LETTERS A genome­wide association study identifies susceptibility loci for nonsyndromic sagittal craniosynostosis near BMP2 and within BBS9 Cristina M Justice1,24, Garima Yagnik2,24, Yoonhee Kim1, Inga Peter3, Ethylin Wang Jabs3, Monica Erazo3, Xiaoqian Ye3, Edmond Ainehsazan3, Lisong Shi3, Michael L Cunningham4, Virginia Kimonis5, Tony Roscioli6, Steven A Wall7, Andrew O M Wilkie7,8, Joan Stoler9, Joan T Richtsmeier10, Yann Heuzé10, Pedro A Sanchez-Lara11, Michael F Buckley12, Charlotte M Druschel13, James L Mills14, Michele Caggana15, Paul A Romitti16, Denise M Kay15, Craig Senders17, Peter J Taub18, Ophir D Klein19–21, James Boggan22, Marike Zwienenberg-Lee22, Cyrill Naydenov23, Jinoh Kim2, Alexander F Wilson1 & Simeon A Boyadjiev2 Sagittal craniosynostosis is the most common form of sutures of dense fibrous tissue that accommodate the growing brain. craniosynostosis, affecting approximately one in 5,000 newborns. Bone is added at these sutures during growth, and the skull eventu­ We conducted, to our knowledge, the first genome-wide ally ossifies completely. Craniosynostosis, the premature closure of association study for nonsyndromic sagittal craniosynostosis one or more of the cranial vault sutures, is a common congenital (sNSC) using 130 non-Hispanic case-parent trios of European anomaly. Approximately 80% of craniosynostosis occurs as an iso­ ancestry (NHW). We found robust associations in a 120-kb region lated anomaly, called nonsyndromic craniosynostosis (NSC), with­ downstream of BMP2 flanked by rs1884302 (P = 1.13 × 10−14, out major associated malformations1. Rare mutations in the FGFR2, odds ratio (OR) = 4.58) and rs6140226 (P = 3.40 × 10−11, TWIST1, FREM1, LRIT3, EFNA4 and RUNX2 duplications have been OR = 0.24) and within a 167-kb region of BBS9 between reported in a minor fraction of individuals with NSC2–9. The remain­ rs10262453 (P = 1.61 × 10−10, OR = 0.19) and rs17724206 der (~20%) of craniosynostosis cases are syndromic, occurring with (P = 1.50 × 10−8, OR = 0.22). We replicated the associations to one or more additional major malformations caused by single­gene −31 © 2012 Nature America, Inc. All rights reserved. America, Inc. © 2012 Nature both loci (rs1884302, P = 4.39 × 10 and rs10262453, P = 3.50 mutations in one of at least eight genes (FGFR1, FGFR2, FGFR3, × 10−14) in an independent NHW population of 172 unrelated TWIST1, EFNB1, POR, MSX2 and RAB23), involving primarily the probands with sNSC and 548 controls. Both BMP2 and BBS9 are coronal sutures10–12. genes with roles in skeletal development that warrant functional The most common type of NSC is sNSC, which accounts for npg studies to further understand the etiology of sNSC. 40–58% of all cases13,14. The etiology of sNSC is not well understood; however, the published literature suggests that it is a multifactorial Skull development is a complex process that involves ongoing inter­ condition in which both genetic and environmental factors have a action between the bones of the skull and cranial soft tissues. The role, as indicated by a higher rate of concordance in monozygotic as cranial vault is comprised of intramembranous bones joined by compared to dizygotic twins (30% compared to 0%, respectively)15, 1Genometrics Section, Inherited Disease Research Branch, Division of Intramural Research, National Human Genome Research Institute, US National Institutes of Health (NIH), Baltimore, Maryland, USA. 2Section of Genetics, Department of Pediatrics, University of California Davis, Sacramento, California, USA. 3Department of Genetics and Genomic Sciences, Mount Sinai School of Medicine, New York, New York, USA. 4Department of Pediatrics, Division of Craniofacial Medicine, University of Washington and Seattle Children’s Research Institute, Seattle, Washington, USA. 5Department of Pediatrics, Division of Genetics, University of California Irvine, Irvine, California, USA. 6School of Women’s and Children’s Health, Sydney Children’s Hospital, University of New South Wales, Sydney, New South Wales, Australia. 7Craniofacial Unit, Oxford University Hospitals National Health Service Trust, John Radcliffe Hospital, Oxford, UK. 8Weatherall Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Oxford, UK. 9Division of Genetics, Children’s Hospital Boston, Harvard University, Boston, Massachusetts, USA. 10Department of Anthropology, Pennsylvania State University, University Park, Pennsylvania, USA. 11Department of Pathology and Pediatrics, Children’s Hospital Los Angeles, University of Southern California, Los Angeles, California, USA. 12Department of Haematology and Genetics, South Eastern Area Laboratory Services, Sydney, New South Wales, Australia. 13Congenital Malformations Registry, New York State Department of Health, Albany, New York, USA. 14Division of Epidemiology, Statistics and Prevention Research, Eunice Kennedy Shriver National Institute of Child Health and Human Development, NIH, Department of Health and Human Services, Bethesda, Maryland, USA. 15Division of Genetics, Wadsworth Center, New York State Department of Health, Albany, New York, USA. 16Department of Epidemiology, College of Public Health, University of Iowa, Iowa City, Iowa, USA. 17Department of Otolaryngology, University of California Davis, Sacramento, California, USA. 18Division of Plastic and Reconstructive Surgery, Kravis Children’s Hospital, Mount Sinai Medical Center, New York, New York, USA. 19Department of Orofacial Sciences, University of California San Francisco, San Francisco, California, USA. 20Department of Pediatrics, University of California San Francisco, San Francisco, California, USA. 21Program in Craniofacial and Mesenchymal Biology, University of California San Francisco, San Francisco, California, USA. 22Department of Neurological Surgery, University of California Davis, Sacramento, California, USA. 23Department of Chemistry and Biochemistry, Medical University, Sofia, Bulgaria. 24These authors contributed equally to this work. Correspondence should be addressed to S.A.B. ([email protected]). Received 3 May; accepted 4 October; published online 18 November 2012; doi:10.1038/ng.2463 1360 VOLUME 44 | NUMBER 12 | DECEMBER 2012 NATURE GENETICS LETTERS 1,16 an increased male­to­female ratio (3:1) 14 and a high risk of recurrence in affected families17. Several environmental factors 12 have also been associated with sNSC, includ­ ing parity, prematurity, intrauterine con­ straint and maternal tobacco or nitrosatable 10 drug use18–21. In an attempt to identify susceptibility 8 P loci for sNSC, we recruited 201 case­ 10 parent trios and 13 nuclear families with –log 6 at least one individual with sNSC (68% NHW and 32% mixed­ethnicity families). These participants were enrolled and evalu­ 4 ated through the collaborative effort of the International Craniosynostosis Consortium 2 (https://genetics.ucdmc.ucdavis.edu/icc. cfm). Samples were genotyped on the 0 Illumina 1M Human Omni1­Quad array. 1 2 3 4 5 6 7 8 9 The final criteria for the current analyses 10 11 12 13 14 15 16 17 18 19 20 21 22 were based on strict sNSC phenotype (no Chromosome additional synostoses, congenital anomalies Figure 1 Manhattan plot of the P values obtained from the genome-wide TDT of 130 trios or developmental delay in the proband or (N = 914,402). The x axis corresponds to the genomic position of the autosomes, and the y axis any affected sibling) and NHW ethnicity as shows the −log10 of the P value. The horizontal dashed line corresponds to the genome-wide significance threshold of 5 × 10−8. determined by self­reported ancestry and P ≤ confirmed by principal component analy­ sis using EIGENSOFT 3.0 (ref. 22). Application of these criteria additional pseudoautosomal marker reaching a low, but not genome­ reduced our discovery population to 130 case­parent trios. wide statistically significant, P value (rs2522623, P = 5.75 × 10−5) and After stringent quality­control procedures, we retained genotypes thus did not meet our criteria for replication. This marker is located for 915,307 SNPs (914,402 autosomal and 905 pseudoautosomal), in PCDH11X (encoding protocadherin 11 X­linked), which has not which we analyzed using the transmission disequilibrium test (TDT) previously been reported to have a functional role in sNSC. Analysis of as implemented in PLINK v1.07 (ref. 23). This resulted in the iden­ a larger cohort may show a stronger association to this locus. tification of 21 SNPs, 18 on 20p12.3 and 3 on 7p14.3, that reached From a total of 46 autosomal SNPs that showed significant and the genome­wide significance threshold of P < 5 × 10−8 (Fig. 1 and suggestive associations with P < 1 × 10−5, we prioritized 25 SNPs Supplementary Fig. 1). The 18 SNPs on chromosome 20 were in two for the replication study (Supplementary Table 2). To confirm linkage disequilibrium (LD) blocks (92 kb and 19 kb on chromosome the observed signals, we genotyped a NHW replication popula­ 20p12.3; Fig. 2a, Table 1 and Supplementary Fig. 2), with the most tion of 172 unrelated cases with sNSC and 548 unaffected controls. © 2012 Nature America, Inc. All rights reserved. America, Inc. © 2012 Nature significant SNP, rs1884302 (P = 1.13 × 10−14), located approximately We considered our data as a replication when the direction of 345 kb downstream of BMP2. Conditional analysis using extended effect of the alleles surpassing nominal significance was consistent TDT in UNPHASED24 confirmed that rs1884302 led these signals between the discovery and replication datasets. All seven genotyped npg