Freie Universität Berlin
Charité University Medicine Berlin
Humboldt University Berlin
Max-Delbrück-Center for Molecular Medicine, Berlin-Buch
|| Home || Positions || Contact ||

GRK 1123:

Cellular Mechanisms of Learning and Memory Consolidation
in the Hippocampal Formation

| Home
| News
| Research
| Study Program
| Events
| Faculty
| Students
| Links

This Research Training Group is funded by the German Research Council DFG

©2006, FU Berlin
GRK 1123

Imprint & Disclaimer
AG Ahnert-Hilger AG Behr AG Geiger AG Haucke AG Heinemann/
Kempter
AG Multhaup AG Wulczyn AG Rosenmund AG Schmitz/
Brecht 		AG Sigrist

Prof. Dr. Uwe Heinemann
Institut für Neurophysiologie
Charité Universitätsmedizin Berlin
Oudenarder Str. 16
13347 Berlin
uwe.heinemann@charite.de

Dr. Richard Kempter
ITB Theoretical Neuroscience
Humboldt Universität zu Berlin
Tucholskystr. 2
10117 Berlin
r.kempter@biologie.hu-berlin.de


Topic

The role of SPW-R complexes in memory consolidation in the hippocampal formation


Title

Induction, incidence and stability of sharp wave ripple complexes in the hippocampal formation

Question of the project

Long-term storage of memory is not hippocampus-dependent. Therefore any information has to be transferred from the hippocampus into the cortical mantle in a process termed memory consolidation. This process is believed to depend on the generation of sharp wave-ripple complexes (SPW-Rs) during which previously stored information is read out in a temporally compressed form characterized by network oscillations of 120 - 200 Hz. SPW-Rs might induce LTP in remote areas and thus may serve for memory consolidation. The stability of such events and the question whether SPW-Rs can induce LTP in remote structures are yet unknown. Regularly, SPW-Rs are preceded by short periods of theta activity, and the mechanisms underlying transition from theta to SPW-R activity are also unknown. Similarly the relationship between global oscillations in the theta or gamma frequency range and induction of SPW-Rs needs attention.

Current State of Research

Storage of information in the hippocampus requires novelty detection and easy propagation of neuronal activity from the peri- and postrhinal cortex through the entorhinal cortex into the hippocampus (Buzsaki, 1989) . Neurons in the hippocampus fire rarely. Therefore most of storage of information is probably dependent on spike timing plasticity requiring a temporal template (Frerking et al., 2005) . Indeed, pairing of presynaptic activity with postsynaptic spikes can easily lead to long lasting strengthening of synaptic coupling in form of long-term potentiation. The temporal template for the induction of synaptic plasticity in networks with rarely firing neurons is likely provided by ensemble gamma activity superimposed on theta cycles (Bartos et al., 2007) . Such activity can be easily induced by acetylcholine trough activation of muscarinic receptors in hippocampal slices, and the induction of LTP is facilitated in presence of acetylcholine, carbachol but also kainate (Weiss et al., 2003) . For memory consolidation the information flow from the peri- and postrhinal cortex must be interrupted and the hippocampus permitted to replay stored information (Buzsaki, 1989) . This is likely occurring during short episodes of neuronal ensemble activity, characterized by sharp waves which are superimposed by high frequency ripples (~200 Hz) (Buzsaki, 1986) . During such events a multitude of neurons in the hippocampus are typically reactivated in the reverse order in which they have been activated during a preceding explorative behavior (Foster and Wilson, 2006) . High frequency network oscillations occurring during ripples provide a pattern of activity, which is well feasible to induce long-term potentiation (LTP) in remote areas thereby strengthening select synaptic connections believed to underlie the formation of memory traces (Buzsáki et al., 1987) . In addition, during SPW-R activity previously stored activity may also occur in the same order as during explorative behavior (Lee and Wilson, 2002;Diba and Buzsaki, 2007) . Such a forward replay potentially might serve to make predictions on a future outcome of a current behavior based on self-gained prior behavioral experience. SPW-Rs usually occur during consummatory behavior and predominantly during slow wave sleep (SWS) (Buzsaki, 1986) . The transition from combined theta-gamma activity to the expression of SPW-Rs and vice versa is poorly understood and thought to be regulated by acetylcholine (ACh). Hereby, low ACh levels would favor the expression of SPW-Rs whereas high levels result in the generation of combined theta/gamma activity (Hasselmo, 1999) . Any specific mechanisms, which trigger the occurrence of recurrent SPW-Rs, are not yet known. Likewise mechanisms which regulate the incidence of SPW-Rs are also unknown. Understanding of such mechanisms might provide means by which loss of memory in patients developing dementia may be ameliorated.


Previous work of the group in the field

We have previously shown that SPW-Rs can be induced in vitro by protocols, which induce LTP upon short stimulus trains using either high frequency or theta burst stimulation (Behrens et al., 2005) . The induction was independent on the site of dorsal versus ventral hippocampus as well as the orientation of a given slice with respect to the lamellar organization (Behrens et al., 2005) . SPW- Rs, once induced, show very stable properties. In fact, the interruption of SPW-R activity either by ACh or by dopamine, norepinephrine or serotonin, indicating a change in the neuronal, activity-dependent state, led to very similar SPW-R activities following washout (Behrens et al, under revision and Haq et al., in preparation).

The induction of SPW-Rs could depend on the occurrence of network oscillations, which might serve as a temporal template for the induction of spike timing-dependent plasticity. In a first step, we therefore asked whether ACh would facilitate the induction of SPW-Rs. This was indeed the case and seemed to depend on both NMDA receptor and Cam Kinase-II activation. Whether BDNF and the protein kinase z M will also facilitate the induction is presently unknown. The Induction of SPW-Rs was also facilitated in presence of nicotine and seemed to depend on alpha 7 containing nicotinic ACh receptors (Liotta et al., submitted). These are prominently expressed on CCK-positive basket cells suggesting a role of such receptors in the induction of SPW-R activity.

The incidence of SPW-Rs was strongly increased when the extracellular potassium concentration was elevated to levels known to occur during slow wave sleep. The reasons for the increase in incidence, which might support memory consolidation, are still unknown but may involve potassium currents. We tested so far for effects of 4-AP, XE-991 and astemizole blockers of Kv1, 3, 7 or 11 family members, respectively. Although some of these agents affected some properties of SPW-Rs, they did not increase the incidence of SPW-R activity (Richter et al., 2008) . Likewise a blocker of SK2 channels did not have any marked effects. We therefore hypothesize that blockers, which affect potassium inward rectifying currents or 2 poor domain potassium currents (K + 2P ) will influence the incidence of such events.

The incidence of SPW-Rs may also depend on the theta activity preceding a given SPW-R. The inducibility of such events may depend on stimulus-induced changes in membrane potential oscillations and resonance. Theta oscillations in isolated hippocampal slices depend on properties of OLM interneurons and pyramidal cells, and may involve changes in H- or M-currents overall resulting in augmented resonance. Work from our lab shows that pharmacologically-induced theta activity can be readily blocked by blockers of H- and M-currents (Boehlen et al., submitted).

If SPW-R activity is a way to read out stored information in a condensed form, and if LTP mechanisms are underlying the formation of memory that SPW-Rs, once induced, should induce LTP-like changes in remote areas is currently under investigation. In a first step we looked therefore for changes in SPW-Rs, once they started to occur, and, in fact, first results consistently show that there seems be a run up until the properties of SPW-Rs are stabilized. Likewise, when the formerly blockade of synaptic transmission in area CA1 or the subiculum is removed, SPW-Rs start getting expressed in these areas with some latency, and then may indeed induce LTP-like alterations in the synaptic transmission.

In a theoretical framework, we have calculated the capacity of a sparsely connected network, such as CA3, to store sequences (Leibold and Kempter, Neural Comput., 2006). We also have derived how synaptic meta-plasticity affects the memory capacity, which equals the memory lifetime for an ongoing storage of new sequences. We found that synaptic meta-plasticity indeed can prolong the lifetime of a memory, but only under the restriction that the neuronal population code is not too sparse. For sparse codes, on the other hand, meta-plasticity may actually hinder memory longevity. This is im portant because in memory-related brain regions, such as the hippocampus, population codes are indeed sparse (Leibold and Kempter, 2008) .

Goals

We wish to understand which mechanisms faciliate generation of SPW-Rm which regulate their stability, their incidence and whether they can indeed induce LTP and other forms of plasticity in remote areas. Using a computational approach, hypotheses can be tested in silico and obtained data tested for plausibility.

Methods

Dissertation topics


Cooperation with other Members  


Scholarship Holders:

 

AG Ahnert-Hilger AG Behr AG Geiger AG Haucke AG Heinemann/
Kempter
AG Multhaup AG Nitsch/
Wulczyn		 AG Rosenmund AG Schmitz/
Brecht 		AG Sigrist