Passive and active ion permeation through lipid bilayers investigated by second harmonic water imaging

Lipid membranes are complex and dynamic systems which are known to mediate signaling processes between cells and their environment. To do this multiple ion channels and pumps are involved in controlling the in- and out-flux of various ions (K+, Na+, Mg2+, Ca2+, etc.) and help maintain and regulate a concentration gradient of ions across the membrane. Additionally, membranes are semi-permeable for some species including water and even ions. However, molecular level information about passive and active transport of ions across lipid membranes is either missing or incomplete and ignores membrane hydration without which a membrane would not self-assemble.In this project, we use high-throughput wide-field second harmonic (SH) microscopy to learn about the molecular level structure of membrane interfaces during those processes by imaging the non-resonant response of interfacial water. We show that this technique is extremely sensitive to the transmembrane distribution of interfacial ions which disturb membrane hydration and thus change SH contrast. We apply this technique to probe the molecular structure(orientational order of water, strength of ion binding) at the interfaces of freestanding lipid bilayers and giant unilamellar vesicles (GUVs). We start with SH imaging of operational voltage-gated alamethicin ion channels in freestanding lipid membranes surrounded by an aqueous solution on either side. We observe a change in SH intensity upon channel activation that is traced back to a change in the orientational distribution of water molecules caused by transported ions. Of the transported ions, a fraction of 10-4 arrives at the hydrated membrane interface, leading to interfacial electrostatic changes on the time scale of a second. Later on, in order to improve SH contrast from low-emitting samples (such as interfacial water) and thus be able to study even more fundamental and delicate surface processes, we develop and build a femtosecond optical parametric amplifier. This laser system generates a tunable output in the range from 670 nm to 1000 nm at a 1 MHz repetition rate with a minimum pulse duration of 23 fs at 900 nm. Comparing our current femtosecond laser system with the new one at equal fluences we experimentally observe a 30 times improvement in SH contrast of low-emitting samples.Then, we present a new approach of label-free second harmonic imaging of GUV hydration which directly reports on the cross-membrane distribution of divalent cations bound to negatively charged lipid headgroups. We use this property to study passive transport of Ca2+ ions through lipid membranes. By varying the hydrophobic core of the bilayer, we observe Ca2+ translocation for mono-unsaturated (DOPC:DOPA) membranes which is reduced upon adding cholesterol. A complete inhibition of translocation is observed for branched (DPhPC:DPhPA) and poly-unsaturated (SLPC:SLPA) lipid membranes. In order to gain more insight into the mechanism of divalent cation transport, we perform a series of experiments with different ions (Mg2+, Ba2+, Ca2+and Cu2+). Surprisingly, we observe translocation of Ba2+, Ca2+ and Cu2+ ions through unsaturated DOPC:DOPA membranes which fully neutralize the inner membrane leaflet. Translocation time increases in the order Cu2+< Ca2+< Ba2+, while Mg2+ ions do not show any permeation. The observed trend in passive transport correlates with a trend in ion binding strength measured on impermeable GUVs with the same headgroup composition.