Structural insights enable drug discovery for the neuronal NBCn2 carbonate transporter

Yang et al., Nature Communications, 2026, set out to describe the neuronal carbonate transporter NBCn2 at atomic resolution and to build the first rationally designed compounds that block it. Cryo-electron microscopy gave three structures, one with no substrate bound, one with substrate bound at 2.42 Angstrom, and one with the lead inhibitor bound. From those structures the group docked a virtual library of 2.4 million molecules, bought 37 of the top hits, and worked the best one into a compound series.

The biology arm is where AUM BioTech comes in. To ask what the transporter does in real neurons rather than in an over-expressing cell line, the group prepared a neuron and glia co-culture from embryonic day 16 C57BL/6J mouse embryos and plated it onto multi-electrode arrays. An AUMsilence sdASO against NBCn2 and a matched AUMscramble control oligo were diluted into the culture medium and layered straight over the cells at 1 μM, on day 14 and again on day 17. No transfection reagent appears anywhere in that protocol. By day 19, quantitative PCR showed NBCn2 messenger RNA down by approximately 50%.

The result the group did not expect is the one worth reading twice. Silencing NBCn2 raised the two related transporters in the same culture, NBCn1 by about 50% and NDCBE by about 30%, and spontaneous spiking on the array was unchanged. The authors read that as the culture compensating for the transporter it had lost. When the knockdown was combined with 100 μM of the new inhibitor, spike counts fell further than with the inhibitor alone at 30 minutes, and that added effect was gone by 24 hours. The knockdown therefore did two jobs in this paper: it showed that a single transporter can be removed from a primary neuronal culture without stopping it, and it exposed the compensation that would otherwise have been invisible.