Glia Form the Insulating Sheaths for Axons
A major function of oligodendrocytes and Schwann cells is to provide the insulating material that allows rapid conduction of electrical signals along the axon. These cells produce thin sheets of myelin that wrap concentrically, many times, around the axon. Central nervous system myelin, produced by oligodendrocytes, is similar, but not identical to peripheral nervous system myelin, produced by Schwann cells.
Both types of glia produce myelin only for segments of axons. This is because the axon is not continuously wrapped in myelin, a feature that facilitates propagation of action potentials (see Chapter 6). One Schwann cell produces a single myelin sheath for one segment of one axon, whereas one oligodendrocyte produces myelin sheaths for segments of as many as 30 axons (Figure 4–13).
Myelin insulates the axons of both central and peripheral neurons.
A. Axons in the central nervous system are wrapped in several layers of myelin produced by oligodendrocytes. Each oligo dendrocyte can myelinate many axons. (Reproduced, with permission, from Raine 1984.)
B. This electron micrograph of a transverse section through an axon (Ax) in the sciatic nerve of a mouse shows the origin of a sheet of myelin (MI) at a structure called the inner mesaxon (IM). The sheet arises from the surface membrane (SM) of a Schwann cell, which is continuous with the outer mesaxon (OM). The Schwann cell cytoplasm (Sc Cyt) still surrounds the axon; eventually, it is squeezed out and the myelin layers become compact. (Reproduced, with permission, from Dyck et al. 1984.)
C. A peripheral nerve fiber is myelinated by a Schwann cell in several stages. In stage 1 the Schwann cell surrounds the axon. In stage 2 the outer aspects of the plasma membrane have become tightly apposed in one area. This membrane fusion reflects early myelin membrane formation. In stage 3 several layers of myelin have formed because of continued rotation of the Schwann cell cytoplasm around the axon. In stage 4 a mature myelin sheath has formed; much of the Schwann cell cytoplasm has been squeezed out of the innermost loop. (Adapted, with permission, from Williams et al. 1989.)
The number of layers of myelin on an axon is proportional to the diameter of the axon—larger axons have thicker sheaths. Axons with very small diameters are not myelinated; nonmyelinated axons conduct action potentials much more slowly than do myelinated axons because of their smaller diameter and lack of myelin insulation (see Chapter 6).
The regular lamellar structure and biochemical composition of the sheath are consequences of how myelin is formed from the glial plasma membrane. In the development of the peripheral nervous system, before myelination takes place, the axon lies within a trough formed by Schwann cells. Schwann cells line up along the axon at regular intervals that become the myelinated segments of axon. The external membrane of each Schwann cell surrounds the axon to form a double membrane structure called the mesaxon, which elongates and spirals around the axon in concentric layers (Figure 4–13C). As the axon is ensheathed, the cytoplasm of the Schwann cell is squeezed out to form a compact lamellar structure.
The regularly spaced segments of myelin sheath are separated by unmyelinated gaps, called nodes of Ranvier, where the plasma membrane of the axon is exposed to the extracellular space for approximately 1 μm (Figure 4–14). This arrangement greatly increases the speed at which nerve impulses are conducted (up to 100 m/s in humans) because the signal jumps from one node to the next, a mechanism called saltatory conduction (see Chapter 6). Nodes are easily excited because they have a low threshold. In the axon membrane at the nodes the density of Na+ channels, which generate the action potential, is approximately 50 times greater than in myelin-sheathed regions of membrane. Several cell adhesion molecules in the paranodal regions keep the myelin boundaries stable.
The myelin sheath of axons has regular gaps called the nodes of Ranvier.
A. Electron micrographs show the region of nodes in axons from the peripheral nervous system and spinal cord. The axon (Ax) runs vertically in both micrographs. The layers of myelin (M) are absent at the nodes (Nd), where the axon's membrane (axolemma, Al) is exposed. (Reproduced, with permission, from Peters et al. 1991.)
B. Regions on both sides of a node of Ranvier are rich in stable contacts between myelinating cells and the axon, to ensure that the nodes do not move or change in size and to restrict the localization of K+ and Na+ channels in the axon. Potassium permeable channels and the adhesion protein Caspr2 are concentrated in the juxtaparanode. Paranodal loops (PNL) of Schwann cell or oligodendrocyte cytoplasm form a series of stable junctions with the axon. The paranode region is rich with adhesion proteins such as Caspr2, contactin, and neurofascin (NF155). At the nodes in central axons, perinodal astroglial processes (PNP) contact the axonal membrane, which is enormously rich with Na+ channels. This localization of Na+ permeability is a major basis for the saltatory conduction in myelinated axons. The membrane-cytoskeletal linker ankyrin G (ankG) and the cell adhesion molecules NrCAM and NF186 are also concentrated at the nodes. (Reproduced, with permission, from Peles and Salzer 2000.)
In the human femoral nerve the primary sensory axon is approximately 0.5 m long and the internodal distance is 1 to 1.5 mm; thus approximately 300 to 500 nodes of Ranvier occur along a primary afferent fiber between the thigh muscle and the cell body in the dorsal root ganglion. Because each internodal segment is formed by a single Schwann cell, as many as 500 Schwann cells participate in the myelination of each peripheral sensory axon.
Myelin has bimolecular layers of lipid interspersed between protein layers. Its composition is similar to that of the plasmalemma, consisting of 70% lipid and 30% protein with high concentrations of cholesterol and phospholipid. In the central nervous system myelin has two major proteins: myelin basic protein, a small, positively charged protein that is situated on the cytoplasmic surface of compact myelin, and proteolipid protein, a hydrophobic integral membrane protein. Presumably, both provide structural stability for the sheath. Both have also been implicated as important autoantigens against which the immune system can react to produce the demyelinating disease, multiple sclerosis. In the peripheral nervous system myelin contains a major protein, P0,, as well as the hydrophobic protein PMP22. Autoimmune reactions to these proteins produce a demyelinating peripheral neuropathy, the Guillain-Barré syndrome. Mutations in myelin protein genes also cause a variety of demyelinating diseases in both peripheral and central axons (Box 4–3). Demyelination slows down, or even stops, conduction of the action potential in an affected axon, because it allows electrical current to leak out of the axonal membrane. Thus, demyelinating diseases have devastating effects on neuronal circuits in the brain, spinal cord, and peripheral nervous system.
Box 4–3 Defects in Myelin Proteins Disrupt Conduction of Nerve Signals
Because in myelinated axons normal conduction of the nerve impulse depends on the insulating properties of the myelin sheath defective myelin can result in severe disturbances of motor and sensory function.
Many diseases that affect myelin, including some animal models of demyelinating disease, have a genetic basis. The shiverer (or shi) mutant mice have tremors and frequent convulsions and tend to die young. In these mice the myelination of axons in the central nervous system is greatly deficient and the myelination that does occur is abnormal.
The mutation that causes this disease is a deletion of five of the six exons of the gene for myelin basic protein, which in the mouse is located on chromosome 18. The mutation is recessive; a mouse develops the disease only if it has inherited the defective gene from both parents. Shiverer mice that inherit both defective genes have only approximately 10% of the myelin basic protein found in normal mice.
When the wild-type gene is injected into fertilized eggs of the shiverer mutant with the aim of rescuing the mutant, the resulting transgenic mice express the wild-type gene but produce only 20% of the normal amounts of MBPs. Nevertheless, myelination of central neurons in the transgenic mice is much improved. Although they still have occasional tremors, the transgenic mice do not have convulsions and have a normal life span (Figure 4–15).
In both the central and peripheral nervous systems myelin contains a protein termed myelin-associated glycoprotein (MAG). MAG belongs to the immunoglobulin superfamily that includes several important cell surface proteins thought to be involved in cell-to-cell recognition, eg, the major histocompatibility complex of antigens, T-cell surface antigens, and the neural cell adhesion molecule (NCAM).
MAG is expressed by Schwann cells early during production of myelin and eventually becomes a component of mature (compact) myelin. Its early expression, subcellular location, and structural similarity to other surface recognition proteins suggest that it is an adhesion molecule important for the initiation of the myelination process. Two isoforms of MAG are produced from a single gene through alternative RNA splicing.
More than half of the protein in myelin in central axons is the proteolipid protein (PLP), which has five membrane-spanning domains. Proteolipids differ from lipoproteins in that they are insoluble in water. Proteolipids are soluble only in organic solvents because they contain long chains of fatty acids that are covalently linked to amino acid residues throughout the proteolipid molecule. In contrast, lipoproteins are noncovalent complexes of proteins with lipids and often serve as soluble carriers of the lipid moiety in the blood.
Many mutations of PLP are known in humans as well as in other mammals, eg, the jimpy mouse. One example is Pelizaeus-Merzbacher disease, a heterogeneous X-linked disease in humans. Almost all PLP mutations occur in a membrane-spanning domain of the molecule. Mutant animals have reduced amounts of (mutated) PLP, hypomyelination, and degeneration and death of oligodendrocytes. These observations suggest that PLP is involved in the compaction of myelin.
The major protein in mature peripheral myelin, myelin protein zero (MPZ or P0), spans the plasmalemma of the Schwann cell. It has a basic intracellular domain and, like MAG, is a member of the immunoglobulin superfamily. The glycosylated extracellular part of the protein, which contains the immunoglobulin domain, functions as a homophilic adhesion protein during myelin ensheathing by interacting with identical domains on the surface of the apposed membrane. Genetically engineered mice in which the function of P0 has been eliminated have poor motor coordination, tremors, and occasional convulsions.
Observation of trembler mouse mutants led to the identification of peripheral myelin protein 22 (PMP22). This Schwann cell protein spans the membrane four times and is normally present in compact myelin. PMP22 is altered by a single amino acid in the mutants. A similar protein is found in humans, encoded by a gene on chromosome 17.
Mutations of the PMP22 gene on chromosome 17 produce several hereditary peripheral neuropathies, while a duplication of this gene causes one form of Charcot-Marie-Tooth disease (Figure 4–16). Charcot-Marie-Tooth disease, the most common inherited peripheral neuropathy, and is characterized by progressive muscle weakness, greatly decreased conduction in peripheral nerves, and cycles of demyeli nation and remyelination. Because both duplicated genes are active, the disease results from increased production of PMP22 (a two- to three-fold increase in gene dosage). Mutations in a number of genes expressed by Schwann cells can produce inherited peripheral neuropathies.
A genetic disorder of myelination in mice can be partially cured by transfection of the normal gene that encodes myelin basic protein.
A. Electron micrographs show the state of myelination in the optic nerve of a normal mouse, a shiverer mutant, and a mutant transfected with the gene for myelin basic protein.
B. The shiverer mutant exhibits poor posture and weakness. Injection of the wild-type gene into the fertilized egg of the mutant improves myelination; the treated mutant looks as perky as a normal mouse. (Reproduced, with permission, from Readhead et al. 1987.)
Charcot-Marie-Tooth disease (type 1A) results from increased production of peripheral myelin protein 22.
A. A patient with Charcot-Marie-Tooth disease shows impaired gait and deformities. (Reproduced, with permission, from Charcot's original description of the disease, Charcot and Marie 1886.)
B. The disordered myelination in Charcot-Marie-Tooth disease (type 1A) results from increased production of peripheral myelin protein 22 (PMP22).
1. Sural nerve biopsies from a normal individual (Reproduced, with permission, from A. P. Hays) and from a patient with Charcot-Marie-Tooth disease. (Reproduced, with permission, from Lupski and Garcia 1992.) In the patient's biopsy the myelin sheath is slightly thinner than normal and is surrounded by concentric rings of Schwann cell processes. These changes are typical of the recurrent demyelination and remyelination seen in this disorder.
2. The increase in PMP22 is caused by a duplication of a normal 1.5 megabase region of the DNA on the short arm of chromosome 17 at 17p11.2-p12. The PMP22 gene is flanked by two similar repeat sequences (CMT1A-REP), as shown in the normal chromosome 17 on the left. Normal individuals have two normal chromosomes. In patients with the disease (right) the duplication results in two functioning PMP22 genes, each flanked by a repeat sequence. The normal and duplicated regions are shown in the expanded diagrams indicated by the dashed lines. (The repeats are thought to have given rise to the original duplication, which was then inherited. The presence of two similar flanking sequences with homology to a transposable element is believed to increase the frequency of unequal crossing-over in this region of chromosome 17 because the repeats enhance the probability of mispairing of the two parental chromosomes in a fertilized egg.)
3. Although a large duplication (3 megabases) cannot be detected in routine examination of chromosomes in the light microscope, evidence for the duplication can be obtained using fluorescence in situ hybridization. The PMP22 gene is detected with an oligonucleotide probe tagged with the dye Texas Red. An oligonucleotide probe that hybridizes with DNA from region 11.2 (indicated by the green segment close to the centromere), is used for in situ hybridization on the same sample. A nucleus from a normal individual (left) shows a pair of chromosomes, each with one red site (PMP22 gene) for each green site. A nucleus from a patient with the disease (right) has one extra red site, indicating that one chromosome has one PMP22 gene and the other has two PMP22 genes. (Adapted, with permission, from Garcia et al. 1991.)
Astrocytes Support Synaptic Signaling
Astrocytes are star-shaped glia found in all areas of the brain; indeed, they constitute nearly half the number of brain cells. They play important roles in nourishing neurons and in regulating the concentrations of ions and neurotransmitters in the extracellular space. But astrocytes and neurons also communicate with each other to modulate synaptic signaling in ways that are still poorly understood.
Astrocytes have large numbers of thin processes that enfold all the blood vessels of the brain, and ensheath synapses or groups of synapses. By their intimate physical association with synapses, often closer than 1 μm, astrocytes are positioned to regulate extracellular concentrations of ions, neurotransmitters, and other molecules (Figure 4–17).
Astrocyte processes are intimately associated with synapses.
A. Each of four adjacent astrocytes appears to occupy a distinct volume, with only a small overlap at the ends of their processes. In this overlap area astrocytes are connected to each other by gap junctions. Bar = 20 μm. (Reproduced, with permission, from Bushong et al. 2002.)
B. This high-voltage electron micrograph shows several thick processes emanating from the cell body of an astrocyte and branching into extraordinarily fine processes. The typical astrocyte envelopment of a blood vessel is shown at lower right. (Reproduced, with permission, from Hama, Ari, and Kosaka 1994.)
C. The processes of astrocytes are intimately associated with both presynaptic and postsynaptic elements. 1. The close association between astrocyte processes and synapses is seen in this electron micrograph of hippocampal cells. (Reproduced, with permission, from Ventura and Harris 1999.) 2. Glutamate released from the presynaptic neuron activates not only receptors on the postsynaptic neuron but also α-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptors on astrocytes. Astrocytes remove glutamate from the synaptic cleft by uptake through high-affinity transporters. (Reproduced, with permission, from Gallo and Chittajallu 2001.)
How do astrocytes regulate axonal conduction and synaptic activity? The first recognized physiological role was that of K+ buffering. When neurons fire action potentials they release K+ ions into the extracellular space. Because astrocytes have high concentrations of K+ channels in their membranes, they can act as spatial buffers: They take up K+ at sites of neuronal activity, mainly synapses, and release it at distant contacts with blood vessels. Astrocytes can also accumulate K+ locally within their cytoplasmic processes along with Cl– ions and water. Unfortunately, accumulation of ions and water in astrocytes can contribute to severe brain swelling after head injury.
Astrocytes also regulate neurotransmitter concentrations in the brain. For example, high-affinity transporters located in the astrocyte's plasma membrane rapidly clear the neurotransmitter glutamate from the synaptic cleft (Figure 4–17C). Once within the glial cell, glutamate is converted to glutamine by the enzyme glutamine synthetase. Glutamine is then transferred to neurons, where it serves as an immediate precursor of glutamate (see Chapter 13). Interference with these uptake mechanisms results in high concentrations of extracellular glutamate that can lead to the death of neurons, a process termed excitotoxicity. Astrocytes also degrade dopamine, norepinephrine, epinephrine, and serotonin.
Astrocytes sense when neurons are active because they are depolarized by the K+ released by neurons and have neurotransmitter receptors similar to those of neurons. For example, Bergmann glia in the cerebellum express glutamate receptors. Thus, the glutamate released at cerebellar synapses affects not only postsynaptic neurons but also astrocytes near the synapse. The binding of these ligands to glial receptors increases the intracellular free Ca2+ concentration, which has several important consequences. The processes of one astrocyte connect to those of neighboring astrocytes through gap junctions, allowing transfer of ions and small molecules between many cells. An increase in free Ca2+ within one astrocyte increases Ca2+ concentrations in adjacent astrocytes. This spread of Ca2+ through the astrocyte network occurs over hundreds of micrometers. It is likely that this Ca2+ wave modulates nearby neuronal activity by triggering the release of nutrients and regulating blood flow. An increase in Ca2+ in astrocytes leads to the secretion of signals that enhance synaptic function, but the specific molecular components of these signals are not understood.
Astrocytes also are important for the development of synapses. They prepare the surface of the neuron for synapse formation and stabilize newly formed synapses. For example, astrocytes secrete substances called thrombospondins that promote the formation of new synapses. In pathological states, such as chromatolysis produced by axonal damage, astrocytes and presynaptic terminals temporarily retract from the damaged postsynaptic cell bodies. Astrocytes release neurotrophic and gliotrophic factors that promote the development and survival of neurons and oligodendrocytes. Astrocytes also protect other cells from the effects of oxidative stress. For example, the gluta thione peroxidase in astrocytes detoxifies toxic oxygen free radicals released during hypoxia, inflammation, and neuronal degeneration.
Finally, astrocytes ensheath small arterioles and capillaries throughout the brain, forming contacts between the ends of astrocyte processes and the basal lamina around endothelial cells. The central nervous system is sequestered from the general circulation so that macromolecules in the blood do not passively enter the brain and spinal cord (the "blood-brain barrier"). The barrier is largely the result of tight junctions between endothelial cells and cerebral capillaries, a feature not shared by capillaries in other parts of the body (see Appendix D). Nevertheless, endothelial cells have a number of transport properties that allow some molecules to pass through them into the nervous system. Because of the intimate astrocyte–blood vessel contacts, the transported molecules, such as glucose, come into contact with and can be taken up by astrocyte end-feet.
Choroid Plexus and Ependymal Cells Produce Cerebrospinal Fluid
Cells of the ependyma and choroid plexus are derived from immature neuroepithelium. The ependyma, a single layer of ciliated cuboidal cells, lines all the ventricles of the brain, helping to move cerebrospinal fluid through the ventricular system (Figure 4–18A). At several places in the lateral and fourth ventricles the ependyma is continuous with cells of the choroid plexus, which covers thin blood vessels that project into the ventricles (Figure 4–18B). These choroid plexus epithelial cells filter plasma from the blood and secrete this ultrafiltrate as cerebrospinal fluid. Cerebrospinal fluid production and the properties of choroid plexus cells are considered in detail in Appendix D.
Ependyma and choroid plexus.
A. The ependyma is a single layer of ciliated, cuboidal cells lining the cerebral ventricles (V). High magnification of the ependymal lining (rectangle in upper image) shows the cilia on the ventricular side of the ependymal cells.
B. The choroid plexus is continuous with the ependyma but projects into the ventricles, where it covers thin blood vessels and forms a highly-branched, papillary structure. This is the site of cerebrospinal fluid formation. High magnification shows the blood vessel core (BV) and overlying choroid plexus (CP). The arrow denotes the direction of fluid flow from capillary into ventricle during the formation of cerebro spinal fluid.
Microglia in the Brain Are Derived from Bone Marrow
Unlike neurons, astrocytes, and oligodendrocytes, microglia do not belong to the neuroectodermal lineage. Instead they derive from bone marrow. Entering the central nervous system early in development, they reside in all regions of the brain throughout life (Figure 4–19). Their functions are not well understood, although they probably play an important role in immunological surveillance in the CNS, poised to react to foreign invaders.
Large numbers of microglia reside in the mammalian central nervous system.
The micrograph on the left shows microglia in the cerebral cortex of an adult mouse (in brown, immunocytochemistry). The blue spots are the nuclei of nonmicroglial cells. The microglial cells have fine, lacy processes, as shown in the higher magnification micrograph on the right. (Reproduced, with permission, from Berry et al. 2002.)
Of all of the cells in the central nervous system, microglia are the best at processing and presenting antigens to lymphocytes and secreting cytokines and chemokines during inflammation. Thus they serve to bring lymphocytes, neutrophils, and monocytes into the central nervous system and expand the lymphocyte population, important immunological activities in infection, stroke, and immune-mediated demyelinating disease. Microglia can also become macrophages, clearing debris after infarcts (strokes) or other degenerative neurological disorders.