MHC and non-MHC genes in the regulation of susceptibility to collagen-induced arthritis among DA, BB, and BN rats

Abstract

OLLAGEN-INDUCED arthritis (CIA) is an experimentally induced model of autoimmune peripheral joint inflammation that is a widely used disease model for rheumatoid arthritis (RA). Like RA, CIA has been found to be a polygenic disease. Susceptibility to CIA is strongly linked to certain MHC class II alleles. However, the severity of arthritis can be viewed as a complex trait resulting from the combined effects of several different individual genesboth MHC and non-MHC encoded. To identify CIA-regulatory genes (and thus candidate genes for RA), our laboratories are carrying out quantitative trait linkage (QTL) analysis of large numbers of F2 progeny derived from matings of CIA-susceptible DA and BB rats with three CIA-resistant strains—BN, F344, ACI. Data on clinical trait severity plus genome wide scans (10 cM) by polymerase chain reaction (PCR) for polymorphic DNA microsatellite markers are being evaluated. The most informative quantitative traits are arthritis severity, weight loss, and the levels, type, and specificity of immune response to porcine and rat type II collagen. Earlier studies with (DA 3 ACI)F2 and (DA 3 F344)F2 hybrids showed a strongly dominant DA-MHC QTL (Cia1) and six nonMHC QTLs (Cia 2, 3, 4, 5, 6, 7) in these two strain combinations. 1‐3 Currently, CIA susceptibility in (DA 3 BN)F2 and (BB 3 BN)F2 rats is being studied. Severe CIA (score 50 to 70) develops in DA and BB rats (100%) after immunization with porcine type II collagen (PII‐CIA). In contrast, BN and DA.1N(BN) rats do not develop CIA (score 0). The test-cross data reveal very different inheritance patterns of PII-CIA susceptibility among the three hybrid populations: 26/27 (96%) of (BB 3 BN)F1 rats and 100% of (DA 3 F344)F1 rats vs only 12% (7/60) of (DA 3 BN)F1 rats develop PII-CIA. Likewise, 67% of 595 (BB 3 BN)F2 and 80% of over 500 (DA 3 F344)F2 rats but only 31% of 1007 (DA 3 BN)F2 rats developed PII-CIA. Clinically affected rats in the (BB 3 BN)F2 and (DA 3 BN)F2 populations show broad ranges of arthritis severity scores (0 to 70) representing the two extremes of parental phenotypes. Importantly, a partial genotyping of the (DA 3 BN)F2 cohort has revealed a very strong influence of non-MHC genes on CIA in this strain combination: approximately 45% of the (DA 3 BN)F2 rats that were DA/DA homozygous for the MHC marker tumor necrosis factor alpha (TNF-a) were CIA resistant (compared to 100% of the parental DA strain); furthermore, approximately 30% of the (DA 3 BN)F2 rats that were DA/BN heterozygous at the MHC were CIA susceptibile [compared to 0% in the (DA 3 BN)F1 rats]. The composit data from all three studies show that non-MHC genes can profoundly influence autoimmune disease expression. More importantly, different MHC alleles vary in their response to the governing influence exerted by non-MHC genes (s) originating in the same strain. Thus, the complex interplay of several MHC and non-MHC genes determine the clinical response to an environmentally imposed arthritogenic and autoimmune challenge. Delineation of the number, strength, and interaction of the multiple alleles that regulate CIA in rats should in the future provide not only specific, individual candidate genes for evaluation in human RA, but also may provide a basis for unraveling the known ethnic variations in the frequency, strength, and identity of susceptibility genes for RA and other polygenic autoimmune diseases.