Fig. 3: Designed receptor stability and function strongly depend on the density of solvent-mediated interactions at switchable TMH interfaces.
a, Constitutive activity (as a percentage of maximal agonist-induced WT A2AR activity) as a function of the calculated conformational energy difference between the active and inactive states relative to A2AR. The reported energy difference was calculated from the average of the ten lowest-energy inactive- and active-state structures (Supplementary Table 3). The best linear fit to the data is shown as a dotted line (correlation coefficient r = 0.91). b, Measured constitutive activity of the designed receptor variants (as a percentage of adenosine-induced A2AR activity) as a function of the number of solvent-mediated interactions at static–switchable TMH interfaces. c,d, Apparent stability of Hyd_low (c) and Hyd_high (d) receptors compared with A2AR, measured as the functional half-life of adenosine-bound receptor variants incubated at 37 °C. The values for all of the designs are given in Supplementary Table 4. In b–d, the data are presented as the mean ± s.e.m. for n = 3 independent experiments.
