Are the integrin binding motifs within SARS CoV-2 spike protein and MHC class II alleles playing the key role in COVID-19?

Abstract

The previous studies on the RGD motif (aa403-405) within the SARS CoV-2 spike (S) protein receptor binding domain (RBD) suggest that the RGD motif binding integrin(s) may play an important role in infection of the host cells. We also discussed the possible role of two other integrin binding motifs that are present in S protein: LDI (aa585-587) and ECD (661-663), the motifs used by some other viruses in the course of infection. The MultiFOLD models for protein structure analysis have shown that the ECD motif is clearly accessible in the S protein, whereas the RGD and LDI motifs are partially accessible. Furthermore, the amino acids that are present in Epstein-Barr virus protein (EBV) gp42 playing very important role in binding to the HLA-DRB1 molecule and in the subsequent immune response evasion, are also present in the S protein heptad repeat-2. Our MultiFOLD model analyses have shown that these amino acids are clearly accessible on the surface in each S protein chain as monomers and in the homotrimer complex and bind to HLA-DRB1 β chain. Therefore, they may have the identical role in SARS CoV-2 immune evasion as in EBV infection. The prediction analyses of the MHC class II binding peptides within the S protein have shown that the RGD motif is present in the core 9-mer peptide IRGDEVRQI within the two HLA-DRB1*03:01 and HLA-DRB3*01.01 strong binding 15-mer peptides suggesting that RGD motif may be the potential immune epitope. Accordingly, infected HLA-DRB1*03:01 or HLA-DRB3*01.01 positive individuals may develop high affinity anti-RGD motif antibodies that react with the RGD motif in the host proteins, like fibrinogen, thrombin or von Willebrand factor, affecting haemostasis or participating in autoimmune disorders.

Keywords: COVID-19; MultiFOLD model; SARS CoV-2 spike protein; autoimmunity; deregulated coagulation; immune evasion; integrin binding motifs.

Figure 1

The 3D structures of SARS CoV-2 spike protein (P0DTC2, SPIKE_SARS2), ADAM17 metalloproteinase (P78536, ADAM17_HUMAN), zinc metalloproteinase-disintegrin-like acurhagin (Q9W6M5, VM3AH_DEIAC) and influenza haemagglutinin H1 (Q9WFX3). The surface views are shown with the key motifs labelled and highlighted in blue. (A) Top view of SARS CoV-2 spikes complex (PDB entry 6VXX). Chain A is shown in green, B in cyan, and C in magenta. (B) Side view of the SARS CoV-2 spike structure shown in (A–C) Chain A of the SARS CoV-2 spike structure. (D) MultiFOLD model of ADAM17 metalloproteinase (plDDT=0.833, pTM=0.760). The disordered residues from 1-33 and 711 onwards are removed for clarity. (E) MultiFOLD model of zinc metalloproteinase-disintegrin-like acurhagin (plDDT=0.912, pTM=0.856). (F) Side view of influenza haemagglutinin H1 complex (PDB entry 4GXX). Images are rendered using PyMOL (http://www.pymol.org/).

Figure 2

The amino acid sequence identity of the single spike and M proteins from different human and bat coronaviruses with the SARS CoV-2 spike (P0DTC2) and M ((P0DTC5) protein as determined by pairwise sequence alignment (%). NA, sequence not available. The alignments were performed using EMBOSS Needle algorithm (71, 72). *Bat SARS-like CoVs. S protein sequences: UniProt data base (17). Performed using STATISTICA, Statsoft.

Figure 3

The alignments of EBV gp42 HLA-DRB1 binding domain sequence with the SARS CoV-2 and SARS CoV-1 S protein HR2 sequence as well with the HR2 sequences from different human and bat coronaviruses. The contact domain to the HLA-DRB1 b chain in the EBV gp42 protein including the binding amino acids: R154, N155, R157 and E160 (76) (* bold letters) aligned with the S protein HR2 amino acid sequence from SARS CoV-2, SARS CoV-1 and different human and bat coronaviruses.

Figure 4

The 3D structures of the C-terminus of the SARS CoV-2 spike protein (P0DTC2, SPIKE_SARS2) and Epstein-Barr Virus (EBV) gp42 protein bound to the MHC class II Receptors HLA-DRB1 (P01911, DRB1_HUMAN) and HLA-DRB3 (P79483, DRB3_HUMAN) β-chains. The surface views are shown with the key residues labelled and highlighted in blue. (A) MultiFOLD model of the SARS CoV-2 spike (P0DTC2 - SPIKE_SARS2) homotrimer complex C-terminus region from residue 1194 onward (plDDT=0.751, pTM=0.599). (B) Crystal structure of EVB gp42 (magenta) bound to the HLA-DR1 complex (PDB ID - 1KG0). (C) MultiFOLD model of the SARS CoV-2 spike (P0DTC2 - SPIKE_SARS2) homotrimer complex C-terminus bound to HLA-DRB1 (P01911, DRB1_HUMAN) (plDDT=0.651, pTM= 0.497).(D) MultiFOLD model of the SARS CoV-2 spike (P0DTC2 - SPIKE_SARS2) homotrimer complex C-terminus bound to HLA-DRB3 (P79483, DRB3_HUMAN) (plDDT=0.627, pTM=0.485). Images of models were rendered using PyMOL (http://www.pymol.org/).