Experimental BrainResearch

Exp Brain Res (1990) 79:225-239

9 Springer-Verlag1990

Effects of lesions of the nucleus of the optic tract on optokinetic nystagmus and after-nystagmus in the monkey D. Schiff, B. Cohen, J. Biittner-Ennever, and V. Matsuo Departments of Neurology and Physiologyand Biophysics, Mount Sinai School of Medicine, City University of New York, 1 East 100th Street, New York, NY 10029, USA and Department of Physiology,Universityof Munich, Munich, Federal Republic of Germany

Summary. 1. The nucleus of the optic tract (NOT) and the dorsal terminal nucleus (DTN) of the accessory optic system were lesioned electrolytically or with kainic acid in rhesus monkeys. When lesions involved NOT and DTN, peak velocities of optokinetic nystagmus (OKN) with slow phases toward the side of the lesion were reduced, and optokinetic after-nystagmus (OKAN) was reduced or abolished. The jump in slow phase eye velocity at the onset of O K N was smaller in most animals, but was not lost. Initially, there was spontaneous nystagmus with contralateral slow phases. O K N and O K A N with contralateral slow phases were unaffected. 2. Damage to adjacent regions had no effect on O K N or O K A N with two exceptions: 1. A vascular lesion in the M R F , medial to NOT and adjacent to the central gray matter, caused a transient loss of the initial jump in OKN. The slow rise in slow phase velocity was prolonged,

but the gain of O K A N was unaffected. There was no effect after a kainic acid lesion in this region in another animal. 2. Lesions of the fiber tract in the pulvinar that inputs to the brachium of the superior colliculus caused a transient reduction in the buildup and peak velocity of O K N and OKAN. 3. In terms of a previous model (Cohen et al. 1977; Waespe et al. 1983), the findings suggest that the indirect pathway that activates the velocity storage integrator in the vestibular system to produce the slow rise in ipsilateral O K N and OKAN, lies in NOT and DTN. Activity for the rapid rise in OKN, carried in the direct pathway, is probably transmitted to the pontine nuclei and flocculus via an anatomically separate fiber pathway that lies in the M R F . A fiber tract in the pulvinar that inputs to the brachium of the superior colliculus appears to carry activity related to retinal slip from the visual cortex to NOT and DTN.

Abbreviations used in figures: BIC, brachium of the inferior

Key words: Optokinetic nystagmus (OKN) - Optokinetic after-nystagmus (OKAN) - Velocity storage Nucleus of the optic tract (NOT) - Dorsal terminal nucleus (DTN) - Monkey

colliculus; BSC,brachium of the superior colliculus;C, caudate nucleus; CG, centralgray; CL, Centralislateralis; dbc, decussation of the brachium conjunctivum;DTN, dorsal terminalnucleus of the accessoryoptic system;IC, inferiorcolliculus;Hb, habenular nucleus; hc, habenular commissure; LD, lateralis dorsalis; LGn, lateral geniculatenucleus; MD, medialisdorsalis; MGn, medial geniculate nucleus; MLF, median longitudinal fasciculus; MRF, mesencephalic reticular formation; cMRF, central mesencephalicreticular formation; NL, nucleus limitans; NLL, nucleus of the lateral lemniscus;NOT, nucleus of the optic tract; PB, parabigeminal nucleus; pc, posterior commissure; Pi, pineal gland; PON, pretectal olivary nucleus; Pt, pretectum; Pulv, pulvinar; R, nucleus reticularis; RN, red nucleus; RpN, raphe nucleus; RTP, nucleusreticularis tegmenti pontis; SC, superior colliculus;SCp, superior cerebellar peduncle; VPL, ventralis postero-lateralis; VPM, ventralis posteromedialis; III, oculomotor nucleus; IV, trochlear nucleus; IVn, trochlear nerve; Vm, mesencephalictrigeminalnucleus Offprint requests to: B. Cohen, Annenberg21-74, Box 1135, Mount Sinai School of Medicine, 1 East 100th Street, New York, NY 10029, USA

Introduction The nucleus of the optic tract (NOT) and the dorsal terminal nucleus of the accessory optic system (DTN) have been implicated in producing O K N in the monkey, cat, rabbit and rat (Collewijn 1975a, b; Hoffmann 1982; Maekawa et al. 1984; Hess et al. 1985; Hoffmann and Distler 1986; Kato et al. 1988; Schiff et al. 1988; see Simpson et al. 1988 for review). In a previous study we showed that electrical stimulation in the region of NOT in the monkey produced nystagmus and after-nys-

226

tagmus whose characteristics were similar to those of the slow component of O K N and O K A N (Schiff et al. 1988). At the onset of stimulation nystagmus velocity rose slowly to a steady state level, generally between 60 and 90 deg/s where it was maintained for the duration of stimulation. The time course of rise of the stimulus-induced nystagmus and the peak velocities were similar to the charging time constant and peak velocities of O K A N (Cohen et al. 1977). At the end of stimulation there was an after-nystagmus that had the same characteristics as O K A N . In terms of a previous model (Cohen et al. 1977; Raphan et al. 1979), we inferred that N O T and D T N are part of the pathway that activates the velocity storage integrator in the vestibular system to produce the slow component of O K N and O K A N . It could not be determined whether the effects of stimulation were due to excitation of cells in N O T and D T N or to stimulation of fibers in passage. In this study we used electrolytic and kainic acid lesions to separate the two possibilities, and to determine whether or not lesions of N O T would produce changes in either the slow or rapid components of O K N and O K A N .

Methods Ten juvenile rhesus monkeys (Macaca mulatta) of 2.5~4 kg were used in these experiments. Under anesthesia using sterile surgical conditions a magnetic scleral search coil was sutured to the left eye to record horizontal and vertical eye position (Robinson 1963; Judge et al. 1980). After bone was removed over NOT at stereotaxic coordinates A 2, L 5.5, a cylinder that accepts a microelectrode carrier (Trent Wells) was implanted on the skull. The cylinder was embedded in dental acrylic cement along with restraining bolts to fix the animal's head during experiments. Animals were treated postoperatively with antibiotics and analgesics and allowed to recover for several weeks. Vestibular and optokinetic testing before and after lesions was done in a three-axis vestibular and optokinetic stimulator. Optokinetic stimulation was provided by moving a large closed cylinder, 89 cm in diameter and 61 cm high, around the animal at a constant velocity. The cylinder had alternating 10 deg black and white stripes and filled its field of vision (see R a p h a n et al. 1981 ; Cohen et al. 1987 and Schiff et al. 1988, for details). Eye position was differentiated and rectified to obtain slow phase eye velocity. The voltages were recorded on F M magnetic tape and on a paper writer with a band width of D C to 30 Hz. Eye velocity was calibrated by rotating the animals at 30 deg/s in light. It was assumed that the eye velocity was close to rotational velocity in this condition (Skavenski and Robinson 1973). Eye movements to the right caused upward deflections in the traces shown in the figures. NOT and D T N were identified by their response to electrical stimulation in darkness (Schiff etal. 1988). M o n o p o l a r pulses of 0.5 ms and 40 microamperes at 250 Hz were delivered through a tungsten microelectrode for 20 s in darkness. Lesions were made at locations where stimulation induced typical horizontal nystagmus. The nystagmus had a slow rise, a steady

state level at or slightly below the saturation level of O K A N and was followed by after-nystagmus that was similar to O K A N (Schiff et al. 1988). Lesions were either electrolytic or were produced by injection of kainic acid (Coyle et al. 1978). For electrolytic lesions a pulse of constant current of 1 milliamp was passed for 30 seconds. Injections of kainic acid or muscimol were made through a micropipette that replaced the tungsten stimulating microelectrode, using the technique described by Cohen and BiittnerEnnever (1984). The optimal location for producing nystagmus was first determined. Then the stimulating microelectrode was removed, leaving the platform and microdrive with the guide tube in place. A micropipette filled with a 1% kainic acid solution was fitted to a Hamilton syringe, attached to the microdrive and lowered into the guide tube until its tip was at the stimulation site. Ten minutes afterwards the lesioning agent was injected at a rate of 0.1 microliter/6 ran, generally to a total of 0.4 microliters. Twenty minutes after injection the micropipette was slowly retracted and removed. In one animal (M991) a vascular lesion developed unexpectedly while electrically stimulating the superior colliculus. The monkey had oculomotor deficits relevant to this study, and its data are included. Another animal (M1176) developed a lesion after a pressure injection of 1 gL of a pharmacological agent (muscimol) into the region of NOT, after the region had first been identified by electrical stimulation. There were permanent alterations in O K N and O K A N after the injection that led us to conclude that we had damaged NOT. This was confirmed histologically, and the results of this lesion are also included in the data in this paper. O K A N and vestibular nystagmus were tested after lesions until the responses had stabilized. Parameters of interest included the velocity of the initial jump in slow phase velocity at the onset of stimulation, mean peak slow phase velocity during steady state OKN, and peak O K A N velocity at the end of stimulation. Measurements for the initial jump in eye velocity and for peak O K A N velocity were taken one to two seconds after the onset and end of stimulation, and steady state O K N values were established by averaging 5-10 slow phases with the highest velocity elicited by each stimulus velocity. These values are shown in the graphs of Figs. 6, 7 and 9. Each point represents data from one or several tests at the times indicated on the side of the graph. We were interested in best performance, and when several trials were available, the highest velocity attained for each of these parameters was chosen. At the conclusion of experiments the animals were deeply anesthetized and perfused through the heart with a 10% formalin-saline solution. After fixation the brains were removed and cut serially. Every fifth section was stained with cresyl violet. The extent of the lesion was estimated by cell loss and invasion of microglia. The lesions boundaries were sharp and were drawn on a microprojector. Weigert stains were used to determine whether there had been damage of myelinated fibers.

Results

Definition of the nucleus of the optic tract (NOT) and the dorsal terminal nucleus (DTN) There has been controversy as to the location of N O T in the monkey (see Simpson et al. 1988 for review). We define N O T as a pretectal nucleus, interstitial to the brachium of the superior colliculus. D T N lies ventral to the brachium (Lin and

227

A# D

F

Fig. 1 A-F. Upper brainstem, including NOT, DTN and the pretectum. The sections are 300 microns apart, and extend from rostral (A) to caudal (F) in this and in subsequent figures. See text for description

Fig. 2 A-D. Microscopic sections through the electrolytic lesions in M1169 and M1166, stained with cresyl violet. A-C Correspond to Fig. 3A-C, and D corresponds to Fig. 3E. The region destroyed in Ml169 in Fig. 2A, B is the approximate site of the lesion in Ml176.

Giolli 1979). These nuclei contain small, medium and large cells, and receive direct input from the retina (Ballas et al. 1981; Hutchins and Weber 1985). The large fusiform cells project to the ipsilateral dorsal cap of Kooy and the beta nucleus

of the inferior olive (Weber and Harting 1980; Hoffmann and Distler 1986). DTN and NOT are interconnected, and DTN also projects to the contralateral NOT (Holstege and Collewijn 1982; Simpson et al. 1988).

228

\

Fig. 3. Diagrams of the electrolytic lesions in Ml169 (AC) and Ml166 (D-F). The sections are separated by 300 microns in A-C, and by 400 microns in D-F. A-C The lesion in Ml169 was about 1 mm in diameter. It lay in the caudal, lateral portion of NOT, in and under the brachium of the superior colliculus (BSC), ventral

and lateral to the pretectal

MII66

MI663-81 / ~

~

~

Sections B-E of Fig. I show the location of N O T and D T N and the surrounding nuclei of the upper mesencephalon, as determined by the criteria of direct retinal input, projection to the inferior olive, and G A D staining (summarized in Cohen et al. 1989). The body of N O T extends caudally, laterally and ventrally in the brachium of the superior colliculus to the lateral edge of the brainstem (Sections B-E), where N O T lies dorsal to D T N (Sections C-E). This corresponds to the spatial coordinates of the region from which nystagmus was produced by electrical stimulation (Fig. 8 of Schiff et al. 1988). According to these criteria N O T and D T N are differentiated primarily by their location. Nucleus limitans (NL), shown by the heavy dots in Sections A and B and the pretectal olivary nucleus (PON) (Section B) are largely rostral to NOT. Lesions

Representative sections from cases which involved NOT and D T N are shown in Figs. 2-5. The elec-

"

PB

olivary nucleus (PON, A) and extended into D T N (C). D - F The lesion in Ml166 lay in the center of NOT (E) in the brachium of the superior colliculus (F), caudal to the pretectal olivary nucleus which was intact (D). The rostrocaudal extent was about 500 microns. D T N was intact (F). The pulvinar and superior colliculus were unaffected by either lesion

trolytic lesions were small (1 m m or less) and lay in the brachium of the superior colliculus. In MI 169 the lesion destroyed the central, caudal and lateral parts of N O T and most of D T N (Figs. 2 A C, 3A-C). Such a lesion would also interrupt the descending output pathways of the rostral medial N O T (D. Schiff, J.A. Biittner-Ennever and B. Cohen, unpublished data). The lesion in M1166 was confined to the central portion of N O T where it was adjacent to but did not involve the pulvinar (Figs. 2D, 3D F). The kainic acid lesions were larger. In one animal the injection site was in the central mesencephalic reticular formation (cMRF; Cohen and Bfittner-Ennever 1984; Cohen et al. 1986) (Ml122, Fig. 4-1, A-C), and in another it was in the ventral caudal thalamus and pulvinar (M1178, Fig. 4-2, A-E). In both animals there was extension into NOT and DTN. The common area of overlap in these two animals is shown in black in Fig. 4-3. It lay over NOT, which was extensively destroyed in both cases. This area of overlap is where nystagmus had been induced by electrical

229

B

C MII22 L

PB

2

MII78

3

A

B ~ 1

C

,~ ~

D

Figi4-1-3. Diagrams of kainic acid lesions in Ml122 (4-1), Ml178 Right (4-2), and the overlap between the two (4-3). 4-1 Injection into c M R F in M1122 extended upward to destroy NOT. At that level the dorsal edge of lesion reached the dorsal edge of the brachium of the superior colliculus (B). NOT was completely destroyed. The pulvinar and the fibers of the brachium of the superior colliculus were intact. 4-2 The kainic acid lesion of the pulvinar in Ml178 extended rostrally into the thalamus (A, B). In the caudal portion (D, E) the lesion reached the medial edge of the brachium of the superior colliculus. A small portion of the rostral medial pole of NOT remained and DTN was not affected (D). A portion of the medial geniculate nucleus and the small and medium sized cells in PON were also destroyed (C). 4-3 The overlap of the two kainic acid lesions, shown in black, covers most of NOT. A The square represents the injection site of kainic acid in the internal capsule in Ml164. This lesion destroyed the lateral third of the pulvinar as well as the dorsal edge of the LGn. Fascicles that join the BSC were damaged at the injection site. B The dot approximates the area of an electrolytic lesion in Ml178 Left. The lesion partially involved fibers passing through the lateral pulvinar

stimulation in a previous study (Schiff et al. 1988). Finally, in one animal (M1176 ) a lesion was made in the region of NOT by a pressure injection of a pharmacological agent (muscimol). The area destroyed corresponds to that shown in Fig. 2A,B.

Each of the preceding lesions caused substantial changes in O K N and OKAN. There were several lesions of this region that did not involve N O T and DTN: 1. Large bilateral electrolytic lesions of the central mesencephalic re-

230

B Pu,v.~? ~ - . _ z - ' ~ Y

I

~ M99I-~73

~

2

- 2 3 3

D

Fig. 5-1, 2. Vascular (5-1; M991) and kainic acid lesions (5-2; Mll12) involving comparable parts of the brainstem. The lesions involved the central gray, the caudal mesencephalic reticular formation, the medial pretectum and the rostral superior colliculus. Deficits in OKN and OKAN occurred only after the vascular lesion of 5-1. NOT and DTN were not affected by either of these lesions

Table 1

Init. jump

OKNss

OKAN

Lesion

Positive cases M 1169 M1166 Ml122 M1178 right M1176 Ml178 left M 1164

Decreased Normal Decreased Decreased Decreased Normal Normal

Strong decrease Transient decrease Strong decrease Strong decrease Strong decrease Transient decrease Transient decrease

Strong decrease Transient decrease Abolished Strong decrease Strong decrease Transient decrease Transient decrease

M991

Absent

Normal

Normal

BSC, caudal half of NOT and DTN (electrolytic) BSC and central NOT (electrolytic) MRF, NOT, DTN, SC (kainic acid) Pulvinar and NOT (kainic acid) BSC and central NOT (pressure injection) Pulvinar, BSC (electrolytic) Int Capsule, Pulvinar, LGn, input to BSC (kainic acid) SC, caudal MRF, central gray, medial pretectum (vascular)

Control cases M 1111 M 1112

Normal Normal

Normal Normal

Normal Normal

OKNss = steady state slow phase velocity, always to the ipsilateral side

cMRF (bilateral) (electrolytic) SC, caudal MRF, central gray, medial pretectum (kainic acid)

231

ticular formation (M1111, not shown), and a kainic acid lesion of the superior colliculus and caudal M R F (Ml112, Fig. 5-2, A - D ) had n o e f f e c t on O K N or O K A N . This supports the general contention that these structures have little to do with production of O K N or O K A N . 2. Two lesions involved the brachium of the superior colliculus in the pulvinar. One was electrolytic on the left side in Ml178. The extent of the damage is approximately the area covered by the dot in Fig. 4-3B. In another animal (Ml164) the fiber tract that inputs to the brachium of the superior colliculus was damaged during the course of a kainic acid injection into the internal capsule. The site of the injection is shown by the square in Fig. 4-3A. The cellular damage in this animal extended medially into the pulvinar and ventrally into the lateral geniculate. N O T was intact. These lesions affected O K N transiently, but it recovered. 3. One animal had a vascular lesion that destroyed a portion of the caudal M R F , the adjacent central gray and the rostral superior colliculus (M991, Fig. 5-1, A-C). The affected area was medial to the pretectum. This lesion affected the rapid rise in O K N , but not the slow rise or O K A N . A summary of the lesions is given in Table 1.

Effects of lesions of NO T and D T N We could not differentiate between N O T and D T N lesions in these experiments, and they will be considered together. N O T and D T N lesions affected O K N and O K A N with slow phases toward the side of the lesion, but caused no change in O K N and O K A N with contralateral slow phases. This is consistent with electrical stimulation of N O T / D T N (Schiff et al. 1988). The lesions caused no obvious postural or behavioral modifications aside from a slight head tilt in one animal (Ml166). Immediately after the N O T / D T N lesions there was spontaneous horizontal nystagmus with contralateral slow phases that subsided within several days. The lesions did not affect vertical O K N or saccadic eye movements. Vestibular nystagmus was intact, and when tested 1-2 weeks after lesion, the slow phase velocities of per- and post-rotatory nystagmus were symmetrical and of normal gain. Characteristic O K N and O K A N after a lesion of N O T and D T N on the left is shown in Fig. 6. It was recorded in Ml176 15 days after a lesion in the region of the left N O T and D T N caused by a pressure injection of muscimol. Nystagmus with slow phases to the contralateral (right) side was normal with an initial jump, a slow rise to a steady state velocity close to 60 deg/s, and pro-

longed O K A N whose peak velocity was close to the stimulus velocity (Fig. 6A). The initial jump of the nystagmus to the ipsilateral (left) side was of the same magnitude as to the contralateral side, but the nystagmus was.irregular, reaching a peak velocity of about 50 deg/s (Fig. 6B). At the end of O K N , eye velocity fell rapidly to about 1012 deg/s (arrow), and the duration of O K A N was short. To determine whether the reduction in O K A N reflected an inability of the animal to charge velocity storage under other conditions, we interacted per- and post-rotatory nystagmus with O K N and O K A N . For this the animal was rotated in light at 60 deg/s and stopped in darkness. In this condition the stored activity generated by the visual and vestibular systems during rotation in light, can be used to counter the post-rotatory response (Ter Braak 1936; Mowrer 1937; Raphan et al. 1979). The normal response is shown in Fig. 6C. Counterclockwise rotation in light elicited slow phase velocities to the right, the side contralateral to the lesion. At the onset of rotation the eyes jumped immediately to a velocity close to 60 deg/s, and this velocity was maintained for the duration of rotation. When rotation was stopped with the animal in darkness, there was a rapid jump back to zero velocity, followed by a few weak beats of postrotatory nystagmus. When the animal was rotated to the right, so that the slow phase velocities were to the left, i.e., ipsilateral to the side of the lesion (Fig. 6D), the initial jump was similar to that for nystagmus with slow phases to the right. Nystagmus velocity fell, however, and was maintained at close to 30 deg/s. At the end of rotation, when the animal was stopped in darkness, there was a vigorous post-rotatory nystagmus with slow phases to the right. The peak velocity of this nystagmus was close to 50 deg/s. The striking difference in the post-rotatory response after rotation in light to the right and left is consistent with the data of Fig. 6A, B. Since the time constant of activity coming from the semicircular canals is approximately 5 sec (Goldberg and Fernandez 1971), by 15 18 seconds after the onset of rotation, the nystagmus of Fig. 6C, D must have been sustained b y activity generated through the visual system. The failure of the animal to generate adequate velocity storage to counter the post-rotatory response in Fig. 6 D suggests that the slow phase velocities toward the end of the period of rotation in light were produced primarily by the direct pathway. The findings in this animal are summarized in the graphs of Fig. 6 E G . The results show tests

232

A

OKN

/

OKAN

SP VEL

YAW POS PHOTO CELL

I :360 __ -: - - : ~ .

.:...- 9 _ :

......

o

DARK

LIGHT 10 see

B

OKN

/

ROTATION IN LIGHT

C

OKAN

STOP IN DARKNESS

" ~ ~ ~ ~ ' ~ ' ~ ' ~ ' ~ ' ~

I 60 ~

r - ~ . j - - . . . j - ~ ~

l ar0 o DARK

DARK u

LIGHT

D

LIGHT

ROTATION IN LIGHT

STOP IN DARKNESS

l a6o ~ DARK LIGHT

E

F

G OK,A N

STEADY STATE OKN 140

140

120

120

120

100

100

80

80

,o

._J LM >

UJ >LU

60

,(~/'o. / "'~

60

40

~_4

/...

/

o ---0-......o.... 9

80

d./ .... /'~5"

/, 2O

/

o D/

-o

50

100

150

200

0

50

100

p'/o' 9

9

-

_

_

20

"-,o/I

~::.9....,,...t..a.....:.=...........y, 0

l~aPre tt Pre RIP01 Lt~01 ~apo2 tlPo2

p

60

40

20

Jump

100

/ o" o/D..

//

0

Initial

140

150

200

0

so

~oo

1io

200

STIMULUS VELOCITY (deg/s)

Fig. 6A-G. Optokinetic nystagmus induced by surround rotation at 60 deg/s in Ml176. 15 days after a left sided NOT lesion. O K N and O K A N to the right (A) were normal. O K N to the left (B) had a good initial jump, but the steady state velocity was irregular, and was followed by weak O K A N (arrow). C, D Vestibular and optokinetic nystagmus induced by rotation at 60 deg/s in light. After rotation that elicited nystagmus with right slow phase velocities, there was little or no after-nystagmus when the animal was stopl~ed in darkness (C). After rotation that elicited left slow phase velocities, however, there was strong after-nystagmus, confirming that the level of O K A N that could be induced to the left was significantly reduced. The first trace in each panel is slow phase eye velocity (SP Vel), the second trace is yaw position about a vertical axis and the third is a photocell showing light and darkness. The thickened portion of the photocell trace represents the response to the passage of stripes, 10 deg apart. The downward deflection of the photocell trace at the end of the period of optokinetic stimulation signifies the onset of darkness. E - G Graphs of steady state velocity (E), peak O K A N (F) and the initial jump (G) recorded on three test occasions, before (Pre), one (POJ) and two months (P02) after lesion. Data from the prelesion test are connected by solid lines, and data from post-lesion tests are connected by the dashed and dotted lines, respectively. Slow phase velocities to the right are shown by the open symbols and slow phase velocities to the left, the lesion side, are shown by the filled symbols. Note the effect of lesion on the O K N steady state velocity (E) and the O K A N peak velocity (F). Effects on the initial jump (G) were irregular, although there was a consistent reduction on several occasions, similar to that shown by the filled squares

233

Findings were similar in Ml169 and Ml166. Their results are summarized in Fig. 7A, B. The peak velocity of steady state O K N was reduced in both animals for stimulus velocities above 30 deg/s (Fig. 7 A 1, 7 B 1, filled circles). O K A N peak velocities fell from 55 and 58 deg/s to 15 and 18 deg/s in the two animals (Fig. 7A2, 7 B 2). The initial jump in O K N was reduced in Ml169 (Fig. 7A3), but was unaffected in M1166 (Fig. 7 B 3). M1166 had recovered when tested two weeks later (Fig. 7 B, squares), probably due to the small size of the lesion. There was no recovery in M1169 at the last test about three weeks after lesion. The kainic acid lesions of M1122 and M1178 (Fig. 4) also produced striking changes in O K N and OKAN. The lateral semicircular canals had been plugged earlier in Ml122, resulting in an enhanced initial jump in slow phase velocity at the onset of O K N and some habituation of the dominant time constant of OKAN, but its OKN and O K A N were otherwise unaffected. Kainic acid, injected into the left cMRF, reached N O T and D T N by upward extension (Fig. 4-1). OKN and O K A N with slow phases to the right were unaffected. O K N with leftward slow phases recorded preoperatively and 50 days after lesion are shown in Fig. 8 A, B and summarized in Fig. 7 C 1-C 3. After lesion (Fig. 8 B) the initial jump in O K N was somewhat lower than the preoperative value (Fig. 8A, 7 C 3), but was close to the initial jump before the

from 3 test sessions, one before and two, one and two months after the lesion had been made. Each data point is the mean of several tests and shows peak performance. Similar results were recorded on other occasions in this animal after the lesion, but the three test sessions are representative and will show the essential features of the changes. The most striking change was a loss of O K A N with slow phase velocities to the ipsilateral side (Fig. 6 F, filled circles and filled squares, arrow). This was present on eight test occasions over a period of three months, and it can be considered as a permanent reduction in OKAN. Reflecting this, the steady state velocity was lower to the lesion than to the normal side (Fig. 6 E), particularly at higher velocities of stimulation. The peak velocity of O K N with slow phases to the lesion side varied considerably, and in some instances was as high as 40-50 deg/s, as in Fig. 6 B. Since there was little O K A N for nystagmus to the lesion side, i.e., little velocity storage, the steady state velocity must have been produced primarily by activation of the direct pathway. The initial jump in velocity was variable for movements to either side (Fig. 6G). More commonly, the velocity of the initial jump was less when the eyes moved to the lesion than to the normal side (Fig. 6 G, filled squares). Values of the initial jump for this animal are close to those that have been reported previously (Matsuo and Cohen 1984).

10KNss A .-~

/'-..

"~ 30

30

"

60

90

"

" (~/S) ~30

"'o.

~O t ~60

""'" '%0,.,........o

30

OKAN

30

60

uJ

30

90 (~'s)

"'"'"ul.

"".... ....

,..::

o 30 / "~ {

3o

:

6o

.

& 90J MII22

"'... "'"'o

9

9

30

. (~'s)

r" 30

~o

60 t

9oJ

o-o pre lesion o e18 days lesion

9

~

"

60 "

60

......

90 .......

9

301

~

pre lesion

$,ov ]

; ; ~days postlesion

u~ 90 J

i-u72 days post lesion

30 60 [~,

n Cn 6O

pre lesion {Day 281-86) 18days post lesion I - n 1,1 days post lesion O-O

to

12G

120

Fig. 9A, B. Graphs showing effects of an electrolytic lesion in the brachium of the superior colliculus in the left pulvinar of Ml178, Left (A), and a kainic acid lesion of the right pulvinar M1178, Right (B). Responses to the ipsi- and contralateral side are shown for each lesion. The scheme is the same as in Fig. 8. Note the transient effect of the lesion in A, and the permanent change of steady state O K N and O K A N in B

rapid component of OKN in M991 can be contrasted to the findings in M1122 shown in Fig. 8B where the predominant effect was on the slow component. Within a few weeks the animal had recovered its ability to accelerate its eyes rapidly at the onset of stimulation (Fig. 8D, arrow). OKN and O K A N with contralateral slow phases were not affected (Fig. 8 E). 4. A kainic acid lesion that destroyed cells in the same area (Fig. 5-2), i.e., in the superior colliculus, caudal MRF and adjacent central gray matter, had no effect on O K N or OKAN. The failure of kainic acid to affect OKN in a similar fashion as a lesion that destroyed both cells and fibers, suggests that a fiber pathway located in the caudal MRF carries activity for the initial jump in OKN.

Discussion The data show that lesions that damaged NOT and D T N in the monkey had a strong effect on OKN and O K A N with ipsilateral slow phases. This occurred after small electrolytic lesions, localized to NOT and DTN, which destroyed cells and fibers, as well as after larger kainic acid lesion that only destroyed cells but extended well beyond the boundaries of N O T and DTN. Lesions of structures outside NOT and DTN, with several exceptions, had no effect on OKN and OKAN. The negative lesions included the superior colliculus, the cMRF and parts of the pulvinar. It seems clear that N O T / D T N are an important part of the pathway from the retina and visual cortex that pro-

236

duces ipsilateral slow phases during O K N and OKAN. This is consistent with findings of Kato et al. (1986) in the monkey, as well as with previous studies in the rabbit (Collewijn 1975a), rat (Cazin et al. 1980) and cat (Hoffmann 1982). The lesion data are also consistent with the findings from our previous stimulation study in indicating that each side apparently functions relatively independently in producing ipsilateral slow phase velocities. There was little evidence of interaction between the two sides in either of our studies. We take this as justification for using each side of an animal as a separate experiment. O K N and O K A N can be simulated by two processes (Cohen et al. 1977; Raphan et al. 1979). One has rapid dynamics and produces rapid changes in slow phase eye velocity. It causes an initial jump in eye velocity at the onset of stimulation and a rapid fall in velocity at the end of stimulation. The second process is dependent on an integrator, postulated to be in the vestibular system, that is responsible for O K A N and for the slow component of OKN. It has sluggish dynamics, causing only slower changes in eye velocity. The first process has been called the " d i r e c t " pathway, and the second the "indirect pathway", and its integrator, the "velocity storage" integrator. (Note that the terms, " d i r e c t " and "indirect" come from modeling and have no anatomical implications as to the number of synapses in the pathway.) The model incorporating these processes simulates O K N and O K A N over a wide variety of conditions (Waespe et al. 1983; Raphan and Cohen 1985; Cohen et al. 1987). It has been possible to identify separate anatomical structures as being involved in producing one or the other components of the optokinetic response that compose the various processes of the model. The direct pathway has been linked to the visual cortex (Zee et al. 1987), pontine nuclei (May et al. 1988) and flocculus (Waespe et al. 1983) and the velocity storage integrator to the vestibular nuclei (Cohen et al. 1977; Waespe and Henn 1977; see Raphan and Cohen 1985 for review) and the nodulus and uvula (Waespe et al. 1985). The results of electric stimulation of N O T and DTN, in which only slow changes in eye velocity were induced, suggested that N O T and D T N are part of the indirect pathway, processing information for the slow component of O K N and for O K A N (Schiff et al. 1988). Findings in the present study are consistent with this idea. O K A N and the slow rise of O K N were strongly affected in every instance in which lesions involved N O T and DTN. In Ml169 and

Ml176 there was a persistent reduction in O K N and O K A N after small N O T / D T N lesions. The rostral half of NOT was preserved in M1169, but by its location the lesion would have interrupted the output pathway from the medial portion of NOT, as well as destroying the caudal and lateral parts of N O T and D T N (D. Schiff, J.A. BfittnerEnnever and B. Cohen, unpublished data). In one animal, Ml122, O K A N was permanently abolished without loss of the rapid component (Fig. 8B, 7C2). In 1178 Rt, O K A N fell to about a quarter of its initial value (Fig. 9 B 2). This was paralleled by a similar fall in steady state eye velocity (Figs. 9 B 1), presumably due mainly to a loss of velocity storage. Findings were similar in 1166, after a small N O T lesion, but here the effects were transient (Fig. 7 B 1, 7 B 2), probably due to the size of the lesion. In this animal the loss of the slow component occurred without any change in the initial jump of O K N (Fig. 7 B 3). An objection to this conclusion might be that the lesions were somewhat disparate, particularly the kainic acid lesions, which only had a small area of overlap. However, small lesions of N O T and D T N in three animals (Ml169, M1176 and Ml166) caused similar effects as the larger kainic acid lesions that involved these and surrounding structures (M1122 and 1178 Rt). Moreover, O K N and O K A N were affected in a stereotyped fashion whenever NOT or D T N were involved, and similar findings did not occur when the surrounding structures were lesioned without involving NOT or D T N or their input pathways. This conclusion is also consistent with the effects of electrical stimulation: There was a striking congruence between the area identified by electric stimulation (Schiff et al. 1988) and the region, which when lesioned, caused the greatest deficits in the slow component of O K N and O K A N (Fig. 4-3). Although the initial jump was generally diminished after N O T / D T N lesions, it was still possible for the animals to alter eye velocity rapidly, and in several animals (Ml176, Ml166 and M1122) the response of the direct pathway was able to compensate for the loss of the slow component in producing steady eye velocities close to those of the stimulus (Fig. 6B; Fig. 7 B I, C 1). This suggests that the major effect of these lesions was on the slow component of OKN, i.e., on the indirect pathway, and that the ability to generate rapid changes in eye velocity was still largely intact. This is somewhat different than the finding of Kato et al. (1986) who originally concluded that all components of O K N were processed in NOT. Their lesions were larger and more rostral. In addition,

237

the species of monkey was different (Macaca fascicularis vs Macaca mulatta), and the area of damage was not specified in some cases. More recently Kato et al. (1988) have reversed their original conclusion and now believe that only the slow component of O K N is processed in NOT. In contrast to the preferential effect of NOT/ D T N lesions on the slow component of OKN and on OKAN, a vascular lesion of the medial pretectum in M991 affected only the rapid component of OKN, leaving the slow component intact. The slow rise in velocity was prolonged after lesion (Fig. 8 C), similar to the prolongation of the slow rise after the loss of the direct pathways after flocculectomy (Zee et al. 1981; Waespe et al. 1983), ablation of the visual cortex (Zee et al. 1987; Cohen et al. 1989) or destruction of the dorsolateral pontine nucleus (May et al. 1988). The long slow rise after removal of the direct pathway can be explained by increased retinal slip plus the effect of a nonlinear processing element that is present in visual-oculomotor pathways, probably at the level of NOT (Waespe et al. 1983). The effects of the vascular lesion in M991 can be contrasted to the effects of a kainic acid lesion of the same region in M1112 which only destroyed cells (Fig. 5-2). There was no effect of the latter on O K N or OKAN. Taken together this suggests that that activity responsible for the rapid component of O K N is probably carried in a fiber tract from MT and MST that courses through the M R F just at the lateral margins of the central gray, anterior to the superior colliculus. Presumably it reaches the dorsolateral pontine nucleus (May et al. 1988) before projecting to the region of the flocculus and paraflocculus. The transient nature of the deficit in the initial jump in O K N slow phase velocity produced by the direct pathway in M991 is consistent with the transient nature of the pursuit deficits reported by Newsome et al. (1985) after MT and MST lesions. As noted, there is both functional and anatomical justification for separating direct and indirect optokinetic pathways. Therefore, it was surprising that the .rapid component of O K N was somewhat diminished in almost every animal that had a lesion " that involved NOT and DTN. At present we do not have a satisfactory explanation for this, since rapid changes in slow phase eye velocity were never induced by stimulation of NOT and D T N (Schiff et al. 1988). It is possible that NOT and D T N may make a contribution to activity in the direct pathway. In support of this, neurons responding to rates of retinal slip as high as 100 deg/s have been described in the N O T of the monkey (Hoffmann

and Distler 1986). Alternatively, activity in these nuclei may indirectly modulate the gain of the direct pathway without being primarily responsible for generating the rapid jump at the onset of OKN. If the indirect pathway lies in NOT, then it is possible to consider how NOT activity is utilized in producing the slow component of O K N and OKAN. The target neurons for the " O K N O K A N " projection from N O T are likely to be Type 1 and 2 cells in the vestibular nuclei that receive input from the lateral semicircular canals. As shown by Waespe and Henn (1977) the firing rates of these vestibular neurons reflect only the slow component of O K N and O K A N and are not related to the rapid component. This indicates that these cells get their visual input primarily through velocity storage, which in turn is activated through NOT. Powerful projections extend from N O T to the inferior olive (Kawamura and Onodera 1984; Maekawa et al. 1984; Sekija and Kawamura 1985; Horn and Hoffman 1987), but the firing rates of the olivary neurons that receive these projections are probably too low to be capable of generating slow phase velocities in the vestibular nuclei. Instead, the NOT projections that reach the vestibular nuclei either directly (Holstege and Collewijn 1982) or indirectly through the prepositus nucleus (Magnin et al. 1983) are likely to be responsible for producing slow phase velocities over the indirect pathway during O K N and OKAN. Whether the projections to the various target areas are separate or are axon collaterals of the NOT-IO projection neurons remains to be determined. The source of the cortical input to NOT is still unknown. Characteristic nystagmus with only a slow component and after-nystagmus was induced by electrical stimulation of the brachium of the superior colliculus or of the fiber tract in the lateral pulvinar that projects to the brachium. Lesions of the brachium of the superior colliculus and of this fiber tract produced similar effects: the rising time constant of O K A N was slowed, and the maximum velocity of steady state O K N and O K A N was transiently reduced. The rapid rise in O K N was unaffected in these animals. This suggests that cortical information, which makes the major contribution to activity in NOT (Hoffmann 1986), may reach NOT through this fiber bundle and the brachium of the superior colliculus. In summary, we have provided evidence from stimulation and lesions experiments that NOT and possibly D T N process activity that reaches the velocity storage integrator in the vestibular system to produce the slow component of O K N and for OKAN. The direct pathway appears to be carried

238

elsewhere, although NOT and D T N may have a modulatory influence on rapid changes in eye velocity during OKN. The importance of the pathway through NOT and D T N has not been widely appreciated in the primate, but if movement of the entire visual field occurs mainly during head movement on the body or during angular body movement in space, then the N O T / D T N pathway may provide the visual information that signals head on body or body in space velocity to the vestibular nuclei. Acknowledgements. Supported by NIH

research grant EY02296, NEI Core Center grant 01867, training grant NS07245, and SFB 220/D 8.

References Ballas I, Hoffmann KP, Wagner HJ (1981) Retinal projection to the nucleus of the optic tract in the cat as revealed by retrograde transport of horseradish peroxide. Neurosci Lett 26:197-202 Cazin L, Precht W, Lannou J (1980) Pathways mediating optokinetic responses of vestibular nucleus neurons in the rat. Pfliigers Arch 384:19 29 Cohen B, Matsuo V, Raphan T (1977) Quantitative analysis of the velocity characteristics of optokinetic nystagmus and optokinetic after-nystagmus. J Physiol (Lond) 270:321-344 Cohen B, Biittner-EnneverJA (1984) Projections from the superior colliculus to a region of the central mesencephalic reticular formation (cMRF) associated with horizontal eye movements. Exp Brain Res 57:167-176 Cohen B, Matsuo V, Fradin J, Raphan T (1985) Horizontal saccades induced by stimulation of the central mesencephalic reticular formation. Exp Brain Res 57:605-616 Cohen B, Waitzman D, Bfittner-Ennever JA, Matsuo V (1986) Horizontal saccades and the central mesencephalic reticular formation. Brain Res 64:243-256 Cohen B, Helwig D, Raphan T (1987) Baclofen and velocity storage: a model of the effects of the drug on the vestibuloocular reflex. J Physiol (Lond) 393 : 703-725 Cohen B, Schiff D, Biittner-Ennever JA (1989) Contribution of the subcortical visual systems to optokinetic nystagrous and optokinetic after-nystagmus. In: Cohen B, BodisWollner I (eds) Vision and the brain: organization of the central visual system. Raven Press, New York, pp 233~55 Collewijn H (1975a) Direction selective units in the rabbit's nucleus of the optic tract. Brain Res 100:48%508 Collewijn H (1975 b) Oculomotor areas in the rabbit's midbrain and pretectum. J Neurobiol 6: 3 22 Coyle JT, Molliver ME, Kuhar MJ (1978) In situ injection of kainic acid : a new method for selectively lesioning neuronal cell bodies while sparing axons of passage. J Comp Neurol 180:301-324 Goldberg JM, Fernandez C (1971) Physiology of peripheral neurons innervating semicircular canal of the squirrel monkey. I. Resting discharge and response to constant angular acceleration. J Neurophysiol 34:635-660 Hess BJM, Precht W, Reber A, Cazin L (1985) Horizontal optokinetic ocular nystagmus in the pigmented rat. Neuroscience 15:97 107 Hoffmann KP (1982) Cortical versus subcortical contributions to the optokinetic reflex in the cat. In: Lennenstrand G,

Zee D, Keller E (eds) Functional basis of ocular motility disorders. Pergamon Press, Oxford, pp 303-310 Hoffmann K-P (1986) Visual inputs relevant for the optokinetic nystagmus in mammals. Prog Brain Res 64:75-84 Hoffmann KP, Distler C (1986) The role of direction selective cells in the nucleus of the optic tract of cat and monkey during optokinetic nystagmus. In: Keller EL, Zee DS (eds) Adaptive processes in visual and oculomotor systems, vol 57. Pergamon Press, Oxford/New York, pp 261-266 Horn AKE, Hoffmann KP (1987) Combined GABA-immunochemistry and TMB-HRP histochemistry of pretectal nuclei projecting to the inferior olive in rats, cats and monkeys. Brain Res 409 : 133-138 Holstege G, Collewijn H (1982) The efferent connections of the nucleus of the optic tract and the superior colliculus in the rabbit. J Comp Neurol 209:139-175 Hutchins R, Weber JT (1985) The pretectal complex of the monkey: a reinvestigation of the morphology and retinal terminations. J Comp Neurol 232:425~442 Judge SJ, Richmond BJ, Chu FC (1980) Implantation of magnetic scleral search coils for measurement of eye position: an improved method. Vision Res 20:535-548 Kato I, Harada K, Hasegawa I, Igarashi T, Koike Y, Kawasaki T (1986) Role of the nucleus of the optic tract in monkeys in relation to optokinetic nystagmus. Brain Res 364:1~22 Kato I, Harada K, Hasegawa T, Ikarashi T (1988) Role of the nucleus of the optic tract of monkeys in optokinetic nystagmus and optokinetic after-nystagmus. Brain Res 474:16-26 Kawamura K, Onodera S (1984) Olivary projections from the pretectal region in the cat studied with horseradish peroxidase and tritiated amino axonal transport. Arch Ital Biol 122:155-168 Lin H, Giolli RA (1979) Accessory optic system of rhesus monkey. Exp Neurol 63:163-176 Maekawa K, Takeda T, Kimura M (1984) Responses of the nucleus of the optic tract neurons projecting to the nucleus reticularis tegmenti pontis upon optokinetic stimulation in the rabbit. Neurosci Res 2:1-25 Magnin M, Courjon JH, Flandrin JM (1983) Possible visual pathways to the cat vestibular nuclei involving the nucleus prepositus hypoglossi. Exp Brain Res 51 : 198-203 Matsuo V, Cohen G (1984) Vertical optokinetic nystagmus and vestibular nystagmus in the monkey: Up-down asymmetry and effects of gravity. Exp Brain Res 53:197-216 May JG, Keller EL, Suzuki DA (1988) Smooth pursuit eye movement deficits with chemical lesions in the dorsolateral pontine nucleus of the monkey. J Neurophysiol 59 : 952-977 Mowrer OH (1937) The influence of vision during bodily rotation upon the duration of post-rotational vestibular nystagmus. Acta Otolaryngol 25:351-364 Newsome WT, Wurtz RH, Dursteler M, Mikami A (1985) Deficits in visual motion processing following ibotenic acid lesion of the middle temporal visual area of the macaque monkey. J Neurosci 5:825-840 Raphan T, Matsuo V, Cohen B (1979) Velocity storage in the vestibulo-ocular reflex arc (VOR). Exp Brain Res 35:229248 Raphan T, Cohen B, Henn V (1981) Effects of gravity on rotary nystagmus in monkeys. Ann NY Acad Sci 374:44-55 Raphan T, Cohen B (1985) Velocity storage and the ocular respone to multidimensional vestibular stimuli. In: Berthoz A, Melvill Jones G (eds) Adaptive mechanisms in gaze control; facts and theories. Elsevier, Amsterdam, pp 123-143 Robinson DA (1963) A method of measuring eye movements using a seleral coil in a magnetic field. IEEE Trans Biomed Eng 10:137-145

239 Schiff D, Cohen B, Raphan T (1988) Nystagmus induced by stimulation of the nucleus of the optic tract in the monkey. Exp Brain Res 70:1 14 Sekija H, Kawamura K (1985) An HRP study in the monkey of olivary projections from the mesodiencephalic structures with particular reference to pretecto-olivary neurons. Arch Ital Biol 123:171 183 Simpson JI, Giolli RA, Blanks HI (1988) The pretectal nuclear complex and the accessory optic system. In: BfittnerEnnever JA (ed) Neuroanatomy of the oculomotor system. Elsevier, Amsterdam, pp 335-364 Skavenski AA, Robinson DA (1973) Role of abducens neurons in vestibulo-ocular reflex. J Neurophysiol 36: 724-738 Ter Braak JWG (1936) Untersuchungen fiber optokinetischen Nystagmus. Arch Neerl Physiol 21:309-376 Waespe W, Henn V (1977) Neuronal activity in the vestibular nuclei of the alert monkey during vestibular and optokinetic stimulation. Exp Brain Res 27:523-538

Waespe W, Cohen B, Raphan T (1983) Role of the flocculus and paraflocculus in optokinetic nystagmus and visuovestibular interactions: effects of lesions. Exp Brain Res 50:9-33 Waespe W, Cohen B, Raphan T (1985) Dynamic modification of the vestibulo-ocular reflex by the nodulus and uvula. Science 228:199 202 Weber JT, Hatting JK (1980) The efferent projections of the pretectal complex: an autoradiographic and horseradish peroxidase analysis. Brain Res 194:1-28 Zee DS, Yamazaki A, Butler PH, Gucer G (1981) Effects of ablation of flocculus and paraflocculus on eye movements in primate. J Neurophysiol 46 : 878-899 Zee DS, Tusa RJ, Butler PH, Herman SJ, Gucer C (1987) Effects of occipital lobectomy upon eye movements in primates. J Neurophysiol 58 : 883-901 Received October 12, 1988 / Accepted July 21, 1989