Ethics of Research involving animals
Disease models in the mouse
7.9 Gene dysfunction is at the root of all genetically determined disease processes. Not all gene dysfunctions are heritable as gene expression is also influenced by injury, infection, ageing, cancer, neural degeneration and neural regeneration. By asking how often mouse mutants reproduce the effect of mutations in the corresponding human gene, it is possible to assess the utility and relevance of disease models. We illustrate this below with several examples (see also Table 7.1), which also show that the implications for the welfare of animals involved in such research are wide ranging.
i) Diabetes: Mutations in the glucokinase gene in humans lead to a form of type II diabetes4 that manifests itself in the young, called maturity-onset diabetes of the young (MODY). Mutations in the glucokinase gene in the mouse also develop a type II diabetes, very similar to that seen in human MODY patients.5 These mutants provide a useful model of MODY and enable scientists to investigate the relationship between mutations in the glucokinase gene and the pathogenesis and severity of the disease. Some of the mouse strains carrying mutations in the glucokinase gene have normal viability and fecundity and there do not appear to be detrimental effects on welfare. Other mutations, however, lead to more severe effects and are lethal during embryonic development.
ii) Deafness: The shaker1 mouse mutant displays a profound hearing loss and was one of the first mouse mutants investigated as a model of human genetic deafness at a time when little was known about the disorder. Researchers identified the mouse gene underlying the shaker1 mutant and then located the corresponding gene in the human genome. It was found that the shaker1 locus was encoded in mice by a gene of the type called myosin VII.6 It was subsequently demonstrated that mutations in the myosin VIIA gene in humans lead to hearing loss. Some of the mutations in this gene in humans can also lead to a syndrome where there is both hearing loss and blindness at around seven or eight years of age, due to the condition retinitis pigmentosa. Yet none of the myosin VIIA mutations isolated in the mouse cause blindness, even in very old mice. This may be a reflection of the short lifespan of the mouse which prevents the retina from receiving sufficient exposure to light to elicit pathological changes. Nevertheless, they do, as the name suggests, show hyperactivity, head-tossing and circling activity in addition to hearing loss.7
iii) Psychiatric disorders: It is probable that the equivalent conditions of many human psychiatric disorders are not exhibited in mice because of differences in the brain structures between the two species. It is also the case that many of the human patients who suffer from these disorders do not inherit them through simple genetic determinants, and that environmental factors play an important role. Scientists are exploring the role of the genes involved in certain inherited psychiatric disorders by examining their function in the mouse, and their influence on other genes and neurotransmitter systems at the level of neurones and the brain. Understanding how these genes function is important for the development of new therapies, although the modification of relevant genes in mice may not necessarily create the neuropsychiatric effects that are exhibited in humans.8 Mutant mice have also been screened for subtle behavioural changes to help identify genes that may be implicated in complex behavioural disorders in humans, such as anxiety or schizophrenia.9 Mice carrying mutations that affect behaviour rarely, if ever, manifest serious welfare problems, although there may be loss of complex subtle behaviours that may be revealed only in the wild or in response to complex stimuli that are not usually available to mice in the laboratory.
iv) Neurodegenerative disorders: Few neurodegenerative disorders, such as Parkinson’s disease and Alzheimer’s disease, are linked to single gene mutations. In Parkinson’s disease, three important mutations in genes responsible for different cellular functions (alpha-synuclein, parkin and a ubiquitin hydrolase) have already been identified. Three different genes with mutations implicated in Alzheimer’s disease (beta-amyloid, presenilin and tau) have also been described. Reproducing the human form of these mutated genes in mice produces comparable pathologies to those in humans. Although there is not yet a model which contains all of the relevant features that characterise the pathology of Alzheimer’s disease, the models available are nevertheless of great interest to researchers.10 A variety of approaches, including histopathological, imaging, electrophysiological and molecular genetic techniques have been particularly helpful for mapping the progression of neurodegenerative disorders in mouse models as well as determining the effects of several of the mutations. With regard to welfare implications, mouse models of neurodegenerative disease may show a variety of neurological impairments including, for example, tremors and ataxia (loss of full control of bodily movements). These symptoms often have significant effects on fecundity and viability and require careful monitoring. The diseases may also affect a mouse’s ability to interact with other animals, and to carry out behaviours such as play, running and climbing.
v) Lesch–Nyhan disease: Mutations in the Hprt gene, which encodes an enzyme involved in metabolism (hypoxanthine-guanine phosphoribosyltransferase), lead to a rare but very severe neurological syndrome in humans known as Lesch–Nyhan disease, the most characteristic feature of which is self-destructive biting. One of the earliest targeted mutations developed in the mouse, applying the reverse genetic approach (see paragraphs 5.19–5.22), resulted in the disruption of the Hprt gene. However, Hprt mouse mutants show none of the phenotype characteristics of Lesch–Nyhan syndrome. Researchers found that in the mouse an alternative enzyme pathway ameliorated the effect of the Hprt mutation, and obvious adverse effects on animal welfare from the generation and study of the mutant model have not been detected.
vi) Cancer: Prior to the sequencing of the mouse genome, investigating spontaneous mutations in genes involved in cancer required approximately 1,000 mice for crossbreeding in order to map a gene to a specific chromosomal region. This region wouldusually contain several genes, all of which needed to be sequenced to determine which one contained the mutation. As a result, it would have taken 15 years to identify ten possible genes that were involved in cancer, whereas this step can now be achieved in months. Moreover, comparisons between the mouse and human genomes help researchers to find related human genes encoding proteins that could be candidates for the development of new medicines. The recent development of a library of some 60,770 full-length cDNAs11 provides researchers with a functional copy of every mouse gene that can be readily genetically modified.12 This library is especially useful for studying human cancers or the role of other human genes involved, where the identity and location of the mouse homologue is unknown. With regard to animal welfare, mouse models of cancer usually demonstrate an increased incidence of tumours and an increased morbidity that will require careful monitoring.
7.10 In assessing the usefulness, relevance and validity of the large amount of data that are already available from studies of GM mice, advocates note that it is important to consider a number of features that characterise the investigation of mouse models, and which apply more generally to the analysis of any genetic animal model of human disease:
- First, when investigating and understanding the mechanistic basis of disease, as with all comparative analyses, the differences may be as instructive as the similarities. This is a feature that pervades not only comparative genetics but also comparative anatomy, physiology and pathology.
- Secondly, all or some of the relevant features of the phenotype arising from any mutation may not be detected by the methods commonly used. Some mutations do not result in any observable consequence. This may be due to: (i) the difficulty of detecting very subtle phenotypes; (ii) the effects of ‘genetic background’ that may modify the phenotypic outcome (see below); and (iii) the redundancy of pathways involved in biological systems.13 The lack of a phenotype may provide relevant information about the genetic pathways involved in any disease process but negative results often go unreported in the scientific literature.
- Thirdly, the disease phenotype resulting from a mutation may be modulated by the person’s genetic makeup. For example, while all siblings in a family might carry a mutation, they may vary in the way in which other genes in their genome affect the manifestation of the disease.14 As we have said, it is similarly true that the effect of mutations in mice can be very significantly altered by their genetic background. Analysis of the mouse genome allows researchers to better understand these interactions and to identify other genes that modify the effects of a particular mutant gene, further elaborating the understanding of the genetic mechanisms of disease.
- Fourthly, scientists do not expect a mutant model to replicate the entire complexity of the process of human disease. This is particularly true in the development and analysis of neurological and neurobehavioural disease models (see paragraph 7.9 (iii)). Rather, the aim is to identify genes that are involved in specific facets of complex neurobehavioural processes, which are called endophenotypes. Study of the separate components in the model system can help to improve understanding of the complexity of the phenotype.
- Finally, the outcomes for animal welfare are very variable, ranging from no immediately noticeable effects to significant effects on welfare and morbidity. They are also very unpredictable (see paragraph 4.57).
In conclusion, mouse models require careful analysis in order to assess their relevance and effects (see Table 7.1). While some animal protection groups remain sceptical about their overall usefulness,15 scientists working in the field maintain that, provided the points above are appropriately considered, their use produces significant information concerning the function of genes in mammalian disease processes and human genetic disease.
Table 7.1: A summary of the contribution
4 Type II diabetes is a late-onset disease that is not necessarily life-threatening and which does not always require control with
insulin administration.
5 Toye AA, Moir L, Hugill A et al. (2004) A New Mouse Model of Type 2 Diabetes, Produced by N-Ethyl-Nitrosourea Mutagenesis, Is the Result of a Missense Mutation in the Glucokinase Gene Diabetes 53: 1577–83.
6 Gibson F, Walsh J, Mburu P et al. (1995) A type VII myosin encoded by the mouse deafness gene shaker-1 Nature 374: 62–4.
7 Gibson F, Walsh J, Mburu P et al. (1995) A type VII myosin encoded by the mouse deafness gene shaker-1 Nature 374: 62–4.
8 Seong E, Seasholtz AF and Burmeister M (2002) Mouse models for psychiatric disorders Trends Genet 18: 643–50.
9 Ohl F and Keck ME (2003) Behavioural screening in mutagenised mice: In search for novel animal models of psychiatric
disorders Eur J Pharmacol 480: 219–28.
10 See, for example, Lee VM, Kenyon TK and Trojanowski JQ (2005) Transgenic animal models of tauopathies Biochim Biophys
Acta 1739: 251–9.
11 Complementary DNA: DNA produced from RNA sequences, which means that it contains only the sequences that code for
proteins.
12 The number of mouse cDNAs identified greatly exceeds the number of genes as some do not in fact code for proteins. See Suzuki M and Hayashizaki Y (2004) Mouse-centric comparative transcriptomics of protein coding and non-coding RNAs
Bioessays 26: 833–43.
13 (iii) ‘Redundancy’ refers to the fact that biological systems do not always fail due to the lack of a particular enzyme, for
example, as another pathway may compensate (see paragraph 7.9 (v)).
14 There may also be environmental effects such as air pollution or exposure to certain chemicals in the workplace which may
influence the expression of the disease phenotype.
15 British Union for the Abolition of Vivisection (2002) Designer Mice (London: BUAV).