Figure 1. – HSQC spectra of the diamagnetic CaLa–calbindin D _9k (a) and of the paramagnetic CaTb–calbindin D _9k (b). Some peaks are labeled.

Figure 2. Structural parameters determining the ^NH PCS (r,θ,ϕ) and PRDC values (Θ,Φ). The angles are provided in the frame established by the principal axes of the magnetic susceptibility anisotropy tensor.

Figure 3. Iso‐PCS surfaces calculated from Equation 8 with Δχ_rh = 0 or 2/3Δχ_ax. Positive shifts are in yellow, negative shifts in red.

Figure 4. PRE causes large line broadening of all NMR signals from nuclei within a sphere, called the “blind” sphere, around the metal ion. The size of the blind sphere changes with the nuclear type, being 1.6 times smaller for than for . The PRE decreases fast with increasing distances from the metal ion, so that it is negligible outside a sphere with detectable paramagnetic effects.

Figure 5. Superimposition of a region of the – HSQC IPAP spectra of calcium‐bound N60D calmodulin (in cyan) and of the thulium‐substituted protein (in black). The J coupling for residue Gly33 changes from −94.93 Hz in the diamagnetic form to −85.07 Hz in the paramagnetic form; the difference in the splitting of the two components provides the PRDC (9.86 Hz). Note the change in the position of the peaks due to a PCS of 2.84 ppm.

Figure 6. paramagnetic line broadening in the presence of various metal ions shown as a function of the typical values of the electron relaxation time, for a molecule with a reorientation time of 10 ns at 298 K and 900 MHz of proton Larmor frequency, and a metal–nucleus distance of 10 Å (HS = high spin, LS = low spin).

Figure 7. Typical values of the longitudinal PRE (calculated at the conditions of Figure 6) and of the magnetic susceptibility anisotropy for various metal ions.

Figure 8. Typical values of the paramagnetic line broadening (calculated at the conditions of Figure 6) and of maximum PCS and ^NH PRDC at 10 (a) and 30 Å (b) from various metal ions. PCSs are detectable when larger than about 0.1 ppm, PRDCs when larger than about 1 Hz. The solid line indicates the condition of paramagnetic line broadening equal to the PRDC value. PCSs and PRDCs are calculated with the axial assumption (i.e., for a proton along the z‐axis of the metal susceptibility tensor and a coupled nitrogen along the metal–proton direction).

Figure 9. PCS measured for the N‐terminal domain of a protein provides the magnetic susceptibility anisotropy tensor; PRDC measured for the C‐terminal domain must agree with the same tensor. Therefore, the relative orientation of the two domains is determined by aligning the tensors calculated for the two domains. The PCS measured for the C‐terminal domain also determines the relative position.

Figure 10. In protein systems experiencing conformational heterogeneity, a target function (TF) is calculated as a function of the weight of a selected conformation. The target function is defined as the minimum squared difference between experimental data (PCS, PRDC, PRE, SAS, etc.) and data calculated by averaging over all conformations of the ensemble. The MO of a conformation is defined as the weight of this conformation corresponding to the intersection of the target function curve and a predefined threshold. In the example shown, the conformations with the C‐terminal domain in dark blue and in red have a MO below 0.1 and around 0.35, respectively.