REPORT Exome Sequencing Identifies CCDC8 Mutations in 3-M Syndrome, Suggesting that CCDC8 Contributes in a Pathway with CUL7 and OBSL1 to Control Human Growth Dan Hanson, 1,2 Philip G. Murray, 1,2,3 James O’Sullivan, 2,3 Jill Urquhart, 2,3 Sarah Daly, 2,3 Sanjeev S. Bhaskar, 2,3 Leslie G. Biesecker, 4,5 Mars Skae, 3 Claire Smith, 6 Trevor Cole, 7 Jeremy Kirk, 8 Kate Chandler, 2,3 Helen Kingston, 2,3 Dian Donnai, 2,3 Peter E. Clayton, 1,3,9 and Graeme C.M. Black 2,3,9, * 3-M syndrome, a primordial growth disorder, is associated with mutations in CUL7 and OBSL1. Exome sequencing now identifies muta- tions in CCDC8 as a cause of 3-M syndrome. CCDC8 is a widely expressed gene that is transcriptionally associated to CUL7 and OBSL1, and coimmunoprecipitation indicates a physical interaction between CCDC8 and OBSL1 but not CUL7. We propose that CUL7, OBSL1, and CCDC8 are members of a pathway controlling mammalian growth. 3-M syndrome (MIM 273750 and 612921), named after the three principle geneticists who first described the condi- tion, 1 is an autosomal-recessive condition characterized by severe postnatal growth restriction that results in signif- icantly short stature (typical final adult height of 120– 130 cm). 1–3 The syndrome is also associated with distinc- tive facial dysmorphism, including triangular facies, frontal bossing, midface hypoplasia, fleshy tipped nose, and full fleshy lips. Prominent heels are also evident in younger individuals with 3-M syndrome. Skeletal abnor- malities including slender tubular bones and relatively tall vertebral bodies are seen in some individuals. The condition is associated with normal intelligence. 1–3 In 2005, autozygosity mapping revealed a 3-M syndrome locus on chromosome 6p21.1, and pathogenic mutations in CULLIN7 (CUL7 [MIM 609577]) were subsequently identified as the primary cause of 3-M syndrome. 4–6 More recently, further autozygosity mapping revealed a second 3-M syndrome locus on chromosome 2q35-q36.1 with the underlying mutations identified in Obscurin-like 1 (OBSL1 [MIM 610991]). 7,8 To date, there have been approx- imately 100 reported cases of 3-M syndrome. 1–8 CUL7 acts as the structural component of an E3 ubiquitin ligase consisting of SKP1, ROC1, and FBXW8, 9 as part of the proteasomal degradation pathway, whose targets include the growth factor signaling molecule IRS-1. 10 Unlike other cullin proteins, CUL7 also has nonproteolytic functions, including interaction with the tumor suppressor p53; loss of CUL7 activity results in inhibition of p53-mediated cell-cycle progression. 11,12 In contrast OBSL1 is highly homologous to the giant muscle protein Obscurin and has been identified as a cytoskeletal adaptor protein 13 that interacts with the giant cytoskeletal proteins titin and myomesin. 14 OBSL1 had not been recognized to have a role in growth. We previously demonstrated that loss of OBSL1 by siRNA knockdown leads to concomitant loss of CUL7 in an HEK293 model. 7 Huber et al. established that individuals with 3-M syndrome and OBSL1 mutations showed significant modulation of Insulin-like growth factor binding protein 2 (IGFBP2 [MIM 146731]) and IGFBP5 (MIM 146734) expression. 8 These findings sug- gested that both CUL7 and OBSL1 are likely to exist in a common pathway 7 and that alterations in IGFBPs and hence changes in IGF levels may contribute to the patho- genesis of 3-M syndrome. 8 We identified individuals of South Asian descent who had 3-M syndrome and a history of consanguinity and did not carry CUL7 or OBSL1 mutations (Figure 1A and Table 1). With institutional ethical approval and informed consent, whole blood was collected from affected individ- uals and genomic DNA extracted via standard procedures. Autozygosity mapping on three such individuals, all unrelated (3M-1, 3M-2, and 3M-3), using the Affymetrix Genome-Wide Human SNP Array 6.0, and analysis of SNP genotype data by the AutoSNPa program 15 revealed a potential third 3-M syndrome locus located on chromo- some 19q13.2-q13.32. The autozygous region of 7.9 Mb is flanked by recombinant SNPs rs535840 (45,373,227 bp) and rs1991722 (53,270,701 bp) (Figure 1B) and contains 301 protein-coding genes. Exome sequencing capture with the Agilent SureSelect in-solution target enrich- ment system using the ABI SOLiD 4.0 platform was per- formed in all three individuals. Filtering of sequence data from the autozygous region against dbSNP, HapMap, and 1 Department of Endocrinology, Manchester Academic Health Sciences Centre (MAHSC), School of Biomedicine, University of Manchester, Manchester, UK M13 9WL; 2 Genetic Medicine Research Group, Manchester Biomedical Research Centre, Manchester Academic Health Sciences Centre, University of Man- chester and Central Manchester Foundation Trust, St Mary’s Hospital, Manchester M13 9WL, UK; 3 Central Manchester University Hospitals Foundation Trust, Manchester M13 9WL, UK; 4 National Institutes of Health (NIH) Intramural Sequencing Center (NISC), National Institutes of Health, Bethesda, MD 20892, USA; 5 National Human Genome Research Institute, NIH, Bethesda, MD 20814, USA; 6 East Lancashire Hospital NHS Trust, Royal Blackburn Hospital, Haslingden Road, Blackburn BB2 3HH ,UK; 7 West Midlands Regional Genetics Service, Birmingham Women’s Hospital, Birmingham B15 2TG, UK; 8 Department of Paediatric Endocrinology, Birmingham Children’s Hospital, Birmingham B4 6NH, UK 9 These authors contributed equally to this work *Correspondence: graeme.black@manchester.ac.uk DOI 10.1016/j.ajhg.2011.05.028. Ó2011 by The American Society of Human Genetics. All rights reserved. 148 The American Journal of Human Genetics 89, 148–153, July 15, 2011