The estimation of risks from the induction of recessive mutations after exposure to ionising radiation.

Since recent assessments of genetic risks from radiation have concentrated on harmful dominant effects, a quantitative assessment of risks from recessives is needed. Presumably, harmful recessives can arise at all loci coding for essential proteins (perhaps 10 000), but mutation to dominant alleles is likely to be a property of relatively few loci. While many recessives doubtless remain to be discovered, those known at present tend to have earlier and more severe effects than dominants. Induced recessive mutations can cause harm by partnership with a defective allele already established in the population; partnership with another recessive mutation induced at the same locus; the formation of homozygous descendants, that is, identity by descent; and heterozygous effects. Calculations based on a combination of data from observations on human populations and from mouse experiments suggest that an extra genetically significant dose of 1 cGy (centiGray, equivalent to 1 rad) X or gamma irradiation received by each parent in a stable population with a million liveborn offspring would induce up to 1200 extra recessive mutations. From partnership effects, about one extra case of recessive disease would be expected in the following 10 generations. Homozygosity resulting from identity by descent could not normally occur until the fourth generation after exposure but, on certain assumptions, about ten extra cases of recessive disease would be expected from this cause by the tenth generation. In the same period, about 250 recessive alleles would be eliminated in heterozygotes (that is, Muller's 'genetic deaths') given 2.5% heterozygous disadvantage. These deleterious heterozygous effects should not be combined with those of dominants, as has been done in some previous risk estimates. It is considered unlikely that many radiation induced recessives would show heterozygous advantage. Certain dominants (combined frequently at least 10(-3)) should be excluded from calculations of mutational risk because they are unlikely to be maintained by mutation.