Background: Orofacial clefts are etiologically heterogeneous malformations. One probable cause is maternal smoking during pregnancy. The effect of maternal smoking may be modified by genes involved in biotransformation of toxic compounds derived from tobacco. We investigated whether polymorphic variants of fetal acetyl-N-transferases 1 (NAT1) and 2 (NAT2) interact with maternal cigarette smoking during early pregnancy to increase the risk of delivering an infant with an orofacial cleft.
Methods: In a California population-based case-control study, we genotyped 421 infants born with an isolated cleft and 299 nonmalformed controls for 2 NAT1 and 3 NAT2 single nucleotide polymorphisms
Results: Although smoking was independently associated with increased risks for both isolated cleft lip ± cleft palate and isolated cleft palate, no independent associations were found for NAT1 1088 or 1095 genotypes or for NAT2 acetylator status. However, the infant NAT1 1088 and 1095 polymorphisms were strongly associated with the risk of clefts among smoking mothers; infants with NAT1 1088 genotype AA versus TT (odds ratio [OR] = 3.9; 95% confidence interval =1.1–17.2) and with NAT1 1095 genotype AA versus CC (OR = 4.2; 1.2–18.0). Infant NAT2 acetylator status did not appreciably affect susceptibility of the fetus to the teratogenic effects of maternal smoking.
Conclusions: Our results suggest that maternal smoking during pregnancy may increase risk for orofacial clefts particularly among smokers whose fetuses have polymorphic variants of NAT1, an enzyme involved in phase II detoxification of tobacco smoke constituents.
From the *Children's Hospital and Research Center, Oakland, California; the †California Birth Defects Monitoring Program, Berkeley, California; the ‡Center for Human Molecular Genetics, University of Nebraska Medical Center, Omaha, Nebraska; and the §Institute of Biotechnology, Texas A&M University, Houston, Texas.
Editors’ note: A commentary on this article appears on page 131. Supplemental material for this article is available with the online version of the journal at www.epidem.com
Supplemental material for this article is available with the online version of the journal at www.epidem.com
Submitted 9 August 2002; final version accepted 17 October 2003.
Funded by the Cigarette and Tobacco Surtax Fund of the California Tobacco-Related Diseases Research Program, University of California, grants 6RT-0360, 1RT466, and 3RT0413. National Institutes of Health, grants DE13613, DE12898, and HD39084; and the Centers for Disease Control and Prevention, Centers of Excellence Award U50/CCU913241.
Correspondence: Edward J. Lammer, Children's Hospital Research Institute, 5700 Martin Luther King, Jr. Way, Oakland, CA 94609. E-mail: email@example.com.
Click on the links below to access all the ArticlePlus for this article.
Please note that ArticlePlus files may launch a viewer application outside of your web browser.
Maternal smoking has been linked to a variety of adverse pregnancy outcomes, including preterm delivery, decreased birth weight, and certain malformations. Oral cleft defects have been consistently associated with smoking during pregnancy, with most studies observing relatively modest risks of 1.5- to 2-fold.1–6 A dose-response is found in some studies.2 Two studies have reported substantially increased risks for oral clefts among infants who have both an uncommon variant of a secretory protein, transforming growth factor-alpha (TGF-α), and a mother who smoked during early pregnancy.2,7 These findings represented the first demonstrations of a gene–environment interaction associated with a common malformation. Recently, van Rooij and colleagues8 reported a suggestive interaction between maternal smoking, and absence of a detoxifying enzyme, with risk of a cleft. Little is known, however, about genetic susceptibility to toxins in cigarette smoke.
Among the many constituents of tobacco smoke, aromatic amines are potentially teratogenic because the initial activation stage of their biotransformation produces highly reactive intermediate compounds that can damage DNA and proteins by creating covalently bound adducts. These unstable intermediates are detoxified by phase II xenobiotic metabolizing enzymes, including the acetyl-N-transferases (NATs).9–12 NATs transfer an acetyl group onto an amine, hydrazine, or hydroxylamine moiety of an aromatic compound.13–16 Humans have only 2 functional NAT isoenzymes, NAT1 and NAT2, each exhibiting overlapping, but well-defined, substrate specificity.15 Substrates of NAT1 include p-aminobenzoate and p-aminosalicylate, whereas NAT2-specific substrates include isoniazid and sulfamethazine.14 A variety of aromatic amines that derive from cigarette smoking must be biotransformed by enzymes expressed by the mother, fetus, and placenta. Because reduced ability to acetylate aromatic amines has been associated with toxicity in adults, it is reasonable to assume that this capacity may also be associated with likelihood of developmental toxicity.13,15–17 Human N-acetyltransferase activity has been detected in both fetal and placental compartments.17–19 Although NAT2 activity has also been detected in early human embryos, it appears that NAT1 contributes most of the N-acetylation in the first trimester.18 Thus, capacity for fetal biotransformation of aromatic amines may contribute to their potential developmental toxicity. For example, decreased NAT1 activity might expose a fetus to increased or prolonged concentrations of an activated intermediate compound because the detoxifying N-acetylation step is impaired. We analyzed data from a population-based case-control study to test the hypothesis that polymorphic variants of NAT1 or NAT2 might affect an infant's susceptibility to the teratogenic effects of maternal smoking during pregnancy.
Details of the population-based case-control study used for these analyses have been described elsewhere.2 Case infants (diagnosed within 1 year of birth) or fetuses with an orofacial cleft (cleft lip ± palate; or cleft palate) were ascertained by reviewing medical records at all hospitals and genetic centers in a known geographic population base. This base comprised a cohort of 552,601 births and fetal deaths occurring in 1987–1989 to women residing in most counties in California (metropolitan areas of Los Angeles and San Francisco were excluded). Diagnostic information from medical records, autopsies, and surgical reports of all infants and fetuses with orofacial clefts or similar orofacial anomalies were reviewed to restrict eligibility to those infants with cleft lip ± palate or cleft palate. Infants with diagnoses of bifid uvula, submucous cleft palate, notching of the alveolar ridge or vermillion border of the upper lip were excluded because ascertainment was thought to be incomplete. Infants cytogenetically diagnosed with any trisomy or Turner syndrome (45,X) were excluded (n = 81). There were 892 infants and fetuses ascertained with an eligible diagnosis. Only live births were included in the present analysis.
The research protocols for this study were approved by the California State Health Department and Children's Hospital Oakland Institutional Review Boards.
A medical geneticist (EJL) classified each case as isolated or multiple based on the nature of any accompanying congenital anomalies. Cleft palate and cleft lip ± palate cases with no other anomaly or with anomalies considered minor (eg, low-set ears) or not true malformations (eg, undescended testicles) were classified as isolated. Only cases with isolated clefts were eligible for these analyses.
A control infant was potentially eligible if: (1) the infant was born alive during 1987–1989; (2) the mother was a resident of one of the counties in which cases were ascertained; and (3) no reportable birth defect was diagnosed before the first birthday. Other than being delivered during a similar time period and within the same geographic area, controls were not matched to cases. A total of 972 controls were electronically selected from California vital records using a pseudorandom number generator among all eligible infants (n = 548,844).
We interviewed mothers of cases and controls in English (91%) or Spanish, nearly all by telephone. Women who only spoke other languages (26 cases and 33 controls) and 3 case mothers who died before interview contact were excluded, yielding 863 eligible cases and 939 eligible controls. Interviews were completed an average of 3.5 years after the date of delivery for cases and 3.6 years for controls. At the beginning of the interview, an interviewer assisted each woman with establishing a 4-month time period (from 1 month before to 3 months after conception) that was referred to throughout the interview to elicit information on exposures and events. This 4-month period encompasses the embryologic period of palate and lip formation, including closure, which is complete by approximately 60 days postconception. Women were asked how many cigarettes they smoked daily for the 4-month period as well as for each month during the period. To assess passive smoke exposures in the 4-month period, women were asked whether anyone smoked inside their homes (including specific questions about paternal smoking), or near their work or school or while commuting to work or school, and whether they regularly frequented (at least weekly) a place such as a restaurant or laundromat where others smoked nearby.
Genomic DNA was obtained from residual dried blood spots on newborn screening specimens (Guthrie filter papers) collected from all liveborn children in California and stored by the Genetic Diseases Branch, Health and Human Services. Genomic DNA was obtained for 83% of eligible cases and 87% of eligible controls. Reasons for the inability to identify a specimen included: (1) no sample remaining on the filter paper, (2) filter paper could not be located, and (3) insufficient information on the child to adequately match records. To minimize the number of samples to be genotyped, the 652 control samples were randomly reduced to 299. DNA was extracted from the residual newborn screening filter papers using a modification of the salting-out method and suspended in 20 μL of TE buffer.20,21 Genotyping was performed blinded to case/control status. Genotyping methods for NAT1 and NAT2 are described in the Appendix, available with the electronic version of this article at www.epidem.com.
Odds ratios (ORs) and their 95% confidence intervals (CIs) were used to estimate risks.
The NAT1 Polymorphisms
Table 1 describes selected maternal and infant characteristics of the cases and controls. Among the 437 isolated cleft cases (309 cleft lip ± palate and 128 cleft palate), we successfully genotyped 98% for the NAT1 1088 polymorphism and 91% for the NAT1 1095 polymorphism. Allele frequencies for the controls were 71% for T1088 and 29% for 1088A, and 67% for C1095 and 33% for 1095A. The observed genotype proportions for controls were consistent with Hardy-Weinberg expectations (χ2 = 0.57 for NAT1 1088 and χ2 = 0.07 for NAT1 1095).
The genotypic frequencies of the NAT1 polymorphisms are shown in Table 2. Among controls, 9% were homozygous for the 1088A variant and 11% homozygous for the 1095A variant. Homozygosity for NAT1 1088A was more frequent among white Hispanic infants (14%) and Asian or other infants (13%) than white non-Hispanic infants (5%). There was no substantial ethnic difference in frequencies of homozygosity for NAT1 1095. Compared with genotype 1088 TT, the OR for isolated cleft lip ± cleft palate among infants homozygous for 1088A or heterozygous was only modestly elevated (Table 2). Odds ratios for isolated cleft palate were not elevated for these same comparisons (Table 2). Compared with genotype 1095 CC, the OR for isolated cleft lip ± cleft palate was increased (OR = 1.7; CI = 0.97–2.9) among infants homozygous for 1095A, but not for heterozygotes. Again, similar comparisons for isolated cleft palate did not show elevated risks.
Twenty-five percent of control mothers smoked during the periconceptional period compared with 35% of case mothers. The odds ratio with maternal smoking was 1.6 for isolated cleft lip ± cleft palate (CI = 1.1 to 2.4) and also for isolated cleft (CI = 1.1 to 2.6).
Tables 3 and 4 show combined results for the infant NAT1 genotypes and maternal smoking. If risks from maternal smoking are modified by genotype, we anticipate that such effects would be strongest in the comparison of homozygous genotypes. Table 3 shows results consistent with this expectation, with a nearly 4-fold increased risk for isolated cleft lip ± palate for the AA genotype in the presence of maternal smoking. We explored this observation further by dividing maternal smoking into 2 categories (Table 3), but the point estimates are imprecise for the greater smokers. For cleft palate alone, 24 mothers had infants who were homozygous for the 1088A variant, but none of these mothers smoked, and so we observed no modification of effect by genotype (data not shown). When we combined heterozygous infants (genotype 1088 TA) with infants homozygous for T1088, the results did not substantially change (data not shown).
Table 4 shows results of the combined influence of NAT1 1095 genotypes and maternal smoking. Similar to the 1088 polymorphism analysis, the AA genotype of the NAT1 1095 polymorphism modified risks for isolated cleft lip ± palate, associated with smoking, with a 4-fold increased risk. There were too few isolated cleft palate cases to adequately assess modification of effect between smoking and genotype (data not shown). Grouping isolated cleft palate heterozygous infants (genotype NAT1 1095 CA) with infants homozygous for C1095, results did not substantially differ from those shown in Table 4 (data not shown).
The risk with the NAT1 genotype 1088 AA + 1095 AA (vs. 1088 TT + 1095 CC) in the presence of maternal smoking was higher than either separately, although the confidence interval was broad (OR = 5.3; CI = 1.0 to 52). No infant with isolated cleft palate whose mother smoked had the 1088 AA + 1095 AA genotype.
We also considered passive smoking exposures during early pregnancy. These analyses were limited to mothers who did not actively smoke during the periconceptional period. Risks for isolated cleft lip ± cleft palate among infants with the NAT1 1088 AA genotype and any passive maternal smoke exposure were not substantially increased (OR = 1.9; CI = 0.75 to 4.9) compared with 1088 TT genotype and no exposure. The same was true for infants with the NAT1 1095 AA genotype (OR = 2.2; CI = 0.92 to 5.5) compared with the 1095 CC genotype and no exposure.
The NAT2 Polymorphisms
We genotyped eligible cases and controls for 3 single nucleotide polymorphisms of NAT2: 857G → A, 481C → T, and 590G → A. We could categorize 96% of the cases and controls into acetylator phenotypes, as described by Ambrosone et al.22; 52% of controls were defined as slow acetylators and 48% rapid acetylators. The distribution of acetylator phenotypes was nearly identical for each ethnic stratum. For isolated cleft lip ± cleft palate cases, 46% were slow acetylators, yielding an OR of 1.3 (CI = 0.90 to 1.8). Similarly, the percent of isolated cleft palate cases categorized as slow acetylators was 45%, yielding an OR of 1.3 (CI = 0.85 to 2.1). Table 4 shows the distribution of acetylator phenotypes according to mothers’ smoking. The ORs are slightly higher among smokers than nonsmokers, but do not substantially differ with respect to acetylator status (Table 5).
We previously reported that maternal cigarette smoking during early pregnancy was associated with increased risks for both isolated cleft lip ± cleft palate and isolated cleft palate.2 These risks, however, were relatively modest (OR = 1.7), with some evidence of higher risk among heavier smokers. The objective of this study was to determine if variants of 2 NAT enzymes known to be involved in biotransformation of tobacco toxins altered the risks with maternal smoking. Neither NAT1 1088 nor 1095 polymorphisms, nor NAT2 acetylator status, was an independent risk factor for clefts. We did, however, observe modifying influences of the NAT1 polymorphisms on clefting risks associated with smoking. Infant homozygosity for uncommon variants of the 1088 and 1095 polymorphisms was associated with 4-fold increased risks for isolated cleft lip ± cleft palate if mothers smoked during early pregnancy. These risks appear to be higher than the 1.6-fold risk associated with smoking, independent of fetal genotypes. No similar risks for isolated cleft palate were identified.
We found no evidence that NAT1 polymorphisms conferred susceptibility to effects of passive smoking. NAT2 acetylator status was not a risk factor for either type of isolated cleft, and we found no interaction between maternal smoking and NAT2 acetylator status for risk of either facial cleft.
Strengths of our study include its relatively large size, population-based design, and systematic collection of data for maternal and passive smoking exposures. We successfully genotyped a high percentage of participating cases and controls. It is difficult to generate these kinds of analyses because the frequency of homozygosity for NAT1 variants is approximately 10% and the percentage of women smoking during pregnancy is 20% to 25%. Statistical power is a challenge and necessitates a large study population. We found little difference in allele frequencies for different ethnic strata, except for NAT1 1088. Our genotype frequencies were similar to those reported from other control groups from the United States.22,25–28 and Europe.14,29–30 In addition, our allele distributions for NAT1 and NAT2 polymorphisms did not differ from Hardy-Weinberg expectations. These observations suggest that our study population was likely representative of the California birth population. A large number of analyses were performed in this study, and thus some of the observed risks could be attributed to random variation.
A limitation of our study is that maternal genotypes were not measured, because we did not have access to this material. Maternal metabolism of toxins in tobacco smoke clearly influences fetal exposures and, without the maternal genotypic information, we have an incomplete picture of all of the genetic components of detoxification in utero. It is important that future studies include assessments of both fetal and maternal metabolism, when possible.
What is the biologic plausibility that variant forms of NAT1 enzyme might alter the biotransformation of tobacco-related toxins? NAT1 substrates probably differ somewhat from those metabolized by NAT2, and so it is not implausible to find a modifying influence with NAT1 variants but not NAT2 variants. Among the known substrates for NAT1 are tobacco carcinogens like arylhydroxamines.30 The biologic mechanism through which altered NAT1 biotransformation might cause a birth defect may be through increased formation of protein or DNA adducts. High levels of 4-aminobiphenyl-hemoglobin adducts, for example, have been measured in term infants born to smoking mothers, demonstrating the production of adducts arising from an aromatic amine that is probably biotransformed by NATs.32 If adducts formed with proteins involved in controlling development of facial structures, lowering their concentrations, this might lead to facial birth defects. NAT1 must be expressed in embryonic tissues for fetal genetic variants to alter risks from maternal smoking during pregnancy. The pattern of developmental expression of NATs has not been extensively investigated in mammals. However, human NAT1 expression has been detected in preimplantation blastocyst-stage embryos and in human placentas from as early as 5.5 gestational weeks.18 In mice, Nat expression at midgestation (day 10) has been shown in both embryonic and placental tissues.33,34
Major reasons for choosing these 2 NAT1 polymorphisms for study are their relatively high frequency and their previously identified associations with colon cancer.30,31 Most human research on NAT variants has examined relationships with various cancers, focusing mainly on genotypic analyses. Few of these studies assessed smoking or other exposures to NAT substrates. Interactions between NAT genotypes and smoking on cancer have all involved NAT2 acetylation variants.22,35,36 A recent review concluded that if NAT genes play a role in causing colon cancer, it is likely one involving modification of a relationship between environmental exposures and disease.30
In conclusion, we found evidence of fetal susceptibility to the teratogenic effects of maternal smoking during pregnancy among infants with NAT1 variants. This risk was for isolated cleft lip ± cleft palate, and not for isolated cleft palate. The magnitude of risk was modest. Further studies will need to confirm these observations and should include information on the contribution of maternal genotypes as well.
Alex Sabugal, Poulina Uddin, and Nam Do made valuable contributions to the laboratory genotyping, and Eric Neri, Wei Yang, and Cecile Laurent contributed programming efforts. We are indebted to George Cunningham and Fred Lorey for making it possible to access stored blood spot specimens.
1.Wyszynski DF, Beaty TH. Review of the role of potential teratogens in the origin of human nonsyndromic oral clefts. Teratology
2.Shaw GM, Wasserman CR, Lammer EJ, et al. Orofacial clefts, parental cigarette smoking, and transforming growth factor-alpha gene variants. Am J Hum Genet
3.Khoury MJ, Weinstein A, Panny S, et al. Maternal cigarette smoking and oral clefts: a population-based study. Am J Public Health
4.Khoury MJ, Gomez-Frias M, Mulinare J. Does maternal cigarette smoking during pregnancy cause cleft lip and palate in offspring? Am J Dis Child
5.Werler MM, Lammer EJ, Rosenberg L, et al. Maternal cigarette smoking in relation to oral clefts. Am J Epidemiol
6.Lieff S, Olshan AF, Werler M, et al. Maternal cigarette smoking during pregnancy and risk of oral clefts in newborns. Am J Epidemiol
7.Hwang SJ, Beaty TH, Panny SR, et al. Association study of transforming growth factor alpha Taq1
polymorphism and oral clefts: indication of gene–environment interaction in a population based sample of infants with birth defects. Am J Epidemiol
8.Van Rooij IALM, Wegerif MJM, Roelofs HMJ, et al. Smoking, genetic polymorphisms in biotransformation enzymes, and nonsyndromic oral clefting: a gene–environment interaction. Epidemiology
9.Badawi AF, Hirvonen A, Bell DA, et al. Role of aromatic amine acetyltransferases, NAT1
, in carcinogen-DNA adduct formation in the human urinary bladder. Cancer Res
10.Blum M, Demierre A, Grant DM, et al. Molecular mechanism of slow acetylation drugs and carcinogens in humans. Proc Natl Acad Sci U S A
11.Shirai T, Fysh FM, Lee M-S, et al. Relationship of metabolic activation of N
-acylarylamines to biological response in the liver and mammary gland of the female CD rat. Cancer Res
12.Vineia P, Caporaso N, Tannenbaum SR. Acetylation phenotype, carcinogen-hemoglobin adducts and cigarette smoking. Cancer Res
13.Hein DW. N
-acetyltransferase genetics and their role in predisposition to aromatic and heterocyclic amine-induced carcinogenesis. Toxicol Lett
14.Upton A, Johnson N, Sandy J, et al. Arylamine. N
-acetyltransferases: of mice, men and microorganisms. Trends Pharmcol Sci
15.Sim E, Payton M, Noble M, et al. An update on genetic, structural and functional studies of arylamine N-acetyltransferases in eucaryotes and procaryotes. Hum Mol Genet
16.King CM, Land SJ, Jones RF, et al. Role of acetyltransferases in the metabolism and carcinogenicity of aromatic amines. Mutat Res
17.Pacifici GM, Bencini C, Rane A. Acetyltransferase in humans: development and tissue distribution. Pharmacology
18.Smelt VA, Upton A, Adjaye J, et al. Expression of arylamine N-acetyltransferases in pre-term placentas and in human pre-implantation embryos. Hum Mol Genet
19.Smelt VA, Mardon HJ, Redman RWG, et al. Acetylation of arylamines by the placenta. Eur J Drug Metab Pharmacokinet
20.Iovannisci DM. Highly efficient recovery of DNA from dried blood using the MasterPure™ complete DNA and RNA purification kit. Epicentre Forum
21.Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res
22.Ambrosone CB, Freudenheim JL, Graham S, et al. Cigarette smoking, N
-acetyltransferase 2 genetic polymorphisms, and breast cancer risk. JAMA
23.Dietz AC, Doll MA, Hein DW. A restriction fragment length polymorphism assay that differentiates human N
) alleles. Anal Biochem
24.Rhodes RB, Lewis K, Shultz J, et al. Analysis of the factor V Leiden mutation using the READIT™ assay. Mol Diagn
25.Probst-Hensch NM, Haile RW, Li DS, et al. Lack of association between the polyadenylation polymorphism in the NAT1
(acetyltransferase 1) gene and colorectal adenomas. Carcinogenesis
26.Lin HJ, Han C-Y, Lin BK, et al. Slow acetylator mutations in the human polymorphic N
-acetyltransferase gene in 786 Asians, blacks, Hispanics, and whites: application to metabolic epidemiology. Am J Hum Genet
27.Millikan RC, Pittman GS, Newman B, et al. Cigarette smoking, N
-acetyltransferases 1 and 2, and breast cancer risk. Cancer Epidemiol Biomarkers Prev
28.Chen J, Stempfer MJ, Hough HL, et al. A prospective study of N
-acetyltransferase genotype, red meat intake, and risk of colorectal cancer. Cancer Res
29.Bunschotoen A, Tiemersma E, Schouls L, et al. Simultaneous determination of polymorphism of N
-acetyltransferases 1 and 2 genes by reverse line blot hybridization. Anal Biochem
30.Brockton N, Little J, Sharp L, et al. N
-Acetyltransferase polymorphisms and colorectal cancer: a HuGE review. Am J Epidemiol
31.Bell DA, Stephens EA, Castranio T. Polyadenylation polymorphism in the acetyltransferase 1 gene (NAT1
) increases risk of colorectal cancer. Cancer Res
32.Pinorini-Godly MT, Myers SR. HPLC and GC/MS determination of 4-aminobiphenyl hemoglobin adducts in fetuses exposed to the tobacco smoke carcinogen in utero
33.Mitchell MK, Futscher BW, McQueen CA. Developmental expression of N
-acetyltransferases in C57Bl/6 mice. Drug Metab Dispos
34.Stanley LA, Copp AJ, Pope J, et al. Immunochemical detection of arylamine N
-acetyltransferase during mouse embryonic development and in adult mouse brain. Teratology
35.Slattery ML, Potter JD, Samowitz W, et al. NAT2
, cigarette smoking, and risk of color cancer. Cancer Epidemiol Biomarkers Prev
36.Welfare MR, Cooper J, Bassendine MF, et al. Relationship between acetylator status, smoking, diet and colorectal cancer risk in the north-east of England. Carcinogenesis
Supplemental Digital Content
© 2004 Lippincott Williams & Wilkins, Inc.