Abstract

The maturation of pyramidal neurons in the primary visual cortex (V1) of marmoset monkeys was investigated using an antibody (SMI-32) to non-phosphorylated neurofilament protein (NNF). Analysis of animals aged between birth and postnatal day 91 (PD 91, which corresponds approximately to the peak of synaptogenesis in this species) revealed discrete changes in both the laminar and the areal distribution of NNF. At PD 0, the upper part of layer 6 contained darkly labelled neurons and associated neuropil, including axons. In this layer a centroperipheral gradient, with more labelled cells in the foveal representation, was apparent at PD 0. This topographic gradient gradually disappeared, and by PD 91 a similar density of labelled layer 6 cells was observed throughout V1. Labelled cells were not apparent in layer 3C until PD 7, and were not distributed according to a topographic gradient. Labelled cells were first observed in layer 3Bα at PD 28, when they formed a centroperipheral gradient similar to that seen in layer 6. This gradient was still evident in an adult animal. These results demonstrate an inside-out profile of postnatal cortical development, with the topographic pattern of maturation of V1 mimicking the centroperipheral gradient of maturation in the retina.

Introduction

The development of the primary visual cortex (V1) of primates continues well into the postnatal period, even though key events such as cellular division and migration of neurons to adult-like layers have already occurred (Rakic, 1974, 1985). In the marmoset monkey (Callithrix jacchus, a New World simian), the volume of V1 doubles between postnatal day 0 (PD 0) and postnatal month 3 (PM 3; the peak of synaptogenesis in this species). This rapid growth is characterized by a significant ‘overshoot’ in the volume of V1, followed by reduction, in such a way that V1 attains adult-like dimensions around PM 12 (Missler et al., 1993a). These marked changes are believed to be the result of an increase in the sizes of cell bodies, primarily in layer 2 and the upper part of layer 3, in such a way that these layers reveal the most dramatic changes in thickness and neuronal density (Fritschy and Garey et al., 1986). In parallel, cells throughout V1 undergo a series of additional morphological changes, including synaptogenesis and pruning (Bourgeois and Rakic, 1993; Missler et al., 1993a,b; Rakic, 2002).

In all monkeys, V1 comprises a single continuous retinotopic map of the contralateral hemifield, which is highly consistent across species (Fritsches and Rosa, 1996). This ordered topography allows the analysis of developmental patterns with respect to the retinotopic map. Here, we used immunohistochemical techniques to map the maturation of V1 neurons located in different layers of V1, with respect to the topographic organization of this area.

Neurofilament (NF), a neuronal cytoskeletal protein, comprises three subunits of differing length (NFH, 200 kDa; NFM, 168 kDa; and NFL, 68 kDa). The monoclonal antibody, SMI-32, which recognizes the non-phosphorylated isoforms of the NFH and to a lesser extent NFM subunits (Sternberger and Sternberger, 1983), has been shown to specifically label the dendrites and perikarya of a subset of pyramidal neurons. Discrete laminar expression of non-phosphorylated neurofilament protein (NNF) has been demonstrated in V1 and extrastriate areas of adult macaques, humans (Campbell and Morrison, 1989), marmosets (Bourne and Rosa, 2003) and cats (van der Gucht et al., 2001). Previous studies have also revealed a greater density of protein in areas and layers associated with the magnocellular (M) pathway (Hof and Morrison, 1995; Chaudhuri et al., 1996; Hof et al., 1996; Bourne and Rosa, 2003).

It has been reported that NFH expression arises relatively late in development (Goldstein et al., 1987; Black and Lee, 1988). This subunit has been implicated in late phases of the morphological maturation of cortical neurons, due to the role of NF in cellular structural dynamics (Lasek, 1981; Sawant et al., 1994). Previous studies in neonatal human, vervet monkey and cat V1 have demonstrated temporal changes in the laminar distribution and morphology of neurons labelled with an antibody to non-phosphorylated neurofilament protein (SMI-32) in early postnatal life (Ang et al., 1991; Liu et al., 1994; Kogan et al., 2000). In this study, we investigated the temporal maturation of the V1 topographical map, by profiling changes in NNF distribution in different parts of V1 of marmoset monkeys. Our results indicate an asynchronous maturation of different parts of V1, with the peripheral representation lagging behind the foveal representation. They also reveal a protracted development of the supragranular layers, which continues for several months postnatally. Finally, comparison of marmosets with humans and macaques reveals a relatively immature organization of V1 at birth, which may prove a significant advantage for future studies of primate cortical development.

Materials and Methods

Animals

Eight marmoset monkeys (Callithrix jacchus), aged PD 0 (2–5 h post-partum), PD 3, PD 7 (two animals), PD 14, PD 28, PD 91 and PM22 (a sexually mature individual), were used in this study. For the neonates, no siblings were used and the largest of the two or three offspring was always chosen for study. Moreover, we compared these individuals against normal developmental growth curves for the species (Hearn, 1983), thereby avoiding the inclusion of abnormally large or small specimens. Experiments were conducted in accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes and were approved by the Monash University Animal Ethics Committee.

Tissue Preparation

Animals were administered an overdose of the opiate sufentanil citrate (0.05 mg/kg, i.m.) and transcardially perfused with 0.1 M heparinized phosphate buffer (PB; pH 7.2, warmed to 37°C) containing 0.1% sodium nitrite, followed by 4% paraformaldehyde in 0.1 M PB at room temperature. Cerebral tissues were immediately removed and postfixed for 24 h in 2% paraformaldehyde in 0.1 M PB containing 5% sucrose. Cryoprotection was achieved by serially transferring the tissue through solutions of 10, 20 and 30% sucrose in 0.1 M PB before freezing (using liquid nitrogen) and storing at −70°C. Parasagittal serial sections, 40 μm thick, were obtained using a cryostat, and were subsequently stored at −20°C in a cryoprotection solution (50% 0.5 M PB, 30% ethylene glycol, 20% glycerol). Adjacent sections to those kept for immunohistochemistry were reacted for cytochrome oxidase and Nissl substance.

Immunohistochemistry

Free-floating sections were initially washed in 0.1 M PB for 20 min, followed by 0.1 M sodium phosphate buffer (PBS; Dulbecco A, Oxoid, Basingstoke, UK) for a further 20 min, and finally PBS containing 0.3% Triton X-100 (PBS–TX; BDH, Poole, UK) for 15 min. Sections were then treated with 0.3% hydrogen peroxide in 50% methyl alcohol made up with PBS for 30 min before being preincubated in a solution of 5% normal rabbit serum (NRS; Vector Laboratories, Burlingame, CA) in PBS–TX for 1 h. Without rinsing, sections were transferred into PBS–TX containing the primary monoclonal antibody, which recognizes the high and medium weight subunits of NNF (SMI-32, 1:2000; Sternberger Monoclonals, Baltimore, MD), and 2% NRS for 16–18 h at 4°C. Sections were then rinsed (3 × 10 min) with 0.3% Tween-20 (Sigma, St Louis, MO) in PBS prior to incubation for 1 h with a biotinylated rabbit anti-mouse secondary antibody (1:500; DAKO A/S, Glöstrup, Denmark) in PBS containing 1% NRS and 1% bovine serum albumin (fraction V; Sigma, St Louis, MO). Following this, they were processed with avidin–biotin–horseradish peroxidase (1:200; Amersham Pharmacia Biotech, Little Chalfont, UK) in PBS for 1 h. Immunoreactivity was revealed using a metal-enhanced chromogen, 3,3′-diaminobenzidine (DAB) and a stable peroxide buffer (Pierce Biotechnology Inc, Rockford, IL) for 3–6 min for visualization. Unless otherwise stated, all incubations were at room temperature with gentle agitation, and tissue was rinsed between steps in PBS (3 × 10 min). Sections were then mounted, dehydrated in graded alcohols, defatted in xylene and coverslipped with DPX (BDH, Poole, UK). Negative and positive controls were performed routinely.

Histology

Sections adjacent to those kept for SMI-32 labelling were reacted for cytochrome oxidase (CO) and Nissl substance. CO reactivity was performed on free-floating sections, using a modification of the original Wong-Riley method (Wong-Riley, 1979). First, sections were pre-treated in 0.1% cobalt chloride solution at room temperature before being incubated at 37°C in 0.1 M PB containing a mixture of cytochrome C oxidase, catalase and DAB chromogen. Sections were reacted until layer 4 of V1 was clearly discernible from the infra/supra granular layers, or for a maximum of 2 h. Nissl substance labelling was achieved using a 0.05% cresyl violet solution. Sections were stained until layer 6 was discernible from other layers. Sections were examined by light microscopy as described for immunohistochemistry.

Qualitative/Quantitative Analysis

Sections were examined under brightfield microscopy using a Zeiss Axioplan imaging microscope. Low- and high-power photomicrographs (1300 × 1030 dpi) were taken on a Zeiss Axiocam digital camera connected to Axiovision software (v 3.1; Zeiss). Images were cropped and sized using Adobe Photoshop 8 and Illustrator 11.

The laminar organization of V1 was assessed using the CO and Nissl-stained sections (Fig. 1). The location of labelled cells was described according to a nomenclature (Spatz et al., 1970) which allows a better comparison between areas and species (Casagrande and Kaas, 1994; Xu et al., 2003), as opposed to the more widely used Brodmann scheme. Of particular relevance for the present paper is the fact that Brodmann's layers 4A and 4B are considered parts of layer 3 (sublayers 3Bβ and 3C, respectively), while Brodmann's layer 3B is referred to as sublayer 3Bα. Cell counts were made in 500 μm linear transects through layers 3Bα, 3C and 6, randomly taken from left hemisphere parasagittal sections (n = 10 for each layer and topographic region; described below).

Figure 1.

Photomicrographs of adjacent sagittal sections from operculum V1 of a postnatal day (PD) 0 marmoset monkey, illustrating how it was possible to demarcate the layers in the NNF immunolabelled section (A) by comparison with more traditional histological staining methods. In the cytochrome oxidase stained section (B), the layers are clearly discernible in terms of their density of staining (layers 3Bβ, 4α, 4β and 6 darker), and the Nissl substance-stained section (C) revealing denser labelling in layers 4 and 6. The cytochrome oxidase stained section also reveals the immature configuration of layer 4, with sublayer 4α darker than 4β (Spatz et al., 1993). Scale bar 200 μm.

Figure 1.

Photomicrographs of adjacent sagittal sections from operculum V1 of a postnatal day (PD) 0 marmoset monkey, illustrating how it was possible to demarcate the layers in the NNF immunolabelled section (A) by comparison with more traditional histological staining methods. In the cytochrome oxidase stained section (B), the layers are clearly discernible in terms of their density of staining (layers 3Bβ, 4α, 4β and 6 darker), and the Nissl substance-stained section (C) revealing denser labelling in layers 4 and 6. The cytochrome oxidase stained section also reveals the immature configuration of layer 4, with sublayer 4α darker than 4β (Spatz et al., 1993). Scale bar 200 μm.

The relationship between labelled neurons and the visuotopic organization of V1 was assessed by comparing the location of the cells with series of parasagittal sections from animals subject to extensive electrophysiological recording from V1 (Fritsches and Rosa, 1996). For analysis, transects were grouped according to three visuotopic regions: central (0–3° eccentricity, corresponding to the occipital operculum, OP), intermediate (up to 30° eccentricity, corresponding to the caudal half of the calcarine sulcus, CC), and peripheral (>30° eccentricity, corresponding to the rostral half of the calcarine sulcus, RC). Care was taken to include equal proportions of transects including the putative upper and lower visual field representations. Counts obtained in different regions were compared in order to establish any statistically significant trends, using a one-way Analysis of Variance, followed by a least significance difference (LSD) post hoc analysis (P = 0.05 as the limit of significance). The results of the ANOVA are shown in Table 1, while the statistical significance values revealed by LSD post hoc analyses are indicated in parentheses in the corresponding section of RESULTS. Finally, comparison of counts obtained in the two PD 7 animals revealed no statistically significant difference between animals of the same age, in any of the analysed visuotopic sectors (two-tailed t-test; P = 0.11).

Table 1

Analyses of variance

Source of variance
 
Layer 6
 
  Layer 3C
 
  Layer 3Bα
 
  

 
df
 
F
 
P
 
df
 
F
 
P
 
df
 
F
 
P
 
PD 0 2, 29 31.365 <0.001  n/a   n/a  
PD 3 2, 29 34.074 <0.001  n/a   n/a  
PD 7 2, 29 51.641 <0.001 2, 29 0.730 0.930  n/a  
PD 14 2, 29 70.440 <0.001 2, 29 0.126 0.882  n/a  
PD 28 2, 29 156.54 <0.001 2, 29 1.195 0.318 2, 29 29.174 <0.001 
PD 91 2, 29 2.096 0.142 2, 29 0.973 0.391 2, 29 141.07 <0.001 
PM 22
 
2, 29
 
1.718
 
0.198
 
2, 29
 
0.275
 
0.762
 
2, 29
 
14.254
 
<0.001
 
Source of variance
 
Layer 6
 
  Layer 3C
 
  Layer 3Bα
 
  

 
df
 
F
 
P
 
df
 
F
 
P
 
df
 
F
 
P
 
PD 0 2, 29 31.365 <0.001  n/a   n/a  
PD 3 2, 29 34.074 <0.001  n/a   n/a  
PD 7 2, 29 51.641 <0.001 2, 29 0.730 0.930  n/a  
PD 14 2, 29 70.440 <0.001 2, 29 0.126 0.882  n/a  
PD 28 2, 29 156.54 <0.001 2, 29 1.195 0.318 2, 29 29.174 <0.001 
PD 91 2, 29 2.096 0.142 2, 29 0.973 0.391 2, 29 141.07 <0.001 
PM 22
 
2, 29
 
1.718
 
0.198
 
2, 29
 
0.275
 
0.762
 
2, 29
 
14.254
 
<0.001
 

One-way ANOVA; values are cell body count per 500 μm transect. df values: corrected model, corrected total.

In order to control for the possibility of topographic differences in the density of NNF-immunoreactive neurons being a reflection of generalized differences in cell density between different parts of V1, we have also investigated the cell density in layers 3Bα, 3C and 6 using adjacent Nissl-stained sections (Table 2). Colour micrographs of 500 × 100 μm transects centred on these layers were adjusted for contrast to optimise counting using Adobe Photoshop 8. As with the SMI-32 cell counts, different transects were obtained in the three visuotopic regions.

Table 2

Nissl positive cells/500×100 μm transect


 
Layer 6
 
Layer 3C
 
Layer 3Bα
 
PD 0    
    OP 759 534 516 
    CC 741 505 540 
    RC 772 534 538 
PD 7    
    OP 323 364 392 
    CC 326 343 412 
    RC 331 345 409 
PD 14    
    OP 399 371 337 
    CC 408 376 343 
    RC 402 383 352 
PD 28    
    OP 506 387 438 
    CC 522 401 422 
    RC 501 402 437 
PD 91    
    OP 335 297 298 
    CC 328 282 294 
    RC 310 286 281 
PM 22    
    OP 478 294 306 
    CC 492 312 310 
    RC
 
471
 
303
 
301
 

 
Layer 6
 
Layer 3C
 
Layer 3Bα
 
PD 0    
    OP 759 534 516 
    CC 741 505 540 
    RC 772 534 538 
PD 7    
    OP 323 364 392 
    CC 326 343 412 
    RC 331 345 409 
PD 14    
    OP 399 371 337 
    CC 408 376 343 
    RC 402 383 352 
PD 28    
    OP 506 387 438 
    CC 522 401 422 
    RC 501 402 437 
PD 91    
    OP 335 297 298 
    CC 328 282 294 
    RC 310 286 281 
PM 22    
    OP 478 294 306 
    CC 492 312 310 
    RC
 
471
 
303
 
301
 

OP, operculum; CC, caudal calcarine; and RC, rostral calcarine.

Results

The maturation of the subset of NNF-immunoreactive pyramidal cells in marmoset V1 is clearly evident during the first three months of postnatal development. Accurate delimitation of the laminar and areal boundaries of the NNF-immunolabelled sections (Fig. 1A) was achieved by comparing adjacent CO (Fig. 1B) and Nissl substance-stained (Fig. 1C) sections.

Laminar Development of V1

Even at PD 0, a subset of NNF-immunoreactive pyramidal cells could be observed in layer 6, near its interface with layer 5 (Fig. 2A). This single-layered distribution was observed throughout the occipital operculum, as well as the caudal and rostral halves of the calcarine sulcus (Fig. 3A). Although no other layer revealed any NNF-immunoreactive cells, within layer 5 there was extensive labelling of the neuropil processes, which also extended into the white matter (Figs 1A and 3A), but not the granular and supragranular layers. The distribution of NNF-immunoreactive neurons stopped abruptly at the outer border of V1, with the adjacent extrastriate cortex being entirely free of label (Fig. 3A).

Figure 2.

Photomicrographs of sagittal NNF immunolabelled sections from similar segments of opercular V1, at different postnatal stages — PD 0 (A), PD 3 (B), PD 7 (C), PD 14 (D), PD 28 (E), PD 91 (F) and PM 22 — revealing the laminar maturation profile of cells in layers 6, 3C and 3Bα. Layer 6 cells were present at PD 0, whereas cells in layers 3C and 3Bα were not observed until PD 7 and PD 28, respectively. Laminar demarcation achieved from adjacent Nissl substance and cytochrome oxidase staining (see Fig. 1). Scale bar 200 μm.

Figure 2.

Photomicrographs of sagittal NNF immunolabelled sections from similar segments of opercular V1, at different postnatal stages — PD 0 (A), PD 3 (B), PD 7 (C), PD 14 (D), PD 28 (E), PD 91 (F) and PM 22 — revealing the laminar maturation profile of cells in layers 6, 3C and 3Bα. Layer 6 cells were present at PD 0, whereas cells in layers 3C and 3Bα were not observed until PD 7 and PD 28, respectively. Laminar demarcation achieved from adjacent Nissl substance and cytochrome oxidase staining (see Fig. 1). Scale bar 200 μm.

Figure 3.

NNF immumoreactivity in approximately corresponding sagittal sections the occipital lobe of animals aged (A) PD 0, (B) PD 3, (C) PD 7, (D) PD 14, (E) PD 28, (F) PD 91 and (G) PM 22. In comparison with the relatively early maturation of V1, there is a lack of immunoreactivity in the adjacent area V2 until PD 28. At PD 28, expression is evident in layer 3B of V2 but is not as dense as that observed at PD 91 or the adult (PM 22). Also note the heavy immunostaining of the white matter at early postnatal ages, and the gradual disappearance of this staining as the axons of presumed corticofugal neurons become phosphorylated (Matus, 1988). The arrow points to the border between V1 and dorsal V2, and the insert (top right) indicates the anatomical axes (C = caudal, D = dorsal, R = rostral, V = ventral). Scale bar = 5 mm.

Figure 3.

NNF immumoreactivity in approximately corresponding sagittal sections the occipital lobe of animals aged (A) PD 0, (B) PD 3, (C) PD 7, (D) PD 14, (E) PD 28, (F) PD 91 and (G) PM 22. In comparison with the relatively early maturation of V1, there is a lack of immunoreactivity in the adjacent area V2 until PD 28. At PD 28, expression is evident in layer 3B of V2 but is not as dense as that observed at PD 91 or the adult (PM 22). Also note the heavy immunostaining of the white matter at early postnatal ages, and the gradual disappearance of this staining as the axons of presumed corticofugal neurons become phosphorylated (Matus, 1988). The arrow points to the border between V1 and dorsal V2, and the insert (top right) indicates the anatomical axes (C = caudal, D = dorsal, R = rostral, V = ventral). Scale bar = 5 mm.

By PD 3, the NNF-immunoreactive pyramidal cells in layer 6 (Fig. 2B) no longer clustered as tightly near the interface with layer 5. As at PD 0, no other cortical layer revealed NNF-immunoreactive cells. However, at this stage there was substantial immunolabelling of neuropil processes within layer 4β, which was denser in comparison with that observed in layer 5; much weaker neuropil staining was also observed in the top part of layer 4 (4α). Axons in the white matter continued to be densely stained for NNF (Fig. 2B).

By PD 7, NNF-immunoreactive pyramidal cells were observed in supragranular layer 3C (Fig. 1C). The network of neuropil processes was more extensive and radially ordered through layers 5 and 4β, probably due to immunolabelling of the apical dendrites of cells resident in layer 6, while in layer 4α it was more punctate (Fig. 2C). The white matter fibres continued to be heavily NNF immunoreactive (Figs. 2 and 3), and no NNF-immunoreactive cells were observed in the second visual area, V2 (Fig. 3C).

Between PD 7 and PD 14, layers 3C and 6 remained the only laminae to contain NNF-immunoreactive cells (Fig. 2D). The white matter continued to be densely labelled, while the cortex adjacent to V1 remained devoid of any immunolabelled neurons (Fig. 3D).

At PD 28 we observed the first evidence of small, lightly labelled cells in layer 3Bα (Fig. 2E), while cells in layers 3C and 6 remained NNF immunoreactive (Fig. 2E). Also apparent at this stage was the first appearance of immunolabelled neurons in V2, which were concentrated at the caudal border with V1 (Fig. 3E).

At the peak of synaptogenesis in V1 (PD 91; Fig. 2F), the pattern of NNF labelling resembled that observed in the adult monkey (PM 22; Fig. 2G), with layer 3Bα being the only lamina to show significant changes in the density of NNF-immunoreactive cells between PD 91 and PM 22 (Fig. 5A; see below). Moreover, this stage was marked by the absence of immunostained white matter fibres (Fig. 3F), and the presence of more substantial NNF immunolabelling of the extrastriate cortex. Both of these characteristics reveal a pattern more akin to that observed in the adult (PM 22; Fig. 3G).

Cellular Maturation

At PD 0, the strongest immunostaining was observed in the perikaryon of layer 6 cells, with minimal staining of the basal dendrites (Fig. 4I). Neurons exhibiting a true apical dendrite directed towards the pia mater were not evident.

Figure 4.

Photomicrographs of NNF labelled cells from layers 3Bα, 3C and 6 of the operculum V1 at the different stages of postnatal development studied, revealing their morphology. Evident are the immaturity of the pyramidal cells at their first appearance, as well as the subsequent growth of the cell body and apical dendrite, and the acquisition of a more complex basal dendritic tree. Scale bar 5 μm.

Figure 4.

Photomicrographs of NNF labelled cells from layers 3Bα, 3C and 6 of the operculum V1 at the different stages of postnatal development studied, revealing their morphology. Evident are the immaturity of the pyramidal cells at their first appearance, as well as the subsequent growth of the cell body and apical dendrite, and the acquisition of a more complex basal dendritic tree. Scale bar 5 μm.

In terms of the morphological characteristics of the layer 6 cells at PD 3 (Fig. 4J), the immunolabelling was denser in the perikaryon, revealing a more pyramid-like structure. When clearly identifiable, apical dendrites were relatively thin.

The morphology of the cells at PD 7 in layer 3C (Fig. 4D) was similar to that observed in layer 6 at PD 0 (Fig. 4I). Cell bodies had a more bulbous shape, were relatively small, and had immunolabelling principally restricted to the perikaryon and proximal dendritic branches. Layer 6 cells now were larger, and typically showed an unmistakable (albeit thin) apical dendrite (Fig. 4K), as well as more extensive labelling of the basal dendrites.

By PD 14, the NNF immunolabelling in layer 6 cells had become much denser (Fig. 4L), including both the perikaryon and the dendrites, while the cells in layer 3C (Fig. 4E) remained immature in appearance.

The lightly labelled cells in 3Bα observed at PD 28 had a rather immature morphology (Fig. 4A), with a small soma and immunolabelling restricted principally to the perikaryon and the proximal portion of the radiating dendrites. Cells in layers 3C and 6 were more strongly labelled, and had larger cell bodies with stouter apical dendrites.

The morphology of the cells in all laminae appeared mature by PD 91 (Fig. 4B,G,N), characterized by an increase in the calibre of the apical dendrite of pyramidal cells in layers 3C and 6 (Fig. 4G,N).

The morphological characteristics observed at PM 22 (Fig. 4C,H,O) were no different from those seen at PD 91 (Fig. 4B,G,N), except for denser labelling throughout the perikaryon and a further increase in the calibre of the apical dendrite.

Topographic Gradients

At PD 0, in layer 6, quantitative analysis revealed a difference in the density of NNF-immunoreactive cell bodies in different parts of V1 (Fig. 5C, PD 0). The number of labelled cell bodies per 500 μm linear transect was significantly higher in the operculum (12.3 ±1.9) than in the caudal (7.9 ± 1.7) or rostral (6.1 ± 1.4) sectors of the calcarine sulcus (Fig. 5C, a, b, c; P < 0.01; see Table 1). Even within the calcarine cortex, a significant difference was found between the intermediate and peripheral representation sectors (Fig. 5C, b, c; P = 0.03).

Figure 5.

Quantitative analysis of the profile of NNF immunolabelled cells relative to the visuotopic map of the marmoset monkey primary visual cortex in (A) layer 3Bα; (B) layer 3C; and (C) layer 6. (D) Schematic of the topographic organisation of the marmoset V1 indicating the topographic subsections used in this analysis. Each histogram represents the mean ± SD of NNF immunoreactive cell bodies in a 500 μm linear transect from three divided topographical subsections of V1, for each animal. Different letters above adjacent histograms indicate that a statistical difference was established (one-way ANOVA followed by LSD post hoc analysis; P ≤ 0.05), while the same letters or no lettering indicates no significant difference.

Figure 5.

Quantitative analysis of the profile of NNF immunolabelled cells relative to the visuotopic map of the marmoset monkey primary visual cortex in (A) layer 3Bα; (B) layer 3C; and (C) layer 6. (D) Schematic of the topographic organisation of the marmoset V1 indicating the topographic subsections used in this analysis. Each histogram represents the mean ± SD of NNF immunoreactive cell bodies in a 500 μm linear transect from three divided topographical subsections of V1, for each animal. Different letters above adjacent histograms indicate that a statistical difference was established (one-way ANOVA followed by LSD post hoc analysis; P ≤ 0.05), while the same letters or no lettering indicates no significant difference.

The centroperipheral gradient along the rostrocaudal axis was still clearly evident in layer 6 at PD 3 (Fig. 5C), with a higher density of NNF-immunoreactive cell bodies in the operculum than in the rostral and caudal calcarine sulcus (Fig. 5C, a, b, c; P < 0.01), and with the caudal calcarine sulcus having a higher density than the rostral segment (Fig. 5C, b, c; P = 0.04).

By PD 7, a significant (P < 0.01) centroperipheral gradient in labelled cell density was still present in layer 6 (Fig. 5C). However, this was primarily due to a clear difference between the opercular and calcarine cortices. From this stage of development on, the intermediate and peripheral subsectors of the calcarine sulcus were indistinguishable (Fig. 5C, PD7, b, b; P = 0.144). Interestingly, unlike in layer 6, even at this stage of development no centroperipheral gradient of NNF-immunoreactive cells could be identified in layer 3C (Fig. 5B, PD 7; P > 0.74).

The same trends observed at PD 7 were apparent at PD 14. In layer 6, the areal density of NNF-immunoreactive cells was significantly different between the operculum and the calcarine sulcus (Fig. 5C, PD 14; a, b, b; P < 0.01), but not between the rostral and caudal subsectors of the calcarine cortex (b, b; P = 0.252). Although limited inferences can be made of trends between cases, given the possibility of technical factors such as differential shrinkage affecting absolute values, there is a suggestion that at this stage of development the density of NNF-immunoreactive cells in the opercular layer 6 had reached levels which are within the range of variation observed in older animals, including the adult case (Fig. 5C). Changes in layer 3C were also primarily in the form of a higher overall density of labelled cells; however, there was still no evidence of a topographic gradient (Fig. 5B, PD 14).

At PD 28, significant differences in the density of NNF immunoreactive cells were observed in layer 3Bα between the central, intermediate and peripheral representations (Fig. 5A, PD 28; a, b, c; P ≤ 0.002), and also between the two calcarine regions (b, c; P < 0.001). In layers 3C and 6 (Fig. 5B,C, PD 28), the cellular densities were similar to those seen at PD 14.

At PD 91, the densities of NNF-immunoreactive cells in layers 3C and 6 resembled the adult pattern (Fig. 5B,C; PD 91 versus PM 22), including the disappearance of the layer 6 centroperipheral gradient. Although there was no significant difference in the number of NNF-immunoreactive cells in layers 3C and 6 along the centroperipheral axis of V1, there remained a definite gradient in the neuropil labelling (Fig. 3F), which was stronger in the operculum. In layer 3Bα, we observed a significant centroperipheral gradient of NNF-immunoreactive cells (P < 0.01; see Fig. 5A, PD 91), coupled with a suggestion of an overall increase in labelled cell density in comparison with PD 28.

Interestingly, in the adult monkey (PM 22), layer 3Bα in the operculum still showed a significantly higher density of NNF-immunoreactive pyramidal cells in comparison with that seen in the calcarine sulcus (Fig. 5C, a, b, b; P ≤ 0.001). However, there was no significant difference between the rostral and caudal sectors of the calcarine cortex.

Cell counts in Nissl-stained sections (Table 2) revealed no differences in cell density between the central (OP), intermediate (CC) and peripheral (RC) regions of V1 at any of the postnatal ages examined. Confirming the results of Fritschy and Garey (1986), the cellular density in each of the three sampled layers was higher at PD0, but this quickly normalized, in such a way that counts observed at PD 7 were within the range of variation found in older animals. Thus, trends in neuronal density observed between topographic regions of the same animal cannot be attributed solely to more general developmental changes in the cellular structure of different parts of V1.

Discussion

We have mapped the sequence of maturation of NNF-immunoreactive pyramidal cells in the striate cortex of marmosets. In addition to its comparative value (this being the first description of the developmental changes in SMI-32 staining in a New World monkey species), we have uncovered new features of the development of primate V1: first, there is a topographic gradient of maturation, with the foveal representation leading the way, and second, this gradient is not apparent in all cortical layers.

Laminar and Morphological Maturation

The maturation of NNF-immunoreactive cells followed the well-described ‘inside-out’ gradient of cortical development along the radial axis of the cortex (Rakic, 1982; Takahashi et al., 1999), with layer 6 cells maturing first. This type of gradient has also been described with respect to the maturation of cortical synapses (Blue and Parnavelas, 1983; Zielinski and Hendrickson, 1992) and the onset of functionality of pyramidal cells (Rumpel et al., 2004).

The first cells observed in layer 6 soon after birth are likely to correspond to the highly structured small/medium-sized pyramidal cells which project to the dorsal lateral geniculate nucleus (dLGN; Lund et al., 1975); these cells also receive geniculate input (Spatz, 1979), which develops prior to geniculate input to layer 4 (Crair et al., 2001). The early maturation of putative corticothalamic layer 6 cells suggests that reciprocal connectivity with the dLGN is a central component of laminar maturation and overall development of V1. As far as it is possible to determine from NNF immunohistochemistry, layer 6 pyramidal cells display an adult-like cellular morphology relatively early (between the second and fourth week of postnatal life), while cells in other layers still show relatively weak staining, particularly of dendrites and thin apical dendrites.

The NNF-immunoreactive subset of cells in layers 3C and 3Bα (which receive geniculate input via layers 4α, 4β and 6) were not morphologically mature until around the peak of synaptogenesis (PD 91). In addition, the up-regulation of the NNF protein in this subset of cells was not evident until the end of the first and fourth postnatal weeks, respectively. Although subplate activity could also instruct early development of layer 2/3 horizontal connections (Grossberg and Seitz, 2003), it is also known that several types of layer 6 pyramidal cells have their apical dendrite and recurrent axon arborization in the supragranular layers (Lund et al., 1977). Thus, they could be influencing maturation of the supragranular layers in an activity-dependent manner.

As soon as it becomes evident in a given layer, the NNF immunostaining is principally restricted to the perikaryon, and progresses radially from the nucleus in a time-dependent manner. This profile is expected considering that NF protein is transcribed in the nucleus and subsequently migrates to the distal extent of dendrites and the axon. Moreover, this same trend has been previously observed in both vervet monkey (Kogan et al., 2000) and human (Ang et al., 1991). Unlike in previous studies (Kogan et al., 2000), in our data there was no clear evidence to suggest that the largest pyramidal cells in a given layer are the first to up-regulate synthesis of NNF protein. However, this may be difficult to assess due to the dynamic changes observed in soma size of marmoset pyramidal cells during this period (Missler et al., 1993b).

The temporal profile of maturation of the NNF immunoreactive layers in V1 differs between primate species. For example, in the marmoset monkey we have shown that the first evidence of labelled neurons in layer 3C does not occur until 3–7 days postnatally, while such cells are already present at birth in the vervet monkey (Kogan et al., 2000; note that 3C corresponds to Brodmann's layer 4B in that study). Conversely, the corresponding label is not seen until 5 months postnatally in humans (Ang et al., 1991). Whether the differences observed between primates relate to the gestational period is not known; however, there is evidence to suggest that they correlate with the duration of the critical period (Ang et al., 1991). Finally, the M-dominated layer 3C, which in marmosets contains pyramidal cells that project directly to the middle temporal area (vogt Weisenhorn et al., 1995), develops relatively early. An early maturation of the cortical regions dominated by the M system has previously been suggested on the basis of anatomical studies in the marmoset (Spatz et al., 1993; Fonta and Imbert, 2002), as well as both anatomical (Gottlieb et al., 1985; Kennedy et al., 1985; Lachica and Casagrande, 1988; Hawken et al., 1997; Kogan et al., 2000) and functional (Mikami and Fujita, 1992; Kovacs et al., 1999) studies in humans and other non-human primates.

Topographical Maturation

In layers 6 and 3Bα, early stages of development were characterized by a distinct centroperipheral gradient in the staining profile, with a higher density of immunoreactive neurons in the central representation. However, at least in the case of layer 6, this gradient was transient, and by the end of the third postnatal month all regions of V1 showed a similar density of labelled cells. Despite the marked increase in the density of labelled cells, in a 22-month-old animal there was still evidence of a topographic gradient in layer 3Bα, which could be interpreted as evidence of an incomplete maturation of the peripheral representation. This is an interesting aspect of the data, as it suggests that the structural development of the top tier of layer 3 could proceed well into adulthood. In marmosets, the period of greatest rearrangement of V1 circuitry is completed within 9 months of birth (Fritschy and Garey, 1986; Spatz, 1989; Spatz et al., 1993); full sexual maturity in this species is reached by 18 months, and the typical lifespan in captivity is 8–12 years (Hearn, 1983). In other species, visuotopic changes induced in adult animals by retinal lesions seem to be more prominent in the supragranular layers (Waleszczyk et al., 2003).

Although centroperipheral gradients are common features of the development of visual structures, it is difficult to attribute causal relationships between events occurring in other parts of the nervous system and the changes observed in V1. For example, during retinal neurogenesis, central (foveal) neurons are generated ahead of peripheral neurons (Rapaport and Stone, 1984; La Vail et al., 1991), a temporal gradient which also seems to have implications for the development of the optic nerve (Reese and Cowey, 1990). In addition, certain features of retinal dendritic morphology (Kirby and Steineke, 1991), synaptogenesis (Hendrickson, 1996) and neurochemistry (Nag and Wadhwa, 1996) also develop along a centroperipheral gradient. However, most of these events happen prenatally, while the changes we observed in V1 extend well into postnatal life. Further, functional changes in the visual system can reveal an opposite gradient, with the peripheral visual field maturing first (Blakemore and Vital-Durand, 1986; Kiorpes and Kiper, 1996).

The correlation between the topographical maturation profile of the retina and that of layer 6 in the early postnatal period could indicate that retinal input is an important driving force in this process, provided that the temporal gradient of innervation observed in the retinothalamic projections ‘carried over’ to the geniculocortical projection. Although we are unaware of direct evidence of this, such a model would carry the prediction that geniculocortical synapses in the foveal representation of V1 would become functional ahead of similar synapses in the peripheral representation (despite the shorter path for axons innervating the calcarine cortex). In contrast, the absence of a topographical density gradient in layer 3C at any stage of development would suggest that a different set of factors are involved in the maturation of this subset, which are not directly dependent on input from layer 6. Indeed, our observations reveal an interesting paradox, in that areas corresponding to the fovea, largely driven by parvocellular (P) input (Schein and de Monasterio, 1987), develop earlier than the more peripheral areas. In addition, the centroperipheral gradient of maturation is only observed in the layers in which there is an input from both the M and P systems. One possibility is that the profile of this gradient is also determined in part by the timing of arrival of retinal ganglion cell axons to the lateral geniculate nucleus. As described by Meissirel et al. (1997), the M layers are innervated at a later stage of development, in comparison with the P layers.

In summary, our data suggest that the maturation of the topographic map of marmoset monkey V1 is a complex process. The laminar and areal gradients observed during the maturation process suggest that foveal representation matures before the peripheral representation, paralleling the developmental profile of the retina. Although NNF-immunoreactive cells are evident at birth, their morphological maturation continues throughout the entire period when V1 cells are forming and refining their synaptic connections.

This work was supported by grants from the Australian Research Council, National Health and Medical Research Foundation of Australian, Clive and Vera Ramaciotti Foundation and the Willliam Buckland Foundation. The authors are grateful to Mr D. Caddy for advice on statistical analysis of these data, Mr A. Hind for his technical assistance with some of the experiments and Ms R. Tweedale for critically reading drafts of the manuscript.

References

Ang LC, Munoz DG, Shul D, George DH (
1991
) SMI-32 immunoreactivity in human striate cortex during postnatal development.
Brain Res Dev Brain Res
 
61
:
103
–109.
Black MM, Lee VM (
1988
) Phosphorylation of neurofilament proteins in intact neurons: demonstration of phosphorylation in cell bodies and axons.
J Neurosci
 
8
:
3296
–3305.
Blakemore C, Vital-Durand F (
1986
) Organization and post-natal development of the monkey's lateral geniculate nucleus.
J Physiol
 
380
:
453
–491.
Blue ME, Parnavelas JG (
1983
) The formation and maturation of synapses in the visual cortex of the rat. I. Qualitative analysis.
J Neurocytol
 
12
:
599
–616.
Bourgeois JP, Rakic P (
1993
) Changes of synaptic density in the primary visual cortex of the macaque monkey from fetal to adult stage.
J Neurosci
 
13
:
2801
–2820.
Bourne JA, Rosa MGP (
2003
) Neurofilament protein expression in the geniculostriate pathway of a New World monkey (Callithrix jacchus).
Exp Brain Res
 
150
:
19
–24.
Campbell MJ, Morrison JH (
1989
) Monoclonal antibody to neurofilament protein (SMI-32) labels a subpopulation of pyramidal neurons in the human and monkey neocortex.
J Comp Neurol
 
282
:
191
–205.
Casagrande VA, Kaas J (
1994
) The afferent, intrinsic and efferent connections of primary visual cortex in primates. In: Cerebral cortex, vol. 10. Primary visual cortex in primates (Peters A, Rockland KS, eds), pp. 201–259. New York: Plenum.
Chaudhuri A, Zangenehpour S, Matsubara JA, Cynader MS (
1996
) Differential expression of neurofilament protein in the visual system of the vervet monkey.
Brain Res
 
709
:
17
–26.
Crair MC, Horton JC, Antonini A, Stryker MP (
2001
) Emergence of ocular dominance columns in cat visual cortex by 2 weeks of age.
J Comp Neurol
 
430
:
235
–249.
Fonta C, Imbert M (
2002
) Vascularization in the primate visual cortex during development.
Cereb Cortex
 
12
:
199
–211.
Fritsches KA, Rosa MGP (
1996
) Visuotopic organisation of striate cortex in the marmoset monkey (Callithrix jacchus).
J Comp Neurol
 
372
:
264
–282.
Fritschy JM, Garey LJ (
1986
) Quantitative changes in morphological parameters in the developing visual cortex of the marmoset monkey.
Brain Res
 
394
:
173
–188.
Goldstein ME, Sternberger LA, Sternberger NH (
1987
) Varying degrees of phosphorylation determine microheterogeneity of the heavy neurofilament polypeptide (Nf-H).
J Neuroimmunol
 
14
:
135
–148.
Gottlieb MD, Pasik P, Pasik T (
1985
) Early postnatal development of the monkey visual system. I. Growth of the lateral geniculate nucleus and striate cortex.
Brain Res
 
349
:
53
–62.
Grossberg S, Seitz A (
2003
) Laminar development of receptive fields, maps and columns in visual cortex: the coordinating role of the subplate.
Cereb Cortex
 
13
:
852
–863.
Hawken MJ, Blakemore C, Morley JW (
1997
) development of contrast sensitivity and temporal-frequency selectivity in primate lateral geniculate nucleus.
Brain Res
 
14
:
86
–98.
Hearn JP (
1983
) The common marmoset (Callithrix jacchus). In: Reproduction in the New World primates: new models in medical science (Hearn JP, ed.), pp. 181–215. London: Kluwer Academic Publishers.
Hendrickson AE (
1996
) Synaptic development in macaque monkey retina and its implications for other developmental sequences.
Perspect Dev Neurobiol
 
3
:
195
–201.
Hof PR, Morrison JH (
1995
) Neurofilament protein defines regional patterns of cortical organization in the macaque monkey visual system: a quantitative immunohistochemical analysis.
J Comp Neurol
 
352
:
161
–186.
Hof PR, Ungerleider LG, Webster MJ, Gattass R, Adams MM, Sailstad CA, Morrison JH (
1996
) Neurofilament protein is differentially distributed in subpopulations of corticocortical projection neurons in the macaque monkey visual pathways.
J Comp Neurol
 
376
:
112
–127.
Kennedy H, Bullier J, Dehay C (
1985
) Cytochrome oxidase activity in the striate cortex and lateral geniculate nucleus of the newborn and adult macaque monkey.
Exp Brain Res
 
61
:
204
–209.
Kiorpes L, Kiper DC (
1996
) Development of contrast sensitivity across the visual field in macaque monkeys (Macaca nemestrina).
Vision Res
 
36
:
239
–247.
Kirby MA, Steineke TC (
1991
) Early dendritic outgrowth of primate retinal ganglion cells.
Vis Neurosci
 
7
:
513
–530.
Kogan CS, Zangenehpour S, Chaudhuri A (
2000
) Developmental profiles of SMI-32 immunoreactivity in monkey striate cortex.
Brain Res Dev Brain Res
 
119
:
85
–95.
Kovacs I, Kozma P, Feher A, Benedek G (
1999
) Late maturation of visual spatial integration in humans.
Proc Natl Acad Sci USA
 
96
:
12204
–12209.
Lachica EA, Casagrande VA (
1988
) Development of the primate retinogeniculate axon arbors.
Vis Neurosci
 
1
:
103
–123.
La Vail MM, Rapaport DH, Rakic P (
1991
) Cytogenesis in the monkey retina.
J Comp Neurol
 
309
:
86
–114.
Lasek RJ (
1981
) The dynamic ordering of neuronal cytoskeletons.
Neurosci Res Program Bull
 
19
:
7
–32.
Liu Y, Dyck R, Cynader M (
1994
) The correlation between cortical neuron maturation and neurofilament phosphorylation: a developmental study of phosphorylated 200 kDa neurofilament protein in cat visual cortex.
Brain Res Dev Brain Res
 
81
:
151
–161.
Lund JS, Lund RD, Hendrickson AE, Bunt AH, Fuchs AF (
1975
) The origin of efferent pathways from the primary visual cortex, area 17, of the macaque monkey as shown by retrograde transport of horseradish peroxidase.
J Comp Neurol
 
164
:
287
–303.
Lund JS, Boothe RG, Lund RD (
1977
) Development of neurons in the visual cortex (area 17) of the monkey (Macaca nemestrina): a Golgi study from fetal day 127 to postnatal maturity.
J Comp Neurol
 
176
:
149
–188.
Matus A (
1988
) Neurofilament protein phosphorylation — where, when and why.
Trends Neurosci
 
11
:
291
–292.
Mikami A, Fujita K (
1992
) Development of the ability to detect visual motion in infant macaque monkeys.
Dev Psychobiol
 
25
:
345
–354.
Meissirel C, Wikler KC, Chalupa LM, Rakic P (
1997
) Early evidence of magnocellular and parvocellular functional subsystems in the embryonic primate visual system.
Proc Natl Acad Sci USA
 
94
:
5900
–5905.
Missler M, Eins S, Merker HJ, Rothe H, Wolff JR (
1993
a) Pre- and postnatal development of the primary visual cortex of the common marmoset. I. A changing space for synaptogenesis.
J Comp Neurol
 
333
:
41
–52.
Missler M, Wolff A, Merker HJ, Wolff JR (
1993
b) Pre- and postnatal development of the primary visual cortex of the common marmoset. II. Formation, remodelling, and elimination of synapses as overlapping processes.
J Comp Neurol
 
333
:
53
–67.
Nag TC, Wadhwa S (
1996
) Calbindin and parvalbumin immunoreactivity in the developing and adult human retina.
Brain Res Dev Brain Res
 
93
:
23
–32.
Rakic P (
1974
) Neurons in rhesus monkey visual cortex: systematic relation between time of origin and eventual disposition.
Science
 
183
:
425
–427.
Rakic P (
1982
) Early developmental events: cell lineages, acquisition of neuronal positions, and areal and laminar development.
Neurosci Res Program Bull
 
20
:
439
–451.
Rakic P (
1985
) DNA synthesis and cell division in the adult primate brain.
Ann N Y Acad Sci
 
457
:
193
–211.
Rakic P (
2002
) Evolving concepts of cortical radial and areal specification.
Prog Brain Res
 
136
:
265
–280.
Rapaport DH, Stone J (
1984
) The area centralis of the retina in the cat and other mammals: focal point for function and development of the visual system.
Neuroscience
 
11
:
289
–301.
Reese BE, Cowey A (
1990
) Fibre organization of the monkey's optic tract. I. Segregation of functionally distinct optic axons.
J Comp Neurol
 
295
:
385
–400.
Rumpel S, Kattenstroth G, Gottmann K (
2004
) Silent synapses in the immature visual cortex: layer-specific developmental regulation.
J Neurophysiol
 
91
:
1097
–1101.
Sawant LA, Hasgekar NN, Vyasarayani LS (
1994
) Developmental expression of neurofilament and glial filament proteins in rat cerebellum.
Int J Dev Biol
 
38
:
429
–437.
Schein SJ, de Monasterio FM (
1987
) Mapping of retinal and geniculate neurons onto striate cortex of macaque.
J Neurosci
 
7
:
996
–1009.
Spatz WB (
1979
) The retino-geniculo-cortical pathway in Callithrix. II. The geniculo-cortical projection.
Exp Brain Res
 
36
:
401
–410.
Spatz WB (
1989
) Loss of ocular dominance columns with maturity in the monkey, Callithrix jacchus.
Brain Res
 
488
:
376
–380.
Spatz WB, Tigges J, Tigges M (
1970
) Subcortical projections, cortical associations, and some intrinsic interlaminar connections of the striate cortex in the squirrel monkey (Saimiri).
J Comp Neurol
 
140
:
155
–174.
Spatz WB, Vogt DM, Illing RB (
1993
) Distribution of cytochrome oxidase in layers IV and V of the striate cortex in neonate monkeys.
Exp Brain Res
 
95
:
183
–186.
Sternberger LA, Sternberger NH (
1983
) Monoclonal antibodies distinguish phosphorylated and nonphosphorylated forms of neurofilaments in situ.
Proc Natl Acad Sci USA
 
80
:
6126
–6130.
Takahashi T, Goto T, Miyama S, Nowakowski RS, Caviness VS Jr (
1999
) Sequence of neuron origin and neocortical laminar fate: relation to cell cycle of origin in the developing murine cerebral wall.
J Neurosci
 
19
:
10357
–10371.
van der Gucht E, Vandesande F, Arckens L (
2001
) Neurofilament protein: a selective marker for the architectonic parcellation of the visual cortex in adult cat brain.
J Comp Neurol
 
441
:
345
–368.
vogt Weisenhorn DM, Illing RB, Spatz WB (
1995
) Morphology and connections of neurons in area 17 projecting to the extrastriate areas MT and 19DM and to the superior colliculus in the monkey Callithrix jacchus.
J Comp Neurol
 
362
:
233
–255.
Waleszczyk WJ, Wang C, Young JM, Burke W, Calford MB, Dreher B (
2003
) Laminar differences in plasticity in area 17 following retinal lesions in kittens or adult cats.
Eur J Neurosci
 
17
:
2351
–2368.
Wong-Riley M (
1979
) Changes in the visual system of monocularly sutured or enucleated cats demonstrable with cytochrome oxidase histochemistry.
Brain Res
 
171
:
11
–28.
Xu L, Tanigawa H, Fujita I (
2003
) Distribution of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate-type glutamate receptor subunits (GluR2/3) along the ventral visual pathway in the monkey.
J Comp Neurol
 
456
:
396
–407.
Zielinski BS, Hendrickson AE (
1992
) Development of synapses in macaque monkey striate cortex.
Vis Neurosci
 
8
:
491
–504.