The Development of lpsilateral Retinal Projections into the Tectum in the Cichlid Fish Haplochromis burfoni: A Dil Study in Fixed Tissue Bernd Fritzsch
*pt
and Claudia Wilm*
Faculty of Biology, University of Bielefeld, 4800 Bielefeld, Germany
SUMMARY The normal development of the retinal projection was studied in a bony fish with Dil. Between 5.5 and 10 days postfertilkation the contralateral retinal projection grows from the rostra1 pole of the tectum across its center. A maximum of 15 retinal fibers reaches the ipsilateral tectum. In 33-day-old juvenile animals, less than 15 ipsilateral fibers terminate in the entire tectum. Ipsilaterally projecting ganglion cells (maximal number = 20 cells) are scattered throughout the entire retina,
and the location of ganglion cells in the retina and axonal terminations in the tectum display a large interindividual variability. This suggests that the small adult contingent of ipsilateral fibers in this bony fish develops without an initial exuberant ipsilateral retinal projection that is later pruned back. 0 1992 John Wile) & Sons, Inc. Keywords: visual system, development, optic nerve, ipsilateral projections, teleosts, ganglion cell distribution.
INTRODUCTION
fish, which belong to derived bony fish, have an almost completely crossed retinal projection ( Wilm and Fritzsch, 1990). This absence ofan ipsilateral projection in the adult stage must be accomplished by modifying somehow the development of the ipsilateral retinal projection, the development of which is insufficiently known for any bony fish (Fritzsch, Wilm, and Crapon de Caprona, 1987; Wilm and Fritzsch, 1990). Among vertebrates at least two distinct patterns of development of an uncrossed, ipsilateral retinal projection have been identified thus far. ( 1 ) During development of the retinal projection in birds and mammals, substantial numbers of retinal fibers project ipsilaterally into the tectum (birds) or colliculus superior (mammals) at about the time a contralateral projection develops. In mammals this ipsilateral projection develops in two phases: at first a circumscribed population of ganglion cells from the ventrotemporal retina projects ipsilaterally ( Drager, 1985: Godement, Salaun, and Metin, 1987; Sretavan, 1990). Later, ganglion cells from the entire retina project ipsilaterally ( Drager, 1985; Guillery, 1989). These ipsilateral projections will subsequently be either altogether
The old notion that an uncrossed, ipsilateral retinal projection is unique to mammals has been disproven for a number of nonmammalian vertebrates ( Wilm and Fritzsch, 1990; for review). Experimental tract tracing data collected in the last 10 years imply instead that a bilateral retinal projection was the primitive feature forjawless (Wicht and Northcutt, 1990) and most likely jawed vertebrates as well (Fritzsch, 1991 ). Consistent with this interpretation is that among bony fish the more derived species have only a very small, if any, ipsilateral projection, whereas the less derived species have a more pronounced, uncrossed retinal projection (von Bartheld and Meyer, 1987). Adult cichlid Received August 27. 1991: accepted May 5. 1992 Journal of Neurobiology, Vol. 23, No. 6. pp. 708-7 19 (1992) Q 1992 John Wiley & Sons. Inc. CCC 0022-3034/92/060708- I 2$04.00 * To whom correspondence should be addressed. Present address: t Department of Biomedical Sciences, Division of Anatomy. Creighton University, Omaha. Nebraska 68178, U.S.A.; $Biologische Forschung. E. Merck, P.O. Box 41 19, 6100 Darmstadt I , Federal Republic of Germany
708
Table 1 Labeling of the Retinal Projection with Dil and FDA Labeling Age of the Animals
N
The Other Optic Nerve
One Optic Nerve
5.5-10 days 33days
64
Dil
-
43
Dil
FDA
33 days
14
1-3 years
11
Dil into one tectum HRP
Preparation Tecta as whole mounts Tecta, and of 10 animals the other retina Both retinae
Brain was cut frozen
eliminated (birds) or reduced (mammals; Provis and Penfold, 1988; for review). Development of the retinal projection thus follows the overall developmental pattern of amniotic vertebrates characterized by an overproduction followed by a phase of sculpting and pruning to achieve the adult pattern (Cowan, Fawcett, O’Leary, and Stanfield, 1984). ( 2 ) In many amphibians and in lampreys an uncrossed projection develops much later, for example, around metamorphosis (Currie and Cowan, 1974; Hoskins and Grobstein, 1985a,b; Kennedy and Rubinson, 1977; Rubinson, 1990), that is, long after a crossed retinal projection has developed. In contrast to amniotic vertebrates, no exuberant ipsilateral projection followed by a phase of pruning has been described in these animals. Table 2 Normal Retinal Projection ( 4 5 Fibers) into the Ipsilateral Tectum Number (N) and Percentage (9%) of Animals with Ipsilateral Retinal Fibers
Groups Age of the Animals 5.5-6 days 6.5-7 days 7.5-8 days 8.5-9 days 9.5-10 days 33 days 1-3 years
Number of Animals 7 11
14 8 13 43 11
N
9%
2 5 5 5 12 41 2
29 45 36 63 92 95 18
Figure 1 Dil-labeled retinal projection into the contralateral tectum o f a 5.5- to 6-day-old larva. The tectum is shown as a whole mount from a dorsal aspect. Retinal fibers project in a broad front into the tectum. C = caudal; V = ventrolateral. Scale bar = 100 win.
Absence of an ipsilateral projection in adult cichlid fish could consequently be brought about either by an excessive reduction of an initially larger ipsilateral retinal projection (i.e., by taking the amniotic developmental pattern towards one extreme, like in birds) or by a complete suppression of any formation of an ipsilateral projection (i.e., by heterochronically modifying the anamniotic developmental pattern). We undertook a study of the development of the retinal projection of the cichlid fish, Huplochromis burtoni, employing the diffusion of Dil in fixed tissue to gain insights into which developmental modification has led to the virtual absence of an ipsilateral projection in this bony fish. In a previous and two accompanying studies, we showed that regenerating retinal axons develop a prominent ipsilateral projection ( Wilm and Fritzsch, 1990, 1992a,b).A second aim ofthe present study is therefore to reveal whether regeneration recapitulates ontogenetic processes, or in other words: Is there an enhanced ipsilateral projection during development as it occurs during regeneration? Preliminary results have been presented in abstract form ( Wilm, 1990).
710
Fr.i/r.vchaiid W'iltn
Figure 2 Dil-labeled retinal projection of a 5.5- to 6-day-old larva. The brain is shown as a whole mount from ventral. Few fibers (arrows) project from the tractus opticus (TrO) to the ipsilateral side. The stippled line marks the midline of the brain. R = rosiral: other abbreviations as per Figure I . Scale bar = 100 Fm.
METHODS For this study, a total of I 2 1 specimens ofthe cichlid fish IIciplochrotvis hiirloni were subjected to postmortem procedures. The breeding animals and the larvae were keptat25" i I"C(light/darkcycle: 13:l 1 h ) . T h e a g e o f the animals was dctermined according to a previously established table of developmental stages ( Wilm and Fritzsch, 1989). Free-swimming juveniles were kept at room temperature ( 2 1 O -t 1 " C )in small aquaria and fed with dry food for young brood (Tetramin).
Anterograde Tracing of the Visual Projection The short distance (only 1-3 m m in 5.5- to 33-day-old animals) of the retinal projection allows the use of Dil ( I . 1 '-dioctadecyl-3,3.3',3'-tetramethyl indocarbocyanine perchlorate, Molecular Probes. Eugene, OR, U.S.A.) in paraformaldehyde-fixed tissue. Table 1 lists the different groups and combinations of the dyes. In 33-day-old juveniles one optic nerve was labeled with fluorescein-isothiocyanate-coupled dextran amine ( F D A ) (Molecular Probes). FDA was applied in deeply anesthetized animals onto the cut optic nerve. The animals survived 14- 16 h for sufficient transport of the dye (Wilm and Fritzsch, 1990). Prior to fixation all animals were anesthetized in 0.0 1 % ethyl p-aminobenzoate (benzocainc, Sigma). The skull was opened, the animals were immersed in 4% paraformaldehyde in 0. I M phosphate buffer (pH 7.3) at room temperature and then stored at 4°C. At 1 day to 4 weeks later, crystals of Dil
were applied to thc optic nerve head inside one eye. The animals werc either stored in the dark at 19" to 2 1 "C for 1 I days (if one retinal projection was labcled with FDA) or at 36°C for 8 h to 2 days depending on the age (size) of the animal. For each age minimal diffusion times to label the retinotectal projection were established to avoid transcellular labeling (Fritzsch and Wilm, 1990). The material was storcd at 4°C prior to examination to minimize further diffusion. The tectum together with the optic tracts were dissected. The tectum was split along the rostrocaudal meridian, sparing the rostral pole and flat-mounted onto a slide in 0.1 .ZZ phosphate buffer ( p H 7.3). Two small coverslips on each side served as spacers. The fluorescent material was viewed with an cpifluorescence microscope and photographed with a rhodamine or fluorescein filter set (TMAX 100 ASA or Ektachromc 200 ASA. Kodak). In I I adult animals, horseradish peroxidase ( H R P ) was applied onto one optic ncrve ( Wilm and Fritzsch, 1990).
Retrograde Tracing of the lpsilaterally Projecting Ganglion Cells Dil was applied into the rostral tcctum ofparaformaldehyde-fixed animals (Table 1 ). A shallow incision was placed transversely across the rostral left tectum from dorsomedial to ventrolateral. Crystals of Dil were stuffed into the incision except at the dorsomedial margin to avoid difusion of Dil into the other tectum. The material was stored in 4%)paraformaldehyde in 0.1 M phosphate buffer, pH 7.3, at 36°C for 2 days. The retinae were dissected and whole mounted in phosphate buffer
Figure 3 Retinal projection of a 6.5- to 7-day-old larva. The left tectum is contralateral and the right tecium is ipsilateral to the labeled eye. Both tecta are shown as a whole mount from a tectum is aspect. Four unbranched retinal fibers (arrow) project into the ipsilateral tectum. The stippled line marks the rostral margin ofthe tectum. Tec = tectum; T r O = tractus opticus; other abbreviations as per Figure 1. Scale bar = 200 pm.
( p H 7.3) onto a glass slide and examined as described above.
RESULTS
Development of the Retinal Projection
Five larval stages (5.5-6 days, 6.5-7 days, 7.5-8 days, 8.5-9 days, and 9.5-10 days old) and one juvenile stage (33 days old) were investigated (Table 2 ) . At 5.5-6 Days. At 5.5 days postfertilization, the larvae hatched. The eyes were pigmented. Retinal fibers reached the rostral pole of the contralateral tectum in a broad front without obvious pioneer fibers (Fig. 1 ). In two out of seven animals, one or two unbranched retinal fibers coursed in the ipsilateral optic tract to the rostral tectal pole. In three animals, some retinal fibers crossed in the ventral diencephalon to the ipsilateral side but did not reach the tectum (Fig. 2).
A t 6.5-7 Days. The contralateral fibers reached the center of the tectum (Fig. 3 ) . In five out of 1 1 animals, fibers ( u p to 12 but mostly less than 5 ) coursed to the rostral pole of the ipsilateral tectum. The ipsilateral fibers were still unbranched (Fig. 3). In another batch, where Dil had been allowed to diffuse over an extended period of time, several cells were labeled transcellularly along the path of retinal fibers (Fig. 4 ) .
At 7.5-8 Days. Retinal fibers covered about onethird of the rostrocaudal extent of the contralateral tectum. No retinal fibers were at the lateral and medial margin. In five out of 14 animals, up to 15 fibers, mostly unbranched or rarely with short branches [Fig. 5 (a,b)], reached the rostral pole of the ipsilateral tectum via the optic tract. In another five animals, few retinal fibers recrossed in the diencephalon, but did not project into the tectum. A t 8.5-9 Days. Retinal fibers covered one-third to one-half of the rostrocaudal extent of the contralateral tectum [Fig. 6( a)]. Compared to 6.5-day-old animals, retinal fibers not only progressed from
(Table 3), and were rare at the caudal margin. Few fibers had crossed the midline between both tecta. These fibers displayed a more profuse branching than did the direct ipsilateral fibers and terminated in deeper layers. Ipsilateral fibers were never doubly labeled with Dil and FDA. In 10 animals, all of which exhibited ipsilaterally coursing retinal fibers, no ganglion cells were labeled in the other retina. Therefore, ipsilateral fibers were not transcellularly labeled by diffusion of Dil into the other optic nerve. Ipsilateral fibers ran in all stages singly towards and within the tectum. Adult Animals. In HRP-labeled projections, two out of 1 1 animals showed less than 10 fibers which coursed via the optic tract into the tectum and terminated in the stratum fibrosum et griseum superficiale ( Wilm and Fritzsch, 1990).
Figure 4 After long-lasting diffusion of Dil, cells along the path of growing axons become labeled, presumably transcellularly (arrows). The stippled line marks the region of the optic tract and the growing retinal fibers. Retinal projection of a 6.5-day-old larva. Abbreviations as per Figures I and 3. Scale bar = 100 fim.
rostral to caudal but also populated more of the lateral and medial margin of the tectum. Up to three retinal fibers coursed in five out of eight animals via the optic tract into the rostral ipsilateral tectum. lpsilateral fibers did not extend as far caudal in the tectum as did contralateral fibers. No fibers crossed the midline between the two tecta. At 9.5-10 Days. The contralateral retinal projection covered two-thirds of the rostrocaudal extent of the tectum [Fig. 6 ( b ) ] . Few fibers had projected across the midline into the ipsilateral tectum. In 12 out of 13 animals, few (< 10) fibers projected via the optic tract into the ipsilateral tectum [Fig. 5 (c)] . These fibers developed several small branches along their course in the tectum. At 33 Days. The contralateral retinal projection covered densely the entire tectum (Fig. 7 ) . In 95% of the animals, few retinal fibers (
*pt
and Claudia Wilm*
Faculty of Biology, University of Bielefeld, 4800 Bielefeld, Germany
SUMMARY The normal development of the retinal projection was studied in a bony fish with Dil. Between 5.5 and 10 days postfertilkation the contralateral retinal projection grows from the rostra1 pole of the tectum across its center. A maximum of 15 retinal fibers reaches the ipsilateral tectum. In 33-day-old juvenile animals, less than 15 ipsilateral fibers terminate in the entire tectum. Ipsilaterally projecting ganglion cells (maximal number = 20 cells) are scattered throughout the entire retina,
and the location of ganglion cells in the retina and axonal terminations in the tectum display a large interindividual variability. This suggests that the small adult contingent of ipsilateral fibers in this bony fish develops without an initial exuberant ipsilateral retinal projection that is later pruned back. 0 1992 John Wile) & Sons, Inc. Keywords: visual system, development, optic nerve, ipsilateral projections, teleosts, ganglion cell distribution.
INTRODUCTION
fish, which belong to derived bony fish, have an almost completely crossed retinal projection ( Wilm and Fritzsch, 1990). This absence ofan ipsilateral projection in the adult stage must be accomplished by modifying somehow the development of the ipsilateral retinal projection, the development of which is insufficiently known for any bony fish (Fritzsch, Wilm, and Crapon de Caprona, 1987; Wilm and Fritzsch, 1990). Among vertebrates at least two distinct patterns of development of an uncrossed, ipsilateral retinal projection have been identified thus far. ( 1 ) During development of the retinal projection in birds and mammals, substantial numbers of retinal fibers project ipsilaterally into the tectum (birds) or colliculus superior (mammals) at about the time a contralateral projection develops. In mammals this ipsilateral projection develops in two phases: at first a circumscribed population of ganglion cells from the ventrotemporal retina projects ipsilaterally ( Drager, 1985: Godement, Salaun, and Metin, 1987; Sretavan, 1990). Later, ganglion cells from the entire retina project ipsilaterally ( Drager, 1985; Guillery, 1989). These ipsilateral projections will subsequently be either altogether
The old notion that an uncrossed, ipsilateral retinal projection is unique to mammals has been disproven for a number of nonmammalian vertebrates ( Wilm and Fritzsch, 1990; for review). Experimental tract tracing data collected in the last 10 years imply instead that a bilateral retinal projection was the primitive feature forjawless (Wicht and Northcutt, 1990) and most likely jawed vertebrates as well (Fritzsch, 1991 ). Consistent with this interpretation is that among bony fish the more derived species have only a very small, if any, ipsilateral projection, whereas the less derived species have a more pronounced, uncrossed retinal projection (von Bartheld and Meyer, 1987). Adult cichlid Received August 27. 1991: accepted May 5. 1992 Journal of Neurobiology, Vol. 23, No. 6. pp. 708-7 19 (1992) Q 1992 John Wiley & Sons. Inc. CCC 0022-3034/92/060708- I 2$04.00 * To whom correspondence should be addressed. Present address: t Department of Biomedical Sciences, Division of Anatomy. Creighton University, Omaha. Nebraska 68178, U.S.A.; $Biologische Forschung. E. Merck, P.O. Box 41 19, 6100 Darmstadt I , Federal Republic of Germany
708
Table 1 Labeling of the Retinal Projection with Dil and FDA Labeling Age of the Animals
N
The Other Optic Nerve
One Optic Nerve
5.5-10 days 33days
64
Dil
-
43
Dil
FDA
33 days
14
1-3 years
11
Dil into one tectum HRP
Preparation Tecta as whole mounts Tecta, and of 10 animals the other retina Both retinae
Brain was cut frozen
eliminated (birds) or reduced (mammals; Provis and Penfold, 1988; for review). Development of the retinal projection thus follows the overall developmental pattern of amniotic vertebrates characterized by an overproduction followed by a phase of sculpting and pruning to achieve the adult pattern (Cowan, Fawcett, O’Leary, and Stanfield, 1984). ( 2 ) In many amphibians and in lampreys an uncrossed projection develops much later, for example, around metamorphosis (Currie and Cowan, 1974; Hoskins and Grobstein, 1985a,b; Kennedy and Rubinson, 1977; Rubinson, 1990), that is, long after a crossed retinal projection has developed. In contrast to amniotic vertebrates, no exuberant ipsilateral projection followed by a phase of pruning has been described in these animals. Table 2 Normal Retinal Projection ( 4 5 Fibers) into the Ipsilateral Tectum Number (N) and Percentage (9%) of Animals with Ipsilateral Retinal Fibers
Groups Age of the Animals 5.5-6 days 6.5-7 days 7.5-8 days 8.5-9 days 9.5-10 days 33 days 1-3 years
Number of Animals 7 11
14 8 13 43 11
N
9%
2 5 5 5 12 41 2
29 45 36 63 92 95 18
Figure 1 Dil-labeled retinal projection into the contralateral tectum o f a 5.5- to 6-day-old larva. The tectum is shown as a whole mount from a dorsal aspect. Retinal fibers project in a broad front into the tectum. C = caudal; V = ventrolateral. Scale bar = 100 win.
Absence of an ipsilateral projection in adult cichlid fish could consequently be brought about either by an excessive reduction of an initially larger ipsilateral retinal projection (i.e., by taking the amniotic developmental pattern towards one extreme, like in birds) or by a complete suppression of any formation of an ipsilateral projection (i.e., by heterochronically modifying the anamniotic developmental pattern). We undertook a study of the development of the retinal projection of the cichlid fish, Huplochromis burtoni, employing the diffusion of Dil in fixed tissue to gain insights into which developmental modification has led to the virtual absence of an ipsilateral projection in this bony fish. In a previous and two accompanying studies, we showed that regenerating retinal axons develop a prominent ipsilateral projection ( Wilm and Fritzsch, 1990, 1992a,b).A second aim ofthe present study is therefore to reveal whether regeneration recapitulates ontogenetic processes, or in other words: Is there an enhanced ipsilateral projection during development as it occurs during regeneration? Preliminary results have been presented in abstract form ( Wilm, 1990).
710
Fr.i/r.vchaiid W'iltn
Figure 2 Dil-labeled retinal projection of a 5.5- to 6-day-old larva. The brain is shown as a whole mount from ventral. Few fibers (arrows) project from the tractus opticus (TrO) to the ipsilateral side. The stippled line marks the midline of the brain. R = rosiral: other abbreviations as per Figure I . Scale bar = 100 Fm.
METHODS For this study, a total of I 2 1 specimens ofthe cichlid fish IIciplochrotvis hiirloni were subjected to postmortem procedures. The breeding animals and the larvae were keptat25" i I"C(light/darkcycle: 13:l 1 h ) . T h e a g e o f the animals was dctermined according to a previously established table of developmental stages ( Wilm and Fritzsch, 1989). Free-swimming juveniles were kept at room temperature ( 2 1 O -t 1 " C )in small aquaria and fed with dry food for young brood (Tetramin).
Anterograde Tracing of the Visual Projection The short distance (only 1-3 m m in 5.5- to 33-day-old animals) of the retinal projection allows the use of Dil ( I . 1 '-dioctadecyl-3,3.3',3'-tetramethyl indocarbocyanine perchlorate, Molecular Probes. Eugene, OR, U.S.A.) in paraformaldehyde-fixed tissue. Table 1 lists the different groups and combinations of the dyes. In 33-day-old juveniles one optic nerve was labeled with fluorescein-isothiocyanate-coupled dextran amine ( F D A ) (Molecular Probes). FDA was applied in deeply anesthetized animals onto the cut optic nerve. The animals survived 14- 16 h for sufficient transport of the dye (Wilm and Fritzsch, 1990). Prior to fixation all animals were anesthetized in 0.0 1 % ethyl p-aminobenzoate (benzocainc, Sigma). The skull was opened, the animals were immersed in 4% paraformaldehyde in 0. I M phosphate buffer (pH 7.3) at room temperature and then stored at 4°C. At 1 day to 4 weeks later, crystals of Dil
were applied to thc optic nerve head inside one eye. The animals werc either stored in the dark at 19" to 2 1 "C for 1 I days (if one retinal projection was labcled with FDA) or at 36°C for 8 h to 2 days depending on the age (size) of the animal. For each age minimal diffusion times to label the retinotectal projection were established to avoid transcellular labeling (Fritzsch and Wilm, 1990). The material was storcd at 4°C prior to examination to minimize further diffusion. The tectum together with the optic tracts were dissected. The tectum was split along the rostrocaudal meridian, sparing the rostral pole and flat-mounted onto a slide in 0.1 .ZZ phosphate buffer ( p H 7.3). Two small coverslips on each side served as spacers. The fluorescent material was viewed with an cpifluorescence microscope and photographed with a rhodamine or fluorescein filter set (TMAX 100 ASA or Ektachromc 200 ASA. Kodak). In I I adult animals, horseradish peroxidase ( H R P ) was applied onto one optic ncrve ( Wilm and Fritzsch, 1990).
Retrograde Tracing of the lpsilaterally Projecting Ganglion Cells Dil was applied into the rostral tcctum ofparaformaldehyde-fixed animals (Table 1 ). A shallow incision was placed transversely across the rostral left tectum from dorsomedial to ventrolateral. Crystals of Dil were stuffed into the incision except at the dorsomedial margin to avoid difusion of Dil into the other tectum. The material was stored in 4%)paraformaldehyde in 0.1 M phosphate buffer, pH 7.3, at 36°C for 2 days. The retinae were dissected and whole mounted in phosphate buffer
Figure 3 Retinal projection of a 6.5- to 7-day-old larva. The left tectum is contralateral and the right tecium is ipsilateral to the labeled eye. Both tecta are shown as a whole mount from a tectum is aspect. Four unbranched retinal fibers (arrow) project into the ipsilateral tectum. The stippled line marks the rostral margin ofthe tectum. Tec = tectum; T r O = tractus opticus; other abbreviations as per Figure 1. Scale bar = 200 pm.
( p H 7.3) onto a glass slide and examined as described above.
RESULTS
Development of the Retinal Projection
Five larval stages (5.5-6 days, 6.5-7 days, 7.5-8 days, 8.5-9 days, and 9.5-10 days old) and one juvenile stage (33 days old) were investigated (Table 2 ) . At 5.5-6 Days. At 5.5 days postfertilization, the larvae hatched. The eyes were pigmented. Retinal fibers reached the rostral pole of the contralateral tectum in a broad front without obvious pioneer fibers (Fig. 1 ). In two out of seven animals, one or two unbranched retinal fibers coursed in the ipsilateral optic tract to the rostral tectal pole. In three animals, some retinal fibers crossed in the ventral diencephalon to the ipsilateral side but did not reach the tectum (Fig. 2).
A t 6.5-7 Days. The contralateral fibers reached the center of the tectum (Fig. 3 ) . In five out of 1 1 animals, fibers ( u p to 12 but mostly less than 5 ) coursed to the rostral pole of the ipsilateral tectum. The ipsilateral fibers were still unbranched (Fig. 3). In another batch, where Dil had been allowed to diffuse over an extended period of time, several cells were labeled transcellularly along the path of retinal fibers (Fig. 4 ) .
At 7.5-8 Days. Retinal fibers covered about onethird of the rostrocaudal extent of the contralateral tectum. No retinal fibers were at the lateral and medial margin. In five out of 14 animals, up to 15 fibers, mostly unbranched or rarely with short branches [Fig. 5 (a,b)], reached the rostral pole of the ipsilateral tectum via the optic tract. In another five animals, few retinal fibers recrossed in the diencephalon, but did not project into the tectum. A t 8.5-9 Days. Retinal fibers covered one-third to one-half of the rostrocaudal extent of the contralateral tectum [Fig. 6( a)]. Compared to 6.5-day-old animals, retinal fibers not only progressed from
(Table 3), and were rare at the caudal margin. Few fibers had crossed the midline between both tecta. These fibers displayed a more profuse branching than did the direct ipsilateral fibers and terminated in deeper layers. Ipsilateral fibers were never doubly labeled with Dil and FDA. In 10 animals, all of which exhibited ipsilaterally coursing retinal fibers, no ganglion cells were labeled in the other retina. Therefore, ipsilateral fibers were not transcellularly labeled by diffusion of Dil into the other optic nerve. Ipsilateral fibers ran in all stages singly towards and within the tectum. Adult Animals. In HRP-labeled projections, two out of 1 1 animals showed less than 10 fibers which coursed via the optic tract into the tectum and terminated in the stratum fibrosum et griseum superficiale ( Wilm and Fritzsch, 1990).
Figure 4 After long-lasting diffusion of Dil, cells along the path of growing axons become labeled, presumably transcellularly (arrows). The stippled line marks the region of the optic tract and the growing retinal fibers. Retinal projection of a 6.5-day-old larva. Abbreviations as per Figures I and 3. Scale bar = 100 fim.
rostral to caudal but also populated more of the lateral and medial margin of the tectum. Up to three retinal fibers coursed in five out of eight animals via the optic tract into the rostral ipsilateral tectum. lpsilateral fibers did not extend as far caudal in the tectum as did contralateral fibers. No fibers crossed the midline between the two tecta. At 9.5-10 Days. The contralateral retinal projection covered two-thirds of the rostrocaudal extent of the tectum [Fig. 6 ( b ) ] . Few fibers had projected across the midline into the ipsilateral tectum. In 12 out of 13 animals, few (< 10) fibers projected via the optic tract into the ipsilateral tectum [Fig. 5 (c)] . These fibers developed several small branches along their course in the tectum. At 33 Days. The contralateral retinal projection covered densely the entire tectum (Fig. 7 ) . In 95% of the animals, few retinal fibers (
