H28 forms state-dependent stabilizing interactions with the GAS belt and T26 with TM1 and TM2b. (A) Side and bottom views of hASIC1a in the predicted resting (upper panel) and open (lower panel) conformations, highlighting interaction networks formed by residues H28, T26 (HG loop), and the lower pore. Three TM1 helices (deep teal/cyan/pale cyan), two TM2 helices (cyan/pale cyan), and one reentrant loop (light pink) are shown. (B) Summary of normalized current, τ_d, and recovery ratio for H28 substitutions. All mutants produce significantly smaller currents than WT (P < 0.00001), have faster desensitization (P < 0.0001), and partial recovery (P < 0.0001). (C) Representative traces of H28R, H28Y, and H28W. Data are presented as mean ± SD of three independent experiments for at least five Xenopus oocytes (n = 5–10) for each construct tested. All oocytes were incubated at least 30 s (30–40 s) in preconditioning buffer (pH 7.4) for channel recovery before the next activation. Recovery ratio (expressed in percent) was calculated as I_{acti(n + 1)}/I_acti(n). Individual measurements and statistical analysis are shown in Data S2.