Some form of genetic quality control of inbred strains is highly desirable. It is not unknown for a supplier to send entirely the wrong strain so somebody setting up new colony would be well advised to have it checked to ensure that it is authentic.
DNA fingerprint of five mouse strains. Note that modern quality programs use single locus markers such as microsatellites or single nucleotide polymorphisms (SNPs) rather than DNA fingerprints.
Prevention is better than cure (in fact there is no cure once a colony is contaminated)
Colonies should, as far as possible, be kept physically separated. If they have to be mixed in the same room, strains with different coat colours should be housed together. From then on, periodic checks should be made to ensure that the colony has not become genetically contaminated by a non-strain mating, paying particular attention to the Foundation Stock colony (see Breeding and Maintenance page) as this is the only one which contributes to the long-term survival of the strain. Generally, it will only be possible to genetically monitor a small sample from the multiplication colonies, so again the main effort should be directed at preventing genetic contamination in the first place.
Technical staff need to be well trained and in particular need to be encouraged to note any changes in the colony, and bring them to the attention of the colony supervisor. A culture of blame should be avoided.
Quality control programs detect genetic contamination, not new mutations.
The aim of genetic quality control programs is to detect genetic contamination of one strain through an inadvertent mating with another strain. Currently it is not possible to monitor strains for new mutations except by observing the phenotype. Mutations which affect invisible characters such as minor changes in immune response, physiology or susceptibility to infectious organisms may go undetected for many generations. However, genetic quality control methods can often be used to distinguish between different sub-strains.
Historically, several methods of genetic quality control have ranged from the use of biochemical polymorphisms to the use of quantitative characters such as the shape of the skeleton and breeding performance [1,2]. Biochemical polymorphisms are often technically difficult to determine and somewhat limited in their distribution among strains of mice and rats. Skeletal morphology and breeding performance have the disadvantage of giving a statistical result, rather than a clear cut positive or negative answer, though breeding performance should be routinely monitored for husbandry purposes, so any change should be investigated.
The development of a large number of microsatellite and other DNA-based genetic markers such as single nucleotide polymorphsms (SNPs) detected using the polymerase chain reaction (PCR) has now completely changed the situation, although changes in phenotype noticed by animal technicians or scientific users of the animals continue to be an important way in which possible genetic contamination is first identified. The great advantage of DNA based methods is that only a small sample of tissue is needed, it can be stored indefinitely in the deep freeze, and the techniques for genotyping are essentially the same for every locus. The only differences are in the sequences of the PCR primers and possibly some minor changes in the conditions for the PCR reaction.
Microsatellites, which are the most widely used markers, are short repetitive DNA sequences with unique flanking regions. They are highly polymorphic in the number of repeats. PCR primers usually consist of about twenty base pairs of the unique flanking DNA for each microsatellite. There are many thousands of microsatellites in the mouse and rat genomes, and primers are commercially available for many of them from Research Genetics (www.resgen.com). The basic technique involves taking a sample of tissue from the animals to be tested, preparing DNA, and amplifying one or more of the microsatellites using PCR with the appropriate primers. The resulting reaction mixture is run on an agarose or polyacrylomide gel with control samples from animals with a known genotype, where necessary. If agarose is used, alleles are usually visualised with UV light after staining with ethidium bromide, with different length alleles running different distances. If, with an inbred strain, the DNA bands are not be aligned, this shows that the genotypes are not identical. Technical methods are given in many publications which use these markers, and by Lit (1991) .
There are several variants on this basic method. The system can be automated using DNA sequencing apparatus, though this is expensive and would normally only be economical if done on a large scale or if the apparatus is already available. As the bottleneck is usually running the gel, another alternative is to pool 5-10 samples of the PCR product in each well. Following electrophoresis this will result in a strong band, with satellite bands if one or more of the samples has a different genotype.
Sample size, number of markers, and sampling frequency
The main difficulty with genetic quality control is in deciding the number of genetic loci to use, the sample size, and sample frequency. Many microsatellite loci have more than one allele at each locus, though sometimes these can only be identified using acrylomide rather than agarose gels. Based on about 7,500 comparisons there is about a 74% chance that two unrelated inbred mouse strains will be the same at any given microsatellite locus, assuming a resolution of six or more base pairs. This means that two unknown strains should be tested at ten loci to give a 95% chance of detecting one or more differences. However, any strain or stock will normally only be at risk of becoming genetically contaminated by other strains in the same animal house, so a critical set of markers can be chosen which will detect any contamination from these strains. Usually, loci should be also chosen which are on different chromosomes to increase their statistical independence, and known alleles should differ by more than about five base pairs so that they may be easily identified.
For routine monitoring the sample size depends primarily on the presumed extent of any genetic contamination. A high level of contamination, say above 20%, can be detected with small sample sizes, but it is virtually impossible to detect a couple of wrong matings in a colony of a thousand breeding cages. Table 6 shows the sample size required to detect different levels of genetic contamination at a specified level of probability. From this it is clear that the best approach is to do everything possible to avoid contamination in the first place.
Inbred strains are relatively easy to monitor because all individuals should be identical at all microsatellite loci, apart from any recent mutations. These will be quite rare. However, at present there are no agreed standards on the number of loci to test, or sample size and frequency. Generally, the effort expended in monitoring each strain should depend on the chance of genetic contamination, with account being taken of the importance of the colony and the likely damage to research or reputation of the breeder from a contamination. Danger arises when colonies are first established because there may be no real assurance that they are what they are supposed to be, and from maintaining animals of several strains in the same animal room as this clearly increases the chance of a mis-mating.
Ideally, newly established colonies should be tested as soon as possible to ensure that they are of the correct genotype. With mice, control DNA from most strains is available from the Jackson Laboratory (www.jax.org). For rats, samples of DNA may be obtained from colleagues or known holders of the strains. If the colony is being established from a small breeding nucleus, it may be possible to test all animals. In this case, the main aim is to test the authenticity of the strain, though the possibility of contamination by one or more non-strain animals should not be ruled out. A sample of 5-10 animals is probably adequate at this stage, and they should probably be tested at about ten microsatellite loci.
Breeding colonies of inbred strains will often be divided up into a stem line and an expansion colony (see above). Ideally, the stem line colony will be physically separated from other colonies. If not, it should at least be kept with strains of a different coat colour and microsatellite profile. With good physical separation and a small colony size, the chance of contamination is low, so the colony would not need to be monitored very frequently. If it is maintained in an isolator with no other strain, then once it has been authenticated it hardly needs any routine monitoring.
Expansion colonies may be large and at risk from other colonies in the same building. The colony might be monitored 2-4 time per year, with sample sizes of about ten animals, using a set of markers which will preclude contamination by all othe strains in the building.
Consult the users
In practice, genetic contamination is often picked up by the users of the animals, who obtain unexpected results. If DNA samples can be obtained from the abnormal animals (scientists should keep frozen tissue samples of their animals), then there is a very good chance that contamination will be obvious using a few microsatellite markers. Conversely, if they match other animals in the colony and reference samples of DNA at 10-12 loci, then there is a good chance that the abnormal response is not due to genetic contamination. It may, however, be due to a new mutation, and should be investigated as such. If live animals are still available, these should be outcrossed to an unrelated strain, and possible Mendelian segregation should be studied in the F2 and/or backcross generations.
1. Nomura, T., Esaki, K. and Tomita, T., ICLAS Manual for genetic monitoring of inbred mice,, University of Tokyo Press, Tokyo, 1984, .
2. Hedrich, H. J., Genetic monitoring of inbred strains of rats,, Gustav Fischer Verlag., Stuttgart, New York, 1990, .
3. Litt, M., PCR of TG microsatellites, in PCR A Practical Approach, McPherson, M. J., Quirke, P. and Taylor, G. R., (Eds.), IRL Press at Oxford University Press, Oxford, New York, 1991, 85.