Arabidopsis UMAMIT24 and 25 are amino acid exporters involved in seed loading

Abstract

Phloem-derived amino acids are the major source of nitrogen supplied to developing seeds. Amino acid transfer from the maternal to the filial tissue requires at least one cellular export step from the maternal tissue prior to the import into the symplasmically isolated embryo. Some members of UMAMIT (usually multiple acids move in an out transporter) family (UMAMIT11, 14, 18, 28, and 29) have previously been implicated in this process. Here we show that additional members of the UMAMIT family, UMAMIT24 and UMAMIT25, also function in amino acid transfer in developing seeds. Using a recently published yeast-based assay allowing detection of amino acid secretion, we showed that UMAMIT24 and UMAMIT25 promote export of a broad range of amino acids in yeast. In plants, UMAMIT24 and UMAMIT25 are expressed in distinct tissues within developing seeds; UMAMIT24 is mainly expressed in the chalazal seed coat and localized on the tonoplast, whereas the plasma membrane-localized UMAMIT25 is expressed in endosperm cells. Seed amino acid contents of umamit24 and umamit25 knockout lines were both decreased during embryogenesis compared with the wild type, but recovered in the mature seeds without any deleterious effect on yield. The results suggest that UMAMIT24 and 25 play different roles in amino acid translocation from the maternal to filial tissue; UMAMIT24 could have a role in temporary storage of amino acids in the chalaza, while UMAMIT25 would mediate amino acid export from the endosperm, the last step before amino acids are taken up by the developing embryo.

Fig. 1.

Secretion of amino acids in the medium by UMAMIT-expressing yeast cells. 22Δ10α cells were transformed with the empty vector pDR196-Ws from which the gateway cassette has been removed, or containing UMAMIT25, UMAMIT24, or UMAMIT23. Cells were grown for 22 h in liquid medium; then the amino acid composition of the medium was determined by UPLC. The Arg peak could not be resolved from the large Gln peak in samples so these amino acids are presented by a single bar. Concentrations found in the medium were divided by the OD of the culture, and the value obtained for the empty vector was subtracted. Error bars correspond to the SE (n=4 biological replicates). Significant differences (P<0.01) compared with the empty vector are indicated by an asterisk according to t-test.

Fig. 2.

Localization of UMAMIT24 in Arabidopsis seeds. (A) Seeds from plants expressing UMAMIT24–GFP were observed with valves removed, under either bright light (top rows) or GFP excitation wavelength (bottom rows). (B) Dissected seeds revealing the embryo (bottom left of the picture) and the seed coat. The white arrowhead points to the chalazal seed coat. DAP, days after pollination.

Fig. 3.

Localization of UMAMIT25 in Arabidopsis seeds. (A) Seeds from plants expressing UMAMIT25–GFP were observed with valves removed, under either bright light (top rows) or GFP excitation wavelength (bottom rows). (B) Dissected seeds revealing the embryo (bottom left of the picture) and the seed coat. DAP, days after pollination.

Fig. 4.

Subcellular localization of UMAMIT24–GFP and UMAMIT25–GFP. (A–C) Protoplasts from seeds expressing UMAMIT24promoter:UMAMIT24-GFP. (A) GFP, (B) bright field, (C) merged images. Scale bars are 5 µm. (D–G) Arabidopsis hypocotyl from 1-week-old seedlings stably expressing 35S:UMAMIT25-GFP. (D) GFP, (E) chlorophyll, (F) merged, and (G) bright field. Scale bars are 20 µm.

Fig. 5.

Seed yields of WT, umamit, and complemented lines. (A) Number of seeds produced per plant. (B) Mass of seeds per plant. Error bars correspond to the SD (n=4). Significant differences (P<0.05) are indicated by different letters according to one-way ANOVA in conjunction with Tukey’s test.

Fig. 6.

Protein and amino acid contents in seeds and pericarp tissue. Plants were grown side by side in soil under long-day conditions for 5 weeks. One biological replicate represents seeds or pericarp tissue from two siliques. Amino acid content in seeds (A) and pericarp tissue (B) or protein content in seeds (C) and pericarp tissue (D) are shown. Samples were harvested at 7, 10, and 14 DAP, or at the mature stage. Error bars correspond to the SD (n≥3 biological replicates). Significant differences (P<0.05) are indicated by different letters according to one-way ANOVA in conjunction with Tukey’s test. Contents of individual amino acids from the same data set are presented in Supplementary Tables S4 and S5.

Fig. 7.

Glutamine and sucrose transfer assay in isolated siliques. ¹⁵N/¹³C ratio of isotopic excess found in seeds at 10 DAP. Siliques were excised 10 DAP, and transport was measured according to the experimental design presented in the Materials and methods. One biological replicate corresponds to three siliques worth of seeds coming from the same plant. ¹⁵N and ¹³C enrichment have been calculated against the non-labeled sample control. Error bars correspond to the SD (n≥3 biological replicates). Significant differences (P<0.05) are indicated by different letters according to one-way ANOVA in conjunction with Tukey’s test.

Fig. 8.

Model of amino acid transport into the seed by UMAMIT24 and UMAMIT25. Amino acids are delivered from the mother plant to the growing embryo through the funicule, then unloaded from the chalazal seed coat. UMAMIT24 localizes at numerous vacuoles in the chalazal seed coat and is involved in this process, possibly acting as a transporter for temporary storage of amino acids in this tissue. Once out of the chalazal, seed coat amino acids are transported radially through the integument cells then delivered to the endosperm, and finally to the embryo. UMAMIT25, expressed in the endosperm cells, unloads amino acids out of those cells which then become available for import by the embryo. Loss of function of UMAMIT25 reduces the amino acid unloading process, and decreases amino acid transfer from the rest of the plant body to the growing embryo.