摘要
Migraine is a common and highly disabling neurological disorder that affect more than 10% of the general population; in 30% of the migraneours migraine headache is preceded by transient neurological symptoms referred as “migraine aura”. Familial Hemiplegic Migraine type 1 (FHM1) is a rare and very severe autosomal dominant subtype of migraine with aura (MA), that is considered a good model for studying migraine: FHM1 typical attacks resemble general MA attacks, but aura symptoms include also motor weakness or paralysis, often but not always unilateral. FHM1 is caused by mutations in the CACNA1A gene, with encodes the α1, pore-forming subunit of neuronal CaV2.1 voltage-gated Ca2+ channels, also referred as P/Q-type Ca2+ channels. FHM1 mutations produce gain-of-function of human recombinant CaV2.1 channels, mainly due to a shift of channel activation to more negative voltages and an increase of the open probability and single channel influx over a wide voltage range (Hans et al., 1999; Tottene et al., 2002, 2005; Pietrobon). Facilitation of experimentally induced cortical spreading depression (CSD) in knock-in (KI) mice carrying the human FHM1 mutation R192Q has been observed in vivo (van den Maagdenberg et al., 2004) and in vitro (Tottene et al., 2009). CSD, a wave of strong neuronal and glial depolarization that slowly progresses across the cerebral cortex, followed by long-lasting neural suppression, is the phenomenon underlying migraine aura and a likely primary cause of migraine headache. Tottene et al. (2009) recently demonstrated enhanced excitatory neurotransmission due to increased action potential-evoked Ca2+ influx through CaV2.1 channels and increased probability of glutamate release at cortical pyramidal cell synapses of R192Q KI mice. Using an in vitro model of CSD Tottene et al. also demonstrated a causative link between enhanced glutamate release and facilitation of CSD induced by brief pulses of high KCl in cortical slices of KI mice. These findings support a key role of CaV2.1 channels in CSD initiation and propagation. In the first part of my Ph. D project I further investigated the mechanisms of CSD induction and propagation and in particular the role of NMDA receptors and voltage-gated calcium channels. Results obtained were published in Tottene el al., 2011 (pdf attached in chapter 3). We adopted the in vitro model of CSD described in Tottene et al., 2009 and measured CSD threshold and velocity in acute cortical slices from WT mice before and after application of specific blockers. CSD was induced by pressure pulses of high KCl of increasing duration; the duration of the first pulse eliciting a CSD was taken as CSD threshold and the rate of horizontal spread of the change in intrinsic optical signal produced by CSD as CSD velocity. We investigated the role of NMDA using the general NMDA receptors inhibitor D-AP5 (50 μM). In the presence of D-AP5 stimuli up to 30-time the duration of the control threshold failed to induce a CSD. We investigated also the role of P/Q, N, R, and L-type voltage-gated calcium channels using the respective specific blockers: ω-agatoxin IVA (300 nM), ω-conotoxin GVIA (1 μM), SNX-482 (250 nM), and nimodipine (50 μM). Similarly to what observed after application of D-AP5, in the presence of ω-AgaIVA stimuli up to 30-time the duration of the control threshold failed to induce a CSD. Application of either ω-CgTxGVIA and SNX-482 only slightly increased CSD threshold (of about 10%) and slightly decreased CSD velocity (of about 15%) suggesting that N- and R-type voltage-gated Ca2+ channels may have a modulatory role in CSD induction and propagation. Application of nimodipine did not affect CSD induction nor propagation. The findings that application of D-AP5 or ω-AgaIVA inhibits CSD induction while ω-CgTxGVIA, SNX-482 and nimodipine have only a poor or null effect are consistent with and support a model of CSD initiation and propagation in which activation of presynaptic (and possibly postsynaptic) CaV2.1 channels and NMDA receptors plays a pivotal role (Tottene et al., 2011). The role of NMDA receptors was further investigated in FHM1 R192Q mice, which show an increased glutamate release from pyramidal neurons. In contrast to what happens in WT mice, application of D-AP5 in brain slices from KI mice did not prevent CSD induction, but significantly increased CSD threshold (90 ± 9%) and significantly decreased CSD velocity (47 ± 1%) relative to control. The different effect of D-AP5 in WT and KI mice suggests that other elements (beside NMDA receptors) could be involved in CSD initiation mechanisms. Current clamp recordings close to the KCl injection pipette (at fixed distance of 100 and 200 μm) allowed to observe different phases in voltage changes after the application of both sub-threshold and threshold KCl stimuli: it was possible, in fact, to distinguish three distinct peaks. While the first two peaks were present in both sub-threshold depolarizations and CSD, the third peak was specifically linked to CSD. Patch-clamp technique allowed to accurately describe different CSD phases but showed the obvious limitation to be restricted to only one cell. Ca2+ imaging of slices loaded with the Ca2+ sensitive indicator OGB1-AM allowed to observe several neurons at different distances from KCl-injecting pipette at the same time for each pulse. Simultaneous patch-clamp recordings and Ca2+ imaging showed a general correlation between voltage changes recorded in current-clamp and fluorescence peaks detected in Ca2+ traces. The physiological meaning of these different phases and their specific role in CSD induction and propagation are not clear and are currently object of study in our laboratory. Ca2+ imaging allowed to detect also intracellular calcium increases from astrocytes in response to KCl pulses. Astrocytes can support and amplify glutamatergic synaptic transmission by releasing glutamate that specifically act on NMDA receptors; astrocytic release of glutamate and other substances is induced by [Ca2+]int increases that are, in turn, stimulated by neuronal release of glutamate and other neurotransmitters. To test whether the neuron-astrocyte crosstalk is involved in CSD induction and propagation, threshold for CSD induction and velocity of CSD propagation were measured before and after inhibiting astrocytic [Ca2+]int increases by the application of cyclopiazonic acid (CPA, 50 μM), that emptied astrocytic Ca2+ internal stores without affecting neuronal response to KCl. CPA application did not affect CSD threshold or velocity in both WT and KI mice. These results show that neuron-astrocyte crosstalk seem not to be relevant in CSD induction and propagation. The combination of Ca2+ imaging and patch clamp recording permitted to observe, in the absence of external stimuli, spontaneous activity that occurs in the neocortical neural networks as slow oscillations between periods of synaptically-driven depolarizations with firing (up-states) followed by decreased synaptic inputs and ceasing of firing (down-states). At the same time spontaneous large Ca2+ oscillations were observed in astrocytes. To check whether neuron-astrocyte crosstalk could be involved in the generation or in the modulation of spontaneous neuronal network activity, astrocytic spontaneous oscillations were inhibited by application of CPA. In presence of CPA up-states were still present in both WT and KI mice; however CPA application led to a significant reduction of up-states frequency in WT (41 ± 3%, p = 0.0001), but not in KI mice (11 ± 6%, p = 0.12). These preliminary results are in agreement with a possible role of neuron-astrocyte crosstalk in spontaneous neuronal network activity.