Substance P stimulates late-stage rat osteoblastic bone formation through neurokinin-1 receptors.

Substance P (SP) is a widely distributed neuropeptide that works as a neurotransmitter and neuromodulator. Recently, SP receptors, particularly neurokinin-1 receptors (NK(1)-Rs) that have a high affinity for SP, have been observed not only in neuron and immune cells, but also in other peripheral cells, including bone cells. To identify the role of SP in bone formation, we investigated the expression of NK(1)-Rs in osteoblastic cells and the effects of SP on bone formation by rat calvarial osteoblastic cells. Rat calvarial osteoblastic cells were isolated and cultured for 3 weeks in alpha-MEM containing 10% serum, ascorbic acid, dexamethasone, and beta-glycerophosphate. We then investigated NK(1)-R expression, SP effects on osteoblastic bone formation, and osteocalcin mRNA expression in osteoblastic cells. RT-PCR and immunocytochemistry showed that NK(1)-R mRNA was expressed and NK(1)-R was present in 14-day, but not 7-day, cultured calvarial osteoblasts. Bone formation by cultured osteoblastic cells significantly increased after the addition of 10(-8)-10(-6)MSP. During 3 weeks of culture, the addition of SP in the first week did not significantly increase bone formation, whereas adding SP during the first and second week or all 3 weeks significantly increased calvarial osteoblastic bone formation. Furthermore, semi-quantitative RT-PCR indicated that SP stimulated osteocalcin mRNA expression in the osteoblasts at day 14 or day 21, whereas SP did not stimulated the runX2 or type I collagen mRNA expression at day 7 but stimulated them at day 14. These results indicate that SP stimulates bone formation by osteoblastic cells via NK(1)-Rs at late-stage bone formation. These effects were dependent on the expression of NK(1)-R in osteoblastic cells. Our findings suggest that SP secreted from sensory neurons may modulate bone formation after the expression of SP receptors.

Comparison in localization between cystatin C and cathepsin K in osteoclasts and other cells in mouse tibia epiphysis by immunolight and immunoelectron microscopy.

We compared the distribution of a cysteine proteinase inhibitor, cystatin C, with that of cathepsin K in osteoclasts of the mouse tibia by immunolight and immunoelectron microscopy. Light microscopically, strong immunoreactivity for cystatin C was found extracellularly along the resorption lacuna and intracellularly in the organelles of osteoclasts. In serial sections, various patterns of cystatin C and cathepsin K localization were seen, specifically: (1) some resorption lacuna were positive for both cystatin C and cathepsin K; (2) others were positive for either cystatin C or cathepsin K, but not both; and (3) some lacuna were negative for both. In osteoclasts, the localization of cystatin C was similar to that of cathepsin K. Furthermore, cystatin C immunoreactivity was detected in preosteoclasts and osteoblasts, whereas cathepsin K was seen only in preosteoclasts. Electron microscopically, cystatin C immunoreactive products were found in the rough endoplasmic reticulum (ER), Golgi apparatus, vesicles, granules, and vacuoles of osteoclasts. These cystatin C-positive vesicles had fused or were in the process of fusion with the ampullar vacuoles (extracellular spaces) containing cystatin C-positive, fragmented, fibril-like structures. The extracellular cystatin C was deposited on and between the cytoplasmic processes of ruffled borders, and on and between type I collagen fibrils. In the basolateral region of osteoclasts, cystatin C-positive vesicles and granules also fused with vacuoles that contained cystatin C-positive or negative fibril-like structures. These results indicate that osteoclasts not only synthesize and secrete cathepsin K from the ruffled border into the bone resorption lacunae, but also a cysteine proteinase inhibitor, cystatin C. Therefore, it is suggested that cystatin C regulates the degradation of bone matrix by cathepsin K, both extracellularly and intracellularly.

Vanilloid receptor expression in the rat tongue and palate.

Capsaicin, the pungent substance in hot peppers, evokes a sensation of burning pain by stimulating the vanilloid receptor 1 (VR1) on primary afferent neurons. Immunohistochemistry revealed that the taste papillae in the tongue and palate are richly innervated by VR1-immunoreactive nerve fibers. Furthermore, VR1 protein expression was seen in the epithelium facing the oral cavity, although taste cells seemed to be devoid of VR1. The most conspicuous VR1 expression was observed in the epithelial cells of the palatal rugae, although there were no VR1-immunoreactive nerves there. The finding that VR1 is expressed not only in primary afferents but also in oral epithelial cells suggests that it is of great importance in the perception of capsaicin, heat, and acid in the mouth. Since VR1 is known to play a key role in nociception and inflammatory pain, it may be a new target for the treatment of oral pain.

An ultrastructural study of the relationship between sensory trigeminal nerves and odontoblasts in rat dentin/pulp as demonstrated by the anterograde transport of wheat germ agglutinin-horseradish peroxidase (WGA-HRP).

Because the ultrastructure of the trigeminal sensory nerves in dentin, especially in relation to odontoblasts, remains to be clarified, we investigated the relationship between the trigeminal sensory nerves and the odontoblast processes using the anterograde axonal transport technique by injecting wheat germ agglutinin-horseradish peroxidase (WGA-HRP) into the rat trigeminal ganglion. Light microscopically, the nerves labeled with WGA-HRP were mainly concentrated at the pulpal horn, forming a nerve plexus at the subodontoblastic region and penetrating the predentin/dentin about 50 to 70 microns. Ultrastructurally, HRP reaction products were observed intra-axonally in the myelinated (A delta) and unmyelinated (C) axons in the subodontoblastic region. Most nerves lost the Schwann sheath and were naked in the predentin/dentin. The labeled varicosities were close to the odontoblast processes in the dentinal tubules. No synaptic structures could be detected between the varicosities and the odontoblasts, but a gap about 20 nm wide was found between them. One type of varicosity was a rich mitochondria-containing varicosity, while the other was a rich vesicle-containing (large dense core vesicles and small clear vesicles) one. The reaction products were also found in the extracellular spaces surrounding the axons. Sometimes the reaction products were seen in the coated pits or the endocytotic vesicles of the odontoblast processes. The present study demonstrated that nerve endings (varicosities) derived from the trigeminal ganglion were present in the dentinal tubules, and that WGA-HRP extracellularly extruded from the sensory nerves in the odontoblastic layer or predentin/dentin. These findings thus suggest that sensory nerves may have some (e.g., trophic) effect on either odontoblasts or the environment around the sensory nerves in the dentin/pulp.

Oxidative stress-induced DNA damage in the synovial cells of the temporomandibular joint in the rat.

Synovial hyperplasia is a feature of degenerative temporomandibular joint (TMJ) disease. However, the mechanism by which hyperplasia progresses in the TMJ is unknown. Based on the hypothesis that the oxidative stress generated by mechanical loading causes degenerative changes in the TMJ synovium, we investigated the generation of the highly reactive species, peroxynitrite, and the occurrence of DNA damage in the synovium. After condylar hypermobility of rat TMJs, a marker of peroxynitrite, nitrotyrosine, was localized to the nuclei and cytoplasm of the synovial lining cells and fibroblasts in synovitis-induced TMJ. DNA single-strand breaks were found in the nuclei of the synovial cells only after enzyme treatment, whereas DNA double-strand breaks were not detected. These findings indicate that condylar hypermovement induces the proliferation of synovial cells, and suggest that oxidative stress leads to the progression of synovial hyperplasia via DNA damage of the synovial cells in TMJs after mechanical loading.

A topographical and ultrastructural study of sensory trigeminal nerve endings in the rat temporomandibular joint as demonstrated by anterograde transport of wheat germ agglutinin-horseradish peroxidase (WGA-HRP).

To extend our previous light microscopic observations concerning the distribution of trigeminal sensory nerves in the synovium of the rat temporomandibular joint, we investigated the detailed distribution and fine structure of sensory nerve endings at the light and electron microscopic level by the anterograde transport method using wheat germ agglutinin-horseradish peroxidase (WGA-HRP) injected into the trigeminal ganglion. At the light microscopic level, HRP-labeled nerve fibers were observed in the joint capsule and peripheral portion of the disc. The anterior portion of the disc was more densely innervated than the posterior portion, while no nerves were found in the central portion. At the electron microscopic level, HRP reaction products were observed intra-axonally in the thinly myelinated (A delta) and unmyelinated (C) axons in the anterior portion of the joint capsule, and were also localized in the extracellular space surrounding the unmyelinated fibers and terminals. In the subsynovial layer of the synovial membrane, the majority of labeled axons located near blood vessels or among the collagenous fibrils were covered by Schwann cell sheaths, although some naked axon terminals without sheaths were also found. These unsheathed terminals contained mitochondria, small clear vesicles, and large granular vesicles, and were close to the synovial A and/or B cells near the joint cavity. The minimum distance between the terminals and synovial cells was 75 nm. This is the first demonstration of trigeminal sensory nerve terminals close to synovial lining cells or joint cavity and suggests that neuropeptides such as substance P may be released close to the synovial lining cells or joint cavity.

Distribution of substance P and calcitonin gene-related peptide-like immunoreactive nerve fibers in the rat temporomandibular joint.

The density and distribution of substance P-like immunoreactive (SP-LI) and calcitonin gene-related peptide-like immunoreactive (CGRP-LI) nerve fibers in rat temporomandibular joint (TMJ) were investigated in whole-mount preparations and frozen sections by immunohistochemistry with the avidin-biotin-peroxidase complex method. Both types of immunoreactive nerves were observed primarily in the joint capsule, the peripheral articular disc, the synovial membrane, and the periosteum. The distribution of CGRP-LI nerves was similar to that of SP-LI nerves. The anterior portion of the joint capsule and disc was most densely innervated, followed by the posterior, lateral, and medial portions. In addition, CGRP-LI nerves were more numerous and more dense in immuno-intensity than SP-LI nerves. In the synovial membrane, many SP- and CGRP-LI nerves terminated in the subsynovial layer, but some branches extended into the superficial synovial lining layer close to the joint cavity. Immunolabeled nerves were prominently located in the disc attachment and peripheral portion of the disc, and occasional nerves were located in the dense collagenous disc band as an actual disc. However, no fibers were detected in the central disc band. Thus, most of the disc was not innervated by any nerves. The present study provides a morphological basis for the possible roles of neuropeptides in endocytosis by synoviocytes, regulation of blood flow in the synovial membrane, nociception mechanisms of the TMJ, and modulation of the inflammatory response in the TMJ.

NF-kappaB activation and iNOS expression in the synovial membrane of rat temporomandibular joints after induced synovitis.

NF-kappaB plays a pivotal role in pathogenesis in general arthritis. However, the participation of NF-kappaB in inflammation of the temporomandibular joint (TMJ) is poorly understood. We examined NF-kappaB expression in rat TMJs with synovitis induced by condyle hypermobility. By immunohistochemistry, NF-kappaB immunoreactivity was found mainly in the cytoplasm, not the nucleus, of the synovial lining cells of induced-synovitis and control TMJs. Southwestern histochemistry, a new method for detecting transcription factors, showed greater NF-kappaB expression in the nucleus of the synovial lining cells in the hypertrophic synovium than in control synovium. Increased numbers of the synovial lining cells with immunoreactivity for inducible nitric oxide synthase (iNOS), which is transcriptionally regulated by NF-kappaB, were also seen in the inflamed synovium. These findings indicate that excess mechanical stress increases NF-kappaB activation in the TMJ and suggest that active NF-kappaB is involved in the progression of TMJ inflammation.

Immunocytochemical localization of cathepsin D in rat junctional epithelium.

Localization of cathepsin D was studied in the junctional epithelium (JE) of healthy rat gingivae by immuno-light and -electron microscopy, by means of both the avidin-biotin-peroxidase complex method and a colloidal gold IgG method. At the light-microscopic level, cathepsin D was demonstrated in the JE and oral sulcular epithelium (OSE). Cathepsin D immunoreactivity was remarkable in the coronal portion of the JE and decreased toward its apical portion. However, cathepsin D immunoreactivity in the basal cell layer of the JE was negligible or negative. In the OSE, the granular layer was positive for cathepsin D. In the adjacent connective tissue, many macrophage-like cells (not clear at this level) close to the basal cell layer showed strong immunoreactivity. At the electron microscopic level, cathepsin D was found in the primary lysosomes and trans-cisternae of Golgi apparatus in the JE cells. These lysosomes were often fused together or were fused with cathepsin D-negative intracytoplasmic vacuoles to form secondary lysosomes, which indicated that intracellular digestion may have been in progress. However, neutrophils contained few gold particles based on cathepsin D. It is likely that the amounts of cathepsin D contained in the JE cells and macrophages are larger than those of cathepsin D contained in the neutrophils. These findings provided morphological evidence that JE cells have the same endocytotic capacity as macrophages and neutrophils, and that JE cells participate in the intracellular digestion that is carried out by lysosomal enzymes such as cathepsin D. It is suggested, in addition, that maximum intracellular digestion occurs in the coronal portion of the JE.

Topography and distribution of sympathetic nerve fibers in the rat temporomandibular joint: immunocytochemistry and ultrastructure.

The distribution and fine structure of nerve fibers containing neuropeptide Y (NPY), tyrosine hydroxylase (TH), and vasoactive intestinal polypeptide (VIP) in the temporomandibular joint were investigated by both the avidin-biotin complex method and an indirect immunofluorescence technique. The innervation pattern of NPY- and TH-positive fibers differed from that of VIP-positive fibers. Specifically, the former was distributed in both the superficial and deep sublining layers, while the latter was mostly located in the deep sublining layer. NPY- and TH-immunoreactive fibers were largely confined to vascular elements; occasional fibers were observed in the synovial lining layer close to the joint cavity. More nerves with NPY and TH immunoreactivity were observed close to the upper joint compartment than near the lower compartment NPY and TH immunoreactivity was dramatically reduced in the TMJ of superior cervical ganglionectomized animals, indicating the sympathetic origin of these nerves. NPY immunoreactivity was found only in unmyelinated axons, which were located in the adventitia and adventitia-medial border of arteries or arterioles. Occasionally, axons were near the joint cavity, in areas free of vascular structures. These observations show that abundant sympathetic nerves supply the temporomandibular joint of the rat and provide a morphological basis for the involvement of different neuropeptides in vascular regulatory and modulatory functions in physiological and pathophysiological conditions.

A histologic evaluation of retrieved hydroxyapatite-coated blade-form implants using scanning electron, light, and confocal laser scanning microscopies.

We histologically examined seven hydroxyapatite-coated (HA) blade implants removed from patients. Four of them radiologically showed severe bone loss and were easily removed with an elevator. Three radiologically showed vertical bone loss and were removed by surgical procedure. Our histological evaluation indicated that coating separation from the HA implants had occurred, and HA coating resorption by bone tissues was suspected in an implant left in situ for 8 years. Several multinucleated giant cells were seen with a few released particles of HA coating at the point lacking bone contact with the HA coating. The presence of microorganisms on and in the HA coating layer was also noted.

The effects of diabetes on the interface between hydroxyapatite implants and bone in rat tibia.

We examined the influence of diabetes on the implant-bone interface of hydroxyapatite (HA) implants inserted transcortically and extending into the medullary canal of rat tibiae, and quantitatively assessed the differences in bone reaction using an image processing system. Forty male Wistar King A rats (aged 5 weeks) were used in this experiment; they were sacrificed 84 days after implant placement. Toluidine blue-stained undecalcified sections were prepared for histological observation and image analysis, and the labeled sections were observed by confocal laser scanning microscopy. The HA implants in the bone marrow area in the control group were completely encapsulated with a bone layer, and there were some osteoblast-like cells in the bone lacunae apposing the implant surface. The HA implants in the diabetes-induced (DI) group were partially surrounded with a thin bone layer, and there were some fibroblasts running parallel to the implant surface at areas of no bone contact. Quantitative evaluation indicated that the control group showed significantly higher bone contact rate, bone contact thickness, and bone contact area than the DI group. The DI group showed approximately 30% reduction in the percentage of bone contact and 50% reduction in the thickness and the area of surrounding bone tissue.

A study of the regional distribution of bone formed around hydroxyapatite implants in the tibiae of streptozotocin-induced diabetic rats using multiple fluorescent labeling and confocal laser scanning microscopy.

The present study was designed to compare the amount and regional distribution of bone formation around hydroxyapatite (HA) implants in normal (control) rats with that of animals with diabetes mellitus (DM), induced by streptozotocin 2 weeks prior to implant placement. Calcein (CAL), alizarin complexone (AL), and tetracycline (TC) were injected on the 7th, 14th, and 21st days after implantation, respectively, and the rats were sacrificed on the 28th day after implantation. Seventy-microns undecalcified sections of the HA-bone interface in both groups were then prepared for confocal laser scanning microscopy (CLSM) observation. In both groups, bone formation developed from the HA surface to the endosteum, periosteum, or bone marrow. In the control group, around the HA close to the endosteum and periosteum, the new bone showed an extensive lamination pattern of three color layers (CAL, AL, and TC), but in the DM group the labeling density of TC on the 21st day was low. In contrast, on the lateral part of the HA surface (away from the endosteum and periosteum), there was considerably less bone formation in the control group, and in the DM group it was almost completely suppressed. These findings indicate that bone formation around the HA was initiated from the HA surface in the control group, while in the DM group, bone formation along the lateral part of the HA away from the endosteum and periosteum was almost completely suppressed. Furthermore, it is also suggested that in the new bone along the HA close to the endosteum and periosteum, only calcification on the 21st day was depressed.

Ultrastructural and immunoelectron microscopic studies of the peri-implant epithelium-implant (Ti-6Al-4V) interface of rat maxilla.

BACKGROUND: The role played by the internal basal lamina (IBL) and hemidesmosomes between an implant and the peri-implant epithelium (PIE) in the adherence of the epithelium to the implant is controversial. This study used rat maxilla implantation models to clarify the ultrastructure of the PIE-implant interface. METHODS: Ti-6Al-4V implants were inserted either immediately or 2 weeks after the extraction of the upper left first molar of 6- or 4-week-old rats, respectively. The junctional epithelium (JE) of the upper right molars in the same animals was used as a control. Four weeks after implantation, the animals were sacrificed to prepare specimens for light and immunoelectron microscopy. RESULTS: Under light microscopy, the PIE appeared to attach to the implant surface. Ultrastructurally, IBL, consisting of the lamina densa and lamina lucida, and hemidesmosomes were formed only in the lower region, and rarely in the middle region, of the PIE-implant interface. In control teeth, the IBL and hemidesmosomes formed throughout the dento-JE interface. Laminin-1 was found in the IBL and also in the vesicles and vacuoles of the PIE and JE cells. Statistical analysis showed that there was also a significant difference in the amount of IBL between the PIE-implant and dento-JE interfaces. CONCLUSIONS: PIE attached to the implant via hemidesmosomes and IBL in the lower region of the PIE-implant interface. Although PIE cells may secrete laminin-1, which contributes to epidermal cell adhesion, the PIE which attaches to implants only in the lower region of the interface is considered to be the poorly adhered epithelium.

An immunohistochemical and monastral blue-vascular labelling study on the involvement of capsaicin-sensitive sensory innervation of the junctional epithelium in neurogenic plasma extravasation in the rat gingiva.

Nerve fibres immunoreactive for substance P (SP) and calcitonin gene-related peptide (CGRP) were located preferentially at the base of the junctional epithelium. Occasional fibres were observed in close proximity to the subepithelial, small blood vessels. The vascular connective tissue papillae projecting into the epithelium were more densely surrounded by SP- and CGRP-immunoreactive fibres in the interdental col than in other regions of the gingiva. In some cases, hyperplasia of the junctional epithelium was noted in the interdental col where the connective tissue papillae were invaded by widened vessels, indicating severe irritation. SP- and CGRP-immunoreactive fibres around these papillae showed increases in their immunoreactivity and thickness, with some fibres terminating as large expansions. Double immunohistochemical staining revealed the co-existence of SP and CGRP in all nerve fibres within and under the junctional epithelium. Capsaicin pretreatment eliminated most of the immunoreactivity for both peptides. Intravenous infusion of capsaicin or SP caused increased permeability in vessels underlying the junctional epithelium, as indicated by Monastral blue labelling. Labelled vessels were arranged not only in a network extending under the epithelium but also in loops protruding into the connective tissue papillae. These labelled vessels were most abundant in the interdental col, where vascular loops with more complex configurations exhibited strong staining in their walls. In the case of hyperplasia of the junctional epithelium in the interdental col, widened vessels showing extensive labelling in their walls were observed. In capsaicin-pretreated animals, capsaicin-induced extravasation was abolished, while the effect of SP was still observed. These findings provide evidence that capsaicin-sensitive sensory nerves supplying the junctional epithelium are involved in neurogenic plasma extravasation in the rat gingiva. The enhancement of neurogenic plasma extravasation in the col may be vascular response associated with a higher susceptibility of this region to gingival inflammation.

The peripheral distribution of trigeminal nerve fibres in the rat temporomandibular joint studied by an anterograde axonal transport method with wheat germ agglutinin-horseradish peroxidase.

Wheat germ agglutinin-horseradish peroxidase was injected into the trigeminal ganglion to trace the peripheral distribution of the nerve fibres in the temporomandibular joint. It was transported anterogradely along trigeminal nerve fibres. Horseradish-peroxidase-labelled nerve fibres were found in the anterior and posterior bands of the articular disc, and terminated as nerve endings near the intermediate zone of the disc. However, the intermediate zone itself did not contain any nerve endings. Other nerve fibres penetrated from the subsynovial layer into the synovial membrane and also terminated as nerve endings close to the articular cavity. Thus, this method is suitable for tracing peripheral nerve fibres and nerve endings originating in the trigeminal ganglion.

Postnatal development of substance P-, calcitonin gene-related peptide- and neuropeptide Y-like immunoreactive nerve fibres in the synovial membrane of the rat temporomandibular joint.

The postnatal (0-24 days) development of substance P (SP)-, calcitonin gene-related peptide (CGRP)- and neuropeptide Y(NPY)-like immunoreactive (LI) nerves in the rat temporomandibular joint (TMJ) was investigated immunohistochemically. Immediately after birth, SP- or CGRP-LI nerves were observed in most disc attachments. A few NPY-LI nerves were observed around the large blood vessels in the joint capsule. From days 3 to 6, the SP- or CGRP-LI nerves were first found close to the anterior, lateral, medial (third day) or posterior (sixth day) peripheral portion of the disc. The synovial cells (type A and B) first appeared at the anterior peripheral portion of the disc (sixth day), and then at the posterior, lateral and medial portions (seventh day). NPY-LI nerves were found around the blood vessels at the disc attachment on the sixth day, and then entered into the peripheral portion of the disc from days 10 to 14. At 14 days a few NPY-LI nerves were first found close to the blood vessels in the sublining layer of the synovial membrane. From days 18 to 24, a few NPY-LI nerves were located in the superficial layer of the synovial membrane. The central portion of the disc did not contain any nerves from days 0 to 24. Thus SP- or CGRP-LI sensory nerves are shown to innervate the rat TMJ at an earlier age than NPY-LI sympathetic nerves, which may modulate the regulation of blood flow in the joint capsule, disc and synovial membrane. However, it is considered that the disc itself does not contribute to the transportation of the afferent sensory information. Furthermore, from the fact that SP- or CGRP-LI nerves were found earlier than the appearance of the synovial cells, it is suggested that these nerves may be associated with the growth and proliferation of synovial cells.

Immunohistochemical localization of cathepsins B and D in the synovial lining cells of the normal rat temporomandibular joint.

In 1 micron-thick serial cryosections, cathepsins B and D were found to coexist in both type A (macrophage-like) and type B (fibroblast-like) cells: the whole cytoplasm of the type A cells showed strong immunoreactivity, while the type B cells contained a few granular reaction products. It is therefore suggested that type A cells have a marked ability for intracellular digestion of organic materials.

Differences in the transport systems between cementocytes and osteocytes in rats using microperoxidase as a tracer.

Microperoxidase (MP) tracer was injected intravenously into rats to investigate any differences in transport pathways of tissue fluids in the lacunae and canaliculi of cementum and bone. Light microscopically, in deep cementum lacunae, pericellular spaces contained a large amount of MP, while close to the cementum surface, the spaces contained scarcely any. In bone, MP was detected throughout all pericellular spaces. MP was detected intracellularly as granular reaction products in most cementocytes and osteocytes. Electron microscopically, MP was found in the pericellular spaces of cementum and bone lacunae, particularly on collagen fibrils and amorphous material. MP deposits were also intense along the plasma membrane of cementocytes in the deep cementum and along the innermost edge of the deep cementum matrix and bone matrix. In uptake of MP by cementocytes, although extracellular tracer was deposited extensively along the plasma membrane of the deeply positioned cementocytes, uptake by these deep cementocytes was less than that of those close to the surface. However, in bone, most osteocytes showed uniform uptake. These results suggest that the transport pathways for tissue fluids in cementum are in the pericellular spaces, but that cementum has an uneven circulation of tissue fluid. In cementum, although there seems to be a well-developed canalicular system to transport tissue fluid into the deep regions, the deep cementocytes had less endocytotic ability than those close to the surface.

Immunocytochemical localization of cathepsin L in the synovial lining cells of the rat temporomandibular joint.

Localization of cathepsin L in the synovial lining cells of the normal rat temporomandibular joint was investigated by the avidin-biotin-peroxidase complex method for semithin (1 microns) cryosections and the colloidal gold-labelled IgG method for ultrathin sections of LR gold resin. At the light-microscopic level, type A (macrophage-like) and B (fibroblast-like) cells formed the synovial lining layer. Extensive immunoreactivity for cathepsin L was observed in many granules and vacuoles of type A cells, while in the type B cells, immunoreactivity was found in very few granules. In the sublining layer, macrophages and a few fibroblasts were positive for cathepsin L. By electron microscopy, at the peripheral cytoplasm of the type A cells close to the lateral intercellular spaces and joint cavity, numerous coated vesicles and vacuoles (probably early endosomes) indicating endocytotic function were found. Gold particles indicating cathepsin L were localized in the vesicles (primary lysosomes) in the perinuclear cytoplasm and in the larger amorphous vacuoles (1 microns dia) as phagolysosomes. In type B cells, gold particles were limited to the vesicles only (primary lysosomes). The cathepsin L-positive primary lysosomes were numerous in a few fibroblasts in the sublining layer. These results indicate that type A cells contain a large amount of cathepsin L, and suggest that these cells endocytose surplus substances such as collagen and proteoglycan fragments in normal rat TMJ, effecting their digestion and degradation by the action of this proteolytic cathepsin.