Edinburgh Research Explorer

Genetic Studies of Quantitative Trait Variability in Drosophila Melanogaster

Project: Research

StatusFinished
Effective start/end date1/04/0331/03/06
Total award£176,356.00
Funding organisationBBSRC
Funder project referenceG18768
Period1/04/0331/03/06

Description

Investigation of the genetic basis of natural variability in female fecundity and lifespan in the fruitfly Drosophila melanogaster

Layman's description

The methods of molecular genetics have proved enormously powerful in telling us about the nature of the genetic material, how genes control development, and how mutations in genes can cause major effects on the organism, including genetic diseases. They have also revealed a huge amount of variation between individuals of a species in the sequence of the ‘letters’ of DNA; several million distinct DNA variants are known to exist in human populations. But we know little about the way in which this variation is connected to variability in traits which we can see or measure, at the level of the organisms themselves (e.g. body size). Most of this variability is quantitative: it is impossible to group individuals into discrete categories, and the population can only be described by its average value and a statistical measure of variation around this average.
We do not know if the genetic variability in such quantitative traits is mainly due to rare mutations with small effects in many different genes, or to a small number of genes with relatively large effects, with variants present at intermediate frequencies. Recently, statistical methods have been developed that allow us to test between these two hypotheses, and to get an estimate of the numbers of genes that contribute to natural variability in a quantitative trait. These rely on measurements both of variability and the effect of inbreeding (matings between close relatives, such as brothers and sisters) on the average value of a trait.
We carried out two experiments on the egg production of young female flies, which is an important aspect of their fitness. One involved comparing the effect of inbreeding with the amount of a measure of genetic variation in the trait. The results suggested that much of the variability was contributed by a relatively small number of genes. The other experiment involved comparing the effects of selection for increased or reduced egg production on the average value and on the effect of inbreeding. The results showed that there must be genes with variants at intermediate frequencies contributing to the variation in the trait. We carried out a third experiment on the mortality rates at different ages of non-reproducing female flies. The results have not yet been fully analysed, but indicate that there is probably a significant contribution from mutations at low frequencies, especially at late ages, where the effects of inbreeding are much larger.
The experiments on egg production suggest strongly that this important trait is affected by some genes with variants at intermediate frequencies in the population, probably maintained by some form of natural selection. This opens up the prospect that these genes can be identified by genetic mapping methods, and the ways in which they control the trait determined.
We carried out a third experiment on the mortality rates at different ages of non-reproducing female flies. The results have not yet been fully analysed, but indicate that there is probably a significant contribution from mutations at low frequencies, especially at late ages, where the effects of inbreeding are much larger.
The experiments on egg production suggest strongly that this important trait is affected by some genes with variants at intermediate frequencies in the population, probably maintained by some form of natural selection. This opens up the prospect that these genes can be identified by genetic mapping methods, and the ways in which they control the trait determined.

Key findings

The main experiment was designed to answer the question of whether loci with deleterious alleles at intermediate frequencies make a major contribution to quantitative trait variability. The effects of a few generations of artificial selection on the mean (M) and the inbreeding load (B) of a trait provide an indicator of the extent to which partially recessive alleles at low frequencies contribute to variation in the traits: if the ratio R of the change in - B to the change in M is large and positive, it is likely that trait variation is mostly due to rare, partially recessive alleles, as postulated in mutational models of quantitative trait variability in fitness components. Selection for early life female fecundity in the fruitfly Drosophila melanogaster gave results that clearly contradicted this expectation, and suggest that variants at intermediate frequencies contribute substantially to variation in this trait.
A second series of experiments, in which components of genetic variance and the level of inbreeding depression were simultaneously estimated, showed that the level of dominance variance compared with the level of inbreeding depression was relatively high. This is also what is expected with important contributions from variants at intermediate frequencies.
Another set of diallel cross experiments concerned the mortality rates of virgin female flies; in this case, the results indicated that low frequency variants play an important role.

Activities

Research outputs