Paranodal Demyelination Research Articles

Structural abnormalities in freeze-fractured sciatic nerve fibres of diabetic mice

Nodal and paranodal regions of myelinated sciatic nerve fibres from diabetic (db/db) mice were examined in freeze fracture replicas. In some fibres, the axolemma was found to display abnormalities in the paranodal region. These include shallow, undifferentiated junctional indentations, thinning of the indentations with widening of the non-junctional grooves between them, particle clusters within the non-junctional grooves, and patches in which axolemmal E-face particles are distributed randomly rather than in the form of linear strings within grooves. Nodal structure, in contrast, is hardly affected. Nodal E-face and P-face particle densities in db/db axons are not significantly different from those in age-matched controls, although we found a few examples in which the E-face density fell slightly below the normal range. Occasional fibres showing evidence of paranodal or segmental demyelination were also seen. The results support paranodal pathology as a potential basis for reduced nerve conduction velocity in diabetic nerves but provide no evidence for significant changes in nodal structure or in nodal Na channel density in sciatic nerve fibres of the db/db mouse.

Nerve conduction studies in experimental non‐freezing cold injury: I. Local nerve cooling

In rabbits, the tibial nerve was exposed in the lower thigh under general anesthesia and cooled in a metal trough at 1 to 2 degrees C or 5 degrees C for 2, 3, or 4 hours. Nerve conduction studies showed local failure of conduction at the site of cooling which persisted after rewarming, and which was followed by distal degeneration of affected fibers. No persistent conduction block was seen. Changes in maximal velocity indicated that the fastest-conducting motor and afferent axons had been preferentially affected. Histological findings in nerves examined at different intervals after cooling confirmed the physiological evidence of primary axonal damage, affecting particularly large diameter fibers. Paranodal demyelination was inconspicuous and restricted to regions just proximal to sites of axonal degeneration. No segmental demyelination was seen. These results clarify previous uncertainties as to the time-course and distribution of nerve damage after local cooling at temperatures just above freezing point.

A computational test of the requirements for conduction in demyelinated axons

Conduction in demyelinated and remyelinating axons has been simulated with a computational model. The calculations made use of recent determinations of ionic channel densities in the internodal axolemma of Xenopus fibers. Several new morphological measurements reduced the number of parameters not directly obtained from experimental data. Action potentials and ionic currents were calculated for a wide range of fiber diameters and internodal lengths. The earliest stage of remyelination, characterized by Schwann cell attachment and extension of processes, was simulated by covering just a small percentage of the internode by a single cell layer. Conduction invariably failed if the internodal Na+ channel density was zero. The minimum density required for successful propagation agreed well with that measured in loose patch clamp experiments. Lateral diffusion of Na+ channels from nodes of Ranvier into the demyelinated internode did not restore conduction in blocked axons, and this was true regardless of the initial internodal Na+ channel density. Decreases in the internodal K+ channel density improved the safety factor for conduction, but this was significant only in the largest axons. Simulating minimal paranodal demyelination by eliminating the axo-glial junctional seals did not result in conduction block, but did produce large conduction delays.

Structural alterations of nerve during cuff compression.

Whether compression nerve injury is due to ischemia, direct mechanical injury, or both remains unsettled. To assess structural changes of nerve during compression, peroneal nerves of rats were compressed at various pressures for different times, and the structural alterations were stopped by simultaneous in situ and perfusion fixation. The structural changes observed during a few minutes of compression cannot be explained by ischemic injury because the pathologic alterations characteristic of ischemia take many hours to develop and in any case are different from the ones found here. The pressure- and time-related structural changes observed in the present study under the cuff were (i) decrease in fascicular area and increase in fiber density due to expression of endoneurial fluid; (ii) compression and expression of axoplasm, sometimes to the point of fiber transection; (iii) lengthening of internodes; and (iv) obscuration of nodes of Ranvier due to cleavage and displacement of myelin and overlapping of nodes by displaced loops of myelin. At the edges of the cuff the changes were (i) increase of fascicular area probably from expressed endoneurial fluid; (ii) widening of nodal gaps, perhaps mainly from translocated axonal fluid; and (iii) disordered structure of axoplasm. We suggest that the process of paranodal demyelination and axonal transection are linked, occur during the act of compression, and are due to shear forces. The initial event is expression of endoneurial fluid, followed by compression and expression of axoplasm and cleavage and displacement of layers of myelin. Conceivably, with prolonged cuff compression ischemic injury might be found to be superimposed on mechanical injury.

A Sequential Study of the Peripheral Nervous System Involvement in Experimental Chagas' Disease

To search for the sequential compromise of the spinal cord, nerves, and skeletal muscle in mice chronically infected with Trypanosoma cruzi, animals were subjected to electromyographic investigation, end-plate recordings, and histological studies at 7, 15, 37, 60, 90, 120, 180, 270, and 360 days postinfection. Electromyographic studies showed signs of motor unit remodeling as early as 15 days postinfection, when diminished duration and amplitude of motor unit potentials pointing to a primary muscle involvement were found. Thereafter, certain features of denervation, reinnervation, and primary muscle involvement were often found to coexist. Low miniature end-plate potentials with normal frequency and acetylcholine quantum content were found in end-plate recordings made at the phrenic-diaphragm in vitro. Double end-plate potentials were observed in most of the tested muscle fibers from day 90 postinfection. All these features suggest post-synaptic damage of the end-plate and the presence of reinnervation after day 90 postinfection. Histological studies disclosed inflammatory infiltrates consisting of lymphocytes and macrophages, with vasculitis as the main lesion in the hamstring muscles; intracellular parasites were seen in 25% of the cases. Neuropathic features, as expressed by type fiber grouping and grouped muscle fiber atrophy, were found. On nerve examination epineural, perineural, and endoneural vasculitis were seen. Digestion chambers and myelin ovoids (axonal degeneration) were observed. In teased fiber preparations, segmental internodal and paranodal demyelination and remyelination were found. The lumbar inflammatory spinal cord failed to show grey or white matter infiltrates. However, spinal roots and dorsal root ganglia were densely affected by inflammatory cells.

Heterogeneous distribution of fast and slow potassium channels in myelinated rat nerve fibres.

1. Potassium currents were measured in voltage-clamped single myelinated rat nerve fibres before and after paranodal demyelination with 0.2% pronase or 0.2% lysolecithin added to the external solution. Sodium currents were blocked by 300 nM-tetrodotoxin. For the purpose of comparison, intact frog nerve fibres were also investigated. 2. Our results suggest the existence of at least two distinct types of K+ channels in the intact node of Ranvier, one with slow and another with fast gating kinetics, in the ratio 4:1. 3. In the rat nodal membrane, slow K+ channels have voltage-dependent time constants of K+ deactivation with tau n = 68 ms at E = -105 mV and tau n = 26 ms at E = -150 mV at 20 degrees C. The activation curve of the slow K+ conductance is sigmoid with an inflexion point at -60 mV. This means that about 35% of the slow K+ channels are in the open state at the resting potential of -77 mV. Slow K+ channels could be blocked by 10 mM-tetraethylammonium chloride, but were insensitive to 4-aminopyridine. 4. After paranodal demyelination the ratio of fast to slow K+ channels increased from 17 to 83%. As in the frog (Dubois, 1981 alpha), the population of fast K+ channels in the rat may consist of two different subgroups, both of which can be blocked by 4-aminopyridine. 5. Demyelination was accompanied by an increase in the capacity current which was used to estimate the exposed membrane area. The density of slow and fast K+ channels was calculated from the quotient of the steady-state K+ conductance to membrane area. The density of the slow K+ channels is maximal in the nodal membrane and decreases to 1/31 in the internode. By contrast, the distribution of the fast K+ channels differs, their density being maximal in the paranode and decreasing to one-sixth in the node and internode.

Myelin sheath remodelling in remyelinated rat sciatic nerve.

In order to elicit de- and remyelination adult rat sciatic nerves were injected with diphtheria toxin dissolved in phosphate buffered saline (PBS). Control nerves were injected with PBS alone. After survival times of 1-10 weeks, the animals were perfused with glutaraldehyde. Specimens from the injected nerves were processed for light microscopic (LM) examination of teased fibres or for electron microscopic (EM) examination of longitudinal thin sections. LM examination of teased fibres after survival times of 6-10 weeks, showed that most remyelinated internodes are 150-300 microns long. In addition, some exceptionally short Schwann cell sheaths, with lengths of 15-150 microns, occur intercalated between conventional remyelinated internodes. EM analysis of thin sections showed that axonal evaginations penetrate in between the terminating myelin lamellae in fibres with nodal widening and/or paranodal demyelination, at early stages of demyelination. Such alterations are not present in relation to myelin sheaths formed during remyelination, which commences about 3 weeks after injection. In addition, some scattered contracted Schwann cell sheaths, which may be as short as 5-10 microns, occur at all stages. These are more frequent shortly after onset of remyelination than at later stages, and they are either composed of a cytoplasmic investment bordered by heminodes, or a more or less distorted myelin sheath bordered by nodes of Ranvier. This picture is very similar to the myelin sheath remodelling observed in regenerated rat sciatic nerves, and in some developing nerves with a mismatch between nerve growth and internodal elongation. It is concluded that a myelin sheath remodelling occurs in de- and remyelinated rat sciatic nerve, presumably as a result of the lack of longitudinal growth.

Model investigations of the temperature dependence of demyelinated and reorganized axonal membrane.

The temperature dependence (from 10 degrees to 50 degrees C) of the intracellular action potentials' parameters in a fiber with a simulated reorganization of the axonal membrane against the background of a systematic paranodal demyelination of the fiber was investigated. The temporal and spatial distribution of the potential as well as the ionic currents' kinetics have been represented. The reorganization of the axonal membrane was achieved by means of potassium channels blocking and increase of the sodium-channel permeability, while the demyelination was achieved by means of elongation of the nodes of Ranvier. In order to account for the temperature dependence of the rate constants and of the maximal sodium and potassium permeabilities, the temperature coefficients (Q10) have been used. It has been shown for the demyelinated and reorganized membrane that increased temperature blocks the conduction at temperatures much higher than the blocking temperature for the demyelinated fiber only. When temperature increases the amplitude of the potential decreases while the velocity increases up to temperatures approaching the blocking temperature after which it abruptly drops. The dependence of the asymmetry and the wavelength of the potential on temperature is complex and nonmonotonic. For the reorganized membrane at the background of a given degree of demyelination with increasing temperature the ionic currents' flow and the membrane conduction respectively increase, but, at lower temperatures, when the temperature increase is combined with the increased degree of the fiber demyelination, the conduction is blocked.

Role of endoneural cells in experimental allergic neuritis and characterisation of a resident phagocytic cell.

Electrophysiological, clinical and histological techniques were used to monitor the time course of events related to experimental allergic neuritis (EAN) in 48 Lewis rats. The primary lesion was found to be paranodal demyelination without cellular infiltration. Endoneural phagocytes derive from hematogenous ED1+ED2- monocytes and possibly from resident ED1-ED2+ monocytic cells, not from Schwann cells and fibroblasts. We demonstrate a population of monocytic Ia-bearing, ED1-ED2+ spindle-shaped cells residing in normal peripheral nerve and provide evidence for their transformation into macrophages in the course of EAN.

ELECTROPHYSIOLOGICAL INVESTIGATIONS IN ADOPTIVELY TRANSFERRED EXPERIMENTAL AUTOIMMUNE ENCEPHALOMYELITIS IN THE LEWIS RAT

Lewis rats with experimental autoimmune encephalomyelitis transferred adoptively with myelin basic protein-specific T line cells (AT-EAE) were studied clinically, electrophysiologically, and histologically. Injection with 5 x 10(6) line cells induced EAE with a rapidly developing tetraplegia after a latent period of 4 days. Electrophysiological testing revealed a profound slowing of afferent conduction within the dorsal column of the spinal cord, and conduction abnormalities in the spinal roots. Injection with a lower cell dose of 1 x 10(6) T line cells caused only moderate clinical signs paralleled by milder conduction slowing and conduction failure. Light microscopy showed marked inflammation with infiltration of mononuclear cells and some demyelination throughout the spinal cord and roots. Inflammation and demyelination were dose dependent and the caudal parts of the spinal cord were more affected than the cranial parts. The peripheral nerves were free of electrophysiological and morphological alterations. Systemic treatment with 4-aminopyridine accelerated and partially restored conduction in the dorsal columns and roots, while increasing the body temperature had a detrimental effect, suggesting demyelination as a prominent pathophysiological mechanism. These findings show that in AT-EAE in Lewis rats the dysfunction of the central nervous system and of spinal roots is cell dose dependent, that the peripheral nervous system distal to the spinal roots is spared, and suggest that paranodal demyelination is an important pathogenic mechanism.

Histopathological heterogeneity of neuropathy in insulin-dependent and non-insulin-dependent diabetes, and demonstration of axo-glial dysjunction in human diabetic neuropathy.

Altered sorbitol and myo-inositol metabolism, (Na,K)-ATPase function, electrochemical sodium gradients, axonal swelling, and distortion and disruption of the node of Ranvier ("axo-glial dysjunction") directly implicate hyperglycemia in the pathogenesis of neuropathy in diabetic rats, but the relevance of this sequence to clinical neuropathy in heterogeneous groups of diabetic patients remains to be established. Fascicular sural nerve morphometry in 11 patients with neuropathy complicating insulin-dependent diabetes revealed a pattern of interrelated structural changes strikingly similar to that of the diabetic rat when compared to age-matched controls. 17 older non-insulin-dependent diabetic patients with comparable duration and severity of hyperglycemia and severity of neuropathy, displayed similar nerve fiber loss, paranodal demyelination, paranodal remyelination and segmental demyelination compared to age-matched controls, but axo-glial dysjunction was replaced by Wallerian degeneration as the primary manifestation of fiber damage, and fiber loss occurred in a spatial pattern consistent with an ischemic component. The mechanistic model developed from the diabetic rat does indeed appear to apply to human diabetic neuropathy, but superimposed hormonal, metabolic, vascular, and/or age-related effects alter the morphologic expression of the neuropathy in non-insulin dependent diabetes.

The neuromuscular pathology of experimental Chagas' disease

In this work, we describe skeletal muscle, neuromuscular junction, nerve and spinal cord lesions in the mouse model system of Chagas' disease. Myositis was a common finding and Trypanosoma cruzi amastigote nests were frequently found in the muscle fibers. Angular atrophy, targetoid fibers, groups of atrophic fibers, fibrosis, myofiber necrosis and phagocytosis of cellular debris were also observed. The neuromuscular junction studies showed degeneration of intramuscular nerve fibers, swelling and distortion of nerve endings and multiple ramifications on the same muscle fiber. Collateral, terminal and ultraterminal axonal sprouts were also present. Inflammatory neuropathy was seen in all of the infected mice. Demyelination, axonal degeneration, remyelination and axonal regeneration were observed in the transverse sections. There was an average reduction of 29% in the total number of myelin fibers. The teasing of single myelin fibers showed segmental and paranodal demyelination and remyelination more frequently than axonal degeneration and regeneration. The lumbar spinal cords presented inflammatory cell infiltration associated with tissue destruction. Amastigote nests were found in 3 out of the 8 infected mice studied. There was a mean loss of 21% of the large cytoneurons of the anterior horn of the lumbar spinal cord.

Apolipoprotein E is released by rat sciatic nerve during segmental demyelination and remyelination.

Apolipoprotein E (apo E) is synthesized and released in greatly increased amounts by peripheral nerve following Wallerian degeneration; it has been suggested that this protein may function in the transport of degenerated myelin lipid. The purpose of this study was to determine if the amount of apo E released by rat peripheral nerve is increased following selective demyelination, in the absence of significant axonopathy. Using an immunoturbidimetric assay, release of apo E from excised sciatic nerve segments was measured during the phases of acute demyelination and remyelination caused by tellurium (Te) toxicity, during segmental demyelination in chronic lead (Pb) poisoning, and during Wallerian degeneration following nerve crush. Morphologic changes were examined in contralateral sciatic nerves by nerve-fiber teasing or by light and electron microscopy of transverse sections. As in previous studies, the amount of apo E released from the nerves was greatly increased following Wallerian degeneration due to nerve crush. In Te neuropathy, increased release of apo E was first detected on the fourth day of Te exposure, corresponding temporally to the acute onset of paralysis and segmental demyelination. Apolipoprotein E release rose steeply to a maximum of ten times the control values by day 9 and then gradually waned during the next five weeks, corresponding to a period of active remyelination and resolution of the neuropathy. In the demyelinating neuropathy of chronic lead poisoning, apo E release was increased four times over control animals after seven weeks of exposure, with less than 10% of teased fibers showing early paranodal demyelination and no evidence of remyelination.(ABSTRACT TRUNCATED AT 250 WORDS)

Extracellular currents and potentials of the active myelinated nerve fiber

This paper is concerned with the accurate and rapid calculation of extracellular potentials and currents from an active myelinated nerve fiber in a volume conductor, under conditions of normal and abnormal conduction. The neuroelectric source for the problem is characterized mathematically by using a modified version of the distributed parameter model of L. Goldman and J. S. Albus (1968, Biophys. J., 8:596-607) for the myelinated nerve fiber. Solution of the partial differential equation associated with the model provides a waveform for the spatial distribution of the transmembrane potential V(z). This model-generated waveform is then used as input to a second model that is based on the principles of electromagnetic field theory, and allows one to calculate easily the spatial distribution for the potential everywhere in the surrounding volume conductor for the nerve fiber. In addition, the field theoretic model may be used to calculate the total longitudinal current in the extracellular medium (I0L(z)) and the transmembrane current per unit length (im(z)); both of these quantities are defined in connection with the well-known core conductor model and associated cable equations in electrophysiology. These potential and current quantities may also be calculated as functions of time and as such, are useful in interpreting measured I0L(t) and im(t) data waveforms. An analysis of the accuracy of conventionally used measurement techniques to determine I0L(t) and im(t) is performed, particularly with regard to the effect of electrode separation distance and size of the volume conductor on these measurements. Also, a simulation of paranodal demyelination at a single node of Ranvier is made and its effects on potential and current waveforms as well as on the conduction process are determined. In particular, our field theoretic model is used to predict the temporal waveshape of the field potentials from the active, non-uniformly conducting nerve fiber in a finite volume conductor.

Cross talk between intraspinal elements during progression of IDPN neuropathy

Unusual electrical interactions between neuronal elements of cat spinal cord were examined during the evolution (7–70 days) of proximal paranodally demyelinated axonal enlargements in alpha motor axons induced by β,β′-iminodipropionitrile (IDPN) treatment, 50 mg/kg ip, once weekly for up to 5 weeks. The electrical cross talk was observed as early as 7 days of the neuropathy, at which time it occurred in 17.0% of all motoneuron perikaryal recordings. The incidence was greater at 14 (39.1%) and 35 (26.1%) days, without preferential involvement of any motoneuron type. The frequency of recordings from axonal swellings increased from 7 to 35 days but without an increase in cross talk per recording. At 70 days of the neuropathy, L7 ventral root was stimulated repetitively to examine a possible influence of potassium on cross talk. Subsequently, action potentials could be elicited in motoneurons by stimulation of additional other ventral root filaments. These studies are in agreement with the lack of direct electrical apposition between excitable membranes in IDPN neuropathy but suggest support for a role for an accumulation of extracellular potassium, due to paranodal demyelination as the axon enlarges, in the pathogenesis of these aberrant electrical interactions.

Nodal and paranodal structure during Wallerian degeneration in frog spinal nerve

The nodal and paranodal regions of myelinated peripheral nerve fibers in frogs were examined at sequential times (1–24 days) during Wallerian degeneration. In the region up to 3 mm distal to the transection, paradonal demyelination and axoplastic degeneration became apparent on day 4 and progressed to involve most of the nodes by day 8. The E-fracture face of the axolemma showed a patchy distribution of nodal particles and some paranodal demyelination on days 4 and 6. On day 8, nodal particles were evenly distributed at low concentration and the adjacent demyelinated paranodal regions showed a corresponding increase in particle density, suggesting redistribution of the nodal particles. The sequence of changes seen is comparable to that in Wallerian degeneration of central nervous system (CNS) fibers but progressed more rapidly in the peripheral nervous system (PNS). In addition a higher proportion of PNS fibers shows pathological changes at corresponding time periods.

Teased fibre studies in leprous neuropathy

Nerve biopsies were taken from four cases of leprosy, which included borderline tuberculoid, borderline and lepromatous types. They were examined in 1-μm resin sections and in teased preparations. The most common finding in teased fibres from each leprosy type was paranodal demyelination affecting successive internodes. In transverse sections some fibres showed changes associated with axonal atrophy. Together, these findings suggest that demyelination in some cases may be secondary to axonal changes. In addition, there was evidence of focal areas of demyelination affecting whole internodes of many fibres at the same level across the nerve and which was possibly caused by local factors.

Ultrastructural observations of organelle accumulation in the equine recurrent laryngeal nerve

The left recurrent laryngeal nerves from five horses with sub-clinical neuropathy were examined by light and electron microscopy in a study designed to examine accumulation of axonal organelles at paranodal and internodal locations. Transverse sections of the nerve showed scattered fibres with split myelin sheaths and axonal accumulation of organelles. On longitudinal sections these collections were seen to result from an axonal outpouching in which dense lamellar bodies and mitochondria had accumulated. These paranodal collections, which could be found on both sides of the node, were often associated with infoldings of the terminal loops of myelin and with occasional paranodal demyelination. The fact that many of the organelles in the outpouches were lysosomal in nature was confirmed by their positive staining for cathepsin D activity. Longitudinal sections demonstrated a number of axons which were swollen over a long distance and which contained focal accumulations of similar organelles. In places, however, there was a clear separation between these organelles and the cytoskeletal proteins. In each case these swollen axons were surrounded by Schwann cell nuclei and their processes, forming well-ordered onion bulbs. The possibility that these two types of changes, i.e. the paranodal accumulations and the axonal swellings could result from a disturbance in axonal transport in this distal axonopathy is discussed.

Schwann cell proliferation and migration during paranodal demyelination

This study examined Schwann cell behavior during paranodal demyelination induced by beta,beta'-iminodipropionitrile (IDPN). The stimuli for Schwann cell proliferation, extensively studied in vitro, are less well understood in vivo. Most in vivo systems previously used to examine Schwann cell proliferation in disease are dominated by loss of internodal myelin sheaths. As used in this study, IDPN administration produces neurofilamentous axonal swellings and paranodal demyelination, without segmental demyelination or fiber degeneration. We asked whether Schwann cells would proliferate following the restricted paranodal demyelination that accompanies the axonal swellings, and if so what the sources and distributions of new Schwann cells might be. IDPN was given as a single large dose (2 ml/kg) to 21-d-old rats. Neurofilamentous axonal swellings formed in the proximal regions of motor axons, reaching their greatest enlargement in the root exit zone 8 d after IDPN administration. These swellings subsequently migrated distally down the nerves at rates approaching 1 mm/d. The axonal enlargement was consistently associated with displacement of the myelin sheath attachment sites into internodal regions, and consequent paranodal demyelination. This stage was associated with perikaryal changes, including nucleolar enlargement, "girdling" of the perikaryon, and formation of attenuated stalks separating the perinuclear region from the external cytoplasmic collar. Schwann cells proliferated abundantly during this stage. Daughter Schwann cells migrated within the endoneurial space (outside the nerve fiber basal laminae) to overlie the demyelinated paranodes of swollen nerve fibers. In these regions, local proliferation of Schwann cells continued, resulting in large paranodal clusters of Schwann cells. As the axonal calibers subsequently returned to normal, the outermost myelin lamellae of the original internodes returned to their paranodal attachment sites and the supernumerary Schwann cells disappeared. Formation of short internodes, segmental demyelination, and nerve fiber loss were rare phenomena. These results indicate that paranodal demyelination is a sufficient stimulus to excite abundant Schwann cell proliferation; neither internodal demyelination nor myelin breakdown is a necessary stimulus for mitosis. The 3H-thymidine incorporation studies indicated that the sources of new Schwann cells included markedly increased division of the Schwann cells of unmyelinated fibers and, as they formed, supernumerary Schwann cells. In addition, there were rare examples of 3H-thymidine incorporation by Schwann cells associated with myelinated nerve fibers.(ABSTRACT TRUNCATED AT 400 WORDS)

Disseminated vasculomyelinopathy in the peripheral nervous system mediated by immune complexes (ICs): Immunohistochemical studies of sciatic nerves in chronic serum sickness (CHSS) in rabbits

Histological examination of 20 sciatic nerves from rabbits with experimental chronic serum sickness (CHSS) revealed patchy vasculitis of the vasa nervorum of various intensity. The vessel lesions ranged from endothelial proliferation to vessel wall necrosis with fibrinoid degeneration and infiltration by lymphocytes, plasma cells, macrophages and, sporadically, by neutrophils. Perivascularly, there were oedema, chronic infiltrates or small haemorrhages. The myelinated fibres in close relation to the vascular system were focally depleted and features of perivascular demyelination were found. Teased fibres showed paranodal and segmental demyelination, axonal degeneration and, sporadically, remyelination. In all cases, immunofluorescent deposits of bovine serum albumin (BSA), IgG and C3 complement were found in and around some vasa nervorum. Other indirect evidence for immune complex (IC) deposition was provided by ultrastructural examination where vascular and endoneurial osmophilic deposits were found; in 4 cases with paracrystalline organization resembling cryoglobulin component. IC-mediated vasculitis led to blood-nerve barrier impairment and leakage of serum proteins into the endoneurial space. The morphological and immunohistochemical changes in this model which develop after a latency period of 2 or more weeks, strongly resemble those observed in human acquired inflammatory demyelinating polyradiculoneuropathies or in connective tissue diseases.