| Sign In to gain access to subscriptions and/or personal tools. |
Anti-cardiolipin Antibodies in Sera from Patients with PeriodontitisClinical Research Center for Periodontal Disease, Virginia Commonwealth University, School of Dentistry, PO Box 980566, Richmond, VA 23298-0566, USA; Correspondence: * corresponding author, haschenk{at}vcu.edu
Antiphospholipid antibodies are commonly found in patients with systemic lupus erythematosus or the antiphospholipid syndrome, and a subset of such antibodies is associated with prothrombotic events such as stroke and with adverse pregnancy outcomes and fetal loss. We examined sera from 411 patients who were clinically characterized as to their periodontal disease status for serum levels of β2-glycoprotein I-dependent anti-cardiolipin autoantibodies (anti-CL). The prevalence of patients with chronic periodontitis (CP) and generalized aggressive periodontitis (GAgP) positive for anti-CL (16.2% and 19.3%, respectively) was greater than that in healthy controls (NP) and localized aggressive periodontitis (LAgP) patients (6.8% and 3.2%). Patients with these autoantibodies demonstrated increased pocket depth and attachment loss compared with patients lacking the antibodies. Analysis of the data indicates that patients with generalized periodontitis have elevated levels of autoantibodies reactive with phospholipids. These antibodies could be involved in elevated risk for stroke, atherosclerosis, or pre-term birth in periodontitis patients.
Key Words: antibody • cardiolipin • periodontitis
Recent reports indicate that patients with periodontitis may be subject to an increased incidence of cardiovascular diseases (Beck et al., 2001; Beck and Offenbacher, 2001; Genco et al., 2002) and a greater chance of bearing pre-term low-birthweight babies (Jeffcoat et al., 2001; Offenbacher et al., 2001). It has been hypothesized that the biological explanation for the association of periodontitis with adverse pregnancy outcomes could be chronic exposure to oral bacterial pathogens, particularly Gram-negative bacteria, which stimulate systemic production of inflammatory mediators and antibodies that deleteriously affect the placenta. With respect to cardiovascular diseases and atherosclerosis, it has been proposed that periodontal pathogens not only enhance systemic inflammatory responses but also can directly and indirectly influence the formation of atheromas. Anti-phospholipid antibodies comprise a class of autoantibodies found in 1-5% of the systemically healthy population (Petri, 2000). Elevated levels of these antibodies can be found in several conditions, including a variety of infectious diseases, and are a hallmark of the Antiphospholipid Syndrome (APS). APS is present in about 30 to 40% of patients with SLE, although there are individuals with the primary form of APS who do not have SLE. The major clinical symptoms of APS include recurrent venous or arterial thrombosis and fetal loss, and patients with APS may demonstrate premature atherosclerosis (Levine et al., 2002). The pathogenesis of APS is related to the prothrombotic activity of some anti-phospholipid antibodies (Levine et al., 2002). Among the major groups of antibodies detected in APS patients are β-2-glycoprotein I-dependent anti-cardiolipin (anti-CL), anti-β-2-glycoprotein I (anti-β2GPI), and lupus anticoagulant (LA). β2GPI is a 50-kDa plasma phospholipid-binding protein that functions as a natural anticoagulant (Kandiah and Krilis, 1994). Immunoassays that measure pathogenic anti-CL require the incorporation of β2GPI bound to CL for detection of anti-CL autoantibodies that promote procoagulant activity (Galli et al., 1990; McNeil et al., 1990). Such antibodies are characteristic of patients with APS. Autoantibodies directed at β2GPI may also be directly detected in immunoassays that omit CL, though these subsets of anti-CL and anti-β2GPI may not be identical (Hojnik et al., 1994). Anti-β2GPI immunoassays have high specificity for the APS, while antibodies that bind directly to CL are also present in patients with syphilis or other infectious diseases. LA antibodies, also characteristic of patients with APS, are associated with thrombo-embolic events and are detected in clotting assays in which they prolong clotting (despite their clear association with in vivo thrombosis) (Levine et al., 2002). In summary, this group of antibodies is heterogeneous, and clinical tests for APS usually involved multiple assays to detect autoimmune anti-CL, anti-β2GPI, and LA. Recent studies strongly implicate bacterial and viral infections in the etiology of APS due to induction of cross-reactive anti-CL autoantibodies. Blank et al.(2002) identified a hexapeptide (TLRVYK) sequence in β2GPI that is recognized by some anti-β2GPI monoclonal antibodies. Mice immunized with microbial pathogens such as Hemophilus influenzae or Neisseria gonorrheae with homologous sequences related to TLRVYK produced cross-reactive anti-β2GPI that induced APS-like symptoms when subsequently purified and passively infused into mice. Thus, bacterial infections could lead to production of pathogenic anti-CL and be responsible for a subset of cases of APS. The remarkable similarity between the hallmark symptoms of APS and the reported systemic sequellae of periodontal infections, as well as the likely infectious origin of the anti-CL antibodies, suggested that patients with periodontitis may have elevated anti-CL antibodies that could partly explain the pathogenesis of these systemic disorders. To assess the possibility that periodontal infections could induce production of autoantibodies to phospholipids, we quantitated a series of antiphospholipid antibodies in sera from patients with a variety of periodontal diagnoses.
Clinical Methods This study was approved by the Internal Review Board of Virginia Commonwealth University. Subjects for this study were identified through the Virginia Commonwealth University School of Dentistry Clinics and the Virginia Twin Registry and included individuals participating in studies of families with aggressive periodontitis, twin pairs, and individually ascertained cases and controls. All subjects were determined, by history, to be systemically healthy. A complete periodontal evaluation—including assessments of pocket depth, attachment loss, plaque index (Silness and Löe, 1964), gingival index (Löe and Silness, 1963), bleeding upon probing (Mühlemann and Son, 1971), and suppuration (Singh et al., 1977)—was performed on each participant. Measurements were performed at 4 sites per tooth (mesiobuccal, midbuccal, distobuccal, and midlingual). At the time of the examination, a blood sample was taken and processed for serum, which was then stored at -70°C until utilized. The subjects were categorized by diagnostic group as follows:
Laboratory Methods
Statistical Analyses
Table 1  depicts the characteristics of the subject population utilized in this study. The groups varied with respect to age, gender, race, smoking habits, and mean attachment loss, due in part to the nature of their periodontal disease categories.
Demographic and clinical data for all subjects with positive and negative tests for anti-CL are shown in Table 2 . Subjects positive for these antibodies were those with either IgG or IgM levels greater than 15 units of antibody. Those subjects who were anti-CL-positive demonstrated significantly greater pocket depth and attachment loss than those who were negative. Additionally, patients who were former or current smokers were more likely to be positive for these antibodies. There were no differences noted in age, sex, or race between subjects positive and those negative for anti-CL.
We calculated the prevalence of positive tests for anti-CL in the 4 diagnostic groups (Table 3 ). Patients with generalized periodontitis had a significantly greater prevalence of IgG or IgM antibody (p = 0.0033). The prevalence in CP patients was 16.2%, while the prevalence in GAgP was 19.3%. In contrast, the overall prevalence of patients with positive tests for IgG or IgM antibodies in the healthy control group was 6.79% and that in the LAgP was 3.2%. The IgG and IgM assays by themselves approached statistical significance.
We also measured levels of anti-β2GPI IgG antibodies in a subset of 239 subjects distributed throughout the diagnostic categories (110 NP, 32 LAgP, 20 GAgP, and 77 CP subjects). Only three subjects (two NP, one CP patient) demonstrated elevated levels of this antibody.
We performed stepwise logistic regression analysis to determine which demographic, clinical, or diagnostic variables (including those variables shown in Table 2
Analysis of our data indicates that patients with generalized periodontal attachment loss, including both chronic periodontitis and generalized aggressive periodontitis, have an increased frequency of β2GPI-dependent anti-CL. Elevated anti-CL levels have been detected in patients with a variety of infections (Hojnik et al., 1994; McNally et al., 1995; Hansen et al., 1998; Celli et al., 1999; Petrovas et al., 1999; Dalekos et al., 2001; Santiago et al., 2001; Uthman and Gharavi, 2002). For the most part, these have not been shown to be β2GPI-specific antibodies and therefore would be classified as non-pathogenic with respect to the clinical manifestations of APS. Exceptions to these observations include a subset of patients with leprosy, HIV, leptospirosis, leishmaniasis, and syphilis in whom anti-β2GPI may be found with clinical manifestations consistent with the pathogenic properties of these antibodies. Further support for a possible microbial induction of pathogenic anti-CL comes from a recent study by Blank and co-workers (Blank et al., 2002). This group demonstrated that pathogenic anti-CL can be induced by immunizing mice with bacterial pathogens such as Hemophilus influenzae or Neisseria gonorrheae, which bear peptide sequences homologous to the TLRVYK peptide of β2GPI. Furthermore, Gharavi and co-workers (Gharavi et al., 2002) demonstrated homology between peptide sequences of β2GPI and sequences in cytomegalovirus, and raised anti-β2GPI antibodies via immunization of mice with such peptides. In a search of the Swiss Prot database, we found that 3 peptide sequences with homology to the TLRVYK peptide are present in the arg-gingipain protease of the periodontal pathogen Porphyromonas gingivalis. One such sequence is located in the propeptide region, but two are found in the hemagglutinin domain of the protease molecule. It is conceivable that P. gingivalis or other oral bacteria may contain homologous domains to peptides in β2GPI and induce cross-reactive autoantibodies. β2GPI is a plasma protein that binds to negatively charged phospholipids and is thought to provide a protective homeostatic mechanism preventing pathological prothrombotic reactions initiated by platelets or endothelial cells (Kandiah and Krilis, 1994). The pathogenesis of APS remains controversial. Pathogenic autoantibodies reactive with phospholipids which target β2GPI are thought to cause arterial and venous thrombosis and fetal loss by interfering with homeostatic β2GPI function. These antibodies are hypothesized to cause disease by several different mechanisms, including activation of endothelial cells, inducement of oxidant-mediated injury due to their reactivity with oxidized LDL, or interference with the natural anticoagulant function of β2GPI (Levine et al., 2002). Our data do not address the pathogenicity or function of the anti-CL found in periodontitis patients. However, β2GPI-dependent anti-CL are frequently pathogenic, and such antibodies may be induced by certain bacterial and viral pathogens. On the other hand, our failure to detect anti-β2GPI does not necessarily indicate a lack of pathogenic anti-CL; for example, several studies have demonstrated the relative insensitivity (but high specificity) of the anti-β2GPI immunoassay in detection of patients with verifiable APS, while the anti-CL assay is more sensitive (Forastiero et al., 1996; Sanmarco et al., 1997; Helbert et al., 2001). We have assessed the levels of antibodies reactive with cardiolipin itself (in the absence of β2GP1) and found that this assay does not distinguish among the periodontal diagnostic groups. We interpret this observation as indicating that the antibody specificity of the anti-CL in the sera from periodontitis subjects is β2GP1-dependent, but somewhat different from that required for binding to β2GP1 alone. Additional examination of these antibodies from patients with periodontitis is needed to determine their pathogenic potential. It is noteworthy that an increased frequency of anti-CL was seen in CP and GAgP patients compared with LAgP patients, even when age was included in the model as a covariate. Analysis of our data indicates that patients with elevated anti-CL have greater mean attachment loss and increased pocket depth. Thus, it appears that the presence, in some patients, of forms of periodontitis demonstrating greater extent and severity of disease and inflammation may lead to the production of anti-CL in those patients. Alternatively, differences in the bacterial flora between localized and generalized disease forms could also explain the lack of anti-CL in LAgP patients. Analysis of these data suggests that a subset of patients with generalized periodontitis produces elevated anti-CL antibodies during the course of chronic periodontal infection. Such antibodies at moderate but abnormal levels could be pathogenic and participate in the induction of the systemic sequellae of periodontitis. It has been proposed that increased incidence of stroke occurs in periodontitis patients (Joshipura, 2002), an observation consistent with the prothrombotic activity of pathogenic anti-CL. Additionally, it has been reported that adverse pregnancy outcomes are more common when the mother has severe generalized periodontitis (Jeffcoat et al., 2001; Offenbacher et al., 2001). The observation that fetal loss and impaired fetal growth are characteristic outcomes of APS is consistent with this characteristic of women with periodontitis. Finally, the known cross-reactivity of anti-CL with some anti-oxLDL antibodies, along with our unpublished observation that the levels of anti-CL and anti-oxLDL are highly correlated in the sera of the subjects reported upon in this study, indicates that these antibodies could also play a role in modifying the development of atherosclerosis. Additional studies are required to examine these possibilities.
The authors gratefully acknowledge Kimberly Lake and Gail Smith for clinical management of the subjects participating in this study. This work was supported by grant DE13102 from the National Institute of Dental and Craniofacial Research, National Institutes of Health. Received for publication May 28, 2003. Revision received August 26, 2003. Accepted for publication September 8, 2003.
Journal of Dental Research, Vol. 82, No. 11,
919-922 (2003) This article has been cited by other articles:
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
