The cerebral cortex is a sheet of neurons, approximately
the size of a small pizza, and about the thickness of
the cardboard underneath. It is massively interconnected,
vertically and especially horizontally. Neurons of the
cerebral cortex have a resting membrane potential, in
the lack of synaptic inputs, approximately 20 mV away
from firing threshold. We hypothesize that the ongoing,
spontaneous activity of the cortex provides the bulk of
this 20 mV of depolarization needed to get the cell near
firing threshold. Thus, the recurrent connectivity of
the cortex is built so that much of the 10,000 to 30,000
synaptic inputs provide a context within which the cell
operates.
That context, generated largely through recurrent network
activity in the cortex, rapidly controls neuronal excitability
and functional connectivity, allowing for the functional
grouping and ungrouping of cortical neuronal networks
in a manner necessary for behavior. Two examples of the
utility of these functional networks are those of working
memory and selective attention. Working memory allows
one to keep in mind features of an ongoing task, putting
together sequences into a logical whole. Neuronal networks
in the prefrontal cortex exhibit persistent activity that
is correlated with working memory. In attention, neuronal
responsiveness rapidly increases and decreases with attention/
inattention. What are the mechanisms by which neuronal
activity and responsiveness may be rapidly regulated within
the cerebral cortex?
Our research has focused on the ability of local networks
within the cerebral cortex to, either spontaneously, or
in response to activation of an afferent input, generate
rapid changes in neuronal activity and responsiveness
through recurrent network activity. During anesthesia
and slow wave sleep, this recurrent network activity is
generated by a fine balance between local feedback excitatory
and inhibitory loops. The activation of these loops results
in a relatively stable state in cortical networks, such
that the neurons are poised near firing threshold and
spontaneously active in what is termed an UP state. We
suggest that a basic operating principle of the cerebral
cortex in the waking animal is the rapid and dynamic reconfiguration
of functional networks based upon the selection and activation
of recurrent networks of cells, in a manner similar to
the generation of UP states during sleep and anesthesia.
In this manner, the ongoing activity of the large cortical
sheet may provide the behavioral and intellectual context
to each cell and local network needed to provide appropriate
responsiveness. This hypothesis is currently being thoroughly
examined and tested.
