Deciphering migraine mechanisms Clues from familial hemiplegic migraine genotypes.код для вставкиСкачать
POINT OF VIEW Deciphering Migraine Mechanisms: Clues from Familial Hemiplegic Migraine Genotypes Michael A. Moskowitz, MD,1 Hayrunnisa Bolay, MD, PhD,2 and Turgay Dalkara, MD, PhD3 The molecular and cellular origins of migraine headache are among the most enigmatic in clinical neuroscience. Most agree that susceptibility is inherited and that its clinical presentation is strongly modulated by both internal and external factors. Polymorphisms in genes regulating ion translocation have been implicated in two subtypes of familial hemiplegic migraine (FHM), a rare migraine disorder. Families with FHM type 1 express point mutations in the Cav2.1 channel,1 whereas type 2 patients express mutations in the ␣ subunit of the Na⫹,K⫹ pump2 (Fig 1). Cav2.1 channels gate Ca⫹⫹, whereas the ATP-utilizing pumps distribute Na⫹,K⫹ ions across plasma membranes. Interestingly, the mutated ␣1A subunit of the P/Q calcium channel is found exclusively on neurons, whereas the ␣2 subunit of the pump is expressed primarily by astrocytes in adult brain.3,4 How then does a coherent migraine phenotype emerge as a consequence of point mutations expressed on distinctive cell types regulating monovalent or divalent cation fluxes? The simple answer is that we do not know. However, human studies strongly implicating cortical spreading depression (CSD) as the generator of migraine aura, together with evidence linking astrocytes and blood vessels to brain metabolism and synaptic activity, provide intriguing possibilities relevant to FHM, and perhaps by extrapolation, to more common forms of migraine headache. Mutations Point mutations in Cav2.1 were first identified in 19961 to provide the first evidence implicating FHM and perhaps more typical migraine as a channelopathy.5,6 Cav2.1 channels are located on presynaptic terminals as well as on somatodendritic membranes and regulate neurotransmitter release in addition to postsynaptic calcium fluxes and excitability. Synaptic release of glutamate from cortical neurons depends primarily on the opening of P/Q-type calcium channels, whereas P/Q-type channels are less significant regulators of GABA release. Accordingly, in leaner mice expressing a spontaneous Cav2.1 mutation, a decrease in depolarization-induced release of glutamate is measurable in microdialysate with almost no change in GABA.7 Interestingly, this mouse exhibits an elevated threshold for initiating CSD and a slower velocity of CSD propagation. These observations clearly show that P/Q channel mutations modify cortical excitability and generate a CSD phenotype. The functional consequences of FHM1 mutations to the membrane properties of Cav2.1 are more complex than initially expected. Despite this complexity, all FHM1 mutations analyzed so far display enhanced Ca2⫹ influx through single Cav2.1 channels for voltages lower than ⫺10mV (single-channel gain-offunction phenotype).8 The lower activation threshold and increased single-channel opening probability may augment Ca2⫹⫹ influx into nerve terminals and enhance glutamate release, which could explain lower CSD threshold and prolonged neurological dysfunction in FHM1 patients (see below). The impact of P/Q mutations on other aspects of channel kinetics (studied as above by patch-clamp techniques) are less consistent and difficult to predict because the net outcome on cortical excitability may vary depending on the dominance of presynaptic or postsynaptic actions and on particular neuronal populations (eg, glutamatergic, GABAergic, serotoninergic). These findings and their implications have been reviewed.9 FHM2 mutations were first reported in two Italian families early in 20032 and more recently in two Dutch families.10 ATP1A2v the mutated gene, encodes the catalytic ␣2 subunit of Na⫹,K⫹-ATPase, and this subunit binds sodium, potassium, and ATP and utilizes ATP hydrolysis to extrude three Na⫹ ions. Na⫹ pumping provides the steep Na⫹ gradient essential for the transport of amino acids (eg, glutamate) and cal- From the 1Neuroscience Center, Department of Radiology and Neurology, Massachusetts General Hospital, Harvard Medical School, Charlestown, MA; 2Gazi University Hospitals, Department of Neurology; and 3Hacettepe University, Faculty of Medicine, Department of Neurology, Ankara, Turkey. Received Sep 5, 2003, and in revised form Oct 31. Accepted for publication Nov 3, 2003. Address correspondence to Dr Moskowitz, Neuroscience Center, Department of Radiology and Neurology, Massachusetts General Hospital, Harvard Medical School, 149 13th Street, Room 6403, Charlestown, MA 02129. firstname.lastname@example.org 276 © 2004 American Neurological Association Published by Wiley-Liss, Inc., through Wiley Subscription Services Fig 1. Topology of the (A) ␣1A subunit of the P/Q calcium channel and (B) ␣2 subunit of Na⫹,K⫹-ATPase implicated in FHM1 and FHM2, respectively. The location of amino acid substitutions are shown. Figures are reprinted from Ducros11 (A) and DeFusco and colleagues2 (B). Moskowitz et al: Deciphering Migraine Mechanisms 277 cium. In neonates, the ␣2 protein is expressed predominantly in neurons and interestingly ATP1A2 mutations are associated with infantile convulsions. Presumably, seizures are caused by the predominant expression of this gene in neurons at early stages of life. In the adult, the ␣2 protein is constitutively expressed predominantly by astrocytes and by pia/arachnoidal cells,4 explaining perhaps the differential clinical expression during maturation: neonatal convulsions and migraine at later ages. In the mutated protein, single substitutions were found in highly conserved amino acids, and when expressed by HeLa cells pump activity was reportedly inhibited, suggesting a loss of function mutation. Clinical Features FHM appears most relevant to migraine with aura (MWA) and lies on one end of a migraine clinical spectrum. FHM1 and FHM2 are characterized by enhanced susceptibility and sustained attacks of visual, somatosensory, and aphasic auras as well as by prolonged motor weakness or paralysis. The two subtypes cannot be easily distinguished phenotypically, except that cerebellar signs and/or atrophy are characteristic of a particular type 1 point mutation.11 Motor weakness and prolonged aura characterize those features typical of types 1 and 2. Photophobia, phonophobia, nausea, and vomiting are also common symptoms. Unlike the FHM prodrome, however, the accompanying headache does not easily differentiate between FHM subtypes or between other more common migraine phenotypes except for its longer duration. The aura merits particular attention in our view. Visual auras are particularly common in FHM types 1 and 2 and MWA, and the appearance and progression of the visual percept are grossly similar for each. Cortical spreading depression or CSD-like events within occipital cortex generate most visual auras and represent transient slowly spreading excitation (depolarization) followed by long-lasting depression (hyperpolarization).12,13,14 (On the basis of its uniquely slow propagation velocity, CSD may underlie the progressive march of aura symptoms arising from primary somatosensory cortex as well.) In MWA patients, Hadjikhani and colleagues found eight characteristics of CSD during visual aura using high field strength nearcontinuous magnetic resonance recordings; these include similar propagation velocity, duration, and suppression of light-induced cortical activation.14 Moreover, CSD activates trigeminal meningeal afferents and contributes to the genesis of headache.15 Of potential relevance to FHM, CSD causes long-lasting mild hemiparesis in experimental animals.16 In more typical migraine subtypes, motor weakness is distinctly unusual, which is perhaps related to unique regionspecific architectonic features or synaptic organization 278 Annals of Neurology Vol 55 No 2 February 2004 rendering, for example, some areas of cortex (eg, visual) more susceptible but human primary motor cortex relatively more resistant to CSD. Reconciling an Apparent Mismatch If CSD causes migraine aura, how do mutations in the P/Q calcium channel and Na⫹,K⫹-ATPase genes modulate susceptibility and sustained attacks? Glutamate, the predominent excitatory amino acid transmitter in brain, and the astrocyte take center stage in this formulation. For example, glutamate or its N-methyl-Daspartate receptor agonist triggers CSD, whereas MK801, an N-methyl-D-aspartate receptor antagonist, blocks its propagation.17 In the wild type, astrocytes terminate synaptic activity by removing glutamate from the synaptic cleft via glutamate transporters. An electrochemical Na⫹ gradient across the plasma membrane is required to drive the astrocytic glutamate transporter. The task of maintaining and reestablishing the Na⫹ gradient is that of Na⫹,K⫹-ATPase, extruding astrocytic sodium in exchange for extracellular potassium. The Na⫹,K⫹ pump requires energy from ATP hydrolysis, and ATP is generated within astrocytes by glycolysis. Lactate then is shuttled from astrocytes to neurons through specific transporters located in each cell type and oxidized to support the energetics of normal and perhaps intense synaptic activity (Fig 2). One attractive hypothesis based on the above dynamics proposes that astrocytes and their foot processes (poised on the surface of blood vessels) couple glucose uptake/ blood flow to glutamatergic synaptic activity.18 It follows from the above that FHM mutations render the brain more susceptible to prolonged CSD caused by either excessive synaptic glutamate release (type 1) or decreased removal of glutamate and K⫹ from the synaptic cleft (type 2) (Fig 2). In FHM1, a lower CSD threshold can be attributed to attendant changes in glutamate release due to enhanced calcium influx into presynaptic terminals.8 In FHM2, one can posit that clearance of synaptic glutamate and K⫹ is slowed because of Na⫹,K⫹-ATPase haploinsufficiency2 as well as insufficient lactate production that may not meet neuronal demand during intense synaptic activity. Extracellular and intracellular K⫹ and Na⫹ levels build up respectively, and the glutamate transporter is slowed or even reversed causing an increase in extracellular glutamate levels. Consistent with this formulation, Na⫹,K⫹- ATPase inhibitors or KCl trigger CSD when applied topically to brain slices, and presumably blockade of astrocytic glutamate transporters would do the same. It is reasonable to postulate that CSD susceptibility and impaired recovery relates to either excessive glutamate release (type 1) or reduced clearance via mutations in astrocytic pumps (type 2) or perhaps even transporters. Other proteins (eg, transporters or channels) relevant to CSD, but not necessarily Cav2.1 or Fig 2. Mutations in genes encoding proteins expressed by neurons and glia are linked to familial hemiplegic migraine (FHM1 and FHM2). A pivotal role for glutamate is proposed to explain the susceptibility to cortical spreading depression, implicated in migraine aura. Depicted is a glutamatergic synapse and cerebral blood vessel, the latter providing a source of glucose. After depolarization, glutamate is released into the synaptic cleft regulated by Cav2.1 (green) gating calcium influx. Synaptic activity is terminated in part by astrocytic uptake of glutamate via transporters (GLAST) driven by sodium gradients. Na⫹ gradients are maintained by activity of Na⫹,K⫹-ATPase removing sodium from inside cells. Energy is required and achieved by glucose utilization after uptake from blood vessels. Lactate so generated is transported and oxidized within neurons to support the excessive energy needs of synaptic activity. Under basal conditions, direct glucose uptake may occur in neurons as well. Susceptibility to cortical spreading depression is enhanced by gain of function mutation in Cav2.1 and increased synaptic release of glutamide from neurons. Loss of function mutation in Na⫹,K⫹-ATPase expressed by astrocytes raise extracellular glutamate and potassium. Modified from Magistretti and colleagues.18 Na⫹,K⫹ pump, also may render the nervous system susceptible to CSD and migraine. In view of its relationship to the Na⫹,K⫹ pump, a role for the Na⫹/ Ca⫹⫹ exchanger and buildup of intracellular Ca⫹⫹ needs to be considered in this formulation as well. Interestingly, single nucleotide polymorphisms in the insulin receptor gene are associated with migraine.19 These receptors are expressed on astrocytes and possibly have an impact on CSD by disrupting glucose utilization and the energetics of synaptic metabolism. Disturbances in synaptic energy metabolism may also relate to migraine headache development in patients with a mitochondrial DNA polymorphism (mitochondrial encephalomyopathy with lactic acidosis and stroke-like episodes).20 It appears possible that during cell stress (conditions not yet defined), a normally compensated but marginally operative glutamate cycle (reduced safety factor) may fail, causing paroxysmal onset and suppression of activity-dependent blood flow augmentation during and after CSD. Obviously, there is much to be learned about the neurobiological consequences of migraine mutations and CSD-associated genes. Nevertheless, using the above formulation, it is theoretically possible to comprehend how a gain-offunction mutation (type 1) and loss-of-function mutation (type 2) expressed on distinct cell types encoding ion channel fluxing genes regulating either monovalent or divalent cations generate a remarkably overlapping migraine phenotype. This work was supported by the National Institutes of Health Interdepartmental Migraine Program Project (NS-35611, M.A.M.). Moskowitz et al: Deciphering Migraine Mechanisms 279 References 1. Ophoff RA, Terwindt GM, Vergouwe MN, et al. Familial hemiplegic migraine and episodic ataxia type-2 are caused by mutations in the calcium channel gene CACNL1A4. Cell 1996; 87:543–552. 2. De Fusco M, Marconi R, Silvestri L, et al. Haploinsufficiency of ATP1A2 encoding the Na⫹,K⫹ pump ␣2 subunit gene is responsible for familial hemiplegic migraine type 2. Nat Genet 2003;33:192–196. 3. Qian J, Noebels JL. Presynaptic Ca2⫹ channels and neurotransmitter release at the terminal of a mouse cortical neuron. J Neurosci 2001;21:3721–3728. 4. McGrail KM, Phillips JM, Sweadner KJ. Immunofluorescent localization of three Na,K-ATPase isozymes in the rat central nervous system: both neurons and glia can express more than one Na,K-ATPase. J Neurosci 1991;11:381–391. 5. Ptacek LJ. The place of migraine as a channelopathy. Curr Opin Neurol 1998;11:217–226. 6. Kors EE, van den Maagdenberg AM, Plomp JJ, et al. Calcium channel mutations and migraine. Curr Opin Neurol 2002;15: 311–316. 7. Ayata C, Shimizu-Sasamata M, Lo EH, et al. Impaired neurotransmitter release and elevated threshold for cortical spreading depression in mice with mutations in the 1A subunit of P/Q type calcium channels. Neuroscience 2000;95:639 – 645. 8. Tottene A, Fellin T, Pagnutti S, et al. Familial hemiplegic migraine mutations increase Ca2⫹ influx through single human Cav2.1 channels and decrease maximal Cav2.1 current density in neurons. Proc Natl Acad Sci USA 2002;99: 13284 –13289. 9. Pietrobon D, Striessnig J. Neurobiology of migraine. Nat Rev Neurosci 2003;4:386 –398. 280 Annals of Neurology Vol 55 No 2 February 2004 10. Vanmolkot KRJ, Kors EE, Hottenga J-J, et al. Novel mutations in the Na⫹, K⫹-ATPase pump gene ATP1A2 associated with familial hemiplegic migraine and benign familial infantile convulsions. Ann Neurol 2003;54:360 –366. 11. Ducros A. The clinical spectrum of familial hemiplegic migraine associated with mutations in a neuronal calcium channel. N Engl J Med 2001;345:17–24. 12. Leao AAP. Spreading depression of activity in the cerebral cortex. J Neurophysiol 1944;7:359 –390. 13. Cao Y, Welch KMA, Aurora S, Vikingstad E. Functional MRIBOLD of visually triggered headache in patients with migraine. Arch Neurol 1999;56:548 –554. 14. Hadjikhani N, Sanchez Del Rio M, Wu O, et al. Mechanisms of migraine aura revealed by functional MRI in human visual cortex. Proc Natl Acad Sci USA 2001;98:4687– 4692. 15. Bolay H, Reuter U, Dunn A, et al. Intrinsic brain activity triggers trigeminal meningeal afferents in a migraine model. Nat Med 2002;8:136 –142. 16. Alexis NE, Back T, Zhao W, et al. Neurobehavioral consequences of induced spreading depression following photothrombotic middle cerebral artery occlusion. Brain Res 1996; 706:273–282. 17. Nellgard B, Wieloch T. NMDA-receptor blockers but not NBQX, an AMPA-receptor antagonist, inhibit spreading depression in the rat brain. Acta Physiol Scand 1992;146: 497–503. 18. Magistretti PJ, Pellerin L, Rothman DL, Shulman RG. Energy on demand. Science 1999;283:496 – 497. 19. McCarthy LC, Hosford DA, Riley JH, et al. Single nucleotide polymorphism alleles in the insulin receptor gene are associated with typical migraine. Genomics 2001;78:135–149. 20. Goto Y, Nonaka I, Horai S. A mutation in the tRNALeu(UUR) gene associated with the MELAS sub-group of mitochondrial encephalomyopathies. Nature 1990;348:651– 653.