The transition from vegetative to reproductive phases during Arabidopsis development is the result of a complex interaction of environmental and endogenous factors. One of the key regulators of this transition is LEAFY (LFY), whose threshold levels of activity are proposed to mediate the initiation of flowers. The closely related APETALA1 (AP1) and CAULIFLOWER (CAL) meristem identity genes are also important for flower initiation, in part because of their roles in upregulating LFY expression. We have found that mutations in the FRUITFULL (FUL) MADS-box gene, when combined with mutations in AP1 and CAL, lead to a dramatic non-flowering phenotype in which plants continuously elaborate leafy shoots in place of flowers. We demonstrate that this phenotype is caused both by the lack of LFY upregulation and by the ectopic expression of the TERMINAL FLOWER1 (TFL1) gene. Our results suggest that the FUL, AP1 and CAL genes act redundantly to control inflorescence architecture by affecting the domains of LFY and TFL1 expression as well as the relative levels of their activities.

Postembryonic development in Arabidopsis proceeds through a series of phases, each characterized by the identity of the lateral primordia produced by the shoot apical meristem (SAM) (Poethig, 1990). During the vegetative phase, the SAM produces closely spaced leaf primordia, each subtending a secondary shoot meristem, to form a rosette. During the reproductive, or inflorescence (I) phase, the SAM produces determinate floral meristems on its flanks. The last few vegetative leaves produced are referred to as cauline leaves and become separated along the inflorescence stem by longer internode distances. Thus, the production of leaves can be considered to occur within two distinct subphases, V1 (rosette) and V2 (cauline).

Genes that promote flowering in Arabidopsis were identified as mutations that extend the duration of the V phase, increasing the number of leaves formed before the development of flowers, but generally not affecting the fate of the lateral primordia produced during the I phase (reviewed by Piñeiro and Coupland, 1998). Another group of genes, including TERMINAL FLOWER1 (TFL1), act by delaying phase change and preventing the normally indeterminate SAM from becoming a flower (Alvarez et al., 1992; Shannon and Meeks-Wagner, 1991). In addition, several meristem-identity genes are responsible for conferring floral characteristics to the lateral primordia produced by the SAM during the I phase. Mutations in floral meristem identity genes cause primordia that would develop into flowers to instead develop shoot characteristics. The best characterized of these genes are LEAFY (LFY), APETALA1 (AP1), APETALA2 (AP2) and CAULIFLOWER (CAL) (for review, see Yanofsky, 1995). Only lfy and ap1 mutants show dramatic flower-to-shoot phenotypes, especially in the most basal nodes. Furthermore, the nearly complete conversion of flowers into shoots observed in lfy ap1 double mutants reveals that they act redundantly to specify meristem fate (Bowman et al., 1993; Huala and Sussex, 1992; Irish and Sussex, 1990; Schultz and Haughn, 1991; Shannon and Meeks-Wagner, 1993; Weigel et al., 1992). Together, the LFY, AP1, CAL and AP2 genes appear to reinforce each other’s activities leading to a sharp transition from vegetative to reproductive development.

The FRUITFULL (FUL) gene encodes a MADS-box protein that has previously been shown to be required for carpel and fruit development (Gu et al., 1998; Mandel and Yanofsky, 1995a). However, in addition to its expression domain during carpel and fruit development, the FUL gene is upregulated in the SAM at around the transition to flowering, suggesting that it may also play a role during this transition (Mandel and Yanofsky, 1995a; Hempel et al., 1997). FUL is closely related to the meristem identity genes AP1 and CAL, suggesting the possibility of functionally redundant activities.

In this work we have undertaken a molecular genetic approach to uncover the possible roles of FUL in the transition to flowering as well as its interactions with different meristem identity genes. We have found that in addition to its role during carpel and fruit development, FUL acts as a flowering-time and meristem-identity gene. These studies provide new insights into the functional redundancy of MADS-box genes during the transition to flowering and on the upregulation of the LFY meristem identity gene.

Plant material and growth conditions

The ap1-1, ful-1, tfl1-2 and lfy-26 alleles have been described previously (Bradley et al., 1997; Gu et al., 1998; Lee et al., 1997; Mandel et al., 1992). The cal-5 allele was generated in a γ-irradiation mutagenesis experiment and contains a single base-pair deletion 33 bp downstream of the translation initiation codon that causes a frame shift and introduces a STOP codon 19 amino acids later (Savidge, 1996). 35S::LFY lines (DW151.2.5, in Landsberg erecta background; Weigel and Nilsson, 1995) and LFY::GUS (DW150.209, in Columbia; Blázquez et al., 1997) were kindly provided by Detlef Weigel. The 35S::AG line was obtained from Hong Ma (Mizukami and Ma, 1992). For all experiments, seeds were vernalized for 3-5 days at 4°C, then germinated and grown at 22-24°C under continuous light conditions.

Characterization of the molecular lesions in the ful alleles

For ful-2, ful-4, ful-5 and ful-6, genomic DNAs were amplified by PCR with the primers OAM25 (5′-GGTCATTTCAGGGTTGT-CGGTT-3′) and OAM14 (5′-AATCATTACCAAGATATGAA-3′), which hybridize respectively 59 ncl upstream of the initiation codon and 202 ncl downstream of the STOP codon of the FUL gene. The amplification products of two independent reactions were sequenced and compared with the wild-type sequence for each allele. For ful-5, since the sequencing of the FUL genomic DNA only showed a silent change in the coding region, we analyzed the sequence of the transcribed RNA by performing a reverse transcription of the ful-5 RNA using OAM14 as a primer, coupled with a PCR amplification using OAM25 and OAM14 as primers.

GUS activity measurements

For quantitative measurements of GUS activity in LFY::GUS ful-2,plants, the assay described by Blázquez et al. (1997) was used.

In situ hybridizations

For in situ experiments at day 12 in ap1 cal and ful ap1 cal plants, genotyping for the presence of the ful-1 allele was necessary since double and triple mutants were indistinguishable (see below).

Tissue was fixed for 2 hours at room temperature in FAE solution (ethanol:acetic acid:formaldehyde:water, 50:5:3.5:41.5, v/v/v/v), dehydrated, embedded and sectioned to 8 μm. After dewaxing in histoclear and rehydration, sections were treated for 20 minutes in 0.2 M HCl, neutralized for 10 minutes in 2′ SSC and incubated for 30 minutes with 1 μg/ml Proteinase K at 37°C. Proteinase action was blocked with 5 minutes incubation in 2 mg/ml Gly and 10 minutes postfixation in 4% formaldehyde. Tissue sections were washed in PBS, dehydrated through an ethanol series and dried under vacuum before applying the hybridization solution (100 μg/ml tRNA; 6′ SSC; 3% SDS; 50% formamide, containing approx. 100 ng/μl of antisense DIG-labeled RNA probe). Sections were hybridized overnight at 52°C, washed twice for 90 minutes in 2′ SSC; 50% formamide at 52°C and the antibody incubation and color detection was performed according to the manufacturer instructions (Boehringer). The probes were synthesized as previously described using plasmids pDW122 (LFY; Weigel et al., 1992), pCIT565 (AG; Drews et al., 1991), pD793 (AP3, Jack et al., 1992) and pSL66 (TFL1; Liljegren et al., 1999).

For double labeling experiments, a DIG-labeled TFL1 probe and a fluorescein-labeled LFY probe were both added to the hybridization solution. Washes, DIG-antibody incubation and color detection with NBT-BCIP as substrates were performed as described above to reveal TFL1 expression as a blue precipitate. Slides were treated in 2′ SSC for 2 hours at 65°C to inactivate the alkaline phosphatase coupled to the DIG-antibody and then incubated with fluorescein-antibody. Color detection was performed according to the manufacturer instructions (Boehringer) using as a substrate Fast Red tablets to reveal LFY expression as a red signal.

Scanning electron microscopy (SEM)

Inflorescences were collected, fixed and observed as previously described (Gu et al., 1998).

Generation and identification of multiple mutants

In all combinations the ful-1 allele was used, except for the generation of LFY::GUS ful-2 lines. ful-1 carries a Ds:GUS element that allows the identification of the mutant allele by assaying GUS activity in cauline leaves. ful-1 was crossed as female to homozygous mutants ap1-1, ap1-1 cal-5 and tfl1-2 and double/triple mutants were identified in F2 populations as new/additive phenotypes segregating in a 1:16 ratio (1:64 for ful ap1 cal).

Since LFY and FUL are closely linked (approximately 1 cM), ful-1 pollen was crossed onto lfy-26 homozygous plants and GUS detection was performed on F2 plants with lfy phenotype until one positive was found, and pollinated with ful-1 pollen. In the F1 from this cross, plants with ful phenotype were selected as ful-1 lfy-26/+ and selfed, and F2 lfy plants were assumed to be homozygous for both mutations.

lfy-26 ap1-1 cal-5 ful-1 were generated by crossing ap1-1 cal-5 onto ful-1 lfy-26. Plants with ap1 cal phenotypes in the F2 generation were tested for GUS activity, and the positives were selfed and assumed to have an ap1 cal (ful lfy)/++ genotype. The F3 populations resulting from these selfed individuals segregated plants with ap1 cal and ap1 lfy-like phenotypes in a 3:1 ratio, and the latter ones were assumed to be the quadruple mutant combination, since given the close linkage between LFY and FUL, 99% of the F3 lfy plants would be ful homozygous.

For the generation of tfl1-2 ap1-1 cal-5 ful-1 quadruple mutants, tfl1-2 ful-1 plants were used as females for ap1-1 cal-5 pollen. ful phenotypes were selected and selfed in the F1 generation and in the F2, since no new phenotypes were observed, the presence of the cal-5 allele was analyzed in plants with ap1 ful tfl phenotype as described below.

For 35S::LFY ap1-1 cal-5 ful-1, 35S::LFY plants were used as pollen donors to fertilize ap1-1 cal-5 ful-1/+ emasculated flowers. 35S::LFY phenotypes with GUS activity were selected and selfed in the F1 generation, and in the F2, plants with ful fruit phenotypes were genotyped for the presence of 35S::LFY, ap1-1 and cal-5 and further reconfirmed by the segregation of the selfed F3 progenies. For 35S::AG ap1-1 cal-5 ful-1, the same strategy was adopted, except that in the F2 population, a new phenotype was identified as the quadruple combination.

Genotyping

To genotype plants for the presence of the ap1-1 allele, we used dCAPS markers (Neff et al., 1998). The ap1-1 mutation introduces a change from G to A at the acceptor site before exon 4 (Mandel et al., 1992). For genotyping, genomic DNA was amplified by PCR using the primers 5′-GCAAGTCTTCCCCAAGATAAGGC-3′ and 5′-GACAGCTTATTGCACCTGAG-3′. The restriction enzyme StuI cleaves only wild-type DNA yielding a 290 and a 22 bp fragment.

For cal-5 genotyping, a FokI RFLP was used. Genomic DNA was amplified by PCR with primers 5′-ATGGGAAGGGG-TAGGGTTG-3′ and 5′-ATTCAGAG-GAGTACTCGAAC-3′. Digestion with FokI generates two fragments of 42 and 140 bp in cal-5 DNA, while wild-type DNA is not digested.

ful-1 plants were genotyped by PCR on genomic DNA with the primer AGL8PG: 5′-TGTATTCACGTCACATACCG-3′, located in the promoter region of the FUL gene, and primers AGL8MG: 5′-CTCATGAGC-TTTCTTGAGC-3′ and GUS3: 5′-CTTGTAACG-CGCTTTCCC-3′, located respectively in the coding regions of the FUL gene and the GUS gene of the Ds-element inserted in the ful-1 allele. ful-1 homozygotes were identified by the detection of an amplification product with primers AGL8PG/GUS3, and no product with the pair AGL8PG/AGL8MG.

The presence of the 35S:LFY transgene was assessed by PCR on genomic DNA with the primers 5′-ACCCAAGCTTCT-TCGCAAGACCCTTCCTCT-3′, located in the 35S promoter region, and 5′-AACTAGAAACGCAAGTCG-3′, located in the LFY coding region, which only amplified the transgene and no endogenous sequences.

The fruitfull phenotypes

The ful-1 mutation affects several aspects of plant development. The most dramatic effect is observed within the carpel where, from stage 12 (Smyth et al., 1990), the valve cells fail to elongate and differentiate, and the replum cells grow with an altered morphology. As a result, silique development is severely affected, leading to short fruits with small crowded seeds (Fig. 1A,J,N; Gu et al., 1998). In addition to the fruit phenotype, ful-1 cauline leaves are also affected in shape, cell organization and vascular differentiation (Gu et al., 1998).

Fig. 1.

Scanning electron micrographs of fruitfull mutants showing fruits 10 days after pollination (late stage 17; according to Smyth et al., 1990). (A-E) ful allelic series. (F-I) Style phenotypes from four different ful alleles. (F and I) ful-1 and ful-4 respectively, both in the Landsberg erecta background. (G) A ful-2 style, in Columbia background and (H) a ful-3 style, in Nossen background. Close-up of (J-M) the valve-replum boundary and (N-Q) the valve epidermal cells in four different ful mutants as indicated. Scale bars, 500 μm (A-E), 200 μm (F-I), 100 μm (J-M) and 50 μm (N-Q).

Fig. 1.

Scanning electron micrographs of fruitfull mutants showing fruits 10 days after pollination (late stage 17; according to Smyth et al., 1990). (A-E) ful allelic series. (F-I) Style phenotypes from four different ful alleles. (F and I) ful-1 and ful-4 respectively, both in the Landsberg erecta background. (G) A ful-2 style, in Columbia background and (H) a ful-3 style, in Nossen background. Close-up of (J-M) the valve-replum boundary and (N-Q) the valve epidermal cells in four different ful mutants as indicated. Scale bars, 500 μm (A-E), 200 μm (F-I), 100 μm (J-M) and 50 μm (N-Q).

We analyzed five new ful alleles (see Materials and Methods) and examined their phenotypes. ful-4, ful-5 and ful-6 were isolated after EMS mutagenesis in the Landsberg erecta (Ler) background. The strong ful-5 allele, which has a phenotype very similar to ful-1, has a single base pair change from C to T in position 156 of the first exon. This mutation results in the generation of an alternate donor splicing site at position 154 that causes a 31 base deletion in the messenger RNA, a frame shift and a truncated protein. ful-4 and ful-6 contain single base pair mutations from G to A in the 5th and 6th exon-intron boundaries respectively. ful-4 has an intermediate phenotype, with a similar arrested development of the valves, but a less pronounced defect in replum morphology (Fig. 1D,L,P). ful-6 is a weak allele, which causes the valve cells differentiate and expand to almost the same extent as the wild type, and the replum cells undergo mild enlargement, allowing the formation of a functional dehiscence zone in the valve-replum boundaries (Fig. 1E,M,Q).

As a result of these two alleles a striking elongation of the style was observed in genetic backgrounds different from Ler. ful-2 is an EMS mutagenized line in the Columbia (Col) background, and contains a nonsense mutation in exon 3 (W91 to STOP). ful-3 is a Nossen line with a Spm transposon element inserted into the last intron. The ful-2 and ful-3 styles of late stage 17 fruits are 2-3 times longer than the Col wild type (Fig. 1F-H). ful-2 siliques have a less severe phenotype in the valves, whose cells elongate to a small extent allowing a modest expansion of the fruit, although, as in ful-1, no differentiation of stomata is observed. ful-3 is a strong allele producing a similar phenotype to ful-1 in valves and replum. Even though a gradation in the silique phenotypes was observed in the allelic series, the cauline leaves in all mutants were affected similarly, suggesting that the FUL function in leaf morphology requires its complete activity (results not shown).

Mutations in FRUITFULL affect flowering time

Since FUL is strongly upregulated in the shoot apical meristem shortly after the transition to flowering, and in response to photoinductive conditions, a possible role for FUL in promoting the initiation of flowers has been proposed (Hempel et al., 1997; Mandel and Yanofsky, 1995a). To investigate this possibility, we quantified flowering time in ful mutants grown under continuous light. ful plants showed a small, significant delay in the time to flower when measured both in number of leaves (Table 1) and days to bolt (results not shown). The increase in leaf number was observed for both rosette and cauline leaves, and no significant effect of the erecta mutation was observed, since both ful-1 (Ler) and ful-2 (Col) mutants flowered later than the corresponding wild-type ecotypes. These results suggest that FUL acts to promote the flowering pathway, although, given the slight effect in flowering time, it may be involved in a highly redundant network of signaling.

Table 1.

Time to flowering under continuous light conditions

Time to flowering under continuous light conditions
Time to flowering under continuous light conditions

FUL and LFY act in parallel pathways

One candidate for a gene that interacts with FUL in this flower-promoting pathway is LFY, as LFY levels have been shown to be important for the transition to flowering (Blázquez et al., 1997). lfy mutants have an extended V2 phase and are defective in floral meristem identity specification (Huala and Sussex, 1992; Schultz and Haughn, 1991; Weigel et al., 1992). In response to photoinductive treatments, LFY and FUL are upregulated at the shoot apex in an overlapping pattern (Blázquez et al., 1997; Hempel et al., 1997).

To investigate whether the late flowering phenotype in ful mutants might be caused by a delay in LFY upregulation or a reduction in its levels, we introduced a LFY::GUS transgene into ful-2 plants (see Methods). We did not observe any effect of the ful mutation on the upregulation of LFY (not shown), suggesting that either LFY is not directly regulated by FUL, or that other genes can substitute for FUL to upregulate LFY. In addition, we generated ful-1 lfy-26 double mutants, which flowered later than the corresponding lfy-26 and ful-1 single mutants (Table 1). Moreover, a similar quantitative effect was observed in ful-1 plants heterozygous for lfy-26, which flowered significantly later than either ful-1 mutants or LFY/lfy plants, with a more dramatic increase in the duration of the V2 phase as scored by the number of cauline leaves produced (Table1), suggesting that LFY and FUL act in parallel pathways or at least influence the same process independently. ful-1 lfy-26 plants also showed an additive phenotype in the floral identity defects, with the same kind of flower-to-shoot transformations as in lfy-26, and the ful-1 phenotype in the carpel-like organs that eventually formed (not shown).

Flowering is eliminated in ful ap1 cal triple mutants

The AP1 gene is required for both flower meristem and flower organ identity. (Irish and Sussex, 1990; Bowman et al., 1993). Mutations in the closely related and partially redundant CAL gene enhance the ap1 mutant phenotype, such that ap1 cal double mutants proliferate inflorescence meristems in positions that would normally be occupied by flowers, resulting in a ‘cauliflower’ appearance. Eventually, however, flower meristems are specified and ap1-like flowers are produced (Bowman et al., 1993; Kempin et al., 1995; Fig. 2A,C,E). The AP1 gene also plays a role in negatively regulating FUL in emerging flower primordia, as FUL is ectopically expressed in ap1 and ap1 cal mutant flower meristems (Mandel and Yanofsky, 1995a; M. A. Mandel, C. F. and M. F. Y, unpublished observations).

Fig. 2.

Mutant phenotypes of ful ap1 cal and ap1 cal plants. (A-B) Apical inflorescences of 3-week-old ap1 cal (A) and ful ap1 cal (B) plants. (C-D) Apical inflorescences of 5-week-old ap1 cal (C) and ful ap1 cal (D) plants. Mature floral organs are visible in ap1 cal, but absent in ful ap1 cal inflorescences. (E) An ap1 cal plant, 8 weeks after germination. Mature siliques have already developed. (F) A ful ap1 cal plant, 8 weeks after germination. The inflorescence structures have branched out producing numerous leafy shoots, but no flower structures are evident.

Fig. 2.

Mutant phenotypes of ful ap1 cal and ap1 cal plants. (A-B) Apical inflorescences of 3-week-old ap1 cal (A) and ful ap1 cal (B) plants. (C-D) Apical inflorescences of 5-week-old ap1 cal (C) and ful ap1 cal (D) plants. Mature floral organs are visible in ap1 cal, but absent in ful ap1 cal inflorescences. (E) An ap1 cal plant, 8 weeks after germination. Mature siliques have already developed. (F) A ful ap1 cal plant, 8 weeks after germination. The inflorescence structures have branched out producing numerous leafy shoots, but no flower structures are evident.

The close sequence similarity of FUL, AP1 and CAL (Mandel and Yanofsky, 1995a), together with the fact that FUL is expressed throughout the proliferating meristems of ap1 cal double mutants, suggested the possibility that the ability of ap1 cal mutants to eventually form flowers could be the result of ectopic FUL activity. In order to test this hypothesis, we introduced the ful-1 mutation into ap1-1 cal-5 mutant plants. We found that mutations in ful dramatically enhanced the ap1 cal double mutant phenotype such that the triple mutant plants proliferated leafy shoots and failed to flower. After bolting, the ful ap1 cal shoot apical meristem gave rise to cauline leaves with associated axillary meristems which repeated this pattern to form leafy shoots with small axillary ‘cauliflowers’ along the main inflorescence, which as a consequence, branched out extensively to adopt a bushy appearance (Fig. 2B,D,F).

A closer inspection by SEM of ap1 cal and ful ap1 cal inflorescences revealed differences in the behavior of the proliferating meristems. Early in their development, ap1 cal double mutants displayed the conversion of flowers into inflorescence meristems. In 21-day-old plants, fifth order inflorescence meristems arranged in a spiral phyllotaxy were apparent and some of the older meristems had visibly acquired a flower identity (Fig. 3A). After 35 days of growth, flowers with developing stamens and carpels were visible in the lower branches, and in the more apical positions new flowers were differentiating (Figs 2C, 3C). After 2 months of growth, the meristems had stopped proliferating and fully developed siliques were observed (Fig. 2E).

Fig. 3.

Scanning electron micrographs of ap1 cal (A,C) and ful ap1 cal (B,D) apical meristems. In 21-day-old ap1 cal inflorescences (A), floral meristems (*) up to stage 3 of development (according to Smyth et al., 1990) are visible. ful ap1 cal apices of the same age (B) initiate a higher number of cauline leaves with axillary meristems (arrows) and no floral meristems are visible. 35-day-old ap1 cal inflorescences (C) are composed of floral meristems at several developmental stages (*), some with characteristics of inflorescence meristems, such a spiral phyllotaxy. 35-day-old ful ap1 cal apices (D) show a proliferation of mainly leaf primordia with axillary meristems (arrows), with no recognizable phyllotactic arrangement. Scale bars, 100 μm.

Fig. 3.

Scanning electron micrographs of ap1 cal (A,C) and ful ap1 cal (B,D) apical meristems. In 21-day-old ap1 cal inflorescences (A), floral meristems (*) up to stage 3 of development (according to Smyth et al., 1990) are visible. ful ap1 cal apices of the same age (B) initiate a higher number of cauline leaves with axillary meristems (arrows) and no floral meristems are visible. 35-day-old ap1 cal inflorescences (C) are composed of floral meristems at several developmental stages (*), some with characteristics of inflorescence meristems, such a spiral phyllotaxy. 35-day-old ful ap1 cal apices (D) show a proliferation of mainly leaf primordia with axillary meristems (arrows), with no recognizable phyllotactic arrangement. Scale bars, 100 μm.

By contrast, the ful ap1 cal meristems at 21 days typically produced cauline leaf primordia with associated axillary meristems as well as two or three additional meristems not subtended by leaves, yielding a phyllotaxy intermediate between spiral and cruciform (Fig. 3B). At these and all later time points, no floral structures were evident, and the proliferating meristems gave rise to new meristems and cauline leaves at a lower rate, losing all phyllotactic arrangement (Fig. 3D). The ful ap1 cal plants kept under our standard growth room conditions failed to produce any kind of flower structures even six months after germination, and remained in this vegetative state until they died. Interestingly, we did observe on one occasion, when accidental overheating of the growth room occurred, that some of these plants formed a few floral organs after several months of vegetative development. These results suggest that there may be a threshold required to induce flowering, and that although the triple mutant is normally just below this threshold, conditions such as heat stress can allow the requirement for FUL/AP1/CAL to be overcome.

In contrast to the enhancement of the ap1 cal double mutant phenotype, we did not observe enhancement of the ap1 mutant phenotype when ful-1 ap1-1 double mutants were generated. The phenotype of the double mutants was found to be strictly additive for both flowering time and flower and fruit development (results not shown). Likewise, ful-1 cal-5 double mutants were indistinguishable from the ful-1 single mutants, consistent with previous observations showing that mutations in CAL do not confer a mutant phenotype in the presence of a functional copy of AP1. These results indicated that both AP1 and CAL are able to compensate for the loss of FUL function in specifying floral meristem identity.

Overexpression of LFY rescues the non-flowering phenotype of ful ap1 cal plants

The non-flowering phenotype of ful ap1 cal mutants is likely the result of reduced activity of the LFY meristem identity gene product, and there are two distinct explanations for how FUL/AP1/CAL can contribute to LFY activity. One scenario is that LFY RNA expression in the ful ap1 cal triple mutant may not reach the threshold required for flower specification. An alternative scenario is that the activity of the LFY protein may require one of the AP1, CAL or FUL functions, for example as a cofactor, even though LFY RNA levels in the triple mutant may be sufficient to promote flowering.

To distinguish between these possibilities, we compared the expression of LFY in ap1 cal double mutants to that in ful ap1 cal triple mutants grown under continuous light. In both ap1 cal and ful ap1 cal plants 14 days after sowing (d14), LFY RNA levels were much lower than in the corresponding wild-type controls (Fig. 4A-C). In d21 ap1 cal inflorescences, floral meristems and flowers up to stage 5 of development could be easily distinguished, whereas in ful ap1 cal mutants, no floral characteristics appeared (Fig. 3A,B). LFY expression was detected at high levels in all floral meristems of ap1 cal plants. In contrast, LFY RNA levels were significantly reduced in ful ap1 cal triple mutants, although the accumulation of LFY RNA was still readily detected (Figs 4E, 6A,B). These results suggests that AP1, CAL and FUL play a redundant role in boosting LFY RNA levels but that other factors are capable of inducing LFY expression initially.

Fig. 4.

Expression of TFL1 and LFY in wild-type, ap1 cal and ful ap1 cal plants. Sections of wild-type apices at day 14 (A), ap1 cal plants at days 14 (B), 21 (D) and 35 (F) and ful ap1 cal plants at days 14 (C), 21 (E) and 35 (G) probed simultaneously with LFY (red label) and TFL1 (blue label) antisense RNA. In day-14 wild-type plants (A) LFY is strongly expressed in flower meristems arising laterally from the inflorescence apex. TFL1 is expressed below the inflorescence meristem in a region not overlapping with LFY. In day-14 ap1 cal plants (B), LFY is weakly expressed and TFL1 is ectopically detected below the lateral meristems initiated. LFY and TFL1 domains of expression overlap in some regions, as described by Ratcliffe et al. (1999). In day-14 ful ap1 cal meristems (C), LFY and TFL1 expression is similar to ap1 cal at the same time point. In day-21 ap1 cal inflorescences (D,E), LFY is detected at high levels in some meristems, presumably those already committed to a floral fate. TFL1 is not found in highly LFY-expressing meristems, and is strongly detected in those putatively behaving as inflorescence meristems. Some regions still show the early pattern of LFY and TFL1 coexpression. Day-21 ful ap1 cal plants still show very low levels of LFY expression, and TFL1 can be detected below most of the meristems initiated, in a similar pattern to earlier time points. At 35 days, ap1 cal inflorescences show high LFY expression in floral meristems and TFL is barely detectable. In 35-day-old ful ap1 cal plants (G), LFY can be detected at similarly low levels, whereas TFL1 expression is reduced and restricted to a few meristems.

Fig. 4.

Expression of TFL1 and LFY in wild-type, ap1 cal and ful ap1 cal plants. Sections of wild-type apices at day 14 (A), ap1 cal plants at days 14 (B), 21 (D) and 35 (F) and ful ap1 cal plants at days 14 (C), 21 (E) and 35 (G) probed simultaneously with LFY (red label) and TFL1 (blue label) antisense RNA. In day-14 wild-type plants (A) LFY is strongly expressed in flower meristems arising laterally from the inflorescence apex. TFL1 is expressed below the inflorescence meristem in a region not overlapping with LFY. In day-14 ap1 cal plants (B), LFY is weakly expressed and TFL1 is ectopically detected below the lateral meristems initiated. LFY and TFL1 domains of expression overlap in some regions, as described by Ratcliffe et al. (1999). In day-14 ful ap1 cal meristems (C), LFY and TFL1 expression is similar to ap1 cal at the same time point. In day-21 ap1 cal inflorescences (D,E), LFY is detected at high levels in some meristems, presumably those already committed to a floral fate. TFL1 is not found in highly LFY-expressing meristems, and is strongly detected in those putatively behaving as inflorescence meristems. Some regions still show the early pattern of LFY and TFL1 coexpression. Day-21 ful ap1 cal plants still show very low levels of LFY expression, and TFL1 can be detected below most of the meristems initiated, in a similar pattern to earlier time points. At 35 days, ap1 cal inflorescences show high LFY expression in floral meristems and TFL is barely detectable. In 35-day-old ful ap1 cal plants (G), LFY can be detected at similarly low levels, whereas TFL1 expression is reduced and restricted to a few meristems.

The reduced levels of LFY RNA accumulation in the triple mutant provide molecular evidence that suggests that the failure to upregulate LFY may be the cause of the non-flowering phenotype. To demonstrate that this is indeed the case, we introduced a constitutively expressed LFY transgene (35S::LFY) into ful ap1 cal triple mutants. In contrast to the triple mutant, 35S::LFY ful ap1 cal plants were able to produce many ful ap1-like flowers, indicating that high levels of LFY expression could overcome the lack of AP1/CAL/FUL functions (Fig. 5A). Taken together, these data demonstrate that the inability of ful ap1 cal plants to flower is due to a reduction of LFY expression.

Fig. 5.

Phenotype of the 35S::LFY ful ap1 cal, tfl1 ful ap1 cal, 35S::AG ap1 cal and 35S::AG ful ap1 cal plants. All plants were grown for 4 weeks. (A) 35S::LFY ful ap1 cal inflorescences produced flowers composed of bract-like sepals, stamens and ful-like carpels, subtended by typical ful cauline leaves. (B) The quadruple tfl1 ful ap1 cal mutants produced flowers similarly to tfl1 ap1 plants. The 35S::AG transgene accelerates flowering in ap1 cal inflorescences (C), which form ap1-like flowers without producing ‘cauliflower’ structures first (as described in Mizukami and Ma, 1997). 35S::AG ful ap1 cal plants (D) did not flower even after several months of growth. The inflorescences resemble those of ful ap1 cal triple mutants, but the leafy organs are reduced in size and curled upwards, a typical effect of the ectopic AG activity.

Fig. 5.

Phenotype of the 35S::LFY ful ap1 cal, tfl1 ful ap1 cal, 35S::AG ap1 cal and 35S::AG ful ap1 cal plants. All plants were grown for 4 weeks. (A) 35S::LFY ful ap1 cal inflorescences produced flowers composed of bract-like sepals, stamens and ful-like carpels, subtended by typical ful cauline leaves. (B) The quadruple tfl1 ful ap1 cal mutants produced flowers similarly to tfl1 ap1 plants. The 35S::AG transgene accelerates flowering in ap1 cal inflorescences (C), which form ap1-like flowers without producing ‘cauliflower’ structures first (as described in Mizukami and Ma, 1997). 35S::AG ful ap1 cal plants (D) did not flower even after several months of growth. The inflorescences resemble those of ful ap1 cal triple mutants, but the leafy organs are reduced in size and curled upwards, a typical effect of the ectopic AG activity.

It has been previously described that ap1 mutations can largely suppress the early flowering phenotype conferred by the 35S::LFY transgene (Mandel and Yanofsky, 1995b; Weigel and Nilsson, 1995). The phenotypes of the 35S::LFY, 35S::LFY ap1, 35S::LFY ap1 cal and 35S::LFY ful ap1 cal showed a gradation in flowering time (Table 2) and number of flowers (not shown) produced by the shoot apical meristem before terminating. These results suggest that the early flowering and shoot-to-flower conversion caused by the 35S::LFY transgene is mostly due to the subsequent activities of FUL/AP1/CAL.

Table 2.

Effect of the ap1/cal/ful mutations on the 35S::LFY phenotype

Effect of the ap1/cal/ful mutations on the 35S::LFY phenotype
Effect of the ap1/cal/ful mutations on the 35S::LFY phenotype

Because low levels of LFY RNA accumulate in ful ap1 cal triple mutants, it was unclear if LFY was still active in the triple mutant. We therefore introduced the strong lfy-26 allele into the triple mutant and found the quadruple mutants to be very similar to lfy ap1 doubles or lfy ap1 cal triples (Bowman et al., 1993). The plants showed enhanced vegetative characteristics, developing secondary shoots in place of flowers, subtended by the typical ful-1 cauline leaves (Gu et al., 1998). The proliferation of ‘leafy cauliflowers’ that occurs in ful ap1 cal triple mutant does not occur in lfy ful ap1 cal quadruple mutants. Instead, the quadruple mutant develops vegetative shoots, indicating that some LFY activity is needed to cause a reiterative pattern of meristem proliferation. In addition, the similar phenotypes of lfy ap1 and lfy ful ap1 cal indicate that in an ap1 lfy background, FUL and CAL are not able to specify floral meristem identity.

Mutations in TFL1 suppress the nonflowering phenotype of ful ap1 cal plants

A number of studies have demonstrated an antagonistic interaction between TFL1 and the floral meristem identity genes (Mandel and Yanofsky, 1995b; Savidge, 1996; Weigel and Nilsson, 1995; Weigel et al., 1992; Lijegren et al., 1999; Ratcliffe et al., 1999). To further explore these interactions, we compared the expression patterns of TFL and LFY in the meristems of ap1 cal mutants to that in the ful ap1 cal triple mutants. LFY is strongly expressed in wild-type flower meristems 14 days after sowing (d14), but is only expressed at low levels in equivalently staged ap1 cal double mutants. One week later (d21), when flower meristems become distinct in ap1 cal mutants, LFY RNA is detected at higher levels and its expression increases at later time points when flowers are already apparent (d35) (Figs 4B,D,F, 6A). In contrast, LFY is only detected at low levels in ful ap1 cal mutants, even after two months of growth (Figs 4C,E,G, 6B).

TFL1 is normally expressed in the center of the inflorescence apex and is not detected in the lateral meristems committed to a floral fate (Fig. 4A). TFL1 expression was found at comparable levels and pattern in ap1 cal and ful ap1 cal plants at d14, below the shoot apex and, ectopically, in the laterally arising meristems (Fig. 4B,C). Similar domains of TFL1 expression were found in ap1 cal and ful ap1 cal meristems at d21; in both cases, TFL1 was expressed more strongly in basal positions and at lower levels in the more apical meristems At d35, TFL1 was almost undetectable in ap1 cal inflorescences, where LFY expression had reached its highest level, whereas some small domains of TFL expression and lower levels of LFY could still be seen in ful ap1 cal (Fig. 4F,G). Thus, the mutually exclusive patterns of LFY and TFL1 expression became overlapping both in ap1 cal and ful ap1 cal meristems, even though the ratio of LFY to TFL1 RNA expression was clearly higher in ap1 cal than in ful ap1 cal meristems from d21 (Fig. 4D,E).

Since TFL1 RNA accumulates in the ful ap1 cal triple mutant, we examined the activity of TFL1 in this context by introducing the tfl1 mutation into the triple mutant. ful ap1 cal tfl1 plants were only distinguishable from ap1 tfl1 plants by the ful fruit phenotype, indicating that the lack of TFL1 completely abolished the proliferation of meristems observed in ful ap1 cal plants (Fig. 5B). We also observed that, as in wild type, in the ful ap1 cal background, the tfl1 mutation had a semidominant phenotype. ful ap1 cal TFL1/tfl1 plants were initially indistinguishable from the ful ap1 cal triples, but, after 2 months of growth, they all were able to form some flowers and set seeds. The quantitative effect of TFL1 on the ful ap1 cal phenotype might reflect a threshold ratio of LFY:TFL1 activities that has to be reached to induce flowering.

ful ap1 cal meristems are not competent to respond to AG floral inductive activity

We have shown that LFY RNA levels are significantly reduced in ful ap1 cal triple mutants, and previous studies have demonstrated that LFY plays a key role in the upregulation of the AP3 and AG organ identity genes. We therefore compared the expression of AP3 and AG in the ful ap1 cal triple mutant to that in the ap1 cal double mutant to determine if the expression of these organ identity genes requires AP1/CAL/FUL. Whereas the accumulation of AP3 and AG RNAs was readily detected in 21-day-old ap1 cal meristems, these RNAs were not detected in ful ap1 cal plants of the same age (Fig. 6C-F). Even after more than 2 months of growth (not shown), no expression of AP3 and AG could be detected in the triple mutant. These results demonstrate that FUL, AP1 and CAL have overlapping functions in the upregulation of AP3 and AG, most likely mediated at least in part through the upregulation of LFY.

Fig. 6.

Expression of LFY, AP3 and AG in ap1 cal and ful ap1 cal plants. Sections of 21-day-old ap1 cal (A,C,E) and ful ap1 cal (B,D,F) plants probed with LFY (A,B), AP3 (C,D) and AG (E,F) antisense RNA are shown. Sections in A,C,E and B,D,F are from single inflorescences. LFY is strongly expressed in the presumed floral meristems of ap1 cal inflorescences. In ful ap1 cal meristems, LFY is detected at much reduced levels. (C,G) Some meristems in ap1 cal inflorescences show AP3 (C) and AG (E) patterns of expression similar to those found in early stages (4-5) of wild-type development. (D,F) AP3 (D) and AG (F) are not detected in ful ap1 cal meristems.

Fig. 6.

Expression of LFY, AP3 and AG in ap1 cal and ful ap1 cal plants. Sections of 21-day-old ap1 cal (A,C,E) and ful ap1 cal (B,D,F) plants probed with LFY (A,B), AP3 (C,D) and AG (E,F) antisense RNA are shown. Sections in A,C,E and B,D,F are from single inflorescences. LFY is strongly expressed in the presumed floral meristems of ap1 cal inflorescences. In ful ap1 cal meristems, LFY is detected at much reduced levels. (C,G) Some meristems in ap1 cal inflorescences show AP3 (C) and AG (E) patterns of expression similar to those found in early stages (4-5) of wild-type development. (D,F) AP3 (D) and AG (F) are not detected in ful ap1 cal meristems.

These results raise the question of whether the inability to produce floral structures in ful ap1 cal mutants is due simply to the reduced LFY expression, or is due in part to the loss of organ-identity gene activation. Related to this is the observation that AG promotes floral identity even in the absence of LFY and AP1 functions and is necessary to maintain floral identity under non-inductive conditions (Mizukami and Ma, 1997). To test whether the lack of the AG floral promoting activity was the direct cause of the non-flowering phenotype, we introduced a 35S::AG transgene in ful ap1 cal plants. Constitutive expression of AG did not induce the ful ap1 cal plants to flower, indicating that the loss of flowering in the triple mutant is not simply due to a failure to upregulate AG (Fig. 5C,D) and that AG requires at least one of the FUL/AP1/CAL activities to promote a floral fate.

FUL controls several aspects of plant development

We have shown that FRUITFULL is involved in several distinct processes during Arabidopsis development, as was suggested by its complex pattern of expression (Mandel and Yanofsky, 1995a). FUL has an early function in controlling flowering time, meristem identity and cauline leaf morphology, and has a later role in carpel and fruit development that affects valve, replum and style morphology. These phenotypic effects correlate well with the biphasic pattern of FUL expression (Gu et al., 1998; Mandel and Yanofsky, 1995a) and reveal a certain degree of non-autonomy of the FUL function since the mutants have a clear phenotype in the replum region where no FUL RNA is detected. Moreover, the study of an allelic series shows that these different roles are separable. All of the mutant alleles have a similar effect on flowering time (C. F. and M. Y., unpublished observations) and cauline leaf morphology. However, the weak alleles display a much less severe phenotype in valve and replum morphology as compared to strong alleles.

It has been proposed that a threshold level of LFY expression is required for the vegetative-to-floral transition (Blázquez et al., 1997). Our results indicate that this threshold level is not reached in ful ap1 cal plants, leading to a dramatic non-flowering phenotype. The initial activation of LFY does not depend on FUL/AP1/CAL, but these activities are necessary for its subsequent upregulation, since LFY expression in ful ap1 cal meristems never exceeds its initially low levels. Moreover, constitutive expression of LFY restored flowering in the ful ap1 cal plants, reinforcing the idea of LFY levels as the switch to reproductive development. Our results agree with previous studies that identified AP1 and CAL as direct or indirect activators of LFY. These studies showed that LFY is ectopically expressed in the converted flower meristems of 35S::AP1 plants (Liljegren et al., 1999) and that the initial expression of LFY is significantly reduced in ap1 cal mutants (Bowman et al., 1993).

It is interesting to note that although constitutive expression of LFY suppressed the non-flowering phenotype of ful ap1 cal triple mutants, many of the phenotypes caused by the ectopic LFY expression were suppressed in the triple mutant background. It has been previously reported that the flower-to-shoot conversion in 35S::LFY plants is largely suppressed by mutations in AP1, although the plants are still early flowering (Weigel and Nilsson, 1995). We found that the flowering time in 35S::LFY ap1 cal plants is further increased and that 35S::LFY ful ap1 cal plants flower significantly later than wild-type plants (Table 2). This suggests that the threshold of LFY required to induce flowering is higher in the ful ap1 cal background, or alternatively, that another factor may be required to accumulate when AP1/CAL/FUL are not present, to act with LFY in promoting flowering.

Similar phenotypes to those found in ful ap1 cal plants have been reported for several mutant combinations. For example, the ft ap1 and fwa ap1 double mutant inflorescences resemble the ful ap1 cal ‘leafy cauliflowers’, although they are able to flower after several months of growth (Ruiz-García et al., 1997). Similarly, ld ap1 cal triple mutants form proliferating leafy shoots at the apex and are unable to flower (Aukerman et al., 1999). The ap1 cal double mutants also show an enhancement of vegetative characteristics when grown in short days or at 16°C (Bowman et al., 1993). LFY fails to be upregulated both in ld ap1 cal and ap1 cal grown in non-inductive conditions (Aukerman et al., 1999; Bowman et al., 1993). The similarities among these mutant phenotypes and the ful ap1 cal inflorescences may reflect a possible role of the FT, FWA and LD genes in the competence of the meristems to respond to reduced levels of LFY, and/or indicate their possible function as LFY activators. It will be interesting to test whether FUL fails to be upregulated at low temperatures or in the ft, fwa or ld backgrounds, thereby preventing LFY activation.

AP1/CAL/FUL may control the transition between developmental phases by modulating the ratio of LFY/TFL activities

It has been suggested that the overlapping expression domains of LFY and TFL1 in emerging lateral primordia cause the meristem proliferation of ap1 cal inflorescences (Ratcliffe et al., 1999). Support for this idea comes from genetic studies, which show that this proliferation does not occur when lfy or tfl1 mutations are introduced into the ap1 cal double mutant. Thus, the role for CAL in an ap1 mutant background is to activate LFY and repress TFL1 expression in lateral meristems, thus preventing their overlapping activities (Bowman et al., 1993; Ratcliffe et al., 1999).

We have found that the expression domains of LFY and TFL1 overlap in the ful ap1 cal triple mutants, as was observed for ap1 cal meristems (Ratcliffe et al., 1999; this work). Accordingly, the ful ap1 cal plants formed meristems in a reiterate pattern, and this proliferation was not observed when LFY or TFL were also mutated. In the ap1 cal background, where FUL is expressed at the apex and ectopically in all lateral meristems, FUL was required for the increase of LFY expression, whereas it seemed to have little effect on TFL1 regulation. Thus, in the ful ap1 cal inflorescences, the ratio of LFY:TFL1 expression was always lower than in ap1 cal double mutants.

Our data suggest that the levels of LFY as well as the relative levels of LFY and TFL1 control plant architecture and meristem behavior. When ap1 cal SAMs enter the inflorescence developmental phase, they give rise laterally to new meristems that also behave as inflorescence apical meristems. In ful ap1 cal plants, the SAMs seem to proliferate as V2 meristems, producing cauline leaves with axillary meristems that in turn repeat this pattern forming the ‘leafy cauliflowers’ (Figs 2, 3). Thus, lower LFY to TFL1 relative levels, together with their overlapping expression, would result in the reiteration of a more vegetative phase in ful ap1 cal plants. In contrast, in ap1 cal double mutants, a slightly higher LFY to TFL1 ratio, together with higher LFY levels, would allow the transition to the inflorescence phase, that in turn would cause the reiterative behavior of the meristems. The subsequent upregulation of LFY in ap1 cal, mediated by FUL, would raise the LFY:TFL1 ratio to the required levels for floral specification. The late formation of floral structures in ful ap1 cal inflorescences in a TFL1 heterozygous background seems to support this hypothesis, since in this situation, a reduced TFL1 activity could eventually lead to the transition to floral commitment.

As noted above, the proliferating ful ap1 cal meristems appear to be arrested in the V2 phase, in contrast to the reproductive character of the ap1 cal cauliflowers. Besides the morphological evidence, additional data support this conclusion. For example, constitutive AG expression failed to promote floral identity in the ful ap1 cal meristems whereas it was able to promote flowering in ap1 cal double mutants (Mizukami and Ma, 1997). The lack of floral specification in 35S::AG ful ap1 cal plants suggests that AG may act through AP1/CAL/FUL to promote floral fate, or, alternatively, that the ful ap1 cal background prevents the meristems from becoming competent to respond to AG, perhaps by keeping them in a vegetative state.

FUL has a floral promoting activity independent of LFY

We have demonstrated that FUL plays a redundant role with AP1 and CAL in LFY upregulation, thus promoting floral meristem specification. Our data also suggest that FUL is involved in phase transition during development in a pathway that is independent of LFY. This is clearly illustrated by the observation that the delay in flowering caused by mutations in LFY is further enhanced in ful lfy double mutants (Table 1). In addition, small but significant increases in V1 and V2 phases are found in ful single mutants, even though levels of LFY expression are not noticeably affected. Furthermore, flowering time is dramatically reduced in plants that constitutively express FUL under the control of the 35S promoter, and this floral-promoting activity is independent of LFY, since is not affected in the lfy mutant background (C.F. and M.F.Y., unpublished results). It is interesting to note that the LFY-independent role of FUL in promoting the phase transition requires AP1, since FUL is unable to promote flowering in a lfy ap1 background. However the small effect on flowering time caused by ful mutations, and the rapid and strong FUL upregulation in the SAM after the induction of the reproductive phase, suggest that its flower promoting activity might be largely obscured by other highly redundant activities. Good candidates for genes that act redundantly with FUL in the SAM are the AGL20 and AGL24 MADS-box genes that share a similar pattern of upregulation (S. Gold and M. Yanofsky; C. Gustafson-Brown, M. Yanofsky and W. Crosby; unpublished results). Regardless of whether additional MADS-box genes or as yet unidentified genes are involved, it is clear that FUL acts in a highly redundant pathway to control the transition to flowering.

We thank Soraya Pelaz and Miguel A. Blázquez for helpful discussions and, together with Allen Sessions, Medard Ng and Detlef Weigel, for critical reading of the manuscript. Detlef Weigel, Hong Ma, Judy Roe, Allen Sessions, Ellen Wisman, David Smyth, John Bowman and Sarah Liljegren for providing seeds, Pilar Cubas for information on double labeling in situ experiments, and Tony Gaba, Amy Chen, Cheryl Wiley and Tad Kawashiwa for excellent technical assistance. This work was supported by grants from the National Science Foundation and the National Institutes of Health to M. F. Y. C. F. was a recipient of a postdoctoral fellowship from the Spanish Ministry of Education and Science.

Alvarez
,
J.
,
Guli
,
C. L.
,
Yu
,
X.-H.
and
Smyth
,
D. R.
(
1992
).
terminal flower, a gene affecting inflorescence development in Arabidopsis thaliana
.
Plant J
.
2
,
103
116
.
Aukerman
,
M. J.
,
Lee
,
I.
,
Weigel
,
D.
and
Amasino
,
R.
(
1999
).
The Arabidopsis flowering-time gene LUMINIDEPENDENS is expressed primarily in regions of cell proliferation and encodes a nuclear protein that regulates LEAFY expression
.
Plant J
.
18
,
195
203
.
Blázquez
,
M. A.
,
Soowal
,
L.
,
Lee
,
I.
and
Weigel
,
D.
(
1997
).
LEAFY expression and flower initiation in Arabidopsis
.
Development
124
,
3835
3844
.
Bowman
,
J. L.
,
Alvarez
,
J.
,
Weigel
,
D.
,
Meyerowitz
,
E. M.
and
Smyth
,
D. R.
(
1993
).
Control of flower development in Arabidopsis thaliana by APETALA1 and interacting genes
.
Development
119
,
721
743
.
Bradley
,
D. J.
,
Ratcliffe
,
O. J.
,
Vincent
,
C.
,
Carpenter
,
R.
and
Coen
,
E. S.
(
1997
).
Inflorescence commitment and architecture in Arabidopsis
.
Science
275
,
80
83
.
Drews
,
G. N.
,
Bowman
,
J. L.
and
Meyerowitz
,
E. M.
(
1991
).
Negative regulation of the Arabidopsis homeotic gene AGAMOUS by the APETALA2 product
.
Cell
65
,
991
1002
.
Gu
,
Q.
,
Ferrándiz
,
C.
,
Yanofsky
,
M. F.
and
Martienssen
,
R.
(
1998
).
The FRUITFULL MADS-box gene mediates cell differentiation during Arabidopsis fruit development
.
Development
125
,
1509
1517
.
Hempel
,
F. D.
,
Weigel
,
D.
,
Mandel
,
M. A.
,
Ditta
,
G.
,
Zambryski
,
P.
,
Feldman
,
L. J.
and
Yanofsky
,
M. F.
(
1997
).
Floral determination and expression of floral regulatory genes in Arabidopsis
.
Development
124
,
3845
3853
.
Huala
,
E.
and
Sussex
,
I. M.
(
1992
).
LEAFY interacts with floral homeotic genes to regulate Arabidopsis floral development
.
Plant Cell
4
,
901
913
.
Irish
,
V. F.
and
Sussex
,
I. M.
(
1990
).
Function of the apetala-1 gene during Arabidopsis floral development
.
Plant Cell
2
,
741
751
.
Jack
,
T.
,
Brockman
,
L. L.
and
Meyerowitz
,
E. M.
(
1992
).
The homeotic gene APETALA3 of Arabidopsis thaliana encodes a MADS-box and is expressed in petals and stamens
.
Cell
68
,
683
697
.
Kempin
,
S. A.
,
Savidge
,
B.
and
Yanofsky
,
M. F.
(
1995
).
Molecular basis of the cauliflower phenotype in Arabidopsis
.
Science
267
,
522
525
.
Lee
,
I.
,
Wolfe
,
D. S.
,
Nilsson
,
O.
and
Weigel
,
D.
(
1997
).
A LEAFY co-regulator encoded by UNUSUAL FLORAL ORGANS
.
Curr. Biol
.
7
,
95
104
.
Liljegren
,
S.
,
Gustafson-Brown
,
C.
,
Pinyopich
,
A.
,
Ditta
,
G.
and
Yanofsky
,
M.
(
1999
).
Interactions among the meristem identity genes APETALA1, LEAFY and TERMINAL FLOWER specify meristem fate
.
Plant Cell
11
,
1007
1018
.
Mandel
,
M. A.
,
Gustafson-Brown
,
C.
,
Savidge
,
B.
and
Yanofsky
,
M. F.
(
1992
).
Molecular characterization of the Arabidopsis floral homeotic gene APETALA1
.
Nature
360
,
273
277
.
Mandel
,
M. A.
and
Yanofsky
,
M. F.
(
1995a
).
The Arabidopsis AGL8 MADS box gene is expressed in inflorescence meristems and is negatively regulated by APETALA1
.
Plant Cell
7
,
1763
1771
.
Mandel
,
M. A.
and
Yanofsky
,
M. F.
(
1995b
).
A gene triggering flower development in Arabidopsis
.
Nature
377
,
522
524
.
Mizukami
,
Y.
and
Ma
,
H.
(
1992
).
Ectopic expression of the floral homeotic gene AGAMOUS in transgenic Arabidopsis plants alters floral organ identity
.
Cell
71
,
119
131
.
Mizukami
,
Y.
and
Ma
,
H.
(
1997
).
Determination of Arabidopsis floral meristem identity by AGAMOUS
.
Plant Cell
9
,
393
408
.
Neff
,
M.
,
Neff
,
J.
,
Chory
,
J.
and
Pepper
,
A.
(
1998
).
dCAPS, a simple technique for the genetic analysis of single nucleotide polymorphisms:experimental aplications in Arabidopsis thaliana genetics
.
Plant J
.
14
,
387
392
.
Piñeiro
,
M.
and
Coupland
,
G.
(
1998
).
The control of flowering time and floral identity in Arabidopsis
.
Plant Physiol
.
117
,
1
8
.
Poethig
,
R. S.
(
1990
).
Phase changes and the regulation of shoot morphogenesis in plants
.
Science
250
,
923
930
.
Ratcliffe
,
O.
,
Bradley
,
D.
and
Coen
,
E.
(
1999
).
Separation of shoot and floral identity in Arabidopsis
.
Development
126
,
1109
1120
.
Ruiz-García
,
L.
,
Madueño
,
F.
,
Wilkinson
,
M.
,
Haughn
,
G.
,
Salinas
,
J.
and
Martínez-Zapater
,
J. M.
(
1997
).
Different roles of flowering time genes in the activation of floral initiation genes in Arabidopsis
.
Plant Cell
9
,
1921
1934
.
Savidge
,
B.
(
1996
).
Floral meristem specification and floral organ development in Arabidopsis
.
University of California at San Diego
,
La Jolla, CA
.
Schultz
,
E. A.
and
Haughn
,
G. W.
(
1991
).
LEAFY, a homeotic gene that regulates inflorescence development in Arabidopsis
.
Plant Cell
3
,
771
781
.
Shannon
,
S.
and
Meeks-Wagner
,
D. R.
(
1991
).
A mutation in the Arabidopsis TFL1 gene affects inflorescence meristem development
.
Plant Cell
3
,
877
892
.
Shannon
,
S.
and
Meeks-Wagner
,
D. R.
(
1993
).
Genetic interactions that regulate inflorescence development in Arabidopsis
.
Plant Cell
5
,
639
655
.
Smyth
,
D. R.
,
Bowman
,
J. L.
and
Meyerowitz
,
E. M.
(
1990
).
Early flower development in Arabidopsis
.
Plant Cell
2
,
755
767
.
Weigel
,
D.
,
Alvarez
,
J.
,
Smyth
,
D. R.
,
Yanofsky
,
M. F.
and
Meyerowitz
,
E. M.
(
1992
).
LEAFY controls floral meristem identity in Arabidopsis
.
Cell
69
,
843
859
.
Weigel
,
D.
and
Nilsson
,
O.
(
1995
).
A developmental switch sufficient for flower initiation in diverse plants
.
Nature
377
,
495
500
.
Yanofsky
,
M.
(
1995
).
Floral meristems to floral organs: genes controlling early events in Arabidopsis flower development
.
Annu. Rev. Plant Physiol. Plant Mol. Biol
.,
46
,
167
188
.