A suite of 19F based relaxation dispersion experiments to assess biomolecular motions

Abstract

Proteins and nucleic acids are highly dynamic bio-molecules that can populate a variety of conformational states. NMR relaxation dispersion (RD) methods are uniquely suited to quantify the associated kinetic and thermodynamic parameters. Here, we present a consistent suite of ¹⁹F-based CPMG, on-resonance R_1ρ and off-resonance R_1ρ RD experiments. We validate these experiments by studying the unfolding transition of a 7.5 kDa cold shock protein. Furthermore we show that the ¹⁹F RD experiments are applicable to very large molecular machines by quantifying dynamics in the 360 kDa half-proteasome. Our approach significantly extends the timescale of chemical exchange that can be studied with ¹⁹F RD, adds robustness to the extraction of exchange parameters and can determine the absolute chemical shifts of excited states. Importantly, due to the simplicity of ¹⁹F NMR spectra, it is possible to record complete datasets within hours on samples that are of very low costs. This makes the presented experiments ideally suited to complement static structural information from cryo-EM and X-ray crystallography with insights into functionally relevant motions.

Keywords: Fluorine; Large complexes; Protein folding; Relaxation dispersion; Structural dynamics..

Fig. 1

The folding-unfolding exchange of the TmCsp protein. a The chemical structure of 5-fluoro tryptophan (5FW; left) and a schematic presentation of the (un-) folding of TmCsp, which contains 5FW residues at positions 7 and 29 (Protein Data Bank, PDB ID 1G6P Kremer et al. 2001). b ¹⁹F spectra of TmCsp at 333 K, 343 K and 373 K with assignments for Trp7 and Trp29. The peak labeled with an asterisk results from a small impurity in the sample. This impurity resonates outside the displayed spectral window at the two lower temperatures

Fig. 2

¹⁹F CPMG experiment. a Pulse sequence for recording ¹⁹F CPMG RD profiles. Narrow (wide) rectangles indicate 90° (180°) pulses, which are applied along the x-axis unless indicated otherwise. The phase cycle is φ1 = φ_rec = [x, x, − x, − x, y, y, − y, − y], φ2 = [y, − y, y, − y, x, − x, x, − x]. The number of CPMG pulses applied during the CPMG time T_CPMG is given by 2n, where n is an integer. The maximum number n_max is chosen so that the highest applied frequency υ_CPMG = n_max/T_CPMG is at or below 5 kHz. b CPMG relaxation dispersion profiles for W7 and W29 recorded at 344 K. The size of the error-bars correspond to 1 standard deviation

Fig. 3

¹⁹F on-resonance R_1ρ experiment. a Pulse sequence for the ¹⁹F on-resonance R_1ρ experiment. Narrow (wide) rectangles indicate 90° (180°) pulses, which are applied along the x-axis unless indicated otherwise. The phase cycle is φ1 = x, φ2 = [− x, x], φ3 = [x, x, − x, − x, y, y, − y, − y], φ4 = [y], φ5 = [− y], φ_rec = [x, − x, − x, x, y, − y, − y, y]. b On-resonance R_1ρ relaxation dispersion profiles for W7 and W29. The size of the error-bars correspond to 1 standard deviation

Fig. 4

¹⁹F off-resonance R_1ρ experiment. a Pulse sequence for the ¹⁹F off-resonance R_1ρ experiment. Narrow rectangles indicate 90° pulses. The phase cycle is φ1 = x, φ2 = [− x, − x, x, x], φ3 = [x, x, x, x, − x, − x, − x, − x, y, y, y, y, − y, − y, − y, − y], φ4 = [y, − y], φ5 = [− y, y], φ6 = [x, − x], φ7 = [x], φ8 = [x, − x], φ_rec = [x, − x, − x, x, y, − y, − y, y]. The flip-angle of pulses that flank the spinlock block is θ, which ensures that the magnetization is aligned at the angle of the effective magnetic field. This angle depends on the offset and the spinlock power. The pulse pairs with phases φ7/φ8 are used to cycle the magnetization to ± z before the spin lock period and back to + z after the spin lock, which ensures that the rotating frame relaxation is symmetrically measured both above and below the transverse plane at all offsets. b R_1ρ off-resonance relaxation dispersion profiles and c corresponding R₂ + R_ex contributions for W7 at spin-lock fields of 100 Hz, 200 Hz, 300 Hz and 400 Hz. Solid lines show the best fit to a two-state Laguerre approximation. The offsets of the folded (F) and the unfolded (U) state are indicated with dotted lines. Error bars show experimental uncertainty (1 standard deviation)

Fig. 5

Kinetic and thermodynamic analysis of the TmCsp folding. a CPMG curves with a global fit, assuming a constant chemical shift difference between the folded and the unfolded state across all temperatures. b Temperature dependence of exchange rates and populations from 328 to 343 K. For comparison, the results from the experiments at 344 K are included. c Temperature dependence of k_u, k_f and K_eq of 5FW-labeled TmCsp from 328 to 343 K. Note that only the data between 335.5 and 343 K was used to extract the thermodynamic parameters

Fig. 6

¹⁹F Relaxation dispersion experiments of the 360 kDa α₇α₇ double heptamer. a Model of the double heptameric α₇α₇ complexes with 14 BTFA labeling sites at position 18C (purple) of each subunit. Position 35 that does not show exchange (Fig. S8) is indicated in cyan. The model of the complex is based on the structure of the 20S proteasome from T. acidophilum (PDB ID 1PMA). b CPMG experiments from 293 to 323 K. c On-resonance R_1ρ experiments from 293 to 323 K. d Plot of exchange rates against temperatures as derived from the global fit of all RD data. e Logarithmic plot of kinetic rates against inverse temperature. f Schematic presentation of ΔG, ΔH and TΔS changes from the ground state GS to the excited state ES via a transition state TS