Chronic Periodontal Disease Is Associated with Single-Nucleotide Polymorphisms of the Human TLR-4 Gene

BRIEF COMMUNICATION Chronic periodontal disease is associated with single-nucleotide polymorphisms of the human TLR-4 gene

NWJ Schro¨der1,5, D Meister1, V Wolff1, C Christan2, D Kaner3, V Haban1, P Purucker3, C Hermann4, A Moter1,UBGo¨bel1 and RR Schumann1 1Institute for Microbiology and Hygiene, Charite´ University Medical Center Berlin, Berlin, Germany; 2Private Dental Office, Berlin, Germany; 3Department of Periodontology and Synoptic Dentistry, Charite´ University Medical Center, Berlin, Germany; 4Biochemical Pharmacology, Department of Biology, University of Konstanz,Konstanz, Germany

Periodontitis is an inflammatory disease affecting the connective tissue surrounding the teeth leading to tooth loss. Pathogens associated with periodontitis interact with Toll-like receptors (TLRs) to induce cytokines causing and aggravating disease. We screened 197 individuals suffering from generalized periodontitis for the presence of Asp299Gly and Thr399Ile of TLR-4 as well as Arg753Gln of TLR-2 in comparison to matched controls. Single-nucleotide polymorphisms (SNPs) of TLR-4 were elevated among patients (odd’s ratio 3.650, 95% CI 1.573–8.467, Pp0.0001), while no difference was observed for TLR-2. TLR-4 SNPs were correlated with chronic periodontitis (odd’s ratio 5.562, 95% CI 2.199–14.04, Pp0.0001), but not with aggressive periodontitis. This observation was confirmed employing a group of periodontally healthy probands over 60 years of age. These data demonstrate that genetic variants of TLR-4 may act as risk factors for the development of generalized chronic periodontitis in humans. Genes and Immunity (2005) 6, 448–451. doi:10.1038/sj.gene.6364221; published online 5 May 2005

Keywords: periodontitis; innate immunity; Toll-like receptors; single-nucleotide polymorphisms; disease susceptibility

Periodontitis is an inflammatory disease affecting gingiva, Porphyromonas gingivalis, have been frequently recovered periodontal ligament, cementum and alveolar bone.1 The from periodontal lesions.6,7 There is evidence that innate major symptoms are formation of periodontal pockets, immune responses aggravate periodontal disease, since bleeding and destruction of connective tissue attachment, the proinflammatory cytokines IL-1 and TNF-a, reported also referred to as attachment loss, eventually leading to to be elevated in patients,8 act synergistically, ie in tooth loss.1 Periodontitis over-exceeds dental caries as causing bone resorption.9 The release of IL-1 and TNF-a leading cause of tooth loss worldwide,2 thus being a major by monocytes and macrophages is induced via a path- burden for health care systems. It is a common disease way involving the family of Toll-like receptors (TLRs).10 with an estimated prevalence of approx 50% among TLR-4 recognizes lipopolysaccharide (LPS), a partial adults, with 10–30% displaying severe forms of disease structure of Gram-negative bacteria, including period- characterized by an involvement of X30% of the teeth ontal pathogens.10 Recent data suggest that P. gingivalis, referred to as generalized periodontitis.3,4 Chronic period- in contrast to other Gram-negative bacteria, exhibits an ontitis (CP), progressing slowly and being predominantly untypical variant of LPS interacting with TLR-4, but also found among the elderly, is distinguished from aggressive TLR-2,11,12 a receptor generally recognizing cell wall periodontitis (AP), where rapid progression of attachment compounds distinct from LPS, such as lipoproteins.13 loss is mainly found in young adults.5 However, aggres- Thus, TLR-2 may be involved in innate immune sive courses of disease may also occur in elder patients.5 responses against periodontal pathogens. Expression of Attachment loss in periodontitis is considered to be both, TLR-2 and -4, has been shown to be increased in a result of both bacterial infection of subgingival macrophages as well as in gingival fibroblasts within tissue as well as host factors primarily directed against inflamed periodontal tissue.14 Furthermore, TLR-4-defi- these pathogens. Gram-negative anaerobe rods, ie cient C3H/HeJ mice exhibit reduced bone resorption in a periodontitis model,15 further indicating that TLRs may play a crucial role in periodontitis. Correspondence: Professor RR Schumann, Institut fu¨r Mikrobiologie und Several studies investigated the impact of single- Hygiene, Charite´—Universita¨tsmedizin Berlin, Dorotheenstrasse 96, nucleotide polymorphisms (SNPs) within genes encod- D-10117 Berlin, Germany. E-mail: [email protected] ing for proinflammatory cytokines on periodontal health, 5Current address: Pediatrics Infectious Diseases, Cedars Sinai Medical Center, Los Angeles 90048, CA, USA. ie SNPs within the IL-1 gene cluster, which lead to Received 3 January 2005; revised and accepted 6 April 2005; higher levels of IL-1b, describing an association of the published online 5 May 2005 occurrence of an SNP with CP and AP, respectively.16,17 TLR-4 polymorphisms in human periodontitis NWJ Schro¨der et al 449 Within the TLR-4 gene two cosegregating SNPs, Asp299Gly was found in two, Thr399Ile in three patients, Asp299Gly and Thr399Ile have been described leading but none of the controls exhibited one of these SNPs. to decreased cytokine release by epithelial cells upon Thus, 43 patients suffering from generalized period- LPS-stimulation in vitro, and are associated with hypo- ontitis displayed TLR-4 variants as compared to 14 found responsiveness to inhaled LPS, as well as with Gram- in the controls (OR 3.650 95% CI 1.573–8.467, P-value negative sepsis, septic shock and Legionnaire’s disease p0.0001). (reviewed in Schro¨der and Schumann18). Within the gene Since CP and AP are regarded as different disease encoding for human TLR-2, an SNP (Arg753Gln) was entities, we stratified the patients for clinical course of described diminishing inflammatory responses to TLR-2 disease. Patients suffering from CP (n ¼ 116) displayed a ligands in vitro and occurring at a frequency of higher frequency of the cosegregating SNPs as compared approximately 10% among Caucasians.19,20 to controls (OR 4.291 95% CI 1.669–11.03, P-value 0.002), To find out whether the genetic variations of TLR-4 while no significant difference was found for AP (n ¼ 81, and -2 influence periodontal disease susceptibility, we Table 2). Rare genotypes of TLR-4 were found exclu- investigated 197 patients suffering from periodontitis, sively within patients suffering from CP, strengthening both AP and CP, for the presence of the SNPs Asp299Gly the association of TLR-4 variants with this clinical picture and Thr399Ile of TLR-4 as well as Arg753Gln of TLR-2. (OR 5.562 95% CI 2.199–14.04, P-value 0.0001). In contrast Inclusion criteria for patients were clinical attachment to the observations made for TLR-4, we did not observe loss (CAL) of X3 mm affecting X30% of the teeth any significant association of the TLR-2 SNP with consistent with moderate to severe generalized period- periodontitis (OR 0.588 95% CI 0.270–1.280, P-value ontitis. Patients were stratified for clinical presentation of 0.2467, Table 2), and this was not affected by stratification AP and CP according to the classification system of the patients for CP and AP. introduced by the American Academy of Periodontology Severe CP is a frequent disease occurring at 10–30% in 1999.5 Patients were also included in the AP group among different populations.3 Although the control when there was clinical evidence for rapid disease probands did not have a history of periodontitis at the progression. Mean CAL of all patients was 4.39 mm time of inclusion in the study, it cannot be excluded that (71.23 mm), with 4.07 mm (71.06 mm) for CP and they will develop this disease later. Thus, we aimed at 5.45 mm (71.19 mm) for AP patients. The patients’ strengthening our results by comparing the frequencies characteristics are shown in Table 1. We found the of TLR-SNPs among CP-patients with controls being combined Asp299Gly/Thr399Ile genotype of TLR-4 in 38 X60 years old having X20 teeth with p1 mm CAL. of the 197 patients examined (19.29% of all individuals, Among our control population, we identified 111 allele frequency 9.64%, Table 2). This frequency was individuals displaying these characteristics, and 75 could significantly higher as compared to matched controls, be matched with CP-patients regarding gender and where 14 individuals displayed this genotype (OR 3.124 smoking-status, with all individuals being nonsmokers. 95% CI 1.633–5.976, P-value 0.0005, Table 2). No Here, we observed a higher frequency of TLR-4 SNPs individuals exhibiting SNPs in a homozygous state were among the CP-patients as compared to these controls identified. A Hardy–Weinberg test in combination with a confirming our previous results (OR 4.814, 95% CI 1.526– w2 test was performed, showing that these frequencies 15.19, P-value 0.007, Table 3), while there was no were in line with the Hardy–Weinberg equilibrium significant association with TLR-2 (data not shown). (P-values 0.3255 and 0.8749 for patients and controls, This association was highly significant when all TLR-4 respectively). Some patients displayed rare genotypes variants were taken into account (OR 6.022, 95% CI with Asp299Gly or Thr399Ile occurring separately: 1.938–18.72, P-value 0.0011).

Table 1 Composition of the patient and control group studied

Periodontitis cases Matched controls

n Mean age (years) Range Male (%) Smoker (%) n Mean age (years) Range Male (%) Smoker (%)

Total 197 49.77711.95 22–81 43.2 34.5 197 43.71716.03 22–83 43.2 34.5 CPa 116 57.0778.82 40–81 38.8 36.2 116 55.0710.53 40–83 38.8 36.2 APb 81 39.3176.94 22–54 49.4 32.1 81 27.5274.48 22–39 49.4 32.1 aChronic periodontitis. bAggressive periodontitis Informed consent was obtained from all individuals included in this study, and the study was performed in concordance with the clinical research guidelines of the Charite´ Medical Center, Berlin, Germany. All patients included were diagnosed by dentists trained at the Department for Periodontology and Synoptic Dentistry, Charite´ University Medical Center, Berlin (DM, VW, CC, DK, VH and PP). In total, 349 healthy individuals with no history of periodontitis were examined as controls. Probands being X60 years old were furthermore examined in detail for their periodontal status, those individuals presenting X20 periodontally healthy teeth (CAL p1 mm) were denoted. Patients and controls with malignant disease and/or diabetes mellitus were excluded from the study. Smokers were defined as individuals with a history of smoking equivalent to 10 years of smoking X20 cigarettes per day. All patients were German Caucasians except of 21 patients displaying AP included by VH being Croatian Caucasians. All controls were German Caucasians from the area of Berlin except of 27 individuals from the area of Constance contributed by CH. Control samples were shuffled randomly and matched for gender and smoking status. Age matching was performed by assigning controls to patients with a corresponding range of age (CP: 40–81 years, AP: 22–54 years).

Genes and Immunity TLR-4 polymorphisms in human periodontitis NWJ Schro¨der et al 450 Table 2 Genotypes of Asp299Gly and Thr399Ile of TLR-4 Table 3 Comparison of CP-patients with periodontally healthy controls Genetic variant All (n ¼ 197) CP (n ¼ 116) AP (n ¼ 81) Genetic variant Patients (n ¼ 75), Controls (n ¼ 75), 7 7 TLR-4 Asp299Gly and Thr399Ile 57.44 8.30 67.45 5.41 Patients n (%) 38 (19.29) 22 (18.97) 16 (19.75) years, 34.67% years, 34.67% Controls n (%) 14 (7.11) 6 (5.17) 8 (9.88) male male OR 3.124 4.291 2.246 95% CI 1.633–5.976 1.669–11.03 0.902–5.593 TLR-4 Asp299Gly and Thr399Ile P-value 0.0005 0.0020 0.1203 n (%) 16 (21.33) 4 (5.33) OR 4.814 TLR-4 Asp299Gly or Thr399Ile 95% CI 1.526–15.19 Patients n (%) 5 (2.54) 5 (2.54) 0 (0) P-value 0.0070 Controls n (%) 0 (0) 0 (0) 0 (0) OR 11.29 11.49 ND TLR-4 Asp299Gly or Thr399Ile 95% CI 0.619–205.6 0.628–210.4 Patients n (%) 3 (4.0) 0 (0) P-value 0.0609 0.0598 ND OR 7.290 95% CI 0.307–143.7 TLR-4 all variants estimated P-value 0.2450 Patients n (%) 43 (21.83) 27 (23.28) 16 (19.75) Controls n (%) 14 (7.11) 6 (5.17) 8 (9.88) TLR-4 all variants estimated OR 3.650 5.562 2.246 Patients n (%) 19 (25.33) 4 (5.33) 95% CI 1.573–8.467 2.199–14.04 0.902–5.593 OR 6.022 P-value o0.0001 0.0001 0.1203 95% CI 1.938–18.72 P-value 0.0011 TLR-2 Arg753Gln Patients n (%) 11 (5.58) 5 (4.31) 6 (7.41) X X Controls n (%) 18 (9.14) 9 (7.76) 9 (11.11) Control probands being 60 years of age with 20 teeth with a OR 0.588 0.536 0.640 clinical attachment loss p1 mm were compared to matched patients 95% CI 0.270–1.280 0.174–1.650 0.217–1.890 regarding the presence of TLR-4 SNPs. All individuals included P-value 0.2467 0.4092 0.5891 here were nonsmokers. Significant P-values are shown in boldface.

DNA was extracted from buccal swabs employing the Qiagen DNA blood mini kit (Qiagen, Hilden, Germany) according to the manufacturer’s protocol for body fluids. Allele-specific PCR was performed in order to detect SNPs of TLR-4 employing Qiagen Taq Furthermore, the fact that rare variants of TLR-4 were DNA polymerase (Qiagen, Hilden, Germany) as well as deoxyr- exclusively found in CP patients is in line with the ibonucleoside triphosphate mix (Clontech, Heidelberg, Germany) described association of rare variants of TLR-4 with on a Biometra Trio-block. We employed the following primers meningitis.24 Our data are in contrast to a recently for Asp299Gly: 50-GATTAGCATACTTAGACTACTACCTCCATG-30 published study investigating Asp299Gly and Thr399Ile (forward) and 50-GATCAACTTCTGAAAAACGATTCCCAC-30 and periodontal disease describing the lack of significant (reverse) and for Thr399Ile: 50-GGTTGCT-GTTCTCAAAGTGATTT differences between patients and controls.25 However, in TGGGAGAA-30 (forward) and 50-CCTGAAGACTGGAGAGTG-AG this study only 52 and 35 patients, respectively, matched 0 TTAAATGCT-3 (reverse) according to the previously published our criteria for moderate or severe generalized period- 29 protocol. A measure of 4 ml of PCR products were digested with ontitis, in contrast to 187 patients studied here, and 1 NcoI(Asp299Gly) and HinfI (Thr399Ile, 0.1 U/sample each) at 37 C statistical analysis failed to imply gender, smoking overnight, followed by electrophoresis on 1% Agarose gels (Roth, status, as well as age. Furthermore, in this study no Braunschweig, Germany) and 0.75% Nusieve (Biozym, Hessisch stratification for AP and CP was performed and—as Oldendorf, Germany). The Arg753Gln SNP was investigated patients as young as 25 years were recruited—the employing allele-specific PCR according to a previously published protocol.20 A 340 bp fragment was amplified via PCR and 4 mlof distinct clinical entities CP and AP were analyzed in a PCR products were digested with AciI (0.125 U/sample) at 371C combined fashion. Since we did not observe any overnight, followed by electrophoresis on gels as described above. association with TLR-4 SNPs and AP in our study, these Statistical analysis was performed employing GraphPad Prism 4.0 patients may have blurred the results. software for Windows. We employed a two-tailed Fisher’s exact test LPS of two bacteria associated with periodontitis, as well as Odds ratio with 95% confidence interval (CI). Since we Prevotella intermedia and Actinobacillus actinomycetemco- did not detect homozygous individuals among patients and mitans, have been identified as TLR-4-ligands,26,27 and controls, statistics were performed considering frequencies of there is evidence that LPS derived from P. gingivalis at heterozygous individuals. Significant P-values are shown in bold- least partially interacts with TLR-4.12 Cytokines released face. by macrophages responding to LPS via TLR-4 may display harmful effects on the periodontal tissue8,9 aggravating disease. On the other hand, macrophages We provide evidence that common variants of TLR-4 play an important role in combating bacteria invading act as risk factors for generalized CP. Regarding AP, our the subgingival space,28 thus preventing the establishing data do not show a significant association, although there of an infection. Our data indicate that the presence of the was an approx two-fold higher frequency of TLR-4 SNPs Asp299Gly/Thr399Ile variant sited within the extra- among AP patients as compared to controls. Our results cellular domain of TLR-4 predominantly involved in are in line with studies describing that these SNPs are ligand interaction, potentially influences susceptibility to associated with infections caused by Gram-negative periodontitis by causing an impaired recognition of bacteria, septic shock and Legionnaire’s disease.21–23 pathogens leading to an aggravated infection.

Genes and Immunity TLR-4 polymorphisms in human periodontitis NWJ Schro¨der et al 451 Acknowledgements from patients with periodontitis. Oral Microbiol Immunol 2003; 18: 54–58. This study was supported by grants given by the 15 Hou L, Sasaki H, Stashenko P. Toll-like receptor 4-deficient Deutsche Forschungsgemeinschaft (DFG) to NWJS and mice have reduced bone destruction following mixed anaero- RRS (Schr 726/1–2), and to DK and PP (GRK 325); and by bic infection. Infect Immun 2000; 68: 4681–4687. the BMBF to RRS (01KI0104). We hereby state that they 16 Kornman KS, Crane A, Wang HY et al. The interleukin-1 do not have any commercial or other association that genotype as a severity factor in adult periodontal disease. 24 might pose a conflict of interest. The excellent technical J Clin Periodontol 1997; : 72–77. 17 Parkhill JM, Hennig BJ, Chapple IL, Heasman PA, Taylor JJ. assistance of Fra¨nzi Creutzburg, Sabine Bobbe and Diana Association of interleukin-1 gene polymorphisms with early- Woellner is acknowledged. onset periodontitis. J Clin Periodontol 2000; 27: 682–689. 18 Schro¨der NWJ, Schumann RR. Single nucleotide polymorphisms of Toll-like receptors and susceptibility to infectious disease. Lancet Infect Dis 2005; 5: 156–164. References 19 Lorenz E, Mira JP, Cornish KL, Arbour NC, Schwartz DA. A novel polymorphism in the toll-like receptor 2 gene and its 1 Williams RC. Periodontal disease. N Engl J Med 1990; 322: potential association with staphylococcal infection. Infect 373–382. Immun 2000; 68: 6398–6401. 2 Shaw JH. Causes and control of dental caries. N Engl J Med 20 Schro¨der NWJ, Hermann C, Hamann L et al. High frequency 1987; 317: 996–1004. of polymorphism Arg753Gln of the Toll-like receptor-2 gene 3 Papapanou PN. Epidemiology of periodontal diseases: an detected by a novel allele-specific PCR. J Mol Med 2003; 81: update. J Int Acad Periodontol 1999; 1: 110–116. 368–372. 4 Brown LJ, Loe H. Prevalence, extent, severity and progression 21 Agnese DM, Calvano JE, Hahm SJ et al. Human toll-like of periodontal disease. Periodontology 2000 1993; 2: 57–71. receptor 4 mutations but not CD14 polymorphisms are 5 American Academy of periodontology. Parameter on aggres- associated with an increased risk of Gram-negative infections. sive periodontitis. J Periodontol 2000; 71: 867–869. J Infect Dis 2002; 186: 1522–1525. 6 Li L, Messas E, Batista Jr EL, Levine RA, Amar S. 22 Lorenz E, Mira JP, Frees KL, Schwartz DA. Relevance of Porphyromonas gingivalis infection accelerates the progression mutations in the TLR4 receptor in patients with Gram- of atherosclerosis in a heterozygous apolipoprotein E-deficient negative septic shock. Arch Intern Med 2002; 162: 1028–1032. murine model. Circulation 2002; 105: 861–867. 23 Hawn TR, Verbon V, Janer M et al. Toll-like receptor 7 Nelson KE, Fleischmann RD, DeBoy RT et al. Complete 4 polymorphisms are associated with resistance to genome sequence of the oral pathogenic Bacterium Porphyr- Legionnaires’ disease. Proc Natl Acad Sci USA 2005; 102: omonas gingivalis strain W83. J Bacteriol 2003; 185: 5591–5601. 2487–2489. 8 Stashenko P, Jandinski JJ, Fujiyoshi P, Rynar J, Socransky SS. 24 Smirnova I, Mann N, Dols A et al. Assay of locus-specific Tissue levels of bone resorptive cytokines in periodontal genetic load implicates rare Toll-like receptor 4 mutations in disease. J Periodontol 1991; 62: 504–509. meningococcal susceptibility. Proc Natl Acad Sci USA 2003; 100: 9 Stashenko P, Dewhirst FE, Peros WJ, Kent RL, Ago JM. 6075–6080. Synergistic interactions between interleukin 1, tumor necrosis 25 Folwaczny M, Glas J, Torok HP, Limbersky O, Folwaczny C. factor, and lymphotoxin in bone resorption. J Immunol 1987; Toll-like receptor (TLR) 2 and 4 mutations in periodontal 138: 1464–1468. disease. Clin Exp Immunol 2004; 135: 330–335. 10 Barton GM, Medzhitov R. Toll-like receptor signaling path- 26 Hashimoto M, Asai Y, Tamai R et al. Chemical structure ways. Science 2003; 300: 1524–1525. and immunobiological activity of lipid A from Prevotella 11 Bainbridge BW, Darveau RP. Porphyromonas gingivalis lipopoly- intermedia ATCC 25611 lipopolysaccharide. FEBS Lett 2003; saccharide: an unusual pattern recognition receptor ligand for 543: 98–102. the innate host defense system. Acta Odontol Scand 2001; 59: 27 Suzuki T, Kobayashi M, Isatsu K et al. Mechanisms involved 131–138. in apoptosis of human macrophages induced by 12 Darveau RP, Pham TT, Lemley K et al. Porphyromonas gingivalis lipopolysaccharide from Actinobacillus actinomycetemcomitans lipopolysaccharide contains multiple lipid A species that in the presence of cycloheximide. Infect Immun 2004; 72: functionally interact with both toll-like receptors 2 and 4. Infect 1856–1865. Immun 2004; 72: 5041–5051. 28 Kornman KS, Page RC, Tonetti MS. The host response to the 13 Kirschning CJ, Schumann RR. TLR2: cellular sensor for microbial challenge in periodontitis: assembling the players. microbial and endogenous molecular patterns. Curr Top Periodontology 2000 1997; 14: 33–53. Microbiol Immunol 2002; 270: 121–144. 29 Lorenz E, Frees KL, Schwartz DA. Determination of the TLR4 14 Mori Y, Yoshimura A, Ukai T et al. Immunohistochemical genotype using allele-specific PCR. Biotechniques 2001; 31: localization of Toll-like receptors 2 and 4 in gingival tissue 22–24.

Genes and Immunity