WO2001002572A1

WO2001002572A1 - Genetic control of flowering using the fwa gene

Info

Publication number: WO2001002572A1
Application number: PCT/NL1999/000414
Authority: WO
Inventors: Wilhelmus Jakobus Johannes Soppe; Antonius Johan Maria Peeters; Maarten Koornneef
Original assignee: Wageningen University
Priority date: 1999-07-02
Filing date: 1999-07-02
Publication date: 2001-01-11
Also published as: AU4803799A; AU6026200A; WO2001002573A1; EP1196582A1

Abstract

This invention relates to the determination, cloning and expression of the flowering time gene FWA and the use of this gene to delay or accelerate flowering in a large variety of plant species. Specifically the gene that was determined is that of Arabidopsis thaliana. Naturally the invention extends to other plants as well. The proper timing of flowering is important in crop plants because it determines the moment that the crop can be harvested when crops are grown for their fruits and seeds. Early flowering is a disadvantage as this leads to reduce yields because insufficient vegetative mass is available. When crop plants flower too late, the harvest may take place too late in the season.

Description

Title: Genetic control of flowering using the FWA gene.

Summary of the invention. This invention relates to the determination, cloning and expression of the flowering time gene FWA and the use of this gene to delay or accelerate flowering in a large variety of plant species. Specifically the gene that was determined is that of Arabidopsis thaliana.

Background to the invention.

The proper timing of flowering is important in crop plants because it determines the moment that the crop can be harvested when crops are grown for their fruits and seeds. Early flowering is a disadvantage as this leads to reduced yields because insufficient vegetative mass is available. When crop plants flower too late, the harvest may take place too late in the season and may therefore result in losses. In plants that are grown for their vegetative part as is the case for many vegetables and tubers such as sugar beets, premature flowering (bolting) leads to losses in yield or may even make the plant useless for the market. One aspect of the problem of premature bolting is that flowering is promoted by specific climatic conditions, such as a cold spring season and that these climatic conditions are not always predictable.

The proper time of flowering, the process required for successful sexual reproduction of plants is controlled by both environmental and endogenous factors. Plant physiologists have studied this process by changing environmental factors and analyzing the subsequent morphological, physiological and biochemical consequences of these treatments. More recently genetics has been used to study the mechanism of flower initiation by analysis of genetic variation in species. Such species are for example the pea and Arabidopsis. A review of the work done in Arabidopsis can be found in Annu. Rev. Plant. Physiol. Plant. Mol. Biol. 1998, 49:345-70 by Koornneef, M. et al. and in The Plant Cell Vol 10, 1973-1989, December 1998 by Levy, Y.Y., and Dean, C.

The response of flowering-time mutants to environmental treatments such as vernalisation and photoperiod combined with genetic analyses of epistasis have established the existance of at least 4 different pathways that control flowering time in Arabidopsis. The floral repression pathway may be a built in mechanism preventing flowering until the plant reaches sufficient maturity, whilst the autonomous promotion pathway probably antagonises the former during development to an increasing degree. So these two pathways monitor the endogenous development stage of the plant. In addition there are two pathways linked to environmental signals one is the photoperiodic pathway which is sensitive to inductive photoperiods for inducing flower development and the other is the vernalisation promotion pathway which allows flowering to commence after extended periods of cold temperature. Table 1 of the Plant Cell review article mentioned above (which is incorporated by reference) shows at least 50 genes known to be involved in the flowering process. Quite clearly it is a complex issue which obviously requires a lot of research in order to arrive at means of actually altering the flowering-time characteristics in a directed manner, rather than remaining in the speculative unclear sphere. The subject invention now surprisingly offers such a method.

Table 1 : Morphological markers, used for the genetic mapping of the FWA locus.

The plant selected as example for such manipulation is the Arabidopsis thaliana This is a facultative long day plant, which means that its flowering is promoted by long day photoperiods Also many strains of this species respond strongly to vernalization, the cold treatment given at the young plant stage

It was known from a previous publication of one of the inventors that dominant mutations at the FWA locus confer reduced sensitivity to the promotive effect of long days (LD) and therefore render these mutants less sensitive to day-length In addition the fwa mutants have a reduced sensitivity to vernalisation (Koornneef et al Mol Gen Genet 229 57-66) The nature and location of the gene were however unknown as were the nature and location of the FWA locus associated mutations It was also unclear whether other mutations were also involved in the phenotypic differences

Description of the invention

The FWA gene, located on the long arm of chromosome 4 of Arabidopsis thaliana has now been cloned

The FWA locus had been identified in Arabidopsis thaliana by its mutant phenotype In comparison to the wild type, theyvrø mutant was delayed in the transition from the vegetative to the generative phase There were two mutant alleles available, fwa-1 was obtained by EMS-mutagenesis andβva-2 by γ-irradiation The FWA locus had already been mapped to chromosome 4 However, nothing was known about the molecular function of FWA To understand this function and the role of the FWA gene in the flowering process an attempt was made to clone this gene There are several strategies to clone a gene, depending on the available molecular information (Gibson and Somerville, 1993) When the function of a gene is known, it is possible to isolate the gene by its ability to complement mutations in bacteria or yeast (Minet et αl, 1992) A gene with a characterised pattern of expression can be cloned by differential screening (Park et αl, 1998) When only the mutant phenotype of a gene is known, other strategies have to be used. A T-DNA or transposon that is inserted in the gene causes the mutant phenotype of tagged mutants. Such a gene can be very effectively cloned by isolation of DNA fragments flanking the insertion (Aarts et al, 1995; Schaffer et al, 1998). In cases where the mutant phenotype is caused by a deletion, the gene can be cloned by subtractive hybridisation. However, cloning by subtractive hybridisation in Arabidopsis is tricky and has only been proven successful for two loci (Silverstone et al, 1998; Sun et al, 1992). Furthermore, if none of the above mentioned methods can be applied, a gene can be cloned by map-based cloning (Putterill et al, 1995; Macknight et al, 1997). In the instant case of FWA the gene product and function were unknown and no tagged alleles were available, so the first methods mentioned above were excluded as cloning strategy. Preliminary attempts to clone the FWA locus by subtractive hybridisation remained unsuccessful. Map-based cloning subsequently seemed the best way for cloning a gene like FWA of which only the mutant phenotype and genetic map position were known. Map-based cloning in Arabidopsis was facilitated by the fact that it has one of the smallest genomes among higher plants with very low levels of repetitive DNA. Furthermore, there are many genetic loci identified by mutations, it has a dense molecular marker map and there is almost a complete yeast artificial chromosome (YAC) coverage (Dean and Schmidt, 1995). The first step in map-based cloning is to locate genetically the locus of interest as accurately as possible with the help of linked markers, either morphological or molecular. The most closely linked molecular markers can be used to isolate clones that contain the region of the genome covering the locus. Thereafter, the gene can be identified by complementation of the mutant phenotype in transgenic plants containing the candidate gene. DNA sequencing of the wild type and mutant alleles can reveal the nature of the mutation and a comparison with sequences in the databases can indicate the putative molecular function of the gene. A set of overlapping cosmid clones isolated from a cosmid library was thus made by the inventors from the mutant allele fwa-1. Agrobacterium mediated transformation of two overlapping cosmids conferred lateness to transformed early wild type plants, similar to the lateness of the fwa mutants. In the overlap between those two cosmids a DNA sequence encoding a homeobox like transcription factor, which is responsible for the lateness conferred to the transformed wild type plants was detected. This DNA sequence is hereafter called the FWA gene. Using the techniques of reverse transcriptase PCR (RTPCR) and northern blot analysis it was shown that the FWA gene is not expressed in the wildtype but is expressed in the mutant.

The lack of expression in the wild type could be attributed to the methylation of a part of the promoter of the gene. The promoter contains two blocks of repeated sequences in which the C residues of the DNA are methylated. Since especially repetitive DNA is vulnerable to methylation, placing this gene under different promoters, vectors that lead to transgenic plants in which flowering is delayed in a genetically stable way. have now become available.

Now the sequence of FWA is known for Arabidopsis it has become possible to detect and isolate other FWA encoding sequences that are homologues of this gene. Using analysis methods of known nucleic acid sequences as present in databases or using the sequence information that is now made available by the subject disclosure the skilled person can readily arrive at genes in other plant species encoding FWA by virtue of homology and hybridisation experiments. This gene can thus also be isolated using standard technology from the other plants. The available nucleic acid sequence will also allow the isolation of homologous genes from other species, either by DNA homology or by their identification in public databases.

These sequences can be used to modify flowering time in those other species by overexpression or by co-suppression. Alternatively once the FWA genes of other plants have been determined the expression regulation of the genes can be amended to the direction required i.e. earlier or later flowering as desired. Alternatively the other FWA genes can be manipulated in other plants analogously to the methodology illustrated herein for Arabidopsis. The same can of course be carried out obviously for Arabidopsis itself as is illustrated herein. The Arabidopsis or any other plant foreign FWA gene can also be introduced into any desired plant. The alteration in other plants can also for example occur by use of the FWA gene of

Arabidopsis or of the plant in question to repress the expression of the plants own FWA genes via the anti-sense or co-suppression technology. This method will be effective as is apparent from our illustration that some transformed late fwa mutant plants, which express FWA, become early after transformation with a mutant copy of the fwa encoding sequence. Silencing of the FWA gene in the fwa mutant due to gene-silencing induced by the transgene is the most likely explanation for this observation. The nucleic acid sequence encoding FWA can also be placed under the control of externally inducible gene promoters, which may allow the exact timing of the crop plant. Thus it has become possible also to alter the flowering-time behavior of other plants in addition to Arabidopsis in a directed and specific manner. Considering the complexity of the subject process it is amazing that such a slight amendmant can have such extreme effect without apparently resulting in detrimental effects for the plant and without being compensated by any of the other numerous control mechanisms for the flowering development.

A detailed description of the cloning of FWA is available. In addition the sequence of the FWA gene and its natural promoter region are provided in the example. Further it is illustrated that indeed loss of expression of the FWA gene results in the phenotype of later flowering. This is achieved by producing revertants of the fwa mutants which exhibit corresponding alteration in phenotype with regard to flowering time. Also the transformation of the FWA gene under control of the 35S promoter illustrates the possibility of replacing the promoter of the FWA gene. Transformation of other plants e.g. Nicotiana tabacum and Nicotiana plumbaginifolia with the same construct as used for Arabidopsis illustrates the possibility to extrapolate the teaching from Arabidopsis to other plant species.

Example

This example describes the map-based cloning of the FWA gene. First a segregating population that was constructed for the fine mapping of FWA is described. Thereafter the YAC and cosmid contigs that were constructed and finally the plant transformation experiments that showed complementation are described

Genetic mapping The mapping population

For the genetic mapping of FWA both morphological and molecular markers were employed. Morphological markers are based on differences in phenotype while molecular markers detect polymorphisms at the DNA level. The latter implies that a mapping population should be derived from a cross between two plants that do not only differ for the locus of interest but that also differ in their DNA sequence. Such DNA polymorphisms can be found between ecotypes in Arabidopsis. For the mapping of FWA the ecotypes Landsberg erecta (Ler) and Columbia (Col) were used. A complication of the use of ecotypes is that they may differ in loci affecting the trait of interest, in this case flowering time. An analysis of recombinant inbred lines derived from the cross Ler x Col revealed genetic variation for flowering time at twelve different loci (Jansen etal, 1995). In this population the effects of individual loci were relatively small, in contrast with the finding of large gene effects in the progeny of the cross between the early flowering ecotypes Ler and Cape Verde Islands (Alonso- Blanco et al, 1998). However, the accumulation of either many early or many late alleles in specific progeny plants from a Ler x Col cross may affect the phenotype for flowering time of these plants in such a way that their classification for the FWA gene cannot be done unambiguously. To solve this problem a mapping population was constructed with a more uniform genetic background (Figure 1). First a cross was made between Ler and Col whereby Ler was homozygous for fwa-1 and for the recessive morphological markers cer2, ga5 and ap2 (Table 1). Late plants homozygous for cer2, ga5,fwa and ap2 were selected in the F2 generation. This selection ensured that in the FWA region of chromosome 4 these plants were homozygous Ler, and thus fwa mutants, whereas the rest of the genome contains both Ler and Col DNA. Some of these F2 individuals were crossed with Col plants, heterozygous for the emb35 mutation, which is linked to FWA. This cross with Col resulted in F2's with a more Col genetic background, leading to less variation in flowering time. Three different F2 populations from this cross were grown to check their variation in flowering time and whether they segregated for the emb35 mutation. The use of a lethal marker linked in repulsion to the semi-dominant late flowering mutant fwa implies that hardly any early plants are expected, unless unlinked flowering time modifiers segregate. This allowed a clear distinction between F2 populations where such modifiers did segregate and those where this was not the case. One F2 population with a clear monogenic segregation for flowering time and segregating emb35 was selected as mapping population and further analysed. The segregation of flowering time in this mapping population showed. The plants were grown under long day light conditions in a greenhouse and in these conditions Col flowered between 27 and 31 days, whereas the progeny of the parental fwa mutant plant that was selected for the cross flowered between 42 and 51 days. As fwa is a semi- dominant mutant, one quarter of the mapping population should flower as early as Col. However, this fraction does not exist in the population because, due to linkage, these plants are homozygous for emb35 and therefore embryo lethal. The early plants found in the population must be the result of a crossover between fwa and emb35. Therefore, the overall shape of the flowering time frequency distribution with two major peaks of different size can be explained because approximately 2/3 of these plants will be heterozygous for fwa (the heterozygous FWAIfwa plant flowers earlier than the homozygous fwa fwa plant). The flowering time of most of the plants of the mapping population is between the values of the parental lines, although a very small fraction of transgresive phenotypes might be present due to the segregation of some other flowering loci of minor effect differing between Ler and Col. All recombinants that were obtained between the different morphological markers were classified according to their flowering time. From this mapping population the recombinants between ga5 and emb35 were selected for the fine mapping of FWA, using molecular markers.

Mapping with morphological markers

The different classes of recombinant F2 plants segregating in the mapping population were used to estimate the genetic distances between the morphological markers cer2, ga5, emb35 and ap2 (Table 2). Table 2: Genetic distances between the morphological markers.

Because of the variation in flowering time, it was not possible to score the FWA phenotype of the F2 plants unambiguously. Therefore, the FWA genotype of the F2 plants was determined by analysing the flowering time of the F3 lines, derived from recombinant plants. These lines were only grown from the 120 recombinants between GA5 and EMB35. F3 lines from homozygous wα mutant F2 plants were completely late flowering; F3 lines from heterozygous FWAIfwa F2 plants segregated flowering time while F3 lines from homozygous wild type FWA F2 plants were early flowering. Out of these 120 recombinants, only two were recombinant between GA5 and FWA and 118 had undergone recombination between FWA and EMB35. This means that FWA maps only 0.1 cM from GA5.

These calculated distances are generally in agreement with the ones in the classical genetic map (Table2, http://genome-www3. stanford.edu/cgi-bin/AtDB/Genintromap). apart from the distance in between GA5 and FWA. In the classical map this distance is 1 cM. This is probably due to the relatively small mapping population previously used and the fact that the distance in this map is based on an integration of distances from different mapping populations.

Fine mapping with molecular markers

The genetic fine mapping of FWA was performed with molecular markers; 16 different restriction fragment length polymorphism (RFLP) markers and one codominant cleaved amplified polymorphism (CAPS) marker were used (Table 3).

Table 3: DNA probes used to detect molecular polymorphisms.

* INRA, Versailles Cedex, France ^** JIC, Norwich, UK Michigan State University, East Lansing, Ml, USA

^* CNRS, Gif-sur-Yvette, France

* ABRC = Arabidopsis Biological Resource Center, Ohio, USA. ^® AtDB = Arabidopsis thaliana database (http://genome- www.stanford.edu/Arabidopsis/)

DNA isolated from F3 lines derived from the 120 F2 plants, showing recombination between GA5 and EMB35, were analysed for polymorphic molecular markers. From this analysis the location of the FWA locus could be limited to a region of 0.7 cM, between the morphological marker gα5, and the molecular marker pcr23 (Figure 3). For gα5 also a molecular marker was available (Xu et αl, 1995), which was used as an extra check for the scoring of this morphological marker, indicating that for one recombinant the gα5 phenotype was misscored. Further fine mapping within this region narrowed the location of FWA down to a small region in between the markers CC128 and pcr28. There was only one recombinant with CC128 and there were two recombinants with pcr28. Recombinants between these flanking markers and the FWA locus were also recombinant between both molecular markers.

Physical mapping The YAC contig

Once FWA was located within a small region between CC128 and pcr28, a YAC contig was constructed in order to locate FWA within a YAC. For this purpose nine YAC's were selected from the published YAC contig of chromosome 4 (Schmidt et al, 1995). The relative positions of these YAC's were refined by hybridising them with all the molecular markers in this region that were used for the mapping. The relative position of a YAC was deduced, according to whether a marker hybridised completely, partially or not at all with the YAC. The YAC contig made in this way is shown in Figure 3. A comparison of this YAC contig with the contig from Schmidt et al (1995) did not show any significant differences in the positions of YAC's. However, the use of more markers for the construction of the YAC contig for FWA did improve its accuracy. For instance, EG2A5 and CIC12H1 are not overlapping in Schmidt's contig but they are in the FWA contig because in the first contig no probe was used located in the overlapping part. Since the molecular sizes of the YAC's could be estimated, it was possible to compare genetic and physical distances in the FWA region. The genetic distance of 0.7 cM between ga5 and pcr23 corresponds to a physical distance between 200 and 250 Kb. This means that the ratio of physical to genetic distance in this region of chromosome 4 is about 300 Kb/cM. The average ratio for this chromosome is 175 Kb/cM, varying from 30 Kb/cM to more than 550 Kb/cM (Schmidt et al, 1995). Therefore the ratio in the FWA region is higher than average, which is not favourable for map-based cloning. If this ratio was lower it would have been possible to further reduce the physical distance where FWA is located, using the same number of recombinants.

The cosmid contig

The/rø-mutant is semi-dominant and probably a gain of function mutation. This raises the possibility that complementation of a mutant plant with the wild type gene might not be possible. Therefore the complementation experiment should be done by transforming a wild type plant with the fwa mutation. In this case a complementing transformant should confer later flowering to wild type plants.

To achieve this, a genomic library was made from fwa-J mutant DNA. This library was constructed in a cosmid binary vector because of the relatively large insert size and the advantage of being able to use the clones directly for plant transformation. The resulting cosmid library consists of 27.264 clones with an average insert size of 16 Kb. Thus, in theory, the library should contain four genome equivalents. However, due to cosmids without a good insert, the library probably contains between two and four genome equivalents. From the contig a YAC, covering the genomic region that contains FWA, was selected that could be used for the screening of this library. This was the case for YAC EG1F12, which contains both markers, CC128 and pcr28, flanking the FWA locus. The library was screened by hybridisation with this YAC and 21 hybridising cosmids were obtained. Four pairs of these clones were identical, which means that the screen yielded in total 17 different cosmids. By hybridising these cosmids with each other and with several YAC's and molecular markers in this region, they could be arranged in a contig (Figure 3). The overlaps between the different cosmids were at least five Kb, apart from the overlap between cosmids 2/5 and 120, which was only a few Kb. Ten of the cosmids covered the region between the markers CC128 and pcr28. Some cosmids were used to find RFLP's that cosegregated with FWA. Indeed such RFLP's were found with both cos28 and cos94. Between these polymorphisms and FWA there were no recombinants. However, it was not possible to limit the region where FWA is located further. The recombination events between FWA and pcr28, detected in two recombinants, occurred between the left end of cos94 and the right end of pcr28. The recombination between FWA and CC128 occurred between the left end of CC128 and the right end of cosl20.

Because the molecular sizes of the cosmids could be estimated, the region in which FWA is located could be delimited to 60 Kb.

Plant transformation and complementation

Nine cosmids were selected for the plant transformation experiment. These cosmids span the complete region where FWA is located, ranging from the left end of CC128 to the right end of pcr28. All these cosmids were introduced in wild type Ler plants. The number of Tl transformants from every cosmid that was checked for flowering time and the flowering time behaviour of these transformants are shown in Table 4

Table 4 Numbers of obtained T1 Transformants

Several Tl transformants of cos20 and cos28 were clearly flowering later than the wild type However, most of the transformants with these two fwα cosmids flowered as early as the wild type and therefore did not show complementation For both cosmids it was shown by PCR analysis that all the late flowering Tl transformants contained the insert but several of the early flowering transformants did not It is also possible that the fwα mutant gene is not always expressed in the plant, depending on the place in the genome where it is inserted

Complementation of the phenotype by transformation is never achieved in 100% of the Tl transformants because several causes might impede the right expression of the inserted genes (positional effect, cosupression, rearrangements, loss of the transgenes) However, in the fwa case the frequency of complementing transformant was rather low compared to literature data for other flowering genes which suggest that these or other reasons might be particularly important in the fwa case.

The cosmids 20, 28 and 31 were also transformed to fwa mutant plants. With cosmids 20 and 28 several early flowering Tl transformants were obtained which is possibly caused by cosuppression of the fwa mutant gene.

The FWA gene

The sequence of the FWA genomic region was obtained from the thus generated database, together with open reading frames (ORF's). In the overlap of cosmids 20 and 28 only one complete ORF was found. This ORF has homology to homeodomain genes and highest homology with ANL, which is a homeobox gene, involved in the accumulation of anthocyanin. The nucleic acid sequences according to the invention exhibit a higher degree of homology and identity with the sequences of Sequence id no 1-4 than with ANL as illustrated in SEQUENCE ID No. 5. They also illustrate a higher degree of homolgy with a fwa gene encoding sequence than with the ANL sequence of Sequence id no 5.

Expression studies of the FWA gene by northern blotting and RTPCR indicated that the gene is only expressed in the mutant and not in the wild-type. Furthermore, the promoter region was analysed and shows two repeating sequences, one of 30 bp and one of 200 bp. Bisulphite sequencing of these repeats revealed that in the wild-type the repeats are hypermethylated but in the mutant they are hypomethylated. Probably this hypermethylation of the repeats prevents expression of the FWA gene in the wild-type plant. Detailed genetic and physical mapping located the FWA locus 0. IcM below ga5 in a region of 60 Kb. Transformation with two overlapping cosmids obtained from, fwa-1 mutant DNA converted late flowering to Ler wild type plants. This indicated hat fwa is located in the overlap between both cosmids and that the mutant allele behaves as a gain of function allele, suppressing flowering.

Materials and methods Plant material Seeds of the Columbia (Col) ecotype and Col containing the emb35 mutation were obtained from David Meinke (Oklahoma State University, Stillwater, OK, USA). The Landsberg erecta (Ler) marker line containing the mutations cer2-l, ga5-l, fwa-1 and ap2-l was generated by Maarten Koornneef.

Seeds were sown in plastic Petri dishes on a filter paper soaked with water and incubated in a cold room (4°C) for three days. After this they were transferred to a climate room (25°C, 16 hours light per day) and incubated for two days. Germinated seeds were planted on potting compost in individual clay pots and grown in a greenhouse with long-day light conditions (at least 14 hours daylight).

Genetic analysis

To estimate the recombination fraction in the mapping population the RECF2 program, which produces maximum likelihood estimates and standard errors, was used (Koornneef and Stam, 1992). For the construction of the linkage map the JOINMAP program (Stam, 1993; Stam and van Ooijen, 1995) was used applying the Kosambi function (Kosambi, 1944) to convert recombination fractions into genetic distances.

DNA isolation

DNA was isolated from plants grown in the greenhouse, following basically the protocol of Bernatzky and Tanksley (1986). Approximately 4 g of fresh leaf material was ground in a mortar filled with liquid nitrogen. The powder was transferred to a tube containing 20 ml extraction buffer (0.1 M Tris pH7.5, 0.35 M Sorbitol, 5 mM EDTA). After centrifuging at 4000rpm for 30 min the supernatant was discarded and 1.25 ml extraction buffer, 1.75 ml nuclei lysis buffer (0.2 M Tris pH7.5, 50 mM EDTA, 2M NaCl, 2% CTAB) and 300 μl 10% sarkosyl were added, mixed with the pellet and incubated at 65°C for 30 min. Then 7.5 ml chloroform isoamylalcohol (24: 1) was added and the tube was rotated for 15 min at room temperature (RT). After centrifuging at 4000 rpm for 30 min, 1 volume of isopropanol was added to the upperphase to precipitate the DNA. The tube was centrifuged again for 30 min; the pellet was dried and dissolved in 400 μl sterile milli-Q water (mQ). RNAase A was added to an end concentration of 10 μg/ml and the tube was incubated at 37°C for 30 min. The solution was extracted twice, first with phenol/chloroform isoamylalcohol (25:24: 1) and then with chloroform/isoamylalcohol. Thereafter the DNA was precipitated with 0.1 volume of 3 M NaAc (pH 5.2) and 2.5 volumes of 96% ethanol, washed with 70% ethanol and dissolved in an appropriate volume of sterile mQ. DNA concentrations were measured with a TKO 100 fluorimeter (Hoefer Scientific Instruments, San Francisco, CA, USA). Plasmid and cosmid DNA was isolated, following the "small-scale preparations of plasmid DNA" protocol of Sambrook et al (1989). When the DNA was used as a probe it was purified with Qiagen-tip 20 columns (Qiagen, Chatsworth, CA, USA) following the manufacturers instructions.

Phage DNA was isolated, following the "rapid analysis of bacteriophage λ isolates, plate lysate method" protocol of Sambrook et α/ (1989). Total genomic YAC DNA was isolated from a 5 ml culture of yeast, which was grown in YPD medium (10 g yeast extract, 20 g peptone and 20 g dextrose per liter) at 30°C. After centrifuging the culture at 4K for 5 min, the pellet was washed in 5 ml of 50 mM EDTA, then washed in 20 mM EDTA, 1 M sorbitol; after this it was resuspended in 150 μl of 20mM EDTA, 1M sorbitol. Hereafter 35 μl lyticase (5U/μl) and 11.5 μl β- mercaptoethanol was added and the solution was incubated for 2 hours at 37°C. After centrifuging at 1200 g for 5 min, the pellet was dissolved in 0.5 ml of 0.1 M EDTA, 0.15 M NaCl, then 25 μl of 20% SDS was added and the solution was incubated at 65°C for 20 min. Next, 200 μl of 5 M KAc was added and the tube was left on ice for 30 min after which it was centrifuged for 3 min. The supernatant was poured in a 1.5 ml Eppendorf tube that was filled with 96% ethanol and then centrifuged for 10 min at RT. The pellet was resuspended in 250 μl of mQ, after which an equal volume of 4.4 M LiCl was added and the tube was left on ice for 30 min. After centrifuging for 5 min the supernatant was taken and the DNA was precipitated with 96% Ethanol and washed twice with 70% Ethanol. Finally the DNA was dissolved in 50 μl mQ. Complete YAC's were isolated from 100 ml cultures of yeast. Cells were pelleted and washed as described above. After washing, the pellet was warmed to 38°C and 14 μl lyticase (5U/μl), 4.6 μl β-mercaptoethanol and 180 μl low melting agarose was added (amounts should be adapted, according to the volume of the pellet). After mixing quickly, the solution was transferred to a mould to cast plugs. The plugs were transferred to a small volume of LET (0.5 M EDTA, 10 mM Tris pH8.0) with 7.5 μl β- mercaptoethanol and 0.1 mg/ml RNAaseA and incubated overnight at 37°C. Hereafter they were washed three times in NDS buffer (0.5 M EDTA, 10 mM Tris pH8.0, 1% sodium N-Lauroylsarcosine) for 15 min. Then they were transferred to NDS with 2 mg/ml proteinase K and incubated overnight at 50°C. Finally they were washed in 50 mM EDTA pH8.0 for 15 min, left overnight in fresh 50 mM EDTA and washed again. The plugs were stored at 4°C in 50 mM EDTA pH 8.0. To separate complete YAC's, the plugs were cast in a 1% agarose gel, which was run by pulsed field gel electrophoresis in a CHEF-DR™II (Bio-Rad, Hercules, CA, USA) apparatus.

Preparation of probes

The insert of a cosmid or plasmid was released by digestion with the appropriate restriction enzymes. The resulting fragments were separated by gel electrophoresis and the band(s) corresponding to the insert were cut out of the gel. YAC's were released as described above.

The DNA was released from these agarose blocks by electro-elution. For this an electro-elution device (Harvard Bio Labs Machineshop, Cambridge, MA, USA) was filled with elution buffer (lOmM Tris pH 7.5, 5mM NaCl and ImM EDTA) and 70μl of 20% NaAc was added to the salt bridge. Two μl of loading buffer was added to the agarose blocks and these were put in the reservoir. Electrophoresis lasted 45 minutes at 80V after which the DNA was pipetted out of the salt trap (two times 175μl). The DNA was first extracted with phenol/chloroform/IAA (25:24: 1), then with chloroform/IAA (24: 1) and finally precipitated with 2.5 volumes of absolute ethanol overnight at -20°C. The precipitate was washed with 70% ethanol and dissolved in mQ water.

Southern blotting and hybridisation

Three μg genomic plant DNA were cut with the appropriate restriction enzymes and the DNA fragments were separated by agarose gel electrophoresis. Thereafter they were transferred to a Hybond-N nylon membrane (Amersham Pharmacia, Uppsala, Sweden) by vacuum blotting following the procedures recommended by the manufacturer (Pharmacia LKB-VacuGeneXL, Amersham Pharmacia, Uppsala, Sweden). The time periods for depurination, denaturation, neutralisation and transfer were respectively 10 min, 10 min, 10 min and 2 hours. After blotting, the blot was soaked in 2 x SSC for 1 min, UV irradiated in an ultraviolet crosslinker (Ultra Lum, Paramount, CA, USA) with 120,000 μJ/cm² and baked at 80°C for 2 hours. Hybridisations were performed in a Hybaid oven (Hybaid, Teddington, UK). A blot was prehybridised with 10 ml of hybridisation solution (5 x SSC, 5 x Denhardt's solution and 0 5% SDS) for 4 hours at 65°C [³ P] Random prime labelled DNA fragments were used as probe for hybridisation overnight Blots were washed at 65°C in 0 1% SDS and respectively 5 x SSC, 3 x SSC and 1 x SSC (every wash step took half an hour) The activity of a blot was visualised with a phosphor imager

Cosmid library

The T-DNA cosmid vector 04541 was used to prepare the genomic library This vector was derived from SLJ1711 (Jones et al, 1992) by the insertion of a fragment containing a cos site between the Bglll sites SLJ1711 was derived from pRK290 (Ditta et al, 1980) The vector contains the kanamycine resistance gene (NPTII), a cos site and a polylinker, with blue/white selection, between T-DNA borders Furthermore it carries a SURE™ bacterial tetracycline resistance gene

To prepare the library, genomic DNA of the fwa-\ mutant was partially digested with the restriction enzyme Sau3AI, treated with calf intestinal phosphatase and size fractionated over a sucrose gradient to obtain fragments in between 15 and 25 Kb

These fragments were ligated into the BamHI site of the cosmid vector After that the DNA was packaged with Gigapack II packaging extract (Stratagene, La Jolla, CA, USA), mixed with SURE™ cells (Stratagene, La Jolla, CA, USA) and plated out on LB (10 g peptone, 5 g yeast extract and 5 g NaCl per liter) plates with tetracycline (lOμg/ml), 0 004% Xgal and 0 2 mM IPTG for blue/white selection Single white colonies were picked and put into wells of high density (384 wells) microtitre plates (Genetix, Dorset, UK) that were filled with freezing medium (LB, containing 36 mM K₂HPO₄, 13 2 mM KH₂PO₄, 1 7 mM Na citrate, 0 4 mM MgSO₄, 6 8 mM (NH4)₂SO₄ and 4 4% glycerol) In total 71 high density plates were filled and stored at -80°C To prepare library filters, cells were transferred from the microtitre plates to agar plates with a replicator and grown overnight Then Hybond-N filters (Amersham Pharmacia, Uppsala, Sweden) were placed on the plates with colonies for 1 minute, denatured and neutralised in trays containing these solutions and baked at 80°C for 2 hours Hybridisation of the filters was similar as mentioned above (southerns, blotting and hybridisation), but the filters were hybridised in trays instead of bottles

Electroporation of Agrobacterium tumefaciens Cosmids that were selected for plant transformation were transferred from Escherichia coli cells (SURE™) to Agrobacterium tumefaciens (AGLO strain; Lazo et al, 1991) by electroporation. To prepare competent cells a 50 ml liquid culture of LB with selective antibiotics was inoculated with A. tumefaciens and grown overnight at 28°C. The next day a 500 ml liquid culture (LB without salts) was inoculated with 25 ml of the overnight culture. Cells were harvested at OD600 by centrifugation (5K, 5min, 4°C) and gently resuspended in 250 ml of ice-cold mQ water. Thereafter cells were centrifuged again and resuspended in 100 ml of ice-cold mQ water. Finally, the cells were resuspended in 10 ml of ice-cold 15% glycerol in mQ water, aliquoted in 100 μl portions and stored at -80°C.

For electroporation an aliquot was thawed and 1 to 5 μl of cosmid DNA was added. The mixture was transferred to a cuvette, which was placed in the cuvetchamber of an electroporator set on 2.2 kV (E. coli pulser from Bio-Rad, Hercules, CA, USA), after which a pulse was given. Immediately after the pulse one ml of SOC (2% bactotrypton, 0.5% yeast extract, lOmM NaCl, 2.5 mM KCl, 10 mM MgCl₂, 10 mM MgSO₄, 20 mM glucose) was added. The SOC medium with cells was transferred to a sterile tube and incubated for 1 to 2 hours (225 rpm, 28°C). Subsequently, the cells were plated on LB plates with selective antibiotics and grown for two days at 28°C.

Transformation of Arabidopsis plants

For transformation of Arabidopsis the protocol of Bechtold et al (1993) was adapted. A. tumefaciens cells of the strain AGLO (Lazo et al, 1991), with the appropriate cosmid, were grown in 15 ml liquid culture (LB with 50 μg/ml kanamycine and 50 μg/ml. rifampicine) at 28°C during 48 hours. One day before transformation 4 flasks with 0.5 liter of liquid medium were inoculated with 0.5 ml of the 15 ml culture and grown overnight. The cells were harvested at an OD600 of 0.8 by centrifugation (5K, 15min, RT) after which the pellet was gently resuspended in 0.5 liter of infiltration medium, pH 5.8 (0.5 X Murashige & Skoog salts, 5% sucrose, 0.05% MES, 0.02% Silwet L-77 (Lehle seeds, Round Rock, TX, USA). The infiltration medium with tumefaciens was put in two jars on top of which the pots with Arabidopsis were placed upside down with the flowering shoots completely submerged in the medium. Thereafter the jars with pots were placed in vacuum for five minutes. Finally, the pots with Arabidopsis were transferred to the greenhouse. The seeds that were harvested from these plants were sterilised for 15 minutes with 20%) bleach in absolute ethanol solution, then they were rinsed two times in absolute ethanol and dried overnight in a flow cabinet. Seeds were sown on plates with selective medium (1 X Murashige & Skoog salts, 1% sucrose, 40μg/ml kanamycine, 0.8% agar, pH 5.8). The plates were kept in the cold room (4°C) for 4 days and then transferred to the growth room (16 hours light, 25°C). After 10 days, transformed seedlings were visible as green plants with several green leaves and a root, whereas untransformed seedlings were yellow and did not develop further than the cotyledons.

PCR analysis

DNA was isolated from a few leaves of a transformant plant and amplified through 35 cycles (10 sec 94°C, 30 sec 54°C and 2 min 72°C) in standard PCR conditions. Presence of the cos20 insert in the plant was confirmed by appearance of a 1.1 Kb band after amplification with the T3 primer, 5'-AATTAACCCTCACTAAAGGG-3' (Sequence id no 6) and the primer 5'-GCTTCGGAACTAAGGAACCCAAGC-3 ' (Sequence id no 7). For cos28 a 0.8 band was amplified, using the T3 primer and the primer 5'-GAGTCTTGCTTTATGCCAAGCCGC-3' (Sequence id no 8).

Additional scientific description of the cloning of FWA Expression analysis:

In addition to the proof that lateness can be conferred by transformation with cos20 and cos28, which contains only one possible ORF (called irø) the expression of this gene was analysed in wild type and various fwa mutants using RT-PCR. For the RT-PCR experiments, genomic DNA and primers derived from the ANL2 gene, which is a gene closely related to FWA and recently cloned in our laboratory (Kubo et al. 1999), were used as controls. RT-PCR analysis was confirmed by northern blot analysis and showed moreover FWA is not expressed in wild type Ler but expressed abundantly in the various fwa mutants. The reason for this lack of expression seems methylation of two repeated regions in the promoter (shown in red in the sequence below). This was confirmed by bisulfite genomic sequencing of this region in wild type and various / v alleles by our collaborator Dr Steve Jacobsen at UCLA in the USA.

Sequence analysis. Sequencing of the FWA region in wild type and mutants revealed that the lack of expression in the wild type (in contrast to the suppressors) is not due to a mutation within the FWA region The phenotype of the suppressor mutants further confirmed that FWA is not essential for normal development of the plant and that its expression results in lateness, which property we expect to be transferred to other plants too A third proof that the fwa sequence confers lateness is that sequencing of two intra- genic suppressors (1R1 and 1R2) showed mutations that either lead to an early stop codon (1R1) or an amino acid change in a, probably for the function, crucial part of the gene (table 6 and gDNA sequence, Sequence id no 1) In these intra-genic suppressor mutants the promoter region is still un-methylated (Jacobsen pers com ), so that we can assume that lack of expression of fwa is due to a non-functional gene in these suppressor mutants Sequencing of the FWA region in wild type and mutants revealed that the lack of expression in the wild type (in contrast to the suppressors) is not due to a mutation within the FWA region The phenotype of the suppressor mutants further confirmed that FWA is not essential for normal development of the plant and that its expression results in lateness, which property can be transferred to other plants too

SEQUENCE INFORMATION LOCUS FWA 5429 bp ds-DNA SYN 12-APR-1999 KEYWORDS flower ng/homeodomam/Arabidops s thaliana SOURCE gDNA

ORGANI SM Arabidopsis thaliana . BASE COUNT 1593 a 940 c 1001 g 1895 t Table 5 Genomic sequence components

Sequence id no 1 FWA1 and Ler gDNA 1 tctagagttacaacttcaaccgagaaagtctatcaatgcatatttccatttaaagtgtgacctgtcattg ttaaaagtttattattgttgaataaatttagttaacaatttcgtttatctgtttttaatattatgggaag gagtcatttttcactaagcatatagattttctaatgagtatccctatatatatattaaacttcttttttc ggtcaatacaattttataatctttcattttttctatcatttcatatcattgtaactataaattttcgtaa atagacctttagtgttaatacaatagatttttattaattttatatcggattttgtttaaaaaagaaaaac cataggatggatgatgattggtacttataagattgtaattgggtatttttggattgttaccaccattaca aagctattaacagagattgaagatatcacacaatgagagcgccacagcttcagcaacgtcccatgcagct gatgtgccttcgcctttctcttcctcatctgcgcttataaataaggcaaagcaactagaaaagattaaaa ccaaaaccaaaacaaaaaactagttaagaccctgattttgtttcataggtacatgcacttttcaacattg atttttgttgttaaaaataaaatccatgtgaaggttctcatcatataccgaaagaatgggaaatttgaaa attccatactttttaaaaaagacaatttgttttatcactttagtttttttatatattcagcgtctaccaa atctacacttttttttctttctcgatttagttaatcttcgttcttatgtcatgtaatagattactatttc aaaacatagatatttagttatctaaataaaactaggccatccatggatggtttcaattttatttttcata tgaactcataaaagaaaagttaaatttcatttcacaataaccattgattactaaatttagtaaagaatca attgggtttagtgtttacttgtttaaggtatttttttttcttttgttatggttctatactaatatcgaag agttatgggcCGAAGCCCATACATCTTTCCGTCGAGAATCTCATATAttctttatCGAAGCCCATACATC TTTCCGTCGAGAATCTCATATAtaccTTATCCCATTCAACATTCATACGAGCGCCGCTCTAGGGTTTTTG CTTTTCGCCATTGGTCCAAGTGCTATTTGGTTGTTTAAGGTTGCTTTTAGCACACAACTTTAATATTATT TTTATGTTTTTTTTCTTACGATTTATCGATTTGTGGGATACTGACAATCAGATTATTGTTGTTTTTTCCA GCCAAAtatcagatcttgcgccgctcTTTATCCCATTCAACATTCATACGAGCACCGCTTTACGGTTTTT GCTTTTCGGCATTGGTCGAACTGCTATTTGGTTGTTTAAGGTTGCTTTTAGCACACAACTTTAATATTAT TTTTATGTTTTCTTCTTACGATTTATCGATTTGTAGGATACTGACAATCAGATTTTTGTTGTTTTTTTCA GCCAAAaatcagattttttttacactttgtttagagaatgattttggttcccgatttgtctgtttttggc ttatgtgtaaagtactttgaaaaatatt||tgttaactctacaatgggtatcccaagttttggagttcttt tgtcttgttcgttgtcgagacactagaaatgttaatttaattctcttcttccaaaaagaaccatttactg gttcattctctacttgaattttattctggttgtatttcttttccagtataaagcagattgttttttgtta tttttcagtttagattggctttgtctcttttgagttgttgcaattgtcaaactttggaatgaaatagtaa ataatcttaggttggtagtaaatcttaacattgtgtttttggggcataatttatcgataaaatcttcagc attaaaaccaaaaagaaaaaactttttaagtcttttttgtttggtggttaaTATAAAgtttatacgtgta ttaatttgatcacactcactatatgtccagggagctaaacctctatatcgagtactaatagtatgcaatg tccaggttattgcattgagggaaaATGAATGGACAAGGTGATTTGGATGCGGTTGGAAACATTCCAAAAC CAGGTGAAGCTGAAGGCGATGAGATTGATATGATTAATGATATGTCTGGTGTTAATGATCAAGATGGTGG AAGGATGAGAAGAACCCATAGGCGCACTGCTTATCAAACTCAAGAACTTGAAaagtttgttcactttctt cttcatttcatcatcatgcaacatttcctattatttttttttatttttttattttgagtttggaatgttt ctctttactttgctctttactttaaaatgagtgtagtTTCTACATGGAAAATCCTCATCCCACTGAAGAA CAGAGGTACGAGCTTGGACAAAGGCTTAATATGGGTGTCAATCAAGTCAAGAATTGGTTCCAGAATAAAA GAAATCTCGAGAAGgtaccaaaaaatacagaagtgtatacatgcatgtggctgtgtcttttcttattctt tgcacgtaccgtattatggaccatgtttctgccctatgttaaaaacaaatactgtatatagtttaccacc tctctctctcatcctcaaaatctgccttgattatggtataattattgagtgtggttttttgggtaagcag ATCAATAATGACCACCTTGAGAATGTAACTCTTAGAGAAGAGCATGACAGATTGCTAGCAACTCAGGATC AGCTTAGAAGCGCAATGCTACGTAGTTTATGCAACATTTGTGGTAAGGCAACTAATTGTGGAGACACTGA ATATGAGGTGCAAAAACTTATGGCTGAGAATGCTAACTTGGAGCGGGAGATAGACCAGTTCAATTCCAGA TACCTTTCCCATCCCAAACAAAGGATGGTCAGTACATCCGAACAGGCGCCTTCTTCTTCCTCTAATCCAG GAATAAATGCAACACCAGTACTTGATTTTAGTGGTGGAACTAGGACGTCTGAGAAGGAGACATCAATTTT TCTGAATCTTGCCATTACGGCTTTGAGAGAGTTGATTACATTGGGAGAAGTGGACTGTCCATTTTGGATG ATAGATCCAATCGTTAGATCCAAAGGAGTATCAAAGATCTATGAGAAGTATAGAAGTTCTTTCAACAATG TCACAAAACCTCCCGGGCAAATTGTGGAGGCTTCAAGAGCTAAAGGTTTAGTTCCCATGACTTGCGTGAC TCTGGTCAAGACTCTTATGGACACAgtaataacctttaaaccaccacccattttttagtgaaatgttgtt acatgaaatgttaacaagttttttttttttttttttaaacagGGCAAATGGGTCAACGTGTTTGCACCTA TAGTCCCTGTGGCATCAACCCATAAAGTGATATCTACCGGTTCTGGTGGAACCAAAAGTGGCTCACTCCA ACAGgtatatatatagttttctgagtcgtctatattggttgataatccaagagaagtagaaacatttcaa atttgtgatcatataattttggcagATTCAAGCAGAATTTCAAGTAATTTCTCCGCTGGTACCAAAGAGA AAAGTAACGTTTATTAGATACTGCAAAGAGATCAGACAGGGCTTATGGGTGGTCGTCGACGTTACTCCTA CTCAAAATCCGACTTTGCTGCCCTATGGTTGTTCTAAGAGGCTACCCTCAGGCCTTATCATAGACGACCT GTCCAATGGGTACTCCCAGgttacttctgaatcctgttctttgaacaatatatgttggttcacttttggt cgcatggtatagatttgataacggctttctatagcctagctcaattctcacatgcttatgagattttttt ttttaaaaactcaaaacgtgttattactcgcaatcataattaatgatctatcagagcatcttggacttgt actgtggagacataaataaatgtgtaatggttattattatattcaccacatgtataacacatagagattt gatttcacagGTTACATGGATTGAACAAGCGGAATATAATGAGAGTCACATCCACCAACTCTACCAGCCT TTGATTGGCTATGGGATCGGGCTAGGTGCAAAGAGATGGCTCGCGACGCTGCAGAGACACTGCGAAAGCC TCTCGACCCTTTCATCTACCAACTTGACTGAAATTAGTCCAggtgaaagctatttaaattatagaataaa tctcattacaactactttttaGGTCATTTTCGGTTCTCCACTAAATTTGGACAGGCCTTGATTGTGATAT ATGTTTTCGATATA§GATTGTCTGCAAAAGGTGCAACTGAAATAGTGAAGCTAGCACAGCGAATGACTCT CAACTACTACAGAGGTATTACGAGTCCTTCGGTGGACAAGTGGCAGAAAATTCAGGTGGAGAATGTGGCA CAAAACATGAGTTTCATGATCCGAAAGAACGTGAATGAGCCTggtgagctaactgggattgtgctgagtg catccacttctgtttggctcccaGTGAACCAGCATACACTCTTTGCTTTCATTAGCCACCTGAGTTTCAG ACACGAGTGGGATATCTTGACCAATGATACTACCATGGAAGAAACAATCCGGATTCAAAAAGCAAAACGC CATGGAAACATCATCTCTCTTCTGAAAATCgttgtgagtttcaaaacatttatttttggtttaatgagtg tcgttttgtctatgtttccattgattttcgtgagttttgcaatgcaaacagAATAATGGTATGCTGGTTC TGCAAGAGATTTGGAATGATGCATCAGGTGCAATGGTGGTGTATGCACCAGTGGAAACCAATTCTATTGA GCTGGTCAAGAGAGGTGAAAATTCAGATTCTGTGAAGTTTCTTCCTTCGGGATTTTCGATAGTGCCAGAT GGAGTAAATGGGTCATATCATAGAGGCAATACTGGTGGAGGATGTCTACTGACATTTGGACTTCAGATCT TGGTGGGCATCAATCCAACTGCTGCACTCATTCAAGGTACTGTCAAAAGTGTCGAGACACTCATGGCTCA TACTATTGTCAAGATCAAATCCGCGTTAGATTTACAGACGTAAccatcagcagctccagtttccgttgtt ccgcaagaaattcgacattggcttgtgccctagtttgcttttccaaccaagactcttctggattaacttt tttatgcgtttgttgatctgtttgctggatatttcttgcttcctttcttctttccttttctgaaactctc aaattgttacaaaccagaattaatagagagttagaacaatatatttcgtattcaacattgatggagatca tgcttcactatatcaggtgaaaaatgcattccagaattc

Table 6 mutations m revertants

Sequence id no 2. cDNA based on computer prediction from the gDNA Created: woensdag 30 ]unι 1999 20:17 D

ATGAATGGACAAGGTGATTTGGATGCGGTTGGAAACATTCCAAAACCAGGTGAAGCTGAAGGCGATGAGA TTGATATGATTAATGATATGTCTGGTGTTAATGATCAAGATGGTGGAAGGATGAGAAGAACCCATAGGCG CACTGCTTATCAAACTCAAGAACTTGAATTCTACATGGAAAATCCTCATCCCACTGAAGAACAGAGGTAC GAGCTTGGACAAAGGCTTAATATGGGTGTCAATCAAGTCAAGAATTGGTTCCAGAATAAAAGAAATCTCG AGAAGATCAATAATGACCACCTTGAGAATGTAACTCTTAGAGAAGAGCATGACAGATTGCTAGCAACTCA GGATCAGCTTAGAAGCGCAATGCTACGTAGTTTATGCAACATTTGTGGTAAGGCAACTAATTGTGGAGAC ACTGAATATGAGGTGCAAAAACTTATGGCTGAGAATGCTAACTTGGAGCGGGAGATAGACCAGTTCAATT CCAGATACCTTTCCCATCCCAAACAAAGGATGGTCAGTACATCCGAACAGGCGCCTTCTTCTTCCTCTAA TCCAGGAATAAATGCAACACCAGTACTTGATTTTAGTGGTGGAACTAGGACGTCTGAGAAGGAGACATCA ATTTTTCTGAATCTTGCCATTACGGCTTTGAGAGAGTTGATTACATTGGGAGAAGTGGACTGTCCATTTT GGATGATAGATCCAATCGTTAGATCCAAAGGAGTATCAAAGATCTATGAGAAGTATAGAAGTTCTTTCAA CAATGTCACAAAACCTCCCGGGCAAATTGTGGAGGCTTCAAGAGCTAAAGGTTTAGTTCCCATGACTTGC GTGACTCTGGTCAAGACTCTTATGGACACAGGCAAATGGGTCAACGTGTTTGCACCTATAGTCCCTGTGG CATCAACCCATAAAGTGATATCTACCGGTTCTGGTGGAACCAAAAGTGGCTCACTCCAACAGATTCAAGC AGAATTTCAAGTAATTTCTCCGCTGGTACCAAAGAGAAAAGTAACGTTTATTAGATACTGCAAAGAGATC AGACAGGGCTTATGGGTGGTCGTCGACGTTACTCCTACTCAAAATCCGACTTTGCTGCCCTATGGTTGTT CTAAGAGGCTACCCTCAGGCCTTATCATAGACGACCTGTCCAATGGGTACTCCCAGGTTACATGGATTGA ACAAGCGGAATATAATGAGAGTCACATCCACCAACTCTACCAGCCTTTGATTGGCTATGGGATCGGGCTA GGTGCAAAGAGATGGCTCGCGACGCTGCAGAGACACTGCGAAAGCCTCTCGACCCTTTCATCTACCAACT TGACTGAAATTAGTCCAGGTCATTTTCGGTTCTCCACTAAATTTGGACAGGCCTTGATTGTGATATATGT TTTCGATATAGGATTGTCTGCAAAAGGTGCAACTGAAATAGTGAAGCTAGCACAGCGAATGACTCTCAAC TACTACAGAGGTATTACGAGTCCTTCGGTGGACAAGTGGCAGAAAATTCAGGTGGAGAATGTGGCACAAA ACATGAGTTTCATGATCCGAAAGAACGTGAATGAGCCTGTGAACCAGCATACACTCTTTGCTTTCATTAG CCACCTGAGTTTCAGACACGAGTGGGATATCTTGACCAATGATACTACCATGGAAGAAACAATCCGGATT CAAAAAGCAAAACGCCATGGAAACATCATCTCTCTTCTGAAAATCAATAATGGTATGCTGGTTCTGCAAG AGATTTGGAATGATGCATCAGGTGCAATGGTGGTGTATGCACCAGTGGAAACCAATTCTATTGAGCTGGT CAAGAGAGGTGAAAATTCAGATTCTGTGAAGTTTCTTCCTTCGGGATTTTCGATAGTGCCAGATGGAGTA AATGGGTCATATCATAGAGGCAATACTGGTGGAGGATGTCTACTGACATTTGGACTTCAGATCTTGGTGG GCATCAATCCAACTGCTGCACTCATTCAAGGTACTGTCAAAAGTGTCGAGACACTCATGGCTCATACTAT TGTCAAGATCAAATCCGCGTTAGATTTACAGACGTAA Translate cDNA Sequence FWA (1,2067) With Standard Genetic Code

Molecular Weight 76767.71 Daltons 688 Ammo Acids

71 Strongly Basic (+) Ammo Acids (K, R) 70 Strongly Acidic (-) Ammo Acids (D,E) 223 Hydrophobic Ammo Acids (A, I ,L, F, W,V) 209 Polar Ammo Acids (N, C, Q, S, T, Y) 7.663 Isolectπc Point 3.369 Charge at PH 7.0

Total number of bases translated is 2067

A = 30.87 [638] G = 23.42 [ 484 ] T = 25.98 [ 537 ] C = 19.74 [ 408 ] Ambiguous = 0 . 00 [ 0 ] A+T = 56.85 [ 1175 ] C+G = 43.15 [ 892 ]

Davi s , Botstem, Roth Melting Temp C . 82 .35

Wallace Temp C 6734.00

Codon usage: gca Ala (A) 17 ft cag Gln(Q) 18 # uug Leu(L) 13 # uaa Ter ( . ) 1 gcc Ala (A) 2 a Gln(Q) 37 (• Leu(L) 61 # uag Ter ( . ) 0 gcg Ala (A) 5 # gaa Glu(E) 21 # aaa Lys (K) 20 # uga Ter ( . ) 0 gcu Ala(A) 10 __ gag Glu(E) 22 ft aag Lys(K) 16 # Ter( . ) 1

Ala(A) 34 # Glu(E) 43 ff Lys( ) 36 # aca Thr (T) 11 aga Arg(R) 20 # gga Gly(G) 18 # aug Met(M) 19 # ace Thr (T) 9 agg Arg(R) 7 # ggc Gly(G) 8 ft Met (M) 19 # acg Thr (T) 6 cga Arg(R) 2 # ggg Gly(G) 5 # uuc Phe(F) 9 # acu Thr (T) 23 cgc Arg(R) 2 # ggu Gly(G) 20 f? uuu Phe(F) 12 ft Thr(T) 49 egg Arg(R) 3 # Gly(G) 51 # — Phe (F) 21 ff ugg Trp(W) 9 cgu Arg(R) 1 # cac Hιs(H) 6 ff cca Pro (P) 11 # Trp(W) 9

Arg(R) 35 # cau Hιs(H) 10 # ccc Pro(P) 6 ft uac Tyr (Y) 8 aac Asn(N) 11 # Hιs(H) 16 # ccg Pro(P) 2 ft uau Tyr (Y) 10 aau Asn(N) 33 # aua Ile(I) 10 ff ecu Pro(P) 10 ft Tyr (Y) 18

Asn (N) 44 # auc Ile(I) 16 # Pro(P) 29 # gua Val (V) 7 gac Asp(D) 10 # auu Ile(I) 23 ft age Ser (S) 3 ft guc Val (V) 13 gau Asp(D) 17 # Ile(I) 49 if agu Ser (S) 11 ft gug Val (V) 21

Asp(D) 27 # cua Leu(L) 6 ft uca Ser (S) 10 # guu Val (V) 8 ugc Cys(C) 4 # cue eu(L) 9 ff ucc Ser (S) 9 # Val (V) 49 ugu Cys(C) 5 # cug Leu(L) 12 fr ucg Se (S) 4 # nnn ??? (X) 0

Cys(C) 9 # cuu Leu(L) 16 ff ucu Ser (S) 15 # TOTAL 689 caa Gln(Q) 19 # uua Leu(L) 5 ff ___ Ser(S) 52 #

Created : woensdag 30 j uni 1999 20 : 17

Seguence id no 3

Predicted Protein 3 letter code MetAsnGlyGlnGlyAspLeuAspAlaValGlyAsnlleProLysPro

GlyGluAlaGluGlyAspGluIleAspMetlleAsnAspMetSerGly

ValAsnAspGlnAspGlyGlyArgMetArgArgThrHisArgArgThr

AlaTyrGlnThrGlnGluLeuGluPheTyrMetGluAsnProHisPro ThrGluGluGlnArgTyrGluLeuGlyGlnArgLeuAsnMetGlyVal

AsnGlnValLysAsnTrpPheGlnAsnLysArgAsnLeuGluLysIle

AsnAsnAspHisLeuGluAsnValThrLeuArgGluGluHisAspArg

LeuLeuAlaThrGlnAspGlnLeuArgSerAlaMetLeuArgSerLeu

CysAsnlleCysGlyLysAlaThrAsnCysGlyAspThrGluTyrGlu ValGlnLysLeuMetAlaGluAsnAlaAsnLeuGluArgGluIleAsp

GlnPheAsnSerArgTyrLeuSerHisProLysGlnArgMetValSer

ThrSerGluGlnAlaProSerSerSerSerAsnProGlylleAsnAla

ThrProValLeuAspPheSerGlyGlyThrArgThrSerGluLysGlu

ThrSerllePheLeuAsnLeuAlalleThrAlaLeuArgGluLeuIle ThrLeuGlyGluValAspCysProPheTrpMetlleAspProIleVal

ArgSerLysGlyValSerLysIleTyrGluLysTyrArgSerSerPhe

AsnAsnValThrLysProProGlyGlnlleValGluAlaSerArgAla

LysGlyLeuValProMetThrCvsValThrLeuValLysThrLeuMet

AspThrGlyLysTrpValAsnValPheAlaProIleValProValAlaSer ThrH sLysVallleSerThrGlySerGlyGlyThrLysSerGlySerLeu

GlnGlnlleGlnAlaGluPheGlnVallleSerProLeuValProLysArg

LysValThrPhelleArgTyrCysLysGluIleArgGlnGlyLeuTrpVal

ValValAspValThrProThrGlnAsnProThrLeuLeuProTyrGlyCys

SerLysArgLeuProSerGlyLeuIlelleAspAspLeuSerAsnGlyTyr SerGlnValThrTrpIleGluGlnAlaGluTyrAsnGluSerHisIleHis

GlnLeuTyrGlnProLeuIleGlyTyrGlylleGlyLeuGlyAlaLysArg

TrpLeuAlaThrLeuGlnArgHisCysGluSerLeuSerThrLeuSerSer

ThrAsnLeuThrGluIleSerProGlyHisPheArgPheSerThrLysPhe

GlyGlnAlaLeuIleVallleTyrValPheAspIleGlyLeuSerAlaLys GlyAlaThrGluIleValLysLeuAlaGlnArgMetThrLeuAsnTyrTyr

ArgGlylleThrSeiProSerValAspLysTrpGlnLysIleGlnValGlu

AsnValAlaGlnAsnMetSe PheMetlleArgLysAsnValAsnGluPro

ValAsnGlnHisThrLeuPheAlaPhelleSerHisLeuSerPheArgHis

GluTrpAspIleLeuThrAsnAspThrThrMetGluGluThrlleArglle GlnLysAlaLysArgHisGlyAsnllelleSerLeuLeuLysIleAsnAsn

GlyMetLeuValLeuGlnGluIleTrpAsnAspAlaSerGlyAlaMetVal

ValTyrAlaProValGluThrAsnSerlleGluLeuValLysArgGlyGlu

AsnSerAspSerValLysPheLeuProSerG yPheSe lleValProAsp

GlyValAsnGlySerTyrHisArgGlyAsnThrGlyGlyGlyCysLeuLeu Th PheGlyLeuGlnlleLeuValGlylleAsnProThrAlaAlaLeuIle

GlnGlyThrValLysSerValGluThrLeuMetAlaHisThrlleValLys

IleLysSerAlaLeuAspLeuGlnThr

Sequence id no 4 Predicted Protein 1 letter code

MNGQGDLDAVGNIPKPGEAEGDEIDMINDMSGVNDQDGGRMRRTHRRTAYQTQELEFYMENPHPTEEQRY ELGQRLNMGVNQVKNWFQNKRNLEKINNDHLENVTLREEHDRLLATQDQLRSAMLRSLCNICGKATNCGD TEYEVQKLMAENANLEREIDQFNSRYLSHPKQRMVSTSEQAPSSSSNPGINATPVLDFSGGTRTSEKETS IFLNIΛITALRELITLGEVDCPFWMIDPIVRSKGVSKIYEKYRSSFNNVTKPPGQIVEASRAKGLVPMTC VTLVKTLMDTGKWVNVFAPIVPYASTHKYISTGSGGTKSGSLQQIQAEFQVISPLVPKRKVTFIRYCKEI RQGLWVWDVTPTQNPTLLPYGCSKRLPSGLIIDDLSNGYSQVTWIEQAEYNESHIHQLYQPLIGYGIGL GAKRWLATLQRHCESLSTLSSTNLTEISPGHFRFSTKFGQALIVIYVFDIGLSAKGATEIVKLAQRMTLN YYRGITS PSVDK QKIQVENVAQNMSFMIRKNVNEPVNQHTLFAFI SHLSFRHEWDILTNDTTMEETIRI QKAKRHGNI I SLLKINNGMLVLQEI NDASGAMWYAPVETNS IELVKRGENSDSVKFLPSGFS IVPDGV NGSYHRGNTGGGCLLTFGLQILVGINPTAALIQGTVKSVETLMAHTIVKIKSALDLQT .

A third proof that the fwa sequence confers lateness is that sequencing of two intra- genie suppressors (1R1 and 1R2) showed mutations that either lead to an early stop codon (1R1) or an amino acid change in a, probably for the function, crucial part of the gene (Table 6 and gDNA sequence, sequence id no 1) In these intra-genic suppressor mutants the promoter region is still un-methylated (Jacobsen pers com ), so that we can assume that lack of expression of fwa is due to a non-functional gene in these suppressor mutants Sequencing of the FWA region in wild type and mutants revealed that the lack of expression in the wild type (in contrast to the suppressors) is not due to a mutation within the FWA region The phenotype of the suppressor mutants further confirmed that FWA is not essential for normal development of the plant and that its expression results in lateness, which property can be transferred to other plants too Description of the Figures

Figure 3 Position of the fwa locus on chromosome 4 The upper part of the figure shows the whole chromosome with some morphological markers Below this the fwa region is shown with morphological and molecular markers that were used for the mapping of the fwa locus The middle of the figure shows the YAC contig from a small part of this region, together with the probes that were used to construct this contig The number of recombinants that \Λ ere left between these probes is indicated The cosmid contig that was generated after screening of the fwa-1 cosmid library with YAC EG1F12 is shown m the bottom of the figure Cosmids in white were used for plant transformation expeπments

Figure 6 representation of the pedigree of the mapping population

Figure 7 shows the segregation of flowering time in this mapping population The plants were grown under long day light conditions in a greenhouse and in these conditions Col flowered between 27 and 31 days, whereas the progeny of the parental fwa mutant plant that was selected for the cross flowered between 42 and 51 days The overall shape of the flowering time frequency distribution with two major peaks of different size can be explained because approximately 2/3 of these plants will be heterozygous for fwa (the heterozygous FWAIfwa plant flowers earlier than the homozygous/vrø/vra plant) The flowering time of most of the plants of the mapping population is between the values of the parental lines, although a very small fraction of transgressive phenotypes might be present due to the segregation of some other flowering loci of minor effect differing between her and Col The Figure shows all recombinants that were obtained between the different morphological markers classified according to their flowering time. From this mapping population the recombinants between ga5 and emb35 were selected for the fine mapping of FWA, using molecular markers.

Claims

I A nucleic acid sequence encoding the amino acid sequence of SEQUENCE ID nr 1,2,3 or 4 said nucleic acid sequence encoding a flowering time associated protein from the vernalisation pathway 2 An equivalent nucleic acid sequence of a nucleic acid sequence according to claim 1

3 A nucleic acid sequence encoding a fragment of the amino acid sequence of SEQUENCE ID nr 1, 2,3 or 4 the nucleic acid sequence encoding a flowering time associated polypeptide from the vernalisation pathway, said polypeptide being at least 4 amino acids in length

4 A nucleic acid sequence according to claim 1 or 2 encoding a homologous amino acid sequence exhibiting flowering time associated activity analogous to that of the protein of SEQUENCE ID No 1,2,3 or 4

5 A nucleic acid sequence according to any of the preceding claims capable of hybridising under stringent to moderate conditions with the encoding nucleic acid sequence of SEQUENCE ID No 1 or 2

6 A mutant of a nucleic acid sequence according to any of claims 1-5, the mutation being such that the expression product of the nucleic acid sequence exhibits reduced flowering time associated activity in comparison to the expression product of the equivalent non mutant nucleic acid sequence, when the sequences are operatively linked to the identical promoters

7 A nucleic acid sequence according to any of the preceding claims operatively linked to an expression regulating sequence not normally associated with the fwa gene 8 A nucleic acid sequence according to any of claims 1-7 operatively linked to an inducible promoter 9 A vector comprising a nucleic acid sequence according to any of the preceding claims, said vector preferably being constructed such that it is suitable for transferring the nucleic acid to a plant, plant cell or plant part 10 A plant, plant part or plant cell into which a nucleic acid sequence according to any of the claims 1-8 or a vector according to claim 9 has been introduced

I I A genetically engineered plant, plant cell or plant part, said genetic engineering comprising genetically altering the plant, plant cell or plant part such that the level of expression of the fwa gene is altered in comparison to that of the non genetically engineered starting material under equivalent conditions Plant, plant part or plant cell other than Arabidopsis thaliana comprising a sequence according to any of claims 1-8 or a vector according to claim 9 Plant, plant part or plant cell comprising an antisense sequence of a nucleic acid sequence according to any of claims 1-8 or a vector according to claim 9 Use of a nucleic acid sequence according to any of the claims 1-8 or a vector according to claim 9 for genetically altering the flowering time characteristics of a plant by introducing in sense direction any of the sequences or vector into the plant, plant part or a plant cell suitable for cultivating into a plant in sense direction Use of a nucleic acid sequence according to any of the claims 1-8 or a vector according to claim 9 for genetically altering the flowering time characteristics of a plant by introducing the nucleic acid sequence into a plant, plant part or a plant cell suitable for cultivating a plant therefrom into a plant in anti-sense direction A method of amending the flowering time characteristics of a plant said method comprising genetically altering the expression level of the fwa gene by introduction of a nucleic acid sequence according to any of the claims 1-8 or a vector according to claim 9 into a plant, plant cell or plant part, said cell or part being suitable for cultivation to a plant A method of lengthening the flowering time of a plant by reducing the level of expression of FWA by genetic manipulation of the fwa gene i e genetic manipulation of the encoding sequence or the expression regulating sequence of fwa A method of lengthening the flowering time of a plant by reducing the level of expression of FWA by introducing sense or antisense copies of nucleic acid sequences capable of exhibiting co suppression or gene silencing of the FWA gene in cells of the plant in question, i e sequences according to any of claims 1-8 or a vector according to claim 9 A method of selectively shortening or lengthening the flowering time of a plant by introducing an inducible expression regulating sequence other than the expression regulating sequence normally associated with fwa via genetic engineering

20. A method according to claim 17, wherein the inducible expression regulating sequence is not a photoperiodically induced expression regulating sequence nor an expression regulating sequence inducible by vernalisation.

21. A method according to claim 17 or 18 wherein the expression regulating sequence is inducible by a signal regulated by human intervention.

22. Use of a nucleic acid sequence according to any of claims 1-8 for detecting and optionally isolating fwa genes.

23. A method of producing flowering of a recombinant plant at a time different to that of the corresponding non-recombinant plant under the same environmental conditions, said method comprising cultivation of a plant, plant part or plant cell according to any of claims 10-14.

24. A method acording to claim 23, comprising applying the inducible signal of the inducible promoter regulating the fwa gene of any of the plants, plant cells or plant parts according to any of claims 10-14.