WO2000047770A1

WO2000047770A1 - Susceptibility locus for osteoarthritis

Info

Publication number: WO2000047770A1
Application number: PCT/GB2000/000493
Authority: WO
Inventors: Bryan Sykes; John Loughlin; Andrew Carr
Original assignee: Catalyst Biomedica Ltd
Current assignee: Catalyst Biomedica Ltd
Priority date: 1999-02-15
Filing date: 2000-02-14
Publication date: 2000-08-17
Anticipated expiration: 2001-08-15

Abstract

It is known that various loci within the genome affect the susceptibility of an individual to osteoarthritis. The present invention relates to the identification of a genetic region that may contain an osteoarthritis susceptibility locus. A genome-wide linkage analysis was carried out using families affected by osteoarthritis. Results were stratified according to sex and joint affected. This produced evidence for linkage of markers on chromosome 6 between 6p21.1 and 6q22, indicating the presence of an osteoarthritis susceptibility locus in this chromosomal region.

Description

SUSCEPTIBILITY LOCUS FOR OSTEOARTHRITIS

Technical Field

This invention relates to the identification of chromosomal regions on chromosome 6 linked to genetic sequences which affect susceptibility to osteoarthritis.

Background of Invention

Osteoarthritis (OA) is a debilitating disease involving degeneration of the articular cartilage of synovial joints^5,6. Although OA has long been considered an inevitable consequence of ageing, it has become increasingly apparent that OA does have a genetic component. Early-onset forms of the disease are associated with several osteochondrodysplasias, rare diseases involving abnormal bone and cartilage development that are transmitted as Mendelian traits⁷. However, the OA in these conditions is secondary to the main dysplastic phenotype. The common late-onset form of the disease (idiopathic OA) often has no obvious environmental (i.e. injury) or characteristic physical cause and does not demonstrate a clear mode of inheritance . Twin and segregation studies demonstrate the role of genetics in the development of the disease, with estimates of the eritability of OA of the hand, knee and hip ranging from 36% to 68% ^x' ⁹' ²⁴.

Over the last 10 years, many genes for single gene or monogenic diseases, which are relatively rare in the population, have been positioned by linkage analysis in families, and localised to a small enough region to allow identification of the gene. The latter sublocalisation and fine mapping can be carried out in single gene rare diseases because recombinations within families define the boundaries of the minimal interval beyond any doubt. In contrast, in common diseases such as osteoarthritis, diabetes or asthma the presence of the disease mutation does not always coincide with the development of the disease: disease susceptibility mutations in common disorders provide risk of developing of the disease, and this risk is usually much less than 100%. Hence, susceptibility genes in common diseases cannot be localised using recombination events within families, unless tens of thousands of families are available to fine map the locus. Because collections of this size are impractical, investigators are contemplating the use of association mapping, which relies on historical recombination events during the history of the population from which the families came from.

Association mapping has been used in over a dozen examples of rare single gene traits, and particularly in genetically isolated populations such as Finland to fine map disease mutations. Nevertheless, association mapping is fundamentally different from straightforward linkage mapping because even though the degree of association between two markers or a marker and a disease mutation is proportional to the physical distance along the chromosome this relationship can be unpredictable because it is dependent on the allele frequencies of the markers, the history of the population and the age and number of mutations at the disease locus . For rare, highly penetrant single gene diseases there is usually one major founder chromosome in the population under study, making it relatively feasible to locate an interval that is smaller than one that can be defined by standard recombination events within living families. The resolution of this method in monogenic diseases in which there is one main founder chromosome is certainly less than 2cM, and in certain examples the resolution is down to 100 kb of DNA (Hastbacka et al . (1994) Cell 78,1-20).

In common diseases like OA, which are caused by a number of genes or polygenes acting together in concert the population frequency of the disease allele may be very high, perhaps exceeding 50%, and there are likely to be several founder chromosomes, all of which impart risk, and not a 100% certainty of disease development. Because association mapping is dependent on unpredictable parameters, and because founder chromosomes will be several and common in frequency in the general population, the task of fine mapping polygenes is currently one of some controversy, and many doubt the feasibility at all of a systematic genetic approach using a combination of linkage and association mapping. Recently, Risch and Marakandis have provided some mathematical background to the feasibility of association mapping in complex diseases ( Science 273 1516-1517, 1996) but they did not take into account the effect of multiple founder chromosomes .

It has often been noted in epidemiological studies that there is a female preponderance for OA^5,6. This may be accounted for by differential effects on the two sexes of environmental factors. However, a Finnish twin study has suggested that genetic susceptibility may be greater in women than men⁹. This result has recently been supported by a segregation analysis³. Not only have differences in risk between females and males been reported but it has also been suggested that there are differences in heritability between joint groups²'^10,11. These differences could be the result of genetic locus heterogeneity between the different joints.

Summary of the Invention

The present inventors have identified a region on chromosome

6 that may a harbour susceptibility locus for OA.

This region was identified following a two-stage non- parametric linkage analysis. The first stage involved a random screen of the genome using 272 microsatellite markers in 297 OA families. The second stage was more selective and involved genotyping an additional 184 families for those markers that demonstrated moderate evidence of linkage in stage 1. This revealed three micro satellites on chromosome 11 and one on chromosome 2 for which the evidence for linkage increased as the number of families studied increased.

Epidemiological, twin-pair and segregation studies²'⁹'^10,22 have demonstrated apparent differences in the heritability of OA between males and females, and between different joint groups^{10, n}' ⁹'²' ³. These categories therefore merit independent analysis in linkage and association studies. In our previous analyses of chromosomes 2q and 11 (see co-pending applications PCT/GB99/03264 and PCT/GB99/03267) , we determined that the suggestive linkage of 2q was of greater impact in hip disease than in knee disease whereas the 11 linkage was restricted to females with hip disease. We therefore studied the results of our whole genome screen using stratification analysis, to investigate whether a susceptibility locus may be obscured in the unstratified data set .

Surprisingly, stratification has also highlighted suggestive linkage to the chromosomal region from 6p21.1 to 6q22.1 in pairs with hip OA, with a corrected MLS of 2.1.

Chromosome 6 has suggested evidence of linkage in hip-only pairs with a MLS of 2.1 between D6S257 and D6S262. Over 50cM of this chromosome has a MLS > 1.5 in this stratum. This region of the genome encompasses the C0L9A1 gene (6ql2-6ql3) . We have demonstrated suggestive linkage of this gene to OA in families with hip-only disease and in female-hip pairs who were drawn from our cohort of 481 families. COL9A1 maps within the 11 cM between D6S257 and D6S286 and encodes the αl chain of type IX collagen. This collagen is a quantitatively minor heterotrimeric FACIT collagen found in cartilage²⁰. Type IX collagen decorates and interacts with the type II collagen fibrils. Two mouse models have demonstrated that this gene is essential for normal cartilage development and that mutations result in an OA phenotype . In the first model, a truncated form of the gene resulted in a mild osteo- chondrodysplasia with OA²⁵ whilst in the second model, a knock-out mouse had no congenital abnormality but developed a severe OA that was comparable in timing and pathology to human primary OA²². COL9A1 is therefore a strong candidate for OA and should be analysed by association analysis for potential predisposing DNA variants . The genes that encode the α2 and α3 chains of type IX collagen, COL9A2 and COL9A3 , map to chromosomes lp32 and 20ql3.3, respectively. There was no evidence of linkage to these two chromosomal regions in any of the six strata that we tested. Furthermore, we had previously targeted COL9A2 as a candidate locus but did not obtain evidence for linkage. It is possible that COL9A2 and COL9A3 are susceptibility genes but are not, by chance, of major impact in the cohort we have screened. Alternatively, COL9A1 may be the only type IX collagen gene that is an OA susceptibility gene. If this is the case functional differences between the αl(IX) polypeptide chain and the α2(IX) and α3 (IX) chains may be the basis for the susceptibility. The αl(IX) chain is distinct from the α2(IX) and α3(IX) chains in that it contains a large cartilage- specific NC4 domain that interacts with additional molecules if the extracellular matrix. This may be a site harbouring OA susceptibility.

Open reading frames (ORFs) are stretches of genetic sequence which are candidates for being expressed genes. They can be identified by the presence of elements in the primary sequence which are characteristic of coding sequence, such as sequence elements from exon-intron boundaries, transcriptional initiation and termination motifs and start and stop codons . Because large amounts of sequence can be screened rapidly for these elements, the identification of ORFs is commonly an initial step in the discovery of novel genes. ORFs can also be identified through comparing sequence from the region of interest with sequence in Expressed Sequence Tag (EST) databases.

Microsatellite marker loci are designated in this application according to the nomenclature conventional in the field of Human Genetics . This provides a unique designation which specifically and unambiguously identifies each marker locus. Mapping data for any particular marker locus is readily available from conventional sources, such as the Gopher server at the Human Genome Mapping Project Resource Centre (Host = gopher . hcmτp .mrc .ac.uk; Port = 70 or URL: gopher : //gopher .hgmp.mrc. ac.us : 70/ or by anonymous ftp from ftp.hgmp .mrc .uk: /Oxford_Primers ) .

A region which is described as 'adjacent' to a microsatellite locus may be within about 1000Kb of the microsatellite, preferably within about 500Kb away, and more preferably within about 100Kb of the microsatellite.

Cytogenetic positions on chromosomes are referenced according to international convention (see Paris Convention 1971) . The arms of a chromosome are designated p and q and each arm is then divided into sections which are given numbers . The sections are then subdivided, with each subsection also being numbered. Sub-sections may be sub-divided further. Regions of a chromosome can be referenced by quoting the chromosome number, the arm and then the section and sub-section numbers.

Numerous polymorphic markers in the chromosomal region from 6p21.1 to 6q22.1 are known and are suitable for use in methods of the present invention. Especially preferred examples include D6S1610,D6S257, D6S286, D6S462 and D6S262. However, those of skill in the art will appreciate that other microsatellite markers within this chromosomal region may be used in an equivalent manner to that described herein for these preferred markers. Similarly, single nucleotide polymorphisms (SNPs) within the region may be used as markers in an equivalent manner.

A first aspect of the present invention is a method for identifying individuals susceptible to osteoarthritis comprising obtaining a sample of genomic DNA and analysing alleles of polymorphic markers located in the chromosomal region from 6p21.1 to 6q22.1.

Another aspect of the present invention is a method for identifying individuals susceptible to osteoarthritis of the hip comprising obtaining a sample of genomic DNA and analysing alleles of polymorphic markers located in the chromosomal region from 6p21.1 to 6q22.1.

Another aspect of the present invention is a method for isolating genetic loci associated with susceptibility to OA comprising screening genomic libraries with genetic sequence derived from polymorphic markers located in the chromosomal region from 6p21.1 to 6q22.1 and identifying open reading frames in regions adjacent to said genetic sequence.

Another aspect of the present invention is a method for isolating genetic loci associated with susceptibility to OA comprising screening genomic libraries with a polymorphic markers located in the chromosomal region from 6p21.1 to 6q22.1 and identifying open reading frames located within 500 Kb, preferably within 100 Kb, more preferably within 50 Kb and most preferably within 10 Kb of the marker.

Another aspect of the present invention is a method for isolating genetic loci associated with susceptibility to OA comprising screening genomic libraries with genetic sequence derived from any one of the polymorphic marker D6S1610,D6S257, D6S286, D6S462 and D6S262 and identifying open reading frames in adjacent regions.

Another aspect of the present invention is a method for isolating genetic loci associated with susceptibility to OA comprising screening genomic libraries with any one of the polymorphic markers D6S1610,D6S257, D6S286, D6S462 and D6S262 and identifying open reading frames located within 500 Kb, preferably within 100 Kb, more preferably within 50 Kb and most preferably within 10 Kb of the marker

Another aspect of the present invention is the use of alleles of polymorphic markers located in chromosomal region from 6p21.1 to 6q22.1 as markers for the identification of loci influencing susceptibility to OA.

Another aspect of the present invention is the use of alleles of polymorphic markers located in chromosomal region from 6p21.1 to 6q22.1 as markers for the ancestral DNA variant conferring enhanced OA susceptibility.

Another aspect of the present invention is the use of alleles of polymorphic markers located in chromosomal region from 6p21.1 to 6q22.1 as markers for mapping associated loci to identify genes associated with OA susceptibility.

Another aspect of the present invention is the use of alleles of polymorphic markers located in chromosomal region from

6p21.1 to 6q22.1 in investigating loci capable of influencing susceptibility to OA.

Another aspect of the present invention is a method for mapping loci which affect susceptibility to OA by comparing a genomic region containing a particular allele of a polymorphic marker located in chromosomal region from 6p21.1 to 6q22.1 with a genomic region containing a different allele of the same marker.

Another aspect of the present invention is the use of alleles of any one of the polymorphic markers D6S1610,D6S257 , D6S286, D6S462 and D6S262 and D16S261 as a marker for the identification of loci influencing susceptibility to OA.

Another aspect of the present invention is the use of alleles of any one of the polymorphic markers D6S1610,D6S257 , D6S286, D6S462 and D6S262 as a marker for the ancestral DNA variant conferring enhanced OA susceptibility.

Another aspect of the present invention is the use of alleles of any one of the polymorphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262 as a marker for mapping associated loci to identify genes associated with OA susceptibility.

Another aspect of the present invention is the use of alleles of any one of the polymorphic markers D6S1610 , D6S257 , D6S286, D6S462 and D6S262 in investigating loci capable of influencing susceptibility to OA.

Another aspect of the present invention is a method for mapping loci which affect susceptibility to OA by comparing a genomic region containing a particular allele of any one of the polymorphic markers D6S1610 , D6S257 , D6S286, D6S462 and D6S262 with a genomic region containing a different allele of the same marker.

Another aspect of the present invention is a method for determining individual susceptibility to osteoarthritis comprising obtaining sample genomic DNA from siblings, at least two of which have clinical symptoms of osteoarthritis. analysing a region of their genomic DNA comprising a polymorphic marker located in chromosomal region from 6p21.1 to 6q22.1, identifying allele sharing between the siblings as defined by a maximum log of the odds (LOD) score of greater than 1 and a p-value of less then 0.25, and determining individual susceptibility to osteoarthritis by reference to the allele sharing.

Another aspect of the present invention is a method for determining individual susceptibility to osteoarthritis comprising obtaining sample genomic DNA from siblings, at least two of which have clinical symptoms of osteoarthritis. analysing a region of their genomic DNA comprising any one of the polymorphic markers D6S1610 , D6S257 , D6S286, D6S462 and D6S262, identifying allele sharing between the siblings as defined by a maximum log of the odds (LOD) score of greater than 1 and a p-value of less then 0.25, and determining individual susceptibility to osteoarthritis by reference to the allele sharing.

Another aspect of the present invention is a method for determining individual susceptibility to osteoarthritis comprising obtaining sample genomic DNA from siblings, at least two of which have clinical symptoms of osteoarthritis. analysing a region of their genomic DNA comprising a polymorphic marker and located within 20cM, preferably within 15 cM, or more preferably within lOcM, even more preferably within 5cM and most preferably within 2cM, of any one of the polymorphic markers D6S1610,D6S257, D6S286, D6S462 and D6S262, identifying allele sharing between the siblings as defined by a maximum log of the odds (LOD) score of greater than 1 and a p-value of less then 0.25, and determining individual susceptibility to osteoarthritis by reference to the allele sharing.

Another aspect of the present invention is a method for determining individual susceptibility to osteoarthritis comprising obtaining sample genomic DNA from siblings, at least two of which have clinical symptoms of osteoarthritis. analysing a region of their genomic DNA comprising one or more of the polymorphic markers D6S1610,D6S257, D6S286, D6S462 and D6S262 and additionally analysing one or more genomic regions comprising the polymorphic markers; D2S202, D2S325, D2S117, D3S1266, D4S415, D4S398, D4S231, D4S250, D4S406, D6S260, D6S273, D6S286, D6S281, D7S669, D7S530, D11S937, D11S907, D11S903, D11S901, D17S807, D17S789 and DXS1068, identifying allele sharing between the siblings as defined by a maximum log of the odds (LOD) score of greater than 1 and a p-value of less then 0.25, and determining individual susceptibility to osteoarthritis by reference to the allele sharing.

Another aspect of the present invention is a method for isolating genetic loci associated with susceptibility to OA comprising screening a genomic library with sequence derived from the region between polymorphic markers D6S1610 and D6S262 and identifying those isolated clones which additionally contain open reading frames

Another aspect of the present invention is a method for isolating genetic loci associated with susceptibility to OA comprising screening a genomic library with sequence derived from the region between polymorphic markers D6S257 and D6S262 and identifying those isolated clones which additionally contain open reading frames.

Another aspect of the present invention is a method for isolating genetic loci associated with susceptibility to OA comprising screening a genomic library with sequence derived from the region between polymorphic markers D6S257 and D6S286 and identifying those isolated clones which additionally contain open reading frames .

Another aspect of the present invention is a method for identifying loci conferring susceptibility to osteoarthritis comprising screening a genomic library with genetic sequence derived from one or more of the polymorphic markers D6S1610,D6S257, D6S286, D6S462 and D6S262 and identifying open reading frames located within 500 Kb, more preferably within 100 Kb, even more preferably within 50 Kb or most preferably within 10 Kb of any of these polymorphic markers .

Another aspect of the present invention is a method for identifying individuals susceptible to osteoarthritis comprising obtaining a sample of genomic DNA and analysing the COL9A1 gene for the presence of alleles associated with OA susceptibility. Another aspect of the present invention is a method for determining individual susceptibility to osteoarthritis comprising sequencing the COL9A1 gene, identifying the allelic form of the gene, correlating the allelic form of the sequenced gene to the known susceptibility to OA of individuals with that allelic form.

Where the present invention relates to the analysis of DNA of an individual, such an individual may be one who has OA, is considered at risk from OA (e.g. by having a sibling with and/or fai ily history of OA) , or may be symptomless. The DNA sample from the individual may be prepared from any convenient sample, for example from blood or skin tissue.

Where the present invention relates to screening for genetic loci, including ORFs, related to OA, such screening is generally performed on a population sample. Desirably the sample includes individuals with and without OA. More desirably the sample comprises at least a proportion of sibling pairs or larger cohorts. Typically, at least 100, preferably at least 500 or more preferably at least 1000 individuals may be sampled. Preferably at least 30%, preferably at least 50% and more preferably at least 70% will exhibit symptoms of OA. The analysis of the individuals may be stratified with respect to a number of factors, such as sex, severity, location or age of onset of disease, and the like.

Where the present invention includes the identification of an ORF, a further aspect of the invention provides isolated nucleic acid comprising the ORF, the use of the ORF in the diagnosis or prognosis of OA, a polypeptide in isolated form encoded by the ORF and the use of the polypeptide in methods of diagnosis or prognosis of OA.

Nucleic acid comprising the ORF will generally be in the form of a recombinant replicable vector, in which nucleic acid, in the form of RNA or DNA, consisting of the ORF is operably linked to a promoter. The nucleic acid may optionally include 5 ' or 3' regions to the ORF containing such things as signal sequences, tags such as polyhistidine tags to allow isolation of expressed product, 5' or 3 ' untranslated regions, and the like. The promoter will be selected to be compatible with a desired host cell, such as a bacterial, yeast, insect or mammalian host cell.

The vector may be constructed by conventional means in the art, for example by providing a pair of PCR primers to a sample of mRNA or genomic DNA, which primer pair is suitable to amplify up the ORF from the sample. The primers may incorporate restriction enzyme sites to facilitate insertion of amplified product into a suitable vector. Various expression vectors are readily available in the art from commercial sources .

Host cells may be used for the production of a protein encoded by the ORF by introducing the expression vectors into the host cell. The host cells may be cultivated using standard techniques known as such in the art under conditions! to bring about expression of the protein encoded by the ORF, and the protein recovered from the host cell or its culture medium.

Antibodies, such as monoclonal, polyclonal or synthetic (e.g. from a phage display library) may be raised or selected against the protein. The term "anitbodies" includes binding fragments thereof, such as Fab and Fv (including scFv) fragments .

ORF nucleic acid, proteins encoded by it, and antibodies thereto may be used in the diagnosis or prognosis of OA. For example, the presence of alleles of the nucleic acid in subjects with or without OA may be determined, the levels of expression of nucleic acid or the amount of protein present in cells in such subjects may also be determined. Such determination may be used to link the presence of alleles of

ORFs, or the presence or absence of particular threshold levels of expression or protein production, to the diagnosis or prognosis of OA.

Description of Drawings

Figure 1 shows a multipoint log of the odds (LOD) analysis of chromosome 6 using stratified data. The analysis shows an MLS score of about 2 for the hip only category in the region between markers D6S1610 and D4S314 and an MLS score of about 1 for the female hip category in the region between D6S422 and D6S265.

Detailed Description of Invention

Linkage Analysis

The present inventors collected 481 families in which at least two siblings have undergone joint replacement surgery of the hip or knee for severe, end-stage idiopathic OA (Table 1) . Due to the late onset of the disease, none of these families contained parents who could participate in the study. In stage 1 272 microsatellite markers were genotyped in 297 of the 481 families. The microsatellites were essentially those used by Reed et al⁸ with the replacement of certain markers that amplified unreliably. Sixteen markers from stage 1 had evidence of linkage at p≤0.05 (Table 2) . These markers were then genotyped in the remaining 184 families. None of the 16 markers had a p≤0.05 in this second stage although three had a p<0.10: D2S202 (p=0.07), D11S903 (p=0.07) and D11S901 (p=0.10) (Table 2). When the data for stages 1 and 2 were combined and compared to stage 1 only, the combined p-value decreased for 4 of the 16 markers: D2S202 (p=0.009 for combined vs. 0.036 for stage 1), D11S907 (p=0.019 for combined vs. 0.05 for stage 1), D11S903 (p=0.004 for combined vs. 0.017 for stage 1) and D11S901 (p=0.0003 for combined vs. 0.0004 for stage 1) increasing the number of families had therefore increased the evidence for linkage at these 4 markers. Further analysis of chromosomes 2 and 11 is described in co-pending applications PCT/GB99/03264 and PCT/GB99/03267.

In the light of stratification analysis of chromosomes 2 and 11, the other results from the genome screen were stratified to determine whether any genomic regions harbour susceptibility loci that were obscured in the unstratified data set.

The results were stratified into six categories: those families that were affected female-only pairs (132 families) , affected male-only pairs (60 families) , hips only (male or female or both) (194 families) knees only (male or female or both) (34 families) , affected female-only pairs who had undergone hip replacement but not knee replacement (female/hip pairs) (85 families) and affected male-only pairs who had undergone hip replacement but not knee replacement (male-hip pairs) (44 families) . We did not stratify for female/knee pairs or male/knee pairs as the number of families were too low (16 and 4 respectively) and thus any significant results may simply be the result of stochastic variation rather than linkage. We adjusted MLS values to correct for the six strata tested by deducting log7 = 0.8 from the original values²³.

Ten of the 20 autosomes have one or more peaks with uncorrected MLS > 1.0 for one or more of the six strata tested (Table 3) . After correcting for multiple testing, chromosome 6 had MLS > 1.0 for hip-only (corrected MLS = 2.1) and female-hip (corrected MLS = 1.0). The corrected multipoint plot of this chromosome is shown in Figure 1.

Over 50 cM of chromosome 6 has a corrected MLS > 1.5 in the hip-only strata. This region is encompassed by D6S257 and D6S262. The female-hip strata also has a corrected MLS > 1.0 on chromosome 6, with an MLS of 1.0 between D6S422 and D6S265. This region encompasses the HLA gene cluster on 6p21.3.

The suggestive linkage indicates that this region harbours a susceptibility locus for OA.

Of the two male strata tested (males-only and male-hip) , no chromosome had a corrected MLS > 1.00.

Methods

Affected sibling-pairs.

Families were recruited which contained at least two siblings two had undergone one or more total hip (THR) and/or total knee replacements (TKR) for idiopathic OA. Clinically these patients are at the severe end of the OA spectrum with advanced radiological changes. The idiopathic diagnosis was supported by clinical, radiological, operative and histological findings: patients who had undergone joint replacement surgery secondary to other factors, such as intra-articular fracture or rheumatoid arthritis, were excluded. Families were ascertained at three centres within the United Kingdom: The Nuffield Orthopaedic Centre in Oxford, the Royal Orthopaedic Hospital in Birmingham and Musgrave Park Hospital in Belfast. Idiopathic OA is typically a late-onset disease and parents of affected siblings are rarely available. Of the 481 families used in the study none contained a parent who was able to participate. We therefore collected siblings who had not undergone joint replacement to assist in the determination of identity-by-descent (IBD) allele transmittance . The 481 families were comprised of 1052 affected individuals plus an additional 302 unaffected siblings (Table 1). 59.3% of the affected individuals were female, 40.7% were male. For each individual, 25ml of venous blood was collected into EDTA tubes and DNA was extracted by conventional techniques . Markers and Genotyping.

Our initial screening panel of 272 microsatellite markers was essentially the panel used by Reed et al⁸ . The markers were amplified using conventional conditions with either the forward or the reverse primer in a PCR pair fluorescently labelled. The amplification products were electrophoresed through 6% acrylamide using an Applied Biosystems 377 Automated DNA Sequencer⁵. Alleles were sized using Applied Biosystems Genescan version 2.0.2. and Genotyper version 1.1 software.

Linkage and Linkage disequilibrium analysis.

Initially error checking procedures were employed for all families for each marker. After identification of straightforward mis-inheritances, more subtle transmission errors were detected using the PedCheck program¹⁷. The entire family data set was tested with Relative which is able to produce a probability calculation (based on 50 or more unlinked markers) that full sibships are in fact half sibs or even unrelated (due to unknown adoption of laboratory error) . All 481 families used in the study successfully progressed through these checks . In addition markers were checked for having excess homozygotes, based on their allele frequencies and eterozygosities . Markers shown to be unreliable were eliminated from the study. Finally the marker data were haplotyped for each chromosome using Simwalk2. This checks for areas on the chromosome where excessive recombination events may alert us to genotyping errors or mis-assignment of a marker position.

Non parametric linkage analysis was performed using the SIBPAIR module of the ANALYZE package¹⁸. This module is able to use data from siblings to determine identity-by-descent (IBD) allele transmittance . In this way it is appropriate to our study since we were unable to collect parents of our affected siblings. In the linkage analysis, siblings who had not undergone joint replacement were given a clinical status of unknown. The SIBPAIR module produces a singlepoint LOD score and its asymptotic P-value. Allele frequencies were calculated from the input data using either GAS or Downfreq (part of the ANALYZE package) . Subsequent multipoint analyses was performed using ASPEX¹⁹ (ftp: //lahmed. stanford.edu /pub/aspec) which calculates its own allele frequencies from the data set, using a maximum likelihood method, and employs marker information across the chromosome simultanenously¹⁹. ASPEX produces maximum LOD score (MLS) under an additive model. In addition it produces an exclusion map along the entire chromosome based on a fixed value for A? .

Stratification.

We stratified for sex, for joint replaced (hip or knee) and for sex combined with joint replaced.

For those families with more than two affected siblings and in which the siblings were not of the same sex, the affected sibling of opposite sex to a same pair was given an affected status of unknown in the linkage analysis. In this way we were stratifying for sex whilst not excluding siblings who could be used to assist in the determination of identity-by- descent (IBD) allele transmittance .

A hip-only pair were siblings who had each undergone joint replacement of the hip only (mono or bi-lateral) whilst a knee-only pair had undergone joint replacement of the knee only (mono or bi-lateral) . If an affected pair was composed of one sibling who had undergone joint replacement of one joint type only (hip or knee) whilst their affected sibling had undergone joint replacement of the hip and knee then that pair were excluded. For an affected trio, if a pair of the siblings had both undergone joint replacement of the same joint type only (hip or knee) whilst the third sibling had undergone both hip and knee replacement, then the concordant pair were used in the stratification study whilst the third sibling was given an unaffected status in the linkage analysis. Again, we were attempting to maximise our determination of IBD allele transmittance.

References . 1. Spector T.D., Cicuttini F., Baker J., Loughlin J. & Hart D. Genetic influences on osteoarthritis in women: a twin study. Br. Med. J. 312, 940-943 (1996) .

2. Chitnavis J., Sinsheimer J.S., Clipsham K., Loughlin J., Sykes B., Burge P.D. & Carr A.J. Genetic influences in end- stage osteoarthritis. Sibling risks of hip and knee replacement for idiopathic osteoarthritis, J. Bone. Joint. Surge 79, 660-664 (1997) .

3. Felson D.T., Couropmitree N.N. , Chaisson C.E., Hannan

M.T., Zhang Y., McAlindon T.E., LaValley M., Levy D. & Myers R.H. Evidence for a Mendelian gene in a segregation analysis of generalized radiographic osteoarthritis. Arth. Rheum. 41, 1064-1071 (1998) .

4. Hirsch R. , Lethbridge-Cejku M., Hanson R., Scott W.W., Reichle R., Plato C.C., Tobin J.D. & Hochberg M.C. Familial aggregation of osteoarthritis. Arth. Rheum. 41, 1227-1232

(1998) .

5. Creamer P. & Hochberg M.C. Osteoarthritis. Lancet 350, 503-508 (1997) .

6. Jones A. & Doherty M. ABC of Rheumatology. Osteoarthritis. Br. Med. J. 310, 457-460 (1995) .

7. Horton W.A. Evolution of the bone dysplasia family. Am. J. Med. Genet. 63, 4-6 (1996).

8. Reed P.W., Davies J.L., Copeman J.B., et al . Chromosome- specific microsatellite sets for fluorescence-based, semi- automated genome mapping. Nature Genet. 7, 390-395 (1994) . 9. Kaprio J., Kujala U.M. , Peltonen L. & Koskenvuo M. Genetic liability to osteoarthritis may be greater in women than men. Br. Med. J. 313, 232, (1996) .

10. Lindberg H. Prevalence of primary coxarthrosis in siblings of patients with primary coxarthrosis. Clin. Orthop. 203, 273-275 (1986).

11. Cooper C, McAlindon T., Snow S., et al . Mechanical and constitutional risk factors for symptomatic knee osteoarthritis: differences between medial tibio-femoral and patello-femoral disease. J. Rheumatol . 21, 307-313 (1994).

12. Wright G.D., Hughes A.E., Regan M. & Doherty M. Association of two loci on chromosome 2q with nodal osteoarthritis. Ann. Rheum. Dis. 55, 317-319 (1996) .

13. Leppavuori J.K., Kujala U.M. , Kaprio J., Nissila M., Helliovaara M., Kinnunen J., Koskenvuo M. & Peltonen L.

Genome screen for predisposing loci of distal interphalangeal joint osteoarthritis. Am. J. Hum. Genet. Suppl. 63, A1715 (1998).

14. Pattrick M. , Manhire A., Ward A.M. & Doherty M. HLA-A, B antigens and αl-antitrypsin phenotypes in nodal generalised osteoarthritis and erosive osteoarthritis. Ann. Rheum. Dis. 48, 470-475 (1989) .

15. Mundlos S. & Olsen B.R. Heritable diseases of the skeleton. Part II: Molecular insights into skeletal development-matrix components and their homeostasis. FASEB J. 11, 227-233 (1977) .

16. Clayton D. Transmit, http://www.mrc- bsu.cam.ac.uk/pub/methodology/genetics/transmit (1997) . 17. O'Connell J.R. & Weeks D.E. PedCheck: a program for identifying marker typing incompatibilities in linkage analysis. Am. J. Hum. Genet. Suppl . 60, A288 (1997).

18. Terwilliger J. Program SIBPAIR: sibpair analysis on nuclear families. 1996 ,- ftp : //linkage . cpmc . Columbia . edu .

19. Hinds D. & Risch N. The ASPEX package: affected sibpair mapping.

1996 ; ftp: //lahmed. stanford.edu/pub/aspex.

20. Ayad S, Boot-Handford RP, Humphries MJ, Kadler KE, Shuttleworth CA. The extracellular matrix facts book. Academic Press Limited, London, 1994.

21. Chapman K, Mustafa Z, Irven CM, Carr AJ, Clipsham K, Smith A, Chitnavis J, Sinsheimer JS, Bloomfield VA, McCartney M, Cox 0, Cardon LR, Sykes B, Loughlin J. Osteoarthritis susceptibility locus on chromosome llq detected by linkage. American Journal of Human Genetics, 1999, 65:167-174.

22. Fassler R, Schnegelsberg PNJ, Dausman J, et al . Mice lacking αl(IX) collagen develop noninflammatory degenerative joint disease. Proc Natl Acad Sci USA 1994; 91:5070-4.

23. Kidd KK, Ott J. Power and sample-size in linkage stidues. Cytogenet Cell Genet 1984; 37:5710-511.

24. MacGregor AJ, Antoniades L, Matson M, Spector TD . The genetic contribution to radiographic hip osteoarthritis: a population-based twin study. Arthritis and Rheumatism 1998; 41 supplement: 1954 (abstract).

25. Nakata K, Ono K, Miyazaki J-I, et al. Osteoarthritis associated with mild chondrodysplasia in transgenic mice expressing αl(IX) collagen chains with a central deletion. Proc Natl Acad Sci USA 1993' 90:2870-4.

26. Roby P, Eyre S, Worthington J, Ramesar R, Cilliers H, Beighton P, Grant M, Wallis G. Autosomal dominant (Beukes) premature degenerative osteoarthropathy of the hip joint maps to an 11-cM region on chromosome 4q35. American Journal of Human Genetics 1999; 64:904-908.

Tabl e 1 ( a )

Table 1 a) The families used in stages 1 and 2 together with th figure. Also listed are the pairs concordant for differei A list of the individuals in the study. a)

Families

Stage 1 Stage 2 Total

Families 297 184 481 sibling pair 265 150 415 sibling trio 23 27 50 sibling quattro 7 5 12 other* 2 2 4

other* 1 0 1

Knee only 34 20 54 pair 33 19 52 trio I 1 2

Female hip only 85 47 132 pair 77 46 123 trio 8 1 9

Female knee only r 16 5 21 pair 16 5 21

Male knee only 4 4 8 pair 4 4 8 Z k

Tabl e 1 ( b )

b )

Individuals

Stage 1 Stage 2 Total

Affected individuals 641 41 1 1052 Female 394 230 624 Male 247 181 428

Hip only 479 309 788 Knee only 121 77 198 Hip & Knee 41 25 66

Female hip only 287 172 459 Female knee only 77 42 1 19 Female hip & knee 30 16 46

Males/hip only 192 137 329 Males/knee only 44 35 79 Males/hip & knee 1 1 9 20

additional siblings^b 21 1 91 302 Female 107 49 156 Male 104 42 146

Other relative pairs such as cousins, uncles, aunts. ^bSince our families lack parents, siblings who had not undergone joint replacement surgery were collected to assist in the determination of parental genotypes. These siblings were given an "unknown" clinical status in the linkage analysis.

Table 2 LOD scores and p-values for all markers that had a p<0.05 in screen 1 , for these markers in screen 2 and for screens 1 and 2 combined (* = p<0.05).

STAGE 1 STAGE 2 COMBINED

Marker p-value LOD p-value LOD p-value LOD

D2S202 0.036* 0.70 0.07 0.49 0.009* 1.21

D3S1266 0.017* 0.96 0.5 0.00 0.082 0.42

D4S231 0.040* 0.67 0.5 0.00 0.33 0.04

D4S415 0.018* 0.95 0.33 0.04 0.025* 0.83

D6S260 0.050* 0.58 0.5 0.00 0.13 0.29

D6S273 0.016* 0.98 0.5 0.00 0.077 0.44 ID tr

D6S286 0.030* 0.77 0.5 0.00 0.081 0.42 to

D6S281 0.046* 0.61 0.45 0.00 0.062 0.52

D7S669 0.018* 0.94 0.25 0.10 0.021* 0.90

D7S530 0.006* 1.33 0.41 0.01 ^•0.013* 1.09

D11S907 0.050* 0.58 0.12 0.31 0.019* 0.92

D11S903 0.017* 0.97 0.07 0.49 0.004* 1.45

D11S901 0.0004* 2.45 0.10 0.37 0.0003* 2.51

D17S807 0.014* 1.03 0.5 0.00 0.15 0.24

D17S789 0.010* 1.16 0.5 0.00 0.071 0.47

DXSU068 0.020* 0.84 0.5 0.00 0.10 0.35

Table 3

Maximum multipoint LOD scores (MLS)

Cht \ MLS Corrected Cvtogenetic Between markers Strata

MLS Position

1 1.3 <1.0 Iq31-Iq44 D1S238&D1S103 female-only

1.5 <1.0 3p25-3p21 D3S 1263 &D3S 1289 female-hip

1.5 <1.0 3p25-3p21 D3S 1263 &D3S 1289 female-only

1.0 <1.0 3p21-3pl4 D3S 1289 &D3S 1285 hip-only

3.9 3.1 4ql2-4q21.2 D4S398 & D4S250 female-hip

1.7 <1.0 4ql2-4q21.2 D4S398 & D4S250 female-only

1.3 <1.0 5pl3.3-5pll.l D5S419 & D5S407 female-only

2.9 2.1 6p21.1-6q22.1 D6S1610&D6S31 hip-only

1.8 1.0 6p23-6p21.3 D6S422 & D6S265 female-hip

1.3 <1.0 6p21.3-6ql5 D6S291&D6S286 female-hip

1.2 <1.0 6ql3-6q22.1 D6S462 & D6S314 male-hip

1.1 <1.0 6p23-6p21.3 D6S422 & D6S265 female-only

1.5 <1.0 7qll.23-7q21.2 D7S502 & D7S524 hip-only

1.3 <1.0 7q22.1-7q32 D7S2502 & D7S684 hip-only

<1.0

1.3 <1.0 12pl2.1-12qll D12S358&D12S87 female-only

1.4 <1.0 12ql3.3- 12q23 D12S43&D12S338 female-hip

1.3 <1.0 12ql3.3- 12q23 D12S43&D12S338 female-only

<1.0

1.2 <1.0 14q24.3-14q32.2 D14S74&D14S51 male-hip

1.1 <1.0 14q24.3-14q32.2 D14S74&D14S51 male-only

<1.0

2.1 1.3 16pl3.1-16ql2.1 D16S407&D16S261 female-hip

1.7 <1.0 16pl3.1-16ql2.1 D16S407&D16S261 female-only

1.1 <1.0 16pl3.1-16ql2.1 D16S407&D16S261 hip-only

2.0 1.2 16q21-16q24.1 D16S265 & D16S516 female-only

<1.0

1.1 <1.0 18pll.32-18pll.21 D18S63&D18S53 female-hip

<1.0

<1.0 — -— .. ____-

Claims

1. A method for identifying individuals susceptible to osteoarthritis comprising obtaining a sample of genomic DNA and analysing alleles of polymorphic markers located in the chromosomal region from 6p21.1 to 6q22.

2. A method for identifying individuals susceptible to osteoarthritis of the hip comprising obtaining a sample of genomic DNA and analysing alleles of polymorphic markers located in the chromosomal region from 6p21.1 to 6q22.

3. A method for isolating genetic loci associated with susceptibility to OA comprising screening genomic libraries with genetic sequence derived from polymorphic markers located in the chromosomal region from 6p21.1 to 6q22 and identifying open reading frames in regions adjacent to said genetic sequence.

4. A method for isolating genetic loci associated with susceptibility to OA comprising screening genomic libraries with a polymorphic markers located in the chromosomal region from 6p21.1 to 6q22 and identifying open reading frames located within 500 Kb, preferably within 100 Kb, more preferably within 50 Kb and most preferably within 10 Kb of the marker.

5. A method for isolating genetic loci associated with susceptibility to OA comprising screening genomic libraries with genetic sequence derived from any one of the polymorphic marker D6S1610 ,D6S257 , D6S286, D6S462 and D6S262 and identifying open reading frames in adj acent regions .

A method for isolating genetic loci associated with susceptibility to OA comprising screening genomic libraries with any one of the polymorphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262 and identifying open reading frames located within 500 Kb, preferably within 100 Kb, more preferably within 50 Kb and most preferably within 10 Kb of the marker.

7. The use of alleles of polymorphic markers located in chromosomal region from 6p21.1 to 6q22 as markers for the identification of loci influencing susceptibility to OA.

8. The use of alleles of polymorphic markers located in chromosomal region from 6p21.1 to 6q22 as markers for the ancestral DNA variant conferring enhanced OA susceptibility.

9. The use of alleles of polymorphic markers located in chromosomal region from 6p21.1 to 6q22 as markers for mapping associated loci to identify genes associated with OA susceptibility.

10. The use of alleles of polymorphic markers located in chromosomal region from 6p21.1 to 6q22 in investigating loci capable of influencing susceptibility to OA.

11. A method for mapping loci which affect susceptibility to OA by comparing a genomic region containing a particular allele of a polymorphic marker located in chromosomal region from 6p21.1 to 6q22 with a genomic region containing a different allele of the same marker.

12. The use of alleles of any one of the polymorphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262 and D16S261 as a marker for the identification of loci influencing susceptibility to OA.

13. The use of alleles of any one of the polymorphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262 as a marker for the ancestral DNA variant conferring enhanced OA susceptibility.

14. The use of alleles of any one of the pol morphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262 as a marker for mapping associated loci to identify genes associated with OA susceptibility.

15. The use of alleles of any one of the polymorphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262 in investigating loci capable of influencing susceptibility to OA.

16. A method for mapping loci which affect susceptibility to OA by comparing a genomic region containing a particular allele of any one of the polymorphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262 with a genomic region containing a different allele of the same marker.

17. A method for determining individual susceptibility to osteoarthritis comprising obtaining sample genomic DNA from siblings, at least two of which have clinical symptoms of osteoarthritis. analysing a region of their genomic DNA comprising a polymorphic marker located in chromosomal region from 6p21.1 to 6q22, identifying allele sharing between the siblings as defined by a maximum log of the odds (LOD) score of greater than 1 and a p-value of less then 0.25, and determining individual susceptibility to osteoarthritis by reference to the allele sharing.

18. A method for determining individual susceptibility to osteoarthritis comprising obtaining sample genomic DNA from siblings, at least two of which have clinical symptoms of osteoarthritis . analysing a region of their genomic DNA comprising any one of the polymorphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262 , identifying allele sharing between the siblings as defined by a maximum log of the odds (LOD) score of greater than 1 and a p-value of less then 0.25, and determining individual susceptibility to osteoarthritis by reference to the allele sharing.

19. A method for determining individual susceptibility to osteoarthritis comprising obtaining sample genomic DNA from siblings, at least two of which have clinical symptoms of osteoarthritis. analysing a region of their genomic DNA comprising a polymorphic marker and located within 20cM, preferably within 15 cM, or more preferably within lOcM, even more preferably within 5cM and most preferably within 2cM, of any one of the polymorphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262, identifying allele sharing between the siblings as defined by a maximum log of the odds (LOD) score of greater than 1 and a p-value of less then 0.25, and determining individual susceptibility to osteoarthritis by reference to the allele sharing.

20. A method for determining individual susceptibility to osteoarthritis comprising obtaining sample genomic DNA from siblings, at least two of which have clinical symptoms of osteoarthritis. analysing a region of their genomic DNA comprising one or more of the polymorphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262 and additionally analysing one or more genomic regions comprising the polymorphic markers; D2S202, D2S325,

D2S117, D3S1266, D4S415, D4S398, D4S231, D4S250, D4S406 D6S260, D6S273, D6S286, D6S281, D7S669, D7S530, D11S937, D11S907, D11S903, D11S901, D17S807, D17S789 and DXS1068, identifying allele sharing between the siblings as defined by a maximum log of the odds (LOD) score of greater than 1 and a p-value of less then 0.25, and determining individual susceptibility to osteoarthritis by reference to the allele sharing.

21. A method for isolating genetic loci associated with susceptibility to OA comprising screening a genomic library with sequence derived from the region between polymorphic markers D6S1610 and D6S262 and identifying those isolated clones which additionally contain open reading frames .

22. A method for isolating genetic loci associated with susceptibility to OA comprising screening a genomic library with sequence derived from the region between polymorphic markers D6S257 and D6S262 and identifying those isolated clones which additionally contain open reading frames .

23. A method for isolating genetic loci associated with susceptibility to OA comprising screening a genomic library with sequence derived from the region between polymorphic markers D6S257 and D6S286 and identifying those isolated clones which additionally contain open reading frames .

24. A method for identifying loci conferring susceptibility to osteoarthritis comprising screening a genomic library with genetic sequence derived from one or more of the polymorphic markers D6S1610, D6S257, D6S286, D6S462 and D6S262 and identifying open reading frames located within 500 Kb, more preferably within 100 Kb, even more preferably within 50 Kb or most preferably within 10 Kb of any of these polymorphic markers.

25. A method for identifying individuals susceptible to osteoarthritis comprising obtaining a sample of genomic DNA and analysing the C0L9A1 gene for the presence of alleles associated with OA susceptibility.

6. A method for determining individual susceptibility to osteoarthritis comprising sequencing the COL9A1 gene, identifying the allelic form of the gene, correlating the allelic form of the sequenced gene to the known susceptibility to OA of individuals with that allelic form.