Lessons from the vegetative shoot apex - Intercellular Communication in Plants - page 168

fate arising from a high concentration of the signal, and therefore closest to the
meristem.
An alternative model for ad-abaxial patterning of organs has recently been pro-
posed following the discovery of two short ( 22 bp) microRNA (miRNA165, 166)

complementary to RNA from the PHB and PHV loci and a third member of this
gene family, REVOLUTA ( REV ; Reinhart et al. , 2002; Rhoades et al. , 2002). This
model proposes that the miRNAs accumulate in the abaxial domain of the incip-
ient primordium where they promote the degradation of HD-ZIP RNA. Absence
of the miRNA from the adaxial domain allows PHB and PHV to promote adaxial
cell identity and REV to promote vascular development. This model does not ex-
clude the possibility that HD-ZIPs are activated by a meristem-derived adaxialising
signal. Experimental support for this model comes from the observation that PHV
RNA is targeted for miRNA-mediated degradation, and that RNA produced by the
PHV gain-of-function mutant allele is resistant to this degradation because of an
imperfect match with the miRNA (Tang et al. , 2003). The lack of degradation may
explain why PHV , and by inference PHB , persists in the abaxial domain of mutant
organs, where it could promote ectopic adaxial fate. Recent work has also shown
that REV is the likely target of miRNA degradation (Emery et al. , 2003).
Also consistent with this model is the recent finding that the Arabidopsis and
maize miRNA166 accumulates in the abaxial domain of developing organ primordia
(Juarez et al. , 2004; Kidner & Martienssen, 2004). This implies that the organ is
already patterned or that the miRNA itself acts as the polarising signal. There is some
support for the latter scenario, as the maize miRNA166 is first expressed below the
initiating organ primordium and subsequently accumulates, possibly by movement
through the phloem, in the abaxial domain of the growing organ (Juarez et al. , 2004).
Two gene families promote abaxial cell identities in developing organs, YABBY s
and KANADI s (reviewed by Bowman et al. , 2002). The KANADI family were first
identified as mutants that enhance the polarity defects of the crabs claw ( CRC ,
the founding member of the YABBY family) mutant and were later shown to encode
likely transcription factors with a GARP domain (Eshed et al. , 1999, 2001; Kerstetter
et al. , 2001). Mutant and ectopic expression studies suggest that KAN s and HD-ZIP s
function as mutual repressors in developing organs. For instance, kan1 kan2 mutants
have adaxialised organs with a broader domain of PHB expression, whereas ubiqui-
tous KAN expression in developing seedlings leads to abaxialisation of cotyledons
and a loss of the shoot apical meristem, phenotypes similar to plants triply mutant
for phb phv rev loss-of-function mutation (Eshed et al. , 2001; Kerstetter et al. ,
2001; Emery et al. , 2003). It is not clear whether KAN s repress HD-ZIPs directly,
or indirectly by promoting miRNA expression in the abaxial domain (Fig. 6.5).
As lateral growth of organs requires a juxtaposition of adaxial and abaxial
cell types, it is likely that the HD-ZIP s and KAN sregulate a signalling pathway
that directly controls cell proliferation at the junction between the different cell
types (see Fig. 6.5). Signalling from the adaxial domain of a developing organ is
also required to promote meristem formation and maintenance. The Antirrhinum
phantastica mutant initially has leaves with both adaxial and abaxial cell types,
butassuccessive leaves arise from the meristem there is an increasing loss of

∼