Natural variation in shade- and ethyleneinduced differential growth and transcription in Arabidopsis

Natural variation exists for many traits, including those that enable plants to cope with environmental stress. The genetic mechanisms behind coping with these stresses have been studied extensively and some regulatory networks have been unraveled. However, due to the complex polygenic nature of many stress-related traits a lot remains unknown. A promising new technique called “genetical genomics” proofed to be feasible in Arabidopsis during this thesis project. The data generated through genetical genomics enables the construction of putative regulatory networks. The general research question of this thesis is therefore to unravel regulatory networks operating during ethylene- and low light-induced hyponastic growth in Arabidopsis. To this end QTL analyses of phenotypic leaf angle traits and global transcription analysis (genetical genomics) will be used to explain and dissect natural variation in hyponastic growth and rosette compactness. Chapter two describes hyponastic growth in detail and has a focus on a QTL study to find large effect controlling loci for leaf angle related traits. Two leaf regions of differential growth were identified and could be separated as genetically (partly) independent. Furthermore, a difference in response to ethylene compared to low-light was found. A QTL involved in the response to both ethylene and low-light was found to be caused by the allelic difference of the ERECTA gene. Some other QTLs could also be narrowed down to a single gene. In Chapter three the light influenced growth traits; rosette area, rosette diameter (fermax), rosette compactness and Relative Growth Rate (RGR) are investigated by QTL analysis. Several QTLs were found and for some the underlying gene could be identified. ERECTA again proved to play an important role in these traits, but also PHYB had a functional allelic difference. In chapter four a method is described to construct a network for transcript regulation by combining genetical genomics data with single gene perturbation transcript profiling. By using the difference in transcript abundance between the ERECTA mutant Ler and its wild type Lan as a start we constructed (part) of the downstream signaling cascade. In chapter five we concentrate on the comparison of two genetical genomics experiments using the Ler x Cvi RIL population. The results of a experiment, where the RIL population is treated with 3 hours of low light, are compared with a previous experiment on the same population (Keurentjes et al., 2007). Apart from network construction, the genetical genomics data of low light treated Ler x Cvi RILs is used to explain variations in light affected traits measured in chapters two and three. In chapter six a NIL is used to study the possibilities to extend the method of network construction described in chapter four. Although in chapter four the effects of a single gene were used, often the gene underlying a QTL is not known and the whole locus needs to be studied. By the transcript profiling of a NIL the effects of a single “small” introgression on transcript levels are known and can be used as a start for constructing the regulatory network. Chapter seven contains a general discussion about the research described in this thesis.