OBSERVATORIO DE MOVILIDAD VIAL

The original studies that pointed toward a role for the PAG in migraine were actually clinical case reports of the effects of stimulation-produced analgesia in the PAG. There is evidence suggesting that electrical or pharmacological activation of PAG neurons can inhibit pain in humans (Hosobuchi et al., 1977; Richardson and Akil, 1977; Gybels and Kupers, 1990) and nonhuman primates (Pert and Yaksh, 1974; Gerhart et al., 1984; Lin et al., 1994). In humans, stimulation produced analgesia has been controversial. Some studies have reported profound analgesic effects without any unpleasant side effects by stimulation of the rostral PAG (Hosobuchi et al., 1977; Baskin et al., 1986). In these studies, stimulation of the PAG produced analgesia that was associated with a feeling of warmth and a sense of well being. On the other hand, stimulation of the PAG has also been shown to produce analgesia associated with an unpleasant feeling that was not well tolerated. In both locations, the effect of stimulation outlasted the duration of the stimulation. This is an interesting correlation with the similar finding in the Weiller MRI study, in which brainstem activation outlasted the headache and its relief. In the stimulation-produced analgesia studies, patients received pain relief by stimulating the PAG four times a day for 20 minutes per stimulation time (Hosobuchi, 1983). Other studies however have not confirmed these results and have reported poor analgesic effects (Young et al., 1985). The difference in the results of these studies is most likely due to placement of the stimulating electrode. Also, differences in patient populations may have contributed to the disparity of the various study results.

Although numerous studies appear to indicate a role for the PAG specifically in migraine, it is important to note than non-head pain is also associated with the PAG.

1.8.3. Current theories o f pathophysioiogy

Included in the IHS classification criteria of migraine without aura is the mention that the ‘mechanisms of the attack are as yet poorly understood. Regional cerebral blood flow remains

normal or is perhaps slightly increased during an attack. Changes in blood composition and platelet function initiated endogenously or by environmental influences may play a triggering role. The pathophysiological process of the attack is presumed to occur in the brain, which via trigemino-vascular and other systems interacts with intra- and extracranial vasculature and perivascular spaces’.

This complex, but still nebulus description of the pathophysiological processes of migraine indicates how little was, and in some ways still is, known of the disease. Nevertheless, it opens a window onto the myriad of experimental studies that are performed into the various themes of trigeminal nociception and migraine, and that constitute the present day theories of pathophysiology. Current opinion is that migraine is a centrally driven neurovascular disorder. Although the central processes involved in the initiation of a migraine attack are poorly understood, much is known about the factors involved in the processes of head pain. Details of these ideas as pathophysiological mechanisms are described below, with emphasis on theories pertaining to this thesis.

1.8.3.1. Molecular gen etics

The tendency of migraine to have familial aggregation, its early onset, and twin studies has provided data on genetic susceptibility as the cause for the disease. Concordance rates in monozygotic twins suggest that up to 50% of the contribution to migraine is genetically based and indicates a multifactorial, genetically complex aetiology (Gardner, 1999). The advent of molecular techniques in the 1990’s brought concrete evidence of the genetic component of migraine. The present molecular focus in migraine is towards the calcium channelopathy. Aside from calcium channel mutations, genetic predispositions to migraine have also been linked to

the following: dopamine receptors (Peroutka et al., 1997), Notch 3 (Davous, 1998),

mitochondrial genes (Montagna et al., 1988; Klopstock et al., 1996), endothelin type A receptor gene (Tzourio et al., 2(X)1), Xq24-28 linked gene (Nyholt et al., 2(X)1), insulin receptor gene (McCarthy et al., 2001) and prothrombotic factors (Iniesta et al., 2001).

1.8.3.1.1. Calcium channelopathy

As summarised in the section on voltage gated calcium channels, theCa^^-selective pore of P/Q- type Ca^^ channels is formed by the ttiA subunit, which also contain the voltage sensor. The

human gene CACNAIA on chromosome 19pl3 encodes ttiA- CACNAIA mutations have been

described which are responsible for at least three different neurological human diseases: episodic ataxia type 2 (EA2), spinocerebellar ataxia type 6, and familial hemiplegic migraine (FHM) with and without cerebellar ataxia. FHM is a rare autosomal dominant form of migraine with prolonged aura, associated with ictal hemi paresis and, in some families, with cerebellar ataxia and atrophy. FHM shows genetic heterogeneity involving at least three different loci on chromosome 19p (Toumier-Lasserve et al., 1993), chromosome lq21-23 (Ducros et al., 1997), and chromosome lq31 (Gardner et al., 1997). Some regular migraine families appear to be

associated with ch rom osom e 1 9 p l3 (M ay et al., 1995; N y h o lt et al., 1998) and a new locu s on X q 28 has been proposed for regular m igraine with X dom inant inheritance (N y h o lt et al., 200 1 ). FH M is thought to be a gen etically sim p le (autosom al dom inant) disorder, w hereas the variants o f m ore com m on m igraine are lik ely to be gen etically co m p lex (m ultifactorial p olygen etic).

A t least thirteen m issen se m utations have been identified in fa m ilies with FH M (O p h o ff et al.,

1996; D ucros et al., 2 0 0 1 ). T he location s on the Œia subunit o f th ese thirteen m utations is pictured in Figure 16; they are D 7 1 5 E , 1 1 8 1 IL, K 1 3 3 6 E , R 5 83Q , R 1 9 2 Q , R 1 9 5 K , R 1668W , T 6 6 6 M , V 1 4 5 7 L , V 7 1 4 A , V 1 6 9 6 1 , W 1 684R , Y 1 3 8 5 C .

Figure 16. M utations in C A C NA I A linked w ith hem iplegic m igraine.

T he v o ltag e g a te d calciu m c h an n el is c o m p rise d o f 5 su b u n its: a , , «2, P, y, Ô. T h e « i a su b u n it c o n tain s fo u r re p ea te d d o m ain s, each in clu d in g six se g m en ts (S 1 -S 6 ). T he S 4 se g m en ts form the v o ltag e sensor, the S5 and S 6 form the in n er p ore and the P loops line the pore. T he th irte en id en tifie d m u tatio n s linked w ith h e m ip leg ic m ig rain e are in d icated by a red sq u a re and lab elled .

a l E x t r a c e llu la r I n t r a c e l lu la r R S 83Q T 6 6 6 M K 1336E Y 1 3 8 5 C V 1 4S 7L R 1 6 6 8 W R195K R 1 9 2 Q E x tra c e llu la r

V V

Two independent mechanisms altered by these mutations can affect P/Q-type currents in FHM patients: altered expression density and changes in channel gating. Functional expression of rabbit «ia subunits containing the FHM mutations D715E, T666M, V714A, 1181IL, R583Q,

D715E, and V1457L in Xenopus laevis oocytes revealed mutation-induced changes in gating

kinetics altering the extent to which P/Q-type channels accumulate in inactivation during trains

of depolarising pulses (Kraus et al., 1998; Kraus et al., 2(XX)). This could alter influx and

signalling during episodes of high neuronal activity. This in turn might result in a long-term activation of neurons within the structure or system of neurons where the mutations manifest their phenotypic dysfunction.

Essentially the same changes in gating kinetics were reported by Hans et a l after introduction of

the T666M, V714A, 1181IL, and R192Q mutations in human aiA followed by heterologous expression in human embryonic kidney 293 cells and patch clamp analysis (Hans et al., 1999). In addition, they reported mutation-induced changes in single channel kinetics and expression

density. It is difficult to predict if these changes of expression density also occur in vivo where,

in addition to the accessory (Xg-Ô and P subunits, ttiA interacts with a number of other modulatory subunits such as receptor mediated G proteins, protein kinases, presynaptic proteins and synaptic vesicle proteins. Irrespective of changes in expression density, changes in gating properties represent an elementary functional alteration underlying P/Q-type Ca^^ channel dysfunction in FHM.

In the context of migraine pathophysiology, the neurons in which these varying effects might occur remains an open question. To date, research on dysfunctional P/Q-type calcium channels and their clinical manifestations has focused on non-headache features. In migraine and familial hemiplegic migraine a link between the mutations and features such as cerebellar impairment and impairment of neuromuscular transmission was shown (Ambrosini et al., 2001; Sandor et al., 2001). In mice, mutations in the gene are found in a variety of mutants named for the specific neurological characteristics and motor behaviour produced by the mutation, such as

tottering, leaner, rocker and rolling mouse (Zwingman et al., 2001). In these, as well as the null mutant mouse, a variety of effects have been demonstrated, primarily in cerebellar and neuromuscular factors, but not in head pain. However, a raised threshold for activation of

cortical spreading depression has been demonstrated in the tottering and leaner mouse (Ayata et

al., 2000).

Whilst an extrapolation between the mouse data and the human phenotype must be viewed with caution, altered levels of neurotransmitter release may contribute to the pathophysiological

consequences of CACNAIA mutations. The phenotypes may also represent compensatory

mechanisms in response to the physiological consequences of the CACNAIA mutations. This is

Purkinje and cerebellar granule cells (amongst other possible and unidentified structural deficits), is bom alive, albeit with rapidly progressing neurological deficits characterised by dystonia, ataxia and cerebellar involvement (Jun et al., 1999). Although these mice die within 3- 4 weeks, it is surprising that they don’t exhibit other, more gross and debilitating deficits, since the ttiA subunit is present on all neurons throughout the body and is responsible for P/Q-type currents in such a wide range of neurons (Hoffman, 2001).

Other channelopathies, such as episodic ataxia and hypokalemic periodic paralysis, exhibit the same characteristic episodicity as migraine and FHM. That is, in these disorders patients are normal between attacks then have marked clinical problems during attacks. The episodic feature of migraine may, in part, indicate a dysfunctional gating mechanism contributed by calcium channel dysfunction providing a continuous background of neuronal instability. However, the findings of the patch clamp studies, that the functional effect of the mutations on single channels was not present in some patches and in some periods of activity, suggests the possibility that some unknown factor can precipitate an attack by directly switching the abnormal channel on or off.

It has been hypothesised that genetic abnormalities (a calcium channel functional defect for example) result in a lowered threshold of response to specific triggers in migraineurs (e.g. certain foods, hormones, changes in levels of stress). This hyper-responsivity of migraineur’s brains may be a consequence of ion channel dysfunction, dysregulating neuronal excitability (Ferrari, 1998).