https://orcid.org/0000-0001-6045-5865
Abstract
FLOWERING LOCUS T (FT) promotes flowering by integrating six genetic pathways. In Arabidopsis, the FT protein is transported from leaves to shoot apices and induces flowering. However, contradictory conclusions about floral induction via graft-transmitted FT in trees were reported in previous studies. We obtained extremely early-flowering transgenic woody Jatropha curcas L. by overexpression of J. curcas FT using Arabidopsis thaliana SUCROSE TRANSPORTER 2 (SUC2) promoter (SUC2:JcFT) and non-flowering transgenic J. curcas by RNA interference (RNAi), which were used to investigate the function of graft-transmitted JcFT in floral induction in woody perennials. Scions from five wild-type species of the Jatropha genus and from JcFT-RNAi transgenic J. curcas were grafted onto SUC2:JcFT rootstocks. Most grafted plants produced flowers in 1–2 months, and the flowering percentage and frequency of various grafted plants decreased with increasing scion length. Consistently, FT protein abundance in scions also decreased with increasing distance from graft junctions to the buds. These findings suggest that FT proteins can be transmitted by grafting and can induce the floral transition in woody perennials, and the efficiency of graft-transmitted JcFT for floral induction depends on the scion length, which may help explain previous seemingly contradictory observations regarding floral induction via graft-transmitted FT in trees.
Introduction
Flowering, which involves a developmental phase change from a vegetative state to a reproductive state, is of fundamental importance to the plant life cycle. In the model species Arabidopsis thaliana, six pathways, i.e., the photoperiod, vernalization, plant hormone, autonomous, ambient temperature and age pathways, coordinate flowering time with environmental inputs to optimize plant adaptation and reproductive success (Albani and Coupland 2010). These pathways converge via a small number of key genes recognized as floral integrator genes, such as FLOWERING LOCUS T (FT), SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1) and LEAFY (LFY), to initiate the early stages of flowering (Blázquez and Weigel 2000, Pena et al. 2001, Kobayashi and Weigel 2007, Xu et al. 2012, Khan et al. 2014). Under long days, FT is rhythmically activated in leaves by CONSTANS (CO), a central activator involved in the photoperiod pathway (Kardailsky et al. 1999, An et al. 2004, Valverde et al. 2004). FT proteins antagonize TERMINAL FLOWER 1 (TFL1) proteins during inflorescence development by competition for complex formation with 14-3-3 and Flowering Locus D (FD) proteins (Turck et al. 2008, Kaneko-Suzuki et al. 2018, Zhu et al. 2021). Recently, a bHLH transcription factor MYC3 was found to repress flowering by antagonizing CO to regulate FT expression (Bao et al. 2019).
The florigen FTs are small, globular and phosphatidylethanolamine binding proteins (PEBPs), which play a crucial role in the transition from vegetative growth to flowering and the integration of flowering signals (Putterill and Varkonyi-Gasic 2016). In annuals, FT proteins have been identified as mobile flowering signals that are produced in the leaves and translocated long distances through the phloem to the shoot apex, triggering flower initiation, as shown in tomato and tobacco (Lifschitz et al. 2006), Arabidopsis (Huang et al. 2005, Corbesier et al. 2007, Notaguchi et al. 2008, Yoo et al. 2013a, 2013b), rice (Tamaki et al. 2007) and cucurbits (Lin et al. 2007, Turck et al. 2008, Turnbull 2011, Wigge 2011, Liu et al. 2013, Yoo et al. 2013a). Studies in Arabidopsis have shown that the long-distance movement of FT proteins from leaves to shoot apices is controlled by several regulators. The FT INTERACTING PROTEIN 1 (FTIP1), an endoplasmic reticulum-localized protein of the multiple C2 domains and transmembrane protein (MCTPs) family, is crucial for the selective transport of FTs from companion cells to sieve elements, thus affecting FT transport to the shoot apical meristem (Liu et al. 2012, 2013). Furthermore, MCTP-SNARE complex-mediated endosomal trafficking is necessary for FT protein export from phloem companion cells to induce flowering (Liu et al. 2019). This process is controlled in a temperature-dependent manner, and the transcriptional inhibition of FT is induced under low temperature to ensure the optimal flowering conditions of plants (Liu et al. 2020). In addition, SODIUM POTASSIUM ROOT DEFECTIVE 1 (NaKR1), a heavy metal-associated (HMA) domain-containing protein, plays an essential role in regulating the long-distance movement of FT. Loss of function of NaKR1 compromises FT transport to shoot apices via sieve elements, resulting in delayed flowering under long-day conditions (Zhu et al. 2016).
With respect to perennials, contradictory observations regarding floral induction via graft-transmitted FT in trees were reported in previous studies. Grafting studies of poplar and apple with rootstocks expressing FT transgenes under the control of a heat-shock inducible promoter GmHsp17.6-L (Zhang et al. 2010) or Gmhsp17.5-E (Wenzel et al. 2013) did not result in precocious flowering of the receptor scions (Zhang et al. 2010, Wenzel et al. 2013). Bull et al. (2017) demonstrated the overexpression of Arabidopsis FT under control of a strong constitutive cauliflower mosaic virus (CaMV) 35S RNA (35S) promoter to trigger early flowering in cassava (Manihot esculenta), but failed to induce precocious flowering in non-transgenic scions grafted onto the FT transgenic rootstocks. Similarly transgenic overexpression of endogenous MeFT1 driven by the cassava vein mosaic virus (CsVMV) promoter produced early flowering in cassava, but the MeFT1 transgenic rootstock was also unable to induce flowering in long-term observations of non-transgenic scions (Odipio et al. 2020). However, Ye et al. (2014) showed, by using a weak synthetic G10-90 promoter, the early-flowering trait of the JcFT-overexpressing transgenic Jatropha curcas was graft-transmissible. Song et al. (2019) also reported that transgenic 35S:VcFT rootstocks promoted the flowering of non-transgenic scions in blueberry. Recently, Soares et al. (2020) observed that transgenic citrange rootstocks expressing the Citrus clementina FT3 gene under the control of the A. thaliana phloem-specific SUCROSE TRANSPORTER 2 (SUC2) promoter induced precocious flowering in non-transgenic sweet orange scions. Although floral promotion via graft transmission of FT has not been well demonstrated in many tree species, FT-like genes have been successfully applied to reduce the long juvenile (pre-flowering) phase of several tree species, enabling rapid breeding (Putterill and Varkonyi-Gasic 2016, Sinn et al. 2021).
To further demonstrate the efficiency of graft-transmitted JcFT for floral induction in perennial plants, in this study, we employed the extremely early-flowering transgenic J. curcas overexpressing JcFT (SUC2:JcFT; Li et al. 2014) and the non-flowering transgenic J. curcas obtained by RNA interference (RNAi), together with five perennial woody species in genus Jatropha of the family Euphorbiaceae, including J. curcas, J. gossypifolia L., J. integerrima Jacq., J. multifida L. and J. Podagrica Hook., which are important sources of lipids and secondary metabolites (Sujatha et al. 2013, Cavalcante et al. 2020). Here, we demonstrate that JcFT proteins can be transmitted by grafting and promote flowering in five woody perennial species of the Jatropha. The abundance of JcFT in the buds of the scions decreased with increasing scion length, and thus the efficiency of graft-transmitted JcFT for floral induction depends on scion length. Our findings help explain the previous seemingly contradictory observations regarding floral induction via graft-transmitted FT in trees.
Materials and methods
Vector construction and J. curcas transformation
To construct the JcFT-RNAi expression vector, the sense and antisense fragments of JcFT were amplified using the primers XK220/221 with XhoI and KpnI restriction sites and XK222/223 with XbaI and BamHI restriction sites, respectively. The two fragments were then ligated to a pHANNIBAL vector (Smith et al. 2000, Wesley et al. 2001) in opposing orientations on either side of a PDK intron to produce a single self-complementary hairpin RNA (hpRNA) of JcFT driven by the 35S promoter, after which the expression cassette was cut by NotI and subsequently ligated into a pJL10 binary expression vector (Li et al. 2017). The expression vector was transformed to J. curcas via the Agrobacterium-mediated transformation method (Fu et al. 2015). The sequences of primers used were as follows: XK220 (GCTCGAGTTTTGGGCAAGAGATAGT), with an XhoI site; XK221 (AGGTACCAGTGTTGAAATTCTGACG), with a KpnI site; XK222 (GTCTAGATTTTGGGCAAGAGATAGT), with an XbaI site and XK223 (GATCGATAGTGTTGAAATTCTGACG), with a ClaI site.
Plant materials and growth conditions
Seeds of J. curcas, J. gossypifolia, J. integerrima, J. multifida and J. podagrica were collected during the summer from the Xishuangbanna Tropical Botanical Garden (XTBG; 21°54′ N, 101°46′ E, 580 m above sea level) of the Chinese Academy of Sciences located in Mengla County, Yunnan Province, Southwest China. All seeds were germinated, after which all the seedlings were grown in the same greenhouse (at 28 °C, 14 h of light and 10 h of darkness). SUC2:JcFT plants were obtained from our previous study (Li et al. 2014). Since the regenerated SUC2:JcFT shoots with flower buds did not produce roots in root induction media, we grafted the SUC2:JcFT shoots onto rootstocks of wild-type (WT) seedlings. Twenty-five lines (T0) of SUC2:JcFT transgenic plants were obtained from more than 100 grafted shoots. Only the SUC2:JcFT shoots with both flower buds and leaves survived and grew to maturity on rootstocks of WT plants. We obtained T1 transgenic seeds from SUC2:JcFT plants by bagging their parents and T1 transgenic seeds from JcFT-RNAi by grafting JcFT-RNAi scions onto SUC2:JcFT rootstocks. Mature transgenic J. curcas plants were transplanted into the field within the XTBG and grown from 2012 to 2021. All tissues for real-time quantitative reverse transcription polymerase chain reaction (qRT-PCR) and western blots were immediately frozen in liquid N2 and stored at −80 °C until use. The phenotypes of the grafted plants were analyzed. For each Jatropha genotype, more than 25 plants were used for characterization.
Plant grafting
Two-month-old JcFT-RNAi J. curcas plants and five species of wild-type (WT) Jatropha (J. curcas, J. gossypifolia, J. integerrima, J. multifida and J. podagrica) plants were used as scions. Two-month-old WT plants and SUC2:JcFT J. curcas plants were used as rootstocks. Stem segments (5 cm in length) were grafted onto WT and SUC2:JcFT rootstocks via cleft grafting with a double-edged razor blade. The wound was wrapped up in wrapping film, which was subsequently removed 10 days later. The grafted plants were grown under low-light intensity conditions (<100 μmol m−2 s−1) for the first 10 days, after which they were grown under high-light intensity conditions (325 μmol m−2 s−1).
Overexpression of JcFT promotes flowering in transgenic J. curcas. (A–C) Wild-type (WT) regenerated shoot in vitro (A), WT shoot grafted onto WT rootstock grown for 40 days (B) and 15-day-old seedling (C); (D–F): SUC2:JcFT transgenicshoot (D), SUC2:JcFT shoot grafted onto WT rootstock grown for 40 days (E) and 15-day-old T1 SUC2:JcFT transgenic seedling (F). The red arrow indicates the first inflorescence. Bar =1 cm.
qRT-PCR analysis
Total RNA was extracted from frozen Jatropha tissues as described by Ding et al. (2008). First-strand cDNA was synthesized from 1 μg of total RNA using a PrimeScript® RT Reagent Kit together with gDNA Eraser (Takara, Dalian, China). The cDNA templates were diluted five times with first-strand cDNA using sterilized double-distilled water; qRT-PCR was performed using SYBR® Premix Ex Taq™ II (Takara) on a Roche 480 Real-Time PCR Detection System (Roche Diagnostics, Indianapolis, IN, USA). The primers used for qRT-PCR are listed in Table S1 available as Supplementary data at Tree Physiology Online. qRT-PCR was performed on three independent biological replicates, with three technical replicates per sample. The data were analyzed via the 2−ΔΔCT method as described by Livak and Schmittgen (2001). The transcript levels of specific genes were normalized to those of the Jatropha Actin1 gene.
Flowering time statistics
The flowering time of the grafted plants grown in the greenhouse was analyzed, and the time from grafting to the first visible inflorescence bud was recorded. The percentage of flowering grafted plants was calculated on the basis of the number of flowering plants and non-flowering plants. The number of inflorescences produced by each grafted plant was recorded and analyzed, and the distance from the junction to the inflorescence buds was also measured. Statistical Package for the Social Science (SPSS) software was used for the data analysis, and SigmaPlot 11.0 software was used for generating graphs.
JcFT-RNAi transgenic J. curcas did not produce flowers. (A) Ten-month-old WT J. curcas plant that produced flowers. (B, C) Inflorescences produced from WT plants in (A). (D) Eight-year-old T0 JcFT-RNAi transgenic J. curcas plant that did not produce flowers. (E) Comparison of flowering time among WT, SUC2:JcFT and JcFT-RNAi plants. The arrows at the top of bar indicate the plants that have not flowered. (F) Analysis of JcFT expression levels in WT, SUC2:JcFT and JcFT-RNAi plants by qRT-PCR. RNA was extracted from mature leaves, and the transcript levels were normalized using the JcACTIN1 gene as a reference. (G) Western blot analysis of JcFT protein abundance in WT, SUC2:JcFT and JcFT-RNAi plants. The values are the means ± standard deviations of three plants per line (three independent experiments). The protein level in the WT was set as the standard, with a value of 1. Scale bars = 50 cm.
Protein extraction
The shoots (10 cm) from plants of which scions from J. gossypifolia, J. integerrima, J. multifida or J. podagrica were grafted onto WT and SUC2:JcFT seedlings grown in an artificial climate chamber were collected to isolate their total protein. Apical and lateral buds from different scion positions were collected and flash frozen in liquid nitrogen. After grinding, the powder was dissolved in 1 ml of extraction buffer to extract the proteins. The total soluble proteins were isolated using a Plant Total Protein Extraction Kit (No. C500053, Sangon Biotech, Shanghai, China) following the manufacturer’s instructions. The protein concentration was subsequently measured using a Lowry Protein Assay Kit (No. C504031, Sangon Biotech) according to the manufacturer’s instructions.
SDS-PAGE and western blot analysis
Two hundred micrograms of total protein extracted from Jatropha buds and 5x Laemmli buffer (50 mM Tris, 1% SDS, 0.05% bromophenol blue and 10% glycerol [pH 6.8]) were mixed together and incubated at 100 °C for 5 min. The samples were then resolved on 5% spacer and 15% separation discontinuous polyacrylamide SDS gels. Electrophoresis was subsequently conducted at 60 V for 30 min and, followed by 120 V for 60 min. The proteins were transferred onto NC membranes (No. F619511, BBI, Shanghai, China) with a wet electroblotting device (Bio-Rad, Hercules, CA, USA). The transfer conditions included a constant 55 V for 2 h. After transfer of the proteins, the membrane was saturated with 5% skim milk powder (Merck, Darmstadt, Germany) in Tris-buffered saline containing Tween-20 (TBST; 10 mM Tris, 150 mM NaCl, 0.2% Tween-20 [pH 7.5]). Incubation with the primary antibody (polyclonal rabbit anti-FT/Actin antibody, PytoAB, San Jose, CA 95131, USA) was performed overnight at 4 °C; after washing three times in TBST, incubation with the secondary antibody (goat anti-rabbit IgG H&L [HRP]; Dianova, Hamburg, Germany) in TBST was performed at room temperature for 60 min. Bound antibodies were detected with an Immobilon Western Chemiluminescent HRP Detection Kit (Millipore Corporation, Billerica, MA, USA) in conjunction with a Gel Doc™ XR+ system (Bio-Rad).
JcFT accelerates J. gossypifolia flowering via grafting. (A) Wild-type (WT) J. gossypifolia scions grafted onto WT and SUC2:JcFT J. curcas rootstocks. The triangles indicate the graft junctions. Scale bar = 10 cm. (B) The shoot apex of J. gossypifolia grafted onto WT J. curcas rootstock. Scale bar = 1 cm. (C, D) Flowers produced on the shoot of J. gossypifolia grafted onto SUC2:JcFT J. curcas rootstock. Scale bar = 1 cm. (E) Comparison of flowering plant percentage of J. gossypifolia scions grafted onto WT and SUC2:JcFT J. curcas rootstocks. (F) Comparison of flowering time of J. gossypifolia scions grafted onto WT and SUC2:JcFT J. curcas rootstocks. (G) Analysis of inflorescences generated on each grafted plant. (H) Analysis of the distance from the graft junction to the flower buds. (I) Western blot analysis of JcFT protein abundance in J. gossypifolia scions grafted onto WT and SUC2:JcFT rootstocks. The values are the means ± standard deviations of three plants per line (three independent experiments). The protein level in the WT was set as the standard, with a value of 1.00. JcFT (20 kDa) and JcActin (45 kDa) were quantified with AtFT and AtActin antibodies. The values are the means ± standard deviations of 15 plants per line (three independent experiments). The asterisks indicate significant differences in comparison with the WT at P < 0.05 according to Student’s t-test.
JcFT accelerates J. integerrima flowering via grafting. (A) Wild-type (WT) J. integerrima scions grafted onto WT (two plants on the left) and SUC2:JcFT (two plants on the right) J. curcas rootstocks. The triangles indicate the graft junctions. Scale bar = 10 cm. (B, C) The shoot apex of J. integerrima grafted onto WT J. curcas rootstocks. Scale bar = 1 cm. (D, E) Flowers produced on the shoot of J. integerrima grafted onto SUC2:JcFT J. curcas rootstocks. Scale bar = 1 cm. (F) Comparison of flowering plant percentage of J. integerrima scions grafted onto WT and SUC2:JcFT J. curcas rootstocks. (G) Comparison of flowering time of J. integerrima scions grafted onto WT and SUC2:JcFT J. curcas rootstocks. (H) Analysis of inflorescences generated on each grafted plant. (I) Analysis of the distance from the junction to the flower buds. (J) Western blot analysis of JcFT protein abundance in J. integerrima scions grafted onto WT and SUC2:JcFT rootstocks. The values are the means ± standard deviations of three plants per line (three independent experiments). The protein level in the WT was set as the standard, with a value of 1.00. JcFT (20 kDa) and JcActin (45 kDa) were quantified with AtFT and AtActin antibodies. The values are the means ± standard deviations of 15 plants per line (three independent experiments). The asterisks indicate significant differences in comparison with the WT at P < 0.05 according to Student’s t-test.
JcFT accelerates J. multifida flowering via grafting. (A) Wild-type (WT) J. multifida scions grafted onto WT and SUC2:JcFT J. curcas rootstocks. The triangles indicate the graft junctions. Scale bar = 10 cm. (B) The shoot apex of J. multifida grafted onto WT J. curcas rootstock. Scale bar = 1 cm. (C, D) Flowers produced on the shoot of J. multifida grafted onto SUC2:JcFT J. curcas rootstock. Scale bar = 1 cm. (E) Comparison of flowering plant percentage of J. multifida scions grafted onto WT and SUC2:JcFT J. curcas rootstocks. (F) Comparison of flowering time of J. multifida scions grafted onto WT and SUC2:JcFT J. curcas rootstocks. (G) Analysis of inflorescences generated on each grafted plant. (H) Analysis of the distance from the junction to the flower buds. (I) Western blot analysis of JcFT protein abundance in J. multifida scions grafted onto WT and SUC2:JcFT rootstocks. The values are the means ± standard deviations of three plants per line (three independent experiments). The protein level in the WT was set as the standard, with a value of 1.00. JcFT (20 kDa) and JcActin (45 kDa) were quantified with AtFT and AtActin antibodies. The values are the means ± standard deviations of 15 plants per line (three independent experiments). The asterisks indicate significant differences in comparison with the WT at P < 0.05 according to Student’s t-test.
Results
JcFT is essential for floral initiation in J. curcas
To investigate the functions of graft-transmitted JcFT in floral induction in woody perennial species of the Jatropha genus, we used transgenic J. curcas plants overexpressing JcFT driven by the phloem-specific A. thaliana SUC2 promoter (designated SUC2:JcFT; Li et al. 2014). Normally, the regenerated shoots in the shoot induction media, the grafted plantlets and seedlings of WT did not produce any flowers (Figure 1A–C). However, SUC2:JcFT plants flowered extremely early. Flower buds appeared on the regenerated shoots in the shoot induction media (Figure 1D). Since the shoots with flower buds failed to produce roots in root induction media, we grafted the shoots with flower buds onto WT seedling rootstocks and the scions produced flowers continually after grafting (Figure 1E). In addition, the T1 seedlings derived from self-pollinated SUC2:JcFT transgenic scions produced flowers at 15 days of age (Figure 1F). These results indicate that JcFT plays a key role in promoting flowering in J. curcas.
To further test whether JcFT is essential for J. curcas floral induction, we generated JcFT-RNAi constructs and transformed them into J. curcas cotyledons via Agrobacterium-mediated transformation (Fu et al. 2015). We obtained JcFT-RNAi transgenic plants, which did not produce any flowers for 8 years, whereas the WT plants usually produced flowers within the first year (Figure 2A–D). The flowering times of the WT and SUC2:JcFT plants were ~325 and 15 days, respectively (Figure 2E).
These results showed that the SUC2:JcFT plants flowered extremely early, whereas the JcFT-RNAi plants did not flower during an 8-year period. In good agreement with these observations, we measured the mRNA expression level and protein abundance in the transgenic plants by qRT-PCR and western blotting, respectively, and found that both the mRNA and protein levels increased in the SUC2:JcFT plants but decreased in JcFT-RNAi plants. The expression level of JcFT in SUC2:JcFT plants was 250-fold as WT, and the protein abundance was 19.58-fold as WT; the expression level of JcFT in JcFT-RNAi plants was only 30% as WT, and the protein abundance was 18% as WT (Figure 2F and G).
Graft-transmitted JcFT promotes flowering in species of the Jatropha genus
Previous research involving ectopic expression in WT and ft-10 mutant Arabidopsis plants indicated that JcFT controls flowering (Li et al. 2014). Nevertheless, no research has revealed that the function of JcFT is conserved in woody perennial plant species. In this study, we grafted 2-month-old shoots of four other Jatropha species, J. gossypifolia, J. integerrima, J. multifida and J. podagrica, onto WT and SUC2:JcFT transgenic J. curcas seedlings, respectively. The results showed that the majority of non-transgenic shoots of the four Jatropha species produced flowers within a month in an artificial climate chamber after being grafted onto SUC2:JcFT transgenic seedlings, whereas the other non-transgenic shoots grafted onto WT rootstocks did not produce any flowers under the same conditions (Figures 3A–D, 4A–E and 5A–D, and see Figure S1A available as Supplementary data at Tree Physiology Online). The number of inflorescences produced on each chimeric plant ranged from 1.5 to 2.6; although these plants produced more than one inflorescence, they did not bloom continually in the chamber (Figures 3G, 4H and 5G and see Figure S1D available as Supplementary data at Tree Physiology Online). We further analyzed the distance from the graft junction to inflorescence buds, and the data showed that the inflorescence buds clustered within a region from 8 to 25 cm from the graft junction (Figures 3H, 4I, 5H and see Figure S1E available as Supplementary data at Tree Physiology Online). Finally, using western blots, we measured the JcFT protein abundance in the shoot apices of different Jatropha scions grafted onto WT and SUC2:JcFTplants, and the results showed that the amount of FT in the shoots grafted onto SUC2:JcFT plants was 11.5–28.5 folds that of the shoots grafted onto the WT plants (Figures 3I, 4J and 5I, and see Figure S1F available as Supplementary data at Tree Physiology Online). These results suggest that, by being transported from the rootstock to the scion, JcFT also functions as a flowering accelerator in other woody perennial plant species of the Jatropha genus.
Graft-transmitted JcFT rescues the non-flowering phenotype of JcFT-RNAi transgenic J. curcas
In this study, we found that decreased JcFT mRNA and protein levels in JcFT-RNAi transgenic J. curcas prevented flowering (Figure 2F and G). These JcFT-RNAi plants and SUC2:JcFT plants with phloem-specific JcFT expression are therefore ideal materials to determine whether FT can be transported in woody perennials. Therefore, we grafted 35S:JcFT-RNAi transgenic shoots onto SUC2:JcFT transgenic seedlings. As shown in Figure 6A, WT scions produced flowers ~30 days after being grafted onto SUC2:JcFT transgenic seedlings. As expected, the 35S:JcFT-RNAi transgenic scions did not produce flowers after being grafted onto WT rootstocks (Figure 6B). However, the 35S:JcFT-RNAi transgenic scions produced flowers ~40 days after grafting onto SUC2:JcFT transgenic rootstocks (Figure 6C). Hence, through grafting, the SUC2:JcFT transgenic rootstocks successfully rescued the non-flowering phenotype of JcFT-RNAi transgenic J. curcas scions, which confirmed that the FT protein but not FT mRNA， can be transported from rootstocks to scions by grafting in woody perennials.
Overexpression of JcFT rescued the non-flowering phenotype of JcFT-RNAi transgenic J. curcas. (A) Wild-type (WT) J. curcas scions grafted onto SUC2:JcFT J. curcas rootstocks. (B) JcFT-RNAi J. curcas scions grafted onto WT J. curcas rootstocks. (C) JcFT-RNAi J. curcas scions grafted onto SUC2:JcFT J. curcas rootstocks. The red triangles indicate the graft junctions. Scale bar = 10 cm.
JcFT protein levels decreased with increasing translocation distance
After the young shoots of J. gossypifolia, J. integerrima, J. multifida and J. podagrica were grafted onto SUC2:JcFT seedlings, the scions produced flowers in 35.0–57.5 days but did not produce inflorescences continually and the grafted plants produced inflorescences only 1.5–2.5 times in the growth chamber (Figures 3 and 5 and Figure S1 available as Supplementary data at Tree Physiology Online). The SUC2:JcFT rootstocks rescued the JcFT-RNAi scions, but no flowers were produced during the second and third years (see Figure S2 available as Supplementary data at Tree Physiology Online). These results indicated that the transportability of FT may be limited by travel distance. To confirm this hypothesis, we grafted JcFT-RNAi scions of different lengths onto SUC2:JcFT rootstocks. The results showed that both the percentage of flowering scions and number of inflorescences produced per plant decreased with increasing scion length. When the length of the scion was 40 cm, no flowers were produced (Table 1). Furthermore, FT protein levels in the lateral buds at different positions were detected by western blot using AtFT antibody, together with AtActin antibody used as an internal reference. The results showed that FT abundance decreased with increasing distance from the junction to the buds. The FT abundance in the grafting joint was set as a value of 100, then after 100-cm transporting the FT abundance was decreased to 4.9–5.6 (Figure 7A). In particular, the decreasing tendency was obvious from 0 to 50 cm; however, there was no significant change in the region from 60 to 100 cm (Figure 7A and B). As such, we constructed a fitted curve according to the protein abundance data (inclusive of three independent replications), and an approximation formula (y = −1.7536x + 90.802) of the matched curve was proposed (Figure 7B). Based on the trend line, it may be difficult to produce flowers when the length of the scions reached 50 cm, which is consistent with our observation that no flowers were produced when the length of the scions was 40 cm (Table 1). Interestingly, some grafted scions produced inflorescences from the lateral buds near the rootstock but did not flower from the apical buds far away from the rootstock (Figure 7C–E). In addition, the grafted trees produced inflorescences in the first year (Figure 6C), but large plants higher than 2 m did not bloom from the grafted scions during the second, third and all subsequent years, whereas flowers and fruits were continually produced from the rootstocks every year (see Figure S2 available as Supplementary data at Tree Physiology Online). However, when the length of the scions was shortened to <20 cm, the new regenerated branches produced inflorescences again (see Figure S3 available as Supplementary data at Tree Physiology Online).
Comparison of flowering phenotypes in grafted plants
| Rootstocks | Scions | Scion length (cm) | Percentage of flowering plants (%) | Flowering time (day) | Flowering frequency | Distance from graft junction to flower buds (cm) |
|---|---|---|---|---|---|---|
| WT | JcFT-RNAi | 5 | 0.00 ± 0.00e | >250a | 0.00 ± 0.00d | – |
| SUC2:JcFT | WT | 5 | 90.21 ± 7.51a | 35.00 ± 8.57c | 2.95 ± 0.73a | 21.15 ± 9.87a |
| JcFT-RNAi | 5 | 63.74 ± 8.37b | 53.15 ± 13.90b | 1.80 ± 0.46b | 9.08 ± 0.64c | |
| 10 | 46.06 ± 6.35c | 57.50 ± 14.11b | 1.11 ± 0.31c | 19.33 ± 3.33a | ||
| 20 | 10.92 ± 2.91d | 50.76 ± 20.34b | 1.08 ± 0.14c | 25.58 ± 7.55a | ||
| 40 | 0.00 ± 0.00e | >250a | 0.00 ± 0.00d | – |
| Rootstocks | Scions | Scion length (cm) | Percentage of flowering plants (%) | Flowering time (day) | Flowering frequency | Distance from graft junction to flower buds (cm) |
|---|---|---|---|---|---|---|
| WT | JcFT-RNAi | 5 | 0.00 ± 0.00e | >250a | 0.00 ± 0.00d | – |
| SUC2:JcFT | WT | 5 | 90.21 ± 7.51a | 35.00 ± 8.57c | 2.95 ± 0.73a | 21.15 ± 9.87a |
| JcFT-RNAi | 5 | 63.74 ± 8.37b | 53.15 ± 13.90b | 1.80 ± 0.46b | 9.08 ± 0.64c | |
| 10 | 46.06 ± 6.35c | 57.50 ± 14.11b | 1.11 ± 0.31c | 19.33 ± 3.33a | ||
| 20 | 10.92 ± 2.91d | 50.76 ± 20.34b | 1.08 ± 0.14c | 25.58 ± 7.55a | ||
| 40 | 0.00 ± 0.00e | >250a | 0.00 ± 0.00d | – |
WT and SUC2:JcFT transgenic plants were used as rootstocks, and WT and JcFT-RNAi transgenic shoots were used as scions. The percentage of flowering plants, flowering time and flowering frequency of plants were calculated, and the distance from graft junction to flower buds was analyzed. The length of the rootstocks was 15 cm. The numbers in the table are presented as the means ± standard deviations. The values with different letters are significantly different (P < 0.05, Tukey’s test). The experiments were replicated three times, and 20 plants were analyzed in each experiment.
Comparison of flowering phenotypes in grafted plants
| Rootstocks | Scions | Scion length (cm) | Percentage of flowering plants (%) | Flowering time (day) | Flowering frequency | Distance from graft junction to flower buds (cm) |
|---|---|---|---|---|---|---|
| WT | JcFT-RNAi | 5 | 0.00 ± 0.00e | >250a | 0.00 ± 0.00d | – |
| SUC2:JcFT | WT | 5 | 90.21 ± 7.51a | 35.00 ± 8.57c | 2.95 ± 0.73a | 21.15 ± 9.87a |
| JcFT-RNAi | 5 | 63.74 ± 8.37b | 53.15 ± 13.90b | 1.80 ± 0.46b | 9.08 ± 0.64c | |
| 10 | 46.06 ± 6.35c | 57.50 ± 14.11b | 1.11 ± 0.31c | 19.33 ± 3.33a | ||
| 20 | 10.92 ± 2.91d | 50.76 ± 20.34b | 1.08 ± 0.14c | 25.58 ± 7.55a | ||
| 40 | 0.00 ± 0.00e | >250a | 0.00 ± 0.00d | – |
| Rootstocks | Scions | Scion length (cm) | Percentage of flowering plants (%) | Flowering time (day) | Flowering frequency | Distance from graft junction to flower buds (cm) |
|---|---|---|---|---|---|---|
| WT | JcFT-RNAi | 5 | 0.00 ± 0.00e | >250a | 0.00 ± 0.00d | – |
| SUC2:JcFT | WT | 5 | 90.21 ± 7.51a | 35.00 ± 8.57c | 2.95 ± 0.73a | 21.15 ± 9.87a |
| JcFT-RNAi | 5 | 63.74 ± 8.37b | 53.15 ± 13.90b | 1.80 ± 0.46b | 9.08 ± 0.64c | |
| 10 | 46.06 ± 6.35c | 57.50 ± 14.11b | 1.11 ± 0.31c | 19.33 ± 3.33a | ||
| 20 | 10.92 ± 2.91d | 50.76 ± 20.34b | 1.08 ± 0.14c | 25.58 ± 7.55a | ||
| 40 | 0.00 ± 0.00e | >250a | 0.00 ± 0.00d | – |
WT and SUC2:JcFT transgenic plants were used as rootstocks, and WT and JcFT-RNAi transgenic shoots were used as scions. The percentage of flowering plants, flowering time and flowering frequency of plants were calculated, and the distance from graft junction to flower buds was analyzed. The length of the rootstocks was 15 cm. The numbers in the table are presented as the means ± standard deviations. The values with different letters are significantly different (P < 0.05, Tukey’s test). The experiments were replicated three times, and 20 plants were analyzed in each experiment.
JcFT protein abundance decreased with increasing transport distance. (A) Western blot analysis of JcFT protein abundance in the scions of JcFT-RNAi J. curcas grafted onto SUC2:JcFT rootstocks (three independent experiments). The protein level in the rootstock was set as the standard, with a value of 100, and the values of the bands were quantified by Image lab software (Bio-Rad, Hercules, CA, USA.). JcFT (20 kDa) and JcActin (45 kDa) were quantified by AtFT and AtActin antibodies. (B) Relationship between distance from graft junctions to the buds and JcFT protein level. The values were calculated on the basis of three independent experiments, and the numbers are presented as the means ± standard deviations. (C) One-year-old plant derived from a JcFT-RNAi scion grafted onto a SUC2:JcFT rootstock. Scale bar =10 cm. (D) Shoot apex of a JcFT-RNAi scion grafted onto the SUC2:JcFT rootstock in (C). No flowers were produced. Scale bar = 5 cm. (E) Lateral buds of a JcFT-RNAi scion grafted onto the SUC2:JcFT rootstock in (C). Inflorescences were produced. Scale bar = 5 cm. The triangle indicates the graft junction. Scale bar = 10 cm.
Schematic diagram of JcFT transcription, translation and graft transmission from rootstock to scion. JcFT-RNAi scions were grafted onto SUC2:JcFT rootstocks. JcFT protein abundance decreased with increasing distance from the graft junction (yellow rectangle in the middle of the trunk) to the buds. The highest abundance was located in the buds of SUC2:JcFT rootstocks, which is marked as red, and the varying degrees of blackness in the stem indicate different levels of JcFT protein. The transport directions of the JcFT protein are shown with yellow arrows, and the thickness of the arrows shows the abundance of JcFT protein. 35S, the 35S RNA promoter of cauliflower mosaic virus; GA, gibberellin; JcFT, Jatropha curcas FLOWERING LOCUS T; OE, overexpression; RNAi, RNA interference; siRNA, small interfering RNA; SUC2, the promoter of Arabidopsis thaliana SUCROSE TRANSPORTER 2.
Discussion
FT plays a conservative role in initiating the vegetative to floral transition via grafting in both perennial and annual plants
In this study, we analyzed the function and transferability of JcFT by constructing SUC2:JcFT and JcFT-RNAi transgenic plants and 13 kinds of grafted plants derived from five species of Jatropha genus. The SUC2:JcFT transgenic plants flowered extremely early (Li et al. 2014, Ye et al. 2014), whereas JcFT-RNAi transgenic plants did not bloom (Figure 1D–F and 2D,E). Both mRNA and protein levels consistently increased in the SUC2:JcFT plants and decreased in the JcFT-RNAi plants (Figure 2F and G). These results clearly indicate that the reduced expression of JcFT by RNAi prevented the transition from vegetative to reproductive growth. The JcFT-RNAi transgenic plants did not bloom, which also indicates that FT plays a vital role in the regulation of flowering in woody perennial plants and that no other genes that are functionally redundant with JcFT are present in J. curcas.
A large number of grafting experiments performed with many annual species, such as Arabidopsis (Corbesier et al. 2007, Lu et al. 2012), tobacco (Freiman et al. 2015), rice (Tamaki et al. 2007, Song et al. 2017), Cucurbita moschata (Yoo et al. 2013a), cucurbits (Lin et al. 2007), tomato (Lifschitz et al. 2006) and potato (Navarro et al. 2011), have demonstrated that FT proteins are translocated to the shoot apex to initiate floral morphogenesis. Previous studies have shown that overexpression of FT orthologs in other woody plant species, such as orange (Endo et al. 2005), poplars (Zhang et al. 2010), plum (Srinivasan et al. 2012), apple (Wenzel et al. 2013) and citrus (Sinn et al. 2021), also results in an early-blooming phenotype. In this study, our results show that JcFT promotes flowering not only in J. curcas (Figure 1D–F) but also in other woody perennial plant species, including J. gossypifolia (Figure 3C–F), J. integerrima (Figure 4D–G), J. multifida (Figure 5C–F) and J. podagrica (see Figure S1A–C available as Supplementary data at Tree Physiology Online). The floral initiation of young WT shoots excised from J. curcas (Table 1), J. gossypifolia (Figure 3F), J. integerrima (Figure 4G), J. multifida (Figure 5F) and J. podagrica (see Figure S1C available as Supplementary data at Tree Physiology Online) was quickly activated after the shoots were grafted onto SUC2:JcFT rootstocks. Furthermore, through grafting, SUC2:JcFT transgenic rootstocks successfully rescued the non-flowering phenotype of JcFT-RNAi scions (Figure 6C). These results indicate that the mobility of JcFT was extensively existed in perennials of Jatropha genus. Taken together, these results clearly demonstrate that the FT protein can be translocated from rootstocks to scions in woody perennials, and also indirectly confirm that the graft-transmitted florigen is FT protein, not FT mRNA, because JcFT mRNA would be degraded in JcFT-RNAi scions. Overall, we conclude that the transportability and the function in initiating the vegetative to floral transition of florigen FTs are broadly conserved in both annual and perennial plants.
FT transportability is limited by translocation distance in trees
Although the graft-transportability of FTs has been well demonstrated in annual species, there are some contradictory reports on floral induction via graft-transmitted FTs in trees (Putterill and Varkonyi-Gasic 2016, Zhu et al. 2021). Grafting experiments with rootstocks overexpressing FT transgenes revealed no floral induction ability in the receptor scions of poplar (Zhang et al. 2010), apple (Wenzel et al. 2013) and cassava (Bull et al. 2017, Odipio et al. 2020). However, floral induction via graft-transmitted FTs was found in J. curcas (Ye et al. 2014), blueberry (Song et al. 2019) and citrus (Soares et al. 2020). To resolve this apparent contradiction, we demonstrate in this study that JcFT transportability was limited by translocation distance from graft junctions to the buds, and JcFT abundance decreased with increasing scion length (Figure 7A and B). The results suggest that FT cannot be transported long distances in woody perennial plants.
The FT expression level in the SUC2:JcFT plants was more than 200-fold higher than that in the WT (Figure 2F), which resulted in that the flowering time of the T1 transgenic seedlings was only 15 days after germination (Figure 1F). Both the percentage of flowering scions and the number of inflorescences produced per grafted plant decreased with increasing scion length (from 5 to 20 cm), and no flowers were produced when the length of the JcFT-RNAi scions was 40 cm (Table 1). We speculate that some JcFT might be degraded by serine/cysteine proteases during transportation (Kim et al. 2016, Qin et al. 2017), resulting in a decrease in amount of JcFT transported to buds of scions. Hence, the key to successfully inducing flowering in scions by the graft-transmitted FT from transgenic rootstocks is to shorten the length of the scions. This was further supported by our observation that the new regenerated branches produced inflorescences again when the length of the scions was shortened to <20 cm (see Figure S3 available as Supplementary data at Tree Physiology Online). Thus, we propose that the failure of floral induction via graft-transmitted FTs in poplar, apple and cassava (Zhang et al. 2010, Wenzel et al. 2013, Bull et al. 2017, Odipio et al. 2020) may result from a relatively low level of transgene FT expression in rootstocks and/or relatively long scions used for grafting experiments. According to the above results, we propose a schematic diagram of JcFT transcription, translation and transport from rootstock to scion (Figure 8).
In addition, to obtain FT transgenic rootstocks with the ability to induce precocious flowering in non-transgenic scions, the selection of a suitable promoter for driving FT transgene expression is crucial. In our previous study (Li et al. 2014), the strong constitutive 35S promoter-driven JcFT transgenic shoots with flower buds failed to develop normally, whereas the viable transgenic Jatropha shoots were obtained by using the phloem-specific SUC2 promoter, which were grafted onto rootstocks of WT seedlings. Consistently, Ye et al. (2014) obtained a weak synthetic G10-90 promoter-driven JcFT transgenic Jatropha rootstocks, which were capable of promoting flowering in the recipient scions, but failed to generate transgenic Jatropha plants overexpressing JcFT from a strong constitutive 35S promoter. Recently, Soares et al. (2020) examined four different types of promoters for efficient expression of the CcFT3 transgene in citrus, including two constitutively expressed promoters (a strong 35S promoter and a weaker NOPALINE SYNTHASE [NOS] promoter), the phloem-specific Arabidopsis SUC2 and the heat inducible A. thaliana HEAT SHOCK PROTEIN 18.2 (AtHSP18.2) promoter. Only the phloem-specific SUC2 promoter-driven CcFT3 transgenic citrange rootstocks showed normal morphological characteristics, exhibited normal vigor and were capable of inducing precocious flowering in juvenile non-transgenic scions (Soares et al. 2020). Hence, the Arabidopsis SUC2 promoter could be an excellent phloem specific promoter driving transgene expression of graft-transmitted JcFT for floral induction in woody perennial species.
Modulation of FT abundance for plant genetic improvement and breeding
For crops, flowering and reproductive cycling are keys for field performance. Modifications of the expression of FT-like genes shifted the proportion of vegetative growth (shoots and leaves) to reproductive growth (flowers, fruits and seeds), providing a range of variation in shoot architecture, plant size, flowering time, fruit set and seed production (Pin and Nilsson 2012, Moraes et al. 2019). This ample variation can be exploited for the genetic studies and breeding of woody perennials. The long juvenile period of woody perennial trees has hindered both genetic studies and traditional breeding. To reduce the generation time in trifoliate orange, a citrus homolog of the FT gene was introduced into trifoliate orange, and the transgenic plants showed extremely early flowering and fruiting (Endo et al. 2005), which was used recently in a fast-track breeding system to introduce citrus tristeza virus (CTV) resistance of trifoliate orange into citrus germplasm (Endo et al. 2020). In poplar, overexpression of Populus trichocarpa FT1 and FT2 driven by a soybean heat-inducible promoter successfully shortened the generation cycle for breeding (Zhang et al. 2010). A previous study showed that expressing A. thaliana or citrus FT genes can successfully promote the transition from the vegetative to the reproductive phase in juvenile citrus plants (Velázquez et al. 2016). Recently, Sinn et al. (2021) found moderate expression of a chimeric FT protein resulted in precocious blooming largely without negative effects in edible citrus cultivars, suggesting an additional valuable tool for rapid-cycle citrus breeding. Similarly, constitutive expression of A. thaliana FT gene (AtFT) in Eucalyptus led to very early flowering, which was also considered as an effective means for acceleration of eucalypt tree breeding and genetic studies (Klocko et al. 2016). By taking advantage of the early-continuous flowering transgenic plums overexpressing poplar FT (PtFT1) (Srinivasan et al. 2012), a novel breeding strategy, termed FasTrack breeding, has been developed to shorten the breeding cycle of plums (Petri et al. 2018).
The early flowering transgenic J. curcas plants overexpressing JcFT obtained in this study may have potential for reducing the length of the breeding cycle of Jatropha species via grafting or crossing (van Nocker and Gardiner 2014). In addition, JcFT-RNAi transgenic J. curcas has not flowered in 8 years, and the JcFT-RNAi plants showed a significant increase in plant height and biomass as compared with WT (Figure 2D and see Figure S4C–F available as Supplementary data at Tree Physiology Online). A similar strategy could also be applied to increase the biomass of certain other economically important tree species.
Data availability statement
No new sequence data were published in the present paper. Sequence data included in our manuscript can be obtained from the publicly available genome of Jatropha curcas (https://www.ncbi.nlm.nih.gov/bioproject/38697) under the following accession numbers: JcFT (NP_001295681), JcACTIN1 (NM_112764).
Acknowledgments
We thank Ping Huang, Purui Li, Yuan Wu, Jiapeng Ke, Zhongjiang Liu and Chuanjia Xu for helping transplant the transgenic Jatropha seedlings. The authors gratefully acknowledge the Central Laboratory of the Xishuangbanna Tropical Botanical Garden (XTBG) and the National Forest Ecosystem Research Station at Xishuangbanna for providing the research facilities.
Conflicts of interest
The authors declare that they have no competing financial interests.
Funding
This work was supported by Yunnan Fundamental Research Projects (grant nos 2018FB060, 202003AD150011), the West Light Foundation of Chinese Academy of Science (CAS) (grant no. Y9XB091), the Natural Science Foundation of China (grant no. 31700273, 31771605), the Young Elite Scientists Sponsorship Program by Chinese Society of Tropical Crops (grant no. CSTC-QN201701) and the Key Deployment Program of the CAS (grant no. ZDRW-ZS-2017-2).
Authors’ contributions
M.T. designed and performed the experiments, analyzed the data, and revised the paper. X.B. performed the experiments, analyzed the data and wrote the paper. J.W. and C.L. revised the paper. T.C., X.M. and H.D. helped collect the data. Z.-F.X. conceived the study and revised the manuscript. All authors reviewed and approved the final manuscript.
Reference
Author notes
Mingyong Tang, Xue Bai contributed equally to this work.
