Ken Harris (Rutgers), 11/25/03
Organization of cell assemblies in the hippocampus
It is now possible to record from hundreds of neurons simultaneously
in behaving animals. However, advances in spike train recording
technology have not been matched by advances in the statistical
methods needed to draw biological conclusions from this new data. In
this talk I will describe some recent research using multiple
simultaneous recordings in hippocampus (Harris et al, Nature 2003),
paying particular attention to the novel statistical methods used.
According to Hebb's "cell assembly" hypothesis, information is
represented in the brain by groups of anatomically distributed neurons
which come together briefly in synchronous activity, and whose
activity underlies both processing of external sensory input, and
internal cognitive processes. Accordingly, neuronal populations
should show an arrangement into synchronous groups, and the firing
patterns of these groups should show coordination beyond that
predicted by common modulation by sensory input.
Using parallel recordings in the behaving rat, we found it was
possible to predict the exact spike times of hippocampal pyramidal
neurons from the spike times of simultaneously recorded neurons,
better than from the animal's trajectory in space, or from a
spatially-dependent theta phase modulation (O'Keefe and Recce,
Hippocampus, 1993; Harris et al, Nature, 2002). This suggests that
hippocampal neurons are organized into assemblies whose activity is
not controlled strictly by the time-course of sensory input. The time
window within which spike times were best predicted from simultaneous
peer activity was 20-30ms, suggesting that this is the timescale at
which cell assemblies are synchronized. Because this temporal window
matches the membrane time constant of pyramidal neurons, the period of
the hippocampal gamma oscillation, and the time window for synaptic
plasticity, we suggest that cooperative activity at this timescale is
optimal for information transmission and storage in cortical circuits.