Answers are grounded in this paper and cite their sources.
Sir, Osteoarthritis and osteoporosis are disabling diseases in the elderly. Clinical observations suggest an inverse relationship between these disorders (for review see [1]). This hypothesis is still controversial and no molecular basis for this observation has been demonstrated. Genome-wide linkage analysis and subsequent association study identified a single-nucleotide polymorphism (SNP) (Arg324Gly) in the frizzled-related protein (FRZB) gene associated with hip osteoarthritis in Caucasian females [2] and with generalized osteoarthritis at multiple sites in Caucasians [3]. The haplotype coding for substitutions of two arginines (Arg200Trp and Arg324Gly) in FRZB appears to be a strong risk factor for primary hip osteoarthritis in Caucasian females. FRZB is a secreted antagonist of Wnt signalling molecules [4]. The Wnt pathway is of critical importance in development, homeostasis and pathology [5]. Wnt signalling, in particular Wnt co-receptor LRP-5 and Wnt antagonist sFRP1, has also been associated with osteoporosis [6–8]. Genetic differences within the Wnt pathway may provide a molecular basis for the inverse relationship between osteoarthritis and osteoporosis. We analysed two non-synonymous SNPs in FRZB in three populations: elderly female controls, females with osteoporotic hip fracture, and females with hip osteoarthritis requiring joint replacement surgery. The selection of female Caucasian patients was reported previously [9, 10]. Briefly, 76 women admitted for joint replacement surgery were included in the osteoarthritis group. In the osteoporosis group, 135 females consecutively admitted for low-trauma hip fracture were studied. Secondary osteoporosis was excluded. Elderly controls (n = 245) were women over 70 who had to be non-institutionalized, functionally independent and free from diseases or medications known to affect bone metabolism. Informed consent was obtained and all procedures were approved by the Leuven University Hospitals Ethics Committee. SNPs were analysed by pyrosequencing (PSQ™ HS-96A system; Biotage, Uppsala, Sweden) at the Genetic Service Facility (Flanders Interuniversity Institute for Biotechnology, Antwerp, Belgium). Both SNPs were in Hardy–Weinberg equilibrium in controls (P = 0.16, χ2 test; P = 0.49, exact test). No differences in genotypes were found for the Arg200Trp SNP between osteoarthritis, osteoporosis and controls (P = 0.09, 3 × 3 Fisher's exact test) (Table 1). No differences in allelic frequencies were found (P = 0.085, 3 × 2 χ2 test). In contrast, variant allele carrier status (T allele) was significantly different between the three groups (P = 0.028, 3 × 2 χ2 test). Twenty-five of 76 osteoarthritis patients carried at least one variant compared with 64/238 of controls (P = 0.39, Yates’ χ2) and 23/133 of patients with osteoporosis (P = 0.016, Yates’ χ2). This association remained significant after Bonferroni correction for multiple testing. For osteoporosis, an opposite trend was observed as only 23/133 patients carried a mutation compared with controls (64/238 controls patients) (P = 0.049, Yates’ χ2) (Table 1). This association did not remain significant after Bonferroni correction. Analysis of Arg200Trp and Arg324Gly SNP in three patient groups Data are number of patients (%). aBonferroni correction for multiple testing was applied by multiplying P values by the number of tests (n = 3); bgenotyping for the Arg200Trp SNP was successful in 76/76 osteoarthritis, 133/135 osteoporosis patients and 238/245 controls; genotyping for the Arg324Gly SNP was successful in76/76 osteoarthritis, 132/135 osteoporosis patients and 237/245 controls; cP value determined by exact test; dP value determined by χ2 test. Analysis of Arg200Trp and Arg324Gly SNP in three patient groups Data are number of patients (%). aBonferroni correction for multiple testing was applied by multiplying P values by the number of tests (n = 3); bgenotyping for the Arg200Trp SNP was successful in 76/76 osteoarthritis, 133/135 osteoporosis patients and 238/245 controls; genotyping for the Arg324Gly SNP was successful in76/76 osteoarthritis, 132/135 osteoporosis patients and 237/245 controls; cP value determined by exact test; dP value determined by χ2 test. No differences were seen for the Arg324Gly SNP (Table 1). Combined analysis of both SNPs showed a significant difference between the groups (P = 0.036, 3 × 3 χ2 test) when patients and controls were stratified as carriers of none, one or two SNPs (Table 1). Significant differences were found between osteoarthritis and osteoporosis (P = 0.009, 3 × 2 χ2 test) as well as between osteoporosis patients and controls (P = 0.021, 3 × 2 χ2 test). The latter significance disappeared after Bonferroni correction. In the control group, no differences in bone mineral density were found between individuals carrying the variant and those with the normal allele at position 200 (Student's t-test P>0.5). The association of the Arg200Trp SNP in FZRB with hip osteoarthritis became apparent when osteoarthritis was compared with osteoporosis. We could not demonstrate a difference from controls. The small number of osteoarthritis and osteoporosis patients limits the power to detect association. Nevertheless, the careful selection of patients and the use of three groups may add sensitivity to the analysis. We could not confirm the association of the Arg324Gly SNP with hip osteoarthritis. The differences between this study and the original report of Loughlin et al. [2] may partly relate to differences in study population. It is not clear to what extent osteoporosis patients were excluded from the control samples in the latter study [2]. In addition, there is evidence that populations from the UK and Belgium are genetically different [11]. The lack of significance may also be due to the small sample size and therefore lack of power. These observations provide molecular support for the inverse relationship between osteoarthritis and osteoporosis. The Arg324Gly substitution in the FRZB gene results in loss of Wnt antagonism [2]. A similar or additional effect of the Arg200Trp substitution could not be demonstrated [2]. Gain-of-function mutations in the LRP5 Wnt co-receptor result in higher bone mass [6, 7]. Genetic deletion of sFRP1 in mice results in increased trabecular bone mineral density [8]. As increased Wnt signalling is associated with high bone mass, loss of function in FZRB may lead to higher bone density and be inversely associated with osteoporosis. Overall, our results support the concept that the Wnt signalling pathway is of particular interest in both joint and bone diseases. The authors wish to thank Dr Jeroen Aerssens and Dr Jan Dequeker for DNA collections, Dr Séverine Vermeire for critical reading of the manuscript and Inge Derese for technical assistance. R.J.L. is the recipient of a postdoctoral fellowship and S.B. of a clinical investigator fellowship from the Fund for Scientific Research Flanders. The authors have declared no conflicts of interest. 1Laboratory for Skeletal Development and Joint Disorders, Department of Rheumatology and 2Leuven University Center for Metabolic Bone Diseases, University Hospitals Leuven, Katholieke Universiteit Leuven, Belgium
No takes yet. Share an insight, caveat, or question.
Lories et al. (2005) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: