Differentiated Human Odontoblasts Reducing Dentin Formation in In

Lipoteichoic Acid Increases TLR and Functional Chemokine Expression while Reducing Dentin Formation in In Vitro Differentiated Human Odontoblasts This information is current as of September 30, 2021. Stéphanie H. Durand, Vincent Flacher, Annick Roméas, Florence Carrouel, Evelyne Colomb, Claude Vincent, Henry Magloire, Marie-Lise Couble, Françoise Bleicher, Marie-Jeanne Staquet, Serge Lebecque and Jean-Christophe Farges Downloaded from J Immunol 2006; 176:2880-2887; ; doi: 10.4049/jimmunol.176.5.2880 http://www.jimmunol.org/content/176/5/2880 http://www.jimmunol.org/

References This article cites 51 articles, 10 of which you can access for free at: http://www.jimmunol.org/content/176/5/2880.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 30, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Lipoteichoic Acid Increases TLR and Functional Chemokine Expression while Reducing Dentin Formation in In Vitro Differentiated Human Odontoblasts

Ste´phanie H. Durand,* Vincent Flacher,‡ Annick Rome´as,* Florence Carrouel,* Evelyne Colomb,† Claude Vincent,† Henry Magloire,* Marie-Lise Couble,* Franc¸oise Bleicher,* Marie-Jeanne Staquet,* Serge Lebecque,‡ and Jean-Christophe Farges1*

Gram-positive bacteria entering the dentinal tissue during the carious process are suspected to influence the immune response in human dental pulp. Odontoblasts situated at the pulp/dentin interface are the first cells encountered by these bacteria and therefore could play a crucial role in this response. In the present study, we found that in vitro-differentiated odontoblasts constitutively expressed the pattern recognition receptor TLR1–6 and 9 genes but not TLR7, 8, and 10. Furthermore, lipoteichoic Downloaded from acid (LTA), a wall component of Gram-positive bacteria, triggered the activation of the odontoblasts. LTA up-regulated the expression of its own receptor TLR2, as well as the production of several chemokines. In particular, an increased amount of CCL2 and CXCL10 was detected in supernatants from LTA-stimulated odontoblasts, and those supernatants augmented the migration of immature dendritic cells in vitro compared with controls. Clinical relevance of these observations came from immunohistochemical analysis showing that CCL2 was expressed in vivo by odontoblasts and blood vessels present under active carious lesions but not in healthy dental pulps. In contrast with this inflammatory response, gene expression of major dentin matrix components http://www.jimmunol.org/ (type I collagen, dentin sialophosphoprotein) and TGF-␤1 was sharply down-regulated in odontoblasts by LTA. Taken together, these data suggest that odontoblasts activated through TLR2 by Gram-positive bacteria LTA are able to initiate an innate immune response by secreting chemokines that recruit immature dendritic cells while down-regulating their specialized functions of dentin matrix synthesis and mineralization. The Journal of Immunology, 2006, 176: 2880–2887.

he human tooth is the target of a substantial number of tion, bacteria and/or components released from their wall such as oral bacterial agents that are responsible for the develop- lipoteichoic acid (LTA)2 (5, 6) can gain access to odontoblasts (3). T ment of carious lesions. These agents induce demineral- Therefore, they are the first cells encountered by bacteria entering by guest on September 30, 2021 ization of enamel that normally constitutes an impermeable barrier dentin from the oral cavity. Their peripheral situation in the dental that protects the underlying dentin and the connective tissue situ- pulp makes the tooth a unique example where neural crest-derived ated in the center of the tooth, the dental pulp (1). When the mesenchymal cells, instead of epithelial cells, may represent the enamel barrier is disrupted, the dentin exposed to the oral envi- first line of defense for the host. For these reasons, we hypothe- ronment is degraded by Gram-positive bacteria, including strepto- sized that odontoblasts could play a central role in the dental pulp cocci, lactobacilli, and actinomyces, that largely dominate the car- innate and adaptive immune responses. ious lesion microflora (2). This leads to the development of The initial step of an innate immune response is the detection of inflammatory and immune events in the dental pulp, the molecular pathogens through specialized pattern recognition receptors and cellular determinants of which remain unknown (3). present in the cell membrane of immune and nonimmune cells, Odontoblasts are neural crest-derived mesenchymal cells orga- among which TLRs are key participants (7, 8). To date, 10 TLR nized as a densely packed layer at the dentin-pulp interface. Their family members (TLR1–10) have been identified in the human main functions are synthesis and extracellular deposition of a type genome, and different TLRs appear to play crucial roles in the I collagen-rich matrix referred to as predentin, and the mineraliza- early activation of the innate immune response by different patho- tion of this matrix to form dentin (4). They send long cytoplasmic gen-associated molecular patterns (PAMPs). TLRs trigger the ef- processes into dentinal tissue. Thus, during dentin demineraliza- fector phase of the innate immune response, mainly through the activation of the NF-␬B pathway (9). This includes the secretion of proinflammatory chemokines and cytokines that recruit and acti- *Laboratory “Development and Regeneration of Dental Tissues,” University Lyon 1, vate blood borne inflammatory cells (10, 11). In the dental pulp, Faculty of Odontology, Institut National de la Sante´ et de la Recherche Me´dicale (INSERM) ERi76 Equipe d’Accueil 1892, Institut Fe´de´ratif de Recherche 62, Lyon, when dentin is being destroyed by caries, immature dendritic cells France; †Equipe d’Accueil 3732, University Lyon 1, E. Herriot Hospital, Lyon, (iDCss) accumulate into the odontoblast layer close to the lesion in France; and ‡INSERM Unite´503, IFR 128, University Lyon 1, Centre d’Etudes et de strategic location to sample foreign Ags (3, 12). Factors favoring Recherches en Virologie et Immunologie, Lyon, France this accumulation are unknown, but chemokines may be Received for publication July 27, 2005. Accepted for publication December 13, 2005. involved (13). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Address correspondence and reprint requests to Dr. Jean-Christophe Farges, Labo- ratory “Development and Regeneration of Dental Tissues,” University Lyon 1, Fac- 2 Abbreviations used in this paper: LTA, lipoteichoic acid; PAMP, pathogen- ulty of Odontology, 11 G. Paradin Street, F-69372 Lyon Cedex 08, France. E-mail associated molecular pattern; iDC, immature dendritic cell; DSPP, dentin address: [email protected] sialophosphoprotein.

In the present study, we used odontoblasts generated in vitro DNA Master SYBR Green I kit, according to the manufacturer’s specifi- from cultures of human dental pulp explants (14). In these condi- cations. Primer sets and annealing temperatures are listed in Table I. The tions, dental pulp cells give rise to highly differentiated cells that real-time PCR conditions were as follows: 10 min at 95°C, followed by 45 cycles of 95°C for 15 s, annealing temperature for 5 s, 72°C for 20 s. The exhibit many odontoblast features such as cell body polarization, increase in PCR product was monitored for each amplification cycle by formation of a typical process at the cell pole opposite to the nu- measuring the increase in fluorescence caused by the binding of SYBR cleus, and strong expression of specific markers, including type I Green dye to dsDNA. Amplicon specificity was confirmed by analysis of collagen, dentin sialophosphoprotein (DSPP), TGF-␤1, phosphate melting curves. Standard curves for target and cyclophilin A housekeeping genes were constructed with serial dilutions of reverse transcription sam- regulating gene with homologies to endopeptidases on the X chro- ples. For each target gene, mRNA levels were normalized to the cyclophi- mosome, osteoadherin, and reelin (14, 15, 16, 17). Although some lin A housekeeping gene with the RelQuant software (Roche). Results were differences may exist with in vivo cells, in vitro-differentiated expressed as fold change values relative to control odontoblast samples odontoblasts are referred to as odontoblasts from herein. nonstimulated with LTA. All runs were performed in duplicates and in- Our aim here was to determine the TLR gene profile of odon- cluded a negative control without template. PCR amplification was also verified for expected products, mispairing, and primer dimer formation on toblasts and their innate immune response to LTA, a PAMP from 2% agarose gels. Gram-positive bacteria and a ligand for TLR2. We stimulated Flow cytometry odontoblasts with LTA and analyzed consequences on their phe- notype in terms of TLR and chemokine expression, induction of Odontoblasts were obtained following trypsin/EDTA treatment of cultures iDCs migration, dentin matrix synthesis, and mineralization. and incubated for 30 min with anti-TLR2 mouse mAb (clone TL2.1; Santa Cruz Biotechnology) or isotype-matched mouse IgG2a (clone UPC 10; Sigma-Aldrich) at the same concentration. After washing, cells were in- Ј Downloaded from Materials and Methods cubated with FITC-conjugated F(ab )2 goat anti-mouse Ab (Caltag Labo- Cell culture ratories) for 30 min. Cells were washed and resuspended in PBS containing 1% formaldehyde. Data were acquired on a DakoCytomation cytometer Odontoblasts were generated in vitro from human dental pulp explants as (DakoCytomation) and analyzed using the WinMDI 2.8 software (The described by Couble et al. (14). Briefly, 20 healthy nonerupted third molars Scripps Institute). were collected with informed consent of the patients, in accordance with the French Public Health Code and following a protocol approved by the Gene arrays local ethics committee. The pulp tissue was separated from the dentin/ ␮ http://www.jimmunol.org/ enamel mineralized complex, and its apical end was removed to prevent RNA samples (1 g each) from LTA-stimulated and control odontoblast cultures were amplified with the MessageAmp aRNA kit (Ambion), ac- periodontal cell contamination. Pulp explants (ϳ2mm3) were grown in cording to the manufacturer’s instructions. Amplified RNAs (500 ng) were Eagle’s basal medium (Invitrogen Life Technologies) supplemented with ␣32 10% FCS (Eurobio), 50 ␮g/ml acid ascorbic, 10 mM ␤-glycerophosphate then used as a template to generate P-dCTP-labeled cDNA probes with use of the AmpoLabeling-LPR kit (SuperArray Bioscience). Nylon mem- (Sigma-Aldrich), and 100 IU/ml penicillin-50 mg/ml streptomycin (In- brane arrays containing cDNA fragments from human chemokine and re- vitrogen Life Technologies). After 4 wk to allow odontoblast differentia- ceptor genes (GEArray Q Series, catalog no. HS-005.2) were purchased tion, half the cultures were stimulated with 1 ␮g/ml Bacillus subtilis LTA from SuperArray. Detailed information about these arrays is available on- (InvivoGen) for 8 h. LTA was tested for LPS content using the Limulus line (͗www.superarray.com/͘). After prehybridization with GEAhyb Hy- amoebocyte lysate assay. Aliquots showed a LPS contamination that did bridization Solution (SuperArray) containing 100 ␮ not exceed 0.12 ng/␮g LTA. Control experiments using 0.12 ng/ml LPS g/ml denatured salmon sperm DNA (Sigma-Aldrich) for 4 h at 60°C, membranes were hybridized (InvivoGen) failed to show signs of odontoblast activation (data not by guest on September 30, 2021 ϫ 6 - shown), so that we considered LPS as not responsible for the observed at 60°C for 15 h with radioactive probes (50–200 10 cpm each) pre viously denaturated at 95°C for 4 min. Membranes were washed twice with biological phenomena. 2ϫ SSC—1% SDS for 15 min and twice with 0.1ϫ SSC—0.5% SDS for Real-time PCR 15 min, then aligned on a PhosphorImaging screen (Molecular Dynamics) and exposed for 72 h at room temperature. Array images were obtained by Total RNA was extracted from LTA-stimulated or control odontoblast cul- scanning the screen, then files were converted into numerical data using the tures with the Nucleospin RNA II kit (Macherey-Nagel), according to the ScanAlyze software (͗www.microarrays.org/software.html͘) and analyzed manufacturer’s instructions. RNA samples (1 ␮g) were then converted to with the GEArray Analyzer software (͗www.superarray.com/͘). Hybridiza-

ﬁrst-strand cDNAs using 500 ng oligo (dT)15 primers (Roche Diagnostics) tion signals were normalized for each gene to the cyclophilin A signals and Power Script Reverse Transcriptase (BD Clontech). Real-time PCR observed on the same membrane. Background was given by the mean value was performed in a Light Cycler instrument (Roche) with the Fast Start of the hybridization signals recorded for pUC18 plasmid. Results were

Table I. Primers used for real-time PCR analysis

Annealing Gene Forward Primer (5Ј–3Ј) Reverse Primer (5Ј–3Ј) Temperature (°C)

TLR1 CCCATTCCGCAGTACTCCAT TTTTCCTTGGGCCATTCCA 60 TLR2 CCCATTGCTCTTTCACTGCT CTTCCTTGGAGAGGCTGATG 60 TLR3 TGGTTGGGCCACCTAGAAGTA TCTCCATTCCTGGCCTGTG 60 TLR4 CTGCAATGGATCAAGGACCA TTATCTGAAGGTGTTGCACATTCC 60 TLR5 TGCCTTGAAGCCTTCAGTTATG CCACCACCATGATGAGAGCA 60 TLR6 CCCTCAACCACATAGAAACG GAGATATTCCACAGGTTTGG 60 TLR7 TTACCTGGATGGAAACCAGCTACT TCAAGGCTGAGAAGCTGTAAGCTA 60 TLR8 AACTTTCTATGATGCTTACATTTCTTATGAC GGTGGTAGCGCAGCTCATTT 60 TLR9 TGAAGACTTCAGGCCCAACTG TGCACGGTCACCAGGTTGT 60 TLR10 TTTGATCTGCCCTGGTATCTCA AGTTGTTCTTGGGTTGTTTTCCTAAC 60 Coll1␣1 TGACCTCAAGATGTGCCACT ACCAGACATGCCTCTTGTCC 62 DSPP GCATCCAGGGACAAGTAAGCA CTTGGACAACAGCGACATCCT 58 TGF-␤1 GTCACCCGCGTGCTAATG CACGTGCTGCTCCACTTTTA 60 Cyclophilin A ATGGCACTGGTGGCAAGTCC TTGCCATTCCTGGACCCAAA 58 CCL2 GATCTCAGTGCAGAGGCTCG AAGCAATTTCCCCAAGTCTC 68 CCL7 GCACTTCTGTGTCTGCTGCT TAGCTCTCCAGCCTCTGCTT 66 CXCL2 AACCGAAGTCATAGCCACAC CAGGAACAGCCACCAATAAG 58 CXCL10 TGAGCCTACAGCAGAGGAA TACTCCTTGAATGCCACTTAGA 62 2882 INNATE IMMUNE RESPONSE IN HUMAN ODONTOBLASTS

expressed as a percentage of cyclophilin A gene expression. Data were reported for genes whose expression level was Ͼ5% of the cyclophilin A gene (this threshold corresponding approximately to the visual detection of the spot) and that exhibited at least a 2-fold change of expression between LTA-stimulated and control samples. Only the matched positive and negative results of three experiments are presented. Statistical analysis Results were expressed as mean values Ϯ SD obtained from three independent experiments. Statistical analysis was determined with Student’s t test. Protein array A Human Chemokine Ab Array I kit was purchased from Raybiotech and used according to the manufacturer’s instructions. Briefly, membranes were blocked for 30 min at room temperature before being incubated for 16 h at 4°C with supernatants recovered from odontoblasts cultured with or without LTA. They were then washed, incubated with biotin-conjugated primary Abs, HRP-conjugated streptavidin, and finally developed by using ECL-type solution. Membranes were scanned with a VersaDoc Imaging FIGURE 1. TLR expression in resting and LTA-activated odontoblasts. System (Bio-Rad), and semiquantitative analysis of the comparative inten- A, Real-time PCR analysis of TLR gene expression in human odontoblasts sity of the spots was performed with the Quantity One 4.4.1 software Downloaded from stimulated or not with LTA. Results were normalized to cyclophilin A gene (Bio-Rad). expression and expressed as fold change values relative to control non- Immunohistochemistry stimulated samples. Data represent the mean Ϯ SD obtained from three independent experiments. Odontoblasts in culture constitutively expressed Three healthy and seven carious human third molars were fixed in 4% ␮ TLR1, 2, 3, 4, 5, 6, and 9 but not TLR7, 8, and 10. A significant increase paraformaldehyde-PBS solution for 16 h, rinsed, and cut into 750- m- ϭ ϭ thick slices. After demineralization in acetic acid or EDTA for 25 days, of gene expression for TLR2 (4.1-fold; p 0.004), TLR3 (2.4-fold; p 0.001), TLR5 (2.1-fold; p ϭ 0.02), and TLR9 (2.1-fold; p ϭ 0.03) was slices were routinely treated for paraffin embedding (18). Five-micrometer http://www.jimmunol.org/ serial sections were deparaffinized, rehydrated, and treated with an Ag observed in LTA-stimulated odontoblasts (Od ϩ LTA) compared with unmasking solution (Vector Laboratories) for 15 min. They were then in- controls (Od). B, Flow cytometry analysis of TLR2 protein expression in cubated with 20 ␮g/ml anti-CCL2 goat polyclonal Ab (R&D Systems). Ab human odontoblasts stimulated or not with LTA. Isotype controls are detection was performed using the Vectastain Elite ABC kit (Vector Lab- shown as open histograms. TLR2 was induced in LTA-stimulated odon- oratories), according to the manufacturer’s protocol. Sections were slightly toblasts (Od ϩ LTA) compared with controls (Od). Results are represen- counterstained with toluidine blue. tative of two independent experiments. Generation of iDCss ϩ iDCss were generated from human CD34 hemopoietic progenitors iso- 0.02), and TLR9 (2.1-fold; p ϭ 0.03). Other TLR genes remained lated from human umbilical cord blood mononuclear fraction by immuno- unchanged. TLR2 was then selected for protein analysis given its by guest on September 30, 2021 magnetic selection with miniMACS (Miltenyi Biotec). Purified progenitors (90–98% purity) were cultured in RPMI 1640 supplemented with 5% heat- crucial role in LTA signaling. Flow cytometry revealed that LTA inactivated FBS (Myoclone super plus), 10 mM HEPES, 2 mM L-glu- up-regulated TLR2 in odontoblasts (Fig. 1B). tamine, 100 U/ml penicillin, 100 ␮g/ml streptomycin (all from Invitrogen Life Technologies), 5 ϫ 10Ϫ5 M 2-ME (Sigma-Aldrich), 200 U/ml recom- LTA treatment of odontoblasts modulates chemokine gene and binant human GM-CSF (kindly provided by Schering-Plough Laborato- protein expression ries), and 50 U/ml recombinant human TNF-␣ (Genzyme) for 7 days. Un- der these conditions, cells remain poorly differentiated and display features Given that one major consequence of TLR activation is the up- of iDCs (19). regulation of chemokine synthesis, we then examined the chemo- Chemotaxis assay kine expression profile of odontoblasts by cDNA array analysis (Table II and Fig. 2A). Data revealed that odontoblasts constitu- ␮ Cell migration was assessed using Costar Transwell devices with an 8- m tively expressed 17 genes related to the chemokine pathway, in- pore size. Supernatants recovered from odontoblasts cultured with or without LTA were diluted (1/4) in RPMI 1640–5% FBS and added to 24-well cluding genes encoding for the chemokines CCL2, CCL26, plates. iDCss (1.5 ϫ 105) suspended in RPMI 1640 supplemented with 5% CXCL4, CXCL12, and CXCL14. The remainder of genes include FBS were applied to Transwell inserts. Plates were incubated for4hat chemokine receptors (CXCR2, CCRL1, and CCRL2), leukotriene 37°C. After removal of the Transwell inserts, migrated cells were recov- B4 receptor, brain-derived neurotrophic factor, endothelial mono- ered and counted. Supernatants of three different control and LTA-stimu- cyte-activating protein 2, stromal cell-derived factor 2, hypoxia- lated odontoblast cultures were analyzed in duplicate. Buffer alone and ␣ buffer ϩ LTA were used as controls. Results were expressed as the number inducible factor 1 , and genes of still unknown function. The other of migrated iDCs in percentage of the input cell number introduced into the genes spotted on the membrane were not expressed by odonto- upper compartment. blasts or at a level Ͻ 5% of the cyclophilin A gene and/or in a nonreproducible manner between the three tested samples. Of Results note, neither TLR2 nor TLR4 expression was detected on the gene TLR expression in resting and LTA-activated odontoblasts array in contrast with their easy amplification by PCR (Fig. 1A). We first set out to determine the expression of TLR mRNAs in Such a negative result is likely due to the lower sensitivity result- resting odontoblasts and in response to LTA, a ligand for TLR2. ing from the generation of the probes for the gene array, as re- Real-time PCR investigation of the TLR profile of odontoblasts ported previously (20, 21). showed that these cells constitutively expressed genes coding for After LTA stimulation, three genes expressed in nonstimulated TLR1, 2, 3, 4, 5, 6, and 9 (Fig. 1A). TLR7, 8, and 10 mRNA could odontoblasts were found to be up-regulated Ͼ2-fold: CCL2 (2.4- not be amplified in odontoblasts, although easily detected in plas- fold; p ϭ 0.01), CCRL1, (10.7-fold; p ϭ 0.002), and CCRL2 (3.9- macytoid or myeloid DCs (data not shown). Upon LTA activation, fold; p ϭ 0.02). CCL7, CXCL2, and CXCL10 were only found to an increase in gene expression level was observed for TLR2 (4.1- be expressed in LTA-stimulated odontoblasts (Fig. 2, B and C). fold; p ϭ 0.004), TLR3 (2.4-fold; p ϭ 0.001), TLR5 (2.1-fold; p ϭ Other genes exhibited no change or low changes that were not The Journal of Immunology 2883

Table II. Expression proﬁle of chemokine-related genes in odontoblasts as determined by gene array analysisa

Position Gene Name GenBank Accession No. Expression Level

Chemokines 5C CXCL4/PF4/SCYB4 NM_002619 ϩ 6G CCL2/MCP-1/SCYA2 NM_002982 ϩϩ 7F CCL26/Eotaxin-3/SCYA26 NM_006072 ϩϩϩ 9B CXCL14/BRAK/SCYB14 NM_004887 ϩ 10A CXCL12/SDF1/SCYB12 NM_000609 ϩ Other genes 1B C10 NM_138425 ϩϩϩ 2E L-CCR/CCRL2 NM_003965 ϩϩ 4H CXCR2/IL8RB NM_001557 ϩϩϩ 5A LTB4R NM_181657 ϩ 6A BDNF NM_001709 ϩ 9H EMAP2/p43/SCYE1 NM_004757 ϩϩϩ 10B SDF2 NM_006923 ϩ 10E CCRL1 NM_016557 ϩ 10G CKLFSF3 NM_144601 ϩ 11B CKLFSF6 NM_017801 ϩϩ 11C CKLFSF7 NM_138410 ϩ Downloaded from 11H HIF1A NM_001530 ϩϩ

a Only the matched positive and negative results of three experiments are presented. The gene expression level was normalized to the cyclophilin A gene and presented as the following ratio: (chemokine signal Ϫ background)/(cyclophilin A signal Ϫ background). ϩ, between 5 and 25% of the expression level of the cyclophilin A gene; ϩϩ, between 25 and 100%; ϩϩϩ, Ͼ100%. http://www.jimmunol.org/ significant statistically. The four chemokine genes up-regulated by CXCL10, respectively. The CCL7 level was very low and showed LTA were selected for verification by real-time PCR with the same no significant modification (Fig. 4). CXCL2 could not be assessed cultured samples. Stimulation of genes encoding CCL2 (3-fold; because of the absence of a specific Ab on the membrane. p ϭ 0.01), CCL7 (7.6-fold; p ϭ 0.02), CXCL2 (10.3-fold; p ϭ 0.006), and CXCL10 (6-fold; p ϭ 0.04) in LTA-treated cultures CCL2 is expressed in vivo in carious teeth was confirmed (Fig. 3). Protein array analysis of culture superna- Data from the literature indicating that CCL2 was able to stimulate tants showed an increase of 2.3- and 2.7-fold for CCL2 and iDCs migration in vitro (11), and our results showing that CCL2 by guest on September 30, 2021

FIGURE 2. Gene expression profile of odontoblasts determined by array analysis and changes induced by LTA treatment. Complete list of genes spotted on the Q Series Human Chemokines and Receptors Gene Array nylon membrane (catalog no. HS-005.2) is available online (͗www.superarray.com/͘). A, Odontoblasts (Od) constitutively expressed 17 chemokine-related genes (listed in Table II), including CCL2 (position 6G on the membrane), CCL26 (7F), CXCL4 (5C), CXCL12 (10A), and CXCL14 (9B) chemokine genes. B, After LTA stimulation (Od ϩ LTA), hybridization signals were significantly increased for CCL2 (6G), CCL7 (8E), CXCL2 (4C), CXCL10 (8G), CCRL1 (10E), and CCRL2 (2E) genes. Cyclophilin A tetraspots (position 14A-D) were used for normalization. The three pUC18 tetraspots used for background subtraction are in the position 13A-C. The autoradiograms shown are representative of three independent experiments. (C) Quantification of hybridization signals on gene arrays. Statistical analysis of three independent experiments showed that six genes were up-regulated Ͼ2-fold: three genes whose expression was detected in nonstimulated odontoblasts and three genes whose expression was only detected in LTA-stimulated samples. The former were CCL2 (2.4- fold; p ϭ 0.01), CCRL1 (10.7-fold; p ϭ 0.002) and CCRL2 (3.9-fold; p ϭ 0.02), the latter were CCL7, CXCL10, and CXCL2. Other genes exhibited no change or low changes that were statistically not significant. 2884 INNATE IMMUNE RESPONSE IN HUMAN ODONTOBLASTS

FIGURE 3. Real-time PCR analysis of CCL2, CCL7, CXCL2, and CXCL10 chemokine gene expression in human odontoblasts stimulated or not with LTA. Results were normalized to cyclophilin A expression and expressed as fold change values relative to control odontoblasts nonstimulated with LTA (Od). Data represent the mean Ϯ SD obtained from three independent experiments. Statistical analysis confirmed the significant up- regulation of gene expression for CCL2 (3-fold; p ϭ 0.01), CCL7 (7.6- fold; p ϭ 0.02), CXCL2 (10.3-fold; p ϭ 0.006), and CXCL10 (6-fold; p ϭ Downloaded from 0.04) in LTA-stimulated cultures (Od ϩ LTA). FIGURE 5. In vivo immunohistochemical analysis of third molars with an anti-CCL2 polyclonal Ab. A and B, A healthy tooth and a carious one was up-regulated at the mRNA and protein levels in LTA-stimu- with an active dentinal lesion (dark brown aspect, black arrow). White lated odontoblasts, prompted us to examine expression of this mol- arrows indicate the precise localization, at the dentin-pulp interface, of the odontoblasts (Od) shown in C and D. No staining was observed with the ecule in vivo in tooth specimens. Analysis of healthy (Fig. 5A) and http://www.jimmunol.org/ anti-CCL2 Ab in the healthy tissue (C). In the carious tooth, a moderate carious (Fig. 5B) third molars with an anti-CCL2 polyclonal Ab cytoplasmic staining was observed in odontoblasts (Od) (D), as well as in showed an absence of staining in healthy teeth (Fig. 5C), whereas endothelial cells from blood vessels (BV) present in the underlying inflam- a moderate staining was observed in odontoblasts under active matory pulp tissue (E). E, enamel; D, dentin; P, pulp. carious lesions (Fig. 5D). Endothelial cells present in the underlying inflammatory tissue were also stained (Fig. 5E), as was previously shown in various inflammatory conditions (22). Down-regulation of dentin matrix protein genes by LTA LTA-stimulated odontoblasts attract iDCss Since the main steady functions of odontoblasts are the secretion by guest on September 30, 2021 To further assess the biological relevance of LTA-induced chemo- of the extracellular matrix of dentin and its mineralization, and that kine response of odontoblasts, LTA-conditioned culture superna- this matrix is strongly reduced under active dentin carious lesions, tants were tested for their chemotactic effect on iDCs in a Trans- we determined whether LTA stimulation influences these odonto- well migration assay. In experiments with buffer alone, buffer ϩ blast specialized functions. Thus, we analyzed gene expression of ␤ LTA (data not shown), and supernatants from nonstimulated odon- two dentin matrix components and of TGF- 1, considered to be a toblasts, a mean number of 37 Ϯ 9% iDCs migrated (Fig. 6). The crucial regulatory element for dentinogenesis, by real-time PCR. A migratory response was enhanced significantly (69 Ϯ 11.8%, p ϭ significant down-regulation of expression was observed for type I ␣ ϭ Յ 0.001) when supernatants from LTA-stimulated odontoblasts were collagen 1 chain (3.4-fold; p 0.001), DSPP (98-fold; p ␤ ϭ added in the lower compartment. Checkerboard analysis estab- 0.0001), and TGF- 1 (3.5-fold; p 0.02) genes in LTA-stimu- lished that cells attracted by LTA-stimulated odontoblasts mi- lated cells (Fig. 7). grated in a directional rather than a random fashion (data not shown). Discussion The interface between human teeth and oral cavity presents several unique features that must be balanced to maintain dental pulp ste- rility. The dentin-pulp complex is protected by an impermeable cap of enamel and confronted to the oral microflora predominantly composed of Gram-positive saprophyte bacteria that are usually well tolerated. However, these bacteria release products that can destroy the enamel and the underlying dentin during the carious process. Dentin-producing odontoblasts then become the first cells run into by pathogens penetrating the living dental tissue. Hence, these cells represent a unique example where neural crest-derived FIGURE 4. Protein array analysis of chemokine release from odonto- mesenchymal cells must fulfill the role devoted elsewhere to epi- blasts stimulated or not with LTA. Each anti-chemokine Ab is present on thelial cells (8). The recent localization of ␤-defensins 1 and 2 in the membrane in duplicate. Chemokines whose gene expression was found healthy odontoblasts (23) also argues in favor of this hypothesis. to be up-regulated in gene array and real-time PCR experiments are shown, To study the expression of TLRs by odontoblasts, we used a except CXCL2, which could not be assessed because of the absence of a specific Ab on the membrane. After background subtraction (Blank), val- pure population of cells generated in vitro that are similar in many ues were adjusted based on the intensity of control spots on the filters (Pos). aspects to in vivo odontoblasts (14, 16). We observed the consti- CCL2 and CXCL10 showed a protein level increase of 2.3- and 2.7-fold, tutive expression of TLR1–6 and 9 genes but not TLR7, 8, and 10 respectively. The CCL7 level was very low and showed no significant mod- genes. This large range of TLRs expressed by odontoblasts appears ification (N.S.). Data are representative of two independent experiments. comparable to what has been reported for cultured epithelial cells, The Journal of Immunology 2885

to TLR3 and TLR4 ligands (our preliminary data). It remains to be determined whether these cells can actually detect and react to TLR5 and 9 ligands. Given the predominant role of Gram-positive bacteria in dental caries, we analyzed further the odontoblast response to LTA. LTA was found to strongly up-regulate the expression of its own receptor, TLR2, and to a lesser extent, of TLR3, 5, and 9. The up- regulation of TLR2 by its ligand may increase the sensitivity of odontoblasts, as previously reported in hemopoietic cells (28). The absence of TLR2 detection in resting odontoblasts is probably because the protein is present at a level below the sensitivity threshold of flow cytometry, as reported for iDCs (29). One well-recognized consequence of TLR2 activation is the up- regulation of chemokine synthesis (30). Using gene arrays, we compared the expression of chemokines by resting vs TLR2-activated odontoblasts. We observed the expression by resting odontoblasts of CCL2, CXCL12, and CXCL14, three chemokines known to drive iDCs recruitment in vitro (11, 31, 32, 33). How- ever, the chemokine the most strongly expressed by resting odon- Downloaded from FIGURE 6. Migration assay of iDCss in the presence of LTA-activated toblasts was CCL26, a natural antagonist for CCR1, CCR2, and ϩ odontoblast supernatants. CD34 cells were cultured in the presence of CCR5 (34). Therefore, it is tempting to speculate that the balance ␣ GM-CSF and TNF- . At day 7, their response to supernatants from odon- of odontoblast-derived chemokines may control the homing of toblasts treated or not with LTA was determined by migration assays in iDCs in the human dental pulp in normal conditions. Whether Transwell chambers. The migratory response was significantly enhanced when supernatants from the LTA-treated odontoblasts (Od ϩ LTA) were CCL2 could also contribute to iDCs recruitment remains uncertain http://www.jimmunol.org/ added in the lower compartment compared with those of nonstimulated because we could not detect this protein in the healthy dental pulp odontoblasts (Od). Results are expressed as the number of migrated cells in by immunohistochemistry. percentage of the input cell number and are the mean Ϯ SD of duplicates Following LTA stimulation, four chemokine genes (CCL2, from three independent experiments. CCL7, CXCL2, and CXCL10) and two corresponding proteins (CCL2 and CXCL10) were clearly up-regulated. Furthermore, including keratinocytes (24), intestinal epithelial cells (25), bron- odontoblasts present under active dentin carious lesions were chial epithelial cells (26), and gingival epithelial cells (27). Inter- found to express the CCL2 protein by immunohistochemistry. estingly, the pattern of TLRs expressed by odontoblasts was sim- CCL2 is a key inflammatory chemokine produced during micro- ilar to the one reported for gingival fibroblasts in primary cultures bial infection that attracts iDCs and also monocytes, activated T by guest on September 30, 2021 (27). Whether a common TLR expression profile exists for all oral cells, NK cells, and basophils through CCR1 and CCR2 (11). This mesenchymal cell types remains to be determined. chemokine might be responsible for iDCs recruitment into the Thus, odontoblasts might be involved in the recognition of bac- odontoblast layer, thereby facilitating their interaction with invad- terial products such as triacetylated lipoproteins (TLR1ϩTLR2), ing Gram-positive bacteria. The detection in vitro of a strong che- LTA (TLR2), diacetylated lipoproteins, peptidoglycans motactic activity for iDCs in the supernatant of LTA-stimulated (TLR2ϩTLR6), LPS (TLR4), flagellin (TLR5), and unmethylated odontoblasts supports this hypothesis. Experiments are going on to CpG motif-containing DNA (TLR9), and also of viral dsRNA identify the chemokines responsible for this migration. Moreover, through TLR3 (7, 8). Indeed, we found in the present work that through the up-regulation of CXCL2 and CXCL10 expression, odontoblasts responded in vitro to the TLR2 ligand LTA but also odontoblasts are likely to contribute to the recruitment of neutrophils and lymphocytes, respectively, during infection (11). CXCR2, a chemokine receptor primarily present on hemopoietic cells (35), was expressed at a high level by odontoblasts. Strik- ingly, CXCL2, for which the only known receptor is CXCR2, was also expressed by odontoblasts and strongly increased after LTA stimulation. Although this increase remains to be confirmed at the protein level, an autocrine control might occur in odontoblasts via CXCL2 regarding the expression of genes involved in predentin remodeling, as CXCR2 activation was shown to induce matrix metalloprotease-3 release in human chondrocytes (36). Chemokines not only induce cell locomotion but also influence angiogenesis (37, 38). Among chemokines expressed by odonto- FIGURE 7. Real-time PCR analysis of gene expression of dentin matrix blasts, CCL2, CXCL2, and CXCL12 are proangiogenic (11, 39), components (type I collagen ␣1 chain, DSPP, and TGF-␤1) in human whereas CXCL4, CXCL10, and CXCL14 are angiostatic (40, 41). odontoblasts stimulated or not with LTA. Results were normalized to cy- In vivo, the production of angiostatic chemokines in the healthy clophilin A expression and expressed as fold change values relative to dental pulp might be involved in the maintenance of blood vessels nonstimulated odontoblasts (Od). Data represent the mean Ϯ SD obtained out of the odontoblast layer. During inflammation of carious ori- from three independent experiments. Statistical analysis indicated a significant decrease of gene expression for the two dentin matrix components gin, the number of capillaries augments in the pulp under the le- (type I collagen ␣1 chain (3.4-fold; p ϭ 0.001), DSPP (98-fold; p Յ sion, and some of them penetrate into the odontoblast layer (42). 0.0001)) and TGF-␤1 (3.5-fold; p ϭ 0.02) in LTA-stimulated odontoblasts Regarding the proangiogenic chemokine CXCL2, its expression (Od ϩ LTA). was strongly up-regulated in LTA-stimulated odontoblasts, and 2886 INNATE IMMUNE RESPONSE IN HUMAN ODONTOBLASTS this chemokine might thus contribute to the increased vasculariza- 6. Telles, P. D. S., C. T. Hanks, M. A. A. M. Machado, and J. E. No¨r. 2003. tion by binding to CXCR2 that is highly expressed on endothelial Lipoteichoic acid up-regulates VEGF expression in macrophages and pulp cells. J. Dent. Res. 82: 466–470. cells (41). Altogether, these data suggest that TLR2-activated 7. Takeda, K. T., T. Kaisho, and S. Akira. 2003. Toll-like receptors. Annu. Rev. odontoblasts play a key role in the recruitment of iDCs and in- Immunol. 21: 335–376. 8. Iwasaki, A., and R. Medzhitov. 2004. Toll-like receptor control of the adaptive flammatory leukocytes and in the neoangiogenesis featured by the immune responses. Nat. Immunol. 5: 987–995. dental pulp during the carious infection. 9. Li, Q., and I. M. Verma. 2002. NF-␬B regulation in the immune system. Nat. Rev. The main steady functions of odontoblasts are the secretion of Immunol. 2: 725–734. 10. Yoshie, O., T. Imai, and H. Nomiyama. 2001. Chemokines in immunity. Adv. the extracellular matrix of dentin, i.e., predentin, and its mineral- Immunol. 78: 57–110. ization (4). Predentin is deposited by odontoblasts at a relatively 11. Mantovani, A., A. Sica, S. Sozzani, P. Allavena, A. Vecchi, and M. Locati. 2004. constant rate throughout the tooth life, but it is strongly reduced in The chemokine system in diverse forms of macrophage activation and polarization. Trends Immunol. 25: 677–686. active dentin carious lesions (43). We found that the major pre- 12. Yoshiba, N., K. Yoshiba, H. Nakamura, M. Iwaku, and H. Ozawa. 1996. Immu- dentin structural component, type I collagen, was down-regulated nohistochemical localization of HLA-DR-positive cells in unerupted and erupted by LTA. This finding is in agreement with the fact that NF-␬B normal and carious human teeth. J. Dent. Res. 75: 1585–1589. 13. Farges, J.-C., A. Romeas, M. Melin, J.-J. Pin, S. Lebecque, M. Lucchini, pathway activation decreases expression of ␣1(I) and ␣2(I) colla- F. Bleicher, and H. Magloire. 2003. TGF-␤1 induces accumulation of dendritic gen genes (44, 45). It also indicates that our model of odontoblast cells in the odontoblast layer. J. Dent. Res. 82: 652–656. 14. Couble, M.-L., J.-C. Farges, F. Bleicher, B. Perrat-Mabillon, M. Boudeulle, and stimulation by LTA is relevant to the in vivo situation of active H. Magloire. 2000. Odontoblast differentiation of human dental pulp cells in carious lesions. Events could be similar to what occurs in inflamed explant cultures. Calcified Tissue Int. 66: 129–138. joints where matrix synthesis by chondrocytes was also shown to 15. Buchaille, R., M.-L. Couble, H. Magloire, and F. Bleicher. 2000. A substractive PCR-based cDNA library from human odontoblast cells: identification of novel Downloaded from be down-regulated (46). We also observed an almost complete genes expressed in tooth forming cells. Matrix Biol. 19: 421–430. abolition of gene expression of DSPP, a glycoprotein that has a 16. Lucchini, M., A. Romeás, M.-L. Couble, F. Bleicher, H. Magloire, and J.-C. crucial role in the predentin mineralization process (47). Thus, it is Farges. 2002. TGF-␤1 signaling and stimulation of osteoadherin in human odontoblasts in vitro. Connect. Tissue Res. 43: 345–353. possible that coordinated regulatory mechanisms are activated to 17. Maurin, J.-C., M.-L. Couble, M. Didier-Bazes, C. Brisson, H. Magloire, and decrease both predentin matrix synthesis and mineralization during F. Bleicher. 2004. Expression and localization of reelin in human odontoblasts. the initial phase of the dental pulp immune response. In parallel, Matrix Biol. 23: 277–285. 18. Lucchini, M., M.-L. Couble, A. Romeás, M.-J. Staquet, F. Bleicher, H. Magloire,

␤ ␣ ␤ http://www.jimmunol.org/ we observed a clear diminution of TGF- 1 gene expression in and J.-C. Farges. 2004. v 3 integrin expression in human odontoblasts and odontoblasts. This might result in a decrease of TGF-␤1 amount in co-localization with osteoadherin. J. Dent. Res. 83: 552–556. ␤ 19. Noirey, N., M.-J. Staquet, M.-J. Gariazzo, M. Serres, C. Dezutter-Dambuyant, the odontoblast microenvironment in vivo. As TGF- 1 is known to C. Andre´, D. Schmitt, and C. Vincent. 2003. Withdrawal of TNF-␣ after the fifth attenuate TLR signaling (48, 49, 50, 51), the TGF-␤1 decrease day of differentiation of CD34ϩ cord blood progenitors generates a homogeneous observed in LTA-stimulated odontoblasts might be necessary for population of Langerhans cells and delays their maturation. Exp. Dermatol. 12: 96–105. an effective innate immune response to take place. Also, as 20. Luo, Y., J. Cai, Y. Liu, H. Xue, F. J. Chrest, R. P. Wersto, and M. Rao. 2002. TGF-␤1 inhibits Th1 and Th2 cell responses (52) and B cell ac- Microarray analysis of selected genes in neural stem and progenitor cells. J. Neu- tivation (50), its reduction would allow a protective adaptive im- rochem. 83: 1481–1497. 21. Fuke, S., T. Betsuyaku, Y. Nasuhara, T. Morikawa, H. Katoh, and M. Nishimura. mune response to happen within the infected dental pulp. 2004. Chemokines in bronchiolar epithelium in the development of chronic ob- In summary, the present work provides insights about the role of structive pulmonary disease. Am. J. Respir. Cell Mol. Biol. 31: 405–412. by guest on September 30, 2021 neural crest-derived odontoblasts in the initiation of immune re- 22. Charo, I. F., and M. B. Taubman. 2004. Chemokines in the pathogenesis of vascular disease. Circ. Res. 95: 858–866. sponses that occur in the human dental pulp in response to cario- 23. Dommisch, H., J. Winter, Y. Acil, A. Dunsche, M. Tiemann, and S. Jepsen. 2005. genic bacteria. Our results suggest that, upon TLR2 activation, Human ␤-defensin (hBD-1, -2) expression in human dental pulp. Oral Microbiol. odontoblasts undergo a functional switch to decrease their special- Immun. 20: 163–166. 24. Ko¨llisch, G., B. N. Kalali, V. Voelcker, R. Wallich, H. Behrendt, J. Ring, ized functions of dentin matrix synthesis and mineralization while S. Bauer, T. Jakob, M. Mempel, and M. Ollert. 2005. Various members of the orienting their metabolic activity toward the production of mole- toll-like receptor family contribute to the innate immune response of human epidermal keratinocytes. Immunology 114: 531–541. cules responsible for the triggering and the development of innate 25. Cario, E., and D. K. Podolsky. 2005. Intestinal epithelial TOLLerance versus and adaptive immune responses. The signaling pathways driving inTOLLerance of commensals. Mol. Immunol. 42: 887–893. this switch remain to be identified. 26. Sha, Q., A. Q. Truong-Tran, J. R. Plitt, L. A. Beck, and R. P. Schleimer. 2004. Activation of airway epithelial cells by Toll-like receptor agonists. Am. J. Resp. Cell Mol. 31: 358–364. Acknowledgments 27. Kusumoto, Y., H. Hirano, K. Saitoh, S. Yamada, M. Takedachi, T. Nozaki, We gratefully acknowledge the staff of the Stomatology Department, Saint Y. Ozawa, Y. Nakahira, T. Saho, H. Ogo, et al. 2004. Human gingival epithelial cells produce chemotactic factors interleukin-8 and monocyte chemoattractant Joseph Hospital; we also thank Dr. P. Exbrayat, Faculty of Odontology, protein-1 after stimulation with Porphyromonas gingivalis via Toll-like receptor Lyon, France, for collecting tooth samples. Flow cytometry was performed 2. J. Periodontol. 75: 370–379. at the Centre Commun d’Imagerie de Laennec (Institut Fe´de´ratif de 28. Ray, A., N. S. Chatterjee, S. K. Bhattacharya, and T. Biswas. 2003. Porin of Recherche 62, Lyon, France). Shigella dysenteriae enhances mRNA levels for Toll-like receptor 2 and MyD88, up-regulates CD80 of murine macrophage, and induces the release of interleukin- 12. FEMS Immunol. Med. Microbiol. 39: 213–219. Disclosures 29. Visintin, A., A. Mazzoni, J. H. Spitzer, D. H. Wyllie, S. K. Dower, and The authors have no financial conflict of interest. D. M. Segal. 2001. Regulation of Toll-like receptors in human monocytes and dendritic cells. J. Immunol. 166: 249–255. 30. Cook, D. N., D. S. Pisetsky, and D. A. Schwartz. 2004. Toll-like receptors in the References pathogenesis of human disease. Nat. Immunol. 5: 975–979. 1. Hamilton, I. R. 2000. Ecological basis for dental caries. In Oral Bacterial Ecol- 31. Cao, X., W. Zhang, T. Wan, L. He, T. Chen, Z. Yuan, S. Ma, Y. Yu, and G. Chen. ogy: The Molecular Basis. Horizon Scientific Press, Wymondham, U.K., pp. 2000. Molecular cloning and characterization of a novel CXC chemokine mac- 219–274. rophage inflammatory protein-2 ␥ chemoattractant for human neutrophils and 2. Love, R. M., and H. F. Jenkinson. 2002. Invasion of dentinal tubules by oral dendritic cells. J. Immunol. 165: 2588–2595. bacteria. Crit. Rev. Oral Biol. Med. 13: 171–183. 32. Shellenberger, T. D., M. Wang, M. Gujrati, A. Jayakumar, R. M. Strieter, 3. Jontell, M., T. Okiji, U. Dahlgren, and G. Bergenholtz. 1998. Immune defense M. D. Burdick, C. G. Ioannides, C. L. Efferson, A. K. El-Naggar, D. Roberts, et mechanisms of the dental pulp. Crit. Rev. Oral Biol. Med. 9: 179–200. al. 2004. BRAK/CXCL14 is a potent inhibitor of angiogenesis and a chemotactic 4. Qin, C., O. Baba, and W. T. Butler. 2004. Post-translational modifications of factor for immature dendritic cells. Cancer Res. 64: 8262–8270. sibling proteins and their roles in osteogenesis and dentinogenesis. Crit. Rev. 33. Shurin, G. V., R. Ferris, I. L. Tourkova, L. Perez, A. Lokshin, L. Balkir, Oral Biol. Med. 15: 126–136. B. Collins, G. S. Chatta, and M. R. Shurin. 2005. Loss of new chemokine 5. Rolla, G., R. V. Oppermann, W. H. Bowen, J. E. Ciardi, and K. W. Knox. 1980. CXCL14 in tumor tissue is associated with low infiltration by dendritic cells High amounts of lipoteichoic acid in sucrose-induced plaque in vivo. Caries Res. (DC), while restoration of human CXCL14 expression in tumor cells causes 14: 235–238. attraction of DC both in vitro and in vivo. J. Immunol. 174: 5490–5498. The Journal of Immunology 2887

34. Petkovic, V., C. Moghini, S. Paoletti, M. Uguccioni, and B. Gerber. 2004. effect on ␣2(I) collagen (COL1A2) gene transcription in human dermal fibro- Eotaxin-3/CCL26 is a natural antagonist for CC chemokine receptors 1 and 5. blasts. J. Immunol. 162: 4226–4234. J. Biol. Chem. 279: 23357–23363. 45. Rippe, R. A., L. W. Schrum, B. Stefanovic, J. A. Solis Herruzo, and 35. Pelus, L. M., D. Horowitz, S. C. Cooper, and A. G. King. 2002. Peripheral blood D. A. Brenner. 1999. NF-␬B inhibits expression of the ␣1(I) collagen gene. DNA stem cell mobilization: a role for CXC chemokines. Crit. Rev. Oncol. Hemat. 43: Cell Biol. 10: 751–761. 257–275. 46. Deschner, J. H., C. R. Hofman, N. P. Piesco, and S. Agarwal. 2003. Signal 36. Baggiolini, M. 2001. Chemokines in pathology and medicine. J. Intern. Med. transduction by mechanical strain in chondrocytes. Curr. Opin. Clin. Nutr. 6: 250: 91–104. 289–293. 37. Rollins, B. J. 1998. Chemokines. Blood 90: 909–904. 47. Sreenath, T., T. Thyagarajan, B. Hall, G. Longenecker, R. D’Souza, S. Hong, 38. Rosenkilde, M. M., and T. W. Schwartz. 2004. The chemokine system: a major J. T. Wright, M. MacDougall, J. Sauk, and A. B. Kulkarni. 2003. Dentin sialo- regulation of angiogenesis in health and disease. APMIS 112: 481–495. phosphoprotein knockout mouse teeth display widened predentin zone and develop defective dentin mineralization similar to human dentinogenesis imperfecta 39. Belperio, J. A., M. P. Keane, M. D. Burdick, B. Gomperts, Y. Y. Xue, K. Hong, type III. J. Biol. Chem. 278: 24874–24880. J. Mestas, A. Ardehali, B. Mehrad, R. Saggar, et al. 2005. Role of CXCR2/ 48. Arsura, M., M. Wu, and G. E. Sonenshein. 1996. TGF-␤1 inhibits NF-␬B /Rel CXCR2 ligands in vascular remodeling during bronchiolitis obliterans syndrome. activity inducing apoptosis of B cells: transcriptional activation of I␬B␣. Immu- J. Clin. Invest. 115: 1150–1162. nity 5: 31–40. 40. Bikfalvi, A., and G. Gimenez-Gallego. 2004. The control of angiogenesis and 49. Nagarajan, R. P., F. Chen, W. Li, E. Vig, M. A. Harrington, H. Nakshatri, and tumor invasion by platelet factor-4 and platelet factor-4-derived molecules. Se- Y. Chen. 2000. Repression of transforming growth factor ␤-mediated transcrip- min. Thromb. Hemost. 30: 137–144. tion by nuclear factor ␬B. Biochem. J. 348: 591–596. 41. Schwarze, S. R., J. Luo, W. B. Isaacs, and D. F. Jarrard. 2005. Modulation of 50. Roes, J., B. K. Choi, and B. B. Cazac. 2003. Redirection of B cell responsiveness CXCL14 (BRAK) expression in prostate cancer. Prostate 59: 1–8. by transforming growth factor ␤ receptor. Proc. Natl. Acad. Sci. USA 100: 42. Heyeraas, K. J., and I. A Mjo¨r. 2001. Pulp-dentin biology in restorative dentistry. 7241–7246. Part 3: pulpal inflammation and its sequelae. Quintessence Int. 32: 611–625. 51. Naiki, Y., K. S. Michelsen, W. Zhang, S. Chen, T. M. Doherty, and M. Arditi. 43. Bjorndal, L., and I. A. Mjo¨r. 2001. Pulp-dentin biology in restorative dentistry. 2004. TGF-␤ differentially inhibits MyD88-dependent but not TRAM- and TRIF- Part 4: dental caries—characteristics of lesions and pulpal reactions. Quintes- dependent LPS-induced TLR4 signaling. J. Biol. Chem. 280: 5491–5495. sence Int. 32: 717–736. 52. Wahl, S. M., N. Vasquez, and W. Chen. 2004. Regulatory T cells and transcrip- Downloaded from 44. Kouba, D. J., K. Y. Chung, T. Nishiyama, L. Vindevoghel, A. Kon, J. F. Klement, tion factors: gatekeepers in allergic inflammation. Curr. Opin. Immunol. 16: J. Uitto, and A. Mauviel. 1999. Nuclear factor-␬B mediates TNF-␣ inhibitory 768–774. http://www.jimmunol.org/ by guest on September 30, 2021