Biology Reference
In-Depth Information
motor information to the Purkinje cells of the cerebellar cortex. It is believed that implicit
motor learning is mediated by synaptic plasticity in the cerebellar cortex and/or the deep
cerebellar nuclei (Marr, 1969; Ito & Kano, 1982; McCormick & Thompson, 1984; Attwell
et
al.
, 2002). These structures are organized as two-dimensional topographical maps of the
body, and it is possible to target specific microzones that mediate particular skeletal muscular
responses (Andersson & Oscarsson, 1978; Garwicz & Ekerot, 1994). The best-studied
example of this functional organization is probably classical conditioning of the nictitating
membrane/eyeblink response in rabbits (Mauk
et al.
, 1986; Attwell
et al.
, 2002) Delgado-
Garcia and Gruart, TINS 2006. With such regular, tractably organized and well-characterized
circuitry, there will be a higher probability of demonstrating models of plasticity in learning
paradigms in cerebellum than in the functionally-less understood hippocampal system. The
knowledge about cerebellar error-dependent plasticity could be applied for other brain
systems and particularly for hippocampal region through “vicarious trial and error” behaviors
(Muenzinger, 1938; Tolman, 1939; Hu & Amsel, 1995; Hu
et al.
, 2006), which are mediated
at least partially by hippocampal place fields (Johnson & Redish, 2007).
Figure 9. Motor control systems. A. Two-degrees-of-freedom control system. Two-degrees-of-freedom
adaptive control system for voluntary movement combines feedback control by the cerebral cortex with
feed-forward control by the cerebellum. Signal transfer characteristics of the controller (g) and of the
controlled object (G); instruction of movement (IM); command signals for movement (CM) and actual
movement performed (AM). AM becomes close to IM if g is sufficiently large if f(G) becomes
equivalent to an inverse of G. In a typical feed-forward control, the controller converts instruction for a
movement to command signals that act on the controlled object. The controlled object in turn converts
the command signals to an actual movement. If the instruction/command conversion is inversely
equivalent to the command/movement conversion by the controlled object, the actual movement
becomes equivalent to the instruction (adapted from (Ito, 2000); (Kawato
et al.
, 1987) and (Ito, 2001)).
B. Another way of performing a precise control in a seemingly feed-forward control system is to utilize
an internal loop through a model that simulates the command/movement conversion by the controlled
object and thereby predicts the movement to be produced by the controlled object. AM becomes
equivalent to IM if the signal transfer characteristics of the forward model G' = G. If the internal loop
contains not only dynamic properties of the controlled object but also the delay time involved in the
external feedback, exactly the same effect as the external feedback from the actual movement will be
reproduced. This model was applied to interpret functional meanings of the cerebrocerebellar
communication loop (adapted from (Ito, 2000) and (Ito, 2001)).
8.2. Visual cortex plasticity
Episodic memories are encoded through the hippocampus, but the experimental tools to
modulate synaptic weights at a spatially distributed set of hippocampal synapses are restricted
