Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement

The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.

RA Susceptibility Is Associated with An Epistatic Interaction Between Shared Epitope Alleles and MMEL1-TNFRSF14

El-Gabalawy1,  Hani S., Daha2,  Nina A., Robinson1,  David B., Oen1,  Kiem G., Smolik1,  Irene, Hart1,  Donna, Willemze2,  Annemiek

University of Manitoba, Winnipeg, MB
Leiden University Medical Center, Leiden, Netherlands
University of Toronto, Toronto


North American Native (NAN) populations have a high prevalence rates for RA. We have studied a Cree/Ojibway population in Central Canada for disease risk, and our previously published data indicate an early age of onset, high rates of multi-case families, high prevalence and titers of anti-citrullinated protein antibodies (ACPA) and rheumatoid factor (RF), and prevalent shared epitope (SE) alleles in the background control population. We therefore asked the question of whether additional genetic factors may interact with SE alleles in predisposing this population to the development of RA


RA patients (n=317), their first degree relatives (FDR, n=295), and unrelated controls (C, n=300) were tested for 21 single nucleotide polymorphisms (SNP) that have been shown to be associated with RA in whole genome scans, including PTPN22, TRAF1-C5, CTLA4, PADI4, STAT4, FCRL3, CCL21, MMEL1-TNFRSF14, CDK6, PRKCQ, KIF5A-PIP4K2C, IL2RB, TNFAIP3, IL10 –1082G/A, REL. HLA-DRB1 alleles were tested by sequencing. Significant SNP associations detected in the NAN population were tested for replication in a Dutch RA population (n= 880) and controls (n= 546)


The prevalence of SE alleles in the NAN population was RA = 85%, FDR = 79%, and C = 66%. The most common SE alleles were *0404 (38%) and *1402 (28%). With the exception of rs3890745 (MMEL1-TNFRSF14), there were no significant associations with any of the RA SNP tested, including the 1858T allele of PTPN22, which was present at a frequency of 9% in all groups. The data indicate that homozygosity for the major allele of MMEL1-TNFRSF14 was significantly higher in RA compared to both the C and FDR groups (OR 1.31, CI 1.05 –1.63, p=0.02). Further analysis indicated that this association was present only in SE+ individuals (OR 1.46 CI 1.15 –1.89, p=0.004), with no association detectable in SE- individuals. Binary logistic regression indicated that SE and MMEL1-TNFRSF14 were both independently associated with an RA diagnosis. Using a recessive model for the major allele of MMEL1-TNFRSF14 (AA vs AG or GG), there was clear evidence of an interaction between MMEL1-TNFRSF14 and SE (OR 1.84 p<0.0001 in SE+, and OR 0.75, p=NS in SE- individuals). Analysis in the Dutch population using the rs6684865 SNP (MMEL1), which is in complete linkage disequilibrium with rs3890745, demonstrated a similar trend in RA association (OR 1.16 CI 0.89–1.52), but there was no statistical evidence for interaction with SE.


The MMEL1-TNFRSF14 SNP has recently been shown to be associated with RA in a large genome wide analysis of primarily Caucasian populations. Here we demonstrate a novel interaction between SE and MMEL1-TNFRSF14 that is highly associated with RA susceptibility in the NAN population. This epistatic interaction is particularly important since SE alleles are prevalent in this population yet most individuals remain disease free.

To cite this abstract, please use the following information:
El-Gabalawy, Hani S., Daha, Nina A., Robinson, David B., Oen, Kiem G., Smolik, Irene, Hart, Donna, et al; RA Susceptibility Is Associated with An Epistatic Interaction Between Shared Epitope Alleles and MMEL1-TNFRSF14 [abstract]. Arthritis Rheum 2009;60 Suppl 10 :577
DOI: 10.1002/art.25657

Abstract Supplement

Meeting Menu