Search

Choroid plexus ependymal cells (CPECs) were known to promote axonal growth when choroid plexus is grafted into the adult rat spinal cord. The present study was carried out to examine whether CPECs promote axonal outgrowth from neurons derived from the CNS in vitro. Hippocampal neurons were cocultured on CPEC monolayers. After 24 h, neurite extension was evaluated using various parameters in comparison with cultures grown on poly-L-lysine (PLL)-coated plates and cocultures grown on astrocyte monolayers. The primary neurite length and total neurite length were longest in the cocultures with CPECs. The number of primary neurites and the number of branches were larger in the cultures with CPECs than in the cultures on PLL-coated plates, but almost the same as in the cocultures with astrocytes. Next, we examined whether the neurite extension-promoting effect occurring within 24 h is due primarily to contact with the CPECs or to factors secreted by CPECs into the culture medium. The CPEC monolayers were killed by ethanol fixation, and neurons cultured on them. The neurons extended long neurites with elaborate branching, as in the case of cocultures grown on living CPECs. On the other hand, CPEC-conditioned medium exhibited less promoting effect on neurite outgrowth from hippocampal neurons. These results indicate that CPECs have a capacity to promote neurite outgrowth from CNS neurons in vitro , and that surface plasma membrane-bound components of CPECs strongly contribute to the enhancement of neurite outgrowth in the present coculture system.

Annulate lamellae (AL) are observed as a stack of specialized parts of the endoplasmic reticulum (ER) in intestinal epithelial cells. At mitosis, the components of the nuclear pore (NP) are known to disassemble into ER or cytoplasm, then are recruited back to the nuclear envelope and AL. To investigate structural contribution of ER to AL and interlamellar relationship of the AL stack, we observed serial ultrathin and thick sections of rat intestinal epithelial cells with conventional and high voltage electron microscopes. We found three distinct types of interlamellar connections of AL at the transitions between AL and their extended rough endoplasmic reticulum (RER). Type 1; several lamellae forming a single AL stack were connected by the reflection or rolling of single ER cisterna. Type 2; the RER often bifurcated into successive AL cisternae. Type 3; several lamellae of a single AL stack were connected by smooth tubular structures. We also found multiple bifurcations from ER to AL on freeze-fracture replicas. Thus, annulate lamellae appeared to be interconnected by three distinct cisternal structures. The presence of these various interlamellar connections suggests that ER would be an essential basis for concentration of integral nuclear pore complex proteins to the peripheral subdomains of ER and for formation of AL.

The high-affinity IgE receptor (FcepsilonRI) on mast cells and basophils consists of a ligand-binding alpha-chain and two kinds of signaling chains, a beta-chain and disulfide-linked homodimeric gamma-chains. Crosslinking by multivalent antigen results in the aggregation of the bound IgE/alpha-chain complexes at the cell surface, triggering cell activation, and subsequent internalization through coated pits. However, the precise topographical alterations of the signaling beta- and gamma-chains during stimulation remain unclarified despite their importance in ligand binding/signaling coupling. Here we describe the dynamics of FcepsilonRI subunit distribution in rat basophilic leukemia cells during stimulation as revealed by immunofluorescence and immunogold electron microscopy. Immunolocalization of beta- and gamma-chains was homogeneously distributed on the cell surfaces before stimulation, while crosslinking with multivalent antigen, which elicited optimal degranulation, caused a distinct aggregation of these signaling chains on the cell membrane. Moreover, only gamma- but not beta-chains were aggregated during the stimulation that evoked suboptimal secretion. These findings suggest that high-affinity IgE receptor beta- and gamma-chains do not co-aggregate but for the most part form homogenous aggregates of beta-chains or gamma-chains after crosslinking.

The epithelial cells of the choroid plexus are a continuation of the ventricular ependymal cells and are regarded as modified ependymal cells. The present study was carried out to determine the influence of choroid plexus ependymal cells (CPECs) on axonal growth in vitro. Choroid plexuses were dissected from the fourth ventricle of postnatal day-1–10 mice, mechanically dissociated, and plated in fibronectin-coated culture dishes. CPECs had spread into monolayers with few endothelial cells in 3-week cultures. Some macrophages were scattered on the monolayer of CPECs. Dorsal root ganglia (DRG) were excised from mouse fetuses of 14-day gestation, dissociated with trypsin and cocultured on the CPEC monolayers. For comparison, dissociated DRG neurons were cocultured on astrocyte monolayers or cultured on laminin-coated plates. After 4.5 h culturing, the cultures were fixed and immunohistochemically double-stained for neurites and CPECs using antibodies against β-tubulin III and S-100 β, respectively. It was demonstrated that neurons extended many long neurites with elaborate branching on the surface of S-100-stained CPECs. In contrast, DRG neurons cultured on the astrocytes and on the laminin-coated plates had much shorter primary neurites with fewer branches than those cultured on the CPECs. The total length of neurites including primary neurites and their branches, of a single DRG neuron was 285 ± 14, 395 ± 15 and 565 ± 12 μm on the laminin-coated plates, on astrocytes and on CPECs, respectively. Scanning electron microscopy revealed extension of neurites with well-developed growth cones on the ependymal cells. These results suggest that CPECs have a great capacity to promote neurite outgrowth from DRG neurons in vitro.

Schwann cell basal lamina tubes serve as attractive conduits for regeneration of peripheral nerve axons. In the present study, by using basal lamina tubes prepared by in situ freeze-treatment of rat saphenous nerve, the effects of exogenously applied basic fibroblast growth factor (bFGF) on peripheral nerve regeneration was examined 2 and 5 days after bFGF administration. Regenerating axons were observed by light and electron microscopy using PGP9.5-immunohistochemistry for specific staining of axons. In addition, the localizations of bFGF and its receptor (FGF receptor-1) were examined by immunohistochemistry using anti-bFGF antibody and anti-FGF receptor-1 antibody, respectively. Regenerating axons extended further in the bFGF-administered segment than in the bFGF-untreated control segment. Electron microscopy showed that regenerating axons grew out unaccompanied by Schwann cells. Findings concerning angiogenesis and Schwann cell migration were very similar between the bFGF treated and control nerve segment. bFGF-immunoreactivity was not detected in the control nerve segment. In contrast, bFGF-immunoreactivity was detected on the basal lamina tubes as well as on the plasmalemma of regenerating axons facing the basal lamina in the bFGF treated nerve segment up to 5 days after administration, suggesting that exogenous bFGF can be retained in the basal lamina for several days after administration. FGF receptor was detected on the plasma membrane of regenerating axons where they abutted the basal lamina. These results indicate that bFGF could promote the extension of early regenerating axons by directly influencing the axons, but not via Schwann cells or angiogenesis.

The Ca(2+)-dependent intercellular adhesion molecule cadherin is known to be linked to the cytoskeleton by the protein catenin, an association of which appears to be important for the cell-adhesion function of cadherin. Catenin consists of three subtypes-alpha, beta, and gamma. In our previous study, N-cadherin was shown to be localized on the plasmalemma of normal and regenerating chick peripheral nerve. Thus, as alpha N-catenin is a subtype of alpha-catenin (which is specifically associated with N-cadherin), we investigated the immunolocalization of alpha N-catenin in normal and regenerating chick sciatic nerve. In normal nerve, unmyelinated axons exhibited either intense or weak alpha N-catenin immunoreactivity throughout the axoplasm, whereas myelinated axons were completely immunonegative. Regenerating axons, including those derived from parent myelinated axons, showed alpha N-catenin immunoreactivity of variable intensities in growth cones and axon shafts. Schwann cells were invariably devoid of immunoreactivity. Thus alpha N-catenin is not necessarily bound to the surface plasmalemma, but is distributed throughout the cytoplasm, suggesting that most alpha N-catenin molecules are dissociated from N-cadherin.

Localization of synaptotagmin, a Ca2+-binding protein associated with synaptic vesicles, in regenerating axons was investigated by immunocytochemistry in the injured rat sciatic nerves. The early regenerating axonal sprouts emanating from nodes of Ranvier exhibited synaptotagmin immunoreactivity on vesicles, vacuoles and surface plasma membranes. In the well-developed regenerating sprouts extending through the space between Schwann cell basal lamina and myelin sheath of the parent axon, the growing tips, i.e., typical growth cones, exhibited an intense immunoreaction on vesicles, vacuoles and surface plasma membranes, while the stem regions where the sprouts were continuous with the parent axon exhibited almost no immunoreaction on any organelles including plasma membranes. These findings suggest that synaptotagmin-immunoreactive vesicles and vacuoles might be utilized for the supply of membrane components to the surface plasma membrane in the growth cone. Synaptotagmin which is known to regulate membrane fusion of synaptic vesicles with presynaptic plasma membranes in a Ca2+-dependent manner in synapses may participate in the regulation of the membrane fusion between the vesicles and plasma membranes in growth cones.

The localization of protein kinase C (PKC) α, β and γ subspecies in sensory axon terminals of muscle spindles in the plantar lumbrical muscles of rat was investigated by light and electron microscopic immunocytochemistry using monoclonal and polyclonal antibodies. Immunoreactivity for these subspecies was detected specifically in sensory axon terminals which wound spirally around the intrafusal muscle fibres of the muscle spindle. Immunostaining was found to be stronger with polyclonal than with monoclonal antibodies. By electron microscopy, immunoreactivity for α, β and γ subspecies was almost diffusely distributed in the cytoplasm of the axon terminal, and the overall pattern of distribution of immunoreactivity was similar for all three subspecies. In the cases of a and β subspecies, some intensely immunostained regions were found in the cytoplasm, but no definite subcellular structures corresponding to such regions could be identified. Considering that PKC plays a crucial role in the regulation of ion channels, it is suggested that PKC might be involved in the control of mechanoelectric transduction in sensory axon terminals.

Immunocytochemical demonstration of protein kinase C (PKC) subspecies (α, β, γ) was carried out in Pacinian corpuscles of rat hind feet using monoclonal or polyclonal antibodies against each of these subspecies. The inner core cells and lamellae and the Schwann cell cytoplasm of the nerve fiber innervating the corpuscle were strongly positive for PKC α-immunoreactivity (IR). In contrast, the axon terminal and the outer core did not display any positive α-IR. Very weak PKC β-IR was detected in the ultraterminal region of the axon terminal, while the trunk region showed no immunoreactivity. Very faint PKC β-IR was found also in the lamellar cells located at the periphery of the inner core and the endoneurial fibroblasts in the intermediate layer. PKC γ-IR was not detected in any part of the corpuscle. The strong PKC α-IR in the inner core and the presence of absence of PKC α-, β-, and γ-IR in the axon terminal are discussed from the point of view of the functional aspects of each part.

Traumatic degeneration of myelinated fibers was studied by electron microscopy over 5 days following transection of mouse sciatic nerve. Special attention was paid to the mechanism which separates the degenerating part, while preserving the viable part of the axon. Immediately after transection, the opened end of the proximal stump revealed extensive subcellular changes including the disorganization of neurofilaments, and disruption of mitochondria and axonal endoplasmic reticulum (SER). Subsequently, vesicles of round and tubular profiles filled up the whole area of the stump end, and proximal to it appeared a neurofilament-predominant area characterized by randomly oriented neurofilaments and normally appearing mitochondria and SER. Characteristic membranous demarcations occurred in early periods at the border between the vesicle accumulation and the neurofilament-predominant areas, and later also within these areas. The demarcation membranes formed both by invagination of the surface plasma membrane and, probably, by fusion of the large vesicles. These became prominent with time, dividing the axoplasm into compartments of varying sizes, which gradually underwent degeneration and were liberated from the parent axon. Occurrence of autophagic vacuoles was characteristic of the degenerating portions of the parent axon. Thus, by the function of demarcation membranes, the parent axon to be preserved could remain membrane-bound, while the degenerating parts were shed off.

The organization of collagen fibrils in the rat sciatic nerve was studied by scanning electron microscopy after digestion of cellular elements by sodium hydroxide treatment, and by conventional transmission electron microscopy. The epineurium consisted mainly of thick bundles of collagen fibrils measuring about 10-20 microns in width; they were wavy and ran slightly obliquely to the nerve axis. Between these collagen bundles, a very coarse meshwork of randomly oriented collagen fibrils was present. In the perineurium, collagen fibrils occupied the interspaces between the concentrically arranged perineurial cells; in each interspace, they formed a sheet of characteristic lacework elaborately interwoven by thin (about 3 microns or less in width) bundles of collagen fibrils. In the subperineurial region, there was a distinct sheet of densely woven collagen fibrils between the perineurium and underlying endoneurial fibroblasts. In the endoneurium, collagen fibrils surrounded individual nerve fibers in two layers as scaffolds: the inner layer was made up of a delicate meshwork of very fine collagen fibrils, and the outer one consisted of longitudinally oriented bundles of about 1-3 microns in width. The collagen fibril arrangement described above may protect the nerve fibers against external forces.

The three-dimensional architecture of filamentous components in the lamina densa of the Schwann cell basement membrane was studied in the mouse sciatic nerve by transmission and scanning electron microscopy after osmium-maceration treatment, and also by conventional electron microscopy. In conventionally prepared specimens, the lamina densa of the basement membrane was the most electron-dense, showing up as a felt-like layer 20-30 nm thickness. The interstitial surface of this layer had a spongy appearance with numerous shallow pits. Maceration of the specimens with 0.1% OsO4 for 2-4 days effectively removed amorphous, non-filamentous components from the basement membrane, thus exposing fine filamentous structures embedded in the lamina densa; these were about 10-15 nm in diameter and elaborately interwoven and/or connected with each other to form the framework of the lamina densa. Occasionally, some of them appeared to twine around adjoining collagen fibrils. The nature of these filamentous structures is discussed in terms of the chemical components of the basement membrane.

Recent studies have suggested that the soluble N-ethylmaleimide-sensitive factor attached protein (SNAP) receptor (SNARE)-mediated membrane fusion system is involved in vesicle fusion with the surface plasma membrane, which leads to neurite elongation. There have been several reports analyzing the effects of neurite outgrowth by inhibition of SNAREs. We studied this mechanism by overexpressing GFP-fusion SNAREs including VAMP-2, SNAP-25A, and syntaxin1A in PC12 cells to investigate the role of SNAREs in neurite outgrowth. When overexpressed in PC12 cells, VAMP-2 promoted neurite elongation, whereas SNAP-25A stimulated neurite sprouting. On the other hand, overexpression of syntaxin1A neither promoted nor inhibited neurite outgrowth. Thus, VAMP-2 and SNAP-25A play different roles in neurite elongation and sprouting.

The basal laminae of muscle fibers, when treated by denaturing methods including freeze thawing, have been used as conduits for regenerating nerves. In this study, we developed a new method for denaturing skeletal muscle fibers through treatment with a biological detergent, sodium dodecyl sulfate. Laminin and type IV collagen proteins of muscle fiber basal laminae were preserved after the detergent treatment. A segment of detergent-denatured muscle was grafted to a 1-cm defect of the rat sciatic nerve. One week after grafting, regenerating axons immunostained for neurofilaments were seen extending within laminin-positive muscle fiber basal lamina tubes. Four weeks after grafting, numerous myelinated axons at a much higher level than the control unoperated sciatic nerve, were found in the middle of the graft. They were smaller in diameter than those in the control nerve. Distal host nerves were well reinnervated 4 weeks after grafting. These findings suggest that the basal laminae of detergent-denatured muscle fibers provide effective conduits for regenerating axons.

It has been reported that in the chick embryonic retina, N-cadherin first appears at the very early stages and is subsequently substituted by R-cadherin at the middle to late stages of development. To examine the role of R-cadherin in the morphogenesis of chick retinal pigment epithelium (RPE), the distribution of this adhesion molecule was studied by immunofluorescence cytochemistry and immunoelectron microscopy from embryonic day (E) 6 to hatching. R-cadherin immunoreactivity was detected at E6, and was strongest at E12–13. During these stages, R-cadherin was expressed uniformly on the lateral plasma membranes of RPE cells in contact with each other. Thereafter, R-cadherin immunoreactivity was markedly decreased, with intense immunoreactivity restricted to zonulae adherentes in latero-apical regions at E16. R-cadherin immunoreactivity was no longer detectable in the newly hatched chick RPE, even though morphologically well developed zonulae adherentes were present in latero-apical regions. No immunoreactivity was detected on the apical side facing the neural retina or on the basal side facing the basal lamina at any stage of development. These findings indicate that R-cadherin plays an important role as a major cadherin subtype in the morphogenesis of chick embryo RPE, and is involved initially in non-specific cell-cell adhesions, and subsequently in the formation and maintenance of developing zonulae adherentes.

The localization of synapsin I, a synaptic vesicle-associated protein, was investigated immunocyto-chemically in normal nerve fibers and regenerating axonal sprouts following crush-injuries to the rat sciatic nerve. In normal myelinated axons, weak synapsin I immunoreactivity was found in the axoplasmic/smooth endoplasmic domains, but not in the cytoskeletal domains comprising neurofilaments and microtubules. In non-myelinated axons without dense cytoskeletal structures, moderate immunoreactivity was distributed diffusely throughout the axoplasm. In the crush-injured nerves, intense synapsin I immunoreactivity was demonstrated by light microscopy in early regenerating sprouts emerging from nodes of Ranvier. These nodal sprouts subsequently elongated as regenerating axons through the space between the basal lamina and the myelin sheath (or Schwann cell plasma membrane). Intense synapsin I immunoreactivity was also found in the growth cones of such long regenerating axons. Electron microscopy revealed that synapsin I immunoreactivity was associated mainly with vesicular organelles in the nodal sprouts and growth cones of regenerating axons. Long regenerating axons exhibited no synapsin I immunoreactivity in the shaft, which contained an abundance of neurofilaments. However, vesicle accumulations remaining in the periphery of the shaft still exhibited intense synapsin I immunoreactivity. Thus, it can be concluded that synapsin I is localized at especially high density in the domains comprising vesicular organelles, which are characteristic of early nodal sprouts, as well as in growth cones of regenerating axons. These findings, together with the proposed functions of synapsin I investigated in other studies, suggest that synapsin I may play important roles in vesicular dynamics including the translocation of vesicles to the plasma membrane in sprouts and growth cones of regenerating axons.

Normal intact and regenerating axon terminals up to 30 days after nerve crushing were studied in the rat flexor carpi ulnaris muscle by confocal laser scanning microscopy (CLSM) and electron microscopy using immunohistochemistry for protein gene product 9.5 (PGP 9.5). The motor axons were intensely and homogeneously stained along their entire length. "Three dimensional" organizations of elaborate axon terminals were clearly demonstrated by reconstructing serial optical images obtained by CLSM in normal neuromuscular junctions. In the injured nerve, the earliest regenerating axons could be identified at endplate regions six days after nerve crushing as intensely immunoreactive thin processes which bifurcated in T-shape and formed delicate lace-like terminals. Such lace-like terminals were composed of fine thread-like portions 0.2-0.8 microm in diameter and expanded portions of 2-3 microm in diameter. Electron microscopy revealed that all the axon terminals in the cytoplasm were stained almost homogeneously by PGP 9.5 immunohistochemistry up to their extreme tips. Axon terminals were in direct contact with the basal lamina of the postsynaptic folds, and showed occasional branching. The thin thread-like portions contained no mitochondria but only a few vesicles, where as the expanded portions, abundant mitochondria. And preterminal axons and some expanded portions were abutted by Schwann cells, while thin thread-like portions were exposed with no association with Schwann cells. Twenty to 30 days after crushing injury, regenerating motor axon terminals resumed their mature form in terms of branching elaborations and ultrastructural features. Thus, CLSM of PGP 9.5 immunocytochemistry combined with electron microscopy was able to demonstrate the "three-dimensional" organization of elaborate axon terminals at high resolution in the normal and regenerating neuromuscular junctions. Using this technique, extremely fine processes of axonal terminals were identifiable at the earliest stage of reinnervation.

The role of basal laminae as conduits for regenerating axons in an allogeneic graft was examined by transplanting a 3 cm long segment of the sciatic nerve from the Brown Norway to the Fischer 344 strain of rat. These strains are not histocompatible with each other. In order to compare the nerve regeneration in variously treated grafts, three different types of graft were employed: non-treated (NT), predenervated (PD), and predenervated plus freeze-treated (PDC) grafts. The cytology of nerve regeneration through these grafts was examined by electron microscopy at four, seven, 14, 30 and 60 days after grafting. In the PDC graft, in which Schwann cells were dead on grafting, basal laminae were well preserved in the form of tubes after Schwann cells and myelin sheaths had been removed at seven days after grafting. Regenerating axons accompanied by immature host Schwann cells grew out through such basal lamina tubes in the same fashion as observed in our previous studies. By day 14, axons extended as far as the middle of the graft. In the proximal part they were separated into individual fibres and even thinly myelinated by Schwann cells. On the other hand, in the NT and PD grafts in which Schwann cells were alive on grafting, most Schwann cells and myelin sheaths appeared to undergo autolytic degeneration by day 14, while Schwann cell basal laminae were left almost intact in the form of tubes. A few regenerating axons were seen associated with Schwann cells in the proximal portion by day seven. It is probable that host Schwann cells moved into the graft after donor cells had been degraded. Schwann cell basal laminae tended to be damaged at the site of extensive lymphoid cell infiltration. By day 30, regenerating axons had arrived at the distal end of the graft in all three types of graft: in the PDC graft thick axons were fully myelinated, whereas in the PD graft they were only occasionally myelinated and in the NT graft most axons were still surrounded by common Schwann cells. By 60 days after grafting, regenerating axons were well myelinated in the host nerve as observed 1 cm distal to the apposition site in all the three types of graft. These findings show that Schwann cell basal laminae can serve as pathways (most efficiently in the PDC graft) for regenerating axons in a 3 cm long allograft in the rat.

Cholinesterase (ChE) activity was ultracytochemically demonstrated in mouse digital (Meissner-like) corpuscles by Karnovsky's or Koelle-Friedenwald's method. Intense enzyme activity was found in the caveolae of lamellar cell processes, in the interlamellar spaces and in the narrow perineural spaces between the nerve endings and the apposing lamellar processes. Prominent enzyme activity was also found in the cistern of the rough endoplasmic reticulum as well as in the nuclear envelope of the lamellar cell body. These findings suggest that the enzyme is synthesized in the lamellar cells and released into the interlamellar and perineural spaces in the corpuscle. The enzyme activity was inhibited by eserine 10-4M, DFP 10-5M, or iso-OMPA 10-4M, but not by BW284C51 at a concentration lower than 10-4M. This result indicates that the enzyme activity studied is not that of acetylcholinesterase (AChE) but of nonspecific cholinesterase.

Schwann cell basal laminae of mouse sciatic nerve were studied histochemically using tannic acid (TA) and ruthenium red (RR) in tissue sections. In addition, basal laminae isolated by sonication were examined using ferritin-conjugated lectin for carbohydrate residues and by immunohistochemistry for laminin. Isolated basal laminae were also examined by RR. The outer (interstitial) surface of the basal laminae were observed by scanning electron microscopy (SEM).By TA staining, the basic lattice-like structure of lamina densa was revealed, and fine and thick filamentous fuzzy structures were seen to project from both sides of this lattice.By RR staining, there were dark-stained spots of proteoglycans, arranged at 30-100 nm intervals, on the outer surface of the basal lamina. These spots are considered to be primarily responsible for the attachment of basal lamina to the surrounding connective tissue matrix.Out of 12 lectins examined, WGA, RCA-I and ConA bound to the isolated basal laminae: WGA tended to bind to the inner surface, while the other two lectins bound almost equally to both sides.Laminin appeared to be more abundant on the inner surface than on the outersurface of the basal lamina.SEM revealed a somewhat fuzzy but unexpectedly smooth outer surface of basal lamina. There were small pits scattered randomly on the basal lamina.These findings are discussed from the point of view of the polarity of the basal lamina, and its role in cell adhesion to the surrounding extracellular matrices.

The development of the lamellar cells of mouse digital corpuscles (Meissner corpuscle) was studied by light and electron microscopic histochemistry for cholinesterase (ChE). The materials used were the hind limbs taken from fetuses at 14, 17 and 20 days of gestation, and from young mice at 1, 5, 7, 15 and 20 days after birth. Embryonal Schwann cells had non-specific ChE activity in the cisternae of the rough endoplasmic reticulum and the nuclear envelope, suggesting that they had the ability of synthesizing the enzyme. After birth, such ability gradually decreased and by five days of age non-specific ChE activity was no longer demonstrable in Schwann cells at the time when myelin sheath formation began. However, Schwann cells which were associated with the axonal tips penetrating into the epidermis still had an intense non-specific ChE activity. Such Schwann cells surrounded the axons in gradually increasing numbers of cytoplasmic processes, which later became the lamellae around the axon terminals; thus by 20 days after birth they had differentiated into mature lamellar cells of Meissner corpuscles. These lamellar cells had, as in the embryonal Schwann cells, an intense ChE activity in the cytoplasm. These findings indicate that the lamellar cell is a specialized form of Schwann cell which still retains the embryonal characteristics for synthesizing non-specific ChE.

Acetylcholinesterase (AChE) activity in the developing chick retina was histochemically studied by electron microscopy. The enzyme activity in the 1-day-old chick retina was confined within the ganglion, amacrine and horizontal cells as well as in the inner plexiform layer. In the inner plexiform layer reaction product was found in the intercellular spaces between nerve processes. Synapses in this layer did not always display reaction product. In the developing chick retina, AChE activity was found in the innermost layer of 6-day-old retina in ovo. The stage of the appearance of histochemically demonstrable AChE activity in the developing ganglion and amacrine cells mostly coincided with that of the beginning of morphological maturity of these cells. These findings indicate that the ganglion, amacrine and horizontal cells are functionally different from other types of neurons, and that synthesis of the enzyme starts at almost the same time as those cells begin to manifest the morphological characteristics of each cell type.

The presence of macrophages in the outer bulb region of mouse, monkey and human Pacinian corpuscles was demonstrated by light and electron microscopy. In the normal, nontreated, Pacinian corpuscles, a few particular cells were located in the spaces between lamellae of the outer bulb. These cells contained numerous vesicles and vacuoles, and various cytoplasmic processes. When horseradish peroxidase (HRP) was injected locally or systemically, many HRP-positive cells, which were considered to be similar to the particular cells described above, were found in the outer bulb region of the corpuscles. Electron microscopy revealed that these cells contained HRP in vesicles and vacuoles, suggesting that they were macrophages vigorously taking up exogenous HRP. Macrophages in the Pacinian corpuscles are considered to work as scavengers to keep the inner environment of the corpuscles clear and constant with regard to its macromolecular content.

The purpose of this study was to examine whether the basal laminae of Schwann cells in allografts could survive immunological rejection and serve as a conduit for regenerating nerves, as in the case of autogenic nerve grafts. Allografts of nerves were carried out using sciatic nerves of mice after the grafts had been repetitively frozen to kill their Schwann cells. Two mouse strains, C57BL/6N and C3H/HeN, were used, as they are known to differ in major histocompatibility complex. The mid-portion of the grafted nerve segments was examined by electron microscopy. In addition, the toe pad skin and lumbrical muscles were examined for determining whether regenerating nerves reinnervate sensory end organs and motor endplates. The process of nerve regeneration in the allograft was the same as that seen in the autograft. Cells in the graft disintegrated into cell debris and were phagocytized by macrophages, whereas the basal laminae of Schwann cells were not removed by macrophages, remaining in the form of tubes or scaffolds. Regenerating nerve fibers grew out through such basal lamina scaffolds, keeping in contact with the inner surface. Digital sensory corpuscles and motor endplates of the operated side were well reinnervated. The results indicate that the basal laminae of Schwann cells of the allograft may survive and serve as a conduit for regenerating axons in the same way as in the case of an autograft.

The endoneurium of the mouse sciatic nerve was studied by scanning and transmission electron microscopy to elucidate their three-dimensional architecture. The endoneurium consisted of collagen fibrils, and occasional fibroblasts and blood vessels. Collagen fibrils surrounded individual nerve fibers, forming two distinct layers of connective tissue sheath: the outer one was composed of bundles of longitudinally oriented collagen fibrils and the inner one was of a delicate network of interwoven thin collagen fibrils. These outer and inner layers of collagen fibrils correspond to the two fibrous sheaths known as the sheath of Key and Retzius and the sheath of Plenk and Laidlaw, respectively, revealed on nerve fibers by silver impregnation. Some of the finest collagen fibrils forming the inner network are closely attached to the basal lamina of Schwann cells, suggesting that these fibrils are concerned with the connection between the basal lamina and the inner collagen network. These two layers of collagen fibrils were found on all the nerve fibers, suggesting that they represent a general structure ensheathing the peripheral nerve fibers. Although these layers occur also on unmyelinated nerves, they are most developed on the largest myelinated fibers.

The three-dimensional architecture of the autonomic nerve terminals in the rat exocrine pancreas was investigated by scanning electron microscopy using the HCl-digestion method as well as by transmission electron microscopy. Unmyelinated nerves presumed to be autonomic in nature were found in networks covering the outer layers of arterioles and their capillary extensions, with nerve fibers often leaving the capillaries to surround acini in the interacinar spaces. Schwann cells formed scaffolds for axons of the networks. No other distinct type of cells such as the so-called interstitial cells of Cajal were found to be associated with the formation of the networks. Although nerve fibers of the networks were locally in close association with the walls of the blood vessels or with the bases of acinar cells, no specialized axonal contacts with these tissues were found. However, local swellings, presumably varicosities, were observed by scanning electron microscopy on the surface of nerves. These findings suggest that the networks of unmyelinated nerves represent terminal apparatuses of the autonomic nerves in the pancreas. Schwann cells in the terminal networks were characterized by an unusual abundance of cell organelles. These "terminal Schwann cells" well correspond in location and reticular extension to the "interstitial cells of Cajal" as demonstrated by silver impregnation and vital methylene blue staining. The occurrence of well developed Golgi apparatuses, rough endoplasmic reticulum, and numerous ribosomes suggests that the cells are specialized Schwann cells which most likely require high levels of cellular activity in order to maintain their elaborate cytoplasmic processes extending along the terminal networks, and also to sustain the specific functions of axon terminals.

The localization of laminin and fibronectin was examined on the basal laminae of Schwann cells. Basal laminae from sciatic nerves were isolated by sonication, and the localization of laminin and fibronectin on such isolated basal laminae was studied by immunoferritin histochemistry. Laminin was localized mainly on the cellular side (i. e. the side originally facing the Schwann cell plasma membrane) of the basal laminae. On the other hand, fibronectin was found to be present as aggregates only on the interstitial side (i. e., the side originally facing the endoneurial connective tissue) of the basal laminae. Thus, the locations of laminin and fibronectin were distinctly different. It is presumed that laminin might be involved in the attachment of axons and Schwann cells to the basal laminae, while fibronectin mediates the adhesion of the basal laminae to connective tissue elements, including the collagen fibrils. These findings are discussed from a standpoint of nerve regeneration through the basal laminae scaffolds of Schwann cells.

The three-dimensional morphology of the surface of myelinated nerve fibres in the mouse sciatic nerve was studied by scanning electron microscopy after combined potassium hydroxide treatment and collagenase digestion (to remove the surrounding collagen fibrils and basal laminae from nerve fibres) as well as by transmission electron microscopy. The myelinated nerve fibre appeared as a long cylinder with sporadic annular constrictions corresponding to the nodes of Ranvier. Slight swellings of the surface due to Schwann cell nuclei were usually found at the middle of each internode. The surface of the nerve fibre clearly exhibited a network of bulges, which consisted of longitudinal bands extending from the nuclear swelling to the nodes of Ranvier through the internode, and transverse trabeculae bridging between these longitudinal bands. These bulges on the surface of nerve fibres were the site of the retained Schwann cell cytoplasm external to the myelin lamellae. These cytoplasmic networks on myelinated fibres presumably corresponded to the networks described by Cajal following silver impregnation. In addition, other thin elevations and focal round swellings were also found associated with these longitudinal bands and transverse trabeculae. These networks of Schwann cell cytoplasm are considered to be cytoplasmic channels for nutrition. The two apposing paranodal bulbs of nodes of Ranvier were often asymmetrical in their structure. The networks of the paranodal region were more complicated than those in the internode. The networks of Schwann cell cytoplasm converged into a continuous circumferential collar toward the node, which in turn gave rise to finger-like projections into the nodal gap.

The cholinesterase activity of the Vater-Pacini corpuscle in the periosteum of mouse fibula was histochemically studied by light and electron microscopy. The enzyme activity was confined to the inner core of the corpuscle. It was revealed that the reaction product was found in the perineural spaces between the nerve ending and apposing inner core lamellae, and also in the interlamellar spaces. The reaction product was also found in the nuclear envelope and in the cisterns of the rough endoplasmic reticulum of the inner core cells. The enzyme activity was totally suppressed by eserine 10-4 M and by iso-OMPA 10-4 M. No enzyme activity was demonstrated in the outer core and capsule. The nerve ending showed no distinct enzyme activity. These findings indicate that the mouse Pacini corpuscle has non-specific cholinesteraase activity in its inner core, and that the enzyme may be synthesized in the inner core cells and released into the perineural and interlamellar spaces.

ATPase and alkaline phosphatase (ALPase) activities were histochemically examined in the sensory corpuscle of mouse digital pads at light and electron microscopic levels. ATPase activity was found on the plasma membranes of the lamellar cells. Moderate activity was present in the caveolae, but little activity could be demonstrated in the other portions of the plasma membrane. ALPase activity was found in the caveolae but not on other portions of the plasma membranes of the lamellar cells. ALPase activity was also observed on the nerve terminals. Neither ATPase nor ALPase activity was found on the pre-terminal portions of the nerve fibres. These findings indicate that the caveolae of the lamellar cells may be involved in a transport function and that the plasma membrane of the nerve terminals may have some permeability properties different from those of the pre-terminal portions of nerve fibres.

Cryo-treated nerves whose Schwann cells had been killed by repeated freezing and thawing were xenogenically grafted into sciatic nerves from rats (Wistar, as donor) to mice (ddy strain, as recipient) to examine whether Schwann cell basal lamina tubes of cryo-treated xenogeneic grafts were effective conduits for regenerating axons. For comparison and evaluation of the effectiveness of this technique, experiments using grafts without the cryo-treatment were carried out. Cells in cryo-treated xenografts degraded into cell debris immediately after grafting and then were phagocytized by macrophages. After the cellular components had been removed from the graft, Schwann cell basal laminae remained intact in situ, serving as conduits for the regenerating axons. The process of nerve regeneration was almost the same as that observed in cryo-treated auto- and allografts, except that the regeneration was slightly delayed in the xenogeneic graft. In contrast, an extensive cell infiltration occurred in the non-treated grafts. It appeared that the donors Schwann cells in the graft deteriorated due to immunological reactions and were finally eliminated by macrophages, leaving their basal laminae undamaged in situ. The initiation of nerve regeneration including perineurial sheath formation in non-treated grafts was, therefore, significantly delayed, but once begun, it proceeded in the same manner as in the cryo-treated grafts. These findings strongly indicate that Schwann cell basal laminae can serve as effective pathways for regenerating axons even in the xenograft. Moreover, cryo-treated xenogeneic grafts are more desirable than non-treated ones, since dead Schwann cells in the former can be removed in the early period (4-14 days) from the graft without causing any immunological reaction, thus resulting in the facilitation of nerve regeneration.

Changes in the diameter of collagen fibrils were observed with an electron microscope in the endo- and epineurium of sciatic nerves of mice during development from 12 days of gestation to 5 months after birth. It was noted that endoneurial collagen fibrils appeared in embryonic mice at 15 days of gestation, and at the same time, basal laminae began to appear sporadically on the Schwann cell plasmalemma. No fibroblasts were seen at this developmental stage. Collagen fibrils in the endoneurium remained as thin as they were when they first appeared, being in the narrow range of 250–300 A in diameter, while those in the epineurium became much thicker (400–450 A, 5 months after birth) as is also the case in dermal connective tissues. The present study shows that the endoneurial collagen fibrils were different in their developmental pattern from those of the epineurial or of other connective tissues, lending support to the concept that the endoneurial collagen fibrils are particular in nature, being so-called histological reticular fibers.

The freeze-fracture replication technique was used to study the mechanoreceptive digital corpuscles in toe pads of mice. The axon terminal plasmalemma had intramembranous particles (IMPs) at a density of 2367 +/- 517 microns-2 (mean +/- S.E.M.) in the P-face and 84 +/- 4 microns-2 in the E-face. Particles were 10 +/- 1.8 nm in diameter in the P-face and 10 +/- 1.5 nm (mean +/- S.D.) in the E-face. Particle-rich and particle-free areas were noted in the P-face. The lamellar cell plasmalemma had IMPs at a density of 3359 +/- 224 microns-2 in the P-face and 265 +/- 95 microns-2 in the E-face. Particles were 10 +/- 1.4 nm in diameter in the P-face and 10 +/- 1.6 nm in the E-face. Non-terminal unmyelinated fibres in the connective tissue compartment of toe pads were also examined: the P-faces of the axolemma and Schwann cell plasmalemma had IMPs at a density of 1356 +/- 283 microns-2 and 1514 +/- 514 microns-2, respectively, while the E-face of these membranes had only a few particles. Particles were 9 +/- 1.2 nm and 10 +/- 1.6 nm in diameter in the P-faces of axon and Schwann cell plasmalemmata, respectively. The results show that the IMPs in terminal axolemma and in lamellar cell plasmalemma have a much higher density than those of non-terminal axons or Schwann cells in myelinated and unmyelinated fibres. In addition, IMPs in the terminal axolemma are larger than those in non-terminal axolemma except for the nodal axolemma. It can be said that plasmalemmata of both the axon terminals and lamellar cells of digital corpuscles are specialized in terms of IMPs, suggesting that they have specific physiological properties in mechanoreceptive functions including mechano-electric transduction.

Peculiar myelin figures were abundantly found in some basal-granulated cells including S, D1 and I cells in human Brunner's glands. Intense acid phosphatase activity was found in the periphery of the myelin figures, indicating that they were secondary lysosomes or residual bodies. The acid phosphatase activity was also found in some secretory granules. There were some secretory granules which were partly membranous in content, suggesting the initial stage of their degradation into myelin figures. There were also features indicating the fusion of secretory granules with the myelin figures. All these findings suggest that the myelin figures are the products of lysosomal degradation of secretory granules. The rate of occurrence of basal-granulated cells containing myelin figures in Brunner's glands tended to be higher in subjects with duodenal ulcer than in cases of gastric cancer or ulcer.

Nerve segments approximately 6-7 mm long were excised from the predegenerated sciatic nerves of mice, and treated 5 times by repetitive freezing and thawing to kill the Schwann cells. Such treated nerve segments were grafted into the original place, being in contact with the proximal stump of the sciatic nerve. The animals were sacrificed 2, 3, 5, 7 and 10 days, 2, 3, 5 and 8 weeks after the grafting. The grafts were examined at the middle level, i.e., about 3-4 mm distal to the proximal end of the graft, by light and electron microscopy. Within 2-3 days after the grafting, the dead Schwann cells were disintegrated into fragments and gradually phagocytized by macrophages. However, the basal laminae of the Schwann cells remained as empty tubes (basal lamina scaffolds). The notable finding was that the regenerating axons always grew through these basal lamina scaffolds. New Schwann cells seemed to migrate along these axons from the proximal stumps. The number of axons growing through the basal lamina scaffolds gradually increased with time. These axons were surrounded in a bundle by Schwann cells. About 1 week after the grafting, axons began to be segregated into smaller bundles by Schwann cells. Axons with a relatively large diameter (about 2 microns) tended to be sorted out and surrounded by their own Schwann cells. The myelination began about 2 weeks after the grafting on such large diameter axons. The basal lamina scaffolds, through which the regenerating axons had grown, were gradually disintegrated into fragments by the expansive forces due to the increase in number and volume of the regenerating axons and Schwann cells. Groups of axons, which had been derived from the same basal lamina scaffolds, were enclosed with the cells resembling perineurial epithelial cells. These perineurial epithelial cells proliferated and further separated groups of axons into smaller ones or even into single axons. The number of myelinated axons increased with the advancement of regeneration. These results show that the basal lamina scaffolds of Schwann cells serve as efficient conduits for the elongation, maintenance and maturation of regenerating axons.

Summary Part of the dorsal funiculus of the adult male rat (Wistar) spinal cord was treated for l h at the thoracolumbar level by running hot water, at approximately 48–50 ° C, through a polyethylene tube 2 mm in diameter in contact with the dura. Animals were fixed 1 day to 4 weeks later and the spinal cords were examined by light and electron microscopy. The affected area in the dorsal funiculus was approximately 1 mm long and less than 1 mm wide at the dorsal surface, and varied from 0.4 to 0.7 mm in depth. Within 3 days after treatment, almost all the myelin sheaths in the affected area were degraded, leaving the axons denuded, and at the same time astrocyte endfeet at the glial limiting membrane were swollen and partly destroyed. Almost all the denuded axons remained intact, exhibiting no noticeable morphological changes. There was evidence of a moderate vasogenic oedema, but minimal signs of haemorrhage in the lesion. Seven days after treatment, many immature Schwann cells but no oligodendrocytes were found between the denuded axons. By 2 weeks many of the denuded axons were remyelinated, and by 4 weeks almost all of those axons located near the pial and perivascular surfaces had been remyelinated by Schwann cells, while most of those located in the deep and marginal zones bordering the adjoining intact areas were remyelinated by oligodendrocytes. Longitudinal sections revealed that at nodes of Ranvier PNS-type myelin sheaths were apposed by either intact or newly formed CNS-type myelin sheaths. A typical glial limiting membrane was not reformed beneath the pial surface, but an inconspicuous one was found between the PNS- and CNS-type fibre areas.

The distribution of carbonic anhydrase (CA) activity was studied by electron microscopic histochemistry in rat Pacinian corpuscles, Meissner corpuscles and Merkel cell-neurite complexes using the cobalt bicarbonate method. The distribution of CA activity in these axon terminals was compared to the activity in sciatic nerve axons. An intense enzymatic CA activity was demonstrated in axon terminals of both Pacinian and Meissner corpuscles, while a weak activity was found within the axoplasm of terminals abutting Merkel cells. Some large- and medium-sized axons in sciatic nerves exhibited an intense activity. These findings indicate that large- or medium-diameter sensory axons innervating corpuscular endings have an intense CA activity extending from their somata to their sensory terminals. Axons to Merkel-neurite complexes differ in CA activity from those innervating Meissner and Pacinian corpuscular endings.

The ultrastructure of cat mesenteric Pacinian corpuscles in cross and longitudinal sections has been examined. The terminal ends of lamellar cells of the inner core have been identified in longitudinal sections through the proximal portion of the inner core. These terminal bulbous expansions contain characteristics concentric membranes of rough endoplasmic reticulum and in some cases masses of oval membranous inclusions. The central axon as seen in cross section is oval in profile, having X-(short) and Y-(long) axes, and each axonal face is characterized by specializations of the axolemma. At the X-axis, the inner lamellae of the inner core tightly abut a smooth axolemma, with no intervening connective tissue matrix, in a manner reminiscent of a neuroepithelium. The axolemma of the Y-axis has numerous axonal spines (microspikes) that project into the cleft in the inner core. The extent of the axolemma having axonal spines can only be appreciated in longitudinal sections. The clefts contain a specialized connective tissue with elastic and collagen fibrils. The connective tissue compartment of fibers and matrix separating individual inner core lamellae is unique, in that it contains extremely thin collagen fibrils measuring approximately 15 nm in diameter. The diameter of collagen fibrils increases as the cleft is approached. Here the fibrils resemble typical endoneural collagen.

The present review of cutaneous sensory receptors begins with a consideration of free nerve endings (FNEs) that can be considered as sensory terminals evidencing the least structural specialization of the axon and associated cells. Using the criteria established by KRUGER et al (1981), FNEs of both A delta and C fibers can be identified on the basis of ultrastructural characteristics that include an intimate relationship between axons and the associated epithelium, the lack of a complete Schwann cell investment, the accumulation of numerous vesicles and other cytoplasmic organelles, and for A delta terminals a 1:1 relationship between axon and investing Schwann cell. Using these criteria, the so-called genital end bulbs of the human glans penis are merely a skein of FNEs based on the ultrastructural study of HALATA and MUNGER (1986).Hair follicles of most species studied to date (the exception being the rabbit and to some extent the guinea pig) are multiply innervated with lanceolate, Ruffini and FNEs. The lanceolate terminals are the rapidly adapting terminals that are numerous in guard hairs. Ruffini terminals of hairs resemble those of the periodontal ligament or joint capsules and both are remarkably similar to Golgi tendon organs in terms of ultrastructural characteristics. The key ultrastructural characteristic is the encircling of collagen bundles by axons and associated Schwann and connective tissue cells.Axons frequently enter the epidermis either to terminate as FNEs or become associated with Merkel cells in glabrous skin at the base of the papillary ridges or in clusters of Merkel cells in hairy skin in touch domes or Haarscheiben. Merkel cells have clusters of apparent secretory granules polarized toward the axon and the axon is typically a slowly adapting mechanoreceptor. The function of the granules is not known.Pacinian corpuscles are the largest of the corpuscular receptors of the dermis and are characterized by an elaborate inner core of stacks of numerous thin lamellae arranged in a bilaterally symmetrical manner. Based on the fact that the lamellae are coupled with gap junctions and the outer core lamellae isolated by numerous tight junctions, the authors have proposed that the unique ionic environment may be in part responsible for the remarkable sensitivity of Pacinian corpuscles (MUNGER and IDE, 1987).Meissner corpuscles are a typical corpuscular receptor of murine (IDE, 1976, 1977), marsupial and primate glabrous skin (MUNGER, 1971). The axons typically weave back and forth between stacks of lamellae. Simple corpuscles are common in many mammals especially in the glabrous snout skin (MUNGER, 1965, HALATA and MUNGER, 1983). The lamellae around the axon are typically arranged circumferentially around the axon.The authors have previously noted that axons in mechanoreceptors including lanceolate terminals, Pacinian and Meissner corpuscles, and muscle spindles frequently have axonal spines projecting from the elongated or Y-axis of the axon (MUNGER and IDE, 1987). Axonal spines typically are filled with filaments and lack other cytoplasmic organelles. The common presence of axonal spines provided the basis for the speculation that they may be involved in mechano-electric transduction.

The present report is the second of two studies re-evaluating the cytological characteristics of Pacinian corpuscles. The extreme tip of the axon of a Pacinian corpuscle has been identified and is quite different from the previously described ultraterminal region. The latter is the site where the inner core lamellae begin to terminate and is characterized by a smooth axolemma. The extreme tip lacks inner core lamellae directly abutting the axolemma and is instead characterized by the presence of many axonal spines projecting into a matrix of basal lamina-like material. The extreme tip of the axon thus resembles the organization of the axolemma facing the clefts of the inner core. The axonal spines at the cleft and extreme tip are proposed as a site of restricted current flow due to the tight apposition of inner core lamellae to the axolemma of X-axis. The hemi-inner cores thus could restrict current flow to the cleft. These anatomical specializations could represent both a source and a sink for K+ ions during mechano-electric transduction and account in part for the exquisite sensitivity of Pacinian corpuscles to complex pressure waves.