TYPES OF ANIMAL MODELS
Mouse strains that exhibit spontaneous development of syndromes that closely resemble human autoimmune rheumatic diseases have been identified by experienced investigators carefully observing mouse colonies. In addition, some efforts to induce mutation using chemical muta-gens have generated mice that express immune system defects or pathology that provide clues to human diseases with simi-lar problems. Advantages of spontaneous models of disease are that the phenotype is generally quite reproducible from ani-mal to animal (although, in some cases, the disease is only partially penetrant) and development of the disease does not require intervention by the investigator.
A disadvantage of spontaneous models is that the disease often takes months to develop, slowing the pace of investigation.
Some of the most useful animal models are those that are induced by the investigator by administration of a drug, an antigen, an adjuvant, an antibody, or through surgi-cal manipulation of the immune system. For example, demethylating drugs alter the structure of chromatin, resulting in increased accessibility of positive or nega-tive regulatory elements in gene promoters or enhancers and subsequent alteration of the gene expression profile of the animal, sometimes resulting in disease. Removal of the thymus at day 3 after birth has been demonstrated to promote development of several organ-targeted autoimmune diseases, most likely based on removal of an important regulatory T-cell population. Induction of immune system activation, as in the pristane model of lupus, can lead to altered patterns of cytokine production and autoimmunity. In some cases, autoimmune disease can be simply transferred from one animal to another by administration of an autoantibody or autoreactive T-cell population. Under these circumstances, the cells and molecules of the immune system that are required for disease expression can be identified and the investigation narrowed to one or at least a more narrow range of immune system components. The contribu-tion of background genetic factors to development of a disease phenotype can also be studied in inducible animal models.
Many investigators have used the tech-nology that generates animals that either overexpress or are deficient in a single gene to investigate the relevance of that gene product to development of a dis-ease phenotype. This approach has been particularly fruitful in the study of lupus, with transgenic and knockout mice dem-onstrating that any number of genetic modifications that alter self-antigen acces-sibility or threshold for immune system activation can lead to production of the classic lupus autoantibody, anti-DNA antibody, and deposition of those autoan-tibodies in the kidney, another character-istic of human lupus. Deficiencies in the complement system, members of which help to clear apoptotic debris and solu-bilize immune complexes; increased or decreased expression of cell surface mol-ecules, kinases, phosphatases, or adaptor molecules relevant to intracellular signal-ing and that modulate the threshold for lymphocyte activation; and altered expres-sion of death pathway molecules, such as Fas, can result in increased targeting of the immune response toward self compo-nents and manifestations of autoimmu-nity. An advantage of these sophisticated animal systems is that the modification is often restricted to one gene, or at least a small genomic region adjacent to the gene of interest, allowing the study of the impact of that particular molecular prod-uct and its relevant pathway on disease expression. However, there are inherent dangers in relying solely on transgenic and knockout models. That the gene of interest is absent during embryonic and fetal development means that other gene products can take over some of the func-tions of the modified gene, obfuscating the predicted phenotype. The role of a gene in development of a disease can also be over-estimated if background genomic factors in the neighborhood of the modified gene also contribute to disease. Finally, the fact that overexpression or deletion of a given gene results in a disease phenotype does not necessarily indicate that the same human gene is altered in the correspond-ing human disease. In fact, few of the genes that result in disease when modified in murine systems have been confirmed as disease genes in humans. Nonetheless, the transgenic and knockout approaches help to identify important pathways for further study and that may be eventually targeted therapeutically.
Like the transgenic and knockout approaches, congenic mice have proved very useful in studying the specific con-tributions of narrow genomic regions to aspects of disease pathology. The util-ity of congenic mouse strains, in which a segment of a chromosome is bred into a desired genetic background, is nicely illus-trated by the NZM 2410 strain. That mouse model has allowed identification of at least three chromosomal loci that confer distinct immune system alterations, which when present together result in severe lupus-like disease. As in the case of the transgenic and knockout models, gene sequences in the chromosomal segment of interest can interact with background genes to compli-cate interpretation of the role of gene prod-ucts in the congenic region.