Statistics from Altmetric.com
Rheumatoid arthritis (RA) shares some features with osteoarthritis (OA); in terms of synoviocyte activation, synovial hyperplasia and release of proinflammatory cytokines in the synovial fluid, OA seems a quantitatively dampened down variant of RA rather than a totally different entity.1 Moreover, lesions of deep cartilage and subchondral bone have been identified in RA prior to other defects.2
Recently, an association of a functional GDF5 polymorphism with OA was reported.3 GDF5 (growth differentiation factor 5) is a secreted prochondrogenic growth factor detected in regions of future joints during early development. Mutations in mouse and human GDF5 cause defective joints,4 several skeletal dysplasias have been described. RA pathogenesis was speculated to involve an attempt to recapitulate the embryonic limb development programme.5 We hypothesise that RA may share with OA an underlying defect in prematurely aged connective tissue; therefore, we aimed at testing the putative role of the functional polymorphism of the GDF5 gene previously found to increase OA risk (rs143383). We also studied an additional single nucleotide polymorphism (rs143384) selected to uncover most (>95%) of the variation within the linkage disequilibrium block where the GDF5 gene is located. The major +104T allele of the 5′UTR in GDF5 is over-represented in OA, both in Asian3 and in European populations.6 The associated allele showed significantly lower in vitro promoter activity in different cell lines3 and this altered expression pattern was confirmed in vivo in articular cartilage of patients suffering severe OA.6
Bone morphogenetic proteins, such as GDF5, were identified for their capacity to induce bone formation when injected subcutaneously in rodents.7 8 Skeletal repair seems to involve the same molecular pathways occurring during embryonic development.9 Articular cartilage fails to heal adequately after damage from trauma or arthritis and GDF5 might enhance repair. The present work shows that the functional GDF5 polymorphism associated with OA risk demonstrates a similar role in RA susceptibility. The over-represented allele in patients with RA correlated with reduced in vitro transcriptional activity and this difference in allelic expression was also observed using RNA extracted from the articular cartilage of patients with OA. Defective GDF5 expression seems to confer a higher predisposition to both OA and RA. However, while increased OA susceptibility is detected in carriers of the T allele, indicative of a dominant effect of this allele, the influence on RA risk evidences only in TT homozygotes (table 1). This fact may reflect a dosage effect: the homozygote genotype at GDF5 +104TT renders individuals more susceptible to RA (OR = 1.29) to the same degree observed for T carriers in patients with OA of Caucasian background6 (OR = 1.28). The study of the other tagging single nucleotide polymorphism or of the haplotypes in our Spanish cohorts adds nothing to the effect conferred by this separate allele.
Consecutively recruited Spanish patients with RA (69% women) and ethnically matched healthy controls (51% women) were included in a case–control study approved by the Hospital Clínico, Madrid. RA diagnosis was based on the American College of Rheumatology (ACR) criteria.10 Mean age at onset was 54 (SD 14); 61% patients carried the shared epitope; 66% and 50% were positive for rheumatoid factor and for cyclic citrullinated peptide, respectively, and 32% presented nodular disease.
Genotyping was performed by using TaqMan Assays, under the conditions recommended by the manufacturer (Applied Biosystems, Foster City, CA, USA). Statistical analysis performed with SPSS v 12.0.
The right level of GDF5 expression is tightly controlled during early life because altered expression can result in failure to develop normal synovial joints. The polymorphism associated with OA/RA susceptibility influences GDF5 promoter activity; therefore, it is a plausible aetiological candidate, and replication in independent cohorts is warranted.
The authors thank Carmen Martínez for her skilful technical assistance. AM holds a research contract of the Spanish Health Ministry (04/00175). JV is a fellow and EU works for the “Fundación para la Investigación Biomédica-Hospital Clínico San Carlos”. This work was supported by grants from: “Fundación Mutua Madrileña”, FIS PI041698, FIS 04/1691.
Competing interests: None.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.