Biomedical Engineering Reference
In-Depth Information
to carry out a specific task, the large number of
exceptions clearly argue against this hypothesis.
Certainly, the physical structure of the dendritic
arbors has a notable influence in the processing
of inputs, and even some computational capabili-
ties arise from it (see Segev and London, 1999).
Nevertheless, most dendrites are active neuron
elements, not mere passive cables. The number
and variety of ion channels in their membranes
modulate synaptic currents with a strength and
complexity largely overcoming that of the cable
structure. These “modulatory” dendritic cur-
rents have the same voltage-dependent nature as
those in the soma and axon, and have been called
intrinsic currents to differentiate them from the
synaptic currents that co-activate and co-localize
with them.
The presence of active electrogenic machin-
ery in dendrites was first reported by founder of
modern electrophysiology Lorente de Nó (1947)
in the first half of the past century. He observed
active backpropagation of action potentials into
dendrites. But it was not until early nineties that
researchers began to grasp the magnitude of the
presence of V-dependent channels in sites other
than the soma and axon. Today, we know that
intrinsic dendritic currents participate in basic
neuron functions as the mentioned conduction
of action potentials from soma to dendrites and
the plastic events born out of their local interac-
tion with synaptic currents. This issue has been
a major object of study in the late years and we
refer the readers to specific literature on the issue
(Johnston et al., 1996; Reyes, 2001).
Our focus here is on the role of dendritic chan-
nels on the forward transmission of inputs, an
issue that has been poorly studied due to technical
difficulties in recording from very thin dendrites.
It is however highly relevant as it concerns to the
very mechanisms of synaptic integration. An
important move forward was the recognition of
the participation of intrinsic dendritic currents in
the reshaping of postsynaptic potentials, which
has been proved in all cell types studied so far
(Canals et al., 2005; Stuart and Sakmann, 1995).
This finding has shaken vigorously the theoretical
basis upon which neuronal integration is based.
According to the classic doctrine of passive den-
dritic cables, the synaptic currents propagate along
dendrites with a constant decay produced by the
filtering properties of the membrane capacitance
and resistance. Far from this, synaptic potentials
are constantly modified in their propagation to
the trigger zone/s. The most common role at-
tributed to intrinsic currents is the amplification
of excitatory postsynaptic potentials generated
in remote sites to compensate for cable filtering.
In some cells (but not others), this mechanism
serves to equalize the impact of synapses located
at different distances from the axon (Magee and
Cook, 2000). This way, the traditional view that
relegated distal inputs to a mere modulatory role of
the proximal ones is no longer accepted. Therefore,
the site of an input is not necessarily related to a
genetically determined weight in deciding out-
put. In fact, a delicate balance between opposing
intrinsic currents determines the varying shape
of synaptic potentials on their spread along the
dendritic cable. It is also possible that the same
synaptic input causes a totally different voltage
envelope at the soma/axon trigger under different
physiological situations due to the interference of
multiple third parties. Though the actors (intrinsic
currents) are relatively new, the concept is not.
Let's bring to mind that inhibitory synapses trade
with excitatory inputs in a similar way.
From the point of view of synaptic integration
in dendrites one might think that little has changed
with the discovery of dendritic channels since their
currents should produce fixed alterations of the
synaptic inputs. The intrinsic dendritic currents
would only be translating the incoming binary
messages into an internal graded code. However,
that being the case, the non-linear properties of
intrinsic currents would already multiply the
computational possibilities of integration: among
others, boosting, thresholding, coincidence detec-
tion, frequency-dependent filtering are functions
Search WWH ::




Custom Search