Ethanolamine-phosphate on the second mannose is a preferential bridge for some GPI-anchored proteins

Structures of liganded glycosylphosphatidylinositol transamidase illuminate GPI-AP biogenesis

Many eukaryotic receptors and enzymes rely on glycosylphosphatidylinositol (GPI) anchors for membrane localization and function. The transmembrane complex GPI-T recognizes diverse proproteins at a signal peptide region that lacks consensus sequence and replaces it with GPI via a transamidation reaction. How GPI-T maintains broad specificity while preventing unintentional cleavage is unclear. Here, substrates- and products-bound human GPI-T structures identify subsite features that enable broad proprotein specificity, inform catalytic mechanism, and reveal a multilevel safeguard mechanism against its promiscuity. In the absence of proproteins, the catalytic site is invaded by a locally stabilized loop. Activation requires energetically unfavorable rearrangements that transform the autoinhibitory loop into crucial catalytic cleft elements. Enzyme-proprotein binding in the transmembrane and luminal domains respectively powers the conformational rearrangement and induces a competent cleft. GPI-T thus integrates various weak specificity regions to form strong selectivity and prevent accidental activation. These findings provide important mechanistic insights into GPI-anchored protein biogenesis.

(Patho)Physiology of Glycosylphosphatidylinositol-Anchored Proteins I: Localization at Plasma Membranes and Extracellular Compartments

Glycosylphosphatidylinositol (GPI)-anchored proteins (APs) are anchored at the outer leaflet of plasma membranes (PMs) of all eukaryotic organisms studied so far by covalent linkage to a highly conserved glycolipid rather than a transmembrane domain. Since their first description, experimental data have been accumulating for the capability of GPI-APs to be released from PMs into the surrounding milieu. It became evident that this release results in distinct arrangements of GPI-APs which are compatible with the aqueous milieu upon loss of their GPI anchor by (proteolytic or lipolytic) cleavage or in the course of shielding of the full-length GPI anchor by incorporation into extracellular vesicles, lipoprotein-like particles and (lyso)phospholipid- and cholesterol-harboring micelle-like complexes or by association with GPI-binding proteins or/and other full-length GPI-APs. In mammalian organisms, the (patho)physiological roles of the released GPI-APs in the extracellular environment, such as blood and tissue cells, depend on the molecular mechanisms of their release as well as the cell types and tissues involved, and are controlled by their removal from circulation. This is accomplished by endocytic uptake by liver cells and/or degradation by GPI-specific phospholipase D in order to bypass potential unwanted effects of the released GPI-APs or their transfer from the releasing donor to acceptor cells (which will be reviewed in a forthcoming manuscript).

Sequential hydrolysis of FAD by ecto-5' nucleotidase CD73 and alkaline phosphatase is required for uptake of vitamin B2 into cells

Extracellular hydrolysis of flavin-adenine dinucleotide (FAD) and flavin mononucleotide (FMN) to riboflavin is thought to be important for cellular uptake of vitamin B₂ because FAD and FMN are hydrophilic and do not pass the plasma membrane. However, it is not clear whether FAD and FMN are hydrolyzed by cell surface enzymes for vitamin B₂ uptake. Here, we show that in human cells, FAD, a major form of vitamin B₂ in plasma, is hydrolyzed by CD73 (also called ecto-5' nucleotidase) to FMN. Then, FMN is hydrolyzed by alkaline phosphatase to riboflavin, which is efficiently imported into cells. We determined that this two-step hydrolysis process is impaired on the surface of glycosylphosphatidylinositol (GPI)-deficient cells due to the lack of these GPI-anchored enzymes. During culture of GPI-deficient cells with FAD or FMN, we found that hydrolysis of these forms of vitamin B₂ was impaired, and intracellular levels of vitamin B₂ were significantly decreased compared with those in GPI-restored cells, leading to decreased formation of vitamin B₂-dependent pyridoxal 5'-phosphate and mitochondrial dysfunction. Collectively, these results suggest that inefficient uptake of vitamin B₂ might account for mitochondrial dysfunction seen in some cases of inherited GPI deficiency.