The development of the nervous system is highly dynamic, and involves the migration of hundreds of millions of cells through a three-dimensional environment. After this migration, the immature neurons extend finger-like processes which then form connections (synapses) with other cells. While some neuronss form only a very small number of connections, others, as cerebellar Purkinje cells, form synapses with 10,000 or more other cells. When one considers that the human brain is estimated to contain a quadrillion synapses, the incredible complexity of neural development becomes clear.
The nervous system is formed from a flat sheet of neuroepithelial cells on the dorsal (upper) surface of the embryo. This neuro- epithelium folds in on itself to form the neural tube, which will eventually develop to form the brain at one end and the spinal cord at the other. In the centre of the tube is a hollow cavity, called the lumen, which will eventually form the brain’s ventricular system and the spinal canal, in which the cerebrospinal fluid will circulate. Neurons are generated by cell division at the ventricular zone, a proliferative tissue found at the interface between the neural tube and the lumen.
During development of the cerebral cortex, the first neurons to be produced are radial glial cells. These are bipolar cells – they have two processes, which together span the radius of the developing neural tube. One of the processes is attached to the ventricular zone on the inner surface of the neural tube; the other extends out towards, and attaches itself to, the outer pial surface of the tube. As immature neurons are generated at the ventricular zone, they migrate outwards along the processes of the glial cells, and themselves form radially-aligned processes. Radial glia can also divide to generate daughter cells, which, like those generated in the ventricular zone, migrate along the radial processes formed by the “mother” cell.
As a cell reaches the end of its migration, the process attached to the ventricular zone collapses, so that the cell becomes unipolar. Subsequently, the cell bodies are translocated along the remaining process towards the pial surface. The cells migrate to their final destination in one of the six layers of the cerebral cortex, then differentiate into cortical projection neurons. The cerebral cortex forms as wave after wave of neuronal progenitors cease their outward migration. The cortex is a six-layered structure, and forms in an inside-out manner – the first cells to migrate form the inner layers, and subsequent waves of cells migrate over them to form the outer layers. The radial glia are a transient population of cells – once the migration of neuronal progenitors is complete, the scaffold is pulled down, and the glial cells differentiate into astrocytes.
Mechanical forces are known to be important in neural tube formation, and have long been thought to be involved in cell migration as well. However, studies of cell migration in the developing nervous system have focused largely on the chemical cues that govern the movements of neurons, and the involvement of mechanical forces has not been investigated directly. Now, a Japanese team have examined the role of mechanical forces in cell migration, and identifiy a novel mechanism for the migration of nascent neocortical cells. They report their findings in Current Biology.
In earlier work, Takaki Miyata and Masaharu Ogawa noticed that, during the transition from a bipolar to a unipolar morphology, the processes of migrating progenitors often contained bends, curls or hairpin loops, and suspected that this twisting may be what causes the translocation of cell bodies their processes. In order to investigate further, they isolated slices of cerebral tissue from mice that were halfway through their 28-day gestation period. The tissue was grown in culture dishes, and a fluorescent dye called DiI was used to label the processes of migrating neuronal progenitors. The migration of the cells was then visualized with time-lapse confocal microscopy.
These experiments confirmed that migrating cells do indeed have curled processes. These curls were often masked by the stretching of the processes. It was also observed that the processes were curled more during the latter stages of migration, when they are unipolar, than earlier on when they still had a second process attached to the ventricular surface. It was also noted that the processes extending towards the pial surface lacked centrosomes. (Centrosomes are structures which act as organizing centres for the cell’s cytoskeleton, the dynamic network of filamentous proteins involved in maintaining the shape of the cell and in transporting materials.)
A hairpin loop in the pial process of a migrating neuronal progenitor cell (from Miyata & Ogawa, 2007).
The authors suggest two different mechanisms by which the movements of the cells occur. One possibility is that the contraction of the pial process causes detachment of the ventricular process, leading both processes to buckle. This had been observed in an earlier study of cultured cortical progenitors. Alternatively, the stretching and twisting of the pial process causes detachment of the ventricular process. This mechanism is more consistent with the coiled, helical appearance of the processes seen in the slice cultures.
A number of microsurgical procedures were performed to determine which of these mechanisms was taking place. First, the pial process of bipolar migrating cells were severed. As a result, the migrating cells adopted a curled morphology which became less prominent with time. In about 40% of cases, the cell body was observed to migrate back towards the ventricular zone. When the ventricular processes were severed, the cells again adopted a coiled appearance, and most of the cell bodies underwent translocation toward the pial surface within 2 minutes of the procedure.
This short film clip shows the mechanism in a single migrating cell. The film proceeds at 30 frames per second, some 9,000 times faster than actual cell movement. The retraction of the process from the ventricular surface of the neural tube occurred over a period of 4 hours. The twisting of the other process, which apparently causes the retraction, is marked by flashing asterisks.
Left: scanning electron micrograph of the radial processes of migrating progenitors, with twisted pial processes pseudocoloured yellow; film clip from the Supplemental Data.
Finally, to confirm the role of cytoskeleton in the newly-identified mechanism, the function of the intermediate filaments – cytoskeletal components – was disrupted by the addition of calyculin-A, a pharmacological agent which causes the filaments to disassemble. As a result, all the bipolar cells in the culture were shortened. This was due to retraction of the pial processes, and not the ventricular processes, which remained intact. This provided some evidence that the twisting of the pial processes is dependent upon re-arrangement of the cytoskeletal intermediate filaments.
Thus, it appears that a novel spring-like mechanism for cell migration during corticogenesis has been identified. Miyata and Ogawa provide evidence that the twisting and curling of the pial processes of migrating cells causes detachment of the ventricular process, and that this detachment generates a force which propels the cell body towards the pial surface. The findings will be very useful to those researching the role of mechanical forces in brain development. However, they are not universally applicable, because not all migrating neurons behave in the same way as those investigated in this study. For example, some cells migrate without the aid of a scaffold to guide them; it is, however, possible that other mechanical forces are involved in these other migrations.
Miyata, T. & Ogawa, M. (2007). Twisting of neocortical progenitor cells underlies a spring-like mechanism for daughter cell migration. Curr Biol. 17: 146-151.
Miyata, T. et al (2001). Asymmetric inheritance of radial glial fibres by cortical neurons. Neuron 31: 727-741.