Hostname: page-component-6898cfcb55-cbk24 Total loading time: 0 Render date: 2026-11-11T12:06:00.812Z Has data issue: false hasContentIssue false
  • English
  • Français

Insights into the genetics of osteoporosis from recent genome-wide association studies

Published online by Cambridge University Press:  26 August 2011

Hou-Feng Zheng,
Timothy D. Spector and
J. Brent Richards
Hou-Feng Zheng
Affiliation:
Departments of Medicine, Human Genetics, Epidemiology and Biostatistics, Jewish General Hospital, McGill University, Montreal, Quebec, Canada.
Timothy D. Spector
Affiliation:
Department of Twin Research ad Genetic Epidemiology, King's College London, London, UK.
J. Brent Richards*
Affiliation:
Departments of Medicine, Human Genetics, Epidemiology and Biostatistics, Jewish General Hospital, McGill University, Montreal, Quebec, Canada.Department of Twin Research ad Genetic Epidemiology, King's College London, London, UK.
*
*Corresponding author: J. Brent Richards, 3755 Côte Ste-Catherine Road, Montreal, Québec, CanadaH3T 1E2. E-mail:brent.richards@mcgill.ca
Rights & Permissions [Opens in a new window]

Abstract

Osteoporosis, which is characterised by reduced bone mineral density (BMD) and an increased risk of fragility fractures, is the result of a complex interaction between environmental factors and genetic variants that confer susceptibility. Heritability studies have shown that BMD and other osteoporosis-related traits such as ultrasound properties of bone, skeletal geometry and bone turnover have significant inheritable components. Although previous linkage and candidate gene studies have provided few replicated loci for osteoporosis, genome-wide association approaches have produced clear and reproducible findings. To date, 20 genome-wide association studies (GWASs) for osteoporosis and related traits have been conducted, identifying dozens of genes. Further meta-analyses of GWAS data and deep resequencing of rare variants will uncover more novel susceptibility loci and ultimately provide possible therapeutic targets for fracture prevention.

Information

Type
Review Article
Information
Expert Reviews in Molecular Medicine , Volume 13 , March 2011, e28
Copyright
Copyright © Cambridge University Press 2011

Osteoporosis is a common skeletal disease characterised by low bone mineral density (BMD) and defects in the microarchitecture of bone, with a consequent increase in bone fragility and susceptibility to fracture (Ref. 1). The burden of osteoporosis on the healthcare system is increasing with the ageing of society (Ref. Reference Burge2), and current direct costs exceed $19 billion per year in the United States alone (Ref. Reference Organization3).

Osteoporosis is a common complex disease caused by the interplay of genetics and environmental factors, with adverse environmental exposures acting on a genetically susceptible individual to produce disease (Ref. Reference Chakravarti and Little4). Many factors influence the risk of osteoporosis, including diet, physical activity, medication use and coexisting diseases, but one of the most important clinical risk factors is a positive family history, emphasising the importance of genetics in its pathogenesis (Refs Reference Sigurdsson5, Reference Torgerson6). The allelic architecture of osteoporosis is likely to be multifactorial, with each imparting a relatively small effect, but their combined effect, as well as their interaction with environmental factors, probably leads to clinical disease (Ref. Reference Manolio and Collins7).

In the past decade, much progress has been made in reproducibly identifying genes that influence osteoporosis. Since 2007, we have experienced an acceleration in the identification of osteoporosis loci with the advent of large-scale, high-density genome-wide association studies (GWASs). GWASs, in which several hundred thousand to more than a million single-nucleotide polymorphisms (SNPs) are assayed in thousands of individuals, have been successfully used thus far to identify common genetic variants (those with a minor allele frequency greater than ~5%) that are associated with common human diseases (Ref. Reference Hindorff8).

Osteoporosis is defined clinically through the measurement of BMD, which is primarily evaluated by a dual-emission X-ray absorptiometry scan. Other osteoporosis-related phenotypes such as bone geometry and quantitative ultrasonography can also provide information on fracture prediction (Ref. Reference Kaptoge9) and assessment of the response to osteoporosis treatment (Refs Reference Greenspan10, Reference Uusi-Rasi11).

BMD is a highly heritable trait, with heritability estimates of 0.6–0.85 (Ref. Reference Ng12). Twin and family studies have shown that peak BMD, as well as BMD loss, is genetically determined (Refs Reference Slemenda13, Reference Smith14, Reference Gueguen15, Reference Krall and Dawson-Hughes16). Heritability studies have also confirmed that other phenotypes relevant to the pathogenesis of osteoporosis, such as ultrasound properties of bone and bone geometry, have significant heritable components (Ref. Reference Arden17). Segregation analysis in families has shown that regulation of BMD and other osteoporosis-related phenotypes is polygenic and determined by the effects of several genes, each with relatively small effects, rather than a small number of genes with large effects (Ref. Reference Gueguen15).

This article reviews the results of linkage and association studies for osteoporosis, with particular focus on the new findings revealed by GWASs since 2007, and discusses future directions for the dissection of the genetic determinants of osteoporosis in the post-GWAS era.

Linkage studies

Linkage analysis has long been used to identify disease-associated loci in inherited diseases. This approach relies on the principle of identity by descent combined with phenotypic information to identify loci conferring disease susceptibility. Linkage analysis is a powerful methodology to detect the causative gene in monogenic bone diseases; however, for complex diseases, the identification of susceptibility genes located within the linkage peak region has proved difficult, and the identified loci have not often been replicated for osteoporosis (Ref. Reference Richards18). More than a dozen genome-wide linkage scans have been performed on BMD and other osteoporosis-related phenotypes, including femoral neck (FN) geometry (Ref. Reference Peacock19), ultrasound properties of bone (Ref. Reference Wilson20) and bone loss. However, even a very large-scale meta-analysis of nine genome-wide linkage studies (combined n = 11 842) did not yield any genome-wide significant loci for BMD (Ref. Reference Ioannidis21). This probably reflects the fact that common variants that regulate BMD have modest effects that are difficult to detect reproducibly by conventional linkage analysis.

Candidate gene association studies

Before GWASs, candidate gene association studies (CGASs) were widely used to identify genes that plausibly have an important role in the disease process, and well-designed CGASs, in combination with fine mapping of known osteoporosis loci, remain an attractive and efficient way of identifying new susceptibility genes or variations until the costs of high-throughput sequencing decrease. Although CGASs are attractively simple in methodology, they have often produced spurious and nonreplicated results. This may be because the statistical power to detect these associations is limited unless causal SNPs or very good surrogates have been typed. Additionally, other factors such as differences in sample characteristics, including ethnicity, sex, age and menopausal status, might produce different results in different studies.

A recent CGAS attempted to overcome these difficulties by standardising genotyping and phenotype definitions across 19 195 European subjects in five international, multicentre population-based studies. This study (Ref. Reference Richards18) used genome-wide genotyping data, but the authors restricted their analysis to 150 candidate genes and 36 016 SNPs within these loci. They found evidence for association in only 9 of 150 (6%) candidate genes previously tested for their association with osteoporosis. Details of these loci can be found on the HUgeNet website (http://www.hugenavigator.net/). These results suggest that candidate genes tested for BMD rarely replicate in large consortia with standardised phenotyping and genotyping. Table 1 lists the candidate genes for osteoporosis.

Table 1.

Promising candidate genes in osteoporosis

Genome-wide association studies

GWASs represent a powerful and efficient study design for investigating the genetic architecture of complex diseases arising from common base-pair variants (Refs Reference Hardy and Singleton50, Reference Manolio51). GWASs use dense maps of common SNPs (minor allele frequency >~5%) that cover the human genome to look for allele-frequency differences between cases and controls (diseases study design) or genotype variation, which can explain trait variation (quantitative traits study design). A significant observation in the variant site is taken to indicate that the corresponding region of the genome contains functional DNA-sequence variants that influence the disease or trait in question (Ref. Reference Kruglyak52). GWASs have proved to be a powerful approach in screening the susceptibility genes (loci) for common diseases. An updated list of published GWASs can be found at the National Cancer Institute–National Human Genome Research Institute catalogue of published GWASs (http://www.genome.gov/26525384). A major advantage of GWASs over candidate gene studies is that they offer the possibility of identifying novel susceptible genes and pathways.

In 2007, Kiel et al. (Ref. Reference Kiel53) published the first GWAS of osteoporosis. In this study, they used the early and low-density commercial Affymetrix GeneChip and genotyped ~100 000 SNPs in 1141 Framingham Heart Study subjects, and identified 40 SNPs that could potentially be associated with several bone phenotypes. However, owing to sample size limitations and other factors, none of the P-values exceeded the threshold of genome-wide significance. In 2008, two GWASs (Refs Reference Richards54,Reference Styrkarsdottir55) reported nine genome-wide significant hits for BMD. In these two studies, more than 300 000 SNPs were genotyped in much larger cohorts. The following 2 years (2009 and 2010) witnessed a deluge of GWASs conducted on osteoporosis and related phenotypes (Refs Reference Guo38, Reference Zhao56, Reference Paternoster57, Reference Kung58, Reference Koller59, Reference Karasik60, Reference Hsu61, Reference Guo62, Reference Xiong63, Reference Timpson64, Reference Styrkarsdottir65, Reference Rivadeneira66, Reference Liu67, Reference Cho68, Reference Liu86, Reference Duncan92, Reference Kou93).

GWASs for osteoporosis and related phenotypes validated several candidate genes for osteoporosis, including LRP5 (Refs Reference Richards54, Reference Rivadeneira66), ESR1 (Refs Reference Styrkarsdottir55, Reference Styrkarsdottir65, Reference Rivadeneira66), SOST (Ref. Reference Styrkarsdottir65), TNFRSF11B (Refs Reference Richards54, Reference Styrkarsdottir55, Reference Rivadeneira66), TNFSF11 (Refs Reference Styrkarsdottir55, Reference Paternoster57, Reference Styrkarsdottir65, Reference Rivadeneira66, Reference Cirulli and Goldstein91), TNFRSF11A (Refs Reference Styrkarsdottir55, Reference Rivadeneira66), PTH (Ref. Reference Guo38) and FOXC2 (Ref. Reference Rivadeneira66). Novel loci that have been identified as being associated with osteoporosis and related phenotypes by recent original GWASs and subsequent meta-analysis of GWAS data are discussed in this section (Fig. 1). Note that the SNPs from each study are the reported SNPs either near or within the genes; however, they probably do not represent the causal variants.

Figure 1.

Illustration of novel hits revealed by genome-wide association studies for osteoporosis and related traits. Novel bone mineral density loci are in blue, new loci for other osteoporosis-related phenotypes are in yellow and candidate loci validated by GWASs are in green. Please refer to Table 1 and the main text for full gene names and references.

BMD loci revealed by GWASs

JAG1

rs2273061, located near the JAG1 (jagged 1) locus on 20p12.2, showed significant association with spine BMD (P = 5 × 10−8) in a GWAS carried out on Southern Chinese subjects in Hong Kong and replicated in other ethnic groups (Ref. Reference Kung58). The study also confirmed association of the high BMD-related allele G of rs2273061 with higher JAG1 mRNA expression in both human bone-derived cells and peripheral blood mononuclear cells. The function of JAG1 seems related to an increase in bone mineral deposition (Ref. Reference Nobta69). JAG1 protein was detected in trabecular osteoblasts and endosteal osteoblasts, and its expression was increased with intermittent parathyroid hormone treatment (Ref. Reference Weber70).

IL21R

Guo et al. (Ref. Reference Guo38) performed a GWAS for FN BMD in a discovery sample consisting of 983 unrelated white subjects and replicated the top 175 SNPs (with P < 5 × 10−4) in a family-based sample of 2557 white subjects. Three IL21R (interleukin 21 receptor) SNPs, rs8057551, rs8061992 and rs7199138, showed consistent association with FN BMD in discovery and replication samples, with combined P-values of 2.31 × 10−6, 8.62 × 10−6 and 1.41 × 10−5, respectively, in the total sample. IL21R is a cytokine receptor that is important to bone biology and has been identified negatively to be correlated with the destruction of cartilage and bone (Ref. Reference Jungel71).

ALDH7A1

This susceptibility gene was first found to be associated with osteoporotic fractures (OFs) in a case–control GWAS in 700 elderly Chinese Han subjects, with a follow-up replication study in an independent Chinese sample containing 390 cases with hip OFs and 516 controls (Ref. Reference Guo62). SNP rs13182402 within the ALDH7A1 (aldehyde dehydrogenase seven family, member A1) gene on chromosome 5q31 was strongly associated with OFs (P = 2.08 × 10−9, odds ratio = 2.25). This SNP was confirmed as consistently associated with hip BMD in both Chinese subjects and Caucasians (P = 6.39 × 10−6). The ALDH7A1 gene encodes an enzyme of the acetaldehyde dehydrogenase superfamily, which degrades and detoxifies acetaldehyde generated by alcohol metabolism. Acetaldehyde has been shown to inhibit osteoblast proliferation and to decrease bone formation (Ref. Reference Giuliani72). Another member of the acetaldehyde dehydrogenase family, ALDH2, has also been found to be significantly associated with osteoporosis (Ref. Reference Yamaguchi73).

SP7

The SP7 (also known as OSTERIX) locus on 12q13 was first found to be associated with spine BMD (Ref. Reference Styrkarsdottir65). Common SNPs near the SP7 region were also associated with total body BMD in a GWAS in children (Ref. Reference Timpson64). SNPs in this region were genome-wide significant in the GEFOS (genetic factors for osteoporosis) meta-analysis (Ref. Reference Rivadeneira66). SP7 encodes a transcription factor that has an essential role in regulating osteoblast differentiation (Ref. Reference Baek74).

ADAMTS18 and TGFBR3

These two susceptible genes were identified by a GWAS performed in different ethnic groups (Ref. Reference Xiong63). One thousand unrelated white US subjects and 1972 subjects from white US pedigrees were genotyped, along with a Chinese hip fracture sample comprising 350 cases and controls, a Chinese BMD sample with 2955 subjects and a Tobago cohort of 908 males of African ancestry as a replication sample. ADAMTS18 (ADAM metallopeptidase with thrombospondin type 1 motif, 18) and TGFBR3 (transforming growth factor, beta receptor III) were significantly associated with BMD variation in the three major ethnic groups. They also provided in silico replication by using publicly available Framingham GWAS data (2953 whites). In addition, ADAMTS18 variants were also found to be associated with hip fracture.

The allele change from T to C of SNP rs16945612 in ADAMTS18 produced a new TEL2 (a member of the E26 transformation-specific family of transcription factors) binding site in ADAMTS18. TEL2 represses two genes, BMP6 and RARa, which are involved in regulating osteoblast differentiation and bone remodelling (Ref. Reference Gu75). Electrophoretic mobility-shift analysis confirmed the potential changes of TEL2 binding to ADAMTS18 caused by rs16945612. TGFBR3 appears to modulate the biological function of BMP2 (bone morphogenetic protein 2) (Ref. Reference Kirkbride76), and BMP2 has key roles in bone biology and is significantly associated with BMD and other bone phenotypes (Ref. Reference Styrkarsdottir77). Tgfbr3 double-null mice suffer severe abnormal skeleton defects (Ref. Reference Stenvers78).

ZBTB40

SNP rs7524102 in the 1p36 region, previously implicated in the genetic regulation of BMD by linkage analysis in families (Refs Reference Devoto79, Reference Devoto80), was identified as a potential locus for regulation of both hip and spine BMD by a GWAS with a significant association signal (P = 5 × 10−16) (Ref. Reference Styrkarsdottir55). This association was confirmed by other GWASs (Ref. Reference Styrkarsdottir65, Reference Rivadeneira66, Reference Cirulli and Goldstein91). In the region, ZBTB40 (zinc finger and BTB domain containing 40), which is located 80 kb downstream from the signal, may be the susceptible gene. ZBTB40 is expressed in bone; however, the function of this gene is as yet unknown and requires more functional studies to elucidate its role in bone physiology.

SOX6

By genotyping 380 000 SNPs in 1000 homogeneous unrelated Caucasians (female n = 501, male n = 499), Liu and colleagues (Ref. Reference Liu67) performed the first bivariate GWAS of obesity and osteoporosis. SNPs rs297325 and rs4756846 in intron 1 of the SOX6 (sex-determining region Y-box 6) gene were found to be associated with hip BMD in the male subjects. These findings were further validated in the 1370 male subjects of the Framingham Heart Study cohort. The SOX6 locus was also identified as the FN BMD locus by a meta-analysis (Ref. Reference Rivadeneira66). SOX6 is a member of the SOX gene family and was previously reported to have an important role in both cartilage formation (chondrogenesis) and obesity-related insulin resistance. Animal model studies show that Sox6 single-null mice present mild skeletal abnormalities, and Sox6-knockout mice fetuses die with a severe, generalised chondrodysplasia (Ref. Reference Smits81).

FAM3C and SFRP4

These two loci were identified in a GWAS in Asian populations (Ref. Reference Cho68). By genotyping 352 228 SNPs in 8842 subjects and involving eight quantitative traits, this study found that FAM3C (family with sequence similarity 3, member C) and SFRP4 (secreted frizzled-related protein 4) were associated with BMD. A FAM3C SNP rs7776725 on chromosome 7q31 was found to be associated with BMD at the radius (P = 1.0 × 10−11), tibia (P = 1.6 × 10−6) and heel (P = 1.9 × 10−10). The function of FAM3C gene in bone biology is unclear to date. SNP rs1721400, mapping close to SFRP4, on chromosome 7p14 was consistently associated with BMD at the three sites mentioned above (P = 2.2 × 10−3, 1.4 × 10−7 and 6.0  ×  10−4, respectively). SFRP4 is a member of the SFRP family, which acts as soluble modulators of Wnt signalling (Ref. Reference Cho82). Moreover, overexpression of Sfrp4 in mice has been associated with a reduction in bone density (Refs Reference Nakanishi83, Reference Nakanishi84).

MARK3

The MARK3 (MAP/microtubule affinity-regulating kinase 3) gene on chromosome 14q32.3 was significantly associated with total hip BMD (rs2010281, located in intron 1 of MARK3, had a P-value of 1.8 × 10−9) (Ref. Reference Styrkarsdottir65). This SNP was also significantly associated with FN BMD in a meta-analysis (Ref. Reference Rivadeneira66). The MARK3 gene encodes mitogen-activated protein/microtubule affinity-regulating kinase 3, a member of the adenosine monophosphate-activated protein kinase superfamily of proteins. MARK3 might be involved in cell cycle regulation, and alterations in MARK3 might lead to carcinogenesis. The function on the regulation of BMD needs to be explored further.

MHC region

Styrkarsdottir and colleagues (Ref. Reference Styrkarsdottir55) first found that SNP rs3130340 in the MHC (major histocompatibility complex) region was associated with spine BMD (P = 1.2 × 10−7). The SNP is located downstream of the uncharacterised chromosome 6 open reading frame 10 gene C6orf10. The association significance of this SNP was also confirmed by meta-analysis (Ref. Reference Rivadeneira66). The MHC locus on chromosome 6p21 is associated with a number of immune-related diseases; however, the precise mechanism responsible for osteoporosis in this region is unknown.

LRP4, GPR177 and CTNNB1

Three genes, all involved in the Wnt/β-catenin signalling pathway, that is, LRP4 (low-density lipoprotein receptor-related protein 4), GPR177 (G-protein-coupled receptor 177) and CTNNB1 (catenin beta 1), were found to be strongly associated with BMD. This pathway has an essential role in the regulation of bone mass. Two SNPs, rs2306033 and rs7935346, which are within or close to the LRP4 locus at 11p11.2, were found to be associated with hip BMD, but were not genome-wide significant (Ref. Reference Styrkarsdottir55). This association was further studied by a meta-analysis of five GWASs (Ref. Reference Rivadeneira66) and a collaborative meta-analysis (Ref. Reference Richards18) and confirmed that SNPs within the LRP4 locus were associated with FN BMD.

Two common SNPs, rs1430742 and rs2566755, were associated with both FN and lumbar spine (LS) BMD (Ref. Reference Rivadeneira66). The two SNPs are located within an intron of GPR177 (G-protein-coupled receptor 177) on chromosome 1p31.3. The study also showed that SNP rs87939, which is located 103 kb upstream of the CTNNB1 gene, was associated with FN BMD (Ref. Reference Rivadeneira66).

MEF2C, CRHR1, DCDC5, FLJ42280 and STARD3NL

These five novel genes were identified with genome-wide significance in a very large-scale meta analysis of five GWASs (Ref. Reference Rivadeneira66). The meta-analysis identified eight genome-wide significant loci. (In addition to the five genes mentioned above, another three genes – GRP177, CTNNB1 and SOX6 – are described earlier in this paper.) The study also reported eight known loci and four suggestive loci for BMD.

New loci for other osteoporosis-related phenotypes

TBC1D8, OSBPL1A and RAP1A

An integration of the GWAS for osteoporosis-related traits, including BMD, femoral neck-shaft angle (NSA), femoral neck length and narrow-neck width (NW), was performed in 7633 Caucasian women and 3657 men (Ref. Reference Hsu61). SNP rs2278729 located in intron 4 of TBC1D8 (TBC1 domain family, member 8) on chromosome 2q11.2, and SNP rs7227401 located in intron 4 of OSBPL1A (18q11.2, oxysterol-binding protein-like 1A) were significantly associated with NSA (P = 1.48 × 10−7) and NW (P = 4.22 × 10−7) in men. SNP rs494453, located in intron 2 of RAP1A (member of RAS oncogene family) on chromosome 1p13.2, was found to be associated with NW (P = 2.80 × 10−7) in women, and the association became more significant when analysing women and men together (P = 3.6 × 10−8). A gene expression study showed that rs2278729, allele A, is associated with smaller NSA in men and with a lower expression of TBC1D8 transcript (Ref. Reference Hsu61).

RTP3

A GWAS for FN bone geometry was performed by analysing 379 000 SNPs in 1000 Caucasians and was replicated in 1488 independent Caucasians and 2118 Chinese subjects (Ref. Reference Zhao56). SNP rs7430431 in the RTP3 (receptor transporter protein 3) gene on chromosome 3p21 was found to be significantly associated with buckling ratio and femoral cortical thickness (CT). Region 3p21 was previously reported to be linked with CT (LOD = 2.19, P = 0.0006) in 3998 subjects from 434 pedigrees (Ref. Reference Xiong85). RTP3 is a newly identified gene and its function in regulating bone geometry is still not known.

PLCL1

Four SNPs in the PLCL1 (phospholipase C-like 1) gene were found, in a GWAS performed for hip bone size (Ref. Reference Liu86), to be associated with hip bone size by testing 380 000 SNPs in 1000 homogeneous unrelated Caucasians. The association evidence of the region was validated both in an independent UK cohort comprising 1216 Caucasian females and in a Chinese sample with 403 females. The study (Ref. Reference Liu86) also suggested the association of the PLCL1 gene with hip fractures.

Perspectives

The study of common base-pair variants has provided evidence that osteoporosis is determined by a large number of common variants, each imparting a modest effect. To date, several dozen susceptibility genes have been identified. Promising genes with known function, confirmed by CGASs or GWASs, can now be classified into four biological pathways: the vitamin D endocrine pathway (VDR, DBP), the oestrogen endocrine pathway (ESR1, ESR2, CYP19A1), the Wnt–β-catenin signalling pathway (LRP5, SOST, WNT10B, SFRP1, FOXC2, LRP4, GPR177, CTNNB1) and the RANKL–RANK–OPG pathway (TNFSF11, TNFRSF11A, TNFRSF11B). New genes identified by recent GWASs are expected to reveal new biological pathways as their biological functions are carefully explained through functional studies. Although GWASs have proved to be a reliable approach for investigating the genetic basis of osteoporosis, emerging strategies, such as applications of next-generation resequencing technologies, are required to better understand the majority of heritability, which still has to be accounted for despite these advances.

Although the identification of susceptibility genes for osteoporosis by recent GWASs highlights contributions of common variants, their interacting effects have not been well described. However, note that a significant interaction between ESR1, ESR2, IGF-I and osteoporotic fracture (OF) and other phenotypes, including BMD and aspects of FN structure, has been reported in women (Ref. Reference Rivadeneira30).

Because the above-described GWASs were carried out in different populations with varying sample size, meta-analytic efforts remain an efficient way of combining signals across tens of thousands of genome-wide genotyped individuals, with follow-up genotyping in an equally well-powered sample. This is especially important when the effect of any common variant on osteoporosis risk is small and the statistical power to identify reproducible signals is insufficient in each GWAS. Such studies will yield more insights into this polygenic trait. For example, a recent meta-analysis of five GWASs of BMD with a sample size of 19 195 identified 13 novel loci and confirmed seven known loci (Ref. Reference Rivadeneira66).

The most useful aspect of GWASs is that they are able to identify entirely novel proteins that previously had no known role in bone disease. The above GWASs have identified many such loci, but an explanation of the exact mechanism whereby they exert their effect has yet to be described in most cases. Thus further functional experiments interrogating these loci are warranted.

It is now apparent that the variance explained by identified SNPs through GWASs is very low. For example, in one study, the top 15 SNPs associated with LS BMD combined explained ~2.9% of the variance of the phenotype, and the top 10 SNPs associated with FN BMD explained ~1.9% of the variance (Ref. Reference Rivadeneira66). Given the high heritability of BMD and the low variance explained by common variants, it is reasonable to assume that there are other sources of variation in the genome that account for this missing heritability.

Rare base-pair variants with a frequency less than 1% are one such source of variation. With the emergence of affordable resequencing technologies, the ability to identify and accurately measure such rare variants is now possible. Rare variants are also more likely to impart large risks for common disease (Ref. Reference Bodmer and Bonilla87). The study of the functional implications of rare causal variants has already led to an improved understanding of the aetiology of common forms of disease. For example, fine mapping of families with a rare form of localised bone remodelling a decade ago highlighted TNFRSF11A as a regulator of bone turnover (Ref. Reference Hughes88). Characterisation of this pathway led to the development of a monoclonal antibody against the protein, which when tested in large Phase III trials effectively treated osteoporosis (Ref. Reference McClung89) and prevented fractures in the general population (Ref. Reference Cummings90). We and others (Ref. Reference Cirulli and Goldstein91) suggest that the interrogation of rare base-pair variants will lead to important and biologically tractable targets for osteoporosis therapy.

Genetic variation in identified genes might also help us understand whether certain drugs are best suited for individuals with a certain genotype through pharmacogenetic studies. Although such studies are still in their infancy, they hold promise for identification of groups of individuals better suited for particular interventions. Furthermore, studies might now be undertaken to understand whether the side effects of osteoporosis therapies can be assigned to individuals harbouring certain genotypes. These two avenues of research provide potential ways of improving osteoporosis patient care through genetic information.

The final aim of genetic studies, such as those described above, is much more than the identification of susceptibility variants. First, identified genes will provide therapeutic targets after thorough functional studies. Second, efforts are also required to understand whether information on genotype profile can improve the prediction of OF risk. However, the small effects identified by GWASs arising from common variants suggest that the second aim will be difficult to fulfil. Although these studies are ongoing, the description of a large set of common base-pair variants reproducibly associated with osteoporosis represents a large step forward in understanding the genetic and biological basis of this disease.

Since submission of this manuscript, two additional GWASs have been published. Duncan and co-workers (Ref. Reference Duncan92) identified four new genes, GALNT3, IBSP, RSPO3 and SOX4, showing genome-wide significant association with BMD in a GWAS that focused on 1955 age-specific women, and with replication in 20,898 women from the general population. Kou and colleagues (Ref. Reference Kou93) conducted a GWAS with 190 osteoporosis cases and 1557 controls followed by replication of 2092 cases and 3114 controls in a Japanese population, they reported that an SNP in the FONG gene on chromosome 2q33.1 was associated with osteoporosis.

Acknowledgements and Funding

We thank the Canadian Institutes of Health Research (CIHR), the Jewish General Hospital, the Lady Davis Institute and the Fonds de la Recherche en Sante du Québec for funding this research, and the peer reviewers for their thorough and helpful review of this manuscript.

References

1Noauthorslist (1993) Consensus development conference: diagnosis, prophylaxis, and treatment of osteoporosis. American Journal of Medicine 94, 646-650CrossRefGoogle Scholar
2Burge, R. et al. (2007) Incidence and economic burden of osteoporosis-related fractures in the United States, 2005–2025. Journal of Bone and Mineral Research 22, 465-475CrossRefGoogle ScholarPubMed
3Organization, W.H. (2007) WHO scientific group on the assessment of osteoporosis at primary health care level. Summary Meeting Report (Brussels, Belgium, 2007)Google Scholar
4Chakravarti, A. and Little, P. (2003) Nature, nurture and human disease. Nature 421, 412-414CrossRefGoogle ScholarPubMed
5Sigurdsson, G. et al. (2008) Impact of genetics on low bone mass in adults. Journal of Bone and Mineral Research 23, 1584-1590CrossRefGoogle ScholarPubMed
6Torgerson, D.J. et al. (1996) Prediction of perimenopausal fractures by bone mineral density and other risk factors. Journal of Bone and Mineral Research 11, 293-297CrossRefGoogle ScholarPubMed
7Manolio, T.A. and Collins, F.S. (2007) Genes, environment, health, and disease: facing up to complexity. Human Heredity 63, 63-66CrossRefGoogle ScholarPubMed
8Hindorff, L.A. et al. (2009) Potential etiologic and functional implications of genome-wide association loci for human diseases and traits. Proceedings of the National Academy of Sciences of the United States of America 106, 9362-9367CrossRefGoogle ScholarPubMed
9Kaptoge, S. et al. (2008) Prediction of incident hip fracture risk by femur geometry variables measured by hip structural analysis in the study of osteoporotic fractures. Journal of Bone and Mineral Research 23, 1892-1904CrossRefGoogle ScholarPubMed
10Greenspan, S.L. et al. (2005) Effect of hormone replacement, alendronate, or combination therapy on hip structural geometry: a 3-year, double-blind, placebo-controlled clinical trial. Journal of Bone and Mineral Research 20, 1525-1532CrossRefGoogle ScholarPubMed
11Uusi-Rasi, K. et al. (2005) Effects of teriparatide [rhPTH (1–34)] treatment on structural geometry of the proximal femur in elderly osteoporotic women. Bone 36, 948-958CrossRefGoogle ScholarPubMed
12Ng, M.Y. et al. (2006) Effect of environmental factors and gender on the heritability of bone mineral density and bone size. Annals of Human Genetics 70, 428-438CrossRefGoogle ScholarPubMed
13Slemenda, C.W. et al. (1996) The genetics of proximal femur geometry, distribution of bone mass and bone mineral density. Osteoporosis International 6, 178-182CrossRefGoogle ScholarPubMed
14Smith, D.M. et al. (1973) Genetic factors in determining bone mass. Journal of Clinical Investigation 52, 2800-2808CrossRefGoogle ScholarPubMed
15Gueguen, R. et al. (1995) Segregation analysis and variance components analysis of bone mineral density in healthy families. Journal of Bone and Mineral Research 10, 2017-2022CrossRefGoogle ScholarPubMed
16Krall, E.A. and Dawson-Hughes, B. (1993) Heritable and life-style determinants of bone mineral density. Journal of Bone and Mineral Research 8, 1-9CrossRefGoogle ScholarPubMed
17Arden, N.K. et al. (1996) The heritability of bone mineral density, ultrasound of the calcaneus and hip axis length: a study of postmenopausal twins. Journal of Bone and Mineral Research 11, 530-534CrossRefGoogle ScholarPubMed
18Richards, J.B. et al. (2009) Collaborative meta-analysis: associations of 150 candidate genes with osteoporosis and osteoporotic fracture. Annals of Internal Medicine 151, 528-537CrossRefGoogle ScholarPubMed
19Peacock, M. et al. (2005) Sex-specific quantitative trait loci contribute to normal variation in bone structure at the proximal femur in men. Bone 37, 467-473CrossRefGoogle ScholarPubMed
20Wilson, S.G. et al. (2004) A genome-screen of a large twin cohort reveals linkage for quantitative ultrasound of the calcaneus to 2q33-37 and 4q12-21. Journal of Bone and Mineral Research 19, 270-277CrossRefGoogle ScholarPubMed
21Ioannidis, J.P. et al. (2007) Meta-analysis of genome-wide scans provides evidence for sex- and site-specific regulation of bone mass. Journal of Bone and Mineral Research 22, 173-183CrossRefGoogle ScholarPubMed
22Mullin, B.H. et al. (2008) Identification of a role for the ARHGEF3 gene in postmenopausal osteoporosis. American Journal of Human Genetics 82, 1262-1269CrossRefGoogle ScholarPubMed
23Yazdanpanah, N. et al. (2007) The -1997 G/T and Sp1 polymorphisms in the collagen type I alpha1 (COLIA1) gene in relation to changes in femoral neck bone mineral density and the risk of fracture in the elderly: the Rotterdam study. Calcified Tissue International 81, 18-25CrossRefGoogle ScholarPubMed
24Grant, S.F. et al. (1996) Reduced bone density and osteoporosis associated with a polymorphic Sp1 binding site in the collagen type I alpha 1 gene. Nature Genetics 14, 203-205CrossRefGoogle ScholarPubMed
25Riancho, J.A. et al. (2009) Association of the aromatase gene alleles with BMD: epidemiological and functional evidence. Journal of Bone and Mineral Research 24, 1709-1718CrossRefGoogle ScholarPubMed
26Fang, Y. et al. (2009) Vitamin D binding protein genotype and osteoporosis. Calcified Tissue International 85, 85-93CrossRefGoogle ScholarPubMed
27Wang, C.L. et al. (2007) Association of estrogen receptor alpha gene polymorphisms with bone mineral density in Chinese women: a meta-analysis. Osteoporosis International 18, 295-305CrossRefGoogle ScholarPubMed
28Lai, B.M. et al. (2008) Estrogen receptor alpha CA dinucleotide repeat polymorphism is associated with rate of bone loss in perimenopausal women and bone mineral density and risk of osteoporotic fractures in postmenopausal women. Osteoporosis International 19, 571-579CrossRefGoogle ScholarPubMed
29Sano, M. et al. (1995) Association of estrogen receptor dinucleotide repeat polymorphism with osteoporosis. Biochemical and Biophysical Research Communications 217, 378-383CrossRefGoogle ScholarPubMed
30Rivadeneira, F. et al. (2006) Estrogen receptor beta (ESR2) polymorphisms in interaction with estrogen receptor alpha (ESR1) and insulin-like growth factor I (IGF1) variants influence the risk of fracture in postmenopausal women. Journal of Bone and Mineral Research 21, 1443-1456CrossRefGoogle ScholarPubMed
31Wilson, S.G. et al. (2009) Common sequence variation in FLNB regulates bone structure in women in the general population and FLNB mRNA expression in osteoblasts in vitro. Journal of Bone and Mineral Research 24, 1989-1997CrossRefGoogle ScholarPubMed
32Yamada, Y., Ando, F. and Shimokata, H. (2006) Association of polymorphisms in forkhead box C2 and perilipin genes with bone mineral density in community-dwelling Japanese individuals. International Journal of Molecular Medicine 18, 119-127Google ScholarPubMed
33Yerges, L.M. et al. (2009) High-density association study of 383 candidate genes for volumetric BMD at the femoral neck and lumbar spine among older men. Journal of Bone and Mineral Research 24, 2039-2049CrossRefGoogle ScholarPubMed
34Lee, H.J. et al. (2007) Polymorphisms and haplotypes of integrinalpha1 (ITGA1) are associated with bone mineral density and fracture risk in postmenopausal Koreans. Bone 41, 979-986CrossRefGoogle ScholarPubMed
35Urano, T. et al. (2004) Association of a single-nucleotide polymorphism in low-density lipoprotein receptor-related protein 5 gene with bone mineral density. Journal of Bone and Mineral Metabolism 22, 341-345CrossRefGoogle ScholarPubMed
36Riancho, J.A., Valero, C. and Zarrabeitia, M.T. (2006) MTHFR polymorphism and bone mineral density: meta-analysis of published studies. Calcified Tissue International 79, 289-293CrossRefGoogle ScholarPubMed
37Tenne, M. et al. (2008) Genetic variation in the PTH pathway and bone phenotypes in elderly women: evaluation of PTH, PTHLH, PTHR1 and PTHR2 genes. Bone 42, 719-727CrossRefGoogle ScholarPubMed
38Guo, Y. et al. (2010) IL21R and PTH may underlie variation of femoral neck bone mineral density as revealed by a genome-wide association study. Journal of Bone and Mineral Research 25, 1042-1048CrossRefGoogle ScholarPubMed
39Mullin, B.H. et al. (2009) Further genetic evidence suggesting a role for the RhoGTPase-RhoGEF pathway in osteoporosis. Bone 45, 387-391CrossRefGoogle ScholarPubMed
40Sims, A.M. et al. (2008) Genetic analyses in a sample of individuals with high or low BMD shows association with multiple Wnt pathway genes. Journal of Bone and Mineral Research 23, 499-506CrossRefGoogle ScholarPubMed
41Uitterlinden, A.G. et al. (2004) Polymorphisms in the sclerosteosis/van Buchem disease gene (SOST) region are associated with bone-mineral density in elderly whites. American Journal of Human Genetics 75, 1032-1045CrossRefGoogle ScholarPubMed
42Balemans, W. et al. (2002) Lack of association between the SOST gene and bone mineral density in perimenopausal women: analysis of five polymorphisms. Bone 31, 515-519CrossRefGoogle ScholarPubMed
43Hsu, Y.H. et al. (2006) Variation in genes involved in the RANKL/RANK/OPG bone remodeling pathway are associated with bone mineral density at different skeletal sites in men. Human Genetics 118, 568-577CrossRefGoogle ScholarPubMed
44Choi, J.Y. et al. (2005) Genetic polymorphisms of OPG, RANK, and ESR1 and bone mineral density in Korean postmenopausal women. Calcified Tissue International 77, 152-159CrossRefGoogle ScholarPubMed
45Langdahl, B.L. et al. (2002) Polymorphisms in the osteoprotegerin gene are associated with osteoporotic fractures. Journal of Bone and Mineral Research 17, 1245-1255CrossRefGoogle ScholarPubMed
46Grundberg, E. et al. (2007) Vitamin D receptor 3′ haplotypes are unequally expressed in primary human bone cells and associated with increased fracture risk: the MrOS Study in Sweden and Hong Kong. Journal of Bone and Mineral Research 22, 832-840CrossRefGoogle ScholarPubMed
47Moffett, S.P. et al. (2007) Association of the VDR translation start site polymorphism and fracture risk in older women. Journal of Bone and Mineral Research 22, 730-736CrossRefGoogle ScholarPubMed
48Morrison, N.A. et al. (1992) Contribution of trans-acting factor alleles to normal physiological variability: vitamin D receptor gene polymorphism and circulating osteocalcin. Proceedings of the National Academy of Sciences of the United States of America 89, 6665-6669CrossRefGoogle ScholarPubMed
49Zmuda, J.M. et al. (2009) Association analysis of WNT10B with bone mass and structure among individuals of African ancestry. Journal of Bone and Mineral Research 24, 437-447CrossRefGoogle ScholarPubMed
50Hardy, J. and Singleton, A. (2009) Genomewide association studies and human disease. New England Journal of Medicine 360, 1759-1768CrossRefGoogle ScholarPubMed
51Manolio, T.A. et al. (2009) Finding the missing heritability of complex diseases. Nature 461, 747-753CrossRefGoogle ScholarPubMed
52Kruglyak, L. (2008) The road to genome-wide association studies. Nature Reviews. Genetics 9, 314-318CrossRefGoogle ScholarPubMed
53Kiel, D.P. et al. (2007) Genome-wide association with bone mass and geometry in the Framingham Heart Study. BMC Medical Genetics 8(Suppl 1), S14CrossRefGoogle ScholarPubMed
54Richards, J.B. et al. (2008) Bone mineral density, osteoporosis, and osteoporotic fractures: a genome-wide association study. Lancet 371, 1505-1512CrossRefGoogle ScholarPubMed
55Styrkarsdottir, U. et al. (2008) Multiple genetic loci for bone mineral density and fractures. New England Journal of Medicine 358, 2355-2365CrossRefGoogle ScholarPubMed
56Zhao, L.J. et al. (2010) Genome-wide association study for femoral neck bone geometry. Journal of Bone and Mineral Research 25, 320-329CrossRefGoogle ScholarPubMed
57Paternoster, L. et al. (2010) Genome-wide association meta-analysis of cortical bone mineral density unravels allelic heterogeneity at the RANKL locus and potential pleiotropic effects on bone. PLoS Genetics 6, e1001217CrossRefGoogle ScholarPubMed
58Kung, A.W. et al. (2010) Association of JAG1 with bone mineral density and osteoporotic fractures: a genome-wide association study and follow-up replication studies. American Journal of Human Genetics 86, 229-239CrossRefGoogle ScholarPubMed
59Koller, D.L. et al. (2010) Genome-wide association study of bone mineral density in premenopausal European-American women and replication in African-American women. Journal of Clinical Endocrinology and Metabolism 95, 1802-1809CrossRefGoogle ScholarPubMed
60Karasik, D. et al. (2010) Genome-wide pleiotropy of osteoporosis-related phenotypes: the Framingham Study. Journal of Bone and Mineral Research 25, 1555-1563CrossRefGoogle ScholarPubMed
61Hsu, Y.H. et al. (2010) An integration of genome-wide association study and gene expression profiling to prioritize the discovery of novel susceptibility Loci for osteoporosis-related traits. PLoS Genetics 6, e1000977CrossRefGoogle ScholarPubMed
62Guo, Y. et al. (2010) Genome-wide association study identifies ALDH7A1 as a novel susceptibility gene for osteoporosis. PLoS Genetics 6, e1000806CrossRefGoogle ScholarPubMed
63Xiong, D.H. et al. (2009) Genome-wide association and follow-up replication studies identified ADAMTS18 and TGFBR3 as bone mass candidate genes in different ethnic groups. American Journal of Human Genetics 84, 388-398CrossRefGoogle ScholarPubMed
64Timpson, N.J. et al. (2009) Common variants in the region around Osterix are associated with bone mineral density and growth in childhood. Human Molecular Genetics 18, 1510-1517CrossRefGoogle ScholarPubMed
65Styrkarsdottir, U. et al. (2009) New sequence variants associated with bone mineral density. Nature Genetics 41, 15-17CrossRefGoogle ScholarPubMed
66Rivadeneira, F. et al. (2009) Twenty bone-mineral-density loci identified by large-scale meta-analysis of genome-wide association studies. Nature Genetics 41, 1199-1206Google ScholarPubMed
67Liu, Y.Z. et al. (2009) Powerful bivariate genome-wide association analyses suggest the SOX6 gene influencing both obesity and osteoporosis phenotypes in males. PLoS One 4, e6827CrossRefGoogle ScholarPubMed
68Cho, Y.S. et al. (2009) A large-scale genome-wide association study of Asian populations uncovers genetic factors influencing eight quantitative traits. Nature Genetics 41, 527-534CrossRefGoogle ScholarPubMed
69Nobta, M. et al. (2005) Critical regulation of bone morphogenetic protein-induced osteoblastic differentiation by Delta1/Jagged1-activated Notch1 signaling. Journal of Biological Chemistry 280, 15842-15848CrossRefGoogle ScholarPubMed
70Weber, J.M. et al. (2006) Parathyroid hormone stimulates expression of the Notch ligand Jagged1 in osteoblastic cells. Bone 39, 485-493CrossRefGoogle ScholarPubMed
71Jungel, A. et al. (2004) Expression of interleukin-21 receptor, but not interleukin-21, in synovial fibroblasts and synovial macrophages of patients with rheumatoid arthritis. Arthritis and Rheumatism 50, 1468-1476CrossRefGoogle Scholar
72Giuliani, N. et al. (1999) Ethanol and acetaldehyde inhibit the formation of early osteoblast progenitors in murine and human bone marrow cultures. Alcoholism, Clinical and Experimental Research 23, 381-385CrossRefGoogle ScholarPubMed
73Yamaguchi, J. et al. (2006) ALDH2 polymorphisms and bone mineral density in an elderly Japanese population. Osteoporosis International 17, 908-913CrossRefGoogle Scholar
74Baek, W.Y. et al. (2009) Positive regulation of adult bone formation by osteoblast-specific transcription factor osterix. Journal of Bone and Mineral Research 24, 1055-1065CrossRefGoogle ScholarPubMed
75Gu, X. et al. (2001) Tel-2 is a novel transcriptional repressor related to the Ets factor Tel/ETV-6. Journal of Biological Chemistry 276, 9421-9436CrossRefGoogle Scholar
76Kirkbride, K.C. et al. (2008) Bone morphogenetic proteins signal through the transforming growth factor-beta type III receptor. Journal of Biological Chemistry 283, 7628-7637CrossRefGoogle ScholarPubMed
77Styrkarsdottir, U. et al. (2003) Linkage of osteoporosis to chromosome 20p12 and association to BMP2. PLoS Biology 1, E69CrossRefGoogle ScholarPubMed
78Stenvers, K.L. et al. (2003) Heart and liver defects and reduced transforming growth factor beta2 sensitivity in transforming growth factor beta type III receptor-deficient embryos. Molecular and Cellular Biology 23, 4371-4385CrossRefGoogle ScholarPubMed
79Devoto, M. et al. (2001) Variance component linkage analysis indicates a QTL for femoral neck bone mineral density on chromosome 1p36. Human Molecular Genetics 10, 2447-2452CrossRefGoogle ScholarPubMed
80Devoto, M. et al. (1998) First-stage autosomal genome screen in extended pedigrees suggests genes predisposing to low bone mineral density on chromosomes 1p, 2p and 4q. European Journal of Human Genetics 6, 151-157CrossRefGoogle ScholarPubMed
81Smits, P. et al. (2001) The transcription factors L-Sox5 and Sox6 are essential for cartilage formation. Developmental Cell 1, 277-290CrossRefGoogle ScholarPubMed
82Cho, S.W. et al. (2008) Differential effects of secreted frizzled-related proteins (sFRPs) on osteoblastic differentiation of mouse mesenchymal cells and apoptosis of osteoblasts. Biochemical and Biophysical Research Communications 367, 399-405CrossRefGoogle ScholarPubMed
83Nakanishi, R. et al. (2006) Secreted frizzled-related protein 4 is a negative regulator of peak BMD in SAMP6 mice. Journal of Bone and Mineral Research 21, 1713-1721CrossRefGoogle ScholarPubMed
84Nakanishi, R. et al. (2008) Osteoblast-targeted expression of Sfrp4 in mice results in low bone mass. Journal of Bone and Mineral Research 23, 271-277CrossRefGoogle ScholarPubMed
85Xiong, D.H. et al. (2006) Genome-wide scan identified QTLs underlying femoral neck cross-sectional geometry that are novel studied risk factors of osteoporosis. Journal of Bone and Mineral Research 21, 424-437CrossRefGoogle ScholarPubMed
86Liu, Y.Z. et al. (2008) Identification of PLCL1 gene for hip bone size variation in females in a genome-wide association study. PLoS One 3, e3160CrossRefGoogle ScholarPubMed
87Bodmer, W. and Bonilla, C. (2008) Common and rare variants in multifactorial susceptibility to common diseases. Nature Genetics 40, 695-701CrossRefGoogle ScholarPubMed
88Hughes, A.E. et al. (2000) Mutations in TNFRSF11A, affecting the signal peptide of RANK, cause familial expansile osteolysis. Nature Genetics 24, 45-48CrossRefGoogle ScholarPubMed
89McClung, M.R. et al. (2006) Denosumab in postmenopausal women with low bone mineral density. New England Journal of Medicine 354, 821-831CrossRefGoogle ScholarPubMed
90Cummings, S.R. et al. (2009) Denosumab for prevention of fractures in postmenopausal women with osteoporosis. New England Journal of Medicine 361, 756-765CrossRefGoogle ScholarPubMed
91Cirulli, E.T. and Goldstein, D.B. (2010) Uncovering the roles of rare variants in common disease through whole-genome sequencing. Nature Reviews. Genetics 11, 415-425CrossRefGoogle ScholarPubMed
92Duncan, E.L. et al. (2011) Genome-wide association study using extreme truncate selection identifies novel genes affecting bone mineral density and fracture risk. PLoS Genet 7, e1001372CrossRefGoogle ScholarPubMed
93Kou, I. et al. (2011) Common variants in a novel gene, FONG on chromosome 2q33.1 confer risk of osteoporosis in Japanese. PLoS One 6, e19641CrossRefGoogle Scholar
Figure 0

Table 1.Promising candidate genes in osteoporosis

Figure 1

Figure 1.Illustration of novel hits revealed by genome-wide association studies for osteoporosis and related traits. Novel bone mineral density loci are in blue, new loci for other osteoporosis-related phenotypes are in yellow and candidate loci validated by GWASs are in green. Please refer to Table 1 and the main text for full gene names and references.