Peripheral nerve as an osmometer: role of the perineurium in frog sciatic nerve

Primary perineuritis, a rare but treatable neuropathy: Review of perineurial anatomy, clinicopathological features, and differential diagnosis

The perineurium surrounds each fascicle in peripheral nerves, forming part of the blood-nerve barrier. We describe its normal anatomy and function. "Perineuritis" refers to both a nonspecific histopathological finding and more specific clinicopathological entity, primary perineuritis (PP). Patients with PP are often assumed to have nonsystemic vasculitic neuropathy until nerve biopsy is performed. We systematically reviewed the literature on PP and developed a differential diagnosis for histopathologically defined perineuritis. We searched PubMed, Embase, Scopus, and Web of Science for "perineuritis." We identified 20 cases (11 M/9F) of PP: progressive, unexplained neuropathy with biopsy showing perineuritis without vasculitis or other known predisposing condition. Patients ranged in age from 18 to 75 (mean 53.7) y and had symptoms 2-24 (median 4.5) mo before diagnosis. Neuropathy was usually sensory-motor (15/20), painful (18/19), multifocal (16/20), and distal-predominant (16/17) with legs more affected than arms. Truncal numbness occurred in 6/17; 10/18 had elevated cerebrospinal fluid (CSF) protein. Electromyography (EMG) and nerve conduction studies (NCS) demonstrated primarily axonal changes. Nerve biopsies showed T-cell-predominant inflammation, widening, and fibrosis of perineurium; infiltrates in epineurium in 10/20 and endoneurium in 7/20; and non-uniform axonal degeneration. Six had epithelioid cells. 19/20 received corticosteroids, 8 with additional immunomodulators; 18/19 improved. Two patients did not respond to intravenous immunoglobulin (IVIg). At final follow-up, 13/16 patients had mild and 2/16 moderate disability; 1/16 died. Secondary causes of perineuritis include leprosy, vasculitis, neurosarcoidosis, neuroborreliosis, neurolymphomatosis, toxic oil syndrome, eosinophilia-myalgia syndrome, and rarer conditions. PP appears to be an immune-mediated, corticosteroid-responsive disorder. It mimics nonsystemic vasculitic neuropathy. Cases with epithelioid cells might represent peripheral nervous system (PNS)-restricted forms of sarcoidosis.

A Brief Review of In Vitro Models for Injury and Regeneration in the Peripheral Nervous System

Nerve axonal injury and associated cellular mechanisms leading to peripheral nerve damage are important topics of research necessary for reducing disability and enhancing quality of life. Model systems that mimic the biological changes that occur during human nerve injury are crucial for the identification of cellular responses, screening of novel therapeutic molecules, and design of neural regeneration strategies. In addition to in vivo and mathematical models, in vitro axonal injury models provide a simple, robust, and reductionist platform to partially understand nerve injury pathogenesis and regeneration. In recent years, there have been several advances related to in vitro techniques that focus on the utilization of custom-fabricated cell culture chambers, microfluidic chamber systems, and injury techniques such as laser ablation and axonal stretching. These developments seem to reflect a gradual and natural progression towards understanding molecular and signaling events at an individual axon and neuronal-soma level. In this review, we attempt to categorize and discuss various in vitro models of injury relevant to the peripheral nervous system and highlight their strengths, weaknesses, and opportunities. Such models will help to recreate the post-injury microenvironment and aid in the development of therapeutic strategies that can accelerate nerve repair.

Electrochemical activation and inhibition of neuromuscular systems through modulation of ion concentrations with ion-selective membranes

Conventional functional electrical stimulation aims to restore functional motor activity of patients with disabilities resulting from spinal cord injury or neurological disorders. However, intervention with functional electrical stimulation in neurological diseases lacks an effective implantable method that suppresses unwanted nerve signals. We have developed an electrochemical method to activate and inhibit a nerve by electrically modulating ion concentrations in situ along the nerve. Using ion-selective membranes to achieve different excitability states of the nerve, we observe either a reduction of the electrical threshold for stimulation by up to approximately 40%, or voluntary, reversible inhibition of nerve signal propagation. This low-threshold electrochemical stimulation method is applicable in current implantable neuroprosthetic devices, whereas the on-demand nerve-blocking mechanism could offer effective clinical intervention in disease states caused by uncontrolled nerve activation, such as epilepsy and chronic pain syndromes.

Homeostatic regulation of the endoneurial microenvironment during development, aging and in response to trauma, disease and toxic insult

The endoneurial microenvironment, delimited by the endothelium of endoneurial vessels and a multi-layered ensheathing perineurium, is a specialized milieu intérieur within which axons, associated Schwann cells and other resident cells of peripheral nerves function. The endothelium and perineurium restricts as well as regulates exchange of material between the endoneurial microenvironment and the surrounding extracellular space and thus is more appropriately described as a blood-nerve interface (BNI) rather than a blood-nerve barrier (BNB). Input to and output from the endoneurial microenvironment occurs via blood-nerve exchange and convective endoneurial fluid flow driven by a proximo-distal hydrostatic pressure gradient. The independent regulation of the endothelial and perineurial components of the BNI during development, aging and in response to trauma is consistent with homeostatic regulation of the endoneurial microenvironment. Pathophysiological alterations of the endoneurium in experimental allergic neuritis (EAN), and diabetic and lead neuropathy are considered to be perturbations of endoneurial homeostasis. The interactions of Schwann cells, axons, macrophages, and mast cells via cell-cell and cell-matrix signaling regulate the permeability of this interface. A greater knowledge of the dynamic nature of tight junctions and the factors that induce and/or modulate these key elements of the BNI will increase our understanding of peripheral nerve disorders as well as stimulate the development of therapeutic strategies to treat these disorders.

A mechanical model for collagen fibril load sharing in peripheral nerve of diabetic and nondiabetic rats

Peripheral neuropathy affects approximately 50% of the 15 million Americans with diabetes. It has been suggested that mechanical effects related to collagen glycation are related to the permanence of neuropathy. In the present paper, we develop a model for load transfer in a whole nerve, using a simple pressure vessel approximation, in order to assess the significant of stiffening of the collagenous nerve sheath on endoneurial fluid pressure. We also develop a fibril-scale mechanics model for the nerve, to model the straightening of wavy fibrils, producing the toe region observed in nerve tissue, and also to interrogate the effects of interfibrillar crosslinks on the overall properties of the tissue. Such collagen crosslinking has been implicated in complications in diabetic tissues. Our fibril-scale model uses a two-parameter Weibull model for fibril strength, in combination with statistical parameters describing fibril modulus, angle, wave-amplitude, and volume fraction to capture both toe region and failure region behavior of whole rat sciatic nerve. The extrema of equal and local load-sharing assumptions are used to map potential differences in diabetic and nondiabetic tissues. This work may ultimately be useful in differentiating between the responses of normal and heavily crosslinked tissue.

An electrophysiological method for measuring the potassium permeability of the nerve perineurium

An electrophysiological method is described for measuring the potassium permeability (PK) of the perineurium of the sciatic nerve of the frog. The method is based on the principle of grease-gap recording, in which an insulating compartment separates two surface recording electrodes. The sciatic nerves of frogs Rana temporaria and R. pipiens were isolated and mounted across a five compartment chamber, with Vaseline grease seals on the partitions between compartments. Compartments #1, #2 and #5 contained frog Ringer solution, #4 was filled with Vaseline and formed the grease gap, and #3 was the test compartment in which solutions could be changed. The nerve was stimulated via platinum electrodes in compartments #1 and #2, and DC potentials and compound action potentials (CAP) were recorded between Ag/AgCl electrodes connected through Ringer-agar bridges to compartments #3 and #5. In nerves with undamaged perineurium, changing from normal Ringer to high [K+] Ringer (100 mM, KCl replacing NaCl) for 2 min caused negligible change in DC potential or CAP, indicating that raised [K+] was not reaching the axon surface, and hence that the perineurium was exerting a diffusional restriction on K+ entry. In nerves damaged by stretching or drying, K+ pulses caused a depolarising change in DC potential (delta DC), and corresponding decline in CAP amplitude, consistent with a leaky perineurium allowing K+ entry and axonal depolarisation. Ringer made hypertonic by the addition of 2.5 M sucrose or 5 M NaCl caused increased perineurial permeability to K+. The method was calibrated by measuring the delta DC in response to raised [K+] in the range 5-100 mM [K+] in desheathed nerves; from this calibration curve relating delta DC to endoneurial [K+] it was possible to calculate the change in endoneurial [K+] occurring in intact preparations. The calculations showed that the undamaged perineurium had a PK of < 6.3 x 10(-7) cm.s-1, similar to the value calculated for in situ nerves using radioisotopic techniques, but less than the value reported for isolated perineurial cylinders. The method gives real-time information on the K+ permeability of the nerve perineurium and its modulation by experimental treatments.

Perineurial permeability to sodium during Wallerian degeneration in rat sciatic nerve

In rat sciatic nerves, the effect of Wallerian degeneration on the rate of transperineurial passage of sodium between the endoneurium and the epineurial extracellular space was investigated. In nerves transected and ligated at the sciatic notch, an in situ technique was used to measure the permeability coefficient-surface area product (PS) of the mid-thigh portion of the perineurium to 22Na. Sampling times ranged from one day to sixteen weeks after the lesion. Additionally, endoneurial water content (an indicator of nerve edema) was also measured in transected, degenerating nerves at the same sampling times. Endoneurial water content increased significantly by the fourth day after transection, peaked at four weeks, and then remained elevated through 16 weeks of post-lesion measurement. The PS of the perineurium to 22Na on the 4th day after transection was significantly greater than that of control animals. This increase then declined to normal levels through the 2nd week, and finally increased to values that were 3-fold to 4-fold of control values for the remainder of the observation period. The earlier, short lasting increase in perineurial PS is probably associated with the inflammatory response to nerve section, and proliferation of perineurial layers and cells. The later increase in perineurial permeability is proposed to play a role in the dissipation of endoneurial hydrostatic pressure and clearance of myelin debris from the endoneurium. In view of the complex changes in perineurial permeability described herein, it would seem inappropriate to consider these phenomena merely as passive breakdowns of the barrier properties of the perineurium.

Blood-nerve transfer of albumin and its implications for the endoneurial microenvironment

Blood-nerve transfer of plasma albumin was studied by measuring the permeability coefficient-surface area (PS) product of the blood-nerve barrier (BNB) to 125I-albumin in rat sciatic nerve using the i.v. bolus injection method. The calculated PS was 6.3 +/- 0.5 (S.E.M.) x 10(-7) ml.g-1.s-1. This value is smaller by more than an order of magnitude of that measured for sucrose and confirms the relative impermeance of the BNB to blood-borne solutes. From a review of the available evidence, it is concluded that normal blood-nerve exchange occurs predominantly across the endoneurial microvasculature, and the PS of the BNB reflects the permeability of capillaries to a greater extent than that of the perineurium. The only capillaries found to be less permeable than these are the cerebral capillaries. Proximo-distal differences (sciatic vs tibial) of the PS could not be detected. Blood-nerve albumin transfer was calculated at 1.2 mg.g-1.day-1, and the daily turnover of endoneurial albumin to be about 30%. It is postulated that small increases in PS of BNB to albumin lead to an elevation of endoneurial albumin concentration and, through the operation of Starling forces, subsequently produce endoneurial oedema. A major question posed by the results of this study is the identity of pathways for clearance of albumin and other macromolecules from the endoneurium.

Calcium transfer at the blood-nerve barrier of the frog sciatic nerve

Calcium transfer across the blood-nerve barrier of the frog sciatic nerve was studied using an in situ perfusion technique and an in vivo i.v. bolus injection technique. The permeability-surface area product of 45Ca at the blood-nerve barrier, (PA)BNB, calculated from radioactivity in the desheathed nerve segment after 5 min of circulation of tracer, and corrected for the residual radioactivity in the blood space, equaled 4.4 +/- 0.4 (S.E.M.) X 10(-5) ml.s-1.g-1 wet wt. The (PA)BNB of 45Ca was independent of [Ca2+] in the perfusion medium between 0.18 and 18 mM. The permeability-surface area products of 45Ca across the perineurium [(PA)per] also was measured by an in situ incubation technique, and equaled 1.45 +/- 0.41 X 10(-5) ml.s-1.g-1 wet wt. (n = 8). The half time (t 1/2) for nerve calcium to equilibrate with plasma calcium was calculated to be 60 min. The low, passive permeability to calcium of the blood-nerve barrier probably limits marked calcium concentration changes in nerve endoneurium following transient changes of plasma calcium, but should not alter steady-state responses.

Permeability of endoneurial capillaries to K, Na and Cl and its relation to peripheral nerve excitability

The permeability coefficient-surface area products (PA) of frog sciatic nerve endoneurial capillaries to K, Na and Cl were measured with an in situ perfusion technique and found to be 40.3, 24.6 and 32.8 X 10(-5) ml . g-1 . s-1, respectively. PAs to [14C]sucrose and 42K, when measured simultaneously, and their ratio were independent of perfusate K concentration (0.1-10.0 mM). Simultaneous measurements with 36Cl and 42K indicated that the Cl/K permeability ratio was significantly smaller than the mobility ratio of these two ions in free solution. On the other hand, comparable experiments with 22Na and 42K revealed that the K/Na permeability ratio was not significantly different from its respective mobility ratio. Thus, these results provide no evidence of facilitated transport of K by endoneurial capillaries, and suggest that K, Na and Cl traverse the endoneurial capillary wall by a paracellular route which is weakly selective for cations. The minimum extracellular K concentration (Ke) capable of producing a depolarization conduction block in frog sciatic nerve was between 12.5 and 15.0 mM. When the vasculature of this nerve was perfused with a hyperkalaemic (20.0 mM) Ringer solution, a conduction block developed in 7.9 min. Comparison of this time with the theoretically predicted rate of change of endoneurial Ke (induced by a comparable change of intravascular K concentration) indicated that an increase of endoneurial Ke is transmitted directly to the paranodal spaces of nerve fibres so as to immediately influence axonal excitability.(ABSTRACT TRUNCATED AT 250 WORDS)

Calcium in rat peripheral nerve during chronic alterations in plasma calcium

The calcium content in desheathed tibial nerve was compared to that in cerebellum in rats fed diets containing either 0.01% (low), 0.67% (control) or 3.0% (high) Ca, for 8 weeks. For changes in concentration of plasma ionized Ca, 48% below and 35% above the control mean, percent change in endoneurial Ca content is linearly related, with a slope of 0.80, to percent change in plasma ionized Ca. A line with a slope of 0.21 describes the relation between percent change in cerebellum Ca and percent change in plasma ionized Ca. Plasma, cerebellum and nerve concentrations of Na, K and Cl were similar in the control compared with the two experimental groups of animals. The concentration of plasma Mg varied 20% below and 17% above the control mean, inversely with plasma Ca, but nerve and cerebellum Mg did not change from control values. The results of this study fail to demonstrate Ca homeostasis in rat peripheral nerve endoneurium during chronic hypo- and hypercalcemia. Endoneurial Mg, however, appears to be regulated.

Endoneurial capillary permeability to [14C]sucrose in frog sciatic nerve

Using an in situ perfusion technique, we measured the [14C]sucrose permeability-surface area product (PA) in endoneurial capillaries of 48 frog sciatic nerves, as 6.6 +/- 0.6 (S.E.M.) X 10(-5) s-1, and the vascular space as 1.31 +/- 0.10%. Assuming A = 30 cm2/g, P = 2.2 X 10(-6) cm/s. P for sucrose was greater than P in some barrier tissues with tight junctions, but was less than P in all capillaries examined so far except rat cerebral capillaries. These observations demonstrate that endoneurial capillaries are an effective part of the blood-nerve barrier to water-soluble non-electrolytes. The findings are consistent with capillary impermeability to microperoxidase and with capillary ultrastructure.

Effects of cold adaptation and starvation on sciatic nerve fibers in the frog

The conditions under which frogs are kept prior to experimentation were found to have a measurable effect on peripheral nerve structure. Frogs kept for 12 weeks at 4 degree C had markedly shrunken sciatic nerve fibers compared with frogs kept at 19 degrees C. Intermediate fiber shrinkage was found for frogs kept at 19 degrees C without feeding. Counts of neurofilaments and microtubules showed that fiber shrinkage was from a preferential loss of filaments, indicating cold- or starvation-induced atrophy of the axon's cyto-skeleton. This effect, however, was superimposed with additional osmotic axonal shrinkage, causing filament densities to increase per area. There were no changes in myelin sheath thickness due to cold adaptation or fasting.

Morphology of endoneurial blood vessels of frog sciatic nerve during vascular perfusion

In order to determine if increased injection pressures can alter the permeability and ultrastructure of blood vessels of the frog blood-nerve barrier, these vessels were examined following perfusion of the iliac artery at rates of 0.21 or 0.82 ml/min. At either perfusion rate, endoneurial blood vessel profiles were clearly evident and the surface area of these vessels amounted to 60% of the surface area of the perineurium. In all vessels a large number of vesicles were present within the endothelial cells. Many were attached by necks to one or the other plasma membrane, but no transcellular channels were evident. At the higher flow rate no changes in vesicles or junctions were seen, but blebs and blisters were evident at the luminal membranes of the endoneurial endothelium. When microperoxidase was perfused at 0.82 ml/min, reaction product frequently flooded the endothelial cells, was found as clumps on the cell surface, and was distributed within the endoneurial space. These changes represent the only ultrastructural evidence of endothelial cell damage and altered permeability in response to increased rate of perfusion.

AC impedance of the perineurium of the frog sciatic nerve

The AC impedance of the isolated perineurium of the frog sciatic nerve was examined at frequencies from 2 Hz to 100 kHz. A Nyquist plot of the imaginary and real components of the impedance demonstrated more than 1 capacitative element, and a DC resistance of 478 +/- 34 (SEM, n = 27) omega cm2. Transperineurial potential in the absence of externally applied current was 0.0 +/- 0.5 mV. The impedance data were fitted by nonlinear least squares to an equation representing the generalized impedance of four equivalent circuits each with two resistive and two capacitative elements. Only two of these circuits were consistent with perineurial morphology, however. In both, the perineurial cells were represented by a resistive and capacitative element in parallel, where capacitance was less than 0.1 microF/cm2. The extracellular matrix and intercellular junctions of the perineurium were represented as single resistive and capacitative elements in parallel or in series, where capacitance exceeded 2 microF/cm2. Immersion of the perineurium in low conductance Ringer's solution increased DC resistive elements as compared with their values in isotonic Ringer's solution, whereas treatment for 10 min with a hypertonic Ringer's solution (containing an additional 1.0 or 2.0 mol NaCl/liter of solution) reduced DC resistive elements, consistent with changes in perineurial permeability. The results indicate that (a) perineurial impedance contains two time constants and can be analyzed in terms of contributions from cellular and extracellular elements, and (b) transperineurial DC resistance, which is intermediate between DC resistance for leaky and nonleaky epithelia, represents intercellular resistance and can be experimentally modified by hypertonicity.

Conduction velocity varies with osmotically induced changes of the area of the axon's profile

The conduction velocity of frog ischiadic nerves incubated in vitro in osmolarities between 220 and 1000 mOsm decreased with the degree of fiber shrinkage. The latter (non-circularity factor) was determined from computer-assisted measurements in freeze-substituted or in chemically fixed fibers. Freeze-substituted normal nerves had a non-circularity factor of 0.91 for fibers of all calibers, which likely reflects the in vivo state of the fiber population. Chemically fixed nerves had a non-circularity factor near 0.68, consistent with previous data. Non-circularity factors decreased with increasing osmolarities of the media, regardless of the type of tissue preparation. Conduction velocity decreased with decreasing non-circularity. Restoration of the nerves to normotonic media increased conduction velocity. The rates of change were accelerated in nerves chemically desheathed with Triton. The decrease in the conduction velocity in osmotically shrunken nerves did not correspond to changes in the absolute refractory period for the propagation of the impulse, used as a sensitive index of non-specific damage. These experimental observations corroborate data from computer simulation of relative sensitivities of nodal and internodal parameters.