Search

In cells, membranes are highly bent in many places in order to fulfill different functions. This is the case of membrane trafficking, which is accompanied by the formation of tubes or vesicles with diameter below 100 nm that transport proteins and lipids throughout the cell. Moreover, during vesicle budding, curvature at the neck becomes even higher before final scission. Structures at the cell periphery, such as lamellipodia’ edge or filopodia used for migration and environment sensing, also exhibit high membrane curvatures. Many peripheral or integral proteins have an intrinsic shape that produces membrane bending or can deform membrane upon binding. In membrane physics, this property is described by a phenomenological parameter, the protein spontaneous curvature Cp, which represents the capability of the protein to produce membrane spontaneous bending. It is expected that proteins that deform membranes can reciprocally be enriched in curved areas. We have addressed the question of the role of membrane curvature in protein sorting both experimentally and theoretically. We have tested the sorting hypothesis using in vitro membrane system (GUV) and different proteins with positive Cp (BAR-domain), negative Cp (I-BAR-domain) and with integral proteins reconstituted in the GUVs (a potassium channel KvAP). Membrane nanotubes with a controlled diameter (15-500 nm) are pulled out of the GUV with optical tweezers; the relative protein sorting is measured as a function of curvature (inverse of tube radius) from fluorescence measurements with a confocal microscope. We show that our mechanical model based on spontaneous curvature induction is in good agreement with these experiments. All together, this demonstrates that membrane shape is an important parameter that might drive the lateral distribution proteins in membranes, independently of more specific sorting mechanisms.

Biophysical characterization of voltage dependent proteins in-vivo is challenging because of the complex feedback and active regulation in living systems. One approach to overcome these difficulties is to reconstitute the proteins in bio-mimetic model systems such as Giant Unilamellar Vesicles (GUVs), solvent-free, cell-sized proteo-liposomes in which membrane composition, tension and curvature can be readily controlled. The membrane potential of a small patch of the GUV membrane can be controlled by aspirating it into a micropipette (“GUV-attached” patch-clamp), thereby allowing measurements of protein function [1].However, electrophysiological and fluorescence studies of the structural rearrangements mediating protein function would clearly benefit from the ability to control the membrane potential of entire, intact GUVs. For cells, electrical access to the cell interior can be obtained either by inserting a micro-electrode or breaking a “cell-attached” membrane patch to go “whole-cell”. Unfortunately the lack of a cytoskeleton makes GUVs far more sensitive to mechanical stresses than cells, which makes the “whole-cell” configuration quite challenging.Examination of failed “whole-cell” GUV experiments suggested key roles for micropipette pressure and adhesion between the GUV membrane and micropipette, and by controlling these we were able to maintain the “whole-GUV” configuration for experimentally useful periods (tens of minutes). We will describe efforts to further optimize the “whole-GUV” technique, and to use it to study the effects of protein-lipid interactions and protein curvature preferences on the function of KvAP, a voltage-gated potassium channel.[1] S. Aimon et al. Functional Reconstitution of a Voltage-Gated Potassium Channel in Giant Unilamellar Vesicles (2011), PLOSone 6:e25529

P2X receptors are trimeric cation channels that are activated by extracellular ATP. These purinergic receptors have been reported to undergo pore dilation following activation by high concentrations of ATP, a phenomenon thought to mediate apoptosis in the immune system. Here we demonstrate that the widely reported slow time-dependent increase in the relative permeability of NMDG to Na, thought to reflect pore dilation, results from a gradual inhibition of small-cation-selective channels following depletion of intracellular alkali ions. Moreover, we find that P2X receptors enter both small-cation-selective open states and large-cation-permeable open states within milliseconds of ATP application. Taken together, our results demonstrate that P2X receptors can rapidly enter a large-cation-permeable open state without requiring either high ATP concentrations or high channel density, indicating that the ability of ATP to permeabilize cells can occur under physiologically realistic conditions.

The flow of ions through channels and transporters depends sensitively on the concentration of the ionic species and voltage at the membrane. In the widely used patch-clamp technique, voltage and current are measured using electrodes in the pipette and bath that are distant from the membrane. Although the importance of voltage differences between the membrane and electrodes is widely appreciated, concentration gradients arising from the flow of ions are often neglected. In this study, we modeled changes in voltage and ion concentrations during patch-clamp experiments using the Nernst-Plank equation and we derived simple formulas for estimating the timescale and extent of ion accumulation and depletion. For excised patch experiments, ions crossing the membrane can directly diffuse into or out of the patch pipette and ion concentrations stabilize on the millisecond timescale after a change in membrane current. In contrast, in whole-cell experiments the cytosol acts as a reservoir and ion concentrations change on the timescale of seconds. In either configuration, ion accumulation or depletion at steady-state depends primarily on the electrode access resistances and currents carried by each ionic species. As a practical illustration, simulations were performed for bi-ionic protocols previously used to characterize the dynamic ionic selectivity of P2X and TRPV channels. Importantly, even when the net current was small and the membrane voltage effectively clamped, ion accumulation and depletion could cause significant, time-dependent changes in current resembling reported examples of “pore dilation”. Thus, limitations for clamping ion concentrations should be considered when performing and interpreting patch-clamp experiments.