Single Unit and Population Analysis of Saccade-related Fastigial Nucleus Activity in the Alert Monkey

A remarkable role in saccade control is played by the cerebellar
caudal fastigial nucleus, FN (fastigial oculomotor region, FOR),
which lies under the cerebellar vermis. FOR bursting activity is
related to many saccade properties, such as direction, amplitude
and duration. The precise mechanism by which FOR bursting
influences a saccade is under debate. It has been assumed on the
basis of microelectrode recordings that saccade-related FOR
neurones act without receiving information about the eye position.
Similarly, discharge activity of P-cells in the oculomotor vermis
does not reveal a prominent eye position-dependency. On the other
hand, it is proposed that eye position supports saccadic control
via at least some FOR neurones, as some authors found that a
minority of FOR neurones do react to a change in horizontal eye
position. Likewise, studies on saccade metrics in vermal oculomotor
areas indicate that the effects of lesions or of electrical
stimulations vary with the starting position of the eye. This
suggests that these areas are involved in the neuronal compensation
for non-linearities in orbital mechanics. In any case, it is
generally agreed that there is a consistent difference in the
timing of saccade-related bursts produced by FOR neurones,
resulting in shorter latencies for contralateral than for
ipsilateral saccades. This pattern of burst timing indicates that
the bursts are associated with the beginning of contralateral
saccades and with the end of ipsilateral ones. We have investigated
(1) whether or not the initial eye position has an effect on bursts
for ipsilateral and contralateral saccades, (2) if there is a
difference between centripetal and centrifugal saccades, and (3) if
FOR neurones influence acceleration and deceleration of a saccade
and, if so, in what way. For this purpose, a 9-point horizontal
starting position training paradigm was applied, using a 3x3 square
grid spaced at 16° intervals. Based on observations on 75
saccade-related FOR neurones, our results permit the following
conclusions. (1) Ipsilateral saccades clearly differ from
contralateral ones. Individual neurone analysis as well as
population burst analysis based on averaging the data across
neurones, show that neither the vertical nor the horizontal
component of the initial eye position substantially influences
saccade-related bursting of the FOR neurone population. (2)
Centripetal and centrifugal saccades do not differ as to FOR
bursting. (3) Although most FOR neurones do not receive information
about eye position, they seem to play a crucial role in
acceleration and/or deceleration of saccades. Finally, as the
discharge patterns of individual neurones are highly variable, with
prominent differences in both latency and amplitude, we propose
that modification of the relative contributions of individual FOR
neurones to the total FOR neurone population response may provide a
mechanism for the adaptive control of saccade metrics in the Rhesus
monkey.