Radial glia, the neuronal stem cells of the embryonic cerebral cortex, reside deep within the developing brain and extend radial fibres to the pial surface, along which embryonic neurons migrate to reach the cortical plate. Here we show that the gap junction subunits connexin 26 (Cx26) and connexin 43 (Cx43) are expressed at the contact points between radial fibres and migrating neurons, and acute downregulation of Cx26 or Cx43 impairs the migration of neurons to the cortical plate. Unexpectedly, gap junctions do not mediate neuronal migration by acting in the classical manner to provide an aqueous channel for cell-cell communication. Instead, gap junctions provide dynamic adhesive contacts that interact with the internal cytoskeleton to enable leading process stabilization along radial fibres as well as the subsequent translocation of the nucleus. These results indicate that gap junction adhesions are necessary for glial-guided neuronal migration, raising the possibility that the adhesive properties of gap junctions may have an important role in other physiological processes and diseases associated with gap junction function.

Significance statement

One of the most remarkable features of human brain evolution is the enormous increase in neuron number and the transformation of the lissencephalic cortex into a highly folded gyrencephalic cortex. This expansion is established primarily by large numbers of a specific type of neural stem cell, the outer radial glial (oRG) cell, present in the developing human brain. The developmental origins of oRG cells, and the mechanisms regulating their proliferation, are unknown. Using time-lapse imaging of human fetal cortical slices, we find that oRG cells are progeny of ventricular radial glial (vRG) cells, the primary neural stem cell present in all mammals. Horizontal vRG divisions, which are far more common in human than in mouse, lead to oRG cell generation, while vertical divisions do not. These results suggest that an increase in horizontal vRG divisions might have expanded the oRG cell population during primate brain evolution. We next use dissociated human fetal cortical cultures to interrogate the molecular pathway controlling a unique oRG cell behavior called called mitotic somal translocation (MST). During MST, the oRG cell body rapidly translocates a distance of several cell diameters towards the cortical plate immediately prior to cytokinesis. We find that oRG cell MST and cleavage angle regulation are cell intrinsic, and that MST is independent of mitosis. Activation of the Rho effector ROCK and of non-muscle myosin II, but not microtubule polymerization or centrosomal guidance, is required for MST. Furthermore, oRG cell MST contributes to radial oSVZ expansion. Finally, we describe several genetic mutations that target the RhoA-ROCK-myosin II pathway and cause cortical malformations in humans, including microcephaly, periventricular heterotopia, and lissencephaly. Interestingly, many of these mutations produce very minimal phenotypes in mice. We propose that MST may be the primary target of these diseases, as oRG cells compose a very small proportion of neural progenitor cells in mice. Given our discovery of the origin of oRG cells and the molecular motor driving MST, the field will now be in a better position to understand the function of oRG cells and oRG-specific behaviors such as MST in human brain development and disease.

Many of the extraordinary intellectual capabilities of humans are attributed to the proper development and function of the neocortex. This region of the brain is responsible for higher cognitive functions such as sensory perception, spatial reasoning, and conscious thought, and is distinguished in humans by its large size and extensively folded pattern. However, the developmental events that underlie this structure have mostly been inferred from characterizations of the mouse or rat, whose neocortex is small and not folded.

Here, we directly characterize the progenitor cells present in the developing human neocortex. We find that the developing human neocortex utilizes a proliferative cell population that forms the outer subventricular zone (OSVZ), a region removed from the cerebral ventricles that is not present in the mouse. We further identify an OSVZ progenitor cell type known as an outer radial glial (oRG) cell and demonstrate using real-time imaging that they undergo self-renewing asymmetric divisions to generate neuronal progenitor cells. oRG cells function as neural stem cells and underlie an OSVZ lineage of progenitor cells, which over the course of many cell cycles amplifies neuron number and likely modifies the trajectory of migrating neurons.

Because the establishment of non-ventricular radial glia may have been a critical evolutionary advance underlying increased cortical size and complexity in the human brain, we further discuss the prevalence of oRG cells and OSVZ proliferation in different species throughout evolution. We learn that OSVZ proliferation is present in all superorders of placental mammals, but only in those species with folded brains are these mechanisms appreciably utilized. To gain insight into the molecular and evolutionary bases for these phenotypic differences, we performed serial microrarray analysis to uncover gene co-expression patterns present in the developing human neocortex. From these data, we identified co-expression modules that correspond to human radial glia and further determined genes within these modules that have undergone human-specific evolutionary changes. This data serves as a molecular substrate for comparisons of radial glia between species and will help elucidate the molecular mechanisms that drive OSVZ proliferation.

The evolutionary expansion of the mammalian cerebral cortex is responsible for our unique higher cognitive abilities such as language and conscious thought. Developmental changes in neural progenitor composition likely contributed to the increased size and number of neurons in the primate cortex. For instance, a germinal region, referred to as the subventricular zone (SVZ), is greatly expanded in primates and contains a heterogeneous population of neural progenitors, including outer radial glia (oRGs). Since oRGs are more abundant in the primate than the rodent and contribute to neurogenesis, an increase in this cell population is thought to underlie cortical expansion and folding. However, the specific contribution of oRGs and other cell populations, such as neurons, to the size and shape of the adult mammalian cortex remains unknown. Since the ferret is gyrencephalic, possesses an enlarged SVZ with an abundance of oRGs, and is experimentally tractable, it provides an attractive model for studying cortical expansion and folding. Here, I provide a detailed characterization of progenitor and migrating neuron behaviors in the developing ferret cortex. I present data demonstrating the mitotic behaviors, proliferative capacity, and daughter cell fates of ferret oRGs and highlight morphological diversity within this progenitor population. I go on to present a technique for visualizing neurogenic divisions and neuronal migration in the ferret cortex. I describe neuronal migration properties that appear to differ from the rodent, such as greater variation in the direction of migration and increased meandering, specifically at later ages when the ferret cortex has begun to fold. These data provide a framework for future studies using the ferret to elucidate the cellular and molecular mechanisms of cortical expansion and folding, as well as neurodevelopmental diseases.