Zongcheng Lin, Fei Xie, Marina Triviño, Maurice Bosch, Vernonica E. Franklin-Tong, Moritz K. Nowack,

Current Biology; March 21, 2022; DOI:https://doi.org/10.1016/j.cub.2022.02.072

Glycosylphosphatidylinositol-anchored proteins (GPI-APs) are tethered to the outer leaflet of the plasma membrane where they function as key regulators of a plethora of biological processes in eukaryotes. Self-incompatibility (SI) plays a pivotal role regulating fertilization in higher plants through recognition and rejection of “self” pollen. Here, we used Arabidopsis thaliana lines that were engineered to be self-incompatible by expression of Papaver rhoeas SI determinants for an SI suppressor screen. We identify HLD1/AtPGAP1, an ortholog of the human GPI-inositol deacylase PGAP1, as a critical component required for the SI response. Besides a delay in flowering time, no developmental defects were observed in HLD1/AtPGAP1 knockout plants, but SI was completely abolished. We demonstrate that HLD1/AtPGAP1 functions as a GPI-inositol deacylase and that this GPI-remodeling activity is essential for SI. Using GFP-SKU5 as a representative GPI-AP, we show that the HLD1/AtPGAP1 mutation does not affect GPI-AP production and targeting but affects their cleavage and release from membranes in vivo. Our data not only implicate GPI-APs in SI, providing new directions to investigate SI mechanisms, but also identify a key functional role for GPI-AP remodeling by inositol deacylation in planta.

See: https://www.cell.com/current-biology/fulltext/S0960-9822(22)00342-6?_returnURL=https%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS0960982222003426%3Fshowall%3Dtrue

Figure 3: AtPGAP1 is a GPI-inositol deacylase, and this function is required for SI

(A and B) A phylogenetic tree of 631 predicted PGAP1 protein homologs from >300 eukaryotic species identified two major phylogenetic clades in plants; two putative Arabidopsis homologs AtPGAP1 and At5g17670 classified into different clades (red triangles). The red star indicates a genome duplication event. Detailed information on the 631 PGAP1-like domain-containing proteins used for this phylogeny tree is contained in Table S6.

(C) Upper panel: cartoon of AtPGAP1 protein secondary structure. The PGAP1-like domain is indicated by the gray box. Black boxes indicate transmembrane domains. Middle panel: amino acid sequence alignment around the conserved lipase motif in the PGAP1-like domain in several higher plants, human, and yeast. The gray box indicates the lipase motif with the catalytic site serine 218 indicated (star). Bottom panel: An amino acid sequence logo of the lipase motif of the PGAP1-like domain shows that it is conserved across eukaryotic kingdoms.

(D) PI-PLC assays used pelleted membrane proteins (P) separated from soluble (S) proteins after centrifugation. GFP-SKU5 (∼120 kDa) was enriched in the pellet (P) fractions in both wild type (WT) and atpgap1 mutant in the mock (buffer control) treatment. After PI-PLC treatment, GFP-SKU5 was present in the soluble (S) fraction in WT samples, demonstrating cleavage by PI-PLC, whereas GFP-SKU5 from the atpgap1 mutants was found in the pellet, demonstrating that in the atpgap1 mutant this GPI-AP was resistant to PI-PLC treatment. This supports the idea that in the atpgap1 mutant a persistent inositol-linked acyl chain makes GPI-APs insensitive to cleavage by PI-PLC. Aquaporin (∼26 kDa) is shown as a control for membrane extraction; loading control is stain-free signals of PROTEAN TGX stain-free precast gels (Bio-Rad).

(E) Hld1 mutants were transformed with AtPGAP1_pro:mCherry-cAtPGAP1 or AtPGAP1_pro:mCherry-cAtPGAP1(S218A) cloned in the pFAST-Green plasmid vector backbone, which expresses GFP in the seeds. Therefore, GFP signals report transgene transmission in the seeds in the F1 generation. Representative images of segregating seeds resulted from pollinations by pollen from T1 heterozygous plants onto self-stigmas (upper panel), At-PrsS₁ stigmas (middle panel) or Col-0 stigmas (lower panel). Numbers indicate the ratio of GFP-positive seeds (number of GFP seeds/total seeds). Bar, 5 mm. Unlike the WT HLD1 that rescued the hld1 phenotype and restored SI, the inositol deacylase-defective S218A HLD1 allele did not, therefore resulting in normal Mendelian segregation of GFP signals in the T2 seeds.