key: cord-0704108-f8k4agdg authors: Singh, Jasdeep; Kar, Sudeshna; Hasnain, Seyed Ehtesham; Ganguly, Surajit title: SARS-CoV-2 ORF8 can fold into human factor 1 catalytic domain binding site on complement C3b: Predict functional mimicry date: 2020-08-05 journal: bioRxiv DOI: 10.1101/2020.06.08.107011 sha: b6f791f9a1314c92071faff76c2d693d81aa4d92 doc_id: 704108 cord_uid: f8k4agdg Pathogens are often known to use host factor mimicry to take evolutionary advantage. As the function of the non-structural ORF8 protein of SARS-CoV-2 in the context of host-pathogen relationship is still obscure, we investigated its role in host factor mimicry using computational protein modelling techniques. Modest sequence similarity of ORF8 of SARS-CoV-2 with the substrate binding site within the C-terminus serine-protease catalytic domain of human complement factor 1 (F1; PDB ID: 2XRC), prompted us to verify their resemblance at the structural level. The modelled ORF8 protein was found to superimpose on the F1 fragment. Further, protein-protein interaction simulation confirmed ORF8 binding to C3b, an endogenous substrate of F1, via F1-interacting region on C3b. Docking results suggest ORF8 to occupy the binding groove adjacent to the conserved “arginine-serine” (RS) F1-mediated cleavage sites on C3b. Comparative H-bond interaction dynamics indicated ORF8/C3b binding to be of higher affinity than the F1/C3b interaction. Hence, ORF8 is predicted to inhibit C3b proteolysis by competing with F1 for C3b binding using molecular mimicry with a possibility of triggering unregulated complement activation. This could offer a mechanistic premise for the unrestrained complement activation observed in large number of SARS-CoV-2 infected patients. The world is currently facing an unprecedented pandemic caused by the transmission of a novel corona virus, named as severe acute respiratory syndrome coronavirus 2 or SARS-CoV-2 [1, 2] . Following the identification of the virus, the early phase of clinical characterization of the disease revealed that a substantial number of the patients progresses towards acute respiratory distress syndrome (ARDS) eventually leading to multi-organ failure [3] [4] [5] [6] . Though the sequence of events leading to multiple organ failure is still not ascertained, it appears that the virus infection-induced hyperactivation of the complement system along with robust pro-inflammatory responses, vascular thrombus formation and coagulation could play a major role [7] [8] [9] [10] . An emerging hypothesis suggests that the SARS-CoV-2 non-structural ORF8 (Open reading frame 8) protein could participate in immune evasion strategy [11] . However, it does not explain the tissue damaging immune storm that accompanies the pathogenesis. This led us to investigate whether the structural elements of ORF8 can mimic any human host factor that can elicit robust proinflammatory responses as a bystander effect. Diverse pathogens, including viruses, are believed to use molecular mimicries that resemble host factors to acquire evolutionary advantage [12, 13] . Towards this goal, we observed that ORF8 has a modest level of amino acid sequence similarity with the substrate binding site located within the C-terminal domain of human complement factor 1 (F1; PDB ID: 2XRC; Ref. 14) . Here, we report that ORF8 has optimal structural resemblance (tertiary fold architecture) that might allow it to interact with one of the F1 substrates -human complement C3b, by adopting F1 mimicry. Using protein-protein interaction modelling, we predict ORF8 to bind to C3b via a site that appears to overlap with the F1 interacting interface on C3b. The possibility of competitive sequestration of F1 from C3b binding interface could form the basis of uncontrolled complement activation in the host. 4 Searching (Blastp suite, NCBI) the human (Taxid 9606) PDB protein database using the full length 121 amino acid long SARS-CoV-2 ORF8 protein did not generate any significant hits with E (expected threshold) value set at 1.0. However, increasing the E to 10 (default) revealed a lower level of sequence similarity of ORF8 (48% similarity, 25% identity, E value 1.4) with C-terminal domain (amino acid residues 500 to 557) of F1. The resemblance of the ORF8 sequence was with the substrate binding region of F1, embedded within the catalytic domain, as schematically shown in Figure 1A . According to the NCBI Conserved Domain database, F1 protein is a trypsin-like serine protease with the domains as described ( Figure 1A ). The catalytic domain extends from 322 to 557 residues after the zymogen cleavage site between R321 and I 322 residues. This F1 catalytic domain is subdivided into an active site, composed of consensus triad residues -H 363, D 411, and S 507; and, a substrate binding site from residues D 501 to R 557 [15] . In comparison, these triad catalytic residues are missing in ORF8. However, it has resemblance from E 57 (corresponding D 501 in F1) to R 115 (corresponding R 557 in F1) in the substrate binding region ( Figure 1A ). Modelling of ORF8, followed by overlaying it on the template F1 (PDB id: 2xrc), the sequence similarity was reflected at the structural level as well ( Figure 1B) ORF8 interactions with C3b were found to occur via multiple hydrogen bonding with residues including N 939, H 1349, Y 1482 and others ( Figure 2B ; Tables 1 and 2 ). These interacting residues are encompassed by F1 interacting region ( Figure 2A ) on chain B of C3b. Furthermore, ORF8 appears to be docked in close proximity to the R 1303 / S1304 and R1321 / S 1322 (Blue spheres in Figure 2B ), the target of F1 catalytic site for proteolysis. Moreover, the lack of conserved serine protease catalytic triad residues H, D and S, as present in F1, might allow ORF8 to act like a competitive inhibitory ligand, hindering access to F1 binding. Thus, ORF8 might subsequently shield proteolytic cleavage between residues R 1303 and S 1321 (Figure 2A ) or other sequential cleavage sites on C3b CUB (C1r/C1s, Uegf, Bmp1) domains, as described previously [14] , with a potential to hinder C3b down-regulation and attenuate opsonin generation. For validation of the accuracy of the docking method adopted and in an effort to minimize the false positive docking solutions, we employed two approaches. In the first approach, we used a mutant form of ORF8, mORF8 bearing amino acid substitutions S24L, 6 V62L and L84S. These mutations were found naturally in viral isolates ( Supplementary Fig. S1 ). Mutation induced structural stability analysis displayed S24L and V62L substitutions to have minor stabilization effects on ORF8 architecture (ΔΔG wild-type→mORF8 +0.36 and +0.09 kcalmol -1 , respectively). In contrast, the high entropy (globally dominant) L84S mutation showed much higher destabilization effect on ORF8 (ΔΔG wild-type→mORF8 -1.07 kcalmol -1 ) ( Supplementary Fig.2 ). In the top ranked solution for mORF8-C3b docking, mORF8 was bound to a site different from F1 and wild-type ORF8 interacting interfaces (Global energy -15.82) ( Figure 3A and 3B). However, the docking solution which showed binding of mORF8 at the proximity of F1 and ORF8 binding site yielded a much lesser global energy score (-2.9) ( Figure 3A and 3C). This suggests that by rigidly fitting the mORF8 in the F1-C3b interface would make the binding energetically less favourable. Although both ORF8 and mORF8 were structurally conserved (RMSD ~0.2 Å; Figure 3A ), the mutations have clearly perturbed mORF8 binding with C3b. It appears that the preferential interaction interface of mORF8 with C3b are dissimilar from F1-C3b binding interface, indicating that mORF8 is less likely to compete with F1 for the same binding site on C3b. Thus, it appears that the wild type ORF8 conformation is required for precisely fitting into the binding groove of F1-C3b interface. In the second validation approach, we used two randomly selected SARS-CoV-2 proteins -ORF7a (121 residue length; green) protein and 1 -121 amino acid fragment of NSP10 (orange) of SARS-CoV-2, with matching amino acid length as ORF8, as negative controls for modelling interactions with C3b ( Figure 4 ). The top ranked docking results for both proteins show surprising associations with C3b at a common region between amino acid residues 771 -850. The interacting regions appear to be located at an exterior surface, away from the ORF8/F1 binding groove on the C3b protein. Anti-parallel β -sheets sometimes provide stable binding interface for protein-protein interactions [17] . Since both unrelated 7 proteins bind to same site on C3b supported by two partial anti-parallel sheets, it appears that these docking solutions are more of a reflection of algorithmic limitations than representing physiological states. In contrast, the binding of ORF8 with C3b is driven by structural resemblance with F1, and is a testimony to protein mimicry action. It is to be noted that in spite of the projected higher affinity of ORF8 for a shared docking region on C3b, the effort here is to identify the "near-native" conformation that dictates the physiological impact of the ORF8-C3b complex. The top-ranked solution does not always replicate the physiological states. Hence, in the absence of crystal structure of the native ORF8, these docking solutions only present a comparative evaluation of ORF8 in the context of F1 mimicry functions. To The current study highlights immunomodulatory potential of ORF8 protein of SARS-CoV-2 through molecular mimicry of a portion of F1 protein, the negative-feedback regulator of complement pathway. Using a combination of MD simulations and protein-protein docking approach, we have compared binding interactions of ORF8 and F1 proteins with C3b. Our results predict a plausible mechanism of infection-induced host complement 8 hyperactivation coupled with evasion strategy adopted by SARS-CoV-2 through sustained protection of C3b protein via competitive disruption of F1 binding. F1 is known to function as a negative regulator of the host complement system by regulating the levels of C3b complexes [14, 16, 18] . In this context, the resemblance of SARS-COV-2 ORF8 protein with the substrate binding domain of human F1 is intriguing. It Recognition and elimination of viral particles by the host complement system has been known since 1930 [21] . During the adoption of humans as host, viruses have also evolved unique strategies to subvert the complement mediated innate immunity [22] [23] [24] . Encoding orthologs of the complement family of proteins or co-opting the host complement system are documented extensively [25] . Our results predicting the use of molecular mimicry as a strategy by SARS-CoV-2 to deregulate complement system is one such action designed to gain advantage during infection. The prediction of ORF8 as a binding partner of the C3b and competitively blocking the actions of F1, the negative regulator of complement homeostasis, could serve as an intriguing hypothesis explaining the immune surge in SARS-CoV-2 infected patients. Thus, in spite of the absence of experimental correlations, the current work presents stimulating insights into functional role of ORF8 at host-viral interface and warrants further investigation. followed by system minimization by Steepest descent protocol. Equilibration and production runs (25 ns) were carried using parameters detailed in our previous reports [31, 32] . Images were constructed using PyMol while data was analysed using standard gromacs tools. Conflict of interest: None 1 7 The arrows (Black) define outcomes from protein-protein docking of C3b with mORF8 at its two different sites. The species Severe acute respiratory syndrome-related coronavirus: Classifying 2019-nCoV and naming it SARS-CoV-2 Outbreak of pneumonia of unknown etiology in Wuhan, China: The mystery and the miracle Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: A descriptive study Genomic characterisation and epidemiology of 2019 novel coronavirus: Implications for virus origins and receptor binding Emerging genetic diversity among clinical isolates of SARS-CoV-2: Lessons for today SARS-CoV-2 and COVID-19: A genetic, epidemiological, and evolutionary perspective Complement as a target in COVID-19? COVID-19: a collision of complement, coagulation and inflammatory pathways The complement system in COVID-19: friend and foe? COVID-19, microangiopathy, hemostatic activation, and complement Immune evasion via SARS-CoV-2 ORF8 protein? The evolutionary conundrum of pathogen mimicry Host-related sequences in RNA viral genomes Structural basis for complement factor I control and its disease-associated sequence polymorphisms Sequence and evolutionary analysis of the human trypsin subfamily of serine peptidases Regulators of Complement Activity Mediate Inhibitory Mechanisms Through a Common C3B-Binding Mode The supramolecular chemistry of β -sheets Complement: a key system for immune surveillance and homeostasis A Protective Role for Complement C3 Protein during Pandemic 2009 H1N1 and H5N1 Influenza A Virus Infection Baric Complement Activation Contributes to Severe Acute Respiratory Syndrome Coronavirus Pathogenesis mBio A study of vaccinal immunity in rabbits by means of in vitro methods Complement and viral pathogenesis Inhibition of the complement cascade by the major secretory protein of vaccinia virus Innate Immune Evasion by Human Respiratory RNA Viruses Complement Evasion Strategies of Viruses: An Overview SeaView Version 4: A Multiplatform Graphical User Interface for Sequence Alignment and Phylogenetic Tree Building Efficient Unbound Docking of Rigid Molecules PatchDock and SymmDock: servers for rigid and symmetric docking FireDock: Fast Interaction Refinement in Molecular Docking FireDock: a web server for fast interaction refinement in molecular docking L-Asparaginase of Leishmania donovani: Metabolic target and its role in Amphotericin B resistance Arjunolic acid, a peroxisome proliferatoractivated receptor α agonist, regresses cardiac fibrosis by inhibiting non-canonical TGF-β signalling Table 1: Protein protein interaction network between pre-simulated F1 and C3b. Residues involved in inter-molecular H-bonds and salt-bridge formation are indicated Evolution of H-bonds between F1 factor and C3b complement complex (Cyan) and ORF8-C3b complex (Red) over the simulation period of 25 ns