Page 1

Control of Gúze by Btaín Stem Nturons, Derelopments in

Neuracience, Voluhe 1, ed¡ted by Baker and Berthoz

@ 1977 ElsevíeilNotth-Holland Éiomedicat P¡ess

291

THE ACTIVITY OF INTERNUCLEAR NEURONS IDENTIFIED I^IITHIN THE

ABDUCENS NUCLEUS OF THE ALERT CAT

.

J. Delgado-Garcia, R. Baker & S.M. Highsteín+

Department of Physiology

New York University Medical center, New York, N.Y.

and

+Departnent

of Neurosciences

Kennedy Center for Research

Albert Einstein College of Medicine, Bronx' N.Y.

INTRODUCTION AND METHODS

The injection of HRP into the nedial rectus subdivision of the oculomotor complex

has suggested the presence of a specific population of non{otoneuronal, internu-

clear neurons vithin the abducens nucleus of the ca! 9t l0 1Bütt."r-Ennever and

Graybiel, this volume), Subsequent electrophysiological studies demonstrated that

these internuclear neurons responded like abducens motoneurons to orthodromic vesti-

bular and reticular inputs 2'3 ¡ttigh"t"i., this volume). The internuclear neurons

have been shown to lerninate on contralateral medial rectus motoneurons exclusively

in an excitatory fashion 1l 1ttigh"a.i. and Baker, this volune). Idhen the latter

studies are combined vith recent work showing horizontal burst-tonic activity in

the MLT t5' t6 an.. internuclear neurons are strongly inplicated as the cells of

origin for the latter actjvity.

Houever, it is clear that only by recording Int

responses within the Abd nucleus could the latter inference be tested. More signi-

ficantly, by actually doing such experiments, ve could place in better perspective

príor work describing the deficits in horizontal eye movement following MLI' le-

.

4. 5. 6.7

s rons

In o¡der to clearly identify and characterize both internuclear (lnt) neurons

and abducens (Abd) notoneurons \1e employed lhe alert cat paradigm complete with

chronically inplanted stimulating electrodes on the left abducens nerve and within

the right medial rectus subdivísion of the oculomotor courplex. Methodological de-

rails wíll appear in a future paper in which Ehe activity of all antidronically

identified neurons from the Abd nucleus will be presented. We have recorded from

more than 80 antídromically isolated extracellular units in the left Abd nuclei of

Lhree cats, whose activities to date are for the most part sti11 being analyzed.

The present report will center itself around fhe qualitative and quantitative re-

sponse profile of one Int neuron selected from an experinent in vhich propitious

localization of both aforementioned stinulating electrodes and the recording site

has already been ascertained histologically.

Spontaneous and/or voluntary induced

eye movements were enployed in this study and a special aLterpt vas made to initiate

both versional and vergence eye movements. Individual right and left eye oculograms

(Frg. 2- 7) were recorded and horizontal gains were adjusted to be of equal nalni-

292

tude. The vertical EOG was recorded from the right eye.

RESÜLTS AND DISCUSSION

The Int neuron illustrated in Lhis paper was selected from an experimenL in which

more than 100 microelectrode tracts were made in and around the lef! Abd nucleus

over a period of one month. In brief sumary, we conclude that within the confinés

of the Abd nucleus as defined by the antidronic field potential depth profile, there

isnit any unit activity qualitatively distinguishable from that of either an Abd Mn

or an Int neuron. This evaluaLion is in direct reference to identified cellular

(i,e. somatic) recording sítes (see Fig. l). 0n the other hand, in the dorsal part

of the Abd nucleus (noticeably posterior or lateral), we often isolated units which

could not be antidromícal1y identified by stimulation of either the MR or the Abd

nerve. These responses were qualitatively different from those seen in Abd Mns and

and Int neurons and comonly consisted of either horizontal (out of phase) or ver-

tica1, burst-tonic responses. In addition, clear burst responses often associated

vith either horizontal or vertical eye movement vere frequently recorded. Since we

now are aware that a non-antidronically identifiable cell may be easily overlooked

(e.g., the Int neuron itself for many years l), then there may be yet another con-

A rR +l

a(il

ln ,.J

.2m¡cc

D

B

c

ABD

1

I'ig. l. Antidrornic identification of an Int neuron within the Abd nucleus. A, MR

sLinulaLion at ¿xon-straddLing intensity.

Double shock MR stinulation at 1.2 x Thr

(B) and 3 x Thr (C). Stimul-ation of the left Abd nerve (in D and E) indicated by

fi11ed círcles antl right MR subdivision bÍ-!-tars (axon-straddling intensity in A

and E). Polarity and time as indicated. Description ín text.

E

I

1

293

cealed response profile within the Abd nucleus. Even so, afrer carefully sanpling

all the activity that we could isolate from every glass microelectrode rrack througb

the Abd nucleus (i.e. concertedly looking for disparate responses), we surrepti-

tiously conclude that other neuronal patlerns than that described below renain

h i dden.

The Int neuron depicted in Figs. 1-7 was recorded dorsally in a microelecLrode

track near the center (anterior-posterior) of the Abd nucleus and was itself iso-

lated from a cluster of Int units. It is our physiological feeling that Int neur-

ons occur in snall groups within the Abd nucleus (similar for Abd Mns also). The

all-or-nothing activation of the Int neuron is shom at axon-straddling intensity

Risht

l ro'

I

Left

Fig. 2. Activity of Int neurori during conjugate saccadic eye movements.. Descrip-

tion in text.

Up

lro'

I

Down

Right

Ito'

.l

Left

Up

],0'

Down

I'ig. 3. Close correlation of Int activity

interval of right MR subdivision indicated

to versional

by stars.

eye movement. Stimul-ation

294

stimulation of the MR subdivision in Fig, lA. If we assurae a conduction distance

of I cm and a stimulus-response latency of .3 msec, then the rnt unit is certainly

antidromically activated (with a velocity of 33m/sec), Double shock MR stimulation

al 1 '2 x Thr indicates a ninimum interval of I msec (B) and at 3 x Thr the interval

is reduced to .6nsec(c).

The arrows in B and c show thaL the stressful double ,.

shock sLimulation produces a clear IS-SD inflection on the extracellular profile.

In order to demonstrate lhat this antidromically activated InL neuron did not have

axon-collaterals extendlng out into the Abd nerve MR stimulation was preceded by an

Abd nerve stiuulus. At various condition test intervals (D-E) the Int response was

always superimposed on the Abd field potential indicating no occlusion. hre never

recorded any Abd Mns or Int neurons which could be activated by stimulation of both

the Abd nerve and MR subdivision 2 ¡Maciewicz and Spencer, this volume). Therefore

we conclude that there are aL least two distinct classes of neurons within the Abd

nucleus which nay be differentiated on the basis of their activation following stim-

ro0r0

Eye Polition [d69]

Fig. 4, Firing frequency of the Int neuron during horizontal and vertical eye move-

ments. Discharge rate is plotted againsr RH (right eye), LH (left eye) and V (ver*

tical) fixations. In the plot on the lower right, the results shown in the upper

Ieft. diagran (RH) were represented with two regression lines which best fit the data

points. Imediately above and to the left of the regression lines a number of vo1-

untary saccades initiated are indicated by horizontal línes. Eye position aE onset

is indicated by a fi11ed circle and terminal eye position by an arrow. Intraburst

frequency is thus represented by the 1evel of the horizontal line.

E

295

ulation of either the Abd or MR subdivision of the oculomotor complex.

The activity of the Int neuron antidromically identified in I'ig. I is depicted

during a sequence of versional eye movements exhibiting frequent saccadic activity

(in Fig. 2) and acconpanying longer fixation periods (in Fig. 3). The near pelfect

conjugacy of the horizontal eye movefients (especially Tig. 3) is indicatecl by the

close correlation between Ehe right and left eye EOG records. Quite noticeable is

the constant relationship betr^reen the Int neuronal response and all aspects of the

eye movements. The Int unit pauses distinctly for all rightward saccades and ex-

hibirs a high-frequency bursr for leftward saccades (fi1led stars in Fig. 2 and

arrows in l.ig. 3).. In addition, a cháracteristic maintained firing frequency is

associated with periods of fixation with the on-direction being leftward' gaze. Such

correlation is outstanding in Fig. 3 which contains uninterrupted periods of fixa-

tion following brisk symetrical versional eye movements. Finally as indicated by

the stars in Fig, 3 the Int neuron could be antidromically activated lhroughout the

normal sequence of eye movements even in its silent period during eccentric riglrt-

ward gaze (third star from left).

The distinct correlation between the ac¡ivity of the Int neuron and movements of

the right as opposed to lefL eye are shown in I'igs, 4 and 5, As illustrated by the

upper two i¡sets in Fig. 4, the,firiog frequency of the Int neuron (ordínate) is

closely correlated to changes in eye posítion (abcissa)" In the latter case' aver-

age discharge was determined from consecutive 100 msec bins of activity during the

periods of naintained fixation. Certainly, the plot of firing frequency versus

either the right or left EOG would indicate a clear correlation in lespect to eye

position; hovever, the scatter of the data points is clearly less when the firing

frequency versus eye position was chosen for the right (RH) as opposed to the left

(LH) eye. Furthermore' as may be seen in the lower left inset (V), there is clearly

1itt1e correlation between firing frequency anal different l-evels of vertical fixa-

tion. Finally, the dot scattergram in the upper left inseE (RIl) was fitted by tvo

regression lines best describing the data points. Such would not have been the case

if the data points reflecting lnt firing frequency versus left eye position were

chosen instead. Interestingly, the two regression lines intersect at approximately

a central gaze position. The latter vas arbitrarily defined as the horizontal- eye

position during attenpted central gaze (Cromrelinck et a1" this volume)' This re-

sul! suggesls that there may be two monotonic firing frequencies for the Int neuront

one extending from a central gaze position to 20o in the off-direction and the other

to 20o in the on-direction. Sinilar data from othel Int neurons has 1ed to several

generalizalions regarding the role of the Int population in horizontal gaze' which

r¡i1l be presented ín a later paper (Delgado-Garcia et al', in preparation)' More

important fot the present context is the observalion that nearly all Int neurons

exhibited eye position thresholds lover and maxinun firiog frequencies higher

-than

Abd Mns. In addition, plots of eye position versus firing frequency for Int neurons

296

always shoüed a steep non-linear increase (as illustrated by Fig. 4) in contrast to

a linear relationship exhibited by Abd Mns.

During saccadic activity, modulation of Int neuron discharge preceded the sacca-

dic eye movement in the on- and off-directions by averages of l0 and 20 msec respect-

ively. We conclude thaE Int spíke activity thereby actually precedes the responses

observed in Abd Mns on the average of 3-5 msec. The clear velocity sensitivity of

the rnt neurons is depicted schernatically in the lower right inset of Fig. 4. The

initial eye position preceding a saccade has been indicated by a filled circle.

This point is then connecEed via a line to an arrowhead. and Lhe 1evel of the line

indicates intraburst frequency during the saccade. This diagrm shows that intra-

burst frequen.y ir.t"r""" as a function of the saccade being initiated from an eye

position more in an on-direction but is independent of the duration of the saccade.

Final1y, intra-burst frequency is nearly constant throughout the saccade.

The nearly perfec! correlation of Int neuronal activity from the left Abd nucleus

with movement of the right eye was completely confirmed by seeking out isolated

REOG t¿*1-

ñr¡rErf

LEOG

c

A

'M

-

Risht

l::;

Fig. 5. Int neuron response during rnonocular ihanges in horizontal rye position.

A-D, selected records from Int neuron identifíed in Fig. l.

%

t:

\i\ñ

120"

J

.l sec Down

297

of the Int neuron. As demonstrated in Fig. 5 (A and B) isolated monocular eye

movements by the right eye, Lo the right (indicated by curved arrows), are clearly

assocíaLed with pauses in Int neuron activity. In I'ig. 5D, a leftward, monocular

saccade by the left eye was clearly not associated vith either a burst or a subse-

quently naintained increase of firing frequenc¡r during the fixation (curved arrow).

Fig. 5C shows a versional (conjugate) followed by vergence eye movement in which

the InL activity continued to íncrease gradually during the very slow velocity sac-

cade (indicaled by double doumward curved arrow). This occurs in spite of a clear

rightward deflectíon of the left eye to another maintained gaze position further in

the off-direction.

An absolutely Darvelous sequence of records was obtained from this Int neuron

(fig. 6), which totally demonstrates the close correlation of Int neuronal activity

selectively with conjugate eye movement and not vergence eye movements. To begin

with, Abd l{n firing was modulated during all vergence and versional eye movements

as has been previously reported 12' 13. In contrast, Int neuron firing frequency

was not altered during or following the change in eye position accompanying neither

complex sequences of vergence eye movements (fig. 6) nor those of the more classical

type. The latter class of movments consists of l-4' amplitude convergence-diver-

gence sequences with 50-100 nsec time course for the step and 200-500 msec maintained

periods of fixation (fig. 7). As nay be seen by beginning on the lefr side of Fig.

6, normal versional eye movements are performed and then suddenly (indicated by

first set of outward curved arrowg) a slow velocity divergence occurs. Note the

absence of any modulation in discharge rate of the Int neuron. This sequence is

followed by two step-like vergence eye movements, then a conjugate rightward eye

movemenL again followed by a divergence sequence and finally a resumption of con-

jugate activity.

f'rom this data it is clear that at no tine during either the slow

or high velocity saccadic-like or step changes in right-eye position there occurs

a correlative modification of Int neuron activity, Thís is best seen by comparing

two comparable amplitude conjugate saccades (indicated by fi11ed circles on the left

in Fíg. 6) with those during vergence (unfi11ed circles on rhe righr). Sinilar1"y,

Risht

I zo.

lr.u

,uo

I

20'

Down

Fig. 6. Int neuron response to versional and vergence eye movements. Description

in text.

I 36c

298

the initial two saccades shom in the REOG of I'ig. 7 may be compared to those taking

place during the vergence movement (on the right side). Secondly, by conparing the

firing frequency of Int neurons for eye positions achieved following a conjugaEe

versus a vergence movement in either Fig. 6 or 7, one distincLly nay appreciate

fhe absence of a step change in firing frequency following the latter moveúent as.

compared to the former. This is especially clear in Fig. 7, where, in fact; the

new eye position is now more than 3" further in the on-direction (left); but there

is actually a decrease in the firing frequency of the Int neuron. Finally, at the

intervals indicated by the stars in Fig. 6, this Int neuron continued to be anti-

dromically activated following MR stimulation and such activation did not interfere

at all with the on-going sequence of conjugate or vergence eye movements,

As a general sumary statement concerning the present study, we would like Lo

mention that not all Int neurons exhibited sinilar response profiles to that depicted

for the Int neuron in this paper. For instance, during the higher velocity ver-

gence movements some Int neurons were modulated for both the on and off stéps, but

did not reflect a new discharge rate accompanying the fixation. In addition, Int

nodulation was more pronounced duriog sinusoidal vergence Eracking (Delgado- Garcia,

in preparation). Other characteristics often varied and sone of the parameters

already mentioned, like recruitment threshold, position sensitivity, conduction vel-

ocity, etc., are being examined for statistical correlation before final corments

are mde. In spite of these variations, the uniformity of the data pemics us to

propoae that vergence eye movements are definitely produced by a separate supra-

nuclear source to MR Mns in contrast to the opposíte opinion expressed in prior

t3. l4

la terature

We now feel that a convincing dernonstration for tvo separate populations of

neurons within the confines of the Abd nucleus exhibiting qualitatively similar

characteristics has been successfu1ly conpleted. Irom the response profiles as

described in this paper, we believe that the sumise of Fuchs and Luschei (1970)

(based on no available evidence af the tine), that a population of interneurons

were mixed LTith that of Abd l"lns, turns out to be a prophetic one indeed; unfortunate-

1y, this view was not always shared by others 3' 5' 13' 14' 17' 18. The most sig-

R¡ght

]'o'

left

Up

]

,0"

Down

I sec

Fig. 7. Int neuron activity during a symetríca1 vergence movement.

299

nificant observation contained within the present study is that only qualitatively

similar responses to that of Abd motoneurons were obtained; hovever, quantitatively

characteristic differences separate the two populations of neurons (sone have been

mentioned, and others wíll be detailed in a later paper; Delgado-Garcia, in prepara-

tion) .

Since recent HRP experiments have indicated that axons of al-1 Int neurons cross

the midlíne at the level of the Abd nucleus and ascend (without axon-co1laterals)

in the dorsal uredial part of the contraláteral MLF to terminate exclusively in an

excitatory fashion on MR Mns 2' 9' lo' ll, r.

"or.hde,

inferentially, tbat this

population of ce11s represents the horizontal burst-tonic fibers recorded in the

MLF 15' l6 tod th.r.fore is the source for relaying conjugate as opposed to vergence

horizontal eye movement signals to contralateral MR Mns for all saccadic (vestibular,

voluntary, or optokinetic), pursuit ancl fixation eye moverents

l5' l6' 18.

Gíven

the absence of any other horizontal sígna1 recorded from the ascending ¡&F 15' 16

and the recent data indicating that the secondary excicatory vestibular neuronal

pathway to the MR Mn is via the Ascending tract of Deiter's in the cat I and monkey

(Büttner-Ennever, personal comunication; note that the latter tract is separate

from the ML¡') then we suggesL Ehat it may possibly be superfluous to carry out fur-

ther quantítative studies of Int neurons, because their activity (i.e. MLF burst-

toníc fibers) during all horizontal- induced eye movements has been reported

so extensively by King et a1. (1976) and by Pola and Robinson (1977). luture em-

phasis is better spent on ascertaining unifornity and/or variation in response pro-

files within the Abd nucleus to better elucidate the classes of activity residíng

Eherein. Finally, lhere can now be little doubt that the deficit produced by an

interruption of the MLF represents the absence of ascending physiological activity

in the Int neurons of the Abd nucleus

4' 6' 7

and is not due to either direct com-

promise of ascending ponto-Í€dullary reticular or other posterior brainstem nuclear

axons .

ACKNOTLEDGEMENT

Research was supported by USPHS research grants NS - 13742 and EY- 02007.

REFERENCES

l. Baker, R. and Highstein, S.U. (1977), submitted for publication.

2. Baker, R. and Highstein, S.U. (1975) Brain Res. 9l'. 292-298.

3. Baker, R.G., Mano, N. and Shimazu, H, (1969) Brain Res. 15: 577-580.

4. Bender, M.B. and l.leinstein, E. (1944) Arch. Neurol. Pyschíatr.52: 106-113.

5. Carpenter, M.B. and McM¿sters, R.E. (1963) Arch. Neurol. 8: 347-368.

6. Cohen, B. (1971) Control of Eye Movements (Bach-y-Rita and Co11ins, Eds,),

Ner^¡ York Acadenic Press, 105-148.

7. Evinger, L.C., f'uchs, A.F., and Baker, R. (1977) Exp. Brain Res. 28: l-20.

8.

Fuchs, A.tr'. and Luschei, E.S. (1970) J. Neurophysiol. 33: 382:392.

300

9. Graybiel, A.M. (1977) Journal Conp. Neurol, 175: 37-78.

10. Graybiel, A.M. and Hartwieg, E.A" (1974) Brain Res. 8l: 543-551.

ll. Highstein, S.M. and Baker, R. (1971), submitted for pubtication.

12. Highstein, S.M., Maekarira, K., Steinacker, A. and Cohen, B. (1976) Brain Res.

ll2z t62-167

13. Ke1ler, E.L. (1973) Vis. Res. l3: 1565-1575.

14. Keller, E.L. and Robinson, D.A. (1972) Vis. Res. l2z 369-382.

15. King, W.M., lisberger, S.G. and Fuchs, A.T. (1976) J. Neurophysiol. 39: lt35-

1149..

16. Pola, J. and Robinson, D.A. (1977) J. Neurophysiol., in press.

ll. Schiller, P.H. (1970) Exp. Brain Res. l0: 341-362.

18. Skavenski, A.A. and Robinson, D.A. (1973) J, Neurophysiol. 36: 724-738.