Development of the articular cavity in the rat temporomandibular joint with special reference to the behavior of endothelial cells and macrophages.код для вставкиСкачать
THE ANATOMICAL RECORD PART A 286A:908 –916 (2005) Development of the Articular Cavity in the Rat Temporomandibular Joint With Special Reference to the Behavior of Endothelial Cells and Macrophages AKIKO SUZUKI,*1 KAYOKO NOZAWA-INOUE,1 NOBUYUKI IKEDA,2 NORIO AMIZUKA,1,3 KAZUHIRO ONO,2 RITSUO TAKAGI,2 AND TAKEYASU MAEDA1,3 1 Division of Oral Anatomy, Department of Oral Biological Science, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan 2 Division of Oral Maxillofacial Surgery, Department of Oral Health Science, Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan 3 Center for Transdisciplinary Research, Niigata University, Niigata, Japan ABSTRACT Previous developmental studies on the temporomandibular joint (TMJ) have proposed several hypotheses on the formation of its articular cavity. However, detailed information is meager. The present study examined the formation process of the articular cavity in the rat TMJ by immunocytochemistry for CD31, RECA-1, and ED1, which are useful cellular markers for endothelial cells and monocyte/macrophage lineages, respectively. The upper articular cavity formation had begun by embryonic day 21 (E21) and was completed at postnatal day 1 (P1) in advance of the lower cavitation; the latter took place from P1 to P3. The occurrence and distribution pattern of the CD31-, RECA-1-, and ED1-positive cells differed between the upper and lower articular cavity-forming areas: the ED1-positive cells exclusively occurred in the area of the prospective upper articular cavity prior to its formation, while no ED1-positive cell appeared in the lower cavity-forming area. In contrast, the CD31- and RECA-1-positive endothelial cells were restricted to the lower cavity-forming area (never the prospective upper cavity) at E19 and diminished thereafter. Throughout the cavity formation, we failed to ﬁnd any apoptotic cells in the cavity formation area, indicating no involvement of apoptosis in the cavity formation in TMJ. The present ﬁndings on the behaviors of endothelial cells and ED1-positive cells show a possibility of different mechanism in the cavity formation between the upper and lower articular cavities in the rat TMJ. The appearance of ED1-reactive cells and temporal vascularization may play crucial roles in the upper and lower articular cavity formation, respectively. © 2005 Wiley-Liss, Inc. Key words: temporomandibular joint; development; articular cavity formation; endothelial cell; monocyte/macrophage lineages. The temporomandibular joint (TMJ) is a bilateral synovial arthrosis between the mandibular fossa of the temporal bone and the mandibular condyle. The articular cavity of TMJ is completely separated by the articular disk to divide into two cavities, the upper and lower articular cavities. Both the upper articular cavity and articular disk involve a sliding movement of the condyle, whereas the lower one plays a role in the rotation of the condyle (Walmsley, 1964). These cavities are ﬁlled with the viscous synovial ﬂuid, which makes these smooth jaw movements possible. Previous developmental studies on the TMJ have paid attention to the general development, including the developmental sequence of compositional elements such as the mandibular condyle and articular disk, of the TMJ (Symons, 1952; Morimoto et al., 1987; Van der Linden et al., © 2005 WILEY-LISS, INC. 1987; Mérida-Velasco et al., 1993). Furthermore, a majority of developmental studies have focused on the individual development of the components— bony elements Grant sponsor: Japanese Ministry of Education, Culture, Sports, Science, and Technology; Grant number: 16659498. *Correspondence to: Akiko Suzuki, Division of Oral Anatomy, Department of Oral Biological Science, Niigata University Graduate School of Medical and Dental Sciences, 2-5274 Gakkochodori, Niigata 951-8514, Japan. Fax: 81-25-223-6499. E-mail: email@example.com Received 3 November 2004; Accepted 20 June 2005 DOI 10.1002/ar.a.20228 Published online 18 August 2005 in Wiley InterScience (www.interscience.wiley.com). ARTICULAR CAVITY FORMATION IN TMJ (Bach-Petersen et al., 1994), lateral pterygoid muscle (Ögütcen-Toller and Juniper, 1993, 1994), and ligaments (Ögütcen-Toller, 1995)—around the TMJ in various animals. However, little information is available on the development of articular cavity in TMJ in detail in contrast to the accumulation of knowledge on joints of other long bones, such as limb joint. Several hypotheses have been proposed for the mechanism of articular cavity formation during limb development (cf. Archer et al., 2003). The cavitation may be caused by several factors, including cell death, enzymatic degradation in the prospective cavity, differential growth of opposing elements, differential matrix synthesis, and mechanical inﬂuences (Andersen and Bro-Rasmusen, 1961; Mitrovic, 1977, 1978; Nalin et al., 1995; Abu-Hijleh et al., 1997). These factors have been considered to adapt to the mechanism of cavity formation in TMJ (Murray and Drachman, 1969; Mitrovic, 1977; Linck and Porte, 1978; Okada et al., 1981; Mérida-Velasco et al., 1999). However, TMJ that belongs to a category of a secondary cartilage with covering ﬁbrous tissue is quite different from other long bone joints at phylogenical and ontogenical aspects, indicating that data on the other limb joints cannot directly adapt on the TMJ (for review, see Nozawa-Inoue et al., 2003). Ohnuki (2000) demonstrated the penetration of blood vessels into the lower, never upper, articular cavity during TMJ development in the human fetus by observation of the serial sections. In our recent developmental study of the synovial membrane (Ikeda et al., 2004), we noticed communication between the lower articular cavity and the surrounding blood vessels during the development of rat TMJ. These ﬁndings indicate the possibility that the vascularization around the TMJ is closely related to the articular cavity formation. However, the involvement of vascularization in the cavity formation in TMJ remains unclear due to the scarcity of developmental investigations of this issue. On the other hand, Linck and Porte (1978) suggested crucial roles for macrophage-like type A cells in the synovial lining layer in the formation of the articular cavity as they serve in the absorption of cell fragments and wastes, which are produced through the articular cavity formation. However, the behavior of the macrophage-like type A cells during the cavity formation process in TMJ remains unknown. The present study was therefore undertaken to investigate the formation process of the articular cavity in the rat TMJ by immunocytochemistry for cellular markers for endothelial cells and macrophages during the prenatal and postnatal stages. It will focus on the chronological changes in the appearance and distribution pattern of endothelial cells and macrophages during upper and lower articular cavity formation. At the early stage of the articular cavity formation, we further applied an Indian-ink injection method to demonstrate vascularization as well as in situ identiﬁcation of fragmented DNA using TdTmediated dUTP-biotin nick end labeling (TUNEL) to detect apoptotic cells. MATERIALS AND METHODS All experiments were performed under guidelines of the Niigata University Faculty of Dentistry Intramural Animal Use and Care Committee. 909 Tissue Preparation A total of 48 Wistar rats was obtained at embryonic day 18 (E18), E19, and E21 (n ⫽ 10 each), as well as postnatal day 1 (P1), P3, and P5 (n ⫽ 6 each). The onset of pregnancy was determined by vaginal smearing, and the day sperm was found in the smear was regarded as E0. We deﬁned the day of birth (P1) as 24 – 48 hr after birth. In order to avoid developmental variations among the animals within each day, suitable embryos and postnatal rats were selected by their crown-rump length (CRL; E18 ⫽ 19.8 –21.2 mm; E19 ⫽ 24.6 –26.4 mm; E21 ⫽ 33.8 –36.2 mm) and their body weight (P1 ⫽ 7.0 – 8.0 g; P3 ⫽ 9.0 –11.0 g; P5 ⫽ 11.0 –14.0 g) according to a report by González (1932). Under anesthesia by an intraperitoneal injection of 8% chloral hydrate (400 mg/kg), the postnatal animals were perfused with a ﬁxative containing 4% paraformaldehyde and 0.0125% glutaraldehyde in a 0.1 M phosphate buffer (pH 7.4). The fetuses were deeply anesthetized in the same manner, decapitated, and ﬁxed in the same ﬁxative. The removed heads were decalciﬁed with a 5% ethylene diamine tetra-acetic acid disodium (EDTA-2 Na) solution at 4°C. One side of each head was equilibrated in a 30% sucrose solution at 4°C overnight for cryoprotection and embedded in an OCT compound (Leica, Nussloch, Germany). Serial sagittal sections were cut at a thickness of 35 m in a cryostat (HM-500; Carl Zeiss, Jena, Germany) and mounted onto silane-coated glass slides. The other half of each head was embedded in parafﬁn. Serial parafﬁn sections were sagitally cut at a thickness of 5 m. Some sections were stained with hematoxylin and eosin (H&E) for histological observations. Immunohistochemistry The cryostat sections were processed for immunocytochemistry using the avidin-biotin complex (ABC) method according to Hsu et al. (1981). After an inhibition of endogenous peroxidase with 0.3% H2O2 in absolute methanol for 30 min, the sections were incubated for 24 hr at 4°C with the primary antibodies. We used two kinds of monoclonal antibodies to CD31 (1:50; BD Pharmingen, NJ) (DeLisser et al., 1994; Williams et al., 1996) and rat endothelial cell antigen (RECA-1; 1:80; Serotec, Oxford, U.K.) (Duijvestijn et al., 1992) for endothelial cells. A monoclonal ED1 antibody (1:500; Serotec), for detection of monocyte/macrophage lineages (Dijkstra et al., 1985), was applied for detection of macrophages. Prior to incubation with an antibody to CD31, sections were pretreated with proteinase K (1:10; Dako, Carpinteria, CA) for 20 min at room temperature. The bound primary antibody was localized using biotinylated antimouse IgG for 2 hr and subsequently ABC conjugated with peroxidase for 90 min at room temperature (ABC kit; Vector Lab, Burlingame, CA). Final visualization used 0.04% 3,3⬘-diaminobenzidine tetrahydrochloride and 0.0125% H2O2 in a 0.05 M Tris-HCl buffer (pH 7.6). Some immunolabeled sections were counterstained with methylene blue. Immunohistochemical Controls Immunohistochemical controls were performed by replacing the primary antibodies with nonimmune mouse sera or PBS, and omitting the antimouse IgG or the ABC conjugated with peroxidase. These control sections did not reveal any immunoreaction. The characterization of the antibodies and the origin of antigens have been previously 910 SUZUKI ET AL. reported elsewhere (cf. Dijkstra et al., 1985; Duijvestijn et al., 1992; DeLisser et al., 1994). Vascular Indian-Ink Perfusion Additional rats at E19, E21, and P1 (n ⫽ 8 each) were perfused with the same ﬁxative mentioned above and followed with Indian ink containing 5% gelatin solution (50°C) via left ventricle under deep anesthesia. The perfused heads were immediately frozen in liquid nitrogen, embedded in an OCT compound (Leica), and sectioned at a thickness of 35 m in a cryostat. They were slightly counterstained with methylene blue. In Situ Identiﬁcation of Fragmented DNA Using TdT-Mediated dUTP-Biotin Nick End Labeling Dewaxed sections were end-labeled with terminal deoxynucleotidyl transferase using a TACS 2 TdT-Blue Label in situ Apoptosis Detection Kit (Trevigen, Gaithersburg, MD). According to the manufacturer’s instructions, the sections were treated with proteinase K (1:50) for 15 min, 3% H2O2 in methanol for 5 min, and 1 ⫻ TdT labeling buffer for 5 min at room temperature; Twenty l labeling reaction mixture containing TdT, dNTP mix, 50 ⫻ cation stock (Mn2⫹), TdT enzyme, and 1 ⫻ TdT labeling buffer was then applied to each specimen at 37°C for 1 hr. The slides were incubated with a 1 ⫻ TdT stop buffer for 5 min and subsequently labeled with a streptavidin-HRP solution containing blue-streptavidin diluent and streptavidin-HRP (750 l: 1 l) for 15 min at room temperature. Finally, they were subjected to the blue label solution. The specimens were counterstained with nuclear fast red. RESULTS Embryonic Day 18 The anlagen of the condyle and the temporal bone were recognizable as a condensation of the mesenchymal cells, and the mesenchymal cells were sparsely distributed between their anlagen (Fig. 1a, c, and e). No tissue separation occurred in future joints, indicating the absence of any formation of the articular cavity at this stage. Observation of the sections showed the presence of a few blood capillaries ﬁlled with red blood cells at the surface of the developing condyle (Fig. 1b). The endothelial cells of these capillaries exhibited a weak immunoreaction for CD31 (Fig. 1d) and RECA-1 (data not shown) at this stage. However, no blood capillaries existed in the mesenchymal tissue between the anlagen of the condyle and temporal bone (Fig. 1c). ED1 immunostaining demonstrated a sparse distribution of immunoreactive cells in the mesenchymal tissue between the anlagen of the condyle and temporal bone (Fig. 1e). These positive cells, ovoid in shape, had a rich cytoplasm and large nucleus (Fig. 1f). surface of the developing condyle. No tissue separation was found between the prospective articular disk and the developing condyle. However, we failed to ﬁnd any TUNEL-positive cells, indicative of apoptotic cells, in these areas (Fig. 1m) at this stage, in spite of the existence of several apoptotic cells at the chondro-osseous junction of mandible (Fig. 1n). The endothelial cells in some ﬂat blood capillaries immunopositive for CD31 and RECA-1 were often observed to run along the developing condyle (Fig. 1g–j). However, no blood capillary was recognizable in the area between the developing temporal bone and prospective articular disk. On the other hand, ED1-positive elements increased drastically compared with the previous stage; numerous ED1-immunoreactive cells occurred in the mesenchymal tissue between the developing temporal bone and prospective articular disk (Fig. 1k). These cells changed in their proﬁles from ovoid to spindle at this stage (Fig. 1l). Their long axis appeared to lie in parallel to the convexity of the condylar surface. However, ED1-positive cells occurred rarely near the surface of the developing condyle (Fig. 1k), which contained CD31- and RECA-1-positive endothelial cells (Fig. 1g–j). Embryonic Day 21 An apparent tissue separation was clearly recognizable between the mandibular fossa and the condensation of mesenchymal cells, i.e., prospective articular disk, indicating the commencement of formation of the upper articular cavity (Fig. 2a and e). However, no formation of such a tissue separation of the mesenchymal tissues was yet discernible in the prospective lower articular cavity (Fig. 2a and e). Even at this stage, no apoptotic cell with TUNEL reaction was identiﬁed between the temporal bone and developing condyle (data not shown). In contrast, several CD31- and RECA-1-immunopositive endothelial cells remained between the surface of the developing condyle and the prospective articular disk (Fig. 2a– c). In the vicinity of the condylar surface, these endothelial cells were observed either to run along the condylar surface or to climb up perpendicularly in the prospective ﬁbrous layer of the condyle (Fig. 2b and c). The Indianink-injected sections could conﬁrm such a characteristic distribution of the blood capillaries (Fig. 2d). A dense distribution of ED1-positive cells was discernable at the region of the forming upper articular cavity, notably an intense immunoreaction in the temporal bone (Fig. 2e). The ED1-immunoreactivity existed in neither the prospective articular disk nor the area between the condylar surface and disk, which had endothelial cells. At the upper anterior region of the articular cavity, the ED1-positive cells changed their proﬁles to appear ovoid or irregular in shape (Fig. 2f), these proﬁles being similar to those of the synovial lining cells in the mature synovium (Ikeda et al., 2004). Embryonic Day 19 Intramembranous and endochondral ossiﬁcation had already begun in the temporal bone and the developing condyle, respectively (Fig. 1g, k, and m). The mesenchymal tissues near the mandibular fossa of the temporal bone appeared to become sparse due to the expansion of the intercellular spaces of the mesenchymal cells. The slender mesenchymal cells aggregated to form a few cell layers, which might be regarded as a prospective articular disk (asterisk in Fig. 1g, k, and m), in the vicinity of the Postnatal Day 1 The upper articular cavity had expanded in the anterior-posterior direction (Fig. 2g and i). The formation of the lower articular cavity also had begun between the articular disk and the condyle, being especially prominent at the anterior and posterior regions (Fig. 2g and i). However, the articular disk was not separated from the condylar surface at the central position of the lower articular cavity. The blood capillaries with CD31- and RECA-1- ARTICULAR CAVITY FORMATION IN TMJ 911 Fig. 1. Sagittal sections of the developing rat TMJ at E18 (a–f) and E19 (g–n). H&E staining (a and b), immunolabeled with antisera against CD31 (c, d, g, and h), RECA-1 (i and j), and ED1 (e, f, k, and l), and TUNEL reaction (m and n). The left side of all ﬁgures indicates the anterior direction. a: Mesenchymal cells are observed to gather at the anlagen of the condyle (C) and the temporal bone (T). Neither the articular cavity nor the disk is formed at this stage. b: Higher magniﬁcation of the boxed area in a. A capillary with red blood cells is seen to run on the surface of the condylar anlage. c and d: CD31 immunoreactivity is found in the endothelial cells with slender proﬁles near the condylar anlage (C). e: The mesenchymal tissue between the anlagen of condyle (C) and temporal bone (T) contains a few ED1-immunopositive cells. f: In a higher-magniﬁed view, the ED1-positive cells display ovoid proﬁles, and the immunoreactions take on a granular appearance in their cytoplasm. g: Compared with the previous stage, the mesenchymal cells at the superior portion near the mandibular fossa (T) appear to become sparse due to the expansion of their intercellular spaces. Near the surface of the condylar anlage (C), the mesenchymal cells, spindle in shape, form a few cell layers (asterisk), an analogue to the future articular disk. g–j: The CD31- (g and h) and RECA-1- (i and j) positive endothelial cells running between the surface of the developing condyle (C) and the future articular disk (asterisk). RECA-1-positive endothelial cells continue to the blood capillaries at the posterior portion (j). k and l: The ED1-positive cells that display irregular or slender shapes (l) increase in number in the mesenchymal tissue between the temporal bone (T) and the future articular disk (asterisk). m and n: No TUNEL-positive cell is recognizable in the mesenchymal tissues between the temporal bone (T) and condylar anlage (C; m), although a few TUNEL-positive cells (arrows in n) are scattered only at the chondro-osseous junction of the mandible. Scale bars ⫽ 150 m (a); 20 m (b, d, f, and l); 100 m (c, e, g, k, m, and n); and 50 m (h–j). immunoreactivity drastically decreased in number at the forming lower articular cavity compared with the previous stage. Interestingly, the immunopositive endothelial cells remained in the central portion of the lower articular cavity, where no cavitation had occurred (Fig. 2g and h). In contrast, the endothelial cells diminished in the area in which the lower articular cavity had been clearly formed (Fig. 2g). Many immunopositive endothelial cells were also observed in the synovial membrane. Numerous ED1positive cells were found in the articular disk and the synovial membrane but not in the lower articular cavityforming area (Fig. 2i). Postnatal Day 3 The upper and lower articular cavity had further expanded in anterior and posterior directions (Fig. 3a). Since the lower articular cavity also had been formed at the central portion, the articular disk was completely separated from the condylar surface at this stage. A part of the synovial membrane had protruded into the articular cavity in the posterior portion of the upper articular cavity to form a synovial fold (Fig. 3a, c, and d). A few immunopositive endothelial cells remained in the ﬁbrous layer of the condylar surface (Fig. 3b) and the synovial membrane 912 SUZUKI ET AL. Fig. 2. Photomicrographs at E21 (a–f) and P1 (g–i). Sections immunolabeled for CD31 (a, b, g, and h), RECA-1 (c), ED1 (e, f, and i), and Indian-ink perfusion (d). a: An upper articular cavity (UC) has been formed between the temporal bone (T) and the articular disk anlage (asterisk). The blood capillaries with CD31 immunoreaction are found as either running along the condylar surface or invading the condyle (C). b and c: Higher magniﬁcation of the boxed area in a. The blood capillaries consisting of CD31- (b) and RECA-1- (c) positive endothelial cells along the condylar surface anastomose them ascending perpendicularly in the condylar surface. d: Several blood capillaries containing Indian ink are recognizable at the mesenchymal tissue between the condyle (C) and prospective articular disk (asterisk). These capillaries run along the condylar surface (arrow) and climb up into the condyle (arrowhead). e and f: Many ED1-positive cells are distributed around the upper articular cavity. The area between the articular disk anlage (asterisk) and the condylar surface (C) does not contain any ED1-positive cells. The ED1-positive cells in the upper articular cavity-forming area change their proﬁles from spindle to ovoid (f). g and h: The enlargement of the upper articular cavity (UC) has proceeded in the anterior-posterior direction. Formation of the lower articular cavity has also commenced, but the articular disk (D) remains in contact with the condyle (C) in the central portion of the forming lower articular cavity. In comparison with the previous stage, the CD31-positive cells have decreased in number. The capillaries entering the condylar surface perpendicularly have disappeared (h). However, many capillaries with CD31 immunoreaction are observed in both the anterior and posterior portions of the articular disk and the synovial membrane. i: In contrast to a dense distribution of ED1-positive cells in the synovial membrane and the articular disk (D), no positive cells exist in the lower articular cavity. Scale bars ⫽ 200 m (a, e, g, and i); 50 m (b– d); 20 m (f); and 40 m (h). (Fig. 3c). In contrast to the decrease of endothelial cells, ED1-immunoreactive cells increased in number daily to be distributed widely in the synovial membrane (Fig. 3d). In the synovial fold, ED1-immunopositive and -negative cells were arranged on the synovial surface (Fig. 3d), so we could easily distinguish the synovial lining layer from the sublining layer. previous stage, and ED1-positive cells with ovoid proﬁles were arranged there. The behaviors of endothelial cells and ED1-positive cells during development of the rat TMJ are given in Figure 4. Postnatal Day 5 No remarkable change in the histological structures was found at P5 except for the development of the synovial fold at the posterior-superior portion of the articular cavity. The distribution of CD31- and RECA-1-immunopositive endothelial cells was unchanged; the cells remained in the ﬁbrous layer of the condylar surface and the synovial membrane. In the surface of the synovial fold, the synovial lining cell layer thickened compared with the DISCUSSION The present immunocytochemical study was clearly able to demonstrate the formation process of the articular cavity in the rat TMJ and region-speciﬁc expression patterns of CD31, RECA-1, and ED1 immunoreactions in the cavity-forming area, suggesting the possibility of the different mechanism of the cavitations between the upper and lower articular cavities. The mechanism on the formation of the articular cavities is controversial. To date, several hypotheses have been proposed in limb joints; they include apoptosis, differential growth, vascularization, enzymatic degradation ARTICULAR CAVITY FORMATION IN TMJ 913 Fig. 3. Micrographs at P3 (a–d) and P5 (e). Stained with hematoxylin and eosin (a), CD31 (b and c) and ED1 immunoreactions (d and e). a: Both upper (UC) and lower articular cavities (LC) have been widely expanded in the anterior and posterior directions. A part of the synovial membrane protrudes into the articular cavity in the posterior portion of the upper articular cavity to form a synovial fold (arrow). b: A few ﬂat capillaries with CD31 immunoreaction (arrow) are observed in the ﬁbrous layer of the condyle (C), but the articular disk (D) lacks positive cells. c: The sublining tissue of the synovial membrane in the synovial fold contains many CD31-positive capillaries (arrows). d: Many ED1-positive cells are widely distributed in the synovial membrane. In particular, ED1-positive (arrows) and -negative cells are arranged alternatively in the synovial lining cell layer. The synovial sublining layer also has many ED1-positive cells with irregular or ovoid proﬁles (arrowheads). e: The thick synovial lining cell layer is clearly distinguishable from the sublining layer. The synovial lining (arrows) and sublining (arrowhead) layer contain several ED1-positive cells with irregular proﬁles. Scale bars ⫽ 100 m (a); 20 m (b– e). of the cavity-forming area, and mechanical inﬂuences (Andersen and Bro-Rasmusen, 1961; Mitrovic, 1977, 1978; Kawai et al., 1982; Kajikawa, 1984; Nalin et al., 1995; Abu-Hijleh et al., 1997). In spite of lack of detailed information, these hypotheses have been directly adopted on the mechanism of cavity formation in the TMJ, which is categorized in secondary cartilaginous joints such as sternoclavicular and acromioclavicular joints. For instance, many researchers have failed to ﬁnd apoptosis at the cavity-forming area in long bone joints (Ballard and Holt, 1968; Murray and Drachman, 1969; Mitrovic, 1977, 1978; Mori et al., 1995; Nalin et al., 1995; Kimura and Shiota, 1996; Kavanagh et al., 2002), while some studies have detected it (Abu-Hijleh et al., 1997; Ito and Kida, 2000). Matsuda et al. (1997) revealed no detection of apoptosis in this area by biochemical and histochemical analyses with the TUNEL method in TMJ development, consistent with the present observations. These evidences lead us to suppose considerably low possibility of the involvement of apoptosis in cavity formation in TMJ. Therefore, detailed mechanism remains unclear in the articular cavity formation of the TMJ. The present immunostaining with the CD31 and RECA-1 antibodies demonstrated the region- and stage- speciﬁc localizations of blood capillaries throughout the cavity formation process. Some reports have shown changes in vascularization during the articular cavity formation; in the knee joint of the mouse and chick embryo, the capillaries invaded the forming articular cavity from the surrounding mesenchymal tissue into the prospective articular cavity, and the endothelial-like cells lining the cavity developed slender protoplasmic projections to overspread the joint cavity (Kawai et al., 1982; Kajikawa, 1984). In TMJ development, the blood vessels have been reported to run posteroanteriorly on the lower surface of the articular disk at the early stages of articular cavity formation (Morimoto et al., 1987; Ohnuki, 2000) in the human fetus. In particular, Ohnuki (2000) found the disintegration of the blood vessels as merged with the forming lower articular cavity, suggesting that blood vessels serve as a partition between the articular disk and the condyle to produce a space for the lower articular cavity in the mesenchymal tissue. The current observations by CD31 and RECA-1 immunocytochemistry and Indian-ink perfusion techniques suppose the involvement of blood vessels in the lower articular cavity formation. Another interesting ﬁnding is the disappearance of the blood capillaries in the lower cavity-forming area after 914 SUZUKI ET AL. Fig. 4. A schematic drawing of the distribution of endothelial cells and macrophages in the developing rat TMJ. cavitation. One possible explanation may be the commencement of jaw movements because the capillaries decreased in number in the lower articular cavity after birth. We can readily suppose that the mechanical pressure from jaw movements induces the disappearance of endothelial cells. This idea is supported by ﬁndings that spinal cordinjured animals had defective articular cavities (Drachman and Sokoloff, 1966). Thus, we can regard mechanical stimulus and blood vessels as one of the essential factors to induce cavitation and to make a space, respectively. However, mechanism of the elimination process of vascular elements remains unknown. It is noteworthy that ED1 immunocytochemistry demonstrated the aggregation of monocyte/macrophage lineages in the prospective upper articular cavity in TMJ, in contrast to the occurrence of CD31- and RECA-1-positive blood capillaries in the future lower cavity. As far as we know, this is the ﬁrst report to demonstrate the aggregation of monocyte/macrophage lineages in the future upper articular cavity in TMJ. The appearance of an ED1-positive cells seems to be closely related to the period of the commencement of the upper cavity formation. The predominant distribution of ED1-positive cells suggested the presence of apoptosis in the upper cavity-forming area. In spite of careful observations, however, we failed to ﬁnd any apoptotic cells in this area, matching the ﬁndings by Matsuda et al. (1997). This means that ED1-positive cells do not participate in the phagocytosis of apoptotic cells during the formation of the upper joint cavity. On the other hand, experimental data have suggested the involvement of hyaluronic acid synthesis in joint cavity formation (Craig et al., 1990; Archer et al., 1994, 2003; Edwards et al., 1994; Pitsillides et al., 1995; Dowthwaite et al., 1998). Thus, it is better to consider that the ED1positive cells play crucial roles in the elimination of intercellular substances such as a hyaluronic acid, not apoptotic cellular debris, from the upper articular cavity-forming area. It is generally accepted that the synovial lining layer contains two kinds of lining cells, macrophage-like type A cells and ﬁbroblast-like type B cells, according to their ultrastructural conﬁgurations (Barland et al., 1962; ARTICULAR CAVITY FORMATION IN TMJ Graabæk, 1984; Nozawa-Inoue et al., 1998, 2003; Iwanaga et al., 2000). Observation of the synovial membrane in osteopetrotic (op/op) mice showed a lack of type A cells (Naito et al., 1991), strongly suggesting that type A cells are differentiated from a monocyte lineage. Our previous report (Ikeda et al., 2004) on the development of the rat TMJ revealed the ﬁrst detection of macrophage-like type A cells in the synovial lining at P3 by electron microscope. This observation suggests that the ED1-immunopositive cells arranged on the synovial surface at P3 in this study might be macrophage-like type A cells. In conclusion, the region-speciﬁc and temporal occurrences of CD31-, RECA-1-, and ED1-positive cells during development in rat TMJ suggest a possibility of a different mechanism between the upper and lower cavity formation. Further investigations are needed to clarify the mechanisms for the disappearance of endothelial cells and the roles of ED1-positive cells in articular cavity formation. ACKNOWLEDGMENTS The authors thank Mr. M. Hoshino and Mr. K. Takeuchi, Division of Oral Anatomy, Department of Oral Biological Science, Niigata University Graduate School of Medical and Dental Sciences, for their technical assistance. LITERATURE CITED Abu-Hijleh G, Reid O, Scothorne RJ. 1997. Cell death in the developing chick knee joint: I, spatial and temporal patterns. Clin Anat 10:183–200. Andresen H, Bro-Rasmusen F. 1961. Histochemical studies on the histogenesis of joints in human fetuses with special reference to the development of the joint cavities of the hand and foot. Am J Anat 108:111–122. Archer CW, Morrison H, Pitsillides AA. 1994. Cellular aspects of the development of diarthrodial joints and articular cartilage. J Anat 184:447– 456. Archer CW, Dowthwaite GP, Francis-West P. 2003. Development of synovial joints. Birth Defects Res C Embryo today 69:144 –155. Bach-Petersen S, Kjaer I, Fischer-Hansen B. 1994. Prenatal development of the human osseous temporomandibular region. J Craniofac Genet Dev Biol 14:135–143. Ballard KJ, Holt SJ. 1968. Cytological and cytochemical studies on cell death and digestion in the foetal rat foot: the role of macrophages and hydrolytic enzymes. J Cell Sci 3:245–262. Barland P, Novikoff AB, Hamermam D. 1962. Electron microscopy of the human synovial membrane. J Cell Biol 14:207–220. Craig FM, Bayliss MT, Bentley G, Archer CW. 1990. A role for hyaluronan in joint development. J Anat 171:17–23. DeLisser HM, Newman PJ, Albelda SM. 1994. Molecular and functional aspects of PECAM-1/CD31. Immunol Today 15:490 – 495. Dijkstra CD, Dopp EA, Joling P, Kraal G. 1985. The heterogeneity of mononuclear phagocytes in lymphoid organs: distinct macrophage subpopulations in rat recognized by monoclonal antibodies ED1, ED2 and ED3. Adv Exp Med Biol 186:409 – 419. Dowthwaite GP, Edwards JC, Pitsillides AA. 1998. An essential role for the interaction between hyaluronan and hyaluronan binding proteins during joint development. J Histochem Cytochem 46:641– 651. Drachman DB, Sokoloff L. 1966. The role of movement in embryonic joint development. Dev Biol 14:401– 420. Duijvestijn AM, van Goor H, Majoor GD, van Bussel E, van Breda Vriesman PJ. 1992. Antibodies deﬁning rat endothelial cells: RECA-1, a pan-endothelial cell-speciﬁc monoclonal antibody. Lab Invest 66:459 – 466. Edwards JC, Wilkinson LS, Jones HM, Soothill P, Henderson KJ, Worrall JG, Pitsillides AA. 1994. The formation of human synovial joint cavities: a possible role for hyaluronan and CD44 in altered interzone cohesion. J Anat 185:355–367. 915 González AWA. 1932. The prenatal growth of the albino rat. Anat Rec 52:117–138. Graabæk PM. 1984. Characteristics of the two types of synoviocytes in rat synovial membrane: an ultrastructural study. Lab Invest 50: 690 –702. Hsu SM, Raine L, Fanger H. 1981. Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29:557–580. Ikeda N, Nozawa-Inoue K, Takagi R, Maeda T. 2004. Development of the synovial membrane in the rat temporomandibular joint as demonstrated by immunocytochemistry for heat shock protein 25. Anat Rec A Discov Mol Cell Evol Biol 279:623– 635. Ito MM, Kida MY. 2000. Morphological and biochemical re-evaluation of the process of cavitation in the rat knee joint: cellular and cell strata alterations in the interzone. J Anat 197:659 – 679. Iwanaga T, Shikichi M, Kitamura H, Yanase H, Nozawa-Inoue K. 2000. Morphology and functional roles of synoviocytes in the joint. Arch Histol Cytol 63:17–31. Kajikawa K. 1984. Synovial membrane. In: Kajikawa K, editor. Connective tissue. Tokyo: Kanehara Press. p 467– 490. Kavanagh E, Abiri M, Bland YS, Ashhurst DE. 2002. Division and death of cells in developing synovial joints and long bones. Cell Biol Int 26:679 – 688. Kawai S, Sakai H, Oka T, Sanda F, Kubo K. 1982. On the formation of the articular cavity of the jaw articulation in chick embryos. Okajimas Folia Anat Jpn 58:467– 484. Kimura S, Shiota K. 1996. Sequential changes of programmed cell death in developing fetal mouse limbs and its possible roles in limb morphogenesis. J Morphol 229:337–346. Linck G, Porte A. 1978. B-cells of the synovial membrane: II, differentiation during development of the synovial cavity in the mouse. Cell Tissue Res 195:251–265. Matsuda S, Mishima K, Yoshimura Y, Hatta T, Otani H. 1997. Apoptosis in the development of the temporomandibular joint. Anat Embryol 196:383–391. Mérida-Velasco JR, Rodrı́guez-Vázquez JF, Jiménez Collado J. 1993. The relationships between the temporomandibular joint disc and related masticatory muscles in humans. J Oral Maxillofac Surg 51:390 –395. Mérida-Velasco JR, Rodrı́gez-Vázquez JF, Mérida-Velasco JA, Sánchez-Montesinos I, Espı́n-Ferra J, Jiménez-Collado J. 1999. Development of the human temporomandibular joint. Anat Rec 255:20 –33. Mitrovic DR. 1977. Development of the metatarsophalangeal joint of the chick embryo: morphological, ultrastructural and histochemical studies. Am J Anat 150:333–347. Mitrovic D. 1978. Development of the diarthrodial joints in the rat embryo. Am J Anat 151:475– 485. Mori C, Nakamura N, Kimura S, Irie H, Takigawa T, Shiota K. 1995. Programmed cell death in the interdigital tissue of the fetal mouse limb is apoptosis with DNA fragmentation. Anat Rec 242:103–110. Morimoto K, Hashimoto N, Suetsugu T. 1987. Prenatal developmental process of human temporomandibular joint. J Prosthet Dent 57:723–730. Murray PD, Drachman DB. 1969. The role of movement in the development of joints and related structures: the head and neck in the chick embryo. J Embryol Exp Morphol 22:349 –371. Naito M, Hayashi S, Yoshida H, Nishikawa S, Shuitz LD, Takahashi K. 1991. Abnormal differentiation of tissue macrophage populations in “osteopetrosis” (op) mice defective in the production of macrophage colony-stimulating factor. Am J Pathol 139:657– 667. Nalin AM, Greenlee TK, Jr Sandell LJ. 1995. Collagen gene expression during development of avian synovial joints: transient expression of type II and XI collagen genes in the joint capsule. Dev Dyn 203:352–362. Nozawa-Inoue K, Takagi R, Kobayashi T, Ohashi Y, Maeda T. 1998. Immunocytochemical demonstration of the synovial membrane in experimentally induced arthritis of the rat temporomandibular joint. Arch Histol Cytol 61:451– 466. 916 SUZUKI ET AL. Nozawa-Inoue K, Amizuka N, Ikeda N, Suzuki A, Kawano Y, Maeda T. 2003. Synovial membrane in the temporomandibular joint: its morphology, function and development. Arch Histol Cytol 66:289 – 306. Ögütcen-Toller M, Juniper RP. 1993. The embryologic development of the human lateral pterygoid muscle and its relationships with the temporomandibular joint disc and Meckel’s cartilage. J Oral Maxillofac Surg 51:772–779. Ögütcen-Toller M, Juniper RP. 1994. The development of the human lateral pterygoid muscle and the temporomandibular joint and related structures: a three-dimensional approach. Early Hum Dev 39:57– 68. Ögütcen-Toller M. 1995. The morphogenesis of the human discomalleolar and sphenomandibular ligaments. J Craniomaxillofac Surg 23:42– 46. Ohnuki H. 2000. On the formation of the temporomandibular joint cavity in the human fetus. Kaibogaku Zasshi 75:325– 336. Okada Y, Nakanishi I, kajikawa K. 1981. Ultrastructure of the mouse synovial membrane: development and organization of the extracellular matrix. Arthritis Rheum 24:835– 843. Pitsillides AA, Archer CW, Prehm P, Bayliss MT, Edwards JC. 1995. Alterations in hyaluronan synthesis during developing joint cavitation. J Histochem Cytochem 43:263–273. Symons NB. 1952. The development of the human mandibular joint. J Anat 86:326 –332. Van der Linden EJ, Burdi AR, de Jongh HJ. 1987. Critical periods in the prenatal morphogenesis of the human lateral pterygoid muscle, the mandibular condyle, the articular disc, and medial articular capsule. Am J Orthod Dentofac Orthop 91:22–28. Walmsley R. 1964. Syndesmology or arthrology: syndesmology or arthrology. In: Romanes GJ, editor. Cunningham’s textbook of anatomy, 10th ed. London: Oxford university press. p 213–264. Williams KC, Zhao RW, Ueno K, Hickey WF. 1996. PECAM-1 (CD31) expression in the central nervous system and its role in experimental allergic encephalomyelitis in the rat. J Neurosci Res 45:747–757.