key: cord-0752708-kxsab0gk
authors: Heffron, Anna S.; McIlwain, Sean J.; Baker, David A.; Amjadi, Maya F.; Khullar, Saniya; Sethi, Ajay K.; Shelef, Miriam A.; O’Connor, David H.; Ong, Irene M.
title: The landscape of antibody binding to SARS-CoV-2
date: 2020-10-11
journal: bioRxiv
DOI: 10.1101/2020.10.10.334292
sha: 5121ae49785f716031123131ed269af5cfba05b4
doc_id: 752708
cord_uid: kxsab0gk

The search for potential antibody-based diagnostics, vaccines, and therapeutics for pandemic severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has focused almost exclusively on the spike (S) and nucleocapsid (N) proteins1–8. Coronavirus membrane (M), orf3a, and orf8 proteins are also humoral immunogens in other coronaviruses (CoVs)8–11 but remain largely uninvestigated for SARS-CoV-2. Here we show that SARS-CoV-2 infection induces robust antibody responses to epitopes throughout the SARS-CoV-2 proteome, particularly in M, in which one epitope achieved near-perfect diagnostic accuracy. We map 79 B cell epitopes throughout the SARS-CoV-2 proteome and demonstrate that anti-SARS-CoV-2 antibodies appear to bind homologous peptide sequences in the 6 known human CoVs. Our results demonstrate previously unknown, highly reactive B cell epitopes throughout the full proteome of SARS-CoV-2 and other CoV proteins, especially M, which should be considered in diagnostic, vaccine, and therapeutic development.

CoV-2 proteins, though data from other CoVs suggest they may be important. Antibodies against 48 SARS-CoV M can be more potent than antibodies against SARS-CoV S 9-11 , and some 49 experimental SARS-CoV and MERS-CoV vaccines elicit responses to M, E, and orf8 8 . 50

Additionally, previous work has demonstrated humoral cross-reactivity between CoVs 7,14,27-30 51

and suggested it could be protective 24,30 , although full-proteome cross-reactivity has not been 52

investigated. We designed a peptide microarray tiling the proteome of SARS-CoV-2 and 8 other 53

human and animal CoVs in order to assess antibody epitope specificity and potential cross-54 reactivity with other CoVs, and we used this microarray to profile IgG antibody responses in 40 55

COVID-19 convalescent patients and 20 SARS-CoV-2-naive controls. 56 57

CoV reactivity in uninfected controls 58

Greater than 90% of adult humans are seropositive for the "common cold" CoVs (CCCoVs: 59

HCoV-HKU1, HCoV-OC43, HCoV-NL63, and HCoV-229E) 31,32 , but it is unknown how these 60 pre-existing antibodies might affect reactivity to SARS-CoV-2 or other CoVs. We measured IgG 61 reactivity in sera from 20 SARS-CoV-2-naïve control subjects to CoV linear peptides, 62

considering reactivity that was >3 standard deviations above the mean for the log2-quantile 63

normalized array data to be indicative of antibody binding. All sera exhibited binding in known 64 epitopes of at least 1 of the control non-CoV strains (poliovirus vaccine and rhinovirus; Fig. 1 , 65

Extended data 1, Extended data 2) and were collected in Wisconsin, USA, where exposure to 66 SARS-CoV or MERS-CoV was extremely unlikely. We found that at least one epitope in 67 structural or accessory proteins had binding in 100% of controls for HCoV-HKU1, 85% of 68 controls for HCoV-OC43, 65% for HCoV-NL63, and 55% for HCoV-229E (Fig. 2 , Extended 69 data 2). Apparent cross-reactive binding was observed in 45% of controls for MERS-CoV, 50% 70

for SARS-CoV, and 50% for SARS-CoV-2. 71 72 SARS-CoV-2 proteome humoral profiling 73 We aimed to map the full extent of binding of antibodies induced by SARS-CoV-2 infection and 74

to rank the identified epitopes in terms of likelihood of importance and immunodominance. We 75 defined epitope recognition as antibody binding to contiguous peptides in which the average 76 log2-normalized intensity for patients was at least 2-fold greater than for controls with t-test 77 statistics yielding adjusted p-values <0.1. We chose these criteria, rather than the 3 standard 78 deviation cut-off, in order to ensure that binding detected would be greater than background 79 binding seen in controls (2-fold greater) and to remove regions of binding that were not at least 80 weakly significantly different from controls (adjusted p<0.1). These criteria identified 79 B cell 81 epitopes (Fig. 3 , Extended data 3) in S, M, N, orf1ab, orf3a, orf6, and orf8. We ranked these 82 epitopes by minimum adjusted p-value for any 16-mer in the epitope in order to determine the 83 greatest likelihood of difference from controls as a proxy for immunodominance. The highest-84 ranking epitope occurred in the N-terminus of M (1-M-24). Patient sera showed high-magnitude 85 reactivity (up to 6.7 fluorescence intensity units) in other epitopes in S, M, N, and orf3a, with 86 lower-magnitude reactivity (<3.3 fluorescence intensity units) epitopes in other proteins. The 87

epitopes with the greatest reactivity in S occurred in the fusion peptide (residues 788-806), with 88 less reactivity in the receptor binding domain (residues 319-541) 6 ( Fig. 3) . The highest specificity (100%) and sensitivity (98%), determined by linear discriminant analysis 96

leave-one-out cross-validation, for any individual peptide was observed for a 16-mer within the 97 1-M-24 epitope: ITVEELKKLLEQWNLV (Extended data 4). Fifteen additional individual 98 peptides in M, S, and N had 100% measured specificity and at least 80% sensitivity. 99

Combinations of 1-M-24 with 1 of 5 other epitopes (384-N-33, 807-S-26, 6057-orf1ab-17, 227-100

N-17, 4451-orf1b-16) yielded an area under the curve receiver operating characteristic of 1 101

(Extended data 5) based on linear discriminant analysis leave-one-out-cross-validation.

Human, animal CoV cross-reactivity 104

We defined cross-reactivity as binding by antibodies in COVID-19 convalescent sera to non-105 SARS-CoV-2 peptides at an average log2-normalized intensity at least 2-fold greater than in 106

controls with t-test statistics yielding adjusted p-values <0.1. Antibodies in COVID-19-107

convalsecent sera appeared to be cross-reactive with homologous epitopes in S, M, N, orf1ab, 108 orf3, orf6, and orf8 in other CoVs (Fig. 4 , Extended data 6, Extended data 7, Extended data 8).

The greatest number of cross-reactive epitopes (70) were in the RaTG13 bat betacoronavirus ( -110 CoV), the closest known relative of SARS-CoV-2 (96% nucleotide identity) 39,40 , then the 111 pangolin CoV (51 epitopes, 85% nucleotide identity with SARS-CoV-2) 41 , then SARS-CoV (40 112 epitopes, 78% identity 39 ). One region, corresponding to SARS-CoV-2 epitope 807-S-26, was 113

cross-reactive across all CoVs, and one, corresponding to SARS-CoV-2 epitope 1140-S-25, was 114

cross-reactive across all -CoVs (Fig. 4) . other CoVs to be a small, glycosylated ectodomain that protrudes outside the virion and interacts 121 with S, N, and E 20 , while the rest of M resides within the viral particle. Full-length SARS-CoV 122 M has been shown to induce protective antibodies 11,43 , and patterns of antibodies binding to 123 SARS-CoV M are similar to those we found in SARS-CoV-2 36 . SARS-CoV anti-M antibodies 124

can synergize with anti-S and anti-N antibodies 11,43 , and M has been used in protective SARS-125

CoV and MERS-CoV vaccines 8 . However, the mechanism of protection of anti-M antibodies 126 remains unknown, and this protein remains largely understudied and underutilized as an antigen. 127

Other groups have not previously identified the high magnitude binding we observed in M, 128

though that may be due to using earlier sample timepoints or different techniques or 129 algorithms 44,45 . Our results, in concert with prior knowledge of anti-SARS-CoV antibodies, 130 strongly suggest that M, particularly the 1-M-24 epitope, as well as other novel epitopes that we 131

identified should be investigated further as potential targets in SARS-CoV-2 diagnostics, 132

vaccines, and therapeutics. 133 134

We also found that antibodies produced in response to SARS-CoV-2 infection appeared to cross-135 react with homologous epitopes throughout the proteomes of other human and non-human CoVs.

Hundreds of CoVs have been discovered in bats and other species 24,39-41,46,47 , making future 137 spillovers inevitable. The broad cross-reactivity we observed in some homologous peptide 138 sequences may help guide the development of pan-CoV vaccines 18 , especially given that 139 antibodies binding to 807-S-26 and 1140-S-25, epitope motifs cross-reactive across all CoVs and 140

all -CoVs, respectively, are known to be potently neutralizing 33,34 . We cannot determine 141

whether the increased IgG binding to CCCoVs in COVID-19 convalescent sera is due to newly 142 developed cross-reactive antibodies or the stimulation of a memory response against the original 143

CCCoV antigens. However, cross-reactivity of anti-SARS-CoV-2 antibodies with SARS-CoV or 144

MERS-CoV is likely real, since our population was very unlikely to have been exposed to those 145 viruses. A more stringent assessment of cross-reactivity as well as functional investigations into 146 these cross-reactive antibodies will be vital in determining their capacity for cross-protection. 147

Further, our methods efficiently detect antibody binding to linear epitopes 48 , but their sensitivity 148

for detecting parts of conformational epitopes is unknown, and additional analyses will be 149 required to determine whether epitopes identified induce neutralizing or otherwise protective 150

antibodies.

Many questions remain regarding the biology and immunology of SARS-CoV-2. Our extensive 153

profiling of epitope-level resolution antibody reactivity in COVID-19 convalescent subjects 154

provides new epitopes that could serve as important targets in the development of improved 155 diagnostics, vaccines, and therapeutics against SARS-CoV-2 and dangerous human CoVs that 156 may emerge in the future. 157 158 260

The study was conducted in accordance with the Declaration of Helsinki and approved by the We included sequences on the array of viruses which we expected all adult humans to be likely 303

to have been exposed to as positive controls: one poliovirus strain (measuring vaccine exposure), 304

and seven rhinovirus strains. Any subject whose sera did not react to at least one positive control 305 would be considered a failed run and removed from the analysis. All subjects in this analysis 306 reacted to epitopes in at least one control strain (Fig. 1 , Extended data 1, Extended data 2). 307 308

Peptide microarray data analysis and data availability 309

The raw fluorescence signal intensity values were log2 transformed. Clusters of fluorescence 310

intensity of statistically unlikely magnitude, indicating array defects, were identified and 311

removed. Local and large area spatial corrections were applied, and the median transformed 312

intensity of the peptide replicates was determined. The resulting median data was cross-313 normalized using quantile normalization. All peptide microarray datasets and code used in these 314

analyses can be downloaded from https://github.com/Ong-Research/Ong_UW_Adult_Covid-315

19.git.

Statistical analysis 318

Statistical analyses were performed in R (v 4.0.2) using in-house scripts. For each peptide, a p-319

value from a two-sided t-test with unequal variance between sets of patient and control 320 responses were calculated and adjusted using the Benjamini-Hochberg (BH) algorithm. To 321 determine whether the peptide was in an epitope (in SARS-CoV-2 proteins) or cross-reactive for 322

anti-SARS-CoV-2 antibodies (in non-SARS-CoV-2 proteins), we used an adjusted p-value cutoff 323 of <0.1 (based on multiple hypothesis testing correction for all 119,487 unique sequences on the 324 array) and a fold-change of greater than or equal to 2 and grouped consecutive peptides as a 325

represented epitope. Linear discriminant analysis leave-one-out cross validation was used to 326 determine specificity and sensitivity on each peptide and from each epitope using the average 327 signal of the component peptides.

To identify cross reactive epitopes, we used each SARS-CoV-2 epitope sequence as a query, 330

searched the database of proteins from the sequences in the peptide array using blastp (-word-331 size 2, num-targets 4000) to find homologous sequences in the bat, pangolin, and other human 332

CoV strains, then determined whether the average log2-normalized intensity for these sequences 333

in patients was at least 2-fold greater than in controls with t-test statistics yielding adjusted p-334 values <0.1. Each blast hit was then mapped back to the corresponding probe ranges.

The clinical and demographic characteristics of convalescent subjects were compared to those of 337 the controls using 2 tests for categorical variables and Wilcoxon rank-sum tests for non-338

normally distributed continuous measures.

Heatmaps were created using the gridtext 50 and complexheatmap 51 packages in R. Alignments 341

for heatmaps were created using MUSCLE 52 . 342 343 344

Evaluation of Nucleocapsid and Spike Protein-Based Enzyme-Linked 389

Immunosorbent Assays for Detecting Antibodies against SARS-CoV-2

Analytical and clinical validation of an ELISA for specific SARS-392

CoV-2 IgG, IgA, and IgM antibodies

Diagnostic accuracy of serological tests for covid-19: systematic 394 review and meta-analysis

Multiplex detection and dynamics of IgG antibodies to SARS-CoV2 and 396 the highly pathogenic human coronaviruses SARS-CoV and MERS-CoV

Structural and functional properties 399 of SARS-CoV-2 spike protein: potential antivirus drug development for COVID-19

Promise and challenges in the development of COVID-19 vaccines

Using Reverse Vaccinology and Machine Learning

Specific epitopes of the structural and hypothetical proteins elicit 406 variable humoral responses in SARS patients

Identification of immunodominant 408 epitopes on the membrane protein of the severe acute respiratory syndrome-associated 409 coronavirus

Protective humoral responses to severe acute respiratory syndrome-411 associated coronavirus: implications for the design of an effective protein-based vaccine

Potent neutralization of severe acute respiratory syndrome (SARS) coronavirus 414 by a human mAb to S1 protein that blocks receptor association

Human monoclonal antibody as prophylaxis for SARS coronavirus 417 infection in ferrets

Structural basis for potent cross-neutralizing human monoclonal antibody 419 protection against lethal human and zoonotic severe acute respiratory syndrome 420 coronavirus challenge

A SARS DNA vaccine induces neutralizing antibody and cellular 422 immune responses in healthy adults in a Phase I clinical trial

Potent neutralization of MERS-CoV by human neutralizing monoclonal 425 antibodies to the viral spike glycoprotein

Human Neutralizing Monoclonal Antibody Inhibition of Middle East 427 Respiratory Syndrome Coronavirus Replication in the Common Marmoset

Rational Vaccine Design in the Time of COVID-19

Prospects for a MERS-CoV spike vaccine

Coronaviruses : methods and protocols

Anti-spike IgG causes severe acute lung injury by skewing macrophage 436 responses during acute SARS-CoV infection

Antibody-dependent 438 enhancement and SARS-CoV-2 vaccines and therapies

Antibody-dependent 440 enhancement occurs upon re-infection with the identical serotype virus in feline infectious 441 peritonitis virus infection

Emerging Pandemic Diseases: How We Got to COVID-19

Immunization with SARS coronavirus vaccines leads to pulmonary 445 immunopathology on challenge with the SARS virus

SARS CoV subunit vaccine: antibody-mediated neutralisation and 447 enhancement

Potent binding of 2019 novel coronavirus spike protein by a SARS 449 coronavirus-specific human monoclonal antibody

Cross-reactive Antibody Response between SARS-CoV-2 and SARS-CoV 452 Infections

Cross-neutralization of SARS-CoV-2 by a human monoclonal SARS-CoV 454 antibody

Epidemiology of Seasonal Coronaviruses: Establishing the Context 456 for the Emergence of Coronavirus Disease

The receptor binding domain of the viral spike protein is an 461 immunodominant and highly specific target of antibodies in SARS-CoV-2 patients

Two linear epitopes on the SARS-CoV-2 spike protein that elicit 464 neutralising antibodies in COVID-19 patients

Mining of epitopes on spike protein of SARS-CoV-2 from COVID-19 466 patients

Linear epitopes of SARS-CoV-2 spike protein elicit neutralizing antibodies in 468 COVID-19 patients

A Sequence Homology and Bioinformatic Approach Can Predict 470

Candidate Targets for Immune Responses to SARS-CoV-2

Preliminary Identification of Potential 473 Vaccine Targets for the COVID-19 Coronavirus (SARS-CoV-2) Based

Immunoinformatic 476 identification of B cell and T cell epitopes in the SARS-CoV-2 proteome

A pneumonia outbreak associated with a new coronavirus of probable bat 479 origin

Coexistence of multiple coronaviruses in several bat colonies in an 481 abandoned mineshaft

Isolation of SARS-CoV-2-related coronavirus from Malayan pangolins. 483

Coronavirus membrane 485 fusion mechanism offers a potential target for antiviral development

The expression of membrane protein augments the specific responses 488 induced by SARS-CoV nucleocapsid DNA immunization

Viral epitope profiling of COVID-19 patients reveals cross-reactivity and 491 correlates of severity

Immunoreactive peptide maps of SARS-CoV-2 and other human 493 coronaviruses

Coronavirus diversity, phylogeny and 496 interspecies jumping

Global patterns in coronavirus diversity

Antibody responses to Zika virus proteins in pregnant and non-499 pregnant macaques

Reduced IgG titers against pertussis in rheumatoid arthritis: Evidence 501 for a citrulline-biased immune response and medication effects

Improved Text Rendering Support for 'Grid' Graphics

Complex heatmaps reveal patterns and correlations in 506 multidimensional genomic data

MUSCLE: multiple sequence alignment with high accuracy and high 508 throughput