Molecular mechanisms of cardiac actomyosin transforming from rigor state to post-rigor state

Abstract

Sudden cardiac death contributed to half of all deaths from cardiovascular diseases. The mechanism of the kinetic cycle of cardiac myosin is crucial for heart protection and drug development. The state change in the myosin kinetic cycle from the rigor state to the post-rigor state is fundamental to explain binding and dissociation. Here, we used β-cardiac myosin in the rigor and post-rigor states to model the actomyosin complexes. Molecular dynamics simulations, electrostatic analysis, and energetic analysis of actomyosin complexes were performed in this work. The results showed that there are fewer interactions and lower electrostatic binding strength in the post-rigor state than in the rigor state. In the post-rigor state, there were higher free binding energy, fewer salt bridges, and fewer hydrogen bonds. The results showed a lower binding affinity in the post-rigor state than in the rigor state. The decrease in the binding affinity provided important conditions for dissociation of the myosin from the actin filament. Although previous studies focused mostly on the binding process, this study provides evidence of dissociation, which is even more important in the myosin kinetic cycle. This research on the mechanism of myosin kinetic cycles provides a novel direction for future genetic disease studies.

I. INTRODUCTION

Sudden cardiac death (SCD) contributes to half of all deaths from cardiovascular diseases.¹ Hypertrophic cardiomyopathy (HCM), the main cause of SCD,^2,3 is a polygenic disease,⁴ which afflicts at least one in 500 people.⁵ In addition to SCD, HCM causes embolic stroke and heart failure in young and middle-aged people. To date, genetic research has located 33 HCM-related genes; MYBPC3, MYH7, TNNT2, TNNI3, TPM1, ACTC1, MYL2, and MYL3 have been categorized as definitive genes.⁶ These definitive genes encode human cardiac sarcomeres, including myosin-binding protein C (MyBP-C), β-cardiac myosin, cardiac troponin T, troponin I, tropomyosin, cardiac actin, and myosin light chains 2 and 3. β-cardiac myosin, a type of motor protein, is assembled to form myosin filaments (thick filaments). These periodically overlap with actin filaments to form the basic unit of contraction.⁷ The actin filament (thin filament) is composed of globular actin (G-actin),⁸ tropomyosin (Tpm), troponin, and MyBP-C.⁹ Binding between thick and thin filaments is caused by the hydrophobic interactions between the myosin heavy chains with G-actins.^10,11 Tropomyosins, two pairs of long helix chains on the two sides of the thin filament, regulate the exposure of the binding interface of G-actins.^12,13 Movement of the Tpms is regulated by the troponin complex, a transformable protein including troponin I, troponin T, and the Tpm-binding subunit.¹² The complex is regulated by the concentration of Ca²⁺ and moves Tpms into open, close, or blocked state.^12,13

In the kinetic cycle, the instantaneous increase in Ca²⁺ concentration causes a huge conformational change in the troponin complex with the azimuthal movement of Tpm,¹² exposing the binding interface of the G-actins. Then, the myosins approach and bind to the actin filament. The binding state between the thin filaments and the myosin is called the rigor state. In this state, both Pi and MgADP are released, forming a nucleotide-free state with a strong binding. Once Pi and MgADP have been released, ATP rapidly rebinds to the myosin, causing conformational change and dissociation^14,15 known as the recovery stroke. Meanwhile, myosin hydrolyzes ATP into ADP and Pi but does not release them directly.¹⁶ In this state, the pre-powerstroke (PPS), myosin is ready to undergo the powerstroke process. The release of Pi, the initial step in powerstroke, is followed by the release of MgADP, which closes the cleft of the myosin motor domain and forms a strong binding between the myosin and actin filaments. After the release of MgADP, myosin reaches the rigor state. It restarts ATP binding and hydrolysis, waiting for the next kinetic cycle.¹⁵

The powerstroke and recovery stroke are the two fundamental processes in the kinetic cycle. The powerstroke has attracted many investigators, whose studies range from the release of Pi^10,17 and MgADP^18,19 to the lever arm movement,^20,21 making the mechanism of the powerstroke much clearer than before. However, far fewer studies on the recovery stroke exist.^22,23 The cycle powerstroke and recovery stroke are similar mechanisms beneath opposite phenomena. Transformation of myosin from the rigor to the post-rigor state is the initial step for the recovery stroke. The necessity of the post-rigor state for the detachment of myosin needs more study. Knowledge of structural differences between the rigor and post-rigor states is fundamental for understanding the importance of the post-rigor state. The energetic and electrostatic analyses are crucial in understanding binding and dissociation.

Here, we used SWISS-MODEL²⁴ and Chimera²⁵ to build human β-cardiac actomyosin complexes in rigor and post-rigor states for a structural comparison. Afterward, we ran Molecular Dynamic (MD) simulations and calculated the electrostatic features and free binding energy to explore the differences between the rigor and post-rigor states, revealing the mechanism of the dissociation. The results showed that the actomyosin complex in the post-rigor state possesses a lower binding affinity compared with that in the rigor state. The energetic analysis also confirmed that myosin in the post-rigor state is not suitable for binding compared with that in the rigor state. These changes in the actomyosin complex in the post-rigor states provide more suitable conditions for the detachment of myosin. This study shows the importance of the post-rigor state for myosin detachment and sheds light on the mechanism of dissociation of the actomyosin complex.

II. METHOD

A. Modeling and MD simulations

The model of β-cardiac myosin in the rigor state was built using a template of rigor-like squid myosin S1 (PDB: 3I5G²⁶) based on the sequence of β-cardiac myosin²² through the SWISS-MODEL.²⁴ The sequence identity is 62.41%. The rigor myosin model was then assembled with the cardiac actin filament (PDB: 5NOJ¹³) based on the actomyosin complex model of 5JLH¹⁰ to build the β-cardiac actomyosin complex in the rigor state through the model alignment by Chimera.²⁵ The model of β-cardiac myosin in the post-rigor state was taken from the Protein Data Bank (PDB: 6FSA²²). The missing cardiomyopathy loop and strut domain were built according to the SWISS-MODEL.²⁴ Then, the β-cardiac myosin in the post-rigor state and the cardiac actin filament were taken to build the β-cardiac actomyosin complex in the post-rigor state using the actomyosin complex model of 5JLH.¹⁰

The solvation box of the TIP3P water model²⁷ with 150 mM of NaCl was built by Visual Molecular Dynamics²⁸ (VMD). The MD simulations were applied by NAMD 2.12.²⁹ In MD simulations, the end residue (Leu 807) of the myosin and the residues in the actin filament, which were 10 Å from the myosin, were restrained. The initial temperature and pH were set at 300 K and 7.0, respectively. The periodic boundary conditions and CHARMM 36³⁰ force field were applied to the systems. 20 000 steps of minimization were applied ahead of the 50 ns simulations. All simulations were repeated twice.

B. Electrostatic potential calculation

After simulations, the frames at 50 ns of the actomyosin complexes in the rigor and post-rigor states were taken for electrostatic potential calculations. The electrostatic potential calculations were performed using Delphi.³¹ The myosin motor domain and actin filament were separated³² by 20 Å. The charge and the radius of atoms were calculated by the force field CHARMM36 and assigned by pdb2pqr.³³ The resolution was set as 1.5 grid/Å, and the dielectric constants were set at 2.0 and 80.0 for protein and water, respectively. The probe radius for generating the protein surface was 1.4 Å, and the protein filling percentage in the calculation box was 70%. The concentration of NaCl was set at 0.15 M, and the boundary condition for the Poisson Boltzmann equation was set as the dipolar boundary condition. The calculated electrostatic potential on surfaces was visualized by Chimera.²⁵ VMD²⁸ was used to show the electrostatic field lines. In both visualizations, the color range was set from red to blue, corresponding to the potential range from −1.0 to 1.0 kT/e.

C. Electrostatic force

The electrostatic force on the myosin motor domain was calculated by DelphiForce³⁴ every 0.05 ns in the last 20 ns of MD simulations (30–50 ns). The parameters for electrostatic force calculation were the same as those for potential calculation. Then, the electrostatic force was split into the binding component and the sliding component, as shown in the following equations, respectively:

$$F_{Bi}=F_i\cdot \cos {\alpha }_i,$$ (1)

$$F_{Si}=F_i\cdot \sin {\alpha }_i,$$ (2)

where F_Bi and F_Si are the binding component and the sliding component in the i-th frame, respectively; F_i is the total electrostatic force on myosin at the i-th frame; and α_i is the angle between the total force and the axis, which passed the mass centers of the myosin motor domain and the actin filament in the i-th frame.

D. MM/PBSA free binding energy calculation

The binding energy (ΔE_bind) is calculated by the following equation, which was applied to Coulombic, van der Waals, polar solvation, and nonpolar solvation binding energies:

$$\mathrm{\Delta }{E}_{bind}=E_{actomyosin}-E_{myosin}-E_{actin},$$ (3)

where ΔE_bind is the binding energy (electrostatic, van der Waals, polar solvation, or nonpolar solvation binding energy); and E_actomyosin is the corresponding energy of the actomyosin complex, while E_myosin and E_actin are the corresponding energy of the myosin and the actin filament, respectively.

The free binding energy ΔE_{total_bind} was calculated by the following equation:

$$\mathrm{\Delta }{E}_{total\text{\_}bind}=\mathrm{\Delta }{E}_{cbind}+\mathrm{\Delta }{E}_{vbind}+\mathrm{\Delta }{E}_{pbind}+\mathrm{\Delta }{E}_{npbind},$$ (4)

where ΔE_{total_bind} is the free binding energy and ΔE_cbind, ΔE_vbind, ΔE_pbind, and ΔE_npbind are the Coulombic, van der Waals, polar solvation, and nonpolar solvation binding energies, respectively. The Coulombic energy and the polar solvation energy were calculated by Delphi.³¹ The van der Waals energy was calculated by NAMD 2.12.²⁹ The nonpolar solvation energy was calculated by the solvent-accessible surface area (SASA) using the following equation:

$$E_{np}=\alpha \cdot SASA+\beta ,$$ (5)

where E_np is the nonpolar solvation energy, α = 0.0054, and β = 0.92 kcal/mol. The SASA was calculated by NACCESS.³⁵

E. Salt bridges and hydrogen bonds

The salt bridge and hydrogen bond analyses were performed by VMD²⁸ based on 1000 frames from the simulations of 30–50 ns. The cutoff distance for salt bridge analysis was set as 4 Å,³⁶ while the cutoff distance and the angle for hydrogen bond analysis were set as 3.5 Å and 20°, respectively.^11,37

III. RESULTS AND DISCUSSION

A. Structural analysis

In previous works on myosin V,²³ the ATP hydrolysis led to the opening of the cleft, which is called the post-rigor state. In the post-rigor state, the myosin detaches from the actin filament.²³ The conformational change in the myosin from the rigor to the post-rigor state includes the movements of cardiomyopathy (CM) loop, loop 3, and loop 4.^10,23 The conformational changes caused by ATP hydrolysis between the rigor and post-rigor states are shown in Fig. S1. The cleft is open in the post-rigor state and closed in the rigor state. The simulations for the actomyosin complex in the rigor state reached stability after 10 ns, while those in the post-rigor state reached stability after 30 ns (Fig. S2). The separated RMSF (Root Mean Square Fluctuation) (each 10 ns for a stage) was also calculated to indicate the stabilization of the structure. In both complexes, stages 4 and 5 [Figs. S2(D) and S2(E)] show the lowest RMSF, which means the structure is stabilized. It is consistent with the result of RMSD (Root Mean Square Deviation). In addition, the slight movement of the post-rigor state myosin motor domain was also observed [Fig. S4(B)]. It indicates the instability of the post-rigor state.

The RMSD of repeated simulations also shows that all simulations reached stability after 30 ns (Fig. S3). There is a clear movement of myosin in the post-rigor state [Fig. S2(B)], which is caused by weak binding. The RMSF analysis for the myosin is shown in Fig. S4. The residues with high RMSF were GLN 571, THR 211, and GLN 734 in the rigor state myosin. The high RMSF of these residues is because of the formation of loops, which are very flexible [marked in Fig. S4(B)]. The GLN 571 is on loop 2, which is located on the interface and possesses high flexibility.

B. Electrostatic potential surface calculation

We took the structures of actomyosin complexes in the rigor and post-rigor states from the last frame (50 ns) for calculations of the electrostatic potential. The electrostatic potential surfaces are shown in Fig. 1, while the electrostatic field lines are shown in Fig. 2. The myosin and actin filament in both figures were separated by 20 Å for better visualization of the interfaces. In Fig. 1, the myosin-binding interface is positively charged, while the actin-binding interface is negatively charged. Compared with the myosin-binding interface in the rigor stage, the positively charged area decreased in the post-rigor state [Figs. 1(a) and 1(b)]. By contrast, the actin-binding interface was not affected by the state changes. In Figs. 1(c) and 1(d), tropomyosins are also negatively charged. They can interact with the positively charged myosin, contributing to electrostatic attraction. We further rendered the electrostatic field lines in Fig. 2 by VMD.²⁸ In Fig. 2, the decreased positively charged area in the myosin-binding interface directly caused fewer interactions with the actin filament in the post-rigor state. The decrease in interactions is shown in Figs. 2(c) and 2(d). Consistent with the electrostatic potential surface, the tropomyosins (Tpms) in Fig. 2 also interact with the myosins. This reflects our previous work on interactions between Tpms and myosins.³⁸ In detail, the electrostatic field lines between the myosin motor domain and tropomyosin are even denser than those with G-actin, especially in the post-rigor state. The electrostatic interactions between the myosin and actin filaments are thought to be important contributors to the binding or movements of myosin in the kinetic cycles. The weakened interactions in the post-rigor state contributed to the dissociation of the actomyosin complex.

FIG. 1.

Coulombic potential surface of the actomyosin in the rigor [(a) and (c)] and post-rigor [(b) and (d)] states.

FIG. 2.

Electrostatic field lines of actomyosin in the rigor [(a) and (c)] and post-rigor [(b) and (d)] states.

C. Electrostatic force analysis

In the rigor state, the actin filament acts an electrostatic force of 125.11 kT/e on the myosin; the electrostatic force decreases to 73.52 kT/e in the post-rigor state (Fig. S5). In Fig. 3, the forces were analyzed as the binding and sliding components [Eqs. (1) and (2)]. From the rigor to the post-rigor state, the binding component decreases [Fig. 3(a)] (from 120.24 to 61.62 kT/e) along with the decrease in the total force. By contrast, there is no obvious difference for the sliding component [Fig. 3(b)] from the rigor to the post-rigor state. Combined with analyses of the surfaces of the electrostatic potential, the conformational changes in the myosin from the rigor to the post-rigor state shrink the positively charged area and decrease the interactions with the actin filament. The decrease in the interaction further suggested the decrease in the electrostatic binding force. The decrease in binding helps detach the myosin from the actin filament.

FIG. 3.

Mass-center (a) and actin-filament (b) components of the electrostatic binding force between the myosin motor domain and the actin filament.

D. Free binding energy calculation

In addition to testing the electrostatic features, we further tested the free binding energy by MM/PBSA.^39,40 The MM/PBSA may result in higher binding energy than expected, but the comparisons of relative binding energy are reliable. In Fig. 4 and Table I, the actomyosin complex in the rigor state is at a lower state (−124.25 kcal/mol) than that in the post-rigor state (−114.78 kcal/mol). The difference infers that the binding in the rigor state is stronger than that in the post-rigor state. To further analyze the contributions among each component, the electrostatic, van der Waals, polar solvation, and nonpolar solvation binding energies were calculated separately [Eq. (4)]. The results are shown in Table I and Fig. S5. van der Waals and solvation binding energies were lower in the rigor state; this is preferable for binding. The polar solvation binding energy contributes most to the rigor state, which is 22.18 kcal/mol lower than in the post-rigor state. The difference is due to the conformational change in the post-rigor myosin, which decreases the area of the interface while extending the solvent-accessible surface area.

FIG. 4.

Free binding energy between the myosin motor domain and the actin filament.

TABLE I.

Free binding energy (kcal/mol).

	Electrostatic binding		van der Waals binding		Polar solvation binding		Nonpolar solvation binding		Free binding
		energy		energy		energy		energy		energy
Rigor		Post-rigor	Rigor	Post-rigor	Rigor	Post-rigor	Rigor	Post-rigor	Rigor	Post-rigor
Average	−1631.87	−1646.70	−80.83	−79.56	1603.37	1625.55	−14.91	−14.06	−124.25	−114.78
SD	58.86	52.18	14.12	8.73	51.92	48.26	0.82	0.88	14.28	13.81

E. Salt bridges and hydrogen bonds

To better understand the reason for the decrease in the binding affinity between the rigor and post-rigor states, the salt bridges and hydrogen bonds from the last 20 ns were calculated by VMD. The results are shown in Tables II and III. In the rigor state, there were eight salt bridges with occupancy higher than 0.5 and three salt bridges with occupancy higher than 0.9. By contrast, in the post-rigor state, there were five salt bridges with occupancy higher than 0.5 and one salt bridge with occupancy higher than 0.9. The positions of salt bridges with higher occupancy (>0.8) are shown in Fig. 5. Most salt bridges in the rigor state disappeared in the post-rigor state. However, the salt bridge (GLU 574-ARG 95) also appeared in the post-rigor state, and the occupancy was increased. Similarly, in the rigor state, a salt bridge was formed by LYS 367 and GLU 180 (Tpm X), while LYS 367 in the post-rigor formed a salt bridge with GLU 184 (Tpm X). These changes in salt bridges and related residues are related to the transformation from the rigor state to the post-rigor state. In the rigor state (Table III), there were nine hydrogen bonds with occupancy higher than 0.5 and one hydrogen bond with occupancy higher than 0.9. By contrast, there were six hydrogen bonds in the post-rigor state with occupancy higher than 0.5 and no hydrogen bonds with occupancy higher than 0.9. In summary, more salt bridges and hydrogen bonds appear in the rigor state than in the post-rigor state. The fewer salt bridges and fewer hydrogen bonds infer the higher binding energy and instability in the post-rigor state, contributing to the detachment of the myosin from the actin filament.

TABLE II.

Salt bridges between the myosin motor domain and the actin filament.

Rigor				Post-rigor
Myosin	Occupancy	Actin filament		Myosin	Occupancy	Actin filament
LYS367	0.994a	GLU180	Tpm X	LYS570	0.993a	GLU100	G-actin E
LYS413	0.958a	ASP175	Tpm Y	GLU574	0.869b	ARG95	G-actin E
GLU632	0.936a	ARG95	G-actin C	LYS640	0.836b	ASP24	G-actin C
GLU574	0.734b	ARG95	G-actin E	LYS367	0.728b	GLU184	Tpm X
LYS639	0.721b	ASP25	G-actin C	ARG369	0.545b	GLU192	Tpm X
LYS397	0.682b	GLU173	Tpm X	LYS635	0.441b	ASP25	G-actin C
GLU370	0.532b	ARG178	Tpm Y	LYS542	0.283	GLU167	G-actin C
ARG403	0.519b	GLU164	Tpm Y	GLU409	0.272	ARG178	Tpm Y
ARG369	0.229	GLU181	Tpm Y	LYS635	0.183	GLU334	G-actin C
GLU371	0.054	LYS328	G-actin C	LYS640	0.009	ASP25	G-actin C
LYS365	0.032	GLU184	Tpm X	ARG369	0.002	GLU195	Tpm X
ASP554	0.024	LYS50	G-actin E
LYS639	0.021	ASP24	G-actin C
GLU632	0.02	ARG28	G-actin C
ARG567	0.015	GLU99	G-actin E
ARG369	0.013	ASP311	G-actin C

^a >0.90.

^b >0.50.

TABLE III.

High-occupancy (>0.1) hydrogen bonds between the myosin motor domain and the actin filament.

Rigor			Post-rigor
Donor	Acceptor	Occupancy	Donor	Acceptor	Occupancy
G-actin C ARG95	Myosin GLU632	0.90a	G-actin E ARG95	Myosin GLU574	0.83
G-actin C SER350	Myosin GLU536	0.82	Myosin LYS570	G-actin E GLU100	0.79
G-actin C SER350	Myosin GLU536	0.82	G-actin C THR351	Myosin GLU536	0.78
Myosin LYS367	Tpm-X GLU180	0.71	G-actin C THR351	Myosin GLU536	0.77
G-actin E ARG95	Myosin GLU574	0.71	Myosin ARG369	Tpm-X GLU192	0.65
Tpm-Y ARG178	Myosin GLU370	0.68	Myosin LYS640	G-actin C ASP24	0.50
Myosin SER643	G-actin C ASP25	0.65	Myosin LYS367	Tpm-X GLU184	0.48
Myosin ARG369	Tpm-Y GLU181	0.55	Tpm-Y ARG178	Myosin GLU409	0.33
Myosin ARG403	Tpm-Y GLU164	0.53	Myosin ASN568	G-actin E GLU100	0.31
Myosin LYS639	G-actin C ASP25	0.44	Myosin LYS635	G-actin C ASP25	0.27
Myosin TYR410	Tpm-Y GLU163	0.40	G-actin E GLN49	Myosin ASP554	0.27
Myosin LYS413	Tpm-Y ASP175	0.38	G-actin C SER350	Myosin SER532	0.17
Myosin LYS397	Tpm-X GLU173	0.33	Myosin LYS542	G-actin C GLU167	0.16
G-actin C ARG28	Myosin GLU632	0.33	Myosin LYS635	G-actin C GLU334	0.13
Myosin LYS640	G-actin C ASP24	0.32	Myosin ARG369	Tpm-X GLU195	0.12
G-actin E GLN49	Myosin ASP554	0.27	G-actin E ARG95	Myosin HSD576	0.10
Myosin LYS639	G-actin C GLY23	0.14

^a >0.90.

FIG. 5.

High-occupancy (>0.8) salt bridge pairs between the myosin motor domain and the actin filament in the rigor (a) and post-rigor (b) states.

IV. CONCLUSION

The actomyosin complex includes a myosin and an actin filament. We assembled actomyosin complexes in the rigor and post-rigor states. Electrostatic features were calculated to analyze the difference between the two states. The MM/PBSA, salt bridges, and hydrogen bonds were also calculated to specify the differences between the two states. The results show that the positively charged area in the myosin interface decreases from the rigor to the post-rigor state, causing fewer interactions and lower electrostatic binding strength. In the MM/PBSA, free binding energy is higher in the post-rigor state. It infers that the post-rigor state is less conducive than the rigor state to the binding process. The fewer salt bridges and hydrogen bonds in the post-rigor state also suggest a lower binding affinity between the myosin domain and the actin filament. However, the free binding energy and the electrostatic binding force in the post-rigor state were still supportive of binding between the myosin and the actin filament. Generally, detachments were not seen in the simulations without the application of external forces. It suggests that the detachment between the myosin and the actin filament is assisted by other forces. These forces could be from the lever arms,²² the light chain, myosin-binding protein C,⁴¹ and so on. Our findings show that the binding affinity is lower in the post-rigor state than that in the rigor state. It is the important character for the dissociation of actomyosin complexes. In the future, we aim to simulate the whole kinetic circle of myosin and to deeply analyze the functions of key residues.

SUPPLEMENTARY MATERIAL

See the supplementary material for the following: Fig. S1: Structural comparison between the myosin motor domains in the rigor (forest green) and post-rigor (orange) states. Figure S2: RMSD (A) and the structural comparison between the structures at the 0 ns (pink and orange) and 50 ns (green and cyan) in the rigor (B) and post-rigor (C) models. RMSFs of Cα in the rigor (D) and post-rigor (E) models were calculated by each 10 ns in 5 stages. Figure S3: RMSD (A) of actomyosin in the rigor and post-rigor states for triple pairs of simulations. Figure S4: RMSF of Cα in the rigor and post-rigor models (A) and rigor model with marked residues with high RMSF (B). Figure S5: Total force between the myosin motor domain and the actin filament. Figure S6: Coulombic (A), van der Waals (B), polar solvation (C), and nonpolar solvation (D) free binding energies between the myosin motor domain and the actin filament.

AUTHOR DECLARATIONS

Conflict of Interest

The authors have no conflicts to disclose.

DATA AVAILABILITY

The data that support the findings of this study are available within the article and its supplementary material.