key: cord-0980298-9fmmh34f
authors: Riahi, Saleh; Lee, Jae Hyeon; Wei, Shuai; Cost, Robert; Masiero, Alessandro; Prades, Catherine; Olfati-Saber, Reza; Wendt, Maria; Park, Anna; Qiu, Yu; Zhou, Yanfeng
title: Application of an integrated computational antibody engineering platform to design SARS-CoV-2 neutralizers
date: 2021-03-23
journal: bioRxiv
DOI: 10.1101/2021.03.23.436613
sha: 9fa10cf4da3a293cf0f0291af79bd7fa957eeb30
doc_id: 980298
cord_uid: 9fmmh34f

As the COVID-19 pandemic continues to spread, hundreds of new initiatives including studies on existing medicines are running to fight the disease. To deliver a potentially immediate and lasting treatment to current and emerging SARS-CoV-2 variants, new collaborations and ways of sharing are required to create as many paths forward as possible. Here we leverage our expertise in computational antibody engineering to rationally design/optimize three previously reported SARS-CoV neutralizing antibodies and share our proposal towards anti-SARS-CoV-2 biologics therapeutics. SARS-CoV neutralizing antibodies, m396, 80R, and CR-3022 were chosen as templates due to their diversified epitopes and confirmed neutralization potency against SARS. Structures of variable fragment (Fv) in complex with receptor binding domain (RBD) from SARS-CoV or SARS-CoV2 were subjected to our established in silico antibody engineering platform to improve their binding affinity to SARS-CoV2 and developability profiles. The selected top mutations were ensembled into a focused library for each antibody for further screening. In addition, we convert the selected binders with different epitopes into the trispecific format, aiming to increase potency and to prevent mutational escape. Lastly, to avoid antibody induced virus activation or enhancement, we applied NNAS and DQ mutations to the Fc region to eliminate effector functions and extend half-life.

responses that are particularly effective against highly evolving pathogens [8] . Multi-specific 60 antibody engineering based on a combination of broadly neutralizing antibodies is another 61 highly effective method to target constantly evolving viruses. This design rationale was used to 62 generate a trispecific antibody against HIV [14] . The underlying hypothesis is that targeting 63 different regions of the antigen prevents resistance and escape and further enhances cross 64 reactivity. Similar strategy using tandem linked single domain camelid antibodies showed 65 significant efficacy against both influenza A and B viruses [15] . and P462L, did not eliminate CR3022 neutralization potency [19] . Previous investigations 80 reported that only CR3022 has detectable binding to the SARS-CoV-2 RBD region [20] . P384A 81 mutation in the SAR-CoV-2 RBD was able to return the binding affinity to SARS-CoV levels 82 which suggests that this location plays a vital role in CR3022 neutralization activity. These 83 observations highlight the importance of optimizing the properties of these mAbs to be used for 84 therapeutic or prophylactic purposes against SARS-CoV-2 virus.

Discovery of antibody therapeutics has rapidly evolved in the past few years, and research in 86 lead generation and optimization faces strong challenges in needing high success rates and 87 short timelines. Structure-based rational engineering of antibodies has been shown fast and highly effective in optimizing features of lead candidates, including cross reactivity, potency, 89 developability, and safety profile. Hereto we selected the above mentioned three structurally 90 known anti-SARS-CoV monoclonal antibodies with established neutralization potency and fed 91 them into our computational design pipeline to propose SARS-CoV-2 neutralizing antibodies.

Moreover, combinations of those binders are designed into a multi-specific format aiming to 93 further enhance the anti-viral potency and tolerance to viral evolution in the RBD. from SARS-CoV or SARS-CoV-2 are available; we select three clones, m396, 80R, and CR3022 as 101 our templates, with the filtering criteria of continuously overlapping epitopes, ranging from 102 highly conserved RBD surface to more mutation prone ( Figure 1A&B ).

Developability assessment and engineering at Fv level. The Fv of the candidates were isolated 104 from their complex structure and subjected to computational prediction of developability 105 features including surface patches, chemical degradation of Asp and Asn, and oxidation of Met.

Patch calculation included spatial aggregation propensity (SAP) [21] using Discovery Studio 107 (BIOVIA, Dassault Systèmes) with a 5 Å radius and clustering of residues in the patch analysis 108 using Molecular Operation Environment (MOE) version 2019.0102 [22] . Patches larger than 50 109 Å 2 were selected for further visual inspection. Deamidation and isomerization motifs were 110 analyzed with bioMOE using structure-based prediction models developed by Sydow et al. [23] 111 and Robinson et al. [24] . Risk of methionine oxidation was predicted using sulfur solvent-112 accessible area and 2-shell models with bioMOE [25] . Residue scanning on the patch residues 113 or chemical liability motifs were manually inspected and mutation strategies were made 114 following two criteria: 1) mutation does not impact binding; and 2) mutation reduces patch 115 area.

Structures preparation for SARS-CoV-2 reactivity engineering. All antibody sequences 118 reported here are renumbered using continuous peptide numbering. The RBD from SARS-CoV-119 2 spike structure is used to replace the RBD in the m396 and 80R complexes. For 80R, the single 120 chain Fv (scFv) was split to Fv with standard VH-VL pairing and the linker between VH and VL 121 in the scFv was removed. Antibody residues that are within 6 Å of the RBD are selected and 122 fed to residue scanning in MOE, Rosetta, TopNetTree, and SAAMBE3D. 

In silico mutagenesis and consensus Z-score. For each complex structure, antibody residues 199 within 6 Å from the RBD were selected for ΔΔG calculations upon mutation to all 20 amino 200 acids. This resulted in 48, 35, and 34, mutation sites corresponding to 80R, m396, and CR3022, 201 respectively. Figures 2D, 3D , and 4D depict the results of ΔΔG calculations performed on 80R, 202 m396, and CR3022, respectively, using the 4 computational methods discussed before. Due to the mutational structure sampling algorithms, the binding affinity scores comparing mutations 204 to wild type (e.g. H:S101S) can be nonzero. For normalization, the ∆∆G value for each mutation 205 is offset so that the wild type mutations are zero. Interestingly, predicted ΔΔG values obtained 206 from SAAMBE-3D are mainly unfavorable (positive values), and the range of predicted values 207 is smaller than other methods. Another observation is the large variation of predicted values 208 among these four methods, reflecting the need of an approach to effectively rank the mutations.

Previous studies in binding affinity predictions suggest that using a consensus approach over 210 different methods can improve prediction accuracy [35, [40] [41] [42] [43] [44] [45] . Following this rationale, we 211 applied a similar strategy to rank the single mutations from the four computational predictions 212 for each antibody. We used relative ranking instead of absolute score due to different 213 magnitudes and scales of the four methods. A Z-score describes a value's relationship to the 214 mean of a group of values, which is useful for normalization of raw scores. Here we used a Z-215 score based on the median value instead of the mean value for each scoring function, which 216 reduces the sensitivity of Z-scores to outliers. By averaging the Z-score from the four methods, 217 consensus Z-scores were computed, and the top ranked mutations were visualized to validate 218 the predictions. For each system we selected the top 60 mutations as presented in Table 1 . Pro residues were taken into consideration. As shown in Figure 2C and Table 1 , selected 222 mutations for 80R belong to positions D50, A51, S52, S67, S92 in the light chain; and N57, R100, 223 S101 in the heavy chain. Since A51 is in the vicinity of Y489, F490, and Q493, it is expected that 224 mutations to Phe, Trp, or Tyr will promote formation of π-π interactions, while mutations to 

As shown in Figure 3C and Table 1 

As shown in Figure 4C and Table 1 

Trispecific antibody design and Fc selection. SARS-CoV-2 has shown fast mutation rates 286 among discovered variants, therefore combining neutralizing antibodies with different epitopes 287 into a multi-specific format can benefit both potency and breadth, especially for future variants.

We therefore proposed to engineer the three mAbs, after affinity optimization against SARS- 

E484K is within the 80R epitope, while N501Y is within both 80R and m396 epitopes ( Figure  1B ). This emphasizes the importance of combining multiple antibodies with different epitopes, 349 especially to include antibodies with conserved epitopes, such as CR3022. Given the success 350 shown in the HIV study, our trispecific format is one of the suitable formats for 3-in-1 antibody 351 design. However, it requires careful geometry modeling and sequence optimization for further 352 developability.

In this study, we used computational protein engineering tools to optimize SARS-CoV 

Underlined residues are potential developability labile sites. 

Underlined residues are potential developability labile sites. 

Cell entry mechanisms of SARS-CoV-2

Targeted therapy strategies against SARS-CoV-2 cell entry 454 mechanisms: A systematic review of in vitro and in vivo studies

Dynamic Regulation of SARS-Cov-2 Binding and Cell Entry 456 Mechanisms in Remodeled Human Ventricular Myocardium

Structural basis of receptor recognition by SARS-CoV-2

Neutralization of SARS-CoV-2 spike pseudotyped virus by recombinant 461 ACE2-Ig

Structure, Function, and Antigenicity of the SARS-CoV-2 Spike 463 Glycoprotein

Structure of the SARS-CoV-2 spike receptor-binding domain bound to the 465 ACE2 receptor

REGN-COV2, a Neutralizing Antibody Cocktail, in Outpatients 467 with Covid-19

SARS-CoV-2 Neutralizing Antibody LY-CoV555 in Outpatients with 469 Covid-19

Escape from neutralizing antibodies by SARS-CoV-2 spike protein 471 variants. Elife, 2020. 9

Comprehensive mapping of mutations to the SARS-CoV-2 receptor-473 binding domain that affect recognition by polyclonal human serum antibodies. bioRxiv

Neutralization of SARS-CoV-2 spike 69/70 deletion, E484K and N501Y 476 variants by BNT162b2 vaccine-elicited sera

Introduction of the South African SARS-CoV-2 variant 501Y.V2 into 478 the UK

Trispecific broadly neutralizing HIV antibodies mediate potent SHIV 480 protection in macaques

Universal protection against influenza infection by a multidomain 482 antibody to influenza hemagglutinin

Structural basis of neutralization by a human anti-severe acute 484 respiratory syndrome spike protein antibody, 80R

Potent cross-reactive neutralization of SARS coronavirus isolates by 487 human monoclonal antibodies

Structure of severe acute respiratory syndrome coronavirus 489 receptor-binding domain complexed with neutralizing antibody

A highly conserved cryptic epitope in the receptor binding domains of 494 SARS-CoV-2 and SARS-CoV

Design of therapeutic proteins with enhanced stability

Medicinal chemistry and the molecular operating 498 environment (MOE): application of QSAR and molecular docking to drug discovery

Structure-based prediction of asparagine and aspartate degradation 501 sites in antibody variable regions

Prediction of protein deamidation rates from 503 primary and three-dimensional structure

Modeling the oxidation of methionine residues by peroxides in 506 proteins

Force fields for protein simulations

The generalized Born/volume integral implicit solvent model: estimation of 510 the free energy of hydration using London dispersion instead of atomic surface area

LowModeMD--implicit low-mode velocity filtering applied to conformational 513 search of macrocycles and protein loops

Scientific benchmarks for guiding macromolecular energy function 515 improvement

Flex ddG: Rosetta Ensemble-Based Estimation of Changes in 517

Protein-Protein Binding Affinity upon Mutation

A topology-based network tree for the prediction of 520 protein-protein binding affinity changes following mutation. Nature Machine 521 Intelligence

SKEMPI 2.0: an updated benchmark of changes in protein-protein 523 binding energy, kinetics and thermodynamics upon mutation

AB-Bind: Antibody binding mutational database for computational 526 affinity predictions

SAAMBE-3D: Predicting Effect of Mutations on Protein-Protein 528 Interactions

Application of Assisted Design of Antibody and Protein Therapeutics 530 (ADAPT) improves efficacy of a Clostridium difficile toxin A single-domain antibody

Potent binding of 2019 novel coronavirus spike protein by a SARS 533 coronavirus-specific human monoclonal antibody

Key residues of the receptor binding motif in the spike protein of SARS-536

CoV-2 that interact with ACE2 and neutralizing antibodies

Engineering an anti-CD52 antibody for enhanced deamidation stability

Site-specific glycan analysis of the SARS-CoV-2 spike

An extensive test of 14 scoring functions using the PDBbind refined set 543 of 800 protein-ligand complexes

Consensus scoring: A method for obtaining improved hit rates 545 from docking databases of three-dimensional structures into proteins

3D-Jury: a simple approach to improve protein structure predictions

Meta-DP: domain prediction meta-server

Assessment of Solvated Interaction Energy Function for Ranking 552

Antibody-Antigen Binding Affinities

Assisted Design of Antibody and Protein Therapeutics (ADAPT)

CODV-Ig, a universal bispecific tetravalent and multifunctional 556 immunoglobulin format for medical applications

Antibody-mediated disruption of the SARS-CoV-2 spike 558 glycoprotein

Antibody-dependent enhancement and SARS-CoV-2 vaccines and 560 therapies

The Fc-mediated effector functions of a potent SARS-CoV-2 562 neutralizing antibody, SC31, isolated from an early convalescent COVID-19 patient, are 563 essential for the optimal therapeutic efficacy of the antibody. bioRxiv

Engineered Fc-glycosylation switch to eliminate antibody effector 566 function

Antibody Fc engineering for enhanced neonatal Fc receptor 568 binding and prolonged circulation half-life

Potential Neutralizing Antibodies Discovered 570 for Novel Corona Virus Using Machine Learning. bioRxiv

Rapid <em>in silico</em> design of antibodies targeting SARS-572

CoV-2 using machine learning and supercomputing