TJ. Moderate loss of function of cyclic AMP-modulated KCNQ2/KCNQ3 Kþ channels causes epilepsy. Nature 1998;396:687–90. [11] Delgado-Escueta AV, Medina MT, Serratosa JM, et al. Mapping and positional cloning of common idiopathic generalized epilepsies: juvenile myoclonic epilepsy and childhood absence epilepsy. In: Olsen RW, Porter RJ, editors. Jasper’s basic mechanisms of the epilepsies. 3rd ed. Advances in neurology, vol. 79. Philadelphia, PA: Lippincott, Williams & Wikins; 1999. p. 351–74.
A unifying concept of seizure onset and termination [12] Patel AJ, Lauritzen I, Lazdunski M, Honore E. Disruption of mitochondrial respiration inhibits volume-regulated anion channels and provokes neuronal cell swelling. J Neurosci 1998;18(9):3117–23. [13] James AM, Wei YH, Pang CY, Murphy MP. Altered mitochondrial function in fibroblasts containing MELAS or MERRF mitochondrial DNA mutations. J Biochem 1996;318:401–7. [14] Brian JE, Faraci FM, Heistad DD. Recent insights into the regulation of cerebral circulation. Clin Exp Pharmacol Physiol 1996;23(6–7):449–57. [15] Leao AAP. Spreading depression of activity in the cerebral cortex. J Neurophsyiol 1944;7:359–89. [16] Guillon B, Duncan R, Biraben A, et al. Correlation between interictal regional ceebral blood flow and depth-recorded interictal spiking in temporal lobe epilepsy. Epilepsia 1998;39(1):67–76. [17] De Simone R, Silvestrini M, Mariani MG, Curatolo P. Changes in cerebral blood flow velocities during childhood absence seizures. Pediatr Neurol 1998;18(2):132–5. [18] Hwang PA, Otsubo H, Koo BK, et al. Infantile spasms: cerebral blood flow abnormalities correlate with EEG, neuroimaging and pathologic findings. Pediatr Neurol 1996;14(3):220–5. [19] Okuchi K, Nagata K, Otsuka H, et al. Regional cerebral blood flow after status epilepticus. Keio J Med 2000;40(S1):A75–6. [20] Karagol U, Deda G, Uysal S, et al. Cerebral blood flow abnormalities in symptomatic West syndrome: a single
745
[21]
[22]
[23] [24]
[25] [26]
[27]
[28]
photon emission computed tomography study. Pediatr Int 2001;43(1):66–70. Sakagami M, Takahashi Y, Matsuoka H, et al. A case of early-onset benign occipal seizure susceptibility syndrome: decreased cerebral blood flow in the occipiatal region detected by interictal single photon emission computed tomography, corresponding to the epileptogenic focus. Brain Develop 2001;23:4217–30. Weinand ME, Labiner DM, Ahern GL. Temporal lobe seizure interhemispheric propagation time depends on nonepileptic cortical cerebral blood flow. Epilepsy Res 2001;44(1): 33–9. Duncan JS. Imaging and epillepsy. Brain 1997;120:339–77. Reid KH, Marrannes R, Waliquier A. Spreading depression and central nervous system pharmacology. J Pharmocol Meth 1988;19:1–21. Amery WK. Brain hypoxia: the turning point in the genesis of the migraine attack? Cephalalgia 1982;2:83–109. DeLuca B, Shibata M, Brozek G, Bures J. Faciliation of Leao’s spreading depression by a pyropyrimidine derivative. Neuropharmacology 1975;14:537–45. Bures J, Buresova O, Krivanek J. The mechanism and applications of Leao’s spreading depression of electroencephalographic activity. New York: Academic Press; 1974. Willmorr LJ, Wheless JW. Epilepsy. In: Dale DC, Federman DD, editors. Scientific American Medicine. New York, NY: WebMD corporation; 2001.
http://intl.elsevierhealth.com/journals/mehy
A unifying concept of seizure onset and terminationq Glenn Doman, Ralph Pelligra* The Institutes for the Achievement of Human Potential, 8801 Stenton Avenue, Wyndmoor, PA 19038, USA Received 4 March 2003; accepted 22 October 2003
Summary Recent discoveries in molecular biology and human genetics have contributed greatly to an understanding of the nature of seizure (ictal) activity. However, two questions of fundamental clinical importance continue to resist scientific inquiry: when and why does a seizure begin; and when and why does a seizure end? This paper cites evidence from the medical literature in support of two counterintuitive concepts that address this issue. First, that despite the diversity of conditions that are associated with seizures, the ictal response results from disturbances of a mitochondrial metabolic pathway that is common to them all. Second, that the seizure is not inherently harmful but is, instead, associated with massive intracerebral circulatory changes that are intended to restore impaired mitochondrial function. We hypothesize that the protogenic pathophysiological condition leading to neuronal hyperexitability and seizures results from inadequate mitochondrial energy production due to hypoxia or a hypoxia-equivalent state. Failure to generate sufficient adenosine triphosphate compromises ionic pump function and the ability to maintain neuronal homeostasis and stability. The seizure cascade is a heroic effort to perfuse the brain when local mechanisms fail to restore energy production and ionic equilibrium. In summary, a seizure starts when the neuron’s aerobic machinery fails to maintain effective ionic pump function and terminates when increased cerebral perfusion, associated with the seizure response, restores adequate supplies of metabolic nutrients required for mitochondrial respiration. This unorthodox unifying concept that views ictogenesis as part of a restorative process rather than as a life threatening event may provide the basis for a much needed paradigm shift in the management of seizures. Current antiepileptic drugs are associated with many serious side effects, including death, and fail to control seizures in 20% of patients with primary generalized epilepsy and 35% of patients with partial epilepsy. We propose that efforts to prevent and control seizures should be directed away from pharma-chemical suppression towards removing the causes of disturbed neuronal energy production and developing methods and bioactive agents that promote an optimized physiological milieu within the brain. c 2004 Elsevier Ltd. All rights reserved.
Introduction q The Matsuzawa-Samoto Research Fund, the institutes for the Achievement of Human Potential. * Corresponding author. Present address: The National Aeronautics and Space Administration (NASA), Moffett Field, CA 94035-1000, USA. Tel.: +1-650-604-5163; fax: +1-650-604-1273. E-mail address: [email protected] (R. Pelligra).
Seizure behavior in humans has been observed and recorded since ancient times and remains a clinical enigma to this day. Modern medical science has had some limited success in controlling seizures by pharma-chemical suppression of the central
0306-9877/$ - see front matter c 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.mehy.2003.10.020
A unifying concept of seizure onset and termination nervous system (CNS) and by the surgical ablation of epileptigenic foci in the brain. Progress in cellular and molecular biology has identified various classes of ion channels and neurotransmitter receptors that are major determinants of neuronal excitability and synchronization, the presumed ‘‘causes’’ of seizures. However, despite these advances, there is still no satisfactory explanation as to when and why a seizure begins, and when and why a seizure terminates. A clear understanding of these fundamental processes has obvious therapeutic implications. The search for a unifying concept of seizure onset and termination is compelled by the observation that the convulsive symptom complex is ubiquitous and stereotypical in nature. Seizures can be induced in any human and throughout the animal world. Moreover, the full convulsive response is essentially the same whether elicited by electroshock, toxic stimulus, or clinical pathology and whether it occurs in a child or adult. It has been proposed that these features define the seizure as a primitive, innate reflex mechanism that is provoked by a lethal threat to the brain [1,2]. Evidence is cited from the current medical literature that supports both this view and the related notion that hypoxia or a hypoxia equivalent state is the protogenic neuronal condition that precedes virtually every convulsive response. A theoretical model is presented that views seizure activity as both a consequence of mitochondrial metabolic insufficiency and as an integral part of the protective process that attempts to restore mitochondrial and neuronal homeostasis.
Discussion Ion channels and ictogenesis Nerve cell excitability has long been known to be associated with the flow of ions through specialized protein aggregates within neuronal membranes. These ion channels, whose primary function is to generate transient electrical signals, are present in great abundance and variety in central neurons. The characteristic electrical activity of neurons – their ability to conduct, transmit, and receive electric signals – results from the opening and closing of specific ion channel proteins in the neuron plasma membrane [3]. Epileptic discharges are a pathological extreme of normal neuronal excitability and synchrony that are modulated by ion channel function. Impaired ion channel function associated with marked changes in extracellular potassium concen-
741 tration ([Kþ ]o), was initially shown to be correlated with the onset of seizures both in vivo [4] and in vitro [5]. Later studies demonstrated that increases in [Kþ ]o can induce seizure activity [6] and increase the frequency of spontaneous ictal discharges [7]. The effects of increased [Kþ ]o on neuronal excitability are complex but include the increased release of excitatory neurotransmitters [8]. The indispensable role of neuronal excitability for normal CNS function, and the abundant presence in the brain of excitatory ligand-gated channels such as fast inotropic glutamate (NMDA) receptors, raise a puzzling question: why is not every brain epileptic? Some investigators, in an effort to resolve this apparent paradox, posit the existence of specific ‘‘epileptic’’ ion channels that result from physical or functional alterations of the ion channels [9]. And, in fact, mutations have been identified in two novel and related voltage-gated Kþ channel genes (KCNQ2 and KCNQ3) that are associated with a form of epilepsy known as benign familial neonatal convulsions (BFNC) [10]. However, in BFNC, the seizures disappear spontaneously in the post-natal period even though the gene defect persists. Genetic factors are also implicated in the ‘‘idiopathic generalized epilepsies’’ (IGE), a heterogeneous seizure group that constitutes a large proportion of all epilepsy as classified by the International League Against Epilepsy (ILAE) [11]. Yet, despite the presence of genetic alterations at birth, seizure activity in IGE is often not manifest until adolescence. It would appear then that the genetic alterations of ion channel and neuronal receptor function are icto-predisposing rather than icto-causal factors. It is our belief that every brain is ‘‘epileptic, ’’ in that every brain is capable of repeated seizure activity. We take issue with the term ‘‘unprovoked’’ as a criterion for ‘‘epilepsy’’ since there must be a transitory event between the seizurefree and the seizure-active state. We propose that some brains, for genetic and other predisposing factors, are more vulnerable to external and internal perturbations of biological functions that ultimately interfere with neuronal energy production. Disturbances of mitochondrial respiration, adenosine triphosphate (ATP) synthesis, and ionic equilibrium are, in our view, the ultimate determinants of seizure occurrence and frequency.
Hypoxia, hypoxia equivalents and ATP synthesis Ion channels, although diverse in structure and function, share a common feature. They depend on energy from the binding or hydrolysis ATP to drive
742 metabolic pumps that actively transport ions against their concentration gradients. The energy expended by nerve cells to maintain the concentration gradients of Naþ , Kþ , Hþ and Ca2þ across the plasma and intracellular membranes is considerable, requiring approximately 25% of the total ATP produced by the cell [3]. Failure to provide sufficient energy to ensure effective ion channel function results in disturbed ionic homeostasis, which, as mentioned in the previous section, may lead to enhanced neuronal excitability and seizure activity. In support of this concept, Hagland and Schwartskroin [8], building on the work of Traynelis and Dingledine [6], have developed a convincing hypothetical model that illustrates how low Na–K pump activity can lead to increased [Kþ ]o and seizures. It is well known that ATP synthesis by mitochondrial respiration, and, therefore, Na–K pump function, is critically dependent on the presence of oxygen and oxidative substrate. Ischemia (diminished oxygen and glucose), hypoglycemia (diminished glucose) and hypoxia (resulting from a disturbance of oxygen availability, transport or utilization), all diminish the supply of ATP available to the cell. Hypoxia equivalents produce the identical ATP depleting effects as hypoxia by interfering with the oxidation of glucose to CO2 at various metabolic sites. This can occur in the glycolytic pathway that converts glucose to pyruvate, in the production of nicotinamide adenine dinucleotide (NADH) in the citric acid cycle, and anywhere in the electron transport chain where these products are re-oxidized by oxygen to form ATP. Some examples of hypoxia equivalents include, toxins and poisons such as 2,4-dinitrophenol (DNP) and cyanide that act as uncouplers allowing oxidation of NADH and the reduction of O2 to continue at high levels, but do not permit ATP synthesis [3]; metabolic inhibitors such as oligomycin and ouabain that interfere with mitochondrial ATPase; antimycin A and rotenone which are proximal inhibitors of the mitochondrial electron transport chain [12]; and an inherited disease, myoclonic epilepsy and ragged-red fibers (MERRF) that is associated with a genetic defect of mitochondrial function [13]. The end result of ischemia, hypoxia and hypoxia equivalents is the same; bioenergetically incompetent mitochondria that cannot support the functional needs of ion channels and their ability to harness and orchestrate normal neuronal excitability and synchronization. If local compensatory mechanisms fail to relieve the hypoxia/hypoxia equivalent state, a critical threshold of neuronal excitability is exceeded, and a seizure ensues.
Doman, Pelligra
The seizure as a response to, and protective reflex against, mitochondrial metabolic insufficiency There are two lines of reasoning and evidence that defend the counter-intuitive notion that a seizure is a physiological defense mechanism intended to restore impaired mitochondrial energy production. They are: (1) the consistent, often dramatic changes in cerebral blood flow prior to and during a seizure, and (2) a teleological justification of the purpose and value of the convulsive symptom complex. 1. Cerebral blood flow: Several potent vasodilator mechanisms have been identified in the cerebral circulation that are activated in response to hypoxia and seizures. These include nitric oxide (NO), trigeminal peptides and potassium channels [14]. A disturbance of potassium channel function can lead to hyperpolarizaton of cerebral vascular smooth muscle which appears to be a major mechanism for dilation of cerebral arteries. Calcitonin gene-related peptide (CGRP) and NO are also associated with the marked vasodi- latation during the phenomenon of cortical spreading depression [15], a condition analogous to seizures in humans [8]. Interictal hypoperfusion and cerebral blood flow alterations have been demonstrated in a wide range of seizure states including temporal lobe epilepsy [16], childhood absence seizures [17], infantile spasms [18], status epilepticus [19], West syndrome [20], and early onset benign childhood occipital seizure susceptibility syndrome (EBOSS) [21]. Weinand et al. [22] demonstrated that progressive hypoperfusion of the temporal lobe epileptic focus correlated with increased epileptogenicity and that the increased perfusion of the focus correlated with decreased epileptogenicity. Duncan [23], in an extensive review of imaging and epilepsy, states that, ‘‘the hallmark of an epileptic focus is an area of reduced glucose metabolism. . .that is commonly more extensive than the underlying anatomical abnormality.’’ It is reasonable to speculate, in view of these observations, that the profound vasodilatation and hyperperfusion associated with seizure activity is a protective response to pre-ictal hypoperfusion/ hypometabolism and reduced mitochondrial energy production. But why should these presumably protective, intracerebral vascular changes be associated with a violent, gross motor convulsive reaction?
A unifying concept of seizure onset and termination 2. A teleological basis for the concurrent intracerebral and convulsive components of a seizure: Fay [1,2], observes that the typical pattern of a convulsive seizure includes the turning of the head and eyes, the twisting of the trunk or extension of the back and the repetitive flexor–extensor movements of the prime muscles of the extremities. Most notably, it is devoid of the skilled rotary and prehensile coordinating movements of the motor cortex. He recognizes this pattern as characteristic of the amphibian level of motor development. Fay postulates that the convulsive symptom complex, retained in all humans, is a primitive, innate reflex inherited from our evolutionary forebears, the amphibians. It evolved as an emergency reflex response to ensure availability of life sustaining water, oxygen, salts and electrolytes as organisms struggled to transition from the sea to the land environment. Primitive life forms embarking on this journey, evolved simple motor patterns concerned with swimming, paddling and crawling about on the belly aided by crude legs responding to flexor–extensor impulses. A ‘‘defense’’ pattern in the CNS involving tail twisting, flapping, jerking and slapping motions ultimately evolved that could return the adventurous vertebrate to the life-nurturing sea if it wandered beyond the safe harbor of a water hole on land. A familiar example is the fisherman’s catch whose violent convulsive behavior will on occasion propel it from the bottom of the rowboat back into sea. The marked intracerebral vasodilation combined with the convulsive attempt to return the organism to a more biochemically hospitable environment are seen as a final, desperate, attempt to compensate for failure of local mechanisms to restore mitochondrial energy production and neuronal homeostasis.
A unifying model of seizure onset and termination The intracerebral (vascular) and extracerebral (convulsive) consequences of mitochondrial metabolic insufficiency described above, support the following theoretical schema of seizure onset and termination: Cerebral hypoxia/hypoxia equivalent state ! diminished mitochondrial ATP synthesis ! impaired Na–K pump function ! increased extracellular [Kþ ]o ! enhanced neuronal excitability ! seizure
743 response ! marked cerebral vasodilation ! increased cerebral perfusion (renewed access to oxygen) ! restored metabolic synthesis of ATP ! restored Na–K pump function ! restored ion channel homeostasis ! reduced neuronal excitability ! termination of seizure activity (cerebral normoxia state). It can be seen from this model that the seizure response is inherently self-limiting. However, if the seizure-associated hyperperfusion is unable to override the hypoxic stimulus, continuing inability of the Na–K pump to restore ionic homeostasis will lead to persistent ictal discharge, i.e. status epilepticus.
Brain trauma, electroshock and spreading depression Seizures that are associated with non-metabolic events such as brain trauma and electroshock may be related to the phenomenon of Leao’s spreading cortical depression (SD) [15]. SD is in concert with our own model and allows for seizures of nonmetabolic origin to be incorporated within it. SD is a reversible, protective response of brain tissue to a local blow, compression or other insult [15]. The most potent stimulus for SD is extracellular [Kþ ]o which may be due to cell membrane injury [24], hypoxia [25] or, inability of the Na–K pump to maintain ionic homeostasis. In electroshock, stimulation of non-synaptic mechanisms (gap junction) develop intense electrical fields [8] that generate SD. SD is not in itself pathological and may be repeated indefinitely without neural damage [26]. It is associated with marked vasodilatation [15] mediated by nitric oxide [14]. SD is manifest as a wave form that moves away from the insult site at a known rate [25] which coincides with the rate of propagation across motor cortex during the characteristic ‘‘Jacksonian March’’ of some forms of epilepsy [8]. Local increase in extracellular [Kþ ]o due to injury or any cause is immediately countered by a concurrent increase in Na–K pumping activated by an influx of Naþ into the cytosol. Consequently, a small region of Kþ release may increase or decrease depending on the local balance of forces. If [Kþ ]o diminishes, the region of cortex involved is stabilized. If Kþ cannot be contained by re-uptake, diffusion or transport away from the active site through the glial cell network, SD ensues [26]. Recovery from SD occurs primarily by re-uptake of Kþ . The rate of recovery is limited by the metabolic energy available and is reduced by hypoxia or hypoglycemia [27].
744
Concluding remarks We submit that an eventual solution to the perplexing clinical problem of seizure control will depend on a clearer understanding of the pathogenic role of mitochondrial metabolic insufficiency and a recognition that the seizure is both a consequence of mitochondrial dysfunction and an integral part of the protective process that attempts to restore mitochondrial energy production. The simplified model of seizure onset and termination presented here does not explore the many, complex, pathophysiological processes that comprise its component parts. However, its value may reside in providing insights that question the wisdom and validity of the current direction of research and therapeutic intervention. Antiepileptic drugs fail to control seizures in 20% of patients with primary generalized epilepsy and 35% of patients with partial epilepsy and are associated with many serious side effects, including death [28]. If, in fact, seizures are a protective reflex response to mitochondrial metabolic insufficiency, pharma-chemical suppression may not only be potentially harmful, but counterproductive as well. Efforts to prevent and control seizures should be directed towards removing the causes of disturbed neuronal energy production and developing methods and bioactive agents that create an optimum physiologic milieu within the CNS. At the gross level, for example, by maintaining fluid and electrolyte balance and correcting or facilitating cardiovascular, pulmonary and hematological functions that affect the delivery, uptake and utilization of oxygen by the brain; at the cellular level, by providing adequate substrate and biochemicals that affect function of the neuronal aerobic machinery. Nutritional pharmaceuticals such as ubiquinone (coQ10), alpha-lipoic acid, and acetyl-L -carnitine can help to promote efficient electron transport and oxidative phosphorylation and to reduce oxidative stress. Physiological ‘‘antihypoxic’’ methods and agents should be developed which protect against episodic tissue hypoxia and ischemia. This multilayered therapeutic approach may not be as convenient or appealing as a single pharmaceutical drug that can eliminate the visual offense of a seizure by inhibiting the CNS response. But, it is medically sound, and may ultimately offer the best hope for resolving this clinical dilemma that has plagued humanity for centuries. We recognize that a concept must be testable in order to qualify as a valid hypothesis. Evidence from the medical literature in support of the ideas
Doman, Pelligra presented in this paper, although compelling, is indirect. It may not be possible to construct a basic research model due to the inherent flaws in inducing seizures in animals and their random occurrence in humans. However, advanced imaging techniques such as functional magnetic resonance imaging and positron emission tomography offer the best opportunity for understanding metabolic and biochemical changes associated with pre- and post-ictal activity. Clinical validation studies that incorporate these tools and detailed outcome measures of seizure prevention and control are needed.
Acknowledgements The authors wish to thank Dr. Henry Leon, Dr. Leland Green, Dr. Ernesto Vasquez, Mr. Teruki Uemura, Mr. Phillip Phillips and Mr. David Bergner for their critical reviews of the manuscript and insightful suggestions.
References [1] Fay T. The other side of a fit. Amer J Psychiatry 1942; 39:196–200. [2] Doman G, Pelligra R. Ictogenesis: the origin of seizures in humans. A new look at an old theory. Med Hypoth 2003;60(1):129–32. [3] Lodish HF, Berk A, Zipursky S, Masudaira P, et al. editors. Nerve Cells. In: Molecular cell biology, 4th ed. New York: W.H. Freeman & Co; 1999. p. 647, 917 [Chapter 21]. [4] Futamachi KJ, Mutani R, Prince DA. Potassium activity in rabbit cortex. Brain Res 1974;75:5–25. [5] Hablitz JJ, Heinemann U. Extracellular Kþ and Ca2þ changes during epileptiform discharges in the immature cortex. Dev Brain Res 1987;36:299–303. [6] Traynelis SF, Dingledine R. Potassium induced spontaneous electrographic seizures in the rat hippocampal slice. J Neurophysiol 1988;59:259–76. [7] Rutecki PA, Lebeda F, Johnston D. Epileptiform activity induced by changes in extracellular potassium in hippocampus. J Neurophysiol 1985;54:1363–74. [8] Haglund NM, Schwartskroin PA. Role of Na–K pump potassium regulation and IPSPs in seizures and spreading depression in immature rabbit hippocampal slices. J Neurophysiol 1990;63(2):225–39. [9] Mody I. Ion channels in epilepsy. Int Rev Neurobiol 1998;42:199–225. [10] Schroeder BC, Kubisch C, Stein V, Jentsch TJ. Moderate loss of function of cyclic AMP-modulated KCNQ2/KCNQ3 Kþ channels causes epilepsy. Nature 1998;396:687–90. [11] Delgado-Escueta AV, Medina MT, Serratosa JM, et al. Mapping and positional cloning of common idiopathic generalized epilepsies: juvenile myoclonic epilepsy and childhood absence epilepsy. In: Olsen RW, Porter RJ, editors. Jasper’s basic mechanisms of the epilepsies. 3rd ed. Advances in neurology, vol. 79. Philadelphia, PA: Lippincott, Williams & Wikins; 1999. p. 351–74.
A unifying concept of seizure onset and termination [12] Patel AJ, Lauritzen I, Lazdunski M, Honore E. Disruption of mitochondrial respiration inhibits volume-regulated anion channels and provokes neuronal cell swelling. J Neurosci 1998;18(9):3117–23. [13] James AM, Wei YH, Pang CY, Murphy MP. Altered mitochondrial function in fibroblasts containing MELAS or MERRF mitochondrial DNA mutations. J Biochem 1996;318:401–7. [14] Brian JE, Faraci FM, Heistad DD. Recent insights into the regulation of cerebral circulation. Clin Exp Pharmacol Physiol 1996;23(6–7):449–57. [15] Leao AAP. Spreading depression of activity in the cerebral cortex. J Neurophsyiol 1944;7:359–89. [16] Guillon B, Duncan R, Biraben A, et al. Correlation between interictal regional ceebral blood flow and depth-recorded interictal spiking in temporal lobe epilepsy. Epilepsia 1998;39(1):67–76. [17] De Simone R, Silvestrini M, Mariani MG, Curatolo P. Changes in cerebral blood flow velocities during childhood absence seizures. Pediatr Neurol 1998;18(2):132–5. [18] Hwang PA, Otsubo H, Koo BK, et al. Infantile spasms: cerebral blood flow abnormalities correlate with EEG, neuroimaging and pathologic findings. Pediatr Neurol 1996;14(3):220–5. [19] Okuchi K, Nagata K, Otsuka H, et al. Regional cerebral blood flow after status epilepticus. Keio J Med 2000;40(S1):A75–6. [20] Karagol U, Deda G, Uysal S, et al. Cerebral blood flow abnormalities in symptomatic West syndrome: a single
745
[21]
[22]
[23] [24]
[25] [26]
[27]
[28]
photon emission computed tomography study. Pediatr Int 2001;43(1):66–70. Sakagami M, Takahashi Y, Matsuoka H, et al. A case of early-onset benign occipal seizure susceptibility syndrome: decreased cerebral blood flow in the occipiatal region detected by interictal single photon emission computed tomography, corresponding to the epileptogenic focus. Brain Develop 2001;23:4217–30. Weinand ME, Labiner DM, Ahern GL. Temporal lobe seizure interhemispheric propagation time depends on nonepileptic cortical cerebral blood flow. Epilepsy Res 2001;44(1): 33–9. Duncan JS. Imaging and epillepsy. Brain 1997;120:339–77. Reid KH, Marrannes R, Waliquier A. Spreading depression and central nervous system pharmacology. J Pharmocol Meth 1988;19:1–21. Amery WK. Brain hypoxia: the turning point in the genesis of the migraine attack? Cephalalgia 1982;2:83–109. DeLuca B, Shibata M, Brozek G, Bures J. Faciliation of Leao’s spreading depression by a pyropyrimidine derivative. Neuropharmacology 1975;14:537–45. Bures J, Buresova O, Krivanek J. The mechanism and applications of Leao’s spreading depression of electroencephalographic activity. New York: Academic Press; 1974. Willmorr LJ, Wheless JW. Epilepsy. In: Dale DC, Federman DD, editors. Scientific American Medicine. New York, NY: WebMD corporation; 2001.