Millisecond Time-Resolved Solid-State NMR Reveals a Two-Stage Molecular Mechanism for Formation of Complexes between Calmodulin and a Target Peptide from Myosin Light Chain Kinase

Abstract

Calmodulin (CaM) mediates a wide range of biological responses to changes in intracellular Ca²⁺ concentrations through its calcium-dependent binding affinities to numerous target proteins. Binding of two Ca²⁺ ions to each of the two four-helix-bundle domains of CaM results in major conformational changes that create a potential binding site for the CaM binding domain of a target protein, which also undergoes major conformational changes to form the complex with CaM. Details of the molecular mechanism of complex formation are not well established, despite numerous structural, spectroscopic, thermodynamic, and kinetic studies. Here, we report a study of the process by which the 26-residue peptide M13, which represents the CaM binding domain of skeletal muscle myosin light chain kinase, forms a complex with CaM in the presence of excess Ca²⁺ on the millisecond time scale. Our experiments use a combination of selective ¹³C labeling of CaM and M13, rapid mixing of CaM solutions with M13/Ca²⁺ solutions, rapid freeze-quenching of the mixed solutions, and low-temperature solid state nuclear magnetic resonance (ssNMR) enhanced by dynamic nuclear polarization. From measurements of the dependence of 2D ¹³C-¹³C ssNMR spectra on the time between mixing and freezing, we find that the N-terminal portion of M13 converts from a conformationally disordered state to an α-helix and develops contacts with the C-terminal domain of CaM in about 2 ms. The C-terminal portion of M13 becomes α-helical and develops contacts with the N-terminal domain of CaM more slowly, in about 8 ms. The level of structural order in the CaM/M13/Ca²⁺ complexes, indicated by ¹³C ssNMR line widths, continues to increase beyond 27 ms.

Figure 1.

(A) Schematic representation of CaM/M13/Ca²⁺ complex formation. Residues of M13 that are isotopically labeled in time-resolved ssNMR experiments are shown in magenta. N-terminal and C-terminal domains of CaM are blue and beige. The apo-CaM structure is based on Protein Data Bank file 1CFC. The fully formed complex structure is based on Protein Data Bank file 2LV6, with four bound Ca²⁺ ions shown in green. (B) Schematic representation of the rapid mixing and freeze-quenching apparatus used to prepare frozen CaM/M13/Ca²⁺ solutions for time-resolved ssNMR measurements. The evolution time τ_e of a given sample is determined by the velocity of a jet of CaM/M13/Ca²⁺ solution from the mixer and by the flight distance d_f before freezing. (C) Double-quantum-filtered ¹³C ssNMR spectra of frozen solutions with the indicated τ_e values. Spectra were obtained at 25 K and 9.4 T, with 7.00 kHz MAS and signal enhancements from DNP.

Figure 2.

(A) 2D ¹³C-¹³C ssNMR spectra of a frozen M13-KVAF/Ca²⁺ solution (τ_e = 0) and a frozen, premixed CaM/M13-KVAF/Ca²⁺ solution (τ_e = ∞). Color scales for crosspeak intensities are explained in the text. ¹³C_α/¹³C_β crosspeak assignments for labeled residues in M13 at τ_e = ∞ are shown in orange. Dashed grey lines illustrate ₁₃C chemical shift changes that accompany the conversion of M13 from a conformationally disordered unbound state to an α-helical CaM-bound state. Spectra were acquired with 7.00 kHz MAS and ¹³C-¹³C mixing periods τ_SD = 20 ms. (B) 2D ¹³C-¹³C ssNMR spectra of frozen M13-KVAF/Ca²⁺ solutions with the indicated values of τ_e (left column) and 2D difference spectra (right column), in which the spectrum at τ_e = ∞ is subtracted from spectra at the indicated values of τ_e as explained in the text. Crosspeak signals in the difference spectra are attributed to disordered components of the conformational distribution of M13 at each value of τ_e. The notation "(x ms)-(∞)" represents the difference between 2D spectra with τ_e = x and τ_e = ∞, including the scaling factor η defined in the text.

Figure 3:

Same as Fig. 2, but for frozen CaM/M13-FAI/Ca²⁺ solutions. 2D ¹³C-¹³C ssNMR spectra were recorded with MAS at 7.00 kHz (A,C) and 5.00 kHz (B,D).

Figure 4:

(A,B,C) 2D ¹³C-¹³C ssNMR spectra of frozen solutions of CaM/M13-KVAF/Ca²⁺, CaM/M13-FAI/Ca²⁺ at 7.00 kHz MAS frequency, and CaM/M13-FAI/Ca²⁺ at 5.00 kHz MAS frequency, respectively, with τ_e = ∞ and τ_SD = 20 ms. Regions that are used to evaluate crosspeak volumes are indicated. (D) Residue-specific build-up curves for crosspeak signals attributable to conformationally ordered M13 molecules, determined from ratios of crosspeak volumes in the 2D spectra and 2D difference spectra in Figs. 2 and 3. Build-up times τ_bu were determined from single-exponential fits, shown as dashed lines.

Figure 5:

(A,B) 2D ¹³C-¹³C ssNMR spectra of frozen CaM/M13-KVAF/Ca²⁺ solutions and frozen CaM/M13-FAI/Ca²⁺ solutions, respectively acquired with τ_SD = 1.0 s. Intermolecular crosspeaks between ¹³C-labeled sites in M13 and ¹³C-labeled methionine methyl carbons of CaM are indicated in orange.

Figure 6:

(A,B) 2D ¹³C-¹³C ssNMR spectra of frozen solutions of CaM/M13-KVAF/Ca²⁺ and CaM/M13-FAI/Ca²⁺, respectively, with τ_e = ∞ and τ_SD = 1.0 s. Regions that are used to evaluate crosspeak volumes are indicated. (C) Residue-specific build-up curves for intermolecular contacts between M13 and methionine sidechains of CaM, determined from ratios of intermolecular and intramolecular crosspeak volumes in the 2D spectra in Fig. 5. Build-up times τ_bu were determined from single-exponential fits, shown as dashed lines.