I use negative-stain EM to characterize membrane remodeling activity of the protein. The assumption is that protein pulls tubules out of PIP2-containing vesicles. To my great surprise tubulated vesicles present in the liposome samples prior to protein addition! After protein addition tubulation seems to be more abundant but there is no qualitative difference with the liposome-only specimen.
Liposomes contain (%mol) POPC 80: POPE 20: brain PIP2:3. Prepared by 6 freeze-thaw cycles between acetone-dry ice and 70C water bath followed by 21 extrusion through 400nm carbon membrane at 65C. For EM carbon-coated formvar grids were glow-discharged, placed on the top of 30 ul drop of 0.2 mg/ml liposome suspension ( sample was preincubated at 30c for 5 min) for 5 minutes, washed in water and stained with 1% UA.