Abstract

The genetics underlying onion development are poorly understood. Here the characterization of onion homologs of Arabidopsis photoperiodic flowering pathway genes is reported with the end goal of accelerating onion breeding programs by understanding the genetic basis of adaptation to different latitudes. The expression of onion GI, FKF1 and ZTL homologs under short day (SD) and long day (LD) conditions was examined using quantitative reverse transcription–PCR (qRT–PCR). The expression of AcGI and AcFKF1 was examined in onion varieties which exhibit different daylength responses. Phylogenetic trees were constructed to confirm the identity of the homologs. AcGI and AcFKF1 showed diurnal expression patterns similar to their Arabidopsis counterparts, while AcZTL was found to be constitutively expressed. AcGI showed similar expression patterns in varieties which exhibit different daylength responses, whereas AcFKF1 showed differences. It is proposed that these differences could contribute to the different daylength responses in these varieties. Phylogenetic analyses showed that all the genes isolated are very closely related to their proposed homologs. The results presented here show that key genes controlling photoperiodic flowering in Arabidopsis are conserved in onion, and a role for these genes in the photoperiodic control of bulb initiation is predicted. This theory is supported by expression and phylogenetic data.

Introduction

The onion (Allium cepa L.) belongs to the order Asparagales, the second most important monocot order (Stevens 2001 onwards, Kuhl et al. 2004). It is a diploid plant (2n = 2x = 16) with a very large genome (32 pg/2n), about 36 and 107 times larger than the rice and Arabidopsis genomes, respectively (McCallum et al. 2001, Kuhl et al. 2004). Onions are farmed worldwide and, in 2007, 68 million tonnes of onions were produced across the world (FAOSTAT 2008). The onion is a biennial plant, the bulb being an overwintering stage of the life cycle (Lancaster et al. 1996). Flowering and seed production will occur following a period of vernalization, provided the juvenile phase has been passed (Brewster 1997). In terms of crop production, onions tend to be grown as annual crops.

The physiology of bulb initiation has been studied extensively. It is a process which is photoperiodically driven in temperate onions, drawing parallels with the photoperiodic control of flowering in other plant species (Mettananda and Fordham 1997). Long days (LDs, ≥16 h of light) will initiate bulbing in temperate onions. Commercially, onion cultivars are classified as long, short and intermediate daylength varieties (Brewster 2008). The exact daylength required will vary between cultivars, but the broad classification gives an indication of which cultivar would be suited to growth at a particular latitude. Short day (SD) onion varieties require a daylength of at least 12 h to initiate bulbing. However, these varieties perform poorly in longer daylengths as they produce a bulb after only one or two leaves, leading to a very small final product (Brewster 2008). Therefore, a daylength of 12 h is optimal for crop production.

Flowering has been well characterized at molecular and genetic levels. The flowering time genes in Arabidopsis mainly function in six different pathways: autonomous; vernalization; gibberellin; temperature; light quality; and photoperiod (Jack 2004). In photoperiodic flowering, light interacts with the circadian clock [through PHYTOCHROME (PHY) and CRYPTOCHROME (CRY) genes] and the timing of the clock is controlled by feedback loops involving TIMING OF CAB EXPRESSION 1 (TOC1). CONSTANS (CO) expression is high at the end of LDs and the CO protein is degraded at night (Valverde et al. 2004). CO regulates the expression of floral integrating genes such as FLOWERING LOCUS T (FT), leading to floral initiation (Massiah 2007, Jackson 2009). Flowering takes place when CO transcription and a blue or far-red (FR) light signal occur simultaneously. The CO gene is an integral part of this pathway and has been isolated from several species including both SD and LD plants (Griffiths et al. 2003).

GIGANTEA (GI) and FLAVIN-BINDING, KELCH REPEAT, F-BOX 1 (FKF1) have been shown to regulate CO expression. Both are circadian regulated (Fowler et al. 1999, Park et al., 1999, Nelson et al. 2000) and control flowering by regulating CO transcription through the degradation of CYCLING DOF FACTOR 1 (CDF1), a repressor of CO (Imaizumi et al. 2005, Sawa et al. 2007). Recent work has shown that additional CDF genes (CDF2, CDF3, and CDF5) act redundantly with CDF1 to repress CO (Fornara et al. 2009). GI shows no homology with any gene of known function (Fowler et al. 1999). The FKF1 protein has three characteristic domains: the light, oxygen and voltage (LOV)-sensing domain, the F-box and the Kelch repeat (Nelson et al. 2000). FKF1 has been shown to regulate the precise timing of CO expression, a function which is light dependent, leading to flowering only in LDs (Imaizumi et al. 2003). Studies have shown that GI and FKF1 form a protein complex in blue light (Sawa et al. 2007) dependent on blue light absorption within the LOV domain of FKF1 (Imaizumi et al. 2003). The complex binds to CDF1 and forms on the CO promoter, regulating CO expression. This occurs in the late afternoon in LDs, leading to CO activation of FT expression and subsequently flowering. GI has been shown additionally to regulate photoperiodic flowering through a mechanism independent of CO, which involves regulation of miR172 abundance and hence its targets, leading to activation of FT (Jung et al. 2007). GI is also involved in the maintenance of circadian rhythms (Mizoguchi et al. 2005). This is mediated through an interaction with ZEITLUPE (ZTL) under blue light conditions (Kim et al. 2007). ZTL is a circadian clock-associated protein which is very closely related to FKF1. It is involved in the control of proteosome-dependent degradation of TIMING OF CAB EXPRESSION 1 (TOC1) (Mas et al. 2003). GI is required to establish and maintain the oscillation of the ZTL protein through protein–protein interactions which are enhanced by blue light acting through the LOV domain. It has also been shown that ZTL and another member of the same gene family, LOV KELCH PROTEIN2 (LKP2), can act redundantly with FKF1 in the degradation of CDF proteins (Fornara et al. 2009).

Two further genes involved in photoperiodic flowering are CONSTITUTIVE PHOTOMORPHOGENIC 1 (COP1) and EARLY FLOWERING 3 (ELF3). COP1 and ELF3 have been shown to control flowering by regulating GI stability (Yu et al. 2008). COP1 is also involved in the degradation of CO protein at night (Jang et al. 2008).

At the physiological level, bulb initiation in LD onions is regulated in a similar way to the photoperiodic regulation of flowering in LD plants such as Arabidopsis (Thomas and Vince-Prue 1997). Bulb initiation requires the perception of light with an appreciable component of FR in the second half of the LD, implying the involvement of phytochrome (Lercari 1984, Quail et al. 1995). Perception of the LD signal takes place in the leaves (Sobeih and Wright 1986, Brewster 1990), as with flowering (Knott 1934), while the response is in the meristem, requiring the transport of a signal within the plant. Other parallels between bulb initiation and floral initiation include a juvenile phase during which plants cannot respond to daylength and the involvement of a homeotic conversion of photosynthetic leaves to either floral organs or swollen bulb scales at the responsive meristem (Sobeih and Wright 1986, Lancaster et al. 1996, Komeda 2004, Massiah 2007). Recent work has identified the FLOWERING LOCUS T (FT) protein and its homolog as a part of the mobile signal in Arabidopsis and rice (Corbesier et al. 2007, Jaeger and Wigge 2007, Tamaki et al. 2007).

This study shows that genes controlling the daylength response are conserved between the model plant Arabidopsis and onion, and supports the hypothesis that these genetic components regulate both bulbing and flowering end-processes.

Results

Conservation of flowering time genes

A clone of a putative onion CO homolog, identified in the A. cepa gene index, was assigned the name Allium cepa CO-like (AcCOL). Full-length sequence information was obtained which showed the gene to be more closely related to Arabidopsis COL4 than CO (Table 1). A phylogenetic analysis based on B-box proteins (Supplementary Fig. S1) showed that this gene is a group 1b CO-like gene (according to the groupings described by Griffiths et al. 2003). This gene did not exhibit a discernible diurnal expression pattern (Supplementary Fig. S2), as shown by both Arabidopsis CO and the rice homolog Heading date 1 (Hd1) (Suárez-López et al. 2001, Izawa et al. 2002). Attempts to clone other CO-like sequences using degenerate primers or screening a normalized cDNA library consistently resulted in isolation of the same sequence as AcCOL.

Table 1

Conservation of Arabidopsis flowering time genes in onion

Arabidopsis/ rice gene Annotation Accession No. Percentage amino acid identity (Arabidopsis) Percentage amino acid identity (rice) Percentage of coding region covered 
FKF1 AcFKF1 GQ232754 63.4 69.7 100 
GI AcGI GQ232756 69.0 75.8 100 
ZTL AcZTL GQ232755 72.1 75.5 100 
CO/Hd1 AcCOL GQ232751 34.9 (48.1 for COL4) 40.6 100 
FT/Hd3a AcFTL CF438000 52.9 56.2 (56.9 for OsFT586 
PHYA AcPHYA GQ232753 55.8 56.3 33 
COP1 AcCOP1 CF451443 79.4 81.9 35 
Arabidopsis/ rice gene Annotation Accession No. Percentage amino acid identity (Arabidopsis) Percentage amino acid identity (rice) Percentage of coding region covered 
FKF1 AcFKF1 GQ232754 63.4 69.7 100 
GI AcGI GQ232756 69.0 75.8 100 
ZTL AcZTL GQ232755 72.1 75.5 100 
CO/Hd1 AcCOL GQ232751 34.9 (48.1 for COL4) 40.6 100 
FT/Hd3a AcFTL CF438000 52.9 56.2 (56.9 for OsFT586 
PHYA AcPHYA GQ232753 55.8 56.3 33 
COP1 AcCOP1 CF451443 79.4 81.9 35 

The percentage of the coding region covered was estimated based on Arabidopsis and rice gene sequences. Arabidopsis accessions: FKF1, NM_105475; GI, NM_102124; ZTL, NM_125119; CO, X94937; COL4, NM_122402; FT, NM_105222; PHYA, NM_001123784; COP1, NM_128855.3. Rice accessions: FKF1, NM_001074600; GI, NM_001048755; ZTL, NM_001064973; Hd1, AB041840; Hd3a, Os01g06320; OsFT5, Os0239064; PHYA, AB109891; COP1, AB040053.

Partial sequence information was obtained for an onion putative PHYA homolog. This gene was assigned the name AcPHYA and was shown to share a high level of nucleotide and amino acid identity with Arabidopsis and rice PHYA (Table 1). A phylogenetic analysis (Supplementary Fig. S3) supported its identity as the onion PHYA homolog. In onion, FR is essential for bulb initiation (Lercari 1982) whilst in Arabidopsis PHYA mediates the response of seedlings to FR, which is consistent with a role for AcPHYA in mediating bulb initiation in response to FR (Thomas 2006). Further analysis of AcPHYA is required to elucidate the exact function of this gene in onion bulbing.

A full-length cDNA clone bearing homology to GI was identified following a normalized library screen and 5′ RACE (rapid amplification of cDNA ends) PCR. The gene represented by this clone was assigned the name Allium cepa GIGANTEA (AcGI) and was shown to share very high nucleotide and amino acid identities with both Arabidopsis and rice GI (Table 1). This gene was further analyzed using expression and phylogenetic analyses.

Full-length cDNA clones bearing homology to FKF1 and ZTL were identified using degenerate PCR and 5′ RACE PCR. Genes representing these cDNAs were assigned the names Allium cepa FKF1 (AcFKF1) and Allium cepa ZTL (AcZTL). Both genes showed very high nucleotide and amino acid identities with Arabidopsis and rice FKF1 and ZTL, respectively (Table 1) and were further characterized through expression and phylogenetic analyses.

The Allium cepa Gene Index also contains expressed sequence tag (EST) sequences with sequence similarity to Arabidopsis FT and COP1. These genes were assigned the names Allium cepa FT-like (AcFTL) and AcCOP1. AcCOP1 showed very high nucleotide and amino acid identity to Arabidopsis and rice COP1, suggesting a potentially similar role for this gene in onion. AcFTL appeared to be more closely related to a group of FT-like genes than FT/Hd3a (Supplementary Fig. S4). This gene showed no clear diurnal expression pattern, although its expression level was much higher in LDs than in SDs (Supplementary Fig. S5).

Phylogenetic analysis of GI homologs

A phylogenetic analysis was carried out in order to establish the relationship between AcGI and other GI proteins. AcGI clustered with other monocot GI homologs, with the support of high bootstrap values (Fig. 1). As there are no other genes known to show homology with GI in Arabidopsis or other species (Fowler et al. 1999), we conclude that AcGI is the onion GI homolog.

Fig. 1

Section of a neighbor–joining (NJ) tree showing the evolutionary relationships (based on amino acid sequences) among GI homologs. Full-length and partial gene sequences were included in this analysis and numbers represent bootstrap values from 1,000 replicates. The tree was constructed using Clustal X (Thompson et al. 1997) and the bar indicates 0.1 substitutions per site. Accession numbers: PtGI (Pinus taeda), TC105492; NtGI (Nicotiana tabacum), DV162002; VvGI (Vitis vinifera), TC75660; LsGI (Lactuca sativa), TC27622; AtGI (Arabidopsis thaliana), NM_102124; Pop GI (Populus), TC124305; AfxAp GI (Aquilegia formosa × pubescens), TC20877; Lpa GI (Lemna paucicostata), AB210843; Lg GI (Lemna gibba), AB210848; Ac GIa (Allium cepa), GC232756; Ac GIb, GC232757; Hv GI (Hordeum vulgare), AY740523; Ta GI (Triticum aestivum), AF543844; Lp (Lolium perenne), DQ534010; Os GI (Oryza sativa), NM_001048755; Zm GI1A (Zea mays), BK006299; Zm GI1B: BK006298; Sb GI (Sorghum bicolor), TC113678. TC numbers refer to unigenes made up of ESTs held by the Dana-Farber Cancer Institute (http://compbio.dfci .harvard.edu/tgi/).

Fig. 1

Section of a neighbor–joining (NJ) tree showing the evolutionary relationships (based on amino acid sequences) among GI homologs. Full-length and partial gene sequences were included in this analysis and numbers represent bootstrap values from 1,000 replicates. The tree was constructed using Clustal X (Thompson et al. 1997) and the bar indicates 0.1 substitutions per site. Accession numbers: PtGI (Pinus taeda), TC105492; NtGI (Nicotiana tabacum), DV162002; VvGI (Vitis vinifera), TC75660; LsGI (Lactuca sativa), TC27622; AtGI (Arabidopsis thaliana), NM_102124; Pop GI (Populus), TC124305; AfxAp GI (Aquilegia formosa × pubescens), TC20877; Lpa GI (Lemna paucicostata), AB210843; Lg GI (Lemna gibba), AB210848; Ac GIa (Allium cepa), GC232756; Ac GIb, GC232757; Hv GI (Hordeum vulgare), AY740523; Ta GI (Triticum aestivum), AF543844; Lp (Lolium perenne), DQ534010; Os GI (Oryza sativa), NM_001048755; Zm GI1A (Zea mays), BK006299; Zm GI1B: BK006298; Sb GI (Sorghum bicolor), TC113678. TC numbers refer to unigenes made up of ESTs held by the Dana-Farber Cancer Institute (http://compbio.dfci .harvard.edu/tgi/).

Expression of AcGI

The expression of AcGI over a 24 h period under both LD and SD conditions was determined using quantitative reverse transcription–PCR (qRT–PCR). In an LD onion variety (Renate F1) the expression of AcGI was seen to peak around ZT 10 (Zeitbeger time) in LDs compared with ZT 7 in SDs (Fig. 2a). This expression pattern is very similar to that shown by Arabidopsis GI, which peaks at ZT 10 in LDs and at ZT 8 in SDs (Fowler et al. 1999). AcGI has a diurnal expression pattern, characteristic of genes involved in the photoperiod response (Jackson 2008), implying it is circadian regulated, although expression in constant light conditions would be required to confirm this.

Fig. 2

Average expression of AcGI over 24 h in onion leaves. (a) LD (open circles) and SD (filled circles) expression in Renate F1 (LD variety), relative to EF1α. (b) LD expression in Agrifound Dark (SD variety, open circles) and Candy F1 (ID variety, filled circles), relative to β-tubulin. (c) SD expression in Agrifound Dark (open circles) and Candy F1 (filled circles), relative to β-tubulin. Error bars represent the SEM of six replicates. White and black bars represent light/dark cycles.

Fig. 2

Average expression of AcGI over 24 h in onion leaves. (a) LD (open circles) and SD (filled circles) expression in Renate F1 (LD variety), relative to EF1α. (b) LD expression in Agrifound Dark (SD variety, open circles) and Candy F1 (ID variety, filled circles), relative to β-tubulin. (c) SD expression in Agrifound Dark (open circles) and Candy F1 (filled circles), relative to β-tubulin. Error bars represent the SEM of six replicates. White and black bars represent light/dark cycles.

The expression of AcGI was also examined in Candy F1 and Agrifound Dark, intermediate day (ID) and SD onion varieties, respectively. When grown under LD and SD conditions, patterns of GI expression were similar in both varieties to those observed for the LD variety Renate F1 when grown under similar conditions, peaking at around ZT 10 and ZT 7–8, respectively (Fig. 2).

Analysis of AcFKF1 and AcZTL proteins

The predicted protein sequences of AcZTL and AcFKF1 were compared with those of Arabidopsis ZTL and FKF1. Both AcZTL and AcFKF1 contain an F-box, six kelch repeats and a LOV domain, as observed for the homologous Arabidopsis genes (Supplementary Fig. S6).

A phylogenetic analysis of FKF1 and ZTL family proteins was carried out. A neighbor–joining (NJ) tree was constructed and rooted through a clade containing F-box genes which lack the other domains essential for FKF1 and ZTL function (Fig. 3). The FKF1 and ZTL homologs form a large clade which is clearly divided into two smaller clades, with strong support from high bootstrap values. AcFKF1 is present in the clade containing the FKF1 homologs, providing further evidence that this gene is the onion FKF1 homolog. AcZTL is present in the clade which contains the ZTL homologs, suggesting that it is the onion ZTL homolog. Both the FKF1 and ZTL clades show a clear split between monocot and dicot sequences. In both cases, the onion putative homolog clusters with monocot gene sequences. The high level of sequence conservation, mirrored by the relationships seen in the phylogenetic tree, may also suggest a level of conservation of function.

Fig. 3

Neighbor–joining (NJ) tree showing the evolutionary relationships (based on amino acid sequences) between members of the FKF1/ZTL gene family. The tree was constructed using Clustal X (Thompson et al. 1997) and is rooted through a clade containing other F-box genes. Numbers represent bootstrap values from 1,000 replicates and the bar indicates 0.05 substitutions per site. Accession numbers: Ac_FBox (Allium cepa), GQ232752; Os_FBox (Oryza sativa), NM_001067833; At_FBox (Arabidopsis thaliana), NM_104033; At_ZTL, NM_105475; In_ZTL (Impomoea nil), DQ309278; Mc_ZTL (Mesembryanthemum crystallinum), AY371291; Ac_ZTL, GQ232755; Os_ZTLa, NM_001064973; Os_ZTLb, AK111850; Zm_ZTL (Zea mays), AY104996; At_LKP2, NM_179652; Ac_FKF1, GQ232754; Os_FKF, NM_001074600; Mc_FKF1, AY371291; At_FKF1, NM_105475; St_FKF1 (Solanum tuberosum), DR751881; Mt_FKF1 (Medicago truncatula), TC130448; Pv_FKF1 (Phaseolus vulgaris), EF643234. TC numbers refer to unigenes made up of ESTs held by the Dana-Farber Cancer Institute (http://compbio.dfci.harvard.edu/tgi/).

Fig. 3

Neighbor–joining (NJ) tree showing the evolutionary relationships (based on amino acid sequences) between members of the FKF1/ZTL gene family. The tree was constructed using Clustal X (Thompson et al. 1997) and is rooted through a clade containing other F-box genes. Numbers represent bootstrap values from 1,000 replicates and the bar indicates 0.05 substitutions per site. Accession numbers: Ac_FBox (Allium cepa), GQ232752; Os_FBox (Oryza sativa), NM_001067833; At_FBox (Arabidopsis thaliana), NM_104033; At_ZTL, NM_105475; In_ZTL (Impomoea nil), DQ309278; Mc_ZTL (Mesembryanthemum crystallinum), AY371291; Ac_ZTL, GQ232755; Os_ZTLa, NM_001064973; Os_ZTLb, AK111850; Zm_ZTL (Zea mays), AY104996; At_LKP2, NM_179652; Ac_FKF1, GQ232754; Os_FKF, NM_001074600; Mc_FKF1, AY371291; At_FKF1, NM_105475; St_FKF1 (Solanum tuberosum), DR751881; Mt_FKF1 (Medicago truncatula), TC130448; Pv_FKF1 (Phaseolus vulgaris), EF643234. TC numbers refer to unigenes made up of ESTs held by the Dana-Farber Cancer Institute (http://compbio.dfci.harvard.edu/tgi/).

Expression of AcFKF1 and AcZTL

Expression of AcFKF1 was examined in LD- and SD-grown Renate F1 plants, and a clear diurnal rhythm of AcFKF1 expression was observed under both growth conditions, with a peak at ZT 10 (Fig. 4a). Whilst Arabidopsis FKF1 is similarly diurnally expressed, expression peaks at around ZT 10 in LDs and ZT 7 in SDs (Imaizumi et al. 2003).

Fig. 4

Average expression of AcFKF1 over 24 h in onion leaves, relative to β-tubulin. (a) LD (open circles) and SD (filled circles) expression in Renate F1 (LD variety). (b) LD expression in Agrifound Dark (SD variety, open circles) and Candy F1 (ID variety, filled circles) compared with Renate F1 (open squares). (c) SD expression in Agrifound Dark (open circles) and Candy F1 (filled circles). Error bars represent the SEM of six replicates. White and black bars represent light/dark cycles.

Fig. 4

Average expression of AcFKF1 over 24 h in onion leaves, relative to β-tubulin. (a) LD (open circles) and SD (filled circles) expression in Renate F1 (LD variety). (b) LD expression in Agrifound Dark (SD variety, open circles) and Candy F1 (ID variety, filled circles) compared with Renate F1 (open squares). (c) SD expression in Agrifound Dark (open circles) and Candy F1 (filled circles). Error bars represent the SEM of six replicates. White and black bars represent light/dark cycles.

The expression of AcFKF1 was investigated using onion varieties with different daylength responses. Under LD conditions, AcFKF1 is seen to peak around ZT 7–8 in Agrifound Dark, the SD variety (Fig. 4b) compared with ZT 10 in Renate F1, the LD variety. Expression peaks at an intermediate time (around ZT 9) in Candy F1, the ID variety. There is therefore a distinct difference between the timing of the peak of expression in varieties showing different daylength responses.

Under SD conditions, the expression of AcFKF1 peaked around ZT 7–8 in both the SD and ID varieties (Fig. 4c) in contrast to the peak seen in the LD variety which was around ZT 10 (Fig. 4a). Thus, as was observed for LD-grown plants, there is a difference in expression profiles between varieties with different daylength responses.

The expression of AcZTL was examined in an SD onion variety (Agrifound Dark) and was shown to be constitutively expressed in both LD- and SD-grown plants (Fig. 5). There is no diurnal expression pattern and no obvious expression peaks, which is similar to Arabidopsis where ZTL does not show cyclic expression, although ZTL protein levels oscillate, with a peak at ZT 10–13 (Kim et al. 2007). The expression profile, coupled with the phylogenetic data, is consistent with AcZTL being the onion ZTL homolog and a component of the photoperiod pathway.

Fig. 5

Average 24 h expression of AcZTL in the leaves of Agrifound Dark (SD variety) in LD (open circles) and SD (filled circles) conditions relative to β-tubulin. Error bars represent the SEM of six replicates. White and black bars denote light/dark cycles.

Fig. 5

Average 24 h expression of AcZTL in the leaves of Agrifound Dark (SD variety) in LD (open circles) and SD (filled circles) conditions relative to β-tubulin. Error bars represent the SEM of six replicates. White and black bars denote light/dark cycles.

Discussion

The genetic network controlling photoperiodic flowering in Arabidopsis is proposed to be broadly conserved across plant species, including legumes, Brassicas, rice, potato and wheat (Robert et al. 1998, Kojima et al. 2002, Martínez-García et al. 2002, Hecht et al. 2005, Miller et al. 2008). In the case of potato, the photoperiod genes control a different end process, namely tuberization (Martínez-García et al. 2002). Our working hypothesis is that the same genes also control daylength-dependent bulb initiation in onion. This is supported by the results obtained with the onion homologs of GI, FKF1 and ZTL presented herein. It is possible that expressed paralogs of the onion Arabidopsis flowering time gene homologs discussed may exist. However, their spatial or temporal expression may be such that they were not represented in the cDNA library synthesized and screened.

In Arabidopsis, the timing of expression of GI is an essential component of the photoperiodic control of flowering (Fowler et al. 1999). This gene shows a later expression peak in LDs compared with SDs, leading to increased expression of CO and induction of flowering when daylength increases. Similarly, the expression of AcGI in the LD onion variety (Renate F1) peaked later under inductive (LD) conditions than in non- inductive (SD) conditions (Fig. 2a). This is consistent with a role for AcGI in the photoperiodic control of bulb initiation. In addition, AcGI shows the same expression pattern in ID and SD onion varieties. This suggests that a key circadian rhythm component of the photoperiod pathway is active in LD, ID and SD onion varieties and that if a difference in daylength response is associated with a change in the photoperiod pathway it should be downstream of AcGI.

In contrast, the expression of AcFKF1 was seen to vary in varieties showing different daylength responses. A consequence of this variation is that in SDs, the peaks of expression of AcFKF1 in the SD and ID varieties occur during the light period, whereas in the LD variety the peak occurs in the dark period. It is possible that the differential expression of this gene contributes towards the different daylength responses seen in the three varieties tested. The earliest LD expression peak was seen in the SD variety, a variety which quickly initiates bulbing under LDs (Brewster 2008). This precocious bulbing response could be partly due to a build up of AcFKF1 protein. Later peaks were seen in the ID and LD varieties, which will initiate bulbing in LDs, but only after a certain number of leaves have been produced.

In Arabidopsis, GI forms a complex with FKF1, leading to the degradation of CDF1, a repressor of CO, and eventually floral initiation (Sawa et al. 2007). The GI–FKF1 complex appears to regulate CDF1 stability directly in the afternoon. In onion, AcFKF1 and AcGI show very high percentage identities with Arabidopsis FKF1 and GI, respectively (Table 1). Therefore it is feasible that the same interactions occur in onion. In the LD onion variety (Renate F1) it was shown that the expression profiles of AcGI and AcFKF1 are very similar in LD conditions, peaking around ZT 10 (Figs. 2a, 4a). This would allow the two gene products to form a complex to control the expression of onion CO (or an equivalent), allowing bulb initiation in LDs. Under SD conditions, there is a difference in the timing of the expression peaks. The expression of AcGI is seen to peak around ZT 7.5, compared with ZT 10 for AcFKF1. Moreover, the expression of AcFKF1 is seen to peak in the dark period. This would mean that an FKF1–GI protein complex would only be formed in the dark period and CO expression would be repressed by CDF1 during the day. However, the GI protein is degraded at night in Arabidopsis so it is possible that the GI–FKF1 complex is not formed at all in SDs (David et al. 2006). This could explain, at least partly, why this variety does not initiate bulbing under SD conditions.

The expression patterns of AcGI and AcFKF1 were very similar in an ID onion variety (Figs. 2, 4). This variety (Candy F1) was seen to produce bulbs in both LD and SD conditions and is described by seed companies as ‘day neutral’. This would allow for a daytime complex to form, leading to CO transcription and hence bulb initiation.

In an SD onion variety (Agrifound Dark), the expression profiles of AcFKF1 and AcGI were very similar under SD conditions, peaking at ZT 7.5 (Figs. 2c, 4c). This would allow for a daytime complex to form and bulbing to be initiated in SDs, i.e. the phenotype which was observed in this variety. Under LD conditions, the expression patterns of AcFKF1 and AcGI are seen to differ (Figs. 2b, 4b). A peak in AcFKF1 expression is seen around ZT 7–8 compared with ZT 10 for AcGI. This may lead to an accumulation of the FKF1–GI protein complex to occur at an earlier time of day than is seen for LD and ID varieties. This could then explain the precocious bulbing response seen in this variety.

In Arabidopsis, GI also forms a protein complex with ZTL (Kim et al. 2007). The formation of this complex is enhanced in blue light. ZTL is involved in controlling circadian rhythm through TOC1. It has been shown that GI is required to maintain the oscillation of the ZTL protein. AcZTL shows a high level of sequence identity with Arabidopsis ZTL, suggesting a conservation of function. Therefore, it is hypothesized that AcGI also forms a complex with AcZTL in order to control circadian rhythm. It was shown that AcZTL is constitutively expressed. The formation of the complex in Arabidopsis is predicted to allow the rapid deployment of ZTL during the light period (Kim et al. 2007). In all the varieties tested (and in all daylengths), the expression of AcGI was seen to peak in the light period. This would allow a peak in AcZTL protein levels to occur in the light period and hence circadian rhythm to be controlled through the onion equivalent of TOC1.

Onion plants will also flower, usually following a period of vernalization (Brewster 1997), raising the question of which genes are involved in flowering and which are involved in bulbing. It is clear that bulb initiation is photoperiodically controlled (Lancaster et al. 1996). Under inductive daylength conditions, onions will initiate bulbing, and flowering will be inhibited (Brewster 2008). Gaining a greater knowledge of the genetic control of flowering in onion would be useful for controlling and stopping flowering during bulb production.

The rationale behind this study is that knowledge of the daylength response in onion is important for adapting new varieties for growth at different latitudes. Molecular genetic studies with onion are difficult because of its very large genome size and biennial habit. The results in this paper indicate that expression patterns of genes involved in the daylength response in Arabidopsis are also seen in onion. Furthermore, variations in AcFKF1 expression in onions with different daylength responses are consistent with a role for those genes in establishing daylength sensitivities. If confirmed in a wider range of germplasm, this information may be useful in accelerating onion breeding programs.

Materials and Methods

Gene isolation

Renate F1 onions, an LD variety from the UK which requires a daylength of approximately 16 h or more to initiate bulbing (Elsoms Seeds Ltd.), grown in the glasshouse in spring and summer were used as the starting material for gene isolation. Initially, the A. cepa Gene Index was screened for genes which showed sequence similarity to Arabidopsis photoperiod pathway genes. Clones homologous to CO and PHYA (AcCO and AcPHYA) were obtained (cloned into the EcoRV site of a pCMV.SPORT 6 vector, Invitrogen) and sequenced.

The onion GI homolog (AcGI) was obtained by screening a normalized cDNA library (produced by vertis Biotechnologie AG) produced using RNA extracted from onion Renate F1 leaf and bulb tissue harvested every 3 h over a 24 h period. The probe was generated from cDNA by PCR using the primers 5′-CAGGCCGAGAAGGATTTACAAC-3′ and 5′-CAAAACTC CGGTTCTGACAGTG-3′ at an annealing temperature of 61°C and a digoxigenin (DIG) non-radioactive nucleic acid labeling kit (Roche) following the manufacturer's guidelines. RACE PCR was used to obtain sequence for the 5′ end of the gene (Invitrogen, Gene Racer Kit) using the gene-specific primer 5′-GGCACGAAGAAGAAGATCCGAGGCACTA-3′ and the nested primer 5′-CAACATCACAAAGCGCATCCACTACCT-3′ following the manufacturer's instructions. Full-length clones were generated using primers designed to the untranslated regions (5′-GCCTTCTTCACGAAAAATCGCAGTG-3′ and 5′-CCAAGACGATTACAAGGATGATAGA-3′).

Members of the FKF1/ZTL gene family were obtained by degenerate PCR using a protocol adapted from the 3′ RACE PCR method (Borson et al. 1992), communicated by Dr. Ken Manning, University of Warwick, UK. Superscript™ II Reverse Transcriptase (Invitrogen) was used to synthesize cDNA using a modified poly(dT) primer (5′-GCGAGCACAGAATTAATA CGACTCACTATAGGTTTTTTTTTTTTVN-3′). PCR was carried out using a primer designed to amplify both FKF1/ZTL homologs (5′-ATGGTHTGTCARAAYGCDTGGGG-3′) and a primer specific to the modified poly(dT) primer (5′-GCGA GCACAGAATTAATACGAC-3′). Sequences for the 5′ end of AcFKF1 and AcZTL genes were obtained using 5′ RACE PCR (Invitrogen) with the gene-specific primer 5′-CCACCCGCTT CCAAGTGGGCTGGTTTG-3′ and the nested primer 5′-CGA GGGTGGTGAGTTCGCGGGAGAGT-3′ for AcFKF1, and the gene-specific primer 5′-GCCCTTGCCTACCACATCCACCA AAT-3′ and nested primer 5′-CCTGCTGCTGGCATGGTTTCT AACGC-3′ for AcZTL. Primers were designed to the untranslated regions of both genes to obtain full-length clones (5′-TCCAAATCCCAAACCAATTACAGC-3′ and 5′-GCATGAA AACGAGCACAATCAGA-3′ for AcFKF1, and 5′-CACAACCA CACACTGATTTTCACA-3′ and 5′-CTCGTTCCTTCTCCAATC GATCA-3′ for AcZTL).

Growth and harvesting of plants for expression analyses

Renate F1 onions were grown from May to August 2006 at Wellesbourne (latitude 52°12′). All plants received 8 h of natural daylight within a glasshouse. Plants in LDs were subjected to a daylength extension of 8 h using low-level incandescent light within a photoperiod chamber with an average photosynthetic photon flux density (PPFD) of 5 μmol m−2 s−1 whilst plants in SDs were subjected to 16 h of darkness in a photoperiod chamber. At bulb initiation in LD-grown plants (99 d after sowing), harvests were carried out at set times over a 48 h period. Middle sections of the youngest fully expanded leaves were harvested, chopped into small sections, flash frozen in liquid nitrogen and stored at −80°C until required for analysis. Leaf material was harvested from three separate plants and pooled. Plants were selected for harvest using a random number generator.

Agrifound Dark, an SD onion variety from India which requires a daylength of approximately 12 h for optimal bulb formation, and Candy F1, an ID variety from the USA which requires a daylength of approximately 14 h or more to initiate bulbing, were grown from April to September 2007 at Wellesbourne (latitude 52°12′). Seed was sourced from the Warwick HRI Genetic Resources Unit, Wellesbourne, UK. Plants were subjected to 8 h of natural daylight plus a daylength extension of 8 h (LD) or 4 h (SD) using low-level incandescent light within a photoperiod chamber with a mean PPFD of 5 μmol m−2 s−1. Plants were placed in specific locations using a Latin square design (Mead et al. 1993). Leaf material was harvested at set times over a 48 h period as described for the LD onion expression experiment. Harvests were carried out when bulbing had been initiated under inductive daylengths (62 d after sowing).

Quantitative expression analyses

Total RNA was extracted from 100 mg of leaf material harvested at each time point using Trizol® reagent (Invitrogen) following the manufacturer's guidelines. Samples were DNase treated using TURBO DNA-free™ (Ambion) and first-strand cDNA synthesized from 2 μg of total RNA using Superscript™ II Reverse Transcriptase (Invitrogen) following the manufacturer's guidelines. Quantitative real-time PCR was carried out using an I-Cycler (Bio-Rad Laboratories, iCycler Thermal Cycler). To assess the expression of AcGI in an LD onion variety (Renate F1), reaction volumes of 25 μl were used, containing 1 μl of cDNA, 1× PCR Mastermix containing SYBR green (Eurogentec) and 0.4 μM of each primer (5′-CACAGATGG ATTGCTTGTTGATG-3′ and 5′-ATTGGCTACGAGATGAACT GCTC-3′). Cycling was carried out as described in the manufacturer's guidelines with an annealing temperature of 61°C. All samples were run in triplicate and data normalized to the expression of Elongation Factor 1 Alpha (AcEF1α, accession No. CF437531) using the primers 5′-TGGCATCC AACTCTAAGGACGAT-3′ and 5′-AATGTGAGATGTGTGGC AATCCA-3′.

For all other qRT–PCR analyses, a MESA GREEN qPCR MasterMix for SYBR® green with fluorescein (Eurogentec) was used, following the manufacturer's guidelines. Reactions were carried out in 15 μl volumes, containing 0.5 μl of cDNA. All samples were run in triplicate and all data normalized to Acβ-tubulin (accession No. AA451549) using the primers 5′-GTCTTCAGAGGCAAGATGAGCAC-3′ and 5′-TCAGTCC AGTAGGAGGAATGTCG-3′. For analysis of AcFKF1, the primers 5′-CCGGTGCAGTTGTTTATGTTGGAT-3′ and 5′-TCCCA CCCACCACACAGGTACTAT-3′ with an annealing temperature of 65°C were used. For AcZTL, the primers 5′-GTTTGG TGGTCTGGCTAAGAGTG-3′ and 5′-CTCCAGGCATACTGCT ACCTGTT-3′ with an annealing temperature of 65°C were used. Data from both cycles of the 48 h time course were combined to calculate average expression over a 24 h period.

Phylogenetic analyses

Phylogenetic analyses of GI and FKF1/ZTL homologs were conducted in the same way. Published sequences were collated, converted to predicted amino acid sequences using the EditSeq package of DNAStar, and alignments carried out using Clustal X (Thompson et al. 1997). NJ trees were constructed using Clustal X, and bootstrap values were calculated using 1,000 replicates. Phylogenetic trees were viewed and edited using NJPlot (Perrière and Gouy 1996).

Supplementary data

Supplementary data are available at PCP online.

Funding

This work was supported by the Biotechnology and Biological Sciences Research Council (BBSRC).

Acknowledgments

We would like to thank Professor Mike Havey (USDA, University of Wisconsin, USA) for providing the onion EST clones. Thanks also to Dr Ken Manning for communicating a method for the isolation of genes and gene families using a single degenerate primer, and Dr Robin Allaby for his assistance on phylogenetic analyses (both University of Warwick, UK).

Abbreviations

    Abbreviations
  • AcCOL

    Allium cepa CONSTANS-LIKE

  • AcCOP1

    Allium cepa CONSTITUTIVE PHOTOMORPHOGENIC 1

  • AcEF1α

    Allium cepa ELONGATION FACTOR 1 ALPHA

  • AcGI

    Allium cepa GIGANTEA

  • AcFKF1

    Allium cepa FLAVIN-BINDING, KELCH REPEAT, F-BOX 1

  • AcFTL

    Allium cepa FLOWERING LOCUS T-LIKE

  • AcPHYA

    Allium cepa PHYTOCHROME A

  • AcZTL

    Allium cepa ZEITLUPE

  • COL4

    CONSTANS-LIKE 4

  • CDF1

    CYCLING DOF FACTOR 1, CO, CONSTANS

  • COP1

    CONSTITUTIVE PHOTOMORPHOGENIC 1

  • CRY

    CRYPTOCHROME

  • EF1α

    ELONGATION FACTOR 1 ALPHA

  • ELF3

    EARLY FLOWERING 3

  • EST

    expressed sequence tag

  • FKF1

    FLAVIN-BINDING, KELCH REPEAT, F-BOX 1

  • FR

    far-red

  • FT

    FLOWERING LOCUS T

  • GI

    GIGANTEA

  • Hd1

    Heading date 1

  • ID

    intermediate day

  • LD

    long day

  • LKP2

    LOV KELCH PROTEIN2

  • LOV

    light, oxygen or voltage

  • NJ

    neighbor–joining

  • PPFD

    photosynthetic photon flux density

  • PHY

    PHYTOCHROME

  • PHYA

    PHYTOCHROME A

  • qRT–PCR

    quantitative reverse transcription–PCR

  • RACE

    rapid amplification of cDNA ends

  • SD

    short day

  • TOC1

    TIMING OF CAB EXPRESSION 1

  • ZT

    Zeitgeber time

  • ZTL

    ZEITLUPE.

References

Borson
ND
Salo
WL
Drewes
LR
A lock-docking oligo(dT) primer for 5' and 3' RACE PCR
Genome Res.
 , 
1992
, vol. 
2
 (pg. 
144
-
148
)
Brewster
JL
Rabinowitch
HD
Brewster
JL
Physiology of crop growth and bulbing
Onions and Allied Crops, Volume I: Botany, Physiology and Genetics
 , 
1990
Boca Raton
CRC Press
(pg. 
53
-
88
)
Brewster
JL
Environmental physiology of the onion: towards quantitative models for the effects of photoperiod, temperature and irradiance on bulbing, flowering and growth
Acta Hort.
 , 
1997
, vol. 
433
 (pg. 
347
-
373
)
Brewster
JL
Onions and Other Vegetable Alliums
 , 
2008
2nd edn
Wallingford, UK
CAB International
Corbesier
L
Vincent
C
Jang
S
Fornara
F
Fan
Q
Searle
I
, et al.  . 
FT protein movement contributes to long-distance signaling in floral induction of Arabidopsis
Science
 , 
2007
, vol. 
316
 (pg. 
1030
-
1033
)
David
KM
Armbruster
U
Tama
N
Putterill
J
Arabidopsis GIGANTEA protein is post-transcriptionally regulated by light and dark
FEBS Lett.
 , 
2006
, vol. 
580
 (pg. 
1193
-
1197
)
FAOSTAT
Food and Agriculture Organisation of the United Nations–Agricultural Data
2008
 
Fornara
F
Panigrahi
KCS
Gissot
L
Sauerbrunn
N
Ruhl.
M
Jarillo
JA
, et al.  . 
Arabidopsis DOF transcription factors act redundantly to reduce CONSTANS expression and are essential for a photoperiodic flowering response
Dev. Cell
 , 
2009
, vol. 
17
 (pg. 
75
-
86
)
Fowler
S
Lee
K
Onouchi
H
Samach
A
Richardson
K
Morris
B
, et al.  . 
GIGANTEA: a circadian clock-controlled gene that regulates photoperiodic flowering in Arabidopsis and encodes a protein with several possible membrane-spanning domains
EMBO J.
 , 
1999
, vol. 
18
 (pg. 
4679
-
4688
)
Griffiths
S
Dunford
RP
Coupland
G
Laurie
DA
The evolution of CONSTANS-like gene families in barley, rice and Arabidopsis
Plant Physiol.
 , 
2003
, vol. 
131
 (pg. 
1855
-
1867
)
Hecht
V
Foucher
F
Ferrandiz
C
Macknight
R
Navarro
C
Morin
J
, et al.  . 
Conservation of Arabidopsis flowering genes in model legumes
Plant Physiol.
 , 
2005
, vol. 
137
 (pg. 
1420
-
1434
)
Imaizumi
T
Schultz
TF
Harmon
FG
Ho
LA
Kay
SA
FKF1 F-box protein mediates cyclic degradation of a repressor of CONSTANS in Arabidopsis
Science
 , 
2005
, vol. 
309
 (pg. 
293
-
297
)
Imaizumi
T
Tran
HG
Swartz
TE
Briggs
WR
Kay
SA
FKF1 is essential for photoperiodic-specific light signalling in Arabidopsis
Nature
 , 
2003
, vol. 
426
 (pg. 
302
-
306
)
Izawa
T
Oikawa
T
Sugiyama
N
Tanisaka
T
Yano
M
Shimamoto
K
Phytochrome mediates the external light signal to repress FT orthologs in photoperiodic flowering of rice
Genes Dev.
 , 
2002
, vol. 
16
 (pg. 
2006
-
2020
)
Jack
T
Molecular and genetic mechanisms of floral control
Plant Cell
 , 
2004
, vol. 
16
 (pg. 
S1
-
S17
)
Jackson
SD
Plant responses to photoperiod
New Phytol.
 , 
2009
, vol. 
181
 (pg. 
517
-
531
)
Jaeger
KE
Wigge
PA
FT protein acts as a long-range signal in Arabidopsis
Curr. Biol.
 , 
2007
, vol. 
17
 (pg. 
1050
-
1054
)
Jang
S
Marchal
V
Panigrahi
KC
Wenkel
S
Soppe
W
Deng
X
, et al.  . 
Arabidopsis COP1 shapes the temporal pattern of CO accumulation conferring a photoperiodic flowering response
EMBO J.
 , 
2008
, vol. 
27
 (pg. 
1277
-
1288
)
Jung
J-H
Seo
Y-H
Seo
PJ
Reyes
JL
Yun
J
Chua
N-H
, et al.  . 
The GIGANTEA-regulated MicroRNA172 mediates photoperiodic flowering independent of CONSTANS in Arabidopsis
Plant Cell
 , 
2007
, vol. 
19
 (pg. 
2736
-
2748
)
Kim
W-Y
Fujiwara
S
Suh
S-S
Kim
J
Kim
Y
Jan
L
, et al.  . 
ZEITLUPE is a circadian photoreceptor stabilized by GIGANTEA in blue light
Nature
 , 
2007
, vol. 
449
 (pg. 
356
-
360
)
Knott
JE
Effect of a localised photoperiod on spinach
Proc. Amer. Soc. Hort. Sci.
 , 
1934
, vol. 
31
 (pg. 
152
-
154
)
Kojima
S
Takahashi
Y
Kobayashi
Y
Monna
L
Sasaki
T
Araki
T
, et al.  . 
Hd3a, a rice ortholog of the Arabidopsis FT gene, promotes transition to flowering downstream of Hd1 under short-day conditions
Plant Cell Physiol.
 , 
2002
, vol. 
43
 (pg. 
1096
-
1105
)
Komeda
Y
Genetic regulation of time to flower in Arabidopsis thaliana
Annu. Rev. Plant Biol.
 , 
2004
, vol. 
55
 (pg. 
521
-
535
)
Kuhl
JC
Cheung
F
Yuan
Q
Martin
W
Zewdie
Y
McCallum
J
, et al.  . 
A unique set of 11,008 onion expressed sequence tags reveals expressed sequence and genomic differences between the monocot orders Asparagales and Poales
Plant Cell
 , 
2004
, vol. 
16
 (pg. 
114
-
125
)
Lancaster
JE
De Ruiter
JM
Triggs
CM
Gandar
PW
Bulbing in onions: photoperiod and temperature requirements and prediction of bulb size and maturity
Ann. Bot.
 , 
1996
, vol. 
78
 (pg. 
423
-
430
)
Lercari
B
The promoting effect of far-red light on bulb formation in the long-day plant Allium cepa L
Plant Sci. Lett.
 , 
1982
, vol. 
27
 (pg. 
243
-
254
)
Lercari
B
Role of phytochrome in photoperiodic regulation of bulbing and growth in the long day plant Allium-cepa
Physiol. Plant.
 , 
1984
, vol. 
60
 (pg. 
433
-
436
)
Martínez-García
JF
Virgós-Soler
A
Prat
S
Control of photoperiod-regulated tuberization in potato by the Arabidopsis flowering-time gene CONSTANS
Proc. Natl Acad. Sci, USA
 , 
2002
, vol. 
99
 (pg. 
15211
-
15216
)
Mas
P
Kim
W-Y
Somers
DE
Kay
SA
Targeted degradation of TOC1 by ZTL modulates circadian function in Arabidopsis thaliana
Nature
 , 
2003
, vol. 
426
 (pg. 
567
-
570
)
Massiah
AJ
2007
 
Understanding flowering time. CAB Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural Resources 2
McCallum
J
Leite
D
Pither-Joyce
M
Havey
MJ
Expressed sequence markers for genetic analysis of bulb onion (Allium cepa L.)
Theor. Appl. Genet.
 , 
2001
, vol. 
103
 (pg. 
979
-
991
)
Mead
R
Curnow
RN
Hasted
AM
Statistical Methods in Agriculture and Experimental Biology
 , 
1993
2nd edn
London
Chapman & Hall
Mettananda
KA
Fordham
R
The effects of 12 and 16 h daylength treatments on the onset of bulbing in 21 onion cultivars (Allium cepa L) and its application to screening germplasm for use in the tropics
J. Hort. Sci. Biotechnol.
 , 
1997
, vol. 
72
 (pg. 
981
-
988
)
Miller
TA
Muslin
EH
Dorweiler
JE
A maize CONSTANS-like gene, conz1, exhibits distinct diurnal expression patterns in varied photoperiods
Planta
 , 
2008
, vol. 
227
 (pg. 
1377
-
1388
)
Mizoguchi
T
Wright
L
Fujiwara
S
Cremer
F
Lee
K
Onouchui
H
, et al.  . 
Distinct roles of GIGANTEA in promoting flowering and regulating circadian rhythms in Arabidopsis
Plant Cell
 , 
2005
, vol. 
17
 (pg. 
2255
-
2270
)
Nelson
DC
Lasswell
J
Rogg
LE
Cohen
MA
Bartel
B
FKF1, a clock-controlled gene that regulates the transition to flowering in Arabidopsis
Cell
 , 
2000
, vol. 
101
 (pg. 
331
-
340
)
Park
DH
Somers
DE
Kim
YS
Choy
YH
Lim
HK
Soh
MS
, et al.  . 
Control of circadian rhythms and photoperiodic flowering by the Arabidopsis GIGANTEA gene
Science
 , 
1999
, vol. 
285
 (pg. 
1579
-
1582
)
Perrière
G
Gouy
M
WWW-Query: an on-line retrieval system for biological sequence banks
Biochimie
 , 
1996
, vol. 
78
 (pg. 
364
-
369
)
Quail
P
Boylan
M
Parks
B
Short
T
Xu
Y
Wagner
D
Phytochromes: photosensory perception and signal transduction
Science
 , 
1995
, vol. 
268
 (pg. 
675
-
680
)
Robert
LS
Robson
F
Sharpe
A
Lydiate
D
Coupland
G
Conserved structure and function of the Arabidopsis flowering time gene CONSTANS in Brassica napus
Plant Mol. Biol.
 , 
1998
, vol. 
37
 (pg. 
763
-
772
)
Sawa
M
Nusinow
DA
Kay
SA
Imaizumi
T
FKF1 and GIGANTEA complex formation is required for day-length measurement in Arabidopsis
Science
 , 
2007
, vol. 
318
 (pg. 
261
-
265
)
Sobeih
WY
Wright
CJ
The photoperiodic regulation of bulbing in onions (Allium-cepa L). 2. Effects of plant-age and size
J. Hort. Sci.
 , 
1986
, vol. 
61
 (pg. 
337
-
341
)
Stevens
PF
Angiosperm Phylogeny Website
2001
 
Suárez-López
P
Wheatley
K
Robson
F
Onouchi
H
Valverde
F
Coupland
G
CONSTANS mediates between the circadian clock and the control of flowering in Arabidopsis
Nature
 , 
2001
, vol. 
410
 (pg. 
1116
-
1120
)
Tamaki
S
Matsuo
S
Hann
LW
Yokoi
S
Shimamoto
K
Hd3a protein is a mobile flowering signal in rice
Science
 , 
2007
, vol. 
316
 (pg. 
1033
-
1036
)
Thomas
B
Light signals and flowering
J. Exp. Bot.
 , 
2006
, vol. 
57
 (pg. 
3387
-
3393
)
Thomas
B
Vince-Prue
D
Photoperiodism in Plants
 , 
1997
2nd edn
San Diego
Academic Press
Thompson
JD
Gibson
TJ
Plewniak
F
Jeanmougin
F
Higgins
DG
The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools
Nucleic Acids Res.
 , 
1997
, vol. 
25
 (pg. 
4876
-
4882
)
Valverde
F
Mouradov
A
Soppe
W
Ravenscroft
D
Samach
A
Coupland
G
Photoperiodic regulation of CONSTANS protein in photoperiodic flowering
Science
 , 
2004
, vol. 
303
 (pg. 
1003
-
1006
)
Yu
J-W
Rubio
V
Lee
N-Y
Bai
S
Lee
SY
Kim
SS
, et al.  . 
COP1 and ELF3 control circadian function and photoperiodic flowering by regulating GI stability
Mol. Cell
 , 
2008
, vol. 
32
 (pg. 
617
-
630
)