Choice of species is largely determined by the nature of the experiment, available facilities and the suitability for a particular type of research. Anyone using mice or rats also has the choice of many strains.
This page lists the main species used in research with their frequency of use (as a percent) based on UK statistics for 2003 (no comparable statistics are available for the USA).
It also gives brief notes about isogenic, mutant, transgenic, and "knockout" strains and outbred stocks. The companion Isogenic web site explains why outbred stocks are usually poor experimental material where isogenic strains are available.
Species used in biomedical research
Mice (Mus musculus) have been domesticated for thousands of years and are easily the most widely used vertebrate. They are mammals and are small, prolific, and easy to maintain.
There are several hundred isogenic (inbred) strains which are like immortal clones of genetically identical individuals. Each strain has its own unique characteristics, and some of them get diseases such as cancer, diabetes, atherosclerosis and heart disease, so are used as models of similar human conditions. There is extensive background data on the characteristics of these strains, which have made an enormous contribution to biomedical research. For example, at least nineteen Nobel prizes have been awarded for work which would have been difficult or impossible without them (Festing, and Fisher 2000 Nature 404:815).
There are also a wide range of mutants such as obese, nude, hairless and dwarf mice.
It is relatively easy to produce transgenic mice by injecting DNA into fertilised eggs, and there is an extensive technology to inactivate specific genes either altogether or in specific tissues or at specific times. Mouse DNA has been fully sequenced, and several inbred strains are currently being individually sequenced.
There is strong genetic linkage homology between mice and humans so that once a gene location is known in mice it's position can be predicted in humans. Mice are widely used in fundamental research, but are also used in applied research such as in toxicity testing and drug development, although the rat is generally preferred by toxicologists. The use of mice is increasing at the expense of other species. The more that is known about the mouse, the more valuable it becomes as a research animal.
Rats (Rattus norvegicus) have only been domesticated since the mid-eighteenth centaury and were unknown in Europe until that time, having migrated from the far East. Rats are also small and prolific, although they are substantially larger than mice. Although about 3.7 as many mice than rats are used in research, this is not a direct reflection of their importance. Until recently more papers were published using the rat than the mouse.
Most research workers use outbred, genetically heterogeneous, rats such as Sprague-Dawley, Wistar or Long-Evans stocks. However, about 200 inbred strains of rats have been described, with about a dozen strains being commercially available in the USA, Europe and Japan. Where there is a choice, the use of isogenic strains is strongly recommended.
The rat genome has been fully sequenced and there are several interesting mutants such as the athymic Rowett nude and the obese "fatty" rat.
The rat tends to be more widely used in applied rather than fundamental studies. It is particularly favoured for toxicological screening, physiology and behaviour, possibly because it is larger than the mouse.
Fish (several species) are overwhelmingly used in fundamental studies. Although there is no breakdown by species in UK statistics, a large proportion are probably Zebra fish which are widely used in fundamental studies of development. The species is oviparous and the egg and developing embryo is transparent so that development can be observed directly in the live animal. Trout are also used in toxicity testing.
Chickens (Gallus domesticus) account for the vast majority of birds used in research, and most of these are used in applied studies in veterinary medicine. A large proportion of these are probably embryonated eggs which are counted in the UK once they reach half the incubation period. These are widely used to culture viruses.
Most of these are used in applied studies in human medicine.
Other rodents (1%)
These include hamsters, gerbils and others. These seem to be used when no other species is found to be suitable.
Most of these are beagles which are used in applied studies in human medicine, probably largely in toxicity testing and drug development.
Non-human primates (0.1%)
About three quarters are Old-World monkeys such as Macaques, with about a quarter of them are New-World monkeys such as marmosets. A large majority of them are used in applied studies in human medicine, and in particular in toxicity testing. Much smaller numbers are used in academic research, some of which involves neuroscience and the study of diseases such as Parkinson's disease which can not always be adequately modelled in rodents.
All other species
These include sheep (0.7%), pigs (0.4%), rabbits (0.6%), amphibians (0.3%) and others (0.2%). It is difficult to generalise about why they are used. However, sheep and pigs are often used in surgical studies because of their large size and ready availability, rabbits are probably largely used for the production of polyclonal antibodies, and amphibians are mostly used in fundamental developmental research.
Strains of mice and rats:
Isogenic strains include inbred strains produced by many generations of brother x sister mating, and the F1 (first generation) offspring of a cross between two inbred strains. They are like immortal clones of genetically identical individuals. The same genotype can be reproduced indefinitely, though over a period of time there may be some genetic drift due to the accumulation of new mutations. Such genetic drift is much slower than that seen in outbred stocks. Even this can be virtually eliminated using frozen embryo banks. A single individual can be genotyped at any locus, and this will serve to genotype all animals of that strain (because all are identical) so that a genetic profile can be developed for each strain. Inbred strains (but not F1 hybrids) are homozygous at all genetic loci, so will breed true, with no "hidden" recessive genes segregating within the population which may confuse experimental results. Genetic homogeneity leads to phenotypic uniformity, which means that smaller numbers of animals are needed to achieve a given level of statistical precision.
Strains and individuals can usually be identified from a small sample of DNA, so that an investigator can check to see whether the animals used in a particular study were of the correct strain. Each inbred strain has a unique set of phenotypic characteristics, such as types of spontaneous disease, response to xenobiotics, behaviour etc. Some of these are valuable for a particular research project, while others would preclude the use of a particular strains from some studies. Searchable lists of inbred strains of mice and rats and their characteristics are maintained by the Jackson Laboratory. On average isogenic strains are more sensitive than outbred stocks to experimental treatments, which also increases the power of experiments which use them. They are internationally distributed, so that work can be replicated all over the world.
Many mutations have arisen in mouse and, to a lesser extent, in rat colonies either spontaneously or as a result of irradiation or chemical treatment. These have been preserved for research. For example mutant athymic "nude" mice and rats have a defective cell mediated immune system and will grow human tumour xenografts. Various obese mutations have made a fundamental contribution to our understanding of obesity and metabolism.
Genetic background may influence the expression of a mutation. It is good practice to backcross them to an inbred genetic background to reduce genetic drift which may change their phenotype.
These strains are made by the incorporation of DNA from another source into the genome. This is usually done by microinjection into an early embryo. The aim is usually to get an animal to express a foreign gene. over-express a gene, or express a gene in an abnormal tissue or at an abnormal time. It is beyond the scope of this web to discuss the many uses for such strains.
These are strains in which one or more host genes have been inactivated either altogether or in a particular tissue or at a particular time. They are widely used in fundamental research. A full discussion of their properties is beyond the scope of this web.
These are closed (usually) breeding colonies within which there is some degree of genetic variation. The amount of genetic variation depends on the previous breeding history of the particular colony. If it has been maintained in relatively small numbers (say less than 25 breeding pairs) for several generations, then it may have become relatively homogeneous ( inbred). However, the extent of the genetic variation will not be known without a special genetic investigation. Investigators using species other than mouse or rats will usually have to use outbred stocks, as isogenic strains of other species are rarely available.
Although outbred stocks are widely used, they are not ideal experimental material because nothing is known about the genotype of any individual, and no two individuals are alike. This means that a "control" and a "treated" animal will differ not only in treatment, but also to an unknown extent in genotype. As a result, sample sizes will need to be greater than if isogenic had been used.
However, crossover "within-subject" experimental designs, using outbred stocks should be as powerful as ones involving isogenic strains (see 12. Experimental designs. Scientists sometimes try to justified the use of outbred stocks on the grounds that the aim is to model humans, and humans are genetically variable. However, a model does not have to represent it's target in every respect. If it did so it would not be a model. Mice differ markedly from humans in size, for example, and if they did not do so they would be a very much less valuable model because they would cost so much to keep. The fact that the use of outbred stocks may lead to less powerful experiments, less able to detect the effect of a treatment makes them less valuable than isogenic strains in research.