Membrane protein mediated bilayer communication in networks of droplet interface bilayers

Abstract

Droplet interface bilayers (DIBs) are model membranes formed between lipid monolayer-encased water droplets in oil. Compared to conventional methods, one of the most unique properties of DIBs is that they can be connected together to generate multi-layered 'tissue-like' networks, however introducing communication pathways between these compartments typically relies on water-soluble pores that are unable to gate. Here, we show that network connectivity can instead be achieved using a water-insoluble membrane protein by successfully reconstituting a chemically activatable mutant of the mechanosensitive channel MscL into a network of DIBs. Moreover, we also show how the small molecule activator can diffuse through an open channel and across the neighbouring droplet to activate MscL present in an adjacent bilayer. This demonstration of membrane protein mediated bilayer communication could prove key toward developing the next generation of responsive bilayer networks capable of defining information flow inside a minimal tissue.

Fig. 1. MscL purification and reconstitution in lipid vesicles.

a SDS–PAGE gel (pre-cast NuPAGE 12% bis-tris protein gel) containing purified protein. Lane 1 contains Novex Sharp Protein Standard Ladder and lanes 2–4 contain 0.27 mg ml⁻¹, 0.11 mg ml⁻¹ and 0.05 mg ml⁻¹ of purified protein, respectively. MscL monomer is indicated at 17 kDa. b Calcein release assay showing the activation of reconstituted MscL by the chemical activator MTSET. MscL is activated after 10 min by the addition of MTSET (red circles), which leads to an increase in fluorescence caused by the diffusion and subsequent dilution of quenched calcein as it diffuses out of the vesicle through the open MscL channel (red circles). No increase in fluorescence is observed when MTSET is replaced with pure buffer solution (black squares). Data were normalized to the value at time = 0 min. Errors are shown as the standard deviation of four measurements.

Fig. 2. Assembly of droplet interface bilayers.

Water droplets comprised of lipid vesicles are placed onto the agar-coated tips of silver/silver chloride electrodes and lowered into a well of oil. After a brief incubation period, a highly packed lipid monolayer assembles at the water–oil interface and a lipid bilayer is formed when the droplets are manipulated into contact (as highlighted in the zoom).

Fig. 3. MscL activity in droplet interface bilayers.

DIBs formed with reconstituted protein and MTSET give rise to discrete channel activity. A schematic of the experimental set-up is provided together with an equivalent electrical circuit, where C_m depicts the membrane capacitance and G_MscL the channel conductance. The indicated regions of a representative 30 min recording obtained at 100 mV are magnified in the 3 s segments shown in a–c. The recordings show rapid ~20 pA opening and closing events in a, more stable ~14 pA events in b and two clearly defined ~40 pA closing events in c. The histograms supplied with each trace reflect the different open states observed.

Fig. 4. MscL activity in networks of droplet interface bilayers.

Three droplet DIB networks were formed with reconstituted protein in the first and last droplet and MTSET in the middle droplet. A representative 20 min bilayer current recording shows multiple gating events including successive openings. A schematic of the experimental set-up is provided together with an equivalent electrical circuit, which highlights that the droplets are connected in series, leading to a reduction in the total capacitance and an increase in the effective resistance, resulting in a reduction in the bilayer current. The indicated regions of the trace are magnified in the 50 s segments shown in a–c. The recording shows an ~20 pA opening and closing event that lasts ca. 12 s in a, an ~30 pA opening that leads to multiple opening events in b and an ~20 pA third-level burst of activity c. The histograms supplied with each trace reflect the different open states recorded.

Fig. 5. MscL-mediated bilayer communication in DIB networks.

Three droplet DIB networks were formed with MTSET in the first droplet and MscL vesicles in the second and third as highlighted in the schematic. As lipid bilayers are impermeable to the free diffusion of MTSET, the only route for network activation is for passive diffusion through the open channel and across the adjacent droplet to activate MscL in the second membrane. The time taken for this to take place is reflected in the equivalent circuit through the addition of an extra resistor labelled G_Diff (this is not intended to indicate a further reduction in the bilayer current). A representative 30-min bilayer current recording shows a short burst of activity at ~6 min and more significant activity from 20 min onward. The indicated regions of the trace are magnified in the 1 s segments shown in a–c. The zooms shows a brief ~7 pA burst of activity in a, a similar ~7 pA burst of activity in b and two brief ~24 pA opening and closing events in c that lead to the channel being held open. The histograms supplied with each trace reflect the different open states observed in the recordings and emphasize the low number of open events in a, b.