BMP4 Was Associated with NSCL/P in an Asian Population

Background The Bone Morphogenetic Protein 4 gene (BMP4) is located in chromosome 14q22-q23 which has shown evidence of linkage for isolated nonsyndromic cleft lip with or without cleft palate (NSCL/P) in a genome wide linkage analysis of human multiplex families. BMP4 has been shown to play crucial roles in lip and palatal development in animal models. Several candidate gene association analyses also supported its potential risk for NSCL/P, however, results across these association studies have been inconsistent. The aim of the current study was to test for possible association between markers in and around the BMP4 gene and NSCL/P in Asian and Maryland trios. Methodology/Principal Findings Family Based Association Test was used to test for deviation from Mendelian assortment for 12 SNPs in and around BMP4. Nominal significant evidence of linkage and association was seen for three SNPs (rs10130587, rs2738265 and rs2761887) in 221 Asian trios and for one SNP (rs762642) in 76 Maryland trios. Statistical significance still held for rs10130587 after Bonferroni correction (corrected p = 0.019) among the Asian group. Estimated odds ratio for carrying the apparent high risk allele at this SNP was 1.61 (95%CI = 1.20, 2.18). Conclusions Our results provided further evidence of association between BMP4 and NSCL/P.


Introduction
Cleft lip with or without cleft palate (CL/P) is the most common human craniofacial malformation with a prevalence in newborns between 1/500-1/2000 worldwide [1,2], accounting for approximately one third of all congenital abnormalities [3]. In etiologic studies, CL/P is usually classified as syndromic or nonsyndromic based on the concurrent existence of other congenital malformations or developmental abnormalities. Almost 70% of the children born with a CL/P phenotype are nonsyndromic cases [4][5][6][7]. As the lip and primary palate have distinct developmental origins from the secondary palate, clefts of these areas can be further subdivided into cleft lip with or without cleft palate and cleft palate alone to maximize homogeneity of cleft phenotype [8]. Although a large number of candidate gene studies have been conducted in this field, and several genome-wide association studies (GWAS) [9][10][11][12] have recently provided more clues, specific causal variants and biological mechanisms responsible for occurrence of nonsyndromic cleft lip with or without cleft palate (NSCL/P) remain unclear.
The bone morphogenetic protein 4 gene (BMP4) on chromosome 14q22-q23 is one of the promising candidate genes for NSCL/P. Animal models show it is predominately expressed in epithelia of the maxillary and mandibular primordia [13,14], as well as the mesenchymal tissue around the frontonasal process and lateral maxillary prominence in the C578L/6J mice [15]. When the exogenous antagonist Noggin for BMP2 and BMP4 was administered to the region of beak fusion, visible clefts of the upper beak were shown in 10 of 13 chick embryos [14]. Further, bilateral fusion delay of the lip was reported in all nine mice with BMP4 conditional null allele at 12 embryonic days [16]. In humans, the chromosomal region around 14q22-q23 gave evidence of linkage (significant at a genome-wide level) to a gene controlling NSCL/P in multiplex NSCL/P families recruited from various populations [17,18]. However, results from published association studies have been inconsistent [9][10][11][12][19][20][21][22] suggesting a need for further research.
Here we tested for linkage and association between markers in and around BMP4 gene and NSCL/P using 297 case parent trios from Maryland and three Asian populations to further clarify the potential role BMP4 may play in the etiology of this common and complex disorder. Our results provided further evidence of association between BMP4 and NSCL/P.

Ethics statement
Study protocols were reviewed and approved by the institutional review boards in all participating institutions. All adult participants (including adult probands and probands' parents) and parents or guardians of pediatric participants provided written informed consent.

Sample description
A total of 297 NSCL/P probands and their parents were recruited from four cleft lip and palate treatment centers:

SNP selection, DNA, and genotyping
A total of 13 SNPs (Table 2  SNPs with ''SNP scores'' .0.6 (an assessment of design quality of the Illumina assay based on a proprietary algorithm), high validation levels in dbSNP (including validation on multiple platforms), and high heterozygosity levels were given priority when selecting SNPs. All 3 SNPs released by the HapMap project (HapMap Data Rel 24/phaseII Nov08, on NCBI B36 assembly, dbSNP b126) were included in our marker panel. The genomic DNA was extracted from peripheral blood using the protein precipitation method and stored at 220uC. Aliquots (4 mg) of each genomic DNA sample were dispensed into a bar-coded 96-well microtiter plate at a concentration of 100 ng/ml and genotyped by Illumina's GoldenGate chemistry [23] at the Genetic Resources Core Facility (GRCF) of Johns Hopkins University. Two duplicates and four controls from the Centre d'Etude du Polymorphisme Humain (CEPH) collection were included on each plate to evaluate genotyping consistency within and between plates.

Preliminary analysis
Among the 13 SNPs genotyped in and around BMP4, rs2855527 was the only one with MAF,1% in Maryland (0.5%) and in Asian parents, and was excluded from further analysis. The remaining 12 SNPs were eligible because all were compatible with HWE (at p.0.01) and their genotyping call rates were .80% (83.6% to 100%).
The minor allele differed for three SNPs (rs17563, rs10130587 and rs2855530) between Asian and Maryland parents among these 12 SNPs. MAF for rs2761880 was low (3.9%) in Maryland parents compared to Asian parents (26.7%). MAFs for the other 8 SNPs were similar in Asian and Maryland parents, ranging from 42.3%-45.9% and 39.3%-46.1%, respectively (Table 2 and 3). The relative difference ratios for MAFs were less than 10% for these 8 SNPs (Table 3). LD patterns among these 12 SNPs were similar in Asian and Maryland parents (Fig S1). In Asian parents, the first 3

Association analysis
When deviation from strict Mendelian inheritance of individual SNPs was analyzed in Asian trios, three SNPs (rs10130587, rs2738265 and rs2761887) showed nominally significant evidence of linkage and association (p,0.05) with NSCL/P. Significance held for rs10130587 after strict Bonferroni correction (corrected p = 0.019). The overall empiric P value for rs10130587 by Haploview was 0.018. Estimated odds ratio (OR) for carrying the apparent high-risk allele (MAF = 48.7%) at rs10130587 was 1.61(95%CI = 1.20, 2.18). No significant evidence of linkage and association with NSCL/P was particularly evident for rs10130587 in any of the three Asian groups or in a specific cleft type (cleft lip and palate or cleft lip only) in combined Asian population after Bonferroni correction (Table S2).
Sliding window haplotype analysis of these 12 SNPs in and around BMP4 generated a total of 38 tests. All haplotypes including rs10130587 were significantly associated with NSCL/P in combined Asian group. The Bonferroni corrected P value for the most significant haplotype (rs2761880-rs10130587) was not significant (0.0626) (Table S1). No significant linkage and association was seen in haplotype analysis of SNPs in each LD block (data not shown).
When 76 Maryland trios were analyzed, SNP rs762642 in the second LD block gave nominally significant evidence of linkage and association (p = 0.0469), however, this result was not significant after Bonferroni correction (Table S1). In sliding window haplotype analysis, majority of the haplotypes including rs762642 was nominally significant, but the most significant haplotype (rs762643-rs762642-rs2761887) was no longer significant after Bonferroni correction (Table S1). No significant linkage and association was seen in haplotype analysis of SNPs in each LD block (data not shown).

Discussion
BMP4 is located in 14q22-q23, and has 4 exons and 3 introns among which exons 3 and 4 encode protein (http://genome.ucsc. edu/cgi-bin). The protein encoded by this gene is a member of the bone morphogenetic protein family, part of the transforming growth factor-beta superfamily [28]. Animal experiments and genome wide linkage analysis in multiplex NSCL/P families repeatedly suggested BMP4 play a potential role in risk for NSCL/P [13][14][15][16][17][18]. Several candidate gene association analyses [19][20][21][22] have also contributed evidence that BMP4 is an important candidate gene for NSCL/P, although there has been no consistency across studies.
SNP rs762642 is the most frequently reported SNP in literature on BMP4 and its influence on NSCL/P [9][10][11][12]19,22]. One caseparent trio study from Chile tested the haplotype of rs1957860-rs762642 and found significant evidence for linkage and association [19]. For individual SNP analysis, as with all previously published reports, no evidence was shown for rs762642 either in our Asian or Maryland trios (after Bonferroni correction). And we did not find evidence in our sliding window haplotype analysis or in haplotype analysis for SNPs in identified LD block that included rs762642 after Bonferroni correction.
SNP rs17563 lies in exon 4 of BMP4, but evidence from association analysis with this marker has not been consistent. Two case-control studies conducted mainly in Han Chinese population [20,21] and one resequencing study [29] using samples from six different populations showed evidence of association between rs17563 and NSCL/P. However, evidence of linkage and association was not seen for rs17563 in a case-parent trio study conducted in two Scandinavia populations using log-linear models [22]. Our negative results are consistent with Jugessur's analysis [22] although our sample size was quite limited.
Although no evidence of linkage and association was seen for rs762642 and rs17563 with NSCL/P, intriguing result was found for another SNP (rs10130587) in our Asian trios (which was significant after Bonferroni correction). SNP rs10130587 was recently identified by the 1000 genomes program (http://www. ncbi.nlm.nih.gov/projects, 2011.12.8) with frequencies for allele C as 48.4% and 38.5% in combined Asian (Han Chinese in Beijing, Han Chinese South, and Japanese individuals) and CEPH samples separately (Table S3) [30]. Because MAF for this intronic SNP was high (48.7%) in our Asian parents, this marker should be informative and may be in LD with an unknown causal variant. The significant linkage and association observed in Asian NSCL/P trios was not replicated in our smaller sample of Maryland trios which also had a slightly lower MAF (40.8%) compared to our Asian founders. Because rs10130587 was not genotyped in any previous candidate gene and GWAS studies [9][10][11][12][19][20][21][22], our result can be considered an independent analysis although 52.9% (157 trios) of our sample [31] went into one GWAS [10]. It is possible that association for NSCL/P could have been identified with this SNP were it included in marker panels in previous association studies, including those failing to detect any evidence of linkage and association between BMP4 and NSCL/P [9][10][11][12]17,22]. The arrays are always changing and it is likely that the version used for GWAS was optimized for Caucasians because most of the early HapMap sequencing was done with CEPH samples. Our results supported the published evidence for some association between BMP4 and NSCL/P, and builds upon both experimental animal and linkage studies in humans. Figure S1 Linkage disequilibrium as measured by r 2 in BMP4 among parents of NSCL/P probands from Asian and Maryland trios. White: r 2 = 0. Shades of gray: 0,r 2 ,1. Black: r 2 = 1. BMP4, Bone Morphogenetic Protein 4; NSCL/P, nonsyndromic cleft lip with or without cleft palate. (DOC)