What if strain differences are found?

If  strain differences are found in a  multi-strain study this implies that the outcome is genetically determined.  Further action depends on individual circumstances. For example, in a screening experiment there may not be any great need to study any differences in detail, although the reasons why strains differ is always of general interest. However, in some cases the reasons for the strain differences could be of great importance in assessing relevance to humans. Two situations are discussed briefly below.

Use of a multi-strain study to find the most suitable strain for future work

    Multi-strain studies can be used to find one strain which is "susceptible" or in some way more suitable than other  strains for further study. Research can then be concentrated on this  single strain. This is a perfectly acceptable strategy  but the findings may not generalise to other strains. The strain used should be  noted explicitly in the title or abstract of the paper.

    Pharmaceutical companies often use animal models for screening potential drugs, and may carry out essentially the same assay many times with slightly different compounds. If they can find a highly sensitive strain of mice or rats, then sample sizes can be reduced or the power of their  screens will be enhanced. A multi-strain study may, therefore, be used  as part of the process of optimising such an experiment, with the most  sensitive  strain being chosen for the screen (see Shaw R, Festing MF, Peers I, Furlong L. 2002. Use of factorial designs to optimize animal experiments and reduce animal use. ILAR J 43:223-232., available on-line at ILAR web site).

Use of a multi-strain study to find loci controlling a character of interest

    Strain differences may  be due to a single genetic locus or, more usually, to multiple loci with environmental influences, i.e. the character has a polygenic mode of inheritance. There are several approaches to  identifying the loci involved.

    The first step is usually to see if the strain difference is due to a  single gene. If several strains have been phenotyped, and they fall into two distinct classes, or one strain is very different from all the others, then this may suggest that the difference is due to a single Mendelian locus. In this case crosses of a "positive" x "negative" strain can be made. The F1 hybrid should then be phenotyped to see whether there is a dominant/recessive mode of inheritance. The F1 hybrids can then either be inter-crossed and the progeny phenotyped to look for the 3:1 or 1:2:1 Mendelian ratios, or they can be backcrossed  to the recessive strain (assuming a dominant mode of inheritance) to look for the Mendelian 1:1 ratio. Tissue samples or DNA should be saved  from all animals.

    If Mendelian ratios are found, then the next step is usually to map the  chromosome location of the locus. If two common inbred strains have been  used then data on many DNA-based genetic markers such as microsatellites  and single nucleotide polymorphisms will be available on various web sites. Basically, the method is to choose 3-4 of these marker loci per mouse chromosome at which the parental strains differ  and type all  the F2 and/or backcross animals at these loci. All these markers can be  typed using PCR-based methods. Any association between a marker locus and phenotype will be indicative of genetic linkage.

    Once an approximate chromosome location has been found, larger numbers of animals and markers on that chromosome will be used to fine-map the  gene. The final step is to identify the polymorphic gene. There may be some good candidate loci in the chromosomal region, or it may be necessary to clone and sequence the gene. At this stage, expert advice is usually advisable, as there may be several different approaches which  could be considered, depending on circumstances.

    If several inbred strains have been phenotyped and they do not seem to fall into distinct groups, but have a range of "susceptibility", then the character probably has a polygenic mode of inheritance. In this case  the aim is to identify the Quantitative Trait Loci or QTLs that control  the character. The methods used are very similar to those described above. Usually two strains with the highest and lowest "susceptibility" are chosen, and the analysis aims to map the loci as which they differ. As above, the strains are crossed to produce a backcross or F2 generation and all animals are phenotyped and typed at marker loci at  which the parental strains differ. Again, any association between a genetic marker and phenotype is suggestive of genetic linkage, but in this case there may be linkage to several chromosome locations.

    The main problem with the QTL analysis is going from map location to the actual locus, so expert advice is essential.

    Derived strains
    A whole range of specialised strains have been developed which may help  in the genetic analysis of a character. Sets of Recombinant inbred  strains (RIS) can help to determine whether a character is due to a  single genetic locus, and if so it's approximate chromosomal location.  Sets or recombinant congenic (RCS) strains have  been developed to help in the genetic analysis of complex traits. Similarly, sets of consomic strains in which a whole chromosome has been substituted for one of a different strain can indicate which chromosomes contribute  polymorphic genes to which characters. These strains are discussed in  more detail under the Isogenic strains page.