The genetic basis for differences in leaf form
between Arabidopsis thaliana and its wild
relative Cardamine hirsuta
Angela Hay & Miltos Tsiantis
A key question in biology is how differences in gene function
or regulation produce new morphologies during evolution.
Here we investigate the genetic basis for differences in leaf
form between two closely related plant species, Arabidopsis
thaliana and Cardamine hirsuta. We report that in C. hirsuta,
class I KNOTTED1-like homeobox (KNOX) proteins are
required in the leaf to delay cellular differentiation and
produce a dissected leaf form, in contrast to A. thaliana, in
which KNOX exclusion from leaves results in a simple leaf
form. These differences in KNOX expression arise through
changes in the activity of upstream gene regulatory sequences.
The function of ASYMMETRIC LEAVES1/ROUGHSHEATH2/
PHANTASTICA (ARP) proteins to repress KNOX expression
is conserved between the two species, but in C. hirsuta the
ARP-KNOX regulatory module controls new developmental
processes in the leaf. Thus, evolutionary tinkering with KNOX
regulation, constrained by ARP function, may have produced
diverse leaf forms by modulating growth and differentiation
patterns in developing leaf primordia.
Morphological innovations are often associated with altered expression
of key developmental regulators1. However, it is unclear how such
divergent expression arises, how it modifies growth to produce
differences in form, or how the potentially pleiotropic effects of
altered regulatory gene activity are constrained during evolution.
Leaves of seed plants provide an attractive system to study the
evolution of developmental mechanisms because they present considerable
morphological variation. Leaf form can be described as
simple (if the leaf blade is undivided) or dissected (if the blade is
divided into distinct leaflets). Both simple and dissected leaves are
initiated at the flanks of a pluripotent structure termed the shoot
apical meristem (SAM). In simple-leafed species, such as A. thaliana
and maize, ARP myb proteins act in the leaf to confine KNOX
transcription factors to the meristem2–6. Conversely, many dissectedleafed
species accumulate KNOX proteins in the leaf and ARP proteins
in the meristem7,8. However, it is not known whether KNOX activity is
required to produce a dissected leaf, or whether differences in ARP
function or regulation are responsible for the divergent patterns of
KNOX expression seen in different species.
To answer these questions, we analyzed dissected leaf development
in C. hirsuta, a small crucifer related to the simple-leafed model
species A. thaliana (Fig. 1a–d). Unlike many A. thaliana relatives,
C. hirsuta has the distinct advantages of being a diploid, self-compatible
plant that can be used for genetic analyses and transformed, thus
allowing parallel genetic studies of leaf development to be conducted
in species that diverged relatively recently9. To understand whether
KNOX expression in the leaf is associated with dissected leaf form, we
examined KNOX protein accumulation patterns in C. hirsuta and
A. thaliana shoot apices. Class I KNOX proteins were expressed in the
SAM but were excluded from the cells that comprise an initiating leaf
primordium in both A. thaliana and C. hirsuta (Fig. 1e,f). However,
in contrast to what we observed in A. thaliana, we observed nuclear
expression of KNOX proteins in later leaf primordia of C. hirsuta
(Fig. 1e,f), associated with leaflet initiation (Fig. 1g,h).
To investigate whether KNOX activity is required for leaflet initiation,
we reduced expression of the KNOX gene SHOOTMERISTEMLESS
(STM) in C. hirsuta by RNA interference (RNAi) (Fig. 2).
C. hirsuta STM is expressed in the SAM of the embryo and mature
plant and is repressed in the majority of cells that comprise initiating
leaf primordia (Fig. 2a,b). However, in contrast to STM expression in
A. thaliana (Fig. 2c), we also observed C. hirsuta STM expression
throughout the outer cell layers at the base of initiating leaf primordia
(Fig. 2b). In comparison to wild-type plants (Fig. 2d), strong RNAi
lines produced shootless plants with fused cotyledons that often
initiated leaves from ectopic positions (Fig. 2e). Thus, as in
A. thaliana10, C. hirsuta STM is required for SAM initiation and
cotyledon separation in the embryo. Furthermore, in weak RNAi lines
that developed a functional SAM, leaflet initiation was severely
reduced (Fig. 2f–h, 0.8 ± 0.2 leaflets per leaf in C. hirsuta STM
RNAi lines compared with 4.4 ± 0.2 in wild-type plants), demonstrating
that C. hirsuta STM is required to initiate leaflets.
To investigate whether the control of leaflet initiation by C. hirsuta
STM involves regulation of cell division, we assayed the effects of
reducing C. hirsuta STM activity on expression of HISTONE 4 (H4),
Received 10 March; accepted 2 June; published online 2 July 2006; doi:10.1038/ng1835
Department of Plant Sciences, University of Oxford, South Parks Road, Oxford OX1 3RB, UK. Correspondence should be addressed to
M.T. (miltos.tsiantis@plants.ox.ac.uk).
942 VOLUME 38 [ NUMBER 8 [ AUGUST 2006 NATURE GENETICS
L E T T E R S©2006NaturePublishingGrouphttp://www.nature.com/naturegenetics
which has previously been used to monitor cell cycle activity in
A. thaliana lateral organs11. We observed that in C. hirsuta STM
RNAi plants, fewer cells in developing leaf primordia express C.
hirsuta H4 (Fig. 2j,k, 23.5 ± 2.1 cells in C. hirsuta STM RNAi lines
compared with 76.5 ± 3.2 cells in wild-type plants), and the epidermal
cells are much larger in these leaves than in wild-type plants
(Fig. 2l,m). This reduction in C. hirsuta H4 expression and increased
cell expansion suggests that C. hirsuta STM prevents the precocious
exit of tissues from the cell cycle into differentiation pathways, thus
promoting leaflet initiation. To investigate whether KNOX expression
a c e g
hfdb
4 2 3
*
3
1
5
*
*
*
4!
2
03
1
d e f j k l m
ihgc
2
0
3
1
ba
* *
Figure 2 C. hirsuta STM expression in young leaf primordia is required for leaflet initiation. (a,b) In situ localization of C. hirsuta STM mRNA. C. hirsuta
STM expression is restricted to the SAM of embryos (longitudinal section, (a) and vegetative shoot apices (transverse section, (b) but absent from most cells
that comprise leaf primordia. However, expression is observed throughout the outer cell layers of P0 to P2 leaf primordia (arrowheads, 0, 1, 2) and in some
cells of P3 leaf primordia (arrows, 3). (c) In situ localization of STM mRNA. Transverse section through an A. thaliana shoot apex shows expression is
restricted to the SAM and absent from leaf primordia (0–3). (d–f) Vegetative C. hirsuta plants. (d) Wild-type. (e) Scanning electron micrograph (SEM) of a
strong C. hirsuta STM RNAi line with fused cotyledons (arrowhead) and ectopic leaf initiation (arrow). (f) A weak C. hirsuta STM RNAi line with simple
leaves. (g–i) C. hirsuta rosette leaves. (g) Wild-type with four lateral leaflets. (h) A weak C. hirsuta STM RNAi line lacking lateral leaflets. (i) 35S::KN1-GR
induced with 10–6 M dexamethasone with ectopic leaflets initiated upon leaflets (arrowheads). (j,k) In situ localization of C. hirsuta H4 mRNA. Longitudinal
sections through vegetative apices of wild-type (j) and C. hirsuta STM RNAi (k). * indicates meristem. (l,m) SEM of epidermal cells of the terminal leaflet of
leaf three in wild-type (l) and C. hirsuta STM RNAi (m). Scale bars: 20 mm (a–c,l,m), 1 cm (d, f–i), 500 mm (e), 50 mm (j,k).
Figure 1 KNOX proteins accumulate in the
dissected leaf of C. hirsuta but are excluded from
the simple leaf of A. thaliana. (a,b) Mature plants
of C. hirsuta (a), and A. thaliana (b). (c,d) Rosette
(left) and cauline (right) leaves of C. hirsuta (c)
and A. thaliana (d). C. hirsuta leaves are
dissected into leaflets, each of which is borne on
a petiolule attached to the rachis (c), whereas
A. thaliana leaves are simple (d). (e,f) Immunolocalization
of class I KNOX proteins in transverse
sections of shoot apices shows nuclear expression
of KNOX proteins in the SAM (*) but no expression
in initiating leaf cells (arrow) of C. hirsuta (e)
and A. thaliana (f). In C. hirsuta (e), nuclear
expression of KNOX proteins is seen throughout
plastochron (P) 2 (2), localized to the initiating
leaflets and vasculature in P3 (arrowheads, 3)
and limited to vascular-associated cells in P4 (4).
No KNOX expression is seen in leaves of
A. thaliana (f). (g,h) Scanning electron micrographs
of the shoot apex of C. hirsuta (g) shows
that the youngest leaf primordium (1) initiates on
the flanks of the SAM (*) with a simple shape,
and leaflets initiate at the leaf margins 1–2
plastochrons later (arrowheads, 3,4; the distal
leaflet of P4 has been removed) in a basipetal
manner. By contrast, leaf primordia in A. thaliana
(h) initiate at the SAM (*) and continue to
develop with a simple shape. Scale bars: 2 cm
(a,b), 0.5 cm (c,d), 50 mm (e,f), 100 mm (g,h).
NATURE GENETICS VOLUME 38 [ NUMBER 8 [ AUGUST 2006 943
L E T T E R S©2006NaturePublishingGrouphttp://www.nature.com/naturegenetics
is sufficient for leaflet initiation, we expressed the maize KNOTTED1
(KN1) protein, which is able to rescue A. thaliana stm mutants12, in
C. hirsuta in a dosage-sensitive manner using a fusion with the rat
glucocorticoid receptor (KN1-GR)13. A single induction of KN1
activity with 10–6 M dexamethasone resulted in reiteration of a second
order of leaflets along the elongated petiolules of first-order leaflets
(Fig. 2i). These results demonstrate that KNOX activity is not only
necessary but also sufficient for leaflet initiation in C. hirsuta. Elevated
KNOX expression in the dissected-leaf tomato plant can also increase
leaflet number14,15, suggesting that the requirement for KNOX activity
in C. hirsuta leaf development may extend to other species where
dissected leaf morphology has evolved independently7.
We have shown that differences in KNOX expression contribute to
the different leaf forms observed in A. thaliana and C. hirsuta,
indicating that distinct mechanisms of KNOX gene regulation evolved
in these two closely related species. In A. thaliana, AS1 represses
KNOX gene expression in the leaf; therefore, loss of this regulation
could be responsible for KNOX expression in C. hirsuta leaves. This
scenario would be consistent with the coexpression of KNOX and ARP
proteins observed in the shoot meristems of many dissected-leaf
plants8. To investigate this possibility, we determined the extent of
functional equivalence between C. hirsuta AS1 and A. thaliana AS1.
Expression of C. hirsuta AS1 under the control of the broadly
expressed CaMV 35S promoter complemented the A. thaliana as1
mutant phenotype (Fig. 3a–c) and repressed expression of the KNOX
gene BREVIPEDICELLUS (BP) in as1 leaves (Fig. 3d–f), indicating
that the function of the two proteins is conserved. Moreover, C.
hirsuta AS1 mRNA was expressed in leaves and excluded from the
SAM (Fig. 3g) in an equivalent pattern as AS1 in A. thaliana2. Thus, it
is unlikely that changes in either the function or expression of AS1
account for the differences in KNOX expression and leaf shape
between A. thaliana and C. hirsuta.
We next investigated whether the differences in KNOX gene
expression observed between the two species are attributable to
differential activity of KNOX gene regulatory sequences. To test this
idea, we analyzed 5¢ upstream regions of the KNOX genes STM and BP
(Supplementary Fig. 1 online) and performed promoter swap experiments
with these regions between C. hirsuta and A. thaliana. We
reasoned that if the regulatory information necessary for speciesspecific
expression is contained within the promoter regions, then
each promoter should drive reporter gene expression regardless of the
species into which it is transformed. If, however, species-specific
activity of trans regulatory factors is required for correct KNOX
gene expression, then each reporter should reflect the expression
pattern of the species into which it is transformed. We found that
each reporter reflected the endogenous gene expression pattern of its
promoter in both the native and heterologous context (Fig. 3). That is,
the A. thaliana BP promoter generated GUS expression in the SAM of
both A. thaliana and C. hirsuta (Fig. 3h,i), and the C. hirsuta BP
promoter generated GUS expression in both the SAM and leaves of
both species (Fig. 3j,k). Swapping the C. hirsuta STM promoter region
between C. hirsuta and A. thaliana gave similar results: the C. hirsuta
promoter generated GUS expression in the SAM and the abaxial side
of developing leaves in both species (Fig. 3l,m), whereas the
A. thaliana promoter generated expression in the SAM only (Fig. 3n).
These results indicate that differences in KNOX gene expression
between A. thaliana and C. hirsuta are at least in part determined
by differential activity of promoter sequences. KNOX activity in
a
b
c
e
d h j
ki
ml
*
f
g
n
Figure 3 AS1 function is conserved between A. thaliana and C. hirsuta, whereas 5¢ upstream regions of KNOX genes are sufficient to drive species-specific
expression. (a–c) Rosettes of A. thaliana wild-type (Columbia) (a), as1-1 showing an asymmetrical leaf lamina (arrowhead) and short, broad petioles (arrow)
(b) and as1-1;35S::ChAS1 transformants showing restoration of leaf shape (arrowhead) and petiole length (arrow) (c). (d–f) Seedlings of the same genotypes
stained for BP::GUS expression. BP::GUS expression is observed only in the SAM of wild-type (d); ectopic expression is observed in as1 leaves (arrow, e),
and this ectopic expression is repressed in as1-1;35S::cAS1 (f). (g) In situ localization of C. hirsuta AS1 mRNA in the shoot apex of C. hirsuta, showing
expression in initiating leaf primordia (arrowhead) on the flanks of the SAM (*) and in leaflets (arrow). (h–n) Seedlings stained for GUS expression. BP::GUS
expression is restricted to the SAM in A. thaliana (h) and C. hirsuta (i). ChBP::GUS is expressed in both the SAM and leaves in A. thaliana (j) and C. hirsuta
(k). Two leaves have been removed for clarity in i and k. ChSTM::GUS is expressed in both the SAM and the abaxial side of young leaves (arrowheads) in
C. hirsuta (l) and A. thaliana (m), whereas STM::GUS expression is restricted to the SAM in A. thaliana (n). Scale bars: 0.5 cm (a–c, h–k), 1 mm (d–f),
50 mm (g), 0.5 mm (l–n).
944 VOLUME 38 [ NUMBER 8 [ AUGUST 2006 NATURE GENETICS
L E T T E R S©2006NaturePublishingGrouphttp://www.nature.com/naturegenetics
developing leaf primordia is sufficient to elicit leaflet formation in the
simple leaf of A. thaliana (Supplementary Fig. 2 online), suggesting
that cis regulatory changes may be sufficient to determine the
differences in leaf form between A. thaliana and C. hirsuta. However,
future work will determine if this is the case or if additional factors
facilitate KNOX-dependent leaflet formation in C. hirsuta.
To understand how the pleiotropic effects (such as compressed
proximodistal axis and supernumerary leaflets; Fig. 4a) of widespread
KNOX expression in C. hirsuta leaves are constrained, we conducted a
genetic screen to isolate recessive mutants that phenocopy the effects
of KNOX overexpression. One such mutant initiated leaflets close
together along an extremely compressed proximodistal axis and had
additional orders of leaflets (Fig. 4b,c). This mutant phenotype was
very similar to that observed in transgenic C. hirsuta lines when we
reduced C. hirsuta AS1 activity using RNAi (data not shown),
suggesting that these phenotypic effects were a consequence of
loss of C. hirsuta AS1 function. Molecular analysis confirmed
that the mutant contained a premature stop codon at amino
acid residue 170 in the C. hirsuta AS1 protein sequence that
cosegregated with the mutant phenotype, and the allele was hence
designated chas1-1 (Supplementary Fig. 3 online). Thus, C. hirsuta
AS1 activity is required for development of the proximodistal axis of
the leaf and for determining number and positioning of leaflets along
this axis.
To investigate whether C. hirsuta AS1 controls
dissected leaf form by defining the
domain and level of KNOX expression, we
analyzed KNOX protein and mRNA accumulation
in C. hirsuta AS1 RNAi lines and chas1-
1 mutants. We observed increased KNOX
protein accumulation in C. hirsuta AS1
RNAi leaves (Supplementary Fig. 4 online)
and observed ectopic expression of C. hirsuta
BP but not C. hirsuta STM in chas1-1 leaves
(data not shown and Fig. 4). Notably, the
pattern of ectopic C. hirsuta BP expression
correlated well with the phenotypic perturbations
observed in chas1-1. For example, in
contrast to the wild-type expression of
C. hirsuta BP (Fig. 4d), in chas1-1, we
observed intense C. hirsuta BP expression in the adaxial domain and
base of developing leaves (Fig. 4e), correlating with repression of
growth and differentiation along the proximodistal axis (Fig. 4b,c)
and adaxial rachis (Fig. 4h,i) of the leaf. The activity of
C. hirsuta AS1 in controlling differentiation of the adaxial side of
the leaf is shared by ARP proteins in other plant species8,16,17.
Additionally, these results indicate that, at least in C. hirsuta, the
roles of ARP proteins in axial patterning and KNOX repression are
intimately intertwined.
To investigate whether this reduction in growth along the chas1-1
leaf rachis reflects a reduction in cell division or cell expansion, we
analyzed C. hirsuta H4 gene expression and cell size in developing
leaves. Compared with wild-type (Fig. 4f), a greater proportion of
cells in chas1-1 leaf primordia express C. hirsuta H4, particularly at the
leaf base (Fig. 4g). In addition, epidermal cells along the adaxial
surface of the chas1-1 leaf rachis fail to elongate, or differentiate a
striated cell wall, as occurs in wild-type (Fig. 4h,i). Similar defects
were observed in the leaves of C. hirsuta plants expressing BP under
the control of the 35S promoter (Supplementary Fig. 5 online). These
observations suggest that C. hirsuta AS1 regulates C. hirsuta BP
expression within the C. hirsuta leaf, thereby defining the correct
timing for leaf cells to exit the cell cycle and enter differentiation
pathways. Thus, changes in KNOX promoter activity underpin differences
in leaf shape between A. thaliana and C. hirsuta by allowing
a b c
d f h
e g i
*
*
*
*
Figure 4 C. hirsuta AS1 delimits C. hirsuta BP
expression and controls leaflet positioning by
regulating growth. (a–c) Silhouettes of rosette
(left) and cauline (right) leaves of 35S::KN1-GR
induced once with 10–6 M dexamethasone (a),
wild type (b), and chas1-1 (c). Arrows denote
the length of the leaf rachis and arrowheads
indicate extra leaflets. (d,e) In situ localization
of C. hirsuta BP mRNA. Longitudinal sections of
wild-type (d) and chas1-1 (e) showing ectopic
expression in the adaxial side and the base of
developing leaves (arrowheads). (f,g) In situ
localization of C. hirsuta H4 mRNA. Longitudinal
sections of wild-type (f) and chas1-1 (g) showing
C. hirsuta H4–expressing cells at the base of
developing leaves in chas1-1 but not wild-type
(arrowheads). * indicates shoot meristem.
(h,i) Scanning electron micrographs of epidermal
cells on the adaxial surface of the leaf rachis at
the position of leaflet insertion in wild-type
(h) and chas1-1 (i). Scale bars: 1 cm (a–c),
20 mm (d–i).
NATURE GENETICS VOLUME 38 [ NUMBER 8 [ AUGUST 2006 945
L E T T E R S©2006NaturePublishingGrouphttp://www.nature.com/naturegenetics
KNOX proteins to become part of the regulatory toolkit that controls
leaf growth and differentiation. However, potentially pleiotropic
effects arising from KNOX activity in leaves are constrained by the
repressive action of C. hirsuta AS1.
We found it striking that although the molecular function of
C. hirsuta AS1 and A. thaliana AS1 to repress KNOX gene expression
is conserved, the developmental significance of this repression is
different for the two species. In A. thaliana, AS1 acts to safeguard
leaf fate by maintaining the repression of KNOX expression in leaves.
By contrast, in C. hirsuta, KNOX gene regulatory sequences drive
expression in the leaf where KNOX activity is required for dissected
leaf development. Within this different developmental context of the
dissected leaf, C. hirsuta AS1 constrains the spatiotemporal domain of
KNOX expression and hence leaflet number and arrangement. Our
work identifies two processes that underpin the evolution of new
morphologies in multicellular eukaryotes. First, changes in the expression
domain of key developmental regulators offer the potential to
alter morphology by changing tissue growth. Second, conserved
molecular interactions of these regulators, within their new expression
domains, can acquire new developmental significance and mold
morphology to its final state. Differences between C. hirsuta and
A. thaliana extend to many other aspects of their growth and development,
including shoot branching and floral organ morphogenesis.
Therefore, future research in these species will test how robustly these
principles apply to the evolution of relevant developmental pathways.
METHODS
Plant growth conditions. Plants were grown in a greenhouse with supplemental
lighting (days: 18 h, 20 1C; nights: 6 h, 16 1C).
Genetic stocks. Wild-type C. hirsuta seed was collected from wild populations
in Oxford, UK; verified by internal transcribed spacer sequencing; and selfpollinated
for seven generations before use (specimen voucher Hay 1 (OXF)).
Wild-type C. hirsuta seed was X-ray–irradiated at 16 kR, sown and harvested in
pools of five plants. Approximately 100 seed of 150 M2 pools, giving a total of
1,500 plants, were screened. Mutant characterization was performed after
backcrossing to wild-type C. hirsuta twice.
Transgenic construction. All primers are listed in Supplementary Table 1
online. To construct the C. hirsuta STM RNAi vector, a 310-bp fragment was
amplified from C. hirsuta shoot cDNA by PCR with the primers ChSTMrnai-F
and ChSTMrnai-R. This fragment was cloned in both sense and antisense
orientations in the PHANNIBAL vector18 using the restriction enzyme pairs
XbaI/ClaI and EcoRI/KpnI. This RNAi cassette was transferred as a NotI
fragment into the binary vector pMLBART19, transformed into the Agrobacterium
tumifaciens strain GV3101 and used to transform wild-type C. hirsuta
plants by a modified floral dipping protocol. We analyzed 14 independent T1
lines. We constructed a C. hirsuta AS1 (hereafter, ChAS1) RNAi vector in an
identical manner using a 343-bp PCR fragment amplified from C. hirsuta shoot
cDNA with the primers ChAS1rnai-F and ChAS1rnai-R, and we used it to
transform wild-type C. hirsuta plants as above. We analyzed 16 independent T2
lines. The pMLBART vector alone was used to transform wild-type C. hirsuta
plants as above, and all T1 lines were phenotypically wild-type. We used three
independent T2 lines as wild-type comparisons for analyses of RNAi lines.
Transcript levels were analyzed by RT-PCR (Supplementary Fig. 6 online). A
previously described 35S::KN1-GR translational fusion13 was used to transform
wild-type C. hirsuta plants as described above. We analyzed 16 independent T2
lines. To construct the 35S::ChAS1 vector, we amplified a 1,117-bp fragment of
the C. hirsuta AS1 coding region by PCR using a proofreading Taq polymerase
(Pyrobest, Takara) from a full-length cDNA clone using the primers ChAS1-F
and ChAS1-R. The PCR product was cloned into the pCR Blunt vector
(Invitrogen), sequenced to confirm fidelity and cloned as an EcoRI fragment
behind the CaMV 35S promoter of the pART7 vector19. The 35S::ChAS1::ocs
cassette was transferred as a NotI fragment into the binary vector pMLBART
and transformed into as1-1 mutant plants by floral dipping. We analyzed 100
independent T1 lines for recovery of the as1 mutant phenotype as described
previously for 35S::AS1 (ref. 20). Five T2 lines with a single transgene copy were
crossed to as1-1;BP::GUS plants3, and GUS expression was analyzed in the F1.
To make transcriptional fusions of C. hirsuta BP and C. hirsuta STM to the
uidA (GUS) gene, a BAC library of C. hirsuta genomic clones was screened (a
full description of library construction will be given elsewhere), and B6-kb
EcoRI and XbaI restriction fragments of C. hirsuta BP and C. hirsuta STM DNA,
respectively, were cloned into pBluescript (Stratagene). We amplified 4 kb of
upstream sequence, including the 5¢ UTR, by PCR using a proofreading Taq
polymerase (Pyrobest, Takara) with the primers M13 reverse and ChBP-R with
a PstI restriction site introduced at the ATG, and cSTMpst-F and cSTMbam-R
with a BamHI restriction site introduced at the ATG. These sequences were
transferred, as a PstI fragment for C. hirsuta BP (hereafter ‘ChBP’) and as a PstI/
BamHI fragment for C. hirsuta STM (hereafter ‘ChSTM’), upstream of GUS in
the pRITA vector19. Orientation and integrity of the sequence junctions were
confirmed by sequencing. Transcriptional fusions were generated in a similar
manner using B5 kb of upstream sequence, including the 5¢ UTR of BP and
STM. All four promoter-GUS cassettes were transferred as NotI fragments into
the binary vector pMLBART, transformed into A. tumefaciens, as above, and
used to transform both wild-type A. thaliana (Columbia ecotype) and
C. hirsuta. GUS expression was analyzed in 97 independent T1 lines for
ChBP::GUS and 96 lines for BP::GUS in A. thaliana, ten lines for ChBP::GUS
and 12 lines for BP::GUS in C. hirsuta, 67 lines for ChSTM::GUS and eight lines
for STM::GUS in A. thaliana, and eight lines for ChSTM::GUS in C. hirsuta. A
previously described 35S::BP construct21 was transformed into A. tumefaciens
and used to transform wild-type C. hirsuta plants as described above. We
analyzed 18 independent T2 lines. PHV44BP, FIL44BP and ANT44BP
lines were generated by constructing pVTOp::BP (ref. 22) and transforming
PHV::LhG4, FIL::LhG4 (gift from Y. Eshed, Weizmann Institute of Science,
Israel) and ANT::LhG4 (ref. 23) plants with this construct. We analyzed 30
independent T2 lines for each construct.
5¢ and 3¢ RACE. C. hirsuta AS1 full-length cDNA sequence was determined in
wild-type lines and chas1-1 mutants by 5¢ and 3¢ RACE. cDNA was generated
using a SmartRace kit (BD Biosciences) according to manufacturer’s protocols.
We used 1 mg of total shoot RNA per reaction. PCR amplification was
performed for 5¢ RACE with the primer ChAS1-R1 and for 3¢ RACE with
the primer ChAS1-F1. C. hirsuta AS1-specific products were cloned, and two
clones from each RACE reaction were sequenced for each genotype. In chas1-1
mutants, a premature stop codon at amino acid 170 of ChAS1 introduces an
AccI site that is not present in wild-type plants. This sequence polymorphism
was used to generate a cleaved amplified polymorphic sequence marker by
amplifying a 600-bp product with primers ChAS1-F2 and ChAS1-R2, which
yielded products of 425 bp and 175 bp after AccI digestion of chas1-1 but not
after digestion of wild-type amplicons.
Leaflet and cell measurements. Average number of leaflets per leaf was
determined for ten C. hirsuta STM RNAi plants and ten wild-type plants.
Average number of C. hirsuta H4–expressing cells in adjacent longitudinal
sections of the two youngest leaf primordia at the shoot apex was determined
for C. hirsuta STM RNAi and wild-type plants (as described in ref. 11). Error
shown in all cases is standard error.
Immunocytochemistry. Fixation and hybridization were carried out as previously
described24 on 8-mm paraffin sections using a previously described
polyclonal antibody to KNOX25 that detects class I KNOX proteins (encoded by
a four-member gene family in A. thaliana and C. hirsuta (data not shown)).
Scanning electron microscopy. Fixation and dehydration were carried out as
previously described26. Scanning electron microscopy was performed using a
JSM-5510 microscope (Jeol).
In situ RNA localization. Fixation and hybridization were carried out as
previously described10 on 8-mm paraffin sections using probes for A. thaliana
STM10, C. hirsuta STM, C. hirsuta H4, C. hirsuta AS1 and C. hirsuta BP. To
generate a probe to C. hirsuta AS1, a 373-bp fragment was amplified from
C. hirsuta shoot cDNA using the primers AS1-F and AS1-R. To generate a
probe to C. hirsuta H4, a 297-bp fragment was amplified from C. hirsuta shoot
946 VOLUME 38 [ NUMBER 8 [ AUGUST 2006 NATURE GENETICS
L E T T E R S©2006NaturePublishingGrouphttp://www.nature.com/naturegenetics
cDNA using the primers H4-F and H4-R. To generate a probe to C. hirsuta
STM, a 986-bp fragment was amplified from a cDNA clone using the primers
ChSTM-F and ChSTM-R. Three probes were generated to C. hirsuta BP by
amplifying 306-bp, 387-bp and 351-bp fragments from a cDNA clone using the
primer pairs ChBP457-F and ChBP762-R, ChBP700-F and Ch1086-R and
ChBP5¢-F and ChBP5¢-R, respectively. All fragments were cloned into the
pGEM T-Easy vector (Promega) and sequenced to determine orientation.
Antisense and sense probes were transcribed and DIG labeled as
previously described10.
Leaf silhouettes. Leaves were flattened onto clear adhesive, adhered to white
paper and digitally scanned.
Chemical treatments. Dexamethasone (Sigma) was dissolved in water and
applied at a concentration of 10–6 M with 0.02% silwet using a paintbrush.
Accession codes. GenBank: C. hirsuta STM mRNA, complete coding sequence
(cds), DQ512732; C. hirsuta AS1 mRNA, complete cds, DQ512733; C. hirsuta
BP mRNA, complete cds, DQ630764; C. hirsuta BP gene, 5¢ upstream region,
DQ526379; C. hirsuta STM gene, 5¢ upstream region, DQ526380.
Note: Supplementary information is available on the Nature Genetics website.
ACKNOWLEDGMENTS
We thank S. Hake for 35S::BP and 35S::KN1-GR plasmids and helpful discussions;
M. Scanlon for anti-KNOX antibody, J. Long for STM plasmid and advice on
in situ hybridization; I. Furner for mutagenesis, T. Rich, S. Hiscock and D. Bailey
for help with C. hirsuta identification; Y. Eshed and J. Bowman for cloning
vectors, FIL::LhG4 and PHV::LhG4 lines; M. Lenhard for ANT::LhG4; J. Craft
for VTOp::BP and VTOP::STM lines; I. Moore for discussions on the LhG4
transcription activation system and S. Langer and M. Anezaki for technical
assistance. We also thank J. Langdale for comments on the manuscript,
A. Hudson for helpful discussions, J. Baker for photography and Genome and
Agricultural Biotechnology for support of library construction. This work was
funded by a Biotechnology and Biological Sciences Research Council grant and
Oxford University Pump Priming grant to M.T. A.H is the recipient of a
University of Oxford Glasstone Research Fellowship and a Balliol College Junior
Research Fellowship. We also acknowledge the support of the Gatsby Charitable
foundation and the Royal Society.
COMPETING INTERESTS STATEMENT
The authors declare competing financial interests (see the Nature Genetics website
for details).
Published online at http://www.nature.com/naturegenetics
Reprints and permissions information is available online at http://npg.nature.com/
reprintsandpermissions/
1. Carroll, S.B. Evolution at two levels: on genes and form. PLoS Biol. 3, e245 (2005).
2. Byrne, M.E. et al. Asymmetric leaves1 mediates leaf patterning and stem cell function
in Arabidopsis. Nature 408, 967–971 (2000).
3. Ori, N., Eshed, Y., Chuck, G., Bowman, J.L. & Hake, S. Mechanisms that control knox
gene expression in the Arabidopsis shoot. Development 127, 5523–5532 (2000).
4. Schneeberger, R., Tsiantis, M., Freeling, M. & Langdale, J.A. The rough sheath2 gene
negatively regulates homeobox gene expression during maize leaf development.
Development 125, 2857–2865 (1998).
5. Tsiantis, M., Schneeberger, R., Golz, J.F., Freeling, M. & Langdale, J.A. The maize
rough sheath2 gene and leaf development programs in monocot and dicot plants.
Science 284, 154–156 (1999).
6. Timmermans, M.C., Hudson, A., Becraft, P.W. & Nelson, T. ROUGH SHEATH2: a Myb
protein that represses knox homeobox genes in maize lateral organ primordia. Science
284, 151–153 (1999).
7. Bharathan, G. et al. Homologies in leaf form inferred from KNOXI gene expression
during development. Science 296, 1858–1860 (2002).
8. Kim, M., McCormick, S., Timmermans, M. & Sinha, N. The expression domain of
PHANTASTICA determines leaflet placement in compound leaves. Nature 424,
438–443 (2003).
9. Koch, M., Haubold, B. & Mitchell-Olds, T. Molecular systematics of the Brassicaceae:
evidence from coding plastidic matK and nuclear Chs sequences. Am. J. Bot. 88,
534–544 (2001).
10. Long, J.A., Moan, E.I., Medford, J.I. & Barton, M.K. A member of the KNOTTED class
of homeodomain proteins encoded by the SHOOTMERISTEMLESS gene of Arabidopsis.
Nature 379, 66–69 (1996).
11. Dinneny, J.R., Yadegari, R., Fischer, R.L., Yanofsky, M.F. & Weigel, D. The role of
JAGGED in shaping lateral organs. Development 131, 1101–1110 (2004).
12. Kim, J.Y., Yuan, Z. & Jackson, D. Developmental regulation and significance of KNOX
protein trafficking in Arabidopsis. Development 130, 4351–4362 (2003).
13. Hay, A., Jackson, D., Ori, N. & Hake, S. Analysis of the dompetence to respond to
KNOTTED1 activity in Arabidopsis leaves using a steroid induction system. Plant
Physiol. 131, 1671–1680 (2003).
14. Hareven, D., Gutfinger, T., Parnis, A., Eshed, Y. & Lifschitz, E. The making of a compound
leaf: Genetic manipulation of leaf architecture in tomato. Cell 84, 735–744 (1996).
15. Chen, J-J., Janssen, B-J., Williams, A. & Sinha, N. A gene fusion at a homeobox locus:
Alterations in leaf shape and implications for morphological evolution. Plant Cell 9,
1289–1304 (1997).
16. Waites, R., Selvadurai, H.R., Oliver, I.R. & Hudson, A. The PHANTASTICA gene
encodes a MYB transcription factor involved in growth and dorsoventrality of lateral
organs in Antirrhinum. Cell 93, 779–789 (1998).
17. Xu, L. et al. Novel as1 and as2 defects in leaf adaxial-abaxial polarity reveal the
requirement for ASYMMETRIC LEAVES1 and 2 and ERECTA functions in specifying
leaf adaxial identity. Development 130, 4097–4107 (2003).
18. Wesley, S.V. et al. Construct design for efficient, effective and high-throughput gene
silencing in plants. Plant J. 27, 581–590 (2001).
19. Gleave, A.P. A versatile binary vector system with a T-DNA organisational structure
conducive to efficient integration of cloned DNA into the plant genome. Plant Mol.
Biol. 20, 1203–1207 (1992).
20. Theodoris, G., Inada, N. & Freeling, M. Conservation and molecular dissection of
ROUGH SHEATH2 and ASYMMETRIC LEAVES1 function in leaf development. Proc.
Natl. Acad. Sci. USA 100, 6837–6842 (2003).
21. Chuck, G., Lincoln, C. & Hake, S. Knat1 induces lobed leaves with ectopic meristems
when overexpressed in Arabidopsis. Plant Cell 8, 1277–1289 (1996).
22. Craft, J. et al. New pOp/LhG4 vectors for stringent glucocorticoid-dependent transgene
expression in Arabidopsis. Plant J. 41, 899–918 (2005).
23. Lenhard, M., Jurgens, G. & Laux, T. The WUSCHEL and SHOOTMERISTEMLESS
genes fulfil complementary roles in Arabidopsis shoot meristem regulation. Development
129, 3195–3206 (2002).
24. Smith, L.G., Greene, B., Veit, B. & Hake, S. A dominant mutation in the maize
homeobox gene, Knotted-1, causes its ectopic expression in leaf cells with altered
fates. Development 116, 21–30 (1992).
25. Scanlon, M.J., Schneeberger, R.G. & Freeling, M. The maize mutant narrow sheath
fails to establish leaf margin identity in a meristematic domain. Development 122,
1683–1691 (1996).
26. Bowman, J.L., Smyth, D.R. & Meyerowitz, E.M. Genetic interactions among floral
homeotic genes of Arabidopsis. Development 112, 1–20 (1991).
NATURE GENETICS VOLUME 38 [ NUMBER 8 [ AUGUST 2006 947
L E T T E R S©2006NaturePublishingGrouphttp://www.nature.com/naturegenetics