Several murine strains have served as classic models of lupus-like disease, with each having many features that mimic immuno-logical and pathological aspects of human lupus. Although several spontaneous mouse models of lupus have striking simi-larities to human SLE, none of the animal models reproduces human SLE perfectly. What makes these animal models particu-larly interesting to study is that the dif-ferences may reflect different forms of the human disease.
The F1 hybrid is the earliest recognized spontaneous model of lupus. The mouse was developed from a cross between NZB mice, which develop a spontaneous auto-immune disease resembling hemolytic anemia, and NZW mice, which are phe-notypically normal. The B/W F1 strain is one of the best-studied models of human SLE and is considered by many to be the murine model most closely resembling human SLE. The disease is characterized by autoantibodies, including antinuclear antibodies (ANAs), IgG autoantibodies to double-stranded DNA (dsDNA), as well as lupus erythematosus (LE) cells, repre-senting phagocytic cells that have engulfed antibody-opsonized cell debris. Death results from a severe glomerulonephri-tis (GN) due to the deposition of immune complexes, including those containing anti-DNA antibodies. The disease is more severe and occurs earlier in female mice than in their male counterparts.
The B/W F1 is a good representative of human SLE because female B/W F1 mice manifest similar autoantibody profiles, and the disease develops in the absence of any known disease-accelerating gene, unlike other mouse models. Production of anti-dsDNA antibodies in this mouse, as in human SLE, is thought to be T-cell depen-dent as suggested by the IgG isotype, by somatic mutations in the antibody genes, and by experiments in which depletion of T cells suppresses antibody production and ameliorates disease. As is true for humans with lupus, the correlation between nephri-tis and anti-DNA antibody levels is not universal. In addition to anti-dsDNA anti-bodies, the mice develop other autoanti-bodies found in human disease, including those that bind single-stranded (ssDNA), transfer RNA (tRNA), polynucleotides,
histones, and nucleic acid–protein com-plexes. Anti-erythrocyte antibodies occur in 35–57 percent of B/W F1 females, but the mice rarely develop hemolytic anemia. Consistent with the important pathogenic role for IFN-α in human lupus, disease and death are accelerated in B/W F1 mice who receive IFN-α.
MRL/MpJ-Faslpr (MRL/lpr) is another spontaneous model of SLE. The MRL/lpr mouse has a genetic mutation, termed lpr (lymphoproliferation), which leads to mas-sive lymphoid organ enlargement and severe early-life lupus-like disease. The congenic MRL/+/+ strain lacks the muta-tion and develops a mild lupus-like dis-ease. MRL/lpr mice exhibit many clinical manifestations found in human SLE. Auto-antibodies produced by these mice are similar in spectrum to those seen in human lupus, including anti-dsDNA and anti-Sm antibodies. The mice also make rheuma-toid factors.MRL/lpr mice develop pro-liferative GN at an early age (4–5 months), and renal failure is the presumed primary cause of death in these mice.
Although the exact mechanisms involved in the pathogenesis of lupus nephritis are as yet unclear, there is gen-eral agreement that disease is mediated by glomerular deposition of autoantibodies as immune complexes formed in situ or by direct binding of the antibodies to an intrin-sic renal antigen or a self-antigen deposited in the kidney. This deposited immunoglob-ulin then induces renal injury primarily through complement activation, leading to recruitment and activation of inflamma-tory mediators. These mechanisms appear to exist in both human and murine lupus. However, a recent study in MRL/lpr mice suggested that the production of autoanti-
bodies was not required for renal disease. Mice that expressed a mutant transgene that did not permit secretion of circulat-ing immunoglobulin still developed renal disease, suggesting that immunoglobulin-expressing B cells might also participate in the disease process either as antigen-presenting cells or as part of the local inflammatory process.
In contrast to B/W F1 mice, both male and female MRL/lpr mice develop high serum levels of immunoglobulins, ANAs, and immune complexes, as well as dis-ease. Other antibodies in their repertoire include IgG2a anti-chromatin, anti-RBCs, anti-thyroglobulin, antilymphocyte, anti-ribosomal P, and anti-RNA polymerase I.
Polyarthritis occurs in MRL/lpr mice in some but not all colonies. The preva-lence varies between 15 and 20 percent. The arthritis is a destructive arthropathy, characterized by synovial cell proliferation with early subchondral bone destruction and marginal erosions. This is unlike the nonerosive arthritis that occurs in some patients with SLE.
Unlike human SLE, MRL/lpr mice are characterized by massive lymphadenopa-thy and splenomegaly, which is due to a defect in the fas gene, a key mediator of apoptosis. The fas defect alone is sufficient to induce autoantibody production but is not sufficient to induce renal disease. This is demonstrated in experiments in which the lpr gene is bred onto a normal back-ground, resulting in congenic lpr mice that produce autoantibodies but do not develop renal disease. Thus, genes in the MRL back-ground, independent of fas, are necessary for disease development, including renal disease, in MRL/lpr mice. Fas defects in humans are generally not seen in patients with SLE, although a population of patients with genetic defects in the fasgene develop a disease known as autoimmune lympho-proliferation syndrome, characterized by massive lymphadenopathy along with some autoimmune features.
The BXSB mouse is a recombinant inbred strain that spontaneously devel-ops an autoimmune syndrome similar to human SLE. This mouse model is charac-terized by the production of autoantibod-ies, hypergammaglobulinemia with class switching to IgG3 and IgG2b, hypocomple-mentemia, splenomegaly, and GN. BXSB mice develop a wide range of autoantibod-ies to nuclear components, typical of SLE, including ANAs, anti-dsDNA, anti-ssDNA, and antichromatin antibodies, with accom-panying splenomegaly and lymphadenop-athy. In addition, a small proportion make anti-erythrocyte antibodies.
The unique features of BXSB mice are that their disease is much worse in the male than the female, and the disease-accelerating gene responsible for that dif-ference (called Yaa for Y chromosome auto-immunity accelerator) is located on the Y chromosome and manifests in a male mor-tality of 50 percent by the age of six months. These mice have recently been documented to have a duplicated segment of the X chro-mosome that has been translocated to the Y chromosome. That segment includes TLR7, encoding a toll-like receptor that responds to ssRNA. The female BXSB mouse gets late-life lupus with death occurring at four-teen months. This suggests that additional genes contribute to disease in female mice.
By three months of age the mice have elevated levels of circulating immune complexes and hypocomplementemia. They are the only lupus mouse strain that has serum levels of C4 that diminish as clinical disease appears. Death is caused by immune complex GN. Histologically, the disease is more exudative than in other mouse models, with neutrophils invading glomeruli along with IgG and C3 deposi-tion, proliferative changes in mesangia and endothelial cells, and basement membrane thickening. The progression from nephritis to death is rapid. Of the most widely stud-ied SLE mouse models, the BXSB has the most fulminant disease.
The NZM2410 mouse is another spon-taneous model of SLE and is a congenic recombinant inbred strain (termed NZM for New Zealand Mixed) produced by inbreeding (NZB × NZW)F1 × NZW back-cross progeny. This mouse strain displays systemic autoimmunity with a highly penetrant, early-onset acute GN. The mice develop GN with a penetrance of about 85 percent and a 50 percent mortality at about six months of age. This represents an earlier onset of disease than that of B/W F1 mice.
The NZM2410 mouse strain differs sig-nificantly from the classical B/W F1 model and human SLE, where there is a strong female preponderance, in that both males and females develop the disease equally. It is possible either that the genomic regions responsible for the strong gender dimor-phism in B/W F1 were not included in NZM2410 or that NZM2410 contains a collection of homozygous susceptibil-ity alleles that are so severe that they can override the effects of sex hormones on the immune system.
Linkage analysis of susceptibility to GN and ANA production in the NZM2410 strain by Morel and Wakeland (2000) have identified three prominent loci termed Sle1, 2, and 3 that are strongly associated with lupus susceptibility. Subsequent congenic strains made by moving each locus onto the lupus-resistant C57Bl/6J (B6) background have determined (1) that Sle1 mediates a spontaneous loss of immu-nological tolerance to nuclear antigens; (2) that Sle2 lowers the activation threshold of the B-cell compartment and mediates poly-clonal/polyreactive antibody production; and (3) that Sle3mediates a dysregulation of the T-cell compartment that potentiates polyclonal IgG antibody production and decreases activation-induced cell death in CD4 T cells. Sle1 and Sle3 in combina-tion revealed that these two susceptibility genes are sufficient to mediate the develop-ment of severe humoral autoimmunity and fatal lupus nephritis on the B6 background. B6 bicongenic mice with these two alleles spontaneously develop high titers of auto-antibody directed against a broad spec-trum of nuclear chromatin autoantigens and die of kidney failure (penetrance >55 percent) due to autoimmune GN within twelve months of age.
The lethal phenotype produced by the combination of Sle1 and Sle3 on the B6 background was somewhat surprising, in that both of these genes are derived from the relatively unaffected NZW strain. This suggested that their severe autoimmune phenotypes must be suppressed in some manner by the NZW genome. Genetic anal-ysis of this epistatic suppression in NZW identified the presence and locations of four SLE suppressor loci (designatedSles1 through Sles4) that account for the absence of fatal disease in NZW.
Murine chromosome 1 contains the Sle1 locus as well as other loci of lupus susceptibility. This region in humans, spe-cifically 1q41-q42, also shows evidence of linkage with SLE, including the production of antichromatin antibodies. Thus, impor-tant susceptibility genes for autoimmunity appear to be conserved between mice and humans, supporting approaches to look for candidate human SLE-associated genes based on mouse studies.