The model of CIA, a T cell- and cytokine-dependent disease, in DBA/1 mice has led to increased understanding of RA and has facilitated the development of novel biologics, such as TNF-blocking therapies [16
]. However, the apparent resistance of strains normally used to carry modified genes has impeded our ability to rapidly ask basic questions about disease pathogenesis, as a 1- to 2-year backcross to DBA/1 mice is needed. Our aim was to comprehensively assess the susceptibility of B6 mice to CIA and compare the disease with the 'classic' model in DBA/1 mice.
Our studies showed that chicken, and not bovine, CII was capable of inducing disease in B6 mice, with an incidence of 50% to 75%, an incidence similar to that previously described [10
]. This is in contrast to DBA/1 mice, in which bovine, mouse, and chicken CII all induced disease, with an incidence of 80% to 100%. This may account for reports of resistance to CIA in B6 mice, in which bovine CII was used for immunisation [8
]. Other confounding factors could include the quality of collagen preparation, or substrains of B6 mice, and it is important to note that our study was carried out with B6 mice purchased from Harlan UK, although we have obtained similar results with B6 mice from Charles River UK Ltd. (Margate, Kent, UK).
The phenotype of arthritis was milder in B6 mice than in DBA/1 mice, with less swelling and a more gradual increase in clinical score. Histological assessment of the hind paws from arthritic DBA/1 and B6 mice revealed that, in early arthritis (up to 2 weeks after onset), there was a similar degree of inflammatory cell infiltration in the two strains. In contrast, in late arthritis (6 to 8 weeks after onset), inflammatory cell infiltration was reduced in DBA/1 mice compared with B6 mice, although it remains to be established which cell types are present in the joints of B6 CIA.
Assessment of lymph node responses showed that in the B6 mouse, both early and late after immunisation, proliferation and IFN-γ production in response to collagen occurred. Also of note, a strong response was observed in B6 mice to mouse collagen, suggesting the autoimmune nature of the model. It must be noted that bovine and murine collagen did not induce arthritis in B6 mice.
The reason why chicken, and not mouse or bovine, CII is arthritogenic in B6 mice is presumably due to recognition by B6 T cells of a peptide of chicken CII in the context of H-2b class II molecules. This suggests that differences in the amino acid sequence between chicken and mouse/bovine CII are required to break tolerance and induce arthritis. It is intriguing that the T-cell response was greater and more sustained in B6 mice compared with DBA/1 mice, but the reasons for this are unknown. The number and activity of CD4+CD25+ regulatory T cells were found to be similar in the two strains (G. Criado, M. Medghalchi, R.O. Williams, unpublished observations). Therefore, we cannot attribute sustained T-cell responses to any obvious defect in regulatory T cells in B6 mice.
It has been proposed that pepsin (used to purify collagen) plays an important role in breaking T-cell tolerance to collagen and that much of the T-cell response in some strains of mice and rat is directed against pepsin or pepsin-modified epitopes of collagen [14
]. By showing equivalent responses to CII prepared with and without pepsin using lathyritic collagen [14
], we have shown that the T-cell response is not dependent on pepsin in this model, in contrast to rat strains, in which T-cell responses have been shown to be directed mainly against contaminating pepsin [15
]. However, we cannot exclude the possibility that pepsin contributes to the breaking of tolerance during immunisation, and we were unable to obtain lathyritic chicken type II collagen in order to test this hypothesis. However, the mycobacterial component of CFA provides many factors that are able to break tolerance via activation of Toll-like receptors.
The therapeutic profile of CIA in the B6 mouse was similar to that of RA, with a therapeutic action of methotrexate at a dose comparable to human therapy. This is in contrast to CIA in DBA/1 mice, in which methotrexate had no effect. One of the anti-inflammatory mechanisms of methotrexate is thought to be due to increased adenosine production [17
]. Adenosine acts via G-protein-coupled receptors to increase cAMP levels, which is known to reduce inflammation [18
]. It was recently reported that DBA/1 mice, but not B6 mice, are genetically resistant to the effects of methotrexate, due to defective adenosine accumulation [19
]. This is of particular significance as methotrexate is now regarded as the 'gold standard' small-molecular-weight drug for RA and is frequently used in combination with biologics, such as anti-TNF therapy [20
]. There is, therefore, an increasing need to model the anti-arthritic effects of methotrexate in combination with other therapies in order to optimise treatment regimens and to identify possible interactions. Likewise, indomethacin did not slow the disease progression of CIA in B6 mice, as in RA, but significantly reduced the disease severity of CIA in DBA/1 mice [21