Can chromosomal heterosis inDrosophilabe explained by deleterious recessive genes? Negative results from a dichromosomal population test

Abstract

SummaryHigh levels of chromosomal heterosis have previously been detected inDrosophilausing the balancer chromosome equilibration (BE) technique, in which single wild-type chromosomes are introduced into population cages along with a dominant/lethal balancer chromosome. The balancer chromosome is rarely eliminated in such populations, showing that the fitness of chromosome homozygotes must be low by comparison with chromosomal heterozygotes. As with all cases of chromosomal heterosis, the underlying cause could either be deleterious recessives at various loci or generalized overdominance. The experiment of the present paper examines the first of these explanations. Population cages containing just two wild-type chromosomes (dichromosomal populations) were set up and allowed to run for many generations. Single chromosomes were then re-extracted from these populations, and their fitness measured using the BE technique. Our expectation was that the gradual elimination of recessive genes from the dichromosomal populations ought to result in an increase in the fitness of such re-extracted chromosome homozygotes. Yet in two replicated experiments we were unable to demonstrate an; unequivocal increase in fitness. We have estimated the rate of increase of fitness under multiple locus dominance and partial dominance models. The principal unknown parameter in these calculations is the selection intensity per locus,s. The expected increase is approximately proportional tos, and we estimate that values ofsaround 1/64 should be detectable in our experiments. However linkage is expected to reduce the efficiency of the dichromosomal procedure We show by computer simulation that this reduction is by a factor of approximately 2, thus increasing the detectable level ofsto approximately 1/32. Consideration of mutation-selection balance models shows that this is a feasible selection intensity only if dominance is nearly complete. Thus we are unable to rule out the notion that the genes responsible for heterosis are maintained by a simple mutation-selection balance, but the experimental results constrain the parameters of such a model to a narrow range.