The calponin regulatory region is intrinsically unstructured: novel insight into actin-calponin and calmodulin-calponin interfaces using NMR spectroscopy

Abstract

Calponin is an actin- and calmodulin-binding protein believed to regulate the function of actin. Low-resolution studies based on proteolysis established that the recombinant calponin fragment 131-228 contained actin and calmodulin recognition sites but failed to precisely identify the actin-binding determinants. In this study, we used NMR spectroscopy to investigate the structure of this functionally important region of calponin and map its interaction with actin and calmodulin at amino-acid resolution. Our data indicates that the free calponin peptide is largely unstructured in solution, although four short amino-acid stretches corresponding to residues 140-146, 159-165, 189-195, and 199-205 display the propensity to form α-helices. The presence of four sequential transient helices probably provides the conformational malleability needed for the promiscuous nature of this region of calponin. We identified all amino acids involved in actin binding and demonstrated for the first time, to our knowledge, that the N-terminal flanking region of Lys(137)-Tyr(144) is an integral part of the actin-binding site. We have also delineated the second actin-binding site to amino acids Thr(180)-Asp(190). Ca(2+)-calmodulin binding extends beyond the previously identified minimal sequence of 153-163 and includes most amino acids within the stretch 143-165. In addition, we found that calmodulin induces chemical shift perturbations of amino acids 188-190 demonstrating for the first time, to our knowledge, an effect of Ca(2+)-calmodulin on this region. The spatial relationship of the actin and calmodulin contacts as well as the transient α-helical structures within the regulatory region of calponin provides a structural framework for understanding the Ca(2+)-dependent regulation of the actin-calponin interaction by calmodulin.

Figure 1

(A) ¹H-¹⁵N HSQC two-dimensional spectrum of ¹⁵N-labeled CaP131–228. The experiment is recorded at 600 MHz on a sample of 0.7 mM in actin binding buffer at pH 7.0 and a temperature of 278 K (5°C). Each peak in the spectrum represents one amino acid. (B) Strips from an ¹H-¹⁵N 3D NOESY-HSQC experiment recorded at a temperature of 278 K. The backbone amide regions for two groups of sequential residues are extracted from the three-dimensional spectrum and arranged in sequential order. Weak sequential amide-amide NOESY crosspeaks can be identified that are suggestive of some nascent turn or helical conformations.

Figure 2

Heteronuclear ¹H-¹⁵N NOE experiments for CaP131–228 recorded at 600 MHz in actin-binding buffer at two temperatures: (A) 298 K and (B) 278 K. The signs of the peaks are color-coded with blue for positive and red for negative peaks. Although all peaks in the spectrum are negative at 298 K there is a significant number peaks with positive sign at 278 K. (C) (Top line) Sequence of mouse h1-calponin 131–228 region. (Graph in the middle) Heteronuclear NOEs on a scale of −0.6 to 0.3. (Arrows at end of the bars) Values that go beyond that range. (Bottom line) All residues for which sequential amide-amide crosspeaks were identified in the three-dimensional ¹H-¹⁵N NOESY-HSQC experiment. These residues have a propensity to form an α-helical structure.

Figure 3

Interaction experiments of CaP131–228 with F-actin at a temperature of 288 K monitored by ¹H-¹⁵N TROSY spectra. To a sample of 200 μM ¹⁵N labeled CaP131–228, F-actin stock of 17 mg/mL (400 μM) was added in five steps (0 μM, dark blue; 17 μM, yellow; 34 μM, blue; 51 μM, brown; 68 μM, green; and 85 μM, red) and a TROSY spectrum was recorded at each step. (A) Complete spectrum. (B and C) Smaller sections. (Dashed boxes) Locations of the smaller sections in the overall spectrum. (Amino acids in the smaller sections are labeled in black; those that show significant changes are labeled in red.) (D) Overall analysis of the CaP131–238-F-actin NMR binding data. Peak intensities in the first and the last step of the titration were measured and the ratio of the intensity of CaP131–228 peaks in the presence of F-actin divided by the intensity of CaP131–228 in the absence of F-actin was calculated for each residue and plotted against the sequence of CaP131–228. A bar is drawn at a value of 0.3 on the Y abscissa corresponding to a loss of intensity of 70%. Values < 0.3 indicate residues involved in binding whereas values >> 0.3 suggest no interaction with F-actin.

Figure 4

Titration of ¹⁵N-labeled CaP131–228 with unlabeled calmodulin in calmodulin-binding buffer at a temperature of 298 K at 500 MHz. The concentration of CaP131–228 was 700 μM and calmodulin was added in eight steps (dark blue, 0 μM; green, 350 μM; yellow, 700 μM; pink, 1050 μM; blue, 1400 μM; black, 1600 μM; and red, 2100 μM) until a ratio of CaP131–228:Calmodulin of 1:3 was reached. An HSQC experiment was recorded at each step. A large number of residues show significant chemical shift perturbations in fast exchange. (A) Overall spectrum. (B and C) Smaller sections. (Dashed boxes) Locations of the smaller sections in the overall spectrum. (Amino acids in the smaller sections are labeled in black whereas those that show significant changes are labeled in red.) (D) Summary of the titration of ¹⁵N-labeled CaP131–228 with calmodulin. The combined ¹H and ¹⁵N chemical shift differences between the first (0 μM calmodulin) and the last step (1600 μM calmodulin) of the titration are plotted against the sequence. Two bars are drawn at 0.3 ppm and 0.7 ppm chemical shift perturbation.

Figure 5

Titration of ¹⁵N-labeled calmodulin with unlabeled CaP131–228 monitored by ¹H-¹⁵N HSQC experiments. The sample concentration of calmodulin was 0.47 mM and CaP131–228 was added in several steps to give concentrations of 0 (blue), 94 (yellow), 187 (black), 281 (magenta), 374 (green), 470 (violet), 702 (brown), and 936 μM (red). Changes in peak positions are shown in two selected regions of the spectrum and displayed in panels B and C whereas panel A corresponds to the whole spectrum. (Amino acids with large changes are labeled in red.) (D) Plot of chemical shift perturbations in calmodulin upon interaction with CaP131–228 against the sequence of calmodulin.

Figure 6

Plot of the chemical shift perturbations of Lys¹¹⁵ in calmodulin upon titration with CaP131–228 fitted to a quadratic equation for a two-state binding equilibrium. (Solid squares) Experimental data points. (Continuous line) Fitted curve. The fitting gives a stoichiometry of 1:1 and a dissociation constant of 200 μM.

Figure 7

Summary of the structure-function relationship of the central region of calponin as obtained from our NMR data and prediction of disordered regions by two prediction programs, RONN (top) and IVLSP (bottom). Residues that showed a loss of intensity upon actin binding are labeled with the letter A, residues that gave chemical shift perturbations upon Ca²⁺-calmodulin binding are labeled with the letter C, and residues that showed a propensity to form an α-helical structure in NOESY experiments (Fig. 2) and NOE experiments (Fig. 3) are labeled with the letter H. P indicates phosphorylation sites Ser¹⁷⁵ and Thr¹⁸⁴.