THE JOURNAL OF COMPARATIVE NEUROLOGY 313~604-612 (1991)

Horizontal Optokinetic Nystagmus in Unilaterally Enucleated Pigmented Rats: Role of the Pretectal Commissural Fibers ANNIE REBER, JEAN MARIE SARRAU, JANICK CARNET, MICHEL MAGNIN, AND THERESE STELZ Universitk de Rouen, Facult6 des Sciences, 76134 Mont-Saint-Aignan Cedex (A.R.J.M.S., J.C., T.S.), and Vision et Motricite, Unite 94 INSERM, 69500 Bron (M.M.) France

ABSTRACT Monocular enucleation reduces the asymmetry of horizontal optokinetic nystagmus (H-OKN) in afoveate mammals by increasing responses to naso-temporal visual stimulation. The origin of these larger responses was investigated in adult pigmented rats monocularly enucleated as neonates or as adults by analyzing retinal and commissural projections to the deafferented nucleus of the optic tract (NOT) and the functional role of this nucleus before and after section of the posterior commissure. Anatomically, monocular enucleation reduces the volume of the contralateral deafferented NOT. Anterograde tracers injected in the intact eye reveal a crossed projection of the retina to the NOT and to the dorsal (DTN) and medial (MTN) terminal nuclei of the accessory optic system as in normal rats. In addition, there is an uncrossed projection to the MTN in the rats enucleated as neonates. Retrograde tracer injected in the deafferented NOT confirms the absence of an uncrossed retinal projection but reveals connections between both NOT via the posterior commissure as in normal rats. Electrophysiologically, the larger naso-temporal optokinetic responses in monocularly enucleated rats return to normal after posterior commissurotomy. This study demonstrates that no anatomical remodelling takes place to increase nasotemporal responses in monocularly enucleated rats. The larger responses must then result from functional changes. The role of exclusive contralateral projections of the retina to the NOT and of the commissural connections in mediating the asymmetry of the optokinetic nystagmus in afoveate mammals is discussed. Key words: optokinetic reflex, pretectum, posterior commissure, enucleation

The optokinetic system of afoveate mammals generates an asymmetrical horizontal optokinetic nystagmus (HOKN) with preference for a temporo-nasal direction of visual stimulation as revealed by tests in monocular viewing conditions (Ter Braack, '36; Tauber and Atkin, '68; Hess et al., '85). Crossed retinal projections convey the most efficient visual input for H-OKN as shown by lesion of the optic nerve or the optic chiasm (Collewijn, '75a; Cowey and Franzini, '79; Russel et al., '87). Retino-recipient structures in the meso-diencephalic area, i.e., the dorsal terminal nucleus (DTN) and the nucleus of the optic tract (NOT), are crucial relays for horizontal optokinetic responses (Simpson, '84; Simpson, '88). Other relays, such as the medial terminal nucleus (MTN), mainly involved in vertical optokinetic responses, also seem to contribute to the horizontal optokinetic reflex (Clement and Magnin, '84). The prime role of the NOT has been demonstrated by different experimental approaches. First, destruction of the NOT abolishes the H-OKN in the rat (Cazin et al., '80a). O

1991 WILEY-LISS, 1°C.

Second, stimulation experiments demonstrate that, among the retino-recipient relays deafferented from crossed visual inputs by degeneration of optic fibers, only the NOT responds to electrical stimulation by inducing an H-OKN (Collewijn, '75b). Finally, recording experiments in the NOT show that a large majority of NOT cells are activated by temporo-nasal visual stimulation. Fewer cells respond to naso-temporal stimulation (Collewijn, '75b; Cazin et al., '80b). Consequently the larger responses of the optokinetic nystagmus to temporo-nasal stimulation are readily understood by the output of NOT, but the smaller responses to naso-temporal direction remain less well explained. Interestingly, these smaller naso-temporal responses become larger, and consequently the asymmetry of the H-OKN Accepted August 12,1991. Address reprint requests to Dr. A. Reber, Universite de Rouen, Faculte des Sciences, Laboratoire de neurophysiologie Sensorielle, 76134-Mont-SaintAignan cedex, France.

605

H-OKN IN UNILATERALLY ENUCLEATED RATS decreases, after monocular enucleation (Reber et al., '89) or even after closure of the lids of one eye Cyucell et al., '90). Even though they are similar in results, such increases could originate from rather different processes in these two different experimental approaches. Most ipsilateral responses of NOT cells are lost during development in normal rats (Reber-Pel& '84). A larger proportion persists after neonatal enucleation (Reber et al., '89) and could be responsible for an increase in naso-temporal responsiveness. After enucleation in the adult or after closure of one eye, an increase in naso-temporal responses could be interpreted as a larger contribution of the posterior commissure that connects the NOT bilaterally (Teresawa et al., '79). It could also originate entirely from neurons located in the normal NOT not deafferented from its crossed retinal inputs. To test these hypotheses we have anatomically studied the retinal and commissural inputs of the NOT dederented from its crossed retinal inputs in adult rats enucleated as neonates or as adults. Correlated with these findings, electrophysiological studies have examined the functional role of this nucleus in the generation of nasotemporal responses before and after section of the posterior commissure.

fluorescent tracer (0.2 p1 of a 3%solution in sterile water) was injected with a micropipette (5 pm diameter). After elevent days of survival time, the rats were deeply anesthetized and perfused transcardially with saline, fixative and buffer, successively. The fixed brain was cut into coronal frozen sections (50 pm thick) to examine commissural inputs. The retina was separated from the pigmented layer and placed on a glass plate to search for a possible ipsilateral retinal input. In both cases, the fluorescent tracer was detected under ultraviolet illumination (Axioscope20 Zeiss, filter 355-425 nm). To complete this study, the commissural projections between the two NOT were studied in normal rats (group 1)using autoradiography. The NOT in deeply anesthetized animals was injected with 0.1 p1 of amino acid solution (300 pCi/ p1 of 3H-Leucine-3H-Proline) over a period of 20 min. After 24-48 hrs of survival time, the rats were deeply anesthetized, perfused transcardially, and their brains processed for autoradiography as described above. (3) Morphological study was made in the three groups of rats using the nomenclature of Palkovits and Brownstein ('88). The area of the NOT, as well as the size and the density of its cells were measured bilaterally on the sections prepared for the above autoradiographic study.

MATERIALS AND METHODS

Electrophysiological methods

Anatomical and electrophysiological analyses were performed in adult pigmented rats (DA/HAN). Normal rats served as controls (group 1).The other rats were unilaterally enucleated on the left side as neonates (group 2) or as adults (group 3) under deep anesthesia as previously described (Reber et al., '89) and were studied three months later.

To evaluate the functional strength of the dederented NOT in the generation of naso-temporal optokinetic responses, the H-OKN was measured before and after posterior commissurotomy in enucleated rats (groups 2 and 3), and compared to the H-OKN recorded in normal adult rats (group 1).Under deep anesthesia, the posterior commissure was sectioned on the mid-line with a slender scalpel inserted from the dorsal surface of the brain posterior to the corpus callosum. After recovery, a new recording session occurred. During both recording sessions, the optokinetic stimuli were delivered by rotation of a projected random dot pattern onto a drum cylinder around the immobilized rat. This visual surrounding motion was done in velocity steps ranging from 0.3 to 4Oo/s. The eye movements were recorded with the magnetic search coil technique using amplitude detection (Sarrau et al., '89). The eye coil was centered in the magnetic field but off-centered relative to the pupil to allow vision as previously described (Reber et al., '89). The eye position was obtained from coil voltage output converted by an analog-digital device. The eye movements (position and velocity) were analyzed by an appropriate Fortran computer program. This program calculated the slope of the eye movements and displayed both curves (the acquired one and the calculated one) using a GKS graphic library.

Anatomical methods (1)To reveal retinal inputs to the NOT, DTN and MTN, the right eye of anesthetized rats in each of the three groups received an intraocular injection of 1 p1 saline solution containing an anterograde tracer of either 30 pCi of amino acid (3H-Leucineand 3H-Proline)for autoradiographic study or a 10% solution of wheat germ agglutinin and horseradish peroxidase (WGA-HRP; Sigma type IV)for histochemical study. After a survival time of 18 to 24 hours, the animals were anesthetized, perfused transcardially with saline, then with fixative (2.5%glutaraldehyde and 1%paraformaldehyde in phosphate buffer) and finally with buffer. For autoradiography, the brains were embedded in Paraplast and cut into transverse sections 10 pm thick. Every fourth section was immersed for a few seconds in Ilford K5 emulsion, exposed for three months in black boxes at 4"C, and finally developed,fixed, and counterstained with cresylviolet. For histochemistry, the WGA-HRP was revealed on frozen transverse sections (50 pm thick) with the tetramethy1benzidine (TMB)technique (Mesulam, '78) and further counterstained with neutral red and mounted in Depex resin. In both studies, topographic analysis of the labeled area was done referring to the cytoarchitectural study of Scalia ('72). (2)The ipsilateral retinal input and commissural input to the deafferented NOT was studied in neonatally enucleated rats (group 2) and compared with inputs to the normal NOT of normal rats (group 1).For this purpose, the right NOT (deafferented in group 2 and normal in group 1)was localized using the stereotaxic coordinates adapted from the atlas of Konig and Klippel('63), and a retrograde fast blue

RESULTS Anatomical results In normal adult rats, each eye predominantly innervates the contralateral retino-recipient structures. Enucleation of one eye results in a degeneration of optic fibers which originate from this eye and project contralaterally. Consequently, the optic nerve becomes reduced in diameter and the optic chiasm becomes asymmetric in volume. Figure 1 illustrates this well-pronounced asymmetry in the animals enucleated three months earlier as neonates (group 2) or as adults (group 31, as well as the lack of asymmetry in the adults enucleated two weeks before as adults. Similarly,

606

Fig. 1. Optic chiasm on the ventral surface of a freshly-fixedbrain in rats enucleated on the left side or in normal rats. The brains are obtained from adult normal rats (a)and from adult rats enucleated on the left side as neonates (b)or as adults (c,d). The delays between enucleation and observation are 2-3 months in b and c, and 18 days in d. The asymmetry of the optic chiasm is characterized by a larger branch of the nerve crossing from the intact eye to the other side of the contralateral brain. It is strongly pronounced 2-3 months after the enucleation whatever the age of the animals at the time of this enucleation (neonatal: b or adult c), but not seen 18 days postoperatively (d). In this case the optic chiasm remains symmetrical as in normal rats (a). Scale bar = 1em.

three months after enucleation, the deafferented NOT appears reduced in volume compared to the non-deafferented NOT which remains similar to that of normal rats (Fig. 2). Quantitatively, the size of the deafferented NOT in adult rats enucleated as neonates (group 2) is particularly small along the dorso-ventral axis (108 pm) compared to the non-deafferented one (236 pm) or to the NOT of normal (group 1)rats (208 pm). Cell counts show a slight increase in cell density in this deafferented NOT compared to the contralateral non-deafferented NOT in group 2 (400 versus 360 cells/mm2,respectively).Measurements of the few cells with a large nucleus do not reveal differences in cell size (89 pm versus 82 pm for deafferented and non-dederented NOT, respectively). In rats enucleated as adults (group 31, the dimensions are 259 pm versus 112 pm for NOT size, and 88 pm versus 82 pm for cell size on the deafferented and non-dederented side, respectively. The cell density is also slightly increased in the deafferented NOT compared to the non-deafferented NOT (440 versus 340 cells/mm2, respectively). Consequently, quantitative examination reveals virtually no morphological differences in the NOT asymmetry between rats enucleated as neonates and rats enucleated as adults. The ipsilateral retinal inputs which are present at birth in some primary visual relays but normally withdraw during development, could have persisted if unilateral enucleation was performed before this period of retraction (Smith, '77; Land and Lund, '79). This possibility was investigatedby tracing the optic pathways followingintraoc-

A. REBER ET AL. ular injection of an anterograde tracer. Radioactive amino acid (3H-Leucine-3H-Proline)was used in eleven adult rats of the three groups (four normal, five enucleated as neonates and two as adults) and WGA-HRP in eight adult rats (three normal rats, two enucleated as neonates and three as adults). In the rats enucleated as neonates (group 21, the two tracers show that the terminations of the crossed retinal fibers overwhelmingly predominate over the uncrossed ones in the primary visual relay areas. Where the crossed fibers terminate in the pretectal area (Fig. 2), mainly three nuclei are strongly labeled: the pretectal olivary nucleus (PON), the lateral and medial nucleus of the optic tract (NOT) and the posterior pretectal nucleus (PP). In the accessory optic system, crossed retinal projections are strongly labeled in the dorsal terminal nucleus (DTN) and the medial terminal nucleus (MTN). The uncrossed projections to the pretectum are rare and only concentrated in the medial and posterior part of the PON. None are detected in the NOT. In the accessory optic system, weak uncrossed projections are visible in the MTN, as illustrated in Figure 3, but none are seen in the DTN. A comparison between the three groups of rats indicates a similar distribution and localization of the label in the pretectal nuclei (Fig. 2) and in the DTN of the accessory optic system. However, a noticeable difference appears in the MTN, which receives a bilateral retinal projection when adult rats have been enucleated as neonates (group 2) and only a contralateral one when animals are normal (group 1) or have been enucleated as adults (group 3). Consequently, this result shows a smaller asymmetry in retinal projections to this nucleus of the accessory optic system in neonatally lesioned rats. In an alternate approach, the fluorescent retrograde tracer fast blue was injected in the dederented NOT to see if ganglion cells in the normal ipsilateral eye are labeled and/or if this NOT still receives inputs via the posterior commissure from the non-dederented contralateral NOT. Eleven neonatally enucleated rats (group 2 ) were compared to four normal ones (group 1).The injected area involved the rostral NOT in two normal and four lesioned rats and the caudal NOT in two normal and four lesioned rats. In the remaining cases, the site of injection was misplaced (too superficial in two rats and too deep in one rat). Only two normal and one monocularly enucleated rat showed an injection site strictly localized in the NOT. In the two normal rats, labeled cells are located mainly in the naso-ventral quadrant of the contralateral retina. In all three rats (two normal and one enucleated) the retina ipsilateral to the injected NOT is devoid of labeled cells. In the other nine rats, the more superficial injection site encroaches the brachium of the superior colliculus and results in labeled ganglion cells located not only in the temporo-ventral quadrant of the contralateral retina in normal rats, but also in the homologous part of the ipsilateral retina in all animals. This is in agreement with the results obtained by Cowey and Perry ('791, showing that this part of the retina projects to the ipsilateral superior colliculus. After fast blue injections in the dederented NOT, labeled cells are also observed in the contralateral NOT (Figs. 4a,b). Two of the four normal and seven of the eight enucleated rats show a small population of fluorescent cells located slightly posterior to the posterior commissure in which some labeled commissural fibers also are seen.

607

H-OKN IN UNILATERALLY ENUCLEATED RATS

Fig. 2. Dark-field autoradiomicrophotographs of coronal sections through the pretectum showing the asymmetry of the projections from the eye injected with radioactive amino acid. The sections are obtained from normal rats (a)and from adult rats enucleated as neonates (b)or as adults (c). The label is strong in the nucleus of the optic tract (NOT) contralateral to the normal injected eye and absent in the ipsilateral NOT in normal as well as in enucleated rats. This observation demon-

strates the exclusive presence of crossed projections and the absence of uncrossed ones even after neonatal enucleation. Note the reduction in volume of the unlabeled NOT resulting from the disappearance of crossed retinal projections (b) and (c). Limits of pretectal area are marked with dotted lines. NOT: nucleus of the optic tract; PON: pretectal olivary nucleus; LGD: lateral geniculate nucleus, dorsal part. Scale bar = 1 mm.

Comparison of the two groups (1and 2) shows no difference in the localization and number of cells. Three normal adult rats received radioactive amino acid injections in one NOT. In two of these rats, the injection was perfectly placed in the rostral NOT with diffusion of the tracer into the PON and the dorsal pole of the anterior pretectal nucleus (Fig. 4c). In the third rat, the injection site was larger and includes the above mentioned pretectal nuclei plus the rostro-lateral border of the superior colliculus. In these three rats, particularly in the former, some of the pretectal efferent labeled fibers are clearly seen to cross the posterior commissure toward the contralateral pretectum (Fig. 4c). In this pretectal area, the NOT has the highest density of silver grain accumulations, typical of

labeled terminals (Fig. 4c), throughout the entire extent of this nucleus.

Electrophysiological results Even though a commissural connection between the bilateral NOT of enucleated rats appears to exist, the functional significance of the deafferented NOT can be questioned. Responses evoked by naso-temporal stimulation in monocularly enucleated rats could originate entirely from naso-temporal sensitive neurons in the non-deafferented NOT. To test the functional role of the commissural input to the dederented NOT, we compared H-OKN in both directions before and after commissurotomy in rats

608

A. REBER ET AL.

Fig. 3. Autoradiomicrophotographs of coronal sections showing the medial terminal nucleus (MTN) after intraocular injection of radioactive amino acids. The sections are obtained from normal rats (a),rats enucleated as neonates (b),and as adults (c). Arrows indicate the position of the MTN. In all the rats the MTN contralateral to the

injected eye is strongly labeled. In contrast, the ipsilaterd one does not appear labeled in normal and in rats enucleated as adults (a and c); a weak but well-distinguishable labeling is only seen in rats enucleated as neonates (b: general view at the top and details at the bottom). Scale bar = 1mm.

enucleated as neonates (5 rats of group 2) or as adults (5 rats of group 3). Figure 5 illustrates examples of H-OKN in these situations. Before mid-line section, H-OKN presents the characteristics previously described in unilaterally enucleated rats (Reber et al., '89). In brief, H-OKN is elicited for a large range of velocity steps (O.4-4O0/s.) of horizontal optokinetic stimulation directed temporo-nasally or nasotemporally. The closed-loop gain (eye velocitylstimulus velocity) of slow phases directed naso-temporally is augmented compared to that of normal adult rats with monocular vision (one eye patched), as illustrated in Figure 6. On the other hand, the gain of slow phases directed temporonasally is unchanged in rats enucleated as neonates and is slightly reduced in rats enucleated as adults. However, gain asymmetry (naso-temporal versus temporo-nasal) is reduced following monocular enucleation in both groups. One day after the mid-line section, naso-temporal visual stimulation generates a weak or no response as shown in Figure 5 . When it exists, only an irregular alternation of slow and quick phases is observed. The slow phases are often curvilinear with velocity running from a high value (lOOo/s.) to a low one. Figure 6 illustrates the very low closed-loop gain obtained for all the stimulus velocities tested. This gain returns to the normal values observed in normal rats under monocular vision. In contrast, H-OKN elicited by temporo-nasal visual stimulation is strong and almost regular. There are fluctuations only in the mean eye position. The closed-loop gain is not well-modified compared with that observed before the mid-line section (Fig. 6). Consequently, the strong asymmetry of H-OKN with preference for temporo-nasal direction is re-established

following mid-line section as in normal rats in monocular viewing conditions.

DISCUSSION Monocular enucleation or closure of one eye is known to produce transsynaptic degeneration in the visual pathways that is characterized by a reduction in volume of the contralateral superior colliculus (for review see Smith, '77) and the lateral geniculate nucleus (LGN) (rat: Fifkova, '701, shrinkage of LGN cells deafferented from retinal inputs (cat: Wiesel and Hubel, '63; rat: Robertson et al., '89), and a significant reduction in the connectivity of neurons in the contralateral visual cortex (rat: Rothblat and Schwartz, '79). This study demonstrates that a similar phenomenon also occurs in the optokinetic system, in which there is a significant reduction in the volume of the NOT deafTerented from crossed retinal inputs and a slight increase in cell density. However, there is no change in cell size. In these conditions, the shrinkage of the NOT probably results mainly from degeneration of the retinal terminals. Although degeneration of the optic fibers appearsmicroscopically complete 24 hours after the deafferentation (rat: Sefton and Lam, '84), our study shows that the subsequent reduction in size only becomes apparent several weeks later both in the optic chiasm and in the NOT. Taken together with degeneration observed in other visual relay areas after monocular deprivation (Rothblat and Schwartz, '79), these results show that the loss of visual input produces degenerative changes in the majority of retino-recipient structures in the rat. Our findings further demonstrate that such

H-OKN IN UNILATERALLY ENUCLEATED RATS

609

Fig. 4. Photomicrographs of coronal sections through the pretectum showing commissural relations between both nucleus of the optic tract (NOT) after injection of fast blue (a and b) or radioactive amino acids (c) in one NOT. a: Site of fast blue injection (#) in the deafferented NOT (left) of adult rats enucleated as neonates and fluorescent cells in the contralateral NOT (right). b: Details of those fluorescent NOT cells. Their labeling demonstrates relations between the two NOT. NOT: nucleus of the optic tract. Scale bar = 1mm in a and 10 pm in b. c: Left: light field microphotograph showing the injection site centered on the

NOT (arrow). Right: dark-field microphotograph of the pretectum

changes occur irrespective of the age of the rats when the lesion was performed, and thus are independent of the state of the functional development of the optokinetic system. The asymmetric degeneration in the NOT, seen only on the side contralateral to the enucleated eye, is consistent

with the predominantly contralateral retinal projection to the NOT. Ipsilateral retinal projections are very small (5-10%) in normal adult rats (Yamadori, '77). They innervate principally the lateral geniculate nucleus and the superior colliculus (Toga and Collins, '81; Linden and

contralateral to the injection showing the pretectal commissural labeled fibers crossing in the posterior commissure, some of them being directed toward the contralateral NOT where they terminate (area located between the two white arrows). LGd: lateral geniculate nucleus, dorsal part; LGv: lateral geniculate nucleus, ventral part; LP: lateral posterior nucleus; NOT: nucleus of the optic tract; PAN: pretectal anterior nucleus; PON: pretectal olivary nucleus. Scale bar = 500 pm.

610

A. REBER ET AL.

57s T N

57s N T

a

b 1

2

c 1

2

Fig. 5. Example of horizontal optokinetic nystagmus (H-OKN) recorded in normal rats and in monocularly enucleated rats before (1) and after (2) posterior commissurotomy. The visual stimulus has a constant velocity of 5"is and a temporo-nasal (TN) or naso-temporal (NT) direction with respect to the normal eye bearing the coil. The onset and the end of the stimulation are indicated by downward and upward arrows, respectively. The trace represents the eye position during the stimulation. In normal rats under monocular vision (a)the

H-OKN appears strongly asymmetrical with a preference for temporonasal direction of visual stimulation. Before posterior commissurotomy, the asymmetry of the responses to naso-temporal direction in the adult rats monocularly enucleated as neonates (b,l)or as adults (c,l) is much less pronounced. After the posterior commissurotomy a strong asymmetry similar to that observed in normal rats reappears (b,2and

Perry, '83). They are not seen in the NOT (rat: Scalia, '72) or in the DTN or MTN (Yamadori and Yamauchi, '83) when using degeneration or HRP histochemical (Terubayashi and Fujisawa, '84) techniques. The present neuroanatomical study demonstrates identical results, since ipsilateral retinal projections to the NOT, DTN and MTN of normal animals are not seen using anterograde neuroanatomical tracing techniques. The use of a retrograde tracer (fast blue) supports the above results, confirming the absence of an ipsilateral projection to the NOT. From all of these observations, it appears that these projections to the NOT, DTN, and MTN either do not exist or are so small that they cannot be revealed by the techniques used so far.

Neonatal unilateral enucleation is known to prevent the natural elimination of ipsilateral optic axons during development (rat: Land and Lund, '79; Sefton and Lam, '84; Chan and Jen, '88; Shirokawa et al., '83; hamster: Campbell and Lieberman, '85) and to provoke an enlargement of the ipsilateral projection area in the lateral geniculate nucleus (Reese, '86) and superior colliculus (Land and Lund, '79; Lam et al., '82; Takeuchi et al., '82). Consistent with these observations, our findings reveal ipsilateral projections to the MTN in the neonatally enucleated rats. These projections, which are present at birth in normal rats (Bunt et al., '83)'apparently survive after neonatal enucleation instead of withdrawing during development.

c,2).

H-OKN IN UNILATERALLY ENUCLEATED RATS A Enucleated as neonates

C I ' 1 0 S

e d L 0 0

P G

a i

n 0 Stimulus Velocity (d./s)

B Enucleatedas adults

a i

n Stimulus Velocity (d./s)

I

NT TN

Fig. 6. Closed-loop gain of H-OKN as a function of visual stimulus velocity in the rats enucleated as neonates or as adults. The mean gain (eye velocity/stimulus velocity) is obtained from 5 rats enucleated as neonates (A) and from 5 rats enucleated as adults (B). For temporonasal direction of visual stimulations the gain is large for low stimulus velocities and decreases for higher velocities before (dashed linesiblack dots) as well as after the commissurotomy (dashed-dotted linesiwhite circles). It is close to the high gain obtained for temporo-nasal visual stimulation in normal rats under monocular vision (indicated by the upper thin dotted line). For naso-temporal direction the gain is high before the commissurotomy (continuous line/black triangles), and low after the commissurotomy (continuous line/white triangles) as in normal rats under monocular vision (thin continuous line).

In contrast, the absence of ipsilateral retinal projections to the NOT and DTN following contralateral neonatal enucleation suggests that any anatomical remodeling, if present, cannot be attributed to a significant persistence of a projection that normally disappears during development, or the exuberance of an abnormal projection to another area. The absence of changes in the projections to the NOT

611 is not specific to the rat, in which ipsilateral retinal projections to the NOT are not seen in normal adults. It has also been observed in species with normal ipsilateral projections to the NOT in adults (rabbit: Collewijn and Holstege, '84; cat: Olavarria et al., '84). These results demonstrate that a definitive adult projection pattern to the NOT is acquired independently of binocular vision. Furthermore, remodelling of neuroanatomical projections might be restricted by the number and/or location of retinal neurons in the intact eye available to establish abnormal connections with target structures. When enucleation is performed in adult rats, uncrossed retinal projections are not detected to either the NOT or the DTN and MTN. This result is in accordance with the weak change described in other retino-recipient areas after enucleation of adults (Rothblat and Schwartz, '79). This can be easily understood for the DTN and the NOT which remain unchanged after neonatal enucleation. The failure of ipsilatera1 projections to the MTN of rats enucleated as adults, however, indicates that the distribution of projection patterns cannot be modified substantially, if at all, by monocular enucleation in the adult animals. Moreover, taken together with the presence of projections to this nucleus following neonatal enucleation, these results suggest that anatomical remodelling of these paths is only possible during the early period of development. The functional involvement of these retinal projections to the ipsilateral MTN in horizontal optokinetic responses is not indicated by our results. The H-OKN of rats with ipsilateral projections to the MTN (group 2) does not differ markedly from that in rats of group 3, in which this projection is absent. Electrophysiological analyses have shown that nasotemporal responses are mediated by the ipsilateral NOT in foveate animals. They are as strong as temporo-nasal responses. Consequently, the gain is symmetric in monocular vision. This symmetry results from visual cortical projections to the NOT (Montarolo et al., '81). In afoveate animals the naso-temporal responses are small but become larger after unilateral enucleation (Reber et al., '89). These stronger responses in enucleated rats do not result from cortical projections but from commissural inputs, since, as shown in this study, they are abolished by commissurotomy. Such a commissural transfer of optokinetic information between the two homologous pretectal areas is further supported by the findings from a deoxyglucose study (Bird et al., '82) and by our anatomical data revealing in both normal and enucleated adult rats the existence of a commiss u r d projection between the two NOT. The mechanisms underlying the increase of nasotemporal responses after commissurotomy probably originate from a disinhibition of naso-temporal NOT responses, since they are obtained either by enucleation on one side or by picrotoxin injection in the NOT on the same side (pigeon: Gioanni et al., '83; frog: Yucell, '90). It is possible that the deafferented NOT is no longer inhibited by NOT neurons on the non-deafferented side, thus leading to an increased excitation of these neurons during naso-temporal stimulation. The increased responses of neurons elicited by nasotemporal stimulation in the bilateral NOT after enucleation (Reber et al., '89) suggests a release from a reciprocal inhibition between the two NOT. This release is not obtained by patching one eye, and therefore probably results from a modification in the resting activity of NOT cells induced by loss of ganglion cells. In intact rats, this resting activity would inhibit naso-temporal responses (Reber et al., '89). The remaining weak naso-temporal responses persisting after commissurotomy are similar to those of normal rats

612

A. REBER ET AL.

and could result either from stimulation of the nondeafferented NOT or from the deafferented NOT by indirect afferent connections to the NOT arising from ipsilateral subcortical structures and cortical areas (for references see Simpson et al., ’88).

ACKNOWLEDGMENTS We are grateful to Professor N. Dieringer for reading our earlier version of the manuscript and for fruitful suggestions. We thank B. Delahaye for his valuable assistance and G . Lietout and M. Dechamps for their help with computer time. Microscopes were loaned by the Carl Zeiss Company and Professor J. Boisard.

LITERATURE CITED Biral, G.P., M. Cavazzuti, R. Ferrari and R. Corazza (1982) Optokinetic study. visual detection in the rat visual centres: A[14Cl-2-deoxy-D-glucose Arch. Int. Physiol. Biochem. 90:141-144. Bunt, S.M., R.D. Lund, and P.W. Land (1983) Prenatal development of the optic projection in albino and hooded rats. Dev. Brain Res. 6:149-168. Campbell, G., K.F. So, and A.R. Lieberman (1985) Identification of synapses formed by aberrant, uncrossed retinogeniculate projection in the hamster after neonatal monocular enucleation. Dev. Brain Res. 211137-140. Cazin, L., W. Precht, and J. Lannou (198Oa) Pathways mediatingoptokinetic responses of vestibular nucleus neurons i n the rat. Pflugers Arch. 384: 19-29. Cazin, L., W. Precht, and J. Lannou (1980b) Firing characteristics of neurons mediating optokinetic responses to rat’s vestibular neurons. Pfliigers Arch. 386:221-230. Chan, S.O., and L.S. Jen (1988) Enlargement of uncrossed retinal projections in the albino rat: additive effects of neonatal eye removal and thalamectomy. Brain Res. 461:163-168. Clement, G., and M. Magnin (1984) Effects of accessory optic system lesions on vestibulo-ocular and optokinetic reflexes in the cat. Exp. Brain Res. 55:49-59. Collewijn, H. (1975a) Oculomotor areas in the rabbit’s midbrain and pretectum. J. Neurobiol. 6:3-22. Collewijn, H. (1975b) Direction-selective units in the rabbit’s nucleus of the optic tract. Brain Res. 100:489-508. Collewijn, H., and G. Holstege (1984) Effects of neonatal and late unilateral enucleation on optokinetic responses and optic nerve projections in the rabbit. Exp. Brain Res. 57t138-150. Cowey, A. and C. Franzini (1979) The retinal origin of uncrossed optic nerve fibres in rats and their role in visual discrimination. Exp. Brain Res. 35.443-455. Cowey, A. and V.H. Perry (1979) The projections of the temporo-nasal retina in the rat, studied by retrograde transport of horseradish peroxydase. Exp. Brain Res. 35457-464. Fifkova, E. (1970) The effect of unilateral deprivation on visual centers in the rat. J. Comp. Neurol. 140:431438. Gioanni, H., J. Rey, J. Villalobos, D. Richard, and A. Dalbera (1983) Optokinetic nystagmus in the pigeon (Colunba Ziuia). I1 Role of the pretectal nucleus of the accessory optic system (AOS). Exp. Brain Res. 50:237-247. Hess, B.J.M., W. Precht, A. Reber, and L. Cazin (1985) Horizontal optokinetic ocular nystagmus in the pigmented rat. Neuroscience 1597-107. Konig, J.F., and R.A. Klippel(1963) The rat brain: a stereotaxic atlas of the forebrain and lower parts of the brainstem. Baltimore: Williams and Wilkins, pp. 1-66. Lam, K., A.J. Sefton, and M.R. Bennett (1982) Loss of axons from the optic nerve of the rat during early postnatal development. Dev. Brain Res. 3:487-49 1. Land, P.W., and R.D. Lund (1979) Development of the rat’s uncrossed retinotectal pathway and its relation to plasticity studies. Science 205698-700. Linden, R., and V.H. Perry (1983) Massive retinotectal projection in rats. Brain Res. 272:145-149. Mesulam, M.M. (1978) Tetramethyl benzidine for horseradish peroxidase neurohistochemistry: a non-carcinogenic blue reaction-product with superior sensitivity for visualizing neural afferents and efferents. J. Histochem. Cytochem. 2:106-117.

Montarolo, P.G., W. Precht, and P. Strata (1981) Functional organization subserving the optokinetic nystagmus in the cat. Neuroscience 6:231246. Olavarria, J., R. Malach, and R.C. Van Sluyters (1984) Retino-pretectal projections in monocularly deprived cats. Dev. Brain Res. 12:141-145. Palkovits, M., and M.J. Brownstein (1988) Maps and guide to microdissection of the rat brain. New York Elsevier. Reber-Pelle, A. (1984) Postnatal development of pretectal and NRTP neuron responses to optokinetic stimulation in the rat. Develop. Brain Res. 12:lll-120. Reber, A., J. Lannou, and B.J.M. Hess (1989) Development of optokinetic neuronal responses in the pretectum and horizontal optokinetic nystagmus inunilaterally enucleated rats. Arch. It. Biol. 127t225-242. Reese, B.E. (1986) The topography of expanded uncrossed retinal projections following neonatal enucleation of one eye: differing effects in dorsal lateral geniculate nucleus and superior colliculus. J. Comp. Neurol. 250s-32. Robertson, R.T., H.K. Poon, M.R. Duran, and J. Yu (1989) Neonatal enucleations reduce number, size, and acetylcholinesterase histochemical staining of neurons in the dorsal lateral geniculate nucleus of developing rats. Dev. Brain Res. 47:209-225. Rothblat, L.A., and M.L. Schwartz (1979) The effect of monocular deprivation on dendritic spines in visual cortex of young and adult albino rats: evidence for a sensitive period. Brain Res. 161:156161. Russell, IS., M.W. van Hof, J. van der Stein, and H. Collewijn (1987) Visual and oculomotor function in optic chiasma-sectioned rabbit. Exp. Brain Res. 66%-73. Sarrau, J.M., A. Reber, B. Mazari, and J. Carnet (1989) Capteurangulaire de mouvements oculaires. Paris: Conference internat. Capteurs, pp. 54-62. Scalia, F. (1972) The termination of retinal axons in the pretectal region of mammals. J. Comp. Neurol. 145223-257. Sefton, A.J., and K. Lam (1984) Quantitative and morphological studies on developing optic axons in normal and enucleated albino rats. Exp. Brain Res. 57:107-1 17. Shirokawa, T., Y. Fukuda, and T. Sugimoto (1983) Bilateral reorganization of the r a t optic tract following enucleation of one eye at birth. Exp. Brain Res. 51:172-178. Simpson, J.I. (1984) The accessory optic system. Ann. Rev. Neurosci. 7:13-41. Simpson, J.I., R.A. Giolli, and R.H.I. Blanks (1988) The pretectal nucleus complex and the accessory optic system. I n J.A. Buttner-Ennever (ed): Progress in Oculomotor Research. Neuroanatomy of the Oculomotor System. Amsterdam:Elsevier, pp 335-364. Smith, D.E. (1977) The effect of deafferentation on the development ofbrain and spinal nuclei. Progress in neurobiol. London: Pergamon Press 8:349-367. Takeuchi, H., Y. Fukuda, Y. Hara, and C. Hsiao (1982) Physiological properties of expanded ipsilateral retinocollicular projection in neonatally enucleated albino rats. Brain Res. 231:191-196. Tauber, E.S., and A. Atkin (1968) Optomotor responses to monocular stimulation: Relation to visual system organisation. Science 160:13651367. Teresawa, K., K. Otani, and J. Yamada (1979) Descending pathways of the nucleus ofthe optic tract in the rat. Brain Res. 173:405417. Ter Braak, J.W.G. (1936) Untersuchungen uber optokinetischen Nystagmus. Arch. Neerl. Physiol. 21:309-376. Terubayashi, H., and H. Fujisawa (1984) The accessory optic system of rodents: a whole-mount HRP study. J. Comp. Neurol. 227t285-295. Togga, A.W., and R.C. Collins (1981) Metabolic response of optic centers to visual stimuli in the albino r a t Anatomical and physiological considerations. J. Comp. Neurol. 199:443464. Wiesel, T.N., and D.H. Hubel (1963) Effect of visual deprivation on morphology and physiology of cells in the cat’s lateral geniculate body. J. Neurophysiol. 26:978-993. Yamadori, T. (1977) An experimental anatomical study on the optic nerve fibers in the rat by using a new selective silver impregnation technique: Termination of the main optic tract. Okajimas Folia Anat. Jap. 54:229246. Yamadori, T., and K. Yamauchi (1983) An experimental anatomical study of the optic nerve fibers in the rat: Courses of the accessory optic tract. Brain Res. 269:41-46. Yucell, H.Y., M.S. Kim, B. Jardon, and N. Bonaventure (1990) Abolition of monocular optocinetic nystagmus directional asymmetry after unilateral visual deprivation in adult vertebrates: Involvement of the gabaergic mechanism. Dev. Brain Res. 53:179-185.