PEG is used to mediate cell-cell fusion in the production of somatic cell hybrids and in the fusion injection of macromolecules into cultured cells from erythrocytes or liposomes. The detailed molecular mechanism of PEG-mediated cell membrane fusion is not known, nor is the mechanism by which fusion proteins induce cell membrane fusion. The long-term goal of this project is to define the lipid molecule rearrangements necessary for cell membrane fusion. The immediate focus is on understanding the molecular details of the fusion of model lipid membranes as mediated by poly(ethylene glycol) [PEG]. The ultimate goals are to advance PEG-mediated cell fusion technologies and to provide insight into the mechanism(s) of such natural cell fusion processes as endocytosis, exocytotic excretion, protein sorting, and viral budding and infection.
Our work has shown that addition of PEG forces model membranes into close contact but will not cause them to fuse. We have shown that disrupted lipid molecular packing in the contacting monolayers of PEG-aggregated membranes is both necessary and sufficient for PEG-mediated fusion. Now that we have defined the structural perturbation that triggers fusion, we have begun to define the sequence and kinetics of molecular events leading to fusion. A recent paper (Lee & Lentz, 1997a) established the time course of PEG-mediated model membrane fusion. This bears remarkable similarity to the sequence of events observed by electo-physiological and fluorescence microscopic methods in influenza virus hemaglutinen (HA)-mediated fusion and secretory granule fusion. However, many more molecular details can be defined in model membrane experiments than can be defined in experiments on bio-membranes.
We are trying now to define the molecular details of the fusion process
as they occur during fusion of PEG-aggregated model
membranes and to compare these details with what is known about biomembrane
fusion. Our hypothesis is that these two
processes share basic molecular mechanisms: fusion occurs by a three-step
mechanism between two curvature-stressed membranes brought into close
contact at the points of maximal curvature. We have recently
obtained evidence to support this hypothesis by showing that the activation
energies of the first and last steps of biomembrane and model membrane
are the same (Lee & Lentz, 1997b). We are currently trying to
define the lipid structural rearrangements that occur during fusion and
to determine how different lipids or membrane structural perturbations
alter the fusion process. One aspect of this part of our work is
asking how the fusion process resolves and creates membrane stress and
lipid rearrangements so that one stage of the process can move smoothly
to the next stage. We are also very interested in the role that membrane
curvature stress plays in fusion and are exploring ways to produce uniform
vesicles preparations with different degrees of curvature stress.
We want also to know how lipid composition alters the rate and extent of
PEG-mediated vesicle fusion, not only to prepare highly fusogenic vesicles
but also to compare more closely to natural membranes that must undergo
fusion.
Since our results suggest a reasonable model for
cellular membrane fusion, we plan to determine how peptide fragments
from the fusion proteins of lipid-sheathed viruses (influenza and
human immuno-deficiency
viruses) disrupt bilayers, and how this enhances membrane fusion.
We aim to determine which, if any, of the three steps in the fusion process
is enhanced by fusion peptide. Our current hypothesis is that the viral
fusion peptide will not induce fusion by favoring the first step of the
process, but may enhance fusion by affecting the second or third step in
the process.
Our current views of membrane fusion are declared in a review.