As described above, the use of periclinal mosaics has been invaluable in ascertaining the contributions of various cell-layers to organ development. However, they have also been used in addressing the potential roles of specific cell layers/tissue types in the control of organ shape (reviewed in Szymkowiak and Sussex 1996; Marcotrigiano 2001). The generation and study of pericli-nal mosaics (either full periclinal chimeras or clones induced during development), where cell layers carry mutations in developmental genes or are derived from different species, has been an important approach. The use of periclinal chimeras in such experiments has the associated disadvantage that certain chimeric combinations can affect the contributions of different cell layers to organs, giving a false impression of normal development. An alternative has been to analyse clones, usually marked by irradiating plants heterozygous for a recessive developmental marker at a known point in development. Because this provides unequivocally clonal groups of marked cells (mericlinal chimeras) from progenitors at defined developmental stages it can give more accurate information about cellular contributions, and has been used for multiple studies.
Relatively large numbers of both types of study have not provided a consensus as to what extent a given cell layer controls organ shape and growth, as the results vary between organs, species, and according to the developmental marker studied. Interestingly however, several studies on leaves and other organs with an extended lamina (such as sepals and petals) using plants chimeric at just one locus (rather than intraspecific combinations with significant degrees of genetic divergence), have shown that the presence of a wild-type allele in the L1 cell layer can play a greater role in restoring wildtype phenotypes than expression in underlying cell-layers (Hantke et al. 1995; Perbal et al. 1996; Vincent et al. 2003). However, this is not always the case as some genes appear to be required in underlying cell layers for normal development, including the Nicotiana LAM1 gene (McHale and Marcotrigiano 1998), the OKRA gene from cotton (Dolan 1998) and the LIGULELESS-1(LG1) gene from maize (Becraft et al. 1990).
How can the differences in behaviour of different developmental genes be explained? It may be that in organs with extended laminas, the epidermal cell-layer plays a significant role in expressing the information which determines final organ size and shape, signalling to underlying cell-layers to control their development. Developmental genes such as LAM1 or LG1, which are needed in underlying cell-layers for normal development, may be involved in these communication processes. Indeed, some indication that LG1 may be required to perpetuate positional signalling across the leaf blade has been gained from analysing the developmental effects of lg1 mutant sectors (Becraft and Free-ling 1991).
Another possible explanation for the observed differences lies in the ability of adjacent cell-layers to accommodate discrepancies in growth, and ultimately co-ordinate their development. The molecular mechanisms underlying this accommodation probably include the symplastic movement of developmentally important molecules, such as transcription factors and regulators of the cell-cycle, between cell-layers in the meristem and in young organ primordia via plasmodesmatal connections (reviewed in Lucas and Lee 2004). The products of both the different developmental genes used in clonal studies, and downstream effectors of the markers, may move from layer-to-layer to different extents during development. Such differential capacity for movement has been observed for several proteins including the transcription factors LEAFY and APETALA1 in Arabidopsis (Sessions et al. 2000; Wu et al. 2003).
Additionally, as pointed out by Vincent et al., chimeras are affected not only in the localisation of developmental gene expression, but also in the absolute amounts of their product (Vincent et al. 2003). Perception of protein concentrations may be more important for the actions of some developmen-tally important genes than for others.
Interestingly, in organs lacking laminas, such as stamens and ovaries, the importance of the L1 appears more often to be overridden by that of the L2
and L3 (Szymkowiak and Sussex 1992; Vincent et al. 2003). One explanation for this is that the L1 cell layer may contribute proportionally more cells to the total volume of an organ with an extended lamina than to an effectively cylindrical organ such as a stamen. Protein dosage could therefore also be an important factor in the perceived differences in cell-layer contribution to organ development in mosaic laminar and non-laminar organs.
Was this article helpful?