Authors: Moskowitz, Roland W.; Altman, Roy D.; Hochberg, Marc C.; Buckwalter, Joseph A.; GoldberG, Victor M.
Title: Osteoarthritis: Diagnosis and Medical/Surgical Management, 4th Edition
Copyright 2007 Lippincott Williams & Wilkins
> Table of Contents > I - Basic Considerations > 6 - Molecular Genetics of Osteoarthritis
Molecular Genetics of Osteoarthritis
There has long been a suspicion that genes play a role in the development of osteoarthritis (OA). However, it is only in the last 10 years that comprehensive epidemiological studies have enabled scientists to construct a reasonably clear picture of the extent and likely nature of the OA genetic component. We now know that OA is a multifactorial disease with a major polygenic element. This genetic susceptibility shows heterogeneity between different skeletal sites and, possibly, between the two sexes. Genes harboring common OA risk alleles are now being reported in reasonably powerful, and therefore robust, linkage and association studies. Several of these genes code for proteins that regulate cartilage homeostasis, but effects on other tissues of the synovial joint cannot be ruled out. Identifying OA susceptibility genes will enhance our understanding of this common arthritis and will assist in the development of new treatments. It will also help in the identification of individuals at increased risk and will aid the characterization of the environmental factors that can also influence OA aetiology. This chapter will focus on the most compelling findings from the numerous linkage and association studies that have been performed. It will bring the reader up-to-date on the molecular genetics of this common disease.
OA can exist in two main forms: primary OA and secondary OA. Primary, or idiopathic, OA is the common late-onset form of the disease with radiographic evidence first detectable in the fifth decade. It has no obvious cause and is either localized to a particular joint group or more generalized. Secondary OA arises in response to clearly identifiable factors such as trauma, or a congenital or a developmental abnormality. In a small number of cases, secondary OA is associated with developmental abnormalities that are transmitted as mendelian traits. These diseases are members of the osteochondrodysplasia class of skeletal dysplasias and the OA in these familial cases is often early onset, precocious, and severe. Linkage and positional cloning has identified the disease genes and causal mutations in several of the osteochondrodysplasias. Because primary OA is the form of OA that impacts most significantly in the population, it is the form that we will concentrate on in this chapter.
Geneticists studying OA have used several strategies to try and identify susceptibility genes. These include candidate gene studies based solely on biological clues, systematic and model-free genome-wide linkage and association scans, and gene expression studies. These investigations are beginning to yield compelling data. They will each be discussed in turn.
Candidate Gene Studies
Without any prior linkage data to guide them toward particular chromosomal locations, a number of investigators have used their biological understanding of OA to select candidate genes for association studies. The initial candidate gene studies focused on genes that code for cartilage extracellular matrix (ECM) structural proteins. These included COL2A1, which codes for the 1 polypeptide chain of type II collagen, the principal collagenous component of articular cartilage. Other cartilage collagen genes studied included the type IX and type XI collagen genes and genes coding for noncollagenous components of the ECM, such as the cartilage oligomeric matrix protein gene (COMP) and the aggrecan gene AGC1. On the whole, these studies did not yield convincing data to support a role for common, nonsynonymous mutations in cartilage ECM structural protein genes as risk factors for primary OA.1,2 It must be concluded, therefore, that these genes have never undergone a mutational event within their amino acid coding sequence that predisposes to primary OA or that if such mutations have occurred, these mutations do not have a sufficiently high frequency to confer a significant population risk.3
Concurrent with the analysis of ECM structural protein genes, a number of investigators focused on genes coding for proteins influencing bone density. This was driven by the observation that subchondral sclerosis is an early observation in some OA joints and the subsequent suggestion that this sclerosis might damage the cartilage by adversely affecting the transmittance of mechanical load. The vitamin D receptor gene VDR and the estrogen receptor gene (ESR1) have both been investigated by a number of groups. Early studies demonstrated association of both genes with OA. However, these associations were not replicated in all subsequent investigations.1,2 This is to be expected for a common complex disease such as OA and highlights the fact that any single locus will have at most only a moderate effect on disease occurrence in the population under investigation. A reasonable number of positive reports are now being published, particularly for ESR1.4 Common variants within this gene are therefore likely to be OA risk factors.
As with other complex disease investigations, OA association studies often suffer from a number of shortcomings. These include the analysis of cohorts of relatively small size and the genotyping of only a few of the known DNA variants within the gene under investigation. The former should be unnecessary for a common disease like OA and makes many studies underpowered and liable to false positives. The latter is becoming avoidable as the cost of genotyping falls and as databases of common variants become more comprehensive in their coverage. Future OA association studies should be designed with these two considerations in mind.
Candidate gene studies have shed some light on OA genetic susceptibility. However, they are constrained by our incomplete knowledge of the biology of OA, which makes candidate selection a fallible act. Investigators have therefore taken a more systematic approach, including genome-wide linkage scans.
Genome-Wide Linkage Scans
Four OA genome-wide scans have so far been published, based on small families of affected relatives collected in the U.K.5,6 Finland,7 Iceland,8 and the USA.9,10 The U.K. scan was performed on patients ascertained by hip or knee OA, the other scans on patients with hand OA. The hand OA scans were performed using either a global hand OA score or by focusing on particular joints of the hand.
The United Kingdom Study
The first OA genome-wide linkage scan was published by an Oxford group in 1999 and was performed on 481 pedigrees that each contained at least one OA-affected sibling pair (ASP) ascertained by total joint replacement of the hip or the knee.5,6 With this ascertainment, the investigators were treating the disease as a discreet trait, since subjects either had or had not undergone joint replacement. The investigators were also focussing on pedigrees that had severe, end-stage OA. Linkages were initially reported to chromosomes 2 and 11, and these were found to be particularly relevant to ASPs concordant for hip OA (chromosome 2) and to female ASPs concordant for hip OA (chromosome 11). A subsequent stratification of the genome scan by sex and by joint replaced (hip or knee) uncovered additional linkages, on chromosome 4 in female ASPs, chromosome 6 in hip ASPs, and chromosome 16 in female ASPs.6 These were important findings not only because they pointed toward areas of the genome that may encode for OA susceptibility but because they also suggested differences in the nature of the susceptibility between males and females and between different skeletal sites, something that had been suggested by previous epidemiological studies.1,2
This original genome scan employed an average of just one microsatellite marker every 15 centiMorgan (cM). This medium density meant that the linkage intervals were relatively large. The Oxford investigators therefore subjected each locus to finer linkage mapping in an expanded cohort of 571 OA pedigrees. This analysis, which employed an average marker density of one marker every 5 cM, succeeded in narrowing each of the linkages and also confirmed or refined their restriction to particular strata. For example, the chromosome 6 linkage was narrowed from a 50-cM interval in hip ASPs to a 12-cM interval in female-hip ASPs. The results of the finer linkage mapping have been published11,12,13,14,15 and are summarized in Table 6-1. The most significant evidence for linkage from the Oxford study is a logarithm of the odds (LOD) score of 4.8 on chromosome 6.
The Finnish Study
The Finnish scan was performed on 27 pedigrees that each contained at least two affected siblings with radiographic distal interphalangeal (DIP) OA.7 Nine genomic regions supported linkage and the genotyping of these in additional pedigree members confirmed linkage to chromosomes 2q, 4q, 7p, and the Xcen (Table 6-1). The 2q and 4q linkages do not show overlap with the finer linkage mapped 2q and 4q intervals from the Oxford study and therefore probably represent different loci. The Finnish 2q linkage encompasses the interleukin-1 (IL-1) receptor and ligand gene cluster at 2q12-q13. This cluster has now been associated with OA (discussed later).
The Iceland Study
A genome-wide linkage scan utilizing 1000 microsatellites was performed on 329 Icelandic families containing 1143 individuals with primary hand OA and 939 of their relatives.8 Each family contained at least two affected individuals related to each other at or within five meioses. Individuals were classed as having hand OA if they exhibited at least two nodes at the DIP joints of each hand or if they demonstrated squaring or dislocation of the first carpometacarpal (CMC1) thumb joint. The highest LOD
TABLE 6-1 LOCI IDENTIFIED FROM THE FOUR OA GENOME-WIDE LINKAGE SCANS
The USA Study
The USA study involved 296 small, two-generational families composed of 684 parents and 793 of their offspring from the Framingham cohort.9,10 Hand OA was characterized radiographically and was assessed in the late 1960s and again in the early 1990s for the parents, and in the 1990s for the offspring. The investigators performed a quantitative linkage analysis that separately tested the degree of joint space narrowing (JSN) and the frequency of osteophytes. They also investigated an overall radiographic score based on the Kellgren and Lawrence (K/L) grading scheme. Linkage was reported to eight chromosomal regions, with the highest LOD score being 2.96 for chromosome 1 in the JSN criteria (Table 6-1). For the overall radiographic score, no LOD exceeded 1.9, and so phenotypic definition was reconsidered by stratifying the linkage scan according to OA disease status in particular joints of the hand that are most susceptible to OA, namely, the DIP joints, the thumb interphalangeal (IP) joint, the proximal interphalangeal (PIP) joint, the metacarpophalangeal (MCP) joints, the wrist joints, and the CMC1 joint at the base of each thumb.10 This revealed two new loci, one on chromosome 7q in the DIP stratum with a LOD score of 3.06, and one on chromosome 15q in the CMC1 stratum with a LOD score of 6.25 (Table 6-1). Stratification by joint and by sex revealed two further loci in females, one on chromosome 1q in the DIP stratum with a LOD score of 3.03, and one on chromosome 20p in the CMC1 stratum with a LOD score of 3.74. Overall, the USA study had identified at least 12 loci, several of which are restricted to particular joints of the hand.
Intriguingly, the chromosome 2p locus identified in the USA study directly overlapped with the 2p locus identified in the Iceland scan. This might therefore represent an independent confirmation, although a skeptic could argue that
Other Osteoarthritis Linkage Studies
As well as the four genome-wide linkage scans reported above, a number of investigators have targeted specific chromosomal regions based on the data from the genome scans. Two groups have investigated the chromosome 6 locus reported by the Oxford study. A Netherlands group investigated 100 small pedigrees composed of probands and their siblings whose OA disease status was assessed radiographically using the K/L scale.16 Linkage to chromosome 6 was reported in sibling pairs concordant for hip OA but not in sibling pairs concordant for knee, hand, or spinal OA. An Irish group investigated 109 small pedigrees composed of sibling pairs ascertained by hip replacement.17 No evidence for linkage to chromosome 6 was obtained. The Oxford study had reported that their chromosome 6 linkage was restricted to female sibling pairs concordant for hip OA. Of the 109 pedigrees in the Irish study, only 32 were female-hip sibling pairs. The Irish study therefore had little power to confirm the Oxford report and is likely to represent a false negative result. A group from Israel has investigated linkage to the chromosome 11 locus reported by the Oxford group.18 This study was conducted on a panel of 295 pedigrees derived from a relatively isolated population of Chuvashians from southern Russia whose hand-OA status had been previously determined using the K/L scale. Moderate evidence of linkage was obtained to the same chromosome 11 locus as that reported by the Oxford group.
The OA genome-wide linkage scans so far reported have revealed some compelling loci based on reasonably high LOD scores for a complex disease (Table 6-1). Several loci also show overlap between the different scans. An important observation from the Iceland and USA scans is that stratification of the hand into its component joints provides more significant evidence for linkage. This highlights the subtle and highly complex way in which OA genetic susceptibility is acting. The U.K. and USA studies stratified their scans by sex and subsequently identified novel loci that were particularly relevant to OA occurrence in women. This supports several epidemiological studies that have suggested that the role of genes in OA may vary between the two sexes.1,2 More targeted linkage studies have provided some confirmatory evidence for some of the loci detected through the genome scans; however, not all follow-up studies have been positive. This is to be expected for a complex trait where susceptibility loci will have varying frequencies between different cohorts and may be interacting with other loci, or nongenetic factors, in a population-specific manner. Complex traits are, as the name suggests, complex and our understanding of the innumerable factors that interact in their etiology is still at a very basic level.
Osteoarthritis Susceptibility Genes Detected via the Genome-Wide Linkage Scans
The genome-wide scans reviewed above have identified a number of relatively broad genomic intervals that may harbor OA susceptibility loci. Several of these intervals have now been subjected to association analyses principally candidate gene-based studies and have yielded a number of associated genes.
IL1 Gene Cluster: Chromosome 2q11.2-q13
Although OA is not an inflammatory arthritis, there are instances in which an inflamed synovium can exacerbate the disease. Furthermore, inflammatory mediators such as interleukins are synthesised by articular cartilage chondrocytes and can act in a paracrine and autocrine manner to influence cartilage tissue homeostasis.19,20 Genes encoding interleukins are therefore plausible OA-susceptibility loci. This notion was supported by the Finnish genome-wide scan discussed above, which reported evidence for linkage to an interval on chromosome 2q that encompassed the interleukin 1 gene cluster. This cluster, which resides at 2q11.2-q13 and covers approximately 12 Mb, contains at least 11 family members. The archetypal members, and those that have been subjected to considerable investigation, are IL1A (encoding IL-1 ), IL1B (encoding IL-1 ), IL1RN (encoding IL-1 receptor antagonist), and IL1R1 (encoding the signalling receptor for the above proteins).
The early OA association studies of the IL-1 gene cluster that were reported in the late 1990s and early 2000s were, on the whole, ambiguous, with only marginal evidence for association to OA reported.1,2 These studies had several weaknesses, including small sample sizes and the analysis of only a small proportion of the known gene variants. In 2004, two new studies were reported that alleviated some of these weaknesses.
The first study, by a group from the Netherlands,21 was a prospective, population-based analysis of Caucasians from the Rotterdam Elderly Study. A radiographic assessment of the OA status of the hip, knee, hand, and spine was conducted on 886 individuals (347 women, 520 men, and 19 subjects of unrecorded sex) aged between 55 and 65 years. The frequencies of OA were 8% for the hip, 16% for the knee, 4% for the hand, and 62% for the spine, with 17% having no radiographic OA at any of these sites. The patients were genotyped for a single-nucleotide polymorphism (SNP) located within exon 5 of IL1B (3953), for a SNP located upstream of IL1B (511), and for a variable number of tandem repeat (VNTR) polymorphism located within intron 2 of IL1RN. None of the variants was associated with OA when the cohort was analyzed unstratified. However, following stratification by site, the three variants each demonstrated association with hip OA, with P-values ranging from 0.004 to 0.001. A haplotype analysis identified a risk and a protective haplotype. Further analyses implied that the three variants themselves were not directly responsible for the associations but were instead in linkage disequilibrium with as yet unidentified functional variants.
The second study, conducted by a group from the U.K. was a case-control association analysis.22 The cases had radiographically diagnosed knee OA and were collected from two U.K. centers: 141 patients from Bristol (76 females and 65 males, mean age of 62 years) and 163 patients from London (125 females and 38 males, mean age of 71 years). All cases had OA as defined by the American College of Rheumatology (ACR) criteria with at least grade 2 radiological changes. The controls were 195 unrelated healthy blood donors (98 females and 97 males with an unspecified mean age). All individuals were U.K. Caucasians. The cases and controls were genotyped for nine IL1R1 promoter variants (seven SNPs, one insertion/deletion [indel], and one microsatellite), for two IL1A variants (a promoter SNP and a microsatellite located in intron 4), for three IL1B variants (the 3953 and 511 SNPs genotyped in the Netherlands study, and a promoter SNP), and for three IL1RN variants (the VNTR genotyped in the Netherlands study, a SNP located in intron 3, and a SNP located in exon 4). A linkage disequilibrium analysis revealed that the nine IL1R1 variants were in strong linkage disequilibrium with each other, as were the eight variants within IL1A, IL1B, and IL1RN. However, IL1R1 showed only weak linkage disequilibrium with the IL1A-IL1B-IL1RN complex, and so these were investigated separately. A haplotype analysis of the nine IL1R1 variants provided no evidence for association (P >0.05) in either the Bristol or the London cases. However, an analysis of the eight variants within the IL1A-IL1B-IL1RN complex provided evidence for an associated risk haplotype in both the Bristol (P = 0.00043) and the London (P = 0.02) cases. In addition, a protective haplotype was also identified in this complex in the Bristol (P = 0.0036) and the London (P = 0.0000008) cohorts. These associations were not restricted by sex.
Following the original publications, the U.K. and the Netherlands groups both expanded their studies using each other's findings.23 The U.K. group investigated hip OA, ascertained by joint replacement. Although the case cohort was small, with only 22 individuals, the use of the eight variants within the IL1A-IL1B-IL1RN complex provided very strong evidence for association (P = 0.0000003). The Netherlands group increased the number of variants genotyped, thus enabling them to do a comprehensive haplotype analysis akin to the U.K. group. This enabled the Netherlands group to increase the evidence for their hip association. The use of complex haplotypes therefore enabled both groups to identify alleles within the IL1 gene cluster that increase the risk of hip OA.
MATN3: Chromosome 2p24.1
As noted earlier, a linkage to chromosome 2p has been reported in an Icelandic hand OA cohort.8 This linkage encompassed MATN3, which encodes matrilin 3, an oligomeric protein present in cartilage ECM. The Icelandic group screened the exons and promoter of this candidate gene for common variants in 76 patients and 18 control individuals. Seven SNPs and one indel were identified. The six variants with frequencies greater than 0.05 were genotyped in a larger cohort of 745 patients and 368 controls. Only one variant showed a significantly greater frequency (P <0.05) in patients versus controls: a nonsynonymous SNP within exon 3 that is predicted to encode the substitution of a threonine by a methionine in the first epidermal growth factor domain of matrilin 3. A subsequent genotyping of a total of 2162 patients and 873 controls reaffirmed the association. However, when the original linkage analysis for chromosome 2p was performed following removal of the mutation carriers, the LOD score remained relatively high, at 3.8. This suggests that other variants, either within the regulatory elements or noncoding regions of MATN3 or within other nearby genes, must be coding for a significant proportion of the chromosome 2p OA susceptibility. This story is, therefore, still a work in progress.
IL4R: Chromosome 16p12.1
As noted above, an Oxford group has reported a linkage to chromosome 16p in affected sibling-pair families containing females with hip OA. A search of public databases within the linkage interval highlighted the IL-4 receptor -chain gene IL4R (16p12.1) as a strong candidate.14 IL-4 is a pleiotropic cytokine that is expressed, along with its receptor, in many cell types including adult articular cartilage. Cartilage integrity is partly regulated by mechanotransduction, and IL-4 and its receptor have a pivotal role in the cartilage chondrocyte response to mechanical signals.24 Molecular genetic analyses of IL4R have identified several common coding polymorphisms. As a first stage in the analysis of IL4R, the Oxford group genotyped nine common IL4R SNPs in the 146 female-hip probands from the families that had provided the 16p linkage and in 399 age-matched female controls.25 Two of the nine SNPs were located in the IL4R promoter, with the remaining seven located in the coding sequence (six nonsynonymous and one synonymous). Two nonsynonymous SNPs were associated in the 146 probands at P <0.05. These two SNPs were also associated in an independent cohort of 310 females with hip OA, as were two other SNPs, while a third approached significance (P = 0.07). Five of the nine variants therefore showed some evidence for association to hip OA in female Caucasians. These five positive SNPs defined two distinct groups, with members of each group being in relatively strong linkage disequilibrium with each other. Possessing a copy of an associated allele from both SNP groups was a particular risk factor, with an odds ratio (OR) of 2.4 (95% confidence interval [CI] 1.5 4.1) and a P-value of 0.0008.
FRZB: Chromosome 2q32.1
The chromosome 2q locus identified by the Oxford genome-wide linkage scan mapped to 2q24.3-q31.1.11 This linkage was restricted to affected sibling-pair families concordant for hip OA and encompassed eight plausible candidate genes: the TNF- -induced protein 6 gene TNFAIP6, the activin A receptor gene ACVR1, the fibroblast activation protein gene FAP, the integrin alpha 6 gene ITGA6, the activating transcription factor 2 gene ATF2, the integrin alpha 4 gene ITGA4, the secreted frizzled-related protein 3 gene SFRP3 (more commonly termed FRZB),
The FRZB gene product acts as an antagonist of extracellular Wnt ligands.27 The Wnt signaling pathway has a crucial role in chondrogenesis and secreted frizzled-related protein 3, which is synthesised by adult articular chondrocytes, has been shown to control chondrocyte maturation.28 Wnt signaling regulates the accumulation of cytoplasmic -catenin. In the absence of Wnt, the -catenin is rapidly degraded, whereas in the presence of Wnt, the -catenin accumulates, is translocated to the nucleus, and instigates gene transcription. The ability of wild-type secreted frizzled-related protein 3 and of the Arg200Trp and Arg324Gly substituted proteins to antagonise Wnt-signalling was therefore investigated by transient transfection of HEK293 cells. Whereas the wild-type protein efficiently inhibited Wnt activity, the Arg324Gly substitution and the Arg200Trp/ Arg324Gly double substitution had diminished activity. Similarly, HEK293 cells transfected with the plasmid containing the Arg324Gly substitution required higher levels of expressing plasmid to modestly decrease free cytosolic and nuclear levels of -catenin. These results clearly demonstrated that the conserved arginines are functionally important, with their substitution reducing the ability of secreted frizzled-related protein 3 to antagonize Wnt signaling.
A Netherlands group very recently genotyped the two FRZB SNPs in a random sample of 1,369 subjects from a population-based cohort scored for radiographic OA in the hip, hand, spine, and knee and in a patient population of 191 ASPs with symptomatic OA at multiple sites.29 Neither SNP demonstrated association in subjects with hip OA. However, the G-allele of the Arg324Gly SNP was associated (P <0.05) in individuals with a generalized OA phenotype. This phenotype constituted OA in at least two of four joint sites (hand, knee, hip, and spine). This is potentially a very important report as it may represent an independent replication, albeit in generalized OA rather than in hip OA, of the original FRZB association. Replicating associations for complex traits is extremely important in that it not only helps to distinguish true positives from false positives but also provides information regarding the global relevance of a reported find.
LRP5: Chromosome 11q13.2
The chromosome 11 linkage reported by the Oxford and Israel groups encompasses another member of the Wnt-signaling pathway: the low-density lipoprotein receptor-related protein 5 gene LRP5. The musculoskeletal community has subjected this gene to considerable investigation since its identification as the susceptibility locus for the osteoporosis-pseudoglioma syndrome30 and for a high bone mass phenotype.31 A U.K. group genotyped five LRP5 SNPs in a cohort of 268 cases with knee OA defined using ACR criteria and in 187 controls.32 No SNP showed evidence of association. However, a haplotype analysis revealed significant differences in haplotype frequencies between the cases and the controls. Since the samples sizes used in this study are relatively small, the results are best considered at this stage as a preliminary indication of an LRP5 association with OA.
BMP5: Chromosome 6p12.1
The Oxford chromosome 6 linkage was centered at 6p12.3-q13 and was restricted to affected sibling-pair families containing females with hip OA.13 The linkage encompassed two strong candidate genes: BMP5 (6p12.1), which encodes bone morphogenetic protein 5, and COL9A1 (6q13), which encodes the 1 polypeptide chain of type IX collagen. Bone morphogenetic protein 5 is a regulator of articular chondrocyte development,33 whereas type IX collagen is a quantitatively minor cartilage collagen required for maintaining cartilage integrity.34 The Oxford group detected all common coding polymorphisms within the two genes but none was associated with OA.13,35 They subsequently genotyped a relatively high density of microsatellite markers within the linked interval, at an average marker interval of 0.36 cM.36 Linkage was confirmed, with a LOD score of 4.8. When each marker was tested for association, a marker within intron 1 of BMP5 was associated (P <0.05), as were two markers located immediately downstream of the gene. Mouse studies have revealed that the regulation of the expression of BMP5 is complex and involves a number of cis elements, some of which can reside at some distance from the gene.37 The Oxford group therefore concluded that variation in cis regulatory elements of BMP5 that influence expression of the gene, as opposed to variants that result in nonsynonymous changes, which they had previously examined and excluded, could account for the linkage result. Any such cis variants have not yet been identified, so this conclusion must be considered speculative at this stage.
Candidate gene studies within regions highlighted by genome-wide linkage scans are beginning to yield OA-associated genes. However, several outstanding issues need
Osteoarthritis Susceptibility Genes Detected via Genome-Wide Association Studies
Genome-wide linkage scans have been the bedrock for the genetic analysis of common diseases. They provide broad but manageable genomic intervals that can then be interrogated in detail. Their time, however, is drawing to an end, with genome-wide association studies offering a more comprehensive and powerful alternative. This has been driven by major advances in bioinformatics, in the categorization of all common DNA variants within the human genome, in the development of relatively cheap and reliable genotyping platforms, and in the development of robust statistical tools for analyzing the enormous amount of data generated by large association studies. For OA, there has so far been one report of a genome-wide association scan, from a large collaborative group based in Japan. Two exciting finds have emerged from this scan, namely, associations to the asporin gene ASPN38 and to the calmodulin 1 gene CALM1.39
Asporin is an ECM macromolecule belonging to the small leucine-rich proteoglycan (SLRP) protein family, other members of which include decorin, biglycan, fibromodulin, and chondroadherin.40,41 SLRP family members are able to bind other structural components of the ECM, such as collagen, as well as growth factors that temporarily reside in the ECM, such as transforming growth factor (TGF- ). Asporin is expressed in a number of tissues, including adult articular cartilage. The ASPN gene comprises eight exons and resides on chromosome 9q22.31, a location that had not previously been reported to harbor OA susceptibility in any of the genome-wide linkage scans conducted on European and U.S. families. The Japanese group initially demonstrated that ASPN was expressed in OA articular cartilage. They then sequenced the ASPN exons and flanking regions and identified 21 DNA variants. Eight variants had frequencies greater than 5% and were subsequently genotyped in a population-based cohort of 371 Japanese, comprising 137 individuals (mean age of 75.3, 72% female) diagnosed as having radiographic knee OA and 234 individuals (mean age of 73.6, 61% female) diagnosed radiographically as unaffected. Only one of the eight variants demonstrated association: a triplet repeat within exon 2, coding for a polymorphic stretch of aspartic acid residues in the N-terminal region of asporin. This repeat polymorphism (given the moniker D-repeat after the one-letter code for aspartic acid) had ten alleles encoding 10-19D residues. The D14 allele was more common in the knee OA individuals than the unaffected individuals (P = 0.0013). The D-repeat was then genotyped in a Japanese case-control cohort of 393 cases with knee OA (mean age of 72.5, 84% female) and 374 controls (mean age of 28.8, 56% female). The D14 allele was also associated in this cohort (P = 0.018). Combining the two cohorts generated a P-value of 0.00024 and an OR for the D14 allele of 1.87 (95% CI 1.3 2.6). The investigators subsequently genotyped the D-repeat in 593 Japanese individuals with hip OA (mean age of 58.3, 93% female). Again, the D14 allele was associated (P = 0.0078). As well as the association to D14, the investigators also noticed that one allele, D13 (i.e., encoding one aspartic acid residue fewer than D14), was consistently underrepresented in the affected individuals. It appeared therefore that an OA-susceptibility allele (D14) and an OA-protective allele (D13) had been detected.
The investigators finally conducted a number of functional studies. These revealed that asporin inhibited the expression of the AGC1 and COL2A1 genes, which code for aggrecan and type II collagen the principal structural components of cartilage ECM. They also demonstrated that TGF- induces transcription of AGC1 and COL2A1 and that asporin interacts with TGF- and inhibits its signaling effect. This inhibitory effect was particularly strong for asporin encoded by the D14 allele and significantly less so for D13-encoded protein. These functional studies therefore provide a plausible model of how the D-repeat polymorphism of ASPN influences susceptibility to OA: firstly, asporin inhibits TGF- signaling and therefore indirectly regulates the synthesis of aggrecan and type II collagen, critical components of articular cartilage ECM; secondly, this inhibition is strongest for the D14 allele, leading to insufficient quantities of these proteins and therefore a cartilage that is structurally compromised; and finally, D13-encoded asporin has the weakest TGF- inhibitory effect, resulting in a more structurally resilient cartilage. What still needs determining is exactly how the size of the D-repeat influences protein activity: is the effect via influences on the conformational structure of the protein or does the repeat itself bind directly to TGF- ? Another important question is whether asporin modulates the signalling of other members of the TGF- superfamily such as the bone morphogenetic proteins, which are also regulators of cartilage development and maintenance.
Calmodulin is an intracellular protein that binds to Ca2+ and interacts with a number of cellular proteins.42,43 The Japanese group initially identified an association in patients with hip OA to a SNP within intron 3 of the
Of the five associated SNPs that were in linkage disequilibrium, the core promoter SNP was considered the most likely to have a functional effect on calmodulin 1. The investigators demonstrated that CALM1 was expressed in human articular chondrocytes and also showed higher levels of CALM1 expression in OA cartilage compared to normal cartilage. They subsequently assessed the effect that the two alleles of the core promoter SNP had on the expression of CALM1. This revealed that the associated allele (the T-allele) resulted in reduced transcriptional activity relative to the unassociated allele (the C-allele). The investigators then demonstrated that calmodulin 1 increases the expression of the aggrecan gene AGC1 and of the type II collagen gene COL2A1. These functional studies provide a model of how the core promoter SNP of CALM1 could influence OA susceptibility: firstly, calmodulin 1 naturally increases the synthesis of aggrecan and type II collagen in articular cartilage; and secondly, this synthesis is reduced for the T-allele, particularly in those individuals who are TT homozygotes, leading to insufficient quantities of aggrecan and type II collagen and therefore a structurally compromised cartilage.
Since calmodulin 1 and asporin both regulate the expression of AGC1 and COL2A1, the investigators finally assessed whether the associated allele at CALM1 (the T-allele of the core promoter SNP) and the risk allele at ASPN (the D14-allele of the D-repeat) interacted with each other in an epistatic manner to further increase the risk of developing OA. This revealed that individuals who had inherited two copies of the T-allele and at least one copy of the D14-allele were at a particularly high risk of hip OA, with an OR of 13.16 (95% CI 1.66 104.06). This makes sense, since both the T-allele of CALM1 and the D14-allele of ASPN lead to a reduction in expression of AGC1 and COL2A1. It should be noted however that the broad confidence interval of the OR means that this result is of low statistical certainty.
The Japanese association study has so far identified two very interesting genes associated with the development of large joint OA. Other OA susceptibility loci are likely to follow from this study, since of the 71,880 SNPs genotyped by the Japanese group, 2,219 demonstrated associations at P <0.01.39 The majority of these will be false positives, but as the Japanese group works through these SNPs by genotyping additional cohorts, the genuine positives will emerge. It will be intriguing to see what pathways and mechanisms these highlight, and to note whether the associations to the IL1 gene cluster, MATN3, IL4R, FRZB, LRP5, and BMP5, reported by European groups are observed in the Japanese study. It will also be interesting to determine whether the ASPN and CALM1 associations detected in Japan have a role in OA development outside of Asia.
Osteoarthritis Susceptibility Genes Detected via Gene Expression Analysis
An alternative strategy to genome-wide linkage or association studies for detecting functional candidates is the identification of those genes that are significantly up- or downregulated in disease versus normal tissue; a difference in expression would imply a role in either disease development or disease progression. Identified genes can then be tested for association. A group based at St. Thomas Hospital in London recently carried out such an analysis for OA.44 Using four cDNA libraries constructed from normal cartilage, OA cartilage, normal synovium, and OA synovium, the investigators initially identified 54 genes that showed differential expression between normal and diseased tissue. The investigators next tested the genes for association to OA. They focused on 12 of the 54 genes, with the selection criteria being those that code for proteins whose actions could be therapeutically modified, such as receptors, enzymes, and secreted molecules. They also investigated 12 additional genes that other groups had previously suggested could have a role in OA aetiology. The cohort for the association analysis comprised 749 females (age range of 43-67) who were participating in the U.K. Chingford Study, a population-based investigation of joint diseases. The individuals had undergone a radiographic assessment, including measures of JSN and the occurrence of osteophytes, of knee OA in 1988-1989 and then a decade later. Of the 749 females, 469 were classified as normal (they did not have OA in either knee at both examinations) and 280 were classified as affected. Twenty-six SNPs from the 24 genes were genotyped in the cohort and were tested for association to both OA prevalence and OA progression. SNPs from four genes were associated with OA prevalence: BMP2 (chromosome 20p12.3), which encodes bone morphogenetic protein 2; CD36 (7q21.11), which encodes a thrombospondin and collagen receptor; COX2 (1q25), which encodes a cyclooxygenase; and NCOR2 (12q24.31), which encodes a nuclear receptor co- repressor. Four genes were associated with OA progression: CILP (15q22.31), which encodes a cartilage intermediate-layer protein; OPG (8q24.12), which encodes osteoprotegerin; TNA (3p21.31), which encodes tetranectin; and ESR1 (6q25.1), which encodes estrogen receptor . One gene was associated with prevalence and progression: ADAM12 (chromosome 10q26.2), which encodes a metalloproteinase. When the association data were corrected for the large number of tests performed, only ADAM12 was still associated, with a corrected P of 0.014.
TABLE 6-2 GENES REPORTED TO SHOW EVIDENCE OF ASSOCIATION TO OA
This study has demonstrated an alternative strategy for identifying potential OA susceptibility genes. It has some weaknesses, including the small number of variants tested per gene and the fact that the cDNA libraries used did not provide a complete coverage of all genes. Despite this, a novel locus in ADAM12 has been identified that merits more comprehensive investigation.
Considerable progress has been made in recent years in our understanding of the molecular genetic basis of primary OA. Using a variety of techniques and strategies, a number of genes have been identified that are likely to harbor risk alleles for this common arthritis (Table 6-2).
Many of the linkage and association studies do, however, have weaknesses, principally the use of relatively small cohorts and the study of only a small proportion of the genetic variants within the gene of interest. These are potentially major deficiencies. The use of small sample sizes limits the power of the study to detect subtle effects and can increase the likelihood of detecting a false positive. The genotyping of only a proportion of the common gene variants provides only a limited view of the gene and may well lead to loci being disregarded when in fact they do harbor susceptibility. The OA research community needs to be more comprehensive in its approach to mapping OA susceptibility loci. The Japanese reports of associations to ASPN and CALM1 provide a model for others to follow: large sample sizes, all common variants interrogated, and subsequent functional studies to investigate the effect of associated variants on gene activity or protein function.
Of the genes that have so far been implicated as OA susceptibility loci, only two code for structural components of articular cartilage: MATN3, which encodes matrilin 3, and ASPN, which encodes asporin. The majority of the associated genes code for proteins that regulate joint tissue biology, either as signaling molecules, receptors, or enzymes. This is an important observation as it implies that effects on joint tissue development, maintenance, and homeostasis are the likely paths through which OA genetic susceptibility is acting. Such paths are likely to be more accessible to modification and treatment than are structural defects. One of the major challenges now is how this genetic insight can be used to enhance the care and treatment of patients at risk of developing the disease.
We are grateful to Research into Ageing and the Arthritis Research Campaign who fund our research.
1. Loughlin J. Genetic epidemiology of primary osteoarthritis. Curr Opin Rheum.13:111 116, 2001.
2. Loughlin J. Genome studies and linkage in primary osteoarthritis. Rheum Dis Clin North Am 28:95 109, 2002.
3. Jakkula E, Melkoniemi M, Kiviranta I, et al. The role of sequence variations within the genes encoding collagen II, IX and XI in non-syndromic, early-onset osteoarthritis. Osteoarthritis Cartilage 13:497 507, 2005.
4. Bergink AP, van Meurs JB, Loughlin J, et al. Estrogen receptor gene haplotype is associated with radiographic osteoarthritis of the knee in elderly men and women. Arthritis Rheum 48:1913 1922, 2003.
5. Chapman K, Mustafa Z, Irven CM, et al. Osteoarthritis-susceptibility locus on chromosome 11q, detected by linkage. Am J Hum Genet 65:167 174, 1999.
6. Loughlin J, Mustafa Z, Irven C, et al. Stratification analysis of an osteoarthritis genome screen suggestive linkage to chromosomes 4, 6 and 16. Am J Hum Genet 65:1795 1798, 1999.
7. Lepp vuori J, Kujala U, Kinnunen J, et al. Genome scan for predisposing loci for distal interphalangeal joint osteoarthritis: evidence for a locus on 2q. Am J Hum Genet 65:1060 1067, 1999.
8. Stef nsson SE, J nsson H, Ingvarsson T, et al. Genomewide scan for hand osteoarthritis: a novel mutation in Matrilin-3. Am J Hum Genet 72:1448 1459, 2003.
9. Demissie S, Cupples LA, Myers R, et al. Genome scan for quantity of hand osteoarthritis. Arthritis Rheum 46:946 952, 2002.
10. Hunter DJ, Demissie S, Cupples LA, et al. A genome scan for joint-specific hand OA susceptibility. Arthritis Rheum 50:2489 2496, 2004.
11. Loughlin J, Dowling B, Mustafa Z, et al. Refined linkage mapping of a hip osteoarthritis susceptibility locus on chromosome 2q. Rheumatology 41:955 956, 2002.
12. Chapman K, Mustafa Z, Dowling B, et al. Finer linkage mapping of primary hip osteoarthritis susceptibility on chromosome 11q in a cohort of affected female sibling pairs. Arthritis Rheum 46:1780 1783, 2002.
13. Loughlin J, Mustafa Z, Dowling B, et al. Finer linkage mapping of a primary hip osteoarthritis susceptibility locus on chromosome 6. Eur J Hum Genet 10:562 568, 2002.
14. Forster T, Chapman K, Marcelline L, et al. Finer linkage mapping of primary osteoarthritis susceptibility loci on chromosomes 4 and 16 in families with affected women. Arthritis Rheum 50:98 102, 2004.
15. Southam L, Dowling B, Ferreira A, et al. Microsatellite association mapping of a primary osteoarthritis susceptibility locus on chromosome 6p12.3-q13. Arthritis Rheum 50:3910 3914, 2004.
16. Alizadeh BZ, Njajou OT, Bijkerk C, et al. Evidence for a role of the genomic region of the gene encoding for the 1 chain of type IX collagen (COL9A1) in hip osteoarthritis. Arthritis Rheum 52:1437 1442, 2005.
17. Meenagh GK, McGibbon D, Nixon J, et al. Lack of support for the presence of an osteoarthritis susceptibility locus on chromosome 6p. Arthritis Rheum 52:2040 2043, 2005.
18. Kalichman L, Kobyliansky E, Malkin I, et al. Search for linkage between hand osteoarthritis and 11q12-13 chromosomal segment. Osteoarthritis Cartilage 11:561 568, 2003.
19. Pelletier J-P, Martel-Pelletier J, Abramson SB. Osteoarthritis, an inflammatory disease. Arthritis Rheum 44:1237 1247, 2001.
20. Goldring MB. The role of the chondrocyte in osteoarthritis. Arthritis Rheum 43:1916 1926, 2000.
21. Meulenbelt I, Seymour AB, Nieuwland M, et al. Association of the interleukin-1 gene cluster with radiographic signs of osteoarthritis of the hip. Arthritis Rheum 50:1179 1186, 2004.
22. Smith AJP, Keen LJ, Billingham MJ, et al. Extended haplotypes and linkage disequilibrium in the IL1R1-IL1A-IL1B-IL1RN gene cluster: association with knee osteoarthritis. Genes Immun 5: 451-460, 2004.
23. Smith AJP, Elson CJ, Perry MJ, et al. Accuracy of haplotype association studies is enhanced by increasing number of polymorphic loci examined: comment on the article by Meulenbelt et al. Arthritis Rheum 52:675, 2005.
24. Salter DM, Millward-Sadler SJ, Nuki G, et al. Integrin-interleukin-4 mechanotransduction pathways in human chondrocytes. Clin Orthop 391:S49-S60, 2001.
25. Forster T, Chapman K, Loughlin J. Common variants within the interleukin 4 receptor gene (IL4R) are associated with susceptibility to osteoarthritis. Hum Genet 114:391 395, 2004.
26. Loughlin J, Dowling B, Chapman K, et al. Functional variants within the secreted frizzled-related protein 3 gene are associated with hip osteoarthritis in females. Proc Natl Acad Sci USA 101:9757 9762, 2004.
27. Jones SE, Jomary C. Secreted frizzled-related proteins: searching for relationships and patterns. Bioessays 24:811 820, 2002.
28. Enomoto-Iwamoto M, Kitagaki J, Koyama E, et al. The wnt antagonist frzb-1 regulates chondrocyte maturation and long bone development during limb skeletogenesis. Dev Biol 251:142 156, 2002.
29. Min JL, Meulenbelt I, Riyazi N, et al. Association of the frizzled-related protein gene with symptomatic osteoarthritis at multiple sites. Arthritis Rheum 52:1077 1080, 2005.
30. Gong Y, Slee RB, Fukai N, et al. LDL receptor-related protein 5 (LRP5) affects bone accrual and eye development. Cell 107: 513-523, 2001.
31. Little RD, Carulli JP, Del Mastro RG, et al. A mutation in the LDL receptor-related protein 5 gene results in the autosomal dominant high-bone-mass trait. Am J Hum Genet 70:11 19, 2002.
32. Smith AJP, Gidley J, Sandy JR, et al. Haplotypes of the low-density lipoprotein receptor-related protein 5 (LRP5) gene: are they a risk factor in osteoarthritis? Osteoarthritis Cartilage 13:608 613, 2005.
33. Bail n-Plaza A, Lee AO, Veson EC, et al. BMP-5 deficiency alters chondrocytic activity in the mouse proximal tibial growth plate. Bone 24:211 216, 1999.
34. Eyre D. Collagen of articular cartilage. Arthritis Res 4:30 35, 2002.
35. Southam L, Chapman K, Loughlin J. Genetic association analysis of BMP5 as a potential osteoarthritis susceptibility gene. Rheumatology 42:911 912, 2003.
36. Southam L, Dowling B, Ferreira A, et al. Microsatellite association mapping of a primary osteoarthritis susceptibility locus on chromosome 6p12.3-q13. Arthritis Rheum 50:3910 3914, 2004.
37. DiLeone RJ, Marcus GA, Johnson MD, et al. Efficient studies of long-distance Bmp5 gene regulation using bacterial artificial chromosomes. Proc Natl Acad Sci USA 97:1612 1617, 2000.
38. Kizawa H, Kou I, Iida A, et al. An aspartic acid repeat polymorphism in asporin inhibits chondrogenesis and increases susceptibility to osteoarthritis. Nat Genet 37:138 144, 2005.
39. Mototani H, Mabuchi A, Saito S, et al. A functional single nucleotide polymorphism in the core promoter region of CALM1 is associated with hip osteoarthritis in Japanese. Hum Mol Genet 14:1009 1017, 2005.
40. Lorenzo P, Aspberg A, nnerfjord P, et al. Identification and characterization of asporin. J Biol Chem 276:12201 12211, 2001.
41. Henry SP, Takanosu M, Boyd TC, et al. Expression pattern and gene characterization of asporin. J Biol Chem 276:12212 12221, 2001.
42. Chin D, Means AR. Calmodulin: a prototypical calcium sensor. Trends Cell Biol 10:322 328, 2000.
43. Haeseleer F, Imanishi Y, Sokal I, et al. Calcium-binding proteins: intracellular sensors from the calmodulin superfamily. Biochem Biophys Res Commun 290:615 623, 2002.
44. Valdes AM, Hart DJ, Jones KA, et al. Association of candidate genes for the prevalence and progression of knee osteoarthritis. Arthritis Rheum 50:2497 2507, 2004.