key: cord-0975667-3mpnwwu3 authors: Tang, Juanjie; Lee, Youri; Ravichandran, Supriya; Grubbs, Gabrielle; Huang, Chang; Stauft, Charles B.; Wang, Tony; Golding, Basil; Golding, Hana; Khurana, Surender title: Epitope diversity of SARS-CoV-2 hyperimmune intravenous human immunoglobulins and neutralization of variants of concern date: 2021-08-20 journal: iScience DOI: 10.1016/j.isci.2021.103006 sha: fb050c9c919fbb7c049bcbab0c53c4e0825b034d doc_id: 975667 cord_uid: 3mpnwwu3 Hyperimmune immunoglobulin (hCoV-2IG) generated from SARS-CoV-2 convalescent plasma (CP) are under evaluation in clinical trials. Here we explored the antibody epitope repertoire, and virus neutralizing capacity of six hCoV-2IG batches as well as nine convalescent plasma (CP) against SARS-CoV-2 and emerging variants of concern (VOC). Epitope-mapping by Gene-Fragment-Phage display library (GFPDL) spanning the SARS-CoV-2 spike demonstrated broad recognition of multiple antigenic sites spanning the entire spike, that was higher for hCoV-2IG than CP, with predominant binding to the fusion peptide. In the pseudovirus neutralization assay (PsVNA) and in the wild-type SARS-CoV-2 PRNT assay, hCoV-2IG lots showed higher titers against the WA-1 strain compared with CP. Neutralization of VOCs were reduced to different extent by hCoV-2IG lots but were higher than CP. Significant reduction of hCoV-2IG binding was observed to RBD-E484K followed by RBD-N501Y (but not RBD-K417N). This study suggests that post-exposure treatment with hCoV-2IG could be preferable to CP. In the current study we probed the antibody quality of 6 hCoV-2IG products. In addition 82 The spike protein is the antigen of choice for development of vaccines and therapeutics 83 against SARS-CoV-2. To decipher the epitope-specificity of the SARS-CoV-2 spike-specific 84 antibodies in an unbiased manner, we subjected the six hCoV-2IG lots to antibody epitope 85 profiling with a highly diverse SARS-CoV-2 spike GFPDL with >10 7.1 unique phage clones 86 displaying epitopes of 18-500 amino acid residues across the SARS-CoV-2 spike. In preliminary 87 studies to characterize the SARS-CoV-2 spike GFPDL, epitope-mapping of monoclonal 88 antibodies (MAbs) targeting SARS-CoV-2 spike or RBD identified the expected linear or 89 conformation-dependent epitopes recognized by these MAbs. Recently, we showed that SARS- The spike proteins mutations in the VOCs used for production of the pseudovirions are shown in 121 Table S2 . All sixteen pre-pandemic 2019-IVIG preparations demonstrated titers of <20 PsVNA50 123 against SARS-CoV-2 strains ( Fig. 2A and Table S3 ). Among the nine CP lots tested against WA-124 1, variable PsVNA50 titers were observed, including one negative, one low (<1:80), six medium 125 (>1:160<1:640) and two high (>1:640). In contrast, all six hCoV-2IG lots exhibited high 126 PsVNA50 titers against WA-1 ranging between 1:1238-1:3309. PsVNA80 titers for hCoV-2IG 127 ranged between 1:168-1:593, but none of the CP lots showed PsVNA80 titers above 1:80 (range 128 <20 to 1:74) against WA-1 (Table S3) . Neutralization of the VOCs showed variable loss of titers 129 as determined by either PsVNA50 or PsVNA80 for the hCoV-2IG and the CPs with the greatest 130 reduction in titers measured against the SA VOC ( Fig. 2A and Table S3 ). For confirmation of the PsVNA neutralization titers, the six hCoV-2IG lots were also 132 evaluated in a classical PRNT assay using VERO-E6 cells against authentic SARS-CoV-2 viruses (Table S5 and Fig. 2D ). Compared to the WA-1 strain, the average PsVNA50 of the hCoV-2IG against CA, UK, 141 JP and SA VOC were reduced by 1.7, 1.9, 3.5, and 9.2-fold respectively (Fig. 2E ). Since the 142 amount of SARS-CoV-2 specific IgG in CP lots is more variable and 5-10 fold lower compared 143 with the hyperimmune hCoV-2IG, the CPs exhibited greater loss of neutralizing activities against 144 the variants in comparison with hCoV-2IG. The average PsVNA50 of the CP against the CA, UK, 145 JP, and SA VOC were reduced by 3.1, 3.3, 3.9, and 18.7-fold, respectively (Table S3 and Fig. 2F ). PsVNA80 titers against the UK and JP VOC for all 9 CPs were lower (~2-fold) and were minimal 147 or negligible against the SA VOC ( Fig. 2A and Table S3 ). For hCoV-2IG, the PsVNA80 titers (Fig. S2) , but a few key mutations among these strains are shared by VOCs as shown in Table S3 . N501Y is shared among the UK, JP, and SA variants. E484K is shared between the JP and SA and E484K) were analyzed in SPR based antibody binding assays (Fig. 3A) . The K417N had 163 minimal to no impact on hCoV-2IG binding. The hCoV-2IG binding to RBD-N501Y was reduced 164 by ~2-fold compared with WA-1 RBD. However, binding to RBD-E484K resulted in an average 165 19-fold reduction in hCoV-2IG binding compared with the WA-1 RBD (Fig. 3B) . In the current study we evaluated six lots of hCoV-2IG that were manufactured from CP collected Table S3 : Neutralization titers of convalescent plasma, IVIG and hCoV-2IG against SARS-CoV- The objective of this study was to investigate various therapeutic polyclonal CP or purified hCoV-536 2-IG antibody preparations, which are being evaluated in clinical trials, for antibody binding and 537 neutralizing capacity of important emerging SARS-CoV-2 variant of concern (VOC). Such were collected in 2020 prior to emergence of variants of concern in the U.S. Nine random CP lots were obtained from recovered COVID-19 patients between May-July 552 2020, at least 30-days post-recovery from U.S. plasma donors. They were not selected based on 553 any a priory criterion in order to represent a broad spectrum of antibody titers as reported in the Table S1 ). Briefly, human codon-optimized cDNA encoding SARS-CoV-2 S glycoprotein of the WA- Extremely potent human monoclonal antibodies from COVID-19 315 convalescent patients Immunodominant antibody germlines in COVID-19 After 1 h the mixture 596 was removed and replenished with fresh MEM containing 2% FBS. Cells were incubated at 37°C 597 for an additional 48 hours, then fixed with 4% paraformaldehyde, followed by staining of cells 598 with 0.1% crystal violet in 20% methanol. The PRNT50 and PRNT90 titers were calculated as the 599 last serum dilution resulting Wild-type SARS-CoV-2 CA and JP strains were not available to test in the BSL3 PRNT assay SARS-CoV-2 Gene Fragment Phage Display Library (GFPDL) construction GenBank: 604 MN908947.3) was chemically synthesized and used for cloning. A gIII display-based phage 605 vector, fSK-9-3, was used where the desired polypeptide can be displayed on the surface of the 606 phage as a gIII-fusion protein. Purified DNA containing spike gene was digested with DNaseI to 607 obtain gene fragments of 50-1500 bp size range (18 to 500 amino acids) and used for GFPDL 608 construction Affinity selection of SARS-CoV-2 GFPDL phages 611 Prior to panning of GFPDL with polyclonal hCoV-2IG antibodies, Ig components, which 612 could non-specifically interact with phage proteins, were removed by incubation with UV-killed 613 M13K07 phage-coated Petri dishes Briefly, the hCoV-2IG lot was incubated with the GFPDL and the protein A/G resin, the unbound 616 phages were removed by PBST (PBS containing 0.1 % Tween-20) wash followed by PBS. Bound 617 phages were eluted by addition of 0.1 N Gly-HCl pH 2.2 and neutralized by adding 8 µL of 2 M 618 After panning, antibody-bound phage clones were amplified, the 619 inserts were sequenced, and the sequences were aligned to the SARS-CoV-2 spike gene The GFPDL affinity selection was performed in duplicate (two independent experiments 622 by a research fellow in the lab, who was blinded to sample identity) Recombinant purified RBD proteins 629 used in the study were produced in HEK-293 mammalian cells. The native receptor-binding 630 activity of the spike RBD proteins was determined by binding to the 5 µg/mL human ACE2 protein 631 ACE2 (AC2-H82E6) contains amino acid residues Gln 18 -Ser 740 expressed from HEK 293 cells 633 was purchased from Acro Biosystems. using F-moc chemistry, biotinylated at C-terminus, and purified using RP Antibody binding kinetics to SARS-CoV-2 RBD mutants or SARS-CoV-2 peptides by 639 Surface Plasmon Resonance (SPR) The purified recombinant SARS-CoV-2 RBD proteins were 642 captured to a Ni-NTA sensor chip with 200 resonance units (RU) in the test flow channels. The 643 protein density on the chip was optimized so as to measure monovalent interactions independent 644 of the antibody isotype. The biotinylated SARS-CoV-2 peptides were captured on a NLC chip and 645 used for peptide antibody profiling of hCoV-2IG Serial dilutions (1 mg/mL, 0.33 mg/mL and 0.11 mg/mL) of freshly prepared hCoV-2IG 647 or 2019-IVIG or 10-fold dilution of CP in BSA-PBST buffer BSA) were injected at a flow rate of 50 µL/min (120 sec contact duration) for association, and 649 disassociation was performed over a 600-second interval. Responses from the protein surface were 650 corrected for the response from a mock surface and for responses from a buffer-only injection All SPR experiments were 653 performed twice, and the researchers performing the assay were blinded to sample identity. The 654 maximum resonance units (Max RU) data shown in the figures are the calculated RU signals for 655 the 1 mg/mL hCoV CA) or R package. Differences between groups were analyzed using 660 multiple group comparisons by non-parametric (Kruskal-Wallis) statistical test using Dunn's post-661 hoc analysis. The difference within each group was performed using one-way ANOVA using 662 Tukey's pairwise multiple comparison test. The differences were considered statistically 663 significant with a 95% confidence interval when the p value was less than 0.05. (*, P values of 664 Correlation analysis of 665 PRNT and PsVNA titers were performed by computing Pearson's correlation coefficient in SARS-CoV-2 hCoV-2IG demonstrate highly diverse antibody epitope profile. 2. SARS-CoV-2 hCoV-2IG lots neutralized SARS-CoV-2 variants better than CP. 3. Significant reduction of hCoV-2IG binding to RBD-E484K compared to unmutated RBD Higher hCoV-2IG dose would be required for SARS-CoV-2 variant infected patients 669 We thank Keith Peden and Marina Zaitseva for their insightful review of the manuscript. We thank 670 Carol Weiss for providing plasmid clones expressing SARS-CoV-2 variants and Dorothy Scott for