Skip to main content
NCBI home page
Human Molecular Genetics logoLink to Human Molecular Genetics
. 2015 Jun 3;24(17):5053–5059. doi: 10.1093/hmg/ddv210

A trans-ethnic genome-wide association study identifies gender-specific loci influencing pediatric aBMD and BMC at the distal radius

Alessandra Chesi 1,*, Jonathan A Mitchell 5, Heidi J Kalkwarf 7, Jonathan P Bradfield 2, Joan M Lappe 8, Shana E McCormack 1,3,6, Vicente Gilsanz 9, Sharon E Oberfield 10, Hakon Hakonarson 1,2,6, John A Shepherd 11, Andrea Kelly 6,4, Babette S Zemel 6,4,†, Struan FA Grant 1,2,3,6,†
PMCID: PMC4527490  PMID: 26041818

Abstract

Childhood fractures are common, with the forearm being the most common site. Genome-wide association studies (GWAS) have identified more than 60 loci associated with bone mineral density (BMD) in adults but less is known about genetic influences specific to bone in childhood. To identify novel genetic factors that influence pediatric bone strength at a common site for childhood fractures, we performed a sex-stratified trans-ethnic genome-wide association study of areal BMD (aBMD) and bone mineral content (BMC) Z-scores measured by dual energy X-ray absorptiometry at the one-third distal radius, in a cohort of 1399 children without clinical abnormalities in bone health. We tested signals with P < 5 × 10−6 for replication in an independent, same-age cohort of 486 Caucasian children. Two loci yielded a genome-wide significant combined P-value: rs7797976 within CPED1 in females [P = 2.4 × 10−11, β =− 0.30 standard deviations (SD) per T allele; aBMD-Z] and rs7035284 at 9p21.3 in males (P = 1.2 × 10−8, β = 0.28 SD per G allele; BMC-Z). Signals at the CPED1-WNT16-FAM3C locus have been previously associated with BMD at other skeletal sites in adults and children. Our result at the distal radius underscores the importance of this locus at multiple skeletal sites. The 9p21.3 locus is within a gene desert, with the nearest gene flanking each side being MIR31HG and MTAP, neither of which has been implicated in BMD or BMC previously. These findings suggest that genetic determinants of childhood bone accretion at the radius, a skeletal site that is primarily cortical bone, exist and also differ by sex.

Introduction

Osteoporosis and osteopenia are common conditions affecting up to 54 million individuals in the USA, and constitute a major public health concern related to an estimated 2 million fractures every year (1).

Bone accretion (a consequence of bone formation occurring at a faster pace than resorption resulting in both increasing size and greater mineral content and density of skeletal components) is rapid in childhood and adolescence, and ultimately dictates peak bone mass in adulthood (2). Suboptimal peak bone mass may contribute to low bone mass and increased osteoporotic risk later in life. However, even if childhood is considered a critical period for the determination of osteoporotic risk, relatively little is known about the genetic determinants of bone accretion during growth.

BMD and osteoporosis have a strong heritable component. Genome-wide association studies (GWAS) have identified more than 60 loci associated with BMD in adults (3–5), but despite recent progress genetic influences specific to bone in childhood remain largely to be elucidated. Identifying the factors that influence bone mineral accretion during childhood and adolescence has important implications for the prevention of this common, disabling disorder. An advantage of pediatric cohorts is that distillation of the genetic component of complex phenotypes such as bone mass (which depends on several environmental factors, such as dietary intake, physical activity and smoking) should be easier in children, where environmental exposure has been present for a relatively short period of their lifetime. Demonstrating the promise of this approach, GWAS using pediatric cohorts have recently identified three novel loci associated with total body BMD [WNT16 (6), SP7/Osterix (7) and RIN3 (8)].

In children, as in adults, fracture rates are higher in individuals with lower BMD. Femoral neck and lumbar spine aBMD as measured by dual-energy X-ray absorptiometry (DXA) is the main diagnostic marker of osteoporosis in adults (9,10).

In children, the most common site of fracture is the upper extremities, accounting up to 45% of all fractures (11); in turn, the majority of these (20–30%) involve the distal end of the radius (12). Prior wrist fracture also strongly predicts 3-year risk of any future osteoporotic fracture for older and younger postmenopausal women, independent of baseline BMD and common osteoporosis risk factors (13).

To date, other GWAS studies of childhood bone accretion have used DXA scans of the total body to test for associations. Given the high forearm fracture rate in children and adults, we elected to identify novel genetic factors that influence pediatric bone strength specifically at the one-third distal radius, by performing a GWAS of aBMD and BMC (both routinely used to assess bone mineralization in children) on participants from the Bone Mineral Density in Childhood Study.

Since sex plays an important role in determining loss of peak bone mass during lifetime [30–50% in women versus 20–30% in men (14)] and osteoporotic risk, we reasoned that these differences may already be operating during pediatric age affecting bone mass accrual, and decided to perform a sex-stratified analysis. Indeed, we recently uncovered sex-specific effects of several adult bone loci on aBMD and BMC at different skeletal sites in Caucasian children (15).

Results

We performed a GWAS of BMC and aBMD Z-scores at the one-third distal radius in a multicenter, multi-ethnic cohort of 1399 healthy children (720 girls and 679 boys; 53% Caucasian, 22% African American, 14% Hispanic) aged 5–20 years old from the Bone Mineral Density in Childhood Study. Details of the study were previously published (16). The study population is outlined in Supplementary Material, Tables S1 and S2.

Phenotypes were expressed as sex-, age-, population ancestry (black versus non-black) specific Z-scores, adjusted for height Z-score (16). Z-scores were used to express BMC and aBMD in children to account for age-specific increases that occur in these outcomes; Z-scores were further adjusted for height Z-score, as recommended, so that the size-related effects of DXA BMC and aBMD measurements were minimized (17,18).

To account for the effect of sex differences in maturational timing on bone accrual, we performed a sex-stratified analysis. We tested ∼7.2 million SNPs (MAF > 5%; directly genotyped or imputed with INFO >0.4) using a univariate linear mixed model as implemented in the software GEMMA (19) to account for population stratification and relatedness. The results showed appropriate control for population structure with genomic inflation factors (λ) approaching unity and Q–Q plots revealing no inflation of the test statistic and gain of power when correction for height Z-scores was applied (Fig. 1).

Figure 1.

Figure 1.

Open in a new tab

Q–Q plots for radius aBMD in females (A) and radius BMC in males (B) showing appropriate control for population structure (no inflation of the test statistics) and gain of power when height correction is applied (black dots versus red dots).

Among females, multiple SNPs at a locus on 7q31 (CPED1-WNT16-FAM3C) already known to influence bone density at other skeletal sites in children and adults (4,6,8,20,21) reached genome-wide significance at P < 5.0 × 10−8 (sentinel SNP rs7797976, P = 1.2 × 10−10, β =− 0.340 standard deviations (SD) per T allele, SE = 0.052, MAF = 0.353; aBMD-Z height-adjusted). In males, a novel locus on 9p21.3 yielded suggestive association (sentinel SNP rs7035284, P = 2.8 × 10−7, β = 0.279 SD per G allele, SE = 0.054, MAF = 0.314; BMC-Z height-adjusted). All signals with P < 5 × 10−6 were tested for replication in an independent, same-age cohort of 486 Caucasian children (245 girls and 241 boys) (Supplementary Material, Tables S3 and S4). Both CPED1-WNT16-FAM3C and the 9p21.3 locus achieved at least nominal significance in this cohort within the relevant gender (rs7797976; P = 0.047, β =− 0.163 SD per T allele, SE = 0.081, MAF = 0.374; rs7035284; P = 0.015, β = 0.281 per G allele, SE = 0.114, MAF = 0.180), and reached genome-wide significance when combining the cohorts (rs7797976, combined P = 2.4 × 10−11, β =− 0.298 SD per T allele, SE = 0.044, MAF = 0.358; rs7035284, combined P = 1.2 × 10−8, β = 0.277 per G allele, SE = 0.048, MAF = 0.279, Fig. 2 and Table 1). Results for the converse, namely radius aBMD in males and radius BMC in females, followed a similar trend, which was expected since BMD and BMC Z-scores are highly correlated in our data set (correlation ∼0.62), but did not reach statistical significance (Supplementary Material, Table S5).

Figure 2.

Figure 2.

Open in a new tab

Regional Manhattan plots of the two genome-wide significant signals identified in this study: the CPED1-WNT16-FAM3C locus is associated with radius aBMD in females (A) and 9p21.3 with radius BMC in males (B). Circles show P-values and positions of SNPs within the locus. The top SNPs (rs7797976 and rs7035284) are denoted by a diamond. Colors indicate varying degrees of pairwise linkage disequilibrium (HapMap CEU release 22) between the top SNP and all other SNPs.

Table 1.

Two loci are associated with radius bone density or content in children

Locus SNP Position Nearest genes Major allele Minor allele MAF Beta SE P-value Trait Cohort sex N
Discovery 7q31.31 rs7797976 120843516 CPED1 C T 0.353 −0.340 0.052 1.20 × 10−10 Radius aBMD Females 706
9p21.3 rs7035284 21678973 MIR31HG, MTAP A G 0.314 0.279 0.054 2.83 × 10−7 Radius BMC Males 665
Replication 7q31.31 rs7797976 120843516 CPED1 C T 0.374 −0.163 0.081 4.71 × 10−2 Radius aBMD Females 242
9p21.3 rs7035284 21678973 MIR31HG, MTAP A G 0.180 0.281 0.114 1.47 × 10−2 Radius BMC Males 239
Combined 7q31.31 rs7797976 120843516 CPED1 C T 0.358 −0.298 0.044 2.42 × 10−11 Radius aBMD Females 948
9p21.3 rs7035284 21678973 MIR31HG, MTAP A G 0.279 0.277 0.048 1.21 × 10−8 Radius BMC Males 904
Open in a new tab

7q31.31 (CPED1-WNT16-FAM3C) in females reached genome-wide significance (P < 5 × 10−8) in the discovery cohort alone; 9p21.3 yielded a suggestive association (P < 5 × 10−6) in males in the discovery cohort. Both loci achieved at least nominal significance in the replication cohort within the relevant gender and were genome-wide significant when combining the cohorts. Effect sizes (betas) refer to the minor allele.

In a joint analysis including both males and females, only the CPED1-WNT16-FAM3C locus remained genome-wide significant, for both aBMD (rs7797976; P = 2.6 × 10−10) and BMC (rs7797976; P = 2.6 × 10−8). A test of heterogeneity of the betas indicated sex-specificity for both loci (CPED1-WNT16-FAM3C: rs7797976, P = 8.1 × 10−3, aBMD; 9p21.3: rs7035284, P = 1.5 × 10−4, BMC), justifying our sex-stratified analysis approach and in agreement with what we have previously reported (15).

To assess which of the two main ethnicities (Caucasians and African Americans) were the primary drivers for the two genome-wide significant associations, we determined the population substructure from genotyped data using ADMIXTURE (22). Children were assigned to the African American (N = 349), Caucasian (N = 1389) or Other/Not classified (N = 114) group based on their highest fraction of estimated ancestry proportions (i.e. >0.50, see Materials and Methods).

It was clear that the CPED1-WNT16-FAM3C signal was principally from the Caucasian portion of the female cohort (Caucasian β =− 0.333, SE = 0.049, N = 716; African American β =− 0.149, SE = 0.114, N = 185) while the 9p21.3 locus yielded a very consistent magnitude of effect in both Caucasian (β = 0.318, SE = 0.063, N = 673) and African American (β = 0.248, SE = 0.103, N = 164) males. This was confirmed by a test of heterogeneity of the betas in the combined (males plus females) cohort (CPED1-WNT16-FAM3C: rs7797976, P = 0.014; 9p21.3: rs7035284, P = 0.379; N African American =349, N Caucasian =1389).

Subsequently, we assessed whether variants in the identified loci were correlated with eQTLs using publicly available genome-wide expression data sets (23). Despite leveraging multiple data sources, we found no compelling evidence linking the GWAS signals to expression of specific transcripts. This negative result may be due to the fact that the tissues and cell types available (adipose and skin tissues, fibroblasts, lymphoblastoids and T-cells) were not very relevant to BMD and bone biology, and further analyses will be required to identify the target transcripts regulated by the GWAS signals.

Next, by leveraging GRAIL (24), we explored the degree of functional similarity between the genes nearest the SNPs yielding P-values < 10−5 in our GWAS with 169 genes known to be associated with bone density and osteoporosis in adults [extracted from the GWAS catalog (3), see Materials and Methods].

Among the 97 female candidate genes, 34 achieved nominal significance (P < 0.05, Supplementary Material, Table S6). Among the top scoring were four known bone genes (WNT16, CCDC170, CPED1 and ESR1). 26 female genes were highly connected (P < 0.05, Supplementary Material, Table S6 and Supplementary Material, Fig. S1) with the top five keywords linking the genes being: ‘cadherin’, ‘neurons’, ‘ampa’, ‘synaptic’ and ‘receptor’. Among the 83 male candidate genes, 27 were significantly associated with known bone genes (P < 0.05, Supplementary Material, Table S6). Sixteen male genes were highly connected (P < 0.05, Supplementary Material, Table S6 and Supplementary Material, Fig. S1) with the top five keywords being: ‘cadherin’, ‘neurons’, ‘cadherins’, ‘wide’ and ‘GDNF’.

Discussion

Two loci achieved genome-wide significance in our study: the CPED1-WNT16-FAM3C locus in females and a novel locus on 9p21.3 in males.

The female signal is located in an intron of CPED1, in close proximity to WNT16 and FAM3C.

CPED1 (cadherin-like and PC-esterase domain containing 1, also known as C7orf58) encodes a poorly characterized protein that is conserved from chimpanzee to frog, containing a ATP-Grasp domain fused to a PC-esterase domain, and is the first identified secreted tubulin-tyrosine ligase (TTL) -like enzyme in eukaryotes (25).

WNT16 is a member of the WNT gene family, secreted glycoproteins that signal through both the Wnt–β-catenin pathway (canonical Wnt pathway) and noncanonical Wnt pathways, and are implicated in oncogenesis and in several developmental processes. Activation of canonical β-catenin signaling increases bone mass in mice and humans (26), while activation of noncanonical Wnt signaling by osteoblast-produced WNT5a has been shown to increase osteoclastogenesis (27). Wnt16-deficient mice develop spontaneous fractures as a result of low cortical thickness and high cortical porosity, while trabecular bone volume is not altered (20).

FAM3C is an uncharacterized member of the family with sequence similarity 3 (FAM3) family and encodes a secreted protein with a GG domain.

Signals at the CPED1-WNT16-FAM3C locus have been previously reported to be associated with BMD at other skeletal sites in adults (28) and children (skull and total body aBMD) of European ancestry (6). Associations with forearm BMD and fractures in adults of European ancestry have recently been reported by Zheng et al. (21). Our top signal (rs7797976) is genome-wide significant in that study (P = 7.1 × 10−10), and is in strong LD (r2 > 0.5) with 23 additional reported signals in CPED1. However, rs7797976 is not in LD with the signals in WNT16 and FAM3C reported in the study (rs7776725; r2 = 0.001).

Our result at the distal radius in children underscores the importance of this locus during growth at multiple skeletal sites, again primarily in Caucasians.

It is worth noting that the signal in our study reached genome-wide significance with a relatively small number of individuals compared with the sample size of previous studies; also, the effect sizes we see in children for radius aBMD/BMC (β =− 0.30 for the female locus and β = 0.28 for the male locus) are substantially higher than the corresponding signals reported previously in adults [rs7797976; β =− 0.117 for forearm BMD (21)] or in children for total body BMD [rs7797976; β =− 0.123 or −0.164, depending on the cohort (6)] or for upper limb BMD derived from total body DXA scans [rs7798060; r2 = 0.97 with rs7797976; β =− 0.170 (8)]. Although we cannot exclude that this is due to the winner's curse effect given the relative size of our cohorts, this may also indicate that the distal radius is the preferred site of action for CPED1-WNT16-FAM3C in children, or that cortical bone phenotypes are more accurately characterized by a dedicated forearm scan in children, i.e. the radius may be the best measure to gain genetic insight in to the pediatric skeleton. Consistently, a recent study in children reports that variants at CPED1-WNT16-FAM3C exerted a larger influence on upper limb aBMD than upon lower limb aBMD derived from total body scans (8). The mechanism of this difference remains to be investigated, since the bone composition (cortical versus trabecular) and the developmental origins of upper and lower limbs are largely similar. Differential weight loading between the upper and lower limbs might play a role.

The signal on 9p21.3, driven equally by both main ethnicities in the cohort, resides within a gene desert, with the nearest gene flanking each side being MIR31HG (miR-31 host gene, non-protein coding) and MTAP (methylthioadenosine phosphorylase), neither of which has been implicated in aBMD or BMC previously. miR-31 has been studied in the context of different cancers, where it plays a key role in progression and metastasis (29). In ex vivo assays, miR-31 expression affects proliferation and invasion of cell lines of Ewing sarcoma (the second most common bone tumor in children and young adults) (30). Interestingly, mutations in a previously uncharacterized terminal exon of MTAP have been shown to result in diaphyseal medullary stenosis with malignant fibrous histiocytoma (DMS-MFH), an autosomal-dominant syndrome characterized by bone dysplasia, myopathy and bone cancer (31).

Many of the signals yielding at least suggestive association in our GWAS (P < 10−5) were functionally connected in the literature to known adult bone loci, suggesting that adult bone genes may play an important role in bone growth during pediatric age. Interestingly, the keyword ‘cadherin’ was enriched in both males and females candidate loci. Cadherins are transmembrane proteins that mediate adhesive junctions between cells and play an important role during morphogenesis, tissue formation and integrity. Cadherin and Wnt/β-catenin signaling are pathways well known to control osteoblastogenesis and bone formation (32).

Our approach underscores the utility of using pediatric cohorts, where environmental factors are less predominant, for bone gene discovery. Use of aBMD and BMC at a specific skeletal site, adjustment for height and stratification by gender further increased discovery power. The reasons for the observed sex differences remain to be investigated. The rapid bone accretion that occurs with onset and progression through puberty, and the emergence of sex differences in bone accretion during our age range are cause for considering sex differences in genetic determinants of bone accretion during growth.

In summary, we report a novel locus on 9p21.3 for pediatric BMC at the distal radius in males, and implicate the CPED1-WNT16-FAM3C locus in pediatric radius aBMD in females. Further functional characterization of these signals is required to elucidate the precise mechanism underlying these observations.

Materials and Methods

Study sample

Discovery cohort

The multi-center, multi-ethnic longitudinal study of BMDCS was established to determine norms for BMC and areal-BMD for US children aged 5–20 years old. Subjects were recruited from the Children's Hospital of Philadelphia (Philadelphia, PA, USA), Creighton University (Omaha, NE, USA), Columbia University (New York, NY, USA), the Children's Hospital of Los Angeles (Los Angeles, CA, USA) and the Cincinnati Children's Hospital Medical Center (Cincinnati, OH, USA).

The study design has been previously published elsewhere (16,33); in brief, both females and males were enrolled between 2002 and 2003, aged 6–15 and 6–16 years old, respectively. These subjects were followed for 6 years (up to 7 visits), involving an annual measurement. Furthermore, older (age 19 years) and younger (age 5 years) subjects were subsequently recruited between 2006 and 2007 and followed for 2 years (up to three visits), with an annual measurement in order to extend the reference percentiles from 5 to 20 years old. Enrollment criteria were established to identify research subjects with normal development, including healthy bones. The key criteria included: term birth (≥37 week gestation), birth weight >2.3 kg, no evidence of precocious or delayed puberty and height, weight or BMI within the 3rd to 97th percentile for age. Children were excluded if they had presented with multiple fractures (>two fractures if <10 years old or > three fractures if >10 years old). Further exclusion criteria included current or previous medication use or a medical condition known to influence bone health, and extended bed rest. As separate aBMD references curves were being generated for males and females, opposite sex siblings were included but same sex siblings were excluded. Study subjects were invited to participate in this genetic study by providing a blood or saliva sample on their final visit. Participants 18 years old and older provided written informed consent while consent was obtained from the parent or guardian of participants younger than 18 years old, who also provided assent. The Institutional Review Board of each Clinical Center approved this protocol.

Replication cohort

As a replication cohort, 486 additional Caucasian children aged 5–18 years old were subsequently enrolled for a one-time visit in the Creighton and Cincinnati centers. These subjects were specifically recruited, in a cross-sectional manner, for validation of any signals coming out of such genetic studies and were not involved in the BMDCS study itself. All study procedures were the same as for BMDCS. This research was approved by CHOP's Institutional Review Board (IRB# IRB 10–007544_AM10, Protocol Title: genome Wide Association Study of Bone Mineral Accretion in Childhood).

Genotyping

We performed high throughput genome-wide SNP genotyping using the Illumina Infinium™ II OMNI Express plus Exome BeadChip technology (Illumina, San Diego) at the Children's Hospital of Philadelphia Center for Applied Genomics, leveraging the same pipeline as previously described (34).

Bone densitometry

Hologic, Inc. (Bedford, MA, USA) bone densitometers (QDR4500A, QDR4500W, Delphi A and Apex models) were used to obtain DXA scans. Scans were adjusted for differences among devices and longitudinal drift in performance by cross-calibration of DXA devices and longitudinal calibration stability, using anthropomorphic spine and hip phantoms, and the Hologic whole body phantom. Forearm scans were acquired according to the manufacturer's guidelines for patient positioning. Hologic software version Discovery 12.3 was used for analysis of baseline scans and Apex 2.1 was used for follow-up scan analysis using the ‘compare’ feature at the DXA Core Laboratory (University of California, San Francisco). aBMD/BMC Z-scores were calculated using the BMDCS reference values (16) to account for increases and sex differences in aBMD/BMC during growth and development. aBMD/BMC Z-scores were also adjusted for height-for-age Z-scores to minimize potential confounding by skeletal size, as previously described (16).

Statistical methods

The PLINK software (v1.07) package (35) was employed to carry out quality control measures. Prior to imputation, we excluded individuals with incorrect gender assignments or whose gender could not be determined by genotype, and individuals with missing rate per person >5%. SNPs with a call rate <95%, and minor allele frequency <0.5% were removed. We did not filter by HWE due to population heterogeneity in out cohort. After quality control, 1885 individuals who were genotyped at 739 284 SNPs were available for analysis.

Genotypes were imputed to the 1000 Genome Phase I Integrated Release Version 3 reference panel (36). A two-step imputation process was performed using SHAPEIT for haplotype phasing and IMPUTE2 for imputation, yielding a total of 39 345 920 imputed SNPs. Imputed genotypes were only used where directly assayed genotypes were unavailable. SNPs with minor allele frequency <5% or INFO <0.4 were excluded from the analysis, yielding a total of 7 238 679 SNPs. The association between each SNP and the quantitative trait was assessed using a univariate linear mixed model as implemented in the software GEMMA (19) to account for population stratification and relatedness, using the Wald test and the restricted maximum likelihood estimate (REML) of β. We performed a sex-stratified analysis and included a covariate for visit site in the model. We also included a covariate for discovery or replication cohort when performing the analysis of the combined cohorts.

Estimation of genomic ancestral components

Autosomal genotyped SNPs of all samples were pruned with PLINK (35), so that no pair of SNPs within a window of 200 markers were in linkage disequilibrium (LD r2 = 0.05). Based on these ∼35 000 SNPs, we performed an estimation of the genetic ancestry components of each individual on basis of the maximum likelihood using the ADMIXTURE software package (22). This program models the probability of observed genotypes using ancestry proportions and ancestral population allele frequencies. The clustering method was set to group individuals in three ancestral populations (K = 3), corresponding to the expected main African, European and Asian ancestry components. Children were assigned to one of the three ancestry groups, based on their highest fraction of estimated ancestry (i.e. >0.50) proportions. In case none of the three ancestral populations reached this proportion, the subject was excluded from further analyses. Children were assigned to one of the three particular ethnic groups based on the major component of genetic ancestry using the HapMap Phase 2 population labels as reference.

eQTL analyses

We assessed whether variants in the identified loci were involved in the regulation of messenger RNA levels via expression quantitative trait loci (eQTLs) using publicly available genome-wide expression datasets (23). We queried three tissue types (adipose, lymphoblastoid cell lines and skin) from 856 healthy females twins of the MuTHER resource (37), lymphoblastoid cell lines from 726 HapMap3 individuals (38) and three cell types (fibroblasts, lymphoblastoids and T-cells) derived from umbilical cords of 75 Geneva GenCord individuals (39).

GRAIL analyses

We explored the degree of functional similarity between SNPs with P < 10−5 in our GWAS and 169 genes reported to be associated with bone density and osteoporosis in adults using the text-mining tool, GRAIL [‘Gene Relationships Across Implicated Loci’ (24)]. Genes were extracted from the GWAS catalog (accessed 10-29-2014) without filtering on P-value, using the following keywords for Disease/Trait: ‘Bone mineral density’; ‘Bone mineral density (hip)’; ‘Bone mineral density (interaction)’; ‘Bone mineral density (spine)’; ‘Bone mineral density (wrist)’; ‘Bone properties (heel)’; ‘Femoral neck bone geometry’; ‘Femoral neck bone geometry and menarche (age at onset)’; ‘Hip bone size’; ‘Obesity and osteoporosis’; ‘Osteoporosis’; ‘Osteoporosis-related phenotypes’; ‘Spine bone size’. The search yielded 233 SNPs that mapped to 169 genes.

We mapped the candidate SNPs from our GWAS (232 for females and 148 for males) to their two nearest genes, and used these gene lists as input for the GRAIL analyses (based on the most recent PubMed abstracts, published up to October 2014).

Supplementary Material

Supplementary Material is available at HMG online.

Funding

This work was supported by the National Institutes of Health (grant numbers HD58886, HD076321); the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) contracts (N01-HD-1-3228, -3329, -3330, -3331, -3332, -3333); and the National Institutes of Health (NIH) Clinical and Translational Science Award (CTSA) program (Grant 8 UL1 TR000077).

Supplementary Material

Supplementary Data
supp_24_17_5053__index.html (1.1KB, html)

References

  • 1.US Department of Health and Human Services (2004) Bone Health and Osteoporosis: A Report of the Surgeon General. Rockville, MD: US Department of Health and Human Services, Office of the Surgeon General, Washington DC. [Google Scholar]
  • 2.Heaney R.P., Abrams S., Dawson-Hughes B., Looker A., Marcus R., Matkovic V., Weaver C. (2000) Peak bone mass. Osteoporos. Int., 11, 985–1009. [DOI] [PubMed] [Google Scholar]
  • 3.Welter D., MacArthur J., Morales J., Burdett T., Hall P., Junkins H., Klemm A., Flicek P., Manolio T., Hindorff L. et al. (2014) The NHGRI GWAS Catalog, a curated resource of SNP-trait associations. Nucleic Acids Res., 42, D1001–D1006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Estrada K., Styrkarsdottir U., Evangelou E., Hsu Y.H., Duncan E., Ntzani E., Oei L., Albagha O., Amin N., Kemp J. et al. (2012) Genome-wide meta-analysis identifies 56 bone mineral density loci and reveals 14 loci associated with risk of fracture. Nat. Genet., 44, 491–501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Richards J.B., Zheng H.F., Spector T.D. (2012) Genetics of osteoporosis from genome-wide association studies: advances and challenges. Nat. Rev. Genet., 13, 576–588. [DOI] [PubMed] [Google Scholar]
  • 6.Medina-Gomez C., Kemp J.P., Estrada K., Eriksson J., Liu J., Reppe S., Evans D.M., Heppe D.H., Vandenput L., Herrera L. et al. (2012) Meta-analysis of genome-wide scans for total body BMD in children and adults reveals allelic heterogeneity and age-specific effects at the WNT16 locus. PLoS Genet., 8, e1002718. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Timpson N.J., Tobias J.H., Richards J.B., Soranzo N., Duncan E.L., Sims A.M., Whittaker P., Kumanduri V., Zhai G., Glaser B. et al. (2009) Common variants in the region around Osterix are associated with bone mineral density and growth in childhood. Hum. Mol. Genet., 18, 1510–1517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Kemp J.P., Medina-Gomez C., Estrada K., St Pourcain B., Heppe D.H., Warrington N.M., Oei L., Ring S.M., Kruithof C.J., Timpson N.J. et al. (2014) Phenotypic dissection of bone mineral density reveals skeletal site specificity and facilitates the identification of novel loci in the genetic regulation of bone mass attainment. PLoS Genet., 10, e1004423. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Cummings S.R., Black D.M., Nevitt M.C., Browner W., Cauley J., Ensrud K., Genant H.K., Palermo L., Scott J., Vogt T.M. (1993) Bone density at various sites for prediction of hip fractures. The Study of Osteoporotic Fractures Research Group. Lancet, 341, 72–75. [DOI] [PubMed] [Google Scholar]
  • 10.Kroger H., Lunt M., Reeve J., Dequeker J., Adams J.E., Birkenhager J.C., Diaz Curiel M., Felsenberg D., Hyldstrup L., Kotzki P. et al. (1999) Bone density reduction in various measurement sites in men and women with osteoporotic fractures of spine and hip: the European quantitation of osteoporosis study. Calcif. Tissue Int., 64, 191–199. [DOI] [PubMed] [Google Scholar]
  • 11.Clark E.M., Ness A.R., Bishop N.J., Tobias J.H. (2006) Association between bone mass and fractures in children: a prospective cohort study. J. Bone Miner. Res., 21, 1489–1495. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Bae D.S. (2008) Pediatric distal radius and forearm fractures. J. Hand Surg. Am., 33, 1911–1923. [DOI] [PubMed] [Google Scholar]
  • 13.Barrett-Connor E., Sajjan S.G., Siris E.S., Miller P.D., Chen Y.T., Markson L.E. (2008) Wrist fracture as a predictor of future fractures in younger versus older postmenopausal women: results from the National Osteoporosis Risk Assessment (NORA). Osteoporos. Int., 19, 607–613. [DOI] [PubMed] [Google Scholar]
  • 14.Riggs B.L., Melton L.J. III (1986) Involutional osteoporosis. N. Engl. J. Med., 314, 1676–1686. [DOI] [PubMed] [Google Scholar]
  • 15.Mitchell J.A., Chesi A., Elci O., McCormack S.E., Kalkwarf H.J., Lappe J.M., Gilsanz V., Oberfield S.E., Shepherd J.A., Kelly A. et al. (2015) Genetics of bone mass in childhood and adolescence: effects of sex and maturation interactions. J. Bone Miner. Res., in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Zemel B.S., Kalkwarf H.J., Gilsanz V., Lappe J.M., Oberfield S., Shepherd J.A., Frederick M.M., Huang X., Lu M., Mahboubi S. et al. (2011) Revised reference curves for bone mineral content and areal bone mineral density according to age and sex for black and non-black children: results of the bone mineral density in childhood study. J. Clin. Endocrinol. Metab., 96, 3160–3169. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Crabtree N.J., Arabi A., Bachrach L.K., Fewtrell M., El-Hajj Fuleihan G., Kecskemethy H.H., Jaworski M., Gordon C.M. and International Society for Clinical, D. (2014) Dual-energy X-ray absorptiometry interpretation and reporting in children and adolescents: the revised 2013 ISCD Pediatric Official Positions. J. Clin. Densitom., 17, 225–242. [DOI] [PubMed] [Google Scholar]
  • 18.Zemel B.S., Leonard M.B., Kelly A., Lappe J.M., Gilsanz V., Oberfield S., Mahboubi S., Shepherd J.A., Hangartner T.N., Frederick M.M. et al. (2010) Height adjustment in assessing dual energy x-ray absorptiometry measurements of bone mass and density in children. J. Clin. Endocrinol. Metab., 95, 1265–1273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Zhou X., Stephens M. (2012) Genome-wide efficient mixed-model analysis for association studies. Nat. Genet., 44, 821–824. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Moverare-Skrtic S., Henning P., Liu X., Nagano K., Saito H., Borjesson A.E., Sjogren K., Windahl S.H., Farman H., Kindlund B. et al. (2014) Osteoblast-derived WNT16 represses osteoclastogenesis and prevents cortical bone fragility fractures. Nat. Med., 20, 1279–1288. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Zheng H.F., Tobias J.H., Duncan E., Evans D.M., Eriksson J., Paternoster L., Yerges-Armstrong L.M., Lehtimaki T., Bergstrom U., Kahonen M. et al. (2012) WNT16 influences bone mineral density, cortical bone thickness, bone strength, and osteoporotic fracture risk. PLoS Genet., 8, e1002745. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Alexander D.H., Novembre J., Lange K. (2009) Fast model-based estimation of ancestry in unrelated individuals. Genome Res., 19, 1655–1664. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Yang T.P., Beazley C., Montgomery S.B., Dimas A.S., Gutierrez-Arcelus M., Stranger B.E., Deloukas P., Dermitzakis E.T. (2010) Genevar: a database and Java application for the analysis and visualization of SNP-gene associations in eQTL studies. Bioinformatics, 26, 2474–2476. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Raychaudhuri S., Plenge R.M., Rossin E.J., Ng A.C., International Schizophrenia C., Purcell S.M., Sklar P., Scolnick E.M., Xavier R.J., Altshuler D. et al. (2009) Identifying relationships among genomic disease regions: predicting genes at pathogenic SNP associations and rare deletions. PLoS Genet., 5, e1000534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Anantharaman V., Aravind L. (2010) Novel eukaryotic enzymes modifying cell-surface biopolymers. Biol. Direct, 5, 1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Baron R., Kneissel M. (2013) WNT signaling in bone homeostasis and disease: from human mutations to treatments. Nat. Med., 19, 179–192. [DOI] [PubMed] [Google Scholar]
  • 27.Maeda K., Kobayashi Y., Udagawa N., Uehara S., Ishihara A., Mizoguchi T., Kikuchi Y., Takada I., Kato S., Kani S. et al. (2012) Wnt5a-Ror2 signaling between osteoblast-lineage cells and osteoclast precursors enhances osteoclastogenesis. Nat. Med., 18, 405–412. [DOI] [PubMed] [Google Scholar]
  • 28.Rivadeneira F., Styrkarsdottir U., Estrada K., Halldorsson B.V., Hsu Y.H., Richards J.B., Zillikens M.C., Kavvoura F.K., Amin N., Aulchenko Y.S. et al. (2009) Twenty bone-mineral-density loci identified by large-scale meta-analysis of genome-wide association studies. Nat. Genet., 41, 1199–1206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Valastyan S., Weinberg R.A. (2010) miR-31: a crucial overseer of tumor metastasis and other emerging roles. Cell Cycle, 9, 2124–2129. [DOI] [PubMed] [Google Scholar]
  • 30.Karnuth B., Dedy N., Spieker T., Lawlor E.R., Gattenlohner S., Ranft A., Dirksen U., Jurgens H., Brauninger A. (2014) Differentially expressed miRNAs in Ewing sarcoma compared to mesenchymal stem cells: low miR-31 expression with effects on proliferation and invasion. PLoS One, 9, e93067. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Camacho-Vanegas O., Camacho S.C., Till J., Miranda-Lorenzo I., Terzo E., Ramirez M.C., Schramm V., Cordovano G., Watts G., Mehta S. et al. (2012) Primate genome gain and loss: a bone dysplasia, muscular dystrophy, and bone cancer syndrome resulting from mutated retroviral-derived MTAP transcripts. Am. J. Hum. Genet., 90, 614–627. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Marie P.J., Hay E., Saidak Z. (2014) Integrin and cadherin signaling in bone: role and potential therapeutic targets. Trends Endocrinol. Metab., 25, 567–575. [DOI] [PubMed] [Google Scholar]
  • 33.Kalkwarf H.J., Zemel B.S., Gilsanz V., Lappe J.M., Horlick M., Oberfield S., Mahboubi S., Fan B., Frederick M.M., Winer K., Shepherd J.A. et al. (2007) The bone mineral density in childhood study: bone mineral content and density according to age, sex, and race. Clin. Endocrinol. Metab., 92, 2087–2099. [DOI] [PubMed] [Google Scholar]
  • 34.Hakonarson H.e.a. (2007) A genome-wide association study identifies KIAA0350 as a type 1 diabetes gene. Nature, 448, 591–594. [DOI] [PubMed] [Google Scholar]
  • 35.Purcell S., Neale B., Todd-Brown K., Thomas L., Ferreira M.A., Bender D., Maller J., Sklar P., de Bakker P.I., Daly M.J. et al. (2007) PLINK: a tool set for whole-genome association and population-based linkage analyses. Am. J. Hum. Genet., 81, 559–575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Pennisi E. (2010) Genomics. 1000 Genomes Project gives new map of genetic diversity. Science, 330, 574–575. [DOI] [PubMed] [Google Scholar]
  • 37.Grundberg E., Small K.S., Hedman A.K., Nica A.C., Buil A., Keildson S., Bell J.T., Yang T.P., Meduri E., Barrett A. et al. (2012) Mapping cis- and trans-regulatory effects across multiple tissues in twins. Nat. Genet., 44, 1084–1089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Stranger B.E., Montgomery S.B., Dimas A.S., Parts L., Stegle O., Ingle C.E., Sekowska M., Smith G.D., Evans D., Gutierrez-Arcelus M. et al. (2012) Patterns of cis regulatory variation in diverse human populations. PLoS Genet., 8, e1002639. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Dimas A.S., Deutsch S., Stranger B.E., Montgomery S.B., Borel C., Attar-Cohen H., Ingle C., Beazley C., Gutierrez Arcelus M., Sekowska M. et al. (2009) Common regulatory variation impacts gene expression in a cell type-dependent manner. Science, 325, 1246–1250. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_24_17_5053__index.html (1.1KB, html)
supp_ddv210_ddv210supp.docx (1.8MB, docx)

Articles from Human Molecular Genetics are provided here courtesy of Oxford University Press

RESOURCES