Several recent reports (8, 10, 11, 13) have established the biological and molecular genetic similarity between the endogenous AKV virus of strain AKR, and an N-ecotropic endogenous virus found in the genome of feral Japanese mice, Mus musculus molossinus. The similarities are so striking as to suggest a common origin of these viruses, which are present in some, but not all, inbred mouse strains. The virogenes of AKR mice may have been acquired by either: (a) common descent of AKR (and other AKV(+) strains) from a common ancestor of AKR and molossinus animals, or (b) horizontal germ line infection of the AKR strains by molossinus virus at 1;he strain's inception followed by fixation through inbreeding. The sexual descent model carries with it a prediction of relative consanguinity of the AKR strain and molossinus, whereas the horizontal infection model does not. We have examined the polymorphic allozyme (allelic isozyme) genotype of 51 nonvirus-related loci in 17 strains of mice including AKR, C58, BALB/c, Swiss, and molossinus. By comparing the composite allozyme genotype of different inbred and outbred mouse strains, the "genetic distance" statistic was derived. Genetic distance measures the degree of allelic substitution between populations and increases proportionately with the amount of time the populations have been reproductively isolated. The genetic distance computed between molossinus and AKR is large, nearly 5-10 times the distance between known related populations and strains (e.g., C57L vs. C57BL/6). Molossinus had a similarly large distance from AKV negative strains (Swiss, C57L) as it did from AKV- positive strains.
Cellular DNA sequences that flank the integrated AKV provirus were analyzed by restriction enzyme digestion of liver DNA from molossinus, AKR, and additional inbred strains that express ecotropic murine leukemia virus. The integration flanks of three AKR provirus sequences, Akv-1, Akv-2, and a third uncharacterized sequence, were not evident in molossinus cell DNA, which contained at least six different proviral integration fragments. These data effectively exclude the interpretation of consanguinity of AKR and molossinus and support the notion of acquisition of the endogenous virus in AKR by horizontal infection of the molossinus virus.