Scientists have used rodent models of learning to investigate the brain circuitry and cellular mechanisms underlying memory. In rats and mice, spatial memory depends upon activity in the hippocampus. Spatial memory is the ability to use sensory cues to navigate and find a remembered location.
The Morris Water Maze is a commonly used test for spatial learning. The rat or mouse must swim in a circular pool that is filled with an opaque liquid. The reward is a platform hidden just below the surface, which, when located, allows the animal to climb out of the water. The animals use visual cues in the surrounding environment to remember where the platform is located. Normally the animal learns the location of the platform and swims directly to it after several trials. This kind of learning is impaired in animals with lesions to the hippocampus Furthermore, spatial learning has been linked to a particular kind of synaptic plasticity seen in the hippocampus (see below).
See Figure 25.4 in Neuroscience by
Purves et al. (2nd edition, 2002) for a depiction of the Morris
[For off-campus access, copy this URL: http://www.ncbi.nlm.nih.gov/books/NBK10878/figure/A1711/?report=objectonly
Next open LINK TO PROXY SERVER, paste in the URL, and click "Log in".]
In order for learning to occur, there must be stable, long-lasting changes that occur in networks of neurons to support that learning. The term plasticity is used to describe the ability for the brain to change. Synaptic plasticity refers specifically to changes that occur at synapses.
One particular type of synaptic plasticity that readily occurs in the hippocampus is called long-term potentiation (LTP). LTP is a type of enhanced synaptic activity that occurs in response to a particular pattern of stimulation. It is possible to study LTP in slices of hippocampal tissue so that scientists can stimulate and record from individual cells, and also observe the effects of drugs.
The figure illustrates what occurs in LTP. Two neurons, A and B, are connected by an excitatory synapse. When cell A is stimulated to fire an action potential, it elicits an EPSP in cell B, as shown in the recording at the bottom left.
If cell A is activated to fire a high frequency train of action potentials (multiple action potentials over the course of several seconds), the synapse between two cells is “potentiated”, which means that there is an increase in the size of EPSP in the postsynaptic cell. Depending on how it is induced, LTP lasts for hours or even days.
There are several ways to induce LTP. Another way LTP is induced is when neuron A is active while another neuron strongly excites neuron B. In each case, what is required is that the synapse between neuron A and neuron B is active when neuron B is strongly depolarized.
As you may have guessed, LTP involves the activity of NMDA receptors. NMDA receptors are ligand-gated ion channels that require two things to happen in order for channel opening: glutamate (the ligand) must bind to the receptor, and strong depolarization must occur in order to dislodge the Mg++ ion plugging the channel. When NMDA channels open, they allow Ca++ ions to enter the cell.
Ca++, an important 2nd messenger, acts via several mechanisms to lead to the increased EPSP seen with LTP. One rapid effect is an increase in the number of AMPA receptors in the postsynaptic membrane. AMPA receptors are ligand-gated ion channels that open in response to glutamate to cause a fast EPSP (see the table below).
Increased Ca++ in the postsynaptic cell also leads to other long lasting changes that increase the size of the EPSP. The strengthening of a synapse between two cells increases the likelihood that they will be active together. In this way, neurons become linked in networks that support learning.
Another intriguing finding is that the hippocampus is a site of adult neurogenesis. Neurogenesis means the formation of new neurons; neurogenesis mainly occurs during development. However, it is now well documented that new neurons are born and added to a particular region of the hippocampus throughout life, in humans as well as rats and mice. So far, there is conflicting data relating neurogenesis to learning and there is no clear model regarding what role new neurons might play in learning or other forms of plasticity in the brain.