- Email: [email protected]
Visual Substitution of Labyrinthine Defects
Visual Substitution of Labyrinthine Defects J. H. COURJON and M.JEANNEROD Laboratoire de Neuropsychologie Exp6rimentale, I.N.S.E.R.M. Unit6 94, 69500 Bron (France)
INTRODUCTION Labyrinthine and visual inputs contribute in maintaining the original angle of the retina in space when the body is tilted laterally. In the static conditions, counterrolling of the eyes compensates only for a small amount of the tilt (Krejcova et al., 1971) and preservation of the retinal angle is mostly achieved by a righting of the head. Magnus (1924, 1926) in his extensive studies in cats and dogs had identified labyrinthine and optical head righting reflexes, which he considered as additive components of the same predetermined reaction. However, a more modem conception of visual and vestibular interactions tends to admit that postural adjustments could be achieved through either input, each one being able to substitute for the other. Adaptation of postural mechanisms to natural, experimental or pathological situations, which is now well documented (see Dichganset al., 1973a; Talbott, 1974)infact requiresalacgeamountof flexibility,where any available input may prevail in each particular case in achieving the correct adjustment. Accordingly, the aim of our study was to evaluate the role of vision in reestablishing a normal head posture after the input from one of the two labyrinths has been suppressed. Hemilabyrinthectomy is known to produce a strong but transient postural asymmetry, which is easily recognizable in all studied vertebrates since the classical work of Flourens (1842), Ewald (1892) andMagnus (1924). In the cat, the postoperative syndrome is first marked by a critical stage lasting more than 24 and less than 48 h. The animal cannot stand up and lies on the side corresponding to the lesioned labyrinth. With a right hemilabyrinthectomy its body axis will be tonically bent to the left and twisted in the clockwise direction, so that its head will be tilted right side down. Within about 2 days, the animal is able to walk though its attempts are limited to inept leftward circling. After the third day, its head-body axis tends to become rectilinear. After 5-8 days gait is almost normal except for a hypotonia of the right limbs, and for occasional falls on the right side during spontaneous head-shakings. Finally, lateral head tilt may persist for several weeks or months. A study of spontaneous head posture and of righting of the head after hemilabyrinthectomy thus appears to be a good way to dGscribethe time course of postural compensation. We have used this index in animals rebvering in normal conditions, and in animals submitted to a temporary exclusion of visual input during the postoperative stage. By varying the duration of the visual deprivation period, and its position in time with respect
784 to the operation, it becomes possible to alter the normal process of recovery and consequently to assess the role of vision in re-acquisition and/or in maintenance of the new postural compensation (see also Putkonen et al., 1977). METHODS Experiments were conducted in adult cats. Destruction of the right labyrinth was performed under Nembutal anaesthesia. The bulla was opened through a ventral approach, and the bony labyrinth was destroyed under visual control with a dissecting microscope. After the operation, the animals were directed to one out of two experimental groups. In the L (light) group (3 cats), hemilabyrinthectomized animals were kept in cages in a normal laboratory environment. Normal illumination was provided during day hours. In the D (dark) group (3 cats), animals were put in a light proof room immediately after surgery. Total darkness was severely controlled, including during feeding and cleaning procedures. Duration of the postoperative dark period was varied from 10 days to 6 months according to different animals. The same dark room was used when animals from either group were put in the dark for short periods (2-15 days) during the late course of recovery. Head posture was measured in standard conditions by using serial photography of cats' heads-Theanimalswereplacedinabox(45 x 1 4 x 18cm)leavingtheheadfreetomove. The box, at a fixed distance from the camera, could be placed either in the upright position (O"), or tilted 45" to the right (+45") or to the left (-45"). For each session, its position was alternated in a fixed sequence (O", +45", -45") repeated 3 times. The cat's head was photographed in each position of the box (9 photographs per session). Sessionswere repeated at fixed intervals of time (see below). In the case of animals from the D group, the same procedure was used in the dark, except that single brief flashes of light were given to take the photographs. However, to ensure a further control of the lack of visual input during the photography sessions, two animalsfrom thisgrouphad thelidssuturedonthesameday as the hemilabyrinthectomy. Finally, unlesioned animals were also photographed in the box, in order to obtain control values of head posture and righting reflexes. Each photograph was then analysed by tracing the interocular axison the cat's face, and by measuring its angulation with respect to the physical horizontal. Values obtained for a given position (e.g., upright 0") were averaged for each session. RESULTS Head posture and righting reflexes in normal cats Normal cats tend to keep a symmetrical head posture when they have their body in the upright position. However, during static lateral body tilt, the head does not fully compensate for the inclination, and remains undemghtened. Fig. 1A shows data from four intact animals observed under normal illumination. For a 45" body tilt to the right or to the left, the head rightens on both sides by about 25" on the average, which represents only a 5096 compensation of body tilt. Undemghting of the head might be partly explained by the existence of a counter rolling of the eyes, thus compensating for
785
I
B
450, 0" ,-45',
40
+
45
0
B 0D Y
- 45 P0S I T I 0 N
+
45
0
- 45
(DEGREES)
Fig. 1. Head position re body position in normal and hemilabyrinthectomizedcats. A) Mean head position and SD in 4 normal cats tested in the box as indicated by the drawings in the lower row (redrawn from original photographs). +45, 45" body tilt to the right; 0, upright body position; -45, 45" tilt to the left. B) Same data from 2 right hemilabyrinthectomizedcats, tested on the second postoperative day, i.e., the day with the maximum head tilt.
the residual head tilt, and bringing the retinal coordinates back to their original position. Though we have not measured ocular torsion in our animals, we know from experiments in monkeys by Krejcova et al., (1971) that under static conditions it should not exceed 10% of the amount of head tilt. In the dark also, righting of the head in response to body tilt is incomplete. This is less surprising, however, since the need for a preservation of retinal position in space no longer exists. Postoperative evolution of head posture in cats from the L group After hemilabyrinthectomy, the normal pattern of static head posture in relation to body position was dramatically altered. When the cats had the body in the upright position, the head appeared permanently tilted to the right (lesioned) side. The amount of spontaneous head tilt was usually maximum on the second postoperative day. Concomitantly, the head righting was abolished when the body was tilted to the right. In this case, the resulting head posture was a passive addition of the spontaneous head tilt, plus the body tilt. By contrast, the righting reflex was preserved, at least partly, when the
786
B (7)
,+45:
(100)
U
0..-15’,
BODY
POSITION
Fig. 2. Postoperative evolution of head posture in one hemilabyrinthectomid cat, recovering in normal conditions (L group). A) Postoperative evolution of head tilt with the body in the upright position. B) Head position re body position in postoperative days 2 , 7 and 100.
body was tilted tothe left. Fig. 1Bshows averagevalues of head tilt in the three positions of the box, obtained in two animals from the L group on their second postoperative day. Recovery from thisposturaldeficitwas first marked by asteepdecreasein spontaneous head tilt during the first 7-10 postoperative days, down to close to normal valuesof static head position. Righting reflex partly reappeared in response to body tilt toward the lesioned side. This postural improvement was only temporary, however. As exemplified by the animal shown in Fig. 2, static head-tilt reincreased during the third postoperative week, up to thevaluesof the firstdays, andrightingreflexesdeteriorated.From this point, a long-lasting process took place, leading progressively to a stable compensationwithin about 3months(Fig. 2). Indication of such a two-stage recovery processwas alsofound in the other animals of the L group. Examinationof hemfiabyrinthectomizedcats at avery late postoperative stage (1year) revealed a small residual head tilt to the right (average value in 4 animals: 9.4”, see Fig. 5B). Postoperative evolution of head position in cats porn the D group Lack of visual input during the postoperative stage, resulted in “freezing” postural recovery at the level of the first or second postoperative day. In the two lid-sutured cats, the postlabyrinthectomysyndrome could be observed as described in the Introduction, throughout the whole period of light deprivation. Once vision was restored (onthe 16th and 28th days, respectively), compensation began immediately and proceeded with an accelerated time course. Head posture was examined in these two animals during the dark period. It was found that the initial head tilt in the upright body position was less pronounced than in animals using visual cues during recovery. However a progressive deterioration occurred over
787
A
z c
-
5
10 POST
20
15
O P E R A T I V E
28
35
m,$. (28 D)
50
DAYS
B
(28 L)
BODY POSITION Fig. 3. Postoperative evolution of head posture in one hemilabyrinthectomizedcat recovering in the dark ( D group). A) Post-operative evolution of head tilt with the body in the upright position. Dashed line on day 28 On indicates the return to thelight condition. B)Headpositionre bodypositioninpostoperativedays2,28,50. day 28, compare the data obtained in the dark (28 D) with those obtained in the light, a few min later (28 L).
time, so that, finally head tilt peaked at 42.7" on the 28th day in one animal, and at 112.5" on the 16th day in the other. This evolution is shown for the first animal in Fig. 3A. Head righting in response to rightward body tilt was absent during the first days. It tended to improve, however, in that the resulting value of head tilt in that position was less than the summation of spontaneous head tilt and of body tilt (Fig. 3B). Thus, the pattern of head posture in animalsof the D group at the end of the dark period differed somewhat from that of animals of the L group during the critical stage. Though they still had their head strongly tilted in the upright body position, the righting in response to lateral body tilt appeared to be normal in amplitude, in both directions. Restoration of normal visual conditions (which had been preceded two days before by a reopening of the lids under a small dose of ketamine) resulted in an immediate decrease in head asymmetry in the upright body position. In the animal shown in Fig. 3, a few min spent in the light were sufficient to restore a quasi-normal pattern. However, this improved level of postural symmetry was not stable and could not be maintained. It deteriorated within 1or 2 days and finally improved progressively over about 2 weeks, before compensated values of head posture could be reached and maintained. This temporal pattern was very reminiscent, though with a shorter time course, of the two stage recovery process observed in the animals from the L group. In a different animal, the duration of the dark period was prolonged for 6 months. Spontaneous head tilt in the upright body position was about 50" after 3 months and about 55" after 6 months (Fig. 4), a value which would never be observed at this stage in animals recovering in normal conditions. Righting reflexes were also very poor and asymmetrical. Restoration of visual input produced an immediate improvement in head posture, to a lesser extent, however, than after shorter visual deprivation periods.
788
':h
B (188)
20 0
1
,45,
0 ,-IS,
BODY
POSlTlOW
Fig. 4. Postope~ativeevolutionofheadpostureinonehemiabyrinthectomizedca thedarkfor6months. A) Postoperative evolution of head tilt with the body in the upright position. Dashed tine on day 180 indicates the return to the tight condition. B) Head position re body position on postoperative days 90,180, 188. On day 180 compare the data obtained in the dark (180 D) with those obtained in the light (180 L).
Role of visual input in maintenance ofa symmetrical head posture in compensated cats Animals from either group which had reached a sufficient level of postural recovery, were submitted to short (2-15 days) periods of visual deprivation.As arule, head posture was examinedin the dark on the last day of the period, and in the light on the first and the following days after restoration of normal vision. In all cases, a deterioration of postural symmetry was observed at the end of the dark period. This effect occurred irrespectivelyof the duration of the visual deprivation, or of its position in postoperative time. For example, Fig. 5A shows the effect of a 15 day dark period in one animal, at the end of the second postoperative month. Head tilt had increased by a factor of 2 or 3 at the end of the period. An effect of a similar amplitude could be obtained in another animal after a dark period of only 2 days. Four of these animals underwent another light deprivation period (duration, 10 days) more than 1year after hemilabyrinthectomy.Mean angular position of the head in the four animals (as measured in the upright body position) increased from 9.4" prior to the deprivation period up to 24" after 10 days (Fig. 5B). The specificity of the role of vision on this effect was ascertained by the immediate return to predeprivation values of head posture, when normal vision was restored. DISCUSSION Postural asymmetry following hemilabyrinthectomy reflects the imbalance between the activityofthe vestibular nuclei on the twosides. According to Precht et al. (1966) and McCabe et al. (1972), the activity of the nuclei on the deafferented side is strongly
789 ........-
-
52
0
c
-
60
A
68
76
O P E R A T I V E
POST
O P E R A T I V E
108
100
92
84
POST
DAYS
I
360
370
372
374
376
DAYS
Fig. 5. Effect of late dark exposure on the postoperative evolution of head posture in hemilabyrinthectomized cats. A) Effect of a 15 day dark exposure in one animal between postoperative days 52 and 67. Compare the 2 values of head tilt obtained on day 67, in the dark (0) and in the light (0). B) Effects of a 10 days dark exposure in 4 animals 1 year after the hemilabyrinthectomy. 0 , average value of head tilt (and SD) in the light. 0 , in the dark. Compare the two values obtained on day 370.
depressed after the operation, while in the contralateral nuclei the resting discharge increases as a result of a reduced contralateral inhibition. Hence, the hyperactivity of neck muscles on the side of the lesion, leading to the head tilt, may be explained by the increased discharge from the intact vestibular nuclei, via crossed excitatory vestibulospinal projections onto neck motoneurons (Wilson and Yoshida, 1969; Wilson and Peterson, 1978). Compensation in normal postoperative conditions begins during the first week with a strong inhibition of the resting activity in the vestibular nuclei on both sides. The evolution is then marked by a progressive regeneration of activity in the vestibular nuclei on the deafferented side (presumably via an increase in excitatory synaptic efficacy of the commissural system) (Dieringer and Precht, 1977), and later by a disinhibition on the normal side (Precht et al., 1966; Precht, 1974). These neural changes may represent the substrate for the 2-stage recovery pattern observed in cats from the L group. The first stage would correspond to a rapid motor learning, tending to reduce a critical dysfunction through a massive inhibition of postural reactions. The second, long-term stage would correspond to the learning and to the fixation of new motor sets, based on the reorganization of activity within the vestibular system (Llinas et al., 1975). Postoperative evolution of animals from the D group seems to indicate that normal recovery heavily relies upon visual cues. First, the fact that postural asymmetry is less marked during the early postoperative stage in animals kept in the dark shows that the
790 visual factor aggravates the motor imbalance produced by the lesion. One possible explanation for this difference is that hemilabyrinthectomy would also bias central mechanisms responsible for the detection of visual coordinates (Bisti et al., 1972), thus resulting in an increased head tilt. If this hypothesiswere correct, then it would become questionable whether the rapid decrease in postural asymmetry observed in the L group is due to motor learning per se, or to some adaptation within the visual system. Second,we know from the animalsof the D group that the motor imbalance cannot be compensatedfor in the absenceof vision even if visual deprivationhas been prolongedfor as long as 6 months. This fact shows that the long-term learning which occurs during the second stage of normal recovery also requires staticvisualcues in order to re-equilibrate and to stabilize the postural system. Finally, whether postural symmetry is recovered normally or whether it is "frozen" during visual deprivation, we know from our experiments that the visual vestibular interactionresponsible for postural compensationremains fully flexible. Restoration of visual input at any postoperative stage in animalsfrom the D group is alwaysfollowed by recovery. We have no reason to believe that a visual deprivation longer than 6 months would not be followed alsoby some degreeof recovery. On the other hand, deprivationof visual input in already compensated animals invariably produces a deterioration of postural symmetry, even 1 year after the operation. This flexibility is a well known fact from the literature in animals (Ewald, 1892; Magnus, 1924; Dow, 1938; Schaefer and Meyer, 1974), as well as in man (Andre-Thomaset al., 1941), showing that blind-folding abolishes temporarily postural compensation after labyrinthine lesions. These experimental data lead to the conclusionthat vision isjust another input channel feeding into the vestibular system. In normal animals, electrophysiological evidence has been found that vision can influence vestibular neurons (Dichgans et al., 1973b; Azzena et al., 1974; Henn et al., 1974). An anatomical pathway has been identified, from the accessory optic tract and via the inferior olive and the cerebellar flocculus (Maekawa and Simpson, 1973) or more directly via the flocculus only (Brauth and Karten, 1977; Winfield et al., 1978), which may account for this influence. Our suggestionis that the gain of this pathway increasesunder pathological conditions, where vision becomes predominant and is able to fully substitute for the labyrinthine input. In fact, totally delabyrinthed animals may have normal righting reflexes when allowed to use their vision (Magnus, 1926). This would mean either that fibers carrying visual input to the brain stem form new synapseswith vestibular neurons responsiblefor postural control, or alternatively that already existing synapses are derepressed by the destruction of labyrinthine afferents. The latter hypothesis, postulated by Merrill and Wall (1972) for the somestheticsystemwould accountfor the rapid recovery when visual input is available to the animal.
SUMMARY The evolution of lateral head-tilt following hemilabyrinthectomyhas been studied in adult cats. Animalswere maintained postoperativelyin normally lit conditions(Lgroup) or in total darkness (D group). In cats from the L group, the head-tilt peaked at 45" (with the lesioned side down) on the second postoperative dAy, and decreased to about 0"within about 10 days. This evolution was followed by rebounds of head-tilt to larger angles before a stable
791
compensated head position could be maintained (approximately at the end of the third postoperative month). In cats from the D group, the head remained tilted by a large angle throughout the duration of the dark period up to 6 months. Re-exposure to light was followed by a rapid decrease of head-tilt. Finally, when put back in the dark at a late postoperative stage (up to 1year), already compensated animals were found to lose their symmetrical head position, and to re-acquire a strong head-tilt. This effect resumed on re-exposure to light. It is inferred that static visual input is a necessary condition for compensation of the postural deficits of hemilabyrinthectomy in the cat. Maintenance of a stable head posture also depends upon continuous availability of visual input. REFERENCES Andre-Thomas, Sorrel, E. and Sorrel-Dejerine, M. (1941) Fracture du rocher, troubles vestibulaires,attitude de la t&te,reactions meningees. Rev. Neurol., 6 k 7 3 . Azzena, G . B., Azzena, M. T. and Marini, R. (1974) Optokinetic nystagmus and the vestibular nuclei. Exp. Neurol., 42: 158-168. Bisti, S., Maffei, L. and Piccolino, M. (1972) Variations of the visual responses of the superior colliculus in relation to body roll. Science, 175: 456-457. Brauth,S.E.andKarten,M. J. (1977)Directaccessoryopticprojectionstothevestibulo-cerebellum: apossible channel for oculomotor control systems. Exp. Brain Res., 28: 73-84. Dichgans, J., Bizzi, E., Morasso, P. and Tagliasco, V. (1973a) Mechanisms underlying recovery of eye-head coordination following bilateral labyrinthectomy in monkeys. Exp. Brain Res., 18: 548-562. Dichgans, J., Schmidt, C. L. and Graf, W. (1973b) Visual input improves the speedometer function of the vestibular nuclei in the goldfish. Exp. Brain Res., 18: 319-322. Dieringer, N. and Precht, W. (1977) Modification of synaptic input following unilateral labyrinthectomy. Nature (Lond.), 269: 431-433. Dow, R. S. (1938) The effects of bilateral and unilateral labyrinthectomyin monkey, baboon and chimpanzee. Amer. J. Physiol., 121: 392-399. Ewald, J. R. (1892) Physwlogische Untersuchungen iiber das Endorgan des N. Oktavus. Wiesbaden, Bergmann. Flourens, J.P.N. (1842) Recherches ExpCrimentales sur les PropriktCs et les Fonctions du SystPme Nervew, duns les Animaw VertkbrCs. Baillike, Paris. Hem, V., Young, L. R. and Finley, C. (1974) Vestibular nucleus units in alert monkeys are also influenced by moving visual fields. Brain Res., 71: 144-149. Krejcova, H., Highstein, S. and Cohen, B. (1971) Labyrinthine and extralabyrinthine effects on ocular counter-rolling. Acta oto-laryngol. (Stockh.), 72: 165-171. Llinhs, R., Walton, R., Hillman, D. E. and Sotelo, C. (1975) Inferior olive: its role in motor learning. Science, 190: 1230-1231.
Maekawa, K. and Simpson,J. L. (1973) Climbingfiber response evokedin vestibulocerebellumof rabbit from visual system. J. Neurophysiol., 36: 649-666. Magnus, R. (1924) Korperstellung. J. Springer, Berlin. Magnus, R. (1926) Some results of studies in the physiology of posture. Lancet, 211: 531-536 and 585-588. McCabe, B. F., Ryu, J. H. and Sekitani, T. (1972) Further experiments on vestibular compensation. Laryngoscope, 82: 381. Merill, E. G. and Wall, P. D. (1972) Factors forming the edge of a receptive field: the presence of relatively ineffective afferent terminals. J. Physiol. (Lond.), 226: 825. Precht, W. (1974) Characteristics of vestibular neurons after acute and chronic labyrinthine destruction. In Handbook ofSensory Physiology, Vol. VIll, VestibularSystem, Part2, Psychophysics,AppliedAspects and General Interpretations. H. H. Kornhuber (Ed.), Springer-Verlag, Berlin, pp. 45 1 6 6 2 . Precht, W., Shimazu, H. and Markham, C. H. (1966) A mechanism of central compensation of vestibular function following hemilabyrinthectomyJ. Neurophyswl., 29: 996-1010. Putkonen, P. T. S., Courjon, J. H. and Jeannerod, M. (1977) Compensationforposturaleffectsof hemilabyrinthectomy in the cat. A sensory substitution process? Exp. Brain Res., 28: 249-257.
792 Schiifer, K. P. and Meyer, D. L. (1974) Compensation of vestibular lesions. In Handbook of Sensory Physiology, Vol. VIII. Vestibular System, Part 2, Psychophysics, Applied Aspects and General Interpretations, H. H. Kornhuber (Ed.), Springer-Verlag, Berlin, pp. 463-490. Talbott, R. E. (1974) Modification of the postural response of the normal dog by blindfolding. 1. Physiol., (Lond.), 243: 309-320. Wilson, V. J. and Yoshida, M. (1969) Bilateral connectionsbetween labyrinths and neck motoneurons. Brain Res., 13: 603-607. Wilson,V. J. andPeterson,B. W. (1978) Peripheralandcentralsubstratesofvestibulospinalreflexes.Physio1. Rev., 58: 80-105. Winfield,J. A., Hendrickson, A. and Kimm, J. (1978) Anatomicalevidencethat the medial terminal nucleus of the accessoryoptic tract in mammalsprovides avisual mossy fiber input to the flocculus.Brain Res., 15 1: 175-182.
INTRODUCTION Labyrinthine and visual inputs contribute in maintaining the original angle of the retina in space when the body is tilted laterally. In the static conditions, counterrolling of the eyes compensates only for a small amount of the tilt (Krejcova et al., 1971) and preservation of the retinal angle is mostly achieved by a righting of the head. Magnus (1924, 1926) in his extensive studies in cats and dogs had identified labyrinthine and optical head righting reflexes, which he considered as additive components of the same predetermined reaction. However, a more modem conception of visual and vestibular interactions tends to admit that postural adjustments could be achieved through either input, each one being able to substitute for the other. Adaptation of postural mechanisms to natural, experimental or pathological situations, which is now well documented (see Dichganset al., 1973a; Talbott, 1974)infact requiresalacgeamountof flexibility,where any available input may prevail in each particular case in achieving the correct adjustment. Accordingly, the aim of our study was to evaluate the role of vision in reestablishing a normal head posture after the input from one of the two labyrinths has been suppressed. Hemilabyrinthectomy is known to produce a strong but transient postural asymmetry, which is easily recognizable in all studied vertebrates since the classical work of Flourens (1842), Ewald (1892) andMagnus (1924). In the cat, the postoperative syndrome is first marked by a critical stage lasting more than 24 and less than 48 h. The animal cannot stand up and lies on the side corresponding to the lesioned labyrinth. With a right hemilabyrinthectomy its body axis will be tonically bent to the left and twisted in the clockwise direction, so that its head will be tilted right side down. Within about 2 days, the animal is able to walk though its attempts are limited to inept leftward circling. After the third day, its head-body axis tends to become rectilinear. After 5-8 days gait is almost normal except for a hypotonia of the right limbs, and for occasional falls on the right side during spontaneous head-shakings. Finally, lateral head tilt may persist for several weeks or months. A study of spontaneous head posture and of righting of the head after hemilabyrinthectomy thus appears to be a good way to dGscribethe time course of postural compensation. We have used this index in animals rebvering in normal conditions, and in animals submitted to a temporary exclusion of visual input during the postoperative stage. By varying the duration of the visual deprivation period, and its position in time with respect
784 to the operation, it becomes possible to alter the normal process of recovery and consequently to assess the role of vision in re-acquisition and/or in maintenance of the new postural compensation (see also Putkonen et al., 1977). METHODS Experiments were conducted in adult cats. Destruction of the right labyrinth was performed under Nembutal anaesthesia. The bulla was opened through a ventral approach, and the bony labyrinth was destroyed under visual control with a dissecting microscope. After the operation, the animals were directed to one out of two experimental groups. In the L (light) group (3 cats), hemilabyrinthectomized animals were kept in cages in a normal laboratory environment. Normal illumination was provided during day hours. In the D (dark) group (3 cats), animals were put in a light proof room immediately after surgery. Total darkness was severely controlled, including during feeding and cleaning procedures. Duration of the postoperative dark period was varied from 10 days to 6 months according to different animals. The same dark room was used when animals from either group were put in the dark for short periods (2-15 days) during the late course of recovery. Head posture was measured in standard conditions by using serial photography of cats' heads-Theanimalswereplacedinabox(45 x 1 4 x 18cm)leavingtheheadfreetomove. The box, at a fixed distance from the camera, could be placed either in the upright position (O"), or tilted 45" to the right (+45") or to the left (-45"). For each session, its position was alternated in a fixed sequence (O", +45", -45") repeated 3 times. The cat's head was photographed in each position of the box (9 photographs per session). Sessionswere repeated at fixed intervals of time (see below). In the case of animals from the D group, the same procedure was used in the dark, except that single brief flashes of light were given to take the photographs. However, to ensure a further control of the lack of visual input during the photography sessions, two animalsfrom thisgrouphad thelidssuturedonthesameday as the hemilabyrinthectomy. Finally, unlesioned animals were also photographed in the box, in order to obtain control values of head posture and righting reflexes. Each photograph was then analysed by tracing the interocular axison the cat's face, and by measuring its angulation with respect to the physical horizontal. Values obtained for a given position (e.g., upright 0") were averaged for each session. RESULTS Head posture and righting reflexes in normal cats Normal cats tend to keep a symmetrical head posture when they have their body in the upright position. However, during static lateral body tilt, the head does not fully compensate for the inclination, and remains undemghtened. Fig. 1A shows data from four intact animals observed under normal illumination. For a 45" body tilt to the right or to the left, the head rightens on both sides by about 25" on the average, which represents only a 5096 compensation of body tilt. Undemghting of the head might be partly explained by the existence of a counter rolling of the eyes, thus compensating for
785
I
B
450, 0" ,-45',
40
+
45
0
B 0D Y
- 45 P0S I T I 0 N
+
45
0
- 45
(DEGREES)
Fig. 1. Head position re body position in normal and hemilabyrinthectomizedcats. A) Mean head position and SD in 4 normal cats tested in the box as indicated by the drawings in the lower row (redrawn from original photographs). +45, 45" body tilt to the right; 0, upright body position; -45, 45" tilt to the left. B) Same data from 2 right hemilabyrinthectomizedcats, tested on the second postoperative day, i.e., the day with the maximum head tilt.
the residual head tilt, and bringing the retinal coordinates back to their original position. Though we have not measured ocular torsion in our animals, we know from experiments in monkeys by Krejcova et al., (1971) that under static conditions it should not exceed 10% of the amount of head tilt. In the dark also, righting of the head in response to body tilt is incomplete. This is less surprising, however, since the need for a preservation of retinal position in space no longer exists. Postoperative evolution of head posture in cats from the L group After hemilabyrinthectomy, the normal pattern of static head posture in relation to body position was dramatically altered. When the cats had the body in the upright position, the head appeared permanently tilted to the right (lesioned) side. The amount of spontaneous head tilt was usually maximum on the second postoperative day. Concomitantly, the head righting was abolished when the body was tilted to the right. In this case, the resulting head posture was a passive addition of the spontaneous head tilt, plus the body tilt. By contrast, the righting reflex was preserved, at least partly, when the
786
B (7)
,+45:
(100)
U
0..-15’,
BODY
POSITION
Fig. 2. Postoperative evolution of head posture in one hemilabyrinthectomid cat, recovering in normal conditions (L group). A) Postoperative evolution of head tilt with the body in the upright position. B) Head position re body position in postoperative days 2 , 7 and 100.
body was tilted tothe left. Fig. 1Bshows averagevalues of head tilt in the three positions of the box, obtained in two animals from the L group on their second postoperative day. Recovery from thisposturaldeficitwas first marked by asteepdecreasein spontaneous head tilt during the first 7-10 postoperative days, down to close to normal valuesof static head position. Righting reflex partly reappeared in response to body tilt toward the lesioned side. This postural improvement was only temporary, however. As exemplified by the animal shown in Fig. 2, static head-tilt reincreased during the third postoperative week, up to thevaluesof the firstdays, andrightingreflexesdeteriorated.From this point, a long-lasting process took place, leading progressively to a stable compensationwithin about 3months(Fig. 2). Indication of such a two-stage recovery processwas alsofound in the other animals of the L group. Examinationof hemfiabyrinthectomizedcats at avery late postoperative stage (1year) revealed a small residual head tilt to the right (average value in 4 animals: 9.4”, see Fig. 5B). Postoperative evolution of head position in cats porn the D group Lack of visual input during the postoperative stage, resulted in “freezing” postural recovery at the level of the first or second postoperative day. In the two lid-sutured cats, the postlabyrinthectomysyndrome could be observed as described in the Introduction, throughout the whole period of light deprivation. Once vision was restored (onthe 16th and 28th days, respectively), compensation began immediately and proceeded with an accelerated time course. Head posture was examined in these two animals during the dark period. It was found that the initial head tilt in the upright body position was less pronounced than in animals using visual cues during recovery. However a progressive deterioration occurred over
787
A
z c
-
5
10 POST
20
15
O P E R A T I V E
28
35
m,$. (28 D)
50
DAYS
B
(28 L)
BODY POSITION Fig. 3. Postoperative evolution of head posture in one hemilabyrinthectomizedcat recovering in the dark ( D group). A) Post-operative evolution of head tilt with the body in the upright position. Dashed line on day 28 On indicates the return to thelight condition. B)Headpositionre bodypositioninpostoperativedays2,28,50. day 28, compare the data obtained in the dark (28 D) with those obtained in the light, a few min later (28 L).
time, so that, finally head tilt peaked at 42.7" on the 28th day in one animal, and at 112.5" on the 16th day in the other. This evolution is shown for the first animal in Fig. 3A. Head righting in response to rightward body tilt was absent during the first days. It tended to improve, however, in that the resulting value of head tilt in that position was less than the summation of spontaneous head tilt and of body tilt (Fig. 3B). Thus, the pattern of head posture in animalsof the D group at the end of the dark period differed somewhat from that of animals of the L group during the critical stage. Though they still had their head strongly tilted in the upright body position, the righting in response to lateral body tilt appeared to be normal in amplitude, in both directions. Restoration of normal visual conditions (which had been preceded two days before by a reopening of the lids under a small dose of ketamine) resulted in an immediate decrease in head asymmetry in the upright body position. In the animal shown in Fig. 3, a few min spent in the light were sufficient to restore a quasi-normal pattern. However, this improved level of postural symmetry was not stable and could not be maintained. It deteriorated within 1or 2 days and finally improved progressively over about 2 weeks, before compensated values of head posture could be reached and maintained. This temporal pattern was very reminiscent, though with a shorter time course, of the two stage recovery process observed in the animals from the L group. In a different animal, the duration of the dark period was prolonged for 6 months. Spontaneous head tilt in the upright body position was about 50" after 3 months and about 55" after 6 months (Fig. 4), a value which would never be observed at this stage in animals recovering in normal conditions. Righting reflexes were also very poor and asymmetrical. Restoration of visual input produced an immediate improvement in head posture, to a lesser extent, however, than after shorter visual deprivation periods.
788
':h
B (188)
20 0
1
,45,
0 ,-IS,
BODY
POSlTlOW
Fig. 4. Postope~ativeevolutionofheadpostureinonehemiabyrinthectomizedca thedarkfor6months. A) Postoperative evolution of head tilt with the body in the upright position. Dashed tine on day 180 indicates the return to the tight condition. B) Head position re body position on postoperative days 90,180, 188. On day 180 compare the data obtained in the dark (180 D) with those obtained in the light (180 L).
Role of visual input in maintenance ofa symmetrical head posture in compensated cats Animals from either group which had reached a sufficient level of postural recovery, were submitted to short (2-15 days) periods of visual deprivation.As arule, head posture was examinedin the dark on the last day of the period, and in the light on the first and the following days after restoration of normal vision. In all cases, a deterioration of postural symmetry was observed at the end of the dark period. This effect occurred irrespectivelyof the duration of the visual deprivation, or of its position in postoperative time. For example, Fig. 5A shows the effect of a 15 day dark period in one animal, at the end of the second postoperative month. Head tilt had increased by a factor of 2 or 3 at the end of the period. An effect of a similar amplitude could be obtained in another animal after a dark period of only 2 days. Four of these animals underwent another light deprivation period (duration, 10 days) more than 1year after hemilabyrinthectomy.Mean angular position of the head in the four animals (as measured in the upright body position) increased from 9.4" prior to the deprivation period up to 24" after 10 days (Fig. 5B). The specificity of the role of vision on this effect was ascertained by the immediate return to predeprivation values of head posture, when normal vision was restored. DISCUSSION Postural asymmetry following hemilabyrinthectomy reflects the imbalance between the activityofthe vestibular nuclei on the twosides. According to Precht et al. (1966) and McCabe et al. (1972), the activity of the nuclei on the deafferented side is strongly
789 ........-
-
52
0
c
-
60
A
68
76
O P E R A T I V E
POST
O P E R A T I V E
108
100
92
84
POST
DAYS
I
360
370
372
374
376
DAYS
Fig. 5. Effect of late dark exposure on the postoperative evolution of head posture in hemilabyrinthectomized cats. A) Effect of a 15 day dark exposure in one animal between postoperative days 52 and 67. Compare the 2 values of head tilt obtained on day 67, in the dark (0) and in the light (0). B) Effects of a 10 days dark exposure in 4 animals 1 year after the hemilabyrinthectomy. 0 , average value of head tilt (and SD) in the light. 0 , in the dark. Compare the two values obtained on day 370.
depressed after the operation, while in the contralateral nuclei the resting discharge increases as a result of a reduced contralateral inhibition. Hence, the hyperactivity of neck muscles on the side of the lesion, leading to the head tilt, may be explained by the increased discharge from the intact vestibular nuclei, via crossed excitatory vestibulospinal projections onto neck motoneurons (Wilson and Yoshida, 1969; Wilson and Peterson, 1978). Compensation in normal postoperative conditions begins during the first week with a strong inhibition of the resting activity in the vestibular nuclei on both sides. The evolution is then marked by a progressive regeneration of activity in the vestibular nuclei on the deafferented side (presumably via an increase in excitatory synaptic efficacy of the commissural system) (Dieringer and Precht, 1977), and later by a disinhibition on the normal side (Precht et al., 1966; Precht, 1974). These neural changes may represent the substrate for the 2-stage recovery pattern observed in cats from the L group. The first stage would correspond to a rapid motor learning, tending to reduce a critical dysfunction through a massive inhibition of postural reactions. The second, long-term stage would correspond to the learning and to the fixation of new motor sets, based on the reorganization of activity within the vestibular system (Llinas et al., 1975). Postoperative evolution of animals from the D group seems to indicate that normal recovery heavily relies upon visual cues. First, the fact that postural asymmetry is less marked during the early postoperative stage in animals kept in the dark shows that the
790 visual factor aggravates the motor imbalance produced by the lesion. One possible explanation for this difference is that hemilabyrinthectomy would also bias central mechanisms responsible for the detection of visual coordinates (Bisti et al., 1972), thus resulting in an increased head tilt. If this hypothesiswere correct, then it would become questionable whether the rapid decrease in postural asymmetry observed in the L group is due to motor learning per se, or to some adaptation within the visual system. Second,we know from the animalsof the D group that the motor imbalance cannot be compensatedfor in the absenceof vision even if visual deprivationhas been prolongedfor as long as 6 months. This fact shows that the long-term learning which occurs during the second stage of normal recovery also requires staticvisualcues in order to re-equilibrate and to stabilize the postural system. Finally, whether postural symmetry is recovered normally or whether it is "frozen" during visual deprivation, we know from our experiments that the visual vestibular interactionresponsible for postural compensationremains fully flexible. Restoration of visual input at any postoperative stage in animalsfrom the D group is alwaysfollowed by recovery. We have no reason to believe that a visual deprivation longer than 6 months would not be followed alsoby some degreeof recovery. On the other hand, deprivationof visual input in already compensated animals invariably produces a deterioration of postural symmetry, even 1 year after the operation. This flexibility is a well known fact from the literature in animals (Ewald, 1892; Magnus, 1924; Dow, 1938; Schaefer and Meyer, 1974), as well as in man (Andre-Thomaset al., 1941), showing that blind-folding abolishes temporarily postural compensation after labyrinthine lesions. These experimental data lead to the conclusionthat vision isjust another input channel feeding into the vestibular system. In normal animals, electrophysiological evidence has been found that vision can influence vestibular neurons (Dichgans et al., 1973b; Azzena et al., 1974; Henn et al., 1974). An anatomical pathway has been identified, from the accessory optic tract and via the inferior olive and the cerebellar flocculus (Maekawa and Simpson, 1973) or more directly via the flocculus only (Brauth and Karten, 1977; Winfield et al., 1978), which may account for this influence. Our suggestionis that the gain of this pathway increasesunder pathological conditions, where vision becomes predominant and is able to fully substitute for the labyrinthine input. In fact, totally delabyrinthed animals may have normal righting reflexes when allowed to use their vision (Magnus, 1926). This would mean either that fibers carrying visual input to the brain stem form new synapseswith vestibular neurons responsiblefor postural control, or alternatively that already existing synapses are derepressed by the destruction of labyrinthine afferents. The latter hypothesis, postulated by Merrill and Wall (1972) for the somestheticsystemwould accountfor the rapid recovery when visual input is available to the animal.
SUMMARY The evolution of lateral head-tilt following hemilabyrinthectomyhas been studied in adult cats. Animalswere maintained postoperativelyin normally lit conditions(Lgroup) or in total darkness (D group). In cats from the L group, the head-tilt peaked at 45" (with the lesioned side down) on the second postoperative dAy, and decreased to about 0"within about 10 days. This evolution was followed by rebounds of head-tilt to larger angles before a stable
791
compensated head position could be maintained (approximately at the end of the third postoperative month). In cats from the D group, the head remained tilted by a large angle throughout the duration of the dark period up to 6 months. Re-exposure to light was followed by a rapid decrease of head-tilt. Finally, when put back in the dark at a late postoperative stage (up to 1year), already compensated animals were found to lose their symmetrical head position, and to re-acquire a strong head-tilt. This effect resumed on re-exposure to light. It is inferred that static visual input is a necessary condition for compensation of the postural deficits of hemilabyrinthectomy in the cat. Maintenance of a stable head posture also depends upon continuous availability of visual input. REFERENCES Andre-Thomas, Sorrel, E. and Sorrel-Dejerine, M. (1941) Fracture du rocher, troubles vestibulaires,attitude de la t&te,reactions meningees. Rev. Neurol., 6 k 7 3 . Azzena, G . B., Azzena, M. T. and Marini, R. (1974) Optokinetic nystagmus and the vestibular nuclei. Exp. Neurol., 42: 158-168. Bisti, S., Maffei, L. and Piccolino, M. (1972) Variations of the visual responses of the superior colliculus in relation to body roll. Science, 175: 456-457. Brauth,S.E.andKarten,M. J. (1977)Directaccessoryopticprojectionstothevestibulo-cerebellum: apossible channel for oculomotor control systems. Exp. Brain Res., 28: 73-84. Dichgans, J., Bizzi, E., Morasso, P. and Tagliasco, V. (1973a) Mechanisms underlying recovery of eye-head coordination following bilateral labyrinthectomy in monkeys. Exp. Brain Res., 18: 548-562. Dichgans, J., Schmidt, C. L. and Graf, W. (1973b) Visual input improves the speedometer function of the vestibular nuclei in the goldfish. Exp. Brain Res., 18: 319-322. Dieringer, N. and Precht, W. (1977) Modification of synaptic input following unilateral labyrinthectomy. Nature (Lond.), 269: 431-433. Dow, R. S. (1938) The effects of bilateral and unilateral labyrinthectomyin monkey, baboon and chimpanzee. Amer. J. Physiol., 121: 392-399. Ewald, J. R. (1892) Physwlogische Untersuchungen iiber das Endorgan des N. Oktavus. Wiesbaden, Bergmann. Flourens, J.P.N. (1842) Recherches ExpCrimentales sur les PropriktCs et les Fonctions du SystPme Nervew, duns les Animaw VertkbrCs. Baillike, Paris. Hem, V., Young, L. R. and Finley, C. (1974) Vestibular nucleus units in alert monkeys are also influenced by moving visual fields. Brain Res., 71: 144-149. Krejcova, H., Highstein, S. and Cohen, B. (1971) Labyrinthine and extralabyrinthine effects on ocular counter-rolling. Acta oto-laryngol. (Stockh.), 72: 165-171. Llinhs, R., Walton, R., Hillman, D. E. and Sotelo, C. (1975) Inferior olive: its role in motor learning. Science, 190: 1230-1231.
Maekawa, K. and Simpson,J. L. (1973) Climbingfiber response evokedin vestibulocerebellumof rabbit from visual system. J. Neurophysiol., 36: 649-666. Magnus, R. (1924) Korperstellung. J. Springer, Berlin. Magnus, R. (1926) Some results of studies in the physiology of posture. Lancet, 211: 531-536 and 585-588. McCabe, B. F., Ryu, J. H. and Sekitani, T. (1972) Further experiments on vestibular compensation. Laryngoscope, 82: 381. Merill, E. G. and Wall, P. D. (1972) Factors forming the edge of a receptive field: the presence of relatively ineffective afferent terminals. J. Physiol. (Lond.), 226: 825. Precht, W. (1974) Characteristics of vestibular neurons after acute and chronic labyrinthine destruction. In Handbook ofSensory Physiology, Vol. VIll, VestibularSystem, Part2, Psychophysics,AppliedAspects and General Interpretations. H. H. Kornhuber (Ed.), Springer-Verlag, Berlin, pp. 45 1 6 6 2 . Precht, W., Shimazu, H. and Markham, C. H. (1966) A mechanism of central compensation of vestibular function following hemilabyrinthectomyJ. Neurophyswl., 29: 996-1010. Putkonen, P. T. S., Courjon, J. H. and Jeannerod, M. (1977) Compensationforposturaleffectsof hemilabyrinthectomy in the cat. A sensory substitution process? Exp. Brain Res., 28: 249-257.
792 Schiifer, K. P. and Meyer, D. L. (1974) Compensation of vestibular lesions. In Handbook of Sensory Physiology, Vol. VIII. Vestibular System, Part 2, Psychophysics, Applied Aspects and General Interpretations, H. H. Kornhuber (Ed.), Springer-Verlag, Berlin, pp. 463-490. Talbott, R. E. (1974) Modification of the postural response of the normal dog by blindfolding. 1. Physiol., (Lond.), 243: 309-320. Wilson, V. J. and Yoshida, M. (1969) Bilateral connectionsbetween labyrinths and neck motoneurons. Brain Res., 13: 603-607. Wilson,V. J. andPeterson,B. W. (1978) Peripheralandcentralsubstratesofvestibulospinalreflexes.Physio1. Rev., 58: 80-105. Winfield,J. A., Hendrickson, A. and Kimm, J. (1978) Anatomicalevidencethat the medial terminal nucleus of the accessoryoptic tract in mammalsprovides avisual mossy fiber input to the flocculus.Brain Res., 15 1: 175-182.