You are here

Article Text

Article menu
Complex traits
A meta-analysis identifies adolescent idiopathic scoliosis association with LBX1 locus in multiple ethnic groups
  1. Douglas Londono1,
  2. Ikuyo Kou2,
  3. Todd A Johnson3,
  4. Swarkar Sharma4,5,
  5. Yoji Ogura2,6,
  6. Tatsuhiko Tsunoda3,
  7. Atsushi Takahashi7,
  8. Morio Matsumoto6,
  9. John A Herring8,9,
  10. Tsz-Ping Lam10,11,
  11. Xingyan Wang12,13,
  12. Elisa M S Tam10,11,
  13. You-Qiang Song14,
  14. Yan-Hui Fan14,
  15. Danny Chan14,
  16. Kathryn S E Cheah14,
  17. Xusheng Qiu15,
  18. Hua Jiang15,
  19. Dongsheng Huang16,
  20. Japanese Scoliosis Clinical Research Group, TSRHC IS Clinical Group, the International Consortium for Scoliosis Genetics,
  21. Peiqiang Su17,
  22. Pak Sham18,
  23. Kenneth M C Cheung19,
  24. Keith D K Luk19,
  25. Derek Gordon1,
  26. Yong Qiu15,
  27. Jack Cheng10,11,
  28. Nelson Tang12,13,
  29. Shiro Ikegawa2,
  30. Carol A Wise4,9,20,21
  1. 1Department of Genetics, Rutgers, The State University of New Jersey, Piscataway, USA
  2. 2Laboratory for Bone and Joint Diseases, Center for Integrative Medical Sciences, RIKEN, Tokyo, Japan
  3. 3Laboratory for Medical Science Mathematics, Center for Integrative Medical Sciences, RIKEN, Yokohama, Japan
  4. 4Seay Center for Musculoskeletal Research, Texas Scottish Rite Hospital for Children, Dallas, USA
  5. 5School of Biotechnology, SMVDU, Katra, India
  6. 6Department of Orthopaedic Surgery, School of Medicine, Keio University, Tokyo, Japan
  7. 7Laboratory for Statistical Analysis, Center for Integrative Medical Sciences, Riken, Yokohama, Japan
  8. 8Department of Orthopedic Surgery, Texas Scottish Rite Hospital for Children, Dallas, USA
  9. 9Department of Orthopaedics, University of Texas Southwestern Medical Center at Dallas, Dallas, USA
  10. 10Department of Orthopaedics and Traumatology, The Chinese University of Hong Kong, Hong Kong, China
  11. 11Joint Scoliosis Research Center, The Chinese University of Hong Kong and Nanjing University, Hong Kong, China
  12. 12Functional Genomics and Biostatistical Computing Laboratory, Shenzhen Research Institute of the Chinese University of Hong Kong, Hong Kong, China
  13. 13Department of Chemical Pathology, the Chinese University of Hong Kong, Shatin, Hong Kong SAR, China
  14. 14Department of Biochemistry, University of Hong Kong, Hong Kong, China
  15. 15Department of Spine Surgery, The Affiliated Drum Tower Hospital of Nanjing University Medical School, Nanjing, China
  16. 16Department of Orthopedics, The Memorial Hospital of Sun Yat-Sen University, Guangzhou, China
  17. 17Department of Spine Surgery, The First Affiliated Hospital of Sun Yat-Sen University, Guangzhou, China
  18. 18Department of Psychiatry, University of Hong Kong, Hong Kong, China
  19. 19Department of Orthopaedics and Traumatology, University of Hong Kong, Hong Kong, China
  20. 20Department of Pediatrics, University of Texas Southwestern Medical Center at Dallas, Dallas, USA
  21. 21McDermott Center for Human Growth and Development, University of Texas Southwestern Medical Center at Dallas, Dallas, USA
  1. Correspondence to Dr Carol Ann Wise, Sarah M and Charles E Seay, Seay Center for Musculoskeletal Research, Texas Scottish Rite Hospital for Children, Dallas, TX 75219, USA; carol.wise{at}tsrh.org Dr Shiro Ikegawa, Laboratory for Bone and Joint Diseases, Center for Integrative Medical Sciences, RIKEN, Tokyo, Japan; sikegawa{at}ims.u-tokyo.ac.jp

Abstract

Background Adolescent idiopathic scoliosis (AIS) is a common rotational deformity of the spine that presents in children worldwide, yet its etiology is poorly understood. Recent genome-wide association studies (GWAS) have identified a few candidate risk loci. One locus near the chromosome 10q24.31 LBX1 gene (OMIM #604255) was originally identified by a GWAS of Japanese subjects and replicated in additional Asian populations. To extend this result, and to create larger AIS cohorts for the purpose of large-scale meta-analyses in multiple ethnicities, we formed a collaborative group called the International Consortium for Scoliosis Genetics (ICSG).

Methods Here, we report the first ICSG study, a meta-analysis of the LBX1 locus in six Asian and three non-Asian cohorts.

Results We find significant evidence for association of this locus with AIS susceptibility in all nine cohorts. Results for seven cohorts containing both genders yielded P=1.22×10–43 for rs11190870, and P=2.94×10–48 for females in all nine cohorts. Comparing the regional haplotype structures for three populations, we refined the boundaries of association to a ∼25 kb block encompassing the LBX1 gene. The LBX1 protein, a homeobox transcription factor that is orthologous to the Drosophila ladybird late gene, is involved in proper migration of muscle precursor cells, specification of cardiac neural crest cells, and neuronal determination in developing neural tubes.

Conclusions Our results firmly establish the LBX1 region as the first major susceptibility locus for AIS in Asian and non-Hispanic white groups, and provide a platform for larger studies in additional ancestral groups.

  • meta-analysis
  • scoliosis
  • LBX1
  • 10q24.31

https://doi.org/10.1136/jmedgenet-2013-102067

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Idiopathic scoliosis (IS), a lateral rotational deformity of the spine, occurs in 2–3% of children, typically at the time of the adolescent growth spurt.1 Unlike other forms of scoliosis, children with adolescent IS (AIS) present without coexisting diagnoses and bear no obvious structural deficiencies in the spinal column itself. Progression to severe deformity can be rapid and is a major clinical concern. This is particularly relevant for girls, whose risk of progression is at least five times greater than for boys, for reasons that are not understood.2 Other progression risk factors are initial curve pattern and severity of curvature in relation to remaining growth. However, the aetiological relationships between developmental stage, patterning and disease are poorly understood.3 Severe IS warrants surgical correction to avoid increasing deformity and pulmonary compromise. An estimated three billion dollars are spent annually for hospital charges in the USA alone. Hence, treatment and risks to AIS patients pose a significant health burden worldwide.

    Familial clustering, twin studies and heritability estimates for IS are well documented and suggest significant genetic contributions.4–7 Two genome-wide association studies (GWAS) have been reported, a family based analysis of a mostly non-Hispanic white (NHW) cohort and a case-control analysis of a Japanese cohort.8 ,9 In the Japanese study, three single nucleotide polymorphism (SNPs) at a single locus in the proximity of the LBX1 gene on chromosome 10q24.31 surpassed genome-wide significance (p<5×10−8). SNP rs11190870 displayed strongest effects in discovery, replication and combined cohorts (OR=1.56, 95% CI 1.41 to 1.71). Subsequently, association was replicated in additional Asian cohorts.10–12 LBX1 encodes ladybird homeobox 1, orthologous to Drosophila ladybird-late, and is involved in proper migration of muscle precursor cells, specification of certain cardiac neural crest cells,13 ,14 and determination of dorsal spinal neurons and hindbrain somatosensory neurons of the developing neural tube.15–17 Of the three SNPs in the Japanese study, one was among the top 100 SNPs (of >370 000 genotyped) in the prior family based GWAS, suggesting the locus could be relevant in non-Asian cohorts.

    Despite high heritability and its burden to public health, genetic risk factors in AIS have remained understudied. To address this issue, and to particularly foster trans-ethnic population studies of AIS, we have formed the International Consortium for Scoliosis Genetics (ICSG). As a first question, we focused on the role of the chromosome 10q24.31 locus across ethnic groups. Our results refine the boundaries of the locus and establish it as a common, transethnic AIS risk factor in males and females. These data continue to support a likely role for LBX1 in pathogenesis of AIS, and raise questions about its role in later human axial development.

    Results

    The nine cohorts included in this study are summarised in table 1. Affected cases were diagnosed with AIS in specialty paediatric orthopaedic centres. One GWAS, ‘Japan’, was previously reported.9 A second GWAS, TSRHC I, analysed 702 family trios (parents and affected offspring) of various ethnicities, including 422 from a previously reported GWAS.9 TSRHC II is a third GWAS comparing NHW cases and controls, and will be reported in its entirety separately. Genotyping details are in the ‘Materials and methods’.

    View this table:
    Table 1

    Subjects included in rs11190870/chr 10 meta-analysis

    We screened the previously reported chromosome 10q24.31 region8 and observed evidence for association with SNPs in each GWAS. Given that each study used different genotyping platforms, and to study the region in more detail, we also imputed SNP genotypes using 1000 genomes data and tested association with AIS (figure 1). In the Japanese GWAS, SNP rs11190870 was the most significantly associated with AIS as expected. The top SNPs in the region for the TSRHC I dataset were rs11598564, the third-ranked SNP in the Japanese GWAS,8 and rs11190870 (see online supplementary figure S1). The two top-ranked SNPs in the TSRHC II dataset were rs594791, which was not genotyped in the Japanese or TSRHC I GWASs, and rs11190870. Overall associations with SNPs in the region were somewhat less significant in TSRHC I compared to TSRHC II or Japanese GWASs. As shown in figure 1A, restricting the TSRHC I analysis to 522 NHW families, the largest ethnic subset, reduced the association further and suggested that the weaker signal in this cohort was not due to mixed ethnicities. For all three studies, the chromosome 10q24.31 peak of association coincided closely.

    Recent follow-up studies in Asian cohorts have yielded strong evidence of AIS association with rs11190870 (table 2).10–12 Given the three GWAS results, and the correlation of rs11190870 with other SNPs in the region, we elected to test this SNP as a surrogate for the 10q24.31 locus in additional Asian and non-Asian cohorts. Individual results for each of the nine cohorts, and combined results for seven cohorts containing both genders, are given in table 2. Results suggest that the 10q24.31 locus, represented by rs11190870, confers substantial effects on AIS susceptibility, with an increased risk of ∼1.6-fold for the populations as a whole. We also analysed these results by gender (table 2). On average, we found similar ORs for females and males (1.59 compared to 1.42); however, results were far more variable between male datasets, an effect we attribute to modest sample size in the male groups.

    View this table:
    Table 2

    AIS association with rs11190870 by cohort and gender

    Figure 1
    Figure 1

    Regional plots of associated SNPs at the 10q24.31 locus (hg19 bp102957900-102997800). Association/linkage disequilibrium plots for TSRHC GWAS I (top), TSRHC GWAS II (middle) and Japanese GWAS (bottom). GWAS, genome-wide association studies.

    To refine the boundaries of the chromosome 10q24.31 associated interval, we compared linkage disequilibrium (LD) structures for northern European from Utah (CEU), and east Asian (ASN) populations in the region chr10: 102 957 900–102 997 800 using 1000 Genomes data (see online supplementary figure S2). As expected, pair-wise LD was more extensive in the ASN populations, and we therefore examined LD structure blocks in the CEU population to tentatively refine the boundaries of the AIS-associated interval. The top TSRHC I and TSRHC II SNPs, rs11598564 and rs594791, were correlated with rs11190870 (r2=0.79 and 0.97, respectively). We further noted that rs11190870 mapped within a ∼25 kb LD block that harbours several highly conserved sequence elements, including LBX1 and an uncharacterised transcript FLJ41350 (LOC399806), as defined by Genomic Evolutionary Rate Profiling (GERP) scores18 (figure 2). In fact, rs11190870 itself is encoded in a ∼1.7 kb sequence block (chr10:102 978 000–102 979 700) that is as evolutionarily constrained as LBX1 exons themselves, suggesting a potential biologic role for this region. According to ENCODE data, rs11190870 and other correlated SNPs (r2>0.8) in and near this sequence block are predicted to occur in, or disrupt, transcription factor-binding sites (table 3) (see online supplementary figure S3).21

    View this table:
    Table 3

    SNPs in predicted functional elements and SNPs in rs11190870-containing conserved sequence block

    Figure 2
    Figure 2

    Physical and linkage disequilibrium map of the 10q24.31 critical region. Linkage disequilibrium (LD) plot was constructed for population CEU using Haploview and is displayed as a heat map, where darker red corresponds to greater LD. Correlation coefficients (r2) are shown in the boxes. Physical map above denotes sequence conservation and ENCODE annotations.

    Discussion

    Using rs11190870 as a surrogate, we find that variation of the chromosome 10q24.31, LBX1-encoding locus increases genetic risk for AIS in multiple Asian and NHW cohorts ascertained in North America and eastern Asia, with ORs ranging from 1.32 to 1.85 combined 1.6. The single cohort (TSRHC I) that included additional ethnic groups (mostly Hispanic and African/African–American) in fact yielded the lowest ORs for SNP rs11190870 as noted. Our analyses suggested that these ethnic groups may also contribute to the association, but clearly larger, independent studies are required to address this question. In terms of gender, most cohorts yielded evidence for genetic susceptibility in females and males. The effect is clearest in females (combined OR=1.59, 95% CI 1.49 to 1.68), whereas we observed more variable results in males (combined OR=1.42, 95% CI 1.23 to 1.65). One cohort, in fact, produced OR less than one for males, but these findings could be a reflection of modest sample sizes. Our study underscores that larger male cohorts are needed to quantify the genetic risk conveyed by this locus in males versus females. It is also interesting that two cohorts, from Guangzhou and Hong Kong, were clearly associated with rs11190870 despite more stringent inclusion criteria (>20 and >35°, respectively). These data suggest that the chromosome 10q24 association may be independent of curve severity. Additional studies of cohorts with characterised, defined endpoints (ie, skeletal maturity) are warranted to identify loci that may contribute quantitatively to the degree, or rate, of deformity progression in AIS.

    The AIS-associated region encompasses LBX1, but we could not discover possible causal variant(s) in LBX1 itself in AIS patients despite resequencing of its coding region (data not shown). The overlap of evolutionarily conserved sequences 3’ of LBX1 with the peak of AIS association suggests the possibility of an enhancer function that is disrupted by AIS mutations. Using Haploreg (http://www.broadinstitute.org/mammals/haploreg/haploreg.php) to search data from the ENCODE project,19 we found that rs11190870 and other correlated SNPs (r2>0.8) are predicted to alter multiple predicted transcription factor-binding sites. Thus, further resequencing in AIS subjects carrying risk haplotypes targeting conserved non-coding elements in the region is warranted to discover possible causal variants. LBX1, indirectly linked to AIS by these data, is an interesting candidate for disease susceptibility. Studies of targeted mouse mutants have revealed that ladybird homeobox 1 is involved in proper migration of muscle precursor cells, specification of certain cardiac neural crest cells13 ,14 and determination of dorsal spinal neurons and hindbrain somatosensory neurons of the developing neural tube.15–17 AIS has long been proposed to be a problem in muscle or nerve development, or both.5 However, the role of factors such as LBX1 in any of these processes at later developmental stages, that is, adolescent growth, is unknown. Developing appropriate systems that model the course of human axial development will be critically important to ongoing AIS genetic research.

    Our data illustrate the benefit of GWAS meta-analyses and consortium-driven efforts to discover risk factors for common diseases such as AIS. Clearly, the chromosome 10q24.31 susceptibility locus is not unique to a particular ethnic group and similar risk factors surely await discovery. Indeed, published AIS risk loci are estimated to account for about 1% of the overall trait variance.20 Future efforts should target discovery of additional AIS risk factors, and more comprehensive studies in males and in under-represented ethnic groups.

    Materials and methods

    Cohorts

    All study subjects were recruited with institutional ethical approval as detailed in the online supplemental methods. All affected cases met accepted clinical criteria for IS, including deformity greater than or equal to 10° measured by the Cobb angle method from standing radiographs, and no other coexisting disorders. The minimum severity for affected cases in these cohorts ranged from 16° to 35° Cobb angle measurement at the time of ascertainment. Ethnically matched controls were recruited from local communities. TSRHC I (422 of the 702 families), Guangzhou, Nanjing, and University of Hong Kong cohorts were as described.9–12 Ascertainment details for Chinese University Hong Kong, TSRHC II, and TSRHC III are given in the online supplementary materials.

    Genotyping

    The Japanese and TSRHC I GWASs were performed as previously described.8 ,9 TSRHC GWAS II samples (457 cases, 744 controls) were genotyped using the Illumina HumanOmniExpress-12 V.1.0 beadchip containing 730 498 markers. PLINK V.1.0721 was used to perform quality control, which includes removing individuals with call rates <95%, individuals with ambiguous gender information, and SNPs with >5% missing data. Full genotyping and statistical details of this study will be reported separately. Guangzhou, Nanjing and University of Hong Kong genotyping methods were as described.10–12 Genotyping methods for the Chinese University of Hong Kong and TSRHC III are given in the online supplementary materials.

    Statistical methods

    Single and combined tests of association were performed using Fisher's exact test as previously described.9 To predict individual genotypes at untyped loci, we used statistical tools in the software application MaCH V.1.0,22 applying reference genotypes from the 1000 Genomes Project June 2011 for populations CEU or East Asian (ASN) as appropriate. Regional plots of chromosome 10 associations for actual and imputed genotypes were created using Locus Zoom software.23 Plots of pair-wise LD were constructed using Haploview software V.4.1.24

    Acknowledgments

    We are grateful to all study participants and collaborating institutions.

    References

    1. ↵
      1. Herring JA
      . Tachdjian's pediatric orthopaedics. Vol I. Philadelphia: WB Saunders, 2008:265–411.
    2. ↵
      1. Karol LA,
      2. Johnston CE,
      3. Brown RH,
      4. Madison M
      . Progression of the curve in boys who have idiopathic scoliosis. J Bone Joint Surg 1993;75-A:1804–10.
      OpenUrlPubMed
    3. ↵
      1. Little DG,
      2. Song KM,
      3. Katz D,
      4. Herring JA
      . Relationship of peak height velocity to other maturity indicators in idiopathic scoliosis in girls. J Bone Joint Surg Am 2000;82:685–93.
      OpenUrlAbstract/FREE Full Text
    4. ↵
      1. Alf L,
      2. Nachemson M,
      3. Sahlstrand T
      . Etiologic factors in adolescent idiopathic scoliois. Spine 1977;2:176–84.
      OpenUrlCrossRef
    5. ↵
      1. Sharma S,
      2. Wise C
      . Current understanding of genetic factors in idiopathic scoliosis. In: Dunwoodie KA, ed. The genetics and development of scoliosis. Springer, 2010:167–90.
    6. ↵
      1. Ward K,
      2. Ogilvie J,
      3. Argyle V,
      4. Nelson L,
      5. Meade M,
      6. Braun J,
      7. Chettier R
      . Polygenic inheritance of adolescent idiopathic scoliosis: a study of extended families in Utah. Am J Med Genet A 2010;152A:1178–88.
      OpenUrl
    7. ↵
      1. Tang NL,
      2. Yeung HY,
      3. Hung VW,
      4. Di Liao C,
      5. Lam TP,
      6. Yeung HM,
      7. Lee KM,
      8. Ng BK,
      9. Cheng JC
      . Genetic epidemiology and heritability of AIS: a study of 415 Chinese female patients. J Orthop Res 2012;30:1464–9.
      OpenUrlCrossRefPubMed
    8. ↵
      1. Takahashi Y,
      2. Kou I,
      3. Takahashi A,
      4. Johnson TA,
      5. Kono K,
      6. Kawakami N,
      7. Uno K,
      8. Ito M,
      9. Minami S,
      10. Yanagida H,
      11. Taneichi H,
      12. Tsuji T,
      13. Suzuki T,
      14. Sudo H,
      15. Kotani T,
      16. Watanabe K,
      17. Chiba K,
      18. Hosono N,
      19. Kamatani N,
      20. Tsunoda T,
      21. Toyama Y,
      22. Kubo M,
      23. Matsumoto M,
      24. Ikegawa S
      . A genome-wide association study identifies common variants near LBX1 associated with adolescent idiopathic scoliosis. Nat Genet 2011;43:1237–40.
      OpenUrlCrossRefPubMed
    9. ↵
      1. Sharma S,
      2. Gao X,
      3. Londono D,
      4. Devroy SE,
      5. Mauldin KN,
      6. Frankel JT,
      7. Brandon JM,
      8. Zhang D,
      9. Li QZ,
      10. Dobbs MB,
      11. Gurnett CA,
      12. Grant SF,
      13. Hakonarson H,
      14. Dormans JP,
      15. Herring JA,
      16. Gordon D,
      17. Wise CA
      . Genome-wide association studies of adolescent idiopathic scoliosis suggest candidate susceptibility genes. Hum Mol Genet 2011;20:1456–66.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Fan YH,
      2. Song YQ,
      3. Chan D,
      4. Takahashi Y,
      5. Ikegawa S,
      6. Matsumoto M,
      7. Kou I,
      8. Cheah KS,
      9. Sham P,
      10. Cheung KM,
      11. Luk KD
      . SNP rs11190870 near LBX1 is associated with adolescent idiopathic scoliosis in southern Chinese. J Hum Genet 2012;57:244–6.
      OpenUrlCrossRefPubMed
    11. ↵
      1. Gao W,
      2. Peng Y,
      3. Liang G,
      4. Liang A,
      5. Ye W,
      6. Zhang L,
      7. Sharma S,
      8. Su P,
      9. Huang D
      . Association between Common Variants near LBX1 and Adolescent Idiopathic Scoliosis Replicated in the Chinese Han Population. PLoS ONE 2013;8:e53234.
      OpenUrlCrossRefPubMed
    12. ↵
      1. Jiang H,
      2. Qiu X,
      3. Dai J,
      4. Yan H,
      5. Zhu Z,
      6. Qian B,
      7. Qiu Y
      . Association of rs11190870 near LBX1 with adolescent idiopathic scoliosis susceptibility in a Han Chinese population. Eur Spine J 2013;22:282–6.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Gross MK,
      2. Dottori M,
      3. Goulding M
      . LBX1 specifies somatosensory association interneurons in the dorsal spinal cord. Neuron 2002;34:535–49.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Cheng L,
      2. Samad OA,
      3. Xu Y,
      4. Mizuguchi R,
      5. Luo P,
      6. Shirasawa S,
      7. Goulding M,
      8. Ma Q
      . Lbx1 and Tlx3 are opposing switches in determining GABAergic versus glutamatergic transmitter phenotypes. Nat Neurosci 2005;8:1510–15.
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Jagla K,
      2. Dollé P,
      3. Mattei MG,
      4. Jagla T,
      5. Schuhbaur B,
      6. Dretzen G,
      7. Bellard F,
      8. Bellard M
      . Mouse Lbx1 and human LBX1 define a novel mammalian homeobox gene family related to the Drosophila lady bird genes. Mech Dev 1995;53:345–56.
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Schafer K,
      2. Braun T
      . Early specification of limb muscle precursor cells by the homeobox gene Lbx1h. Nat Genet 1999;23:213–16.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Schäfer K,
      2. Neuhaus P,
      3. Kruse J,
      4. Braun T
      . The homeobox gene Lbx1 specifies a subpopulation of cardiac neural crest necessary for normal heart development. Circ Res 2003;92:73–80.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Cooper GM,
      2. Stone EA,
      3. Asimenos G,
      4. Green ED,
      5. Batzoglou S,
      6. Sidow A
      . Distribution and intensity of constraint in mammalian genomic sequence. Genome Res 2005;15:901–13.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      1. Ward LD,
      2. Kellis M
      . HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Res 2012;40:D930–4.
      OpenUrlAbstract/FREE Full Text
    20. ↵
      1. Kou I,
      2. Takahashi Y,
      3. Johnson TA,
      4. Takahashi A,
      5. Guo L,
      6. Dai J,
      7. Qiu X,
      8. Sharma S,
      9. Takimoto A,
      10. Ogura Y,
      11. Jiang H,
      12. Yan H,
      13. Kono K,
      14. Kawakami N,
      15. Uno K,
      16. Ito M,
      17. Minami S,
      18. Yanagida H,
      19. Taneichi H,
      20. Hosono N,
      21. Tsuji T,
      22. Suzuki T,
      23. Sudo H,
      24. Kotani T,
      25. Yonezawa I,
      26. Londono D,
      27. Gordon D,
      28. Herring JA,
      29. Watanabe K,
      30. Chiba K,
      31. Kamatani N,
      32. Jiang Q,
      33. Hiraki Y,
      34. Kubo M,
      35. Toyama Y,
      36. Tsunoda T,
      37. Wise CA,
      38. Qiu Y,
      39. Shukunami C,
      40. Matsumoto M,
      41. Ikegawa S
      . Genetic variants in GPR126 are associated with adolescent idiopathic scoliosis. Nat Genet 2013;45:676–9.
      OpenUrlCrossRefPubMed
    21. ↵
      1. Purcell S,
      2. Neale B,
      3. Todd-Brown K,
      4. Thomas L,
      5. Ferreira MA,
      6. Bender D,
      7. Maller J,
      8. Sklar P,
      9. de Bakker PI,
      10. Daly MJ,
      11. Sham PC
      . PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007;81:559–75.
      OpenUrlCrossRefPubMed
    22. ↵
      1. Li Y,
      2. Abecasis G
      . Mach 1.0: rapid haplotype reconstruction and missing genotype inference. Am J Hum Genet 2006;S79:2290.
      OpenUrl
    23. ↵
      1. Prium R,
      2. Welch R,
      3. Sanna S,
      4. Teslovich T,
      5. Chines P,
      6. Gliedt T,
      7. Boehnke M,
      8. Abecasis G,
      9. Willer C
      . LocusZoom: regional visualization of genome-wide association scan results. Bioinformatics 2010;26:2336–7.
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Barrett J,
      2. Fry B,
      3. Maller J,
      4. Daly M
      . Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005;21:263–5.
      OpenUrlAbstract/FREE Full Text

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • The authors wish it to be known that DL, IK, TAJ, and SS, in their opinion, should be regarded as joint first authors.

    • Collaborators Japanese Scoliosis Clinical Research Group the participants list is provided in the online supplementary material, Notes. TSRHC (Texas Scottish Rite Hospital for Children) IS Collaborative Group – collaborators list is provided in the online supplementary material, Notes. International Consortium for Scoliosis Genetics the membership list is provided in the online supplementary material, Notes. A Alanay; A Child; A Moreau; A Santiago-Cornier; A Zaidman; B Alman; B Dahl; B S Richards; B Yeung; C Eberson; C Gurnett; C Johnston; C Raggio; D Rousie; D Sucato; E Acaroglu; E Clark; F Berry; F Moldovan; G Liu; H Iwinski; H Sudo; H K Wong; H Yanagida; I Yonezawa; J Birch; J C Channing; J P Dormans; J Fairbank; J Ogilvie; J C Tassone; J Yu; K Kono; K Kusumi; K Patten; K Rathjen; K Uno; K Ward; K Watanabe; L Karol; M Dobbs; M Ito; N Ahituv; N Hadley-Miller; N Kawakami; O Pourquie; P C Edery; P F Giampietro; P Turnpenny; P Vidal; R Blank; R M Castelein; R Marcucio; R Shindell; S Dunwoodie; S Edelstein; S F A Grant; S Minami; T Kotani; T Kotwicki; T Milbrandt; T Tsuji; V Talwakar; W Schrader; W Skalli; X Liu; Y Qiu; Y Toyama; Z Zhu.

    • Contributors CW: conception/design, data collection, interpretation, and guarantor; DL, SS, P Sham and DG: conception/design, analysis, and interpretation; IK, Y-QS: conception/design, EAS data collection, analysis, and interpretation; TAJ: analysis; YO, TT, AT: data collection; MM, JH, T-PL, XW, EMST, DH, P Su, YQ, JC: data collection and interpretation; Y-HF: conception/design and analysis; DC, KSEC: conception/design and interpretation; XQ, HJ: data collection and analysis; KMCC, KL: conception/design, data collection, and interpretation; NT: data collection, analysis, and interpretation; SI: conception/design, data collection, and guarantor.

    • Funding This work was supported by a Direct Grant from The Chinese University of Hong Kong (to NT and JC); the Excellent Young Scientist of the New Century of Educational Ministry in China (to PS); the Natural Science Foundation of China (81171767 and 30901570 to YQ); the National Institutes of Health (R01 HD052973), TSRHC Research Fund project 867, Crystal Charity Ball, Scoliosis Research Society, and Cain Foundation (to CAW); and a Direct Grant from The Chinese University of Hong Kong (to NT and KC).

    • Ethics approval The enrollment of patients, into the study cohort at the Chinese University of Hong Kong, was approved by the Institute of Clinical Research Ethics Committee. Subjects included in the GWAS, TSRHC I, II, and III cohorts were enrolled in a study approved by the University of Texas Southwestern Medical Center Institutional Review Board. Otherwise, all study subjects were recruited with institutional ethical approval as detailed in the online supplemental methods.

    • Provenance and peer review Not commissioned; externally peer reviewed.