Patches, pegs and piggies


In his letter, Korn (2010) describes several instances of ectopic outgrowths of lamina on leaves and suggests how they could reveal a link between vascular tissue and dorsiventral polarity. Here, we suggest how the phenotypes he describes fit into a subset of the range of different ectopic outgrowths on leaves, and that the different types of outgrowth tell us different things about leaf development.

Ectopic outgrowths on leaves, as described in Korn (2010), are fascinating phenotypes and have been key in understanding the development of leaf dorsiventral polarity. Table 1 lists several examples: these can be divided into three groups. The first are cases of reversed polarity. The classical example of this is the phantastica (phan) mutant, whose interpretation was essential for understanding the relationship between polarity and outgrowth (Waites & Hudson, 1995). In these mutants, defects in the leaf polarity pathway lead to patches of abaxial cell fate on the adaxial side of the leaf. The boundary of these patches, where abaxial cell fate abuts adaxial cell fate, becomes specified as leaf margin. Outgrowth occurs along this new margin, leading to troughs, or paired flaps of leaf blade. The blade produced is fully polarized, with the polarity reflecting the patch from which it has arisen (Fig. 1).

Table 1.   Ectopic lamina phenotypes
NameReferenceSidePositionOutgrowth polar?Outgrowth formOutgrowth surrounds reverse polarity?
  1. D/V, dorsiventral.

Antirhinnum PhanWaites & Hudson (1995)AdaxialBetween veinsD/VPatchYes
Maize lax midribSchichnes et al. (1997)AbaxialOver veinsD/VPatchYes
Maize lbl1-refTimmermans et al. (1998)AdaxialOver veinsD/VPatchYes
Maize ig1-mumEvans (2007)AdaxialOver veinsD/VPatchYes
Maize Rld1-dJuarez et al. (2003)AbaxialOver veinsNo outgrowthPatch
Maize mwp1Candela et al. (2008)AbaxialBetween veinsD/VPatchYes
Tobacco ntPhanMcHale & Koning (2004)AdaxialOver veinsD/VPatchYes
Begonia‘Cathedral’Korn (2010)AbaxialBetween veinsD/VPatchYes
Xanthosoma atrovirens‘Appendiculatum’Korn (2010)AbaxialOver veinsD/VPatchYes
Xanthosoma atrovirens‘Variegatum Monstrosum’Korn (2010)AdaxialOver veinsD/VPatchYes
Xanthosoma jacquinii‘Linearum’Korn (2010)AdaxialBetween veinsLittle outgrowthPatchYes
Arabidopsis 35S:AS2Lin et al. (2003)AbaxialAnywhereAbaxializedPointNo
Arabidopsis ett arf3Pekker et al. (2005)AbaxialAnywhereAbaxializedPointNo
Arabidopsis kan1 kan2Eshed et al. (2001, 2004)AbaxialAnywhereAbaxializedPointNo
Arabidopsis PiggiesPinon et al. (2008)AdaxialOver veinsD/VPointNo
Begonia hispida var. cuculliferaSattler & Maier (1979)AdaxialOver veinsD/VPointNo
Begonia luxuriansC. A. Kidner (unpublished)AdaxialOver veinsD/VPointNo
Begonia F1 hybridsC. A. Kidner (unpublished)AdaxialOver veinsD/VPointNo
HelianthusFambrini et al. (2000)AdaxialOver veinsD/VPointNo
Figure 1.

 The generation of ectopic lamina from reversal of polarity. The boundary between adaxial and abaxial cell types promotes outgrowth of the lamina along the leaf primordia. Unstable dorsiventral polarity can lead to patches of reversed polarity or ectopic abaxial cell identity (a). Outgrowth along all the boundaries between adaxial and abaxial cell types results in ectopic lamina (b). Cross-sections through the leaf show that ectopic lamina surround patches of reversed polarity (c).

In phan, Begonia‘Cathedral’ and Xanthosoma jacquinii‘Linearum’ the troughs occur between main veins, but in the other cases patches of reversed polarity are associated with veins. In Lax midrib l-O, Schichnes & Freeling (1998) suggest that the outgrowth is associated with heterochronic effects, modifying how the cells respond to a signal from the vein. They hypothesize that during early leaf development lamina outgrowth is promoted by a signal from the vein. In these mutants, cells remain competent to respond to this signal for longer and therefore form ectopic blades over main veins (Schichnes & Freeling, 1998). In some cases ectopic outgrowth may be associated with ectopic expression of meristem-specific genes. Some class I KNOX genes are ectopically expressed in the leaf in Ig mutants in maize and tobacco plants with downregulation of NtPhan, both of which have ectopic lamina (McHale & Koning, 2004; Evans, 2007). This could be construed as heterochronic reversion to an earlier developmental stage, but a direct link between ectopic class I KNOX genes and leaf outgrowths has not been made and the cloning of several other reverse-polarity mutants has revealed that many are dorsiventral polarity genes rather than clear regulators of developmental timing.

Korn (2010) suggests that the vasculature could be the source of patterning for dorsiventrality in the leaf. This is unlikely to be the case in early development as initial dorsiventrality appears to derive from gene-expression domains in the peripheral zone of the meristem, which are then reinforced by a meristem-derived signal through the L1 (Husbands et al., 2009). Surgical experiments show that leaves isolated from the meristem (but not from the stem vasculature) do not develop dorsiventral polarity (Sussex, 1955; Reinhardt et al., 2005). The dorsivental polarity of the vasculature could explain why some of the reversions occur over veins or flank them. Both leaf and vasculature dorisventral patterning involves KANADI and HD-ZIPIII genes (Emery et al., 2003). These genes are expressed for longer in the veins than they are in the rest of the leaf. In the vasculature they regulate polarity and cambium activity via auxin flow (Ilegems et al., 2010). The expression of these polarity pathway genes in the vasculature could induce polarity reversals in adjacent competent ground tissue. However, vascular signals are not an absolute requirement for polarity reversals, as shown by the cases where polarity-reversed patches occur independently of veins (phan and Begonia‘Cathedral’).

A second class of dorsiventral polarity mutation leads to the production of peg-like outgrowths lacking dorsiventral polarity. These are seen in the kan1 kan2 double mutant, in 35S:AS2 and in ett arf4 double mutants (Sessions et al., 1997;Eshed et al., 2001, 2004; Lin et al., 2003; Pekker et al., 2005). These ectopic outgrowths may be related to how KANADI affects auxin flow (Izhaki & Bowman, 2007). Ectopic auxin maxima could result in ectopic outgrowth without a reversal of polarity. Alternatively, very small reversals of polarity could result in a peg-like outgrowth rather than a trough. This interpretation is supported by the marginal cell identity of the whole outgrowth in kan1 kan2 double mutants (Eshed et al., 2004). milkweed pod1 (mwp1) is a mutant in a maize KANADI gene required for outgrowth at the margins in early leaf development and for maintenance of abaxial cell fate during later leaf development (Candela et al., 2008). mwp1 plants have patches of reversed polarity with dorsiventral outgrowths, rather than the pegs seen in Arabidopsis loss-of-KANADI function. The different phenotypes of the outgrowths may be related to cell-division patterns during later leaf development, which affects the size of the patch of polarity reversal.

The third set of phenotypes are typified by those of piggyback mutants in Arabidopsis and Begonia hispida var. cucullifera. Individual fully polarized leaves are formed from the adaxial side of the leaf. These ‘piggies’ are fully polarized leaves with midveins and margins. Their polarity reflects the proximal–distal polarity of the parent leaf. Outgrowth is from a true ectopic leaf primordium, formed without any underlying meristem. The primordia appear to self-organize from subepidermal cell divisions (Maier & Sattler, 1976; Sattler & Maier, 1977; Pinon et al., 2008). In Begonia hispida var. cucullifera the leaf primordia form a continuum with multicellular trichomes and a similar phenotype is seen in the ectopic leaves produced on the disc of insertion of the compound leaved Begonia luxurians (J. Cavers & C. A. Kidner, unpublished). Begonia urophylla has multicellular trichomes at the petiole attachment point and occasional cell divisions in the subepidermal hypodermal layer. In F1 hybrids with either Begonia cardiocarpa or Begonia nelumbiifolia these trichomes and subepidermal divisions are replaced with ectopic leaves, suggesting that the changed developmental context in the hybrids is sufficient to confer leaf primordium identity on these dividing cells (D. Wrigley & C. A. Kidner, unpublished).

These ‘piggies’ are always adaxial, proximal and over veins. The adaxial side of the leaf is closer to the meristem in space and in gene-expression patterns. Cell divisions continue for longer on this side, as they do in proximal regions of the leaf. A stronger case can be made for this type of ectopic leaf to be associated with heterochronic effects. The meristematic competency in the vascular cambium of the main veins may expand to include the still-dividing adaxial proximal cells late in leaf development. This gives the leaf cells the ability to respond to the positional information around them, conferring leaf identity in a way suitable to a much earlier stage in leaf development (Fig. 2).

Figure 2.

 Signals that could result in the production of an ectopic leaf primordia. Cell division ceases in a gradient along the leaf from distal to proximal and from adaxial to abaxial. Cell-division competency is represented by shades of red, from dark (competent for further division) to white (no longer competent for cell division). Cell divisions continue longest in the vascular tissues (a). Some cells at the proximal, adaxial region of the leaf retain competency for cell division. This could be promoted by signals from the vascular cambium (red). Positional signals in the leaf confer leaf rather than meristem cell identity on the dividing cells (black). This results in the organization of leaf primordium (b). The leaf primordium generates a normal leaf that takes its polarity from the parent leaf (c).

Ectopic outgrowths on leaves are relatively late events, but they can be informative about the processes occurring during earlier leaf development. The three different types of ectopic outgrowths on leaves are each telling us different things. The patches were key in understanding that the boundary between adaxial and abaxial cell fates leads to lamina outgrowth. The pegs suggest a role for auxin transport and auxin response regulation in locating outgrowth position. The piggies may be able to help us understand what signals define a group of dividing cells as a leaf primordium. Although these phenotypes may seem just curiosities, it is curiosity that leads us to new understanding.


Work in C.K.’s laboratory is funded by the BBSRC and the M. L. MacIntyre Begonia Research Trust. We would like to acknowledge Mary Byrne for useful discussions on these phenotypes.