The GWA approach is a state-of-the-art approach to uncover modest genetic variants contributing to common diseases or phenotypes. Using a GWA approach, our group has reported two candidate genes—ADAMTS18
(16q23) and TGFBR3
(1p22)—for spine or hip BMD previously.(18
) In addition, three other GWA studies on BMD have been published,(13
) and they successfully identified several candidate genes for BMD, including RANKL
(14q32), and SP7
(12q13). However, these loci in combination can explain only a small fraction of BMD variation, leaving the majority of the genetic factors that influence BMD variation unknown. In addition, most published GWA studies focused only on the genes or SNPs of top-ranking statistical significance, which may ignore some useful information. In this study, by using available GWA data sets from two white populations, we identified two susceptibility genes—PTH
(11p15) and IL21R
(16p11)—associated with FN BMD variation. These two genes were not in the top-significance list in either of the populations and were not identified by our previous GWA study on BMD.(18
) However, combining the two data sets by meta-analysis revealed the promising significance of these two genes because the meta-analysis could improve the power to detect more associations and investigate the consistency of those associations across different populations.(33
) Moreover, both genes have potential biologic functions that are important to bone metabolism. Thus our findings added more information to the overall understanding of the genetic basis of osteoporosis.
plays a pivotal role in calcium homeostasis and bone remodeling. In experimental animals and patients with osteoporosis, intermittent administration of PTH can increase bone mass by stimulating de novo bone formation.(32
) However, genetic studies testing for association between polymorphisms in PTH
and osteoporosis are lacking, and most of them are underpowered and show inconsistent results.(37
) Our study found a consistent association between PTH
and FN BMD in two independent white populations, thereby supporting the conclusion that PTH
is an important candidate gene for BMD and osteoporosis. Although the significant SNPs we identified are located in the upstream of the PTH
gene, they are clustered in the same LD block as the SNPs within the PTH
gene. In addition, intergenic transcription now has been recognized as an active and common cellular process. Extensive transcription has been observed in unannotated genomic regions that are related to genotype-phenotype correlations.(42
) As an important function, intergenic transcription can regulate expression of the nearby genes.(44
) In particular, SNPs rs9630182 and rs2036417 are located at potential transcription factor binding sites predicted by the FASTSNP program (http://fastsnp.ibms.sinica.edu.tw
). Thus we hypothesized that those SNPs potentially might regulate PTH
gene expression through intergenic transcription, although the real molecular mechanisms await further investigation.
Cytokins are potent mediators regulating homeostasis of the immune system and pathophysiologic processes. As a member of the type I cytokine receptors, IL21R
has multiple functions. For example, IL21R
plays an important role in the proliferation and differentiation of various immune cells, such as T cells and B cells. Studies have shown that B cells may participate in osteoclastogenesis.(46
induces the growth-promoting signals of its ligand, IL21
, which might be involved in the maturation and function of myeloid cells.(47
have been revealed to be involved in a variety of human diseases, including cancers, inflammatory bowel disease and Crohn's disease, and multiple autoimmune diseases. Especially, IL21R
has been identified as associated with the activated phenotype of rheumatoid arthritis fibroblasts and correlates negatively with the destruction of cartilage and bone.(48
) With this information taken together, we suggested that IL21R
may be a new candidate gene for BMD.
We compared the results for the key SNPs identified in previous BMD GWA studies(13
) with our current GWA study. Since replication analysis was the specific hypothesis driven, p
< .05 was considered significant. We confirmed associations for several SNPs located in the previously well-studied candidate genes, such as ESR1
, and LRP5
(). However, some SNPs were not able to be replicated in our study (Supplemental Table 3
), which might be affected by many factors. First, the effect sizes of variants were very small and thus easily lead to failure of replication under current statistical power. Second, some SNPs identified in previous studies were for spine BMD, and our study focused only on FN BMD. BMDs at different skeletal sites may have different genetic mechanisms. Third, the differences in gene-gene and gene-environment interactions between the two study sets may result in inconsistency in replication. In addition, other factors, such as differential LD and allele frequencies across populations, also may significantly influence the chance of replicating GWA results.
It is worth emphasizing that population stratification is unlikely to be a major concern in this GWA study. This is so first because we used EIGENSTRAT to perform GWA analyses in the discovery sample, which can control for potential population stratification effectively. Second, we used a family-based sample to perform replication analyses. Family-based samples are ideal for the follow-up validation of initial GWA findings(49
) because they are robust to population stratification and essentially can eliminate the possible impact of population stratification. Thus our GWA results are not likely to be plagued by spurious associations owing to population stratification.
In summary, we identified two susceptibility bone mass candidate genes, PTH and IL21R, that may influence FN BMD variation. Although additional functional studies are required to elucidate the detailed roles and potential functional variants of these loci, our findings provide some new insights into the understanding of the genetic architecture of BMD and osteoporosis.