CRISPR/Cas9‐mediated mutations of FANTASTIC FOUR gene family for creating early flowering mutants in tomato

Summary Flowering time is of great agricultural importance and the timing and extent of flowering usually determines yield and availability of flowers, fruits and seeds. Identification of genes determining flowering has important practical applications for tomato breeding. Here we demonstrate the roles of the FANTASTIC FOUR (FAF) gene family in regulating tomato flowering time. In this plant‐specific gene family, SlFAF1/2a shows a constitutive expression pattern during the transition of the shoot apical meristem (SAM) from vegetative to reproductive growth and significantly influences flowering time. Overexpressing SlFAF1/2a causes earlier flowering compared with the transformations of other genes in the FAF family. SlFAF1/2c also positively regulates tomato flowering, although to a lesser extent. The other members of the SlFAF gene family, SlFAF1/2b, SlFAF3/4a and SlFAF3/4b, are negative regulators of tomato flowering and faf1/2b, faf3/4a and faf3/4b single mutants all display early flowering. We generated a series of early flowering mutants using the CRISPR/Cas9 editing system, and the faf1/2b faf3/4a faf3/4b triple mutant flowering earliest compared with other mutants. More importantly, these mutants show no adverse effect on yield. Our results have uncovered the role of the FAF gene family in regulating tomato flowering time and generated early flowering germplasms for molecular breeding.


Introduction
In plants, a pool of pluripotent cells in the growing tips forms the shoot apical meristems (SAM) and the activities of cells in the SAMs determine the shoot developmental process.During the vegetative phase, SAMs generate leaves and stems (Zhang et al., 2022).After floral initiation, SAMs develop into an inflorescence meristem (IM) and give rise to flowers (Benlloch et al., 2007;Kwiatkowska, 2008).The development of vegetative and reproductive meristems along the main axes varies in different flowering plants and the growth habit of higher plants can be divided into two categories, monopodial and sympodial (Weberling, 1992).The model species Arabidopsis thaliana presents a typical monopodial growth habit.The SAM undergoes vegetative growth until sensing external cues and is then transformed into an inflorescence meristem (IM), which grows continuously and produces floral meristems (FM) laterally (Andr es and Coupland, 2012;Benlloch et al., 2007).In this way, the main axis of Arabidopsis grows indeterminately.By contrast, the main shoot of sympodial plants terminates in flowers.Tomato plants have classical determinate inflorescences where the primary shoot meristem (PSM) switches into a FM after producing 7-12 leaves and terminates in an inflorescence.A new shoot (called sympodial shoot meristem, SYM) arises below the inflorescence on the PSM and generates three leaves before terminating in an inflorescence.Again, a new SYM develops below the inflorescence from the previous SYM and generates three leaves with a terminal inflorescence.This process occurs repeatedly so that the tomato plants display an indefinite growth habit (Park et al., 2011;Pnueli et al., 1998).Therefore, flowering process relies on the dynamic development of the SAM, and the timely switching from vegetative to reproductive growth is regulated by many internal and external factors.
The importance and influence of photoperiod on flowering were noticed and examined many years ago (Garner andAllard, 1920, 1923).Later, grafting experiments showed that, in photoperiodic plants, leaves perceive day length changes and emit a systemic hormone signal, called florigen, which moves in the phloem from leaves to the vegetative shoot apical meristem that initiates flower development (Chailakhyan, 1936).After years of dedicated research, it was demonstrated that florigen is a protein hormone encoded by the FLOWERING LOCUS T (FT) gene in Arabidopsis (An et al., 2004).The FT protein functions as a transcriptional cofactor that is produced in leaves after exposure to long-days (LDs) conditions and moves to the SAM to form a complex with the basic leucine zipper domain transcription factor FD, and this complex transcriptionally activates the expression of APETALA1 (AP1) and SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1) to initiate the flowering process (Andr es and Coupland, 2012;Kobayashi and Weigel, 2007;Turck et al., 2008).The tomato FT ortholog is SINGLE FLOWER TRUSS (SFT), and sft mutants display several phenotypes including lateflowering under LDs and short-days (SDs) conditions, defects in the flower meristem identity and abnormal developments of flower organs and sympodial shoots (Molinero-Rosales et al., 2004).The sft phenotypes can be complemented by grafttransmissible SFT signals, which suggests that the florigen pathway is conserved in tomato, and SFT is an important regulator of flowering time, like FT in Arabidopsis (Lifschitz et al., 2006;Lifschitz and Eshed, 2006).However, the regulatory network of SFT shows differences compared with FT.In Arabidopsis, FD is specifically expressed in the SAM, but the tomato FD homologue, SPGB (a bZIP G-box protein), is expressed in leaves.Therefore, unlike FT and FD, SFT may not bind with FD specifically in the SAM, which indicates that the SFT-integrator genes and their targets may be different in tomato and Arabidopsis (Lifschitz et al., 2006).Thus, the model of tomato flowering regulation favours a florigen-like model, in which the downstream pathway is relatively incomplete and requires further investigation.
The FANTASTIC FOUR gene family encode a class of plantspecific protein with unknown functions, which includes four members in Arabidopsis, that is, FAF1, FAF2, FAF3 and FAF4.Initially, the FAF1 and FAF2 genes were observed to respond strongly and rapidly to photoperiod changes (Schmid et al., 2003).Further investigations of the four members of the FAF family revealed the potential function of FAF proteins.It was shown that FAF2 and FAF4 are expressed in the central zone of shoot meristem that overlaps with the expression region of WUSCHEL (WUS), and the FAF2 and FAF4 proteins can arrest the expression of WUS.The expression of FAF2 and FAF4 is also under negative control by CLAVATA3 (CLV3).These results suggest that in Arabidopsis the FAF genes are involved in a classical CLV3-WUS feedback loop that regulates the shoot meristem development (Wahl et al., 2010).The four FAF genes are also, however, expressed in the developing and mature vasculature (Wahl et al., 2010).Further, through analysing the transcriptome of Arabidopsis floral transition meristem, it was found that FAF2 expression is significantly increased in the meristem after changing the growth conditions from SDs to LDs.The activation of FAF2 in the floral transition meristems is dependent on the presence of FT and TSF (TWIN SISTER OF FT) (Torti et al., 2012).
FAF genes are also involved in other processes.The hormone cytokinin (CK) activates the growth of buds whereas auxin inhibits growth through apical dominance.Supplying buds with CK overcomes the auxin-mediated bud inhibition.During this process, the expression levels of FAF1, FAF2 and FAF3 are significantly decreased under apical auxin and increased under CK treatment, suggesting that the FAF1-3 genes are involved in bud development (Bhargava et al., 2013;Muller et al., 2015).FAF1 also responds to GR-REV and KAN1-GR induction, which may function as a meristem regulator involved in an Ad/Abaxial regulatory network (Reinhart et al., 2013).Therefore, the four FAF genes show functional divergence in Arabidopsis.Furthermore, the numbers of FAF genes differ between monocotyledonous and dicotyledonous species (Wahl et al., 2010).In tomato, the CELL SIZE REGULATOR (CSR) underlies the fw11.3locus, and CSR is mainly expressed in fruit tissues and vascular bundles during the fruit maturation process, which ultimately lead to increases in the mesocarp cell size and fruit weight (Mauxion et al., 2021;Mu et al., 2017).Phylogenetic analysis showed that the SlCSR gene is orthologous to AtFAF-like gene, and protein sequence analysis showed that there are 13 proteins containing FAF domain in the tomato genome, three more than in Arabidopsis.It has been suggested that the expansion of the number of FAF gene family might only occur in the Solanaceace family (Mu et al., 2017).Therefore, we explored the function of the tomato FAF orthologs.
During tomato flowering, the transition meristem will eventually develop into a flower meristem, and previous studies have thoroughly characterized this process using single-meristem transcriptomes technology (Meir et al., 2021).We noticed that the expression level of a FAF family gene was significantly downregulated in the sft mutant during floral transition stages (Meir et al., 2021).According to the phylogenic relationship in Arabidopsis, the FAF gene family are divided into two branches in tomato, the FAF1/FAF2 branch including SlFAF1/2a, SlFAF1/2b and FAF1/2c and the FAF3/FAF4 branch including SlFAF3/4a and FAF3/4b.SlFAF1/2a is expressed in all SAM development stages with high expression levels in floral transition meristems.Furthermore, we showed that overexpressing SlFAF1/2a will lead to the earliest flowering.We also characterized the function of other genes from FAF family, SlFAF1/2c is another positive regulator and SlFAF1/2b, SlFAF3/4a and SlFAF3/4b are three negative regulators.Based on that, we generated a series of early flowering mutants between SlFAF1/2b, SlFAF3/4a and FAF3/4b using the CRISPR/Ca9 editing system.

SlFAF1/2a is mainly expressed during floral transition phases
In Arabidopsis, the FAF gene family includes four members: FAF1, FAF2, FAF3 and FAF4, and FAF1 is paralogous with FAF2, FAF3 is paralogous with FAF4 (Wahl et al., 2010).Phylogenetic analysis showed that there are five FAF genes in tomato and Solyc06g084280, Solyc06g008990 and Solyc09g065140 are homologous with FAF1/FAF2 paralog, therefore, they were named as SlFAF1/2a, SlFAF1/2b and SlFAF1/2c, respectively.Solyc01g079740 and Solyc06g054310 are homologous with FAF3/FAF4 paralog and were named SlFAF3/4a and SlFAF3/4b, respectively (Figure S1).The nomenclature of the SlFAFs genes follows a previous study (Mu et al., 2017).We compared the amino acid sequences of these five SlFAF proteins and sequence alignment results showed that in addition to the FAF conserved domain, their amino acid sequences display with many other similarities (Figure S2).These structural similarities suggested that the SlFAF gene family members may exhibit similar regulatory properties in tomato flowering.But there are also some differences in their amino acid sequences, so it is also possible of functional divergences between these genes.
Previous study sampled hundreds of individual tomato SAMs to perform single-meristem transcriptome (SMT) profiling and establish a detailed dynamic transcriptome map of the floral transition process (Meir et al., 2021).To capture the temporal changes in genes expression levels at different developmental stages, Meir et al. ordered  Engineering early flowering tomato by editing FAF genes 775 and transition III), flower initiation phase and floral meristem formed phase.In the released data, we found that among these five SlFAF genes, only SlFAF1/2a is highly expressed in the SAMs at all developmental stages in a specific pattern, with the other four SlFAF genes showing much lower expression levels without clear expression pattern (Figure 1a, Figure S3).Thus, first we focus on the gene SlFAF1/2a.
Specifically, the expression level of SlFAF1/2a is lowest in the early vegetative phases (i.e.veg1-2) and gradually increases in the later vegetative phase (i.e.veg3).Its highest expression level was in SAMs that were dooming (showing enlargement characteristic of floral meristem development) in the transition I phase (i.e.tr1) and subsequently decreased in the transition II phase (i.e.tr2).Following SAM bifurcation, the expression level of SlFAF1/2a decreased slightly in the late transition meristems (i.e.tr3 phase) and was maintained at a relatively stable level during flower initiation and formation (i.e.FI and flo phases) (Figure 1a).Intriguingly, we also noticed that in the sft mutant, the expression level of SlFAF1/2a was significantly lower in all sft meristems from the five developmental phases (Figure 1b, c).The most notable reduction (5.91-fold) was in the early transition I phase in the sft mutant compared with the wild type.In the vegetative, transition II, transition III and floral initiation phases, SlFAF1/2a was also downregulated approximately 2.43-fold, 2.32-fold, 1.32-fold and 1.15-fold, respectively (Figure 1a).The high expression level of SlFAF1/2a in flower transition meristems indicates that this gene may be involved in the regulation of tomato flowering, and SlFAF1/ 2a may function in the downstream regulatory network of SFT.
All WT (AC) plants produced more than 10 leaves before the first inflorescence, with a majority (85%) of plants producing more than 12 leaves before the first inflorescence (Figure 2c-e).Remarkably, the number of leaves formed before the first inflorescence was significantly reduced in the SlFAF1/2aoverexpressing lines (Figure 2d).For all overexpressing lines, the maximum number of leaves before the first inflorescence was 10 (Figure 2d).The majority, 95.65%, 88.46% and 94% in 35S: FAF1/2a-1, 35S: FAF1/2a-5 and 35S: FAF1/2a-18 lines, respectively, produced fewer than nine leaves before the first inflorescence (Figure 2e).In addition to the effects of SlFAF1/2a on the number of leaves produced before the first inflorescence, the days to flowering were also significantly reduced in the SlFAF1/2a-overexpressing lines.The average number of days before flowering was 61 in the WT (AC) plants, which was reduced to 37.19, 38.75 and 39.88 in the three SlFAF1/2aoverexpressing lines, respectively (Figure 2f).Therefore, these results show that the increased expression level of SlFAF1/2a resulted in marked earlier flowering in tomato.When we investigated the CR-faf1/2a mutants, there are no significant differences in the number of leaves produced before the first inflorescence and the number of days before flowering compared with WT (AC) plants (Figure 2c-f).Thus, loss of SlFAF1/2a function did not alter the flowering time significantly.Taken together, these results demonstrated that SlFAF1/2a positively regulate tomato flowering time.

Different members of the SlFAF gene family have contrasting effects on tomato flowering
To explore the function of other SlFAF genes, we overexpressed SlFAF1/2b, SlFAF1/2c, SlFAF3/4a and SlFAF3/4b separately in the AC background and obtained three independent overexpressing lines for each gene (Figure S4A-D).We also mutagenized SlFAF1/ 2b, SlFAF1/2c, SlFAF3/4a and SlFAF3/4b in the same background using CRISPR/Csa9 technology (Figure S4E-H).We sequenced the resulting faf1/2b, faf1/2c, faf3/4a and faf3/4b alleles in independent faf1/2b, faf1/2c, faf3/4a and faf3/4b mutants, respectively, and selected two homozygous mutants for each gene (Figure S4E-H).In the FAF1/FAF2 evolutionary branch, the regulatory effect of SlFAF1/2c is similar to SlFAF1/2a.Overexpressing SlFAF1/2c leads to early flowering (Figure 3a).The number of leaves before the first inflorescence of SlFAF1/2c-overexpressing lines was between 7 and 10, which is significantly reduced compared with WT (AC) (Figure 3d).The number of days before flowering was also significantly reduced to 37 days in the SlFAF1/2c-overexpressing lines compared with 61 days in WT (AC) (Figure 3e).And no differences were observed in the flowering time comparing CR-faf1/2c mutants with WT (AC) (Figure 3d,e).In contrast, overexpressing SlFAF1/2b did not alter the flowering time (Figure 3b,c).In the CR-faf1/2b mutants, however, the number of leaves before first inflorescence developed was lower to 10 compared to 13 leaves in WT (AC), which is significantly reduced (Figure 3a,b).The number of days before flowering also significantly reduced to 51 days in the faf1/2b mutants (Figure 3c), which represents a significant acceleration of the flowering process compared with WT (AC).Therefore, the SlFAF1/2a and SlFAF1/2c display similar regulations on tomato flowering, and their action is different from SlFAF1/2b.
In the FAF3/FAF4 branch, the transgenic lines of SlFAF3/4a and SlFAF3/4b exhibited similar phenotypic changes as the SlFAF1/2b transgenic lines.A majority of plants from SlFAF3/4a-overexpressing lines produced over 12 leaves before the first inflorescence, Engineering early flowering tomato by editing FAF genes 777 which is similar to WT (AC) (Figure 3f), and the number of days before flowering are also similar to WT (AC) (Figure 3g).But in the CR-faf3/4a mutants, the number of leaves before the first inflorescence developed was lower to 9 compared to 13 leaves in WT (AC), which is significantly reduced (Figure 3a,f).The number of days before flowering also significantly reduced to 48 days in the faf3/4a mutants (Figure 3g).Similarly, the flowering time of SlFAF3/4b-overexpression lines is not altered (Figure 3h,i).In the CR-faf3/4b mutants, the number of leaves before first inflorescence developed was lower to 9, which is significantly reduced compared with WT (AC) (Figure 3a,h).The number of days before flowering also significantly reduced to 47 days in the faf3/4a mutants (Figure 3i).Thus, the loss of the function of SlFAF3/4a and SlFAF3/4b leads to early flowering.Taken together, these transformation experiments demonstrated that in the SlFAF gene family, the SlFAF1/2a and SlFAF1/2b are the positive regulators and SlFAF1/2b, SlFAF3/4a, SlFAF3/4b are the negative regulators for tomato flowering time.
Additive effects among the negative regulators in SlFAF gene family on tomato flowering Knocking out SlFAF1/2b, SlFAF3/4a and SlFAF3/4b caused early flowering (Figure 3b,c,f-i) and the average number of leaves before the first inflorescences were 10, 9 and 9 in CR-faf1/2b, CR-faf3/4a and CR-faf3/4b lines, respectively, which was 3-4 leaves earlier than in WT (AC) that produces an average of 13 leaves before the first inflorescence develops.To explore whether knocking out these three genes simultaneously would result in further early flowering, we took advantage of the CRISPR/Cas9 multiplex editing capability (Xie et al., 2015) to generate the double mutants faf1/2b faf3/4b, faf3/4a faf3/4b (Figure S5A,B) and the triple mutants faf1/2b faf3/4a faf3/4b (Figure S5C), respectively.These mutants were generated in the AC background.For each mutant combination, we selected two homozygous lines separately (Figure S5).
The number of leaves produced before the first inflorescence and the number of days before flowering in WT (AC) and different mutant lines were quantified.In WT (AC), approximately 83.82% of plants formed more than 12 leaves before flowering and the average number is 13 leaves, the average number of days before flowering is 62 days (Figure 4b-d).As described above, the single mutants of faf1/2b, faf3/4a and faf3/4b cause early flowering compared with WT (AC) (Figure 3).Among these three single mutants, in the faf1/2b mutant, the reduction in number of leaves before flowering is to a lesser degree compared with faf3/ 4a and faf3/4b single mutant, the number of days before flowering is not significantly different from faf3/4a and faf3/4b single mutant (Figure 4c,d).When we combined the faf1/2b with faf3/4b, and the faf1/2b faf3/4b double mutants exhibited similar flowering time distributions compared with faf3/4b single mutants (Figure 4b-d).Thus, the faf1/2b did not show an additive effect on faf3/4b.The combination between faf3/4a and faf3/4b mutations also shows no additional effect on flowering time compared with the single mutant (Figure 4b-d).Further, we generated the faf1/2b faf3/4a faf3/4b triple mutants.In the triple mutants, the average number of leaves before first inflorescence was 8 and the average number of days before flowering was 39 days, which is significantly reduced compared with WT (AC), the single and double mutants (Figure 4c,d).Thus, the triple mutants exhibit the most pronounced early flowering phenotype.Taken together, we showed that in the FAF gene family, the combined mutations between the single mutants of the negative regulators will cause the earliest flowering.

No adverse effects on plant development and yield in early flowering slfafs mutants
To explore whether the SlFAF gene family show other effects on plants following the alteration of flowering time, we investigated the plant growth, flower and inflorescence development, fruit set rates, fruit development and plant yield of the SlFAFs transgenic lines and WT (AC).
At 110 days, plants from early flowering SlFAFs transgenic lines grow similarly with WT (AC) plants without defections (Figure S6).For the flower development, there are no significant differences in their flower size and structure comparing different transgenic lines with the WT (AC) (Figure S7A), and the flower number from each inflorescence of the transgenic lines is similar to WT (AC), which is mainly distributed between 8 and 10 (Figure S7B-G).The inflorescences from the transgenic lines still develop one branch, which are same as the WT (AC) (Figure S7H).Therefore, the flower and inflorescence development remain unaffected in the SlFAF transgenic lines.In addition, we investigated the fruit set rates in the transgenic lines, which are similar to the WT (AC) (Figure S8).The fruits from SlFAF1/2a and SlFAF1/2coverexpression lines are smaller than the WT (AC) (Figure 5a,c).However, the total yield from SlFAF1/2a-overexpression plants is not changed, while they are reduced in the SlFAF1/2coverexpression plants (Figure 5g,i).For other transgenic lines, no differences were observed in the fruit size or plant yield comparing with WT (AC) (Figure 5).Besides that, the soluble solids content of fruits from SlFAF1/2c-overexpression lines also reduced relative to WT (AC) (Figure 6c), but not in other transgenic lines (Figure 6a,b,  d-f).Regarding fruit shape, fruits from SlFAF3/4a and SlFAF3/4boverexpression lines are longer than the WT (AC), but not in other transgenic lines (Figure S9).Taken together, the early flowering SlFAF1/2a-overexpression lines and the CR-faf1/2b, faf3/4a, faf3/ 4b, faf1/2b faf3/4b, faf3/4a faf3/4b, faf1/2b faf3/4a faf3/4b mutants have no side effect on plant yield.

Discussion
In this study, we characterized the function of the FAF gene family in tomato and showed that they have different regulatory effects on tomato flowering time.Increasing the expression level of SlFAF1/2a had the most significant effect on tomato flowering.Most plants from the SlFAF1/2a-overexpressing lines produced less than 8 leaves before the first inflorescence, and it took about 30 days before flowering, which is drastically shortened compared with WT (AC), which took on average 61 days and 13 leaves before flowering (Figure 2).The other positive regulator is SlFAF1/2c and its effect on early flowering was also significant, albeit to a lesser extent than SlFAF1/2a (Figure 3d,e).Mutations in the remaining three genes, SlFAF1/2b, SlFAF3/4a and SlFAF3/4b, caused early flowering (Figure 3a).SlFAF1/2b forms a branch with SlFAF1/2a and SlFAF1/2c, but display the opposite regulation on flowering.This may be caused by the different expression pattern between SlFAF1/2b and SlFAF1/2a or SlFAF1/2c (Figure S3).SlFAF3/4a and SlFAF3/4b formed the other branch in SlFAF gene family (Figure S1).Both of them are negative regulators of tomato flowering (Figure 3f-i).
In tomato, most natural mutants related to flowering time show delayed flowering, such as falsiflora (fa), compound inflorescence (s), jointless (j), single flower truss (sft) and uniflora (uf) (Dielen et al., 2004;Lifschitz et al., 2006;Lippman et al., 2008;Molinero-Rosales et al., 1999;Szymkowiak and Irish, 2006).In contrast, the terminating flower (tmf) is the only identified early flowering mutant in tomato (MacAlister et al., 2012).Early flowering usually means early yield, which is important for tomato production.To acquire early flowering germplasms, previous studies used the CRISPR/Cas9 system to engineer mutations in the flowering repressor SELF-PRUNING 5G (SP5G) (Soyk et al., 2017).These sp5g mutants displayed rapid flowering and the quick flower production subsequently translated to early yield (Soyk et al., 2017).We showed that individually mutagenizing SlFAF1/2b, SlFAF3/4a and SlFAF3/4b also leads to early flowering.To explore the genetic effects on tomato flowering between these three genes, we generated the faf1/2b faf3/4b, faf3/4a faf3/4b and faf1/2a faf3/4a faf3/4b Engineering early flowering tomato by editing FAF genes 779 multiple gene mutants.The double mutants did not show an addictive effect compared with the single mutants (Figure 4c,d).Simultaneous mutations in SlFAF1/2b, SlFAF3/4a and SlFAF3/4b significantly accelerated flowering time compared with the single and double mutants, which suggests that there are possible additive effects underlying the action of these three genes (Figure 4b-d).Taken together, we created a series of tomato mutants with different degrees of early flowering, which assist breeders to select germplasms of different flowering times.
In the early flowering faf1/2b, faf3/4a, faf3/4b single mutants and the faf1/2b faf3/4b, faf3/4a faf3/4b, faf1/2a faf3/4a faf3/4b multiple gene mutants, we found that their plant growths were unaffected and the fruits mature earlier comparing with WT (AC) owing to early flowering in the mutants (Figure S6).In addition, the flowers, inflorescences and fruit development from those mutants were also unaffected (Figures S7, S9F and Figure 6f).More importantly, since the fruit set rates and fruit weight still were similar to WT (AC), which the plant yield was not affected In summary, we discovered a gene belonging to FAF gene family (i.e.SlFAF1/2a) that is highly expressed during floral transition stages, and its expression level is influenced by SFT.Overexpressing SlFAF1/2a caused early flowering, knocking out this gene had no influence on flowering time, compared to WT (AC).Characterization of the remaining genes from this family demonstrated that SlFAF1/2a and SlFAF1/2c promote early flowering, whereas SlFAF1/2b, SlFAF3/4a and SlFAF3/4b are flowering repressors.Knocking out either SlFAF1/2b, SlFAF3/4a or SlFAF3/4b caused early flowering, and simultaneously mutagenizing these three genes resulted in the earliest flowering time without adverse effects on yield compared with the other single and double mutants.Engineering early flowering tomato by editing FAF genes 781

Plant materials and growth conditions
The cultivated tomato, Solanum lycopersicum cv.Ailsa Craig (AC), was selected for Agrobacterium tumefaciens (strain C58)mediated transformation experiments (Ouyang et al., 2005).The transgenic plants from T 2 generations were investigated.All plants were grown under the same environmental conditions: photoperiod consisting of 16 h of light and 8 h of darkness, temperature at 25 AE 2 °C, and a relative humidity of 70%.

Plants phenotyping
Individual plants from the transgenic lines and wild-type (AC) were used to count the number of leaves formed before the first inflorescence developed.Days to flowering were counted from the first day of sowing seeds.At least seven plants were investigated for each line.For flower and inflorescence development, they were investigated using the second or third inflorescence from each transgenic line and WT (AC).At least six inflorescences were investigated for each line.Regarding fruit development, at least six fruits were investigated for fruit size, fruit shape and soluble solids content (SSC).SSC was quantified as Brix from red-ripe fruits using a digital refractometer (PAL-BX|ACID3).Fruit set rates were counted from the second or third inflorescence of each line.At least six inflorescences were investigated.Plant yields were counted using at least five plants for each line.For each phenotype, the exact investigated number was presented in corresponding Figures.

Phylogenetic analysis
Phylogenetic relationships between FAFs proteins in S. lycopersicum and Arabidopsis thaliana were inferred using the neighbourjoining method.The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (2000 replicates) are shown next to the branches.The full-length amino acid sequences of SlFAFs paralogous and orthologous genes were downloaded from EnsemblPlants and aligned using Clustal W2.Evolutionary analyses were conducted in MEGA7 (Kumar et al., 2016).

Quantitative real-time PCR analysis
To detect the expression levels of FAF genes, total RNA from leaves of FAFs-overexpression lines was extracted using TRIzol reagent (Aidlab, China).AHiScript II 1st Strand cDNA Synthesis Kit (+gDNA wiper) (Vazyme, China) was used to synthesize the firststrand cDNA according to the manufacturer's protocol.The relative transcript levels of FAFs genes were determined by quantitative real-time PCR (qRT-PCR) on a QuantStudio TM 6 Flex System (ABI, USA).Solyc11g008430 (Q-actin) was used as an internal control.Primer sequences are listed in Table S1.

Statistical analysis
GraphPad Prism version 8.3.0 (GraphPad Software, Inc., La Jolla, CA: http://www.graphpad.com/)was used to perform the statistical analysis.Statistically significant differences between groups were determined using One-way ANOVA with Tukey's post hoc test.

Figure 1
Figure 1 Transcriptional dynamics of SlFAF1/2a.(a), Expression levels of SlFAF1/2a in different SAM developmental phases from WT and the sft mutant.Vegetative phase was divided into three subphases (i.e.veg1-3).Three transition phases: tr1, tr2 and tr3.Flower initiation phase: FI.Floral meristem formed phase: flo.Expression levels on the y axis were computed as the unique molecular identifiers (UMIs) per 100 000 UMIs. (b, c), Expression levels of SlFAF1/2a in individual meristems from WT (b) and the sft mutant (c).A total of 379 meristems were measured in WT and 162 meristems were measured in the sft mutant.The background colours represent different developmental phases, which are consistent with the colours in A. The transcriptome data and graphs were automatically produced on the website: https://tanaylab.weizmann.ac.il/SMT/.

Figure 2
Figure 2 Functional characterizations of SlFAF1/2a.(a) Relative transcript levels of SlFAF1/2a in WT (AC) and the SlFAF1/2a-overexpressing lines.Three biological replicates were analysed for each line.Error bars indicate SE.(b) Generation of faf1/2a mutants by CRISPR/Cas9.The sgRNAs sequences of SlFAF1/2a in the WT (AC) and in the faf1/2a-2, faf1/2a-11 mutants are shown.(c) Plants from WT (AC), SlFAF1/2a-overexpressing lines and CR-faf1/2a mutants.The first inflorescences are indicated with red arrows and the number of leaves before the first inflorescences are indicated on the Figures.(d) Distribution of number of leaves before the first inflorescences in WT (AC), SlFAF1/2a-overexpressing lines and CR-faf1/2a mutants.(e) Percentages of number of leaves before the first inflorescences of WT (AC), SlFAF1/2a-overexpressing lines and CR-faf1/2a mutants.(f) Distribution of days to flowering in WT (AC), SlFAF1/2a-overexpressing lines and CR-faf1/2a mutants.n, number of plants investigated.All distributions are shown as box plots of the range of percentiles from the total data, as determined using Tukey's method.Central line, median; whiskers, interquartile range; outer dots, outliers.Different letters indicate significant differences (P < 0.0001, one-way ANOVA with Tukey's post hoc test).

Figure 4
Figure 4 Additive effects among the negative regulators in the SlFAF gene family.(a) Plants from WT (AC), CR-faf1/2b, faf3/4a and faf3/4b single mutants, faf1/2b faf3/4b, faf3/4a faf3/4b double mutants and faf1/2b faf3/4a faf3/4b triple mutants.The first inflorescences are indicated with red arrows and the number of leaves before the first inflorescences are indicated on the Figures.(b) Percentages of number of leaves before the first inflorescences in WT (AC) and different single, double and triple mutants.(c) Distribution of number of leaves before the first inflorescences in WT (AC) and different single, double and triple mutants.(d) Distribution of days to flowering in WT (AC) and different single, double and triple mutants.n, number of plants investigated.All distributions are shown as box plots of the range of percentiles from the total data, as determined using Tukey's method.Different letters indicate significant differences (P < 0.05, one-way ANOVA with Tukey's post hoc test).
these individual SAMs into six phases, including vegetative phase (further divided into three subphases, i.e. veg1-3), three transition phases (i.e.transition I, transition II ª 2023 The Authors.Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd., 22, 774-784

Table S1
List of primers used in this study.ª 2023 The Authors.Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd., 22, 774-784