MADS on the move : a study on MADS domain protein function and movement during floral development in Arabidopsis thaliana

In this thesis we investigated the behaviour of fluorescently-tagged MADS domain proteins during floral development in the model plant Arabidopsis thaliana, and explored the importance of intercellular transport via plasmodesmata for MADS domain transcription factor functioning. The MADS domain transcription factor family plays an important regulatory role in the development of flowers, among others by establishing the identities of the different floral organs. Although genetic screens and in vitro and in vivo studies on protein-protein and protein-DNA interactions provide important information on how MADS domain transcription factor complexes are able to regulate downstream target genes, understanding of the behaviour of MADS domain transcription factors in planta is still limited. Also, the extent to which intercellular movement of MADS domain transcription factors via plasmodesmata plays a role in developmental processes is poorly understood. Since the discovery of the GREEN FLUORESCENT PROTEIN (GFP) and the subsequent development of similar fluorescent tags, it has become possible to observe the subcellular localisation and behaviour of fluorescently-tagged proteins in living tissues with confocal laser scanning microscopy. In Chapter 2 of this thesis, different methods of tagging the MADS domain transcription factors AGAMOUS (AG), SEPALLATA3 (SEP3), and FRUITFULL (FUL) for chromatin immunoprecipitation, chromatin affinity purification and in planta imaging are described. This research shows that the addition of a small peptide tag or a fluorescent tag to MADS domain proteins easily leads to transgene silencing and specific loss-of-function mutant phenotypes, especially when the tagged MADS box genes are expressed under the control of the constitutive CaMV35S promoter. Plants that express tagged MADS box genes from genomic fragments that include all or most of the regulatory elements, and therefore mimic the natural expression pattern as much as possible, show lower levels of loss-of-function phenotypes. In addition, these plants are also more useful for investigating biological relevant behaviour of the MADS domain proteins. In Chapter 3, the spatio-temporal localisation patterns of GFP-tagged MADS domain transcription factors AG, SEP3, FUL and APETALA1 (AP1) during floral development are reported. These analyses demonstrate that there are several tissues, often epidermal cell layers, where MADS domain proteins could be detected, while the available literature describes an absence of mRNA in those tissues. This could indicate that there is intercellular transport of MADS domain proteins in meristematic tissues during floral development. The implications of the observed behaviour of the different MADS domain proteins for MADS domain protein functioning are discussed in this chapter. In Chapters 4 and 5 we describe the different methods that were used to investigate whether MADS domain proteins are indeed able to transport between cells during floral development. The difficulties that we encountered in our attempts to investigate intercellular MADS domain protein transport with microinjection techniques and by using the photoconvertible fluorescent mEosFP-tag are discussed. In plants that specifically overexpress GFP-tagged MADS domain transcription factors AG, SEP3, APETALA3 (AP3), or PISTILLATA (PI) in the epidermis, we demonstrated with a photobleaching technique that all tested proteins were able to move within the epidermal cell layer. This mechanism of lateral epidermal movement provides an explanation for most of the unexpected MADS domain protein localisations that we found in the spatio-temporal localisation analyses in Chapter 3. Additionally, we demonstrate that epidermis-expressed GFP-tagged AG is able to move from the epidermis to the subepidermis in the centre of the floral meristem, which provides proof for the suggestions that AG acts non-cell-autonomously in the floral meristem. In these plants we also analyzed the effects of epidermal MADS domain protein expression on the plant phenotype. This showed, among others, that epidermis-expressed AG is able to fully complement its own mutant background, while epidermis-expressed AP3 is not. In Chapter 6, we explore the mechanisms underlying the behaviour of GFP-tagged SEP3 during petal and stamen development that was observed in the spatio-temporal localisation studies described in Chapter 3. Just prior to the initiation of petal and stamen primordia GFP-tagged SEP3 proteins change their subcellular localisation from predominantly nuclear to more cytoplasmic, and at later stages GFP-tagged SEP3 protein seems to disappear in the middle of the primordia without the loss of SEP3 mRNA expression. These two processes could be regulated at a post-transcriptional level by two mechanisms that are discussed, namely 26s proteasome mediated SEP3 protein degradation and epidermal-oriented intercellular transport of SEP3 proteins. Additionally, we demonstrate that there are no clear indications that the observed GFP-tagged SEP3 behaviour is due to the presence of F-box protein UNUSUAL FLORAL ORGANS (UFO), which regulates petal and stamen development. In Chapter 7, this thesis finishes with some concluding remarks on in planta imaging of MADS domain transcription factors and the possible mechanisms of MADS domain protein movement in the floral meristem. Furthermore, we speculate on the importance of MADS domain protein movement in establishing MADS box gene expression patterns and MADS domain protein gradients, and on the need for symplastically isolated domains for proper floral development.