Mechanically-induced cortical spreading depression associated regional cerebral blood flow changes are blocked by Na+ ion channel blockade

BR A IN RE S E A RCH 1 2 29 ( 20 0 8 ) 2 7 –3 6

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / b r a i n r e s

Research Report

Mechanically-induced cortical spreading depression associated regional cerebral blood flow changes are blocked by Na + ion channel blockade Simon Akerman, Philip R. Holland, Peter J. Goadsby⁎ Department of Neurology, University of California, San Francisco, 505 Parnassus Avenue, San Francisco CA 94143-0114, USA

A R T I C LE I N FO

AB S T R A C T

Article history:

Migraine with aura occurs in up to 20–30% of all migraineurs. The regional cerebral blood flow

Accepted 3 July 2008

(rCBF) changes that occur during cortical spreading depression (CSD) are considered to be an

Available online 12 July 2008

experimental correlate of aura. CSD is synchronous with a failure in brain ion homeostasis and efflux of excitatory amino acids from nerve cells. Therefore studying the mechanisms

Keywords:

that underlie CSD, such as ion channel manipulation, and observing rCBF changes may help

Cortical spreading depression

our understanding of migraine aura. In this study we used mechanical stimulation to induce

Aura

oligemia and hyperemia, in surgically prepared cats and rats, using laser Doppler probes to

Sodium channel

measure the cerebral blood flow and single cell cortical recording to measure the spike/

Migraine

neuronal burst, both generated as a consequence of CSD. We looked at the response of ion channel blockers directed at sodium, voltage-dependent calcium and ATP-activated potassium ion channels. The sodium ion channel blocker was able to inhibit rCBF changes in both the cat and rats. Voltage-dependent calcium channel blockers had little effect on the initiation or propagation of the spread, as did the ATP-activated potassium channel blocker. The data are consistent with what is known of human aura in that sodium ion channels are those predominantly involved in mechanical stimulation-induced rCBF changes and thus may represent therapeutic targets for the aura response in migraine. © 2008 Elsevier B.V. All rights reserved.

1.

Introduction

Migraine is a common (Lipton et al., 2001) and disabling (Menken et al., 2000) neurological disorder involving activation, or the perception of activation, of trigeminal neurons (Goadsby et al., 2002). It has been estimated to be the most costly neurological disorder in the European Community at more than €27 billion per year, affecting 14% of the population (Andlin-Sobocki et al., 2005). Up to 30% of sufferers report focal

neurological disturbances: the ‘aura’ (Rasmussen and Olesen, 1992). The aura symptoms develop over 5–20 min and usually last less than an hour, taking the form of visual, sensory and motor disturbances. Aura is thought to be a wave of oligemia (Lauritzen, 1994), or deficiency in blood flow preceded by a brief hyperemia, that spreads across the cortex at 2–6 mm min− 1 (Headache Classification Committee of The International Headache Society, 2004). Indeed the aura phase has been shown to be accompanied by a slowly spreading

⁎ Corresponding author. E-mail address: [email protected] (P.J. Goadsby). Abbreviations: ANOVA, analysis of variance; AU, arbitrary units; AUC, area under curve; CSD, cortical spreading depression; FHM, familial hemiplegic migraine; HR, hyperemic response; rCBF, regional cerebral blood flow; SP, speed of propagation; TTX, tetrodotoxin; VDCC, voltage-dependent calcium channel 0006-8993/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2008.07.014

28

BR A IN RE S EA RCH 1 2 29 ( 20 0 8 ) 2 7 –36

reduction in regional cerebral blood flow (rCBF) (Olesen, 1991; Olesen et al., 1981). The hyperemia followed by oligemia occurring during the migraine aura is similar to the changes that take place during the cortical spreading depression (CSD) of Leao (Leao, 1944, 1944). CSD is believed, primarily, to be an electrophysiological phenomenon involving a wave of depolarisation followed by a short-lasting depression that spreads across the cortex at a similar rate to aura, and is also associated with regional cerebral blood flow changes and a sustained hypoperfusion in the same brain region as the depolarisation. The vascular changes are largely (Lauritzen, 1994), but not exclusively due to vasoneuronal coupling (Brennan et al., 2007). It is the blood flow changes that have been observed in migraine aura (Lauritzen, 1994). The changes are synchronous with a failure in brain ion homeostasis and efflux of excitatory amino acids from nerve cells (Kraig and Nicholson, 1978; Lauritzen, 1994, 2001); (Martins-Ferreira et al., 2000; Somjen, 2001). This similarity helps provide important support for the theory that the mechanism of CSD may underlie migraine aura. Studying the mechanisms involved in CSD, such as ion homeostasis and excitatory amino acid efflux, using the observation of vascular and neuronal changes in animals may help in our understanding of the migraine aura. Cortical hyperemia can be triggered by electrical, mechanical and chemical stimuli in vivo (Ebersberger et al., 2001, Kaube and Goadsby, 1994, Kaube et al., 1999, Lauritzen et al., 1988; Nellgard and Wieloch, 1992, Richter et al., 2002, Somjen, 2001, van den Maagdenberg et al., 2004). Many experiments have used compounds that interfere with the ion homeostasis and excitatory amino acid efflux and measured the blood flow response in order to understand its mechanism more fully. At present only MK801, an NMDA glutamate receptor blocker, halothane and topiramate have been shown to block mechanically induced vascular changes and the neuronal

burst of firing coincident with CSD (Akerman and Goadsby, 2005, Kaube and Goadsby, 1994, Zhang et al., 1990). While it has recently been shown that some migraine preventives may have actions in inhibiting CSD with longer term administration (Ayata et al., 2006). In the present study we used whole animal in vivo techniques to study vascular changes, and used mechanical stimulation of the cortex to induce the wave of spreading depression, a correlate of the blood flow changes that occur during aura. We sought to investigate the underlying mechanism of the rCBF changes that occur by targeting potential specific ion changes that may take place in CSD. Studies looking at these targets have been done previously using electrical stimulation and KCl induced CSD in rodents (Fabricius et al., 1995, Kow and van Harreveld, 1972, Lauritzen and Hansen, 1992, Lauritzen et al., 1988, Marrannes et al., 1988, Nellgard and Wieloch, 1992; Richter et al., 2002; Shimazawa and Hara, 1996; Shimizu et al., 2000; Sugaya et al., 1975; van den Maagdenberg et al., 2004; Wu et al., 2003). We used Na+, ATPactivated K+ and voltage dependent Ca2+ ion channel blockers, and carefully observed the rCBF changes. If the rCBF changes were significantly affected by the addition of the test compound we proceeded to observe the single unit electrical changes as well, to confirm that the compound is affecting the electrical activity and not just blood flow. The work has been presented in preliminary form at the XIth International Headache Congress [Kyoto, October 2005 (Akerman et al., 2005)].

2.

Results

In all cats the respiratory parameters were maintained at physiological levels throughout the experiments: pH 7.43 ± 0.01, pCO2 2.72 ± 0.1 kPa and O2 33.5 ± 1.2 kPa. To avoid hypovolemia in the rat experiments arterial samples were

Table 1 – Summary of raw data for the ion channel modulators that highlights whether cerebral blood flow changes have occurred Experimental plunge

Did rCBF change occur?

Statistical significance Probe 1

Saline control Post ω-conotoxin-GVIa (20 μg kg− 1) Saline control Post ω-agatoxin-IVa (20 μg kg− 1) Saline control Post calciseptine (20 μg kg− 1) Saline control Post TTX (60 μg kg− 1, cat) 60 min post TTX Saline Post TTX (10 μg kg− 1, in rat) 45 min post TTX Control plunge Post cadmium chloride (20 μg kg− 1) Saline control Post glibenclamide (30 mg kg− 1, ip)

✓ ✓ ✓ ✓ ✓ ✓ ✓

 ✓ ✓

 ✓ ✓ ✓ ✓ ✓

Probe 2

⁎(t7 = 6.22, n = 8) ⁎(t10 = 4.1, n = 11) ⁎(t6 = 3.78, n = 7) ⁎(t6 = 2.90, n = 7) ⁎(t5 = 4.51, n = 6) ⁎(t5 = 2.74, n = 6) ⁎(t5 = 3.31, n = 6)

⁎(t9 = 7.75, n = 10) ⁎(t10 = 3.48, n = 11) ⁎(t6 = 8.02, n = 7) ⁎(t6 = 5.44, n = 7) ⁎(t5 = 4.09, n = 6) ⁎(t5 = 2.83, n = 6) ⁎(t5 = 4.71, n = 6)

(t5 = 1.55, n = 6, p = 0.18) ⁎(t4 = , n = 5) ⁎(t7 = − 6.46, n = 8) (t7 = −1.35, p = 0.22, n = 8) ⁎(t7 = − 6.77, n = 8) ⁎(t4 = 3.81, n = 5) ⁎(t4 = 3.62, n = 5) ⁎(t6 = 4.02, n = 7) ⁎(t5 = 3.22, n = 6)

(t5 = 1.55, n = 6, p = 0.18) ⁎(t3 = 3.3, n = 4) – – – ⁎(t4 = 3.29, n = 5) ⁎(t4 = 3.33, n = 5) ⁎(t6 = 11.79, n = 7) ⁎(t5 = 3.23, n = 6)

✓ — CSD occurred,  — CSD did not occur (or inhibited). TTX — tetrodotoxin. ⁎ p b 0.05 significance compared to the baseline in that series of experiments, indicating significant blood flow change.

BR A IN RE S E A RCH 1 2 29 ( 20 0 8 ) 2 7 –3 6

29

not taken, but blood pressure and end-expiratory CO2 levels were observed and physiological throughout the studies. Baseline data are summarised and included in Table 1 for each test compound. Briefly, on each occasion for the control response to mechanical stimulation and the response after saline there was a significant increase in the area under the curve for each probe compared to the baseline recording, thus indicating that the hyperemic response of cortical spreading depression had occurred on each occasion.

2.1.

Cat studies

2.1.1.

Calcium channels

There was a significant increase in AUC after needle plunge when ω-conotoxin-GVIa was administered, for each probe (Table 1). When we tested the response across the entire cohort using ANOVA for repeated measures probe 1 proved significant (F1,22 = 7.77, p b 0.05, n = 11) while probe 2 was not (F1,23 = 0.00, p = 0.98, n = 11). When each hyperemic response was compared to the control using a Student's paired t-test no significances were found for either probe. When ω-agatoxinIVa was used as a pre-treatment to needle plunge there was a significant increase in AUC of rCBF for both probes indicating that a hyperemic response similar to that found in CSD had occurred. When ANOVA for repeated measures tested the entire cohort, there was no significance across the cohort, probe 1 (F2,12 = 1.71, p = 0.23, n = 7) and probe 2 (F1,12 = 0.43, p = 0.63, n = 7). In the case of calciseptine, there was a significant increase in AUC for each probe, the hyperemic response was significant. When ANOVA for repeated measures tested the entire cohort, there was no significance across the cohort, probe 1 (F3,15 = 1.43, p = 0.29, n = 6) and probe 2 (F3,15 = 0.22, p = 0.76, n = 6). Cadmium chloride did not prevent the rCBF changes: there was a significant increase in AUC of rCBF for each probe when it was used as a pre-treatment. Only a control and treatment group was included for cadmium, when the control response was compared to the cadmium pre-treated response no significances occurred for either probe, probe 1 (t4 = 0.29, p = 0.98, n = 5) and probe 2 (t4 = 0.35, p = 0.75, n = 5).

2.1.2.

Sodium channels

In the case of tetrodotoxin in two of six experiments a degree of hyperemia was observed, however there was no significance of the AUC between the baseline response and that after the needle plunge, indicating the hyperemic response had been inhibited, probe 1 (t5 = 1.55, p = 0.18, n = 6) and probe 2 (t5 = 1.56, p = 0.18, n = 6; Fig. 1A). When we compared the results after needle plunges across the cohort there was also a significance for each probe, probe 1 (F3,9 = 3.99, p b 0.05) and probe 2 (F3,12 = 5.15, p b 0.05).

2.1.3.

Atp-dependent potassium channels

Glibenclamide did not prevent the HR as the AUC was significant when compared to the baseline response. When the response across the entire cohort was tested there was also no significance, probe 1 (F2,12 = 0.53, p = 0.59), and probe 2 (F2,12 = 2.04, p = 0.2).

2.1.4.

Speed of propagation

We measured both the time taken to start the hyperemic response and also the speed of the propagation. When ω-

Fig. 1 – Effect of tetrodotoxin, on cerebral blood flow changes induced in a model of cortical spreading depression in the (A) cat or (B) rat. Following control and saline control cats were injected with either 60 μg kg− 1 tetrodotoxin, rats with 10 μg kg− 1 tetrodotoxin 5 min prior to spreading depression induction by needle plunge into the cortex, and the cerebral blood flow changes recorded. * Represents blood flow responses that were not significant from baseline recordings (baseline data not shown).

conotoxin-GVIa was used as a pre-treatment there was no significant difference in either the time to start the hyperemic response (F1,21 = 0.95, p = 0.34) or the speed of propagation (F1,19 = 0.12, p = 0.73) across the cohort of needle plunges. This was similar for ω-agatoxin-IVa (F 1,14 = 2.91, p = 0.11 and F1,14 = 1.32, p = 0.27, respectively), calciseptine (F1,9 = 0.02, p = 0.94 and F1,9 = 0.14, p = 0.68, respectively) and cadmium chloride (t4 = 0.93, p = 0.40 and t4 = 1.33, p = 0.25, respectively). In the two examples where hyperemia occurred with TTX, it was unable to affect the time to hyperemic response (F2,9 = 3.2, p = 0.092) and the SP (F2,9 = 0.04, p = 0.96). Glibenclamide did not significantly effect either time to hyperemic response (F2,12 = 0.06, p = 0.85) or the speed of propagation (F1,7 = 3.69, p = 0.1).

2.2.

Rat studies

Additionally cortical spreading depression, using mechanical stimulation, was induced in rats and they were treated with tetrodotoxin (10 μg kg− 1) in order to assess the effects it might

30

BR A IN RE S EA RCH 1 2 29 ( 20 0 8 ) 2 7 –36

Fig. 2 – A schematic top view (not to scale) of the experimental set-up in the (A) cat and (B) rat. (A) It shows the two laser Doppler probes on the drilled closed cranial window or (B) the single laser Doppler probe with the recording electrode occipital to this, with the bone and dura removed. The needle stick site is anterior to bregma. MMA—middle meningeal artery. (C) Example tracing of a single laser Doppler probe and single cell cortical electrical activity occurring during a mechanical stimulation to induce spreading depression in the rat. A needle plunge was followed by a hyperemic response and a cortical spike and subsequent electrical depression that spreads across the cortex. Intravenous injection of tetrodotoxin (10 μg kg− 1) inhibits the spreading depression and cortical spiking, after 45 min the spreading depression response returns.

have on the electrical discharge as well as the blood flow changes, as has been done previously (Akerman and Goadsby, 2005; Kaube and Goadsby, 1994; Kaube et al., 1999). Again tetrodotoxin inhibited the cerebral blood flow response to mechanical stimulation induced cortical spreading depression (F3,15 = 16.91, p b 0.05, n = 6) and also when compared to the saline control group (t7 = 5.03, p b 0.05, n = 8; Fig. 1B). Also, when the neuronal recording was being monitored there was a spike of electrical activity that was time-matched to the hyperemic response, followed by a depression of electrical activity (Fig. 2B) and this spike/neuronal discharge was similarly inhibited by tetrodotoxin across the cohort (F3,15 = 17.81, p b 0.05, n = 6). Both the hyperemic response and the cortical spike of electrical activity returned on a subsequent needle prick.

2.2.1.

Blood pressure changes

Arterial blood pressure was monitored throughout, and there is a summary of changes that occurred after drug intervention

in Table 2. Briefly, with the ion channel modulators, ωconotoxin-GVIa (t10 = 3.08, p b 0.05, n = 11) and cadmium chloride (t3 = 6.00, p b 0.05, n = 4), both caused a significant decrease in mean arterial blood pressure, while glibenclamide (t6 = 5.53, p b 0.05, n = 7) caused a significant increase in the blood pressure. Tetrodotoxin caused a significant decrease in mean arterial blood pressure in the cat (t4 = 5.44, p b 0.05, n = 5) and rat (t7 = 6.52, p b 0.05, n = 8).

3.

Discussion

The vascular changes and neuronal discharge/spike activity that occur during cortical mechanical stimulation (CMS) were inhibited by the sodium ion channel blocker, tetrodotoxin. The various voltage-dependent calcium channel (VDCC) blockers and an ATP-dependent K+ channel blocker were unable to inhibit the induced changes, or affect the speed of propagation

31

BR A IN RE S E A RCH 1 2 29 ( 20 0 8 ) 2 7 –3 6

Table 2 – Summary of mean arterial blood pressure changes with drug intervention Drug intervention

Baseline blood pressure (mm Hg)

Blood pressure change (mm Hg)

Statistics

ω-conotoxin-GVIa

85 ± 5

↓13.5 ± 4

ω-agatoxin-IVa

78 ± 5

↓1 ± 2

Calciseptine

77 ± 8

↓7.1 ± 7

Cadmium chloride

104 ± 14

↓41.0 ± 7

Tetrodotoxin (cat)

87 ± 8

↓38.0 ± 16

Tetrodotoxin (rat)

100 ± 7

↓43.8 ± 7

87 ± 6

↑26.0 ± 5

⁎t10 = 3.08, p b 0.05, n = 11 t6 = 0.45, p = 0.67, n=7 t5 = 0.98, p = 0.37, n=6 ⁎t3 = 6.00, p b 0.05, n=4 ⁎t4 = 5.44, p b 0.05, n=5 ⁎t7 = 6.52, p b 0.05, n=8 ⁎t6 = 5.53, p b 0.05, n=7

Glibenclamide

↓↑ direction of blood pressure change. ⁎ p b 0.05 significant change from baseline.

and timing of the hyperemic response. The data indicate that sodium ions are involved in the mechanical stimulation induced vascular changes and neuronal discharge that accompanies spreading depression. The VDCCs and the ATPdependent K+ channels seem to be somewhat less important in these vascular changes that accompany spreading depression, although they are still believed to be important in the state of depolarisation and subsequent hyperpolarisation that defines spreading depression. A summary of responses of these compounds directed against the different modes of CSD induction can be found in Table 3. Familial hemiplegic migraine (FHM) is a rare autosomal dominant form of migraine with aura that is due to either missense mutations of the Cav2.1 gene, which encodes the α1A subunit of the voltage-dependent P/Q-type calcium channel (Ophoff et al., 1996), or to mutations in ATP1A2 gene that encodes the α2 subunit of the Na+/K+ pump (De Fusco et al., 2003), or to mutations in the gene for the voltage-gated sodium channel SCN1A (Dichgans et al., 2005). Therefore VDCCs may be involved in the genesis of migraine aura. The data presented, and that of Richter et al. (2002) indicate that N, P/ Q or L-type VDCCs are not involved in the rCBF and DC shift changes caused by CMS in the cortex, using either systemic administration or direct application by bathing the cortex with VDCCs (Richter et al., 2002). However, in transgenic mice with the human FHM-1 R192Q mutation knocked in, the mutant mice required a much lower electrical current threshold to induce CSD and the speed of propagation was much greater than in wild-types (van den Maagdenberg et al., 2004). When KCl crystals were placed on the pia mater to induce CSD,

bathing the area with both P/Q and N-type voltage-dependent calcium channel (VDCC) blockers reduced the number of repetitions of SD, as measured by DC shifts (Richter et al., 2002). An L-type VDCC blocker had only a small effect on the repetition rate. The implication being that the N and P/Q-type VDCCs may be involved in CSD. These experiments begin to highlight potential differences in the methods of CSD induction and the outcomes depending on treatments. VDCC blockers seem to have less effect on the wave generated by mechanical stimulation of the cortex, but appear effective in the KCl induced CSD model. This could be a response to the type of stimulus used. Previous studies have shown that using different levels of CSD induction stimulus with either varying concentrations of KCl or changing electrical currents they were able to dissect a susceptibility to manipulate CSD with VDCCs (Ayata et al., 2000; Richter et al., 2002; van den Maagdenberg et al., 2004). The mechanical method with the use of the needle prick is less variable in this way, however, the use of von Frey hairs, which could be used to apply increasing forces to induced CSD and rCBF changes through mechanical stimulation may be a novel approach to this issue and particularly useful in comparing transgenic models with wild-type animals, where within subject controls are not possible. ω-Conotoxins are known to penetrate the blood–brain barrier, therefore intravenously administered compounds are likely to have reached the brain (Newcomb et al., 2000). It is less certain whether ω-agatoxin crosses the blood–brain barrier, however Richter and colleagues (Richter et al., 2002) did not see significant inhibition of CMS-induced CSD when ωagatoxin was applied locally. As such the route of administration of VDCCs did not seem to affect significantly the response to CMS-induced rCBF changes. Calciseptine was not effective with either direct (bathing the cortex) or indirect (intravenous administration) application and therefore unlikely to be responsive in this model. VDCCs may play a role in the migraine aura response but CMS-induced CSD method may not be a good model for observing their role. Sodium ion channels may be involved in the initiation of CSD (Somjen, 2001) and agents acting on these channels, such as lamotrigine, have been suggested to be useful in migraine with aura or persistent aura (Chen et al., 2001), but less so in migraine without aura (Steiner et al., 1997). In our study we used tetrodotoxin (TTX), a sodium ion channel blocker used to

Table 3 – Summary of ion channel interventions that affect CSD/rCBF changes using different modes if induction Ion channel pharmacology N-type VDCC P/Q-type VDCC L-type VDCC Na+ channels ATP-activated K+

Mechanical stimulation

KCl induction

Electrical stimulation

  

✓ ✓ ✓



✓



 ✓

✓ — CSD/rCBF affected,  — CSD/rCBF unaffected. VDCC — voltage-dependent calcium channels. ND — no data.

✓

 ✓ ND

32

BR A IN RE S EA RCH 1 2 29 ( 20 0 8 ) 2 7 –36

characterise their presence in the brain and on peripheral nerves, as opposed to TTX-resistant sodium ion channels present in the heart. TTX was able to inhibit the hyperemia induced by mechanical stimulation of the cerebral cortex in both cats and rats, and in each case the hyperemic response returned after 45 min. We also showed that the electrical spiking/neuronal discharge that often accompanies cortical spreading depression was inhibited. These data indicate that during CMS sodium ion channels appear to be involved in the initiation of rCBF wave. Evidence from in vitro studies using tissue slices from the hippocampus shows that TTX is also able to inhibit and delay CSD (Somjen, 2001), and the description of a mutation in the SCN1A gene in FHM type III (Dichgans et al., 2005), are consistent with some role for sodium channels in aura and offer a potential target for new therapies. Other evidence in vivo using TTX is slightly more conflicting. Sugaya and colleagues (Sugaya et al., 1975) found that the neuronal discharge disappeared after TTX treatment, as well as the spreading depression after repetitive stimulation. If, however, the electrical stimulus was increased up to 50 V spreading depression could still be evoked (Sugaya et al., 1975). Electrical stimulation induced spreading depression has also been found by others to be inhibited by TTX (Tobiasz and Nicholson, 1982) When KCl was used to initiate spreading depression and the propagation analysed, TTX was able to inhibit the neuronal discharge but unable to inhibit the spread when it came from an untreated region, and also when it was initiated in a treated region, although the shape was deformed with quicker recovery (Kow and van Harreveld, 1972; Sugaya et al., 1975; Tobiasz and Nicholson, 1982). They did however find that when TTX treatment duration was increased it became more difficult for KCl to induce the spread. Others have also found that TTX was unable to inhibit the propagation of spreading depression, again induced by KCl (Kow and van Harreveld, 1972; Tobiasz and Nicholson, 1982). It therefore seems that some role for Na+ ions in spreading depression is reasonably clear. Potassium ion exchange also occurs during CSD and with evidence that some FHM patients possess a mutation in the ATP1A2 gene that encodes the α2 subunit of the Na+/K+ pump (De Fusco et al., 2003), ATP-activated potassium channels may be involved in the aura response. Potassium ion channels may be a useful target in understanding the rCBF changes that take place during the wave that occurs as a result of mechanical stimulation (Somjen, 2001). Glibenclamide, an ATP-activated potassium channel blocker, was unable to inhibit rCBF changes or significantly alter the speed of propagation. There is previous evidence that glibenclamide is able to affect cerebral blood flow after i.p. administration, albeit in a stroke model (Xiong et al., 2003), therefore it seems glibenclamide reaches the brain. Glibenclamide has previously been shown to increase the peak of the hyperemic response of blood flow changes in the KCl method of CSD induction (Shimizu et al., 2000) and potassium channel openers of the KCNQ2 gene family have also been found to reduce the number of repetitions (Wu et al., 2003). Potassium channels do appear to have importance in KCl-induced models, but they do not appear to play an important role in either the initiation or the propagation of the hyperemia found with mechanical stimulation of the

cortex as used in the current experiments. This perhaps highlights important differences in the methodology, the KCl method using potassium ions specifically to initiate the ion changes, whereas the needle prick method uses direct neural activation, which may require sodium ion exchange as the crucial element in the ion changes, rather than potassium, or maybe supramaximal stimulation with the needle prick response. The vascular changes and neuronal discharge that accompany spreading depression have always been described as a passive response to the metabolic changes that occur during spreading depression, because they occur during or after the spread (Brennan et al., 2007; Leao, 1944), therefore ultimately measurements of electrical changes have underpinned CSD research. A recent publication may change this view point and also alter the perception of the importance of the vascular changes that occur. With regard to migraine, the CSD hypothesis postulated that neuronal responses were the primary cause and vascular changes were secondary (Lauritzen, 1994). Indeed, subsequent fMRI work has implied a neuronal component to migraine aura in patients (Cutrer et al., 1998; Hadjikhani et al., 2001). Blood flow changes as a consequence of spreading depression are undeniable, as measured by laser Doppler flowmetry, but they lack the parallel spatial and temporal resolution of optic intrinsic signal imaging (Brennan et al., 2007; Goadsby, 2007). Such studies by Brennan and colleagues have found that vasomotor changes, including blood flow changes travel at significantly greater velocities than the neuronal changes and may even precede the neuronal changes (Brennan et al., 2007). This data suggests dissociation between the vascular response and metabolic/neuronal response in the brain in the surface cortical changes that take place during spreading depression, with possibly independent mechanisms of action. Separate mechanisms of action would certainly go some way to explain the differences in responses to TTX with regard vascular and neuronal changes. Whilst the Brennan paper is not advocating a complete rethink on the causes of cortical spreading depression, the suggestion of more than one mechanism for the induction and propagation of the response is possible. This may be particularly important with migraine aura, it has always been considered that aura is a response to the cortical blood flow wave that has been observed through various techniques (Brennan et al., 2007; Cutrer et al., 1998; Hadjikhani et al., 2001); (Lauritzen, 1994), but it is cortical spreading depression caused by the neuronal metabolic changes that underlies the vascular changes. This theory potentially provides a fresh view of studies that have employed vascular changes to measure spreading depression. Further investigations are necessary probably with a combination of both neuronal and vascular measurements to disentangle these issues. Thus far it is clear that using different induction methods for CSD produces differential response to therapeutics. Mechanical stimulation and electrical stimulation appear to be responsive to Na+ ion channel inhibitors, while KCl and electrical induction are responsive to VDCCs. Perhaps, again we may find that different modes of induction rely on different mechanisms of spreading depression, and may explain the differential therapeutic response.

BR A IN RE S E A RCH 1 2 29 ( 20 0 8 ) 2 7 –3 6

In clinical studies migraine aura is primarily studied as changes in cerebral blood flow (Lauritzen, 1994; Olesen, 1991; Olesen et al., 1981), so the study of cerebral blood flow as a surrogate where the primary purpose of the investigation is to probe for new therapeutic opportunities seems appropriate. Mechanical stimulation as an approach relies on the lengthy refractory period as the wave continues to spread across the cortex as similar to the visual aura experienced by patients. Whilst it is acknowledged that some migraineurs can experience multiple aura attacks within the one migraine, the fact that aura more often occurs as a single incidence in migraine, and hyperemia cannot be induced during this wave, supports the use of this method as a model of migraine aura. Limitations are the potential neuronal damage caused by induction, although it has been shown that repeated inductions do not affect time to initiate CSD and therefore no additional harm is caused to the brain (Akerman and Goadsby, 2005). While it is clear that none of the models, including mechanical, electrical or chemical fully represent the etiology of aura, it seems attractive to have more than one approach to the question. Despite the differences in the models, what seems to be clear is that no single ion channel or action is exclusively responsible for the events that take place. Although much of this may be explained by different mechanisms that underlie vascular and metabolic changes (Brennan et al., 2007). Using mechanical stimulation no compound has been able to inhibit the response in every animal, only alter the response in a significant proportion. Interestingly a similar result is seen with topiramate (Akerman and Goadsby, 2005) as was observed with TTX in this study, in that not all animals responded to the TTX treatment, this has also been shown previously (Sugaya et al., 1975). In summary, sodium ion channels seem to play a role in the initiation of rCBF changes induced by mechanical stimulation of the cerebral cortex, although they do not appear to be solely responsible for the wave that occurs, as in other models of spreading depression TTX is less successful at inhibiting the depolarisation that occurs in CSD and also sometimes the rCBF changes (Fabricius et al., 1995; Sugaya et al., 1975). ATP-dependent K+ and VDCC ion channels do not appear to be crucial to the changes induced by this method. From the data presented here and from the KCl and mutant mouse studies, therapies that are able to manipulate sodium, potassium and calcium ion channels are each worthy of careful exploration in the treatment of migraine aura.

4.

Experimental procedures

4.1.

Surgical preparation for cats

Eleven male and eight female cats of mass 3.0 ± 0.1 kg were anesthetized using α-chloralose (60 mg kg− 1, intraperitoneally) and prepared for physiological monitoring. Halothane (0.5–2% in 40% O2-enriched air) was administered during surgical procedures and discontinued during experimental protocols. The left femoral artery and vein were cannulated for blood pressure recording and infusion of anesthetic and test

33

compounds, respectively. Supplementary doses of α-chloralose in 2-hydroxy-β-cyclodextrin (Storer et al., 1997) were given at a rate of 10 mg kg− 1hr− 1. The cats were intubated after local anesthesia with lidocaine and placed in a stereotaxic frame. Temperature was maintained throughout using a homeothermic blanket system (CWE Inc., USA). The cats were ventilated with oxygen-enriched air, 35 ml, 24 strokes per minute (Ugo Basile, UK). End-tidal CO2 was monitored (Capstar-100, CWE Inc., USA) and kept between 2.5–3.5% and blood pressure was monitored continually. This allows one to monitor for changes to respiration and blood pressure due to long-term anesthetic maintenance. Depth of anesthesia was monitored by testing the sympathetic (papillary and cardiovascular) responses to noxious stimulation and withdrawal reflexes. All experiments were conducted consistent with the UK Home Office Animals (Scientific Procedures) Act (1986). For the preparation of the cranium for blood flow monitoring the pericranial musculature was retracted laterally following a midline skin incision. A small unilateral rectangular burr hole was made (20 × 5 mm) over the left parietal cortex, 5 mm from the midline with a low rotational speed, saline-cooled drill, to prevent thermal injury. At the occipital end of the burr hole the dura mater was carefully removed to provide access for a 26 gauge needle, mounted in a stereotaxic micromanipulator, for mechanical stimulation. More rostrally two laser Doppler flow probes (Moor Instruments, UK) were lowered perpendicular to the brain surface on to the dura mater, 5– 10 mm apart, to measure cortical flow, known as probe 1 and probe 2 (Fig. 2A).

4.2.

Surgical preparation for rats

Eight male Sprague–Dawley rats (315–470 g) were anesthetized throughout the experiments with sodium pentobarbitone (Sigma-Aldrich, UK, 60 mg kg− 1 i.p. and then 18 mg− 1kg− 1hri.v. infusion). Again the left femoral artery and vein were cannulated for blood pressure recording and intravenous infusion of anesthetic and test compounds, respectively. Temperature was maintained throughout using a homeothermic blanket system (CWE Inc., USA). The rats were placed in a stereotaxic frame and ventilated with oxygenenriched air, 3–5 ml, 60–80 strokes per minute (Small Rodent Ventilator – Model 683, Harvard Instruments, UK). End-tidal CO2 was monitored (Capstar-100, CWE Inc., USA) and kept between 3.5–4.5% and blood pressure was monitored continually. The parietal bone was drilled so the dural blood vessels were clearly visible but the skull remained intact. At the occipital end of the drilled area a small region of bone and dura were removed to allow insertion of a tungsten recording electrode (WPI, UK, impedance 1.0 MΩ, tip diameter 0.5 μm) to a depth of 500–1000 μm below the surface to monitor cortical single-cell activity. Electrical signals were amplified and passed through filters and a 50 Hz noise eliminator (Humbug, Quest Scientific, Canada). The signal was fed into a gated window discriminator and an analog-to-digital converter (Power 1401plus, CED, Cambridge, UK) to a personal computer. Filtered and amplified signal is also fed for audio monitoring and displayed on an oscilloscope. A laser Doppler probe was placed anterior to the recording electrode, perpendicular to

34

BR A IN RE S EA RCH 1 2 29 ( 20 0 8 ) 2 7 –36

the skull surface, to measure CBF changes. Also, just anterior to bregma a burr hole was created on the bone and dura mater removed so a 26 gauge needle could be inserted to cause cortical lesioning and induce spreading depression, this method has been used previously to reliably induce cortical blood flow changes as a correlate of migraine aura (Akerman and Goadsby, 2005; Bolay et al., 2002; Lambert et al., 2008) (Fig. 2B). In all experiments cerebral blood flow was continuously measured using the laser Doppler probes (Moor Instruments, UK) and displayed with electrical and physiological monitoring on an online data analysis system (CED spike2 v5 software).

4.4.

Data analysis

In some cats the CBFLDF response to hypercapnia was monitored to assess whether CSD had been accidentally elicited during surgery, since CSD prevents blood flow responses to hypercapnia for up to 8 h in the cat (Kaube and Goadsby, 1994; Piper et al., 1991). The animals were left for a least one hour for physiological monitoring to stabilise, and to allow a refractory period in case spreading depression and blood flow changes had been elicited during surgery. A 26 gauge needle was plunged into the cortex to a depth of 1 mm using the micromanipulator, and immediately withdrawn. In the cat there was a 60 min interval, in the rat a 45 min interval, between needle plunges to avoid cortical refractory periods (Akerman and Goadsby, 2005; Kaube and Goadsby, 1994; Kaube et al., 1999). Each animal received a control plunge, followed by a plunge after 0.25 ml saline (i.v.) to ensure the reproducibility of the response.

Regional cerebral blood flow changes were observed and recorded online and evidence of cortical spreading depression observed as a spike of electrical activity caused by depolarization, followed by electrical depression (Fig. 2C). It is important to note that previous studies have shown the time-relation of neuronal discharges/spiking and cortical spreading depression are not always coincident (Sugaya et al., 1975; Tobiasz and Nicholson, 1982), indicating that the spread is probably not due to the discharge, but might be another result of the spreading depression and glial depolarization itself (Sugaya et al., 1975). To quantify the rCBF changes we measured the area under the curve (AUC) between the beginning of the hyperemic response until blood flow had returned to baseline, and the change in electrical activity from single cell recordings. These were compared to a baseline level of recording, taken as the AUC of the cerebral blood flow output or cell firing (Hz) for the minute prior to the needle plunge. These measurements were compared in all animals using a Student's paired t-test to determine that there was significant change in rCBF. The response across the cohort of control, saline, test scenario and post needle plunge (where necessary) were tested using analysis of variance (ANOVA) for repeated measures with Bonferroni correction applied, followed by Student's paired ttest if any significance was found. Speed of propagation was derived by the difference in the latencies to the beginning of the hyperemic phase of the probes and was measured using ANOVA for repeated measure with Bonferroni correction applied, and followed by Student's paired t-test if any significance was found.

4.3.1.

4.5.

4.3.

Experimental protocols

Cat studies

In the cat the N-type voltage-dependent calcium channel (VDCC) blocker ω-conotoxin-GVIa (20 μg kg− 1), the P/Q-type VDCC blocker ω-agatoxin-IVa (20 μg kg− 1) and the L-type VDCC blocker calciseptine (20 μg kg− 1) as well as the non-selective calcium channel blocker cadmium chloride (20 μg kg− 1) were all administered 5 min prior to a needle plunge. If a hyperemic response was not observed the needle plunge was repeated 60 min later to see if the hyperemic response returned. If the hyperemic response had occurred a further dose of blocker was administered and needle plunge repeated. Tetrodotoxin (60 μg kg− 1), the sodium ion channel blocker was given as an intravenous bolus 5 min prior to needle plunge. Needle plunge was repeated 60 min later and every 60 min if the hyperemic response was inhibited, until the response returned. Glibenclamide (30 mg kg− 1), the blocker of ATP-dependent potassium channels was given intraperitoneally 15 min prior to needle plunge. Again the needle plunge was repeated 60 min later if the hyperemic response was inhibited.

4.3.2.

Rat studies

A second series of experiments took place in the rat to measure the effects of the Na+ ion channel blocker, tetrodotoxin (10 μg kg− 1) on the electrical activity that can accompany cortical spreading depression (Fig. 2C). Tetrodotoxin was again given 5 min prior to the needle plunge and the needle plunge was repeated after 45 min if the spreading depression was inhibited, or did not occur.

Drugs

The infusion of anesthetic and test compounds was via the same catheter, however, the line was always flushed with saline first, several minutes before administering the different compounds. ω-Conotoxin-GVIa, which is isolated in the marine snail, ω-agatoxin-Iva, isolated in the funnel web spider (both Peptide Institute, UK), calciseptine, which is isolated from the black mamba, and tetrodotoxin (isolated from the puffer fish) (both Sigma-Aldrich, UK) were all dissolved in deoxygenated water, aliquotted and frozen until required. They were then re-dissolved in 0.9% saline to make up to a volume of 0.2 ml before intravenous administration. The doses used were as used previously in a model of trigeminovascular nociception (Akerman et al., 2003), except tetrodotoxin, where the maximum tolerated dose was given in each animal, and it was found that in the rat this was much lower. Cadmium chloride (Sigma-Aldrich, UK) was dissolved in 0.9% NaCl. Glibenclamide (Sigma-Aldrich Ltd, UK) was dissolved in dimethyl sulphoxide (DMSO), and was administered intraperitoneally as the solution did not dissolve in aqueous solution, and because intravenous DMSO can be fatal in high enough volume. Therefore it was administered 15 min prior to CSD induction with needle prick, to allow time for absorption. α-chloralose (SigmaAldrich, UK) was dissolved in either water for injection or 2hydroxy-β-cyclodextrin (Sigma-Aldrich, UK) (Storer et al., 1997). Sodium pentobarbitone (Sigma-Aldrich, UK) dissolved

BR A IN RE S E A RCH 1 2 29 ( 20 0 8 ) 2 7 –3 6

in 0.9% NaCl. Lidocaine hydrochloride (Intubeaze, Arnolds, UK) and halothane (May and Poulenc, Dagenham, UK) were also used.

Acknowledgments The authors would like to thank Kevin Shields and Paul Hammond for both assistance and technical support during these experiments.

REFERENCES

Akerman, S., Goadsby, P.J., 2005. Topiramate inhibits cortical spreading depression in rat and cat: impact in migraine aura. NeuroReport 16, 1383–1387. Akerman, S., Holland, P.R., Goadsby, P.J., 2005. 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 25, 871. Akerman, S., Williamson, D., Goadsby, P.J., 2003. Voltage-dependent calcium channels are involved in neurogenic dural vasodilation via a pre-synaptic transmitter release mechanism. Br. J. Pharmacol. 140, 558–566. Andlin-Sobocki, P., Jonsson, B., Wittchen, H.U., Olesen, J., 2005. Cost of disorders of the brain in Europe. Eur. J. Neurol. 12, 1–27. Ayata, C., Jin, H., Kudo, C., Dalkara, T., Moskowitz, M.A., 2006. Suppression of cortical spreading depression in migraine prophylaxis. Ann. Neurol. 59, 652–661. Ayata, C., Shimizu-Sasamata, M., Lo, E.H., Noebels, J.L., Moskowitz, M.A., 2000. Impaired neurotransmitter release and elevated threshold for cortical spreading depression in mice with mutations in the alpha 1A subunit of P/Q type calcium channels. Neuroscience 95, 639–645. Bolay, H., Reuter, U., Dunn, A.K., Huang, Z., Boas, D.A., Moskowitz, M.A., 2002. Intrinsic brain activity triggers trigeminal meningeal afferents in a migraine model. Nat. Med. 8, 136–142. Brennan, K.C., Beltran-Parrazal, L., Lopez Valdes, H.E., Theriot, J., Toga, A., Charles, A.C., 2007. Distinct vascular conduction with cortical spreading depression. J. Neurophysiol. 97, 4143–4151. Chen, W.T., Fuh, J.L., Lu, S.R., Wang, S.J., 2001. Persistent migrainous visual phenomena might be responsive to lamotrigine. Headache 41, 823–825. Cutrer, F.M., Sorensen, A.G., Weisskoff, R.M., Ostergaard, L., Sanchez del Rio, M., Lee, J., Rosen, B.R., Moskowitz, M.A., 1998. Perfusion-weighted imaging defects during spontaneous migrainous aura. Ann. Neurol. 43, 25–31. De Fusco, M., Marconi, R., Silvestri, L., Atorino, L., Rampoldi, L., Morgante, L., Ballabio, A., Aridon, P., Casari, G., 2003. Haploinsufficiency of ATP1A2 encoding the Na+/K+ pump a2 subunit associated with familial hemiplegic migraine type 2. Nat. Genet. 33, 192–196. Dichgans, M., Freilinger, T., Eckstein, G., Babini, E., Lorenz-Depiereux, B., Biskup, S., Ferrari, M.D., Herzog, J., van den Maagdenberg, A.M., Pusch, M., Strom, T.M., 2005. Mutation in the neuronal voltage-gated sodium channel SCN1A causes familial hemiplegic migraine. Lancet 366, 371–377. Ebersberger, A., Schaible, H.-G., Averbeck, B., Richter, F., 2001. Is there a correlation between spreading depression, neurogenic inflammation, and nociception that might cause migraine headache? Ann. Neurol. 41, 7–13. Fabricius, M., Akgoren, N., Lauritzen, M., 1995. Arginine-nitric oxide pathway and cerebrovascular regulation in cortical spreading depression. Am. J. Physiol. 269, H23–H29.

35

Goadsby, P.J., 2007. Cortical spreading depression— better understanding and more questions. J. Neurophysiol. 97, 3827. Goadsby, P.J., Lipton, R.B., Ferrari, M.D., 2002. Migraine — current understanding and treatment. N. Engl. J. Med. 346, 257–270. Hadjikhani, N., Sanchez del Rio, M., Wu, O., Schwartz, D., Bakker, D., Fischl, B., Wong, K.K., Cutrer, F.M., Rosen, B.R., Tootell, R.B. H., Sorensen, A.G., Moskowitz, M.A., 2001. Mechanisms of migraine aura revealed by functional MRI in human visual cortex. Proc. Natl. Acad. Sci. U. S. A. 98, 4687–4692. Headache Classification Committee of The International Headache Society, 2004. The international classification of headache disorders (second edition). Cephalalgia 24, 1–160. Kaube, H., Goadsby, P.J., 1994. Anti-migraine compounds fail to modulate the propagation of cortical spreading depression in the cat. Eur. Neurol. 34, 30–35. Kaube, H., Knight, Y.E., Storer, R.J., Hoskin, K.L., May, A., Goadsby, P.J., 1999. Vasodilator agents and supracollicular transection fail to inhibit cortical spreading depression in the cat. Cephalalgia 19, 592–597. Kow, L.M., van Harreveld, A., 1972. Ion and water movements in isolated chicken retinas during spreading depression. Neurobiology 2, 61–69. Kraig, R.P., Nicholson, C., 1978. Extracellular ionic variations during spreading depression. Neuroscience 3, 1045–1059. Lambert, G.A., Hoskin, K.L., Zagami, A.S., 2008. Cortico-NRM influences on trigeminal neuronal sensation. Cephalalgia 28, 640–652. Lauritzen, M., 1994. Pathophysiology of the migraine aura. The spreading depression theory. Brain 117, 199–210. Lauritzen, M., 2001. Cortical spreading depression in migraine. Cephalalgia 21, 757–760. Lauritzen, M., Hansen, A.J., 1992. The effect of glutamate receptor blockade on anoxic depolarization and cortical spreading depression. J. Cereb. Blood Flow Metab. 12, 223–229. Lauritzen, M., Rice, M.E., Okada, Y., Nicholson, C., 1988. Quisqualate, kainate and NMDA can initiate spreading depression in the turtle cerebellum. Brain Res. 475, 317–327. Leao, A.A.P., 1944. Pial circulation and spreading activity in the cerebral cortex. J. Neurophysiol. 7, 391–396. Leao, A.A.P., 1944. Spreading depression of activity in cerebral cortex. J. Neurophysiol. 7, 359–390. Lipton, R.B., Stewart, W.F., Diamond, S., Diamond, M.L., Reed, M., 2001. Prevalence and burden of migraine in the United States: data from the American Migraine Study II. Headache 41, 646–657. Marrannes, R., Willems, R., De Prins, E., Wauquier, A., 1988. Evidence for a role of the N-methyl-D-aspartate (NMDA) receptor in cortical spreading depression in the rat. Brain Res. 457, 226–240. Martins-Ferreira, H., Nedergaard, M., Nicholson, C., 2000. Perspectives on spreading depression. Brain Res. Rev. 32, 215–234. Menken, M., Munsat, T.L., Toole, J.F., 2000. The global burden of disease study — implications for neurology. Arch. Neurol. 57, 418–420. Nellgard, B., Wieloch, T., 1992. NMDA-receptor blockers but not NBQX, an AMPA-receptor antagonist, inhibit spreading depression in the rat brain. Acta Physiol. Scand. 146, 497–503. Newcomb, R., Abbruscato, T.J., Singh, T., Nadasdi, L., Davis, T.P., Miljanich, G., 2000. Bioavailability of Ziconotide in brain: influx from blood, stability, and diffusion. Peptides 21, 491–501. Olesen, J., 1991. Cerebral and extracranial circulatory disturbances in migraine: pathophysiological implications. Cerebrovasc. Brain Metab. Rev. 3, 1–28. Olesen, J., Larsen, B., Lauritzen, M., 1981. Focal hyperemia followed by spreading oligemia and impaired activation of rCBF in classic migraine. Ann. Neurol. 9, 344–352. Ophoff, R.A., Terwindt, G.M., Vergouwe, M.N., van Eijk, R., Oefner, P.J., Hoffman, S.M.G., Lamerdin, J.E., Mohrenweiser, H.W., Bulman, D.E., Ferrari, M., Haan, J., Lindhout, D., van Ommen,

36

BR A IN RE S EA RCH 1 2 29 ( 20 0 8 ) 2 7 –36

G.J.B., Hofker, M.H., Ferrari, M.D., Frants, R.R., 1996. Familial hemiplegic migraine and episodic ataxia type-2 are caused by mutations in the Ca2+ channel gene CACNL1A4. Cell 87, 543–552. Piper, R.D., Lambert, G.A., Michalicek, J., 1991. Inhalational anesthetic agents decrease susceptibility to cortical spreading depression in the cat. Proc. Soc. Neurosci. 17, 865. Rasmussen, B.K., Olesen, J., 1992. Migraine with aura and migraine without aura: an epidemiological study. Cephalalgia 12, 221–228. Richter, F., Ebersberger, A., Schaible, H.G., 2002. Blockade of voltage-gated calcium channels in rat inhibits repetitive cortical spreading depression. Neurosci. Lett. 334, 123–126. Shimazawa, M., Hara, H., 1996. An experimental model of migraine with aura: cortical hypoperfusion following spreading depression in the awake and freely moving rat. Clin. Exp. Pharmacol. Physiol. 23, 890–892. Shimizu, K., Bari, F., Busija, D.W., 2000. Glibenclamide enhances cortical spreading depression-associated hyperemia in the rat. NeuroReport 11, 2103–2106. Somjen, G.G., 2001. Mechanisms of spreading depression and hypoxic spreading depression-like depolarization. Physiol. Rev. 81, 1065–1096. Steiner, T.J., Findley, L.J., Yuen, A.W.C., 1997. Lamotrigine versus placebo in the prophylaxis of migraine with and without aura. Cephalalgia 17, 109–112. Storer, R.J., Butler, P., Hoskin, K.L., Goadsby, P.J., 1997. A simple method, using 2-hydroxypropyl-b-cyclodextrin, of administering a-chloralose at room temperature. J. Neurosci. Methods 77, 49–53.

Sugaya, E., Takato, M., Noda, Y., 1975. Neuronal and glial activity during spreading depression in cerebral cortex of cat. J. Neurophysiol. 38, 822–841. Tobiasz, C., Nicholson, C., 1982. Tetrodotoxin resistant propagation and extracellular sodium changes during spreading depression in rat cerebellum. Brain Res. 241, 329–333. van den Maagdenberg, A.M.J.M., Pietrobon, D., Pizzorusso, T., Kaja, S., Broos, L.A.M., Cesetti, T., van de Ven, R.C.G., Tottene, A., van der Kaa, J., Plomp, J.J., Frants, R.R., Ferrari, M., 2004. A Cacna1a knock-in migraine mouse model with increased susceptibility to cortical spreading depression. Neuron 41, 701–710. Wu, Y.J., Boissard, C.G., Greco, C., Gribkoff, V.K., Harden, D.G., He, H., L'Heureux, A., Kang, S.H., Kinney, G.G., Knox, R.J., Natale, J., Newton, A.E., Lehtinen-Oboma, S., Sinz, M.W., Sivarao, D.V., Starrett Jr, J.E., Sun, L.Q., Tertyshnikova, S., Thompson, M.W., Weaver, D., Wong, H.S., Zhang, L., Dworetzky, S.I., 2003. (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. 46, 3197–3200. Xiong, L., Zheng, Y., Wu, M., Hou, L., Zhu, Z., Zhang, X., Lu, Z., 2003. Preconditioning with isoflurane produces dose-dependent neuroprotection via activation of adenosine triphosphateregulated potassium channels after focal cerebral ischemia in rats. Anesth. Analg. 96, 233–237. Zhang, E., Hansen, A.J., Wieloch, T., Lauritzen, M., 1990. Influence of MK-801 on brain extracellular calcium and potassium activities in severe hypoglycemia. J. Cereb. Blood Flow Metab. 10, 136–139.