Redundancy scheme

Francois Coulier coulier at
Sun Sep 22 11:18:47 EST 1996


I am working on a gene family encoding growth factors (the FGF family).
Several mouse genes have already been knocked out. While one of the KO
lead to very severe phenotype (embryonic lethal, due to lack of
implantation), most ko's lead to very mild to no detectable phenotypes. By
mild phenotype, I mean for example the angora phenotype associated with KO
of FGF5. 
In addition, in most cases the tissues involved in the observed phenotypes
are only one (and minor, in term of RNA level for example) site where
expression of the gene has been observed. For example, FGF5 is highly
expressed in the blastocyst, and the only observed phenotype is long hair.
To explain this, one called for redundancy between genes of the family,
that is other FGF genes expressed in the same tissues and at the same time
as the knocked out gene, are able to rescue its function. For example,
other FGFs are expressed at the blastocyst stage, and may replace FGF5.
I believe this redundancy scheme has been call for in the case of other
multigene families.

The question I have is how such redundancy scheme, if real, can be
selected for. 
If one assumes the gene is indeed dispensable, should'nt have it been lost
during evolution (or not kept at all in the first place)?
If one assumes the mild or undetected phenotypes are such in laboratory
conditions only, and strongly selected against in the wild (for example
long hairs may favour parasites), why will redundancy at, say, the
blastocyst stage, be selected for?

Any comment on this redundancy scheme will be highly appreciated! Thanks
to all of you.

Francois Coulier
INSERM Unite 119
27 bd Lei Roure
13009 Marseille

tel (33) 91 75 84 11 (October 18, 1996=> 33 (0) 4 91 75 84 11)
fax (33) 91 26 03 64 (October 18, 1996=> 33 (0) 4 91 26 03 64)
email: coulier at

More information about the Mol-evol mailing list