Wim Van Hul

Identification of Sex-Specific Associations Between Polymorphisms of the Osteoprotegerin Gene, TNFRSF11B, and Paget's Disease of Bone†

We studied the role of TNFRSF11B polymorphisms on the risk to develop Pagets disease of bone in a Belgian study population. We observed no association in men, but a highly significant association was found in women, and this was confirmed in a population from the United Kingdom.

Genetic variation in the **TNFRSF11A** gene encoding RANK is associated with susceptibility to Paget's disease of bone

RANK (receptor activator of nuclear factor‐κB), encoded by TNFRSF11A, is a key protein in osteoclastogenesis. TNFRSF11A mutations cause Pagets disease of bone (PDB)–like diseases (ie, familial expansile osteolysis, expansile skeletal hyperphosphatasia, and early‐onset PDB) and an osteoclast‐poor form of osteopetrosis. However, no TNFRSF11A mutations have been found in classic PDB, neither in familial nor in isolated cases. To investigate the possible relationship between TNFRSF11A polymorphisms and sporadic PDB, we conducted an association study including 32 single‐nucleotide polymorphisms (SNPs) in 196 Belgian sporadic PDB patients and 212 control individuals. Thirteen SNPs and 3 multimarker tests (MMTs) turned out to have a p value of between .036 and 3.17 × 10−4, with the major effect coming from females. Moreover, 6 SNPs and 1 MMT withstood the Bonferroni correction (p < .002). Replication studies were performed for 2 nonsynonymous SNPs (rs35211496 and rs1805034) in a Dutch and a British cohort. Interestingly, both SNPs resulted in p values ranging from .013 to 8.38 × 10−5 in both populations. Meta‐analysis over three populations resulted in p = .002 for rs35211496 and p = 1.27 × 10−8 for rs1805034, again mainly coming from the female subgroups. In an attempt to identify the underlying causative SNP, we performed functional studies for the coding SNPs as well as resequencing efforts of a 31‐kb region harboring a risk haplotype within the Belgian females. However, neither approach resulted in significant evidence for the causality of any of the tested genetic variants. Therefore, further studies are needed to identify the real cause of the increased risk to develop PDB shown to be present within TNFRSF11A.

Indications for a genetic association of a VCP polymorphism with the pathogenesis of sporadic Paget's disease of bone, but not for TNFSF11 (RANKL) and IL-6 polymorphisms

Pagets disease of bone (PDB) is, after osteoporosis, the second most common metabolic bone disorder in the elderly Caucasian population. Mutations in the sequestosome 1 gene (SQSTM1) are responsible for the etiology of PDB in a subset of patients, but the disease pathogenesis in the remaining PDB patients is still unknown. Therefore association studies investigating the relationship between genetic polymorphisms and sporadic PDB have been performed in order to find the susceptibility polymorphisms. In this paper, we sought to determine whether polymorphisms in 3 functional candidate genes play a role in the development of sporadic PDB: TNFSF11 (receptor activator of nuclear factor κB ligand, RANKL), VCP (valosin-containing protein) and IL-6 (interleukin 6). Analyzing 9 tag SNPs and 2 multi-marker tests (MMTs) in TNFSF11, 3 tag SNPs and 1 MMT in VCP and 8 tag SNPs in IL-6 in a population of 196 Belgian patients with sporadic PDB and 212 Belgian control individuals revealed that one VCP SNP (rs565070) turned out to be associated with PDB in this Belgian study population (p=5.5×10(-3)). None of the tag SNPs or MMTs selected for TNFSF11 or IL-6 was associated with PDB. Still, replication of our findings in the VCP gene in other populations is important to confirm our results. However, when combining data of VCP with those from other susceptible gene regions from previous association studies (i.e. TNFRSF11A, CSF1, OPTN and TM7SF4), independent effect of each gene region was found and the cumulative population attributable risk is 72.7%.

Single nucleotide polymorphisms in sFRP4 are associated with bone and body composition related parameters in Danish but not in Belgian men.

The senescence accelerated mouse P6 (SAMP6) has a low bone mass and has previously shown to be a good model for senile osteoporosis in humans. In addition to a reduced bone mass, SAMP6 mice are obese and have hyperlipidemia. Using positional cloning and expression studies, an increased expression of sfrp4 was found in these mice. SFRP4 is a modulator of the Wnt signalling pathway. This pathway has been previously shown to be involved in regulating bone mass. Additional evidence that sFRP4 has an influence on BMD was delivered by linkage and association studies mostly performed in Asian populations. Based on these data we decided to perform an association study between common variants in sFRP4, BMD, hip geometry parameters and body composition parameters in a population consisting of 1383 Danish men (783 aged 20-29 years; 600 aged 60-74 years). Afterwards we tried to replicate the significant results in a population of 994 Belgian men. In the Danish population we found 6 SNPs associated with BMD at the hip and/or femoral neck. Furthermore, all 6 SNPs were associated with several hip geometry parameters. The homozygous presence of the minor allele resulted for all SNPs (except rs4720265) in a decrease in bone density and bone strength. Finally, we observed in the Danish population age specific associations with height and fat mass. In the Belgian population we tried to replicate the results of three SNPs with BMD and body composition parameters. Unfortunately, we were not able to replicate the results found in the Danish cohort but we found one SNP (rs2598116) associated with height. In conclusion, genetic variation in sFRP4 has an influence on hip fracture risk, percentage body fat and height in a Danish male population. However, we were unable to replicate these results in an independent Belgian population.

Association study of common variants in the sFRP1 gene region and parameters of bone strength and body composition in two independent healthy Caucasian male cohorts

Bone mineral density (BMD) and bone strength are predictive parameters for the development of osteoporosis and related fracture later in life. Although it is well known that BMD and bone strength have a high heritability, not much of the variation is already explained. Mice models showed that sFRP1 has an influence on bone formation. Therefore this study aimed to investigate the effect of common genetic variation on BMD and bone strength in Caucasian men of different ages. Using HapMap we selected 13 tagSNPs which tag most common genetic variation in and around sFRP1 and we genotyped these SNPs in the young cohort of the Odense Androgen Study (OAS). The OAS includes a total of 1383 Danish men from two different age groups ([20-29 years]: N=783; [60-74 years]: N=600) and is well characterised. The subjects were phenotyped for BMD at several sites, and additionally for body composition and hip geometry parameters. Based on the results of the young cohort we selected three SNPs for further analysis in the complete OAS population. To conclude we tried to replicate the results of two SNPs in an independent population of 994 Belgian men. We found a strong association for rs9694405 with BMI as well in both cohorts separately as in the whole OAS population. Further we found rs4736965 associated with several hip geometry parameters in the same population. However we were not able to replicate those results in the Belgian population. At last we found in the OAS population age specific effects for rs10106678 with whole body BMD and waist to hip ratio.