In article 1308950129000001 at genetics.comd.uic.edu, u58563 at uic.edu (Tommy the Terrorist) writes:
>I've been involved with a project in which the human form of a gene is 50% GC, and the mouse form is 60%. The coding sequence is only about 80% conserved on protein and DNA levels; the changes are primarily GC ---> AT transitions. Interestingly, about 75% of the CpG doublets in the mouse have been lost from the human (but almost all of the human CG's are present in the mouse). Nevertheless, the other transitions are predominantly in the same direction.
> Thus, we have a pair of homologues, one of which seems to differ from the other primarily due to (I presume) methylcytosine formation and conversion to thymine. It would appear to be a case in which evolution is driven neither by natural selection nor in a strictly random fashion.
> We have mulled a few notions of imprinting, proximity to heterochromatin, and so forth, but haven't really come up with any compelling ideas for experiments that could explain this odd observation.
> Any recommendations or comments would be most welcome.
Mammalian genomes are composed of isochores, that is large regions with
homogeneous GC content. Genes have different GC contents according to which
isochore they're in. The GC of genes tends to be very conserved between
different mammalian species, but exceptions exist. The most important
known difference is between mouse and rat vs other mammals (usually human) :
Mouse and rat genomes have a lower GC variance between isochores than other
mammals. Thus a gene which has a high GC content in human tends to have a
lower one in mouse and rat, and a gene which has a low GC content in human
tends to have a higher one in mouse and rat.
IMO what you observe typically falls into this pattern, general to the
genome, and probably does not come from something particular in the
history of your gene. Indeed a 50% to 60% GC change is very usual when
scanning the data bases for homologous genes between human and mouse or rat.
see : Mouchiroud and Bernardi (1993) Compositional properties of coding
sequences and mammalian phylogeny, J Mol Evol 37: 109-116, and references
I (and Dominique Mouchiroud, my boss) would be very interested in any
experimental results you may obtain in this field.
hope this helps, Marc