Phenotypic traits are determined by three components; genetics, environment, and the interaction between genes and the environment (when genes produce different phenotypes dependent upon the environment). The genetic variance underlying phenotypes can be further divided in to subcomponents, the additive, dominance, and epistatic variances.
Additive genetic variance of a trait is the amount of variance caused by additive loci – where the number of each allele at the loci determines the phenotype. For an illustrative example, imagine two loci determine height in a diploid organism and there are two alleles each A/a and B/b respectively. An individual with the genotype aabb has a height of 1 metre, and each copy of A or B adds 20 centimetres to the height. Therefore Aabb and aaBb genotypes are 120 cm, AAbb, AaBb and aaBB are 140 cm, AABb and AaBB are 160 cm, and AABB are 180 cm. In summary , in an additive locus the heterozygote is the intermediate of the two homozygotes.
Box: Additive genetic variance
The amount of variation in phenotype determined by additive genetic variance is also known as the (narrow sense) heritability, a measure of how much offspring will resemble parents. The value is constrained to be between one and zero where a value of one means the offspring will perfectly resemble the midparent value (be an average of both parents) and zero when the phenotype of the parents has no relation to offspring phenotype. Analysing additive genetic variance, for the purpose of heritability estimation, is central to quantitative genetics. Better estimations will allow greater insight in to the heritability of disease, improvements in breeding programs (useful for both commercial and conservation purposes) and the theory of evolution.
Additive genetic variance has classically been estimated using experimental designs like full- and half-sibling breeding or pedigree analyses. One problem with these approaches, though I firmly believe they are still useful, is their limited power. For example a half-sib experiment might take a small number of males, each mated to just a handful of females therefore just capturing a small amount of the genetic variance within the population within each set of half-siblings. Further, while an individual in the offspring generation shares on average 25% of it’s genetic variance with its half-siblings (50% probability of having the same genetic material from one of the two parents), the same individual compared to one half-sibling selected at random could share anything from 0-50% of its genetic material. Therefore to get a good estimate of additive genetic variance from a half-sib design it is necessary to get large numbers of samples, both within half-families (many females per male) to maximise background genetic variation, and from many males to maximise the amount of genetic variance being measured. Obviously this is limited by feasibility – unfortunately money and time can be a huge constraint in science.
Half-sib experimental design: each father mated to multiple females
Hemiclonal analysis in Drosophila melanogaster, pioneered by the Rice lab in the 1990’s, offers a solution to this problem. I use hemiclones extensively for my PhD work and a recent paper by our lab demonstrates how useful they can be in quantitative genetic analysis. A hemiclone is analagous to taking one sperm, cloning it, and fertilizing many different eggs with them. All of the offspring produced from those crosses would share exactly half of their genetic material, a clonally amplified haploid genome, while the other half is random. This will give a very good precision in the estimate of additive genetic variance because it does away with one of the above problems (variance in the degree of genetic relatedness between relatives).
To make a hemiclone line we use genetic tools and take advantage of a lack of recombination in males.A randomly selected wild type male is crossed to a DXCG female, a female that has a double X-chromosome (two X-chromosomes fused together allowing her to also carry a Y-chromosome) and with both major autosomes translocated (one autosome is split and fused to half of the other autosome, leaving two chromosomes composed half of autosome 2 and half of autosome 3). This means three quarters of the offspring do not hatch (those with three X chromosomes, two Y-chromosomes, or 2.5 copies of either major autosome) and the surviving quarter are either brown eyed males or white eyed females (the translocated chromosomes are also marked with recessive and dominant eye genes which, when homozygous give white eyes, and when heterozygous give brown eyes). The brown eye males are then crossed to fresh DXCG females, and this cross can be repeated many times to maintain lines.
Hemiclone production: the first step shows how the line is produced, the second shows how it is maintained, and the final part is the production of many wild type focal flies each carrying the clonally amplified haploid genome.
By crossing the brown eyed males to wild type females, or females with a double X-chromsome (double X also gives a yellow body colour) and wild type autosomes, we can produce many offspring, female and male respectively. Because many fathers and mothers can be used to produce each line a high degree of replication can be achieved, and much of the standing genetic variance can be captured in the background genetic variance within lines. Furthermore, because lines can be maintained for many generations, many lines can be assayed in an experiment (using many small experimental blocks), enabling a better capture of genetic variation across the lines.
Chromosome diagram of the hemiclone production process
In the ~20 years since the advent of hemiclonal analyses in Drosophila there was initially a slow rate in the use of this method, however in recent years the publication rate is on the rise. Over half of the 34 papers that reference the Drosophila hemiclone analysis pioneered by Rice were published in just the last six years. The growing trend is indivicative that finally this powerful tool is being utilised on a greater scale within evolutionary biology.