Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

TWIN SISTER OF FT (TSF) Acts as a Floral Pathway Integrator Redundantly with FT

TWIN SISTER OF FT (TSF) Acts as a Floral Pathway Integrator Redundantly with FT Abstract In Arabidopsis, several genetic pathways controlling the floral transition (flowering) are integrated at the transcriptional regulation of FT, LFY and SOC1.TSF is the closest homolog of FT in Arabidopsis. TSF expression was induced rapidly upon activation of CONSTANS (CO). The mRNA levels of TSF and FT showed similar patterns of diurnal oscillation and response to photoperiods: an evening peak, higher levels in long day (LD) than in short day (SD) conditions, and immediate up-regulation upon day-length extension. These observations suggest that TSF is a direct regulatory target of CO. tsf mutation delayed flowering in SD conditions and enhanced the phenotype of ft in both LD and SD conditions. TSF and FT also shared similar modes of regulation by FLC, an integrator of autonomous and vernalization pathways, and other factors such as EBS and PHYB. Consistently, TSF overexpression caused a precocious flowering phenotype independent of photoperiods or CO, or FLC. These observations suggest that TSF is a new member of the floral pathway integrators and promotes flowering largely redundantly with FT but makes a distinct contribution in SD conditions. TSF and FT seem to act independently of each other and of LFY, and partially upstream of SOC1. Interestingly, the expression patterns of TSF and FT in seedlings did not overlap, although both were expressed in the phloem tissues. Our work revealed additional complexity and spatial aspects of the regulatory network at the pathway integration level. We propose that the phloem is the site where multiple regulatory pathways are integrated at the transcriptional regulation of FT and TSF. Introduction Plants monitor multiple environmental and endogenous signals to determine when to flower. Genetic and molecular analyses of flowering time mutants in Arabidopsis have led to the current model, in which four major genetic pathways regulate the transition from vegetative to reproductive phase (floral transition or flowering). The photoperiod and vernalization pathways mediate responses to environmental inputs, day lengths and low temperature, respectively. The autonomous and gibberellin pathways act independently of these external signals. These four pathways are integrated at the transcriptional regulation of the floral pathway integrators, FLOWERING LOCUS T (FT), SUPRESSOR OF OVEREXPRESSION OF CO 1 (SOC1) and LEAFY (LFY) (reviewed by Araki 2001, Simpson and Dean 2002, Boss et al. 2004). FT acts as a floral promoter mainly in the photoperiod pathway downstream of CONSTANS (CO). ft mutants show a late flowering phenotype under long-day (LD) conditions, with flowering time only slightly affected under short-day (SD) conditions (Koornneef et al. 1991). Plants overexpressing FT have a precocious flowering phenotype that is independent of photoperiods and the CO function (Kardailsky et al. 1999, Kobayashi et al. 1999). Expression of FT is directly regulated by CO protein, the activity of which is controlled by a circadian clock and light conditions (Kobayashi et al. 1999, Samach et al. 2000, Suárez-López et al. 2001, Yanovsky and Kay 2002, Valverde et al. 2004). Light also affects flowering time independently of photoperiods through the phyB function, as phyB negatively regulates the transcription of FT (Blázquez and Weigel 1999, Cerdán and Chory 2003, Halliday et al. 2003). The floral repressor FLOWERING LOCUS C (FLC), which is the convergence node of the autonomous and vernalization pathways, acts, in part, by repressing the transcription of FT (Samach et al. 2000, Hepworth et al. 2002). Other than these pathways, two chromatin-related factors, TERMINAL FLOWER 2 (TFL2) and EARLY BOLTING IN SHORT DAYS (EBS), are involved in repression of FT transcription, mainly in SD conditions (Kotake et al. 2003, Piñeiro et al. 2003, Takada and Goto 2003). FT encodes a protein similar to phosphatidylethanolamine-binding protein (PEBP), which is also known as Raf kinase inhibitor protein (RKIP) (Kardailsky et al. 1999, Kobayashi et al. 1999). The biochemical functions of the PEBP/RKIP family proteins in plants, including FT, remain to be elucidated. The Arabidopsis genome contains six genes for the PEBP/RKIP family proteins (Kobayashi et al. 1999, Mimida et al. 2001), of which two other genes are also involved or implicated in the regulation of floral transition. TERMINAL FLOWER 1 (TFL1) plays an important role in floral transition in an antagonistic manner to FT, in addition to its pivotal role in the maintenance of inflorescence meristem identity (Bradley et al. 1997, Ratcliffe et al. 1998, Ratcliffe et al. 1999). A recent work suggested that MOTHER OF FTAND TFL1 (MFT), the most distantly related member of the family, might act to promote flowering (Yoo et al. 2004). The regulation of the floral transition by the PEBP/RKIP family is conserved in other plants, such as an SD plant rice and an LD plant Pisum. A quantitative trait locus for the photoperiod sensitivity in rice, named Heading-date 3a (Hd3a), is an FT ortholog and plays an important role in the promotion of flowering in SD conditions (Kojima et al. 2002). Several Hd3a homologs, such as RFT1/FT-L 3 and FTL/FT-L 1, are also implicated in the regulation of flowering (Izawa et al. 2002, Kojima et al. 2002). Like TFL1 in Arabidopsis, the activity of LATE FLOWERING (LF), a TFL1 homolog, in Pisum is important to prevent precocious floral transition (Foucher et al. 2003). TSF (TWIN SISTER OF FT) is the closest homolog of FT, with 82% identity in the deduced amino acid sequence (90% similarity, if amino acid residues with similar biochemical properties are considered). Transgenic plants overexpressing TSF show a precocious flowering phenotype, as do those overexpressing FT (Kobayashi et al. 1999). These observations suggest that FT and TSF may share a similar role in promoting the floral transition. In this study, we investigated the role of TSF in the regulation of floral transition. Our work suggests that TSF is a new member of the floral pathway integrators and promotes flowering redundantly with FT. TSF and FT share similar modes of regulation by major genetic pathways and other factors. They have a similar tissue preference, with expression in the phloem, but have distinct spatial patterns: TSF mainly in hypocotyl and FT in cotyledons and leaves. These observations suggest redundant roles with subtle functional differentiation between the two closely related genes. Among the floral pathway integrators, TSF and FT seem to act independently of each other and of LFY, and partially upstream of SOC1. Phloem seems to represent the site where multiple regulatory pathways are integrated at the transcriptional regulation of FT and TSF. Our work, together with recent reports by others, reveals additional complexity and intricacy of the regulatory network of flowering at the pathway integration level, which may contribute to fine control of the floral transition. Results Effect of tsf mutation on flowering time To investigate whether inactivation of TSF affects the timing of phase transition, we isolated a loss-of-function allele of TSF. We found a line that carries a T-DNA insertion in the second intron of TSF (SALK_087522) (Fig. 1A). No large deletion was found associated with the T-DNA insertion (data not shown). Since severe reduction of TSF transcript levels in homozygous plants was confirmed by reverse transcription–PCR (RT–PCR) (Fig. 1B), we named the insertion allele tsf-1. In addition to the tsf-1 allele, we found that ecotype Wassilewskija (Ws) carries a TSF allele that gives rise to greatly reduced levels of mRNA of the correct size (Fig. 1C). By sequence comparison between Columbia-0 (Col) and Ws, an A to T substitution was found at the splicing acceptor site of the second intron in Ws. It is likely that this substitution results in mis-splicing and is the cause of decreased mRNA levels. Details of natural variation of the TSF locus are described in the Supplementary material (Supplementary text, S-Fig. 1 and S-Table 1). RT–PCR analysis of transcripts in Ws and comparison of flowering time between Landsberg erecta (Ler) and Ws are described in S-Fig. 2 and S-Table 2, respectively. In LD (16 h light/8 h dark) conditions, tsf-1 did not show an obvious difference in flowering time from wild-type plants (Fig. 1D, Table 1). In contrast, tsf-1 flowered later than wild type in SD (8 h light/16 h dark) conditions (Table 1). A similar difference was observed between Ler wild type and tsf-1 introgressed into Ler by three crosses (data not shown). We generated a tsf-1; ft-1 double mutant. tsf mutation enhanced the late flowering phenotype of ft in both LD and SD conditions (Fig. 1D, Table 1). These results suggest that TSF plays a role as a floral promoter in the photoperiod pathway redundantly with FT, but also makes a distinct contribution to flowering in SD conditions. Photoperiodic regulation of TSF transcription The phenotype of tsf and tsf; ft mutants suggests that TSF promotes flowering redundantly with FT in LD conditions. FT acts downstream of the photoperiod pathway through transcriptional regulation by CO, which encodes a nuclear protein with two B-boxes and a CCT domain (Putterill et al. 1995, Samach et al. 2000, Robson et al. 2001). We examined whether transcription of TSF is regulated by CO protein in a photoperiod-dependent manner. Similarly to the case of FT, mRNA levels of TSF were reduced in co-2 (Fig. 2A). To demonstrate further that transcription of TSF is under the control of CO, we used an inducible system in which CO fused to a rat glucocorticoid receptor (GR) is expressed by the cauliflower mosaic virus 35S RNA (35S) promoter (35S::CO:GR) (Simon et al. 1996). Use of this system has demonstrated that FT is an early direct target of CO without requiring intermediary protein synthesis (Samach et al. 2000). TSF expression was induced within 1 h after application of dexamethasone (Dex) in the presence of cycloheximide (Cyc), as was FT expression (Fig. 3A). In contrast, mRNA levels of a late downstream gene APETALA1 (AP1) were not affected. Similarly, LFY and TFL1 expression was not induced by Dex treatment (S-Fig. 3). These results strongly suggest that CO regulates TSF as a direct target. Interestingly, ubiquitous activation of CO by Dex treatment of 35S::CO:GR plants did not result in ubiquitous induction of β-glucuronidase (GUS) under the control of the FT promoter (pFT::GUS, Takada and Goto 2003) or genomic sequences of TSF (gTSF::GUS) (see below and S-Fig. 4B, C). FT expression is barely detectable in SD conditions, but in LD conditions FT mRNA accumulates during the illuminated part of the day with a peak at dusk (Suárez-López et al. 2001, Yanovsky and Kay 2002). TSF mRNA levels showed diurnal oscillation similar to that of FT with a clear peak at dusk in LD conditions (Fig. 3B). In SD conditions, similar diurnal rhythms with a small peak at dusk were also observed, although levels were much lower as compared with LD (Fig. 3B). Transcription of CO is regulated by a circadian clock (Suárez-López et al. 2001, Yanovsky and Kay 2002) and light through the action of FLAVIN-BINDING, KELCH REPEAT, F-BOX 1 (FKF1) (Imaizumi et al. 2003). In addition to the regulation at the transcription level, it was proposed that light regulates CO protein stability to generate the daily rhythm of protein abundance and resulting in the CO activity being restricted to the evening in LD conditions. This, in turn, is responsible for the diurnal oscillation of FT (Imaizumi et al. 2003, Valverde et al. 2004). We performed day-length shift experiments to examine the effect of day-length change on regulation of TSF expression. Plants were entrained in SD conditions for 6 d after imbibition, and day-length conditions were changed from SD to LD by extending the light period on day 7. As expected, expression of TSF was induced immediately upon day-length extension, as was that of FT (Fig. 3C). Light in the subjective night of SD-entrained plants [Zeitgeber time (ZT) 8 to ZT16] on day 7 affected the oscillation pattern of CO mRNA levels (Fig. 3C). As previously described (Imaizumi et al. 2003), CO expression in SD-to-LD-shifted plants showed a biphasic pattern similar to that of the LD-entrained plants, with an earlier daytime peak due to the action of FKF1 (ZT10 vs. ZT13). The shifted daytime peak of CO mRNA accumulation is likely to be responsible for the earlier peak of TSF mRNA levels on day 7 than on day 8. These results strongly suggest that, as in the case of FT, the diurnal oscillation of TSF expression is controlled by CO activity regulated by a circadian clock and light conditions. Photoperiod- and CO-independent precocious flowering phenotype by constitutive overexpression of TSF We previously reported that plants overexpressing TSF (35S::TSF) showed a precocious flowering phenotype in LD conditions (Kobayashi et al. 1999). This was confirmed in representative homozygous lines (Fig. 4A). If TSF promotes flowering under the regulation of CO, it is expected that the early flowering phenotype of the 35S::TSF plants is independent of photoperiods and CO function. As expected, 35S::TSF plants flowered with the same number of leaves in LD and SD conditions (Table 2). Similarly, co-1 mutation did not affect the precocious flowering phenotype of 35S::TSF plants (Table 3). It has been reported recently that the expression levels of SOC1 were elevated in seedlings of FT and TSF activation-tag lines (Michaels et al. 2005) and 35S::FT seedlings (Moon et al. 2005). In good agreement with these reports, we observed that the amount of SOC1 mRNA was increased in 35S::TSF and 35S::FT seedlings as early as on day 4 (Fig. 4G). In addition, mRNA levels of another floral pathway integrator LFY were increased on day 4 in 35S::TSF and 35S::FT. A MADS-box gene AGL24 has been assigned an intermediate position between SOC1 and LFY, since AGL24 expression is reduced in soc1 and LFY expression is reduced in agl24 (Yu et al. 2002). However, no change in the expression level of AGL24 was observed in 35S::TSF and 35S::FT (Fig. 4G). No difference of AP1 expression level was observed until day 7. Consistent with the increase in 35S::FT and 35S::TSF, a slight decrease in SOC1 mRNA levels was observed in tsf and ft seedlings (Fig. 4G). TSF and FT did not affect each other’s expression (Fig. 4G, 5Za–Zf). It has been reported that the precocious flowering phenotype of 35S::FT was dramatically enhanced by 35S::LFY so that seedlings formed a single terminal flower immediately after germination (Kardailsky et al. 1999, Kobayashi et al. 1999, Moon et al. 2005). Similarly, the phenotype of 35S::TSF was greatly enhanced by 35S::LFY, although enhancement was not as strong as in the case of 35S::FT. After forming two severely up-curled rosette leaves, the primary shoot terminated with a single flower (Fig. 4D–F). In contrast, no enhancement of phenotype was observed in 35S::TSF; 35S::FT plants as compared with either parental plant (data not shown). Multiple environmental and endogenous factors that regulate TSF expression In Arabidopsis, floral transition is regulated by multiple genetic pathways such as autonomous, vernalization and gibberellin pathways, as well as the photoperiod pathway. FT is regulated at the transcriptional level by the inputs from these pathways and acts as one of the pathway integrators. As shown above, TSF is likely to act as a floral promoter in the photo-period pathway downstream of CO in a similar manner to FT. We are interested in whether TSF is also regulated by inputs from multiple pathways. We first examined TSF expression in flowering time mutants, which have been reported to affect FT mRNA levels (Fig. 2). In the following experiments, samples were collected at ZT14 in LD, and ZT7 or 8 in SD conditions, when TSF levels are highest in the respective day-length conditions. The vernalization and autonomous pathways promote the floral transition by reducing the expression levels of the FLC, which encodes a MADS-box transcription factor and acts as a repressor of flowering (Michaels and Amasino 1999, Sheldon et al. 1999, Sheldon et al. 2000, Michaels and Amasino 2001). FLC delays floral transition through repression of FT and SOC1 expression (Lee et al. 2000, Samach et al. 2000, Hepworth et al. 2002). A defect in the autonomous pathway results in the increased expression of FLC, which is counteracted by vernalization treatment (Michaels and Amasino 1999, Sheldon et al. 1999). Functional alleles at the FRIGIDA (FRI) locus promote the accumulation of FLC mRNA and confer vernalization requirement in the winter-annual accessions (Michaels and Amasino 1999, Sheldon et al. 1999, Johanson et al. 2000). In fca-1 defective in the autonomous pathway, TSF and FT mRNA levels were reduced (Fig. 2A). Similarly, Col plants with a functional FRI allele from ecotype San Feliu-2 (FRI-Sf2) [hereafter referred to as FRI-Sf2 (Col)] had reduced mRNA levels of TSF and FT (Fig. 2A). As described below, vernalization treatment restored TSF (or gTSF::GUS) expression in fca-1 and FRI-Sf2 (Col) (Fig. 6). These results suggest that TSF expression is negatively regulated by FLC, as is FT expression. In agreement with these results, 35S::TSF was able to suppress the extreme late flowering phenotype conferred by the active FRI-FLC system (Table 4). In Arabidopsis, phyB has a role in the shade-avoidance response which includes accelerated flowering. It has been reported that FT mRNA levels are higher in early flowering phyB mutants than in wild type in LD conditions (Cerdán and Chory 2003). We tested whether TSF mRNA levels are also affected in phyB mutants. A clear increase in mRNA levels of TSF, as well as FT, was observed in 5-day-old phyB-9 mutants in LD conditions (Fig. 2B). Similar differences in TSF and FT mRNA levels were also observed between Ler and phyB-5 (data not shown). Arabidopsis mutants that exhibit the early flowering phenotype have revealed the existence of genes involved in the repression of flowering. Some of these repressors act independently of environmental factors to prevent plants from precocious transition to flowering during early developmental stages or in inappropriate conditions. In non-inductive SD conditions, it has been reported that FT expression is repressed by EBS which encodes a nuclear protein with a bromoadjacent homology domain and a plant homeodomain Zn finger (Piñeiro et al. 2003). In the early flowering ebs-1 mutant, expression levels of TSF were increased as well as those of FT in SD conditions (Fig. 2C). Another floral repressor, TFL2, which encodes a protein with homology to heterochromatin protein 1 (HP1), represses FT expression (Kotake et al. 2003, Takada and Goto 2003). We examined the effect of tfl2 mutation on TSF mRNA levels by RT–PCR. Although FT mRNA levels were increased in the tfl2 mutant in SD conditions as previously reported, TSF expression was not affected (Fig. 2D). Spatial patterns of TSF expression In LD-grown seedlings, TSF was expressed mainly in hypocotyls (Fig. 5Y). In contrast, FT expression was detected mainly in cotyledons (Fig. 5Y). In mature plants, high levels of TSF expression were observed in flowers and developing siliques (data not shown) as in the case of FT (Kobayashi et al. 1999). In situ RNA hybridization has been tried, but did not succeed. Similar trials to detect FT expression in wild-type seedlings have not been successful (data not shown; see Takada and Goto 2003, An et al. 2004). To examine spatial patterns of expression in detail, we generated transgenic plants expressing GUS under the control of regulatory sequences of a TSF genomic fragment (gTSF::GUS). Since extensive natural variations in the TSF locus and the adjacent intergenic region were found, some of which involve insertion of a retroelement in the 3′-untranslated region (3′-UTR), we selected the simplest Ler sequence as the source for the reporter construct (see Materials and Methods and S-Fig.1). Consistent with the RT–PCR analysis described above (Fig. 3B), expression of gTSF::GUS was at higher levels and over a broader range in LD than in SD conditions. In 6-day-old seedlings in LD conditions, expression of gTSF::GUS was observed in the vascular tissues of hypocotyl and petiole and the basal part of cotyledons (Fig. 5A, I). A cross-section of hypocotyls of an LD-grown seedling at day 10 showed staining in the phloem parenchyma (Fig. 5L, M). Signals were also detected near the shoot apical meristem (SAM) (Fig. 5Q–S). Close examination showed that the expression is confined to a small number of cells between the SAM and young leaf primordia (Fig. 5V–X). gTSF::GUS expression was barely observable in 6-day-old plants in SD conditions (Fig. 5C, O). It was reported previously that pFT::GUS is expressed in phloem tissues within the apical part of cotyledons and leaves (Takada and Goto 2003) and we confirmed this (Fig. 5B, D, F, H). Therefore, the expression of the two genes does not seem to overlap in young seedlings (compare Fig. 5A, C, E, G, Za for gTSF::GUS and Fig. 5B, D, F, H, Zd for pFT::GUS). gTSF::GUS expression in hypocotyls and near SAM gradually increased with growth in LD conditions (Fig. 5I–K, N). Expression in true leaves was first detected in phloem at the apical-most region from day 8 (see the insets in Fig. 5E and G). In SD conditions, the signals in hypocotyls and leaves were low or almost undetectable (Fig. 5O, P, T). However, in 28-day-old plants grown under SD conditions, clear signals were detected near SAM (Fig. 5U). In mature plants, gTSF::GUS was expressed in the vascular tissues of inflorescence stems, pedicels, floral organs and roots (data not shown). These expression patterns of TSF observed in later developmental stages were, at least in part, similar to those of FT (data not shown). We made use of the gTSF::GUS reporter to examine the effect of induction of CO activity on spatial patterns of TSF expression. Although transient ubiquitous activation of CO is expected with Dex treatment of 35S::CO:GR plants, neither gTSF:GUS nor pFT:GUS were induced ubiquitously (S-Fig. 4). These results suggest that there may be additional factor(s) other than CO that restrict spatial patterns of TSF and FT expression upon activation of CO. We also analyzed the response of TSF expression to vernalization using the gTSF::GUS reporter. In F1 plants with FRI-Sf2 (Col) (gTSF::GUS/-; FRI-Sf2/fri-Col), gTSF::GUS expression was greatly reduced compared with plants without functional FRI (gTSF::GUS; fri-Col/fri-Col) (compare Fig. 6A, D and Fig. 6B, E). This was consistent with the results obtained by RT–PCR analysis of TSF expression (Fig. 2A). After 4 weeks vernalization treatment, gTSF::GUS expression in an FRI-active background was released from repression by FLC [compare Fig. 6B, E (without vernalization) and Fig. 6C, F (with vernalization)]. Again, these observations were consistent with the RT–PCR analysis of the endogenous gene expression (Fig. 6J). Expression of pFT::GUS showed essentially the same responses (Fig. 6G–J). Genetic interactions with tfl1 and fwa mutations We investigated whether TSF and FT behave in a similar manner in genetic interaction with tfl1 and fwa mutations. FT and TFL1 have opposite roles in regulation of flowering despite their high degree (55%) of sequence identity (Kardailsky et al. 1999, Kobayashi et al. 1999). Loss of TFL1 results in an early flowering phenotype, in addition to conversion of inflorescences to flowers (Shannon and Meeks-Wagner 1991). ft and tfl1 are additive so that ft; tfl1 plants flower slightly earlier than ft, and inflorescences produce terminal flowers (Ruiz-García et al. 1997). Similarly, tsf and tfl1 were additive in phenotype. The tsf-1; tfl1-17 double mutant flowered as early as tfl1-17 and produced terminal flowers (Table 1 and data not shown). As in the case of FT overexpression, the tfl1-17 enhanced the early flowering and inflorescence to flower conversion phenotype of 35S::TSF. 35S::TSF; tfl1 flowered earlier than 35S::TSF, and axillary shoots in 35S::TSF; tfl1 were replaced with solitary flowers (Fig. 4B, C). In semi-dominant fwa ‘epimutants’, a homeobox gene FWA is ectopically expressed in seedlings due to hypomethylation of the direct repeat sequences in the promoter (Soppe et al. 2000). It was reported previously that the precocious flowering phenotype of 35S::FT was strongly suppressed by fwa (Kardailsky et al. 1999, Kobayashi et al. 1999). It has been suggested that ectopically expressed FWA interferes with the FT function by an unknown mechanism (Kardailsky et al. 1999, Kobayashi et al. 1999). In contrast to 35S::FT, the phenotype of 35S::TSF was little affected by fwa. The flowering time of F1 plants derived from a cross between fwa-101D and 35S::TSF was as early as that of 35S::TSF hemizygotes (Table 4). As in the case of FT, TSF mRNA levels were not affected by fwa-2 mutation (Fig. 2A). Discussion Genetic and molecular studies in Arabidopsis have led to a model in which the photoperiod, vernalization, autonomous and gibberellin pathways regulate the floral transition through the transcriptional regulation of the floral pathway integrators, FT, SOC1 and LFY (reviewed by Araki 2001, Simpson and Dean 2002, Boss et al. 2004). It has been shown that co; fca; ga1 triple mutant plants impaired in the photoperiod, autonomous and gibberellin pathways, without vernalization, do not flower even under inductive LD conditions. This indicates that the three pathways are essential for flowering to occur and that the vernalization pathway is absolutely required in the absence of the three major pathways (Reeves and Coupland 2001). In contrast, a recent work demonstrated that the floral transition is not completely blocked in an ft; soc1; lfy triple mutant lacking all of the three known pathway integrators (Moon et al. 2005). These facts suggest that there are as yet unidentified gene(s) acting at the pathway integration level to promote the floral transition. Among the three pathway integrators, FT is a member of a small gene family with TSF as the closest homolog (Kardailsky et al. 1999, Kobayashi et al. 1999, Mimida et al. 2001), and transgenic analysis has suggested a potential role for TSF in the promotion of flowering (Kobayashi et al. 1999). In rice, the FT ortholog Hd3a is involved in photoperiodic regulation of flowering, acting downstream of the CO ortholog Hd1 (Kojima et al. 2002, Hayama et al. 2003). Some of the Hd3a homologs have also been implicated in the regulation of flowering based on expression profiles and overexpression phenotypes (Izawa et al. 2002, Kojima et al. 2002). However, the actual roles of TSF and Hd3a homologs in the regulation of flowering remain to be demonstrated. In the present work, we investigated the role of TSF in floral transition. TSF promotes flowering as a floral pathway integrator Phenotypes of mutant and transgenic plants (Fig. 1D, 4A, Tables 1–3) suggest that TSF promotes flowering redundantly with FT in the photoperiod pathway. Functional FT in LD conditions can compensate for the loss of TSF. Since FT mRNA levels did not change in tsf-1 (Fig. 1B, see also Fig. 4G), compensation is not due to up-regulation of FT in response to the loss of TSF. On the other hand, reduced levels of FT expression in SD conditions cannot fully substitute for TSF. There seems to be a discrepancy between our observation that the tsf single mutant flowered later than the wild type in SD conditions and a recent report by Michaels et al. (2005), who observed no difference using the same T-DNA insertion allele. Because the difference was fully reproducible in our conditions (similar results were obtained twice in the original Col background and once in the Ler background after introgression), we think the discrepancy was due to the difference in growth conditions between the two groups. Interestingly, ecotype Ws with a natural variation causing severe reduction in TSF mRNA levels (see Supplementary material) flowered as early as Ler in LD conditions, but much later than Ler in SD conditions (S-Table 2). It will be interesting to explore the association between the natural variation at the TSF locus and the flowering time phenotype in SD conditions. Severe reduction of mRNA levels in a co mutant (Fig. 2A), rapid induction of mRNA accumulation by treatment of 35S::CO:GR seedlings with Dex in the presence of Cyc (Fig. 3A), very low mRNA amounts in SD conditions (Fig. 3B) and an immediate increase in mRNA levels in SD-entrained seedlings upon day-length extension (Fig. 3C), taken together, strongly suggest that transcription of TSF is directly regulated by CO, which is under the control of a circadian clock and light at the transcription and protein activity levels. In agreement with this, gTSF::GUS expression was higher in LD than in SD conditions (Fig. 5) and gTSF::GUS expression in 35S::CO:GR seedlings was induced by treatment with Dex (S-Fig. 4C). As expected, constitutive expression of TSF caused a precocious flowering phenotype independently of photoperiods and CO function (Tables 2, 3). TSF and FT, therefore, share direct regulation by CO in the photoperiod pathway. In a previous attempt to identify CO targets, TSF was not identified (Samach et al. 2000). This may be due to the fact that TSF has much lower mRNA levels than FT even after induction by CO (data not shown). Recently, several sequence motifs responsible for regulation by CO have been identified by analysis of the FT promoter (H. Nakagawa and T. Izawa, personal communication). Interestingly, sequences similar to at least some of these motifs are also found in the TSF promoter (T. Izawa, personal communication). Since the sequence upstream of the transcribed region is much shorter in TSF (1.5 kb) than in FT (7.5 kb) (see Kaya et al. 2000 for the FT region on chromosome 1 and S-Fig.1 for the TSF region on chromosome 4), use of the TSF promoter fragment will facilitate further analysis. Repression of TSF expression in the presence of high levels of FLC expression (Fig. 2A, 6B, E, J) and de-repression of TSF by vernalization through reduction of FLC levels (Fig. 6C, F, J) suggest that TSF is a regulatory target of the floral repressor FLC. Consistently, constitutive expression of TSF suppressed the extreme late flowering phenotype conferred by the active FRI-FLC system (Table 4), which is in good agreement with the recent identification of an activation-tagged TSF allele as a suppressor of FRI-Sf2 (Col) (Michaels et al. 2005). In addition, sequences similar to the CArG box motif (MADS-domain protein-binding site) were found near the coding region of TSF (data not shown). The presence of additional regulatory targets of FLC and vernalization other than FT and SOC1 has recently been suggested, since the flowering time of the ft; soc1 double mutant was accelerated by flc-3 and further delayed by an active FRI-Sf2 allele, and the ft; soc1; lfy triple mutant still responds to vernalization (Moon et al. 2005). Our results indicate that TSF represents the suggested target. It will be interesting to examine the phenotypes of ft; tsf; soc1 and ft; tsf; soc1; lfy mutants in terms of their interaction with the FRI-FLC system and their response to vernalization. Although TSF, with FT, is clearly placed at the convergent point of the photoperiod and autonomous/vernalization pathways, the relationship of TSF with the gibberellin pathway remains unclear. EBS, which encodes a chromatin-related protein, has been implicated in mediating signals in the gibberellin pathway (Gómez-Mena et al. 2001, Piñeiro et al. 2003). Expression of TSF, as well as FT, in SD conditions was de-repressed in an ebs mutant (Fig. 2C). Based on the effect of ebs on FT expression, Piñeiro et al. (2003) discussed a possible contribution of the gibberellin pathway to the regulation of FT. The same argument may be applicable to TSF as well. However, since there is no direct experimental evidence for the involvement of the gibberellin pathway in the regulation of FT (Moon et al. 2003) and TSF, further investigation will be required before a firm conclusion can be reached. Besides the four major genetic pathways, other pathway(s) or factor(s) contribute to the regulation of flowering. Light quality, through the action of phyB, is one such environmental factor (Blázquez and Weigel 1999). phyB seems to negatively regulate TSF and FT expression, since mRNA levels of TSF and FT were elevated in phyB mutants (Fig. 2B and see Cerdán and Chory 2003, Halliday et al. 2003). It has been suggested recently that phyB acts through PFT1, which encodes a putative nuclear protein, to regulate FT at least partly in a CO-independent manner (Cerdán and Chory 2003). It is likely that PFT1 also mediates regulation of TSF by phyB. Ambient temperature is another environmental factor that affects the timing of flowering (Blázquez et al. 2003, Halliday et al. 2003). It has been reported that TSF expression is influenced by ambient temperatures, with mRNA levels higher at 16°C than at 23°C, and that FT has an opposite response (Blázquez et al. 2003). The relationship of TSF to the ‘thermosensory pathway’ is an interesting problem to be investigated further. TFL2, an Arabidopsis counterpart of HP1 (Gaudin et al. 2001, Kotake et al. 2003), is another candidate for a regulator of TSF expression. However, in sharp contrast to FT, expression of TSF was not affected by the tfl2 mutation (Fig. 2D). It is likely that TFL2 regulates genes in euchromatin regions through localized modification of chromatin structure. It is to be noted that while the FT locus is rather isolated from the neighboring genes (see Fig. 7A of Kaya et al. 2000 for gene distribution in the FT region), TSF is a typical Arabidopsis gene with short intergenic sequences (S-Fig. 1). This might be reflected in the difference in TFL2-mediated regulation. Interestingly, ABA and ABSCISIC ACID-INSENSITIVE 3 (ABI3) recently have been implicated in the regulation of TSF based on the observation that TSF was repressed in plants overexpressing the maize ABI3 ortholog, VIVIPAROUS1 (VP1), when treated with ABA (Suzuki et al. 2003). It has been known that abi3 mutants flower earlier than the wild type (Kurup et al. 2000). Therefore, possible involvement of ABA in the regulation of TSF is an interesting problem to be investigated. In conclusion, TSF is a new pathway integrator regulated by various pathways and factors largely in a similar manner to FT (Fig. 7). The regulatory relationship of TSF (and FT) with the gibberellin pathway and signals from ambient temperatures and ABA (through ABI3) remain interesting issues requiring further investigation. Recent studies have revealed regulatory interactions among the pathway integrators. For example, it has been reported that mRNA levels of SOC1 were decreased in the ft mutant (Lee et al. 2000, Schmid et al. 2003) and increased in FT and TSF activation-tag lines and 35S::FT plants (Michaels et al. 2005, Moon et al. 2005). Our observation that the amount of SOC1 mRNA was increased in 35S::TSF and 35S::FT plants and slightly decreased in tsf and ft mutants (Fig. 4G) confirmed these recent reports. These findings strongly suggest that SOC1 is, in part, positively regulated by FT and TSF (Fig. 7). Interestingly, it was demonstrated that activation-tagged alleles of TSF and FT can overcome the repressive effect of the active FRI-FLC system on SOC1 (Michaels et al. 2005). This is in contrast to the case of 35S::CO which cannot overcome the repression of SOC1 by 35S::FLC (Hepworth et al. 2002). FT and TSF seem to regulate SOC1 independently of the upstream regulators FLC and CO. It has been shown that overexpression of SOC1 has little effect on FT and TSF mRNA levels, indicating the absence of reciprocal regulation of FT and TSF by SOC1 (Michaels et al. 2005, Moon et al. 2005). Another pathway integrator, LFY, has the least inter-dependence with FT (Kardailsky et al. 1999, Kobayashi et al. 1999, Moon et al. 2005). Similarly, LFY and TSF are likely to act rather independently (Fig. 7). The lack of enhancement of the phenotype in 35S::FT; 35S::TSF indicates that FT and TSF act through common downstream factors. It is unlikely that FT and TSF regulate each other at the transcription level (see discussion below). Spatial pattern of TSF and FT expression and functional differentiation Although our knowledge of genetic interactions among regulators of flowering has improved significantly, how these regulators and pathways are spatially organized is still poorly understood. The spatial patterns of expression of several important regulators such as CO and FT have only recently been described in detail (Takada and Goto 2003; An et al. 2004). We analyzed and compared spatial patterns of expression of TSF and FT by using reporter constructs (gTSF::GUS and pFT::GUS) (Fig. 5). gTSF::GUS expression was observed mainly in the vascular (phloem) tissues of hypocotyl and petiole and the basal part of cotyledons (Fig. 5). As previously reported (Takada and Goto 2003), pFT::GUS expression was observed mainly in vascular (phloem) tissues of cotyledons and leaves, where it was stronger in the apical region than in the basal region and petiole (Fig. 5). Therefore, TSF and FT share similar tissue specificity of expression in phloem, but have largely non-overlapping patterns of expression in young seedlings. The difference in spatial patterns of TSF and FT is probably due, in part, to the difference in negative regulation by TFL2. Expression of TFL2 as demonstrated by gTFL2::GUS is stronger in the basal region of the leaves (Takada and Goto 2003), and TFL2 seems to be responsible for down-regulation of FT, but not TSF, in the basal region of cotyledons and leaves. A mutual regulation model of TSF and FT expression recently proposed by Michaels et al. (2005) might also explain the difference in the spatial patterns. However, we could not reproduce their observation that TSF mRNA levels are reduced in 35S::FT and FT mRNA levels in 35S::TSF (Fig. 4G). Analysis with GUS reporters also failed to provide evidence for mutual, either positive or negative, regulation (Fig. 5Za–Zf). Furthermore, we observed effects of neither ft mutation on TSF mRNA levels nor tsf on FT mRNA levels (Fig. 4G). Therefore, we do not think it likely that mutual repression between TSF and FT is the cause of the non-overlapping patterns. Another important difference in spatial patterns of expression of TSF and FT was the presence vs. the absence of expression in the shoot apical region. Whereas gTSF::GUS expression was consistently observed in a small number of cells near the shoot apex, pFT::GUS expression could not be detected associated with the shoot apex in similar experimental conditions for growth and GUS staining (Fig. 5 and data not shown, see also Takada and Goto 2003). Although the possibility that this might be due to the difference in the design of reporter constructs cannot be ruled out, we think that it reflects an actual difference in the patterns of endogenous gene expression. Observation of whole-mount preparations as well as cross- and longitudinal sections of gTSF::GUS plants showed that GUS staining is not in the shoot apical meristem per se, but surrounds the meristem and is sometimes associated with the base of young leaf primordia. Expression of gTSF::GUS in the shoot apical region seems to be regulated by photoperiods (possibly via CO) and the FRI-FLC system. To our knowledge, there are no reports of genes with similar patterns of expression in the shoot apical region. On the other hand, similar vascular expression in hypocotyl was reported for Arabidopsis thaliana CENTRORADIALIS homologue (ATC), which also belongs to the PEBP/RKIP gene family (Mimida et al. 2001). These differences in spatial patterns of expression are likely to be responsible for the fact that the loss of FT function is not compensated for by functional TSF in LD conditions (Fig. 1D, Table 1). Since the late flowering phenotype of the ft mutation was fully suppressed by 35S::TSF, resulting in 35S::TSF; ft-1 plants flowering as early as 35S::TSF plants (Table 3), TSF can substitute for FT activity when it is expressed in the proper tissues and/or in amounts above a certain threshold level. Although it is difficult to compare the mRNA levels of the two genes quantitatively, we attempted to gain some clues by RT–PCR with a common set of primers for TSF and FT followed by digestion with restriction enzymes to distinguish the two gene products. The amount of TSF mRNA was transiently higher than that of FT immediately after germination, but soon afterwards it became much lower than that of FT (data not shown), an observation consistent with our previous results (Kobayashi et al. 1999). Therefore, lower levels of TSF expression may also be responsible for the lack of compensation for ft. On the other hand, most observations including a similar genetic interaction with TFL1, which encodes another protein of the PEBP/RKIP family and plays an opposite role in flowering, suggest that FT and TSF proteins have equivalent functions (Fig. 4B, C and Table 1; Kobayashi et al. 1999, Kardailsky et al. 1999). However, since 35S::FT and 35S::TSF had different interactions with fwa, there might be subtle functional differentiation at the protein level as well (Table 4). Experiments to express TSF under the control of the FT regulatory sequence or FT under the control of the TSF regulatory sequence in ft mutants should provide important clues. The fact that the tsf single mutant showed a late flowering phenotype in SD conditions (actually tsf-1 was later than ft-1; see Table 1) suggests that TSF makes a greater relative contribution to the promotion of flowering in SD than in LD conditions. So far, the reason for this is not fully clear, and detailed analysis of TSF and FT expression in plants prior to the floral transition in SD conditions will be necessary to gain some insight. A greater relative contribution of TSF may be observed in other conditions, such as lower ambient temperatures, as well (see discussion above). Through differential contributions under various environmental conditions, FT and TSF may fine-tune the timing of the floral transition and provide robustness for the process that integrates the multiple floral signals. Further investigation to identify the environmental conditions in which TSF makes a distinct contribution is necessary to explore this interesting possibility. In rice, there are nine homologs of FT, and the expression of three of them (Hd3a, RFT/FT-L 3 and FTL/FT-L 1), which belong to the same clade with FT and TSF, was analyzed in detail by RT–PCR (Izawa et al. 2002). Interestingly, certain differences in response to photoperiods and developmental regulation were observed among the three genes, which otherwise were similar in regulation, suggesting that a certain degree of differentiation in their roles, likely, in flowering (Izawa et al. 2002). Therefore, the redundant roles of FT and TSF, with subtle functional differentiation, will not be a unique situation in Arabidopsis. Genome projects in species other than Arabidopsis and rice have revealed multiple members of the FT clade in the PEBP/RKIP family (Brunner et al. 2003). It is likely that similar redundancy and differentiation of multiple FT homologs contribute to the fine regulation of flowering in other species as well. Spatial aspects of pathway integration and future perspectives The expression of TSF and FT in phloem is likely to be determined by the direct regulator CO, which is expressed mainly in phloem (Takada and Goto 2003, An et al. 2004), although additional factor(s) may also contribute to the restriction of TSF and FT expression to phloem tissue. Expression of TSF and FT in phloem of hypocotyl and cotyledons, respectively, is also under the control of FLC and vernalization (Fig. 6). Furthermore, phyB seems to exert its negative regulatory effect on FT and TSF mainly in phloem, since ectopic GUS staining was observed neither in gTSF::GUS; phyB-9 nor in pFT::GUS; phyB-9 (S-Fig. 5). By quantitative RT–PCR analysis of FT expression in mesophyll and vascular cells of wild-type and phyB seedlings, a similar conclusion has been reached for FT (Endo et al. 2005). Taken together, it is likely that phloem represents one of the tissues where signals from the photoperiod, autonomous, vernalization and light quality pathways are integrated. One interesting aspect of integration is the possibility of release from repression by FLC in presumably non-dividing cells in the vasculature of cotyledons and hypocotyl, an interesting issue to be investigated further with the aid of the GUS reporters. That the phloem represents the tissue where signals from various pathways are integrated at the transcriptional regulation of FT and TSF raises the interesting question of the possible relationship of these two genes to a hitherto unidentified leaf-generated long-distance floral stimulus acting at the shoot apex (called ‘florigen’). We have accumulated evidence for FT acting at the shoot apical meristem (Abe et al. 2005) and there is some evidence for TSF also acting at the shoot apex. We also found that the effect of 35S::FT to promote flowering is graft transmissible in Arabidopsis (M. Notaguchi, Y. Daimon, M. Abe and T. Araki, unpublished results), and similar graft transmissibility is expected for 35S::TSF. These findings support the recent hypothesis that FT is involved in the generation of, or its possible identity with, the long-distance signal(s) promoting flowering (Takada and Goto 2003, An et al. 2004). An et al. (2004) also suggested the presence of CO-dependent, but FT-independent component(s) in the long-distance signals. TSF is likely to represent at least part of the suggested components. Considering all this together, a logical step forward is to explore whether FT and TSF, with some differentiation, actually represent the long-sought ‘florigen’. Materials and Methods Plant materials and growth conditions Col and Ler were used as wild types. Ecotype Ws was also used in some experiments. A T-DNA insertion allele of TSF in the Col background (SALK_087522) (Alonso et al. 2003) was obtained from the Arabidopsis Biological Resource Center and backcrossed with wild-type Col three times. 35S::CO:GR; co-2 was formerly obtained from G. Coupland (currently at Max Planck Institute for Plant Breeding Research, Cologne, Germany) and Plant BioScience Ltd, UK. ft-1 introgressed into the Col background, referred to as ft-1 (Col) (described in Yoo et al. 2004), was obtained from J. H. Ahn (Korea University, Korea). Col carrying a functional FRI allele from ecotype San Feliu-2 (FRI-Sf2), referred to as FRI-Sf2 (Col), was obtained from R. Amasino (University of Wisconsin, Madison, WI, USA). ebs-1 and phyB-9 were obtained from J. Martínez-Zapater (Universidad Autonoma de Madrid, Spain) and A. Nagatani (Kyoto University, Japan), respectively. A strong line (#11–1) of 35S::FT was previously described (Kobayashi et al. 1999). 35S::TSF transgenic lines were previously generated (Kobayashi et al. 1999), and homozygotes from two strong lines (#1–2-B and #2–2) and a weak line (#4–1), all with a single-locus T-DNA insertion, were chosen for analysis. A 35S::LFY transgenic line (DW151.2.5L), previously obtained from D. Weigel (currently at Max Planck Institute, Germany) and used for constructing 35S::FT; 35S::LFY in the previous work (Kobayashi et al. 1999), was used for making 35S::TSF; 35S::LFY plants. tsf-1 was crossed with ft-1 (Col) twice and double mutant plants were identified in the F2 populations. A pFT::GUS line (#6–16) was described previously (Takada and Goto 2003). tfl1-17 is an RNA null allele (Kobayashi et al. 1999, Mimida et al. 2001). fwa-101D is a newly identified epiallele of FWA and will be described elsewhere. co-2, ft-2, fwa-2, fca-1 and ebs-1 are in the Ler background, and phyB-9, tfl2-1 and tfl1-17 are in the Col background. For analysis of flowering time phenotype, plants were grown on soil at 22°C under LD (16 h light/8 h dark) conditions with white fluorescent lights (∼60 µmol m–2 s–1) or SD (8 h light/16 h dark) conditions with white fluorescent lights (∼100 µmol m–2 s–1). For expression analysis, plants were grown on Murashige and Skoog (MS) medium or 1/2 MS medium supplemented with 1 or 0.5% sucrose. Seeds were stratified by keeping at 4°C for 2–4 d and then transferred to 22°C, which was defined as day 0. Dexamethasone treatment 35S::CO:GR; co-2 plants were grown for 7 d under constant light (CL) conditions on MS medium with 1% sucrose. Seedlings were submerged for 30 min in water containing 1 µM Dex (Wako, Osaka, Japan) and/or 10 µM Cyc (Nacalai, Kyoto, Japan), after which excess liquid was removed. RNA was extracted from seedlings harvested at 0.5, 1, 2 and 9 h after the start of the treatment. The effectiveness of Cyc treatment was confirmed by comparing patterns of GUS staining (see below for procedure) of pFT::GUS; 35S::CO:GR seedlings after Dex treatment without Cyc, Dex treatment with Cyc and mock treatment without Cyc (see S-Fig. 4A). Vernalization treatment Seeds were sown on MS or 1/2 MS medium supplemented with 1 or 0.5% sucrose and incubated at 4°C under CL conditions for 4 weeks. Seeds germinated during this incubation period. Seedlings were then transferred to 22°C under LD conditions. RT–PCR analysis RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and was treated with RNase-free DNase (Invitrogen) according to the manufacturer’s instructions. Total RNA (0.5 µg) was reverse-transcribed in a 20 µl reaction mixture using Superscript II (Invitrogen). After reaction, the mixture was diluted with 30 µl of water, and 1 µl aliquots were analyzed. Primers used in this study were as follows: TSF, 5′-GAGTCCAAGCAACCCTCACCAA-3′, 5′-CACAATACGATGAATTCCCGAG-3′; FT, 5′-TACGAAAATCCAAGTCCCACTG-3′, 5′-AAACTCGCGAGTGTTGAAGTTC-3′; mTUB2, 5′-CTCAAGAGGTTCTCAGCAGTA-3′, 5′-TCACCTTCTTCATCCGCAGTT-3′; CO, 5′-GACCACTCTACTCACCACCAAAG-3′, 5′-CAACCTCCTTGGCATCCTTATC-3′; AP1, 5′-GCACATCCGCACTAGAAAAAA-3′, 5′-CTTCTTGATACAGACCACCC-3′; FLC, 5′-TTAGTATCTCCGGCGACTTGAACCCAAACC-3′, 5′-AGATTCTCAACAAGCTTCAACATGAGTTCG-3′; SOC1, 5′-TGAGGCATACTAAGGATCGAG-3′, 5′-GCGTCTCTACTTCAGAACTTGGGC-3′; AGL24, 5′-TCCATCGAAGTCAACTCTGCTGGATC-3′, 5′-GTCTTCATGCAAGTAACATCAAC-3′; LFY, 5′-AAGGTTTCACGAGTGGCTTATTCCG-3′, 5′-GATGCTCCCTCTGTCTCTCTGTCC-3′. PCR products were electrophoresed on an agarose or a polyacrylamide gel, and visualized by ethidium bromide staining or Southern blotting. Southern blotting was performed with AlkPhos Direct (Amersham, Piscataway, NJ, USA) according to the manufacturer’s protocol. Probes were made from fragments amplified with the same primer sets for each gene. Blot images were quantified using the public domain NIH Image program for Fig. 3B and C. Plasmid construction and transgenic plants To construct gTSF::GUS, the GUS coding sequence from pBI 101 was inserted into the 4.8 kb TSF genomic fragment (including a 1.5 kb sequence upstream of ATG and a 1.36 kb sequence downstream of the stop codon) cloned from Ler by PCR. Because TSF in ecotype Ler lacks insertion sequences, including a 1 kb insertion in the 3′-UTR and a 5 kb copia-like retroelement (At4g20365) in the intergenic region between TSF (At4g20370) and At4g20360 (TUFA) found in Col and other ecotypes (see Supplementary material for natural variations), we have chosen Ler as a source of the TSF sequence for the gTSF::GUS construct. The stop codon of TSF was replaced with the GUS open reading frame and a linker sequence in front of ATG. This chimeric gene on the binary vector pBI 101 was introduced into Agrobacterium strain pMP90 and transformed into Col plants by the floral-dip procedure (Clough and Bent 1998). Histological analysis of GUS staining gTSF::GUS lines #21–1, #8–1 and #39–1 and pFT::GUS line #6–16 (Takada and Goto 2003) were chosen for analysis. In vernalization experiments, F1 plants between FRI-Sf2 (Col) and gTSF (#21–1 and #39–1) or pFT::GUS were compared with gTSF or pFT::GUS parental lines. Samples were collected at ZT14 for LD-grown plants or at ZT7 for SD-grown plants. For GUS staining, tissues were incubated for 15 min in fixing solution (0.3% formaldehyde, 50 mM sodium phosphate buffer, pH 7.2, 0.2% Triton X-100), rinsed with water, infiltrated with staining solution (0.5 mg ml–1 X-Gluc, 50 mM sodium phosphate buffer, pH 7.2, 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 0.2% Triton X-100) under vacuum for 5 min and incubated at 37°C for about 48 h in the dark. After staining, for whole-mount observation, samples were cleared in a mixture of chloral hydrate, glycerol and water solution (8 g : 1 ml : 2 ml). For sectioning, samples were dehydrated through an ethanol series, embedded in Technovit 7100 (Heraeus Kulzer, Wehrheim, Germany) and sectioned at a thickness of 4 µm with a microtome. Supplementary material Supplementary material mentioned in the article is available to online subscribers at the journal website www.pcp.oupjournals.org. Acknowledgments We thank J. H. Ahn, R. Amasino, G. Coupland and Plant Bioscience Ltd, J. Martínez-Zapater, A. Nagatani, D. Weigel, and the NSF-supported ABRC (Ohio, USA) for materials, T. Izawa for communicating unpublished results, J. H. Ahn for discussion, Y. Tomita for excellent technical assistance and analysis of natural variations, S. Yamamoto and Y. Ikeda for help in RNA extraction, and members of the Araki lab for comments and advice. Supported by a Grant-in-Aid for Scientific Research on Priority Areas (to T.A.), a Grant for Scientific Research (B) (to T.A.), a Grant for Biodiversity Research of the 21st Century COE (A14) from MEXT, Japan, a grant by the CREST program of the Japan Science and Technology Agency (to T.A.), and a grant by PROBRAIN (to M.A.). A.Y. was supported in part by a Grant for Biodiversity Research of the 21st Century COE (A14) from MEXT, Japan. 5 Present address: Department of Molecular Biology, Max Planck Institute for Developmental Biology, D-72076 Tübingen, Germany The nucleotide sequences reported in this paper have been submitted to DDBJ under accession numbers AB204805–AB204809. View largeDownload slide Fig. 1 Isolation of a tsf-1 allele. (A) TSF gene structure in ecotype Col and position of the T-DNA insertion (represented by a box on an inverted triangle, not in scale) in tsf-1 (SALK_087522). Boxes and lines indicate exons and introns of TSF; gray and yellow boxes represent untranslated regions and coding regions, respectively. Orange flags represent positions of the primer set used in (B) and (C). Analysis of the sequence flanking the T-DNA left border (filled box) and successful PCR amplification with primers shown with blue flags indicated that there was no large deletion associated with the T-DNA insertion. In ecotype Ws, an A to T substitution at the splicing acceptor site (ag) of the second intron was found. (B) RT–PCR analysis showing that tsf-1 had reduced mRNA levels. Upper and lower panels of TSF represent Southern blot images after 24 and 28 cycles of amplification by PCR, respectively. (C) Expression levels of TSF among three representative ecotypes. Plants were grown on MS medium in LD conditions and harvested at ZT14 on day 7 (B, C). (D) Phenotype of tsf-1 plants. From left to right: wild type (Col), tsf-1, ft-1 and tsf-1; ft-1 plants grown for 8 weeks under CL conditions at 22°C. View largeDownload slide Fig. 1 Isolation of a tsf-1 allele. (A) TSF gene structure in ecotype Col and position of the T-DNA insertion (represented by a box on an inverted triangle, not in scale) in tsf-1 (SALK_087522). Boxes and lines indicate exons and introns of TSF; gray and yellow boxes represent untranslated regions and coding regions, respectively. Orange flags represent positions of the primer set used in (B) and (C). Analysis of the sequence flanking the T-DNA left border (filled box) and successful PCR amplification with primers shown with blue flags indicated that there was no large deletion associated with the T-DNA insertion. In ecotype Ws, an A to T substitution at the splicing acceptor site (ag) of the second intron was found. (B) RT–PCR analysis showing that tsf-1 had reduced mRNA levels. Upper and lower panels of TSF represent Southern blot images after 24 and 28 cycles of amplification by PCR, respectively. (C) Expression levels of TSF among three representative ecotypes. Plants were grown on MS medium in LD conditions and harvested at ZT14 on day 7 (B, C). (D) Phenotype of tsf-1 plants. From left to right: wild type (Col), tsf-1, ft-1 and tsf-1; ft-1 plants grown for 8 weeks under CL conditions at 22°C. View largeDownload slide Fig. 2 Expression levels of TSF in various flowering-time mutants. (A) Effects of late flowering mutations in LD conditions. FRI-Sf2: FRI-Sf2 (Col). Plants were harvested at ZT14 on day 7. (B) Effect of phyB-9 mutation in LD conditions. Plants were harvested at ZT14 on day 5. (C) Effect of ebs-1 mutation in SD conditions. (D) Effect of tfl2-1 mutation in SD conditions. Plants were harvested at ZT7 (C) or ZT8 (D) on day 14. Plants were grown on MS medium except for the left panel of A (on 1/2 MS). View largeDownload slide Fig. 2 Expression levels of TSF in various flowering-time mutants. (A) Effects of late flowering mutations in LD conditions. FRI-Sf2: FRI-Sf2 (Col). Plants were harvested at ZT14 on day 7. (B) Effect of phyB-9 mutation in LD conditions. Plants were harvested at ZT14 on day 5. (C) Effect of ebs-1 mutation in SD conditions. (D) Effect of tfl2-1 mutation in SD conditions. Plants were harvested at ZT7 (C) or ZT8 (D) on day 14. Plants were grown on MS medium except for the left panel of A (on 1/2 MS). View largeDownload slide Fig. 3 Regulation of TSF expression by CO and photoperiods. (A) Up-regulation of TSF expression by activation of the CO–GR fusion protein. Numbers indicate hours after application of dexamethasone (+Dex) or the solvent (–Dex) in the presence of cycloheximide (+Cyc). (B) Diurnal oscillation of TSF mRNA levels in LD or SD conditions. Plants were harvested from ZT0 on day 7 to ZT4 on day 8 at 2 h intervals. Southern blot images and the quantified data are shown. The vertical axis shows relative mRNA levels of TSF and FT normalized by those of a β-tubulin gene (TUB2). White and black bars represent light and dark periods, respectively. (C) Response of TSF mRNA levels to SD to LD shift. Plants were entrained in SD conditions, and were shifted from SD to LD on day 7 by extending the light period. Seedlings were harvested from ZT1 on day 7 to ZT1 on day 9 at 3 h intervals. Relative mRNA levels of TSF, FT and CO were normalized by TUB2. White and black bars represent light and dark periods, respectively. The extended light period in the subjective night on day 7 is shown with a gray bar. Plants were grown on MS (A, C) or 1/2 MS (B) media at 22°C. View largeDownload slide Fig. 3 Regulation of TSF expression by CO and photoperiods. (A) Up-regulation of TSF expression by activation of the CO–GR fusion protein. Numbers indicate hours after application of dexamethasone (+Dex) or the solvent (–Dex) in the presence of cycloheximide (+Cyc). (B) Diurnal oscillation of TSF mRNA levels in LD or SD conditions. Plants were harvested from ZT0 on day 7 to ZT4 on day 8 at 2 h intervals. Southern blot images and the quantified data are shown. The vertical axis shows relative mRNA levels of TSF and FT normalized by those of a β-tubulin gene (TUB2). White and black bars represent light and dark periods, respectively. (C) Response of TSF mRNA levels to SD to LD shift. Plants were entrained in SD conditions, and were shifted from SD to LD on day 7 by extending the light period. Seedlings were harvested from ZT1 on day 7 to ZT1 on day 9 at 3 h intervals. Relative mRNA levels of TSF, FT and CO were normalized by TUB2. White and black bars represent light and dark periods, respectively. The extended light period in the subjective night on day 7 is shown with a gray bar. Plants were grown on MS (A, C) or 1/2 MS (B) media at 22°C. View largeDownload slide Fig. 4 Effects of TSF overexpression. (A) Phenotype of TSF-overexpressing plants. From left to right: wild type (Col), 35S::FT #11–1, 35S::TSF #1–2-B and 35S::TSF #2–2 plants grown for 18 d in CL conditions. (B) Phenotype of 35S::TSF; tfl1-17 plants. Left, 35S::TSF #1–2-B; right, 35S::TSF #1–2-B; tfl1-17 grown for 13 d in LD conditions. (C) Axillary shoots of the first two leaves in 35S::TSF; tfl1-17 which are converted to single flowers. (D–F) Phenotype of 35S::TSF #2–2; 35S::LFY plants grown for 14 d in LD conditions. (F) A close-up of the primary shoot of the plant in (E), which is converted into a single terminal flower. One cotyledon was removed to reveal the short pedicel (arrowhead). Plants were grown on soil (A) or on 1/2 MS medium (B–F). Scale bars: 1 cm in (A), 0.5 cm in (B), 1 mm in (C) and (E), 2 mm in (D) and 0.5 mm in (F). (G) Effects of TSF and FT overexpression on the expression of SOC1, AGL24, LFY and AP1. Plants were grown on MS medium in LD conditions and harvested at ZT14 on days 4 and 7. View largeDownload slide Fig. 4 Effects of TSF overexpression. (A) Phenotype of TSF-overexpressing plants. From left to right: wild type (Col), 35S::FT #11–1, 35S::TSF #1–2-B and 35S::TSF #2–2 plants grown for 18 d in CL conditions. (B) Phenotype of 35S::TSF; tfl1-17 plants. Left, 35S::TSF #1–2-B; right, 35S::TSF #1–2-B; tfl1-17 grown for 13 d in LD conditions. (C) Axillary shoots of the first two leaves in 35S::TSF; tfl1-17 which are converted to single flowers. (D–F) Phenotype of 35S::TSF #2–2; 35S::LFY plants grown for 14 d in LD conditions. (F) A close-up of the primary shoot of the plant in (E), which is converted into a single terminal flower. One cotyledon was removed to reveal the short pedicel (arrowhead). Plants were grown on soil (A) or on 1/2 MS medium (B–F). Scale bars: 1 cm in (A), 0.5 cm in (B), 1 mm in (C) and (E), 2 mm in (D) and 0.5 mm in (F). (G) Effects of TSF and FT overexpression on the expression of SOC1, AGL24, LFY and AP1. Plants were grown on MS medium in LD conditions and harvested at ZT14 on days 4 and 7. View largeDownload slide Fig. 5 Spatial patterns of TSF expression. (A–K) and (N–U) Whole-mount analyses of GUS expression patterns of gTSF::GUS (A, C, E, G, I–K and N–U) and pFT::GUS (B, D, F and H). Cotyledons of LD-grown seedlings (A and B) and SD-grown seedlings (C and D) on day 6. The first leaves of LD-grown seedlings on days 8 (E and F) and 10 (G and H). Insets in (E) and (G) show higher magnification images of the apical region of the leaves. Hypocotyls of LD-grown seedlings on days 6 (I), 8 (J), 10 (K) and 12 (N) and SD-grown seedlings on days 6 (O) and 12 (P). Shoot apical region of LD-grown seedlings on days 6 (Q), 8 (R) and 10 (S) and SD-grown seedlings on days 12 (T) and 28 (U). Q, R, T and U are side views and S is the view from above. Arrowheads in S indicate vascular bundles to cotyledons. (L), (M) and (V–X) Histological analysis of gTSF::GUS expression patterns. A cross-section of hypocotyl of LD-grown plants at day 10 (L). The inside region of the frame in (L) is magnified in (M). Longitudinal sections (V and W) and a cross-section (X) through SAM of 6-day-old gTSF::GUS plants grown in LD conditions. For gTSF::GUS, line #21–1 was used except for (W) (#8–1). Scale bars: 0.5 mm in (A–H) and (S), 0.2 mm in (I–K), (N–R), (T) and (U), 50 µm in (L) and (V–X), and 10 µm in (M). (Y) RT–PCR analysis of expression in LD-grown seedlings on day 7; SA, shoot apex; RL, rosette leaves; Co, cotyledons; Hy, hypocotyls; Ro, roots. (Za–Zf) Effects of TSF and FT overexpression on gTSF::GUS and pFT::GUS expression. F1 plants of gTSF::GUS×Col (Za), gTSF::GUS×35S::FT #11–1 (Zb), gTSF::GUS×35S::TSF #2–2 (Zc), pFT::GUS×Col (Zd), pFT::GUS×35S::FT #11–1 (Ze) and pFT::GUS×35S::TSF #2–2 (Zf). Plants were grown in LD conditions and subjected to GUS staining at ZT14 on day 4. Scale bar: 1 mm. View largeDownload slide Fig. 5 Spatial patterns of TSF expression. (A–K) and (N–U) Whole-mount analyses of GUS expression patterns of gTSF::GUS (A, C, E, G, I–K and N–U) and pFT::GUS (B, D, F and H). Cotyledons of LD-grown seedlings (A and B) and SD-grown seedlings (C and D) on day 6. The first leaves of LD-grown seedlings on days 8 (E and F) and 10 (G and H). Insets in (E) and (G) show higher magnification images of the apical region of the leaves. Hypocotyls of LD-grown seedlings on days 6 (I), 8 (J), 10 (K) and 12 (N) and SD-grown seedlings on days 6 (O) and 12 (P). Shoot apical region of LD-grown seedlings on days 6 (Q), 8 (R) and 10 (S) and SD-grown seedlings on days 12 (T) and 28 (U). Q, R, T and U are side views and S is the view from above. Arrowheads in S indicate vascular bundles to cotyledons. (L), (M) and (V–X) Histological analysis of gTSF::GUS expression patterns. A cross-section of hypocotyl of LD-grown plants at day 10 (L). The inside region of the frame in (L) is magnified in (M). Longitudinal sections (V and W) and a cross-section (X) through SAM of 6-day-old gTSF::GUS plants grown in LD conditions. For gTSF::GUS, line #21–1 was used except for (W) (#8–1). Scale bars: 0.5 mm in (A–H) and (S), 0.2 mm in (I–K), (N–R), (T) and (U), 50 µm in (L) and (V–X), and 10 µm in (M). (Y) RT–PCR analysis of expression in LD-grown seedlings on day 7; SA, shoot apex; RL, rosette leaves; Co, cotyledons; Hy, hypocotyls; Ro, roots. (Za–Zf) Effects of TSF and FT overexpression on gTSF::GUS and pFT::GUS expression. F1 plants of gTSF::GUS×Col (Za), gTSF::GUS×35S::FT #11–1 (Zb), gTSF::GUS×35S::TSF #2–2 (Zc), pFT::GUS×Col (Zd), pFT::GUS×35S::FT #11–1 (Ze) and pFT::GUS×35S::TSF #2–2 (Zf). Plants were grown in LD conditions and subjected to GUS staining at ZT14 on day 4. Scale bar: 1 mm. View largeDownload slide Fig. 6 Effect of vernalization on TSF and FT expression. (A–I) Whole-mount analyses of GUS expression patterns of gTSF::GUS #21–1(A–C), gTSF::GUS #39–1 (D–F) and pFT::GUS (G–I). A, D and G are in the Col background. B, C, E, F, H and I are in the FRI-Sf2 (Col) background. Seedlings were grown in LD conditions for 10 d without vernalization (A, B, D, E, G and H) or after 4 weeks vernalization (C, F and I). (J) RT–PCR analysis of TSF and FT expression in fca-1 with vernalization treatment. Plants were grown on 1/2 MS medium with (+Vern) or without (–Vern) vernalization treatment and harvested at day 7 after being transferred to 22°C. Scale bars: 0.2 mm in (A–F) and 1 mm in (G–I). View largeDownload slide Fig. 6 Effect of vernalization on TSF and FT expression. (A–I) Whole-mount analyses of GUS expression patterns of gTSF::GUS #21–1(A–C), gTSF::GUS #39–1 (D–F) and pFT::GUS (G–I). A, D and G are in the Col background. B, C, E, F, H and I are in the FRI-Sf2 (Col) background. Seedlings were grown in LD conditions for 10 d without vernalization (A, B, D, E, G and H) or after 4 weeks vernalization (C, F and I). (J) RT–PCR analysis of TSF and FT expression in fca-1 with vernalization treatment. Plants were grown on 1/2 MS medium with (+Vern) or without (–Vern) vernalization treatment and harvested at day 7 after being transferred to 22°C. Scale bars: 0.2 mm in (A–F) and 1 mm in (G–I). View largeDownload slide Fig. 7 A model for the integration of regulatory pathways of flowering in Arabidopsis. Four main (photoperiod, vernalization, autonomous and gibberellin) and additional (‘light quality’ and ‘repression’) pathways are shown with representative genes. TSF is placed as a floral pathway integrator with FT, SOC1 and LFY. Regulation of floral pathway integrators by input pathways and regulatory relationships among floral pathway integrators is depicted. TFL1 and FWA (mentioned in the text) were omitted from the model for simplicity. Regulation at the transcriptional level is represented by solid arrows (promotion) and T-bars (repression). Arrows and T-bars to or from a dotted box with TSF and FT in it mean similar regulation of or by the two genes, while a T-bar from TFL2 to FT alone means regulation of only FT. Dotted arrows mean modulation of activity of the gene products or the machinery. Gray arrows represent regulatory relationships that are not directly analyzed in the present work. Because the role of the gibberellin pathway in the regulation of FT remains controversial (Moon et al. 2003, Piñeiro et al. 2003), we tentatively assign an independent position to EBS, which was placed to act through the gibberellin pathway (Piñeiro et al. 2003, Boss et al. 2004). The thickness of lines from CO and FLC is intended as a speculative measure of the strength of promotion and repression. Similarly, the thickness of shaded arrows from floral pathway integrators shows the relative contribution of each integrator to the floral transition. Phloem is the site of the pathway integration for FT and TSF (see text for detail). View largeDownload slide Fig. 7 A model for the integration of regulatory pathways of flowering in Arabidopsis. Four main (photoperiod, vernalization, autonomous and gibberellin) and additional (‘light quality’ and ‘repression’) pathways are shown with representative genes. TSF is placed as a floral pathway integrator with FT, SOC1 and LFY. Regulation of floral pathway integrators by input pathways and regulatory relationships among floral pathway integrators is depicted. TFL1 and FWA (mentioned in the text) were omitted from the model for simplicity. Regulation at the transcriptional level is represented by solid arrows (promotion) and T-bars (repression). Arrows and T-bars to or from a dotted box with TSF and FT in it mean similar regulation of or by the two genes, while a T-bar from TFL2 to FT alone means regulation of only FT. Dotted arrows mean modulation of activity of the gene products or the machinery. Gray arrows represent regulatory relationships that are not directly analyzed in the present work. Because the role of the gibberellin pathway in the regulation of FT remains controversial (Moon et al. 2003, Piñeiro et al. 2003), we tentatively assign an independent position to EBS, which was placed to act through the gibberellin pathway (Piñeiro et al. 2003, Boss et al. 2004). The thickness of lines from CO and FLC is intended as a speculative measure of the strength of promotion and repression. Similarly, the thickness of shaded arrows from floral pathway integrators shows the relative contribution of each integrator to the floral transition. Phloem is the site of the pathway integration for FT and TSF (see text for detail). Table 1 Flowering time of tsf mutants under LD and SD conditions Genotype  No. of rosette leaves  No. of cauline leaves  n  Experiment 1  Long day  Wild type (Col)  12.0 ± 1.2  3.2 ± 0.7  23  tsf-1  13.9 ± 2.7  2.5 ± 0.8  27  ft-1  44.9 ± 8.1†  7.2 ± 1.2  18  ft-1; tsf-1  54.4 ± 6.6†  9.6 ± 1.5  14  Short day  Wild type (Col)  35.1 ± 5.5  7.1 ± 1.2  21  tsf-1  48.3 ± 4.7  8.0 ± 1.9  15  ft-1  44.0 ± 5.7‡  6.8 ± 1.0  18  ft-1; tsf-1  53.0 ± 5.1‡  10.5 ± 0.9  8  Experiment 2  Long day  Wild type (Col)  11.9 ± 1.2  2.5 ± 0.5  11  tsf-1  11.9 ± 1.8  2.9 ± 0.5  15  tfl1-17  7.2 ± 1.5  1.3 ± 0.5  11  tfl1-17; tsf-1  7.9 ± 1.6  1.0 ± 0.5  15  Genotype  No. of rosette leaves  No. of cauline leaves  n  Experiment 1  Long day  Wild type (Col)  12.0 ± 1.2  3.2 ± 0.7  23  tsf-1  13.9 ± 2.7  2.5 ± 0.8  27  ft-1  44.9 ± 8.1†  7.2 ± 1.2  18  ft-1; tsf-1  54.4 ± 6.6†  9.6 ± 1.5  14  Short day  Wild type (Col)  35.1 ± 5.5  7.1 ± 1.2  21  tsf-1  48.3 ± 4.7  8.0 ± 1.9  15  ft-1  44.0 ± 5.7‡  6.8 ± 1.0  18  ft-1; tsf-1  53.0 ± 5.1‡  10.5 ± 0.9  8  Experiment 2  Long day  Wild type (Col)  11.9 ± 1.2  2.5 ± 0.5  11  tsf-1  11.9 ± 1.8  2.9 ± 0.5  15  tfl1-17  7.2 ± 1.5  1.3 ± 0.5  11  tfl1-17; tsf-1  7.9 ± 1.6  1.0 ± 0.5  15  The number of leaves is presented as the average ± SD. There was a statistically significant difference (Student’s t-test, P < 0.002) between two pairs of genotypes marked with † or ‡, respectively. View Large Table 2 Flowering time of 35S::TSF plants under LD and SD conditions Genotype  No. of rosette leaves  No. of cauline leaves  n  Long day  Wild type (Col)  12.1 ± 1.6  3.2 ± 0.7  16  35S::FT #11–1  2.0 ± 0  1.7 ± 0.5  18  35S::TSF #1–2-B  2.1 ± 0.3  1.1 ± 0.3  16  35S::TSF #2–2  2.1 ± 0.5  1.5 ± 0.5  17  35S::TSF #4–1  5.9 ± 0.9  1.2 ± 0.4  13  Short day  Wild type (Col)  35.5 ± 3.0  6.5 ± 0.5  11  35S::FT #11–1  2.0 ± 0  1.6 ± 0.5  16  35S::TSF #1–2-B  2.0 ± 0  1.2 ± 0.4  16  35S::TSF #2–2  2.0 ± 0  1.5 ± 0.5  16  35S::TSF #4–1  6.4 ± 0.9  1.5 ± 0.6  16  Genotype  No. of rosette leaves  No. of cauline leaves  n  Long day  Wild type (Col)  12.1 ± 1.6  3.2 ± 0.7  16  35S::FT #11–1  2.0 ± 0  1.7 ± 0.5  18  35S::TSF #1–2-B  2.1 ± 0.3  1.1 ± 0.3  16  35S::TSF #2–2  2.1 ± 0.5  1.5 ± 0.5  17  35S::TSF #4–1  5.9 ± 0.9  1.2 ± 0.4  13  Short day  Wild type (Col)  35.5 ± 3.0  6.5 ± 0.5  11  35S::FT #11–1  2.0 ± 0  1.6 ± 0.5  16  35S::TSF #1–2-B  2.0 ± 0  1.2 ± 0.4  16  35S::TSF #2–2  2.0 ± 0  1.5 ± 0.5  16  35S::TSF #4–1  6.4 ± 0.9  1.5 ± 0.6  16  The number of leaves is presented as the average ± SD. View Large Table 3 Effects of photoperiod pathway mutations on flowering time of 35S::TSF plants Genotype  No. of rosette leaves  No. of cauline leaves  n  Long day  Wild type (Col)  12.8 ± 1.2  3.9 ± 0.5  12  35S::TSF #2–2  2.7 ± 0.5  1.1 ± 0.6  30  35S::TSF #2–2; co-1  2.0 ± 0  1.0 ± 0  30  35S::TSF #2–2; ft-1  2.6 ± 0.5  0.8 ± 0.4  27  Short day  Wild type (Col)  50.7 ± 5.3  9.3 ± 1.2  15  35S::TSF #2–2  2.6 ± 0.5  0.9 ± 0.4  27  35S::TSF #2–2; co-1  2.0 ± 0  1.0 ± 0  28  35S::TSF #2–2; ft-1  2.5 ± 0.5  0.6 ± 0.5  20  Genotype  No. of rosette leaves  No. of cauline leaves  n  Long day  Wild type (Col)  12.8 ± 1.2  3.9 ± 0.5  12  35S::TSF #2–2  2.7 ± 0.5  1.1 ± 0.6  30  35S::TSF #2–2; co-1  2.0 ± 0  1.0 ± 0  30  35S::TSF #2–2; ft-1  2.6 ± 0.5  0.8 ± 0.4  27  Short day  Wild type (Col)  50.7 ± 5.3  9.3 ± 1.2  15  35S::TSF #2–2  2.6 ± 0.5  0.9 ± 0.4  27  35S::TSF #2–2; co-1  2.0 ± 0  1.0 ± 0  28  35S::TSF #2–2; ft-1  2.5 ± 0.5  0.6 ± 0.5  20  The number of leaves is presented as the average ± SD. View Large Table 4 Flowering time of Fl plants between fwa or FRI-Sf2 (Col) and 35S::TSF Genotype  No. of rosette leaves  No. of cauline leaves  n  Wild type (Col)  10.2 ± 1.2  2.6 ± 0.5  10  35S::TSF #1–2-B×Col (F1)  2.6 ± 0.5  1.1 ± 0.5  15  35S::TSF #2–2×Col (F1)  2.8 ± 0.4  1.1 ± 0.3  14  35S::TSF #1–2-B×fwa-101D (F1)  2.7 ± 0.5  0.9 ± 0.3  10  35S::TSF #2–2×fwa-101D (F1)  2.9 ± 0.5  0.9 ± 0.3  18  35S::TSF #1–2-B×FRI-Sf2 (Col) (F1)  3.9 ± 0.8  1.1 ± 0.3  12  35S::TSF #2–2×FRI-Sf2 (Col) (F1)  5.1 ± 0.9  1.1 ± 0.5  15  fwa-101D×Col (F1)  22.4 ± 2.3  5.7 ± 1.2  10  FRI-Sf2 (Col)×Col (F1)  >63 a  ND  5  Genotype  No. of rosette leaves  No. of cauline leaves  n  Wild type (Col)  10.2 ± 1.2  2.6 ± 0.5  10  35S::TSF #1–2-B×Col (F1)  2.6 ± 0.5  1.1 ± 0.5  15  35S::TSF #2–2×Col (F1)  2.8 ± 0.4  1.1 ± 0.3  14  35S::TSF #1–2-B×fwa-101D (F1)  2.7 ± 0.5  0.9 ± 0.3  10  35S::TSF #2–2×fwa-101D (F1)  2.9 ± 0.5  0.9 ± 0.3  18  35S::TSF #1–2-B×FRI-Sf2 (Col) (F1)  3.9 ± 0.8  1.1 ± 0.3  12  35S::TSF #2–2×FRI-Sf2 (Col) (F1)  5.1 ± 0.9  1.1 ± 0.5  15  fwa-101D×Col (F1)  22.4 ± 2.3  5.7 ± 1.2  10  FRI-Sf2 (Col)×Col (F1)  >63 a  ND  5  The number of leaves is presented as the average ± SD. ND, not determined. a Visible leaves were counted when plants had been grown for 4 months after sowing. Plants were grown under LD conditions. View Large Abbreviations CL constant light CO CONSTANS Col Columbia-0 Cyc cycloheximide Dex dexamethasone GR glucocorticoid receptor GUS β-glucuronidase LD long day Ler Landsberg erecta RT–PCR reverse transcription–PCR 35S cauliflower mosaic virus 35S RNA promoter SAM shoot apical meristem SD short day 3′-UTR 3′-untranslated region Ws Wassilewskija ZT Zeitgeber time References Abe, M., Kobayashi, Y., Yamamoto, S., Daimon, Y., Yamaguchi, A., Ikeda, Y., Ichinoki, H., Notaguchi, M., Goto, K. and Araki, T. ( 2005). FD, a bZIP protein mediating signals from the floral pathway integrator FT at the shoot apex. Science  (in press). Google Scholar Alonso, J.M., Stepanova, A.N., Leisse, T.J., Kim, C.J., Chen, H. et al. ( 2003) Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science  301: 653–657. Google Scholar An, H., Roussot, C., Suárez-López, P., Corbesier, L., Vincent, C., et al. ( 2004) CONSTANS acts in the phloem to regulate a systemic signal that induces photoperiodic flowering of Arabidopsis. Development  131: 3615–3626. Google Scholar Araki, T. ( 2001) Transition from vegetative to reproductive phase. Curr. Opin. Plant Biol.  4: 63–68. Google Scholar Blázquez, M.A., Ahn, J.H. and Weigel, D. ( 2003) A thermosensory pathway controlling flowering time in Arabidopsis thaliana. Nat. Genet.  33: 168–171. Google Scholar Blázquez, M.A. and Weigel, D. ( 1999) Independent regulation of flowering by phytochrome B and gibberellins in Arabidopsis. Plant Physiol.  120: 1025–1032. Google Scholar Boss, P.K., Bastow, R.M., Mylne, J.S. and Dean, C. ( 2004) Multiple pathways in the decision to flower: enabling, promoting, and resetting. Plant Cell  16: S18–S31. Google Scholar Bradley, D., Ratcliffe, O., Vincent, C., Carpenter, R. and Coen, E. ( 1997) Inflorescence commitment and architecture in Arabidopsis. Science  275: 80–83. Google Scholar Brunner, A.M., Busov, V.B. and Strauss, S.H. ( 2003) Poplar genome sequence: functional genomics in an ecologically dominant plant species. Trends Plant Sci.  9: 49–56. Google Scholar Cerdán, P.D. and Chory, J. ( 2003) Regulation of flowering time by light quality. Nature  423: 881–885. Google Scholar Clough, S.J. and Bent, A.F. ( 1998) Floral dip: a simplified method for Arabidopsis thaliana. Plant J.  16: 735–743. Google Scholar Endo, M., Nakamura, S., Araki, T., Mochizuki, N, and Nagatani, A. ( 2005) Phytochrome B in the mesophyll delays flowering by suppressing FLOWERING LOCUS T expression in Arabidopsis vascular bundles. Plant Cell  17: 1941–1952. Google Scholar Gaudin, V., Libault, M., Pouteau, S., Juul, T., Zhao, G., Lefebvre, D. and Grandjean, O. ( 2001) Mutations in LIKE HETEROCHROMATIN PROTEIN 1 affect flowering time and plant architecture in Arabidopsis. Development  128: 4847–4858. Google Scholar Gómez-Mena, C., Piñeiro, M., Franco-Zorrilla, J.M., Salinas, J., Coupland, G. and Martínez-Zapater, J.M. ( 2001) early bolting in short days: an Arabidopsis mutation that causes early flowering and partially suppresses the floral phenotype of leafy. Plant Cell  13: 1011–1024. Google Scholar Foucher, F., Morin, J., Courtiade, J., Cadioux, S., Ellis, N., Banfield, M.J. and Rameau, C. ( 2003) DETERMINATE and LATE FLOWERING are two TERMINAL FLOWER1/CENTRORADIALIS homologs that control two distinct phases of flowering initiation and development in pea. Plant Cell  15: 2742–2754. Google Scholar Halliday, K.J., Salter, M.G., Thingnaes, E. and Whitelam, G.C. ( 2003) Phytochrome control of flowering is temperature sensitive and correlates with expression of the floral integrator FT. Plant J.  33: 875–885. Google Scholar Hayama, R., Yokoi, S., Tamaki, S., Yano, M. and Shimamoto, K. ( 2003) Adaptation of photoperiodic control pathways produces short-day flowering in rice. Nature  422: 719–722. Google Scholar Hepworth, S.R., Valverde, F., Ravenscroft, D., Mouradov, A. and Coupland, G. ( 2002) Antagonistic regulation of flowering-time gene SOC1 by CONSTANS and FLC via separate promoter motifs. EMBO J.  21: 4327–4337. Google Scholar Imaizumi, T., Tran, H.G., Swartz, T.E., Briggs, W.R. and Kay, S.A. ( 2003) FKF1 is essential for photoperiodic-specific light signalling in Arabidopsis. Nature  426: 302–306. Google Scholar Izawa, T., Oikawa, T., Sugiyama, N., Tanisaka, T., Yano, M. and Shimamoto, K. ( 2002) Phytochrome mediates the external light signal to repress FT orthologs in photoperiodic flowering of rice. Genes Dev.  16: 2006–2020. Google Scholar Johanson, U., West, J., Lister, C., Michaels, S., Amasino, R. and Dean, C. ( 2000) Molecular analysis of FRIGIDA, a major determinant of natural variation in Arabidopsis flowering time. Science  290: 344–347. Google Scholar Kardailsky, I., Shukla, V.K., Ahn, J.H., Dagenais, N., Christensen, S.K., Nguyen, J.T., Chory, J., Harrison, M.J. and Weigel, D. ( 1999) Activation tagging of the floral inducer FT. Science  286: 1962–1965. Google Scholar Kaya, H., Sato, S., Tabata, S., Kobayashi, Y., Iwabuchi, M. and Araki, T. ( 2000) hosoba toge toge, a syndrome caused by a large chromosomal deletion associated with a T-DNA insertion in Arabidopsis. Plant Cell Physiol.  41: 1055–1066. Google Scholar Kobayashi, Y., Kaya, H., Goto, K., Iwabuchi, M. and Araki, T. ( 1999) A pair of related genes with antagonistic roles in mediating flowering signals. Science  286: 1960–1962. Google Scholar Kojima, S., Takahashi, Y., Kobayashi, Y., Monna, L., Sasaki, T., Araki, T. and Yano, M. ( 2002) Hd3a, a rice ortholog of the Arabidopsis FT gene, promotes transition to flowering downstream of Hd1 under short-day condition. Plant Cell Physiol.  43: 1096–1105. Google Scholar Koornneef, M., Hanhart, C.J. and van der Veen, J.J. ( 1991) A genetic and physiological analysis of late flowering mutants in Arabidopsis thaliana. Mol. Gen. Genet.  229: 57–66. Google Scholar Kotake, T., Takada, S., Nakahigashi, K., Ohto, M. and Goto, K. ( 2003) Arabidopsis TERMINAL FLOWER 2 gene encodes a heterochromatin protein 1 homolog and represses both FLOWERING LOCUS T to regulate flowering time and several floral homeotic genes. Plant Cell Physiol.  44: 555–564. Google Scholar Kurup, S., Jones, H.D. and Holdsworth, M.J. ( 2000) Interactions of the developmental regulator ABI3 with proteins identified from developing Arabidopsis seeds. Plant J.  21: 143–155. Google Scholar Lee, H., Suh, S.S., Park, E., Cho, E., Ahn, J.H., Kim, S.G., Lee, J.S., Kwon, Y.M. and Lee, I. ( 2000) The AGAMOUS-LIKE 20 MADS domain protein integrates floral inductive pathways in Arabidopsis. Genes Dev.  14: 2366–2376. Google Scholar Michaels, S.D. and Amasino, R.M. ( 1999) FLOWERING LOCUS C encodes a novel MADS domain protein that acts as a repressor of flowering. Plant Cell  11: 949–956. Google Scholar Michaels, S.D. and Amasino, R.M. ( 2001) Loss of FLOWERING LOCUS C activity eliminates the late-flowering phenotype of FRIGIDA and autonomous pathway mutations but not responsiveness to vernalization. Plant Cell  13: 935–941. Google Scholar Michaels, S.D., Himelblau, E., Kim, S.Y., Schomburg, F.M. and Amasino, R.M. ( 2005) Integration of flowering signals in winter-annual Arabidopsis. Plant Physiol.  137: 149–156. Google Scholar Mimida, N., Goto, K., Kobayashi, Y., Araki, T., Ahn, J.H., Weigel, D., Murata, M., Motoyoshi, F. and Sakamoto, W. ( 2001) Functional divergence of the TFL-like gene family in Arabidopsis revealed by characterization of a novel homologue. Genes Cells  6: 327–336. Google Scholar Moon, J., Lee, H., Kim, M. and Lee, I. ( 2005) Analysis of flowering pathway integrators in Arabidopsis. Plant Cell Physiol.  46: 292–299. Google Scholar Moon, J., Suh, S.S., Lee, H., Choi, K.R., Hong, C.B., Paek, N.C., Kim, S.G. and Lee, I. ( 2003) The SOC1 MADS-box gene integrates vernalization and gibberellin signals for flowering in Arabidopsis. Plant J.  35: 613–623. Google Scholar Piñeiro, M., Gómez-Mena, C., Schaffer, R., Martínez-Zapater, J.M. and Coupland, G. ( 2003) EARLY BOLTING IN SHORT DAYS is related to chromatin remodeling factors and regulates flowering in Arabidopsis by repressing FT. Plant Cell  15: 1552–1562. Google Scholar Putterill, J., Robson, F., Lee, K., Simon, R. and Coupland, G. ( 1995) The CONSTANS gene of Arabidopsis promotes flowering and encodes a protein showing similarities to zinc finger transcription factors. Cell  80: 847–857. Google Scholar Ratcliffe, O.J., Amaya, I., Vincent, C.A., Rothstein, S., Carpenter, R., Coen, E.S. and Bradley, D.J. ( 1998) A common mechanism controls the life cycle and architecture of plants. Development  125: 1609–1615. Google Scholar Ratcliffe, O.J., Bradley, D.J. and Coen, E.S. ( 1999) Separation of shoot and floral identity in Arabidopsis. Development  126: 1109–1120. Google Scholar Reeves, P.H. and Coupland, G. ( 2001) Analysis of flowering time control in Arabidopsis by comparison of double and triple mutants. Plant Physiol.  126: 1085–1091. Google Scholar Robson, F., Costa, M.M.R., Hepworth, S., Vizir, I., Piñeiro, M., Reeves, P.H., Putterill, J. and Coupland, G. ( 2001) Functional importance of conserved domains in the flowering-time gene CONSTANS demonstrated by analysis of mutant alleles and transgenic plants. Plant J.  28: 619–631. Google Scholar Ruiz-García, L., Madueño, F., Wilkinson, M., Haughn, G., Salinas, J. and Martínez-Zapater, J.M. ( 1997) Different roles of flowering-time genes in the activation of floral initiation genes in Arabidopsis. Plant Cell  9: 1921–1934. Google Scholar Samach, A., Onouchi, H., Gold, S.E., Ditta, G.S., Schwarz-Sommer, Z., Yanofsky, M.F. and Coupland, G. ( 2000) Distinct roles of CONSTANS target genes in reproductive development of Arabidopsis. Science  288: 1613–1616. Google Scholar Schmid, M., Uhlenhaut, N.H., Godard, F., Demar, M., Bressan, R., Weigel, D. and Lohmann, J.U. ( 2003) Dissection of floral induction pathway using global expression analysis. Development  130: 6001–6012. Google Scholar Shannon, S. and Meeks-Wagner, D.R. ( 1991) A mutation in the Arabidopsis TFL1 gene affects inflorescence meristem development. Plant Cell  3: 877–892. Google Scholar Sheldon, C.C., Burn, J.E., Perez, P.P., Metzger, J., Edwards, J.A., Peacock, W.J. and Dennis, E.S. ( 1999) The FLF MADS box gene: a repressor of flowering in Arabidopsis regulated by vernalization and methylation. Plant Cell  11: 445–458. Google Scholar Sheldon, C.C., Rouse, D.T., Finnegan, E.J., Peacock, W.J. and Dennis, E.S. ( 2000) The molecular basis of vernalization: the central role of FLOWERING LOCUS C (FLC). Proc. Natl Acad. Sci. USA  97: 3753–3758. Google Scholar Simon, R., Igeño, M.I. and Coupland, G. ( 1996) Activation of floral meristem identity genes in Arabidopsis. Nature  384: 59–62. Google Scholar Simpson, G.G. and Dean, C. ( 2002) Arabidopsis, the rosetta stone of flowering time? Science  296: 285–289. Google Scholar Soppe, W.J.J., Jacobsen, S.E., Alonso-Blanco, C., Jackson, J.P., Kakutani, T., Koornneef, M. and Peeters, A.J.M. ( 2000) The late flowering phenotype of fwa mutants is caused by gain-of-function epigenetic alleles of a homeodomain gene. Mol. Cell  6: 791–802. Google Scholar Suárez-López, P., Wheatley, K., Robson, F., Onouchi, H., Valverde, F. and Coupland, G. ( 2001) CONSTANS mediates between the circadian clock and the control of flowering in Arabidopsis. Nature  410: 1116–1120. Google Scholar Suzuki, M., Ketterling, M.G., Li, Q.B. and McCarty, D.R. ( 2003) Viviparous1 alters global gene expression patterns through regulation of abscisic acid signaling. Plant Physiol.  132: 1664–1677. Google Scholar Takada, S. and Goto, K. ( 2003) TERMINAL FLOWER2, an Arabidopsis homolog of HETEROCHROMATIN PROTEIN1, counteracts the activation of FLOWERING LOCUS T by CONSTANS in the vascular tissues of leaves to regulate flowering time. Plant Cell  15: 2856–2865. Google Scholar Valverde, F., Mouradov, A., Soppe, W., Ravenscroft, D., Samach, A. and Coupland, G. ( 2004) Photoreceptor regulation of CONSTANS protein in photoperiodic flowering. Science  303: 1003–1006. Google Scholar Yanovsky, M.J. and Kay, S.A. ( 2002) Molecular basis of seasonal time measurement in Arabidopsis. Nature  419: 308–312. Google Scholar Yoo, S.Y., Kardailsky, I., Lee, J.S., Weigel, D. and Ahn, J.H. ( 2004) Acceleration of flowering by overexpression of MFT (MOTHER OF FT AND TFL1). Mol. Cells  17: 95–101. Google Scholar Yu, H., Xu, Y., Tan, E.L. and Kumar, P.P. ( 2002) AGAMOUS-LIKE 24, a dosage-dependent mediator of the flowering signals. Proc. Natl Acad. Sci. USA  99: 16336–16341. Google Scholar JSPP © 2005 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Plant and Cell Physiology Oxford University Press

TWIN SISTER OF FT (TSF) Acts as a Floral Pathway Integrator Redundantly with FT

Download PDF
15 pages

Loading next page...
 
/lp/oxford-university-press/twin-sister-of-ft-tsf-acts-as-a-floral-pathway-integrator-redundantly-FGxGDdzTTm

References (58)

Publisher
Oxford University Press
Copyright
JSPP © 2005
ISSN
0032-0781
eISSN
1471-9053
DOI
10.1093/pcp/pci151
pmid
15951566
Publisher site
See Article on Publisher Site

Abstract

Abstract In Arabidopsis, several genetic pathways controlling the floral transition (flowering) are integrated at the transcriptional regulation of FT, LFY and SOC1.TSF is the closest homolog of FT in Arabidopsis. TSF expression was induced rapidly upon activation of CONSTANS (CO). The mRNA levels of TSF and FT showed similar patterns of diurnal oscillation and response to photoperiods: an evening peak, higher levels in long day (LD) than in short day (SD) conditions, and immediate up-regulation upon day-length extension. These observations suggest that TSF is a direct regulatory target of CO. tsf mutation delayed flowering in SD conditions and enhanced the phenotype of ft in both LD and SD conditions. TSF and FT also shared similar modes of regulation by FLC, an integrator of autonomous and vernalization pathways, and other factors such as EBS and PHYB. Consistently, TSF overexpression caused a precocious flowering phenotype independent of photoperiods or CO, or FLC. These observations suggest that TSF is a new member of the floral pathway integrators and promotes flowering largely redundantly with FT but makes a distinct contribution in SD conditions. TSF and FT seem to act independently of each other and of LFY, and partially upstream of SOC1. Interestingly, the expression patterns of TSF and FT in seedlings did not overlap, although both were expressed in the phloem tissues. Our work revealed additional complexity and spatial aspects of the regulatory network at the pathway integration level. We propose that the phloem is the site where multiple regulatory pathways are integrated at the transcriptional regulation of FT and TSF. Introduction Plants monitor multiple environmental and endogenous signals to determine when to flower. Genetic and molecular analyses of flowering time mutants in Arabidopsis have led to the current model, in which four major genetic pathways regulate the transition from vegetative to reproductive phase (floral transition or flowering). The photoperiod and vernalization pathways mediate responses to environmental inputs, day lengths and low temperature, respectively. The autonomous and gibberellin pathways act independently of these external signals. These four pathways are integrated at the transcriptional regulation of the floral pathway integrators, FLOWERING LOCUS T (FT), SUPRESSOR OF OVEREXPRESSION OF CO 1 (SOC1) and LEAFY (LFY) (reviewed by Araki 2001, Simpson and Dean 2002, Boss et al. 2004). FT acts as a floral promoter mainly in the photoperiod pathway downstream of CONSTANS (CO). ft mutants show a late flowering phenotype under long-day (LD) conditions, with flowering time only slightly affected under short-day (SD) conditions (Koornneef et al. 1991). Plants overexpressing FT have a precocious flowering phenotype that is independent of photoperiods and the CO function (Kardailsky et al. 1999, Kobayashi et al. 1999). Expression of FT is directly regulated by CO protein, the activity of which is controlled by a circadian clock and light conditions (Kobayashi et al. 1999, Samach et al. 2000, Suárez-López et al. 2001, Yanovsky and Kay 2002, Valverde et al. 2004). Light also affects flowering time independently of photoperiods through the phyB function, as phyB negatively regulates the transcription of FT (Blázquez and Weigel 1999, Cerdán and Chory 2003, Halliday et al. 2003). The floral repressor FLOWERING LOCUS C (FLC), which is the convergence node of the autonomous and vernalization pathways, acts, in part, by repressing the transcription of FT (Samach et al. 2000, Hepworth et al. 2002). Other than these pathways, two chromatin-related factors, TERMINAL FLOWER 2 (TFL2) and EARLY BOLTING IN SHORT DAYS (EBS), are involved in repression of FT transcription, mainly in SD conditions (Kotake et al. 2003, Piñeiro et al. 2003, Takada and Goto 2003). FT encodes a protein similar to phosphatidylethanolamine-binding protein (PEBP), which is also known as Raf kinase inhibitor protein (RKIP) (Kardailsky et al. 1999, Kobayashi et al. 1999). The biochemical functions of the PEBP/RKIP family proteins in plants, including FT, remain to be elucidated. The Arabidopsis genome contains six genes for the PEBP/RKIP family proteins (Kobayashi et al. 1999, Mimida et al. 2001), of which two other genes are also involved or implicated in the regulation of floral transition. TERMINAL FLOWER 1 (TFL1) plays an important role in floral transition in an antagonistic manner to FT, in addition to its pivotal role in the maintenance of inflorescence meristem identity (Bradley et al. 1997, Ratcliffe et al. 1998, Ratcliffe et al. 1999). A recent work suggested that MOTHER OF FTAND TFL1 (MFT), the most distantly related member of the family, might act to promote flowering (Yoo et al. 2004). The regulation of the floral transition by the PEBP/RKIP family is conserved in other plants, such as an SD plant rice and an LD plant Pisum. A quantitative trait locus for the photoperiod sensitivity in rice, named Heading-date 3a (Hd3a), is an FT ortholog and plays an important role in the promotion of flowering in SD conditions (Kojima et al. 2002). Several Hd3a homologs, such as RFT1/FT-L 3 and FTL/FT-L 1, are also implicated in the regulation of flowering (Izawa et al. 2002, Kojima et al. 2002). Like TFL1 in Arabidopsis, the activity of LATE FLOWERING (LF), a TFL1 homolog, in Pisum is important to prevent precocious floral transition (Foucher et al. 2003). TSF (TWIN SISTER OF FT) is the closest homolog of FT, with 82% identity in the deduced amino acid sequence (90% similarity, if amino acid residues with similar biochemical properties are considered). Transgenic plants overexpressing TSF show a precocious flowering phenotype, as do those overexpressing FT (Kobayashi et al. 1999). These observations suggest that FT and TSF may share a similar role in promoting the floral transition. In this study, we investigated the role of TSF in the regulation of floral transition. Our work suggests that TSF is a new member of the floral pathway integrators and promotes flowering redundantly with FT. TSF and FT share similar modes of regulation by major genetic pathways and other factors. They have a similar tissue preference, with expression in the phloem, but have distinct spatial patterns: TSF mainly in hypocotyl and FT in cotyledons and leaves. These observations suggest redundant roles with subtle functional differentiation between the two closely related genes. Among the floral pathway integrators, TSF and FT seem to act independently of each other and of LFY, and partially upstream of SOC1. Phloem seems to represent the site where multiple regulatory pathways are integrated at the transcriptional regulation of FT and TSF. Our work, together with recent reports by others, reveals additional complexity and intricacy of the regulatory network of flowering at the pathway integration level, which may contribute to fine control of the floral transition. Results Effect of tsf mutation on flowering time To investigate whether inactivation of TSF affects the timing of phase transition, we isolated a loss-of-function allele of TSF. We found a line that carries a T-DNA insertion in the second intron of TSF (SALK_087522) (Fig. 1A). No large deletion was found associated with the T-DNA insertion (data not shown). Since severe reduction of TSF transcript levels in homozygous plants was confirmed by reverse transcription–PCR (RT–PCR) (Fig. 1B), we named the insertion allele tsf-1. In addition to the tsf-1 allele, we found that ecotype Wassilewskija (Ws) carries a TSF allele that gives rise to greatly reduced levels of mRNA of the correct size (Fig. 1C). By sequence comparison between Columbia-0 (Col) and Ws, an A to T substitution was found at the splicing acceptor site of the second intron in Ws. It is likely that this substitution results in mis-splicing and is the cause of decreased mRNA levels. Details of natural variation of the TSF locus are described in the Supplementary material (Supplementary text, S-Fig. 1 and S-Table 1). RT–PCR analysis of transcripts in Ws and comparison of flowering time between Landsberg erecta (Ler) and Ws are described in S-Fig. 2 and S-Table 2, respectively. In LD (16 h light/8 h dark) conditions, tsf-1 did not show an obvious difference in flowering time from wild-type plants (Fig. 1D, Table 1). In contrast, tsf-1 flowered later than wild type in SD (8 h light/16 h dark) conditions (Table 1). A similar difference was observed between Ler wild type and tsf-1 introgressed into Ler by three crosses (data not shown). We generated a tsf-1; ft-1 double mutant. tsf mutation enhanced the late flowering phenotype of ft in both LD and SD conditions (Fig. 1D, Table 1). These results suggest that TSF plays a role as a floral promoter in the photoperiod pathway redundantly with FT, but also makes a distinct contribution to flowering in SD conditions. Photoperiodic regulation of TSF transcription The phenotype of tsf and tsf; ft mutants suggests that TSF promotes flowering redundantly with FT in LD conditions. FT acts downstream of the photoperiod pathway through transcriptional regulation by CO, which encodes a nuclear protein with two B-boxes and a CCT domain (Putterill et al. 1995, Samach et al. 2000, Robson et al. 2001). We examined whether transcription of TSF is regulated by CO protein in a photoperiod-dependent manner. Similarly to the case of FT, mRNA levels of TSF were reduced in co-2 (Fig. 2A). To demonstrate further that transcription of TSF is under the control of CO, we used an inducible system in which CO fused to a rat glucocorticoid receptor (GR) is expressed by the cauliflower mosaic virus 35S RNA (35S) promoter (35S::CO:GR) (Simon et al. 1996). Use of this system has demonstrated that FT is an early direct target of CO without requiring intermediary protein synthesis (Samach et al. 2000). TSF expression was induced within 1 h after application of dexamethasone (Dex) in the presence of cycloheximide (Cyc), as was FT expression (Fig. 3A). In contrast, mRNA levels of a late downstream gene APETALA1 (AP1) were not affected. Similarly, LFY and TFL1 expression was not induced by Dex treatment (S-Fig. 3). These results strongly suggest that CO regulates TSF as a direct target. Interestingly, ubiquitous activation of CO by Dex treatment of 35S::CO:GR plants did not result in ubiquitous induction of β-glucuronidase (GUS) under the control of the FT promoter (pFT::GUS, Takada and Goto 2003) or genomic sequences of TSF (gTSF::GUS) (see below and S-Fig. 4B, C). FT expression is barely detectable in SD conditions, but in LD conditions FT mRNA accumulates during the illuminated part of the day with a peak at dusk (Suárez-López et al. 2001, Yanovsky and Kay 2002). TSF mRNA levels showed diurnal oscillation similar to that of FT with a clear peak at dusk in LD conditions (Fig. 3B). In SD conditions, similar diurnal rhythms with a small peak at dusk were also observed, although levels were much lower as compared with LD (Fig. 3B). Transcription of CO is regulated by a circadian clock (Suárez-López et al. 2001, Yanovsky and Kay 2002) and light through the action of FLAVIN-BINDING, KELCH REPEAT, F-BOX 1 (FKF1) (Imaizumi et al. 2003). In addition to the regulation at the transcription level, it was proposed that light regulates CO protein stability to generate the daily rhythm of protein abundance and resulting in the CO activity being restricted to the evening in LD conditions. This, in turn, is responsible for the diurnal oscillation of FT (Imaizumi et al. 2003, Valverde et al. 2004). We performed day-length shift experiments to examine the effect of day-length change on regulation of TSF expression. Plants were entrained in SD conditions for 6 d after imbibition, and day-length conditions were changed from SD to LD by extending the light period on day 7. As expected, expression of TSF was induced immediately upon day-length extension, as was that of FT (Fig. 3C). Light in the subjective night of SD-entrained plants [Zeitgeber time (ZT) 8 to ZT16] on day 7 affected the oscillation pattern of CO mRNA levels (Fig. 3C). As previously described (Imaizumi et al. 2003), CO expression in SD-to-LD-shifted plants showed a biphasic pattern similar to that of the LD-entrained plants, with an earlier daytime peak due to the action of FKF1 (ZT10 vs. ZT13). The shifted daytime peak of CO mRNA accumulation is likely to be responsible for the earlier peak of TSF mRNA levels on day 7 than on day 8. These results strongly suggest that, as in the case of FT, the diurnal oscillation of TSF expression is controlled by CO activity regulated by a circadian clock and light conditions. Photoperiod- and CO-independent precocious flowering phenotype by constitutive overexpression of TSF We previously reported that plants overexpressing TSF (35S::TSF) showed a precocious flowering phenotype in LD conditions (Kobayashi et al. 1999). This was confirmed in representative homozygous lines (Fig. 4A). If TSF promotes flowering under the regulation of CO, it is expected that the early flowering phenotype of the 35S::TSF plants is independent of photoperiods and CO function. As expected, 35S::TSF plants flowered with the same number of leaves in LD and SD conditions (Table 2). Similarly, co-1 mutation did not affect the precocious flowering phenotype of 35S::TSF plants (Table 3). It has been reported recently that the expression levels of SOC1 were elevated in seedlings of FT and TSF activation-tag lines (Michaels et al. 2005) and 35S::FT seedlings (Moon et al. 2005). In good agreement with these reports, we observed that the amount of SOC1 mRNA was increased in 35S::TSF and 35S::FT seedlings as early as on day 4 (Fig. 4G). In addition, mRNA levels of another floral pathway integrator LFY were increased on day 4 in 35S::TSF and 35S::FT. A MADS-box gene AGL24 has been assigned an intermediate position between SOC1 and LFY, since AGL24 expression is reduced in soc1 and LFY expression is reduced in agl24 (Yu et al. 2002). However, no change in the expression level of AGL24 was observed in 35S::TSF and 35S::FT (Fig. 4G). No difference of AP1 expression level was observed until day 7. Consistent with the increase in 35S::FT and 35S::TSF, a slight decrease in SOC1 mRNA levels was observed in tsf and ft seedlings (Fig. 4G). TSF and FT did not affect each other’s expression (Fig. 4G, 5Za–Zf). It has been reported that the precocious flowering phenotype of 35S::FT was dramatically enhanced by 35S::LFY so that seedlings formed a single terminal flower immediately after germination (Kardailsky et al. 1999, Kobayashi et al. 1999, Moon et al. 2005). Similarly, the phenotype of 35S::TSF was greatly enhanced by 35S::LFY, although enhancement was not as strong as in the case of 35S::FT. After forming two severely up-curled rosette leaves, the primary shoot terminated with a single flower (Fig. 4D–F). In contrast, no enhancement of phenotype was observed in 35S::TSF; 35S::FT plants as compared with either parental plant (data not shown). Multiple environmental and endogenous factors that regulate TSF expression In Arabidopsis, floral transition is regulated by multiple genetic pathways such as autonomous, vernalization and gibberellin pathways, as well as the photoperiod pathway. FT is regulated at the transcriptional level by the inputs from these pathways and acts as one of the pathway integrators. As shown above, TSF is likely to act as a floral promoter in the photo-period pathway downstream of CO in a similar manner to FT. We are interested in whether TSF is also regulated by inputs from multiple pathways. We first examined TSF expression in flowering time mutants, which have been reported to affect FT mRNA levels (Fig. 2). In the following experiments, samples were collected at ZT14 in LD, and ZT7 or 8 in SD conditions, when TSF levels are highest in the respective day-length conditions. The vernalization and autonomous pathways promote the floral transition by reducing the expression levels of the FLC, which encodes a MADS-box transcription factor and acts as a repressor of flowering (Michaels and Amasino 1999, Sheldon et al. 1999, Sheldon et al. 2000, Michaels and Amasino 2001). FLC delays floral transition through repression of FT and SOC1 expression (Lee et al. 2000, Samach et al. 2000, Hepworth et al. 2002). A defect in the autonomous pathway results in the increased expression of FLC, which is counteracted by vernalization treatment (Michaels and Amasino 1999, Sheldon et al. 1999). Functional alleles at the FRIGIDA (FRI) locus promote the accumulation of FLC mRNA and confer vernalization requirement in the winter-annual accessions (Michaels and Amasino 1999, Sheldon et al. 1999, Johanson et al. 2000). In fca-1 defective in the autonomous pathway, TSF and FT mRNA levels were reduced (Fig. 2A). Similarly, Col plants with a functional FRI allele from ecotype San Feliu-2 (FRI-Sf2) [hereafter referred to as FRI-Sf2 (Col)] had reduced mRNA levels of TSF and FT (Fig. 2A). As described below, vernalization treatment restored TSF (or gTSF::GUS) expression in fca-1 and FRI-Sf2 (Col) (Fig. 6). These results suggest that TSF expression is negatively regulated by FLC, as is FT expression. In agreement with these results, 35S::TSF was able to suppress the extreme late flowering phenotype conferred by the active FRI-FLC system (Table 4). In Arabidopsis, phyB has a role in the shade-avoidance response which includes accelerated flowering. It has been reported that FT mRNA levels are higher in early flowering phyB mutants than in wild type in LD conditions (Cerdán and Chory 2003). We tested whether TSF mRNA levels are also affected in phyB mutants. A clear increase in mRNA levels of TSF, as well as FT, was observed in 5-day-old phyB-9 mutants in LD conditions (Fig. 2B). Similar differences in TSF and FT mRNA levels were also observed between Ler and phyB-5 (data not shown). Arabidopsis mutants that exhibit the early flowering phenotype have revealed the existence of genes involved in the repression of flowering. Some of these repressors act independently of environmental factors to prevent plants from precocious transition to flowering during early developmental stages or in inappropriate conditions. In non-inductive SD conditions, it has been reported that FT expression is repressed by EBS which encodes a nuclear protein with a bromoadjacent homology domain and a plant homeodomain Zn finger (Piñeiro et al. 2003). In the early flowering ebs-1 mutant, expression levels of TSF were increased as well as those of FT in SD conditions (Fig. 2C). Another floral repressor, TFL2, which encodes a protein with homology to heterochromatin protein 1 (HP1), represses FT expression (Kotake et al. 2003, Takada and Goto 2003). We examined the effect of tfl2 mutation on TSF mRNA levels by RT–PCR. Although FT mRNA levels were increased in the tfl2 mutant in SD conditions as previously reported, TSF expression was not affected (Fig. 2D). Spatial patterns of TSF expression In LD-grown seedlings, TSF was expressed mainly in hypocotyls (Fig. 5Y). In contrast, FT expression was detected mainly in cotyledons (Fig. 5Y). In mature plants, high levels of TSF expression were observed in flowers and developing siliques (data not shown) as in the case of FT (Kobayashi et al. 1999). In situ RNA hybridization has been tried, but did not succeed. Similar trials to detect FT expression in wild-type seedlings have not been successful (data not shown; see Takada and Goto 2003, An et al. 2004). To examine spatial patterns of expression in detail, we generated transgenic plants expressing GUS under the control of regulatory sequences of a TSF genomic fragment (gTSF::GUS). Since extensive natural variations in the TSF locus and the adjacent intergenic region were found, some of which involve insertion of a retroelement in the 3′-untranslated region (3′-UTR), we selected the simplest Ler sequence as the source for the reporter construct (see Materials and Methods and S-Fig.1). Consistent with the RT–PCR analysis described above (Fig. 3B), expression of gTSF::GUS was at higher levels and over a broader range in LD than in SD conditions. In 6-day-old seedlings in LD conditions, expression of gTSF::GUS was observed in the vascular tissues of hypocotyl and petiole and the basal part of cotyledons (Fig. 5A, I). A cross-section of hypocotyls of an LD-grown seedling at day 10 showed staining in the phloem parenchyma (Fig. 5L, M). Signals were also detected near the shoot apical meristem (SAM) (Fig. 5Q–S). Close examination showed that the expression is confined to a small number of cells between the SAM and young leaf primordia (Fig. 5V–X). gTSF::GUS expression was barely observable in 6-day-old plants in SD conditions (Fig. 5C, O). It was reported previously that pFT::GUS is expressed in phloem tissues within the apical part of cotyledons and leaves (Takada and Goto 2003) and we confirmed this (Fig. 5B, D, F, H). Therefore, the expression of the two genes does not seem to overlap in young seedlings (compare Fig. 5A, C, E, G, Za for gTSF::GUS and Fig. 5B, D, F, H, Zd for pFT::GUS). gTSF::GUS expression in hypocotyls and near SAM gradually increased with growth in LD conditions (Fig. 5I–K, N). Expression in true leaves was first detected in phloem at the apical-most region from day 8 (see the insets in Fig. 5E and G). In SD conditions, the signals in hypocotyls and leaves were low or almost undetectable (Fig. 5O, P, T). However, in 28-day-old plants grown under SD conditions, clear signals were detected near SAM (Fig. 5U). In mature plants, gTSF::GUS was expressed in the vascular tissues of inflorescence stems, pedicels, floral organs and roots (data not shown). These expression patterns of TSF observed in later developmental stages were, at least in part, similar to those of FT (data not shown). We made use of the gTSF::GUS reporter to examine the effect of induction of CO activity on spatial patterns of TSF expression. Although transient ubiquitous activation of CO is expected with Dex treatment of 35S::CO:GR plants, neither gTSF:GUS nor pFT:GUS were induced ubiquitously (S-Fig. 4). These results suggest that there may be additional factor(s) other than CO that restrict spatial patterns of TSF and FT expression upon activation of CO. We also analyzed the response of TSF expression to vernalization using the gTSF::GUS reporter. In F1 plants with FRI-Sf2 (Col) (gTSF::GUS/-; FRI-Sf2/fri-Col), gTSF::GUS expression was greatly reduced compared with plants without functional FRI (gTSF::GUS; fri-Col/fri-Col) (compare Fig. 6A, D and Fig. 6B, E). This was consistent with the results obtained by RT–PCR analysis of TSF expression (Fig. 2A). After 4 weeks vernalization treatment, gTSF::GUS expression in an FRI-active background was released from repression by FLC [compare Fig. 6B, E (without vernalization) and Fig. 6C, F (with vernalization)]. Again, these observations were consistent with the RT–PCR analysis of the endogenous gene expression (Fig. 6J). Expression of pFT::GUS showed essentially the same responses (Fig. 6G–J). Genetic interactions with tfl1 and fwa mutations We investigated whether TSF and FT behave in a similar manner in genetic interaction with tfl1 and fwa mutations. FT and TFL1 have opposite roles in regulation of flowering despite their high degree (55%) of sequence identity (Kardailsky et al. 1999, Kobayashi et al. 1999). Loss of TFL1 results in an early flowering phenotype, in addition to conversion of inflorescences to flowers (Shannon and Meeks-Wagner 1991). ft and tfl1 are additive so that ft; tfl1 plants flower slightly earlier than ft, and inflorescences produce terminal flowers (Ruiz-García et al. 1997). Similarly, tsf and tfl1 were additive in phenotype. The tsf-1; tfl1-17 double mutant flowered as early as tfl1-17 and produced terminal flowers (Table 1 and data not shown). As in the case of FT overexpression, the tfl1-17 enhanced the early flowering and inflorescence to flower conversion phenotype of 35S::TSF. 35S::TSF; tfl1 flowered earlier than 35S::TSF, and axillary shoots in 35S::TSF; tfl1 were replaced with solitary flowers (Fig. 4B, C). In semi-dominant fwa ‘epimutants’, a homeobox gene FWA is ectopically expressed in seedlings due to hypomethylation of the direct repeat sequences in the promoter (Soppe et al. 2000). It was reported previously that the precocious flowering phenotype of 35S::FT was strongly suppressed by fwa (Kardailsky et al. 1999, Kobayashi et al. 1999). It has been suggested that ectopically expressed FWA interferes with the FT function by an unknown mechanism (Kardailsky et al. 1999, Kobayashi et al. 1999). In contrast to 35S::FT, the phenotype of 35S::TSF was little affected by fwa. The flowering time of F1 plants derived from a cross between fwa-101D and 35S::TSF was as early as that of 35S::TSF hemizygotes (Table 4). As in the case of FT, TSF mRNA levels were not affected by fwa-2 mutation (Fig. 2A). Discussion Genetic and molecular studies in Arabidopsis have led to a model in which the photoperiod, vernalization, autonomous and gibberellin pathways regulate the floral transition through the transcriptional regulation of the floral pathway integrators, FT, SOC1 and LFY (reviewed by Araki 2001, Simpson and Dean 2002, Boss et al. 2004). It has been shown that co; fca; ga1 triple mutant plants impaired in the photoperiod, autonomous and gibberellin pathways, without vernalization, do not flower even under inductive LD conditions. This indicates that the three pathways are essential for flowering to occur and that the vernalization pathway is absolutely required in the absence of the three major pathways (Reeves and Coupland 2001). In contrast, a recent work demonstrated that the floral transition is not completely blocked in an ft; soc1; lfy triple mutant lacking all of the three known pathway integrators (Moon et al. 2005). These facts suggest that there are as yet unidentified gene(s) acting at the pathway integration level to promote the floral transition. Among the three pathway integrators, FT is a member of a small gene family with TSF as the closest homolog (Kardailsky et al. 1999, Kobayashi et al. 1999, Mimida et al. 2001), and transgenic analysis has suggested a potential role for TSF in the promotion of flowering (Kobayashi et al. 1999). In rice, the FT ortholog Hd3a is involved in photoperiodic regulation of flowering, acting downstream of the CO ortholog Hd1 (Kojima et al. 2002, Hayama et al. 2003). Some of the Hd3a homologs have also been implicated in the regulation of flowering based on expression profiles and overexpression phenotypes (Izawa et al. 2002, Kojima et al. 2002). However, the actual roles of TSF and Hd3a homologs in the regulation of flowering remain to be demonstrated. In the present work, we investigated the role of TSF in floral transition. TSF promotes flowering as a floral pathway integrator Phenotypes of mutant and transgenic plants (Fig. 1D, 4A, Tables 1–3) suggest that TSF promotes flowering redundantly with FT in the photoperiod pathway. Functional FT in LD conditions can compensate for the loss of TSF. Since FT mRNA levels did not change in tsf-1 (Fig. 1B, see also Fig. 4G), compensation is not due to up-regulation of FT in response to the loss of TSF. On the other hand, reduced levels of FT expression in SD conditions cannot fully substitute for TSF. There seems to be a discrepancy between our observation that the tsf single mutant flowered later than the wild type in SD conditions and a recent report by Michaels et al. (2005), who observed no difference using the same T-DNA insertion allele. Because the difference was fully reproducible in our conditions (similar results were obtained twice in the original Col background and once in the Ler background after introgression), we think the discrepancy was due to the difference in growth conditions between the two groups. Interestingly, ecotype Ws with a natural variation causing severe reduction in TSF mRNA levels (see Supplementary material) flowered as early as Ler in LD conditions, but much later than Ler in SD conditions (S-Table 2). It will be interesting to explore the association between the natural variation at the TSF locus and the flowering time phenotype in SD conditions. Severe reduction of mRNA levels in a co mutant (Fig. 2A), rapid induction of mRNA accumulation by treatment of 35S::CO:GR seedlings with Dex in the presence of Cyc (Fig. 3A), very low mRNA amounts in SD conditions (Fig. 3B) and an immediate increase in mRNA levels in SD-entrained seedlings upon day-length extension (Fig. 3C), taken together, strongly suggest that transcription of TSF is directly regulated by CO, which is under the control of a circadian clock and light at the transcription and protein activity levels. In agreement with this, gTSF::GUS expression was higher in LD than in SD conditions (Fig. 5) and gTSF::GUS expression in 35S::CO:GR seedlings was induced by treatment with Dex (S-Fig. 4C). As expected, constitutive expression of TSF caused a precocious flowering phenotype independently of photoperiods and CO function (Tables 2, 3). TSF and FT, therefore, share direct regulation by CO in the photoperiod pathway. In a previous attempt to identify CO targets, TSF was not identified (Samach et al. 2000). This may be due to the fact that TSF has much lower mRNA levels than FT even after induction by CO (data not shown). Recently, several sequence motifs responsible for regulation by CO have been identified by analysis of the FT promoter (H. Nakagawa and T. Izawa, personal communication). Interestingly, sequences similar to at least some of these motifs are also found in the TSF promoter (T. Izawa, personal communication). Since the sequence upstream of the transcribed region is much shorter in TSF (1.5 kb) than in FT (7.5 kb) (see Kaya et al. 2000 for the FT region on chromosome 1 and S-Fig.1 for the TSF region on chromosome 4), use of the TSF promoter fragment will facilitate further analysis. Repression of TSF expression in the presence of high levels of FLC expression (Fig. 2A, 6B, E, J) and de-repression of TSF by vernalization through reduction of FLC levels (Fig. 6C, F, J) suggest that TSF is a regulatory target of the floral repressor FLC. Consistently, constitutive expression of TSF suppressed the extreme late flowering phenotype conferred by the active FRI-FLC system (Table 4), which is in good agreement with the recent identification of an activation-tagged TSF allele as a suppressor of FRI-Sf2 (Col) (Michaels et al. 2005). In addition, sequences similar to the CArG box motif (MADS-domain protein-binding site) were found near the coding region of TSF (data not shown). The presence of additional regulatory targets of FLC and vernalization other than FT and SOC1 has recently been suggested, since the flowering time of the ft; soc1 double mutant was accelerated by flc-3 and further delayed by an active FRI-Sf2 allele, and the ft; soc1; lfy triple mutant still responds to vernalization (Moon et al. 2005). Our results indicate that TSF represents the suggested target. It will be interesting to examine the phenotypes of ft; tsf; soc1 and ft; tsf; soc1; lfy mutants in terms of their interaction with the FRI-FLC system and their response to vernalization. Although TSF, with FT, is clearly placed at the convergent point of the photoperiod and autonomous/vernalization pathways, the relationship of TSF with the gibberellin pathway remains unclear. EBS, which encodes a chromatin-related protein, has been implicated in mediating signals in the gibberellin pathway (Gómez-Mena et al. 2001, Piñeiro et al. 2003). Expression of TSF, as well as FT, in SD conditions was de-repressed in an ebs mutant (Fig. 2C). Based on the effect of ebs on FT expression, Piñeiro et al. (2003) discussed a possible contribution of the gibberellin pathway to the regulation of FT. The same argument may be applicable to TSF as well. However, since there is no direct experimental evidence for the involvement of the gibberellin pathway in the regulation of FT (Moon et al. 2003) and TSF, further investigation will be required before a firm conclusion can be reached. Besides the four major genetic pathways, other pathway(s) or factor(s) contribute to the regulation of flowering. Light quality, through the action of phyB, is one such environmental factor (Blázquez and Weigel 1999). phyB seems to negatively regulate TSF and FT expression, since mRNA levels of TSF and FT were elevated in phyB mutants (Fig. 2B and see Cerdán and Chory 2003, Halliday et al. 2003). It has been suggested recently that phyB acts through PFT1, which encodes a putative nuclear protein, to regulate FT at least partly in a CO-independent manner (Cerdán and Chory 2003). It is likely that PFT1 also mediates regulation of TSF by phyB. Ambient temperature is another environmental factor that affects the timing of flowering (Blázquez et al. 2003, Halliday et al. 2003). It has been reported that TSF expression is influenced by ambient temperatures, with mRNA levels higher at 16°C than at 23°C, and that FT has an opposite response (Blázquez et al. 2003). The relationship of TSF to the ‘thermosensory pathway’ is an interesting problem to be investigated further. TFL2, an Arabidopsis counterpart of HP1 (Gaudin et al. 2001, Kotake et al. 2003), is another candidate for a regulator of TSF expression. However, in sharp contrast to FT, expression of TSF was not affected by the tfl2 mutation (Fig. 2D). It is likely that TFL2 regulates genes in euchromatin regions through localized modification of chromatin structure. It is to be noted that while the FT locus is rather isolated from the neighboring genes (see Fig. 7A of Kaya et al. 2000 for gene distribution in the FT region), TSF is a typical Arabidopsis gene with short intergenic sequences (S-Fig. 1). This might be reflected in the difference in TFL2-mediated regulation. Interestingly, ABA and ABSCISIC ACID-INSENSITIVE 3 (ABI3) recently have been implicated in the regulation of TSF based on the observation that TSF was repressed in plants overexpressing the maize ABI3 ortholog, VIVIPAROUS1 (VP1), when treated with ABA (Suzuki et al. 2003). It has been known that abi3 mutants flower earlier than the wild type (Kurup et al. 2000). Therefore, possible involvement of ABA in the regulation of TSF is an interesting problem to be investigated. In conclusion, TSF is a new pathway integrator regulated by various pathways and factors largely in a similar manner to FT (Fig. 7). The regulatory relationship of TSF (and FT) with the gibberellin pathway and signals from ambient temperatures and ABA (through ABI3) remain interesting issues requiring further investigation. Recent studies have revealed regulatory interactions among the pathway integrators. For example, it has been reported that mRNA levels of SOC1 were decreased in the ft mutant (Lee et al. 2000, Schmid et al. 2003) and increased in FT and TSF activation-tag lines and 35S::FT plants (Michaels et al. 2005, Moon et al. 2005). Our observation that the amount of SOC1 mRNA was increased in 35S::TSF and 35S::FT plants and slightly decreased in tsf and ft mutants (Fig. 4G) confirmed these recent reports. These findings strongly suggest that SOC1 is, in part, positively regulated by FT and TSF (Fig. 7). Interestingly, it was demonstrated that activation-tagged alleles of TSF and FT can overcome the repressive effect of the active FRI-FLC system on SOC1 (Michaels et al. 2005). This is in contrast to the case of 35S::CO which cannot overcome the repression of SOC1 by 35S::FLC (Hepworth et al. 2002). FT and TSF seem to regulate SOC1 independently of the upstream regulators FLC and CO. It has been shown that overexpression of SOC1 has little effect on FT and TSF mRNA levels, indicating the absence of reciprocal regulation of FT and TSF by SOC1 (Michaels et al. 2005, Moon et al. 2005). Another pathway integrator, LFY, has the least inter-dependence with FT (Kardailsky et al. 1999, Kobayashi et al. 1999, Moon et al. 2005). Similarly, LFY and TSF are likely to act rather independently (Fig. 7). The lack of enhancement of the phenotype in 35S::FT; 35S::TSF indicates that FT and TSF act through common downstream factors. It is unlikely that FT and TSF regulate each other at the transcription level (see discussion below). Spatial pattern of TSF and FT expression and functional differentiation Although our knowledge of genetic interactions among regulators of flowering has improved significantly, how these regulators and pathways are spatially organized is still poorly understood. The spatial patterns of expression of several important regulators such as CO and FT have only recently been described in detail (Takada and Goto 2003; An et al. 2004). We analyzed and compared spatial patterns of expression of TSF and FT by using reporter constructs (gTSF::GUS and pFT::GUS) (Fig. 5). gTSF::GUS expression was observed mainly in the vascular (phloem) tissues of hypocotyl and petiole and the basal part of cotyledons (Fig. 5). As previously reported (Takada and Goto 2003), pFT::GUS expression was observed mainly in vascular (phloem) tissues of cotyledons and leaves, where it was stronger in the apical region than in the basal region and petiole (Fig. 5). Therefore, TSF and FT share similar tissue specificity of expression in phloem, but have largely non-overlapping patterns of expression in young seedlings. The difference in spatial patterns of TSF and FT is probably due, in part, to the difference in negative regulation by TFL2. Expression of TFL2 as demonstrated by gTFL2::GUS is stronger in the basal region of the leaves (Takada and Goto 2003), and TFL2 seems to be responsible for down-regulation of FT, but not TSF, in the basal region of cotyledons and leaves. A mutual regulation model of TSF and FT expression recently proposed by Michaels et al. (2005) might also explain the difference in the spatial patterns. However, we could not reproduce their observation that TSF mRNA levels are reduced in 35S::FT and FT mRNA levels in 35S::TSF (Fig. 4G). Analysis with GUS reporters also failed to provide evidence for mutual, either positive or negative, regulation (Fig. 5Za–Zf). Furthermore, we observed effects of neither ft mutation on TSF mRNA levels nor tsf on FT mRNA levels (Fig. 4G). Therefore, we do not think it likely that mutual repression between TSF and FT is the cause of the non-overlapping patterns. Another important difference in spatial patterns of expression of TSF and FT was the presence vs. the absence of expression in the shoot apical region. Whereas gTSF::GUS expression was consistently observed in a small number of cells near the shoot apex, pFT::GUS expression could not be detected associated with the shoot apex in similar experimental conditions for growth and GUS staining (Fig. 5 and data not shown, see also Takada and Goto 2003). Although the possibility that this might be due to the difference in the design of reporter constructs cannot be ruled out, we think that it reflects an actual difference in the patterns of endogenous gene expression. Observation of whole-mount preparations as well as cross- and longitudinal sections of gTSF::GUS plants showed that GUS staining is not in the shoot apical meristem per se, but surrounds the meristem and is sometimes associated with the base of young leaf primordia. Expression of gTSF::GUS in the shoot apical region seems to be regulated by photoperiods (possibly via CO) and the FRI-FLC system. To our knowledge, there are no reports of genes with similar patterns of expression in the shoot apical region. On the other hand, similar vascular expression in hypocotyl was reported for Arabidopsis thaliana CENTRORADIALIS homologue (ATC), which also belongs to the PEBP/RKIP gene family (Mimida et al. 2001). These differences in spatial patterns of expression are likely to be responsible for the fact that the loss of FT function is not compensated for by functional TSF in LD conditions (Fig. 1D, Table 1). Since the late flowering phenotype of the ft mutation was fully suppressed by 35S::TSF, resulting in 35S::TSF; ft-1 plants flowering as early as 35S::TSF plants (Table 3), TSF can substitute for FT activity when it is expressed in the proper tissues and/or in amounts above a certain threshold level. Although it is difficult to compare the mRNA levels of the two genes quantitatively, we attempted to gain some clues by RT–PCR with a common set of primers for TSF and FT followed by digestion with restriction enzymes to distinguish the two gene products. The amount of TSF mRNA was transiently higher than that of FT immediately after germination, but soon afterwards it became much lower than that of FT (data not shown), an observation consistent with our previous results (Kobayashi et al. 1999). Therefore, lower levels of TSF expression may also be responsible for the lack of compensation for ft. On the other hand, most observations including a similar genetic interaction with TFL1, which encodes another protein of the PEBP/RKIP family and plays an opposite role in flowering, suggest that FT and TSF proteins have equivalent functions (Fig. 4B, C and Table 1; Kobayashi et al. 1999, Kardailsky et al. 1999). However, since 35S::FT and 35S::TSF had different interactions with fwa, there might be subtle functional differentiation at the protein level as well (Table 4). Experiments to express TSF under the control of the FT regulatory sequence or FT under the control of the TSF regulatory sequence in ft mutants should provide important clues. The fact that the tsf single mutant showed a late flowering phenotype in SD conditions (actually tsf-1 was later than ft-1; see Table 1) suggests that TSF makes a greater relative contribution to the promotion of flowering in SD than in LD conditions. So far, the reason for this is not fully clear, and detailed analysis of TSF and FT expression in plants prior to the floral transition in SD conditions will be necessary to gain some insight. A greater relative contribution of TSF may be observed in other conditions, such as lower ambient temperatures, as well (see discussion above). Through differential contributions under various environmental conditions, FT and TSF may fine-tune the timing of the floral transition and provide robustness for the process that integrates the multiple floral signals. Further investigation to identify the environmental conditions in which TSF makes a distinct contribution is necessary to explore this interesting possibility. In rice, there are nine homologs of FT, and the expression of three of them (Hd3a, RFT/FT-L 3 and FTL/FT-L 1), which belong to the same clade with FT and TSF, was analyzed in detail by RT–PCR (Izawa et al. 2002). Interestingly, certain differences in response to photoperiods and developmental regulation were observed among the three genes, which otherwise were similar in regulation, suggesting that a certain degree of differentiation in their roles, likely, in flowering (Izawa et al. 2002). Therefore, the redundant roles of FT and TSF, with subtle functional differentiation, will not be a unique situation in Arabidopsis. Genome projects in species other than Arabidopsis and rice have revealed multiple members of the FT clade in the PEBP/RKIP family (Brunner et al. 2003). It is likely that similar redundancy and differentiation of multiple FT homologs contribute to the fine regulation of flowering in other species as well. Spatial aspects of pathway integration and future perspectives The expression of TSF and FT in phloem is likely to be determined by the direct regulator CO, which is expressed mainly in phloem (Takada and Goto 2003, An et al. 2004), although additional factor(s) may also contribute to the restriction of TSF and FT expression to phloem tissue. Expression of TSF and FT in phloem of hypocotyl and cotyledons, respectively, is also under the control of FLC and vernalization (Fig. 6). Furthermore, phyB seems to exert its negative regulatory effect on FT and TSF mainly in phloem, since ectopic GUS staining was observed neither in gTSF::GUS; phyB-9 nor in pFT::GUS; phyB-9 (S-Fig. 5). By quantitative RT–PCR analysis of FT expression in mesophyll and vascular cells of wild-type and phyB seedlings, a similar conclusion has been reached for FT (Endo et al. 2005). Taken together, it is likely that phloem represents one of the tissues where signals from the photoperiod, autonomous, vernalization and light quality pathways are integrated. One interesting aspect of integration is the possibility of release from repression by FLC in presumably non-dividing cells in the vasculature of cotyledons and hypocotyl, an interesting issue to be investigated further with the aid of the GUS reporters. That the phloem represents the tissue where signals from various pathways are integrated at the transcriptional regulation of FT and TSF raises the interesting question of the possible relationship of these two genes to a hitherto unidentified leaf-generated long-distance floral stimulus acting at the shoot apex (called ‘florigen’). We have accumulated evidence for FT acting at the shoot apical meristem (Abe et al. 2005) and there is some evidence for TSF also acting at the shoot apex. We also found that the effect of 35S::FT to promote flowering is graft transmissible in Arabidopsis (M. Notaguchi, Y. Daimon, M. Abe and T. Araki, unpublished results), and similar graft transmissibility is expected for 35S::TSF. These findings support the recent hypothesis that FT is involved in the generation of, or its possible identity with, the long-distance signal(s) promoting flowering (Takada and Goto 2003, An et al. 2004). An et al. (2004) also suggested the presence of CO-dependent, but FT-independent component(s) in the long-distance signals. TSF is likely to represent at least part of the suggested components. Considering all this together, a logical step forward is to explore whether FT and TSF, with some differentiation, actually represent the long-sought ‘florigen’. Materials and Methods Plant materials and growth conditions Col and Ler were used as wild types. Ecotype Ws was also used in some experiments. A T-DNA insertion allele of TSF in the Col background (SALK_087522) (Alonso et al. 2003) was obtained from the Arabidopsis Biological Resource Center and backcrossed with wild-type Col three times. 35S::CO:GR; co-2 was formerly obtained from G. Coupland (currently at Max Planck Institute for Plant Breeding Research, Cologne, Germany) and Plant BioScience Ltd, UK. ft-1 introgressed into the Col background, referred to as ft-1 (Col) (described in Yoo et al. 2004), was obtained from J. H. Ahn (Korea University, Korea). Col carrying a functional FRI allele from ecotype San Feliu-2 (FRI-Sf2), referred to as FRI-Sf2 (Col), was obtained from R. Amasino (University of Wisconsin, Madison, WI, USA). ebs-1 and phyB-9 were obtained from J. Martínez-Zapater (Universidad Autonoma de Madrid, Spain) and A. Nagatani (Kyoto University, Japan), respectively. A strong line (#11–1) of 35S::FT was previously described (Kobayashi et al. 1999). 35S::TSF transgenic lines were previously generated (Kobayashi et al. 1999), and homozygotes from two strong lines (#1–2-B and #2–2) and a weak line (#4–1), all with a single-locus T-DNA insertion, were chosen for analysis. A 35S::LFY transgenic line (DW151.2.5L), previously obtained from D. Weigel (currently at Max Planck Institute, Germany) and used for constructing 35S::FT; 35S::LFY in the previous work (Kobayashi et al. 1999), was used for making 35S::TSF; 35S::LFY plants. tsf-1 was crossed with ft-1 (Col) twice and double mutant plants were identified in the F2 populations. A pFT::GUS line (#6–16) was described previously (Takada and Goto 2003). tfl1-17 is an RNA null allele (Kobayashi et al. 1999, Mimida et al. 2001). fwa-101D is a newly identified epiallele of FWA and will be described elsewhere. co-2, ft-2, fwa-2, fca-1 and ebs-1 are in the Ler background, and phyB-9, tfl2-1 and tfl1-17 are in the Col background. For analysis of flowering time phenotype, plants were grown on soil at 22°C under LD (16 h light/8 h dark) conditions with white fluorescent lights (∼60 µmol m–2 s–1) or SD (8 h light/16 h dark) conditions with white fluorescent lights (∼100 µmol m–2 s–1). For expression analysis, plants were grown on Murashige and Skoog (MS) medium or 1/2 MS medium supplemented with 1 or 0.5% sucrose. Seeds were stratified by keeping at 4°C for 2–4 d and then transferred to 22°C, which was defined as day 0. Dexamethasone treatment 35S::CO:GR; co-2 plants were grown for 7 d under constant light (CL) conditions on MS medium with 1% sucrose. Seedlings were submerged for 30 min in water containing 1 µM Dex (Wako, Osaka, Japan) and/or 10 µM Cyc (Nacalai, Kyoto, Japan), after which excess liquid was removed. RNA was extracted from seedlings harvested at 0.5, 1, 2 and 9 h after the start of the treatment. The effectiveness of Cyc treatment was confirmed by comparing patterns of GUS staining (see below for procedure) of pFT::GUS; 35S::CO:GR seedlings after Dex treatment without Cyc, Dex treatment with Cyc and mock treatment without Cyc (see S-Fig. 4A). Vernalization treatment Seeds were sown on MS or 1/2 MS medium supplemented with 1 or 0.5% sucrose and incubated at 4°C under CL conditions for 4 weeks. Seeds germinated during this incubation period. Seedlings were then transferred to 22°C under LD conditions. RT–PCR analysis RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and was treated with RNase-free DNase (Invitrogen) according to the manufacturer’s instructions. Total RNA (0.5 µg) was reverse-transcribed in a 20 µl reaction mixture using Superscript II (Invitrogen). After reaction, the mixture was diluted with 30 µl of water, and 1 µl aliquots were analyzed. Primers used in this study were as follows: TSF, 5′-GAGTCCAAGCAACCCTCACCAA-3′, 5′-CACAATACGATGAATTCCCGAG-3′; FT, 5′-TACGAAAATCCAAGTCCCACTG-3′, 5′-AAACTCGCGAGTGTTGAAGTTC-3′; mTUB2, 5′-CTCAAGAGGTTCTCAGCAGTA-3′, 5′-TCACCTTCTTCATCCGCAGTT-3′; CO, 5′-GACCACTCTACTCACCACCAAAG-3′, 5′-CAACCTCCTTGGCATCCTTATC-3′; AP1, 5′-GCACATCCGCACTAGAAAAAA-3′, 5′-CTTCTTGATACAGACCACCC-3′; FLC, 5′-TTAGTATCTCCGGCGACTTGAACCCAAACC-3′, 5′-AGATTCTCAACAAGCTTCAACATGAGTTCG-3′; SOC1, 5′-TGAGGCATACTAAGGATCGAG-3′, 5′-GCGTCTCTACTTCAGAACTTGGGC-3′; AGL24, 5′-TCCATCGAAGTCAACTCTGCTGGATC-3′, 5′-GTCTTCATGCAAGTAACATCAAC-3′; LFY, 5′-AAGGTTTCACGAGTGGCTTATTCCG-3′, 5′-GATGCTCCCTCTGTCTCTCTGTCC-3′. PCR products were electrophoresed on an agarose or a polyacrylamide gel, and visualized by ethidium bromide staining or Southern blotting. Southern blotting was performed with AlkPhos Direct (Amersham, Piscataway, NJ, USA) according to the manufacturer’s protocol. Probes were made from fragments amplified with the same primer sets for each gene. Blot images were quantified using the public domain NIH Image program for Fig. 3B and C. Plasmid construction and transgenic plants To construct gTSF::GUS, the GUS coding sequence from pBI 101 was inserted into the 4.8 kb TSF genomic fragment (including a 1.5 kb sequence upstream of ATG and a 1.36 kb sequence downstream of the stop codon) cloned from Ler by PCR. Because TSF in ecotype Ler lacks insertion sequences, including a 1 kb insertion in the 3′-UTR and a 5 kb copia-like retroelement (At4g20365) in the intergenic region between TSF (At4g20370) and At4g20360 (TUFA) found in Col and other ecotypes (see Supplementary material for natural variations), we have chosen Ler as a source of the TSF sequence for the gTSF::GUS construct. The stop codon of TSF was replaced with the GUS open reading frame and a linker sequence in front of ATG. This chimeric gene on the binary vector pBI 101 was introduced into Agrobacterium strain pMP90 and transformed into Col plants by the floral-dip procedure (Clough and Bent 1998). Histological analysis of GUS staining gTSF::GUS lines #21–1, #8–1 and #39–1 and pFT::GUS line #6–16 (Takada and Goto 2003) were chosen for analysis. In vernalization experiments, F1 plants between FRI-Sf2 (Col) and gTSF (#21–1 and #39–1) or pFT::GUS were compared with gTSF or pFT::GUS parental lines. Samples were collected at ZT14 for LD-grown plants or at ZT7 for SD-grown plants. For GUS staining, tissues were incubated for 15 min in fixing solution (0.3% formaldehyde, 50 mM sodium phosphate buffer, pH 7.2, 0.2% Triton X-100), rinsed with water, infiltrated with staining solution (0.5 mg ml–1 X-Gluc, 50 mM sodium phosphate buffer, pH 7.2, 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 0.2% Triton X-100) under vacuum for 5 min and incubated at 37°C for about 48 h in the dark. After staining, for whole-mount observation, samples were cleared in a mixture of chloral hydrate, glycerol and water solution (8 g : 1 ml : 2 ml). For sectioning, samples were dehydrated through an ethanol series, embedded in Technovit 7100 (Heraeus Kulzer, Wehrheim, Germany) and sectioned at a thickness of 4 µm with a microtome. Supplementary material Supplementary material mentioned in the article is available to online subscribers at the journal website www.pcp.oupjournals.org. Acknowledgments We thank J. H. Ahn, R. Amasino, G. Coupland and Plant Bioscience Ltd, J. Martínez-Zapater, A. Nagatani, D. Weigel, and the NSF-supported ABRC (Ohio, USA) for materials, T. Izawa for communicating unpublished results, J. H. Ahn for discussion, Y. Tomita for excellent technical assistance and analysis of natural variations, S. Yamamoto and Y. Ikeda for help in RNA extraction, and members of the Araki lab for comments and advice. Supported by a Grant-in-Aid for Scientific Research on Priority Areas (to T.A.), a Grant for Scientific Research (B) (to T.A.), a Grant for Biodiversity Research of the 21st Century COE (A14) from MEXT, Japan, a grant by the CREST program of the Japan Science and Technology Agency (to T.A.), and a grant by PROBRAIN (to M.A.). A.Y. was supported in part by a Grant for Biodiversity Research of the 21st Century COE (A14) from MEXT, Japan. 5 Present address: Department of Molecular Biology, Max Planck Institute for Developmental Biology, D-72076 Tübingen, Germany The nucleotide sequences reported in this paper have been submitted to DDBJ under accession numbers AB204805–AB204809. View largeDownload slide Fig. 1 Isolation of a tsf-1 allele. (A) TSF gene structure in ecotype Col and position of the T-DNA insertion (represented by a box on an inverted triangle, not in scale) in tsf-1 (SALK_087522). Boxes and lines indicate exons and introns of TSF; gray and yellow boxes represent untranslated regions and coding regions, respectively. Orange flags represent positions of the primer set used in (B) and (C). Analysis of the sequence flanking the T-DNA left border (filled box) and successful PCR amplification with primers shown with blue flags indicated that there was no large deletion associated with the T-DNA insertion. In ecotype Ws, an A to T substitution at the splicing acceptor site (ag) of the second intron was found. (B) RT–PCR analysis showing that tsf-1 had reduced mRNA levels. Upper and lower panels of TSF represent Southern blot images after 24 and 28 cycles of amplification by PCR, respectively. (C) Expression levels of TSF among three representative ecotypes. Plants were grown on MS medium in LD conditions and harvested at ZT14 on day 7 (B, C). (D) Phenotype of tsf-1 plants. From left to right: wild type (Col), tsf-1, ft-1 and tsf-1; ft-1 plants grown for 8 weeks under CL conditions at 22°C. View largeDownload slide Fig. 1 Isolation of a tsf-1 allele. (A) TSF gene structure in ecotype Col and position of the T-DNA insertion (represented by a box on an inverted triangle, not in scale) in tsf-1 (SALK_087522). Boxes and lines indicate exons and introns of TSF; gray and yellow boxes represent untranslated regions and coding regions, respectively. Orange flags represent positions of the primer set used in (B) and (C). Analysis of the sequence flanking the T-DNA left border (filled box) and successful PCR amplification with primers shown with blue flags indicated that there was no large deletion associated with the T-DNA insertion. In ecotype Ws, an A to T substitution at the splicing acceptor site (ag) of the second intron was found. (B) RT–PCR analysis showing that tsf-1 had reduced mRNA levels. Upper and lower panels of TSF represent Southern blot images after 24 and 28 cycles of amplification by PCR, respectively. (C) Expression levels of TSF among three representative ecotypes. Plants were grown on MS medium in LD conditions and harvested at ZT14 on day 7 (B, C). (D) Phenotype of tsf-1 plants. From left to right: wild type (Col), tsf-1, ft-1 and tsf-1; ft-1 plants grown for 8 weeks under CL conditions at 22°C. View largeDownload slide Fig. 2 Expression levels of TSF in various flowering-time mutants. (A) Effects of late flowering mutations in LD conditions. FRI-Sf2: FRI-Sf2 (Col). Plants were harvested at ZT14 on day 7. (B) Effect of phyB-9 mutation in LD conditions. Plants were harvested at ZT14 on day 5. (C) Effect of ebs-1 mutation in SD conditions. (D) Effect of tfl2-1 mutation in SD conditions. Plants were harvested at ZT7 (C) or ZT8 (D) on day 14. Plants were grown on MS medium except for the left panel of A (on 1/2 MS). View largeDownload slide Fig. 2 Expression levels of TSF in various flowering-time mutants. (A) Effects of late flowering mutations in LD conditions. FRI-Sf2: FRI-Sf2 (Col). Plants were harvested at ZT14 on day 7. (B) Effect of phyB-9 mutation in LD conditions. Plants were harvested at ZT14 on day 5. (C) Effect of ebs-1 mutation in SD conditions. (D) Effect of tfl2-1 mutation in SD conditions. Plants were harvested at ZT7 (C) or ZT8 (D) on day 14. Plants were grown on MS medium except for the left panel of A (on 1/2 MS). View largeDownload slide Fig. 3 Regulation of TSF expression by CO and photoperiods. (A) Up-regulation of TSF expression by activation of the CO–GR fusion protein. Numbers indicate hours after application of dexamethasone (+Dex) or the solvent (–Dex) in the presence of cycloheximide (+Cyc). (B) Diurnal oscillation of TSF mRNA levels in LD or SD conditions. Plants were harvested from ZT0 on day 7 to ZT4 on day 8 at 2 h intervals. Southern blot images and the quantified data are shown. The vertical axis shows relative mRNA levels of TSF and FT normalized by those of a β-tubulin gene (TUB2). White and black bars represent light and dark periods, respectively. (C) Response of TSF mRNA levels to SD to LD shift. Plants were entrained in SD conditions, and were shifted from SD to LD on day 7 by extending the light period. Seedlings were harvested from ZT1 on day 7 to ZT1 on day 9 at 3 h intervals. Relative mRNA levels of TSF, FT and CO were normalized by TUB2. White and black bars represent light and dark periods, respectively. The extended light period in the subjective night on day 7 is shown with a gray bar. Plants were grown on MS (A, C) or 1/2 MS (B) media at 22°C. View largeDownload slide Fig. 3 Regulation of TSF expression by CO and photoperiods. (A) Up-regulation of TSF expression by activation of the CO–GR fusion protein. Numbers indicate hours after application of dexamethasone (+Dex) or the solvent (–Dex) in the presence of cycloheximide (+Cyc). (B) Diurnal oscillation of TSF mRNA levels in LD or SD conditions. Plants were harvested from ZT0 on day 7 to ZT4 on day 8 at 2 h intervals. Southern blot images and the quantified data are shown. The vertical axis shows relative mRNA levels of TSF and FT normalized by those of a β-tubulin gene (TUB2). White and black bars represent light and dark periods, respectively. (C) Response of TSF mRNA levels to SD to LD shift. Plants were entrained in SD conditions, and were shifted from SD to LD on day 7 by extending the light period. Seedlings were harvested from ZT1 on day 7 to ZT1 on day 9 at 3 h intervals. Relative mRNA levels of TSF, FT and CO were normalized by TUB2. White and black bars represent light and dark periods, respectively. The extended light period in the subjective night on day 7 is shown with a gray bar. Plants were grown on MS (A, C) or 1/2 MS (B) media at 22°C. View largeDownload slide Fig. 4 Effects of TSF overexpression. (A) Phenotype of TSF-overexpressing plants. From left to right: wild type (Col), 35S::FT #11–1, 35S::TSF #1–2-B and 35S::TSF #2–2 plants grown for 18 d in CL conditions. (B) Phenotype of 35S::TSF; tfl1-17 plants. Left, 35S::TSF #1–2-B; right, 35S::TSF #1–2-B; tfl1-17 grown for 13 d in LD conditions. (C) Axillary shoots of the first two leaves in 35S::TSF; tfl1-17 which are converted to single flowers. (D–F) Phenotype of 35S::TSF #2–2; 35S::LFY plants grown for 14 d in LD conditions. (F) A close-up of the primary shoot of the plant in (E), which is converted into a single terminal flower. One cotyledon was removed to reveal the short pedicel (arrowhead). Plants were grown on soil (A) or on 1/2 MS medium (B–F). Scale bars: 1 cm in (A), 0.5 cm in (B), 1 mm in (C) and (E), 2 mm in (D) and 0.5 mm in (F). (G) Effects of TSF and FT overexpression on the expression of SOC1, AGL24, LFY and AP1. Plants were grown on MS medium in LD conditions and harvested at ZT14 on days 4 and 7. View largeDownload slide Fig. 4 Effects of TSF overexpression. (A) Phenotype of TSF-overexpressing plants. From left to right: wild type (Col), 35S::FT #11–1, 35S::TSF #1–2-B and 35S::TSF #2–2 plants grown for 18 d in CL conditions. (B) Phenotype of 35S::TSF; tfl1-17 plants. Left, 35S::TSF #1–2-B; right, 35S::TSF #1–2-B; tfl1-17 grown for 13 d in LD conditions. (C) Axillary shoots of the first two leaves in 35S::TSF; tfl1-17 which are converted to single flowers. (D–F) Phenotype of 35S::TSF #2–2; 35S::LFY plants grown for 14 d in LD conditions. (F) A close-up of the primary shoot of the plant in (E), which is converted into a single terminal flower. One cotyledon was removed to reveal the short pedicel (arrowhead). Plants were grown on soil (A) or on 1/2 MS medium (B–F). Scale bars: 1 cm in (A), 0.5 cm in (B), 1 mm in (C) and (E), 2 mm in (D) and 0.5 mm in (F). (G) Effects of TSF and FT overexpression on the expression of SOC1, AGL24, LFY and AP1. Plants were grown on MS medium in LD conditions and harvested at ZT14 on days 4 and 7. View largeDownload slide Fig. 5 Spatial patterns of TSF expression. (A–K) and (N–U) Whole-mount analyses of GUS expression patterns of gTSF::GUS (A, C, E, G, I–K and N–U) and pFT::GUS (B, D, F and H). Cotyledons of LD-grown seedlings (A and B) and SD-grown seedlings (C and D) on day 6. The first leaves of LD-grown seedlings on days 8 (E and F) and 10 (G and H). Insets in (E) and (G) show higher magnification images of the apical region of the leaves. Hypocotyls of LD-grown seedlings on days 6 (I), 8 (J), 10 (K) and 12 (N) and SD-grown seedlings on days 6 (O) and 12 (P). Shoot apical region of LD-grown seedlings on days 6 (Q), 8 (R) and 10 (S) and SD-grown seedlings on days 12 (T) and 28 (U). Q, R, T and U are side views and S is the view from above. Arrowheads in S indicate vascular bundles to cotyledons. (L), (M) and (V–X) Histological analysis of gTSF::GUS expression patterns. A cross-section of hypocotyl of LD-grown plants at day 10 (L). The inside region of the frame in (L) is magnified in (M). Longitudinal sections (V and W) and a cross-section (X) through SAM of 6-day-old gTSF::GUS plants grown in LD conditions. For gTSF::GUS, line #21–1 was used except for (W) (#8–1). Scale bars: 0.5 mm in (A–H) and (S), 0.2 mm in (I–K), (N–R), (T) and (U), 50 µm in (L) and (V–X), and 10 µm in (M). (Y) RT–PCR analysis of expression in LD-grown seedlings on day 7; SA, shoot apex; RL, rosette leaves; Co, cotyledons; Hy, hypocotyls; Ro, roots. (Za–Zf) Effects of TSF and FT overexpression on gTSF::GUS and pFT::GUS expression. F1 plants of gTSF::GUS×Col (Za), gTSF::GUS×35S::FT #11–1 (Zb), gTSF::GUS×35S::TSF #2–2 (Zc), pFT::GUS×Col (Zd), pFT::GUS×35S::FT #11–1 (Ze) and pFT::GUS×35S::TSF #2–2 (Zf). Plants were grown in LD conditions and subjected to GUS staining at ZT14 on day 4. Scale bar: 1 mm. View largeDownload slide Fig. 5 Spatial patterns of TSF expression. (A–K) and (N–U) Whole-mount analyses of GUS expression patterns of gTSF::GUS (A, C, E, G, I–K and N–U) and pFT::GUS (B, D, F and H). Cotyledons of LD-grown seedlings (A and B) and SD-grown seedlings (C and D) on day 6. The first leaves of LD-grown seedlings on days 8 (E and F) and 10 (G and H). Insets in (E) and (G) show higher magnification images of the apical region of the leaves. Hypocotyls of LD-grown seedlings on days 6 (I), 8 (J), 10 (K) and 12 (N) and SD-grown seedlings on days 6 (O) and 12 (P). Shoot apical region of LD-grown seedlings on days 6 (Q), 8 (R) and 10 (S) and SD-grown seedlings on days 12 (T) and 28 (U). Q, R, T and U are side views and S is the view from above. Arrowheads in S indicate vascular bundles to cotyledons. (L), (M) and (V–X) Histological analysis of gTSF::GUS expression patterns. A cross-section of hypocotyl of LD-grown plants at day 10 (L). The inside region of the frame in (L) is magnified in (M). Longitudinal sections (V and W) and a cross-section (X) through SAM of 6-day-old gTSF::GUS plants grown in LD conditions. For gTSF::GUS, line #21–1 was used except for (W) (#8–1). Scale bars: 0.5 mm in (A–H) and (S), 0.2 mm in (I–K), (N–R), (T) and (U), 50 µm in (L) and (V–X), and 10 µm in (M). (Y) RT–PCR analysis of expression in LD-grown seedlings on day 7; SA, shoot apex; RL, rosette leaves; Co, cotyledons; Hy, hypocotyls; Ro, roots. (Za–Zf) Effects of TSF and FT overexpression on gTSF::GUS and pFT::GUS expression. F1 plants of gTSF::GUS×Col (Za), gTSF::GUS×35S::FT #11–1 (Zb), gTSF::GUS×35S::TSF #2–2 (Zc), pFT::GUS×Col (Zd), pFT::GUS×35S::FT #11–1 (Ze) and pFT::GUS×35S::TSF #2–2 (Zf). Plants were grown in LD conditions and subjected to GUS staining at ZT14 on day 4. Scale bar: 1 mm. View largeDownload slide Fig. 6 Effect of vernalization on TSF and FT expression. (A–I) Whole-mount analyses of GUS expression patterns of gTSF::GUS #21–1(A–C), gTSF::GUS #39–1 (D–F) and pFT::GUS (G–I). A, D and G are in the Col background. B, C, E, F, H and I are in the FRI-Sf2 (Col) background. Seedlings were grown in LD conditions for 10 d without vernalization (A, B, D, E, G and H) or after 4 weeks vernalization (C, F and I). (J) RT–PCR analysis of TSF and FT expression in fca-1 with vernalization treatment. Plants were grown on 1/2 MS medium with (+Vern) or without (–Vern) vernalization treatment and harvested at day 7 after being transferred to 22°C. Scale bars: 0.2 mm in (A–F) and 1 mm in (G–I). View largeDownload slide Fig. 6 Effect of vernalization on TSF and FT expression. (A–I) Whole-mount analyses of GUS expression patterns of gTSF::GUS #21–1(A–C), gTSF::GUS #39–1 (D–F) and pFT::GUS (G–I). A, D and G are in the Col background. B, C, E, F, H and I are in the FRI-Sf2 (Col) background. Seedlings were grown in LD conditions for 10 d without vernalization (A, B, D, E, G and H) or after 4 weeks vernalization (C, F and I). (J) RT–PCR analysis of TSF and FT expression in fca-1 with vernalization treatment. Plants were grown on 1/2 MS medium with (+Vern) or without (–Vern) vernalization treatment and harvested at day 7 after being transferred to 22°C. Scale bars: 0.2 mm in (A–F) and 1 mm in (G–I). View largeDownload slide Fig. 7 A model for the integration of regulatory pathways of flowering in Arabidopsis. Four main (photoperiod, vernalization, autonomous and gibberellin) and additional (‘light quality’ and ‘repression’) pathways are shown with representative genes. TSF is placed as a floral pathway integrator with FT, SOC1 and LFY. Regulation of floral pathway integrators by input pathways and regulatory relationships among floral pathway integrators is depicted. TFL1 and FWA (mentioned in the text) were omitted from the model for simplicity. Regulation at the transcriptional level is represented by solid arrows (promotion) and T-bars (repression). Arrows and T-bars to or from a dotted box with TSF and FT in it mean similar regulation of or by the two genes, while a T-bar from TFL2 to FT alone means regulation of only FT. Dotted arrows mean modulation of activity of the gene products or the machinery. Gray arrows represent regulatory relationships that are not directly analyzed in the present work. Because the role of the gibberellin pathway in the regulation of FT remains controversial (Moon et al. 2003, Piñeiro et al. 2003), we tentatively assign an independent position to EBS, which was placed to act through the gibberellin pathway (Piñeiro et al. 2003, Boss et al. 2004). The thickness of lines from CO and FLC is intended as a speculative measure of the strength of promotion and repression. Similarly, the thickness of shaded arrows from floral pathway integrators shows the relative contribution of each integrator to the floral transition. Phloem is the site of the pathway integration for FT and TSF (see text for detail). View largeDownload slide Fig. 7 A model for the integration of regulatory pathways of flowering in Arabidopsis. Four main (photoperiod, vernalization, autonomous and gibberellin) and additional (‘light quality’ and ‘repression’) pathways are shown with representative genes. TSF is placed as a floral pathway integrator with FT, SOC1 and LFY. Regulation of floral pathway integrators by input pathways and regulatory relationships among floral pathway integrators is depicted. TFL1 and FWA (mentioned in the text) were omitted from the model for simplicity. Regulation at the transcriptional level is represented by solid arrows (promotion) and T-bars (repression). Arrows and T-bars to or from a dotted box with TSF and FT in it mean similar regulation of or by the two genes, while a T-bar from TFL2 to FT alone means regulation of only FT. Dotted arrows mean modulation of activity of the gene products or the machinery. Gray arrows represent regulatory relationships that are not directly analyzed in the present work. Because the role of the gibberellin pathway in the regulation of FT remains controversial (Moon et al. 2003, Piñeiro et al. 2003), we tentatively assign an independent position to EBS, which was placed to act through the gibberellin pathway (Piñeiro et al. 2003, Boss et al. 2004). The thickness of lines from CO and FLC is intended as a speculative measure of the strength of promotion and repression. Similarly, the thickness of shaded arrows from floral pathway integrators shows the relative contribution of each integrator to the floral transition. Phloem is the site of the pathway integration for FT and TSF (see text for detail). Table 1 Flowering time of tsf mutants under LD and SD conditions Genotype  No. of rosette leaves  No. of cauline leaves  n  Experiment 1  Long day  Wild type (Col)  12.0 ± 1.2  3.2 ± 0.7  23  tsf-1  13.9 ± 2.7  2.5 ± 0.8  27  ft-1  44.9 ± 8.1†  7.2 ± 1.2  18  ft-1; tsf-1  54.4 ± 6.6†  9.6 ± 1.5  14  Short day  Wild type (Col)  35.1 ± 5.5  7.1 ± 1.2  21  tsf-1  48.3 ± 4.7  8.0 ± 1.9  15  ft-1  44.0 ± 5.7‡  6.8 ± 1.0  18  ft-1; tsf-1  53.0 ± 5.1‡  10.5 ± 0.9  8  Experiment 2  Long day  Wild type (Col)  11.9 ± 1.2  2.5 ± 0.5  11  tsf-1  11.9 ± 1.8  2.9 ± 0.5  15  tfl1-17  7.2 ± 1.5  1.3 ± 0.5  11  tfl1-17; tsf-1  7.9 ± 1.6  1.0 ± 0.5  15  Genotype  No. of rosette leaves  No. of cauline leaves  n  Experiment 1  Long day  Wild type (Col)  12.0 ± 1.2  3.2 ± 0.7  23  tsf-1  13.9 ± 2.7  2.5 ± 0.8  27  ft-1  44.9 ± 8.1†  7.2 ± 1.2  18  ft-1; tsf-1  54.4 ± 6.6†  9.6 ± 1.5  14  Short day  Wild type (Col)  35.1 ± 5.5  7.1 ± 1.2  21  tsf-1  48.3 ± 4.7  8.0 ± 1.9  15  ft-1  44.0 ± 5.7‡  6.8 ± 1.0  18  ft-1; tsf-1  53.0 ± 5.1‡  10.5 ± 0.9  8  Experiment 2  Long day  Wild type (Col)  11.9 ± 1.2  2.5 ± 0.5  11  tsf-1  11.9 ± 1.8  2.9 ± 0.5  15  tfl1-17  7.2 ± 1.5  1.3 ± 0.5  11  tfl1-17; tsf-1  7.9 ± 1.6  1.0 ± 0.5  15  The number of leaves is presented as the average ± SD. There was a statistically significant difference (Student’s t-test, P < 0.002) between two pairs of genotypes marked with † or ‡, respectively. View Large Table 2 Flowering time of 35S::TSF plants under LD and SD conditions Genotype  No. of rosette leaves  No. of cauline leaves  n  Long day  Wild type (Col)  12.1 ± 1.6  3.2 ± 0.7  16  35S::FT #11–1  2.0 ± 0  1.7 ± 0.5  18  35S::TSF #1–2-B  2.1 ± 0.3  1.1 ± 0.3  16  35S::TSF #2–2  2.1 ± 0.5  1.5 ± 0.5  17  35S::TSF #4–1  5.9 ± 0.9  1.2 ± 0.4  13  Short day  Wild type (Col)  35.5 ± 3.0  6.5 ± 0.5  11  35S::FT #11–1  2.0 ± 0  1.6 ± 0.5  16  35S::TSF #1–2-B  2.0 ± 0  1.2 ± 0.4  16  35S::TSF #2–2  2.0 ± 0  1.5 ± 0.5  16  35S::TSF #4–1  6.4 ± 0.9  1.5 ± 0.6  16  Genotype  No. of rosette leaves  No. of cauline leaves  n  Long day  Wild type (Col)  12.1 ± 1.6  3.2 ± 0.7  16  35S::FT #11–1  2.0 ± 0  1.7 ± 0.5  18  35S::TSF #1–2-B  2.1 ± 0.3  1.1 ± 0.3  16  35S::TSF #2–2  2.1 ± 0.5  1.5 ± 0.5  17  35S::TSF #4–1  5.9 ± 0.9  1.2 ± 0.4  13  Short day  Wild type (Col)  35.5 ± 3.0  6.5 ± 0.5  11  35S::FT #11–1  2.0 ± 0  1.6 ± 0.5  16  35S::TSF #1–2-B  2.0 ± 0  1.2 ± 0.4  16  35S::TSF #2–2  2.0 ± 0  1.5 ± 0.5  16  35S::TSF #4–1  6.4 ± 0.9  1.5 ± 0.6  16  The number of leaves is presented as the average ± SD. View Large Table 3 Effects of photoperiod pathway mutations on flowering time of 35S::TSF plants Genotype  No. of rosette leaves  No. of cauline leaves  n  Long day  Wild type (Col)  12.8 ± 1.2  3.9 ± 0.5  12  35S::TSF #2–2  2.7 ± 0.5  1.1 ± 0.6  30  35S::TSF #2–2; co-1  2.0 ± 0  1.0 ± 0  30  35S::TSF #2–2; ft-1  2.6 ± 0.5  0.8 ± 0.4  27  Short day  Wild type (Col)  50.7 ± 5.3  9.3 ± 1.2  15  35S::TSF #2–2  2.6 ± 0.5  0.9 ± 0.4  27  35S::TSF #2–2; co-1  2.0 ± 0  1.0 ± 0  28  35S::TSF #2–2; ft-1  2.5 ± 0.5  0.6 ± 0.5  20  Genotype  No. of rosette leaves  No. of cauline leaves  n  Long day  Wild type (Col)  12.8 ± 1.2  3.9 ± 0.5  12  35S::TSF #2–2  2.7 ± 0.5  1.1 ± 0.6  30  35S::TSF #2–2; co-1  2.0 ± 0  1.0 ± 0  30  35S::TSF #2–2; ft-1  2.6 ± 0.5  0.8 ± 0.4  27  Short day  Wild type (Col)  50.7 ± 5.3  9.3 ± 1.2  15  35S::TSF #2–2  2.6 ± 0.5  0.9 ± 0.4  27  35S::TSF #2–2; co-1  2.0 ± 0  1.0 ± 0  28  35S::TSF #2–2; ft-1  2.5 ± 0.5  0.6 ± 0.5  20  The number of leaves is presented as the average ± SD. View Large Table 4 Flowering time of Fl plants between fwa or FRI-Sf2 (Col) and 35S::TSF Genotype  No. of rosette leaves  No. of cauline leaves  n  Wild type (Col)  10.2 ± 1.2  2.6 ± 0.5  10  35S::TSF #1–2-B×Col (F1)  2.6 ± 0.5  1.1 ± 0.5  15  35S::TSF #2–2×Col (F1)  2.8 ± 0.4  1.1 ± 0.3  14  35S::TSF #1–2-B×fwa-101D (F1)  2.7 ± 0.5  0.9 ± 0.3  10  35S::TSF #2–2×fwa-101D (F1)  2.9 ± 0.5  0.9 ± 0.3  18  35S::TSF #1–2-B×FRI-Sf2 (Col) (F1)  3.9 ± 0.8  1.1 ± 0.3  12  35S::TSF #2–2×FRI-Sf2 (Col) (F1)  5.1 ± 0.9  1.1 ± 0.5  15  fwa-101D×Col (F1)  22.4 ± 2.3  5.7 ± 1.2  10  FRI-Sf2 (Col)×Col (F1)  >63 a  ND  5  Genotype  No. of rosette leaves  No. of cauline leaves  n  Wild type (Col)  10.2 ± 1.2  2.6 ± 0.5  10  35S::TSF #1–2-B×Col (F1)  2.6 ± 0.5  1.1 ± 0.5  15  35S::TSF #2–2×Col (F1)  2.8 ± 0.4  1.1 ± 0.3  14  35S::TSF #1–2-B×fwa-101D (F1)  2.7 ± 0.5  0.9 ± 0.3  10  35S::TSF #2–2×fwa-101D (F1)  2.9 ± 0.5  0.9 ± 0.3  18  35S::TSF #1–2-B×FRI-Sf2 (Col) (F1)  3.9 ± 0.8  1.1 ± 0.3  12  35S::TSF #2–2×FRI-Sf2 (Col) (F1)  5.1 ± 0.9  1.1 ± 0.5  15  fwa-101D×Col (F1)  22.4 ± 2.3  5.7 ± 1.2  10  FRI-Sf2 (Col)×Col (F1)  >63 a  ND  5  The number of leaves is presented as the average ± SD. ND, not determined. a Visible leaves were counted when plants had been grown for 4 months after sowing. Plants were grown under LD conditions. View Large Abbreviations CL constant light CO CONSTANS Col Columbia-0 Cyc cycloheximide Dex dexamethasone GR glucocorticoid receptor GUS β-glucuronidase LD long day Ler Landsberg erecta RT–PCR reverse transcription–PCR 35S cauliflower mosaic virus 35S RNA promoter SAM shoot apical meristem SD short day 3′-UTR 3′-untranslated region Ws Wassilewskija ZT Zeitgeber time References Abe, M., Kobayashi, Y., Yamamoto, S., Daimon, Y., Yamaguchi, A., Ikeda, Y., Ichinoki, H., Notaguchi, M., Goto, K. and Araki, T. ( 2005). FD, a bZIP protein mediating signals from the floral pathway integrator FT at the shoot apex. Science  (in press). Google Scholar Alonso, J.M., Stepanova, A.N., Leisse, T.J., Kim, C.J., Chen, H. et al. ( 2003) Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science  301: 653–657. Google Scholar An, H., Roussot, C., Suárez-López, P., Corbesier, L., Vincent, C., et al. ( 2004) CONSTANS acts in the phloem to regulate a systemic signal that induces photoperiodic flowering of Arabidopsis. Development  131: 3615–3626. Google Scholar Araki, T. ( 2001) Transition from vegetative to reproductive phase. Curr. Opin. Plant Biol.  4: 63–68. Google Scholar Blázquez, M.A., Ahn, J.H. and Weigel, D. ( 2003) A thermosensory pathway controlling flowering time in Arabidopsis thaliana. Nat. Genet.  33: 168–171. Google Scholar Blázquez, M.A. and Weigel, D. ( 1999) Independent regulation of flowering by phytochrome B and gibberellins in Arabidopsis. Plant Physiol.  120: 1025–1032. Google Scholar Boss, P.K., Bastow, R.M., Mylne, J.S. and Dean, C. ( 2004) Multiple pathways in the decision to flower: enabling, promoting, and resetting. Plant Cell  16: S18–S31. Google Scholar Bradley, D., Ratcliffe, O., Vincent, C., Carpenter, R. and Coen, E. ( 1997) Inflorescence commitment and architecture in Arabidopsis. Science  275: 80–83. Google Scholar Brunner, A.M., Busov, V.B. and Strauss, S.H. ( 2003) Poplar genome sequence: functional genomics in an ecologically dominant plant species. Trends Plant Sci.  9: 49–56. Google Scholar Cerdán, P.D. and Chory, J. ( 2003) Regulation of flowering time by light quality. Nature  423: 881–885. Google Scholar Clough, S.J. and Bent, A.F. ( 1998) Floral dip: a simplified method for Arabidopsis thaliana. Plant J.  16: 735–743. Google Scholar Endo, M., Nakamura, S., Araki, T., Mochizuki, N, and Nagatani, A. ( 2005) Phytochrome B in the mesophyll delays flowering by suppressing FLOWERING LOCUS T expression in Arabidopsis vascular bundles. Plant Cell  17: 1941–1952. Google Scholar Gaudin, V., Libault, M., Pouteau, S., Juul, T., Zhao, G., Lefebvre, D. and Grandjean, O. ( 2001) Mutations in LIKE HETEROCHROMATIN PROTEIN 1 affect flowering time and plant architecture in Arabidopsis. Development  128: 4847–4858. Google Scholar Gómez-Mena, C., Piñeiro, M., Franco-Zorrilla, J.M., Salinas, J., Coupland, G. and Martínez-Zapater, J.M. ( 2001) early bolting in short days: an Arabidopsis mutation that causes early flowering and partially suppresses the floral phenotype of leafy. Plant Cell  13: 1011–1024. Google Scholar Foucher, F., Morin, J., Courtiade, J., Cadioux, S., Ellis, N., Banfield, M.J. and Rameau, C. ( 2003) DETERMINATE and LATE FLOWERING are two TERMINAL FLOWER1/CENTRORADIALIS homologs that control two distinct phases of flowering initiation and development in pea. Plant Cell  15: 2742–2754. Google Scholar Halliday, K.J., Salter, M.G., Thingnaes, E. and Whitelam, G.C. ( 2003) Phytochrome control of flowering is temperature sensitive and correlates with expression of the floral integrator FT. Plant J.  33: 875–885. Google Scholar Hayama, R., Yokoi, S., Tamaki, S., Yano, M. and Shimamoto, K. ( 2003) Adaptation of photoperiodic control pathways produces short-day flowering in rice. Nature  422: 719–722. Google Scholar Hepworth, S.R., Valverde, F., Ravenscroft, D., Mouradov, A. and Coupland, G. ( 2002) Antagonistic regulation of flowering-time gene SOC1 by CONSTANS and FLC via separate promoter motifs. EMBO J.  21: 4327–4337. Google Scholar Imaizumi, T., Tran, H.G., Swartz, T.E., Briggs, W.R. and Kay, S.A. ( 2003) FKF1 is essential for photoperiodic-specific light signalling in Arabidopsis. Nature  426: 302–306. Google Scholar Izawa, T., Oikawa, T., Sugiyama, N., Tanisaka, T., Yano, M. and Shimamoto, K. ( 2002) Phytochrome mediates the external light signal to repress FT orthologs in photoperiodic flowering of rice. Genes Dev.  16: 2006–2020. Google Scholar Johanson, U., West, J., Lister, C., Michaels, S., Amasino, R. and Dean, C. ( 2000) Molecular analysis of FRIGIDA, a major determinant of natural variation in Arabidopsis flowering time. Science  290: 344–347. Google Scholar Kardailsky, I., Shukla, V.K., Ahn, J.H., Dagenais, N., Christensen, S.K., Nguyen, J.T., Chory, J., Harrison, M.J. and Weigel, D. ( 1999) Activation tagging of the floral inducer FT. Science  286: 1962–1965. Google Scholar Kaya, H., Sato, S., Tabata, S., Kobayashi, Y., Iwabuchi, M. and Araki, T. ( 2000) hosoba toge toge, a syndrome caused by a large chromosomal deletion associated with a T-DNA insertion in Arabidopsis. Plant Cell Physiol.  41: 1055–1066. Google Scholar Kobayashi, Y., Kaya, H., Goto, K., Iwabuchi, M. and Araki, T. ( 1999) A pair of related genes with antagonistic roles in mediating flowering signals. Science  286: 1960–1962. Google Scholar Kojima, S., Takahashi, Y., Kobayashi, Y., Monna, L., Sasaki, T., Araki, T. and Yano, M. ( 2002) Hd3a, a rice ortholog of the Arabidopsis FT gene, promotes transition to flowering downstream of Hd1 under short-day condition. Plant Cell Physiol.  43: 1096–1105. Google Scholar Koornneef, M., Hanhart, C.J. and van der Veen, J.J. ( 1991) A genetic and physiological analysis of late flowering mutants in Arabidopsis thaliana. Mol. Gen. Genet.  229: 57–66. Google Scholar Kotake, T., Takada, S., Nakahigashi, K., Ohto, M. and Goto, K. ( 2003) Arabidopsis TERMINAL FLOWER 2 gene encodes a heterochromatin protein 1 homolog and represses both FLOWERING LOCUS T to regulate flowering time and several floral homeotic genes. Plant Cell Physiol.  44: 555–564. Google Scholar Kurup, S., Jones, H.D. and Holdsworth, M.J. ( 2000) Interactions of the developmental regulator ABI3 with proteins identified from developing Arabidopsis seeds. Plant J.  21: 143–155. Google Scholar Lee, H., Suh, S.S., Park, E., Cho, E., Ahn, J.H., Kim, S.G., Lee, J.S., Kwon, Y.M. and Lee, I. ( 2000) The AGAMOUS-LIKE 20 MADS domain protein integrates floral inductive pathways in Arabidopsis. Genes Dev.  14: 2366–2376. Google Scholar Michaels, S.D. and Amasino, R.M. ( 1999) FLOWERING LOCUS C encodes a novel MADS domain protein that acts as a repressor of flowering. Plant Cell  11: 949–956. Google Scholar Michaels, S.D. and Amasino, R.M. ( 2001) Loss of FLOWERING LOCUS C activity eliminates the late-flowering phenotype of FRIGIDA and autonomous pathway mutations but not responsiveness to vernalization. Plant Cell  13: 935–941. Google Scholar Michaels, S.D., Himelblau, E., Kim, S.Y., Schomburg, F.M. and Amasino, R.M. ( 2005) Integration of flowering signals in winter-annual Arabidopsis. Plant Physiol.  137: 149–156. Google Scholar Mimida, N., Goto, K., Kobayashi, Y., Araki, T., Ahn, J.H., Weigel, D., Murata, M., Motoyoshi, F. and Sakamoto, W. ( 2001) Functional divergence of the TFL-like gene family in Arabidopsis revealed by characterization of a novel homologue. Genes Cells  6: 327–336. Google Scholar Moon, J., Lee, H., Kim, M. and Lee, I. ( 2005) Analysis of flowering pathway integrators in Arabidopsis. Plant Cell Physiol.  46: 292–299. Google Scholar Moon, J., Suh, S.S., Lee, H., Choi, K.R., Hong, C.B., Paek, N.C., Kim, S.G. and Lee, I. ( 2003) The SOC1 MADS-box gene integrates vernalization and gibberellin signals for flowering in Arabidopsis. Plant J.  35: 613–623. Google Scholar Piñeiro, M., Gómez-Mena, C., Schaffer, R., Martínez-Zapater, J.M. and Coupland, G. ( 2003) EARLY BOLTING IN SHORT DAYS is related to chromatin remodeling factors and regulates flowering in Arabidopsis by repressing FT. Plant Cell  15: 1552–1562. Google Scholar Putterill, J., Robson, F., Lee, K., Simon, R. and Coupland, G. ( 1995) The CONSTANS gene of Arabidopsis promotes flowering and encodes a protein showing similarities to zinc finger transcription factors. Cell  80: 847–857. Google Scholar Ratcliffe, O.J., Amaya, I., Vincent, C.A., Rothstein, S., Carpenter, R., Coen, E.S. and Bradley, D.J. ( 1998) A common mechanism controls the life cycle and architecture of plants. Development  125: 1609–1615. Google Scholar Ratcliffe, O.J., Bradley, D.J. and Coen, E.S. ( 1999) Separation of shoot and floral identity in Arabidopsis. Development  126: 1109–1120. Google Scholar Reeves, P.H. and Coupland, G. ( 2001) Analysis of flowering time control in Arabidopsis by comparison of double and triple mutants. Plant Physiol.  126: 1085–1091. Google Scholar Robson, F., Costa, M.M.R., Hepworth, S., Vizir, I., Piñeiro, M., Reeves, P.H., Putterill, J. and Coupland, G. ( 2001) Functional importance of conserved domains in the flowering-time gene CONSTANS demonstrated by analysis of mutant alleles and transgenic plants. Plant J.  28: 619–631. Google Scholar Ruiz-García, L., Madueño, F., Wilkinson, M., Haughn, G., Salinas, J. and Martínez-Zapater, J.M. ( 1997) Different roles of flowering-time genes in the activation of floral initiation genes in Arabidopsis. Plant Cell  9: 1921–1934. Google Scholar Samach, A., Onouchi, H., Gold, S.E., Ditta, G.S., Schwarz-Sommer, Z., Yanofsky, M.F. and Coupland, G. ( 2000) Distinct roles of CONSTANS target genes in reproductive development of Arabidopsis. Science  288: 1613–1616. Google Scholar Schmid, M., Uhlenhaut, N.H., Godard, F., Demar, M., Bressan, R., Weigel, D. and Lohmann, J.U. ( 2003) Dissection of floral induction pathway using global expression analysis. Development  130: 6001–6012. Google Scholar Shannon, S. and Meeks-Wagner, D.R. ( 1991) A mutation in the Arabidopsis TFL1 gene affects inflorescence meristem development. Plant Cell  3: 877–892. Google Scholar Sheldon, C.C., Burn, J.E., Perez, P.P., Metzger, J., Edwards, J.A., Peacock, W.J. and Dennis, E.S. ( 1999) The FLF MADS box gene: a repressor of flowering in Arabidopsis regulated by vernalization and methylation. Plant Cell  11: 445–458. Google Scholar Sheldon, C.C., Rouse, D.T., Finnegan, E.J., Peacock, W.J. and Dennis, E.S. ( 2000) The molecular basis of vernalization: the central role of FLOWERING LOCUS C (FLC). Proc. Natl Acad. Sci. USA  97: 3753–3758. Google Scholar Simon, R., Igeño, M.I. and Coupland, G. ( 1996) Activation of floral meristem identity genes in Arabidopsis. Nature  384: 59–62. Google Scholar Simpson, G.G. and Dean, C. ( 2002) Arabidopsis, the rosetta stone of flowering time? Science  296: 285–289. Google Scholar Soppe, W.J.J., Jacobsen, S.E., Alonso-Blanco, C., Jackson, J.P., Kakutani, T., Koornneef, M. and Peeters, A.J.M. ( 2000) The late flowering phenotype of fwa mutants is caused by gain-of-function epigenetic alleles of a homeodomain gene. Mol. Cell  6: 791–802. Google Scholar Suárez-López, P., Wheatley, K., Robson, F., Onouchi, H., Valverde, F. and Coupland, G. ( 2001) CONSTANS mediates between the circadian clock and the control of flowering in Arabidopsis. Nature  410: 1116–1120. Google Scholar Suzuki, M., Ketterling, M.G., Li, Q.B. and McCarty, D.R. ( 2003) Viviparous1 alters global gene expression patterns through regulation of abscisic acid signaling. Plant Physiol.  132: 1664–1677. Google Scholar Takada, S. and Goto, K. ( 2003) TERMINAL FLOWER2, an Arabidopsis homolog of HETEROCHROMATIN PROTEIN1, counteracts the activation of FLOWERING LOCUS T by CONSTANS in the vascular tissues of leaves to regulate flowering time. Plant Cell  15: 2856–2865. Google Scholar Valverde, F., Mouradov, A., Soppe, W., Ravenscroft, D., Samach, A. and Coupland, G. ( 2004) Photoreceptor regulation of CONSTANS protein in photoperiodic flowering. Science  303: 1003–1006. Google Scholar Yanovsky, M.J. and Kay, S.A. ( 2002) Molecular basis of seasonal time measurement in Arabidopsis. Nature  419: 308–312. Google Scholar Yoo, S.Y., Kardailsky, I., Lee, J.S., Weigel, D. and Ahn, J.H. ( 2004) Acceleration of flowering by overexpression of MFT (MOTHER OF FT AND TFL1). Mol. Cells  17: 95–101. Google Scholar Yu, H., Xu, Y., Tan, E.L. and Kumar, P.P. ( 2002) AGAMOUS-LIKE 24, a dosage-dependent mediator of the flowering signals. Proc. Natl Acad. Sci. USA  99: 16336–16341. Google Scholar JSPP © 2005

Journal

Plant and Cell Physiology – Oxford University Press

Published: Aug 1, 2005

There are no references for this article.