Neurons differentiate from the neuroectoderm
Although the embryology of the nervous system may seem like an odd place to begin studying the physiology of the brain, there are a number of reasons to start here. Knowledge of the embryology of the nervous system greatly facilitates comprehension of its complex organization. Events in the development of the nervous system highlight how different neuronal cell types evolve from a single type of precursor cell and how these neurons establish astonishingly specific connections. Finally, the characteristics of brain cell proliferation as well as the growth of neuronal processes during development provide insight into the consequences of brain injury.
At the stage of gastrulation, the vertebrate embryo consists of three primitive tissue layers: endoderm, mesoderm, and ectoderm (Fig. 10-4). The entire nervous system arises from ectoderm, which also gives rise to the skin. Underlying the ectoderm is a specialized cord of mesodermal cells called the notochord. Cells of the notochord somehow direct or “induce” the overlying ectoderm, or neuroectoderm, to form the neural tube in a complex process called neurulation.
FIGURE 10-4 Development of the nervous system. The left column provides a dorsal view of the developing nervous system at three different time points. The right column shows cross sections of the dorsal portion of the embryo at five different stages, three of which correspond to the dorsal views shown at the left.
The first step in neurulation is formation of the neural plate at about the beginning of the third fetal week. Initially, the neural plate is only a single layer of neuroectoderm cells. Rapid proliferation of these cells, especially at the lateral margins, creates a neural groove bordered by neural folds. Continued cell division enlarges the neural folds, and they eventually fuse dorsally to form the neural tube. The neural tube is open at both ends, the anterior and posterior neuropores. The neural tube ultimately gives rise to the brain and spinal cord. The lumen of the neural tube, the neural canal, becomes the four ventricles of the brain and the central canal of the spinal cord. Congenital malformations of the brain commonly arise from developmental defects in the neural tube (Box 10-2).
Abnormalities of Neural Tube Closure
Closure of the neural tube in humans normally occurs between 26 and 28 days of gestation. A disturbance in this process results in a midline congenital abnormality called a dysraphism (from the Greek dys [abnormal] + rhaphe [seam or suture]). The defect can be so devastating that it is incompatible with life or, alternatively, have so little consequence that it goes unnoticed throughout life. These midline embryonic abnormalities also involve the primitive mesoderm and ectoderm associated with the neural tube. Therefore, the vertebral bodies or skull (derived from mesoderm) and the overlying skin (derived from ectoderm) may be affected along with the nervous system.
The most serious neural tube defect, occurring in 1 of 1000 deliveries, is anencephaly, in which the cerebral hemispheres are absent and the rest of the brain is severely malformed. Overlying malformations of the skull, brain coverings, and scalp are present (Table 10-3). Affected fetuses are often spontaneously aborted.
Defects of Neural Tube Closure
Absence of the brain, with massive defects in the skull, meninges, and scalp
Partial brain herniation through a skull defect (cranium bifidum)
Meningeal herniation through a skull or spine defect
Spina Bifida Defects
Spina bifida occulta
Vertebral arch defect only
Spina bifida cystica
Herniation of the dura and arachnoid through a vertebral defect
Herniation of the spinal cord and meninges through a vertebral defect
The most common dysraphisms affect formation of the spinal vertebral bodies and are called spina bifida. The problem may be slight and cause only a minor abnormality in closure of the vertebral arch, called spina bifida occulta (Fig. 10-7A). This malformation affects ~10% of the population, usually at the fifth lumbar or first sacral vertebra, and generally causes no significant sequelae. If the dura and arachnoid membranes herniate (i.e., protrude) through the vertebral defect, the malformation is called spina bifida cystica (see Fig. 10-7B); if the spinal cord also herniates through the defect, it is called myelomeningocele (see Fig. 10-7C). These latter problems are often more significant and may cause severe neurological disability.
FIGURE 10-7 Variations of spina bifida. A, An incomplete vertebral arch with no herniation. B, The dura and arachnoid membranes herniate through the vertebral defect. C, The spinal cord and meninges herniate through the vertebral defect.
Genetic and nongenetic factors can cause dysraphism. Some severe forms of this condition appear to be inherited, although the genetic pattern suggests that multiple genes are involved. Nongenetic factors such as folic acid deficiency may also play a role. Mothers taking folic acid (see p. 1143) before and during the periconceptional period have a decreased risk of having a fetus with a neural tube closure defect. Current medical recommendations are that women contemplating becoming pregnant receive folic acid supplementation, and it has been suggested that bread products should be enriched with folic acid to ensure that women will have the protective advantage of this vitamin if they become pregnant. Other factors that increase the risk of these defects are maternal heat exposure (e.g., from a hot tub) and certain drugs such as the anticonvulsant valproate. Neural tube disorders can be detected during pregnancy by measuring the concentration of α-fetoprotein in maternal blood or amniotic fluid. α-fetoprotein is synthesized by the fetal liver and, for unclear reasons, increases in concentration abnormally with failure of neural tube closure.
The neural crest derives from lateral portions of the neural plate. Neural crest cells migrate to sites in the body where they form the vast majority of the PNS and most of the peripheral cells of the ANS, including the sympathetic ganglia and the chromaffin cells of the adrenal medulla. On the sensory side, these neural crest derivatives include unipolar neurons whose cell bodies are in the dorsal root ganglia, as well as the equivalent sensory cells of cranial nerves V, VII, IX, and X. Neural crest cells also give rise to several non-neuronal structures, including Schwann cells, satellite glial cells in spinal and cranial ganglia, and pigment cells of the skin.
The human brain begins to exhibit some regional specialization around the fourth gestational week (Fig. 10-5A, B). By then, it is possible to discern an anterior part called the prosencephalon, a midsection called the mesencephalon, and a posterior part called the rhombencephalon. Rapid brain growth ensues, and important new regions emerge in just another week (see Fig. 10-5C). Distinct regions called brain vesicles, which are destined to become separate parts of the adult brain, are set apart as swellings in the rostral-caudal plane (see Fig. 10-5B, C). The prosencephalon is now divisible into the telencephalon, which will give rise to the basal ganglia and cerebral cortex, and the diencephalon, which becomes the thalamus, subthalamus, hypothalamus, and neurohypophysis (the posterior or neural portion of the pituitary). Similarly, the rhombencephalon can now be divided into the metencephalon, which will give rise to the pons and cerebellum, and the myelencephalon, which becomes the medulla. Robust development of the cerebral cortex becomes apparent in mammals, especially humans, after the seventh week. This structure gradually expands so that it enwraps the rostral structures.
FIGURE 10-5 Embryonic development of the brain.
As the neural tube thickens with cell proliferation, a groove called the sulcus limitans forms on the inner, lateral wall of the neural tube (Fig. 10-6A). This anatomical landmark extends throughout the neural tube except in the farthest rostral area that will become the diencephalon and cortex. The sulcus limitans divides the neural tube into a ventral area called the basal plate and a dorsal area called the alar plate.
FIGURE 10-6 Development of the spinal cord and medulla. A, In this cross section through the neural tube, the sulcus limitans is the landmark that separates the ventral basal plate from the dorsal alar plate. The basal plate will form efferent (or motor-type) structures, whereas the alar plate will form afferent and associative (or sensory-type) functions. B, The true afferent neurons are those in the dorsal root ganglion, which derive from neural crest cells. These afferents will contact the neurons in the alar plate, which will become associative. C, The basal plate has developed into the ventral horn and intermediolateral column (motor), whereas the alar plate has developed into the dorsal horn (associative). D, The basal plate has developed into nuclei with motor functions, whereas the alar plate has developed into nuclei with sensory functions. The roof of the rostral medulla becomes the fourth ventricle. E, This cross section shows the same gross separation between motor and associative/sensory functions as is seen with the rostral medulla and the spinal cord. CN, cranial nerve.
Structures that derive from the basal plate mediate efferent functions, and structures that arise from the alar plate mediate afferent and associative functions. Efferent neurons are mainly motor neurons that convey information from the CNS to outside effectors (i.e., muscles or secretory cells). In a strict sense, the only true afferent neurons are those that derive from neural crest cells and that convey sensory information from various kinds of receptors to the CNS. In the CNS, these afferent neurons synapse on other neurons derived from the alar plate; these alar plate neurons may be referred to as afferent because they receive sensory information and pass it along to other parts of the CNS. However, it is also appropriate to call these alar plate–derived neurons associative.
The development of the spinal cord and medulla illustrates how this early anatomical division into alar and basal plates helps make sense of the final organization of these complex regions. Neurons of the alar and basal plates proliferate, migrate, and aggregate into discrete groups that have functional specificity. In the spinal cord (see Fig. 10-6B, C), the basal plate develops into the ventral horn, which contains the cell bodies of somatic motor neurons, and the intermediolateral column, which contains the cell bodies of autonomic motor neurons. Both regions contain interneurons. The alar plate in the spinal cord develops into the dorsal horn, which contains the cell bodies onto which sensory neurons synapse.
In the medulla (see Fig. 10-6D, E), as well as in the rest of the brain, aggregates of neurons are called nuclei. Nuclei that develop from the alar plate are generally afferent, such as the nucleus tractus solitarii, which plays an important sensory role in the ANS. Nuclei that develop from the basal plate are generally efferent, such as the dorsal motor nucleus of the vagus nerve, which plays an important motor role in the ANS. The choroid plexus that invaginates into the lumen of the central canal is responsible for secreting cerebrospinal fluid (see pp. 278–279).
Neurons and glial cells originate from cells in the proliferating germinal matrix near the ventricles
The trillions of neurons and glial cells that populate the brain arise from rapidly dividing stem cells called neuroepithelial cells N10-3 located near the ventricles (which derive from the neural canal) of the embryonic CNS. This germinal area is divided into two regions, the ventricular zone (VZ) and the subventricular zone (SVZ). Most of the neurons in the human brain are generated during the first 120 days of embryogenesis. Growth factors such as epidermal growth factor and platelet-derived growth factor and hormones such as growth hormone influence the rate of cell division of the neuroepithelial cells. The signals that direct one immature neuron to become a cortical pyramidal cell and another to become a retinal ganglion cell are not understood. Neuroepithelial cells generate different classes of neural precursor cells that develop into different mature cell types. In the developing brain, radial glial cells (or just radial cells), so called because their processes extend from the ventricular surface to the brain's outer surface (Fig. 10-8A), appear very early in neurogenesis and generate most of the projection neurons in forebrain cortex. Inhibitory interneurons, in contrast, arise from neural precursor cells located in the SVZ. Neurons are probably not fully differentiated when first created, and their mature characteristics may depend on their interactions with the chemical environment or other cells in a specific anatomical region of the nervous system.
FIGURE 10-8 Arrangement of radial glial cells and migrating neurons. A, The upper portion is a coronal section of the developing occipital cerebral lobe of fetal monkey brain. The lower portion is a magnified view. The VZ contains the germinal cells that give rise to the neurons as well as to the cell bodies of the radial glial cells (or, simply, radial cells). These radial cells extend from the ventricular surface to the pial surface, which overlies the developing cortex. B, This more magnified view shows the cell bodies of radial cells as well as their processes that extend upward toward the cortex. Also shown are two migratory neuroblasts moving from the VZ toward the cortex along the fibers of the radial cells. The right part of the panel shows that, in the postnatal VZ, the radial cell can give rise to several major classes of cells. (Data from Rakic P: Mode of cell migration to the superficial layers of fetal monkey neocortex. J Comp Neurol 145:61–84, 1972; and Tramontin AD, García-Verdugo JM, Lim DA, Alvarez-Buylla A: Postnatal development of radial glia and the ventricular zone [VZ]: A continuum of the neural stem cell compartment. Cereb Cortex 13:580–587, 2003.)
Contributed by Bruce Ransom
The general rule that neurons are created only during embryonic development and are never replaced is valid in a practical sense for all parts of the mammalian CNS except for the olfactory bulb and the dentate gyrus of the hippocampus, which may retain a population of true stem cells. Stem cells are cells that have the ability to do the following:
2. Renew themselves over the life of the organism
3. Create fully differentiated cells through progenitor cells
4. Retain their multilineage potential throughout life
5. Replace cells lost to injury or disease
These stem cells create mature brain cells by engaging in asymmetrical cell division, which yields one stem cell and one cell that begins on the path to terminal differentiation. This latter cell is called a progenitor cell. It may continue to divide, but its progeny are committed to a particular line of cell differentiation (e.g., neurons or astrocytes, but not both). Stem cells can also engage in symmetrical division and simply create two new stem cells.
The stem cells of the nervous system are capable of generating neurons, astrocytes, and oligodendrocytes. A slowly dividing population of presumed stem cells resides in the subependymal area of the lateral ventricles (i.e., analogous to the germinal matrix of the fetal brain; see pp. 263–267). These cells are apparently the source for the olfactory neurons that are continually renewed during life. Adult stem cells from the brain proliferate in response to epidermal growth factor. It is not understood how to make these cells produce progenitor cells for neurons or glial cells. Eventually it is hoped that, by sending stem cells the right sequence of signals, it might be possible to replace neurons lost to injury or disease.
The VZ appears to produce separate progenitor cells that produce only neurons, oligodendrocytes, astrocytes, and ependymal cells (see Fig. 10-8B). The VZ does not contribute to the population of Schwann cells, which derive from neural crest tissue, or to microglial cells, which arise from the mesodermal cells that briefly invade the brain during early postnatal development. Recent work shows that the embryonic and perinatal VZ and SVZ may give rise to the adult SVZ, which is in part responsible for limited adult neurogenesis.
Neuronal progenitor cells appear earliest and produce nearly the entire complement of adult neurons during early embryonic life. Glial cells arise later in development. Neurons are confined to specific locations of the brain, whereas glial cells are more evenly distributed.
Many more neurons are created during fetal development than are present in the adult brain. Most neurons, having migrated to a final location in the brain and differentiated, are lost through a process called programmed cell death, or apoptosis (Greek for “falling off”; see p. 1241). Apoptosis is a unique form of cell death that requires protein synthesis and can be triggered by removal of specific trophic influences, such as the action of a growth factor. In contrast to necrotic cell death, which rapidly leads to loss of cell membrane integrity after some insult causes a toxic increase in [Ca2+]i, apoptosis evolves more slowly. For example, in the retina, ~60% of the ganglion cells and thus ~60% of the retinal axons are lost in the first 2 weeks of extrauterine life as a result of programmed retinal ganglion cell death. This process of sculpting the final form of a neuronal system by discarding neurons through programmed death is a common theme in developmental biology.
The number of glial cells in different areas of the brain appears to be determined by signals from nearby neurons or axons. For example, in the optic nerve, the final number of glial cells in the nerve is closely determined by the number of axons. When programmed cell death is prevented by expression of the bcl-2 gene in transgenic animals, the number of axons in the optic nerve, as well as the number of astrocytes and oligodendrocytes, is dramatically increased. Thus, glial cell–axon ratios remain relatively constant. The axon-dependent signal or signals responsible for these adjustments in glial cell number are not known, but the process appears to operate by influencing both survival and proliferation of glial cells.
Neurons migrate to their correct anatomical position in the brain with the help of adhesion molecules
During embryogenesis, the long processes of radial cells create an organized, cellular scaffolding on which neurons can migrate to their final position in the brain shortly after they appear. Migrating neurons contact radial cells (see Fig. 10-8B) and move along their processes toward their final positions in the developing cortex. Thus, the prearranged positions of these radial processes determine the direction of neuronal migration. The importance of the radial framework for assisting neuronal migration is illustrated by the failure of neurons to populate the cortex normally when the radial processes are interrupted by hemorrhage in the fetal brain.
The navigation mechanisms used by migrating cells in the nervous system and elsewhere in the body are only partially understood. Proteins that promote selective cellular aggregation are called cell-cell adhesion molecules (CAMs; see p. 17) and include the Ca2+-dependent cadherins and Ca2+-independent neural cell adhesion molecules (N-CAMs). These molecules are expressed by developing cells in an organized, sequential manner. Cells that express the appropriate adhesion molecules have a strong tendency to adhere to one another. These Velcro-like molecules can assemble cells in a highly ordered fashion; experimentally, disrupted germ cells can properly reorganize themselves into a three-layered structure that replicates the normal embryonic pattern.
Another mechanism that assists migrating cells is the presence of extracellular matrix molecules such as laminin and fibronectin (see p. 17). These glycoproteins are selectively secreted by both neurons and astrocytes and form a kind of extracellular roadway with which migrating cells can interact. Growing axons express at their surface cell matrix adhesion molecules called integrins that bind laminin and fibronectin. As a result, growing axons move together in fascicles.
Perhaps the least understood mechanism related to cell migration is chemotaxis, the ability of a cell to follow a chemical signal emitted from a target cell. The tips of developing or regenerating axons, called growth cones, follow such chemical cues as they grow toward their specific targets, propelled by coordinated extension of microtubules and thin filaments. Coordinated growth-cone movement also involves continuous polymerization of tubulin and actin at the leading edge of the growth cone and depolymerization at the trailing edge. This process requires that the recognition proteins on growth cones generate intracellular signals when they encounter path-finding signals. For example, a molecule called netrin, secreted by midline cells, attracts developing axons destined to cross the midline. On the other hand, molecules like slit repel axons by interacting with specific receptors on the growth cone. Such signals steer axon growth cones, perhaps by localized changes in intracellular [Ca2+], which leads to the strategic insertion of new patches of membrane on the surface of the growth cone.
Neurons do not regenerate
Most human neurons arise in about the first 4 months of intrauterine life. After birth, neurons do not divide, and if a neuron is lost for any reason, it is generally not replaced, which is the main reason for the relatively limited recovery from serious brain and spinal cord injuries. It has been argued that this lack of regenerative ability is a design principle to ensure that learned behavior and memories are preserved in stable populations of neurons throughout life. A notable exception to this rule is olfactory bulb neurons, which are continually renewed throughout adult life by a population of stem cells or neuronal progenitor cells. N10-3 As noted above (see pp. 263–265), cells in the adult SVZ have the capacity to generate neurons and may do so to a limited extent throughout life. Learning how to induce these cells to make functional new CNS neurons after severe neural injury is the holy grail of regeneration research.
Another reason that relatively little recovery follows severe brain and spinal cord injury is that axons within the CNS do not regenerate effectively. This lack of axon regeneration in the CNS is in sharp contrast to the behavior of axons in the PNS, which can regrow and reconnect to appropriate end organs, either muscle or sensory receptors. For example, if the median nerve of the forearm is crushed by blunt trauma, the distal axon segments die off in a process called wallerian degeneration (see Box 10-3) because the sustaining relationship with their proximal cell bodies is lost. These PNS axons can slowly regenerate and connect to muscles and sensory receptors in the hand. It is believed that the inability of CNS axons to regenerate is the fault of the local environment more than it is an intrinsic property of these axons. For example, on their surface, oligodendrocytes and myelin carry molecules, such as myelin-associated glycoprotein, that inhibit axon growth. Experiments have shown that if severed CNS axons are given the opportunity to regrow in the same environment that surrounds axons in the PNS, they are capable of regrowth and can make functional connections with CNS targets. The remarkable ability of damaged peripheral nerves to regenerate, even in mammals, has encouraged hope that CNS axons might, under the right conditions, be able to perform this same feat. It would mean that victims of spinal cord injury might walk again.
Axonal Degeneration and Regeneration
Axons have their own mitochondria and produce the ATP that they need to maintain the steep ion gradients necessary for excitability and survival. In this sense, they are metabolically independent of the cell body. However, they cannot make proteins and are unable to sustain themselves if separated from the cell body (see Fig. 10-2A). If an axon is cut in either the PNS or the CNS, a characteristic series of changes takes place (Fig. 10-9):
FIGURE 10-9 Nerve degeneration. A, Normal neuron. B, Degenerating neuron. ER, endoplasmic reticulum.
Step 1: Degeneration of the synaptic terminals distal to the lesion. Synaptic transmission occurring at the axon terminal fails within hours because this complex process is dependent on material provided by axonal transport. Visible changes in the degenerating terminal are seen a few days after the lesion. The terminal retracts from the postsynaptic target.
Step 2: Wallerian degeneration. The lesion divides the axon into proximal and distal segments. The distal segment degenerates slowly during a period of several weeks in a process named after its discoverer, Augustus Waller. Eventually, the entire distal segment is destroyed and removed.
Step 3: Myelin degeneration. If the affected axon is myelinated, the myelin degenerates. The myelinating cell (i.e., the Schwann cell in the PNS and the oligodendrocyte in the CNS) usually survives this process. Schwann cells are immediately induced to divide, and they begin to synthesize trophic factors that may be important for regeneration.
Step 4: Scavenging of debris. Microglia in the CNS and macrophages and Schwann cells in the PNS scavenge the debris created by the breakdown of the axon and its myelin. This step is more rapid in the PNS than in the CNS.
Step 5: Chromatolysis. After axonal injury, most neuron cell bodies swell and undergo a characteristic rearrangement of organelles called chromatolysis. The nucleus also swells and moves to an eccentric position. The endoplasmic reticulum, normally close to the nucleus, reassembles around the periphery of the cell body. Chromatolysis is reversible if the neuron survives and is able to re-establish its distal process and contact the appropriate target.
Step 6: Retrograde transneuronal degeneration. Neurons that are synaptically connected to injured neurons may themselves be injured, a condition called transneuronal or trans-synaptic degeneration. If the neuron that synapses on the injured cell undergoes degeneration, it is called retrograde degeneration.
Step 7: Anterograde transneuronal degeneration. If a neuron that received synaptic contacts from an injured cell degenerates, it is called anterograde degeneration. The magnitude of these transneuronal effects (retrograde and anterograde degeneration) is quite variable.
Unlike neurons, glial cells can be replaced if they are lost or injured in an adult. Such repopulation depends on progenitor cells committed to the glial cell lineage. Either the progenitor cells reside in a latent state (or are slowly turning over) in adult brains or true multipotential stem cells are activated by specific conditions, such as brain injury, to produce de novo glial progenitors. The most typical reaction of mammalian brains to a wide range of injuries is the formation of an astrocytic glial scar. This scar is produced primarily by an enlargement of individual astrocytes, a process called hypertrophy, and increased expression of a particular cytoskeleton protein, glial acidic fibrillary protein. Only a small degree of astrocytic proliferation (i.e., an increase in cell number) accompanies this reaction. Microglial cells, which derive from cells related to the monocyte-macrophage lineage in blood and not from neuroepithelium, also react strongly to brain injury and are the main cells that proliferate at the injury site.