mise à jour du
5 mai 2007
Eur J Neurosci.
Effect of cortical spreading depression on synaptic transmission of rat hippocampal tissues.
Wernsmann B, Pape HC, Speckmann EJ, Gorji A.
Institut fur Physiologie I, Westfalische Wilhelms-Universitat Munster, Germany.
Yawning and migraine


Abstract : Cortical spreading depression (CSD) is believed to be a putative neuronal mechanism underlying migraine aura and subsequent pain. In vitro and ex vivo/in vitro brain slice techniques were used to investigate CSD effects on the field excitatory postsynaptic potentials (fEPSP) and tetanus-induced long-term potentiation (LTP) in combined rat hippocampus-cortex slices. Induction of CSD in combined hippocampus-cortex slices in which DC negative deflections did not propagate from neocortex to hippocampus significantly augmented fEPSP amplitude and LTP in the hippocampus.
Propagation of CSD to the hippocampus resulted in a transient suppression followed by reinstatement of fEPSP with amplitude of pre-CSD levels. LTP was inhibited when DC potential shifts were recorded in the hippocampus. Furthermore, CSD was induced in anaesthetized rats and, thereafter, hippocampal tissues were examined in vitro. LTP was significantly enhanced in hippocampal slices obtained from ipsilateral site to the hemisphere in which CSD was evoked.
The results indicate the disturbances of hippocampal synaptic transmission triggered by propagation of CSD. This perturbation of hippocampal synaptic transmission induced by CSD may relate to some symptoms occurring during migraine attacks, such as amnesia and hyperactivity.
Spreading depression (SD) is a pronounced self-propagating depolarization of neurons and glia with a transient massive redistribution of ions between intracellular and extracellular compartments.
SD spreads slowly from the site of onset as a radial wave across the neuronal tissues followed by a transient period of depressed bioelectrical activity (Leao, 1944; Somjen, 2001).
Studies on the human brain suggested that SD might be a clinical phenomenon. SD occurred following head injury or intracranial haemorrhage in human neocortex (Mayevsky et al., 1996; Strong et al., 2002), and SD-like waves were observed during the aura phase of migraine attacks (Hadjikhani et al., 2001).
Furthermore, experimental investigations indicated that SD may also play a role in transient global amnesia, epilepsy (Gorji, 2001; Gorji & Speckmann, 2004) and spinal cord disorders (Vinogradova et al., 1991; Gorji et al., 2004). SD was originally linked to the aura phase of migraine (Lashley, 1941; Leao & Morrison, 1945; Pearce, 1985).
However, some evidence suggested that SD is also implicated in migraine pain as well as other signs and symptoms such as amnesia, hyperactivity, drowsiness and yawning, sexual arousal, nausea and vomiting, pupil reaction, and fluid retention (Gorji, 2001).
The potential interrelation of SD and migraine has usually involved neocortical tissues, and the possible roles in other brain regions including the hippocampus were not fully investigated. The hippocampus has direct and important functional interactions with brain areas likely to be important to migraine, such as the areas associated with vision, emotions and neuroendocrine homeostasis.
The connection between entorhinal cortex and hippocampus is regarded as an important loop responsible for the processing of sensory information (Vaisanen et al., 1999). Thus, these medial temporal lobe structures may play a crucial role in the development of somatosensory and neuropsychotic symptoms in neurological disorders such as epilepsy and migraine (Eid et al., 1995).
As memories in humans depend initially on the medial temporal lobe system, including the hippocampus, it was suggested that interictal memory dysfunction in patients with migraine might be attributed to the hippocampus involvement (Kupfermann, 1966; Kapp & Schneider, 1971). Furthermore, propagation of SD in the hippocampus was believed to play a role in migraine pain by triggering nociceptive activation of the caudal trigeminal nucleus (Kunkler & Kraig, 2003).
Classical studies investigated hippocampal SD more often by implantation of KCl into the hippocampus and induction of SD directly in the tissue. Little information is available on the effects of cortical spreading depression (CSD) on hippocampal activity. Because altered neural circuit function can be seen remote from the SD propagation site (Bures et al., 1961; Albe-Fessard et al., 1984; Moskowitz et al., 1993; Kunkler & Kraig, 2003; Gorji et al., 2004), using in vitro and ex vivo in vitro brain models, the effects of neocortical SD on the synaptic plasticity of hippocampal tissues were tested.
The present study shows how differently CSD could affect hippocampal activity. Regarding our findings, abortive CSD, i.e. SD that travelled the whole neocortex and entorhinal cortex but stopped entering the hippocampus, enhanced the fEPSP as well as LTP in CA1 area of combined hippocampal&endash;entorhinal cortex slices. `
On the contrary, CSD spreading from the temporal neocortex all the way to CA3 area transiently suppressed evoked fEPSP and reduced LTP in the hippocampal tissues. Intrinsic optical imaging also revealed different patterns of SD propagation in hippocampal and entorhinal slices (Buchheim et al., 2002). Although the exact mechanisms responsible for different propagation patterns of SD are not clear, some hypotheses can be derived from experimental data. A traditional view assumes that the entorhinal cortices faithfully transmit neocortical inputs to the hippocampus and vice versa (Naber et al., 1999).
More recent evidence suggests that the entorhinal cortices are more than a simple relay between the neocortex and hippocampus. Entorhinal cortices contribute to the gating of impulses between these structures. Local inhibition and intrinsic membrane properties of entorhinal neurons are major factors limiting impulse traffic across the entorhinal cortex (Pelletier et al., 2004).
Consistent with this, physiological studies have disclosed the existence of powerful inhibition in the entorhinal cortex (Finch et al., 1986; Jones & Buhl, 1993; Funahashi & Stewart, 1998), which may act to abort the propagation of SD. Furthermore, some studies indicated a relative resistance of SD occurrence in the hippocampus compared with entorhinal cortex (Dalby & Mody, 2003; Faria & Mody, 2004). `
The failure of cortical SD to propagate to the hippocampus was reported earlier (Fifkova, 1964). The release of glutamate is essential to the propagation of cortical SD (Van Harreveld & Fifkova, 1973). Several studies have shown that glutamate acts via NMDA receptors during the generation and propagation of SD (Mody et al., 1987; Gorji, 2001). The NMDA receptor is a heterotetramer assembled from NR1 subunits and at least one subtype of the four members of the NR2(A&endash;D) subunits family. NR2B subunits are essential to the generation and propagation of SD in entorhinal cortical slices (Faria & Mody, 2004). The physiological characteristics and possibly the localization of NR2B subunits at synapses differ between the entorhinal cortex and the hippocampus (Gordey et al., 2001; Faria & Mody, 2004), which, in turn, may influence SD penetration into the hippocampus.
Seventy per cent of CSD waves propagating from temporal cortex slices penetrated to adjacent entorhinal cortex slices and stopped there, whereas the remaining 30% reached CA1 and CA3 regions of the hippocampal slices. On the other hand, CSD elicited from the somatosensory neocortex of anaesthetized rats did not penetrate into the hippocampus.
This suggests that the CSD recording in slices offers better conditions for SD propagation probably due to weakening of intrahippocampal inhibitory mechanisms. The entorhinal cortex, a palaeocortical area, receives projections from secondary and higher associative areas of the neocortex. Regions of both ipsilateral frontal and temporal lobes are found to contribute afferents to this region of the brain.
The association areas from the primary sensory modalities of vision, audition and somesthesis project to multimodal convergence areas in the frontal and parietal lobes (Pandya & Kuypers, 1969). Both multimodal regions project in turn to the cingulate gyrus on the medial surface of the hemisphere, which contributes a heavy supply of afferents to the presubiculum and entorhinal cortex (Jones & Powell, 1970). Thus, the entorhinal cortex is a final cortical link between the sensory systems of the neocortex and the hippocampus of the limbic system. From the entorhinal input, the hippocampus receives highly complex and differentiated signals, coding information about the properties of the applied stimuli. The entorhinal cortical neurons constitute the direct perforant path and the crossed temporoammonic path to the hippocampus. They terminate on dendritic branches of CA1&endash;CA3 and the dentate fascia neurons (Van Hoesen et al., 1972). In the present study, transient sensory cortical dysfunction induced by abortive SD enhanced hippocampal activity.
This suggests an inhibitory tone mediated through neocortical influence on hippocampal plasticity. Our conclusion is supported by recent evidence indicating that elimination of cortical input resulted in increased reactivity and complete disappearance of habituation, with prolongation of tonic responses in the hippocampus (Vinogradova, 2001). Lesions of the entorhinal cortex in adolescent rats also resulted in augmented spontaneous locomotor activity, an effect possibly mediated by postsynaptic hypersensitivity (Sumiyoshi et al., 2004). LTP is an experimental phenomenon, which can be used to demonstrate the repertoire of long-lasting modifications of which individual synapses are capable.
LTP remains one of the prime candidates for mediating learning and memory as well as many other forms of experience-dependent plasticity (Malenka & Bear, 2004). In the present study, functional disruption of neocortical input to the hippocampus induced by abortive SD in both in vitro and ex vivo experiments enhanced the LTP in the CA1 hippocampal area ipsilateral to SD initiation.
Enhancement of LTP by abortive SD was NMDA receptor dependent, as APV blocked LTP induction. Further propagation of SD to the hippocampus, conversely, inhibits LTP. These data indicate the modulatory role of SD on the efficacy of the hippocampal synaptic transmission. In line with our results, CSD visualized using manganese-enhanced MRI following topical application of KCl to the exposed rat cortex revealed signal enhancement in CA1&endash;3 areas, the subiculum and the dentate gyrus of the hippocampus (Henning et al., 2005).
The increase in synaptic responsiveness associated with LTP has been discussed to be detrimental to neuronal function in the sense that it can promote seizure activity (Johnston, 1996). It is thus interesting to hypothesize that the facilitatory effect of abortive SD on evoked synaptic activity and LTP observed in the hippocampus may contribute to seizure generation or epileptogenesis. Indeed, the expression of neuronal hyperexcitability associated with SD has been suggested to link migraine and epilepsy (Leniger et al., 2003).
In line with this are preliminary data indicating that SD precede epileptic activity in human brain tissue (Gorji & Speckmann, 2005). Some evidence implicates the hippocampus in spatial memory and navigation, learning and emotion (Jensen & Lisman, 2005). This structure is also related primarily to the control of gross movements, such as locomotion and changes in posture, and involved in certain aspects of the pituitary&endash;adrenocortical system. Amnesia, emotional disturbances, hyperactivity, yawning and fluid retention were observed in hippocampal dysfunction as well as during migraine attacks (Bures et al., 1974; Isaacson & Pribram, 1975; Dalessio, 1980; Daquin et al., 2001). SD in animal experiments also elicits similar symptoms (Gorji, 2001). SD-like changes occur with visual aura in patients with migraine (Hadjikhani et al., 2001). Propagation of depolarizing waves in sensory systems of the neocortex may directly affect primary sensory modalities and induce aura symptoms such as visual hallucinations. SD, either indirectly via the effect on entorhinal input to the hippocampus or directly by propagation to the hippocampal structure, may disturb the hippocampal function and lead to symptoms such as amnesia or hyperactivity during migraine attacks.