Trends in Neurosciences
Mice, gene targeting and behaviour: more than just genetic background
Section snippets
How many mutations might the flanking DNA from strain 129 contain?
To answer this question we must ask - how different is the 129-mouse genome from that of other strains of mouse commonly used in transgenic experiments? While few systematic studies have been done, it is well established that gene-targeting constructs based on non-129 genomic DNA give poor efficiencies of homologous recombination in ES cells from 129 mice32, 33. It is presumed that even rare mismatches between incoming and resident DNA either provide sites for repair enzymes to act (preventing
Evidence suggesting that the flanking DNA from strain 129 can make a significant contribution to the phenotype
What evidence is there to support this view? One prediction this line of analysis makes is that phenotypes should evolve during successive backcrosses as markers from strain 129 segregate away from the targeted-gene locus. Indeed, heterogeneity of phenotype has been observed previously. For example, Silva et al.[2]take care to note that the deficit in long-term potentiation (LTP) recorded in mice deficient for a type of Ca2+-calmodulin kinase type II, alpha-CaMKII, was only present in 14 of 16
Many factors other than genetic background complicate interpretation
While all of these are sensible and pragmatic options, none would appear to address all the possible factors that are likely to complicate interpretation (Table 1). The most significant and unavoidable of these is, possibly, our inability to attribute a phenotypic alteration (for example, impaired spatial learning or deficits in hippocampal LTP) to the loss of gene function in the brain region under study (in this case the hippocampus). Loss of gene function in ‘knockout’ mice occurs in all
Are there any solutions?
Most, or all, of these caveats apply across the board to the analysis of physiological processes using gene-knockout technology. What then is the answer? Two approaches might be envisaged. One will be to attempt to restore function of the targeted-gene product by microinjecting mRNA (or the mature protein) into cells derived from mutant mice, or derived from double-null ES cells and differentiated in vitro. If normal function is restored then the phenotypic deficit might be attributed to the
Acknowledgements
The author is indebted to the many colleagues who have contributed to this discussion. J.J. Mullins, S. Prusiner and J. Wehner are thanked for communicating unpublished data.
References (55)
Pharmacol. Biochem. Behav.
(1984)- et al.
Pharmacol. Biochem. Behav.
(1994) Pharmacol. Biochem. Behav.
(1986)- et al.
Pharmacol. Biochem. Behav.
(1979) - et al.
Pharmacol. Biochem. Behav.
(1989) Pharmacol. Biochem. Behav.
(1989)Brain Res.
(1990)- et al.
Dev. Brain Res.
(1994) - et al.
Behav. Brain Res.
(1994) Behav. Neural Biol.
(1992)