Directionally-specific effects of afferent signals from the extraocular muscles upon responses in the pigeon brainstem to horizontal vestibular stimulation

0306-4522/90 S3.00 f 0.00 Pergamon Press plc C 1990 IBRO

Nuums&rlce Vol. 38. No. I. pp. 145-161.1990

Primd in Great Britain

DIRECTIONALLY-SPECIFIC EFFECTS OF AFFERENT SIGNALS FROM THE EXTRAOCULAR MUSCLES UPON RESPONSES IN THE PIGEON BRAINSTEM TO HORIZONTAL VESTIBULAR STIMULATION I.

Department

of Pha~acoiogy,

M. L. DONALDSON* and P. C. KNOX

University of Edinburgh, 1 George Square, Edinburgh EH8 9JZ, U.K.

Abstract-The responses of single units in the brainstem of the decerebrate, paralysed, pigeon were studied. Natural vestibular stimulation was provided by horizontal, sinusoidal, oscillation of the bird and extraocular muscle afferents of the ipsilateral eye were activated by passive eye-movement. Unit responses to vestibular and/or orbital stimuli were examined in sets of peristimulus time histograms interteaved in time. Of 352 units in the brainstem, in the region of the vestibular nuclei, which were exposed to the effects of both vestibular stimuli and passive eye-movement, 40 (I I%) responded only to the latter; the other 312 units (89%) responded to vestibular stimulation at 0.4 Hz (amplitude +8”). Of these 312 units, 129 (41%) were affected only by vestibular stimuli; in the other I83 units (59%) passive eye-movement produced clear modification of the vestibular responses by adding excitation or inhibition, or both. There were phasic modifications in most units; in 77 there were longer-lasting changes in the vestibular responses, often following a phasic response. In I24 units whose responses were subjected to statistical analysis, the vestibular responses of 42 (34%) were modified only by horizontal eye-movement and eight (6%) were affected only by vertical movement. A further 18% showed larger effects from horizontal than from vertical eye-movement; in 2% vertical eye-movement was preferred. Further examination of the specificity of the effects of eye-movement in planes between the vertical and horizontal was possible in 29 units which showed various degrees of “tuning” of the effect. In some units there was additional specificity for eye-movement in (a) particular directions (towards the beak rather than towards the tail, for example); (b) in particular arcs of the orbit (centre-to-temporal rather than nasal-to-centre, for example). Note that all these effects were upon the responses of the units to horizontal vestibular stimulation. Thus, the m~i~cations of the vestibular responses depended upon specific characteristics of the passive eye-movement. The exact recording sites of 29 units were determined histologically; some were in the medial vestibular nucleus but many were in the adjacent reticular formation. The principal interest of the results is that they provide more detailed information than was available previously on the specificity of the effects of afferent signals from the extraocular muscles upon the vestibular responses of units in regions of the brainstem known to be involved in oculomotor control. The decerebrate pigeon proves to be a particularly good preparation in which to study these effects. Possible actions of the afferent signal from the eye-muscles on the control of eye-movement are discussed and the results from the pigeon are compared with those from our previous work on the vestibulo-ocular system and the visual cortex. The results in this paper add further weight to the hypothesis that a proprioceptive signal from the extraocular muscles is likely to play a part in the control of eye-movement by the vestibulo-ocular system and they suggest further experiments to pursue the hypothesis.

The extrinsic ocular muscles (extraocular muscles. EOMs) of many species of vertebrate contain stretch receptors.” In a few species, including Man,” these receptors include true mu~ie-spindles, but in most animals there are only simple spirals and a specialized type of tendon organ (see Spencer and Porte? for review).

*To whom correspondence should be addressed. EOM, extraocuiar muscle; PEM, passive eye-movement; PSTH peristimulus time histogram; Sl-S3. phases l-3 of passive eye-movement stimulus; RLR3, responses to St-S3, respectively; VOR vestibulo-ocular reflex.

Ab~re~.iu~iu~s:

In 1918 Sherrington35 gave an account of his belief that EOM proprioceptors were concerned both with the control of eye-movement and with the elaboration of external space. This view fell out of favour for reasons which we have examined elsewhere,’ and, more recently, the general opinion has been that EOM proprioceptors play little or no role in oculomotor control. However, in recent years, we have shown that, in an amphibian (the giant toad, Sufo marinu.@), a bony fish (the trout’), and in the cat: afferent signals from the EOM reach the vestibular nuclei and other centres involved with the control of eye-movements. The vestibular nuclei project both monosynaptically I45

146

I.

M. L. DONALDSON

and through polysynaptic pathways to the oculomotor nuclei-the final common path for control of eye-movement. Furthermore. in the fish’ and pigeon.” we have found that the vestibular responses of cells in the oculomotor nuclei are modified by EOM afferent signals. Thus. in several species with very different oculomotor and visual behaviour, it appears that afferent signals from the EOM enter the vestibulo-ocular system and affect signal processing there. There are also strong indications that the effect of the signal is more specific than would be required simply to indicate that an eye-movement had occurred. Thus. some of the units have been found to be differentially affected by passive movement of the eye in different planes; in particular. horizontal and vertical movements of the globe affect some units differently.‘,’ Successful as our previous experiments have been, it has been difficult to amass a large enough sample of units whose vestibular responses are affected by passive eye-movement (PEM) to allow detailed analysis of these preferences for eye-movement in one plane rather than in another. In the toad and the fish. ceil-packing in the brainstem is loose and responsive units are difficult to isolate. In the cat, the effects of anaesthetics lead to uncertainty in interpretation of the results and may also contribute to a low yield of responses. Thus. we sought a species which could be used as a decerebrate preparation without compromising the afferent input from the orbit and which also had high visual acuity and a wide repertoire of eye-movements so that it would be likely to provide a good model for the higher vertebrates. Some birds fulfil these criteria. Preliminary experiments showed that the decerebrate pigeon is an excellent preparation for experiments on the effects of EOM afferent signals on the vestibulo-ocular system. The pigeon has excellent visual acuity” and makes a wide range of eye-movements.‘.” Although pigeons do not appear to have true muscle-spindles in their EOM.2R afferent projections from intramuscular receptors have been demonstrated by transport of horseradish peroxidase” from the EOM to the brainstem. The experiments reported in this paper were undertaken to establish whether the pigeon vestibular nuclei and neighbouring structures in the brainstem receive afferent input from the EOM and, if so, whether these signals affect the processing of vestibular input. We also hoped that the new preparation would allow us to collect a large enough group of responses to study the details of the differential effects of eye-movement in various planes. A preliminary account of some of the results has already been given.” EXPERIMENTAL

PROCEDURES

Preparotron Adult plgeons ether. The skull

(C‘olumha

/irk)

were anaesthetized

wtth

was opened and the cerebral hemispheres

and P. C.

KNOX

removed by aspirauon, leaving the opuc lobes. cerebellum and brainstem intact. After recover). evidenced b! the return of vestibular reflexes. the bird was sedated with

pentobarbitone sodium (Sagatal. May and Baker. 6mg. 1.m.). and the trachea cannulated. It was then placed m a stereotaxlc frame and the dorsal surface of the cerebellum cleared of blood and debris to expose folia V and Vla After the frame had been clamped to a servo-controlled vestihular turntable, muscular paralysis was induced with gallaminc (Flaxedil. May and Baker, 1-3 mg. i.m.) and intermittent positive pressure ventilation with room av was Instituted. Air was blown into the tracheal cannula and escaped through punctures made m the abdominal au sacs The heart rate was monitored

throughout

the experlmenr

Slimularion

Natural vestibular stimulation was used and EOM proprioceptive signals were induced by PEM. The vestibular stimulation consisted of sinusoidal oscillation in the honzontal plane with an amplitude of k8 at a frequency of 0.4 Hz. The head was held at an appropriate angle’ to bring the horizontal canals into the Earth-horizontal plane. The left eye was moved passively. either when the turntable was stationary or during vestibular stimulation. by an electromagnetic servo-controlled device which acted upon a stalk carried by an opaque contact-lens held in position on the cornea by suction. The device was similar in its action to that which we have described previously.’ The eye-mover was arranged so that the eye could be moved from a central resting position along any desired rachus. The amplitude and velocity of PEM were approximately 14 at 127’,~ this is within the saccadic range for the pigeon.’ The eye-movements were generated by the computer system (see below) which also selected the orbital arc along which each PEM was made. The PEM stimulus consisted of a rapid movement 01’the eye (velocity approximately 127 is) from the centre of the orbit to an eccentric position at which It was held for 2OOms. and a second rapid movement back 10 the resting position. For convenient reference these three phases of the displacement of the eye are labelled Sl. S2 and S3. respectively (see Fig. 2). The eye started and finished approximately centred in the orblt. Responses measured from the peristumulus time histograms (PSTHs) to the three elements of the stimulus were labelled RI. R2 and R3. respectively. Occasionally, responses were measured at a time after the eye had been returned to the resting posiuon: these were labelled R4. Recording and collection of daru

Glass-coated tungsten mrroelectrodes” were directed vertically downward through the anterior folia of the cerebellum between 0.5 and 1.2 mm lateral to the midline. Recordings were usually made from the left brain (ipsilateral to the eye which was moved) but. on some occasions. the electrode was inserted in the right brainstem. Extracellular records were made from single units which responded to vestibular stimulation, or to PEM. or to both stimuli. Sets of interleavedI PSTHs of IO-ms bin width were constructed using a computer system on-line (CED 1401 interface controlled by a Tandon PCA30 personal computer). This system was similar in principle. although quite different in construction. to that which we used in pt’WiOUS experiments.4.5.’ Sets of either four. or of eight interleaved PSTHs were collected. Construction of sets of four ms was a rapid method of obtaining information about the response of a unit to vestibular stimulation alone. and on its responses to three different combinations of vestibuiar and PEM stimuli. Sets of eight PSTHs contained one test of vestibular stimulation alone and seven tests of combinations of vestibular and PEM stimuli: these sets took longer 10 compile but did allow more detailed study of the “tuning” of the Interactions observed

Extraocular muscle aBerent signals in pigeon brainstem

In many experiments recording sites in the brainstem were marked by electrolytic lesions (20pA for 10s). At the end of the experiment the bird was killed with pentobarbitone sodium (30mg Sagatal) and the head removed. The skull was dissected to expose the cerebellum further and was then left for a minimum of 24 h in buifered formalin. In later experiments the brain was fixed by transcardiac carotid perfusion of buffered formatin using the method of Eden and Correia.‘6 After at least 24 h in formalin the brain was removed from the skull and placed in fresh buffered formalin. After further fixation 50qm serial, frozen, parasagittal sections were cut and stained for Nissl substance. Electrode tracks were reconstructed from projected sections and the sites of units were marked on the drawings. Structures were identified with the help of the atlas of the pigeon brain by Karten and Hodos.”

Further analysis of responses to passive eye-movement The responses to PEM alone, and the modification of vestibular responses by PEM, were examined for evidence of preference for movement of the eye along one, rather than along another orbital arc using the method described previously.4.f.7 When only four PsTHs had been collected we compared the effect of PEM in the horizontal plane with that of vertical PEM. When tuning data had been collected, with PEM in up to seven different orbital arcs, the effect of PEM along particular arcs in the orbit was examined in more detail. Briefly, the size of the responses to PEM was estimated by counting the total number of impulses in a time-window (a standard number of PSTH bins, usually seven) containing the response. For each set of histograms. whether of four or of eight PSTHs, the same window was used. The sizes of these responses were then compared in pairs as we have described previously$S using a statistical test devised by Dorrscheidt” and based on the binomial theorem. A unit was regarded as having “planar” preference if its response to PEM along one arc in the orbit differed significantly from that along at least one other arc in another plane, The criterion of statistical significance was P < 0.025. This definition corresponds to that which we previously used for “radially-selective” responses to PEM in the visual cortexI except that in the previous experiments only arcs lying in orthogonal planes were used. RESULTS

The responses of units to horizontal vestibular stimulation and PEM in 122 pigeons are described below. Units which responded to either or both of these stimuli, but which also responded to other stimuli such as stretching the eye lids (polymodal units. Ashton et uk4,‘). were not examined in detail and are not included in the results presented here. Of 352 units located in the brainstem which were tested with vestibular and PEM stimuli. 40 responded only to PEM and 312 responded to horizontal vestibular stimulation. Of these 312 units, PEM produced clear modifications of the vestibular response in 183 (.59%), while 129 (41%) responded to vestibular stimulation only. The criterion for response to vestibular stimulation was cyclic modulation of the activity of a unit during sinusoidal table oscillation, as judged by listening to di~riminat~ spike activity on a Ioudspeaker and observing the PSTH of the unit’s activity collected over a number of cycles. Of the 312 units recorded in

the

brainstem

which

responded

to

vestibular

147

s~i~~ia~ion in the horizontal plane, ah but one responded to rotation in one direction only, either towards or away from the recording side. Thus 154 units responded to ipsilateral rotation and 157 to contralateral rotation; examples are shown in Fig. 1a and b, respectively; one unit responded to both ipsi- and contralateraj rotations. The same amplitude and frequency of vestibular stimulation were used throughout these experiments, approximately If:8” at 0.4 Hz (as shown in Fig. 1). Responses to passive eye -movement alone In the first series of ex~~ments, the results of which are described below, the eye was moved through three standard arcs, two in the horizontal plane (90°, PEM towards the tail, and 270”, PEM towards the beak), and one in the vertical plane (180”, downwards). fn 40 units which responded to PEM alone, the responses were excitatory in 25, and inhibitory in eight; four units had a combination of excitatory and inhibitory effects. Most units (21) gave phasic responses during the eye-movement. In the remainder the responses had both phasic and longer-lasting components. The example shown in Fig. 3 illustrates some of these features. In response to PEM there was both phasic excitation and long-lasting suppression of activity. This unit showed a clear preference for PEM in a particular direction, towards the beak (270”). Although all three PEMs produced responses significantly different (P 6 0.005) from the control, the response to PEM towards the tail was also significantly larger than that produced by movement downward (P Q 0.025) and significantly smaller than that produced by movement towards the beak (P =$ 0.005). Thus this unit showed a preference for PEM in the horizontal plane, and within that plane, a preference for movement towards the beak. The responses of 29 of the 40 units which responded to PEM alone were examined statistically (see Experimental Procedures). Of these, nine (3 1%) showed statistically significant responses (P < 0.025 or less) only to PEM in the ho~zontal plane, and three (10%) responded only when the movement was in the vertical plane. Of the 17 units which responded to PEM in both horizontal and vertical planes, seven (24%) responded best to horizontal PEM, while only two (7%) showed a preferenee for vertical PEM. The other eight units (27%) in the group which responded to PEM in both planes showed no planar preference. Interaction between the effects of horizontat vestibular stimulation and those of passive eye-movement The effects of PEM on vestibular responses were similar to the types of responses observed to PEM alone. Of 183 units, the responses of 106 (58%) to vestibular stimulation were modified in a phasic manner; 67 (37%) showed both phasic and longlasting effects and 10 (5%) showed long-lasting

t. M. L. DONALDSON and P. C. KNOX

148 24

sweeps

Eln

width

10

Sweep

msec.

Uurat

ion

2.5

5ec

a

0

0.5

1.0

1.5

‘2.0

’ 2.5

sec.

Fig. 1. Examples of responses to horizontal, sinusoidal, vutibular stimulation 8t 0.4 Hz (amplitude -+ 8”). Upper traces show tabk position; downward dcfkction indicates movement towards the kft (ipsilatarai) side. Lower traces are P!WHs of responses of units in the bminstcm; conventions as for Fig. 3. (a) Unit fires in response to ipsWera1 mtation with a phase lead on tabk velocity of 12”.(b) Unit Rm in response to contralateral rotation with a phase kad on tabk velocity of 38”. effects only. In the 106 units which showed phasic interaction, the effect was inhibitory in 42 and excitatory in 48. The remaining 16 units had a combination of inhibitory and excitatory effects. An example of a unit which responded to contralaterai rotation and in which PEM produced a pronounced phasic inhibition of the vestibuiar response is shown in Fig. 4. PEM in each of the three directions tested produced a statistically significant modification in the vestibular response (90” and 180”. P d 0.005; 270”. P G 0.025) compared with the response to vestibuiar stimulation alone. However, the largest e&et was produced by PEM towards the tail (Fig. 4, top right); there was a significant difference between the response at 90’ (PEM towards the tail) and the responses at 180” and 270’ (P C 0.005in each case). Note also that this unit responded only to movement of the eye from the centre of the orbit to eccentric positions (Sl): the return movements (S3) had Pttk effect. The return movement, S3, at 270” and the initial movement. SI. at 90- were both directed towards the tail.

Responses of a unit which. in contrast to that shown in Fig. 4, exhibited a complex interaction with several distinct features are shown in Fig. 5. Two main effects are clear; the first eye-movement (Sl) produced a marked excitation, followed by a clear inhibition. Note that the test histograms differ in important details though the inhibition is signifkant (P Q 0.005) compared with the control in all cases. Firstly, the inhibition in response to PEM at 90” was longer-lasting than that seen in response to PEM at 270”. and at 180”. it iaated until the return PEM. There was no response to the return movement (S3) at 90” other than the removal of the inhibition of the vestibutar response; however, at 270” (and to a lesser extent at 180”) the return movement produced a marked phask response (R3) consisting of both excitation and inhibition. At 270 this movement was towards the tail, that is, in the same direction as the first PEM, Sl, at !Ml”, but starting at an eccentric position and returning to the central position. Thus, in Fig. 5, PEM in the same direction but through different arcs produced similar responses.

149

Extraocular muscle affcrent signals in pigeon brainstem

Sl 270’

y 53

180’ \

\

/

\ 8li\

83

/f

83

Fig. 2. Diagrammatic illustration of the PEM stimulus used and its relation to the analogue trace on the histograms. The eye was held centred in the orbit from which it was moved rapidly (Sl) to an eccentric position through one of three arcs in the orbit as shown in the upper part of the figure. After being held at the new position for 200 ms (S2), it was returned to the centre of the orbit (S3). The lower part of the figure shows the phases of the stimulus as they appear on the eye-position record in succeeding figures. The arrangements were similar in experiments in which four and in which eight PSTHs were collected except that, in the latter, movements along seven different orbital radii (rather than three) were tested. PBBA0102.VEM

24

sweeps

Bin

width

110.0

0

0.S

L.0

1.5

sec.

2.0

10

msec.

Sweep

duration

2.5

sec.

0.0

2.s

0

0.1

.s.s

i.0

‘2.0

sec.

Fig. 3. Set of four PSTHs of the response of a unit to PEM of the ipsilateral eye by 14” at 127”/s. The PSTHs, which are made up of 24 stimulus presentations, have a bin width of 10 ms and a total duration of 2.5s. The PSTHs were interleaved in time so that the four histograms were recorded effectively simultaneously. Upper traces show the PEM in the directions indicated; 270” and 90” correspond to movements of the eye through arcs in the horizontal plane towards the beak and the tail, respectively; 180’ indicates movement downward through an arc in the vertical plane. Top left: control histogram, no PEM. Note that PEM produces phasic excitation followed by long-lasting suppression of activity, and that this unit responds best to PEM in a particular plane (horizontal, 90” and 270”) and, within the horizontal plane, its maximum response is given to movement towards the beak (270”).

8.5

IS0

I.M.

P98A040

1. VEM

24

sweeps

L.

Bin

DONALDSON width

sec.

and P. C. KNOX 10

msec.

Sweep

duration

2.5

set

sec.

Fig. 4. Set of four interleaved PSTHs showing e&cts of combining vzstibukr and PEM stimuli. Upper traces in each panel show the stimulus; tabk position only in top left. tabk position with eye-position superimposed in the other three panels. Details of PSTHs as for Fig. 3. Top kftz control; vestibukr stimulation alone. Unit teeponds to contrahxteral rotation. Top right: combination of vestibular stimulation and PEM at 90“; initial movement (SI) in the horizontal plane directed towards the tail. Bottom right: vatibular stimulation and PEM at 270”; Sl in the horizontal plane directed towards the beak. Bottom kft: vestibular stimulation and PEM at 180”; SI in vertical plane directed down. Note elfect of PEM on vestibular msponse; a clear phmic inhibition in all three test histograms,with the largest effect at 90” (in the horizontal plane, towards the tail).

The unit of Fig. 6 responded differently. The first PEM at 90” (Sl, toward the tail) produced a small but ciear inhibition in the vestibular response; the same direction of movement, but through a different arc (the return PEM, S3, at 270”), produced a marked excitation. At 180” (downward PEM) there was both an inhibition in RI, which was smaller than that observed at 90”, and a small excitation in R3, which was smaller than the excitation observed at 270”. The interaction at 180’ was intermediate with respect to the interactions at 90” and 270”. Thus, in this unit, PEM in the same direction but through different arcs produced different responses-inhibitory in one arc and excitatory in the other. The behaviour of this unit differs from that of Fig. 4 in showing marked excitation to the S3 movement in two of the three arcs tested. Thus, the unit of Fig. 6 is sekcti~e for a combination of direction of movement and of orbital arc. We observed some units in which PEM, as wdl as giving phasic modifications of vestibular response, had long-lasting effects which were distinct from the phasic actions. One exam+ is shown in Fig. 5 where the response to PEM towards the tail (90”) contained

an in~bition of the vestibular response (centred at about I .25 s) which was quite distinct from the phasic inhibitory e&ts seen at other angles. Another exampk of long-lasting effects is illustrated in Fig. 7; PEM produced small phasic effects but only PEM towards the tail caused a statistically significant phasic inhibition (90”, Rl P < 0.025). However, PEM in the horizontal plane, both towards the tail (90’) and towards the beak (270”) caused significant reductions in firing from the control level 170 ms after the eye had been mtumed to the central position (W’, R4, P 6 0.005; 270’, R4, P G 0.025). PEM at 180” produced no statist~caily signitican~ modifir%ion of the vestibular response after the eye had been returned to the eentfai position. Long-lasting effects of PEM are a common feature of interactions between PEM and vestibular responses. Over all, some 42% of units in which interactions were observed showed long-lasting effects either alone, or along with phasic e%cts. The interaction between vesdbular and PEM stimuli depands not only upon the dire&on of PEM, but also on the time of occurrence of PEM within the vestibular cycle. In the experiment of Fig. 8 the

151

Extraocutar muscle aRerent signals in pigeon brainstem P79A0201.VEM

0

24

0.5

I.0

0.1

1.0

sweeps

Bin

10

width

msec.

Sweep

duration

2.5

sec.

1.s

2.0

2.5

0

0.5

1.0

1.0

a.0

Z.S

,.a

2.0

*.LI

0

0.5

1.0

1.5

2.0

Z.5

sec.

sec.

Fig. 5. Set of four interleaved PSTHs showing response of a unit to combinations of vestibular and PEM stimuli. Conventions as for Fig. 4. Note that, in addition to the effect of Sl, the initial eccentricalIy directed PEM, other effects are now &ear. Contrast with Fig. 4. See text for detailed discussion. This unit is “direction-selective”. P03DOIOl.VEM

24

sweeps

Bin

width

10

msec.

Sweep

duration

2.5

sec.

90.0

9

I, -_:.~, 0

0.2

1.0

,.?I

2.0

2.5

1 0

0.Z

1.0

0

0.*

1.0

a.0.

2.0

8.1

:: d

100.0

0

0.e

1.0 sec.

5.s

2.0

d.2.

2.3

l.Z

.2.0

sec.

Fig. 6. Set of four interleaved PSTHs showing response of a unit to combinations of vestibular and PEM stimuli. Conventions as for Fig. 4. Note differences in the responses in the three test histograms. The responses of this unit are different from those of the units of Figs 4 and 5. See text for detailed discussion. This unit is selective for both “arc” and “direction”.

‘2.2

152

and

I.M.L.D~NALDsoN P72AQ2Oi.VEM

24

sweeps

Bin

width

10

P.C. msec.

KNOX

Sweep

duration

2.5

SAC

z-90.0 -3.D

1 0

0.5

L.0

1.5

2.0

2,s

sec.

11 0

0.1

1.0

I.5

9.0

a.5

sec.

Fig. 7. Set of four interleaved ISTHs showing response of a unit to combinations of vcstibular and PEM stimuli. Conventions as for Fig. 4. Note that, in this unit, PEA4 has only a small physic effect, but there is a clear long-lasting reduction in-the v&War response, which is most pronounced at 90” (horizontal movement towards the tail).

direction of PEM was kept constant (horizontally towards the tail for all the histograms) but the time in the cycle at which the eye was moved varied between histograms. Early in the cycle, when there was very little vestibular activity, PEM caused a marked phasic excitation. During the early part of the vestibular response (Fig. 8b-d), the phasic excitation remained clear and was slightly enhanced as the vestibular response increased. At a slightly later time, as the vestibular response peaked and began to decline, the phasic excitation cotrId not be distinguished from the background (Fig. 8e). There was also a second effect of PEM; that of long-lasting suppression of the vestibular response, which is particularly clear (and statistically signifkant, P G 0.005) in Fig. 8d. Thus the occurrence of both the phasic and the long-lasting inhibitory effects depended upon the time at which PEM was delivered during the vestibular cycle (see also Discussion). As is clear from the above figures, the veatibular response of many units was modified by PEM in both horizontal and vertical pIarms. However, in some cases, PEM in only one pIsme was e&etive, and in others a preference for one plane was ckar. Of the 183 units in which interactions were &arved, the responses of 124 were suI&cted to furtbcr statiatkal tests (see Experimental Pmcedurca). ‘PIw vc++r responses of 42 (34%) units were modifkd by PEM only in the horizontal plane; just eight (6%) were

modified only by vertical PEM. Of those units in which PEM in both vertical and horizontal planes produced an effect, 23 (18%) showed larger effects with PEM in the horizontal plane, while three (2%) preferred vertical PEM. No statistically significant preference for one plane of PEM over the other was found in 48 units (39%). Tuning of passive eye-movement

interacrion effects

A second type of experiment was performed to investigate the tuning of responses to PEM in planes between the horizontal and vertical. Instead of four P!STHs (one control and three with PEM). sets of eight PSIHs were constructed allowing the coikction of data for PEMs in arcs lying in several planes between the horizontal and vertical. Tuning information was collected on 29 of the units described above which showed interactions between vestibular and PEM stimuli. In ahnost all units the efiect of PEM remained consistent over a number of expwimcntal tests over long periods of time. Compriacms of the preliminary sets of four PSTHs with the sets of eight (tuning) PSTHs mveaIod only three units of 29 in which there was incousistency between the two sets of data. The tuning experiments reveakd that PEh4 produced various patterns of modulation of the vestibular response. An exampk of a unit in which PEM produced an inhibitory modulation of the

153

Extraocular muscle afTerem signals in pigeon brainstem 24

sweeps

Bin

width

10 msec.

Sweep

duration

2.5

sec.

e

b

0.5

L.0

L.5

2.0

2.5

sec.

sec.

Fig. 8. Set of interleaved PSTHs showing response of a unit in the anterior part of the medial vestibular nucleus to vestibular and PEM stimuli. In each test histogram the PEM is at 90” (Sl towards the tail) but the time of occurrence of the PEM within the vestibuiar cycle is different in each. Conventions as for Fig. 4. (a) Control; vestibular stimulation only; (b) PEM 0.2 s after start of vestibular cycie; (c) PEM 0.5 s after start of vestibular cycle; (d) PEM 1.0 s after start of vestibular cycle; (e) PEM 1.4s after start of vestibular cycle. Note the change in size of the phasic excitation due to PEM at different times in the vestibular cycle and the presence of long-lasting inhibition of the vestibular response at one time only (d). See text for further details.

vestibular response is shown in Fig. 9. Supe~cially it seemed that any eccentrically directed PEM (Sl) was effective in producing a large inhibitory effect; the responses at all angles were significantly different (P G 0.005) from the response to vestibular stimulation alone. However, the tuning plots reveal that those PEMs with a horizontal com~nent directed towards the tail produced the largest inhibitory effects. The inhibition due to PEM towards the beak (RI 270”) was significantly smaller than that at all angles other than at 225”. Responses towards the tail and beak (90” and 270’, respectively) were signifi~ntly different (P < 0.055); the largest inhibition was at 90” (movement towards tail). Thus,

for this unit, the most powerful and the least powerful effects, which were inhibitory in both cases, were produced by movements in the horizontal plane; the largest effect arose from PEM towards the tail and the smallest effect was produced by PEM towards the beak. The vestibular response recovered dung S2 after which there was a less pronounced, but still statistically significant, inhibitory modulation due to the centrally directed return PEM at each angle (R3, P G 0.005for all test values compared with the control). Again, PEM towards the tail was the most effective in inhibiting the vestibular response. Thus, in this unit, as in that illustrated in Fig. 5, movements

I. M. L. DWALDSON and P. C. KNOX

154

in similar directions, but through different arcs. produced similar responses. Six of the 29 units subjected to tuning analysis showed no evidence of tuning; they were modulated to the same extent by PEM in any direction. The remaining 23 units demonstrated various degrees of tuning. The unit discussed above, and whose response is illustrated in Fig. 9, showed an intermediate level of tuning. An example of a unit tightly tuned to a particular direction of PEM is shown in Fig. IO in which PEM produced excitatory modulation of the vestibular response. The largest effect was produced by PEM directed towards the beak but with a downward component (240 in Fig. IOa). To investigate this pattern of response at higher resolution, PSTHs were constructed for seven PEMs equally spaced betwen 180 and 360 , so concentrating on the angle at which we observed the best response. The RI response (Fig. lob) was consistent with, and very similar to, that in the previously collected set of PSTHs (Fig. IOa). Between 205 and 230’ there was a dramatic increase in the effectiveness of PEM. This unit responded not only to downward movement with a component towards the beak, but also to downward movement with a tailward component as the R3 response shows (Fig. 10~). A further nine units showed similar tight tuning, although it was not always possible to hold units long enough to construct high resolution plots around the angles of principal response. Areas in brainstem,from

which responses were recorded

The recording sites of 29 units in the brainstem were established by making electrolytic lesions and reconstructing electrode tracks from histolo~cal sections (Table I). Although some regions in the brainstem were easily identifiable, for example the medial vestibular nucleus and the motor nuclei, the boundaries of others, such as the dorsal part of the central diBinucleus of the medulla, were more cult to delineate. However, units which responded strongly to horizontal vestibular stimulation were not limited to the medial vestibular nucleus. Many of the units which were not located histologically were

recorded deep in the brainstem; it is probable that these were in the reticular formation deep to the medial vestibular nucleus. Controfs

Several control experiments were carried out to eliminate sources other than orbital afferents as the effective source of signal during PEM. Units which responded to scraping the beak or head. or stretch of the eye-lids, were immediately rejected. These units were usually only weakly modulated by the vestibular stimulus but some of them appeared to respond well to PEM. Application of local anaesthetic to the cornea of the eye being moved did not affect the response to PEM. No units were found which responded to auditory or vibratory stimuli. It was important to eliminate retinal signals as a source of effective input. The suction-lens which moved the eye was opaque and many recordings were made in darkness. This did not alter the effectiveness of PEM in prodding responses. Several experiments were also performed in which retinal function was blocked; the result of one of these is shown in Fig. Il. An averaged field-potential in response to a brief flash was recorded from the surface of the optic tectum contralateral to the eye being moved (Fig. I la). and the response of a single unit to vestibular and PEM stimuli was recorded (Fig. I lb). Injection of barbiturate into the globe of the eye which was moved passively abolished the tectal visual response (Fig. I Id) while the effect of PEM was unchanged (Fig. I lc). This confirms that PEM remains effective in the absence of retinal signals, as we have shown in previous series of experiments.5.h.’ DISCUSSION

Pigeons and eve -mooemenrs Pigeons have been used extensively in studies of both the visual”*2’*z7 and the vestibular systems.z~3*z6 It has been reported that birds depend to only a limited

extent

on eye-movements

and

rely mainly

Fig. 9. Set of eight interleaved “tuning” PSTHs for responses of a unit to a combination of vestibular and PEM stimuli and polar plots derived from these. Because of lack of space the table-position and eye-position ate not shown on this figure. A set of four PSTHs from sameunit is shown in Fig. 4. (a) Set of eight PSTHs contains one control (top left: vestibular stimulation only) and seven test histograms with PEM in the following directions: 0” vertical plane, Sl upwards: 45’1 intermediate ph~e, SI up and tailward; 90’ horizontal plane. Sl t&ward; 135” intermediite plane, St down and tailward: 186” vertical plane, St downwards; 225” intetiiate plane. Sl down and towards bmk; 270’ horizontal phone. SI towards beak: 3 15‘ intermediate phme, St up and towards beak. (b and c) The polar pIots were ~~t~~ by counting the number of spikes Falling in a time window in each test histogram. Solid &&es indicate the number of spikes in the widow in the control histogram. The responsesare plotted as vectors in which the distance of a point from the cesrtre of the plot reptesents the number of spikes and ade of the vector shows the direction of PEM. In tbesc polar plots the distance of each cross from the surround&g circle indicates the strength of the inhibitory effect prod& by PEM at that angle. (b) polar plot of RI. the response to Sl. The window was set fram bins I50 to 156 (70ms). (cc)Polar plot of R3. the cc~pons~to S3. The window was set from bins 175 to 181 (70ms). Note that RI is the response to the &titialf eccentrically directed PEM, whereas R3 is the response to the (final) centrahy directed PEM.

Extraocular muscle afferent signals in pigeon brainstem

155

a

Fig. 9.

156

1. M. L.

D~NALDXIN

and P. C.

KNOX

on movements of the head for visual exploration of the surroundings. However, it is clear that birds can and do make eye-movements. Even in owls, which have enclosed bony orbits and extremely mobile heads, spontaneous eye-movements can be recorded, some of which occur in response to visual stimuli.” Pigeons possess most of the types of eye-movement observed in the higher vertebrates. Saccadic eyemovements of up to 15” have been observed,’ and pigeons exhibit a form of vergence used in binocular fixation during feeding. There is also a well-developed vestibulo-ocular reflex (VOR).’ So it seems that eyemovements may be more important in birds in general, and in pigeons in particular, than was believed previously.” In the decerebrate pigeon the afferent pathway taken by signals from the EOM proprioceptors is intact. The advanced development of central vestibular and related structures, with relatively high cell densities in fairly discrete centres, allied with the removal of the complicating influence of general anaesthetics, has allowed us to accumulate a larger amount and wider range of data than was possible in our studies on EOM afferent signals in other species such as the cat,4 toad6 and trout.’ Having established the existence of interactions between EOM proprioceptive signals and vestibular responses in the pigeon, we have gone on to look at some of the characteristics of the interactions.

exclude retinal and somatosensory signals as the effective input during PEM. We found no units which responded to vibratory or auditory stimuli. Application of local anaesthetic to the eye has no effect on the responses so stimulation of cornea1 or conjunctival receptors can be eliminated as the source of signal. Thus, we conclude that the effects produced by PEM were due to stimulation of receptors in the EOMs themselves (see also discussion in Refs 4 and 7). There is little information about the receptors in the pigeon’s eye-muscles. It appears that birds in general lack muscle-spindles in their EOMS’*.~~but do have unencapsulated receptors. In the cat, which also lacks muscle-spindles in its eye-muscles, afferent information is provided by simple spirals and tendon organs’,” (see also Ref. 36). In the pigeon. injection of HRP into EOMs is reportedI to produce labelling of cell-bodies in the descending nucleus of the trigeminal nerve. This contrasts with findings in cat9.j4 and monkey32.3’ that primary afferent cell-bodies are labelled in the semi-lunar ganglion. with their central terminals in the spinal trigeminal nucleus. The reasons for this apparent, and striking, difference in the primary afferent path between cat and pigeon are not clear and the point needs further examination. Even in the cat, however, in which the EOM afferent pathway has been studied quite extensively, there is still disagreement about its details (see Spencer and Porter36 and Donaldson and Porter”).

Choice of stimuli

Nature of interactions

Our reasons for using PEM to stimulate EOM afferents, rather than using muscle stretch or electrical stimulation, have been presented in detail elsewhere.’ Briefly, stretch of a single eye-muscle, or even of several EOMs, simulates natural eye-movement poorly since any natural eye-movement involves simultaneous changes in the length of several, or of all, eye-muscles. PEM simulates natural eye-movement better than does the stretch of individual muscles, although still imperfectly. In the present experiments, the vestibular stimulus was provided by sinusoidal oscillation of the bird in the horizontal plane. Modulation of activity by the vestibular stimulus was very clear in most units.

In the pigeon we have been able to obtain recordings from large numbers of units which were held for long periods of time. In the units which we have examined the response to vestibular and PEM stimuli over a period of time is remarkably constant. This is particularly clear in those experiments in which we examined the tuning of responses: in some cases this involved a number of tests on a particular unit over several hours. It is clear that the two stimuli used in these experiments, horizontal rotation of the bird f producing horizontal canal stimulation) and PEM, interact in a number of distinct ways. The effects of PEM on vestibular responses were either purely phasic or showed a combination of phasic and longer-lasting effects. Although our experimental arrangement is not well-suited to the accurate measurement of short latencies it is clear that most of the phasic interactions begin within tens of milliseconds of the start of the PEM. Occasional units, though, responded at latencies up to approximately IOOms. Thus it is

Source of afferent signal during passive eye-movement As in our earlier experiments on other species,” we have sought to establish the location of the receptors which are stimulated by PEM by a process of elimination. The results of control experiments which used mechanical stimulation or retinal block allow US to

Fig. 10. Polar plots from tuning data of interaction between PEM and vestibular stimuli. Unit in nucleus reticularis parvoceIh&ris. Conventions as for Fig. 9. Note that PEM produced an excitatory interaction. All plots from the same unit. (a) PEN at angles between 0” and 360’-, plot of RI. Note that PEM at 0 and at 360”. the same stimulus, produced approximately the same effect. (b) The same unit, again RI, but with PEM at seven angles between 180” and 360” to give better resolution in the region of the best response. (c) Plot of R3 from the same PSTHs as b. Note inversion of axes compared with b.

157

Extraocular muscle afferent signals in pigeon brainstem

a

b

+ 180

up

Fig. 10.

1.M.L.

158

DONALDSON and P.C. KNOX

Table I Histological location MVN CND RF MLF Number of units responding to: I Passive eye-movement only Vestibular stimulation only 13-Both vestibular stimulation and passive eye-movement 7 Totals

9

I-7

4

5

I1

4

5

MVN, medial vestibular nucleus; CND, central nucleus of the medulla (dorsal part); RF, reticular formation of the brainstem; MLF, medial longitudinal fasciculus.

possible that these modifications in the vestibular responses serve to adjust the vestibular drive to EOM motoneurons in the short term so compensating for perturbations in the intended eye-movement. It has often been assumed that a role for afferent feedback in the control of eye-movements is excluded by the lack of a dynamic stretch reflex in EOMs. In one well-known experiment” activity was recorded from abducens motoneurons in alert rhesus monkeys and modifications to the motor output were sought either when the eye was prevented from moving or during a spontaneous or imposed eye-movement. No change in motoneuron discharge was observed. However, the authors point out that their findings deal with only one possible function of EOM proprioceptive feedback, and by analogy with spinal reflexes, a

a

“low order” reflex at that. Our experiments are concerned with a different aspect of oculomotor control. It is known that cells which are driven by canal afferents, and thus respond to vestibular stimulation, in turn drive oculomotor motoneurons; this is the basis of the VOR. We have shown that the vestibular responses of cells in various brainstem centres can be modified by signals from EOM proprioceptors. Thus these data suggest that, during the VOR, afferent signals from EOM receptors might modify the motor output to the EOMs by modifying the vestibular drive to their motoneurons. This in no way conflicts with earlier findings about the absence of a dynamic stretch reflex in EOMs. It is interesting that, when EOM afferents were blocked in the rab.bit,2’ the gain of the horizontal VOR fell by 2440% and the slow phases of the horizontal VOR were disrupted.23 It is also very significant that, following our earlier similar observations in trout,’ we have found recently” that the responses of units in the abducens nucleus of the pigeon to vestibular drive are, indeed, modified by passive eye-movement; the details of these interactions are being studied at present. Taken together, all these observations suggest that it is probable that EOM afferents do act to modify the characteristics of the VOR though direct testing of this hypothesis in the pigeon awaits further experiments using slow eye-movements rather than movements at saccadic velocities.

Y V.

.

a

Y

!‘a

Iam

10s

Isa

au

25s

Fig. 11. Control experiment using retinal block. Time scclla in miltiac~nds. (a) Avera& lidd poCmtial from contralateral optic taztum to br#f &ah of light to lcR eye. (b) EItaa of ~~tib&r stimuli and PEM of the ipsihtcr8l (I&) aye on tbc response of a unit in the brainstem. (c) ElTect of the UISDC vestibular stimuli and PEM after bbcking ~&MI fun&on with barbiturate. (d) Afidd patmretinal block. M made iwnsdLtely &cr tial from contri3latcral optic tccUnn dcawnrtntiag collecting PSTH of bottom kft. Clearly, blocking the retina of the eye which is moved does not alter the effects of PEM.

Extraocular muscle afferent signals in pigeon brainstem Another aspect of the interactions in the present experiments is the “long-term” effect of PEM on vestibular responses. We have avoided the use of the term “tonic” as this suggests an effect which lasts as long as the eye remains held at a new position. The long-term effects which we have found seem to be due to the dynamic component of the stimulus, the movement, as opposed to the static component, since increasing or decreasing the time for which the eye is held at the deflected position has little effect on the long-term response. In Fig. 7, for example, the longterm effects are best seen after the eye has been returned to the central position and consist of a general reduction in the vestibular response. Interestingly, this effect of PEM often also has planar selectivity; in the unit of Fig. 7 PEM in the horizontal plane produced the largest modification in the vestibular response. Given that the long-term effects are clearly not simply the phasic modifications delayed in time, they may serve a different function. If the phasic modifi~tions could serve to provide correction for perturbations in the intended movement, perhaps the long-term modifications might indicate storage of afferent signals; this, in turn, could be the first step in parametric adjustment of the behaviour of the oculomotor system produced by the aBerent feedback from the EOMs. Of course, these suggestions are speculative at the moment. The absence of a clear tonic type of modification need not mean that no position signal is generated from the EOM afferents. Indeed, the idea that a tonic position signal could be derived from a phasic signal (a velocity signal) forms a central part of many models of the oculomotor control system.” As we have pointed out elsewhere,’ the principle of integration would seem to be equally credible whether the velocity signal were of afferent or of central origin. The details of the EOM afferent signal as it enters the central nervous system in the pigeon still need to be established; for example, recording from the first central synapse should establish whether the EOM receptors provide only phasic signals or whether there is a tonic signal which could supply afferent information about eye-position. The interaction between the signal from EOM proprioceptors and the vestibuiar responses is certainly not always linear. As is clear from Fig. 8, the effect observed is dependent on when, during the vestibufar response, PEM occurs. This observation confirms an earlier result in the cat.’ This aspect of the modification of the vestibular response needs to be studied in more detail but the non-iinearity will pose a difficulty in attempts to make mathematical models of the process of the interaction. “Tuning” of effects of passive eye-fnoveinerzt on vestibular responses Our experiments select units which respond to horizontal vestibular stimulation. It is significant that, in a large proportion of the units in which PEM NK

M/I-F

159

modlfles the vestibular response, it is PEM in, or near, the horizontal plane which produces the largest modifications. In the present experiments we have been able to take our analysis beyond the comparison of the effects of PEM in only horizontal and vertical planes which we were able to make in earlier studies.4*7 experiments have revealed that The “tuning” there are different degrees of sharpness of tuning of interactions. There are some units whose response to vestibular stimulation is profoundly modified by PEM in any direction. However in others (see Fig. lo), the modifications observed are clearly very specific to PEM in certain planes. The responses of these highly-tune units are not consistent with a hypothesis that these interactions are the effects of a non-specific signal which indicates no more than that the eye-movement has occurred. Responses to PEM have also been recorded in the cat primary visual cortex (area 17).’ Responses were phasic and excitatory, and, as with the responses described in this paper, many were “radiallyselective”, that is to say that PEM often produced a greater effect in one plane than in another. Within the radially-selective group, directionally-selective and arc-selective units were described. Directionally-selective units responded to PEM in a particular direction; the arc in the orbit through which the eye was moved was of less importance. The unit of the present Fig. 5 responded in this way. In contrast, arc-selective units responded in a similar way to PEM along given arcs, even when the directions of movement were different. The unit of the present Fig. 9 responded in this way. Although tuning curves were not constructed for the “radially-selective” cortical responses, we pointed outs that they showed degrees of tuning ranging from “well-tuned” to non-selective. Taking together the findings from the cat and the pigeon, it is now clear that information about eye-movements in particular orbital arcs, as well as in particular directions, is available from EOM proptioceptors and travels to nuclei in the brainstem as well as to the visual cortex. The similarity in the effects suggests that the EOM receptors provide similar patterns of information from the orbit to these different levels of the nervous system. The implications of different degrees of tuning are as yet uncertain but their interpretation in any particular case must depend on how much processing of the two signals there has been at the level in the system at which the responses were recorded. If we are dealing with a level near the input of the system, where the proprioceptive inflow is relatively unprocessed, the tuning we observe may reveal something about the spatial organ~tion of the muscle afferents in the orbit. On the other hand, if we are recording nearer the output of the system, with the modified vestibular drive feeding forward to EOM motoneurons with little or no further processing, the different degrees of tuning may reveal different

160

1. M.L.DONALDSON

degrees of involvement of units in particular directions of eye-movements. It is, of course, not unlikely that both of these suggestions may hold. At the moment, the histological location of the units from which we recorded does not tell us whether an individual unit lies nearer the input or closer to the output of the control system. Localions of units which show interactions

The responses described above were recorded from a number of sites in the brainstem. Some units were certainly in the vestibular nuclei, in particular in the medial vestibular nucleus, as shown by the lesions made at the recordings sites and identified on histological sections. The electrode tracks in many of our earlier experiments passed through the medial nucleus but we did not have lesions to mark the sites of units on the tracks. However, it was clear in these early experiments that many of the units recorded lay in the reticular formation deep to the anterior portion of the vestibular nucleus, in part of the nucleus centralis medullae (pars dorsalis). This was confirmed in later experiments in which lesions were available. We also recorded units with interactions in the nucleus reticularis gigantocellularis; this camp lements a similar finding in the cat.’ The details of the connections and function of this area do not yet seem to be known in the pigeon but, by analogy with mammals, it is likely to be involved in the control of eye-movement and, in particular, in the control of horizontal gaze.

and P.C.

KNOX

CONCLUSION

These experiments have shown that, in the pigeon, signals from EOM proprioceptors, stimulated by passive eye-movement, reach various brainstem centres involved with oculomotor control. Furthermore, these signals modify responses to horizontal vestibular stimulation in a directionallyselective manner. Thus the orbital proprioceptive signal is not merely an “eye-movement” signal. Some of the modifications observed are both longlatency and long-lasting. These findings, taken along with those in several other species, reinforce the hypothesis that orbital proprioceptive signals are likely to play an important role in oculomotor control. The suggestions that these signals may have an action on both the short-term control of eye-movements and the longer-term adjustment of the oculomotor system need to be tested, by experiments of a rather different design than those reported here, both on the vestibular nuclei and on the final common path of the system, the oculomotor nuclei.

Acknowle&menrs-This work was supported by grants from The Welkome Trust and The !%eoce and Engin&ng Research Council. The Sir Stauky and Lady Davidson Research Fund provided some of the equipment. We arc grateful to Mrs J. P. Donaldson for the preparation of all the histological material. Mrs C. Wdluton provided expert technical assistance. We are indebted to May and Baker, Ltd for the gift of Plaxedil.

REFRRRNCES

1. Alvarado-Mahart R.-M. and Piuam-Raymond M. (1979) The pa&xade endings of cat extraocular muscks: a tight and ekctroo m&ncope study. T&s. crll11,567-584. 2. AnasUsio A. J. and Correia M. J. (1988) A frequancy and time domain study of the horizontal and vertical vcxtibulo-ocular &la in the pipoo. 1. NewopIrysiol. S9, 1143-1161. 3. Aoastasio T. J., Corruia M. J. aad I%ach.io A. A. (1985) Spoo~us and driven responses of semicircular canal primary a!Tuents in the M pigam. 1. Neurophysiol.54, 335-347. 4. Ashton J. A., Boddy A., Daan S. R., M&ret C. and DonakIson I. M. L. (1988) Affercnt signals from cat extraocular muscks in the medial vcstibular nucleus, the nucleus praepositus hypoglossi and adjacent brainstem structures. Neuroscience 24, 131-145. 5. Ashton J. A., Boddy A. and Donaldson I. M. L. (1984) Directional sensitivity in the responses of units io cat primary visual cortex to passive eye movcmant. Neurosciersce 13, 653462. 6. Ashton J. A., Be&y A. and Donrldroa I. M. L. (1984) Input from proprioapton in the extrinsic ocular muszks to thevcstibularnuckiinthegiultto&~ 0 tMrinu.r. Expl Brain Res. sj, 409419. 7. Ashton J. A., Milkret C. aud Doua&ou 1. M. L. (1989) Effects of atTerentsignals from the extraocular mu&as upon units in ths cembcllum, v&ii m&at compkx and on&motor nuck.us of the trout. Newroscisnce 31, 529-541. 8. Blnch S.. Martinoya C. and Rivatxi S. (19gl) Eye movurrmts in piens: participation in binocular fhtaticmand visual pursuit. J. Pkysbl., Load. 3S, 20-21P. 9. Bortolami R., Luazhi M. L. and Raaoai V. E. (1987) Locahxation and somatotopy of fry calls irmarvating the cxtraocuhr mum&s of lunb, pir and at. Hirtochcrnial and ckctmphysiologiml iovatigation. A&s irol. &I. 125, l-15. q,frhe Eyes, 2nd aio. Pioo. London. 10. Cupenter R. H. S. (1988) Mm 11. Cooper S. arid Fi M. (1955) A&t& di&ugcs io response to stretch from the extraocular mu&s of the cat and monkey and the inacrvrtioa of t&n mr&es. J. Physttd...Lmd. 127, 400-413. 12. Cooper S. and Da& P. M. (19@) Mu& spiadks in human extrinsic eye muscles. Brain 72, l-24. 13. D&ra&idt D. G. @X31) The statistial si@uace of the pcristimulus time histogram (PSTH). B&n Rex 22& 397401. 14. Donakisoo 1. M. L. and Knox, P. C. (1998) Units in pigeon brainstem whose vestibular responacs arc modulated by passive eye-movement. J. Physiol., Land. m, 35P.

Extraocular muscle afferent signals in pigeon brainstem

161

15. Donaldson I. M. L. and Knox P. C. (1990) Alferent signals frotn the extraocular muscles affect vestibular responses in oculomotor nuclei of the pigeon. J. Physiol., Land. 420, 106P. 16. Donaldson I. M. L. and Long A. C. (1980) Interactions between extraocular proprioceptive and visual signals in the superior colliculus of the cat. J. Physiol., Land. 298, 85-110. 17. Donaldson I. M. L. and Porter J. D. (1990) Brainstem terminations of extraocular muscle primary afferents in the cat. J. Physiol., Land. 420, 105P. 18. Eden A. R. and Correia M. J. (1987) Improved fixation of the pigeon brain by transcardiac carotid catheterization. Physiol. Behav. 27, 947-949. 19. Eden A. R., Correia M. J. and Steinkuller P. G. (1982) Medullary proprioceptive neurons from extraocular muscles in the pigeon identified with horseradish peroxidase. Brain Res. 237, 15-21. 20. Hodos W.. Bessette B. B.. Macko K. A. and Weiss S. R. B. (1985) uiaeon vision. Vis. Res. 25. . , Normative data for _ _

1525-1527:

21. Hodos W. and Bonbright J. C. (1972) The detection of visual intensity differences by pigeons. J. exp. analyt. Behau. 18, 471-479. 22. Karten H. J. and Hodos W. (1967) A Sfereotaxic

At/as of the Brain of the Pigeon (Columba hvia). Johns Hopkins University Press, Baltimore, MD. 23. Kashii S., Matsui Y., Honda Y., Ito J., Sasa M. and Takaori S. (1989) The role of extraocular proprioception in vestibulo-ocular reflex of rabbits. Inuesr. Ophthal. uis. Sci. 30, 2258-2264. 24. Keller E. L. and Robinson D. A. (1971) Absence of a stretch reflex in extraocular muscles of the monkey. J. Neurophysiol. 34, 908-919. 25. Kimura M., Takeda T. and Maekawa K. (1981) Functional role of extraocular muscle afferents in the control of eye movements in rabbits. J. Physiol. Sot. Japan 43, 317. 26. Lifschitz W. S. (1973) Responses from the first order neurons of the horizontal semicircular canal in the pigeon. Brain Res. 63, 43-57. 27. Macko K. A. and Hodos W. (1984) Near-field visual acuity changes after visual system lesions in pigeons. I. Thalamus. Behav. Brain Res. 13, I-14. 28. Maier A., De Santis M. and Eldred E. (1971) Absence of muscle spindles in avian extraocular muscles. Expl Eye Res. 12, 251-253. 29. Maier A., De Santis M. and Eldred E. (1974) The occurrence of muscle spindles in extraocular muscles of various vertebrates. J. Morph. 143, 397-408. 30. Merrill E. G. and Ainsworth A. (1972) Glass-coated platinum-plated tungsten microelectrodes. Med. Biol. Engng 10, 662672. 31. Nye P. W. (1969) The monocular eye movements of the pigeon. Vis. Res. 9, 133-144. 32. Porter J. D. (1986) Brainstem terminations of extraocular muscle primary afferent neurons in the monkey. J. camp. Neurol. 247, 133-143.

33. Porter J. D., Guthrie B. L. and Sparks D. L. (1983) Innervation of monkey extraocular muscles: localization of sensory and motor neurons by retrograde transport of horseradish peroxidase. J. camp. Neurol. 218, 208-219. 34. Porter J. D. and Spencer R. F. (1982) Localization and morphology of cat extraocular muscle afferent neurones identified by retrograde transport of horseradish peroxidase. J. camp. Neural. 204, 56-64. 35. Sherrington C. S. (1918) Observations on the sensual role of the proprioceptive nerve supply of the extrinsic ocular muscles. Brain 41, 332-343. 36. Spencer R. F. and Porter J. D. (1988) Structural organisation of the extraocular inuscles. In Neuroanaromy of the Oculomotor System (ad. Biittner-Ennever J.), pp. 33-79. Ehevier, Amsterdam. 37. Steinbach M. J. and Money K. E. (1973) Eye movements of the owl. Vis. Res. 13, 889-891. (Accepted 26 March 1990)