Methodology for rigorous modeling of protein conformational changes by Rosetta using DEER distance restraints

Abstract

We describe an approach for integrating distance restraints from Double Electron-Electron Resonance (DEER) spectroscopy into Rosetta with the purpose of modeling alternative protein conformations from an initial experimental structure. Fundamental to this approach is a multilateration algorithm that harnesses sets of interconnected spin label pairs to identify optimal rotamer ensembles at each residue that fit the DEER decay in the time domain. Benchmarked relative to data analysis packages, the algorithm yields comparable distance distributions with the advantage that fitting the DEER decay and rotamer ensemble optimization are coupled. We demonstrate this approach by modeling the protonation-dependent transition of the multidrug transporter PfMATE to an inward facing conformation with a deviation to the experimental structure of less than 2Å Cα RMSD. By decreasing spin label rotamer entropy, this approach engenders more accurate Rosetta models that are also more closely clustered, thus setting the stage for more robust modeling of protein conformational changes.

Fig 1

A) Distribution of pseudo-rotamers, shown as spheres, at four representative residues in T4 Lysozyme prior to (top, gold) and following (bottom, teal) refinement by multilateration. A flow chart detailing the iterative steps of pseudo-rotamer refinement using RosettaDEER is shown between the two T4L structures. B) Five representative DEER traces in T4 Lysozyme used for multilateration, alongside simulated DEER traces prior to (yellow) and following (teal) refinement. Insets: Simulated DEER distributions following pseudo-rotamer refinement alongside reference distributions with 95% confidence bands (calculated using GLADDvu and shown in grey). C) Goodness-of-fit evaluated from the RMSD between simulated and experimental DEER traces comparing RosettaDEER to other analysis programs.

Fig 2

Evaluation of average distances (A) and distribution widths (B) between pseudo-rotamers prior to (top) and following (bottom) refinement by multilateration. T4 lysozyme distributions omitted from multilateration are shown in light green. C and D) Boxplots showing the difference between values obtained using GLADDvu and values simulated between pseudo-rotamer ensembles prior to and following refinement.

Fig 3. Modeling the outward-to-inward conformational change in the multidrug transporter PfMATE.

(A) Outward-facing and (B) inward-facing crystal structures of PfMATE. N- and C-terminal domains are shown in purple and green, respectively, and TM7 is shown in red. (C) RMSD values of the ten best-scoring models for each of four sets of restraints relative to the inward-facing conformation using either pseudo-rotamers refined by multilateration (teal) or unrefined pseudo-rotamers available by default (yellow).

Fig 4. Models of PfMATE obtained using multilaterated rotamers more closely resemble the inward-facing crystal structure than those obtained using default rotamers.

Deviation between C_α-C_α distances observed between representative pairs of residues on the A) extracellular and B) intracellular sides of the crystal structure (PDB: 6FHZ) and the corresponding distances predicted from each of the best-scoring models. (C and D) Best-scoring inward-facing models of PfMATE obtained using ten restraints either with pseudo-rotamers available by default (left) or with those refined by multilateration (right). Inward-facing crystal structure shown in black. Ribbon thickness corresponds to the C_α root mean squared fluctuation among the top ten models. Bottom: The best-scoring models obtained using default rotamers (left) were less inward-open than those obtained using multilaterated rotamers (right).