Nanostructured phase-separated domains—so-called lipid rafts—are widely thought to act as control elements in the plasma membrane. They have been implicated in a wide variety of biological processes, including signal transduction, viral infection, and antibiotic activity. The ability to control and manipulate lipid nanostructures in cell membranes is a potential novel therapeutic approach for a wide variety of diseases. Insulin resistance underlying type-2 diabetes, for instance, has been linked to a pathological association between the insulin receptor protein and membrane domains enriched in the lipid ganglioside GM3. Understanding the biophysical bases of such interactions and developing pharmaceutical approaches to altering them would be greatly aided by the development of simple in vitro systems for studying lipid nanostructure formation and raft/protein interactions.

Simplified model systems for studying lipid biophysics in vitro are typically formed from synthetic or purified natural lipids that spontaneously self-assemble into bilayer structures. Given the proper lipid compositions, such systems display the phase separation behavior thought to underlie the lipid raft phenomenon. This phase separation, however, takes place on the wrong scale: existing biomimetic membrane systems produce separated domains tens of microns wide rather than the physiologically relevant nanoscale lipid domains. Recent work has suggested that this scale discrepancy originates with the mechanical connectivity between the cell membrane and the cytoskeleton. Membrane anchoring to the cytoskeleton results in the formation of membrane compartments between which lipids cannot freely diffuse, limiting the long-range diffusivity of lipid molecules and restricting the growth of phase-separated domains to nanoscale rafts.

We are building lipid-based systems that properly recapitulate biological raft formation by mimicking the mechanical attachments between the lipid bilayer membrane and the cytoskeleton. These systems are based on giant unilammelar vesicles (GUVs) that contain hydrophilic monomer solutions in their interior and lipids with vinyl-modified headgroups in the membrane. Upon initiation of free-radical polymerization, a cross-linked polymer hydrogel is formed and covalently attached to the modified lipids. The hydrogel then serves as a biomimetic cytoskeleton, mechanically restricting diffusion in the bilayer.

Nanoscale lipid behavior is observed using a combination of fluorescence resonance energy transfer techniques and total internal reflection fluorescence microscopy. These techniques allow us to probe supermolecular complexes containing only a few molecules, and to observe early events in raft formation. These observations promise insights on the thermodynamics and kinetics of raft formation.

We are further developing a microfluidic system that facilitates the automated production of GUVs with highly tailored structures, including asymmetric bilayer compositions and precise control over vesicle contents. This system is based on a multiphase droplet flow configuration that allows for the sequential assembly of the bilayer from two independently constituted lipid monolayers. Together with the technology for engineering biomimetic cytoskeletons, this system promises to serve as a flexible platform for studying raft/protein interactions and developing pharmaceuticals that target lipid nanostructures.