A molecular basis for synexin-driven, calcium-dependent membrane fusion.

Abstract

Membranes of secretory vesicles fuse with each other and with plasma membranes during exocytosis in many different cell types. The probable role of calcium in the process is now widely accepted, and it is possible that at least one cytosolic mediator of calcium action is synexin. Synexin is a 47,000 Mr calcium-binding protein, initially discovered in the bovine adrenal medulla, which binds to granule membranes and to inner aspects of chromaffin cell plasma membranes. Synexin causes chromaffin granules to aggregate, and such aggregates can be caused to fuse in the additional presence of arachidonic acid. Synexin also mediates the direct fusion of liposomes and chromaffin granule ghosts. To understand better the mechanisms of membrane fusion promoted by synexin we have attempted to define the primary sequence of the protein. Our initial efforts were directed towards purification of bovine synexin in sufficient amounts to allow us to sequence tryptic peptides. However, as the project progressed we also directed our attention to human synexin, preparing peptides from this protein as well. From analysis of bovine peptides we learned that the synexin molecule might be closely related to a class of proteins including lipocortin I, calpactin (p36), endonexin II, protein II and calelectrin 67K. Complete analysis of a human synexin cDNA clone revealed strong homology with bovine synexin. The analysis also showed that synexin contained a unique, long, highly hydrophobic N-terminal leader sequence followed by a characteristic four-fold repeat homologous with those found in other members of the synexin gene family. The highly hydrophobic character of synexin seems consistent with information previously obtained that synexin is able to insert directly into the interior of bilayers prepared not only from purified phosphatidylserine but also from biological membranes. The evidence for such insertions is a dramatic increase in the capacitance of the membrane, formed at the tip of a patch pipette, when calcium-activated synexin is applied to the bilayer. Additional evidence is the fact that synexin also forms calcium-selective channels when the protein is applied to the cytosolic aspect of the plasmalemma when that side is also exposed to calcium at sub-millimolar concentrations. Thus, the synexin molecule not only enters the membrane, but also spans it. From these and other data we have developed the concept that the fusion process may involve synexin forming a 'hydrophobic bridge' between two fusing membranes. Lipid movement across this bridge may then be the material basis for final fusion.(ABSTRACT TRUNCATED AT 400 WORDS)