The dentate-CA3 regions as reciprocally connected networks that produce sequence recall (phase precession):  CA3 is the autoassociative network while heteroassociation (links between sequential memories) occur in backprojecting synapses in the dentate. Another function of dentate is to establish a representation that combines information about self (from the medial entorhinal region) with information about non-self (from the lateral entorhinal cortex).

The hippocampal-VTA loop: novelty detection in the hippocampus activates the dopamine system, thereby activating late LTP and gating the flow of information into long-term memory.

Theta/gamma nested oscillations form a neural code:
: an item is represented by firing within a gamma cycle; different items are represented in order at different discrete phases of a theta cycle. This code is used for memory recall in the hippocampus. It may also be used in cortex and explain the basis of the 7+-2 capacity limit on short-term memory (there are ~7 gamma cycles in a theta cycle).

Role of NMDA-mediated EPSP's: the voltage-dependence of such EPSP's can give rise to bistable signals that provide critical mechanisms for working memory and integrator networks.

Different roles of different Ca2+ levels in synaptic plasticity: high Ca triggers LTP (via kinase reactions); moderate Ca triggers LTD / depotentiation (via phosphatases).

Memory storage by CaMKII: CaMKII as molecular switch with the stability, reversibility and structural involvement necessary to store synaptic memory.

The quantal response: Monte-Carlo simulations show that the quantal response at synapses is generated by AMPA channels in a "hotspot" of 200 nm diameter, a result that has important implications for transmission and LTP.

The equivalent light hypothesis for retinal degeneration: Many mutations that underlie the disease produce a common physiological change: the persistent activation of the transduction cascade. This, like constant real light, triggers photoreceptor degeneration.

CaMKII as molecular memory: CaMKII as molecular memory: We monitor CaMKII in living neurons using GFP-labelling and relate this to synaptic events triggered by 2-photon glutamate uncaging. These methods provide strong tests of the CaMKII hypothesis of synaptic memory.

The key test for any hypothesis regarding the molecular basis of memory is to turn the memory on and then attempt to turn if off by attacking a molecule. We have now succeeded in doing this for synaptic memory: CaMKIIN-tide, a molecule that interferes with the binding of CaMKII to the NMDA channel, can reverse LTP. Our working hypothesis is that synaptic memory is encoded structurally and is proportional to the amount of the CaMKII/NMDAR complex at the synapse. This is now being tested in a variety of ways. 

NMDAR, dopamine and schizophrenia:  NMDA antagonists cause schizophrenia-like symptoms in humans. We are exploring the mechanisms by which these antagonists affect cellular and network processes in slices of rat cortex and thalamus. Antagonists appear to selectively reduce the excitation of interneurons. The resulting disinhibition of pyramidal cells, particularly in the hippocampal region, may lead to overexcitation of the dopamine system (through the hippocampal-VTA loop) and thereby produce psychosis.


Pictures of my colleagues in science that I've taken over the last 25 years.


FastCounter by bCentral
Copyright © 2000, Steven Erat of TalkingTree Web Design and Development