Jump to ContentJump to Main Navigation
The Axon$

Stephen G. Waxman, Jeffery D. Kocsis, and Peter K. Stys

Print publication date: 1995

Print ISBN-13: 9780195082937

Published to Oxford Scholarship Online: May 2009

DOI: 10.1093/acprof:oso/9780195082937.001.0001

Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2015. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 02 May 2016

Morphology of normal peripheral axons

Morphology of normal peripheral axons

(p.13) 2 Morphology of normal peripheral axons
The Axon



Oxford University Press

Abstract and Keywords

This chapter discusses peripheral axons. Topics covered include classification and general organization of peripheral axons, components of a peripheral nervous system (PNS) axon, unmyelinated PNS axons, and myelinated PNS axons.

Keywords:   peripheral nervous system, unmyelinated PNS axons, myelinated PNS, classification

An axon leaves the nerve cell body (or a proximal dendrite) at the axon hillock and gives rise to a varying number of collateral branches, each ending in sets of terminal branches equipped with presynaptic complexes. An axon consists of a relatively firm gelatinous cord of neuronal cytoplasm, the axoplasm, which is enclosed by the axonal part of the neuronal cell membrane, the axolemma.

The axon effects point-to-point transmission of electrically and chemically coded information and transports cytoplasmic materials. Propagation of electrically coded information is in practice unidirectional and non-decremental. It involves the axolemma and proceeds somatofugally at a velocity that depends on axon caliber and varies from approximately 10−1 m/s in thin axons to more than 102 m/s in thick axons (see Chapter 4). Transmission of chemical signals and transportation of cytoplasmic materials are bidirectional and take place inside the axon at several different velocities independent of axon caliber, varying from approximately 1 mm/d to more than 400 mm/d (see Grafstein and Forman, 1980; Ochs and Brimijoin, 1993; Vallee and Bloom, 1991; see also Chapter 9).

Axons instruct and control differentiation of accompanying satellite cells and of effector and receptor cells in the terminal fields. Their branching patterns are fundamental for the integration of activities between widespread sets of receptors, neurons, and effectors. Mature axons remain plastic and possess a high regenerative ability, as discussed by Bisby (Chapter 28).

Most axons reside in the central nervous system (CNS). In mammals less than 1/1000 appear in the peripheral nervous system (PNS). Of these, most cross the PNS-CNS border and should, strictly speaking, be referred to as PNS-CNS compound axons. Genuine peripheral axons (i.e., axons that remain wholly outside the CNS) are the thin axons of autonomic postganglionic neurons (Figure 2-1).

In discussing peripheral axons here, we will omit axons of the first and second cranial nerves. Nerve cell bodies of cranial and spinal ganglia send their axons into the CNS via the corresponding afferent roots. The nerve cell elongations that leave the cranial and spinal ganglia via peripheral nerves and finally terminate in various receptor fields can be classified, from the ontogenetic point of view, as dendrites (Bodian, 1962). In the adult animal these dendrites appear both structurally and functionally, except for their terminals and their somatopetal transmission of electric signals, similar to genuine axons. Conventionally, these dendrites, disguised as axons, are classified as axons when met with in peripheral nerves.

Summaries of earlier literature on the microscopic anatomy of peripheral axons are to be found in previous reviews (e.g., Berthold, 1978; Landon, 1981; Landon and Hall, 1976; Lehman, 1959; Ochoa, 1976; Sunderland, 1968; Thomas et al., 1993). Informative “updatings” about Schwann cell biology are given by Gould et al. (1992) and Kleitman and Bunge (Chapter 5). Useful historical reviews also have been presented recently (Koppenhöfer et al., 1992; Thomas et al., 1993), and the microscopic anatomy of myelinated axons in the CNS has been reviewed by Hildebrand et al. (1992) and Hirano and Llena (Chapter 3). This chapter will refer mainly to mammalian axons, with particular emphasis on the specialized paranode-node-paranode (PNP) region.


Peripheral axons are accompanied by Schwann cells and rigorously demarcated from the surrounding connective tissue compartment (i.e., from the endoneurial space) by the Schwann cell basement membrane. According to the organization of their attending Schwann cells, mature peripheral axons are classified as unmyelinated or myelinated (Figures 2-2 and 2-3). In the former case, one or several axons are more or less submerged in longitudinal troughs on the outside of trains of interconnected Schwann cells. A string of longitudinally arranged and closely apposed Schwann cells and the axons lodging therein constitute an unmyelinated or a Remak fiber (see Ochoa, 1976). In a myelinated nerve fiber, a single axon is associated with a train of longitudinally arranged Schwann cells. Each of these Schwann cells supports, in cooperation with the axon, the integrity of a well-defined segment of the myelin, (p.14)

                      Morphology of normal peripheral axons

fig. 2-1. Main types of axons in the PNS. 1, α, β, and γ axons; myelinated efferent axons. 2, Unmyelinated and myelinated afferent axons; a = genuine axon in sensory root, b = dendrite disguised as axon. 3, Myelinated preganglionic axon. 4, Unmyelinated postganglionic axon. Thick lines indicate the “stereotype” part of an axon. Axons of the first, second, mesencephalic portion of the fifth, and eighth cranial nerves are not represented in the figure.

sheath, a semicrystalline lipid-protein-water layer that tightly holsters the axon.

Axon diameter (caliber) commonly is referred to as “d.” The caliber spectrum of peripheral mammalian axons extends from 0.1 μm to about 20 μm, with unmyelinated axons being less than 2 μm and myelinated ones being more than 1 to 2 μm in peripheral nerve. Length of PNS axons varies from less than a centimeter to many meters, with the maximum length depending on animal size. In most neurons with axons in the PNS, more than 90% of the cell volume is axonal.

Branching of PNS axons is more or less restricted to effector and receptor fields (Zenker and Hohberg, 1973a, b). The degree of branching is nil or very small in the spinal roots (see, however, Langford and Coggeshall, 1979), increases gradually along the proximodistal course of the axon, and becomes high or very high close to or inside the terminal field (Eccles and Sherrington, 1930; Gilliatt, 1966). Axon caliber remains fairly constant between branching points (Thiel, 1957). Because axon branches in most cases are thinner than their parent axon, the shift to the left of the caliber spectrum, noted when tracing a peripheral nerve trunk in the distal direction, should depend on branching rather than tapering (Quilliam, 1956).

The part of an axon that extends in the PNS from the CNS border or from a ganglion to the field of termination is here referred to as the stereotype axon. The stereotype axon shows a monotonously repeated modular pattern, defined by the extent of individual Schwann cells and interrupted only by a few branch points. Morphological interest centers at five regions: (1) the modular stereotype axon; (2) branching points (see Pfeiffer and Friede, 1985); (3) the PNS-CNS transition (see Berthold et al., 1993a; Fraher, 1992); (4) terminal branches of the effector and receptor fields, including motor endplates, terminals of pre- and postganglionic axons (Trimble et al., 1991), and receptor terminals (Bannister, 1976); and (5) axon hillock–initial segment regions of sensory (Spencer et al., 1973) and autonomic neurons. We will deal with the first of these regions, the modular stereotype axon.



The axolemma (Figures 2-3 and 2-5) appears, after conventional electron microscopic preparation, like an 8 nm thick, asymmetric, triple-layered unit membrane, the inner electron-dense layer being the thicker and more osmiophilic one; these features are not seen in the Schwann cell plasma membrane, which is symmetric (Elfvin, 1963). Freeze-fracturing has seen in the outside of the inner leaflet of the axolemma, the P-face, is comparatively rich in intramembranous particles (IMPs). The inside of the outer leaflet, the E-face, displays, except at nodes of Ranvier (see below), relatively few IMPs. Many E-face IMPs are about 10 nm in size and may represent voltage-sensitive Na+ channels (Stolinsky and Breathnach, 1982; Tao-Cheng and Rosenbluth, 1980; see also Chapters 11 and 21).

The very much restricted extracellular space found between the axolemma and the facing cell membrane (the adaxonal membrane) of an accompanying Schwann cell is referred to as the periaxonal space. According to the preparatory procedure used, the periaxonal space has been reported to measure from about 30 down to just a few nanometers (Figures 2-3 and 2-5). At some sites the two facing cell membranes appear to fuse and form five-layered membrane complexes, reminiscent of tight junctions (Figure 2-5). Freeze-etching studies do not reveal tight junctions between the axolemma and the adaxonal Schwann cell membrane, however. The interaction between the neuron and its Schwann cells takes place via the axolemma, which transfers the signals necessary to control proliferative and myelin-producing behaviors of the Schwann cells and to some extent the setting of axon size and clustering of voltage-dependent sodium channels in the future nodal axolemma (Joe and Angelides, 1992; Ledeen et al., 1992; Neuberger and DeVries, 1992; Poduslo, 1993; Suter et al., 1993; Trapp et al., 1988; Waxman, 1987; see also Chapter 6).


                      Morphology of normal peripheral axons

fig. 2-2. Transversely sectioned dorsal root of adult cat (× 13,000). The “duck”-shaped Schwann cell unit—Remak fiber—contains 82 unmyelinated axons. These axons are distributed in groups (*) or solitary (×) in troughs formed by the Schwann cell. Encircled axons lack surrounding Schwann cell cytoplasm but are demarcated from the endoneurial space (E) by a basement membrane. M, myelinated axons. (From Berthold, 1978.)

                      Morphology of normal peripheral axons

fig. 2-3. Crosscut Schwann cell unit of five unmyelinated axons (× 55,000). The axoplasm (A) contains mitochondria (m), neurofilaments, microtubules, and axoplasmic reticulum. The largest axon (A) is surrounded by three to four thin, helically wrapped layers of Schwann cell cytoplasm separated by a spiraling mesaxon (dots). Arrow, periaxonal space; arrowheads, the basement membrane that surrounds the Schwann cell unit; E, endoneurial space.

A large number of the functionally most important axonal proteins are integral parts of the axolemma. Among those identified, the following should be mentioned: ion channel proteins (see Chapters 11 and 13), pump proteins, enzymes, and cell adhesion molecules. The additional presence of receptor and transducer proteins involved in axon-Schwann cell communication can be postulated (e.g., Evans et al., 1992a, b; Lieberman et al., 1989). However, isolation of PNS axolemma in quantities large and pure enough to allow biochemical (p.16)

                      Morphology of normal peripheral axons

fig. 2-4. Crosscut axoplasm of a myelinated ventral root axon in adult cat (× 120,000). The sectioning plane is through a proximal paranodal main segment. Microtubular (MT) and neurofilament (NF) domains are well separated. Arrows point at axoplasmic reticulum profiles; arrowheads, vesiculotubular membranous profiles; m, mitochondriaon.

                      Morphology of normal peripheral axons

fig. 2-5. Crosscut myelinated axon in adult cat. STIN region (× 150,000). The specimen was fixed in glutaraldehyde + tannic acid followed by OsO4 + potassium ferrocyanide. With this method, organelles appear coated by a diffuse layer of electron-dense material. The lipid (intermediate) layer of the cytomembranes appears like a “white” (clear) line, approximately 3 nm in width. There is a thick axoplasmic cortex (arrowheads) inside the axolemma (thin arrow). Neurofilaments (NF) are interconnected by diffuse electron-dense bridges. One of several microtubules is encircled. Asterisks indicate axoplasmic reticulum profiles associated with the axoplasmic cortex. Periaxonal space is marked by one dot (•) and the inner adaxonal-cytoplasmic Schwann cell compartment by two dots (••). The axolemma and the Schwann cell membrane seem to fuse at several sites (thick arrows). My, myelin sheath; m, mitochondria.

(p.17) analysis is difficult because of low recovery and contamination from myelin and Schwann cell plasma membranes (Calderon et al., 1993; DeVries, 1984; Yoshino and DeVries, 1993). The various membrane proteins seem to be rather evenly distributed in the axolemma of unmyelinated axons, but show a tendency to concentrate in the PNP regions in myelinated axons (see below). The axolemma is stabilized by the immediate subjacent part of the axoplasm, which here forms the axoplasmic cortex (Figure 2-5), an outer condensed part of the cytoskeletal microtrabecular matrix containing ankyrin, fodrin, actin, and A-60 (Baines, 1991; Fath and Lasek, 1988; Heriot et al., 1985; Hirokawa, 1982; Hirokawa et al., 1985; Levine and Willard, 1981; Rayner and Baines, 1989; Schnapp and Reese, 1982; Tsukita et al., 1986).


The viscosity of mammalian axoplasm is five times that of water (Haak et al., 1976). The elemental composition of axoplasm in myelinated rat fibers has been reported by LoPachin et al. (1993). The axoplasm can be described as consisting of a cytosol and formed elements. The formed elements are (1) the axoplasmic organelles and (2) axoplasmic inclusions.

Axoplasmic Organelles.

This group of formed elements includes mitochondria, the axoplasmic–smooth endoplasmic reticulum, the cytoskeleton, vesiculotubular membranous (VT) profiles, dense lamellar bodies, multivesicular bodies, and membranous cisterns. Most organelles show their highest occurrence and/or their most elaborate arrangement in the PNP regions of large myelinated axons.


Axonal mitochondria are 0.1 to 0.3 μm in diameter and 0.5 to more than 10 μm in length (Figures 2-3 through 2-5; see also Figures 2-30 through 2-33). Their flat or tube-like cristae are as a rule oriented longitudinally. Electron-dense intramitochondrial granules are scarce. The concentration of axonal mitochondria decreases with increasing axon diameter from about 2/μm2 of cross-sectioned axoplasm in the thinnest unmyelinated axons to about 0.1/μm2 in the thickest myelinated ones (Tables 2-1 and 2-2). Transportation of mitochondria takes place in a discontinuous manner: Periods of antero- or retrograde movement at a velocity similar to that of other organelles are interrupted by periods of rest (Forman, 1987). The outer membrane of the mitochondria is considered multivalent in terms of sites capable of interacting with the force-generating apparatus. Two thirds of the mitochondria moving in an average myelinated axon seem to be transported retrogradely and the rest anterogradely (Takenaka et al., 1990). An interesting relation between mitochondria and the axoplasmic reticulum has been demonstrated ultrastructurally by Spacek and Lieberman (1980), who discussed the possibility that some mitochondria may be transported inside the reticulum, from which they eventually bud off.

The axoplasmic reticulum.

The axoplasmic reticulum (Figures 2-3 through 2-5; see also Figures 2-13, 2-14, 2-25, and 2-31 through 2-33) forms a continuous system of delicate membranous tubules about 20 to 60 nm in diameter and flattened sacks up to 100 nm in width, which branch and interconnect frequently as they extend lengthwise throughout the axon from the axon hillock into the terminals (Lindsey and Ellisman, 1985; Rambourg and Droz, 1980). The axoplasmic reticulum connects to the rough endoplasmic reticulum of the soma in the proximal part of the axon hillock via a “transitional endoplasmic reticulum” (Lindsey and Ellisman, 1985; see also Quatacker, 1981; Stelzner, 1971). It occupies just a few percent of the axonal cross-sectional area. There are about 25 axoplasmic reticulum profiles per μm2 of cross-sectioned axoplasm in thin axons, which corresponds on the average to one profile in a 0.2-μm axon. In a thick, myelinated axon there are two

table 2-1. This table cannot be displayed for copyright reasons.

This table is available in the print edition


table 2-2. This table cannot be displayed for copyright reasons.

This table is available in the print edition

to three profiles per μm2 and 300 to 500 in the whole transverse axon area. In amphibian PNS axons, the axoplasmic reticulum appears quite distinct after incubation in ruthenium red, and it has been speculated that endocytotic vesicles, generated at the nodes of Ranvier in myelinated axons, could be the vehicle interconnecting the endoneurial space and the axoplasmic reticulum (Singer et al., 1972).

Droz et al. (1975) depicted the axoplasmic reticulum as a continuous meshwork that at some sites forms “subaxolemmal plates” and splits into a primary and a secondary system as it reaches into the preterminal branches. The primary system runs just inside the axolemma and extends to the presynaptic grids. The secondary system occupies the core of the axoplasm and breaks up into spherical units the size of synaptic vesicles (Droz, 1976). In view of the branching and a minimum caliber of axoplasmic reticulum tubules that is less than 20 nm, the proposition that the axoplasmic reticulum extends uninterrupted all along the axon has been hard to contradict. Results of cold-block experiments indicate that the axoplasmic reticulum “itself is not rapidly transported” either anterogradely or retrogradely (Ellisman and Lindsey, 1983). Most likely the axoplasmic reticulum is a vector in slow anterograde transport (Lindsey and Ellisman, 1985), is not involved in retrograde transport (Rambourg and Droz, 1980), and acts as a reservoir for Ca2+ ions (Duce and Keen, 1978).

The cytoskeleton.

Modern electron microscopic techniques (see below) reveal the cytoskeleton as the most conspicuous of the axoplasmic organelles (Burgoyne, 1991). It consists of the microtubules, the neurofilaments, and the microtrabecular matrix, including the microfilaments (Figures 2-4 and 2-5; see also Figures 2-13, 2-14, 2-24, and 2-31 through 2-33) (see Ellisman and Porter, 1980; Hirokawa, 1991). It determines the growth pattern, the size, and the shape of the axon, stabilizes the axolemma, and contains the machinery necessary for interaction with the anterograde transport “motor” (kinesin) and the retrograde “motor” (dynein) (Sheetz and Martenson, 1991). Bidirectional transport and its molecular correlates in axons differ in several aspects from those in the cell body and the dendrites (p.19) (Bunge, 1986; Hollenbeck, 1989). There are also a number of differences in the composition of the axonal cytoskeleton with regard to type of peripheral nerve, distance from cell body, and age (Oblinger et al., 1987; Tashiro and Komiya, 1991; Watson, 1991).

Microtubules are about 25 nm thick, are unbranched, and vary from a few to more than 1000 μm in length (Bray and Bunge, 1981; Tsukita and Ishikawa, 1981). They lay solitary or in small groups forming bundles (Lewis et al., 1989) that extend longitudinally in the axon. The occurrence of microtubules decreases with increasing axon size from about 100/μm2 of cross-sectioned axoplasm in the thinnest axons to about 10/μm2 in the largest ones (Tables 2-1 and 2-2). The microtubule packing also varies along the course of an axon as well as between afferent and efferent axons (Pannese et al., 1984; Saitua and Alvarez, 1989; Zenker and Hohberg, 1973a, b; Zenker et al., 1975). The distal end of a microtubule (i.e., the end furthest from the cell body) is referred to as the plus-end. The opposite end is the minus-end. Microtubules grow by the assembly of tubulin monomers and oligomers at the plus-end and shorten by disassembling of the minus-end. Microtubules form the tracks along which fast axoplasmic transport of membranous organelles takes place (Weiss et al., 1987); anterograde transport is toward the plus-end and retrograde the other way. Microtubules are vulnerable structures, and about 50% of the microtubule mass is particularly labile and disassembles easily (Baas and Black, 1990; Baas et al., 1991). However, microtubules are well preserved after glutaraldehyde-tannic acid fixation.

Neurofilaments are 10 nm thick, of undefined length, and unbranched. They form diffuse bundles running longitudinally in the axon, often with a spiraling course. There are about 150 to 300 neurofilaments per μm2 of cross-sectioned axoplasm, a density rather unaffected by the size of the axon, and neurofilaments are considered to be the “major determinant” of axon size (Griffin et al., 1988; Hoffman and Griffin, 1993; Hoffman et al., 1985). Three kinds of neurofilament monomers polymerize together to form a neurofilament: NF-L (68 kD), NF-M (150 kD), and NF-H (200 kD), all three existing in many different isoforms depending on degree of phosphorylation (Nixon and Sihag, 1991). The less phosphorylated the neurofilament, the higher is the probability that the neurofilament units exist as mobile monomers or oligomers and the more labile are polymers of such units. Neurofilament density and degree of phosphorylation increase with distance from the cell body and with age.

Conventional electron microscopic preparation preserves little of the microtrabecular matrix, which then appears like thin strings of a wispy material that radiate from and interconnect the neurofilaments and some microtubules (Figure 2-4). Modern preparatory procedures, including quick freeze in combination with deep etching and rotatory shading, examination of unfixed tissue in the high-voltage electron microscope, and the use of tannic acid as a mordant in combination with glutaraldehyde fixation (Figure 2-5), all have added new features to an otherwise rather empty-looking axoplasm (e.g., Berthold, 1982; Ellisman and Porter, 1980; Tsukita et al., 1982, 1986; see also McDonald, 1984). The most conspicuous of these features is the abundance of microtrabecules—cross-linkers—radiating from the shafts of the neurofilaments and the microtubules linking them together into a dense lattice that extends throughout the axoplasm (Hirokawa, 1982; Hirokawa et al., 1985; Meller, 1987; Schnapp and Reese, 1982). The membranous axoplasmic organelles are suspended in the neurofilament-microtubule lattice and cross-linked to it by microtrabecules. Three main types of microtrabecules, all 4 to 6 nm thick, have been identified in frog myelinated axons (Hirokawa, 1982). Those between microtubules and between microtubules and membranous organelles are less than 20 nm long (Hirokawa, 1982; Langford et al., 1987). The links between neurofilaments and between neurofilaments and other formed elements are 20 to 50 nm long (Hirokawa et al., 1984) and those between the axolemma and nearby microtubules and neurofilaments are 50 to 150 nm long (Levine and Willard, 1981). After tannic acid treatment, all membranous organelles, the neurofilaments, the microtubules, and the inside of the axolemma are coated by a 3 to 30 nm thick uneven, at some sites coarsely granular, layer of material from which the microtrabecules seem to emanate (Figure 2-5; see also Figures 2-14 and 2-24).

The earlier idea that the whole neurofilament-microtubule lattice slowly moves down the axon (see Bamburg, 1988; Lasek, 1982) is now under debate (Nixon, 1993; Ochs et al., 1989; Okabe and Hirokawa, 1990, 1993). The slow component (SCa) of the axoplasmic transport, containing tubulin and the neurofilament proteins, probably consists of monomers and oligomers that become incorporated in a stationary polymer lattice and gradually renew it by balanced local disassembling and assembling (Okabe and Hirokawa, 1988). The same is probably true also for parts of the other slowly transported component (SCb), which, besides proteins heading for the terminals (e.g., synapsin and clathrin-uncoating protein complexes [Black et al., 1991]), contains proteins considered as parts of the microtrabecular matrix (e.g., actin, which during transport probably is complexed with an actin-depolymerizing factor [Bray et al., 1992]), fodrin (brain spectrin), myosin, tau (Kosik, 1993), tropomyosin, and calmodulin (Fath and Lasek, 1988; Lasek et al., 1984; Lewis et al., 1989).

Based on the distribution of the components of the (p.20) cytoskeleton, the axoplasm can be separated into three principal domains: (1) a subaxolemmal domain (i.e., the axon cortex), (2) a neurofilament domain, and (3) domains characterized by microtubules and diffuse granular material (Schnapp and Reese, 1982). The domains are not well defined in thin axons; axoplasmic reticulum is present in all three domains. The subaxolemmal domain (Figure 2-5), mainly studied in squid giant axons, where it is several μm thick, should in vertebrate axons correspond to the axoplasmic cortex and the immediate subjacent axoplasmic zone. The domain interconnects the axolemma and the cytoskeleton. In myelinated fibers it includes the outer compartment of the axoplasmic reticulum (see Figure 2-13), which covers 10% to 20% of the inner aspect of the axolemma like a coarse meshwork of flat membranous tubules associated with scattered vesicles (Berthold, 1982). Fast anterograde axoplasmic transport seems to have a predilection for the subaxolemmal domain (Rambourg and Droz, 1980; see also Smith and Forman, 1988). The neurofilament domain (see Figure 2-4) is the shape- and size-supporting framework of the axon. It is, as seen in the transversal aspect, pierced and separated into interconnecting subdomains by the “channels, tunnels, or streets,” formed by the microtubule domains (Figure 2-4) along which rapid anterograde and retrograde transport of membranous organelles takes place.

Vesiculotubular (VT) membranous profiles.

These elements are 50 to 100 nm in diameter and up to 0.5 μm (the tubes) in length (see Figures 2-30 through 2-33). Some profiles look empty. Others contain different amounts of a finely granular electron-dense material and a few are of high density. As a rule, tubular profiles appear varicose or even distinctly beaded. VT profiles often are associated with axoplasmic reticulum profiles (Ellisman et al., 1984; Tsukita and Ishikawa, 1976). They accumulate preferentially proximal to a transport blockage and are considered to be transported anterogradely in the axon (Ellisman and Lindsey, 1983; Tsukita and Ishikawa, 1980; Waxman and Black, 1985). Most VT profiles probably derive from the Golgi apparatus of the cell body and act as fast transport vectors for newly synthesized protein and lipid (Ellisman et al., 1984; Hammerschlag and Stone, 1987; Janetzko et al., 1989; Stone and Hammerschlag, 1987; Toews et al., 1987). Some may have budded off from the axoplasmic reticulum (Berthold et al., 1993b; Tsukita and Ishikawa, 1976) and some may represent endosomes.

Dense lamellar bodies and multivesicular bodies.

The shape and size of these bodies vary from short tubes of 0.1 × 0.2 to 0.1 × 0.5 μm to rounded entities about 0.5 μm in diameter (see Figures 2-30 and 2-31). Dense lamellar bodies and multivesicular bodies are rare except in the PNP regions of myelinated axons. The ultrastructural appearance and the occasional content of acid hydrolases in the dense lamellar bodies and the multivesicular bodies indicate that these organelles belong to a heterogeneous group of prelysosomal, lysosomal, and residual bodies (Gatzinsky and Berthold, 1990; Gatzinsky et al., 1988; Holtzman, 1992). Both kinds of bodies gather distal to a transport blockage and are considered to be transported retrogradely (Ellisman and Lindsey, 1983; Smith, 1980; Tsukita and Ishikawa, 1980). They accumulate endocytotically retrieved materials (Hollenbeck, 1993).

Membranous cisterns.

Membrane-demarcated and empty-looking entities more than 0.15 μm in size are classified here as cisterns (see Figure 2-30). They make up an ill-defined heterogeneous group that includes vacuoles, some of which probably are preparatory artifacts, and large endosomes.

Axoplasmic Inclusions

Granular material.

Granular material (in the present context, this refers to electron-dense granules as seen after conventional preparative procedures) is sparse and observed inconstantly. There are two main types of granules (see Figure 2-29), coarse ones of high electron density and tiny ones that aggregate to areas of a “pepper-like” appearance (see Berthold, 1978).

Coarse high-electron-dense granules measure from about 25 nm to more than 100 nm. The smaller ones are classified as glycogen granules (Phillips et al., 1972; Raine et al., 1983; Zelena, 1980). They usually form clusters and become more common with increasing age (Thomas, 1980). The larger ones have irregular shapes and lay scattered more or less at random.

Tiny granules are less than 10 nm and accumulate to form territories of a pepper-like texture in which the components of the cytoskeleton are more or less obscured. Such territories seem to be present only in the PNP regions of myelinated axons (see below). Most likely the accumulations of these tiny granules give rise to the strong and striking staining as seen light microscopically in the PNP regions of some myelinated axons (Berthold et al., 1993b; Mohammed and Landon, 1983; Raine et al., 1983). The significance of the granular material is unknown.

Lipid inclusions.

These inclusions (droplets) do not seem to occur in normal mammalian axoplasm, but are present in certain invertebrate axons (Dumont et al., 1965).


In mammals, unmyelinated axons comprise approximately 75% of axons in cutaneous nerves and dorsal spinal roots, approximately 50% in muscle nerves (see Ochoa, 1976), and approximately 30% in ventral roots (Coggeshall et al., 1974; Risling and Hildebrand, 1982). (p.21) Visceral nerve trunks (e.g., inferior cardiac nerve and gray rami communicantes) contain few myelinated nerves (Coggeshall et al., 1976; Emery et al., 1976). White rami communicantes contain two thirds unmyelinated axons (Coggeshall et al., 1976).

Stereotype Axon Port

The caliber spectrum of unmyelinated axon ranges from approximately 0.1 to approximately 2 μm (to ∼3 μm in humans). It is unimodal and has, as a rule, a peak between 0.4 and 1 μm (1.6 μm in human cutaneous nerves). In most nerves, less than 15% of the unmyelinated axons are more than 1 μm in diameter. Serial section analyses show large variations in the diameter of individual axons even over short intervals, indicating a clear varicose axon shape as seen in the longitudinal domain. For instance, over a distance of 5 to 10 μm, the largest d value of an axon may exceed the smallest by three to four times or even more (Greenberg et al., 1990). Obviously, mean d values of several measurements collected at well-separated intervals along individual axons should be used when estimating the caliber spectrum of unmyelinated axons. This would decrease the span of the spectrum and shift its upper limit toward 1 μm, the classic and controversial borderline value claimed to separate the unmyelinated part of the PNS axon population from the myelinated part (see Matthews, 1968).

Based on ultrastructural serial section analysis, Greenberg et al. (1990) have performed an elegant computer-aided three-dimensional reconstruction of several unmyelinated axon segments, thus demonstrating both their varicose shape and the distribution of axoplasmic organelles. Freeze-fracture studies of the axolemma show approximately 500 to approximately 1000 IMPs in the P-face and approximately 100 to approximately 200 in the E-face. There are no particular IMP aggregations (Black et al., 1982; Gabriel et al., 1986).

Unmyelinated axons are more or less completely submerged in longitudinal troughs formed along the Schwann cell surface (Figure 2-2). Some superficially positioned axons are separated from the endoneurial space only by the Schwann cell basement membrane. Other more deeply positioned axons are, in addition to the basement membrane, also covered by cytoplasmic Schwann cell “tongues” or septa. Here the only communication open between the periaxonal space and the outside of the Schwann cell is the 20 to 30 nm wide slit, the mesaxon, that separates the abutting Schwann cell edges, or lips, which “close” the trough. Horseradish peroxidase (Böck et al., 1970) and ferritin (Hall and Williams, 1971) enter the periaxonal space freely via the mesaxon. The width of the periaxonal space as shown by electron microscopy depends on the preparatory procedure; conventional techniques give the 20 to 30 nm space and freeze substitution or freeze-drying may obliterate it (Elfvin, 1963; Malhotra and van Harreveld, 1965). A reaction product develops in the periaxonal space after histochemical incubation for a number of esterases and phosphatases.

The number of axons in a Remak fiber (the term used to describe a nonmyelinating Schwann cell and its associated axons) varies at different transversal levels. This is an effect of the very common exchange of axons between Remak fibers. Although a Schwann cell extension usually covers the axon as it changes fibers, it is not uncommon that crossing axons are covered by just a Schwann cell basement membrane (Figure 2-2). Crisscrossing is especially prominent close to the PNS-CNS border and distally in peripheral nerves.

One way to present the number of unmyelinated axons and their related Schwann cells is to estimate the number and sizes of so-called Schwann cell units (Schwann cell unit = a basement membrane-demarcated cross-sectional profile through any transverse level of a Schwann cell plus the number of axons in that profile; the latter variable determines the size of the unit). Schwann cell units vary substantially in size. They can be large, and may reach a maximum of 100 or even more in dorsal roots (Figure 2-2), or small—only one, distally in peripheral nerves. In mouse sural nerve, for instance, approximately 15% of the Schwann cell units contain 10 or more axons and approximately 15% just one axon (Aguayo and Bray, 1975).

The length of Schwann cells in Remak fibers is difficult to measure in view of the almost imperceptible transitions between consecutive Schwann cells. Consecutive cells simply interdigitate or telescope into one another, leaving a winding, empty-looking space of 20 to 30 nm in between. At some sites, there are desmosomes and increased cytoplasmic electron density between the neighboring Schwann cells (Eames and Gamble, 1970).


In striking contrast to the situation along unmyelinated axons, the meeting points between Schwann cells that accompany a myelinated axon show a most spectacular ultrastructural organization. The meeting points are conspicuous even when examined in a low-power light microscope (Figure 2-6). They appear, irrespective of axon size, like short, constricted, and approximately 1 μm long myelin-free fiber segments. They are referred to as the nodes of Ranvier. The intervening fiber segments correspond to the extension of one Schwann cell and are called internodes. Internodes are from approximately (p.22)

                      Morphology of normal peripheral axons

fig. 2-6. a, Schematic drawing to show the terminology used. CON, constricted axon segment; FLUT, fluted paranodal axon segment (main paranodal segment); IN, internode; MYSA, myelin sheath attachment axon segment (paranodal end segment); NoR, node of Ranvier; PN, paranode; PNP, paranode-node-paranode region; STIN, stereotype internodal region. FLUT and MYSA together form the paranode. Asterisk, the axon-Schwann cell network. (Adapted from Berthold and Rydmark, 1983a; Thomas et al., 1993.) b, Variation in the length of the axon circumference of an internode of an α motor axon (d = 11 μm). Horizontal line shows the mean value of the STIN region (n = 15). The d value is obtained by division by π. (Adapted from Berthold, 1982; Thomas et al., 1993.)

200 to approximately 2000 μm long; their length is positively correlated to axon size (Figure 2-11c, d).

An internode can be separated into three main parts: a central stereotype internodal (STIN) region that forms about 95% of the internode, and a proximal and a distal end region (i.e., the proximal and the distal paranodes), each corresponding to 2% to 3% of the internode. The transition between a STIN region and the proximal and distal paranodes is indicated by an increasing outer cytoplasmic Schwann cell compartment that is very rich in mitochondria, and by irregularities in the myelin sheath contour. As a rule, the ends of an internode are dilated, the distal one slightly more than the proximal one, forming the paranodal bulbs (Lubinska and Lukaszewska, 1956; Williams and Hall, 1970, 1971; Williams and Kashef, 1968).

The organization of STIN regions appears somewhat trivial compared to that of regions formed by a node of Ranvier and its two bordering paranodes (Figures 2-7 through 2-10; see also Figure 2-12). These regions (i.e., the PNP regions) are of major functional interest. For instance, the PNP regions are the sites of generation of the action potentials necessary for impulse conduction (e.g., see Rogart, 1984; see also Chapters 4 and (p.23)

                      Morphology of normal peripheral axons                      Morphology of normal peripheral axons                      Morphology of normal peripheral axons                      Morphology of normal peripheral axons

figs. 2-7 through 2-10. These electron micrographs illustrate the appearance of large myelinated ventral root axons crosscut at different levels.

fig. 2-7. STIN regions (× 4700). Fiber Ma is cut through a level where a Golgi-Rezzonico spiral originates from the outer Schwann cell compartment. At this level, the axon contour is more or less circular. (From Berthold, 1978.)

fig. 2-8. Paranodal main segment (× 3100). Single-headed arrow points at one of six mitochondrion bags; double-headed arrow, one of the six myelin crests. (From Berthold, 1968.)

fig. 2-9. Paranodal end segment (× 3000). The MYSA part of the constricted axon segment is rather circular in outline. Asterisk, axon-Schwann cell network inside myelin crests; arrow, mitochondrial bag. (From Berthold, 1968.)

fig. 2-10. Node of Ranvier (× 13,300). N, nodal axon segment; NG, node gap; Sc, Schwann cell collar; asterisks, tuft-like domains in the node gap brush border; E, endoneural space. (From Berthold, 1978.)

(p.24) 11), they interfere with axoplasmic transport (Berthold and Mellström, 1986; Berthold et al., 1993b; Raine et al., 1983; Smith, 1989), and they are primary reactive centers in a large number of neuropathies (see Hirano, 1984; Jacobs, 1984), in Wallerian degeneration (Webster, 1962; Williams and Hall, 1971), and in collateral sprouting (Gorio, 1984; Hopkins and Brown, 1982; McQuarrie, 1985). With this background, we found it practical to describe the peripheral myelinated axon as being organized, both structurally and functionally, according to a modular principle wherein long STIN regions alternate with short PNP regions. Each of these two basic modules can be visualized as consisting of a proximal and a distal half (proximal and distal is used with reference to the cell body), with the midpoints at the Schwann cell nucleus (Bremner and Smart, 1965) and at the node of Ranvier, respectively.

In the transverse aspect, the axon forms the inner of the two concentric zones that constitute a myelinated nerve fiber; the Schwann cell forms the outer zone. The transverse shape of the axon varies according to section level. It is more or less rounded in the STIN region (Figure 2-7); whether it is circular or not is a matter of discussion (Arbuthnott et al., 1980; Berthold et al., 1982; Okamura and Tsukita, 1986). The contour of the axon usually is indented at the level of the Schwann cell nucleus, a phenomenon most striking in thin axons, which here may appear U-shaped. The axon is fluted paranodally (Figures 2-8, 2-9 and 2-12) and close to circular at the node of Ranvier (Figures 2-10 and 2-12). The variation in axon size (circumference) as seen along the internode of a cat α motor axon is illustrated in Figure 2-6.

The STIN Region

The size (i.e., the thickness), of a myelinated axon conventionally is measured in the STIN region, avoiding the Schwann cell perikaryon, and given as the d (diameter) value, assuming a circular axon contour. “d” values are nowadays, as a rule, calculated by a computer (Ewart et al., 1989) from estimations, also made by a computer, of the transverse axon area and/or the length of the transverse axon contour (perimeter). The size of mammalian peripheral myelinated axons, given as d values, is in the interval 1 to 20 μm with very few being more than 15 μm thick (Figure 2-11).

The surrounding Schwann cell zone can be separated into three concentric layers: the outer (abaxonal) and inner (adaxonal) cytoplasmic layers and, between them the myelin layer. The Schwann cell generates the myelin during development in a process of continuous spiral wrapping, close packing, and cell membrane condensation outside the axon. The innermost turn of this Schwann cell spiral corresponds to the inner cytoplasmic

                      Morphology of normal peripheral axons

fig. 2-11. Plots of the number of myelin sheath lamellae (nl), internodal length (il) and myelin sheath cross-sectional area (MYca) versus axon diameter (d) in myelinated fibers of lumbar ventral roots of kittens and cats. a, Pooled data from five animals 10 weeks to 8 months old. b, Pooled data from five cats 1 to 11 years old. c, One cat 1 year old; note lin- and log-function adaption through regression analysis, d, one cat 5 years old; regression line similar to c. e and f, same material as in a and b. Note linearity in distribution, contrary to the observations in a through d. (a, b, e, and f from Berthold et al., 1983; c and d from Nilsson and Berthold, 1988.)

                      Morphology of normal peripheral axons

fig. 2-12. Three-dimensional rendering of PNP region of a large (d = 14 μm) cat dorsal root myelinated axon, a, Longitudinal aspect. Schwann cell and myelin have been “removed”. D, distal; P, proximal. Numbers on the distal paranode refer to axon crests (see b). Hyphens indicate nodal mid-level, “x” and “xx” on the proximal paranode mark out two particularly conspicuous axonal crests (see e). b, Distal paranodal axon viewed head on from the proximal side. N, transverse “cut” through the nodal segment, c, Distal paranodal myelin sheath viewed head on from its distal end. Note inside bulgings corresponding to the Schwann cell mitochondrion bags. White dot indicates opening for the constricted axon segment. Note perspective and compare b with c. d, Outer Schwann cell cytoplasmic compartment of the distal paranode viewed from the distal end. e, f, and g, Proximal paranode displayed as in b, c, and d, respectively. The rendering is based on a series of consecutive transverse electron microscopic images through the approximately 70-μm long PNP region. The reconstruction was performed on a Crimson Elan work station in cooperation with Drs. R. Pascher and C. Malmeström at the MEDNET Laboratory, Medical Faculty, Göteborgs Universitet.

(p.26) layer and the outermost to the outer layer. Intervening turns form the compact myelin lamellae. A large number of thin cytoplasmic cords spiral through the compact myelin layer and connect the outer and inner Schwann cell cytoplasmic compartments. These cords are referred to as Golgi-Rezzonico spirals (Figure 2-7) and are the structure underlying the Schmidt-Lanterman incisures, empty-looking shearing defects that develop as preparatory artifacts in the normal myelin sheath. Calculation shows that the length of the myelin spiral (i.e., of the hypothetically unrolled myelin sheath of an axon 10 μm in diameter and with a myelin layer consisting of 140 turns each 18 nm thick) would be approximately 5500 μm, which is in contrast to the apparent distance of 2.5 μm (thickness of the myelin layer) that separates the outer and inner cytoplasmic compartments of the Schwann cell.

The numeric relations between the three fundamental structural variables in myelinated nerve fibers—axon size (d), myelin sheath thickness, and internodal length—have interested neuroanatomists for more than a century (Figure 2-11). In modern studies, using electron microscopy, myelin sheath thickness is substituted by the number of myelin lamellae multiplied by lamellar thickness. Current concepts regarding these relations in adult mammals are based partly on the careful studies of Friede and co-workers (e.g., see Friede and Beuche, 1985; Friede and Bischhausen, 1982; Friede et al., 1981; also see Arbuthnott et al., 1980; Berthold et al., 1983; Nilsson and Berthold, 1988; Yates et al., 1976).

In general, there is a linear correlation between the length of the myelin spiral (or the myelin cross-sectional area) and the circumference of the axon (Figure 2-11). The slope of the regression line varies in different species, but the correlation coefficients are about the same (.90 to .99). This suggests that the control of myelin production in peripheral nerve depends on the area of contact between axon and Schwann cell, as it does in the CNS (see Chapter 6).

Adult internodal length along PNS axons corresponds on the average to the product obtained when the longitudinal growth of a body region taking place after the passing axons acquired their first myelin (at a d value of about 1 μm) is multiplied by the length of the Schwann cells when they began myelination. Because the prospective largest axons usually are the first to myelinate, there is a positive correlation between internodal length and d (Figure 2-11). The classical ideas of rectilinear proportionality between myelin sheath thickness and internodal length versus fiber diameter must be interpreted in the light of this complex developmental interaction.

In this context, it should be noted that the bond between the axon early during myelination and its expanding Schwann cells still may be a weak one that can loosen if there is a discrepancy in longitudinal growth between axon and Schwann cells. Thus, for example, in cat lumbar spinal roots, the average increase in Schwann cell length during the first month after the α axons become myelinated is twice that of the root. The result is a wave of segmental demyelination and subsequent elimination of as many as 50% of the Schwann cells originally present along the immature α axons (see Berthold and Nilsson, 1987; Nilsson and Berthold, 1988). There is virtually no trace of this phenomenon in rat and mouse lumbar spinal roots, where longitudinal growth of axons and their myelinating Schwann cells is well matched (Berthold, 1974). With this background, the intussusception of one internode into its neighbor, as described by Kidd and Heath (1988a, and b; Kidd et al., 1992) in the sympathetic nervous system of the mouse, can be explained as a similar discrepancy in growth but with a more firmly attached and/or less axonally dependent myelin-supporting Schwann cell. The phenomenon of segmental demyelination and Schwann cell elimination is also known to take place during regeneration in adult nerves. In this situation, longitudinal growth of myelinating Schwann cells takes place along axons that already have finished their longitudinal growth (see Hildebrand et al., 1987).

The axolemma of the STIN region contains about 1000 to 1500 randomly scattered IMPs per μm2 in the P-face. In the E-face, there are about 100 randomly scattered IMPs per μm2 and a few rare particle clusters (Gabriel et al., 1986; Miller and Da Silva, 1977; Stolinski et al., 1981).

There are frequent close appositions between the axolemma and the adaxonal Schwann cell membrane (Figure 2-5). Freeze-fracture studies do not report tight junctions between the axolemma and the adaxonal Schwann cell membrane. However, a tight junctional complex has been demonstrated in the mouth of the inner mesaxon (Schnapp and Mugnaini, 1978). Thus the periaxonal space is probably sealed off from the fluid mesaxonal space enclosed in the minor dense line of the compact myelin. The periaxonal space seems to open only at the nodes, where it apparently communicates more or less freely with the endoneurial space via the node gaps, as indicated by the accessibility of various tracer substances (Hall and Williams, 1971). However, not even lanthanum can enter the periaxonal space from the nodes in large fibers (MacKenzie et al., 1984).

Differences in the organization of the axoplasm between myelinated and unmyelinated axons are illustrated further in Tables 2-1 and 2-2. One noteworthy qualitative difference is the organization of the axoplasmic reticulum. In the STIN region, this membranous network is clearly separated into two communicating compartments: a minor outer subaxolemmal (Figures (p.27) 2-5 and 2-13), and a major inner central compartment. The outer compartment is more or less embedded in the axoplasmic cortex just inside the axolemma, where it forms a single layer of membranous sacks and tubes. It covers 10% to 20% of the inner aspect of the axolemma and takes part in the rapid anterograde axoplasmic transport (Berthold, 1982; Rambourg and Droz, 1980). Vesicles of a texture similar to that of the axoplasmic reticulum and approximately 50 to 100 nm in size are common just inside the axolemma. Table 2-2 shows that the number of axoplasmic reticulum profiles associated with a unit length of the axon perimeter is roughly the same irrespective of d, whereas the number of axoplasmic reticulum profiles of the inner compartment per unit transverse area of axoplasm decreases with an increasing d value. Consequently, the myelin volume of an internode should be strongly and linearly correlated to the outer axoplasmic reticulum compartment (see Nilsson and Berthold, 1988).

The relative occurrences of mitochondria, dense lamellar bodies, multivesicular bodies, and VT profiles calculated for a large “average” cat α axon are given in Table 2-3.

The PNP Region

A PNP region includes a central node of Ranvier flanked by a proximal and a distal paranode. Each paranode can be subdivided in two segments, a main and an end segment. The latter is defined by the termination and attachment of the myelin sheath onto the axon. We will refer to this part of the PNP axon as the myelin sheath attachment (MYSA) segment. The two MYSA axon segments and the intervening nodal axon segment constitute the constricted (CON) axon segment (Figures 2-6 through 2-10 and 2-12). It should be noted that the paranodal end segment or the MYSA axon segment corresponds to the notion “paranode” as used in a number of recent works on nerve fiber morphology, in particular those dealing with freeze-fracturing. We use the word “paranode” in its extended and more traditional meaning (see Landon and Hall, 1976). The following description is based mainly on observations in mammals.

The Paranode.

The paranodal myelin sheath is free of Golgi-Rezzonico spirals and its contour is irregular. The outer cytoplasmic Schwann cell compartment is comparatively voluminous and usually forms longitudinal cords that extend with an increasing volume from the STIN segment to the node of Ranvier. The cords, one or two in small fibers and up to seven in the largest ones, “indent” the underlying myelin and axon (Figures 2-8, 2-9, and 2-12; see also Figures 2-18 and 2-19). This gives the typical transverse appearance of a deeply crenated paranodal myelin sheath (Landon and Willams, 1963; Williams and Landon, 1963). The cords of paranodal Schwann cell cytoplasm are extraordinarily rich in mitochondria (Figures 2-8 and 2-9; see also Figures 2-18 and 2-19; Table 2-3). We will refer to them as the “mitochondrion bags.” Mitochondrion bags are rich in glycogen (see Figure 2-18), several contain lipid droplets (see Figure 2-19), and some hold Marchi-positive myelinoid bodies and crystalline lamellar bodies (Reich granules). Schwann cell mitochondria become scarce close to the node. Here the outer Schwann cell compartment, adjacent to where it forms the nodal collar and the nodal brush border (see below), is rich in moderately electron-dense, round or tube-like “juxtanodal Schwann cell bodies” 50 to 200 nm in size and of unknown significance (see Figure 2-19) (Berthold and Rydmark, 1983b). Their position in a cytoplasmic domain between numerous mitochondria and a brush border, as well as their general appearance, recalls the “dense apical tubules” that appear in proximal tubule cells after oil infusion (Sui et al., 1991).

The shape of the paranodal main axon segment is a cast of the myelin sheath (Figure 2-12). It is best described as fluted. The slender longitudinal axon ridges cease at the level of the end segment, where the myelin sheath turns to the axon in order to terminate on and attach to the “trimmed down” remaining axon core. In many paranodes, particularly in those of large axons in old animals, the more nodal parts of the axon ridges break up in thin, winding, irregularly shaped processes that lay embedded in the inner cytoplasmic Schwann cell compartment (Figure 2-15). This is the axon-Schwann cell network (see Berthold, 1978; Spencer and Thomas, 1974). Usually the Schwann cell component of the network forms an exceedingly thin cytoplasmic film that separates an more or less encloses the axonal parts of the network (Figure 2-16). Both axon and Schwann cell profiles contain dense lamellar bodies and multivesicular bodies, some of which are acid phosphatase positive and should be classified as lysosomes (Figures 2-16 and 2-17) (Gatzinsky and Berthold, 1990; Gatzinsky et al., 1988). The more voluminous parts of the Schwann cell compartment also contain membranous debris and residual bodies. The axon-Schwann cell network probably reflects the ability of the PNP region to entrap, sequester, and degrade worn-out constituents. The network also may be a site where there is local lysosomal degradation of retrogradely transported endocytotically retrieved materials, and of substances that have turned around after anterograde transport from the soma (Gatzinsky and Berthold, 1990; Snyder, 1989; Spencer and Thomas, 1974). Whether there is a mutual exchange between axon and Schwann cell of whole cell fragments in the network is not known (see Buchheit and Tytell, 1992). The network appears strikingly hypertrophic in a number of neuropathies.


                      Morphology of normal peripheral axons

fig. 2-13. Longitudinal section through STIN region of large myelinated cat ventral root axon (× 36,000). This electron micrograph illustrates the network of subaxolemmal axoplasmic reticulum (arrows). The diffuse electron-dense domains represent tangentially cut myelin. The axoplasmic reticulum network is well developed in the myelin domain marked with an asterisk. Arrowheads, vesicular profiles, many in close association with the axoplasmic reticulum.

                      Morphology of normal peripheral axons

fig. 2-14. Longitudinal section through PNP region of a large ventral root fiber in the cat (× 50,000). The axoplasm (lower half of figure) is electron dense. The nodal axolemma extends between the two white vertical bars and faces the node gap (NG) between the arrowheads. The part of the axon between a bar and an arrowhead belongs to the nodal recess (R). The axon cortex of the left MYSA segment contains an approximately 3 μm long axoplasmic reticulum profile. Serial section analysis indicated that this profile is part of a flat axoplasmic reticulum sac. Asterisks in the myelin sheath (My) indicate the tips of “ear of barley”-like arrangements of the terminal cytoplasmic pockets. Preparation as in Figure 2-5.


table 2-3. The paranode and node of a large cat α axon: A morphometric descriptiona

General Fiber Characteristics:

Diameter (D) = 17.5 μm; axon diameter (d) = 12.5 μm; number of myelin lamellae = 140; myelin sheath thickness = 2.5 μm; internodal length = 1800 μm

Paranode (values refer to one paranode)

Main Segment


75 μm



mean cross-sectional area

102 μm2


maximum circumference

63 μm


mean circumference

49 μm


membrane area

3680 μm2


Myelin sheath

maximum cross-sectional area

188 μm2


maximum circumference

88 μm


Schwann cell

adaxonal membrane area

3680 μm2


adaxonal cytoplasmic maximum cross-sectional area

0.25 μm2


adaxonal cytoplasmic volume

15 μm3


endoneurial membrane area

4330 μm2


outer cytoplasmic compartment maximum cross-sectional area

28 μm2


outer cytoplasmic compartment volume

1214 μm3


mitochondria, maximum number noted in a cross section


> 2000%

mitochondria, calculated total number (assumed mitochondrion size = 0.15 × 0.5 μm)


End Segment


4 μm




4.7 μm


cross-sectional area

17.7 μm2



14.9 μm


membrane area

60 μm2



71 μm3


Terminal cytoplasmic spiral

number of turns


number of turns attached to axolemma



2530 μm

diameter of cord

0.1 μm

membrane area

795 μm2


20 μm3

Schwann cell

outer cytoplasmic compartment cross-sectional area

8 μm2


outer cytoplasmic compartment volume

32 μm3


endoneurial membrane area

200 μm2


Node of Ranvier

Schwann Cell Compartment (values represent the sum of two Schwann cells)

Extracellular interconnection area between endoneurial space and node gap

0.2 μm2

Node gap


0.3 μm


1 μm

extracellular volume

0.7 μm3

Schwann cell microvilli



mean length

1 μm

mean diameter

84 nm

total membrane area

210 μm2

total volume

4.4 μm3

walls, membrane area

43 μm2

recesses, ceiling area

11 μm2

Axon Compartment

Axon length

1 μm


Axon diameter

5 μm


Axon cross-sectional area

20 μm2


Axon circumference

16 μm


Axon membrane area

24 μm2


Axon volume

19 μm3


Paranode-Node-Paranode Axon Region



Multivesicular bodies


Dense lamellar bodies


Vesiculotubular membranous profiles


(a) Data are taken from Rydmark and Berthold (1983), Berthold and Rydmark (1983a), and Berthold et al. (1993b) and are compensated for preparative dimensional changes (Berthold et al., 1982).

(b) Percentage values represent parts of the corresponding internodal total.


                      Morphology of normal peripheral axons

fig. 2-15. Cross section of cat ventral root through paranodal main segment of medium-sized myelinated fiber (× 12,600). The axon crests inside the myelin crests have fragmented and form, together with the inner cytoplasmic Schwann cell compartment, typical axon-Schwann cell networks (asterisks). Arrow points at myelin crest covered by a substantial layer of Schwann cell cytoplasm; arrowhead, mitochondrion bag. (From Berthold, 1978.)

                      Morphology of normal peripheral axons

fig. 2-16. Cross-section through axon-Schwann cell network (× 70,000). Axonal profiles (A) are surrounded by extensions of the inner cytoplasmic Schwann cell compartment. At some sites the Schwann cell layer is just a thin film (arrowheads). At other sites it is more conspicuous (asterisks) and contains membranous debris (arrow). One axonal profile contains a multivesicular body (MVB) and another a lamellar body (+). My, myelin sheath. Dots in upper part of picture mark out the periaxonal space (•) and the inner cytoplasmic Schwann cell compartment (••). Preparation as in Figures 2-5, 2-14, and 2-19.

                      Morphology of normal peripheral axons

fig. 2-17. Detail from axon-Schwann cell network after incubation for demonstration of acid phosphatase (× 33,000). One of the larger axon profiles contains a multivesicular body filled with reaction product. Courtesy of Dr. K. Gatzinsky.

(p.31) The main changes in the organization of the axon, noted when tracing along the main paranodal segment toward the node (Figure 2-12), are the gradual reduction of the transverse axon area, the increasing delicacy and final disintegration of the axon ridges, and the increase in microtubules and organelles in the more central parts of the axon. Thus the microtubule concentration increases in proportion to the reduction of the axon transverse area (Table 2-2) (Tsukita and Ishikawa, 1981).

The termination of the myelin sheath defines the paranodal end segment. Here the axon is reduced to a narrow cylinder—the MYSA axon segment—3 to 5 μm long irrespective of axon size, and of a diameter that is one half to one third that of the STIN region (Figure 2-12) (Rydmark, 1981). As seen in a median longitudinal section, each terminating myelin lamella splits into two leaflets, forming a terminal cytoplasmic pocket that encloses a drop-shaped portion of Schwann cell cytoplasm (Figures 2-14, 2-19, and 2-20). The whole set of pockets arranged along a MYSA segment is the cross-sectional representation of one single, very thin, and continuous cord of Schwann cell cytoplasm (the lateral belt) that encircles the MYSA axon segment in a complex helical manner. Each turn of this terminal cytoplasmic spiral (TCS) corresponds to the termination of one myelin lamella. The length of the TCS is about 2500 μm in a large cat α motor fiber. The TCS represents the nodal margin—the lateral belt—of the hypothetically unrolled myelin sheath and interconnects the inner and the outer cytoplasmic Schwann cell compartments of the paranode with one another. In this respect, the TCS is equivalent to the cytoplasmic (Golgi-Rezzonico) spiral associated with an incisure of Schmidt-Lanterman, although it is less than half the length of the latter.

The innermost myelin lamella attaches through the TCS to the MYSA segment furthest from the node. Subsequent lamellae then terminate in consecutive order up to the node. However, only some 10% to 20% of the lamellae of a thick fiber attach directly to the axolemma. The turns of the TCS of the remaining 80% to 90% of the myelin lamellae coil up on top of one another in sets of 10 to more than 25 turns. In a longitudinal section, such a set gives the characteristic picture of an “ear-of-barley”-like aggregation of terminal cytoplasmic pockets (Figure 2-20) that penetrates the surrounding compact myelin for a depth of 0.5 to 1.5 μm. The specific staining properties of the TCS explain the light microscopic appearance of the “spinous cuffs of Nageotte” (Nageotte, 1911).

The turns of the TCS that reach the axolemma indent it and give it a serrated outline. The turns are closely attached to the axolemma by gap junction-like membrane complexes that, when viewed after freeze-etching, form “transverse bands” (Hirano and Dembitzer, 1982; Livingston et al., 1973; Rosenbluth, 1984; Wiley and Ellisman, 1980) and complex patterns of IMPs in both P- and E-faces. The various types of membrane connections noted between the MYSA segment and the TCS are as a whole referred to as the glial cell–axon junction (GAJ) complex of peripheral myelin. The periaxonal space in the GAJ complex is 3 to 5 nm wide and is, at least in its more nodal parts, open to the recessed node gap (see below). The turns of the TCS seem to be joined together by a system of more or less continuous tight junctions (Schnapp and Mugnaini, 1978). The outermost two to five turns of the TCS are not part the GAJ complex in large fibers, but are separated from the axolemma by a space 20 to 50 nm in width and form the “ceiling” of the recessed node gap (Figure 2-20; see also Figure 2-27). The most nodal turn of the TCS is large and high. It forms the node gap wall (Figure 2-20) and contains juxtanodal Schwann cell bodies. One study has been interpreted as suggesting that the TCS is rich in Na+ and Ca2+ (Ellisman et al., 1980). Flat axoplasmic reticulum sacks 2 × 2 μm in size, are common in the axoplasmic cortex of the MYSA segment (Figure 2-14). Whether these components are involved in the impulse-generating machinery is not known (Moran and Mateu, 1983; Wurtz and Ellisman, 1988).

The Node of Ranvier.

Nodes of Ranvier are the only sites along myelinated nerve fibers that can support the fast (see Chapter 11) de- and repolarization processes necessary for generation of action potentials. Nodes of Ranvier are thus the very sites that make rapid electrical signaling possible. This ability is mainly due to the high concentration of voltage-sensitive Na+ channels in the nodal axolemma where it is free of myelin insulation.

The morphological study of nodes of Ranvier is hampered by the their vulnerability to manipulation and chemical influences. The most serious artifacts develop when the preparatory protocol involves dehydration, which unfortunately is included in most protocols (Berthold et al., 1982; Uhrik and Stämpfli, 1981; see also Wurtz and Ellisman, 1988). In particular, the node gaps become distorted as the two myelin segments that meet at the node shrink, move apart, and split, and a node gap appears as a well-defined and relatively large opening in the myelin sheath; this is a picture not at all consistent with the in vivo or predehydration appearance, wherein a node gap is hard to define (Berthold et al., 1982; Gerhardt, 1991). The thicker the myelin sheath, the more serious are the preparatory artifacts. These difficulties probably explain why there are few systematic studies dealing with quantitative aspects of nodal ultrastructure in large fibers (see, however, Mohammed et al., 1983; Raine, 1982; Rydmark and Berthold, 1983; (p.32)

                      Morphology of normal peripheral axons

fig. 2-18. Longitudinal section through PNP region of a large ventral root fiber in the cat (× 7000). Section plane is outside the node gap. P, perinodal space between the nodal ends of the two meeting paranodes. The myelin sheaths are crenated and equipped with longitudinal crests and intervening furrows. The latter are filled with mitochondrion-rich Schwann cell cytoplasm forming the mitochondrion bags (B). Note electron-dense glycogen granules dispersed among the mitochondria. A, axoplasm inside a myelin crest.

                      Morphology of normal peripheral axons

fig. 2-19. Longitudinal section of cat ventral root through node of Ranvier and associated mitochondrion bag (B) (× 19,000). The mitochondrion bag contains, besides mitochondria (m), well-preserved lipid droplets (×) and, close to the node, aggregations of juxtanodal bodies (J,*). The two arrows delineate the nodal axolemma. The black dots are in the nodal gap walls. NG, node gap. Preparation as in Figures 2-5 and 2-14.

                      Morphology of normal peripheral axons

fig. 2-20. Longitudinal section of cat ventral root through nodal gap (NG) containing densely packed microvilli emanating from the Schwann cell nodal collars (SC) (× 70,000). The nodal collars separate the nodal gap from the perinodal space (P). Asterisks in the myelin sheath indicate the tips of “ear of barley”-like arrangements of the terminal cytoplasmic pockets. The nodal axolemma extends between the two bars (note thick electron-dense coating inside the axolemma in this GMA embedding), while the node gap proper extends between the two arrowheads. A nodal recess extends between a bar and an arrowhead. The nodal gap walls (W) overlie each recess and delineate the node gap from the myelin sheath. (From Berthold and Rydmark, 1983b.)

                      Morphology of normal peripheral axons

fig. 2-21. Longitudinal section of cat ventral root through the nodal gap of a large nerve fiber (× 100,000). The sectioning plane is outside the axon. The Schwann cell microvilli (V) are crosscut. A floccular extracellular node gap (×) surrounds the microvilli. The node gap is limited proximally and distally by the node gap walls (W). (From Berthold, 1978.)

                      Morphology of normal peripheral axons

fig. 2-22. Longitudinal section of cat ventral root through the nodal gap of a larger nerve fiber (× 100,000). The sectioning plane is just outside the nodal axon. The Schwann cell microvilli (V) are crosscut and almost circular and close to maximally packed in a hexagonal pattern. In this GMA embedding (leaving out the dehydration step in the preparation), the extracellular space is minute. W, node gap walls. (From Berthold and Rydmark 1983b.)

(p.34) Uhrik and Stämpfli, 1981). Morphometric data are presented in Table 2-3.

Like other parts of a myelinated fiber, a node of Ranvier can be said to consist of two concentrically arranged compartments: an outer Schwann cell compartment, here without myelin, and an inner axon compartment. Cell adhesion molecules (CAMs) both of the neuron–glial cell and of the neuronal type coexist specifically in the node of Ranvier, where they are associated with the Schwann cell compartment as well as with the axon. The localization here of both types of CAMs in adult myelinated nerve fiber emphasizes the importance of nodal structural integrity (Kordeli et al., 1990; Rieger et al., 1986). The part of the endoneurial compartment that surrounds a node of Ranvier is denoted the “perinodal space.”

The nodal Schwann cell compartment.

This compartment consists of three regions: (1) the outer demarcation, (2) the node gap, and (3) the node gap walls (Figures 2-19 and 2-20; see also Figures 2-27, 2-29, and 2-30).

The outer demarcation of a node of Ranvier is formed by delicate extensions (the nodal collars) of the outer cytoplasmic Schwann cell compartments of the two meeting paranodes (Figures 2-10, 2-19, 2-20, 2-23, 2-26, and 2-27). The nodal collars send a brush border of radially arranged microvilli into the node gap to the axolemma. The brush border is as a rule particularly well developed in those sectors of the outer demarcation where the nodal collars are direct continuations of a mitochrondrion bag. At some sites the nodal collars of the converging paranodes overlap and join, forming five-layered membrane complexes reminiscent of tight junctions (Figures 2-23 and 2-27). At other sites the collars are joined by clumps of lamellar material. In some sectors of the same demarcation region, the nodal collars are minimal, located 0.1 to 0.2 μm apart, and the node gap is separated from the perinodal space only by the Schwann cell basement membrane (Figure 2-27). Small tufts of Schwann cell microvilli are common outside the node gap facing the periaxonal space and occasionally observed elsewhere on the outer cytoplasmic Schwann cell compartment (see Pannese et al., 1989).

Viewed in a median longitudinal section (Figure 2-20; see also Figures 2-29 and 2-30), the node gap appears like an isosceles trapezoid whose longer and shorter bases correspond to the axolemma and the plane of the outer demarcation, respectively. Its sides are formed by the node gap walls. The node gap contains the Schwann cell brush border embedded in the node gap matrix substance (Figures 2-10 and 2-20; see also Figures 2-29 and 2-30). The matrix substance occupies the extracellular space of the gap and has the properties of a cation exchanger (Landon and Langley, 1971; Langley, 1971; Langley and Landon, 1969), with strong affinity to heavy cations such as Ag+, Cu2+, Fe2+, and Pb2+, a fact noteworthy in connection with use of histochemical methods based on the precipitation of heavy metal ions (Krammer and Lischka, 1973; Zagoren, 1984).

The node gap microvilli come close to the axolemma, and some end less than 5 nm from it. The height of the node gap increases with increasing axon size. In large axons (d = 10 μm), the gap height varies in different sectors of a node from 0.1 to as much as 2 μm (Figure 2-27). There are in such nodes about 800 to 1000 microvilli, each 70 to 80 nm in diameter and on the average 1 μm long (see Table 2-3). The seven to eight microfilaments noted in a microvillus and the lack of a terminal web in the nodal collars suggest structural kinship with the brush border of the proximal tubule cell of the kidney (see Kenny and Booth, 1978). Freezeetched nodal microvilli show P- and E-faces wherein 45% of the intramembranous particles are about 10 nm, a size fitting that of voltage-sensitive Na+ channels (Waxman and Black, 1987).

The node gap walls are formed by the most nodal turn in the TCS of each of the two meeting paranodes. Low recesses extend from the main node gap for a distance of 0.1 to 0.4 μm both proximally and distally between the axolemma and the most nodal turns of the TCS (Figures 2-20 and 2-27). We have calculated, for a large cat α fiber, that the Schwann cell faces the node gap with a cell membrane area of about 250 μm2. Approximately 75%, 20%, and 5% of this area is provided by the brush border, the node gap walls, and the “ceiling” of the gap recesses, respectively (Table 2-3). Completely “closed” node gaps with maximal packing of the microvilli are seen after embedding in water-soluble media (Figures 2-20 through 2-22) or after use of glutaraldehyde-tannic acid fixation followed by post-fixation in osmic acid and potassium ferrocyanide.

The nodal axon compartment.

This compartment, also known as the nodal axon segment, forms the mid-region of the CON segment (Figures 2-20, 2-29, and 2-30). It is approximately 1.0 to 1.5 μm long regardless of fiber size (Table 2-3) and extends between the most nodally and closely attached turn of the TCS of each of the two meeting paranodes. The nodal axon segment is as a rule barrel shaped and a few percent thicker than the adjacent MYSA segments. It is demarcated by the nodal axolemma, the area of which increases linearly with increasing axon size; in cat spinal roots, the nodal membrane area varies from about 4 μm2 in the smallest fibers to about 30 μm2 in the largest ones (Figure 2-28).

Two features in particular characterize the nodal axolemma: a uniquely high concentration, about 1500 per μm2, of E-face IMPs (Rosenbluth, 1984) considered to represent the voltage-sensitive Na+ channels; and an inside coating of electron-dense material (Figures 2-19, 2-29, and 2-30). The inside coating, which should be (p.35)

                      Morphology of normal peripheral axons

fig. 2-23. Longitudinal section of cat ventral root showing the outer aspect of the nodal gap (NG) of a larger nerve fiber (× 75,000). The axon is outside the picture to the right, the perinodal space (P) is to the left, and the nodal gap walls (W) are in the upper and lower parts of the picture. One nodal collar (Sc2) covers the nodal gap (thick arrow) and the collar of its neighbor Schwann cell (Sc1) with a delicate cytoplasmic extension that continues its extent up the perinodal wall of its neighbor. The converging Schwann cell membranes form five-layered complexes (arrows) suggesting a tight-junction contact. Between the overlapping Schwann cell collars and the nodal gap proper, there is a thin dark lamellar body (asterisk). (From Berthold and Rydmark, 1983b.)

                      Morphology of normal peripheral axons

fig. 2-24. α-axon (× 240,000) of cat ventral root. Detail from cross section through node of Ranvier. Dot is in axolemma. Large arrowheads, axolemma + axon cortex, which here includes the nodal undercoating; arrows point at axoplasmic reticulum profiles. Note small size of axoplasmic reticulum profile, marked out by arrow at the lower middle of the picture. One microtubule is encircled. Small arrowheads point at neurofilaments. NG, node gap, ×, vesiculotubular profile. Preparation as in Figure 2-5.

                      Morphology of normal peripheral axons

fig. 2-25. α-axon (× 90,000) of cat ventral root. Detail from a series of cross sections through a node of Ranvier. About 100 nm separates a from b. The axon (A) emits a nodal spine (×) containing several axoplasmic reticulum profiles. Thin arrows in a point out axolemma; thick arrows indicate coated pits. Note lack of nodal undercoating in relation to the pits. Microvilli (*) of the node gap (NG) brush border terminate close to the axolemma.

                      Morphology of normal peripheral axons

fig. 2-26. Cross section of cat dorsal root through the nodal axon segment (A) of a very small myelinated fiber (the myelin sheath, not visible at this level, contains only 14 lamellae) (× 70,000). The nodal gap is covered by Schwann cell nodal collars (NC) containing clusters of Schwann cell juxtanodal bodies (*). The nodal gap (NG) contains only a few randomly oriented Schwann cell microvilli (arrowhead). The inside electron-dense coating of the axolemma is restricted to the stretch between the arrows and at the arrowhead, i.e., to those regions where the microvilli appose the nodal axon. (From Berthold and Rydmark 1983b.)

                      Morphology of normal peripheral axons

fig. 2-27. Reconstructions of node of Ranvier. a, Cross-sectional “nodal map” reconstructed from a series of 120 consecutive longitudinal sections through the nodal region of a cat ventral root nerve fiber about 15 μm in diameter. The horizontal line represents the projection of the reconstructed median longitudinal plane into a cross-sectional plane. Stippling outside the outer contour of the nodal gap indicates sectors where the nodal gap is separated from the perinodal space by the Schwann cell nodal collars and basement membrane. Thick stippling indicates that the nodal collars are joined by tight junctions. Thin stippling means that the meeting collars are separated by a 10 to 30 nm wide gap. Lack of stippling means that the nodal gap is only delimited from the perinodal space by a basement membrane. Black bars mark places where “plugs” of lamellar material is found sealing off the outer part of the nodal gap. Scale bar: 1 μm. b, “Rolled out” semischematic reconstruction of the nodal axolemma (inner circle in a). The stippled parts (40%) indicate the nodal recesses, whereas the clear parts (60%) are those facing the node gap proper (containing microvilli). Small circles mark out coated pits. The nodal axon membrane area is 17 μm2. Scale bar: 1 μm. c, “Rolled out” semischematic reconstruction of the outer nodal gap region (outer perimeter in a). The meaning of stippling and bars is given in the legend for a. Closely packed round profiles indicate sectors of the outer part of the nodal gap where microvilli face the basement membrane. The demarcation area of the outer part of the nodal gap is 4.8 μm2, but only 0.5 μm2 is extracellular and a free passage between the nodal gap and the perinodal (endoneural) space. Because the areas indicated by thick stippling (33%) constitute regions with tight junctions, the area open for “free passage” is reduced further to 0.3 μm2 (i.e., 2% of the nodal axon membrane area). Scale bar: 1 μm. (From Berthold and Rydmark, 1983b.)

                      Morphology of normal peripheral axons

fig. 2-28. Plots of nodal variables (y-axes) versus internodal axon diameter (d) of cat spinal root myelinated fibers, a, The nodal axolemma area (AxA). b, The Schwann cell membrane area facing the nodal gap (ScA) (i.e., of microvilli, walls, and recesses), c, The nodal gap extracellular volume (GECV) (i.e., the space between microvilli and in the recesses), d, The ratio (ScA/AxA) between the nodal axolemma area (a) and the Schwann cell membrane area facing the node gap (b). The values are compensated for preparatory changes (see Berthold et al., 1982). ○, ventral root; •, dorsal root. (From Rydmark and Berthold, 1983.)

regarded as a specialized part of the axoplasmic cortex, is about 30 nm thick after conventional electron microscopic preparation and up to 100 nm thick after embedding in a water-soluble resin (GMA) (Figure 2-20) or after glutaraldehyde-tannic acid fixation and postfixation in osmic acid-potassium ferrocyanide (Figure 2-24). It contains actin (Zimmerman and Vogt, 1989) and an isoform of ankyrin—“ankyrinR”—that probably binds directly to the voltage-gated Na+ channels (see Wiley and Ellisman, 1980). After staining with ferric ions and ferrocyanide, the nodal axolemma develops a thick electron-dense precipitate, a reaction taken to indicate the presence of Na+ ion channels (Quick and Waxman, 1977; see also Zagoren, 1984). The nodal axolemma of small axons (d = 1 to 2 μm) lacks a continuous inside coating (Figure 2-26). It is instead supplied with circumscribed electron-dense patches at sites related to the tips of the few microvilli present.

Besides being finely corrugated, the nodal axolemma forms spine- and crest-like outgrowths (Figure 2-25) (Berthold and Rydmark, 1983b; Uhrik and Stämpfli, 1981). The latter run across the longitudinal fiber axis and are 1 to 4 μm in length, up to 0.6 μm high, and 50 to 400 nm thick. The spines are 50 to 100 nm in diameter and up to 0.5 μm high. Membranous profiles, similar to axoplasmic reticulum profiles, are common in both crests and spines. There are approximately 0.5 to 2.0 coated pits per μm2 of nodal axolemma. The pits are particularly common at the bases of spines and crests and in relation to the nodal recesses (Figures 2-25 and 2-27). Their presence suggests a significant exchange of material between the nodal axon and the node gap. The additional observation of coated invaginations in the nodal collars (Berthold and Rydmark, 1983b; Le Beau et al., 1987) invites speculation of an interplay between axon and Schwann cell mediated via the node gap.

The Constricted Axon Segment.

The constricted axon segment and adjacent parts of the paranodal main segment constitute together the functionally and structurally most dynamic parts of the myelinated axon. The molecular machinery necessary to generate trains of action potentials (p.38)

                      Morphology of normal peripheral axons

fig. 2-29. α-axons of cat ventral root; longitudinal sections through PNP regions from different cats. D, distal; P, proximal; asterisk, node gap. a, The axoplasm contains few organelles, (× 8000). The proximal MYSA segment and the adjacent part of the proximal paranode is rich in clusters of electron-dense glycogen-like granules 25 to 30 nm in size, b, The axoplasm of the CON segment is crammed with organelles (× 10,000). The proximal MYSA segment contains longitudinal strands of vesiculotubular profiles and several swollen mitochondria. The distal MYSA segment and the adjacent paranode are rich in dense lamellar and multivesicular bodies. Note dilation of the proximal MYSA segment. This is the typical axoplasmic picture of a segregated PNP region. In addition, the proximal paranode contains a finely granular material that gives the axoplasm a “pepper”-like appearance and obscures the cytoskeleton.

                      Morphology of normal peripheral axons

fig. 2-30. α-axon of cat ventral root. Longitudinal section through organelle-rich and segregated PNP axoplasm (× 8000). Finely granular material is scanty and concentrated to the nodal segment. D, distal; P, proximal; asterisk, node gap.

(p.39) is localized here (for review and references, see Waxman and Ritchie, 1993). The axoplasm of the CON segment is involved in antero- and retrograde transport through an axon region whose transverse area is reduced by 75% to 90% as compared to that of the adjacent STIN regions, a reduction shown to be advantageous for conduction velocity (Halter and Clark, 1993).

A CON segment is about 6 to 10 μm long and increases only slightly with fiber size. The axolemma of the CON segment consists of a median nodal field, flanked by a proximal and a distal MYSA field. Fast-depolarizing Na+ currents are generated at mammalian nodes during impulse conduction. The fast-repolarizing K+ current, well known at active amphibian nodes, is missing (Brismar, 1979, 1980; Kocsis, 1984). A fast K+ current appears at the active mammalian node if the myelin is removed from the MYSA segments. This fact, and results from electrophysiology in combination with different ion channel and enzyme-blocking experiments, freeze-fracturing, cytochemistry, and immuno-cytochemistry, give the picture (for references, see Waxman and Ritchie, 1993; see also Chapters 4 and 11) of a nodal axolemma devoid of fast, 4-aminopyridine (4-AP)-blockable K+ channels but rich in Na+ channels, slow tetraethylammonium (TEA)-blockable K+ channels, and Na,K-ATPase. The axolemma of the MYSA segment, in contrast seems to be rich in fast, 4-AP-blockable, K+ channels but poor in Na+ channels and slow K+ channels. The axolemma adjacent to a node may contain chloride channels (Strupp and Grafe, 1991). Other channels present in the axon membrane are discussed by Vogel and Schwarz (see Chapter 13).

The Schwann cell plasma membrane related to the node gap and the MYSA segment may contain, in addition to acetylcholine receptors, Na+ and fast and slow K+ channels (Chiu, 1988; Chiu et al., 1984; Waxman and Ritchie, 1993). It has been speculated that some of the Na+ channels integrated in the nodal axolemma may have been synthesized in the Schwann cell and delivered to the nodal axolemma via the node gap microvilli (Gray and Ritchie, 1985).

The nodal axolemma is characterized, in addition to its numerous IMPs and thick axoplasmic cortex, by high Na,K-ATPase activity (Ariyasu and Ellisman, 1987; Ariyasu et al., 1985; Shinogami, 1989; Wood et al., 1977), α-bungarotoxin-binding sites (Freedman and Lentz, 1980), and numerous filipin-sterol complexes (Blanchard et al., 1985). The presence of Ca2+-activated ATPase (Mata and Fink, 1988) has been demonstrated in the axolemma of the MYSA segment and in the cytoplasm of the adjoining TCS. The axolemma of the whole CON segment is 5′-nucleotidase positive, binds ruthenium red and lectins, and becomes impregnated when treated with bismuth iodide (Dolapchieva and Ovtscharoff, 1985; Dolapchieva et al., 1986a, b, 1988, 1989).

The ratio between the Schwann cell membrane area and the area of the axolemma that faces the extracellular space (periaxonal and node gap spaces) outside the CON segment can be calculated to about 10:1 in axons with diameters greater than 3 μm (Berthold and Rydmark, 1983a). Elsewhere along the fiber this ratio is 1:1. The Schwann cell/axon membrane area ratio of the nodal region as plotted against axon diameter is shown in Figure 2-28. If a high ratio reflects the ability of the Schwann cell to control the ionic milieu outside the axolemma in a way favorable for impulse conduction, nerve fibers with the highest conduction stamina would appear in the middle of the caliber spectrum (d = 5 to 9 μm). This is the size of large γ and small α axons and type II afferents, all axons known to fire tonically (i.e., have high conduction stamina) (Berthold and Rydmark, 1983b). This and the suggested accumulation of K+ channels in the MYSA axolemma are similar to the purported role of Schwann cells as potassium-clearing units outside squid giant axons (Abbott et al., 1988; Brown and Abbott, 1993).

The diversity in the organization of the axoplasm in different CON segments of the same nerve and even in different CON segments of the same axon is a most remarkable feature of myelinated peripheral axons (Berthold, 1982; Berthold et al., 1993b; Raine et al., 1983). The axoplasmic patterns range from the comparatively trivial axoplasmic appearance met with in the STIN region (Figure 2-29a) to that of an axoplasm crammed with organelles and granular material (Figures 2-29b and 2-30). Retrogradely transported organelles—dense lamellar bodies and multivesicular bodies—occupy the axoplasm in the distal half of the CON segment and in the adjacent part of the paranodal main segment (Figure 2-31). Organelles transported anterogradely—VT profiles—occupy the proximal half of the CON segment but are also numerous in the nodal segment, and some appear in the distal MYSA segment (Figures 2-32 and 2-33). Such a proximodistal segregation of anterogradely and retrogradely transported elements can be understood in terms of a local reduction of the axonal stream. The organelle distribution corresponds to that noted at a blockage of axoplasmic transport (Ellisman and Lindsey; 1983; Smith, 1980; Tsukita and Ishikawa, 1980; see also Dahlström et al., 1992). The striking variations noted in the proximodistal segregation pattern, the presence of lysosomes, and the very low occurrence of dense lamellar bodies and multivesicular bodies outside the PNP regions (Table 2-3) (Berthold and Mellström, 1986; Berthold et al., 1986; Gatzinsky and Berthold, 1990; Gatzinsky et al., 1988), however, are difficult to understand as passive rheologic phenomena. One possible explanation could (p.40)

                      Morphology of normal peripheral axons                      Morphology of normal peripheral axons                      Morphology of normal peripheral axons

figs. 2-31 through 2-33. Details from different levels of the PNP-region of cat ventral root in Figure 2-30. D, distal; P, proximal.

FIG. 2-31. Transition between the main and MYSA segments of the distal paranode (× 30,000). Arrowheads, dense lamellar bodies; MT, bundle of microtubules; asterisks, strands of vesiculotubular profiles; m, mitochondrion. Inset, two multivesicular bodies (× 35,000).

FIG. 2-32. Proximal MYSA segment (× 30,000). The axoplasm is dominated by longitudinal strands of vesiculotubular profiles associated with microtubular bundles (MT).

FIG. 2-33. Proximal paranodal main segment (× 30,000). Arrows point to axoplasmic reticulum profiles; asterisks, vesiculotubular profiles; m, mitochondrion.

be that these variations indicate differences in the functional/metabolic state of a neuron in general and of a PNP region in particular. The various patterns may reflect different operative steps in a local interaction between, on the one hand, the axoplasm of a CON segment and its immediate paranodal extensions, including the axon-Schwann cell network, and, on the other hand, various transported elements. Recent observations indicate that the axoplasm of the CON segment interacts actively both with material transported retrogradely from (p.41)
                      Morphology of normal peripheral axons                      Morphology of normal peripheral axons                      Morphology of normal peripheral axons

figs. 2-34 through 2-36. Image-processed electron micrographs of cat medial gastrocnemic nerve; longitudinal sections through PNP regions of axons transporting intramuscularly injected HRP. Axoplasmic organelles that contain HRP appear as black profiles. Hyphens, the nodal mid-level; asterisks, node gaps; P, proximal; D, distal (closest to the muscle). Each picture is the result obtained when a stack of six consecutive and aligned electron microscopic images has been processed by an arithmetic minimum filter and the minima collected in a single plane. The operations were performed on a Teragon 4000 image analysis system at the MEDNET Laboratory.

fig. 2-34. PNP region collected approximately 1 cm proximal to muscle hilus 12 hours after HRP injection (× 8100.) HRP-containing bodies form a “plug” in the distal opening of the CON segment. We refer to this as PNP pattern type A.

fig. 2-35. PNP region collected at hip level 48 hours after injection (× 10,000). Light microscopically, this PNP region showed a distal plug-like accumulation of HRP activity and a proximal disk-like one in the proximal part of the CON segment. Electron microscopically, the disk corresponds to a large number of small HRP-positive bodies associated with strands of vesicotubular profiles. Very few HRP-positive bodies appear proximal to the CON segment. This is PNP pattern type B. Compare with the PNP region of Figure 2-29b.

fig. 2-36. PNP region collected close to muscle hilus 48 hours after injection (× 8100). In the light microscope, this PNP region showed a diffuse, nonsegregated staining of the CON segment and adjacent paranodal region. Electron microscopically, many HRP-positive bodies appeared proximal to the CON segment. This is PNP pattern type C.

                      Morphology of normal peripheral axons

fig. 2-37. Schematic, to-scale representation of two cross-sectioned nodal axon segments (cat α-axons). Microtubules are marked out as dots, a, Node containing a large number of vesiculotubular profiles (not shown); approximately 5% of the microtubules appear within 0.2 μm from the axolemma. b, Node containing only a few organelles; more than 30% of the microtubules occupy the zone just inside the axolemma.

the terminal region, such as intramuscularly injected horseradish peroxidase (HRP) (Berthold and Mellström, 1986; Berthold et al., 1986; Gatzinsky and Berthold, 1990), and with materials transported anterogradely from the soma, such as newly synthesized proteins (Armstrong et al., 1987; Janetzko et al., 1989; Price and Griffin, 1980; Zimmermann and Vogt, 1989). Our studies of PNP segments that take part in retrograde transport of intramuscularly injected HRP have shown organelle patterns similar to those of unexposed animals but with a specific temporal order of appearance after the injection (Figures 2-34 through 2-36). No doubt the CON segments are sites where elements of the retrograde endocytotic stream are forced into close contact with elements that belong to the anterograde exocytotic (lysosomal) stream (see Holtzman 1992; Huttner and Dotti, 1991; see also van Deurs et al., 1989). Extreme axoplasmic segregation appears in the PNP segments of the distal stump in the first days after a crush injury (Martinez and Friede, 1970; Webster, 1962).

In organelle-rich and segregated CON segments, microtubules form coherent bundles of up to 20 members. Such bundles support longitudinally running strands of organelles. Typically, the proximal end of a strand extends for a varying distance in the proximal paranodal main segment, where it is thin and consists of just one or a few longitudinal profiles of axoplasmic reticulum. Close to the opening of the CON segment, these profiles appear to be decorated with a distally increasing number of VT profiles. The strands become conspicuous as they enter and pass the proximal half of the CON segment. More distally, the strands are poor in VT profiles and consist, before they dissolve in the distal paranodal main segment of trains of mitochondria, dense lamellar bodies, and multivesicular bodies (Figures 2-29b and 2-30 through 2-33). In organelle-poor and vaguely segregated CON segments, the microtubule bundling is less distinct and the microtubules more dispersed. In some such CON segments, most microtubules are concentrated in a layer just inside the undercoating of the axolemma (compare Figure 2-37a and b). The distribution of microtubules in the axonal cross section will, in view of their fundamental role as transport rails, influence the transport capacity of the CON segment with regard both to amount and size of transported bodies. A mechanism that could rearrange the distribution of microtubules here would give a CON segment the combined properties of a variable sieve and a throttle valve. The recent claims that the Schwann cells may influence the axon, and in particular the neurofilaments and the microtubules of the PNP regions, are indeed noteworthy in connection with axonal transport (de Waegh and Brady, 1990, 1991; de Waegh et al., 1992; Nixon, 1993). The low phosphorylation degree of neurofilaments in the PNP regions (Mata et al., 1992) may explain the drastic local reduction in number, not in concentration, of neurofilaments in the PNP regions in general and in the CON segments in particular (Table 2-2) (Reles and Friede, 1991).

The node-paranode (NP) apparatus.

The term “paranodal apparatus,” as introduced by Williams and Landon (1963) over 30 years ago, included the Schwann cell mitochondrion bags, the node gap microvilli, and the node gap substance. The concept emphasizes the significance of the Schwann cell in nodal function, which at that time assumed that a fast K+ current repolarized the nodal axolemma during the action potential and accounted only marginally for axoplasmic transport. Current knowledge and ideas indicate that the concept of a paranodal apparatus should be expanded to include also the TCS of the Schwann cell, the whole CON segment, and the immediate adjoining parts of the proximal and distal paranodal main segments and their associated axon-Schwann cell networks. With this background, the notion of a NP apparatus could be more informative. Although there are many speculations (p.43) with regard to the precise functional significance of the NP apparatus and the ways its various components may cooperate (e.g., Ellisman et al., 1980; Muller-Mohnssen et al., 1974; Tippe and Muller-Mohnssen, 1975; Vascilescu and Filip, 1979; for reviews and references see Landon, 1981; Landon and Hall, 1976), little precise information is available about this. This is not surprising in view of the rich structure expressed by the PNP regions.

This work was supported by the Swedish Medical Research Council; Project No. 03157, Torsten och Ragnar Söderbergs Research Fund; and Ingabritt och Arne Lundbergs Research Fund.


Bibliography references:

ABBOTT, N. J., LIEBERMAN, E. M., PICHON, Y., HASSAN, S. AND LARMET, Y. 1988. Periaxonal K+ regulation in the small squid Alloteuthis: Studies on isolated and in situ axons. Biophys. J. 53:275–279.

AGUAYO, A. J. AND BRAY, G. M. 1975. Pathology and pathophysiology of unmyelinated nerve fibers. In: Peripheral Neuropathy, Vol. 1. P. J. DYCK, P. K. THOMAS, AND H. L. LAMBERT, eds. Philadelphia, W. B. Saunders Company, pp. 363–390.

ARBUTHNOTT, E. R., BOYD, I. A. AND KALU, K. U. 1980. Ultrastructural dimensions of myelinated peripheral nerve fibres in the cat and their relation to conduction velocity. J. Physiol. 308:125–157.

ARIYASU, R. G. AND ELLISMAN, M. H. 1987. The distribution of (Na+ and K+)ATPase is continuous along the axolemma of unensheathed axons from spinal roots of ‘dystrophic’ mice. J. Neurocytol. 16:239–248.

ARIYASU, R. G., NICHOL, J. A. AND ELLISMAN, M. H. 1985. Localization of sodium/potassium adenosine triphosphatase in multiple cell types of the murine nervous system with antibodies raised against the enzyme from kidney. J. Neurosci. 5:2581–2596.

ARMSTRONG, R., TOEWS, A. D. AND MORELL, P. 1987. Axonal transport through nodes of Ranvier. Brain Res. 412:196–199.

BAAS, P. W. AND BLACK, M. M. 1990. Individual microtubules in the axon consist of domains that differ in both composition and stability. J. Cell Biol. 111:495–509.

BAAS, P. W., SLAUGHTER, T., BROWN, A. AND BLACK, M. M. 1991. Microtubule dynamics in axons and dendrites. J. Neurosci. Res. 30:134–153.

BAINES, A. J. 1991. Neuronal plasma membrane associated cytoskeleton. In: The Neuronal Cytoskeleton. R. D. BURGOYNE, ed. New York, Wiley-Liss, Inc., pp. 161–183.

BAMBURG, J. R. 1988. The axonal cytoskeleton: Stationary or moving matrix? TINS 11:248–249.

BANNISTER, L. H. 1976. Sensory terminals of peripheral nerves. In: The Peripheral Nerve. D. N. Landon, ed. London, Chapman and Hall, pp. 396–454.

BERTHOLD, C-H. 1968. Ultrastructure of the node-paranode region of mature feline ventral spinal root nerve fibers. Acta Soc. Med. Upsal. 73(Suppl. 9):37–70.

BERTHOLD, C-H. 1974. A comparative morphological study of the developing node-paranode region in lumbar spinal roots. I. Electron-microscopy. Neurobiology 4:82–104.

BERTHOLD, C-H. 1978. Morphology of normal peripheral axons. In: Physiology and Pathobiology of Axons. S. G. WAXMAN, ed. New York, Raven Press, pp. 3–63.

BERTHOLD, C-H. 1982. Some aspects of the ultrastructural organization of peripheral myelinated axons in the cat. In: Axoplasmic Transport. D. G. WEISS, ed. Berlin, Springer-Verlag, pp. 40–54.

BERTHOLD, C-H., CARLSTEDT, T. AND CORNELIUSON, O. 1993a. The central-peripheral transition zone. In: Peripheral Neuropathy, 3rd ed., Vol. 1. P. J. DYCK, AND P. K. THOMAS, eds. Philadelphia, W. B. Saunders Company, pp. 73–80.

BERTHOLD, C-H., CORNELIUSON, O. AND MELLSTRÖM, A. 1986. Peroxidase activity at nodes of Ranvier in lumbosacral ventral spinal roots and in the PNS-CNS transitional region after intramuscular administration of horseradish peroxidase. J. Neurocytol. 15:253–260.

BERTHOLD, C-H., CORNELIUSON, O. AND RYDMARK, M. 1982. Changes in shape and size of cat spinal root myelinated nerve fibres during fixation and Vestopal-W embedding for electron microscopy. J. Ultrastruct. Res. 80:23–41.

BERTHOLD, C-H., FABRICIUS, C., RYDMARK, M. AND ANDERSEN, B. 1993b. Axoplasmic organelles at nodes of Ranvier. I. Distribution and occurrence in large myelinated spinal root axons of the adult cat. J. Neurocytol. 22:925–940.

BERTHOLD, C-H. AND MELLSTRÖM, A. 1986. Peroxidase activity at consecutive nodes of Ranvier in the nerve to the medial gastrocnemius muscle after intramuscular administration of horseradish peroxidase. Neuroscience 19:1349–1362.

BERTHOLD, C-H. AND NILSSON, I. 1987. Redistribution of Schwann cells in developing feline L7 ventral spinal roots. J. Neurocytol. 16:811–828.

BERTHOLD, C-H., NILSSON, I. AND RYDMARK, M. 1983. Axon diameter and myelin sheath thickness in nerve fibers of the ventral spinal root of the seventh lumbar nerve of the adult and developing cat. J. Anat. 136:483–508.

BERTHOLD, C-H. AND RYDMARK, M. 1983a. Anatomy of the paranode-node-paranode region in the cat. Experientia 39:964–976.

BERTHOLD, C-H. AND RYDMARK, M. 1983b. Electron microscopic serial section analysis of nodes of Ranvier in lumbosacral spinal roots of the cat: Ultrastructural organization of nodal compartments in fibres of different sizes. J. Neurocytol. 12:475–505.

BLACK, J. A., FOSTER, R. E. AND WAXMAN, S. G. 1982. Freeze-fracture ultrastructure of rat CNS and PNS nonmyelinated axolemma. J. Neurocytol. 10:981–993.

BLACK, M. M., CHESTNUT, M. H., PLEASURE, I. T. AND KEEN, J. H. 1991. Stable clathrin: Uncoating protein (hsc70) complexes in intact neurons and their axonal transport. J. Neurosci. 11:1163–1172.

BLANCHARD, C. E., MACKENZIE, M. L., SIKRI, K. AND ALLT, G. 1985. Filipin-sterol complexes at nodes of Ranvier. J. Neurocytol. 14:1053–1062.

BÖCK, P., HANAK, H. AND STOCKINGER, L. 1970. The penetration of exogenous protein into mesaxonal and periaxonal space: An electron microscopic study on unmyelinated nerves using horseradish peroxidase as a tracer. J. Submicrosc. Cytol. 2:189–192.

BODIAN, D. 1962. The generalized vertebrate neuron. Science 137:323–326.

BRAY, D. AND BUNGE, M. B. 1981. Serial analysis of microtubules in cultured rat sensory axons. J. Neurocytol. 10:589–605.

BRAY, J. J., FERNYHOUGH, P., BAMBURG, J. R. AND BRAY, D. 1992. Actin depolymerizing factor is a component of slow axonal transport. J. Neurochem. 58:2081–2087.

BREMNER, D. AND SMART, I. 1965. The position of Schwann cell nuclei in relation to the internodal mid-point of myelinated nerve fibres. J. Anat. 99:194–195.

BRISMAR, T. 1979. Potential clamp experiments on myelinated nerve fibres from alloxan diabetic rats. Acta Physiol. Scand. 105:384–386.

BRISMAR, T. 1980. Potential clamp analysis of the membrane currents in rat myelinated nerve fibres. J. Physiol. (Lond.) 298:171–184.

BROWN, E. R. AND ABBOTT, N. J. 1993. Ultrastructure and permeability of the Schwann cell layer surrounding the giant axon of the squid. J. Neurocytol. 22:283–298.

(p.44) BUCHHEIT, T. E. AND TYTELL, M. 1992. Transfer of molecules from glia to axon in the squid may be mediated by glial vesicles. J. Neurobiol. 23:217–230.

BUNGE, M. B. 1986. The axonal cytoskeleton: Its role in generating and maintaining cell form. TINS 9:477–481.

BURGOYNE, R. D. 1991. Cytoskeleton is a major neuronal organelle. In: The Neuronal Cytoskeleton. R. D. BURGOYNE, ed. New York, Wiley-Liss, Inc., pp. 1–3.

CALDERON, R. O., MAGGIO, B., NEUBERGER, T. J. AND DEVRIES, G. H. 1993. Surface behavior of axolemma monolayers: Physicochemical characterization and use as supported planar membranes for cultured Schwann cells. J. Neurosci. Res. 34:206–218.

CHIU, S. Y. 1988. Changes in excitable membrane properties in Schwann cells of adult rabbit sciatic nerves following nerve transection. J. Physiol. 396:173–178.

CHIU, S. Y., SHRAGER, P. AND RITCHIE, J. M. 1984. Neuronal-type Na+ and K+ channels in rabbit cultured Schwann cells. Nature 311:156–157.

COGGESHALL, R. E., COULTER, J. D. AND WILLIS, W. D. JR. 1974. Unmyelinated axons in the ventral roots of the cat lumbo-sacral enlargement. J. Comp. Neurol. 153:39–58.

COGGESHALL, R. E., HANCOCK, M. B. AND APPLEBAUM, M. L. 1976. Categories of axons in mammalian rami communicantes. J. Comp. Neurol. 167:105–124.

DAHLSTRÖM, A. B., CZERNIK, A. J. AND LI, J.-Y. 1992. Organelles in fast axonal transport. Mol. Neurobiol. 6:157–177.

DEVRIES, G. H. 1984. The axonal plasma membrane. In: Handbook of Neurochemistry. A. LAJTHA, ed. New York, Plenum Press, pp. 341–360.

DEWAEGH, S. M. AND BRADY, S. T. 1990. Altered slow axonal transport and regeneration in a myelin-deficient mutant mouse: The trembler as an in vivo mode for Schwann cell-axon interactions. J. Neurosci. 10:1855–1865.

DEWAEGH, S. M. AND BRADY, S. T. 1991. Local control of axonal properties by Schwann cells: Neurofilaments and axonal transport in homologous and heterologous nerve grafts. J. Neurosci. Res. 30:201–212.

DEWAEGH, S. M., LEE, V. M.-Y. AND BRADY, S. T. 1992. Local modulation of neurofilament phosphorylation, axonal caliber and slow axonal transport by myelinating Schwann cells. Cell 68:451–463.

DOLAPCHIEVA, S., ICHEV, K. AND OVTSCHAROFF, W. 1988. Ultrastructural cytochemical localization of 5′-nucleotidase activity in axon-myelin-Schwann cell complex. Acta Histochem. 83:125–135.

DOLAPCHIEVA, S., ICHEV, K. AND OVTSCHAROFF, W. 1989. Cytochemical localization of ATPase in axon-myelin-Schwann cell complex. Z. Mikrosk. Anat. Forsch. 103:151–165.

DOLAPCHIEVA, S. AND OVTSCHAROFF, W. 1985. Bismuth iodide impregnation of the peripheral nerve fibers. Z. Mikrosk. Anat. Forsch. 99:804–812.

DOLAPCHIEVA, S., OVTSCHAROFF, F. AND ICHEV, K. 1986a. Lectin binding sites in axon-myelin-Schwann cell complex. Acta Histochem. Cytochem. 19:253–261.

DOLAPCHIEVA, S., OVTSCHAROFF, W. AND ICHEV, K. 1986b. Localizations of ruthenium red positive material in rabbit peripheral nerves. Acta Histochem. 78:19–30.

DROZ, B. 1976. Synthetic machinery and axoplasmic transport: Maintenance of neuronal connectivity. In: The Nervous System, Vol. 1. B. D. Tower, ed. New York, Raven Press, pp. 111–127.

DROZ, B., RAMBOURG, A. AND KOENIG, H. L. 1975. The smooth endoplasmic reticulum: Structure and role in the renewal of axonal membrane and synaptic vesicles by fast axonal transport. Brain Res. 60:93–127.

DUCE, I. R. AND KEEN, P. 1978. Can neuronal smooth endoplasmic reticulum function as a calcium reservoir? Neuroscience 3:837–848.

DUMONT, J. N., ANDERSON, E. AND CHOMYN, E. 1965. The anatomy of the peripheral nerve and its ensheathing artery in the horseshoe crab, Xiphosura (Limulus) polyphemus. J. Ultrastruct. Res. 13:38–64.

EAMES, R. A. AND GAMBLE, H. J. 1970. Schwann cell relationship in normal human cutaneous nerves. J. Anat. 106:417–435.

ECCLES, J. C. AND SHERRINGTON, S. C. 1930. Number and contraction values of individual motor-units examined in some muscles of the limb. Proc. R. Soc. B 106:326–357.

ELFVIN, L. G. 1963. The ultrastructure of the plasma membrane and myelin sheath of peripheral nerve fibers after fixation by freeze-drying. J. Ultrastruct. Res. 8:283–304.

ELLISMAN, M. H., FRIEDMAN, P. L. AND HAMILTON, W. J. 1980. The localization of sodium and calcium to Schwann cell paranodal loops at nodes of Ranvier and of calcium to compact myelin. J. Neurocytol. 9:185–205.

ELLISMAN, M. H. AND LINDSEY, J. D. 1983. The axoplasmic reticulum within myelinated axons is not transported rapidly. J. Neurocytol. 12:393–411.

ELLISMAN, M. H. AND PORTER, K. R. 1980. Microtrabecular structure of the axoplasmic matrix: Visualization of cross-linking structures and their distribution. J. Cell Biol. 87:464–479.

ELLISMAN, M. H., WILEY, C. A., LINDSAY, J. D. AND WURTZ, C. C. 1984. Structure and function of the cytoskeleton and endomembrane systems at the node of Ranvier. In: The Node of Ranvier. J. C. ZAGOREN AND S. FEDOROFF, eds. New York, Academic Press, pp. 153–181.

EMERGY, D. G., FOREMAN, R. D. AND COGGESHALL, R. E. 1976. Fiber analysis of the feline inferior cardiac sympathetic nerve. J. Comp. Neurol. 166:457–468.

EVANS, P. D., REALE, V., MERZON, R. M. AND VILLEGAS, J. 1992a. The effect of a glutamate uptake inhibitor on axon-Schwann cell signalling in the squid giant nerve fibre. J. Exp. Biol. 173:251–260.

EVANS, P. D., REALE, V., MERZON, R. M. AND VILLEGAS, J. 1992b. N-Methyl-d-aspartate (NMDA) and non-NMDA (metabotropic) type glutamate receptors modulate the membrane potential of the Schwann cell of the squid giant nerve fibre. J. Exp. Biol. 173:229–249.

EWART, D. P., KUZON, W. M. JR., FISH, J. S. AND MCKEE, N. H. 1989. Nerve fiber morphometry: A comparison of techniques. J. Neurosci. Meth. 28:143–150.

FATH, K. R. AND LASEK, R. J. 1988. Two classes of actin microfilaments are associated with the inner cytoskeleton of axons. J. Cell Biol. 107:613–621.

FORMAN, D. S. 1987. Axonal transport of mitochondria. In: Axonal Transport. R. S. SMITH AND M. A. BISBY, eds. New York, Alan R. Liss, Inc., pp. 155–163.

FRAHER, J. P. 1992. The CNS-PNS transitional zone of the rat: Morphometric studies of cranial and spinal levels. Prog. Neurobiol. 38:216–316.

FREEDMAN, S. D. AND LENTZ, T. L. 1980. Binding of horseradish peroxidase-alfa-bungarotoxin to axonal membranes at the node of Ranvier. J. Comp. Neurol. 193:179–185.

FRIEDE, R. L. AND BEUCHE, W. 1985. A new approach toward analyzing peripheral nerve fiber populations. I. Variance in sheath thickness corresponds to different geometric proportions of the internode. J. Neuropathol. Exp. Neurol. 44:60–72.

FRIEDE, R. L. AND BISCHHAUSEN, R. 1982. How are sheath dimensions affected by axon clibre and internodal length? Brain Res. 235:335–350.

FRIEDE, R. L., MEIER, T. AND DIEM, M. 1981. How is the exact length of an internode determined? J. Neurol. Sci. 50:217–228.

GABRIEL, G., THOMAS, P. K., KING, R. H. M., STOLINSKI, C. AND BREATHNACH, A. S. 1986. Freeze-tracture observations on human peripheral nerve. J. Anat. 146:153–166.

(p.45) GATZINSKY, K. P. AND BERTHOLD, C-H. 1990. Lysosomal activity at nodes of Ranvier during retrograde axonal transport of horseradish peroxidase in alpha-motor neurons of the cat. J. Neurocytol. 19:989–1002.

GATZINSKY, K. P., BERTHOLD, C-H. AND CORNELIUSON, O. 1988. Acid phosphatase activity at nodes of Ranvier in alpha-motor and dorsal root ganglion neurons of the cat. J. Neurocytol. 17:531–544.

GERHARDT, G. 1991. Are Ranvier’s nodes artifacts by preparation? Acta Anat. 141:132–138.

GILLIATT, R. W. 1966. Axon branching in motor nerves. In: Control and Innervation of Skeletal Muscle. D. B. TOWER, ed. New York, Raven Press, pp. 53–60.

GORIO, A. 1984. Sprouting and regeneration of peripheral nerve. In: The Node of Ranvier. J. C. ZAGOREN AND S. FEDOROFF, eds. New York, Academic Press, pp. 353–388.

GOULD, R. M., JESSEN, K. R., MIRSKY, R. AND TENNEKOON, G. 1992. The cell of Schwann: An update. In: Myelin: Biology and Chemistry. R. E. MARTENSEN, ed. Boca Raton, FL, CRC Press, pp. 123–171.

GRAFSTEIN, B. AND FORMAN, D. S. 1980. Intracellular transport in neurons. Physiol. Rev. 4:1167–1283.

GRAY, P. T. A. AND RITCHIE, J. M. 1985. Ion channels in Schwann and glial cells. TINS 8:411.

GREENBERG, M. M., LEITAO, C., TROGADIS, J. AND STEVENS, J. K. 1990. Irregular geometries in normal unmyelinated axons: A 3D serial EM analysis. J. Neurocytol. 20:978–988.

GRIFFIN, J. W., ROSENFELDT, J., HOFFMAN, P. N., GOLD, B. G. AND TRAPP, B. D. 1988. The axonal cytoskeleton: Influences on nerve fiber form and Schwann cell behavior. In: Intrinsic Determinants of Neuronal Form and Functions. R. J. LASER AND M. M. BLACK, eds. New York, Alan R. Liss, Inc., pp. 403–439.

HAAK, R. A., KLEINHAUS, F. W. AND OCHS, S. 1976. The viscosity of mammalian axoplasm measured by electron spin resonance. J. Physiol. (Lond.) 263:115–137.

HALL, S. M. AND WILLIAMS, P. L. 1971. The distribution of electrondense tracers in peripheral nerve fibres. J. Cell Sci. 8:541–555.

HALTER, J. A. AND CLARK, J. W. JR. 1993. The influence of nodal constriction on conduction velocity in myelinated nerve fibers. NeuroReport 4:89–92.

HAMMERSCHLAG, R. AND STONE, G. C. 1987. Further studies on the initiation of fast axonal transport. In: Axonal Transport. R. S. SMITH AND M. A. BISBY, eds. New York, Alan R. Liss, Inc., pp. 37–51.

HERIOT, K., GAMBETTI, P. AND LASEK, R. 1985. Proteins transported in slow components a and b of axonal transport are distributed differently in the transverse plane of the axon. J. Cell Biol. 100:1167–1172.

HILDEBRAND, C., MUSTAFA, G. Y., BOWE, C. AND KOCSIS, J. D. 1987. Nodal spacing along regenerated axons following a crush lesion of the developing rat sciatic nerve. Dev. Brain Res. 32:147–154.

HILDEBRAND, C., REMAHL, S., PERSSON, H. AND BJARTMAR, C. 1992. Myelinated nerve fibers in the CNS. Prog. Neurobiol. 44:319–384.

HIRANO, A. 1984. Nodes of Ranvier in pathological conditions. In: The Node of Ranvier. J. C. ZAGOREN AND S. FEDOROFF, eds. New York, Academic Press, pp. 213–243.

HIRANO, A. AND DEMBITZER, H. A. 1982. Further studies on the transverse bands. J. Neurocytol. 11:861–866.

HIROKAWA, N. 1982. Cross-linker system between neurofilaments, microtubules, and membranous organelles in frog axons revealed by the quick-freeze, deep-etching method. J. Cell Biol. 94:129–142.

HIROKAWA, N. 1991. Molecular architecture and dynamics of the neuronal cytoskeleton. In: The Neuronal Cytoskeleton. R. D. BURGOYNE, ed. New York, Wiley-Liss, Inc., pp. 5–74.

HIROKAWA, N., BLOOM, G. S. AND VALLEE, R. B. 1985. Cytoskeletal architecture and immunocytochemical localization of microtubule-associated proteins in regions of axons associated with rapid axonal transport: The beta, beta′-iminodipropionitrile-intoxicated axon as a model system. J. Cell Biol. 101:227–239.

HIROKAWA, N., GLICKSMAN, M. A. AND WILLARD, M. B. 1984. Organization of mammalian neurofilament polypeptides within the neuronal cytoskeleton. J. Cell Biol. 98:1523–1536.

HOFFMAN, P. V. AND GRIFFIN, J. W. 1993. The control of axonal calibre. In: Peripheral Neuropathy, 3rd ed., Vol. 1. P. J. DYCK AND P. K. THOMAS, eds. Philadelphia, W. B. Saunders Company, pp. 73–80.

HOFFMAN, P. V., GRIFFIN, J. W., GOLD, B. G. AND PRICE, D. L. 1985. Slowing of neurofilament transport and the radial growth of developing nerve fibres. J. Neurosci. 5:2920–2929.

HOLLENBECK, P.J. 1989. The transport and assembly of the axonal cytoskeleton. J. Cell Biol. 108:223–227.

HOLLENBECK, P. J. 1993. Products of endocytosis and autophagi are retrieved from axons by regulated retrograde organelle transport. J. Cell Biol. 121:305–315.

HOLTZMAN, E. 1992. Membrane trafficking in neurons. Curr. Opin. Neurobiol. 2:607–612.

HOPKINS, W. G. AND BROWN, M. C. 1982. The distribution of nodal sprouts in a paralysed or partly denervated mouse muscle. Neuroscience 7:37–44.

HUTTNER, W. B. AND DOTTI, C. G. 1991. Exocytotic and endocytotic membrane traffic in neurons. Curr. Opin. Neurobiol. 1:388–392.

JACOBS, J. M. 1984. Toxic effects on the node of Ranvier. In: The Node of Ranvier. J. C. ZAGOREN AND S. FEDOROFF, eds. New York, Academic Press, pp. 245–272.

JANETZKO, A., ZIMMERMANN, H. AND VOLKNANDT, W. 1989. Intraneuronal distribution of a synaptic vesicle membrane protein: Antibody binding sites at axonal membrane compartments and trans-Golgi netvork and accumulation at nodes of Ranvier. Neuroscience 32:65–77.

JOE, E. AND ANGELIDES, K. 1992. Clustering of voltage-dependent sodium channels on axons depend on Schwann cell contact. Nature 356:333–335.

KENNY, A. J. AND BOOTH, A. G. 1978. Microvilli: Their Ultrastructure, enzymology and molecular organization. Essays Biochem. 14:1–44.

KIDD, G. J. AND HEATH, J. W. 1988a. Double myelination of axons in the sympathetic nervous system of the mouse. I. Ultrastructural features and distribution. J. Neurocytol. 17:245–261.

KIDD, G.J. AND HEATH, J. W. 1988b. Double myelination of axons in the sympathetic nervous system of the mouse. II. Mechanisms of formation. J. Neurocytol. 17:263–276.

KIDD, G. J., HEATH, J. W., TRAPP, B. D. AND DUNKLEY, P. R. 1992. Myelin sheath survival after guanethidine-induced axonal degeneration. J. Cell Biol. 116:395–403.

KOCSIS, J. D. 1984. Functional organization of potassium channels in normal and pathological mammalian axons. In: The Node of Ranvier. J. C. ZAGOREN AND S. FEDOROFF, eds. New York, Academic Press, pp. 183–212.

KOPPENHÖFER, E., SOMMER, R.-G. AND BOHUSLAVIZKI, K. H. 1992. Die Erforschung der Funktion peripherer Nerven. TW Neurol. Psychiatr. 6:25–39.

KORDELI, E., DAVIS, J., TRAPP, B. AND BENNETT, V. 1990. An isoform of ankyrin is localized at nodes of Ranvier in myelinated axons of central and peripheral nerves. J. Cell Biol. 110:1341–1352.

KOSIK, K. S. 1993. The molecular and cellular biology of Tau. Brain Pathol. 3:39–43.

KRAMMER, E. B. AND LISCHKA, M. F. 1973. Schwermetallaffine Strukturen des peripheren Nerven. I. Potentieller Stürfaktor beim cytochemischen AChE-Nachweis. Histochemie 36:269–282.

LANDON, D. N. 1981. Structure of normal peripheral myelinated nerve fibres. In: Demyelinating Disease: Basic and Clinical Electrophysiology. S. G. WAXMAN AND J. M. RITCHIE, eds. New York, Raven Press, pp. 25–49.

(p.46) LANDON, D. N. AND HALL, S. 1976. The myelinated neve fibre. In: The Peripheral Nerve. D. N. London, ed. London, Chapman and Hall, pp. 1–105.

LANDON, D. N. AND LANGLEY, O. K. 1971. The local chemical environment of nodes of Ranvier: A study of cation binding. J. Anat. (Lond.) 108:419–432.

LANDON, D. N. AND WILLIAMS, P. L. 1963. Ultrastructure of the node of Ranvier. Nature 199:575–577.

LANGFORD, G. M., ALLEN, R. D. AND WEISS, D. G. 1987. Substructure of sidearms of squid axoplasmic vesicles and microtubules visualized by negative contrast electron microscopy. Cell Motil. cytoskel. 7:20–30.

LANGFORD, L. A. AND COGGESHALL, R. E. 1979. Branching of sensory axons in the dorsal root and evidence for the absence of dorsal root efferent fibers. J. Comp. Neurol. 184:193–204

LANGLEY, O. K. 1971. A comparison of the binding of Alcian blue and inorganic cations to polyanions in peripheral nerve. Histochem. J. 3:251–260.

LANGLEY, O. K. AND LANDON, D. N. 1969. Copper binding at nodes of Ranvier: A new electron histochemical technique for the demonstration of polyanions. J. Histochem. Cytochem. 17:66–69.

LASEK, R. J. 1982. Translocation of the neuronal cytoskeleton and axonal locomotion. Phil. Trans. R. Soc. Lond. B 299:313–327.

LASEK, R. J., GARNER, J. A. AND BRADY, S. T. 1984. Axonal transport of the cytoplasmic matrix. J. Cell Biol. 99:212–221.

LE BEAU, J. M., POWELL, H. C. AND ELLISMAN, M. H. 1987. Node of Ranvier formation along fibres regenerating through silicone tube implants: A freeze-fracture and thin-section electron microscopic study. J. Neurocytol. 16:347–358.

LEDEEN, R. W., GOLLY, F. AND HALEY, J. E. 1992. Axon-myelin transfer of phospholipids and phospholipid precursors. Mol. Neurobiol. 6:179–190.

LEHMAN, H. J. 1959. Die Nervenfaser. In: Handbuch der mikroskopischen Anatomie des Menschen, Vol. 4. V. VON MÖLLENDORFF, ed. Berlin, Springer-Verlag, pp. 515–701.

LEVINE, J. AND WILLARD, M. 1981. Axonally transported polypeptides associated with the internal periphery of many cells. J. Cell Biol. 90:631–643.

LEWIS, S. A., IVANOV, I. E., LEE, G-H. AND COWAN, N. J. 1989. Organization of microtubules in dendrites and axons is determined by a short hydrophobic zipper in microtubule-associated proteins MAP2 and tau. Nature 342:498–505.

LIEBERMAN, E. M., ABBOTT, N. J. AND HASSAN, S. 1989. Evidence that glutamate mediates axon-to-Schwann cell signaling in the squid. Glia 2:94–102.

LINDSEY, J. D. AND ELLISMAN, M. H. 1985. The neuronal endomembrane system. III. The origins of the axoplasmic reticulum and discrete axonal cisternae at the axon hillock. J. Neurosci. 5:3135–3144.

LIVINGSTON, R. B., PFENNINGER, K., MOOR, H. AND AKERT, K. 1973. Specialized paranodal and interparanodal glial-axonal junctions in the peripheral and central nervous system: A freeze-etching study. Brain Res. 58:1–24.

LOPACHIN, R. L., CASTIGLIA, C. M., LEHNING, E. AND SAUBERMANN, A. J. 1993. Effects of acrylamide on subcellular distribution of elements in rat sciatic nerve myelinated axons and Schwann cells. Brain Res. 608:238–246.

LUBINSKA, L. AND LUKASZEWSKA, I. 1956. Shape of myelinated nerve fibres and proximodistal flow of axoplasm. Acta Biol. Exp. XVII:115–133.

MACKENZIE, M. L., GHABRIEL, M. N. AND ALLT, G. 1984. Nodes of Ranvier and Schmidt-Lanterman incisures: An in vivo lanthanum tracer study. J. Neurocytol. 13:1043–1055.

MALHOTRA, S. K. AND VAN HARREVELD, A. 1965. Dorsal roots of the rabbit investigated by freeze-substitution. Anat. Rec. 152:182–192.

MATA, M. AND FINK, D. J. 1988. Cytochemical localization of calcium-ATPase activity in peripheral nerve. Brain Res. 445:47–54.

MATA, M., KUPINA, N. AND FINK, D. J. 1992. Phosphorylation dependent neurofilament epitopes are reduced at the node of Ranvier. J. Neurocytol. 21:199–210.

MARTINEZ, A. J. AND FRIEDE, R. L. 1970. Accumulation of axoplasmic organelles in swollen nerve fibers. Brain Res. 19:183–198.

MATTHEWS, M. A. 1968. An electron microscopic study of the relationship between axon diameter and the initiation of myelin production in the peripheral nervous system. Anat. Rec. 161:337–352.

MCDONALD, K. 1984. Osmium ferricyanide fixation improves microfilament preservation and membrane visualization in a variety of animal cell types. J. Ultrastruct. Res. 86:107–118.

MCQUARRIE, I. G. 1985. Effect of a conditioning lesion on axonal sprout formation at nodes of Ranvier. J. Comp. Neurol. 231:239–249.

MELLER, K. 1987. Early structural changes in the axoplasmic cytoskeleton after axotomy studied by cryofixation. Cell Tissue Res. 250:663–672.

MILLER, R. G. AND DA SILVA, P. P. 1977. Particle rosettes in the periaxonal Schwann cell membrane and particle clusters in the axolemma of rat sciatic nerve. Brain Res. 130:135–141.

MOHAMMED, U. H. M. AND LANDON, D. N. 1983. Axoplasmic asymmetry at the node of Ranvier. J. Anat. (Lond.) 137:D21.

MOHAMMED, U. H. M., LANDON, D. N. AND LOVE, S. 1983. Ultrastructural morphometry of frog nodes of Ranvier. J. Anat. (Lond.) 137:D22.

MORAN, O. AND MATEU, L. 1983. Loosening of paranodal myelin by repetitive propagation of action potentials. Nature 304:344–345.

MULLER-MOHNSSEN, H., TIPPE, A., HILLENKAMP, F. AND UNSOLD, E. 1974. Is the rise of the action potential at the Ranvier node controlled by a paranodal organ? Naturwissenshaften 61:369–370.

NAGEOTTE, J. 1911. Betrachtung uber den tatsachlichen Bau und die kunstlisch hervorgeruften Deformationen der markhaltigen Nervenfaser. Arch. Mikrosk. Anat. 77:245–279.

NEUBERGER, T. J. AND DEVRIES, G. H. 1992. Axonal contact as a determinant of oligodendrocyte and Schwann cell function. In: Myelin: Biology and Chemistry. R. E. MARTENSON, ed. Boca Raton, FL, CRC Press, pp. 173–193.

NILSSON, I. AND BERTHOLD, C-H. 1988. Axon classes and internodal growth in the ventral spinal root L7 of adult and developing cats. J. Anat. 156:71–96.

NIXON, R. A. 1991. Slow axonal transport. Curr. Opin. Cell Biol. 4:8–14.

NIXON, R. A. 1993. The regulation of neurofilament protein dynamics by phosphorylation: Clues to neurofibrillary pathobiology. Brain Pathol. 3:29–38.

NIXON, R. A. AND SIHAG, R. K. 1991. Neurofilament phosphorylation: A new look at regulation and function. TINS 14:501–506.

OBLINGER, M. M., BRADY, S. T., MCQUARRIE, I. G. AND LASEK, R. J. 1987. Cytotypic differences in the protein composition of the axonally transported cytoskeleton in mammalian neurons. J. Neurosci. 7:453–462.

OCHOA, J. 1976. The unmyelinated nerve fibre. In: The Peripheral Nerve. D. N. LANDON, ed. London, Chapman and Hall, pp. 106–158.

OCHS, S. and BRIMIJOIN, W. S. 1993. Axonal transport. In: Peripheral Neuropathy, 3rd ed., Vol. 1. P. J. DYCK AND P. K. THOMAS, eds. Philadelphia, W. B. Saunders Company, pp. 331–360.

OCHS, S., JERSILD, R. A. JR. AND LI, J-M. 1989. Slow transport of freely movable cytoskeletal components shown by beading partition of nerve fibers in the cat. Neuroscience 33:421–430.

OKABE, S. AND HIROKAWA, N. 1988. Microtubule dynamics in nerve cells: Analysis using microinjection of biotinylated tubulin into PC12 cells. J. Cell Biol. 107:651–664.

(p.47) OKABE, S. AND HIROKAWA, N. 1990. Turnover of fluorescently labelled tubulin and actin in the axon. Nature 343:279–282.

OKABE, S. AND HIROKAWA, N. 1993. Do photobleached fluorescent microtubules move?: Re-evaluation of fluorescence laser photobleaching both in vitro and in growing Xenopus axon. J. Cell Biol. 120:1177–1186.

OKAMURA, Y. AND TSUKITA, S. 1986. Morphology of freeze-substituted myelinated axon in mouse peripheral nerves. Brain Res. 383:146–158.

PANNESE, E., PROCACCI, P., LEDDA, M., ARCIDIACONO, G. AND RIGAMONTI, L. 1984. A quantitative study of microtubules in motor and sensory axons. Acta Anat. 118:193–200.

PANNESE, E., PROCACCI, P., LEDDA, M., ARCIDIACONO, G. AND RIGAMONTI, L. 1989. Internodal microvilli of Schwann cells of myelinated fibers in lizard spinal roots project onto unmyelinated axons. J. Neurocytol. 18:295–302.

PFEIFFER, G. AND FRIEDE, R. L. 1985. The axon tree of rat motor fibers: Morphometry and fine structure. J. Neurocytol. 14:809–824.

PHILLIPS, D. D., HIBBS, R. G., ELLISON, J. P. AND SHAPIRO, H. 1972. An electron microscopic study of central and peripheral nodes of Ranvier. J. Anat. 111:229–238.

PRICE, D. L. AND GRIFFIN, J. W. 1980. Neurons and ensheathing cells as targets of disease processes. In: Experimental and Clinical Neurotoxtcology. P. S. SPENCER AND H. H. SCHAUMBURG, eds., Baltimore, Williams & Wilkins, pp. 1–48.

PODUSLO, L. F. 1993. Regulation of myelin gene expression in the peripheral nervous system. In: Peripheral Neuropathy, 3rd ed., Vol. 1. P. J. DYCK AND P. K. THOMAS, eds. Philadelphia, W. B. Saunders Company, pp. 282–289.

QUATACKER, J. 1981. The axonal reticulum in the neurons of the superior cervical ganglion of the rat as a direct extension of the Golgi apparatus. Histochem. J. 13:109–124.

QUICK, D. C. AND WAXMAN, S. G. 1977. Specific staining of the axon membrane at nodes of Ranvier with ferric ion and ferrocyanide. J. Neurol. Sci. 31:1–12.

QUILLIAM, T. A. 1956. Some characteristics of myelinated fibre populations. J. Anat. 90:172–187.

RAINE, C. S. 1982. Differences between the nodes of Ranvier of large and small diameter fibres in the P.N. S. J. Neurocytol. 11:935–947.

RAINE, C. S., FINCH, H. AND MASONE, A. 1983. Axoplasmic asymmetry at the node of Ranvier. J. Neurocytol. 12:533–536.

RAMBOURG, A. AND DROZ, B. 1980. Smooth endoplasmic reticulum and axonal transport. J. Neurochem. 35:16–25.

RAYNER, D. A. AND BAINES, A. J. 1989. A novel component of the axonal cortical cytoskeleton, A60, defined by a monoclonal antibody. J. Cell Sci. 94:489–500.

RELES, A. AND FRIEDE, R. L. 1991. Axonal cytoskeleton at the nodes of Ranvier. J. Neurocytol. 20:450–458.

RIEGER, F., DANILOFF, J. K., PINCON-RAYMOND, M., CROSSIN, K. L, GRUMET, M. AND EDELMAN, G. M. 1986. Neuronal cell adhesion molecules and cytotactin are colocahzed at the node of Ranvier. J. Cell Biol. 103:379–391.

RISLING, M. AND HILDEBRAND, C. 1982. Occurrence of unmyelinated axon profiles at distal middle and proximal levels in the ventral root L7 of cats and kittens. J. Neurol. Sci. 56:219–231.

ROGART, R. 1984. Structural and functional relationships of ion conduction in the myelinated and demyelinated axon. In: The Node of Ranvier. J. C. ZAGOREN AND S. FEDOROFF, eds. New York, Academic Press, pp. 109–152.

ROSENBLUTH, J. 1984. Membrane specializations at the nodes of Ranvier and paranodal and juxtaparanodal regions of myelinated central and peripheral nerve fibers. In: The Node of Ranvier. J. C. ZAGOREN AND S. FEDOROFF, eds. New York, Academic Press, pp. 31–67.

RYDMARK, M., 1981. Nodal axon diameter correlates linearily with internodal axon diameter in spinal roots of the cat. Neurosci. Lett. 24:247–250.

RYDMARK, M. AND BERTHOLD, C-H. 1983. Electron microscopic serial section analysis of nodes of Ranvier in lumbar spinal roots of the cat: A morphometric study of nodal compartments in fibres of different sizes. J. Neurocytol. 12:537–565.

SAITUA, F. AND ALVAREZ, J. 1989. Microtubular packing varies along the course of motor and sensory axons: Possible regulation of microtubules by environmental cues. Neurosci. Lett. 104:249–252.

SCHNAPP, B. AND MUGNAINI, E. 1978. Membrane architecture of myelinated fibers as seen by freeze-fracture. In Physiology and Pathobiology of Axons. S. G. WAXMAN, ed. New York, Raven Press, pp. 83–123.

SCHNAPP, B. J. AND REESE, T. S. 1982. Cytoplasmic structure in rapid-frozen axons. J. Cell. Biol. 94:667–679.

SHEETZ, M. P. AND MARTENSON, C. H. 1991. Axonal transport: Beyond kinesin and cytoplasmic dynein. Curr. Opin. Neurobiol. 1:393–398.

SHINOGAMI, H. 1989. Cytochemical localization of K+-NPPase in the heminode of Ranvier in the rat optic nerve. Acta Histochem. Cytochem. 22:207–213.

SINGER, M. KRISHNAN, N. AND FYFE, D. A. 1972. Penetration of ruthenium red into peripheral nerve fibers. Anat. Rec. 173:375–390.

SMITH, R. S. 1980. The short term accumulation of axonally transported organelles in the region of localized lesions of single myelinated axons. J. Neurocytol. 9:39–65.

SMITH, R. S. 1989. Real-time imaging of axonally transported sub-resolution organelles in vertebrate myelinated axons. J. Neurosci. Meth. 26:203–209.

SMITH, R. S. AND FORMAN, D. S. 1988. Organelle dynamics in lobster axons: Anterograde and retrograde particulate organelles. Brain Res. 446:26–36.

SNYDER, R. E. 1989. Loss of material from the retrograde axonal transport system in frog sciatic nerve. J. Neurobiol. 20:81–94.

SPACEK, J. AND LIEBERMAN, A. R. 1980. Relationships between mitochondrial outer membranes and agranular reticulum in nervous tissue: Ultrastructural observations and a new interpretation. J. Cell Sci. 46:129–147.

SPENCER, P. S., RAINE, C. S. AND WISNIEWSKI, H. 1973. Axon diameter and myelin thickness—unusual relationships in dorsal root ganglia. Anat. Rec. 176:225–244.

SPENCER, P. S. AND THOMAS, P. K. 1974. Ultrastructural studies of the dying-back process II. The sequestration and removal by Schwann cells and oligodendrocytes of organelles from normal and diseased axons. J. Neurocytol. 3:763–783.

STELZNER, D. J. 1971. The relationship between synaptic vesicles, Golgi apparatus and smooth endoplasmic reticulum: A developmental study using the zinc iodide-osmium technique. Z. Zellforsch. 120:332–345.

STOLINSKY, C. AND BREATHNACH, A. S. 1982. Freeze-fracture replication of mammalian peripheral nerve—a review. Neurol. Sci. 57:1–28.

STOLINSKY, C., BREATHNACH, A. S., MARTIN, B. THOMAS, P. K., KING, R. H. M. AND GABRIEL, G. 1981. Associated particle aggregates in juxtaparanodal axolemma and adaxonal Schwann cell membrane of rat peripheral nerve. J. Neurocytol. 10:679–691.

STONE, G. C. AND HAMMERSCHLAG, R. 1987. Molecular mechanisms involved in sorting of fast-transported proteins. In: Axonal Transport. R. S. SMITH AND M. A. BISBY, eds. New York, Alan R. Liss, Inc., pp. 15–36.

STRUPP, M. AND GRAFE, P. 1991. A chloride channel in rat and human axons. Neurosci. Lett. 133:237–240.

SUI, S., CHRISTENSEN, E. K. AND NIELSEN, S. 1991. Membrane traffic after inhibition of endocytosis in renal proximal tubules. J. Struct. Biol. 107:201–210.

(p.48) SUNDERLAND, S. 1968. Nerves and Nerve Injuries. Edinburgh, Livingstone.

SUTER, U., WELCHER, A. W. AND SNIPES, G. J. 1993. Progress in the molecular understanding of hereditary peripheral neuropathies reveals new insights in the biology the peripheral nervous system TINS 16:50–56.

TAKENAKA, T., KAWAKAMI, T., HIKAWA, N. AND GOTOH, H. 1990. Axoplasmic transport of mitochondria in cultured dorsal root ganglion cells. Brain Res. 528:285–290.

TAO-CHENG, J-H. AND ROSENBLUTH, J. 1980. Nodal and paranodal membrane structure in complementary freeze-fracture replicas of amphibian peripheral nerves. Brain Res. 199:249–265.

TASHIRO, T. AND KOMIYA, Y. 1991. Maturation and ageing of the axonal cytoskeleton: Biochemical analysis of transported tubulin. J. Neurosci. Res. 30:192–299.

THIEL, W. 1957. Morphologische Ergebnisse an einzelnen markhaltigen Nervenfasern und ihre funktionelle Bedeutung. Acta Anat. 31:156–192.

THOMAS, P. K. 1980. Changes with age in the peripheral nerves in the rat: An ultrastructural study. Acta Neuropathol. 52:1.

THOMAS, P. K., BERTHOLD, C.-H., AND OCHOA, J. 1993. Microscopic anatomy of the peripheral nervous system. In: Peripheral Neuropathy, 3rd ed., Vol. 1. P. J. DYCK AND P. K. THOMAS, eds. Philadelphia, W. B. Saunders Company, pp. 28–91.

TIPPE, A. AND MULLER-MOHNSSEN, H. 1975. Further experimental evidence for the synapse hypothesis of Na+ current activation and inactivation at the Ranvier node. Naturwissenshaften 62:490–491.

TOEWS, A. D., ARMSTRONG, R., HOLSHEK, J., GOULD, R. M. AND MORELL, P. 1987. Unloading and transfer of axonally transported lipids. In: Axonal Transport. R. S. SMITH AND M. A. BISBY, eds. New York, Alan R. Liss, Inc., pp. 327–354.

TRAPP, B. D., HAUER, P. AND LEMKE, G. 1988. Axonal regulation of myelin protein mRNA levels in actively myelinating Schwann cells. J. Neurosci. 8:3515–3521.

TRIMBLE, W. S., LINIAL, M. AND SCHELLER, R. H. 1991. Cellular and molecular biology of the presynaptic nerve terminal. Annu. Rev. Neurosci. 14:93–122.

TSUKITA, S. AND ISHIKAWA, H. 1976. Three-dimensional distribution of smooth endoplasmic reticulum in myelinated axons. J. Electron Microsc. 25:141–149.

TSUKITA, S. AND ISHIKAWA, H. 1980. The movement of membranous organelles in axons: Electron microscopic identification of anterogradely and retrogradely transported organelles. J. Cell. Biol. 84:513–530.

TSUKITA, S. AND ISHIKAWA, H. 1981. The cytoskeleton in myelinated axons: Serial section study. Biomed. Res. 2:424–437.

TSUKITA, S., TSUKITA, S., KOBAYASHI, T. AND MATSUMOTO, G. 1986. Subaxolemmal cytoskeleton in squid giant axon. II. Morphological identification of microtubule- and microfilament-associated domains of axolemma. J. Cell Biol. 102:1710–1725.

TSUKITA, S., USUKURA, J., TSUKITA S. AND ISHIKAWA, H. 1982. The cytoskeleton in myelinated axons: A freeze-etch replica study. Neuroscience 7:2135–2147.

UHRIK, B. AND STÄMPFLI, R. 1981. Ultrastructural observations on nodes of Ranvier from isolated single frog peripheral nerve fibres. Brain Res. 215:93–101.

VALLEE, R. B. AND BLOOM, G. S. 1991. Mechanisms of fast and slow axonal transport. Annu. Rev. Neurosci. 14:59–92.

VAN DEURS, B., PETERSEN, O. W., OLSNES, S. AND SANDVIG, K. 1989. The way of endocytosis. Int. Rev. Cytol. 117:131–177.

VASILESCU, V. AND FILIP, D. A. 1979. The Ranvier node as a chemo-electric pulsatory unit: A study of its structure-functions relations. Rev. Roum. Morphol. Embryol. Physiol. 16:81–91.

WATSON, D. 1991. Regional variation in the abundance of axonal cytoskeletal proteins. J. Neurosci. Res. 30:226–231.

WAXMAN, S. G. 1987. Rules governing membrane reorganization and axo-glial interactions during the development of myelinated fibers. Prog. Brain Res. 71:121–142.

WAXMAN, S. G. AND BLACK, J. A. 1985. Membrane structure of the vesicotubular complexes in developing axons in rat optic nerve. Proc. R. Soc. B 225:357–364.

WAXMAN, S. G. AND BLACK, J. A. 1987. Macromolecular structure of the Schwann cell membrane: Perinodal microvilli. J. Neurol. Sci. 77:23–34.

WAXMAN, S. G. AND RITCHIE, J. M. 1993. Molecular dissection of the myelinated axon. Ann. Neurol. 33:121–136.

WEBSTER, H. DEF. 1962. Transient, focal accumulation of axonal mitochondria during the early stages of Wallerian degeneration. J. Cell Biol. 12:361–383.

WEISS, D. G., SEITZ-TUTTER, D., LANGFORD, G. M. AND ALLEN, R. D. 1987, The native microtubule as the engine for bidirectional organelle movements. In: Axonal Transport. R. S. SMITH AND M. A. BISBY, eds. New York, Alan R. Liss, Inc., pp. 91–111.

WILEY, C. A. AND ELLISMAN, M. H. 1980. Rows of dimeric-particles within the axolemma and juxtaposed particles within glia, incorporated into a new model for the paranodal glial-axonal junction at the node of Ranvier. J. Cell Biol. 84:261–280.

WILLIAMS, P. L. AND HALL, S. M. 1970. In vivo observations on mature, myelinated nerve fibres of the mouse. J. Anat. 107:31–38.

WILLIAMS, P. L. AND HALL, S. M. 1971. Prolonged in vivo observations of normal peripheral nerve fibres and their acute reactions to crush and deliberate trauma. J. Anat. 108:397–408.

WILLIAMS, P. L. AND KASHEF, R. 1968. Asymmetry of the node of Ranvier. J. Cell Sci. 3:341–356.

WILLIAMS, P. L. AND LANDON, D. N. 1963. Paranodal apparatus of peripheral nerve fibres of mammals. Nature 198:670–673.

WOOD, J. G., JEAN, D. H., WHITAKER, J. N., MCLAUGHLIN, B. J. AND ALBERS, R. W. 1977. Immunocytochemical localization of the sodium, potassium activated ATPase in knifefish brain. J. Neurocytol. 6:571–581.

WURTZ, C. C. AND ELLISMAN, M. H. 1988. Activity associated ultrastructural changes in peripheral nodes of Ranvier are independent of fixation. Exp. Neurol. 101:87–98.

YATES, A. J., BOUCHARD, J.-P. AND WHERRETT, J. R. 1976. Relation of axon membrane to myelin membrane in sciatic nerve during development: Comparison of morphological and chemical parameters. Brain Res. 104:261–271.

YOSHINO, J. E. AND DEVRIES, G. H. 1993. Isolation and characterization of axolemma-enriched fractions from rabbit and bovine peripheral nerve. Neurochem. Res. 18:297–303.

ZAGOREN, J. C. 1984. Cation binding at the node of Ranvier. In: The Node of Ranvier. J. C. ZAGOREN AND S. FEDOROFF, eds. New York, Academic Press, pp. 69–91.

ZELENA, J. 1980. Arrays of glycogen granules in the axoplasma of peripheral nerves at pre-ovoid stages of Wallerian degeneration. Acta Neuropatbol. (Berl.) 50:227–232.

ZENKER, W. AND HOHBERG, E. 1973a. α-motorische Nervenfaser: Axonquerschnittsfläche von Stammfaser und Endästen. Z. Anat. Entwicklungsgesch. 139:163–172.

ZENKER, W. AND HOHBERG, E. 1973b. α-Nerve-fibre: Number of neurotubules in the stem fibre and the terminal branches. J. Neurocytol. 2:143.

ZENKER, W. MAYR, R. AND GRUBER, H. 1975. Neurotubules: Different densities in peripheral motor and sensory nerve fibres. Experientia 31:318–320.

ZIMMERMAN, H. AND VOGT, M. 1989. Membrane proteins and synaptic vesicles and cytoskeletal specializations at the node of Ranvier in electric ray and rat. Cell Tissue Res. 258:617–629.