Identification and functional characterization of genetic loci involved in osteoporosis and atypical femoral fracture

Abstract

Osteoporosis is a complex disease, determined by genetic and environmental factors, characterized by a low bone mass and microarchitectural deterioration of bone tissue, with a consequent decrease of bone strength and increase in bone fragility and fracture risk. It is the most common bone metabolic disorder, and a major worldwide public health concern. Bone mineral density (BMD) heritability ranges from 50% to 85% and genetic factors have a crucial role in the pathogenesis of osteoporosis. To date, many genes and variants associated with osteoporosis have been identified, but they only explain 20% of the phenotypic variance. In addition, the underlying mechanisms of these associations are poorly understood. Hence, it is necessary to identify new variants/genes and to deepen the knowledge about osteoporosis pathogenesis. Nitrogen-containing bisphosphonates (N-BPs) are the first-line pharmacological treatment for osteoporosis. They are cost-effective anti-resorptive drugs that inhibit the mevalonate pathway, preventing osteoclasts’ function and promoting their apoptosis. Very rarely, atypical femoral fractures (AFFs) can occur after a long-term therapy with N-BPs. AFFs are located in the subtrochanteric region or the femoral diaphysis and arise after minimal or no trauma. The pathogenic mechanisms underlying AFF remain largely unknown.

This PhD thesis aimed at contributing to the elucidation of the genetic determinants of osteoporosis and AFF.

On the one side, we deeply studied a GWAS signal in the C7ORF76 locus, in 7q21.3, including its dissection in the BARCOS cohort of Spanish postmenopausal women and the functional characterization of the associated variants and regulatory elements within the locus. We identified 2 variants associated with BMD and osteoporotic fracture and showed that they are cis-eQTLs for the neighbouring gene SLC25A13 in human primary osteoblasts. One of the associated variants lay in an upstream putative regulatory element, which we named UPE. We have demonstrated the regulatory capacity of UPE in bone cells, and its interaction with a lncRNA and other regulatory elements within the region. We also studied a previously described mouse Dlx5/6 enhancer (eDlx#18) within the C7ORF76 locus. We demonstrated that it is able to activate transcription in an osteoblastic context and that it interacts with the DLX5 promoter, as well as with other tissue-specific DLX5/6 enhancers. An SNP within this enhancer has been shown to be a cis-eQTL for DLX6 in human primary osteoblasts. Finally, the homozygous deletion of eDlx#18 in mice resulted in a reduced viability, a decreased Dlx5 expression in otic vesicle and branchial arches in E11.5 embryos. In E17.5 embryos, skeletal defects were noted, including a smaller dentary and deficient ossification at several sites.

On the other side, a small cohort of N-BP-associated AFF patients was analyzed by whole exome sequencing, and the most conserved and interesting mutation found, in the GGPS1 gene, was functionally characterized using molecular and cellular approaches. We found 37 rare mutations in 34 genes shared by 3 sisters with N-BP-associated AFF, including mutations in GGPS1 and CYP1A1, which was also mutated in one unrelated patient. The p.Asp188Tyr mutation in GGPS1 affected oligomerization of the enzyme, leading to a severe reduction in enzyme activity. GGPS1 depletion in osteoblastic cells resulted in a strong mineralization reduction and a decreased expression of some osteoblastic markers, such as osteocalcin, osterix and RANKL. The depletion in osteoclast precursors led to increased osteoclast number but with reduced resorption activity.