Fig. 2: Burst-wise analysis of the structural heterogeneity and dynamics of A_2AAR.

a Schematic drawings of the PR traces expected for a molecule switching between low-FRET (LF) and high-FRET (HF) states. Left, “apparent static heterogeneity” scenario: if the switching is too slow to occur within the observation time of individual molecules, the data is indistinguishable from the “static heterogeneity” scenario, where low-FRET and high-FRET species do not interchange. Right, “apparent static homogeneity” scenario: if the switching is fast enough to average over the binning time of the experiments, the data is indistinguishable from the “static homogeneity” scenario, where all molecules have the same PR that remains constant over time. b Experimental distributions of the burst-wise PR measured for A_2AAR without ligands (APO, N = 1815 bursts), with antagonist ZM241385 (ZM, N = 1436 bursts), with agonist NECA (NECA, N = 1741 bursts), and with agonist adenosine (ADN, N = 1271 bursts) are shown (purple lines). Simulated distributions for the “static homogeneity” scenario are also given (green lines) to show the expected broadening of the PR distributions due to shot noise. In the “static homogeneity” simulations, the ground-truth burst-wise PR matches the ensemble-averaged burst-wise PR in the A_2AAR data. c Standard deviations of the PRs measured in 1-ms time-bins were averaged in the groups of bursts equally spaced along the burst-wise PR axis, and were plotted for the apo and ligand-bound A_2AARs (purple lines) and the simulated “static heterogeneity” scenario (orange line). In the “static heterogeneity” simulations, the distribution of the PR matches those of the A_2AAR data. PR bins with less than 20 bursts were not analyzed. d Autocorrelation functions for the PRs calculated in 1-ms time-bins are plotted for the apo and ligand-bound A_2AAR (purple lines), and for simulated “static heterogeneity” (orange) and “static homogeneity” (green) scenarios. In the simulated data, the numbers of bursts, their durations, and the levels of the shot noise match those of the experimental data for A_2AAR shown on the same panel. The 95% confidence intervals were obtained via statistical bootstrapping (shaded colored areas)¹¹⁰.