Brain Research, 130 (1977) 239-252 .,c) Elsevier/North-Holland Biomedical Press
239
V E R T I C A L EYE M O V E M E N T R E L A T E D U N I T ACTIVITY IN T H E R O S T R A L M E S E N C E P H A L I C R E T I C U L A R F O R M A T I O N OF T H E A L E R T M O N K E Y
U. BI]TTNER, J. A. B(JTTNER-ENNEVER and V. HENN Department of Neurology and Brain Research hTstitute, University of Zfirich, Ziirich (Switzerland)
(Accepted November 3rd, 1976)
SUMMARY Eye movement related unit activity was recorded in the rostral mesencephalic reticular formation ( M R F ) of the alert monkey. Most units (78 out of 117) were activated with a short activity burst starting before the eye movement and were otherwise silent. The activity was the same whether movements occurred spontaneously in the light or dark, or were fast phases of vestibular or optokinetic nystagmus, and could be related to parameters of a vector representing the eye movement such as amplitude, position changes along certain planes or direction of movements. Units coding position changes or direction of movement had their preferred direction always close to the vertical. Other units (18 out of 117) showed some tonic activity, which was also only related to vertical eye position. It is suggested that this region of the rostral M R F acts as an immediate supranuclear structure, mediating eye movements in the vertical plane.
INTRODUCTION Paralysis of vertical or horizontal eye movements can occur independently in man and monkeys. This implies that there are separate anatomical sites for the generation of vertical and horizontal eye movements. However, as many movements have a vertical and horizontal component, the execution of both these components must be time-locked, and hence these two areas must be functionally connected. Clinical evidence has shown that bilateral lesions in the medial mesencephalic tegmentum rostral to the 3rd nucleus lead to impairment of vertical eye movements 7. Lesion experiments in the monkey are in good agreement with these observations zS. There is also anatomical6,3z and electrophysiologica131 evidence that the interstitial nucleus of Cajal, which lies in this area, is directly connected to the oculomotor nuclei involved in vertical eye movements.
240 A necessary condition for the hypothesis that eye movements ii~ tt~e vertica! plane are generated in the mesencephalic reticular formation (MRF) is tts~;~tthe unil activity found there leads eye movements and can be quantitatively related to the parameters of eye movement. So far there is little evidence for this available2~, e:.. However several lines of evidence suggest that the paramedian pontine reticular formation (PPRF) is the immediate supranuclear structure responsible fol° eye movements in the horizontal planeLg,to,'-'0,".L This report presents evidence that unit activity in the rostral M R F ~>t"the alert monkey is indeed related to vertical eye movements, and that the strategies of quar> titative analysis applied to units in the PPRF, which allowed the prediction of single eye movements .7, are also suitable for analyzing activity in the rostral M R F METHODS A metal ring, that could accept a microdrive, was chronically implanted over a trephine hole in the skull of 3 juvenile rhesu; monkeys (Macaca mulatto). Extracellular unit potentials were recorded with tung;ten microelectrodes of 2-8 MQ. The horizontal position of the eye was measured with DC silver-silver chloride electrodes a inserted at the outer canthi of the boney orbit. In order to measure ver~icat eye position, altogether 4 DC-electrodes were implanted above and below each eye, the upper and lower pair being connected. This arrangement minimizes cross-coupling. Eye movements were calibrated by measuring slow-phase velocity during whole field optokinetic stimulation. During recording sessions the monkey sat upright in a primate chair with its head rigidly fixed to a headholder. To achieve a constant level of alertness the monkey usually received small doses of amphetamine (max. dose, 0.3 mg/kg i.m.). Unit recordings were taken while the monkey made spontaneous eye movements in the dark and the light. In addition horizontal nystagmus was induced either optokinetically (rotation of a striped cylinder around the stationary monkey) or by vestibular stimulation (rotation of the monkey on a servocontrolled turntable in the dark). Single unit activity, horizontal and vertical eye position, light-on and -off signal, the cylinder and turntable velocity, and a digital time code were stored on an FM tape recorder. At the end of the experiments a selected recording site was marked by electrolytic lesions or by injection of tracer materials. Animals were perfused under deep anesthesia. Frozen sections of the midbrain, taken every 320/~m, were stained with cresylecht violet and recording sites were identified. Quantitative analysis was performed using the same computer programs as in earlier studies 16,17. At least 120 sec of consecutive eye movements (sample rate 1.25 kHz each channel) and unit data (sample rate 20 kHz) were stored on digital magnetic tape. The beginning and end of eye movements were marked. Blinks were marked separately, and excluded from further analysis. Statistical programs related unit activity to parameters of movement and to eye position. As in an earlier study of PPRF units, eye movements will be described using vectors. A vector is defined by an amplitude (A) and a direction (a) (Fig. 1). An im-
241 110 °
90 °
70 °
d--pos --'---
18o_ °
~
I
/', g
o o
270 ° Fig. 1. Schematic drawing of an eye movement represented by a vector, characterized by an amplitude (A) and a direction (u). The vector projection of this eye movement on the planes 70° (d-pos 70') and 110' (d-pos 110) are demonstrated. ~ 0' is defined as right, a ~ 90~ as up, ct 180~ as left, and ~ 270 as down. In subsequent calculations unit activity will systematically be related to each of these parameters, A unit coding position changes along the direction of 70 would be related to the other parameters by d-pos 70c A - cos(a-70 ). In this equation 70 is subtracted from ~, because u is always measured in reference to the horizontal right direction and not in reference to the ondirection of the units. p o r t a n t derivative of these parameters is the vector projection (d-pos) onto specific planes. Two examples for the planes 70 ° (up, right), and 110 ° (up, left), are shown in Fig. 1. The vector projections in these planes are related to amplitude and direction by d-pos 110,~ A • cos (a-110 °) or d-pos 70o -- A • cos (ct-70°). In the analysis unit activity before and during eye movements ( n u m b e r of spikes per burst and mean frequency of each burst) was systematically related to each of these 3 parameters: amplitude (A), angle of m o v e m e n t (a), and position changes along particular planes (d-pos). R ES U LTS Recording sites
All units were recorded in the rostral M R F in a n area anterior to the oculom o t o r nucleus, dorsal to the red nucleus and ventral to the thalamus. Laterally this area extends from a b o u t 1 m m from the midline to the fields of Forel and the zona incerta. It is centered a r o u n d stereotaxic coordinates A 8.0, L 2-3 (ref. 26) and extends in the stereotaxic plane 1-2 mm ventrodorsal[y, 2 m m mediolaterally and 3 mm anteroposteriorly. This area coincides mainly with the most rostral part of the M R F anterior to the nucleus of Darkschewitsch and the interstitial nucleus of Cajal. It is also the projection site of a pathway arising in the ipsilateral P P R F 5. Fig. 2 shows the localization of two lesions from two different monkeys in this region. Useful landmarks were the third nerve rootlets which were encountered 3-4 m m ventrally (see
242
Fig. 2. Section of brain taken in the stereotaxic transverse plane at A 8 and stained ailh cre.sy/ech~ violet, showing the electrode track from the final recording session in which an efecirolyfic lesh~ was made in the area were burst units were found. The arrow indicates the location of a similar poiiH, marked in a second monkey, and lying in exactly the same location. LGN, lateral genicH!ate mlcletx,;: MRF, mesencephalic reticular formation ; N I1 I, 3rd nerve rootlets ; RN, red nucleus : St. st~bthalanms : VL, n. ventralis lateralis; VPI, n. ventralis posterior inferior: V PL, n. ventralis poste.~ior lateral is: ZI, zona incerta. Magnification 3.5.
Fig. 2). The m o s t p o s t e r i o r recordings were t a k e n at the level o f the a n t e r i o r pole o f the interstitial nucleus o f Cajal. N o records were taken as far caudal as the nuclei o f the p o s t e r i o r commissure.
Unit activity M o s t units in the a r e a described a b o v e altered their activity with s p o n t a n e o u s eye m o v e m e n t s . A b o u t 30 o//o d i d not, and will not be considered further. The results are based on the analysis o f 117 units, all on the right side. O f these 117 units, 78 were b u r s t units, 18 tonic units and the r e m a i n i n g 21 fell in o t h e r categories. Tonic and b u r s t units were intermingled, a n d could sometimes be r e c o r d e d simultaneously.
Burst units ( N ~- 78) Burst units were active only with saccades or fast phases o f nystagmus. O t h e r wise they were silent, even d u r i n g slow phases o f o p t o k i n e t i c or vestibular nystagmus. W h e n m o n k e y s b e c a m e d r o w s y a n d m a d e slow drifting eye m o v e m e n t s without saccades, these units often h a d irregular activity o f 10-20 imp./sec. This was t a k e n to indicate t h a t units were inhibited d u r i n g gaze in the alert state. Saccade related bursts consisted o f 1-40 spikes which e n d e d a p p r o x i m a t e l y at
243 the end o f a saccade. The highest frequencies usually occurred at the very beginning o f the burst and fell monotonically• There were no differences in activity d u r i n g eye m o v e m e n t s in light or dark. Activity patterns were also i n d e p e n d e n t o f the starting p o s i t i o n o f the eye and the hemifield in which the eye movements were executed. Burst units could be divided into two m a i n g r o u p s : Those activated with eye movements in all directions (Fig. 3A), a n d those activated with eye movements in a preferred direction. Activity o f the first 57 units was analyzed quantitatively and the r e m a i n i n g 21 units qualitatively.
Units coding amplitude of eye movements ( N = 12 of 57) Activity o f a burst unit which e n c o d e d the a m p l i t u d e o f eye m o v e m e n t s is shown in Fig. 3A. There was a short b u r s t o f activity which started j u s t before each saccade. The n u m b e r o f spikes in consecutive bursts could be related to the a m p l i t u d e o f eye m o v e m e n t (Fig. 3B). In a few units there was a large v a r i a t i o n in this relationship, a n d only averaged d a t a showed a clear correlation• M e a n frequency o f firing in single bursts had no consistent relation to any o f the other m o v e m e n t p a r a m e t e r s tested (direction or d-pos).
Units with preferred directions of activity ( N -- 45 of 57) In these units the preferred direction was close to the vertical plane. N o units were f o u n d with on-directions in the h o r i z o n t a l plane. Twenty units fired m a x i m a l l y with upward, and 25 with d o w n w a r d movements. These units were intermingled and sometimes could be recorded simultaneously. M o s t units fired only with m o v e m e n t s in their respective on-directions and were otherwise silent• Others fired with every eye movement, and the n u m b e r o f spikes or frequency o f the burst was m o d u l a t e d with
A
B
++ t
L Illlt
_iLl
-
H-
i
[deg]
60
v
40.
,.--,...r-"-----~ 30 °
.:, . : . ; : : •
E
•
v ....
~_~,__
"i
.°
:::,ti;|.": w.* w •
.t
''. •
."
•
_d_,~-~.~
2985 500
msec
•
• ":1q1i .•~' . . . . ~2
0
1'0
s
20 30 number of spikes
Fig. 3. A : original record of a unit exhibiting a short activity burst with every saccade during spon+ taneous eye movements in the light. The quantitative analysis showed that this unit was coding the amplitude of the eye movement by the number of spikes. H = horizontal, and V vertical, eye position. Right and upward movements are indicated by upward deflection. B: computer analysis of a unit coding the amplitude of the eye movement by the number of spikes, independent of the direction of the eye movements. The values of 278 consecutive eye movements into all directions are displayed.
244 [degree] +40-
0
+20i
0-
.-i,''
.'.;~ e~"." : ,.
.
"[i.
CZ
20-
if"
40-
370 4
6
2'0
3'o
[number of spikes]
Fig. 4. Quantitative analysis of one unit coding position changes in the plane of I l0 (d-pos l l0 ) through the number of spikes per burst. Data from 253 consecutive spontaneous eye movements, in all directions, are displayed. For each eye movement the movement-component in the 110° direction (ordinate) was calculated and related to the number of spikes per burst (abscissa). Note that there a re also a few spikes with eye movements in the opposite direction (290°; i.e. negative values on the ordinate). The correlation coefficient for the linear regression line is r := 0.81, if data points for movements in the on-direction are used. If unit activity was related to planes other than the 110~' plane, correlation coefficients fell.
the direction o f m o v e m e n t . Q u a n t i t a t i v e analysis showed that, as in the P P R F 17, unit activity c o u l d be related to p o s i t i o n changes in certain planes (d-pos) or to the direction o f m o v e m e n t (a). Units coding position changes (d-pos) ( N := 41 0[45) did so by the n u m b e r o f spikes p e r burst. C o m p o n e n t s o f m o v e m e n t s in p a r t i c u l a r directions were calculated f r o m consecutive s p o n t a n e o u s eye m o v e m e n t s a n d related to the n u m b e r o f spikes in each burst. Fig. 4 shows d a t a f r o m a unit which codes p o s i t i o n changes in a plane o f l l0 °. The more spikes in a burst, the higher the p r o b a b i l i t y that there was a large m o v e m e n t c o m p o n e n t in the 1 l0 ° direction. N o t e t h a t there were also a few spikes when the eyes m o v e d d o w n w a r d (negative values on ordinate). Thus a l t h o u g h this unit fired with m o v e m e n t s in all directions, there was a linear relation between the n u m b e r o f spikes and the m o v e m e n t - c o m p o n e n t into the direction o f 110 ~. Several d a t a points f r o m bursts with 1-4 spikes lie on a line o f zero position change in the l l0 ° plane. I n these instances m o v e m e n t c o m p o n e n t s p e r p e n d i c u l a r to 110 ° occurred which are roughly in the horizontal plane. C o r r e l a t i o n coefficients were calculated for linear regression lines. Only eye m o v e m e n t s in the on-direction were used, because there was usually a change o f slope between values for m o v e m e n t s into the on- a n d off-direction (positive a n d negative values on the ordinate). Values o f c o r r e l a t i o n coefficients for linear regression lines were between 0.75 a n d 0.90. These calculations were also used to d e t e r m i n e the on-directions o f units. The planes o f vector projections were systematically shifted in steps o f 10 ° and c o r r e l a t i o n coefficients were d e t e r m i n e d for the r e l a t i o n s h i p between n u m b e r o f spikes a n d the vector projections o n t o each o f the different planes. T h e c o r r e l a t i o n was highest in only one plane. This plane was then considered to be
245
angle A
B
360o .
.
.°
.:.,
M+
.I
":,~. . .....,.-." .~,.°
270 °-
*
•.:°, ,.a~:" • .,. ~,,,~...
9 0 o.
.-...i::
270 °
.
' i';'~
180 °-
angle 360 °
180 °
• 90 °
t +°
.
0 o.
0 ....
260
400
660 frequency[Hz]
0 200 ....
460
660 frequency[Hz]
Fig. 5. Analysis o f a unit coding direction (~1) of eye m o v e m e n t s (ordinate) by the m e a n frequency during each burst (abscissa). In A the data from 248 eye m o v e m e n t s are displayed. The relationship between frequency and direction of m o v e m e n t was most closely a p p r o x i m a t e d by a cosine function, having a m a x i m u m at 265 c' (correlation coefficient for the linear line of regression r 0.85). T h e similarity to a cosine function is m o r e clearly d e m o n s t r a t e d in B where the averaged values of the data points in A are plotted. T h e observed frequencies were independent of the size o f the eye movements.
the on- or preferred direction. Correlation coefficients for planes perpendlcmar to the on-direction were always close to zero. The units coding direction (a) ( N -- 4 of 45) did so through frequency changes. If the average frequency of each burst was related to the direction of eye movement, a characteristic curve was obtained with a maximum and minimum separated by 180+ (Fig. 5). Values for the maximum frequency in Fig. 5 occurred with movements in a direction of 265 °. This curve was well described by a cosine function, whose average values are shown in Fig. 5B. Correlation factors for the regression line relating frequencies to the cosine of direction of movement (cos (a - - x)) were used to determine on-directions.
Planes of on-directions As described, correlation coefficients were used to determine on-directions for units coding both position changes (d-pos) and direction of movements (a). In Fig. 6 these on-directions are in planes which are tilted about 10-20 ° from the vertical. Few units had strictly vertical on-directions. They did not form a separate group, but appeared to be part of a population scattered around an off-vertical maximum. Most of the units, which were all recorded from the right side of the brain stem, had a 100° or 280 + on-direction. These two directions form one plane. This is the plane of the pulling direction of vertical eye muscles whose motoneurons are located on the left side of the brain stem15,17,TM. These include the inferior oblique and inferior rectus of the left eye, and the superior rectus and superior oblique of the right eye 35.
246 90* ~ up
\ 180"
00
left
right
1 unil
270* I down
Fig. 6. Display of on-directions for 45 burst units, all recorded on the right side of the brain. These planes are tilted by 10-20° from the vertical plane. The plane of 100° (or 280° in the opposite direction) coincides with the average pulling direction of vertical eye muscles with their mot,neurons on the left side.
Latencies The onset o f activity in the burst units was closely related to the beginning o f saccades, and occurred earlier than the 4-6 msec latency o f mot,neurons13,!5,29, z0, which is consistent with the idea that these are p r e m o t o r neurons. This was certainly the case for most neurons (N = 35 o f 45), which had preferred directions of activity. Their latency was 5-15 msec before the eye m o v e m e n t if there was a c o m p o n e n t of m o v e m e n t in the preferred or on-direction (Fig. 7A). Only two units began firing 3 ~ 4 0 msec before the beginning o f eye movements. If there was any activity for the movements in the opposite direction, its onset was usually after the beginning o f a saccade (Fig. 7A). The variation o f latencies in the off-direction was large; often activity started only 30-50 msec after the saccade began, but it usually did not occur after the end o f the movement. F o r other units the observed latencies were generally too short to have allowed them to participate in the initiation o f eye movements. A n example is shown in Fig. 7B. This was a unit which coded the amplitude o f eye movement. Its latencies were distributed a r o u n d 1-2 msec before the beginning o f the saccade. A similar latency distribution was observed for all units coding amplitude and for 10 o f the units with a preferred direction. As in the other units with preferred directions, latencies for movements into off-directions were between 10 and 50 msec after the beginning o f the movement.
247
N U u
15-
A
u u
1C
D
u
D
U
D
U
D
u
DD
D
u
DD
D
U
D
DDD D
D
U
D D D D D D
U u
D D D D D D
DD
u
D D D D D D D
D D D D D D D D D D -is
15-
. . . . . .- ~ .o . .
u
-'s
u
u
. . . .
0 saccade onset
u +5
U < (msec)
N
B 10
0 0 0
0
0 0 0
0 0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
o 0
o
0
0
0
0
0
o
0
0
0
0
0
0
0
0
0 |
-5
'
,
0
,
,
,
+5
.
.
.
.
,
--
+10 (msec)
160 a
saccade onset
Fig. 7. Latencies (first spike relative to the onset of the saccade) for two different burst units (A and B). The unit in A was coding d-position 280". For A, eye movements were grouped in movements with downward or upward components. For downward movements (D) the latencies were between 6 and 15 msec before the onset of the saccade, whereas for the majority of upward movements (U) the first spike occurred more than 6 msec after the onset of the saccade. In B the latencies for a unit coding the amplitude of eye movements are displayed. Most of the latencies are between 1 and 5 msec before the saccade onset, but often the first spike occurs only after the start of the movement.
248
Tonic units ( N -- 18) Tonic units are characterized by a continuous discharge during periods of fixation and the discharge rate can be related to eye position. On-directions fbr tonic units were similar to those of the burst units: 8 units had upward and 10 had downward directions. As with the burst units, the on-directions were tilted t0-20 ° from the vertical. No unit was encountered whose activity could be related to position changes in the horizontal plane. No anatomical separation was found for upward or downward related tonic units. The activity was less regular than for motoneuronsl~,l.~,,29,:~o, and the highest frequencies were less than 300 imp./sec. Some units had a threshold, whereas other fired with all positions. During saccades some units had a weak phasic increase in firing which started from 5 to 10 msec before eye movements in the preferred direction and paused for saccades in the opposite direction. In others no phasic component could be detected. Quantitative analysis showed that in most cases the correlation between position and firing rate was weak. A better correlation was usually found for the relationship between change in frequency and change in position, but it was still poor in comparison to the precision found in oculomotor neurons I~. Other units ( N =- 21) One unit was classified as a pause unit. It fired with a constant frequency and paused with each saccade. The pause started before the eye movement and its duration was related to the amplitude. Two units, which were otherwise silent, started to fire more than 100 msec before all eye movements with upward components. The frequency reached a maximum shortly before the onset of eye movement and dropped to zero during the saccade. The remaining 18 units showed some irregular activity which was not related to eye position. The increase in activity was always associated with an eye movement, but was irregular, and often occurred after the beginning of movement. With quantitative analysis we could not determine a consistent correlation between eye movements and the activity increase. DISCUSSION
Physiological characterization of M R F units The activity of most units in the rostral M R F could be related to eye movements. The preferred direction of movement for these units was only vertical. This activity was the same for eye movements in light and in darkness, and for fast phases of vestibular or optokinetic nystagmus. This suggests that the rostral M R F is an immediate supranuclear structure mediating fast eye movements in the vertical plane. It seems remarkable that all burst units could be quantitatively related to a vector parameter of eye movement and that the relation to the number of spikes, or average frequency per burst could always be described by a linear function, when curve segments above threshold were considered. Nearly all the units with a preferred direction (all recorded on the right side) had their on-directions tilted from the vertical. Most of these directions were close to the 100°/280 ° plane. This is also the plane of ondirections for motoneurons which lie on the left side of the brain stem. This suggests
249 a decussation of the projection from the M R F to the motoneurons. We would then predict that on-directions from burst units, recorded from the left MRF, would mostly be in the 80°/260 ° plane. It has been shown that in order to move the eyes along the plane of their respective pulling direction, motoneurons need an excitatory input which is equivalent to the position change. The linear relation between position change (d-pos) and the number of spikes in many M R F burst units could provide this information and directly relay it to the motoneurons 15,16. In most units with a preferred direction, latencies for movements into the on-direction were appropriate for their action as premotor units, starting and controlling eye movements. Activity during movements in the off-direction consisted of a small number of spikes, which usually started considerably (10- 50 msec) after the onset of the saccade. For most amplitude units, the onset of activity was usually too late to allow them to initiate eye movements, although they could still play a role in determining amplitude of movements. Alternatively the units could also convey information about eye movements to other systems, like the frontal eye fields or the pregeniculate body, where saccade related activity starts after the beginning of the eye movement2,4.
Anatomy The part of the rostral M R F that chiefly contained cells related to vertical eye movements has received very little attention in anatomical studies, compared with the nucleus of Darkschewitsch and the interstitial nucleus of Cajal, which lie directly behind it. According to Crosby and Woodburne 11 the region belongs to the tegmental area, situated at the transition between the diencephalon and the mesencephalon, where large cells lie among the bundles of the medial longitudinal fasciculus (MLF), rostral to the third nucleus. Preliminary studies show a particularly strong input to this cell group from the pontine reticular formation 5, suggesting that this region is itself a functional entity, and is a specialization of the MRF, like the nucleus of Darkschewitsch and the interstitial nucleus of Cajal z3. The latter nuclei, which are classically related to vertical eye movements, may have contributed to the records from only the most caudal penetrations, but have not been studied in detail in the present study. From degeneration studies it has been shown that the interstitial nucleus of Cajal projects to the oculomotor and trochlear nucleus6, 33, whereas such projections have not yet been investigated specifically from the rostral MRF. The localization of eye movement related neurons is also in good agreement with available clinical data, but is not sufficient to explain all clinical phenomena like occurrences of isolated upward or downward paresis 7. However most of our units were located relatively rostrally, in the area which from the few clinical reportsS,19, 34 appears to be concerned with downward eye movements. Stimulation, unit and lesion data Unilateral lesions, more posterior and lateral to the area which we recorded
250 from, cause gaze preference to the contralateral side, spontaneous nystagmus in the dark to the ipsilateral side, and a severe disturbance of horizontal optokinetic nystagmus to the contralateral side2L Hess and Hassler (summarized in ref. 14) also generated strong rotatory eye movements by unilateral stimulation in the M R E This emphasizes that the rostral M R F might also be associated with eye movements in other planes. Matsunami 25 localized units in the central gray, adjacent and caudal to our recording sites, which fired before and during saccades.
Comparison of M R F and PPRF units Activity in the rostral P P R F has been explored in several studies 13,~7,~°. If the M R F plays a similar role for generating eye movements in the vertical plane as the P P R F does for horizontal movements, then unit behavior should be similar. A common feature of both structures is that burst units are found, whose activity can be related to vector components of eye movements. In the M R F only units related to vertical movements are encountered, whereas P P R F units had on-directions in horizontal as well as in vertical planes. Another difference is that only one out of more than 100 M R F units was a pause unit. In P P R F pause units are much more commonly encountered: these are units which maintain a high firing frequency, only interrupted prior and during eye movements. In the P P R F a fair percentage of burst units start firing more than 20 msec, in some cases more than 100 msec before the onset of a saccade. The activity of these 'long-lead burst units' is also related to vector parameters of eye movements, so that information about the next movement is available in the P P R F about 100 msec before its onset. In the M R F we found only two units whose activity started more than 100 msec before the next eye movement. Thus, in general, information about eye movements seems to be available in the M R F only immediately before the movement is executed. The important question of how the horizontal and vertical components of eye movements are temporally coordinated is still unanswered. It seems reasonable to assume that there might be a center where information about horizontal as well as vertical components of eye movements is coordinated. The single unit data suggest that this could be in the rostral part of the PPRF. This would imply that the coding of vertical eye movements may originally occur in the PPRF and then be transferred from the P P R F to the M R F . There is evidence to support this hypothesis; Bender and Shanzer 1 report that bilateral lesions in the P P R F cause vertical as well as horizontal gaze paralysis and that bilateral stimulation in the tegmentum of the pons causes vertical eye movements 27. If this hypothesis is correct, a pathway from the P P R F to the M R F should be present, and this has been demonstrated in the monkey 5. Involvement of the pontine reticular formation in vertical gaze is supported by clinical observations as well 7,12,a2. In summary, we have shown that unit activity in the rostral M R F can be related to vertical eye movements, and that a vector description of neuronal activity allows the prediction of amplitude and direction of spontaneous eye movements as well as fast phases of nystagmus.
251 ACKNOWLEDGEMENTS T h e c o m p u t e r p r o g r a m s w e r e w r i t t e n by C h a r l e s L a s n e r ( D e p t . o f N e u r o l o g y , M t . Sinai S c h o o l o f M e d i c i n e , N e w Y o r k , U . S . A . ) . T h e a u t h o r s wish to t h a n k M s V. l s o v i i t a f o r t e c h n i c a l assistance a n d M s Schill f o r t y p i n g the m a n u s c r i p t . S u p p o r t e d by Swiss N a t i o n a l F u n d 3.044.76.
REFERENCES 1 Bender, M. B. and Shanzer, S., Oculomotor pathways defined by electrical stimulation and lesions in the brain stem of the monkey. In M. B. Bender (Ed.), The Oculomotor System, Harper and Row, New York, 1964, pp. 81-140. 2 Bizzi, E., Discharge of frontal eye field neurons during saccadic and following eye movements in unanesthetized monkeys, Exp. Brain Res., 6 (1968) 69-80. 3 Bond, H. W. and Ho, P., Solid miniature silver-silver chloride electrodes for chronic implantation, Electroenceph. clin. NeurophysioL, 28 (1970) 206-208. 4 Bfittner, U. and Fuchs, A. F., Influence of saccadic eye movements on unit activity in simian lateral geniculate and pregeniculate nuclei, J. NeurophysioL, 36 (1973) 127 141. 5 B~ttner-Ennever, J. A. and Henn, V., An autoradiographic study of the pathways from the pontine reticular formation involved in horizontal eye movements, Brain Research, 108 (1976) 155-164. 6 Carpenter, M. B., Harbison, J. W. and Peter, P., Accessory oculomotor nuclei in the monkey: projections and effects of discrete lesions, J. comp. NeuroL, 140 (1970) 131-154. 7 Christoff, N., A clinicopathological study of vertical eye movements, Arch. Neurol. (Chic.), 3l (1974) 1-8. 8 Cogan, D. C., Paralysis of down-gaze, Arch. Ophthalmol., 91 (1974) 192 199. 9 Cohen, B. and Henn, V., Unit activity in the pontine reticular formation associated with eye movements, Brain Research, 46 (1972) 403410. 10 Cohen, B. and Komatsuzaki, A., Eye movements induced by stimulation of the pontine retiuclar formation. Evidence for integration in oculomotor pathways, Exp. NeuroL, 36 (1972) 101-117. 11 Crosby, E. C. and Woodburne, R. T., The nuclear pattern of the non-tectal portions of the midbrain and the isthmus in primates, J. comp. NeuroL, 78 (1943) 441-482. 12 Freeman, W., Paralysis of associated lateral movements of the eyes. A sy~ ~ ~ ~n of infrapontile lesion, Arch. Neurol. Psyehiat. (Chic.), 7 (1922) 454487. 13 Fuchs, A. F. and Luschei, E. S., Firing patterns of abducens neurons of alert monkeys in relationship to horizontal eye movements, J. NeurophysioL, 33 (1970) 382 392. 14 Hassler, R., Supranuclear structures regulating binocular eye and head movements. In J. Dichgans and E. Bizzi (Eds.), Cerebral Control of Eye Movernents and Motion Perception, Bibl. Ophthal., VoL 82, Karger, Basel, 1972, pp. 207-220. 15 Henn, V. and Cohen, B., Quantitative analysis of activity in eye muscle motoneurons during saccadic eye movements and positions of fixation, J. NeurophysioL, 36 (1973) 115 126. 16 Henn, V. and Cohen, B., Activity in eye muscle motoneurons and brain stem units during eye movements. In J. Lennerstrand and P. Bach-Y-Rita (Eds.), Basic Mechanisms of Ocular Motility and Their Clinical Implications, Pergamon Press, Oxford, 1975, pp. 303-345. 17 Henn, V. and Cohen, B., Coding of information about rapid eye movements in the pontine reticular formation of alert monkeys, Brain Research, 108 (1976) 307-325. 18 Hering, E., Die Lehre vom binokularen Sehen, Engelmann, Leipzig, 1868. 19 Jacobs, L., Anderson, P. J. and Bender, M. B., The lesions producing paralysis of downward but not upward gaze, Arch. Neurol. (Chic.), 28 (1973) 319-323. 20 Keller, E. L., Participation of medial pontine reticular formation in eye movement generation in monkey, J. Neurophysiol., 37 (1974) 316-332. 21 King, W. M., Magnin, M. and Fuchs, A. F., Mesencephalic reticular neurons related to visual and vestibular evoked eye movements. In Neuroscience Abstracts, Vol. 1, Society for Neuroscience, Bethesda, Md., 1975, Abstr. 353.
252 22 Komatsuzaki, A., Alpert, J., Harris, H. E. and Cohen, B., Effect of mesencephalic reticular formation lesions on optokinetic nystagmus, Exp. NeuroL, 24 (1972) 522-534. 23 Krieg, W. J. S., Functional Neuroanatomy, Brain Books, Evanston, Ill., 1966. 24 Luschei, E. S. and Fuchs, A. F., Activity of brain stem neurons during eye movements of alert monkeys, J. Neurophysiol., 35 (1972) 445-461. 25 Matsunami, K., Saccadic eye movements and neurons in the central gray area in awake monkeys, Brain Research, 38 (1972) 217-221. 26 Olszewski, J., The Thalamus o[ the Macaea mulatta, Karger, Basel, 1952. 27 Pasik, P. and Pasik, T., Oculomotor functions in monkeys with lesions of the cerebrum and superior colliculi. In M. B. Bender (Ed.), The Oculomotor System, Harper and Row, New York, 1964, pp. 40-80. 28 Pasik, P., Pasik, T. and Bender, M. B., The pretectal syndrome in monkeys. I. Disturbances (~f gaze and body posture, Brain, 92 (1969) 521-534. 29 Robinson, D. A., Oculomotor unit behaviour in the monkey, J. Neurophysiol., 33 (1970) 393-404. 30 Schiller, P.H., The discharge characteristics of single units in the oculomotor and abducens nuclei of the unanesthetized monkey, Exp. Brain Res., l0 (1970) 347-362. 31 Schwindt, P. C., Precht, W. and Richter, A., Monosynaptic excitatory and inhibitory pathways from medial midbrain nuclei to trochlear motoneurons, Exp. Brain. Res., 20 (1974) 223-238. 32 Spiller, W. G., The importance in clinical diagnosis of paralysis of the eyeballs (B!icklfihmung), especially of upward and downward associated movements, J. Nerv. ment. Dis., 32 (1905), 4t7-456. 33 Szentagothai, J., Die zentrale Innervation der Augenbewegungen, Arch Psyehiat: Nervenkr., 116 (1943) 721-760. 34 Thomas, A., Schaefer, M. et Pertrand, F., Paralysie de 1,abaissement du regard: Paralysie des inferogyres, hypertonie des superogyres et des releveurs du regard, Rev. NeuroL, 11 (t933) 535-541. 35 Warwick, R., Representation of the extraocular muscles in the oculomotor nuclei of the monkey, J. comp. Neurol., 98 (1953) 449-504.
239
V E R T I C A L EYE M O V E M E N T R E L A T E D U N I T ACTIVITY IN T H E R O S T R A L M E S E N C E P H A L I C R E T I C U L A R F O R M A T I O N OF T H E A L E R T M O N K E Y
U. BI]TTNER, J. A. B(JTTNER-ENNEVER and V. HENN Department of Neurology and Brain Research hTstitute, University of Zfirich, Ziirich (Switzerland)
(Accepted November 3rd, 1976)
SUMMARY Eye movement related unit activity was recorded in the rostral mesencephalic reticular formation ( M R F ) of the alert monkey. Most units (78 out of 117) were activated with a short activity burst starting before the eye movement and were otherwise silent. The activity was the same whether movements occurred spontaneously in the light or dark, or were fast phases of vestibular or optokinetic nystagmus, and could be related to parameters of a vector representing the eye movement such as amplitude, position changes along certain planes or direction of movements. Units coding position changes or direction of movement had their preferred direction always close to the vertical. Other units (18 out of 117) showed some tonic activity, which was also only related to vertical eye position. It is suggested that this region of the rostral M R F acts as an immediate supranuclear structure, mediating eye movements in the vertical plane.
INTRODUCTION Paralysis of vertical or horizontal eye movements can occur independently in man and monkeys. This implies that there are separate anatomical sites for the generation of vertical and horizontal eye movements. However, as many movements have a vertical and horizontal component, the execution of both these components must be time-locked, and hence these two areas must be functionally connected. Clinical evidence has shown that bilateral lesions in the medial mesencephalic tegmentum rostral to the 3rd nucleus lead to impairment of vertical eye movements 7. Lesion experiments in the monkey are in good agreement with these observations zS. There is also anatomical6,3z and electrophysiologica131 evidence that the interstitial nucleus of Cajal, which lies in this area, is directly connected to the oculomotor nuclei involved in vertical eye movements.
240 A necessary condition for the hypothesis that eye movements ii~ tt~e vertica! plane are generated in the mesencephalic reticular formation (MRF) is tts~;~tthe unil activity found there leads eye movements and can be quantitatively related to the parameters of eye movement. So far there is little evidence for this available2~, e:.. However several lines of evidence suggest that the paramedian pontine reticular formation (PPRF) is the immediate supranuclear structure responsible fol° eye movements in the horizontal planeLg,to,'-'0,".L This report presents evidence that unit activity in the rostral M R F ~>t"the alert monkey is indeed related to vertical eye movements, and that the strategies of quar> titative analysis applied to units in the PPRF, which allowed the prediction of single eye movements .7, are also suitable for analyzing activity in the rostral M R F METHODS A metal ring, that could accept a microdrive, was chronically implanted over a trephine hole in the skull of 3 juvenile rhesu; monkeys (Macaca mulatto). Extracellular unit potentials were recorded with tung;ten microelectrodes of 2-8 MQ. The horizontal position of the eye was measured with DC silver-silver chloride electrodes a inserted at the outer canthi of the boney orbit. In order to measure ver~icat eye position, altogether 4 DC-electrodes were implanted above and below each eye, the upper and lower pair being connected. This arrangement minimizes cross-coupling. Eye movements were calibrated by measuring slow-phase velocity during whole field optokinetic stimulation. During recording sessions the monkey sat upright in a primate chair with its head rigidly fixed to a headholder. To achieve a constant level of alertness the monkey usually received small doses of amphetamine (max. dose, 0.3 mg/kg i.m.). Unit recordings were taken while the monkey made spontaneous eye movements in the dark and the light. In addition horizontal nystagmus was induced either optokinetically (rotation of a striped cylinder around the stationary monkey) or by vestibular stimulation (rotation of the monkey on a servocontrolled turntable in the dark). Single unit activity, horizontal and vertical eye position, light-on and -off signal, the cylinder and turntable velocity, and a digital time code were stored on an FM tape recorder. At the end of the experiments a selected recording site was marked by electrolytic lesions or by injection of tracer materials. Animals were perfused under deep anesthesia. Frozen sections of the midbrain, taken every 320/~m, were stained with cresylecht violet and recording sites were identified. Quantitative analysis was performed using the same computer programs as in earlier studies 16,17. At least 120 sec of consecutive eye movements (sample rate 1.25 kHz each channel) and unit data (sample rate 20 kHz) were stored on digital magnetic tape. The beginning and end of eye movements were marked. Blinks were marked separately, and excluded from further analysis. Statistical programs related unit activity to parameters of movement and to eye position. As in an earlier study of PPRF units, eye movements will be described using vectors. A vector is defined by an amplitude (A) and a direction (a) (Fig. 1). An im-
241 110 °
90 °
70 °
d--pos --'---
18o_ °
~
I
/', g
o o
270 ° Fig. 1. Schematic drawing of an eye movement represented by a vector, characterized by an amplitude (A) and a direction (u). The vector projection of this eye movement on the planes 70° (d-pos 70') and 110' (d-pos 110) are demonstrated. ~ 0' is defined as right, a ~ 90~ as up, ct 180~ as left, and ~ 270 as down. In subsequent calculations unit activity will systematically be related to each of these parameters, A unit coding position changes along the direction of 70 would be related to the other parameters by d-pos 70c A - cos(a-70 ). In this equation 70 is subtracted from ~, because u is always measured in reference to the horizontal right direction and not in reference to the ondirection of the units. p o r t a n t derivative of these parameters is the vector projection (d-pos) onto specific planes. Two examples for the planes 70 ° (up, right), and 110 ° (up, left), are shown in Fig. 1. The vector projections in these planes are related to amplitude and direction by d-pos 110,~ A • cos (a-110 °) or d-pos 70o -- A • cos (ct-70°). In the analysis unit activity before and during eye movements ( n u m b e r of spikes per burst and mean frequency of each burst) was systematically related to each of these 3 parameters: amplitude (A), angle of m o v e m e n t (a), and position changes along particular planes (d-pos). R ES U LTS Recording sites
All units were recorded in the rostral M R F in a n area anterior to the oculom o t o r nucleus, dorsal to the red nucleus and ventral to the thalamus. Laterally this area extends from a b o u t 1 m m from the midline to the fields of Forel and the zona incerta. It is centered a r o u n d stereotaxic coordinates A 8.0, L 2-3 (ref. 26) and extends in the stereotaxic plane 1-2 mm ventrodorsal[y, 2 m m mediolaterally and 3 mm anteroposteriorly. This area coincides mainly with the most rostral part of the M R F anterior to the nucleus of Darkschewitsch and the interstitial nucleus of Cajal. It is also the projection site of a pathway arising in the ipsilateral P P R F 5. Fig. 2 shows the localization of two lesions from two different monkeys in this region. Useful landmarks were the third nerve rootlets which were encountered 3-4 m m ventrally (see
242
Fig. 2. Section of brain taken in the stereotaxic transverse plane at A 8 and stained ailh cre.sy/ech~ violet, showing the electrode track from the final recording session in which an efecirolyfic lesh~ was made in the area were burst units were found. The arrow indicates the location of a similar poiiH, marked in a second monkey, and lying in exactly the same location. LGN, lateral genicH!ate mlcletx,;: MRF, mesencephalic reticular formation ; N I1 I, 3rd nerve rootlets ; RN, red nucleus : St. st~bthalanms : VL, n. ventralis lateralis; VPI, n. ventralis posterior inferior: V PL, n. ventralis poste.~ior lateral is: ZI, zona incerta. Magnification 3.5.
Fig. 2). The m o s t p o s t e r i o r recordings were t a k e n at the level o f the a n t e r i o r pole o f the interstitial nucleus o f Cajal. N o records were taken as far caudal as the nuclei o f the p o s t e r i o r commissure.
Unit activity M o s t units in the a r e a described a b o v e altered their activity with s p o n t a n e o u s eye m o v e m e n t s . A b o u t 30 o//o d i d not, and will not be considered further. The results are based on the analysis o f 117 units, all on the right side. O f these 117 units, 78 were b u r s t units, 18 tonic units and the r e m a i n i n g 21 fell in o t h e r categories. Tonic and b u r s t units were intermingled, a n d could sometimes be r e c o r d e d simultaneously.
Burst units ( N ~- 78) Burst units were active only with saccades or fast phases o f nystagmus. O t h e r wise they were silent, even d u r i n g slow phases o f o p t o k i n e t i c or vestibular nystagmus. W h e n m o n k e y s b e c a m e d r o w s y a n d m a d e slow drifting eye m o v e m e n t s without saccades, these units often h a d irregular activity o f 10-20 imp./sec. This was t a k e n to indicate t h a t units were inhibited d u r i n g gaze in the alert state. Saccade related bursts consisted o f 1-40 spikes which e n d e d a p p r o x i m a t e l y at
243 the end o f a saccade. The highest frequencies usually occurred at the very beginning o f the burst and fell monotonically• There were no differences in activity d u r i n g eye m o v e m e n t s in light or dark. Activity patterns were also i n d e p e n d e n t o f the starting p o s i t i o n o f the eye and the hemifield in which the eye movements were executed. Burst units could be divided into two m a i n g r o u p s : Those activated with eye movements in all directions (Fig. 3A), a n d those activated with eye movements in a preferred direction. Activity o f the first 57 units was analyzed quantitatively and the r e m a i n i n g 21 units qualitatively.
Units coding amplitude of eye movements ( N = 12 of 57) Activity o f a burst unit which e n c o d e d the a m p l i t u d e o f eye m o v e m e n t s is shown in Fig. 3A. There was a short b u r s t o f activity which started j u s t before each saccade. The n u m b e r o f spikes in consecutive bursts could be related to the a m p l i t u d e o f eye m o v e m e n t (Fig. 3B). In a few units there was a large v a r i a t i o n in this relationship, a n d only averaged d a t a showed a clear correlation• M e a n frequency o f firing in single bursts had no consistent relation to any o f the other m o v e m e n t p a r a m e t e r s tested (direction or d-pos).
Units with preferred directions of activity ( N -- 45 of 57) In these units the preferred direction was close to the vertical plane. N o units were f o u n d with on-directions in the h o r i z o n t a l plane. Twenty units fired m a x i m a l l y with upward, and 25 with d o w n w a r d movements. These units were intermingled and sometimes could be recorded simultaneously. M o s t units fired only with m o v e m e n t s in their respective on-directions and were otherwise silent• Others fired with every eye movement, and the n u m b e r o f spikes or frequency o f the burst was m o d u l a t e d with
A
B
++ t
L Illlt
_iLl
-
H-
i
[deg]
60
v
40.
,.--,...r-"-----~ 30 °
.:, . : . ; : : •
E
•
v ....
~_~,__
"i
.°
:::,ti;|.": w.* w •
.t
''. •
."
•
_d_,~-~.~
2985 500
msec
•
• ":1q1i .•~' . . . . ~2
0
1'0
s
20 30 number of spikes
Fig. 3. A : original record of a unit exhibiting a short activity burst with every saccade during spon+ taneous eye movements in the light. The quantitative analysis showed that this unit was coding the amplitude of the eye movement by the number of spikes. H = horizontal, and V vertical, eye position. Right and upward movements are indicated by upward deflection. B: computer analysis of a unit coding the amplitude of the eye movement by the number of spikes, independent of the direction of the eye movements. The values of 278 consecutive eye movements into all directions are displayed.
244 [degree] +40-
0
+20i
0-
.-i,''
.'.;~ e~"." : ,.
.
"[i.
CZ
20-
if"
40-
370 4
6
2'0
3'o
[number of spikes]
Fig. 4. Quantitative analysis of one unit coding position changes in the plane of I l0 (d-pos l l0 ) through the number of spikes per burst. Data from 253 consecutive spontaneous eye movements, in all directions, are displayed. For each eye movement the movement-component in the 110° direction (ordinate) was calculated and related to the number of spikes per burst (abscissa). Note that there a re also a few spikes with eye movements in the opposite direction (290°; i.e. negative values on the ordinate). The correlation coefficient for the linear regression line is r := 0.81, if data points for movements in the on-direction are used. If unit activity was related to planes other than the 110~' plane, correlation coefficients fell.
the direction o f m o v e m e n t . Q u a n t i t a t i v e analysis showed that, as in the P P R F 17, unit activity c o u l d be related to p o s i t i o n changes in certain planes (d-pos) or to the direction o f m o v e m e n t (a). Units coding position changes (d-pos) ( N := 41 0[45) did so by the n u m b e r o f spikes p e r burst. C o m p o n e n t s o f m o v e m e n t s in p a r t i c u l a r directions were calculated f r o m consecutive s p o n t a n e o u s eye m o v e m e n t s a n d related to the n u m b e r o f spikes in each burst. Fig. 4 shows d a t a f r o m a unit which codes p o s i t i o n changes in a plane o f l l0 °. The more spikes in a burst, the higher the p r o b a b i l i t y that there was a large m o v e m e n t c o m p o n e n t in the 1 l0 ° direction. N o t e t h a t there were also a few spikes when the eyes m o v e d d o w n w a r d (negative values on ordinate). Thus a l t h o u g h this unit fired with m o v e m e n t s in all directions, there was a linear relation between the n u m b e r o f spikes and the m o v e m e n t - c o m p o n e n t into the direction o f 110 ~. Several d a t a points f r o m bursts with 1-4 spikes lie on a line o f zero position change in the l l0 ° plane. I n these instances m o v e m e n t c o m p o n e n t s p e r p e n d i c u l a r to 110 ° occurred which are roughly in the horizontal plane. C o r r e l a t i o n coefficients were calculated for linear regression lines. Only eye m o v e m e n t s in the on-direction were used, because there was usually a change o f slope between values for m o v e m e n t s into the on- a n d off-direction (positive a n d negative values on the ordinate). Values o f c o r r e l a t i o n coefficients for linear regression lines were between 0.75 a n d 0.90. These calculations were also used to d e t e r m i n e the on-directions o f units. The planes o f vector projections were systematically shifted in steps o f 10 ° and c o r r e l a t i o n coefficients were d e t e r m i n e d for the r e l a t i o n s h i p between n u m b e r o f spikes a n d the vector projections o n t o each o f the different planes. T h e c o r r e l a t i o n was highest in only one plane. This plane was then considered to be
245
angle A
B
360o .
.
.°
.:.,
M+
.I
":,~. . .....,.-." .~,.°
270 °-
*
•.:°, ,.a~:" • .,. ~,,,~...
9 0 o.
.-...i::
270 °
.
' i';'~
180 °-
angle 360 °
180 °
• 90 °
t +°
.
0 o.
0 ....
260
400
660 frequency[Hz]
0 200 ....
460
660 frequency[Hz]
Fig. 5. Analysis o f a unit coding direction (~1) of eye m o v e m e n t s (ordinate) by the m e a n frequency during each burst (abscissa). In A the data from 248 eye m o v e m e n t s are displayed. The relationship between frequency and direction of m o v e m e n t was most closely a p p r o x i m a t e d by a cosine function, having a m a x i m u m at 265 c' (correlation coefficient for the linear line of regression r 0.85). T h e similarity to a cosine function is m o r e clearly d e m o n s t r a t e d in B where the averaged values of the data points in A are plotted. T h e observed frequencies were independent of the size o f the eye movements.
the on- or preferred direction. Correlation coefficients for planes perpendlcmar to the on-direction were always close to zero. The units coding direction (a) ( N -- 4 of 45) did so through frequency changes. If the average frequency of each burst was related to the direction of eye movement, a characteristic curve was obtained with a maximum and minimum separated by 180+ (Fig. 5). Values for the maximum frequency in Fig. 5 occurred with movements in a direction of 265 °. This curve was well described by a cosine function, whose average values are shown in Fig. 5B. Correlation factors for the regression line relating frequencies to the cosine of direction of movement (cos (a - - x)) were used to determine on-directions.
Planes of on-directions As described, correlation coefficients were used to determine on-directions for units coding both position changes (d-pos) and direction of movements (a). In Fig. 6 these on-directions are in planes which are tilted about 10-20 ° from the vertical. Few units had strictly vertical on-directions. They did not form a separate group, but appeared to be part of a population scattered around an off-vertical maximum. Most of the units, which were all recorded from the right side of the brain stem, had a 100° or 280 + on-direction. These two directions form one plane. This is the plane of the pulling direction of vertical eye muscles whose motoneurons are located on the left side of the brain stem15,17,TM. These include the inferior oblique and inferior rectus of the left eye, and the superior rectus and superior oblique of the right eye 35.
246 90* ~ up
\ 180"
00
left
right
1 unil
270* I down
Fig. 6. Display of on-directions for 45 burst units, all recorded on the right side of the brain. These planes are tilted by 10-20° from the vertical plane. The plane of 100° (or 280° in the opposite direction) coincides with the average pulling direction of vertical eye muscles with their mot,neurons on the left side.
Latencies The onset o f activity in the burst units was closely related to the beginning o f saccades, and occurred earlier than the 4-6 msec latency o f mot,neurons13,!5,29, z0, which is consistent with the idea that these are p r e m o t o r neurons. This was certainly the case for most neurons (N = 35 o f 45), which had preferred directions of activity. Their latency was 5-15 msec before the eye m o v e m e n t if there was a c o m p o n e n t of m o v e m e n t in the preferred or on-direction (Fig. 7A). Only two units began firing 3 ~ 4 0 msec before the beginning o f eye movements. If there was any activity for the movements in the opposite direction, its onset was usually after the beginning o f a saccade (Fig. 7A). The variation o f latencies in the off-direction was large; often activity started only 30-50 msec after the saccade began, but it usually did not occur after the end o f the movement. F o r other units the observed latencies were generally too short to have allowed them to participate in the initiation o f eye movements. A n example is shown in Fig. 7B. This was a unit which coded the amplitude o f eye movement. Its latencies were distributed a r o u n d 1-2 msec before the beginning o f the saccade. A similar latency distribution was observed for all units coding amplitude and for 10 o f the units with a preferred direction. As in the other units with preferred directions, latencies for movements into off-directions were between 10 and 50 msec after the beginning o f the movement.
247
N U u
15-
A
u u
1C
D
u
D
U
D
U
D
u
DD
D
u
DD
D
U
D
DDD D
D
U
D D D D D D
U u
D D D D D D
DD
u
D D D D D D D
D D D D D D D D D D -is
15-
. . . . . .- ~ .o . .
u
-'s
u
u
. . . .
0 saccade onset
u +5
U < (msec)
N
B 10
0 0 0
0
0 0 0
0 0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
o 0
o
0
0
0
0
0
o
0
0
0
0
0
0
0
0
0 |
-5
'
,
0
,
,
,
+5
.
.
.
.
,
--
+10 (msec)
160 a
saccade onset
Fig. 7. Latencies (first spike relative to the onset of the saccade) for two different burst units (A and B). The unit in A was coding d-position 280". For A, eye movements were grouped in movements with downward or upward components. For downward movements (D) the latencies were between 6 and 15 msec before the onset of the saccade, whereas for the majority of upward movements (U) the first spike occurred more than 6 msec after the onset of the saccade. In B the latencies for a unit coding the amplitude of eye movements are displayed. Most of the latencies are between 1 and 5 msec before the saccade onset, but often the first spike occurs only after the start of the movement.
248
Tonic units ( N -- 18) Tonic units are characterized by a continuous discharge during periods of fixation and the discharge rate can be related to eye position. On-directions fbr tonic units were similar to those of the burst units: 8 units had upward and 10 had downward directions. As with the burst units, the on-directions were tilted t0-20 ° from the vertical. No unit was encountered whose activity could be related to position changes in the horizontal plane. No anatomical separation was found for upward or downward related tonic units. The activity was less regular than for motoneuronsl~,l.~,,29,:~o, and the highest frequencies were less than 300 imp./sec. Some units had a threshold, whereas other fired with all positions. During saccades some units had a weak phasic increase in firing which started from 5 to 10 msec before eye movements in the preferred direction and paused for saccades in the opposite direction. In others no phasic component could be detected. Quantitative analysis showed that in most cases the correlation between position and firing rate was weak. A better correlation was usually found for the relationship between change in frequency and change in position, but it was still poor in comparison to the precision found in oculomotor neurons I~. Other units ( N =- 21) One unit was classified as a pause unit. It fired with a constant frequency and paused with each saccade. The pause started before the eye movement and its duration was related to the amplitude. Two units, which were otherwise silent, started to fire more than 100 msec before all eye movements with upward components. The frequency reached a maximum shortly before the onset of eye movement and dropped to zero during the saccade. The remaining 18 units showed some irregular activity which was not related to eye position. The increase in activity was always associated with an eye movement, but was irregular, and often occurred after the beginning of movement. With quantitative analysis we could not determine a consistent correlation between eye movements and the activity increase. DISCUSSION
Physiological characterization of M R F units The activity of most units in the rostral M R F could be related to eye movements. The preferred direction of movement for these units was only vertical. This activity was the same for eye movements in light and in darkness, and for fast phases of vestibular or optokinetic nystagmus. This suggests that the rostral M R F is an immediate supranuclear structure mediating fast eye movements in the vertical plane. It seems remarkable that all burst units could be quantitatively related to a vector parameter of eye movement and that the relation to the number of spikes, or average frequency per burst could always be described by a linear function, when curve segments above threshold were considered. Nearly all the units with a preferred direction (all recorded on the right side) had their on-directions tilted from the vertical. Most of these directions were close to the 100°/280 ° plane. This is also the plane of ondirections for motoneurons which lie on the left side of the brain stem. This suggests
249 a decussation of the projection from the M R F to the motoneurons. We would then predict that on-directions from burst units, recorded from the left MRF, would mostly be in the 80°/260 ° plane. It has been shown that in order to move the eyes along the plane of their respective pulling direction, motoneurons need an excitatory input which is equivalent to the position change. The linear relation between position change (d-pos) and the number of spikes in many M R F burst units could provide this information and directly relay it to the motoneurons 15,16. In most units with a preferred direction, latencies for movements into the on-direction were appropriate for their action as premotor units, starting and controlling eye movements. Activity during movements in the off-direction consisted of a small number of spikes, which usually started considerably (10- 50 msec) after the onset of the saccade. For most amplitude units, the onset of activity was usually too late to allow them to initiate eye movements, although they could still play a role in determining amplitude of movements. Alternatively the units could also convey information about eye movements to other systems, like the frontal eye fields or the pregeniculate body, where saccade related activity starts after the beginning of the eye movement2,4.
Anatomy The part of the rostral M R F that chiefly contained cells related to vertical eye movements has received very little attention in anatomical studies, compared with the nucleus of Darkschewitsch and the interstitial nucleus of Cajal, which lie directly behind it. According to Crosby and Woodburne 11 the region belongs to the tegmental area, situated at the transition between the diencephalon and the mesencephalon, where large cells lie among the bundles of the medial longitudinal fasciculus (MLF), rostral to the third nucleus. Preliminary studies show a particularly strong input to this cell group from the pontine reticular formation 5, suggesting that this region is itself a functional entity, and is a specialization of the MRF, like the nucleus of Darkschewitsch and the interstitial nucleus of Cajal z3. The latter nuclei, which are classically related to vertical eye movements, may have contributed to the records from only the most caudal penetrations, but have not been studied in detail in the present study. From degeneration studies it has been shown that the interstitial nucleus of Cajal projects to the oculomotor and trochlear nucleus6, 33, whereas such projections have not yet been investigated specifically from the rostral MRF. The localization of eye movement related neurons is also in good agreement with available clinical data, but is not sufficient to explain all clinical phenomena like occurrences of isolated upward or downward paresis 7. However most of our units were located relatively rostrally, in the area which from the few clinical reportsS,19, 34 appears to be concerned with downward eye movements. Stimulation, unit and lesion data Unilateral lesions, more posterior and lateral to the area which we recorded
250 from, cause gaze preference to the contralateral side, spontaneous nystagmus in the dark to the ipsilateral side, and a severe disturbance of horizontal optokinetic nystagmus to the contralateral side2L Hess and Hassler (summarized in ref. 14) also generated strong rotatory eye movements by unilateral stimulation in the M R E This emphasizes that the rostral M R F might also be associated with eye movements in other planes. Matsunami 25 localized units in the central gray, adjacent and caudal to our recording sites, which fired before and during saccades.
Comparison of M R F and PPRF units Activity in the rostral P P R F has been explored in several studies 13,~7,~°. If the M R F plays a similar role for generating eye movements in the vertical plane as the P P R F does for horizontal movements, then unit behavior should be similar. A common feature of both structures is that burst units are found, whose activity can be related to vector components of eye movements. In the M R F only units related to vertical movements are encountered, whereas P P R F units had on-directions in horizontal as well as in vertical planes. Another difference is that only one out of more than 100 M R F units was a pause unit. In P P R F pause units are much more commonly encountered: these are units which maintain a high firing frequency, only interrupted prior and during eye movements. In the P P R F a fair percentage of burst units start firing more than 20 msec, in some cases more than 100 msec before the onset of a saccade. The activity of these 'long-lead burst units' is also related to vector parameters of eye movements, so that information about the next movement is available in the P P R F about 100 msec before its onset. In the M R F we found only two units whose activity started more than 100 msec before the next eye movement. Thus, in general, information about eye movements seems to be available in the M R F only immediately before the movement is executed. The important question of how the horizontal and vertical components of eye movements are temporally coordinated is still unanswered. It seems reasonable to assume that there might be a center where information about horizontal as well as vertical components of eye movements is coordinated. The single unit data suggest that this could be in the rostral part of the PPRF. This would imply that the coding of vertical eye movements may originally occur in the PPRF and then be transferred from the P P R F to the M R F . There is evidence to support this hypothesis; Bender and Shanzer 1 report that bilateral lesions in the P P R F cause vertical as well as horizontal gaze paralysis and that bilateral stimulation in the tegmentum of the pons causes vertical eye movements 27. If this hypothesis is correct, a pathway from the P P R F to the M R F should be present, and this has been demonstrated in the monkey 5. Involvement of the pontine reticular formation in vertical gaze is supported by clinical observations as well 7,12,a2. In summary, we have shown that unit activity in the rostral M R F can be related to vertical eye movements, and that a vector description of neuronal activity allows the prediction of amplitude and direction of spontaneous eye movements as well as fast phases of nystagmus.
251 ACKNOWLEDGEMENTS T h e c o m p u t e r p r o g r a m s w e r e w r i t t e n by C h a r l e s L a s n e r ( D e p t . o f N e u r o l o g y , M t . Sinai S c h o o l o f M e d i c i n e , N e w Y o r k , U . S . A . ) . T h e a u t h o r s wish to t h a n k M s V. l s o v i i t a f o r t e c h n i c a l assistance a n d M s Schill f o r t y p i n g the m a n u s c r i p t . S u p p o r t e d by Swiss N a t i o n a l F u n d 3.044.76.
REFERENCES 1 Bender, M. B. and Shanzer, S., Oculomotor pathways defined by electrical stimulation and lesions in the brain stem of the monkey. In M. B. Bender (Ed.), The Oculomotor System, Harper and Row, New York, 1964, pp. 81-140. 2 Bizzi, E., Discharge of frontal eye field neurons during saccadic and following eye movements in unanesthetized monkeys, Exp. Brain Res., 6 (1968) 69-80. 3 Bond, H. W. and Ho, P., Solid miniature silver-silver chloride electrodes for chronic implantation, Electroenceph. clin. NeurophysioL, 28 (1970) 206-208. 4 Bfittner, U. and Fuchs, A. F., Influence of saccadic eye movements on unit activity in simian lateral geniculate and pregeniculate nuclei, J. NeurophysioL, 36 (1973) 127 141. 5 B~ttner-Ennever, J. A. and Henn, V., An autoradiographic study of the pathways from the pontine reticular formation involved in horizontal eye movements, Brain Research, 108 (1976) 155-164. 6 Carpenter, M. B., Harbison, J. W. and Peter, P., Accessory oculomotor nuclei in the monkey: projections and effects of discrete lesions, J. comp. NeuroL, 140 (1970) 131-154. 7 Christoff, N., A clinicopathological study of vertical eye movements, Arch. Neurol. (Chic.), 3l (1974) 1-8. 8 Cogan, D. C., Paralysis of down-gaze, Arch. Ophthalmol., 91 (1974) 192 199. 9 Cohen, B. and Henn, V., Unit activity in the pontine reticular formation associated with eye movements, Brain Research, 46 (1972) 403410. 10 Cohen, B. and Komatsuzaki, A., Eye movements induced by stimulation of the pontine retiuclar formation. Evidence for integration in oculomotor pathways, Exp. NeuroL, 36 (1972) 101-117. 11 Crosby, E. C. and Woodburne, R. T., The nuclear pattern of the non-tectal portions of the midbrain and the isthmus in primates, J. comp. NeuroL, 78 (1943) 441-482. 12 Freeman, W., Paralysis of associated lateral movements of the eyes. A sy~ ~ ~ ~n of infrapontile lesion, Arch. Neurol. Psyehiat. (Chic.), 7 (1922) 454487. 13 Fuchs, A. F. and Luschei, E. S., Firing patterns of abducens neurons of alert monkeys in relationship to horizontal eye movements, J. NeurophysioL, 33 (1970) 382 392. 14 Hassler, R., Supranuclear structures regulating binocular eye and head movements. In J. Dichgans and E. Bizzi (Eds.), Cerebral Control of Eye Movernents and Motion Perception, Bibl. Ophthal., VoL 82, Karger, Basel, 1972, pp. 207-220. 15 Henn, V. and Cohen, B., Quantitative analysis of activity in eye muscle motoneurons during saccadic eye movements and positions of fixation, J. NeurophysioL, 36 (1973) 115 126. 16 Henn, V. and Cohen, B., Activity in eye muscle motoneurons and brain stem units during eye movements. In J. Lennerstrand and P. Bach-Y-Rita (Eds.), Basic Mechanisms of Ocular Motility and Their Clinical Implications, Pergamon Press, Oxford, 1975, pp. 303-345. 17 Henn, V. and Cohen, B., Coding of information about rapid eye movements in the pontine reticular formation of alert monkeys, Brain Research, 108 (1976) 307-325. 18 Hering, E., Die Lehre vom binokularen Sehen, Engelmann, Leipzig, 1868. 19 Jacobs, L., Anderson, P. J. and Bender, M. B., The lesions producing paralysis of downward but not upward gaze, Arch. Neurol. (Chic.), 28 (1973) 319-323. 20 Keller, E. L., Participation of medial pontine reticular formation in eye movement generation in monkey, J. Neurophysiol., 37 (1974) 316-332. 21 King, W. M., Magnin, M. and Fuchs, A. F., Mesencephalic reticular neurons related to visual and vestibular evoked eye movements. In Neuroscience Abstracts, Vol. 1, Society for Neuroscience, Bethesda, Md., 1975, Abstr. 353.
252 22 Komatsuzaki, A., Alpert, J., Harris, H. E. and Cohen, B., Effect of mesencephalic reticular formation lesions on optokinetic nystagmus, Exp. NeuroL, 24 (1972) 522-534. 23 Krieg, W. J. S., Functional Neuroanatomy, Brain Books, Evanston, Ill., 1966. 24 Luschei, E. S. and Fuchs, A. F., Activity of brain stem neurons during eye movements of alert monkeys, J. Neurophysiol., 35 (1972) 445-461. 25 Matsunami, K., Saccadic eye movements and neurons in the central gray area in awake monkeys, Brain Research, 38 (1972) 217-221. 26 Olszewski, J., The Thalamus o[ the Macaea mulatta, Karger, Basel, 1952. 27 Pasik, P. and Pasik, T., Oculomotor functions in monkeys with lesions of the cerebrum and superior colliculi. In M. B. Bender (Ed.), The Oculomotor System, Harper and Row, New York, 1964, pp. 40-80. 28 Pasik, P., Pasik, T. and Bender, M. B., The pretectal syndrome in monkeys. I. Disturbances (~f gaze and body posture, Brain, 92 (1969) 521-534. 29 Robinson, D. A., Oculomotor unit behaviour in the monkey, J. Neurophysiol., 33 (1970) 393-404. 30 Schiller, P.H., The discharge characteristics of single units in the oculomotor and abducens nuclei of the unanesthetized monkey, Exp. Brain Res., l0 (1970) 347-362. 31 Schwindt, P. C., Precht, W. and Richter, A., Monosynaptic excitatory and inhibitory pathways from medial midbrain nuclei to trochlear motoneurons, Exp. Brain. Res., 20 (1974) 223-238. 32 Spiller, W. G., The importance in clinical diagnosis of paralysis of the eyeballs (B!icklfihmung), especially of upward and downward associated movements, J. Nerv. ment. Dis., 32 (1905), 4t7-456. 33 Szentagothai, J., Die zentrale Innervation der Augenbewegungen, Arch Psyehiat: Nervenkr., 116 (1943) 721-760. 34 Thomas, A., Schaefer, M. et Pertrand, F., Paralysie de 1,abaissement du regard: Paralysie des inferogyres, hypertonie des superogyres et des releveurs du regard, Rev. NeuroL, 11 (t933) 535-541. 35 Warwick, R., Representation of the extraocular muscles in the oculomotor nuclei of the monkey, J. comp. Neurol., 98 (1953) 449-504.
