Gene - Environment Interaction in Birth Defects

In the U.S., birth defects occur in 3-5% of live-births and comprise the leading cause of infant mortality, producing a substantial economic, social, and emotional burden. The case-control studies presented in this dissertation tested two major questions: 1) does an association exist between environmental factors, candidate genes and birth defects?; and 2) are there interactions between these genes and environmental factors? Article 1, Testing for Gene-Environment Interaction and Oral Clefts using Two Control Groups, combined case groups from two studies and compared them to two separate control groups: (1) non-cleft malformations and (2) nonmalformed infants. Genotype at the TGFA locus did not increase risk for oral clefts. There was an increased risk for cleft lip and palate (CL/P) and cleft palate only (CP) when compared to the non-malformed controls if the mother reported smoking during pregnancy. There was also evidence for gene-environment interaction among the CP group (malformed controls: OR=3.47, 95% CI=1.51-7.88; non-malformed controls: OR=4.21, 95% CI=1.48-12.79). In Research Study 2, Genetic and Environmental Risk Factors for Sagittal Craniosynostosis there were no associations with any of the examined environmental factors. All cases that were sequenced were negative for mutations at the following candidate genes: FGFR1 exon IIIa, FGFR2 (exons IIIa and IIIc,), FGFR3 exon IIIa , and TWIST exon I. In the third study, Methylenetetrahydrofolate Reductase and Congenital Heart Defects: Evidence for Gene-Environment Interaction, congenital heart defect cases and controls were genotyped for two MTHFR variants, C677T and A1298C. There were no differences between cases and controls for either of these MTHFR variants. Heavy alcohol use (OR=6.76, 95% CI=2.61-17.51) and heavy smoking (OR=4.70, 95% CI=1.38-16.04) increased risk for pulmonic stenosis among Caucasians. There was evidence of gene-environment interaction among the African-American pulmonic stenosis cases with maternal smoking and the MTHFR A1298C Cgenotype (nonsmokers: OR=0.98, 95%CI=0.27-3.42; smokers: OR=3.03, 95%CI=1.09-8.52). These studies strengthen our understanding of the complex interaction between genes and environment in the etiology of birth defects. BACKGROUND AND SIGNIFICANCE Birth defects represent an important public health issue, in the United States birth defects comprise the leading cause of infant mortality (20%) and years of potential life lost before age 65 (Yang et al., 1997). No major changes in frequency of congenital anomalies have occurred in the last 50 years; however, there is a declining trend in mortality rates from birth defects due to improved medical care and increased use of prenatal diagnosis. Congenital anomalies occur in about three to five percent of all live births. Defects range from non-life threatening malformations such as extra digits to fatal anomalies such as anencephaly. Many different factors play a role in the etiology of congenital malformations: genetic causes, 28%; multifactorial inheritance, 23%; teratogens, 3%; unknown, 43% (estimated by Nelson and Holmes, 1989). It is evident that birth defects are heterogeneous in their etiology; with an estimated 23% of birth defects considered multifactorial in nature, and the challenge is to determine the genetic and environmental factors that are involved. Genetic linkage studies, association studies, and family studies have shown associations between birth defects and oncogenes (e.g. BCL3 and oral clefts), growth factor genes (TGFA and oral clefts), and homeobox genes (CSX and heart defects). Environmental exposures that have been implicated in causing birth defects include maternal infections [rubella and cardiovascular defects, CNS and hearing defects, use of alcohol [oral clefts, cigarettes [oral clefts, craniosynostosis, heart defects, and drug use [antiepileptics and cardiovascular defects, oral clefts, and neural tube defects. Other studies have shown that genetic factors may interact with environmental factors to produce birth defects [smoking, TGFA and oral clefts. The last several years have seen advances in both molecular genetics and statistical methods. This has enabled researchers to more adequately assess not only environmental risk factors and their association with disease, but also genetic risk factors and how genes and environment may work together to increase risk to disease. Several study designs exist to examine the combined effects of genes and environment, ranging from traditional case-control study designs which incorporates standard epidemiologic data as well as marker information on candidate genes to family based study designs. The case-control study design has been commonly utilized to measure geneenvironment interaction in birth defects, as the studied outcomes are rare, and multiple exposures can be assessed. Two conditions must be met: (1) the environmental exposure is independent of the genotype, and (2) misclassification of the environmental exposure is non-differential with respect to the genotype. Appropriate selection of cases and controls is paramount to the case-control study design. When selecting cases, a clear definition of the case must first be devised, and then methods of identifying the cases must be developed. Improper or inconsistent diagnosis of cases can lead to misclassification and ultimately bias the results. In addition, different studies may devise their own case definition, which could lead to conflicting results and an apparent lack of reproducibility. Selecting a suitable control group in studies of gene-environment interaction is crucial; an inappropriate control group could lead to spurious findings as a result of confounding due to population stratification. For this reason, it is important that controls have a similar genetic background to cases, and when doing studies of candidate genes, analyses should be stratified by race. Population based studies provide an effective means of obtaining a representative sample of cases and controls. Several limitations exist when performing case-control studies. Participation rates may be low and often differ between cases and controls. There is also the potential for both differential and nondifferential misclassification. With nondifferential misclassification, both cases and controls make errors in exposure recall; there is increased potential for this to occur if the time between the exposure and the interview is prolonged. Differential misclassification, also referred to as recall bias, is said to occur when cases remember past exposures more thoroughly than controls causing an apparent effect. Many factors, both genetic and environmental, contribute to the occurrence of most, if not all, congenital malformations. This has created considerable obstacles in understanding their pathogenesis. Studying candidate genes elected for their biological relevance to the disease and identifying environmental risk factors may provide new information to understand the etiology of these birth defects and hopefully open up opportunities for effective strategies for primary prevention and genetic counseling. Genetic and Environmental Risk Factors for Sagittal Craniosynostosis Zeiger J, MS; Beaty T, PhD; Hetmanski J, MS; Wang H, MD, PhD; et al. ABSTRACT Objective: To investigate whether genetic and environmental factors influence risk for sagittal craniosynostosis. Design: Data were derived from a clinic based case-control study. Setting: Cases were ascertained from craniofacial clinics in the Baltimore-Washington region. Controls were recruited from a pediatric practice in Baltimore County. Patients: Fifty-one Caucasian probands with isolated, non-syndromic sagittal craniosynostosis born in the Mid-Atlantic region were included in this analysis. Controls are Caucasian infants born in Maryland without any known birth defect (n=109). Main Outcome Measures: Odds ratios (OR) and corresponding 95% confidence intervals (CI) were calculated. Cases were genotyped at several loci implicated in malformation syndromes that include craniosynostosis. Results: There were no elevated risks for craniosynostosis when considering maternal or paternal smoking and maternal vitamin usage. Case mothers consumed less alcohol (OR=0.10, 95% CI 0.04-0.26) and had less education than control mothers (p<0.001). All cases sequenced were negative for mutations at the following genes: FGFR1 exon IIIa 755C->G, FGFR2 (exons IIIa and IIIc,), FGFR3 exon IIIa, and TWIST exon I. Conclusions: These findings suggest that while TWIST and the FGFR genes are important in syndromic craniosynostosis; they are unlikely to be involved in isolated sagittal craniosynostosis. Parental education and alcohol consumption were associated with sagittal craniosynostosis. Testing for Gene-Environment Interaction and Oral Clefts using Two Control Groups J. Zeiger, T. Beaty, J. Hetmanski, I. McIntosh, M.Mracna, S-J. Hwang, K-Y. Liang, S. Panny, and C. VanderKolk ABSTRACT Purpose: This case-control analysis examines the effects of maternal smoking and infant genotype at the Taq1 polymorphism in the transforming growth factor alpha locus on risk to oral clefts using data from two published studies. Methods: Cases were Caucasian infants with an isolated cleft born in Maryland. Two controls groups were used: infants with isolated non-cleft birth defects and healthy infants. Polytomous logistic regression analyses were performed to estimate odds ratios and their 95% confidence intervals. All individuals were genotyped for the Taq1 polymorphism. Results: Genotype at the transforming growth factor alpha locus did not increase risk for oral clefts. There was an increased risk for cleft lip and palate (CL/P) and cleft palate only (CP) when compared to the healthy controls if the mother reported smoking during pregnancy (CL/P: OR=2.38, 95% CI=1.16-4.89; CP: OR=3.05, 95% CI=1.61-5.79). There was also evidence for statistically significant gene-environment interaction among the CP group. Conclusions: These findings suggest that maternal smoking during pregnancy may be a general risk factor for oral clefts independent of genotype and there may be etiology heterogeneity between types of clefts.Purpose: This case-control analysis examines the effects of maternal smoking and infant genotype at the Taq1 polymorphism in the transforming growth factor alpha locus on risk to oral clefts using data from two published studies. Methods: Cases were Caucasian infants with an isolated cleft born in Maryland. Two controls groups were used: infants with isolated non-cleft birth defects and healthy infants. Polytomous logistic regression analyses were performed to estimate odds ratios and their 95% confidence intervals. All individuals were genotyped for the Taq1 polymorphism. Results: Genotype at the transforming growth factor alpha locus did not increase risk for oral clefts. There was an increased risk for cleft lip and palate (CL/P) and cleft palate only (CP) when compared to the healthy controls if the mother reported smoking during pregnancy (CL/P: OR=2.38, 95% CI=1.16-4.89; CP: OR=3.05, 95% CI=1.61-5.79). There was also evidence for statistically significant gene-environment interaction among the CP group. Conclusions: These findings suggest that maternal smoking during pregnancy may be a general risk factor for oral clefts independent of genotype and there may be etiology heterogeneity between types of clefts. Methylenetetrahydrofolate Reductase and Congenital Heart Defects: Evidence for Gene-Environment Interaction Zeiger J, Hetmanski J, McIntosh I, Mracna M, Dey S, Ochiai L, Loffredo C, Beaty TH ABSTRACT Congenital heart defects represent a major public health issue; collectively they are the most common birth defect creating a considerable medical, emotional and economic burden. This case-control study examined two common poylmorphisms, C677T and A1298C, in the gene encoding for methylenetetrahydrofolate reductase. We tested for their possible involvement in the etiology of four common types of congenital heart defects: aortic stenosis (n=36), coarctation of the aorta (n=64), pulmonic stenosis (n= 120), and atrial septal defect (n=108). Odds ratios with 95% confidence intervals were calculated to test for associations between genotype and case/control status, and to test for interactions between genotype and environmental factors. There were no differences between cases and controls (n=480) for either of the methylenetetrahydrofolate reductase variants. Heavy alcohol use (OR=6.76, 95% CI=2.61-17.51) and heavy smoking (OR=4.70, 95% CI = 1.38-16.04) increased risk for pulmonic stenosis among Caucasians. We did not find strong evidence of association with these genotypes with these specific types of CHD, but there was some evidence of interaction between maternal smoking and the 1298C genotype among African American pulmonic stenosis cases (OR nonsmokers: 0.98, 95% CI= 0.27-3.42; OR smokers: 3.03, 95% CI=1.09-8.52).Congenital heart defects represent a major public health issue; collectively they are the most common birth defect creating a considerable medical, emotional and economic burden. This case-control study examined two common poylmorphisms, C677T and A1298C, in the gene encoding for methylenetetrahydrofolate reductase. We tested for their possible involvement in the etiology of four common types of congenital heart defects: aortic stenosis (n=36), coarctation of the aorta (n=64), pulmonic stenosis (n= 120), and atrial septal defect (n=108). Odds ratios with 95% confidence intervals were calculated to test for associations between genotype and case/control status, and to test for interactions between genotype and environmental factors. There were no differences between cases and controls (n=480) for either of the methylenetetrahydrofolate reductase variants. Heavy alcohol use (OR=6.76, 95% CI=2.61-17.51) and heavy smoking (OR=4.70, 95% CI = 1.38-16.04) increased risk for pulmonic stenosis among Caucasians. We did not find strong evidence of association with these genotypes with these specific types of CHD, but there was some evidence of interaction between maternal smoking and the 1298C genotype among African American pulmonic stenosis cases (OR nonsmokers: 0.98, 95% CI= 0.27-3.42; OR smokers: 3.03, 95% CI=1.09-8.52). Gene-Environment Interaction and Risk to Oral Clefts J.S. Zeiger and T.H. Beaty INTRODUCTION Oral clefts include cleft lip (CL), cleft palate (CP) and cleft lip and palate (CLP) and collectively these constitute a heterogeneous group of non-fatal birth defects known to be multifactorial in origin, in that both genes and environmental factors contribute to their etiology. It is possible to reduce heterogeneity in a sample of oral cleft cases by eliminating infants with recognized malformation syndromes (genetic or teratogenic) that can include oral clefts, and infants with multiple anomalies that may not fall into any recognized malformation syndrome. Even so isolated, non-syndromic oral clefts represent a complex and heterogeneous group of birth defects where there is strong evidence of an etiologic role for both genetic and environmental factors. It is imperative, therefore, to design studies of oral clefts to consider the effects of both genes and environmental factors, as well as their possible interaction. A variety of study designs are available, each with their own advantages and limitations. Here we review how these different study designs can be used to test for gene-environment interaction and then describe published studies of oral clefts that incorporate gene-environment interactions to one degree or another. In epidemiologic studies, interaction occurs when the joint effect of two risk factors is greater (synergism) or less than a simple combination of their individual effects. In the context of gene-environment interaction (GxE), the isolated effects of a genotype (or of an allele) and of some observed environmental exposure fail to predict the effect seen when both are present. While interactions are commonly described as synergistic effects, it is important to realize that a number of patterns of risk in the presence of the gene alone, environment alone, and both are possible, including negative or protective effects. Investigators have described five “biologically plausible” models of relationships between genotype, environmental exposure, and their effect on disease risk. These models include the following: (1) the genotype increases expression of the risk factor, but which can act on its own; (2) the high risk genotype exacerbates the effect of the risk factor, but the genotype has no effect in unexposed individuals; (3) the exposure exacerbates the effect of the high-risk genotype, but there is no effect in individuals who are exposed if they have the low-risk genotype; (4) both exposure and the high risk genotype are needed to alter risk; and (5) the exposure and the genotype both effect disease risk separately, but the risk is higher or lower when they occur simultaneously. The marginal effects of the gene or the environment may or may not be apparent when interaction does exist. A number of epidemiologic study designs can be used to test for gene-environment interaction, as discussed below, and several of these have been used to demonstrate gene-environment interaction in studies of oral clefts. CONCLUSIONS Studying the causes of birth defects is difficult; both genetic and environmental factors contribute to risk, and the possibility that genes and environment act together to determine risk cannot be overlooked . To address this latter issue, gene-environment interaction has become a primary focus in the study of birth defects. While pharmacogenetic studies have led the field in demonstrating that different genotypes respond variously to environmental exposures, the field of birth defects research has not yet been able to identify biologic mechanisms by which exogenous factors (such as smoking or alcohol) affect gene function to increase risk to the developing fetus. Several different study designs and statistical tests exist to assess the joint effects of genes and environment, with case-control studies being the traditional method by which gene-environment interaction can be assessed in the context of birth defects. Cases with the disease of interest are compared to unaffected controls. Both groups are genotyped for the markers of interest and environmental exposure information is collected. The ability to test for the marginal effects of genotype and environmental exposures are two advantages to this case-control study design; however, the ever present possibility of population stratification which can lead to spurious associations is perhaps the biggest drawback to this design. Case-only and family-based study designs have been devised to help overcome the problems inherent in the case-control study design. Case-only analyses involve comparing cases without the susceptibility genotype (i.e. the control group) to cases with the susceptibility genotype with respect to the presence or absence of the environmental exposure. The main effects of either genotype or the environmental risk factor cannot be tested separately, but the case-only design offers an inexpensive way to test for gene-environment interaction. Family-based designs utilize sibs, parents or cousins to test for geneenvironment interaction. In these studies, each case is compared to an unaffected relative who shares genes to some extent. Case-parent trio designs allow for a case/control analysis using parents to define a “matched” control as the allele (or genotype) not transmitted to the affected child, such as the transmission disequilibrium test (TDT). This type of analysis has the advantage of controlling for heterogeneity between populations which can lead to false associations, but requires DNA from the parents. It is also important to realize that detecting an association between a genetic marker and a disease does not imply evidence of a biological mechanism, as the association could arise from: (1) the genetic marker playing a role in disease etiology, (2) an indirect association between the genetic marker and a susceptibility locus nearby, whereby the two are linked and associated with each other, and (3) confounding due to population stratification. Summary of studies The studies presented here are examples of different ways to test for geneenvironment interaction in the context of association studies. These studies tested two major questions: 1) Is there an association between environmental factors, candidate genes and birth defects?; and 2) are there interactions between these genes and the observable environmental factors (e.g. maternal smoking or alcohol use)? Additionally, the three original studies presented in this dissertation exemplify the difficulty in identifying risk factors for birth defects. Research Article 1, Testing for Gene-Environment Interaction and Oral Clefts using Two Control Groups, combined case groups from two studies to form larger case groups and compared them to two separate control groups, one with non-cleft birth defects and the other a group of non-malformed infants. Combining the case groups from two sources (a passive state-wide registry and a prospective study) allowed for more power to assess the impact of the genotype, cigarette smoking and their interaction.