• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Sep 1, 1998; 95(18): 10746–10750.

New susceptibility locus for rheumatoid arthritis suggested by a genome-wide linkage study


Rheumatoid arthritis (RA), the most common autoimmune disease, is associated in families with other autoimmune diseases, including insulin-dependent diabetes mellitus (IDDM). Its genetic component has been suggested by familial aggregation (λs = 5), twin studies, and segregation analysis. HLA, which is the only susceptibility locus known, has been estimated to account for one-third of this component. The aim of this paper was to identify new RA loci. A genome scan was performed with 114 European Caucasian RA sib pairs from 97 nuclear families. Linkage was significant only for HLA (P < 2.5[center dot]10−5) and nominal for 19 markers in 14 other regions (P < 0.05). Four of the loci implicated in IDDM potentially overlap with these regions: the putative IDDM6, IDDM9, IDDM13, and DXS998 loci. The first two of these candidate regions, defined in the RA genome scan by the markers D18S68-D18S61-D18S469 (18q22–23) and D3S1267 (3q13), respectively, were studied in 194 additional RA sib pairs from 164 nuclear families. Support for linkage to chromosome 3 only was extended significantly (P = 0.002). The analysis of all 261 families provided a linkage evidence of P = 0.001 and suggested an interaction between this putative RA locus and HLA. This locus could account for 16% of the genetic component of RA. Candidate genes include those coding for CD80 and CD86, molecules involved in antigen-specific T cell recognition. In conclusion, this first genome scan in RA Caucasian families revealed 14 candidate regions, one of which was supported further by the study of a second set of families.

Rheumatoid arthritis (RA), a chronic disease that leads to progressive joint destruction (1), is associated with other autoimmune diseases in families, such as thyroid diseases and insulin-dependent diabetes mellitus (IDDM) (2, 3). RA preferentially affects women (1), with a sex ratio of 2 to 4. Its mean age of onset is between 45 and 50 years of age, and its prevalence may be as high as 1% in adults (1). RA is believed to be a multifactorial disease resulting from a T cell-driven autoimmune process aimed primarily at the joints; its pathophysiology is unknown (4). Its genetic component has been suggested by familial aggregation (RA prevalence ratio of sibs to the general population: λs = 5) (2, 5), twin studies (68), and segregation analysis (9). HLA, which is the only susceptibility locus known (10, 11), has been estimated to account for one-third of this component (λHLA=1.8) (12, 13). To identify new RA susceptibility loci, ECRAF (the European Consortium on Rheumatoid Arthritis Families) collected affected sib pair families for analysis by genome scanning, followed by further testing of suggested loci.


Ascertainment of Families.

Families were recruited through the two RA siblings and then extended to parents and eventually other siblings when one or two parents were unavailable, which is often the case in this disease of relatively late onset. Potential families either were identified by clinicians (all countries) or were self-reported after media announcements (France, Italy, Spain, and Belgium). At least two siblings in each family met the 1987 American College of Rheumatology criteria for RA (14), as reported by one of the rheumatologists participating in this study (The Netherlands and Belgium) or the clinician in charge of the patient (other countries) who completed a standardized questionnaire.


DNA was extracted from whole blood by standard methods in the various centers and used at Généthon for genotyping of microsatellite markers with the fluorescence-based technology on ABI373 automatic sequencers (15), using the programs genescan analysis 2.0.0 and genotyper 1.1. The list of markers (with a mean heterozygosity of 77%) is available on request. The software interpretation of each allelic profile was checked by two different individuals before verification of Mendelian segregation. Any discrepancy was resolved by S.F. or C.D., or the data were excluded for the analysis. Three families from the first group of 100 families were excluded for non-Mendelian segregation, as was one unaffected sib from the additional 164 families. Molecular HLA-DRB1 genotyping was performed with the Inolipa kit (Murex, Chatillon, France).

Analytical Methods.

Affected sib pair analysis was conducted for the bipoint linkage analysis (one marker locus and the disease locus) with J. Terwilliger’s analyze package, using the program sibpalna 1.1 for autosomes and sibpair 2.1 for the X chromosome. For the few families with multiple affected “sibships,” we used all-possible pairs as if they were independent. The use of all-possible pairs has little effect on the P values under H0 (null hypothesis: no linkage) when the majority of the sample is independent sib pairs (16). Allele frequencies were estimated from the data with ilink of the linkage package (17). Multipoint linkage analysis (several marker loci and the disease locus) was performed with the program genehunter 1.1 (ref. 18).

HLA partition was done according to the identity by state (IBS) of genotypes available for 3 HLA markers (TNFa, D6S276, and HLA-DRB1 typing), as an approximation of identity by descent (IBD), because most parents were unavailable for genotyping; HLA-concordant applies to the pairs with complete IBS, and HLA-discordant applies to the remaining pairs (only the first affected pair of each family was used in the analysis). Homogeneity between subsamples was tested by using a χ2 comparing the numbers of shared and unshared IBD alleles.

The parameter λ for a susceptibility locus M is defined as the ratio of the expected proportion of sib pairs sharing no alleles IBD (25%) to the observed proportion among affected sib pairs. For the markers with the best linkage evidence, this observed proportion was estimated with the program sibpalna. The relative contribution of locus M was computed assuming a multiplicative model for the interaction between M and other susceptibility loci (logλM/logλs) (19). From our data, the relative contribution of HLA and the chromosome 3 region (3.1) can be estimated to be 33% and 16%, respectively, using λs = 5 (ref. 1), λHLA = 1.7 (for TNFa), and λ3.1 = 1.3 (for D3S3576).


A genome scan was performed outside of the HLA region with 114 sib pairs from 97 nuclear families (Table (Table1)1) with a mean spacing of 12 cM, using 309 microsatellite markers. In the HLA region, 17 markers spanning a 43-cM interval (D6S443– D6S282) were used. Linkage to HLA was significant (P = 2.0 × 10−5 for the TNFa microsatellite marker). Evidence for linkage to HLA (P < 0.05) was observed in a region spanning 31 cM, illustrating the power of the 12-cM scan. Outside of the HLA region, using bipoint linkage analysis, 30 markers showed nominal evidence of linkage to RA (P < 0.05), defining 23 candidate regions (Table (Table22).a A second analysis of the results was performed with a multipoint analysis program (Fig. (Fig.1).1). This showed similar trends, but nominal evidence was obtained for only 19 of 30 markers (Table (Table2).2). No additional regions were suggested by the multipoint analysis. Altogether, 14 regions showed consistent nominal evidence of linkage. Four of the loci implicated in IDDM (2125) potentially overlap with these regions: the putative IDDM6, IDDM9, IDDM13, and DXS998 loci with the regions 18.2, 3.1, 2.2, and X.2 defined by the markers D18S68-D18S61-D18S469, D3S1267, D2S377-D2S2354, DXS1068, and DXS998, respectively (Table (Table2).2).

Table 1
Rheumatoid arthritis-affected sib-pair families
Table 2
Evidence for linkage (P < 0.05) from the RA genome scan performed with 309 markers outside of the HLA region
Figure 1
Multipoint analysis of the chromosomes 3 and 18 scan. The genetic map is plotted on the x axis, and the likelihood ratio score calculated by the genehunter program is plotted on the y axis. The associated linkage P value is reported for the most linked ...

The first two regions (18.2 and 3.1) on chromosome 18q22–23 and chromosome 3q13 (Fig. (Fig.1)1) were studied further. Additional markers in both regions were analyzed in the same families, which increased the evidence for linkage only for the 18.2 region (P = 10−4) (Table (Table3).3). Markers from both regions then were studied in 194 new sib pairs from 164 nuclear families (Table (Table1).1). A small excess of allele sharing was observed in the chromosome 18 region, which did not reach significance (P = 0.08) (Table (Table4).4). Evidence for RA linkage to the chromosome 3 region was found (P = 0.002), providing strong support for a new RA locus, although the overall evidence from all families was only suggestive [P = 0.001, Table Table4,4, the proposed threshold for significance (26) is 2.5 × 10−5]. Multipoint analysis was consistent with these results (Table (Table5).5).

Table 3
Evidence for linkage to chromosome 18 region 18.2 and chromosome 3 region 3.1 (Table (Table2)2) in genome scan families
Table 4
Study of a second set of families for chromosome 18 and 3 regions (18.2 and 3.1, Table Table2)2) and search for interaction with HLA
Table 5
Multipoint affected sib-pair analysis

To search for possible interactions with the HLA locus, an analysis of all families was then performed after partition for HLA (see Materials and Methods). The results revealed a higher evidence for linkage in the HLA-concordant RA sib pairs for the chromosome 3 region (P = 0.001) compared with the remaining RA pairs (P = 0.08) (Table (Table4).4). No consistent trend was observed in the chromosome 18 region (Table (Table44).

We believe that our data support RA linkage to chromosome 3, but not chromosome 18, placing more emphasis on consistency than on P value. We hypothesize that RA and IDDM could share a putative chromosome 3 autoimmunity factor. However, definitive proof must await further evidence for linkage or discovery of the underlying genetic factor. Interestingly, obvious candidate genes are located in this region. They code for the antigen-presenting cell molecules CD80 and CD86, which interact with T cell molecules CTLA-4 and CD28 in the costimulatory pathway that determines the fate of T cells (27). A model involving this pathway has been proposed recently for RA, IDDM, and other autoimmune diseases (28).

The initial evidence for the chromosome 18 locus was either spurious, or was a chance finding that revealed a minor RA locus. Indeed, the recent refinement of the IDDM6 locus (just centromeric to the 18.2 region) at the level of allelic association will permit testing the implication of this locus in RA (29).

From our data, the relative contribution of HLA and the chromosome 3 region (3.1) can be estimated to be 33% and 16%, respectively (see Materials and Methods). To search for the remaining factor(s), the other candidate regions (Table (Table2)2) deserve high priority. Confirmation of some of these regions would rely on further testing in additional families. While this paper was under revision, two of the regions reported here on chromosome 1 and chromosome X (regions 1.1 and X.2, Table Table2)2) were implicated in Japanese RA families (30).


We reported a genome scan in Caucasian RA families. It showed similarities with the associated autoimmune disease, IDDM (21, 22), with no other locus than HLA showing significant linkage, but 14 regions with nominal evidence. Additional study for one of these candidate regions on chromosome 3 provided further support for a new RA locus.


We greatly thank the families and their clinicians for their participation. Funding was provided by the European Union (Biomed2); the Association Française contre les Myopathies; the Association de Recherche sur la Polyarthrite; the Association Française des Polyarthritiques; the Société Française de Rhumatologie; the Assistance-Publique-Hôpitaux de Paris (PHRC94); and the Ministère de la Recherche (ACC-SV2). We thank M. Besnard, S. Boisson, S. Duverneuil, E. Gardais, S. Larget, and M. Sennecheau for technical assistance at Généthon, and S. Cure, P. Brooks, and the reviewers for improving the manuscript.


insulin-dependent diabetes mellitus
identity by descent
identity by state


aThe candidate regions 3.1, 1.1, and X.2 (Table (Table2)2) were referred to as RA2, RA3, and RA4 loci (retaining RA1 for the HLA locus) at the American Society of Human Genetics, the American College of Rheumatology, and the Société Française de Rhumatologie 1997 meetings. However, these names, even if provisional, could be confusing, as none of these loci has been definitively linked to RA susceptibility.


1. Harris E. In: in Textbook of Rheumatology. Kelley W, Harris E, Ruddy S, Sledge C, editors. Philadelphia: Saunders; 1989. pp. 943–974.
2. Thomas D, Young A, Gorsuch A, Bottazzo G, Cudworth A. Ann Rheum Dis. 1983;42:297–300. [PMC free article] [PubMed]
3. Buchanan W, Crooks J, Alexander D, Koutras D, Wayne E, Gray K. Lancet. 1961;i:245–248.
4. Feldmann M, Brennan F, Maini R. Cell. 1996;85:307–310. [PubMed]
5. Deljunco D, Luthra H, Annegers J, Worthington J, Kurland L. Am J Epidemiol. 1984;119:813–829. [PubMed]
6. Lawrence J. Ann Rheum Dis. 1970;29:357–379. [PMC free article] [PubMed]
7. Aho K, Koskenvuo M, Tuominen J, Kaprio J. Arthritis Rheum. 1986;13:899–902. [PubMed]
8. Silman A, MacGregor A, Thomson W, Holligan S, Carthy D, Farhan A, Ollier W. Brit J Rheumatol. 1993;32:903–907. [PubMed]
9. Lynn A, Kwoh C, Venglish C, Aston C, Chakravarti A. Am J Hum Genet. 1995;57:150–159. [PMC free article] [PubMed]
10. Stastny P. N Engl J Med. 1978;28:869–871. [PubMed]
11. Gregersen P, Silver J, Winchester R. Arthritis Rheum. 1987;30:1205–1213. [PubMed]
12. Wordsworth P, Bell J. Springer Semin Immunopathol. 1992;14:59–78. [PubMed]
13. Ollier W. In: in Genetics of Common Diseases. Day I, Humphries S, editors. Oxford: Bios; 1997. pp. 171–183.
14. Arnett F, Edworthy S, Bloch D, McShane D, Fries J, Cooper N, Healey L, Kaplan S, Liang M, Luthra H, et al. Arthritis Rheum. 1988;31:315–324. [PubMed]
15. Reed P, Davies J, Copeman J, Benett S, Palmer S, Pritchard L, Gough S, Kawaguchi Y, Cordell H, Balfour K, et al. Nat Genet. 1994;7:390–395. [PubMed]
16. Meunier F, Philippi A, Martinez M, Demenais F. Genet Epidemiol. 1997;14:1107–1111. [PubMed]
17. Lathrop G, Lalouel J, Julier C, Ott J. Proc Natl Acad Sci USA. 1985;81:3443–3446. [PMC free article] [PubMed]
18. Kruglyak L, Daly M, Reeve-Daly M, Lander E. Am J Hum Genet. 1996;58:1347–1363. [PMC free article] [PubMed]
19. Risch N. Am J Hum Genet. 1987;40:1–14. [PMC free article] [PubMed]
20. Dib C, Fauré S, Fizames C, Samson D, Drouot N, Vignal A, Milasseau P, Marc S, Hazan J, Seboun E, et al. Nature (London) 1996;380:152–154. [PubMed]
21. Davies J, Kawaguchi Y, Bennett S, Copeman J, Cordell H, Pritchard L, Reeds P, Gough S, Jenkins S, Palmer S, et al. Nature (London) 1994;371:130–135. [PubMed]
22. Hashimoto L, Habita C, Béressi JP, Delepine M, Besse C, Cambon-Thomsen A, Deschamps I, Rotter J, Djoulah S, James M, et al. Nature (London) 1994;371:161–163. [PubMed]
23. Todd J. Proc Natl Acad Sci USA. 1995;92:8560–8565. [PMC free article] [PubMed]
24. Todd J, Farrall M. Hum Mol Genet. 1996;5:1443–1448. [PubMed]
25. Cordell H, Kawaguchi Y, Todd J, Farrall M. Ann Hum Genet. 1995;59:435–449. [PubMed]
26. Lander E, Kruglyak L. Nat Genet. 1995;11:241–247. [PubMed]
27. Reiser H, Stadecker M. N Engl J Med. 1996;335:1369–1377. [PubMed]
28. Thomas R, Lipsky P. Immunol Today. 1996;17:559–564. [PubMed]
29. Merriman T, Twells R, Merriman M, Eaves I, Cox R, Cucca F, McKinney P, Shield J, Baum D, Bosi E, et al. Hum Mol Genet. 1997;6:1003–1010. [PubMed]
30. Shiozawa, S., Hayashi, S., Tsukamoto, Y., Yasuda, N., Goko, H., Kawasaki, H., Wada, T., Shimizu, K., Takasugi, K., Tanaka, Y., Shiozawa, K. & Imura, S. (1997) Arthritis Rheum. 40, Suppl., S329.

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • MedGen
    Related information in MedGen
  • OMIM
    OMIM record citing PubMed
  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...