Mutations in the receptor-binding domain of human SARS CoV-2 spike protein increases its affinity to bind human ACE-2 receptor

Abstract The severe acute respiratory syndrome virus-2 (SARS CoV-2) infection has resulted in the current global pandemic. The binding of SARS CoV-2 spike protein receptor-binding domain (RBD) to the human angiotensin converting enzyme-2 (ACE-2) receptor causes the host infection. The spike protein has undergone several mutations with reference to the initial strain isolated during December 2019 from Wuhan, China. A number of these mutant strains have been reported as variants of concern and as variants being monitored. Some of these mutants are known to be responsible for increased transmissibility of the virus. The reason for the increased transmissibility caused by the point mutations can be understood by studying the structural implications and inter-molecular interactions in the binding of viral spike protein RBD and human ACE-2. Here, we use the crystal structure of the RBD in complex with ACE-2 available in the public domain and analyse the 250 ns molecular dynamics (MD) simulations of wild-type and mutants; K417N, K417T, N440K, N501Y, L452R, T478K, E484K and S494P. The ionic, hydrophobic and hydrogen bond interactions, amino acid residue flexibility, binding energies and structural variations are characterized. The MD simulations provide clues to the molecular mechanisms of ACE-2 receptor binding in wild-type and mutant complexes. The mutant spike proteins RBD were associated with greater binding affinity with ACE-2 receptor. Communicated by Ramaswamy H. Sarma


Introduction
The severe acute respiratory syndrome coronavirus-2 (SARS CoV-2) has caused a global pandemic of the coronavirus disease 2019  during the last 20 months.  has been the single major cause of death due to any disease within a short span of time. SARS CoV-2 was first reported in individuals known to have been in contact with wildlife animals at the live animal and seafood market in Jianghan District, Wuhan (Zhu et al., 2020). SARS CoV-2 is similar to SARS CoV (2003to 2005, Middle East respiratory syndrome coronavirus (MERS CoV) (2012 to 2013) and other human CoVs in the 20 th century that has led to epidemics resulting in severe respiratory diseases and deaths (Guruprasad, 2021a). These viruses harbour $ 30 K bp single stranded positive-sense RNA genome. SARS CoVs enter human cells through fusion of viral and host cellular membranes mediated by the interaction between viral spike protein and human angiotensin converting enzyme-2 (ACE-2) (Guruprasad, 2020;2021b;Li et al., 2003;Shang et al., 2020).
The SARS CoV-2 spike protein is a heavily glycosylated homo-trimeric protein with $1,273 amino acids and the sequence region (amino acids 333-520) constitutes the receptor-binding domain (RBD) that interacts with human ACE-2 receptor. The three-dimensional structures of the spike protein apo and RBD bound forms to human ACE-2 receptor are available in the public domain Xiao et al., 2021;Xu et al., 2021). Viruses acquire mutations over a period of time during the host infection giving rise to new sequence variants. RNA viruses have much higher mutation rate compared to DNA viruses. The viruses that undergo favourable mutations continue to persist in the host. Due to mutations, the viruses might gain ability to evade detection by specific viral diagnostic tests, or decreased susceptibility to therapeutic agents, such as, monoclonal antibodies and small molecule drugs. Some mutations can produce viruses with new antigenic determinants and the antigenically altered viruses may be able to cause disease in previously resistant or immune hosts or cause vaccine rejection (Fleischmann, 1996).
The wild-type refers to first reported strain of the human SARS CoV-2 virus isolated from patient in Wuhan, China (NCBI_id: NC_045512) . With respect to wildtype proteins, the mutations and sequence variants are collected through complete genome sequence and epidemiological studies of SARS CoV-2 strains across populations from various geographical locations and different times. These sequences have been deposited in the NCBI (www.ncbi.nlm. nih.gov/) and GISAID (https://www.gisaid.org/) databases. Sequence analyses have reported deletions, insertions and substitution mutations in all SARS CoV-2 proteins including the spike protein (Guruprasad, 2021b;Mohammadi et al., 2021) demonstrating that SARS CoV-2 has an innate ability to undergo mutations rapidly. SARS CoV-2 vaccines show protective efficacy towards humans by providing neutralizing antibodies which recognize the viral spike protein (Kyriakidis et al., 2021). The effects of spike protein mutations in SARS CoV-2 on the neutralization of antibodies have been studied (Rees-Spear et al., 2021).
Genome sequencing and protein sequence analyses have shown the emergence and persistence of some SARS CoV-2 spike protein mutations in subsequent generations during human infection. One of the early identified mutations in spike protein, D614G is associated with higher viral loads but not with increased disease severity (Korber et al., 2020). The D614G mutant spike protein increases SARS CoV-2 infection of multiple human cell types compared to the wild-type strain (Daniloski et al., 2021) and efficiency of viral entry with enhanced ACE-2 binding affinity (Ozono et al., 2021) by assembling more functional spike protein into the virion . Epidemiological evidence suggests that the D614G variant has increased ability to spread more quickly than viruses without this mutation. Therefore, the D614G mutant has become dominant in the SARS CoV-2 spike protein.
The Phylogenetic Assignment of Named Global Outbreak Lineages (PANGOLIN) software tool (Rambaut et al., 2020) implements a dynamic and rational nomenclature of the SARS CoV-2 strains. The PANGO lineages available at website https://www.cov-lineages.org refers to the cluster of sequences associated with distinct geographical locations with evidence of onward spread and captures the emerging trends of mutations from genomic epidemiological surveillance and outbreak investigations. The US government SARS CoV-2 Interagency Group (SIG), initially developed a variant classification scheme that defines three classes of SARS CoV-2 variants; variant of interest, variant of concern and variant of high consequence. A variant of interest has specific genetic markers associated with changes in receptor binding, increased disease severity, reduced efficacy of treatments, reduced neutralization by antibodies generated against previous infection or vaccination, potential diagnostic impact, increased proportion of clusters of cases and therefore increased transmissibility. A variant of concern must display evidence of increased transmissibility, more severe disease with evidence of reduced effectiveness of treatments or vaccines, or failure in diagnosis detection, treatments, or vaccines leading to increased hospitalizations or deaths.
Recently, the SIG has re-classified SARS CoV-2 variants into four classes; variant being monitored, variant of interest, variant of concern and variant of high consequence (https:// www.cdc.gov/coronavirus/2019-ncov/cases-updates/variantsurveillance/variant-info.html). The lineages for variant being monitored and variant of concern along with their World Health Organization (WHO) label and mutations in the spike protein are listed in Table 1. Currently, there are no mutations classified under variant of interest or variant of high consequence.
The B.1.1.7 lineage first detected in the UK during September 2020 was subsequently reported in several countries including India. The substitution mutations in the RBD of spike protein in this lineage were E484K, S494P, and N501Y. This lineage is attributed to 50% increased transmissibility (Davies et al., 2021), increased severity based on hospitalizations and fatality rates compared to other variants (Horby et al., 2021). The B.1.351 lineage was first identified in South Africa during early October 2020, Zambia during late December 2020 and subsequently reported in several countries. The substitution mutations in the RBD of spike protein in this lineage were; K417N, E484K, N501Y. This lineage is also attributed to 50% increased transmissibility (Patone et al., 2021). The E484K mutation may affect neutralization by some polyclonal and monoclonal antibodies . The P.1 variant was first reported in Japan in travellers from Brazil, subsequently, Manaus, in the Amazon region and also in the United States at the end of January 2021. The P.1 lineage mutations in the RBD of spike protein were K417T, E484K, N501Y. The P.1 variant mutations may affect its transmissibility and antigenic profile resulting in its decreased ability to recognise antibodies generated through a previous viral natural infection or through vaccination (Harvey et al., 2021 and variant of concern comprises mutations; K417N, L452R, T478K. The P.2 lineage first reported in Brazil during October 2020 has four mutations in the entire spike protein with the E484K mutation in RBD. The B.1.621 lineage that originated in Colombia during January 2021 has the mutations; E484K and N501Y in RBD. The mutations corresponding to the entire spike protein across the different lineages are shown in Table 1. The three-dimensional crystal structure of the spike protein RBD (residues 333-520) complexed with human ACE-2 is available in the protein structure data bank (PDB) (Lan et al., 2020;Wang et al., 2020). The structure comprises a fivestranded antiparallel b-sheet. According to the PDBSum (Laskowski et al., 2018), the amino acids region 440 À 506 (67 amino acids) located between b4 and b5 strands folds into an extended loop that comprises short stretches of two a-helices The substitution mutations in RBD were shown to increase the transmissibility of COVID-19 and decreased protection from vaccines (Bian et al., 2021;Chen et al., 2020;Gomez et al., 2021;Harvey et al., 2021;Noh et al., 2021;Zhou et al., 2021). This prompted us to analyse the mutations and solvated interaction energy (SIE) corresponding to variants of concern and variants being monitored using MD simulations of the spike protein RBD domain and human ACE-2 receptor complex. These studies on host-virus protein-protein interactions at the atomic level suggest molecular mechanisms of their binding.

Generation of mutants in human SARS CoV-2 spike protein RBD
The crystal structure of the human SARS CoV-2 spike protein RBD complexed with human ACE-2 receptor (PDB id: 6LZG)  was used to generate single amino acid substitution mutations; K417N, K417T, N440K N501Y, L452R, T478K, E484K, S494P, K417A, Q498A, T500I in the spike protein RBD using Discovery Studio 3.5. The N-acetylglucosamines, located at the sites of glycosylation were deleted from the experimental structure. In all, 11 single amino acid mutant human SARS CoV-2 spike protein RBD and human ACE-2 complexes were generated and studied by MD simulations. The wild-type RBD-ACE-2 complex was also studied as the reference molecular system.

Molecular dynamics simulations
The MD simulations of the wild-type and single amino acid substitution mutant RBD -ACE-2 receptor heterodimer complexes were carried out using AMBER (ver.18.14) (Gotz et al., 2012;Salomon-Ferrer et al., 2013). The GAFF2 force fields charges and Amber ff14sb protein atomic positions of all the systems were generated with Antechamber using am1bcc method (Lindorff-Larsen et al., 2010). All input parameter files for MD simulations were generated for the heterodimer by adding hydrogen atoms with Hþþ server (Anandakrishnan et al., 2012;Gordon et al., 2005). Sodium ions were added to the system in order to neutralize excessive charge generated by solvating the complex in a 10 Å cubic box. The final ionic concentration of the systems was set to 100 mM. The Amber ff14sb-idln force fields were used with TIP3P water model (Jorgensen et al., 1983;Mark & Nilsson, 2001). The topology input.crd parameter files were generated with tLEaP module using the Amber suite (Hornak et al., 2006). The MD simulations were run at 300 K temperature and 1 atmospheric pressure. A 2 fs time step was considered for friction coefficient of 1/ps of Langevin integrator. Energy minimization was carried out using steepest descent method for 40,000 cycles in order to overcome short range null contacts in the system. The longrange electrostatic interactions were handled using the particle-mesh Ewald (PME) method with a 9 Å real-space cut-off and with PME order 4. The systems were double minimized under NPT ensemble at interval of 5 frames to maintain pressure of 1 atmosphere with Monte Carlo barostat. The minimization was carried out under NVT ensemble at interval of 5 frames, in order to maintain volume and temperature at 300 K using Monte Carlo thermostat (Darden et al., 1993;Wang et al., 2006). The systems were equilibrated for 7 ns before the production run. The coordinates from production runs were saved for every 10 ps. All molecular systems, one wild-type and 11 mutant complexes were executed for 250 ns MD simulations using AMBER.

Post molecular dynamics data analyses
The MD simulations trajectory data analysis was carried out using cpptraj and pytraj in Amber tools. The average structures, root mean square deviation (RMSD) and root mean square fluctuation (RMSF) for all systems were derived from .trr analysis. The cpptraj hbond sub-level trajectory analysis provides the average hydrogen bonding distance between human ACE-2 (chain-A) and spike protein RBD (chain-B). The SIE is an indirect method to calculate the binding free energies between protein-protein or protein-ligand complexes simulated with explicit solvent models (Sulea & Purisima, 2012). In order to understand the protein-protein affinities for wild-type and mutant protein complexes, the SIE-traj analysis (Naïm et al., 2007) was performed between ACE-2 receptor and RBD on the entire Figure 1. Amino acid residues in 4.5 Å vicinity between human SARS CoV-2 spike protein RBD (blue) and human ACE-2 receptor (magenta). The hydrogen bonding interactions involved residues; Ala475-Ser19, Asn487-Gln24, Thr500-Tyr41, Lys417-Asp30, Tyr449-Asp38, Tyr449-Gln42, Asn487-Tyr83, Gln498-Gln42, Asn501-Tyr41, Gly502-Lys353. 25,000 frames generated from 250 ns MD simulations. It approximates the protein-protein binding free energy in aqueous solution, DG bind , by an interaction energy contribution and a desolvation free energy contribution (Cui et al., 2008). The SIEs and non-bonded interaction energies were computed for the wild-type and mutant dimeric RBD -ACE-2 complexes.

Molecular dynamics simulations
The MD simulations showed enhanced binding of human SARS CoV-2 spike protein RBD to human ACE-2 receptor, suggesting the possible role of mutations in leading the virus to become variants of concern or variants being monitored. The structural superimposition of initial and final MD simulations structures showed relative displacement of chain-B (RBD) compared to chain-A (ACE-2), except for the protein with E484K mutation as shown in Figure 2. The RMSD plots for representative structures shown in Figure 3 and for all the structures (Supplementary material, Figures 1 and 7) demonstrated that all structures were stabilised within 4 Å RMSD during

Impact of mutations on the secondary structure and protein folding
The helices (H1, H2, H17 and H19) and b-hairpin 'B' in the ACE-2 interact with the spike protein RBD. In the wild-type complex, both H1 and H2 helices retain their secondary structure during the 250 ns MD simulations and is similar to the initial crystal structure as can be seen from Figure 5. However, in all the mutant complexes a broken helix H1 is observed. The short helix H17 has lost the secondary structure in the wild-type and N510Y mutant complexes during the MD simulations. The b-hairpin 'B' is retained in all the complexes studied. In the spike protein RBD, the RBM between the amino acids 440 and 506 comprise four strands that form two sets of anti-parallel b-sheets and two short helices. These helical structures are lost in the RBD due to the single amino acid substitution mutations ( Figure 6). In addition to the secondary structural changes in the regions involving the RBD and VBM in spike protein and human ACE-2 respectively, minor secondary structural alterations were observed in both the chains of the heterodimeric complex. Despite the changes in the secondary structural regions, we observed the retention of the three-dimensional fold in all the mutant complexes as can be seen from

Ionic, hydrophobic, hydrogen bond interactions in wild-type and mutant complexes
Several inter-molecular interactions were observed to be mediated between the SARS CoV-2 spike protein RBD and human ACE-2 receptor in the wild-type and mutant protein complexes during the 250 ns MD simulations. Accordingly, these were classified as the hydrogen bond interactions, ionic interactions and hydrophobic interactions. A list of these non-bonding interactions is shown in Table 2. The K417 residue is mutated to N417 in B.1.351 lineage variant of concern. This mutation is reported to contribute to the loss of serum antibody neutralization (Collier et al., 2021). The K417 residue is located on a3-helix between b3 and b4 strands and is close to the insertion loop in RBD. The sidechain Nf-atom of K417 makes ionic interactions with sidechain atom of Asp30 located on a1-helix in human ACE-2. The Lys417-Asp30 side chain ionic interaction observed in the wild-type and all the other mutants is lost owing to the K417N and K417T mutation. The N501Y mutation associated with lineages; B.1.1.7, B.1.351, P.1 binds human ACE-2 receptor with increased binding affinity (Luan et al., 2021). The N501 is located in the loop connecting the a-helix preceding b5-strand and its side-chain amide nitrogen makes intermolecular hydrogen bond interactions with side-chain OH Hydrogen bonds: Asn487-Gln24, Gln493-Lys31, Gln493-His34, Gln493-Glu35, Tyr505-Glu37, Asn501-Tyr41, Thr500-Tyr41, Gln498-Gln42, Asn487-Tyr83, Tyr495-Lys353, Gly496-Lys353, Gly502-Lys353, Thr500-Asp355. Hydrophobic: Tyr489-Phe28, Phe486-Met82 and Phe486-Tyr83.
atom of Tyr41 in ACE-2. The SARS CoV-2 spike protein threedimensional structure of the UK variant (B.1.1.7) in complex with ACE-2 ectodomain showed p-p interaction between Tyr501 mutant and Tyr41 in ACE-2 that enhances the binding of spike protein to the receptor and abolishes binding of a potent neutralizing antibody . The N501Y mutant spike protein RBD forms hydrophobic interactions with Tyr41 and Lys353 in ACE-2 thus increasing the overall binding affinity of the RBD -ACE-2 complex (Luan et al., 2021). Tian et al., have shown that the N501Y mutation exhibited a stronger interaction, with a faster association rate and a slower dissociation rate using surface plasmon resonance, also the atomic force microscopy studies showed a higher binding probability and binding strength for interactions with human ACE-2. The computational studies using MD simulations revealed additional p-p and p-cation interactions in the RBD-ACE-2 complexes (Tian et al., 2021). We observed that the N501Y mutant leads to hydrophobic interactions in the mutant complex involving Tyr501-Tyr41 (Tyr41 is indicated in Figure 7) at the protein-protein interface, in addition to the Phe486-Met82 hydrophobic interaction. The L452 is associated with RBM's anti-parallel b-sheet, it is exposed to solvent and is not directly involved in interaction with ACE-2. The Leu452 together with Phe490 and Leu492 forms a hydrophobic surface on RBD. The E484Q, L452R and double mutant E484Q and L452R were studied using MD simulations (Antony & Vijayan, 2021). It was observed that Arg452 interacts more with neighbouring residues Ser349, Tyr351, Phe490, Leu492 and Ser494 when compared to the wild-type and propose that the increased intra-molecular interactions could lead to the increased stability of the SARS CoV-2 spike protein. The mutation; L452R is observed in lineages; B.1.427 and B.1.429. The T478 is associated with a loop in RBM. The hydrophobic cluster formed by Phe486 in RBD and its interactions with residues; Leu79, Met82, Tyr83 in ACE-2 is observed in the complexes with the mutants; T478K, E484K, N440K. The E484 is within a loop in RBM and does not interact with human ACE-2 in wild-type structure. The E484K mutant results in the loss of serum antibody neutralization . Wang et al., 2021 have shown that the E484K mutation resulted in more favorable electrostatic interactions and significantly improved binding affinity with ACE-2. Further, the E484K mutation is shown to cause conformational rearrangements of the loop region containing the mutant residue that leads to a tighter binding with ACE-2 and formation of some new hydrogen bonds (Wang et al.,

2021
). Similar to E484, the S494 residue is also exposed to the solvent and is not involved in interactions with ACE-2. The S494 residue is associated with an antiparallel b-sheet in RBM. The L452 and S494 are located on the individual strands of antiparallel b-sheet and the N440 is located on a helical turn at the end of b4-strand distant from the ACE-2 binding site. Two single amino acid substitution mutations (E484K, N501Y) and a triple mutant (K417N þ E484K þ N501Y) in the RBD domain in complex with human ACE-2 was studied using protein-protein docking and MD simulations (Istifli et al., 2021). The South African (K417N-E484K-N501Y) and Brazilian (K417T-E484K-N501Y) triple mutants have been shown to be lethal due to the inter-protein contacts specifically mediated via the electrostatic interactions from the results of molecular docking and MD simulations studies (Istifli et al., 2021;Khan et al., 2021). Single amino acid point mutations in the RBD and C-terminus of spike protein were studied using MD simulations (Ahamad et al., 2022;;Istifli et al., 2021). These authors have shown that mutation brings about higher fluctuations mainly in the spike protein RBD region around 400-544 and heptad repeat 1 around 930-940 (Ahamad et al., 2022). Mutations in the SARS CoV-2 spike protein RBD are responsible for strong ACE-2 binding and poor anti-SARS CoV-2 monoclonal antibodies cross-neutralization (Shah et al., 2020). The alanine scanning mutagenesis and computational binding affinity studies of certain residues in SARS CoV-2 RBD complex with human ACE-2 showed that the mutations in conserved receptor binding motif (RBM) affects the structural dynamics of the complex. The charge distribution disturbs the inter-molecular non-bonded contacts thereby perturbing the strength of binding to host cell ACE-2 receptor (Dehury et al., 2021). A pictorial representation of the non-bonding interactions between the human spike protein RBD and human ACE-2 described above and in Table 2 are shown in Figure 8. In summary, Phe486 and Tyr489 in the human SARS CoV-2 spike protein were observed to be involved in hydrophobic interactions with residues; Ile21, Phe28, Tyr41, Leu79, Met82 and Tyr83 present on helices (H1, H2) of human ACE-2 as shown in Figure 7. A list of the more common inter-molecular hydrogen bonding interactions for the wild-type and mutant SARS CoV-2 spike protein RBD -ACE-2 complex is provided in Table 1 (supplementary material). The predominant hydrogen bonding interactions were observed between; Asn487-Tyr83, Asn487-Gln24, Gln498-Gln24, Gln498-Gln42, Gln498-Tyr41, Gln498-Lys353, Gly496-Lys353, Gln493-Lys31, Ala475-Ser19 and Ala475-Gln24. The interactions involving the same amino acid residue with different partner residues in the complex during the MD simulations, suggests the promiscuous nature of hydrogen bonds in the complex of human SARS CoV-2 spike protein RBD and human ACE-2 receptor reflecting the structural plasticity associated with the RBM. Several hydrogen bonds, hydrophobic and ionic interactions mediate the intermolecular interactions between the SARS CoV-2 spike protein RBD and human ACE-2 in wild-type and mutant complexes. Further, despite the mutations observed, the spike protein RBD is capable of interacting with the ACE-2 receptor leading to the host infection. Each of the single amino acid substitution mutations independently have significant effect on the nature of interactions with human ACE-2. It has already been reported that single and double mutants in the RBD do not disrupt the interactions with ACE-2, but reduce the binding free energies because of the multiple interactions in the inter-molecular interactions and the extended molecular surface (Taka et al., 2021).

Cluster analysis
The flexible partners stabilising inter-molecular interactions between the heterodimeric complex was further analysed using cluster analyses shown in Figure 9. The population of clusters and their standard deviations is provided in Table 2 (supplementary material). This table indicates 6 clusters with greater than 0.05 fraction (1,250 populations in a given cluster out of the 25,000 frames) in wild-type protein, whereas, the number of clusters associated with the different mutations were; K417N (8 clusters), K417T (9), N440K (10), L452R (9), T478K (8), E484K (9), S494P (10), N501Y (7). The members of each cluster comprise structurally similar conformations. The presence of only 6 clusters in the wild-type protein with a higher population of frames suggests the structural stability of wild-type complex, which is in contrast to the relatively higher dynamics observed for the mutant complexes associated with larger number of clusters.

Solvated interaction energies
Based on the studies from nine types of amino acid mutations in the spike protein RBD in complex with human ACE-2 using MD simulations, it was reported that some of the RBD mutants showed enhanced binding affinities with ACE-2 and the mutants N354D/D364Y, V367F and W436R displayed lower relative binding free energies compared to the wild-type . In order to quantify the strength of intermolecular interactions, the binding free energies of the heterodimers were analysed. The SIE binding free energy calculations were carried out on the AMBER MD simulations trajectories shown in Table 3. Among all complexes studied, lower SIE values were observed for S494P (-31.24 kcal/mol), T478K (-29.67 kcal/mol), K417N (-29.59 kcal/mol), L452R mutant (-27.94 kcal/mol), N440K (-20.18 kcal/mol), E484K (-19.15 kcal/ mol), N501Y (-18.98 kcal/mol) and K417T (-16.58 kcal/mol) compared to the binding free energy for the wild-type heterodimeric complex (-13.75 kcal/mol). It was observed that all mutations in the spike protein RBD were associated with lower binding-free energies compared to the wild-type proteins indicating better binding efficiency to human ACE-2. All mutations attributed as variants of concern or variants being monitored in RBD are known to increase transmissibility. These mutations cause greater infectivity to the host and may be under positive selection pressure. However, for some mutations, such as, K417A, Q498A, T500I in the spike protein RBD previously reported (Guruprasad, 2021b), we observed relatively higher binding free energies compared to the wild-type hetero-dimeric complex (Table 3). These mutant proteins with implied reduced binding affinity to human ACE-2 may therefore not be significant mutations. Thereby, mutations resulting in greater infectivity to host seem to have been selected in the evolution of human SARS CoV-2 spike protein.
Such mutations have therefore become prominent and have resulted as variants of concern or variants being monitored.

Conclusions
The MD simulations studies of the human SARS CoV-2 spike protein RBD and ACE-2 receptor complex for wild-type and mutants; K417N, K417T, N440K, L452R, T478K, E484K, S494P and N501Y reveal the molecular interactions underlying their binding affinities. The promiscuous nature of the non-bonding interactions is facilitated by the structural plasticity of the RBD that is accompanied by large conformational changes during the MD simulations. The mutant proteins are characterized by larger number of clusters indicating greater conformational variability among different clusters. This suggests that the mutant proteins undergo relatively greater conformational changes compared to the wild-type proteins. . Cluster analyses for wild-type and mutant SARS CoV-2 spike protein RBD and human ACE-2 receptor for 250 ns MD simulations data. Table 3. Solvent interaction energies (SIE) in kcal/mol calculated from sietraj for the 250 ns MD simulations trajectories for wild-type and mutant human SARS CoV-2 spike protein RBD -ACE-2 complexes. The SIE binding free energy (DG) and its components; van der Waals interaction energy (vdW); Coulomb interaction energy (Coul); Reaction Field (RF); Constant (Const). The SIE analyses of human SARS CoV-2 spike protein RBD and ACE-2 complex suggests the basis for positive selection of mutants that have led to more infectious variants resulting in rapid spread of the COVID-19 disease.