key: cord-0844449-4jh8b6wj
authors: Cao, Yufei; Ge, Jun
title: Study of Specific Receptor Binding Mode Suggests a Possible Enzymatic Disinfectant for SARS-CoV-2
date: 2021-01-29
journal: Langmuir
DOI: 10.1021/acs.langmuir.0c02911
sha: ba9a1ec13612a2dca099361e53f16d72d1a183fc
doc_id: 844449
cord_uid: 4jh8b6wj

[Image: see text] The outbreak of the coronavirus disease 2019 (COVID-19) caused by SARS-CoV-2 has spread globally. SARS-CoV-2 enters human cells by utilizing the receptor-binding domain (RBD) of an envelope homotrimeric spike (S) glycoprotein to interact with the cellular receptor angiotensin-converting enzyme 2 (ACE2). We thoroughly studied the differences between the two RBDs of SARS-CoV and SARS-CoV-2 when they bind with ACE2 through molecular dynamics simulations. The peculiarities of the SARS-CoV-2 RBD are obvious in several aspects such as fluctuation of the binding interface, distribution of binding free energy on residues of the receptor-binding motifs, and the dissociation process. Based on these peculiarities of SARS-CoV-2 revealed by simulations, we proposed a strategy of destroying the RBD of SARS-CoV-2 by employing enzymatic digestion. This unique strategy is promising for developing a skin-friendly, nontoxic, and convenient disinfectant to protect people from infection by SARS-CoV-2.

The emergence of the highly pathogenic coronavirus SARS-CoV-2 poses a serious global health emergency, causing a devastating societal and economic burden. 1, 2 The virus has spread worldwide. 3 Similar to individuals infected by SARS-CoV in 2003, patients with COVID-19 showed symptoms of fever, severe respiratory illness, and pneumonia. 4,5 SARS-CoV-2 entry into human cells is mediated by an envelope homotrimeric spike (S) glycoprotein protruding from the viral surface. 2, 6, 7 There are two subunits in S glycoprotein: one is responsible for binding to the host cell receptor angiotensinconverting enzyme 2 (ACE2) and the other promotes fusion of the viral and cellular membranes. 8, 9 The S glycoprotein of SARS-CoV-2 shares 80% amino acid sequence identity with SARS-CoV. 10 A recent high-resolution structure of the SARS-CoV-2 receptor-binding domain (RBD) bound to ACE2 has shown that the ACE2-binding mechanism is apparently identical between SARS-CoV-2 and SARS-CoV RBDs. 7, 10, 11 Although SARS-CoV-2 shares many similarities with SARS-CoV, it is more likely to spread from human to human. 12 The origin of its higher infectivity remains puzzled. McLellan et al. previously observed that ACE2 can bind to the SARS-CoV-2 S ectodomain with ∼15 nM affinity which is approximately 10to 20-fold higher affinity than ACE2 binding to SARS-CoV S. 7 This high affinity of SARS-CoV-2 S for ACE2 was considered to contribute to the apparent ease of spread between humans. To the contrary, some experiments obtained similar binding affinities for ACE2 binding to RBDs of two viruses. 10, 13 Different simulations also showed distinct results, 14−16 probably because of different free energy calculation methods, different simulation settings (temperature and force field), and different complex structures employed. Beyond this, experiments showed that the SARS-CoV-2 RBD does not have antibody cross-reactivity with RBDs of SARS-CoV and MERS-CoV. 13,17−19 This indicated that antiRBD antibodies are viral species-specific inhibitors and inherent differences between coronavirus exist. Therefore, thoroughly understanding the differences between two RBDs of SARS-CoV and SARS-CoV-2 can provide a structural basis for the rational design of vaccine and disinfectant. Previous studies mainly focused on simulation of the binding structure of the complex, 20 comparison of the binding affinity with ACE2 between two RBDs, 21 and analysis of hot spots of the binding interface 22,23 among others. Our simulations aimed to elucidate the different interface dynamics and binding interaction distributions of two RBDs and its influence on the dissociation process of complexes. Some information of the mechanism about the recognition of two proteins was also shown.

As the available treatment or vaccine has not obtained a complete success, this virus might be with us for a long time and the ability to protect people from infection by SARS-CoV-2 is of great concern. Many hard surface disinfectants have been proved to be effective against SARS-CoV-2 but are not suitable for use on skin or personal protective equipment (PPE). 24 For example, PPE should be effectively disinfected to extend the use and reuse due to shortages of N95 masks and other protective PPE. A previous study suggested that ethanol and chlorine-based disinfectants are not suitable for decontamination because of their capacity to adsorb on the fabric material and interfere with their electrostatic properties, thereby reducing the filtration capacity. 25 Some recommended PPE decontamination methods (UV, dry heat, and hydrogen peroxide vapors) cannot, however, be easily applied for daily use of PPE, warranting the need for new or additional disinfection methods that can be applied in times of need in these settings without causing harm to adjacent skin. In addition to this, the use of nasal sprays to minimize person-toperson transmission of the virus is also needed, because of the dispersal of the SARS-CoV-2 virus from asymptomatic individuals carrying high viral loads in their nasal epithelium. Therefore, a skin-friendly, nontoxic, and convenient disinfectant used for decontamination of PPE or nasal sprays is urgently needed.

Based on thorough study of the binding mode between SARS-CoV-2 and ACE2, we proposed a strategy that can principally destroy the SARS-CoV-2 RBD by employing endopeptidase. This enzymatic strategy has a great potential for the disinfection of skin and PPE and for the development of nasal sprays.

■ RESULTS AND DISCUSSION Different Interface Dynamics and Binding Interaction Distributions. The RBDs of SARS-CoV and SARS-CoV-2 can be structurally divided into two parts: a core region and a receptor-binding motif (RBM). The core region which consists of five antiparallel strands of β-sheet shares 87% sequence identity between the two viruses and is highly conserved. In contrast, there are more mutations in the RBM, only sharing 50% homology. As shown in Figure 1a ,b, there are three binding contact regions in the RBM named R1, R2, and R3. These three regions bind with different sites of ACE2 and anchor the RBD firmly with ACE2. The homology between the two RBMs in R1, R2, and R3 is 30, 70, and 55%, respectively. The sequence variations mainly exist in R1 and R3, two loop regions.

Mutations in the RBM might bring obvious differences between the two RBDs of SARS-CoV and SARS-CoV-2 when they bind with ACE2. To understand the two modes of binding at the atomic level, we analyzed 10 μs-long trajectories of the complexes of SARS-CoV S and SARS-CoV-2 S bound to ACE2 provided by Shaw group. 26 The overall root mean square deviation (RMSD) of both the complexes approximately reached an equilibrium during the last 2-μs-long simulations ( Figure S1 ). However, the RMSD of the RBM of the two RBDs ( Figure 1d ) reveals different interface dynamics. The RBM of the SARS-CoV-2 RBD is more active than that of Langmuir pubs.acs.org/Langmuir Article the SARS-CoV RBD, which demonstrates that the SARS-CoV-2 RBM in the complex is more flexible and can change conformation easily with lower energy barrier. The per-residue root mean square fluctuations (RMSF) of ACE2 in the two complexes did not show obvious difference (Figure 1e ), while more fluctuation can be observed in RMSF of the SARS-CoV-2 RBD than that of the SARS-CoV RBD ( Figure 1f ). Residues from Ser443 to Asn450 in the SARS-CoV-2 RBD are especially more fluctuant than the corresponding residues in the SARS-CoV RBD, which contributes to the higher flexibility of the SARS-CoV-2 RBD (Figure 1c , regions masked by black circles). Most of these residues are located in the R3 of the RBM, and therefore, we can also observe that the RMSD of the R3 in the SARS-CoV-2 RBM is more fluctuant than SARS-CoV ( Figure S2 ). This predicts that the binding affinity of R3 may be weakened in SARS-CoV-2 bound to ACE2. This was verified in the latter analysis of binding free energy. The origin of the above difference between the two RBDs in this region is attributed to the four mutations (Ser443, Lys444, Val445, Gly446) of the SARS-CoV-2 RBD (Figures 1a and S3 ). The above results indicate that different interaction patterns at the binding interface are adopted for these two complexes. We calculated the binding free energy of the two complexes and decomposed the binding energy into each pair of residues at the interface. The SARS-CoV-2 binds to ACE2 stronger than SARS-CoV by 5.6 kcal/mol due to enhanced electrostatic and van der Waals interactions (Table 1 ). Figure 2a shows the 10 pairs of residues with the highest binding affinity and the distribution of these residues on the RBM of SARS-CoV and SARS-CoV-2. The higher binding affinity of SARS-CoV-2 with ACE2 is also manifested in higher pairwise interactions. Four residue pairs were preserved in SARS-CoV-2, including Asp38-Tyr436 (449), Lys353-Tyr491 (505), Glu37-Tyr491 (505), and Tyr 83-Asn473 (487) (the number in parentheses is the corresponding residue id of the SARS-CoV-2 RBD). The interaction of the two pairs is strengthened such as Lys353-Tyr491 (505) and Tyr 83-Asn473 (487), while the interaction of the other two pairs is weakened. This indicated that even the preserved pair interaction in SARS-CoV-2&ACE2 is also strengthened or weakened compared with SARS-CoV&ACE2. Some pair interaction originates from the mutations of the SARS-CoV-2 RBD, which has no corresponding pairs in SARS-CoV&ACE2. Three of the four residue pairs with the strongest interaction resulted from the mutations of the SARS-CoV-2 RBD, including a strong interaction of salt bridge Asp30&Lys417 (Figures 1a and 2a) . These mutations make a significant contribution to the higher binding affinity of the SARS-CoV-2 RBD.

Beyond this, we found that a different interaction mode is adopted in SARS-CoV-2&ACE2. There are three residue pairs with strong interaction located at R2 in SARS-CoV-2&ACE2, while there is only one pair located at R2 in SARS-CoV&ACE2. Therefore, the interaction distribution on the binding interface changes in the SARS-CoV-2&ACE2 complex.

To show this change more clearly, we decomposed the free energy contributions in a per-residue basis as shown in Figure  2b . Compared with the SARS-CoV RBD, more binding energy is distributed in R1 and R2 in the SARS-CoV-2 RBD, especially in R2. Nearly 50% of the binding energy of the SARS-CoV-2 RBD was distributed in R2, far more than the 23% in the SARS-CoV RBD (Figure 2b ). Lys 417 (in the salt bridge Asp30&Lys417), Leu455, Phe456, and Gln493 located at R2 of the SARS-CoV-2 RBD make a significant contribution to this change. The stronger interaction between the middle part of the helical bundle of ACE2 and R2 makes the SARS-CoV-2 RBD more closely integrate with ACE2 than the SARS-CoV RBD, with less water molecules accessible to the binding interface ( Figure S3 ). This can also be illustrated by a larger E GB contribution of SARS-CoV-2&ACE2 binding free energy (Table 1) . Also, the interaction distributed on the R3 bound to beta-hairpin of ACE2 is weak, which contributes to the larger fluctuation of this region (Figure 1c,f) .

Different Dissociation Process. We constructed an "artificial" dissociation model to gather some information on the recognition mechanism in the binding process of ACE2 and the two RBDs. The mass centers of ACE2 and RBD separated along the x direction at a constant speed, as shown in Figure 3a . During the dissociation, the force applied on SARS-CoV-2&ACE2 lasts longer than that applied on SARS-CoV&ACE2 (Figure 3b ). This demonstrates that more effort is needed to dissociate the complex of the SARS-CoV-2 RBD and ACE2. This can also be verified in Figure 3c , which shows the interaction energy between ACE2 and the two RBDs. The higher binding affinity of the SARS-CoV-2 RBD made it difficult to be dissociated and start dissociating after longer simulation. SARS-CoV&ACE2 and SARS-CoV-2&ACE2 started dissociating at 1.3 ns (red dotted line) and 3.5 ns (blue dotted line), respectively. We found that the dissociation of SARS-CoV-2&ACE2 is obviously controlled by the Coulomb interaction. The long-range interaction from Coulomb distribution could make SARS-CoV-2 RBD easier to recognize ACE2 compared with the SARS-CoV RBD. The distances of Cα of 10 residue pairs obtained in Figure 2 during SARS-CoV-2&ACE2 dissociation show the coordinated distance variation of first 8 pairs and the changes are smaller than SARS-CoV&ACE2 (Figure 3d,e) . Interestingly, the two residue pairs maintain approximately constant distance during the dissociation in both cases, which are Y-N (Tyr83&Asn473) and Q-N (Gln24&Asn473) in SARS-CoV&ACE2 and Y-N (Tyr83&Asn487) and Q-A (Gln24&Ala475) in SARS-CoV-2&ACE2, respectively (Figures 3d,e and S4 ). The residues Tyr83 and Gln24 lie near the N-terminal of the helical bundle of ACE2. The residue Asn473 of the SARS-CoV RBD and the residue Asn487, Ala475 of SARS-CoV-2 RBD are located at the R1 region of the RBD (Figures S4 and 3h,k) . Therefore, we could assume that the N-terminal of the helical bundle of ACE2 might be first recognized by R1 of the RBD in the binding dynamics and plays an important role in the whole binding process. These results also show that SARS-CoV and The van der Waals contribution, electrostatic contribution, polar solvation term (E GB ), and nonpolar solvation term (E Surf ) were also listed. Difference of binding affinity (ΔG bind MM-GBSA) between two RBDs is 5.6 kcal/mol. Langmuir pubs.acs.org/Langmuir Article SARS-CoV-2 adopt similar recognization mechanisms to bind with ACE2.

Design of a New Disinfectant. Based on the understanding of the binding mode of SARS-CoV-2 RBD and ACE2, we proposed a strategy to destroy the SARS-CoV-2 RBD by employing endopeptidase. Enzyme preparation for disinfection is a safe, skin-friendly, convenient way to block the spread of SARS-CoV-2. Endoproteinases are proteolytic peptidases that can break the peptide bonds of nonterminal residues. They are usually very specific for certain residues and can selectively cleave certain positions of a protein. For example, glutamyl endopeptidase 27 can hydrolyze peptide bonds after the negatively charged Glu or Asp residues, with a higher preference to the former. We used five kinds of endopeptidases, which have preferential cleavage of Glu, Gly, Lys, Pro, or Ala, to cut the SARS-CoV-2 RBD (Figure 4 ). There are 5, 14, 8, 9, and 10 cleavage sites on the RBD of Glu, Gly, Lys, Pro, and Ala, respectively. Therefore, the SARS-CoV-2 RBD will be cut into several fragments by specific endopeptidases. Results of molecular dynamics simulation showed that the conformation of the RBM of the cleaved RBD has an obvious change except for the case of Ala cleavage. For the cases of Glu, Gly, and Pro cleavage, the R1 of the RBM, which plays an important role in ACE2 recognition, is almost completely destroyed. Hence, Glu, Gly, and Pro cleavage can result in the failure of its binding with ACE2. Although the R1 of the RBM in the case of Pro is mainly preserved after cleavage, it bends a lot with a smaller angle between R1 and R2, also inhibiting its binding with ACE2 ( Figure S6 ). The Ala cleavage brings a tiny change of the RBM conformation and its binding with ACE2 cannot be blocked (Figures 4 and S7 ). Our simulations indicated that endopeptidases which can selectively cleave Glu, Gly, Lys, and Pro are able to destroy the structure of the RBM of SARS-CoV-2. These endopeptidases can be utilized as the major components of new disinfectants to block the virus spread.

We thoroughly studied the different characteristics of SARS-CoV-2 compared with SARS-CoV using molecular dynamics simulations. A different binding mode is adopted in the complex of SARS-CoV-2&ACE2. The dissociation process of the binding complexes reveals that R1 of the RBD is vital in ACE2 recognition. The conformation destruction of R1 might lead to the failure of binding with ACE2. These findings inspire us to propose a endopeptidase-based disinfectant for the disinfection of SARS-CoV-2, which is suitable for application on skin, PPE, and in nasal sprays. The peculiarities of SARS-CoV-2 revealed by our simulations and the proposed strategy of designing the disinfectant provide some guidance for the development of clinical interventions in the future.

■ METHODS MD Simulation. All molecular dynamics simulations were performed using GROMACS 2019.3. 28 The simulations used the Amber ff99SB-ILDN force field for proteins and the TIP3P model for water. 29 The initial structures of SARS-CoV and SARS-CoV-2 bound to ACE2 were from equilibrium structures in 10 μs-long trajectories Figure 2 . Binding free energy distribution on the binding interface. (a) Pairwise interaction between the two residues obtained from binding free energy decomposition. Here, just 10 pairs of residues with the highest binding energy were listed. In each pair, the first residue is from ACE2 and the second is from RBD. The residues located at R1, R2, and R3 were masked with a blue circle. The residue pairs with green color contains mutated residues of the SARS-CoV-2 RBD compared with SARS-CoV. (b) Decomposing the free energy contributions in a per-residue basis. The contribution of R1, R2, and R3 to the binding free energy of the RBD and the contribution of the helical bundle and beta-hairpin to the binding free energy of ACE2 were listed. Langmuir pubs.acs.org/Langmuir Article provided by Shaw group. 26 The simulations from Shaw group 26 were initiated from ACE2 in complex with the RBD of S protein from SARS-CoV (PDB ID: 2AJF) 30 and SARS-CoV-2 (PDB ID 6M17). 31 In order to mimic physiological conditions, the system was neutralized and salted with NaCl, with a final concentration of 0.15 M. The default setting of protonation states of the charged residues in GROMACS was adopted and adjusted on the basis of local hydrogenbonding interactions, salt bridge, and Zinc finger. The protein was placed in a periodic cubic box of water molecules represented by the three-point charge TIP3P model, whose boundary is at least 15 Å Langmuir pubs.acs.org/Langmuir Article from any protein atoms. The total system was energy minimized by a succession of steepest descent and conjugate gradient methods. Thereafter, the solvent was equilibrated for 10 ns at constant temperature (310 K) and pressure (1 bar) (NPT) by restraining the positions of the protein atoms followed by NPT equilibration for another 100 ns without position restrain. The final production runs (NPT) lasted for 100 ns. The binding free energies (ΔG bind ) were calculated with the Molecular Mechanics Generalized Born Surface Area (MM-GBSA) method by using MMPBSA.py 32 on 200 frames taken from the final production runs at constant temperature (310 K) and pressure (1 bar) (NPT). In our calculations, the entropic contribution of the free energy was not considered. We used V-rescale thermostat 33 and Parrinello−Rahman barostat 34 to keep the temperature and pressure constant, respectively. The cutoff radius for neighbor searching and nonbonded interactions was taken to be 10 Å, and all bonds were constrained using the LINCS algorithm. 35 All the simulation results were visualized by VMD. 36 Disinfectant Design. Five kinds of endopeptidases, which have preferential cleavage of Glu, Gly, Lys, Pro, or Ala, to cut the SARS-CoV-2 RBD, were selected as candidates for disinfectant design. The peptide bond of the corresponding cleavage sites on the RBD of Glu, Gly, Lys, Pro, and Ala was broken in MD simulation settings. The cleaved structure was equilibrated for 200 ns of NPT simulation with the same settings in the section of MD Simulation. The equilibrated structure of SARS-CoV-2 RBD was aligned to the noncleaved structure to display the destructiveness of the disinfectant.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.langmuir.0c02911.

Graphical representations of RMSD of two complexes versus simulation time; residues in SARS-CoV-2 RBD different from SARS-CoV RBD; water molecules accessible to the binding interface of SARS-CoV&ACE2 and SARS-CoV-2&ACE2; residue pairs with approximately constant distance during dissociation; and structural superimposition of SARS-CoV-2 RBD and cleaved RBD by glycyl endopeptidase (PDF)

A new coronavirus associated with human respiratory disease in China

A novel coronavirus outbreak of global health concern

China Novel, C. A Novel Coronavirus from Patients with Pneumonia in China

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation

Cryo-electron microscopy structure of a coronavirus spike glycoprotein trimer

Tectonic conformational changes of a coronavirus spike glycoprotein promote membrane fusion

Veesler, D. Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein

Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor

A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster

Potent binding of 2019 novel coronavirus spike protein by a SARS coronavirus-specific human monoclonal antibody

Comparing the binding interactions in the receptor binding domains of SARS-CoV-2 and SARS-CoV

The SARS-CoV-2 exerts a distinctive strategy for interacting with the ACE2 human receptor

Is the Rigidity of SARS-CoV-2 Spike Receptor-Binding Motif the Hallmark for Its Enhanced Infectivity? Insights from All-Atoms Simulations

Human neutralizing antibodies elicited by SARS-CoV-2 infection

A noncompeting pair of human neutralizing antibodies block COVID-19 virus binding to its receptor ACE2

A highly conserved cryptic epitope in the receptor binding domains of SARS-CoV-2 and SARS-CoV

Computational simulations reveal the binding dynamics between human ACE2 and the receptor binding domain of SARS-CoV-2 spike protein

Does SARS-CoV-2 Bind to Human ACE2 More Strongly Than Does SARS-CoV?

Computational alanine scanning and structural analysis of the SARS-CoV-2 spike protein/angiotensin-converting enzyme 2 complex

Enhanced receptor binding of SARS-CoV-2 through networks of hydrogen-bonding and hydrophobic interactions

In vitro efficacy of a copper iodine complex PPE disinfectant for SARS-CoV-2 inactivation

Can N95 Respirators Be Reused after Disinfection? How Many Times?

Molecular Dynamics Simulations Related to SARS-CoV-2. D. E. Shaw Research Technical Data

Algorithms for highly efficient, load-balanced, and scalable molecular simulation

Comparison of simple potential functions for simulating liquid water

Structure of SARS coronavirus spike receptor-binding domain complexed with receptor

Structural basis for the recognition of SARS-CoV-2 by full-length human ACE2

py: An Efficient Program for End-State Free Energy Calculations

Canonical sampling through velocity rescaling

Crystal Structure and Pair Potentials: A Molecular-Dynamics Study

LINCS: A linear constraint solver for molecular simulations

VMD: Visual molecular dynamics

The authors declare no competing financial interest.

This work was supported by the Beijing Natural Science Foundation (JQ18006) and the National Key Research and Development Plan of China (2016YFA0204300). The calculations were completed on the "Explorer 100" cluster system of Tsinghua HPC Platform.