Fig. 4: Pore lipids can reside stably within the pore.

a Time traces from backbone unrestrained (left) and backbone restrained (right) simulations in 9:1 POPC:cholesterol membranes, indicating the number of lipids that remain bound upright within the pore (salmon; simulations sampled every 1 ns) and the measured radius of the pore (Å) at the constriction site (blue; simulations sampled every 10 ns, with missing time points reflecting structural snapshots that lacked plausible pores, as described in the Methods). The radius for a bare Cl^– ion (1.8 Å) is indicated with a gray dashed line. Individual simulation replicates (#1–6) are plotted successively from top to bottom. b Side-views of the pore immediately after the end of equilibration (left, 1 ns) and near the end of the total simulation (900 ns) for an unrestrained simulation where lipids remained stably bound (center-left), an unrestrained simulation where lipids dissociated (center-right), and a restrained simulation (right). LRRC8 subunits, white; lipids, salmon; water surface, blue. To the left of each 900-ns snapshot are pore profiles determined across simulation frames. Pore radii (Å) are plotted as a function of z-position within the pore (Å), with pore lipids included (solid lines colored from blue to green with increasing simulation time) or excluded (gray lines) from the pore radius calculation. Pore profiles were calculated every 10 ns, as described above.