Sage Journals: Discover world-class research

Abstract

Despite the relatively well-characterized headache mechanisms in migraine, upstream events triggering individual attacks are poorly understood. This lack of mechanistic insight has hampered a rational approach to prophylactic drug discovery. Unlike targeted abortive and analgesic interventions, mainstream migraine prophylaxis has been largely based on serendipitous observations (e.g. propranolol) and presumed class effects (e.g. anticonvulsants). Recent studies suggest that spreading depression is the final common pathophysiological target for several established or investigational migraine prophylactic drugs. Building on these observations, spreading depression can now be explored for its predictive utility as a preclinical drug screening paradigm in migraine prophylaxis.

Keywords

Spreading depression migraine prophylaxis experimental models drug screening

INTRODUCTION

Almost 70 years after Lashley's own account of visual aura implicated occipital cortex as the source (1), cortical spreading depression (SD) is now widely recognized as the neurophysiological substrate of classical migraine aura, and may be involved in migraines without a perceived aura as well. Since its discovery by Leão in the 1940s (2), SD has been a focus of intense investigation in relation to apparently diverse disease states including ischaemic or haemorrhagic stroke, head trauma and, importantly, migraine. The similarities between cortical SD and migraine aura were pointed out by Leão based on their characteristic slow spread (3), and a direct pathophysiological link has been theorized since the 1950s (4, 5). The evidence for an association between SD and migraine aura, albeit indirect, has been steadily accumulating over the past few decades with the help of technological advances.

Spreading depression

SD, observed across a broad range of species from locust to man (6–9), is a transient depression of all spontaneous and evoked activity in neural tissue, often preceded by a brief burst of action potential firing, which slowly propagates in brain tissue centrifugally from the point of origin. Despite the name depression, the hallmarks of SD are the complete and prolonged membrane depolarization of both neurons and glia, and the associated massive K⁺ efflux increasing the extracellular K⁺ concentration ([K⁺]_e) from a resting level of 3–4 mm to > 40 mm. This is accompanied by a slow negative extracellular potential shift (d.c. shift), which is 10–30 mV in amplitude (Fig. 1). Whereas the membrane potential and [K⁺]_e are usually restored within less than a minute via Na⁺/K⁺ ATPase activation, as well as K⁺ clearance by astrocytes and diffusion, complete recovery of synaptic activity may take up to 10 min in the wake of an SD. Although the channels mediating the ion fluxes during SD are still a matter of debate (10), it is believed that the surge in local [K⁺]_e depolarizes adjacent neurons and glia, and in this way the depolarization spreads. Membrane depolarization is accompanied by Na⁺ and water influx leading to cell swelling and shrinkage of extracellular space (11), and Ca²⁺ influx triggering glutamate release (e.g. presynaptic P/Q-type Ca²⁺ channels). Glutamate by activating N-methyl-D-aspartate (NMDA) receptors, and gap junctions by promoting intercellular Ca²⁺ and/or K⁺ waves, are believed to facilitate SD propagation. The speed of SD propagation ranges between 2 and 5 mm/min in cerebral cortex in most species under physiological conditions. At least in the mammalian brain, slower propagation speeds probably culminate in propagation failure; therefore, drugs that significantly reduce propagation speed can lead to propagation block.

Figure 1

Representative tracings of electrocorticogram (ECoG) and extracellular d.c. potential during a cortical spreading depression (SD) wave in rat brain. SD is characterized by transient depression of cortical electrical activity and a slow d.c. shift 20–30 mV in amplitude lasting < 1 min. When triggered by an intense stimulus that simultaneously depolarizes a minimum critical volume of brain tissue, SD spreads centrifugally to be detected first at the proximal (E1) and, after a latency, at the distal (E2) microelectrode. In otherwise normal tissue, ECoG activity starts to recover a few minutes after the onset of SD, but may take up to 10 min to return to normal. The speed of propagation is calculated from the distance between the electrodes and the latency. SD was triggered by 1 M KCl briefly applied on to the occipital cortex, and recorded by two intracortical glass microelectrodes placed in series with the KCl application site.

SD is triggered when a minimum critical volume of brain tissue (estimated to be ∼1 mm³ in rat cortex; (12)) is simultaneously depolarized by an intense stimulus, such as concentrated KCl application, direct electrical stimulation, trauma, ischaemia or epileptic activity. The critical event to initiate an SD is [K⁺]_e exceeding a critical threshold (∼10–12 mm) for depolarization (13, 14). The minimum critical volume of depolarized tissue, the magnitude of [K⁺]_e rise during stimulation (i.e. K⁺ release/clearance mismatch), and the [K⁺]_e threshold determine the ease with which SD occurs after a stimulus (15, 16). This is often termed SD susceptibility, and varies depending upon species, tissue morphology and physiological state, and can be genetically and hormonally modulated (17, 18). More importantly, modulation of SD susceptibility by drugs appears to be the basis for migraine prophylaxis, and provides an opportunity for developing preclinical drug screening models.

Although initially described in cortex, SD is known to occur readily in subcortical structures such as striatum, thalamus and hippocampus, with the exception of brainstem, which is resistant to SD unless tested in immature animals or after pharmacological preconditioning (e.g. K⁺ channel blockade) (19–23). Because of its dependence upon the K⁺ surge and to a lesser extent glutamate efflux, SD requires dense neuronal populations and limited extracellular space to propagate, and thus cannot enter white matter. Nevertheless, cortical SD can propagate into subcortical structures that have direct grey matter contiguity with the cortex (17, 24). For purposes of migraine drug screening, cortical SD susceptibility is most relevant, since cortex is most accessible, and the relationship of subcortical SD to migraine aura has not been established.

EVIDENCE SUPPORTING A ROLE FOR SPREADING DEPRESSION IN MIGRAINE

Several lines of evidence strongly, albeit indirectly, support a role for cortical SD in migraine aura:

(a)

Symptomatology of aura: Visual or somatosensory symptoms of migraine aura propagate at a rate that corresponds well to the propagation speed of experimental cortical SD, based on the known somato/retinotopic organization of human cortex (1, 4, 25). Moreover, negative symptoms of aura (e.g. expanding scotoma or hypoaesthesia) are often preceded by positive symptoms (e.g. visual scintillations or paraesthesia), resembling the brief period of neuronal burst firing preceding the suppression of activity during SD.

(b)

Cerebral haemodynamics during aura: Experimental SD is associated with a brief and variable hyperaemia lasting a few minutes, followed by an oligaemic phase lasting up to an hour during which functional coupling and hypercapnic hyperaemia are impaired. A long-lasting oligaemia similar in magnitude and duration has been detected during migraine attacks in a few studies, during which functional coupling and hypercapnic hyperaemia were diminished (26–28). A preceding hyperaemia has been more difficult to demonstrate in migraineurs (29), presumably because of the unpredictable onset of attacks and the insufficient temporal and spatial resolution of clinical imaging techniques. More recently, functional magnetic resonance imaging (MRI) during visual aura revealed a brief focal increase in brain oxygen level-dependent (BOLD) signal that propagated within the occipital cortex at a speed of ∼3 mm/min and was retinotopically congruent with the patient's visual experience (25). The BOLD signal increase was followed by a longer lasting decrease and impaired BOLD response to functional activation. These abnormalities were strictly unilateral, consistent with the unilaterality of migraine headache.

(c)

Downstream pain mechanisms: Cortical SD can activate brainstem trigeminal nucleus caudalis (30), dilate meningeal arteries with a latency fairly consistent with that of headache after aura (31), and cause plasma extravasation and inflammation (32). Furthermore, cortical SD stimulates calcitonin gene-related peptide and nitric oxide release and activates perivascular nerves (33–35).

(d)

Genetic modulation of migraine and SD: Human mutations associated with familial hemiplegic migraine type 1 (FHM1) (17, 36) and cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (Eikermann-Haerter and Ayata, unpublished data) enhance SD susceptibility in mutant mouse models. Furthermore, FHM1 mutations also augment the severity of post-SD neurological deficits, thus mimicking the phenotype in this severe human migraine syndrome (17).

(e)

Hormonal modulation of migraine and SD: Migraine is three times more prevalent in women than in men during reproductive age (i.e. menarche to menopause). Recently, SD susceptibility was shown to be significantly higher in female mice compared with male, and the sex difference was abolished by gonadectomy and ageing (17, 18).

(f)

Pharmacological modulation: Five clinically effective migraine prophylactic drugs (valproate, topiramate, propranolol, amitriptyline, methysergide) were shown to suppress SD susceptibility (37). These drugs belong to different pharmacological classes, making it unlikely that SD suppression is an incidental class effect. Furthermore, SD suppression required chronic treatment for weeks to months reminiscent of the temporal profile of their clinical efficacy in migraineurs. Interestingly, these drugs are clinically effective not only in migraine with aura but also in migraine without aura, implicating SD in the latter condition as well.

In summary, evidence strongly implicates cortical SD as the electrophysiological event underlying migraine aura, and suggests that it can trigger downstream pain pathways at least in a subset of migraineurs. Another form of spreading ionic shift, astrocytic Ca²⁺ waves, is unlikely to be relevant for migraine, since neither the propagation speed (≤ 1 mm/min) nor the propagation distance (< 500 μm) resembles migraine aura. Although the occurrence of SD-like propagating depolarization events has recently been shown in injured human cortex using subdural electrode strips (38), it is not yet known how cortical SD is triggered in migraineurs with aura in the absence of an underlying injury.

EVIDENCE FOR SPREADING DEPRESSION AS A SERENDIPITOUS THERAPEUTIC TARGET IN MIGRAINE

The similarities between SD and migraine aura have led a number of laboratories to test migraine drugs on SD, with mixed results. The experimental models and species used, drug doses/concentrations, administration routes and durations markedly differed, making comparisons and reconciliation difficult.

In vitro studies

Retinal SD has been a common in vitro model for pharmacological studies. Migraine drugs tested in this model, even the abortive or analgesic drugs (e.g. sumatriptan, acetylsalicylic acid), appeared to suppress retinal SD induction, duration or speed (39–41); however, the effective concentrations were often supraphysiological. Notably, only propranolol and methysergide were effective at submillimolar concentrations, perhaps suggesting that this in vitro model is relevant for migraine prophylaxis only when SD suppression is achieved at low concentrations.

In vivo studies

A total of 16 studies from eight different laboratories tested 12 prophylactic and five abortive or analgesic drugs on cortical SD after systemic administration, in vivo (Table 1). Experimental paradigms to trigger and monitor SD, and the efficacy end-points varied significantly, and all but two (37, 42) explored acute drug effects (i.e. single dose, or short-term daily treatment ≤ 1 week). In contrast to retinal SD, only three drugs, all anticonvulsant prophylactics, appeared to acutely suppress cortical SD. Tonabersat (SB-220453) suppressed KCl-evoked SDs in all four studies from one group of investigators (43–45) and their collaborators (46), whereas flunarizine showed a somewhat dose-dependent SD suppression by one laboratory (47–50), but was ineffective at a lower daily dose for 5 days in the hands of another (15). Topiramate (30 mg/kg) suppressed repetitive pinprick-induced SDs in one study when administered 30 min before testing, and the effect was reversed within 60 min (51). In contrast, a higher dose of topiramate (80 mg/kg) given as a single dose 60 min before testing, or daily for 1 week, was ineffective on SD in another study (37). Although the sources of discrepancy between the two studies are not clear, differences in the method of SD induction (pinprick vs. KCl or electrical) and monitoring (cortical blood flow and single unit activity vs. d.c. potential) may be responsible. Apart from the three drugs, acute in vivo studies did not support the overall notion that SD is a target in migraine prophylaxis.

Table 1

Systemic administration of selected migraine drugs on cortical spreading depression (SD), in vivo

Prophylactic	Species	Monitor	Dose, route	Pretreatment	Susceptibility	Speed	Dur	Amp	Ref
Valproate	Cat	LDF+SUA	3.5–7 mg/kg, i.v.	60 min	ND	↔	ND	ND	(52)
	Rat	DC	200 mg/kg/day, i.p.	1 week	↔ (E, KCl)	↔	↔	↔	(37)
	Rat	DC	25–200 mg/kg/day, i.p.	≥ 4 weeks	↓ (E, KCl)	↓	↔	↔	(37)
Topiramate	Cat, rat	LDF+SUA	30 mg/kg, i.v.	30 min	↓ (M)	↔	↔	↔	(51)
	Rat	DC	80 mg/kg/day, i.p.	1 week	↔ (E, KCl)	↔	↔	↔	(37)
	Rat	DC	60–80 mg/kg/day, i.p.	≥ 4 weeks	↓ (E, KCl)	↓	↔	↔	(37)
Tonabersat	Cat	DC	3–10 mg/kg, i.p.	90 min	↓ (KCl)	ND	↔	↔	(43)
	Cat	DC	3–10 mg/kg, i.p.	90 min	↓ (KCl)	ND	↔	↔	(44)
	Cat	MRI	10 mg/kg, i.p.	90 min	↓ (KCl)	↑	↓	ND	(46)
	Rat	DC	10 mg/kg, i.p.	90 min	↓ (KCl)	ND	ND	ND	(45)
Flunarizine	Rat	DC	20–40 mg/kg, i.p.	105 min	↔ (E)	ND	↔	↔	(47)
	Rat	DC	20 mg/kg, p.o. b.i.d.	36 h	↔ (E)	ND	↓	↔	(47)
	Rat	DC	20 mg/kg/day, p.o.	5 days	↔ (M)	↔	ND	↔	(15)
	Rat	DC	30 mg/kg, p.o.+10 mg/kg, i.p.	24 h	↓ (M, E)	ND	ND	ND	(48)
	Rat	DC	40 mg/kg, i.p.	60 min	↓ (M)	ND	ND	ND	(49)
	Rat	DC	10–20 mg/kg. i.p.	105 min	↔ (E)	ND	↓	ND	(50)
	Rat	DC	20 mg/kg, p.o. b.i.d.	36 h	↓ (KCl)	ND	↓	ND	(50)
Propranolol	Rat	DC	20 mg/kg, i.p.	105 min	↔ (E)	ND	↔	↔	(47)
	Rat	DC	20 mg/kg, p.o. b.i.d.	5 days	↔ (M)	↔	ND	↔	(15)
	Rat	DC	20 mg/kg, i.p.	60 min	↔ (E, KCl)	↔	↔	↔	(37)
	Rat	DC	20 mg/kg/day, i.p.	≥ 4 weeks	↓ (E, KCl)	↓	↔	↔	(37)
Metoprolol	Cat	LDF+SUA	25 mg/kg, i.v.	60 min	ND	↔	ND	ND	(52)
Clonidine	Rat	DC	1.25–10 mg/kg, i.p.	105 min	↔ (E)	ND	↑	↔	(47)
Methysergide	Rat	DC	10 mg/kg. i.p.	105 min	↔ (E)	ND	↔	↔	(47)
	Rat	DC	10 mg/kg/day, p.o.	5 days	↔ (M)	↔	ND	↔	(15)
	Rat	DC	0.1–1 mg/kg/day, i.p.	≥ 6 weeks	↓ (KCl)	↔	↔	↔	(37)
Pizotifen	Rat	DC	10 mg/kg, i.p.	105 min	↔ (E)	ND	↔	↔	(47)
Amitriptyline	Rat	DC	20 mg/kg, i.p.	60 min	↔ (E, KCl)	↔	↔	↔	(37)
Amitriptyline	Rat	DC	10–20 mg/kg/day, i.p.	≥ 4 weeks	↓ (E, KCl)	↓	↔	↔	(37)
Fluoxetine	Rat	DC	10–40 mg/kg/day, s.c.	≥ 25 days	ND	↓	↔	↔	(42)
Citalopram	Rat	DC	20 mg/kg/day, i.p.	1 week	ND	↓	↔	↔	(53)
Abortive/Analgesic
Sumatriptan	Cat	MRI	0.3 mg/kg, i.v.	15 min	↔ (KCl)	↔	↔	ND	(46)
	Cat, rat	DC	0.3 mg/kg, i.v.	15 min	↔ (KCl)	ND	ND	↔	(54)
	Rat	DC	0.3 mg/kg, i.v.	30 min	↔ (KCl)	ND	↓	↔	(30)
	Rat	DC	0.3 mg/kg, i.v.	15 min	↔ (KCl)	ND	ND	ND	(45)
Dihydroergotamine	Cat	LDF+SUA	15 μg/kg, i.v.	60 min	ND	↔	ND	ND	(52)
Ergotamine	Rat	DC	0.31–20 mg/kg, i.p.	105 min	↔ (E)	ND	↑	↔	(47)
Acetylsalicylic acid	Cat	LDF+SUA	30 mg/kg, i.v.	60 min	ND	↔	ND	ND	(52)
Lignocaine	Cat	LDF+SUA	5 mg/kg, i.v.	60 min	ND	↔	ND	ND	(52)

Dur, SD duration; Amp, SD amplitude; ND, not determined or reported; DC, slow extracellular potential shift; LDF, laser Doppler flowmetry measuring blood flow changes; SUA, single unit spike activity; E, electrically-induced SD; M, mechanical pinprick-induced SD; KCl, KCl-induced SD.

A number of potential caveats could explain the lack of efficacy of prophylactic drugs on SD, in vivo, such as anaesthetic influences, differences in tissue properties (e.g. no blood flow in retina preparation), and the sensitivity of techniques used to trigger and monitor SD. It has also been suggested that a predisposition to SD (e.g. injury) might be required for migraine drugs to suppress SD (47). More importantly, however, higher tissue drug concentrations may have been attained in vitro, since brain levels upon systemic administration would depend on pharmacokinetic factors (e.g. plasma levels and half-life) as well as drug penetration across the blood–brain barrier. In support of this, topical application of serotonin (5-HT) reuptake inhibitors citalopram (2.5–12.5 mm) and fluoxetine (15–30 mm) for 10 min appeared to suppress cortical SD more potently compared with systemic administration (42, 53). Topical application of propranolol (0.25–1 mm), in vivo, slowed, and at higher concentrations blocked cortical SD propagation triggered by pinprick in rats (55), whereas systemic administration of a single dose was ineffective (15, 37). Similarly, phenytoin suppressed retinal SD in vitro (8–80 mm for 20 min) but not cortical SD in vivo (80 mg/kg, intraperitoneally, 45 min before) (56–58). Clonidine (0.56 mm), noradrenaline (0.1–1 mm) and yohimbine (1.75 mm) all suppressed cortical SD after topical application (55), perhaps suggesting that more drugs could modulate SD if sufficient tissue levels can be attained without dose-limiting side-effects. This is an important consideration for the predictive value of any SD model for drug screening.

Chronic treatment (weeks to months) is required for some prophylactic drugs to suppress SD, in vivo

In all the negative studies with classical prophylactics reviewed above, drugs have been administered either as a single dose or as daily doses for up to a week. However, it is well recognized that migraine prophylactic drugs show a gradual build up of clinical efficacy over a few months, and a possible need for chronic treatment to achieve efficacy on experimental SD has been proposed (47). Indeed, recent data strongly suggest that SD suppression is a mechanism of action shared by several migraine prophylactic drugs after prolonged treatment. Five established prophylactic drugs (valproate, topiramate, propranolol, amitriptyline and methysergide) are now shown to suppress cortical SD susceptibility after daily treatment for 4 weeks or more in rats (37). All five drugs elevated the electrical threshold for SD induction, reduced the frequency of SDs triggered by topical KCl application (Fig. 2), and decreased the SD propagation speed in a dose-dependent manner, without altering SD duration or amplitude. More importantly, longer treatment durations (up to 4 months) produced more potent SD suppression, reminiscent of the need for prolonged treatment with these drugs in migraineurs to reduce attack frequency and/or severity (59, 60). Consistent with most previous reports, single doses of these prophylactic drugs did not alter SD susceptibility or electrophysiological properties when administered 60 min before testing. D-propranolol, deemed clinically ineffective (61, 62), did not suppress SD susceptibility even after prolonged treatment, perhaps demonstrating the specificity and predictive value of the models used in this study.

Figure 2

Representative tracings of extracellular d.c. potential during continuous topical KCl application for 2 h, recorded from two rats treated with intraperitoneal injections of saline or valproate (200 mg kg⁻¹ day⁻¹) for 4 weeks. Seventeen cortical spreading depressions (SDs) are detected at the proximal recording site (E1) in the saline-treated rat; only nine SDs reached the distal electrode (E2), suggesting propagation block between the two electrodes. The number of SDs detected at E1 was nine in the valproate-treated rat, and only two propagated to E2. This experimental model (i.e. frequency of continuous KCl-induced SDs) shows good correlation with electrical threshold of cortical SD (see Fig. 3). The electrode configuration was identical to that shown in Fig. 1.

It is important to note that not too many drugs have been tested as rigorously in SD susceptibility models, and independent confirmation by other laboratories will help build confidence in the predictive ability of these models. Nevertheless, these data suggest that prolonged treatment is a requisite for some migraine prophylactic drugs to suppress SD. Even with drugs that do suppress SD after brief treatment, longer treatment durations may augment their efficacy, as shown for fluoxetine (42). The reasons for the delayed appearance of efficacy on SD are not clear. As noted above, for some drugs it may simply be a matter of achieving therapeutic brain tissue concentrations, in vivo; those drugs are then expected to suppress SD via the same mechanisms when tested at higher concentrations in vitro, or topically in vivo. However, ≥ 4 weeks' treatment may be too long for tissue levels to reach a plateau, and alternative mechanisms should be considered for some drugs. It is known that prolonged treatment with some migraine prophylactic drugs alters gene expression (63). Detailed discussion of individual gene products altered by each drug is beyond the scope of this review, except to note that many of the altered genes are mechanistically related to SD such as those involved in glutamatergic neurotransmission (e.g. tissue glutamate levels), brain water homeostasis (e.g. aquaporin-4), neurotransmitter release, K⁺ channels and gap junctions (64, 65). Therefore, drugs that require weeks to suppress SD susceptibility may do so not directly by interfering with receptors, ion channels or pumps, but by altering gene expression. A potential consequence of altered gene expression, one that may require even longer treatment durations, is a change in tissue morphology or ultrastructure, such as synaptic or mitochondrial density, and even glia/neuron ratios. Such possibilities remain hypothetical at this time, although differences in cortical thickness have been reported in migraineurs (66). Drugs acting exclusively by altering gene expression or ultrastructure after prolonged prophylactic therapy would not be expected to suppress SD acutely even when tested at higher concentrations in vitro, or topically in vivo.

PHARMACOLOGICAL TARGETS MODULATING SPREADING DEPRESSION SUSCEPTIBILITY AND THEIR RELEVANCE IN MIGRAINE PROPHYLAXIS

Migraine drugs that have been shown to suppress SD belong to different pharmacological and therapeutic classes (e.g. anticonvulsant, β-adrenergic blocker, antidepressant, and serotonergic), suggesting that their proximate pharmacological effects converge onto SD. Multiple mechanisms of action have been defined for most migraine prophylactic drugs, and a review of their numerous pharmacological targets exceeds the scope of this review. However, it is important to consider some of the potential drug targets that suppress SD, as well as relevant others that have been ineffective.

Glutamate receptors

Both competitive (e.g. _DL-2-amino-7-phosphonoheptanoic acid) and non-competitive NMDA receptor blockers (e.g. ketamine, phencyclidine, MK-801, CP-101,606) are potent inhibitors of SD; they dose-dependently elevate electrical SD threshold, slow down or block SD propagation, and shorten SD duration (56, 67–76). None of the clinically efficacious migraine drugs (prophylactics in particular) appear to interact ‘directly’ with the NMDA receptor, but may suppress excitatory neurotransmitter release (77). It is worth noting, however, that ketamine and the inhalational analgesic nitrous oxide, both of which inhibit NMDA receptors (78) and suppress SD (56, 70, 71, 79, 80), reportedly show clinical efficacy in aborting migraine aura (81) and migraine headache (82, 83), respectively. NMDA receptor blockers have a number of side-effects that have thus far limited their usefulness in neurotherapeutics, and their role in migraine prophylaxis depends on the development of agents that are clinically better tolerated (e.g. glycine site blockers, NR2B inhibitors). Mg²⁺ augments voltage-dependent block of NMDA receptors, and suppresses SD susceptibility in vitro (8, 84–86). Mg²⁺ showed both abortive and prophylactic efficacy in some but not all trials (87–89). Acute intravenous administration of Mg²⁺ reportedly suppressed KCl-induced SDs in the rat (90); however, whether chronic Mg²⁺ supplementation can elevate brain tissue levels to have an impact on SD is unknown.

Unlike NMDA antagonists, inhibitors of non-NMDA subtypes of ionotropic glutamate receptors have generally been ineffective on SD (67, 91–96), except when SD was triggered by agonists of these receptors or by high-frequency afferent stimulation (94, 95, 97). For this reason, topiramate-induced SD suppression is unlikely to be related to its inhibition of non-NMDA receptor-mediated neurotransmission (98, 99). It should be noted, however, that it is not known how SD is triggered in migrainous human brain, and it may in part involve intense synaptic activity to elevate [K⁺]_e above the SD threshold. If so, inhibitors of non-NMDA subtypes of ionotropic glutamate receptors may still suppress SD in migraineurs.

Some prophylactic drugs may interfere with glutamatergic neurotransmission indirectly by decreasing glutamate release (e.g. amitriptyline, gabapentin). One such drug is lamotrigine, which may be more efficacious on migraine with aura (100–102), but has not been tested on SD.

Voltage-gated Ca²⁺ channels

Non-selective inhibition of voltage-gated Ca²⁺ channels (e.g. by divalent cations Co²⁺, Ni²⁺, Mn²⁺) suppresses SD (16, 47, 95, 103, 104). The relevant channel subtypes appear to be the P/Q- and N-types (95, 105), although in two studies inhibitors of these channels failed to suppress pinprick-induced SD (105, 106). The importance of P/Q-type voltage-gated Ca²⁺ channels in SD susceptibility is further underscored by modulation of SD susceptibility in mouse models expressing mutations in this channel (17, 36, 107). Because P/Q and N-types are the major channels mediating presynaptic Ca²⁺ influx, the mechanism of SD suppression may relate to inhibition of glutamate release (107, 108). Among the drugs with clinical efficacy in migraine, only gabapentin has been shown to interact directly with P/Q-type channels and reduce glutamate release as a major mode of action (109). Whether gabapentin can suppress experimental SD has not yet been tested. Flunarizine, which shows efficacy in SD suppression as well as in migraine prophylaxis, is a relatively non-selective inhibitor of voltage-gated Ca²⁺ channels (110–112). Topiramate also inhibits L-, R- and P/Q-type channels (113–115). Valproate inhibits T-type channels, but there is no evidence at this time to support a role for T-type channels in SD. Both lamotrigine and levetiracetam reportedly suppress N- and P/Q-type Ca²⁺ currents in cortical neurons; however, their efficacy in migraine is not well established (100, 116, 117), and it is not known whether they suppress SD.

Adrenergic receptors

Because of the known efficacy of some β-adrenergic receptor blockers in preventing migraine, a few studies have tested the role of adrenergic receptors on SD. Both noradrenaline (0.1–1 mm topical superfusion), a mixed α- and β-adrenergic receptor agonist, and clonidine (0.56 mm topical superfusion), an α₂-adrenergic receptor agonist, inhibit SD initiation and/or propagation when topically applied in the rat (55). Therefore, inhibition of noradrenaline uptake by amitriptyline may be a relevant mechanism in SD suppression.

Propranolol is a non-selective β-adrenergic receptor blocker with no intrinsic sympathomimetic activity; β-adrenergic receptor blockers that possess partial agonistic activity are ineffective in migraine prophylaxis (118). Propranolol also has potent membrane stabilizing actions; however, β-blockers without this action appear to be equally effective clinically. As mentioned above, propranolol suppresses SD both in vitro and in vivo. Interestingly, chronic treatment with amitriptyline decreases β-adrenergic receptor expression (63), which may be relevant in SD suppression after prolonged administration of this drug in rats (37).

The mechanism of adrenergic modulation of SD may relate to modulation of presynaptic glutamate release. Activation of β-adrenergic receptors augments glutamatergic neurotransmission via increased presynaptic cyclic adenosine-3′,5′-monophosphate levels, and enhanced Ca²⁺ influx possibly through P/Q-type Ca²⁺ channels (119–122). Consistent with this, propranolol inhibits the stimulatory effect of β-adrenergic receptor activation on evoked glutamate release, in vitro (120, 123). Conversely, α₂-adrenergic receptors inhibit depolarization-induced glutamate release (124–126), and may explain SD suppression by clonidine.

Serotonin receptors

5-HT receptors also modulate SD susceptibility. Migraine attacks have been associated with a relatively 5-HT-depleted state, and reserpine, a drug that depletes 5-HT in the brain, can acutely precipitate migraine attacks, although long-term use reduces the attack frequency (127, 128). Consistent with this, 5-HT depletion by tryptophan hydroxylase inhibitor para-chlorophenylalanine augments SD susceptibility (129), while 5-HT reuptake inhibitors amitriptyline, d-fenfluramine, citalopram and fluoxetine suppress cortical SD in the rat, in vivo (37, 42, 53, 130). The 5-HT₂ receptor antagonist and 5-HT₁ receptor agonist methysergide is clinically effective in migraine prophylaxis, and suppresses retinal SD after topical application, in vitro, and cortical SD susceptibility after prolonged systemic treatment, in vivo (37, 131). As with β-adrenergic receptors, chronic amitriptyline treatment reduces 5-HT₂ receptor expression as a potential mechanism of SD suppression (63).

Although in general 5-HT₂ receptors stimulate and 5-HT_1A receptors inhibit neuronal excitation, there are data suggesting that the more relevant receptor subtype in SD suppression is the 5-HT_1A receptor. For example, 5-HT_1A receptor agonist 8-OH-DPAT reduced SD duration in rat cortical slices (67). Topical application of yohimbine, an α₂-antagonist and 5-HT_1A agonist, suppressed CSD in the rat in vivo, and this effect was blocked by 5-HT_1A antagonist NAN-190 (55). NAN-190 also increased the frequency of SDs triggered by topical KCl, suggesting that the endogenous 5-HT exerts an inhibitory effect on SD via this receptor. The mechanisms of SD suppression by 5-HT_1A receptors may relate to the activation of G-protein-coupled K⁺ channels and hyperpolarization, inhibition of adenylate cyclase, and to reduced glutamate release possibly via inhibition of N- and P/Q-type Ca²⁺ channels (67, 132–137).

Gap junction blockers

Although, based on indirect evidence, a role for neuronal synchronization via gap junctions at the onset of SD has been proposed (138), it has been difficult to ascertain the role of astrocytic or neuronal gap junctions in SD, mainly due to the lack of selective inhibitors. In retina and rat hippocampal and neocortical slices, non-selective gap junction blockers (e.g. aliphatic alcohols) block SD propagation at millimolar concentrations (96, 139–142). However, more recent data using the more selective gap junction inhibitor carbenoxolone (100 mm) have shown that the Ca²⁺ waves propagating via astrocytic gap junctions are neither required for SD initiation or propagation, nor necessarily coupled to it, and that gap junction blockade by this agent does not block SD (141, 143). Quinine, quinidine and mefloquine (all 200 mm), which selectively inhibit neuronal connexin 36 junctions among their many other effects, only weakly reduced SD duration and propagation speed in rat neocortical slices (144). Gap junction blockers have not been tested on SD systemically, in vivo, although halothane and isoflurane, both known also to block gap junctions, do suppress cortical SD in anaesthetized rats and cats (79, 80, 145, 146). It appears that SD suppression by these non-specific inhibitors cannot be easily attributed to gap junction blockade, and that at this time there is no solid evidence to support the notion that functional gap junctions are required for SD, in vivo. Nonetheless, the anticonvulsant tonabersat (147), which reportedly blocks gap junctions (118, 148), was shown to suppress cortical SD (43–46). Recent clinical trials of tonabersat in migraine prophylaxis have yielded conflicting results. In one trial, tonabersat appeared to be ineffective in migraine prophylaxis, although both the primary and the secondary end-points tended to favour tonabersat over placebo (149). The negative results of this study, however, should be interpreted with caution because of high placebo response rate and possibly suboptimal drug dosing. The yet unpublished results of another larger trial (Tonabersat Evaluation in Migraine Prevention in the US, TEMPUS) also failed to achieve primary end-points, casting further doubt on both the utility of tonabersat in migraine and the hypothesis that SD is a relevant therapeutic target in migraine. However, another randomized, double-blind, placebo-controlled crossover trial did show a significant reduction in the number of migraine aura attacks both with and without headache, but attacks without aura were not reduced, in keeping with the SD-suppressive effect of tonabersat (150). Clearly, clinical results of tonabersat efficacy in migraine are inconclusive and a larger trial, perhaps with higher doses, is needed.

K⁺ channels

In general, an increase in membrane K⁺ conductance decreases neuronal excitability by virtue of hyperpolarization. Consistent with this, blockade of K⁺ channels (e.g. tetraethylammonium, 4-aminopyridine, Ba²⁺, Cs⁺) has been shown to augment SD susceptibility and even trigger spontaneous SDs (151–154). Conversely, suppression of repetitive KCl-evoked SDs and their propagation speed has been shown acutely after systemic treatment with a K⁺ channel (K_v7) opener with good brain penetration (155). In the absence of studies with selective inhibitors or activators, it is unclear whether these effects are specific for particular channel subtypes or a non-specific consequence of membrane hyperpolarization. Nevertheless, K⁺ channels may be a target for some migraine prophylactic drugs (e.g. valproate).

Sigma-1 receptors

Sigma-1 receptors (σ1R) are unique non-opioid, non-phencyclidine receptors that reside in the endoplasmic reticulum and are mainly expressed in the central nervous system (156). They modulate intracellular Ca²⁺ signalling, K⁺ channels, glutamatergic transmission, and neurotransmitter release and neuronal excitability in a complex manner (157). Several σ1R agonists have been shown to suppress KCl-induced and hypoxic SD, in vitro, and this effect appeared to be independent of NMDA receptors (158). As with the K⁺ channels, future in vivo work with more selective σ1R agonists and antagonists will help clarify the potential role of this receptor in SD suppression, and whether currently available migraine prophylactic drugs that are known to interact with this receptor (e.g. fluoxetine, fluvoxamine) suppress SD by this mechanism.

Tissue pH

Low pH, whether caused by direct acidification, or indirectly by carbonic anhydrase inhibition or hypercapnia, is a potent suppressor of neuronal excitability in general, and SD susceptibility in particular (152, 159–164). Although the efficacy of carbonic anhydrase inhibitor acetazolamide in migraine is unclear (116, 165), there is anecdotal evidence for its benefit in hyperexcitable channelopathies such as FHM (166–168). Topiramate also inhibits carbonic anhydrase and causes metabolic acidosis, a well-recognized side-effect, which may contribute to its SD suppression and migraine prophylactic efficacy (169, 170). Although no correlation between arterial blood pH and the degree of SD suppression was found after chronic topiramate treatment in rats (37), brain tissue pH is more relevant and was not measured in this study.

Na⁺/K⁺ ATPase

Because SD initiation requires elevation of [K⁺]_e above a critical threshold (∼10–12 mm), both resting and stimulated Na⁺/K⁺ ATPase activity are expected to be critical in determining SD occurrence. In support of this, FHM2 is associated with loss-of-function mutations in the α₂ subunit of Na⁺/K⁺ ATPase (171–174), and pharmacological inhibitors of Na⁺/K⁺ ATPase enhance SD susceptibility (162, 175). Although there is no specific activator of Na⁺/K⁺ ATPase, it is worth noting that ketogenic diet suppresses both neuronal excitability and SD (176, 177), and although the precise mechanisms of this effect are unclear, they may be related to enhanced oxidative metabolism and Na⁺/K⁺ ATPase activity (178). Ketogenic diet has not been explored in migraine prophylaxis.

GABA_A receptors

Drugs that augment GABAergic transmission, despite being potent anticonvulsants, do not inhibit SD. For example, diazepam did not alter electrical SD threshold and increased propagation speed and duration (56). Similarly, clonazepam did not inhibit pinprick-induced SD propagation in the cat monitored by laser Doppler flowmetry and single unit activity recordings (52). Barbiturates also did not suppress SD, in vitro or in vivo (16). For example, diethylbarbiturate (0.8 mm) did not block retinal SD, and instead increased propagation speed and amplitude (179). In the rabbit, pentobarbital doses (50 mg/kg, intraperitoneally) sufficient to depress electrocorticogram and breathing failed to inhibit SD, although higher doses apparently increased the tetanic electrical stimulation intensity required to evoke a SD (180). Similar results were obtained in rats (80). Consistent with these, clonazepam was ineffective in migraine prophylaxis (116, 181). Therefore, although topiramate and valproate enhance GABAergic transmission, this does not appear to be a relevant mechanism of action in migraine prophylaxis (182, 183).

Voltage-gated Na⁺ channels

It is generally agreed that voltage-gated Na⁺ channels, while involved in the process, are not critical for SD (94, 184–188). One exception, of course, is when SD is evoked by high-frequency afferent stimulation requiring action potential firing. For example, topical tetrodotoxin superfusion increased the electrical stimulation (20 Hz) threshold of SD, but did not block SD initiation by KCl or its propagation once initiated (189). Conflicting data have recently been reported using pinprick to trigger cortical SD after systemic tetrodotoxin administration, suggesting model differences in sensitivity to pharmacological modulation (106). Inhibition of voltage-gated Na⁺ channels may still be a relevant mechanism of action by some migraine prophylactic drugs (e.g. topiramate, valproate, amitriptyline, and fluoxetine), if intense synaptic activity is a requisite for SD to develop in migraineurs. Recent identification of Na⁺ channel mutations in FHM3 provides further support for a potential role for these channels in SD initiation (190).

Non-pharmacological interventions

Transcranial magnetic stimulation (TMS) has recently attracted attention as a potential migraine therapy. TMS decreased headache intensity when delivered shortly after attack onset (191), but although prophylactic TMS anecdotally reduced attack frequency and intensity (192), a larger, controlled, blinded trial showed no difference from placebo in primary and secondary end-points (193). TMS does alter cortical excitability, and its experimental equivalents in rodents have been shown to modulate SD (194–196). In a recent preliminary report, TMS administered shortly before or immediately after cortical pinprick appeared to acutely suppress SD susceptibility monitored using surrogate blood flow changes (197). However, both the experimental and clinical data are too preliminary to draw conclusions on the utility of TMS to suppress SD or migraine.

In summary, a multitude of direct and indirect pharmacological effects has been described for migraine therapeutics, and there appears to be no single unifying mechanism to explain or predict their efficacy in SD suppression. To complicate the matter further, mechanisms of action of drugs that require prolonged treatment for weeks to months may differ from those that acutely suppress SD after a single dose, underscoring the need to develop experimental SD susceptibility models for high-throughput drug screening.

EXPERIMENTAL MODELS OF SPREADING DEPRESSION FOR DRUG SCREENING

Since its discovery and initial characterization by Leão in anaesthetized rabbits, numerous SD models have been described, each with its own merits and shortcomings. A detailed account of technical aspects of SD models can be found in the volume by Bures et al. (16). Quantitative assessment of SD susceptibility for the purpose of drug screening requires consideration of some practical aspects of experimental models to improve intra- and interobserver reproducibility, as well as the sensitivity and specificity (i.e. predictive value) of the assays.

In vitro vs. in vivo models

In vitro, SD has been studied in retina, brain slices and neuronal cultures. Much of the SD work was done using amphibian and later chick retina models; chick retina forms a natural slice without a need for cutting, with a few well-characterized neuronal layers and no blood vessels. Subsequent in vitro work largely relied upon cortical, hippocampal and cerebellar slices. The advantages of in vitro SD models include bypassing the blood–brain barrier by drug superfusion, and lack of systemic physiological, pharmacokinetic and anaesthetic confounders, allowing high-throughput standardized drug testing as has been done in retina in a few studies. The disadvantages include exposure to trauma and hypoxia during tissue preparation and, more importantly, the difficulty in estimating clinically relevant drug concentrations in vitro. Because of these caveats, pharmacological SD suppression in vitro has not always been predictive of in vivo efficacy.

In vivo SD models have been described in many species from pigeons to cats and monkeys, and there is considerable interspecies variability in SD susceptibility. A great deal of variability also exists among the models used, particularly in the methods used to trigger and detect SD, route of drug administration, type of anaesthetic, and the extent of systemic physiological monitoring. Although most brain regions can sustain SD, in vivo drug testing has been carried out almost exclusively in cerebral cortex. In vivo assays of drug efficacy are expected to be more predictive of clinical efficacy (except when drugs are topically applied on the cortex). However, systemic physiological state, which has so often been neglected, has a significant impact on SD susceptibility. Most importantly, body temperature, plasma glucose, arterial blood gases and pressure, and possibly pH, are important determinants of SD occurrence, duration, refractory period and propagation speed, and need to be routinely monitored throughout the experiments. Drugs with systemic physiological or cerebrovascular effects may indirectly alter SD properties, sometimes without even crossing the blood–brain barrier. For example, hypotension and hypertension may have been the cause of prolonged and shortened SD durations by clonidine and sumatriptan, respectively (30, 47). Blood glucose can be standardized across groups by overnight fasting, which tends to facilitate SD. Anaesthesia is a potential confounder, in vivo. Although certain commonly used anaesthetics can directly impact SD susceptibility, there is no evidence at this time to suggest that the choice of anaesthesia alters the efficacy of a drug to suppress SD susceptibility, except when the anaesthetic and the drug are known to act on the same target (e.g. barbiturate anaesthesia when testing a GABA_A receptor agonist, or ketamine anaesthesia when testing an NMDA receptor antagonist). In general, it is advisable to avoid anaesthetics that strongly suppress SD susceptibility, such as ketamine. While barbiturates, urethane and α-chloralose do not significantly suppress SD, inhalational anaesthetics and nitrous oxide have mild but reproducible inhibitory effects (80).

Methods to trigger SD

Both in vitro and in vivo, the most common and reliable methods to trigger an SD have been topical or intraparenchymal application of KCl, electrical stimulation, and trauma. Although ultimately they all evoke SD by elevating the [K⁺]_e above the critical threshold, the way this is achieved by each stimulus type is highly relevant for drug efficacy.

KCl can be used to assess SD susceptibility in several ways. For example, KCl concentration threshold to trigger an SD (as a surrogate for [K⁺]_e threshold) can be determined by applying escalating concentrations of KCl topically, or intraparenchymally via microdialysis or iontophoresis (107). Alternatively, escalating volumes of KCl solution at a fixed concentration can be applied by intraparenchymal microinjection (18), which probably is also a measure of the minimum critical volume of depolarized tissue to trigger SD. A third approach is to determine the frequency of repetitive SDs triggered by continuous topical application of KCl at a fixed concentration (17, 37, 80); SD frequency increases in a KCl concentration-dependent manner in this model. This method not only measures SD susceptibility, but is also sensitive to the absolute refractory period after an SD (16), which is about 2–3 min depending on the physiological state of the tissue.

Direct electrical stimulation using cathodal currents may be a better method to determine actual SD threshold, but requires careful and reproducible control of stimulus geometry, which is a function of electrode shape and impedance characteristics as well as cortical surface properties (13, 16, 198). The stimulus can be delivered by monopolar or bipolar needle, or cortical cup electrodes; the latter provides better control over stimulus geometry, but all methods give thresholds between 10 and 10³ μC, and can be used as long as the electrode configuration and impedance are consistent when comparing groups. Both single square pulses (typically 10²–10⁴ ms, 10–10³ μA) and high-frequency stimulus trains have been used to evoke SD locally at the stimulation site. Alternatively, a high-frequency stimulus can be delivered to an afferent axonal pathway to evoke SD by intense synaptic activity (i.e. mimicking seizure activity). It is important to note that the mechanisms of [K⁺]_e elevation, and therefore the pharmacological sensitivity profiles, may differ between a single prolonged square pulse and high-frequency tetanic stimulation. For example, the latter is likely to require action potentials and synaptic transmission, and may be suppressed by inhibitors of voltage-gated Na⁺ channels or non-NMDA type ionotropic glutamate receptors. For the same reason, glutamate receptor agonists (e.g. NMDA), which can evoke SD when applied directly onto the tissue (199), are not suitable to trigger SD for drug screening as they yield an obviously different pharmacological sensitivity profile. Because the mechanisms of SD initiation in human brain are not known, the experimental SD induction method most predictive of clinical prophylactic drug efficacy is not clear.

Trauma, usually in the form of pinprick or gentle mechanical pressure onto the tissue, is also a common method to evoke SD. However, unlike the KCl and electrical stimulation, it cannot be easily graded to determine SD thresholds, and may be insensitive to detect small changes in SD susceptibility. The suppression of pinprick-induced SD is often an all-or-none response in a given animal, quantified as percent of animals in which SD is blocked (51). Moreover, cumulative injury may prohibit repetitive SD induction using this method, further limiting its statistical power. In some studies, pinprick-induced SD has shown a decidedly different pharmacological profile compared with KCl-induced SD, in that it was effectively suppressed by inhibitors of voltage-gated Na⁺ channels (and by an inhibitor of non-NMDA subtype of ionotropic glutamate receptors and an agonist of GABA_A receptor in a preliminary report), but not by inhibitors of N- or P/Q-type voltage-gated Ca²⁺ channels (105, 106, 200), further underscoring the importance of identifying the most relevant SD model for drug screening.

Lastly, ischaemia is a potent trigger for SD. For example, endothelin has been used to evoke SD, probably due to vasoconstrictive ischaemia (201). A role for transient ischaemic events as triggers for migraine aura in a subset of patients is further supported by the association between migraine with aura and intracardiac or pulmonary right-to-left shunts (e.g. patent foramen ovale) (202). Paradoxical microembolism as a migraine trigger is a plausible hypothesis yet to be tested, and further work is required to develop experimental models to assess ischaemic SD susceptibility.

Methods to monitor SD

The standard method to detect SD and study its electrophysiological properties is extracellular d.c. potential recordings. This is best accomplished by two or more intraparenchymal glass micropipettes placed serially along the path of propagation, but surface recordings using metal electrodes also suffice, albeit with less precision. The SD attributes most often measured in this way are amplitude and duration (at half maximal amplitude) of the d.c. potential shift, as well as the propagation speed or failure between the two recoding sites. The depression of synaptic activity (e.g. electrocorticogram, or single unit spike activity) can supplement but should not substitute for d.c. potential recording, as they can be suppressed by drugs that have no impact on SD itself.

Alternative methods to detect SD, such as intrinsic optical signal changes reflecting intracellular/extracellular volume shifts, in vitro, or cerebral blood flow changes, in vivo, can supplement the d.c. potential recordings. However, they should not substitute for d.c. potential recording to detect SD, since their mechanisms are poorly understood, and at least in case of cerebral blood flow changes, they can be uncoupled from electrophysiological measures of SD (203). Finally, MRI can be used to detect SD reliably (46, 204, 205), but this method will probably be cost-effective only in larger gyrencephalic animals.

EMPIRICAL CONSIDERATIONS

Developing experimental drug screening models in migraine prophylaxis requires validation, i.e. demonstration of predictive value of the model for clinical efficacy. Ideally in a given model, all drugs known to be effective in migraine prophylaxis should suppress SD, and those tested and found to be ineffective should not. Congruence, however, may be incomplete for several reasons, including differences between experimental animals and humans (e.g. intrinsic brain factors, drug pharmacokinetics), sensitivity and specificity of the employed experimental models, and importantly, the quality of clinical trials of migraine prophylaxis. On the latter point, only a handful of drugs have been conclusively tested for their clinical efficacy in migraine prophylaxis, and lack of efficacy has been particularly difficult to ascertain (118, 206). All these underscore the need to characterize and validate relevant and predictive preclinical SD models and attributes to use as end-points for drug development in migraine. Some broad outlines of a drug screening paradigm can be empirically drawn to seek validation of the models initially, and then to test candidate drugs. Far from having been validated, however, the outlines below are intended to be recommendations rather than guidelines, and further work will undoubtedly help highlight some models among others towards that goal.

Experimental models

Despite the advantages of in vitro models mentioned above, drug screening ultimately needs to demonstrate efficacy in vivo. Because different types of stimuli to evoke an SD have yielded different pharmacological profiles for its suppression, it is critical to identify the method that is most relevant for clinical efficacy. In our experience, quantitative assessment of SD susceptibility, a poorly defined term in itself, is best achieved using KCl or electrical stimulation, because they both allow precise titration of stimulus intensity, and because SD susceptibility measured using these stimuli was sensitive to suppression by clinically effective migraine prophylactic drugs, suggesting good predictive value (37). Regardless of the trigger, however, it is generally agreed that SD propagating into adjacent regions acquires a rather uniform character, provided that the stimulus intensity is comparable. Hence, the best method to monitor SD is less ambiguous, since the characteristic d.c. potential shift accompanied by depression of cortical activity is historically the gold standard. Last but not least, careful choice of anaesthesia and systemic physiological monitoring are critical for proper data interpretation.

Relevant SD attributes

The most direct attribute of SD susceptibility is probably electrical or KCl threshold for triggering an SD. One has to keep in mind that the experimental models detecting SD at a distance assess not only SD initiation but also successful spread to the recording site. Because the ionic shifts sustaining the momentum of spread may be weaker or more prone to inhibition than the external stimulus used to initiate the SD, drugs may preferentially slow down and block propagation rather than initiation of SD. Based on our own experience, this is often the case especially when suprathreshold stimuli are used (e.g. 1 m KCl, or pinprick), so that the farther apart the stimulation and recording sites are, the stronger the apparent SD suppression (Fig. 2). Therefore, the speed of SD propagation, or its failure, are also important attributes for drug screening. Changes in other SD attributes, the duration and particularly the amplitude, are harder to interpret as they do not always correspond to SD susceptibility. The frequency of repetitive SDs triggered by continuous topical KCl was also a surprisingly reproducible and reliable indicator of SD susceptibility (17, 37, 72, 80, 90, 105, 145), and correlated well with the electrical SD threshold when both were tested in the same animal (Fig. 3). Corroborating multiple SD attributes that can be measured in the same animal would probably enhance the predictive value of any model for drug screening.

Figure 3

Graph showing the inverse relationship between the electrical threshold and the frequency of spreading depressions (SDs) evoked by continuous topical KCl application, when these two were measured in the same animal (n = 65 rats). The data are pooled vehicle and drug treatment groups from Ayata et al. (37). Vertical axis, mean ±s.d.; horizontal axis, median electrical threshold. Spearman rank order correlation test.

Experimental animals

SD susceptibility varies among species (e.g. lissencephalic brains are generally more susceptible than gyrencephalic brains), often attributed to glia/neuron ratios (164, 207). Moreover, it is not known whether potential drug targets (e.g. gap junctions, voltage-gated Ca²⁺ channels, NMDA receptors) contribute to SD occurrence and propagation to the same extent in different species. Therefore, initial high-throughput studies in more cost-effective and accessible rodents should ideally be followed by demonstration of efficacy in gyrencephalic species (e.g. cat). In addition, some drugs may show efficacy or lose their efficacy in the presence of certain genetic risk factors (e.g. P/Q-type Ca²⁺ channel inhibitors in FHM1 with mutations in this channel); it might therefore be of interest to test drug candidates in genetic mouse models of migraine as well, as they become available. Because migraine shows a female preponderance, it will be important to screen drugs in both sexes.

Treatment dose and protocol

In general, drug doses used in rodents are approximately an order of magnitude higher than those used in humans, mainly due to faster rates of metabolism and elimination. However, it is difficult to predict the clinically relevant animal doses for SD suppression for a given drug. Measurement of plasma levels may help guide dose selection, when this information is available from human studies. Further complicating the matter, brain tissue levels may not always correlate with plasma levels, and may show gradual build-up upon prolonged treatment as a requisite for efficacy in migraine prophylaxis. Even when target plasma levels are achieved, prolonged treatment may be necessary for efficacy to become manifest, possibly via altered gene expression and/or ultrastructural changes. Therefore, a thorough drug screening should seek a dose–response relationship, and start by testing the acute effects of a single administration, followed by chronic treatment in search of increased efficacy and/or reduced dose requirements.

CONCLUDING REMARKS

Although there is little doubt that SD is the electrophysiological substrate of migraine aura, whether it is the inciting event for migraine headache is as yet unresolved. If SD is indeed the trigger for migraine headache, SD inhibitors will probably be useful in prophylaxis rather than abortive treatment, because the brief and episodic nature of SD (i.e. aura) is unlikely to provide a wide enough window of opportunity to intervene pharmacologically with the process before pain mechanisms are activated. Exceptionally, if an attack can be anticipated (e.g. exposure to known triggers, hormonal fluctuations), then a drug that acutely suppresses SD after a single or few doses can be used as a ‘pre-emptive abortive’ therapy (i.e. episodic rather than chronic treatment). On a more speculative note, prophylactic suppression of SD susceptibility is expected to reduce migraine attack frequency by preventing SD occurrence. In addition, headache severity might be reduced by limiting the duration or distance of propagation of breakthrough cortical SDs (i.e. propagation block), although there is yet no evidence that these attributes correlate with headache intensity.

It is possible that multiple yet unidentified genetic factors, modulated by hormonal and environmental influences, determine SD susceptibility and response to different prophylactic drugs in migraineurs. Identification of biomarkers for SD susceptibility (e.g. electrophysiological, genetic) that have better sensitivity and specificity than symptomatic aura might therefore help predict the response to anti-SD therapy and tailor patient selection.

Since the earlier attempts exploring SD as a target, it has taken more than two decades to develop the concept and put it to use prospectively as a drug-screening paradigm for migraine. Only a handful of prophylactic drugs have been vigorously tested in relevant SD models. Tonabersat may be considered the first migraine drug candidate that is being tested in clinical trials, in part based on its efficacy to suppress experimental SD. Although its precise mechanism of action is unknown and clinical efficacy remains to be proven, the momentum needs to be carried forward by screening new drugs and candidates in standardized SD models. It is equally important, however, to test drugs that are conclusively ineffective in migraine prophylaxis to select and refine relevant SD models and optimize their predictive value. In this respect, the quality of clinical trial design, conduct, analysis and reporting will be critical (206).

References

1 Lashley

. Patterns of cerebral integration indicated by the scotomas of migraine. Arch Neurol Psychiatry 1941; 46:331–9.

2 Leao

AAP

. Spreading depression of activity in cerebral cortex. J Neurophysiol 1944; 7:359–90.

3 Leao

AAP

Morison

. Propagation of spreading cortical depression. J Neurophysiol 1945; 8:33–45.

4 Milner

. Note on a possible correspondence between the scotomas of migraine and spreading depression of Leao. Electromyogr Clin Neurophysiol 1958; 10:705.

5 Lauritzen

. On the possible relation of spreading cortical depression to classical migraine. Cephalalgia 1985; 5 (Suppl. 2):47–51.

Strong

Fabricius

Boutelle

Hibbins

Hopwood

Jones

Spreading and synchronous depressions of cortical activity in acutely injured human brain. Stroke 2002; 33:2738–43.

7 Rodgers

Armstrong

Shoemaker

Labrie

Moyes

Robertson

. Stress preconditioning of spreading depression in the locust CNS. PLoS ONE 2007; 2:e1366.

8 Lauritzen

Rice

Okada

Nicholson

. Quisqualate, kainate and NMDA can initiate spreading depression in the turtle cerebellum. Brain Res 1988; 475:317–27.

9 Van Harreveld

Stamm

Christensen

. Spreading depression in rabbit, cat and monkey. Am J Physiol 1956; 184:312–20.

10.

10 Somjen

. Mechanisms of spreading depression and hypoxic spreading depression-like depolarization. Physiol Rev 2001; 81:1065–96.

11.

11 Phillips

Nicholson

. Anion permeability in spreading depression investigated with ion-sensitive microelectrodes. Brain Res 1979; 173:567–71.

12.

12 Matsuura

Bures

. The minimum volume of depolarized neural tissue required for triggering cortical spreading depression in rat. Exp Brain Res 1971; 12:238–49.

13.

13 Reid

Marrannes

Wauquier

. Spreading depression and central nervous system pharmacology. J Pharmacol Methods 1988; 19:1–21.

14.

14 Nicholson

Kraig

. The behaviour of extracellular ions during spreading depression. In: Zeuthen

, ed. The application of ion-selective microelectrodes. Amsterdam: Elsevier 1981; 217–38.

15.

15 Hansen

Lauritzen

Tfelt-Hansen

. Spreading cortical depression and antimigraine drugs. In: Amery

Van Nueten

Waquier

, eds. The pharmacological basis of migraine therapy. London: Putman 1984; 161–70.

16.

16 Bures

Buresova

Krivanek

. The mechanism and applications of Leao's spreading depression of electroencephalographic activity. New York: Academic Press 1974.

17.

Eikermann-Haerter

Dilekoz

Kudo

Savitz

Waeber

Baum

Genetic and hormonal factors modulate spreading depression and transient hemiparesis in mouse models of familial hemiplegic migraine type 1. J Clin Invest 2009; 119:99–109.

18.

18 Brennan

Romero Reyes

Lopez Valdes

Arnold

Charles

. Reduced threshold for cortical spreading depression in female mice. Ann Neurol 2007; 61:603–6.

19.

19 Richter

Bauer

Lehmenkuhler

Schaible

. Spreading depression in the brainstem of the adult rat: electrophysiological parameters and influences on regional brainstem blood flow. J Cereb Blood Flow Metab 2008; 28:984–94.

20.

20 Richter

Rupprecht

Lehmenkuhler

Schaible

. Spreading depression can be elicited in brainstem of immature but not adult rats. J Neurophysiol 2003; 90:2163–70.

21.

21 Fifkova

. Spreading depression in subcortical structures in rabbit. Brain Res 1966; 2:61–70.

22.

22 Fifkova

Bures

. Spreading depression in the mammalian striatum. Arch Int Physiol Biochim 1964; 72:171–9.

23.

23 Fifkova

. Thalamic spreading depression in the rat. Electromyogr Clin Neurophysiol 1966; 20:68–76.

24.

24 Vinogradova

Koroleva

Bures

. Re-entry waves of Leao's spreading depression between neocortex and caudate nucleus. Brain Res 1991; 538:161–4.

25.

Hadjikhani

Sanchez Del Rio

Schwartz

Bakker

Fischl

Mechanisms of migraine aura revealed by functional MRI in human visual cortex. Proc Natl Acad Sci USA 2001; 98:4687–92.

26.

26 Lauritzen

Skyhoj Olsen

Lassen

Paulson

. Changes in regional cerebral blood flow during the course of classic migraine attacks. Ann Neurol 1983; 13:633–41.

27.

27 Olesen

. Cerebral and extracranial circulatory disturbances in migraine: pathophysiological implications. Cerebrovasc Brain Metab Rev 1991; 3:1–28.

28.

28 Lauritzen

Olsen

Lassen

Paulson

. Regulation of regional cerebral blood flow during and between migraine attacks. Ann Neurol 1983; 14:569–72.

29.

29 Olesen

Larsen

Lauritzen

. Focal hyperemia followed by spreading oligemia and impaired activation of rCBF in classic migraine. Ann Neurol 1981; 9:344–52.

30.

30 Moskowitz

Nozaki

Kraig

. Neocortical spreading depression provokes the expression of c-fos protein-like immunoreactivity within trigeminal nucleus caudalis via trigeminovascular mechanisms. J Neurosci 1993; 13:1167–77.

31.

31 Bolay

Reuter

Dunn

Huang

Boas

Moskowitz

. Intrinsic brain activity triggers trigeminal meningeal afferents in a migraine model. Nat Med 2002; 8:136–42.

32.

32 Buzzi

Bonamini

Moskowitz

. Neurogenic model of migraine. Cephalalgia 1995; 15:277–80.

33.

33 Wahl

Schilling

Parsons

Kaumann

. Involvement of calcitonin gene-related peptide (CGRP) and nitric oxide (NO) in the pial artery dilatation elicited by cortical spreading depression. Brain Res 1994; 637:204–10.

34.

34 Read

Smith

Hunter

Parsons

. The dynamics of nitric oxide release measured directly and in real time following repeated waves of cortical spreading depression in the anaesthetised cat. Neurosci Lett 1997; 232:127–30.

35.

Reuter

Weber

Gold

Arnold

Wolf

Dreier

Perivascular nerves contribute to cortical spreading depression-associated hyperemia in rats. Am J Physiol 1998; 274:H1979–87.

36.

van den Maagdenberg

Pietrobon

Pizzorusso

Kaja

Broos

Cesetti

A Cacna1a knockin migraine mouse model with increased susceptibility to cortical spreading depression. Neuron 2004; 41:701–10.

37.

37 Ayata

Jin

Kudo

Dalkara

Moskowitz

. Suppression of cortical spreading depression in migraine prophylaxis. Ann Neurol 2006; 59:652–61.

38.

Dreier

Woitzik

Fabricius

Bhatia

Major

Drenckhahn

Delayed ischaemic neurological deficits after subarachnoid haemorrhage are associated with clusters of spreading depolarizations. Brain 2006; 129:3224–37.

39.

39 Peixoto

Fernandes de Lima

Hanke

. Correlation of the electrical and intrinsic optical signals in the chicken spreading depression phenomenon. Neurosci Lett 2001; 299:89–92.

40.

40 Wiedemann

de Lima

Hanke

. Effects of antimigraine drugs on retinal spreading depression. Naunyn Schmiedebergs Arch Pharmacol 1996; 353:552–6.

41.

41 Maranhao-Filho

Martins-Ferreira

Vincent

Ribeiro

Novis

. Sumatriptan blocks spreading depression in isolated chick retina. Cephalalgia 1997; 17:822–5.

42.

42 dos Santos

Pinheiro

de Lima

Ozias

de Oliveira

Guimaraes

Guedes

. Fluoxetine inhibits cortical spreading depression in weaned and adult rats suckled under favorable and unfavorable lactation conditions. Exp Neurol 2006; 200:275–82.

43.

43 Read

Smith

Hunter

Upton

Parsons

. SB-220453, a potential novel antimigraine agent, inhibits nitric oxide release following induction of cortical spreading depression in the anaesthetized cat. Cephalalgia 2000; 20:92–9.

44.

44 Smith

Read

Chan

Thompson

Hunter

Upton

Parsons

. Repetitive cortical spreading depression in a gyrencephalic feline brain: inhibition by the novel benzoylamino-benzopyran SB-220453. Cephalalgia 2000; 20:546–53.

45.

45 Read

Hirst

Upton

Parsons

. Cortical spreading depression produces increased cGMP levels in cortex and brain stem that is inhibited by tonabersat (SB-220453) but not sumatriptan. Brain Res 2001; 891:69–77.

46.

Bradley

Smith

Netsiri

Smith

Bockhorst

Hall

Diffusion-weighted MRI used to detect in vivo modulation of cortical spreading depression: comparison of sumatriptan and tonabersat. Exp Neurol 2001; 172:342–53.

47.

47 Marrannes

Wauquier

Reid

De Prins

. Effects of drugs on cortical spreading depression. In: Amery

Wauquier

, eds. Prefude to the Migraine Attack. London: Bailliere Tindall 1986; 158–73.

48.

48 Reid

Marrannes

Wauquier

. Effects of flunarizine on rat cerebral cortex. Drug Development Res 1987; 10:107–16.

49.

49 Wauquier

Ashton

Marrannes

. The effects of flunarizine in experimental models related to the pathogenesis of migraine. Cephalalgia 1985; 5 (Suppl. 2):119–23.

50.

50 Marrannes

De Prins

Fransen

Clincke

. Neuropharmacology of spreading depression. In: Lehmenkuhler

Grotemeyer

Tegtmeier

, eds. Migraine: basic mechanisms and treatment. Baltimore: Urban & Schwarzenberg 1993; 431–43.

51.

51 Akerman

Goadsby

. Topiramate inhibits cortical spreading depression in rat and cat: impact in migraine aura. Neuroreport 2005; 16:1383–7.

52.

52 Kaube

Goadsby

. Anti-migraine compounds fail to modulate the propagation of cortical spreading depression in the cat. Eur Neurol 1994; 34:30–5.

53.

53 Guedes

Amancio-Dos-Santos

Manhaes-De-Castro

Costa-Cruz

. Citalopram has an antagonistic action on cortical spreading depression in well-nourished and early-malnourished adult rats. Nutr Neurosci 2002; 5:115–23.

54.

54 Read

Parsons

. Sumatriptan modifies cortical free radical release during cortical spreading depression. A novel antimigraine action for sumatriptan? Brain Res 2000; 870:44–53.

55.

55 Richter

Mikulik

Ebersberger

Schaible

. Noradrenergic agonists and antagonists influence migration of cortical spreading depression in rat—a possible mechanism of migraine prophylaxis and prevention of postischemic neuronal damage. J Cereb Blood Flow Metab 2005; 25:1225–35.

56.

56 Marrannes

Willems

De Prins

Wauquier

. Evidence for a role of the N-methyl-D-aspartate (NMDA) receptor in cortical spreading depression in the rat. Brain Res 1988; 457:226–40.

57.

57 Chebabo

Do-Carmo

Martins-Ferreira

. The effect of diphenylhydantoin on spreading depression. Braz J Med Biol Res 1988; 21:603–5.

58.

58 Chebabo

Do Carmo

. Phenytoin and retinal spreading depression. Brain Res 1991; 551:16–9.

59.

59 Silberstein

Neto

Schmitt

Jacobs

. Topiramate in migraine prevention: results of a large controlled trial. Arch Neurol 2004; 61:490–5.

60.

60 Silberstein

Collins

. Safety of divalproex sodium in migraine prophylaxis: an open-label, long-term study. Long-term Safety of Depakote in Headache Prophylaxis Study Group. Headache 1999; 39:633–43.

61.

61 Stensrud

Sjaastad

. Short-term clinical trial of phopranolol in racemic form (Inderal), D-propranolol and placebo in migraine. Acta Neurol Scand 1976; 53:229–32.

62.

62 Tfelt-Hansen

. Efficacy of beta-blockers in migraine. A critical review. Cephalalgia 1986; 6 (Suppl. 5):15–24.

63.

63 Peroutka

Snyder

. Regulation of serotonin2 (5-HT2) receptors labeled with [3H]spiroperidol by chronic treatment with the antidepressant amitriptyline. J Pharmacol Exp Ther 1980; 215:582–7.

64.

64 O'Donnell

Rotzinger

Ulrich

Hanstock

Nakashima

Silverstone

. Effects of chronic lithium and sodium valproate on concentrations of brain amino acids. Eur Neuropsychopharmacol 2003; 13:220–7.

65.

65 Bosetti

Bell

Manickam

. Microarray analysis of rat brain gene expression after chronic administration of sodium valproate. Brain Res Bull 2005; 65:331–8.

66.

66 DaSilva

Granziera

Snyder

Hadjikhani

. Thickening in the somatosensory cortex of patients with migraine. Neurology 2007; 69:1990–5.

67.

67 Kruger

Heinemann

Luhmann

. Effects of ionotropic glutamate receptor blockade and 5-HT1A receptor activation on spreading depression in rat neocortical slices. Neuroreport 1999; 10:2651–6.

68.

68 Martin

Warner

Todd

. Effects of glycine receptor antagonism on spreading depression in the rat. Neurosci Lett 1994; 180:285–9.

69.

69 Amemori

Bures

. Ketamine blockade of spreading depression: rapid development of tolerance. Brain Res 1990; 519:351–4.

70.

70 Hernandez-Caceres

Macias-Gonzalez

Brozek

Bures

. Systemic ketamine blocks cortical spreading depression but does not delay the onset of terminal anoxic depolarization in rats. Brain Res 1987; 437:360–4.

71.

71 Gorelova

Koroleva

Amemori

Pavlik

Bures

. Ketamine blockade of cortical spreading depression in rats. Electromyogr Clin Neurophysiol 1987; 66:440–7.

72.

72 Peeters

Gunthorpe

Strijbos

Goldsmith

Upton

James

. Effects of pan- and subtype-selective N-methyl-D-aspartate receptor antagonists on cortical spreading depression in the rat: therapeutic potential for migraine. J Pharmacol Exp Ther 2007; 321:564–72.

73.

73 Faria

Mody

. Protective effect of ifenprodil against spreading depression in the mouse entorhinal cortex. J Neurophysiol 2004; 92:2610–4.

74.

74 Obrenovitch

Zilkha

. Inhibition of cortical spreading depression by L-701,324, a novel antagonist at the glycine site of the N-methyl-D-aspartate receptor complex. Br J Pharmacol 1996; 117:931–7.

75.

75 Menniti

Pagnozzi

Butler

Chenard

Jaw-Tsai

Frost White

. CP-101,606, an NR2B subunit selective NMDA receptor antagonist, inhibits NMDA and injury induced c-fos expression and cortical spreading depression in rodents. Neuropharmacology 2000; 39:1147–55.

76.

76 Willette

Lysko

Sauermelch

. A comparison of (+)SK & F 10047 and MK-801 on cortical spreading depression. Brain Res 1994; 648:347–51.

77.

77 Ishii

Sumi

. Amitriptyline inhibits striatal efflux of neurotransmitters via blockade of voltage-dependent Na+ channels. Eur J Pharmacol 1992; 221:377–80.

78.

Jevtovic-Todorovic

Todorovic

Mennerick

Powell

Dikranian

Benshoff

Nitrous oxide (laughing gas) is an NMDA antagonist, neuroprotectant and neurotoxin. Nat Med 1998; 4:460–3.

79.

79 Saito

Graf

Hubel

Taguchi

Rosner

Fujita

Heiss

. Halothane, but not alpha-chloralose, blocks potassium-evoked cortical spreading depression in cats. Brain Res 1995; 699:109–15.

80.

80 Kudo

Nozari

Moskowitz

Ayata

. The impact of anesthetics and hyperoxia on cortical spreading depression. Exp Neurol 2008; 212:201–6.

81.

81 Kaube

Herzog

Kaufer

Dichgans

Diener

. Aura in some patients with familial hemiplegic migraine can be stopped by intranasal ketamine. Neurology 2000; 55:139–41.

82.

82 Stanley

. Migraine treated by inhalation sedation using nitrous oxide and oxygen. Br Dent J 1980; 149:54.

83.

83 Triner

Bartfield

Birdwell

Raccio-Robak

. Nitrous oxide for the treatment of acute migraine headache. Am J Emerg Med 1999; 17:252–4.

84.

84 Van Harreveld

. The nature of the chick's magnesium-sensitive retinal spreading depression. J Neurobiol 1984; 15:333–43.

85.

85 Martins-Ferreira

do Carmo

. Retinal spreading depression and the extracellular milieu. Can J Physiol Pharmacol 1987; 65:1092–8.

86.

86 Mody

Lambert

Heinemann

. Low extracellular magnesium induces epileptiform activity and spreading depression in rat hippocampal slices. J Neurophysiol 1987; 57:869–88.

87.

Pfaffenrath

Wessely

Meyer

Isler

Evers

Grotemeyer

Magnesium in the prophylaxis of migraine—a double-blind placebo-controlled study. Cephalalgia 1996; 16:436–40.

88.

88 Peikert

Wilimzig

Kohne-Volland

. Prophylaxis of migraine with oral magnesium: results from a prospective, multi-center, placebo-controlled and double-blind randomized study. Cephalalgia 1996; 16:257–63.

89.

89 Bigal

Bordini

Tepper

Speciali

. Intravenous magnesium sulphate in the acute treatment of migraine without aura and migraine with aura. A randomized, double-blind, placebo-controlled study. Cephalalgia 2002; 22:345–53.

90.

90 van der Hel

van den Bergh

Nicolay

Tulleken

Dijkhuizen

. Suppression of cortical spreading depressions after magnesium treatment in the rat. Neuroreport 1998; 9:2179–82.

91.

91 Nellgard

Wieloch

. NMDA-receptor blockers but not NBQX, an AMPA-receptor antagonist, inhibit spreading depression in the rat brain. Acta Physiol Scand 1992; 146:497–503.

92.

92 Lauritzen

Hansen

. The effect of glutamate receptor blockade on anoxic depolarization and cortical spreading depression. J Cereb Blood Flow Metab 1992; 12:223–9.

93.

93 Christensen

Jorgensen

Diemer

. Impairment of Fos protein formation in the rat infarct borderzone by MK-801, but not by NBQX. Acta Neurol Scand 1993; 87:510–15.

94.

94 Sheardown

. The triggering of spreading depression in the chicken retina: a pharmacological study. Brain Res 1993; 607:189–94.

95.

95 Kunkler

Kraig

. P/Q Ca2+ channel blockade stops spreading depression and related pyramidal neuronal Ca2+ rise in hippocampal organ culture. Hippocampus 2004; 14:356–67.

96.

96 Nedergaard

Cooper

Goldman

. Gap junctions are required for the propagation of spreading depression. J Neurobiol 1995; 28:433–44.

97.

97 Kertesz

Kapus

Levay

. Interactions of allosteric modulators of AMPA/kainate receptors on spreading depression in the chicken retina. Brain Res 2004; 1025:123–9.

98.

98 Poulsen

Simeone

Maar

Smith-Swintosky

White

Schousboe

. Modulation by topiramate of AMPA and kainate mediated calcium influx in cultured cerebral cortical, hippocampal and cerebellar neurons. Neurochem Res 2004; 29:275–82.

99.

99 Qian

Noebels

. Topiramate alters excitatory synaptic transmission in mouse hippocampus. Epilepsy Res 2003; 55:225–33.

100.

100 Lampl

Katsarava

Diener

Limmroth

. Lamotrigine reduces migraine aura and migraine attacks in patients with migraine with aura. J Neurol Neurosurg Psychiatry 2005; 76:1730–2.

101.

101 Steiner

Findley

Yuen

. Lamotrigine versus placebo in the prophylaxis of migraine with and without aura. Cephalalgia 1997; 17:109–12.

102.

102 D'Andrea

Granella

Cadaldini

Manzoni

. Effectiveness of lamotrigine in the prophylaxis of migraine with aura: an open pilot study. Cephalalgia 1999; 19:64–6.

103.

103 Jing

Aitken

Somjen

. Role of calcium channels in spreading depression in rat hippocampal slices. Brain Res 1993; 604:251–9.

104.

104 Footitt

Newberry

. Cortical spreading depression induces an LTP-like effect in rat neocortex in vitro . Brain Res 1998; 781:339–42.

105.

105 Richter

Ebersberger

Schaible

. Blockade of voltage-gated calcium channels in rat inhibits repetitive cortical spreading depression. Neurosci Lett 2002; 334:123–6.

106.

106 Akerman

Holland

Goadsby

. Mechanically-induced cortical spreading depression associated regional cerebral blood flow changes are blocked by Na+ ion channel blockade. Brain Res 2008; 1229:27–36.

107.

107 Ayata

Shimizu-Sasamata

Noebels

Moskowitz

. Impaired neurotransmitter release and elevated threshold for cortical spreading depression in mice with mutations in the alpha1A subunit of P/Q type calcium channels. Neuroscience 2000; 95:639–45.

108.

108

Tottene

Conti

Fabbro

Vecchia

Shapovalova

Santello

Enhanced excitatory transmission at cortical synapses as the basis for facilitated spreading depression in Ca(v)2.1 knockin migraine mice. Neuron 2009; 61:762–73.

109.

109 Dooley

Taylor

Donevan

Feltner

. Ca2+ channel alpha2delta ligands: novel modulators of neurotransmission. Trends Pharmacol Sci 2007; 28:75–82.

110.

110 Shin

Kim

Min

Ogawa

Tanaka

Akaike

. A selective T-type Ca2+ channel blocker R(-) efonidipine. Naunyn Schmiedebergs Arch Pharmacol 2008; 377:411–21.

111.

111 Takahashi

Akaike

. Calcium antagonist effects on low-threshold (T-type) calcium current in rat isolated hippocampal CA1 pyramidal neurons. J Pharmacol Exp Ther 1991; 256:169–75.

112.

112 Chronicle

Mulleners

. Anticonvulsant drugs for migraine prophylaxis. Cochrane Database Syst Rev 2004; (3):CD003226.

113.

113 Kuzmiski

Barr

Zamponi

MacVicar

. Topiramate inhibits the initiation of plateau potentials in CA1 neurons by depressing R-type calcium channels. Epilepsia 2005; 46:481–9.

114.

114 Martella

Costa

Pisani

Cupini

Bernardi

Calabresi

. Antiepileptic drugs on calcium currents recorded from cortical and PAG neurons: therapeutic implications for migraine. Cephalalgia 2008; 28:1315–26.

115.

115 Zhang

Velumian

Jones

Carlen

. Modulation of high-voltage-activated calcium channels in dentate granule cells by topiramate. Epilepsia 2000; 41 (Suppl. 1):S52–60.

116.

116 Mulleners

Chronicle

. Anticonvulsants in migraine prophylaxis: a Cochrane review. Cephalalgia 2008; 28:585–97.

117.

117 Brighina

Palermo

Aloisio

Francolini

Giglia

Fierro

. Levetiracetam in the prophylaxis of migraine with aura: a 6-month open-label study. Clin Neuropharmacol 2006; 29:338–42.

118.

118 Silberstein

. Preventive treatment of migraine. Trends Pharmacol Sci 2006; 27:410–15.

119.

119 Huang

Hsu

Gean

. Isoproterenol potentiates synaptic transmission primarily by enhancing presynaptic calcium influx via P- and/or Q-type calcium channels in the rat amygdala. J Neurosci 1996; 16:1026–33.

120.

120 Wang

Coutinho

Sihra

. Presynaptic cross-talk of beta-adrenoreceptor and 5-hydroxytryptamine receptor signalling in the modulation of glutamate release from cerebrocortical nerve terminals. Br J Pharmacol 2002; 137:1371–9.

121.

121 Gereau

RWt

Conn

. Presynaptic enhancement of excitatory synaptic transmission by beta-adrenergic receptor activation. J Neurophysiol 1994; 72:1438–42.

122.

122 Herrero

Sanchez-Prieto

. cAMP-dependent facilitation of glutamate release by beta-adrenergic receptors in cerebrocortical nerve terminals. J Biol Chem 1996; 271:30554–60.

123.

123 Lynch

Bliss

. Noradrenaline modulates the release of [14C]glutamate from dentate but not from CA1/CA3 slices of rat hippocampus. Neuropharmacology 1986; 25:493–8.

124.

124 Kamisaki

Hamada

Maeda

Ishimura

Itoh

. Presynaptic alpha 2 adrenoceptors inhibit glutamate release from rat spinal cord synaptosomes. J Neurochem 1993; 60:522–6.

125.

125 Kamisaki

Hamahashi

Hamada

Maeda

Itoh

. Presynaptic inhibition by clonidine of neurotransmitter amino acid release in various brain regions. Eur J Pharmacol 1992; 217:57–63.

126.

126 Kamisaki

Hamahashi

Okada

Itoh

. Clonidine inhibition of potassium-evoked release of glutamate and aspartate from rat cortical synaptosomes. Brain Res 1991; 568:193–8.

127.

127 Ferrari

Odink

Tapparelli

Van Kempen

Pennings

Bruyn

. Serotonin metabolism in migraine. Neurology 1989; 39:1239–42.

128.

128 Ferrari

. Biochemistry of migraine. Pathol Biol (Paris) 1992; 40:287–92.

129.

129 Supornsilpchai

Sanguanrangsirikul

Maneesri

Srikiatkhachorn

. Serotonin depletion, cortical spreading depression, and trigeminal nociception. Headache 2006; 46:34–9.

130.

130 Cabral-Filho

Trindade-Filho

Guedes

. Effect of d-fenfluramine on cortical spreading depression in rats. Braz J Med Biol Res 1995; 28:347–50.

131.

131 Silberstein

. Methysergide. Cephalalgia 1998; 18:421–35.

132.

132 Sun

Dale

. Differential inhibition of N and P/Q Ca2+ currents by 5-HT1A and 5-HT1D receptors in spinal neurons of Xenopus larvae. J Physiol 1998; 510:103–20.

133.

133 Sun

Dale

. G-proteins are involved in 5-HT receptor-mediated modulation of N- and P/Q- but not T-type Ca2+ channels. J Neurosci 1999; 19:890–9.

134.

134 Czyrak

Czepiel

Mackowiak

Chocyk

Wedzony

. Serotonin 5-HT1A receptors might control the output of cortical glutamatergic neurons in rat cingulate cortex. Brain Res 2003; 989:42–51.

135.

135 Schmitz

Empson

Heinemann

. Serotonin and 8-OH-DPAT reduce excitatory transmission in rat hippocampal area CA1 via reduction in presumed presynaptic Ca2+ entry. Brain Res 1995; 701:249–54.

136.

136 Andrade

Malenka

Nicoll

RAA

. G protein couples serotonin and GABAB receptors to the same channels in hippocampus. Science 1986; 234:1261–5.

137.

137 Schmitz

Empson

Gloveli

Heinemann

. Serotonin reduces synaptic excitation of principal cells in the superficial layers of rat hippocampal-entorhinal cortex combined slices. Neurosci Lett 1995; 190:37–40.

138.

138 Herreras

Largo

Ibarz

Somjen

Martin del Rio

. Role of neuronal synchronizing mechanisms in the propagation of spreading depression in the in vivo hippocampus. J Neurosci 1994; 14:7087–98.

139.

139 Martins-Ferreira

Ribeiro

. Biphasic effects of gap junctional uncoupling agents on the propagation of retinal spreading depression. Braz J Med Biol Res 1995; 28:991–4.

140.

140 Largo

Tombaugh

Aitken

Herreras

Somjen

. Heptanol but not fluoroacetate prevents the propagation of spreading depression in rat hippocampal slices. J Neurophysiol 1997; 77:9–16.

141.

141 Vilagi

Klapka

Luhmann

. Optical recording of spreading depression in rat neocortical slices. Brain Res 2001; 898:288–96.

142.

142 Aitken

Tombaugh

Turner

Somjen

. Similar propagation of SD and hypoxic SD-like depolarization in rat hippocampus recorded optically and electrically. J Neurophysiol 1998; 80:1514–21.

143.

143 Peters

Schipke

Hashimoto

Kettenmann

. Different mechanisms promote astrocyte Ca2+ waves and spreading depression in the mouse neocortex. J Neurosci 2003; 23:9888–96.

144.

144 Margineanu

Klitgaard

. The connexin 36 blockers quinine, quinidine and mefloquine inhibit cortical spreading depression in a rat neocortical slice model in vitro . Brain Res Bull 2006; 71:23–8.

145.

145 Kitahara

Taga

Abe

Shimoji

. The effects of anesthetics on cortical spreading depression elicitation and c-fos expression in rats. J Neurosurg Anesthesiol 2001; 13:26–32.

146.

146 Piper

Lambert

. Inhalational anesthetics inhibit spreading depression: relevance to migraine. Cephalalgia 1996; 16:87–92.

147.

147

Chan

Evans

Hadley

Herdon

Jerman

Parsons

Identification of (-)-cis-6-acetyl-4S-(3-chloro-4-fluoro-benzoylamino)- 3,4-dihydro-2,2-dimethyl-2H-benzo[b]pyran-3S-ol as a potential antimigraine agent. Bioorg Med Chem Lett 1999; 9:285–90.

148.

148 Damodaram

Thalakoti

Freeman

Garrett

Durham

. Tonabersat inhibits trigeminal ganglion neuronal–satellite glial cell signaling. Headache 2009; 49:5–20.

149.

149 Goadsby

Ferrari

Csanyi

Olesen

Mills

. Randomized, double-blind, placebo-controlled, proof-of-concept study of the cortical spreading depression inhibiting agent tonabersat in migraine prophylaxis. Cephalalgia 2009; 29:742–50.

150.

150 Hauge

Asghar

Schytz

Christensen

Olesen

. Effects of tonabersat on migraine with aura: a randomised, double-blind, placebo-controlled crossover study. Lancet Neurol 2009; 8:718–23.

151.

151 Fujimoto

Yanase

. The effect of K(+)-conductance-blocking substances on the occurrence of retinal spreading depression. Exp Eye Res 1991; 53:333–6.

152.

152 Chen

Gao

Reinert

Popa

Hendrix

Ross

Ebner

. Involvement of kv1 potassium channels in spreading acidification and depression in the cerebellar cortex. J Neurophysiol 2005; 94:1287–98.

153.

153 Psarropoulou

Avoli

. CPP, an NMDA-receptor antagonist, blocks 4-aminopyridine-induced spreading depression episodes but not epileptiform activity in immature rat hippocampal slices. Neurosci Lett 1992; 135:139–43.

154.

154 Koroleva

Oitzl

Bures

. Threshold of synaptically elicited cortical spreading depression: drug-induced changes in rats. Electromyogr Clin Neurophysiol 1985; 60:55–64.

155.

155

Boissard

Greco

Gribkoff

Harden

(S)-N-[1-(3-Morpholin-4-ylphenyl)ethyl]- 3-phenylacrylamide: an orally bioavailable KCNQ2 opener with significant activity in a cortical spreading depression model of migraine. J Med Chem 2003; 46:3197–200.

156.

156 Su

Hayashi

. Understanding the molecular mechanism of sigma-1 receptors: towards a hypothesis that sigma-1 receptors are intracellular amplifiers for signal transduction. Curr Med Chem 2003; 10:2073–80.

157.

157 Hayashi

. Sigma-1 receptor ligands: potential in the treatment of neuropsychiatric disorders. CNS Drugs 2004; 18:269–84.

158.

158 Anderson

Andrew

. Spreading depression: imaging and blockade in the rat neocortical brain slice. J Neurophysiol 2002; 88:2713–25.

159.

159 Martins-Ferreira

De Oliveira Castro

Struchiner

Rodrigues

. Circling spreading depression in isolated chick retina. J Neurophysiol 1974; 37:773–84.

160.

160 Tong

Chesler

. Endogenous pH shifts facilitate spreading depression by effect on NMDA receptors. J Neurophysiol 1999; 81:1988–91.

161.

161 Tong

Chesler

. Modulation of spreading depression by changes in extracellular pH. J Neurophysiol 2000; 84:2449–57.

162.

162 Menna

Tong

Chesler

. Extracellular pH changes and accompanying cation shifts during ouabain-induced spreading depression. J Neurophysiol 2000; 83:1338–45.

163.

163 Skelton

Gardner-Medwin

George

. The effects of carbon dioxide, oxygen and pH on spreading depression in the isolated chick retina. Brain Res 1983; 288:229–33.

164.

164 Gardner-Medwin

. Possible roles of vertebrate neuroglia in potassium dynamics, spreading depression and migraine. J Exp Biol 1981; 95:111–27.

165.

165

Vahedi

Taupin

Djomby

El-Amrani

Lutz

Filipetti

Efficacy and tolerability of acetazolamide in migraine prophylaxis: a randomised placebo-controlled trial. J Neurol 2002; 249:206–11.

166.

166 Athwal

Lennox

. Acetazolamide responsiveness in familial hemiplegic migraine. Ann Neurol 1996; 40:820–1.

167.

167

Battistini

Stenirri

Piatti

Gelfi

Righetti

Rocchi

A new CACNA1A gene mutation in acetazolamide-responsive familial hemiplegic migraine and ataxia. Neurology 1999; 53:38–43.

168.

168 Baloh

Foster

Yue

Nelson

. Familial migraine with vertigo and essential tremor. Neurology 1996; 46:458–60.

169.

169 Dodgson

Shank

Maryanoff

. Topiramate as an inhibitor of carbonic anhydrase isoenzymes. Epilepsia 2000; 41 (Suppl. 1):S35–9.

170.

170 Wilner

Raymond

Pollard

. Topiramate and metabolic acidosis. Epilepsia 1999; 40:792–5.

171.

171

De Fusco

Marconi

Silvestri

Atorino

Rampoldi

Morgante

Haploinsufficiency of ATP1A2 encoding the Na+/K+ pump alpha2 subunit associated with familial hemiplegic migraine type 2. Nat Genet 2003; 33:192–6.

172.

172 Capendeguy

Horisberger

. Functional effects of Na+,K+-ATPase gene mutations linked to familial hemiplegic migraine. Neuromolecular Med 2004; 6:105–16.

173.

173 van den Maagdenberg

Haan

Terwindt

Ferrari

. Migraine: gene mutations and functional consequences. Curr Opin Neurol 2007; 20:299–305.

174.

174 Pietrobon

. Familial hemiplegic migraine. Neurotherapeutics 2007; 4:274–84.

175.

175 Haglund

Schwartzkroin

. Role of Na-K pump potassium regulation and IPSPs in seizures and spreading depression in immature rabbit hippocampal slices. J Neurophysiol 1990; 63:225–39.

176.

176

de Almeida Rabello Oliveira

da Rocha Ataide

de Oliveira

de Melo Lucena

de Lira

Soares

Effects of short-term and long-term treatment with medium- and long-chain triglycerides ketogenic diet on cortical spreading depression in young rats. Neurosci Lett 2008; 434:66–70.

177.

177 Bough

Schwartzkroin

Rho

. Calorie restriction and ketogenic diet diminish neuronal excitability in rat dentate gyrus in vivo . Epilepsia 2003; 44:752–60.

178.

178 Bough

. Energy metabolism as part of the anticonvulsant mechanism of the ketogenic diet. Epilepsia 2008; 49 (Suppl. 8):91–3.

179.

179 Brand

Fernandes de Lima

Hanke

. Pharmacological modulation of the refractory period of retinal spreading depression. Naunyn Schmiedebergs Arch Pharmacol 1998; 357:419–25.

180.

180 Van Harreveld

Stamm

. Effect of pentobarbital and ether on the spreading cortical depression. Am J Physiol 1953; 173:164–70.

181.

181 Stensrud

Sjaastad

. Clonazepam (rivotril) in migraine prophylaxis. Headache 1979; 19:333–4.

182.

182 White

Brown

Woodhead

Skeen

Wolf

. Topiramate enhances GABA-mediated chloride flux and GABA-evoked chloride currents in murine brain neurons and increases seizure threshold. Epilepsy Res 1997; 28:167–79.

183.

183 Cutrer

Limmroth

Moskowitz

. Possible mechanisms of valproate in migraine prophylaxis. Cephalalgia 1997; 17:93–100.

184.

184 Tobiasz

Nicholson

. Tetrodotoxin resistant propagation and extracellular sodium changes during spreading depression in rat cerebellum. Brain Res 1982; 241:329–33.

185.

185 Kow

van Harreveld

. Ion and water movements in isolated chicken retinas during spreading depression. Neurobiology 1972; 2:61–9.

186.

186 Olsen

Miller

. Spontaneous slow potentials and spreading depression in amphibian retina. J Neurophysiol 1977; 40:752–67.

187.

187 Mori

Miller

Tomita

. Microelectrode study of spreading depression (SD) in frog retina—general observations of field potential associated with SD. Jpn J Physiol 1976; 26:203–17.

188.

188 Fabricius

Akgoren

Lauritzen

. Arginine-nitric oxide pathway and cerebrovascular regulation in cortical spreading depression. Am J Physiol 1995; 269:H23–9.

189.

189 Sugaya

Takato

Noda

. Neuronal and glial activity during spreading depression in cerebral cortex of cat. J Neurophysiol 1975; 38:822–41.

190.

190

Dichgans

Freilinger

Eckstein

Babini

Lorenz-Depiereux

Biskup

Mutation in the neuronal voltage-gated sodium channel SCN1A in familial hemiplegic migraine. Lancet 2005; 366:371–7.

191.

191 Clarke

Upton

Kamath

Al-Harbi

Castellanos

. Transcranial magnetic stimulation for migraine: clinical effects. J Headache Pain 2006; 7:341–6.

192.

192 Brighina

Piazza

Vitello

Aloisio

Palermo

Daniele

Fierro

. rTMS of the prefrontal cortex in the treatment of chronic migraine: a pilot study. J Neurol Sci 2004; 227:67–71.

193.

193

Teepker

Hotzel

Timmesfeld

Reis

Mylius

Haag

Low-frequency rTMS of the vertex in the prophylactic treatment of migraine. Cephalalgia 2009; Jun 8 [Epub ahead of print].

194.

194

Fregni

Liebetanz

Monte-Silva

Oliveira

Santos

Nitsche

Effects of transcranial direct current stimulation coupled with repetitive electrical stimulation on cortical spreading depression. Exp Neurol 2007; 204:462–6.

195.

195 Liebetanz

Fregni

Monte-Silva

Oliveira

Amancio-dos-Santos

Nitsche

Guedes

. After-effects of transcranial direct current stimulation (tDCS) on cortical spreading depression. Neurosci Lett 2006; 398:85–90.

196.

196 Fregni

Monte-Silva

Oliveira

Freedman

Pascual-Leone

Guedes

. Lasting accelerative effects of 1 Hz and 20 Hz electrical stimulation on cortical spreading depression: relevance for clinical applications of brain stimulation. Eur J Neurosci 2005; 21:2278–84.

197.

197 Holland

Schembri

Fredrick

Goadsby

. Transcranial magnetic stimulation for the treatment of migraine aura. American Academy of Neurology 61st Annual Meeting. Seattle, WA, 2009.

198.

198 Gardner-Medwin

. A study of the mechanisms by which potassium moves through brain tissue in the rat. J Physiol 1983; 335:353–74.

199.

199 Ayata

Moskowitz

. Cortical spreading depression confounds concentration-dependent pial arteriolar dilation during N-methyl-D-aspartate superfusion. Am J Physiol Heart Circ Physiol 2006; 290:H1837–41.

200.

200 Akerman

Holland

Goadsby

. Needle-stick induced regional cerebral blood flow change (rCBF) is affected by Na+ channel blockade, AMPA/kainate antagonism and GABA agonists, but not by voltage-dependent Ca2+ channel or ATP-dependent K+ channel blockade. Cephalalgia 2005; 25:871.

201.

201

Dreier

Kleeberg

Petzold

Priller

Windmuller

Orzechowski

Endothelin-1 potently induces Leao's cortical spreading depression in vivo in the rat: a model for an endothelial trigger of migrainous aura?

Brain 2002; 125:102–12.

202.

202 Schwedt

Demaerschalk

Dodick

. Patent foramen ovale and migraine: a quantitative systematic review. Cephalalgia 2008; 28:531–40.

203.

203 Brennan

Beltran-Parrazal

Lopez-Valdes

Theriot

Toga

Charles

. Distinct vascular conduction with cortical spreading depression. J Neurophysiol 2007; 97:4143–51.

204.

204 Latour

Hasegawa

Formato

Fisher

Sotak

. Spreading waves of decreased diffusion coefficient after cortical stimulation in the rat brain. Magn Reson Med 1994; 32:189–98.

205.

205

Bockhorst

Smith

Bradley

Houston

Carpenter

A quantitative analysis of cortical spreading depression events in the feline brain characterized with diffusion-weighted MRI. J Magn Reson Imaging 2000; 12:722–33.

206.

206 Silberstein

Tfelt-Hansen

Dodick

Limmroth

Lipton

Pascual

Wang

. Guidelines for controlled trials of prophylactic treatment of chronic migraine in adults. Cephalalgia 2008; 28:484–95.

207.

207 Tower

Young

. The activities of butyrylcholinesterase and carbonic anhydrase, the rate of anaerobic glycolysis, and the question of a constant density of glial cells in cerebral cortices of various mammalian species from mouse to whale. J Neurochem 1973; 20:269–78.

Spreading Depression: From Serendipity to Targeted Therapy in Migraine Prophylaxis

Abstract

Keywords

INTRODUCTION

Spreading depression

EVIDENCE SUPPORTING A ROLE FOR SPREADING DEPRESSION IN MIGRAINE

EVIDENCE FOR SPREADING DEPRESSION AS A SERENDIPITOUS THERAPEUTIC TARGET IN MIGRAINE

In vitro studies

In vivo studies

Chronic treatment (weeks to months) is required for some prophylactic drugs to suppress SD, in vivo

PHARMACOLOGICAL TARGETS MODULATING SPREADING DEPRESSION SUSCEPTIBILITY AND THEIR RELEVANCE IN MIGRAINE PROPHYLAXIS

Glutamate receptors

Voltage-gated Ca2+ channels

Adrenergic receptors

Serotonin receptors

Gap junction blockers

K+ channels

Sigma-1 receptors

Tissue pH

Na+/K+ ATPase

GABAA receptors

Voltage-gated Na+ channels

Non-pharmacological interventions

EXPERIMENTAL MODELS OF SPREADING DEPRESSION FOR DRUG SCREENING

In vitro vs. in vivo models

Methods to trigger SD

Methods to monitor SD

EMPIRICAL CONSIDERATIONS

Experimental models

Relevant SD attributes

Experimental animals

Treatment dose and protocol

CONCLUDING REMARKS

References

Voltage-gated Ca²⁺ channels

K⁺ channels

Na⁺/K⁺ ATPase

GABA_A receptors

Voltage-gated Na⁺ channels