Probing the binding nature and stability of highly transmissible mutated variant alpha to omicron of SARS‐CoV‐2 RBD with ACE2 via molecular dynamics simulation

Currently, no approved drug is available as a causative agent of coronavirus disease 2019 (COVID‐19) except for some repurposed drugs. The first structure of severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) was reported in late 2019, based on that some vaccines and repurposed drugs were approved to prevent people from COVID‐19 during the pandemic situation. Since then, new types of variants emerged and notably, the receptor binding domain (RBD) adopted different binding modes with angiotensin‐converting enzyme 2 (ACE2); this made significant changes in the progression of COVID‐19. Some of the new variants are highly infectious spreading fast and dangerous. The present study is focused on understanding the binding mode of the RBD of different mutated SARS‐CoV‐2 variants of concern (alpha to omicron) with the human ACE2 using molecular dynamics simulation. Notably, some variants adopted a new binding mode of RBD with ACE2 and formed different interactions, which is unlike the wild type; this was confirmed from the comparison of interaction between RBD‐ACE2 of all variants with its wild‐type structure. Binding energy values confirm that some mutated variants exhibit high binding affinity. These findings demonstrate that the variations in the sequence of SARS‐CoV‐2 S‐protein altered the binding mode of RBD; this may be the reason that the virus has high transmissibility and causes new infections. This in‐silico study on mutated variants of SARS‐CoV‐2 RBD with ACE2 insights into their binding mode, binding affinity, and stability. This information may help to understand the RBD‐ACE2 binding domains, which allows for designing newer drugs and vaccines.


| INTRODUCTION
Covid-19 is an infectious disease, which was first reported in Wuhan, China on December 2019 and created high pandemic situation in worldwide. This disease is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and it is called as coronavirus disease 2019 (Covid- 19). As of September 2022, globally 6.5 million people were died due to Covid-19 (https://covid19.who.int/). To date there is no clinically approved drug is available to cure Covid-19. To fight against this disease, vaccination developments and repurposed drugs are the immediate solution to resolve this problem. Moreover, the main challenge in the current pandemic situation is, facing the newly generating different variants of SARS-CoV-2 virus, accordingly design new vaccines and drugs, and gain immunity for the affected people. SARS-CoV-2 virus is a single-strand RNA virus, and the genome encodes four structural proteins such as Spike (S) protein, Small Envelope (E) protein, Membrane (M) protein, and Nucleocapsid protein. Spike protein is a type I fusion protein that builds trimmers on virion surface. It has two subdomains such as S1 and S2; S1 domain is responsible for receptor binding which includes receptor binding domain (RBD) and S2 domain is for membrane fusion. 1 Currently, spike (S) protein is undergoing different structural mutations, and it is very important to investigate the structural and biological significance of mutations. 2 Viruses are regularly changing their structure called as mutation. When a virus has one or more mutations on its structure is known as a variant of wild type/original virus. In viruses, generally, the mutations occur in their surface proteins; this can significantly modify the viral infection function as well as the interactions with the neutralizing antibodies. In RNA viruses, the rate of mutations are higher than the DNA viruses; mutations are the significant building blocks of evolution of new infections. 3,4 During the infection, SARS-CoV-2 RBD of S-protein binds with the human angiotensin-converting enzyme 2 (ACE2) protein form SARS-CoV-2 RBD-ACE2 complex. Different mutations in SARS-CoV-2 RBD cause new infections when they bind with ACE2. Understanding the structure and conformation of this RBD-ACE2 complex, and the interactions of different mutated SARS-CoV-2 RBD with the ACE2 are very crucial for the design of new drugs or vaccines. Mutated form of this virus emerges as a new form virus. In the present study, we have investigated the structure, interaction and the dynamic nature of reported different mutated forms of virus such as alpha to omicron SARS-CoV-2 RBD with ACE2 through in silico analysis. The SARS-CoV-2 RBD-ACE2 complex structure was reported and the coordinates and structure factor files are deposited in the Protein Data Bank (PDB); which is also similar to the previously reported SARS-CoV RBD-ACE2 complex structure. 1,5,6 The evaluation of SARS-CoV-2 has been characterized by the development of mutations, in the context of "variants of concern," that effective virus natures, including transmissibility and antigenicity, apparently in response to the immune profile modify the human population. [7][8][9] The transmissibility and infecting nature of COVID-19 is relatively much higher than the SARS-CoV and MERS viruses, this made a situation of pandemic of respiratory syndromes in worldwide. 10 To understand the infections and pathogeneses nature of different variants of SARS-CoV-2 the in-vivo, in-vitro and clinical studies have been conducted 8,[11][12][13][14][15][16] and still the same is also in progress. However, there is no detailed structural report about the dynamical nature and interactions of the alpha to omicron SARS-CoV-2 RBD-ACE2 complex of different SARS-CoV-2 variants. Importantly, the knowledge of dynamical binding nature of RBD with ACE2 is also essential to identify the drug targets for the variants which has mutations, this often alter the binding mode of virus and entry mechanism to the host cells. Several reports outline the evidence of conformational change of binding domain that occurs due to the mutations in SARS-CoV-2 variants; these previous studies which demonstrate the same with their experimental and computational reports. [16][17][18][19][20] Protein structure of RBD bound with ACE2 have been reported (PDB code: 6M0J) and the binding nature also revealed by J Lan et al which states that the sequence of this protein is highly similar to the SARS-CoV RBD and has improved binding nature. 6 P Han et al reported an insights into RBD-ACE2 binding for emerging SARS-CoV-2 variants which characterize the binding using cytometry. 21 As per the reported study, the bindings of Alpha (82.7%), Beta (48.1%), Gamma (57.7%) are higher than the wild type (33.4%). In this study the binding sites, poses and its affinities of selective COVID-19 variant of concern such as B. are changed its binding sties with ACE2 during the simulation confirmed by the intermolecular interactions analysis. The present study also explores the analysis of their structure, intermolecular interactions and conformational modification using the molecular dynamics simulation and binding free energy studies of alpha to omicron variants (mutated) of SARS-CoV-2 RBD bound with human ACE2 from; and the results of all the mutated complexes are compared with the wild type structure.
Recently, the B.1.1.7 variant (alpha) was dominant in UK, reports mapped the structure and functional analysis of mutant N501Y of RBD region of S-protein and outline that it is harder to neutralize with high transmission property. 22 The nine modifications of S-protein amino acids of this variant are, deletion of 69-70 and 144, N501Y, A570D, D614G, P681H, T716I, S982A, and D1118H 23,24 ( Figure 1). And the study also found that the N501Y residue mutation of RBD enhanced the binding affinity to the murine and human ACE2 receptor and this mutation plays a key role in virus realization and neutralization. [25][26][27] Studies show that during the pandemic period, this variant raised the death rate to 61%, which is higher than previously reported variants and also caused severe illness to people. 28 Further, the B.1.351 variant (beta) was identified in South Africa on October 2020, found this variant has some deletion and substitutions in S-protein. The mutations occurred in S-protein as D80A, deletion of L242-244, K417N, E484K, N501Y, D614G, and A701V; these mutations in RBD gives tight binding to ACE2 and widespread escape from monoclonal antibody neutralization ( Figure 2). Some reports also indicated that the B.1.351 may increase the risk of disease infection in immunized individuals. 24,29 The next variant B.1.617 reportedly known as an Indian variant, which was first identified in India on February 2021, in this variant, the substitutions in S-protein are L452R, E484Q, and D614G ( Figure 3). The main attribute of this variant is, potential for neutralization reduction by some monoclonal antibody treatments and reduces the neutralization by the post vaccinations. Although a result of the reduced neutralizing capacity of sera is caused an increase in breakthrough. [30][31][32][33] The variant B.1.617.2 is also an Indian variant named as delta variant, in which the S-protein substitutions are T19R and G142D and the deletions are E156 and F157; R158G, L452R, T478K, D614G, P681R, and D950N substitutions ( Figure 4). This variant has some main attributes, such as increased transmissibility, reduced neutralization by post vaccination sera and monoclonal antibody treatments. 24 The early in vitro studies reveal the influence of delta variant, largely it is attributed to the effect of combination of evasion of neutralizing antibodies in previously infected persons and its increased virus infectivity. 13 Figure 5). This variant was characterized as highly infectious and reduces the potency of vaccinations. Further, this mutation also associated with the resistance to recently developed monoclonal antibody drug treatments (https://www.cdc.gov/coronavirus/2019ncov/variants/variant-info.html). 34 The B.1.618 (triple mutant) was identified in December 2020 in West Bengal, India; the mutations presence in the S-protein are deletion of Y145, H146, E484K, and D614G 31 ( Figure 6). In RBD, there is only one residue has been mutated (E484K). Reportedly, the S-protein of this variant B.1.618 neutralized with 2-5 fold cutback by recovering sera and vaccine-elicited antibodies. 32,33 The P.1 variant (gamma) was initially identified in Amazonas state, Brazil and Japan. Generally, this variant is also known as Brazil variant. The substitutions in the S-protein of this variant are L18F, T20N, P26S, D138Y, R190S, K417T, E484K, N501Y, D614G, H655Y, and T1027I 36 (Figure 7). And the main attributes are, significant reduction of susceptibility to the combination of bamlanivimab and etesevimab monoclonal antibody treatments and less neutralization by convalescent and post vaccination sera collected during the previous epidemic waves. [37][38][39] The B.1.1.529 variant (omicron) is the recent emergence in worldwide countries, which consists of more than 30 mutations in the S-protein. Particularly, in RBD of S-protein mutated with G339D, S371L, S373P, S375F, K417N, N440K, G446S, S477N, T478K, E484A, Q493R, G496S, Q498R, N501Y, and Y505H ( Figure 8). Recent reports outline that any mutations in the S-protein of RBD would cause immediate concerns about the efficacy of existing vaccines, mAbs and the potential of reinfection. 40,41 On compared with the other variants of emerged strains, relatively this variant exhibits a greater number of mutations which includes deletions, insertions and substitutions in S-protein residues.
In the present study, we performed the molecular dynamics simulations and binding free energy

| Protein preparation
The crystal structure of the complex form of SARS-CoV-2 S-protein RBD with ACE2 has been downloaded from the protein data bank (PDB Code: 6M0J). 6 As per the variants data reports in corresponding RBD domain sequences (333-526), the wild type residues were replaced with mutated residues by using residue mutate option in Maestro panel. The mutated RBD residues are B. In this phase, missing atoms has been added, ionization states were allocated and the atom types, bond orders tautomer were adjusted to optimize the hydrogen bonding networks. In the initial step of protein preparation, the structure of variants has been pre-processed by Prime module to predict the protein structures and proceed for optimization. Further, the optimized protein structures were energy minimized with OPLS4 force field. 43

| Molecular dynamics simulations
The energy minimized protein structures are subjected for molecular dynamics (MD) simulation to understand their conformational stability, intermolecular interactions, binding mode and binding energy. In Schrödinger Maestro package, Desmond system builder module was used to prepare the complex of alpha to omicron variant of SARS-COV-2 RBD with ACE2 for MD simulation. In this process, for each complex a predefined TIP3P orthorhombic 10x10x10 Å water box was chosen as solvent model. 44 To neutralize each complex variant, the Na + /Cl − ions were added into the TIP3P solvent box and all the MD systems were prepared with OPLS4 force field for MD simulation. Further, the MD simulation was performed for each wild and mutated complexes with NPT ensembles at the temperature 310 K and 1 atm pressure with the time scale of 500 ns using Desmond MD package module incorporated with the Schrödinger. 45 Further, the RMSD, RMSD plottings and binding free energy calculations have been done to understand the conformational modifications of complexes during the MD simulation. And the intermolecular interactions of each complex have been analyzed and compared the same at before and after MD simulation. Its helps to understand and identify the modified binding sites of some variants as presented in the results and discussion.

| Binding free energy
In biological system, most of the biological activities are regulated by the intermolecular interactions between the proteins components present in each cell. These intermolecular interactions play an important role in the protein-protein contact; the strength of these interactions can be estimated from the binding free energy of protein-protein interactions. In the present study, the binding free energy of RBD-ACE2 for wild and all mutated complex of alpha to omicron have been calculated. PRODIGY online (PROtein binDIng enerGY) web server was used to calculate the binding free energy (ΔG) values 46 reported here. For this analysis, both initial and end frame of the MD simulation of all the complexes were taken as pdb format file. While uploading the pdb files into the server, the chains were identified and the temperature was fixed with 37°C. Further, the binding free energy was calculated between RBD and ACE2 of all the variants which we have taken to study, in both energy minimized state (after mutation) and MD simulation state. And also, these results were compared with the wild type to understand its binding affinities.

| Intermolecular interactions of wild and mutated SARS-CoV-2 RBD with ACE2
The MD simulations were performed for wild and mutated SARS-CoV-2 RBD with ACE2 complex B.  (Tables S1 and S2). The mutation in SARS-CoV-2 RBD altered the binding mode of the virus with the human ACE2 protein, this may affect the binding affinity between RBD and ACE2. On comparing the interactions and the binding energy of each mutated variant of the SARS-CoV-2 RBD-ACE2 complex with the wild SARS-CoV-2 RBD-ACE2 complex allows to understand the binding nature and dynamics of mutant RBD-ACE2 complex. This allows to predict the variability of transmission of differently mutated RBD-ACE2. In the present study, the MD simulation was performed for each RBD-ACE2 complex up to 500 ns.

| Wild type: Interactions between the wild SARS-CoV-2 RBD and ACE2
In before the MD simulation, the wild SARS-CoV-2 RBD with ACE2 complex was energy minimized. Furthermore, the MD simulation was performed at 500 ns time scale for the wild SARS-CoV-2 RBD with ACE2 complex. The interactions of MD simulated wild complex have compared with the corresponding energy minimized complex of wild complex. The MD simulation shows that, some of the interactions exist between RBD and ACE2 in the initial state of RBD-ACE2 complex are more stable during the simulation and the interactions are Asn487···Gln24, Tyr489···Tyr83, Thr500···Asn330, Gln498···Lys353, and Gly502···Lys353 and the interaction distances are 1.94, 1.86, 3.03, 1.82, and 1.84 Å, respectively (energy minimized state) as shown in Figure 9.
During the MD simulation, the above interactions were maintained; the corresponding distances in the final frames are 1.98, 1.86, 3.02, 2.0, and 2.06 Å respectively. And these interactions are also comparable with the reported complexes by Jun Lan et al. 6 3.3 | Alpha variant (B.1.1.7): Interactions between alpha variant of SARS-CoV-2 RBD and ACE2 The recent report outlines that there are nine modifications among the amino acids of S-protein of alpha variant, this includes the deletion of 69-70 and 144, and the mutants N501Y, A570D, D614G, P681H, T716I, S982A, and D1118H. 25,29 Notably, the mutation of N501Y residue within the RBD shows enhanced binding affinity towards the murine and human ACE2 receptor. 1,14 The energy minimized mutated RBD-ACE2 complex shows that the RBD forms hydrogen bonding interactions with ACE2 ( Figure 10) and the interactions are Asn487···Tyr83, Gly496···Lys353, Asn487···Gln24, and Tyr489···Tyr83; these interactions are largely different from the interactions found in wild type. However, the Tyr489···Tyr83 interaction found in wild type also present in this alpha variant B.1.1.7. However, this interaction was disappeared during the MD simulation and resulted with the interactions Thr333···Glu589, Glu340···His239, and Lys444···Asp609, the corresponding interaction distances are 2.08, 1.84, and 1.95 Å, respectively. Overall, it indicates that the above said mutations are significantly modified the binding mode of RBD of Sprotein with the ACE2. This could be the reason that the alpha variant B.1.1.7 has high transmission ability into the human body.

| Beta variant (B.1.351): Interactions between the beta variant of SARS-CoV-2 RBD and ACE2
This variant was identified in South Africa and it is called as beta variant. This variant has some deletions and substitutions in S-protein, namely D80A, ΔL242-244, K417N, E484K, N501Y, D614G, and A701V. Beta variant is the mutated form of the wild type SARS-CoV-2 RBD. In the present study, the MD simulation was performed for this mutated RBD-ACE2 complex. In prior to the MD simulation, the complex has been energy minimized. In the energy minimized mutated RBD-ACE2 complex, the RBD forms interactions with ACE2 ( Figure 11) and the interactions are Asn487···Tyr83, Gly496···Lys353, Tyr449···-Asp38, Tyr489···Tyr83, and Gly502···Lys353, the interaction distances are 1.64, 2.68, 1.83, 1.86, and 1.94 Å, respectively. Among these interactions, notably, some of them are same as found in the wild type. As found in wild type, the Asn501 residue forms interactions with Lys353 and beta mutant (N501Y) at the distance 2.49 and 1.97 Å respectively. During the MD simulation, the interactions in the complex have been modified in different binding regions, and the new interactions are Thr333···Glu589, Asn334···Thr593, and Glu340···Lys596 and the corresponding distances are 1.96, 2.23, and 2.58 Å, respectively. These interactions are neither comparable with wild type nor with the above energy minimized complex; perhaps, this difference may be attributed to the large motion and conformational modification of amino acids of binding region during the MD simulations; this leads the RBD to form new interactions with ACE2. However, the presence of the above said mutations in RBD provides tight binding (interactions) with ACE2 and this facilitates for widespread escape from monoclonal antibody neutralization. The previous results indicate that the B.1.351 may increase the risk of disease infection among immunized individuals. 24,28

| Delta variant (B.1.617): Interactions between the delta variant of SARS-CoV-2 RBD and ACE2
The variant B.1.617 is known as Indian variant, which was first identified in India. The S-protein substitutions of this variant are L452R, E484Q, and D614G. The main attribute of this variant is, potential neutralization reduction by some monoclonal antibody treatments and reduced neutralization by the post vaccinations, although an increase in breakthrough infections may arise as a result of the reduced neutralizing capacity of sera. [28][29][30][31] This Indian variant is also known as double mutant, which is mutated in the RBD region of S-protein. Before perform the MD simulation, the structure of this mutated RBD with ACE2 complex was energy minimized. In the energy minimized complex, the RBD forms several interactions with ACE2, some of the interactions are Lys417···Asp30, Thr500···Tyr41, Gln498···Lys353, Asn501···Lys353, Tyr449···Asp38, Asn487···Gln24, Gln498···Asp38, Gly502···Lys353, and Tyr505···Glu37, and these are very strong, the corresponding distances are 1.98, 2.69, 2.35, 3.03, 1.78, 2.27, 2.21, 1.97, and 2.01 Å, respectively. Further, the MD simulation was carried out for this mutated RBD-ACE2 complex. Surprisingly, all the above said interactions are intact even after the MD simulation, however the distances are slightly modified, the new interaction distances are 2.65, 1.98, 2.67, 2.99, 1.87, 2.05, 2.28, 1.62, and 1.85 Å, respectively. Among these, notably, the Asn487···Gln24 and Gly502···Lys353 interactions are similar to the interactions found in the wild type RBD-ACE2 complex as well as the minimized state of the mutated structure ( Figure 12); interestingly, these interactions are highly stable during the MD simulation.

| Delta variant (B.1.617.2): Interactions between this delta variant of SARS-CoV-2 RBD and ACE2
The variant B.1.617.2 is also an Indian variant and it is another delta variant exhibits different substitutions in the S-protein, namely T19R, G142D, R158G, L452R, T478K, D614G, P681R, and D950N, and the deletion of E156 and F157. Notably, this variant has some attributes, such as increased transmissibility, reduced neutralization by some post vaccination sera and monoclonal antibody treatments. 24 The in vitro studies reveals that the influence of the delta variant has been mostly directed by a combination of evasion of neutralizing antibodies in the previously infected persons and increased virus infectivity. 12,[32][33][34] The results of MD simulation studies of mutated RBD-ACE2 complex ( Figure 13) shows that the Arg346···Glu483, Lys444···Asp609, Asn343···Ser602, and Thr345···Val604 interactions exist between RBD and ACE2 and are found strong, the corresponding interaction distances are 1.84, 2.06, 1.86, and 1.66 Å, respectively. Due to mutation in the in the RBD, the interacting amino acids with the ACE2 of wild complex is different; furthermore, during the MD simulation large conformational modification took place in both RBD and ACE2 regions, this can be well understood when compare with the results of the above said energy minimized delta variant RBD-ACE2. And it is also noteworthy that the mutated residues L452R and T478K are not forming any interaction with the ACE2.

| Delta plus variant (B.1.617.2.1): Interactions between the delta plus variant of SARS-CoV-2 RBD and ACE2
The variant B.1.617.2.1 (delta plus/AY.1) also an Indian variant, which was identified in India. This variant exhibit one main mutation in S-protein known as K417N and particularly this mutation is also found in the RBD region and this is mutated from the Indian variant B.1.617.2. Furthermore, this variant was characterized as reduced potency for vaccinations and it is highly infectious. The results of MD simulation of mutated RBD-ACE2 complex reveals that RBD forms strong Tyr489···Tyr83, Gln493···Glu35, and Thr500···Tyr41 types of interactions with ACE2, the interaction distances are 1.77, 1.94, and 1.97 Å, respectively ( Figure 14). Interestingly, these interactions are also found in the energy minimized complex, where the interaction distances are 2.58, 2.04, and 1.80 Å, respectively; this confirms that during the MD simulation these interactions are intact and the distances are not much altered. Whereas with the wild type, the only Tyr489···Tyr83 interaction is commonly found and no other interaction is comparable. The results of MD simulation indicate that this mutated RBD forms closer interactions with the neighboring amino acids of ACE2.

| Triple mutant (B.1.618): Interactions between the variant triple mutant SARS-CoV-2 RBD and ACE2
The variant B.1.618 is the triple mutant, which was identified in India, the mutations present in the S-protein are the deletion of Y145, H146, E484K, and D614G. 31 In the RBD region, there is only one residue has been mutated. The variant B.1.618 S-protein is neutralized with 2-5 fold cutback by recovering sera and vaccine-elicited antibodies. 32,33 Before the MD simulation, in minimized state, for B.1.618 variant, the intermolecular interactions are formed as obtained from the wild type with some less distance differences. The MD simulation reveals that the RBD forms Glu340···His239, Asn343···Val604, Asn334···Leu585, Asn343···Lys600, and Thr345···Val604 ( Figure 15) interactions with ACE2 at the distances 1.98, 1.79, 1.88, 1.86, and 2.26 Å, respectively. These interactions are not comparable with the wild type RBD-ACE2 complex and are different.

| Gamma variant (P.1): Interactions between the variant gamma variant SARS-CoV-2 RBD and ACE2
The variant P.1 is also known as gamma variant, which was identified in Brazil and Japan initially, the mutations presented as L18F, T20N, P26S, D138Y, R190S, K417T, E484K, N501Y, D614G, H655Y, and T1027I in the S-protein. 39 In the RBD region, E484K only has been mutated. For P.1 variant the intermolecular interactions are formed similar to the wild type at minimized state. The MD simulation shows that the interaction between RBD and ACE2 are Glu340···Lys596, Arg346···Asp609, Asn334···Asn586, Thr345···Val604, and Tyr449···Asp609 at the distances 1.96, 2.01, 2.13, 2.64, and 1.84 Å, respectively ( Figure 16). As these interactions totally found with new binding sites of the proteins, they are not comparable with the wild type RBD-ACE2 complex. These new interactions are formed due to the changes between binding sites of RBD and ACE2 during the MD simulation.

| PHYLOGENETIC TREE ANALYSIS
The phylogenetic tree analysis shows the characteristics of wild and variants proteins amino acid sequences ( Figure 18). The protein structures are converted into fasta format using Open Babel tool and the files are merged as single fasta file for further phylogenetic analysis. 47 These sequences are aligned with default parameters and subjected to the phylogenetic tree analysis using NGPhylogeny.fr server. 48 This relationship analysis reveals how all the variants of SARS-CoV-2 RBD with ACE2 are relatively close to each other. This tree shows that the connections between the sequences which are based on the mutations are commonly raised since the remaining residues are the same as presented in the wild type.

| ROOT MEANS SQUARE DEVIATION (RMSD) ANALYSIS OF WILD TYPE AND MUTATION VARIANTS
The RMSD of protein-protein complexes of variants illustrates their deviations during the MD simulation. Figure 19 shows the variation of RMSD values of wild type and all the mutation variants complexes during the MD simulations up to 500 ns. For wild type, the average RMSD range is 2.0-4.0 Å. Notably, during the 225-270 ns period of MD simulation, the RMSD peaks are found in the range of 3.2-6.6 Å.
The MD simulation of B.1.1.529 (Omicron) shows the variation of RMSD in the range of 1.8-4.2 Å, in which, up to 60 ns the deviations maintained in the range 1.8-3.0 Å, beyond that the RMSD slightly increased up to 4.7 Å and this trend was found up to 130 ns. Further, the whole remaining fickle peaks of RMSD range from 2.7 to 3.7 Å and this trend maintained up to 500 ns. Notably, the variation RMSD of omicron is similar to the wild type, and no significant deviation is found, this confirms that the omicron RBD-ACE2 complex is stable during the entire MD simulations.
In B.1.617 (Indian variant/double mutant), the RMSD variation is found in the range of 2.5-3.5 Å. During the simulation, some deviations are noticed, however they are not very significant. In B.1.617.2 (Indian/Delta variant) up to 200 ns, the RMSD value is slightly increased from 1.8 to 4.8 Å and further some fickle peaks were found until 500 ns simulation and are similar trend is also found notably in the loop regions residues have high fluctuations. On compare with the loop regions, the fluctuation of helix and sheet residues are relatively less. However, the average RMSF fluctuation is found in the range of the variants are within 0.8-3 Å. From these results, it is confirmed that no significant RMSF variations is found in all reported variants in sheets and helix region residues whereas loop region residues exhibit large fluctuations.

| HYDROGEN BONDING ANALYSIS
MD trajectories were analyzed to plot the number of hydrogen bonds using VMD molecular visualization program. 49 The complexes are stabilized through the hydrogen bonding interactions during the MD simulation. In the RBD-ACE2 complexes formation, the conformational changes with alterations in the structural properties like flexibility, compactness, surface area exposed to water molecules and the number of hydrogen bonds leads to the activation of inactivation process.
The number of hydrogen bonds formed during the MD simulation were plotted to confirm the stability of the RBD-ACE2 complexes. The average number of hydrogen bonds were shown Figure 21  F I G U R E 19 RMSD plots for wild type and selected variants of concern (Alpha to Omicron) of SARS-CoV-2 RBD-ACE2 complexes. ACE2, angiotensin-converting enzyme 2; RBD, receptor binding domain; RMSD, root means square deviation; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

| BINDING FREE ENERGY CALCULATIONS
The binding free energy values (ΔG) allow to understand the binding affinity of variants with ACE2. The binding energy of all the variants with ACE2 has been calculated and the results are compared with the wild type RBD-ACE2 complex (Table 1 and Figure 22). The ΔG values are calculated from the final frames of the 500 ns scale of MD simulations and the energy minimized complexes of all mutant variants and wild F I G U R E 20 RMSF plot for wild type and selected variants of concern (Alpha to Omicron) of SARS-CoV-2 RBD-ACE2 complexes. ACE2, angiotensin-converting enzyme 2; RBD, receptor binding domain; RMSF, roots means square fluctuations; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.  1 kcal/mol respectively. Relatively, the binding energy ΔG of wild type found to be similar to all variants and the same trend is also observed in all energy minimized RBD-ACE2 complexes of the variants. Hence, we can understand that from these ΔG values, some of the mutant variants sustain the affinity with less differences as presented in minimized state after the MD simulation ( Figure 22). As well as, some of the variants shows less affinity in MD simulation than presented in minimized state. However, the overall results confirm that SARS-CoV-2 RBD is binding with ACE2 with good binding affinity.

| CONCLUSION
This in silico study has been performed to understand the binding nature of some SARS-CoV-2 RBD variants with ACE2. Among the variants of concern, the present dynamics simulation found that both B. protein; in consequence of that the interactions between RBD and ACE2 are found different; however, some interactions are found common. The binding free energy analysis of B.1.617 (delta), B.1.617.2.1 (delta plus), B.1.618 (triple mutant) shows higher binding affinity compare with other variants similar to the wild type (<−10 kcal/mol). This may be the main reason for the inactivity of newly designed vaccines or repurposed drugs against the target, which was initially reported as a binding site of the spike protein. Further RMSD and RMSF analysis shows that in all the variants, the complexes are very stable during the MD simulation. The results of this in silico study suggest that during the MD simulations, some of the RBD variants bind with ACE2 at different locations; these new binding sites can be considered as new target sites as already reported as binding sites in the drug discovery process. Based on the results of this in silico study, the suggested new binding sites in ACE2 may be further confirmed by experimental studies to design new drug to effectively treat COVID-19.

AUTHORS CONTRIBUTIONS Jaganathan Ramakrishnan and Kumaradhas
Poomani designed the research work and did the MD simulation works. Jaganathan Ramakrishnan, Archana Chinnamadhu, and Suganya Suresh performed the analysis. All authors equally contributed to preparing and reviewing the manuscript.