GmFT2a, a soybean homolog of FLOWERING LOCUS T, is involved in flowering transition and maintenance

Abstract

Background: Flowering reversion can be induced in soybean (Glycine max L. Merr.), a typical short-day (SD) dicot, by switching from SD to long-day (LD) photoperiods. This process may involve florigen, putatively encoded by FLOWERING LOCUS T (FT) in Arabidopsis thaliana. However, little is known about the potential function of soybean FT homologs in flowering reversion.

Methods: A photoperiod-responsive FT homologue GmFT (renamed as GmFT2a hereafter) was cloned from the photoperiod-sensitive cultivar Zigongdongdou. GmFT2a gene expression under different photoperiods was analyzed by real-time quantitative PCR. In situ hybridization showed direct evidence for its expression during flowering-related processes. GmFT2a was shown to promote flowering using transgenic studies in Arabidopsis and soybean. The effects of photoperiod and temperature on GmFT2a expression were also analyzed in two cultivars with different photoperiod-sensitivities.

Results: GmFT2a expression is regulated by photoperiod. Analyses of GmFT2a transcripts revealed a strong correlation between GmFT2a expression and flowering maintenance. GmFT2a transcripts were observed continuously within the vascular tissue up to the shoot apex during flowering. By contrast, transcripts decreased to undetectable levels during flowering reversion. In grafting experiments, the early-flowering, photoperiod-insensitive stock Heihe27 promotes the appearance of GmFT2a transcripts in the shoot apex of scion Zigongdongdou under noninductive LD conditions. The photothermal effects of GmFT2a expression diversity in cultivars with different photoperiod-sensitivities and a hypothesis is proposed.

Conclusion: GmFT2a expression is associated with flowering induction and maintenance. Therefore, GmFT2a is a potential target gene for soybean breeding, with the aim of increasing geographic adaptation of this crop.

Figure 1. The expression pattern of GmFT2a in soybean.

The expression level of GmFT2a in different tissues or organs was detected by real-time quantitative PCR at 7 or 25 days after the first pair of unifoliate leaves fully expanded under short days (SD) or long days (LD). GmACTIN was used as a control gene. The data of flower and pod were not available for 7 SD, 7 LD or 25 LD, as cultivar Zigongdongdou did not flower at that time. SA, shoot apex (not including any manually removable leaves). For each real-time quantitative PCR point, the averages and standard errors are the result of three replications.

Figure 2. The diurnal circadian rhythm of GmFT2a gene expression under SD and LD conditions.

Unifoliate leaves were sampled every 3 hours since 1 hour after dawn. White bars at the bottom represent light phases, and black bars dark phases. ZG SD, Zigongdongdou under SD conditions; ZG LD, Zigongdongdou under LD conditions; HH SD, Heihe27 under SD conditions; and HH LD, Heihe27 under LD conditions.

Figure 3. GmFT2a expression during flower development.

(a) Global floral primordia. (b) Floral primordia with emerging sepals. (c) Floral primordia with sepals. (d) Flower structure initially formed with primordia of petal, stamen and pistil. (e) Developing carpel and stamen. (f) Initially formed ovary and stamen. (g) Ovary with ovules and developed stamen with pollen. (h) Flower bud. The profile of the flower primordia is enclosed with yellow dashes, and the GmFT2a signals are contained mainly in the region enclosed with red dashes. a, anther; br, bract; c, carpel; cp, carpel primordium; f, filament; fp, flower primordium; p, pistil; pe, petal; po, pollen; se, sepal; sp, stamen primordium; st, stamen. Bar, 250 µM.

Figure 4. The expression pattern of GmFT2a in soybean cultivar Zigongdongdou shoot apices under different photoperiod conditions by in situ hybridization.

(a) Shoot apex detected by the sense GmFT2a probe as a control. All other shoot apices were detected by the antisense GmFT2a probe shown in (b)–(o). (a) Shoot apex at the 22nd day of SD conditions (briefed as SD22). (b) and (c) Shoot apexes at LD1 and LD13, respectively. (d)–(i) Shoot apices at SD1, SD7, SD13, SD16, SD19, and SD22, respectively. (j)–(o) Shoot apices treated by LD for 3, 6, 9, 12, 15, and 27 days after SD13, respectively. am, apical meristem; axm, axillary meristem; br, bract; c, carpel; fm, floral meristem; fp, floral primordium; im, inflorescence meristem; ip, inflorescence primordium; le, leaf/trifoliolate leaf; p, pistil; pe, petal; pep, petal primordium; pp, pistil primordium; rf, reversed flower; s, stamen; se, sepal; sep, sepal primordium; sp, stamen primordium; tlp, trifoliolate leaf primordium. Bar, 250 µM.

Figure 5. Early-flowering stock promotes the appearance of GmFT2a transcripts in late-flowering scion by grafting.

(a–b) Shoot apices of self-grafted Heihe27 (HH/HH, scion/stock) at the 3rd or 6th day after grafting (DAG), respectively. (c) Shoot apex of self-grafted Zigongdongdou (ZG/ZG) at 18 DAG. (d–f) Shoot apices of scion Zigongdongdou grafted on stock Heihe27 (ZG/HH) at 3, 9 or 14 DAG, respectively. am, apical meristem; axm, axillary meristem; br, bract; fm, floral meristem; fp, floral primordium; pp, pistil primordium; tlp, trifoliolate leaf primordium. Bar, 250 µM.

Figure 6. Transgenic analysis of GmFT2a in Arabidopsis thaliana.

(a) Upper, Arabidopsis plants grown under long days (LD); Lower, Arabidopsis plants grown under short days (SD) when the transgenic plants flowered. From left to right, plants represent wildtype, transgenic line 17 and 18. (b) Transgenic overexpression of GmFT2a fully suppresses the late-flowering phenotype of a FT mutant ft10. From left to right, plants represent wild type, mutant ft10, and transgenic line 2 with mutant ft10 genetic background. (c) The days to flowering of Arabidopsis plants. (d) The number of cauline and rosette leaves at flowering. SR, rosette leaves under SD conditions. SC, cauline leaves under SD conditions. LR, rosette leaves under LD conditions. LC, cauline leaves under LD conditions. (e) Detection of GmFT, AtFT, AtAP1, AtSOC1, and AtFLC by real-time quantitative PCR. Asterisks show that the data was not available at flowering. For each real-time quantitative PCR point, averages and standard errors are the result of three replications.

Figure 7. Transgenic GmFT2a induces precocious flowering in soybean cultivar Zigongdongdou.

(a) A wild-type Zigongdongdou plant. (b) A transgenic Zigongdongdou plant showing precocious flowering at the axils of the unifoliate leaves. (c) The overview of wild-type (left) and transgenic (right) plants, including the plants shown in (a) and (b). (d) The close shot of the transgenic plant in (b) shows the precocious flowers at the axils of the unifoliate leaves. (e) Identification of transgenic GmFT2a plants by reverse transcription PCR with primers specific to 35S::GmFT. (f) Days to flowering from emergence of the transgenic plants and wild-type plants. The asterisk indicates that wild-type plants cannot flower at LD conditions during this experiment.

Figure 8. GmFT2a expression under different combinations of photoperiod and temperature.

GmFT2a expression was detected in Zigongdongdou (a), Zigongdongdou pretreated by 13 SD (b), and Heihe27 (c) under short days at 30°C (SD30°C) or 20°C (SD20°C), and under long days 30°C (LD30°C) or 20°C (LD20°C). Zigongdongdou plants, including pretreated plants, did not flower throughout this experiment under LD conditions. For all other treatments, sampling ceased when the plants began blooming. For each real-time quantitative PCR point, averages and standard errors are the result of three replications.