Using epigenomic data to inform genome-wide association studies of bone mineral density

Using epigenomic data to inform genome-wide association studies of bone mineral density

John A. Morris1,2

1Department of Human Genetics, McGill University, Montreal, Canada;2Centre for Clinical Epidemiology, Lady Davis Institute, Jewish General Hospital, McGill University, Montreal, Canada

Correspondence to: John A. Morris, PhD Candidate. Centre for Clinical Epidemiology, Lady Davis Institute, Jewish General Hospital, 3755 Côte Ste-Catherine, Montréal, Québec H3T 1E2, Canada. Email:

Provenance: This is a Guest Editorial commissioned by Section Editor Hongfei Shi, MD, PhD (Associate Chief Surgeon, Department of Orthopaedics, Nanjing Drum Tower Hospital, Nanjing, China).

Comment on: Guo Y, Dong SS, Chen XF, et al. Integrating Epigenomic Elements and GWASs Identifies BDNF Gene Affecting Bone Mineral Density and Osteoporotic Fracture Risk. Sci Rep 2016;6:30558.

Submitted Sep 27, 2016. Accepted for publication Sep 29, 2016.

doi: 10.21037/atm.2016.11.09

Osteoporosis is primarily an aging-related disease, characterized by decreased bone strength and increased fracture risk. The incidence of osteoporosis is increasing worldwide due to aging populations and is a major healthcare burden. For instance, osteoporosis-related healthcare costed China $9.45 billion USD in 2010, will likely double by 2035, and increase to $25 billion USD by 2050 (1). Bone mineral density (BMD) is used to predict fracture risk and is the primary clinical measurement used to diagnose osteoporosis. Genome-wide association studies (GWAS) of BMD have successfully identified many genetic loci that influence osteoporosis (2-12), but like most GWAS, these studies used stringent statistical significance thresholds to limit false positive results at the expense of false negatives.

Large collaborative efforts such as ENCODE (13) and the Epigenome Roadmap (14) projects have come to completion in recent years, generating genome-wide epigenetic data from hundreds of cell lines and primary tissues. This resource is immensely useful, particularly to help investigate whether non-protein coding single nucleotide polymorphisms (SNPs) associated with a disease are functional (e.g., via perturbation of regulatory elements such as transcription factors). Guo et al. (15) integrated epigenomic data from ENCODE and BMD GWAS data from GEFOS (2) to identify a novel gene, brain-derived neurotrophic factor (BDNF), demonstrating how large epigenomic datasets can be utilized to explore putative false negative GWAS results and inform experimental follow-up studies.

Guo et al. (15) reasoned that enriched regulatory elements for known BMD genes can be used to predict which genes from a pool of putative false negative GWAS results may have a role in osteoporosis etiology. Guo et al. (15) identified 259 previously reported BMD and osteoporosis-related trait associated genes (16) and performed pathway analyses to identify which biological pathways were enriched, resulting in several known pathways such as the Wnt, Hedgehog, MAPK, and osteoclast differentiation pathways. Guo et al. (15) then integrated data from the ENCODE project; using data from multiple cell lines to determine which transcription factor binding sites (TFBSs), chromatin states, or histone marks were enriched within the 259 known genes. By developing a scoring system and prioritizing genes with the highest rankings for gene set enrichment analysis, Guo et al. (15) validated their gene list by identifying similarly enriched pathways with the known pathways, and further performed network analyses to evaluate shared pathways between their gene list and the known genes.

The high scoring genes, BDNF and phosphodiesterase 4D (PDE4D), both demonstrated evidence to be implicated in osteoporosis etiology. GWAS data showed that BDNF and PDE4D had suggestive association with BMD measured at the lumbar spine, with association P values of approximately 5×10−4 (2). In addition, the SNPs mapping to BDNF were found to have suggestive association with fracture (rs11030119: P=0.024; rs7124442: P=0.042) in the Chinese Fractures Study (CFS) of 350 cases and 350 controls. However, these SNPs have relatively low minor allele frequencies (MAF) in cases (rs11030119: MAF =0.037; rs7124442: MAF =0.062) and controls (rs11030119: MAF =0.062; rs7124442: MAF =0.089), thus limiting the power of the study, resulting in a confidence interval of the odds ratio for rs7124442 encompassing the null [OR =0.67 (0.45–1.00)]. As such, fracture results of rs11030119 and rs7124442 would require replication in a larger sample size to conclusively state their association with fracture.

To functionally link the BDNF SNPs to BDNF function, Guo et al. (15) performed eQTL analyses and short interfering RNA (siRNA) knock-down experiments. Cis-eQTL analysis of BDNF SNPs in 462 unrelated lymphoblastoid cell line (LCL) samples identified proximal SNPs in high linkage disequilibrium significantly associated with BDNF mRNA, suggesting that the effect of these SNPs on BMD are through BDNF. siRNA knockdown assays of BDNF in mouse-derived osteoblasts provided evidence that BDNF interacts with osteoblast differentiation factors, and its inhibition may stunt bone formation.

Guo et al. (15) have therefore shown that genes with suggestive association to BMD from population-based studies may still be relevant to osteoporosis etiology, and have identified BDNF as one such candidate for a novel BMD gene. Studies using epigenomic data are not without caveats, and it is important to consider these points when conducting integrative analyses. Data from the ENCODE project were entirely derived from cell lines, many of which many not be directly relevant to the trait or disease under study. ENCODE project data did include a human osteoblast cell line, but in the study of bone biology, there is a clear deficit of epigenomic data from other important bone-related cells such as osteoclasts and osteocytes. Studies integrating epigenomic data and GWAS of BMD will be greatly improved by using primary tissues or cell cultures directly related to bone biology, once these become available. Whether or not BDNF becomes clinically utile in diagnosing osteoporosis or predicting fracture risk in humans will be of interest, as Guo et al. (15) have presented intriguing functional evidence in murine osteoblast.


I thank Dr. Vincenzo Forgetta for his helpful input on the editorial.


Conflicts of Interest: The author has no conflicts of interest to declare.


  1. Si L, Winzenberg TM, Jiang Q, et al. Projection of osteoporosis-related fractures and costs in China: 2010-2050. Osteoporos Int 2015;26:1929-37. [Crossref] [PubMed]
  2. Estrada K, Styrkarsdottir U, Evangelou E, et al. Genome-wide meta-analysis identifies 56 bone mineral density loci and reveals 14 loci associated with risk of fracture. Nat Genet 2012;44:491-501. [Crossref] [PubMed]
  3. Zheng HF, Forgetta V, Hsu YH, et al. Rosello-Diez A.Whole-genome sequencing identifies EN1 as a determinant of bone density and fracture. Nature 2015;526:112-7. [Crossref] [PubMed]
  4. Styrkarsdottir U, Thorleifsson G, Gudjonsson SA, et al. Sequence variants in the PTCH1 gene associate with spine bone mineral density and osteoporotic fractures. Nat Commun 2016;7:10129. [Crossref] [PubMed]
  5. Styrkarsdottir U, Thorleifsson G, Sulem P, et al. Nonsense mutation in the LGR4 gene is associated with several human diseases and other traits. Nature 2013;497:517-20. [Crossref] [PubMed]
  6. Duncan EL, Danoy P, Kemp JP, et al. Genome-wide association study using extreme truncate selection identifies novel genes affecting bone mineral density and fracture risk. PLoS Genet 2011;7:e1001372. [Crossref] [PubMed]
  7. Koller DL, Ichikawa S, Lai D, et al. Genome-wide association study of bone mineral density in premenopausal European-American women and replication in African-American women. J Clin Endocrinol Metab 2010;95:1802-9. [Crossref] [PubMed]
  8. Xiong DH, Liu XG, Guo YF, et al. Genome-wide association and follow-up replication studies identified ADAMTS18 and TGFBR3 as bone mass candidate genes in different ethnic groups. Am J Hum Genet 2009;84:388-98. [Crossref] [PubMed]
  9. Rivadeneira F, Styrkársdottir U, Estrada K, et al. Twenty bone-mineral-density loci identified by large-scale meta-analysis of genome-wide association studies. Nat Genet 2009;41:1199-206. [Crossref] [PubMed]
  10. Styrkarsdottir U, Halldorsson BV, Gretarsdottir S, et al. New sequence variants associated with bone mineral density. Nat Genet 2009;41:15-7. [Crossref] [PubMed]
  11. Richards JB, Rivadeneira F, Inouye M, et al. Bone mineral density, osteoporosis, and osteoporotic fractures: a genome-wide association study. Lancet 2008;371:1505-12. [Crossref] [PubMed]
  12. Styrkarsdottir U, Halldorsson BV, Gretarsdottir S, et al. Multiple genetic loci for bone mineral density and fractures. N Engl J Med 2008;358:2355-65. [Crossref] [PubMed]
  13. Dunham I, Kundaje A, Aldred SF, et al. An integrated encyclopedia of DNA elements in the human genome. Nature 2012;489:57-74. [PubMed]
  14. Kundaje A, Meuleman W, Ernst J, et al. Integrative analysis of 111 reference human epigenomes. Nature 2015;518:317-30. [Crossref] [PubMed]
  15. Guo Y, Dong SS, Chen XF, et al. Integrating Epigenomic Elements and GWASs Identifies BDNF Gene Affecting Bone Mineral Density and Osteoporotic Fracture Risk. Sci Rep 2016;6:30558. [Crossref] [PubMed]
  16. Welter D, MacArthur J, Morales J, et al. The NHGRI GWAS Catalog, a curated resource of SNP-trait associations. Nucleic Acids Res 2014;42:D1001-6. [Crossref] [PubMed]
Cite this article as: Morris JA. Using epigenomic data to inform genome-wide association studies of bone mineral density. Ann Transl Med 2016;4(24):487. doi: 10.21037/atm.2016.11.09

Download Citation