In many of the model systems used to study rhythmical movements, the behavior to be explained (locomotion, respiration, mastication) is generated by the same organism being studied in the slice or dish. Studies of the stomatogastric ganglion seek explanations for behavior observed in crustaceans. For the oculomotor system, cellular and molecular data usually come from slices of the mouse or rat brain, whereas data concerning the movements to be explained usually were collected from nonhuman primates. This discrepancy will be eliminated as we acquire more information about the properties of eye movements generated by rodents (van Alphen et al. 2001; Sakatani and Isa 2004). The link between behavior and the properties of cells and circuits will be more direct if both types of data are obtained from the same animal. However, in the case of eye movements, the strategy of selecting the experimental preparation that permits the use of the most powerful analytical tools available yields different choices: The mouse will be selected for studies of cellular and molecular mechanisms, the monkey for behavioral studies (rodents are nocturnal and do not have a fovea; thus, the level of cognitive control over eye movements may be very different in rodents and monkeys). Cellular and molecular mechanisms of neuronal and circuit behaviors will develop quickly in mouse studies, but experiments designed to test the generality of these findings and putative mechanisms in animals with other visual demands on the oculomotor system will be highly valued.
An additional limitation of current isolated preparations is the inability to study the interactions of visual, auditory, and somatosenory stimuli. Because sensory input is not preserved, cellular and molecular mechanisms involved in converting sensory signals organized in different coordinate frames into the same reference frame, so that they can share a common motor circuitry, cannot be examined. Mechanisms involved in the dynamic mapping of auditory receptive fields (Jay and Sparks 1987) are also unlikely to emerge from data obtained in slice.
Was this article helpful?