FLOWERING LOCUS T (FT) encodes a member of the phosphatidylethanolamine-binding protein (PEBP) family that functions as the mobile floral signal, playing an important role in regulating the floral transition in angiosperms. We isolated an FT-homolog (GhFT1) from Gossypium hirsutum L. cultivar, Xinluzao 33 GhFT1 was predominantly expressed in stamens and sepals, and had a relatively higher expression level during the initiation stage of fiber development. GhFT1 mRNA displayed diurnal oscillations in both long-day and short-day condition, suggesting that the expression of this gene may be under the control of the circadian clock. Subcellular analysis revealed that GhFT1 protein located in the cytoplasm and nucleus. Ectopic expression of GhFT1 in transgenic arabidopsis plants resulted in early flowering compared with wild-type plants. In addition, ectopic expression of GhFT1 in arabidopsis ft-10 mutants partially rescued the extremely late flowering phenotype. Finally, several flowering related genes functioning downstream of AtFT were highly upregulated in the 35S::GhFT1 transgenic arabidopsis plants. In summary, GhFT1 is an FT-homologous gene in cotton that regulates flower transition similar to its orthologs in other plant species and thus it may be a candidate target for promoting early maturation in cotton breeding.

FLOWERING LOCUS T (FT) encodes a mobile signal protein, recognized as major component of florigen, which has a central position in regulating flowering, and also plays important roles in various physiological aspects. A mode is recently emerging for the balance of indeterminate and determinate growth, which is controlled by the ratio of FT-like and TERMINAL FLOWER 1 (TFL1)-like gene activities, and has a strong influence on the floral transition and plant architecture. Orthologs of GhFT1 was previously isolated and characterized from Gossypium hirsutum. We demonstrated that ectopic overexpression of GhFT1 in tobacco, other than promoting flowering, promoted lateral shoot outgrowth at the base, induced more axillary bud at the axillae of rosette leaves, altered leaf morphology, increased chlorophyll content, had higher rate of photosynthesis and caused flowers abscission. Analysis of gene expression suggested that flower identity genes were significantly upregulated in transgenic plants. Further analysis of tobacco FT paralogs indicated that NtFT4, acting as flower inducer, was upregulated, whereas NtFT2 and NtFT3 as flower inhibitors were upregulated in transgenic plants under long-day conditions, but downregulated under short-day conditions. Our data suggests that sufficient level of transgenic cotton FT might disturb the balance of the endogenous tobacco FT paralogs of inducers and repressors and resulted in altered phenotype in transgenic tobacco, emphasizing the expanding roles of FT in regulating shoot architecture by advancing determine growth. Manipulating the ratio for indeterminate and determinate growth factors throughout FT-like and TFL1-like gene activity holds promise to improve plant architecture and enhance crop yield.

transcripts ranged from 1000 and 2000 bp, and 12.6% transcripts were longer than 2000 bp (Fig. 2b). The number of transcripts annotated with seven public databases is summarised in Fig. 2c. A total of 311,224 (73.01%) and 227,392 (53.34%) transcripts received annotations from NCBI (https://www.ncbi.nlm.nih.gov/) non-redundant protein (NR) and Swiss-Prot databases (http://www.expasy.org/sprot/), respectively. Gene ontology (http:// www.geneontology.org), protein family (http://pfam.xfam.org/) and Kyoto Encyclopedia of Genes and Genomes (http://www.genome.jp/kegg) annotation analysis indicated that 201,698 (47.31%), 200,757 (47.09%) and 127,605 (29.93%) transcripts matched these databases, respectively. The genomic similarity analysis by BLASTx showed that the most abundant transcripts were annotated as Sesamum indicum, which accounted for 59.6%, followed by Erythranthe guttata (15.7%); only 1.7%, 1.7% and 1.5% sequences had hit against in Coffea canephora, Salvia miltiorrhiz and Vitis vinifera, respectively; 19.8% of annotated sequences were similar to other plant species (Fig. 2d). A box diagram of fragments per kilobase of transcript sequence per million base pairs (FPKM) indicated that gene expression levels (FPKM > 0.3) were not evenly distributed at the three developmental stages (Supplementary Fig. S1). There was an obvious increase at the Bud I and Bud II stages compared to the flowering stage 50%.