SNP–SNP interactions within APOE gene influence plasma lipids in postmenopausal osteoporosis

Abstract

Biological revelations have gradually started clearing the complex relationships of bones with lipids. Studies have shown that lipid proWles are related to bone mass, bone fragility and fracture risk [1, 2]. It has been observed that oxidized lipids inhibit mineralization of bone, induce osteoblastic diVerentiation in vascular cells, and hyperlipidemia reduces bone density in mice [3]. To understand the genetic link between lipids and osteoporosis risk, apolipoprotein E gene (APOE) is a potential candidate because of its vital contribution to both lipid and vitamin K metabolism. APOE is located on the chromosome 19q13.2 and has three codominant alleles i.e. APOE2, APOE3 and APOE4. It serves as a ligand for receptor-mediated uptake of lipoprotein particles, which are the main carriers of vitamin K, and availability of vitamin K is an essential step for the carboxylation of glutamic acid residues of osteocalcin, an important bone protein. A study has shown that decrease in bone mineral density (BMD) in hemodialysis patients is correlated with hyperlipidemia, and the risk of fractures in such patients varies according to APOE genotypes [4]. Some preliminary Wndings suggest that APOE contributes to the variation in cholesterol increase with menopause [5], and APOE-E4-negative subjects respond better to hormone replacement therapy (HRT) for lowering lipid levels than APOE-E4-positive subjects [6]. However, in what way, genomic changes within APOE inXuence lipids for the risk of osteoporosis remains unclear. The impact of individual APOE SNP or allelic variant on lipid levels may vary if another nearby SNP is also participating or in linkage disequilibrium with another functional SNP. Studies comprising single SNP would have overlooked coordinated eVects of such SNPs. The present study explored the SNP–SNP interactions of four pertinent APOE SNPs (rs440446, rs769450, rs429358 and rs7412) as the genetic mediators of lipids in postmenopausal osteoporotic women of Punjab Province of Northwest India. Out of 366 randomly selected postmenopausal women who were not using HRT and were aged between 45 and 70 (Mean age 53.2 § 9.2), 154 osteoporotic women (71-lumbar spine and 83-femoral neck cases) were enrolled after veriWcation of the fractures from original radiographs and validation by DEXA (dual energy X-ray absorptiometry) unless they met the exclusion criteria: hyperlipidemia, irregular cycles or premature ovarian failure (menopause < 38 years), women with surgical menopause, any systemic disease or receiving medications known to inXuence calcium metabolism. DEXA veriWed population-based 83 postmenopausal women were enrolled as controls, according to a random-digit invitation from residents of the same geographic suburban and rural areas from Northwest India with three main exclusion criteria: hyperlipidemia, any fracture before or after menopause or any anti-osteoporotic treatment. Detailed information on age, medical history, years since menopause and gynecological history was obtained from these subjects. Height and weight were measured, and BMI (kg/m) was calculated. In the present study, none of the subjects were found to be alcohol users or smokers. Bone mineral density was measured by DEXA using Hologic QRR 4500 (Hologic Inc. Waltham, MA, M. Singh (&) · P. Singh Molecular Genetics Laboratory, Department of Human Biology, Punjabi University, Patiala, Punjab 147002, India e-mail: singh_m12@rediffmail.com