Download Developmental Changes Revealed by Immunohistochemical

Developmental Changes Revealed by
Immunohistochemical Markers in Human
Cerebral Cortex
The developing human cerebral cortex is distinguished by a
particularly wide subplate. a transient zone in which crucial cell-cell
interactions occur. To further understand the role of the subplate in
human brain development, w e have studied the immunohistochemical expression of certain neuronal (GAP-43, MAP-2,
parvalbumin) and astroglial (vimentin, GFAP) markers in the
developing visual cortex from gestational ages of 14 weeks to 9
months post-term. At 14-22 weeks, immunoreactivity to GAP-43, a
protein involved in axonal outgrowth, was most prominent in the
subplate and marginal zone neuropil and in the fibers of the
radiations running near the ventricular zone; at 22-42 weeks,
GAP-43 immunoreactive fibers were observed in the maturing
cortical plate. Immunoreactivity for the microtubule-associated
protein MAP-2 was present in the differentiating cortical plate at 14
weeks, but at 22-42 weeks was most prominent in the somata and
dendrites of differentiated neurons, particularly the Cajal-Retzius
neurons of the marginal zone, in neurons of the subplate and in those
forming cortical layer 5. Parvalbumin immunoreactivity did not
appear until 26 weeks, when stained neurons were in a sparse band
of cells in layer 6 and upper subplate. Vimentin and GFAP did not
-stain differentiated neuronal cells. Vimentin immunoreactivity
appeared early in neuroepithelial and radial glial cells, decreasing
after 35 weeks, with a concomitant increase in GFAP
immunoreactivity in radial glial and maturing astrocytic cells. Our
results show that despite the greater complexity of the developing
human neocortex, molecular markers are expressed in spatial and
temporal patterns similar to those observed in non-human primates,
carnivores and rodents. These protein markers should prove useful
in developmental staging, and in providing a framework in which to
examine congenital disorders of cerebral development
The mammalian neocortex is a multilaminar structure that
develops in an inside-first, outside-last pattern (Caviness et al,
1981; Sidman and Rakic, 1982). Neurons of the innermost
cortical layer 6 are born earlier than those of the more superficial
layers 2 and 3 (see reviews by McConnell, 1991; Allendoerfer
and Shatz, 1994). In addition to the neurons constituting the
permanent adult cortical layers 1-6, a transient population of
neurons is present in the subplate layer of the developing
cerebral wall (Allendoerfer and Shatz, 1994). Studies in
experimental mammals (carnivores and primates) have shown
that subplate neurons are among the first generated post-mitotic
neurons of the telencephalon (Kostovic and Rakic, 1980, 1990;
Luskin and Shatz, 1985a,b; Chun and Shatz, 1989a; Kostovic,
1990; McConnell, 1991; Allendoerfer and Shatz, 1994). These
subplate neurons differentiate morphologically and function
physiologically during fetal stages, prior to the maturation of the
definitive neurons of the mature cortex (Friauf etal,\ 990; Friauf
and Shatz, 1991). During normal postnatal development the
subplate neurons largely disappear (Valverde and Facal-Valverde,
1987; Chun and Shatz, 1989b), although some persist as
Lawrence S. Honig1-2-3, Kathrin Herrmann1-4 and
CarlaJ. Shatz15
Departments of 'Neurobiology and 2Neurology, Stanford
University Medical Center, Stanford, CA 94305, USA
Current addresses: 'Department of Neurology (F2-318),
University of Texas Southwestern Medical Center at Dallas,
5323 Harry Hines Blvd, Dallas, TX 75235-9036, ''Laboratory of
Neurophysiology, NIMH/NIH Animal Center, POB 608,
Poolesville, MD 20837 and 5Howard Hughes Medical Institute
& Department of Molecular & Cell Biology (LSA #3200),
University of California, Berkeley, CA 94720-3200, USA
interstitial neurons of the adult subcortical white matter
(Valverde and Facal-Valverde, 1988; Chun and Shatz, 1989b;
Kostovic and Rakic, 1990; Meyer etal, 1992).
The subplate is a region through which growing axons as well
as migrating neurons travel. It is thought to function as an early
scaffold involved in the formation of connections between
thalamus and cortex (Shatz et al, 1988; McConnell et al., 1989;
Ghosh and Shatz, 1992a; Allendoerfer and Shatz, 1994). The
subplate layer is more pronounced in width as one ascends the
mammalian phylogenetic tree (reviewed in Allendoerfer and
Shatz, 1994). For example, it is larger in cats than in rodents
(Uylings and van Eden, 1990; Allendoerfer and Shatz, 1994), and
in monkeys than in cats (Kostovic and Rakic, 1980,1990), and it
is particularly prominent in the developing human brain
(Kostovic and Rakic, 1980, 1990). Clinically, this region is
susceptible to injury in utero or perinatally. For example, it is
affected in both of the two major neonatal brain conditions,
periventricular leukomalacia and germinal matrix hemorrhage,
which are responsible for significant congenital developmental
delay and subsequent mental retardation (Volpe, 1989, 1990).
We have undertaken a developmental study of human visual
cortex to examine the spatio-temporal pattern of immunostaining for selected markers of neuronal and glial development.
To follow neuronal development, antibodies to GAP-43, MAP-2
and parvalbumin were used. GAP-43 is a protein linked to axon
outgrowth during both development and regeneration (Skene,
1989; Biffo et al., 1990). With the exception of the earliest
developmental stages, at which time it is also present at low
levels in neuronal somata, the protein is present at high levels in
axons and neuronal growth cones. MAP-2 (microtubuleassociated protein 2) is a microtubule- associated protein found
exclusively in neuronal dendrites and somata (Caceres et al,
1984a,b; Sims et al., 1988; Chun and Shatz, 1989a; Crandall,
1989; Schoenfeld et al, 1989). MAP-2 immunoreactivity can be
detected in the cortical and subplate neurons of developing
mammals and non-human primates. Parvalbumin is a calciumbinding protein associated with certain subpopulations of
GABA-ergic cells in the developing mammalian cerebral cortex,
including subplate neurons (Antonini and Shatz, 1990;
Hendrickson et al, 1991; Van Brederode et al, 1991). In
addition, to confirm the specificity of our immunocytochemistry
and correlate with the neuronal markers, we examined the
staining patterns for two astroglial markers, GFAP and vimentin.
Vimentin is an intermediate filament protein of the cytoskeleton
known to be present early in development in primitive
neuroepithelial cells and radial glia (Sasaki et al, 1988; Bignami
and Dahl, 1989; Stagaard and M0llgard, 1989; Sarnat, 1992).
GFAP, a different intermediate filament protein, is well
recognized as a marker for astroglial cells including mature
astrocytes of the central nervous system (Antanitus et al, 1976;
Cerebral Cortex Nov/Dec 1996;6:794-806; 1047-32Il/96/$4.00
Levitt and Rakic, 1980; Stagaard and Mellgard, 1989; Sarnat,
1992).
The immunostaining patterns of these five antibodies at
different ages can provide information about the progression of
developmental changes in the neurons and glia of the subplate
and cortical plate. We studied this progression in human
cerebral cortex at developmental stages ranging from 14
gestational weeks to early postnatal life. The patterns of distribution of these antigen markers observed here in the human
cortex are highly similar to those observed in the developing
cerebral cortex of other mammals, implying similarities in the
underlying developmental events of neuronal maturation and
formation of connections.
vimentin (clone Vim 3B4; Boehringer-Mannheim; 1:40). These antibodies
all cross-reacted with their human counterpart protein antigens,
consistent with the strong amino acid sequence conservation exhibited
by these proteins. Antibody directed against GFAP was a rabbit
anti-human polyclonal antibody (1:250; Dr L. Eng, Stanford University).
Following incubation with primary antibody, sections were washed with
PBS, incubated with appropriate anti-mouse or anti-rabbit biotinylated
secondary antibody, washed and stained using the Vectastain ABC-elite
kit (Vector) avidin-biotin linked peroxidase method. Diaminobenzidine
(Sigma) was used as chromogen. All slides were dehydrated and mounted
using Permount (Fisher Scientific; Pittsburgh, PA). Control sections, using
normal mouse serum (1:500) instead of 'primary antibody' or omitting
either primary antibody or biotinylated secondary antibody, did not show
reaction product, except for some occasional stain in blood cells
persisting despite peroxide pre-treatment. Sections near those used for
immunohistochemistry were stained histologically by the Nissl method
using cresyl violet.
Materials and Methods
Tissue
Hunan fetal brains were obtained at autopsy, or following therapeutic
abortions, and immersion-fixed in buffered 10% formalin for 14-21 days
at 4°C. Use of autopsy specimens received approval by the Department of
Pathology, Stanford University School of Medicine. The Stanford
Institutional Review Board for Human Subjects in Medical Research gave
Human Subjects Approval. Gestational age (GA) is defined using the
clinical convention, as total weeks (including postnatal survival time)
from the last menstrual period. When this information was not known,
specimens were dated by use of the prior physical measurements
obtained upon ultrasound in utero, or those made at post-mortem
autopsy. By the convention used here, GA is 2 weeks greater than the
actual conceptional (post-ovulatory, or fertilization) embryonic age;
normal term is GA 40 weeks (40 weeks). Any brains showing anatomical
anomalies, parenchymal hemorrhages or macroscopic autolytic changes
were excluded. The average post-mortem interval was 18 h (range 1-36
h). With the exclusion of several brains showing microscopic
post-mortem change on hcmatoxylin & eosin staining, there were no
effects of post-mortem interval noted. Twenty brains were examined in
this study, including several post-term (>40 weeks GA) brains; GA ranged
from 14 to 79 weeks GA. The nomenclature of the developing cortical
layers is a modification (Allendoerfer and Shatz, 1994) of that of the
Boulder Committee (1970).
After brain fixation, blocks (-5 mm thick slabs) of left or right
occipital cortex were manually cut in the coronal plane. The calcarine
fissure was included in these slabs for those stages at which the fissure
was evident (after -27 weeks). The tissue blocks were infiltrated for 1-3
days at 4°C in 25% sucrose in 0.1 M phosphate buffer, pH 7.4, and then
freeze-embedded in Tissue-Tek® O.C.T. Compound (Mites; Elkhart, IN)
on a block of dry ice before storage at -80°C or sectioning. Cryostat
sections were cut at 12 urn thickness and mounted onto slides previously
made adhesive using a subbing solution of 0.1% chrome alum-1.5%
gelatin-30% ethanoL Slides were then stained as below, or in some cases
stored desiccated at -20°C prior to staining.
Immunohistochemistry
Sections were usually pretreated with 3% hydrogen peroxide in
Dulbecco's sodium phosphate-buffered saline (PBS), pH 7.4, to attenuate
endogenous peroxidase staining associated with blood cells and vessels
(Nilaver and Kozlowski, 1989), then washed in PBS. For the antibodies
used in this study, this pretreatment did not affect the antibody-specific
staining. Prior to treatment with primary antibody, slides were
preincubated for 1-3 h in blocking solution consisting of 2% horse serum
(Vector, Burlingame, CA), 1% bovine serum albumin (Sigma Chemical
Co., St Louis, MO) and 0.1% Triton X-100 (Sigma) in PBS. This step was
followed by incubation with primary antibody, diluted in blocking
solution, for 12-18 h at 4°C. Antibodies (with their dilutions) included
mouse monoclonal antibodies generated against: rat GAP-43 (clone
9-1E12; 1:10 000), originally provided by Drs D. J. Schreyer and J. H. P.
Skene (Schreyer and Skene, 1991) and now available from
Boehringer-Mannheim (Indianapolis, IN); rat MAP-2 (clone HM-2; Sigma;
1:2000): carp parvalbumin (clone PA-135; Sigma; 1:1000); and bovine
Results
During development, the human cerebral hemisphere, like that
of all mammals, can be subdivided into several embryonic zones.
Initially, the neuroepithelium is a single layer, which develops
into a trilayered structure consisting of a ventricular zone, a
preplate and a marginal zone. The first appearance of the cortical
plate occurs at -10 weeks gestational age (Sidman and Rakic,
1982). Thus by 14 weeks, the earliest stage studied here, four
zones are present (Fig. IB). From ventricle to outer brain surface
these are: the ventricular zone (VZ), which contains progenitor
cells undergoing cell division; the intermediate zone (IZ), which
can be further subdivided into the radiations (RA) and the
subplate (SP); the cell-dense cortical plate (CP), which contains
the accumulating postmigratory neurons of the adult cerebral
cortex; and the marginal zone (MZ) (Boulder Committee, 1970;
Allendoerfer and Shatz, 1994). In humans at -40 weeks GA, the
generative ventricular zone has disappeared, leaving a
differentiated ependymal layer, whereas the intermediate zone
has become the definitive white matter. In experimental
mammals, including nonhuman primates, it is known that the
subplate and marginal zones contain the first postmitotic
neurons of the cerebral cortex. However, by early maturity many
of these have been eliminated by cell death. The subplate
becomes the mature white matter, which has few neurons, and
the marginal zone becomes the molecular layer, which contains
very few neurons and is known as cortical layer 1.
Distribution of GAP-43 Immunoreactivity
GAP-43 immunoreactivity is present in the cerebral cortex
throughout the stages examined in this study. At the earliest
stage examined (14 weeks), GAP-43 is present through nearly
the full width of the cerebral wall (Fig. L4), excepting the
ventricular zone (VZ) (compare Fig. L4 and B). The most intense
regions of GAP-43 immunoreactivity at the 14 week stage are the
marginal zone and intermediate zone. In the intermediate zone,
staining is present in dense fiber fascicles running generally
parallel to the ventricular zone (Figs \A and 2/4). However, the
forming cortical plate at 14 weeks, which spans -30% of the
width of the cerebral wall, does show some modest
immunoreactivity. This is evident in cell somata and neuropil,
with some fine fibers running radially within the cerebral wall,
most likely axonal processes of differentiating cortical cells.
Thus, at the 14 week stage, with the exception of the minor cell
body immunoreactivity found in the cortical plate, anti-GAP-43
antibody is predominantly a fiber stain.
At stages 20-25 weeks (Figs ID and 2B), fiber staining in the
Cerebral Cortex Nov/Dec 1996, V 6 N 6 795
14w
sp
IZ
*6w
1
2/3
sp
1mm
Figure 1 . Visual cortex at gestational ages 14-36 weeks. This figure of nine panels consists of three rows showing sections of brains from different fetal ages and three columns
with different stains. The ages are 14 weeks in the top row (4, B, C), 22 weeks in the middle row (0, E. F) and 36 weeks in the bottom row (6, H. I). Immunohistochemical stains
are for GAP-43 (4. D. G) and for MAP-2 [C. F. I), and comparable sections are shown with cresyl violet (CV) staining (fl. f, tf). Brain regions are labeled in the central column panels:
mz. marginal zone (future cortical layer 1); cp. cortical plate (future cortical layers 2-6); iz, intermediate zone (containing the subplate, sp, a transient zone containing subplate
neurons, and the radiations, ra), destined to become the definitive white matter; vz, ventricular zone (early proliferative zone and future ependyma). All panels are at the same
magnification; the scale bar shown in G is 1 mm.
796 Immunohistochemtstry in Developing Cortex • Honig et at.
W^^
'^
vz
vz
B
Figure 2. GAP-43 immunoreactivity in the intermediate zone. Dense fiber staining is evident in the optic radiations and developing white matter. Sections through the inner half of
the cerebral wall of 14 week Ifl) and 22 week (8) brains, and the central portion of a gyrus of a 36 week brain (C) are shown. At 14 weeks ifl) there is a relative absence of staining
in the ventricular zone (vz), but marked staining in the optic radiations (ra) present in the intermediate zone, which are shown at higher magnification in the inset (*). At 22 weeks
(fl) the intermediate zone is wider and the layered arrangement of the radiations can be appreciated; the staining pattern is similar, with more intense staining in the radiations (ra)
than in the subplate (sp) or ventricular zone. At 36 weeks (C) layers 4-6 and the developing white matter (wm) are labeled; dense fiber staining is evident in the white matter. Scale
bars are 100 jim.
intermediate zone increases in prominence. A dense layer of
axon fascicles coursing superficial to the ventricular zone
becomes distinct from the overlying subplate; this represents the
radiations portion of the intermediate zone (Fig. IE). At 20-30
weeks (Figs ID and 2B) the GAP-43 immunoreactive fascicles of
the radiations continue to be prominent. These fascicles
presumably contain corticofugal fibers as well as the corticopetal
(thalamocortical) fibers of the optic radiations emanating from
the lateral geniculate nucleus. At 36 weeks and later, following
deep sulcation of the developing cortex with gyrus formation,
the GAP-43 immunostained fibers are present in the center of the
gyri, anatomically representing the subcortical white matter
(Figs lGand2QAt early stages, GAP-43 staining is very low in the cortical
plate (Fig. 34, B); no immunoreactivity is visible in cell bodies at
19 weeks. The staining is markedly less than that in the
intermediate zone, where there is robust fiber staining (Fig. 5A).
However, between 21 and 40 weeks, increasing amounts of
GAP-43 immunoreactivity appear in the cortical plate present in
fine radial fiber bundles (Fig. 3C, E, G). At 22 weeks these fibers
are evident only in the deep regions of the cortical plate
(Fig. ID); subsequently, at 26-27 weeks (Fig. 3C-F), immunostaining extends throughout the entire plate, reaching the
outermost cortical cellular layers 2/3 (Fig. 3E, F). By 36 weeks
(Figs \G and 3G, H) the cortical plate staining is non-uniform,
with a laminar accumulation of immunoreactivity. The most
intense staining is present in three bands: a top band located in
the marginal zone, a middle band corresponding to layer 4 and a
bottom band corresponding to upper layer 6/lower layer 5 (Figs
Distribution ofMAP-2 Immunoreactivity
At 14 weeks gestation, anti-MAP-2 monoclonal antibody stains
neuronal cell bodies and processes in the forming cortical plate
(Figs \C and 44, B, D). The ventricular zone is essentially
unstained (Fig. 4A), while the marginal zone (Fig. 4A, B) and
subplate (Fig. 4A, B, D) contain immunostained cell bodies at
low density. This distribution of MAP-2 immunoreactivity (Fig.
1O is almost complementary to that of GAP-43 (Fig. L4)
staining: at 14 weeks, the densest MAP-2 immunoreactivity is
located in a closely packed lower part of the marginal zone and
within the neurons of the cortical plate (Figs \C and 4A, B),
whereas GAP-43 staining is most intense in the marginal and
intermediate zones and lowest in the cortical plate (Fig. 1/4). At
22 weeks, intense MAP-2 staining is detectable in the sparsely
distributed neurons of layer 1, known as the Cajal-Retzius
neurons (Figs 4B, C, E and 5A), and in maturing cortical layers 5
and 6 (Figs AC and 5E). By 36 weeks, at which time the cortical
layers are clearly evident (Caviness et al., 1981; Sidman and
Rakic, 1982; Kostovic and Rakic, 1984; Kostovic et al, 1988,
1989), staining is present not only in layer 1 neurons and the
large pyramidal cells of layers 5 and 6, but also in the maturing
neurons of layers 2-4 (Fig. 5C, D, G, H). This immunostaining is
prominent in cell somata and dendrites, especially the apical and
parabasal dendrites of the layer 5 pyramidal cells (Fig. 5D, E).
Subplate neurons show MAP-2 immunoreactivity throughout
Cerebral Cortex Nov/Dcc 1996, V 6 N 6 797
36w
Figure 3. GAP-43 immunoreactivity in the developing cortical plate. Photographs of GAP-43-stained sections (left: A C, E, G) are paired with corresponding cresyl violet (CV)-stained
sections (right: B. D.F.H). These latter are labeled with abbreviations as in Figure 1, and with the gestational ages: 19 weeks (4, B), 26 weeks (C. 0), 27 weeks (£, F) and 36 weeks
(G, H). GAP-43-immunoreactive fibers running perpendicular to the pial surface increase in prominence during development, as described in the text. Magnification is the same in all
panels; the scale bar shown in E is 500 urn.
the period under study. In the 14 weeks subplate, neuronal cell
body and process staining is most evident in the upper subplate,
immediately below the cortical plate (Figs \C and 44, Q. At
22-27 weeks MAP-2 somatic staining is most prominent in the
subplate cells, as well as the pyramidal neurons of layer 5 (Figs IF
and 5A, B, E, F). The population of immunostained bipolar or
multipolar subplate neurons appears denser in the upper,
compared with the lower, portions of the subplate. By 36 weeks,
subplate somata and processes are intensely immunoreactive for
MAP-2 (Figs 1/ and 5C, G, H), and can be seen scattered
throughout the subplate and developing white matter (Fig. 5G,
H).
Parvalbumin
Immunoreactivity
At the earliest ages studied here (14-25 weeks), no parvalbumin
798 Immunohistochcmistry in Developing Cortex • Honigetal.
immunostaining could be detected in the cerebral wall. At
subsequent stages, from -26 weeks onwards, increasing
numbers of neurons staining for parvalbumin are present. These
neurons consist of a sparse population located in the upper
subplate or deepest part of cortical layer 6 (Fig. 6/1, C, E, F).
Comparison with adjacent MAP-2 immunostained (not shown)
or Nissl-stained sections (Fig. 6B, D) reveals that these
Immunoreactive cells are only a small fraction of the total
neuronal population in this region. The number of parvalbumin
immunoreactive cells increases during development such that a
prominent band is visible at stages 36 weeks (Fig. 6A) and later.
However, this broad band has rather low cell density, and
consists of cells separated at some lateral distance from each
other, located in layer 6 and the upper subplate. The neurons are
often spindle shaped at 26-30 weeks. At later stages, the labeled
mz.
Figure 4. MAP-2 immunoreactivity in the developing cerebral wall. Sections from 14 week (4, B. D) and 22 week (C, E] brains are stained with antibody to MAP-2. For the 14 week
brain the entire cerebral wall is shown 14), along with higher magnification views of the cortical (6) and subplate zones [D, magnified view of box \nA). Prominent staining of subplate
neurons can be seen (0, arrowheads). The Cajal-fietzius neurons of the marginal layer also stain densely (£, short arrows). These neurons are somewhat obscured at 14 weeks by
extensive neuropil staining [B). but they are obvious at 22 weeks (C and E. which is a magnified view of box in C). Staining is also observed in the forming cortical plate. Abbreviations
are as in Figure 1. All scale bars are 100 |xm.
neurons appear larger, more polygonal and with more processes
(Fig. 6E, F). At term and during the postnatal period, these cells
are more numerous, with greater abundance of more ramified
processes, than present during previous stages (e.g. 26-36
weeks). In the postnatal 79 weeks brain shown in Figure 6C, D,
F a greater number of parvalbumin immunoreactive cells are
evident than at 36 weeks; while at 79 weeks these cells are still
mostly located in layer 6, scattered cells are seen throughout
other layers of the cortex, particularly in layers 3, 5 and the white
matter. The ramified cell processes stain prominently (Fig. 6F).
Vimentin and GFAP Immunoreactivity
In agreement with previous studies (Stagaard and Mollgard,
1989; Wilkinson et al, 1990; Sarnat, 1992), vimentin immunoreactivity at the ages studied here is evident in cells that appear
morphologically to be neuroepithelial cells and radial glia. At 14
weeks there is some immunoreactivity for vimentin distributed
throughout the entire width of the cerebral wall (Fig 7A). The
most dense staining is in the cells of the ventricular zone. There
is moderately dense staining in the intermediate zone, in both
cell bodies and 'radially' oriented cell processes (Fig. 7A). The
cortical plate contains only light staining, present in radially
oriented fibers. A thin band of moderate staining is present at the
boundary between the marginal zone and the cortical plate. At
later ages (19-36 weeks) vimentin immunoreactivity persists in
cells of the neuroepithelium, and in cell bodies and processes in
the intermediate zone that morphologically fit criteria for radial
glia (Rakic, 1978; Schmechel and Rakic, 1979; Voigt, 1989).
Radial glia have their cell bodies situated in the ventricular and
subventricular zones, and long fibers extending from ventricular
surface to marginal zone (Fig. IB, C). At 36 weeks the most
intense staining for vimentin persists in the periventricular
region adjacent to the ventricular zone and in the forming
subventricular zone (Fig. 7O- Subsequently, particularly during
postnatal development, the vimentin immunoreactivity evident
in the brain diminishes. The remaining vimentin staining
Cerebral Cortex Nov/Dcc 1996, V 6 N 6 799
- stt~^r*H+»* 90> f>,Sk^%:
m
wm
Figure 5. MAP-2 immunoreactivity in the developing cortical plate. Anti-MAP-2 immunostaining is seen in low magnification views of the cortical plate of 22 week ifi), 27 week (B)
and 36 week (C) brains. Higher magnification panels show selected regions from sections of the same brains: 22 weeks layer 5 (£), 27 weeks subplate If). 36 weeks cortical plate
layers 5-6 (D), and 36 weeks upper subplate (6) and subplate and forming white matter (H). At 22 weeks staining is most prominent in the Cajal-Retzius cells of the marginal zone
IA. arrowheads), the pyramidal cells of layer 5-6 and subplate cells (A, £). While staining of subplate cells and pyramidal layer 5 cells remains strong at subsequent stages (5, C. D.
F. G. H), prominent staining is also seen in differentiating neurons of layers 4,3 and 2 (5, C). Abbreviations are as in Figure 1. All scale bars are 100 \un.
observed at 79 weeks includes intense staining of the ependyma
itself, as well as some cell bodies and tortuous fibers located in
the periventricular region (Fig. ID). Staining in blood vessels and
capillaries is also pronounced.
GFAP immunostaining by our methods is very light at the
earlier developmental stages studied here (brains of 14-26
weeks of age); there is some staining of cells in the ventricular
and subventricular rones (Fig. &4). Immunostaining increases in
the subventricular zone during the period 26-36 weeks, and the
cells and processes that are stained are morphologically
characteristic of radial glia. However, during this period, GFAP
staining by our methods appears less intense and in fewer cells
than that observed with antibody to vimentin. By 36 weeks,
differentiated ramified astrocytes are evident throughout tfee
intermediate zone and cortical plate (Fig. 8B). Postnatally, there
is prominent staining of process-bearing astrocytes as well as
persistent ventricular zone immunoreactivity (Fig. 8QDiscussion
We have studied the development of human cerebral cortex
during the second and third trimesters of fetal development,
from gestational ages 14-40 weeks and postnatally, by
examining the distribution of certain neuronal and astroglial
protein markers (summarized in part in Table 1), whose
distribution and developmental regulation during development
has previously been investigated in lower animals. Compared
with lower mammals of similar stages, human cerebral cortex
800 Immunohtstochcmistry in Developing Cortex • Honig et a].
has a much more prominent subplate and a slower, more
elaborate pace of development of connections and cortical
layers. However, the spatio-temporally patterned sequence of
immunostaining for the neuronal and astroglial proteins studied
here, while prolonged in time, shows broad similarities in
pattern to that of other mammals. In particular, GAP-43 staining
was present mostly in axonal fiber tracts, and in locations filled
with growth cones and with ingrowing axons (Table 1). MAP-2
stained a number of neuronal cell bodies and dendrites in the
developing brain, particularly those of subplate, marginal zone
and large cortical projection neurons. Parvalbumin was present
in a much more restricted set of upper subplate and lower
cortical plate neurons. The astroglial marker proteins vimentin
and GFAP were seen in primitive radial glia cell bodies near the
ventricular zone and their fibers. Astrocytes in the subplate and
cortical plate were evident in later stages, using antibody to
GFAP. These observations indicate that the patterns of cortical
development as assessed by these markers are highly similar in
humans and lower mammals.
GAP-43 is a Marker for Axonal Development
GAP-43 immunoreactivity was present in the developing human
cortex predominantly in the thalamic radiations, and in the
neuropil of the marginal zone and subplate. These zones are
known from previous anatomical studies in experimental
mammals to contain many growing axons (Marin-Padilla, 1971;
Molliver etal,l973; Kostovic and Rakic, 1980, 1990; Ghosh and
1
'
• •-.'.
—
B
'
*
•
'
' .
• ' • .
'
•
•
'
'
sp-: -
• • • " L _ v b :• >:.
i
>
:
.
. ' ^
«:
4
Figure 6. Parvalbumin immunoreactivity in the developing brain. Sections from late fetal and postnatal brains of gestational ages 36 weeks (A, B, E] and 79 weeks (C, D, F)
respectively are shown. Low magnification views show parvalbumin immunostained sections (4, C) adjacent to comparable Nissl-stained sections (ft D), labeled with abbreviations
as in Figure 1. Parvalbumin-immunoreactive neurons are mostly confined to the subplate and lower cortical plate at 36 weeks {A, B. E), although more numerous than when first
observed at 26-27 weeks (not shown). At 79 weeks (C, D, F) parvalbumin-immunoreactive cells are distributed throughout other cortical layers as well, although there remains a
higher density of these cells in the deeper layers of cortex and white matter (C, F). The higher magnification pictures show the morphology of the stained subplate neurons at 36
weeks (£) and 79 weeks (f). All scale bars are 100 \m.
Shatz, 1992a). This pattern of dominant fiber staining is
consistent with the fact that GAP-43 is synthesized in neurons
extending processes and axonally transported to growth cones
(Skene, 1989). Since the protein does not accumulate in cell
bodies but is rapidly transported, there are relatively
complementary, or mutually exclusive, patterns of mRNA
expression and protein distribution for GAP-43 (Biffo et aL,
1990).
The presence of GAP-43 protein or mRNA in the developing
brain has been demonstrated biochemically in a number of
systems, including cat visual cortex (Mclntosh et aL, 1990) and a
20-22 week human fetal brain cDNA library (Neve et aL, 1987).
The anatomic distribution of GAP-43 protein in developing brain
has also been established previously using immunohistochemistry in embryonic rodents, including mice (Biffo et aL,
1990) and rats (Dani et aL, 1991). Rodents do not have a very
well developed subplate. However, the prominent concentration
of GAP-43 in deep fiber tracts seen in our study of developing
human brains is consistent with similar patterns seen in the
developing mouse and rat (Biffo et aL, 1990; Dani et al., 1991).
It is also consistent with a recent report in which GAP-43 was
studied in human brain development in several embryos at
earlier gestational stages (6-10 weeks GA) than those in this
report (Milosevic et aL, 1995). These authors noted no GAP-43
immunoreactivity at 6-8 weeks, but the development of some
fiber staining below the forming cortical plate at 10 weeks GA
(Milosevic et aL, 1995).
The striking intensity of staining in the neuropil of the
marginal zone in humans resembles that seen in small mammals.
At the earliest age studied, 14 weeks, GAP-43 immunoreactivity
was observed in the cortical plate. The presence of
immunostaining in the cortical plate at this stage, by analogy
with experimental mammals (De Carlos and O'Leary, 1992;
Ghosh and Shatz, 1994), likely reflects GAP-43 production by
early resident cortical neurons of layers 5 and 6 that are
beginning to extend axons towards their targets. At subsequent
stages, staining in the cortical plate was minimal. This relative
absence of staining in the cortical plate during the major period
of neuronal migration has also been noted in rat (Dani et aL,
1991). The gradual pattern of intrusion of GAP-43 immunoCerebral Cortex Nov/Dcc 1996, V 6 N 6 801
I
<*•*#*"
r-
<
14w
Figure 7. Changes in the pattern of vimentin immunoreactivity from gestational age 14 weeks to 79 weeks. The entire cerebral wall at 14 weeks (4) shows immunoreactivity for
vimentin, including intense immunoreactivity at the ventricular surface (v), and in fibers coursing perpendicular to the ventricular surface (arrowheads). Subsequent stages shown
include 19 weeks (S). 36 weeks (C) and 79 weeks (D). Staining for vimentin continues to be present at the ventricular surface, including the differentiated ependyma (e) at 79 weeks,
and in fibers coursing from ventricular zone to the base of the cortical plate at 19 and 36 weeks. Blood vessels (b) also show immunoreactrvity. As described in the text, the fibers
are likely those of radial glia. All scale bars are 200 UJTI.
reactive fibers into the differentiating cortical plate that we have
been able to elucidate in the developing human has not been
demonstrated clearly in lower mammals, although a transient
period of increased GAP-43 reactivity in the cortical plate has
indeed been shown to occur during the first postnatal week in
the rat (Erzurumlu et aL, 1990; Dani et aL, 1991), a time when
thalamocortical axons are known to invade the cortical plate
(Catalano et aL, 1991; Agmon et aL, 1993)- This accumulation of
GAP-43 in the rat is in layers 4 and 6 (Erzurumlu et aL, 1990;
Dani et aL, 1991), as we observed in the human neocortex.
Thetiming of the appearance of GAP-43 staining in fine
perpendicular fibers in the deep cortical plate at 22-28 weeks
correlates well with the proposed period of ingrowth of the
waiting fibers of the thalamocortical afferents (Kostovic and
Rakic, 1990). The appearance at later ages of increased GAP-43
in the more superficial cortical layers, with particularly dense
staining in layer 4, is also consistent with the ingrowth of the
thalamocortical axonal projection, which ramifies extensively in
this layer (Shatz et aL, 1988).
Axon pathway tracing studies of experimental animals have
revealed that in the cat and monkey, thalamocortical axons do
not immediately invade the cortical plate. Rather, they
accumulate in the subplate, beneath the forming cortical plate,
802 Immunohistochcmistry in Developing Cortex • Honigetal.
for weeks in the cat (Shatz and Luskin, 1986; Ghosh and Shatz,
1992b) or months in the monkey (Rakic, 1983; Kostovic and
Rakic, 1990). The extensive GAP-43 immunostaining within the
subplate at ages 14-40 weeks likely reflects in part the
accumulation of axons waiting in this zone prior to their
invasion of the cortical plate. In addition, the intense
immunoreactivity observed in the marginal zone is consistent
with the very early generation of marginal zone neurons in all
species studied (reviewed in Allendoerfer and Shatz, 1994), and
the presence there of an extensive synaptic network (Molliver et
aL, 1973)- Taken together, these observations on the appearance
and timing of GAP-43 immunoreactivity suggest that the
presence of GAP-43 protein is correlated with periods and routes
of axon growth within the developing human cortex.
MAP-2 Staining Marks Subplate and Other Large
Neurons
The pattern of MAP-2 immunostaining differs markedly from the
fiber labeling of GAP-43- Antibody to MAP-2 stains somata and
dendrites of the differentiating subplate and cortical neurons.
While anti-MAP-2 stains mature neuronal cells throughout the
width of the developing human cortex, at early stages (14 weeks)
prior to differentiation of the cortical plate the staining of cell
•
>
r
Figure 8. Immunoreactivity for the astroglial marker GFAP in the prenatal and postnatal brain. Immunostaining for GFAP is most intense in the periventricular (v) regions, as can be
seen in 26 week (4) and 79 week (C) brains. However, from mid-gestation onwards, prominent staining of cellular elements is also observed in the intermediate zone, as seen in
panel B (36 weeks) and in the upper portions of panels/1 and C, as well as even in the subpial region postnatally, at 79 weeks (not shown). The morphology of the stained parenchymal
cells is that of astrocytes. All magnifications are the same; the scale bar shown in C is 100 urn.
Table 1
Summary of immunohistochemical data for neuronal markers
Antibody immunoreactivity
MAP-2
GAP-43
14 weeks
19-20 weeks
21-26 weeks
27-32 weeks
33-40 weeks
MZ
CP
SP
neuropil
neuropS
+
-
neuropi
neuropB
neuropi
-L fibers
X fibers
neuropH:4,6
|| fibers
II fibers
|| fibers
II fibers
II fibers
VZ
Parvalbumin
MZ
CP
SP
CR
CR
CR
CR
CR
+++
5,6
3,5,6
3,5,6
2,3,4,5,6
SP
SP
SP
SP
SP
VZ
MZ
CP
6
5,6
2,3,4.5.6
SP
VZ
uSP
uSP
uSP
For cerebral layers MZ (marginal zone). CP (cortical plate). SP (subplate). and VZ (ventricular zone), as defined in the text, the table entries show the type of staining obtained with antibodies directed
towards neuronal markers GAP-43, MAP-2. and parvalbumin. The symbols and abbreviations are: -I- (staining present), - (staining absent), 1 (perpendicular to ventricular wall), 11 (parallel to ventricular
wall); CR (Cajal-Retzius cells), and |u]SP ([upper) subplate cells). The numbers in the CP columns refer to the cortical layers.
somata is most marked in the subplate neurons and the
Cajal-Retzius cells of the marginal zone, the two populations of
neurons with earliest-known birthdates (Luskin and Shatz,
1985b; Kostovic and Rakic, 1990). Previous studies in the
developing cat brain have similarly shown prominent MAP-2
immunoreactivity in the subplate, where it is exclusively
associated with subplate neurons (Chun and Shatz, 1989a). As
the cortical plate matures, deeper layers such as layers 5 and 6,
and then later in turn layers 4 and especially 3 and 2 (each of
which have pyramidal cell populations), stain very darkly. The
polymorphic cells of layer 6 and small granular cells of layer 4
stain less intensely. The abundance of MAP-2-positive cells
within the subplate, which do not stain with astroglial markers
vimentin or GFAP, provides further evidence for their neuronal
identity and strengthens the analogy between these neurons in
humans and similar neurons described in cats (Shatz et al, 1988)
and monkeys (Kostovic and Rakic, 1990). Furthermore, the very
prominent early MAP-2 staining of the cell somata and dominant
apical dendrites of the pyramidal cells of future layer 5 may
provide a useful histochemical landmark for this cortical lamina
in humans.
Parvalbumin-reactive Neurons Form a Distinctive
Population
At early human stages, <26 weeks, no cells immunoreactive for
parvalbumin were seen in the cerebral cortex. At analogous
Cerebral Cortex Nov/Dcc 1996, V 6 N 6 803
early stages in monkeys, parvalbumin immunoreactivity is
similarly absent (Hendrickson et al., 1991). Wefindthat from 26
weeks through to term (40 weeks) in humans a layer of initially
sparsely distributed, but strongly immunoreactive, cells was
seen in layer 6 and the upper part of the subplate. This
restriction of parvalbumin-immunoreactive cells in the human
third trimester to layer 6 is similar to the limited deep
localization of these cells observed in early postnatal rats and
gerbils, and in late fetal (E54) and neonatal cats (Stichel et al.,
1987; Hogan and Berman, 1994). hi comparison to macaque
monkeys, the parvalbumin-positive cells may appear earlier in
human development, since in the monkey they have been
reported as first appearing at the very end of gestation
(Hendrickson et al, 1991), a time-point considerably later than
the occurrence of complete formation of the definitive cortical
layers. Following the initial appearance of the parvalbuminreactive cells in the deep layers, these cells are subsequently
seen increasingly superficially, in layers 2-5, a progressive
spread also noted in studies of rodents (Seto-Ohshima et al,
1990), cats (Stichel et al, 1987; Hogan and Berman, 1994) and
non-human primates (Hendrickson et al, 1991). Indeed, in the
postnatal human specimens, parvalbumin-reactive cells were
widely distributed throughout the layers of the cerebral cortex,
consistent with the more widespread distribution noted in
mature postnatal experimental mammals (Stichel et al, 1987;
Hendrickson et al, 1991; Van Brederode etal, 1991; Hogan and
Berman, 1994). Parvalbumin-reactive cells are GABA-ergic
interneurons (Van Brederode et al, 1990). The developmental
appearance and maturation that we observe in humans is in
an'inside-out' pattern, like that reported for experimental
animals. The timecourse of the marked increase in parvalbuminpositive neurons during human cortical maturation suggests a
relationship between their appearance and cortical synaptogenesis, as has previously been suggested for experimental
primates (Hendrickson et al., 1991). The period in which the
earliest parvalbumin-reactive cells appeared—26 weeks GA—is
the same as that in which the earliest synapses within the
cortical plate have been observed electron microscopically—25
weeks GA (Molliver et al, 1973).
Notes
We thank Drs D. J. Schreyer, J. H. P. Skene and L. Eng for antibodies. We
acknowledge with thanks the support of NIH grant EY02858 (C.J.S.), the
Alzheimer's Association (C.J.S.), Fight for Sight/Prevent Blindness
Fellowship (K.H.), Nato Fellowship (K.H.), Dana Fellowship in the
Neurosciences (L.S.H.), Walter V. and Idun Berry Fellowship (L.S.H.), and
UT President's Research Council Award (L.S.H.). C J.S. is an investigator of
the Howard Hughes Medical Institute.
Address correspondence to Dr L. S. Honig, Department of Neurology
(F2-318), University of Texas Southwestern Medical Center at Dallas, 5323
Harry Hines Blvd (MC-9036), Dallas, TX 75235-9036, USA.
Astroglial Markers Expression in Humans is Like That
in Other Mammals
Previous studies in lower mammals and in developing humans
have shown that vimentin and GFAP are both expressed in the
developing brain, but have mostly focused on the ventricular
zone. We studied the human brain specimens here for the
appearance of these glial markers, with particular attention to
the subplate, to confirm the reliability of the immunochemical
approach used, to confirm the results of previous studies and to
be able to compare the astroglial cell populations with the
observed changes in the neuronal markers. Vimentin staining
was observed in the cells of the generative neuroepithelium
(ventricular zone) and in the cells and fibers of the radial glia.
This staining is like that previously reported in rodents (Bignami
and Dahl, 1989) and prenatal human brain (Sasaki et al, 1988;
Stagaard and M0llg&rd, 1989; Sarnat, 1992). In contrast, using
our fixation and staining protocol, we did not observe major
GFAP staining until ages >26 weeks. This staining was present in
the radial glia, and in maturing astrocytes of the subplate and
cortical layers. Our results are consistent with those of previous
investigations. Some studies have demonstrated the appearance
of GFAP staining in radial glia in the human brain as early as 10
weeks (Antanitus etal, 1976; Stagaard and Mollgard, 1989), like
References
Agmon A, Yang LT, O'Dowd DK, Jones EG (1993) Organized growth of
thalamocortical axons from the deep tier of terminations into layer IV
of developing mouse barrel cortex. J Neurosci 13:5365-5382.
AUendoerfer KL, Shatz CJ (1994) The subplate, a transient neocortical
structure: its role in the development of connections between
thalamus and cortex. Annu Rev Neurosci 17:185-218.
Antanitus DS, Choi BH, Lapham LW (1976) The demonstration of glial
fibrillary acidic protein in the cerebrum of the human fetus by
indirect immunofluorescence. Brain Res 103:613-616.
Antonini A, Shatz CJ (1990) Relation between putative transmitter
phenotypes and connectivity of subplate neurons during cerebral
cortical development. Eur J Neurosci 2:744-761.
Biffo S, Verhaagen J, Schrama LH, Schotman P, Danho W, Margolis FL
(1990) B-50/GAP43 expression correlates with process outgrowth in
the embryonic mouse nervous system. Eur J Neurosci 2:487-499.
Bignami A, Dahl D (1989) Vimentin-GFAP transition in primary
dissociated cultures of rat embryo spinal cord. Int J Devi Neurosci
7:343-357.
Boulder Committee (1970) Embryonic vertebrate central nervous system:
revised. Anat Rec 166:257-262.
Caceres A, Banker G, Steward O, Binder L, Payne M (1984a) MAP2 is
localized to the dendrites of hippocampal neurons which develop in
culture. Brain Res 315:314-318.
Caceres A, Binder LI, Payne MR, Bender P, Rebhun L, Steward O (1984b)
Differential subcellular localization of tubulin and the
micro tubule-associated protein MAP2 in brain tissue as revealed by
804 Immunohistochemistry in Developing Cone
Honig et al.
that seen in rhesus monkey (Levitt and Rakic, 1980). Other
studies of human brain have reported little or no GFAP staining
until -15-25 weeks (Sasaki et al, 1988; Sarnat, 1992). These
differences may relate to differing fixation techniques and
antibodies. Our results are consistent with a generally observed
increase in GFAP staining during cerebral development, and
reciprocally, a maturational decrease in vimentin staining of
astroglial elements, as previously noted in studies of
experimental animals (Chiu and Goldman, 1985; Bignami and
Dahl, 1989; Voigt, 1989).
Conclusions
The changing patterns of immunoreactivity observed for the
neuronal and astroglial markers studied here in a series of human
embryos from mid-gestation to postnatal are similar to those
observed in other mammals. Because of the greater size of the
human subplate and the slower pace of human development, we
suggest that GAP-43 can provide a valuable measure of the
extent of axonal penetration into the differentiating cortical
plate. This may be useful for developmentally staging human
fetal brains. The overall similarity between the behavior of
theimmunocytochemical markers between humans and experimental animals permits greater confidence in extrapolations
made from studies in experimental mammals to human
development. The use of these immunohistochemical markers
may also aid in the investigation of the role of subplate neurons
in normal and pathological conditions.
immunocytochemistry with monoclonal hybridoma antibodies. J
Neurosci 4:394-410.
Catalano SM, Robertson RT, Killackey HP (1991) Early ingrowth of
thalamocortical afferents to the neocortex of the prenatal rat. Proc
Nad Acad Sci USA 88:2999-3003.
Caviness VS Jr, Pinto-Lord MC, Evrard P (1981) The development of
laminated pattern in the mammalian neocortex. In: Morphogenesis
and pattern formation (Connelly TG, Brinkley LL, Carlson BM, eds),
pp 103-126. New York: Raven Press.
Chiu FC, Goldman JE (1985) Regulation of glial fibrillary acidic protein
(GFAP). J Neuroimmunol 8:283-292.
Chun JJM, Shatz CJ (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 JJM, Shatz CJ (1989b) Interstitial cells of the adult neocortical white
matter are the remnant of the early generated subplate neuron
population. J Comp Neurol 282:555-569.
Crandall JE (1989) Developmental regulation of microtubule-associated
protein 2. J Neurochem 52:1910-1917.
Dani JW, Armstrong DM, Benowitz LI (1991) Mapping the development of
the rat brain by GAP-43 immunocytochemistry. Neuroscience
40:277-287.
De Carlos JA, O'Leary DDM (1992) Growth and targeting of subplate
axons and establishment of major cortical pathways. J Neurosci
12:1194-1211.
Erzurumlu RS, Jhaveri S, Benowitz U (1990) Transient patterns of GAP-43
expression during the formation of barrels in the rat somatosensory
cortex. J Comp Neurol 292:443-456.
Friauf E, Shatz CJ (1991) Changing patterns of synaptic input to subplate
and cortical plate during development of visual cortex. J Neurophysiol
66:2059-2071.
Friauf E, McConnell SK, Shatz CJ (1990) Functional synaptic circuits in
the subplate during fetal and early postnatal development of cat visual
cortex. J Neurosci 10:2601-2613.
Ghosh A, Shatz CJ (1992a) Involvement of subplate neurons in the
formation of ocular dominance columns. Science 255:1441-1443.
Ghosh A, Shatz CJ (1992b) Pathfinding and target selection by developing
geniculocortical axons. J Neurosci 12:39-55.
Ghosh A, Shatz CJ (1994) Segregation of geniculocortical afferents during
the critical period: a role for subplate neurons. J Neurosci 14:
3862-3880.
Hendrickson AE, Van Brederode JF, Mulligan KA, Celio MR (1991)
Development of the calcium-binding protein parvalbumin and
calbindin in monkey striate cortex. J Comp Neurol 307:626-646.
Hogan D, Berman NE (1994) The development of parvalbumin and
calbindin-D28k immunoreactive interneurons in kitten visual cortical
areas. Dev Brain Res 77:1 -21.
Kostovic I (1990) Structural and histochemical reorganization of the
human prefrontal cortex during perinatal and postnatal life. Prog
Brain Res 85:223-240.
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.
Kostovic I, Rakic P (1984) Development of prestriate visual projections in
the monkey and human fetal cerebrum revealed by transient
cholinesterase staining. J Neurosci 4:25-42.
Kostovic I, Rakic P (1990) Developmental history of the transient
subplate zone in the visual and somatosensory cortex of the macaque
monkey and human brain. J Comp Neurol 297:441-470.
Kostovic I, Skavic J, Strinovic D (1988) Acetylcholinesterase in the human
frontal associative cortex during the period of cognitive development:
early laminar shifts and late innervation of pyramidal neurons.
Neurosci Lett 90:107-112.
Kostovic I, Lukinovic N, Judas M, Bogdanovic N, Mrzljak L, Zecevic N,
Kubat M (1989) Structural basis of the developmental plasticity in the
human cerebral cortex: the role of the transient subplate zone. Metab
Brain Dis 4:17-23.
Levitt P, Rakic P. (1980) Immunoperoxidase localization of glial fibrillary
acidic protein in radial glial cells and astrocytes of the developing
rhesus monkey brain. J Comp Neurol 193:815-840.
Luskin MB, Shatz CJ (1985a) Neurogenesis of the cat's primary visual
cortex. J Comp Neurol 242:611-631.
Luskin MB, Shatz CJ (1985b) Studies of the earliest generated cells of the
cat's visual cortex: cogeneration of subplate and marginal zones. J
Neurosci 5:1062-1075.
Marin-Padilla M (1971) Early prenatal ontogenesis of the cerebral cortex.
Z Anat Entwickl 134:125-142.
McConnell SK (1991) The generation of neuronal diversity in the central
nervous system. Annu Rev Neurosci 14:269-300.
McConnell SK, Ghosh A, Shatz CJ (1989) Subplate neurons pioneer the
first axon pathway from the cerebral cortex. Science 245:978-982.
Mclntosh H, Daw N, Parkinson D (1990) GAP-43 in the cat visual cortex
during postnatal development. Vis Neurosci 4:585-593.
Meyer G, Wahle P, Castaneyra-Perdomo A, Ferres-Torres R (1992)
Morphology of neurons in the white matter of the adult human
neocortex. Exp Brain Res 88:204-212.
Milosevic A, Kanazir S, Zecevic N (1995) Immunocytochemical
localization of growth-associated protein GAP-43 in early human
development. Dev Brain Res 84:282-286.
Molliver ME, Kostovic I, van der Loos H (1973) The development of
synapses in cerebral cortex of the human fetus. Brain Res 50:403-407.
Neve RL, Perrone-Bizzozero NI, Finklestein S, Zwiers H, Bird E, Kurnit
DM, Benowitz LI (1987) The neuronal growth-associated protein
GAP-43 (B-50, Fl): neuronal specificity, developmental regulation and
regional distribution of the human and rat mRNAs. Mol Brain Res
2:177-183.
Nilaver G, Kozlowski GP (1989) Comparison of the PAP and ABC
immunocytochemical techniques. Tech Immunocytochem 4:
199-215.
Rakic P (1978) Neuronal migration and contact guidance in the primate
telencephalon. Postgrad MedJ 1:25-40.
Rakic P (1983) Geniculo-cortical connections in primates: normal and
experimentally altered development. Prog Brain Res 58:393-404.
Sarnat HB (1992) Regional differentiation of the human fetal ependyma:
immunocytochemical markers. J Neuropathol Exp Neurol 51:58-75.
Sasaki A, Hirato J, Nakazato Y, Ishida Y (1988) Immunohistochemical
study of the early human fetal brain. Acta Neuropathol 76:128-134.
Schmechel DE, Rakic P (1979) A Golgi study of radial glial cells in
developing monkey telencephalon: morphogenesis and transformation into astrocytes. Anat Embryol (Bert) 156:115-152.
Schoenfeld TA, McKerracher L, Obar R, Vallee RB (1989) MAP 1A
and MAP IB are structurally related micro tubule associated proteins
with distinct developmental patterns in the CNS. J Neurosci
9:1712-1730.
Schreyer DJ, Skene JH (1991) Fate of GAP-43 in ascending spinal axons of
DRG neurons after peripheral nerve injury: delayed accumulation and
correlation with regenerative potential. J Neurosci 11:3738-3751.
Seto-Ohshima A, Aoki E, Semba R, Emson PC, Heizmann CW (1990)
Appearance of parvalbumin-specific immunoreactivity in the
cerebral cortex and hippocampus of the developing rat and gerbil
brain. Histochemistry 94:579-589.
Shatz CJ, Luskin MB (1986) The relationship between the
geniculocortical afferents and their cortical target cells during
development of the cat's primary visual cortex. J Neurosci
6:3655-3668.
Shatz CJ, Chun JJM, Luskin MB (1988) The role of the subplate in the
development of the mammalian telencephalon. In: Cerebral cortex
(Peters A.Jones EG, eds), vol 7, pp 35-58. New York: Plenum Press.
Sidman RL, Rakic P (1982) Development of the human central nervous
system. In: Histology and histopathology of the nervous system
(Haymaker W, Adams RD, eds), pp 3-145. Springfield, IL: Charles C.
Thomas.
Sims KB, Crandall JE, Kosik KS, Williams RS (1988) Microtubule-associated protein 2 (MAP2) immunoreactivity in human fetal
neocortex. Brain Res 449:192-200.
Skene JHP (1989) Axonal growth-associated proteins. Annu Rev Neurosci
12:127-156.
Stagaard M, Mollgird K (1989) The developing neuroepithelium in human
embryonic
and
fetal
brain
studied
with
vimentinimmunocytochemistry. Anat Embryol (Berl) 180:17-28.
Stichel CC, Singer W, Heizmann CW, Norman AW (1987)
Immunohistochemical localization of calcium-binding proteins,
parvalbumin and calbindin-D 28k, in the adult and developing visual
cortex of cats: a light and electron microscopic study. J Comp Neurol
262:563-577.
Uylings HB, van Eden CG (1990) Qualitative and quantitative comparison
of the prefrontal cortex in rat and in primates, including humans. Prog
Brain Res 85:31-62.
Valverde F, Facal-Valverde MV (1987) Transitory population of cells in the
temporal cortex of kittens. Brain Res 429:283-288.
Cerebral Cortex Nov/Dec 1996, V6N 6 805
Valverde F, Facal-Valverde MV (1988) Postnatal development of interstitial
(subplate) cells in the white matter of the temporal cortex of kittens:
a correlated Golgi and electron microscopic study. J Comp Neurol
269:168-192.
Van Brederode JF, Mulligan KA, Hendrickson AE (1990) Calcium-binding
proteins as markers for subpopulations of GABAergic neurons in
monkey striate cortex J Comp Neurol 298:1-22.
Van Brederode JF, Helliesen MK, Hendrickson AE (1991) Distribution of
the calcium-binding proteins parvalbumin and calbindin-D28k in the
sensorimotor cortex of the rat. Neuroscience 44:157-171.
806 Immunohistochcmistry in Developing Cortex • Honigctal.
Voigt T (1989) Development of glial cells in the cerebral wall of ferrets. J
Comp Neurol 289:74-88.
Volpe JJ (1989) Current concepts of brain injury in the premature infant.
Am J Roentgenol 153:243-251.
Volpe JJ (1990) Brain injury in the premature infant: is it preventable?
Pediat Res 27:S28-S33.
Wilkinson M, Hume R, Strange R, Bell JE (1990) Glial and neuronal
differentiation in the human fetal brain 9-23 weeks of gestation.
Neuropathol Appl Neurobiol 16:193-204.