Bacteria think like humans do

The researchers engineered a nanopore that mimics synaptic plasticity — it “learns” from repeated voltage pulses, showing basic memory-like behavior similar to neural synapses. 

1. Discovery by EPFL scientists: Researchers from Ecole Polytechnique Fédérale de Lausanne (led by Matteo Dal Peraro and Aleksandra Radenovic) have explained why biological nanopores sometimes behave unpredictably.

2. Focus on aerolysin: The study used the bacterial nanopore aerolysin, a common tool in biotechnology, and created 26 engineered variants by modifying its internal charged amino acids.

3. Two key puzzling behaviors explained:

   - Rectification: Ion flow changes depending on the direction (sign) of the applied voltage.

   - Gating: Sudden decrease or complete stop of ion flow through the pore.

4. Root cause identified: Both rectification and gating are governed by the electrical charges lining the inside (lumen) of the nanopore and how they interact with passing ions.

5. Mechanism of rectification: Internal charges act like a one-way valve, making ion flow easier in one direction than the other.

6. Mechanism of gating: Strong ion flow disrupts the charge balance, causing temporary structural destabilization (collapse) of the flexible pore, which blocks ion passage until it resets.

7. Experimental approach: The team combined experiments, molecular simulations, and theoretical modeling. They used alternating voltage signals to separate fast rectification from slower gating effects.

8. Control through engineering: Reversing the sign of internal charges or increasing pore rigidity could control or completely eliminate gating, showing these behaviors are tunable.

9. Brain-like learning demonstrated: The researchers engineered a nanopore that mimics synaptic plasticity — it “learns” from repeated voltage pulses, showing basic memory-like behavior similar to neural synapses.

10. Future applications: The findings enable the design of smarter, more reliable nanopores for DNA sequencing and sensing, and open the door to bio-inspired computing and ion-based processors that harness molecular “learning.” 

Methodology for the characterization of nonlinearities using time-varying voltage drop across single- and multi-pore membranes.

a, Triangular forcing at 2 Hz, 200 mV, with 6 aerolysin wt nanopores used to extract the single open-pore IV curve. b, IV characteristics of a single aerolysin wt nanopore in the lipid membrane measured with sinusoidal potential (0.1 Hz). Fifty consecutive cycles are summed to retrieve the smooth hysteresis loop. c, Normalized IV characteristics of 26 pores in the same membrane showing the smooth hysteresis loop (0.1 Hz). Pores in the membrane close and open in a voltage-dependent manner, giving rise to the hysteresis loop. Arrows indicate the direction of the sweep. Inset: gating quantification. The solid blue line represents the closed-state probability, p, and the solid black line represents the sinusoidal voltage. The calculation of p is ill-defined around the origin, resulting in a discontinuous line in the probability plot, which did not affect our conclusions (‘Data analysis’ in Methods). The maximum slope indicated by the red lines and dots corresponds to kX for both polarities, where kX is the maximum closing rate. All experiments were conducted in 1 M KCl buffered with 10 mM phosphate to pH 6.2.
Credits: Mayer, S.F., Mitsioni, M.F., Robin, P. et al. Lumen charge governs gated ion transport in β-barrel nanopores. Nat. Nanotechnol. 21, 116–124 (2026). https://doi.org/10.1038/s41565-025-02052-6