Alterations in ultrastructural localization of growth-associated protein-43 (GAP-43) in periodontal Ruffini endings of rat molars during experimental tooth movement

A Shift from a Pivotal to Supporting Role for the Growth-Associated Protein (GAP-43) in the Coordination of Axonal Structural and Functional Plasticity

In a number of animal species, the growth-associated protein (GAP), GAP-43 (aka: F1, neuromodulin, B-50, G50, pp46), has been implicated in the regulation of presynaptic vesicular function and axonal growth and plasticity via its own biochemical properties and interactions with a number of other presynaptic proteins. Changes in the expression of GAP-43 mRNA or distribution of the protein coincide with axonal outgrowth as a consequence of neuronal damage and presynaptic rearrangement that would occur following instances of elevated patterned neural activity including memory formation and development. While functional enhancement in GAP-43 mRNA and/or protein activity has historically been hypothesized as a central mediator of axonal neuroplastic and regenerative responses in the central nervous system, it does not appear to be the crucial substrate sufficient for driving these responses. This review explores the historical discovery of GAP-43 (and associated monikers), its transcriptional, post-transcriptional and post-translational regulation and current understanding of protein interactions and regulation with respect to its role in axonal function. While GAP-43 itself appears to have moved from a pivotal to a supporting factor, there is no doubt that investigations into its functions have provided a clearer understanding of the biochemical underpinnings of axonal plasticity.

Changes of myelinated nerve and myelin basic protein expression in rats' periodontal ligaments after experimental tooth movement

INTRODUCTION

Information about the effect of tooth movement on the myelinated nerve in the periodontal ligament is limited. In this study, we aimed to investigate what responses of the periodontal myelinated nerve can be evoked during experimental tooth movement.

METHODS

In experimental-I group, the maxillary left and mandibular right third molars were moved distally. In experimental-II group, the maxillary left third molar but not the right one was moved, and the bilateral mandibular third molars were extracted. The ultrastructures of the myelinated nerve in the periodontal ligament of the bilateral maxillary third molars were observed under a transmission electron microscope. The expression of myelin basic protein was evaluated by immunohistochemistry.

RESULTS

Degenerative ultrastructural changes of the myelinated nerve in the periodontal ligament were noticed mainly in the myelin sheath; these were observed earlier and were recoverable in the experimental-I group. In contrast, the ultrastructural changes of the myelinated nerve occurred mainly in the axons, were observed later, and were unrecoverable in the experimental-II group. A concomitant decrease of myelin basic protein expression was observed in both groups.

CONCLUSIONS

Both experimental tooth movement and occlusal changes accompanying it caused changes of the myelinated nerve in the periodontal ligament.

Reactions of periodontal ligament epithelial cell clusters and OX6-immunopositive cells to experimental tooth movement and periodontitis

BACKGROUND AND OBJECTIVE

The aim of this study was to investigate reactions of periodontal ligament epithelial cell clusters and major histocompatibility complex class II (OX6)-immunopositive cells to simultaneously induced tooth movement and periodontitis employing Waldo's method.

MATERIAL AND METHODS

Elastic gums were inserted between the right upper first and second molars of rats. Animals were killed by intracardiac perfusion on days 1, 3, 7 and 14 after the experimental procedures, and maxillary molars were decalcified and processed for OCT compound. Cytokeratin and OX6 antibodies to detect epithelial and immunocompetent cells were used for double-fluorescence immunohistochemistry. Immunostained sections of rat upper molar regions were examined with a fluorescence microscope.

RESULTS

Large periodontal ligament epithelial cell clusters appeared and became contiguous with each other, and OX6-immunopositive cells surrounded the clusters over time in the periodontal ligament near the gum insertion site. In the periodontal ligament distant from the gum insertion site, epithelial cell clusters and OX6-immunopositive cells were scattered. After 14 d, thickened epithelium and elongated rete pegs were found close to large epithelial cell clusters in the periodontal ligament near the gum insertion site.

CONCLUSION

These findings suggest proliferation and/or aggregation of periodontal ligament epithelial cells, and interaction between OX6-immunopositive cells and the periodontal ligament epithelial cells, in response to tooth movement and periodontal inflammation. This method may be a useful experimental model to elucidate the relationship between rete pegs and periodontal ligament epithelial cell clusters in inflammatory conditions.

Involvement of neurotrophin-4/5 in regeneration of the periodontal Ruffini endings at the early stage

Little is known about the role of neurotrophin-4/5 (NT-4/5) in the regeneration of mechanoreceptors. Therefore, the present study examined the regeneration process of Ruffini endings in the periodontal ligament in nt-4/5-deficient and wildtype mice following transection of the inferior alveolar nerve by immunohistochemistry for protein gene product 9.5 (PGP 9.5), a general neuronal marker, and by computer-assisted quantitative image analysis. Furthermore, rescue experiments by a continuous administration of recombinant NT-4/5 were performed and analyzed quantitatively. At postoperative day 3 (PO 3d), almost all PGP 9.5-positive neural elements had disappeared; they began to appear in both types of animals at PO 7d. At PO 10d, almost all nerve fibers showed a beaded appearance, with fewer ramifications in both types of mice. Although the regeneration proceeded in the wildtype, a major population of the periodontal Ruffini endings continued to display smooth outlines at PO 28d in the nt-4/5 homozygous mice. The reduction ratio of neural density reached a maximum at PO 3d, decreased at PO 10d, and later showed a plateau. In a rescue experiment, an administration of NT-4/5 showed an acceleration of nerve regeneration in the homozygous mice. These findings indicate that the nt-4/5-depletion causes a delay in the regeneration of the periodontal Ruffini endings, but the delay is shortened by an exogenous administration of NT-4/5. Combined with our previous findings of bdnf-deficient mice (Harada et al. [2003] Arch Histol Cytol 66:183-194), these morphological and numerical data suggest that multiple neurotrophins such as NT-4/5 and brain-derived neurotrophic factor (BDNF) play roles in their regeneration in a stage-specific manner.

Cellular, molecular, and tissue-level reactions to orthodontic force

Remodeling changes in paradental tissues are considered essential in effecting orthodontic tooth movement. The force-induced tissue strain produces local alterations in vascularity, as well as cellular and extracellular matrix reorganization, leading to the synthesis and release of various neurotransmitters, cytokines, growth factors, colony-stimulating factors, and metabolites of arachidonic acid. Recent research in the biological basis of tooth movement has provided detailed insight into molecular, cellular, and tissue-level reactions to orthodontic forces. Although many studies have been reported in the orthodontic and related scientific literature, a concise convergence of all data is still lacking. Such an amalgamation of the rapidly accumulating scientific information should help orthodontic clinicians and educators understand the biological processes that underlie the phenomenon of tooth movement with mechanics (removable, fixed, or functional appliances). This review aims to achieve this goal and is organized to include all major findings from the beginning of research in the biology of tooth movement. It highlights recent developments in cellular, molecular, tissue, and genetic reactions in response to orthodontic force application. It reviews briefly the processes of bone, periodontal ligament, and gingival remodeling in response to orthodontic force. This review also provides insight into the biological background of various deleterious effects of orthodontic forces.

Nerve-epithelium association in the periodontal ligament of guinea pig teeth

Several lines of evidence have suggested that periodontal nerves have other roles besides sensory function. Exploring the distribution pattern of nerves in relation to other structures within the periodontal ligament of various species should be important to understand their roles within the ligament. This study investigated whether any association exists between the nerves and the epithelial cells in the periodontal ligament of continuously erupting guinea pig molars, which show distinct enamel epithelium layers among the cementum pearls. Ten guinea pigs were fixed by vascular perfusion and jaw sections were processed for immunohistochemistry of protein gene product 9.5 (PGP 9.5), growth-associated protein-43 (GAP-43) and glia-specific S-100 protein, and for enzyme histocytochemistry of cholinesterase. Nerves that were immunopositive for the above neuronal markers were located predominantly in the alveolus-related part of the periodontal ligament. Some nerves, immunoreactive for PGP 9.5 and GAP-43, were also found in the tooth-related part (TRP) of the periodontal ligament close to the tooth surface. PGP 9.5-positive nerves in the TRP appeared very thin and terminated by making loops or plexus-like structures in close apposition to the epithelium layers, overlying the enamel surface in between cementum pearls. Such an intimate association between nerves and the enamel epithelium was not found in the labial periodontal tissue of incisors or the apical growing end of the molar, where periodontal fibre attachment was indistinct. The association between nerves and epithelium in the periodontal ligament of guinea pig molar is site specific and is only seen in the presence of cementum, suggesting that this association is related to the attachment function of the ligament.

C-fos expression in rat brain nuclei following incisor tooth movement

In the rat experimental model, molar tooth movement induced by Waldo's method is known to cause a temporally and spatially defined pattern of brain neuronal activation. Since orthodontic correction usually involves the entire dental arch, we used a spring-activated appliance to extend the investigation to incisors, and we included brain regions related to antinociception. Adjustment of the non-activated appliance on incisors resulted in c-fos expression in the dorsal raphe, peri-aqueductal gray matter, and the locus coeruleus, in addition to trigeminal sensory subnuclei and the parabrachial nucleus, where neuronal activation has already been detected in previous studies on molar tooth movement. Appliance activation with a 70-g force resulted in a further increase in Fos-immunoreactive neurons in the trigeminal sensory subnucleus caudalis and in the dorsal raphe. This result suggests that there is a recruitment of neurons related to nociception and to antinociception when tooth movement is increased.

Expression of c-Fos protein in the trigeminal nuclear complex resulting from quantified force application to the rat molar

This study was conducted to investigate the expression and distribution of c-Fos-like immunoreactive neurones (Fos-neurones), in the rat trigeminal sensory nuclear complex, produced by mechanical forces with various magnitudes and durations applied to the left upper first molar. The magnitudes of forces applied to the tooth were 25, 50 and 100 g and the duration was 2 h. A quantified force of 100 g was also applied to the upper molar for varying durations [short-time (1-2 min)], 2, 4, 8 and 12 h. Fos-neurones distributed in the bilateral superficial laminae of the subnucleus caudalis, and the ipsilateral dorsomedial part of subnucleus oralis (Sp5Odm). The number of Fos-neurones increased in the subnucleus caudalis (Sp5C) according to the force magnitude. In the Sp5C, the number of Fos-neurones exhibited maximum level, 2 or 4 h after the application. In the Sp5Odm, however, the number of Fos-neurones reached the maximum level at 8 h. These data suggest that the change in the number of nociceptive neurones in Sp5C reflect changes in encoding the magnitude of force to tooth, and that the nature of pain response to orthodontic forces might have some relation to the delayed expression of c-Fos protein in the Sp5Odm.

The involvement of brain-derived neurotrophic factor (BDNF) in the regeneration of periodontal Ruffini endings following transection of the inferior alveolar nerve

The present study employed immunohistochemistry for protein gene product 9.5 (PGP 9.5) to examine the regeneration process of Ruffini endings, the primary mechanoreceptor in the periodontal ligament, in heterozygous mice with targeted disruption of the brain-derived neurotrophic factor (BDNF) gene and their littermates, following transection of the inferior alveolar nerve. When immunostained for PGP 9.5, periodontal Ruffini endings appeared densely distributed in the periodontal ligament of the heterozygous mice, but the density of the positively stained nerve fibers in the ligament was 20% lower than that in the control littermates. At 3 days after surgery, the PGP 9.5-positive neural elements had disappeared; they began to appear in the periodontal ligament of both animals at 7 days. However, the recovery pattern of the PGP 9.5-positive nerves differed between heterozygous and wild type mice, typical periodontal Ruffini endings morphologically identical to those in the control group appeared in the wild-type mice at 7 days, whereas such Ruffini endings were detectable in the heterozygous mice at 28 days, though much smaller in number. On day 28, when PGP 9.5-positive nerves were largely regenerated in wild type mice, their distribution was much less dense in the ligament of the heterozygous mice than in the non-treated heterozygous mice. The density of PGP 9.5-positive nerve fibers was significantly lower in the heterozygous mice than in wild type mice at any stage examined. These data showing that a reduced expression of BDNF causes delayed regeneration of the periodontal Ruffini endings suggest the involvement of BDNF in the regeneration process of these mechanoreceptors.

Regeneration of periodontal Ruffini endings in adults and neonates

We reviewed the regeneration of periodontal Ruffini endings, primary mechanoreceptors in the periodontal ligament, following injury to the inferior alveolar nerve (IAN) in adult and neonatal rats. Morphologically, mature Ruffini endings are characterized by an extensive arborization of axonal terminals and association with specialized Schwann cells, called lamellar or terminal Schwann cells. Following injury to IAN in the adult, the periodontal Ruffini endings of the rat lower incisor ligament regenerate more rapidly than Ruffini endings in other tissues. During regeneration, terminal Schwann cells migrate into regions where they are never found under normal conditions. The development of periodontal Ruffini endings of the rat incisor is closely associated with the eruption of the teeth; the morphology and distribution of the terminal Schwann cells became almost identical to those in adults during postnatal days 15-18 (PN 15-18d) when the first molars appear in the oral cavity, while the axonal elements showed extensive ramification around PN 28d when the functional occlusion commences. When the IAN was injured in neonates, the regeneration of periodontal Ruffini endings was delayed compared with the adults. The migration of terminal Schwann cells is also observed following IAN injury, after which the distribution of terminal Schwann cells became almost identical to that of the adults, i.e., PN 14d. Since the interaction between axon and Schwann cell is important during regeneration and development, further studies are required to elucidate its molecular mechanism during the regeneration as well as the development of the periodontal Ruffini endings.

Growth-associated protein-43 immunohistochemical and ultrastructural changes in jaw muscle spindles of the rat following loss of occlusion

The effects of complete loss of occlusion on the structural and functional status of these muscle spindles were investigated by immunohistochemistry either for protein gene product 9.5 (PGP 9.5) or growth-associated protein-43 (GAP-43) by light and electron microscopy. All the upper molars of 4-week-old Wistar rats were extracted and the erupted portions of the upper and lower incisors of the same animals were cut-off at the level of the interdental papilla every other day. In a control group, immunoreactivity for GAP-43 was positive in the developing annulospiral endings of 2-week-old rats, but was not detected in any of the muscle spindles after 3 weeks of age. At 4 weeks of age, the PGP 9.5 immunostained spindles had well-differentiated annulospiral endings. Ultrastructurally, these afferent endings showed lenticular or circular profiles in cross-sections, and were differentially indented into the intrafusal-fibres. The inner surfaces of the terminals formed rather smooth myoneural junctions, while the outer surfaces were covered only by basal lamina continuous with that of the underlying intrafusal muscle fibres. After the experimental elimination of occlusal contact, GAP-43 immunoreactivity reappeared in some nerve endings of muscle spindles by 3 days, and persisted for at least 28 days. During this period, the afferent-terminals exhibited various fine structural abnormalities such as irregular outlines and invaginated neuromuscular interfaces. Some sensory-terminal (ST) profiles were completely engulfed by intrafusal-fibres. However, GAP-43 expression and ultrastructural alterations became undetectable within a week of the end of incisal cutting and the recovery of incisal-contact. These data indicate that remodelling of nerve terminals in muscle spindles, as assessed by GAP-43 expression and ultrastructural changes, occurs soon after a loss of occlusion, and ceases if incisal-contact is restored. It is concluded that possible changes in jaw muscle function, as well as a sudden loss of proprioceptive sensory input from the periodontal mechanoreceptors of molars and incisors, induce the structural reorganisation of nerve terminations in jaw muscle spindles that is associated with the appearance and disappearance of GAP-43 immunoreactivity.

Temporal and spatial distribution of Fos protein in the parabrachial nucleus neurons during experimental tooth movement of the rat molar

The present study was undertaken to reveal spatio-temporal changes in the distribution of Fos-like immunoreactive (-IR) neurons in the parabrachial nucleus (PBN), one of the important relay nuclei for processing autonomic and somatosensory information from the oro-facial regions, following the induction of experimental tooth movement in rat upper molars. The experimental tooth movement was induced by the insertion of elastic rubber between the first and second upper molars. In normal animals, the PBN contained a smaller number of Fos-IR neurons. Following experimental tooth movement, the Fos-IR neurons increased in number significantly on both the ipsilateral and contralateral PBN, reaching a maximum at 4 h (about 10 times that of normal animals), and then decreased gradually. However, a significant number of Fos-IR neurons remained at 24 h post-operation. Remarkable side-by-side differences in the number of Fos-IR neurons were recognized at 1 to 4 h following the experimental tooth movement. Their number returned to normal (basal) levels at 5 days post. All subnuclei of PBN showed similar temporal changes in the number of Fos-IR neurons, this being particularly apparent in lateral PBN. Administrations of morphine (3 and 10 mg/kg, i.p.) drastically reduced the induction of Fos-IR neurons in all subnuclei of both the ipsilateral and contralateral PBN in a dose-dependent manner, and its effect was antagonized by pretreatment with naloxone (2 mg/kg, i.p.). The reduction of Fos-IR neurons by morphine pretreatment suggests that the appearance of Fos-IR neurons in the PBN may be partly due to the noxious stimulation and/or stress arising from tooth movement. The bilateral expression of Fos-IR neurons in the PBN indicates that the experimental tooth movement causes the activation of PBN neurons for the processing of somatosensory as well as autonomic information. The prolonged expression of Fos-IR neurons in all the subnuclei of bilateral PBN reflects clinical features of the transient discomfort and/or abnormal sensations, which many patients often complain about during orthodontic treatment.

Time-related changes in periodontal mechanoreceptors in rat molars after the loss of occlusal stimuli

The effect of a loss of occlusal stimuli upon the distribution and structure of the periodontal mechanoreceptors of the rat mandibular molar was examined after extracting opposing molars. The hypofunctional periodontal ligament narrowed significantly two weeks after tooth extraction, associated with an altered morphology of the Ruffini endings that showed typical dendritic profiles in normal controls. At four weeks and later periods after extraction, the Ruffini endings-including those without light microscopic changes demonstrated unusual ultrastructural features such as the eccentric localization of mitochondria along the axonal membrane and loss of other cell organelles, unusual elongation of axonal microprojections, or a deep invagination of the Schwann sheath into the axoplasm. Immunoreactivity for the growth-associated protein-43 (GAP-43) in the Ruffini endings was restricted to the Schwann element in both the normal and hypofunctional periodontal ligament, but the reaction was weaker and even negligible in some cases in the latter ligament. The present results suggest that occlusal stimuli are essential for maintaining the structural integrity of the periodontal ligament, including that of periodontal mechanoreceptors. A decreased immunoreactivity for GAP-43 in the Schwann sheaths supports the notion of a possible functional alteration in the Ruffini endings that showed no structural abnormality.

Morphological and cytochemical characteristics of periodontal Ruffini ending under normal and regeneration processes

Current knowledge on the Ruffini endings, primary mechanoreceptors in the periodontal ligament is reviewed with special reference to their cytochemical features and regeneration process. Morphologically, they are characterized by extensive ramifications of expanded axonal terminals and an association with specialized Schwann cells, called lamellar or terminal Schwann cells, which are categorized, based on their histochemical properties, as non-myelin-forming Schwann cells. Following nerve injury, the periodontal Ruffini endings of the rat incisor ligament can regenerate more rapidly than Ruffini endings in other tissues. During regeneration, terminal Schwann cells associated with the periodontal Ruffini endings migrate into regions where they are never found under normal conditions. Also during regeneration, alterations in the expression level of various bioactive substances occur in both axonal and Schwann cell elements in the periodontal Ruffini endings. Neuropeptide Y, which is not detected in intact periodontal Ruffini endings, is transiently expressed in their regenerating axons. Growth-associated protein-43 (GAP-43) is expressed transiently in both axonal and Schwann cell elements during regeneration, while this protein is localized in the Schwann sheath of periodontal Ruffini endings under normal conditions. The expression of calbindin D28k and calretinin, both belonging to the buffering type of calcium-binding proteins, was delayed in periodontal Ruffini endings, compared to their morphological regeneration. As the importance of axon-Schwann cell interactions has been proposed, further investigations are needed to elucidate their molecular mechanism particularly the contribution of growth factors during the regeneration as well as development of the periodontal Ruffini endings.

The Ruffini ending as the primary mechanoreceptor in the periodontal ligament: its morphology, cytochemical features, regeneration, and development

The periodontal ligament receives a rich sensory nerve supply and contains many nociceptors and mechanoreceptors. Although its various kinds of mechanoreceptors have been reported in the past, only recently have studies revealed that the Ruffini endings--categorized as low-threshold, slowly adapting, type II mechanoreceptors--are the primary mechanoreceptors in the periodontal ligament. The periodontal Ruffini endings display dendritic ramifications with expanded terminal buttons and, furthermore, are ultrastructurally characterized by expanded axon terminals filled with many mitochondria and by an association with terminal or lamellar Schwann cells. The axon terminals of the periodontal Ruffini endings have finger-like projections called axonal spines or microspikes, which extend into the surrounding tissue to detect the deformation of collagen fibers. The functional basis of the periodontal Ruffini endings has been analyzed by histochemical techniques. Histochemically, the axon terminals are reactive for cytochrome oxidase activity, and the terminal Schwann cells have both non-specific cholinesterase and acid phosphatase activity. On the other hand, many investigations have suggested that the Ruffini endings have a high potential for neuroplasticity. For example, immunoreactivity for p75-NGFR (low-affinity nerve growth factor receptor) and GAP-43 (growth-associated protein-43), both of which play important roles in nerve regeneration/development processes, have been reported in the periodontal Ruffini endings, even in adult animals (though these proteins are usually repressed or down-regulated in mature neurons). Furthermore, in experimental studies on nerve injury to the inferior alveolar nerve, the degeneration of Ruffini endings takes place immediately after nerve injury, with regeneration beginning from 3 to 5 days later, and the distribution and terminal morphology returning to almost normal at around 14 days. During regeneration, some regenerating Ruffini endings expressed neuropeptide Y, which is rarely observed in normal animals. On the other hand, the periodontal Ruffini endings show stage-specific configurations which are closely related to tooth eruption and the addition of occlusal forces to the tooth during postnatal development, suggesting that mechanical stimuli due to tooth eruption and occlusion are a prerequisite for the differentiation and maturation of the periodontal Ruffini endings. Further investigations are needed to clarify the involvement of growth factors in the molecular mechanisms of the development and regeneration processes of the Ruffini endings.

Effects of morphine on the distribution of Fos protein in the trigeminal subnucleus caudalis neurons during experimental tooth movement of the rat molar

The present study was undertaken to disclose temporal changes in the distribution of Fos-like immunoreactive (-IR) neurons in the trigeminal subnucleus caudalis (SpVc), one of the important relay nuclei for processing the nociceptive information from the oro-facial regions, following induction of experimental tooth movement in rat upper molars. Furthermore, the effect of morphine and naloxone on the levels of Fos-IR neurons in the SpVc was examined. The experimental tooth movement was induced by insertion of an elastic rubber between the first and second upper molars. In normal animals, Fos-IR neurons were rarely observed in the SpVc. Immediately after insertion of the elastic band, the distribution of Fos-IR neurons was comparable to that observed in normal animals. The number of Fos-IR neurons increased significantly from 1 to 4 h following the induction of experimental tooth movement, reaching a maximum at 2 h, and then decreasing gradually. Most of the neurons were localized in the dorsomedial portion of the superficial layers of the ipsilateral SpVc near the obex, but a few were observed at the ventral portion of the SpVc. The neurons at the superficial layers and ventral portion of the contralateral SpVc also showed Fos-like immunoreactivity, but their numbers were significantly smaller than those on the ipsilateral side. Pretreatment with morphine (3 and 10 mg/kg, i.p.) significantly reduced the induction of Fos-IR neurons at the superficial layers of the ipsilateral SpVc in a dose-dependent manner, and its effect was antagonized by the subsequent treatment of naloxone (2 mg/kg, i.p.). Naloxone pretreatment enhanced the expression of Fos-IR neurons on the ipsilateral SpVc. The present results of a reduction of Fos-IR neurons by morphine pretreatment suggest that the induction of Fos-IR neurons may be due to the noxious stimulation caused by induction of experimental tooth movement.

Growth-associated protein-43 (GAP-43) in the regenerating periodontal Ruffini endings of the rat incisor following injury to the inferior alveolar nerve

Alterations in the levels of growth-associated protein 43 (GAP-43)-like immunoreactivity (-LI) were examined in the lingual periodontal ligament of the rat incisor following two types of injury (resection and crush) to the inferior alveolar nerve (IAN). In normal animals, GAP-43-like immunoreactive (IR) structures were observed as tree-like ramifications in the alveolar half of the lingual periodontal ligament of incisors. Under immunoelectron microscopy, GAP-43-LI appeared in the Schwann sheaths associated with periodontal Ruffini endings; neither cell bodies of the terminal Schwann cells nor axonal profiles showed GAP-43-LI. During regeneration of the periodontal Ruffini endings following resection of the IAN, GAP-43-LI appeared in the cytoplasm of the terminal Schwann cell bodies and axoplasm of the terminals. The distribution of GAP-43-LI in the Ruffini endings returned to almost normal levels on days 28 and 56 following the injury. The changes in the distribution of GAP-43-LI following the crush injury were similar to those following resection; however, expression of GAP-43-LI was slightly higher for the entire experimental period compared with the resection. The transient expression of GAP-43 in the terminal Schwann cells and axonal profiles of the periodontal Ruffini endings following nerve injury suggests that GAP-43 is closely associated with axon-Schwann cells interactions during regeneration.