Fig. 4: Direct observation of POR biased conformational sampling by small-molecule ligands using smFRET.

A Illustration of smFRET assay using TIRF microscopy. Top; SbPOR2b is site-specifically labeled with Cy3/Cy5 fluorophores, reconstituted in lipid nanodiscs and tethered on a passivated microscope surface. Bottom; representative smFRET traces displaying FRET states and dynamic transitions between them (see Supplementary Fig. 12 for more examples). Top row: Donor (green) and acceptor (red) intensities over time (s). Middle row: acceptor only intensity (red), bottom row: E_FRET values (orange) calculated with calibration factors, and idealized FRET value determined from HMM fitting (blue). B Distribution of FRET efficiencies in the absence and presence of ligands. Distributions are optimally fit with 5 states for all conditions as determined from BIC (see Supplementary Methods and Supplementary Fig. 13) with average distances ranging from ~40 to ~80 Å. Rifampicin, cyclophosphamide and dhurrin alter the occupancies of each of the five FRET states indicating biased conformational sampling. Colored bars on top of histograms represent occupancies of each state. N denotes of the number of single molecules at each experimental condition. C FRET efficiencies and converted inter-dye distances obtained from five-state gaussian mixture models. Each FRET state may reflect an equilibrium between multiple conformations. D Homology modeling of SbPOR2b from crystal structures of POR isoforms in a compact, intermediate conformation and a human-yeast chimera in a fully extended conformation (PDBs: 3QE2, 3ES9 and 3FJO respectively) with Monte Carlo simulated inter-dye distances (bold) and Cα-Cα distances (brackets). Source data are provided as a Source Data file.