Probing the mechanical properties of ORF3a protein, a transmembrane channel of SARS-CoV-2 virus: Molecular dynamics study

Abstract

SARS-CoV-2-encoded accessory protein ORF3a was found to be a conserved coronavirus protein that shows crucial roles in apoptosis in cells as well as in virus release and replications. To complete the knowledge and identify the unknown of this protein, further comprehensive research is needed to clarify the leading role of ORF3a in the functioning of the coronavirus. One of the efficient approaches to determining the functionality of this protein is to investigate the mechanical properties and study its structural dynamics in the presence of physical stimuli. Herein, performing all-atom steered molecular dynamics (SMD) simulations, the mechanical properties of the force-bearing components of the ORF3a channel are calculated in different physiological conditions. As variations occurring in ORF3a may lead to alteration in protein structure and function, the G49V mutation was also simulated to clarify the relationship between the mechanical properties and chemical stability of the protein by comparing the behavior of the wild-type and mutant Orf3a. From a physiological conditions point of view, it was observed that in the solvated system, the presence of water molecules reduces Young’s modulus of TM1 by ∼30 %. Our results also show that by substitution of Gly49 with valine, Young’s modulus of the whole helix increases from 1.61 ± 0.20 to 2.08 ± 0.15 GPa, which is consistent with the calculated difference in free energy of wild-type and mutant helices. In addition to finding a way to fight against Covid-19 disease, understanding the mechanical behavior of these biological nanochannels can lead to the development of the potential applications of the ORF3a protein channel, such as tunable nanovalves in smart drug delivery systems, nanofilters in the new generation of desalination systems, and promising applications in DNA sequencing.

1. Introduction

The cells of living organisms are constantly exposed to various physicochemical environmental stimuli [1], [2], [3]. To date, significant advances have been made in understanding the direct relationship between cell receptors and intracellular electrochemical signaling. These receptors, which include a variety of voltage-gated, ligand-gated, and mechanosensitive ion channels, are responsible for the cell's perception of the physical and chemical stimuli around it [4], [5], [6]. Among these stimuli, mechanical forces, including osmotic pressure, gravity, touch, fluid-cell interactions, and sound waves, target various types of cell membrane proteins, causing intracellular changes and subsequent responses in the cell nucleus which is an essential part of the field of mechanobiology [7].

Since the outbreak of the SARS-CoV-2 virus in late 2019, which led to the pandemic of Covid-19 in the world, many efforts have been made by researchers in various fields such as virus propagation, vaccine development, drug design, and the field of diagnosis and treatment [8], [9], [10], [11], [12], [13], [14]. Besides all aspects of research about SARS-CoV-2, it is crucial to expand researchers' knowledge of the target proteins of the virus and to develop alternative therapies to reduce the resistance of new variants or viruses that may appear in the future [15]. Among the various proteins encoded in the structure of SARS-CoV-2, some of them, such as proteins E, 8a, and 3a, are so far known as ion channels [16], [17]. The crystal structure of the ORF3a protein was obtained by Cryogenic electron microscopy (cryo-EM) with a resolution of 2.1 Ã… [18]. As shown in Fig. 1 , the cryo-EM data reveals that this channel consists of three helices named TM1, TM2, and TM3 and a neighboring lateral helix (LH). Based on electrophysiological techniques and ion flux measurements, it is claimed that the ORF3a protein is an ion channel in the category of cation channels [18], [19], [20].

Fig. 1.

The ORF3a protein components in the side and top views obtained by cryo-EM. The ORF3a channel consists of three helices, TM1, TM2, and TM3. The lateral helix (LH) is located horizontally in the peripheral part of the protein and can be considered as a linkage between the protein and the bilayer membrane.

The study of ion channels is critical because a significant part of the biological behavior of living cells is directly related to these proteins. The presence of these channels is essential for hearing, vision, touch, bone and muscle growth, and blood pressure regulation, and malfunction of these channels is one of the main causes of various diseases and disorders, including irregular heartbeat, hearing loss, muscle resorption, cancers, and neurological disorders [21]. In the field of therapy, although currently, ∼15 % of clinically used drugs target ion channels [22], however, there are novel potentials for considering ion channels both as drug targets and as a complementary part of the smart drug delivery systems [1], [23], [24]. Therefore, investigating the mechanism of action of these channels and studying their components in detail can be a bridge to identifying the hidden dimensions of these biological nanostructures.

Up to now, various theoretical and experimental techniques have been used to study the mechanical behavior of biomolecules and biological material [25], [26], [27], [28], [29], [30]. Using molecular dynamics simulation and computational methods based on elasticity approaches, Bavi et al. have calculated the mechanical properties of the components of the Mechanosensitive ion channel of large conductance (MscL) of E-Coli and Mycobacterium tuberculosis and studied the function of these proteins under external mechanical forces [28]. These studies show that the function of these channels is significantly dependent on how the force is distributed in the cell membrane as well as the mechanical properties of the components of the MscL. The role of local curvatures, the effect of amphipathic molecules on the composition and structural order of the membrane, and the physical and chemical properties of the membrane have been studied by researchers to reveal the behavior and gating mechanism of the ion channels [31], [32], [33], [34], [35], [36], [37], [38].

Over the past years, ion channels have been put forward to be used as controlled release valves for specific and targeted drug delivery purposes. Recent research has shown that these ion channels, such as MscL, have the potential to be converted into nanovalves through the creative use of magnetic nanoparticles and magnetic forces [23], [24], [39], [40]. This has only been possible though when the external force was applied to the parts of the channel that has a key role in the gating process and a robust elastic range in response to external force before they unfold. As the gating mechanism of these channels becomes discovered, the root is being paved for the use of these channels as tunable nanovalves in smart drug delivery systems, as well as some novel applications in water filtration, desalination process, and biomolecules sequencing [41], [42].

As mentioned at the beginning of the Introduction, one of the ways to fight the spread of coronavirus is to identify effective proteins and study their mechanism of action. One of these proteins, known as the cationic ion channel, is the ORF3a channel and its components and sequence are shown in Fig. 1. Due to the presence of the ORF3a ion channel in infected tissue samples and the presence of ORF3a antibodies in the plasma of these patients [43], this protein was considered the basis of this study. Nowadays, it is known the ORF3a ion channel is involved in the viral functions of the virus [44] and cell death [45], and its removal reduces viral titers and disease in mice, increasing the likelihood that ORF3a could be an effective candidate for vaccine drug-target [46], [47]. These data indicate that the ORF3a plays a key role in the pathogenicity of Covid-19 [10], and therefore, it seems that knowing the behavior and physical/mechanical properties of the components of the channel provides useful information to researchers. In this study, performing all-atom steered molecular dynamics (SMD) simulations, the mechanical properties of the force-bearing components of the ORF3a channel are calculated in different physiological conditions.

2. Materials and methods

The Cryo-EM structure of the ORF3a ion channel in lipid nanodiscs (PDB ID: 7KJR) was used to extract the alpha helices TM1, TM2, TM3, and LH. The NAMD 3.0 package and CHARMM36 force field [48] were used for all molecular dynamic simulations. Visual Molecular Dynamics (VMD) and PYMOL were used to prepare simulation models, apply constraints and external forces, visualizations, and post-processing the output data.

To calculate the mechanical properties of channel components, the constant force (CF) mod of steered molecular dynamics (SMD) was preferred to the constant velocity (CV) because the CF method, despite the need for a significant number of simulations, is an equilibrium method and does not depend on the rate of the applied tension or stiffness constant of the dummy springs connecting the SMD atom [28].

For all cases, one end of the helix was fixed by applying harmonic constraints (with elastic constant: 12 kcal/mol/Ų), and an external axial constant force was applied to the other end to obtain the resultant strain of the elongated helix. All the cases were simulated in 300 K and 1 atmosphere, using Langevin dynamic and modified Nosé-Hoover Langevin piston. In the solvated cases, the physical TIP3 model was used to model water molecules. The details of the physical properties and bioinformatic calculations of each channel component are given in Table 1 .

Table 1.

Amino acid sequence and helical information of all the α-helices in ORF3a protein. The bioinformatics calculations are made by HeliQuest [49].

ORF3a alpha helices	TM1	TM2	TM3	LH
Amino acid sequence	FGWLIVGVALLAVFQSASK	KRWQLALSKGVHFVCNLLLLFVTVYSHLLLVAAG	FLYLYALVYFLQSINFVRIIMRLWLCWKCR	SKLREQLGPVTQEFWDNLEKETE
Length (Å)	28.4	50.3	46.2	33.1
Hydrophobicity < H>	0.84	0.80	1.02	0.16
Hydrophobic moment	0.16	0.27	0.11	0.45
Polar residue % (n)	31.6 (6)	35.3 (12)	23.3 (7)	69.57 (16)
Charged residue % (n)	5.2 (1)	8.8 (3)	13.3 (4)	39.1 (9)
Net charge (z)	1	3	4	−3

In all cases, the simulation systems reach equilibrium, and then the external forces are applied to the system. For each applying external force in the CF method, sufficient time (approximately 25–30 ns) is needed to determine the final deformation of the helix. The simulation steps (equilibrium and loading) were performed in NPT and NVT ensembles in all cases, respectively. Setting the NVT ensemble in the loading step is due to the fact that the change in the volume of the simulation system should not affect the length of the helix. If so, the change in the length of the helix will only be ​​ affected by the applied external force. In addition to monitoring of length change of the helix visually (See Fig. S1), the root mean square deviation (RMSD) diagram and the variations of the system’s energy are also considered to ensure the equilibrium (See Figs. S1-S3).

After the equilibrium of the system, the values of stress and strain corresponding to the applied force are recorded as a point of the stress–strain diagram. Dividing each force by the equivalent cross-sectional area of the helix (see Equation (2)), the equivalent stress value is obtained, and by dividing the elongation by the initial length of the helix, the equivalent strain is obtained. By logging the average of these values and applying other forces with different magnitudes, the sufficient number of stress–strain points in the stress–strain diagram are obtained and the necessary reliability is provided to calculate Young's modulus and elastic stiffness.

The slope of this diagram is the helix stiffness constant, K, which can be converted to the Young modulus of the helix, E, as [28].

$$E=k\times L_0/A$$ (1)

where L₀ and A are the initial length and the mean cross-sectional area of the helix, respectively. To calculate the cross-sectional area of the helix, it is necessary to calculate the gyration radius of the helix, R_gyr, as [28]:

$$R_{\mathit{gyr}}=\sqrt{{\frac{\sum \left(\left(\sqrt{x^2+y^2}\right)-\left(x_{\mathit{mean}}^2+y_{\mathit{mean}}^2\right)\right)}{N}}^2}$$ (2)

where N is the total number of atoms of the helix (excluding the hydrogen atoms), and ×, y, x_mean, and y_mean are the coordinates of each atom, and the coordinates of the helix cross-sectional center of mass, respectively.

It is worth noting that to get reliable mechanical properties of the ORF3a components in both vacuum and water-soluble environments, different values for the applied force (∼7 to 240 pN) are required. Taking into account the repetition of the simulations (to calculate the mean values of the results), more than 700 molecular dynamics simulations have been performed in this study.

3. Results and discussion

The results of applying mechanical forces to each solvated ORF3a channel component are presented in Fig. 2 . Using equation (2), the value of the gyration radius of the studied helices is calculated to be about 2.35 Å on average. According to this value of the radius, the cross-sectional area of the helices is 17.35 Å². With the help of this value and Equation (1), the mechanical properties of the helixes can be calculated.

Fig. 2.

Stress–strain curve of ORF3a ion channel solvated components, A) TM1, B) TM2, C) TM3, and D) LH. The horizontal lines represent the standard error of the mean (SEM), and the slope of the graph is drawn with 95 % confidence from the stress and strain axes. Based on these graphs, Young's modulus, E, of TM1- TM3 and LH helices are obtained as 1.6 ± 0.2, 4.9 ± 0.8, 3.1 ± 0.35, and 2.2 ± 0.1 GPa, respectively (Mean ± SEM, n = 3).

All calculations are based on at least three simulation repetitions for each applied force. In Fig. 2, the stress–strain curves of each helix are plotted. Blue circles represent the average value of three runs for each applied force. Horizontal lines represent the standard error of the mean (SEM), and the slope of the graph is drawn with 95 % confidence from both stress and strain axes. For helix TM1, with external stretching tension in the range of 0.1–1.65 kcal/mol/Å (i.e., from ∼7.0 to 115 pN), Young's modulus of this helix was obtained as 1.6 ± 0.2 GPa. As can be seen from part B, the slope of the stress–strain curve of the TM2 helix was obtained by applying forces in the range of 0.1–2.0 kcal/mol/Å (∼7.0–140 pN), and gives E_TM2= ​​4.9 ± 0.8 GPa. For the TM3 helix, the constant forces in the range of 0.1–2.3 kcal/mol/Å (∼7.0–160 pN) were applied to the system, and Young's modulus of this helix was obtained as 3.1 ± 0.35 GPa. Young's modulus of the LH helix was calculated as 2.2 ± 0.1 GPa by exerting external forces in the range of 0.1–1.6 kcal/mol/Å (∼7.0–110 pN). It should be noted that these forces are within the range of biological forces. Obviously, in complex cellular processes, the channels will be affected by such values with a high probability [23].

In Fig. 3 , the elastic constant and Young's modulus of the ORF3a α-helices for the solvated system using the steered molecular dynamic simulation are listed. As can be seen from the data, the lowest value of the helix stiffness (elastic constant), K, is 9.7 pN/Å (for helix TM1), and the highest value is 16.9 pN/Å (for helix TM2). Also, Young's modulus, E, values ​​vary between 1.6 and 4.9 GPa for different α-helices, and the lowest amount of Young's modulus belongs to the TM1 helix, and the highest amount of Young's modulus belongs to TM2 helix. As can be clearly seen from these results, Young's modulus of the TM2 helix is ​​about three times harder than that of the TM1 helix. This significant distinction can be due to the difference in physical properties, including the length and electrostatic charges of these two helices (see Table 1). Of note is that the significant stiffness of helix TM2 compared to the TM1 helix has been observed and reported before for other protein channels such as MscL channels of E.coli and Mycobacterium tuberculosis bacteria [28].

Fig. 3.

Mechanical properties of ORF3a α-helices in the presence of water. The constant-force (CF) steered molecular dynamic simulation was used here to estimate (A) the stiffness (elastic constant), K, and (B) Young’s modulus, E, of the helices. The elasticity constants are between 9.7 and 16.9 pN/Å (Young’s moduli are between 1.6 and 4.9 GPa). The values of Young’s moduli are Mean ± SEM for n = 3.

For a better understanding of the difference in mechanical stiffness of TM2 and TM1, we investigated the structural change of TM1 and TM2 helices during the simulation time based on the RMSD as well as the number of hydrogen bonds. As can be seen from the Figs. S4-S7, the number of hydrogen bonds breaking in Helix TM1 during the stretching process (for F = 1.5 kcal/mol/Å) is much higher than that of TM2. This means that by applying the tensile force, the number of hydrogen bonds is reduced, and because the hydrogenation bonds bear a significant part of the applied load, reducing their number leads to more softening of the structure and a decrease in Young's modulus. Since the hydrogen bonds are a function of the distance between the donor and acceptor atoms, the changes in the RMSD over time also confirm the decrease in the number of hydrogen bonds, especially in larger applied forces.

3.1. Effect of hydration on mechanical properties

It is clear that the behavior of biological molecules, including the constituent components of the channel, can be completely different in different environments due to the peptide net charges, hydrophobic/hydrophilic residues, and the dominant role of hydrogen bonds in the protein structure under mechanical loading. As mentioned in the Introduction, protein channels have various potential applications. Therefore, it is a good idea to investigate the effect of the protein’s environment on the mechanical behavior of the helices.

This research is specifically focused on two environments, water and vacuum. These environments have been taken into account because, in the closed-state of the channel, the internal helices forming the wall of the central pore of the channel (pore-lining helices) have no direct contact with water molecules. But as the opening process of the channel under physicochemical stimuli initiates, some water molecules enter the channel, and the internal helix gets wet. Bavi et al. [28] have shown that this process, which can significantly increase the opening speed of MscL channels, is directly related to the changing the mechanical properties of the pore-lining helices in the presence or absence of water molecules. Hence, it is essential to examine the difference in the elasticity of pore-lining helices in aqueous and vacuum environments.

For this purpose, the mechanical properties of one of the pore-lining helices of the channel (TM1) have been calculated in the vacuum system. To simulate the mechanical behavior of the helix in the vacuum environment, external forces up to 3.5 kcal/mol/Å (approximately equivalent to 240 pN) have been applied. In Fig. 4 , the stress–strain curve of the TM1 in the vacuum system.

Fig. 4.

The stress–strain curve of the TM1 helix in the vacuum system (external forces up to 3.5 kcal/mol/Å). (Mean ± SEM, n = 3). To show more clearly the deviation values from the data mean, a sample area of the graph is shown as a magnified inset.

By measuring the slope of the diagrams in Fig. 4, Young's modulus value of the TM1 helix can be calculated in the vacuum. The slope of the graph with 95 % confidence from both stress, and strain axes gives the value of Young's modulus as 2.32 ± 0.15 GPa. In Fig. 5 , Young's modulus values of the TM1 helix in the vacuum and aqueous systems are compared.

Fig. 5.

The effect of hydration on the mechanical properties of TM1 helix. The Young's modulus of the helix is 2.32 ± 0.15 GPa and 1.61 ± 0.20 GPa for vacuum and aqueous systems, respectively. The student’s t-test was applied in the systems, and the difference was considered significant for *p-value < 0.05.

Although various parameters can affect the elasticity of helices in the presence of water, it seems that the role of the formation/breaking of hydrogen bonds in the protein structure is more prominent than all other factors. To investigate this issue, the number of hydrogen bonds of the TM1 helix has been monitored during the simulation for each vacuum and aqueous system. As Fig. 6 shows, the average number of hydrogen bonds in the vacuum system is about 20. In contrast, this number is 15 for the aqueous system. This result shows that the effect of water molecules on the mechanical behavior of helices is quite apparent. The reduction of mechanical strength and softening of the helix in the presence of water molecules have already been shown in similar studies on spectrin-like proteins, immunoglobulin, and collagen microfibrils [28], [50], [51]. So, in applications where protein channels may operate in non-aqueous environments (for example, the use of channels as controllable nanovalves for the targeted release of drugs soluble in non-aqueous fluids), knowing the effect of the environment on the channel's behavior will be beneficial to increase the efficiency of the designed system.

Fig. 6.

Comparison of the average number of hydrogen bonds of the pore-lining helix in the vacuum and hydrated systems. An angle cut-off of 40° and a donor–acceptor distance of 3.5 Å were used for hydrogen bond calculations. The displayed time corresponds to the last 10 ns of the simulations.

3.2. Chemical stability and mechanical properties

According to the available reports, various genetic mutations in the ORF3a protein have been identified so far, and its genome is constantly evolving by acquiring mutations [52], [53]. These mutations significantly change the physicochemical stability of the protein, and even in some cases, they can change the structure and function of the protein. The purpose of this part of the research is to investigate the relationship between the free chemical energy of protein and its elasticity properties. For this purpose, a series of all-atom MD simulations were performed to obtain the mechanical properties of the G49V mutant TM1 helix, and the results have been compared with the wild-type helix. This particular mutation was chosen because it would result in the most significant positive change in the free energy difference (ΔΔG) of the protein [54].

The mutant (TM1_G49V) helix was modeled using web-based SWISS-MODEL software [55], and more than 200 independent simulations have been performed to obtain the mechanical properties of this helix for both water and vacuum systems.

In Fig. 7 , Young's modulus of the TM1_G49V and wild-type TM1 are compared for both solvated and vacuum systems. As it is clear from the data in this figure, for the vacuum system, Young's modulus of TM1G49V is 2.94 ± 0.20 GPa, nearly 25 % higher than the modulus of the wild-type TM1 (2.32 ± 0.15 GPa). This increase is also observed for the solvated system. Considering these results, the mutual relationship between chemical stability and mechanical properties of helices will be pretty manifest. This relationship is regarded as an essential parameter for studying the ORF3a protein channel, especially in the design of drug delivery systems, fluid impurity filtration, desalination, and other potential applications.

Fig. 7.

The Young's modulus of the mutant (TM1_ G49V) and wild-type TM1 for hydrated and vacuum systems. The Young's modulus of TM1_G49V is obtained as 2.94 ± 0.20 and 2.08 ± 0.15 GPa for vacuum and hydrated systems, respectively. These values are nearly 25 % higher than those of the wild-type TM1 (2.32 ± 0.15 and 1.61 ± 0.20 GPa for vacuum and hydrated systems, respectively). The values of Young’s moduli are Mean ± SEM for n = 3. The student’s t-test was applied in the systems, and differences were considered significant for *p-value < 0.05.

4. Conclusion

Performing all-atom steered molecular dynamics simulations, the mechanical behavior of ORF3a channel components under the physiological range of external forces was studied. The results showed that the lowest Young's modulus corresponds to the TM1 and the highest belongs to the TM2. This significant distinction can be due to the difference in physical properties including the length and electrostatic charges of these two helices. To investigate the relationship between mechanical properties and chemical stability, independent simulations were performed to model and calculate the properties of the TM1_G49V mutant. It was seen that mutant that has experienced more free energy changes have more robust elasticity properties. The effect of the environment on the elasticity of the pore-lining α-helix was also investigated by comparing the results of the mechanical behavior of the helix in two aqueous and vacuum systems. It was observed that in the solvated system, the presence of water molecules breaks the hydrogen bonds of the helix and reduces Young’s modulus of TM1 by ∼30 %. The biophysical consequences of these findings can be helpful for understanding the physical principles underlying the mechanotransduction process of the ORF3a with the participation of other cellular components. Obviously, in addition to finding a way for fighting against Covid-19 disease, understanding the mechanical behavior of these biological nanochannels can lead to the development of their potential in other applications such as tunable nanovalves in smart drug delivery systems, nanofilters in the new generation of desalination systems, and promising applications in biomolecules sequencing. Although still far from being implemented, this is another potential path for further research and development of novel applications of ORF3a protein channel in various fields such as tunable nanovalves in smart drug delivery systems, nanofilters in the new generation of desalination systems, and promising applications in biomolecules sequencing.

CRediT authorship contribution statement

Vahid Mahmoudi Maymand: Methodology, Software, Formal analysis, Validation, Visualization. Omid Bavi: Conceptualization, Methodology, Validation, Supervision, Project administration, Writing – review & editing. Abbas Karami: Methodology, Validation.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

The following are the Supplementary data to this article:

Data availability

No data was used for the research described in the article.