Different plant organs have characteristic sizes and shapes, which are the outcomes of specific growth patterns. For example, cell proliferation in developing leaves of dicotyledonous plants stops first at the leaf tip and sequentially further towards the leaf base, whereas in petals cells continue to divide for the longest period in the central and distal regions that give rise to the petal blade (Dinneny et al. 2004; Disch et al. 2006; Donnelly et al. 1999). Ultimately, these different growth patterns are controlled by the transcriptional networks that determine organ identity (Krizek and Fletcher 2005), but how these modify growth control remains largely unknown. In principle, they could modulate the activity of common growth control pathways, as many of the factors discussed above affect both leaves and floral organs, albeit to different extents or even in opposite directions.
With respect to petals, some initial progress has been achieved in understanding the link between organ identity and growth control. One downstream target that is activated by the B-class homeotic proteins APETALA3 (AP3) and PISTILLATA (PI) in stamens and in petals at the beginning of cell expansion is NAP, a homolog of the petunia NO APICAL MERISTEM gene (Sablowski and Meyerowitz 1998). Overexpression of NAP leads to a failure of petal cells to enlarge properly, suggesting that temporally regulated NAP expression is important for normal petal cell expansion. Another downstream target of AP3 and PI, the petal-specific splice form of BPE, prevents excess cell enlargement in petals (Szecsi et al. 2006). Thus, the combined activities of NAP and the petal-specific form of BPE appear to mediate between organ identity factors and patterns of post-mitotic growth.
In contrast to the radially symmetric flowers of Arabidopsis, Antirrhinum forms zygomorphic (i.e., bilaterally symmetric) flowers with dorsal, lateral and ventral petals of distinct sizes and shapes (Schwarz-Sommer et al. 2003). The class II TCP proteins CYCLOIDEA (CYC) and DICHOTOMA (DICH) are required for dorsal petal identity in a partially redundant manner, and double mutants form radially symmetric flowers (Luo et al. 1996, 1999). Also, in cyc mutants, growth of the dorsal stamen, which normally does not develop, is derepressed, correlating with upregulated expression of cell cycle genes (Gaudin et al. 2000). Overexpression of CYC in Arabidopsis inhibited cell proliferation and expansion in leaves, but promoted late-stage cell expansion in petals (Costa et al. 2005). Thus, similar to CIN (see above) and in light of the findings on TCP20 (Li et al. 2005), these factors may have a rather direct influence on growth. CYC and DICH activate expression of the small MYB-domain protein RADIALIS in the dorsal region of the flower (Corley et al. 2005), but how this modifies growth patterns in developing organs remains unknown.
The widely accepted model of Waites and Hudson for lateral organ growth, which was derived from the analysis of Antirrhinum mutants in the myb-domain protein PHANTASTICA (PHAN) (Waites et al. 1998; see below), states that outgrowth of the lamina requires the juxtaposition of adaxial (towards the shoot, usually the upper side) and abaxial (away from the shoot, usually the lower side) identities (Bowman et al. 2002; Canales et al. 2005; Waites and Hudson 1995). Loss of either adaxial or abaxial identity causes the formation of radialized organs without a lamina, whereas ectopic boundaries between cells with ad- and abaxial identities can induce ectopic lamina outgrowth. How the ad/abaxial boundary induces lamina outgrowth is largely unknown. In addition, recent work has uncovered further complexity in the relation between polarity control and growth regulation, as a number of genes that were originally identified because of growth defects have now been shown to also interact with the pathway determining organ polarity. For example, the growth stimulator ANT was found to influence organ polarity, acting together with the abaxializing YABBY proteins FILAMENTOUS FLOWER (FIL) and
YABBY3 (YAB3) (Nole-Wilson and Krizek 2006). Triple mutants of antfilyab3 show very severe polarity defects that correlate with strongly reduced expression of the homeodomain zipper (HD-ZIP) protein PHABULOSA (PHB), normally expressed on the adaxial side. In addition, JAG was shown to pattern fruit tissues in cooperation with FIL and YAB3, and NUB is only expressed on the adaxial side of developing organs (Dinneny et al. 2005, 2006). Although the exact mechanistic links are far from clear, these intriguing findings may open the door to detailed studies of the connections between the control of growth and organ polarity.
The putative transcriptional co-repressors LUG and SEUSS (SEU), originally described as repressors of the floral homeotic gene AGAMOUS (AG) in outer flower whorls, also appear to function in the regulation of floral organ cell proliferation and ad/abaxial patterning, independently of their influence on organ identity (Franks et al. 2002, 2006; Sridhar et al. 2004). Petals of ag lug seu triple mutants are strongly reduced in size because of insufficient cell numbers and show defects in ad/abaxial patterning. These are correlated with reduced expression of major regulators of ad/abaxial patterning, PHB and FIL. Although it is not clear from this study whether LUG and SEU act independently on proliferation and on patterning or whether the growth defects result from the polarity defects, the changes in leaf shape and size without apparent polarity defects in the rotunda2 alleles of LUG argue against the latter alternative (Cnops et al. 2004). Thus, ANT, LUG and SEU may both stimulate growth directly and also indirectly by maintaining a robust juxtaposition of the gene expression domains that determine organ polarity.
Patterning of different tissues or domains within organs often involves differential growth. An example is provided by the subdivision of Arabidopsis rosette leaves into a bladeless petiole and the distal lamina region. Outgrowth of lamina tissue from the leaf petiole is suppressed by the redundant BLADE-ON-PETIOLE1 (BOP1) and BOP2 genes, coding for BTB/POZ- and ankyrin-domain proteins that appear to regulate transcription by interacting with additional factors (Ha et al. 2003, 2004; Hepworth et al. 2005; Nor-berg et al. 2005). Similar to bop loss-of-function mutants, JAG overexpression causes lamina outgrowth from the petiole. Indeed, bop mutants show ectopic expression of JAG and its homolog NUB in the petiole region, suggesting that exclusion of these growth stimulators from the prospective petiole by BOP activity contributes to the petiole's distinct growth pattern.
Lamina outgrowth from the petiole is also suppressed by the ASYMMETRIC LEAVES1 (AS1) and AS2 genes. AS1 encodes a myb-domain protein, while AS2 codes for a protein with a leucin-zipper, which can interact with AS1 (Byrne et al. 2000; Iwakawa et al. 2002). Like mutants in the maize AS1 ortholog ROUGH SHEATH2 (RS2) and presumably in the Antirrhinum ortho-log PHANTASTICA (PHAN), Arabidopsis as1 and as2 mutants show ectopic expression of class I Knotted1-like homeobox (KNOX) genes in developing leaves (reviewed in Kessler and Sinha 2004). This defect is also found in bop1
mutants (Ha et al. 2003). Class I KNOX genes act to maintain meristem cells in an undifferentiated state, and their ectopic expression leads to lobed leaves because of prolonged cell proliferation (Scofield and Murray 2006). In addition to repressing KNOX genes in leaves, AS1/PHAN/RS2 also contribute to ad/abaxial polarity establishment by promoting adaxial cell fate and thus stimulate lamina outgrowth, as postulated by the above model.
Thus, there are interconnected pathways of factors that determine regional and organ identity, e.g., in the case of KNOX gene repression by the BOP and AS genes, leading to different growth patterns that account for the characteristic organ sizes and shapes. Yet, as before, more work will be required to unravel the molecular mechanisms by which these regulators modulate cellular growth and division patterns.
Was this article helpful?