Macaque-human differences in SARS-CoV-2 Spike antibody response elicited by vaccination or infection

Macaques are a commonly used model for studying immunity to human viruses, including for studies of SARS-CoV-2 infection and vaccination. However, it is unknown whether macaque antibody responses recapitulate, and thus appropriately model, the response in humans. To answer this question, we employed a phage-based deep mutational scanning approach (Phage-DMS) to compare which linear epitopes are targeted on the SARS-CoV-2 Spike protein in humans and macaques following either vaccination or infection. We also used Phage-DMS to determine antibody escape pathways within each epitope, enabling a granular comparison of antibody binding specificities at the locus level. Overall, we identified some common epitope targets in both macaques and humans, including in the fusion peptide (FP) and stem helix-heptad repeat 2 (SH-H) regions. Differences between groups included a response to epitopes in the N-terminal domain (NTD) and C-terminal domain (CTD) in vaccinated humans but not vaccinated macaques, as well as recognition of a CTD epitope and epitopes flanking the FP in convalescent macaques but not convalescent humans. There was also considerable variability in the escape pathways among individuals within each group. Sera from convalescent macaques showed the least variability in escape overall and converged on a common response with vaccinated humans in the SH-H epitope region, suggesting highly similar antibodies were elicited. Collectively, these findings suggest that the antibody response to SARS-CoV-2 in macaques shares many features with humans, but with substantial differences in the recognition of certain epitopes and considerable individual variability in antibody escape profiles, suggesting a diverse repertoire of antibodies that can respond to major epitopes in both humans and macaques.

2 23 11 Center for Innate Immunity and Immune Disease, University of Washington, Seattle, WA, USA 24 25 * Corresponding author 26 Email: joverbau@fredhutch.org 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 130 a Within each group of macaques, subgroups received slightly different treatments (described in Table S1).
131 Enrichment of wildtype peptides 132 To compare which regions of Spike protein are recognized by human and macaque antibodies, 133 we examined the enrichment of wildtype peptides by antibodies from each individual ( Fig 1A). (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted December 3, 2021. ; https://doi.org/10.1101/2021.12.01.470697 doi: bioRxiv preprint 9 142 visualized on the structure of a Spike protein monomer in Fig 1B. In addition to these defined 143 regions, we noted that one convalescent rhesus macaque appeared to weakly recognize an 144 epitope at the beginning of the S2 subunit (amino acid 686-710, Fig 1A). 145 In general, we did not detect responses in the RBD because many epitopes in this region are 146 known to be conformational, and Phage-DMS only has the power to detect epitopes that 147 include linear sequences. Epitopes in the RBD have been extensively detailed elsewhere [62, 148 63]. However, we did detect strong binding to an RBD epitope in some vaccinated pigtail 149 macaques (Fig 1A). This same region was enriched in samples from before vaccination in four of 158 convalescent humans (Fig 2).
162 Meanwhile, convalescent macaques recognized the following epitope regions more than 163 convalescent humans: CTD-N' (p ≤ 0.01), pre-FP (p ≤ 0.001), and post-FP (p ≤ 0.01) (Fig 2B). All 164 groups consistently recognized the SH-H epitope region (Fig 2). While vaccination appeared to 165 induce a stronger response against HR2 than infection (Fig 1A), there were no significant 166 differences in response driven by species (Fig 2). Within each group of macaques (vaccinated 167 and convalescent), subgroups received slightly different treatments (Table S1) 170 Taken together, these findings indicate: 1) vaccinated humans were the only group to 171 consistently recognize peptides from both the NTD and CTD-C' epitope regions, which are in 172 close physical proximity to one another ( Fig 1B); 2) convalescent humans had a limited 173 response to the CTD-N'; 3) compared to other groups, convalescent macaques had a notably 174 more robust response to regions upstream and downstream of the main FP epitope region; 4) 175 vaccinated macaques did not recognize the FP as strongly as other groups; and 5) vaccination 176 seemed to induce a stronger response against HR2 than infection in both macaques and 177 humans.
178 Defining and comparing escape pathways 179 To assess differences in the binding characteristics of human and macaque antibodies on a 180 more granular level, we next examined the mutations in Spike that reduced antibody binding in 181 each epitope region of interest. Because the antibody escape pathways for vaccinated humans 182 have been described previously [60], we did not examine the NTD and CTD-C', which are 183 exclusively recognized by this group. Instead, we focused on comparing escape profiles 184 between groups in the following epitope regions: CTD-N', FP, and SH-H. We first represent the 185 data as scaled differential selection values in logo plot form, as commonly shown in previous 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted December 3, 2021. ; https://doi.org/10.1101/2021.12.01.470697 doi: bioRxiv preprint 12 207 To summarize the trends observed in the individual findings, we calculated summed differential 208 selection values for each individual at each site and generated boxplots by group (Fig 3A). In 209 addition to the aforementioned regions of escape common to all groups, convalescent 210 macaques also showed considerable escape between AA 529-535, with vaccinated macaques 211 also showing a less consistent response in this area (Figs. 3A and S5). The complexity and 212 variability of the escape pathways also prompted us to quantify the similarity in escape 213 between and within groups. Escape similarity scores largely corresponded to areas of high 214 magnitude of escape. Sites with low-magnitude summed differential selection values indicate 215 loci where mutations have no notable impact, and therefore those escape profiles reflect 216 fluctuations in peptide enrichments due to noise, which drives a lower escape similarity score 217 at those sites ( Fig 3A, lower panel). At some sites (e.g., 560, as described above), low scores 218 were also the result of some samples showing negative differential selection and others 219 showing positive differential selection, a comparison that was assigned the highest cost in our 220 escape similarity score algorithm. 221 To test the similarity of escape profiles across the CTD-N' epitope region, escape similarity 222 scores were aggregated across the region and computed both within and between groups.
223 These are shown as boxplots, with each point representing a pairwise comparison between 224 individual samples ( Fig 3B). For example, every vaccinated macaque was compared to every 225 other vaccinated macaque (a within-group comparison) and to every vaccinated human (a 226 between-group comparison). We included a comparison of convalescent macaques and 227 vaccinated humans, given visual similarities between their patterns of escape ( Fig 3A).
228 Convalescent macaques showed the highest within-group similarity in escape profiles, meaning 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted December 3, 2021. ; https://doi.org/10.1101/2021.12.01.470697 doi: bioRxiv preprint 13 229 their escape profiles were more consistent than those of the vaccinated macaques or 230 vaccinated humans ( Fig 3B). Between-group escape similarity scores were on par with the 231 within-group scores for the vaccinated macaques and humans, indicating that although there 232 was substantial variability in individual profiles, this was not driven by sample groups. (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted December 3, 2021. ; https://doi.org/10.1101/2021.12.01.470697 doi: bioRxiv preprint 14 251 escape similarity scores for vaccinated humans and convalescent macaques were also higher in 252 the SH-H than in the CTD-N' or FP, confirming a more concordant response. The median 253 between-group escape similarity score for vaccinated humans and convalescent macaques was 254 on par with their median within-group scores, indicating that the escape profile of a vaccinated 255 human looks as similar to that of a convalescent macaque as it does to another vaccinated 256 human (Fig 5B). The similarity between these two groups was higher than the similarity 257 between convalescent macaques and humans, as well as between vaccinated macaques and 258 humans (Fig 5B). Despite this overall trend, two vaccinated humans had more unique escape  285 Discussion 286 In this study, we aimed to assess whether the antibody binding specificities to SARS-CoV-2 287 Spike in macaques are a useful model for the human response. Our results indicate important 288 similarities between macaques and humans; for example, both have antibodies that recognize 289 major epitopes in the CTD, FP, and SH-H. However, many differences are also apparent, with 290 some groups showing responses to unique epitopes, such as two physically proximal epitopes in 291 the NTD and CTD that are recognized by antibodies from vaccinated humans but not macaques.
292 Additionally, epitope regions flanking the FP were recognized by antibodies from convalescent 293 macaques, while antibodies from convalescent humans did not recognize the flanking regions 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted December 3, 2021. ; https://doi.org/10.1101/2021.12.01.470697 doi: bioRxiv preprint 16 294 but showed a strong response within the FP itself. We found considerable diversity in the 295 pathways of escape between individuals, and this was not specific to either macaques or 296 humans, suggesting a diverse repertoire of antibodies that can respond to the major epitopes in 297 both groups. Overall, these results suggest that macaques and humans share recognition of 298 certain major epitopes. The differences that exist could be due to species (macaque vs. human), 299 but could also be influenced by differences in the specific type and number of exposures to 300 antigen in each group. 353 While our focus was on understanding how macaques and humans respond to a similar 354 exposure (i.e., vaccination or infection), we also noted similarities in response between re-355 infected macaques and vaccinated humans. These groups both exhibited the broadest 356 recognition across Spike, although the epitope regions they targeted were somewhat different. 357 As described above, these groups also had highly similar antibody escape profiles in the SH-H. 358 The vaccinated humans and re-infected macaques both received two exposures to high doses 359 of antigen. It is plausible that re-exposure directed initially diverse antibodies to converge on a 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 365 Additionally, epitopes that may normally be glycosylated are exposed for antibody binding in 366 Phage-DMS. There also are known germline-encoded differences in the properties of 367 immunoglobulin subclasses and Fc receptors between macaques and humans, leading to 368 differences in antibody function that cannot be assayed using Phage-DMS [90]. Additionally, 369 our sample set includes variables that limit our ability to draw conclusions about species-370 specific (macaque vs. human) differences in antibody response. The vaccinated macaques and 371 humans both received RNA vaccines encoding full-length Spike protein, but there were 372 differences in vaccine technology, including: 1) the use of mRNA in the human vaccine vs. 373 repRNA in the macaque vaccine, 2) the stabilization of Spike in its pre-fusion state in the human 374 vaccine, 3) the dosage and number of doses delivered, and 4) the formulation used to deliver 375 the RNA. Despite these differences, we found commonalities in some of the epitopes targeted 376 by antibodies from both groups. Additionally, the convalescent rhesus macaques were 377 experimentally infected twice with high titers of virus, compared to the convalescent humans 378 who were naturally infected once. This important discrepancy could be the reason why the 379 response in re-infected macaques aligned more closely with vaccinated humans than 380 convalescent humans. Studies of re-infected humans would help address this possibility. 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted December 3, 2021. ; https://doi.org/10.1101/2021.12.01.470697 doi: bioRxiv preprint 20 381 Our findings suggest that while vaccinated and convalescent macaques and humans share 382 recognition of some major epitopes, each group has a unique antibody binding profile.
383 Antibody escape profiles suggest a diversity of individual responses to most epitopes.
384 Important avenues for future study include comparing macaque and human responses to the 385 RBD and evaluating species differences in antibody function. Continued investigation of 386 immunogenic epitopes in conserved regions of Spike is also warranted to inform the 387 development of immunity that is more robust in the face of viral escape. 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC  454 The effect of a mutation on antibody-peptide binding was quantified as "differential selection," 455 which is the log fold change in the enrichment of a mutation-containing peptide compared to 456 the wildtype peptide. This number is multiplied by the average of the wildtype peptide 457 enrichments at that site and its two adjacent sites to get a "scaled differential selection" value, 458 as described previously [60]. The enrichment values of the adjacent wildtype peptides are 459 included in this calculation to make the analysis less susceptible to noise. Negative differential 460 selection values represent reduced binding compared to wildtype, while positive differential 461 selection values indicate that the mutation enhanced binding. "Summed differential selection" 462 is the sum of the 19 scaled differential selection values for all mutations at a site, and gives a 463 sense of the overall magnitude of escape at that site.
24 464 The comparison of two escape profiles is quantified by an escape similarity score computed in 465 the framework of an optimal transport problem [93]; this algorithm was described in detail at 466 https://matsengrp.github.io/phippery/esc-prof.html. An overview of the method is shown in S4 105 and is also made available for use under a CC0 license.
(which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC The copyright holder for this preprint this version posted December 3, 2021. ; https://doi.org/10.1101/2021.12.01.470697 doi: bioRxiv preprint