Researchers at EPFL have explained the unpredictable behaviors of biological nanopores, attributing them to internal electrical charges. By engineering variants of the bacterial pore aerolysin, they demonstrated how these pores can mimic brain-like learning. The findings suggest potential for bio-inspired computing applications.
Biological nanopores, tiny molecular holes essential in nature and biotechnology, have long puzzled scientists with their complex behaviors. Pore-forming proteins play key roles in human immune defense and bacterial toxins that puncture cell membranes, allowing precise control of ion and molecule movement. Adapted for uses like DNA sequencing and molecular sensing, these pores sometimes exhibit rectification—where ion flow varies with voltage polarity—and gating, where flow abruptly halts.
A team led by Matteo Dal Peraro and Aleksandra Radenovic at EPFL investigated these phenomena using the bacterial pore aerolysin. They engineered 26 variants by modifying charged amino acids inside the pore, observing ion travel under various conditions. Applying alternating voltage signals, they differentiated rectification, which occurs rapidly, from slower-developing gating. Biophysical models helped reveal the mechanisms.
Rectification arises from charges along the pore's inner surface influencing ion movement, acting like a one-way valve favoring one direction. Gating, however, results from intense ion flow disrupting charge balance, destabilizing the pore's flexible structure and causing temporary blockage until reset. Reversing charge signs allowed control over gating timing, while enhancing pore rigidity eliminated it entirely, underscoring structural flexibility's importance.
These insights enable custom nanopore design to reduce unwanted gating in sensing or exploit it for computing. In a key experiment, the team created a nanopore mimicking synaptic plasticity, "learning" from voltage pulses akin to neural synapses. This points to future ion-based processors harnessing molecular learning. The study appears in Nature Nanotechnology (2025; DOI: 10.1038/s41565-025-02052-6).