Lasse Klausen2 1 Thomas Fuhs3 2 Xiaojun Han3 Mingdong Dong2 1

2, Aarhus University, Aarhus, , Denmark
1, Stanford University, Stanford, California, United States
3, Harbin Institute of Technology, Harbin, , China

Biological membranes contain a sizeable fraction of charged lipids and proteins that separate into domains essential to cell functions. Phase separated lipid bilayers are routinely prepared in vitro, and the average surface charge or surface potential is characterized by methods such as electrophoresis, streaming potential or electroacoustics. The local charge of individual domains has however not previously been quantitatively described. This stands in stark contrast to a widespread focus on zeta potential values, especially for drug delivery.

In this work we prepare phase separated lipid bilayers by mixing an unsaturated lipid of distinct charge (DOPG, DOPC or DOTAP) with a zwitterionic saturated lipid (DPPC). Electrostatic repulsion between lipids of same charge favors a uniform mixing of the two lipids, while lipids with similar tail groups are attracted by an entropically favored packing configuration. The two lipids are mixed at a temperature above TM and slowly cooled, leading to a demixing and formation of micrometer sized domains with distinct charge. We then characterize these domains using the recently developed method Quantitative Surface Charge Microscopy (QSCM). QSCM is based on scanning ion conductance microscopy (SICM), and uses fluctuations in ionic current passing through a nanopipette as it is placed near the lipid bilayer.

Lipid domains are clearly distinguishable from both topography and charge maps in all three model membranes. We furthermore find that the standard sample preparation, where a mixture of lipids with different TM is slowly cooled to induce mixing, will not produce a complete division between the lipids. Instead we estimate that at least 30% of disordered domains in DOPG:DPPC and DOTAP:DPPC will be composed of DPPC. This ratio could present a limit for the formation of charged domains in lipid membranes. In summary we demonstrate that QSCM is capable of distinguishing domains of charged and uncharged lipids with spatial resolution on the nanoscale, and we show that the precision of the measured charge is good enough to deduce the mixing ratios between charged and uncharged lipids.