A cascade of sequentially expressed sucrose transporters in the seed coat and endosperm provides nutrition for the Arabidopsis embryo

Abstract

Developing plant embryos depend on nutrition from maternal tissues via the seed coat and endosperm, but the mechanisms that supply nutrients to plant embryos have remained elusive. Sucrose, the major transport form of carbohydrate in plants, is delivered via the phloem to the maternal seed coat and then secreted from the seed coat to feed the embryo. Here, we show that seed filling in Arabidopsis thaliana requires the three sucrose transporters SWEET11, 12, and 15. SWEET11, 12, and 15 exhibit specific spatiotemporal expression patterns in developing seeds, but only a sweet11;12;15 triple mutant showed severe seed defects, which include retarded embryo development, reduced seed weight, and reduced starch and lipid content, causing a "wrinkled" seed phenotype. In sweet11;12;15 triple mutants, starch accumulated in the seed coat but not the embryo, implicating SWEET-mediated sucrose efflux in the transfer of sugars from seed coat to embryo. This cascade of sequentially expressed SWEETs provides the feeding pathway for the plant embryo, an important feature for yield potential.

Figure 1.

Expression and Localization of SWEET11, 12, and 15 in Seeds. (A) Tissue-specific SWEET gene expression from microarray analysis (Dean et al., 2011) and protein accumulation as assessed by translational SWEET-GFP fusions during seed development. Representative images of developing seeds are shown above the panels. MSC, micropylar end of seed coat; MCE, micropylar endosperm; OI, outer integument; S, suspensor. (B) Confocal images of eGFP fluorescence in transgenic Arabidopsis seeds expressing translational SWEET11-, 12-, or 15-eGFP fusions under control of their native promoters. The white arrow points to red autofluorescence of the cotyledon. The blue arrow points to red propidium iodide staining of cell walls. The red arrow points to the suspensor. Bar = 50 μm.

Figure 2.

Spatial and Temporal Expression of SWEET15. (A) to (C) Distribution of GUS activity in whole plants at different ages: (A) 25 d old, (B) 38 d old, and (C) 44 d old. (D) High GUS activities detected in the inflorescence. (E) and (F) GUS activity (E) in leaves at different developmental stages and in seed (F). Two individual transgenic lines were tested. One of the lines giving stronger GUS activity is shown here. GC, green cauline leaf; GR, green rosette leaf and seed; PS, partially senescent leaf; LS, late stage senescent leaf.

Figure 3.

SWEET11, 12, and 15 Are Required for Embryo Development. Embryo phenotype of sweet single mutants at 8 DPA (A), double mutants at 8 DPA (B), and triple mutants at 8 and 13 DPA (C). Bars = 0.2 mm.

Figure 4.

Comparison of the Developmental Stages of Embryos in Col and sweet11;12;15 and Determination of the Maternal Contribution to Embryo Growth. (A) and (B) Images of cleared seeds taken by DIC microscopy. Col-0 (A) and sweet11;12;15 (B) seeds at similar stages of embryogenesis. Bar = 0.1 mm. (C) Embryo area measured at the corresponding stages from images in (A) and (B) (mean ± sd, 3 DPA, n ≥ 167; 4 DPA, n ≥ 61; 5 to 8 DPA, n ≥ 18). (D) Embryo area measured at 9 d after pollination from the indicated reciprocal crosses (mean ± sd, n ≥ 13).

Figure 5.

Characterization of Other Seed Development-Related Phenotypes of sweet11;12;15. (A) Comparison of dry seed weight per plant from Col-0, sweet11;12, and sweet11;12;15 (mean ± sd, n ≥ 14 plants, from four independent experiments). (B) Comparison of dry seed weights of 20 siliques from Col-0, sweet11;12, and sweet11;12;15 (mean ± sd, n ≥ 13 plants, from four independent experiments). (C) Comparison of number of seeds from each silique from Col-0, sweet11;12, and sweet11;12;15 (mean ± sd, n = 30, from three plants). Two biological repeats were done. (D) Total fatty acid content (measured as FAME) of dry seeds from Col-0, sweet11;12, and sweet11;12;15 (mean ± sd, n = 14 plants, from four independent experiments). (E) and (F) Scanning electron microscopy images of dry seeds from Col-0 (E) and from sweet11;12;15 (F).

Figure 6.

Effect of Sucrose on Early Root Growth in Seedlings and on in Vitro-Grown Embryos. (A) Comparison of root growth of Col-0, sweet11;12, and sweet11;12;15 seedlings grown on sugar-free and sugar-supplied media. (B) Embryos dissected from seeds at 3.5 DPA grown in vitro for 5.5 d with and without supply of sucrose. (C) Embryo size as approximated by area measurements of embryo from in vitro culture with 5% sucrose and dissected seed of intact plant at 9 DPA. Area of sweet11;12;15 is normalized to Col-0 (mean ± sd, n ≥ 20, from three independent experiments).

Figure 7.

Analysis of Silique and Embryo Growth for in Vitro-Cultured Flowers in Medium with or without Sucrose. (A) to (D) Comparison of silique development ([A] and [C]) and embryo growth ([B] and [D]) from Col-0 ([A] and [B]) and sweet11;12;15 ([C] and [D]) cultured for 9 d in sugar-free medium. (E) to (H) Comparison of silique development ([E] and [G]) and embryo growth ([F] and [H]) from Col-0 ([E] and [F]) and sweet11;12;15 ([G] and [H]) cultured for 9 d in medium with 3% sucrose. (B), (D), (F), and (H) Embryos were dissected from three siliques cleared with 0.2 NaOH and 1% SDS solution by gently pressing cover slide. Yellow coloration was observed in Col-0, while sweet11;12;15 showed reduced coloration. Black bar = 5 mm and white bar = 0.5 mm. (I) Embryo area was measured for data shown in (B), (D), (F), and (H) (mean ± sd, n = 7 [B], n = 24 [F], and n = 21 [H]). None of embryos of sweet11;12;15 cultured in medium without sucrose could be measured due to abortion. The data shown are from a single experiment. Two independent experiments showed similar results.

Figure 8.

Analysis of Starch Accumulation in Embryos and Seed Coats. (A) and (B) Comparison of starch accumulation in sweet11;12;15 and Col-0 at 9, 10, and 11 DPA. (C) and (D) Cross section of seeds stained with Lugol's iodine solution at 9 DPA. (E) Enzymatic quantification of starch content of seed coats and embryos from Col-0 or sweet11;12;15 at 10 DPA (mean ± sd, n = 4).

Figure 9.

Model for Multistep Sequential Apoplasmic Transport Steps during Seed Development Initiated by SWEET Sucrose Transporters. Sucrose, which arrives in the seed coat via the funicular phloem, can enter the outer integument of the seed coat likely through plasmodesmata. SWEET15, expressed in the outer integument (OI), exports sucrose into the apoplasm (AP). Possibly, SWEET11 is involved in the release of sucrose from the inner integument (II). SWEET12 appears to play a role in transport of sucrose out of cells at the micropylar end of the seed coat (MSC) and SWEET11 and 15 from the micropylar endosperm (MCE) to supply the embryo proper (EM).