AtPIG-S, a predicted Glycosylphosphatidylinositol Transamidase Subunit, is critical for pollen tube growth in Arabidopsis

Abstract

AbstractBackgroundGlycosylphosphatidylinositol (GPI) addition is one of the several post-translational modifications to proteins that increase their affinity for membranes. In eukaryotes, the GPI transamidase complex (GPI-T) catalyzes the attachment of pre-assembled GPI anchors to GPI-anchored proteins (GAPs) through a transamidation reaction. A mutation in AtGPI8 (gpi8-2), the putative catalytic subunit of GPI-T in Arabidopsis, is transmitted normally through the female gametophyte (FG), indicating the FG tolerates loss of GPI transamidation. In contrast, gpi8-2 almost completely abolishes male gametophyte (MG) function. Still, the unexpected finding that gpi8-2 FGs function normally requires further investigation. Additionally, specific developmental defects in the MG caused by loss of GPI transamidation remain poorly characterized.ResultsHere we investigated the effect of loss of AtPIG-S, another GPI-T subunit, in both gametophytes. Like gpi8-2, we showed that a mutation in AtPIG-S (pigs-1) disrupted synergid localization of LORELEI (LRE), a putative GAP critical for pollen tube reception by the FG, yet is transmitted normally through the FG. Conversely, pigs-1 severely impaired male gametophyte (MG) function during pollen tube emergence and growth in the pistil. A pPIGS:PIGS-GFP transgene complemented these MG defects and enabled generation of pigs-1/pigs-1 seedlings, but seemingly failed to rescue the function of AtPIG-S in the sporophyte, as pigs-1/pigs-1, pPIGS:PIGS-GFP seedlings died soon after germination.ConclusionsCharacterization of pigs-1 provided further evidence that the FG tolerates loss of GPI transamidation more than the MG and that the MG compared to the FG may be a better haploid system to study the role of GPI-anchoring. pigs-1 pollen develops normally and thus represent a tool in which GPI anchor biosynthesis and transamidation of GAPs have been uncoupled, offering a potential way to study free GPI in plant development. While previously reported male fertility defects of GPI biosynthesis mutants could have been due either to loss of GPI or GAPs lacking the GPI anchor, our results clarified that the loss of mature GAPs underlie male fertility defects of GPI-deficient pollen grains, as pigs-1 is defective only in the downstream transamidation step. Our study also provided further evidence that GPI transamidation is essential in seedling development.