MADS interactomics : towards understanding the molecular mechanisms of plant MADS-domain transcription factor function

Protein-protein and protein-DNA interactions are essential for the molecular action of transcription factors. By combinatorial binding to target gene promoters, transcription factors are able to up- or down-regulate the expression of these genes. MADS-domain proteins comprise a large family of transcription factors present in all eukaryotes. In plants, and especially in seed plants, this family has significantly expanded. For example, more than 100 representatives are found in the <em>Arabidopsis</em> genome. MADS-box genes have initially been shown to play major roles in flower development, however their emerging functional characterization revealed functions in almost all developmental processes throughout the plant life cycle. How MADS-domain transcription factors acquire their functional specificity remains unresolved. The goal of this thesis was to characterize some of the molecular mechanisms by which MADS-domain proteins act in <em>Arabidopsis</em>. Chapter 1 comprehensively reviews functions of MADS-domain transcription factors in flowering plants, with a main focus on <em>Arabidopsis</em>. Major classes of MADS-domain proteins are introduced, and their modular structures are described. Additionally, it is shown that several distinctive subfamilies of MADS-box genes can be inferred from the phylogenetic analysis of the whole gene family. By compiling recent studies on MADS-domain protein-protein and protein-DNA interactions, we present a hypothetical model of MADS-domain protein action that combines higher-order protein complex formation and active chromatin remodeling by large transcriptional machineries. Chapter 2 describes MADS-domain protein complexes that are potentially formed during <em>Arabidopsis</em> flower development. By using a targeted proteomics approach we were able to characterize the protein interactome of major floral homeotic MADS-domain proteins (APETALA1, APETALA3, PISTILLATA, AGAMOUS, SEPALLATA3 and FRUITFULL) in native plant tissues, confirming interactions suggested in the ‘floral quartet’ model. Additionally, we discovered transcription factors from other families and chromatin-associated proteins as possible interaction partners of MADS-domain proteins. These interactions shed light on the combinatorial modes of action of MADS-domain transcription factors and suggest that they can act by recruiting or redirecting the chromatin remodeling machinery to control the expression of their target genes. In Chapter 3 we review recent advances in proteomics approaches used to study cellular signaling and developmental processes in plants. We mention the emerging tools for of whole plant proteome characterization as well as sub-cellular protein localization. The major focus, though, is on the description of complete cellular signaling cascades in plants, starting from the characterization of signaling mobile molecules (e.g. peptide or protein), through identification of receptors and receptor protein complexes, ending with identification of intermediate signaling pathway members. Two examples of biochemical procedures used to identify complexes of membrane-bound receptors and transcriptional regulators from nuclei are described in Chapter 4. In our optimized method we make use of fluorophore-tagged single step affinity purification of protein complexes and label-free mass spectrometry-based protein quantification to distinguish true complex partners from non-specifically precipitated proteins. The exact molecular mechanisms of DNA sequence recognition by MADS-domain transcription factors are still unknown. Particularly intriguing is the question whether various MADS-domain protein complexes possess different DNA-binding specificities. We address this question in Chapter 5. We used systematic evolution of ligands by exponential enrichment (SELEX) followed by high-throughput sequencing (seq) approach to discriminate DNA-binding specificities of several MADS-domain protein homo- and heterodimers. Finally, in Chapter 6, we aimed to identify the molecular features of different DNA-binding specificities of MADS-domain transcription factors.With help of bioinformatics tools and <em>in vitro</em> DNA-binding assays we found that structural characteristics of the DNA play an important role in DNA-binding of MADS-domain proteins. Taken together, research described in this thesis advances our knowledge on the molecular mechanisms of MADS-domain transcription factor action in plants. Chapter 7 concludes the thesis and describes future perspectives in MADS-domain protein research. Highlighted are the advances of high-throughput (proteomics and genomics) technologies that could be used to unravel not only the static characteristics of transcriptional regulation but also the dynamic and stoichiometric changes of complex protein and gene regulatory networks during plant development.