Introduction

Studies on the maturation of the human neocortex represent an important area of neurobiological research. The gray matter is crucial for integrating sensory information, memory formation, and higher-order thinking skills. Understanding the mechanisms of cerebral cortex maturation may help identify the causes of neurological and psychiatric disorders such as autism spectrum disorder and schizophrenia. Investigations into age-related and individual differences in cortical maturation could contribute to developing effective diagnostic and intervention strategies for cognitive and emotional development.

A key characteristic of human neocortex development is its pronounced gyrification. Broader cortical maturation and the processes underlying sulcus and gyrus formation remain poorly understood in neuroscience.

Our previous findings indicate that cortical plate formation progresses more slowly within sulci than within gyri of the same cytoarchitectonic field during early and middle fetal periods [1]. The study focused on the visual cortex, particularly the calcarine sulcus (primary visual field 17) and parieto-occipital sulcus (secondary visual field 18). This region was selected due to its accelerated synaptogenesis observed within the visual cortex [2, 3] and stable association between these sulci and their respective cytoarchitectonic fields [4].

Neuronal maturity markers include synaptic activity markers, NeuN, GAD65/67, and GAT-1.

Glutamate decarboxylase (GAD), an enzyme responsible for synthesizing γ-aminobutyric acid (GABA), is the most prevalent inhibitory neurotransmitter in the central nervous system (CNS) of higher vertebrates [5]. GAD exists in two isoforms – GAD65 and GAD67, named for their molecular weights, which function synchronously to produce and regulate physiological GABA levels [5].

The GABA transporter 1 (GAT-1) is the primary GABA transporter in the brain that plays a crucial role in modulating GABAergic transmission. It is predominantly found in axon terminals forming symmetrical synapses and astrocytic processes [6].

NeuN is a marker for most postmitotic neuroblasts and mature neurons in vertebrate animals [7].

To assess cortical maturity in the sulcus, gyrus, and sulcal banks of the visual cortex, we analyzed the distribution of GAD65/67, NeuN, and GAT-1 using immunohistochemistry.

Materials and methods

This study used autopsy specimens from the archival collection of the Laboratory of Nervous System Development of Avtsyn Research Institute of Human Morphology of FSBSI “Petrovsky National Research Center of Surgery.” The study was approved by the local Ethics Committee (No. 33(9), February 7, 2022). Fourteen post-mortem human brain samples were analyzed: (1) six late-fetal period specimens (29–40 gestational weeks); (2) six pediatric samples (1 day–4 years); and (3) two adult samples (48 and 70 years). Gestational age was calculated from the last menstrual period, verified by ultrasound examination, and supplemented with anthropometric parameters (weight/height and crown-rump length) [8]. Our analysis included samples without clinical records indicating chromosomal anomalies or CNS pathology. Samples for immunohistochemical analysis were selected after histological assessment of material preservation. We excluded samples with hemorrhages, necrosis, or signs of post-mortem autolysis, as well as samples with prolonged storage in formalin due to the potential impact on immunoreactivity distribution. The data on human occipital lobe formation at earlier developmental stages are available on the Human Brain Development Atlas project website at https://brainmorphology.science [9, 10].

Histological preparation and immunohistochemistry

All brain samples underwent histological processing. Depending on the fetal age, we examined either the whole brain or a region of the medial wall of the hemisphere encompassing the calcarine sulcus (sulcus calcarinus, Cas) and parieto-occipital sulcus (sulcus parietooccipitalis, Pos). Samples were immersion-fixed in either 10% neutral buffered formalin (pH 7.0–7.4) or Bouin’s solution, followed by dehydration in graded ethanol solutions and dioxane. Tissues were then paraffin-embedded and sectioned at 10-μm thickness. Sections were stained with classic histological cresyl violet (Nissl stain) for cytoarchitectural analysis and mounted on Superfrost Plus slides (Thermo Fisher Scientific, USA) and stored at 4°C. Immunohistochemical analysis was performed using primary antibodies targeting nervous system antigens: NeuN (Millipore, RRID: AB_2298772, Germany) at a dilution of 1:100, GAT-1 (Millipore, RRID: AB_90791, Germany) at a dilution of 1:100, and GAD65/67 (Thermo Fisher Scientific, RRID: AB_930942, USA) at a dilution of 1:50. For immunohistochemical labeling, sections were deparaffinized, rehydrated, boiled in citrate buffer (pH 6.0; DiaGene, Russia) for 10 min, and incubated with primary antibodies diluted in 0.01 M phosphate-buffered saline (PBS, pH 7.3–7.5; Biolot, Saint Petersburg, Russia) for 1 h at room temperature for NeuN and GAT-1 and for 18–20 h at 8°C for GAD65/67. Visualization was achieved with The Ultra Vision LP Detection System (Thermo Fisher Scientific, USA). Negative control sections were obtained by replacing the primary antibodies with PBS (0.01 M). Non-specific staining was absent in all control sections.

Statistical analysis

We analyzed all specimens under a light microscope (DM 2500; Leica Microsystems, Germany) connected to a digital camera (Lomo, Russia). Images were acquired and processed using McrA-View 7.1.1.2 software (Lomo, Russia). Measurements were performed on photomicrographs of the specimens. We scanned some specimens using a modified MEKOS-C2 complex (MEKOS, Russia) based on a Zeiss Axio Imager 1 microscope (Carl Zeiss, Germany) with a 20x objective. We analyzed 10 non-overlapping observation fields from 3–5 sections per region. The immunoreactivity coefficient was calculated using ImageJ software ver. 1.43 as the ratio of the sum of immunopositive (colored) pixels to the total number of pixels in the selected area.

For statistical analysis, we divided cases into three groups: late fetal (29–40 weeks), postnatal, and adult periods. The immunoreactivity coefficient values were compared in the cortical sulci, sulcal banks, and gyri. Statistical analysis was performed using Statistica 10 software (StatSoft, USA) employing the non-parametric Kruskal–Wallis ANOVA with post-hoc multiple comparisons of mean ranks.

Results

Distribution of NeuN immunoreactivity

In the late fetal period, NeuN-positive cells were in all layers of the cortical plate on the brain surface, both in gyri and sulcal banks (Fig. 1 A). The NeuN immunoreactivity coefficient was significantly higher in sulcal banks than in the sulci [p=0.000000] (Fig. 1 B). Within the sulci, the frequency of NeuN-positive cells gradually decreased from layer II to layer VI. Notably, NeuN-positive cells were predominantly localized in layer VI of the calcarine sulcus. The cortex within the parieto-occipital sulcus showed no specific layer in NeuN immunoreactivity.

In the postnatal period, the above-described trends in marker distribution were detected, which persisted from birth until 4 years postnatally. More intense staining was observed in the gyri cortex, where labeled neurons were found in all cortical layers. In the depth of the sulci, NeuN immunoreactivity was low in both areas 17 and 18.

A distinct pattern of NeuN-immunopositive cell distribution was observed at 10 months postnatally (Fig. 1 C, D). Within the sulcus, NeuN-immunopositive cells formed morphological and functional neural assemblies, predominantly in layer II with partial extension to layer III of the neocortex, and additional clusters in layers V and VI. These assemblies measured 250 μm in width in layer II and 210 μm in layer VI.

In adults, NeuN-positive neurons were found throughout the primary and secondary visual fields, with no difference in distribution in the gyri, sulcal banks, and sulcal depths.

Thus, we detected asynchronous maturation of NeuN immunoreactivity in the sulci and gyri, persisting until 4 years postnatally (Fig. 1 B, E, F).

Distribution of GAD65/67-immunoreactivity

In the late fetal period, a small number of GAD65/67-immunoreactive cells were observed in the cortex. Approaching birth, GAD65/67-immunoreactivity intensified, with a significantly higher coefficient being in the sulcal bank cortex compared to that in the gyri cortex and the cortex within the sulci within the same cytoarchitectonic fields (Fig. 2 A, B, C). In area 17, immunoreactivity was mainly localized in layers IVC and V–VI. Layers IVB and IVA had dense groups of neurons and revealed the structure of functional neuronal assemblies. Areas 18 and 19 showed strong immunoreactivity in layers II and upper layer III (Fig. 2 A, B, C), with occasional morphological and functional assemblies found in the lower layer III and layer IV. The width of the assemblies at layer IV averaged 190 μm.

Postnatally, the GAD65/67-immunoreactivity coefficient increased, particularly in gyri and sulcal banks (Fig. 2 D, E), with its statistically significant differences observed in the cortex of sulcal banks, gyri, and sulci (Fig. 2 F).

These findings demonstrate that GAD65/67 successfully captured transient cortical assemblies in newborns and revealed heterogeneity of cortical maturation in sulci, gyri, and sulcal banks.

Distribution of GAT-1 immunoreactivity

In the late fetal period, GAT-1-immunopositive fibers were distributed across the entire hemisphere wall, mainly localizing in the marginal zone. A cytoarchitectonically specific pattern of immunopositive fiber arrangement in the marginal zone was noticeable. GAT-1 immunoreactivity of the cortex was the highest in the sulcal banks and gyri, with GAT-1 immunoreactivity decreasing when moving deeper into the sulci (Fig. 3 A, D). In area 17, the lower cortical layers and marginal zone showed prolonged immunohistochemical labeling, whereas in areas 18–19, the upper layers and marginal zone maintained stronger immunoreactivity over time. Postnatally, GAT-1 immunoreactivity in the sulcal bank cortex and the gyri cortex was comparable (Fig. 3 B, E). GAT-1 immunoreactivity in the sulcal cortex was significantly lower until 4 years postnatally (Fig. 4 A, B, C).

In adults, GAT-1-immunopositive fibers were primarily found in the cortex, with predominant labeling being in layer I. In adults, we observed no reduction in labeling within the sulci (Fig. 3 C, F).

GAT-1 immunoreactivity demonstrates that the gyri cortex and the sulcal bank cortex mature earlier than the cortex within the sulci. The delayed maturation of the sulci cortex persists until at least 4 years of age.

Discussion

Distribution of GAT-1, GAD65/67, and NeuN markers

Several studies have identified age-related changes in the levels of the synaptic activity markers GAT-1 [11] and GAD [12–14].

Our findings demonstrate predominant localization of the GAT-1 transporter to the marginal zone, consistent with earlier studies of the human neocortex [11]. However, due to the limited sample size, age-related changes could not be confirmed.

Literature reports significant changes in the GAD65 isoform expression but not in the GAD67 one [12]. Our study demonstrates the predominance of neuron body labeling, i.e., the GAD67 isoform labeling, in the late fetal period and newborns. According to the literature, this isoform appears earlier than others during human and animal ontogenesis [13, 14]. In the adult human cortex, immunolabeling revealed a predominance of fine, small fibers, i.e., GAD65 isoform, which is consistent with published data [13]. The quantity of the GAD67 isoform is likely to depend on the functional load of certain neurons. A similar dependence was confirmed for the GAD65 isoform [15].

This paper reveals differing immunoreactivity to GAT-1, GAD65/67, and NeuN in sulci, sulcal banks, and gyri. We showed higher immunoreactivity coefficient values in sulcal banks and gyri in the late fetal period, which continued in the postnatal period until 4 years postnatally. In the adult samples, we observed no heterogeneity in the distribution of the markers in the cortex of sulci and gyri.

Neocortical neurons from vertical columnar assemblies span all cortical layers. Neurons within a column process signals of a single modality [16], a functional architecture established in early cortical development. Each column contains cells derived from multiple clones of a common origin [17]. In the adult brain, these columns group into larger second-order assemblies—macrocolumns—which typically consist of 3–12 columns [18].

Previous studies visualized macrocolumns with GAD67 immunohistochemistry in both the developing and mature rat brain [19]. In cats, transient visualization of second-order morphological and functional structures was observed with the labeling of excitatory receptors NMDAR1 during normal brain development [20].

Our findings reveal second-order morphological and functional assemblies in the human neonatal cortex using GAD labeling. In newborns, GAD-immunopositive neuron bodies were found in area 17.

Additionally, NeuN labeling has temporarily revealed morphological and functional assemblies in a 10-month-old infant, localized within the sulci of the neocortex. In vivo imaging in mice demonstrated that prospective interneurons form patches of correlated activity that merge during postnatal development. This functional organization changes rapidly in two steps: by the end of the postnatal week, GABA assemblies form; two days later, these assemblies merge into a fully connected functional network [21]. As a result, the clustered functional organization of the cortex is a transient developmental stage in mice. The persistence of morphological features of the clustered organization within the sulci in the human neocortex at 10 months may indirectly indicate a slower maturation rate in these regions.

Thus, immunohistochemical labeling of GAD65/67 and NeuN shows that cortical maturation progresses in clusters, representing second-order functional neural assemblies. Comparative analysis using GAT-1, GAD65/67, and NeuN markers indicates a more accelerated cortical maturation in the sulcal banks and gyri than in the cortex within the sulci.

Conclusion

This research studied the complex maturation of the human cerebral cortex within sulci and gyri during late fetal and postnatal development, thus addressing a significant gap in our understanding of neocortical formation. Cortical maturation progresses heterogeneously and is characterized by the formation of second-order functional neural assemblies, as visualized using GAD65/67 and NeuN markers. Specifically, cortical regions of sulcal banks and gyri mature earlier than the sulci, as confirmed by the distribution patterns of GAT-1, GAD65/67, and NeuN. These data indicate a differentiated maturation pattern in the gyrified cortex, where regions at the boundaries of sulci and gyri were shown to mature earlier than sulcal regions.

The paper highlights the complex interplay between gyrification and cortical maturation, suggesting a strong link between the cortical structural organization and its functional development. Future research should elucidate the mechanisms underlying heterogeneity in cortical maturation and investigate their functional consequences. These studies could provide a more comprehensive understanding of normal brain development and help uncover diagnostic and therapeutic targets for neurodevelopmental disorders.