development of commissural neurons in the wallaby (macropus eugenii)

Development of Commissural Neuronsin the Wallaby (Macropus eugenii)

F. SHANG,1 K.W.S. ASHWELL,1 L.R. MAROTTE,2 AND P.M.E. WAITE1*1School of Anatomy, University of New South Wales, 2052, New South Wales, Australia

2Developmental Neurobiology, Research School of Biological Sciences, Australian NationalUniversity, ACT 2601, Australia

ABSTRACTWe have examined the development of the laminar and areal distribution of cortical

commissural neurons in a marsupial mammal, the wallaby Macropus eugenii. In this species,commissural axons approach the major cerebral commissure, the anterior commissure, viaeither the internal capsule or the external capsule and first cross the midline at postnatal day14 (P14). By retrogradely labelling these axons with 1,18-dioctadecyl-3,3,38,38-tetramethylin-docarbocyanine (DiI) at P15, we show here that the cell bodies of these neurons are restrictedto a region of cortex adjacent to the rhinal fissure. Most of these labelled neurons are located inthe compact cell zone of the cortical plate, with only a few labelled cells found in the zone ofloosely packed cells deep to this layer. Over the subsequent 66 days, commissural neurons arefound progressively more dorsally, rostrally, and caudally, so that, by P80, they are presentthroughout the extent of the neocortex. At this age, they are mainly pyramidal in morphologyand form a single band within the deeper part of layer 5 of the developing cortex. From P80 toadulthood, the distribution of commissural neurons has been assessed in the visual cortex byusing retrograde transport of horseradish peroxidase. At P80, labelled neurons with imma-ture pyramidal morphology are present throughout the occipital cortex; as in DiI material,somata are located in deep layer 5. At P165, previously shown to be the age when commissuralaxon numbers peak, widespread labelling is present in the occipital region, with labelled cellsnow found in two bands corresponding to layers 3 and 5. After this age, neurons become morerestricted in distribution, so that, by adulthood, commissural neurons are no longer apparentthroughout area 17 but are restricted to a localised region around the area 17/18 boundary.Within this region, labelling is still present in layers 3 and 5 but is more dense in layer 3. Thegradual restriction of commissural fields seen here in the wallaby is similar to that reported inthe neocortex in many eutherians. These findings also support studies in eutheria, suggestingthat subplate neurons do not appear to play a major role in commissural development.J. Comp. Neurol. 387:507–523, 1997. r 1997 Wiley-Liss, Inc.

Indexing terms: anterior commissure; callosum; cortical plate; subplate; marsupials

The development of commissural connections requiresthe directed outgrowth of axons from the appropriatecortical neurons through the cerebral midline to terminatein the suitable cortical region of the contralateral hemi-sphere. The mechanisms responsible for guiding commis-sural axons into and through the corpus callosum, themajor cerebral commissure of eutherian (placental) mam-mals, have received considerable attention (Silver et al.,1982; Valentino and Jones, 1982; Lent et al., 1990). Inmarsupials, which lack a corpus callosum, all corticalconnections pass through the anterior commissure (Ebner,1969). A few studies have examined the development of thecommissural connections traversing the anterior commis-sure in both eutheria (mouse: Sturrock, 1975; hamster:Lent and Guimaraes, 1991; Pires-Neto and Lent, 1993;

rhesus monkey: LaMantia and Rakic, 1994) and poly-protodontid metatheria (opossum: Cabana and Martin,1985).

Studies have reported that early generated neurons ofthe neocortical subplate are responsible for sending thefirst axons out of the cortex; thus, these cells have beenascribed the role of ‘‘pioneering’’ cortical outflow (McCon-nell et al., 1989; DeCarlos and O’Leary, 1992). It has alsobeen noted that some subplate neurons may be retro-

Grant sponsor: Australian Research Council.*Correspondence to: Dr. P. Waite, School of Anatomy, University of New

South Wales, Sydney 2052, Australia. E-mail: [email protected] 17 October 1996; Revised 6 May 1997; Accepted 16 May 1997

THE JOURNAL OF COMPARATIVE NEUROLOGY 387:507–523 (1997)

r 1997 WILEY-LISS, INC.

gradely labelled from the contralateral hemisphere, rais-ing the possibility that these neurons may also play a rolein initial formation of the commissural pathways (Chun etal., 1987; Antonini and Shatz, 1990). Alternatively, be-cause subcortical and commissural projections are formedby topographically discrete populations of neocortical neu-rons, with different laminar locations and afferent inputs,it is feasible that different developmental mechanismsmay be responsible for initial subcortical vs. commissuraltract formation. Recent results obtained by Koester andO’Leary (1994) show that the eutherian corpus callosum ispioneered by early-born neurons in the adjacent cingulatecortex and that subplate neurons do not appear to pioneerthe callosal pathway. Pires-Neto and Lent (1993) havefound that pioneer axons in the hamster anterior commis-sure arise from morphologically immature neurons in theolfactory peduncle and superficial olfactory cortex.

We have examined the early development of commis-sural connections in the tammar wallaby (Macropus euge-nii), a member of the diprotodontid metatheria, which isunique to Australasia. The diprotodontids differ frompolyprotodontids, such as the opossum (Didelphis virgin-iana), in the pathways taken by neocortical commissuralaxons. In the polyprotodontid forebrain, all neocorticalcommissural axons are directed through the externalcapsule to reach the anterior commissure, the sole commis-sural pathway available to neocortical and paleocorticalinterhemispheric connections (Martin, 1967; Ebner, 1969).In the diprotodontids, neocortical commissural axons mayapproach the midline either by the internal or the externalcapsule (Heath and Jones, 1971). Those commissuralfibres traversing the internal capsule join the anteriorcommissure via a discrete bundle that is unique to thisgroup, known as the fasciculus aberrans (Elliot Smith,1902; see Fig. 1). Previous studies (Ashwell et al., 1996a)have shown that the formation of interhemispheric connec-tions is confined to the postnatal period, whereas theyoung wallaby is attached to a teat in the pouch. Theanterior commissure does not appear until postnatal day14 (P14), and the fasciculus aberrans is first visible at P18.The number of axons in the anterior commissure peaks atapproximately P150 at 49–62 million; about 60% are lostbetween this age and adulthood (Ashwell et al., 1996b).

Our present aims were to map the laminar and arealdistribution of the neo- and paleocortical commissuralneurons throughout the establishment of interhemi-spheric connections in the wallaby. Apart from the remark-able arrangement of commissural pathways in this spe-cies, the wallaby also has the experimental advantage ofprolonged postnatal development in the pouch (Renfree etal., 1982). We were particularly interested in determiningwhether there is a topographically and morphologicallydiscrete group of neurons in areas like the subplate region,which is responsible for initial establishment of commis-sural connections, or whether initial connections are sim-ply formed by the earliest generated neurons of the

Fig. 1. Diagram illustrating the pathways taken by commissuralaxons in three mammalian groups. Top: Eutherians (e.g., Rattusnorvegicus). Middle: Polyprotodontid metatherian (e.g., Didelphisvirginiana). Bottom: Diprotodontid metatherian (e.g., Macropus euge-nii). Note that the diprotodont commissural connection is unique inthe use of the internal capsule and fasciculus aberrans by somecommissural axons.

508 F. SHANG ET AL.

nearest cortical region to the commissure. Because thecommissural axons in this marsupial pass through boththe internal and the external capsules, it was of interest tosee whether the path might be pioneered by subplate cells(as postulated for the internal capsule in eutheria) or bycortical plate neurons (as for the eutherian callosal fibres).

We were also interested in determining whether theorganisation of commissural neurons into topographicallydiscrete cortical regions develops by progressive restric-tion or is apparent from the earliest stages of interhemi-spheric connection. Studies in eutheria using bulk injec-tions of the contralateral hemisphere have demonstratedthat, during early development, commissural neurons maybe found throughout the tangential extent of the neocor-tex, even in regions that do not contribute commissuralconnections in the adult (Ivy and Killackey, 1982). In thevisual system of eutherians, the widespread distribution ofcallosal projection cells throughout areas 17 and 18 inimmature animals becomes relatively restricted to thearea 17/18 border in adults (Innocenti and Caminiti, 1980;Innocenti and Clarke, 1984; Olavarria and van Sluyters,1985). Thus, our second aim was to map the areal distribu-tion of commissural neurons in a marsupial mammalduring pouch development to test the hypothesis thatearly, widespread distribution followed by restriction oc-curs, as found in eutheria. The areal distribution has beendemonstrated by using 1,18-dioctadecyl-3,3,38,38-tetra-methylindocarbocyanine (DiI) labelling of the midlineanterior commissure in young wallabies (P15–P80) andcortical injections of horseradish peroxidase (HRP) intothe visual cortex in older animals (P80 to adult).

MATERIALS AND METHODS

Animals

The wallabies (M. eugenii) used in this study included 15pouch young ranging in age from P15 to P150, twojuveniles (P290), and two adults. All animals were ob-tained from a breeding colony; in this species, gestationlasts 28 days. The ages of young wallabies were deter-mined either directly, by noting elapsed time from anobserved birth, or by measuring head length and referringto a chart of head lengths of animals of known age. Fromour experience, the error associated with the latter methodof dating is minimal (no more than 62 days at all ages).The experiments were approved by the Animal EthicsExperimentation Committee of the Australian NationalUniversity.

Carbocyanine dye labelling

Nine pouch young were used for this part of the study(two animals at P15, one animal each at P20, P27, P28,P50, P51, P54, and P79). Animals younger than P35 wereanaesthetised by hypothermia, whereas older animalswere anaesthetised with intramuscular ketamine (1–3mg/100 g body weight, depending on age; Parnell, Silverwa-ter, New South Wales, Australia) and xylazine (0.2 mg/100g; Bayer, Sydney, New South Wales, Australia). Animalswere perfused through the ascending aorta with normalsaline followed by 4% paraformaldehyde in 0.1 M phos-phate-buffered saline (PBS), pH 7.4. The forebrains werebisected sagittally, and the anterior commissure was iden-tified anterior and dorsal to the lamina terminalis. Smallcrystals of DiI (Molecular Probes, Eugene, OR) wereapplied to the cut surface of the anterior commissure. To

standardise the quantity of label applied, approximatelytwo to three 150-µm-diameter fragments of DiI wereapplied to each 0.25 mm2 of anterior commissural area.The tissue was stored in the dark at either room tempera-ture or in a 37°C oven for 4–12 weeks.

Forebrains were embedded in 6% noble agar and werecoronally sectioned at a thickness of 150 µm with the aid ofa Vibratome. The sections were counterstained with 0.001%bisbenzimide in 0.1 M PBS and examined by using a ZeissAxiophot microscope with epifluorescence (excitation filter546 nm, barrier filter 590 nm for DiI; excitation filter 365nm, barrier filter 420 nm for bisbenzimide). Sections werephotographed on Tri-X Pan or colour transparency film(Ektachrome 400).

HRP labelling

Because retrograde labelling with carbocyanine dyes infixed tissue was impractical after P80 due to the increasingbrain size, HRP or HRP conjugated to wheat germ aggluti-nin (WGA-HRP) was used. Wallabies at ages P80 (twoanimals), P150 (four animals), P290 (two animals), andadult (two animals) were used; these ages were chosen tospan the period of axon overproduction and elimination inthe anterior commissure (Ashwell et al., 1996b). Animalsat ages P80 and P150 were anaesthetised with ketamineand xylazine, as described above. Older animals wereanaethetised by intramuscular ketamine (0.2 mg/kg) andxylazine (1.5 mg/kg) followed by intravenous 5% sodiumthiopentone. The region of the area 17/18 border wasinjected with either 30% HRP or 2% WGA-HRP. Eachanimal received 7 to 14 injections of 0.02–0.05 µl perinjection. Alternatively, small crystals of WGA-HRP wereinserted into the cortex on the tip of a glass micropipette.After 24–48 hours, the animals were deeply anaesthetisedwith ketamine (6 mg/100 g) and xylazine (2 mg/100 g) andwere perfused with 1% paraformaldehyde/1.25% glutaral-dehyde in 0.1 M phosphate buffer, pH 7.4, for 30 minutes,followed by 10% sucrose in 0.1 M phosphate buffer. Thebrains were placed in 0.1 M phosphate buffer containing20% sucrose overnight at 4°C and were then sectionedcoronally at 50 µm or 60 µm with the aid of a freezingmicrotome. Sections were processed according to the TMBtechnique (Mesulam, 1982), and alternate sections werecounterstained with neutral red or thionin.

Quantitative analysis and computerreconstruction

Sections were examined with epifluorescence, directillumination, or darkfield with the aid of a Zeiss Axiophot,and the distributions of all DiI- and HRP-labelled neuronswere plotted with the aid of an IBM computer running theMagellan 3.1 program (Halasz and Martin, 1985). Theoutlines of the sections, profiles of the anterior commis-sure, injection sites, and commissural axons were entered,and the locations of all labelled neurons were plotted. Thenumbers of labelled neurons were counted automaticallyfrom the entered data. For the three-dimensional recon-structions, every second (P15 and P28), fourth (P50), 12th(P79 and P150), 20th (P80, P150, and P300), or 24th(adult) section was selected, and the series was displayed.The reconstructed image was exported to Coreldraw forcolouring and enhancement.

DEVELOPMENT OF WALLABY COMMISSURAL NEURONS 509

RESULTS

DiI labelling of commissural neurons

Previous studies in silver-stained sections (Ashwell etal., 1996a) have shown that the anterior commissure firstappears at P14. When DiI was applied to the anteriorcommissure at P15, labelled neurons with the appearanceof immature pyramidal cells were found in the neo- andpaleocortex adjacent to the rhinal fissure (Fig. 2a). At thisage, this region of the cortex consists of a marginal zone, athin cell compact zone (CCZ) at the top of the cortical plate,and, below this, a loosely packed zone (LPZ) of cells (Fig.2b). Commissurally labelled neurons were mainly confinedto the CCZ of the cortical plate and showed prominentapical dendrites, which ramified extensively in the mar-ginal zone (Fig. 2a). Only occasional labelled cells wereseen in the LPZ. Axons arising from the labelled cells weredirected radially until they reached the intermediate zone,where they turned sharply toward the external capsule(Fig. 2a). The path of the commissural axons was re-stricted to the external capsule at this age. Figure 3 showsthe areal distribution of commissural neurons at P15.Labelled cells were restricted to the region around therhinal fissure and did not extend into the frontal oroccipital poles.

At P28–P30, commissural axons could be seen in thefasciculus aberrans and internal capsule (Fig. 4a,b). Thecell bodies that give rise to these axons are located in thelateral cortex, dorsal and ventral to the rhinal fissure.Labelled neurons in both neo- and paleocortex have acharacteristic pyramidal morphology, with tufts of apicaldendrites projecting into the marginal zone (Fig. 4c,d).Labelled somata are located predominantly in the CCZ ofthe upper cortical plate and frequently form tightly packedclusters (Fig. 4c). Labelled cells with multipolar morphol-ogy and primary dendrites extending horizontally werealso seen in the upper half of the LPZ, but these wererelatively rare and inconsistently present (Fig. 4d). Thedistribution of all labelled cells at this age is shown inFigure 3. Far more cells are labelled than at P15 (P28,2,500; P15, 80) and they extend farther dorsally andventrally through the cortex. However, commissural label-ling is still absent in the most dorsal and ventral cortexand in the frontal and occipital poles. Cells labelled in theventricular zone, immediately adjacent to the ventricle(Fig. 3, dorsal parietal cortex), are due to contamination ofthe ventricle by the direct spread of DiI from the deposit onthe anterior commissure.

At P50–P54, labelled neurons were found in the frontaland parietal neocortex and in the adjacent paleocortex.Sparse labelling was also present in the occipital cortex.Figure 3 illustrates the number (48,000) and distributionof labelled commissural neurons in a typical experiment atthis age. Commissural neurons adjacent to the rhinalfissure and the lateral olfactory tract were of a morphologysimilar to neurons at younger ages, with somata lyingprimarily in the CCZ and with short, stout apical dendritesramifying in the marginal zone (Fig. 5b–e). On the otherhand, neocortical commissural neurons from both dorsal(Fig. 5f) and lateral (Fig. 5g) regions were now foundmainly in the superficial layers of the LPZ and typicallyresembled immature pyramidal cells. Lateral neocorticallabelled cells (Fig. 5g) were more mature in appearancethan those in the dorsal cortex (Fig. 5f), with long apicaldendrites extending to the marginal zone before branching

and well-developed basal dendrites that ramified in thesuperficial LPZ. Occasional labelled cells with stellateproximal dendrites, as at P28, were noted in the deeperlayers of the LPZ and the presumptive subplate, but theseneurons comprised only a small percentage of the total.Axons arising from commissural neurons in both theneocortex and the paleocortex adopted radial coursesthrough the LPZ before turning sharply into the develop-ing white matter at the superficial intermediate zone.

At around P80, layers 5 and 6 are apparent, and layer 4is just recognisable beneath the CCZ of the cortical plate(Fig. 6e). The majority of labelled neocortical neurons werepyramidal in morphology, with somata confined to a tightband in the deeper part of layer 5 (Fig. 6d,e) and apicaldendrites extending through the cortical plate to themarginal zone before ramifying. Similarly labelled neu-rons in the paleocortex (Fig. 6b,c) were also located mainlybelow the CCZ at this stage and had a morphology similarto neurons in the neocortex. The developing white matterbelow layer 6 was filled with a dense plexus of efferentaxons and collaterals (Fig. 6d,f). Very few axons were seento enter the superficial cortex above the level of labelledperikarya. The number (approximately 160,000 cells) anddistribution of commissural neurons were greater than atprevious ages, extending widely throughout the corticalcircumference (Fig. 10). Labelling was particularly densein the frontal and parietal cortex, becoming more sparsedorsally and caudally.

HRP labelling of visual commissural neurons

For this part of the study, we confined our analysis to thevisual cortex, because extensive injections of the entireneocortex (as required for assessment of commissuralneuron distribution in the entire cortex) result in highmortality of pouch young. Widespread injections were usedto cover all of area 17 and the area 17/18 border andextended through the thickness of the cortex and into theunderlying white matter (see, e.g., Fig. 7a). Cytoarchitec-tonic features of areas 17 and 18 in the wallaby have beendescribed before (Sheng et al., 1990). The border betweenarea 17 and 18 can be recognised by the decrease inthickness of layer 3 (Sheng et al., 1990), as in othermammals (Innocenti and Caminiti, 1980; Cusick andLund, 1981).

In the animals that received dorsal occipital injections atP80, labelled commissural neurons were identified in acontinuous band in the contralateral dorsal occipital cor-tex. A representative section is shown in Figure 8a. Thisband of labelled cells extended laterally from the cingulatecortex, through presumptive area 17, and into area 18.These neurons were located in the deeper part of layer 5(Fig. 8b,c), as previously seen in the dorsal and lateralneocortex in the DiI-labelled material (Fig. 6d–f).

At P150–P165, the age at which the number of axons inthe anterior commissure peaks (Ashwell et al., 1996b),labelled commissural neurons still showed a widespreaddistribution throughout the dorsal occipital cortex follow-ing contralateral hemisphere injections (Fig. 9a). At thisage, all layers present in the adult cortex can be recognised(Fig. 9b). This is the first age in our series at which thedistribution of commissural neurons showed the bilaminararrangement (Fig. 9a) characteristic of these neurons inthe adult (Sheng et al., 1990), with labelled cells present inlayers 3 and 5. Labelled neurons in both layers showedcharacteristic pyramidal cell morphology with well-

510 F. SHANG ET AL.

Fig. 2. a: Photomontage of 1,18-dioctadecyl-3,3,38,38-tetramethylin-docarbocyanine (DiI)-labelled commissural neurons and axons in thelateral paleocortex of a postnatal day 15 (P15) wallaby. The insetbelow shows the position of the region photographed. Labelled cellsare in the compact cell zone (CCZ) of the cortical plate. b: Nissl-stained

coronal section showing the structure of this region of the cortex at thisage. hi, Hippocampus; lv, lateral ventricle; ac, anterior commissure; ec,external capsule; lot, lateral olfactory tract; MZ, marginal zone; LPZ,loosely packed zone. Scale bars 5 100 µm.

Fig. 3. Distribution of labelled neurons at P15, P28, and P51 following DiI insertion into the anteriorcommissure. Arrows indicate the rhinal fissure, and arrowheads indicate the lateral olfactory tract. Eachdot represents one labelled cell.

512 F. SHANG ET AL.

developed apical and basal dendrites similar to those seenat P290 (Fig. 10d–f).

By P290, the number of axons in the anterior commis-sure is falling toward adult values (Ashwell et al., 1996b).Figure 10a shows that, by this age, the distribution ofcommissural neurons is more restricted than at P160, withmore medial regions of area 17 that lack any labelled cells,especially in caudal sections. Labelled cells were clusteredaround the area 17/18 border and still extended into area17 in more rostral sections. Figure 10b–f shows examplesof labelled neurons at this age. The bilaminar arrange-ment first seen at P150 is still apparent, but the numberand density of labelling in layer 5 is less than in layer 3(Fig. 10c). Neurons in layer 3 (Fig. 10d,f) are generallylarger than those in layer 5 and are typical of pyramidalcells, with prominent apical dendrites; those in layer 5(Fig. 10e) are less densely packed and less homogeneous in

structure, with roundish or inverted pyramidal somata aswell as typical pyramidal morphologies.

The bilaminar distribution of commissural somata per-sists into adulthood (Fig. 11), with more numerous labelledcells in layer 3 than in layer 5 and with cells showingmorphologies similar to those seen at P290. In the adult,labelled commissural neurons show a further restriction intangential distribution, with the majority of labelled cellslocated around the area 17/18 border, as found previouslyfor the mature visual interhemispheric connections (Shenget al., 1990). Labelled cells are absent from most of caudalarea 17 and are sparsely distributed or confined to smallregions within rostral area 17. The labelled cells lyingventrolateral to area 18 in caudal sections probably repre-sent cells in extrastriate visual areas (Sheng et al., 1990),whereas, more rostrally, they probably lie in auditoryregions (Mayner, 1989).

Fig. 4. a,b: Low-power view of the lateral telencephalon at P28after staining with bisbenzimide (a) or labelling with DiI application tothe anterior commissure (b). Axons project though both the external(thin arrow) and internal (star) capsules. The arrow in a indicates theposition of the rhinal fissure. c: High-power view of labelled cells in theCCZ of the lateral neocortex. The axons descend through the LPZ

before turning sharply (broad arrows). Note the presence of occasionallabelled cells below the CCZ (thin arrow). d: Labelled cells in thepaleocortex; occasional deeper labelled cells are present (arrow). Forabbreviations, see Figure 2. Scale bars 5 500 µm in b (also applies toa), 100 µm in c,d.


Fig. 5. Examples of commissural neurons at P51. a: Schematicdiagram of a coronal section showing the location of photomicrographsb–g. b,c: Medium-power view of the paleocortex exposed for bisbenz-imide (b) and DiI (c). Labelled cells are present in the CCZ. d,e:Low-power view of the neocortex adjacent to the rhinal fissure. The

majority of labelled somata are located in the CCZ. f,g: Labelled cellsin the dorsal (f) and lateral (g) neocortex. Cells of the lateral cortex aremore mature. ic, Internal capsule. For other abbreviations, see Figure2. Scale bars 5 200 µm (c also applies to b; e also applies to d).

DISCUSSION

The present studies have shown that the earliest commis-sural neurons in the wallaby are located in a restrictedregion of the neocortex and paleocortex around the rhinalfissure and project exclusively via the external capsule.These earliest cells are found in the CCZ of the corticalplate and appear to be immature pyramidal cells. ByP28–P30, projections also pass via the external capsuleand fasciculus aberrans. The distribution of commissuralneurons gradually becomes more extensive, so that, byP80, most of the cortex provides input to the commissure.This widespread distribution continues in the visual cor-tex until at least P165 but is followed by a period ofrestriction, so that, in the adult, relatively few neurons inarea 17 project to the commissure, and these are mainly inregions adjacent to the border with area 18. Both theidentity of the pioneering cells and the developmentalphases of widespread connectivity followed by selectiverestriction parallel results on callosal development ineutherians and suggest that common mechanisms mayoperate across widely divergent mammalian species. How-ever, before discussing some of these parallels in moredetail, a number of technical limitations should be consid-ered.

Technical limitations

For the earliest ages examined here, the present resultsdepend on the diffusion of DiI along axons of the anteriorcommissure. Although every effort was made to labelaxons across the whole commissure, it is likely that someaxons were missed. Thus, the distribution of label through-out the cortex represents a minimal extent of projectionsat each age. That such incomplete labelling has notsignificantly altered the results presented here is indi-cated by two findings. First, the overall pattern of labellingwas consistent in animals of similar ages (e.g., two at P15;P27 and P28; P50, P51, and P54). Second, the distributionof labelled cells gradually extended throughout the cortex;isolated patches of labelled cells that were disconnectedfrom other areas of labelling were seldom seen. Thus, itseems likely that the DiI placements had not missed largeareas of the commissure. Contamination of surfaces otherthan the anterior commissure by inadvertent DiI spreadresulted in some artifactual labelling (e.g., of cells in theventricular zone; see Fig. 3) but did not pose a problem forinterpretation of commissural cells, because axon path-ways could be followed through successive sections. Slowtransport of DiI can also cause technical problems in thistype of experiment, but the present results show clearlylabelled dendrites extending to the marginal zone in allareas where labelling is present. Animals that showedgradually fading label, characteristic of insufficient trans-port time, were not included in the results. For animalsolder than P80, distances were too great for reliable use ofDiI. Finally, it should be mentioned that DiI can some-times give transcellular labelling, and this will be dis-cussed in relation to the occasional labelling seen in cellsin the LPZ in young animals.

For the HRP injections, incomplete labelling of allcortical layers can result in incomplete distribution ofprojection cells. However, we were careful to extend ourinjections down to the white matter, so that label wasdistributed to all cortical layers. For analysis, only regions

of cortex contralateral to intense deposits of HRP wereretained.

Nature of pioneer axons in the wallabyanterior commissure

The concept of pioneer axons has been borrowed fromthe body of literature on invertebrate neural development,in which special classes of developing neurons with dis-tinct morphologies are responsible for making the initialconnections in the nervous system. These pioneer fibresperform this task at stages of neural development whenthe nervous system is small and when the distances thatmust be traversed are modest. Subsequently, the growth ofthe bulk of axons that comprise the tract is directed andfacilitated by the pioneer axons in ways that are still notcompletely clear.

Within the developing eutherian cortex, candidates forthe role of pioneer neurons have been identified as theneurons of the subplate region. The axons of these cellshave been said to pioneer the pathway through the inter-nal capsule and into the thalamus in the cat (McConnell etal., 1989) and the rat (De Carlos and O’Leary, 1992;however, see below). These cells are among the earliestborn neurons of the cortex and are largely transient in thecat (Luskin and Shatz, 1985), whereas, in the rat, they giverise to layer 6b, the deeper part of layer 6 (Valverde et al.,1989). Subplate neurons have now been described ashaving a range of morphologies, including pyramidal,inverted pyramidal, multipolar, and horizontally and radi-ally oriented bipolar (Mrzljak et al., 1988; McConnell et al.,1989; Valverde et al., 1989; De Carlos and O’Leary, 1992),whereas those in the cortical plate are more commonlyoriented radially and are pyramidal in morphology (DeCarlos and O’Leary, 1992; Erzurumlu and Jhaveri, 1992;Clasca et al., 1995). In the wallaby neocortex, a presump-tive subplate region has been identified below the LPZ atthe top of the intermediate zone (Sheng et al., 1991), basedon the original cytoarchitectonic criteria of Kostovic andMollivar (1974). However, neurons from this region do notextend the first axons into the thalamus (Sheng et al.,1991). Our present findings clearly indicate that theseneurons do not appear to be responsible for extending thefirst axons into the anterior commissure. Somata of axonsfirst reaching the midline lie within the CCZ at the top ofthe cortical plate. They have been labelled at P15, within aday of the initial appearance of the anterior commissure,as seen in silver-stained material (Ashwell et al., 1996a).Thus, it is unlikely that transient axons from the subplatepioneer the anterior commissure and withdraw before thefirst age examined in our series. Along with not acting aspioneers, subplate cells in the wallaby also failed tocontribute substantially to commissural projections dur-ing development, in contrast to results in the cat (Chun etal., 1987; Antonini and Shatz, 1990). Occasional DiI-labelled cells were seen below the CCZ, some of which hadmultipolar or horizontal morphologies similar to thoseascribed for subplate cells. However, these cells were rareand were inconsistently present; they were sometimescompletely absent throughout a large region of cortex. Itwas thought that these cells were most likely to have beentranscellularly labelled, presumably by having mem-branes closely adjacent to those of labelled cells. Suchtranscellular labelling is known to occur (Godement et al.,1987).


Figure 6

The argument against the earliest commissural cellsbeing subplate neurons is further supported by the mor-phology of the majority of the cells labelled here at P15.The morphology of these earliest labelled cells does notresemble that described for subplate neurons (see above),

but it closely resembles the morphology of the first callo-sally projecting neurons described in the rat cingulatecortex (Koester and O’Leary, 1994). These cells have aradial orientation, with superficially projecting dendritesbranching in tufts in the marginal zone. In the rodent, itwas suggested that they mature into callosally projectingcells with classical pyramidal morphology, and a similarfate is likely for many of these cells in the wallaby.Commissurally projecting cells with stellate morphologyhave been described in the cat (Innocenti and Caminiti,1980) and can develop from pyramidal cells by regressionof the apical dendrite (Vercelli et al., 1992). In the adultwallaby, most commissurally labelled cells in layer 3 werepyramidal in structure, but cells in layer 5 were morevaried, with pyramidal, inverted pyramidal, and stellatemorphologies present. The large size of labelled pyramidalcells in layer 3 has also been noted as a feature of the area17/18 boundary in the cat (Shoumura, 1974).

Fig. 7. Distribution of retrogradely labelled commissural neurons in the cortex of a P79 wallabyfollowing application of DiI to the anterior commissure. White arrows indicate the rhinal fissure, blackarrows indicate the area 17/18 border, and each dot represents six labelled cells.

Fig. 6. Examples of commissural neurons at P79. a: Schematicdiagram of a coronal section showing the location of photomicrographsb–f. Arrow marks the rhinal fissure. b,c: Low-power view of the rhinalfissure stained with bisbenzimide (b) or labelled by DiI (c) to show thelocation of labelled commissural cells in both the paleo- and neocortex.In both regions, the labelled cells form a single layer, several cellsdeep, below the CCZ (star). d–f: Labelling in the dorsal (d) and lateral(f) neocortex. In e, the section from f was exposed for bisbenzimide toshow the cell distribution. Labelled cells are located in the deep part oflayer 5. wm, white matter. Scale bars 5 100 µm in c,f (c also applies tob; f also applies to e), 500 µm in d.


Ontogeny and phylogenyof commissural pathways

The report by Koester and O’Leary (1994) has shownthat, although a few subplate neurons may form interhemi-spheric connections during eutherian development, theydo not pioneer those connections; that role appears to betaken by pyramidal neurons of the adjacent cingulatecortex. This contrasts with the reported role of subplateneurons in pioneering subcortical pathways, as mentionedabove. This finding has prompted Koester and O’Leary(1994) to postulate that the phylogenetically older subcor-tical connections (through the internal capsule) are depen-dent upon the transient population of subplate neurons,which may represent remnants of a primordial or reptiliancortex (Marin-Padilla, 1978). They propose that the phylo-genetically younger corpus callosum appeared after thedevelopmental role of the subplate neurons had beenestablished and perhaps genetically fixed. Thus, the mecha-nisms controlling axon extension along the callosal trajec-tory would have evolved without access to subplate facilita-tion and pathway pioneering.

If this hypothesis is correct, then one would expect thatthe development of the phylogenetically older anteriorcommissural connections of the metatherian brain wouldrequire assistance from the transient subplate neurons.This would be particularly likely in the wallaby, becausemany commissural axons in this species follow the inter-nal capsule. Our findings indicate that, in fact, commis-sural pathways in the wallaby are pioneered by neurons ofdistinctively pyramidal morphology located in the superfi-cial parts of the cortical plate. Furthermore, in the wallaby,subplate neurons do not pioneer the pathways from thevisual (Sheng et al., 1991) or somatosensory (Leamey etal., 1996) cortex to the thalamus. Rather, they are pio-neered by neurons in layer 5 in the dorsal visual cortex andthe somatosensory cortex and in layers 5 and 6 in thelateral visual cortex. The recent studies of Clasca et al.(1995) indicate that, even in a eutherian species, theferret, it is not the subplate neurons but cells in layer 5that first extend axons to the thalamus, brainstem, andspinal cord. In another eutherian, the hamster, Miller etal. (1993) report that it is both the subplate cells and cellsof layer 6 together that first project to the thalamus.Clasca et al. suggest that the early labelling of subplatecells from the thalamus reported previously was due totransneuronal transport of carbocyanine dye from labelledthalamic afferents. Rather than pioneering commissuraland subcortical pathways, subplate neurons are the first toproject into the underlying white matter (Clasca et al.,1995), and they appear to make a major contribution toipsilateral corticocortical connections (Miller et al., 1993;Koester and O’Leary, 1994).

Progressive refinement of commissuralneuron fields during development

Studies in eutherians have shown that commissurallyprojecting neurons are much more widespread in theneocortex early in development compared with the maturedistribution. In the rat, for example (Ivy and Killackey,1982), neurons in lamina Va of the barrel field area, whichdo not project intercortically in the mature rat, initiallycontribute axons to the corpus callosum (at about P8).These inappropriately projecting commissural neuronssimultaneously project to their appropriate target, the

Fig. 8. a: Diagram of a representative coronal section from a P80wallaby showing the injection site (on the right) and the correspondingretrograde labelling in the contralateral cortex. Each dot representsthree labelled cells, and the arrow indicates the 17/18 border, witharea 17 medial to the arrow. The three levels of shading at the injectionsite correspond to intense, moderate, and light levels of horseradishperoxidase (HRP) staining. b,c: Photomicrographs taken from the siteindicated in a showing HRP labelling (darkfield; b) and an adjacentNissl-stained section (c). Cortical layers 4, 5, and 6 and the whitematter (wm) can be seen below the CCZ. Labelled cells are present inthe deep part of layer 5. Scale bar 5 200 µm.

518 F. SHANG ET AL.

ipsilateral motor cortex. The gradual restriction in thedistribution of callosally projecting neurons appears to bedue to pruning of commissural collaterals of immatureprojection neurons rather than the death of inappropri-ately situated neurons (Ivy and Killackey, 1982; Chalupaand Killackey, 1989). A sequence of early widespreaddistribution followed by restriction has been reported inthe sensory-motor cortex in the opossum (Cabana andMartin, 1985). Similarly, in the visual cortex in the rat,shrew, cat, and rabbit, cells are present throughout areas17 and 18 early in development but become more restrictedin the adult (Innocenti et al., 1977; Innocenti and Cami-niti, 1980; Chow et al., 1981; Kretz and Rager, 1992). Thepresent results and those of Sheng et al. (1990) suggestthat the wallaby has a comparable sequence. The visualcortex is one of the latest regions to develop commissuralconnections. The present study found relatively few com-

missurally projecting cells present in the occipital cortexbefore P80, whereas Sheng et al. (1990) first found recipro-cal connections between areas 17 the two sides at P111.The present study showed connections to be more exten-sive at P165 before they started a gradual restriction to theadult pattern. The process proceeds slowly; labelling ofcells at P290 still shows a relatively immature pattern.

The gradual restriction in distribution of callosallyprojecting neurons in eutherians is accompanied by loss ofaxons from the corpus callosum (Koppel and Innocenti,1983; Berbel and Innocenti, 1988), and a similar reductionin axon numbers has been shown recently to occur in thewallaby anterior commissure (Ashwell et al., 1996b). Ineutheria, regression in the anterior commissure varieswith the species studied and may correlate with the extentof neocortical fibres present. Thus, in the hamster, wherethe commissure mainly connects regions of paleocortex, no

Fig. 9. a: Distribution of labelled commissural neurons in theoccipital cortex of a P165 wallaby following contralateral injection ofHRP. Labelling is apparent throughout area 17 at this age. Arrowsindicate the 17/18 border, and each dot corresponds to one labelled cell.

b: Nissl-stained coronal section through the cortex at this age showingthat all six cortical layers are now identifiable. Labelled cells shown ina are in layers 3 and 5. wm, white matter. Scale bar 5 0.5 mm.


Fig. 10. a: Distribution of labelled commissural neurons in theoccipital cortex of a P290 wallaby following contralateral injection ofHRP. Labelled cells are no longer continuous throughout area 17 butare relatively restricted to its more lateral region, near the 17/18border (arrows), particularly in caudal sections. Each dot correspondsto one labelled cell. b,c: Low-power photomicrographs from the areaindicated in (a) with Nissl staining (b) and darkfield (c) to show the

location of HRP-labelled neurons. A bilaminar distribution in layers 3(arrow) and 5 (star) is apparent. d–f: Higher power views of labelledcells from layer 3 (d,f) and 5 (e) showing normal pyramidal cells as wellas cells with inverted pyramidal morphology (thin arrow) and roundedsomata (thick arrow). wm, white matter. Scale bars 5 1 mm in c (alsoapplies to b), 100 µm in d,e, 20 µm in f.

520 F. SHANG ET AL.

regression of distribution was seen (Lent and Guimaraes,1991). In primates, the anterior commissure connectsneocortical regions, such as the temporal lobe (Jouandetand Gazzaniga, 1979; Jouandet et al., 1984), as well aspaleocortex, and axon elimination has been noted (LaMan-tia and Rakic, 1994).

The present findings with the use of retrograde trans-port of HRP indicate two changes that occur in the

distribution of visual commissural neurons after P80. Thefirst of these is the appearance of a bilaminar grouping oflabelled neurons, which emerge between P80 and P150.The cortex of metatheria, like eutheria, develops in an‘‘inside-out’’ fashion, with deeper layers being born anddifferentiating from the cortical plate prior to the superfi-cial layers (Reynolds et al., 1985; Marotte and Sheng,1995). At P80, the single layer of labelled cells is in layer 5,

Fig. 11. a: Distribution of labelled commissural neurons in theadult wallaby occipital cortex following multiple injections of HRP intothe visual areas of the contralateral hemisphere. The arrows indicatethe area 17/18 border, and each dot corresponds to one labelled cell.b,c: HRP-labelled cells from the region of the 17/18 border indicated by

the boxed region in a. Darkfield, low-power view (b) shows thebilaminar distribution of labelling in layers 3 (arrowhead) and deep 5(star), and higher power brightfield view (c) shows labelled pyramidalcells in layer 3. Scale bar 5 200 µm in b, 100 µm in c.


identified both cytoarchitectonically here and by birth-dating studies (Marotte and Sheng, 1995), and layer 3 hasyet to differentiate from the cortical plate. By P150, layer 3is well differentiated, and the bilaminar pattern corre-sponds to cells in cortical layers 3 and 5. Like the wallaby,cells projecting through the corpus callosum are alsolocated in layers 3 and 5 in the rat (Jacobson and Tro-janowski, 1974; Wise and Jones, 1976) and the opossum(Cabana and Martin, 1985; Granger et al., 1985). However,this is not the case for all mammals. The kitten shows abilaminar distribution with dense projections from layers3 and upper 4 and layer 6. However, by adulthood, thelabelled cells are located mainly in layer 3, with onlyoccasional labelled cells in layer 6. Similarly, in primates,most callosal cells arise from the supragranular layers(Killackey and Chalupa, 1986; Schwartz and Goldman-Rakic, 1991), although the contribution from infragranu-lar cells varies with cortical region (Jouandet et al., 1984).

The second change in distribution, occurring after P165,is the gradual restriction of commissural neurons to area18, as discussed above. Our previous finding of peak axonnumbers at P150 (Ashwell et al., 1996b) is consistent withthe elimination of connections during this stage. Presum-ably, axon elimination for the occipital cortex is delayedrelative to that for the frontal cortex, so it is not surprisingthat the juvenile distribution of commissural neuronsthroughout area 17 is retained until after P165.

The change in distribution seen here from P165 toadulthood indicates that commissural neurons in thewallaby undergo a developmental restriction similar tothat previously reported for neocortical connections inmany eutherians. Thus, although they are widely diver-gent in evolution, marsupial and placental mammalsappear to share similar developmental programs for theestablishment of their cortical hemispheric connections.

ACKNOWLEDGMENTS

We thank Ms. A. Devlin and Mr. K. Williams for invalu-able technical assistance and Dr. R. Meischke for veteri-nary advice.

LITERATURE CITED

Antonini, A., and C.J. Shatz (1990) Relation between putative transmitterphenotypes and connectivity of subplate neurons during cerebralcortical development. Eur. J. Neurosci. 2:744–761.

Ashwell, K.W.S., P.M.E. Waite, and L.R. Marotte (1996a) Ontogeny of theprojection tracts and commissural fibres in the forebrain of the wallaby(Macropus eugenii): A comparison of timing with other mammals. BrainBehav. Evol. 47:8–22.

Ashwell, K.W.S., L.R. Marotte, L. Li, and P.M.E. Waite (1996b) Anteriorcommissure of the wallaby (Macropus eugenii): Adult morphology anddevelopment. J. Comp. Neurol. 366:478–494.

Berbel, P., and G.M. Innocenti (1988) The development of the corpuscallosum in cats: A light- and electron-microscope study. J. Comp.Neurol. 276:132–156.

Cabana, T., and G.F. Martin (1985) The development of commissuralconnections of somatic motor-sensory areas of neocortex in the NorthAmerican opossum. Anat. Embryol. 171:121–128.

Chalupa, L.M., and H.P. Killackey (1989) Process elimination underliesontogenetic change in the distribution of callosal projection neurons inthe postcentral gyrus of the fetal rhesus monkey. Proc. Natl. Acad. Sci.USA 86:1076–1079.

Chow, K.L., H.D. Baumbach, and R. Lawson (1981) Callosal projections ofthe striate cortex in the neonatal rabbit. Exp. Brain Res. 42:122–126.

Chun, J.J.M., M.J. Nakamura, and C.J. Shatz (1987) Transient cells of thedeveloping mammalian telencephalon are peptide-immunoreactive neu-rons. Nature 325:617–620.

Clasca, F., A. Angelucci, and M. Sur (1995) Layer-specific programs ofdevelopment in neocortical projection neurons. Proc. Natl. Acad. Sci.USA 92:11145–11149.

Cusick, C.G., and R.D. Lund (1981) The distribution of the callosalprojection to the occipital visual cortex in rats and mice. Brain Res.214:239–259.

DeCarlos, J.A., and D.D.M. O’Leary (1992) Growth and targeting ofsubplate axons and establishment of major cortical pathways. J.Neurosci. 12:1194–1211.

Ebner, F.F. (1969) A comparison of primitive forebrain organisation inmetatherian and eutherian mammals. Ann. NY Acad. Sci. 167:241–257.

Elliott Smith, G. (1902) On a peculiarity of the cerebral commissure incertain marsupialia, not hitherto recognised as a distinctive feature ofthe diprotodontia. Proc. R. Soc. London B 70:226–231.

Erzurumlu, R.S., and S. Jhaveri (1992) Emergence of connectivity in theembryonic rat parietal cortex. Cerebral Cortex 2:336–352.

Godement, P., J. Vanselow, S. Thanos, and F. Bonhoeffer (1987) A study indeveloping visual systems with a new method of staining neurons andtheir processes in fixed tissue. Development 101:697–713.

Granger, E.M., R.B. Masterton, and K.K. Glendenning (1985) Origin ofinterhemispheric fibres in acallosal opossum (with a comparison tocallosal origins in rat). J. Comp. Neurol. 241:82–98.

Halasz, P., and P.R. Martin (1985) A microcomputer based system forsemiautomatic analysis of histological sections. Proc. R. Microsc. Soc.19:312P.

Heath, C.J., and E.G. Jones (1971) Interhemispheric pathways in theabsence of a corpus callosum. An experimental study of commissuralconnections in the marsupial phalanger. J. Anat. 109:253–270.

Innocenti, G.M., and R. Caminiti (1980) Postnatal shaping of callosalconnections from sensory areas. Exp. Brain Res 38:381–394.

Innocenti, G.M., and S. Clarke (1984) The organization of immaturecallosal connections. J. Comp. Neurol. 230:287–309.

Innocenti, G.M., L. Fiore, and R. Caminiti (1977) Exuberant projection intothe corpus callosum from the visual cortex of new born cats. Neurosci.Lett. 4:237–242.

Ivy, G.O., and H.P. Killackey (1982) Ontogenetic changes in the projectionsof neocortical neurons. J. Neurosci. 2:735–743.

Jacobson, S., and J.Q. Trojanowski (1974) The cells of origin of the corpuscallosum in rat, cat and rhesus monkey. Brain Res. 74:149–155.

Jouandet, M.L., and M.S. Gazzaniga (1979) Cortical field of origin of theanterior commissure of the rhesus monkey. Exp. Neurol. 66:381–397.

Jouandet, M.L., L.J. Garey, and H.-P. Lipp (1984) Distribution of the cells oforigin of the corpus callosum and anterior commissure in the marmosetmonkey. Anat. Embryol. 169:45–59.

Killackey, H.P., and L.M. Chalupa (1986) Ontogenetic change in thedistribution of callosal projection neurons in the postcentral gyrus ofthe fetal rhesus monkey. J. Comp. Neurol. 244:331–348.

Koester, S.E., and D.D.M. O’Leary (1994) Axons of early generated neuronsin cingulate cortex pioneer the corpus callosum. J. Neurosci. 14:6608–6620.

Koppel, H., and G.M. Innocenti (1983) Is there a genuine exuberancy ofcallosal projections during development? A quantitative electron micro-scopic study in the cat. Neurosci. Lett. 41:33–40.

Kostovic, I., and M.E. Molliver (1974) A new interpretation of the laminardevelopment of cerebral cortex: Synaptogenesis in different layers ofneopallium in the human fetus. Anat. Rec. 178:395.

Kretz, R., and G. Rager (1992) Postnatal development of area 17 callosalconnections in Tupaia. J. Comp. Neurol. 326:217–228.

LaMantia, A.-S., and P. Rakic (1994) Axon overproduction and eliminationin the anterior commissure of the developing rhesus monkey. J. Comp.Neurol. 340:328–336.

Leamey, C.A., L.R. Marotte, and P.M.E. Waite (1996) The initial projectionfrom somatosensory cortex to the thalamus in the wallaby (Macropuseugenii). Proc. Aust. Neurosci. Soc. 7:119.

Lent, R., and R.Z.P. Guimaraes (1991) Development of paleocortical projec-tions through the anterior commissure of hamsters adopts progressive,not regressive, strategies. J. Neurobiol. 22:475–498.

Lent, R., C. Hedin-Pereira, J.R.L. Menezes, and S. Jhaveri (1990) Neurogen-esis and development of callosal and intracortical connections in thehamster. Neuroscience 38:21–37.

Luskin, M.B., and C.J. Shatz (1985) Studies of the earliest generated cellsof the cat’s visual cortex: Cogeneration of subplate and marginal zones.J. Neurosci. 5:1062–1075.

Marin-Padilla, M. (1978) Dual origin of the mammalian neocortex and theevolution of the cortical plate. Anat. Embryol. 152:109–126.

522 F. SHANG ET AL.

Marotte, L.R., and X.-M. Sheng (1995) The relationship between developingvisual cortical layers and thalamocortical afferents in the wallaby(Macropus eugenii). Proc. Aust. Neurosci. Soc. 6:80.

Martin, G.F. (1967) Interneocortical connections in the opossum, Didelphisvirginiana. Anat. Rec. 157:607–616.

Mayner, L. (1989) A cytoarchitectonic study of the cortex of the tammarwallaby, Macropus eugenii. Brain Behav. Evol. 33:303–316.

McConnell, S.K., A. Ghosh, and C.J. Shatz (1989) Subplate neurons pioneerthe first axon pathway from the cerebral cortex. Science 245:978–982.

Mesulam, M.M. (1982) Tracing Neural Connections With HorseradishPeroxidase. Bath, United Kingdom: John Wiley and Sons.

Miller, B., L. Chou, and B.L. Finlay (1993) The early development ofthalamocortical and corticothalamic projections. J. Comp. Neurol.335:16–41.

Mrzljak, L., H.B.M. Uylings, I. Kostovic, and C.G. Van Eden (1988)Prenatal development of neurons in the human prefrontal cortex: I. Aqualitative Golgi study. J. Comp. Neurol. 271:355–386.

Olavarria, J., and R.C. van Sluyters (1985) Organization and postnataldevelopment of callosal connections in the visual cortex of the rat. J.Comp. Neurol. 239:1–26.

Pires-Neto, M.A., and R. Lent (1993) The prenatal development of theanterior commissure in hamsters: Pioneer fibres lead the way. Dev.Brain Res. 72:59–66.

Renfree, M.B., A.B. Holt, S.W. Green, J.P. Carr, and D.B. Cheek (1982)Ontogeny of the brain in a marsupial (Macropus eugenii) throughoutpouch life. I. Brain growth. Brain Behav. Evol. 20:57–71.

Reynolds, M.L., M.E. Cavanagh, K.M. Dziegielewska, L.A. Hinds, N.R.Saunders, and C.H. Tyndale-Biscoe (1985) Postnatal development ofthe telencephalon of the tammar wallaby (Macropus eugenii). Anat.Embryol. 173:81–94.

Schwartz, M.L., and P.S. Goldman-Rakic (1991) Prenatal specification ofcallosal connections in rhesus monkey. J. Comp. Neurol. 307:144–162.

Sheng, X.-M, L.R. Marotte, and R.F. Mark (1990) Development of connec-tions to and from the visual cortex in the wallaby (Macropus eugenii). J.Comp. Neurol. 300:196–210.

Sheng, X.-M, L.R. Marotte, and R.F. Mark (1991) Development of thelaminar distribution of thalamocortical axons and corticothalamic cellbodies in the visual cortex of the wallaby. J. Comp. Neurol. 307:17–38.

Shoumana, K. (1974) An attempt to relate the origin and distribution ofcommissural fibres to the presence of large and medium pyramids inlayer III in the cat’s visual cortex. Brain Res. 67:13–25.

Silver, J., S.E. Lorenz, D. Wahlsten, and J. Coughlin (1982) Axonalguidance during development of the great cerebral commissures: De-scriptive and experimental studies, in vivo, on the role of preformedglial pathways. J. Comp. Neurol. 210:10–29.

Sturrock, R.R. (1975) A quantitative electron microscopic study of myelina-tion in the anterior commissure of the mouse brain. J. Anat. 119:67–75.

Valentino, K.L., and E.G. Jones (1982) The early formation of the corpuscallosum: A light and electron microscopic study in foetal and neonatalrats. J. Neurocytol. 11:583–609.

Valverde, F., M.V. Facal-Valverde, M. Santacana, and M. Heredia (1989)Development and differentiation of early generated cells of sublayerVIb in the somatosensory cortex of the rat: A correlated Golgi andautoradiographic study. J. Comp. Neurol. 290:118–140.

Vercelli, A., F. Assal, and G.M. Innocenti (1992) Emergence of callosallyprojecting neurons with stellate morphology in the visual cortex of thekitten. Exp. Brain Res. 90:346–358.

Wise, S.P., and E.G. Jones (1976) The organization and postnatal develop-ment of the commissural projection of the rat somatic sensory cortex. J.Comp. Neurol. 168:313–344.


development of commissural neurons in the wallaby (macropus eugenii)

Documents