232
S. Gleissberg et al.
the closely approached marginal ridges. These subunifacial structures with a narrow adaxial furrow could be converted into unifacial ones by a congenitallongitudinal fusion of the contacting marginal ridges. Although this process has not been
documented developmentally, the morphology of some mature leaves suggests this.
Leaves in Oxypolis filiformis (Fig. 3) produce a single median row of pinnae indicating the presence of a single adaxial marginal blastozone (Fig. 3) (Kaplan 1970a;
Eberwein 1995). Kaplan (1980) suggested that the adaxial ridge of the Acacia
phyllodes corresponds to longitudinally fused margins as they occur in bifacial,
leaflet-producing blade segments. Acacia phyllodes correspond developmentally to
sword-like leaves known from many monocotyledons (Iris, Acorus), where adaxial
meristem activity is more extensive than in cylindrical unifacial leaves, leading to
medially flattened leaves (Troll 1939; Kaplan 1975). However, the adaxial ridge of
sword leaves is generally not homologized with the leaf margins of transversely
flattened leaves.
Yet another model for the developmental origin of abaxialized leaf segments
was proposed by Goebel (1905) and later elaborated by Thielke (1948) and Roth
(1949). According to these authors in, for example, Iris, a secondary leaf apex is
formed on the abaxial side of the primordium early in ontogeny; this subsequently
forms the new leaf axis, while the activity of the primary apex ceases. As a result of
this "sympodial" mode of development, the leaf portion formed by the secondary
apex is entirely surrounded by abaxial tissue. Adaxial tissue remains confined to a
sector below the primary apex, which develops into the ligule. This model was
criticized initially by Troll and Meyer (1955) because the model implied that unifacial
segments arise postgenitally from bifacial primordia. Kaplan's (1970b) studies indicated that the supposed secondary apex was in fact the primary apex that becomes
dislocated abaxially by early onset of adaxial meristem activity.
The phenotype of the phantastica mutant inAntirrhinum led Waites and Hudson
(1995) to propose a new model for transverse leaf asymmetry. In these mutants, leaf
margins do not differentiate in the absence of adaxial surface identity. Completely
abaxialized leaves formed at higher nodes in phantastica plants are cylindrical in
shape and lack a margin. Early formed leaves are bifacial, but show patches of
abaxial tissue on the upper surface that are surrounded by ectopic marginal ridges.
In leaves showing abaxialization only proximally, this region is separated from the
bifacial distal part by an adaxially fused margin. This is similar to cross zones
developed at the border of unifacial and bifacial leaves and is also seen in lbll
mutants of maize. These data suggest that the determination of margins in the
primordium requires a border of adaxial and abaxial domains and that marginal
identity of bifacialleaves cannot be established in the absence of adaxial identity. It
suggests, furthermore, that unifacial, abaxialized leaf sectors can form by a localized and controlled downregulation of a factor conferring adaxial identity. This
could be correlated with enhanced thickening growth, the dedifferentiation (loss)
of the developing margin, and the subsequent fusion of the remaining marginal
ridges at the new borders of adaxial and abaxial identity, forming cross zones.
S. Gleissberg et al.
the closely approached marginal ridges. These subunifacial structures with a narrow adaxial furrow could be converted into unifacial ones by a congenitallongitudinal fusion of the contacting marginal ridges. Although this process has not been
documented developmentally, the morphology of some mature leaves suggests this.
Leaves in Oxypolis filiformis (Fig. 3) produce a single median row of pinnae indicating the presence of a single adaxial marginal blastozone (Fig. 3) (Kaplan 1970a;
Eberwein 1995). Kaplan (1980) suggested that the adaxial ridge of the Acacia
phyllodes corresponds to longitudinally fused margins as they occur in bifacial,
leaflet-producing blade segments. Acacia phyllodes correspond developmentally to
sword-like leaves known from many monocotyledons (Iris, Acorus), where adaxial
meristem activity is more extensive than in cylindrical unifacial leaves, leading to
medially flattened leaves (Troll 1939; Kaplan 1975). However, the adaxial ridge of
sword leaves is generally not homologized with the leaf margins of transversely
flattened leaves.
Yet another model for the developmental origin of abaxialized leaf segments
was proposed by Goebel (1905) and later elaborated by Thielke (1948) and Roth
(1949). According to these authors in, for example, Iris, a secondary leaf apex is
formed on the abaxial side of the primordium early in ontogeny; this subsequently
forms the new leaf axis, while the activity of the primary apex ceases. As a result of
this "sympodial" mode of development, the leaf portion formed by the secondary
apex is entirely surrounded by abaxial tissue. Adaxial tissue remains confined to a
sector below the primary apex, which develops into the ligule. This model was
criticized initially by Troll and Meyer (1955) because the model implied that unifacial
segments arise postgenitally from bifacial primordia. Kaplan's (1970b) studies indicated that the supposed secondary apex was in fact the primary apex that becomes
dislocated abaxially by early onset of adaxial meristem activity.
The phenotype of the phantastica mutant inAntirrhinum led Waites and Hudson
(1995) to propose a new model for transverse leaf asymmetry. In these mutants, leaf
margins do not differentiate in the absence of adaxial surface identity. Completely
abaxialized leaves formed at higher nodes in phantastica plants are cylindrical in
shape and lack a margin. Early formed leaves are bifacial, but show patches of
abaxial tissue on the upper surface that are surrounded by ectopic marginal ridges.
In leaves showing abaxialization only proximally, this region is separated from the
bifacial distal part by an adaxially fused margin. This is similar to cross zones
developed at the border of unifacial and bifacial leaves and is also seen in lbll
mutants of maize. These data suggest that the determination of margins in the
primordium requires a border of adaxial and abaxial domains and that marginal
identity of bifacialleaves cannot be established in the absence of adaxial identity. It
suggests, furthermore, that unifacial, abaxialized leaf sectors can form by a localized and controlled downregulation of a factor conferring adaxial identity. This
could be correlated with enhanced thickening growth, the dedifferentiation (loss)
of the developing margin, and the subsequent fusion of the remaining marginal
ridges at the new borders of adaxial and abaxial identity, forming cross zones.