FBP1 Research Articles

Transposon Display identifies individual transposable elements in high copy number lines

The dTph1 transposable element family of Petunia hybrida line W138 consists of between 100 and 200 members. A strategy that allows simultaneous detection of individual elements is described. Sequences flanking dTph1 elements are amplified by means of a ligation-mediated PCR. The resulting fragments are locus-specific and can be analysed by polyacrylamide gel electrophoresis. One of the applications of Transposon Display is the isolation of dTph1-tagged genes. Fragments that co-segregate with a mutant phenotype can be extracted from the gel and reamplified, providing access to tagged genes, as demonstrated in a reconstruction experiment. Data on the molecular identification of a phenotypic mutant, isolated in a random tagging experiment is also presented. Upon sequencing, the obtained candidate fragment was found to be identical to part of the previously identified Fbp1 gene.

Analysis of the structure of a natural alternating d(TA)n sequence in yeast chromatin.

We address here the question of the in vivo structure of a natural alternating d(TA)n sequence found at the 3' region of the Saccharomyces cerevisiae FBP1 gene. This sequence consists of 13 TA pairs interrupted by a TT dinucleotide in the middle of the tract. Previous experiments with cruciform-specific nucleases S1 and Endonuclease VII demonstrated the presence in vitro of a cruciform in this region. We also showed this region to be part of a nuclease hypersensitive site flanked by nucleosomes in yeast chromatin. Here we demonstrate, by means of S1 in vivo footprinting, that in yeast plasmids also adopts in vivo a non B-DNA structure where is not a cruciform. A theoretical analysis of this region that it contains a site susceptible to superhelical stress duplex destabilization. The locations and conditions under which alternative structures form in the wild-type sequence and in deletion mutants agree with these theoretical predictions, suggesting that some kind of denaturation is the alternative structure adopted by the sequence in vivo. This suggests that negative superhelical stress sufficient for local denaturation exists in nucleosomal DNA. We also demonstrate by micrococcal nuclease digestions that the deletion of the alternating d(TA)n sequence modifies the chromatin hypersensitive site but does not affect nucleosome positioning.

The wis1 signal transduction pathway is required for expression of cAMP-repressed genes in fission yeast.

The wis1 protein kinase of Schizosaccharomyces pombe is a member of the MAP kinase kinase family. Loss of wis1 function has previously been reported to lead to a delay in the G2-mitosis transition, loss of viability in stationary phase, and hypersensitivity to osmotic shock. It acts at least in part by activating the MAP kinase homologue sty1; loss-of-function sty1 mutants share many phenotypes with wis1 deletion mutants. We show here that, in addition, loss of wis1 function leads to defective conjugation, and to suppression of the hyperconjugation phenotype of the pat1-114 mutation. Consistent with this, the induction of the mei2 gene, which is normally induced by nitrogen starvation, is defective in wis1 mutants. In wild-type cells, nitrogen starvation leads to mei2 induction through a fall in intracellular cyclic AMP (cAMP) level and activity of the cAMP-dependent protein kinase. We show here that wis1 function is required for mei2 induction following nitrogen starvation. Expression of the fbp1 gene is negatively regulated by cAMP in response to glucose limitation: induction of fbp1 also requires wis1 and sty1 function. Loss of wis1 is epistatic over increased fbp1 expression brought about by loss of adenylate cyclase (git2/cyr1) or cAMP-dependent protein kinase (pka1) function. These observations can be explained by a model in which the pka1 pathway negatively regulates the wis1 pathway, or the two pathways might act independently on downstream targets. The latter explanation is supported, at least as regards regulation of cell division, by the observation that loss of function of the regulatory subunit of the cAMP-dependent protein kinase (cgs1) brings about a modest increase in cell length at division in both wis1+ and wis1 delta genetic backgrounds.

The Schizosaccharomyces pombe pyp1 protein tyrosine phosphatase negatively regulates nutrient monitoring pathways.

The Schizosaccharomyces pombe pyp1+ gene, encoding a protein tyrosine phosphatase (pyp1), was isolated as a high copy number suppressor of a mutation that results in reduced cAMP-dependent protein kinase (PKA) activity. Overexpression of pyp1+ inhibits both transcription of the fbp1 gene, which is negatively regulated by a glucose-induced activation of PKA, and sexual development, which is negatively regulated by PKA through a nitrogen- and glucose-monitoring mechanism. Overexpression of a catalytically inactive form of pyp1 has little effect on either process. Previous studies suggest that overexpression of pyp1+ results in a mitotic delay by positively regulating wee1 activity. We show that pyp1 repression of fbp1 transcription is independent of wee1. The direct role of the pyp1 protein is to dephosphorylate and inactivate the sty1/spc1 mitogen-activated protein kinase (MAPK) that is activated by the wis1 MAPK kinase. As overexpression of pyp1+ has no further effect upon the mitotic delay observed in a wis1 deletion strain, the role of pyp1 appears to be restricted to negative regulation of the sty1/spc1 MAPK. This study indicates that pyp1 negatively regulates fbp1 transcription, sexual development and mitosis by inactivation of the sty1/spc1 MAPK, but that bifurcations downstream of the MAPK separate these processes as seen by the differential role for the wee1 gene.

Direct repeats bind the EcR/USP receptor and mediate ecdysteroid responses in Drosophila melanogaster.

The steroid hormone 20-hydroxyecdysone plays a key role in the induction and modulation of morphogenetic events throughout Drosophila development. Previous studies have shown that a heterodimeric nuclear receptor composed of the EcR and USP proteins mediates the action of the hormone at the transcriptional through binding to palindromic ecdysteroid mediates the action of the hormone at the transcriptional level through binding to palindromic ecdysteroid response elements (EcREs) such as those present in the promoter of the hsp27 gene or the fat body-specific enhancer of the Fbp1 gene. We show that in addition to palindromic EcREs, the EcR/USP heterodimer can bind in vitro with various affinities to direct repetitions of the motif AGGTCA separated by 1 to 5 nucleotides (DR1 to DR5), which are known to be target sites for vertebrate nuclear receptors. At variance with the receptors, EcR/USP was also found to bind to a DR0 direct repeat with no intervening nucleotide. In cell transformation assays, direct repeats DR0 to DR5 alone can render the minimum viral tk or Drosophila Fbp1 promoter responsive to 20-hydroxyecdysone, as does the palindromic hsp27 EcRE. In a transgenic assay, however, neither the palindromic hsp27 element nor direct repeat DR3 alone can make the Fbp1 minimal promoter responsive to premetamorphic ecdysteroid peaks. In contrast, DR0 and DR3 elements, when substituted for the natural palindromic EcRE in the context of the Fbp1 enhancer, can drive a strong fat body-specific ecdysteroid response in transgenic animals. These results demonstrate that directly repeated EcR/USP binding sites are as effective as palindromic EcREs in vivo. They also provide evidence that additional flanking regulatory sequences are crucially required to potentiate the hormonal response mediated by both types of elements and specify its spatial and temporal pattern.

Multiple signalling pathways trigger the exquisite sensitivity of yeast gluconeogenic mRNAs to glucose.

The transcription of the yeast FBP1 and PCK1 genes, which encode the gluconeogenic enzymes fructose-1,6-bisphosphatase and phosphoenolpyruvate carboxykinase, is repressed by glucose. Here, we show that this repression is both very strong and exceptionally sensitive to glucose, being triggered by glucose at concentrations less than 0.005% (0.27 mM). This repression remains operative in yeast mutants carrying any one of the three hexose kinases, but is lost in a triple hxk1, hxk2, glk1 mutant. In addition, 2-deoxyglucose can trigger the repression, but 6-deoxyglucose cannot, suggesting that internalization and phosphorylation of the glucose is essential for repression to occur. While gluconeogenic gene transcription is subject to the Mig 1p-dependent pathway of glucose repression, the exquisite response to glucose is maintained in hxk2 and mig1 mutants, suggesting that this pathway is not essential for the response. The response can also be triggered by the addition of exogenous cAMP, suggesting that the Ras/cAMP pathway can mediate repression of the FPB1 and PCK1 mRNAs. However, the response is not dependent upon this pathway because it remains intact in Ras, adenyl cyclase and protein kinase A mutants. The data show that yeast cells can detect very low glucose concentrations in the environment, and suggest that several distinct signalling pathways operate to repress FPB1 and PCK1 transcription in the presence of glucose.

Analysis of Positive Elements Sensitive to Glucose in the Promoter of the FBP1 Gene from Yeast

We have identified in the promoter of the FBP1 gene from Saccharomyces cerevisiae, which codes for fructose-1,6-bisphosphatase, two elements which can form specific DNA.protein complexes and which confer glucose-repressed expression to an heterologous reporter gene. Complex formation and activation of transcription by either element require a functional CAT1 gene and are not blocked by a hap2-1 mutation, although this mutation interferes with maximal expression of the FBP1 gene. A sequence from one of the elements acts as a weak upstream activating sequence, but its activity can be stimulated up to 10-fold by neighboring sequences. A further element of the promoter has been characterized, which forms a specific DNA.protein complex only when a nuclear extract from derepressed cells is used. This element does not activate transcription in a heterologous promoter. The DNA sequences of the three elements involved in protein binding, defined by DNase I footprinting, have no homology with consensus sequences for known activating factors.

CAT8, a new zinc cluster-encoding gene necessary for derepression of gluconeogenic enzymes in the yeast Saccharomyces cerevisiae.

The expression of gluconeogenic fructose-1,6-bisphosphatase (encoded by the FBP1 gene) depends on the carbon source. Analysis of the FBP1 promoter revealed two upstream activating elements, UAS1FBP1 and UAS2FBP1, which confer carbon source-dependent regulation on a heterologous reporter gene. On glucose media neither element was activated, whereas after transfer to ethanol a 100-fold derepression was observed. This gene activation depended on the previously identified derepression genes CAT1 (SNF1) (encoding a protein kinase) and CAT3 (SNF4) (probably encoding a subunit of Cat1p [Snf1p]). Screening for mutations specifically involved in UAS1FBP1 derepression revealed the new recessive derepression mutation cat8. The cat8 mutants also failed to derepress UAS2FBP1, and these mutants were unable to grow on nonfermentable carbon sources. The CAT8 gene encodes a zinc cluster protein related to Saccharomyces cerevisiae Gal4p. Deletion of CAT8 caused a defect in glucose derepression which affected all key gluconeogenic enzymes. Derepression of glucose-repressible invertase and maltase was still normally regulated. A CAT8-lacZ promoter fusion revealed that the CAT8 gene itself is repressed by Cat4p (Mig1p). These results suggest that gluconeogenic genes are derepressed upon binding of Cat8p, whose synthesis depends on the release of Cat4p (Mig1p) from the CAT8 promoter. However, gluconeogenic promoters are still glucose repressed in cat4 mutants, which indicates that in addition to its transcription, the Cat8p protein needs further activation. The observation that multicopy expression of CAT8 reverses the inability of cat1 and cat3 mutants to grow on ethanol indicates that Cat8p might be the substrate of the Cat1p/Cat3p protein kinase.

The ecdysone response enhancer of the Fbp1 gene of Drosophila melanogaster is a direct target for the EcR/USP nuclear receptor.

The transcription of the Drosophila melanogaster Fbp1 gene is induced by the steroid hormone 20-hydroxyecdysone and restricted to the late-third-instar fat body tissue. In a previous study we showed that the -68 to -138 region relative to the transcription start site acts as an ecdysone-dependent third-instar fat body-specific enhancer in a transgenic assay. Here we report that seven nucleoprotein complexes are formed in vitro on this enhancer when a nuclear extract from late-third-instar fat body is used in a gel shift assay. Accurate mapping of the binding sites of the complexes revealed a remarkably symmetrical organization. Using specific antibodies, one of the complexes was identified as a heterodimer consisting of the ecdysone receptor (EcR) and Ultraspiracle (USP) proteins. The binding site of the heterodimer as defined by mutagenesis and methylation interference experiments bears strong sequence similarity to the canonical hsp27 ecdysone response element, including an imperfect palindromic structure. The two elements diverge at three positions in both half-sites, indicating that the structure of an active EcR/USP binding site allows considerable sequence variations. In vivo footprinting experiments using ligation-mediated PCR and wild-type or ecdysteroid-deficient larvae show that occupancy of the Fbp1 EcR/USP binding site and adjacent region is dependent on a high concentration of ecdysteroids. These results provide strong evidence for a direct role of the EcR/USP heterodimer in driving gene expression in response to changes of the ecdysteroid titer during Drosophila larval development.

The ecdysone response enhancer of the Fbp1 gene of Drosophila melanogaster is a direct target for the EcR/USP nuclear receptor.

The transcription of the Drosophila melanogaster Fbp1 gene is induced by the steroid hormone 20-hydroxyecdysone and restricted to the late-third-instar fat body tissue. In a previous study we showed that the -68 to -138 region relative to the transcription start site acts as an ecdysone-dependent third-instar fat body-specific enhancer in a transgenic assay. Here we report that seven nucleoprotein complexes are formed in vitro on this enhancer when a nuclear extract from late-third-instar fat body is used in a gel shift assay. Accurate mapping of the binding sites of the complexes revealed a remarkably symmetrical organization. Using specific antibodies, one of the complexes was identified as a heterodimer consisting of the ecdysone receptor (EcR) and Ultraspiracle (USP) proteins. The binding site of the heterodimer as defined by mutagenesis and methylation interference experiments bears strong sequence similarity to the canonical hsp27 ecdysone response element, including an imperfect palindromic structure. The two elements diverge at three positions in both half-sites, indicating that the structure of an active EcR/USP binding site allows considerable sequence variations. In vivo footprinting experiments using ligation-mediated PCR and wild-type or ecdysteroid-deficient larvae show that occupancy of the Fbp1 EcR/USP binding site and adjacent region is dependent on a high concentration of ecdysteroids. These results provide strong evidence for a direct role of the EcR/USP heterodimer in driving gene expression in response to changes of the ecdysteroid titer during Drosophila larval development.

In vivo functional characterization of an ecdysone response enhancer in the proximal upstream region of the Fbp1 gene of D. melanogaster

Transcription of the D. melanogaster Fat-body-protein-1 (Fbp1) gene is induced by the steroid hormone 20-hydroxyecdysone and is restricted to the fat body tissue at the end of the third larval instar. The location and functional properties of the Fbp1 cis-acting regulatory sequences contained in the region from −1386 to +80 relative to the transcription start were examined by transformation using hybrid constructs with the Adh or lacZ genes as reporters. Regulatory element(s) required for the full level of transcription of the Fbp1 gene were located between positions −1386 and −138. Sequences between −138 and −68 were able to drive transcription from a heterologous minimal promoter in the fat body of late third instar larvae. Remarkably, these sequences also conferred 20-hydroxyecdysone inducibility and behaved as an enhancer-like element. These results provide the first functional characterization, at the level of the whole organism, using a direct in vivo ecdysone induction assay, of a discrete ecdysone response element.

Deletion scanning of the regulatory sequences of the Fbp1 gene of Drosophila melanogaster using P transposase-induced deficiencies.

A procedure permitting deletion scanning of potential cis-regulatory sequences within a transgene whose genomic position remains fixed was applied to the study of the upstream sequences of the ecdysteroid-inducible Fat-body-protein-1 (Fbp1) gene. Deficiencies were induced in a Fbp1:Adh fusion transgene by means of a secondary P transposase mutagenesis. Phenotypic and molecular screens were used to select mutant transposons that retained their original genomic location and carried a deletion affecting the Fbp1 sequences but not the Adh reporter gene. Molecular mapping of the deletion breakpoints was achieved by sequence analysis and expression of the reporter gene was quantified by measurement of ADH activity. This procedure was efficient in detecting cis-acting elements, even those with moderate effects on levels of gene expression. For example, we have succeeded in identifying a negative regulatory element. Deletion of this element leads to a 50% increase in the reporter ADH activity. This element binds the transcription factor AEF-1. In addition, we have detected a strong, positively acting element contained within a 32-bp region located immediately upstream of an ecdysone-response element.

Chromatin structure of the yeast FBP1 gene: Transcription‐dependent changes in the regulatory and coding regions

We have studied the chromatin structure of the Saccharomyces cerevisiae FBP1 gene, which codes for fructose-1,6-bisphosphatase. A strong, constitutive, DNase I, micrococcal nuclease and S1 nuclease hypersensitive site is present close to the 3' end of the coding region. In the repressed state, positioned nucleosomes exist around this site, and subtle changes occur in this nucleosomal organization upon derepression. A DNase I hypersensitive region is located within the promoter between positions -540 and -400 and its extends towards the gene in the derepressed state, leading to an alteration of nucleosomal positioning. Psoralen crosslinking of chromatin, which is used for the first time to study the mobility of restriction fragments from an RNA polymerase II gene, revealed that part of the promoter is nucleosome-free, in accordance with the results of DNase I digestion. A model is presented that, based on the chromatin structure, puts forward the hypothesis that the promoter UAS is located between -540 and -340. Finally, psoralen crosslinking, as well as digestions with micrococcal nuclease or restriction endonucleases suggests that most if not all of the copies of the active FBP1 gene are covered by nucleosomes.

Six git genes encode a glucose-induced adenylate cyclase activation pathway in the fission yeast Schizosaccharomyces pombe

An important eukaryotic signal transduction pathway involves the regulation of the effector enzyme adenylate cyclase, which produces the second messenger, cAMP. Previous genetic analyses demonstrated that glucose repression of transcription of the Schizosaccharomyces pombe fbp1 gene requires the function of adenylate cyclase, encoded by the git2 gene. As mutations in git2 and in six additional git genes are suppressed by exogenous cAMP, these 'upstream' git genes were proposed to act to produce a glucose-induced cAMP signal. We report here that assays of cAMP levels in wild-type and various mutant S. pombe cells, before and after exposure to glucose, show that this is the case. The data suggest that the cAMP signal results from the activation of adenylate cyclase. Therefore these 'upstream' git genes appear to encode a glucose-induced adenylate cyclase activation pathway. Assays of cAMP on a strain carrying a mutation in the git6 gene, which acts downstream of adenylate cyclase, indicate that git6 may function to feedback regulate adenylate cyclase activity. Thus git6 may encode a cAMP-dependent protein kinase.

A Natural A/T-Rich Sequence from the Yeast FBP1 Gene Exists as a Cruciform in Escherichia coli Cells

Palindromic or semipalindromic sequences can adopt cruciform structures in DNA in vitro. It has been demonstrated in some cases that A/T-rich cruciforms exist also in vivo in Escherichia coli. The biological function of those structures is not understood although putative cruciforms have been found in interesting locations on replication origins, operators, or transcriptional termination regions. Here we show by means of the use of structure-dependent nucleases that the 3′ end of the yeast FBP1 gene contains a stable cruciform both in vitro and in E. coli cells and that in both cases, its extrusion depends on the DNA supercoiling state.

Molecular analysis of the neutral trehalase gene from Saccharomyces cerevisiae.

Neutral trehalase (EC 3.2.1.28) is a trehalose hydrolyzing enzyme of the yeast Saccharomyces cerevisiae (App, H., and Holzer, H. (1989) J. Biol. Chem. 264, 17583-17588). The gene of neutral trehalase was cloned by complementation of a neutral trehalase-deficient yeast mutant which was obtained by ethylmethanesulfonate mutagenesis. Three mutants without detectable neutral trehalase activity were obtained and characterized by tetrad analysis and found to belong to the same complementation group. The mutants were transformed with a S. cerevisiae genomic library in YEp24. Two overlapping plasmids were isolated, containing the neutral trehalase gene NTH1 with an open reading frame of 2079 base pairs (bp), encoding a protein of 693 amino acids, corresponding to a molecular mass of 79,569 Da. Several putative TATA boxes were found in the 5'-nontranslated region of the NTH1 gene. In positions -652 to -641 a possible binding sequence for the MIG1 protein, a multicopy inhibitor of the GAL1 promotor, which also binds to the promotor sequences of the SUC2 and the FBP1 gene, was found. The start codon of the neutral trehalase is located about 2500 bp upstream of the centromere 4 consensus sequence elements I, II, and III (Mann, C., and Davis, R. W. (1986) Mol. Cell. Biol. 6, 241-245). Vicinity to a centromere is known to have a depressing influence on the number of plasmid copies per cell. This probably explains why transformation with pNTH does not lead to overexpression of neutral trehalase. The four consensus sequences AATAAA contained in the centromeric elements and reconfirmed by our sequencing data might be polyadenylation signals for NTH1-mRNA transcription termination. Northern blot analysis yielded a single mRNA species of approximately 2.3 kilobase(s). The neutral trehalase protein has a putative cAMP-dependent phosphorylation consensus sequence RRGS from amino acid positions 22-25. Therefore, the previously described activation of neutral trehalase by cAMP-dependent phosphorylation is probably due to phosphorylation of serine 25. Three potential N-glycosylation sites (Asn-X-Ser/Thr) occur in the open reading frame of the neutral trehalase gene. However, no evidence for glycosylation could be detected by Western blotting.(ABSTRACT TRUNCATED AT 400 WORDS)

Structural features critical to the activity of an ecdysone receptor binding site

Two ecdysone-response elements from the hsp27 (hsp27 EcRE) and the Fbp1 (D EcRE) genes of Drosophila melanogaster were used as probes in a gel shift assay to investigate the interactions of the ecdysone receptor (EcR) with its cognate DNA response element. The source of EcR was a nuclear extract from the late third-larval instar fat body. The hsp27 and D EcREs share a sequence similarity at 12 positions over a 15bp region including an imperfect palindromic structure consisting of two pentamer half-sites separated by a single intervening nucleotide. We have shown that a short oligonucleotide containing this 11bp imperfect palindrome of the hsp27 EcRE and three flanking bp on each side is an efficient EcR binding site. Mutational analysis confirms that the integrity of both these half-sites as well as their 1bp spacing are critical for binding of the ecdysone receptor. The D EcRE behaved as a much weaker EcR binding sapacity equivalent to that of the hsp27 EcRE. These results have led us to propose the sequence PuG(G/T)T(C/G)A(N)TG(C/A)(C/A)(C/t)Py as a revised version of a previously proposed EcRE consensus sequence.

Identification of UAS elements and binding proteins necessary for derepression of Saccharomyces cerevisiae fructose-1,6-bisphosphatase.

Fructose-1,6-bisphosphatase is a key enzyme in gluconeogenesis and the FBP1 gene is not transcribed during growth with glucose. Genetic analysis indicated a positive regulation of FBP1 expression after exhaustion of glucose. By linker-deletion analysis, two upstream activation sites (UAS1 and UAS2) were localized and the respective UAS-binding factors (DAP I and DAP II for derepression activating protein) were identified by gel retardation. UAS1 and UAS2 span about 30 bp each, and are separated by approximately 30 bp. Both UAS sites act synergistically. Although UAS1 showed some similarities to the DNA-binding consensus for the general yeast activator Rap1, competition experiments and DEAE-chromatography proved that DAP I and Rap1 correspond to different proteins. Gel retardation by DAP I depended on carbon sources and did not occur in cells growing logarithmically with glucose, whereas a strong retardation signal was obtained with ethanol-grown cells. The present results suggest that DAP I and DAP II are the final regulatory elements for glucose derepression.

Regulatory regions in the yeast FBP1 and PCK1 genes

By deletion analysis of the fusion genes FBP1-lacZ and PCK1-lacZ we have identified a number of strong regulatory regions in the genes FBP1 and PCK1 which encode fructose-1,6-bisphosphatase and phosphoenolpyruvate carboxykinase. Lack of expression of β-galactosidase in fusions lacking sequences from the coding regions suggests the existence of downstream activating elements. Both promoters have several UAS and URS regions as well as sites implicated in catabolite repression. We have found in both genes consensus sequences for the binding of the same regulatory proteins, such as yAP1, MIG1 or the complex HAP2/HAP3/HAP4. Neither deletion nor overexpression of the MIG1 gene affected the regulated expression of the FBP1 or PCK1 genes.

Differential expression of two MADS box genes in wild-type and mutant petunia flowers.

We isolated and characterized two flower-specific genes from petunia. The protein products of these genes, designated floral binding protein 1 (FBP1) and 2 (FBP2), are putative transcription factors with the MADS box DNA binding domain. RNA gel blot analysis showed that the fbp1 gene is exclusively expressed in petals and stamen of petunia flowers. In contrast, the FBP1 protein was only detectable in petals and not in stamens, suggesting post-transcriptional regulation of the fbp1 gene in these tissues. The fbp2 gene is expressed in petals, stamen, carpels, and at a very low level in sepals but not in vegetative tissues. We analyzed the spatial expression of these fbp genes in floral organs of two homeotic flower mutants. In the blind mutant, whose flower limbs are transformed into antheroid structures on top of normal tubes, identical expression levels of both genes were observed in the antheroid structures as in normal anthers. In the homeotic mutant green petals, the petals are replaced by sepaloid organs in which the expression of fbp1 is strongly reduced but not completely abolished. Our results suggest a regulation of the fbp1 gene expression by the green petals (gp) gene. Expression of the fbp2 gene was not affected in the green petals mutant. In contrast to the proposed models describing floral morphogenesis, our data indicated that homeotic genes can be functional in one whorl only.