Golgi vesicle formation: Flipping coincidence

Lipidomics Gateway (25 November 2009) [doi:10.1038/lipidmaps.2009.32]

Golgi flippase activity and transport vesicle budding are regulated by coincident binding of phosphatidylinositol-4-phosphate and an ArfGEF to the Saccharomyces cerevisiae ATPase Drs2p.

Model for the stimulation of Drs2p flippase activity by PI(4)P and ArfGEF. For full figure and legend, see Supplementary Information, Nat. Cell Biol. doi:10.1038/ncb1989

Phosphatidylinositol-4-phosphate, PI(4)P, recruits adaptor proteins to trans-Golgi network (TGN) membranes to facilitate the budding of transport vesicles. In S. cerevisiae the PI(4)P needed for TGN vesicle formation is generated by the phosphatidylinositol-4-kinase Pik1p. Also required is Drs2p, an ATPase with phospholipid translocase (flippase) activity that is proposed to function in the same pathway as Pik1p. In Nature Cell Biology Todd Graham and colleagues now show that Drs2p activity is stimulated by PI(4)P binding, and suggest that this, with binding of an ArfGEF, forms a coincidence detection system that regulates vesicle formation.

To characterize the relationship between Pik1p and Drs2p, the authors first showed that Drs2p does not act upstream of Pik1p, as phosphoinositide levels and localization are both unaffected by a lack of drs2. Next, the effect of PI(4)P on Drs2p flippase activity was investigated using a short-chain, fluorescently tagged phosphatidylserine probe. This is incorporated into the outer leaflet of purified membranes, and a proportion gradually diffuses, by spontaneous flip-flop, to the inner leaflet from where it cannot be back-extracted onto bovine serum albumin. Flippase activity increases the rate of flip-flop in an ATP-dependent manner. The authors purified TGN membranes from cells carrying temperature-sensitive alleles of pik1 and drs2, and demonstrated that Drs2p flippase activity is primarily stimulated by PI(4)P produced by Pik1p.

The carboxy-terminal tail (C-tail) of Drs2p is required for its function in Golgi trafficking, and Graham and colleagues observed a region of homology with a phosphoinositide binding domain from another protein. This region includes a patch of basic residues so they purified C-tail constructs where this patch was intact, deleted, or lacking basic residues. The wild-type C-tail preferentially bound immobilized PI(4)P, and this was disrupted by the mutations. In vivo using full-length constructs, only the wild-type DRS2 gene was fully able to rescue drs2-deleted cells from an associated growth defect. The basic patch was not required for stability or TGN localization of Drs2p, but its mutation strongly abrogated flippase activity.

The Drs2p basic patch overlaps with a previously-mapped binding site for the ArfGEF Gea2p. TGN membranes were purified from a double mutant with temperature-sensitive pik1 and deleted gea2. Whereas PI(4)P stimulated flippase activity in these membranes, Gea2p had only a marginal effect. However, adding both molecules produced a synergistic activation of Drs2p. In pik1 mutant cells, with reduced PI(4)P, mutation of the Drs2p PI(4)P-binding basic patch was not lethal, but loss of the interaction between Drs2p and Gea2p could not be tolerated. These data indicate that Drs2p flippase activity at the TGN, required for vesicular traffic, is strongly stimulated by PI(4)P and ArfGEF binding to the C-tail, and that coincidence of both activators at the TGN may be required for Drs2p activity.

The order of these activating interactions at the TGN and the structural basis for stimulation of Drs2p flippase activity are not yet known. Nevertheless, this work identifies Drs2p as an effector of phosphoinositides, describes the first function for an ArfGEF other than activation of Arf, and increases understanding of the vesicle budding machinery of the TGN.

Emma Leah