Abstract
Apoptosis of dental pulp cells of rat incisors was investigated by the TUNEL method and electron microscopy. The results showed that a considerable amount of apoptosis occurred in the pulp, increasing in extent with incisal direction. OX6, ED1, and ED2 antibodies were used to detect macrophages and dendritic cells in combination with immunoelectron microscopy. Apoptotic fragments were eliminated mainly by MHC Class II-expressing cells, including dendritic cells positive for the OX6 antibody, and by MHC Class II-negative macrophages. Macrophages and dendritic cells positive for OX6, ED1, or ED2 increased from the apical to incisal direction of the incisor. These results indicate that apoptosis contributes to normal pulp formation and maintenance.
I
Incisor dental pulp provides a space for newly formed dentin to attach to the dentin wall and for the vascular and nerve supply to odontoblasts and the remaining pulp. TUNEL-positive cells have been reported in the pulp of rodent teeth (Bronckers et al. 1996; Vaahtokari et al. 1996). However, their extent and characterization remain to be clarified.
Macrophages and MHC Class II-expressing cells in dental pulp have been widely studied. By using probes for MHC Class II antigens and several markers for dendritic cells and macrophages, these cells have been localized in human (Jontell et al. 1987), rat (Jontell et al. 1988,1991; Okiji et al. 1992,1996; Ohshima et al. 1994) and mouse (Nagahama et al. 1998) dental pulp. Especially in the case of rat teeth, several useful antibodies have been found: the OX6 antibody directed to MHC Class II antigens in dendritic cells and macrophages, the ED1 antibody directed to the phagolysosomal membrane of dendritic cells and macrophages, and the ED2 antibody, which is used as a differentiation marker of macrophages but not dendritic cells (Dijkstra et al. 1985). These studies have revealed substantial numbers of macrophages and dendritic cells in dental pulp (Jontell et al. 1988; Jontell and Bergenholtz 1992; Okiji et al. 1992,1996). Because there are OX6-positive cells and considerable numbers of ED1-positive and ED2-positive macrophages in the pulp of the first molar of rats between 3 and 4 weeks after birth, when active tooth eruption still occurs (Okiji et al. 1996), these immunocompetent cells have been interpreted to be “strategically” located to detect “incoming” antigens, e.g., bacterial antigens enter via exposed dentin. Alternatively, the presence of immunocompetent cells in young tooth pulp is believed to play a role in eliminating apoptotic fragments to create the structure of dental pulp.
This study focused on whether or not apoptotic cell death really occurs in continuously erupting incisors of rats and on which cells eliminate apoptotic fragments. The results show that a considerable amount of apoptosis occurs in the pulp and that it increases with incisal direction. Apoptotic cell fragments are eliminated mainly by either MHC Class II-expressing cells, including dendritic cells, or MHC Class II-negative macrophages. Fibroblasts and odontoblasts are rather minor factors.
Materials and Methods
Sixteen male Wistar rats (7-8 weeks, 180-260 g) (Jcl Wistar, Clea Japan, Tokyo, Japan; institutional guidelines were followed) were used. Twelve rats were used for light microscopic immunocytochemistry and TUNEL tests and four rats were used for electron microscopy, one for conventional fixation, and three for fixation of immunocytochemistry. They were anesthetized with sodium pentobarbital (Nembutal; Abbott, North Chicago, IL) and perfused with an appropriate fixative (described below) via the left ventricle for 10-15 min. Maxillary and mandibular incisors were dissected and then immersed in the same fixative at 4C for 2 hr. The fixative used for immunoelectron microscopy of the OX6 antibody (Serotec; Oxford, UK) was composed of 4% paraformaldehyde and 0.5% glutaraldehyde in 0.1 M phosphate buffer solution (PB), pH 7.2. The fixative used for light microscopic immunocytochemistry of ED1 (Serotec) and OX6 antibodies was composed of 4% paraformaldehyde in 0.1 M PB. For the ED2 antibody (Serotec), three rats were sacrificed by diethyl ether inhalation and their maxillary and mandibular incisors were dissected. Their teeth were split open and dental pulp was immediately frozen in a cryotome (HM505E; Microm, Walldorf, Germany). Fixation protocols for conventional electron microscopy were as described elsewhere (Nishikawa and Sasaki 1995). After being washed with 0.1 M PB, the specimens were decalcified with 5% EDTA in 0.1 M PB in a cold room for 3-4 weeks, except for freshly prepared dental pulp.
Immunocytochemistry
Decalcified incisors and freshly prepared dental pulp were cut longitudinally (6-8 μm thick) using the cryotome. The cryosections were laid on glass microscopic slides and labeled at room temperature (RT) for 30 min with OX6, ED1, or ED2 antibodies diluted to 1:100, 1:50, or 1:50, respectively, with 1% bovine serum albumin in PBS (1% BSA-PBS). Cryosections of freshly prepared specimens were fixed on slides with pure acetone at 4C for 5 min. Before immunolabeling, cryosections were processed with 0.6% hydrogen peroxide in 80% methanol to inhibit endogenous peroxidase activity at RT for 20 min and 1% BSA-PBS to block nonspecific adsorption at RT for 30 min. The specimens were further labeled at RT for 30 min with HRP-conjugated anti-mouse IgG (Cappel; West Chester, PA) diluted 1:100 with 1% BSA-PBS. They were then developed with an AEC substrate chromogen system (AEC; 3-amino-9-ethylcarbazole as chromogen; DAKO, Carpinteria, CA) or HRP conjugate substrate kit (HSK; 4-chloro-1-naphthol as chromogen; BioRad, Hercules, CA). They were further labeled with Hoechst 33342 (1 μg/ml in PBS; Molecular Probes, Eugene, OR) for DNA localization at RT for 20 min. Control experiments were processed in the same way as described above, except the primary antibody was replaced by 1% BSA-PBS.
For double labeling with OX6 and ED1 antibodies, cryosections were labeled with either primary antibody, followed by the HRP-conjugated secondary antibody and then HSK or AEC development. The labeled sections on the slides were incubated in 0.1 M glycine-HCl, pH 2.2, for 1 hr to remove the labeled antibodies. The second primary antibody was then labeled in the same way as described above. Control experiments were performed in two ways: (a) the first and second primary antibodies were replaced by 1% BSA-PBS alone, or (b) sections were labeled by one of the primary antibodies and the HRP-conjugated secondary antibody, and then incubated in 0.1 M glycine-HCl, pH 2.2, for 1 hr and developed by HSK or AEC.
For pre-embedding immunoelectron microscopy, apical halves of decalcified incisors, which included pulp regions facing secretory and early maturation ameloblasts, were cryosectioned (20-30 μm thick). The sections were labeled at 4C overnight with OX6 diluted 1:100. After being washed with PBS, they were further labeled with HRP-conjugated antimouse IgG (Cappel) diluted 1:100 with 1% BSA-PBS. After being washed with 0.05 M Tris-HCl buffer (pH 7.6) twice for 15 min each, specimens were fixed for 1 hr with 1% glutaraldehyde in a Tris-HCl buffer. They were processed with diaminobenzidine (DAB, 0.2 mg/ml in 0.05 M Tris-HCl) at RT for 30 min and then with DAB plus 0.005% H2O2 for 5 min. The specimens were postfixed with osmium tetroxide, dehydrated with a series of ethanol, and embedded in TAAB Epon 812 (TAAB; Reading, UK). Ultrathin sections were stained with lead citrate alone and examined with a Jeol 1200EXII electron microscope. Control experiments were processed in the same way as described above, except that the primary antibody was replaced by 1% BSA-PBS.
TUNEL Method
Paraformaldehyde-fixed and decalcified specimens were cryosectioned (6-8 μm thick). The sections were washed with 0.1 M cacodylate buffer (pH 7.3), and incubated for 5 min with 1 × labeling buffer provided by the supplier (TdT in situ detection kit; Genzyme, Cambridge, MA). They were labeled with a mixture composed of one part 0.25 mM biotinylated dNTP, one part 50 mM Co2+, one part 15 U/μl TdT, and 50 parts 1 X labeling buffer at 37C for 1 hr. For positive control experiments, some sections were pretreated with DNase I (0.2 mg/ml, Grade I; Boehringer, Mannheim, Germany) at RT for 30 min, and then incubated with the labeling reaction mixture. For negative control experiments, other sections were incubated with an incomplete reaction mixture lacking either TdT enzyme or biotinylated dNTP. The reaction was completed with a 1 X stop buffer provided by the supplier and then washed with streptavidin-fluorescein (Genzyme) or streptavidin-HRP (Genzyme) at RT for 20 min. HRP-labeled sections were developed by HSK (Bio-Rad). The labeled sections were further incubated with 1 μg/ml Hoechst 33342 at RT for 20 min.
Numerical Analysis
The frequency of apoptosis in the dental pulp was examined using five male rats (7-8 weeks, 180-250 g). Apoptotic nuclear fragments were examined by the TUNEL method. Pulp regions were divided into six zones: Zone I, pulp zone facing early secretory ameloblasts (n = 7); Zone II, pulp zone facing middle and late secretory ameloblasts (n = 6); Zone III, pulp zone facing transitional ameloblasts (n = 8); Zone IV, pulp zone facing early maturation ameloblasts (n = 9); Zone V, pulp zone facing late maturation ameloblasts (n = 6); Zone VI, pulp zone facing pigmented ameloblasts close to reduced enamel epithelia (n = 6). The number of TUNEL-positive cells per unit area was counted using photographic prints enlarged 210 times (330,000 μm2) in each pulp region. At least two different rats were used for counting of each pulp region.
For counting phagocytes that engulfed apoptotic nuclear fragments, sections from three male rats (8 weeks, 220-240 g) were doubly labeled with OX6 and Hoechst 33342. Cells engulfing Hoechst dye-positive apoptotic nuclear fragments were counted randomly and examined to see whether they were OX6-positive or -negative.
Results
Distribution of TUNEL-positive Cells and Frequency of Apoptosis in Dental Pulp
The developmental stages of pulp were roughly classified on the basis of ameloblasts they encounter: (I) pulp zone facing early secretory ameloblasts with thin enamel; (II) middle and late secretory ameloblasts; (III) transitional ameloblasts; (IV) early maturation ameloblasts; (V) late maturation ameloblasts; and (VI) pigmented ameloblasts close to reduced enamel epithelial cells. TUNEL-positive cells were rare in the apical end of the pulp and in Zone I and were increased in Zones II-VI (Figure 1). TUNEL-positive cells were rare in the odontoblast layer in Zones III and IV, and most of the positive cells were scattered throughout the rest of pulp (Figure 2). However, in Zones V and VI, positive cells were noted in the odontoblast layer and in the rest of the pulp (Figure 2).
Frequency of TUNEL-positive cells in rat incisor pulp from the apical pulp (left) to the incisal pulp (right). (Abscissa) Pulp regions: Zone I, pulp zone facing early secretory ameloblasts (n = 7); Zone II, pulp zone facing middle and late secretory ameloblasts (n = 6); Zone III, pulp zone facing transitional ameloblasts (n = 8); Zone IV, pulp zone facing early maturation ameloblasts (n = 9); Zone V, pulp zone facing late maturation ameloblasts (n = 6); Zone VI, pulp zone facing pigmented ameloblasts close to reduced enamel epithelia (n = 6). (Ordinate) TUNEL-positive cell number per unit area (mm2) of each zone in the dental pulp. The bar in each column shows the standard deviation. p values of Student's t-test between two different zones are as follows: p<0.01 between I and II; p<0.001 between IV and V; and p<0.2 between V and VI.
TUNEL micrographs (a,c) and corresponding Hoechst 33342 fluorescence micrographs (b,d). Double labeling. TUNEL-positive cells are visualized as dark dots by streptavidin-HRP and HSK (arrowheads). (a,b) Zone II, TUNEL-positive cells are limited in the central pulp. Peripheral odontoblast layer lacks positive cells. (c,d) Zone VI, TUNEL-positive cells are scattered throughout the entire pulp zone, including the odontoblast layer. Broken lines show the boundary between the odontoblast layer and the rest of the pulp zone. Bars = 100 μm.
When dental pulp was doubly labeled by TUNEL and Hoechst 33342 for total DNA labeling, small globular structures were labeled by both probes, indicating that those globular structures were apoptotic nuclear fragments (Figure 3). Dental pulp facing secretory, transitional, and early maturation ameloblasts were examined by conventional electron microscopy. Typical apoptotic figures were observed in the cell fragmentation and nuclear chromatin condensation in a crescent formation in the periphery of the nuclei (Figure 4). However, it was difficult to identify the origin of the apoptotic cells (Figure 4). Apoptotic fragments were often ingested by the macrophage-like phagocytes but rarely by fibroblastic cells. There were dark cells, with many processes from the cell bodies, which were distinguishable from fibroblasts and macrophages (Figure 4). These were considered to be dendritic cells.
Fluorescence micrographs (a,b) and corresponding phase-contrast micrograph (c). TUNEL-positive nuclear fragments (a) were also detected by DNA staining with Hoechst 33342 (b). Double labeling. TUNEL-positive cells are visualized as bright dots by streptavidin-fluorescein. Arrows show the same apoptotic nuclear fragments. In b, TUNEL-positive or similar DNA-positive globular fragments are closely associated with a large nucleus, indicating engulfment of apoptotic bodies by some unknown cell located near a blood vessel (c). Bar = 10 μm.
Conventional electron micrographs of dental pulp. (a) A typical apoptotic cell (asterisk) and a dendritic cell (D). The dendritic cell extends a cell process to the apoptotic cell. (b) A macrophage (M) contains many lysosomes. (c) A macrophage (M) engulfs an apoptotic fragment (arrow). F, fibroblast. Bars = 2 μm.
Immunohistochemistry of Dental Pulp by OX6, ED1, and ED2 Antibodies
Whereas OX6-, ED1-, or ED 2-positive cells were rare in immature apical pulp, they increased in proportion to incisal direction, showing more macrophages and dendritic cells in the incisal pulp than the apical pulp (Figure 5). OX6 and ED2 continuously labeled the cell surfaces, and ED1 labeled cytoplasm in a patchy pattern. OX6-labeled cells were distributed around the blood capillaries located between odontoblasts, and also in the rest of the pulp. Double labeling of OX6 and ED1 antibodies showed considerable numbers of doubly positive cells and some singly labeled cells. When neighboring thin sections were labeled with either OX6 or ED2, the number of OX6-positive cells was smaller than that of ED2-positive macrophages (Figure 5).
Neighboring sections of dental pulp of an incisor labeled with OX6 (a,c,e) and ED2 (b,d,f) followed by HRP-conjugated secondary antibodies and AEC. Zone of a, c, and e exactly corresponds to zone of b, d and f. Acetone fixation. (a,b) Apical pulp zone; (c,d) middle pulp zone; (e,f) incisal pulp zone. The number of cells positive for each antibody is greater in the more incisally located pulp zone than in the apically located pulp zone. Bars = 100 μm.
Macrophages and dendritic cells localized by immunohistochemical staining using OX6 antibodies for MHC Class II antigen were combined with Hoechst dye staining to examine the nature of cells that eliminate apoptotic fragments in dental pulp. Both OX6-positive and -negative cells engulfed apoptotic nuclear fragments stained by Hoechst dye (Figure 6). Numerical analysis showed that the proportion of OX6-negative cells engulfing apoptotic bodies was greater than that of OX6-positive cells engulfing apoptotic bodies (Table 1). Some OX6-positive cells exhibited many slender processes and were considered to be dendritic cells (Figure 6). ED1-positive cells also ingested apoptotic nuclear fragments (Figure 7).
OX6 immunocytochemistry visualized by HRP-conjugated secondary antibodies and AEC (a,c,e,g) and corresponding Hoechst 33342 labeling (b,d,f,h). Double labeling. OX6-positive (a-d) or -negative (e-h) cells engulf apoptotic nuclear fragments (arrowheads). Arrows show the nuclei of phagocytes. An OX6-positive cell (a,b) is highly dendritic. Bars = 10 μm.
An ED1-positive cell in dental pulp ingesting apoptotic nuclear fragments (arrowheads). Arrows show a nucleus of phagocyte. (a) ED1 immunocytochemistry visualized by HRP-conjugated secondary antibodies and AEC. (b) DNA labeling of the same section by Hoechst 33342. Bar = 10 μm.
Cells phagocytosing apoptotic nuclear fragments in the dental pulp of the rat incisors a
Apoptotic nuclear fragments were detected by Hoechst 33342 labeling.
Ll, mandibular incisor; Ul, maxillary incisor.
Immunoelectron Microscopy of Dental Pulp Examined with OX6 Antibodies
OX6-positive cells were located close to the blood capillaries in the odontoblast layer but were distributed only sparsely in the rest of the pulp. Some of the OX6-positive cells had similar morphological characteristics, including eccentric nuclear localization, a slender cell body, thin processes projecting from the cell body, and fewer lysosomal inclusions, whereas other weakly OX6-positive cells that had more lysosomal inclusions were also present in the dental pulp. The former was considered to be a dendritic cell and the latter to be a macrophage. A number of OX6-negative macrophages were also found in the dental pulp, and these were distinguished from fibroblastic cells by their larger proportion of heterochromatin in nuclei and the presence of blebs at the cell periphery (Figures 8b and 8c). Dendritic cells engulfed apoptotic bodies (Figure 8a). OX6-negative macrophages also frequently ingested apoptotic bodies (Figure 8b). Fibroblastic cells only occasionally ingested cell debris (Figure 8c).
Immunoelectron micrographs of dental pulp labeled with OX6 antibodies followed by HRP-conjugated secondary antibodies and DAB/H2O2. (a) A dendritic cell near a blood vessel is heavily labeled with OX6 in the plasma membrane and contains several apoptotic bodies (asterisks). (b) An OX6-negative macrophage contains apoptotic nuclear fragments (asterisks). (c) A fibroblastic cell engulfs cell debris (arrow). N, nuclei. Bars = 2 μm.
Discussion
During mouse tooth morphogenesis, many TUNEL-positive cells are found in the enamel organs and dental follicles of early tooth germs (Vaahtokari et al. 1996). It has been interpreted that apoptotic events are involved in tooth shape formation, in addition to extensive proliferation activities during these stages. On the other hand, dental papillae contained very few TUNEL-positive apoptotic figures (Vaahtokari et al. 1996). In the postnatal incisors of mice and hamsters, dental pulp contained TUNEL-positive cells (Bronckers et al. 1996). In this study, TUNEL-positive cells were examined throughout entire incisors. It was found that very few TUNEL-positive cells were observed in the pulp at the apical end and that of the thin enamel layer but that they increased in the pulp facing middle-thickness enamel during the secretory ameloblast stage. This tendency continued until the incisal end of the pulp. OX6-, ED1-, or ED2-positive cells also increased in proportion to incisal direction. Consistent with the results described above, typical apoptotic figures and their engulfment by other cells were often observed in the dental pulp facing secretory ameloblasts and early maturation ameloblasts by conventional electron microscopy.
Fibroblasts and odontoblasts appear to play a limited role in the elimination of apoptotic cell fragments. Macrophages and OX6-positive cells, including dendritic cells, may play a major role in their elimination. On the basis of numerical analysis, OX6-negative cells, probably mostly macrophages, contributed to apoptotic fragment elimination more than OX6-positive cells. It is well known that macrophages are scavenger cells that phagocytose apoptotic fragments. On the other hand, dendritic cells have been reported to be quiescent cells without extensive phagocytosis. However, recent studies have shown that juvenile dendritic cells are able to phagocytose and, in fact, ingest apoptotic fragments to present antigens to cytotoxic T-lymphocytes via MHC Class I, and that dendritic cells are a part of the regulation of self reactivity in the periphery (Matsuno et al. 1996; Inaba et al. 1997; Albert et al. 1998; Banchereau and Steinman 1998). In our previous report, apoptotic fragments that were epithelial cells were phagocytosed by MHC Class II-positive macrophage-like cells (Nishikawa and Sasaki 1996). On the basis of these results, it would be possible to generalize a close relationship between apoptosis and immunological responses.
Several questions remain to be clarified. First, which kind of pulp cell actually undergoes apoptosis? The pulp at the incisal end exhibited a number of apoptotic figures both in the odontoblast layer and in the remaining pulp, indicating that at least odontoblasts may undergo apoptosis. The nature of apoptotic cells in the remaining pulp is unknown. Second, what triggers apoptosis in the pulp? It is possible that compressive forces occurring during pulp cavity reduction in the incisal direction may stimulate mechanical stress-related programmed cell death (Cheng et al. 1995; Leri et al. 1998). Third, MHC Class II-positive cells which ingest apoptotic bodies are considered to move to the regional lymph nodes where antigen presentation is likely to occur. It has been reported that isolated pulp cell populations rich in dendritic cells stimulated T-cell proliferation in vitro (Jontell et al. 1994). Any evidence, however, suggesting lymphatic pathways from pulp to regional lymph nodes has not been found in dental pulp. In conclusion, apoptosis contributes to normal pulp formation and maintenance. MHC Class II-positive cells, including dendritic cells and macrophages, and -negative macrophages are major constituents that eliminate apoptotic fragments.