The nature of the developmental signals required for early embryo patterning have long been a source of speculation to plant scientists. When considering protodermal specification, it is first interesting to ask when cells first acquire protodermal characteristics. There is a certain amount of evidence to suggest that protodermal identity is a default during the development of the embryoproper and that in at least some species cells have protodermal identity well before a defined dermatogen/protoderm is set aside by restrictions in cell division planes. Bruck and Walker, in their study of embryogenesis in Citrus jambhiri, conclude that even at the very early globular stage, (well before protoderm demarcation) a cuticle has been secreted on the outside of the developing embryo (Bruck 1985a). Cuticle secretion is considered to be a strictly epidermal function, and this suggests that cells at the surface possess epidermal characteristics which are then presumably lost from the more internal progeny of a periclinal division. Interestingly, this hypothesis is borne out at the gene expression level in Arabidopsis. At least two protodermal markers (ATML1 and ACR4) are expressed in all cells of the embryo proper at the eight cell stage (when all eight cells are in contact with the periphery), and their expression is then lost in "internal" cells subsequent to the periclinal divisions which give rise to the dermatogen embryo (Lu et al. 1996; Gifford et al. 2003). It is fair then, to suggest that the signals required for protoderm development are present from very early on in embryogenesis.
The "loss" of protodermal characteristics in embryonic cells which loose contact with the periphery, is very reminiscent of the process of aleurone proliferation where most starch endosperm (internal) cells are derived by periclinal divisions from cells on the endosperm periphery, which will eventually differentiate as aleurone. These similarities are backed up by gene expression patterns which are common to the two tissues. For example some of the Outside Cell Layer (OCL) genes from maize, which are homologous to the HDZipIV class of transcription factors to which ATML1 and PDF2 belong, are expressed in both the developing embryonic protoderm and the incipient aleurone layer during caryopsis development (Ingram et al. 2000). Although the precise function of these genes in maize has not yet been ascertained, they do appear to play a role in the growth of the developing caryopsis (Khaled et al. 2005), and, extrapolating from the role of their Arabidopsis homologues in protoderm development it is possible that aleurone development is also specified by these genes. The agronomic importance of the cereal endosperm, coupled to the characterisation of a wide range of easily visualised aleurone specific pigment markers, has meant that many loci involved in the specification of the aleurone layer have been extensively characterised by maize geneticists over the past decades. Several of these genes appear to play parallel roles in the specification of embryonic protoderm, including DEK1, which is required for aleurone specification and normal embryogenesis in maize (Be-craft et al. 2002). As mentioned previously, the Arabidopsis DEK1 orthologue AtDEK1 also appears to be necessary for embryonic protoderm specification and endosperm development (Johnson et al. 2005; Lid et al. 2005). DEK1 encodes putative membrane localised calpain protease, suggesting that it acts by cleaving other protein molecules (Lid et al. 2002). However, to date, no confirmed targets of this cleavage activity have been identified. Additionally, the non cell-layer specific expression pattern and severe phenotypes associated with DEK1 and orthologues in Arabidopsis and tobacco (Ahn et al. 2004), has led to the suggestion that it may play a general role in plant cell differentiation rather than a specific role in endosperm development. The maize Extra cell layers (Xcl) mutant, previously discussed for its defects in the regulation of epidermal cell division planes in leaves, also shows an analogous increase in aleurone cell layer number (Kessler et al. 2002). Similarly the disorganised aleurone layer 1 and 2 mutants of maize both show disorganised epidermal cell arrangement in rescued homozygous plants (Lid et al. 2004). Finally, the maize supernumerary aleurone layer 1 mutant also affects embryo development, although in this case the embryonic defects have not been characterised in depth (Shen et al. 2003).
Another maize gene, CRINKLY4 (CR4), has also been implicated both in aleurone cell fate specification and the development of the embryo, with weak alleles showing defects in leaf epidermal fate (Becraft et al. 1996, 2001). CR4 encodes an RLK and again Arabidopsis homologues, including the suspected orthologue ACR4, have been identified (Tanaka et al. 2002; Gifford et al. 2003; Watanabe et al. 2004; Cao et al. 2005). Unlike CR4, ACR4 expression is restricted to the L1 cells layer of the embryo and all shoot meristems and organ primordia. The null acr4 mutant phenotype and the sub cellular localisation of ACR4 protein are consistent with a role in signalling between L1 cells during development, and/or between L1 cells and surrounding tissue (Gifford et al. 2003; Watanabe et al. 2004). The acr4 mutant phenotype is restricted to ovule integuments and sepal margins, with slight defects in leaf surface formation reported. Although no ligands for CR4 or ACR4 have been identified, it has been postulated that the ABNORMAL LEAF SHAPE1 (ALE1) encoding a subtilisin-like serine protease is involved in generating an ACR4 ligand from a peptide precursor, due to the genetic interaction between acr4 and ale1 mutant alleles (Watanabe et al. 2004). If they are not protected from desiccation, mutants in ALE1 die after germination due to aberrant cuticle formation on the seedling cotyledons and early leaves. ALE1 is expressed in the Embryo-Surrounding Region (ESR) of the endosperm during embryogenesis, and may also be expressed weakly in the very young embryo. Double mutants between ALE1 and ACR4 show a dramatically exacerbated embryo/seedling pheno-type. Recent results have shown that ACR4 may act with a second unrelated RLK, ALE2, during the development of the embryonic epidermis (Tanaka et al. 2007).
The analysis of ALE1 has fuelled the ongoing debate regarding whether the correct specification and differentiation of the embryonic protoderm is dependent on signals from the surrounding endosperm. Interestingly, a parallel debate exists regarding the role of maternal tissues in the differentiation of the endosperm aleurone layer in cereals. As described previously, considerable evidence that "outside" or "surface" position is critical both for protoderm and for aleurone cell specification has been published. As a result it is natural that the finger should be pointed at surrounding tissues (the ESR for the protoderm, and the maternal nucellus for the aleurone) as a source of developmental signals. Indeed studies in maize have shown that several genes including small secreted peptides, including those encoding the secreted ESR-proteins (Opsahl-Ferstad et al. 1997; Bonello et al. 2000, 2002), are expressed in the embryo surrounding region of maize. ESR proteins belong to the CLE (CLV3/ESR) family of secreted peptides which includes CLV3 a putative ligand of the CLAVATA1 RLK (Cock and McCormick 2001; Sharma et al. 2003). Moreover, considerable evidence now exists that these peptides could be cleaved as part of their activation, potentially by secreted proteases such as that encoded by ALE1 (Ito et al. 2006; Kondo et al. 2006; Ni and Clark 2006). However, debate still exists as to whether embryogenesis (either zygotic or somatic) can occur in plants in the absence of endosperm-like cells. In many studies in vitro, embryogenesis appears to depend upon the presence of non-embryogenic cells with some endosperm-like characteristics (Magnard et al. 2000, and references therein). However, it is difficult to distinguish between the requirements for initiation of embryogenesis, and those for cell-fate specification in many of these systems. A recent publication addressed the question of whether aleurone specification can occur in the absence of maternal nucel-lus tissue, and concluded that it could, although sucrose concentrations used in the culture systems were abnormally high (Gruis et al. 2006). Interestingly, recent research in Arabidopsis has shown the probable presence of specific sucrose transporters in both the ESR (Baud et al. 2005), and the maternal en-
dothelium (the inner-most maternal cell layer which directly juxtaposes the developing aleurone) (Lauterbach 2007).
If positional signals are not derived from the tissue surrounding the developing protoderm and aleurone, then two further possibilities exist. The first is that a signal present and immobilized within the zygote, egg or central cell, possibly in the cell wall or membrane, is perceived by cells at the embryo/endosperm periphery. Such signals are known to play roles in the development of algae, such as Fucus (Kropf et al. 1988; Quatrano et al. 1991; Shaw and Quatrano 1996). The second possibility is that embryo and/or endosperm cells have a mechanism for sensing the presence of their neighbours, and those cells which sense that they lack neighbours on one face, respond by expressing protoderm/aleurone specific genes. For the moment neither of these possibilities can be either proven or disproved.
Was this article helpful?