| Sign In to gain access to subscriptions and/or personal tools. |
TGF-β1 Induces Accumulation of Dendritic Cells in the Odontoblast Layer
1 Laboratory of Development of Dental Tissues, EA MENRT 1892, IFR 62, Faculty of Odontology, Lyon 1 University, G. Paradin Str., 69372 Lyon Cedex 08, France; and Correspondence: * corresponding author, farges{at}laennec.univ-lyon1.fr
TGF-β1 released from dentin degraded by bacterial or iatrogenic agents is suspected to influence dental pulp response, including the modulation of cell migration. To determine the consequences of TGF-β1 action on pulp immune cells, we analyzed, by immunohistochemistry, the effect of transdentinally diffusing TGF-β1 on their localization in a human tooth slice culture model. TGF-β1 induced an accumulation of HLA-DR-positive cells in both odontoblast and subodontoblast layers of the stimulated zone. Together with HLA-DR, these cells co-expressed Factor XIIIa and CD68, two features of immature antigen-presenting dendritic cells (DC), as well as the TGF-β1 specific receptor TβRII. In contrast, no effect could be detected on the localization of either mature DC-LAMP-positive DC or of T- and B-lymphocytes. Analysis of these data suggests that TGF-β1 released from dentin degraded by bacterial or iatrogenic agents could be involved in the immune response of the dental pulp resulting from tooth injury.
Key Words: tooth pulp • immune response • HLA-DR • TGF receptor • chemokine
Dental pulp immune defense and repair are the result of successive and interrelated processes—including chemotaxis, proliferation, angiogenesis, extracellular matrix remodeling, and cell differentiation—that ultimately lead to tertiary dentin formation. How these processes are triggered at the molecular level remains to be elucidated, and the identification of specific controlling factors is a major challenge for future therapeutic treatments (Melin et al., 2000; Smith and Lesot, 2001). Growth factors released from dentin which is degraded by bacterial or iatrogenic agents might influence pulp response and healing (Finkelman et al., 1990; Magloire et al., 2001; Smith and Lesot, 2001). Among these molecules, Transforming Growth Factor-β1 (TGF-β1) has the ability to function as an inducer of tertiary dentin production (Tziafas et al., 1998) as well as a potent pulpal immunosuppressor (D’Souza et al., 1998). The stimulatory function of TGF-β1 on tertiary dentin synthesis is beginning to be elucidated (Smith and Lesot, 2001), but little is known about the role of this growth factor in the immune response of the pulp to carious infection, even if TGF-β1 is considered as a key player in inflammatory and immune reactions (Ashcroft, 1999; Wahl, 1999). Generally, TGF-β1 has a pro-inflammatory function during the initial stages of inflammation, while having anti-inflammatory effects during the later stages. The pro-inflammatory properties of TGF-β1 include immune cell recruitment, adhesion, induction of matrix metalloproteinase secretion, and activation, whereas inflammation is suppressed through repression of lymphocyte proliferation and inhibition of antigen-presenting dendritic cells (DC) and macrophage activation (Strobl and Knapp, 1999; Wahl, 1999). When dentin is being destroyed by caries or operative procedures, immature DC accumulate in the odontoblast layer and the subjacent pulp tissue close to the lesion in locations strategic to sample foreign antigens entering the dentinal tissue (reviewed in Jontell et al., 1998; Sakurai et al., 1999). Immature DC present in the interstitial (connective) tissues are characterized by a dendritic morphology, constitutive expression of class II major histocompatibility complex molecules that confer on them their antigen-presenting capacity, and expression of Factor XIIIa and CD68 markers (Banchereau and Steinman, 1998; Liu, 2001). After capture of foreign antigens at the dentin-pulp interface, DC migrate, while undergoing a process of maturation, via the afferent lymphatic to regional lymph nodes, to stimulate naïve T-lymphocytes, thus initiating a primary immune response (Banchereau and Steinman, 1998). Factors favoring the accumulation of immature DC in the peripheral pulp in response to dentin injury are unknown, but TGF-β1 diffusing intratubularly from damaged dentin could be implicated in this process, given its crucial role in the control of immune response and DC behavior (Strobl and Knapp, 1999). Here, we used a culture system of thick-sliced human teeth allowing for the transdentinal diffusion of TGF-β1 (Melin et al., 2000; Lucchini et al., 2002) to analyze, by immunostaining, the distribution and activation status of DC in odontoblast, subodontoblast, and pulp core compartments.
Tissue Culture Thirty human third molars were collected from 15- to 18-year-old patients (with their informed consent) and following an informed-consent protocol approved by the Local Ethics Committee. Samples presented no clinical or radiographic evidence of caries and no histologic evidence of tissue inflammation. Tooth tissue cultures were performed as previously described (Melin et al., 2000; Lucchini et al., 2002). Briefly, teeth were cut into 3 750-µm-thick slices. One slice was immediately fixed and processed for immunohistochemistry, whereas the other 2 slices were used for culture. In this case, a small polypropylene tube was glued, with cyanoacrylate glue, onto the dentin close to a pulp horn (Fig. 1  ). Slices were placed in 12-well culture plates (Falcon, Becton Dickinson, Oxford, England) and covered with 1 mL of Eagle’s Basal Medium (Gibco BRL, Life Technologies Inc., Grand Island, NY, USA) containing ascorbic acid (50 mg/mL), 2% fetal calf serum (Roche Diagnostics, Mannheim, Germany), and antibiotics. Tubes were filled with 50 µL of this culture medium, supplemented or not (for controls) with 20 ng/mL human recombinant TGF-β1 (BioVision Research Products, Palo Alto, CA, USA). Slices were cultured for 3–4 days without medium change to limit the diffusion of the factor to the nearest pulp horn.
Immunohistochemistry Immunohistochemistry was performed by routine procedures as previously described (Lucchini et al., 2002). Slices were fixed in 4% paraformaldehyde-PBS solution, immersed in 7% saccharose-PBS, then in 15% saccharose-8% glycerol-PBS. The hard tissue was gently discarded, after which the pulpal tissue was embedded in Tissue Tek compound (EMS, Washington, PA, USA) and frozen in liquid-nitrogen-cooled isopentane. Cryostat serial sections (10 µm) were treated with 0.3% H2O2-methanol, incubated in PBS containing normal goat or normal horse serum (dilution 1/50), then with specific polyclonal or monoclonal antibodies (Table 1 ). For antibody detection, sections were treated with immunoperoxidase with an Elite Vectastain kit used according to the protocol of the manufacturer (Vector Labs, Burlingame, CA, USA) and stained with di-amino-benzidine. For fluorescence detection, double-immunostained sections were treated with both FITC-conjugated anti-mouse and rhodamine-conjugated anti-rabbit IgG antibodies (BioRad, Hercules, CA, USA, and Sigma, St. Louis, MO, USA, respectively) (Melin et al., 2000).
Histomorphometry HLA-DR-positive dendritic cell density was evaluated in non-cultured human tooth slices and in TGF-β1-stimulated and non-stimulated slices by means of an image analysis workstation (x 25 objective, Biocom 200, France)(AQ) equipped with specialized software (Histo, Biocom) (Melin et al., 2000). Twenty teeth originating from seven patients were used, each given one slice for in vivo analysis, one for TGF-β1 stimulation, and one for control. Three pulpal areas were analyzed from HLA-DR-immunostained thin sections, namely, the odontoblast and the subodontoblast layers, and the pulp core tissue situated just beneath the subodontoblast layer. Two or three random fields were examined for each area. Results are given as the mean of cell density (cell number/0.01 mm2) for each area, and statistical analysis was performed by Student’s t test (p  0.05).
The experimental approach used allowed for the intradentinal diffusion of TGF-β1 from the base of the tube glued onto the dentin to the closest pulp horn (Melin et al., 2000) (Fig. 1  ). In vivo—that is, in non-cultured tooth slices—HLA-DR-positive DC were detected in the whole pulp tissue at a higher density in the subodontoblast layer (Fig. 2a) and along nerve fibers (Fig. 2b). In some sections, rare B-lymphocytes, stained with anti-CD19 antibody, were present in the central core of the pulp (Fig. 2c). Immunolocalization of T-lymphocytes with anti-CD3 antibody showed some positive cells scattered in the pulp tissue (Figs. 2d, 2e). Rare maturing DC-LAMP-positive DC were localized close to the pulp chamber floor and in the central part of the forming roots but not in the peripheral subodontoblast and odontoblast layers (Figs. 2f, 2g).
In slices stimulated with TGF-β1, DC density was increased in odontoblast and subodontoblast layers (Fig. 2h ). Statistical analysis revealed increases of 125% in the odontoblast layer and 38% in the subodontoblast layer (Table 2). No significant increase was detected in the pulp core. In control tooth slices, no significant variation was observed in the odontoblast layer, while DC density decreased in the subodontoblast layer and pulp core (Fig. 2i, Table 2). The distribution of T- and B-lymphocytes and mature DC in TGF-β1-treated sections was similar to that observed in vivo and in non-stimulated control tooth slices (not shown). Double-immunostaining analysis showed the presence of HLA-DR/Factor XIIIa-positive (Figs. 2j, 2k) and Factor XIIIa/CD68-positive (Figs. 2l, 2m) immature DC in the whole pulp, including the stimulated area, whereas Factor XIIIa/DC-LAMP-positive mature cells (Figs. 2n, 2o) were detected only in the pulp close to the floor and in roots. Co-localization of CD68 with the type II TGF-β1 receptor (TβRII) indicated that immature DC expressed this receptor (Figs. 2p, 2q), whereas mature DC-LAMP-positive cells present in the deep pulp failed to do so (Figs. 2r, 2s).
TGF-β1 is a pleiotropic growth factor that influences immune response and repair in mammalian tissues by controlling cell proliferation, chemotaxis, extracellular matrix synthesis and degradation, cell differentiation, and apoptosis (O’Kane and Ferguson, 1997; Wahl, 1999). We recently reported that transdentinally diffusing TGF-β1 induces the accumulation of non-proliferating cells in the pulp periphery, suggesting migration of cells from the pulp core (Melin et al., 2000). Here we show that DC likely represent at least part of these recruited cells, based on the observation that the density of HLA-DR-positive cells significantly increased in odontoblast and subodontoblast layers of the pulp horn stimulated by TGF-β1. Cells expressing HLA-DR molecules in sound tissues are DC, B-lymphocytes, and some macrophages (Dong et al., 2001). In our experiments, B-lymphocytes were not detected in the odontoblast and subodontoblast layers with or without TGF-β1 stimulation. Only rare CD19-positive B-lymphocytes were localized in the pulp core, as previously reported (Sakurai et al., 1999). While the vast majority of sound pulp resident macrophages fail to express detectable levels of HLA-DR molecules (Izumi et al., 1996; Okiji et al., 1996; Jontell et al., 1998), few activated ones could also have been stained with our anti-HLA-DR antibody, in addition to DC. However, in agreement with previous data indicating that about 90% of HLA-DR-expressing pulp cells with dendritic morphology co-express immunoreactivity to CD68 and Factor XIIIa (Jontell et al., 1998), double-immunostaining analysis indicated that most HLA-DR-positive cells were also positive for these two markers, confirming the identity of DC over macrophages.
In our in vitro model, DC, or their precursors, were pre-existing in the pulp, given that there is no vascularization in the tooth slice culture system, and different mechanisms, not mutually exclusive, may have contributed to the apparent accumulation of DC in the odontoblast and subodontoblast layers: DC precursors, including monocytes residing in the tissue, may have differentiated into migratory immature DC under the influence of TGF-β1 (Randolph et al., 2002). Furthermore, TGF-β1 may have modified the spatial distribution of DC in the tooth. Indeed, for many years, TGF-β1 has been known not only to inhibit DC proliferation, maturation, and apoptosis (Strobl and Knapp, 1999; Liu, 2001), but also to stimulate migration of human immature DC in vitro by increasing the expression of specific chemokine receptors constitutively present in the DC membrane (Sato et al., 2000). The migratory responses to Macrophage Inflammatory Protein-1 Like most cells present in human teeth (Sloan et al., 2001; Lucchini et al., 2002), pulp DC expressed TβRII, the type II receptor necessary for TGF-β1 cell binding and signal transduction (Jayaraman and Massagué, 2000). Thus, pulp immature DC could also respond directly to TGF-β1 stimulation. A few mature DC were detected, in vivo and after culture, in both the deep pulp and the central part of the roots, but never in pulp horns. Those HLA-DR-positive cells were CD68-negative but expressed the lysosome-associated membrane protein DC-LAMP that is "turned on" upon activation of DC (de Saint-Vis et al., 1998). Mature DC-LAMP-positive DC are constitutively present in lymphoid organs but not in normal peripheral tissue. However, they have been observed at inflammatory lesions of both the skin and the joints. Interestingly, DC-LAMP-positive DC present in the tooth pulp in vivo still expressed Factor XIIIa, suggesting that they had been recently activated. The reason for the presence of those few mature DC in apparently sound pulp tissue is not known, and their potential role in either immunization or peripheral tolerance remains to be determined (Banchereau and Steinman, 1998; Huang et al., 2000). In humans, TGF-β1 is produced by secretory odontoblasts in sound teeth (Sloan et al., 2000; Lucchini et al., 2002), and this expression is increased under caries lesions (Sloan et al., 2000). Moreover, TGF-β1 is stored in the dentinal matrix (Finkelman et al., 1990; Cassidy et al., 1997; Zhao et al., 2000), and, when dentin is being destroyed by caries or operative procedures, it is released from the matrix and could diffuse to the pulp through dentinal tubules (Finkelman et al., 1990; Magloire et al., 2001; Smith and Lesot, 2001). Thus, odontoblasts and/or damaged dentin, via a gradient of TGF-β1 radiating from the pulp-dentin interface to the pulp core, could contribute, through differentiation from precursors and/or chemotactic recruitment, to the accumulation of immature DC in the peripheral pulp. In this way, TGF-β1 would stimulate the host’s ability to respond rapidly to infection or injury and initiate repair, allowing for tertiary dentin formation by primary or replacement odontoblasts. In conclusion, we present evidence that immature DC are attracted into the odontoblast layer by TGF-β1 originating from dentin. This factor could thus direct DC trafficking in pathological conditions resulting from dentin injury. From a clinical point of view, the use of TGF-β1 as a dentin/pulp-capping agent might represent a therapeutic strategy for modulating the early immune response and favoring healing in inflamed pulps, with the accumulation of immature DC at the pulp periphery further minimizing the risk of occurrence of a novel infection.
The authors express their gratitude to the staff of the Stomatology Department, Saint Joseph Hospital, Lyon, and to Dr. P. Exbrayat, Faculty of Odontology, Lyon, for collecting tooth samples. This work was supported by grants from the French Ministry of National Education, Research and Technology. Received for publication November 22, 2002. Revision received March 31, 2003. Accepted for publication May 7, 2003.
Journal of Dental Research, Vol. 82, No. 8,
652-656 (2003) This article has been cited by other articles:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
(MIP-1
