NMR contributions to structural dynamics studies of intrinsically disordered proteins

Abstract

Intrinsically disordered proteins (IDPs) are characterized by substantial conformational plasticity. Given their inherent structural flexibility X-ray crystallography is not applicable to study these proteins. In contrast, NMR spectroscopy offers unique opportunities for structural and dynamic studies of IDPs. The past two decades have witnessed significant development of NMR spectroscopy that couples advances in spin physics and chemistry with a broad range of applications. This article will summarize key advances in basic physical-chemistry and NMR methodology, outline their limitations and envision future R&D directions.

Keywords: Biomolecular NMR; EPR spectroscopy; Intrinsically disordered proteins; NMR spin relaxation; Protein meta-structure; Structural biology.

Graphical abstract

Fig. 1

NMR spectral overlap in intrinsically disordered proteins. ¹H^N–¹⁵N HSQC spectra of IDPs are characterized by narrow spectral ranges and thus complicating sequential signal assignment. Spectra for the following IDPs are shown: (A) BASP1; (B) Osteopontin and (C) Tcf4.

Fig. 2

2D spectral planes for consecutive amino acids in MAP1B obtained by SMFT processing of the 5D randomly sampled signal. 2D cross-sections of (top) 5D hNCOncaCONH (N_i–CO_i−1 and N_i−1–CO_i−2) and (bottom) 5D HNcaCONH (HN_i–N_i and HN_i+1–N_i+1). Spins that were used for coherence transfer are depicted in lower case, while spins that are recorded during the indirect dimensions are given in upper case.

Fig. 3

Comparison between experimental NMR data and primary sequence-derived local secondary structure elements in the microtubule binding domain of the IDP MAP1B light chain. (top) ¹³C secondary chemical shifts (ΔδCα); (bottom) meta-structure derived local 2nd structure. The meta-structure derived values are defined according to NMR convention (positive values: α-helical elements; negative: extended conformations or β-strands).

Fig. 4

Long-range structural probing in IDPs using paramagnetic relaxation enhancements (PREs). PRE effects are quantified by intensity ratios of cross peaks selected from (A) diagmagnetic reference and (B) paramagnetic (spin-tagged) form of the protein. Selection of suitable spin label attachment sites is based on meta-structure derived compactness values (see text). Large compactness values are found for compact regions of the proteins, whereas small compactness values indicate conformationally flexible residue positions (and thus suitable for attaching the spin label). (C) Meta-structure derived compactness values and (D) experimental PREs obtained on the IDP Osteopontin. Spin label attachment sites were selected using small compactness values .

Fig. 5

Probing of structural compaction of IDPs using SOFAST-HMQC techniques . SOFAST-HMQC spectra of chicken BASP1 obtained under acidic (pH = 2) conditions without (A) or with (B) inversion of aliphatic side-chain protons. (C) enlarged view (circled) showing differential compaction of the polypeptide chain in BASP1. The intensity ratio λ_noe between reference (without, I_ref) and by applying the aliphatic proton inversion pulse (I_noe) is indicative of structural compaction of the polypeptide chain.

Fig. 6

Adiabatic fast passage NOESY as a valuable tool for studies of structural dynamics of IDPs. (A) Adiabatic spin-lock frame with offset Δω(t), and different effective fields ω_eff(t). Increasing the RF field strength ω₁(t) (from red to blue) leads to an increase of the tilt angle θ(t) between the static magnetic B₀ (along z-axis) and the effective field. Depending on the field strength cross-relaxation during adiabatic fast passage is predominantly longitudinal (red, NOESY-type) for low field strength and dominated by transverse contributions (blue, ROESY-type) at high field strengths. Depending on the effective correlation time (B) AFP-NOESY cross peaks can change sign. (C) ¹³C–¹H filtered ¹H^N–¹⁵N detected cross-relaxation in the IDP BASP1. Increasing the AFP field strength (from left to right) leads to increased contributions of ROESY-type cross-relaxation pathways. It should be noted that individual residues in the IDP BASP1 display significantly different side-chain backbone cross-relaxation behavior indicating differential mobilities along the protein backbone.

Fig. 7

Structural and dynamical adaptations in the IDP OPN upon heparin binding are probed by differential PREs and ¹⁵N relaxation parameters. (top) Structural changes are probed by differential ΔPRE. ΔPRE > 0 indicates “on average” increasing distance between labeling site and a residue upon binding (for ΔPRE < 0 the opposite). The red bar indicates the heparin binding site (binding site residues were excluded from the analysis because of signal overlap). The location of spin label attachment sites is indicated by blue triangles. (bottom) Conformational dynamics were monitored via ¹⁵N T₂ and ¹H^N–¹⁵N NOEs.

Fig. 8

Automated backbone dihedral angle determination using cross-correlated relaxation (CCR). The figure shows a schematic description how the combined usage of different, complementary cross-correlation rates can be used to unambiguously identify the backbone dihedral angles φ and ψ. Details can be found elsewhere . For illustration purposes different backbone dihedral angle arrangements are shown: (A) α-helical; (B) β-strand. Conformational averaging – as expected for an IDP – is illustrated with an equal distribution between α-helical and β-strand, shown in (C).

Fig. 9

Solution structural probing of IDPs using EPR-based double electron–electron resonance (DEER) spectroscopy . (A) DEER time traces of the double Cys-mutant C108–C188 of the IDP Osteopontin (OPN) at different urea concentrations. The modulation depth, Δ_eff = 1.0 − V(t = 3 μs)/V(0) is a direct measure of structural compaction . Decreased Δ_eff at higher urea concentration is due to global unfolding of the protein. (B) Identification of cooperatively folded substates in the ensemble of the OPN by measuring Δ_eff for the OPN double Cys-mutant C54–C247 as a function of urea concentration compaction. The sigmoidal dependence of Δ_eff vs urea concentration clearly shows the existence of a cooperatively folded substate . Error bars stem from signal noise.

Fig. 10

Domain elongation strategy for dynamic studies of IDPs. The presence of collective dynamic modes in IDPs does not allow for the separation of overall and internal dynamics. Significant separation of time scales can be achieved by (covalently) coupling the IDP to a large internal reference frame (fusion protein). (A) The IDP Osteopontin (OPN) (blue) is covalently linked (via Bismaleimide) to the C-terminal interaction domain of the oncogenic transcription factor Myc (C34). Binding of Myc (red) to its cognate protein binding partner Max (gray) leads to the formation of a stable α-helical protein complex displaying significant motional anisotropy (PDB_ID: 1NKP). (B) Differential changes (OPN_Myc-Max − OPN_Wild-Type) of ¹⁵N relaxation parameters (T₁,T₂, ¹H–¹⁵N NOE) upon domain elongation (C54@OPN was chosen as the Myc attachment site). The heterogenous distribution of relaxation parameters clearly indicates that OPN does not exist as a random coil in solution but populates compact substates . The covalent Bismaleimide linker is indicated in orange.

Fig. 11

The modular architecture of proteins. (A) Crystal structure of the Saccharomyces Cerevisiae Skp1-Cdc4-pSic1 peptide complex (PDB-ID: 3V7D); red: Cdc4:263-744; green: Skp1; blue: pSic1:67-85. The 3D structure of Cdc4 comprises eight repeats of a four β-strand motif which itself results from a duplication of a basic β-hairpin structure. (B) and (C) internal structural superpositions indicating the repeating structure (B: rotation angle = 360/8 = 45°; C: rotation angle = 5 * 360/8 = 225. (D) 8-fold structural symmetry in Cdc4 (the different motifs are color coded). (E) The basic four β-strand structural motif is composed of two fundamental β-hairpins. Internal structural symmetries were identified using the program TopMatch .