Developmental responses of bread wheat to changes in ambient temperature following deletion of a locus that includes FLOWERING LOCUS T1

Abstract FLOWERING LOCUS T (FT) is a central integrator of environmental signals that regulates the timing of vegetative to reproductive transition in flowering plants. In model plants, these environmental signals have been shown to include photoperiod, vernalization, and ambient temperature pathways, and in crop species, the integration of the ambient temperature pathway remains less well understood. In hexaploid wheat, at least 5 FT‐like genes have been identified, each with a copy on the A, B, and D genomes. Here, we report the characterization of FT‐B1 through analysis of FT‐B1 null and overexpression genotypes under different ambient temperature conditions. This analysis has identified that the FT‐B1 alleles perform differently under diverse environmental conditions; most notably, the FT‐B1 null produces an increase in spikelet and tiller number when grown at lower temperature conditions. Additionally, absence of FT‐B1 facilitates more rapid germination under both light and dark conditions. These results provide an opportunity to understand the FT‐dependent pathways that underpin key responses of wheat development to changes in ambient temperature. This is particularly important for wheat, for which development and grain productivity are sensitive to changes in temperature.


| INTRODUCTION
Flowering and the subsequent production of gametes enable the introduction and maintenance of genetic diversity to a plant species. The timing of flowering has evolved to best suit individual plant species, with respect to coincidence with pollinators and the local environment.
Utilizing this naturally highly regulated response has been essential in enabling the production of high yielding crops and has been crucial for the adaptation of crops to geographically diverse regions of the world. Therefore, understanding the mechanisms by which flowering time is controlled is important for our sustained agricultural production. Research conducted in numerous plant species has identified photoperiod and temperature as key environmental triggers, whose signals converge on a few genes that control the transition from vegetative to reproductive development (Jung & Müller, 2009). One of the central and final genetic regulators in this process is FLOWERING LOCUS T (FT; Pin & Nilsson, 2012). The FT gene encodes a protein that has sequence homology to phosphatidylethanolamine binding proteins or RAF kinase inhibitor proteins and has been shown to bind proteins and lipids (Nakamura et al., 2014) but has not yet been observed to directly bind DNA. Genes with high sequence similarity to FT from Arabidopsis have been identified across the angiosperm species, and these form relatively large families of FT-like genes, many of which have yet to be characterized (Faure, Higgins, Turner, & Laurie, 2007). The conservation of the FT-like genes is also observed through a central role in flowering time regulation and its action in forming at the least part of florigen, the mobile flowering signal (Putterill & in Arabidopsis and rice that FT is expressed in the leaf phloem vasculature and that FT protein is then transported to the shoot apical meristem where it exerts its effect as the florigen signal, triggering the change from vegetative to reproductive development (Jaeger & Wigge, 2007;Mathieu, Warthmann, Kuttner, & Schmid, 2007;Tamaki, Matsuo, Wong, Yokoi, & Shimamoto, 2007).
In the monocot crop plants, such as wheat and barley, multiple FT-like genes have been identified and shown to have distinct expression patterns and specific interactions with FD-like and 14-3-3 proteins (Li, Lin, & Dubcovsky, 2015;Taoka et al., 2011). This enables the formation of protein complexes containing FT which bind specific DNA sequences and mediate regulation of gene expression in an FTdependent manner (Faure et al., 2007;Kikuchi, Kawahigashi, Ando, Tonooka, & Handa, 2009;Li et al., 2015;Li & Dubcovsky, 2008).
Although this regulation is very similar to that observed in Arabidopsis and rice, other aspects of FT regulation differ between the monocots and dicots. In particular, in barley, the expression of HvFT1-like is not directly regulated by high temperature (Hemming, Walford, Fieg, Dennis, & Trevaskis, 2012).
In hexaploid wheat, at least five FT-like genes, with a copy on each the A, B, and D genomes within wheat, have been identified (Lv et al., 2014). However, the distinct roles of the FTs in flowering regulation remain largely unknown. FT-B1 on Chromosome 7B is the most characterized wheat FT gene, mainly because it was identified to have a central role in the vernalization pathway (FT-B1 is denoted VRN3 in the vernalization pathway; Yan et al., 2006). The notion of FT-B1 as the central FT1 was subsequently supported through homologue specific gene expression analysis which identified that FT-B1 was expressed at a higher level than either FT-A1 or FT-D1 in the Paragon cultivar (Shaw, Turner, & Laurie, 2012). Transgenic overexpression of FT1 bypassed all signals of flowering repression to rapidly promote flowering, with the transformed wheat plants flowering whilst still in the callus stage (Lv et al., 2014). An allele that causes higher expression of FT-B1, first identified in the Hope cultivar, also led to an early flowering phenotype under long days (LD) and bypassing the vernalization requirement (Yan et al., 2006). Notably, this earlier flowering did not lead to any significant decrease in yield (Nitcher, Pearce, Tranquilli, Zhang, & Dubcovsky, 2014). Development of FT1 RNAi lines in Brachypodium further supported the central role of FT1 in flowering regulation in grasses, as FT1 RNAi lines showed dramatic flowering delays. Unfortunately, wheat FT1 RNAi lines were infertile and so prevented confirmation of this result (Lv et al., 2014). Subsequently, an FT-B1 null has been identified and confirmed to be late flowering, this genotype also showed an altered spikelet architecture with the formation of paired spikelets (Boden et al., 2015).
Here, we report the identification of a late flowering hexaploid wheat line which contained a deletion across the FT-B1 region. In addition, we have generated a FT-B1 overexpressor, using the Hope allele, in a highly similar genetic background (Paragon cultivar). This provided the opportunity to further understand the role of FT-B1 in flowering regulation and to identify whether FT-B1 in wheat fulfilled similar roles, specifically relating to temperature responses and development, to those identified in other plant species. Understanding this will provide valuable information for the application of FT-B1 alleles into an agricultural environment and provide important information for the identification of other genes involved in temperature dependent regulation of development and flowering.

| Plant material
FT-B1 null, hereafter referred to as ft-b1, was identified from a cross between the elite UK spring hexaploid wheat, Paragon (Triticum aestivum L.; Vrn-A1a and Vrn-B1c), and a hexaploid landrace wheat from the Watkins collection, denoted W352 (Wingen et al., 2014). An initial cross was conducted between these parents, and the resulting plant was allowed to self-fertilize. From these seeds, a population of 94 individuals was developed through single seed descent to, at least, the F 5 generation (Wingen et al., 2017). All experiments reported in this paper used F 5 or subsequent self-fertilized generations. The ft-b1 line was identified during a detailed phenotyping experiment at the F 5 stage.
The ft-b1 line was single nucleotide polymorphism (SNP)genotyped using the iSelect 90K array  and identified to contain a near equal genetic contribution from both parents.
To generate the ft-b1 near isogenic line (NIL), hereafter referred to as the ft-b1 NIL, a backcross was made using F 6 ft-b1 to the recurrent parent, Paragon, and the ft-b1 allele was selected using Taqman copy number assay (Díaz, Zikhali, Turner, Isaac, & Laurie, 2012)   NILs were developed under LD glasshouse conditions.

| Molecular markers and sequence information
Initial polymerase chain reaction (PCR) analysis suggested that ft-b1 was a gene deletion. To assess how far this deletion extended, PCR primers were designed based on synteny alignment of 7BS with Brachypodium and rice genomes. Oligonucleotide sequences and annealing temperatures are provided in Table S1. Nucleotide sequences of amplicons were confirmed by Sanger sequencing. Using the markers described, the deletion region of Chromosome 7BS was delimited (details provided in Table S2).

| Germination assay
Paragon, ft-b1 NIL, and FT-B1 OX were grown under natural photoperiods in glasshouse conditions (Paragon and FT-B1 OX) or controlled photoperiods in growth cabinets (Paragon and ft-b1), and inflorescences harvested at maturity. Threshed seeds were placed at 37°C for at least 4 weeks to ensure that they were fully after-ripened. The germination assays were constant dark (except for when phenotype scores were taken). The assay was conducted as per Barrero et al. (2015) with up to 18 seeds per genotype per Petri dish. Three separate Petri dishes were phenotyped, with one of these being conducted in an independent experiment.

| Apex characterization (short day [SD] and LD)
Paragon, W352, ft-b1, and FT-B1 OX were grown under either SD (8 hr light/16 hr dark) or LD photoperiods at a constant 20°C and humidity set to 70% in Gallenkamp-controlled environment chambers. Two plants were dissected per genotype for the most time-points every 3 days once the most developmentally advanced genotype had started floral apex development. A second independent experiment was performed to confirm the result (data not included). The sampling frequency was much less for the SD cabinet due to the slower developmental rate. Apex length was measured in ImageJ from the base of the spikelet forming region of the apex to its tip, and error is shown as standard deviation between two samples.

| Quantitative real-time PCR (qRT-PCR)
For the identification of ft-b1 ( Figure S2), leaf tissue samples were taken from Paragon, W352, and PxW352 line74, and RNA was extracted using Tri Reagent (Sigma-Aldrich) and DNase I treated (NEB) before cDNA was synthesized using Superscript II reverse transcriptase (Invitrogen) and random and Oligo dT primers. For the expression analysis of FT-B1 NILs ( Figure 5), the fifth leaf was harvested at Zeitgeber time (ZT) 13 hr for Paragon, ft-b1 NIL, and FT-B1 OX NIL. RNA was extracted using the Spectrum Plant RNA extraction kit (Sigma-Aldrich) before being treated with DNase I, and cDNA synthesized as described above and diluted (1:10) for use in qRT-PCR. For the expression analysis of FT-B1 NILs ( Figure S3), Paragon, ft-b1, and FT-B1 OX were germinated in 5 × 5 cm pots containing peat and sand mix under SD conditions. Leaf samples were collected from six plants at the third leaf stage (SD sample), and the remaining plants were transferred to LD conditions for 7 days when the third leaf was then collected (LD sample). This stage was used because it is when the meristem is at the transition apex stage and is therefore competent to receive the flowering signal. RNA extraction and cDNA synthesis were performed as described above for NILs. For the expression analysis from seeds (Figure 4), RNA was extracted from imbibed seeds.
Seeds were placed on Whatman filter paper with 5 ml distilled water in 9 cm diameter Petri dishes and incubated at 20°C overnight (for ft-b1 and some Paragon) and for a further 24 hr (remaining Paragon and FT-B1 OX) until the seeds had broken dormancy and the definition of the embryo could be identified. The radicle and the plumule of the embryo were dissected and pooled (with three individuals per replicate) and flash frozen in liquid nitrogen. RNA extraction and cDNA synthesis were performed as described above. Oligonucleotide sequences used for qRT-PCR are given in Table S3 with all data being expressed as normalized product according to 2^-ΔCT . The identity of the normalization gene for each experiment is provided in the respective figure legends.  Figure 1c). Using recent advancements in wheat genome assembly, it is evident that the deletion contains more genes than identified through synteny with Brachypodium and rice, and the deletion is estimated to cover a region containing 408 predicted gene models (listed in Table S2), and so a possible contribution of other genes to the flowering-time phenotype cannot be discounted.

| Lines carrying an FT-B1 deletion have altered developmental rate and plant architecture
The flowering delay observed in plants carrying the ft-b1 genotype demonstrates that flowering was affected but provides no indication of the developmental stage that this delay is established. In many plant species, FT forms part of the mobile florigen signal that travels from the leaf to the shoot apical meristem to signal the vegetative to reproductive transition. We hypothesized that this conserved function for   (Figure 3a), but the degree of this delay was similar under both temperature conditions, 23.6 days under higher temperatures (Figure 3a) and 24.6 days under lower temperatures ( Figure 3d). As previous reports had shown that FT has additional roles in regulating plant architecture, we also assessed spikelet and mature tiller number (Figure 3b,c, e,f; Boden et al., 2015;Tsuji et al., 2015). We observed that FT-B1 does influence plant architecture in wheat and that this response is under environmental control (Figure 3b,c,e,f). At the lower temperature, there was no significant difference between Paragon and the FT-B1 OX NIL for spikelet number, on average, Paragon formed 21.2 spikelets, and FT-B1 OX formed 20.6 spikelets. In ft-b1 NIL, the spikelet number significantly increased to 30.2/inflorescence, with on average one more tiller (Figure 3g,h). A similar pattern was observed under the higher

| FT-B1 influences wheat seed germination
FT and FT-family genes have also been identified as key regulators in controlling the dormancy and germination of seeds. In wheat, a homologue of FT, MOTHER OF FT, was identified to increase the dormancy of seeds under lower temperatures (Nakamara et al., 2011).
To investigate the possibility of FT-B1 being involved in regulating seed germination, we used the ft-b1, FT-B1 OX NILs, and Paragon in seed germination assays. The assays were conducted under two temperatures because FT1 is known to be involved in aspects of temperature regulation of seed germination in Arabidopsis (Chen et al., 2014). Seed germination is promoted by warm, dark conditions, and under these conditions, ft-b1 NIL showed a higher percentage of germinated seeds on the first day compared to Paragon and the FT-B1 OX, with all genotypes showing a near 100% germination by the second day (Figure 4b).
A very similar pattern was observed under the dark low-temperature conditions but with a delay of a day before the seeds began to germinate (Figure 4a). Under constant light conditions, ft-b1 NIL also showed faster germination which was more pronounced under the higher temperature conditions (Figure 4c,d). Although we cannot discount that other genes within the 7BS deletion contribute to the germination phenotype of the ft-b1 NIL, a role for FT-B1 in seed germination was supported by detection of FT-B1 transcripts in the radicle and plumule of germinating seeds (Figure 4e). This suggests that FT-B1 has a developmental role beyond flowering-time regulation.

| The role of FT-B1 in temperature regulated gene expression
The physiological analysis of the role of FT in apex transition, germination, and plant architecture (Figures 2, 3, and 4) indicated that FT fulfils different functions at different stages of plant development. The data suggest that the temperature dependency of apex transition and flowering acceleration was occurring at least semi-independently of FT. To try to identify any role of FT-B1 in the temperature dependent signalling, we looked at gene expression at the fifth leaf stage under the two ambient temperature conditions, as this is a stage when stem elongation starts, and which has been shown to be a more temperature dependent process (Kiss et al., 2017). At this stage, a low, very similar level of expression is observed under both ambient temperature conditions in Paragon (Figure 5a), and in the overexpressor, an increased level of FT-B1 is measured which is substantially higher at elevated ambient temperatures. Interestingly, this higher level of FT-B1 at the fifth leaf stage under higher ambient temperatures does not lead to an earlier flowering-time, which may indicate that the levels of FT-B1 at this stage are not important in determining to final flowering time.
The gene expression of FT-B1 under standard regulation (Paragon) suggested that FT-B1 was not involved in the ambient temperature response of wheat plants. If this was the case, we could anticipate that the expression of genes known to be involved in the temperature response in cereals would be similar with or without FT-B1. To test this, we measured the gene expression of RNase-S-like (RSH1) which has been shown to respond to different ambient temperatures (Hemming et al., 2012). The expression of RSH1 was extremely similar in Paragon and ft-b1 NIL, identifying that both genotypes could still respond to ambient temperature signals (Figure 5b).
To determine if altered levels of FT-B1 activity affects the expression of FT1 homologues or FT paralogs, we measured transcript levels of FT-A1 and FT-D1 and FT2, FT3, FT4, FT5, and FT6 (sequences provided in Figure S4). Paragon, ft-b1 NIL, and FT-B1 OX plants were grown under SD conditions until the third leaf stage, before being transferred to LD conditions for 7 days when leaves were harvested for RNA extraction. Using qRT-PCR, we were not able to detect transcripts for FT-A1, FT-D1, FT5, and FT6 in all three genotypes.
Transcripts for FT2, FT3, and FT4 were detected; however, there was no significant difference in levels identified between the three genotypes ( Figure S3b). Based on these results, we conclude that of the FT-like genes in wheat, FT-B1 uniquely contributes to the flowering-time phenotype of the ft-b1 and FT-B1 OX NILs at this stage.

| DISCUSSION
The FT and FT-like genes are known to be central in the regulation of vegetative to reproductive transition in plants; however, it is becoming apparent that these genes also have central roles in the regulation of a number of other processes, many of which are key to agricultural productivity (Pin & Nilsson, 2012). Here, we have examined regulatory roles for FT-B1 during processes including flowering, germination, and tiller development through genetic characterization of the FT-B1 locus in a spring, photoperiod sensitive wheat. By growing genotypes that differ in FT-B1 activity under different ambient growth temperatures, we have also increased our understanding of the contribution for FT to thermally responsive developmental processes in wheat.
Our results show that the absence of FT-B1 significantly delays flowering under both high and low ambient temperatures (Figure 3), which confirms that FT has a key role during the vegetative to reproductive transition in wheat. Absence of FT-B1 was also found to delay   (Lv et al., 2014). This suggests that expression from the Hope allele is not constitutive, which is supported by gene expression analysis shown here ( Figure 5) and by previous reports showing that transcript levels of the FT-B1 OX allele and its effect on flowering-time are dependent on the developmental age of the plant and the genetic background, respectively . Similarly, our characterization of flowering-time in the ft-b1 NILs shows that the FT-B1 locus is not singularly responsible for integrating the environmental and developmental information required to trigger transition from vegetative to floral development in hexaploid wheat and that some redundancy either with other FT genes or an FT-independent pathway exists to ensure floral progression without FT-B1, as previously reported (Boden et al., 2015;Hemming et al., 2012;Lv et al., 2014).
The characterization of the FT-B1 NILs also provides insights into the contribution of FT-B1 to plant architecture and development beyond regulation of flowering-time. We observed that tiller numbers were lower in the FT-B1 OX NIL, that this effect was significant under high but not low growth temperatures and associated with accelerated flowering and reduced spikelet numbers. Given that tillers develop from vegetative nodes, this result may indicate that a higher level of FT-B1 accelerates development and thus ultimately results in fewer tillers as the plants reach maturity faster. This result is consistent with the observation that the ft-b1 generally had a higher tiller number, indicating that without FT-B1 more tillers could grow out, possibly at a slower rate.
Interestingly, however, the phenotypic result is distinct from reports in rice that showed the FT homologue, Hd3a, promotes tiller development such that in an Hd3a RNAi line tiller number is decreased even though the plant flowers later (Tsuji et al., 2015). This result may reflect a different developmental strategy or greater developmental dependency on Results presented here provide insights that could be used to further understand the flowering response of wheat to increased ambient temperatures and to identify genes that underpin these responses.
The data presented show that FT-B1's role in regulating flowering-time is largely exerted through acting early in inflorescence development and therefore highlights the importance of considering early developmental stages, in addition to the later stages of inflorescence emergence, anthesis, and grain development, when considering the effects of increased growth temperatures on reproductive development in wheat. Our observation that the flowering time delay was greatly decreased by elevated ambient temperatures when the ft-b1 allele was in a background with a higher proportion of landrace genetic material ( Figure S1) suggests that genetic resources such as the Watkin's population (Wingen et al., 2014) could be useful for identifying genes that are crucial for ambient temperature responses in wheat.
This approach could be especially useful for identifying flowering-time genes whose effect has been lost from modern wheat cultivars or is masked by key regulators, such as VERNALIZATION1 and PHOTOPE-RIOD1. Furthermore, the ft-b1 allele could potentially be used as an alternative vernalization allele in areas which do not always meet vernalization requirement due to warming climate. In a spring background, the ft-b1 allele shows a developmental delay which could allow winter drilling but proceeds fast enough, with the additional benefit of higher spikelet number, to still flower within a standard season. Field trials would be required to ascertain the most suited environmental conditions to utilize this allele. The research presented here identifies new applications for a wellknown gene in flowering regulation and links these applications to the role key environmental parameter, of ambient temperature. It uncovers differences in FT-B1 function to that of FT's in other plant species and highlights the need to study the genes destined for agricultural application in as close to agricultural genetic material as possible, exampled through the differences we observed in the ft-b1 response between landrace and elite wheat cultivars. Identifying and understanding the role of FT-dependent and independent pathways will be critical for our ability to regulate growth under warmer climates, something which will be required to develop a robust wheat yield.