Biology Reference
In-Depth Information
that while, in the same inflorescence, most of the flowers did not set or set very
few seeds, some were able to develop normal siliques filled up with seeds (Fig-
ure 3C and 3D). Cross pollination showed the pollens produced by myb21-t1
myb24-t1 myb57-t1 triple mutant plants were partial viable (data not shown),
suggesting the short stamens in the triple mutants are the main cause of the steril-
ity. It is possible that environmental factors (e.g. temperature) may influence male
fertility in these mutants, an observation also reported for MYB33 and MYB65
[24]. Therefore, MYB21, MYB24 and MYB57 function redundantly to control
the stamen filament development in the late developed flowers.
Since MYB21 and MYB24 are also expressed in sepals and petals, we exam-
ined the sepal and petal development in the single, double and triple myb mu-
tants. As shown in Figure 3D, sepal development appeared normal in all mutants
whereas petal development varied in different mutants. Petals in the myb24-t1
and myb57-t1 two single mutants grew to a final length longer than the pistils, as
that did the WT petals. Petals in the myb21-t1 single, myb24-t1 myb57-t1 and
myb21-t1 myb57-t1 two double mutants grew to a final height parallel to the
pistil (Figure 3D). Petals in the myb21-t1 myb24-t1 double mutant grew just out
of the sepals but ended at a lower level than the stigma (Figure 3D). The growth
of petals in the myb21-t1 myb24-t1 myb57-t1 triple mutant was arrested and the
petals never grew out of the sepals (Figure 3D).
myb21-t1 myb24-t1 Is Epistatic to gai-t6 rga-t2 rgl1-1 rgl2-1 in
Controlling Stamen Filament Elongation
MYB21 and MYB24 were repressed in ga1-3 but their expressions were restored
to the WT level in the ga1-3 gai-t6 rga-t2 rgl1-1 rgl2-1 penta mutant, suggest-
ing that GA regulates MYB21 and MYB24 through inactivating DELLA pro-
teins. Application of exogenous GA could not rescue the stamen development in
myb21 myb24 mutant (data not shown), suggesting that MYB21 and MYB24 are
needed in GA-mediated stamen development. To further confirm this hypothesis,
we crossed myb21 myb24 with ga1-3 gai-t6 rga-t2 rgl1-1 rgl2-1 to generate two
hexa mutants (hexa1: ga1-3 gai-t6 rga-t2 rgl1-1 rgl2-1 myb21-t1; hexa2: ga1-3
gai-t6 rga-t2 rgl1-1 rgl2-1 myb24-t1) and one hepta mutant (ga1-3 gai-t6 rga-t2
rgl1-1 rgl2-1 myb21-t1 myb24-t1). The two hexa mutants overall appeared simi-
lar to each other and had wildtype-like stamens and were largely fertile (Figure
4A). The hepta mutant plant displayed no difference from the penta mutant plant
in its vegetative growth. However, its mature flowers showed a short filament
phenotype identical to that in the myb21-t1 myb24-t1 double mutant (Figure
4A). This observation demonstrated that myb21-t1 myb24-t1 double mutations
are epistatic to DELLA mutations. SEM analysis showed that the short stamen
