Biology Reference
In-Depth Information
information storage. Exposure of an animal to a novel spatial configurations, for example
results in the expression of LTD in CA1 region (Manahan-Vaughan & Braunwell, 1999). It
has been postulated that the mechanisms underlying long-term depression in the
hippocampus, together with the mechanisms of long-term potentiation, are responsible for
information storage by the hippocampus (Martin
et al.
, 2000). In the following section we
will summarize the main models that explain the long-term changes in synaptic efficacy.
2.
M
ODELS FOR
S
YNAPTIC
P
LASTICITY
I
NDUCTION
Generally, synaptic plasticity can involve a presynaptic component as well as the
postsynaptic one. Synaptic modifications also can be divided on homosynaptic, involving
only one pre- and postsynaptic interaction, and heterosynaptic, where synaptic alterations
involve more than one presynaptic component. Regarding the temporal effect, synaptic
changes can be defined as short-lasting (seconds, minutes) and long-lasting (days, months).
The most common object of plasticity research is the homosynaptic long-term change of
synaptic strength as it is believed to be a necessary component in hippocampal network
modifications. Here we will emphasize the two main models explaining how this type of
synaptic plasticity can be induced.
2.1. Frequency-dependent plasticity
The BCM (Bienenstock, Cooper and Munro) rule states that synaptic strengths are
increased when the activity of the pre- and post-synaptic neurons exceeds a particular
threshold and weakened when activity is below it (Bienenstock
et al.
, 1982). Crucially, this
modification of threshold varies according to the mean activity of both neurons, which
prevents runaway scaling up or down of all the synapses (Bear
et al.
, 1987; Kirkwood
et al.
,
1995). In the BCM model, correlated pre- and postsynaptic activity evokes LTP when the
postsynaptic firing rate is higher than the threshold value and LTD when it is lower (Fig 1A).
To stabilize the model, the threshold is shifting as a function of the average postsynaptic
firing rate. For example, the threshold increases if the postsynaptic neuron is highly active,
making LTP more difficult and LTD easier to induce. Frequency-dependent plasticity relies
on the natural fast and very fast cortical oscillations timed on a slower theta rhythm (Larson
& Lynch, 1988, , 1989; Gray & McCormick, 1996).
Most forms of LTP are glutamatergic and the most prominent form is induced following
activation of the
N-
methyl- D-aspartate (NMDA) receptor. NMDA-dependent LTP occurs
only if there is both presynaptic firing and substantial postsynaptic depolarization sufficient to
open the NMDA channel. Postsynaptic activity can occur with a delay, the duration of which
is the time-constant of decay of the NMDA conductance (100-200ms) (Gustafsson
et al.
,
1987). Frequency-dependent models require repetitive firing for LTP induction. Scaling of
the NMDA-mediated component has implications for Hebbian plasticity, because LTP and
LTD are activated by calcium entry through NMDA receptors. It is accepted that large
amounts of calcium entry induce LTP, while smaller amounts cause LTD (Lisman, 1994). If
