Dear Editor,
New SARS-CoV-2 variants started to evolve and spread worldwide in late 2020 and until today. The new variants bear many modifications (insertion, deletion, and mutations) in the spike protein. Some lie in the receptor-binding domain (RBD), which mediates viral entry to the host cell utilizing different host-cell agents. The new variant Omicron (B.1.1.529) shed scientific concern as its spike bear many mutations (A67V, Δ69–70, T95I, G142D/Δ143–145, Δ211/L212I, ins214EPE, G339D, S371L, S373P, S375F, K417N, N440K, G446S, S477N, T478K, E484A, Q493R, G496S, Q498R, N501Y, Y505H, T547K, D614G, H655Y, N679K, P681H, N764K, D796Y, N856K, Q954H, N969K, and L981F) that alter vaccination strategy to reduce the infection spread. Angiotensin-converting enzyme 2 (ACE2) was reported as the principal entry agent for SARS-CoV-2, but not the only gate. Different host-cell receptors are defined to mediate SARS-CoV-2 recognition and entry (1). In this Journal, we previously published a predicted SARS-CoV-2 spike-host cell surface receptor, glucose-regulated protein 78 (GRP78), binding site (2), which has been confirmed experimentally by Carlos et al. lately (3). The receptor-binding domain region of the spike protein can be recognized by GRP78 substrate-binding domain β (SBDβ), where the C480-C488 was the predicted binding motif of the spike. Mutation at D484 (to Q, K, D, G, V, and A) was detected in delta and Omicron variants, among other mutations which impacted the binding of Cs-GRP78 to the viral spike protein as reported by our group previously (Beta and Gamma variants) (4).
In this report, we simulate the cell-surface GRP78 (Cs-GRP78) recognition to the spike of Omicron variant of SARS-CoV-2 after performing 100 ns molecular dynamics simulation (MDS) to the wildtype (WT) & Omicron spikes and the GRP78 structure (PDB ID: 5E84, Chain A). PyMOL v2.2.2 was utilized to perform the mutations in the RBD of the spike before the minimization and MDS performed by Nanoscale molecular dynamics software (NAMD) 2.13 software using CHARMM 36 force field and TIP3P water model (5, 6). The MDS calculations and analysis were performed on the King Abdullah University of Science and Technology (KAUST) supercomputing facility, SHAHEEN (project no. k1482), and a local workstation. The input files for MDS were generated using the CHARMM-GUI webserver (7). Temperature, pressure, and salt concentration were set to be 310 K, 1 atm, and 0.154 M NaCl as the physiological conditions. Before the simulation, the system was minimized for 20,000 steps in a constant number of atoms, constant volume, constant temperature (NVT) ensemble using a conjugate gradient algorithm. The system was then equilibrated in an NPT ensemble for one nanosecond period before the 100 ns production run.
Additionally, the binding of GRP78 to the spikes was predicted using HADDOCK 2.4 webserver (8). We docked GRP78 with both WT and mutant SARS-CoV-2 spike RBD. GRP78 and SARS-CoV-2 spike RBD's active sites were selected to be T428, V429, V432, T434, F451, S452, V457 & I489, and C480-C488, respectively. Other options of HADDOCK were kept as default during the docking. The carbohydrate moieties (NAG) attached to the proteins were held in the structure during the simulations. After docking, the best-scored complexes were used to predict the binding energies using PRODIGY of the WeNMR portals (9).
RBD dynamics of the WT spike versus the Omicron variant
Fig. 1 shows the MDS analysis of the WT spike (blue curve) and the Omicron variant (orange curve) spike. The root-mean-square deviation (RMSD) in Å (A), the radius of gyration (RoG) in Å (B), surface accessible surface area (SASA) in Å2 (C), and the number of H-bonds (D) is plotted against the simulation time in ns. Additionally, the per-residue root-mean-square fluctuations are plotted for the two spike RBDs (Fig. 1E). The two systems are equilibrated at around 3.4 Å (WT) and 2.3 Å (Omicron) during the first 50 ns as reflected from the RMSD curves. Additionally, the two systems are stable during the simulation period as reflected from the RoG, SASA, and the total number of H-bonds values (17.5 Å, 11,200 Å2, and 225, for the WT RBD and 18 Å, 11,000 Å2, and 240, for the Omicron RBD, respectively).
Fig. 1.
The molecular dynamics simulation analysis of the WT RBD spike (blue curves) and the Omicron RBD variant (orange curves). (A) the root-mean-square deviation versus the simulation time. (B) the radius of gyration versus the simulation time. (C) the surface accessible surface area versus the simulation time. (D) the total number of H-bonds versus the simulation time. (E) the per-residue root-mean-square fluctuations among Omicron RBD representative structure taken at 43.1 ns. GRP78 binding site of the spike is labeled and depicted in the red cartoon, while the region of high deviation from the WT is shown in the yellow cartoon.
On the other hand, the RMSF (see Fig. 1E) of the WT RBD (blue curve) show a significantly elevated level of fluctuations in the region 470–490 of the protein (yellow cartoon) and from which the GRP78 recognition site (C480-C488) in the WT RBD is at least two-fold more flexible than that of the Omicron RBD. This indicates the stabilization of the Omicron variant of the spike RBD at the GRP78 binding site exerted by the RDB mutants.
The binding affinity of GRP78 to WT spike versus the Omicron variant
After MDS, the trajectories are subjected to clustering using TTclust software (10). Two different clusters are found in GRP78 and Omicron spike RBD trajectories, while three are found in the WT spike RBD. A representation member from each cluster is used in the protein-protein docking of HADDOCK. Fig. 2 A shows the average HADDOCK score values (columns) and their corresponding binding affinity values (points) calculated using PRODIGY software for the WT RBD-GRP78 (blue) and the Omicron RBD-GRP78 (green) complexes. The average binding affinity of the Omicron RBD to GRP78 is lower (−9.68 ± 0.63 kcal/mol) compared to the binding affinity of the WT RBD to GRP78 (−8.83 ± 0.60 kcal/mol). This reflects the higher probability of the association between the Omicron spike and the host-cell surface receptor GRP78 than the WT spike. A result we reported before in alpha and beta variants of SARS-CoV-2 as well (4).
Fig. 2.
The average binding affinity (in kcal/mol) calculated by PRODIGY (line) and the average HADDOCK scores (columns) for the WT RBD-GRP78 complexes (blue) and the Omicron RBD-GRP78 complexes (green) calculated from the representative cluster members of each protein after the MDS trajectory analysis.
Table 1 summarizes the established interactions upon docking the two GRP78 representative structures into the three WT RBD conformations and the two Omicron RBD conformations. The average number of the formed interactions is increased in the case of Omicron RBD variants compared to the WT RBDs. On average, eight hydrophobic contacts and 8 H-bonds are formed in the case of Omicron RBD docking against GRP78. Those numbers are 4.67 (hydrophobic contacts) and 9.33 (H-bonds) for the docking of GRP78 against WT RBD variants. The mutant E484A has an impact on the recognition of the C480-C488 region of the spike as it raises the average hydrophobicity index (Kyte & Doolittle) (see Fig. 2B) of the peptide to be closer to the value of the Pep42 cyclic peptide that was confirmed before to bind Cs-GRP78 over cancer cells (11). In addition, the E484A mutant increased the number of hydrophobic contacts formed between the spike and the GRP78 SBD β, as reflected in Table 1 (bold residues).
Table 1.
The interactions established upon docking the GRP78 into WT RBD and Omicron RBD of SARS-CoV-2 spike. Red residues are the amino acids involved in salt bridge formation, while blue residues form π-cation interactions upon docking.
| CLUSTER NUMBER | NUMBER OF HYDROGEN BONDS | AMINO ACIDS IN GRP78 | AMINO ACIDS IN RBD | NUMBER OF HYDROPHOBIC INTERACTIONS | AMINO ACIDS IN GRP78 | AMINO ACIDS IN RBD | |
|---|---|---|---|---|---|---|---|
| WT RBD | C1_grp1 | 111 | V245(2), D348, S349, D350(2), D350, K435, N440, Q449(2), and S452 | T345, R346(4), K444, N450, N481, V483(2) R509, and R509 | 2 | E347 and F451 | K444 and F486 |
| C1_grp2 | 11 | E427, T428, V429, S452(2), G454, P487, R488(2), and G489(2) | P479, C480, N481, G485, N487, C488, Q493(4), and S494 | 5 | F451, V453, V457(2), and P485 | Y449, P479, and F486(3) | |
| C2_grp1 | 8 | T428, G430, S452(2), T456(2), T458, and R488 | Q474, N481(2), E484(2), G485, N487, and Y505 | 5 | I426, V429, F451, I459, and V490 | T478, F486(3), and Y489 | |
| C2_grp2 | 71 | T428, T434, K435, K435, S452(3), and Q492 | N481(4), E484(2), E484, and G485 | 6 | T428, V429, T434, F451, and V457(2) | N481, E484(2), F486(2), and Y489 | |
| C3_grp1 | 9 | V429, T434, K447, S448(2), I450, S452(2), and G454 | G446, Y449(2), T478, N481(2), V483, E484, and G485 | 3 | I450, V457, and I459 | V483 and F486(2) | |
| C3_grp2 | 101 | T428, T434, K435, V442, T445(2), K446, Q449, S452(2), and Q492 | Y449, T478, P479, N481, V483, E484, E484, G485, Q498(2), and T500 | 71 | I426, T434, V442, K447, K447, F451, V457, and I459 | Y449, Y449, E484, F486(4), and T500 | |
| Omicron RBD | C1_grp1 | 10 | T428, V429, G430, Q449(2), S452(2), Q492(2), and T514 | Y449, N481, G482, V483, Y489(2), F490, and R493(3) | 12 | I426(2), T434, L436, Q449, I450, F451(2), V457, I459, and V490(2) | A484(3), F486(6), Y489, and F490(2) |
| C1_grp2 | 61 | E121, E121, V429, S452(2), T458, and D525 | N487(2), C488, R493, R498, T500, and Y501 | 7 | V429(2), V432(2), T434, F451, and V453 | I472(2), V483, A484, F486(2), and Y489 | |
| C2_grp1 | 7 | V429, Q449, S452(4), and G454 | K478, C480, N481(3), V483, and C488 | 6 | I426, V429, T434, V453(2), and V457 | K478(2), A484(2), F486, and Y489 | |
| C2_grp2 | 9 | E347(2), V429, K435, I450, S452(2), T458, and Q492 | K444, V445, E471(4), V483, F490, and S494 | 7 | E347, T428, V432(2), T434, I459, and P467 | K444, Y449, V483, and F490(4) | |
Conclusively, in this letter, we shed light on the modified affinity of the spike RBD of the new variant Omicron against the host cell-surface GRP78. This recognition could be targeted by peptides, antibodies, or phytochemicals (12).
Declaration of Competing Interest
All the authors declare no competing interest in this work.
References
- 1.Nassar A., Ibrahim I.M., Amin F.G., Magdy M., Elgharib A.M., Azzam E.B., et al. A review of human coronaviruses’ receptors: the host-cell targets for the crown bearing viruses. Molecules. 2021;26(21):6455. doi: 10.3390/molecules26216455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ibrahim I.M., Abdelmalek D.H., Elshahat M.E., Elfiky A.A. COVID-19 spike-host cell receptor GRP78 binding site prediction. J Infect. 2020;80(5):554–562. doi: 10.1016/j.jinf.2020.02.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Carlos A.J., Ha D.P., Yeh D.W., Van Krieken R., Tseng C.C., Zhang P., et al. The chaperone GRP78 is a host auxiliary factor for SARS-CoV-2 and GRP78 depleting antibody blocks viral entry and infection. J Biol Chem. 2021;296 doi: 10.1016/j.jbc.2021.100759. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ibrahim I.M., Elfiky A.A., Elgohary A.M. Recognition through GRP78 is enhanced in the UK, South African, and Brazilian variants of SARS-CoV-2; An in silico perspective. Biochem Biophys Res Commun. 2021;562:89–93. doi: 10.1016/j.bbrc.2021.05.058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Phillips J.C., Braun R., Wang W., Gumbart J., Tajkhorshid E., Villa E., et al. Scalable molecular dynamics with NAMD. J Comput Chem. 2005;26(16):1781–1802. doi: 10.1002/jcc.20289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Seeliger D., de Groot B.L. Ligand docking and binding site analysis with PyMOL and Autodock/Vina. J Comput Aided Mol Des. 2010;24(5):417–422. doi: 10.1007/s10822-010-9352-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Lee J., Cheng X., Swails J.M., Yeom M.S., Eastman P.K., Lemkul J.A., et al. CHARMM-GUI Input Generator for NAMD, GROMACS, AMBER, OpenMM, and CHARMM/OpenMM Simulations Using the CHARMM36 Additive Force Field. J Chem Theory Comput. 2016;12(1):405–413. doi: 10.1021/acs.jctc.5b00935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.van Dijk A.D., Bonvin A.M. Solvated docking: introducing water into the modelling of biomolecular complexes. Bioinformatics. 2006;22(19):2340–2347. doi: 10.1093/bioinformatics/btl395. [DOI] [PubMed] [Google Scholar]
- 9.Wassenaar T.A., van Dijk M., Loureiro-Ferreira N., van der Schot G., de Vries S.J., Schmitz C., et al. WeNMR: structural Biology on the Grid. J Grid Comput. 2012;10(4):743–767. [Google Scholar]
- 10.Tubiana T., Carvaillo J.C., Boulard Y., Bressanelli S. TTClust: a versatile molecular simulation trajectory clustering program with graphical summaries. J Chem Inf Model. 2018;58(11):2178–2182. doi: 10.1021/acs.jcim.8b00512. [DOI] [PubMed] [Google Scholar]
- 11.Kim Y., Lillo A.M., Steiniger S.C., Liu Y., Ballatore C., Anichini A., et al. Targeting heat shock proteins on cancer cells: selection, characterization, and cell-penetrating properties of a peptidic GRP78 ligand. Biochemistry. 2006;45(31):9434–9444. doi: 10.1021/bi060264j. [DOI] [PubMed] [Google Scholar]
- 12.Elfiky A.A., Baghdady A.M., Ali S.A., Ahmed M.I. GRP78 targeting: hitting two birds with a stone. Life Sci. 2020;260 doi: 10.1016/j.lfs.2020.118317. [DOI] [PMC free article] [PubMed] [Google Scholar]