axon order in the visual pathway of the quokka wallaby

Axon Order in the Visual Pathwayof the Quokka Wallaby

D.K. CHELVANAYAGAM, S.A. DUNLOP, AND L.D. BEAZLEY*Department of Zoology, University of Western Australia, Nedlands 6907, Australia

ABSTRACTAxon order throughout the visual pathway of the quokka wallaby (Setonix brachyurus)

was determined after localised retinal applications of the tracers DiI and/or DiASP. Postnataldays (P) 22–90 were studied to encompass the development and refinement of retinalprojections. Order was essentially similar at all stages. Axons entered the optic nerve headtrue to their sector of retinal origin. In the optic nerve, nasal and temporal axons continued toreflect their retinal origin, dominating, respectively, the medial and lateral halves. Bycontrast, dorsal and ventral axons exchanged locations between the retrobulbar level andone-third the distance along the nerve; thus, the inversion of the dorsoventral retinal axis,imposed by the lens, was corrected. Decussating axons maintained their relative locationsthrough the chiasm. At the base of the optic tract, nasal and temporal axons underwent anaxial rotation to lie on the medial and lateral sides, respectively; thus nasal overlapped withventral axons and temporal with dorsal axons. Axons maintained their alignments through-out the tract, and as a result, nasal and ventral axons invaded the superior colliculus medially,whereas temporal and dorsal axons invaded laterally. Each retinal quadrant terminatedpreferentially in its retinotopically appropriate sector of the colliculus.

The arrangement of axons in the quokka visual pathway displays several novel features.Axon order is distinct throughout, involving a well-demarcated exchange of dorsal and ventralaxons in the nerve and an axial rotation of nasal and temporal axons at the base of the tract;these relocations suggest decision regions for growing axons. The organisation presumablyunderlies the less extensive searching within the developing superior colliculus to generateretinotopic maps in the quokka and also in tammar wallaby [Marotte, J. Comp Neurol.293:524–539, 1990] than in the rat [Simon and O’Leary, J. Neurosci. 12:1212–1232, 1992].J. Comp. Neurol. 390:333–341, 1998. r 1998 Wiley-Liss, Inc.

Indexing terms: visual projections; macropod; axonal tracing; retinotopicity

It is well established that, in mature vertebrates, theretina projects as a series of topographically ordered mapswithin the primary visual brain centres (Siminoff et al.,1966). However, less is known about order amongst opticaxons en route to the brain. If axons were ordered appropri-ately along the length of the developing visual pathway, socalled ‘pre-ordering’, only limited searching for target cueswould be required to establish retinotopic maps in thebrain (Horder and Martin, 1978). A degradation in theorder would require extensive searching within targettissue (Sperry, 1951; Thanos et al., 1984; Kaethner et al.,1992).

Axons appear to be well ordered along the primaryvisual pathway of the nonmammalian vertebrates (fish:Scholes, 1979; Easter et al., 1981; Stuermer, 1988; Springerand Mednick, 1985; Fraley and Sharma, 1986; amphib-ians: Fujisawa et al., 1981; Scalia and Arango, 1982; Reh etal., 1983; reptiles: Beazley et al., 1997; birds: Thanos andBonhoeffer, 1983). By contrast, a relaxation of order has

been reported along the optic nerve of some mammals,both during development (rat: Simon and O’Leary, 1990,1991, 1992a; Colello and Guillery, 1992; Chan and Guil-lery, 1994) and at maturity (rat: Baker and Jeffery, 1989;cat: Horton et al. 1979; Naito, 1986). Order, albeit modest,is also found within the optic tract (rat: Chan and Guillery,1994; cat: Aebersold et al., 1981; Torrealba et al., 1982;Voigt et al., 1983). By contrast, the primate visual pathwayseems to be much more highly ordered along its length(Hoyt and Luis, 1962; Naito, 1989).

Grant sponsor: National Health & Medical Research Council (Australia),Australian Neuromuscular Research Institute; Grant number: 331710.

Dr. Chelvanayagam’s current address is Psychology Department, McMas-ter University, Hamilton, Ontario, Canada.

*Correspondence to: Professor L.D. Beazley, Department of Zoology,University of Western Australia, Nedlands 6907, Australia.E-mail: [email protected]

Received 17 April 1997; Revised 28 July 1997; Accepted 4 August 1997

THE JOURNAL OF COMPARATIVE NEUROLOGY 390:333–341 (1998)

r 1998 WILEY-LISS, INC.

Although axon order within the visual pathway has beenreported for several species of eutherian mammals, com-monly thought of as the placental mammals, it has yet tobe reported for the other major mammalian radiation,namely to marsupials. Differences might be anticipatedbecause the chiasmal organisations seem to be markedlydifferent between marsupials and eutherians (Guillery,1995). Within the marsupials, there are two major divi-sions, namely the diprotodonts and the polyprotodonts. Inboth the diprotodont quokka (Jeffery and Harman, 1992;Harman and Jeffery, 1992, 1995) and the polyprotodontopossum Monodelphis domestica (Taylor and Guillery,1994), ipsilaterally projecting temporal axons are located

laterally along the length of the optic nerve and retain alateral location through the chiasm. Thus, these axons donot come in contact with contralaterally projecting axonsfrom the other eye until both sets reach the optic tract. Bycontrast, in rodents, ipsilaterally projecting temporal axonsbecome widespread within the optic nerve and approachthe midline at the chiasm, allowing them to interact withaxons from the other eye before returning to the ipsilateralside (Baker and Jeffery, 1989; Jeffery, 1989; Godement etal., 1990; Sretevan, 1990; Marcus et al., 1995).

To address the issue of axon order from each retinalquadrant along the length of the visual pathway of adiprotodont marsupial, we here report a tracing study in

TABLE 1. Summary of Experimental Animal

Age AnimalRetina1

(% Labelled) Tracer R-B Nerve 1/3 Nerve P-C Nerve Chiasm Contra-Tract Ipsi-Tract

22 Q29 Temporodorsal (6%) Di-ASP Dorsal & lateral D/V strip & lateral Ventral & lateral Ventral & caudal Ventral & lateral Lateral22 Q34 Dorsal (10%) Di-I Dorsal D/V strip Ventral Ventral Lateral —22 Q34 Ventronasal (22%) Di-ASP Ventral & medial Lateral Dorsal & lateral Dorsal & rostral Dorsal & medial —23 Q90 Nasal (15%) Di-ASP Medial Medial Medial Rostral Medial Medial (sparse)23 Q90 Temporal (18%) Di-I Lateral Lateral Lateral Caudal Lateral Lateral25 Q20 Nasal (21%) Di-ASP Medial Medial Medial Rostral Medial —26 Q31 Temporal (14%) Di-ASP Lateral Lateral Lateral Caudal Lateral Lateral29 Q21 Dorsal (5%) Di-I Dorsal D/V strip Ventral Ventral Ventral & lateral —29 Q21 Ventral (14%) Di-ASP Ventral Lateral & medial Dorsal Dorsal Dorsal & medial —29 Q95 Dorsal (9%) Di-I Dorsal D/V strip Ventral Ventral Ventral —29 Q95 Ventral (8%) Di-ASP Ventral Lateral & medial Dorsal Dorsal Medial —30 Q97 Dorsal (20%) Di-I Dorsal D/V strip Ventral Ventral Ventral —30 Q97 Ventral (24%) Di-ASP Ventral Lateral & medial Dorsal Dorsal Medial —33 Q56 Nasal (2%) Di-ASP Medial Medial Medial Medial Rostral Medial (sparse)40 Q50 Ventronasal (5%) Di-ASP Ventral & medial Medial Dorsal & medial Rostral Medial Medial (sparse)40 Q50 Nasal (6%) Di-I Medial Medial Medial Dorsal & rostral Dorsal & medial Medial (sparse)45 Q40 Dorsal (5%) Di-ASP Dorsal D/V strip Ventral Ventral Ventral —55 Q39 Temporal (9%) Di-ASP Lateral Lateral Lateral n.a. n.a. n.a.60 Q38 Ventral (3%) Di-ASP Ventral Lateral & medial Dorsal Dorsal Medial —68 Q18 Temporodorsal (15%) Di-ASP Dorsal & lateral D/V strip & lateral Ventral & lateral Ventral & caudal Ventral & lateral Lateral68 Q18 Nasal (10%) Di-I Medial Medial Medial Rostral Medial —71 Q16 Dorsal (12%) Di-ASP Dorsal D/V strip n.a. n.a. n.a. n.a.71 Q16 Dorsonasal (20%) Di-I Dorsal & medial D/V strip & medial Ventral & medial n.a. n.a. n.a.74 Q17 Dorsonasal (11%) Di-I Dorsal & medial D/V strip & medial Ventral & medial n.a. n.a. n.a.90 Q19 Dorsal (10%) Di-I Dorsal D/V strip Ventral n.a. n.a. n.a.90 Q19 Temporal (4%) Di-I Lateral Lateral Lateral n.a. n.a. n.a.

1% labelled indicates the percentage of the retinal area to which tracer was applied, including retina labelled retrogradely peripheral to the tracer implant; in this way, we includedretinal ganglion cells with somata peripheral to the implant whose optic axons were severed when tracer was applied. These areal values suggest that the maximum number of axonslabelled fell in the range of 10,000–40,000. R-B, retrobulbar; 1/3, one-third of the way along the nerve; P-C, prechiasmal, Contra, contralateral; Ipsi, ipsilateral; D/V, dorsoventral; Q,quokka, n.a., not available.

Fig. 1. Histologic preparations viewed by fluorescence (B,C,F-L) andconfocal (A,D,E,M-P) microscopy. Left column (A-G) shows projectionsof dorsal (red) and ventral (green) axons; the right column (H-P) showsprojections of nasal (green) and temporal (red) axons; insets in A andH show diagrammatically the retinal sites of tracer application withD, dorsal; V, ventral; T, temporal; N, nasal. Areas of axon overlapappear yellow. A,H: Whole-mount preparations of optic nerves (retro-bulbar to left, viewed from the side in A, and from below in H). In A,both dorsal and ventral axons are labelled; in H, nasal axons arelabelled. B–D,I,J: Cross-sectioned optic nerves, dorsal being up andmedial to the right, at the retrobulbar level (Bi,Ci,Ii,Ji), one-third ofthe way along the nerve (Bii,Cii,Iii,Jii), prechiasmally (Biii,Ciii,Iiii,Jiii), and as axons enter the chiasm (D). The same sections, butdifferently illuminated, is shown in Bi and Ci, in Bii and Cii, in Biii andCiii, in Ii and Ji, in Iii and Jii, and in Iiii and Jiii. A-D: Dorsal andventral axons exchange locations one-third of the way along the nerveand maintain their territories thereafter. H-J: Nasal and temporalaxons maintain medial and lateral territories, respectively, along thenerve. E: Chiasm sectioned coronally, shown with dorsal up, toillustrate that dorsal axons lie ventrally and ventral axons lie dorsally.K,L: A single section of chiasm sectioned horizontally (its extentindicated by dots), shown with rostral down and illuminated differ-ently to illustrate that nasal axons (K) lie rostrally and temporal axons(L) lie caudally. M: coronal section of caudal chiasm (arrow points tomidline) and base of the contralateral optic tract to illustrate the axialrotation of nasal and temporal axons. N: Base of the ipsilateral (Ni)

and contralateral (Nii) optic tracts seen in cross-section as axons exitthe chiasm, to illustrate that nasal and temporal axons have under-gone an axial rotation to align nasal axons dorsal to temporal ones. F,Gand O,P are whole-brain preparations with the cerebrum removed andseen from the side; rostral is to the left and is dorsal up, the outline ofsuperior colliculus is indicated by dots. F and G are the comparableviews of the same brain but with different illuminations, as are O andP. Ventral and nasal axons share the medial optic tract (F,O); dorsaland temporal axons share the lateral optic tract (G,P). In the superiorcolliculus, ventral retina terminates rostromedially (F), dorsal retinacaudolaterally (G), nasal retina caudomedially (O), and temporalretina rostrolaterally (P). Ages shown are the following: A,D,E,postnatal day (P) 29 (Q95); B,C, P29 (Q21); F,G, P22 (Q34); H, P40(Q50); I-L, P68 (Q18); M-P, P23 (Q90). Sections are 100-µm thick.Micrographs in this Figure and Figures 2 and 4 are produced by usinga Hewlett Packard Scan Jet S4c/T scanner with Adobe Photoshop andAdobe Pagemaker programs. A,D,E,M,N: Green highlight increasedby 90%, midtone red by 100%, and magenta by 19%. B,C,F-L:Brightness decreased by 70, contrast increased by 32; colour balancechanged such that midtone green is increased by 74% and cyan by 50%with green highlights increased by 80%; colourised with a hue of 120(B,F,H,I,K), 0 (C,G,J,L). O,P: Brightness decreased by 35, contrastincreased by 25; O is colourised with a hue of 120, P with a hue of 0.Scale bars 5 500 µm in A,H,K,L, 100 µm in B,C,F,G,O,P, 250 µm inD,E, 200 µm in I,J, 1 mm in M,N.

334 D.K. CHELVANAYAGAM ET AL.

which tracers were applied locally to the quokka retina.Part of this research has been reported in abstract form(Chelvanayagam and Beazley, 1994).

MATERIALS AND METHODS

Animals and anaesthesia

The experiments conformed to the guidelines of Na-tional Health and Medical Research Council (Australia)and the Animal Welfare Committee of the University ofWestern Australia. Animals were collected and held underlicense from Conservation and Land Management (West-ern Australia).

Quokkas were maintained at the University of WesternAustralia, and those between postnatal days (P) 22 and 90were studied, animals being staged by tail and pes (heel tobase of claw) lengths and by body weight (Shield andWooley, 1961). The developmental period selected for studyspanned from when central projections were first estab-lished in visual centres (P20), became maximally exuber-ant (up to P50), and had reached maturity by P90 (Har-man and Beazley, 1986). In this period, the distribution ofganglion cells changed from almost uniform to one exhibit-ing an area centralis in midtemporal retina, embedded in aweak visual streak (Dunlop and Beazley, 1985); a minordeviation of axons around the area centralis into a slightly

Figure 1

OPTIC AXON ORDER IN THE WALLABY 335

arcuate arrangement also develops with age (Dunlop,personal observation). Animals were overdosed by inhala-tion of halothane or by injection of Saffan (30 mg/kg, i.m.)or Nembutal (150 mg/kg, i.p.).

Axonal tracing

In some cases, animals were transcardially perfusedwith saline, followed by paraformaldehyde in 0.1 M, pH7.2, phosphate buffer. On one side, the retina was exposedby removing the cornea, lens, and vitreous and a crystal ofa lipophilic dye (Molecular Probes, Inc., Eugene, OR) DiI orDiASP applied to the nerve fibre layer within the dorsal,ventral, nasal, or temporal quadrants. Tracers were usedeither singly or for double labelling, with each tracerusually being restricted to separate retinal quadrants. Asan alternative labelling procedure with comparable re-sults, animals were decapitated, the heads immediatelywere immersed in culture medium (Dunlop, 1990) anddissected to expose the primary visual pathway. On oneside, the eye was dissected to reveal the retina, which wasbriefly drained of culture medium to allow the applicationof tracers as before. Two to six hours later, the prepara-tions were placed in fixative as above.

Preparations were kept in the dark at 37°C or at roomtemperature for several weeks before retinae were whole-mounted, ganglion cell layer uppermost, whilst wet. Theoptic nerves were dissected, maintaining their orientation,and severed immediately prechiasmally. In favourablepreparations, nerves and brains were photographed intactbefore sectioning. The nerves and brains were then embed-ded separately in 4% agar and Vibratome (Oxford Instru-ments, Oxford, England) sectioned at 100 µm thickness;nerves were sectioned transversely and brains horizon-tally or coronally. Sections were wet mounted from ice-coldphosphate buffer (0.05 M, pH 7.2) onto glass slides,examined directly, and photographed whilst wet or tempo-rarily coverslipped in buffer. Some sections were counter-stained with toluidine blue (0.025% in phosphate buffer at0.05 M) to reveal the underlying cytoarchitecture (Chel-vanayagam and Beazley, 1997).

Material was viewed and photographed by using aLaborlux Leica epifluorescent microscope with E3, N2.1,and L3 filters as appropriate. For dual labelling, an L3filter was used to discriminate DiASP, rejecting even astrong DiI signal. In addition, tissue was viewed in theconfocal microscope (Bio-Rad MRC 1000 with wavelengthsof 488 and 514 nm, A1 and A2 blocks, Kalman filtering;Bio-Rad, Richmond, CA).

We analysed the locations of labelled axons along theentire length of the visual pathway and of their terminalfields in the superior colliculus. Difficulty in defining theborders of the lateral geniculate nucleus at younger stagesprecluded an analysis of the topography of the projectionswithin this nucleus.

The extent of retinal area labelled by a tracer and thepercentage of the cross-sectional areas of the nerve occu-pied by labelled axons were estimated from photographsby using a Jandel graphics tablet (Corte Madera, CA) andSigma-plot software. Two independent observers checkedthe location of the border between labelled and unlabelledtissue. We assessed whether the territory occupied bylabelled axons increased or decreased along the nerve. Todo so, we compared values for the percentage of nerve areaoccupied by labelled axons at the retrobulbar level withthose values midway along the nerve and prechiasmally.

The percentage of nerve area labelled at the retrobulbarlevel was standardised to an ‘occupancy value’ of 1. Forexample, if the percentage of nerve area labelled at themidway or prechiasmal levels was the same as that at theretrobulbar level, the occupancy values along the nervewould be 1:1:1. Occupancy values greater than 1 at themidway or prechiasmal levels would indicate that labelledaxons had spread out; values less than 1 would indicatethat axons had become more tightly packed. The standardi-sation allowed for minor changes in cross-sectional areaalong the length of the nerve and thus enabled comparisonof values at the three levels. Moreover, allowance could bemade for differences in the extent of retinal labellingbetween individuals.

RESULTS

We analysed projections from 26 tracer placements.Eight quokkas had a single tracer and nine had dualplacements of tracers. Tracers were restricted to between 5and 22% of retinal area. The findings for all stages aredescribed together because the results were essentiallysimilar, except that the larger size of the older animalsresulted in less robust labelling of the tract (Table 1).

Retina and optic nerve

Labelled axons projected to the corresponding region ofthe optic nerve head, travelling within the reticular net-work of axonal fascicles as described in chick (Nakamuraand O’Leary, 1989) and rat (Simon and O’Leary, 1991).

Considerable order was seen along the length of the opticnerve, with characteristic axon distributions being associ-ated with each retinal quadrant (Fig. 1). The organisationwas reflected in the arrangement of glial septa, a featurethat was particularly pronounced behind the eye (Fig. 2)and became more marked with age. Nevertheless in thequokka, as in zebra fish (Bodick and Levinthal, 1980) andmonkey (Williams and Rakic, 1985), axons did not maintaintheir original neighbour relations nerve, and order becamesomewhat less stringent nearer the brain. The relaxationof order was indicated by the tendency for the ‘occupancyvalues’of axons from each retinal quadrant to rise, particu-larly between the midnerve and prechiasmally (Fig. 3).

Nasal and temporal axons maintained medial and lat-eral domains, respectively, along the length of the opticnerve (Fig. 1H-J). By contrast, dorsal and ventral axonsunderwent a major rearrangement. Dorsal axons (Fig.1A-C) were located dorsally in the retrobulbar nerve.Between this level and one-third of the way along thenerve’s length, axons dived through the nerve core tooccupy the ventral region thereafter. Ventral axons (Fig.1A,B) entered the nerve at its ventral aspect as twogroups. The groups straddled a septum, containing theophthalmic blood vessels and located slightly temporal tothe ventral pole; we assume that this structure is theremnant of the optic fissure. The two groups of ventralaxons relocated via the periphery of the nerve, one groupmoving clockwise and the other counterclockwise. Byone-third of the distance between the eye and the brain,the two groups lay at the medial and lateral extremes ofthe nerve; they continued to ascend to reunite dorsally.

Chiasm, optic tract, and superior colliculus

The locations of axons in the prechiasmal optic nervewere retained through the chiasm. In other words, nasal


axons were located rostrally in the chiasm (Fig. 1K) andtemporal axons caudally (Fig. 1L); dorsal and ventralaxons favoured ventral and dorsal chiasm, respectively(Fig. 1D,E).

As axons exited the chiasm and entered the base of thecontralateral optic tract, the nasal and temporal popula-tions underwent an axial rotation to change their align-ment from rostrocaudal to mediolateral (Fig. 1M). As aresult, axons from rostral chiasm came to lie medially inthe tract, those from caudal chiasm laterally (Fig. 1Nii).By contrast, dorsal and ventral axons were appropriatelyplaced in the chiasm to allow them to retain their relativepositions and enter the optic tract with dorsal axons lyinglaterally and ventral axons medially (Fig. 1F,G). As aresult of the rotation of the nasotemporal axis but the

retention of the dorsoventral one, nasal axons came tooccupy the rostromedial tract, in territory which over-lapped that of ventral axons (Fig. 1F,O). Similarly, tempo-ral axons came to lie in the caudolateral tract, with theirterritory overlapping that of dorsal axons (Fig. 1G,P). Thearrangement ensured that axons approached the superiorcolliculus appropriately aligned into retinal halves to forma map, with dorsal axons accompanying temporal ones andventral axons accompanying nasal ones (see Fig. 1F,G,O,P).

Once within the superior colliculus, most axons wereseen to terminate only in the sector that was topographi-cally appropriate for them (Fig. 1F,G,O,P). In other words,ventral axons terminated rostromedially, nasal axons cau-domedially, temporal axons rostrolaterally, and dorsalaxons caudolaterally. The restriction of the majority ofaxons to the appropriate quarter of the colliculus wasapparent from the earliest stages.

A minority of temporal axons, having entered the chiasmcaudally, did not decussate with their fellows but peeled offinto the ipsilateral optic tract (Fig. 4B), as describedpreviously (Jeffery and Harman, 1992; Harman and Jef-fery, 1992, 1995). Once within the optic tract, the un-crossed temporal axons relocated ventrolaterally, a ma-noeuvre that presumably ensured their proximity to thecrossed temporal axons from the other eye (Fig. 4B).

At early stages, two aberrant projections were present.One, seen at P23–40, was a sparse nasal projection in theipsilateral optic tract, located immediately medial to theipsilaterally projecting temporal axons (Figs. 1N, 4B;Table 1). No similar projection was present at P68, a resultcompatible with the projection being removed by celldeath, as described for rat (Jeffery, 1989). The otheraberrant projection, most prominent up to P26, was aretinoretinal projection that extended along the oppositeoptic nerve and invaded the opposite retina, matchingprevious descriptions (quokka: Beazley et al., 1995; rat:Bunt et al., 1983).

DISCUSSION

We here report that axon order is maintained along thelength of the visual pathway in the quokka and does notchange throughout development. There is a clearly demar-cated exchange in the location of dorsal and ventral axonswithin the optic nerve and an axial rotation at the base ofthe optic tract for nasal and temporal axons (summarizedin Fig. 5).

Axon order in the visual pathway:interspecies comparison

The literature points to a more stringent axon orderwithin the visual pathway of fish, amphibia, and reptilesthan that of birds or the eutherian mammals, other thanprimates (Bunt and Horder, 1982). The difference is pres-ent even when one allows for the range of techniques used.It has been suggested that order is less apparent whensmall numbers of axons are traced (Walsh, 1986); electro-physiologic recording represents an extreme example ofthis principle, allowing very few axons to be located(Horton et al., 1979).

In all vertebrates studied, axons approach the opticnerve head true to their sectors of origin. However, thereare considerable differences in the subsequent extent oforder and arrangement of axons from different retinalquadrants. These differences are probably best addressed

Fig. 2. Cross-section of the optic nerve, viewed by fluorescencemicroscopy, one-quarter of the way along its length at P60 (Q38) withdorsal up and medial to the right, showing the arrangement of glialsepta (outlined in inset). The section is morphologically similar to thatshown in Figure 1Jii. Labelled ventral axons lie ventrolaterally andventromedially. Toluidine blue; sections are 100-µm thick. Brightnessdecreased by 20, contrast increased by 15. Mode changed to black/white. Scale bar 5 350 µm.

Fig. 3. Relative ‘occupancy values’ in the optic nerve at the retrobul-bar, midnerve, and prechiasmal (chiasm) levels.


by considering the nasal and temporal axons separatelyfrom the dorsal and ventral ones.

Nasal and temporal axons enter the retrobulbar nervetrue to their retinal origins in most nonmammals studied(cichlid fish: Scholes, 1979; goldfish: Easter et al., 1981;Springer and Mednick, 1985; Fraley and Sharma, 1986;newt: Fujisawa et al., 1981; Anura: Scalia and Fite, 1974;lizard: Beazley et al., 1997). As an exception, in the frogRana pipiens (Scalia and Arango, 1982; Reh et al., 1983),some axons from both nasal and temporal retina cross overto the opposite side of the optic nerve head. Nasal andtemporal axons maintain their territories subsequentlyalong the length of the nerve. At or beyond the chiasm, inboth fish (Springer and Mednick, 1985) and amphibia(Fujisawa et al., 1981; Reh et al., 1983), there is a complexreorganisation such that nasal axons become split into twogroups that occupy the medial and lateral flanks of thetract. The arrangement allows both the nasal and tempo-

ral axons to enter both optic brachia. In chick, nasal andtemporal axons are restricted within their appropriatehalf of the optic nerve but overlap in the rostral part of thetract (Ehrlich and Mark, 1984).

In all mammals studied, including the quokka, thequadrantic origin of retinal axons is retained with nasaland temporal axons occupying the medial and lateralsides, respectively, of the retrobulbar optic nerve (rat:Collelo and Guillery, 1992; Chan and Guillery, 1994;hamster: Baker and Jeffery, 1989; cat: Naito, 1986; mon-key: Hoyt and Luis, 1962; Naito, 1989). However, theextent of subsequent order varies markedly between spe-cies.

Quokka, along with the primates (Hoyt and Luis, 1962;Naito, 1989), exhibits considerable axon order along thelength of the visual pathway. The orderliness of nasal andtemporal axons along the nerve and through the chiasm inquokka ensures that these axons remain segregated asthey enter the contralateral optic tract. The rotation of thenasotemporal axis aligns nasal axons within the medialtract and temporal axons within the lateral tract; thesubsequent maintenance of axon order allows nasal andtemporal axons to enter the superior colliculus appropri-ately, at its medial and lateral aspects, respectively. Inprimates also, nasal and temporal axons undergo an axialrotation in the base of the tract and remain well demar-cated thereafter.

Order is apparent along the length of the cat optic nervewith nasal and temporal axons occupying their respectivehemi-nerves (Naito, 1986); the arrangement of axons inthe chiasm awaits anatomical investigation. It may bethat, as in quokka, order is maintained and an axial shiftbrings about the modest order seen in the tract (Aebersoldet al., 1981; Torrealba et al., 1982; Voigt et al., 1983).

Fig. 4. Whole-mount brain, seen from below and viewed by fluores-cence microscopy, showing the chiasm and the base of the ipsilateraloptic tract, with rostral being down. Illuminated reveals temporalaxons in (A) and nasal axons in (B) within the same section.Ipsilaterally projecting temporal axons (A) undergo an axial rotationto relocate them to the lateral optic tract. Aberrant ipsilaterallyprojecting nasal axons are seen in B. Sections are 350-µm thick.Brightness decreased by 10, contrast increased by 20. Mode changedto black/white. Scale bar 5 500 µm.

Fig. 5. Diagrammatic representation of optic axon order along thecrossed visual pathway of the quokka from immediately behind theeye to the superior colliculus. D, dorsal; V, ventral; T, temporal; N,nasal.


Axon order within the visual pathway is weakest in therodents. The retrobulbar segregation of nasal from tempo-ral axons degrades along the length of the nerve and iscompletely absent from the chiasm (Chan and Guillery,1994). The finding accords with the observation thatipsilaterally projecting temporal axons are widespread inthe chiasm (Baker and Jeffery, 1989). A new but impover-ished order is set in place at the base of the tract (Colelloand Guillery, 1992; Chan and Guillery, 1994). However, theorder must be even further degraded along the tractbecause, for example, temporal axons enter the superiorcolliculus across its width (Simon and O’Leary, 1991).

As with the nasal and temporal axons, the extent towhich dorsal and ventral axons retain separate territoriesalong the visual pathway varies across the vertebrateclasses with considerable order in fish, amphibia, andreptiles (cichlid fish: Scholes, 1979; goldfish: Easter et al.,1981; Springer and Mednick, 1985; Fraley and Sharma,1986; newt: Anura: Scalia and Fite, 1974; Fujisawa et al.,1981; lizard: Beazley et al., 1997). Lesser order is seen inbirds (Thanos and Bonhoeffer, 1983) and those eutherianmammals studied other than primates (rat: Silver andSapiro, 1981; Simon and O’Leary, 1991; Collelo and Guil-lery, 1992; Chan and Guillery, 1994; ferret: Guillery andWalsh, 1987; cat: Aebersold et al., 1981; Torrealba et al.,1982; Naito, 1986; Voigt et al., 1983 ).

However, irrespective of the stringency of axon order,one feature is common, namely the exchange in thelocation of dorsal and ventral axons (summarised in Fig.6). Amongst the nonmammalian classes, the exchangetakes place prechiasmally in frog (Reh et al., 1983), hasoccurred before axons entering the base of the tract innewt (Fujisawa et al., 1981), and within the optic tract offish (Scholes, 1979; Fraley and Sharma, 1986) and lizards(Beazley et al., 1997). The representation of the dorsoven-tral retinal axis appears to be already inverted in the optictract of the chick, although the location of the exchangehas yet to be defined (Thanos and Bonhoeffer, 1983;Ehrlich and Mark, 1984).

Amongst the mammals, the quokka demonstrates anunusually clear translocation. It takes place at a novel

location, namely approximately in the one-third of thenerve behind the eye and is maintained subsequentlythroughout the visual pathway. The exchange seems totake place midway along the optic nerve in cat (Naito,1986) and is maintained in the tract (Aebersold et al.,1981; Torrealba et al., 1982; Voigt et al., 1983). Dorsal andventral axons exchange locations as they exit the chiasmin the ferret (Guillery and Walsh, 1987) and at the base ofthe optic tract in primates (Naito, 1989; Hoyt and Luis,1962). An exchange of dorsal and ventral axons is also seenin rat but it extends along the length of the optic nerve(Silver and Sapiro, 1981; Simon and O’Leary, 1991, 1992a;Chan and Guillery, 1994). However, axon order is lostcompletely in the chiasm and a new weak order emergespostchiasmally, tending to re-establish a separation ofdorsal from ventral axons (Chan and Guillery, 1994).

An exchange in the locations of dorsal and ventral axonswithin the visual pathway corrects for the inversion in therepresentation of the visual world, brought about by thelens. A maintenance of order along the remainder of thevisual pathway allows axons to enter the superior collicu-lus aligned appropriately to form the dorsoventral axis ofthe retinal projection (Siminoff et al., 1966); in turn, thealignment of the visual maps matches those of othersensory modalities such as audition (guinea pig: Withing-ton et al., 1990).

The procedures adopted here precluded an examinationof chronotopy in the visual pathway. This feature ispronounced in fish (Easter et al., 1981; Springer andMednick, 1985; Fraley and Sharma, 1986), reflecting theirlife-long centroperipheral generation of ganglion cells(Johns, 1977). In mammals, chronotopy (ferret: Walsh andGuillery, 1985; Walsh, 1986) probably reflects the genera-tion of different ganglion cell classes in chronologicallydistinct waves (cat: Walsh and Polley, 1985) and explainsthe characteristic depth profiles of axons in the optic tract(opossum: Cavalcante et al., 1978; rat: Reese, 1987; ferret:Baker, 1990; cat: Guillery et al., 1982; monkey: Reese andGuillery, 1987; Reese and Cowey, 1990).

Axonal navigation

Studies of axon order in the visual pathway provideinsights into axonal navigation. Growing axons probablyuse a series of guidance cues, which may attract or repelthem (Karlstrom et al., 1997). Some cues may be in theform of gradients, whereas others might be strategicallyplaced to influence axons within a restricted region. In thepast few years, many genes involved in axon pathfindingin invertebrates have been identified, some of whichproduce proteins that directly guide growth cones (Ishii etal., 1992).

More recently, genetic screening in zebrafish has identi-fied genes involved in pathfinding within the visual path-way. We presume that, similarly, genetically programmedcytochemical cues underlie the exchange of dorsal andventral axons within the quokka optic nerve and at otherkey sites along the visual pathway in other vertebrates. Asearch for putative agents that guide axons in the verte-brate visual pathway might profitably be directed at suchdecision regions. However, it is likely that cues at any sitewill be read only by particular subpopulations of axons.For example, the relocation of ventral axons from theventral to the dorsal part in the quokka optic nerve mustbe driven by cues read only by ventral axons. The nasaland temporal axons, through whose territory the ventral

Fig. 6. Diagrammatic representation of the point along the opticpathway at which axons from dorsal and ventral retina exchangelocations. Broken lines indicate a loss of order. References are asfollows: 1–3, Easter et al., 1981; Springer and Mednick, 1985; Fraleyand Sharma, 1986; 4, Scholes, 1979; 5, Reh et al., 1983; 6, Chan andGuillery, 1994; 7, Reese and Baker: 1993; 8, Naito, 1986; 9,10, Hoytand Luis, 1962; Naito, 1989.


axons grow, must ignore the cues because they do notdeviate from their trajectory along the length of the nerve.

Map formation

The signals for map formation are likely to be distinctfrom those for axonal navigation because, in other sys-tems, each can function independently of the other. As anexample, in the zebrafish mutant the optic pathway isnormally ordered but the retinotectal map is disturbed(Karlstrom et al., 1997). Moreover, in X. laevis, axonsforced to enter visual centres from abnormal directions,and as a result deprived of the normal pathfinder cues, canstill form an organised projection (Beazley and Lamb,1979). Regenerating optic axons in the lizard illustrate thereverse situation. Axons must still be able to read naviga-tional cues because they follow the visual pathway faith-fully, but cues for map restoration must be absent (orremain unread) because at 1 year after lesion the regener-ated projection lacks topographic order (Beazley et al.,1997).

The order displayed within the quokka visual pathwayallows temporal along with dorsal axons and nasal alongwith ventral axons to approach the visual centres asdemarcated units. The arrangement ensures that axonsare located appropriately into the correct side of the optictract of the dorsal/temporal and ventral/nasal hemi-retinae. However, as axons enter the target tissue, signalsfor map formation (Walter et al., 1987; Dreschner et al.,1995; Logan et al., 1996) must become paramount toensure that temporal and ventral axons innervate rostralcolliculus whilst dorsal and nasal axons continue morecaudally. The process of selecting terminal fields mayinvolve differential adhesion between axons from differentretinal quadrants and cells in visual centres (Boxberg etal., 1993) because temporal axons have been shown invitro to adhere to tectal cells more strongly than do nasalaxons (chick: Walter et al., 1987; rat: Simon and O’Leary,1992b). The result could explain the mechanism by whichthe nasotemporal axis of the retinotectal projection isestablished in vivo. Temporal axons terminate rostrallyand nasal axons caudally although both populations enterthe tectum rostrally. The greater adhesion between tempo-ral axons and tectal cells would ensure that these axonspreferentially innervate rostral tectum, leaving the lessadhesive nasal axons to continue their growth into caudaltectum.

The degree of order amongst ingrowing optic axons willdetermine the extent to which ingrowing axons need tosearch within visual centres such as the superior colliculusbefore refining their connections (Debski et al., 1990;Schmidt, 1993) onto the appropriate partner cell(s) andestablishing a mature organised visual projection. Thestricter the retinotopic order of the axons, the lesser theterritory in which they must search (Horder and Martin,1978). As we have argued above, axon order in the quokkavisual pathway exceeds that found in rat (Simon andO’Leary, 1992a; Chan and Guillery, 1994). In line with thisdifference, we report here that the search by ingrowingaxons for postsynaptic partner cells in the quokka superiorcolliculus is largely limited to one quadrant, as previouslydescribed for the tammar wallaby (Marotte, 1990, 1993).By contrast, in embryonic rat, optic axons are known tosearch widely for partner collicular cells (Simon andO’Leary, 1991). The rat is therefore probably an excellentmammalian model in which to search for cues for mapformation. Conversely, the quokka, although not the ani-

mal of choice for studying cues for map formation, offersopportunities to investigate decision points for growingaxons.

ACKNOWLEDGMENTS

Dr. Lisa Tee prepared some material. Wendy Rossprepared the figures. Bob McNeice, Noreen Underwood,and Peter Cowl maintained the marsupial facility. D.K.C.held a studentship of the Australian Neuromuscular Re-search Institute.

LITERATURE CITEDAebersold, H., O.D. Creutzfeldt, U. Kuhnt, and D. Sanides (1981) Represen-

tation of the visual field in the optic tract and optic chiasma of the cat.Exp. Brain Res. 42:127–145.

Baker, G.E. (1990) Prechiasmatic reordering of fibre diameter classes in theretinofugal pathway of ferrets. Eur. J. Neurosci. 1:224–233.

Baker, G.E., and G. Jeffery (1989) Distribution of uncrossed axons along thecourse of the optic nerve and chiasm of rodents. J. Comp. Neurol.289:455–461.

Beazley, L.D., and A.H. Lamb (1979) Rerouted optic axons in Xenopustadpoles form normal visuotectal projections. Brain Res. 179:373–378.

Beazley, L.D., S.A. Dunlop, D.K. Chelvanayagam, and W.M. Ross (1995)Wiring up the visual system. Clin. Exp. Pharmacol. Physiol. 22:550–558.

Beazley, L.D., P.W. Sheard, M. Tennant, D. Starac, and S.A. Dunlop (1997)The optic nerve regenerates but does not restore topographic projec-tions in the lizard Ctenophorus ornatus. J. Comp. Neurol.377:105–120.

Bodick, N., and C. Levinthal (1980) Growing optic nerve fibres followneighbours during embryogenesis. Proc. Natl. Acad. Sci. USA 77:4374–4378.

Boxberg, Y., S. Deiss, and U. Schwarz (1993) Guidance and topographicstabilisation of nasal chick retinal axons on target-derived componentsin vitro. Neuron 10:345–357.

Bunt, S.M., and T.J. Horder (1982) Evidence of an orderly arrangement ofoptic axons within the optic nerves of the major non-mammalianvertebrate classes. J. Comp. Neurol. 213:94–114.

Bunt, S.M., R.D. Lund, and P.W. Land (1983) Prenatal development of theoptic projection in albino and hooded rats. Dev. Brain Res. 6:149–168.

Cavalcante, L.A., S. Allodi, and C.E. Rocha-Miranda (1978) Fiber order inthe opossum’s optic tract. Anat. Embryol. 186:589–600.

Chan, S.O., and R.W. Guillery (1994) Changes in fibre order in the opticnerve and tract of rat embryos. J. Comp. Neurol. 344:20–32.

Chelvanayagam, D.K., and L.D. Beazley (1994) Optic axons are orderedfrom early development in the visual pathway of the quokka Setonixbrachyurus. Proc. Aust. Neurosci. Soc. 5:121.

Chelvanayagam, D.K., and L.D. Beazley (1997) Toluidine blue-O, a bright-field Nissl stain compatible with the lipophilic fluorescent tracers. J.Neurosci. Methods 75:49–56.

Colello, R.J., and R.W. Guillery (1992) Observations on the early develop-ment of the optic nerve and tract in mouse. J. Comp. Neurol.317:357–378.

Debski, E.A., H.T. Cline, and M. Constantine-Paton (1990) Activity-dependent tuning and the NMDA receptor. J. Neurobiol. 1:18–32.

Dreschner, U., C. Kremoser, J. Loschinger, M. Noda, and F. Bonhoeffer(1995) In vitro guidance of retinal ganglion cell axons by RAGS, a25kDa tectal protein related to ligands for Eph receptor tyrokinases.Cell 82:359–370.

Dunlop, S.A. (1990) Early development of retinal ganglion cell dendrites inthe marsupial Setonix brachyurus, quokka. J. Comp. Neurol. 293:425–447.

Dunlop, S.A., and L.D. Beazley (1985) Changing distribution of retinalganglion cells during area centralis and visual streak formation in themarsupial Setonix brachyurus. Dev. Brain Res. 23:81–90.

Easter, S.S., A. Russoff, and P.E. Kish (1981) The growth and organisationof the optic nerve and tract in juvenile and adult goldfish. J. Neurosci.1:793–811.

Ehrlich, D, and R.F. Mark (1984) The course of axons of retinal ganglioncells within the optic nerve and tract of the chick (Gallus gallus). J.Comp. Neurol. 223:583–591.

Fraley, S.M., and S.C. Sharma (1986) Retinal topography in the optic tractof adult goldfish. Neuroscience 19:1363–1380.


Fujisawa, H., K. Watanabe, N. Tani, and Y. Ibata (1981) Retinotopicanalysis of fibre pathways in amphibians: I. The adult newt Cynopspyrrhogaster. Brain Res. 206:9–20.

Godement, P.J., J. Salatin, and C.A. Mason (1990) Retinal axon pathfindingin the optic chiasm: Divergence of crossed and uncrossed fibers. Neuron5:173–186.

Guillery, R.W. (1995) A comparison of the eutherian and marsupial opticchiasms: A brief review. Rev. Bras. Biol.1:1–10.

Guillery, R.W., E.H. Polley, and F. Torrealba (1982) The arrangement ofaxons according to fibre diameter in the optic tract of the cat. J.Neurosci. 2:714–721.

Guillery, R.W., and C. Walsh (1987) Changing glial organisation relates tochanging tibre order in the developing optic nerve of ferrets. J. Comp.Neurol. 265:203–217.

Harman, A.M., and L.D. Beazley (1986) Development of visual projectionsin the marsupial, Setonix brachyurus. Anat. Embryol. 175:181–188.

Harman, A.M., and G. Jeffery (1992) Distinctive pattern of organisation inthe retinofugal pathway of a marsupial: I. Retina and optic nerve. J.Comp. Neurol. 323:1–10.

Harman, A.M., and G. Jeffery (1995) Development of the chiasm of amarsupial, the quokka wallaby. J. Comp. Neurol. 359:507–521.

Horder, T.J., and K.A.C. Martin (1978) Morphogenetics as an alternative tochemospecificity in the formation of nerve connections. In Curtis, A.S.G.(ed): Cell-Cell Recognition, Soc. Exp. Biol. Symp. Cambridge: Cam-bridge University Press, pp. 275–358.

Horton, J.C., M.M. Greenwood, and D.H. Hubel (1979) Non-retinotopicarrangement of fibres in the cat optic nerve. Nature 282:720–722.

Hoyt, W.E., and O. Luis (1962) Visual fiber anatomy in the infra-geniculatepathway of the primate: Uncrossed and crossed retinal quadrant fiberprojections studied with Nauta stain. Arch. Ophthalmol. 68:94–106.

Ishii, N., W.G. Wadsworth, B.D. Stern, J.G. Culotti, and E.M. Hedgecock(1992) UNC-6, a laminin-related protein, guides cell and pioneer axonmigrations in C. elegans. Neuron 9:873–881.

Jeffery, G. (1989) Distribution and trajectory of uncrossed axons in the opticnerves of pigmented and albino rats. J. Comp. Neurol. 289:462–466.

Jeffery, G., and A.M. Harman (1992) The distinctive pattern of organisationin the retinofugal pathway of a marsupial: II. Optic chiasm. J. Comp.Neurol. 323:1–11.

Johns, P.R. (1977) Growth of the adult goldfish eye: III. The source of newretinal cells. J. Comp. Neurol. 176:343–358.

Kaethner, R.J., and C.A.O. Stuermer (1992) Dynamics of terminal arborformation and target approach of retinotectal axons in living zebrafishembryos: A time-lapse study of single axons. J. Neurosci. 12:3257–3271.

Karlstrom, R.O., T. Trowe, and F. Bonhoeffer (1997) Genetic analysis ofaxon guidance and mapping in the zebrafish. Trends Neurosci. 20:3–8.

Logan, C., A. Wizenmann, U. Dreschner, M. Monschau, F. Bonhoeffer, andA. Lumsden (1996) Rostral tectum acquires caudal characteristicsfollowing ectopic engrailed expression. Curr. Biol. 6:1006–1014.

Marcus, R.C., R. Blazeski, and C.A. Mason (1995) Retinal axon divergencewithin the optic chiasm: Uncrossed axons diverge from crossed fiberswithin a midline glial specialisation. J. Neurosci. 15:3716–3729.

Marotte, L.R. (1990) Development of retinotopy in projections from the eyeto the dorsal lateral geniculate nucleus and superior colliculus of thewallaby (Macropus eugenii). J. Comp. Neurol. 293:524–539.

Marotte, L.R. (1993) Location of retinal ganglion cells contributing to theearly impression in the retinotopic order of the developing projection tothe superior colliculus of the wallaby (Macropus eugenii). J. Comp.Neurol. 331:1–13.

Naito, J. (1986) Course of retinogeniculate projection fibres in the cat opticnerve. J. Comp. Neurol. 251:376–387.

Naito, J. (1989) Retinogeniculate projection fibres in the monkey opticnerve: A demonstration of fibre pathways by retrograde axonal trans-port of WGA-HRP. J. Comp. Neurol. 284:174–186.

Nakamura, H., and D.D.M. O’Leary (1989) Inaccuracies in initial growthand arborization of chick retinotectal axons followed by course correc-tions and axon remodelling to develop topographic order. J. Neurosci.9:3776–3795.

Reese, B.E. (1987) The distribution of axons according to diameter in theoptic nerve and optic tract of the rat. Neuroscience 22:1015–1024.

Reese, B.E., and G.E. Baker (1993) The re-establishment of the dorsoven-tral retinal axis in the chiasmatic region of the ferret. Vis. Neurosci.10:957–968.

Reese, B.E., and A. Cowey (1990) Fibre organisation of the monkey’s optictract: I. Segregation of functionally distinct optic axons. J. Comp.Neurol. 295:385–400.

Reese, B.E., and R.W. Guillery (1987) Distribution of axons according todiameter in the monkey’s optic tract. J. Comp. Neurol. 260:453–460.

Reh, T.A., E. Pitts, and M. Constantine-Paton (1983) The organisation offibres in the optic nerve of normal and tectum-less Rana pipiens. J.Comp. Neurol. 218:282–296.

Scalia, F., and V. Arango (1982) The anti-retinotopic organisation of centralconnections in the frog’s optic nerve. Brain Res. 266:121–126.

Scalia, F., and K.V. Fite (1974) A retinotopic analysis of the centralconnections of the optic nerve in the frog. J. Comp. Neurol. 158:455–478.

Schmidt, J.T. (1993) Activity-driven mechanisms for sharpening retinotopicprojections: Correlated activity, NMDA receptors, calcium entry, andbeyond. In S.C. Sharma and J.W. Fawcett (eds): Formation andRegeneration of Nerve Connections. Boston: Birkhauser, pp. 185–204.

Scholes, J.H. (1979) Nerve fibre topography in the retinal projection to thetectum. Nature 278:620–624.

Shield, J.W., and P. Wooley (1961) Age estimation by measurement of pouchyoung in the quokka, Setonix brachyurus. Aust. J. Zool. 9:14–23.

Silver, J., and J. Sapiro (1981) J. Axonal guidance during development ofthe optic nerve: The role of pigmented epithelia and other extrinsicfactors. J. Comp. Neurol. 202:521–538.

Siminoff, R., H.O. Schwassmann, and L. Kruger (1966)An electrophysiologi-cal study of the visual projection to the superior colliculus of the rat. J.Comp. Neurol. 127:435–444.

Simon, D.K., and D.D.M. O’Leary (1990) Limited topographic specificity inthe targeting and branching of mammalian retinal cells. Dev. Biol.137:125–134.

Simon, D.K., and D.D.M. O’Leary (1991) Relationship of retinotopic order-ing of axons in the optic pathway to the formation of visual maps incentral targets. J. Comp. Neurol. 307:393–404.

Simon, D.K., and D.D.M. O’Leary (1992a) Development of topographicorder in the mammalian retinocollicular projection. J. Neurosci. 12:1212–1232.

Simon, D.K., and D.D.M. O’Leary (1992b) Responses of retinal ganglioncells in vivo and in vitro to position-encoding molecules in the embry-onic superior colliculus. Neuron 9:977–989.

Sperry, R.W. (1951) Mechanisms of neural maturation. In S.S. Stevens (ed):Handbook of Experimental Psychology. New York: Wiley, pp. 236–280.

Springer, A.D., and A.S. Mednick (1985) Topography of the goldfish optictracts: Implications for the chronological clustering model. J. Comp.Neurol. 239:108–116.

Sretevan, D.W. (1990) Specific routing of retinal ganglion cell axons at themammalian optic chiasm during embryonic development. J. Neurosci.10:1995–2007.

Stuermer, C.A.O. (1988) Retinotopic organisation of the developing retino-tectal projection in the zebrafish embryo. J. Neurosci. 8:4513–4530.

Taylor, J.S.H., and R.W. Guillery (1994) Early development of the opticchiasm of the gray short-tailed opossum Monodelphis domestica. J.Comp. Neurol. 350:109–121.

Thanos, S., and F. Bonhoeffer (1983) Investigations on development andtopographic order in retinotectal axons: Anterograde and retrogradestaining of axons and their perikarya with rhodamine in vivo. J. Comp.Neurol. 219:420–430.

Thanos, S., F. Bonhoeffer, and U. Rutishauser (1984) Fiber-fiber interactionand tectal cues influence the development of the chick retinotectalsystem. Proc. Natl. Acad. Sci. USA. 78:2145–2149.

Torrealba, F., R.W. Guillery, U. Eysel, and E.H. Polley (1982) Studies ofretinal representations within the cat’s optic tract. J. Comp. Neurol.211:377–396.

Voigt, T., J. Naito, and H. Wassle (1983) Retinotopic scatter of optic tractfibres in the cat. Exp. Brain Res. 52:25–33.

Walsh, C. (1986) Age-related order in the ferret’s optic nerve and opticchiasm. J. Neurosci. 6:1635–1642.

Walsh, C., and R.W. Guillery (1985) Age-related order in the optic tract ofthe ferret. J. Neurosci. 5:3061–3069.

Walsh, C., and E.H. Polley (1985) The topography of ganglion cell produc-tion in the cat’s retina. J. Neurosci. 5:741–750.

alter, J., B. Kern-Veits, J. Huf, B. Stolze, and F. Bonhoeffer (1987)Recognition of position-specific properties of tectal membranes byretinal axons in vivo. Development 101:685–696.

Williams, R.W., and P. Rakic (1985) Dispersion of growing axons within theoptic nerve of the embryonic monkey. Proc. Natl. Acad. Sci. USA82:3906–3910.

Withington, D.J., K.E. Binns and M.J. Keating (1990). The maturation ofthe superior collicular map of auditory space is disrupted by developmen-tal visual deprivation. Eur. J. Neurosci. 2:693–703.


axon order in the visual pathway of the quokka wallaby

Documents