Functional Redundancy of FLOWERING LOCUS T 3b in Soybean Flowering Time Regulation

Photoperiodic flowering is an important agronomic trait that determines adaptability and yield in soybean and is strongly influenced by FLOWERING LOCUS T (FT) genes. Due to the presence of multiple FT homologs in the genome, their functions in soybean are not fully understood. Here, we show that GmFT3b exhibits functional redundancy in regulating soybean photoperiodic flowering. Bioinformatic analysis revealed that GmFT3b is a typical floral inducer FT homolog and that the protein is localized to the nucleus. Moreover, GmFT3b expression was induced by photoperiod and circadian rhythm and was more responsive to long-day (LD) conditions. We generated a homozygous ft3b knockout and three GmFT3b-overexpressing soybean lines for evaluation under different photoperiods. There were no significant differences in flowering time between the wild-type, the GmFT3b overexpressors, and the ft3b knockouts under natural long-day, short-day, or LD conditions. Although the downstream flowering-related genes GmFUL1 (a, b), GmAP1d, and GmLFY1 were slightly down-regulated in ft3b plants, the floral inducers GmFT5a and GmFT5b were highly expressed, indicating potential compensation for the loss of GmFT3b. We suggest that GmFT3b acts redundantly in flowering time regulation and may be compensated by other FT homologs in soybean.


Introduction
The change from vegetative to reproductive growth is a critical developmental transition in the life of flowering plants. Time to flowering directly influences crop maturity and determines adaptability to diverse geographic regions [1]. Proper flowering time is a prerequisite for soybean yield, and a series of studies have focused on optimizing soybean flowering time to maintain productivity during introduction to new regions [1][2][3].
In Arabidopsis thaliana, a long-day (LD) plant, flowering is induced by external and endogenous cues such as photoperiod, gibberellin levels, vernalization, and autonomous flowering signaling [4]. Among these cues, photoperiodical variation directly affects not only flowering time but also the podding stage and time to maturity [5]. Florigen is a compound produced in leaves and transmitted to the shoot apical meristem (SAM) to initiate flowering [6,7]. In Arabidopsis, florigen is the key regulatory integration factor in flowering induction pathways [7]. Recent research has shown that FLOWERING LOCUS T (FT) homologs (FTs), a family of phosphatidylethanolamine-binding proteins (PEBPs), have florigen function [6,7]. Several photoperiodical regulatory pathways determine flowering: GIGANTEA (GI), CONSTANS (CO), and FT function as central components in triggering flowering under LD conditions [7][8][9]. In brief, the circadian clock GI activates CO expression by binding to its promoter under LD photoperiod conditions, but not under short-day (SD) photoperiod conditions [10]. Accumulated CO protein directly activates the expression of FT in leaves [9,10]. Furthermore, FT moves to the SAM, where it forms a complex by interacting with FLOWERING LOCUS D (FD) [11,12]. The FT-FD complex induces expression of flowering-related genes, such as SUPPRESSOR OF OVEREXPRESSION OF CO 1 (SOC1), FRUITFULL (FUL), and LEAFY (LFY), and finally initiates the floral meristem identity gene APETALA1 (AP1) to stimulate flowering [11,[13][14][15][16]. Likewise, in rice, the FT homolog protein Heading date 3a (Hd3a) interacts with the FD homolog OsFD1 in the SAM, assisted by 14-3-3 proteins [17]. These complexes then bind to the promoter of OsMADS15 (an AP1 homolog) to activate floral transition in rice, which highlights conservation of the flowering regulatory module (FT/FD-AP1) between different photoperiodic plants [17,18].
Soybean, a typical short-day (SD) plant, is particularly sensitive to photoperiods and is considered a classical photoperiodic model plant [5]. Day length determines plant flowering time; for example, in soybean, SD accelerates flowering, whereas LD represses flower bud formation [19,20]. Soybeans can be cultivated at latitudes ranging from~20 • N in the south to~50 • N in the north [21]. Unfortunately, most soybean varieties have limited latitude adaptability, resulting in a narrow cultivation area for each variety [5]. Flowering time and maturity period are the key agronomic traits that directly determine the yield and quality of soybeans [5,19]. Thus, an in-depth understanding of the roles of photoperiod genes at the molecular level is of great significance for adaptation of soybean varieties to diverse geographic regions.
To date, several key genetic loci have been identified that have large effects on flowering time and maturity period in soybean. These include E1-E11, J, Tof11, and Tof12, which comprise the phytochrome-clock-related gene E1-FTs flowering pathway [3,5,20,[22][23][24]. FTs redundantly control photoperiod-regulated flowering in soybean. Specifically, GmFT2a (Glyma16g26660) and GmFT5a (Glyma16g04830) have been proven to effectively promote flowering through photoperiod regulation [20,25]. Interestingly, GmFT2a has a stronger effect on floral initiation under SD conditions, whereas GmFT5a has a stronger effect on flower induction under LD conditions [25]. GmFT2a and GmFT5a interact with both GmFDL12 and GmFDL19, but only GmFT5a interacts with GmFDL06 [26]. This may cause functional differentiation of FT genes. In addition, GmFT2b (Glyma16g26690) promotes flowering under LD [2]. GmFT5b (Glyma19g28400) promotes early flowering in Arabidopsis [27]. In contrast, GmFT1a (Glyma18g53680) was shown to be upregulated by E1 and to delay flowering and maturity, confirming the identity of flowering inhibitors [28]. Similarly, GmFT4 (Glyma08g47810) also delays flowering in Arabidopsis plants and may be the relevant gene in the E10 locus [29,30]. Duplication and divergence of ancestral FT genes have produced multiple flowering regulatory proteins, some of which have antagonistic functions in flowering in sugar beet [31], apple [32], and onion [33]. This demonstrates that FTs have undergone diversified functional changes during the evolution of various crops and that photoperiod-dependent flowering is strictly controlled by coordinated expression of FT family genes.
Soybean FTs have been evaluated in various species. For example, ectopic expression of GmFT3b induces early flowering in Arabidopsis [27]. In this study, we found that GmFT3b redundantly participated in soybean photoperiodic flowering. First, GmFT3b expression was confirmed to be photoperiod-dependent and more responsive to LD conditions. We generated GmFT3b-overexpressing soybean plants and ft3b knockout plants and evaluated them under various photoperiods. Based on the expression profiles of flowering-related genes, we here propose a model where FTs redundantly regulate flowering in soybean.

Identification of the FT Homolog GmFT3b in Soybean
According to the Glyma19g28390 gene sequence in the Phytozome database, GmFT3b was cloned from the soybean variety Jack and sequenced. The 2336 bp genomic sequence of GmFT3b contained four introns and three exons (Appendix A), including an open reading frame (ORF) 546 bp in length. The ORF encoded a product 175 aa residues in length with a molecular weight of 19.72 kDa.
Some FTs have developed opposing functions during evolution, antagonizing plant flowering processes in Arabidopsis [27], soybean [28], and tobacco [34]. Thus, we performed multiple sequence alignment of GmFT3b with other FTs that have been well characterized in multiple species [35]. The results show that GmFT3b belongs to the FT family, with a highly conserved PEBP domain from 32 aa to 162 aa. GmFT3b also contains a tyrosine residue at position 134, which is consistent with all flowering inducer FTs except GmFT5a. In contrast, repressor FTs (AtFTL1, GmFT4, and NtFT1) are not tyrosine residue at position 134 ( Figure 1A,B).
of GmFT3b contained four introns and three exons (Appendix A), including an open reading frame (ORF) 546 bp in length. The ORF encoded a product 175 aa residues in length with a molecular weight of 19.72 kDa.
Some FTs have developed opposing functions during evolution, antagonizing plant flowering processes in Arabidopsis [27], soybean [28], and tobacco [34]. Thus, we performed multiple sequence alignment of GmFT3b with other FTs that have been well characterized in multiple species [35]. The results show that GmFT3b belongs to the FT family, with a highly conserved PEBP domain from 32 aa to 162 aa. GmFT3b also contains a tyrosine residue at position 134, which is consistent with all flowering inducer FTs except GmFT5a. In contrast, repressor FTs (AtFTL1, GmFT4, and NtFT1) are not tyrosine residue at position 134 ( Figure 1A,B).
Phylogenetic analysis was performed with GmFT3b and selected PEBP proteins from soybean, Arabidopsis, Beta vulgaris, and Malus domestica. Both GmFT3b and GmFT3a clustered with inducer FTs such as GmFT2a, GmFT2b, and Arabidopsis FT ( Figure 1C). GmFT3b therefore appears evolutionarily conserved and closely related to soybean flowering inducer GmFT2a, suggesting that GmFT3b contains conserved elements that may positively regulate flowering.  Phylogenetic analysis was performed with GmFT3b and selected PEBP proteins from soybean, Arabidopsis, Beta vulgaris, and Malus domestica. Both GmFT3b and GmFT3a clustered with inducer FTs such as GmFT2a, GmFT2b, and Arabidopsis FT ( Figure 1C). GmFT3b therefore appears evolutionarily conserved and closely related to soybean flowering in-ducer GmFT2a, suggesting that GmFT3b contains conserved elements that may positively regulate flowering.

GmFT3b Is Localized to the Nucleus
Previous studies showed that the FT protein was localized to the nucleus and acted as an integration factor in soybean [2,28]. We generated a construct containing a GmFT3b-GFP fusion gene driven by the 35S-CaMV promoter (the PTF101-GFP-GmFT3b vector), then assessed the subcellular localization of GmFT3b-GFP in tobacco leaves with transient expression of the plasmid. As expected, GmFT3b-GFP was mainly expressed in the nucleus, as demonstrated by colocalization with the red nuclear marker fluorescent fusion protein NM-mCherry ( Figure 2A). In addition, immunoblot analysis of total nuclear proteins confirmed that GmFT3b-GFP/GFP were expressed in the nucleus of tobacco leaves as expected ( Figure 2B,C).

GmFT3b Is Localized to the Nucleus
Previous studies showed that the FT protein was localized to the nucleus and acted as an integration factor in soybean [2,28]. We generated a construct containing a GmFT3b-GFP fusion gene driven by the 35S-CaMV promoter (the PTF101-GFP-GmFT3b vector), then assessed the subcellular localization of GmFT3b-GFP in tobacco leaves with transient expression of the plasmid. As expected, GmFT3b-GFP was mainly expressed in the nucleus, as demonstrated by colocalization with the red nuclear marker fluorescent fusion protein NM-mCherry ( Figure 2A). In addition, immunoblot analysis of total nuclear proteins confirmed that GmFT3b-GFP/GFP were expressed in the nucleus of tobacco leaves as expected ( Figure 2B,C).

Day Length and Circadian Rhythm Regulate the Expression Pattern of GmFT3b
FT functions as a florigen to induce floral transition, and its expression patterns are regulated by photoperiod and circadian rhythm [20,28]. Previously, we evaluated two varieties with extreme photoperiod response phenotypes, the early-flowering variety Heihe 27 (HH27) and the late-flowering variety ZiGongDongDou (ZGDD). Here, the diurnal expression patterns of GmFT3b were analyzed in leaves of HH27 and ZGDD plants under various photoperiodic conditions. In both varieties, GmFT3b showed diurnal circadian

Day Length and Circadian Rhythm Regulate the Expression Pattern of GmFT3b
FT functions as a florigen to induce floral transition, and its expression patterns are regulated by photoperiod and circadian rhythm [20,28]. Previously, we evaluated two varieties with extreme photoperiod response phenotypes, the early-flowering variety Heihe 27 (HH27) and the late-flowering variety ZiGongDongDou (ZGDD). Here, the diurnal expression patterns of GmFT3b were analyzed in leaves of HH27 and ZGDD plants under various photoperiodic conditions. In both varieties, GmFT3b showed diurnal circadian rhythm under SD and LD conditions ( Figure 3A,B). Under SD conditions, GmFT3b remained highly expressed in both varieties and peaked at 4 h after dark in HH27, but not in ZGDD plants.
GmFT3b expression patterns were also comparable to one another in HH27 and ZGDD under LD conditions, although GmFT3b levels peaked 2 h earlier in ZGDD than in HH27. The results suggested that GmFT3b expression was regulated by circadian rhythm and was more sensitive to the induced LD photoperiod compared to the SD. rhythm under SD and LD conditions ( Figure 3A,B). Under SD conditions, GmFT3b remained highly expressed in both varieties and peaked at 4 h after dark in HH27, but not in ZGDD plants. GmFT3b expression patterns were also comparable to one another in HH27 and ZGDD under LD conditions, although GmFT3b levels peaked 2 h earlier in ZGDD than in HH27. The results suggested that GmFT3b expression was regulated by circadian rhythm and was more sensitive to the induced LD photoperiod compared to the SD.

Evaluation of GmFT3b-Overexpressing Soybean Plants under Different Photoperiods
To investigate the effects of GmFT3b in plant flowering, we created three GmFT3boverexpressing soybean lines (named OE3, OE5, OE6) via Agrobacterium-mediated transformation. The GmFT3b-overexpressing plants were identified via PCR and LibertyLink strips ( Figure 4A,B). Western blot analysis indicated that the GFP-GmFT3b fusion protein was expressed in the T2 generations of OE3, OE5, and OE6 plants, but not in WT plants ( Figure 4C). These results indicated that GmFT3b was inserted into the soybean genome and successfully translated. The

Evaluation of GmFT3b-Overexpressing Soybean Plants under Different Photoperiods
To investigate the effects of GmFT3b in plant flowering, we created three GmFT3boverexpressing soybean lines (named OE3, OE5, OE6) via Agrobacterium-mediated transformation. The GmFT3b-overexpressing plants were identified via PCR and LibertyLink strips ( Figure 4A,B). Western blot analysis indicated that the GFP-GmFT3b fusion protein was expressed in the T 2 generations of OE3, OE5, and OE6 plants, but not in WT plants ( Figure 4C). These results indicated that GmFT3b was inserted into the soybean genome and successfully translated.

ft3b Knockout Did Not Affect Flowering Time
To further investigate the function of GmFT3b in soybean, we used CRISPR/Cas9 to generate an ft3b knockout soybean line. First, the genome editing target was placed near the start codon of the first exon of GmFT3b ( Figure 5A). The genomic target sequence of  Figure 4G).

ft3b Knockout Did Not Affect Flowering Time
To further investigate the function of GmFT3b in soybean, we used CRISPR/Cas9 to generate an ft3b knockout soybean line. First, the genome editing target was placed near the start codon of the first exon of GmFT3b ( Figure 5A). The genomic target sequence of GmFT3b near the cleavage site was amplified and confirmed by sequencing. Finally, we obtained a homozygous ft3b mutant with a 72 bp deletion that resulted in a missing start codon, preventing normal translation ( Figure 5B,C).

Expression of Downstream Flowering-Related Genes
It was previously reported that FT positively regulates expression of SOC1, FUL, and LFY homologs in the SAM of Arabidopsis and soybean [13][14][15]25]. We therefore used quantitative reverse-transcription PCR (qRT-PCR) to assess the expression profiles of FTs and downstream flowering-related genes. Under NLD photoperiod, as expected, GmFT3b expression was significantly higher in GmFT3b overexpressors and lower in ft3b knockouts  Figure 5G). These results indicated that mutation of GmFT3b alone did not alter soybean flowering performance under several photoperiodic environments.

Expression of Downstream Flowering-Related Genes
It was previously reported that FT positively regulates expression of SOC1, FUL, and LFY homologs in the SAM of Arabidopsis and soybean [13][14][15]25]. We therefore used quantitative reverse-transcription PCR (qRT-PCR) to assess the expression profiles of FTs and downstream flowering-related genes. Under NLD photoperiod, as expected, GmFT3b expression was significantly higher in GmFT3b overexpressors and lower in ft3b knockouts compared with WT plants (Figure 6). Expression levels of GmAP1d, GmLFY1, and GmFUL1 (a, b) were significantly lower in ft3b mutant plants but not in GmFT3b overexpressors, with the exception of FUL1b (Figure 6).

Discussion
Soybean provides more than a quarter of the total protein consumed by humans and animals worldwide, meaning that soybean directly affects the quality of human life. Cultivated soybeans are paleo tetraploids that were domesticated from wild soybean (Glycine soja Sieb. et Zucc.) ~5000 years ago [3,36]. Thus, the soybean genome is complex, and ~75% of protein-coding genes have multiple copies [37]. For example, more than 11 FTs are pre- The genetic compensation response (GCR) mechanism can significantly increase expression of other FTs in single ft soybean mutants [1]. We therefore also measured expression of FTs in the FT3b knockouts and overexpressors. In GmFT3b-overexpressing soybean plants, no significant differences in FT levels were observed compared with WT plants. As expected, GmFT5a and GmFT5b, two flowering inducers, were upregulated in ft3b mutant plants. Taken together, the results showed that neither increasing nor decreasing GmFT3b expression levels affected expression of downstream flowering-related genes in soybean.

Discussion
Soybean provides more than a quarter of the total protein consumed by humans and animals worldwide, meaning that soybean directly affects the quality of human life. Cultivated soybeans are paleo tetraploids that were domesticated from wild soybean (Glycine soja Sieb. et Zucc.)~5000 years ago [3,36]. Thus, the soybean genome is complex, and~75% of protein-coding genes have multiple copies [37]. For example, more than 11 FTs are present in the soybean genome [19,20,38]. However, most studies have focused on only two of these, GmFT2a and GmFT5a [25,26,39]. The other FTs were discovered more than ten years ago, at which time studies were limited in Arabidopsis [27]. Therefore, in-depth research to fully explain the function of other FTs (such as GmFT3b) is necessary to facilitate the understanding of the photoperiodic flowering pathways in soybean.
After gene duplication events, the redundant genes can undergo functional conservation, neofunctionalization, subfunctionalization, or functional degradation during evolution [40,41]. As a result, some FT homologs display opposing biochemical actions [28,29,32,33,35]. GmFT3b is a member of the PEBP homologs and shows high sequence similarity with flower-inducing FTs ( Figure 1A,B). The GmFT3b-GFP fusion protein was primarily observed in the nucleus (Figure 2A-C), indicating that GmFT3b was localized to the nucleus with other FT proteins [2,28]. Photoperiod and circadian rhythm regulate the overall transcription and diurnal expression patterns of FT genes, thereby systematically and accurately controlling the transition from vegetative to reproductive growth [20,28]. Three expression patterns of FT were preliminarily identified in cultivated soybean varieties, and GmFT3b belonged to the photoperiod-independent group [28]. However, GmFT3b was determined in the present study to be photoperiod-dependent, with expression patterns depending on day length and changing more strongly under LD than SD conditions ( Figure 3A,B). In addition, considering that GmFT3b promotes flowering in Arabidopsis [27], GmFT3b might retain the function of controlling flowering in soybean.
Several recent studies showed that GmFT2a and GmFT5a are functionally equivalent to the Arabidopsis FT that induce early flowering, with both able to rescue the Arabidopsis ft-10 mutant [20,37]. Likewise, Lee et al. found that GmFT3b acted as a strong flowering inducer in Arabidopsis [27]. In contrast, GmFT3b expression had a negligible influence on flowering in soybean; there were no significant differences in floral transition time between WT, three GmFT3b-overexpressing soybean lines, and an ft3b knockout line under NLD, SD, or LD conditions. Interestingly, soybean plants overexpressing GmFT5a failed to induce early flowering under SD conditions because only GmFT2a also promoted flowering [25]. Similarly, we speculated that the expression levels of GmFT2a, GmFT5a and GmFT5b were sufficient to induce flowering under LD photoperiods, although the GmFT3b was overexpressed in soybean. It would intuitively be expected that knocking out FT genes would drastically affect downstream flowering-related genes [1,25], but the ft3b knockouts did not show altered expression of AP1 (a-c), LFY2, FULa, or SOC1 (a, b) compared with the WT (Figure 6). GCR is widespread in animals and plants, which leads to weaker phenotypes in single mutants [1,42]. Strikingly, Gmft2a and Gmft5a single mutants displayed weak roles in activating flowering compared with the double mutants, while the expression levels of other FT genes were increased [1]. Similarly, GmFT5a and GmFT5b were upregulated in ft3b plants in this study, suggesting GmFT5a and GmFT5b compensate for the function of GmFT3b. In particular, the double mutations of ft3b with Gmft5a or Gmft5b should be created to further investigate the function of GmFT3b in future studies. Taken together, the results suggested that a loss-of-function mutation in GmFT3b was compensated by other FT genes in photoperiodic flowering.
Short, synchronous flowering time is considered a critical trait for crop yield. Interestingly, the domesticated FT alleles slightly delay flowering compared with wild alleles, which may be a result of losing FT duplicates [37,43]. Moreover, loss of function in FTs by natural or artificial mutation clearly determined the ecological adaptation range of soybean [1,20,25]. Thus, variation in FTs is an important basis for diversity in flowering time and maturity, which contribute to soybean geographical adaptability. We previously showed that GmFT3b may have undergone breeding selection, and its haplotypes were associated with flowering time and maturity [44]. However, GmFT3b was functionally redundant in regulation of flowering time under our experimental conditions. In addition, almost all identified polymorphisms were distributed in gene regulatory regions (the 5 UTR, 3 UTR, and intron regions) [44], implying that differences in GmFT3b between soybean varieties are at the transcriptional level. Therefore, it is necessary to study the function of GmFT3b in different soybean backgrounds in future studies. Besides, environmental stresses including drought, salinity, heat, and nutrient stress can also affect the flowering [45]. Although we explored the function of GmFT3b under different photoperiods, it would be interesting to study the role of GmFT3b under environmental stress in the future. Taken together, we propose a model where FTs redundantly regulate flowering in soybean ( Figure 7). Under LD photoperiod conditions, the light signal is received by photoreceptors (E3 and E4), then transmitted to FTs through a photoperiod-dependent pathway [3,7,19,20]. GmFT5a, GmFT2a, GmFT5b, and GmFT3b act as floral activators, with GmFT5a having a decisive influence and the ability to compensate for the function of GmFT3b. Floral activators are required to counteract flowering inhibitors to activate downstream flowering-related gene expression and subsequently induce flowering. Further studies are needed to determine how GmFT3b is compensated in soybean. Taken together, we propose a model where FTs redundantly regulate flowering in soy bean ( Figure 7). Under LD photoperiod conditions, the light signal is received by photo receptors (E3 and E4), then transmitted to FTs through a photoperiod-dependent pathway [3,7,19,20]. GmFT5a, GmFT2a, GmFT5b, and GmFT3b act as floral activators, with GmFT5 having a decisive influence and the ability to compensate for the function of GmFT3b. Flo ral activators are required to counteract flowering inhibitors to activate downstream flow ering-related gene expression and subsequently induce flowering. Further studies ar needed to determine how GmFT3b is compensated in soybean.

Plant Materials and Growth Conditions
Soybean variety Jack was used in this study. The seeds of Jack, GmFT3b-overexpress ing, and ft3b mutants were sown in plastic pots, which were placed in standard long-day

Plant Materials and Growth Conditions
Soybean variety Jack was used in this study. The seeds of Jack, GmFT3b-overexpressing, and ft3b mutants were sown in plastic pots, which were placed in standard long-day (

Bioinformatics Analysis
The amino acid sequence of FT and TFL homologs was retrieved from NCBI (https: //www.ncbi.nlm.nih.gov/; accessed on 19 May 2021) and Phytozome (https://phytozome. jgi.doe.gov/pz/portal.html/; accessed on 19 May 2021) database. The multiple sequence alignment was performed using the Clustal Omega (https://www.ebi.ac.uk/Tools/msa/ clustalo/; accessed on 19 May 2021) web program. The phylogenetic tree was constructed using the maximum likelihood method by MEGA X software.

Generation of GmFT3b-Overexpressing Soybean Plants
The soybean variety was used for transformation according to the protocol reported previously [46]. To generate the GmFT3b-overexpressing soybean plants, the full-length GmFT3b sequence was cloned from the cDNA library of soybean shoot apex using GmFT3b-FQ-F/R primers and then subcloned to the overexpression vector PTF101-GFP with the GmFT3b-101F/R primers, named PTF101-GFP-GmFT3b. Subsequently, the constructed vector was directly transformed into Agrobacterium tumefaciens strain EHA101 and generated T 0 generation of GmFT3b-overexpressing plant lines. Next, the T 1 generations of GmFT3boverexpressing seedlings were screened by glufosinate herbicide and confirmed by PCR using the FT3b-JC351F/R primers, and the T 2 generations of GmFT3b-overexpressing plants were used in this study. All PCR primers were listed in Appendix B.

CRISPR-Mediated Mutation of GmFT3b
The sgRNA of GmFT3b (GmFT3b-TS: AGAGGGTTCCTACTACCGCCAGG) was selected by the CRISPR-P web server (http://cbi.hzau.edu.cn/cgi-bin/CRISPR; Accessed on 18 July 2018). Then, the oligo sequence of GmFT3b-TS was synthesized and inserted into the CRISPR/Cas9 vector that was driven by the AtU6 promoter. Next, the CRISPR/Cas9 vector was transformed into the A. tumefaciens strain EHA105 and then introduced into soybean variety Jack through Agrobacterium-mediated transformation as described previously [46]. All T 0 GmFT3b-overexpressing were screened by PCR with the detection primers (FT3b-439F/FT3b-439R) and subsequently confirmed by sequence. In addition, an ft3b homozygous mutant was detected by both PCR and Bar test strip methods as described previously [47]. All primers used in this study are listed in Appendix B.

Gene Expression Analysis
The shoot apex tissues of soybean were harvested at 20 DAE (natural long-day conditions), 14 DAE (short-day conditions), and 35 DAE (long-day conditions), respectively. According to the manufacturer's instructions, total RNA was extracted by TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Then,~1 µg total RNA was reverse transcribed into cDNA via HiScript ® III RT SuperMix for qPCR kits (Vazyme, Nanjing, China). Quantitative realtime PCR (qRT-PCR) was performed on ABI QuantStudio 7 Flex (Applied Biosystems, San Francisco, CA, USA) with ChamQ SYBR qPCR Master Mix (Low ROX Premixed) (Vazyme, Nanjing, China). The qRT-PCR program followed the manufacturer's instructions, and all samples tested in expression analysis were verified with three technical replications. The qRT-PCR data were determined by 2 −∆∆CT methods [48], and the statistical significance of differences was analyzed by Microsoft Excel by using the one-way ANOVA method. All primers are listed in Appendix B.

Subcellular Localization of GmFT3b
The subcellular localization of GmFT3b was conducted in Nicotiana benthamiana plants using Agrobacterium-mediated transformation as described previously [49]. Briefly, N. benthamiana plants were grown in pots under long-day conditions (16 h/light, 8 h/dark; 22-30 • C). Then, the N. benthamiana leaves were injected with A. tumefaciens strain GV3101, which carried PTF101-GFP-GmFT3b/pTF101-GFP and nuclear marker plasmids together, respectively. After 48 h agroinfiltration, the N. benthamiana leaves were collected and imaged by FLUOVIEW FV3000 Confocal Laser Scanning Microscope (Olympus Corporation, Tokyo, Japan).

Western Blot Analysis
The first trifoliate leaf of GmFT3b-overexpressing seedlings was harvested and conserved under −80 • C. The total soluble proteins were extracted using the plant protein extraction protocol as described previously [50]. The total nuclear proteins were extracted using the nuclear and cytoplasmic extraction kit (CWBIO, Beijing, China). After protein denaturation, samples were separated by 10% SDS-PAGE and then analyzed by immunoblot using 1:3000-fold dilution anti-GFP mouse monoclonal antibody. In addition, the PVDF membrane was incubated with the One Step Western Kit HRP (CWBIO, Beijing, China). Finally, the immunized proteins were imaged under the Amersham Imager 600 (GE Healthcare, Little Chalfont, Buckinghamshire, UK) machine.

Conclusions
GmFT3b is a typical FT homolog. Soybean lines overexpressing GmFT3b and ft3b knockout soybean plants were used to investigate the effects of GmFT3b in regulation of photoperiodic flowering for the first time. Neither overexpression nor knockout of GmFT3b significantly affected flowering time or expression of downstream flowering-related genes. Based on these data, we suggest that other FT homologs are functionally redundant with GmFT3b in regulation of photoperiodic flowering and that the homologs may compensate for the loss of GmFT3b .  TATAAATAGCTAGCCTGGTGGGTGTTGTAGTACAATTGTTTGTGACACAAAGGCATA  TTTGTGTTTAGCAGTGTGACTCAGTGAAGTGTGTGAGTTAGTT ATGCCTGGCGGTAG   GAACCCTCTTGTTGTTGGGCGTGTTATAGGGGAAGTAATAGATCCCTTTGAAATTTC   TATTCCTTTCAGGGTCACCTATGGTAATAGAGAAGTGGGCAATGGTTGTGAGCTTAA  ACCTTCCCAAGTTGCCAACCAACCCAGAGTGAGTGTTGGTGGAGATGACCTCAGG  AACTTCTACACTATG GTAATTAGTTACATGAGTTAGTTATCATATTATATTTACATA  TATATAGTGCTTTGCTAAACAAAATTCTGTTTAATTTCGTTTCTTCTATCAAAG GTC  CTGGTGGATCCTGATGCTCCTAGCCCAAGTAACCCTAATTTCAGGGAGTACCTTCAT  TG GTGAGTACATACCTTGCATACCGTTAGATGAACCAAATTTACATATAAGAGAG  ACTTTAATATTACATTATAACTAATACCCTTCCCTCTATGGATTTTTATAG GTTGGTG  ACTGATATTCCAGAAACTACAGGGCCTAATTTCG GTGTGTATATATATAGTTAATT  TTACCATTCATTTAAAATATTACAAATAACTCCCTTGAATTATAAGTTAATTGTAAAA  AGAATATTGAGAATTAGGATTATCTTTTTATTAGTCTAACTCGTGTCTATCAAAAAAT  AAACTGTTAGTATAAATATTATTATGAGTGTATATATAATAAAAATGTGATACATTAT  ACTAACACTTATATATTGATAATATATATTAATTTCTATACTAATAAAATTGTCTATTA  ATATGATTTGTCATACTTTTAATACAGACTTCCAATATTAATACTAAGGATTTATTTGT  TTGTTAGTCAAAGATTATATATACTAACATAGATTAATCCAATGTCGTATCCTATAAG  AAAATAACGTTTCTTTTTATAGTAAATTCAAACAAGTTCTATATATTTATTTATCTGCT  AGATGTTTAATTAGTTTTAAATTACCCAAACCGGGACAAACATTAAATAGTATAAA  AATATTTGCTAAAAAATAACTTAATGGATAATTATTAGTACGTCTAATCATGCTAAAA  GTCAGCGTAGCTACGATATACAAGCATTGTTAAGCCTTTTCTTTTTGTTAAAAATTC  CCTTGCACGTTATACTTTTCTTTTGAGGCTTTTTCATTGTGTTAAAGTCCCTTTATATT  TTTTTCCTAAATAATAACTGAAACTTTCGTTTAGAAGATAGCGGAAATATTACAAGA  AATCATATAAAACAAATGGCAAAATACTTAATTTTATAAACTATATTACAAATTAAA  GCGTGAAAATGTAAATGAACTTTATAATATTAAAAATATTTGTTTTCTTATACCAAAT  TAGTTACTTTTGTGATGTCTTTAAAAAAAAGCTACATAACTCTACTGCTATATATCAT  GTACAAATTAGATTAATATTATAGAGATTTTATATTTGTACAACATTTTTTAACATTAA  TACTGATTATTATTTTTTATTTATTTCACTTTTTATATTTTGTTCGCATATCACGTCACA  AGATCATAAAAAGTGACATGAATATATACAACACCCAAATTATATATTCAATTTGA  ATCTGCTATTGGTAGTTTGTCGACAAGTTTGCGTTACCACAATCCGATTTTGGAACC  TATCACATCATGCGCGAAACACTAGTTTCAAGAAAATCTCTTGTCGTGCAGGGATT  GGAGTATAAAAGTATTGGAATCATTTATGTGGCCAACGTTTCTTGATATATACTTGAA  ATGCAAATAGGGTTTATATGATGATACATAACATCTCATGCAGGAAATTAAAAATG  ATTCCACGAGGCATCAAATTCTAAATTACATGCGATGAGACAAATGACTACATGTC  CCAAAAGATATATAAATATCTAATTCCAATAATAAAAAATATGTCTAACACCAAAAT  ATATGCTAAATTTTGATACACGTTAGTAGTATCTCAGTAGTTTATAAATATATATATAT  ATATATGCTAATTACGAGTTTGCATGATTTGCACGTACGAAG GTAACGAGGTTGTA  AGCTATGAAAGCCCACGACCCACGATGGGGATTCATCGGTTGGTGTTTGTGTTATT  CCGTCAACAGTTTAGACAGAGGGTGTATGCTCCTGGATGGCGACAAAATTTCAATA  CCAGAGAATTTGCTGAACTTTACAACCTTGGATTGCCGGTTGCTGCTGTCTTCTTCA  TGTCAGAGGGAAAGTGGCTCTGGTGGTAGAACATTTTGA   Appendix B   Table A1. Oligonucleotide sequences used in this study.