BABY BOOM-induced somatic embryogenesis in Arabidopsis

Under appropriate tissue culture conditions, somatic plant cells can be induced to form embryos in a process called somatic embryogenesis (SE). SE provides a way to clonally propagate desirable plants and is therefore an important plant breeding tool. SE has also fascinated scientists for decades as an expression of plant ‘totipotency’, the ability to regenerate a whole new individual through embryogenesis. This thesis aims to obtain a deeper understanding of somatic embryo induction in Arabidopsis by the transcription factor BABY BOOM (BBM), through identification and functional analysis of BBM-binding proteins and BBM target genes. <strong>Chapter 1</strong> introduces the concept of somatic embryogenesis, describes the different SE systems in Arabidopsis, and discusses the role of the plant hormone auxin and chromatin modifying proteins in this process. An overview is presented on the current knowledge on SE-induction through ectopic overexpression of certain transcription factor genes. These include <em>BBM</em>, as well as other genes that are studied in this thesis in relation to BBM. BBM is part of the eight member AIL subfamily of AP2/ERF domain transcription factors. <strong>Chapter 2 </strong>reviews the role of AIL proteins during embryogenesis, stem cell niche specification, meristem maintenance and organ positioning and growth. We summarize the gene regulatory networks in which AILs function and describe how these transcription factors integrate multiple hormonal inputs, with special emphasis on the interactions between AILs and auxin. Finally, we conclude that although the functions of AILs in plant development are well described, knowledge on the molecular mode of action of AIL proteins and the identity of AIL target genes is still limited. Transcription factors function in protein complexes and in <strong>Chapter 3 </strong>we show that members of the HOMEODOMAIN GLABROUS (HDG) transcription factor family physically interact with BBM and other AILs. <em>HDG</em> genes are expressed in the epidermis, the outer cell layer of the plant, where they promote differentiation of cells into specialized epidermal cell types, such as trichomes or stomata. We show that ectopic overexpression of <em>HDG1</em> leads to loss of root and shoot meristems, phenotypes that had previously been reported for loss-of-function <em>ail </em>mutants. Conversely, down-regulation of <em>HDG </em>genes led to reduced cell differentiation, enhanced cell proliferation and SE phenotypes, phenotypes that resemble those found in <em>AIL</em> overexpression lines. Moreover, we found that co-overexpression of BBM and HDG1 reduces the overexpression phenotypes of both proteins. These results suggest opposite functions of AIL and HDG transcription factors, with AILs stimulating cell proliferation and HDGs stimulating cell differentiation, with the ratio between the two proteins determining the developmental outcome. Finally, we show that HDGs and AILs regulate each other on a transcriptional level and that they share common target genes. A variety of AIL overexpression phenotypes has been described in the literature, with BBM and PLT5/AIL5 being the only known AILs that induce SE upon overexpression. We show in <strong>Chapter 4</strong> that all AIL proteins except AIL1 and ANT are able to induce SE, but that this phenotype relies on a high AIL protein dosage. Using BBM and PLT2 as AIL representatives, we show that an intermediate AIL concentration induces organogenesis (ectopic root and shoot formation) and that a low concentration inhibits cellular differentiation. In addition, we show that BBM and PLT2 induce direct SE when activated at seed germination, while post-germination activation leads to indirect SE from callus. The <em>LEAFY COTYLEDON</em> (<em>LEC</em>)/<em>LAFL</em> genes, which also encode SE-inducing transcription factors, are direct targets of BBM/PLT2 during direct SE, showing that these two SE pathways are linked. Using <em>LAFL </em>gene mutants, we show that the LAFL pathway is an important downstream component of BBM-mediated SE. <strong>Chapter 5</strong> presents the <em>in vivo</em>, genome-wide analysis of BBM DNA binding sites in somatic embryos using chromatin immunoprecipitation followed by sequencing (ChIP-seq). Our ChIP-seq and gene expression analysis reveal that BBM binds and positively regulates auxin biosynthesis genes and the recently discovered positive regulators of SE, the <em>AT-HOOK MOTIF CONTAINING NUCLEAR LOCALIZED</em> (<em>AHL</em>) genes. Knock-out of either pathway reduced BBM-mediated SE, showing that auxin biosynthesis and the <em>AHL </em>genes are important components of the BBM pathway. We also show that BBM binds to a consensus DNA motif that resembles the reported ANT binding motif. <strong>Chapter 6 </strong>reviews methods for identifying the direct target genes of a plant transcription factor using microarrays, as was done for HDG1 (Chapter 4). We describe which different systems can be used to control transcription factor activity, and how these can be combined with microarray analysis to identify target genes. In addition, we provide guidelines for the statistical analysis of microarray data and for the confirmation of candidate target genes. In plant biology, protein-protein interactions are often studied using bimolecular fluorescence complementation (BiFC) or split-YFP. In my BBM-HDG interaction studies I encountered problems using this method, which lead to the cautionary note on the use of BiFC presented in<strong> Chapter 7</strong>. BiFC is based on the restoration of fluorescence after the two non-fluorescent halves of a fluorescent protein are brought together by a protein-protein interaction event. However, because the fluorescent protein halves are prone to self-assembly, it is crucial to use proper controls and a quantitative read-out of fluorescence to avoid false positive interactions. We present a guideline for the setup of a BiFC experiment, discussing each step in the protocol. <strong>Chapter 8 </strong>discusses how the results presented in this thesis contribute to our knowledge on AIL transcription factors and somatic embryo induction, as well as the questions that still remain. An extended model of dose-dependent AIL function is proposed, as well as mechanisms by which the AIL-HDG interaction could function at the molecular level. Finally, an overview is provided of the molecular-genetic intersection between the different transcription factor-induced SE pathways.