Neuronal Proliferation

Normal Development

Major proliferative events occur initially between 2 and 4 months of gestation, with peak proliferation quantitatively in the third and fourth months ( Box 5.1 ). All radially migrating neurons and glia are derived from the ventricular and subventricular zones, present in the subependymal location at every level of the developing nervous system. A second tangential stream of migration that arises from the ganglionic eminence consists of developing interneurons.

Valuable quantitative information concerning cellular proliferation is derived from studies of the deposition of brain DNA, the chemical correlate of cell number, or direct counting by optical and stereological methods ( Fig. 5.1 ). Two phases can be distinguished: the first, occurring from approximately 2 to 4 months of gestation, is associated primarily with the generation of radial glia and neurons , initially as neuronal-glial progenitors that over time undergo cell fate decisions that define them as neurons and glial cells; the second, occurring from approximately 5 months of gestation to 1 year (or more) of life, is associated primarily with glial multiplication (see Chapters 7 and 8 concerning organizational events and myelination). Similarly, some continued generation of neurons occurs later than 4 months of gestation, principally in the cerebral subventricular zone and the cerebellar external granule cell layer. Finally, proliferation of the vascular tree , arterial before venous, is particularly active during the phase of neuronal proliferation. Initially, a leptomeningeal plexus of vessels appears; this is followed in the third month by radially oriented, primarily unbranched vessels, which in the fourth and later months develop horizontal branching ( Fig. 5.2 ).

The fundamental aspects of cell proliferation in the wall of the neural tube were described first on the basis of morphological observations by Sauer in 1935. They then were delineated further by the use of radioautography with [ ³ H]thymidine-labeled DNA by Sidman, Rakic, Berry, and others 2 to 3 decades later and, still later, with bromodeoxyuridine-labeled DNA by Caviness, Rakic, and co-workers. Most recently they have been demonstrated by immunocytochemistry, computer-assisted serial electron micrographic reconstruction, time-lapse multiphoton imaging, and a variety of molecular genetic techniques ( Fig. 5.3 ).

Cells at the periphery of the ventricular zone (VZ) were shown to replicate their DNA, migrate away from the ventricular (sometimes called apical ) surface, and divide ; the two daughter cells were then noted to migrate back to the periphery of the VZ. This to-and-fro migration , or interkinetic nuclear migration , is repeated each time DNA replication and mitosis occur in the VZ. In some regions of the forebrain, a subventricular zone (SVZ) of proliferating cells can also be identified (see Fig. 5.3B ). In the monkey cerebrum, studied in detail by Rakic and co-workers and by others, the VZ gives birth to most neurons, and the SVZ is the point of origin of some later-appearing neurons (e.g., upper layers of cerebral cortex and later subplate neurons) and most glia. When cells withdraw from the mitotic cycle and cease proliferative activity, they migrate into the intermediate zone on their way to forming the cortical plate (see Chapter 6 concerning neuronal migration). The elegant work of Caviness and co-workers defined the G ₁ phase of the cell cycle as the molecular control point for these critical proliferative events.

Rakic’s studies of cortical development in monkeys led to the conclusion that, in the earliest phases of proliferation, progenitor cells divide symmetrically into two additional progenitor cells and that proliferative units of neuronal progenitor cells develop in this way ( Box 5.1 ). This process determines the number of proliferative units in the ventricular-subventricular zones. Later, at a time comparable to the second half of the second month of gestation in the human, the number of these proliferative units becomes stable as the progenitor cells begin to divide asymmetrically (i.e., each division results in dissimilar cells, one of which is a stem cell and the other of which is a postmitotic neuronal cell). These asymmetrical divisions determine the size of each proliferative unit (see Box 5.1 ). As the proliferative phase progresses, proportionately more postmitotic neuronal cells and fewer stem cells are produced. Rakic concluded that the neurons of these proliferative units migrate together in a column to form the neuronal columns of the cerebral cortex ( Fig. 5.4 ), but there is also evidence, from studies in the developing ferret nervous system, that there is dispersion of cells across the would-be columnar territories arising from each neuronal-glial progenitor cell. Other factors contribute to the complete functional organization of the cerebral cortex (see Chapters 6 and 7 concerning neuronal migration and organizational events), but the general principle is the generation of neuronal units in the ventricular-subventricular zones with subsequent migration of these groups. Rakic showed that the distinguishing features of the kinetics of neuronal proliferation in primates versus species with smaller neocortices are a longer cell cycle duration and, particularly, a more prolonged developmental period of neuronal proliferation. Thus the total number of proliferative units of neuronal cells generated is much greater in the primate compared with nonprimate species with smaller neocortices.

At least two types of neuronal progenitors are present in the VZ: (1) a short neural precursor that has a ventricular endfoot and a leading process of variable length and (2) the radial glial cell that spans the entire cortical plate with contacts at both the ventricular and pial surfaces ( Fig. 5.5 ). The former progenitor previously was considered the principal neuronal precursor cell. An exciting advance in the understanding of neuronal proliferation was the identification of the radial glial cell as another major neuronal progenitor in the VZ . Radial glial cell bodies are mainly located in the VZ (inner radial glial cells) or in the SVZ (outer radial glial cells). The major roles of this cell were initially considered to be a glial guide for migrating neurons and, later, a source of astrocytes. However, more recent studies based on immunocytochemical and molecular techniques indicate that radial glial cells give rise to many neurons generated in the VZ, particularly radially migrating excitatory projection cortical neurons. Comparative studies of outer radial glial cells across species have demonstrated that an expansion in this population of cells may correlate with brain size. Thus the term radial glial cell (which we continue to use) may ultimately be replaced by radial glial progenitor or radial progenitor . When the radial glial cell functions as a progenitor that eventually results in differentiation into a neuron, the clonally related neuron so generated then migrates along the parent radial glial fiber (see Fig. 5.5 ).

These elegant proliferative events involving the radial glial cell as neuronal progenitor are modulated by several key signaling pathways involving the Notch receptor, the ErbB receptor (through the ligand neuregulin), and the fibroblast growth factor receptor. Other critical molecular determinants include beta-catenin, a protein that functions in the decision of progenitors to proliferate or differentiate. Finally, of particular importance in the regulation of radial glial production of neurons are calcium waves propagating through connexin channels of the radial glial cell. Calcium entry is critical in the regulation of the cell cycle. Subsequent to neurogenesis, radial glial cells produce astrocytes and other glial cells (e.g., oligodendroglia). Radial glial cells also facilitate neurite (axon and dendrite) growth and subsequent connectivity. In addition, it appears likely that radial glial cells give rise to cells that persist in the SVZ of the adult brain as stem cells capable of producing neurons. The multiple functions of radial glial cells are summarized in Box 5.2 .

The classical understanding of neuronal proliferation and migration centers on the ventricular and subventricular zones and radially migrating neurons. However, in addition, there are important proliferative centers in the median ganglionic eminence (MGE) that give rise to tangentially migrating cortical and striatal interneurons. Although there was some early evidence from studies in the mouse that interneurons arising from the same MGE progenitor maintain some clustering, more recent evidence suggests that many progenitors in the MGE often give rise to interneurons that disperse widely across the brain. There are some important interspecies issues to consider as animal models continue to inform our understanding of human brain development, particularly regarding interneuron development and circuitry. Cell lineage studies in organotypic slice cultures of human embryonic forebrain provide evidence for two GABAergic subpopulations in humans: the first, which arises from the VZ and SVZ in the dorsal telencephalon, expresses the transcription factors Dlx1/2 and Mash1 and represents about two-thirds of human neocortical GABAergic neurons; the second, which arises from the MGE of the ventral telencephalon, contains neurons that are transcriptionally distinguished from the first in that they are Dlx1/2-positive but Mash1-negative.

Disorders

Disorders of neuronal proliferation would be expected to have a major effect on CNS function. Because of difficulties in quantitating neuronal populations, however, proliferative disorders are often difficult to define by conventional neuropathological examination. Even when the disorder is so extreme that the brain is grossly undersized (as in microcephaly) or oversized (as in macrocephaly), defining the nature and severity of the proliferative derangement is also difficult by conventional techniques. (Although theoretically there is the possibility that the disorders relate to alterations in later-occurring normal apoptotic events, we consider these to be disorders of proliferation unless evidence of an apoptotic disorder is recorded.) In the following discussion, we focus on these two extremes of apparent proliferative disorders, emphasizing that conclusions about the nature of the disorders can be drawn only cautiously.

Microcephaly

Microcephaly means “small head,” as opposed to micrencephaly , which means “small brain.” We will use the former term, because head size in living patients—measured as occipitofrontal circumference—is used as an approximation of brain size. Barring severe cranial defects resulting in premature skull closure, small brain size is generally considered the reason for small head size. We distinguish primary microcephalies , apparently related to impaired neuronal proliferation resulting in too few neurons, from microcephalies secondary to destructive disease ( Box 5.3 ). The latter relate to hypoxic-ischemic, infectious, metabolic, or other destructive events that usually occur following completion of cerebral neuronal proliferative events near the end of the fourth month of gestation (see Chapter 16, Chapter 20, Chapter 25, Chapter 28, Chapter 34, Chapter 35 ). The primary microcephalies that have been shown most clearly to be related to impaired neuronal proliferation include the autosomal recessively inherited disorders, often categorized as microcephaly vera . Thus in the context of this chapter, we discuss these conditions in most detail.