Abstract
purpose. To investigate the role of extraocular muscle afferent signals in the
control of saccadic eye movements.
methods. A suction scleral contact lens was used to impede the movements of the
right eye while subjects executed visually guided saccades to briefly
presented targets. Movements of the left eye were measured using
infrared oculography. Saccade amplitude, peak velocity, and duration
were analyzed trial by trial and compared before, during, and after the
right eye was impeded.
results. When the right eye was impeded, the amplitudes of saccades executed by
the left eye were reduced. There was no alteration in the main sequence
relationships. The amplitude effect had a rapid onset and offset. There
was no evidence that the effects built up over a number of trials, nor
was there evidence that individual saccades were modified on-line.
conclusions. These results are consistent with the hypothesis that extraocular
muscle afferent signals provide a feedback signal of the movements of
the eyes that is used to produce rapid adjustments of oculomotor output
when required.
Precise control of eye movement is important, not just for its
own sake but also for effective vision and interaction with objects in
the visual world. It might be thought that all possible sources of
information would be used to monitor and control the position or
movements of the eyes. However, although the muscles that move the eyes
in the orbits, the extraocular muscles (EOMs), are well endowed with
intramuscular receptors, it is generally believed that afferent signals
from these receptors are not involved in the control of eye movement at
least in the short term.
1 2
Single-unit recording studies in various species have demonstrated that
afferent signals arising from the EOMs reach a wide range of oculomotor
and visual structures in the central nervous system, including
brainstem oculomotor centers (vestibular nuclei, nucleus praepositius
hypoglossi, and oculomotor nuclei
3 4 5 6 ), the
cerebellum,
6 the lateral geniculate
nucleus,
7 8 midbrain oculomotor and visual processing
centers (the superior colliculus in the cat
9 and optic
tectum in the pigeon
10 11 ), and the visual
cortex.
12 13 Furthermore, EOM afferent signals carry
information related to the parameters of eye movement rather than
simply a gross signal indicating the occurrence of eye movement to many
of these centers
4 11 and there modify the processing of
information (for example, visual or vestibular information) in a
functionally specific manner.
11 14 15 Evidence from human
studies also indicates that EOM afferent signals are involved in longer
term adaptive control of eye movement.
16 17 Recent human
studies using single EOM vibration to induce afferent signals have
shown that EOM afferent signals from one eye modify the position of the
other eye
18 and influence the programming of memory-guided
saccades.
19 Although in the monkey the removal of EOM
afferent signals does not disable accurate saccades,
20 this does not rule out the possibility that, when available, EOM
afferent signals contribute to oculomotor control.
One reason a role for a peripheral feedback signal is usually
discounted in the oculomotor system is that it is argued that the EOMs
operate under conditions of fixed load. Thus, a given efferent signal
always has a reliable and predictable effect on the position of the
eyes. We have altered this condition by impeding the movement of one
eye to cause an acute increase in EOM load. We measured the movement of
the other eye during a visually guided saccade task to assess the
response of the oculomotor system to this perturbation.
All procedures conformed to the Declaration of Helsinki for
research involving human subjects and had local ethics board approval.
All subjects gave informed consent. Three adult male subjects with
normal or corrected-to-normal visual acuity and normal eye movements
were tested (PK, 36 years of age, one of the authors; RH, 31; and KB,
27). A suction scleral contact lens
21 was used to impede
the movement of the right eye while subjects performed a visually
guided saccade task with the left eye. The lens had a short stalk that
was placed in an adjustable holder clamped to the experimental table.
Attachment of the lens to the eye by light suction did not cause the
eye to move posteriorly but simply impeded its rotation. Thus, with the
lens in place, when subjects were asked to make voluntary horizontal
saccades, we observed that the right eye continued to move under the
lens, but not by as far as the (free-to-move) left eye. The movements
of the left eye were measured using an infrared corneal reflection
device (Iris; Skalar Medical, Delft, The Netherlands). Eye position
signals were digitized at 1 kHz with 12-bit precision using an
intelligent interface (model μ1401; Cambridge Electronic Design,
Cambridge, UK). The eye position and a time marker of the appearance of
the visual target were displayed on the computer screen; data from 100
msec before to 500 msec after the appearance of the target were stored
on disc for analysis off-line.
Saccade targets, generated by a visual stimulus generator (Cambridge
Research Systems, Rochester, UK), were presented on a monitor, which
subjects viewed from 57 cm with the left eye. Head movement was
prevented by a chin rest and cheek pads. A fixation target appeared in
the center of the screen for a random period of 0.5 to 1.5 sec. This
was extinguished and replaced by a saccade target (0.3° black square
on a light background) which was displayed for 200 msec and appeared
randomly at one of four locations: 5° or 10° to either the left or
right of fixation. Targets were presented in three runs of 52 trials.
In the first and third runs, the right eye was occluded, and in the
second run the right eye was impeded with the suction scleral lens.
Before lens placement, several drops of local anesthetic
(Proxymetacaine Minims; Chauvin, Romford, UK) were instilled in
the right eye. The lens was then placed in the eye and gentle suction
applied. The lens remained in place for approximately 5 minutes, and
once the experiment had been completed, intraocular pressure was
measured, and the corneal surface was examined.
Data were analyzed off-line, using an analysis program that displayed
the recorded eye position, the calculated eye velocity, and the time at
which the target appeared. For each record in which target appearance
was preceded by steady fixation, the amplitude, duration, peak velocity
and latency of the primary saccade were measured. Data from
anticipatory saccades (i.e., latency <80 msec) were not included in
the analysis. A calibration factor was calculated from the first run by
plotting the maximum gaze amplitude (i.e., primary plus subsequent
corrective saccades when these occurred) of each individual trial in
digital-to-analogue converter units against the target
amplitude in degrees and using linear regression analysis to obtain the
slope of the relationship.
Subjects executed monocular saccades to the briefly presented (200
msec) targets with reasonable accuracy. When the right eye was impeded,
subjects reported no discomfort and no perceived difficulty in either
seeing the target with the left eye or executing saccades in response
to the targets.
When movements of the right eye were impeded, the mean saccade
amplitudes of the left eye were reduced in each of the three subjects
in all experimental sessions
(Fig. 1) . For saccades executed in response to targets appearing 5° and 10°
to the right of fixation, mean pooled saccade amplitude was reduced by
23% compared with the original level when the right eye was free to
move. This reduction was statistically significant (Students
t-test,
P < 0.001,
t = 8.88 for
R5,
t = 11.57 for R10). For targets appearing 5° and
10° to the left of fixation, the reductions in mean saccade amplitude
were 15% (
t = 5.39;
P < 0.001) and
17% (
t = 6.33;
P < 0.001),
respectively. After the lens had been removed, the mean saccadic
amplitudes of the left eye increased toward the normal (prelens)
control values
(Fig. 1B) . In response to targets appearing 5° and
10° to the right, mean saccade amplitudes increased by 22%
(
t = 6.86;
P < 0.001) and 25%
(
t = 8.55;
P < 0.001), respectively.
In response to targets appearing 5° and 10° to the left, mean
saccade amplitudes increased by 7% (
t = 2.26;
P < 0.05) and 11% (
t = 3.52;
P < 0.001), respectively.
We noted that the left eye saccade amplitudes after the suction contact
lens had been removed sometimes remained slightly lower than the
original amplitudes recorded before the suction contact lens was
inserted
(Fig. 1C) . For the pooled data rightward saccade amplitudes
for 5° and 10° remained 6% (
t = 2.01;
P < 0.05) and 4% (
t = 1.88; not
significant) lower than the prelens control values. Leftward saccade
amplitude for 5° and 10° were 9% (
t = 3.26;
P < 0.05) and 8% (
t = 3.54;
P < 0.001) lower, respectively.
Recording began within approximately 90 seconds of lens insertion in
most runs, and on one occasion within less than 60 seconds.
Figure 2 shows trial-by-trial mean (±SEM) amplitudes (data pooled across
subjects and sessions). The pooled data are very similar to the
individual data. Note that saccade amplitude was reduced in the first
trial. Linear regressions of amplitude on trial number for both the
pooled and individual data showed either no significant deviation in
the slope from zero or no significant difference in the slope of the
line between the eye-free and eye-impeded conditions. Thus, there was
no evidence of a build up in the effect.
We found no evidence of any alterations in the amplitude–velocity
relationship. Typical data from one subject in a single experiment are
plotted in
Figure 3A . When the right eye was impeded, the peak velocity of left eye
saccades was no lower than would be predicted, given the reduction in
amplitude. There was no evidence from this analysis that saccade
duration was modified inappropriately
(Fig. 3B) . Thus, although the
amplitudes were reduced when the right was impeded, the velocity and
duration scaled by an appropriate amount. There was little evidence
that impeding the movement of the right eye affected the latencies of
left eye saccades.
We examined the velocity profiles of saccades with and without the
right eye impeded. All profiles were aligned using the latency
measurements; for two experiments in two subjects we calculated mean
profiles
(Fig. 4) . Peak velocity when the right eye was impeded was lower, as expected.
The duration of these mean profiles was only slightly reduced. There
was little evidence of the profiles’ being distorted in any way. The
impeded profile diverged from the free profile at or near the beginning
of the saccade. Examination of velocity profiles trial by trial
confirmed that from the first trial, there was a large reduction in
peak velocity, with little evidence of further clear reductions.
In two subjects, a control experiment was run in which the procedures
were identical with those described, including the instillation of
local anesthetic into the right eye. However, the lens was not placed
in the right eye, it was occluded. In these runs we observed no
consistent alteration in any of the saccade parameters. In particular,
there was no alteration in saccade amplitude.
The assumption that the EOMs operate in conditions of unchanging
load, coupled with the absence of a monosynaptic stretch reflex in the
macaque monkey
22 and the ability of monkeys to make
accurate saccades when the EOMs have been deafferentated
20 have been taken to justify the view that EOM afferent signals play
little or no role, at least in the short term, in the control of eye
movement. By impeding one eye, we have acutely altered the load under
which the EOM operate. We do not know, of course, the extent to which
load was increased. Observation demonstrated that the impeded eye did
not move as far as the unimpeded eye, although the extent of this was
variable.
We have shown that in these circumstances the oculomotor system made
rapid adjustments. From the first trial in which the right eye was
impeded, that is within a maximum of a few tens of seconds of lens
placement, saccade amplitude in the other eye was reduced. Note that
this response is quite different from other types of adaptive response
observed in the oculomotor system. These involve internal comparison of
retinal information indicating a difference between desired and actual
eye position
23 or retinal slippage
information.
24 In our experiments there was no retinal
error. The nonseeing eye was impeded, and the brief target presentation
time ensured that when, in the impeded condition, the seeing eye landed
short of the target position, no retinal error was generated.
The clearest evidence of the oculomotor response to impeding one eye
was the effect on saccade amplitude. There was no evidence of this
built up over even a short period of a few seconds, or a small number
of trials, although there was some evidence that when the lens was
removed some residual amplitude reduction remained. Most examples of
adaptation of oculomotor parameters build up over a larger number of
trials, or over a period during which adapting stimuli are presented.
There was, however, little evidence that each saccade was modified
on-line. Had this been the case, we would have expected the velocity
profiles of saccades in the impeded condition to diverge from the
control profiles some short period after the beginning of the saccade.
As
Figure 4 shows, there was little evidence of this.
Finding a reduction in saccade amplitude was unexpected. However, this
could be interpreted as evidence that the saccade system seeks, at
least in the circumstances used in these experiments, to preserve
conjugacy. Thus, because the right eye is not moving as far as
intended, the drive to the left eye is reduced. It remains to be seen
whether the controller responds in this manner in different
circumstances or when different types of eye movement (e.g., smooth
pursuit) are manipulated. Although saccade amplitude was clearly
modified, the amplitude–velocity relationship was apparently
unaffected. The peak velocity of saccades in the impeded condition was
reduced to the extent that might be predicted from the amplitude
reduction. Although we found no statistically significant alteration of
the amplitude–duration relationship, the examination of the velocity
profiles suggested that the duration of saccades in the impeded
condition was not as short as might be predicted.
We have shown that there is a feedback signal, which in the absence of
a retinal error signal induces alterations in visually guided saccades.
In our experiments there was always a period between the placing of the
lens and the beginning of the experimental run. However, this was kept
a short as possible and was usually no longer than 90 seconds. During
this time, little specific visual information was available to aid any
adaptive process. Any saccades executed were not responses to specific
saccade targets but were voluntary saccades made on request to check
the lens position. In the experimental run the amplitude effect was
always present from the very first trial and did not subsequently build
up (see
Fig. 2 ). It seems highly unlikely that an adaptive process
begun during the 90 second prerun period would be completed by the end
of the prerun period. Rather, the results are suggestive of the
operation of an afferent signal that indicates that the eye is impeded
and induces rapid modifications of the oculomotor system.
The impeded eye was anesthetized, and although this does not rule out
entirely the possibility of a mechanoreceptive source for these
signals, it seems unlikely. The more likely candidate source for the
effects we have observed is EOM intramuscular proprioceptors. The human
EOMs are well known to have relatively high numbers of muscle spindles
and also palisade endings, which may be unique to the
EOM.
2 Single-unit recording studies in various animal
species have shown that afferent signals arising from EOM intramuscular
receptors are able to modify the processing of information in the
brainstem “on-line”—that is, as soon as the afferent signals are
induced, information processing is modified. It does not build up over
a number of trials or cycles of stimulation.
3 5 14 However, if a feedback signal were acting on the brainstem gaze centers
directly, we would have anticipated alterations in the some of the main
sequence parameters or their relationships. Furthermore, if afferent
signals were being distributed separately and directly to subareas of
the horizontal gaze center (e.g., to the burst-generating circuitry in
the paramedian pontine reticular formation and the integrator circuitry
in the nucleus praepositus hypoglossi), we might have observed
mismatches between the saccade pulse and the saccade step. There was no
evidence of any of these. We speculate therefore that EOM afferent
signals act at a higher level in the saccade control circuitry to exert
their effects. Two candidate sites would be the cerebellum or superior
colliculus, both of which are known to receive EOM afferent
signals.
6 9 11
An implication of these results is that alteration of, or damage to,
EOM intramuscular receptors or their afferent pathway in one eye
alters, in the course of a few minutes, the movements of the other eye.
Presumably in the presence of the resultant retinal error signals,
recalibration quickly follows. Thus, in experimental circumstances in
which the available visual information is not carefully controlled, it
could be concluded that removal of afferent feedback has no effect,
whereas what is observed is the result of an effective recalibration of
the oculomotor system.
Our results also suggest further experiments that would test the
hypothesis that EOM afferent signals are involved in oculomotor
control. Surgery involving the tendinous insertion of the EOM disrupts
the intramuscular receptors found in this region of the
muscles.
25 If our experiments were repeated in subjects
with altered EOM proprioception—for example, in patients in whom the
horizontal recti in one eye had been operated on to realign the eyes,
we would predict that when the surgically treated eye is impeded, a
different pattern of effect to that reported herein would be observed.
Specifically, when the surgically treated eye is impeded, we would
expect that the saccade amplitude reduction effect in the nontreated
eye would be absent or reduced. We further predict that when the
impeded eye is the nontreated eye, the saccade amplitude effect would
be similar to that reported herein.
Supported by a grant from the Chief Scientist Office, Scotland.
Submitted for publication December 17, 1999; revised March 1, 2000; accepted March 15, 2000.
Commercial relationships policy: N.
Corresponding author: Paul C. Knox, Division of Orthoptics, University of Liverpool, Thompson Yates Building, Brownlow Hill, Liverpool L69 3GB, UK.
[email protected]
Carpenter RHS. Movements of the Eyes. 1988; 2nd ed. Pion London.
Ruskell GL. Extraocular muscle proprioceptors and proprioception. Prog Retinal Eye Res
. 1999;18:269–291.
[CrossRef] Donaldson IML, Knox PC. Directionally-specific effects of afferent signals from the extraocular muscles upon responses in the pigeon brainstem to horizontal vestibular stimulation. Neuroscience
. 1990;38:145–161.
[CrossRef] [PubMed]Donaldson IML, Knox PC. Afferent signals from pigeon extraocular muscles modify the vestibular responses of units in the abducens nucleus. Proc R Soc Lond B Biol Sci
. 1991;244:233–239.
[CrossRef] Ashton JA, Boddy A, Dean SR, Milleret C, Donaldson IML. Afferent signals from cat extraocular muscles in the medial vestibular nucleus, the nucleus praepositus hypoglossi and adjacent brainstem structures. Neuroscience
. 1988;26:131–145.
[CrossRef] [PubMed]Ashton JA, Milleret C, Donaldson IML. Effects of afferent signals from the extraocular muscles upon units in the cerebellum, vestibular nuclear complex and oculomotor nucleus of the trout. Neuroscience
. 1989;31:529–541.
[CrossRef] [PubMed]Donaldson IML, Dixon RA. Excitation of units in the lateral geniculate and contiguous nuclei of the cat by stretch of extrinsic ocular muscles. Exp Brain Res
. 1980;38:245–255.
[PubMed]Lal R, Friedlander MJ. Effect of passive eye position changes on retinogeniculate transmission in the cat. J Neurophysiol
. 1990;63:502–522.
[PubMed]Donaldson I, Long A. Interactions between extraocular proprioceptive and visual signals in the superior colliculus of the cat. J Physiol
. 1980;298:85–110.
[CrossRef] [PubMed]Knox PC, Donaldson IML. The effect of afferent signals from extraocular muscles on visual responses of cells on the optic tectum of the pigeon. Proc R Soc Lond B Biol Sci
. 1995;259:285–291.
[CrossRef] Knox PC, Whalley HC. The directional effects of passive eye movement on the directional visual responses of single units in the pigeon optic tectum. Exp Brain Res
. 1997;116:510–518.
[CrossRef] [PubMed]Ashton JA, Boddy A, Donaldson IML. Interactions, in cat area 17, between visual and orbital proprioceptive signals. J Physiol (Lond)
. 1984;348:19P.
[CrossRef] Enomoto H, Matsumura M, Tsutsui J. Projection of extraocular afferents to the visual cortex in the cat. Neuro-ophthalmology
. 1983;3:49–57.
[CrossRef] Donaldson IML, Knox PC. Evidence for corrective effects of afferent signals from the extraocular muscles on single units in the pigeon vestibulo-oculomotor system. Exp Brain Res
. 1993;95:240–250.
[PubMed]Knox PC, Donaldson IML. Extraocular muscle afferent signals and the vestibulo-ocular reflex. Delgado–Garcia JM Godaux E Vidal PP eds. Information Processing Underlying Gaze Control. 1994;309–318. Pergamon Oxford, UK.
Gauthier GM, Vercher JL, Zee DS. Changes in ocular alignment and pointing accuracy after sustained passive rotation of one eye. Vision Res
. 1994;34:2613–2627.
[CrossRef] [PubMed]van Donkelaar P, Gauthier GM, Blouin J, Vercher J-L. The role of ocular muscle proprioception during modifications in smooth pursuit output. Vision Res
. 1997;37:769–774.
[CrossRef] [PubMed]Lennerstrand G, Tian S, Han Y. Effects of eye muscle proprioceptive activation on eye position in normal and exotropic subjects. Graefes Arch Clin Exp Ophthalmol
. 1997;235:63–69.
[CrossRef] [PubMed]Allin F, Velay J-L, Bouquerel A. Shift in saccadic direction induced in humans by proprioceptive manipulation: a comparison between memory-guided and visually guided saccades. Exp Brain Res
. 1996;110:473–481.
[PubMed]Guthrie BL, Porter JD, Sparks DL. Corollary discharge provides accurate eye position information to the oculomotor system. Science
. 1983;221:1193–1195.
[CrossRef] [PubMed]Al-Hinnawi MF, Barbenel J, Al-Qurainy IA, Dutton GN. Quantitative mechanical assessment of the extraocular muscles. Proc Inst Mech Eng [H]
. 1991;205:101–108.
[PubMed]Keller EL, Robinson DA. Absence of a stretch reflex in extraocular muscles of the monkey. J Neurophysiol
. 1971;34:908–919.
[PubMed]Deubel H, Wolf W, Hauske G. Corrective saccades: effect of shifting the saccadic goal. Vision Res
. 1982;22:353–364.
[CrossRef] [PubMed]Miles FA, Eighmy BB. Longterm adaptive changes in primate vestibulo-ocular reflex, I: behavioral observations. J Neurophysiol
. 1980;43:1406–1425.
[PubMed]Steinbach MJ, Kirshner EL, Arstikaitis MJ. Recession vs marginal myotomy surgery for strabismus: effects on spatial localization. Invest Ophthalmol Vis Sci
. 1987;28:1870–1872.
[PubMed]