Excitatory Inhibitory Balance

8/12/2019 Excitatory Inhibitory Balance

1/14

Feature Review

Migraine: a disorder of brainexcitatoryinhibitory balance?Dania Vecchia 1 and Daniela Pietrobon 1 , 2

1 Department of Biomedical Sciences, University of Padova, 35121 Padova, Italy2 CNR Institute of Neuroscience, 35121 Padova, Italy

Migraine is a common disabling brain disorder whosekey manifestations are recurrent attacks of unilateralheadache and interictal hypersensitivity to sensory sti-muli. Migraine arises from a primary brain dysfunctionthat leads to episodic activation and sensitization of thetrigeminovascular pain pathway and as a consequence

to

headache.

Major

open

issues

concern

the

molecularand cellular mechanisms of the primary brain dysfunc-tion(s) and of migraine pain. We review here our currentunderstanding of these mechanisms, focusing on recentadvances regarding migraine genetics, headache mech-anisms, and the primary brain dysfunction(s) underlyingmigraine onset and susceptibility to cortical spreadingdepression, the neurophysiological correlate of migraineaura. We also discuss insights obtained from the func-tional analysis of familial hemiplegic migraine mousemodels.

IntroductionMigraine is a common episodic neurological disorder withcomplex pathophysiology that manifests itself as recurrentattacks of typically throbbing and unilateral, often severe,headache with associated features such as nausea, phono-phobia and/or photophobia; in a third of patients the head-ache is preceded by transient neurological symptoms thatare most frequently visual but may involve other senses(migraine with aura: MA) [1] (Table 1 ). Migraine is a publichealth problem of great impact upon boththe individual andsociety. It is one of the 20 most disabling diseases (according to World Health Organization ranking [2]). Furthermore, itis remarkably common (e.g., it affects 17% of femalesand 8%of males in the European population [3]) and very costly (EUR 18.5 billion/year in Europe [4]).

It

is

generally

believed

that

migraine

headache

dependson the activation and sensitization of the trigeminovascu-lar pain pathway [57] (Figure 1 ), and that cortical spread-ing depression (CSD)-like events underlie migraine aura[5,8,9] . CSD can be induced in animals by focal stimulationof the cerebral cortex and consists of a slowly propagating (26 mm/min) wave of strong neuronal and glial depolari-zation whose mechanisms of initiation and propagationremain unclear [10,11] . It is also generally recognized thatmost migraine attacks start in the brain. This is suggestedby the premonitory symptoms (such as difculty with

speech and reading, increased emotionality, sensory hypersensitivity) which in many patients are highly predictive of the attack although occurring up to 12 hbefore it [12] as well as by the nature of some typicalmigraine triggers (e.g., stress, sleep deprivation, oversleep-ing, hunger and/or prolonged sensory stimulation) [13] .

Psychophysical

and

neurophysiological

studies

have

provided clear evidence that in the period between attacksmigraineurs show hypersensitivity to sensory stimuli andabnormal processing of sensory information, characterizedby increased amplitudes and reduced habituation of evoked and event-related potentials [14,15] .

The nature and mechanisms of the primary brain dys-function(s) leading to the onset of a migraine attack, toCSD susceptibility, and to episodic activation of the trige-minovascular pain pathway remain largely unknown andare major outstanding issues to be addressed in furthering our understanding of the neurobiology of migraine. Otherimportant open questions concern the mechanisms of mi-graine pain.

Here, we review recent advances regarding (i) the ge-netics of migraine; (ii) the mechanisms of migraine head-ache, focusing on the roles of meningeal inammation,calcitonin gene-related peptide (CGRP), central sensitiza-tion, and dysfunctional central control of pain; and (iii) themechanisms of the primary brain dysfunction(s) leading toepisodic activation of the trigeminovascular pain pathway.We also discuss insights into these mechanisms obtainedfrom the functional analysis of mouse models of familialhemiplegic migraine (FHM), a rare monogenic autosomaldominant form of MA.

Genetics of migraine

Migraine

is

a

complex

genetic

disorder,

with

heritability estimates as high as 50% and probable polygenic multifac-torial inheritance [16,17] . The complexity of the diseasehas hampered the identication of common susceptibility variants; the lack of consensus on most of the identiedsusceptibility loci probably reects clinical and geneticheterogeneity [16,17] .

Most of our current understanding of genetic factorsunderlying migraine comes from studies of FHM. Threecausative genes, all encoding ion channels or transporters,have been identied [16,1821] . Additional FHM genescertainly exist and remain to be identied [22] . Apart fromthe motor weakness or hemiparesis during aura, typicalFHM attacks resemble MA attacks (Table 1 ) and both

Review

Corresponding author: Pietrobon, D. ( [email protected] ). Keywords: migraine; trigeminovascular pain; spreading depression; excitatoryinhibitory balance; channelopathy..

0166-2236/$ see front matter 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.tins.2012.04.007 Trends in Neurosciences, August 2012, Vol. 35, No. 8 507
mailto:[email protected]://dx.doi.org/10.1016/j.tins.2012.04.007http://dx.doi.org/10.1016/j.tins.2012.04.007mailto:[email protected]


2/14

types of attacks may alternate in patients and co-occurwithin families, suggesting that FHM and MA may be partof the same spectrum and may share some pathogeneticmechanisms. Some FHM patients can also have atypicalsevere attacks and/or permanent cerebellar symptoms[16,21] .

FHM1 is caused by missense mutations in CACNA1A ,the gene encoding the pore-forming subunit of neuronalCa V 2.1 (P/Q-type) voltage-gated calcium channels [18,23] .These calcium channels are widely expressed in the nervous system, including all brain regions implicated in thepathogenesis of migraine. Ca V 2.1 channels play a domi-nant role in controlling neurotransmitter release, particu-larly at central synapses. Their somatodendriticlocalization points to additional postsynaptic roles, suchas in neural excitability [23,24] .

Analysis of the single channel properties of mutant re-combinant human Ca V 2.1 channels and of the P/Q-typecalcium current in different neurons [including corticaland trigeminal ganglion (TG) neurons] of knockin mice

carrying FHM1 mutations revealed that the mutationsproduce gain-of-function of Ca V 2.1 channels, mainly dueto increased channel open probability and channel activa-tion at lower voltages [23,2531] . However, the gain-of-function effect may be dependent on the specic Ca V 2.1splice variant and/or auxiliary subunit [32,33] . In TG neu-rons of FHM1 knockin mice the P/Q-type calcium currentwas increased in a subtype of neuron (that does not innervate the dura) but was unaltered in capsaicin-sensitiveneurons innervating the dura; congruently, the FHM1 mu-tation did not alter depolarization-evoked CGRP releasefrom the dura, but increased CGRP release from trigeminalganglia [31] . In the cerebral cortex of FHM1 knockin mice,excitatory synaptic transmission was enhanced as a conse-quence of increased action potential-evoked glutamaterelease at pyramidal cell synapses; in striking contrast,inhibitory neurotransmission at fast-spiking (FS) interneu-ron synapses was unaltered (despite being initiated by P/Q-type calcium channels) [26] (Figure 2a). Neuron sub-type- and synapse-specic effects may help to explain why a

(a) (b) Efferent modulatory pathways

TG

TCC

RVM

vlPAGNCF

A11PH

S1 Ins

TNCC1,C2

Afferent pathways

Dura mater

Cerebralcortex

Pia mater

TCC

RVM

vlPAGNCF

S1 S2 Ins

TNCC1,C2

SSN

VPMPo

Hypothalamus

Thalamus

TG

TRENDS in Neurosciences

Figure 1 . Schematic illustration of important neuronalstructures and connections in the trigeminovascular pathways involvedin migraineheadache. (a) Afferent pathways.The central projections of the trigeminal ganglion (TG) neurons that innervate the meninges terminate in the so-called trigeminocervical complex (TCC) comprising the C1C2 dorsalhorns of thecervical spinalcord and thecaudal division of thespinal trigeminal nucleus (TNC) (C-fibersmainly in superficial layers; A- d fibersin deep layers). TheTCC makes direct ascending connections with different areas in the brainstem (including the superior salivatory nucleus, SSN, the ventrolateral periacqueductal grey,vlPAG, the nucleus cuneiformis, NCF) and with higher structures including several hypothalamic and thalamic nuclei, which in turn make ascending connections with thecortex [99,144,145] . Stimulation of the dural afferents in experimental animals results in activation of second-order trigeminovascular neurons (mainly in laminae I, II and V)in the TCC, as well as neurons in several bra instem (e.g ., SSN, vlPAG, rostral ventromedial medul la, RVM), hypo thalamic and tha lamic ( in part icu lar theventroposteriomedial, VPM and posterior, Po) nuclei receiving connections from the TCC [60,73,92,105,111] ([5,99] for review and older references ). Dura-sensitive VPMthalamic neurons project mainlyin thetrigeminal primary andsecondary somatosensory (S1, S2)and theinsular (Ins) cortex(in theso-called pain matrix areas), andthusarelikely to play a role in the perception of the headache; whereas trigeminovascularPo thalamic neurons project well beyond the pain matrix into non-trigeminal S1,aswell as auditory, visual, retrosplenial, ectorhinal, and parietal association cortices, and thus are likely to contribute to other aspects of the migraine experience whichinclude disturbances in neurological functions involved in vision, auditory, memory, motor, and cognitive performance [105,145] . The trigeminovascular projections to

specific hypothalamic and brainstem nuclei are likely to contribute to other aspects of the complex migraine symptomatology such as loss of appetite, sleepiness,irritability, stress, pursuit of solitude, and autonomic symptoms [99] . (b) Efferent modulatory pathways. The TCC receives descending projections from brainstem andhypothalamic nociceptive modulatory nuclei that may mediate descending modulation of trigeminovascular nociceptive traffic [99] . Experimental evidence of modulationof TCC response to dural stimulation has been obtained for vlPAG, the nucleus raphae magnus in the RVM, the posterior hypothalamus (PH) and the A11 dopaminergichypothalamic nucleus (reviewed in [99] ). The TCC also receives descending cortical projections from layer 5 pyramidal cells of the contralateral S1 (innervating mainlyneurons in deeplaminaeIIIV) and caudal Ins cortex (innervating exclusively laminae I and II) [111] . It shouldbe noted that there areother afferent and efferent connections:this diagram only illustrates those mentioned in the text.

Review Trends in

Neurosciences

August 2012, Vol. 35, No. 8

508


3/14

calcium channel that is widely expressed in the nervoussystem produces the specic dysfunctions that cause FHM(the implications for specic migraine mechanisms are dis-cussed in the sections following).

FHM2 is caused by (mainly missense) mutations in ATP1A2 , the gene encoding the a 2 subunit of the Na + /K +

ATPase [16,19] . In the brain, this isoform is expressedprimarily in neurons during embryonic development andat time of birth but almost exclusively in astrocytes in theadult; its colocalization and functional coupling with glialglutamate transporters in astrocytic processes surround-ing glutamatergic synapses suggests a specic role inglutamate clearance [3436] (Figure 2a), whereas its colo-calization with the Na + /Ca 2+ exchanger in microdomains

that

overlie

subplasmalemmal

endoplasmic

reticulum

sug-gests a specic role in the regulation of intracellular Ca 2+

[21] . FHM2 mutations cause complete or partial loss-of-function of recombinant Na + /K + ATPases due to loss orreduction of catalytic activity (and/or more subtle function-al impairments) or impairment of plasma membrane de-livery [21,3739] . The a 2 Na + /K + ATPase protein wasbarely detectable in the brain of homozygous FHM2knockin mice and strongly reduced in the brain of hetero-zygous mutants [39] .

FHM3 is caused by missense mutations in SCNA1A , thegene encoding the pore-forming subunit of neuronalNa V 1.1 voltage-gated sodium channels [16,20] ; these chan-nels are highly expressed in particular inhibitory inter-neurons where they play an important role in sustaining high-frequency ring [40] (Figure 2 a). Conicting ndingswere obtained from the analysis of mutant recombinanthuman Na V 1.1 channels expressed in non-neuronal cells,pointing to either gain- or loss-of-function effects of FHM3mutations [41,42] . Given the evidence that an epilepsy-causing mutation produced opposite effects on recombi-nant Na V 1.1 channels and native Na V 1.1 channels inneurons of mouse models [40] , functional analysis inFHM3 mouse models appears necessary to shed lighton FHM3 mechanisms.

Whereas most genetic studies indicate that the FHMgenes (except perhaps for ATP1A2 ) are not involved in

common migraines [16,21] , some homozygous mutations in SLC4A4 (the gene encoding the electrogenic Na + /HCO 3cotransporter NBCe1) were recently found to be associatedwith either hemiplegic migraine, MA, or migraine withoutaura (MO), depending on the mutation [43] . Only muta-tions producing near total loss-of-function of the transport-er expressed in glioma cells were associated with migraine,supporting a causative role and the view that hemiplegicand common migraine represent a phenotypic spectrumthat may share at least some genetic basis [43] (Figure 2 a).

A loss-of-function frameshift mutation in KCNK18 [thegene encoding the weakly inward rectifying K + channel(TWIK)-related spinal cord potassium channel (TRESK)],cosegregated perfectly with typical MA in a large multi-

generational

family

[44] . TRESK

channels

are

two-pore-domain K + channels that are broadly expressed in thenervous system, with particularly high expression in TG,where they probably play an important role in control of neuronal excitability [44] . However, a mutation leading tocomplete loss-of-function of the TRESK channel was re-cently found in both migraine and control cohorts, indicat-ing that non-functional TRESK channels are not sufcientto cause typical migraine alone [45] .

Recent genome-wide association studies have identieda few risk factors for both MA and MO that map within ornear transcribed regions of interesting genes. These in-clude metadherin ( MTDH ), a gene that regulates the ex-pression of GLT-1 [46] , an astrocyte glutamate transporterthat plays a major role in removal of glutamate at gluta-matergic synapses [47] , and TRPM8 [48] , a gene thatencodes a cation channel expressed primarily in sensory neurons and that is involved in cold-sensing [49] .

Taken together, genetic ndings suggest that migraineis a nervous system disorder characterized by alteredsynaptic (in particular glutamatergic) transmission and/ or altered neuronal excitability.

Mechanisms of migraine headacheBased on a large body of indirect evidence, it is believedthat the development of migraine headache depends on theactivation and sensitization of trigeminal sensory afferents

Table 1. Main features of migraine without aura (MO), migraine with aura (MA), and familial hemiplegic migraine (FHM)Type Headache symptoms Aura symptomsMO Headache attacks (472 h in duration)

with at least two of:Unilateral locationThrobbingModerate or severe pain intensityAggravation by physical activity

And at least one of:Nausea and/or vomitingPhotophobia and phonophobia

None

MA Headache as in MO begins duringthe aura or follows aura (within 1 h)

At least one of:Fully reversible visual symptoms (e.g., ickering lights, spots, lines and/or loss of vision)Fully reversible sensory symptoms (i.e., pins and needles and/or numbness)Fully reversible dysphasic speech disturbance

Each aura symptom lasts 5 and 60 minFHM Headache as in MA Fully reversible motor weakness and at least one of the MA aura symptoms

Each aura symptom lasts 5 min and < 24 h

Review Trends in

Neurosciences


509


4/14

that innervate cranial tissues, in particular the meningesand their large blood vessels [57] (Figure 1 a). The sensi-tization of mechanosensitive meningeal afferents providesa mechanism that may explain the throbbing nature of themigraine headache (typically attributed to arterial pulsa-tion) as well as the exacerbation of the headache during events that increase intracranial pressure [50] . The pri-mary mechanism(s) leading to activation and sensitizationof the perivascular trigeminal nociceptors remain incom-pletely understood and controversial, particularly in thecase of MO.

Vasodilation Infusion of vasodilator substances such as CGRP or glyceriltrinitrate (GTN) induce in migraineurs (but not in healthy subjects) a delayed migraine attack indistinguishable fromthe spontaneous attacks [51] . However, there is growing evidence that vasodilation of meningeal and/or extracrani-al arteries is neither necessary nor sufcient to causemigraine pain in most patients [52] . In contrast to aprevious study showing lack of signicant dilatation of extracranial and intracranial arteries during GTN-induced migraine [53] , a 912% dilatation of extracranial

(a)

(b)

N a +

HCO3

-

Astrocyte

Na V1.1?

H

Na

GABA AR

Ca V2.1

Inhibitory FS interneuron synapseExcitatory PC synapse

GluR NMDAR

Ca V2.1

FHM1

FHM3

Astrocyte

Na +G L U N a + , H +

K +

K +

Na +

H C O3

-

2 Na+,K+

ATPaseFHM2

FHM4 ? Na+, HCO 3

-

cotransporter

FHM1

FS interneuron

PC PC

FS interneuron

++

++

++

++

+ +

+ +


Figure 2 . Differential alteration of cortical excitatory and inhibitory synaptic transmission in FHM. (a) In FHM1, gain-of-function of presynaptic Ca V 2.1 channels leads toenhanced action potential-evoked glutamate release and enhanced excitatory synaptic transmission at cortical pyramidal cell (PC) synapses (left panel); inhibitory synaptictransmission at FS interneuron synapses is unaltered, despite being initiated by Ca V 2.1 channels (right panel) [26] . In FHM2, given the specific colocalization and functionalcoupling of a 2 Na + /K+ pumps with glial glutamate transporters in astrocyte processes surrounding excitatory, but not inhibitory, synapses, loss-of-function of the pumpmight impair glutamate clearance and lead to specific gain-of-function of excitatory transmission, particularly NMDA receptor (NMDAR)-mediated transmission duringhigh-frequency action potential (AP) trains [39] . A decreased capacity of astrocytes to buffer activity-dependent extracellular alkalosis as a consequence of loss-of-functionof the Na + /HCO3 cotransporter NBCe1 has been proposed to also lead to enhanced NMDAR-mediated excitatory synaptic transmission [43] . The consequences of FHM3mutations on Na V 1.1 channels in FS interneurons, where they are highly expressed, remain unknown. (b) Schematic representationof theeffect of FHM1 mutations on thespecific cortical subcircuit involving recurrent excitatory synapses between PCs and reciprocal excitatory and inhibitory synapses between PCs and FS interneurons.Synaptic transmission is enhanced at excitatory synapses (as indicated by the thicker connection in the right panel compared to the left panel) but is unaltered at theinhibitory synapsesin FHM1 mouse models [26] . One predicts that the gain-of-function of glutamate release at the recurrent synapsesbetweenPCs would certainlyincreasenetwork excitation. By contrast, the gain-of-function of glutamate release at the synapses onto FS interneurons (PCFS synapses) would lead to enhanced recruitment of interneuronsand enhancedinhibition.However, during high-frequency repetitive activity the enhancedrecruitment of FS interneuronsis expectedto ceaserapidlybecausethePCFS synapses have been shown to depress stronglyduring AP trains(muchmore than therecurrent PCPC synapses), and short-term depression waseven strongerin FHM1 knockin mice, particularly at PCFS synapses (whereas short-term plasticity at the inhibitory FSPC synapses was unaltered) [26] . This analysis, despite beingrestricted to a specific subcircuit, makes the important point that the differential effects of FHM1 mutations on excitatory and inhibitory neurotransmission may produce

overexcitation in certain brain conditions, but may leave the E/I balance within physiological limits in others, which is consistent with the episodic nature of the disease.

Review Trends in

Neurosciences


510


5/14

and intracranial arteries was recently measured in CGRP-induced migraine; this modest vasodilation is probably insufcient to activate the perivascular afferents but mightaffect sensitized nociceptors [54] .

Meningeal inammation Based on a large body of indirect evidence from both

clinical

and

animal

studies,

sterile

meningeal

inamma-tion is considered to be a key mechanism that may activateand sensitize perivascular meningeal afferents and lead tomigraine pain [7,55,56] . Indirect clinical evidence is provided by the increased level of various inammatory med-iators in the cephalic venous outow during spontaneousmigraine attacks and by the efcacy of nonsteroidal anti-inammatory drugs in the acute treatment of migraine inmany patients [7,55,56] . In experimental animals, activa-tion of meningeal nociceptors in vivo leads to release of vasoactive proinammatory peptides (such as CGRP andsubstance P) from their peripheral nerve endings. Thesepeptides result in vasodilation of meningeal blood vessels(mainly due to CGRP), plasma extravasation, and localactivation of dural mast cells (MCs), with ensuing releaseof cytokines and other inammatory mediators (resulting in neurogenic inammation, NI) [5,7] . The dural trigemi-nal afferents exhibit properties characteristic of nocicep-tors in other tissues, including chemosensitivity andsensitization [7,31,50,57,58] . In vivo , most dural afferentswere activated and sensitized by an inammatory soup(IF) applied to the dura; nearly all IF-sensitive duralafferents were mechanosensitive and their mechanosensi-tivity was enhanced by IF [59] . Chemical inammation of the dura produced facial and hind-paw cutaneous allody-nia in awake behaving animals [60] , with a time-course of development that is consistent with that seen in migraine

patients [61] . The pharmacology of the IF-induced allody-nia in animals shows important parallels with the clinicalpharmacology of migraine pain [60] . Further support to theinammation hypothesis is provided by the evidence thatdural MC degranulation per se can produce a long-lasting activation and sensitization of rat dural nociceptors [62] , aswell as cephalic tactile hypersensitivity [63] .

However, the endogenous processes that promote men-ingeal inammation and peripheral sensitization during migraine attacks remain incompletely understood. Many investigators consider the NI produced by release of vaso-active proinammatory neuropeptides following activationof peptidergic meningeal nociceptors (by CSD or otherdifferent primary mechanisms; next section) as the endog-enous inammatory process that sustains the activationand causes the long-lasting sensitization of meningealnociceptors in many migraine attacks. Indeed, measure-ments of CGRP levels in the external and internal jugular venous blood have provided evidence that CGRP isreleased during migraine attacks [6466] . Consistent withthe NI hypothesis is also the recent evidence that theheadache-triggering substances ethanol and umbellolone(the major volatile constituent of the Californian headachetree) both activate peptidergic meningeal trigeminalafferents, causing CGRP release and neurogenic durainammatory responses in experimental animals [67,68] .However, direct evidence that the release of inammatory

molecules associated with NI is able to sensitize meningealnociceptors is lacking.

In certain types of migraine, endogenous processesdifferent from NI, and not requiring initial activation of meningeal nociceptors, might promote meningeal inam-mation and cause sensitization, ensuing long-lasting activation of meningeal nociceptors. These processes may

include

direct

activation

of

dural

MCs

by

several

exoge-nous and endogenous migraine triggers, as was shown tooccur in vitro [7,56,69] . They may also include the release of inammatory mediators from brain parenchyma (e.g., as aconsequence of CSD) and/or from meningeal blood vesselsor immune cells, which might directly sensitize meningealnociceptors.

Central sensitization After headache onset, about two-thirds of migrainepatients develop cutaneous allodynia in the periorbitalregion that may spread to extracephalic regions [70,71] .Clinical observations and animal studies support the ideathat facial allodynia reects sensitization of trigeminovas-cular neurons in the trigeminocervical complex (TCC)which receive convergent input from the meningeal noci-ceptors and facial skin [5,70,72] . Extracephalic allodyniareects sensitization of trigeminovascular thalamic neu-rons that process converging sensory information from thecranial meninges and extracephalic skin [5,70,73] . More-over, these studies are consistent with the idea that initia-tion, but not maintenance, of central sensitization dependson the afferent input from sensitized meningeal nocicep-tors [5,61,70,72] . A recent study in rats points to theactivation of the descending facilitatory pathway arising from the rostral ventromedial medulla (RVM) (Figure 1 b)as a key central mechanism involved in IF-induced central

sensitization of trigeminovascular neurons [60] . Function-al magnetic resonance imaging (fMRI) studies in humansubjects indicate that the periacqueductal grey (PAG) andnucleus cuneiformis (NCF), the major sources of input tothe RVM (Figure 1 b), are involved in central sensitizationin humans [74] and that the NCF is hypofunctional inmigraine patients [75] . Interestingly, positron emissiontomography (PET) studies revealed activation of similarareas in the dorsal rostral brainstem during migraineattacks [76] .

CGRP Several ndings support a pivotal role of CGRP inmigraine, including (i) the effectiveness of CGRP receptorantagonists in migraine treatment [77,78] and (ii ) theinduction of a delayed migraine-like headache by intra- venous CGRP administrat ion in a large fraction of mi-graine patients but not in controls [79] , suggesting thatmost migraineurs are hypersensitive to CGRP-mediatedmodulation of nociceptive pathways. However, the mech-anisms underlying this hypersensitivity, the mechanismsof action of CGRP during a migraine attack, and the exactsites of action of CGRP receptor antagonists remainunclear and controversial [6466] . The localization of CGRP receptors in the trigeminovascular system pointsto multiple possible mechanisms at both peripheral andcentral sites [80] .

Review Trends in

Neurosciences


511


6/14

In the periphery, CGRP receptors are expressed in blood vessels, Schwann cells and dural MCs at the meninges andin glial satellite cells and a subpopulation of TG neurons inthe TG [80,81] . A relevant role of CGRP-induced dural vasodilation in migraine is unlikely in view of the evidencethat topical or systemic CGRP does not activate or sensi-tize rat dural afferents [82] and the evidence that vasodi-

lation

is

neither

necessary

nor

sufcient

to

triggermigraine [52] . CGRP-induced dural MC degranulationmight contribute to maintaining an inammatory cycleat the dura. Consistent with this idea are animal studiesshowing that dural MC degranulation (as well as topicalapplication of some individual MC mediators to the dura)preferentially activate and sensitize mechanosensitive Cunits, most of which express CGRP [62,83] , and increaseCGRP release from capsaicin-sensitive dural afferents[84] . The facilitation of CGRP-induced dural vasodilationand/or MC degranulation does not seem to contribute toheadache generation in FHM1 because depolarization-evoked CGRP release from the dura was unaltered inFHM1 knockin mice [31] .

It has been suggested that CGRP-mediated intragan-glionic crosstalk between neurons and satellite glial cells,may promote and maintain a neuron-glia inammatory cycle, that may contribute to persistent peripheral trigem-inal sensitization [64,66] . This suggestion is mainly basedon the evidence that prolonged application of CGRP tocultures of TG neurons and/or satellite glial cells leads toincreased gene expression and/or membrane targeting of specic receptors (e.g., P2X 3 ) in neurons and to increasedexpression of inammatory genes and release of inam-matory mediators from satellite glial cells; these inam-matory mediators can sensitize TG neurons and act back on glial cells further activating them [8590] . It remains

unclear whether similar phenomena indeed occur withinthe TG upon prolonged activation of meningeal nociceptorsin vivo . If they do, they might be facilitated in FHM1, assuggested by ndings in TG neurons of FHM1 knockinmice, that show enhanced P2X 3 receptor activity [91] andenhanced CGRP release from intact trigeminal gangliaand/or cultured neurons [31,90] . Moreover, there is someevidence that facilitation of CGRP-mediated neuron to gliacrosstalk may occur in cultured TG neurons from FHM1knockin mice following exposure to proinammatory sti-muli [90] .

Within the central trigeminovascular system, expres-sion of CGRP receptors has been shown in the trigeminalnucleus caudalis (TNC, laminae III, in a ber network forming irregular glomeruli-like structures but not in sec-ond-order neurons) [80] , and in some neuronal cell bodiesin the ventroposteromedial (VPM) nucleus of the thalamus[92] (Figure 1a). Functional studies indicate that activa-tion of TNC presynaptic CGRP receptors may lead topotentiation of excitatory neurotransmission [93,94] . Thepossibility of central mechanisms of CGRP action during amigraine attack is indirectly supported by animal studiesshowing that systemic CGRP does not activate or sensitizedural afferents [82] . Furthermore, high doses of systemicCGRP receptor antagonists reduce the activity of TNCneurons [95] and VPM thalamic neurons [92] evoked by stimulation of dural afferents, as well as the number of

Fos-positive neurons in TNC (laminae III) after intrave-nous infusion of capsaicin [96] . However, given the very poor permeability of the bloodbrain barrier to CGRP[97,98] , it seems difcult to explain how CGRP infusioncan cause a migraine attack if one excludes a peripheralrole of CGRP.

Dysfunctional

central

control

of

pain Direct evidence in the clinical setting for increased activity of trigeminal neurons during migraine is lacking and aconsistent cephalic pathology has not been detected.Therefore, some investigators propose the alternative viewthat migraine headache arises from dysfunction withinsubcortical brainstem and diencephalic nuclei that modu-late trigeminal nociceptive inputs (Figure 1 b). Dysfunctionin these nuclei (in particular in the PAGRVM circuitry)would lead to abnormal central interpretation of normalsensory input in the trigeminovascular system, causing normal sensory trafc from the meninges to be perceived asmigraine pain [99,100] . Functional imaging studies show-ing increased cerebral blood ow in the dorsal rostralbrainstem and in the hypothalamus during migraineattacks [76,101] are considered to provide indirect supportfor this view [99] . However, it appears more likely thatthese brainstem areas function as migraine headachemodulators rather than generators because their activa-tion does not seem specic for migraine pain [101,102] .Moreover, a recent fMRI study showed activation of dorsalrostral brainstem areas only during the migraine attack and not during the preictal phase [103] . Dysfunction inbrainstem nuclei involved in central control of pain andcentral sensitization [75,104] may lead to hyperexcitability of central trigeminovascular pathways and contribute tothe development of migraine headache. Evidence from

clinical and animal studies that questions the notion thatabnormalities in the PAGRVM circuitry (or other des-cending mechanisms of pain inhibition) can generate mi-graine headache in the absence of peripheral sensory inputhas been discussed in recent reviews [6,7] .

Photophobia A large fraction of migraineurs experience exacerbation of headache by light (i.e., photophobia). A neural mechanismfor migraine photophobia has been recently uncovered[105] . It has been shown that dura-sensitive thalamicneurons in the rat posterior thalamus receive monosynap-tic input from retinal ganglion cells (mainly intrinsically photosensitive cells involved in non-image-forming func-tions) and that light enhances the activity of dura-sensitivethalamic neurons located in the same area (Figure 3). Theidea that a non-image-forming retinal pathway is involvedin migraine photophobia is supported by the nding thatexacerbation of headache by light was preserved in blindmigraineurs who could sense light in the face of severedegeneration of rod and cone photoreceptors [105] .

Primary brain dysfunctions in migraineThe nature and mechanisms of the primary brain dysfunc-tion(s) leading to episodic activation of the trigeminovas-cular pain pathway remain incompletely understoodand controversial. Given the wide genetic and clinical

Review Trends in

Neurosciences


512


7/14

heterogeneity of the disorder, different primary mecha-nisms of migraine onset probably exist.

CSD Increasing evidence from animal studies support the ideathat CSD, the underlying mechanism of aura, can activatetrigeminal nociception and thus trigger the headachemechanisms. A direct nociceptive effect of CSD has beendemonstrated by the nding that a single CSD can lead to a

long-lasting

increase

in

ongoing

activity

of

dural

nocicep-tors and central trigeminovascular neurons in supercialand deep laminae of the TCC [106,107] . In most neurons,activation occurred with a delay consistent with that be-tween the onset of visual aura and the onset of headache;the delay as well as the magnitude and duration of neuro-nal activation were similar in peripheral and central neu-rons, suggesting that CSD-evoked activity of meningealnociceptors is sufcient to activate the central neurons.Immediate neuronal activation by CSD was observed in afraction of neurons, mainly C nociceptors and exclusively laminae I, II TCC neurons. This suggests that it might bemediated by peptidergic nociceptors with axon collateralsextending to the pia, where immediate activation could bemediated by increased K + or other noxious mediatorsreleased in the wake of the CSD wave. This hypothesisis supported by the demonstration that CSD-inducedCGRP release from perivascular trigeminal bers contrib-utes to the transient dilation of pial vessels measuredduring CSD [108] . The mechanism of the long-lasting and, in most neurons, delayed neuronal activation remainsunknown. One possibility is that release of proinamma-tory neuropeptides in the dura promotes NI that sustainsthe activation of meningeal nociceptors and leads to theirsensitization [7,106,107,109] . This idea is supported by thending of CSD-induced plasma protein extravasation fromdural blood vessels, which was abolished by trigeminal

nerve section [109] (but see [63,110] ). Different mecha-nisms that may potentially explain the delayed activationof dural afferents are discussed in [63] . The CSD-inducedactivation of central TCC neurons might be further modu-lated via direct cortexTCC connections, because it hasbeen shown that the reduction of cortical activity in pri-mary somatosensory and insular cortical areas following CSD results in reduced and enhanced responses of TCCneurons to noxious electrical stimulation of the dura,

respectively

[111] . Possibly,

this

top-down

modulation

of meningeal nociception may help to explain why somepeople experience migraine aura without headache.

In general, it remains unclear whether the activation of the trigeminovascular system induced by a CSD is sufcientto elicit the perception of headache in patients. It has beensuggested that it may not be sufcient on the basis of theobservation that freely moving rats do not seem to experi-ence CSD as being aversive because they do not show painbehavior [112,113] or cutaneous allodynia [114] aftera CSD(but see [115] ). The idea that CSD may trigger the headachemechanisms is indirectly supported by the nding that theelectrical stimulation threshold for induction of CSD in therat cortex increases after chronic treatment with ve differ-ent migraine prophylactic drugs thatareequally effective inreducing the frequency of MA and MO attacks [116] . More-over, two drugs ineffective in migraine prophylaxis did notaffect susceptibility to experimental CSD [116,117] . Howev-er, this correlation between inhibition of CSD and effective-ness in migraine prophylaxis does not appear to hold for twoother drugs (although larger clinical trials appear necessary for a denite conclusion) [118,119] .

The analysis of experimental CSD in FHM knockinmouse models has strengthened the view of CSD as akey migraine trigger by demonstrating that both FHM1and FHM2 knockin mice show a lower electrical stimula-tion threshold for CSD induction and a higher velocity of

Dura sensitive ( n = 20)

(a) (b)

3.60 4.16 4.52

4.80

LPMR LDVLLPLR

LPMR

Po

VPM VPL

Dura/light sensitive ( n = 14)Key:

Dura/light insensitive ( n = 14)

9

6

3

0

N u m

b e r o

f n e u r o n s

LP-LD/Po ( n = 11) Po ( n = 15) VPM-VPL ( n = 8)

Dura sensitive ( n = 6)

APTLPMR LPLCLPMC

PoVPM

LPMC

LPLCAPT PLi

PoT

PoVPM

VPL

3V 3V3V

Dura insensitive ( n = 14)

40

M e a n s p i

k e s p e r s 20

0

40

20

0Dark 500 Dark 50,000

Lux Lux

*

*


Figure 3 . A neural mechanism for exacerbation of migraine headache by light. (a) Extracellular single-unit electrophysiological recordings in deeply anesthetized ratsrevealed 20 neurons in the posterior thalamus that responded to stimulation of the dura, 14 of which were also photosensitive [105] . On average, dura-sensitive neuronsincreased their mean firing rate about twofold in responseto ambient fluorescence light (500 lux) and fourfold in response to bright light (50000 lux), By contrast, thalamicneurons unresponsive to stimulation of the dura were also unresponsive to light. (b) Histological analysis of the recording sites indicated that most dura/light-sensitive

neurons were localized at or above the dorsal border of the posterior thalamic nuclear group. Adapted, with permission, from [105] . LDVL, laterodorsal thalamic nucleus,ventrolateral; LPMR, LPLR, LPMC and LPLC, lateral posterior thalamic nuclei, mediorostral, laterorostral, mediocaudal and laterocaudal respectively; Po, posterior thalamicnuclear group; VPM, ventral posteromedial thalamic nucleus; VPL, ventral posterolateral thalamic nucleus; PLi,posterior limitans thalamicnucleus; PoT, posterior thalamicnuclear group, triangular; APT, anterior pretectal nucleus.

Review Trends in

Neurosciences


513


8/14

CSD propagation [27,28,39] (Figure 4a). In FHM1 knockinmice carrying the mild R192Q or the severe S218L muta-tion, the strength of CSD facilitation as well as the severity of the post-CSD neurological motor decits and the pro-pensity of CSD to propagate into subcortical structureswere all in good correlation with the strength of the gain-of-function of the Ca V 2.1 channel and the severity of theclinical phenotype produced by the two FHM1 mutations[27,28,120122] . The velocity of propagation and the fre-quency of CSDs, elicited by continuous epidural high KClapplication, were larger in female than in male FHM1mouse mutants, in agreement with the higher femaleprevalence of migraine; the sex difference was abrogatedby ovariectomy and enhanced by orchidectomy, suggesting that female and male gonadal hormones exert reciprocaleffects [121,123] . However, no gender differences in the

electrical threshold for CSD induction and the velocity of CSD propagation were found in FHM2 knockin mice [39] . Although FHM3 mouse models are not available, thereport that FHM3 in two unrelated families cosegregateswith a new eye phenotype with clinical features similar toexperimental spreading depression in retina [124] sug-gests that, probably, the ability to facilitate CSD is alsoshared by FHM3 mutations. Moreover, a lower electricalthreshold for CSD induction and increased velocity of CSDpropagation were measured in a mouse model of cerebralautosomal dominant arteriopathy, a systemic vasculopa-thy associated with 5-fold increased incidence of MA [125] .

Despite the strong support provided by animal studies,the idea that CSD may initiate the headache mechanismsin migraine (particularly in MO) is not generally accepted,mainly because it seems unable to explain some clinical

(a)

(b)

(i)

Stim1 2

wt1

30

CSD in FHM1 knockin mice in vivo CSD in FHM2 knockin mice in vivo

20

T h r e s

h o l d s t

i m u l u s

( C )

10

0

* * * *

* **10 50 10

8

6

4

20

40

30

20

10

-10

W T / W 8 8 7 R W

T

W T / W 8 8 7 R W

T

Evoked EPSC

20 ms 20 ms

0

8

6

4

V e l o c i

t y ( m m . m

i n - 1 )

C S D t h r e s

h o l d ( m i c r o

C )

C S D v e

l o c i

t y ( m m

/ m i n )

20

*

W T

W T

W T

W T

S 2 1 8 L / W T

S 2 1 8 L / S 2 1 8 L

R 1 9 2 Q / R 1 9 2 Q

S 2 1 8 L / W T

S 2 1 8 L / S 2 1 8 L

R 1 9 2 Q / R 1 9 2 Q

1

2

2

50 sec 2 5 m

V

R192Q KI

(ii) (iii)

(iv)

3.5******

******

WT

R192Q KI R192Q KI

Aga 40 nM

3.5250

200

150

100

50

0

250

200

150

100

C S D t h r e s h o

l d ( m s

)

C S D v e l o c

i t y

( m m

/ m i n )

50

20

500 m

KCIpuffer

Patch clampelectrode

0

-20

V ( m V )

-40

-60

-800

3.0

2.0

2.5

1.5

1.0

0.5

0.0WT

8 82117KIKIKIWT

22 22 10 10

KI KI+AgaKI+Aga

3.0

2.0

1.0

2.5

1.5

0.5

0.0

(iii)(ii)(i)


Figure4 . FHM mutations facilitate theinduction and the propagation of CSD. (a) Theanalysis of the thresholdfor initiation and therate of propagation of CSD, induced inanesthetizedmice by electrical stimulation of the visual cortex through a bipolar electrode, revealeda lower electrical stimulation threshold and a higher rateof propagationin both FHM1 and FHM2 knockin mice compared with wild-type (WT) mice [27,28,39] . Panel (i) shows the location of the stimulating and recording electrodes andrepresentative CSD recordings at sites 1 and 2 in WT and homozygous FHM1 knockin (KI) mice carrying the R192Q mutation; stimulation current pulses of increasingintensitywereapplied at 5 mininterval until a CSDwas observed (the chargedeliveredwith thefirststimulation elicitinga CSDwas taken as theCSD threshold). (ii) A lowerCSD threshold and a higher rate of CSD propagation were measured in KI mice carrying two different human FHM1 mutations (R192Q, causing typical FHM attacks, or

S218L,causing a severe hemiplegic migraine syndrome that is associated with ataxia, seizures, coma andsevere brain edemaoften triggeredby only a

mild head trauma)[27,28] . S218L KI mice showedbotha lowerthreshold forCSD inductionand a fasterrateof CSDpropagation compared with R192QKI mice.The facilitation of CSDwas alsodemonstratedto be dosage-dependent, with moresignificantdifferences observedin micehomozygous for the S218L KI comparedwith heterozygotes. (iii) Similar findingswere observed in KI mice carryingthe humanFHM2 mutationW887R [39] . Adapted, with permission, from [27] (i), [28] (ii) and [39] (iii). (b) The facilitation of induction andpropagation of CSD in acuteslices of somatosensory cortex of R192Q KI micewas completely eliminated whenglutamate release at pyramidal cellsynapseswas reduced toWT values using a subsaturating concentration of thespecific P/Qchannel blocker v -AgaIVA (Aga) [26] . (i) Pressure-ejection pulses of high KCl of increasing duration wereapplied at 8 mininterval through a glass micropipette on layer 2/3of acute slices of somatosensory cortexuntila CSD wasrecorded in a pyramidal cell at 600 mm from thepressure-ejection pipette; the duration of the first pulse eliciting a CSD was taken as the CSD threshold, and the rate of horizontal spread of the change in intrinsic opticalsignal as the velocity of CSD propagation. Similarly to observations made in vivo , the CSD threshold was lower (ii) and the CSD rate of propagation was higher (iii) inKIcompared to WT mice. After perfusion of slices from KI mice with 40nM Aga [a concentration that reduced the evoked EPSC recorded from KI pyramidal cells inmicroculture to the average value recorded from WT pyramidal cells, as shown by the representative traces in (iv) ], the CSD threshold increased (ii) and the CSD velocitydecreased (iii) to values strikingly similar to those measured in WT slices. Adapted, with permission, from [26].

Review Trends in

Neurosciences


514


9/14

observations, in particular the lack of a xed relationshipbetween aura and headache [5,9] . The possibility thatsilent CSDs (i.e., CSDs involving areas of the brain thatwould not generate a perceived aura) may initiate theheadache mechanisms in MO is neither proven nor dis-proven by current evidence [5,8,126] . Another argumentthat has been used against the idea that CSD may initiate

the

headache

mechanisms

is

based

on

the

fact

that

in somepatients migraine premonitory symptoms may occur up to1224 h before the onset of the headache and aura, indi-cating that different brain regions are activated well beforethe onset of CSD [12] .

In this context, it is interesting that the interictalneurophysiological abnormalities in sensory informationprocessing, typical of MO and MA patients, are not con-stant but change in intensity in temporal relation to themigraine attack. In most instances, the intensity of the decit reaches the maximal value in the 1224 h beforethe attack (i.e., the same time when the premonitory symptoms appear) and then normalize a few hours beforeand/or during the attack (with the exception of decits inpain processing) [5,14,15,103,127,128] . Neurophysiologicalreactivity to stress, one of the most common migrainetriggers, increases in the period between attacks and ismaximal (and signicantly higher than in healthy subjects) 13 d before an attack [127] . These data suggest thatin the brain of migraineurs some intrinsic mechanisms areat work during the pain-free interval that progressively increase the dysfunction in central information processing and increase the susceptibility to migraine triggers and theneurophysiological readiness to generate a migraine at-tack. It seems possible that these mechanisms may lead toboth the premonitory symptoms and, above a certainthreshold of cortical dysfunction and/or in response to

migraine triggers, create the conditions for ignition of CSD (see below).

Dysfunctional regulation of cortical excitationinhibition (E/I) balance and sensory information processing The analysis of interictal cortical excitability using psy-chophysical, electrophysiological, transcranial magneticstimulation (TMS) and fMRI has produced contradictory ndings and interpretations regarding the mechanismsunderlying the abnormal processing of sensory information(including trigeminal nociception) in migraineurs. It isbeyond the scope of the present review to discuss in detailthis very large and controversial literature ([5,14,15] forreviews, and e.g., [129133] for some recent studies).Depending on the study, it has been concluded that eitherthe cortex of migraineurs is hyperexcitable as a conse-quence of either enhanced excitation or reduced inhibition,or that it is hypoexcitable and/or has a lower preactivationlevel possibly due to serotonin hypoactivity and/or inef-cient thalamo-cortical drive. Methodological problems, het-erogeneity of the subjects and/or the time period relative tothe last and next migraine attack, and a lack of detailedunderstanding of the underlying mechanisms involved incentral information processing, probably account for thecontradictory ndings and interpretations.

Interestingly, recent TMS studies in MA patients pointto decient regulatory mechanisms of cortical excitability

and ensuing reduced ability to dynamically maintain thecortical E/I balance and to prevent excessive increases incortical excitation rather than merely hypo- or hyperex-citability as the mechanisms that underlie abnormalsensory processing [134136] . The molecular and cellularmechanisms underlying the abnormal regulation of corti-cal function and its periodicity remain largely unknown.

The

extent

to

which

some

of

the

cortical

and/or

subcorticalalterations are affected by disease duration (e.g., repetitiveCSDs) is also unclear. Equally unclear is the extent towhich the abnormal processing of trigeminal nociceptiveinput reects a primary dysregulation of central sensory processing or central sensitization persisting outside theattack (e.g., [104,137] ).

The functional analysis of FHM knockin mouse mod-els supports the view of migraine as a disorder of brainexcitability characterized by decient regulation of thecortical E/I balance, and gives insights into the possibleunderlying molecular and cellular mechanisms and theirrelationship to CSD susceptibility. It has been shownthat the gain-of-function of glutamate release at synap-ses onto cortical pyramidal cells can explain the facilita-tion of experimental CSD in FHM1 knockin mice [26](Figure 4 b). The data are consistent with, and support amodel of, CSD init iat ion in which Ca V 2.1-dependentrelease of glutamate from cortical pyramidal cell synap-ses and activation of NMDA receptors (and possibly postsynaptic Ca V 2.1 channels) play a key role in thepositive feedback cycle that ignites CSD [26,138,139] .The demonstration that FHM1 mutations may different-ly affect synaptic transmission and short-term plasticity at cortical excitatory and inhibitory synapses (noting thelack of effect at FS interneuron synapses) [26] impliesthat, very probably, the FHM1 mutations alter the neu-

ronal circuits that dynamically adjust the E/I balanceduring cortical activity [140,141] (Figure 2 b). It seemsplausible to hypothesize that these alterations may incertain conditions lead to disruption of the E/I balance,overexcitation (due to excessive recurrent excitatory ac-tivity) and neuronal hyperactivity, that may create con-ditions for the initiation of spontaneous CSDs (e.g., by increasing the extracellular [K + ] above a critical value).Similar mechanisms might underlie the susceptibility to CSD in FHM2, given that loss-of-function of the a 2Na + /K + ATPase might impair glutamate clearance andmainly affect excitatory synaptic transmission [39](Figure 2 a).

Thus, ndings from FHM mouse models suggest thatimpairment of the cortical circuits that dynamically adjustthe E/I balance during cortical activity, due to excessiverecurrent glutamatergic neurotransmission, may underlieboth the abnormal regulation of cortical function and thesusceptibility to CSD in FHM (Figure 5). It is certainly possible that FHM mutations produce parallel dysfunc-tions in subcortical areas that might also contribute tothe altered regulation of cortical function and in generalto the disease in a way that remains to be established (e.g.,by altering cortical neuromodulation by monoaminergicprojections and/or by favoring hyperexcitability of centraltrigeminovascular pathways). In this context, CSDmight represent only one manifestation of fundamental

Review Trends in

Neurosciences


515


10/14

alterations

(e.g.,

impairment

of

E/I

balance)

produced

by FHM mutations in different brain areas (Figure 5 ).

Similar mechanisms may underlie the abnormalregulation of cortical (and possibly subcortical) functionin some common migraine subtypes, for which thereis indirect evidence consistent with enhanced corticalglutamatergic neurotransmission [46,136,142] and en-hanced cortico-cortical or recurrent excitatory neuro-transmission [129,130,132,135] . Given the wide clinicaland genetic heterogeneity of migraine, different molecu-lar and cellular mechanisms may well underlie the im-paired regulation of brain function and the susceptibility to CSD in different migraineurs (parallel arrows inFigure 5 ).

Despite recent drug developments, there is a great needfor more efcacious and specic prophylactic migrainemedications [143] . The recent advances in our understand-ing of migraine primary brain dysfunctions support noveltherapeutic strategies that consider cortical E/I dysregula-tion and CSD as key targets of preventive migraine treat-ment. In particular, cortical glutamatergic synapsesappear as key therapeutic targets for novel drugs aimedat counteracting excessive glutamatergic synaptic trans-mission in FHM and some migraine subtypes. Particularly efcacious would be drugs that increase CSD thresholdindependently of the specic cortical dysfunctions under-lying susceptibility to CSD in different migraineurs.

Concluding

remarksTaken together, currently available evidence suggests thatmigraine is a disorder of brain excitability characterized by decient regulation of the E/I balance during corticalactivity. The mechanisms underlying the decient regula-tion of the cortical E/I balance might lead to both (i) thetypical interictal dysfunction in sensory (including trigem-inal nociceptive) information processing, that progressive-ly increases in the period between attacks, and (ii) inparticular conditions, ignition of CSD and activation of the trigeminovascular pain pathway. To verify this hypoth-esis, future studies should investigate the molecular andcellular mechanisms underlying cortical E/I balance dys-regulation and the susceptibility to CSD in migraine, inaddition to how migraine triggers modulate these mecha-nisms. Future research should also elucidate how thecortical and/or subcortical dyfunctions lead to activationand sensitization of the trigeminovascular pain pathway (Box 1 ).

Functional studies in genetic mouse models have begunto unravel the molecular and cellular mechanisms under-lying the dysfunctional regulation of the cortical E/I bal-ance and the susceptibility to CSD in FHM. Thesemechanisms remain largely unknown for the commonforms of migraine, for which the discovery of causativegenes is a key aspect of future research efforts (Box 1 ).Better knowledge of the mechanisms of initiation and

Dysfunctionalsensory

processing

Headache

Aura Activation, sensitizationof the trigeminovascular

pain pathway

Cortical E/I unbalance

Hyperactive cortical circuits

CSD

Migrainetriggers

?

Subcortical areas

?

FHMMigraine

FHM1FHM2

Migraine ?

Glutamatergic synapses

Recurrent excitation

Cortex

?

Cortex

Migraine


Figure5 . Proposed pathophysiological mechanisms in the generation of migraine. It is proposed that migraine is a disorder of brain excitability characterized by deficientregulation of the E/I balance during cortical (and possibly subcortical) activity. In FHM(1,2) and possibly some other migraine subtypes (large blue arrow pathways),excessive recurrent glutamatergic neurotransmission in the cortex (pink boxes) is proposed to lead to alterations in the cortical circuits that dynamically adjust the corticalE/I balance. This results in dysfunctional sensory processing and, under certain conditions (e.g., migraine triggers), in disruption of the E/I balance and neuronalhyperactivity, which creates conditions for ignition of CSD and consequent generation of auraand activation of the trigeminovascular painpathway. Given the wide clinicaland genetic heterogeneity of migraine, different molecular and cellular mechanisms that remain to be elucidated may underlie the impaired regulation of cortical functionandthe susceptibility to CSD in differentmigraine subtypes (asindicated by theparallel thin blue arrow pathways). Theproposed schemealso includesthe possibility that,in both FHMand commonmigraine, dysfunctions in subcortical areas (greenbox) mightcontribute to thealtered regulation of cortical function andto the developmentof migraine headache (e.g., by leading to hyperexcitability of central trigeminovascular pathways).

Review Trends in

Neurosciences


516


11/14

propagation of CSD and of CSD facilitation in mousegenetic models will help the development of novel prophy-lactic migraine medications. The understanding of themolecular and cellular mechanisms underlying corticalE/I balance dysregulation in different migraine forms willbe essential for the development of novel prophylacticmedications tailored to distinct therapeutic targets indifferent patients.

Note added in proof As this review went to press, Freilinger et al. [146] pub-lished the ndings of the rst genome-wide associationstudy of migraine without aura (MO). One of the suscepti-bility loci for MO identied in this study appears particu-larly interesting, since it is within the MEF2D gene thatencodes a transcription factor that mediates neuronalactivity-dependent transcription in neurons and plays akey role in many aspects of synapse and neural circuitdevelopment and function [147] .

AcknowledgmentsWe would like to apologize to the many investigators whose work we wereunable to cite due to space limitations. D.P. is supported by grants from

University of Padova (Strategic Project: Physiopathology of Signaling inNeuronal Tissue) and Fondazione Cariparo (Excellence Project: CalciumSignaling in Health and Disease) and acknowledges the support fromTelethon-Italy (GGP06234).

References1 Lipton , R.B. et al. (2004) Classication of primary headaches.

Neurology 63, 4274352 Leonardi, M. et al. (2005) The global burden of migraine: measuring

di sabili ty in headache disorder s with WHOs Classication of Functioning, Disability and Health (ICF). J. Headache Pain 6,429440

3 Stovner, L.J. and Hagen, K. (2006) Prevalence, burden, and cost of headache disorders. Curr. Opin. Neurol. 19, 281285

4 Olesen, J. et al. (2012)The economic cost of brain disorders in Europe. Eur. J. Neurol. 19, 155162

5 Pietrobon, D. and Striessnig, J. (2003)Neurobiology of migraine. Nat. Rev. Neurosci. 4, 386398

6 Olesen, J. et al. (2009) Origin of pain in migraine: evidence forperipheral sensitisation. Lancet Neurol. 8, 679690

7 Levy, D. (2010) Migraine pain andnociceptor activation where do westand? Headache 50, 909916

8 Ayata, C. (2010) Cortical spreading depression triggers migraineattack: pro. Headache 50, 725730

9 Charles , A. (2010) Does cortica l spreading depression ini tiate amigraine attack? Maybe not. Headache 50, 731733

10 Charles, A. and Brennan, K. (2009) Cortical spreading depression new insights and persistent questions. Cephalalgia 29, 11151124

11 Somjen, G.G. (2001)Mechanisms of spreading depressionand hypoxicspreading depression-like depolarization. Physiol. Rev. 81, 10651096

12 Gifn, N.J. et al. (2003) Premonitory symptoms in migraine: anelectronic diary study. Neurology 60, 935940

13 Hauge, A.W. et al. (2011) Characterization of consistent triggers of migraine with aura. Cephalalgia 31, 416438

14 Coppola, G. et al. (2007) Is the cerebral cortex hyperexcitable orhyperresponsive in migraine? Cephalalgia 27, 14271439

15 Aurora, S.K. and Wilkinson, F. (2007) The brain is hyperexcitable inmigraine. Cephalalgia 27, 14421453

16 de Vries, B. et al. (2009) Molecular genetics of migraine. Hum. Genet.126, 115132

17 Russell, M.B. and Ducros, A. (2011) Sporadic and familial hemiplegicmigraine: pathophysiological mechanisms, clinical characteristics,diagnosis, and management. Lancet Neurol. 10, 457470

18 Ophoff, R.A. et al. (1996) Familial hemiplegic migraine and episodicataxia type-2 are caused by mutations in the Ca 2+ channel geneCACNL1A4 . Cell 87, 543552

19 De Fusco, M. et al. (2003) Haploinsufciency of ATP1A2 encoding theNa + /K + pump a 2 subunit associated with famil ia l hemiplegicmigraine type 2. Nat. Genet. 33, 192196

20 Dichgans, M. et al. (2005) Mutation in the neuronal voltage-gatedsodium channel SCN1A in familial hemiplegic migraine. Lancet 366,371377

21 Pietrobon, D. (2007) Familial hemiplegicmigraine. Neurotherapeutics4, 274284

22 Thomsen, L.L. et al. (2007) The genetic spectrum of a population-based sample of familial hemiplegic migraine. Brain 130, 346356

23 Pietrobon, D. (2010) Ca V 2.1 channelopathies. Pugers Arch. 460, 375393

24 Pietrobon, D. (2005) Function and dysfunction of synaptic calciumchannels: insights from mouse models. Curr. Opin. Neurobiol. 15,257265

25 Tottene, A. et al. (2002) Familial hemiplegic migraine mutationsincrease Ca 2+ inux through single human Ca V 2.1 channels anddecrease maximal Ca V 2.1 current density in neurons. Proc. Natl. Acad. Sci. U.S.A. 99, 1328413289

26 Tottene, A. et al. (2009) Enhanced excitatory transmission at corticalsynapses as the basis for facilitated spreading depression in Ca V 2.1knockin migraine mice. Neuron 61, 762773

27 vanden Maagdenberg,A.M. et al. (2004) A Cacna1a knockin migrainemouse model with increased susceptibility to cortical spreading depression. Neuron 41, 701710

28 van den Maagdenberg, A.M. et al. (2010) High cortical spreading depression susceptibility and migraine-associated symptoms inCa V 2.1 S218L mice. Ann. Neurol. 67, 8598

Box 1. Outstanding questions

Genetics: which genes are involved in common migraine(s) andhow do they cooperate in causing the disease? What are theidentities of the FHM genes that remain to be identified?

Primary brain dysfunctions: although it is clear that most migraineattacks start in the brain, a major general outstanding questionconcerns the nature and mechanisms of the primary brain dysfunc-

tions that cause episodic activation of the trigeminovascular painpathway. To understand the primary brain mechanisms of migraine itseems essential that future studies address the following specificquestions:

( i) How are the cortical circuits that dynamically adjust the E/Ibalance specifically altered in FHM mouse models, and in whichconditions may these alterations lead to disruption of the E/Ibalance in a way that allows CSD ignition?

(ii) What are the molecular and cellular mechanisms underlying thedysfunctional regulation of the cortical E/I balance and theabnormal processing of sensory information, and its periodi-city, in migraineurs? How are they affected by migraine triggersand by repeated attacks?

( iii) Do silent CSDs occur in MO and what are the underlyingmolecular and cellular mechanisms?

Mechanisms of activation and sensitization of the trigeminovascularpain pathway: it is generally recognized that the throbbing migraineheadache is due to long-lasting sensitization of meningeal trigemi-novascular nociceptors (peripheral sensitization) together with, inmost patients, central sensitization of the trigeminovascular painpathway. A major general outstanding question is how the migraineprimary brain dysfunctions lead to activation and sensitization of thetrigeminovascular pathway. Specic questions to address are:(i) What are the mechanisms of the sustained, and in most cases,

delayed activation of dural trigeminal afferents induced by CSDin animal studies? Is NI involved? More generally, is NI theendogenous process underlying peripheral sensitization of meningeal nociceptors? Is a neuronglia inflammatory cycle atthe TG level involved? If other mechanisms are involved, what istheir nature?

(ii) Are subcortical and/or cortical structures that are involved inthe central control of pain dysfunctional in migraine? Whatare the molecular and cellular mechanisms of their specificdysfunctions?

Review Trends in

Neurosciences


517


12/14

29 Inchauspe,C.G. et al. (2010)Gain of function in FHM-1Ca V 2.1 knock-in mice is related to the shape of the action potential. J. Neurophysiol.104, 291299

30 Adams, P.J. et al. (2010) Contribution of calcium-dependentfacilitation to synaptic plasticity revealed by migraine mutations inthe P/Q-type calcium channel. Proc. Natl. Acad. Sci. U.S.A. 107,1869418699

31 Fioretti, B. et al. (2011) Trigeminal ganglion neuron subtype-specicalterations of Ca V 2.1 calcium current and excitability in a Cacna1a

mouse model of migraine. J. Physiol. 589, 5879589532 Mullner, C. et al. (2004) Famil ia l hemiplegic migraine type 1

mutations K1336E, W1684R, and V1696I alter Ca V 2.1 Ca 2+

channel gating: evidence for b -subunit isoform-specic effects. J. Biol. Chem. 279, 5184451850

33 Adams, P.J. et al. (2009) Ca V 2.1 P/Q-type calcium channel alternativesplicing affects the functional impact of familial hemiplegic migrainemutations: implications for calcium channelopathies. Channels(Austin) 3, 110121

34 Cholet, N. et al. (2002) Similar perisynaptic glial localization for theNa + ,K + -ATPase a 2 subunit and the glutamate transporters GLASTandGLT-1in the rat somatosensory cortex. Cereb. Cortex 12, 515525

35 Pellerin, L. and Magistretti, P.J. (1997)Glutamate uptake stimulatesNa + ,K + -ATPase activi ty in astrocytes via act ivat ion of a dis tinctsubunit highly sensitive to ouabain. J. Neurochem. 69, 21322137

36 Rose, E.M. et al. (2009) Glutamate transporter coupling to Na,K-

ATPase. J. Neurosci. 29, 8143815537 Tavraz, N.N. et al. (2008) Diverse functional consequences of mutations in the Na + /K + -ATPase a 2 -subunit causing familialhemiplegic migraine type 2. J. Biol. Chem. 283, 3109731106

38 Tavraz, N.N. et al. (2009) Impaired plasma membrane targeting orprotein stability by certain ATP1A2 mutations identied in sporadicor familial hemiplegic migraine. Channels (Austin) 3, 8287

39 Leo, L. et al. (2011) Increased susceptibility to cortical spreading depression in the mouse model of famil ia l hemiplegic migrainetype 2. PLoS Genet. 7, e1002129

40 Catterall, W.A. et al. (2010) Na V 1.1 channels and epilepsy. J. Physiol.588, 18491859

41 Cestele,S. et al. (2008) Self-limited hyperexcitability:functional effectof a familial hemiplegic migraine mutation of the Na V 1.1 ( SCN1A )Na + channel. J. Neurosci. 28, 72737283

42 Kahlig,K.M. et al. (2008) Divergent sodiumchannel defects in familialhemiplegic migraine. Proc. Nat. Acad. Sci. U.S.A. 105, 97999804

43 Suzuki , M. et al. (2010) Defective membrane expression of the Na + -HCO 3 cotransporter NBCe1 is associated with familial migraine. Proc. Nat. Acad. Sci. U.S.A. 107, 1596315968

44 Lafreniere, R.G. et al. (2010) A dominant-negative mutation in theTRESK potassium channel is linked to familial migraine with aura. Nat. Med. 16, 11571160

45 Andres-Enguix, I. et al. (2012) Functional analysis of missense variants in the TRESK (KCNK18) K + channel. Sci. Rep. 2, 237

46 Anttila, V. et al. (2010) Genome-wide association study of migraineimplicates a common susceptibility variant on 8q22.1. Nat. Genet. 42,869873

47 Tzingounis, A.V. and Wadiche, J.I. (2007) Glutamate transporters:conning runaway excitation by shaping synaptic transmission. Nat. Rev. Neurosci. 8, 935947

48 Chasman, D.I. et al. (2011) Genome-wide association study revealsthree susceptibility loci for common migraine in the generalpopulation. Nat. Genet. 43, 695698

49 Madrid, R. et al. (2006) Contribution of TRPM8 channels to coldtransduct ion in primary sensory neurons and peripheral nerveterminals. J. Neurosci. 26, 1251212525

50 Strassman, A.M. and Levy, D. (2006) Response properties of duralnociceptors in relation to headache. J. Neurophysiol. 95, 12981306

51 Schytz, H.W. et al. (2010) What have we learnt from trigger ing migraine? Curr. Opin. Neurol. 23, 259265

52 Brennan, K.C.and Charles, A.(2010) An update onthe blood vesselinmigraine. Curr. Opin. Neurol. 23, 266274

53 Schoonman, G.G. et al. (2008) Migraine headache is not associatedwith cerebral or meningeal vasodilatation a 3T magnetic resonanceangiography study. Brain 131, 21922200

54 Asghar, M.S. et al. (2011) Evidence for a vascular factor in migraine. Ann. Neurol. 69, 635645

55 Waeber, C. and Moskowitz, M.A. (2005) Migraine as an inammatory disorder. Neurology 64, S9S15

56 Levy, D. (2009) Migraine pain, meningeal inammation, and mastcells. Curr. Pain Headache Rep. 13, 237240

57 Vaughn, A.H. and Gold, M.S. (2010) Ionic mechanisms underlying inammatory mediator-induced sensitization of dural afferents. J. Neurosci. 30, 78787888

58 Yan, J. et al. (2011) Dural afferents express acid-sensing ion channels:a role for decreased meningeal pH in migraine headache. Pain 152,

10611359 Strassman, A.M. et al. (1996) Sensitization of meningeal sensory

neurons and the origin of headaches. Nature 384, 56056460 Edelmayer, R.M. et al. (2009) Medullary pain facilitating neurons

mediate allodynia in headache-related pain. Ann. Neurol. 65, 18419361 Burstein, R. et al. (2000) The development of cutaneous allodynia

during a migraine at tack cl inical evidence for the sequentialrecruitmen t of spinal and supraspinal nociceptive neu rons inmigraine. Brain 123, 17031709

62 Levy,D. et al. (2007) Mast cell degranulation activates a pain pathway underlying migraine headache. Pain 130, 166176

63 Levy, D. (2012) Endogenous mechanisms underlying the activationand sensi tiza tion of meningeal nociceptors: the role of immuno- vascular interactions and cortical spreading depression. Curr. Pain Headache Rep. 16, 270277

64 Villalon, C.M. and Olesen, J. (2009) The role of CGRP in the

pathophysiology of migraine and efcacy

of CGRP receptorantagonistsasacuteantimigrainedrugs. Pharmacol.Ther. 124,30932365 Recober, A. andRusso, A.F. (2009)Calcitonin gene-relatedpeptide:an

update on the biology. Curr. Opin. Neurol. 22, 24124666 Ho, T.W. et al. (2010) CGRP and its receptors provide new insights

into migraine pathophysiology. Nat. Rev. Neurol. 6, 57358267 Nicoletti, P. et al. (2008) Ethanol causes neurogenic vasodilation by

TRPV1 activationand CGRP release in the trigeminovascular systemof the guinea pig. Cephalalgia 28, 917

68 Nassini,R. et al. (2012)The headache treevia umbellulone andTRPA1activates the trigeminovascular system. Brain 135, 376390

69 Baun, M. et al. (2012) Dural mast cell degranulation is a putativemechanism for headache induced by PACAP-38. Cephalalgia 32,337345

70 Burstein, R. et al. (2000) An association between migraine andcutaneous allodynia. Ann. Neurol. 47, 614624

71 Lipton, R.B. et al. (2008) Cutaneous allodynia in the migrainepopulation. Ann. Neurol. 63, 148158

72 Burstein, R. et al. (1998) Chemical stimulation of the intracranialdura induces enhanced responses to facial stimulation in brain stemtrigeminal neurons. J. Neurophysiol. 79, 964982

73 Burstein, R. et al. (2010) Thalamic sensitization transforms localizedpain into widespread allodynia. Ann. Neurol. 68, 8191

74 Lee,M.C. et al. (2008) Identifying brain activity specically related tothe maintenance and perceptual consequence of central sensitizationin humans. J. Neurosci. 28, 1164211649

75 Moulton, E.A. et al. (2008) Interictal dysfunction of a brainstemdescending modulatory center in migraine patients. PLoS ONE 3,e3799

76 Weiller, C. et al. (1995) Brain stem activation in spontaneous humanmigraine attacks. Nat. Med. 1, 658660

77 Olesen , J. et al. (2004) Calcitonin gene-related peptide receptorantagonist BIBN 4096 BS for the acute treatment of migraine. N. Engl. J. Med. 350, 11041110

78 Ho,T.W. et al. (2008)Efcacy and tolerability of MK-0974(telcagepant),a new oral antagonis t of calcitonin gene-related peptide receptor,compared with zolmitriptan for acute migraine: a randomised,placebo-controlled, parallel-treatment trial. Lancet 372, 21152123

79 Lassen,L.H. et al. (2002) CGRP mayplay a causativerole in migraine.Cephalalgia 22, 5461

80 Lennerz, J.K. et al. (2008) Calcitonin receptor-like receptor (CLR),receptor activity-modifying protein 1 (RAMP1), and calcitonin gene-related peptide (CGRP) immunoreactivity in the rat trigeminovascularsystem: differences between peripheral and central CGRP receptordistribution. J. Comp. Neurol. 507, 12771299

81 Eftekhari, S. et al. (2010) Differential distribution of calcitonin gene-related peptide and its receptor components in the human trigeminalganglion. Neuroscience 169, 683696

Review Trends in

Neurosciences


518


13/14

82 Levy,D. et al. (2005) Calcitonin gene-related peptide doesnot exciteorsensitize meningeal nociceptors: implications for the pathophysiology of migraine. Ann. Neurol. 58, 698705

83 Zhang, X.C. et al. (2007) Sensitization and activation of intracranialmeningeal nociceptors by mast cell mediators. J. Pharmacol. Exp.Ther. 322, 806812

84 Dux, M. et al. (2009) Involvement of capsaicin-sensitive afferentnervesin the proteinase-activated receptor 2-mediated vasodilatation in therat dura mater. Neuroscience 161, 887894

85 Fabbretti, E. et al. (2006) Delayed upregulation of ATP P2X 3 receptorsof trigeminal sensory neurons by calcitonin gene-related peptide. J. Neurosci. 26, 61636171

86 Zhang,Z. et al. (2007) Sensitization of calcitonin gene-related peptidereceptors by receptor activity-modifying protein-1 in the trigeminalganglion. J. Neurosci. 27, 26932703

87 Li , J. et al. (2008) Calcitonin gene-related peptide stimulation of nitricoxide synthesis andrelease from trigeminal ganglion glial cells. Brain Res. 1196, 2232

88 Vau se, C.V. an d Durham, P.L. (2010) Calcitonin gene-relatedpeptide differentially regulates gene and prote in expression intrigeminal glia cells: ndings from array analysis. Neurosci. Lett.473, 163167

89 Capuano, A. et al. (2009) Proinammatory-activated trigeminalsat elli te cells promote neuronal sensi tization: relevance formigraine pathology. Mol. Pain 5, 43

90 Ceruti, S. et al. (2011) Calcitonin gene-related peptide-mediatedenhancement of purinergic neuron/glia communication by thealgogenic factor bradykinin in mouse trigeminal ganglia from wild-type and R192Q Ca V 2.1 knock-in mice: implications for basicmechanisms of migraine pain. J. Neurosci. 31, 36383649

91 Nai r, A. et al. (2010) Familial hemiplegic migraine Ca V 2.1 channelmutation R192Q enhances ATP-gated P2X 3 receptor activity of mouse sensory ganglion neurons mediating trigeminal pain. Mol. Pain 6, 48

92 Summ, O. et al. (2010) Modulation of nocioceptive transmission withcalcitonin gene-related peptide receptor antagonists in the thalamus. Brain 133, 25402548

93 Storer, R.J. et al. (2004) Calcitonin gene-related peptide (CGRP)modulates nociceptive trigeminovascular transmission in the cat. Br. J. Pharmacol. 142, 11711181

94 Meng, J. et al. (2009) Activation of TRPV1 mediates calcitonin gene-related peptide release, which excites trigeminal sensory neurons andis attenuated by a retargeted botulinum toxin with anti-nociceptivepotential. J. Neurosci. 29, 49814992

95 Fischer, M.J. et al. (2005) The nonpeptide calcitonin gene-relatedpeptide receptor antagonist BIBN4096BS lowers the activity of neurons with meningeal input in the rat spinal trigeminal nucleus. J. Neurosci. 25, 58775883

96 Six t, M.L. et al. (2009) Calcitonin gene-related peptide receptorantagonist olcegepant acts in the spinal trigeminal nucleus. Brain132, 31343141

97 Edvinsson, L. et al. (2007) Inhibitory effect of BIBN4096BS, CGRP(8-37), a CGRP antibody and an RNA-Spiegelmer on CGRP induced vasodilatation in the perfused and non-perfused rat middle cerebralartery. Br. J. Pharmacol. 150, 633640

98 Asghar, M.S. et al. (2012) Effect of CGRP and sumatriptan on theBOLD response in visual cortex. J. Headache Pain 13, 159166

99 Akerman, S. et al. (2011) Diencephalic and brainstem mechanisms inmigraine. Nat. Rev. Neurosci. 12, 570584

100 Lambert, G.A. andZagami, A.S. (2009)The mode of actionof migrainetriggers: a hypothesis. Headache 49, 253275

101 Denuelle, M. et al. (2007) Hypothalamic activation in spontaneousmigraine attacks. Headache 47, 14181426

102 Mainero,C. et al. (2007) Mapping the spinal and supraspinal pathwaysof dynamicmechanical allodyniain thehuman trigeminal system using cardiac-gated fMRI. Neuroimage 35, 12011210

103 Stankewitz, A. et al. (2011) Trigeminal nociceptive transmissionin migraineurs predicts migraine attacks. J. Neurosci. 31, 19371943

104 Mainero, C. et al. (2011) Altered functional magnetic resonanceimaging resting-state connectivity in periaqueductal gray networksin migraine. Ann. Neurol. 70, 838845

105 Noseda, R. et al. (2010) A neural mechanism for exacerbation of headache by light. Nat. Neurosci. 13, 239245

106 Zhang, X. et al. (2010) Activation of meningeal nociceptors by corticalspreading depression: implicat ions for migraine with aura. J. Neurosci. 30, 88078814

107 Zhang, X. et al. (2011) Activation of central trigeminovascularneurons by cortical spreading depression. Ann. Neurol. 69, 855865

108 Busija, D.W. et al. (2008) Mechanisms involved in the cerebrovasculardilator effects of cortical spreading depression. Prog. Neurobiol. 86,379395

109 Bolay, H. et al. (2002) Intrinsic brain activity triggers trigeminal

meningeal afferents in a migraine model. Nat. Med. 8, 136142110 Ebersberger, A. et al. (2001) Is there a correlation between spreading

depression, neurogenic inammation, and nociception that mightcause migraine headache? Ann. Neurol. 49, 713

111 Noseda, R. et al. (2010) Changes of meningeal excitability mediated by corticotrigeminal networks: a link for the endogenous modulation of migraine pain. J. Neurosci. 30, 1442014429

112 Koroleva, V.I. and Bures, J. (1993) Rats do not experience cortical orhippocampal spreading depression as aversive. Neurosci. Lett. 149,153156

113 Akcali, D. et al. (2010) Does single cortical spreading depression elicitpain behaviour in freely moving rats? Cephalalgia 30, 11951206

114 Fioravanti, B. et al. (2010) Evaluation of cutaneous allodyniafollowing induction of cort ical spreading depres sion in freely moving rats. Cephalalgia 31, 10901100

115 Levy, D. et al. (2012) Activation of the migraine pain pathway by

cortical spreading depression: Do we need more evidence?Cephalalgia 32, 581582116 Ayata,C. et al. (2006) Suppression of cortical spreading depression in

migraine prophylaxis. Ann. Neurol. 59, 652661117 Hoffmann, U. et al. (2011) Oxcarbazepine does not suppress cortical

spreading depression. Cephalalgia 31, 537542118 Hauge, A.W. et al. (2009)Effects of tonabersat on migraine with aura:

a randomised, double-blind, placebo-controlled crossover study. Lancet Neurol. 8, 718723

119 Bogdanov, V.B. et al. (2011) Migraine preventive drugs differentially affect cortical spreading depression in rat. Neurobiol. Dis. 41, 430435

120 Tottene, A. et al. (2005) Specic kinetic alterations of human Ca V 2.1calcium channels produced by mutation S218L causing familialhemiplegic migraine and delayed cerebral edema and coma afterminor head trauma. J. Biol. Chem. 280, 1767817686

121 Eikermann-Haerter, K. et al. (2009) Genetic and hormonal factorsmodulate spreading depression and transient hemiparesis in mousemodels of familial hemiplegic migraine type 1. J. Clin. Invest. 119,99109

122 Eikermann-Haerter, K. et al. (2011) Enhanced subcortical spreading depression in familial hemiplegic migraine type 1 mutant mice. J. Neurosci. 31, 57555763

123 Eikermann-Haerter, K. et al. (2009) Androgenic suppression of spreading depression in familial hemipleg ic migraine type 1mutant mice. Ann. Neurol. 66, 564568

124 Vahedi, K. et al. (2009) Elicited repetitive daily blindness: a newphenotype associated with hemiplegic migraine and SCN1Amutations. Neurology 72, 11781183

125 Eikermann-Haerter, K. et al. (2011) Cerebral autosomal dominantarteriopathy with subcortical infarcts and leukoencephalopathy syndrome mutations increase susceptibility to spreading depression. Ann. Neurol. 69, 413418

126 Denuel le , M. et al. (2008) Posterior cerebral hypoperfusion inmigraine without aura. Cephalalgia 28, 856862

127 Siniatchkin, M. et al. (2006) Neurophysiological reactivity before amigraine attack. Neurosci. Lett. 400, 121124

Documents

Excitatory Inhibitory Balance