| Sign In to gain access to subscriptions and/or personal tools. |
Evidence of a Substantial Genetic Basis for IgG2 Levels in Families with Aggressive Periodontitis
1 Center for Pharmacogenomics and Complex Disease Research, New Jersey Dental School, UMDNJ, 185 South Orange Ave, MSB C-636, Newark, NJ 07101-1709; and Correspondence: * corresponding author, diehlsr{at}umdnj.edu
IgG2 is elevated in localized but not in generalized aggressive periodontitis (AgP). Exposure to pathogenic bacteria is essential for disease. Immune responses are dominated by IgG2 reactive with bacterial surface carbohydrates. We used variance component analyses to assess IgG2 heritability and determine whether genes that influence IgG2 are the same genes that influence disease susceptibility. We studied 17 Caucasian and 43 African American families with two or more localized or generalized AgP-affected members (274 subjects with IgG2 measurements). Only 16% of the variance in IgG2 was attributable to age, race, and smoking. Even with the addition of localized AgP, the model still explained only 19% of IgG2 variance. By contrast, heritability of IgG2 levels was estimated to be 38% and highly significant (P = 0.0006), demonstrating a substantial genetic basis. Bi-trait variance component analyses of IgG2 and quantitative measures of AgP indicate that different genes appear to control IgG2 levels and disease susceptibility.
Key Words: heritability • variance components • genetic correlation • host susceptibility • antibody response
Only a few bacteria species are considered primary agents of disease initiation in aggressive periodontitis (AgP) (Haffajee and Socransky, 1994). Through the action of their toxins, enzymes, or other virulence factors, these bacteria may directly damage periodontal tissues. In some subjects, host immune responses eliminate the infection. In others, however, bacteria persist in large numbers in subgingival biofilms and tissues, and the immune system produces a hyperinflammatory response that results in much greater destruction of periodontal tissues (Offenbacher, 1996). AgP aggregates in families, and this suggests that genetic variations in host responses play an important role in disease susceptibility (Schenkein, 2002). It is likely that AgP has a complex etiology, with interactions of multiple susceptibility genes and environmental factors (Diehl and White, 2001). A major portion of antibody to periodontitis-associated bacteria is of the IgG2 subclass and is reactive with serotype-specific carbohydrate antigens (Wilson and Hamilton, 1992; Lu et al., 1994). Individuals affected by localized AgP have elevated serum IgG2 levels (Zhang et al., 1996). High levels of IgG2 reactive with serotype-specific antigen of A. actinomycetemcomitans correlate with less severe disease in generalized AgP (Califano et al., 1996). These findings suggest that genetic differences in AgP susceptibility may be mediated by antibody responses. A segregation analysis of IgG2 in families with one or more AgP-affected members suggested that IgG2 levels are controlled by a single gene (Marazita et al., 1996). However, studies of other antibodies, such as IgE, indicated that the response is controlled by multiple genes (Mathias et al., 2001). Smoking is associated with reduced IgG2, but this appeared to involve interaction of race, localized vs. generalized AgP (Quinn et al., 1996), and possibly also IgG allotypes (Gunsolley et al., 1997). The association with allotypes was not confirmed in another study (Colombo et al., 1998). Recent advances in algorithms and software allow variance component analysis of quantitative traits to be applied to extended human pedigrees (Almasy and Blangero, 1998). After adjusting for covariates such as age, sex, and smoking, one can use the patterns of similarity among family members to partition the residual variance between an "environmental" component and that which aggregates within families. The latter is formally called "heritability", although studies of twins or molecular markers are needed to rule out other causes of similarity among close relatives. This method enables us to evaluate realistic models of complex inheritance involving multiple genetic and environmental factors. We applied variance component analysis to 60 families with two or more members affected by localized or generalized AgP. We estimated heritability of IgG2 and effects of age, race, sex, smoking, and the diagnosis of localized AgP. We also applied variance component analyses to quantitative measures of AgP [means of attachment loss (AL) for various groups of teeth] and to smoking, an established risk factor for periodontitis (Albandar and Rams, 2002). Our aim was to determine whether these traits have a heritable basis and, if so, whether genes that influence IgG2 are the same genes that control AgP susceptibility or smoking behavior.
Clinical Study Sample We recruited 17 Caucasian and 43 African American families for a study of the genetic basis of susceptibility to AgP. Probands were referred to the Clinical Research Center for Periodontal Disease from School of Dentistry Clinics at Virginia Commonwealth University or by clinicians in the Richmond, Virginia, area. Families were chosen for participation based on finding multiple cases of AgP among first-, second-, or third-degree relatives of the proband. In all, 79 periodontally healthy, 51 localized AgP-affected, 83 generalized AgP-affected, 55 chronic periodontitis, and six edentulous subjects with periodontitis history unknown were evaluated for IgG2. Individual family members were excluded if diabetes or another systemic disorder known to be associated with periodontitis was present. The protocol was reviewed and approved by an Institutional Review Board at Virginia Commonwealth University, and informed consent was obtained from all study subjects. Details about recruitment of the clinical sample and methods of diagnosis are reported elsewhere (Diehl et al., 1999). In brief, calibrated examiners measured AL at 4 locations on each tooth (lingual, buccal, proximal, and distal). As in most studies of AgP, we summarized these quantitative measures of AL and classified subjects as healthy, localized AgP, generalized AgP, or chronic periodontitis (Diehl et al., 1999). In addition, however, we also analyzed AgP as a quantitative trait and conducted bi-trait variance component analyses to determine whether or not the host response genes that influence IgG2 are the same genes that influence susceptibility to AgP. We first selected the maximum AL measurement obtained for each tooth from among the 4 measurements taken. We then calculated means across the individual teeth, either for all teeth or for subsets of teeth, such as only incisors or only first molars. We then analyzed these quantitative variables jointly with the subjects’ antibody levels to assess whether or not the same genetic and environmental factors influence their variation.
Laboratory Assays
Statistical Analyses
Bi-trait analyses of IgG2 and AgP, or IgG2 and smoking, estimated genetic and environmental correlations between these pairs of traits. The bi-trait analysis is available only for quantitative variables, so we analyzed AgP using means of AL for different groups of teeth, such as incisors and first molars. AL measures did not exhibit kurtosis greater than 1.0 and thus were not transformed. The localized AgP covariate was not included in these analyses. Smoking was analyzed as the quantitative measure of serum cotinine, square-root-transformed to reduce kurtosis.
Heritability and Demographic, Smoking, and Localized AgP Effects Differences among individuals in serum IgG2 levels show strong familial aggregation. We estimated heritability of IgG2 to be 38.2% (P = 0.0006) in a model that included demographic factors and smoking but without a parameter for localized AgP (Table 1  ). Although statistically significant, these covariates together accounted for only 16.4% of the variance in IgG2. When localized AgP was included, this parameter was significant (P = 0.002), heritability was slightly higher (39.1%, P = 0.0005), and the significance levels of the other covariates were about the same. Including localized AgP increased the proportion of IgG2 variance explained by less than 3%, to 19.2%.
Black female non-smokers had the highest levels of IgG2 among subjects who did not have localized AgP, followed by black male non-smokers, black female smokers, and black male smokers (Fig. 1A  ). Race had the largest overall effect on IgG2, with about 1000 µg/mL higher IgG2 in blacks compared with whites. Consequently, white subjects of both sexes, including smokers and non-smokers, had lower predicted IgG2 levels than blacks. The curves shown in the Fig. are parallel because only main effects were included in the model. We tested interactions of race and smoking effects because of previous reports in the literature. However, these were not statistically significant and therefore were not included in the final models shown in the Figs.
Subjects with localized AgP have elevated IgG2 levels (Fig. 1B
Since the covariates explain less than 20% of the variance of IgG2, individuals’ actual measures vary widely from the model’s predictions (Fig. 2
Independent Genetic Basis for IgG2, AgP Susceptibility, and Smoking Bi-trait variance component analyses of IgG2 with 5 alternative quantitative measures of AL yielded heritability estimates for IgG2 ranging from 33.1% to 33.4% (Table 2 ). Heritability of AL ranged from 15.8% to 28.6%, with highest heritability obtained for first molars. Genetic correlations between IgG2 and AL measures are positive but low, ranging from 0.02 to 0.17, and none is statistically significant; nor are any of the environmental correlations.
Heritability for smoking was 29.0% in a model without a covariate for subjects’ AgP status. When the AgP covariate (which combined both localized and generalized forms) was added to the model, heritability for smoking was 29.2%. Genetic and environmental correlations between IgG2 and smoking were consistently negative, but not statistically significant.
IgG2 levels show strong familial aggregation, with heritability estimated at 39%. Familial aggregation accounts for a greater proportion of IgG2 variance than the combined effects of age, sex, race, smoking, and localized AgP status (19%). Previously reported segregation analyses of IgG2 suggested the importance of genetic variation. However, segregation analyses attribute all or most of the heritable variation to a single "major" gene. Studies of other antibodies such as IgE indicate that the single-gene model is not appropriate (Mathias et al., 2001). The variance component method we used assesses realistic, complex disease models, with multiple genes (Blangero et al., 2001). Thus, our finding of substantial heritability may be more robust and reliable.
IgG2 levels increase with age until age 50, then slowly decline (Fig. 1A
Bi-trait analyses indicated that genes that influence IgG2 are not the same genes as those that influence AgP susceptibility (Table 2
Previous studies indicate that heritability of smoking behavior ranges from 37% to 59% (Li et al., 2003). Our estimate of 29% with a standard error of 11% (Table 2 Data on antibody responses in periodontitis are somewhat inconsistent. Lehner et al. (1974) found elevated IgA, IgG, and IgM in AgP, and Waldrop et al. (1981) reported higher IgG in AgP, but similar concentrations of subclasses. Lu et al. (1994) observed elevated IgG2 levels in localized but not in all generalized AgP subjects. Quinn et al. (1996) found elevated IgG2 levels only in non-smoking generalized AgP. Albandar et al. (2002) reported higher antibody levels in African-Americans compared with Caucasians. They found no IgG or subclass associations with AgP, although with only 13 localized AgP subjects, statistical power was limited. Localized AgP subjects’ response to A. actinomycetemcomitans leukotoxin is primarily IgG1 (Califano et al., 1997), while both localized AgP and non-smoking generalized AgP subjects’ response to LPS is primarily IgG2 (Tangada et al., 1997). Generalized AgP with very high levels of anti-LPS IgG2 has significantly less AL than generalized AgP with lower levels of these antibodies (Califano et al., 1996). Albandar et al. (2001) found significantly higher levels of IgG and IgA reactive to P. gingivalis and A. actinomycetemcomitans and IgA antibody reactive to P. intermedia in generalized AgP compared with healthy controls, but no differences between controls and 13 localized AgP subjects. Craig et al. (2002) reported that IgG reactive to P. gingivalis was higher in African-Americans compared with Asians and Hispanics, but also found greater probing depth and AL, more missing teeth, and a higher number of unskilled laborers in African-American subjects. They suggested that "environmental and socio-economic variables may have a greater influence on serum IgG antibody levels in these populations." Data on immunoglobulin allotypes are inconsistent. Gunsolley et al. (1997) reported interactions among allotypes, subtypes of AgP, and smoking affecting IgG2. G2m(n)+ allotype subjects with generalized AgP in Korea had higher IgG2 reactive to A. actinomycetemcomitans, while Km(1)+ subjects with localized AgP had higher IgG2 reactive to P. gingivalis (Choi et al., 1996). However, Colombo et al. (1998) reported that neither serum IgG2 nor allotypes were strongly related to refractory periodontal disease. We found no evidence of interaction between race and smoking relative to IgG2, so we did not evaluate the higher-order, three-way interactions of race, smoking, and localized vs. generalized AgP suggested previously (Quinn et al., 1996, 1998). Smoking is associated with reduced IgG2 reactive with A. actinomycetemcomitans in African Americans with generalized AgP, but IgG2 reactive with other antigens appears unaffected by smoking (Tangada et al., 1997). Lower levels of antibodies reactive with A. actinomycetemcomitans, P. intermedia, and T. denticola were found in smoking vs. non-smoking generalized AgP patients on maintenance therapy, but there was no effect of smoking in untreated patients (Mooney et al., 2001). We used variance component analyses to identify factors influencing levels of IgG2. We found significant associations with age, race, smoking, and localized AgP, but these variables explain only 19% of IgG2 variance. We estimated heritability of IgG2 levels to be 38%. This substantial genetic basis for individual differences in IgG2 is similar to findings for other antibodies. We also found that IgG2 and susceptibility to AgP are genetically uncorrelated, indicating that these traits are controlled by different genes.
This work was supported by US Public Health Service grant DE-13102 from the National Institute of Dental and Craniofacial Research, and by resources provided by the University of Medicine and Dentistry of New Jersey and Virginia Commonwealth University. We thank our study families for their cooperation in this research, and express our appreciation for the technical assistance of J. Francis, K. Lake, P. Ober, D. Ruggles, G. Smith, and D. Williams. We thank J. Blangero, C. Peterson, and T. Dyer for technical guidance and interpretation of SOLAR analyses, J. Cole for editorial assistance, and R.P. Erickson for scientific review of the manuscript. Received for publication January 29, 2003. Revision received April 27, 2003. Accepted for publication June 10, 2003.
Journal of Dental Research, Vol. 82, No. 9,
708-712 (2003)
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
