- Top of page
- Materials and Methods
- Supporting Information
Plants have evolved mechanisms tointegrate environmental and developmental cues and precisely control the timing of vegetative and reproductive growth. In Arabidopsis thaliana (Arabidopsis), flowering is initiated by multiple pathways (Amasino, 2010) that converge on a small number of integrator genes, including FLOWERING LOCUS T (FT; Kardailsky et al., 1999; Kobayashi et al., 1999), which encodes a major flowering hormone ‘florigen’ (Chailakhyan, 1968; Zeevaart, 2008). The current model has FT protein transported in the phloem from leaves to the shoot apex (Turck et al., 2008), where it interacts with the bZIP transcription factor FLOWERING LOCUS D (FD), to activate floral meristem identity genes APETALA1 (AP1) and LEAFY (LFY; Abe et al., 2005; Wigge et al., 2005) and SUPPRESSOR OF OVER EXPRESSION OF CONSTANS1 (SOC1; Searle et al., 2006). An endoplasmic reticulum membrane protein, FT-INTERACTING PROTEIN 1 (FTIP1), facilitates FT protein transport (Liu et al., 2012) and 14-3-3 proteins act as intracellular receptors for FT (Taoka et al., 2011). FT-like genes are universally conserved in flowering plants and were demonstrated to perform a role of florigen in plants other than Arabidopsis, including tomato (Lifschitz et al., 2006), squash (Lin et al., 2007) and rice (Tamaki et al., 2007; Komiya et al., 2009; Taoka et al., 2011).
The FT protein is a member of a family of phosphatidylethanolamine-binding proteins (PEBP), initially identified in animals as Raf-1 kinase inhibitors (Yeung et al., 1999). In angiosperms, this family consists of three phylogenetically distinct groups: the FT-like proteins, the TERMINAL FLOWER1 (TFL1)-like proteins (Bradley et al., 1997; Ohshima et al., 1997), and the MOTHER OF FT AND TFL1 (MFT)-like proteins (Mimida et al., 2001; Yoo et al., 2004). The MFT-like genes are the likely basal clade found in angiosperms, gymnosperms, lycophytes and bryophytes, and the ancestor of FT/TFL1-like genes that have diverged in seed plants (Hedman et al., 2009; Karlgren et al., 2011). TFL1 is a key repressor of flowering, which maintains the inflorescence meristem by preventing expression of AP1 and LFY (Ratcliffe et al., 1998, 1999). Further duplication events gave rise to multiple genes within these groups. For example, Arabidopsis has two FT-like genes, FT and TWIN SISTER OF FT (TSF; Yamaguchi et al., 2005), while TFL1 is paralogous to ATC, the Arabidopsis thaliana CENTRORADIALIS (CEN) homologue. A divergent external loop confers antagonistic activity on FT and TFL1 and a single amino acid change is sufficient to convert TFL1 to an activator of flowering (Hanzawa et al., 2005; Ahn et al., 2006). Similarly, two FT homologs from sugar beet perform antagonistic functions, one promoting flowering and the other repressing flowering owing to a mutation in the protein external loop (Pin et al., 2010); antagonistic FT-like paralogs were also recently described in tobacco (Harig et al., 2012).
Studies in species with sympodial growth suggested a more general role for FT-like genes in systemic regulation of growth and termination of meristems (Lifschitz & Eshed, 2006; Lifschitz et al., 2006; Shalit et al., 2009), and implicated the tomato FT homolog SINGLE FLOWER TRUSS (SFT) in heterosis for fruit yield (Krieger et al., 2010).
Unlike annual plants, woody perennials undergo successive growing seasons with vegetative growth before transition to reproductive maturity, followed by cycles of coordinated vegetative and reproductive growth. In temperate regions, these cycles are interrupted by dormancy periods, which ensure survival in unfavourable conditions. Accumulation of chilling during dormancy is often necessary to ensure budbreak and flowering in spring. Studies on FT- and TFL1-like genes in Populus spp. began to shed light on possible mechanisms underlying the regulation of vegetative to reproductive transition, growth and dormancy cycles and the coexistence of vegetative and floral meristems on the same shoot (Böhlenius et al., 2006; Hsu et al., 2006, 2011; Mohamed et al., 2010; Rinne et al., 2011). Two FT genes, FT1 and FT2, have functionally diverged to regulate reproductive onset and vegetative growth, respectively (Hsu et al., 2011). CEN/TFL1 genes were implicated in regulation of maturity, axillary meristem identity and release from dormancy (Mohamed et al., 2010). Similarly, apple has two FT-like genes with differential expression patterns and potentially distinct roles (Kotoda et al., 2010) capable of interacting with transcription factors implicated in cell growth and leaf and fruit development (Mimida et al., 2011). Ectopic expression of apple FT and downregulation of TFL1 both accelerated flowering (Kotoda et al., 2006, 2010). Precocious flowering was also observed upon ectopic expression of a citrus FT homolog in trifoliate orange (Endo et al., 2005). In grape, based on expression irrespective of flowering, VvFT might have a role other than flowering control (Sreekantan & Thomas, 2006; Carmona et al., 2007). From the knowledge gathered to date, it seems that the FT/TFL1 module provides potential to fine-tune developmental regulatory mechanisms and opportunities for improvement of agriculturally important traits (Jung & Müller, 2009; Zhang et al., 2010; Yeoh et al., 2011; Iwata et al., 2012).
The aim of this study was to extend the analysis of FT/TFL1-like genes in woody perennial vines. Kiwifruit (Actinidia spp.) are perennial vines with horticultural importance and features of development specific to woody perennials, including a juvenile period before establishment of flowering competence (Ferguson, 1990), growth spread over two seasons (Brundell, 1975a,b; Walton et al., 1997) and a period of low temperatures (winter chilling) required to resume bud growth (Brundell, 1976). The most important commercial kiwifruit cultivars belong to Actinidia chinensis and Actinidia deliciosa, two closely related species with differences in winter-chilling requirements (Wall et al., 2008). It is unclear if photoperiod plays any role in kiwifruit flowering (Snelgar et al., 2007) and there is discrepancy in reports on the timing of floral commitment in latent shoot buds (Fabbri et al., 1992; Snelgar & Manson, 1992; Snowball, 1996; Walton et al., 1997, 2001). These buds are established in the first growing season and contain undifferentiated, dome-shaped axillary meristems with a potential to differentiate into flowers after the winter dormancy period (Walton et al., 1997). The timing of establishment of dome-shaped meristems (Walton et al., 1997), combined with expression of a LEAFY homolog (Walton et al., 2001) and low levels of accumulation of flower-specific MADS-box genes PISTILLATA and AGAMOUS (Varkonyi-Gasic et al., 2011), support the view that floral commitment occurs in the spring of the first year, followed by differentiation in the spring of the second year.
In this study we characterize a kiwifruit FT and a CEN gene by expression analysis and ectopic transgenic analysis. We demonstrate interaction with a FD-like bZIP transcription factor, which is functionally conserved in Arabidopsis and is downregulated in dormant buds in response to cold treatment when FT is induced in response to winter chilling. This work highlights the conservation and divergence in strategies perennial plants employ to regulate their flowering, growth and dormancy cycles and adds to better understanding of the range of FT/TFL1-like gene action in woody perennials.