Multiple innervation of rat pacinian corpuscles regenerated after neonatal axotomy

Demise of nociceptive Schwann cells causes nerve retraction and pain hyperalgesia

ABSTRACT

Recent findings indicate that nociceptive nerves are not "free", but similar to touch and pressure sensitive nerves, terminate in an end-organ in mice. This sensory structure consists of the nociceptive nerves and specialized nociceptive Schwann cells forming a mesh-like organ in subepidermis with pain transduction initiated at both these cellular constituents. The intimate relation of nociceptive nerves with nociceptive Schwann cells in mice raises the question whether defects in nociceptive Schwann cells can by itself contribute to pain hyperalgesia, nerve retraction, and peripheral neuropathy. We therefore examined the existence of nociceptive Schwann cells in human skin and their possible contribution to neuropathy and pain hyperalgesia in mouse models. Similar to mouse, human skin contains SOX10+/S100B+/AQP1+ Schwann cells in the subepidermal border that have extensive processes, which are intimately associated with nociceptive nerves projecting into epidermis. The ablation of nociceptive Schwann cells in mice resulted in nerve retraction and mechanical, cold, and heat hyperalgesia. Conversely, ablating the nociceptive nerves led to a retraction of epidermal Schwann cell processes, changes in nociceptive Schwann cell soma morphology, heat analgesia, and mechanical hyperalgesia. Our results provide evidence for a nociceptive sensory end-organ in the human skin and using animal models highlight the interdependence of the nerve and the nociceptive Schwann cell. Finally, we show that demise of nociceptive Schwann cells is sufficient to cause neuropathic-like pain in the mouse.

Computational Parametric Analysis of the Mechanical Response of Structurally Varying Pacinian Corpuscles

The Pacinian corpuscle (PC) is a cutaneous mechanoreceptor that senses low-amplitude, high-frequency vibrations. The PC contains a nerve fiber surrounded by alternating layers of solid lamellae and interlamellar fluid, and this structure is hypothesized to contribute to the PC's role as a band-pass filter for vibrations. In this study, we sought to evaluate the relationship between the PC's material and geometric parameters and its response to vibration. We used a spherical finite element mechanical model based on shell theory and lubrication theory to model the PC's outer core. Specifically, we analyzed the effect of the following structural properties on the PC's frequency sensitivity: lamellar modulus (E), lamellar thickness (h), fluid viscosity (μ), PC outer radius (Ro), and number of lamellae (N). The frequency of peak strain amplification (henceforth "peak frequency") and frequency range over which strain amplification occurred (henceforth "bandwidth") increased with lamellar modulus or lamellar thickness and decreased with an increase in fluid viscosity or radius. All five structural parameters were combined into expressions for the relationship between the parameters and peak frequency, ωpeak=1.605×10-6N3.475(Eh/μRo), or bandwidth, B=1.747×10-6N3.951(Eh/μRo). Although further work is needed to understand how mechanical variability contributes to functional variability in PCs and how factors such as PC eccentricity also affect PC behavior, this study provides two simple expressions that can be used to predict the impact of structural or material changes with aging or disease on the frequency response of the PC.

Developing a sense of touch

The sensation of touch is mediated by mechanosensory neurons that are embedded in skin and relay signals from the periphery to the central nervous system. During embryogenesis, axons elongate from these neurons to make contact with the developing skin. Concurrently, the epithelium of skin transforms from a homogeneous tissue into a heterogeneous organ that is made up of distinct layers and microdomains. Throughout this process, each neuronal terminal must form connections with an appropriate skin region to serve its function. This Review presents current knowledge of the development of the sensory microdomains in mammalian skin and the mechanosensory neurons that innervate them.

Ultrastructures of mechanoreceptors in the oral mucosa

The present review describes the fine structures of lamellated mechanoreceptive corpuscles, Merkel cell-neurite complexes and free nerve endings in the oral mucosae of mammals, with special attention to axon terminals and lamellar cells. The mechanoreceptive nerve endings of the oral mucosa were studied using histochemistry, immunohistochemistry and transmission electron microscopy techniques. The organized mechanoreceptive corpuscles are present in the mucosae of gingiva, cheek, tongue and soft and hard palate. They are elongated or globular in shape, being located in the connective tissue papillae. The capsule is composed of several layers of cytoplasmic extensions of perineural cells. Numerous bundles of collagen fibers are noted at the periphery of the corpuscle. The lamellated corpuscles are surrounded by several layers of superimposed flattened capsular cell processes. The interlamellar spaces are 0.2-0.4 micron in width and filled with thin fibrillar collagen fibers embedded in the amorphous substance. The lamellar cells contain rich microtubules and are characterized by the presence of caveolae on the surface plasma membrane. The terminal axon contains an abundance of mitochondria and small clear vesicles (20-50 nm in diameter). There are neurofilaments in the center of the axon terminal. Intermediate-type junctions are seen between the adjacent lamellar cells and between the axon and adjacent lamellae. The free nerve endings are found in the subepithelial regions, very close to the basal laminae of mucosal epithelium. They are surrounded by a thin cytoplasm of Schwann cells. Sometimes Schwann cell basal larinae become multilayered. Merkel cells are present within the basal layer of mucosal epithelium and contain characteristic electron-dense granules that are located almost exclusively at the side of cytoplasm in contact with axon terminals. Intermediate-type junctions are noted between axon terminals and Merkel cells.

Glial growth factor rescues Schwann cells of mechanoreceptors from denervation-induced apoptosis

Golgi tendon organs and Pacinian corpuscles are peripheral mechanoreceptors that disappear after denervation during a critical period in early postnatal development. Even if regeneration is allowed to occur, Golgi tendon organs do not reform, and the reformation of Pacinian corpuscles is greatly impaired. The sensory nerve terminals of both types of mechanoreceptors are closely associated with Schwann cells. Here we investigate the changes in the Schwann cells found in Golgi tendon organs and Pacinian corpuscles after nerve resection in the early neonatal period. We report that denervation induces the apoptotic death of these Schwann cells and that this apoptosis can be prevented by administration of a soluble form of neuregulin, glial growth factor 2. Schwann cells associated with these mechanoreceptors are immunoreactive for the neuregulin receptors erbB2, erbB3, and erbB4, and the sensory nerve terminals are immunoreactive for neuregulin. Our results suggest that Schwann cells in developing sensory end organs are trophically dependent on sensory axon terminals and that an axon-derived neuregulin mediates this trophic interaction. The denervation-induced death of mechanoreceptor Schwann cells is correlated with deficiencies in the re-establishment of these sensory end organs by regenerating axons.

Reinnervation of rat Pacinian corpuscles after nerve crush during the postcritical period of development

The ultrastructure of crural Pacinian corpuscles was examined after sciatic nerve crush performed in 7- to 20-day-old rats, i.e. during the postcritical period of development when the corpuscles no longer degenerate after axotomy but cease growing. The aim of our study was to assess the innervation pattern and structural changes of the corpuscles following transient denervation and subsequent reinnervation during their maturation and growth. Reinnervated corpuscles were examined by electron microscopy from 2.5 months after nerve crush onwards. After sciatic nerve crush at 7 days of age, the corpuscles are mostly reinnervated with multiple axon terminals, each of them enclosed within a newly formed inner core. The axial multiple cores are in part covered by a layer of concentric inner core lamellae and surrounded by a capsule, both structures having survived from the original corpuscle. After nerve crush at 10 days of age, reinnervated Pacinian corpuscles usually contain, in their axial region, a denervated remainder of the original core together with a few regenerated axon terminals enclosed within new inner cores. These axial structures are surrounded by a layer of concentric lamellae of the original core which may accommodate some regenerated terminals. Additional axon terminals with their small inner cores may be found at the outer aspect of the composite core beneath the capsule. When the nerve is crushed in 15-day-old rats, the inner core which is already well developed remains preserved by the time of reinnervation, and regenerating axons grow in between the original lamellae inducing only moderate neoformation of 2-3 lamellar layers which enclose the terminals. After crushing the sciatic nerve in 20-day-old rats, formation of new inner core lamellae is minimal and regenerated terminals become accommodated between the original lamellar of the core as is the case in adult animals. Regeneration of new inner cores and reinnervation of the preserved lamellar structure thus characterize the recovery of Pacinian corpuscles following reinnervation after nerve crush during the postcritical period of their development.

Effacement and regeneration of tactile lamellar corpuscles of rat after postnatal nerve crush

The development of Meissner-like lamellar corpuscles was studied in rat toe pads under normal conditions and after crushing the sciatic nerve in 1- to 15-day-old animals. During normal development, rat lamellar corpuscles begin to differentiate first by postnatal day 8. By this time, sensory axons have grown up to the apex of dermal papillae and form axon terminals beneath epidermis. The terminals are ensheathed by lamellar cells derived from Schwann cells. First thin lamellae are formed around the terminals 8-12 days after birth, and the number of lamellar layers increases until the corpuscles become structurally mature by 20 days after birth. A mature corpuscle consists of two or more terminals, each surrounded by approximately 10 lamellae, all components being enclosed by an incomplete capsule. No lamellar corpuscles develop in toe pads after crushing the sciatic nerve in newborn rats, and only occasional corpuscles regenerate after nerve crush at 5 days of age. The corpuscles fail to develop because dermal papillae remain permanently denervated after crushing the nerve early postnatally. After nerve crush in 10-day-old rats, lamellar corpuscles regenerate by 1 month after the operation, but they remain underdeveloped: their number and size are smaller than normal even 1 year after injury, and their terminals are encircled only by 1-3 lamellar layers. After nerve crush in 15-day-old rats, the corpuscles recover upon reinnervation and their size and lamellation become almost normal.

Immunocytochemical localization of the neural cell adhesion molecules L1, N-CAM, and J1 in Pacinian corpuscles of the mouse during development, in the adult and during regeneration

The immunocytochemical localization of the neural cell adhesion molecules L1, N-CAM and J1/tenascin was investigated by light and electron microscopical techniques in murine Pacinian corpuscles during development, in the adult and in the regenerating state. In adult corpuscles, L1 was present only at contact sites between the sensory axon and inner core lamellae. From birth, the earliest stage tested, until day 7, L1 was additionally expressed on lamellar processes of the inner core cells. N-CAM was expressed in developing and adult corpuscles on lamellae and somata of the inner and outer core cells at their contact sites but was hardly detectable at contact sites between axolemma and inner core lamellae. J1/tenascin was found only in association with the extracellular material of the inner core, especially with the two radial clefts and the boundary space between inner and outer core. In developing corpuscles, J1/tenascin became detectable on extracellular material with the onset of inner core differentiation at approximately day 2. After transection or crush of the sciatic nerve, L1 disappeared from the corpuscles but reappeared with regrowing axons at contact sites between axonal membranes and inner core cells. At any regenerative stage inner core cells remained L1-negative. In denervated and reinnervated corpuscles the expression pattern of N-CAM and J1/tenascin did not differ from the normal adult. These observations suggest that a sensory organ, the Pacinian corpuscle, differs from the sciatic nerve and the neuromuscular junction in that its expression of adhesion molecules remains the same in the denervated state as in the innervated adult. Furthermore, in the denervated Pacinian corpuscle, adhesion molecule expression does not resemble that of any developmental stage tested. Thus, other cures than regulation of adhesion molecule expression patterns might be involved in the successful reinnervation of sensory corpuscles.

The ultrastructure of sensory nerve endings in human anterior cruciate ligament

The sensory innervation of the anterior cruciate ligament (ligamentum cruciatum anterius) of the human knee joint was studied by light- and electron microscopy. The connective tissue between the synovial membrane and the cruciate ligament contains small Ruffini corpuscles and lamellar corpuscles with several inner cores. The connective tissue septa between the individual fascicles of the cruciate ligament contain Ruffini corpuscles and free nerve endings. The free nerve endings are innervated by C-fibres and myelinated A-delta fibres. The afferent axons of Ruffini corpuscles are myelinated and measure 4-6 microns in diameter, those of the lamellar corpuscles with several inner cores measure about 6 microns in diameter. It is discussed, whether these receptors of the anterior cruciate ligament may influence the muscle tone via polysynaptic reflexes.

Grafts of pacinian corpuscles reinnervated by dorsal root axons

In adult rats, a piece of the crural interosseous nerve with several Pacinian corpuscles attached was removed from the crural region, autotransplanted onto the surface of the lumbar spinal cord and connected with the peripheral stump of a transected dorsal root. From 10 days up to 6 months after the operation, the grafts were investigated by light and electron microscopy. The regenerating dorsal root axons grew along the grafted nerves into the attached Pacinian corpuscles. By 1-2 months after the operation, the nerves and their branches became almost completely reinnervated by myelinated and unmyelinated dorsal root axons. In a sample of corpuscles examined 2-6 months after grafting, 75% of corpuscles were found reinnervated; each of them was supplied by 1-5 large myelinated axons that formed multiple axon terminals in the inner core. The maximal number of axonal profiles found in a transverse section through different levels of the inner core varied, in individual corpuscles, from 3 to 17 axons and terminals. The dorsal root terminals formed in the grafted corpuscles were mainly filled with mitochondria and resembled peripheral sensory endings. In some instances, the newly formed endings developed lateral processes and membrane specializations characteristic for peripheral Pacinian terminals. Thus regenerating dorsal root axons recognize a grafted peripheral mechanoreceptor as their target and reinnervate it with axon terminals, most of them structurally transformed into peripheral sensory endings.

Properties of cutaneous mechanosensitive afferents during the early stages of regeneration in man

The technique of percutaneous microneurography was used to record single unit activity from 75 regenerated primary afferents innervating the glabrous skin of the human hand. Thirteen patients were studied, who had suffered complete transection, with subsequent suture or graft, of the median or ulnar nerves. The recordings were obtained from 7 to 23 months postoperatively (early regeneration). Three types of mechanoreceptive afferents (RA, SAI, SAII) and many deep units of unknown origin were found. No regenerated PC units could be identified. The reinnervated receptors were predominantly located in the palm and proximal fingers, comparable to those found 3 years or more postoperatively (late regeneration). Response thresholds and in general, discharge and receptive field characteristics of the majority of afferents were largely comparable to late regeneration and normal. The properties of SAII units were like normal in all respects. However, several distinct abnormalities were encountered early during regeneration: multiple receptive fields innervated by a single afferent (2/9 RA and 2/9 SAI), unusually small or large receptive fields (RA and SAI), pronounced fatigue on repetitive stimulation (7/15 SAI, 4/6 deep). Responses of reinnervated skin to sustained and repeated indentations were found to be similar to those of normal skin, and therefore, could not account for the abnormal discharge behavior. It is suggested that the transitional properties of regenerating afferents reflect unstable axon-end organ connections and immature axonal properties. Both factors would contribute to the slow course of sensory recovery, making prognosis on tactile recovery unpredictable.

Multiple axon terminals in reinnervated Pacinian corpuscles of adult rat

The ultrastructure of Pacinian corpuscles localized beneath the crural interosseous membrane was examined two weeks to 18 months after crushing the sciatic nerve in adult rats. The Pacinian inner core and capsule remained preserved during the transient period of denervation. Regenerating axons reached Pacinian corpuscles approximately three weeks after nerve crush. Up to 15 axonal sprouts entered a single corpuscle at the initial stage of reinnervation, but only 1-3 axons increased in size, myelinated and formed axon terminals in the inner core, the excess sprouts being eliminated. Most corpuscles of the crural group were reinnervated by the end of the first month. Three to 19 months after nerve crush, 10% of corpuscles examined were found to be monoaxonal and monoterminal as before the operation; 74% contained multiple terminals; 16% remained denervated. Over half the multiterminal corpuscles were supplied with a single myelinated axon that branched inside the corpuscles; the rest received two or three myelinated axons which formed several terminals. The terminals were distributed at random, usually in the axial region between the lamellae of the inner core. They were cylindrical, with an oval profile; the larger terminals were filled with mitochondria and microtubules at their circumference and contained a core of neurofilaments. Lateral processes of the terminals were filled with vesicles and had membrane specializations as in normal corpuscles. The mean number of terminals in reinnervated corpuscles was 4.07 +/- 0.37 (S.E.M.) at three months, and 3.26 +/- 0.49 (S.E.M.) 6-18 months after nerve crush. This small decrease was apparently the result of degeneration occasionally observed in some axon terminals at later stages of reinnervation. These experiments thus demonstrate that most rat Pacinian corpuscles become reinnervated with multiple terminals after nerve injury and maintain multiterminal innervation permanently.

Reinnervation of mechanoreceptors in the human glabrous skin following peripheral nerve repair

The technique of percutaneous microneurography was used to record single unit activity from 65 reinnervated and 24 normally innervated mechanoreceptors in the glabrous skin of the human hand. The results were obtained from 20 patients and 5 control subjects. The patients had suffered complete traumatic transsection, with subsequent repair, of the median or ulnar nerves. Three types of mechanoreceptors (RA, SAI, SAII) and many unidentified units located in deep tissues were found to become reinnervated. No reinnervated PC units could be identified. Response thresholds, discharge characteristics and receptive field properties of reinnervated receptors were comparable to normal, with the exception that reinnervated SA I units had slower static discharge rates and smaller receptive fields. No evidence was found for multiple peripheral innervation by a single afferent fiber. The reinnervated mechanoreceptors were predominantly located in the palm and the proximal fingers with few in the finger tips, contrary to normal. The locations and frequency of occurrence of the different types of receptors could be correlated with the goodness of sensory recovery. It is suggested that these differences result from misguidance of regenerating fibers and from poor reinnervation, and that they account for reduced sensitivity and poor tactile discrimination in patients with peripheral nerve injuries.

Sensory electroneurographic parameters and clinical recovery of sensibility in sutured human nerves

A total of 37 patients with traumatic transection of median or ulnar nerves at the wrist (total 41 nerves) were examined clinically and electrophysiologically 4-59 months after primary or secondary suture or grafting. There was a significant increase of cumulative amplitude with the time after suture, whereas maximum sensory nerve conduction velocity and maximum amplitude of nerve action potentials did not reveal such a correlation. The recovery of two-point discrimination, vibration threshold and sensibility scored according to the scale of Nicholson and Seddon were also not related to the passage of time after operation. Though there were significant correlations between cumulative amplitude and both two-point discrimination and recovery of sensibility, electrophysiological parameters were shown to be inadequate predictors of clinical recovery.