This series of reviews on cortical development is based on collaborative work carried out in several European laboratories under the EU 5th Framework Programme research network ‘CONCORDE’ (for ‘CONsortium on CORtical DEvelopment’). Our collaboration arose from the natural need to address fundamental questions related to cerebral cortical development through international partnerships, cooperation and peer review. A European consortium of ten groups with mutual interests was formed after an informal and memorable gathering in Berlin at the FENS meeting in 1998.
The research carried out by this consortium addresses diverse issues which, in most cases, would be beyond the purview of a large laboratory. The range of research topics includes cortical neuronal proliferation, radial and tangential migration of neurons, layer formation and development and plasticity of connectivity. Cortical development is studied in rodents but also in primates including humans, with special attention to human developmental disorders. Other related areas such as thalamus and hippocampus are also examined. Because of the scale of the consortium and the multidisciplinary nature of the participants, the comparative and evolutionary aspects of cortical development also fit nicely into the discussion.
The most complex region of the human brain is the cerebral cortex. This structure is composed of the neocortex and limbic cortex. Neocortex plays a key role in cognitive and intellectual processes and other aspects of our personality, while the limbic cortex and hippocampus are involved in emotional control, learning and memory. Millions of neurons in our brain develop billions of connections with trillions of possible specificities. The complexity of the cortex results from carefully orchestrated developmental interactions, most of which are unknown. Our twenty- to thirty-thousand genes are expressed in different temporal sequences and combinations, partially as a result of environmental signals. It is a marvel how the unfolding genetic program will produce a functional brain. A number of neurological or psychiatric diseases result from disrupted cortical development. Disruptions can manifest a range of symptoms such as gross cortical malformations with profound mental retardation and refractory epilepsy, such as lissencephaly, less severe alterations that result in epilepsy, learning and memory disorders and/or cognitive impairment, or behavioural disorders. Even some psychoses are thought to be partially of developmental origin. While diseases resulting from destructive lesions (e.g. ischemia, infections and metabolic deficiency) are relatively well understood and can sometimes be prevented, the mechanisms and pathophysiology of a wide variety of cortical developmental disorders remain largely unknown and without effective treatment.
Cortical neurons are generated at defined locations and time points in the ventricular (VZ) and subventricular (SVZ) zones which line the telencephalic ventricles. Neurons generated in the cortical VZ migrate radially and nearly all become excitatory glutamatergic neurons. In addition, many inhibitory GABAergic interneurons are generated in the VZ of the ganglionic eminence (GE), which also forms the striatum, and reach the cortex by tangential routes. The VZ and SVZ also produce various types of macroglial cells. Neuronal determination proceeds spatially, neuronal classes being generated with respect to VZ location, and temporally, a given point of the VZ generating different cell types over time. Inducers (such as molecules of the Fgf, Bmp, Shh and Wnt families) are secreted by organizing centres (such as the anterior neural ridge, the choroid–cortical junction or cortical hem and the zona limitans in the thalamus) and modulate stem cell expression of genes that regulate proliferation and determination of neuronal or glial cell fate. Cortical arealization results from a delicate interplay between signalling by extracellular molecules and key transcription factors such as Emx1 and -2, Pax6 and others. Amazing progress has been made in this field during the last 5 years due to increasingly sophisticated murine models, and is reviewed in the first contribution by Mallamaci & Stoykova (2006). A closely related question concerns the differentiation of various cortical cell types and of the acquisition of layer-specific determinants that give rise to the laminated cortex. Multipotent stem cells do not directly form neurons and glial cells, but instead generate intermediate precursors that are then restricted towards the generation of a single cell type. These questions are reviewed in the second paper by Guillemot et al. (2006), focusing particularly on bHLH transcription factors, Pax6 and a few novel genes.
Human brain malformation disorders, although uncommon, are a major medical concern. Studies of rare genetic disorders provide fertile ground for furthering our understanding of basic developmental mechanisms. This approach has been, and continues to be, extremely powerful, as is clearly outlined in the review by Francis et al. (2006), where this approach is framed in a useful historical perspective. Studies of human brain malformation touch on all aspects of neural development and have been particularly useful for investigating neural generation and migration. Within the last 5 years, mutations in the human X-linked Arx gene have been associated with a wide spectrum of brain disorders ranging from nonsyndromic mental retardation to severe malformations. This led to a number of studies of Arx in mice and other organisms; these studies clearly implicate Arx orthologs in many aspects of brain development, principally in specification and migration of GABAergic cortical interneurons. The present status of research on this important gene is reviewed by Friocourt et al. (2006).
In contrast to other tissues, neurons migrate within the neuroepithelium over variable distances along well defined pathways. Migration begins with extension of a leading process (a step controlled mainly by the ‘actin treadmill’) and a vital question concerns differences between the leading processes of radially and tangentially migrating cells. This is followed by nucleokinesis (movement of the cell soma and nucleus), a step primarily dependent on microtubules. Most cortical GABAergic interneurons arise in the medial GE and migrate tangentially to the cortex. Tangentially migrating neurons undergo spectacular morphological modifications and their nucleokinesis is tightly regulated. Mechanisms of tangential interneuron migration have been intensely investigated over the last 10 years and are reviewed by Métin et al. (2006). When neurons reach their destination, they are incorporated into the local architecture. In rodents, the preplate is a loose horizontal network populated with pioneer neurons. These preplate neurons establish the blueprint for layer formation and probably for areal and hodological development. Preplate formation is followed by the migration and condensation of the cortical plate (CP), which splits the preplate into marginal zone (MZ) and subplate. Following migration, neural incorporation into the laminar structures of cortical plate and hippocampus is controlled by the Reelin pathway. Reelin secreted by Cajal–Retzius cells in the marginal zone binds to lipoprotein receptors (VLDLR and ApoER2) on cortical and hippocampal cells. This phosphorylates Dab1 and triggers a cascade that results in the normal architectonic disposition of the adult mammalian brain. Recent data on Reelin signalling and on the cell biological action of Reelin in the dentate gyrus, where it is particularly well studied, are reviewed by Förster et al. (2006).
Functional specialization of cortical areas is closely related to cytoarchitecture and hodological organization. Two models are suggested by this relationship. In the first, areal differences may develop independently of afferents in response to local differences in neuronal determination in the VZ, which forms a ‘protomap’ of the cortical map; this model links together studies of cell determination and arealization. The second model considers the CP a ‘tabula rasa’ on which areal differences are imprinted by afferent fibres, particularly from the thalamus. A host of experimental data show that normal areal and functional development requires both intrinsic area-specific markers and the integrity of thalamocortical (as well as intra- and interhemispheric) connections, suggesting that the two models are complementary. The mechanisms responsible for wiring the cortex are complex and act sequentially at different organizational levels starting with the directional elongation of axons. Although basic mechanisms are probably conserved among mammals, there are significant differences between rodents and primates. This complex field is reviewed by Price et al. (2006).
‘Nothing in Biology makes sense except in light of evolution’ (Dobzhansky), and we thought it fitting to conclude these reviews with comparative aspects of cortical development. A lot of data are available on the comparative anatomical organization of the brain, and the forebrain in particular. Fewer studies have addressed comparative development and it was not until recently that modern techniques such as in situ hybridization could be applied to comparative analysis. Comparison between widely different amniotes shed some light on the evolutionary process. Although much work remains to be done, some central questions are at last beginning to yield to investigation. Some of the recent progress is reviewed in the last paper by Molnár et al. (2006). The mouse cortex as a model has obvious limitations and therefore it is necessary to study cortical development in the human and monkey embryo to define primate-specific features. In man, Cajal–Retzius cells are a highly complex, heterogeneous collection of p73- and Reelin-expressing neurons which are transiently present in the MZ. The cortical hem is the source of a substantial proportion of Cajal–Retzius cells that migrate from medial to lateral before contributing to the MZ in neocortex and hippocampus. This novel migration, more evident in man than mice, complements the tangential migration of interneurons from the GE. Other studies using human fetal cortex have focused on doublecortin and LIS1. Doublecortin is not expressed by human radially migrating neurons that express LIS1. In addition to doublecortin-positive cells derived from the GE, some are prominent in the neocortical subventricular zone, suggesting that, in humans as in rodents, a proportion of cortical interneurons found in the subventricular zone switch from an initial tangential migration mode to a late radial mode. Exploring these differences will be very useful for understanding the limitations of our current models and further comprehending the general biological mechanisms of cortical development and their possible evolutionary origin.
We are most grateful to the EJN for giving us this opportunity to publish our views, and sincerely hope that the general readership of the journal will share some of our excitement and enthusiasm. Understanding the cortex ranks among the quintessential intellectual challenges of this millennium, and we cannot think of a more fascinating subject to attract the best young minds.