Mechanisms of Ageing and Development

123 (2002) 473–479

Development of dendritic bundles of pyramidal neurons inthe rat visual cortex

Roberta Curtetti, Diego Garbossa, Alessandro Vercelli *Department of Anatomy, Pharmacology and Forensic Medicine, Uni�ersity of Torino Medical School, Corso M. D’Azeglio 52,

Torino, Italy

Received 23 January 2001; received in revised form 11 June 2001; accepted 27 July 2001

Abstract

The apical dendrites of pyramidal neurons in the cerebral cortex form vertical bundles whose distribution anddensity vary across species and areas. To understand their relationships with cortical columns, we labeled retrogradelyneurons from the white matter underlying the visual cortex with 1,1�-dioctadecyl-3,3,3�,3�-tetramethylindocarbocya-nine perchlorate (DiI) at P3 and P10 and with biotinylated dextran amine at P30. We also mapped the distributionof apical dendrites in tangential sections, immunostained for microtubule-associated proteins (MAP2). Theircomposition and distribution were studied with Neurolucida and NeuroExplorer software. The apical dendrites ofpyramidal neurons formed different bundle types: at P3 we found bundles formed (a) by neurons located in corticalplate; (b) by layer V neurons; and (c) by upper layer V neurons and cortical plate neurons. At P10, the amount ofsupragranular neurons participating in the bundles increased. The inter-dendritic and inter-bundle distances increasedwith age. These findings confirm that dendritic bundles are present in the rat visual cortex early in development andare formed by neurons belonging to different cortical layers. The existence of different types of bundles relative to thelayer of location of their parent neurons suggests that they are heterogeneous from each other in nature and in thepattern of connectivity. © 2002 Elsevier Science Ireland Ltd. All rights reserved.

Keywords: Apical dendrite; Cluster; Cerebral cortex; Cortical column

www.elsevier.com/locate/mechagedev

1. Introduction

The cerebral cortex is arranged in six layersparallel to its surface, differing from each other inthe types and density of neurons and in the orga-nization of neuronal processes. Moreover, it isorganized in morphologically- and functionally-

specific tangential areas. On a lower scale, verti-cally-oriented, functionally-defined columns canbe identified, even if their anatomical morphologyis less sharply understood than their physiologicalidentity.

Pyramidal neurons represent the great majorityof cortical neurons (�70%): their cell bodies arelocated through all cortical layers excepted I, es-pecially in layers II/III and V, and their apicaldendrites are usually vertically oriented towardsthe cortical surface. Apical dendrites from pyra-

* Corresponding author. Tel.: +39-011-6707-700; fax: +39-011-6707-732.

E-mail address: alessandro.vercelli@unito.it (A. Vercelli).

0047-6374/02/$ - see front matter © 2002 Elsevier Science Ireland Ltd. All rights reserved.

PII: S 0 0 47 -6374 (01 )00357 -8

R. Curtetti et al. / Mechanisms of Ageing and De�elopment 123 (2002) 473–479474

midal neurons tend to join to each other to formbundles ascending to the supragranular layers.These bundles are heterogeneous in the number oftheir apical dendrites and in the laminar locationof their parent neurons and are differently dis-tributed through cortical areas. Dendritic bundlesare found early in development, when their den-drites are electrically coupled by gap junctions.Their function within cortical physiology and cor-tical columns is, as yet, not well understood. Theirmorphology and quantitative aspects in the adultmammal have been extensively studied by Peterset al. (Peters et al., 1985; Peters and Kara, 1987;Peters and Payne, 1993).

In this paper, we describe the postnatal devel-opment of dendritic bundles in the visual cortexof rats. We show that dendritic bundles are al-ready present at P3, when the cortical plate is stilldeveloping, that different types of bundles may bedescribed and that the distance among bundlesand among dendrites in a bundle increases pro-gressively with age.

2. Materials and methods

Wistar rats from our breeding colony werekilled by an overdose of anesthetics (chloral hy-drate i.p.) at P3, P10 or P30 and perfused throughthe left ventricle with saline followed by 4%paraformaldehyde in 0.1 M phosphate buffer (PB)at pH 7.4. Their brains were dissected from theskull and postfixed for 2 h. The brains to be usedfor immunohistochemistry were immersed in 30%sucrose in PB overnight, cut on the cryostat in60-�m thick coronal or tangential sections andreacted free floating with the AP18 antibody (akind gift from BM Riederer, IBCM, Lausanne;Riederer et al., 1995) against MAP2 (Binder et al.,1984). Briefly, sections were incubated for 15 minin 3% fetal calf serum in PB, washed in PB andthen incubated o.n. with the primary antibody(1:100); the following day, after washing, the sec-tions were incubated 1 h in the link antibody(rabbit anti-mouse IgG, 1:75), washed and incu-bated for 4 h in mouse peroxidase-anti-peroxidasecomplex (1:75). Peroxidase was revealed using4-chloro-1-naphthol as a cromogen (0.05%,

Riederer et al., 1995). Incubation times were re-duced for P3 brains, in order to avoid damage tothe sections.

The brains to be used for neural tracing re-ceived one small crystal of DiI (Godement et al.,1987) in the white matter underlying areas 17/18border. The brains were kept in the fixative in theoven at 37 °C, for variable periods depending onthe age of the animal (�10 days for P3 brainsand 20 days for P10), after which they were cut onthe vibratome in 150-�m thick coronal or tangen-tial sections.

Since DiI is not efficient in tracing pathways inmature brains in vitro (Vercelli et al., 2000), toexamine the morphology of dendritic bundles inadult (P30) animals, we used coronal sections ofvisual cortex which were meant for other studies:pyramidal neurons had been retrogradely labeledwith an injection of 5% biotinylated dextranamine (Sigma, MO) in PB (80–100 �l) in thewhite matter underlying area 18. Two weeks fol-lowing the injection of the tracer, the animalswere perfused with a rinse of PB followed by 1%paraformaldehyde (PAF) and 2.5% gluteralde-hyde in PB. The brains were dissected out of theskulls, postfixed 4 h in the same fixative andinfiltrated overnight in 30% sucrose. Some 50-�mthick coronal sections were cut at the cryostat andincubated free-floating overnight at room temper-ature in an avidin-biotin–HRP complex (ABCKit Elite, Vector Labs., Burlingame, CA). TheHRP was revealed using 0.05% 3-3�-diaminoben-zidine (DAB, Sigma) as a chromogen, intensifiedwith 0.2% Nickel ammonium sulfate (Adams,1980).

Coronal sections labeled with DiI or BDA, orimmunoreacted for MAP2 were used to describethe morphology of pyramidal neurons and thedistribution of their apical dendrites through thelayers. Moreover, coronal sections immunore-acted for MAP2 were used to determine the thick-ness of apical dendrites in each cortical layer.Tangential sections were used to map apical den-drites at the middle of layer II/III: apical den-drites (recognized according to their cross-sectional diameter) were mapped with the pro-gram Neurolucida (Glaser and Glaser, 1990) in anarea of at least 0.13 mm2.

R. Curtetti et al. / Mechanisms of Ageing and De�elopment 123 (2002) 473–479 475

3. Results

3.1. Morphology of dendritic bundles in coronalsections

DiI inserted in the underlying white matterlabeled pyramidal neurons through all corticallayers, with a decreasing tangential density de-pending on the distance from the site of insertion.The number of labeled neurons was higher inlayers IV–V than in the cortical plate and insupragranular layers, both at P3 and at P10,respectively. The apical dendrites of these neuronsformed different types of bundles, depending onthe age and on the layer considered. At P3 (Fig.1A) we could find (a) bundles formed by neuronslocated in cortical plate (i.e. developing layersIV–II); (b) bundles formed by layer V neurons(either large lower layer V neurons or small-to-medium sized upper layer V neurons); and (c)bundles formed by small upper layer V neuronsand cortical plate neurons. At P10 (Fig. 2), theamount of supragranular neurons participating inthe bundles increased: bundles formed by layer Vneurons were similar to those observed in youngeranimals and bundles formed by supragranularneurons exclusively or in association with upper

layer V neurons could be observed. Retrogradetransport of BDA in adult rats (Fig. 1B) gavesimilar results to those obtained with DiI in P10rats.

In sections immunoreacted for MAP2, apicaldendrites were clearly identifiable and formedbundles at the earliest stages considered, P3. Inthe P3 animals, in the coronal section layers Vand VI were clearly defined, whereas upper layersconsisted of the cortical plate. Apical dendritesoriginating from infragranular layers entered thecortical plate, where some apical dendrites of thelocal pyramidal neurons joined some bundles.Neurons were densely packed to each others andthe neuropil was poorly developed, such that den-dritic bundles were very close to each other, alsoin tangential sections cut at the level of corticalplate. At later ages, P10, all cortical layers werealready formed and several different types of bun-dles were recognizable on coronal sections. Intangential sections cut through supragranular lay-ers (Figs. 3 and 4), the first type consisted ofmostly thick dendrites, with a few thin dendritesand the second was made of both medium sizeand thin dendrites. In coronal sections, these twotypes of bundles corresponded to dendritic bun-dles of different dendritic composition. The first

Fig. 1. Dendritic bundles in coronal sections of the visual cortex at P3 (A), labeled retrogradely by DiI crystals inserted in the whitematter underlying the cortex and at P30 (B), retrogradely labeled by BDA injected in the cortical white matter. Scale bar=80 (A)and 100 (B) �m.

R. Curtetti et al. / Mechanisms of Ageing and De�elopment 123 (2002) 473–479476

Fig. 2. Computer-reconstructed map of dendritic bundles in coronal section of visual cortex of a P10 rat. In the insert, the entireslice is shown. Scale bar=200 �m (1000 �m in the insert). Arrowhead points to a bundle formed by neurons located insupragranular layers; open arrow to a bundle formed by infragranular layer neurons; and black arrow to a bundle formed by bothinfra- and supragranular layer neurons.

type was formed by layer V neurons and consistedof apical dendrites that extend to supragranularlayers without any relationship with apical den-drites arising from layer II/III cells. The secondtype was formed by the juxtaposition of apicaldendrites originating in layer V (thinner than theprevious ones) and in layer II/III. The third typewas formed by dendrites of neurons located insupragranular layers. Moreover, in infragranularlayers, dendritic bundles formed by a few layer VIneurons appeared to reach layer IV. The distanceamong bundles was clearly increased when com-pared to P3 animals. In adult rats, the morphol-ogy and distribution of dendritic bundles wassimilar to P10, except for a further increase in thedistance among bundles.

3.2. Thickness of apical dendrites in coronalsections

The thickness of apical dendrites was evaluatedin coronal MAP2 immunostained sections in or-der to distinguish apical from basal dendrites andfrom their oblique side branches. Quantitativedata were obtained from sections through corticalplate in P3 rats and through layer III in P10 and

P30. The average cross-sectional diameter of api-cal dendrites was 1.11�0.14 �m at P3, 1.34�0.19 �m at P10 and 1.31�0.18 �m at P30.

4. Discussion

These findings confirm that dendritic bundlesare present in the rat visual cortex early in devel-opment and that they are formed by neuronsbelonging to different cortical layers. The exis-tence of different types of bundles relative to thelayer of location of their parent neurons suggeststhat they are heterogeneous from each other innature and in the pattern of connectivity (bothafferent and efferent).

Dendritic bundles in cerebral cortex are formedby apical dendrites of pyramidal neurons. Wehave shown here that these dendritic bundles areof several different morphological types. Theymay represent morphological units underlying theorganization of cortical columns (Mountcastle,1997).

The heterogeneity of dendritic bundles stemsfrom the heterogeneity of pyramidal neurons. Infact, pyramidal neurons in cerebral cortex consist

R. Curtetti et al. / Mechanisms of Ageing and De�elopment 123 (2002) 473–479 477

of several different cell types, which are character-ized by their position, dendritic morphology andpattern of axonal projection. Associated withthese morphological aspects are different physio-logical properties. Moreover, pyramidal neuronslocated in different cortical layers differ in theirbirthdates.

Pyramidal neurons are found in all corticalareas and layers, excepted for layer I (Cajal, 1894;Lorente de No, 1938). Within a given layer, theyare located at different depths and show differentmorphological types on the grounds of cell sizeand characteristics of the dendritic arbors. Forexample, in layer V of rat visual cortex (area 17),corticocollicular pyramidal neurons are locatedmore deeply, they have more basal dendrites andlarger cell bodies than callosal neurons and bearprominent apical dendrites which always formwide apical tufts in layer I, whereas apical den-drites of callosally-projecting layer V pyramids

Fig. 4. Tangential maps through layer III of the visual cortexat P3 (A), P10 (B) and P30 (C). Each point corresponds to anapical dendrite transversally cut on the plane of section, circlescorrespond to neuronal somata (not shown in C). Scale bar=20 �m.

Fig. 3. Dendritic bundles in tangential sections of visual cortexat P10 and P30, put in evidence by MAP2 immunohistochem-istry. Scale bar=50 �m.

often end in layers II/III (Hallman et al., 1988;Hubener and Bolz, 1988; Koester and O’Leary,1992, 1993; Kasper et al., 1994a; Rumberger etal., 1998). Morphological differences correspondto distinct physiological characteristics (Kasper et

R. Curtetti et al. / Mechanisms of Ageing and De�elopment 123 (2002) 473–479478

al., 1994b; Rumberger et al., 1998), developingafter pathway selection and maturation of theapical dendrite (Kasper et al., 1994c).

Bundles consist of apical dendrites originatingfrom neurons of the same layer or dendrites mayjoin to the bundle both from infra- and supra-granular layers. In some cases, their distribution issuch to avoid the three-dimensional arrays of cellsin layer IV, for example in the mouse barrelcortex where they seem to avoid barrel hollows(Detzer, 1976). Moreover, there is an extremevariability in the number of dendrites in a bundledepending on the layer and on the area which isexamined and, consequently, in its size: bundlesmay extend to 100 �m in diameter and 200 apicaldendrites, such as in the retrosplenial cortex of therat (Wyss et al., 1990), so many to oblige theapical dendrites to angle 45° to join the bundle.

Here we have described three main types ofdendritic bundles in the visual cortex of the rat,one originated exclusively by neurons in layer V,another by neurons located in layer V and II/IIIand a last one formed by neurons located insupragranular layers. Moreover, we have founddendritic bundles at the earliest ages considered,i.e. at P3; at this age, only a few neurons in thecortical plate participated to dendritic bundles,whereas supragranular pyramidal neurons joinedfrequently dendritic bundles from layer V neuronsat P10.

Analysis of the distribution of apical dendritesin tangential sections, immunostained for MAP2,showed that apical dendrites are clustered to-gether at all ages considered. The distance amongdendritic bundles is, in general, lower in supra-granular layers at all ages considered and in-creases with age. Also, the distance amongdendrites in a cluster increases with age: thissuggest that electrical coupling, operated throughgap junctions in dendritic bundles at early stagesof development, might be less common in adultanimals, thus changing the physiological proper-ties of the neurons in the same bundle.

Therefore, we conclude that dendritic bundlesin the cerebral cortex are morphological entitieswhich are highly conserved through species(Schmolke, 1989; Peters et al., 1997; Schmolkeand Kunzle, 1997), through cortical areas (White

and Peters, 1993; Gabbott and Bacon, 1996; Levand White, 1997) and with age (Hirst et al., 1991).Even if their function in cortical physiology re-mains unclear, their highly organized distributionand their early electrical coupling (Lo Turco andKriegstein, 1991; Yuste et al., 1992; Peinado et al.,1993) are suggestive of a functional role withincortical columns and, finally, in cortical activity.

Acknowledgements

The authors are grateful to BM Riederer forthe generous gift of the AP18 antibody. Sup-ported by funds from Compagnia di San Paoloand CRT foundation to AV. RC is a recipientof a fellowship from Cavalieri OttolenghiFoundation.

References

Adams, J.C., 1980. Stabilizing and rapid thionin staining ofTMB-based HRP reaction product. Neurosci. Lett. 17,7–9.

Binder, L.I., Frankfurter, A., Kim, H., Caceres, A., Payne,M.R., Rebhun, L.I., 1984. Heterogeneity of microtubule-associated protein 2 during rat brain development. Proc.Natl. Acad. Sci. USA 81, 5613–5617.

Cajal, S.R., 1894. Les Nouvelles Idees sur la Structure duSysteme Nerveux chez l’Homme et chez les Vertebres.Reinwald, Paris.

Detzer, K., 1976. Course and distribution of apical dendritesof layer V pyramids in the barrel field and area parietalisof the mouse. Anat. Embryol. 149, 251–258.

Gabbott, P.L.A., Bacon, S.J., 1996. The organization of den-dritic bundles in the prelimbic cortex (area 32) of the rat.Brain Res. 730, 75–86.

Glaser, J.R., Glaser, E.M., 1990. Neuron imaging with Neu-rolucida—a PC-based system for image combining mi-croscopy. Comp. Med. Imag. Graph. 14, 307–317.

Godement, P., Vanselow, J., Thanos, S., Bonhoeffer, F., 1987.A study in developing visual systems with a new method ofstaining neurones and their processes in fixed tissue. Devel-opment 101, 697–713.

Hallman, L.E., Schofield, B.R., Lin, C.S., 1988. Dendriticmorphology and axon collaterals of corticotectal, cortico-pontine, and callosal neurons in layer V of primary visualcortex of the hooded rat. J. Comp. Neurol. 272, 149–160.

Hirst, E., Asante, J., Price, J., 1991. Clustering of dendrites inthe cerebral cortex begins in the embrionic cortical plate. J.Neurocytol. 20, 431–438.

R. Curtetti et al. / Mechanisms of Ageing and De�elopment 123 (2002) 473–479 479

Hubener, M., Bolz, J., 1988. Morphology of identified projec-tion neurons in layer 5 of rat visual cortex. Neurosci. Lett.94, 76–81.

Kasper, E.M., Larkman, A.U., Lubke, J., Blakemore, C.,1994a. Pyramidal neurons in layer V of the rat visualcortex. I. Correlation between cell morphology, intrinsicelectrophysiological properties and axon targets. J. Comp.Neurol. 339, 459–474.

Kasper, E.M., Larkman, A.U., Lubke, J., Blakemore, C.,1994b. Pyramidal neurons in layer V of the rat visualcortex. II. Development of electrophysiological properties.J. Comp. Neurol. 339, 475–494.

Kasper, E.M., Lubke, J., Larkman, A.U., Blakemore, C.,1994c. Pyramidal neurons in layer V of the rat visualcortex. III. Differential maturation of axon targeting, den-dritic morphology, and electrophysiological properties. J.Comp. Neurol. 339, 495–518.

Koester, S.E., O’Leary, D.D.M., 1992. Functional classes ofcortical projection neurons develop dendritic distinctionsby class-specific sculpting of an early common pattern. J.Neurosci. 12, 1382–1393.

Koester, S.E., O’Leary, D.D.M., 1993. Connectional disten-tion between callosal and subcortically projection corticalneurons is determined prior to axon extension. Dev. Biol.160, 1–14.

Lev, D.L., White, E.L., 1997. Organization of pyramidal cellapical dendrites and composition of dendritic clusters inthe mouse: emphasis on primary motor cortex. Eur. J.Neurosci. 9, 280–290.

Lo Turco, J.J., Kriegstein, A.R., 1991. Clusters of coupledneuroblasts in embryonic neocortex. Science 252, 563–566.

Lorente de No, R., 1938. The cerebral cortex: architecture,intracortical connections and motor projections. In: Ful-ton, J.F. (Ed.), Physiology of the Nervous System. OxfordUniversity Press, New York, pp. 288–313.

Mountcastle, V.B., 1997. The columnar organization of theneocortex. Brain 120, 701–722.

Peinado, A., Yuste, R., Katz, L.C., 1993. Extensive dye cou-pling between rat neocortical neurons during the period ofcircuit formation. Neuron 10, 103–114.

Peters, A., Kara, D.A., 1987. The neuronal composition ofarea 17 of rat visual cortex. IV. The organization ofpyramidal cells. J. Comp. Neurol. 260, 573–590.

Peters, A., Payne, B.R., 1993. Numerical relationship be-tween geniculocortical afferents and pyramidal cell mod-ules in the cat primary visual cortex. Cereb. Cortex 3,69–78.

Peters, A., Kara, D.A., Harriman, K.M., 1985. The neuronalcomposition of area 17 of rat visual cortex. III. Numericalconsiderations. J. Comp. Neurol. 238, 263–274.

Peters, A., Cifuentes, M.J., Sethares, C., 1997. The organiza-tion of pyramidal cells in area 18 of the Rhesus monkey.Cereb. Cortex 7, 405–421.

Riederer, B.M., Draberova, E., Viklicky, V., Draber, P., 1995.Changes of MAP2 phosphorylation during brain develop-ment. J. Histochem. Cytochem. 43, 1269–1284.

Rumberger, A., Schmidt, M., Lohmann, H., Hoffmann, K.P.,1998. Correlation of electrophysiology, morphology,and functions in corticotectal and corticopretectal pro-jection neurons in rat visual cortex. Exp. Brain Res. 119,375–390.

Schmolke, C., 1989. The ontogeny of dendrite bundles inrabbit visual cortex. Anat. Embryol. 180, 371–381.

Schmolke, C., Kunzle, H., 1997. On the presence of dendritebundles in the cerebral cortex of the Nadagascan lesserhedgehog tenrec and the red-eared pond turtle. Anat.Embryol. 196, 195–213.

Vercelli, A., Repici, M., Garbossa, D., Grimaldi, A., 2000.Recent techniques for tracing pathways in the centralnervous system of developing and adult mammals. BrainRes. Bull. 51, 11–28.

White, E.L., Peters, A., 1993. Cortical modules in the postero-medial barrel subfield (Sm1) of the mouse. J. Comp.Neurol. 334, 86–96.

Wyss, J.M., van Groen, T., Sripanidkulchai, K., 1990. Den-dritic bundles in layer I of granular retrosplenial cortex:intracellular labeling and selectivity of innvervation. J.Comp. Neurol. 295, 33–42.

Yuste, R., Peinado, A., Katz, L.C., 1992. Neuronal domainsin developing neocortex. Science 257, 665–669.