Structural basis for calmodulin as a dynamic calcium sensor

Abstract

Calmodulin is a prototypical and versatile Ca(2+) sensor with EF hands as its high-affinity Ca(2+) binding domains. Calmodulin is present in all eukaryotic cells, mediating Ca(2+)-dependent signaling. Upon binding Ca(2+), calmodulin changes its conformation to form complexes with a diverse array of target proteins. Despite a wealth of knowledge on calmodulin, little is known on how target proteins regulate calmodulin's ability to bind Ca(2+). Here, we take advantage of two splice variants of SK2 channels, which are activated by Ca(2+)-bound calmodulin but show different sensitivity to Ca(2+) for their activation. Protein crystal structures and other experiments show that, depending on which SK2 splice variant it binds to, calmodulin adopts drastically different conformations with different affinities for Ca(2+) at its C-lobe. Such target protein-induced conformational changes make calmodulin a dynamic Ca(2+) sensor capable of responding to different Ca(2+) concentrations in cellular Ca(2+) signaling.

Figure 1. A distinct conformation of the CaM-CaMBD2-b complex

(A & B) Structure of the CaM-CaMBD2-b complex. A 2×2 complex is formed with two horizontal CaMBD2-b peptides, green, and two CaM molecules, salmon (A). Also shown are side chains of the three residues A463, R464 and K465 in CaMBD2-b. A 90°-turn of the structure shows that CaM in CaMBD2-b adopts an “S”-shaped configuration (B). (C & D) Structure of the CaM-CaMBD2-a complex (1G4Y). Two CaMBD2-a peptides, gold, and two CaM molecules, blue, form the 2×2 complex (C). CaM in the CaM-CaMBD2-a complex displays a “C”-like structure (D). (E & F) Comparison of the structures of CaM from CaM-CaMBD2-b (salmon) and CaM-CaMBD-2a (blue). The CaM structures are aligned at the N-lobe. Notice the difference at the linker region (R74-E82). See also Figure S1

Figure 1. A distinct conformation of the CaM-CaMBD2-b complex

(A & B) Structure of the CaM-CaMBD2-b complex. A 2×2 complex is formed with two horizontal CaMBD2-b peptides, green, and two CaM molecules, salmon (A). Also shown are side chains of the three residues A463, R464 and K465 in CaMBD2-b. A 90°-turn of the structure shows that CaM in CaMBD2-b adopts an “S”-shaped configuration (B). (C & D) Structure of the CaM-CaMBD2-a complex (1G4Y). Two CaMBD2-a peptides, gold, and two CaM molecules, blue, form the 2×2 complex (C). CaM in the CaM-CaMBD2-a complex displays a “C”-like structure (D). (E & F) Comparison of the structures of CaM from CaM-CaMBD2-b (salmon) and CaM-CaMBD-2a (blue). The CaM structures are aligned at the N-lobe. Notice the difference at the linker region (R74-E82). See also Figure S1

Figure 2. Ca²⁺ promotes formation of a 2×2 complex for both CaM-CaMBD2-a and CaM-CaMBD2-b in solution

(A & B) A 1×1 complex of CaM-CaMBD2-a and CaM-CaMBD2-b in the absence of Ca²⁺. Upper panels are raw SE data (symbols) at different centrifugation speeds as indicated. Curves are fit of the raw data to a single-component model using Sedphat. Lower panels are residuals of the fit. Without Ca²⁺ both CaM-CaMBD2-a and CaM-CaMBD2-b are 1×1, with a predicted molecular mass of 28.9 kDa and 29.3 kDa respectively. (C & D) A 2×2 complex of CaM-CaMBD2-a and CaM-CaMBD2-b in the presence of Ca²⁺. Upper panels are raw SE data (symbols) with fit to a single-component model (curves), and lower panels are residuals of the fit. Ca²⁺ promotes formation of a 2×2 complex for both CaM-CaMBD2-a and CaM-CaMBD2-b with a predicted molecular mass of 57.8 kDa and 58.6 kDa respectively.

Figure 2. Ca²⁺ promotes formation of a 2×2 complex for both CaM-CaMBD2-a and CaM-CaMBD2-b in solution

(A & B) A 1×1 complex of CaM-CaMBD2-a and CaM-CaMBD2-b in the absence of Ca²⁺. Upper panels are raw SE data (symbols) at different centrifugation speeds as indicated. Curves are fit of the raw data to a single-component model using Sedphat. Lower panels are residuals of the fit. Without Ca²⁺ both CaM-CaMBD2-a and CaM-CaMBD2-b are 1×1, with a predicted molecular mass of 28.9 kDa and 29.3 kDa respectively. (C & D) A 2×2 complex of CaM-CaMBD2-a and CaM-CaMBD2-b in the presence of Ca²⁺. Upper panels are raw SE data (symbols) with fit to a single-component model (curves), and lower panels are residuals of the fit. Ca²⁺ promotes formation of a 2×2 complex for both CaM-CaMBD2-a and CaM-CaMBD2-b with a predicted molecular mass of 57.8 kDa and 58.6 kDa respectively.

Figure 3. Different thermodynamic profiles for CaM-CaMBD2-b and CaM-CaMBD2-a

(A) Surface representation of the CaM-CaMBD2-b complex. CaMBD2-b (green) interacts with CaM (salmon) only at both the N- and C-lobes. There is no physical contact between the CaMBD2-b peptides. (B) Surface representation of the CaM-CaMBD2-a complex. The CaMBD2-a peptides (gold) form extensive contacts with CaM (blue) as well as between themselves. (C) Results of ITC experiments. Ca²⁺-bound CaM was titrated into either CaMBD2-a or CaMBD2-b. For CaM-CaMBD2-a, a two-site model is required to fit the data (the smooth line). On the other hand, a single-site model can adequately fit the data and use of the two-site model did not statistically improve the fitting. (D) Thermodynamic profiles for formation of the CaM-CaMBD2-a and CaM-CaMBD2-b complexes. Plotted are the Gibbs free energy (ΔG), enthalpy (ΔH) and entropy (−TΔS) for formation of both CaM-CaMBD2-a and CaM-CaMBD2-b. ΔG is calculated from ΔH and TΔS (Table S1), which were determined by ITC at 20°C (ΔG = ΔH − TΔS). See also Figure S2 and Table S1

Figure 4. Interactions of CaM and CaMBDs at the CaM N-lobe

(A) Amino acid residues involved in formation of CaM-CaMBD2-b and CaM-CaMBD2-a with Ca²⁺. Buried surface area (BSA, bar graphs) identifies key residues in CaMBD2-b and CaMBD2-a which interact with CaM. A C-terminal fragment, from E472/E469 to L491/L488, interacts with the CaM N-lobe. Insertion of ARK shifts the numbering by three for CaMBD2-b. Also listed are CaM residues, in a smaller type face, that are in contacts (within 5 Å radius) with individual CaMBD residues. Shown in red are the CaM residues which only interact with key residues in CaMBD2-b. CaM residues in cyan form contacts only with CaMBD2-a. CaM residues in a black bold type face interact with both. (B & C) Hydrophobic interfaces of the CaM N-lobe in complex with CaMBD2-b (B) or CaMBD2-a (C). Overlaid are key residues from CaMBD2-b or CaMBD2-a. Both N-lobes are aligned from L4 to R74, with r.m.s.d. = 0.86 Å, to create the graphs. See also Figure S3

Figure 5. Interactions of CaM and CaMBDs at the CaM C-lobe

(A) Amino acid residues involved in formation of CaM-CaMBD2-b and CaM-CaMBD2-a in the presence of Ca²⁺, similar to descriptions in Fig. 3. An N-terminal fragment, from R419 to L440, interacts with the CaM C-lobe. (B & C) Hydrophobic interfaces of the CaM C-lobe in complex with CaMBD2-b (B) or CaMBD2-a (C). Both C-lobes are aligned from I85 to T146, with r.m.s.d. = 2.53 Å, to create the graphs. See also Figure S4

Figure 6. Conformational plasticity of the CaM EF-hands

(A & B) Structures of EF-hands from CaM-CaMBD2-b (salmon) and CaM-CaMBD2-a (blue). EF-hands 1 and 2 are aligned from D20 to E31 and D56 to E67 respectively. EF-hands 3 and 4 are aligned from D93 to E104 and D129 to E140 respectively. Side chains at positions 7 and 9 are not shown. (C & D) Structures of EF-hands from CaM-CaMBD2-b (salmon) and the CaM-edema factor (pale cyan, 1K90). Significant differences exist in the structures of EF-hands 1 and 2, but not EF-hands 3 and 4. Side chains at positions 7 and 9 are not shown. See also Figure S5

Figure 7. The CaM-CaMBD2-b complex has a reduced apparent affinity for Ca²⁺

(A & B) Formation of CaM-CaMBD2-b (n = 4) complex is less sensitive to Ca²⁺, compared to that of CaM-CaMBD2-a (n = 7). Increases in fluorescence intensity (normalized) are plotted against free Ca²⁺ concentrations (A). There is a significant reduction in the apparent Kd for Ca²⁺ (p < 0.001, B). (C & D) Dose-dependent binding of CaMBD2-b (n = 3) and CaMBD2-a (n = 3) to AEDANS labeled CaM(T34C) in the presence of 10 µM Ca²⁺ (C) or 0.3 µM Ca²⁺ (D). CaM(T110C) yielded the same results. (E) The EC50 of CaM for CaMBDs obtained at 10 µM or 0.3 µM Ca²⁺. The EC50 of CaM for CaMBD2-b at 0.3 µM Ca²⁺ could not be determined (N.D.). (F) The EC50 for activation of SK2-b (n=10) and SK2-a (n=8) channels by Ca²⁺ from electrophysiology experiments, (P<0.001). See also Figure S6

Figure 7. The CaM-CaMBD2-b complex has a reduced apparent affinity for Ca²⁺

(A & B) Formation of CaM-CaMBD2-b (n = 4) complex is less sensitive to Ca²⁺, compared to that of CaM-CaMBD2-a (n = 7). Increases in fluorescence intensity (normalized) are plotted against free Ca²⁺ concentrations (A). There is a significant reduction in the apparent Kd for Ca²⁺ (p < 0.001, B). (C & D) Dose-dependent binding of CaMBD2-b (n = 3) and CaMBD2-a (n = 3) to AEDANS labeled CaM(T34C) in the presence of 10 µM Ca²⁺ (C) or 0.3 µM Ca²⁺ (D). CaM(T110C) yielded the same results. (E) The EC50 of CaM for CaMBDs obtained at 10 µM or 0.3 µM Ca²⁺. The EC50 of CaM for CaMBD2-b at 0.3 µM Ca²⁺ could not be determined (N.D.). (F) The EC50 for activation of SK2-b (n=10) and SK2-a (n=8) channels by Ca²⁺ from electrophysiology experiments, (P<0.001). See also Figure S6

Figure 8. A molecular model for changes in CaM’s affinity for Ca²⁺, induced by CaMBD2-b or CaMBD2-a

In the presence of Ca²⁺, CaM forms a 2×2 complex with both CaMBD2-b and CaMBD2-a in solution. Formation of the CaM-CaMBD2-b complex in the presence of Ca²⁺ requires that both the CaM N- and C-lobes become Ca²⁺-bound before they can interact with CaMBD2-b (A). The CaM C-lobe has a lower affinity for Ca²⁺ than the N-lobe (K_1C < K_1N), effectively determining the overall reduced Ca²⁺ sensitivity in formation of the 2×2 CaM-CaMBD2-b complex. In contrast, formation of the CaM-CaMBD2-a complex is Ca²⁺-dependent at the CaM N-lobe and becomes Ca²⁺-independent at the CaM C-lobe. (B). The open symbols represent the Ca²⁺-free EF-hands, while the filled circles and squares represent Ca²⁺-bound EF-hands. The open triangles represent EF-hands which are no longer able to bind Ca²⁺.