Oral Presentation Annual Meetings of the Endocrine Society of Australia and Society for Reproductive Biology and Australia and New Zealand Bone and Mineral Society 2016

Skeletal determinants of bone strength identified in knockout mice are associated with low bone mineral density (BMD) in human cohorts (#158)

Mohammad Ali Moni 1 , Fernando Rivadeneira 2 , Julian M.W. Quinn 1 , Scott E. Youlten 1 , Emma L. Duncan 3 , Paul A. Baldock 1 , J.H. Duncan Bassett 4 , Graham R. Williams 4 , Peter I. Croucher 1
  1. Garvan Institute of Medical Research, Darlinghurst, NSW, Australia
  2. Department of Internal Medicine , Erasmus University of Rotterdam, Rotterdam, Netherland
  3. School of Medicine, Faculty of Medicine and Biomedical Sciences, , University of Queensland, Brisbane, Quuensland, Australia
  4. Department of Medicine, Imperial college, London, UK

BMD is genetically determined. Genome wide association studies have identified genes that control BMD; however, 90% of the genetic variability remains to be defined, suggesting alternative approaches are needed. We hypothesise that mice with altered bone mass and bone strength can be used to identify genes that control BMD in human populations.


The Origins of Bone and Cartilage Disease (OBCD) program examined bone mineral content (BMC) and strength in 200 knockout mouse strains generated by the International Knockout Mouse Consortium. We identified 18 genes with altered bone strength (‘test’ gene-set) and 18 ‘control’ genes with no impact on BMC or strength. We interrogated the femoral neck BMD human genome-wide association study (GWAS) dataset in women (n=32,961), from the Genetic Factors for Osteoporosis (GEFOS) consortium. This identified single nucleotide polymorphisms (SNPs) significantly associated with BMD in ‘control’ and ‘test’ genes, 10kb up-stream or down-stream of these genes, in gene regulatory motifs, in promoter or enhancer regions. To define a threshold p-value we ranked SNPs in the control gene-set, identified those significantly associated with BMD and applied the Holm-Bonferroni method to correct for the number of SNPs examined. This enabled a p value of 5x10-4 to be selected to define SNPs associated with BMD in the ‘test’ gene-set. This identified 21 SNPs significantly associated with BMD in three ‘test’ genes: 19 in AGAP1, an ADP-ribosylation factor GTPase-activating protein involved in membrane trafficking and 1 each in the enhancer regions of KLC2 (Kinesin light chain 2), a molecular motor and SPNS2 (spinster homolog-2), a sphingosine-1-phosphate transporter. Expression quantitative trait loci analysis showed SNPs in AGAP1 to be associated with gene expression (p<8.8x10-5) and in strong linkage disequilibrium (r2>0.8).


These data show that skeletal phenotypes in mice, coupled with a structured analytical methodology can identify SNPs associated with BMD in human populations.