Page 111 - The Vasculitides, Volume 1: General Considerations and Systemic Vasculitis

Experimental Autoimmune Vasculitis 87

These in turn influence the long term clinical management. Moreover, genetic studies that
demonstrate the differing molecular associations of PR3- and MPO highlight the necessity of
considering the two entities separately.

Models of MPO-AAV

There have been many historical attempts at developing models of MPO-AAV, but most
have failed to mirror the clinical findings as they were associated with significant immune
complex deposition, a major deviation from the pauci-immune characteristics of clinical
disease. Spontaneous models of crescentic nephritis and vasculitis characterised by the
presence of circulating MPO-ANCA have been reported, for example the spontaneous
crescentic glomerulonephritis/Kinjoh (SCG/Kinjoh) mice. However, they are not pauci
immune and have an array of other autoantibodies besides MPO-ANCA [4]. Significant
advances were made in this field through use of knockout and genetically manipulated mice
strains, in particular rat strains inherently susceptible to glomerulonephritis. Those strategies
included the passive transfer of pathogenic antibody which proved useful in modelling the
acute vascular injury response, and active immunization with MPO that resulted in activation
of both humoral and cellular immune elements, consequently providing more insight into the
adaptive autoimmune response and potential responses to novel therapies.

Murine Models

Xiao and colleagues [5] developed an acute passive transfer model using purified
antibody or splenocytes taken from MPO-deficient mice immunized with purified murine
MPO. This took advantage of the allo-immune response to MPO in MPO-deficient mice,
generating antibodies and lymphocyte responses that were sufficiently potent to induce
disease. Anti-MPO antibodies injected into wild type or RAG-deficient recipient mice led to
development of necrotizing pauci-immune glomerulonephritis, and in some cases, pulmonary
capillaritis over 6 to 13 days. It was further noted that splenocytes transfer recipients
developed glomerular immune deposits, whereas antibody transfer recipients remained pauci-
immune. Virtually all subsequent work was conducted with the model that used a passive
transfer strategy even though the degree of induced disease was much milder.

The development of glomerulonephritis using cell transfer was dependent on B-cells
rather than T-cells, and required the presence of neutrophils [6]. This model was replicated at
other centres and further refined to produce augmented disease through the addition of
lipopolysaccharide (LPS) [7], mimicking the effects of infection, and exacerbating
autoimmunity via toll like receptor (TLR) activation and up-regulation of tumour necrosis
factor-?. Further investigation of the model using different mouse strains, demonstrated that
the genetic background of the mice had a profound effect on the severity of the model, with
129S6 mice developing worse crescentic disease than C57BL/6 mice in which the model was
first generated [8], and with augmented pulmonary-renal disease in the absence of the
inhibitory Fc?-II receptors [9]. That model demonstrated a critical role for alternative, but not
classical, complement components in disease induction [3], which was surprising given the

Complimentary Contributor Copy