GLUCOSAMINE INOSITOLPHOSPHORYLCERAMIDE TRANSFERASE1 (GINT1) Is a GlcNAc-Containing Glycosylinositol Phosphorylceramide Glycosyltransferase

Abstract

Glycosylinositol phosphorylceramides (GIPCs), which have a ceramide core linked to a glycan headgroup of varying structures, are the major sphingolipids in the plant plasma membrane. Recently, we identified the major biosynthetic genes for GIPC glycosylation in Arabidopsis (Arabidopsis thaliana) and demonstrated that the glycan headgroup is essential for plant viability. However, the function of GIPCs and the significance of their structural variation are poorly understood. Here, we characterized the Arabidopsis glycosyltransferase GLUCOSAMINE INOSITOLPHOSPHORYLCERAMIDE TRANSFERASE1 (GINT1) and showed that it is responsible for the glycosylation of a subgroup of GIPCs found in seeds and pollen that contain GlcNAc and GlcN [collectively GlcN(Ac)]. In Arabidopsis gint1 plants, loss of the GlcN(Ac) GIPCs did not affect vegetative growth, although seed germination was less sensitive to abiotic stress than in wild-type plants. However, in rice, where GlcN(Ac) containing GIPCs are the major GIPC subgroup in vegetative tissue, loss of GINT1 was seedling lethal. Furthermore, we could produce, de novo, "rice-like" GlcN(Ac) GIPCs in Arabidopsis leaves, which allowed us to test the function of different sugars in the GIPC headgroup. This study describes a monocot GIPC biosynthetic enzyme and shows that its Arabidopsis homolog has the same biochemical function. We also identify a possible role for GIPCs in maintaining cell-cell adhesion.

Figure 1.

Representative structures and synthetic pathways of GIPC sugar headgroups in plants. The first (Ins) and second (GlcA) carbohydrate residues are conserved in plants, but the third residue (Hex, HexN, or HexNAc) is variable in plant tissues/species. This study identifies the (N-acetyl)hexosamine-specific transferase GINT1. IPCS, inositolphosphorylceramide synthase (ceramide phosphoinosityl transferase); IPUT1, inositolphosphorylceramide glucuronosyl transferase; GMT1, GIPC mannosyl transferase; P, phosphate group; Ins, inositol; Cer, ceramide.

Figure 2.

Tissue-specific expression of AtGINT1 and AtGMT1 and comparison of GIPC subclasses. A to H, Promoter-GUS staining indicating expression patterns of AtGINT1 (A, C, E, and G) and AtGMT1 (B, D, F, and H) in young seedlings (A and B), flowers (C and D), developing siliques (E and F), and mature seeds (G and H). I and J, GT expression analysis (I) and quantification of GIPC contents (J) during seed germination. Data are means of three or four biological replicates ± sd.

Figure 3.

GIPC composition of Atgint1 plants. A to H, LC-MS/MS was used to determine the GIPC composition of dry seeds (A and B), leaves of 3-week-old plants (C and D), flowers (E and F), and developing green siliques (G and H). A, C, E, and G, Accumulation profiles of all the major GIPC classes. B, D, F, and H, Overall amounts of either Hex_n-only GIPCs (Hex) or HexN(Ac)-containing GIPCs [HexN(Ac)]. Data are means of three biological replicates ± sd. Asterisks indicate significant difference compared to that in the wild type (Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.001).

Figure 4.

Complementation of the HexN-GIPC deficiency in Atgint1. LC-MS/MS chromatograms of the dry seed sphingolipidome in Atgint1-1, Atgint1-1 expressing AtGINT1_pro:AtGINT1, AtGINT1_pro:OsGINT1, and AtGINT1_pro:AtGMT1. Total ion counts corresponding to Hex-GIPCs (blue line), HexN-GIPCs (red line), and Ganglioside GM1 (internal standard, black line) are shown.

Figure 5.

De novo synthesis of HexN-GlcA-IPCs in Arabidopsis leaves following constitutive expression of AtGINT1 or OsGINT1. A, LC-MS/MS chromatograms of Hex-GIPC (t18:1-h24:1, left) and HexN-GIPC (t18:1-h24:1, right) are shown. B, Total quantity of HexN- and HexN(Ac)-GIPC in leaves of Arabidopsis expressing 35Spro:AtGINT1 or 35Spro:OsGINT1.

Figure 6.

Microscopy analysis and cell wall composition of wild-type and Atgint1 seeds. A, Ruthenium red staining of imbibed seeds indicating mucilage layer. Bar = 200 µm. B, Scarlet4B staining of imbibed seeds indicating cellulose rays in mucilage. Bar = 100 µm. C, Scanning electron microscopy. Bar = 200 µm. D, TEM. Bar = 4 µm. E, Seed weight. F, Seed fatty acid content. G, Seed protein content. A to D, representative images are shown. E to G, Data are means of 3 or 4 biological replicates ± sd. Asterisks indicate significant difference compared to the wild type (Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.001).

Figure 7.

Cell wall composition of Atgint1. A, Monosaccharide composition of seed AIR following TFA hydrolysis. B, Glc released from TFA-insoluble AIR from leaves, stems, and seeds following Saeman hydrolysis. Data are means of three biological replicates ± sd. No statistically significant difference (Student’s t test, P < 0.05) was observed between wild-type and Atgint1 plants.

Figure 8.

Effect of ABA and NaCl on wild-type and Atgint1 seed germination. A, Seed germination rates in the presence of ABA; seeds were scored for radicle emergence after 3 d. B, Seed germination rates in the presence of NaCl, with seeds scored for radicle emergence after 2 d. Data are means of three biological replicates (>50 seeds per replicate) ± sd. An asterisk indicates statistically significant difference from the wild type (Student’s t test, *P < 0.05, **P < 0.01).

Figure 9.

Seedling phenotype and GIPC content of OsGINT1-targeted CRISPR/Cas9 rice. A, Regenerated T0 shoots harboring T-DNA with two independent CRISPR/Cas9 OsGINT1 target sequences were further cultured on 0.5× Murashige and Skoog media for 2–3 weeks. Some of the regenerated lines showed severely retarded growth (#5–8), which later proved to be seedling lethal. B, The GIPC content for each shoot shown in A as determined by LC-MS/MS. Hex-containing GIPCs were not detected.

Figure 10.

Heterologous complementation of Atgmt1 with GINT1. A, Heterozygous Atgmt1 plants were transformed with CaMV 35S_pro:AtGMT1, 35S_pro:AtGINT1, or 35S_pro:OsGINT1. Bar = 2 cm. T2 seedlings (two independently transformed lines per construct) were analyzed by RT-PCR (B). Sphingolipidomic analyses using LC-MS/MS (C and D). C, Profiles of all the major classes of GIPCs detected in the tissues tested, and D, amounts of either Hex_n-only GIPCs (Hex) or HexN(Ac)-containing GIPCs [HexN(Ac)]. PP2AA3, PROTEIN PHOSPHATASE 2A SUBUNIT A3; −RT, reaction lacking reverse transcriptase enzyme. Data are means of three biological replicates (>50 seeds per replicate) ± sd.