Logo of plntcellLink to Publisher's site
Plant Cell. 2008 Apr; 20(4): 901–919.
PMCID: PMC2390727

REBELOTE, SQUINT, and ULTRAPETALA1 Function Redundantly in the Temporal Regulation of Floral Meristem Termination in Arabidopsis thaliana[W]


In Arabidopsis thaliana, flowers are determinate, showing a fixed number of whorls. Here, we report on three independent genes, a novel gene REBELOTE (RBL; protein of unknown function), SQUINT (SQN; a cyclophilin), and ULTRAPETALA1 (ULT1; a putative transcription factor) that redundantly influence floral meristem (FM) termination. Their mutations, combined with each other or with crabs claw, the genetic background in which they were isolated, trigger a strong FM indeterminacy with reiterations of extra floral whorls in the center of the flower. The range of phenotypes suggests that, in Arabidopsis, FM termination is initiated from stages 3 to 4 onwards and needs to be maintained through stage 6 and beyond, and that RBL, SQN, and ULT1 are required for this continuous regulation. We show that mutant phenotypes result from a decrease of AGAMOUS (AG) expression in an inner 4th whorl subdomain. However, the defect of AG activity alone does not explain all reported phenotypes, and our genetic data suggest that RBL, SQN, and, to a lesser extent, ULT1 also influence SUPERMAN activity. Finally, from all the molecular and genetic data presented, we argue that these genes contribute to the more stable and uniform development of flowers, termed floral developmental homeostasis.


Flowers are the reproductive structures that characterize angiosperms. They contain the male and female reproductive organs (stamens and carpels), often surrounded by outer sterile structures (sepals and petals). Floral organ identity is controlled by a set of genes that apparently act according to the quartet model (Ferrario et al., 2004; Jack, 2004; Angenent et al., 2005). In Arabidopsis thaliana, this model postulates that, together with the E class genes SEPALLATA1 (SEP1) to SEP4, the A class genes APETALA1 (AP1) and AP2 control sepal and petal identities, the B class genes AP3 and PISTILLATA (PI) control petal and stamen identities, and the C class gene AGAMOUS (AG) controls stamen and carpel identities.

In the eudicot plant Arabidopsis, the flower is determinate: floral organs are in whorls, with a fixed number of whorls and parts per whorl, and internodes between whorls that do not elongate. By contrast, in many angiosperms outside the core eudicots, flowers are somewhat indeterminate and display morphological variations, such as a variable number of whorls and parts per whorl, an extensive elongation of the floral axis, or a spiral phyllotaxy. It is thought that variations of floral axis determinacy and compression (with a decrease of floral organ number) are two major evolutionary innovations that produced compact angiosperm flowers (Baum and Hileman, 2006) and that these two evolutionary innovations are manifestations of an increased floral developmental homeostasis (i.e., the tendency to produce a fixed number of structures in stereotypical positions; Frohlich, 2006). Some angiosperms would have retained the ancestral condition of weak developmental homeostasis, whereas core eudicots, including Arabidopsis, would have evolved strong floral homeostasis. However, underlying cellular and developmental mechanisms remain elusive and, except for AG and WUSCHEL (WUS) (see below), little is known about factors involved in these processes. A better understanding of these processes could result from the isolation of new regulatory factors from model plants.

Flowers develop from flower meristems (FMs), which in turn are produced at the flank of inflorescence shoot apical meristems (iSAMs; Blazquez et al., 2006). Therefore, the FM initially inherits the potential of indeterminacy from the iSAM, manifested in the maintenance of a stem cell population located in the center of the meristem. This maintenance is based on a feedback loop involving WUS and CLAVATA1 (CLV1), CLV2, and CLV3 genes, the mechanistic nature of which is unknown (Carles and Fletcher, 2003; Doerner, 2006). WUS, expressed in the deeper layers of meristems (Mayer et al., 1998), maintains CLV3 expression in stem cells, located in the upper two layers of the central zone. In turn, CLV genes inhibit WUS and thus regulate the size of the stem cell population (Fletcher et al., 1999; Brand et al., 2000; Schoof et al., 2000). FMs become determinate after the initiation of carpel primordia, but determinacy is likely initiated earlier since WUS expression starts declining from stage 3, prior to the initiation of the carpels (Mayer et al., 1998; stages are all defined according to Smyth et al., 1990). AG is of particular importance in specifying FM determinacy since it is part of a second feedback loop, set up in the FM, which is superimposed on the CLV/WUS feedback loop. This second feedback loop starts with the activation of AG expression by WUS together with LEAFY at stage 3 of floral development (Parcy et al., 1998; Busch et al., 1999; Lenhard et al., 2001; Lohmann et al., 2001). In turn, AG turns WUS expression off, likely indirectly (Sieburth et al., 1998), resulting in the arrest of stem cell maintenance in the center of the FM. Mutations in AG and CLV genes or ectopic expression of WUS beyond stage 6 are sufficient to induce floral meristem indeterminacy (Bowman et al., 1989; Clark et al., 1993; Shannon and Meeks-Wagner, 1993; Clark et al., 1995; Lenhard et al., 2001; Lohmann et al., 2001).

Interestingly, CLV and AG illustrate different levels of the control of FM size and determinacy. In a clv null allele, the FM is significantly enlarged. It produces more primary floral organs in all whorls than the wild-type plants and one or two additional whorls of carpels in the center of the flower, inside the primary carpels (Clark et al., 1993; Dievart et al., 2003). The increase of FM size results from an enlarged WUS expression domain and supernumerary whorls of carpels from delayed WUS extinction (Schoof et al., 2000). In ag mutants, the FM is not enlarged. Organ numbers per whorl are not increased, but a large number of extra whorls develop from the center of flower, resulting in a flower-within-a-flower phenotype, or Russian doll (Bowman et al., 1989; Sieburth et al., 1995). Accordingly, WUS expression remains spatially restricted but is not switched off (Lenhard et al., 2001; Lohmann et al., 2001). The temporal regulation of WUS extinction, which prevents the development of supernumerary whorls in the center of the FM, is here defined as FM termination and will be discussed in this article.

Other regulators also regulate FM termination. ULTRAPETALA1 (ULT1) was shown to prevent the formation of extra primary floral organs and of supernumerary carpel whorls (Fletcher, 2001; Carles et al., 2005). Reported data suggest that ULT1 operates in the AG pathway. SUPERMAN (SUP), whose mutations result in an increase of primary stamen and carpel numbers, depending on the allele (Schultz et al., 1991; Bowman et al., 1992; Gaiser et al., 1995), also affects FM termination. AG and SUP act synergistically since the double mutant consists of reiterations of petals produced by an indeterminate and fasciated FM (Schultz et al., 1991; Bowman et al., 1992), but the interconnections between AG and SUP functions remain an enigma (Sakai et al., 2000; Gomez-Mena et al., 2005; Sablowski, 2007). It is postulated that SUP controls FM termination by preventing B class gene expression in the 4th whorl. This is corroborated by the fact that, in Arabidopsis and Antirrhinum majus, B class mutants are overdeterminate and that flowers ectopically expressing B class genes are partially indeterminate (Hill and Lord, 1989; Sommer et al., 1990; Schultz et al., 1991; Bowman et al., 1992; Trobner et al., 1992; Krizek and Meyerowitz, 1996; Sakai et al., 2000). Recently, miR172-mediated regulation of AP2 was also shown to be crucial for floral termination (Chen, 2004; Zhao et al., 2007). Ectopic expression of miR172-resistant AP2 triggers FM indeterminacy with flowers producing an indeterminate number of either petals or stamens. In the transgenic lines producing supernumerary stamens, the B class gene PI is ectopically expressed in the center of the FM and WUS expression is prolonged (Zhao et al., 2007). The authors proposed that this miR172-dependant AP2 regulatory pathway acts through both the AG and CLV pathways.

Examples reported above show that the manifestations of FM termination defects are diverse but can be distinguished as defects that take place before primary carpel development (corresponding to early events of FM termination; e.g., miR172-resistant AP2 and sup phenotypes) or defects that take place after (corresponding to late events of FM termination; e.g., ag and some aspects of clv and ult1 phenotypes). However, since these two types of indeterminacy have not been observed in the same mutant backgrounds, one can wonder whether they reflect a continuous regulation of FM termination.

CRABS CLAW (CRC) is the founding member of the YABBY gene family (Bowman and Smyth, 1999). Its expression is largely limited to carpels and nectaries (Bowman and Smyth, 1999). In crc loss-of-function mutants, carpels are shorter, unfused at their top (the so-called crabs claw phenotype; Alvarez and Smyth, 1999), and display polarity defects (Eshed et al., 1999), while nectaries fail to develop (Bowman and Smyth, 1999). However, the development of supernumerary whorls inside primary carpels observed in crc-1 spatula-2 (spt-2) and crc-1 ag-1/+ double mutants suggested that CRC could also influence FM termination (Alvarez and Smyth, 1999). Based of this observation, we performed a modifier mutagenesis screen in the crc-1 background (Bowman et al., 2001), which allowed us to isolate and analyze three independent factors that act redundantly with CRC and with each other in controlling FM termination. From double and triple mutant combinations, we were able to define two major categories of phenotypes. The first shows a recurrent reset of the FM with the production of extra floral organs inside the gynoecium, corresponding to defects in late events of FM termination, and the second displays a large increase in numbers of primary floral organ whorls, particularly stamens, corresponding to defects in early events of FM termination.


crc-1 Modifiers Reveal a Range of Floral Indeterminacy Syndromes

T-DNA and ethyl methanesulfonate (EMS) mutagenesis experiments were performed in the crc-1 background to screen for modifiers of the crc phenotype (Figure 1). The crc-1 mutation limits style development and disrupts distal carpel fusion, (Figure 1B, left) but very rarely affects FM determinacy (Alvarez and Smyth, 1999). Upon screening several thousand M2 (EMS mutagenesis) and T2 (T-DNA mutagenesis) families, three independent mutants of similar phenotype were recovered. In a crc background, each triggers FM indeterminacy associated with reiterated whorls produced internal to the primary carpels (Figures 1B to 1H). A detailed analysis of the three double mutants was then performed. The names of alleles, rebelote-1 (rbl-1; “once again” in French), squint-4 (sqn-4), and ultrapetala1-4 (ult1-4), are based on the cloning of the genes, which is described below (Figure 2).

Figure 1.
Phenotypes of the crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 Double Mutants.
Figure 2.
Structures of the RBL, SQN, and ULT1 Genes and Mapping of the Allelic Mutations.

All three double mutants develop shorter gynoecia than crc-1 (Figure 1B) and reiterate extra whorls of stamens, carpels, and occasionally petals in the center of the flower (Figures 1C to 1E and 1G). Such reiterations always preserve a wild-type whorl organization (i.e., stamens-carpels or petals-stamens-carpels). Nevertheless, differences exist between the three mutants. Phenotypic scoring (Table 1) shows that 100% of crc-1 rbl-1 flowers are indeterminate, whereas 86% of crc-1 sqn-4 flowers and 83% of crc-1 ult1-4 flowers produce extra whorls. In addition, among the crc-1 rbl-1 indeterminate flowers, 91% break through the gynoecium and grow outside the silique (Figures 1C and 1D). Such elongated axes are observed in ~25% of crc-1 sqn-4 and more rarely in crc-1 ult1-4 flowers. Strong indeterminacy phenotypes show new floral organs developing even while the rest of the plant is senescent (Figure 1D), indicating a permanent loss of determinacy. In cases of less-pronounced indeterminacy, the gynoecium morphology is usually similar to crc-1, except that there is a mild increase of carpel number (Table 1), with ectopic stamens and carpels developing inside the pistil (Figure 1G). Despite a slight variability from one experiment to another, the same tendency in the strength of the phenotype (rbl-1 > sqn-4 > ult1-4) is always observed between the mutations in combination with crc-1. Whatever the strength of the phenotype, internal organs produced inside the ovary greatly compromise fertility. In addition to the indeterminacy observed at the FM center, the crc-1 ult1-4 double mutant develops additional sepals and petals on the two outermost whorls similar to ult1 single mutants (Figure 1F, Table 1), whereas neither crc-1 rbl-1 nor crc-1 sqn-4 display such a phenotype.

Table 1.
Floral Organ Number in crc-1, rbl-1, sqn-4, and ult1-4 Single Mutants and in Double Mutants Generated in Pairwise Combinations

Observations of gynoecium development show that the ectopic floral organs produced in double mutants arise at the base of and adaxial to the primary carpels (Figure 4). At floral stage 6, while carpel primordia appear tightly joined in wild-type flowers (Figure 4A), they are separated by a cell dome, similar to a meristem, in crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 double mutants (Figures 4B to 4D). During later flower development, the ectopic organs or structures stay firmly attached to an elongating axis at the base of the silique (Figure 1H).

Figure 4.
Analysis of WUS and CLV3 Expression Patterns in crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 Double Mutant Backgrounds and Phenotypes of the wus-1 rbl-1, wus-1 sqn-4, and wus-1 ult1-4 Double Mutants.

In summary, in combination with crc-1, the three mutations make the FM indeterminate, with ectopic stamens and carpels developing from the very center of the FM, between primary carpels, and on an elongated axis. RBL, SQN, and ULT1 therefore influence late events (stage 6 and beyond) of FM termination and are collectively essential for normal floral axis compression. The reiteration of stamens and petals implies a reactivation of B class genes after carpel formation.

Cloning of crc Modifier Genes

The molecular identity of the two EMS-induced crc modifiers was determined via map-based cloning. The mutation in RBL was mapped to At3g55510. rbl-1 is caused by the replacement of an Ala residue by a Thr, eight amino acids downstream of the first Met (Figure 2A). Two additional alleles, rbl-2 (FLAG 299B04; Institut National de la Recherche Agronomique, Versailles) and rbl-3 (SALK 059.267; Figure 2A), were crossed to crc-1. The resulting crc-1 rbl-2 double mutants developed indeterminate flowers (see Supplemental Figure 1A online), confirming RBL as At3g55510. rbl-3 causes embryonic lethality with an arrest of embryo development at the heart stage (see Supplemental Figure 2 online). This suggests that rbl-1 and rbl-2 are partial loss-of-function alleles, and rbl-1 reveals only one of RBL's functions. However, indeterminate flowers were also observed in crc-1 rbl-3/+ plants, implying a dose-dependent effect of the rbl-3 mutation in a crc background. Database searches revealed one sequence similar to RBL in Arabidopsis: At2g18220. RBL and At2g18220 share 26% identity and 39% similarity over the full length of the proteins (see Supplemental Figure 3 online). Sequence analyses of RBL and At2g18220 recognized the conserved NOC2 domain, also known as UPF0120 domain (Figure 2B), located between amino acids 256 and 582 in RBL. This domain identifies an uncharacterized protein family and its name is based on yeast Noc2p (Milkereit et al., 2001), the founding member of the family. RBL and At2g18220 share 35% identity and 56% similarity over the NOC2 domain, and RBL and At2g18220 show 25 and 28% identity and 39 and 45% similarity, respectively, with the yeast NOC2 domain (Figure 2B). In yeast, the Noc2 protein is required for the maturation and intranuclear transport of the 60S ribosome subunit and pre-rRNA processing (Milkereit et al., 2001). NOC2 was shown to be essential for embryonic development in Caenorhabditis elegans (Kamath et al., 2003) and zebra fish (Amsterdam et al., 2004). Interestingly, a human NOC2-containing protein has been identified as an inhibitor of histone acetyltransferase activity (Hublitz et al., 2005).

The second EMS-induced crc modifier was mapped to At2g15790, which corresponds to SQN, the Arabidopsis homolog of cyclophilin 40 (CyP-40; Berardini et al., 2001). The new allele, sqn-4, has a G-to-A transition that alters the donor-splicing site of the 7th intron (Figure 2A). To verify that the mutation indeed corresponds to the SQN locus, the double mutant was constructed with crc-1 and another sqn allele, sqn-5 (SALK 033.511; Figure 2A). crc-1 sqn-5 was found to develop additional stamens and carpels inside the primary carpels (see Supplemental Figure 1B online). SQN is the only representative of the CyP-40 family in Arabidopsis and is predicted to have a peptidyl-prolyl cis-trans isomerase activity domain.

The third crc modifier was induced by a T-DNA insertion (Eshed et al., 1999). Sequencing of the flanking regions indicated that the T-DNA disrupted At4g28190, corresponding to ULT1 (Carles et al., 2005). Our mutation was therefore named ult1-4. In ult1-4, the borders of a T-DNA are inserted 95 bp upstream of the translational start site and 25 bp downstream of the stop codon (Figure 2A), resulting in a complete deletion of the coding sequence. RT-PCR experiments confirmed an absence of ULT1 transcripts in ult1-4 (see Supplemental Figure 4 online) that therefore corresponds to a null allele. Here, too, an additional ult1 allele, ult1-3 (SALK 074.642; Figure 2A), was crossed to crc-1, and the resulting double mutant presented the characteristic “Russian doll” gynoecium (see Supplemental Figure 1C online). ULT1 encodes a small protein containing a SAND domain, located between the 18th and the 108th amino acid, and a Cys-rich region. ULT1 and the closely related ULT2 (At2g20825) are the only two SAND domain proteins in the Arabidopsis genome. The SAND domain is a specific DNA binding domain previously characterized in Sp100, AIRE-1, NucP41/75, and DEAF-1 proteins (Gibson et al., 1998).

According to their predicted sequences and motifs, RBL, SQN, and ULT1 control FM termination by seemingly unrelated mechanisms. Moreover, the binary interactions between RBL, SQN, ULT1, and CRC have been tested using the yeast two-hybrid technique, and no physical interactions have been observed (see Supplemental Figure 5 online).

RBL, SQN, and ULT1 Expression Patterns and Subcellular Localization of the Proteins

We used RT-PCR to determine the distribution of RBL, SQN, and ULT1 mRNA transcripts in wild-type tissues (Figure 3A). The three genes are expressed in all tissues tested, the highest levels of expression being observed in tissues with cells undergoing division, especially in the inflorescence, and the lowest in old leaves, except for SQN, whose expression level remained constant during leaf development. Global expression data analyses (Digital Northern, Gene Chronologer, Gene Atlas, and Meta Analyzer; https://www.genevestigator.ethz.ch/at/) also indicated that the three genes are ubiquitously expressed, with higher expression levels during the flowering transition and during seed formation and at the rosette stage for SQN. We also analyzed the global expression data in homeotic flower mutant backgrounds (Wellmer et al., 2004) and found that RBL, SQN, and ULT1 transcript levels were not altered in ap1, ap2, ap3, pi, and ag mutants.

Figure 3.
Expression Patterns of the Three Genes and Subcellular Localizations of RBL and SQN.

We next performed in situ hybridization experiments to localize more precisely RBL, SQN, and ULT1 transcripts during flower development. Regarding ULT1, we obtained similar flower expression patterns to those described by Carles et al. (2005). RBL expression can be detected throughout the iSAM and the FM at young stages 1 and 2 (Figure 3B). Later during flower development, the hybridization signal becomes weaker in differentiating organs and remains strong in the center of FM and young emerging primordia (Figures 3C to 3F). However, this apparent reduction of the signal is likely due to differences in cell density and size in differentiating organs compared with proliferating tissues, rather than to a decrease of RBL expression per se. In mature flowers (stages 11 to 13), a strong signal is detected in ovules (Figure 3G) and a weak one (possibly for the same reasons as those described above) in valves. Fusions of the β-glucuronidase (GUS) reporter gene downstream of RBL promoter sequences confirmed the ubiquitous expression patterns of RBL (Figures 3I to 3L). We did not manage to obtain detectable in situ hybridization signals with SQN, whose expression is likely very low and ubiquitous, making signals difficult to observe compared with the background. Fusions of the GUS reporter gene downstream of SQN promoter sequences indicate ubiquitous expression pattern (Figures 3Q to 3S), with the exception of anthers, which remained unlabeled in several independent transgenic lines.

To determine the in vivo subcellular localization of each protein, we generated transgenic plants stably expressing RBL-green fluorescent protein (GFP) and SQN-GFP fusion proteins under the control of the 35S promoter (Figures 3M and 3T, respectively) and the RBL-monomeric red fluorescent protein (mRFP) fusion protein under the control of its own promoter (Figure 3N). Transgenic plants expressing the SQN-GFP fusion protein under the control of its own promoter were also generated but did not provide detectable signals, likely because of the low expression level. Both ProRBL:RBL-mRFP and ProSQN:SQN-GFP constructs were able to complement the indeterminacy phenotype generated by the corresponding double mutant with crc-1 (see Supplemental Table 1 online). Confocal microscopy observations were performed on roots and show that RBL is localized in the nucleus and SQN in the cytoplasm (Figures 3M, 3N, and 3T, respectively). Since expression levels of the fluorescent proteins were sometimes low (especially with the SQN-GFP construct), we also performed transient expression assays in tobacco (Nicotiana benthamiana) leaves using the same RBL- and SQN-GFP fusions. RBL is mainly localized in the entire nucleus (Figure 3P), and SQN is mainly localized in the cytoplasm (Figure 3U).

In summary, RBL and SQN seem to display ubiquitous expression patterns. RBL, SQN, and ULT1 are probably general regulators of development acting upstream of flower-specific genes to trigger more specific effects on FM termination.

WUS and CLV3 Expression Persists in the Center of the Floral Meristem in the crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 Double Mutants

To investigate the molecular basis of the indeterminacy in crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 backgrounds, we followed WUS and CLV3 expression by in situ hybridization and with a ProCLV3:GFP reporter line, respectively.

In wild-type plants, WUS is not expressed at stage 6, after carpel primordia emerge (Mayer et al., 1998; Figure 4A). In each of the three double crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 mutants, during stage 6, a strong WUS expression is detected in the center of the dome that separates carpel primordia (Figures 4B to 4D). WUS expression remains detectable until much later stages (Figure 4E). The three double crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 mutants were crossed with the ProCLV3:GFP reporter line. Stereomicroscopy GFP observations of F2 plants, homozygous for both mutant alleles and containing the CLV3 reporter construct, show that the apices of indeterminate flowers express CLV3 (Figure 4F), thus confirming the maintenance of a bona fide meristematic activity.

SQN and ULT1 Promote AG Expression in an Inner Subdomain of Whorl 4

The prominent role of AG in the control of FM termination prompted us to monitor its expression pattern by in situ hybridization in the crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 backgrounds.

In the wild type, AG is uniformly expressed throughout the FM center from stage 3 and is restricted to the 3rd and 4th whorls from stage 5 onwards (Figures 5A to 5D). By contrast, AG expression seems to be downregulated in the very center of the FM from stage 3, in both crc-1 sqn-4 (Figures 5E to 5H) and crc-1 ult1-4 (Figures 5I to 5L). In the crc-1 rbl-1 background, we were not able to observe obvious modifications of AG expression. Although RBL does not seem to affect the pattern of AG transcription, it could either positively and uniformly influence AG transcription throughout the FM center and young organ primordia or influence AG activity at a posttranscriptional level.

Figure 5.
AG Expression Patterns in Wild-Type and crc-1 sqn-4 and crc-1 ult1-4 Double Mutant Backgrounds.

In conclusion, the downregulation of AG expression in the very center of the FM delimits a 4th whorl subdomain where specific regulatory pathways (involving SQN and ULT1) operate to control AG expression.

RBL, SQN, and ULT1 Redundantly Control FM Termination

We next evaluated the effects of each of the modifiers, alone or in combination, on FM determinacy. In each of the single mutants, the flowers produce supernumerary carpels (Table 1) that sometimes develop in the upper half of the gynoecium and therefore modify its morphology (Figure 6B). In addition, ult1-4 shows a significant increase of sepal, petal, and stamen number (Table 1, Figure 6C). These data are in accordance with those reported by Berardini et al. (2001) for sqn-1 and by Fletcher (2001) for other ult1 alleles. Note that the gynoecia of the three mutants are also slightly shorter and wider than the wild type, but no additional whorls are produced inside. Thus, the three mutants exhibit mild effects on primary carpel number but show no alteration of FM termination.

Figure 6.
Phenotypes of the rbl-1, sqn-4, and ult1-4 Single, Double, and Triple Mutants.

Double mutants between rbl-1, sqn-4, and ult1-4 were subsequently generated in pairwise combinations. In all of these double mutants, gynoecia are often borne on a gynophore and are shorter and wider compared with the wild type (Figure 6D). Some combinations also show increases in primary carpel numbers compared with single mutants (Table 1). In addition, gynoecia of all double mutants display reiterations of whorls of stamens, carpels, and sometimes petals inside the primary carpels, which are not observed in single mutants (Figure 6E). As observed in corresponding crosses with crc-1, the ectopic organs develop on an elongating axis that almost always grows from the base of the primary gynoecium. The phenotypic similarity (with the exception of the distal carpel fusion trait) of double mutants between crc-1 and the identified modifiers and the modifiers themselves strongly suggests that CRC and its modifiers are required redundantly for FM termination.

In rbl-1 sqn-4 ult1-4 triple mutant flowers, sepal and petal numbers are increased similar to that observed in ult1-4 (Figures 6F and 6I). All triple mutant flowers display additional primary stamen whorls compared with the double mutants, but the strength of indeterminacy is variable between flowers along the inflorescence, with the strongest effects usually observed in the first flowers. In the weak phenotypes, the increase of primary stamen whorl number is moderate, while fused carpels enclosing extra floral organs (stamens and carpels) develop in the FM center (Figure 6F). In addition, chimeric stamen-petal organs regularly develop in place of the primary stamens (Figures 6G and 6H). In intermediate phenotypes, flowers do not display fused carpels. Instead, reiterations of unfused carpel-like structures surrounding stamens are observed (Figure 6I). Chimeric stamen-carpel organs can also be observed in these flowers. These organs resemble stamens but bear ovules on the connective and display stigmatic papillae at their top (Figure 6J). Flowers showing a strong phenotype bear only stamens, organized in a spiral phyllotaxy and produced by an everlasting meristem at the center of mature flowers (Figures 6K and 6L).

In conclusion, our data show that RBL, SQN, and ULT1 are genetically redundant and are involved in the control of FM termination. Their effects on FM termination depend on WUS activity since wus-1 is epistatic to rbl-1, sqn-4, and ult1-4 mutants: each double mutant develops one or two stamens and no carpels, similar to the wus-1 single mutant (Figures 4G and 4H). Double mutant combinations between rbl-1, sqn-4, and ult1-4 show the same type of indeterminacy observed in combinations with crc-1 and show reiterations of ectopic stamens and carpels, suggesting that the effects of pairwise combinations of these mutants are limited to late events (from stage 6 onwards) of FM termination. By contrast, the triple mutant combination produces the completely new phenotype of a loss of FM termination with an expansion of whorl 3, giving rise to a spiral of stamens (Figures 6K and 6L). Thus, the simultaneous malfunctioning of these three genes has effects on earlier events (before stage 6) of FM termination compared with the pairwise combinations of these same mutations. The loss of FM termination, the development of chimeric petal/stamen organs, and the large increase of stamen number are reminiscent of some aspects of the ag and sup phenotypes and therefore suggest that RBL, SQN, and ULT1 genes could function in both AG and SUP pathways.

rbl-1, sqn-4, and ult1-4 Reinforce the ag Phenotype

To further investigate the interactions of RBL, SQN, and ULT1 with AG, we crossed rbl-1, sqn-4, and ult1-4 to ag-4 (weak allele [Sieburth et al., 1995]; Figures 7A to 7D, 7I, and 7J) and ag-6 (strong allele; Figures 7E to 7J). ag-6 results from a T-DNA insertion in AG, and protein gel blot analysis of total protein from wild-type and ag-6 mutants with an AG-specific polyclonal antibody (Ito et al., 1997) demonstrated that the AG protein is not expressed in ag-6, thus confirming ag-6 as a null allele.

Figure 7.
Phenotypes of ag-4 and ag-6 Single Mutants and of ag rbl-1, ag sqn-4, and ag ult1-4 Double Mutants.

First, ag-4 rbl-1 double mutants develop flowers that resemble ag-6 flowers (cf. Figures 7A, 7B, and 7E), whereas the ag-4 sqn-4 double mutant flowers produce more organs than ag-4 (cf. Figures 7A to 7C), and the ag-4 ult1-4 double mutant develops flowers similar to ag-4 flowers but with some of the stamens partially transformed into petals (cf. Figures 7A and 7D). The distinct enhancements of the ag-4 phenotype therefore suggest that RBL, SQN, and ULT1 all positively influence AG activity.

In the progeny from the cross with ag-6, double mutants look only slightly different from ag-6 (cf. Figures 7E and 7F). In ag-6 flowers, the successive whorls of sepals are clearly recognizable and there are 10.22 ± 0.52 (n = 23) petals between the two first sepal whorls. In ag-6 rbl-1 and ag-6 sqn-4 flowers, there are 12.36 ± 1.19 (n = 28) and 12.03 ± 1.28 (n = 32) petals, respectively, before reaching the second whorl of sepals (Figure 7G). ag-6 rbl-1 and ag-6 sqn-4 therefore have phenotypes that resemble those of ag-6 sup-1 (cf. Figures 7F and 7H). A weaker transformation was observed in ag-6 ult1-4, since 11.68 ± 0.82 (n = 19) petals were counted before reaching the second whorl of sepals whose transformation into petal is less pronounced.

Plants inferred to be heterozygous for ag-6 or ag-4 and homozygous for either rbl-1, sqn-4, or ult1-4 display a floral formula identical to that of the rbl-1, sqn-4, and ult1-4 single mutants, but the gynoecia are larger, borne on a short gynophore (Figure 7I), and enclose ectopic stamens and carpels (Figure 7J) produced from the center of the FM. While 100% (n = 69) of ag-4/+ rbl-1 flowers are indeterminate, only 41.6% (n = 100) of ag-4/+ sqn-4 and 31.7% (n = 124) of ag-4/+ ult1-4 are.

Such results show that rbl-1, sqn-4, and ult1-4 mutations enhance, to somewhat different degrees, the ag-4 phenotype and, therefore, that RBL, SQN, and ULT1 play partially redundant roles in AG activity. However, crosses with ag-6 suggest that RBL, SQN, and, to a lesser extent, ULT1 also mediate flower determinacy through an additional pathway that could involve SUP since phenotypes of ag-6 rbl-1 and ag-6 sqn-4 have attributes of those of ag-1 sup-1 (Schultz et al., 1991; Bowman et al., 1992).

rbl-1, sqn-4, and ult1-4 Mutations Increase the sup-1 Phenotype

The spirally initiated stamens observed in rbl-1 sqn-4 ult1-4 and the ag-1 sup-1-like phenotype (Bowman et al., 1992), mimicked by ag-6 rbl-1 and ag-6 sqn-4, prompted us to evaluate RBL, SQN, and ULT1 functions with respect to SUP activity. rbl-1, sqn-4, and ult1-4 mutants were therefore crossed with sup-1.

Compared with the wild type, sup-1 produces more stamens and fewer carpels that are often staminoid (Figure 8A, Table 2; Bowman et al., 1992). The increase in floral organ number results from a delayed termination of the FM due to a delay of the extinction of WUS expression in the center of the FM (Figures 8B and 8C). Remarkably, sup-1 rbl-1 flowers display everlasting meristems that indefinitely produce stamens in a spiral phyllotaxy (Figures 8D and 8E, Table 2). sup-1 sqn-4 flowers also possess many more stamens than sup-1 (Table 2), but the flowers produce a few unfused stamen-carpel organs at their center (Figures 8F and 8G), indicating a strongly delayed termination. In sup-1 sqn-4, stamen phyllotaxy is spiral in the first five flowers and then becomes more whorled. Enhancement of the sup-1 phenotype is less pronounced in the ult1-4 background, where only additional stamens are produced compared with sup-1 (Figure 8H, Table 2).

Figure 8.
Phenotypes of sup-1 rbl-1, sup-1 sqn-4, sup-1 ult1-4, and sup-1 ag-4 Double Mutants.
Table 2.
Floral Organ Number in Single sup-1 and Double sup-1 rbl-1, sup-1 sqn-4, and sup-1 ult1-4 Mutants

To test whether the enhancement of the sup-1 phenotype by rbl-1, sqn-4, and ult1-4 could be due to a slight decrease of AG function compared with the sup-1 single mutant, we generated the sup-1 ag-4 double mutant. As shown in Figures 8I, sup-1 ag-4 flowers are indeterminate and produce an indefinite number of stamens in a spiral phyllotaxy, thus mimicking rbl-1 sup-1, and also the strong rbl-1 sqn-4 ult1-4 phenotypes.

These data suggest that rbl-1, sqn-4, and, to a lesser extent, ult1-4 increase the sup-1 phenotype by decreasing AG function. However, the fact that the rbl-1 sqn-4 ult1-4 triple mutant and the sup-1 ag-4 double mutant produce indistinguishable indeterminate male flowers suggests that both AG and SUP activities are modified in rbl-1 sqn-4 ult1-4.

Mutations in B Class Genes Do Not Significantly Modify Floral Indeterminacy Levels Observed in the crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 Backgrounds

Previous data showed that each crc modifier in the crc background produces extra stamens inside gynoecia (Figure 1), implying that B class genes are reactivated in the FM center. To evaluate whether ectopic expression of B class genes influences FM indeterminacy in our mutant backgrounds, we introduced pi-1 and ap3-3 mutations into each of the crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 double mutants.

In B class single mutants, petals and stamens are transformed into sepals and carpels, respectively, and floral growth is determinate (Figures 9A and 9B; Bowman et al., 1989). In triple mutants (Figures 9C and 9D), floral organ identities are modified similarly to the ap3-3 and pi-1 single mutants, but FM growth remains indeterminate similar to the crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 double mutants. Since all triple mutants behave similarly (only two examples are given in Figure 9), we conclude that B class genes are not essential in promoting indeterminacy in crc-1 rbl-1, crc-1 sqn-4, and crc-1 ult1-4 mutant backgrounds or that effects of B class mutations on FM termination depend on AG, as proposed from previously reported data (Bowman et al., 1989; Schultz et al., 1991; Sakai et al., 2000).

Figure 9.
Phenotypes of Triple Mutants Combining crc-1 rbl-1, crc-1 sqn-4, or crc-1 ult1-4 and ap3-3 or pi-1 Mutations.


Our results address the regulation of the switch from indeterminate to determinate growth during flower development. We report on three genes, RBL (a newly discovered and cloned gene), SQN (to which we assign a new function), and ULT1, which redundantly control FM termination. As single mutants, these genes have minor effects on carpel number, but when combined with each other or with mutations in AG, SUP, and CRC, they trigger a strong and persistent FM indeterminacy with the development of extra floral whorls. These phenotypes have been categorized into two different temporal classes (Figure 10A). We discuss the fact that these two classes reveal a continuous control on FM termination, which is initiated at stages 3 to 4 and maintained through stage 6 and beyond. Our data also point to the existence of an inner 4th whorl subdomain, which keeps a stem cell identity and maintains a meristematic potential, as distinct from the remainder of the 4th whorl, which is competent to produce carpel primordia. We suggest that a decrease of AG expression in this inner 4th whorl subdomain is sufficient to disrupt FM termination. However, we discuss why a decrease of AG activity alone is not sufficient to explain all phenotypes. Finally, we argue that RBL, SQN, and ULT1 features make these genes candidates for contributors to floral developmental homeostasis.

Figure 10.
Model for RBL, SQN, and ULT1 Action in the Spatial and Temporal Control of FM Determinacy.

RBL, SQN, and ULT1 Promote FM Termination from Stages 3 to 4 on and Maintain the Repression of a Meristematic Activity in the Center of the FM at Stage 6 and Beyond

The first class of phenotypes, observed in pairwise combinations of rbl-1, sqn-4, ult1-4, and crc-1 (Class I; Figures 1, ,6,6, and 10A), groups together flowers that show reiterations of stamens, carpels, and occasionally petals inside the primary carpels. Since reiterations are produced after the formation of the primary carpels, we estimate that they result from events at stage 6 and beyond. In addition, the identity of reiterated organs indicates a recurrent resetting of the FM identity after each carpel whorl. Our data therefore suggest that after primary carpels have been initiated as lateral organs on the floral axis (Bossinger and Smyth, 1996), a meristematic potential persists in the intercarpellary space. This would suggest that, in the wild type, stem cells are not entirely consumed to build carpels but rather transiently persist in a repressed state, with RBL, SQN, and ULT1 playing a role in this process.

Flowers from the second class (Class II; Figures 6, ,8,8, and 10A) show an increase of organ or whorl numbers in one or several of the primary floral organ domains. The strongest phenotype observed in this class, shown by the rbl-1 sqn-4 ult-1 and the sup-1 rbl-1 mutant combinations, results in spirally indeterminate male flowers due to an expanding whorl 3. Unlike Class I, Class II phenotypes do not result from the recurrent resetting of the FM identity but rather from its maintenance with a continuing 3rd whorl identity. This indicates that indeterminacy takes place at an earlier stage in Class II mutants than in Class I. Since the reduction of AG expression in the very center of the FM is observed from stages 3 to 4 onward (Figure 5) and extra stamen development in sup mutants results from alteration to the whorl 3/4 boundary from stage 4 onward (Sakai et al., 1995, 2000), we propose that Class II phenotypes represent events occurring at stages 3 to 4.

The two phenotypic classes are arbitrary and, as suggested by Figure 10A, the phenotypic transitions within and between the two classes are progressive. The phenotypes therefore denote a continuum and strongly suggest that FM termination, once initiated at stages 3 to 4, needs to be continually maintained after all primary flower organs have been initiated, including carpels. However, the two classes have been instrumental in dissecting out the spatio-temporal control of FM termination, as discussed below.

RBL, SQN, and ULT1 Promote AG Function in FM Termination

Class I phenotypes result from a reduction of AG expression mainly in the intercarpellary space (Figures 5 and 10C). This is consistent with the transient reduction of AG expression previously observed in indeterminate ult1-1 (Fletcher, 2001) and clv1 (Clark et al., 1993) mutants and also with the Class I–like phenotypes generated in AG antisense experiments (Type III phenotypes in Mizukami and Ma, 1995). Interestingly, FM proliferation and floral reversion observed in Impatiens also correlate with an absence of AG expression in the central proliferating meristem (Chiurugwi et al., 2007). Class I phenotypes, however, differ from the one resulting from a lack of AG expression in the entire 4th whorl (ProAP3:AG ag-3 plants; Jack et al., 1997). The latter is also indeterminate but does not produce carpels. Our data therefore restrict the domain where AG activity is necessary for the maintenance of FM termination to the intercarpellary space and show that SQN and ULT1, and probably RBL, are jointly required to promote AG activity in this domain (Figure 10C). Since AG expression is activated by WUS in a feedback loop (Lenhard et al., 2001; Lohmann et al., 2001) and wus-1 is epistatic to rbl-1, sqn-4, and ult1-4 (Figure 4), RBL, SQN, and ULT1 could function downstream of WUS in the activation of AG in the center of the FM.

The two distinct functions of AG, organ identity and FM termination, therefore seem to correspond to distinct domains of expression, the 3rd whorl and the outer 4th whorl subdomains and the inner 4th whorl subdomain, respectively. Spatial determinants defining this last domain remain to be identified.

In addition, Class I mutants confirm the function of CRC in FM termination (Figures 1 and 10C) proposed by Alvarez and Smyth (1999), based on the fact that crc-1 spt-2 and especially crc-1 ag-1/+ double mutants develop supernumerary whorls internal to the primary carpels. Similar to AG, CRC is therefore involved both in carpel identity (polarity, size, and distal fusion; Alvarez and Smyth, 1999; Eshed et al., 1999) and FM termination. However, the molecular role of CRC in this last regulatory network remains elusive. While it has been shown that CRC expression depends upon AG activity (Gomez-Mena et al., 2005; Lee et al., 2005), no data reported to date have shown the converse. That would suggest that CRC and AG influence floral termination through independent pathways, in agreement with the lack of CRC expression at the center of the stage 6 FM.

Class II phenotypes are also very likely to result from a reduction of AG activity in the inner 4th whorl subdomain, since reiterated organs produced on the flanks of the FM are stamens and occasionally carpels, and no phenotypes similar to strong ag alleles were generated in rbl-1 sqn-4 ult1-4 and sup-1 rbl-1 combinations (Figures 6, ,8,8, and 10B). However, AG misregulation alone cannot explain the Class II phenotypes: strong indeterminacies observed in ag have not been associated with such an extended 3rd whorl (Bowman et al., 1989; Mizukami and Ma, 1995; Sieburth et al., 1995; Chuang and Meyerowitz, 2000). One hypothesis is that Class II phenotypes result from the downregulation of both SUP and AG at stages 3 to 4, corresponding to the onset of their activities (Yanofsky et al., 1990; Schultz et al., 1991; Bowman et al., 1992; Sakai et al., 2000). Several observations support this hypothesis. First, we observed an increase of primary stamen number in the weak and moderate phenotypes of rbl-1 sqn-4 ult1-4 (Figure 6). Second, the ProAP3:AG ag-3 sup-1 mutant (Jack et al., 1997) and the ag-1/+ sup-2 (Schultz et al., 1991) and ag-4 sup-1 double mutants (Figure 8) display completely indeterminate male flowers, similar to the strong Class II phenotype (Figure 10A). Third, ag-6 rbl-1 and ag-6 sqn-4 display an ag-6 sup-1–like phenotype (Figure 7), with both phenotypes resembling the strong Class II phenotype (except for the phyllotaxy) with indeterminacy due to an expanded 3rd whorl, stamens being transformed into petals because of the strong ag-6 allele. Our data highlight the convergence of AG and SUP pathways to control FM termination, as recently discussed by Sablowski (2007), and further support earlier data on the synergistic action of AG and SUP on FM determinacy (Schultz et al., 1991; Bowman et al., 1992). RBL, SQN, and, to a lesser extent, ULT1 appear to be part of this flower termination regulatory pathway (Figure 10B).

Recently, it has been shown that ectopic expression of miR172-resistant AP2 triggers floral indeterminacy with a phenotype similar to our strong Class II flowers (Zhao et al., 2007). Interestingly, the authors propose that this is due to a decrease in AG and CLV expression, together with an ectopic expression of PI in the center of the FM. Our data are consistent with this regulatory scheme, and effects of mutations in RBL and SQN on AP2 and CLV activities deserve to be tested. However, the authors claimed that the shift of the inner boundary of B class gene expression toward the center of the FM, observed in the miR172-resistant AP2 lines, does not depend on SUP activity but rather on the decrease of AG expression in the center of the FM. In our case, it seems that the Class II phenotype is unlikely to result from AG and CLV misregulation alone.

Together, our results, and the results of others, suggest that FM termination is based on a spatial and temporal sequence of events operating in the FM center (Figures 10B and 10C). The initiation phase (from stages 3 to 4) appears to require AG and SUP activities. The maintenance phase (up to and beyond stage 6) appears to require AG, CLV, and probably CRC activities. Based on our results, we propose that RBL, SQN, and ULT1 are factors required during both early and late steps of the FM termination process (Figure 10).

RBL, SQN, and ULT1 Largely Contribute to Flower Termination Homeostasis

The range of floral indeterminacy phenotypes reported here might also have evolutionary implications since they illustrate phenotypic variations and transitions observed between flowers of the same family or during the course of evolution (Ronse de Craene, 2004; Baum and Hileman, 2006; Endress and Doyle, 2007). For instance, the combination of sup-1 and rbl-1 transforms a hermaphrodite whorled and determinate flower into a unisexual male, spiraled, and indeterminate flower. The mutant mimics the hypothetical angiosperm male ancestor flower, according to the out of male (Theissen et al., 2002) and the mostly male (Frohlich, 2002) theories. In addition, several Class I flowers display elongated internodes, an ancestral trait that tends to be repressed in the course of evolution, likely because of its lower reproductive fitness (Baum and Hileman, 2006). The fact that internode elongation only occurs between reiterated organs suggests that floral axis compaction and the competence for FM termination are linked in Arabidopsis.

The range of phenotypes also illustrates the consequences of a decreased floral developmental homeostasis. Floral developmental homeostasis corresponds to the tendency of angiosperms to make more uniform flowers (fixed number of parts developing in the right position), in spite of mutations and fluctuating environments (Frohlich, 2006). Core eudicots show more floral developmental homeostasis than many groups, such as Ranunculaceae (Aquilegia sp, Helleborus foetidus) or Magnoliaceae (Magnolia sp). These other groups have less uniform numbers of whorls and parts per whorl, similar to our mutants. It is conceivable that the evolution of enhanced developmental homeostasis in core eudicots was partly due to functions provided by RBL, SQN, and ULT1 and that these genes might have played some indirect role in the great success of the core-eudicot clade.


Plant Growth, Crosses, and Line Isolations

All mutants are in the Ler ecotype background with the exception of the ult1-3 (SALK 074.642), crc-8 (SALK 065.539), sqn-5 (SALK 033.511), rbl-3 (SALK 059.267), and rbl-4 (SALK 071.349) allelic lines that are in the Columbia-0 (Col-0) background and the rbl-2 (FLAG 299B04) allelic line that is in the Wassilewskija background. Allelic lines were obtained from the Nottingham Arabidopsis Stock Centre (SIGnAL T-DNA-Express, http://www.arabidopsis.org) or the Institut National de la Recherche Agronomique, Versailles (rbl-2; http://dbsgap.versailles.inra.fr). ag-1, ag-4, sup-1, and the reporter line ProCLV3:GFP were provided by E. Meyerowitz (California Institute of Technology) and wus-1 by T. Laux (University of Freiburg, Germany). ag-6 is a new strong allele of ag resulting from a T-DNA insertion in AG.

Plants were grown in soil, first for 15 d under short-day conditions (10 h cool white fluorescent light; 10,000 lux luminosity; 55 to 80% of humidity; 19°C [day] to 15°C [night]) and then under long-day conditions (18 h cool white fluorescent light; 10,000 to 20,000 lux luminosity; 60 to 66% of humidity; 20.5°C [day] to 15°C [night]).

The single mutant phenotypes of crc-1 modifiers were determined by analyzing progeny from 15 to 20 non-crc plants in F2 families (derived from double mutants crossed to the wild type) as F3 families. The homozygous T-DNA–tagged mutants were isolated in F2 progeny by PCR testing using a combination of gene and T-DNA–specific primers. Each mutant was backcrossed three times to the wild type, and in F2 generations, only plants homozygous for crc-1 and either rbl-1 or sqn-4 or ult1-4 displayed the FM indeterminacy phenotype. ult1-4 was tested for cosegregation of the phenotype with the T-DNA and found to be tagged.

Desired mutant combinations were identified among phenotypic categories in the F2 segregants generated by cross-fertilizing heterozygote mutants. Genotypes were confirmed by monitoring Mendelian ratios and by progeny testing. Whenever necessary, theoretical and observed averages were compared using the χ2 test.

Mutagenesis and Transformation

crc-1 seeds were mutagenized with 25 mM EMS for 12 h and phenotypic modifiers selected in the M2 (Eshed et al., 1999). The T-DNA tagging screen used the plasmid pOCA28-15-991 (Eshed et al., 1999). The modifiers were backcrossed three times to wild-type Ler. All transgenic plants were generated by the infiltration method described by Bechtold and Pelletier (1998) with selection on soil for resistance to the herbicide BASTA (Bayer; 0.05% of the commercial solution) or on Murashige and Skoog plates due to resistance to the antibiotic hygromycin (20 μg/L).

Mapping and Molecular Identification of the Modifier of crc Genes

rbl-1 and sqn-4 were mapped by crossing crc-1 rbl-1 and crc-1 sqn-4 (Ler) to crc-8 (Col-0, SALK 065.539). crc-8 contains a T-DNA insertion in the 4th intron of the gene that is associated with a 14-bp deletion from nucleotide positions 820 to 833 (relative to the translational start codon). Using cleaved amplified polymorphic sequence (CAPS) and simple sequence length polymorphism (SSLP) markers (Konieczny and Ausubel, 1993; Bell and Ecker, 1994), we established that rbl-1 was localized on the right arm of chromosome 3 flanked by markers AP3-linked and CIW19 and sqn-4 was on the left arm of chromosome 2 flanked by markers CIW3 and THY1. For the fine mapping, we designed new SSLP or CAPS markers that spanned the region between the respective markers, using the Monsanto Arabidopsis polymorphism and Ler sequence database (http://www.arabidopsis.org/Cereon) as a source for potential single-nucleotide polymorphisms and small insertions/deletions between Ler sequence and Col-0 sequence. The primer sequences, restriction enzymes, and number of restriction sites in the Col-0/Ler ecotypes for the SSLP and CAPS marker sequences generated across these intervals are available upon request.

To isolate the ult1-4 associated gene, we took advantage of the enhancer trap vector pOCA28-15-991 that allows plasmid rescue of flanked DNA sequences (Eshed et al., 1999).

Sequencing was performed on an ABI PRISM 3100 genetic analyzer sequencer (Perkin-Elmer) according to the manufacturer's instructions. Computer-based sequence analysis was performed using VectorNTI Suite (Informax) and Sequencher 4.0 (Gene Codes) software. Multiple alignments were obtained using ClustalX and edited with SeqVu (The Garvan Institute of Medical Research, Australia).


Total RNA was isolated from various tissues using Trizol reagent (Invitrogen) and treated with RNase-free DNaseI (Ambion). The first-strand cDNA synthesis was performed on 0.6 μg of total RNA using RevertAid M-MuLV reverse transcriptase (Fermentas) and an oligo(dT) primer (18-mer) according to the manufacturer's instructions. From 20 μL of RT product, 1 μL was used for each PCR reaction. The annealing temperature was 55°C for all primer pairs, and 25 cycles of PCR were performed for all genes. After amplification, RT-PCR products were separated on agarose gel and stained with ethidium bromide.

In Situ Hybridization, Toluidine and GUS Staining, and GFP Transient Expression

Tissue preparation, for both in situ hybridization and toluidine staining, was as described by de Almeida-Engler et al. (2001). Tissue sections were 10 μm thick. Probes for in situ hybridization were transcribed using digoxigenin labeling mix. WUS and AG templates were generated as previously described (Bowman et al., 1991; Mayer et al., 1998). RBL probe was synthesized in vitro from the PCR fragment containing 5′ ends of RBL. Whenever possible, constructs for the promoter and gene fusions upstream of GUS, GFP, and RFP reporter genes were made using the Gateway system. cDNA and DNA sequences were cloned into pDON207 vector and then transferred into the appropriate destination vector using LR cloning enzyme. We used pMDC162, pMDC107, and pMDC83 destination vectors for the Pro:GUS, gene-GFP, and Pro35S:cDNA-GFP constructs, respectively (Curtis and Grossniklaus, 2003), and pGIINK-GK7B 35Ster destination vector for the gene-mRFP constructs (Rotman et al., 2005). We always used the full intergenic sequences as promoter regions, except for the ProRBL:RBL-mRFP construct, in which we used 1.3 kb upstream of the first Met. GUS activity was assayed according to McConnell and Barton (1998). Prom35S:cDNA-GFP construct were also assayed by transient expression in tobacco (Nicotiana tabacum) leaves. Leaves were infiltrated as described (Rathjen et al., 1999), with an exponential phase culture Agrobacterium tumefaciens strain C58C1, diluted to an OD600 of 0.5 with infiltration medium (10 mm MES, pH 5.6, 10 mm MgCl2, and 200 μm acetosyringone).


Plants were observed with a Leica MZFLIII stereomicroscope coupled to a DC300F digital camera (Leica Microsystems). Images were processed with the FW4000 software (Leica). Fresh flowers were also observed on a Hitachi S-3000N variable pressure scanning electron microscope. In situ hybridization signals and toluidine-stained tissues were visualized using a Nikon epifluorescence microscope (Optiphot-2) coupled with an AxioCam MRc digital camera (Carl Zeiss). Images were processed with the AxioVision software (Zeiss). GFP expression was monitored from homozygote plants using the Leica LSM-510 confocal microscope. All figures were composed with Adobe Photoshop 7.0 (Adobe Systems), and certain flower images have been treated to remove unrelated background.

Accession Numbers

Arabidopsis Genome Initiative locus identifiers are as follows: AP3 (AT3g54340), AG (AT4g18960), CRC (AT1g69180), PI (AT5g20240), RBL (AT3g55510), SQN (AT2g15790), SUP (AT3g23130), and ULT1 (AT4g28190).

Supplemental Data

The following materials are available in the online version of this article.

  • Supplemental Figure 1. Phenotypes of rbl, sqn, and ult1 Allelic Lines Crossed with crc-1.
  • Supplemental Figure 2. Phenotype of rbl-3.
  • Supplemental Figure 3. Alignment of the Translation Products of At3g55510 (RBL) and At2g18220.
  • Supplemental Figure 4. Genomic and RT- PCR on the ult1-4 Allele.
  • Supplemental Figure 5. Yeast Two-Hybrid Analysis of Binary Interactions between RBL, SQN, ULT1, and CRC.
  • Supplemental Table 1. Complementation Tests.
  • Supplemental Data Set 1. Fasta File Corresponding to the Alignment in Figure 2B.

Supplementary Material

[Supplemental Data]


This work was supported by the Agence Nationale de la Recherche through Grant ANR-05-BLAN-0280-01. We thank Christian Dumas and Jan Traas for supporting this project. We also thank Trudi Higginson, François Parcy, Mike Frohlich, Pradeep Das, Cédric Finet, Charlie Scutt, and Stève de Bossoreille for constructive discussions. We thank Isabelle Anselme-Bertrand and Béatrice Burdin for their help during scanning electron microscopy imaging, Claire Lionnet for her help during confocal microscopy, Edwige Moyroud and Cyrille Py for their excellent technical help, and Mathieu Reymond for naming RBL. We acknowledge Elliot Meyerowitz and Thomas Laux for providing the Arabidopsis mutants, Toshiro Ito for the AG antiserum, and Frank Wellmer for access to primary data on transcriptome profiles of flower homeotic mutants. We also thank the anonymous reviewers for their comments.


The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantcell.org) is: Christophe Trehin (rf.noyl-sne@nihertc).

[W]Online version contains Web-only data.



  • 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. [PubMed]
  • Amsterdam, A., Nissen, R.M., Sun, Z., Swindell, E.C., Farrington, S., and Hopkins, N. (2004). Identification of 315 genes essential for early zebrafish development. Proc. Natl. Acad. Sci. USA 101 12792–12797. [PMC free article] [PubMed]
  • Angenent, G.C., Stuurman, J., Snowden, K.C., and Koes, R. (2005). Use of Petunia to unravel plant meristem functioning. Trends Plant Sci. 10 243–250. [PubMed]
  • Baum, D.A., and Hileman, L.C. (2006). A genetic model for the origin of flowers. In Flowering and Its Manipulation, C. Ainsworth, ed (Sheffield, UK: Blackwell Publishing), pp. 3–27.
  • Bechtold, N., and Pelletier, G. (1998). In planta Agrobacterium-mediated transformation of adult Arabidopsis thaliana plants by vacuum infiltration. Methods Mol. Biol. 82 259–266. [PubMed]
  • Bell, C.J., and Ecker, J.R. (1994). Assignment of 30 microsatellite loci to the linkage map of Arabidopsis. Genomics 19 137–144. [PubMed]
  • Berardini, T.Z., Bollman, K., Sun, H., and Poethig, R.S. (2001). Regulation of vegetative phase change in Arabidopsis thaliana by cyclophilin 40. Science 291 2405–2407. [PubMed]
  • Blazquez, M.A., Ferrandiz, C., Madueno, F., and Parcy, F. (2006). How floral meristems are built. Plant Mol. Biol. 60 855–870. [PubMed]
  • Bossinger, G., and Smyth, D.R. (1996). Initiation patterns of flower and floral organ development in Arabidopsis thaliana. Development 122 1093–1102. [PubMed]
  • Bowman, J.L., Drews, G.N., and Meyerowitz, E.M. (1991). Expression of the Arabidopsis floral homeotic gene AGAMOUS is restricted to specific cell types late in flower development. Plant Cell 3 749–758. [PMC free article] [PubMed]
  • Bowman, J.L., Eshed, Y., Baum, S., Emery, J.F., Floyd, S.K., Alvarez, J., Hawker, N.P., Lee, J.Y., Siegfried, K.R., Khodosh, R., Tatom-Jaurez, M., and Perea, J.V. (2001). The story of CRABS CLAW (or How we learned to love the mutagen). Flowering Newsletter 31: 3–11.
  • Bowman, J.L., Sakai, H., Jack, T., Weigel, D., Mayer, U., and Meyerowitz, E.M. (1992). SUPERMAN, a regulator of floral homeotic genes in Arabidopsis. Development 114 599–615. [PubMed]
  • 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. Development 126 2387–2396. [PubMed]
  • Bowman, J.L., Smyth, D.R., and Meyerowitz, E.M. (1989). Genes directing flower development in Arabidopsis. Plant Cell 1 37–52. [PMC free article] [PubMed]
  • Brand, U., Fletcher, J.C., Hobe, M., Meyerowitz, E.M., and Simon, R. (2000). Dependence of stem cell fate in Arabidopsis on a feedback loop regulated by CLV3 activity. Science 289 617–619. [PubMed]
  • Busch, M.A., Bomblies, K., and Weigel, D. (1999). Activation of a floral homeotic gene in Arabidopsis. Science 285 585–587. [PubMed]
  • Carles, C.C., Choffnes-Inada, D., Reville, K., Lertpiriyapong, K., and Fletcher, J.C. (2005). ULTRAPETALA1 encodes a SAND domain putative transcriptional regulator that controls shoot and floral meristem activity in Arabidopsis. Development 132 897–911. [PubMed]
  • Carles, C.C., and Fletcher, J.C. (2003). Shoot apical meristem maintenance: The art of a dynamic balance. Trends Plant Sci. 8 394–401. [PubMed]
  • Chen, X. (2004). A microRNA as a translational repressor of APETALA2 in Arabidopsis flower development. Science 303 2022–2025. [PubMed]
  • Chiurugwi, T., Pouteau, S., Nicholls, D., Tooke, F., Ordidge, M., and Battey, N. (2007). Floral meristem indeterminacy depends on flower position and is facilitated by acarpellate gynoecium development in Impatiens balsamina. New Phytol. 173 79–90. [PubMed]
  • Chuang, C.F., and Meyerowitz, E.M. (2000). Specific and heritable genetic interference by double-stranded RNA in Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 97 4985–4990. [PMC free article] [PubMed]
  • Clark, S.E., Running, M.P., and Meyerowitz, E.M. (1993). CLAVATA1, a regulator of meristem and flower development in Arabidopsis. Development 119 397–418. [PubMed]
  • Clark, S.E., Running, M.P., and Meyerowitz, E.M. (1995). CLAVATA3 is a specific regulator of shoot and floral meristem development affecting the same processes as CLAVATA1. Development 121 2057–2067.
  • Curtis, M.D., and Grossniklaus, U. (2003). A gateway cloning vector set for high-throughput functional analysis of genes in planta. Plant Physiol. 133 462–469. [PMC free article] [PubMed]
  • de Almeida Engler, J., De Groodt, R., Van Montagu, M., and Engler, G. (2001). In situ hybridization to mRNA of Arabidopsis tissue sections. Methods 23 325–334. [PubMed]
  • Dievart, A., Dalal, M., Tax, F.E., Lacey, A.D., Huttly, A., Li, J., and Clark, S.E. (2003). CLAVATA1 dominant-negative alleles reveal functional overlap between multiple receptor kinases that regulate meristem and organ development. Plant Cell 15 1198–1211. [PMC free article] [PubMed]
  • Doerner, P. (2006). Plant meristems: What you see is what you get? Curr. Biol. 16 R56–R58. [PubMed]
  • Endress, P.K., and Doyle, J.A. (2007). Floral phyllotaxis in basal angiosperms: Development and evolution. Curr. Opin. Plant Biol. 10 52–57. [PubMed]
  • Eshed, Y., Baum, S.F., and Bowman, J.L. (1999). Distinct mechanisms promote polarity establishment in carpels of Arabidopsis. Cell 99 199–209. [PubMed]
  • Ferrario, S., Immink, R.G., and Angenent, G.C. (2004). Conservation and diversity in flower land. Curr. Opin. Plant Biol. 7 84–91. [PubMed]
  • Fletcher, J.C. (2001). The ULTRAPETALA gene controls shoot and floral meristem size in Arabidopsis. Development 128 1323–1333. [PubMed]
  • Fletcher, J.C., Brand, U., Running, M.P., Simon, R., and Meyerowitz, E.M. (1999). Signaling of cell fate decisions by CLAVATA3 in Arabidopsis shoot meristems. Science 283 1911–1914. [PubMed]
  • Frohlich, M.W. (2002). The mostly male theory of flower origins: Summary and update regarding the Jurassic pteridosperm Pteroma. In Developmental Genetics and Plant Evolution, Q.C. Cronk, R.M. Bateman, and J.A. Hawkins, eds (London: Taylor and Francis), pp. 85–108.
  • Frohlich, M.W. (2006). Recommendations and goals for evo-devo research: Scenarios, genetic constraint, and developmental homeostasis. Aliso 22 172–187.
  • Gaiser, J.C., Robinson-Beers, K., and Gasser, C.S. (1995). The Arabidopsis SUPERMAN gene mediates asymmetric growth of the outer integument of ovules. Plant Cell 7 333–345. [PMC free article] [PubMed]
  • Gibson, T.J., Ramu, C., Gemund, C., and Aasland, R. (1998). The APECED polyglandular autoimmune syndrome protein, AIRE-1, contains the SAND domain and is probably a transcription factor. Trends Biochem. Sci. 23 242–244. [PubMed]
  • Gomez-Mena, C., de Folter, S., Costa, M.M., Angenent, G.C., and Sablowski, R. (2005). Transcriptional program controlled by the floral homeotic gene AGAMOUS during early organogenesis. Development 132 429–438. [PubMed]
  • Hill, J.P., and Lord, E.M. (1989). Floral development in Arabidopsis thaliana: A comparison of the wild-type and the homeotic pistillata mutant. Can. J. Bot. 67 2922–2936.
  • Hublitz, P., et al. (2005). NIR is a novel INHAT repressor that modulates the transcriptional activity of p53. Genes Dev. 19 2912–2924. [PMC free article] [PubMed]
  • Ito, T., Takahashi, N., Shimura, Y., and Okada, K. (1997). A serine/threonine protein kinase gene isolated by an in vivo binding procedure using the Arabidopsis floral homeotic gene product, AGAMOUS. Plant Cell Physiol. 38 248–258. [PubMed]
  • Jack, T. (2004). Molecular and genetic mechanisms of floral control. Plant Cell 16 (suppl.): S1–S17. [PMC free article] [PubMed]
  • Jack, T., Sieburth, L., and Meyerowitz, E. (1997). Targeted misexpression of AGAMOUS in whorl 2 of Arabidopsis flowers. Plant J. 11 825–839. [PubMed]
  • Kamath, R.S., et al. (2003). Systematic functional analysis of the Caenorhabditis elegans genome using RNAi. Nature 421 231–237. [PubMed]
  • Konieczny, A., and Ausubel, F.M. (1993). A procedure for mapping Arabidopsis mutations using co-dominant ecotype-specific PCR-based markers. Plant J. 4 403–410. [PubMed]
  • Krizek, B.A., and Meyerowitz, E.M. (1996). The Arabidopsis homeotic genes APETALA3 and PISTILLATA are sufficient to provide the B class organ identity function. Development 122 11–22. [PubMed]
  • Laux, T., Mayer, K.F., Berger, J., and Jurgens, G. (1996). The WUSCHEL gene is required for shoot and floral meristem integrity in Arabidopsis. Development 122 87–96. [PubMed]
  • Lee, J.Y., Baum, S.F., Alvarez, J., Patel, A., Chitwood, D.H., and Bowman, J.L. (2005). Activation of CRABS CLAW in the nectaries and carpels of Arabidopsis. Plant Cell 17 25–36. [PMC free article] [PubMed]
  • Lenhard, M., Bohnert, A., Jurgens, G., and Laux, T. (2001). Termination of stem cell maintenance in Arabidopsis floral meristems by interactions between WUSCHEL and AGAMOUS. Cell 105 805–814. [PubMed]
  • Lohmann, J.U., Hong, R.L., Hobe, M., Busch, M.A., Parcy, F., Simon, R., and Weigel, D. (2001). A molecular link between stem cell regulation and floral patterning in Arabidopsis. Cell 105 793–803. [PubMed]
  • Mayer, K.F., Schoof, H., Haecker, A., Lenhard, M., Jurgens, G., and Laux, T. (1998). Role of WUSCHEL in regulating stem cell fate in the Arabidopsis shoot meristem. Cell 95 805–815. [PubMed]
  • McConnell, J.R., and Barton, M.K. (1998). Leaf polarity and meristem formation in Arabidopsis. Development 125 2935–2942. [PubMed]
  • Milkereit, P., Gadal, O., Podtelejnikov, A., Trumtel, S., Gas, N., Petfalski, E., Tollervey, D., Mann, M., Hurt, E., and Tschochner, H. (2001). Maturation and intranuclear transport of pre-ribosomes requires Noc proteins. Cell 105 499–509. [PubMed]
  • Mizukami, Y., and Ma, H. (1995). Separation of AG function in floral meristem determinacy from that in reproductive organ identity by expressing antisense AG RNA. Plant Mol. Biol. 28 767–784. [PubMed]
  • Parcy, F., Nilsson, O., Busch, M.A., Lee, I., and Weigel, D. (1998). A genetic framework for floral patterning. Nature 395 561–566. [PubMed]
  • Rathjen, J.P., Chang, J.H., Staskawicz, B.J., and Michelmore, R.W. (1999). Constitutively active Pto induces a Prf-dependent hypersensitive response in the absence of avrPto. EMBO J. 18 3232–3240. [PMC free article] [PubMed]
  • Ronse de Craene, L.P. (2004). Floral development of Berberidopsis corallina: A crucial link in the evolution of flowers in the core Eudicots. Ann. Bot. (Lond.) 94 741–751. [PMC free article] [PubMed]
  • Rotman, N., Durbarry, A., Wardle, A., Yang, W.C., Chaboud, A., Faure, J.E., Berger, F., and Twell, D. (2005). A novel class of MYB factors controls sperm-cell formation in plants. Curr. Biol. 15 244–248. [PubMed]
  • Sablowski, R. (2007). Flowering and determinacy in Arabidopsis. J. Exp. Bot. 58 899–907. [PubMed]
  • Sakai, H., Krizek, B.A., Jacobsen, S.E., and Meyerowitz, E.M. (2000). Regulation of SUP expression identifies multiple regulators involved in Arabidopsis floral meristem development. Plant Cell 12 1607–1618. [PMC free article] [PubMed]
  • Sakai, H., Medrano, L.J., and Meyerowitz, E.M. (1995). Role of SUPERMAN in maintaining Arabidopsis floral whorl boundaries. Nature 378 199–203. [PubMed]
  • Schoof, H., Lenhard, M., Haecker, A., Mayer, K.F., Jurgens, G., and Laux, T. (2000). The stem cell population of Arabidopsis shoot meristems is maintained by a regulatory loop between the CLAVATA and WUSCHEL genes. Cell 100 635–644. [PubMed]
  • Schultz, E.A., Pickett, F.B., and Haughn, G.W. (1991). The FLO10 gene product regulates the expression domain of homeotic genes AP3 and PI in Arabidopsis flowers. Plant Cell 3 1221–1237. [PMC free article] [PubMed]
  • Shannon, S., and Meeks-Wagner, D.R. (1993). Genetic interactions that regulate inflorescence development in Arabidopsis. Plant Cell 5 639–655. [PMC free article] [PubMed]
  • Sieburth, L.E., Drews, G.N., and Meyerowitz, E.M. (1998). Non-autonomy of AGAMOUS function in flower development: use of a Cre/loxP method for mosaic analysis in Arabidopsis. Development 125 4303–4312. [PubMed]
  • Sieburth, L.E., Running, M.P., and Meyerowitz, E.M. (1995). Genetic separation of third and fourth whorl functions of AGAMOUS. Plant Cell 7 1249–1258. [PMC free article] [PubMed]
  • Smyth, D.R., Bowman, J.L., and Meyerowitz, E.M. (1990). Early flower development in Arabidopsis. Plant Cell 2 755–767. [PMC free article] [PubMed]
  • Sommer, H., Beltran, J.P., Huijser, P., Pape, H., Lonnig, W.E., Saedler, H., and Schwarz-Sommer, Z. (1990). Deficiens, a homeotic gene involved in the control of flower morphogenesis in Antirrhinum majus: The protein shows homology to transcription factors. EMBO J. 9 605–613. [PMC free article] [PubMed]
  • Szecsi, J., Joly, C., Bordji, K., Varaud, E., Cock, J.M., Dumas, C., and Bendahmane, M. (2006). BIGPETALp, a bHLH transcription factor is involved in the control of Arabidopsis petal size. EMBO J. 25 3912–3920. [PMC free article] [PubMed]
  • Theissen, G., Becker, A., Winter, K.U., Munster, T., Kirchner, C., and Saedler, H. (2002). How the land plants learned their floral ABCs: The role of MADS-box genes in the evolutionary origin of flowers. In Developmental Genetics and Plant Evolution, Q.C.B. Cronk, R.M. Bateman, and J.A. Hawkins, eds (London: Taylor and Francis), pp. 173–205.
  • Trobner, W., Ramirez, L., Motte, P., Hue, I., Huijser, P., Lonnig, W.E., Saedler, H., Sommer, H., and Schwarz-Sommer, Z. (1992). GLOBOSA: A homeotic gene which interacts with DEFICIENS in the control of Antirrhinum floral organogenesis. EMBO J. 11 4693–4704. [PMC free article] [PubMed]
  • Wellmer, F., Riechmann, J.L., Alves-Ferreira, M., and Meyerowitz, E.M. (2004). Genome-wide analysis of spatial gene expression in Arabidopsis flowers. Plant Cell 16 1314–1326. [PMC free article] [PubMed]
  • Yanofsky, M.F., Ma, H., Bowman, J.L., Drews, G.N., Feldmann, K.A., and Meyerowitz, E.M. (1990). The protein encoded by the Arabidopsis homeotic gene agamous resembles transcription factors. Nature 346 35–39. [PubMed]
  • Zhao, L., Kim, Y., Dinh, T.T., and Chen, X. (2007). miR172 regulates stem cell fate and defines the inner boundary of APETALA3 and PISTILLATA expression domain in Arabidopsis floral meristems. Plant J. 51 840–849. [PMC free article] [PubMed]

Articles from The Plant Cell are provided here courtesy of American Society of Plant Biologists
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Gene
    Gene links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • HomoloGene
    HomoloGene links
  • MedGen
    Related information in MedGen
  • Pathways + GO
    Pathways + GO
    Pathways, annotations and biological systems (BioSystems) that cite the current article.
  • PubMed
    PubMed citations for these articles
  • Taxonomy
    Related taxonomy entry
  • Taxonomy Tree
    Taxonomy Tree

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...