Neural tube defects (NTDs) have complex etiologies that include both genetic and environmental factors. Some environmental factors such as folic acid intake have been well described in terms of their importance to NTD risk (Berry et al., 1999). However, potential genetic contributions to risk have been more resistant to identification.
The paired-box (PAX) gene family encodes specific DNA-binding transcription factors that play important roles in embryonic development. Of the nine PAX genes, PAX 2, 3, 5, 6, 7, and 8 are expressed during the formation and organization of the neural tube along its entire axis (Hol et al., 1996; Helwig et al., 1995; Tremblay and Gruss, 1994; Balling et al., 1988). PAX genes contain a paired-box DNA binding domain, with or without an octapeptide and homeodomain (Dahl et al., 1997). The paired-box domain is a highly conserved DNA-binding domain of 128 amino acids located at the amino terminus of the protein. Analysis of mouse models and human syndromes has revealed the importance of PAX genes in their role as regulators of normal development (Mansouri et al., 1999; Pani et al., 2002). Heterozygous splotch mice display pigmentary abnormalities while homozygotes die during gestation with spina bifida or exencephaly (Vogan et al., 1993; Auerbach, 1954). Mutations within the human PAX3 gene have been associated with Waardenburg syndrome, a condition occasionally associated with NTDs (Baldwin et al., 1992; Hoth et al., 1993; Hol et al., 1995). In 1995, Hol and colleagues identified a 5-bp deletion in exon 5 of the PAX3 gene in a patient with spina bifida and mild manifestations of Waardenburg syndrome. A subsequent study identified two spina bifida patients who also had small interstitial chromosomal deletions involving PAX3 (Nye et al., 1998).
In the current study, we investigated a California population to determine whether polymorphisms in PAX3 were associated with risk of human spina bifida.
MATERIALS AND METHODS
The study design was a population-based case-control study. Cases and controls were ascertained by the California Birth Defects Monitoring Program, a population-based active surveillance system for collecting information on infants and fetuses with congenital malformations. Diagnostic and demographic information was collected by program staff from multiple sources of medical records of all liveborn and stillborn fetuses. Nearly all major structural anomalies diagnosed within 1 year of delivery were ascertained. Overall ascertainment for major malformations has been estimated as 97% complete (Schulman et al., 1993). Eligible for this study were liveborn infants only, because of the fact that the source of DNA was from newborn screening blood spots.
Included for study were 74 infants with spina bifida and 87 nonmalformed control infants born during the period 1983–1986 in selected counties in California. Among cases, 47.3% were non-Hispanic Whites, 41.9% were Hispanic Whites including U.S. born and foreign born, 4.0% were African American, and 6.8% were other ethnic groups. Among controls, 52.9% were non-Hispanic Whites, 34.5% were Hispanic Whites including U.S. born and foreign born, 4.6% were African American, and 8.0% were other ethnic groups. All newborn blood samples were obtained with approval from the State of California Health and Welfare Agency Committee for the Protection of Human Subjects. Genomic DNA was extracted from dried newborn screening blood spots on filter paper using Puregene DNA Purification kit (Gentra, Minneapolis, MN) according to the manufacturer's instructions and was amplified using PCR (Schwartz et al., 1990).
PAX3 Gene Resequencing
The genomic region of the PAX3 gene, including the paired-box domain, was resequenced with VariantSEQr resequencing system (Applied Biosystems, Foster City, CA). The primer set (RSS000013589-02) for PAX3 included nine pairs of primers and covered a 6-kb genomic region containing five exons and flanking intronic sequences. PCRs (30 s at 94°C, 45 s at 60°C, 60 s at 72°C for 40 cycles) were performed in a final volume of 10 μL containing 60 ng of genomic DNA, 2.0 μL VariantSEQr primer mix, 250 μM of each dNTP, in 2.0 mM MgCl2, 50 mM KCl, 20 mM Tris-HCl (pH 8.4), 8% glycerol, and 1.5 U of Taq DNA polymerase on a PE9700 thermalcycler. Then, PCR products were cleaned by digestion with ExoSAP-IT enzymes (USB Corporation, Cleveland, OH), and 3 μL of final products was applied to sequencing reaction (10 s at 96°C, 5 s at 50°C, 4 min at 60°C for 25 cycles) with BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems). The fragments were precipitated with isopropanol, denatured with 12 μL HiDi formamide solution at 95°C for 5 min, and loaded onto a 3730 DNA Analyzer. Sequencing results were analyzed with SeqScape v2.5 software (Applied Biosystems).
(CA) n Repeats Genotyping
Primers (forward: 5′-CCCAGATGCCTTCTTA and reverse: 5′-FAM-CAGGGAG-ATGGCAGTT-3′) were used to amplify fragments. The fragment was located ∼300 bp upstream of the transcription start site and contained the (CA) n dinucleotide repeats. PCR (35 s at 94°C, 30 s at 55.7°C, 60 s at 72°C for 35 cycles) was performed in a final volume of 25 μL containing 60 ng of genomic DNA, 6.25 pmol of each primer, 200 μM of each dNTP, in 1.5 mM MgCl2, 50 mM KCl, 20 mM Tris-HCl (pH 8.4), 1 μg/mL bovine serum albumin, and 0.4 U of Taq DNA polymerase. Data were collected by 3730 DNA Analyzer (ABI) and analyzed by GeneMapper v3.5 (ABI).
Deviations from Hardy-Weinberg equilibrium among control infants were evaluated using a χ2 test. Genotyping frequencies for detected variants within the PAX3 gene were compared between cases and controls and ORs and 95% CIs were estimated utilizing SAS software (version 9.1). Haplotype analysis was performed with Haploview software (version 3.2) (Barrett et al., 2005) and evaluated with χ2 tests.
Nineteen SNPs were observed by resequencing in the upstream genomic region and exons 1–4 of the PAX3 gene containing the paired-box domain (Table 1, Fig. 1). Two SNPs in coding regions did not result in an amino acid substitution. The flanking sequences and SNPs are shown in Table 2. Fifteen novel SNPs observed in controls had been submitted to the National Center for Biotechnology Information (NCBI) database with allele frequencies. These were: rs28945085, rs16863657, rs28945086, rs28945087, rs28945088, rs28945668, rs28945089, rs28945090, rs28945091, rs12623857, rs28945092, rs28945093, rs28945094, rs28945095, rs28945096.
Table 1. Primer Sequences for Resequencing PAX3 Gene
Sequences (5′ → 3′)
Fragment length (bp)
The 5′ end was flanked with 18 bp of universal M13 sequence.
Table 2. SNPs Observed in Genomic Region of PAX3 Gene Containing Paired-Box Domain
Nucleotide +1 is the A of the ATG-translation initiation codon; the nucleotide 5′ to +1 is numbered −1.
N/A: not applicable.
Deviations from Hardy-Weinberg equilibrium were evaluated in controls for each SNP. Statistical evidence (P < .05) for deviations was observed for rs28945089 and rs28945094. Stratified by maternal race/ethnicity, deviations of those two SNPs remained significant in non-Hispanic Whites.
Analyses focused on (CA) n repeats, single SNPs, and haplotypes. We did not observe significant associations between (CA) n repeats polymorphism and risk of spina bifida (data not shown). For SNP analyses, Table 3 shows allele frequencies by case and control status and by maternal race/ethnic background. Overall, substantial differences between cases and controls in allele frequencies were not observed. One exception, however, was rs16863657. The major allele (T) for this SNP in Hispanic White case infants was observed substantially less frequently than in control infants.
The risk associated with rs16863657 for spina bifida was further explored (Table 4). The OR associated with rs16863657 was 3.5 (95% CI: 1.2–10.0), indicating that SNP, rs16863657, was associated with a substantial increased risk of spina bifida in Hispanic Whites.
Table 4. PAX3_T-1186C (rs28945085) Genotype in Hispanic White Population
Haplotype analyses were performed for various race/ethnic groups. Because all observed SNPs were in linkage disequilibrium, these SNPs were defined as whole block for these analyses. Results of these analyses are shown in Table 5. For African Americans, five haplotypes were investigated, including CCGC, GCGC, CTGC, CTCC, and CTGG, which were represented by four SNPs, rs28945092 (C/G), rs12623857 (T/C), PAX3_G-396C, and rs28945085 (C/G). For Hispanic Whites, seven haplotypes were investigated: TCTCCGGTC, TCTCCGCCC, TCTCCGCTC, TCTTCGCTG, TCTCAGCCC, TCTTCGCTC, and CCTCCGCTC, corresponding to rs28945096 (T/C), rs28945095 (C/T), rs28945094 (T/C), rs12623857 (T/C), rs28945091 (C/A), rs28945090 (G/A), rs28945089 (C/G), rs16863657 (T/C), and rs28945085 (C/G). For non-Hispanic Whites, five haplotypes were investigated: TTCACGCCCTTC, TTCACGCCCTCC, TTCATGCCCTTG, TTCATGCCCTTC, and TCCACGCCCTTC, corresponding to rs28945096 (T/C), rs28945094 (T/C), PAX3_C701A, rs28945093 (A/C), rs12623857 (T/C), PAX3_G-569C, rs28945668 (C/T), rs28945088 (C/G), rs28945087 (C/T), rs28945086 (T/C), rs16863657 (T/C), and rs28945085 (C/G). For the other race/ethnic groups, six haplotypes were investigated: TCGT, TCCT, TCCC, CCGT, TTGC, and TTCC, corresponding to rs28945094 (T/C), rs12623857 (T/C), rs28945089 (C/G), and rs16863657 (T/C). For these various comparisons, only one haplotype appeared to be more frequent among infants with spina bifida than control infants. That is, the haplotype frequency of TCTCCGCCC in Hispanic Whites was 0.35 and 0.15 in cases and controls respectively, including the SNP, rs16863657. Because of the numbers of analytic comparisons made in this study, we cannot exclude the possibility that this latter result is also consistent with expected random variation.
African Americans: rs28945092 (C/G), rs12623857 (T/C), PAX3_G-396C, and rs28945085 (C/G)
Others: rs28945094 (T/C), rs12623857 (T/C), rs28945089 (C/G), and rs16863657 (T/C)
PAX3 belongs to the family of paired domain proteins that bind DNA and regulate gene expression. Previous studies in mouse and chick embryos demonstrate that pax3 plays an important role in neural tube development. Expression of PAX3 is detected in gestation day (GD) 8.5 embryos, with peaks of expression from day 9 to 12 followed by declining levels until no expression is evident in GD 17 embryos. The time window of pax3 high expression is coincident with the critical time for neural tube closure. Homozygotes (splotch mice) have 100% penetrance of NTDs. Our study focused on the conserved, paired-box domain of the PAX3 gene. Nineteen SNPs were found and 15 novel ones were identified and submitted to the NCBI database. Haplotypes including these SNPs were investigated in a study population of California infants. Overall, our analyses indicated that PAX3 SNPs were not strong risk factors for human spina bifida. Despite the fact that our study is strengthened by its population-based design, our lack of observing a modest role for PAX3 SNPs in spina bifida risk may be explicable to limited sample sizes and the exclusion of spina bifida-affected pregnancies that were terminated. Nevertheless, additional follow-up of the PAX3 gene variant T-1186C (rs16863657) and its related haplotype, TCTCCGCCC, may be important in other populations.
The authors are indebted to Dr. George Cunningham, Dr. Fred Lorey, and Terry Kennedy for making it possible to access newborn blood specimens. We also appreciate the technical support of Ms. Consuelo Valdes, Ms. Dia R. Gentile, Ms. Sarah Seth, and Mr. James Ebot Enaw.