Activation of CRABS CLAW in the Nectaries andCarpels of Arabidopsis W

Ji-Young Lee,1 Stuart F. Baum,2 John Alvarez,3 Amita Patel, Daniel H. Chitwood, and John L. Bowman4

Section of Plant Biology, University of California Davis, Davis, California 95616

CRABS CLAW (CRC), a member of the YABBY gene family, is required for nectary and carpel development. To further

understand CRC regulation in Arabidopsis thaliana, we performed phylogenetic footprinting analyses of 59 upstream

regions of CRC orthologs from three Brassicaceae species, including Arabidopsis. Phylogenetic footprinting efficiently

identified functionally important regulatory regions (modules), indicating that CRC expression is regulated by a combination

of positive and negative regulatory elements in the modules. Within the conserved modules, we identified putative binding

sites of LEAFY and MADS box proteins, and functional in vivo analyses revealed their importance for CRC expression. Both

expression and genetic studies demonstrate that potential binding sites for MADS box proteins within the conserved

regions are functionally significant for the transcriptional regulation of CRC in nectaries. We propose that in wild-type

flowers, a combination of floral homeotic gene activities, specifically the B class genes APETALA3 and PISTILLATA and the

C class gene AGAMOUS act redundantly with each other and in combination with SEPALLATA genes to activate CRC in the

nectaries and carpels. In the absence of B and C class gene activities, other genes such as SHATTERPROOF1/2 can

substitute if they are ectopically expressed, as in an A class mutant background (apetala2). These MADS box proteins may

provide general floral factors that must work in conjunction with specific factors in the activation of CRC in the nectaries

and carpels.

INTRODUCTION

Nectaries are organs that produce and secrete nectar, primarily

consisting of sugar and proteins (Fahn, 1979). The main function

of nectaries is to lure pollinators and protectors by providing

sugary foods as rewards. Though there are reports of nectaries in

ferns (Darwin, 1877) and in Gnetales (Porsch, 1910), nectaries

are most widespread in angiosperms, particularly within flowers.

A large proportion of angiosperm species require pollination by

animals (Eriksson and Bremer, 1992), with pollinators in many

cases being attracted to flowers to gather nectar as their food

source. The fossil records of angiosperms and insects suggest

that the timing of the radiation of angiosperms corresponds to

the radiation of insects that consume nectar as a main food

source (Meeuse, 1978; Crepet and Friis, 1987; Pellmyr, 1992).

Therefore, it is likely that innovation or modification of genetic

mechanisms to produce nectaries occurred during the evolution

of flowering plants.

In species of the Brassicaceae, such as Arabidopsis thaliana,

nectaries develop as a ring at the base of the stamens, with

secretory glands associated with the stamens. In Arabidopsis

crabs claw (crc) mutants, all traces of nectary development, both

morphological and molecular, are lacking (Bowman and Smyth,

1999; Baum et al., 2001). CRC encodes a putative transcription

factor containing a zinc finger and helix-loop-helix domain called

the YABBY domain (Bowman and Smyth, 1999; Siegfried et al.,

1999). CRC expression commences in two horseshoe-shaped

domains thought to be the precursors of the nectary, and

expression continues throughout the nectary beyond anthesis

(Baum et al., 2001). Ectopic expression of CRC alone does not

cause the development of ectopic nectaries, implying that other

factors are required for nectary development. Some of these

factors are meristem identity genes, gain- or loss-of-function

alleles of which lead to ectopic nectaries at the base of the flower

pedicel, suggesting that factors both intrinsic and extrinsic to the

flower are required to localize CRC expression. Regardless of

genetic background, CRC is always required for nectaries,

indicating that it is one of the key genes directing nectary

development in Arabidopsis (Baum et al., 2001).

The Arabidopsis genome contains six members of the YABBY

gene family, and all are expressed abaxially in lateral organs

(Bowman and Smyth, 1999; Sawa et al., 1999; Siegfried et al.,

1999; Villanueva et al., 1999). Ectopic expression of two mem-

bers, FILAMENTOUS FLOWER and YABBY3, in the adaxial

regions of leaves is sufficient to promote abaxial cell fates,

indicating a role for these genes in establishing the polarity of

lateral organs (Sawa et al., 1999; Siegfried et al., 1999). Likewise,

CRC is required for proper establishment of adaxial–abaxial

polarity in the carpel. Whereas crc single mutants do not exhibit

1 Current address: Department of Biology, Duke University, Durham, NC27708.2 United States Patent and Trademark Office, 400 Dulany, Alexandria,VA 22312.3 Department of Plant Sciences, Weizmann Institute of Science,Rehovot, 76100 Israel.4 To whom correspondence should be addressed. E-mail jlbowman@ucdavis.edu; fax 530-752-5410.The author responsible for distribution of materials integral to thefindings presented in this article in accordance with the policy describedin the Instructions for Authors (www.plantcell.org) is: John L. Bowman( jlbowman@ucdavis.edu).WOnline version contains Web-only data.Article, publication date, and citation information can be found atwww.plantcell.org/cgi/doi/10.1105/tpc.104.026666.

This article is published in The Plant Cell Online, The Plant Cell Preview Section, which publishes manuscripts accepted for publication after they

have been edited and the authors have corrected proofs, but before the final, complete issue is published online. Early posting of articles reduces

normal time to publication by several weeks.

1 of 12The Plant Cell Preview, www.aspb.orgª 2004 American Society of Plant Biologists

altered carpel polarity, crc kanadi1 carpels develop adaxial

tissues in abaxial positions, and in addition, ectopic adaxial

expression ofCRC in lateral organs, such as leaves and petals, is

sufficient to promote abaxial cell fates (Alvarez and Smyth, 1999;

Eshed et al., 1999). CRC is expressed in the abaxial epidermis of

the carpel, as well as in internal domains, in a complex and

dynamic pattern, and this expression is critical for promotion of

abaxial fates (Bowman and Smyth, 1999; Eshed et al., 1999).

Whereas a role for CRC in carpels may be an ancestral function

within angiosperms since the CRC ortholog in Oryza, DROOP-

ING LEAF, promotes carpel identity, its role in nectaries could be

a derived condition based on the proposed evolutionary origins

of nectaries and the lack of evidence for any other YABBY gene

family member expressing in the nectaries (Brown, 1938; Sawa

et al., 1999; Siegfried et al., 1999; Villanueva et al., 1999;

Yamaguchi et al., 2004).

Previous genetic analyses indicated that the floral homeotic

genes that specify the identity of the other floral organs,

commonly referred to as the ABC genes, influence nectary

development (Baum et al., 2001). In both B (apetala3 [ap3] and

pistillata [pi] ) and C (agamous [ag]) class mutants, a reduction in

nectary development is observed, and in BC double mutants, no

nectaries develop. However, when A function is also compro-

mised, nectary development is restored. CRC is active in the

nectaries and carpels, regions of the flower in which B and C

class genes are also active, raising the possibility that they may

play a role as redundant activators in the nectary development

pathway. To better understand the relationship between CRC

and the floral homeotic genes and to elucidate how CRC is

activated in nectaries and carpels, we undertook an analysis of

its promoter by identifying evolutionarily conserved regulatory

elements by comparison of CRC promoter sequences from

related species, a process called phylogenetic footprinting

(Duret and Bucher, 1997; Tautz, 2000; Colinas et al., 2002).

Based on molecular and genetic analyses, we propose that

a combination of floral homeotic gene activities act redundantly

with each other and in combination with SEPALLATA (SEP)

genes to activateCRC in the nectaries and carpels. TheseMADS

box proteins may provide general floral factors that must work in

conjunction with region-specific factors in the activation of CRC

in the nectaries and carpels.

RESULTS

59 Upstream Regions of CRC Are Conserved in Modules

To identify conserved regions within the CRC promoter, we

chose to analyze CRC orthologs in Lepidium and Brassica, two

Brassicaceae genera closely related to Arabidopsis. Based on

recent phlyogenetic analyses of the Brassicaceae, Lepidium is

more closely related to Arabidopsis than is Brassica (e.g., Koch

et al., 2001a). Carpel structure and morphology in Lepidium and

Brassica are similar to that of Arabidopsis, suggesting that the

CRC orthologs play similar functional roles in the three species

(Polowick and Sawhney, 1986; Bowman and Smyth, 1998; Davis

et al., 1998). That the Lepidium and Brassica genes described

here are CRC orthologs is substantiated by expression patterns,

functional studies, and phylogenetic analyses (J.-Y. Lee and J.L.

Bowman, unpublished data).

Five highly conserved regions were detected in the 59 up-

stream sequences of Arabidopsis, Lepidium, and Brassica CRC

genes, and the regions used for analysis were denoted A, B, C, D,

and E (Figure 1; see also supplemental data online for complete

alignment). The regional analysis was primarily based on the

conservation between Arabidopsis and Lepidium because the

Lepidium CRC promoter driving b-glucuronidase (GUS) expres-

sion in transgenic Arabidopsis exhibited a staining pattern in-

distinguishable from an Arabidopsis CRC:GUS transgene. In

addition, an ArabidopsisCRC cDNA driven by the LepidiumCRC

promoter complements the Arabidopsis crc-1mutant phenotype

in >90% of transformants (data not shown).

To analyze how these conserved regions regulate CRC tran-

scription, transgenic plants in which a GUS reporter gene is

driven by each conserved region were generated. In this pro-

cess, we used region A, which includes 321 bp just upstream of

the start codon, as a proximal transcriptional promoter instead of

using the Cauliflower mosaic virus (CaMV)-derived TATA box.

This was because of preliminary experiments using the CaMV

TATA box resulting in a low frequency of expression in combi-

nation with the other elements and region A being identified as

important for CRC expression in a promoter deletion experiment

(data not shown). In these analyses, we used a two-component

expression system using the chimeric transcription factor, LhG4,

driven by combinations of the identified conserved regions

(Moore et al., 1998). These were transformed into wild-type or

crc-1 plants in which a GUS reporter gene was fused to the pOP

promoter, which is activated by LhG4. Both wild-type and crc-1

plants were analyzed to ascertain whether autoregulation may

play a role and to separately analyze regulation byCRC promoter

regions in early versus late stages of nectary development.

The GUS expression patterns of transgenic plants are sum-

marized in Table 1 and Figure 2. Region A alone does not result in

the transcription of GUS. However, chimeras between A and

other regions result in GUS expression, partially reflecting the

endogenous expression pattern of CRC. Regions C and E were

responsible for most of the positive regulation of CRC. Both C

and E drive GUS expression in the carpels, with E also re-

sponsible for nectary expression. By contrast, regions B and D

did not result in any expression on their own in crc-1, but

influenced expression patterns when combined with C and E

(note the frequency of expression in valves and nectaries driven

by E:D:C:B:A in Table 1), suggesting that the primary roles of

regions B andDmay be to negatively regulateCRC expression in

the sepals and tissues of the gynoecium.

Region C is sufficient to drive reporter gene expression in

carpels at early stages of development, with later expression

restricted to the valves (Figures 2A to 2C). Expression continues

longer than when the GUS reporter gene is driven by the whole

promoter (Figures 2C and 2H). Region E regulates CRC expres-

sion in both nectaries and carpels (Figures 2D to 2F). Expression

in nectaries commences early (stages 8 and 9; stages according

to Smyth et al., 1990) and continues until after anthesis (Figures

2J to 2K and 2O). In a crc-1 background, expression in the

nectary anlagen, the region where nectaries arise in thewild-type

flower, indicates that region E is responsible for initial CRC

2 of 12 The Plant Cell

expression in nectaries before the cell divisions that mark the

initiation of the nectary glands (Figures 2J to 2K). In addition to

driving transcription in nectaries, region E also drives expression

in carpels, initially visible at stage 6 (Figures 2L and 2N) but

conspicuous by stage 7 (Figures 2L and 2M). GUS reporter gene

expression appears to be throughout the entire carpels and

ovules. After stage 13, expression in ovules disappears (Figure

2I), but expression in carpels is still present even after stage 15,

although by this stage it is only present in the upper region of

carpels, including the style and stigma (Figures 2F and 2O). Thus,

the expression pattern driven by region E is in tissues that

endogenously express CRC and, in addition, some tissues

where CRC is not normally expressed. It is likely that the

additional expression observed when the reporter gene is driven

by region E is suppressed by other regions, presumably region B

and/or D because reporter gene expression driven by a chimeric

promoter combining all the conserved regions mimics the

pattern driven by the intact 3.8-kb promoter (Figures 2G, 2H,

2Q, and 2R). GUS expression driven by the intact 3.8-kb pro-

moter and the composite conserved region promoter (E:D:C:B:A)

commences in carpels at stage 6 and continues until stage 14. In

contrast with E:A�GUS, GUS is not expressed in the septum or

ovules at stage 12 (Figure 2P). In nectaries, E:D:C:B:A�GUS

expression commences by stage 9 (Figure 2Q), about the same

stage when the nectary expression is initiated by region E. These

results suggest that the sequences conserved in three closely

related species are sufficient for regulating CRC.

Although region E contains qualitative elements for nectary

expression of CRC, for proper quantitative expression of CRC in

nectaries, region A was also required. As a comparison, when

region E fused with a CaMV-TATA:GUS was introduced into

Arabidopsis, GUS activity in nectaries was reduced in frequency

and intensity (data not shown), suggesting that region Amight act

as a core promoter for transcriptional initiation (Smale, 2001).

Complementation of crcMutants by Promoter Regions

Because regions C and E are largely responsible for CRC

expression in carpels and nectaries, we examined the extent to

which these regions are able to complement crcmutants (Figure

3). Transgenic lines with E:A:LhG4, C:A:LhG4, and E:D:C:B:

A:LhG4 were generated in a line with OP:CRC to determine

the extent of complementation of the phenotypic defects in

carpels and nectaries in a crcmutant (Figure 3). Driving CRC ex-

pression with the composite promoter, E:D:C:B:A, is sufficient to

rescue carpel defects, as exemplified by normal fruit develop-

ment in these transgenic lines (Figure 3D). By contrast, CRC

expression driven by E:A was able to rescue the style defects

associated with crc mutations, but was unable to fully comple-

ment fruit growth, suggesting that rescue of ovary wall defects

was not complete (Figure 3C). Surprisingly, expression of CRC

by C:A was unable to rescue most carpel defects despite C:A

driving expression early in carpel development (Figure 3B).

With regards to nectary development, CRC expression driven

by either E:D:C:B:A or E:A was sufficient to rescue most aspects

of nectary development in a crc background (Figures 3C and 3D).

In both these genotypes, prominent nectary glands develop at

the abaxial base of lateral stamens (Figures 3C and 3D).

Figure 1. Sequence Comparison of 59 Upstream Regions of CRC from Arabidopsis, Lepidium, and Brassica.

VISTA analysis (75% identity and 50 base sliding window) identified five conserved regions in three species (see supplemental data online for complete

alignment). Regionswith >75% identity shared by three species are shaded in pink. The top comparison is betweenArabidopsis and Lepidium and so on.

Five conserved domainswere found betweenArabidopsis and Lepidium, denoted regions A, B, C, D, and E, and used for the functional analyses. A larger

fragment than the conserved region of region A that extends to just before the start codon was used as the transcriptional enhancer instead of a CaMV-

derived TATA box. Sequence comparisonwith Brassica showed that the same regions are conserved but some of the conserved domains do not overlap

with the domains conserved between Arabidopsis and Lepidium. CArG boxes (line with diamond) and putative LEAFY binding sites (line with oval) are

shown along aligned sequences; the top row represents Arabidopsis, the second row represents Lepidium, and the third row represents Brassica. The

CArGboxes andLFYbinding sites functionally analyzed are labeled (EM1, EM2, EL1,CL2, andCL3). The translation start site is identifiedby a vertical line,

and the exons of CRC are denoted by blue boxes. Numbers below denote base pairs of Arabidopsis sequence.

Control of CRC Expression 3 of 12

However, medial nectaries were not present in most flowers, and

although the lateral nectary glands exhibit conspicuous cuticular

thickenings characteristic of wild-type nectary glands, stomata

were not present on most glands as they are in the wild type

(Figure 3A). Expression of CRC using C:A was unable to

complement crc nectary defects (Figures 3B and 3E).

Binding Sites for LEAFY and MADS Box Proteins and the

Regulation of CRC

Previous genetic studies implicated several known transcription

factors in the regulation ofCRC. MADS box genes are candidate

regulators because pi ag flowers lack nectaries (Baum et al.,

2001). In addition to the regulation by MADS box genes, two

genes, LEAFY (LFY) and UFO, thought to regulate both the

homeotic genes and the formation of the third whorl, also affect

nectary development (Baum et al., 2001). In the case of LFY, it

appears to be involved both in inducing nectary development

within the flower and suppressing nectary development outside

the flower.Mutations in anotherMADSbox gene,AP1, also result

in ectopic nectary development (Baum et al., 2001). Thus, we

investigated whether any of these genes may directly regulate

CRC in nectaries.

To begin to address this question, we searched the 59

upstream sequences of CRC in all three species for potential

LFY andMADSbox protein binding sites (Dolan and Fields, 1991;

Treisman, 1992; Busch et al., 1999). In the region spanning 3.8 kb

upstream of the Arabidopsis CRC coding region, the region

found to be necessary and sufficient for proper CRC expression,

four putative LFY binding sites (CCANTG) and two potential

binding sites for MADS box proteins, known as CArG boxes

[CC(A/T)6GG], were identified (Figure 1; see also supplemental

data online). In L. africanum, five LFY binding domains and four

CArG boxes were identified within 5 kb upstream of CRC,

whereas in B. oleracea, two LFY binding domains and four

CArG boxes were identified within 4.5 kb upstream of CRC.

Some of these binding sites are found not only in the regions

conserved among the three species but also in nonconserved

regions (Figure 1). The CArG boxes located in Arabidopsis region

E were identical to those of Lepidium; however, some base

changes in Brassica resulted in slight deviations from the

consensus CArG box sequence (see supplemental data online).

LFY binding domains in regions E and C were also 100%

identical between Arabidopsis and Lepidium but base changes

were found in Brassica. We focused on the binding sites located

in conserved regions because these sites might be functionally

significant for regulating CRC in the context of the other cis-

regulatory elements resident in conserved regions.

To determine the roles of these putative binding sites, site-

directed mutagenesis was performed to alter the sequence of

three LFY binding sites of regions C and E and two CArG boxes

of region E (Figure 1; Tilly et al., 1998; Busch et al., 1999). The

expression of GUS controlled by regions C and E or the fusion of

all five conserved domains containing mutagenized sites was

analyzedand is summarized inTables2 (LFY) and3 (CArGboxes).

Site-directed mutagenesis of the LFY binding site (CCANTG /

AAANTG) in regionE led toslightly decreased levels of expression

but did not affect the pattern of expression. A similar result was

obtained for the LFY binding sites in region C. Mutagenesis of all

three LFY binding sites in the context of the fusion of the five

conserved domains does not change the expression pattern or

the frequency and level of expression (Table 2).

Mutagenesis of the CArG boxes [CC(A/T)6GG/ AA(A/T)6GG]

alters the expression pattern driven by E:A dramatically (Table 3).

No expression was observed in the nectaries in any of the

transgenic lines, with only a few lines exhibiting a low level of

expression in the stigma. The mutagenesis of the CArG boxes

also affected GUS expression driven by the fusion of the five

conserved domains. Only six out of 25 transgenic lines exhibited

expression in nectary anlagen in crc-1, with expression in

nectaries commencing later than with the wild-type fusion pro-

moter. This suggests that additional elements responsible for

CRC nectary expression exist, but for consistent early nectary

expression binding of MADS box protein(s) is essential. The

expression pattern in the carpels was also affected significantly

in the promoters in which the CArG boxes weremutated. Only 10

out of 25 transgenic lines had expression in the valves, similar

to transgenic lines with region C alone (seven out of 23 lines),

suggesting that MADS box proteins also regulate carpel expres-

sion in a redundant manner.

LEAFY Regulation of CRC

Though the site-directed mutagenesis of the LFY binding sites

did not significantly affect the expression pattern in vivo, other

studies indicate the involvement of LFY in the regulation of CRC.

In plants carrying bothCRC:GUSand LFY:LFY:VP16 (Parcy et al.,

1998) transgenes, the level of GUS expression is greatly en-

hanced, whereas the domain of expression is not altered (data

Table 1. Regulation by Conserved Regions of 59 Upstream

Sequences of CRC in crc-1 and Ler

Construct Frequency Expression Pattern

A�GUS 0/12 (crc-1) No expression

B:A�GUS 0/14 (crc-1) No expression

3/10 (Ler) Upper valves

1/10 (Ler)a Pedicel

C:A�GUS 7/23 (crc-1) Valves and sepals

7/19 (Ler) Valves and sepals

D:A�GUS 0/24 (crc-1) No expression

3/17 (Ler) Lateral nectaries

1/17 (Ler)a Pedicel

E:A�GUS 10/24 (crc-1) Upper valves

7/24 (crc-1)b Nectary anlagen, ovule

14/19 (Ler) Style, stigma, and nectaries

E:D:C:B:A�GUS 19/24 (crc-1) Valves

12/24 (crc-1)b Nectary anlagen

pCRC, 3.8kb�GUS 21/24 (crc-1) Valves

15/24 (crc-1)b Nectary anlagen

aOne Ler plant with B:A�GUS showed expression on the pedicel as

well as in upper valves, and the one with D:A�GUS showing expression

on pedicel also showed GUS expression in lateral nectaries.b All the crc-1 transgenic plants with GUS expression in nectary anlagen

had GUS expression in valves.

4 of 12 The Plant Cell

Figure 2. GUS Expression Patterns Regulated by Conserved Regions of CRC Promoter in Arabidopsis Expressed in Wild-Type Landsberg erecta and

crc-1.

All whole-mount images ([A] to [H]) are Landsberg erecta (Ler). C:A�GUS ([A] to [C]); E:A�GUS ([D] to [F] and [I] to [O]); E:D:C:B:A�GUS ([G], [P],

and [Q]); pCRC 3.8kb�GUS ([H] and [R]). Region C drives GUS expression in developing carpels and the margins of sepals. Expression in carpels is

very high and lasts until late stages of flower development (after stage 15) (C). Though at early stages (stage 10) GUS is expressed throughout the

carpels (A), it is restricted to valves at stage 12 (B). Region E drives GUS expression in carpels and nectaries. In carpels, GUS expression is throughout

at the early stages and is restricted to the upper regions of the carpels later ([E] and [F]). Expression in nectaries starts before stage 10 (D) and lasts until

after stage 17 (F). GUS expression regulated by the fusion of all the conserved regions (G) is very similar to the regulation by the 3.8-kb-long 59 upstream

region of CRC (H). Expression in carpels is absent at stage 12 (insets in [G] and [H]) in contrast with the late expression observed with regions E and C

([C] and [F]).

(I) In a cross section of a crc-1 inflorescence, E:A�GUS is expressed in developing carpels and ovules. By stage 13, ovule expression disappears but

expression in the septum is detected (arrow).

(J) In a longitudinal section of a stage 9 flower, E:A�GUS is detected in the nectary anlagen (arrow).

(K) Longitudinal section of a stage 12 crc-1 flower in which E:A�GUS is expressed in cells that would form a nectary in the wild type (arrow).

(L) Longitudinal section of a crc-1 inflorescence meristem; E:A�GUS expression in carpels starts in stage 6, but it is very weak. Stage 7 flowers exhibit

very strong expression in carpels. Numbers refer to floral stages (Smyth et al., 1990).

(M) Longitudinal section of a Ler inflorescence meristem; E:A�GUS expression in the carpel is very similar to (L).

(N) Cross section of a Ler inflorescence; the E:A�GUS expression pattern in wild-type flowers is similar to that of crc-1 flowers.

(O) Longitudinal section of a stage 15 to 16 flower. Nectaries are demarcated with arrows.

(P) Cross section of a crc-1 inflorescence with E:D:C:B:A�GUS. Expression in septum and ovules disappears in stage 11 to 12 flowers (arrows),

whereas septum expression persists with E:A�GUS.

(Q) Stage 10 crc-1 flower; E:D:C:B:A�GUS expression in nectary anlagen is clearly visible (arrows).

(R) Stage 9 crc-1 flower with pCRC 3.8kb�GUS showing expression in the nectary anlagen (arrow).

Control of CRC Expression 5 of 12

not shown). In a lfy-6 background, CRC:GUS is expressed at the

abaxial base of pedicels, where small outgrowths resembling

nectaries arise (Baum et al., 2001). However, whereas LFY

influences the extent and pattern of CRC expression,

a 35S:LFY:VP16 transgene is not sufficient to activate CRC in

seedlings (data not shown), in contrast with the activation of AG

by 35S:LFY:VP16 (Parcy et al., 1998).

Expression of CRC in Plants with Altered MADS Box

Gene Activity

Previous genetic analyses of nectary development in floral

homeotic mutants suggests that although nectary development

depends on the presence of the third whorl in flowers of

Arabidopsis (Figures 4A and 4E), it is not generally affected by

homeotic changes of floral organs (Baum et al., 2001). The single

exception is that nectaries fail to develop in pi ag flowers, in

which both B and C class gene activities are missing (Bowman

et al., 1991), and this phenotypewas attributed to the lack of third

whorl development in this genotype (Baum et al., 2001). Muta-

tions in the SEP genes, which are required for B and C gene

activity (Pelaz et al., 2000, 2001; Honma and Goto, 2001), also

result in a failure in nectary development. The flowers of sep1

sep2 sep3 triple mutants consist entirely of sepals (Figure 4B),

similar to pi ag double mutants, but in contrast with pi ag flowers,

a third whorl appears to be present in sep1 sep2 sep3 flowers

(Pelaz et al., 2000). Despite possessing a third whorl, nectaries

are lacking in sep1 sep2 sep3 flowers (Figure 4H), suggesting

that the SEP genes are required for CRC activation in the third

whorl, consistent with the loss of CArG boxes resulting in a loss

of CRC expression. That ap3 and pi flowers lack a third whorl,

whereas sep1 sep2 sep3 flowers appear to have a third whorl,

suggests a SEP independent role for the B class genes in

patterning the floral ground plan.

When ap2 activity is compromised in an ag pi background,

nectary development is restored, suggesting thatAP2 negatively

regulates a redundant pathway for CRC activation (Figure 4J;

Baum et al., 2001). Because B and C gene products interact with

SEP gene products (Egea-Cortines et al., 1999; Honma and

Goto, 2001), other MADS box encoding genes are attractive

Figure 3. Complementation of the crc Mutant Phenotype.

As compared with wild-type fruits (A), the composite construct EDCBA�CRC fully complements the carpel defects of crc mutants (D). EA�CRC

partially complements crc carpel defects, with carpel fusion restored, but carpel (fruit) growth only partially restored (C). By contrast, CA�CRC fails to

complement crc (B), with the transgenic lines resembling crcmutants (E). Stamens are identified as medial (m) and lateral (l). Nectary glands are found

at the abaxial bases of all stamens in wild-type flowers (A). Both the EA�CRC and EDCBA�CRC transgenes are able to rescue lateral nectaries,

although medial nectaries are rarely present ([C] and [D]). By contrast, the CA�CRC transgene is unable to rescue nectary development (B) (compare

with crc-1 mutants in [E]).

Table 2. Site-Directed Mutagenesis of LFY Binding Sites in crc-1

Construct Frequency Expression Pattern

EL1:A�GUS 7/24 Upper valves

4/24a Nectary anlagen, ovule

CL2L3:A�GUS 7/18 Valves and partially on sepals

EL1:D:CL2L3:B:A�GUS 28/30 Valves

18/30a Nectary anlagen

a All the crc-1 transgenic plants with GUS expression in nectary anlagen

had GUS expression in valves.

6 of 12 The Plant Cell

candidates for the redundant factors. SHATTERPROOF1 (SHP1)

and SHP2 encode proteins similar to AG, are negatively regu-

lated by AP2, and are ectopically expressed in ap2 ag flowers

(Savidge et al., 1995; Flanagan et al., 1996; Pinyopich et al.,

2003). Flowers of ap2 pi ag shp1 shp2 plants consists entirely of

leaf-like organs produced froman indeterminate flowermeristem

(Figure 4C), and these flowers lack nectaries (Figure 4I), suggest-

ing that SHP proteins may interact with SEP proteins to activate

CRC in the third whorl of the flower. Whereas the SHP genes are

expressed in wild-type nectaries (Baum et al., 2001), loss of SHP

activity in shp1 shp2 flowers does not result in morphological

defects in nectary development (Figure 4F).

If SHP1 and SHP2 are acting with SEP to activate CRC in an

ap2 pi ag background, nectary development in a sep1 sep2 sep3

ap2 background is predicted to be absent. Thus, we examined

nectaries in this background. Because of the instability of the

sep2 allele (Pelaz et al., 2000), the interpretation of the carpelloid

organs produced in such flowers (Figure 4D) is ambiguous

because all quadruple mutant plants displayed some reversion

of the sep2mutant allele to the wild type at the molecular level as

assayed by PCR. However, in all flowers examined, no trace of

nectary development was observed (Figure 4G), suggesting that

the primary pathway negatively regulated byAP2 is one that acts

in conjunction with SEP proteins.

DISCUSSION

Modular Conservation of cis-Regulatory Elements

Functional analyses of the CRC promoter sequences demon-

strate that natural selection acts on regions directing proper gene

expression (modules; Arnone and Davidson, 1997) differently

from nonfunctional regions, and this conservation of modules is

recognizable in sequences from species within the same family,

Brassicaceae in this case (Wray et al., 2003). Each of five

modules conserved between Arabidopsis and Lepidium has

a discrete regulatory function, and only in combination do they

reflect the complete regulatory pattern of the endogenous CRC

promoter. Module E primarily directs CRC expression in nectar-

ies and carpels, and C acts redundantly with E to direct CRC

expression in valves of developing gynoecia. However, addi-

tional modules (B and D) are needed to repress expression in

specific tissues of the carpels and sepals. Thus, the regulation of

CRC expression is a combination of positive and negative

regulation acting through different modules of the promoter.

Five regions of the 59 upstream sequences of CRC are

conserved between Arabidopsis and Brassica, but each con-

served region does not precisely overlap with the regions

conserved between Arabidopsis and Lepidium. The nonoverlap-

ping conservation is likely because of differential selection

pressure acting upon the conserved regions after the separation

from the last common ancestor of the three species.Whether the

regions conserved in a nonoverlapping manner reflects func-

tional redundancy or is the result of selection is not clear until

those regions are functionally tested.

Two factors that could contribute to the relatively large size of

the conserved regions are the evolutionary distances between

species compared (e.g., linkage disequilibrium) and the com-

plexity of gene regulation.We compared sequences from closely

related species that have diverged less than ;10 to 14 million

years ago (Arabidopsis and Lepidium) and 20 million years ago

(Arabidopsis andBrassica) based on the fossil record of Rorippa,

and consistent with their phylogenetic relationships, within the

conserved regions, Arabidopsis CRC promoter sequences are

more similar to those of Lepidium than to those of Brassica (Mai,

1995; Koch et al., 2001a). Other studies comparing orthologous

regulatory sequences from Brassicaceae species facilitated

identification of conserved regulatory elements (Hill et al.,

1998; Koch et al., 2001b; Hong et al., 2003). Comparisons of

the AP3 and CHALCONE SYNTHASE promoters, consisting of

;500 bases 59 to the transcription start site, from 22 species

identified conserved elements, some of which have been exper-

imentally tested for functionality, amidst sequences that are

difficult or impossible to align (Hill et al., 1998; Tilly et al., 1998;

Koch et al., 2001b). By contrast, sequences of the second intron

of AG were unambiguously aligned over their entire length of

;3000 bases (Hong et al., 2003). Comparisons of the regulatory

sequences ofCRCwere similar to those of AP3 andCHALCONE

SYNTHESIS in that conserved sequences are flanked by se-

quences that cannot be aligned because of extensive diver-

gence.

Complexity of gene regulation may also be responsible for

maintaining large regions of sequence conservation, especially

within region E, which is almost 500-bp long and is responsible

for most of the nectary and effective carpel expression. Region E

is a combination of three highly conserved regions separated by

very short nonconserved sequences. When region E was dis-

sected further using the three conserved subregions individually

fused with a CaMV-TATA:GUS, we observed nectary expression

only when all three subregions are combined, suggesting pos-

sible interactions between the subregions (J.-Y. Lee, unpub-

lished data). However, because expression levels conferred by

region E are significantly decreased without region A, analysis of

the subregions of E combined with region A is required to verify

these results. Sequencing of CRC regulatory sequences from

additional Brassicaceae species could help discriminate be-

tween linkage disequilibrium and regulatory complexity as

causes for the large size of conserved regions observed in this

Table 3. Site-Directed Mutagenesis of CArG Boxes in crc-1

Construct Frequency Expression Pattern

EM1M2:A�GUS 3/22 Stigma

EM1M2:D:C:B:A�GUS 10/25 Valves

6/25a Nectary anlagen in late

stage (after stage 12)

flowers

EM1M2L1:D:CL2L3:B:A�GUS 10/35 Valves

7/35b Nectary anlagen in late

stage (after stage 12)

flowers

2/35c Base of pedicel

a Five of six had GUS expression in valves.b All seven had GUS expression in valves.c One had GUS expression in valves and nectary anlagen.

Control of CRC Expression 7 of 12

study. Analyses of potential binding sites of MADS box proteins

demonstrate their importance for activating CRC in nectaries

and carpels. However, conservation in the sequence immedi-

ately surrounding the CArG boxes is not as high as in the

remainder of region E. A similar pattern is observed in LFY

binding sites in the AG promoter (Hong et al., 2003). The lack of

sequence conservation may be because of these transcription

factors relying primarily on their specific binding sites and not on

surrounding sequence context.

Complementation of crcMutant Phenotypes

The ability of both the intact 3.8-kb CRC promoter and the

E:D:C:B:A composite promoter to complement the crc mutant

phenotype suggests that all necessary promoter elements are

found within conserved DNA sequences among species. Al-

though at a lower frequency, CRC regulated by region E alone

could partially complement both the carpel and nectary pheno-

types. Neither E:D:C:B:A nor E:A fully complemented the crc

nectary phenotype because medial nectaries were lacking and

stomata were reduced in number on the lateral nectaries. One

possible explanation is that additional sequences, perhaps those

immediately flanking the highly conserved regions, may be

required for fine-tuning quantitative or qualitative CRC expres-

sion. Surprisingly, despite region C driving high levels of expres-

sion in the carpel, it was unable to complement the crc carpel

phenotype, in contrast with E:A�CRC, which partially comple-

ments the crc carpel phenotype. Thus, a combination of se-

quences in C and E is required for proper spatial and temporal

regulation within the carpel, and these promoter elements are at

least partially redundant because both C and E are sufficient to

drive expression independently.

Regulation of CRC by Floral Genes in Arabidopsis

Genetic analysis of nectary development in floral homeotic

mutants demonstrated that nectaries can develop in the absence

of the activity of the ABC genes (Baum et al., 2001). Although

Figure 4. Nectary Phenotypes in Genotypes Lacking Combinations of MADS Box Gene Activity.

(A) to (D) Flower phenotypes.

(A) The wild type.

(B) sep1-1 sep2-1 sep3-2.

(C) ap2-2 pi-1 ag-1 shp1-1 shp2-1.

(D) ap2-2 sep1-1 sep2-1 sep3-2.

(E) to (J) Nectary phenotypes.

(E) The wild type. Stomata and cuticular patterning characteristic of nectary morphology are visible in the inset.

(F) shp1-1 shp2-1.

(G) ap2-2 sep1-1 sep2-1 sep3-2.

(H) sep1-1 sep2-1 sep3-2.

(I) ap2-2 pi-1 ag-1 shp1-1 shp2-1. No cuticular patterning characteristic of nectaries is visible on the pedicel (inset).

(J) ap2-2 pi-1 ag-1.

Nectaries are present in wild-type, shp1 shp2, and ap2 pi ag flowers (arrows) but are lacking in the other genotypes. The slender outgrowths in ap2 sep1

sep2 sep3 and ap2 pi ag shp1 shp2 flowers are stipules that develop at the base of the floral organs because of their leaf-like character.

8 of 12 The Plant Cell

nectaries are normally associated with stamens in wild-type

flowers, when the identity of the third whorl organs is altered to

carpels (as in ap3 and pi flowers) or petals (as in ag flowers),

nectaries still develop at the abaxial base of the third whorl

organs. Conversely, in 35S:PI 35S:AP3 ap2 flowers, in which

stamens occupy all floral whorls, nectaries are only found

associated with the third whorl stamens. In superman mutants

and in 35S:UFO flowers, multiple whorls of nectaries are asso-

ciated with the supernumerary whorls of stamens produced

interior to the third whorl. One interpretation is that nectary

development depends on the formation of the third whorl and is

independent of the development of the other floral organs (Baum

et al., 2001). However, the sizes and positions of nectaries are

affected by mutations in the ABC genes (Baum et al., 2001). For

example, in both B and C class mutants, the extent of nectary

gland development is reduced, and in both lfy and ufo single

mutants, nectaries were rarely found. Finally, in BC double

mutants (ag pi or ag ap3) no sign of nectary development is

observed, but nectary development is restored when AP2

activity is also compromised. Thus, whereas the ABC genes

are not absolutely required, these genes influence the extent of

nectary development.

Because both LFY and MADS box proteins are implicated in

the regulation of CRC, we searched for their respective binding

sites in theCRCpromoter. TwoCArGboxes and four LFYbinding

sites were found in the Arabidopsis CRC promoter. CArG boxes

were found in region E, and site-directed mutagenesis of the two

boxes dramatically disrupted transcriptional regulation by region

E, suggesting that a MADS box protein(s) is critical for the

transcriptional regulation of CRC. Based on the observation that

B and C class gene products interact with SEP proteins in the

specification of floral organ identity (Egea-Cortines et al., 1999;

Honma andGoto, 2001; Pelaz et al., 2001) and the lack of nectary

development in sep1 sep2 sep3 flowers, a compelling scenario is

that a complex including SEP proteins andC and/or BMADSbox

proteins activates CRC through binding of the CArG boxes in

region E. The lack of nectary formation in BCdoublemutants, but

their presence in B and C single mutants, would reflect redun-

dancy of these proteins in the complex with the SEP proteins.

The restoration of nectary development in ap2 pi ag triple

mutants suggests that AP2 represses another redundant path-

way or factor(s) that can act with the SEP proteins to activate

CRC.SHP1 andSHP2 are obvious candidates representing such

a redundant pathway because they are negatively regulated by

AP2, encode proteins similar to AG, and are ectopically ex-

pressed in an ap2 ag background (Savidge et al., 1995; Flanagan

et al., 1996; Pinyopich et al., 2003). In this scenario, in an ap2 pi

ag background a complex of SHP and SEP proteins activates

CRC in the nectaries. Consistent with this hypothesis, ap2 pi ag

shp1 shp2 flowers lack nectaries. That sep123 ap2 flowers also

lack nectaries indicates that the primary pathway negatively

regulated by AP2 is through repression of genes encodingMADS

box proteins. Thus, in wild-type flowers, a complex of SEPs þ B

and/or C would activate CRC, whereas in an ABC triple mutant

SEPs þ SHPs would activate CRC. The control of CRC expres-

sion by complexes of MADS box genes represents a mechanism

by which CRC is activated specifically in the flowers. However,

because the combination of SEPs þ B and/or C would have the

potential to activateCRC throughout the inner three whorls of the

flower, these complexes must act in conjunction with other,

presently unidentified, spatial regulators such that activation is

restricted to the nectary anlagen and specific regions of the

carpels. The spatial regulators must also act through promoter

sequences contained within regions C and E (Figure 5).

Site-directed mutagenesis of putative LFY binding sites did

not generate any dramatic alteration of gene activity in vivo, and

LFY:VP16 was insufficient to activate CRC, despite the genetic

evidence for LFY-mediated regulation of CRC. This is reminis-

cent of LFY-mediated activation of AP3, where putative LFY

binding sites in the AP3 promoter are dispensable (Lamb et al.,

2002). Thus, although LFY activity is critical for proper regulation

of CRC, its action is unlikely to be direct, perhaps acting through

LFY-mediated regulation of MADS box genes (Parcy et al., 1998;

Schmid et al., 2003).

As with other genes whose regulation is complex both in terms

of spatial and temporal patterns, several modules act positively

and negatively in a combinatorial fashion to control CRC ex-

pression, as summarized in Figure 5. The SEP þ B/C MADS box

genes activateCRC in the flower, but otherwhorl-specific factors

are required to restrict CRC expression to the nectaries and

specific tissues of the gynoecium. Finally, that the E module is

not easily dissected suggests that thewhorl-specific factorsmay

be required to interact directly with the more general flower

activation factors to modulate CRC expression.

Figure 5. Model of CRC Regulation in Arabidopsis.

Based on previous genetic analyses (Baum et al., 2001), B (AP3 and PI)

and C (AG) genes are regulating nectary development in combination.

Because sep1/2/3 does not have any nectaries, SEPs are likely interact-

ing with B and/or C proteins directly to activate CRC, whose protein

interaction was shown by Honma and Goto (2001). In wild-type flowers,

SHP1 and 2 are expressed in developing carpels not in the third whorl;

however, in ap2 mutants, they are activated in all the whorls (Savidge

et al., 1995; Flanagan et al., 1996; Pinyopich et al., 2003). It is likely that

SHP1/2 might be redundantly interacting with SEP genes (Favaro et al.,

2003) to activate CRC (shown as dashed line). To restrict the CRC

expression at the base of stamens, there should be other unidentified

floral factors (?) that fine tune the gene regulation. LFY may activate the

CRC expression within flowers by activating BC genes directly and SEPs

directly or indirectly (Schmid et al., 2003).

Control of CRC Expression 9 of 12

The regulation of CRC has implications concerning nectary

evolution in flowering plants. Nectaries in basal angiosperms are

usually associated with the perianth, whose structure is not as

distinct or elaborate as in core eudicots and monocots. By

contrast, in core eudicots, nectaries are usually located near or

on the reproductive organs (Brown, 1938; Fahn, 1953; Endress,

2001). Expression analysis of CRC in developing nectaries of

several speciesof eudicots suggests thatCRCmight beageneral

regulator for nectary development in core eudicot lineages (J.-Y.

Lee, unpublished data). The functional significance of CArG

boxes for the regulation of CRC suggests that the establishment

of the regulatory pathway betweenMADS box proteins andCRC

may have facilitated restricting nectaries to the reproductive

whorls of flowers.

METHODS

Isolation and Sequence Analysis of CRCOrthologs from Other

Brassicaceae Species

The CRC ortholog from Lepidium africanum was cloned by screening

a cosmid library constructed using size-selected genomic DNA. The 4958

bases (6090 bases total, including a portion of the coding sequence)

containing the L. africanum CRC 59 upstream region were subcloned into

pBluescript SKþ from the cosmid and sequenced. The CRC ortholog

from Brassica oleracea was cloned by PCR amplification of CRC coding

sequences from a BAC identified in a BAC library generated by C. Quiros

at the University of California, Davis. The 4540 bases (5462 bases total,

including a portion of the coding sequence) of the B. oleracea CRC 59

upstream region were sequenced. Four domains (B, C, D, and E)

conserved between Arabidopsis and Lepidium CRC 59 upstream se-

quences were identified with a local alignment tool, bl2seq (http://

www.ncbi.nlm.nih.gov/blast/bl2seq/bl2.html; Tatusova and Madden,

1999). They are indicated with black bars on the top of the plots in Figure

1, and they were used for expression analysis. Including Brassica

sequences, three sequences were aligned with AVID, a global alignment

program (Bray et al., 2003), and conserved regions shared by three

species were detected using mVISTA with the parameter of 75% of

identity and a 50 base sliding window (Dubchak et al., 2000; Mayor et al.,

2000), which fits well with the local alignment result between Arabidopsis

and Lepidium (Figure 1). Three-way species comparison used in mVISTA

identifies conserved domains by choosing regions conserved in all three

species using intersection/union analysis (Dubchak et al., 2000). The

conserved regions of 300 to 550 bp were conspicuous relative to

surrounding sequence because in most cases the adjacent sequences

were impossible to align with any degree of confidence (Figure 1).

Transgene Construction for Promoter Analysis

The five domains (A, B, C, D, and E) conserved between the 59 upstream

sequences of Arabidopsis thaliana CRC and L. africanum CRC were

isolated by PCR and subcloned such that they were 59 to the chimeric

transcription factor, LhG4 (Moore et al., 1998), with a 39 octopine

synthase terminator after the LhG4 coding sequences. Six constructs

were generated: A:LhG4, B:A:LhG4, C:A:LhG4, D:A:LhG4, E:A:LhG4, and

E:D:C:B:A:LhG4. These constructs were subcloned into the binary vector

pMLBART and transformed into Agrobacterium tumefaciens strain ASE

by electroporation. Each construct was transformed into wild-type Ler or

crc-1 mutants containing the transgene 2OP:GUS by floral dipping

(Weigel and Glazebrook, 2002), and the transgenic plants were selected

with BASTA.

Site-directed mutagenesis was accomplished using a PCR-based

method. Primers having mutant sequences were used to amplify (Pfu

polymerase) the CRC promoter region of interest. The template plasmid

was removed by treating the PCR mixture with the restriction enzyme

DpnI, then the mixture was transformed into Escherichia coli. The

mutagenesis was confirmed by sequencing.

GUS Staining

Inflorescences were fixed in 90% ice-cold acetone for 20 min and rinsed

twice with a GUS working solution [25 mM phosphate buffer, pH 7.0,

1.25 mM K3Fe(CN)6, 1.25 mM K4Fe(CN)6, 0.25% Triton X-100, and

0.25 mM EDTA] for 20 min each time. After rinsing, tissue was incu-

bated with 5-bromo-4-chloro-3-indoyl b-D-glucuronide cyclohexylamine

salt, added to GUSworking solution to the final concentration of 1.25mg/

mL, at 378C overnight. The reaction was terminated and tissue was

cleared in 70% ethanol added fresh once a day for a week.

Microscopy

Inflorescences incubated with X-Gluc were fixed in formaldehyde-acetic

acid (3.7% formaldehyde, 5% acetic acid, and 50% ethanol) for 2 h and

dehydrated through an ethanol series and embedded in paraffin. Inflo-

rescences were sectioned at 8 mm of thickness. Sectioned tissue was

viewed using dark-field optics. For scanning electron microscopy, tissue

was fixed overnight with 3% glutaraldehyde, phosphate buffered to pH 7,

followed by a second overnight fixation in 0.5% osmium tetraoxide.

Tissue was dehydrated in ethanol and critical point dried. After sputter

coating with gold/palladium, tissue was observed on a Hitachi S-3500N

scanning electron microscope (Tokyo, Japan).

Sequence data from this article have been deposited with the EMBL/

GenBank data libraries under accession numbers AY703987 for 6090

bases of L. africanum CRC 59 upstream and partial coding regions and

AY703986 for 5462 bases of B. oleracea CRC 59 upstream and partial

coding regions.

ACKNOWLEDGMENTS

We thank Carlos Quiros for providing us with B. oleracea BAC library,

Adrienne Roeder, Gary Ditta, and Marty Yanofsky for seed stocks, and

Nathaniel Hawker and Sandy Floyd for critical reading of this manu-

script. This work was made possible with funding from the Department

of Energy, Division of Biosciences (DE-FG03-97ER20272) to J.L.B.

Received August 5, 2004; accepted October 20, 2004.

REFERENCES

Alvarez, J., and Smyth, D.R. (1999). CRABS CLAW and SPATULA, two

Arabidopsis genes that control carpel development in parallel with

AGAMOUS. Development 126, 2377–2386.

Arnone, M.I., and Davidson, E.H. (1997). The hardwiring of develop-

ment: Organization and function of genomic regulatory systems.

Development 124, 1851–1864.

Baum, S.F., Eshed, Y., and Bowman, J.L. (2001). The Arabidopsis

nectary is an ABC-independent floral structure. Development 128,

4657–4667.

Bowman, J.L., and Smyth, D.R. (1998). Patterns of petal and stamen

reduction in Australian species of Lepidium L. (Brassicaceae). Int. J.

Plant Sci. 159, 65–74.

10 of 12 The Plant Cell

Bowman, J.L., and Smyth, D.R. (1999). CRABS CLAW, a gene that

regulates carpel and nectary development in Arabidopsis, encodes

a novel protein with zinc finger and helix-loop-helix domains. De-

velopment 126, 2387–2396.

Bowman, J.L., Smyth, D.R., and Meyerowitz, E.M. (1991). Genetic

interactions among floral homeotic genes of Arabidopsis. Develop-

ment 112, 1–20.

Bray, N., Dubchak, I., and Pachter, L. (2003). AVID: A global alignment

program. Genome Res. 13, 97–102.

Brown, W.H. (1938). The bearing of nectaries on the phylogeny of

flowering plants. Proc. Am. Philos. Soc. 79, 549–595.

Busch, M.A., Bomblies, K., and Weigel, D. (1999). Activation of a floral

homeotic gene in Arabidopsis. Science 285, 585–587.

Colinas, J., Birnbaum, K., and Benfey, P.N. (2002). Using cauliflower

to find conserved non-coding regions in Arabidopsis. Plant Physiol.

129, 451–454.

Crepet, W.L., and Friis, E.M. (1987). The evolution of insect pollination

in angiosperms. In The Origin of Angiosperms and Their Biological

Consequences, E.M. Friis, W.G. Chaloner, and P.R. Crane, eds

(Cambridge, UK: Cambridge University Press), pp. 181–202.

Darwin, F. (1877). On the nectar glands of the common brakeferns. Bot.

J. Linn. Soc. 15, 407–409.

Davis, A.R., Pylatuik, J.D., Paradis, J.C., and Low, N.H. (1998).

Nectar-carbohydrate production and composition vary in relation to

nectary anatomy and location within individual flowers of several

species of Brassicaceae. Planta 205, 305–318.

Dolan, J.W., and Fields, S. (1991). Cell-type-specific transcription in

yeast. Biochim. Biophys. Acta 1088, 155–169.

Dubchak, I., Brudno, M., Loots, G.G., Mayor, C., Pachter, L., Rubin,

E.M., and Frazer, K.A. (2000). Active conservation of noncoding

sequences revealed by 3-way species comparisons. Genome Res.

10, 1304–1306.

Duret, L., and Bucher, P. (1997). Searching for regulatory elements in

human noncoding sequences. Curr. Opin. Struct. Biol. 7, 399–406.

Egea-Cortines, M., Saedler, H., and Sommer, H. (1999). Ternary

complex formation between the MADS-box proteins SQUAMOSA,

DEFICIENS and GLOBOSA is involved in the control of floral archi-

tecture in Antirrhinum majus. EMBO J. 18, 5370–5379.

Endress, P.K. (2001). The flowers in extant basal angiosperms and

inferences on ancestral flowers. Int. J. Plant Sci. 162, 1111–1140.

Eriksson, O., and Bremer, B. (1992). Pollination systems, dispersal

modes, life forms, and diversification rates in angiosperm families.

Evolution 46, 258–266.

Eshed, Y., Baum, S.F., and Bowman, J.L. (1999). Distinct mechanisms

promote polarity establishment in carpels of Arabidopsis. Cell 99,

199–209.

Fahn, A. (1953). The topography of the nectary in the flower and its

phylogenetical trend. Phytomorphology 3, 424–426.

Fahn, A. (1979). Secretory Tissues in Plants. (New York: Academic

Press).

Favaro, R., Pinyopich, A., Battaglia, R., Kooiker, M., Borghi, L., Ditta,

G., Yanofsky, M.F., Kater, M.M., and Colombo, L. (2003). MADS-

box protein complexes control carpel and ovule development in

Arabidopsis. Plant Cell 15, 2603–2611.

Flanagan, C.A., Hu, Y., and Ma, H. (1996). Specific expression of the

AGL1 MADS-box gene suggests regulatory functions in Arabidopsis

gynoecium and ovule development. Plant J. 10, 343–353.

Hill, T.A., Day, C.D., Zondlo, S.C., Thackeray, A.G., and Irish, V.F.

(1998). Discrete spatial and temporal cis-acting elements regulate

transcription of the Arabidopsis floral homeotic gene APETALA3.

Development 125, 1711–1721.

Hong, R.L., Hamaguchi, L., Busch, M.A., and Weigel, D. (2003).

Regulatory elements of the floral homeotic gene AGAMOUS identified

by phylogenetic footprinting and shadowing. Plant Cell 15, 1296–

1309.

Honma, T., and Goto, K. (2001). Complexes of MADS-box proteins are

sufficient to convert leaves into floral organs. Nature 409, 525–529.

Koch, M., Haubold, B., and Mitchell-Olds, T. (2001a). Molecular

systematics of the Brassicaceae: Evidence from coding plastidic

matK and nuclear Chs sequences. Am. J. Bot. 88, 534–544.

Koch, M.A., Weisshaar, B., Kroymann, J., Haubold, B., and Mitchell-

Olds, T. (2001b). Comparative genomics and regulatory evolution:

Conservation and function of the Chs and Apetala3 promoters. Mol.

Biol. Evol. 18, 1882–1891.

Lamb, R.S., Hill, T.A., Tan, Q.K., and Irish, V.F. (2002). Regulation of

APETALA3 floral homeotic gene expression by meristem identity

genes. Development 129, 2079–2086.

Mai, D.H. (1995). Tertiare Vegetationsgeschichte Europas. (Stuttgart,

Germany: Fischer).

Mayor, C., Brudno, M., Schwartz, J.R., Poliakov, A., Rubin, E.M.,

Frazer, K.A., Pachter, L.S., and Dubchak, I. (2000). VISTA: Visual-

izing global DNA sequence alignments of arbitrary length. Bioinfor-

matics 16, 1046–1047.

Meeuse, A.D.J. (1978). Nectarial secretion, floral evolution, and the

pollination syndrome in early angiosperms. Proc. K. Ned. Akad. Wet.

C. 81, 300–326.

Moore, I., Galweiler, L., Grosskopf, D., Schell, J., and Klaus, P.

(1998). A transcription activation system for regulated gene ex-

pression in transgenic plants. Proc. Natl. Acad. Sci. USA 95,

376–381.

Parcy, F., Nilsson, O., Busch, M.A., Lee, I., and Weigel, D. (1998). A

genetic framework for floral patterning. Nature 395, 561–566.

Pelaz, S., Ditta, G.S., Baumann, E., Wisman, E., and Yanofsky, M.F.

(2000). B and C floral organ identity functions require SEPALLATA

MADS-box genes. Nature 405, 200–203.

Pelaz, S., Tapia-Lopez, R., Alvarez-Buylla, E.R., and Yanofsky, M.F.

(2001). Conversion of leaves into petals in Arabidopsis. Curr. Biol. 11,

182–184.

Pellmyr, O. (1992). Evolution of insect pollination and angiosperm

diversification. Trends Ecol. Evol. 7, 46–49.

Pinyopich, A., Ditta, G.S., Savidge, B., Liljegren, S.J., Baumann, E.,

Wisman, E., and Yanofsky, M.F. (2003). Assessing the redundancy

of MADS-box genes during carpel and ovule development. Nature

424, 85–88.

Polowick, P.L., and Sawhney, V.K. (1986). A scanning electron

microscope study on the initiation and development of floral organs

of Brassica napus (cv Westar). Am. J. Bot. 73, 254–263.

Porsch, O. (1910). Ephedra campylopoda C. A. Mey, eine entomophyle

Gymnosperme. Ber. Deutsch. Bot. Ges. 31, 580–590.

Savidge, B., Rounsley, S.D., and Yanofsky, M. (1995). Temporal

relationship between the transcription of two Arabidopsis MADS box

genes and the floral organ identity genes. Plant Cell 7, 721–733.

Sawa, S., Watanabe, K., Goto, K., Kanaya, E., Morita, E.H., and

Okada, K. (1999). FILAMENTOUS FLOWER, a meristem and organ

identity gene of Arabidopsis, encodes a protein with a zinc finger and

HMG-related domains. Genes Dev. 13, 1079–1088.

Schmid, M., Uhlenhaut, N.H., Godard, F., Demar, M., Bressan, R.,

Weigel, D., and Lohmann, J.U. (2003). Dissection of floral induction

pathways using global expression analysis. Development 130, 6001–

6012.

Siegfried, K.R., Eshed, Y., Baum, S.F., Otsuga, D., Drews, G.N., and

Bowman, J.L. (1999). Members of the YABBY gene family specify

abaxial cell fate in Arabidopsis. Development 126, 4117–4128.

Smale, S.T. (2001). Core promoters: Active contributors to combinato-

rial gene regulation. Genes Dev. 15, 2503–2508.

Control of CRC Expression 11 of 12

Smyth, D.R., Bowman, J.L., and Meyerowitz, E.M. (1990). Early flower

development in Arabidopsis. Plant Cell 2, 755–768.

Tatusova, T.A., and Madden, T.L. (1999). Blast 2 sequences, a new

tool for comparing protein and nucleotide sequences. FEMS Micro-

biol. Lett. 174, 247–250.

Tautz, D. (2000). Evolution of transcriptional regulation. Curr. Opin.

Genet. Dev. 10, 575–579.

Tilly, J.J., Allen, D.W., and Jack, T. (1998). The CArG boxes in the

promoter of the Arabidopsis floral organ identity gene APETALA3

mediate diverse regulatory effects. Development 125, 1647–1657.

Treisman, R. (1992). The serum response element. Trends Biochem.

Sci. 17, 423–426.

Villanueva, J.M., Broadhvest, J., Hauser, B.A.,Meister, R.J., Schneitz,

K., and Gasser, C.S. (1999). INNER NO OUTER regulates abaxial-

adaxial patterning in Arabidopsis ovules. Genes Dev. 13, 3160–3169.

Weigel, D., and Glazebrook, J. (2002). Arabidopsis: A Laboratory

Manual. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory

Press).

Wray, G.A., Hahn, M.W., Abouheif, E., Balhoff, J.P., Pizer, M.,

Rockman, M.W., and Romano, L.A. (2003). The evolution of tran-

scriptional regulation of eukaryotes. Mol. Biol. Evol. 20, 1377–1419.

Yamaguchi, T., Nagasawa, N., Kawasaki, S., Matsuoka, M., Nagato,

Y., and Hirano, H.-Y. (2004). The YABBY gene DROOPING LEAF

regulates carpel specification and midrib development in Oryza sativa.

Plant Cell 16, 500–509.

12 of 12 The Plant Cell

DOI 10.1105/tpc.104.026666; originally published online December 14, 2004;Plant Cell

Ji-Young Lee, Stuart F. Baum, John Alvarez, Amita Patel, Daniel H. Chitwood and John L. Bowman in the Nectaries and Carpels of ArabidopsisCRABS CLAWActivation of

 This information is current as of October 18, 2020

 

Supplemental Data /content/suppl/2004/12/17/tpc.104.026666.DC1.html

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists