(B-74) Synapse-Inspired Biomolecular Memcapacitors

Brain-inspired computing offers an efficient alternative to conventional electronics by collocating processing and memory [1]. In this work we build a simple model synapse from biomolecular elements using droplet interface bilayers (DIBs). DIBs form lipid membranes at the intersection of lipid-coated aqueous droplets in an oil reservoir [2]. These lipid membranes may be functionalized by including pore-forming peptides in the solutions, permeabilizing the surrounding membranes. In this project, a three-droplet network was formed with differing concentrations of alamethicin, a voltage-dependent pore-forming peptide. Positive and negative voltages were applied to the system to manipulate the gating characteristics of both membranes and to trap charge within the central droplet. This trapped charge produces electrowetting of the droplet interfaces, providing long-lasting memcapacitance.

Lipid solutions were prepared with 1,2-diphytanoyl-sn-glycero-3-phosphocholine (DPhPC) lipids dispersed in a buffer solution with varying concentrations of alamethicin.  These solutions were deposited directly on silver/silver-chloride electrodes within an oil reservoir and manipulated into contact to form networks of lipid membranes. The electrodes were connected to a patch-clamp amplifier (Axopatch 200B) which was used to measure the network characteristics. 

The properties of alamethicin at varying concentrations were tested by forming a single membrane between two droplets containing alamethicin.  A voltage sweep was performed and the current across the membrane was recorded to produce expected thresholds for alamethicin insertion. Networks of three droplets were then formed with varying distributions of alamethicin, creating an artificial synapse droplet between two membranes.  Varying pulse magnitudes, widths, and sequences were applied to the droplets and the change in properties were measured.

Results and Discussion: Varying the concentration of alamethicin altered the threshold voltage for membrane gating, establishing distinct low and high barriers within the two membranes bordering the artificial synapse. Applying positive pulses charged the central droplet, leading to long-lasting changes in the measured network capacitance. Subsequent negative voltage pulses removed the excess charge from the artificial synapse, returning the network capacitance to equilibrium.

Conclusions: This research produced a novel artificial synapse from two lipid membranes in series with varying concentrations of pore-forming peptides. The material exhibits long-term potentiation, multiple memory states, spike timing-dependent plasticity, and can program and erase memory states through series of positive and negative input pulses.  Future work will optimize and simulate the phenomena.

The authors acknowledge the support of the Natural Science Foundation under the project title REU Site: Interdisciplinary Research Experiences in Nanotechnology and Biomedicine, award number EEC-1950581.

1.         Makhoul-Mansour, M.M., J.J. Maraj, and S.A. Sarles, Bioderived materials for stimuli-responsive, adaptive, and neuromorphic systems: A perspective. Journal of Composite Materials, 2023. 57(4): p. 659-678.

2.         Hwang, W.L., M. Chen, B. Cronin, M.A. Holden, and H. Bayley, Asymmetric droplet interface bilayers. J Am Chem Soc, 2008. 130(18): p. 5878-9.