Experimental Brain Research
Exp. Brain Res. 35,457--464 (1979)
@ Springer-Verlag 1979
The Projection of the Temporal Retina in Rats, Studied by Retrograde Transport of Horseradish Peroxidase A. Cowey and V.H. Perry Department of ExperimentalPsychology,South Parks Road, Oxford OX1 3UD, England
Summary. Horseradish peroxidase (HRP) was injected unilaterally into the lateral geniculate nucleus or tectum, or both, in 26 hooded rats in order to mark the exact extent of the retina from which uncrossed optic axons arise. This region occupied about a quarter of the retina, in the temporal periphery, following thalamic injections, but a much smaller region following tectal injections. By comparing the proportions of HRP positive neurones in nasal and temporal retinae of both eyes it was shown that: (1) within the region supplying uncrossed axons the majority of the ganglion cells nevertheless project contralaterally, (2) a large proportion of the ganglion cells from the temporal crescent project bilaterally, which does not occur from the remainder of the retina, (3) ganglion cells of all sizes contribute to both ipsilateral and contralateral projections. The results also support earlier suggestions that the smallest neurones in the ganglion cell layer do not send an axon into the brain, and are therefore not ganglion cells Key words: Rats - Retina - Ganglion cells
In the preceding paper (Cowey and Franzini, 1979) the region of retina from which uncrossed axons arise in hooded rats was estimated by plotting the position of undegenerated ganglion cells following sagittal section of the optic chiasma or unilateral optic tract section. This region in the temporal and ventral retina proved to be unexpectedly large in view of the estimate that fewer than 10% of optic nerve fibres do not decussate in rats (Polyak, 1957). However, Cowey and Franzini (1979) counted only large ganglion cells with a soma diameter greater than 14-15 ~,m and it is possible that far from all the ganglion cells in this part of the retina project ipsilaterally. We therefore decided to examine the problem by injecting horseradish peroxidase (HRP) unilaterally into the lateral geniculate nuclei (LGN) or superior colliculus (SC) or both.
Offprint requests to:
Dr. A. Cowey(address see above)
0014-4819/79/0035/0457/$ 1.60
458
A. Cowey and V. H. Perry
Fig. 1. Schematicrepresentation of 6 possible arrangements of optic axons. All axons from the larger nasal retina project contralaterally. In the smaller temporal part of the retina branched arrows represent bifurcating axons
The six possible arrangements of ganglion cell projections from the temporal retina are shown in Fig. 1. In each schematic retina the crescent shows the temporal region, in which some ganglion cells survive section of the optic chiasma or contralateral optic tract. The arrows indicate the direction in which ganglion cells project. Dividing arrows indicate bifurcating axons innervating both hemispheres as suggested by Cunningham and Freeman (1977). In the nasal region all neurones larger than 9 ~tm in diameter degenerate following contralateral tract section and most of the remainder are almost certainly not shrunken ganglion cells (Eayrs, 1952; Mantz and Klein, 1951; Bunt et al., 1974; Perry, 1977). If virtually all neurones larger than about 9 gin, i.e., all the ganglion cells, are labelled in the nasal retina of the contralateral eye following a cerebral injection of H R P then the proportion labelled in the temporal region of the ipsilateral eye provides a means of distinguishing between some of the arrangements shown in Fig. 1. For example, if far fewer ganglion cells are labelled in the ipsilateral temporal crescent schemes A, B, and C must be incorrect because in each schema every ganglion cell in the temporal crescent projects solely or via a bifurcation to the ipsilateral hemisphere. Likewise, if the temporal retina of the eye contralateral to the injection contains any labelled ganglion cells scheme A must be incorrect, and if the temporal retinae of the two eyes are labelled to a conspicuously different extent following a unilateral injection scheme B must be wrong. If schemes A - C prove to be untenable it should be possible to discriminate among D - F by comparing the proportions of labelled cells in the temporal crescent of the ipsilateral and contralateral eyes. For example scheme E demands that all ganglion cells in the contralateral temporal region should be labelled but only a proportion of those in the ipsilateral temporal retina. The remaining aim of the experiment was to measure the extent of the retina from which ipsilateral-projecting axons arise and to compare it with that found by Cowey and Franzini (1979) following section of the optic chiasma, which is likely to damage the ipsilateral fibres and may, therefore, yield too low an estimate.
Projections from Temporal Retina
459
Methods Subjects T w e n t y - s i x adult male hooded Lister rats were used. The rats were anaesthetised with an intra-peritoneal injection of chlornembutal (2.5 m g / k g of 2.1 g chloral hydrate + 0.5 g sodium pentobarbitaI in 50 ml) and placed in a stereotaxic headholder. An incision was m a d e in the scalp and a hole drilled in the skull with a dental burr. The hole was enlarged with fine rongeurs, and the dura cut. T h e rats were injected in three groups. The first group of 6 rats received three or four injections of 0.75-1.0 ~tl of a 5 0 % H R P solution (Boehringer, Mannheim), and the second group of 17 rats received 3 injections of 0.75 ~tl of a 3 0 % solution of H R P dissolved in 2 % dimethoxysulphoxide (Lin et al., 1977). The injections of all 23 animals were evenly spaced along the length of the lateral geniculate nucleus and the stereotaxic co-ordinates were chosen with the aid of the atlas of K6nig and Klippel (1963). The injections were delivered hydraulically from a 1 ~tl Hamilton syringe, and each lasted for a period of several minutes. A further group of three animals received single injections of 3 0 % H R P in 2 % dimethoxysulphoxide. The injections were aimed at the anterior half of the superior colliculus in an attempt to label only the ipsilateral retinotectal pathway.
Histology T h e eyes and brains from all the animals were treated in the same manner. The rat was killed with chloroform and a stitch placed in the most dorsal margin of each eye. The eyes were then removed and placed in 0.9% chilled saline at 4 ~ C. The rat was t h e n perfused through the heart with chilled physiological saline followed by 2 % paraformaldehyde + 1% glutaraldehyde in 0.1 M phosphate buffer. The brain was r e m o v e d and stored in the fixative for approximately six hours at 4 ~ C before being washed in 0.1 M phosphate buffer + 3 0 % sucrose overnight. The eyes were opened at the level of the ora serrata and a major cut m a d e in the retina from the optic disc to the stitch to serve as a landmark. The retina was then dissected out and fixed for 3 0 - 6 0 rain in 2 % glutaraldehyde in 0.1 M phosphate b uffer either free floating or flat on a slide with a layer of porous paper on top. The latter technique gave more satisfactory results. Following fixation the retinae were washed in a 0.1 M phosphate buffer overnight. T h e brains were sectioned at 50 ~t on a freezing microtome and a 1-in-5 series saved. Both brain sections and the retinae were then reacted with either o-dianisidine as a substrate for the H R P following the method of Coleman et al. (1976), or with diaminobenzidene (DAB). The retinae were then w e t - m o u n t e d on gelatinised slides and outline drawings made of the retinae before any shrinkage had occurred, using a microscope with an X - Y plotter attached. The part of the ipsilateral retina containing labelled ganglion cells was then similarly plotted. Some of the retinae were also photographed at this point. The retinae were then fixed to gelatinised slides following the method of Stone (1965) and lightly counterstained with cresyl violet. We noted that sections reacted with o-dianisidine showed some fading of the reaction product when counterstained, which did not occur after reaction with D A B . The brain sections were m o u n t e d on gelatinised slides, dehydrated and coverslipped in the usual manner.
Cell Counts In order to discriminate a m o n g schemes A - F we counted labelled neurones as follows. In each of three rats with L G N injections the temporal retina of both eyes and the nasal retina of the eye contraiateral to the injection were examined at a magnification of x 1000. The site chosen was always approximately 500 m m from the edge of the retina and h a l f w a y below the optic disc. Using a square eyepiece grid covering 90 ,am2 and subdivided into 100 smaller squares every neurone within the grid was drawn by m e a n s of a drawing tube attached to the microscope. The H R P labelled neurones were then m a r k e d on the drawing. The grid was then moved to an adjacent region until every neurone in an area 180 x 360 ~tm had been drawn. Each experimenter examined a slightly different region and the results were summed. Glial cells were not drawn and were distinguished from neurones by their small size (usually less than 6 gin), dense staining, and absence of cytoplasm and Nissl substance.
460
A. Cowey and V. H. Perry
Fig. 2. Photographs of ipsilateral and contralateral unstained flat-mounted retinae reacted with o-dianisidine after injection of HRP into right LGN
Results
1. L G N Injections In all 23 rats in w h i c h t h e i n j e c t i o n s w e r e a i m e d at the L G N , b o t h d o r s a l and v e n t r a l divisions o f the nucleus w e r e d e n s e l y s t a i n e d with r e a c t i o n p r o d u c t , which e x t e n d e d for s e v e r a l m m u n i f o r m l y in all d i r e c t i o n s b e y o n d the p e r i p h e r y
Projections from Temporal Retina
461
Table 1. Total number of neurones counted in equivalent regions of far temporal and nasal retinae and the proportion labelled with HRP Ipsilateral Temporal
Contralaterai Temporal
Contralateral Nasal
Rat 24
561 14%
522 39%
506 46%
Rat 25
638 13%
554 32%
557 42%
Rat 26
587 13%
602 42%
534 39%
of these nuclei. Although there is evidence that H R P is taken up chiefly or entirely in the close vicinity of the needle tip (Vanegas et al., 1978) we cannot be certain that there was no uptake in the tectum and pretectum. Moreover, the optic tract adjacent to L G N was densely labelled, and damaged fibres e n r o u t e to the mid-brain may well have accumulated H R P in the tract. For these reasons it is safest to regard the injections as involving both L G N and mid-brain optic pathways. However, the tracer did not cross the mid-line, or only slightly so in some animals, so the results are not contaminated by involvement of the other hemisphere. Figure 2 is a photograph of ipsilateral and contralateral retinae after reacting them with o-dianisidine but before counterstaining. The region of the ipsilateral retina containing labelled ganglion cells is clear, and was similar in the remaining animals of the group. The m e a n extent of the ipsilaterally projecting crescent was 2 4 . 6 % (range 17.6-33.3) of the entire retina, which is very close to the figure of 21.2% calculated by Cowey and Franzini (1979). Apart from an occasional solitary labelled cell outside the temporal crescent but still in the temporal half of the retina the remainder of the ipsilateral retina contained no labelled neurones. Figure 2 also shows that the entire contralaterat retina is labelled; there is no unlabelled temporal crescent as would be expected if the region projected only ipsilaterally. Scheme A of Fig. 1 is, therefore, incorrect and it is clear that the temporal crescent sends axons to both hemispheres. Figure 2 also shows that the retina of the contralateral eye is m o r e densely labelled than the temporal crescent of the ipsilateral eye, indicating that schemes B and C are also incorrect. The results of counting neurones and estimating the proportion of labelled cells are shown in Table 1. In the nasal retina of the eye contralateral to the injection nearly half the neurones were labelled. As there are about twice as many neurones in the ganglion cell layer as there are fibres in the optic nerve of the rat (Hughes, 1977) we conclude that almost every ganglion cell was labelled and that the remainder are interneurones as proposed by Bunt et al. (1974) and Perry (1977). Furthermore, it was obvious from the outline drawings of the cell
462
A. Cowey and V.H. Perry
,,;;;~T: a
@
0
Fig. 3. Outline drawings of right retina following injection of HRP into right LGN (A) or right superior colliculus (C-D). Dashed line indicates extent of region containing HRP-positive neurones
bodies that the smallest neurones were commonly unlabelled whereas those larger than 9 ~tm were invariably labelled. The results in the temporal retina were different. In the eye ipsilateral to the injection only 14 % of the neurones, i.e., about 28 % of the ganglion cells, were labelled and the unlabelled group contained neurones of all sizes. As neurones from the ipsilateral nasal retina do not project ipsilaterally and the temporal retina lies outside the region of highest cell density it is clear that Polyak (1957) was correct in estimating that only 5 - 1 0 % of the optic nerve fibres project ipsilaterally despite the fact that they originate from an area involving 2 4 % of the retina. This result also shows that if scheme D is correct m a n y fewer ganglion cells in the temporal retina project ipsilaterally than contralaterally, and this is consistent with the finding that approximately 3 times as m a n y neurones are labelled in the contralateral temporal retina (Table 1). However, scheme D must be rejected if Cunningham and F r e e m a n (1977) are correct in concluding that the majority of ganglion cells in the temporal retina innervate both hemispheres by bifurcating axons. We are, therefore, left with schemes E and F. Scheme E, in which all neurones in the temporal crescent project contralaterally but only a proportion project ipsilaterally, requires that a similar proportion of ganglion cells should be labelled in the contralateral temporal and nasal retinae. This was observed in rat 26 but not in the other two rats (Table 1). We are, therefore, unable to reject or confirm scheme E without much more extensive quantitative analysis. Scheme F is also compatible with the results shown in Table 1. A further method of discriminating between schemes E and F would be to use a different retrograde tracer in each hemisphere of a single animal. In scheme E some neurones should accumulate only the contralateral tracer and the remainder should accumulate both whereas in scheme F some cells should contain either tracer alone and the remainder should contain both.
Projections fromTemporal Retina
463
2. Tectal Injections The dashed outlines in Fig. 3 show the region of the ipsilateral retina containing labelled neurones in 3 rats with injections aimed at the superior colliculus of one hemisphere. The region was in every case much smaller than that projecting to LGN which is compatible with the finding that only a small region of the ipsilateral rostro-lateral superior colliculus contains terminal label following eye injection of tritiated amino acid (Lund and Lund, 1973; Cowey and Franzini, 1979). Although we did not assess the proportion of labelled neurones in this region of the far temporal retina it was our clear impression that the proportion was definitely much lower than following LGN injections. The ipsilateral tectal projection is, therefore, very sparse. Discussion
One of the principal aims of the experiment was to assess the area of the retina from which uncrossed optic fibres arise, and to compare it with the estimate made in the preceding paper by a less satisfactory method. The two estimates agreed closely and show that about one quarter of the retinal area, concerned with the binocular visual field above the nose, contains ganglion cells that project ipsilaterally. However, it is clear that not all ganglion cells project solely to the ipsilateral hemisphere and that probably the majority do not, a result foreshadowed by the finding that terminal degeneration occurs in the LGN of both hemispheres following a small lesion in temporal retina of one eye (Lund et al., 1974). It is thus possible to reconcile the large extent of the ipsilaterally projecting retina with Polyak's estimate that fewer than 10 % of the optic fibres project ipsilaterally (Polyak, 1957). By comparing the proportions of HRP positive neurones in temporal and nasal retinae of the two eyes it was possible to reject four of the six possible arrangements of projections from the temporal retinae. We were unable to discriminate between the two remaining possibilities (E and F of Fig. 1) although much more extensive cell counts in HRP-labelled retinae or following dual tracer injection should provide an answer. However, it is clear that both schemes require that many ganglion cells in this region give rise to bifurcating axons that innervate both hemispheres, as suggested by Cunningham and Freeman (1977) on the basis of cobalt transport and the electrophysiological collision technique. Unfortunately, the function of bifurcating axons from retinal ganglion cells is mysterious but the fact that they are confined to the temporal retina strongly suggests that they are involved in binocular vision. In the immediately preceding paper Cowey and Franzini (1979) estimated the extent of the ipsilaterally projecting retina by counting only ganglion cells larger than 14-15 btm in rats with section of the optic chiasma or tract. The present results show that ganglion cells of all sizes contribute to the ipsilateral projection and, therefore, presumably to visual discrimination following section of the optic chiasma. The final and still puzzling point is that there is now incontrovertible evidence that nearly half the neurones in the ganglion cell layer of the rat do not
464
A. Cowey and V. H. Perry
s e n d an axon into the brain. T h e n e u r o n e s survive optic n e r v e or tract section (Eayrs, 1952; M a n t z a n d Klein, 1951; Perry, 1977; Cowey a n d F r a n z i n i , 1979) a n d do n o t a c c u m u l a t e H R P applied to the cut optic n e r v e (Perry, 1977) or i n j e c t e d into the b r a i n as here a n d by B u n t et al. (1974). F u r t h e r , H u g h e s (1977) has shown that t h e r e are n e a r l y twice as m a n y n e u r o n e s in the ganglion cell layer as there are fibres in the optic nerve, a n d Perry (1977) has described Golgi i m p r e g n a t e d n e u r o n e s in the g a n g l i o n cell layer with m o r p h o l o g y of a m a c r i n e cells. If these i n t e r n e u r o n e s are a m a c r i n e cells, recordings from single n e u r o n e s in the ganglion cell layer m a y give very m i s l e a d i n g i n f o r m a t i o n a b o u t the p r o p e r t i e s of g a n g l i o n cells in rats.
Acknowledgements. This work was supported by MRC Grant G 971/387/B. We thank Mrs. M. Walker for help with histology.
References Bunt, A.H., Lund, R.D., Lund, J.S.: Retrograde axonal transport of horseradish peroxidase by ganglion cells of the albino rat. Brain Res. 73,215-228 (1974) Coleman, D.R., Scalia, F., Fabrales, E.: Light and electron microscopic observations on the anterograde transport of horseradish peroxidase in the optic pathway in the mouse and rat. Brain Res. 102, 156-163 (1976) Cowey, A., Franzini, C.: The retinal origin of uncrossed optic nerve fibres in rats and their role in visual discrimination. Exp. Brain Res. 35, 443-455 (1979) Cunningham, T.J., Freeman, J.A.: Bilateral ganglion cell branches in the normal rat: a demonstration with electrophysiological collision and cobalt tracing methods. J. Comp. Neurol. 172, 165-176 (1977) Eayrs, J.T.: Relationship between the ganglion cell layer of the retina and the optic nerve in the rat. Br. J. Ophthalmol. 36, 453-459 (1952) Hughes, A.: The pigmented-rat optic nerve: fibre count and fibre spectrum. J. Comp. Neurol. 176, 263-268 (1977) K6nig, J, F. R., Klippel, R.A.: The rat brain: A stereotaxic atlas of the forebrain and lower parts of the brain stem, p. 162. Baltimore: Williams and Wilkins 1963 Lin, C.S., Kratz, K.E., Sherman, S.M.: Percentage of relay cells in the cat's geniculate nucleus. Brain Res. 131, 167-173 (1977) Lund, R. D., Lund, J.S.: Plasticity in the developing visual system: the effects of retinal lesions made in young rats. J. Comp. Neurol. 169, 133-154 (1973) Lund, R.D., Lurid, J.S., Wise, R.P.: The organization of the retinal projection to the dorsal lateral geniculate nucleus in pigmented and albino rats. J. Comp. Neurol. 158, 383-404 (1974) Mantz, J., Klein, M.: Modifications histologues de la r6tine apr6s interruptions du neff optique. C. R. Soc. Biol. (Paris) 145, 921-922 (1951) Perry, V.H.: Studies of the primary and secondary visual pathways. D. Phil. Dissertation, Oxford University 1977 Polyak, S.: The vertebrate visual system, pp. 1380. Chicago: University Press 1957 Stone, J.: A quantitative analysis of the distribution of ganglion cells in the cat's retina. J. Comp. Neurol. 124, 337-352 (1965) Vanegas, H., Hollander, H., Distel, H.: Early stages of uptake and transport of horseradish peroxidase by cortical structures, and its use for the study of local neurons and their processes. J. Comp. Neurol. 177, 193-212 (1978) Received September 4, 1978