Identifying the targets of selection during sunflower domestication
Sunflower poses an interesting model for domestication. Most studies, such as in maize and bean, have shown that a few loci (quantitative trait loci, or QTL) are responsible for the major phenotypic differences between wild and domesticated species, however in sunflower a lot more QTL are involved, with each corresponding to less of the phenotypic variance (Burke et al 2002).
The cloning of QTL, or the identification of genes responsible for domestication phenotypes, has resulted in the identification of genes for branching and fruit encapsulation in maize, fruit weight in tomato, and shattering in rice and wheat. These have been identified by gradually 'zooming in' closer and closer on the piece of the chromosome that corresponds to the phenotype until the actual gene is identified (the so-called 'top-down' approach). This approach relies on good genomic tools (a genome sequence helps!), nice big QTL, lots of time and probably lots of greenhous space greenhouse space! An alternative is the bottom-up approach:
1. Screen many genes (hundreds) for variation in a crop and its wild progenitor. Variation is usually estimated using microsatellites or DNA sequnce polymorphism
2. Determine the strength of the demographic bottleneck resulting from domestication.
3. Identify the genes that have lost significantly more variation than expected based on the bottleneck
These are your candidate domestication-related genes!
This strategy has been carried out in maize (Vigouroux et al 2002) and in sorghum (Casa et al 2005). In both a subset of loci showed the 'signature of selection'. We are doing this in sunflower. Approximately 500 loci have been screened on a panel of wild and domesticated sunflower and our data have indicated 36 loci of interest - 18 lost a significant amount of variation during the domestication of sunflower and another 18 have done so since domestication (i.e. during 'improvement' and intensive breeding). Of these, we have chosen 18 strong candidates based on significance under multple tests. We are now obtaining the full gene sequence for the loci that are putatively under selection. We want to see if traditional measures of selection using DNA polymorphism are also evident in these loci. In addition, did changes at the sequence level or changes in expression occur, giving rise to the domesticated sunflower phenotype? We are carrying out these experiments now!
Reducing the risks of transgene escape
Genetic modification may allow an increase in food production, and thus feed more people, reducing famine and malnutrition, however field trials are necessary to determine if the transgenes are able to escape from the crop and become established in a wild species via gene flow. We recently reviewed the risks of GM crops, and potential strategies to limit the escape of transgenes (Chapman and Burke, 2006a).
One of the most promising approaches to preventing transgene escape is to physically link the transgene to a gene that is unfit in the wild relatives of the crop, but neutral (or beneficial) in the crop (Transgene mitigation; Gressell, 1999).
Our lab is investigating this in a series of field experiments (spearheaded by Dr. Jutta Burger). Our lab recently genotyped inbred lines of wild x cltivated sunflower (Baack et al in review and Chapman and Burke, unpublished data) and these will be grown in the field in two sites over two years to determine if the fitness QTL are constant. If so, we have identified regions of the genome which would be promising for the transgene mitigation strategy.