Shoot epidermal identity is critical for plant survival, growth, and interaction with the environment. Epidermal identity is specified during very early embryogenesis, and maintained in the outermost cells of the plant throughout the entire life cycle. In this work I aimed to generate a model for the establishment of basal epidermal cell fate during embryogenesis based on the analysis of both known and novel regulators. Loss of function of two HD-ZIP IV transcription factors, ATML1 and PDF2 had previously been shown to lead to embryo lethality due to loss of epidermal specification. In this study I uncover dosage dependency of ATML1 and PDF2 function during embryogenesis. By expressing functional ATML1 and PDF2 fusion proteins specifically in the epidermis, I developed a novel tool allowing demonstration of homo- and heterodimerization of these two transcription factors in planta. Using genetic and proteomic analysis I provide evidence that other HD-ZIP IV proteins are involved in epidermal specification together with ATML1 and PDF2, suggesting the presence of multiple regulatory protein complexes. Based on previous published and unpublished work, I tested the hypothesis that ATML1 and PDF2 form part of a regulatory feedback loop necessary for maintenance of epidermal identity, and involving cell-cell signalling mediated by the receptor kinase ACR4. Using a genetic approach I confirm that ATML1 and PDF2 likely act together with ACR4 in the specification of embryonic epidermal identity. I show that ATML1 and PDF2 negatively regulate both ACR4, and their own expression, most likely by binding to L1 box motifs. In contrast, I provide evidence that ACR4-mediated signalling participates in maintaining expression levels of ATML1 and PDF2. Mathematical modelling of the properties of the feedback loop supported by my results, suggests that it is capable of maintaining a robust epidermal cell fate, and predicts possible changes in network interactions during the process of epidermal cell fate specification. Finally I used a combination of bioinformatics approaches to integrate in silico and experimental data with the aim of discovering potential novel epidermal regulators and targets of epidermal fate specifying pathways. This work highlighted potential roles for WOX-family transcription factors in epidermal fate specification, which were further analysed genetically. In addition, bioinformatics analysis pinpointed an intriguing overlap between the targets of epidermal specification pathways and targets of abiotic stresses signalling.

TCP4 as an integrator of key developmental events A striking aspect of plant life is their sedentary life-style. Though it relieves them of the obligation of forming a complex body organization, it exposes them to environmental challenges. Plants have evolved a flexible pattern of post-embryonic growth. The major phases in their life cycle are photomorphogenesis, vegetative growth with phase transitions, reproductive growth and senescence. The phase transitions are coordinated temporally to ensure proper maturation of organism. Flexibility is built in the re-iterated programs of organogenesis, which provides a plant with an option to adopt an architecture best suited to prevailing environmental conditions. Organogenesis occurs by processes of cell division and maturation (expansion). Cell division determines the growth potential by generating the requisite number of cells and cell maturation fulfils the potential by elaborating the organ form. Organ growth requires spatially- and temporally-controlled cellular maturation. The TCP class of plant-specific transcription factors, conserved from bryophytes to angiosperms, control diverse developmental and morphological traits, such as plant architecture, floral asymmetry, seed germination, male and female gametophyte development and photomorphogenesis (Martín-Trillo and Cubas, 2010). Class II TCPs, which are targets of miR319, are best known for their role in leaf morphogenesis. They are believed to function by redundantly regulating the onset of cellular maturation ( Efroni et al., 2008; Koyama et al., 2007; Nath et al., 2003; Ori et al., 2007; Palatnik et al., 2003; Schommer et al., 2008). To establish the link between level of TCP activity and organ growth, we undertook the approach of hyper-activating the function of TCP4, a representative class II TCP, by fusing it with a strong transactivation domain. Enhanced level of TCP4 activity reduced organ growth by causing precocious cellular maturation. It also accelerated the process of organ initiation, maturation and its progression into the final stage of senescence. Hyper-active TCP4-expressing plants underwent faster maturation of shoot apex into reproductive phase. In general, hyper-activation of TCP4 advanced cellular, organ and organism maturation programs in Arabidopsis life cycle (Fig. 1). Traits such as organ initiation rate, organ size, flowering time and seed yield contribute to the fitness of the plant. Faster rate of organ initiation, bigger organ size, early onset of flowering and higher seed yield are obvious desirable traits. However, they rarely occur simultaneously in a mutant or a natural variant, suggesting that there is a trade-off among different traits. Studies have shown that such traits are linked and are controlled by multiple loci that contribute quantitatively to the phenotype. A change that benefits one trait may adversely affect another (Colautti et al., 2011; Kozlowski, 1992; Mendez-Vigo et al., 2010) . Our study shows that TCP4 activity can potentially coordinate these inter-connected traits. Though hyper-active TCP4-expressing plants have faster rate of organ initiation, the final organ size is reduced and senescence is advanced. These plants reach reproductive phase faster, but produce fewer seeds, hence limiting their propagation and lowering their fitness in comparison to the wild type. Such a genetic constraint on the traits limits the phenotypic variation that can be produced in plants and, hence, their adaptation to the environment. Our study suggests that TCP4 can link organ growth with that of the whole organism. It acts as a heterochronic regulator which possibly affects timing of multiple maturation programs. Any perturbation in the TCP activity may have far-reaching effects on plant growth and thus, optimal level of TCP activity is crucial for plant homeostasis. One possible explanation for the developmental pleitropy in TCP4 hyper-activation line is an alteration in hormone biosynthesis or sensitivity. A combination of microarray and hormone application studies on hyper-active TCP4-expressing line has indicated a reduction in the levels of GA and auxin and an increase in cytokinin and MeJA levels. There may also be an inhibition of auxin signaling and upregulation of MeJA and ethylene signaling. In addition, TCP4 appeared to regulate both GA biosynthesis and response in opposing manner. The molecular mechanisms involved in TCP4-mediated integration of hormonal pathways are still unclear. Answering these questions would require identification of its direct downstream targets.