Clock-Controlled and FLOWERING LOCUS T (FT)-Dependent Photoperiodic Pathway in Lotus Japonicus I: Veriﬁcation of the Flowering-Associated Function of an FT Homolog

Biosci. Biotechnol. Biochem., 77 (4), 747–753, 2013

Clock-Controlled and FLOWERING LOCUS T (FT)-Dependent Photoperiodic Pathway in Lotus japonicus I: Veriﬁcation of the Flowering-Associated Function of an FT Homolog

y Takafumi YAMASHINO, Saori YAMAWAKI, Emi HAGUI, Hanayo UEOKA-NAKANISHI, Norihito NAKAMICHI, Shogo ITO, and Takeshi MIZUNO

Laboratory of Molecular and Functional Genomics, School of Agriculture, Nagoya University, Chikusa-ku, Nagoya 464-8601, Japan

Received November 13, 2012; Accepted January 14, 2013; Online Publication, April 7, 2013 [doi:10.1271/bbb.120871]

During the last decade, significant research progress highly conserved throughout the plant kingdom.5–8) in the study of Arabidopsis thaliana has been made in However, it is also assumed that the mechanisms by defining the molecular mechanism by which the plant which the clock regulates flowering time differs consid- circadian clock regulates flowering time in response to erably in detail between species.9–11) In this respect, changes in photoperiod. It is generally accepted that comparative genomics approach might be effective in the clock-controlled CONSTANS (CO)-FLOWERING addressing these issues. LOCUS T (FT)-mediated external coincidence mecha- A. thaliana is classified as a facultative annual long- nism underlying the photoperiodic control of flowering day plant, whose flowering is induced in response to a time is conserved in higher plants, including A. thaliana longer day length (or photoperiod). The circadian clock and Oryza sativa. However, it is also assumed that the can measure the day length, and tells the time to mechanism differs considerably in detail among species. flower.3,4,12) The key player in the photoperiodic Here we characterized the clock-controlled CO-FT induction of flowering is CONSTANS (CO), which pathway in Lotus japonicus (a model legume) in serves as a DNA-binding transcription activator.13,14) comparison with that of A. thaliana. L. japonicus has The transcription of CO is under the control of the at least one FT orthologous gene (named LjFTa), which circadian clock and shows a biphasic diurnal expression is induced specifically in long-days and complements the profile with peaks in late daytime and nighttime mutational lesion of the A. thaliana FT gene. However, specifically in LDs.15–17) It has been found that CO it was speculated that this legume might lack the protiens are stabilized in late daytime.18–20) Hence, CO upstream positive regulator CO. By employing L. japo- can actively promote the transcription of FLOWERING nicus phyB mutant plants, we showed that the photo- LOCUS T (FT) in leaf phloem only in LDs.12) FT receptor mutant displays a phenotype of early flowering gene-products act in the shoot apical meristem to due to enhanced expression of LjFTa, suggesting that induce reproductive growth.21) In SDs, however, CO LjFTa is invovled in the promotion of flowering in is transcribed excluisively in nighttime. Since the CO L. japonicus. These results are discussed in the context gene-products are degrated by CONSTITUTIVE of current knowledge of the flowering in crop legumes PHOTOMORPHOGENIC 1 (COP1) in nighteime, FT such as soybean and garden pea. is not transcribed in short-days (SDs).19) In short, the clock-contolled CO-FT pathway is essential in promot- Key words: Arabidopsis thaliana; circadian clock; flowing flowering predominantely in an appropriate season ering time; Lotus japonicus; photoreceptor (or LDs) in A. thaliana. mutant We have been studying circadian clock-controlled biological mechanisms, including photoperiodic control Since plants are sessile, they must be able to sense of flowering time, employing not only A. thaliana but changes in environmental light conditions and adapt also L. japonicus.22–25) L. japonicus belongs to a large their developmental processes accordingly. In this family of legumes that are present in most ecosystems respect, the circadian clock plays prominent roles and include many important crop species.26) It is a model through providing an adaptive advantage in anticipating legume of choice in conducting comparative genome- daily changes in light/dark conditions and seasonal wide studies,27–29) together with Medicago truncatula,30) changes in the photoperiod.1,2) In Arabidopsis thaliana, Glycine max (soybean),31,32) and Pisum sativum (garden significant progress has been made in defining the pea).33) Among these, Medicago and Pisum are predom- molecular mechanisms by which the circadian clock inantly long day plants from temperate regions, whereas regulates long-days (LDs)-specific promotion of flower- soybean as well as bean (Phaseolus) originating at lower ing in A. thaliana.3,4) It is generally accepted that the latitudes is predominantly short-days plants. L. japoni- molecular bases of clock-controlled flowering time are cus (accessions Gifu and Miyakojima MG-20) is a

y To whom correspondence should be addressed. Fax: +81-52-789-4091; E-mail: [email protected] Abbreviations: CaMV, cauliflower mosaic virus; CO, CONSTANS; COP1, CONSTITUTIVE PHOTOMORPHOGENIC 1; FLC, FLOWERING LOCUS C; FT, FLOWERING LOCUS T; LDs, long-days; SDs, short-days 748 T. YAMASHINO et al. perennial temperate legume that sets flowers preferen- The BLAST conditions used were blastp against all predicted peptides, tially in LDs.34,35) It is important to understand how the filter low complexity, expect value 1, and matrix BLOSUM62. The circadian clock regulates the flowering time of legume best-hit coding sequence in L. japonicus was found, and the results were summarized in Supplemental Table S1 (see Biosci. Biotechnol. species in repose to seasonal changes in the photoperiod, Biochem. Web site). because this has a strong impact on the yields of crops. It would also be important to understand how the Plant materials and growth conditions. A. thaliana Columbia-0 and mechanisms underlying the regulation of flowering time L. japonicus Miyakojima MG-20 were used in this study. The were modified optionally in L. japonicus during evolu- L. japonicus experimental strain, Miyakojima MG-20, was provided tion to adapt to domestic habitates. by the National BioResource Project (L. japonicus and G. max). The L. japonicus phyB Among legumes, pea and soybean have long been mutant (EMS mutagenesis line 01-0017) was provided by Dr. A. Suzuki (Saga University). All analyses were characterized in terms of the control of flowering time, conducted in a cultivation chamber (MLR-350, Panasonic or LPH-350, and recently, a small family of FT homologs in legumes Nippon Medical & Chemical Instruments) under neutral white has been characterized for M. truncatula, pea, and fluorescent light at a constant temperature of 22 C and 25 C for soybean, based on genome-wide information.33,36–40) A. thaliana and L. japonicus respectively. The intensity of the light These FT homologs of legumes are commonly classified was adjusted to 80 mmolm2s1 for A. thaliana, and to more than 2 1 into three clades, designated FTa, FTb, and FTc, 180 mmolm s for L. japonicus. A daily 16 h light/8 h dark cycle was adopted as the long day condition of A. thaliana and L. japonicus. according to phylogenetic analyses. Some of them have Daily 10 h light/14 h dark and 11.5 h light/12.5 h dark cycles were been demonstrated that they have an ability to promote adopted as the short day conditions of A. thaliana and L. japonicus 37) flowering. We also previously showed that a gene respectively. For flowering assay, plants were germinated and grown encoding a putative FT homolog in L. japonicus has an on a gellan gum plate containing MS salts and 1% sucrose for a week, ability to promote flowering in A. thaliana.24) Hence it and then seedlings were transferred to soil. is conceivable that the function of FT homologs is Preparation of RNA and qRT-PCR. conserved between A. thaliana and legumes, but it is Total RNA was purified from frost plant (100 mg) with an RNeasy plant mini kit (Qiagen). To speculated that there might be significant differences in synthesize cDNA, RNA (1 mg of each) was converted to cDNA with detail of the mechanisms underlying the regulation of ReverTra Ace (TOYOBO) and oligo-dT primer. The synthesized flowering time between A. thalana and legumes. It is cDNAs were amplified with SYBR Premix Ex Taq II (Takara Bio) well-known that FLOWERING LOCUS C (FLC), which and the primer set for each target gene, analyzed by using a Stepone is involved in repression of FT in the vernalization Plus Real-Time PCR System (Life technologies). The primer pathway of flowering, belongs to a specialized MIKC- sets used were described in Supplemental Table S2. LjUBC (chr1.LjT04O06.110.r2.m), encoding an ubiquitin carrier protein, was type MADS box transcriptional factor that evolved only used as internal references. The following standard thermal cycling in Brassica species. Moreover, the result of an extensive program was used for all PCR: 95 C for 120 s, 40 cycles of 95 C for inspection of the soybean, L. japonicus, and M. trunca- 10 s, and 60 C for 60 s. The CT values for individual reactions were tula genome seequences suggests that these legumes determined by analysis of the raw fluorescence data (without baseline seem to lack the orthologous gene CO.24,37) Considering correction) using the freely available software PCR Miner (http: 41) these, it is important to clarify whether the canonical //www.miner.ewindup.info). Based on the comparative CT method, CO-FT-mediated flowering pathway is conserved in relative expression level was calculated. L. japonicus. Phylogenetic analysis. Amino acid sequences deduced from the For this purpose, we previously compiled a list of determined cDNA sequences were aligned using the ClustalW genes presumably implicated in the FT-mediated photo- program, and the numbers of amino acid substitutions between each periodic control of flowering time in L. japonicus,24) but pair of family proteins were estimated by the Jones–Taylor–Thornton the list was based on the ver. 1.0 database of the (JTT) model with the complete-deletion option.42) From estimated L. japonicus genome sequence. An up-to-date ver. 2.5 numbers of amino acid substitutions, a phylogenetic tree was reconstructed using the neighbor-joining (NJ) method.43) The bootstrap database has been released to the public (http://www. values were calculated with 1000 replications. These procedures were kazusa.or.jp/lotus/index.html). In the updated database, all performed using MEGA4.1 software (http://www.megasoftware.net/ containing a number of newly annotated genes, it was index.html). revealed that L. japonicus has several additional genes each encoding a putative FT or CO homolog. In the Construction of transgenic Arabidopsis plants. To construct the present study, we reassessed the CO-FT mediated Cauliflower Mosaic Virus (CaMV) 35S promoter::LjFTb1 fusion gene, photoperiodic pathway controlling the flowering of the putative coding sequence of the LjFTb1 gene was amplified using the primers, 50-CATCTAGAGTGAAGGCGCAATGCCTAG-30 (the L. japonicus on the basis of the new database. The italic part: XbaI site) and 50-TTGCGGCCGCTTATCTCCTGCTGTT- results were discussed in the context of current knowl- CCTTGC-30 (the italic part: NotI site). The PCR products were edge of flowering in crop legumes such as soybean and digested by XbaI and NotI, and cloned into the XbaI-NotI cloning site garden pea. of the cloning vector pBluescript SK(þ) (Stratagene). After confirmed by sequencing that the intact genes were cloned, the cloned XbaI-NotI DNA fragments were subjected to subcloning to the binary vector Materials and Methods pSK1 to make the cloned genes under a control of constitutive CaMV 35S promoter.44) This construct was transformed into Agrobacterium Genome-wide analysis of L. japonicus FT/TFL1 and the COL tumefaciens strain EHA101, and then A. thaliana plants were trans- family. A set of Arabidopsis amino acid sequences was retrieved from formed by vacuum infiltration procedures. the Arabidopsis Information Resource (TAIR, http://www.arabidopsis.org/). These sequences were subjected as queries to a BLAST Results and Discussion homology search by the use of the L. japonicus genome browser (miyakogusa.jp, http://www.kazusa.or.jp/lotus/index.html). The database (Lotus japonicus genome assembly build 2.5) provides the L. japonicus has multiple FT homologs genome information of approximately 91.3% of the L. japonicus gene The revised list of genes encoding FT-like proteins space, as predicted from coverage of L. japonicus EST sequences.29) revealed that L. japonicus has at least six coding- Regulation of Flowering Time in L. japonicus 749

A 99 AtFT cloned the corresponding cDNAs, and confirmed their 71 AtTSF nucleotide sequences of LjFTa and LjFTb1. We could 99 LjFTa not clone the LjFTb2 cDNA, most likely because of its 56 93 LjFTb1 LjFTb2 very low expression level. AtBFT Besides the AtFT sub-clade, the phylogenetic analysis 97 AtTFL1 indicated that there are three additional FT-related 93 LjTFL1a proteins in A. thaliana, AtBFT (BROTHER OF FT LjTFL1b 100 AtMFT AND TFL1), AtMFT (MOTHER OF FT AND TFL1), 0.1 LjMFT and AtTFL1 (TERMINAL FLOWER 1) (Fig. 1A). Among them, AtTFL1 is particularly interesting, be- 1.2 B LjFTa cause this gene-product functions as a repressor of 1 flowering.45) It was reported that AtFT (activator of 0.8 LDs 0.6 flowering) and AtTFL1 (repressor of flowering) could be 0.4 distinguished from each other on the basis of their 0.2 SDs primary amino acid sequences (Supplemental Relative expression Relative 0 Fig. S1).46) The opposite activities of AtFT and AtTFL1 0 3 6 9 12 15 18 21 24 are to be attributed to the difference in a single amino Time (h) acid residue (Tyrosine-85 of AtFT as opposed to

1.2 Histidine-88 of AtTFL1). According to this criterion, C LjFTb1 1 L. japonicus appears to have two gene products highly LDs 0.8 homologous to AtTFL1. They were hence named 0.6 LjTFL1a and LjTFL1b (Fig. 1A). The counterpart of 0.4 AtMFT is present in L. japonicus, while that of AtBFT 0.2 SDs

Relative expression Relative 0 appeared to be absent. 0 3 6 9 12 15 18 21 24 Time (h) An LjFT homolog has an ability to promote flowering In a previous study, it was demonstrated that LjFTa is Fig. 1. Characterization of a Set of Genes Encoding FT-Like Proteins in L. japonicus. capable of complementing an A. thaliana ft mutant, 24) A, Phylogenetic analysis of FT/TFL1 family in A. thaliana and suggesting that LjFTa is orthologous to AtFT. The L. japonicus. A phylogenetic tree was constructed according to the question of the present study was: what about LjFTb1? method described in Materials and Methods. See Supplemental The diurnal expression profiles of LjFTa and LjFTb1 Table S1 for locus code and amino acid sequence of each FT-like were compared with each other in L. japonicus plants protein in L. japonicus. The FT-like protein named LjFTa in this study is most similar to the authentic Arabidopsis FT designated grown under LDs and SDs (Fig. 1B and C, respectively). AtFT and they seem to be monophyletic. B, Characterization of the In addition, a pair of 35S-promoter::LjFTa and 35S- diurnal expression profiles of LjFTa and LjFTb1 in L. japonicus. promoter::LjFTb1 fusion genes were introduced to Seedlings were grown in LDs (16 h light/8 h dark cycles) or SDs A. thaliana wild-type Col-0 plants to see the effect on (11.5 h light/12.5 h dark cycles) for 14 d, and mRNA samples were flowering time in the resultant transgenic plants (Fig. 2). prepared at every 3 h interval. The diurnal expression profiles were examined by means of qRT-PCR analyses according to the method LjFTa was expressed with a peak at dusk specifically in described in the Materials and Methods. Relative expression levels LDs, under conditions in which flowering is induced were shown as mean values SD (n ¼ 3), for which the maximum predominantly in L. japonicus (Fig. 1B). The expression value of each graph was taken as 1.0. The dark periods are indicated profile of LjFTb1 was rather constitutive regardless of with shadings. the photoperiod, although enhanced expression was observed during daytime in LDs (Fig. 1C). The results sequences apparently belonging to the FT/TFL1 family of reverse genetics for the heterologous plants showed (Supplemental Table S1). They were subjected to that the LjFTa transgene in A. thaliana exhibited an phylogenetic analysis together with members of the ability to promote flowering, as shown previously FT/TFL family of A. thaliana (Fig. 1A). In the follow- (Fig. 2A).24) Based on these results, we concluded that ing, A. thaliana and L. japonicus genes are described LjFTa, the ortholog of AtFT, is functionally equivalent with the prefixes, At and Lj respectively in order to to AtFT. Interestingly, the LjFTb1-overproducing discriminate each other. In the ver. 1.0 database, only a A. thaliana transgenic lines showed a phenotype of late single gene encoding a FT homolog was found.24) The flowering in LDs (Fig. 2B). It was speculated that LjFTb revised compilation revealed that L. japonicus has at might play an as-yet-unknown role in the control of least two more genes highly homologous to AtFT flowering time in L. japonicus. In any case, it is (Fig. 1A). Of three LiFT homologs (designated LjFTa, suggested that the function of AtFT is conserved in LjFTb1, and LjFTb2), LjFTa, characterized previously, L. japonicus and that LDs-specific induction of LjFTa is was most similar to AtFT and AtTSF (TWIN SISTER the key regulation in the photoperiodic control of OF FT). The existence of a set of multiple FT homologs flowering time in L. japonicus. was reported for other legumes including garden pea (5 FT-like genes), soybean (10 genes), and M. trichocarpa Where has the flowering activator CO gone? (5 genes), and they were commonly classified into three According to the scenario for A. thaliana flowering, sub-groups (designated FTa, FTb, and FTc). Accord- the main activator of AtFT is AtCO (see Introduction), a ingly, the two additional L. japonicus FT homologs member of the large BBX family of proteins with B-box were found to belong to the FTb sub-group. They were motif (Supplemental Fig. S2).47) Among them, AtCO hence named LjFTb1 and LjFTb2, respectively. We belongs to the COL family of proteins, containing the 750 T. YAMASHINO et al.

30 d 45 d 96 Glyma13g01290.1 A 78 Glyma17g07420.1 SDs 94 IF 84 LjCOLb Medtr8g128200.1 IF AtCOL5 98 39 AtCOL3 AtCOL4 99 Medtr1g016230. 1 80 LjCOLc 100 Glyma04g06240.1 44 99 Glyma06g06300.1 Col L3-6 L5-5 Col 60 LjCOLd LjFTa-ox Medtr3g138670.1 100 Medtr4g041790.1 35 d 45 d 94 AtCO B 99 AtCOL2 LDs AtCOL1 100 Glyma13g07030.1 63 IF Glyma19g05170.1 90 LjCOLa 72 Glyma18g51320.1 100 Glyma08g28370.1 Hd1 Col L4-2 L6-1 LjFTb1-ox 0.05

Fig. 2. Effect of Heterologous Expression of LjFTa and LjFTb1 on Fig. 3. Phylogenetic Analysis of CO Family in A. thaliana, L. japo- Flowering Phenotype in A. thaliana. nicus, G. max, and M. truncatula. A, Effect of heterologous expression of LjFTa. Wild-type (Col) A phylogenetic tree was constructed according to the method and two independent A. thaliana homozygous transgenic lines described in the Materials and Methods. See Supplemental Table S1 carrying the 35S-promoter::LjFTa transgene, denoted LjFTa-ox, for locus code and amino acid sequence of each CO-like protein in were grown in SDs (10 h light/14 h dark cycles). Flowering L. japonicus. The amino acid sequences of G. max, and M. trunca- phenotype of 30-d-old plants is shown. Inflorescences with flower tula was retrieved from the phytozome database (http://www. buds were indicated as IF. A 45-d-old reference plant was presented phytozome.net/). A neighbor-joining (NJ) tree with bootstrap values to see the wild-type plant did not set flower even in the prolonged was constructed according to the method described in the Materials growth under SDs. It was confirmed that these 35S-promoter::LjFTa and Methods. transgenic lines expressed a large amount of LjFTa transcripts, which were not detected in Col, by analysis with semi-quantitative RT-PCR (data not shown). B, Effect of heterologous expression of LjFTa. Wild-type (Col) and two independent A. thaliana homozy- LjCOLd were examined in L. japonicus seedlings grown gous transgenic lines carrying the 35S-promoter::LjFTb1 transgene, in LDs and SDs (Fig. 4). LjCOLa showed a shape peak denoted LjFTb1-ox, were grown in LDs (16 h light/8 h dark cycles). at dawn in both LDs and SDs. The expression profile is Flowering phenotype of 45-d-old plants is shown. Inflorescences similar to that of AtCOL2 (Supplemental Fig. S3). with flower buds are indicated as IF. It was confirmed that these Therefore, LjCOLa might be functionally relevant to 35S-promoter::LjFTb1 transgenic lines expressed a large amount of LjFTb1 transcript, which was not detected in Col, by analysis with AtCOL2 rather than to AtCO. LjCOLb showed a robust semi-quantitative RT-PCR (data not shown). These results indicate rhythm with a peak at the middle of the day, and the LjFTa has an ability to promote flowering in A. thaliana, but LjFTb1 expression was observed exclusively in daytime in does not. Scale bar, 2.5 cm. LDs (Fig. 4A). In SDs, the expression of LjCOLb was observed both in daytime and nighttime (Fig. 4B). B-box motif followed by the CCT-motif. Among COL Considering that CO is stabilized and LjFTa is induced sub-families, AtCO belongs to clade-I together with at dusk in LDs, it is not plausible that LjCOLb is an AtCOL1 to AtCOL5 (Supplemental Fig. S2). We activator of LjFTa. The diurnal expression of LjCOLc searched the L. japonicus genome database for genes and LjCOLd was rather constitutive, compared with that encoding proteins homologous to AtCO or AtCOL1-5. of LjCOLa and LjCOLb. Taken all of these results To gain more general idea, the same analysis was made together, we concluded tentatively that L. japonicus through inspecting the databases of soybean and does not have functional homolog of AtCO in clade I of M. truncatula (http://www.phytozome.net/). The CO- the COL family. The same argument has been made for related proteins found in the three legume species, pea.33) According to the phylogenetic analyses, soybean together with those of A. thaliana, were analyzed by and M. truncatula also appeared to have no gene constructing a phylogenetic tree (Fig. 3). The results orthologous to AtCO (Fig. 3). These findings suggest indicated that there are at least four putative LjCOL either that these legumes lack the AtCO ortholog, or that genes (LjCOLa to LjCOLd) in the L. japonicus genome a more distantly related COL homolog plays a sub- database. Among them, only LjCOLa is placed into the stituted role in these legumes. Hence, an activator that same lineage (sub-clade) as AtCO, and the closest integrates photoperiodic signals into the LDs-specific homolog of LjCOLa is AtCOL2. Although AtCOL1 and induction of LjFTa to promote flowering remains to be AtCOL2 are highly similar to AtCO in their amino acid identified in L. japonicus. This issue will be dealt with sequences (identity 62%), the important fact is that both the accompanying paper. of them are nothing to do with the control of flowering time in A. thaliana.48) Another well-known fact is that Is the red light photoreceptor LjphyB implicated in AtCO is diurnally expressed at dusk specifically in LDs the control of flowering time? to induce AtFT, while both AtCOL1 and AtCOL2 So far, we did not succeed to obtain direct exper- showed a shape peak at dawn (Supplemental Fig. S3). imental evidence that LjFTa has the ability to promote Therefore, the diurnal expression profiles of LjCOLa to flowering in L. japonicus, due to the absence of Regulation of Flowering Time in L. japonicus 751

1.2 1.2 LjphyB is implicated in the control of flowering time A LjCOLa LjCOLb 1 1 by modulating the expression of LjFTa 0.8 0.8 Both the wild-type (MG-20) and the isogenic phyB 0.6 0.6 mutant of L. japonicus plants were grown in LDs for 0.4 0.4 45 d (18 h light/6 h dark). The phyB mutant plants 0.2 0.2 Relative expression Relative Relative expression Relative exhibited an elongated morphology as compared with 0 0 49) 0 6 12 18 24 0 6 12 18 24 the wild-type plants, as reported previously (Fig. 5). It Time (h) Time (h) was further observed that the phyB mutant plants set the 1.2 1.2 LjCOLc LjCOLd first flower at the 7th node on average, while wild-type 1 1 plants did so at the 13th node on average. This event 0.8 0.8 0.6 0.6 observed for phyB is a hallmark of the early flowering 0.4 0.4 phenotype. This view was more clearly observed when 0.2 0.2 the same set of plants was grown in SDs (11.5 h light/ Relative expression Relative 0 expression Relative 0 12.5 h dark) (Fig. 6). The resulting plants were photo- 0 6 12 18 24 0 6 12 18 24 graphed 60 d after germination to show that the wild- Time (h) Time (h) type plants never set flowers. In contrast, the phyB 1.2 1.2 B LjCOLa LjCOLb mutant plants set flowers even in SDs. Thus the Lotus 1 1 phyB mutant exhibited a phenotype of early flowering 0.8 0.8 regardless of photoperiod. These were the first evidence 0.6 0.6 that the L. japonicus phyB photoreceptor gene is 0.4 0.4 implicated in the photoperiodic control of flowering

0.2 expression Relative Relative expression Relative 0.2 time. 0 0 It was of more interest to examine whether LjphyB 0 6 12 18 24 0 6 12 18 24 Time (h) Time (h) regulates the ﬂowering time by modulating the expres- 1.2 1.2 LjCOLc LjCOLd sion of LjFTa. For this purpose, both the wild-type and 1 1 phyB mutant seedlings were grown for 14 d in LDs and 0.8 0.8 SDs, and the diurnal expression proﬁles of LjFTa were 0.6 0.6 examined at every 3 h interval. In the wild-type seed- 0.4 0.4 lings, LjFTa was expressed with a peak in the afternoon

Relative expression Relative 0.2 0.2 Relative expression Relative in LDs, while its expression levels were very low 0 0 throughout in SDs, as shown earlier (Fig. 7A and B, see 0 6 12 18 24 0 6 12 18 24 Time (h) Time (h) also Fig. 1B). In the phyB mutant seedlings, LjFTa were expressed at higher levels throughout the day and night, Fig. 4. Characterization of Diurnal Expression Profiles of LjCOLa to regardless of photoperiods, suggesting that the phyB LCOLd in L. japonicus. A and B, Seedlings were grown in LDs (16 h light/8 h dark mutation causes early flowering in both LDs and SDs cycles) (A) and SDs (11.5 h light/12.5 h dark cycles) (B) for 14 d, through enhanced expression of LjFTa (Fig. 7A and B). and total RNA samples were prepared at every 3 h interval. The This result is consistent with the earlier conclusion that diurnal expression profiles were examined by means of qRT-PCR LjFTa promotes flowering in L. japonicus. LjFTb1 was analyses according to the method described in the Materials and also enhanced in phyB mutant background, suggesting Methods. Relative expression levels are shown as mean values SD (n ¼ 3), for which the maximum value of each graph was taken that complicated regulation might be involved in the as 1.0. The dark periods are indicated with shadings. photoperiodic control of flowering time in L. japonicus (Fig. 7C and D). appropriate L. japonicus mutants. Recently, a stable Implications phyB (the gene encoding one of the red light photo- L. japonicus is a perennial temperate legume that receptors) mutant of L. japonicus was isolated and responds to a longer day length. In this study, we carried characterized in terms of a linkage between nodulation out genome-wide studies on the control of flowering in roots and far-red response in shoots.49) The phyB in L. japonicus, based on current knowledge of the mutant allowed us to characterize the L. japonicus plant molecular mechanisms underlying the photoperiodic itself with special reference to the flowering time. It is control of flowering time in A. thaliana (Fig. 1). We well known of A. thaliana that the phyB photoreceptor identified a gene (LjFTa), which would be responsible regulates (or inhibits) flowering through AtFT in for stimulation of flowering in L. japonicus. L. japoni- response to changes in light quality.50) The develop- cus has another AtFT homolog (LjFTb1), which is also mental nature of Arabidopsis phyB mutant plants is suggested to play a flowering-associated role in L. striking in that it shows a phenotype of extremely early japonicus. The role of LjFTb1 in flowering appears to be flowering.51) The target of phyB may be CO at the level distinct from that of LjFTa. Curiously, it was suggested of protein, or FT directly at the level of transcription. In that this legume might lack the ortholog to the canonical any case, if the L. japonicus phyB mutant plants show a regulator for AtFT, namely, AtCO (activator). However, phenotype of early flowing through enhancement of the we do not rule out the possibility that more distantly endogenous LjFTa expression, this is good evidence that related COL homologs might play a related role in LjFTa can promote flowering in L. japonicus plants L. japonicus. Furthermore, we showed that phyB regu- themselves. Based on this, the L. japonicus phyB mutant lates the expression level of LjFTa, thereby controlling was characterized with reference to the control of flowering time in L. japonicus. Clarification of the flowering time. molecular mechanisms of flowering in L. japonicus 752 T. YAMASHINO et al. AB 1.2 1.2 LjFTa LjFTa 1 1

0.8 0.8 phyB 0.6 phyB 0.6 0.4 0.4 Relative expression Relative Relative expression Relative 0.2 MG20 0.2 MG20 0 0 03691215182124 0 3 6 9 1215182124 Time (h) Time (h)

CD 1.2 1.2 LjFTb1 LjFTb1 1 1 phyB 0.8 0.8 0.6 0.6 phyB 0.4 MG20 0.4 Relative expression Relative

Fig. 5. Comparison of Flowering Phenotype between Wild-Type expression Relative 0.2 0.2 MG20 (MG-20) and a phyB Mutant Grown on Soil in LDs (16 h light/8 h dark) for 45 d. 0 0 The positions of the node that set the first flower were counted 03691215182124 0 3 6 9 12 15 18 21 24 (n ¼ 11 for MG-20, n ¼ 14 for phyB), and the average numbers of Time (h) Time (h) nodes were calculated, as indicated (#7 for MG-20, #13 for phyB). Scale bar, 5 cm. Fig. 7. Diurnal Expression Profile of LjFTa and LjFTb in Wild-Type (MG-20) and phyB Grown under LDs and SDs for 14 d. mRNA samples were prepared at every 3 h interval. The diurnal expression profiles of LjFTa (A and B) and LjFTb (C and D) were examined by means of qRT-PCR analyses, according to the method described in the Materials and Methods. Relative expression levels are shown as mean values SD (n ¼ 3), for which the maximum value of phyB was taken as 1.0. The dark periods are indicated with shadings.

awaits extensive future studies, taking into consideration that there might be significant difference between annual and perennial plants with regard to their underlying molecular mechanisms of flowering.52) Together with the current progress in the relevant study of other model legumes (e.g., soybean and garden pea), the results of LDs MG-20 SDs MG-20 this study will provide a platform on which one can address a number of interesting issues with regard to the control of reproduction of legumes, which is of agronomical importance to gain improved crop yields. Acknowledgments

We are grateful to Dr. Akihiro Suzuki (Saga Uni- versity) for L. japonicus phyB mutant seeds. We also thank Dr. Mari Banba (Nagoya University, Japan) and Dr. Masayoshi Kawaguchi (National Institute for Basic Biology, Okazaki, Japan) for protocols of general manipulation of L. japonicus and kind advice as to establishing its growth conditions. This work was supported by Japan Society of the Promotion of Science phyB phyB LDs SDs (no. 23580133 and no. 23012018 to T.Y., no. 20370018 to T.M.). Fig. 6. Flowering Phenotype of 60-d-Old Wild-Type (MG-20) and phyB Mutant L. japonicus Plants. Wild-type (MG-20) plants set flowers specifically in LDs, References whereas phyB mutant plants showed a phenotype of early flowering regardless of photoperiod. LDs denotes long day condition (16 h 1) de Montaigu A, To´th R, and Coupland G, Trends Genet., 26, light/8 h dark), and SDs denotes short day condition (11.5 h light/ 296–306 (2010). 12.5 h dark). Some flowers, but not all, are indicated with arrows. 2) Hotta CT, Gardner MJ, Hubbard KE, Baek SJ, Dalchau N, Scale bar, 5 cm. Suhita D, Dodd AN, and Webb AA, Plant Cell Environ., 30, 333–349 (2007). 3) Fornara F, de Montaigu A, and Coupland G, Cell, 141, 550, 550 e551–e552 (2010). Regulation of Flowering Time in L. japonicus 753 4) Imaizumi T, Curr. Opin. Plant Biol., 13, 83–89 (2010). I, Saurat O, Scarpelli C, Schiex T, Segurens B, Severin AJ, 5) Hut RA and Beersma DG, Philos. Trans. R. Soc. Lond. B Biol. Sherrier DJ, Shi R, Sims S, Singer SR, Sinharoy S, Sterck L, Sci., 366, 2141–2154 (2011). Viollet A, Wang BB, Wang K, Wang M, Wang X, Warfsmann 6) McClung CR, BMC Biol., 8, 55 (2010). J, Weissenbach J, White DD, White JD, Wiley GB, Wincker P, 7) Song YH, Ito S, and Imaizumi T, Curr. Opin. Plant Biol., 13, Xing Y, Yang L, Yao Z, Ying F, Zhai J, Zhou L, Zuber A, 594–603 (2010). Deńarie´ J, Dixon RA, May GD, Schwartz DC, Rogers J, Que´tier 8) Valverde F, J. Exp. Bot., 62, 2453–2463 (2011). F, Town CD, and Roe BA, Nature, 480, 520–524 (2011). 9) Abelenda JA, Navarro C, and Prat S, Curr. Opin. Biotechnol., 31) Kim MY, Lee S, Van K, Kim TH, Jeong SC, Choi IY, Kim DS, 22, 287–292 (2011). Lee YS, Park D, Ma J, Kim WY, Kim BC, Park S, Lee KA, Kim 10) Pin PA, Benlloch R, Bonnet D, Wremerth-Weich E, Kraft T, DH, Kim KH, Shin JH, Jang YE, Kim KD, Liu WX, Chaisan T, Gielen JJ, and Nilsson O, Science, 330, 1397–1400 (2010). Kang YJ, Lee YH, Kim KH, Moon JK, Schmutz J, Jackson SA, 11) Tsuji H, Taoka K, and Shimamoto K, Curr. Opin. Plant Biol., Bhak J, and Lee SH, Proc. Natl. Acad. Sci. USA, 107, 22032– 14, 45–52 (2011). 22037 (2010). 12) Imaizumi T and Kay SA, Trends Plant Sci., 11, 550–558 (2006). 32) Schmutz J, Cannon SB, Schlueter J, Ma J, Mitros T, Nelson W, 13) Putterill J, Robson F, Lee K, Simon R, and Coupland G, Cell, Hyten DL, Song Q, Thelen JJ, Cheng J, Xu D, Hellsten U, May 80, 847–857 (1995). GD, Yu Y, Sakurai T, Umezawa T, Bhattacharyya MK, Sandhu 14) Wenkel S, Turck F, Singer K, Gissot L, Le Gourrierec J, D, Valliyodan B, Lindquist E, Peto M, Grant D, Shu S, Samach A, and Coupland G, Plant Cell, 18, 2971–2984 (2006). Goodstein D, Barry K, Futrell-Griggs M, Abernathy B, Du J, 15) Sawa M, Nusinow DA, Kay SA, and Imaizumi T, Science, 318, Tian Z, Zhu L, Gill N, Joshi T, Libault M, Sethuraman A, Zhang 261–265 (2007). XC, Shinozaki K, Nguyen HT, Wing RA, Cregan P, Specht J, 16) Yanovsky MJ and Kay SA, Nat. Rev. Mol. Cell Biol., 4, 265– Grimwood J, Rokhsar D, Stacey G, Shoemaker RC, and Jackson 275 (2003). SA, Nature, 463, 178–183 (2010). 17) Yanovsky MJ and Kay SA, Nature, 419, 308–312 (2002). 33) Weller JL, Hecht V, Liew LC, Sussmilch FC, Wenden B, 18) Valverde F, Mouradov A, Soppe W, Ravenscroft D, Samach A, Knowles CL, and Vander Schoor JK, J. Exp. Bot., 60, 2493– and Coupland G, Science, 303, 1003–1006 (2004). 2499 (2009). 19) Jang S, Marchal V, Panigrahi KC, Wenkel S, Soppe W, Deng 34) Udvardi MK, Tabata S, Parniske M, and Stougaard J, Trends XW, Valverde F, and Coupland G, EMBO J., 27, 1277–1288 Plant Sci., 10, 222–228 (2005). (2008). 35) Young ND, Mudge J, and Ellis TH, Curr. Opin. Plant Biol., 6, 20) Liu LJ, Zhang YC, Li QH, Sang Y, Mao J, Lian HL, Wang L, 199–204 (2003). and Yang HQ, Plant Cell, 20, 292–306 (2008). 36) Hecht V, Knowles CL, Vander Schoor JK, Liew LC, Jones SE, 21) Corbesier L and Coupland G, J. Exp. Bot., 57, 3395–3403 Lambert MJ, and Weller JL, Plant Physiol., 144, 648–661 (2006). (2007). 22) Mizuno T and Nakamichi N, Plant Cell Physiol., 46, 677–685 37) Hecht V, Laurie RE, Vander Schoor JK, Ridge S, Knowles CL, (2005). Liew LC, Sussmilch FC, Murfet IC, Macknight RC, and Weller 23) Nakamichi N, Plant Cell Physiol., 52, 1709–1718 (2011). JL, Plant Cell, 23, 147–161 (2011). 24) Ono N, Ishida K, Yamashino T, Nakanishi H, Sato S, Tabata S, 38) Liew LC, Hecht V, Laurie RE, Knowles CL, Vander Schoor JK, and Mizuno T, Plant Cell Physiol., 51, 1800–1814 (2010). Macknight RC, and Weller JL, Plant Cell, 21, 3198–3211 25) Yamashino T, Biosci. Biotechnol. Biochem., 77, 10–16 (2013). (2009). 26) Escaray FJ, Menendez AB, Garriz A, Pieckenstain FL, Estrella 39) Sun H, Jia Z, Cao D, Jiang B, Wu C, Hou W, Liu Y, Fei Z, Zhao MJ, Castagno LN, Carrasco P, Sanjuan J, and Ruiz OA, Plant D, and Han T, PLoS ONE, 6, e29238 (2011). Sci., 182, 121–133 (2012). 40) Jung CH, Wong CE, Singh MB, and Bhalla PL, PLoS ONE, 7, 27) Cannon SB, May GD, and Jackson SA, Plant Physiol., 151, e38250 (2012). 970–977 (2009). 41) Zhao S and Fernald RD, J. Comput. Biol., 12, 1047–1064 28) Sato S, Isobe S, and Tabata S, Curr. Opin. Plant Biol., 13, 146– (2005). 152 (2010). 42) Jones DT, Taylor WR, and Thornton JM, Comput. Appl. Biosci., 29) Sato S, Nakamura Y, Kaneko T, Asamizu E, Kato T, Nakao M, 8, 275–282 (1992). Sasamoto S, Watanabe A, Ono A, Kawashima K, Fujishiro T, 43) Saitou N and Nei M, Mol. Biol. Evol., 4, 406–425 (1987). Katoh M, Kohara M, Kishida Y, Minami C, Nakayama S, 44) Kojima S, Banno H, Yoshioka Y, Oka A, Machida C, and Nakazaki N, Shimizu Y, Shinpo S, Takahashi C, Wada T, Machida Y, DNA Res., 6, 407–410 (1999). Yamada M, Ohmido N, Hayashi M, Fukui K, Baba T, 45) Shannon S and Meeks-Wagner DR, Plant Cell, 3, 877–892 Nakamichi T, Mori H, and Tabata S, DNA Res., 15, 227–239 (1991). (2008). 46) Ahn JH, Miller D, Winter VJ, Banfield MJ, Lee JH, Yoo SY, 30) Young ND, Debelle´ F, Oldroyd GE, Geurts R, Cannon SB, Henz SR, Brady RL, and Weigel D, EMBO J., 25, 605–614 Udvardi MK, Benedito VA, Mayer KF, Gouzy J, Schoof H, Van (2006). de Peer Y, Proost S, Cook DR, Meyers BC, Spannagl M, 47) Khanna R, Kronmiller B, Maszle DR, Coupland G, Holm M, Cheung F, De Mita S, Krishnakumar V, Gundlach H, Zhou S, Mizuno T, and Wu SH, Plant Cell, 21, 3416–3420 (2009). Mudge J, Bharti AK, Murray JD, Naoumkina MA, Rosen B, 48) Ledger S, Strayer C, Ashton F, Kay SA, and Putterill J, Plant J., Silverstein KA, Tang H, Rombauts S, Zhao PX, Zhou P, Barbe 26, 15–22 (2001). V, Bardou P, Bechner M, Bellec A, Berger A, Berge`sH, 49) Suzuki A, Suriyagoda L, Shigeyama T, Tominaga A, Sasaki M, Bidwell S, Bisseling T, Choisne N, Couloux A, Denny R, Hiratsuka Y, Yoshinaga A, Arima S, Agarie S, Sakai T, Inada S, Deshpande S, Dai X, Doyle JJ, Dudez AM, Farmer AD, Fouteau Jikumaru Y, Kamiya Y, Uchiumi T, Abe M, Hashiguchi M, S, Franken C, Gibelin C, Gish J, Goldstein S, Gonza´lez AJ, Akashi R, Sato S, Kaneko T, Tabata S, and Hirsch AM, Proc. Green PJ, Hallab A, Hartog M, Hua A, Humphray SJ, Jeong Natl. Acad. Sci. USA, 108, 16837–16842 (2011). DH, Jing Y, Jo¨cker A, Kenton SM, Kim DJ, Klee K, Lai H, 50) Schepens I, Duek P, and Fankhauser C, Curr. Opin. Plant Biol., Lang C, Lin S, Macmil SL, Magdelenat G, Matthews L, 7, 564–569 (2004). McCorrison J, Monaghan EL, Mun JH, Najar FZ, Nicholson C, 51) Cerdan PD and Chory J, Nature, 423, 881–885 (2003). Noirot C, O’Bleness M, Paule CR, Poulain J, Prion F, Qin B, Qu 52) Albani MC and Coupland G, Curr. Top. Dev. Biol., 91, 323–348 C, Retzel EF, Riddle C, Sallet E, Samain S, Samson N, Sanders (2010).