Studies on immunological mechanisms of induced arthritis in the rat

Abstract

Experimental arthritis induced with adjuvants shares many features with the human autoimmune disorder rheumatoid arthritis (RA). The use of arthritis models is therefore a relevant approach in the search for pathogenic mechanisms in human disease.

This thesis has focused on investigation of the mechanisms underlying arthritis induced with adjuvants in rats. The experimental models we have utilized were oil-induced arthritis (OIA), squalene-induced arthritis (SIA) and a new model, denominated CpG-induced arthritis.

OIA and SIA are both induced with non-immunogenic adjuvants, OIA with incomplete Freund's adjuvant and SIA with the cholesterol precursor squalene. They are both T cell dependent diseases, affecting only the peripheral joints with subsequent erosions of bone and cartilage. The influence of MHC as well as non-MHC genes has been demonstrated in both these models.

In the first paper we evaluated the contribution of CD4+ and CD8+ T cells respectively, to development of disease. This was performed using adoptive transfer to irradiated recipients. To further characterise the pathogenic cells we blocked receptors on the transferred population, determined their TCR VP usage and analysed their cytokine expression. We demonstrated that CD4+, but not CD8+ T cells were pathogenic and could transfer OIA to irradiated recipients. These cells were polyclonal and Th1-biased. Blocking of cell surface molecules (CD4, ICAM1 or the IL-2 receptor) on the arthritis-transferring cell population before transfer did not markedly influence passive OIA development.

One important feature with experimental models is the possibility to characterise events before arthritis onset. In the second study we utilised this possibility investigating systemic and local effects after injection with IFA but before disease onset. In an attempt to characterise differences underlying susceptibility we used the susceptible DA strain as well as the MHCcongenic but resistant LEW.1AV1 and PVG.1AV1 strains. Of the selected phenotypes, an early systemic response to the adjuvant exposure as well as lymph node hyperplasia, both not regulated by MHC genes, could be demonstrated. In the susceptible strain, the acute-phase reactant alpha1-acid glycoprotein and mRNA expression of IL-1beta were demonstrated to be potential markers for disease.

Since it is intriguing that non-immunogenic adjuvants can lead to an organ-specific disease and both susceptible as well as resistant strains express a systemic response to the adjuvant injection, we explored the distribution of adjuvant. We demonstrated a high deposition of the 3 H- labelled squalene at the injection site even after 15 days (i.e. when animals are arthritic). Adjuvant was also deposited in draining inguinal lymph nodes but not in affected joints or in non-draining lymph nodes (axillary and brachial). All strains revealed the same distribution of adjuvant regardless of susceptibility. Although no adjuvant could be traced in the non-draining lymph nodes these cells were still pathogenic and could transfer disease to irradiated recipients.

DNA sequences containing CpG motifs are common in bacteria but rare in humans. This immunostimulatory DNA induces an innate immune response through interaction with Tolllike receptor (TLR)-9. Intradermal. administration of CpG-containing oligonucleotides together with IFA induced arthritis in LEW.1AV1 rats, a strain which is resistant to arthritis induced by IFA alone. In this arthritis model IL-1beta, AGP and T cells also appeared to be of importance for the pathogenesis.

In conclusion, this thesis has added to the knowledge of immune responses to diseaseinducing adjuvants as well as tissue distribution of injected adjuvants. We have also taken genetic aspects into consideration when investigating differences in distribution of and response to adjuvants between disease-susceptible and -resistant rat strains.