Lecture 16 Model Systems
Human epilepsy patients
Case studies of human patients with extreme forms of epilepsy that have been treated with lesions or removal of parts of the brain have been very important in defining the neuroanatomy of human memory. Perhaps the most well known of these patients is H.M. His epilepsy was so disruptive that surgery to remove both his medial temporal lobes, the locus of his seizures, was required. The results of this surgery produced specific disruptions in his memory. He retained normal short term memory and perfect long term memory for events that ocurred prior to the surgery, but experienced anterograde amnesia, unable to retain memory for events that ocurred after the surgery. Interestingly, though, he retained his implicit memory, and could improve his skill in motor tasks, even if he did not remember practicing the tasks in the past.
Another famous epilepsy patient was R.B., who underwent a bilateral hippocampal lesion and consequently also experienced anterograde amnesia. Patients like R.B. and H.M. have allowed investigators to establish the medial temporal lobes, which include the hipposcampus, are necessary in the processes of encoding memory.
Xenopus tadpole retinal ganglion cell – tectum synapses
Zhang et al 1998 used the developing tadpole RGC-tectum synapses as a model system to study spike timing dependent synaptic plasticity (see Lecture 15, Slides 12-15). The Xenopus tadpole is a good model system for manipulating synaptic plasticity because it has fewer total neurons, neurons that are easier to patch clamp, and because synapses have not yet been pruned in the young brain.
Zhang et al 1998 were thus able to manipulate synapse strength by inserting two electrodes into two individual RGC neurons that were both connected to a single postsynaptic cell in the tectum, which was patch clamped in order to measure membrane potential following temporal manipulation of electrical firing from the two presynaptic RGCs. The results of the paper resulted in a modification of Hebb’s rule of synaptic plasticity: both the initial synaptic strength and the temporal order of activation of presynaptic neurons are critical for synaptic interactions among convergent synaptic inputs during activity-dependent refinement of developing neural networks (Zhang et al 1998). Specifically, they found that firing right before a postsynaptic spike was rewarded with increased robustness of synaptic strenth, while arriving immediately after spikes was punished and arriving too far or too late had little effect in modifying synaptic strength.
Mouse barrel cortex
The barrel cortex is the part of the somatosensory cortex of rodents where sensory inputs from the whiskers in the contralateral side of the body are represented. Inputs carrying information from a given whisker terminate in discrete areas of layer IV forming anatomically distinguishable areas called barrels.
Drosophila mushroom bodies
Mushroom bodies are known to be involved in learning and memory, particularly for smell. They are largest in the Hymenoptera, which are known to have particularly elaborate olfactory control over behaviour. In fruit flies, studies suggest that mushroom bodies have other learning and memory functions, like associative memory, sensory filtering, motor control, and place memory. The mushroom bodies are currently the subject of intense research. Because they are small compared to the brain structures of vertebrates, and yet many arthropods are capable of quite complex learning, it is hoped that investigations of the mushroom bodies will allow a clear view of the neurophysiology of animal cognition. The most recent research is also beginning to reveal the genetic control of processes within the mushroom bodies.
Aplysia, a type of sea slug, is a model organism for the study of learning and memory because of its simple nervous system and large neurons. It is relatively easy to perform electrophysiological recordings and molecular analyses, given the large amount of mRNAs found in a single Aplysia cell. Kandel et al. examined this animal and found that its learning behaviors are analogous to those of higher animals, and it can also form long-term memories. Kandel et al.'s most useful study involved studying habituation and non-associative learning in the Aplysia. The animal performs a simple reflex (retraction) when its siphon is touched that exhibits three main features of non-associative learning in vertebrates:
1. Habituation: the retraction response of its siphon and gills progressively decreases in intensity after repeated stimulus.
2. Dishabituation: the partial or complete restoration of an already habituated response following the presentation of another stimulus.
3. Sensitization: an enhanced reflex response when a stimulus is paired with a novel, strong and noxious stimulus (a tail shock).
The Aplysia displayed such non-associative learning for time intervals between minutes and hours. With similar stimulus setups, the Aplysia also exhibited classical conditioning behaviors. The molecular mechanisms of this memory behavior was studied in the Aplysia's motor neurons and dopaminergic 5HT neurons.
Experiments coupling food/reward (sugar water) to certain colors show that honeybees can be trained to recognize color through classical conditioning. Initially, the sugar water is always placed with a certain color, so the honeybees go to the sugar water. But later, when the bees detect just the color, they will move toward it (expecting reward/sugar water). At a more subtle level, honeybees can be trained to even distinguish among variations of patterns. This suggests that honeybees may be capable of processing more abstract data/concepts, for instance sameness versus difference. When honeybees show recognition of the "correct" pattern, they in essence are recognizing the same pattern that previously was associated with food/reward. In these experiments, the visual patterning that is associated with reward can be manipulated to be more and more specific/subtle, for instance beginning with a solid black color, then going to black stripes, then testing for horizontal versus vertical stripe orientation, and so forth.
Lecture 16 Techniques
The mirror trace task is used as a test of implicit memory in the form of skill learning. The task involves tracing the outline of a figure while viewing only the mirror image of one's progress (the subject's hands and the actual figure are covered from view). Normal subjects and amnesics with damage to the medial temporal lobe (such as patient H.M.) show improved speed and accuracy with practice on the mirror trace task, demonstrating that the medial temporal lobe is not required for this type of skill learning. Diseases of the basal ganglia, such as Huntington's disease and Parkinson's disease, cause impairment on the mirror trace task and other tests of skill learning.
This technique helps to “jog” the memory by activating certain associations in the brain just prior to carrying out a task or action. These associations can work both unconsciously as well as consciously. In either case, they are thought to activate clusters of neurons, which makes it more likely that the memory will come into consciousness. Though HM had poor long term memory after surgery, he still could perform
well in priming tasks. For example, if presented with a list of words, and asked to memorize them, he would not be able to remember them later. However, if you presented him later with the first three letters of each word he was supposed to remember, he was able to complete the words. He could recall ABSENT if presented with (ABS), INCOME when presented with (INC), etcetera.
Two-photon excitation microscopy is a technique that allows imaging living tissue up to a depth of one millimeter. The concept of two-photon excitation is based on the idea that two photons of low energy can excite a fluorophore in a quantum event, resulting in the emission of fluorescence. The probability of the simultaneous absorption of two photons is extremely low. Therefore a high flux of excitation photons is typically required.
Morris Water Maze
Morris Water Maze is a behavioral procedure developed by Richard Morris in 1984 to test memory function. Specifically, the maze tests hippocampal learning and spatial memory formation. In one paradigm of the procedure, a mouse is placed into a small pool of water that contains a black and visible platform. While mice can swim, they prefer not to and thus are always motivated to get onto the platform and rest. Since the platform in this case is visible, the mouse can use it as a beacon and swim to it easily. After several training sessions the platform is removed but the mouse still swims towards it. This kind of learning does not involve the hippocampus but rather involves motor and visual capabilities.
In a second paradigm, the mouse is placed in a pool of water that contains a clear and thus invisible platform. Several visual cues, such as different colored shapes, are placed around the pool in plain sight to the mouse. When the platform is removed, the mouse is able to remember the position of the platform based on the external visual cues and swims to the usual position where the old platform was placed. This kind of learning does require the hippocampus and spatial memory formation.
Using this model, scientists can explore the mechanisms of memory formation. Lesions of the hippocampus resulted in decreased maze performance. In addition, both NMDA receptor KO mice and those treated with APV, a NMDA receptor antagonist, showed reduced maze proficiency. Based on observed results, scientists were able to eludicate the importance of the hippocampus and NMDA receptors in memory formation.
To determine what occurs in the hippocampus during learning and memory requires recording from many cells. Multielectrode arrays implanted in the hippocampus allowed for electric recording, while mice were likewise video recorded. Different neurons fire when the mouse is at a specific location within the cage, indicating that each cell encodes laocation (in the lecture, each color indicated a different cell). In new environments, a new place map can form within minutes and this can be stable for a period of weeks. Wild-type mice exhibit a stable and precise cell map. On the other hand, however, mice with a constitutive form of CaMKII have an unstable space map, possibly suggesting an inability to consolidate memory without proper CaMKII function. This type of memory defect is reminiscent of patient H.M.
Classical conditioning is an associative learning process where an animal learns to associate a conditioned stimulus with an unconditioned stimulus. The unconditioned stimulus (US) is something the animal naturally has a reaction to, such as food, or a shock. The conditioned stimulus (CS) is an innocuous or neutral stimulus that doesn't trigger any reaction on its own. An example of
classical conditioning is the Pavlovian dog, that associates the bell (CS) with food (US), or the tone-shock training used on mice. Just as the dog learns to salivate when it hears the bell, the mice display a startle response when the tone is played, in anticipation of a noxious stimulus. The CS-US coupling has to be fairly tight, since if they are too temporally distant, the animal will not learn to associate the two. Also, the CS has to precede the US in order to trigger the association.