Studies of the influence of age on regeneration and reinnervation in the peripheral nervous system (PNS) and neuromuscular junction (NMJ) are reviewed, with a particular focus on aged and denervated skeletal muscles. The morphological and functional features of incomplete regeneration and reinnervation are compared between adult and aged animals. In addition, some possible mechanisms of the age-related defects will be discussed. Increased fragmentation or damage in individual components of the NMJ (terminal Schwann cells (TSCs), axon terminals and acetylcholine receptor sites occurs during muscle reinnervation following PNS injury in the aged animals. The capacity to produce ultraterminal sprouting or multiple innervation secondary to PNS injury is maintained, but not the capacity to eliminate such anomalous axonal profiles. The frequency and accuracy of reoccupation of the synaptic sites by TSCs and axon terminals are impaired. Thus, despite the capability of extending neural processes, the rate at which regenerating nerve fibers grow, mature and precisely appose the postsynaptic muscle fiber is impaired, resulting in the failure of re-establishment of the normal single motor innervation in the NMJ. A complex set of cellular interactions in the NMJ are known to participate in the neurotrophism and neurotrophism to support growth of the regenerating and sprouting axons and their pathfinding to direct the target muscle fiber. Besides the capability of α-motoneurons, signaling originating from the TSCs and muscle may be impaired during aging.

In the course of a morphological investigation of age-related changes in the rat spinal cord, using nicotinamide adenine dinucleotide phosphate-diaphorase (NADPH-d) histochemistry, we found abundant NADPH-d positive bodies, which were characteristically expressed in the aged lumbosacral spinal cord. Together with a normally stained fiber network and a few neurons, the dense, spheroidal NADPH-d positive bodies occurred in portions of the sacral dorsal spinal cords, such as the dorsal commissural nucleus, intermediolateral nuclei, and superficial dorsal horn, and were scattered throughout the dorsal white column. These NADPH-d positive bodies were occasionally observed in a fibrous structure. Two morphologically distinctive subsets of NADPH-d positive bodies were noted in the spinal cord of rats aged 8 to 36 months: 1) highly-dense spheroidal shapes with sharp edges; 2) moderately-dense spheroidal or multiangular shapes with a central "core" and a peripheral "halo". The quantitative analysis, particularly the stereological measurement, confirmed a gradual increase in the incidence and size of NADPH-d positive bodies with increasing age. With nNOS immunohistochemistry, no corresponding structures to NADPH-d positive bodies were detected in aged rats; thus NADPH-d activity is not always specific to the NO-containing neural structures. The major distribution of the NADPH-d positive bodies in the aged lumbosacral spinal cord indicates some anomalous changes in the neurite, which might account for a disturbance in the aging pathway of the autonomic and sensory nerve in the pelvic visceral organs.

Neurogenesis in the hippocampus persist throughout life and precursors of neurons reside in the granule cell layer of the dentate gyrus. Until now, the role of nitric oxide (NO) in the phenomenon has been unclear. By using specific antibodies and a confocal laser scanning microscope, the localization of NO synthase (NOS) was examined in the dentate gyrus of the adult guinea pig in relation with the neuronal precursor marker highly polysialylated neural cell adhesion molecule (PSA-N-CAM). Observation of single immunolabeled sections has revealed that both the PSA-N-CAM- and most NOS-positive cells were localized in the granule cell layer of the dentate gyrus. The former were small in size and showed a punctate, clustered immunoreaction with an irregular cellular margin, whereas the latter showed somewhat diverse cellular profiles. Some NOS-positive neurons had elliptical-like morphology with elongated dendrites, whereas others were small, irregularly shaped and mostly lacking dendritic spines. Double immunolabeling has revealed that NOS-immunoreactivity intermingled, as well as colocalized, with that of PSA-N-CAM, particulary in the granule cell layer. The doubly stained cells were morphologically indistinguishable from PSA-N-CAM single positive cells. These results not only suggest the role of NO production in adult hippocampal neurogenesis, but also indicate that some PSA-N-CAM-expressing neuronal precursors produce NO.

It is well known that regenerating axons enter Schwann cell (SC) columns, within which they grow to reinnervate the appropriate targets. The current study detected a marked induction of a 27‐kDa heat shock protein (HSP27) in the SC columns of crush‐injured rat sciatic nerves. Immunohistochemical studies showed the first appearance of strong HSP27‐immunoreactive linear structures in the proximal stump near an injury site 7 h after an operation. The HSP27‐immunoreactive linear structures crossed the injury site to the distal stump 2 days after the operation. They then extended in a more proximal and more distal direction and were found to have propagated through the entire length of the nerve 1 week after the operation. This pattern of expression was maintained until 3 weeks after the operation. Double‐immunofluorescent labeling and confocal laser microscopy confirmed that the linear structures consisted of SC columns and associated multiple axons. The HSP27‐immunoreactive SC columns expressed glial fibrillary acidic protein, but not S‐100 protein. Electron microscopy and immunoelectron microscopy demonstrated that reactive Schwann cells (SCs) and the associated axons with an outgrowing profile exhibited a strong immunoreactivity to HSP27, with the former containing a greater number of bundles of intermediate filaments. It is suggested that HSP27 may play an essential role in axonal outgrowth, especially by contributing to cytoskeletal dynamics in SCs. GLIA 42:1–11, 2003. © 2003 Wiley‐Liss, Inc.

The literature concerning Schwann cells (SCs) and macrophages in myelin phagocytosis during Wallerian degeneration is reviewed. SCs carry out the first step in the removal of myelin by segmenting myelin and then incorporating the degraded myelin. The recruited macrophages then join in the myelin‐phagocytosis event, appearing to make full use of their original phagocyte abilities until the end of myelin clearance. The molecular mechanisms of the two cells underlying myelin phagocytosis are thought to be different; myelin phagocytosis by SCs being lectin‐mediated, i.e., opsonin‐independent, whereas that of macrophages is mainly opsonin‐dependent. It is important to note that SCs and macrophages cooperatively accomplish myelin phagocytosis. Microsc. Res. Tech. 57:541–547, 2002. © 2002 Wiley‐Liss, Inc.

Intercellular coupling of hepatocytes through gap junctions facilitates exchange of small metabolites or ions, and contributes to maintenance of tissue homeostasis. As protein constituents of the liver gap junction channels, connexin32 (Cx32) and Cx26 have been identified. By use of rat cDNA probes, we cloned cDNAs for guinea pig homologs of Cx32 and Cx26, and compared their amino acid sequences with those of other species. The deduced primary structure of guinea pig Cx32 was 283 amino acids long and contained 98% identical amino acids to the rat and human Cx32. Only six amino acid exchanges were detected between the guinea pig and rat Cx32. On the contrary, the deduced amino acid sequence of guinea pig Cx26 (226 amino acids long) was 91 and 89% identical to the rat and human Cx26, respectively. Twenty-one amino acid exchanges were found between the guinea pig and rat, and the divergence was mostly located in cytoplasmic domains of Cx26. These results suggest that Cx26 shows structural diversity between species, while Cx32 is highly conserved.

Cryosections and whole-mount preparations of the guinea pig small intestine and colon were single or double immunolabeled using the anti-c-Kit and protein gene product 9.5 antibodies. Immunolabeled specimens were observed under a confocal laser scanning microscope. The main findings of the present study are: (1) the distribution and profiles of three-dimensional structures of c-Kit-positive cellular networks in the small intestine and colon, and (2) the anatomical relations of c-Kit-positive cells to the enteric nerves in the layers. In the small intestine, c-Kit-positive cellular networks were observed at levels of the deep muscular plexus and myenteric plexus. The c-Kit-positive cellular networks ran along or overlay the nerve fibers at the deep muscular plexus, while they showed the reticular structures intermingled with the nerve elements at the myenteric plexus. In the colon, c-Kit-positive cellular networks were observed at levels of the submuscular plexus and myenteric plexus, and were further identified within the circular and longitudinal muscle layers as well as in the subserosal layer. In the circular muscle layer, c-Kit-positive cells surrounded the associated nerve fibers and extended several long processes toward the adjacent c-Kit-positive cells. The c-Kit-positive cellular networks within the longitudinal muscle layer as well as in the subserosal layer were not associated with the nerve fibers. In the layers of the intestinal wall with c-Kit-positive cells, the cellular networks of the interstitial cells were identified in ultrastructure. The characteristic profiles of c-Kit-positive cellular networks provide a morphological basis upon which to investigate the mechanisms regulating intestinal movement.

Molecular mechanisms of myelin removal by macrophages were explored by examining the immunophenotypes of macrophages following injury of rat sciatic nerve, using a combined method of immunohistochemistry and confocal laser microscopy. In the crush injury model, the involvement in myelin clearance of a cytoplasmic antigen specific for monocytes/macrophages, ED1, was evident. The obvious recruitment of ED1-immunoreactive (-ir) cells was detected first at the crush injury site and then in the distal stump within which Wallerian degeneration had occurred. Double labelling revealed that the ED1-ir cells, except for monocyte-like round cells, always phagocytosed myelin basic protein-ir myelin debris. On the other hand, the expression of ED2, a surface antigen specific for resident macrophages, was significantly different; ED2-ir cells also increased while myelin removal was progressing from day 3 to day 7, but only some of the cells were engaged in myelin phagocytosis. The poor capacity of myelin phagocytosis by ED2-ir cells was supported by the transection model, in which the proximal stump was ligated to suppress regeneration. ED2 may be involved in events other than myelin removal, providing a local environment conducive to axonal regeneration. Our findings thus seem to suggest that ED1 is one of the most reliable markers for cells carrying out myelin phagocytosis, whereas ED2 may participate in entirely different functions. The expression of complement receptor type 3, OX42, was similar to that of ED1 in terms of the swift recruitment of immunopositive cells, their distribution with close association to myelin debris and their high phagocytotic capacity. This supports previously reported in vitro evidence that myelin phagocytosis by macrophages may be complement-mediated.

Cellular networks of pacemaker activity in intestinal movements are still a matter of debate. Because gap-junctional intercellular communication in the intestinal wall may provide important clues for understanding regulatory mechanisms of intestinal movements, we have attempted to clarify the distribution patterns of three types of gap junction proteins. Using antibodies for connexin40, connexin43, connexin45, smooth muscle actin, and vimentin, immunocytochemical observations were made with the confocal laser scanning microscope on cryosections of fresh-frozen small intestine and colon of the dog and rat. Connexin 45 was localized along the deep muscular plexus of the small intestine in both dog and rat. Double labeling studies revealed that connexin45 overlapped with vimentin –, but not actin-positive areas, indicating the fibroblast-like nature of the cells, rather than their being smooth muscle-like. Connexin43 immunoreactivity appeared along the smooth muscle cell surface in the outer circular layer of the small intestine of both animals. Connexin 40 immunoreactivity was not observed in the muscle layer other than in the wall of large blood vessels. It is suggested that connexin45-expressing cells along the deep muscular plexus of dog and rat small intestine are likely to act as a constituent of a pacemaker system, which may include a conductive system, by forming a cellular network operating via specific types of gap junctions.

A monoclonal antibody (mAb) TP-3 has been established by immunizing rats with the BALB/c mouse thymic epithelial cell line TEL-2. The TP-3 antigen is expressed on stroma cells of thymus, spleen, and lymph node in syngeneic BALB/c mice (H-2d). This antigen is also expressed at a low level on the cell surface of immature thymocytes, and at a high level on mature T and B cells. In allogeneic mice such as C57BL/6 (H-2b) or C3H (H-2k), no cells expressed the TP-3 antigen. Using H-2 congenic mice, reactivity with mAb TP-3 was found to map to a region of H-2DdLd or between Dd and Qa, suggesting that TP-3 is a major histocompatibility complex (MHC) class I antigen. However, immunoprecipitation analysis indicated that this antigen is not identical to the classical mouse class I molecules in terms of molecular size, antigenicity, and tissue distribution.

In the course of ultrastructural investigations of motor endplate pathology mediated by calcium ions, intranuclear sarcoplasmic inclusions, either membrane-free (true type) or membrane-delimited (false type), were observed during chronic daily high-dose exposure to the anticholinesterase neostigmine. At the stage in which subjunctional components, including soleplate nuclei, were severely damaged (day 7), the true nuclear inclusions were frequently associated with the disrupted nuclear envelope (fragmentation, vesiculation etc.) and nuclear pores. At a subsequent stage, in which muscle repair was accelerated and most soleplate nuclei were less severely affected (day 21), formation of the false inclusions in these nuclei was enhanced. Analysis of serial sections of the less severely affected nuclei, where only a true inclusion type was present, revealed no sign of invaginated nuclear envelopes or other membranes enclosing the inclusions. Our findings indicate that morphogenesis of true inclusions depends upon the severity of nuclear degeneration, i.e., in severely affected nuclei there is disruption in the nuclear envelope and/or nuclear pores, while in less severely affected nuclei, either a pinched-off invagination or diffusion of excessive sarcoplasmic proteins into the nucleus via nuclear pores occurs.

Pseudointranuclear inclusions resulting from invaginations of the nuclear envelope, characteristic of myopathic muscles, were found in the soleplate nuclei of normal soleus muscles of rats. Such inclusions had annulate lamellae (AL), nemaline or rod bodies, lamellar structures, a concentric system of membranes, as well as tightly packed sarcoplasmic organelles. An accumulative origin of such included structures was suggested. The AL were sometimes connected directly to the outer membrane enclosing the inclusions, dilated into cisterns and positioned perpendicularly against the nuclear surface. AL, some of which were well-developed, were observed in the same loci as above during the course of denervation atrophy.