/. Embryo/, exp. Morph. Vol. 52, pp. 89-103, 1979
Printed in Great Britain © Company of Biologists Limited 1979
§9
Interactions between optic fibres
controlling the locations of their terminals in the
goldfish optic tectum
By JEREMY E. COOK 1
From the University Laboratory of Physiology, Oxford
SUMMARY
Removal of the caudal half of a goldfish optic tectum induces optic fibres from the entire
contralateral retina to terminate retinotopically within the remaining half. This compression
has been viewed by some as the result of competition between the fibres and by others as a
consequence of changes, induced by the surgery, in tectal labels guiding fibres to terminal
sites.
To distinguish between these possibilities, the time-course of compression has been
measured by electrophysiological mapping of the visual projection. In some fish, fibres
terminating in the rostral half-tectum remained intact when the caudal half was removed.
In others, the optic nerve was cut at the time of tectal surgery: even after its regeneration
into a half-tectum, optic terminals were first detected in the regions they normally occupy.
The subsequent reorganization was gradual and retinotopic order was maintained. However,
it was slower where some fibres had never been cut.
In a third series the nerve was cut 18 days before the tectal halving to reveal any dependence
of compression on progressive changes in the halved tectum; but its time-course from nerve
section was found to be independent of the time within the regeneration period at which the
tectum was halved.
In a fourth series the nerve was cut at the time of tectal halving and then cut again after
85-97 days when compression was complete to reveal any permanent change in the halved
tectum. No change was evident: the previous compression did not preclude subsequent
regeneration of an uncompressed projection and its gradual recompression as before.
In a fifth series, repeated crushing of fibres normally ending in the missing caudal tectum
temporarily prevented compression among the remainder, while crushing of fibres destined
for rostral tectum caused transposition of the remaining projection to the rostral half.
Surgically induced changes in the labels which are thought to guide growing fibres to
their normal tectal regions do not account for these results. Indeed, this guidance persists
unchanged for fibres regenerating a second time after compression. Since compression is
delayed while certain fibres are withheld, it appears instead to be the direct result of competition between the fibres. The maintenance of retinotopic order in compression, despite
unchanged tectal guidance, may require selective interactions between fibres from different
retinal regions which could contribute to the refinement of the normal visual projection.
1
Author's present address; c/o Dr T. J. Horder, Department of Human Anatomy, South
Parks Road, Oxford OXl 3QX, Great Britain.
90
J. E. COOK
INTRODUCTION
Regeneration of a severed optic nerve in a lower vertebrate can lead to
restoration of the normal topographic fibre projection from the retinal ganglion
cell array to the contralateral optic tectum. Sperry (1943 a, b, 1944, 1948) found
that the visuomotor responses of newts, frogs and fish were normal after
regeneration of the optic nerve from a normal eye but were permanently and
systematically inverted after regeneration of the nerve from an inverted eye. He
concluded that fibres probably regenerated to their proper tectal regions. They
did this in the inverted case, he argued, despite both the adverse functional
consequences and the fact that they were confronted by inappropriate central
paths because of rotation, displacement and scarring at the nerve lesion site.
He therefore proposed that parallel differentiation within the developing retina
and tectum might generate matching arrays of position-related cytochemical
labels, or specificities, which control optic fibre connexions through selective
affinity between correspondingly labelled retinal and tectal cells.
Electrophysiological recording from optic fibre terminals in the tectum has
confirmed that fibres normally regenerate to their proper tectal regions after
nerve section in adult frogs (Gaze, 1959; Maturana, Lettvin, McCulloch &
Pitts, 1959) and fish (Jacobson & Gaze, 1965; Horder, 1971). Attardi & Sperry
(1963) showed anatomically that newly regenerated fibres from areas of goldfish retina which remained after partial retinal ablation grew almost exclusively
into their proper branch of the optic tract, and could avoid inappropriate tectal
regions to arborize in appropriate regions. Further, fibres from the entire retina
of a 'double-ventral compound eye' in the frog Xenopus reach the tectum by
the branch of the optic tract appropriate to normal ventral retina, the other
branch being empty (Cook & Horder, 1977). This implies active route selection
by fibres from the anatomically dorsal half of the 'compound' retina. All this
is evidence that there are indeed position-related differences between the fibres
of retinal ganglion cells and suggests that complementary differences exist to
guide them in the pathway to their tectal sites.
However, new and unexpected properties of the adult goldfish visual system
were found by Gaze & Sharma (1970). Within 3 months of removal of the
caudal half of a tectum, the entire retina established a retinotopic projection
within the remaining rostral half: this could occur with or without crush and
regeneration of the nerve. Autoradiographic tracing has recently confirmed that
these changes in the electrophysiologically mapped projection are accompanied
by corresponding changes in the positions of the optic fibre terminals (Cook &
Horder, 1977). Gaze & Sharma suggested that the changes might be the result
of competitive reinnervation. Yoon (1972) induced this compression of the
whole visual map on the rostral half-tectum by implanting foil or gelatin
barriers across the tectum: their removal reversed the compression. He suggested, in contrast, that the changes were the direct consequence of a surgically
Interactions controlling optic terminal location
91
induced ^differentiation of the cytochemical labels which Sperry had proposed
for the tectal elements. Such a suggested redifferentiation in the adult fish has
been compared with the embryonic phenomenon of regulation (Gaze, 1974).
The work to be described explores the effects of different surgical procedures,
and of their timing, on compression of the visual projection in goldfish, after
removal of the caudal half-tectum. The procedures were chosen with the aim of
distinguishing between the mechanisms mentioned above. Some of the results
have been published in abstract form (Cook & Horder, 1974\
METHODS
Experimental animals
Goldfish (Carassius auratus) 54-63 mm long (snout to tail-fin base) were kept
in large tanks in natural light, at 28 °C which promotes somewhat faster
regeneration than does room temperature (Springer & Agranoff, 1977) and is
readily controlled for consistent results. All surgery was performed under a
dissecting microscope after immersion of the fish in 5 % urethane solution until
immobile. Tap water was then passed through a mouth tube at 100-400 ml/min,
anaesthesia being deepened if necessary by substituting urethane solution.
The bony plate overlying the tecta was freed with a scalpel and hinged by
skin along its rostral edge. The left optic tectum was cut through to the lateral
ventricle with fine scissors, along a medio-lateral line judged to be exactly midway between the visible rostral and caudal tectal limits. This bisection was very
consistent, which outweighed the disadvantage that it divided the tectum into
two parts of not quite equal area. Electrophysiological maps made immediately
afterwards from five fish showed scotomata covering a mean of 80-3 % of the
dorsotemporal visual field quadrant (measured as detailed below) with a
standard deviation of only 6-0% which includes errors of mapping. The caudal
half was then excised: only its more accessible dorsal and dorsolateral regions
were normally removed but in one series the remaining ventrolateral fragment
was also extracted with a wire loop. The bone flap was hinged into place, where
angled scalpel cuts held it without adhesive.
Right optic nerves were cut midway in the orbit with fine scissors under
direct microscopic vision. Fish in which the accompanying blood vessels were
damaged were not used. The temporal or nasal halves of nerves were crushed
with very slender forceps after the insertion of a needle-point diametrically
through the nerve to define the limit of the lesion. To divide fibres from nasal
and temporal retina as far as possible, the orientation of this thrust was adjusted
by mapping the scotomata immediately after crushing in several fish. Though
not sharply defined, each scotoma was predominantly confined to either the
rostral or the caudal half-tectum, implying that nasal and temporal fibres were
distinct at the lesion site. Judging by the orientation required, the right retinal
projection was rotated some 30° anticlockwise at this point.
92
J. E. COOK
Mapping the visual projection
The tecta were exposed and covered by light mineral oil. The right eye was
centred in a modified Aimark perimeter and the fish oriented so that the plane
of the pupil was perpendicular to the perimeter axis: seen through an axially
mounted ophthalmoscope the corneal reflexion of the ophthalmoscope light
was central in the pupil. In some series the optic disc was also visualized. Its
image lay, using these methods, about 10° dorsonasal to the perimeter axis in
the visual field. A grid was drawn on a Polaroid photograph of the dorsal aspect
of the left tectum and the blood vessels were used to identify sites for electrode
insertion corresponding to intersections on the grid. The most caudal row of
these intersections was always drawn as close to the cut edge of the half-tectum
as possible. A varnished tungsten wire electrode of low impedance with a tip
radius of a few micrometres fed a Tektronix type 122 preamplifier, high-pass
filter, audio amplifier and speaker. Responses to visual stimuli could be heard
at each electrode site from several superficial units, believed to be the terminal
arborizations of optic nerve fibres (Gaze & Sharma, 1970). The visual stimulus,
a white disc (diameter 8°) against a dark ground, was moved in the field to
determine a position of maximum response. This was represented conventionally on a polar field chart, with two changes discussed below. Rows of visual
field positions corresponding to mediolateral rows of electrode positions were
plotted for each fish, which was mapped only once.
Analysing projection maps
To quantify the results and simplify presentation, a method was devised for
measuring the amount of visual field represented on the remaining rostral halftectum in each fish. Since the maps obtained were always retinotopically ordered,
field positions projecting to the row of electrode sites lying as close as possible
to the caudal edge of the half-tectum always represented the temporal limit of
vision. The projection of the nasal limit could not be found reliably, owing to
the tectal curvature, so the position of the temporal limit was compared instead
with arbitrary boundaries in the field by measuring areas defined by lines drawn
on the field charts. The area measured on each chart lay nasal to the line
representing the temporal limit of vision, inside the dorsotemporal quadrant
of the chart and within the circle representing 90°. It was sometimes necessary
to extend the line representing the temporal field limit in order to enclose an
area in this way, but the convention used throughout to do this led to minor
inconsistencies only. The measured area is shaded on each projection map
shown, and any extrapolated line is dashed.
The method depends on careful orientation of the eye in the perimeter, and
two changes to the usual charting procedure improved the consistency of the
results without affecting general trends. First, the naso-temporal axis of the
polar chart was made to represent not the horizontal plane of the perimeter
Interactions controlling optic terminal location
Temporal
93
Nasal
Fig. 1. Chart used to represent the visual field. The radius of each circle is proportional to the sine of half the visual angle with the optic axis that it represents:
the area it encloses is proportional to the solid angle subtended at the eye.
but a reference plane, normally near the horizontal, passing through the optic
axis and parallel to two marks (nasal and temporal) which are a consistent
feature of the iris; this compensates for small varied rotations of different eyes
about the optic axis. Secondly, the radial distance of a point on the chart (representing a field position) from the chart centre (representing the optic axis) was
here made proportional to the sine of half the visual angle with the optic axis
subtended by that field position. Areas measured on the surface of a conventional chart, where radial distance is directly proportional to visual angle,
represent varied solid angles of vision according to their position. The premise
that the relationship between chart area and solid visual angle should instead be
constant throughout the field leads directly to this simple trigonometric proportionality between chart radius and plane visual angle. Using the modified
chart shown in Fig. 1, any chart area is therefore directly proportional to the
corresponding solid angle of vision and provides a better estimate of the retinal
area involved. However, it is important to note that the general appearance of
the projection map is virtually unchanged.
In the graphs the vertical axis shows each measured chart area as a percentage
of the area of the entire dorsotemporal chart quadrant inside the 90° circle. The
dorsotemporal field of a normal goldfish in air covers most or all of the corresponding field quadrant. About 20% of that quadrant normally projects to
what is here called, throughout, the rostral half-tectum (from its appearance
in dorsal view at surgery) but which actually comprises more than half of the
tectal area. Compression, after ablation of the caudal half, leads to an increase
in the proportion of the dorsotemporal field quadrant projecting to the rostral
half. This proportion may sometimes reach 100 %, but caution is essential in
making any inference from its absolute value because of the arbitrary limits
used. Direct comparison with the extent of the normal projection is difficult,
7
EMB 52
94
J. E. COOK
Left tectum
(b)
Right visual field
Right visual field
Fig. 2. Visual projection maps from two fish with halved left tecta. (a) Directly after
tectal halving, (b) 136 days after tectal halving. Each number on the dorsal view of
each left half tectum (copied from a Polaroid photograph) represents an electrode
position; each electrode position is most strongly stimulated from the position with
the same number in the visual field of the contralateral eye. The arc round each field
represents a visual angle of 90° with the optic axis. The arrow by each tectum lies on
the midline of the fish and points rostrally; its length represents about 0-4 mm. The
shaded areas are represented by two points in Fig. 3 (21 % and 73 %).
and retino-tectal maps made in air are always likely to be incomplete because
the retinal image is diminished.
RESULTS
(1) Tectal halving alone
The caudal half of the left optic tectum was excised from each of a series of
goldfish. Twenty-three of these were subsequently mapped as described above
to assess the timing of compression. Immediately after tectal surgery all remaining accessible regions, including the border of the cut, responded to visual
stimuli just as in an intact tectum. The visual projection of the right eye of such
a fish is shown in Fig. 2(a). Comparison with the corresponding half of the
normal projection (from a left eye) shown in Fig. 2 of Jacobson & Gaze (1964),
points 1-33, reveals a close similarity despite the chart modification.
The first few fish mapped after the day of surgery gave similar results but a
change became detectable in fish mapped some twenty days after tectal surgery;
more temporal regions of visual field became represented on the half-tectum as
time passed, until in fish mapped after 100-150 days most of the normal field
was represented there, as shown in Fig. 2(b). No enlargement of the remaining
tectum was seen, grossly or in sagittal sections. The gradual increase in field
95
Interactions controlling optic terminal location
100175
50
o
Ol
25
50
75
100
Days from tectal halving
125
J50
Fig. 3. Graph showing areas measured from the visualfieldmaps of 43fish(e.g. those
shaded in Figs. 2, 4 and 5) as percentages of the field quadrant area, plotted against
time in days from tectal halving. O, Tectum halved only; • , tectum. halved with
optic nerve cut (main series); *, teclum halved with optic nerve cut (ventrolateral
tectal remnant also removed).
representation on the half-tectum, measured from the field charts of different
fish as described above, is shown in Fig. 3, which also shows the extent of
variation between fish. At all stages of compression of the projection on the
half-tectum the maps were retinotopically ordered: a single field position gave
a maximum response at each electrode position and in no case were two discrete
field positions found to give detectable responses at the same tectal point as
sometimes found by Gaze & Sharma (1970) and, after larger but not smaller
tectal lesions, by Horder & Martin (1977). As compression proceeded, new
field became represented at the caudal edge of the half-tectum in an orderly,
sequential fashion, so that the scotoma was gradually eliminated.
(2) Tectal halving with optic nerve cut
The caudal half of the left optic tectum was excised (apart from a small
ventrolateral remnant, as described previously) and the right optic nerve cut
in each of a series of goldfish which were then allowed to recover. Fifteen of
these were mapped after regeneration to assess the timing of compression. The
first responses to visual stimuli were found after 21 days, weaker and more
quickly fatigued than normal but from localized field positions. The first maps
were similar to those from fish with newly halved tecta though much harder to
elicit at this stage, as shown in Fig. 4(a). Measured areas were also similar, as
Fig. 3 shows. Compression soon followed and was essentially finished after
50 days, when most of the normal field was represented on the rostral halftectum. Incomplete compression is represented in Fig. 4(b), and full compression
in Fig. 5 (a). All maps were ordered and were like those obtained by Yoon (1972,
1976) in very similar circumstances and those found late after tectal halving
alone as described above. The estimated sizes of multi-unit fields were similar
throughout to those found at equivalent stages in regeneration without tectal
7-2
96
J. E. COOK
Left tectuin
(b)
Right visual field
Fig. 4. Visual projection maps from two fish with halved left tecta and regenerated
right optic nerves. Conventions as in Fig. 2. (a) 21 days after tectal halving and nerve
cut. (b) 39 days after tectal halving and nerve cut. The shaded areas are represented
by two points in Fig. 3.
Left tectum
Left tectum
(a)
Right visual field
(b)
Right visual field
Fig. 5. Visual projection maps from two fish with halved left tecta and regenerated
right optic nerves. Conventions as in Fig. 2. (a) 62 days after tectal halving and nerve
cut; ventrolateral fragment of caudal tectum not removed (see tectal outline), (b)
105 days after tectal halving and nerve cut; ventrolateral fragment of caudal tectum
also removed. These maps are typical of fish of their respective series mapped 50 or
more days after surgery. The shaded areas are represented by two points in Fig. 3.
Interactions controlling optic terminal location
97
Right visual Held
Fig. 6. Visual projection map from a fish with a halved left tectum and regenerated
right optic nerve, 32 days after nerve cut but only 14 days after delayed tectal halving.
Conventions as in Fig. 2. The shaded area is represented by one point in Fig. 7
(64%).
100 r
75
A*
A
50
25
18 25
50
75
100
Days from optic nerve cut
125
150
Fig. 7. Graph showing areas measured from the visual field maps of 21 fish as percentages of thefieldquadrant area, plotted against time in days from optic nerve cut.
• , Tectum halved with nerve cut on Day 0; A, nerve cut on Day 0, preceding tectal
halving on Day 18 (arrow).
surgery (Horder, 1971). Figure 3 demonstrates the time-course of the whole
process.
In a further five fish the ventrolateral tectal remnant was also excised at the
time of surgery. Figure 5 (b) shows a typical compressed projection in such a fish.
The measured areas, shown in Fig. 3, differ only slightly since the measuring
procedure minimizes the effect of the tectal remnant.
(3) Optic nerve cut preceding tectal halving
The right optic nerves of a series of goldfish were cut and the fish allowed to
recover. Eighteen days later the caudal halves of their left tecta were excised.
Six of these fish were then mapped within 35 days of the optic nerve cut. The
18-day delay was intended to reveal any direct dependence of the course of
compression on the time since tectal halving which might reflect the course of
98
J. E. COOK
Left tectum
Left tectum
(b)
(a)
T
Right visual field
Right visual field
Fig. 8. Visual projection maps from two fish with halved left tecta and regenerated
right optic nerves, following a second right optic nerve cut 85 days after thefirst.Conventions as in Fig. 2. (a) 22 days after second nerve cut; 107 days after tectal halving
and first nerve cut. (b) 68 days after second nerve cut; 153 days after tectal halving
and first nerve cut. These maps are typical offish mapped early and late, respectively,
after a second regeneration. The shaded areas are represented by two points in Fig. 9.
surgically induced changes in cell labels. A delay in compression resulting from
a very short delay in tectal halving would be hard to measure and even harder
to exclude, while too long a delay in halving would permit the cut fibres to
re-establish a projection on the intact tectum first, making a delay in compression
inevitable and of trivial significance.
In the event there was no such delay. Compression in these fish bore the
same relation to the time of the optic nerve cut as in those of the preceding
section in which both cut and halving were performed together. The extent of
compression in the two cases was consequently quite differently related to the
time since halving. Indeed, four substantially compressed projections were
found in the four fish mapped only 13-15 days after tectal halving. One of these
is shown in Fig. 6, and all six areas are plotted in Fig. 7 against the time since
optic nerve cut.
(4) Tectal halving with optic nerve cut: subsequent second nerve cut
The caudal half of the left optic tectum was excised and the right optic nerve
cut in each of a series of goldfish which were allowed to recover and maintained
for 85-97 days, by which time compression had long been essentially complete
in exactly similar fish (taken at random from the same tank after surgery) as
described in section (2) above. After this period their right optic nerves were
again cut. Fourteen fish were subsequently mapped after a second regeneration
of the nerve, to establish whether or not a permanent change in the halved
Interactions controlling optic terminal location
uu
99
•
•
•
75
•
•
•
•
+
•
50
n
i
0
25
i
50
75
100
Days from optic nerve cut
i
125
150
Fig. 9. Graph showing areas measured from the visual field maps of 28 fish as percentages of the field quadrant area, against time in days from the most recent
optic nerve cut. • , Tectum halved with nerve cut on Day 0. +, Tectum halved
with first nerve cut 85-97 days before Day 0; second nerve cut on Day 0.
tectum had accompanied the first compression. The first maps obtained after
this second cut were not compressed but similar to those from fish with newly
halved tecta, though again harder to elicit. Figure 8 (a) shows such a map: not
only the position of the temporal field limit, but also its orientation and shape,
are characteristic of the uncompressed projection. Gradual recompression soon
followed and Fig. 8(6) shows the projection map of a fish completing compression for the second time. The course of the second compression, seen in Fig. 9,
appears similar to that of the first. Scarring and disruption of the nerve after
the first cut may account for varied delays in regeneration after the second,
though fish with obviously swollen, disrupted or scarred nerves were discarded
at the second cut. These results exclude any permanent change in the mechanisms controlling optic fibre growth following tectal halving.
(5) Tectal halving with optic nerve cut: repeated half-nerve crush
The caudal half of the left optic tectum was excised as usual and the right
optic nerve cut in each of a series of goldfish which were allowed to recover.
Further operations on their right optic nerves were then performed after 14-15
days, 25-29 days and 42-44 days, for which the fish were divided into two
groups: in one group the nasal half of the optic nerve, containing fibres predominantly destined for the excised caudal half-tectum, was severed each time;
while in the other group the temporal half, containing fibres predominantly
destined for the rostral half-tectum, was severed. Ten of the first group and six
of the second were later mapped within 15 days of the last operation, to establish whether or not compression depends specifically on the regeneration of
fibres deprived of their usual tectal terminal sites. The repeated operations were
originally intended to prevent fibres in the severed half-nerve from becoming
established in the tectal terminal layer at any time between the initial surgery
and mapping, while permitting the remaining half to regenerate. It now seems
100
J. E. COOK
Left tectum
Left tectum
(.b)
Right visual field
Fig. 10. Visual projection maps from two fish with halved left tecta and partially
regenerated right optic nerves, following repeated lesions of the other part of each
nerve. Conventions as in Fig. 2. (a) 51 days after tectal halving and nerve cut: nasal
half-nerve lesions on Days 14, 29 and 44. (b) 48 days after tectal halving and nerve
cut: temporal half-nerve lesions on Days 15, 28 and 42. The shaded areas are represented by two points in Fig. 11.
that fibres of the severed half-nerve may have reached the tectum briefly between
the operations (Springer & Agranoff, 1977); but the fact that the repeated
interruptions were nevertheless sufficient to prevent compression justifies the
method in practice.
All ten fish mapped within 15 days of the last nasal half-nerve lesion showed
reduced or absent compression, as shown in Figs. 10 (a) and 11. The extent of
compression sometimes differed between the upper and middle parts of the
same field and may have depended on variation in the extent of the half-nerve
lesions. In support of this explanation, unresponsive tectal regions were never
found. A further four fish were mapped later, after complete regeneration also
of the thrice-severed nasal half-nerve. These gave compressed maps, as may be
seen from Fig. 11.
In contrast, all six fish mapped after the last temporal half-nerve lesions gave
maps showing both transposition and varied degrees of partial compression as
illustrated in Fig. 10(b), where temporal field projected to the caudal edge of
the rostral half-tectum as in simple compression, giving chart areas on the same
curve in Fig. 11, while (in contrast to simple compression) varied amounts of
nasal field were unrepresented on the tectum. The effect of the lesion was thus
specific to its site in the nerve, supporting an explanation of compression based
on a competitive interaction between optic fibres.
Interactions controlling optic terminal location
uu
•
•
25
•
•
75
50
* 4
1
0
101
25
i
t
50
75
100
125
Days from nerve cut and tcctal halving
i
150
175
Fig. 11. Graph showing areas measured from the visual field maps of 35 fish as
percentages of thefieldquadrant area, against time in days from nerve cut and tectal
halving. 3 , Tectum halved with nerve cut; nasal half-nerve lesion made in period
under each arrow base. C, Tectum halved with nerve cut; temporal half-nerve lesion
made in period under each arrow base. • , Tectum halved with optic nerve cut only.
DISCUSSION
The findings described in this paper demonstrate the considerable potential
of optic fibre terminals, as detected electrophysiologically, for gradual, extensive
relocation in the tectum. The last section showed how the onset of compression
in a halved tectum can be delayed by temporarily withholding those fibres
deprived by surgery of their usual terminal sites. This implies that the relocation
of fibres in the remaining tectum depends on a competitive influence of the
withheld fibres. The other possibility, that relocation results directly from
modification of specificity labels in the tectum upon halving, would have led in
this case to compression of the regenerated half-nerve projection into a quartertectum, leaving the remaining quarter unoccupied. Further, this possibility also
seems to be disproved by the previous experiment in which the nerve was cut
a second time. Here optic fibres were freshly regenerated into a tectum which
had previously supported a compression projection, and yet they initially
returned to their original, normal terminal locations. This must be taken to
indicate that the previous compression had occurred despite the persistence of
whatever forces guide fibres to their normal terminal sites.
In recent experiments involving the removal of caudal tectum at intervals
after the start of optic nerve regeneration, Yoon (1976) found that compression
was attained later in those cases where tectal halving was most delayed,
apparently suggesting the influence of some change in the tectum itself which
depended on the time since halving. The comparable experiment reported here,
which showed no such effect, differed in one respect: halving was here carried
out after 18 days, before the regenerating fibres were electrically detectable in
the tectum, whereas Yoon found delays in compression only by continuing to
delay tectal halving up to a time (42 days after nerve cut) when, judging by
personal electrophysiological observations on fish at room temperature and the
102
J. E. COOK
behavioural evidence of Springer & Agranoff (1977), fibres may possibly have
been well advanced in their reinnervation of the tectum even though Yoon did
not detect them. Results described in the present paper indicate that fibres of an
uncut nerve innervating a tectum at the time of halving resist compression more
than do freshly regenerating fibres. Yoon's findings might be readily explained
if regenerating fibres had been able to establish functional connexions before a
long-delayed tectal halving, with the consequence that subsequent compression
was both later and slower. In the present experiment, even the rate of compression shows no effect of the 18-day delay in tectal halving; the course of
compression depends only on the regeneration and, presumably, maturation
of the fibre terminals.
Yoon (1976) also mapped both visual projections in each of four fish soon
after bilateral optic tract section close to the tectum: one tectum had just been
halved while the other, halved previously, had already borne a compressed
projection. Compression was achieved later on the newly halved tectum. The
circumstances of the two tecta differed in many ways which might explain this:
for example, fibre regeneration may have been faster in the previously regenerated tract; or, alternatively, remnants of the recently cut fibres in the freshly
divided tectum may have served to stabilize regenerating fibres in their initial
uncompressed state, particularly since the fibres were cut near the tectum and
regained it swiftly. Such an effect was not evident in the present work after the
longer regeneration which follows section of the nerve in the orbit, but would
be consistent with a similar effect seen by Schmidt (1978) in the case of an
expanded projection, and abolished by long-term denervation.
It is remarkable that retinotopic order has been found to be maintained
throughout compression. It must be emphasized that experiments of this type
cannot show directly how this is achieved; however, the evidence from each
of the experiments presented here does indicate that the maintenance of order
should be attributed to the optic fibres themselves. Though it seems that a
variety of factors (including the number and maturity of the fibres present) may
affect the time, rate or extent of compression, the problem is much simplified
if selective interactions occur between the fibres themselves influencing their
behaviour, where they contact each other, according to their relative retinal
origins. Such interactions could, in principle, promote order at any stage in
compression and regardless of the location of the retinal map on the tectum.
It is therefore unnecessary, as well as uneconomical, to assume that all the
various factors influencing compression do so by way of changes (which would
need to be extremely complex) in specificities controlling the exact tectal
location of every optic terminal. Instead, regional specializations in the optic
pathway and tectum need only suffice to orientate the retinal map and could
theoretically comprise as few as two binary distinctions, though more may
exist (Cook & Horder, 1977).
The extensive plasticity which is perhaps the most remarkable feature of
Interactions controlling optic terminal location
103
compression is most economically regarded as a gross exaggeration of a potential for fibre adjustment necessary in normal development. If this is so, then the
establishment of an orderly retinotopic projection in the developing normal fish,
too, may well reflect interactive properties of the fibre array.
The author wishes to express his gratitude to Dr T. J. Horder for continual help, discussion
and advice throughout this work and to the Medical Research Council of Great Britain for
a Research Studentship.
REFERENCES
D. G. & SPERRY, R. W. (1963). Preferential selection of central pathways by
regenerating optic fibers. Expl Neurol. 1, 46-64.
COOK, J. E. & HORDER, T. J. (1974). Interactions between optic fibres in their regeneration
to specific sites in the goldfish tectum. /. Physiol., Lond. 241, 89-90/*.
COOK, J. E. & HORDER, T. J. (1977). The multiple factors determining retinotopic order in
the growth of optic fibres into the optic tectum. Phil. Trans. R. Soc. Lond. B 278, 261-276.
GAZE, R. M. (1959). Regeneration of the optic nerve in Xenopus laevis. Q. Jl exp. Physiol.
44, 290-308.
GAZE, R. M. (1974). Neuronal specificity. Brit. med. Bull. 30, 116-121.
GAZE, R. M. & SHARMA, S. C. (1970). Axial differences in the reinnervation of the goldfish
optic tectum by regenerating optic nerve fibres. Expl Brain Res. 10. 171-181.
HORDER, T. J. (1971). The course of recovery of the retinotectal projection during regeneration of the fish optic nerve. /. Physiol., Lond. Ill, 53-54P.
HORDER, T. J. & MARTIN, K. A. C. (1977). Translocation of optic fibres in the tectum may
be determined by their stability relative to surrounding fibre terminals. /. Physiol., Lond.
Ill, 23-24P.
JACOBSON, M. & GAZE, R. M. (1964). Types of visual response from single units in the optic
tectum and optic nerve of the goldfish. Q. Jl exp. Physiol. 49, 199-209.
JACOBSON, M. & GAZE, R. M. (1965). Selection of appropriate tectal connections by regenerating optic nerve fibres in adult goldfish. Expl Neurol. 13, 418-430.
MATURANA, H. R., LETTVIN, J. Y., MCCULLOCH, W. S. & PITTS, W. H. (1959). Evidence
that cut optic nerve fibers in a frog regenerate to their proper places in the tectum. Science,
N.Y. 130, 1709.
SCHMIDT, J. T. (1978). Retinal fibers alter tectal positional markers during the expansion of
the half retinal projection in goldfish. /. comp. Neurol. Ill, 279-300.
SPERRY, R. W. (1943O). Visuomotor coordination in the newt (Triturus viridescens) after
regeneration of the optic nerve. /. comp. Neurol. 79, 33-55.
SPERRY, R. W. (19436). Effect of 180 degree rotation of the retinal field on visuomotor
coordination. /. exp. Zool. 92, 263-279.
SPERRY, R. W. (1944). Optic nerve regeneration with return of vision in anurans. /. Neurophysiol. 1, 57-69.
SPERRY, R. W. (1948). Patterning of central synapses in regeneration of the optic nerve in
teleosts. Physiol. Zool. 21, 351-361.
SPRINGER, A. D. & AGRANOFF, B. W. (1977). Effect of temperature on rate of goldfish optic
nerve regeneration: a radioautographic and behavioral study. Brain Res. 128, 405-415.
YOON, M. (1972). Reversibility of the reorganization of retinotectal projection in goldfish.
Expl Neurol. 35, 565-577.
YOON, M. G. (1976). Progress of topographic regulation of the visual projection in the halved
optic tectum of adult goldfish. /. Physiol., Lond. 257, 621-643.
ATTARDI,
(Received 1 November 1978, revised 24 November 1978)