Coincidence Detection In Cortical Cells addresses an
issue that is part of the more basic problem of how the
environment is represented in the brain such that an organism
can respond in an appropriate manner when the environment
changes. In other words what are possible cellular and
molecular mechanisms that underlie those brain functions
which are sometimes referred to as "higher" brain functions
such as recognition of an object, learning and retrieval
from memory?
The underlying assumption we make in trying to understand
such functions is that the brain operates by electrical
signals generated by the constituting nerve cells. An
external stimulus, say a visual scene and its different
features, is encoded by a sequence of action potentials
in ensembles of neurons in the six cortical layers. If
we want to understand brain functions mechanistically
we have to be able to monitor the ever changing pattern
of electrical signals in neuronal ensembles. This means
to find out how the complex pattern of signals in the
so called "representational" cortical fields is generated
and how the stimulus is "encoded" by the ever changing
pattern of electrical signals. There are several views
on how this may happen.
One view is referred to as "Rate coding" the other one
as "Temporal coding". Both types of stimulus representation
require an ensemble of "read-out" neurons. These must
be sensitive to the temporal structure of the electrical
activity of those cortical neurons that are lower in the
hierarchical order. For example they must be sensitive
to synchronous activity of the lower-order neurons.
Possible mechanisms of coincidence detection are found
in the large neurons of cortical layer 5 which are the
cortex's main read-out neurons. Multiple simultaneous
recordings from different compartments of single L5 neurons
suggested that at threshold synaptic stimulation sodium
dependent action potentials are initiated in the axon
which propagate actively in two directions - orthogradely
into the axonal arbor and retrogradely into the dendritic
arbor. With stronger stimulation an additional action
potential initiation zone in the distal apical dendrite
is operating. In this dendritic zone calcium dependent
action potentials are initiated that spread along the
dendrite to the axonal initiation region to influence
the output discharge pattern of the neuron in a specific
way. Thus the L5 pyramidal neurons of the neocortex have
two action potential initiation zones. They can interact
in such a way that synaptic input to the basal dendrites
from layer 5, when occurring almost coincident (within
a few milliseconds) with synaptic input to the apical
tuft from layer 1 will generate a high frequency burst
of several action potentials.
The intracortical innervation pattern of L5 pyramidal
neurons is such that the generation of burst firing by
coincident synaptic inputs to the basal and apical dendrites
repectively would be a mechanism to associate local, columnar
synaptic input to the basal dendrites arriving from a
primary sensory area with an input to the apical tuft
dendrites arriving from secondary sensory areas. Furthermore,
synchronized burst activity of connected cortical neurons
can change the strength of their synaptic connections.
Coincident input dependent burst activity can represent
one mechanism that initiates long-term changes in the
synaptic efficacy of wiring of cortical neurons.
In summary, one could speculate that the generation of
bursts of action potentials in layer 5 of the cortex signals
to other parts of the brain that coincident activity in
two sensory areas, say the visual and somatosensory areas,
has occurred. If this happens repeatedly the connection
between the bursting neurons would be strengthened leading
to a lower threshold for the reaction of the organism
to a specific set of external stimuli.
