Short-chain mono-carboxylates as negative modulators of allosteric transitions in GLIC, and impact of a pre-β5 strand (Loop Ω) double mutation on crotonate, not butyrate effect

The bacterial model GLIC remains one of the best known among pentameric ligand-gated ion channels (pLGICs), regarding their structure. GLIC is activated at low extracellular pH, but no agonist compound is known. Van Renterghemet al. (2023) showed that short-chain di-carboxylates potentiate GLIC activity, with strict dependence on two carboxylate binding pockets, previously characterized by crystallography (Sauguetet al., 2013, Fouratiet al., 2015, 2020). An “in series” model was proposed, with compound binding at the inter-subunit pocket [homologous to the pLGICs orthotopic neutotransmitter binding site], and with involvement of the intra-subunit (or vestibular) pocket in coupling binding to gating.

Here we characterize saturated, mono-carboxylates as negative modulators of GLIC, as previously shown for crotonate (Alqazzazet al., 2016). Butyrate and crotonate have indistinguishable properties regarding negative modulation of WT GLIC. However, a double mutation in the pre-β5 strand (Loop Ω) converts crotonate, as well as caffeate, but not butyrate, into positive modulators. We perform a mutational analysis of residue dependency in the two pockets, with the pre-β5 strand either intact or mutated. We propose that positive modulation requires stringent conditions, with integrity of both pockets, whereas negative modulation is less labile. The vestibular pocket may be involved as an accessory binding site leading to negative, but not positive modulation. We propose that the pre-β5 strand is involved in ligand-elicited modulation of GLIC gating, not in pHo-controlled gating. Possible involvement in Eukaryote pLGICs of regions corresponding to the vestibular pocket and the pre-β5 strand/Loop Ω is discussed.