Extended Data Fig. 2: Biochemical and biophysical control experiments show that the labelled GrsA mutants are functional, and that the dipoles of attached dyes can randomize on the millisecond timescale.
From: Subdomain dynamics enable chemical chain reactions in non-ribosomal peptide synthetases
a, Similar thioester formation activities measured by ESI-TOF MS for holo cysteine-free A-PCP (hCF), labelled hCC1 and labelled hCC2. Data plotted are percentages of the first Phe-adduct mass peak and presented as mean ± 1 s.d. from three time-dependent FRET kinetic runs for hCF and labelled hCC1, and two runs for labelled hCC2. Solid lines are single exponential functions with fitting results labelled. b–c, Binding to Phe-AMS, a non-hydrolysable Phe-AMP mimic, results in a more compact state as shown by a native PAGE mobility assay for CC1 (b) and CC2 (c) in both the apo and holo forms. n = 2 in (b) and n = 4 in (c) for independently repeated native PAGE analyses with similar results. d, The His-tag and anti–His-tag antibody interaction does not affect the thioester formation activity of hCF probed by ESI-TOF MS. Mass spectra were normalized to the maximum peak height. The relative ESI-TOF MS peak height was used in calculating the relative activity between the sample of 1:1 hCF:anti-His-tag antibody and the hCF alone. The relative activity is shown as mean ± 1 s.d. from three kinetic runs with 90-s incubation. Inset: a cartoon representation of the immobilization scheme. e, Low bulk anisotropy during FRET process was observed for aCC2-K434A and hCC2-K434A with or without substrates. Data are presented as mean ± 1 s.d. from four repeats of hCC2-K434A with substrates and three repeats for other conditions. f–k, Representative autocorrelation functions of the acceptor channel (f, g), the donor channel (h, i) and the cross-correlation functions between the two channels during smFRET (j, k). Panels (f), (h) and (j) are for single aCC1molecules, and (g), (i) and (k) are for single aCC2-R439A molecules. The excitation was linearly polarized and modulated at 1 kHz so that any unwanted interaction between attached dyes and proteins on millisecond or slower would have obvious signatures in these correlation tests at the oscillation frequency of 1 kHz. However, no significant correlation at such lag time was observed; therefore, these results ruled out the possibility that dyes are transiently immobilized on the protein surface on millisecond or slower time scale. Dashed lines represent 95% confidence level.
