• cytokinins
• embryos
• epiphylly
• helianthus
• IAA
• Interspecific hybrid
• Knotted1-like gene
• Leafy cotyledon1-like gene
• meristems
• totipotency

In higher plants, the zygote divides to produce the embryo, a bipolar structure with one, two or several embryonic leaves (cotyledons), the shoot apical meristem (SAM) and the root apical meristem (RAM). Throughout plant life, apical meristems continually produce new organs (branches, leaves or roots), which are continuously added to the plant body. An important consequence of this unlimited and reiterative growth, defined as “recurrent ontogenesis”, is the absence of separation between germ-line and soma; thus any somatic cell is a potential progenitor of a new individual. This capacity (totipotency) is revealed by processes of vegetative propagation in vivo, such as: i) adventitious embryos from nucellar cells; ii) adventitious embryos from single cells of epidermis of stems and leaves; iii) adventitious buds of multicellular origin. Moreover, somatic cell totipotency is most clearly expressed in the ability of plant cells to develop into a complete fertile plants by in vitro adventitious embryogenesis and/or organogenesis. The concept of totipotency implies that every plant cell, with a normal complement of chromosomes, is able to express its genetic potential by regenerating a whole plant. However, as differentiation proceeds, tissues, initially competent to develop in a number of different patterns, reduce their degree of freedom, becoming eventually determined with respect to a single fate. Differentiated plant cells can also acquire new cell fates in the adult body, often independently of neighbouring cells. Structures as embryos, shoots, leaves and flowers may occur upon a leaf or leaf homologue in any position (epiphylly). The principal aim of this work is the improvement of our information on plant cell totipotency at molecular, cytological and biochemical level through the study of a plant systems featured by epiphylly: the clone EMB-2 of the interspecific hybrid Helianthus annuus x H. tuberosus characterized by the ectopic proliferation of adventitious morphogenetic structures on leaves. Notwithstanding the relative difficulty to identify the genes involved in the acquisition of cell totipotency, it is likely that a regulatory network, analogous to that involved in meristem and zygotic embryo development control the coordinate development of adventitious meristems and somatic embryos. Recent findings seem to support this point of view. In fact, the ectopic expression of genes involved in formation and maintenance of SAM (e.g. KNOTTED1-like homeobox, KNOX) or in embryo development (e.g. LEAFY COTYLEDON, LEC) is sufficient to re-programme in vivo the fate of differentiate cells. In the present work, the LEAFY COTYLEDON1-like gene of H. annuus (HaL1L) was isolated HaL1L expression patterns. HaL1L mRNA peaked at an early stage of embryogenesis (i.e. globular-, heart-, early cotyledon-embryo). More precisely, at the globular and heart stage HaL1L mRNA was prevalent in the outer cell layers of the embryo, in the suspensor and in integumentary tapetum cells. At cotyledon stage, HaL1L transcripts became evenly distributed throughout the embryo, but the positive signal was more evident in both inner cells of the integument and integumentary tapetum cells. In particular, a very strong signal was detected in the chalaza’s window. This result could support the hypothesis that HaL1L plays a role in the control of the nutrient transfer during embryo development. The class I KNOX gene of H. tuberosus, named HtKNOT1, was also isolated. HtKNOT1 was expressed in vegetative shoot apices and stem internodes of H. tuberosus, while leaves (blades and veins) and petioles did not accumulate any detectable HtKNOT1 transcripts. In particular, a higher accumulation of HtKNOT1 mRNA was detected in apical stem internodes in comparison to medial and basal region of the stem. HtKNOT1 transcripts were strongly detected in the meristematic dome of the SAM, but a weak presence of transcripts was also detected in incipient leaf primordia. These findings indicate that, in some species, the transcriptional down-regulation of class I KNOX genes is not a prerequisite for lateral organ initiation. In apical stem internodes of H. tuberosus, HtKNOT1 is expressed in cambial cells, phloem cells and xylematic parenchyma cells of the xylem. In basal internodes, HtKNOT1 expression was restricted to the presumptive initials and recently derived phloem cells. The expression of HtKNOT1 in the cambium is not surprising because the class I KNOX genes are involved in the meristem maintenance and the cambial cells behave as totipotent cells responsible for the stem radial growth. In this work we have also investigated the possible role of HtKNOT1 in controlling the development of the heterogamous inflorescence in the Helianthus genus. Inflorescence meristems of H. annuus and H. tuberosus showed HtKNOT1 expression in the region of the floret meristems and in developing organ primordia (i.e. floral bracts, petals, stamens and carpels). In older flowers strong expression of HtKNOT1 was seen in developing ovule. Notably, in the anthers of H. annuus, HtKNOT1 expression was also seen in pollen mother cells, in the tapetum and in the first and second mitotic division of developing pollen. This expression analysis also strongly suggests that the HtKNOT1 function is employed in pollen morphogenesis: this fact, if confirmed, will be one not yet discovered role of the class I KNOX genes. Several data support the view that when class I KNOX and/or LEC genes are not properly repressed adventitious phenomena of epiphylly became possible in different species. The EMB-2 somaclonal variant, derived by in vitro tissue culture of the tetraploid (2n = 4x = 68) interspecific hybrid Helianthus annuus x H. tuberosus (A-2 clone), shows an unusual pattern of development: it produces both in vitro and in vivo, epiphyllous embryos and shoot-like structures. More precisely, in addition to nonepiphyllous leaves (NEP) that expanded normally, some leaves of EMB-2 plants exhibit on the adaxial surface knobs and a prominent proliferation of ectopic structures (EP leaves), usually arranged in clusters along leaf veins. In EMB-2 plants, HtKNOT1 was highly expressed in both stem internodes and EP leaves. By contrast, HtKNOT1 mRNA accumulation was undetected in roots, petioles and NEP leaves. HtKNOT1 transcripts were confined at the level of the palisade cell layer in EP leaves. Thereafter, transcripts accumulated in developing morphogenetic structures, but the signal did not spread to epidermal cells. However, as development of shoot-like structures proceeded, transcript signal was also spread throughout epidermal cells. Afterwards, the level of Ht-KNOT1 transcripts decreased and mainly accumulated in the external cell layers of ectopic shoot-like structures. In contrast, in adventitious embryos, HtKNOT1 transcripts were confined to the basal portion of structures, whereas in the advanced globular embryos, the signal was restricted to a few scattered cellsHaL1L transcripts were undetected in both A-2 and NEP leaves while it was expressed in EP leaves. More precisely, the accumulation of HaL1L transcripts was evenly detected throughout all stages of somatic embryos. On the contrary, no signal related to HaL1L transcript accumulation was observed in developing shoot-like structures. Histological analyses suggested that epidermal and parenchyma cells around the vascular bundles contributed to the development of epiphyllous structures. Thus, the ectopic accumulation of HaL1L mRNA in this cell type opens the possibility that HaL1L is also involved in switching somatic cell fate towards embryogenic competence. Given the clear link of both cytokinins and auxin with plant organogenesis and embryogenesis, the endogenous levels of these hormones were detected in epiphyllous (EP) and non-epiphyllous (NEP) leaves of EMB-2 plants as well in the non-epiphyllous control genotype (A-2). The endogenous levels of cytokinins and IAA in NEP leaves of EMB-2 plants did not significantly differ in comparison to A-2 leaves. By contrast, the content of both hormones in EP leaves were 1.8-fold higher than NEP ones. The results suggested that epiphylly of EMB-2 plants was tightly related to a localized increase of hormones in EP leaves as compared to NEP ones. The involvement of cytokinins in ectopic morphogenetic events of EMB-2 plants was strengthened by immunolocalization results. The presence of active zeatin marked the development of adventitious shoot-like structures, from the first division of epidermal cells to SAM establishment. In EP leaves, a strong immunostaining was detected wherever the ectopic structures were formed, as well as at the level of the vascular bundles. In addition, a strong immunoreaction marked adaxial epidermal cells, which in some case undergo periclinal division, as well as clusters of epidermal cells, from which morphogenetic structures were likely to originate. As morphogenetic structures developed, a strong immunostaining was clearly detected. By contrast, in epiphyllous embryolike structures, zeatin signals were confined to a few scattered cells. Notably, before ectopic structures became evident, HtKNOT1 expression and zeatin signal were localized in distinct histological domains (mesophyll tissue vs. epidermis). However, as morphogenetic processes occurred, the accumulation of HtKNOT1 transcripts and zeatin clearly overlapped. A-2 leaves (i.e. non epiphyllous clone) displayed a strong accumulation of auxin in vascular bundles but not in the epidermis. Nevertheless, an immunostaining signal of indole-3-acetic acid (IAA) was evident in the mesophyll cells, preferentially in the chloroplasts. By contrast, EP leaves showed an auxin accumulation in a single cell or in a small group of cells of the adaxial epidermis. Indeed, early stages of ectopic embryogenic structures showed a strong immunostaining at the level of the whole globular embryo-like structure. By contrast, as for the zeatin immunolocalization, in advanced stages of embryo development a characteristic scattered immunoreaction was observed. In EP leaves with shoot-like structures, the auxin is preferentially accumulated in the protoderm and in active dividing cells. Interestingly, in the early stages of somatic embryo development, HaL1L expression and auxin were localized in the same histological domains (i.e. epidermis, parenchyma cells of vascular bundles, globular embryos). In conclusion, the epiphylly of EMB-2 plants seems linked to a complex regulatory network involving hormones and transcription factors. At physiological level the epiphylly of EMB-2 can be related to a localized increase of both zeatin and IAA while about the molecular mechanisms involved, HtKNOT1 and HaL1L cooperate during ectopic embryogenesis. On the contrary, in adventitious organogenesis only misexpression of HtKNOT1 has probably specific function. The histological analyses, the localization of auxin and cytokinin in specific histological domains and the expression patterns of HaL1L and HtKNOT1 also identified the epidermal cells located in proximity of vascular bundles as the totipotent cells of the EMB-2 clone. The localized increase of IAA in epidermal layers of EP leaves could determines an alteration in the intracellular polarity an therefore a shift of cell division plane (anticlinal vs. periclinal). Thereafter, citokinins, activated by HtKNOT1, could promote, by active divisions, the development of ectopic structures. Additionally, the two transcription factors could interact with other regulators of embryogenesis and/or organogenesis in the differentiation and organization of ectopic structures