Three structure forms of ACE2-spike protein complexes, namely 2ajf, 6m0j, as well as a chimeric structure 6vw1 have been selected for the study. The chimeric structure is a receptor-binding domain of SARS-CoV that acts as a scaffold, while the receptor-binding motif of SARS-CoV-2 acts as a functional group for the interaction with ACE212. Therefore, the comparison of three structures could be feasible to understand the role of spike mutations on the binding affinity to ACE2.
Analysis of spike (S) glycoprotein-ACE2 receptor interaction
At first, three crystal structures (2ajf, 6m0j and 6vw1) were obtained, and then the differences in their structures and sequences were identified (Figures S1 and S2). The results indicated that the majority of the mutations occurred in the receptor-binding motif of SARS-CoV-2 compared with SARS-CoV which is listed in the table S1. This motif plays an essential role in spike-receptor interaction. Then, we performed in silico comparison among these structures to deeply investigate the importance of mutations in the receptor-binding motif (RBM) and or other points of the receptor-binding domain (RBD) as well as the spike affinity to receptors to understand the binding of the virus to ACE2.
In the second step, hydrophobic, electrostatic, cation-pi interactions and hydrogen bonds formed in the receptor-spike complex were analyzed using Ligplot+. Based on the results, four regions in ACE2 (24-38, 41-42-45, 82-83 and 330-357) interact with the spike glycoprotein of SARS-CoV, including Arg426, Tyr436, Tyr440, Tyr442, Leu443, Leu472, Asn473, Tyr475, Asn479, Gly482, Tyr484, Thr486 Thr487, Gly488, Ile489. Also, a group of residues in the SARS-CoV-2 spike protein are present in the interface region of the four regions in ACE2 including Lys417, Gly446, Tyr449, Tyr453, Leu455, Phe456, Ala475, Phe486, Asn487, Tyr489, Gln493, Gly496, Gln498, Thr500, Asn501, Gly502, Tyr505. Ccomparative analysis of the RBD between SARS-CoV-2 and SARS-CoV showed the presence of several mutations in this area including Tyr442Leu, Leu443Phe, Leu472Phe, Asn479Gln, Tyr484Gln, and Thr487Asn. These mutations could be considered as important players in the binding sites of the spike protein in the ACE2 receptor, leading to an increase in binding affinity of SARS-CoV-2 to ACE2. For instance, mutation Tyr442Leu leads to change in interaction pattern between spike-ACE2 from SARS-CoV to SARS-CoV2. By means that, instead of residue 31, residue 34 of ACE2 is involved in virus-receptor interaction in the binding site which we will further discuss about this area. Accordingly, other mutations can also cause changes in the interacting pattern of SARS-CoV-2-ACE2 when compared with SARS-CoV. The detailed interaction patterns are shown in (Figure 1).
Analysis of the interacting region in the spike-ACE2 complex during MD simulation
PyContact was used for the analysis of non-covalent interactions between the spike and ACE2 receptor during the MD simulation to characterize the critical areas in ACE2 involved in the interaction with the spike. The contact areas between SARS-CoV and CoV-2-ACE2 with ACE2 receptor were depicted in figures S3 and S4. The obtained results showed five and six regions of ACE2 interacted with SARS-CoV and CoV-2-ACE2 respectively and the highest contact area in both complexes were detected in residues 35-194 and 266-344. These two regions are not constantly involved in the interaction, but instead residues 35-50 and 300-330 have the highest participation in the interaction in ACE2. Also, analysis of the interface area demonstrated that a group of critical residues including residues 353-358 from region 346-372 of the ACE2 receptor interacted with the SARS-CoV-2 spike protein, while such interaction was not found in the SARS-CoV-ACE2 complex.
On the other hand, the interface region of the chimeric structure is completely different from two other structures (Fig. S5). Based on our results, there are seven regions in ACE2 that interact with chimeric structure. The chimeric structure has an interface region including 11 residues (495-505), which were non-specific for protein-protein interaction in the virus-receptor complex. In residues 19-33, the interface region has higher contact area than SARS-CoV-2 but lower than SARS-CoV. In three regions of ACE2 (residues 16-175, residues 224-246, residues 248-326), the interface area was about 800 Å2, while in two other regions (347-373 and 542-615) was about 300 Å2. Such a difference may stem from the integration of two regions from two different proteins that cause nonspecific interactions.
Hydrogen bonds were analyzed in the four regions of the ACE2 receptor (19-33, 35-54, 325-331, and 334-339) and receptor-binding motif of the three spike structures during the simulation, and they are summarized in figure 2. Obtained results showed that the four regions of the ACE2 receptor participated in the interaction with the receptor-binding motif of the spike protein from SARS-CoV and SARS-CoV-2 (19-33, 35-54, 353-358, and 325-331).
The number of H-bonds in residues 19-33 of the receptor when interacting with the receptor-binding motif of the spike during the simulation was higher in complexes of chimeric structure and SARS-CoV-2 than that of SARS-CoV. Also, the number of H-bonds in residues 334-339 of the receptor in SARS-CoV-2 and chimeric structure complexes were more than that of SARS-CoV. However, the number of H-bonds in the spike-receptor complex in two regions 35-54 and 325-331 of SARS-CoV-2 and chimeric structure complexes were lower than SARS-CoV. In this way, the number of H-bonds in the four regions of the SARS-CoV-ACE2 complex when interacting with the receptor-binding motif of the spike protein was not the same as SARS-CoV-2 and chimeric structure complexes. Therefore, it appears that the number of H-bonds in the receptor-binding motif of SARS-CoV-2 when interacting with ACE2 was more than SARS-CoV. These four regions of the receptor interact with specific parts of the virus receptor-binding motif, which we will discuss later.
Interaction network analysis of SARS-CoV and SARS-CoV-2 spike protein complexes
The interactions pattern between the spike glycoprotein and ACE2 has been evaluated in SARS-CoV and SARS-CoV-2 using NAPS. The structures were obtained from the initial and final 500 frames of the simulation. Our results showed that the spike is attached to the receptor through two regions at the beginning and end of the receptor-binding motif during the simulation. Although other interactions also were occurred by the end of the simulation, it seems that these two areas play an important role in the attachment of the virus spike to its cognate receptor (Figure 3).
Analysis of binding free energies for three forms of the spike-receptor complex
The MM-PBSA method was utilized for the calculation of the binding energies of SARS-CoV, SARS-CoV-2, and the chimeric structure when bound to ACE2 (Table 1). The lowest binding energy was related to SARS-CoV-2 with -31.5759 ±2.4425 kcal.mol-1. According to the results, electrostatic interactions have an essential role in binding affinities between SARS-CoV-2 and its receptor.
The binding free energy for the spike-receptor was the same for chimeric and SARS-CoV complexes (-13.4810 ± 2.6388 kcal·mol−1 and -12.0104±3.0752 kcal·mol−1 respectively).
The result showed that the mutations in receptor-binding motif have an essential role in the increased affinity of SARS-CoV-2 to ACE2; however, the impact of mutations on the other regions of RBD is not great as much as mutations effects on RBM. According to the above results, interaction mechanism of SARS-CoV and SARS-CoV-2 spike with ACE2 receptor has been investigated in details and roles of mutations in changing the SARS-Co-2 affinity for ACE2 have been also assessed.
Binding free energy decomposition for spike-receptor complexes
The analysis of free energy decomposition was performed on the spike-ACE2 complexes. The results are depicted in figure 4 and tables S2 and S3. Free energy decomposition analysis helps to find contribution of a single residues by summing its interactions over the entire residues.
These results revealed that two mutation including Tyr442Leu and Leu443Phe in SARS-CoV-2 have changed binding free energies from -1.6539±0.4785 kcal·mol−1and -0.5149±0.0363 kcal·mol−1in SARS-CoV to -2.7769±0.17222 kcal·mol−1 and -1.6744±0.2814 kcal·mol−1 in SARS-CoV-2.
Mutations Pro462Ala and Leu472Phe in SARS-CoV-2 altered the binding free energy from -2 kcal·mol−1 in SARS-CoV to -6 kcal·mol−1. These two residues are located at the beginning and end of a loop that interacts with the N-terminal domain of the receptor. The flexibility of this loop might facilitate the binding of the spike protein to its receptor which is shown in figure S6. The Pro462Ala mutation makes the region flexible as a hinge, and therefore, facilitates the binding of the virus to its receptor.
Mutation Asn479Gln also altered the binding free energy of these from -2 kcal·mol−1 in SARS-CoV to -4 kcal·mol−1 in SARS-CoV-2. Therefore, it seems that this region also plays a critical role in the interaction of coronavirus to ACE2. This finding was also confirmed by the native contact result.
Also, binding free energy decompositions for ACE2 residues have been calculated, and the results are shown in figure 5. According to the results, free energy contributions of Lys31 and Lys353 are significantly different between SARS-CoV and SARS-CoV-2. The changes in free energy contribution of Lys31 from -1.9 kcal·mol−1 in SARS-CoV to -4.4 kcal·mol−1 in SARS-CoV-2 and Lys353 from -3.7 kcal·mol−1 in SARS-CoV to -5.5 kcal·mol−1 in SARS-CoV-2 could be related to the different interacting residues form spike protein and the mutations occurring in this region. The results demonstrated that Lys31 has a fundamental role in the interaction of ACE2 with N-terminal and middle regions of the receptor-binding motif. However, residue Lys353 contributes in the interaction with almost the end of the receptor-binding motif.
Contact patterns of the SARS-CoV and SARS-CoV-2 interactions with ACE2
For more details about the interaction between spike-ACE2 and also, to confirm the significance of the mutations that have occurred in the structure of SARS-CoV-2 compared with SARS-CoV, the native contact pattern was analyzed in two structures of SARS-CoV and SARS-CoV-2 during the simulation. The native contact pattern between the receptor-binding motif of the spike protein and ACE2 is shown in figure 6. The native contact contains Gln24, Thr27, Phe28, Asp30, Lys31, His34, Glu37, Asp38, Tyr41, Gln42, Leu45, Leu351, Gly352, Lys353, Gly354, Asp355, Phe356, Arg357, Asn330 from the ACE2 receptor, interacting with Thr436, Tyr440, Tyr442, Leu443, Pro462, Asn373, Thr475, Asn479, Gly482, Tyr484, Thr486 of SARS-CoV and Tyr449, Tyr453, Leu455, Phe456, Tyr473, Ala475, Asn487, Tyr489, Gln493, Gln498, Thr500 of SARS-CoV-2.
The maximum number of native contacts in the interaction between the receptor-binding motif of the spike protein and ACE2 receptor was observed in residues 31, 353 of ACE2 for both structures (SARS-CoV and SARS-CoV-2). These two residues are considered as hotspot points that interact with the beginning and terminal regions of the receptor-binding motif in the spike protein.
In addition to regions at the beginning and terminal of the receptor-binding motif, another region in the middle of the receptor-binding motif was also involved in the interaction between the spike protein and ACE2 receptor. This region acts as a clamp in the binding of the virus to ACE2 (Fig. 6A3, 6B3). Indeed, in this region is including Tyr440, Tyr442, Leu443, and Asn479 of SARS-CoV as well as Tyr453, Phe455, Leu456 and Gln493 of SARS-CoV-2, create a cavity that, through its two edges interacts with Thr27, Asp30, Lys31, His34 and Glu35 of ACE2.
There are three mutations in this region, including Tyr442Lue, Leu443Phe, and Asn479Gln.
These mutations cause an increase in the binding affinity of the ACE2 receptor to SARS-CoV-2 compared with SARS-CoV. For more confidence, selected alanine scanning was done for residues (Tyr440, Tyr442, Leu443, and Asn479 for SARS-CoV and Tyr453, Leu455, Phe456, Gln493 for SARS-CoV-2) and MM-PBSA method was used in order to calculate the binding affinity of each substitution (Table 2 and 3). According to the result, altering each of these residues lead to reduce the binding affinity in SARS-CoV-2 and the lowest binding affinity was observed for Phe456Ala substitution.
In contrast to the selected SARS-CoV-2 alanine scanning results, except for Leu443Ala, the rest of substitution for SARS-CoV have increased the binding affinity to ACE2.