European Journal of Neuroscience, Vol. 19, pp. 2039±2046, 2004
ß Federation of European Neuroscience Societies
Neurons in the corpus callosum of the cat during
postnatal development
Beat M. Riederer,1,2 Pere Berbel3 and Giorgio M. Innocenti4
1
Institut de Biologie Cellulaire et de Morphologie, Universite de Lausanne, Rue du Bugnon 9, 1005 Lausanne, Switzerland
Centre des Neurosciences Psychiatriques, Site de CERY, 1008 Prilly, Switzerland
3
Instituto de Neurosciencias, Universidad Miguel HernaÂndez-CSIC, Apartado de Correos 18, 03550 San Juan (Alicante), Spain
4
Karolinska Institutet, Laboratory of Neuroanatomy and Brain Development, Department of Neuroscience, 17177 Stockholm, Sweden
2
Keywords: cat, corpus callosum, cytoskeleton, development, MAP2, neuron
Abstract
The corpus callosum (CC) is a major telencephalic commissure containing mainly cortico-cortical axons and glial cells. We have
identi®ed neurons in the CC of the cat and quanti®ed their number at different postnatal ages. An antibody against microtubuleassociated protein 2 was used as a marker of neurons. Immunocytochemical double-labelling with neuron-speci®c enolase or gaminobutyric acid antibodies in the absence of glial ®brillary acidic protein positivity con®rmed the neuronal phenotype of these cells.
CC neurons were also stained with anti-calbindin and anti-calretinin antibodies, typical for interneurons, and with an anti-neuro®lament
antibody, which in neocortex detects pyramidal neurons. Together, these ®ndings suggest that the CC contains a mixed population of
neuronal types. The quanti®cation was corrected for double counting of adjacent sections and volume changes during CC development. Our data show that CC neurons are numerous early postnatally, and their number decreases with age. At birth, about 570
neurons are found within the CC boundaries and their number drops to about 200 in the adult. The distribution of the neurons within the
CC also changes in development. Initially, many neurons are found throughout the CC, while at later ages they become restricted to the
boundaries of the CC, and in the adult to the rostrum of the CC close to the septum pellucidum or to the indusium griseum. Although
origin and function of transient CC neurons in development and in adulthood remain unknown, they are likely to be interstitial neurons.
Some of them have well-developed and differentiated processes and resemble pyramidal cells or interneurons. An axon-guiding
function during the early postnatal period can not be excluded.
Introduction
The corpus callosum (CC) is a major telencephalic commissure with
well-de®ned boundaries containing axons that interconnect neurons
in the two cerebral hemispheres (reviewed by Innocenti, 1986). In
addition, in the mature CC there is an important number of glial cells
(in particular astrocytes and oligodendrocytes). However, neurons at
the periphery of the CC were described, whose dendrites intermingle
with callosal axons (Molobabic et al., 1984), suggesting that the CC,
as other regions of the cortical white matter, might contain a
permanent population of interstitial neurons (Kostovic & Rakic,
1980).
Microtubule-associated proteins are known to be intimately
involved in the differentiation of neurons and the growth of axons
and dendrites (reviewed by Matus, 1988; Riederer, 1990). Microtubule-associated protein 2 (MAP2) was already found, at different
postnatal ages, in the earliest generated neurons of the cortical subplate
(Chun & Shatz, 1989a). The high molecular weight forms of MAP2
(MAP2a and b, of 280 and 260 kDa, respectively) are con®ned to
somata and dendrites (Matus et al., 1981; Burgoyne & Cumming,
1984; Caceres et al., 1984; De Camilli et al., 1984). The occurrence of
a low molecular weight form, MAP2c, had been reported (Riederer &
Matus, 1985). The subcellular distribution and particularities of both
isoforms have been described, with a presence also in axons or glial
cells (Tucker et al., 1988; Doll et al., 1993; Sha®t-Zagardo et al.,
2000). In kittens, we have previously observed MAP2-positive processes, often with many varicosities, and occasional somata in the
subcortical white matter, including the CC (Riederer & Innocenti,
1992). These observations strongly suggested that the developing CC
might also contain a population of transient neurons.
Therefore, in the present paper our aim has been to quantify and
characterize the population of neurons in the CC during development
and in the adult animal. We have used the monoclonal antibody AP14
against the high molecular weight forms of MAP2 because it is a good
neuronal marker (Binder et al., 1984; Riederer & Innocenti, 1992;
Riederer et al., 1995). In addition, we used a panel of antibodies
against other proteins, speci®c either for glia or for neuronal subpopulations or neurotransmitters.
Materials and methods
Tissue preparation
Correspondence: Dr B. M. Riederer, as above.1
E-mail: BeatMichel.Riederer@ibcm.unil.ch
Received 28 April 2003, revised 6 February 2004, accepted 10 February 2004
doi:10.1111/j.1460-9568.2004.03305.x
Kittens at birth (postnatal day 0; P0), P1, P3, P11, P19, P28, P39,
P45 and adult cats over 1 year were obtained by timed pregnancies from a breeding colony. For immunohistochemical analysis,
animals were deeply anaesthetized with Nembutal (50 mg/kg) and
2040 B. M. Riederer et al.
perfused for 20 min with 4% paraformaldehyde in phosphatebuffered saline (5 mM phosphate buffer, pH 7.4 and 0.9% sodium
chloride). The two hemispheres (with intact CC) were post®xed for
6 h in the same ®xative used for the perfusion and kept in phosphatebuffered saline with 0.05% Na-azide at 4 8C. Most cerebral hemispheres were cut sagittally, while only a few were cut coronally.
Medial, coronal and parasagittal sections of 50 mm thickness were cut
with a vibratome and processed for immunohistochemistry (see
below).
Electrophoresis and immunoblots
Fig. 1. Western blot of tissue of kittens at postnatal day 3 (P3), P11, P19, P28
and adult animals of the corpus callosum (CC) and P3 and adult animals of the
visual cortex (CX), stained for MAP2 with monoclone AP14. Note that only
little MAP2 is found in white matter tissue. The location of the high molecular
weight form of MAP2b is indicated to the left.
Tissue from the CC was dissected as previously described (Riederer &
Innocenti, 1992). Proteins (50 mg/slot) were separated on a 3.6±15%
sodium dodecyl sulphate±polyacrylamide gel electrophoresis according to Laemmli (1970). Gels were either stained with Coomassie Blue
or electrically transferred to nitrocellulose sheets (Towbin et al., 1979)
and stained with monoclonal antibody AP14 for MAP2a and 2b
(Binder et al., 1984).
Fig. 2. Immunohistochemical location of MAP2-positive cells in sagittal sections close to the midline of corpus callosum (CC) in kitten and adult. The regions shown
are the trunk (A, F and G) and the genu (B±E) of the CC. The samples were taken at P1 (A and C), P11 (F), P39 (D and G) and adult (B, E and H). The dashed line
indicates the border between CC and fornix (FX). The magni®cation bars in A and B are 100 mm and pertain to F; the bar in E is 50 mm and pertains to C, D, G and H.
ß 2004 Federation of European Neuroscience Societies, European Journal of Neuroscience, 19, 2039±2046
Neurons in the corpus callosum of the cat
Immunohistochemistry
Floating sagittal sections of whole series of CC from one hemisphere
were incubated overnight with anti-MAP2 (clone AP14, hybridoma
supernatant, l : 10) or alternate serial CC sections were exposed to one
of the following primary monoclonal antibodies: anti-MAP2 (clone
AP14, hybridoma supernatant, l : 10), anti-medium-sized neuro®lament
subunit (clone FNP7, 1 : 1000), anti-calbindin (1 : 1000), anti-calretinin
(1 : 1000) and anti-vimentin (Boehringer; 1 : 100), and subsequently
incubated for 2 h either with 2% rhodamine-conjugated or 1% peroxidase-conjugated rabbit anti-mouse antibodies (Dako), the last
further reacted with 4-chloro-1-naphthol (Riederer & Innocenti,
1992). Non-speci®c binding was blocked with 3% foetal calf serum
in Tris-buffered saline.
Double-labelling was performed as described before (Riederer et al.,
1990) in coronal sections using monoclonal AP14 against MAP2a and
2b and then one of the following polyclonal antibodies: anti-neuronspeci®c enolase (NSE, Polyscience, Warrington, PA, USA; 1 : 100),
anti-cow glial ®brillary acidic protein (GFAP; Dako: 1 : 200), anti-gaminobutyric acid (GABA, Sera Laboratory; 1 : 100; Romijn et al.,
1992) and the monoclone M22 against NF-M (1 : 10). After, sections
were incubated with 2% rhodamine- and 2% ¯uorescein-conjugated
secondary antibodies. Sections were mounted and coverslipped in a
semisolid mounting medium (Lenette, 1978). Monoclonal antibodies
anti-MAP2 was generously provided by Dr Lester (Skip) Binder
(Birmingham, AL, USA), anti-calbindin and anti-calretinin were gifts
of Dr M. Celio (Fribourg, Switzerland), for references see Schwaller
(1996) and Hunziker (1996). The use of antibodies against calciumbinding proteins as markers for neurons has been reviewed in detail
(Andressen et al., 1993). The monoclonal antibody FNP7 against NFM was a gift of Dr V.Y.-M. Lee (Philadelphia, USA). This antibody
2041
recognizes exclusively pyramidal cells in the cerebral cortex (Hornung
& Riederer, 1999). Another antibody against NF-M, the M22, was
used to show a general neuro®lament distribution (Riederer et al.,
1996).
Quantification
The MAP2-positive neurons in the CC were drawn using a camera
lucida or a computer-assisted microscope (Neurolucida, MicroBrightField, Colchester, VT, USA) and counted in serial sagittal sections of
three animals per age group, except at P39 for which only two animals
were available. To obtain the total number of neurons per half of CC,
serial sections from the callosal midline up to the point where the trunk
of CC fuses with the cingulate cortex were used. Only cell bodies were
counted. The number of sections used for counting MAP2-positive
neurons in half of CC was 16 sections at P1, 18 sections at P11, 20
sections at P19, 22 sections at P29, 28 sections at P39 and 34 sections
in the adult. The number of neurons in each section was correct for
double counting in adjacent sections with the Abercrombie correction
factor, calculated (according to Clarke, 1993) to be 0.70, considering
that the section thickness was 50 mm and the average cell diameter was
20.9 4.6 mm (n ˆ 10 neurons). The number of neurons per half CC
was multiplied by two to obtain the total neuron number. In order to
correct for the volumetric increase of the CC and to test for a dilution
effect during development, a volume correction factor was also
applied. This was obtained from ®g. 6 of Fleischhauer & SchluÈter
(1970) by dividing the volume of each speci®c age with volume at P1.
Thus, the values were at P0±P11 ˆ 1.0; at P19 ˆ 1.25; at P28 ˆ 1.4; at
P39 ˆ 1.7; at P45 ˆ 2.1; and adult ˆ 3.2. However, the correction did
not alter the result, which was that a dramatic decrease in the number
of neurons in the CC occurred, and the corrected values were not
included in the results.
Fig. 3. Double-labelling of coronal sections of CC tissue with anti-MAP2 (monoclonal AP14) and a variety of polyclonal antibodies, followed by ¯uorescein and
rhodamine-conjugated secondary antibodies. MAP2 labelling (A and C) is compared with anti-NSE staining (B and D). Anti-MAP2 staining (E and G) is compared
with anti-GABA labelling (F and H). CC tissue was taken at P5 (E and F), P19 (A and B), P45 (G and H) and adult (C and D). The arrow in G and H points to a neuron
that contains MAP2 but lacks GABA. Scale bars, 40 mm.
ß 2004 Federation of European Neuroscience Societies, European Journal of Neuroscience, 19, 2039±2046
2042 B. M. Riederer et al.
Results
Identification of MAP2b-positive neurons in developing CC
On immunoblots, the AP14 antibody reacted intensely with MAP2b in
P3 and adult cat cortex (Fig. 1). Less intense MAP2b staining was
found in the CC at all ages. This was surprising because MAP2b
localizes in the somatodendritic part of neurons and a location in white
matter is therefore unusual. Therefore, the weak staining in CC blots
suggested the presence of a few neuronal cell bodies.
Immunostained sections of the CC revealed the presence of MAP2positive cells at all studied ages (Fig. 2). More cells and processes were
stained in the developing CC (Fig. 2A), fewer in the adult (Fig. 2B). At
Fig. 4. Double-labelling of a coronal corpus callosum (CC) section of a newborn kitten with monoclonal anti-microtubule-associated protein 2 (MAP2; A, C and E)
and polyclonal anti-glial ®brillary acidic protein (GFAP; B, D and F). The location of the area of interest is indicated with an arrow in all panels. At low magni®cation
the glial scaffold is seen at the lateral border of the CC (B). At medium (C and D) and better visible at higher magni®cation (E and F) no overlap of glial staining in
MAP2-positive cells was seen. CCX, cingulated cortex; E, ependyma; RCC, radiatio corpus callosum. The asterisk indicates the end of the CCX. This point indicates
the transition zone of the CC into its radiation, according to Fleischauer & SchluÈter (1970). Scale bar, 100 mm (B); 30 mm (D and F).
ß 2004 Federation of European Neuroscience Societies, European Journal of Neuroscience, 19, 2039±2046
Neurons in the corpus callosum of the cat
all ages, the morphological characteristics of MAP2-positive cells
were very variable; some cells exhibited short and poorly rami®ed
processes, while others had well developed ones (Fig. 2C±E). Different
morphological types were observed, pyramidal-like neurons with
impressive dendrites and rami®cations (Fig. 2F), interneuron-like cells
with ®ne and varicose processes (Fig. 2G), and bipolar shaped neurons
resembling migrating neurons (Fig. 2H).
The neuronal phenotype of these MAP2 cells was demonstrated by
double-labelling experiments with NSE and GABA. Coronal sections
of CC at various ages were stained with AP14 (Fig. 3A, C, E and G),
and double-labelled either for NSE (Fig. 3B and D) or GABA (Fig. 3F
and H). All MAP2-positive neurons were also positive for NSE but less
than half of them were positive for GABA at an early (P5) and a later
(P45) postnatal age, as indicated with an arrow in Fig. 3G and H. In
coronal sections of the CC, near the callosal radiation a particular
GFAP staining was observed, with glial ®bres forming a scaffold
between cingulated cortex and ependyma (Fig. 4A). Usually, MAP2and GFAP-positive processes ran in parallel trajectories but did not colocalize (Fig. 4C and D). This is better visualized at higher magni®cation (Fig. 4E and F).
Serial sagittal sections with anti-GFAP and monoclonal antibodies
M22 against NF-M (a marker for axons) and MAP2 showed that at
several locations, GFAP-positive processes (Fig. 5B) had similar trajectories to the NF-M-positive axons (Fig. 5C), while MAP2-positive
processes showed more dispersed trajectories (Fig. 5D). In peroxidaselabelled MAP2 sections, the processes were seen more clearly when
using polarized light than with conventional light microscopy; however, this was only possible in early postnatal stages with little
myelination (not shown). At all ages, most of the processes were
above or below the CC, at the border with the indusium griseum and
the fornix. These processes, however, penetrated the CC over a short
distance ( 50 mm or less). Instead, the neurons whose body was
2043
located in the CC established a rich dendritic arbor within it. In the
genu of the CC, near the septum pellucidum, some MAP2 processes
arising from neurons located in the border of the CC and remained in
the adult while they disappeared in other regions of the CC.
The neurons of the CC were further characterized with antibodies
against different neuronal subtypes. The FNP7 antibody against an
epitope of NF-M, which labels exclusively pyramidal cells in the
cerebral cortex (Hornung & Riederer, 1999), also marked several
neurons in the CC. They had their soma in the CC, from which arose
two±three primary dendrites that occasionally gave off secondary
branches. Their axon was occasionally followed into the fornix
(Fig. 6B). This suggests that some of the neurons of the CC are
pyramidal cells somewhat modi®ed in their morphology, as they
are often found in the deep cortical layers and in the subcortical white
matter. Antibodies against calretinin and calbindin (Fig. 6C and D)
labelled spindle-shaped cell bodies and localized at the upper and
lower borders of the CC, from which one or two poorly developed
primary dendrites emerged.
The distribution of MAP2-positive neurons at different ages is
shown in Fig. 7. The density of neurons clearly decreased with age,
and in the adult the few remaining neurons were mainly concentrated
in the rostrum, particularly near the border with the septum pellucidum. Corrected quantitative measurements for double cell counting
and CC growth (see Materials and methods) showed that the total
number of neurons in the CC remains fairly stable during the ®rst
postnatal month (on average 585 80 neurons). In Fig. 8, the slight
and non-statistically signi®cant (P > 0.05) decrease in the number of
cells per section during the same period is most probably an artefact,
due to the increased volume of the CC (see also Fleischhauer &
SchluÈter, 1970). However, this number dropped signi®cantly
(P < 0.001) in the adult CC where fewer neurons were counted (on
average 204 60 neurons).
Fig. 5. (A) A low magni®cation of the distribution of glial ®brillary acidic protein (GFAP; top), medium-sized neuro®lament subunit (NF-M; monoclone M22;
middle) and microtubule-associated protein 2b (MAP2b; monoclone AP14; bottom) in serial, sagittal sections of a P19 kitten CC and surrounding structures
visualized with peroxidase-conjugated secondary antibodies and 4-chloro-1-naphthol. Parts of the fornix (FX) and hippocampus (HI) are also shown. The square
indicates the area shown at higher magni®cation in B, the views in C and D correspond to similar areas, although not squared. Note the different staining patterns with
GFAP-, NF-M- and MAP2-positive ®bres in®ltrating the CC. The arrow points to an area were often MAP2-positive cells extend their processes throughout the
thickness of the callosal trunk from the indusium griseum to the fornix. GE, genu; TR, trunk; SP, splenium. Scale bar, 100 mm (B±D).
ß 2004 Federation of European Neuroscience Societies, European Journal of Neuroscience, 19, 2039±2046
2044 B. M. Riederer et al.
Fig. 6. Immunohistochemical staining of P19 kitten corpus callosum (CC) with anti-microtubule-associated protein 2 (MAP2; A), anti-medium-sized neuro®lament
subunit (NF-M; monoclone FNP7), which marks pyramidal cells in the cerebral cortex (B), anti-calretinin (arrow in C) and anti-calbindin (arrow in D), which mark
interneurons. They were visualized by peroxidase-conjugated secondary antibodies and 4-chloro-1-naphthol. Note the axon of the NF-M-positive neuron passing
towards the fornix (FX). Interneuron-like cells align rather at the borders of the CC and resemble migrating neurons. Borders of the CC are indicated with dashed lines
(B±D). Scale bar, 50 mm.
Discussion
CC neurons in development and adulthood
MAP2 is a good neuronal marker and has been used to identify
neuronal morphology (reviewed by Matus, 1988), neuronal development in the cat visual cortex (Riederer & Innocenti, 1992) or to
characterize the neuronal nature of cells found in subcortical white
matter (Chun & Shatz, 1989a). Here we have used a well-characterized
MAP2 antibody to identify and quantify neurons during postnatal
development of cat corpus callosum. The detection by Western blots of
minute MAP2 quantities in the CC indicated the presence of somatodendritic neuronal components in this structure.
Indeed, MAP2-positive cells were subsequently identi®ed by immunohistochemistry. The MAP2-positive cells were not double-labelled
with GFAP, which excludes their astrocytic nature. Instead, their
neuronal nature was con®rmed with an antibody against neuro®laments and by double-labelling with antibodies against NSE and
GABA. CC neurons are a mixed population with different morphological characteristics. Some exhibit pyramidal-like features, with
elaborated, thick and long dendrites, others have interneuron-like
morphologies with ®ne and varicose dendrites, still others have bipolar
shape and resemble migrating neurons.
The CC is a de®ned morphological structure that changes shape and
volume during development (Fleischhauer & SchluÈter, 1970). These
changes could render the quanti®cation of the total number of neurons
in the CC dif®cult. However, we found that in the developing brain
GFAP-stained ®bres form a scaffold at the junction between the CC
and the hemispheric wall. These ®bres allowed the identi®cation of the
lateral border of the CC, and therefore its volume. In addition we have
applied a correction factor to avoid double counting of neurons in
adjacent sections (see Materials and methods).
It was found that the number of CC neurons remained stable during
the ®rst postnatal month, at about 580 neurons, while the volume of the
CC increased and the neuron number dropped to about 200 cells in the
adult.
The origin, role and fate of CC neurons
Questions regarding the origin, function or fate of CC neurons were
beyond the scope of this paper and remain open for speculations. These
are based on the similarity between the neurons of the CC and those in
other compartments of the subcortical white matter. Additional hints
come from the morphology of the neurons and their changing number
in development.
The occurrence of interstitial neurons in white matter has been
known for a long time (Kostovic & Rakic, 1980). NADPH diaphorase
cells are evenly distributed throughout the cat neocortex and some
were found also in the white matter and the CC (Kuchiiwa et al., 1994).
Unfortunately, their role and fate is still debated. Earliest-generated
ß 2004 Federation of European Neuroscience Societies, European Journal of Neuroscience, 19, 2039±2046
Neurons in the corpus callosum of the cat
Fig. 7. Distribution of MAP2-positive neurons shown in computer-assisted
drawings of representative sagittal CC sections near the midline from kittens
at ages postnatal day 1 (P1), P11 and P39, and an adult cat. Note the dramatic
decrease of labelled neurons in the splenium (SP) and body of the adult CC. In
the adult, almost all labelled neurons in the CC are in the ventral area of the
rostrum. FX, fornix; GE, genu; Spe, septum pellucidum; TR, trunk.
neurons of the cat cerebral cortex have also been characterized by
MAP2 and neurotransmitters during foetal life (Chun & Shatz, 1989a).
In addition to the radial migration of developing neurons, more
recently a tangential migration of GABAergic neurons with extracortical origin has been investigated (Zhu et al., 1999; MarõÂn &
Rubenstein, 2001; Tamamaki et al., 2003). It cannot be excluded
that the CC neurons might be part of the tangentially migrating
cohorts, travelling via the CC. Deng & Elberger (2001) also described
some cells in the CC of the mouse, which they interpreted as migrating
cells.
Another possibility is that the neurons of the CC are subplate
neurons. Subplate neurons, as CC neurons, are a partially transient
and heterogeneous neuronal population. They are found in the
Fig. 8. Numbers of MAP2-positive neurons counted for three series of sagittal
sections of CC [except two for postnatal day 39 (P39)]. The total number of
neurons counted for the whole CC at P1 and P39 was between 504 and 732
neurons and in the adult 204 neurons per CC. These numbers had been corrected
with the Abercrombie correction factor for double counting (scale to the right).
The difference between young and adult was signi®cant, P < 0.001 by Student's
t-test.
2045
subcortical white matter and are involved in the formation of thalamocortical connections (Allendoerfer & Shatz, 1994). Subplate neurons in the cat were also found to be positive for MAP2, GABA, as well
as for the neuropeptide Y and cholecystokinin (Chun et al., 1987; Chun
& Shatz 1989a; see also Allendoerfer & Shatz, 1994 for review). At
least some of the subplate neurons therefore appear to be inhibitory in
nature, although others might be excitatory, as based on the evidence
that they retrogradely transport 3H-aspartate (Antonini & Shatz, 1990).
Excitatory subplate neurons are generated early, at embryonic day 24±
30 (Luskin & Shatz, 1985), in the cat and follow a radial migration.
Unfortunately, the time of generation of the neurons of the CC is not
known.
The hypothesis that the neurons of the CC might be subplate neurons
implies that they might provide cues for axonal growth. This role has
also been ascribed to the neurons of the subplate (Shatz et al., 1988;
Aggoun-Zouaoui & Innocenti, 1994).
Axon number in the CC is highest at birth. The period of axonal
elimination extends over the following 20±40 days when the juvenile
cytoskeleton changes to the adult type (Riederer & Innocenti, 1991). It
is, however, surprising that the number of neurons in the CC decreases
well after 40 days. That seems to be considerably later than the
subplate neurons, most of which are eliminated during the ®rst
postnatal month (Chun & Shatz, 1989b). The mechanism by which
CC neurons disappear is unclear. It seems improbable though that they
might be migrating out of the CC, as no clear topographical displacement towards the periphery of the CC was noticed. However, this
possibility cannot be fully discarded, although it seems more plausible
that the neurons of the CC might disappear by neuronal death. This
awaits con®rmation with speci®c indicators of neuronal death, i.e.
assays for DNA fragmentation, caspase immunohistochemistry and/or
ultrastructural studies.
Other explanations for the presence of neurons in the CC are that
such neurons are part of neighbouring structures, such as the fornix or
septum pellucidum, as they are found close to their borders. In the case
of bipolar cells their processes have the same trajectory as those of the
cells observed in the fornix. CC neurons located close to the septum
may also establish connections through the genu of the CC into which
they extend their dendrites or even as far as the cingulate cortex which
sometimes they reach with their dendrites.
As to the function of the neurons of the CC, the positivity for GABA
indicates that some of them are inhibitory. In other species, GABAergic cells positive for glutamic acid decarboxylase have been identi®ed in the CC (Kaufman et al., 1986; DeDiego et al., 1994). Some CC
neurons contained calretinin or calbindin. In the neocortex, calbindin
is found in double bouquet neurons, and in Martinotti cells while
calretinin is found in bipolar and double-bouquet neurons (Andressen
et al., 1993). Other cells in the CC were labelled with an antibody
against an NF-M epitope which identi®es speci®cally pyramidal
neurons (Hornung & Riederer, 1999). Calbindin was also found in
a subpopulation of layer II/III pyramidal cells (Andressen et al., 1993).
Therefore, the neurons of the CC are neurochemically and probably
functionally a heterogeneous population. This heterogeneity is con®rmed by the morphology of the neurons, mentioned above. In
particular, some neurons exhibited large dendrites and resembled
pyramidal-like cells. These cells mostly disappear from the CC
towards adulthood and might be ectopic neurons, which failed to
complete migration.
In conclusion, we have documented the existence of neurons in the
CC of the cat in development and have provided their detailed
characterization. Concerning the role of these neurons several hypotheses are possible, based on their time of occurrence, morphology and
location.
ß 2004 Federation of European Neuroscience Societies, European Journal of Neuroscience, 19, 2039±2046
2046 B. M. Riederer et al.
Acknowledgements
We thank R. Porchet and I. Riederer for their excellent technical help. B.M.R.
was supported by grants of the Swiss National Science Foundation no. 3153725.98 and 3100-067201.01.
Abbreviations
CC, corpus callosum; GABA, g-aminobutyric acid; GFAP, glial ®brillary acidic
protein; MAP2, microtubule-associated protein 2; NF-M, medium-sized neuro®lament subunit; NSE, neuron-speci®c enolase; P, postnatal day.
References
Aggoun-Zouaoui, D. & Innocenti, G.M. (1994) Juvenile visual callosal axons in
kittens display origin and fate-related morphology and distribution of arbors.
Eur. J. Neurosci., 6, 1846±1863.
Allendoerfer, K.L. & Shatz, C. (1994) The subplate, a transient cortical
structure: its role in the development of connections between thalamus
and cortex. Annu. Rev. Neurosci., 17, 185±218.
Andressen, C.B.I., BluÈmcke, I. & Celio, M.R. (1993) Calcium-binding proteins:
selective markers of nerve cells. Cell Tissue Res., 271, 181±208.
Antonini, A. & Shatz, C.J. (1990) Relation between putative transmitter
phenotypes and connectivity of subplate neurons during cerebral cortical
development. Eur. J. Neurosci., 2, 744±761.
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.
Burgoyne, R.D. & Cumming, R. (1984) Ontogeny of microtubule-associated
protein 2 in rat cerebellum: differential expression of the doublet polypeptides. Neuroscience, 11, 157±167.
Caceres, A., Binder, L.I., Payne, M.R., Bender, P., Rebhun, L. & Steward,
O. (1984) Differential subcellular localization of tubulin and the microtubule-associated protein MAP2 in brain tissue as revealed by immunocytochemistry with monoclonal hybridoma antibodies. J. Neurosci., 4,
394±410.
Chun, J.J.M., Nakamura, M.J. & Shatz, C.J. (1987) Transient cells of the
developing mammalian telencephalon are peptide-immunoreactive neurons.
Nature, 325, 617±620.
Chun, J.M. & Shatz, C.J. (1989a) The earliest-generated neurons of the cat
cerebral cortex: characterization by MAP2 and neurotransmitter immunohistochemistry during fetal life. J. Neurosci., 9, 1648±1667.
Chun, J.J. & Shatz, C.J. (1989b) Interstitial cells of the adult neocortical white
matter are remnant of the early generated subplate neuron population. J.
Comp. Neurol., 282, 555±569.
Clarke, P.G.H. (1993) An unbiased correction factor for cell counts in histological sections. J. Neurosci. Meth., 49, 133±140.
De Camilli, P., Miller, P.E., Navone, F., Theurkauf, W.E. & Vallee, R.B. (1984)
Distribution of microtubule-associated protein 2 in the nervous system of the
rat studied by immuno¯uorescence. Neuroscience, 11, 817±846.
DeDiego, I., Smith-FernaÂndez, A. & FaireÂn, A. (1994) Cortical cells that
migrate beyond area boundaries: characterization of an early neuronal
population in the lower intermediate zone of prenatal rats. Eur. J. Neurosci.,
6, 983±997.
Deng, J. & Elberger, A.J. (2001) The role of pioneer neurons in the development of mouse visual cortex and corpus callosum. Anat. Embryol., 204,
437±453.
Doll, T., Meichsner, M., Riederer, B.M., Honegger, P. & Matus, A. (1993) An
isoform of microtubule-associated protein 2 (MAP2) containing four repeats
of the tubulin-binding motif. J. Cell Sci., 106, 633±640.
Fleischhauer, K. & SchluÈter, G. (1970) Ueber das postnatale Wachstum des
Corpus callosum der Katze (Felis domestica). Z. Anat. Entwick. Gesch., 132,
228±239.
Hornung, J.-P. & Riederer, B.M. (1999) Medium-sized neuro®lament protein
related to maturation of a subset of cortical neurons. J. Comp. Neurol., 414,
348±360.
Hunziker, W. (1996) Calbindin. In Celio, M.R., Pauls, T. & Schwaller, B.
(Eds), Guidebook to the Calcium-Binding Proteins. Oxford University
Press, Oxford, pp. 23±25.
Innocenti, G.M. (1986) General organization of callosal connections in the
cerebral cortex. In Peters, A. & Jones, E.G. (Eds), Cerebral Cortex, Vol. 5.
Plenum Press, New York, pp. 291±353.
Kaufman, D.L., McGinnis, J.F., Krieger, N.R. & Tobin, A.J. (1986) Brain
glutamate decarboxylase cloned in llgt-11: fusion protein produces gaminobutyric acid. Science, 232, 1138±1140.
Kostovic, I. & Rakic, P. (1980) Cytology and time of origin of interstitial
neurons in the white matter in infant and adult human and monkey telencephalon. J. Neurocytol., 9, 219±242.
Kuchiiwa, S., Kuchiiwa, T., Mori, S. & Nakagawa, S. (1994) NADPH diaphorase neurons are evenly distributed throughout cat neocortex irrespective of
functional specialization of each region. Neuroreport, 5, 1662±1664.
Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of
the head of bacteriophage T4. Nature, 227, 680±685.
Lenette, D.A. (1978) An improved mounting medium for immuno¯uorescence
microscopy. Am. J. Clin. Pathol., 69, 647±648.
Luskin, M.B. & Shatz, C.J. (1985) Studies of the earliest generated cells of the
cat's visual cortex: cogeneration of subplate and marginal zones. J. Neurosci.,
5, 1062±1075.
MarõÂn, O. & Rubenstein, J.L.R. (2001) A long, remarkable journey: tangential
migration in the telencephalon. Nature Rev. Neurosci., 2, 780±790.
Matus, A. (1988) Microtubule-associated proteins: their potential role in
determining neuronal morphology. Ann. Rev. Neurosci., 11, 29±44.
Matus, A., Bernhardt, R. & Hugh-Jones, T. (1981) HMWP proteins are
preferentially associated with dendritic microtubules in brain. Proc. Natl.
Acad. Sci. USA, 78, 3010±3014.
Molobabic, S., Bogdanovic, D. & Drekic, D. (1984) On the neurons with
dendrites intermingled with the ®bers of the human corpus callosum: a Golgi
picture. Gegenbaurs Morph Jahrb, 130, 557±564.
Riederer, B.M. (1990) Some aspects of the neuronal cytoskeleton in development. Eur. J. Morph., 28, 347±378.
Riederer, B.M., Draverova, E., Vicklicky, V. & Draber, P. (1995) Changes of
MAP2 phosphorylation during brain development. J. Histochem. Cytochem.,
43, 1269±1284.
Riederer, B.M. & Innocenti, G.M. (1991) Differential distribution of tau
proteins in developing cat cerebral cortex and corpus callosum. Eur. J.
Neurosci., 3, 1134±1145.
Riederer, B.M. & Innocenti, G.M. (1992) MAP2 isoforms in developing cat
cerebral cortex and corpus callosum. Eur. J. Neurosci., 4, 1376±1386.
Riederer, B. & Matus, A. (1985) Differential expression of distinct microtubuleassociated proteins during brain development. Proc. Natl. Acad. Sci. USA,
82, 6006±6009.
Riederer, B.M., Porchet, R. & Marugg, R.A. (1996) Differential expression and
modi®cation of neuro®lament triplet proteins during cat cerebellar development. J. Comp. Neurol., 364, 704±717.
Romijn, H.J., Janszen, A.W.J.W., van Voorst, M.J.D., Buijs, R.M., BalaÂzs, R. &
Swaab, D.F. (1992) Perinatal hypoxic ischemic encephalopathy affects the
proportion of GABA-immunoreactive neurons in the cerebral cortex of the
rat. Brain Res., 592, 17±28.
Schwaller, B. (1996) Calretinin. In Celio, M.R., Pauls, T. & Schwaller B.
(Eds), Guidebook to the Calcium-binding Proteins. Oxford University
Press, Oxford, pp. 26±28.
Sha®t-Zagardo, B., Davies, P., Rockwood, J., Kress, Y. & Lee, S.C. (2000)
Novel microtubule-associated protein-2 isoform is expressed early in human
oligodendrocyte maturation. Glia, 29, 233±245.
Shatz, C.J., Chun, J.J.M. & Luskin, M.B. (1988) The role of the subplate in the
development of the mammalian telencephalon. In Peters, A. & Jones, E.G.
(Eds), Cerebral Cortex, Development and Maturation of Cerebral Cortex,
Vol. 7. Plenum Press, New York, pp. 35±58.
Tamamaki, N., Fujimori, K., Noyo, Y., Kaneko, T. & Takauji, B. (2003)
Evidence that Sema3a and Sema3T regulate the migration of GABAergic
neurons in the developing neocortex. J. Comp. Neurol., 455, 238±248.
Towbin, H., Staehelin, T. & Gordon, J. (1979) Electrophoretic transfer of
protein from acrylamide gels to nitrocellulose sheets: procedure and some
application. Proc. Natl. Acad. Sci. USA, 76, 4354±4356.
Tucker, R.P., Binder, L.I. & Matus, A.I. (1988) Neuronal microtubule-associated
proteins in the embryonic avian spinal cord. J. Comp. Neurol., 271, 44±55.
Zhu, Y., Li, H.-S., Zhou, L., Wu, J.Y. & Rao, Y. (1999) Cellular and molecular
guidance of GABAergic neuronal migration from an extracortical origin to
the neocortex. Neuron, 23, 473±485.
ß 2004 Federation of European Neuroscience Societies, European Journal of Neuroscience, 19, 2039±2046