Transitional neuronal layers are a hallmark of the prenatal and neonatal brain yet their contribution to the development of higher functions is not clear. Evidence accumulated over the last 3 decades shows that early connectivity and functional activity start in a transitional layer called the subplate zone (SPZ). The SPZ is host to a heterogenous population of neurons and its evolutionary complexity peaked in the human brain. In this issue of Cerebral Cortex, three reports (Hoerder-Suabedissen et al., 2008; McKellar and Shatz, 2008; Moore et al., 2008) present new data and evidence in three species (mouse, rat, human) as to the function of the SPZ, to the heterogenity of its cellular composition, and to the genetic basis of its development.
The subplate zone (SPZ) is one of the few new structures in the central nervous system defined in the last 35 years. The SPZ was overlooked by the generation of great neuroanatomists at the onset of the 20th century, perhaps because of the common belief that transitional neuronal layers coalesce with neighboring layers or altogether disappear (His 1874; Campbell 1905; Brodmann 1909; Economo and Koskinas 1925; Cajal 1952). It was first described as a separate region between the cortical plate and the intermediate zone in the human fetal cortex (Kostovic and Molliver 1974) and then in the fetal macaque (Rakic 1977) and carnivores (Luskin and Shatz 1985). It is much smaller but nevertheless clearly defined in rodents (Molnar 2000). In spite of its biomedical significance and enormous evolutionary expansion that culminates in human (Kostovic and Rakic 1990), the SPZ has been the least investigated among transient embryonic cellular compartments. This is apparently changing now. In this issue of Cerebral Cortex, 3 papers combine neuroanatomical and physiological methodology as well as gene expression profiling to advance our understanding of the molecular and genetic control of SPZ development in 3 different species, human, rat, and mouse (Hoerder-Suabedissen et al. 2008; McKellar and Shatz 2008; Moore et al. 2008).
In the last 3 decades, published evidence has established an important role for the pioneering cells and the SPZ they occupy for the proper development of corticocortical and subcortical connections (Kostovic and Rakic 1980; Shatz and Luskin 1986; McConnell et al. 1994). Subplate neurons are among the first-born neurons in the telencephalic wall and start making the first synapses by midgestation. Studying synaptogenesis in vivo in developing neocortical neurons has been a tedious endeavor because of the technical challenges and advanced skills required. In vitro analysis of synaptogenesis, on the other hand, is marred by the uncertainty of its applicability in vivo. Here, McKellar and Shatz (2008) immunopurified SPZ neurons and studied gene expression involved in synaptogenesis. They provide a new, well-characterized in vitro system to explore early circuit formation under different experimental paradigms. McKellar and Shatz (2008) improved on earlier in vitro systems and provided gene expression data from different experimental conditions, thus making a powerful system to dissect synaptic events and the genetic determinants of synaptogenesis. Their experimental system triggers rapid induction of synapses, which seems to be unprecedented, and may shed light on the role of the SPZ and is of potential use in studying cortical patterning. One of the authors’ propositions is that a developmental program for SPZ cells necessitates the interactions with other cell types for their maturation. The nature of interactions, soluble factors, and the cells that SPZ neurons interact with in vivo remain to be explained. It is not clear whether the maturation of SPZ neurons in vivo is initiated through paracrine signaling before the onset of synaptogenesis. Furthermore, it is not known whether cortical plate neurons, radial glial cells, or thalamocortical axons are involved, to what extent, and what molecules drive these interactions. These and other open questions can now be dissected in vitro in the hope of determining the full repertoire of molecules facilitating communication among early-born neurons at or before the onset of synaptogenesis.
Gene expression profiling has been successfully used to determine regional gene expression in the neocortex (Kudo et al. 2007) but never at a layer-specific level. Hoerder-Suabedissen et al. (2008) microdissected and compared gene expression between the SPZ and layer 6 from the somatosensory or the visual cortices of postnatal day 8 mouse brains. This focused approach revealed diverse classes of SPZ neurons, indicating that specific subpopulations of neurons may contribute differentially to cortical development (Hoerder-Suabedissen et al. 2008; McKellar and Shatz 2008). Although SPZ neurons express some genes in common with other cortical layers, they selectively express another unique set of genes (Bayatti et al. 2008; Hoerder-Suabedissen et al. 2008; McKellar and Shatz 2008). For example, expression of CplX3, CTGF, Nurr-1/Nr 4a2, Mox D1, and F-spondin seem to be characteristic for SPZ neurons while other markers label other cortical layers in addition to the SPZ. Indeed, a study done by Osheroff and Hatten (2009) identified a separate set of genes expressed in the preplate and subplate at embryonic day (E12) in the mouse. Some of these genes seem to be expressed in a short developmental window or in a subset of SPZ cells (Osheroff and Hatten 2009). Analysis at this early embryonic stage (E12) uncovered more SPZ-specific genes (Osheroff and Hatten 2009) than reported by Hoerder-Suabedissen et al. (2008) from early postnatal embryos. Age-related differences apart, confirmation of gene expression reported by Osheroff and Hatten, by real-time quantitative polymerase chain reaction and in situ hybridization, is necessary after Genechip hybridization. At first inspection, the difference between the SPZ and layer 6 neurons seems miniscule, considering the difference in their function and morphology (Hoerder-Suabedissen et al. 2008). Sorting out spatiotemporal gene expression is necessary to affirm specificity and contribution of identified genes to neuronal subtype specificiation, axonal guidance, dendritogenesis, synaptogenesis and, in a broader sense, the evolution of the human brain. Furthermore, one cannot expect a complete picture without a comparative genomic analysis to uncover mutations in coding and noncoding regulatory sequences of SPZ-specific genes responsible for the expansion and cellular elaboration of the SPZ during evolution.
Biophysical and immunohistochemical properties of embryonic neurons were studied by Moore et al. (2008) in slices of human fetal brain at different gestational ages. The advanced maturation and electrical excitability of SPZ neurons as reported by Moore et al. (2008) lend support to the proposition that the earliest functional circuit forms in the SPZ by midgestation in the human and rodent neocortex (Allendoerfer and Shatz 1994). The size and prolonged period of SPZ development in the human brain, with respect to the increased number of connections (Kostovic and Rakic 1990) and presence in specific cortical areas, require further studies from the neurological point of view. In this vein, identifying heterotopic cells, which have been observed in several neuropathological conditions, based on their expression of SPZ-specific genes (CplX3, Nurr-1/Nr 4a2, Mox D1, CTGF, and F-spondin) or other cortical markers may help resolve the issue of the source of these cells, as mislocalized cortical plate neurons or remnants of the SPZ, and the time window when the pathology ensued. Studies on the human SPZ using a variety of physiological and molecular genetic approaches (Bayatti et al. 2008; Moore et al. 2008) are essential if we are to understand the etiology of neocortical abnormalities.
The 3 reports published herein confirm the initial findings that the SPZ is composed of 2 basic classes of neuronal phenotypes, glutamatergic and γ-aminobutyric acid (GABA)ergic (Antonini and Shatz 1990; Meinecke and Rakic 1992), serving different functions due to their differential gene expression. Postmigratory GABAergic neurons may coexist with different peptides (somatostatin [Kostovic et al. 1991]; neuropeptide-y [Delalle et al. 1997]), calbindin and calretinin. Laminar shifts of different markers during development (Ina et al. 2007; Stumm et al. 2007; Bayatti et al. 2008) may pose a problem when analyzing different developmental stages in 1 species or comparing different species (Kostovic and Rakic 1990). The most prominent differences in comparative studies were observed in the size and fiber content of the SPZ between primates and rodents (Kostovic and Rakic 1990; Bystron et al. 2008). The primate has a higher activity of extracellular matrix (ECM) production because this extracellular component forms the largest portion of the SPZ (Kostovic et al. 2002). Some of the ECM molecules are putative synaptic genes (McKellar and Shatz 2008) or may represent axon guidance molecules (Kostovic et al. 2002). Functional studies involving in utero overexpression or misexpression of select genes have proved invaluable in explaining the molecular mechanisms underlying the formation of deep cortical layers (Kwan et al. 2008; Voss et al. 2008). Similarly, future molecular genetic analyses should characterize the function of SPZ-specific genes (CplX3, Nurr-1/Nr 4a2, Mox D1, CTGF, and F-spondin) as well as the diversity of cells within this layer during development.
Studies of SPZ development are necessary to assess the perinatal origin of cognitive disorders and shed light on the transient functions of preterm cortex (Kostovic and Judas 2006; Bayatti et al. 2008). Recent advances in magnetic resonance imaging technology allow imaging and assessment of the SPZ in the human fetus in utero (Kostovic et al. 2002; Staudt et al. 2006). The SPZ may be a crucial structure for understanding early functional interactions in the human preterm cerebral cortex (Kostovic and Judas 2006) and essential elements of structural plasticity that are necessary for the development of higher brain function in the human brain (Staudt et al. 2006). The 3 reports published in this issue of Cerebral Cortex and other recently published evidence (Bayatti et al. 2008; Kwan et al. 2008) provide support for earlier hypotheses, which were based primarily on anatomical data, that the SPZ is a substrate for the formation of early molecular and functional neuronal interactions, synaptogenesis, growth, and patterning of thalamocortical pathways and their termination in the cerebral cortex (Rakic 1977; Kostovic and Rakic 1980, 1990; Shatz et al. 1988). Furthermore, the identification of new genes and development of new experimental paradigms are bound to pave the way for exiting new discoveries and deeper understanding of species-specific differences between humans and rodents.
The Patterson Trust Fellowship in Brain Circuitry to AEA; grant UKF06/07 to IK.
Conflict of Interest: None declared.