Florigen Activation Complex Dynamics and SVP-Mediated Repression Orchestrate Temperature-Regulated Flowering in Saffron

18 Saffron, a high-value spice cultivated worldwide for its therapeutic and culinary uses, 19 is a sterile triploid species, rendering conventional breeding approaches ineffective. 20 This limitation underscores the need for molecular and biotechnological strategies for 21 its genetic improvement. Flowering, a key determinant of saffron yield, is strongly 22 influenced by temperature; however, the genetic regulatory networks underlying this 23 process remain poorly understood. Our study identifies key regulators of saffron's 24 flowering, focusing on the Florigen Activation Complex (FAC) components: 25 FLOWERING LOCUS T (FT), bZIP transcription factor FD, and TERMINAL FLOWER-26 1 (TFL -1), and demonstrate their temperature -dependent roles in floral regulation . 27 Spatiotemporal expression analyses suggested that CsatFT3 and CsatFD2, 28 expressed in the floral meristem promote floral induction, while CsatTFL1-3 acts as a 29 floral repressor. Protein interaction studies showed that CsatFT3 and CsatTFL1 -3 30 compete for binding to CsatFD2, and their balance modulates floral induction . 31 Functional validation in Arabidopsis and Saffron confirmed these findings. 32 Furthermore, we identified CsatSVP2, an ortholog of SHORT VEGETATIVE PHASE 33 (SVP), as a low temperature-responsive repressor that directly binds the CsatFT3 34 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint promoter to inhibit its expression. Together, these findings enhance our understanding 35 of temperature mediated floral induction in saffron and provide insights and lay the 36 groundwork for genetic interventions to enhance yield under variable temperature 37 conditions. 38

Florigen activation complex (FAC), Flowering locus T(FT), bZIP 39 transcription factor-FD, TFL1, Geophytes, Thermoresponsive, Thermoperiodic, Floral 40 induction, Phosphatidyl Ethanolamine Binding Protein (PEBP), VIGS, 41 42

43 Saffron ( Crocus sativus L.) is a sterile, clonally propagated, autotriploid geophytic 44 monocot cultivated for its highly valued dried stigmas, which constitute the world’s 45 most expensive spice. While the stigmas are the primary edible and culinary 46 component, other parts of the saffron plant also have diverse applications, including 47 uses in agriculture, traditional medicine, and as a natural colouring agent, making the 48 entire flower economically valuable (Ahrazem et al., 2015) . Saffron has been 49 predominantly cultivated for centuries in select regions such as Iran, India, Greece, 50 Italy, Afghanistan, Morocco, Spain, and various Mediterranean basins (Cardone et al., 51 2020). Flowering is the key determinant of saffron’s crop productivity, and this critical 52 phase is increasingly influenced by environmental changes. As a thermoperiodic plant, 53 saffron’s flowering is primarily regulated by temperature. During the warm summer 54 months, the transition from vegetative to reproductive growth occurs underground, 55 with floral bud emergence only triggered by exposure to cooler temperatures (Molina 56 et al., 2005a) . This biphasic thermal requirement consists of an induction phase at 57 warm temperatures lasting 50 –150 days, followed by an emergence phase under 58 cooler conditions, a pattern that has been well documented (Jose-Santhi et al., 2023; 59 Molina et al., 2005a) . The inappropriate temperature during the flowering transition 60 leads to flower atrophy or no flower, causing yield loss (Wang et al., 2021) . Saffron 61 corms can sense the temperature change and modify the response accordingly 62 (Molina et al., 2005b) . Thus, there exists a thermoresponsive regulation of flowering 63 in saffron which is still not well studied at the molecular level. 64 Flowering is a key transition in the angiosperm life cycle, marking the shift from 65 vegetative to reproductive growth and representing a tightly regulated process vital for 66 survival. This process is orchestrated by a complex network of environmental cues, 67 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint internal signalling pathways, hormonal regulation, and genetic programs that together 68 ensure flowering occurs under optimal conditions (Zik and Irish, 2003) . Key 69 environmental pathways influencing flowering include photoperiod, ambient 70 temperature, light quality, vernalization, and gibberellin signalling. These pathways 71 operate in coordination with endogenous factors along with plant age and nutritional 72 status, allowing plants to synchronize reproduction with favourable environmental 73 windows (Andrés and Coupland, 2012; Freytes et al., 2021; Pyo et al., 2014; Srikanth 74 and Schmid, 2011) . Extensive studies in model organisms such as Arabidopsis 75 thaliana have elucidated major regulatory components of flowering (Kinoshita and 76 Richter, 2020) . Central to this regulation are floral integrator genes, particularly 77 FLOWERING LOCUS T (FT) a member of the phosphatidylethanolamine -binding 78 protein (PEBP) family. FT functions to promote flowering by forming the Florigen 79 Activation Complex (FAC), which includes FD -like bZIP transcription factors and the 80 florigen receptor 14 -3-3 protein (Abe et al., 2005; Taoka et al., 2013) . This complex 81 activates floral meristem identity genes such as LEAFY and APETALA1, initiating floral 82 development (Putterill et al., 2004; Simon et al., 1996; Wigge et al., 2005). In contrast, 83 TERMINAL FLOWER 1 (TFL1), another PEBP family member, acts antagonistically 84 to repress flowering (Wickland and Hanzawa, 2015) . Members of the FD family are 85 key interaction partners of the PEBP family, playing roles in both promoting and 86 repressing meristem differentiation, depending on their specific interacting partners 87 (Zhu et al., 2020) . Concurrently, the flowering program suppresses negative 88 regulators, including FLOWERING LOCUS C (FLC), and SHORT VEGETATIVE 89 PHASE (SVP), which are crucial for ensuring a proper transition from the vegetative 90 to the reproductive phase mainly regulated by temperature (Bowman et al., 2012; 91 Turck et al., 2008) . Temperature-dependent flowering has been widely studied, with 92 FT-like genes playing a central role (Capovilla et al., 2015) . Upstream regulators 93 include MADS-domain transcription factors such as SVP and FLC, which modulate 94 flowering by forming complexes and repressing FT expression via direct binding to its 95 promoter (Balasubramanian et al., 2006; Gu et al., 2013; Lee et al., 2007) . However, 96 the thermoresponsive pathway governing floral induction in saffron remains largely 97 unexplored. 98 In contrast to annuals, the regulation of flowering in perennial geophytes involves 99 modified or additional mechanisms to accommodate their extended life cycles, periods 100 of dormancy and dual reproduction (Khosa et al., 2021) . This has been observed in 101 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint species like Narcissus tazetta (Noy-Porat et al., 2013) , tulip (Tulipa spp.) 102 (Leeggangers et al., 2017b), and saffron (Molina et al., 2005a), where floral induction 103 occurs even in the absence of leaves. In these species, flowering is thought to be 104 mediated directly at the meristem level, independent of traditional photoperiod or 105 vernalization cues. Instead, ambient temperature plays a dominant role in determining 106 flowering time. These observations highlight the need to identify the specific genes 107 and mechanisms that govern flowering in geophytic perennials. Over evolutionary 108 time, the molecular mechanisms controlling flowering have grown increasingly 109 complex, involving intricate interactions between floral regulator genes and their 110 transcription factor partners, which form regulatory complexes (Wickland and 111 Hanzawa, 2015) . Gene duplication and neofunctionalization have contributed to the 112 diversification of these regulatory genes. Both FT and TFL1, the central regulators of 113 flowering, have undergone gene duplication events in various lineages, leading to 114 paralogs with distinct functions that promote or inhibit flowering (Jin et al., 2021) . This 115 evolutionary diversification is evident across many species. In geophytes, for instance, 116 members of the PEBP gene family have extended their functions beyond flowering to 117 include the formation of underground storage organs, contributing to both sexual and 118 asexual reproduction (Khosa et al., 2021) . In species like potato, tulip, and others, 119 gene duplication has led to distinct roles in flowering, vegetative growth, and tuber or 120 bulb formation (Bellinazzo et al., 2025; Jing et al., 2023; Navarro et al., 2011). Similar 121 divergence has been observed in onion (Allium cepa), sugar beet (Beta vulgaris), and 122 hybrid aspen, demonstrating the complexity of flowering regulation in perennials (Hsu 123 et al., 2011; Lee et al., 2013b; Pin et al., 2010) . Beyond the PEBP gene family, 124 functional diversification is seen in related pathways. In hybrid aspen, two FD 125 homologs, FDL1 and FDL2, regulate seasonal growth differently (Tylewicz et al., 126 2015). In rice, OsFD2 contributes to leaf development while OsFD1 regulates 127 flowering (Tsuji et al., 2013) . FD -like transcription factors can have dual roles, 128 promoting or repressing flowering depending on their partners. For instance, CmFDa 129 in Chrysanthemum morifolium represses flowering through epigenetic regulation (Xue 130 et al., 2025). These examples highlight the evolutionary plasticity of flowering networks 131 and the importance of species-specific regulatory mechanisms. 132 Saffron, due to its sterile nature and extremely limited genetic diversity, is not 133 amenable to conventional breeding approaches, which hinders efforts to develop 134 climate-resilient, high-yielding, and stress -tolerant varieties. As a result, there have 135 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint been no significant advancements in improving saffron’s adaptability to climate 136 change, making its cultivation increasingly vulnerable to environmental fluctuations. 137 Consequently, only biotechnological interventions can be employed to introduce 138 desirable traits. Despite its economic and agricultural importance, research on the 139 molecular regulation of flowering in saffron remains limited. Most existing studies have 140 focused on temperature effects on flowering physiology (Molina et al., 2005a; Wang 141 et al., 2021), and while several flowering -related genes have been identified through 142 transcriptomic and genomic analyses (Hu et al., 2020; Kalia et al., 2022; Renau -143 Morata et al., 2021; Singh et al., 2023; Tsaftaris et al., 2013) , these remain largely 144 descriptive and lack functional validation. Expression profiling has indicated the 145 involvement of multiple FT and TFL1 homologs in saffron flowering (Kalia et al., 2022; 146 Renau-Morata et al., 2021; Tsaftaris et al., 2013), yet no functional characterization of 147 these genes has been reported. To fill this gap, we present a comprehensive 148 molecular framework for temperature -dependent floral induction in saffron —a 149 geophytic monocot with a unique flowering phenotype. By integrating gene expression 150 profiling, heterologous functional assays, protein–protein interaction studies, promoter 151 binding analyses, and virus -induced gene silencing, we identify key regulatory 152 components that mediate the floral transition in response to ambient temperature. 153 Overall, these findings advance our understanding of temperature-mediated flowering 154 in saffron and provide a foundation for genetic interventions aimed at improving yield 155 under fluctuating environmental conditions 156 157

158 Plant materials and growth conditions 159 Saffron corms (Crocus sativus L.) were grown under controlled conditions at CSIR -160 IHBT, and samples were collected monthly from January to September, covering both 161 the vegetative growth and flowering phases (Jose-Santhi et al., 2023; Kalia et al., 162 2022). To analyze gene expression patterns, apical buds, axillary buds, and corm 163 tissues were harvested across this period, with three biological replicates per time 164 point, each comprising pooled tissue from ten individual corms. For RT -qPCR 165 analysis, apical buds, axillary bud, corm tissue collected during the vegetative and 166 flowering stages were used, with detailed methods and results provided in the 167 Supplementary Information. 168 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint ViGS experiments were conducted using corms grown under the same controlled 169 conditions as described earlier. Corms with high flowering competency (weighing >10 170 g) were selected for the silencing of CsatFT3 and CsatFD2. To investigate the role of 171 CsatSVP2, two experimental conditions were used: (i) standard growth conditions, and 172 (ii) a temperature -sensitive treatment in which corms were incubated at 8 °C during 173 the flowering phase to assess temperature -dependent gene regulation. For silencing 174 of CsatTFL1-3/CEN1, corms with lower flowering competency (weighing 7 –9 g) were 175 utilized. 176 Arabidopsis thaliana (Col-0) was used as the wild-type background for transformation 177 experiments, and the ft1 mutant in the Ler background was used for functional 178 complementation of CsatFT3. Nicotiana benthamiana was used for transient 179 expression assays. Arabidopsis seeds were surface -sterilized with 4% sodium 180 hypochlorite and germinated on half-strength Murashige and Skoog (MS) medium. To 181 synchronize germination, seeds were stratified at 4°C for two days before being 182 transferred to a soil mixture of peat, vermiculite, and perlite (2:1:1). All plants were 183 cultivated in controlled growth chambers under long -day conditions (16 h light/8 h 184 dark) at 22°C for Arabidopsis and 25°C for Nicotiana benthamiana. 185 186 RNA isolation and expression analysis by RT-qPCR 187 Total RNA was isolated from saffron apical buds, axillary buds, corm tissues, and 188 Arabidopsis using the Plant Total RNA Isolation Kit (Sigma -Aldrich), according to the 189 manufacturer’s protocol. For each sample, 1 µg of total RNA was used to synthesize 190 cDNA. Reverse transcription was performed using the RevertAid cDNA Synthesis Kit. 191 The transcript level of each gene was quantified by quantitative real-time PCR (q-PCR) 192 using gene -specific primers (Supplementary Table 1), generated by SnapGene 193 Viewer 5.3.1 software (version 0.4.0) using the System (BioRad CFX Opus Real time 194 PCR). Each expression analysis included three biological replicates, with each 195 replicate derived from pooled tissue of ten individual corms. RT -qPCR results are 196 presented as the mean ± standard error (SE M) of three technical replicates per 197 biological sample. 198 199 Sequence alignments and phylogenetic analysis 200 Protein sequences were aligned using the ClustalW algorithm with default parameters 201 to generate high-quality multiple sequence alignments. Visualization of the alignments 202 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint was performed using ESPript 3.x (Robert & Gouet, 2014) to highlight conserved 203 regions and structural features. The aligned sequences were exported in Newick 204 (.nwk) format for phylogenetic analysis. Phylogenetic tree construction was carried out 205 using the “build” function of ETE3 v3.1.3 (Huerta-Cepas et al., 2016), implemented via 206 the GenomeNet platform (https://www.genome.jp/tools/ete/). FastTree v2.1.8 (Price et 207 al., 2010) was used with default settings to infer approximately maximum -likelihood 208 phylogenetic trees. 209 210 Plant transformation 211 To generate Arabidopsis transgenic lines, full -length coding sequences (CDS) of the 212 target genes were amplified and first cloned into the pENTR entry vector using 213 Gateway® cloning technology. Subsequently, LR recombination was performed to 214 transfer the CDS into the destination vectors pK2GW7 (for overexpression analysis ), 215 under the control of the Cauliflower Mosaic Virus (CaMV) 35S promoter. The resulting 216 constructs included 35S:: CsatFD1, 35S::CsatFD2, 35S:: CsatFD3, 35S:: CsatFT3, 217 35S::CsatTFL1-1, 35S:: CsatTFL1-2, 35S:: CsatTFL1-3, 35S:: CsatSVP1, and 218 35S::CsatSVP2. Primer sequences used for vector construction are provided in the 219 supplementary table 1 file. The final recombinant plasmids were introduced into 220 Arabidopsis thaliana (Col-0) via Agrobacterium tumefaciens strain GV3101 using the 221 floral dip transformation method. Transgenic seeds were harvested individually and 222 screened based on the appropriate selection marker. Homozygous T3 lines were 223 identified and used for further phenotypic characterization. Flowering time was 224 recorded as the number of days from germination to the emergence of the first floral 225 bud. 226 227 Yeast one hybrid (Y1H) assay 228 To examine the binding of CsatSVP2 to the CsatFT3 promoter, yeast one-hybrid (Y1H) 229 assays were conducted using the Matchmaker® Gold Yeast One -Hybrid System Kit 230 (Clontech). Promoter fragments of CsatFT3 containing the core CArG-box motif were 231 cloned into the pAbAi vector. The resulting recombinant pAbAi -promoter plasmids 232 were linearized with BstBI (NEB) and integrated into the genome of the Y1HGold yeast 233 strain. Transformants were selected on synthetic dextrose (SD) medium lacking uracil 234 (SD/-Ura). The coding sequence of CsatSVP2 was amplified and inserted into the 235 pGADT7-AD vector. The recombinant pGADT7 -AD constructs were then introduced 236 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint into the Y1HGold strains containing the integrated CsatFT3 promoter. Protein –DNA 237 interaction was assessed by culturing the transformed yeast on SD medium lacking 238 leucine (SD/-Leu) and supplemented with Aureobasidin A (AbA) at a minimal inhibitory 239 concentration of 150 ng/mL for four days. Primer sequences used for cloning are 240 provided in Supplementary Table 1. 241 242 Histochemical localization of GUS activity 243 GUS activity was performed using the histochemical staining protocol described by 244 Jefferson(Jefferson et al., 1987) . Transgenic Arabidopsis thaliana lines with 245 ProCsatFT3::GUS construct were incubated in a staining solution composed of 50 mM 246 sodium phosphate buffer (pH 7.2), 2.0 mM potassium ferricyanide [K ₃Fe(CN)₆], 2.0 247 mM potassium ferrocyanide [K₄Fe(CN)₆], 0.1% (v/v) and 1.0 mg/mL X-Gluc (5-bromo-248 4-chloro-3-indolyl-β-D-glucuronide). Samples were incubated at 37 °C for 5 hours to 249 allow GUS expression to develop. Following staining, tissues were cleared with 70% 250 (v/v) ethanol to remove chlorophyll and enhance contrast, then examined under a 251 stereo microscope (ZEISS Stemi 508). To validate GUS expression patterns, at least 252 ten independent transgenic lines were screened for the construct. Plants from three 253 confirmed GUS-positive lines were selected for detailed expression analysis at various 254 developmental stages, including 7, 14, and 21 days after germination (DAG), 255 inflorescence emergence, flowering, and silique formation. For the flowering and 256 silique stages, staining was performed on aerial tissues from soil-grown plants 257 258 Yeast two-hybrid (Y2H) assays 259 Yeast two-hybrid assays were conducted following the manufacturer’s protocol using 260 the Matchmaker ™ Two-Hybrid System (Clontech). The coding sequences of 261 CsatFD1, CsatFD2, and CsatFD3 were cloned into the bait vector pGBKT7, whereas 262 those of CsatFT3, CsatTFL1 -1, CsatTFL1-2, and CsatTFL1-3 were inserted into the 263 prey vector pGADT7. The primers used for amplification are provided below. Bait and 264 prey plasmids were co -transformed into the Y2HGold yeast strain and cultured on 265 synthetic dextrose (SD) medium lacking tryptophan (SD/ -Trp) and leucine (SD/ -Leu) 266 for 3–5 days at 30°C. Protein-protein interactions were assessed by selecting colonies 267 on SD medium lacking Trp, Leu, and histidine (SD/-Trp/-Leu/-His). 268 269 Bimolecular fluorescence complementation (BiFC) 270 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint The coding sequences of CsatFD1, CsatFD2, CsatFD3 Cloned into YFP C -terminal 271 (YFPC) and CsatFT3, CsatTFL1-1, CsatTFL1-2, CsatTFL1-3 were cloned into the 272 YFP N-terminal (YFPN) vectors. Primer sequences used for vector construction are 273 provided in the supplementary table 1. These constructs were then introduced into 274 Agrobacterium tumefaciens strain GV3101 for transient expression in 5 -week-old 275 Nicotiana benthamiana leaves. After co -infiltration for 48 hours, fluorescence was 276 imaged using a confocal laser scanning microscope (Leica TCS SP5, Germany).YFP 277 fluorescence was excited at 514 nm, with emission detected between 522 and 560 278 nm. 279 280 Firefly Complementation Assay 281 The full-length coding sequences (CDSs) of CsatFT3, CsatTFL1-3, and CsatFD2 were 282 cloned into the pCAMBIA1300-nLUC and pCAMBIA1300-cLUC vectors, respectively, 283 and introduced into Agrobacterium tumefaciens strain GV3101. Various combinations 284 were co -infiltrated into Nicotiana benthamiana leaves. The infiltrated plants were 285 incubated at 22 °C for two days. Prior to LUC activity detection, the leaves were 286 sprayed with 1 mM D -luciferin (Potassium Salt, BioVision) and incubated in the dark 287 for 8 minutes to eliminate background fluorescence. Fluorescence was detected using 288 a cooled CCD camera system (Bio -Rad Chemi Doc). Relative LUC intensity was 289 quantified using ImageJ software. Statistical analysis was performed using one -way 290 ANOVA, and significance was assessed using the Compact Letter Display (CLD) 291

to indicate differences at a significant p -value level. Primer sequences used 292 in these assays are provided in Supplementary Table 1. 293 294 Genome walker library for promoter amplification 295 The 5′ upstream region of the CsatFT3 gene from Crocus sativus was obtained using 296 the Genome Walker ™ kit (Clontech, USA). Genomic DNA was extracted from leaf 297 tissue following the protocol provided with the QIAGEN DNA isolation kit, then 298 digested with two blunt-end restriction enzymes, DraI and EcoRV, and further ligated 299 with adapters, to generate Genome Walker libraries. Gene -specific primers (GSPs), 300 designed based on the CsatFT3 cDNA sequence, were used alongside adaptor 301 primers from the kit to carry out primary and nested PCR amplifications as listed in 302 Supplementary Table 1. The amplified PCR products were resolved on a 1% (w/v) 303 agarose gel, purified, and cloned into the pGEM -T Easy vector (Promega, USA) for 304 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint sequencing. The resulting CsatFT3 promoter sequence was analyzed using the 305 PlantCARE database (Lescot et al., 2002) to identify potential cis-regulatory elements 306 involved in transcriptional regulation. 307 308 Virus-induced gene silencing (VIGS) 309 Virus-induced gene silencing (VIGS) was carried out following the protocol established 310 in lab (Kalia et al., 2024) . Gene-specific fragments were amplified for CsatFT3 (300 311 bp, flanked by EcoRI sites), CsatFD2 (292 bp, flanked by BamHI sites), CsatTFL1-3 312 (302 bp, containing both EcoRI and BamHI sites), and CsatSVP2 (318 bp, with 313 EcoR1and BamH1 sites) restriction sites at both termini. These fragments were cloned 314 into the pTRV2 vector to generate the constructs. Agrobacterium-mediated infiltration 315 was performed as previously described (Kalia et al., 2024). To confirm gene silencing, 316 transcript levels of CsatFT3, CsatFD2, CsatTFL1-3, and CsatSVP2 were quantified by 317 RT-qPCR in apical buds collected at the floral induction stage. Phenotypic analysis of 318 flower development was conducted when floral buds reached the floral emergence 319 stage. Primer sequences used for the VIGS assay are listed in Supplementary table 320 1. 321 322 Luciferase assay 323 To investigate the regulatory effect of CsatSVP2 on the CsatFT3 promoter, a 324 luciferase reporter assay was conducted in Nicotiana benthamiana leaves. Promoter 325 sequences of CsatFT3 were cloned into the Gateway -compatible vector pGWB435 326 (Invitrogen) upstream of the luciferase (Luc) reporter gene. Three promoter constructs 327 were generated: (i) the full -length promoter (FL), (ii) a truncated fragment containing 328 the CArG motif(T1), and (iii) a truncated fragment lacking the CArG motif (T2) . The 329 CsatSVP2 coding sequence was separately cloned into pGWB402 under the control 330 of the CaMV35S promoter to serve as an effector. Control treatments included empty 331 vectors (pGWB435 and pGWB402) as a negative control, and ProCsatFT3::Luc alone 332 as a positive control. Constructs were introduced into the Agrobacterium tumefaciens 333 strain GV3101 and co -infiltrated into Nicotiana benthamiana leaves. Each leaf was 334 divided into four quadrants to enable side-by-side comparisons of different treatments 335 within the same biological context. Luciferase activity was visualized and quantified 336 using ImageJ software. The experiment was repeated independently at least three 337 times in separate leaves to ensure reproducibility. One -way ANOVA was used for 338 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint statistical analysis, and differences among treatments were determined using 339 Compact Letter Display (CLD) to indicate statistically significant differences (p < 0.05). 340 341 Statistical analyses 342 All statistical analyses were conducted using GraphPad Prism version 10 for Windows 343 (GraphPad Software, Boston, MA, USA; www.graphpad.com). One -way ANOVA 344 followed by Tukey’s Honest Significant Difference (HSD) test was applied for multiple 345 comparisons. Statistically significant differences between groups at p < 0.05 are 346 indicated by different letters, as determined by Compact Letter Display (CLD). Two-347 way ANOVA to assess the effects of temperature, month, and their interaction, 348 followed by Sidak’s multiple comparisons test and Student’s t -test for one -to-one 349 comparison. (*: p<0.05; **: p<0.01***: p<0.001) (****: p<0.0001). 350 351

352 CsatFT3 functions as a temperature-sensitive regulator of floral induction in 353 saffron 354 Our previous work identified several PEBP family genes implicated in flowering 355 regulation in saffron, with CsatFT3 emerging as a potential key regulator of floral 356 induction (Kalia et al., 2023). However, that study was limited to spatial and temporal 357 expression profiling during the reproductive phase transition. To further elucidate the 358 functional role of CsatFT3, we investigated its expression dynamics across vegetative 359 and reproductive phases in saffron apical buds. We found that CsatFT3 is specifically 360 expressed during the reproductive phase, with transcript accumulation coinciding with 361 the floral induction stage (Figure 1A, Supplementary Figure 1). In contrast, its 362 expression was barely detectable in apices of newly developing corms and during the 363 vegetative phase in winter. Given that low temperatures are known to suppress floral 364 induction in saffron (Jose-Santhi et al., 2023), we next assessed CsatFT3 expression 365 in apical buds of corms stored under low (8°C) and ambient (2 5°C) temperature 366 conditions following the vegetative phase. Consistent with a temperature -responsive 367 role, CsatFT3 transcript levels were markedly reduced in corms stored at 8°C 368 compared to those maintained at 25°C (Figure 1B). 369 To further investigate tissue - and temperature-specific expression of the CsatFT3 370 gene, we isolated a 1. 2 kb fragment of the CsatFT3 promoter (Supplementary figure 371 2) that contains several core, distal, and proximal promoter elements and constructed 372 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint a promoter-GUS fusion construct. GUS-driven transcription reporter lines of CsatFT3 373 displayed meristem and flower-specific expression in Arabidopsis, consistent with our 374 spatiotemporal expression profiling in saffron (Figure 1C and Supplementary figure 3). 375 Furthermore, these ProCsatFT3::GUS fusion lines demonstrated temperature -376 sensitive regulation, with expression gradually decreasing as the temperature was 377 lowered, in line with our previous findings that low temperatures suppress the 378 expression of CsatFT3 (Figure 1 D). At low temperatures (4°C), GUS activity was 379 nearly undetectable, indicating a significant suppression of expression under cold 380 conditions. These findings reinforce the role of CsatFT3 in temperature -sensitive 381 flowering regulation in saffron and suggest that its expression is activated under the 382 relatively high-temperature conditions that promote floral induction. 383 To functionally validate the role of CsatFT3 in floral regulation, we ectopically 384 expressed CsatFT3 in Arabidopsis Col -0 and complemented the late -flowering ft-10 385 mutant (Ler -0). Overexpression in Col -0 resulted in a significant early -flowering 386 phenotype (Figure 1E –G), while ft -10 plants transformed with CsatFT3 exhibited 387 restored or even earlier flowering compared to wild-type (Figure 1H–J). These results 388 confirm that CsatFT3 is functionally conserved and capable of promoting flowering in 389 a heterologous system. Following confirmation of CsatFT3’s involvement in floral 390 induction and flowering, finally, to evaluate the role of CsatFT3 in its native we 391 performed Virus -Induced Gene Silencing (VIGS) of CsatFT3 in saffron corms and 392 monitored flowering. Silencing of CsatFT3 led to a significant reduction in its transcript 393 levels in apical buds (Figure 1K) and was accompanied by a marked decrease in 394 flowering percentage compared to TRV2 mock -infected controls (Figure 1L –M). No 395 significant decrease of CsatFT1 and CsatFT2 was observed in TRV2-silenced 396 CsatFT3 corms (Supplementary Figure 4). Together, these findings establish CsatFT3 397 as a temperature -sensitive floral inducer in saffron, functioning as a key regulatory 398 component during the transition from vegetative to reproductive growth. 399 400 Identification of florigen activation complex component FD, involved in 401 flowering induction in saffron 402 FT like genes requires interaction with bZIP transcription factor -FD to form FT -FD 403 complex that directly activates the floral meristem identity genes. Thus, to identify the 404 FD like genes involved in floral induction of saffron we have cloned full length cDNAs 405 for three FD -like genes (CsatFD1, CsatFD2 and CsatFD3), which encodes for 197, 406 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint 215 and 245 aa of proteins respectively. All the three FDs are similar to FD proteins 407 from other plants and contains the conserved basic region, Leucine zipper region and 408 the functionally important conserved threonine (T)/SAP motif at the C terminus 409 (Supplementary Figure 5A). Phylogeny aligns CsatFD’s with monocot FDs with 410 Asparagus officinalis as the nearest family member (Supplementary Figure 5B). 411 We next investigated the expression profiles of the three saffron FD genes during the 412 vegetative and flowering induction stages. Interestingly, the expression of CsatFD2 413 coincided with the floral induction stage, whereas CsatFD3 showed higher expression 414 at later stages of sprouting, specifically in July, when floral differentiation and 415 vegetative growth begin (Figure 2 A). In contrast, CsatFD1 was expressed at relatively 416 stable levels throughout flower initiation and the vegetative phase, but its expression 417 was generally lower compared to CsatFD2 and CsatFD3. In saffron, the apical bud 418 gives rise to the flower, while the axillary buds contribute only to vegetative growth, 419 thereby distinguishing the reproductive meristem from the vegetative meristems (Kalia 420 et al., 2022) . Notably, CsatFD2 was predominantly expressed in apical buds and 421 flower tissues, indicating a primary role in flowering (Figure 2B, C). Additionally, 422 CsatFD2 showed comparatively higher expression in flower tissues compared to the 423 other two FD genes (Figure 2C). We also examined the temperature -sensitive 424 expression of these genes in corms stored at low and ambient high temperatures. Our 425

showed that only CsatFD2 expression was significantly downregulated at low 426 temperatures (Figure 2D) , while the expression of CsatFD1 and CsatFD3 did not 427 exhibit a similar pattern (Supplementary Figure 6A). Interestingly, CsatFD3 was highly 428 expressed in corms stored at low temperatures (Supplementary Figure 6B). 429 Next, we examined the effect of ectopic expression of the saffron FD genes in 430 Arabidopsis. Ectopic expression of CsatFD2 can only induce early flowering in 431 Arabidopsis, whereas there was no significant phenotypic change observed in 432 CsatFD1 and CsatFD3 expressing Arabidopsis plants (Figure 2E). These findings 433 were further supported by measurements of days to flowering and the number of 434 rosette leaves in the Arabidopsis transformants (Figures 2F -K). These results further 435 suggest that CsatFD2 may be the primary gene involved in flowering regulation in 436 saffron. To confirm the role of CsatFD2 during floral induction, we silenced the gene 437 using Virus -Induced Gene Silencing (VIGS) in flower -competent saffron corms. 438 Compared to the TRV2 control, silencing of CsatFD2 resulted in a reduction in flower 439 formation and overall flowering percentage (Figure 2L). VIGS silencing of CsatFD2 440 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint also led to a significant decrease in its transcript levels (Figure 2M), but did not affect 441 the expression of CsatFD1 and CsatFD3 (Supplementary Figure 7 A and B ). These 442

suggest that CsatFD2 is a key component of the Floral Activator Complex 443 (FAC) involved in floral induction in saffron. 444 445 TFL1-3/CEN1 is a newly identified negative regulator of flowering induction in 446 saffron 447 In our previous study (Kalia et al., 2023) , we identified two homologs of TFL-1/CEN, 448 named CsatTFL1-1 and CsatTFL1-2, and through spatio -temporal expression 449 profiling, we proposed that TFL1 -2 may play a role in determining flowering 450 competency in saffron (Supplementary figure 8A and B) . However, neither of these 451 two TFL-1 genes exhibited a temperature-sensitive response (Supplementary Figure 452 9A and B) . Furthermore, ectopic expression of CsatTFL1-1 and CsatTFL1-2 in 453 Arabidopsis led to delayed bolting in only CsatTFL1-2 plants compared to wild -type 454 (WT) under long -day (LD) conditions, whereas CsatTFL1-1 transformants bolted 455 normally, similar to WT plants (Supplementary figure 10A-F). Additionally, interaction 456 studies, including BiFC and Y2H assays, did not reveal any positive interactions with 457 CsatFD2, our identified positive regulator of flowering (Supplementary Figure 11A and 458 B). These findings prompted us to investigate further, hypothesizing that another 459 homolog of TFL1/CEN might be involved in floral induction in saffron. Through corm 460 development studies in our lab (Jose-Santhi et al., 2023) , we identified a third TFL1 461 homolog, which exhibited higher expression during the vegetative phase. This gene 462 shares similarities with CsatTFL1-1 and CsatTFL1-2, but phylogenetic analysis 463 clustered it within the CEN subgroup. Based on this, we named it CsatTFL1-3/CEN1 464 (Supplementary Figure 12A and B ). Like other TFL1/CEN family members, TFL1 -465 3/CEN1 contains the conserved domains characteristic of this gene family. 466 Quantitative PCR (qPCR) expression profiling revealed that CsatTFL1-3/CEN1 467 expression is elevated during the vegetative phase, with a significant decrease during 468 the reproductive phase (Figure 3 A). Tissue -specific expression analysis confirmed 469 that CsatTFL1-3/CEN1 is predominantly expressed in the axillary bud tissue, further 470 supporting its potential role in negatively regulating flowering (Figure 3 C and D). 471 Next, we examined the expression of CsatTFL1-3/CEN1 in corms stored at low (8°C) 472 and ambient (2 5°C) temperatures. Interestingly, there was no reduction in TFL1-473 3/CEN1 expression in corms stored at low temperatures compared to those stored at 474 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint ambient high temperatures, suggesting that it may be regulated by temperature and 475 could play a role in temperature -mediated floral induction in saffron (Figure 3B). We 476 then performed ectopic expression of CsatTFL1-3/CEN1 in Arabidopsis thaliana (Col-477 0) under the control of the constitutive 35S promoter. Transgenic lines exhibited a 478 significant delay in flowering compared to wild-type plants, as evidenced by increased 479 days to bolting and a higher number of rosette leaves at flowering (Figure 3F-G). 480 These results confirm the ability of TFL1 -3/CEN1 to act as a floral repressor in 481 Arabidopsis, consistent with its proposed function in saffron. Notably, ectopic 482 expression of TFL1 -3/CEN1 also led to a hyper -vegetative shoot phenotype, 483 characterized by the development of an inflorescence composed of leaf primordia 484 surrounded by small, serrated leaves—a phenotype not observed in TFL1-1 or TFL1-485 2 expressing Arabidopsis lines (Figure 3E). Finally, we performed Virus-Induced Gene 486 Silencing (VIGS) of CsatTFL1-3/CEN1 in low -flower-competent saffron corms. In 487 comparison to TRV2 control plants, silencing of CsatTFL1-3/CEN1 led to the formation 488 of floral structures in 7-9 g corms (classified as low -flower-competent), and these 489 flowers appeared earlier than in TRV2 controls (Figure 3G -I). Importantly, VIGS of 490 CsatTFL1-3/CEN1 did not affect the transcript levels of the other two TFL1 homologs, 491 TFL1-1 and TFL1 -2 (Supplementary Figure 13 A and B ), suggesting TFL1 -3/CEN1 492 and not TFL1 -1 or TFL1-2 could play a role in flowering in saffron . This observation 493 suggests that TFL1 -3/CEN1 acts as a negative regulator of floral induction, and its 494 silencing promotes floral transition in otherwise low-flowering competence corms. 495 496 The FT-FD-TFL complex involved in floral regulation 497 Our earlier analyses identified CsatFT3, CsatFD2, and CsatTFL1-3/CEN1 as potential 498 core components of a Floral Activator Complex (FAC) in saffron. To investigate 499 whether these proteins physically interact to form functional regulatory complexes, 500 we employed two complementary approaches: yeast two -hybrid (Y2H) assays and 501 bimolecular fluorescence complementation (BiFC) in Nicotiana benthamiana. In the 502 Y2H assays, the coding sequences (CDS) of the three CsatFD genes were cloned into 503 the pGADT7 prey vector, and co -transformed into yeast with either CsatFT3 or 504 CsatTFL1-3/CEN1 fused to the pGBKT7 bait vector. Interaction screening revealed 505 that both CsatFT3 and TFL1-3/CEN1 specifically interacted with CsatFD2, but not with 506 CsatFD1 or CsatFD3 (Figure 4 A and B ). These results suggest that CsatFD2 may 507 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint serve as a central interacting hub for both activator (FT3) and repressor (TFL1 -3) 508 proteins within the FAC. 509 We further validated these interactions by BiFC assays in Nicotiana benthamiana 510 leaves. YFP fluorescence was observed in nuclei of cells co-expressing the C-terminal 511 half of YFP fused to CsatFD2 (cYFP–CsatFD2) with the N-terminal half of YFP fused 512 to either CsatFT3 (nYFP –CsatFT3) or CsatTFL1 -3 (nYFP–CsatTFL1-3) (Figure 4 C 513 and D ). No YFP fluorescence was detected in negative controls, including co -514 expression of cYFP–CsatFD2 with nYFP alone or cYFP alone with nYFP–CsatFT3 or 515 CsatTFL1-3. Furthermore, consistent with the Y2H results, no BiFC signal was 516 observed when CsatFT3 or TFL1 -3/CEN1 were co -expressed with CsatFD1 or 517 CsatFD3 (Supplementary Figure). Together, these findings strongly support a model 518 in which CsatFD2 specifically interacts with both CsatFT3 and TFL1 -3/CEN1, 519 suggesting that competitive complex formation may modulate floral transition in 520 saffron. 521 Given the antagonistic roles of CsatFT3 and CsatTFL1-3 in the regulation of flowering, 522 and their respective interactions with the transcription factor CsatFD2, it was 523 hypothesized that CsatFT3 and CsatTFL1-3 may compete for binding to CsatFD2. To 524 investigate this, a luciferase complementation imaging (LCI) assay was conducted in 525 Nicotiana benthamiana leaves to evaluate the interaction strength between CsatFD2-526 nLUC and cLUC -CsatTFL1-3 in the presence or absence of SK -CsatFT3.Strong 527 luciferase (LUC) activity was observed when CsatFD2 -nLUC and cLUC-CsatTFL1-3 528 were coexpressed, indicating a robust interaction. However, coexpression with SK -529 CsatFT3 significantly reduced LUC activity, suggesting that CsatFT3 interferes with 530 the interaction between CsatTFL1 -3 and CsatFD2. Moreover, this suppression was 531 dose-dependent, with increasing amounts of CsatFT3 leading to a progressive 532 decrease in LUC signal (Figure 4E). Biological replicates supporting these 533 observations are presented in Supplementary figure 15 A and B . These results 534 strongly suggest that CsatFT3 competes with CsatTFL1 -3 for interaction with 535 CsatFD2, which interferes with CsatTFL1-3's ability to interact with CsatFD2. 536 537 CsatSVP2 Acts Upstream of FT3 in Temperature-Mediated Floral Repression 538 To identify upstream genes potentially involved in temperature -dependent floral 539 induction in saffron, we screened transcriptome data representing the suppression of 540 floral induction under low temperatures (Jose-Santhi et al., 2023) and identified two 541 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint partial SVP-like MADS-box genes that were upregulated in non -flowering samples. 542 SVP-like genes are known to regulate flowering in a temperature -dependent manner 543 (Lee et al., 2007). Using long-read and other in-house transcriptome data, and based 544 on sequence similarity, we obtained full -length sequences of these two SVP -like 545 genes. We isolated and cloned them from saffron, naming them CsatSVP1 and 546 CsatSVP2. Sequence analysis indicated that while these genes are highly similar, they 547 code for different amino acid sequences. Multiple sequence alignment revealed that 548 both CsatSVP1 and CsatSVP2 contain the conserved MADS -box domain, and 549 phylogenetic analysis grouped them with SVP-like genes from other geophytes, such 550 as Narcissus, Zingiber, and Lilium (Supplementary Figure 16 A and B). 551 To further characterize their role in flowering, we performed expression analysis of 552 CsatSVP1 and CsatSVP2 during vegetative and floral induction stages. Both genes 553 showed a significant reduction in expression during the floral induction stage (June); 554 however, the reduction in CsatSVP2 was more pronounced and correlated closely with 555 floral induction and subsequent stages (Figure 5A) . We then evaluated their 556 expression in corms stored under floral inductive (25°C) and non -inductive (8°C) 557 conditions. CsatSVP2 transcript levels remained high at both temperatures, 558 suggesting it may contribute to the floral repression observed at low temperatures 559 (Figure 5C) . In contrast, CsatSVP1 expression decreased under both conditions, 560 showing no temperature -dependent pattern (Figure 5B) . Additionally, CsatSVP2 561 expression was comparatively higher in floral tissues than CsatSVP1, further 562 supporting its potential regulatory role (Supplementary Figure 17). 563 Next, to functionally validate these genes, we ectopically expressed both CsatSVP1 564 and CsatSVP2 in Arabidopsis under a constitutive promoter. Ectopic expression of 565 both genes resulted in delayed flowering, with a more pronounced delay in the 566 CsatSVP2-transformed lines compared to both the control and CsatSVP1 lines (Figure 567 5D-I). Arabidopsis plants expressing CsatSVP2 displayed various phenotypic changes 568 in flower bud and flower morphology, producing smaller, more compact flowers and 569 shorter siliques (Supplementary Figure 18). To investigate the role of SVP in 570 temperature-mediated floral suppression, we silenced CsatSVP2 in saffron corms 571 using Virus -Induced Gene Silencing (VIGS). CsatSVP2 was selected due to its 572 expression pattern and phenotype in Arabidopsis, which correlated with its expected 573 function. Since low temperatures during storage or dormancy are known to suppress 574 flowering, and CsatSVP2 may be involved in this process, we stored the CsatSVP2-575 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint silenced corms at low temperatures, alongside a mock control (TRV2). Silencing of 576 SVP2 resulted in flowering in corms at low temperatures , whereas the TRV2-treated 577 corms stored at low temperatures did not flower (Figure 5J). In parallel, corms stored 578 at 25°C—an inductive condition—showed flowering in both control and SVP2-silenced 579 groups, confirming that CsatSVP2 acts specifically in the floral suppression 580 mechanism under cold conditions (Figure 5K) . Gene expression analysis further 581 revealed that CsatSVP2 silencing prevented the typical low -temperature-mediated 582 suppression of CsatFT3 expression, thereby facilitating flowering (Figure 5P and Q). 583 These findings suggest that CsatSVP2 and CsatFT3 act within the same regulatory 584 pathway, mediating temperature-responsive floral induction in saffron. 585 586 SVP2 Directly Binds to the CsatFT3 Promoter and Represses Its Expression 587 Our previous results indicated that CsatSVP2 and CsatFT3 function in the same 588 temperature-sensitive regulatory pathway: low temperature maintains or upregulates 589 CsatSVP2 expression, which in turn suppresses CsatFT3, thereby repressing floral 590 induction. Notably, CsatFT3 transcript levels were restored in CsatSVP2-silenced 591 corms, suggesting that CsatSVP2 functions upstream of CsatFT3. To further explore 592 the mechanism of this regulation, we analyzed the CsatFT3 promoter sequence to 593 identify potential binding motifs for MADS -box transcription factors. In silico analysis 594 of the 1. 2 kb FT3 promoter, previously shown to drive temperature -responsive 595 expression, revealed a conserved CArG -box motif (CCATTTAAGG) located 596 approximately 827 bp upstream of the ATG start codon that is established for binding 597 of MADS boxes genes including SVP orthologs (Figure 6A). To experimentally test 598 further whether CsatSVP2 directly binds to the CsatFT3 promoter, we performed a 599 yeast one-hybrid (Y1H) assay. The Y1H assay confirmed that CsatSVP2 specifically 600 binds to the CsatFT3 promoter, supporting the hypothesis of direct regulation (Figure 601 6B). To further validate the functional specificity of this interaction, we employed a 602 luciferase reporter assay using Nicotiana benthamiana transient expression (Figure 603 6C) and conducted a luciferase reporter assay using three different FT3 promoter 604 fragments: (1) the full-length promoter (~1.2kb), (2) a truncated version containing the 605 conserved CArG motif (T1), and (3) a truncated version lacking the motif (T2) (Figure 606 6D). Co -expression of CsatSVP2 with these constructs in Nicotiana benthamiana 607 revealed that luciferase activity was significantly reduced in both the full-length and T1 608 constructs (Figure6 D and F) . In contrast, no repression was observed with the T2 609 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint construct, confirming that SVP2 specifically suppresses FT3 expression through direct 610 binding to the CArG motif (Figure 6F). Together, these results establish a mechanistic 611 model in which CsatSVP2 acts as a temperature -responsive floral repressor that 612 directly binds to the FT3 promoter via a conserved CArG motif, thereby repressing 613 FT3 expression and floral induction under low-temperature conditions. 614 615

616 Saffron is a high -value crop cultivated in limited regions due to its strict temperature 617 requirements, which distinctly regulate its vegetative and reproductive growth. High 618 temperatures induce flowering, while subsequent cooling triggers flower formation. In 619 contrast, low temperatures suppress floral induction and support vegetative growth, 620 enabling overwintering and daughter corm development (Molina et al., 2005a) . This 621 temperature sensitivity restricts saffron cultivation to specific climates. Although 622 environmental control of flowering has enabled cultivation in soilless and controlled 623 settings, the molecular mechanisms underlying temperature -regulated flowering 624 remain poorly understood. As a sterile crop, saffron cannot be improved through 625 conventional breeding, making biotechnological interventions essential . However, 626 limited genomic and genetic resources, challenges in genetic transformation, and a 627 scarcity of molecular studies have hindered progress in functional genomics—creating 628 a critical knowledge gap that must be addressed. 629 In this study, we have identified a temperature -sensitive transcriptional network that 630 regulates flowering in saffron, uncovering both conserved and saffron -specific 631 components involved in floral induction (Figure 7). Our data establish a regulatory 632 module involving CsatFT3, CsatFD2, CsatTFL1 -3/CEN1, and CsatSVP2 as key 633 components in mediating floral transition in response to temperature cues. Although 634 the core elements of the Florigen Activation Complex (FAC)—FT, FD, and TFL1—are 635 conserved across angiosperms, their expression patterns, protein interactions, and 636 functional dynamics in saffron suggest neofunctionalization in line with its geophytic 637 lifecycle. For instance, Notably, CsatFT3 and CsatFD2 are upregulated at higher 638 temperatures and co -expressed in the floral meristem, promoting flowering. In 639 contrast, other FT and FD homologs show no such expression correlation (Figure 2; 640 (Kalia et al., 2022) , indicating their limited or no involvement in floral induction. 641 Similarly, CsatTFL1 -3 functions as a floral repressor by competitively binding to 642 CsatFD2, whereas the other two TFL1 homologs neither interact with CsatFD2 nor 643 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint delay flowering in Arabidopsis, suggesting a lack of functional redundancy and 644 functional specificity. Functional validation in both Arabidopsis and saffron confirms 645 that floral induction is governed by the dynamic interplay between CsatFT3, CsatFD2, 646 and CsatTFL1 -3.Given the roles of gene duplication, neofunctionalization, and 647 lineage-specific gene expansion in driving the specialization of reproductive and 648 vegetative pathways in geophytes (Khosa et al., 2021) , our identification of distinct 649 regulatory factors and their molecular interactors from multigene families in saffron 650 represents a significant advance. 651 A distinguishing feature of floral induction observed in saffron is the spatial restriction 652 of flowering -related gene expression predominantly to the meristem. All the genes 653 identified in this study were specifically expressed in the apical (flowering) meristem. 654 During floral induction, saffron corms are either underground or stored in the dark and 655 remain leafless, indicating that the molecular signals initiating flowering are localized 656 within the apical buds. Such meristem-localized control may reflect an adaptation to 657 the geophytic habit of saffron, which undergoes floral induction in the absence of 658 foliage, similar to what has been observed in other geophytes such as tulip s 659 (Leeggangers et al., 2017b) . This localization supports the hypothesis that 660 temperature cues are sensed and interpreted locally within the corm, rather than 661 systemically via leaves, as observed in many other species, including some geophytes 662 (Lee et al., 2013b; Navarro et al., 2011; Wigge, 2011; Wu et al., 2022; Yan et al., 663 2021). 664 Temperature plays a pivotal role in regulating flowering time across plant species, 665 often through conserved thermosensory pathways. In Arabidopsis, warmer 666 temperatures promote flowering via FT activation (Balasubramanian et al., 2006) . 667 while in crops like strawberry and onion, cold exposure induces flowering by triggering 668 FT-like genes (Heide, 1977; Lee et al., 2013b). Interestingly, the effect of temperature 669 on flowering is not always straightforward. Temperatures above 25°C seem to have 670 both inductive and repressive effects on flowering, depending on the species and 671 environmental conditions (Nakano et al., 2013; Noy -Porat et al., 2013; Oda et al., 672 2012). This temperature sensitivity is closely linked to the regulation of FT, a key gene 673 that integrates various environmental signals and controls the timing of flowering. In 674 Narcissus tazetta (daffodils), floral induction is triggered by high ambient temperatures 675 (25°C) through the activation of NtFT, while in tulips, TgTFL1 downregulation and 676 TgFT2 upregulation contribute to temperature -induced floral induction (Leeggangers 677 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint et al., 2017a; Noy-Porat et al., 2013). In contrast, in species like onion and lily, which 678 flower in spring, vernalization (low -temperature treatment) is required to activate the 679 expression of AcFT2 and LiFT, respectively (Lee et al., 2013b). In our study, we also 680 found that CsatFT3, a key flowering -time gene in saffron, is temperature -regulated, 681 with its expression being induced by high ambient temperatures and suppressed by 682 low temperatures (Figure 1B) . Similarly, CsatFD2, a partner of CsatFT3 in the 683 flowering regulatory network, follows a similar expression pattern, suggesting their 684 involvement in the temperature -mediated floral induction pathway in saffron. 685 Interestingly, we observed that CsatTFL1-3, a gene that maintains the vegetative state 686 and prevents early flowering, exhibits a reciprocal expression pattern. In Arabidopsis 687 low and ambient temperature regulation of flowering involves the gene TFL1 (Strasser 688 et al., 2009) . Higher expression of CsatTFL1-3 is observed during the vegetative 689 phase, which coincides with the low winter temperatures during saffron's growth cycle 690 (Figure 3A). This expression pattern is crucial for maintaining the vegetative phase 691 and delaying flowering, as confirmed by the hyper -vegetative shoot phenotype in 692 TFL1-3 overexpressing Arabidopsis plants (Figure 3) similar to what was observed 693 when Arabidopsis TFL1 is overexpressed (Lee et al., 2019) . In contrast, as 694 temperatures rise in the summer, the expression of CsatFT3 increases (Figure 1A), 695 likely triggering the activation of floral pathway genes and the transition to reproductive 696 development. Overall, a balance between CsatFT3, CsatTFL1-3, and their interaction 697 with CsatFD2 governs floral induction in a temperature -dependent manner. This 698 delicate interplay between temperature and gene expression is vital for ensuring 699 proper timing of flowering in saffron, aligning the plant’s reproductive phase with 700 favourable environmental conditions. 701 We also identify CsatSVP2, a MADS -box transcription factor, as a critical floral 702 repressor acting upstream of FT–FD–TFL core module . MADS-box genes are key 703 regulators of flowering time, with SVP -like genes playing central roles in ambient 704 temperature-mediated floral control (Jin and Ahn, 2021) . CsatSVP2 directly binds a 705 conserved CArG-box in the CsatFT3 promoter, repressing its expression under low 706 temperatures. Silencing of SVP2 alleviated low -temperature-induced flowering 707 suppression and restored CsatFT3 expression, thereby establishing a direct 708 mechanistic link between ambient temperature perception and floral induction (Figure 709 5). In Arabidopsis thaliana , a comparable regulatory pathway has been described, 710 where SVP negatively regulates FT expression by directly binding to a CArG motif in 711 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint its promoter (Lee et al., 2007). Additionally, SVP levels are known to increase under 712 low temperatures and are degraded at higher temperatures, facilitating early flowering 713 as SVP abundance declines (Blázquez et al., 2002; Lee et al., 2013a; Sureshkumar 714 et al., 2016) . Interestingly, individual members of the SVP gene family have been 715 shown to play distinct roles in processes such as bud dormancy, flowering initiation, 716 and floral development reflecting adaptations of developmental timing to local 717 environmental conditions and seasonal temperature fluctuations. In tulips, gene 718 expression analyses have similarly identified TgMADS15/16 and TgMADS9 as low-719 temperature-responsive SVP-like genes involved in floral induction (Lu et al., 2023) , 720 reinforcing the conserved role of this gene family in temperature -mediated flowering 721 regulation across geophytic species. SVP homologs in Lilium and Narcissus tazetta 722 have also been reported as flowering repressors, though their temperature 723 responsiveness remains unexplored (Li et al., 2015; Wu et al., 2024). Notably, ectopic 724 expression of CsatSVP2 in Arabidopsis caused floral morphology defects, reminiscent 725 of phenotypes observed with Lilium SVP expression (Wu et al., 2024), suggesting that 726 these homologs may also contribute to flower development in addition to regulating 727 flowering time. 728 729

730 Collectively, our findings establish a detailed molecular framework for temperature -731 regulated flowering in saffron, centered on the dynamic balance between floral 732 activators (CsatFT3, CsatFD2) and repressors (CsatTFL1 -3, CsatSVP2). The 733 spatiotemporal expression patterns and protein interactions among these components 734 provide key insights into how flowering is precisely modulated in response to 735 environmental temperature fluctuations. By uncovering saffron -specific adaptations 736 within otherwise conserved gene networks, this study advances our understanding of 737 floral regulation in geophytes and sterile crops. Importantly, these results lay a 738 foundation for the development of genetic and biotechnological strategies to improve 739 saffron yield, synchronize flowering, and enhance resilience to climate variability. Our 740 work highlights a saffron -specific model of thermoresponsive flowering that builds 741 upon conserved floral regulatory pathways, yet exhibits functional divergence shaped 742 by the geophytic lifecycle. Beyond saffron, these findings offer a valuable framework 743 for translational research in other temperature -sensitive geophytes such as tulip, 744 Narcissus, Hyacinthus, and Iris, where molecular insights remain scarce (Khodorova 745 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint and Boitel-Conti, 2013). By identifying novel regulatory targets, this study opens new 746 avenues for improving flowering and productivity in saffron and related high -value 747 geophytic species. 748 749 Acknowledgments 750 We gratefully acknowledge the financial support from the Council of Scientific and 751 Industrial Research (CSIR) under the Genome Editing Mission for Crop Improvement 752 (MMP25301), the intramural CSIR grant to IHBT (MLP201), the SERB -Start-up 753 Research Grant (SRG/GAP0288), and the Department of Biotechnology 754 (DBT/GAP0307) to Rajesh Kumar Singh. Senior Research Fellowships provided to 755 Diksha Kalia (UGC), Joel Jose-Santhi (CSIR), and Firdous Rasool Sheikh (CSIR) are 756 also gratefully acknowledged. We sincerely thank Dr. Rishikesh Bhalerao, Umeå Plant 757 Science Centre, Swedish University of Agricultural Sciences, Sweden, for his careful 758 reading of the manuscript and for providing constructive feedback, which significantly 759 improved the clarity and quality of the work. Our sincere thanks go to Dr. Rimpy Diman, 760 Technical Officer at CSIR-IHBT, for her valuable assistance with confocal microscopy. 761 This manuscript represents the CSIR-IHBT communication number "5819". 762 763 Author contribution statement 764 R.K.S. conceptualized the research idea and designed the work plan. D.K., J.J.S., and 765 F.R.S. conducted the experiments and collected the data. D.K., J.J.S., and R.K.S. 766 analyzed the data and contributed to writing the manuscript. All authors read and 767 approved the final version of the manuscript. 768 769 Conflicts of Interest 770 The authors declare no conflicts of interest. 771 772 Figure Legends 773 Figure 1. CsatFT3 is involved in temperature mediated floral induction in saffron. 774 (A) CsatFT3 transcript levels in apical buds across developmental stages show 775 specific expression during the reproductive phase. (B) Relative expression of CsatFT3 776 in apical buds of saffron corms stored at ambient (25°C) and low (8°C) temperatures. 777 Data are means ± SEM from 3 biological replications, where n=10 for each replicate. 778 (C) Histochemical GUS assay of the CsatFT3 promoter (proCsatFT3) in transgenic 779 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Arabidopsis in long ‐day conditions (LD) during various developmental stages. 1. 2kb 780 sequence upstream of the CsatFT3 translation start sites (proCsatFT3) was fused with 781 uidA (GUS) reporter gene. Different developmental stages include Arabidopsis plants 782 after 7 Days After Germination (DAG), 14DAG, 21DAG, and 30DAG. FT3 expression 783 is confined to meristem and inflorescence (D) Temperature sensitive GUS activity at 784 3 different temperatures. The experiment was repeated three times independently with 785 similar results. (E-G) Ectopic expression of CsatFT3 in Arabidopsis resulted in early 786 flowering phenotype. Statistical significance was analyzed by Student’s t -test (*P < 787 0.05). (H-J) Functional complementation of the ft -10 mutant with CsatFT3 restores 788 flowering. Representative line expressing number of rosette leaves and number of 789 days to flower. Total leaf number was calculated by combining total rosette leaves. 790 Data represent a minimum of 10 plants scored for each line. Statistical significance 791 was analyzed by Student’s t -test (*P < 0.05). (K) Representative flowering vs. non -792 flowering phenotypes in control and CsatFT3 -silenced saffron corms. The scale bar 793 represents 2 cm. (L) qRT-PCR validation of CsatFT3 transcript knockdown in saffron 794 corm apices following VIGS (n=3 technical replicates, 10 independent corms). Error 795 bars are shown as Standard Error of Mean (SEM) of three technical repeats. (M) 796 Percentage of flowering in VIGS-silenced vs. TRV2 control corms. The data presented 797 includes the mean percentage of flowering, with error bars representing the Standard 798 Error of Mean (SEM) of three technical repeats (n=3) and for each replicate, 10 corms 799 were used. All data in this figure are shown as mean ± SEM. Statistical Significance 800 for Data in panel A, I, & J are analyzed using one -way ANOVA followed by Tukey’s 801 Honest Significant Difference (HSD) test for multiple comparisons. Different letters 802 indicate statistically significant differences between groups at p < 0.05, according to 803 Compact Letter Display (CLD). Data in panels B are analyzed using two-way ANOVA 804 to assess the effects of temperature, month, and their interaction, followed by Sidak’s 805 multiple comparisons test. Asterisks indicate significant differences between 8°C and 806 25°C within each month (**** p < 0.0001 ), F, G, L, and M are analyzed using the 807 student’s t-test (*: p<0.05; **: p<0.01). 808 809 Figure 2: Functional characterization of FD -like genes in saffron (A) Expression 810 profiles of CsatFD1, CsatFD2, and CsatFD3 during vegetative and floral induction 811 stages in the apical bud by RT -qPCR. Data are means ± SE of 3 independent 812 replicates. (B) Expression patterns of CsatFD1, CsatFD2, and CsatFD3 in the Main 813 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint bud (MB), axillary bud (AB), and corm tissue (CT) during the floral induction stage. 814 Values are means ± SE of 3 independent replicates. (C) Tissue specific expression of 815 FDs in different parts of saffron plant. (D) Expression analyses of CsatFD genes in 816 apical buds of corms stored at 25 ◦ C and 8 during months of May and June ( E) 817 Phenotypic comparison of early flowering in CsatFD2 -expressing Arabidopsis 818 compared to control. (F-K) Measurements of days to flowering and rosette leaf number 819 in CsatFD1, CsatFD2, and CsatFD3 expressing Arabidopsis plants. Values are means 820 ± SE (n = 10). Asterisks indicate a significant difference between the transgenic 821 Arabidopsis lines and wild -type plants (Col -0; Student’s t -test, P < 0.05). (L) 822 Phenotypes of CsatFD2 -TRV2 silenced lines and TRV2 lines. The scale bar 823 represents 2 cm. (M) qRT-PCR validation of CsatFD2 knockdown in saffron corm 824 apices after VIGS (n=3). Error bars are shown as SD of three technical repeats (N) 825 Flowering percentage in TRV2 control and CsatFD2 VIGS-silenced saffron corms. The 826 data presented includes the mean percentage of flowering, with error bars 827 representing the standard error of mean (SEM) of three technical repeats (n=3) and 828 for each replicate, 10 corms were used. All data in this figure are shown as mean ± 829 SEM. Statistical Significance for data in panel A & B are analyzed using two-way 830 ANOVA followed by Tukey’s Honest Significant Difference (HSD) test for multiple 831 comparisons. Different letters indicate statistically significant differences between 832 groups at p < 0.05, according to Compact Letter Display (CLD). Data in panel D are 833 analyzed using two-way ANOVA to assess the effects of temperature, month, and their 834 interaction, followed by Sidak’s multiple comparisons test. Asterisks indicate 835 significant differences between 8°C and 25°C within each month (**** p < 0.0001), , 836 F,G,H,I,J,K,M and N are analyzed using the Student’s t-test (*: p<0.05; **: p<0.01). 837 838 Figure 3: CsatTFL1 -3 is negative regulator of floral induction in saffron ( A) 839 Relative expression of CsatTFL1 -3 during vegetative and reproductive phase in the 840 apical bud by RT -qPCR. Data are means ± SE of 3 independent replicates. (B) 841 Expression analyses of CsatTFL1-3 gene in apical buds of corms stored at 25°C and 842 8 °C during floral induction time (C) Corm tissue specific expression of TFL1-3 during 843 floral induction. (D) Tissue Specific Expression of CsatTFL1-3. (E)Phenotypic analysis 844 of Arabidopsis Col -0 plants ectopically expressing TFL1 -3/CEN1, Hyper vegetive 845 phase phenotype observed in CsatTFL1-3 transgenic lines of Arabidopsis. (F and G) 846 Rosette leaves number and number of days to flower of 35S::TFL1 -3 transgenic and 847 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint control plants respectively. Values are means ± SEM (n = 10). Asterisks indicate a 848 significant difference between the transgenic Arabidopsis lines and wild -type plants 849 (Col-0; Student’s t-test, P < 0.05). (H) Virus-Induced Gene Silencing (VIGS) of TFL1-850 3/CEN1 in low-flower-competent saffron corms. The scale bar represents 2 cm. (I) The 851 expression of CsatTFL1-3 in CsatTFL1-3-TRV2 silenced lines and TRV2 lines (n=3). 852 (J) Comparison of flowering percentages in VIGS -silenced versus control saffron 853 corms. The data presented includes the mean percentage of flowering of three 854 technical repeats (n=3) and for each replicate, 10 corms were used. All data in this 855 figure are shown as mean ± SEM. Statistical Significance for Data in panel A,C & D 856 are analyzed using one -way ANOVA followed by Tukey’s Honest Significant 857 Difference (HSD) test for multiple comparisons. Different letters indicate statistically 858 significant differences between groups at p < 0.05, according to Compact Letter 859 Display (CLD). Data in panel B are analyzed using two -way ANOVA to assess the 860 effects of temperature, month, and their interaction, followed by Sidak’s multiple 861 comparisons test. Asterisks indicate significant differences between 8°C and 25°C 862 within each month (**** p < 0.0001), , F,G,I, & J are analyzed using the Student’s t -863 test (*: p<0.05; **: p<0.01). 864 865 Figure 4. Interaction analysis of CsatFT3, CsatTFL1 -3, and CsatFD2. 866 (A, B) Yeast two -hybrid (Y2H) assays showing physical interactions between 867 CsatFT3/CsatTFL1-3 and CsatFD2. The coding sequences of CsatFT3 and 868 CsatTFL1-3 were fused to the activation domain in the pGADT7 vector, and CsatFD2 869 was fused to the DNA -binding domain in the pGBKT7 vector. Yeast transformants 870 were grown on double dropout (DDO) medium lacking tryptophan and leucine 871 (selection for transformation) and quadruple dropout (QDO) medium lacking 872 tryptophan, leucine, histidine, and adenine (selection for interaction). pGBKT7-53 and 873 pGADT7-T were used as positive control; pGBKT7 -lam and pGADT7 -T interactions 874 served as the negative control. (C-D) Bimolecular fluorescence complementation 875 (BiFC) assay showing nuclear localization of the interactions between CsatFT3 or 876 CsatTFL1-3 with CsatFD2 in Nicotiana benthamiana leaf epidermal cells. CsatFD2 877 was fused to the C -terminal half of YFP (cYFP), and CsatFT3 or CsatTFL1 -3 were 878 fused to the N -terminal half of YFP (nYFP). YFP signal indicates interaction. Bars = 879 20 μm. (E) Competitive luciferase complementation assay demonstrating the 880 competition between CsatFT3 and CsatTFL1 -3 for binding to CsatFD2 in N. 881 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint benthamiana leaves. CsatTFL1 was fused to the N-terminal half of luciferase (nLUC), 882 and CsatFD2 to the C -terminal half (cLUC). Increasing amounts of CsatFT3 (non -883 tagged, driven by 35S promoter, in SK vector) were co -infiltrated at different ratios. 884 Images represents three independent replications. (F) Relative luminescence 885 intensities indicate competition for FD2 binding with TFL1 -3 and FT3. Data 886 represented is mean of three replicates. Statistical Significance are analyzed using 887 one-way ANOVA followed by Tukey’s Honest Significant Difference (HSD) test for 888 multiple comparisons. Different letters indicate statistically significant differences 889 between groups at p < 0.05, according to Compact Letter Display (CLD). 890 891 Figure 5: Functional characterization of CsatSVP1 and CsatSVP2 in 892 temperature-dependent floral regulation (A) Expression profiling of CsatSVP1 and 893 CsatSVP2 in apical bud during different vegetative and flowering stages. Data are 894 means ± SE from 3 replications. (B and C) Expression of SVP1 and SVP2 genes in 895 corms stored at inductive (25°C) and non -inductive (8°C) temperatures respectively. 896 (D) Flowering phenotypes of Arabidopsis lines ectopically expressing CsatSVP1. (E 897 and F) Rosette leaves number and number of days to flower of CsatSVP1 transgenic 898 and control plants respectively (G) Phenotypic changes of the floral organs in 899 35S::CsatSVP2 transgenic Arabidopsis plants. (H-I) Rosette leaves number and 900 number of days to flower of CsatSVP2 transgenic and control plants respectively. The 901 total leaf number was calculated by combining total rosette leaves. (J-K) Silencing of 902 CsatSVP2 by VIGS rescues flowering under low -temperature conditions in saffron. 903 The scale bar represents 2 cm. (L-M) qRT-PCR validation of CsatSVP2 knockdown in 904 saffron corm apices after VIGS stored at different temperatures (n=3). Error bars are 905 shown as SD of three technical repeats. (N-O) Flowering percentage in TRV2 control 906 and CsatSVP2 VIGS-silenced saffron corms stored at low (8°C) and ambient (23°C) 907 temperatures. The data presented includes the mean percentage of flowering, with 908 error bars representing the standard deviation (SD) of three technical repeats (n=3) 909 and for each replicate, 10 corms were used. (P-Q) Relative expression of CsatFT3 in 910 SVP2-silenced versus control corms under different temperature regimes. All data in 911 this figure are shown as mean ± SEM. Statistical Significance for data in panel A are 912 analyzed using two-way ANOVA followed by Tukey’s Honest Significant Difference 913 (HSD) test for multiple comparisons. Different letters indicate statistically significant 914 differences between groups at p < 0.05, according to Compact Letter Display (CLD). 915 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Data in panel B & C are analyzed using two -way ANOVA to assess the effects of 916 temperature, month, and their interaction, followed by Sidak’s multiple comparisons 917 test. Asterisks indicate significant differences between 8°C and 25°C within each 918 month (**** p < 0.0001), E, F, H, I, L, M, N, O, P & Q are analyzed using the student’s 919 t-test (*: p<0.05; **: p<0.01). 920 921 Figure 6. CsatSVP2 directly binds to the FT3 promoter via the CArG motif and 922 represses its transcription (A) Schematic diagram of the 1.2 kb CsatFT3 promoter 923 showing the location of a conserved CArG -box motif (CCATTTAAGG) approximately 924 827 bp upstream of the ATG start codon. (B) Yeast one -hybrid assay showing 925 interaction between CsatSVP2 and the FT3 promoter region containing the CArG 926 motif. Growth on selective media confirms binding specificity. (C) Schematic diagram 927 of the effectors and reporters used in dual -luciferase reporter. (D) Schematic of the 928 dual-luciferase constructs used for transient expression in Nicotiana benthamiana: FL 929 (Full length promoter) T1 (with CArG motif) and T2 (without). (E) Relative luciferase 930 activity in leaves co -expressing CsatSVP2 with the full promoter and T1 fragment 931 construct of CsatFT3 gene. (F) Relative luciferase activity in leaves co -expressing 932 CsatSVP2 and CsatFT3 full promoter and with T2 fragment constructs. Significant 933 repression is observed only in the T1 construct, indicating that CsatSVP2 represses 934 CsarFT3 transcription through the CArG motif. (G and H) Quantification of luciferase 935 activity. Bars represent mean ± SEM (n = 3). All data in figure are shown as mean ± 936 SEM. Statistical Significance for Data in panel F & H are are analyzed using one-way 937 ANOVA followed by Tukey’s Honest Significant Difference (HSD) test for multiple 938 comparisons. Different letters indicate statistically significant differences between 939 groups at p < 0.05, according to Compact Letter Display (CLD) 940 941 Figure 7. Proposed model of temperature -mediated flowering regulation in 942 saffron. This schematic model illustrates the temperature -regulated floral induction 943 pathway in saffron, emphasizing the key roles of CsatFT3, CsatFD2, TFL1 -3/CEN1, 944 and CsatSVP2 in transitioning from vegetative to reproductive growth. At low 945 temperatures (e.g., below 10°C), CsatFT3 expression remains low, preventing the 946 formation of the Floral Activator Complex (FAC). TFL1 -3/CEN1 binds to CsatFD2, 947 maintaining the shoot apical meristem in a vegetative state and inhibiting flowering. 948 CsatFD2 expression is minimal, suppressing floral induction. At low temperatures, 949 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint CsatSVP2 is upregulated and directly binds to the FT3 promoter at a conserved CArG 950 motif, repressing its expression and preventing FAC formation. As temperatures rise 951 above the threshold (e.g., 2 5°C), CsatFT3 expression increases, leading to the 952 formation of the FAC by binding to CsatFD2. This complex activates the expression of 953 floral meristem identity genes and promotes floral development. At the same time, 954 TFL1-3/CEN1 expression decreases, allowing the activation of CsatFD2 and the 955 initiation of flowering. 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Figure 1 984 985 986 987 988 989 990 991 992 993 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Figure 2 994 995 996 997 998 999 1000 1001 1002 1003 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Figure 3 1004 1005 1006 1007 1008 1009 1010 1011 1012 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Figure 4 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Figure 5 1024 1025 1026 1027 1028 1029 1030 1031 1032 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Figure 6 1033 1034 1035 1036 1037 1038 1039 1040 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Figure 7 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint

1055 Abe, M., Kobayashi, Y., Yamamoto, S., Daimon, Y., Yamaguchi, A., Ikeda, Y., Ichinoki, H., Notaguchi, M., 1056 Goto, K. and Araki, T. (2005) FD, a bZIP protein mediating signals from the floral pathway 1057 integrator FT at the shoot apex. Science 309, 1052-1056. 1058 Ahrazem, O., Rubio-Moraga, A., Nebauer, S.G., Molina, R.V. and Gomez-Gomez, L. (2015) Saffron: Its 1059 Phytochemistry, Developmental Processes, and Biotechnological Prospects. J Agric Food Chem 1060 63, 8751-8764. 1061 Andrés, F. and Coupland, G. (2012) The genetic basis of flowering responses to seasonal cues. Nature 1062 Reviews Genetics 13, 627-639. 1063 Balasubramanian, S., Sureshkumar, S., Lempe, J. and Weigel, D. (2006) Potent induction of Arabidopsis 1064 thaliana flowering by elevated growth temperature. PLoS genetics 2, e106. 1065 Bellinazzo, F., Manders, I., Heidemann, B., Bolanos, M.A., Stouten, E., Busscher, J., Abarca, D., van der 1066 Wal, F., Dornelas, M.C. and Angenent, G.C. (2025) Differential growth and flowering capacity 1067 of tulip bulbs and the potential involvement of PHOSPHATIDYLETHANOLAMINE -BINDING 1068 PROTEINS (PEBPs). Biology Direct 20, 29. 1069 Blázquez, M.A., Trénor, M. and Weigel, D. (2002) Independent control of gibberellin biosynthesis and 1070 flowering time by the circadian clock in Arabidopsis. Plant Physiology 130, 1770-1775. 1071 Bowman, J.L., Smyth, D.R. and Meyerowitz, E.M. (2012) The ABC model of flower development: then 1072 and now. Development 139, 4095-4098. 1073 Capovilla, G., Schmid, M. and Posé, D. (2015) Control of flowering by ambient temperature. Journal of 1074 Experimental Botany 66, 59-69. 1075 Cardone, L., Castronuovo, D., Perniola, M., Cicco, N. and Candido, V. (2020) Saffron (Crocus sativus L.), 1076 the king of spices: An overview. Scientia Horticulturae 272, 109560. 1077 Freytes, S.N., Canelo, M. and Cerdán, P.D. (2021) Regulation of flowering time: when and where? 1078 Current Opinion in Plant Biology 63, 102049. 1079 Gu, X., Le, C., Wang, Y., Li, Z., Jiang, D., Wang, Y. and He, Y. (2013) Arabidopsis FLC clade members 1080 form flowering -repressor complexes coordinating responses to endogenous and 1081 environmental cues. Nature Communications 4, 1947. 1082 Heide, O.M. (1977) Photoperiod and temperature interactions in growth and flowering of strawberry. 1083 Physiologia Plantarum 40, 21-26. 1084 Hsu, C.-Y., Adams, J.P., Kim, H., No, K., Ma, C., Strauss, S.H., Drnevich, J., Vandervelde, L., Ellis, J.D. and 1085 Rice, B.M. (2011) FLOWERING LOCUS T duplication coordinates reproductive and vegetative 1086 growth in perennial poplar. Proceedings of the National Academy of Sciences 108, 10756 -1087 10761. 1088 Hu, J., Liu, Y., Tang, X., Rao, H., Ren, C., Chen, J., Wu, Q., Jiang, Y., Geng, F. and Pei, J. (2020) 1089 Transcriptome profiling of the flowering transition in saffron (Crocus sativus L.). Scientific 1090 Reports 10, 9680. 1091 Jefferson, R.A., Kavanagh, T.A. and Bevan, M.W. (1987) GUS fusions: beta-glucuronidase as a sensitive 1092 and versatile gene fusion marker in higher plants. Embo j 6, 3901-3907. 1093 Jin, S. and Ahn, J.H. (2021) Regulation of flowering time by ambient temperature: repressing the 1094 repressors and activating the activators. New Phytologist 230, 938-942. 1095 Jin, S., Nasim, Z., Susila, H. and Ahn, J.H. (2021) Evolution and functional diversification of FLOWERING 1096 LOCUS T/TERMINAL FLOWER 1 family genes in plants. Seminars in Cell & Developmental 1097 Biology 109, 20-30. 1098 Jing, S., Jiang, P., Sun, X., Yu, L., Wang, E., Qin, J., Zhang, F., Prat, S. and Song, B. (2023) Long-distance 1099 control of potato storage organ formation by SELF PRUNING 3D and FLOWERING LOCUS T-like 1100 1. Plant communications 4. 1101 Jose-Santhi, J., Sheikh, F.R., Kalia, D. and Singh, R.K. (2023) Sugar metabolism mediates temperature-1102 dependent flowering induction in saffron (Crocus sativus L.). Environmental and Experimental 1103 Botany 206, 105150. 1104 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Kalia, D., Jose-Santhi, J., Kumar, R. and Singh, R.K. (2022) Analysis of PEBP Genes in Saffron Identifies 1105 a Flowering Locus T Homologue Involved in Flowering Regulation. Journal of Plant Growth 1106 Regulation, 1-20. 1107 Kalia, D., Jose-Santhi, J., Kumar, R. and Singh, R.K. (2023) Analysis of PEBP genes in saffron identifies a 1108 flowering locus T homologue involved in flowering regulation. Journal of Plant Growth 1109 Regulation 42, 2486-2505. 1110 Kalia, D., Jose-Santhi, J., Sheikh, F.R., Singh, D. and Singh, R.K. (2024) Tobacco rattle virus-based virus-1111 induced gene silencing (VIGS) as an aid for functional genomics in Saffron (Crocus sativus L.). 1112 Physiol Mol Biol Plants 30, 749-755. 1113 Khodorova, N.V. and Boitel-Conti, M. (2013) The role of temperature in the growth and flowering of 1114 geophytes. Plants 2, 699-711. 1115 Khosa, J., Bellinazzo, F., Kamenetsky Goldstein, R., Macknight, R. and Immink, R.G.H. (2021) 1116 PHOSPHATIDYLETHANOLAMINE-BINDING PROTEINS: the conductors of dual reproduction in 1117 plants with vegetative storage organs. Journal of Experimental Botany 72, 2845-2856. 1118 Kinoshita, A. and Richter, R. (2020) Genetic and molecular basis of floral induction in Arabidopsis 1119 thaliana. Journal of Experimental Botany 71, 2490-2504. 1120 Lee, C., Kim, S.J., Jin, S., Susila, H., Youn, G., Nasim, Z., Alavilli, H., Chung, K.S., Yoo, S.J. and Ahn, J.H. 1121 (2019) Genetic interactions reveal the antagonistic roles of FT/TSF and TFL1 in the 1122 determination of inflorescence meristem identity in Arabidopsis. The Plant Journal 99, 452-1123 464. 1124 Lee, J.H., Ryu, H. -S., Chung, K.S., Posé, D., Kim, S., Schmid, M. and Ahn, J.H. (2013a) Regulation of 1125 Temperature-Responsive Flowering by MADS -Box Transcription Factor Repressors. Science 1126 342, 628-632. 1127 Lee, J.H., Yoo, S.J., Park, S.H., Hwang, I., Lee, J.S. and Ahn, J.H. (2007) Role of SVP in the control of 1128 flowering time by ambient temperature in Arabidopsis. Genes & development 21, 397-402. 1129 Lee, R., Baldwin, S., Kenel, F., McCallum, J. and Macknight, R. (2013b) FLOWERING LOCUS T genes 1130 control onion bulb formation and flowering. Nat Commun 4, 2884. 1131 Leeggangers, H.A., Nijveen, H., Bigas, J.N., Hilhorst, H.W. and Immink, R.G. (2017a) Molecular 1132 regulation of temperature-dependent floral induction in Tulipa gesneriana. Plant physiology 1133 173, 1904-1919. 1134 Leeggangers, H.A., Nijveen, H., Bigas, J.N., Hilhorst, H.W. and Immink, R.G. (2017b) Molecular 1135 Regulation of Temperature -Dependent Floral Induction in Tulipa gesneriana. Plant Physiol 1136 173, 1904-1919. 1137 Li, X.-F., Wu, W.-T., Zhang, X. -P., Qiu, Y., Zhang, W., Li, R., Xu, J., Sun, Y., Wang, Y. and Xu, L. (2015) 1138 Narcissus tazetta SVP-like gene NSVP1 affects flower development in Arabidopsis. Journal of 1139 Plant Physiology 173, 89-96. 1140 Lu, J., Qu, L., Xing, G., Liu, Z., Lu, X. and Han, X. (2023) Genome -Wide Identification and Expression 1141 Analysis of the MADS Gene Family in Tulips (Tulipa gesneriana). Genes (Basel) 14. 1142 Molina, R., Valero, M., Navarro, Y., Guardiola, J. and Garcia -Luis, A. (2005a) Temperature effects on 1143 flower formation in saffron (Crocus sativus L.). Scientia horticulturae 103, 361-379. 1144 Molina, R.V., M., V., Y., N., A., G.-L. and and Guardiola, J.L. (2005b) Low temperature storage of corms 1145 extends the flowering season of saffron (Crocus sativus L.). The Journal of Horticultural Science 1146 and Biotechnology 80, 319-326. 1147 Nakano, Y., Higuchi, Y., Sumitomo, K. and Hisamatsu, T. (2013) Flowering retardation by high 1148 temperature in chrysanthemums: involvement of FLOWERING LOCUS T -like 3 gene 1149 repression. Journal of Experimental Botany 64, 909-920. 1150 Navarro, C., Abelenda, J.A., Cruz-Oró, E., Cuéllar, C.A., Tamaki, S., Silva, J., Shimamoto, K. and Prat, S. 1151 (2011) Control of flowering and storage organ formation in potato by FLOWERING LOCUS T. 1152 Nature 478, 119-122. 1153 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Noy-Porat, T., Cohen, D., Mathew, D., Eshel, A., Kamenetsky, R. and Flaishman, M.A. (2013) Turned on 1154 by heat: differential expression of FT and LFY -like genes in Narcissus tazetta during floral 1155 transition. Journal of experimental botany 64, 3273-3284. 1156 Oda, A., Narumi, T., Li, T., Kando, T., Higuchi, Y., Sumitomo, K., Fukai, S. and Hisamatsu, T. (2012) 1157 CsFTL3, a chrysanthemum FLOWERING LOCUS T-like gene, is a key regulator of photoperiodic 1158 flowering in chrysanthemums. Journal of Experimental Botany 63, 1461-1477. 1159 Pin, P.A., Benlloch, R., Bonnet, D., Wremerth-Weich, E., Kraft, T., Gielen, J.J. and Nilsson, O. (2010) An 1160 antagonistic pair of FT homologs mediates the control of flowering time in sugar beet. Science 1161 330, 1397-1400. 1162 Putterill, J., Laurie, R. and Macknight, R. (2004) It's time to flower: the genetic control of flowering 1163 time. BioEssays 26, 363-373. 1164 Pyo, Y., Park, S., Xi, Y. and Sung, S. (2014) Regulation of flowering by vernalisation in Arabidopsis. In: 1165 Advances in botanical research pp. 29-61. Elsevier. 1166 Renau-Morata, B., Nebauer, S.G., García -Carpintero, V., Cañizares, J., Gómez Minguet, E., de los 1167 Mozos, M. and Molina, R.V. (2021) Flower induction and development in saffron: Timing and 1168 hormone signalling pathways. Industrial Crops and Products 164, 113370. 1169 Simon, R., Igeño, M.I. and Coupland, G. (1996) Activation of floral meristem identity genes in 1170 Arabidopsis. Nature 384, 59-62. 1171 Singh, D., Sharma, S., Jose-Santhi, J., Kalia, D. and Singh, R.K. (2023) Hormones regulate the flowering 1172 process in saffron differently depending on the developmental stage. Front Plant Sci 14, 1173 1107172. 1174 Srikanth, A. and Schmid, M. (2011) Regulation of flowering time: all roads lead to Rome. Cellular and 1175 molecular life sciences 68, 2013-2037. 1176 Strasser, B., Alvarez, M.J., Califano, A. and Cerdán, P.D. (2009) A complementary role for ELF3 and TFL1 1177 in the regulation of flowering time by ambient temperature. The Plant Journal 58, 629-640. 1178 Sureshkumar, S., Dent, C., Seleznev, A., Tasset, C. and Balasubramanian, S. (2016) Nonsense-mediated 1179 mRNA decay modulates FLM -dependent thermosensory flowering response in Arabidopsis. 1180 Nat Plants 2, 16055. 1181 Taoka, K.-i., Ohki, I., Tsuji, H., Kojima, C. and Shimamoto, K. (2013) Structure and function of florigen 1182 and the receptor complex. Trends in plant science 18, 287-294. 1183 Tsaftaris, A., Pasentsis, K. and Argiriou, A. (2013) Cloning and Characterization of FLOWERING LOCUS 1184 T-Like Genes from the Perennial Geophyte Saffron Crocus (Crocus sativus). Plant Mol. Biol. 1185 Report. 31, 1558–1568. 1186 Tsuji, H., Nakamura, H., Taoka, K. and Shimamoto, K. (2013) Functional diversification of FD 1187 transcription factors in rice, components of florigen activation complexes. Plant Cell Physiol 1188 54, 385-397. 1189 Turck, F., Fornara, F. and Coupland, G. (2008) Regulation and identity of florigen: FLOWERING LOCUS 1190 T moves center stage. Annu. Rev. Plant Biol. 59, 573-594. 1191 Tylewicz, S., Tsuji, H., Miskolczi, P., Petterle, A., Azeez, A., Jonsson, K., Shimamoto, K. and Bhalerao, 1192 R.P. (2015) Dual role of tree florigen activation complex component FD in photoperiodic 1193 growth control and adaptive response pathways. Proceedings of the National Academy of 1194 Sciences 112, 3140-3145. 1195 Wang, Z., Li, X., Xu, J., Yang, Z. and Zhang, Y. (2021) Effects of ambient temperature on flower initiation 1196 and flowering in saffron (Crocus sativus L.). Scientia Horticulturae 279, 109859. 1197 Wickland, D.P. and Hanzawa, Y. (2015) The FLOWERING LOCUS T/TERMINAL FLOWER 1 gene family: 1198 functional evolution and molecular mechanisms. Molecular plant 8, 983-997. 1199 Wigge, Philip A. (2011) FT, A Mobile Developmental Signal in Plants. Current Biology 21, R374-R378. 1200 Wigge, P.A., Kim, M.C., Jaeger, K.E., Busch, W., Schmid, M., Lohmann, J.U. and Weigel, D. (2005) 1201 Integration of spatial and temporal information during floral induction in Arabidopsis. Science 1202 309, 1056-1059. 1203 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint Wu, X., Ling, W., Pan, Y., Yang, Z., Ma, J., Yang, Y., Xiang, W., Zhou, L., Sun, M., Chen, J., Chen, H., Zheng, 1204 S., Zeng, J. and Li, Y. (2024) Functional analysis of a lily SHORT VEGETATIVE PHASE ortholog in 1205 flowering transition and floral development. Plant Physiology and Biochemistry 206, 108287. 1206 Wu, Y.-M., Ma, Y.-J., Wang, M., Zhou, H., Gan, Z.-M., Zeng, R.-F., Ye, L.-X., Zhou, J.-J., Zhang, J.-Z. and 1207 Hu, C.-G. (2022) Mobility of FLOWERING LOCUS T protein as a systemic signal in trifoliate 1208 orange and its low accumulation in grafted juvenile scions. Horticulture Research 9. 1209 Xue, W., Shi, J., Li, Z., Zhang, Y., Tang, M., Du, X., An, Z., Chen, H., Yang, D. and Li, X. (2025) CmFDa‐1210 mediated epigenetic regulation of flowering in chrysanthemum. Plant Biotechnology Journal. 1211 Yan, X., Cao, Q. -Z., He, H. -B., Wang, L. -J. and Jia, G. -X. (2021) Functional analysis and expression 1212 patterns of members of the FLOWERING LOCUS T (FT) gene family in Lilium. Plant Physiology 1213 and Biochemistry 163, 250-260. 1214 Zhu, Y., Klasfeld, S., Jeong, C.W., Jin, R., Goto, K., Yamaguchi, N. and Wagner, D. (2020) TERMINAL 1215 FLOWER 1 -FD complex target genes and competition with FLOWERING LOCUS T. Nature 1216 communications 11, 5118. 1217 Zik, M. and Irish, V.F. (2003) Global identification of target genes regulated by APETALA3 and 1218 PISTILLATA floral homeotic gene action. The Plant Cell 15, 207-222. 1219 1220 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted July 2, 2025. ; https://doi.org/10.1101/2025.07.01.662559doi: bioRxiv preprint